- Email: [email protected]
Dynamic behaviors of inorganic all-solid-state electrochromic device: Role of potential
Accepted Manuscript Dynamic behaviors of inorganic all-solid-state electrochromic device: Role of potential Qirong Liu, Qianqian Chen, Qianqian Zhang, Guobo Dong, Xiaolan Zhong, Yu Xiao, Marie-Paule Delplancke-Ogletree, François Reniers, Xungang Diao PII:
S0013-4686(18)30332-3
DOI:
10.1016/j.electacta.2018.02.050
Reference:
EA 31245
To appear in:
Electrochimica Acta
Received Date: 23 December 2017 Revised Date:
25 January 2018
Accepted Date: 8 February 2018
Please cite this article as: Q. Liu, Q. Chen, Q. Zhang, G. Dong, X. Zhong, Y. Xiao, M.-P. DelplanckeOgletree, Franç. Reniers, X. Diao, Dynamic behaviors of inorganic all-solid-state electrochromic device: Role of potential, Electrochimica Acta (2018), doi: 10.1016/j.electacta.2018.02.050. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Dynamic behaviors of inorganic all-solid-state electrochromic device: Role of potential Qirong Liu a,b,c, Qianqian Chen b,c, Qianqian Zhang a,b, Guobo Dong a,b, Xiaolan Zhong a,b, Yu Xiao a,b,c, Marie-Paule Delplancke-Ogletree b,d, François Reniers b,c, Xungang Diao a,b,∗ School of Physics and Nuclear Energy Engineering, Beihang University, Beijing 100191, China
b
ULB-BUAA Plasma Laboratory for Materials and Environment
c
CHANI, Faculté of Sciences, Université Libre de Bruxelles, Brussels 1050, Belgium
d
4MAT, School of Engineering, Université Libre de Bruxelles, Brussels 1050, Belgium
SC
RI PT
a
Abstract
M AN U
Applied potential plays a significant role in the properties of electrochromic devices (ECDs), for example, cyclic property, optical modulation, and response rate. In this study, inorganic all-solid-state
ECDs
with
the
multilayer
structure
of
Glass/ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO were prepared by magnetron sputtering. The potential dependence of charges dynamic behaviors in the ECDs were analyzed on the basis of
TE D
cyclic voltammograms (CVs), chronoamperograms (CAs), multi-potential steps, and in-situ optical transmittance at 550 nm. Results demonstrated that the trapping of Li+ ions in NiOx layer and in WO3 layer were responsible for the degradation of electrochromic properties of the ECDs
EP
operated at different potential ranges. Besides, the dynamic behavior of Li+ ions in WO3 layer, acting as the primary electrochromic layer in the ECDs, had a crucial influence on the response
AC C
characteristics of the ECDs. Excellent optical memory effects at randomly electrochromic extent were observed and the corresponding open-circuit potential was relevant to the chemical potential of the ECDs.
Keywords: Electrochromic
device,
Dynamic
behaviors,
Inorganic,
Applied
potential,
All-solid-state
1 Introduction
∗
Corresponding author at: School of Physics and Nuclear Energy Engineering, Beihang University, Beijing 100191, China. E-mail address: [email protected] (Xungang Diao).
ACCEPTED MANUSCRIPT Electrochromic devices (ECDs) are capable of modulating optical transmittance in visible region in response to the switching of externally applied potential [1-3]. The energy-efficient features, including low power consumption, large optical modulation, high coloration efficiency, and good optical memory effect, enable ECDs to suitably serve as smart windows, skylights, etc.
RI PT
[4-6]. Inorganic all-solid-state ECDs have been of tremendous interest due to some unique advantages over ECDs containing organic materials, such as high resistance to ultraviolet radiation exposure, light weight and no requirement of seal [2,3,7].
As the core layers in ECDs, inorganic electrochromic materials (typically cathodic colored
SC
WO3 and anodic colored NiOx) have been being extensively studied in the presence of electrochemical properties, optical modulation, response rate, cyclic durability, and degradation of
M AN U
electrochromic properties [8-12]. These characteristics are in association with dynamic behaviors of transferred charges (small ions and electrons) such as interfacial transfer dynamics of charges [13-14], ions diffusion [15], and ions trapping in electrochromic layers [11,16]. There are many reports involving the influence of applied potential on the ion-diffusion and electrochromic properties of WO3 and NiOx. Faughnan et al. [13] explicated that the coloration dynamics was
TE D
limited by the interfacial barrier between WO3 and electrolyte at low applied potentials; and ions diffusions in the bulk of WO3 was the decisive factor at relatively high voltages. The imbalance of inserted and extracted charges, induced by the applied potentials, could lead to the formation of
EP
dielectric tungstate phase (Li2WO4), and bring about a fast degradation process of WO3 [17-18]. Niklasson et al. [19-20] reported that the incorporation of titanium (Ti) into WO3 could improve
AC C
its cyclic durability due to the enlargement of the appropriate potential range. In the case of electrochromic NiOx thin films, an inappropriate potential range is able to accelerate the electrochromic degradation of NiOx thin film, even disintegrate the films [21-23]. Response characteristic and optical density of NiOx are strongly reliable on the applied potential [24]. It is noteworthy that all these behaviors are directly relevant to the externally applied potential that is one of most significant factors in the transfer dynamics of charges in electrochromic materials. Nevertheless, there is a lack of studies on the dynamics of charges transfer in inorganic all-solid-state ECDs. Undoubtedly, the dynamic behaviors of charges transfer dominate the electrochromic properties of ECDs. For instances, the optical density is extremely dependent on the amount of the transferred charges [25-27]; the response rate relies on the ions diffusion
ACCEPTED MANUSCRIPT characteristics in electrochromic layers and ions electrolyte layer [28], and the kinetics of transferred charges across the interfaces between layers [14]. Besides, the unbalance between the transferred charges for coloration and bleaching could restrain the optical modulation and readily give rise to a growing degradation process [29]. Granqvist et al. [1,16,19,30] mentioned that an
RI PT
appropriately applied potential played a critical role in long-term cyclic durability of ECDs. The “safe voltage limit” significantly imposes restriction on the potential range applied between two electrodes of ECDs [29,31]. Once the applied potential goes beyond the safe range, ECDs would be subjected to electrochromic degradation [24,32] or even electrical breakdown [33]. With
SC
increasing applied potentials, the residual current densities in coloring and bleaching processes of ECD increase [7], which implied an enhancing electronic leakage current and would cause a poor
M AN U
optical memory effect.
In our previous work [34], ECDs with a superimposed seven-layer structure was designed and monolithically prepared. Compared with the typical five-layer ECDs, the seven-layer ECD with the embedment of two Ta2O5 layers had a distinct merit of excellent optical memory. On the basis of that, the seven-layer ECD is chosen as the object of the present study. This paper focuses
TE D
on the potential dependence of dynamic behaviors of the inorganic all-solid-state ECDs. The correlations of charge-transfer imbalance, optical modulation, response characteristic, and optical memory with potential is unraveled, according to electrochemical measurements and
EP
corresponding in-situ optical transmittance.
AC C
2. Experimental details
2.1 Preparation of inorganic all-solid-state ECDs ECDs with the structure of ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO were monolithically
deposited on the 3 cm × 2 cm indium tin oxide (ITO) coated glass substrates using magnetron sputtering without intentional heating. Before introduced into vacuum chamber, these 25 Ω cm-2 ITO-coated substrates were ultrasonically cleaned in anhydrous ethanol and deionized water for 20 min in sequence. Prior to the deposition of each layer, the corresponding target with 100 mm in diameter was pre-sputtered for 10 minutes to avoid cross-contaminants. During the deposition processes, the substrates holder was rotated at a certain speed and the vertical distances between targets and substrate were approximately 150 mm. The initial base pressure of vacuum chamber
ACCEPTED MANUSCRIPT was 3.0×10-3 Pa. The flow rates of Ar (99.99 %) and O2 (99.99 %) were individually set by mass flow controllers. The ions electrolyte layer (LiNbO3) was prepared by radio frequency magnetron sputtering with 300 W. Other more details on deposition parameters were presented in the previous work [34].
RI PT
2.2 Characterization of ECDs
Structural profile of the as-prepared ECDs was characterized using a Quanta FEG 250 environmental scanning electron microscope (FEI, USA). The In-situ optical transmittance of the
SC
ECDs versus time was acquired at 550 nm and air was used as the baseline and reference by ultraviolet-visible (UV-Vis) spectrophotometer (HITACHI UV-3010) with transmittance model (T% model).
M AN U
The electrochemical measurements for these ECDs were carried out with a two-electrode cell in a CHI 660E electrochemical workstation (CH Instruments) in normal ambient air. For these measurements, the top ITO layer adjacent to WO3 layer was used as the working electrode, and the bottom ITO layer next to NiOx layer linked to the counter electrode and the reference electrode. Prior to these characterizations of the ECDs, 1-2 mm width of all edges except the masked edge
3 Results and discussion
TE D
with vacuum tape was cut off to avoid leakage current caused by edge shorts.
EP
3.1 Physical structure of ECDs
SEM image in Fig. 1 presents the cross section of the inorganic ECD. As observed, the ECD
AC C
is obviously composed of seven superimposed layers: ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO. The thickness of each layer is roughly 110 nm, 200 nm, 40 nm, 280 nm, 40 nm, 620 nm, and 250 nm, respectively. The exhibited clear interfaces between layers could rule out that the diffusion of particles during the deposition process causes the performance decay of the ECDs. Furthermore, the strong adhesion between layers can be observed, which suggests an excellent robust mechanical property of the ECD. And more significantly, it is beneficial to the transfer of charges through these interfaces, such as the transfer of electrons between NiOx or WO3 layers and ITO layers, and the transfer of Li+ ions between LiNbO3 layer and electrochromic layers through Ta2O5 barrier layer. As observed, the top ITO layer with a compact bulk-like structure shows no any
ACCEPTED MANUSCRIPT defects such as voids and cracks, which is conducive to the horizontal and vertical transfer of electrons in the electrode. As a result, it could lead to a low potential drop on the electrode surface and consequently improve the response characteristic. At the same time, it could smoothly transport charge-balancing electrons for electrochromic processes of WO3 layer. The other layers
RI PT
present a column-like structure. More importantly, these columns extends from NiOx layer to WO3 layer without obviously intermediate truncation caused by the interfaces. Therefore, the boundaries between these columns could provide additional channels for the dynamic behavior of
M AN U
SC
small ions, which proved the excellent match between layers in the ECDs.
TE D
Fig. 1 Cross-section SEM image of the inorganic all-solid-state ECDs. 3.2 Imbalance of transferred charges
As mentioned previously, it is significantly important for the ECDs to operate within the safe
EP
potential range. Here, all applied potentials for the electrochemical measurements of ECDs fall into the range between -2.0 V and 1.5 V. Fig. 2 shows CVs obtained in different potential spans
AC C
with the scan rate of 50 mV s-1, and in-situ optical transmittance corresponding to the first two cycles (insets) of the ECDs. As observed from Fig. 2(a-c), the 1st coloring process clearly displays abnormal behaviors, on which they have remarkably larger peak current densities than the following coloring processes. However, for the first two CVs, an increasing optical modulation between the colored and bleached states is observed. Fig 2d exhibits the potential dependence of CV (2nd cycle) and in-situ optical transmittance of the first three cycles. With wider potential range, the CV has a larger enveloped area, which represents more charges taking part in the electrochemical redox reactions. As a result, larger optical modulation is achieved during the CV. Besides, another critical feature is that the current densities closely approximate zero, even though
ACCEPTED MANUSCRIPT the applied potentials increase from a few tenth to 1.5 V at the end of all bleaching processes. This behavior implies an extremely limited leakage current occurring to the ECDs during these cycling processes. Because the Ta2O5 layers allow Li+ ions to smoothly shuttle between two electrochromic layers and block the transfer of electrons contributing to leakage current in the
RI PT
ECDs [34]. Patel et al. [35] mentioned that poor optical memory was induced by electrical leakage current, and the thickness of electrolyte had a significant effect on the behavior. Nagai et al. [36] also attributed poor optical memory to the presence of electrical leakage stemmed from the existence of pinholes, dust and metallic particles in the ECDs. Furthermore, O'Brien et al. [37]
SC
highlighted that a critical issue of inorganic all-thin-film ECDs was the inherent leakage current in the ECD, which was produced by edge shorts and bulk defects in the as-deposited thin films. It
M AN U
follows that the optical memory is strongly associated with the amount of electrical leakage current. These present ECDs show no obvious leakage current in the CVs, which implies an
AC C
EP
TE D
excellent optical memory.
Fig. 2 Evolutions of CVs with different potential ranges, (a) -1.5 V ~ 1.5 V, (b) -1.8 V ~ 1.5 V, and (c) -2.0 V ~ 1.5 V, respectively. Insets are corresponding optical transmittance at 550 nm plotted against potential for first 2 cycles. (d) Comparison of CVs with different potential ranges (2nd cycle) and in-situ transmittance for the first three cycles. Arrows denote the scan direction of potential.
Fig. 3 indicates the evolution of transferred charge densities versus cycle number derived
ACCEPTED MANUSCRIPT from the CVs. As expected, for the first CV, the transferred charge densities distinctly deviate from the subsequent cycles, whether for coloration process or for bleaching process. It must be emphasized that a larger transferred charge density would generally result in a higher optical density for one ECD [5,38]. However, for the first coloration, there is a relatively higher optical
RI PT
transmittance than that for the 2nd coloration, which is incongruous with the less transferred charge densities for the coloring process. There are three possible explanations for the unusual phenomenon. One is relevant to the oxidation reaction of LiNbO3 that lost Li+ ions and was oxidized to Nb2O5. The second one is associated with the distribution of electric field across the
SC
ECDs. It assumes that the potential drop mainly appears to the interfaces between WO3 and LiNbO3 layers for the first coloring process. More Li+ ions could readily pass through the
M AN U
interfaces and Ta2O5 layer, and insert into electrochromic WO3 layer. In contrast, the relatively lower potential drop at the interfaces between NiOx and LiNbO3 layers would limit the pass of Li+ ions for the following cycles. The third one relates to the physical diffusion of Li+ ions in the preparation process of ECDs. Li+ ions have a small radius and readily diffuse in bulk. During the deposition process of Li+ ions electrolyte (LiNbO3) layer, some energetic Li+ ions diffuse
TE D
through/into the Ta2O5 barrier layer and into NiOx layer. Owing to the physical diffusion of Li+ ions, the transferring process of these Li+ ions could not lead to electrochromic behaviors of NiOx layer when they are transferred from NiOx layer to WO3 layer under the application of external potential. For the first bleaching process, there is a less transferred charge density, which is
EP
probably attributed to the retention of Li+ ions in WO3 due to the blocking effect of Ta2O5 barrier
AC C
layer. For the 2nd cycle, since the larger coloration potential is applied, the deviation of transferred charge densities between for coloration and for bleaching becomes less, which, to some extent, suggests that the larger coloration potential promotes the faster transfer of Li+ ions into WO3 layer.
Fig. 3 Evolutions of transferred charge density versus cycle number corresponding to the CVs with different potential ranges (a) -1.5 V ~1.5 V, (b) -1.8 V ~1.5 V, and (c) -2.0 V ~1.5 V. Insets are their
ACCEPTED MANUSCRIPT corresponding in-situ transmittance at 550 nm, respectively.
Fig. 3a presents larger transferred charge densities for coloration than that for bleaching. Correspondingly, as the cycle number increases, the inset in Fig. 3a shows increasing optical transmittances of ECD in colored and bleached states, which is irreconcilable with that more
RI PT
charges are transferred for coloring process than for bleaching process. Two possible reasons could explain the phenomenon. One involves a relatively low potential for bleaching, which could not reversibly extract all these inserted Li+ ions from NiOx layer, and leads to an increasing optical transmittance for bleached states. Besides, as observed in Fig. 3b and 3c, the transferred charge
SC
density is much higher than that in Fig. 3a, which means that Li+ ions in LiNbO3 layer are sufficient for electrochromic behavior in the present potential range, the higher charge densities
M AN U
for coloration are attributed to the transfer of these Li+ ions in LiNbO3 layer. The Li+-ion trapping could result in reduction of nickel vacancies, and a simultaneous variation in dielectric properties of NiOx layer, and brings about an increasing voltage drop of potential across NiOx layer, as a result, the practical potential applied to drive charges becomes low and the transferred charge density decreases with the increasing cycle. Another possible reason for this behavior is the
TE D
trapping of Li+ ions in WO3. The Li+-ion trapping process gives rise to the formation of Li2WO4 [18] or Li2O in the WO3 film matrix [39], which also causes the coloration degradation of WO3 [11,40]. The increasing transmittance in bleached state implies that a growing amount of Li+ ions
EP
contributes to the bleaching of the ECD.
Meanwhile, some emphases should put on the Fig. 3b and 3c. (i) In addition to the first two
AC C
CVs, the imbalances of transferred charge densities for coloration and bleaching processes in each CV of ECDs are slight in these figures. (ii) However, it is of obvious difference between transferred charge density for current coloration process and for last bleaching process. When the former is less than the latter, the optical transmittance of ECDs in colored state becomes higher, as shown in insets (in Fig. 3b and 3c), which could be attributed to the declining charge density transferred into WO3 layer in the coloring processes. (iii) Different from the transmittance in bleached state in Fig. 3a, it keeps constant in Fig. 3b and 3c despite of the increasing transmittance in colored state (except the first few cycles). Taking (i), (ii) and (iii) into consideration, it seems that the Li+-ion trapping in NiOx causes the degradation of ECDs in Fig. 3b and 3c. On one hand, for one complete CV (first coloration, then bleaching), the transferred charge density for bleaching
ACCEPTED MANUSCRIPT processes is slightly larger than that for coloration processes (slight imbalances), which argues that it is less possible for Li+ ions to trap in WO3. On the other hand, (ii) and (iii) indicate that the transfer of Li+ ions from WO3 into NiOx is not completely reversible and a portion of them could not be extracted from NiOx layer. However, WO3 layer generally plays a primary and dominant
RI PT
role in the contribution to the electrochromic behavior of the ECDs. The dynamic behavior of Li+ ions in WO3 could be a very important factor in the degradation of these ECDs. Besides, the interfacial dependence of Li+-ion dynamic behavior in ECDs cannot be neglected. Therefore, there are possibly some more complicated dynamic behaviors pertaining to the elctrochromic processes
SC
of the ECDs. 3.3 Response characteristic
M AN U
Optical modulation, one of the most critical criteria for the assessment of ECDs, is dominated by the applied potential. The inorganic ECD was operated using multi-potential steps technique. Fig. 4 displays the effect of applied potentials on optical modulation of the ECD. As the potential span becomes wider, the optical modulation of ECD is obviously larger (in Fig. 4c). Corresponding photographs of the ECD in colored and bleached states are exhibited in Fig. 4a. As
TE D
shown in Fig. 4b, the induced peak current density is larger with the increasing potential, and the transferred charge density increases from 5.3 mC/cm2 to 19.5 mC/cm2 for coloration and from 6.2 mC/cm2 to 19.4 mC/cm2 for bleaching. These behaviors demonstrate that, as the applied potential
EP
increases, more Li+ ions could get rid of the confinement of relatively high energy barrier [41], and are smoothly transferred in the ECD during the coloration/bleaching processes as a result of
AC C
increasing the potential span.
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Fig. 4 Optical modulation of the ECD driven by step potentials with the duration of 40 s, (a) applied step potentials, (b) corresponding current density, and (c) in-situ optical transmittance at 550 nm.
Response rate of coloring and bleaching processes of the ECD is determined by the time it takes to achieve 90 % of total variation in optical transmittance between both stable electrochromic states [6]. Fig. 5 shows the relaxation behavior of current densities and response
TE D
characteristic of optical transmittance of the ECD operated at different positive potentials during bleaching processes. In Fig. 5a, the analyzed CAs with different potential ranges are 10th cycle where CAs tends to be stable. As observed from insets in Fig. 5a, accompanying with the
EP
increasing positive potential from 0.5 V to 1.5 V, the peak current density increases for bleaching and decreases for coloring, respectively. The relaxation time of corresponding current density
AC C
individually are shorter for bleaching and longer for coloring. As shown in Fig. 5b, the response time gradually decreases for both bleaching and coloration of the ECD.
ACCEPTED MANUSCRIPT Fig. 5 Variations in current density (a) and response times (b) of the ECD operated at different applied potentials during bleaching processes.
Concerning the response characteristics of the ECD operated at different potentials for coloring processes, as observed in Fig. 6a, the relaxation time of current densities during the
RI PT
coloring and bleaching processes increases with the increasing coloration potential. In Fig. 6b, shorter response time for coloring process is observed with the application of larger coloration potential. On the contrary, the response rate of bleaching process becomes slower. Since a larger
SC
coloration potential is applied for coloring process, Li+ ions are more readily extracted from NiOx layer, and simultaneously a larger amount of them are accommodated in the WO3 layer or even reside in the trap sites with higher energy barriers [41-42]. When the reverse potential is constant,
M AN U
the bleaching process, during which all Li+ ions are reversibly transferred from WO3 layer into NiOx layer, has to take a longer polarization duration. However, the increasing response time for the bleaching process seems inconsistent with the results in Fig. 5b. This behavior could be explicated by the following description. WO3 layer is the primary electrochromic layer and the dynamic behavior of Li+ ions inside this layer plays a dominant role in the response rate of the
TE D
ECDs. Furthermore, in one electrochromic layer, the amount of Li+ ions residing in trap sites with higher energy barrier is mostly decided by the potential for insertion of Li+ ions into the electrochromic layer.
EP
Compared to the reported results in our previous work [34], the faster response rate could be ascribed to the thinner electrochromic layers, barrier layers and ions electrolyte layer. In addition,
AC C
as shown in Fig. 5a and Fig. 6a, it should be stressed that the residual current densities at the end of both coloring and bleaching processes are approximately zero, which indicates a negligible electronic leakage current through the inorganic all-solid-state ECDs during these cycling electrochromic processes. This behavior, in accordance with the results in CVs, indicates an excellent optical memory of the ECDs [34].
RI PT
ACCEPTED MANUSCRIPT
Fig. 6 Variations in current density (a) and response times (b) of the ECD operated at different applied
SC
potentials during coloring processes.
3.4 Optical memory effect
M AN U
Excellent optical memory effect is a significant feature of ECDs serving as energy-efficient smart windows. Dynamic electrochromic windows can be readily controlled to provide additional comfort [3,43], which requires that the ECDs at all colored and bleached extents have an excellent optical memory. Fig. 7 and Fig. 8 shows the optical memory and open-circuit potential randomly at three times during the coloring and bleaching processes of the ECD, respectively. As expected,
TE D
since no obvious electronic leakage current is observed in CVs and CAs, the ECD has an excellent optical memory in Fig. 7(a~c) and in Fig. 8(a~c), which implies a nearly perfect interfaces between layers in the ECD. Such phenomenon suggests that the incorporation of Ta2O5 layers
EP
effectively limits the formation of defects such as pinholes at the interface and in the ion electrolyte layer, contributing to the electronic leakage. Accompanying with the enhancing extent
AC C
of coloring, shown in Fig. 7(d~f), the open-circuit potential of the ECD is lowering. During the bleaching process, the initial open-circuit potential is closer to the applied positive potential with the increasing extent of bleaching, illustrated in Fig. 8(d~f). Owing to the transferring of Li+ ions between WO3 and NiOx layers, the chemical potentials (µ) of both the layers are changing. The over-potential (η) between two electrodes of the ECD can be expressed as [13]: η = VA - ∆µ/Ne, wherein, VA and ∆µ denote the applied potential and the variation in the chemical potential throughout the ECD. N and e is Avogadro’s constant and unit charge. The open-circuit potential corresponds to the second term on the right side of the equation. When η is higher than 0, the
ACCEPTED MANUSCRIPT electrochromic behavior occurs to the ECD till the η approaches zero, accompanied by the
SC
RI PT
increasing chemical potential.
M AN U
Fig. 7 Optical memory effect (a~c) and corresponding open-circuit potential (d~f) at random times
EP
TE D
during the coloring process of the ECD operated between -2.0 and 1.0 V.
AC C
Fig. 8 Optical memory (a~c) and corresponding open-circuit potential (d~f) at random times during the bleaching process of the ECD operated between -2.0 and 1.0 V.
4 Conclusions
In summary, the inorganic all-solid-state ECD with seven superimposed layers of
Glass/ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO were monolithically prepared by magnetron sputtering. In the CVs of the ECD, the imbalance of transferred charges during coloring and bleaching processes is mainly attributed to the trapping behavior of Li+ ions in electrochromic layers. Larger optical modulation can be obtained for the ECD operated within wider potential range. In the case of response characteristic, the higher positive potentials during bleaching
ACCEPTED MANUSCRIPT processes would lead to faster response rates for the ECD. However, the larger coloration potentials during coloring processes brings about shorter coloring and longer bleaching response times. These dynamic behaviors are associated with the response characteristics of WO3 layer acting as the primary electrochromic layer in the ECD. The ECD has excellent optical memory
RI PT
effects in coloring and bleaching processes. And corresponding open-circuit potentials link to the variation in the electrochemical potential due to the transfer of charges in the ECD. Acknowledgments
SC
The authors sincerely appreciate the financial support of the National Program on Key Research Project of China (2016YFB0303901), the Beijing Municipal Natural Science Foundation the
Fundamental
Re-search
Funds
for
the
Central
Universities
(Grant
M AN U
(2161001),
No.YWF-16-JCTD-B-03), National Natural Science Foundation of China (Grant No. 21701003), China Postdoctoral Science Foundation Grant (2015M580035 and 2017T100022), and the Fundamental Research Funds for the Central Universities (KG12030401, ZG216S17I3,
References
TE D
ZG226S17Q4).
[1] Claes G. Granqvist, Electrochromics for smart windows: Oxide-based thin films and devices, Thin Solid Films, 2014, 564, 1-38.
EP
[2] K. J. Patel, G. G. Bhatt, J. R. Ray, Priya Suryavanshi, C. J. Panchal, All-inorganic solid-state electrochromic devices: a review, Journal of Solid State Electrochemistry, 2016, 21, 1-11.
AC C
[3] Guofa Cai, Jiangxin Wang, Pooi See Lee, Next-Generation Multifunctional Electrochromic Devices, Acc. Chem. Res., 2016, 49, 1469-1476. [4] Sami Oukassi, Cédric Giroud-Garampon, Christophe Dubarry, Cédric Ducros, Raphaël Salot, All inorganic thin film electrochromic device using LiPON as the ion conductor, Solar Energy Materials & Solar Cells, 2016, 145, 2-7. [5] Guofa Cai, Peter Darmawan, Mengqi Cui, Jingwei Chen, Xu Wang, Alice Lee-Sie Eh, Shlomo Magdassi, Pooi See Lee, Inkjet-printed all solid-state electrochromic devices based on NiO/WO3 nanoparticle complementary electrodes, Nanoscale, 2016, 8, 348-357. [6] Huiying Qu, Hangchuan Zhang, Na Li, Zhongqiu Tong, Jing Wang, Jiupeng Zhao, Yao Li, A
ACCEPTED MANUSCRIPT rapid-response electrochromic device with significantly enhanced electrochromic performance, RSC Adv., 2015, 5, 803-806 [7] Chien-Jen Tang, Jia-Ming Ye, Yueh-Ting Yang, Ju-Liang He, Large-area flexible monolithic ITO/WO3/Nb2O5/NiVOx/ITO electrochromic devices prepared by using magnetron sputter
RI PT
deposition, Optical Materials, 2016, 55, 83-89. [8] Rui-Tao Wen, Sara Malmgren, Claes G. Granqvist, Gunnar A. Niklasson, Degradation Dynamics for Electrochromic WO3 Films under Extended Charge Insertion and Extraction: Unveiling Physicochemical Mechanisms, ACS Appl. Mater. Interfaces, 2017, 9, 12872-12877.
SC
[9] Junling Zhou, Gui Luo, Youxiu Wei, Jianming Zheng, Chunye Xu, Enhanced electrochromic performances and cycle stability of NiO-based thin films via Li-Ti co-doping prepared by sol-gel
M AN U
method, Electrochimica Acta, 2015, 186, 182-191.
[10] Rui-Tao Wen, Gunnar A. Niklasson, Claes G. Granqvist, Strongly Improved Electrochemical Cycling Durability by Adding Iridium to Electrochromic Nickel Oxide Films, ACS Appl. Mater. Interfaces, 2015, 7, 9319-9322.
[11] Rui-Tao Wen, Miguel A. Arvizu, Michael Morales-Luna, Claes G. Granqvist, Gunnar A.
TE D
Niklasson, Ion Trapping and Detrapping in Amorphous Tungsten Oxide Thin Films Observed by Real-Time Electro-Optical Monitoring, Chem. Mater., 2016, 28, 4670-4676. [12] Guofa Cai, Mengqi Cui, Vipin Kumar, Peter Darmawan, Jiangxin Wang, Xu Wang, Alice
EP
Lee-Sie Eh, Kai Qian, Pooi See Lee, Ultra-large optical modulation of electrochromic porous WO3 film and the local monitoring of redox activity, Chem. Sci., 2016, 7, 1373-1382.
AC C
[13] Richard S. Crandall, Brian W. Faughnan, Dynamics of coloration of amorphous electrochromic films of WO3 at low voltages, Applied Physics Letters, 28, 1976, 95-97. [14] Feng Lin, Justin B. Bult, Sanjini Nanayakkara, Anne C. Dillon, Ryan M. Richards, Jeffrey L. Blackburn, Chaiwat Engtrakul, Graphene as an Efficient Interfacial Layer for Electrochromic Devices, ACS Appl. Mater. Interfaces, 2015, 7, 11330-11336. [15] Francisco Fabregat-Santiago, Germà Garcia-Belmonte, Juan Bisquert, Noemí S. Ferriols, Paulo R. Bueno, Elson Longo, Julio Suay Antón, Socorro Castro-García, Dynamic Processes in the Coloration of WO3 by Lithium Insertion, Journal of The Electrochemical Society, 2001, 148, E302-E309. [16] R.-T.Wen, C.G. Granqvist, G.A. Niklasson, Eliminating degradation and uncovering
ACCEPTED MANUSCRIPT ion-trapping dynamics in electrochromic WO3 thin films, Natural Materials, 2015, 14, 996-1002. [17] Miguel A. Arvizu, Rui-Tao Wen, Daniel Primetzhofer, Jolanta E. Klemberg-Sapieha, Ludvik Martinu, Gunnar A. Niklasson, Claes G. Granqvist, Galvanostatic Ion Detrapping Rejuvenates Oxide Thin Films, ACS Appl. Mater. Interfaces, 2015, 7, 26387-26390.
RI PT
[18] P.V. Ashrit, G. Bader, Vo-Van Truong, Electrochromic properties of nanocrystalline tungsten oxide thin films, Thin Solid Films 320, 1998, 324-328.
[19] M. A. Arvizu, C. A. Triana, B. I. Stefanov, C. G. Granqvist, G. A. Niklasson, Electrochromism in sputter-deposited W-Ti oxide films: Durability enhancement due to Ti, Solar
SC
Energy Materials & Solar Cells, 2014, 125, 184-189.
[20] M. Morales-Luna, M. A. Arvizu, C. G. Granqvist, G. A. Niklasson, Sputter deposited
Conference Series, 2016, 682, 012021.
M AN U
W1-x-yNixTiyO3 thin films: Electrochromic properties and durability, Journal of Physics:
[21] Rui-Tao Wen, Claes G. Granqvist, Gunnar A. Niklasson, Cyclic voltammetry on sputter-deposited films of electrochromic Ni oxide: Power-law decay of the charge density exchange, Applied Physics Letters, 2014, 105, 163502.
TE D
[22] Rui-Tao Wen, Claes G. Granqvist, Gunnar A. Niklasson, Anodic Electrochromic Nickel Oxide Thin Films: Decay of Charge Density upon Extensive Electrochemical Cycling, Chem. Electro. Chem., 2016, 3, 266-275.
EP
[23] Yoshijiro Ushio, Akira Ishikawa, Tatsuo Niwa, Degradation of the electrochromic nickel oxide film upon redox cycling, Thin Solid Films, 1996, 280, 233-237.
AC C
[24] A. Pennisi, C. M. Lampert, Optical Properties of Electrochromic Nickel Oxide Devices Utilizing A Polymeric Electrolyte, SPIE Vol. 1016 Optical Materials Technology for Energy Efficiency and Solar Energy Conversion VII, 1988, 131-144. [25] Anneke Georg, Andreas Georg, Electrochromic device with a redox electrolyte, Solar Energy Materials & Solar Cells, 2009, 93, 1329-1337. [26] J. Zhang, J. P. Tu, X. H. Xia, Y. Qiao, Y. Lu, An all-solid-state electrochromic device based on NiO/WO3 complementary structure and solid hybrid polyelectrolyte, Solar Energy Materials & Solar Cells, 2009, 93, 1840-1845. [27] C. Person, I. Porqueras, M. Vives, C. Corbella, A. Pinyol, E. Bertran, Degradation of a solid state electrochromic device, Solid State Ionics, 2003, 165, 73-80.
ACCEPTED MANUSCRIPT [28] Tatsuo Niwa, Osamu Takai, Optical and electrochemical properties of all-solid-state transmittance-type electrochromic devices, Thin Solid Films, 2010, 518, 1722-1727. [29] Kwang-Soon Ahn, Yoon-Chae Nah, Jin-Young Park, Yung-Eun Sung, Ki-Yun Cho, Seung-Shik Shin, Jung-Ki Park, Bleached state transmittance in charge-unbalanced all-solid-state
RI PT
electrochromic devices, Appl. Phys. Lett., 2003, 82, 3379. [30] Claes G.Granqvist, Oxide electrochromics: An introduction to devices and materials, Solar Energy Materials & Solar Cells, 2012, 99, 1-13.
[31] M. Schmitt, M. A. Aegerter, Electrochromic properties of pure and doped Nb2O5 coatings
SC
and devices, Electrochimica Acta, 2001, 46, 2105-2111.
[32] L. Kullman, A. Azens, C. G. Granqvist, Decreased electrochromism in Li-intercalated Ti
M AN U
oxide films containing La, Ce, and Pr, Journal of Applied Physics, 1997, 81, 8002. [33] Mark Samuel Burdis, Bryan D. Greer, Douglas Glenn John Weir, Electrochromic Devices Having Improved Ions Conducting Layers, United States Patent, US 8,004,744 B2, Aug. 23, 2011 [34] Qirong Liu, Guobo Dong, Qianqian Chen, Junji Guo, Yu Xiao, Marie-Paule Delplancke-Ogletree, François Reniers, Xungang Diao, Charge-transfer kinetics and cyclic
TE D
properties of inorganic all-solid-state electrochromic device with remarkably improved optical memory, Solar Energy Materials and Solar Cells, 2018, 174, 545-553. [35] K. J. Patel, M. S. Desai, C. J. Panchal, Studies of ZrO2 electrolyte thin-film thickness on the
275-279.
EP
all-solid thin-film electrochromic devices. Journal of Solid State Electrochemistry, 2014, 19,
AC C
[36] J. Nagai, G. D. McMeeking, Y. Saitoh, Durability of electrochromic glazing, Solar Energy Materials & Solar Cells, 1999, 56, 309-319. [37] Nada A. O'Brien, J. Gordon, H. Mathew, Bryant P. Hichwa, Electrochromic coatings-applications and manufacturing issues, Thin Solid Films, 1999, 345, 312-318. [38] Miao-Syuan Fan, Sheng-Yuan Kao, Ting-Hsiang Chang, R. Vittal, Kuo-Chuan Ho, A high contrast solid-state electrochromic device based on nano-structural Prussian blue and poly(butylviologen) thin films, Solar Energy Materials & Solar Cells, 2016, 145, 35-41. [39] Sung Jong Yoo, Ju Wan Lim, Yung-Eun Sung, Improved electrochromic devices with an inorganic solid electrolyte protective layer, Solar Energy Materials & Solar Cells, 2006, 90, 477-484.
ACCEPTED MANUSCRIPT [40] Rui-Tao Wen, Gunnar A. Niklasson, Claes G. Granqvist, Sustainable Rejuvenation of Electrochromic WO3 Films, ACS Appl. Mater. Interfaces 2015, 7, 28100-28104. [41] Juan Bisquert, Analysis of the kinetics of ion intercalation: Ion trapping approach to solid-state relaxation processes, Electrochimica Acta, 2002, 47, 2435-2449.
RI PT
[42] Juan Bisquert, Vyacheslav S. Vikhrenko, Analysis of the kinetics of ion intercalation. Two state model describing the coupling of solid state ion diffusion and ion binding processes, Electrochimica Acta, 2002, 47, 3977-3988.
[43] N. Sbar, M. Badding, R. Budziak, K. Cortez, L. Laby, L. Michalski, T. Ngo, S. Schulz, K.
AC C
EP
TE D
M AN U
Materials & Solar Cells, 1999, 56, 321-341.
SC
Urbanik, Progress toward durable, cost effective electrochromic window glazings, Solar Energy
S0013-4686(18)30332-3
DOI:
10.1016/j.electacta.2018.02.050
Reference:
EA 31245
To appear in:
Electrochimica Acta
Received Date: 23 December 2017 Revised Date:
25 January 2018
Accepted Date: 8 February 2018
Please cite this article as: Q. Liu, Q. Chen, Q. Zhang, G. Dong, X. Zhong, Y. Xiao, M.-P. DelplanckeOgletree, Franç. Reniers, X. Diao, Dynamic behaviors of inorganic all-solid-state electrochromic device: Role of potential, Electrochimica Acta (2018), doi: 10.1016/j.electacta.2018.02.050. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Dynamic behaviors of inorganic all-solid-state electrochromic device: Role of potential Qirong Liu a,b,c, Qianqian Chen b,c, Qianqian Zhang a,b, Guobo Dong a,b, Xiaolan Zhong a,b, Yu Xiao a,b,c, Marie-Paule Delplancke-Ogletree b,d, François Reniers b,c, Xungang Diao a,b,∗ School of Physics and Nuclear Energy Engineering, Beihang University, Beijing 100191, China
b
ULB-BUAA Plasma Laboratory for Materials and Environment
c
CHANI, Faculté of Sciences, Université Libre de Bruxelles, Brussels 1050, Belgium
d
4MAT, School of Engineering, Université Libre de Bruxelles, Brussels 1050, Belgium
SC
RI PT
a
Abstract
M AN U
Applied potential plays a significant role in the properties of electrochromic devices (ECDs), for example, cyclic property, optical modulation, and response rate. In this study, inorganic all-solid-state
ECDs
with
the
multilayer
structure
of
Glass/ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO were prepared by magnetron sputtering. The potential dependence of charges dynamic behaviors in the ECDs were analyzed on the basis of
TE D
cyclic voltammograms (CVs), chronoamperograms (CAs), multi-potential steps, and in-situ optical transmittance at 550 nm. Results demonstrated that the trapping of Li+ ions in NiOx layer and in WO3 layer were responsible for the degradation of electrochromic properties of the ECDs
EP
operated at different potential ranges. Besides, the dynamic behavior of Li+ ions in WO3 layer, acting as the primary electrochromic layer in the ECDs, had a crucial influence on the response
AC C
characteristics of the ECDs. Excellent optical memory effects at randomly electrochromic extent were observed and the corresponding open-circuit potential was relevant to the chemical potential of the ECDs.
Keywords: Electrochromic
device,
Dynamic
behaviors,
Inorganic,
Applied
potential,
All-solid-state
1 Introduction
∗
Corresponding author at: School of Physics and Nuclear Energy Engineering, Beihang University, Beijing 100191, China. E-mail address: [email protected] (Xungang Diao).
ACCEPTED MANUSCRIPT Electrochromic devices (ECDs) are capable of modulating optical transmittance in visible region in response to the switching of externally applied potential [1-3]. The energy-efficient features, including low power consumption, large optical modulation, high coloration efficiency, and good optical memory effect, enable ECDs to suitably serve as smart windows, skylights, etc.
RI PT
[4-6]. Inorganic all-solid-state ECDs have been of tremendous interest due to some unique advantages over ECDs containing organic materials, such as high resistance to ultraviolet radiation exposure, light weight and no requirement of seal [2,3,7].
As the core layers in ECDs, inorganic electrochromic materials (typically cathodic colored
SC
WO3 and anodic colored NiOx) have been being extensively studied in the presence of electrochemical properties, optical modulation, response rate, cyclic durability, and degradation of
M AN U
electrochromic properties [8-12]. These characteristics are in association with dynamic behaviors of transferred charges (small ions and electrons) such as interfacial transfer dynamics of charges [13-14], ions diffusion [15], and ions trapping in electrochromic layers [11,16]. There are many reports involving the influence of applied potential on the ion-diffusion and electrochromic properties of WO3 and NiOx. Faughnan et al. [13] explicated that the coloration dynamics was
TE D
limited by the interfacial barrier between WO3 and electrolyte at low applied potentials; and ions diffusions in the bulk of WO3 was the decisive factor at relatively high voltages. The imbalance of inserted and extracted charges, induced by the applied potentials, could lead to the formation of
EP
dielectric tungstate phase (Li2WO4), and bring about a fast degradation process of WO3 [17-18]. Niklasson et al. [19-20] reported that the incorporation of titanium (Ti) into WO3 could improve
AC C
its cyclic durability due to the enlargement of the appropriate potential range. In the case of electrochromic NiOx thin films, an inappropriate potential range is able to accelerate the electrochromic degradation of NiOx thin film, even disintegrate the films [21-23]. Response characteristic and optical density of NiOx are strongly reliable on the applied potential [24]. It is noteworthy that all these behaviors are directly relevant to the externally applied potential that is one of most significant factors in the transfer dynamics of charges in electrochromic materials. Nevertheless, there is a lack of studies on the dynamics of charges transfer in inorganic all-solid-state ECDs. Undoubtedly, the dynamic behaviors of charges transfer dominate the electrochromic properties of ECDs. For instances, the optical density is extremely dependent on the amount of the transferred charges [25-27]; the response rate relies on the ions diffusion
ACCEPTED MANUSCRIPT characteristics in electrochromic layers and ions electrolyte layer [28], and the kinetics of transferred charges across the interfaces between layers [14]. Besides, the unbalance between the transferred charges for coloration and bleaching could restrain the optical modulation and readily give rise to a growing degradation process [29]. Granqvist et al. [1,16,19,30] mentioned that an
RI PT
appropriately applied potential played a critical role in long-term cyclic durability of ECDs. The “safe voltage limit” significantly imposes restriction on the potential range applied between two electrodes of ECDs [29,31]. Once the applied potential goes beyond the safe range, ECDs would be subjected to electrochromic degradation [24,32] or even electrical breakdown [33]. With
SC
increasing applied potentials, the residual current densities in coloring and bleaching processes of ECD increase [7], which implied an enhancing electronic leakage current and would cause a poor
M AN U
optical memory effect.
In our previous work [34], ECDs with a superimposed seven-layer structure was designed and monolithically prepared. Compared with the typical five-layer ECDs, the seven-layer ECD with the embedment of two Ta2O5 layers had a distinct merit of excellent optical memory. On the basis of that, the seven-layer ECD is chosen as the object of the present study. This paper focuses
TE D
on the potential dependence of dynamic behaviors of the inorganic all-solid-state ECDs. The correlations of charge-transfer imbalance, optical modulation, response characteristic, and optical memory with potential is unraveled, according to electrochemical measurements and
EP
corresponding in-situ optical transmittance.
AC C
2. Experimental details
2.1 Preparation of inorganic all-solid-state ECDs ECDs with the structure of ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO were monolithically
deposited on the 3 cm × 2 cm indium tin oxide (ITO) coated glass substrates using magnetron sputtering without intentional heating. Before introduced into vacuum chamber, these 25 Ω cm-2 ITO-coated substrates were ultrasonically cleaned in anhydrous ethanol and deionized water for 20 min in sequence. Prior to the deposition of each layer, the corresponding target with 100 mm in diameter was pre-sputtered for 10 minutes to avoid cross-contaminants. During the deposition processes, the substrates holder was rotated at a certain speed and the vertical distances between targets and substrate were approximately 150 mm. The initial base pressure of vacuum chamber
ACCEPTED MANUSCRIPT was 3.0×10-3 Pa. The flow rates of Ar (99.99 %) and O2 (99.99 %) were individually set by mass flow controllers. The ions electrolyte layer (LiNbO3) was prepared by radio frequency magnetron sputtering with 300 W. Other more details on deposition parameters were presented in the previous work [34].
RI PT
2.2 Characterization of ECDs
Structural profile of the as-prepared ECDs was characterized using a Quanta FEG 250 environmental scanning electron microscope (FEI, USA). The In-situ optical transmittance of the
SC
ECDs versus time was acquired at 550 nm and air was used as the baseline and reference by ultraviolet-visible (UV-Vis) spectrophotometer (HITACHI UV-3010) with transmittance model (T% model).
M AN U
The electrochemical measurements for these ECDs were carried out with a two-electrode cell in a CHI 660E electrochemical workstation (CH Instruments) in normal ambient air. For these measurements, the top ITO layer adjacent to WO3 layer was used as the working electrode, and the bottom ITO layer next to NiOx layer linked to the counter electrode and the reference electrode. Prior to these characterizations of the ECDs, 1-2 mm width of all edges except the masked edge
3 Results and discussion
TE D
with vacuum tape was cut off to avoid leakage current caused by edge shorts.
EP
3.1 Physical structure of ECDs
SEM image in Fig. 1 presents the cross section of the inorganic ECD. As observed, the ECD
AC C
is obviously composed of seven superimposed layers: ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO. The thickness of each layer is roughly 110 nm, 200 nm, 40 nm, 280 nm, 40 nm, 620 nm, and 250 nm, respectively. The exhibited clear interfaces between layers could rule out that the diffusion of particles during the deposition process causes the performance decay of the ECDs. Furthermore, the strong adhesion between layers can be observed, which suggests an excellent robust mechanical property of the ECD. And more significantly, it is beneficial to the transfer of charges through these interfaces, such as the transfer of electrons between NiOx or WO3 layers and ITO layers, and the transfer of Li+ ions between LiNbO3 layer and electrochromic layers through Ta2O5 barrier layer. As observed, the top ITO layer with a compact bulk-like structure shows no any
ACCEPTED MANUSCRIPT defects such as voids and cracks, which is conducive to the horizontal and vertical transfer of electrons in the electrode. As a result, it could lead to a low potential drop on the electrode surface and consequently improve the response characteristic. At the same time, it could smoothly transport charge-balancing electrons for electrochromic processes of WO3 layer. The other layers
RI PT
present a column-like structure. More importantly, these columns extends from NiOx layer to WO3 layer without obviously intermediate truncation caused by the interfaces. Therefore, the boundaries between these columns could provide additional channels for the dynamic behavior of
M AN U
SC
small ions, which proved the excellent match between layers in the ECDs.
TE D
Fig. 1 Cross-section SEM image of the inorganic all-solid-state ECDs. 3.2 Imbalance of transferred charges
As mentioned previously, it is significantly important for the ECDs to operate within the safe
EP
potential range. Here, all applied potentials for the electrochemical measurements of ECDs fall into the range between -2.0 V and 1.5 V. Fig. 2 shows CVs obtained in different potential spans
AC C
with the scan rate of 50 mV s-1, and in-situ optical transmittance corresponding to the first two cycles (insets) of the ECDs. As observed from Fig. 2(a-c), the 1st coloring process clearly displays abnormal behaviors, on which they have remarkably larger peak current densities than the following coloring processes. However, for the first two CVs, an increasing optical modulation between the colored and bleached states is observed. Fig 2d exhibits the potential dependence of CV (2nd cycle) and in-situ optical transmittance of the first three cycles. With wider potential range, the CV has a larger enveloped area, which represents more charges taking part in the electrochemical redox reactions. As a result, larger optical modulation is achieved during the CV. Besides, another critical feature is that the current densities closely approximate zero, even though
ACCEPTED MANUSCRIPT the applied potentials increase from a few tenth to 1.5 V at the end of all bleaching processes. This behavior implies an extremely limited leakage current occurring to the ECDs during these cycling processes. Because the Ta2O5 layers allow Li+ ions to smoothly shuttle between two electrochromic layers and block the transfer of electrons contributing to leakage current in the
RI PT
ECDs [34]. Patel et al. [35] mentioned that poor optical memory was induced by electrical leakage current, and the thickness of electrolyte had a significant effect on the behavior. Nagai et al. [36] also attributed poor optical memory to the presence of electrical leakage stemmed from the existence of pinholes, dust and metallic particles in the ECDs. Furthermore, O'Brien et al. [37]
SC
highlighted that a critical issue of inorganic all-thin-film ECDs was the inherent leakage current in the ECD, which was produced by edge shorts and bulk defects in the as-deposited thin films. It
M AN U
follows that the optical memory is strongly associated with the amount of electrical leakage current. These present ECDs show no obvious leakage current in the CVs, which implies an
AC C
EP
TE D
excellent optical memory.
Fig. 2 Evolutions of CVs with different potential ranges, (a) -1.5 V ~ 1.5 V, (b) -1.8 V ~ 1.5 V, and (c) -2.0 V ~ 1.5 V, respectively. Insets are corresponding optical transmittance at 550 nm plotted against potential for first 2 cycles. (d) Comparison of CVs with different potential ranges (2nd cycle) and in-situ transmittance for the first three cycles. Arrows denote the scan direction of potential.
Fig. 3 indicates the evolution of transferred charge densities versus cycle number derived
ACCEPTED MANUSCRIPT from the CVs. As expected, for the first CV, the transferred charge densities distinctly deviate from the subsequent cycles, whether for coloration process or for bleaching process. It must be emphasized that a larger transferred charge density would generally result in a higher optical density for one ECD [5,38]. However, for the first coloration, there is a relatively higher optical
RI PT
transmittance than that for the 2nd coloration, which is incongruous with the less transferred charge densities for the coloring process. There are three possible explanations for the unusual phenomenon. One is relevant to the oxidation reaction of LiNbO3 that lost Li+ ions and was oxidized to Nb2O5. The second one is associated with the distribution of electric field across the
SC
ECDs. It assumes that the potential drop mainly appears to the interfaces between WO3 and LiNbO3 layers for the first coloring process. More Li+ ions could readily pass through the
M AN U
interfaces and Ta2O5 layer, and insert into electrochromic WO3 layer. In contrast, the relatively lower potential drop at the interfaces between NiOx and LiNbO3 layers would limit the pass of Li+ ions for the following cycles. The third one relates to the physical diffusion of Li+ ions in the preparation process of ECDs. Li+ ions have a small radius and readily diffuse in bulk. During the deposition process of Li+ ions electrolyte (LiNbO3) layer, some energetic Li+ ions diffuse
TE D
through/into the Ta2O5 barrier layer and into NiOx layer. Owing to the physical diffusion of Li+ ions, the transferring process of these Li+ ions could not lead to electrochromic behaviors of NiOx layer when they are transferred from NiOx layer to WO3 layer under the application of external potential. For the first bleaching process, there is a less transferred charge density, which is
EP
probably attributed to the retention of Li+ ions in WO3 due to the blocking effect of Ta2O5 barrier
AC C
layer. For the 2nd cycle, since the larger coloration potential is applied, the deviation of transferred charge densities between for coloration and for bleaching becomes less, which, to some extent, suggests that the larger coloration potential promotes the faster transfer of Li+ ions into WO3 layer.
Fig. 3 Evolutions of transferred charge density versus cycle number corresponding to the CVs with different potential ranges (a) -1.5 V ~1.5 V, (b) -1.8 V ~1.5 V, and (c) -2.0 V ~1.5 V. Insets are their
ACCEPTED MANUSCRIPT corresponding in-situ transmittance at 550 nm, respectively.
Fig. 3a presents larger transferred charge densities for coloration than that for bleaching. Correspondingly, as the cycle number increases, the inset in Fig. 3a shows increasing optical transmittances of ECD in colored and bleached states, which is irreconcilable with that more
RI PT
charges are transferred for coloring process than for bleaching process. Two possible reasons could explain the phenomenon. One involves a relatively low potential for bleaching, which could not reversibly extract all these inserted Li+ ions from NiOx layer, and leads to an increasing optical transmittance for bleached states. Besides, as observed in Fig. 3b and 3c, the transferred charge
SC
density is much higher than that in Fig. 3a, which means that Li+ ions in LiNbO3 layer are sufficient for electrochromic behavior in the present potential range, the higher charge densities
M AN U
for coloration are attributed to the transfer of these Li+ ions in LiNbO3 layer. The Li+-ion trapping could result in reduction of nickel vacancies, and a simultaneous variation in dielectric properties of NiOx layer, and brings about an increasing voltage drop of potential across NiOx layer, as a result, the practical potential applied to drive charges becomes low and the transferred charge density decreases with the increasing cycle. Another possible reason for this behavior is the
TE D
trapping of Li+ ions in WO3. The Li+-ion trapping process gives rise to the formation of Li2WO4 [18] or Li2O in the WO3 film matrix [39], which also causes the coloration degradation of WO3 [11,40]. The increasing transmittance in bleached state implies that a growing amount of Li+ ions
EP
contributes to the bleaching of the ECD.
Meanwhile, some emphases should put on the Fig. 3b and 3c. (i) In addition to the first two
AC C
CVs, the imbalances of transferred charge densities for coloration and bleaching processes in each CV of ECDs are slight in these figures. (ii) However, it is of obvious difference between transferred charge density for current coloration process and for last bleaching process. When the former is less than the latter, the optical transmittance of ECDs in colored state becomes higher, as shown in insets (in Fig. 3b and 3c), which could be attributed to the declining charge density transferred into WO3 layer in the coloring processes. (iii) Different from the transmittance in bleached state in Fig. 3a, it keeps constant in Fig. 3b and 3c despite of the increasing transmittance in colored state (except the first few cycles). Taking (i), (ii) and (iii) into consideration, it seems that the Li+-ion trapping in NiOx causes the degradation of ECDs in Fig. 3b and 3c. On one hand, for one complete CV (first coloration, then bleaching), the transferred charge density for bleaching
ACCEPTED MANUSCRIPT processes is slightly larger than that for coloration processes (slight imbalances), which argues that it is less possible for Li+ ions to trap in WO3. On the other hand, (ii) and (iii) indicate that the transfer of Li+ ions from WO3 into NiOx is not completely reversible and a portion of them could not be extracted from NiOx layer. However, WO3 layer generally plays a primary and dominant
RI PT
role in the contribution to the electrochromic behavior of the ECDs. The dynamic behavior of Li+ ions in WO3 could be a very important factor in the degradation of these ECDs. Besides, the interfacial dependence of Li+-ion dynamic behavior in ECDs cannot be neglected. Therefore, there are possibly some more complicated dynamic behaviors pertaining to the elctrochromic processes
SC
of the ECDs. 3.3 Response characteristic
M AN U
Optical modulation, one of the most critical criteria for the assessment of ECDs, is dominated by the applied potential. The inorganic ECD was operated using multi-potential steps technique. Fig. 4 displays the effect of applied potentials on optical modulation of the ECD. As the potential span becomes wider, the optical modulation of ECD is obviously larger (in Fig. 4c). Corresponding photographs of the ECD in colored and bleached states are exhibited in Fig. 4a. As
TE D
shown in Fig. 4b, the induced peak current density is larger with the increasing potential, and the transferred charge density increases from 5.3 mC/cm2 to 19.5 mC/cm2 for coloration and from 6.2 mC/cm2 to 19.4 mC/cm2 for bleaching. These behaviors demonstrate that, as the applied potential
EP
increases, more Li+ ions could get rid of the confinement of relatively high energy barrier [41], and are smoothly transferred in the ECD during the coloration/bleaching processes as a result of
AC C
increasing the potential span.
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Fig. 4 Optical modulation of the ECD driven by step potentials with the duration of 40 s, (a) applied step potentials, (b) corresponding current density, and (c) in-situ optical transmittance at 550 nm.
Response rate of coloring and bleaching processes of the ECD is determined by the time it takes to achieve 90 % of total variation in optical transmittance between both stable electrochromic states [6]. Fig. 5 shows the relaxation behavior of current densities and response
TE D
characteristic of optical transmittance of the ECD operated at different positive potentials during bleaching processes. In Fig. 5a, the analyzed CAs with different potential ranges are 10th cycle where CAs tends to be stable. As observed from insets in Fig. 5a, accompanying with the
EP
increasing positive potential from 0.5 V to 1.5 V, the peak current density increases for bleaching and decreases for coloring, respectively. The relaxation time of corresponding current density
AC C
individually are shorter for bleaching and longer for coloring. As shown in Fig. 5b, the response time gradually decreases for both bleaching and coloration of the ECD.
ACCEPTED MANUSCRIPT Fig. 5 Variations in current density (a) and response times (b) of the ECD operated at different applied potentials during bleaching processes.
Concerning the response characteristics of the ECD operated at different potentials for coloring processes, as observed in Fig. 6a, the relaxation time of current densities during the
RI PT
coloring and bleaching processes increases with the increasing coloration potential. In Fig. 6b, shorter response time for coloring process is observed with the application of larger coloration potential. On the contrary, the response rate of bleaching process becomes slower. Since a larger
SC
coloration potential is applied for coloring process, Li+ ions are more readily extracted from NiOx layer, and simultaneously a larger amount of them are accommodated in the WO3 layer or even reside in the trap sites with higher energy barriers [41-42]. When the reverse potential is constant,
M AN U
the bleaching process, during which all Li+ ions are reversibly transferred from WO3 layer into NiOx layer, has to take a longer polarization duration. However, the increasing response time for the bleaching process seems inconsistent with the results in Fig. 5b. This behavior could be explicated by the following description. WO3 layer is the primary electrochromic layer and the dynamic behavior of Li+ ions inside this layer plays a dominant role in the response rate of the
TE D
ECDs. Furthermore, in one electrochromic layer, the amount of Li+ ions residing in trap sites with higher energy barrier is mostly decided by the potential for insertion of Li+ ions into the electrochromic layer.
EP
Compared to the reported results in our previous work [34], the faster response rate could be ascribed to the thinner electrochromic layers, barrier layers and ions electrolyte layer. In addition,
AC C
as shown in Fig. 5a and Fig. 6a, it should be stressed that the residual current densities at the end of both coloring and bleaching processes are approximately zero, which indicates a negligible electronic leakage current through the inorganic all-solid-state ECDs during these cycling electrochromic processes. This behavior, in accordance with the results in CVs, indicates an excellent optical memory of the ECDs [34].
RI PT
ACCEPTED MANUSCRIPT
Fig. 6 Variations in current density (a) and response times (b) of the ECD operated at different applied
SC
potentials during coloring processes.
3.4 Optical memory effect
M AN U
Excellent optical memory effect is a significant feature of ECDs serving as energy-efficient smart windows. Dynamic electrochromic windows can be readily controlled to provide additional comfort [3,43], which requires that the ECDs at all colored and bleached extents have an excellent optical memory. Fig. 7 and Fig. 8 shows the optical memory and open-circuit potential randomly at three times during the coloring and bleaching processes of the ECD, respectively. As expected,
TE D
since no obvious electronic leakage current is observed in CVs and CAs, the ECD has an excellent optical memory in Fig. 7(a~c) and in Fig. 8(a~c), which implies a nearly perfect interfaces between layers in the ECD. Such phenomenon suggests that the incorporation of Ta2O5 layers
EP
effectively limits the formation of defects such as pinholes at the interface and in the ion electrolyte layer, contributing to the electronic leakage. Accompanying with the enhancing extent
AC C
of coloring, shown in Fig. 7(d~f), the open-circuit potential of the ECD is lowering. During the bleaching process, the initial open-circuit potential is closer to the applied positive potential with the increasing extent of bleaching, illustrated in Fig. 8(d~f). Owing to the transferring of Li+ ions between WO3 and NiOx layers, the chemical potentials (µ) of both the layers are changing. The over-potential (η) between two electrodes of the ECD can be expressed as [13]: η = VA - ∆µ/Ne, wherein, VA and ∆µ denote the applied potential and the variation in the chemical potential throughout the ECD. N and e is Avogadro’s constant and unit charge. The open-circuit potential corresponds to the second term on the right side of the equation. When η is higher than 0, the
ACCEPTED MANUSCRIPT electrochromic behavior occurs to the ECD till the η approaches zero, accompanied by the
SC
RI PT
increasing chemical potential.
M AN U
Fig. 7 Optical memory effect (a~c) and corresponding open-circuit potential (d~f) at random times
EP
TE D
during the coloring process of the ECD operated between -2.0 and 1.0 V.
AC C
Fig. 8 Optical memory (a~c) and corresponding open-circuit potential (d~f) at random times during the bleaching process of the ECD operated between -2.0 and 1.0 V.
4 Conclusions
In summary, the inorganic all-solid-state ECD with seven superimposed layers of
Glass/ITO/NiOx/Ta2O5/LiNbO3/Ta2O5/WO3/ITO were monolithically prepared by magnetron sputtering. In the CVs of the ECD, the imbalance of transferred charges during coloring and bleaching processes is mainly attributed to the trapping behavior of Li+ ions in electrochromic layers. Larger optical modulation can be obtained for the ECD operated within wider potential range. In the case of response characteristic, the higher positive potentials during bleaching
ACCEPTED MANUSCRIPT processes would lead to faster response rates for the ECD. However, the larger coloration potentials during coloring processes brings about shorter coloring and longer bleaching response times. These dynamic behaviors are associated with the response characteristics of WO3 layer acting as the primary electrochromic layer in the ECD. The ECD has excellent optical memory
RI PT
effects in coloring and bleaching processes. And corresponding open-circuit potentials link to the variation in the electrochemical potential due to the transfer of charges in the ECD. Acknowledgments
SC
The authors sincerely appreciate the financial support of the National Program on Key Research Project of China (2016YFB0303901), the Beijing Municipal Natural Science Foundation the
Fundamental
Re-search
Funds
for
the
Central
Universities
(Grant
M AN U
(2161001),
No.YWF-16-JCTD-B-03), National Natural Science Foundation of China (Grant No. 21701003), China Postdoctoral Science Foundation Grant (2015M580035 and 2017T100022), and the Fundamental Research Funds for the Central Universities (KG12030401, ZG216S17I3,
References
TE D
ZG226S17Q4).
[1] Claes G. Granqvist, Electrochromics for smart windows: Oxide-based thin films and devices, Thin Solid Films, 2014, 564, 1-38.
EP
[2] K. J. Patel, G. G. Bhatt, J. R. Ray, Priya Suryavanshi, C. J. Panchal, All-inorganic solid-state electrochromic devices: a review, Journal of Solid State Electrochemistry, 2016, 21, 1-11.
AC C
[3] Guofa Cai, Jiangxin Wang, Pooi See Lee, Next-Generation Multifunctional Electrochromic Devices, Acc. Chem. Res., 2016, 49, 1469-1476. [4] Sami Oukassi, Cédric Giroud-Garampon, Christophe Dubarry, Cédric Ducros, Raphaël Salot, All inorganic thin film electrochromic device using LiPON as the ion conductor, Solar Energy Materials & Solar Cells, 2016, 145, 2-7. [5] Guofa Cai, Peter Darmawan, Mengqi Cui, Jingwei Chen, Xu Wang, Alice Lee-Sie Eh, Shlomo Magdassi, Pooi See Lee, Inkjet-printed all solid-state electrochromic devices based on NiO/WO3 nanoparticle complementary electrodes, Nanoscale, 2016, 8, 348-357. [6] Huiying Qu, Hangchuan Zhang, Na Li, Zhongqiu Tong, Jing Wang, Jiupeng Zhao, Yao Li, A
ACCEPTED MANUSCRIPT rapid-response electrochromic device with significantly enhanced electrochromic performance, RSC Adv., 2015, 5, 803-806 [7] Chien-Jen Tang, Jia-Ming Ye, Yueh-Ting Yang, Ju-Liang He, Large-area flexible monolithic ITO/WO3/Nb2O5/NiVOx/ITO electrochromic devices prepared by using magnetron sputter
RI PT
deposition, Optical Materials, 2016, 55, 83-89. [8] Rui-Tao Wen, Sara Malmgren, Claes G. Granqvist, Gunnar A. Niklasson, Degradation Dynamics for Electrochromic WO3 Films under Extended Charge Insertion and Extraction: Unveiling Physicochemical Mechanisms, ACS Appl. Mater. Interfaces, 2017, 9, 12872-12877.
SC
[9] Junling Zhou, Gui Luo, Youxiu Wei, Jianming Zheng, Chunye Xu, Enhanced electrochromic performances and cycle stability of NiO-based thin films via Li-Ti co-doping prepared by sol-gel
M AN U
method, Electrochimica Acta, 2015, 186, 182-191.
[10] Rui-Tao Wen, Gunnar A. Niklasson, Claes G. Granqvist, Strongly Improved Electrochemical Cycling Durability by Adding Iridium to Electrochromic Nickel Oxide Films, ACS Appl. Mater. Interfaces, 2015, 7, 9319-9322.
[11] Rui-Tao Wen, Miguel A. Arvizu, Michael Morales-Luna, Claes G. Granqvist, Gunnar A.
TE D
Niklasson, Ion Trapping and Detrapping in Amorphous Tungsten Oxide Thin Films Observed by Real-Time Electro-Optical Monitoring, Chem. Mater., 2016, 28, 4670-4676. [12] Guofa Cai, Mengqi Cui, Vipin Kumar, Peter Darmawan, Jiangxin Wang, Xu Wang, Alice
EP
Lee-Sie Eh, Kai Qian, Pooi See Lee, Ultra-large optical modulation of electrochromic porous WO3 film and the local monitoring of redox activity, Chem. Sci., 2016, 7, 1373-1382.
AC C
[13] Richard S. Crandall, Brian W. Faughnan, Dynamics of coloration of amorphous electrochromic films of WO3 at low voltages, Applied Physics Letters, 28, 1976, 95-97. [14] Feng Lin, Justin B. Bult, Sanjini Nanayakkara, Anne C. Dillon, Ryan M. Richards, Jeffrey L. Blackburn, Chaiwat Engtrakul, Graphene as an Efficient Interfacial Layer for Electrochromic Devices, ACS Appl. Mater. Interfaces, 2015, 7, 11330-11336. [15] Francisco Fabregat-Santiago, Germà Garcia-Belmonte, Juan Bisquert, Noemí S. Ferriols, Paulo R. Bueno, Elson Longo, Julio Suay Antón, Socorro Castro-García, Dynamic Processes in the Coloration of WO3 by Lithium Insertion, Journal of The Electrochemical Society, 2001, 148, E302-E309. [16] R.-T.Wen, C.G. Granqvist, G.A. Niklasson, Eliminating degradation and uncovering
ACCEPTED MANUSCRIPT ion-trapping dynamics in electrochromic WO3 thin films, Natural Materials, 2015, 14, 996-1002. [17] Miguel A. Arvizu, Rui-Tao Wen, Daniel Primetzhofer, Jolanta E. Klemberg-Sapieha, Ludvik Martinu, Gunnar A. Niklasson, Claes G. Granqvist, Galvanostatic Ion Detrapping Rejuvenates Oxide Thin Films, ACS Appl. Mater. Interfaces, 2015, 7, 26387-26390.
RI PT
[18] P.V. Ashrit, G. Bader, Vo-Van Truong, Electrochromic properties of nanocrystalline tungsten oxide thin films, Thin Solid Films 320, 1998, 324-328.
[19] M. A. Arvizu, C. A. Triana, B. I. Stefanov, C. G. Granqvist, G. A. Niklasson, Electrochromism in sputter-deposited W-Ti oxide films: Durability enhancement due to Ti, Solar
SC
Energy Materials & Solar Cells, 2014, 125, 184-189.
[20] M. Morales-Luna, M. A. Arvizu, C. G. Granqvist, G. A. Niklasson, Sputter deposited
Conference Series, 2016, 682, 012021.
M AN U
W1-x-yNixTiyO3 thin films: Electrochromic properties and durability, Journal of Physics:
[21] Rui-Tao Wen, Claes G. Granqvist, Gunnar A. Niklasson, Cyclic voltammetry on sputter-deposited films of electrochromic Ni oxide: Power-law decay of the charge density exchange, Applied Physics Letters, 2014, 105, 163502.
TE D
[22] Rui-Tao Wen, Claes G. Granqvist, Gunnar A. Niklasson, Anodic Electrochromic Nickel Oxide Thin Films: Decay of Charge Density upon Extensive Electrochemical Cycling, Chem. Electro. Chem., 2016, 3, 266-275.
EP
[23] Yoshijiro Ushio, Akira Ishikawa, Tatsuo Niwa, Degradation of the electrochromic nickel oxide film upon redox cycling, Thin Solid Films, 1996, 280, 233-237.
AC C
[24] A. Pennisi, C. M. Lampert, Optical Properties of Electrochromic Nickel Oxide Devices Utilizing A Polymeric Electrolyte, SPIE Vol. 1016 Optical Materials Technology for Energy Efficiency and Solar Energy Conversion VII, 1988, 131-144. [25] Anneke Georg, Andreas Georg, Electrochromic device with a redox electrolyte, Solar Energy Materials & Solar Cells, 2009, 93, 1329-1337. [26] J. Zhang, J. P. Tu, X. H. Xia, Y. Qiao, Y. Lu, An all-solid-state electrochromic device based on NiO/WO3 complementary structure and solid hybrid polyelectrolyte, Solar Energy Materials & Solar Cells, 2009, 93, 1840-1845. [27] C. Person, I. Porqueras, M. Vives, C. Corbella, A. Pinyol, E. Bertran, Degradation of a solid state electrochromic device, Solid State Ionics, 2003, 165, 73-80.
ACCEPTED MANUSCRIPT [28] Tatsuo Niwa, Osamu Takai, Optical and electrochemical properties of all-solid-state transmittance-type electrochromic devices, Thin Solid Films, 2010, 518, 1722-1727. [29] Kwang-Soon Ahn, Yoon-Chae Nah, Jin-Young Park, Yung-Eun Sung, Ki-Yun Cho, Seung-Shik Shin, Jung-Ki Park, Bleached state transmittance in charge-unbalanced all-solid-state
RI PT
electrochromic devices, Appl. Phys. Lett., 2003, 82, 3379. [30] Claes G.Granqvist, Oxide electrochromics: An introduction to devices and materials, Solar Energy Materials & Solar Cells, 2012, 99, 1-13.
[31] M. Schmitt, M. A. Aegerter, Electrochromic properties of pure and doped Nb2O5 coatings
SC
and devices, Electrochimica Acta, 2001, 46, 2105-2111.
[32] L. Kullman, A. Azens, C. G. Granqvist, Decreased electrochromism in Li-intercalated Ti
M AN U
oxide films containing La, Ce, and Pr, Journal of Applied Physics, 1997, 81, 8002. [33] Mark Samuel Burdis, Bryan D. Greer, Douglas Glenn John Weir, Electrochromic Devices Having Improved Ions Conducting Layers, United States Patent, US 8,004,744 B2, Aug. 23, 2011 [34] Qirong Liu, Guobo Dong, Qianqian Chen, Junji Guo, Yu Xiao, Marie-Paule Delplancke-Ogletree, François Reniers, Xungang Diao, Charge-transfer kinetics and cyclic
TE D
properties of inorganic all-solid-state electrochromic device with remarkably improved optical memory, Solar Energy Materials and Solar Cells, 2018, 174, 545-553. [35] K. J. Patel, M. S. Desai, C. J. Panchal, Studies of ZrO2 electrolyte thin-film thickness on the
275-279.
EP
all-solid thin-film electrochromic devices. Journal of Solid State Electrochemistry, 2014, 19,
AC C
[36] J. Nagai, G. D. McMeeking, Y. Saitoh, Durability of electrochromic glazing, Solar Energy Materials & Solar Cells, 1999, 56, 309-319. [37] Nada A. O'Brien, J. Gordon, H. Mathew, Bryant P. Hichwa, Electrochromic coatings-applications and manufacturing issues, Thin Solid Films, 1999, 345, 312-318. [38] Miao-Syuan Fan, Sheng-Yuan Kao, Ting-Hsiang Chang, R. Vittal, Kuo-Chuan Ho, A high contrast solid-state electrochromic device based on nano-structural Prussian blue and poly(butylviologen) thin films, Solar Energy Materials & Solar Cells, 2016, 145, 35-41. [39] Sung Jong Yoo, Ju Wan Lim, Yung-Eun Sung, Improved electrochromic devices with an inorganic solid electrolyte protective layer, Solar Energy Materials & Solar Cells, 2006, 90, 477-484.
ACCEPTED MANUSCRIPT [40] Rui-Tao Wen, Gunnar A. Niklasson, Claes G. Granqvist, Sustainable Rejuvenation of Electrochromic WO3 Films, ACS Appl. Mater. Interfaces 2015, 7, 28100-28104. [41] Juan Bisquert, Analysis of the kinetics of ion intercalation: Ion trapping approach to solid-state relaxation processes, Electrochimica Acta, 2002, 47, 2435-2449.
RI PT
[42] Juan Bisquert, Vyacheslav S. Vikhrenko, Analysis of the kinetics of ion intercalation. Two state model describing the coupling of solid state ion diffusion and ion binding processes, Electrochimica Acta, 2002, 47, 3977-3988.
[43] N. Sbar, M. Badding, R. Budziak, K. Cortez, L. Laby, L. Michalski, T. Ngo, S. Schulz, K.
AC C
EP
TE D
M AN U
Materials & Solar Cells, 1999, 56, 321-341.
SC
Urbanik, Progress toward durable, cost effective electrochromic window glazings, Solar Energy