Available online at www.sciencedirect.

com

Journal of the European Ceramic Society 30 (2010) 1139–1144

Electrophoretic deposition (EPD) of WO3 nanorods

for electrochromic application
Eugene Khoo, Pooi See Lee, Jan Ma ∗
School of Materials Science and Engineering, Nanyang Technological University, Singapore 639798, Singapore
Available online 10 June 2009

Abstract
Electrophoretic deposition (EPD) was applied in coating hydrothermally synthesized crystalline tungsten oxide (WO3 ) nanorods onto ITO glass for
electrochromic application. Nanorods suspension of 10 mg/cm3 was used in the EPD with optimum electric field of 5–6 V/cm. Saturation in WO3
deposited amount at electric field >7 V/cm was observed during constant voltage EPD. This could be attributed to the oxide layer shielding effect
on the electric field induced electrophoresis. Constant current EPD from 0.2 mA/cm2 to 1.4 mA/cm2 was also performed for the WO3 nanorods.
The deposited amount of nanorods was found to be proportional to the current density from 0.2 mA/cm2 to 0.8 mA/cm2 under constant deposition
duration. However, the deposited amount decreased at current density >0.8 mA/cm2 . This could be due to the high deposition rate that resulted in
poor adhesion and hence nanorods peel off during the substrate removal. It was noted that the EPD of nanorods followed a linear relationship in I
vs. t−1/2 plot according to Cottrell equation, which implied that the reaction was a diffusion controlled process. The EPD coated substrate was tested
in 1 M LiClO4 /propylene carbonate (PC) electrolyte for electrochromic studies. The porous WO3 nanorods layer exhibited optical modulation
T700 nm of 40%, moderate coloration time tc70% of 28.8 s and improved bleaching time tb70% of 4.5 s, which could be due to the porous oxide layer
with large surface area that facilitates the ion insertion/extraction and the electrolyte penetration in the oxide layer shortens the ionic diffusion
length of Li.
© 2009 Elsevier Ltd. All rights reserved.

Keywords: Transition metal oxides WO3 ; Optical properties; Suspensions; Functional applications; Electrochromic

1. Introduction potential for large scale production and forming crystalline

structure with good stability.11 However, one of the process
Tungsten oxide (WO3 ) is one of the widely studied transition drawbacks is that the hydrothermally synthesized products are
metal oxides for its attractive application such as energy efficient dispersed as a suspension and additional assembly steps are
windows (smart windows), antiglare automobile rear-view mir- required to adhere the dispersed products onto the substrate.
rors, sunroofs and sensors.1–5 WO3 has the optical modulation Electrophoretic deposition (EPD), on the other hand, is an attrac-
between transparent and blue color upon the ion-electron double tive process where the charged particles or nanostructures in a
injection, which is known as electrochromism. The reaction can suspension can be driven onto the electrode surface under the
be expressed as influence of an electric field. This method is efficient in forming
a film or coating on conductive surface (even a complex sur-
WO3 (transparent) + xe− + xM+ ⇔ Mx WO3 (blue) (1) face morphology) with the control in film thickness and porosity
where M+ stands for the cations such as H+ , Li+ Na+ or K+ . through the electric field, pH value and deposition duration. EPD
Various methods such as sol–gel,6 electrodeposition,7 has provided flexibilities in designing electrochromic devices
hydrothermal process,8 evaporation9 and sputtering10 have been with different degrees of optical modulation (film thickness) and
applied in the formation of crystalline and amorphous WO3 . switching time (film porosity) and hence widen their application
Among these methods, hydrothermal process has received huge in advanced technologies.
attention for its advantages such as simple operation, low-cost, EPD has been applied in surface coating, ceramics process-
ing and forming composite materials12–14 where particles are
usually used as the starting material in the deposition. Recently,
∗ Corresponding author. Tel.: +65 67906214; fax: +65 67909081. Girishkumar et al.15 and Boccaccini et al.16 have reported the
E-mail address: [email protected] (J. Ma). 1D nanostructure EPD of CNT and Yu and Zhou17 have success-

0955-2219/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jeurceramsoc.2009.05.014
1140 E. Khoo et al. / Journal of the European Ceramic Society 30 (2010) 1139–1144

fully demonstrated the EPD of titanate nanotubes with diameter ried out in an electrolyte of 1.0 M LiClO4 in propylene carbonate
around 10 nm. These results suggest that the EPD is also (PC). Transmittance variation and color changes were detected
applicable in 1D nanostructures assembly. In this work, WO3 through UV–vis spectrometer (Shimadzu UV-2501PC) at visible
1D nanorods constant voltage and constant current EPD were wavelength ranged from 300 nm to 900 nm. Optical modula-
performed on a transparent conductive substrate (ITO glass). tion and switching time were extracted from the transmittance
Optimum deposition conditions were identified with a discus- spectra.
sion in the deposition mechanism. The resultant electrochromic
performances such as optical modulation and switching time of 3. Results and discussion
the WO3 nanorods coated substrates were also studied.
3.1. Electrophoretic deposition (EPD)
2. Experimental details
WO3 nanorods suspension of 10 mg/cm3 was used in the con-
2.1. WO3 synthesis stant voltage EPD. The applied electric field ranged from 3 V/cm
to 8 V/cm for 900 s deposition and the I–t behaviors are recorded
Tungsten oxide (WO3 ) nanorods were prepared through in Fig. 1a. With the various electric field strength studied, the
hydrothermal process: 0.825 g of Na2 WO4 ·2H2 O (Aldrich, EPD of WO3 nanorods can be characterized into three regions:
99%) and 0.290 g of NaCl (Sigma Ultra, 99.5%) were dissolved no deposition (<4 V/cm), uniform deposition (5–6 V/cm) and
in 20 ml of de-ionized water with a mole ratio of 1:2. The solu- saturated deposition (>7 V/cm). For the electric field <4 V/cm,
tion was adjusted with 3.0 M HCl until a pH of 2, then transferred the electric field induced driving force is not strong enough to
into the Teflon liner of the stainless steel autoclave and heated drive the WO3 nanorods onto the ITO glass. Hence, only small
at 180 ◦ C for 24 h. White precipitate of nanorods was obtained amount of nanorods is observed under the FESEM (Fig. 2a).
and centrifuged with ethanol and de-ionized water for several When the electric field is increased beyond 5 V/cm, uniform
times to remove any left over precursors. The WO3 nanorod WO3 nanorods deposition was is observed as a porous oxide
was found to be a hexagonal single crystal with a growth direc- layer on top of the ITO (Fig. 2b). However, for the electric field
tion of (0 0 0 2). The diameter and length of the nanorod were >7 V/cm, the deposition ceases after a period of time as showed
around 100 nm and 2 ␮m, respectively.18 in Fig. 1a with a sharp drop in the current density until nearly

2.2. Electrophoretic deposition (EPD)

The nanorods suspension was prepared by dispersing the

WO3 nanorods in de-ionized water with 10 mg/cm3 concen-
tration, −40 mV zeta potential and a pH value around 6.77.
Transparent conductive substrate – ITO coated glass (Delta
Technologies, Rs = 15–25 ) was used as the working electrode
in order to study the electrochromic performances of the WO3
nanorods while a Pt wire was used as the counter electrode. The
working electrode (ITO glass) was kept at positive voltage or
current with distance between the two electrodes remained as
1 cm. The deposition was carried out by Autolab Potentiostat
(PGSTAT302) with a maximum voltage limit of 10 V. Con-
stant voltage and current EPD were carried out with an electric
field ranged from 3 V/cm to 8 V/cm and current density ranged
from 0.2 mA/cm2 to 1.4 mA/cm2 , respectively. The current and
voltage responses were recorded with respect to the deposi-
tion duration for further analysis. The coated substrates were
overnight air-dried before morphology and electrochromic stud-
ies. Surface morphology of the coated ITO was characterized
using field emission scanning electron microscopy (FESEM,
JOEL6340F) and the oxide layer thickness was checked by
Alpha-step Surface Profiler.

2.3. Electrochromic studies

Optical modulation and switching studies were carried out

in a two-electrode system for ±3 V. The coated ITO was used Fig. 1. (a) Constant voltage EPD of 10 mg/cm3 WO3 nanorods suspension
as the working electrode and a Pt wire was used as the counter for 900 s deposition and (b) WO3 nanorods deposited amount at 5 V/cm for
and reference electrode. All electrochemical studies were car- 300–1500 s deposition duration.
 E. Khoo et al. / Journal of the European Ceramic Society 30 (2010) 1139–1144 1141

Fig. 2. FESEM images of the deposited WO3 nanorods on ITO glass under (a) 4 V/cm and (b) 6 V/cm and (c) higher magnification of 6 V/cm deposited WO3
nanorods.

zero. This might be due to the electrical resistance on the work- maximum voltage of the potentiostat (10 V) after a period of
ing electrode increases with the oxide layer thickness and hence deposition while optimum deposition was observed to be around
decreases the electric field induced electrophoresis in the sus- 0.8 mA/cm2 . The deposited amount of WO3 is plotted with
pension. The oxide layer will weaken the driving force (due to respect to the current density (Fig. 3b). The deposited amount
the voltage drop across the oxide layer) and hence decrease the is found to increase with respect to the current density, which
nanorods motion.13 The nearly zero current density implies that implies that the passed charges (current) are proportional to the
the electric field induced driving force is not strong enough to electrophoresis deposited nanorods. However, at the higher cur-
drive the nanorods onto the electrode after the build-up of the rent density (>0.8 mA/cm2 ), the deposited amount is recorded
oxide layer. In addition, oxide layer peeling was observed at to be reduced, which could be due to the ITO substrate degra-
7 V/cm and higher electric field during the substrate removal, dation or poor adhesion of the nanorods to the substrate. ITO
which causes non-uniform coverage on the ITO surface and is glass has been demonstrated as the working electrode for aque-
undesirable for the electrochromic application. Thus, it is noted ous system EPD up to 40 V/cm21 and 2.5 mA/cm2 22 and no ITO
that 5–6 V/cm gives the optimum deposition electric field for degradation was observed. Thus, ITO electrochemical degrada-
the present nanorods suspension. Fig. 1b shows the deposited tion at high current density is not the reason for the decrease in
amount with respect to the deposition duration under 5 V/cm. deposited amount. Hence, the decrease of deposited mass can be
The relationship turns non-linear after 900 s with a nearly con- attributed to the accumulation of large amount of WO3 nanorods
stant deposited amount, which corresponds to a thickness of at the electrode within a short duration under a high driving
1.2 ␮m. This result further supports that EPD will cease after force, which results in the formation of a very porous, inho-
the build-up of the oxide layer and longer deposition duration mogeneous film with reduced adhesion at the electrode surface.
does not significantly improve the deposited amount.13,19,20 As a result, peel-off was observed during the deposited sub-
Constant current density EPD was carried out using strate removal and lower deposited amount was remained on the
0.6–1.4 mA/cm2 (Fig. 3a). The voltage increased up to the substrate.
1142 E. Khoo et al. / Journal of the European Ceramic Society 30 (2010) 1139–1144

Fig. 3. (a) Constant current EPD of 10 mg/cm3 WO3 nanorods suspension for
900 s deposition and (b) deposited amount of WO3 nanorods on ITO glass for Fig. 4. Currents density vs. inverse square root of deposition duration plot under
900 s deposition duration at various current densities. (a) 5 V/cm and (b) 6 V/cm deposition.

In summary, the moderate electric field of 5–6 V/cm and detected current containes significant parasitic currents such as
current density 0.8 mA/cm2 are reckoned to be the optimum water electrolysis.
depositions condition for uniform coverage and avoiding poor
nanorods adhesion. Oxide layer shielding effect is observed and 3.2. Electrochromic studies
it could be probably solved by constant current EPD, with fur-
ther increased deposition voltage, a longer deposition duration The EPD coated WO3 nanorods were tested for UV–vis
and optimum current density. transmittance under a switching voltage of ±3 V (Fig. 5a).
As reported by Koura et al.,23 a diffusion controlled process The WO3 nanorods were bleached at +3 V with around 60%
can be described by Cottrell equation: of transmittance at 700 nm and exhibited strong absorption of
KC red-end light at −3 V biasing. The transmittance modulation
i= (2) at 700 nm (T700 nm ) is found to be around 40%, which is
(πDt)1/2 comparable to other crystalline films.24,25 Switching time of
where i is the current, K is the kinetic constant, D is the diffusion the WO3 nanorods is extracted from the transmittance data
coefficient and t is the deposition duration. The equation can be in Fig. 5b by applying a square wave voltage of ±3 V to the
simplified to i = kt−1/2 when K, C and D are kept as constant. oxide layer (simulating the switching condition). The switch-
Fig. 4a and b show the experimental data of 5 V/cm and 6 V/cm ing time – bleaching (tb ) and coloring (tc ) time are calculated
deposition separately. It is observed that both cases follow the as the time required for 70% optical modulation in order to
Cottrell equation where the i–t−1/2 curves shows an almost linear compare with literatures. Coloration time tc70% is found to be
relationship with a similar slope, which suggests the deposition 28.8 s, which is comparable to the reported values (tb ∼ 25 s).26
presented in this work is a diffusion controlled process. Devi- However, the bleaching time tb70% of 4.5 s is found to be much
ation from the slope is observed at long deposition time. This superior than the crystalline26 (tb ∼ 22 s) and amorphous WO3
deviation could be due to the build-up of the oxide layer which (tb ∼ 18–36 s).26,27 This improvement is attributed to the porous
decreases the nanorods electrophoresis in the suspension and the nature of the oxide layer which consists of nanorods. The porous
 E. Khoo et al. / Journal of the European Ceramic Society 30 (2010) 1139–1144 1143

i vs. t−1/2 plot showed a linear relationship. The EPD coated

WO3 nanorods exhibited a moderate coloration time tc70% of
28.8 s and significantly improved bleaching time tb70% of 4.5 s.
The improvement was due to the porous nanorods layer which
facilitated the ion extraction by shortening the ionic diffusion
length and the large surface area acted as the reaction sites for
ion insertion/extraction.

References

1. Niklasson, G. A. and Granqvist, C. G., Electrochromics for smart windows:

thin films of tungsten oxide and nickel oxide, and devices based on these. J.
Mater. Chem., 2007, 17, 127–156.
2. Guérin, J., Aguir, K., Bendahan, M. and Lambert-Mauriat, C., Thermal
modeling of a WO3 ozone sensor response. Sens. Actuators B, 2005, 104,
289–293.
3. Lampert, C. M., Large-area smart glass and integrated photovoltaics. Sol.
Energy Mater. Sol. Cells, 2003, 76, 489–499.
4. Zhao, Y., Feng, Z. C. and Liang, Y., Pulsed laser deposition of WO-
base film for NO gas sensor application. Sens. Actuators B, 2000, 66,
171–173.
5. Granqvist, C. G., Electrochromic tungsten oxide films: review of progress
1993–1998. Sol. Energy Mater. Sol. Cells, 2000, 60, 201–262.
6. Yoo, S. J., Lim, J. W., Sung, Y.-E., Jung, Y. H., Choi, H. G. and Kim, D.
K., Fast switchable electrochromic properties of tungsten oxide nanowire
bundles. Appl. Phys. Lett., 2007, 90, 173126.
7. Deepa, M., Srivastava, A. K., Sood, K. N. and Agnihotry, S. A., Nanostruc-
tured mesoporous tungsten oxide films with fast kinetics for electrochromic
smart windows. Nanotechnology, 2006, 17, 2625–2630.
8. Gu, Z., Li, H., Zhai, T., Yang, W., Xia, Y., Ma, Y. et al., Large-scale synthesis
of single-crystal hexagonal tungsten trioxide nanowires and electrochemical
lithium intercalation into the nanocrystals. J. Solid State Chem., 2007, 180,
98–105.
Fig. 5. (a) UV–vis transmittance of WO3 nanorods at bleaching (+3 V) and 9. Liao, C.-C., Chen, F.-R. and Kai, J.-J., Electrochromic properties of
coloring (−3 V); (b) switching study by applying a square wave signal ±3 V nanocomposite WO3 films. Sol. Energy Mater. Sol. Cells, 2007, 91,
at 632.8 nm. All electrochemical studies were carried out in 1 M LiClO4 /PC 1258–1266.
electrolyte. 10. Subrahmanyam, A. and Karuppasamy, A., Optical and electrochromic prop-
erties of oxygen sputtered tungsten oxide (WO3 ) thin films. Sol. Energy
Mater. Sol. Cells, 2007, 91, 266–274.
oxide layer allows the electrolyte penetration and hence shortens
11. Lee, S.-H., Deshpande, R., Parilla, P. A., Jones, K. M., To, B., Mahan, A. H.
the ionic diffusion length from the electrolyte to the centre of et al., Crystalline WO3 nanoparticles for highly improved electrochromic
the oxide. On the other hand, the high surface to volume ratio of applications. Adv. Mater., 2006, 18, 763–766.
the nanorods oxide layer facilitats the Li+ ion movement across 12. Van der Biest, O. O. and Vandeperre, L. J., Electrophoretic deposition of
the oxide–electrolyte interface by a providing large amount of materials. Annu. Rev. Mater. Sci., 1999, 29, 327–352.
13. Sarkar, P. and Nicholson, P. S., Electrophoretic deposition (EPD): mecha-
reaction sites. Hence, shorter switching time is observed as the
nisms, kinetics, and application to ceramics. J. Am. Ceram. Soc., 1996, 79,
result of these two factors. 1987–2002.
14. Boccaccini, A. R. and Zhitomirsky, I., Application of electrophoretic and
4. Conclusion electrolytic deposition techniques in ceramics processing. Curr. Opin. Solid
State Mater. Sci., 2002, 6, 251–260.
15. Girishkumar, G., Vinodgopal, K. and Kamat, P. V., Carbon nanostruc-
WO3 nanorods electrophoretic deposition (EPD) has been tures in portable fuel cells: single-walled carbon nanotube electrodes for
demonstrated for a thin layer coating ∼1.2 ␮m for elec- methanol oxidation and oxygen reduction. J. Phys. Chem. B, 2004, 108,
trochromic application. The optimum EPD conditions were 19960–19966.
16. Boccaccini, A. R., Cho, J., Roether, J. A., Thomas, B. J. C., Minay, E. J. and
found to be at 0.8 mA/cm2 or 5–6 V/cm in 10 mg/cm3 WO3
Shaffer, M. S. P., Electrophoretic deposition of carbon nanotubes. Carbon,
nanorods suspension. Oxide layer shielding effect was observed 2006, 44, 3149–3160.
in the constant voltage EPD while the nanorods deposition 17. Yu, J. and Zhou, M., Effects of calcination temperature on microstructures
ceased after 900 s of deposition. Constant current EPD has been and photocatalytic activity of titanate nanotube films prepared by an EPD
performed and the deposited amount was found to be propor- method. Nanotechnology, 2008, 19, 1–6.
18. Wang, J., Khoo, E., Lee, P. S. and Ma, J., Synthesis, assembly, and elec-
tional to the charge passed through the electrode. However, oxide
trochromic properties of uniform crystalline WO3 nanorods. J. Phys. Chem.
layer peel off occured at high current density (>0.8 mA/cm2 ) due C, 2008, 112, 14306–14312.
to the poor adhesion of the nanorods at high deposition rate. The 19. Kuwabara, K. and Noda, Y., Electrophoretic preparation of antimonic acid
process was shown to be a diffusion controlled process as the film. J. Mater. Sci., 1993, 28, 5257–5261.
1144 E. Khoo et al. / Journal of the European Ceramic Society 30 (2010) 1139–1144

20. Ma, J. and Cheng, W., Deposition and packing study of sub-micron PZT 24. Liao, C.-C., Chen, F.-R. and Kai, J.-J., WO3−x nanowires based elec-
ceramics using electrophoretic deposition. Mater. Lett., 2002, 56, 721–727. trochromic devices. Sol. Energy Mater. Sol. Cells, 2006, 90, 1147–1155.
21. Jung, S. M., Jung, H. Y. and Suh, J. S., Horizontally aligned carbon nan- 25. Joraid, A. A., Comparison of electrochromic amorphous and crystalline
otube field emitters fabricated on ITO glass substrates. Carbon, 2008, 46, electron beam deposited WO3 thin films. Curr. Appl. Phys., 2009, 9, 73–79.
1973–1977. 26. Sallard, S., Brezesinski, T. and Smarsly, B. M., Electrochromic stability of
22. Yui, T., Mori, Y., Tsuchino, T., Itoh, T., Hattori, T., Fukushima, Y. et al., Syn- WO3 thin films with nanometer-scale periodicity and varying degrees of
thesis of photofunctional titania nanosheets by electrophoretic deposition. crystallinity. J. Phys. Chem. C, 2007, 111, 7200–7206.
Chem. Mater., 2005, 17, 206–211. 27. Deepa, M., Singh, D. P., Shivaprasad, S. M. and Agnihotry, S. A., A
23. Koura, N., Tsukamoto, T., Shoji, H. and Hotta, T., Preparation of vari- comparison of electrochromic properties of sol–gel derived amorphous
ous oxide films by an electrophoretic deposition method: a study of the and nanocrystalline tungsten oxide films. Curr. Appl. Phys., 2007, 7,
mechanism. Jpn. J. Appl. Phys., 1995, 34, 1643–1647. 220–229.