Electrochromic characterization of nickel oxide films grown on ITO/glass by liquid phase deposition

Article abstract of DOI:10.1149/1.3187911

Uniform nickel oxide film is grown on a conducting glass substrate with an aqueous solution of saturated NiF₂·4 H₂O and H ₃ BO₃. The quality of NiO is improved after thermal annealing at 300°C in air from the de

Full text of DOI:10.1149/1.3187911

Journal of The Electrochemical Society, 156 ͑10͒ D395-D399 ͑2009͒
D395
0
013-4651/2009/156͑10͒/D395/5/$25.00 © The Electrochemical Society
Electrochromic Characterization of Nickel Oxide Films
Grown on ITO/Glass by Liquid Phase Deposition
z
Ming-Kwei Lee and Cho-Han Fan
Department of Electrical Engineering, National Sun Yat-sen University, Kaohsiung 80424, Taiwan
Uniform nickel oxide film is grown on a conducting glass substrate with an aqueous solution of saturated NiF ·4H O and H BO .
2
2
3
3
The quality of NiO is improved after thermal annealing at 300°C in air from the decrease in oxygen vacancy and better F ion
passivation on the defects and the dangling bonds. The transmittance of as-deposited NiO/indium tin oxide ͑ITO͒/glass with a
thickness of 100 nm is about 78% and is improved to 88% after annealing at a wavelength of 550 nm. By the electrochromic
cycling test done 50 times on the annealed NiO film, the transparency ratio is kept at 48% between the fully colored state and the
fully bleached state at a wavelength of 550 nm. By the memory time test, the annealed liquid phase deposition–NiO film has a
shorter memory time.
©
2009 The Electrochemical Society. ͓DOI: 10.1149/1.3187911͔ All rights reserved.
Manuscript submitted January 14, 2009; revised manuscript received July 3, 2009. Published July 31, 2009.
The electrochromic ͑EC͒ material is capable of varying its opti-
cal ͑transmittance͒ state under an applied potential. When the ap-
plied potential is stopped, its current optical state is kept. Under an
appropriately reverse potential, it reverses its optical state. The
variation in the optical state is caused by the insertion or the extrac-
tion of cations into the EC material from the electrolyte under the
pure boric acid with 500 mL DI water. The deposition temperature
was kept at 40°C. After the NiO film deposition, part of the NiO
film was removed for probing the ITO substrate. Thermal annealing
was used to improve the characteristics of LPD-NiO from 100 to
500°C for 30 min in air. A scanning electron microscope was used
to measure the film thickness.
The strongest intensity of the sun’s radiation was at 550 nm,
which was used as the wavelength to determine the transmittance of
the film in the bleached state and the colored state. The reflecting
spectrograph was used to measure the transmittance. The transpar-
ency ratio is defined as follows
1
applied potential, and it can be repeated many times.
With the negative potential applied onto the indium tin oxide
+
͑
ITO͒ electrode of the EC/ITO structure, Li ions are inserted into
the EC film ͑coloring state͒ via the “zigzag” rows from the electro-
lyte of LiClO and the propylene carbonate solution. Electrons from
4
+
ITO are injected to the EC film to compensate for the Li ion
⌬
T% = T % − T %
͓1͔
charges. With the positive potential applied onto the EC/ITO, the Li+
ions are extracted out of the EC film. The EC film goes into the
b
c
where Tb% and Tc% are the transmittances of the bleached state and
the colored state, and ⌬T% is the transparency ratio. The transmit-
tance was 70% for the ITO/glass at 550 nm.
2
bleaching state.
Many materials have been used as cathodic and anodic EC films.
3
,4
5,6
7,8
9,10
11,12
13
MoO3, TiO2, V O , Nb O ,
WO3,
and Ta O
were
2
5
2
5
2
5
used as cathodic EC films. Tungsten oxide ͑WO ͒ has been widely
3
Results and Discussion
used due to its clearer bleaching state and higher transparency ratio.
14
15
16-18
19,20
Figure 1 shows the deposition thickness of the NiO film as a
function of deposition time for ͑a͒ 30, ͑b͒ 40, and ͑c͒ 50°C. The
delay times of 24, 8, and 0 h are observed for 30, 40, and 50°C,
respectively. The delay time decreases with the deposition tempera-
ture. The delay time depends on the substrate surface, precursor
concentration, and deposition temperature. During LPD, it needs an
incubation period to nucleate before the film growth. The time for
Anodic EC films include IrO2, VO2, CoO,
and NiO.
NiO
has many advantages, such as ͑i͒ higher transparency ratio between
fully bleached and fully colored states, ͑ii͒ lower cost, and ͑iii͒ pos-
21
sible deposition by a wet process. The EC NiO film is transparent
in visible light in the bleaching state and deep brown in the coloring
state.
Many methods were used to deposit NiO films, such as
2
8,29
2
2
23
24
25
nucleus is shorter at higher deposition temperatures.
evaporation, electrodeposition, sputtering, sol–gel, and liquid
26
Due to different delay times, it is inconvenient to derive the
deposition rate at different deposition temperatures. In the linear
film growth region, the linear film growth rates are 1.91, 25.2, and
phase deposition ͑LPD͒. LPD has many advantages, such as low
temperature deposition ͑room temperature is permissible͒, high uni-
formity, good selectivity on substrate, and low cost.
2
7
2
0.2 nm/h at 30, 40, and 50°C, respectively. It indicates that the
Usually, the deposition rate of LPD-NiO is very low ͑50–75 nm
linear film growth rate increases with the deposition temperature.
Because of a very low linear film growth rate and a very long delay
time, the film prepared at 30°C is not good for EC application. The
film quality prepared at 50°C is not good for EC application due to
a very high linear film growth rate. The samples prepared at 40°C
for 11 h are for the scanning electron microscopy ͑SEM͒, X-ray
photoelectron spectroscopy, and EC measurements shown in Fig.
2-6.
in 16–72 h͒ at 30°C using saturated NiF ·4H O as the growth
2
2
2
6
solution. A quite slow deposition rate is impractical for applica-
tions. Due to the growth of an endothermic reaction, the deposition
rate can be increased by increasing the growth temperature. In this
study, the deposition temperature was increased from 30 to 40°C,
and the EC characteristics of the LPD-NiO films were investigated.
Experimental
A very smooth surface of the NiO film can be obtained by the
LPD method, as shown in Fig. 2, in which the thickness of the
LPD-NiO film is about 100 nm. The interface between the LPD-NiO
film and the ITO/glass substrate is very sharp. The transmittance of
the as-deposited NiO/ITO/glass with a thickness of 100 nm is about
78% and is improved to 88% after annealing at 300°C in air at a
wavelength of 550 nm. The transmittances of both films are higher
than that of the ITO/glass. It could be from the decrease in the
surface light reflection by the lower refractive index of NiO than
that of ITO, and they are 1.14 and 1.8 measured by an ellipsometer,
respectively. The amorphous structure and fluorine incorporation
could contribute to the lower refractive index of NiO.
The 50 nm thick ITO-coated glass with a sheet resistance of
0 ⍀ cm was used as the substrate in our experiments. The growth
2
solution was prepared with saturated NiF ·4H O and 0.6 M boric
2
2
acid solutions. The saturated NiF solution was prepared by mixing
2
1
0.5 g of NiF ·4H O with 300 mL deionized ͑DI͒ water, and it was
2 2
stirred steadily for 16 h at room temperature. After observing the
NiF ·4H O precipitate, the saturated NiF solution was obtained.
2
2
2
The 0.6 M boric acid solution was prepared by mixing 18.9 g of
z
E-mail: mklee@mail.ee.nsysu.edu.tw
Downloaded on 2015-01-31 to IP 128.114.34.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D396
Journal of The Electrochemical Society, 156 ͑10͒ D395-D399 ͑2009͒
1
00
o
3
0 C
8
6
4
2
0
0
0
0
0
Figure 2. SEM cross section of annealed LPD-NiO/ITO/glass film.
0
10 20 30 40 50 60 70
Deposition Time (hr)
with and without thermal annealing at 300°C for 30 min in air are
shown in Fig. 3a and b. The corresponding transmittance spectra are
shown in Fig. 3c. In Fig. 3a and b, the colored state is darker and the
bleached state is clearer for the annealed LPD-NiO film. It could be
from the decrease in oxygen vacancy and defect. Obviously, the
transparency ratio is higher for the thermal annealed LPD-NiO film.
As shown in Fig. 3c, ⌬T% is about 23% for the as-deposited LPD-
NiO film and 48% for the annealed LPD-NiO film.
(
a)
1
40
20
00
o
4
0 C
1
There are some EC parameters for the LPD-NiO films, such as
operation voltage, coloring and bleaching time, and memory time.
1
80
60
40
20
0
0
2
4
6
8
10 12
(
i)
(ii)
(iii)
(
a)
Deposition Time (hr)
(
b)
100
o
5
0 C
80
60
40
20
0
(
i)
(ii)
b)
(iii)
(
0
.9
.8
.7
.6
bleached (annealed)
bleached (as-deposited)
0
0
0
0
1
2
3
4
5
0.5
colored (as-deposited)
colored (annealed)
0.4
0.3
0.2
0.1
Deposition Time (hr)
(
c)
Figure 1. Deposition thickness as a function of deposition times: ͑a͒ 72 h at
0.0
1
00 200 300 400 500 600 700 800 900
3
0°C, ͑b͒ 12 h at 40°C, and ͑c͒ 6 h at 50°C.
Wavelength (nm)
(
c)
LiClO with a propylene carbonate solution and a platinum ͑Pt͒
4
Figure 3. ͑Color online͒ ͑a͒ Pictures of ͑i͒ initial, ͑ii͒ 10 cycle colored, and
iii͒ 10 cycle bleached states for 100 nm LPD-NiO films without thermal
annealing. ͑b͒ The same states as ͑a͒ with thermal annealing. ͑c͒ Correspond-
ing transmittance spectra for 10 cycle colored and bleached states in ͑a͒ and
͑b͒.
foil were used as the electrolyte and the counter electrode for the EC
measurement. The completely colored and bleached states of the
LPD-NiO film were characterized by an EC test under Ϯ5 V for 30
s. The colored and bleached pictures of the 100 nm LPD-NiO films
͑
Downloaded on 2015-01-31 to IP 128.114.34.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 156 ͑10͒ D395-D399 ͑2009͒
D397
45
40
35
30
25
20
15
10
5
0
9
0
5
as-deposited
as-deposited
annealed at 300 C
8
o
o
80
75
70
65
60
55
annealed at 300 C
5
0
5
4
0
5
10 15 20 25 30
Coloring Time (sec)
1
2
3
4
5
Operation Voltage (V)
Figure 4. ͑Color online͒ Transparency ratio as a function of operation volt-
age for as-deposited and annealed LPD-NiO films.
(
a)
80
75
70
65
60
55
50
For the as-deposited and the annealed LPD-NiO films, the transpar-
ency ratio is a function of operation voltage, as shown in Fig. 4. The
as-deposited
annealed at 300 C
o
0
5
10 15 20 25 30
Bleaching Time (sec)
(
b)
Figure 6. ͑Color online͒ Transmittance as a function of coloring and bleach-
ing time for as-deposited and annealed LPD-NiO films.
operation voltage for activating EC is about 4 V. Usually, the opera-
tion voltage is requested to be below 5 V. It indicates that the LPD-
NiO is a promising material for EC applications.
Coloring and bleaching time are mainly associated with the film
structure. That is to say, the change in the speed of the states is
+
determined by the insertion or extraction speed of the Li ions into
3
0
the EC material via vacancies under the applied potential. Usually,
the perfect single-crystalline structure has few vacancies, so the de-
gree of coloring and bleaching is smaller. A polycrystalline structure
+
has many dangling bonds on the grain boundary; Li ions are
trapped in films and make coloring and bleaching difficult. Theoreti-
cally, an amorphous structure can provide even more vacancies for
+
the Li ion diffusion. But it has a higher density of dangling bonds
+
to decrease Li ion diffusion and make coloring and bleaching more
difficult. By the LPD method, fluorine ions from NiF ·4H O can
2
2
passivate the defects and the dangling bonds of the amorphous LPD-
NiO film. It benefits the EC. The defects and the dangling bonds can
be further improved by post-thermal annealing and hence enhance
+
the Li ion diffusion and make coloring and bleaching faster. Figure
5
A shows the electron spectroscopy for chemical analysis spectra
for the as-deposited and the annealed NiO films. The peaks at 857
and 862 eV are from Ni 2p3 , and the peaks at 874 and 882 eV are
Figure 5. ͑Color online͒ ͑A͒ X-ray photoelectron spectrum of NiO. ͑B͒
X-ray photoelectron spectrum of fluorine.
/2
Downloaded on 2015-01-31 to IP 128.114.34.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D398
Journal of The Electrochemical Society, 156 ͑10͒ D395-D399 ͑2009͒
were measured by a reflecting spectrograph. Figure 7 shows the
60
55
50
45
40
35
30
25
transmittance as a function of memory time for the fully colored ͑a͒
as-deposited and ͑b͒ the annealed LPD-NiO films prepared at a
deposition time of 8–12 h at 40°C. Figure 7a shows that the trans-
mittance of the thicker sample is lower than that of the thinner
sample. The transmittance of the colored as-deposited NiO film is
mainly determined by the thickness. The as-deposited NiO film pre-
pared for 8 h seems to be inconsistent with this tendency. Because of
this, the film is not thick enough to be ripened into a stable quality,
and there are high concentrations of the defects and the dangling
bonds.
8
9
1
hr
hr
0hr
11hr
Basically, the transmittance of the fully colored NiO film is de-
12hr
+
termined by the Li ion concentration and the film thickness. The
+
concentration of the Li ion in the annealed NiO films is higher than
0
300
Memory Time (s)
a)
600
900
+
that of the as-deposited films due to the faster diffusion of the Li
+
ion in the annealed NiO film. The concentration of the Li ion in the
thinner film is higher due to a higher electrical field under a fixed
voltage during coloring. Based on the Li ion concentration and the
film thickness, the transmittances of the annealed NiO films are
nearly the same, as shown in Fig. 7b.
+
(
7
7
6
6
5
5
4
4
3
5
0
5
0
5
0
5
0
5
For the as-deposited and the annealed LPD-NiO films, the trans-
mittances are in the ranges of 45–57.5% and 57.5–65% at 600 s,
respectively. In other words, the annealed LPD-NiO film has a
shorter memory time, which indicates that the trapping for the Li
ions is weaker due to lower oxygen vacancy and better F ion passi-
vation.
+
8hr
9hr
Conclusion
The as-deposited LPD-NiO film is uniform. The linear film
growth rate of the LPD-NiO film can be controlled at about 25.2
nm/h. The transmittance of the annealed LPD-NiO film is higher
than that of the as-deposited film. For the annealed LPD-NiO film,
the transparency ratio can reach about 48% at a wavelength of 550
nm under the operation voltage of 4 V. The higher coloring and
bleaching rates are from the higher quality LPD-NiO films, in which
F ions play an important role.
1
1
1
0hr
1hr
2hr
0
200
Memory Time (sec)
b)
400
600
800 1000
Acknowledgments
(
This research was supported by the National Science Council of
Taiwan under grant NSC93-ET-7-110-002-ET.
Figure 7. ͑Color online͒ Transmittance as a function of memory time for ͑a͒
as-deposited and ͑b͒ annealed LPD-NiO films prepared at deposition times
from 8 to 12 h.
National Sun Yat-sen University assisted in meeting the publication costs
of this article.
References
1
2
3
. M. Gómez, A. Medina, and W. Estrada, Sol. Energy Mater. Sol. Cells, 64, 297
͑2000͒.
. N. N. Dinh, N. Th. T. Oanh, P. D. Long, M. C. Bernard, and A. H. L. Goff, Thin
Solid Films, 423, 70 ͑2003͒.
. Y. G. Wu, G. M. Wu, X. Y. Ni, and X. Wu, Sol. Energy Mater. Sol. Cells, 63, 217
͑2000͒.
3
1
from Ni 2p1 , which belong to NiO. The peaks of the annealed
/2
NiO are stronger than that of the as-deposited NiO, which is from
the decrease in oxygen vacancy after thermal annealing in air. Fig-
ure 5B shows a peak at about 685.1 eV from Ni–F that belongs to
3
2
33
NiF2
and a shoulder at about 684.4 eV from Ni–F
for the
4. D. J. De Smet and J. L. Ord, J. Electrochem. Soc., 134, 1734 ͑1987͒.
5. J. Scarminio, A. Lourenco, and A. Gorenstein, Thin Solid Films, 302, 66 ͑1997͒.
6. Y. Ye, J. Zhang, P. Gu, X. Liu, and J. Tang, Thin Solid Films, 298, 197 ͑1997͒.
7. A. Gutarra, A. Azens, B. Stjerna, and C. G. Granqvist, Appl. Phys. Lett., 64, 1604
͑1994͒.
annealed Ni–O film. A peak at 685.7 eV and a shoulder at 684.8 eV
for the as-deposited film are also shown in Fig. 5B. Compared with
the annealed Ni–O film, the blueshifts come from the higher F con-
centration of the as-deposited film. For the annealed film, the peak
intensity of Ni–F is slightly lower, and the peak intensity of Ni–F is
34
8. K. Nagase, Y. Shimizu, N. Miura, and N. Yamazoe, Appl. Phys. Lett., 64, 1059
1994͒.
. A. Talledo, A. M. Anderson, and C. G. Granqvist, J. Mater. Res., 5, 1253 ͑1990͒.
0. T. Maruyama and T. Kanagawa, J. Electrochem. Soc., 141, 2868 ͑1994͒.
͑
2
9
higher than that of the as-deposited film. It supports the fact that
1
Ni–F can be decomposed and transformed into Ni–F at the anneal-
ing temperature. The stronger Ni–F bonds indicate a better F ion
passivation on the annealed NiO film.
11. K. Yoshimura, T. Miki, P. Jin, and S. Tanemura, Reports of the National Industrial
Research Institute of Nagoya, National Industrial Research Institute of Nagoya,
Nagoya, Japan ͑1996͒.
2
1
1
2. S. K. Deb, Trans. Faraday Soc., 65, 3187 ͑1969͒.
3. S. K. Deb, Philos. Mag., 27, 801 ͑1973͒.
For the as-deposited and the annealed LPD-NiO films, the trans-
mittance is a function of coloring and bleaching time, as shown in
Fig. 6. The coloring speed of the as-deposited film is faster than that
14. Z.-W. Fu and Q.-Z. Qin, Electrochem. Solid-State Lett., 2, 600 ͑1999͒.
5. G. Beni, C. E. Rice, and J. L. Shay, J. Electrochem. Soc., 127, 1342 ͑1980͒.
16. J. L. Ord, S. D. Bishop, and D. J. De Smet, J. Electrochem. Soc., 138, 208 ͑1991͒.
7. P. S. Patil, L. D. Kadam, and C. D. Lokhande, Sol. Energy Mater. Sol. Cells, 53,
29 ͑1998͒.
18. P. M. S. Monk and N. Farooq, J. Mater. Sci.: Mater. Electron., 6, 389 ͑1995͒.
1
+
of the annealed film because the Li ions diffuse faster by a higher
1
concentration of oxygen vacancy. But the bleaching speed of the
annealed film is faster than that of the as-deposited film because the
lower concentration of traps is due to the decrease in oxygen va-
cancy and a better F ion passivation.
By the memory time test, the samples were fully colored and
were kept in air for 0, 300, 600, and 900 s, and the transmittances
2
1
9. T. Maruyama and S. Arai, J. Electrochem. Soc., 143, 1383 ͑1996͒.
2
0. J. S. E. M. Svensson and C. G. Granqvist, Sol. Energy Mater. Sol. Cells, 16, 19
͑1987͒.
2
1. J. S. E. M. Svensson and C. G. Granqvist, Appl. Phys. Lett., 49, 1566 ͑1986͒.
22. W. Estrada, A. M. Anderson, and C. G. Granqvist, J. Appl. Phys., 64, 3678 ͑1988͒.
Downloaded on 2015-01-31 to IP 128.114.34.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 156 ͑10͒ D395-D399 ͑2009͒
D399
2
2
3. T. Seike and J. Nagai, Sol. Energy Mater. Sol. Cells, 22, 107 ͑1991͒.
4. C. Natarajan, H. Matsumoto, and G. Nogami, J. Electrochem. Soc., 144, 121
29. M. P. Houng, C. J. Huang, Y. H. Wang, and W. J. Chang, J. Appl. Phys., 82, 5788
͑1997͒.
͑
1997͒.
5. S. R. Jiang, P. X. Yan, B. X. Feng, X. M. Cai, and J. Wang, Mater. Chem. Phys.,
7, 384 ͑2003͒.
30. M. K. Lee, C. L. Ho, and J. Y. Zeng, Phys. Status Solidi A, 11, 2679 ͑2008͒.
31. C. P. Li, A. Proctor, and D. M. Hercules, Appl. Spectrosc., 38, 880 ͑1984͒.
32. S. W. Gaarenstroom and N. Winograd, J. Chem. Phys., 67, 3500 ͑1977͒.
33. M. Y. Wang, C. W. Chang, C. M. Wu, C. T. Lin, C. H. Hsieh, W. S. Shue, and M.
S. Liang, Dig. Tech. Pap. - Symp. VLSI Technol., 2003, 157.
2
7
2
2
2
6. Z. Jiao, M. Wu, Z. Qin, and H. Xu, Nanotechnology, 14, 458 ͑2003͒.
7. T. J. Richardson and M. D. Rubin, Electrochim. Acta, 46, 2119 ͑2001͒.
8. J. H. Wei and S. C. Lee, J. Electrochem. Soc., 144, 1870 ͑1997͒.
34. J. C. Yu, J. Yu, W. Ho, Z. Jiang, and L. Zhang, Chem. Mater., 14, 3808 ͑2002͒.
Downloaded on 2015-01-31 to IP 128.114.34.22 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).