Journal of Electroanalytical Chemistry 661 (2011) 106–112

Contents lists available at SciVerse ScienceDirect

Journal of Electroanalytical Chemistry

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

Enhanced photoelectrochemical activity of Ce doped ZnO nanocomposite thin

films under visible light
M. Yousefi a, M. Amiri b, R. Azimirad c, A.Z. Moshfegh a,d,⇑
a
Department of Physics, Sharif University of Technology, P.O. Box 11155-9161, Tehran, Iran
b
Department of Chemistry, Payame-Noor University, P.O. Box 55135-469, Ardabil, Iran
c
Malek-Ashtar University of Technology, Tehran, Iran
d
Institute for Nanoscience and Nanotechnology, Sharif University of Technology, P.O. Box 14588-89694, Tehran, Iran

a r t i c l e i n f o a b s t r a c t

Article history: Ce-doped ZnO and pure ZnO nanocomposite thin films with different Ce/Zn ratios (0, 2, 5, 10, 15, 20, and
Received 21 November 2010 30 at.%) have been prepared by sol–gel method at optimum annealing temperature of 500 °C. The synthe-
Received in revised form 15 June 2011 sized samples were characterized by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and
Accepted 18 July 2011
UV–vis spectrophotometry. According to our XPS data analysis, there are three major metal ions namely
Available online 12 August 2011
Ce3+, Ce4+ and Zn2+ that coexist on the surface. The XRD measurements indicate that the ZnO thin films
have a hexagonal wurtzite structure, and CeO2 crystallites formed in the Ce-doped ZnO nanocomposite
Keywords:
thin films. Photoelectrochemical property of the samples was studied by three electrode galvanostat/
Ce-doped ZnO
Sol–gel
potentiostat system. The addition of Ce to ZnO thin film increased photocurrent density to about double
Nanocomposite amount at applied potential of 0.6 V as compared to undoped ZnO film in photoelectrochemical water
Photoelectrochemical activity splitting reaction. It was also observed that adding cerium to ZnO nanocomposite thin films resulted
in enhancement of the photoresponsivity of the layers. Among different Ce/Zn ratios examined, the opti-
mum doping concentration at 10 at.% demonstrated the highest photocurrent density as compared to
other investigated ratios under similar experimental conditions.
Ó 2011 Elsevier B.V. All rights reserved.

1. Introduction electrode, solar cell windows, photovoltaic devices, catalysts, gas

sensors, room temperature ultraviolet lasers and photoelectro-
Research on photoelectrochemical (PEC) cells, specially their chemical cells [9,11–13].
application in solar energy conversion, has been gained special Several thin film deposition techniques have been utilized to
attention in last few decades, particularly in recent years [1–5]. produce pure and doped ZnO films, including sputtering [14],
PEC cells convert solar energy into storable chemical energy as pulsed laser deposition [15], vapor–liquid–solid (VLS) [16] and
hydrogen through the photoelectrolysis of water. Several semicon- the sol–gel process [17]. Among these methods, sol–gel process
ductors such as TiO2, SrTiO3, SnO2 and BaTiO3, investigated as pos- has several advantages due to its simplicity, easy control of the film
sible photo anodes, are relatively wide band gap materials and deposition, safety as well as low cost of the apparatus and raw
generally transparent to visible light. However, narrow band gap materials [17].
materials like GaAs, Fe2O3 and CdS get easily corroded in contact Doping is an effective way to improve the properties of ZnO thin
with the electrolyte. ZnO, despite its wide band gap, is a suitable films. Therefore, adding selective elements to ZnO offers an impor-
material for PEC water splitting reaction due to its excellent tant route to enhance and control its optical and electrical proper-
electrochemical stability [1,6–8]. ties, which is crucial to its practical applications [18–20]. Several
ZnO is a high and direct band gap (3.37 eV) semiconducting elements have been added to ZnO for improving its photoelectro-
material with a large exciton binding energy (60 meV) [9,10]. It chemical property [12,21,22]. For example, the doping of ZnO with
has been extensively studied in the last few years owing to its nonmetal (C, N, etc.) could narrow its band gap and might drive the
potential applications in various fields. Some of its important response to visible light [12], whereas the doping of metal could
usages include surface acoustic wave devices (SAW), transparent trap temporarily the photogenerated charge carriers. It can sup-
press the recombination of photo induced electron-hole pairs
⇑ Corresponding author at: Department of Physics, Sharif University of Technol-
[23]. Among different rare earth metals, the cerium doping has re-
ogy, P.O. Box 11155-9161, Tehran, Iran. Tel.: +98 21 6616 4516; fax: +98 21 6601 ceived much attention due to its important properties including:
2983. (1) the redox couple Ce3+/Ce4+ that makes cerium oxide shift
E-mail address: moshfegh@sharif.edu (A.Z. Moshfegh). between CeO2 and Ce2O3 under oxidizing and reducing conditions

1572-6657/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jelechem.2011.07.022
 M. Yousefi et al. / Journal of Electroanalytical Chemistry 661 (2011) 106–112 107

and (2) the easy formation of labile oxygen vacancies with the determine phase formation, crystal structure of the layers and
relatively high mobility of bulk oxygen species [24]. Therefore, average crystalline size. X-ray photoelectron spectroscopy (XPS)
considering these advantages, it is expected to observe an equipped with an Al Ka X-ray source at energy of 1486.6 eV was
enhancement in photocurrent density in cerium doped ZnO thin employed to study surface chemical composition of the films.
films during a photoelectrochemical reaction. Thus, cerium can The hemispherical analyzer (Specs EA 10 Plus) operating in a vac-
modify photoresponsivity of the Ce–ZnO nanocomposite thin film. uum better than 107 Pa was used for binding energy analysis. All
Extensive literature search show no reported work on photo- binding energy values were calibrated by fixing the C (1 s) core le-
electrochemical (PEC) activity of Ce–ZnO system. As seen in Table vel line to the 285.0 eV as a reference energy point. All of the peaks
1, to best of our knowledge, this is the first report on PEC study of were deconvoluted using SDP software (version 4.1) with 80%
Ce doped ZnO system ever investigated. In this work, we have Gaussian-20% Lorentzian peak fitting.
synthesized Ce–ZnO nanocomposite thin films by using sol–gel For investigating the photoelectrochemical properties of all the
dip-coating method. Additionally, the influence of various Ce prepared samples, a three-electrode galvanostat/potentiostat sys-
concentrations on the photoresponsivity of ZnO thin films under tem (Autolab PGSTAT302) was used. The measurements were car-
visible light has been also investigated in different experimental ried out in Na2SO4 aqueous solution as cathodic and anodic
conditions. electrolyte. On one side of our cell, there was a quartz window
for mounting the photoanode (the prepared Ce–ZnO nanocompos-
ite thin films or pure ZnO) behind it and then illuminating the light.
2. Experimental details The counter electrode was a platinum wire and Ag/AgCl was used
as reference electrode. Pure ZnO and Ce doped ZnO thin films
In this experiment, zinc acetate dehydrate (Zn(CH3COO)2.H2O, deposited on the ITO coated glass substrates were used as working
Merck, >99.5%) and cerium nitrate hexahydrate (Ce(NO3)3.6H2O, electrodes with equal surface area (1.5 cm  2 cm). Xenon short
Merck, >98.5%) are used as starting material and dopant sources, arc lamp (OSRAM 5000 W HBM/OFR) was used as irradiation
respectively. Moreover, ethanol (C2H5OH, Merck, >99.9%) and source. To perform the measurements, a glassy IR filter has been
monoethanolamine (MEA) (C2H7NO, Merck, >99.8%) are used as a put in front of the lamp. So, wavelengths (k < 360 nm) was cut from
solvent and stabilizer, respectively. In a typical preparation proce- the lamp spectrum insuring only visible region. Moreover, by
dure, for Ce-doped ZnO thin films, Zinc acetate dihydrate is first changing the input power, the intensity of the lamp was adjusted
dissolved in a mixture which was composed of ethanol and MEA at AM 1.5 (0.1 W/cm2) in the visible range.
at 60 °C. The molar ratio of MEA to zinc acetate dihydrate main-
tained at 1.0 and the concentration of zinc acetate was 0.1 mol/L.
The mixed solution stirred at 60 °C for 1 h and then ethanolic solu- 3. Results and discussion
tion of cerium nitrate hexahydrate with different Ce/Zn ratios (2, 5,
10, 15, 20 and 30 at.%) was added to it and the final solution was 3.1. Optical measurements
stirred for additional 1 h. After stirring, the solution became clear
and homogenous which was used as the coating solution after The optical transmittance spectra of ZnO and Ce-doped ZnO
cooling to room temperature and allowed to age for 24 h before thin films in the UV–visible wavelength (300–800 nm) are pre-
initiating deposition. sented in Fig. 1. As can be seen, the transmittance of pure ZnO
The films are obtained by a dip-coating technique. The deposi- and 5 as well as 10 at.% Ce–ZnO nanocomposite thin films were
tion process was performed by dipping cleaned microscope slide measured all about 97% in the visible range from 400 to 800 nm.
glass and cleaned indium tin oxide (ITO) coated glass (commercial The same results which are not shown here were also obtained
ITO 1 lm thickness with electrical sheet resistance 30 X/h) for 2, 15 and 20 at.% Ce–ZnO nanocomposite thin films. But, for
into the solution for 30 s and pulling them out at a rate of the 30 at.% Ce–ZnO sample, it decreased to about 90% in the visible
1.5 mm/s. After coating step, the films were dried at 180 °C for region. The reduction in transmittance of the Ce doped ZnO thin
30 min. Then, they finally were annealed at 500 °C in air for 1 h. films can be due to the observed morphology for the corresponding
A UV–visible spectrophotometer (Jasco-V530) was used to sample. According to our AFM analysis (not shown here), there is a
investigate the optical properties of the films (with eliminating difference in the measured root mean square (rms) surface rough-
the substrate effect) in wavelength ranging from 300 to 800 nm ness values between samples. We have determined the rms values
with 1 nm resolution. Philips PW 3710 profile X-ray diffractometry of 0.64, 0.52, and 2.77 nm for 5, 10, and 30 at.% Ce doped ZnO sam-
(XRD) with Cu Ka radiation source and step size 0.05° was used to ples, respectively. As can be seen, the sample with the highest Ce-
doping (Ce (30 at.%))-has the largest surface roughness and as a re-
sult it has the highest scattering and therefore the lowest transmit-
Table 1
tance. Moreover, the RMS surface roughness of the 10% Ce doped
Various Ce/ZnO systems and their applications.
ZnO sample is lower than the 5%Ce-doped sample. Thus, the higher
System Growth method Application Reference measured transmittance is due to lower scattering of this sample. A
CeO2–ZnO Co-evaporation Gas-sensing [20] similar result was also obtained for MOCVD grown CeO2–ZnO thin
Ce-doped ZnO Sol–gel Photoluminescence [25] films reported in reference [50].
nanorods
In order to determine the dependence of the band gap energy
CeO2–ZnO Sol–gel Gas-sensing [19]
Ce-doped ZnO Sol–gel/CVD Photoelectric [26] on the amount of cerium in thin films, the position of the transmit-
anobrushes tance’s first derivative peak is measured. It is expected that the
Ce-doped ZnO Homogeneous Catalysis [27] transmittance’s first derivative diverges at kg = hc/Eg, where kg is
nanoparticle precipitation the position of maximum peak for the pure ZnO and doped Ce–
Ce-doped ZnO Co-precipitation Photoluminescence [28]
ZnO thin films. The position of the peak being a measurement of
Ce–ZnO Solid-state Photoluminescence [29]
nanoparticle sintering method the band gap energy, Eg [32,33]. Considering Fig. 2, the first deriv-
Ce4+ doped ZnO Soft chemical Ferromagnetism [30] ative of transmittance with respect to wavelength (dT/dk), deter-
nanorods method mines the band gap energy for the pure ZnO and Ce–ZnO
Ce–ZnO Spray pyrolysis Photoluminescence [31]
nanocomposite thin films. According to our optical data analysis,
Ce–ZnO Sol–gel Photoelectrochemical This work
the band gap energy were measured at 3.349, 3.376 and 3.643 eV
108 M. Yousefi et al. / Journal of Electroanalytical Chemistry 661 (2011) 106–112

100 3.2. XRD analysis

To study the crystalline structure and estimate the size of the

80
Ce–ZnO particle, XRD analysis was used. Due to low thickness of
Transmittance (%)

pure ZnO
the films about 16 nm (obtained by both RBS and optical analysis),
60 XRD measurement on the powder samples was performed. Fig. 4
Ce(30 at.%)-ZnO
shows the typical XRD pattern of pure ZnO and Ce–ZnO powders.
40 The major peaks are identified as (1 0 0), (0 0 2), (1 0 1), (1 0 2),
Ce(5 at.%)-ZnO
(1 1 0), (1 0 3), (2 0 0), (1 1 2), and (2 0 1) plane of reflections for
Ce(10 at.%)-ZnO a single phase wurzite structure of ZnO and the peak located at
20 2h = 28.7° is attributed to (1 1 1) plane for the CeO2 crystalline
structure [9,19]. Therefore, the diffraction peaks of pure ZnO and
0 Ce–ZnO, without characteristic peaks of other impurities, are in-
300 400 500 600 700 800 dexed to the ZnO hexagonal wurtzite structure according based
Wavelength (nm) on its JCPDS standard card (No. 75-0576), and the CeO2 cubic struc-
ture (JCPDS standard card No. 75-0390).
Fig. 1. Transmittance spectra of Ce–ZnO nanocomposite thin films for different The average crystalline size (D) was also estimated by using
cerium concentrations.
scherrer’s formula, D = 0.9k/(b cos h), based on the integral width
of XRD peaks, where k is the wavelength of the incident X-ray radi-
ation, b is the full width at half-maximum (FWHM) of the desired
peak (here (1 0 1) at 2h = 36.3°) in radians and h is the Bragg angle
(c) [37]. The crystalline structure and averaged crystal size for differ-
ent Ce doped ZnO system was summarized in Table 2. It was found
that by increasing cerium concentration in the films, the average
(b)
crystalline size decreased. This is because the ionic radii of Ce3+
and Ce4+ are 0.103 and 0.092 nm, respectively which are larger
than that of Zn2+ ion of 0.074 nm reported by the others [25,38–
40]. Therefore, it is difficult for Ce3+ and Ce4+ to enter the crystal-
(a)
line lattice of ZnO to substitute Zn2+. The Ce3+ and Ce4+ ions seem
more likely to form complex with the surface oxygen of ZnO, which
decreases the growth of ZnO crystallite [25,38,41]. Besides, the

Wavelength (nm)
3
Fig. 2. First derivative (dT/d(k)) plot of Transmittance as a function of wavelength Pure ZnO
for different samples: of (a) pure ZnO, (b) Ce(10 at.%)–ZnO and (c) Ce(30 at.%)–ZnO Ce(2 at.%)-ZnO
(101)

thin films. 2.5 Ce(5 at.%)-ZnO

Ce(10 at.%)-ZnO
(100)
CeO2(111)

(002)

(110)
for the pure ZnO as well as 10 and 30 at.%, respectively. In addition,
(102)

(103)
2

(112)
Intensity ( a.u.)

the band gap energy for the 2, 5, 15 and 20 at.% Ce-doped ZnO

(201)
(d)

(200)

(004)
nanocomposite thin films were determined at 3.349, 3.357, 3.405
1.5
and 3.421 eV, respectively. Based on XPS analysis, which will be
discussed in Section 3.3, the synthesized thin films contain CeO2. (c)
But, the band gap of this oxide (3.6–3.8 eV) is higher than pure 1
ZnO (3.37 eV). Thus, the band gap energy of Ce–ZnO film was mea- (b)
sured greater than pure ZnO as also reported by others [34,35]. The
0.5
variation of band gap energy with cerium concentration is shown
in Fig. 3. A similar result was also observed for Mn doped ZnO thin (a)
films recently [36,34]. 0
15 35 55 75

3.7
Fig. 4. XRD patterns of the ZnO-based powders with different Ce concentrations:
(a) Zero, (b) 2 at.%, (c) 5 at.% and (d) 10 at.%.
Band gap Energy (eV)

3.6

3.5 Table 2
Crystalline properties of pure ZnO and different Ce–ZnO thin films.

Sample Crystalline structure I111/(I101 + I111) Average crystal

3.4
size (nm)
Pure ZnO Hexagonal ZnO Zero 33.0
3.3 2 at.% Ce:ZnO Hexagonal ZnO + 0.045 28.6
0 10 20 30 cubic CeO2
5 at.% Ce:ZnO Hexagonal ZnO + 0.086 27.8
Cerium content (at.%) cubic CeO2
10 at.% Ce:ZnO HexagonalZnO + 0.142 26.3
Fig. 3. Variation of band gap energy with cerium concentrations for different cubic CeO2
Ce–ZnO nanocomposite thin films.
 M. Yousefi et al. / Journal of Electroanalytical Chemistry 661 (2011) 106–112 109

lower peak angle shift of the Ce–ZnO powders was also observed as component (A) with a binding energy of 1022.6 eV corresponds to
compared with the pure ZnO powder. As explained earlier, the dif- Zn–O bond in ZnO lattice and the higher-energy component (B)
ference between cerium and zinc’s ions radii resulted in the ob- with a binding energy of 1023.5 eV is attributed to Zn–OH bond
served shift. When Ce4+ and Ce3+ incorporated into ZnO, the as also reported for the sol–gel deposited ZnO films by other
lattice parameters will slightly increased, leading to a shift of researchers [43].
ZnO main peaks namely (1 0 0), (0 0 2) and (1 0 1) to lower angles The measured Ce(3d) core level binding energy (Fig. 5b) clearly
as also observed by other investigators [25,42]. shows 10 deconvoluted peaks. The complexity of the XPS spectra of
the Ce(3d), which were resolved by Burroughs et al. [44], comes
from the hybridization between Ce(4f) levels and the O(2p) states.
3.3. XPS analysis
The two sets of spin–orbital multiplets, corresponding to the
Ce(3d3/2) and Ce(3d5/2) contributions, are labeled u and v, respec-
In order to determine stoichiometry, purity and surface chemi-
tively [24,45]. The peaks labeled v and v00 have been attributed to
cal state of samples, pure ZnO and 10 at.% Ce–ZnO nanocomposite
a mixing of Ce(3d9 4f2)O(2p4) and Ce (3d9 4f1)O(2p5) Ce(IV) final
thin film were studied by XPS technique. According to the survey
states, and the peak denoted as v000 corresponds to the Ce
spectra (not shown here), only zinc, oxygen, cerium and some trace
(3d9 df0)O(2p6) Ce( IV) final state. On the other hand, v0 and v0
of carbon (due to air contamination and CO/CO2 adsorption on the
are assigned to the Ce(3d9 4f2)O(2p5) and the Ce(3d9 4f1)O(2p6)
surface) were detected on the films with no other surface impuri-
of the Ce(III) state. The same assignation can be applied to the u
ties. Fig. 5 shows high resolution XPS spectra for the Zn(2p), Ce(3d),
structures, which correspond to the Ce(3d3/2) levels. Additionally,
and O(1s) core levels.
many other peaks were also seen in the Ce(3d) spectra, indicating
The Zn(2p3/2) core level spectrum is shown in Fig. 5a. The
coexistence of Ce3+ and Ce4+ as also reported by others[24]. How-
Zn(2p3/2) peak was deconvoluted into two peaks. The lower-energy
ever, the diffraction peaks for the compounds containing Ce3+ were
not observed in the Fig. 4. This probably denotes that the Ce3+com-
pound was either amorphous or had a very low concentration in
the films [24].
The O(1s) spectrum (Fig. 5c) was also deconvoluted into three
peaks. The lower-energy component (A) with a binding energy at
529.3 eV corresponds to the Ce–O bond in the CeO2 compound as
also measured earlier [24,20]. The medium-energy component
(B) with a binding energy at 530.8 eV related to the Zn–O bond
in ZnO film. The high-energy component (C) with a binding energy
at 532.3 eV attributed to oxygen species chemisorbed at the sur-
face [43]. The contribution of each oxygen component is obtained
by dividing the area of that particular peak by the total area of the
O(1s) spectrum. The weight percent of component A, B and C were
obtained about 12%, 50% and 38%, respectively. Thus, the oxygen in
the films is mainly present in the ZnO crystal.
The O(1s) spectrum of the pure ZnO thin film annealed at 500 °C
is illustrated in Fig. 6. As can be seen, the O(1s) peak of the pure
ZnO film only is convoluted from Zn–O (B) and chemisorbed oxy-
gen (C) bonds, and there is not any peak corresponding to Ce–O
bond. Therefore, comparison of O(1s) peaks for the pure ZnO and
10 at.% Ce–ZnO thin films (Fig. 5c) is indicating the existence of
Ce4+ ion in the Ce–ZnO sample as also reported in [20].

3.4. Photoelectrochemical Performance

To investigate the photoelectrochemical activity of the

synthesized films, all experiments have been performed in the

Fig. 5. XPS spectra of the Ce(10 at.%)–ZnO nanocomposite thin film for different
core levels: (a) Zn(2p3/2), (b) Ce(3d) and (c) O(1s). Fig. 6. The deconvoluted O(1s) core level of the pure ZnO film.
110 M. Yousefi et al. / Journal of Electroanalytical Chemistry 661 (2011) 106–112

compartment cell with 0.5 M Na2SO4 aqueous solution as both 2

cathodic and anodic electrolytes. Fig. 7 shows the current–poten-
(a) Ce(20 at.%)-ZnO

Current density (µA)/mm2

tial (I–V) curves of the zinc oxide film annealed at 500 °C under 1.6 Ce(15 at.%)-ZnO
dark and illumination conditions. It indicates oxidation of water Ce(2 at.%)-ZnO
and generated that photocurrent increased strongly by illuminat- 1.2 Ce(5 at.%)-ZnO
ing the film with the visible light intensity at 0.1 W/cm2. It reveals
Pure ZnO
photoelectrochemical activity of the synthesized ZnO film under
0.8 Ce(10 at.%)-ZnO
light illumination as compared with the dark condition clearly.
The electron–hole pairs are generated by the absorption of Ce(30 at.%)-ZnO
photons and separated by the electric field in the depletion region. 0.4
In n-type semiconductors such as ZnO, the excited electrons move
through bulk region to the counter electrode where water reduc- 0
-0.6 -0.3 0 0.3 0.6 0.9 1.2
tion occurs. The generated holes move towards semiconductor/
Voltage (V)
electrolyte interface where water oxidation takes place. -0.4

3.4.1. The effect of Ce doping -0.8

To study the influence of Ce concentration as dopant on photo-
electrochemical performance of ZnO, several samples containing (b) 0.2
different amounts of cerium were prepared and examined.

Current density (µA)/mm2

Fig. 8a exhibits the I–V characteristics for the annealed samples
containing 2, 5, 10, 15, 20 and 30 at.% Ce concentrations as com-
pared with the pure ZnO thin films. It is obvious that the photocur-
rent increased by increasing the Ce content from 0 to 10% in the 0.1
layers. But, further addition of Ce in the ZnO film resulted in signif-
icant reduction of the photocurrent. Therefore, among all the films
with different Ce contents, the sample containing 10 at.% cerium
demonstrated the highest photoelectrochemical performance.
The photocurrent obtained in the chopped I–t curves at con-
stant potential of 0.6 V vs Ag/AgCl has been shown in Fig. 8b. A 0
small amount of current reduction at first few seconds of each illu- Time (s)
mination intervals is due to the charge carrier recombination be-
fore their participation in redox reactions as also observed by Fig. 8. Photocurrent density versus (a) applied voltage and (b) chopping time (on/
off light), for the ZnO films containing various Ce concentrations annealed at 500 °C
other researches [46,47]. Moreover, similar to Fig. 8b, the highest
under 0.1 W/cm2 visible photoillumination (0.5 M Na2SO4 electrolyte).
value of photocurrent is also achieved for the ZnO thin films con-
taining the optimum value (10 at.% Ce).
able for Ce4+ to replace Zn2+ partially in the ZnO crystallites. Thus
3.4.2. Kinetics Ce4+ may substitute the Zn2+ ions according to the following defect
To elucidate the role of Ce doping in ZnO film different samples reaction [19]:
were examined under various conditions. It is known that the Ce
ZnO
addition to a photosensitive semiconductor such as ZnO, may act CeO2 ! Ceoo x
Zn þ Oo þ 1=2O2 þ 2e
0
ð1Þ
in several ways. Firstly, in respect to the size of ionic radius
where Ceoo
is the substitution defect of substituting Ce for Zn in 4+ 2+
(Zn2+ = 0.074 nm, Ce4+ = 0.092 nm) into account, it seems reason- Zn
the ZnO lattice, and Oxo is the interstitial defect of oxygen ions. The
easy formation of labile oxygen vacancies with the relatively high
mobility of bulk oxygen species can improve photoelectrochemical
properties of thin films by increasing electrons and holes [19]
Secondly, the redox couple Ce3+/Ce4+ makes cerium oxide shift
between CeO2 and Ce2O3 under oxidizing and reducing conditions.
Ce4+ doped could act as an electron trap to enhance the electron–
hole pair separation because of its varied valences and special 4f
levels as illustrated in the following reactions [51]:

Ce4þ þ e ! Ce3þ ð2Þ

Ce3þ þ O2 ! O2 þ Ce4þ ð3Þ

To understand the kinetics of carrier recombination process, it
is necessary to study the transient time during the reaction. It is
found that the process of recombination is controlled by the fol-
lowing kinetic equation:

D ¼ et=s ð4Þ
where D defines as
D ¼ ðIðtÞ  Iðf ÞÞ=ðIðiÞ  Iðf ÞÞ ð5Þ
Fig. 7. I–V characteristics of pure ZnO thin film under different conditions: (a)
under dark condition as a reference point and (b) under 0.1 W/cm2 visible range here, t denotes the time, s is a parameter called transient time and I
photoillumination (Electrolyte: 0.5 M Na2SO4). is the photocurrent, indices i and f represent initial and final
 M. Yousefi et al. / Journal of Electroanalytical Chemistry 661 (2011) 106–112 111

(steady) states [48]. Fig. 9 shows the plots of ln (D) as a function of Considering the optimum Ce concentration in ZnO thin films, it
time for the pure as well as 10 and 30 at.% Ce–ZnO films. Using is important to note that further addition of Ce4+ doping resulted in
these linear plots, the transient time was evaluated about 9.8, reduction of photocurrents, that leads to decrease in the rate of PEC
12.0, 14.7, 20.0, 16.1, 15.3 and 7.1 s for the films containing 0, 2, reaction (see Fig. 8b). This is because these sites could become the
5, 10, 15, 20 and 30 at.% Ce, respectively. According to the measured recombination center for the electron–hole pairs. Moreover, the
transient time, further addition of Ce higher the optimum value agglomeration of ZnO caused by additional Ce4+ doping higher than
(10 at.%) led to defect scattering and/or recombination in the films 10 at.% (observed by AFM analysis which is not shown here) re-
which caused a negative effect on the charge separation efficiency. sulted in the further reduction of the PEC efficiency. Furthermore,
Similar to this result has been also reported for the Au doped TiO2 according to our optical data analysis (Figs. 1 and 2), the existence
very recently [46]. of CeO2 in the Ce–ZnO nanocomposite thin films which was also
These reactions can improve the photoelectrochemical and pho- verified by XPS analysis resulted in the increasing of its band gap
tocatalytic activity of semiconductor in presence of cerium as also energy. Therefore, this rising in Eg leads to the reduction in the
reported very recently [24,49]. PEC efficiency of the cerium doped nanocomposite thin films hav-
At last, cerium doping up to 10 at.% increases PEC reaction rate ing higher amount of Ce concentration. Based on all data analysis,
due to increase in the surface area of ZnO by decreasing the crystal it was found that the Ce(10 at.%)-ZnO sample was our optimum
size as also estimated by our XRD analysis (see Table 2). Addition- sample.
ally, the OH hydroxyl on ZnO surface may prevent the recombina-
tion of electrons and holes [24].
An interesting observation in these experiments is the effect of 4. Conclusions
Ce doping on the zero current potential (ZCP) of the ZnO electrode.
At this potential, the driving force required to transport the elec- This study presents data on preparation and characterization of
trons across the particulate film reduced and all the photogenerat- pure ZnO and Ce–ZnO nanocomposite thin films with different
ed charges are lost in the recombination process. For a single- amount of cerium. The influence of cerium dopant on photoelect-
crystal semiconductor electrode, such a ZCP exhibits the flat band rochemical properties of the synthesized ZnO samples has been
potential [47]. However, for a nanostructured semiconductor film investigated using various analytical techniques for the first time.
electrode, this value can be regarded as an apparent flat band po- According to XRD analysis, the crystalline structure of the pure
tential since the kinetic factors control the net flow of electrons synthesized ZnO and Ce–ZnO powders is hexagonal wurtzite.
[43]. Fig. 10 shows that the zero current potential of ZnO shifts Moreover, the cubic CeO2 phase is also identified in the cerium
from 0.1556 to 0.298 V upon modification with 10 at.% Cerium. doped ZnO. The XPS analysis indicated that the cerium doped
This negative shift in the ZCP value describes the ability of the Ce– ZnO thin films contain mainly Ce3+, Ce4+ and Zn2+ ions on their sur-
ZnO nanocomposite films to prevent the charge recombination by faces. Photoelectrochemical properties of the films were investi-
facilitating hole transfer to the redox couple in the electrolyte [47]. gated under 0.1 W/cm2 photo-irradiation as compared to dark
condition. It was also found that the photoresponsivity of ZnO thin
films increased by doping cerium up to optimum concentration (Ce
Time (s)
10 at.%) due to its action as an electron trap in the PEC reaction.
0 3 6 9 12 15 18
0
Acknowledgements

The authors would like to thank Research and Technology

-1 Council of Sharif University of Technology for financial support.
The assistant of Mr. S. Rafiee for XPS measurements and Mrs N. Na-
Ln (D)

ZnO seri for photoelectrochemical analysis is greatly acknowledged.

-2
Ce(30 at.%)-ZnO References

Ce(10 at.%)-ZnO [1] E. Serrano, G. Rus, J. Garcı́a-Martı́nez, Renew. Sustain. Energy Rev. 13 (2009)
2373–2384.
-3 [2] Ch.-H. Yang, H.-L. Chen, Ch.-P. Chen, Sh.-H. Liao, H.-A. Hsiao, Y.-Y. Chuang,
H.-Sh. Hsu, T.-L. Wang, Y.-T. Shieh, L.-Y. Lin, Y.-Ch. Tsai, J. Electroanal. Chem.
Fig. 9. The linear plot of ln(D) versus time for different samples: (a)ZnO, (b) 631 (2009) 43–51.
Ce(30 at.%)–ZnO and (c) Ce(10 at.%)–ZnO thin films. [3] J. Nissfolk, K. Fredin, J. Simiyu, L. H ggman, A Hndersagfeldt, G. Boschloo, J.
Electroanal. Chem. 646 (2010) 91–99.
[4] B. Yoo, K.-J. Kim, S.-Y. Bang, M.J. Ko, K. Kim, N.-G. Park, J. Electroanal. Chem. 638
(2010) 161–166.
0.2
Current density (µA)/mm2

[5] V. Suryanarayanan, K.-M. Lee, J.-G. Chen, K.-Ch. Ho, J. Electroanal. Chem. 633
(2009) 146–152.
0 [6] M. Gupta, V. Sharma, J. Shrivastava, A. Solanki, A.P. Singh, V.R. Satsangi, S. Dass,
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 R. Shrivastav, Bull. Mater. Sci. 32 (2009) 23–30.
[7] A. Agrawal, Y.S. Chaudhary, V.R. Satsangi, S. Dass, R. Shrivastav, Curr. Sci. 85
-0.2 Voltage (V) (2003) 371–374.
[8] K. Rajeshwar, P. Singh, J. Dubow, Electrochim. Acta 23 (1978) 1117–1143.
-0.4 [9] M. Dutta, S. Mridha, D. Basak, Appl. Surf. Sci. 254 (2008) 2743–2747.
[10] A.M.P. Santos, E.J.P. Santos, Thin Solid Films 516 (2008) 6210–6214.
[11] V. Musat, A.M. Rego, R. Monteiro, E. Fortunato, Thin Solid Films 516 (2008)
Pure ZnO -0.6 1512–1515.
[12] K.-S. Ahn, Y. Yan, S. Shet, T. Deutsch, J. Turner, M. Al-Jassim, Appl. Phys. Lett 91
-0.8 (2007) 231909.
Ce(10 at.%)-ZnO [13] A.Z. Moshfegh, J. Phys. D 42 (2009) 233301–233331.
[14] S.T. Tan, B.J. Chen, X.W. Sun, W.J. Fan, H.S. Kwok, X.H. Zhang, S.J. Chua, J. Appl.
-1 Phys. 98 (2005) 013505.
[15] K. Othmer, Encyclopedia of Chemical Technology, John Wiley and Sons, NY,
Fig. 10. Comparing ZCP for both ZnO and Ce(10 at.%)–ZnO thin films. New York, 1992.
112 M. Yousefi et al. / Journal of Electroanalytical Chemistry 661 (2011) 106–112

[16] P. Sangpour, M. Roozbehi, O. Akhavan, A.Z. Moshfegh, Curr. Nanosci. 5 (2009) [35] S. Debnath, M.R. Islam, M.S.R. Khan, Bull. Mater. Sci. 30 (2007) 315.
479–484. [36] U.N. Maiti, P.K. Ghosh, S. Nandy, K.K. Chattopadhyay, Physica B 387 (2007)
[17] K.R. Murali, J. Phys. Chem. Solids 68 (2007) 2293–2296. 103.
[18] X. Yang, A. Wolcott, G. Wang, A. Sobo, R.C. Fitzmorris, F. Qian, J.Z. Zhang, Y. Li, [37] M. El Jouad, M. Alaoui Lamrani, Z. Sofiani, M. Addou, T. El Habbani, N. Fellahi, K.
Nano Lett. 9 (2009) 2331–2336. Bahedi, L. Dghoughi, A. Monteil, B. Sahraoui, S. Dabos, N. Gaumer, Opt. Mater.
[19] Ch. Ge, Ch. Xie, Sh. Cai, Mater. Sci. Eng. B 137 (2007) 53–58. 31 (2009) 1357.
[20] M.F. Al-Kuhaili, S.M.A. Durrani, I.A. Bakhtiari, Appl. Surf. Sci. 255 (2008) [38] Gao-Ren Lu, Xi-Hong Lu, Wen-Xia Zhao, Cheng-Yong Su, Ye-Xiang Tong, Cryst.
3033–3039. Growth Des. 8 (2008) 1276.
[21] W.-J. Lee, H. Okada, A. Wakahara, A. Yoshida, Ceram. Int. 32 (2006) 495–498. [39] Nirmala Chandrasekharan, Prashant V. Kamat, J. Phys. Chem. B 104 (2000)
[22] S. Rani, P. Suri, P.K. Shishodia, R.M. Mehra, Sol. Energy Mater. Sol. Cells 92 10851.
(2008) 1639–1645. [40] Xi-Hong Lu, Gao-Ren Li, Wen-Xiao Zhao, Ye-Xiang Tong, Electrochim. Acta 53
[23] L. Peng, T.-F. Xie, M. Yang, P. Wang, D. Xu, Sh. Pang, D.-J. Wang, Sens. Actuators (2008) 5180.
B 131 (2008) 660–664. [41] Yang Jing-hai, Gao Ming, Zhang Yong-jun, Yang Li-li, Lang Ji-hui, Wang Dan-
[24] Ch. Liu, X. Tang, C. Mo, Zh. Qiang, J. Solid State Chem. 181 (2008) 913–919. dan, Wang Ya-xin, Liu Hui-lian, Fan Hou-gang, We Mao-bin, Liu Fu-zhu, Chem.
[25] J. Yang, M. Gao, L. Yang, Y. Zhang, J. Lang, D. Wang, Y.W.H. Liu, H. Fan, Appl. Res Chin. Univ. 24 (2008) 266.
Surf. Sci. 255 (2008) 2646–2650. [42] Braja Gopal Mishra, G. Ranga Rao, J. Mol. Catal. A: Chem. 243 (2006) 204.
[26] H.-Ch. Gong, J.-F. Zhong, Sh.-M. Zhou, B. Zhang, Zh.-H. Li, Z.-L. Du, Superlattices [43] Young-Sung Kim, Weon-Pil Tai, Su-Jeong Shu, Thin Solid Films 491 (2005) 153.
Microstruct. 44 (2008) 183–190. [44] P. Burroughs, A. Hamnett, A.F. Orchard, G. Thornton, J. Chem. Soc. Dalton Trans.
[27] L. Fen, Y. Bo, Zh. Jie, J. Anxi, Sh. Chunhong, K. Xiangji, W. Xin, J. Rare Earths 25 17 (1976) 1686.
(2007) 306–310. [45] J. Silvestre-Albero, F. Rodŕıguez-Reinoso, A. Seṕulveda-Escribano, J Catál. 210
[28] L.L. Wu, Y.S. Wu, K. Zou, H.D. Gai, Chin. J Lumin 29 (2008) 523. (2002) 127.
[29] Fu.-Shou. Tsai, Su.-Hua. Yang, Wu. Woan-Lin, I.E. EE, Conf. Electron. Devices [46] N. Naseri, M. Amiri, A.Z. Moshfegh, J. Phys. D: Appl. Phys. 43 (2010)
Solid State Circuits (2007) 665. 105405–105412.
[30] Javed Iqbal, Xiaofang Liu, Huichao Zhu, Chongchao Pan, Yong Zhang, Dapeng [47] N. Chandrasekharan, P.V. Kamat, J. Phys. Chem. B 104 (2000) 10851–10857.
Yu, Ronghai Yu1, J. Appl. Phys. 106 (2009) 083515. [48] M. Madecka, M. Wierzbicka, S. Komornicki, M. Rekas, Phys. B 348 (2004)
[31] Z. Sofiani1, B. Derkowska, P. Dalasiński, Z. Łukasiak, K. Bartkiewicz, W. Bała2, 160–168.
M. Addou, A. Lamrani Mehdi, L. Dghughi I. V. Kityk, B. Sahraoui, in: IEEE on 7th [49] M.T.S. Adrián, G.S. Cláudia, G. Drazić, J.L. Faria, Catal. Today 144 (2009) 13–18.
International Conference in Transparent Optical Networks, 2005, pp. 267. [50] A.M. Torres-Huerta, M.A. Domı ´ nguez-Crespo, S.B. Brachetti-Sibaja, H.
[32] Shane O’Brien, L.H.K. Koh, Gabriel M. Crean, Thin Solid Films 516 (2008) 1391. Dorantes-Rosales, M.A. Hernández-Pé rez, J.A. Lois-Corre, J. Solid State Chem.
[33] R.E. Marotti, D.N. Guerra, C. Bello, G. Machado, E.A. Dalchiele, Sol. Energy 183 (2010) 2205–2217.
Mater. Sol. Cells 82 (2004) 85–103. [51] K. Ch. Liu, X. Tang, C. Mo, Zh. Qiang, J. Solid State Chem. 181 (2008) 913.
[34] K.R. Murali, J. Mater Sci: Mater. Electron. 19 (2008) 369.