i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 0

Avai lab le at www.sc iencedi rect .com

journa l homepage : www.e lsev ie r . com/ loca te /he

A novel photoelectrochemical cell with self-organized TiO2

nanotubes as photoanodes for hydrogen generation

Yongkun Li a,b, Hongmei Yu a,*, Wei Song a, Guangfu Li a,b, Baolian Yi a, Zhigang Shao a

a Laboratory of Fuel Cells, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, PR ChinabGraduate University of Chinese Academy of Sciences, Beijing 100039, PR China

a r t i c l e i n f o

Article history:

Received 28 May 2011

Received in revised form

27 July 2011

Accepted 7 August 2011

Available online 10 September 2011

Keywords:

Photoelectrochemical cell

Electrochemical anodization

TiO2 nanotubes

Water splitting

Hydrogen production

* Corresponding author. Tel.: þ86 411 843790E-mail address: hmyu@dicp.ac.cn (H. Yu)

0360-3199/$ e see front matter Copyright ªdoi:10.1016/j.ijhydene.2011.08.026

a b s t r a c t

A photoelectrochemical (PEC) cell with an innovative design for hydrogen generation via

photoelectrocatalytic water splitting is proposed and investigated. It consisted of a TiO2

nanotube photoanode, a Pt/C cathode and a commercial asbestos diaphragm. The PEC

could generate hydrogen under ultraviolet (UV) light-excitation with applied bias in KOH

solution. The Ti mesh was used as the substrate to synthesize the self-organized TiO2

nanotubular array layers. The effect of the morphology of the nanotubular array layers on

the photovoltaic performances was investigated. When TiO2 photocatalyst was irradiated

with UV-excitation, it prompted the water splitting under applied bias (0.6 V vs. Normal

Hydrogen Electrode, NHE.). Photocurrent generation of 0.58 mA/cm2 under UV-light irra-

diation showed good performance on hydrogen production.

Copyright ª 2011, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

reserved.

1. Introduction cathode are used and immersed in an aqueous solution

Photoelectrochemical (PEC) water splitting has attracted

significant attention in the past decades as a promising

renewable energy source due to their low production cost and

simple manufacturing process [1e7]. Since the early 1970s,

Fujishima and Honda reported a single-crystalline TiO2

semiconductor photoanode for photoelectrocatalytic decom-

position of water under UV-excitation and an external bias for

the first time [8], great efforts have been devoted to the

development of photocatalytic hydrogen production tech-

nologies [9e14]. However, the practical applications are still

difficult. The main difficulties can be attributed to the lack of

effective photocatalysts and efficient photoelectrocatalytic

reactors [15]. In photoelectrochemical water splitting for

hydrogen generation, the ITO (indium-tin oxide)/FTO (F-doped

SnO2) glass or metal substrate-based photoanode and the Pt

51; fax: þ86 411 84379185.2011, Hydrogen Energy P

especially in alkaline solutions [16e21]. Porous material, such

as glass frit, is used as the separator in the cell compartment.

The drawback is that such a cell design is only suitable for

small scale experiments, and the use of three-electrode or two

separated electrodes in an undivided cell, or in a one-

compartment cell, can be very problematic due to less effec-

tive separation of the generated hydrogen and oxygen. The

gaseous mixture is hazardous and brings collecting problems.

Recently, Kamat and co-workers have reported a hybrid

fuel cell based on the polymer membrane electrode assembly

(MEA) consisting of a P-25 TiO2 photoanode, a Pt cathode, and

a proton exchange membrane for the hydrogen production,

and there have other literatures about the new PECs’ structure

for hydrogen generation [9,22]. However, the photocatalyst

casted onto the carbon paper or carbon cloth would be peeled

easily after long time run. Therefore, it is not suitable for long-

.

ublications, LLC. Published by Elsevier Ltd. All rights reserved.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 0 14375

term water splitting operations. Whereas, the TiO2 nanotube

arrays incorporated with the Ti meshes will avoid this issue.

Furthermore, the ions can be transferred easily from the

anode to the cathode through the openings of the Ti meshes

while it is difficult for Ti foils. In the Ti meshes substrate

system, the ions transfer distance is short, that the ions

transfer to the counter electrode faster. Besides, the growths

of titanium oxide tubes on the Ti foil/meshes will strength the

adhesion of TiO2 nanotube arrays to the substrate. This can be

a great opportunity for their use in electrically enhanced

photo processes in which the electrical contacts are crucial to

obtain high efficiency [23].

In this paper, a novel PEC cell with an innovative design is

proposed and investigated. The core of the PEC cell is an MEA

incorporated with Ti mesh as the substrates of the photo-

anode and Pt/C carbon paper as the cathode. The MEA serves

as a compact but effective reactor for water splitting as well as

an effective gas separator. A self-organized TiO2 nanotube

arrays around the whole Ti mesh with high stabilities in

alkaline environments is employed as the photoanode.

Besides, a detailed investigation of this photoanode applying

on photoelectrolysis of water under illumination of UV light is

carried out. As for this type of PEC cell, the driving force for the

water electrolysis is the applied potential and light source. In

this way, gases generated at each electrode can be separated,

and the utilization of light can be improved in comparison

with the traditional three-arm reactor.

Fig. 1 e Schematic diagram of the photoelectrolysis cell for

hydrogen generation.

2. Experimental section

2.1. Materials

All chemical reagents were commercially available, including

ethylene glycol, ammonium fluoride (NH4F, AR). The purity of

the titanium meshes was 99.8%. A commercial Pt/C electrode

(Pt/C anchored on the Toray TGP-H-60 carbon paper, Sunrise

Power Co. Ltd.) with Pt loading of 0.4 mg/cm2 was used as the

cathode, and asbestos (pore diameter 160 nm, thickness

0.5 mm, SiChuan, China) was used as the diaphragm. The

electrolyte was KOH solution (1 M, pH z 13.6).

2.2. Photoanode preparation

Ti meshes (0.2 mm thick, 99.8% metal basis, Dexmet Co. Ltd.)

with 58% porosity were used as starting material to obtain the

TiO2 photoelectrodes. Prior to anodization, the Ti mesh was

ultrasonically cleaned and degreased in acetone and ethanol

successively, followed by rinsing with deionized water and

drying in an air stream. Anodization was performed in a two-

electrode configuration with titanium mesh as the working

electrode under constant potential (30 V) at room temperature

(20 �C). A rectangle-shaped carbon electrode (thickness 2 mm,

area 4 � 4 cm2) served as the cathode. The distance between

the two electrodeswas kept at 4 cm in all experiments. A direct

current power supply (HY 1791-20S, YaGuang Electronics Co.

Ltd.) was used as the voltage source to drive the anodization.

The Ti mesh was anodized in the electrolyte (pH z 6.2) with

a mixture of ethylene glycol (97.5 wt%), H2O (2 wt%) and NH4F

(0.5 wt%). During the first 2 min of the anodization, little

bubbles were generated out of the surface of the titanium

mesh and afterward less gradually. The colour of the titanium

mesh changed fromwhite to light purple and finally brown. In

order to achieve different morphologies of nanotubular array

layers, anodization time ranged from 4 h to 10 h. After

oxidation, the anodized samples were properly rinsed with

deionized water to remove the occluded ions, and then dried

in an air oven. A subsequent annealing treatment, performed

at 450 �C (heating rate of 2 �C/min) in air, was needed to

transfer the amorphous structure into crystalline anatase.

2.3. Measurements

A field emission scanning electron microscope (FESEM; Hita-

chi, S-4800) was used to analyze the morphology of the

nanotubes and X-ray diffraction (XRD) measurements were

carried out by using an X’pert PRO (PAN-alytical) with a Cu Ka

tube. Photoelectrochemical features of TiO2 photoelectrode

were tested in a three-arm reactor, in which the TiO2 sample

worked as the anode, a platinum grid worked as the counter

electrode and a saturated calomel electrode (SCE) as the

reference electrode; all the values of potential in the text were

referred to SCE. The light source was an 8 W Deuterium lamp;

1 M KOH solution was employed as the electrolyte. The

effective photoactive area of the cell was 1.3 cm2. A computer-

controlled potentiostat (CHI760, CH Instruments Co. Ltd.,

China) was employed to control the external bias and to

record the photocurrent. The samples were anodically polar-

ized at a scan rate of 5 mV/s under the illumination, and the

photocurrent was recorded.

2.4. Photoelectrochemical generation of hydrogen fromwater

In our early patent [24], a PEC cell was described as Fig. 1. With

the anode (TiO2 nanotube) and the cathode (Pt/C carbon paper)

on two sides of the asbestos diaphragm, the membrane elec-

trode assembly (MEA)was pressed between two stainless steel

flow-field plates. There were parallel channels flow at the

cathode side, and an open chamber with a 24 cm2 quartz

window at the anode side, which allowed the photocatalyst

surface exposed to the light. Rubber rings were positioned in

the cell as the sealing. A potentiostat (PARSTAT 2273,

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 014376

Princeton) was employed to control the external bias and to

record the photocurrent generated. A 300 W Xenon Lamp

(Beijing Perfectlight Co. Ltd.) served as the light source and

optical filters kept the UV light (365 � 15 nm) intensity at

70 mW/cm2. A peristaltic pump circulated 1 M KOH solution

from a reservoir to the reactor chamber. The gas was analyzed

by a gas chromatography (GC-14C, CHIMADZU).

3. Results and discussions

3.1. Characterization of the photoanode

Fig. 2 shows the FESEM image of the highly ordered titania

nanotube arrays prepared by the electrochemical anodization

in an electrolytic solution with 2 wt% water and 0.5 wt% NH4F

in ethylene glycol. As shown in Fig. 2a, the structure of the

nanotube arrays prepared at 30 V for 4 h is well ordered with

circular shape, whereas, the nanotubes prepared for 6 h, 8 h

and 10 h are not well ordered and not exactly circular (Fig. 2b,

c, d). It is also observed in Fig. 2 that the length of the titania

nanotubes varies with the anodization time. The length of the

titania nanotubes increases as the anodization time changes

from 4 h to 8 h, but the length limits at 4 mm which does not

Fig. 2 e FESEM images of titania nanotubes prepared by electro

ethylene glycol at 30 V for (a) 4 h (b) 6 h (c) 8 h (d) 10 h.

continue to increase with longer anodization time. On the

other hand, the pore diameter and the wall thickness of the

titania nanotubes does not completely depend on the anod-

ization time. In all the prepared titania nanotubes, most of

them are straight and highly compact, however, some of them

are slightly bent resulting from the strained environment. The

most interesting phenomenon is the nanowire/belt structure

(Fig. 2b, c, and d), that the TiO2 nanowire/belt likes nanograss

floating on the top of the nanotube arrays. The lower part of

the nanotube arrays is straight andwell ordered, while the top

is with poor uniformity and floating at random orientation.

The possible reasons of this phenomenon are as follows.

Firstly, the nanotube wall thickness of the prepared TiO2

arrays is not even that the corrosion happens on the weak

point, while the nanotubular openings that are exposed for

more time than the bottom parts are firstly corroded. Hence, it

is speculated that the corrosion firstly happens on the weak

point along the nanotubes. Secondly, when increasing the

water content to 10 wt% in the electrolyte and keeping F�

constant and other conditions unchanged, the nanowire/belt

structure does not appear, and the nanotubes is well ordered

with circular shape. Therefore, it can be inferred that the

water content is a key role on the formation of the nanowire/

belt structure. F� transfers faster in water than in ethylene

chemical anodization in 2 wt% water D 0.5 wt % NH4F in

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 0 14377

glycol, and there are more F� on the surface of the TiO2 arrays

in water than that in ethylene glycol which leads to the faster

corrosive speed. Compared with the effect of the nanotube

wall nonuniformity, the formation of nanograss’s structure is

more sensitive to the water content in the electrolyte. This

process is effective to produce highly ordered TiO2 nanotube

arrays with tunable morphology via changing the anodization

conditions.

The nanotube arrays prepared at 30 V for 4 h using aqueous

ethylene glycol and ammonium fluoride solution are selected

for further investigation. The XRD patterns of the as-anodized

titania nanotube arrays before annealing treatment shows

only the amorphous structure (figure not included) in nature

and are annealed to convert them into crystalline phases. A

representative XRD pattern of TiO2 nanotubes annealed at

450 �C for 2 h is shown in Fig. 3, which shows anatase phase

except substrate titanium. However, the tube wall thickness

varies a lot compared with the previous state. Obviously, it

does not vary uniformly along the nanotubes. The possible

reason is that the amorphous nanotube arrays have different

dilatabilities during the phase transition.

Fig. 4 e (a) IeV characteristics of TiO2 electrode operated

with UV-illumination at different applied potentials. Scan

rate was 5 mV/s (b) Potentiostatic (Iet) curve obtained from

the photoelectrolytic cell containing TiO2 nanotube as an

anode and Pt grid as a cathode at a potential of 0.2 V vs.

SCE.

3.2. Photoelectrochemical performance of TiO2

photoanode

The titania nanotube arrays (effective photoactive

area ¼ 1.3 cm2) prepared at 30 V for 6 h in 2 wt% and 10 wt%

water electrolyte are used as the photoanodes in a typical

three-electrode system to evaluate the activity for water

photoelectrolysis, respectively.

Fig. 4a shows the photocurrent response to the light.

Photocurrent reaches 112 mA/cm2 and 90 mA/cm2 under exci-

tation, respectively, even with low external bias. The current

drops to zerowhen the light source is switched off. It indicates

that the photogenerated charges can be separated effectively

in the TiO2 anode and the current generated is from photo-

current. It is worth to note that the photocurrent boosts from

Fig. 3 e XRD pattern of the titania nanotubular arrays

prepared at 30 V for 4 h and annealed at 450 �C showing

anatase phases.

90 mA/cm2 to 112 mA/cm2 (Fig. 4b) as the water content

increases from 2 wt% to 10 wt% at the same condition. This

phenomenon indicates that the photoexcited charges of the

10wt%water sample are separatedmore effectively under the

illuminated conditions.

The FESEM images of titania nanotubes prepared by

electrochemical anodization in 2 wt% water (a) and 10 wt%

water (b) þ0.5 wt% NH4F in ethylene glycol at 30 V for 6 h are

showed in Fig. 5. As it is discussed previously, there is

nanowire/belt formed on the top of the nanotubes during the

preparation in the electrolyte of 2 wt% water. This nanowire/

belt can be easily collapsed after calcinations. The collapsed

nanowire/belt blocks the nanotubular openings and prevents

the light passing through. As a result, it cannot absorb and

utilize the sunlight efficiently. However, the openings of the

TiO2 nanotube arrays prepared in the electrolyte of 10 wt%

water are well ordered and uniform, and the generated

photocurrent is obviously higher than that in 2 wt% water

system.

Fig. 5 e FESEM images of titania nanotubes prepared by

electrochemical anodization in 2 wt% water (a) and 10 wt%

water (b) D0.5 wt% NH4F in ethylene glycol at 30 V for 6 h.

Fig. 6 e Photocurrent response of the PEC cell during oneoff

cycles to UV-illumination under applied potential (0.6 V vs.

NHE), measured in 1 M KOH solution.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 014378

The photo conversion efficiency (h) of the photoanode

which converts the light energy into chemical energy is

calculated using the following equation [10,25e27].

hð%Þ ¼ ½ðtotal power output� electrical power outputÞ=light power input� � 100

¼ JP��E0rev � Eapp

��I0�� 100

where, Jp ¼ photocurrent density (mA/cm2); Jp Erev0 ¼ total

power output; JpEapp ¼ electrical power input; I0 ¼ power

density of incident light (mW/cm2).

Erev0 is the standard reversible potential which is 1.23 V vs.

NHE, and the applied potential Eapp¼ Emeas� Eaoc, where Emeas

is the electrode potential (vs. SCE) of the working electrode at

which photocurrent is measured under illumination. Eaoc is

the electrode potential (vs. SCE) of the sameworking electrode

at open circuit conditions under the same illumination and in

the same electrolyte. With the equation, h of the photoanode

usingUV spectrum can be calculated as 2.6% (at�0.5 V vs. SCE,

Jp ¼ 0.112 mA/cm2, OCP ¼ �0.8 V vs. SCE, I0 ¼ 4 mW/cm2). The

OCP is the potential difference between the working electrode

(anode) and the reference electrode (SCE) at equilibrium.

3.3. PEC cell performance

The principle of the PEC cell has shown in Fig. 1. Excitation of

TiO2 nanoparticles with UV-light leads to the electron-holes

separation, most of them recombine again, and the rest can

induce redox reactions. The excited electrons that promote to

the unoccupied conduction band transfer to the cathode

through an external circuit. The holes which stay at the

valence band transfer to the interface of the electrode/elec-

trolyte, and OH� is oxidized. (Equation (1)e(3)).

Excitation : TiO2 þ hv/e� þ hþ (1)

Photoanode : 2hþ þ 2OH�/1=2O2 þH2O (2)

Cathode : 2e� þ 2H2O/H2 þ 2OH� (3)

The PEC cell with the TiO2 photoanode is operated in the

alkaline electrolyte under the applied potential. The TiO2

nanotube photoanode is prepared by electrochemical anod-

ization in 2 wt% water þ 0.5 wt% NH4F in ethylene glycol at

30 V, and the obtained nanotubes length was 3 mm with

anodization time of 4 h.

Fig. 6 shows the photocurrent response of the PEC cell

with TiO2 electrode under UV irradiation during the on/off

cycles under the potentiostatic mode. During on/off cycles of

the illumination, the photocurrent reaches 0.58 mA/cm2 and

the current drops to almost zero without illumination. It

indicates that the PEC cell has good photo-electricity prop-

erties. A 4-h experiment is performed under UV-illumination

with bias potential of 0.6 V for the PEC cell for hydrogen

production. Deaerated KOH solution (1 M) is cycled by

a pump with constant rate through both the compartments

of the PEC cell. The bubbles coming out of the cell are

collected into a vessel. GC-14C gas chromatography is used to

analyze the trapped gas and confirms it to be hydrogen. The

photocurrent keeps a stable value (about 0.58 mA/cm2)

during the entire 4-h test. It indicates that the TiO2 photo-

anode is stable enough to withstand the mechanical stress

and shows no loss of photoactivity. During the 4-h operation,

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 0 14379

the PEC cell with photoactive area of 10.08 cm2 generates

nearly 7.2 ml of hydrogen under the steady-state condition,

so the hydrogen evolution rate of 0.178 ml/h cm2 can be

detected at the average current of 0.58 mA/cm2. If we

consider an average current of 0.58 mA/cm2 flowing through

the circuit, it would expect to observe H2 formation at

a theoretical rate of 0.25 ml/h cm2 (at 15 �C). The observed H2

evolution rate of 0.178 ml/h cm2 accounts for 71% of the

amount predicted on the basis of current flow. The discrep-

ancy between the theoretical and observed yield is likely to

originate from the H2 collecting.

4. Conclusions

A novel PEC cell was constructed by using Ti mesh as the

substrate of the photoanode and Pt/C carbon paper as the

cathode, respectively. Compared with the typical three-arm

reactor, it has many advantages. The Ti mesh was anodized

to synthesize the self-organized TiO2 nanotubular array layer.

The effect of the morphology of the nanotubular array layers

on the photovoltaic performances of the cell was investigated

and the results indicated that the well ordered nanotubular

array layer is a key factor to achieve high conversion efficiency

in the PEC cell. The highest photocurrent density of the

anodization TiO2 nanotubes was about 0.58 mA/cm2 with a H2

evolution rate up to 0.178ml/h cm2. The designed PEC cell was

effective for hydrogen production andwas able to separate the

evolved gases completely. The Pt/C carbon paper as the

cathode rather than Pt foil minimized the catalyst loading,

which was an important progress to cut down the cost of the

PEC cell. Further works are essential to optimize the material

and the structure of the cell, and the progress will be reported

in the near future.

Acknowledgments

This work was financially supported by the National Natural

Science Foundations of China (No. 20636060, No. 20876154).

We also thank Dr. Jingying Shi, the Molecular Catalysis & In-

situ Characterization Group, Dalian Institute of Chemical

Physics, Chinese Academy of Sciences.

r e f e r e n c e s

[1] Kudo A, Miseki Y. Heterogeneous photocatalyst materials forwater splitting. Chem Soc Rev 2009;38:253e78.

[2] Arai N, Saito N, Nishiyama H, Shimodaira H, Kobayashi H,Inoue Y, et al. Photocatalytic activity for overall watersplitting of RuO2-loaded YxIn2�xO3 (x ¼ 0.9e1.5). J Phys ChemC 2008;112:5000e5.

[3] Mor GK, Prakasam HE, Varghese OK, Shankar K, Shankar K,Grimes CA. Vertically oriented Ti-Fe-O nanotube array films:toward a useful material architecture for solar spectrumwater photoelectrolysis. Nano Lett 2007;7:2356e64.

[4] Gong JJ, Lai YK, Lin CJ. Electrochemically multi-anodized TiO2

nanotube arrays for enhancing hydrogen generation by

photoelectrocatalytic water splitting. Electrochim Acta 2010;55:4776e82.

[5] Li Y, Zhang JZ. Hydrogen generation fromphotoelectrochemical water splitting based onnanomaterials. Laser Photonics Rev 2010;4:517e28.

[6] Palmas S, Polcaro AM, Ruiz JR, Pozzo AD, Mascia M, Vacca A.TiO2 photoanodes for electrically enhanced water splitting.Int J Hydrogen Energy 2010;35:6561e70.

[7] Zhang Z, Hossain MF, Takahashi T. Photoelectrochemicalwater splitting on highly smooth and ordered TiO2 nanotubearrays for hydrogen generation. Int J Hydrogen Energy 2010;35:8528e35.

[8] Fujishima A, Honda K. Electrochemical photolysis of water ata semiconductor electrode. Nature 1972;238:37e8.

[9] Jeng KT, Liu YC, Leu YF, Zeng YZ, Chung JC, Wei TY.Membrane electrode assembly-based photoelectrochemicalcell for hydrogen generation. Int J Hydrogen Energy 2010;35:10890e7.

[10] Mohapatra SK, Misra M, Mahajan VK, Raja KS. Design ofa highly efficient photoelectrolytic cell for hydrogengeneration by water splitting: application of TiO2�xCx

nanotubes as a photoanode and Pt/TiO2 nanotubes asa cathode. J Phys Chem C 2007;111:8677e85.

[11] Park JH, Kim S, Bard AJ. Novel carbon-doped TiO2 nanotubearrays with high aspect ratios for efficient solar watersplitting. Nano Lett 2006;6:24e8.

[12] Maeda K, Domen K. Photocatalytic water splitting: recentprogress and future challenges. J Phys Chem Lett 2010;1:2655e61.

[13] Kelly NA, Gibson TL. Solar energy concentrating reactors forhydrogen production by photoelectrochemical watersplitting. Int J Hydrogen Energy 2008;33:6420e31.

[14] Liu ZW, Hou WB, Pavaskar P, Cronin SB. Plasmon resonantenhancement of photocatalytic water splitting under visibleillumination. Nano Lett 2011;11:1111e6.

[15] Minggu LJ, Daud WRW, Kassim MB, Cronin SB. Anoverview of photocells and photoreactors forphotoelectrochemical water splitting. Int J HydrogenEnergy 2010;35:5233e44.

[16] Shankar K, Basham JI, Allam NK, Varghese OK, Mor GK,Grimes CA, et al. Recent advances in the use of TiO2

nanotube and nanowire arrays for oxidativephotoelectrochemistry. J Phys Chem C 2009;113:6327e59.

[17] Cesar I, Sivula K, Kay A, Zboril R, Gratzel M. Influence offeature size, film thickness, and silicon doping on theperformance of nanostructured Hematite photoanodes forsolar water splitting. J Phys Chem C 2009;113:772e82.

[18] Yang XY, Wolcottt A, Wang G, Sobo A, Zhang JZ, Li Y, et al.Nitrogen-doped ZnO nanowire arrays forphotoelectrochemical water splitting. Nano Lett 2009;9:2331e6.

[19] Jang JS, Joshi UA, Lee JS. Solvothermal synthesis of CdSnanowires for photocatalytic hydrogen and electricityproduction. J Phys Chem C 2007;111:13280e7.

[20] Lee YL, Chi CF, Liau SY. CdS/CdSe co-sensitized TiO2

photoelectrode for efficient hydrogen generation ina photoelectrochemical cell. Chem Mater 2010;22:922e7.

[21] Dholam R, Patel N, Adami M, Miotello A. Physically andchemically synthesized TiO2 composite thin films forhydrogen production by photocatalytic water splitting. Int JHydrogen Energy 2008;33:6896e903.

[22] Seger B, Kamat PV. Fuel cell Geared in reverse: photocatalytichydrogen production using a TiO2/Nafion/Pt membraneassembly with no applied bias. J Phys Chem C 2009;113:18946e52.

[23] Varghese OK, Grimes CA. Appropriate strategies fordetermining the photoconversion efficiency of water photo

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 6 ( 2 0 1 1 ) 1 4 3 7 4e1 4 3 8 014380

electrolysis cells: a review with examples using titaniananotube array photoanodes. Solar Energy Mater Solar Cells2008;92:374e84.

[24] Yu HM. CN patent, application NO. 201010238737.9.[25] Khan SUM, Al-Shahry M, Ingler WB. Efficient photochemical

water splitting by a chemically modified n-TiO2. Science2002;297:2243e5.

[26] Mor GK, Shankar K, Paulose M, Varghese OK, hankar K,Grimes CA. Enhanced photocleavage of water using titaniananotube arrays. Nano Lett 2005;5:191e5.

[27] Mishra PR, Shukla PK, Srivastava ON. Study of modular PECsolar cells for photoelectrochemical splitting of wateremploying nanostructured TiO2 photoelectrodes. Int JHydrogen Energy 2007;32:1680e5.