BiVO4 Heterojunction Films ... - ACS Publications

Naseer Iqbal , Ibrahim Khan , Zain Hassan Abdallah Yamani , Ahsanulhaq Qurashi ..... A. Imam , T. Gries , H. Sezen , M. Amati , D. Mangin , T. Belmont...
0 downloads 0 Views 1019KB Size
LETTER pubs.acs.org/NanoLett

Nanostructured WO3/BiVO4 Heterojunction Films for Efficient Photoelectrochemical Water Splitting Jinzhan Su,† Liejin Guo,*,† Ningzhong Bao,‡ and Craig A. Grimes*,‡ †

State Key Laboratory of Multiphase Flow in Power Engineering, School of Energy and Power Engineering, Xi'an Jiaotong University, Shaanxi 710049, China ‡ State Key Laboratory of Materials-Oriented Chemical Engineering, College of Chemistry and Chemical Engineering, Nanjing University of Technology, Nanjing 210009, China ABSTRACT: We report on a novel heterojunction WO3/ BiVO4 photoanode for photoelectrochemical water splitting. The heterojunction films are prepared by solvothermal deposition of a WO3 nanorod-array film onto fluorine-doped tin oxide (FTO) coated glass, with subsequent deposition of a low bandgap, 2.4 eV, visible light responding BiVO4 layer by spincoating. The heterojunction structure offers enhanced photoconversion efficiency and increased photocorrosion stability. Compared to planar WO3/BiVO4 heterojunction films, the nanorod-array films show significantly improved photoelectrochemical properties due, we believe, to the high surface area and improved separation of the photogenerated charge at the WO3/ BiVO4 interface. Synthesis details are discussed, with film morphologies and structures characterized by field emission scanning electron microscopy and X-ray diffraction. KEYWORDS: WO3, BiVO4, nanorod, nanowire, photoelectrochemical, photoelectrolysis

S

ince the discovery of photoelectrochemical water splitting using n-type TiO2 electrodes,1 hydrogen production by water photoelectrolysis has been a topic of great interest.26 Although many semiconductor materials show photoelectrochemical activity, most of them have limited utility due to being prone to photocorrosion, high charge carrier recombination,7 and band gap energies poorly suited to capturing visible light photons where the bulk of the solar spectrum energy lies.8 Low band gap semiconductors, suitable for capturing visible spectrum light, are prone to photocorrosion while their larger band gap counterparts are not. When a large band gap semiconductor is coupled with a small band gap semiconductor having a more negative conduction band (CB) level, CB electrons can be injected from the small band gap semiconductor into the large band gap semiconductor. BiVO4 has a narrow band gap (2.4 eV) and its conduction band is more negative than that of WO3, so coupled with photocorrosion resistant WO3, WO3/BiVO4 photoanodes9,10 could be suitable for efficient photoelectrochemical cell application under visible irradiation; see Figure 1. Other heterojunction films, such as WO3 /Fe2 O3 ,11 TiO2/CuTiO,12 BiVO 4/Co3O4,13 and TiO2/CdTe14 have shown promising photoelectrochemical activity, with the improvement ascribed to efficient carrier separation in the heterojunction. Metal oxide nanowire/nanorod arrays, which offer large surface areas with physically small distances over which the minority carriers are required to diffuse, offer great promise as photoelectrode materials for photoelectrochemical (PEC) hydrogen generation,15 hence our interest in heterojunction nanoarchitectures that enable efficient light harvesting and photogenerated carrier separation, and thus high r 2011 American Chemical Society

light to chemical (stored energy) conversion efficiencies. In this paper we examine properties of WO3/BiVO4 heterojunctions films, with WO3 nanorod films, perpendicularly aligned to the substrate, grown by solvothermal deposition, upon which BiVO4 layers are deposited by spin-coating. Comparative planar WO3/ BiVO4 heterojunction films are fabricated to help investigate heterojunction properties. Experimental Section. Planar WO3/BiVO4 heterojunctions were synthesized on FTO coated glass substrates (TEC8, 8 Ω/square, Hartford Glass) by spin-coating of the individual layers. Spin-coating was conducted at 4000 rpm for 30 s using a precursor made by dissolving 1.25 g of H2WO4 and 0.5 g of PVA in 10 mL of 50 wt % H2O2. Each spin-coated layer, after annealing in air at 500 °C for 2 h, resulted in a WO3 layer approximately 100 nm thick. BiVO4 was deposited onto the WO3 layers by spin-coating using the procedure described in ref 16, followed by a 400 °C anneal for 2 h in air. Final sample thickness was determined by the number of spin-coating steps. A solvothermal technique was used to grow WO3 nanorod films on FTO coated glass (TEC8). WO3 nanorod growth was assisted by a WO3 seed layer; as described in ref 17, a 100 nm thick seed layer was deposited on a FTO coated glass substrate by spin coating a solution, the same as that used for planar WO3 deposition, followed by 500 °C anneal for 2 h in air. In a typical Received: January 7, 2011 Revised: April 15, 2011 Published: April 22, 2011 1928

dx.doi.org/10.1021/nl2000743 | Nano Lett. 2011, 11, 1928–1933

Nano Letters WO3 nanorod array film deposition, the substrate (with seed layer) was vertically placed in a 23 mL autoclave filled with 3 mL of H2WO4 aqueous solution (0.05 M), 9 mL of ethanol, 4 mL of HCl (6 M), 0.01 g of oxalic acid, and 0.02 g of urea. The H2WO4 aqueous solution was prepared by dissolving 1.25 g of H2WO4 into 30 mL of H2O by adding 10 mL of 50 wt % H2O2 while heating at 95 °C on a hot plate with stirring. The autoclave was then sealed and maintained at 180 °C for 24 h. The substrate was then rinsed with deionized water and dried in a nitrogen stream. The deposited WO3 films were annealed in air at 500 °C for 2 h. Nanostructured WO3/BiVO4 heterojunction films were formed by spin-coating a top layer of BiVO4 onto the nanorod WO3 layer, with a subsequent annealing at 400 °C for 24 h.

Figure 1. Depiction of the energy diagram of the WO3/BiVO4 heterojunction (at pH 7) and electron transport process.

LETTER

Glancing angle X-ray diffraction (GAXRD) analysis was performed using a Scintag X2 diffractometer (Scintag, Inc., CA) to determine the structure and phase of the samples. Linear sweep voltammetry, incident photon-to-current conversion efficiency (IPCE), and intensity modulated photocurrent spectrum (IMPS) measurements were conducted using a two-electrode setup, platinum foil counter electrode, in a Na2SO4 aqueous solution electrolyte. A scanning potentiostat (CH Instruments, model CHI 600C) was used to measure photocurrents at a scan rate of 25 mV/s. During the IMPS measurements a bias of 0.5 V was applied to the photoanode. Sunlight was simulated with a 150 W xenon lamp (Spectra Physics) and AM 1.5 filter (Oriel). The light intensity was set using a NREL calibrated crystalline silicon solar cell, equivalent to global AM 1.5 illumination at 100 mW/cm2. IPCE measurements were performed using a 300 W xenon lamp (Spectra Physics), integrated with a parabolic reflector, passing through an AM 1.5 filter and computer-controlled monochromator with an Oriel calibrated silicon photodiode used for detection. The photocurrent measurement scan was from longer to shorter wavelengths. IMPS data were obtained using a custom built system; a blue light emitting diode (LUXEON III, λ = 455 nm) was used as a light source whose dc illumination was adjusted to 3 mW/cm2. The LED provided both the dc and ac components of the illumination. Light intensity modulation was conducted by current modulation with a depth of 5%. A lock-in amplifier (Stanford Research Systems SR 830) was used to record the photocurrent response as a function of frequency.

Figure 2. Top and cross-sectional (inset) views of (a) planar WO3, (b) planar WO3/BiVO4 heterojunction, (c) WO3 nanorod, and (d) WO3 /BiVO4 nanorod heterojunction films. 1929

dx.doi.org/10.1021/nl2000743 |Nano Lett. 2011, 11, 1928–1933

Nano Letters

LETTER

Figure 3. XRD patterns of annealed (400 °C, 2 h) spin-coated planar WO3 and planar WO3/BiVO4 heterojunction films and solvothermal deposited WO3 nanorod film before and after annealing (500 °C, 2 h).

Figure 4. Currentpotential (with respect to normal hydrogen electrode NHE scale) plots for (a) planar WO3/BiVO4 heterojunction film, (b) planar WO3 film, and (c) planar BiVO4 film illuminated with chopped white light (100 mW/cm2) in an aqueous solution of 0.5 mol 3 L1 sodium sulfate (Na2SO4).

Results and Discussion. Figure 2a shows a field emission scanning electron microscope (FE-SEM) image of a dense planar WO3 film, with thickness of ∼700 nm deposited by seven spin coatingannealing steps. Figure 2b shows a top view image of a planar WO3/BiVO4 heterojunction, mechanically fractured for viewing. The thickness of the BiVO4 layer, formed with a single spin coatinganneal step, is ∼70 nm. Figure 2c shows an asgrown WO3 nanorod array film, with the rods having a square cross section of 60300 nm width, while the length of a nanorod ranges from 500 to 1000 nm. There was no noticeable morphology change after annealing at 500 °C for 2 h. The WO3 rods were covered by nanoporous BiVO4 with one BiVO4 spin-coating step. The crystalline structures of annealed planar WO3, annealed planar WO3/BiVO4, and WO3 nanorod films before and after annealing were characterized by X-ray diffraction (XRD) and are shown in Figure 3. The planar WO3 films showed a monoclinic structure (JSPDF: #00-005-0363). After deposition of a BiVO4 layer new peaks appear at 2θ = 18.6°, 18.9°, and 28.7°, which correspond to the (101), (011), and (112) of monoclinic BiVO4 (JSPDF: #01-75-1867). The as-deposited WO3 nanorod films were indexed as orthorhombic WO3 3 0.33H2O (JSPDF: #00-035-0270). After annealing at 500 °C for 2 h the hydrated tungsten oxide was converted to hexagonal WO3 (JSPDF: #00033-1387) with the highest peak at 2θ = 24.3° corresponding to (110). Variation of temperature between 160 and 200 °C, and reaction times from 16 to 24 h had little influence on the resulting WO3 nanorod array morphology. HCl was found to be essential for nanorod growth; without HCl only particle films were grown. The amount of HCl in the solvothermal growth solution was found to have a significant influence on the nanorod length, while the width of the nanorod was unaffected. With no other changes to the growth chemistry the addition of 1 mL of HCl and 3 mL of H2O resulted in an average nanorod length of 1000 nm. When the quantity of HCl was increased to 4 mL, the nanorod length was found to range between 500 and 1000 nm (Figure 2c). A further increase in the amount of HCl to 6 mL saw a decrease in the nanorod length to ∼300 nm. The additions of oxalic acid and urea were found to facilitate WO3 nanorod growth. With neither, only blocks of WO3, diameters ranging from 200 to 1000 nm,

were grown; with one or both WO3 nanorods of similar dimensions can be grown. For a WO3 film in Na2SO4 a flat-band potential of Vfb = 0.15 V (vs SCE) was reported by Patil,18 while the Vfb of BiVO4 was reported as 0.62 V (vs SCE) by Sayama19 and Li.20 It is generally known that the bottoms of the conduction bands in many n-type semiconductors are more negative by 0.1 V than the flat band potential.21 With the bottom of WO3 conduction band more positive than that of BiVO4, it is favorable for the electrons to travel from the BiVO4 conduction band to WO3, in turn suppressing electronhole recombination. Figure 4 shows a set of linear-sweep voltammagrams (with respect to normal hydrogen electrode NHE scale)22 recorded on planar thin films illuminated through the substrate (back-side illumination) with chopped AM 1.5 light of 100 mW/cm2 intensity. Upon illumination, a single BiVO4 layer (c) showed a photocurrent in the range of 20 μA/cm2, while a ≈700 nm thick WO3 layer (b) showed pronounced photocurrent starting at 0.1 V which continued to increase to 0.4 mA/cm2 at þ1.0 V. With the BiVO4 layer placed upon the WO3 layer, individual layer thicknesses unchanged, the WO3/BiVO4 heterojunction films (a) showed a significant photoresponse enhancement which started at 0.4 V and reached a photocurrent density of 0.8 mA/cm2 at þ1.0 V. The significant photocurrent enhancement was ascribed to improved charge separation due to the WO3/BiVO4 heterojunction structure. Figure 5a shows the IPCE, back-side illumination, of WO3/ BiVO4 heterojunction films comprised of a single BiVO4 layer and multiple WO3 layers. We note that with addition of a BiVO4 layer upon the WO3 base the onset photocurrent wavelength was shifted from 450 to 525 nm, as the low band gap BiVO4 layer extended the absorption edge to longer wavelengths. The conversion efficiency in the region from 400 to 500 nm for the WO3/ BiVO4 heterojunction films increased for up to four WO3 layers, with a decreased value obtained for seven WO3 layers. Under 450525 nm illumination, the photocurrent was mainly generated in the BiVO4 layer, rather than the WO3, thus increasing the WO3 layer thickness should not improve the light absorption over this wavelength region; however the photocurrent was significantly increased. We believe this photocurrent increase 1930

dx.doi.org/10.1021/nl2000743 |Nano Lett. 2011, 11, 1928–1933

Nano Letters

LETTER

Figure 5. IPCE for (a) planar WO3/BiVO4 heterojunction film with a variable number of WO3 layer depositions and (b) planar WO3 film as a function of number of deposition layers; each deposition layer is approximately 100 nm in thickness. Measurements were conducted in an aqueous 0.5 M Na2SO4 solution at a bias of 0.5 V.

due to suitable formation of the WO3/BiVO4 heterojunction, with carrier separation becoming more efficient with a thicker WO3 layer. The optimized thickness of WO3 was determined to be approximately 400 nm, which is thick enough to establish an efficient heterojunction for carrier separation. For WO3 films of higher thickness, ≈700 nm, the IPCE was reduced by unwanted recombination due to the longer transport distance. Figure 5b shows the IPCE of a pure WO3 film as a function of layer thickness. In the short wavelength region the IPCE for a pure WO3 layer was higher than that of WO3/BiVO4 films, with IPCE values increasing with increasing WO3 layer thickness. With back-side illumination, short wavelength light is largely absorbed, within the WO3 layer, close to the back contact allowing efficient collection of the photogenerated electrons, while a BiVO4 top-layer increases the distance required for holes to travel. However, since visible light comprises a greater portion of the solar spectrum energy, the overall AM 1.5 photocurrent was higher for WO3/BiVO4 heterojunction films as seen in Figure 4. IPCE spectra and photocurrents of both planar (≈770 nm thick) and nanorod (≈1000 nm thick) WO3/BiVO4 heterojunction films were measured, see Figure 6, under front (electrolyte to electrode) and back (substrate to electrode) illumination. For the planar WO3/BiVO4 heterojunction films, shown in Figure 6a, the photocurrent under backside illumination is about twice that obtained under front illumination. Hagfeldt et al.23 studied charge separation in sintered colloidal TiO2 film electrodes by

Figure 6. Under front and back illumination, the IPCE for (a) planar WO3/BiVO4 heterojunction film and (b) WO3 nanorod/BiVO4 heterojunction film. IPCE measurements were conducted in a 0.5 M aqueous Na2SO4 solution with a bias of 0.5 V. Insets show currentpotential plots measured with chopped AM 1.5 light (100 mW/cm2) in 0.5 M aqueous Na2SO4 electrolyte.

action spectra in the UV region and concluded that the most efficient charge separation takes place close to the back contact. For backside illumination, the carrier generated by UV light absorbed in the WO3 is close to the back contact, while carriers generated in the BiVO4, due to visible light absorption, is close to the heterojunction; hence higher photocurrents are obtained, see Figure 6a. The IPCE of the WO3 nanorod/BiVO4 heterojunction film is shown in Figure 6b, demonstrating values significantly greater than the planar films. The rough parity between the front and backside illuminated films indicates how the nanorod geometry greatly facilitates charge removal from the interface. For the nanorod WO3/BiVO4 heterojunction films, the photocurrents under backside illumination and front illumination are essentially equal, with both much higher than that of planar WO3/BiVO4 heterojunction films. At a 1 V bias, backside illumination, the nanorod heterojunction gave a photocurrent of 1.6 mA/cm2 compared to 0.8 mA/cm2 for the planar sample. With the nanorod array geometry the WO3/BiVO4 heterojunction is 1931

dx.doi.org/10.1021/nl2000743 |Nano Lett. 2011, 11, 1928–1933

Nano Letters

LETTER

Figure 8. Photocurrent time dependence, under 455 nm, ≈3 mW/cm2, illumination in 0.5 M Na2SO4, of (a) front and (b) back illuminated planar WO3/BiVO4 heterojunction films and (c) BiVO4 nanowire film as described in ref 16.

Figure 7. Complex plane plots of (a) planar WO3/BiVO4 heterojunction film and (b) nanorod WO3/BiVO4 heterojunction film. A 455 nm LED was used as the light source for the IMPS measurement.

formed vertical to the substrate, along which direction light is absorbed. As a result more carriers are generated close to the WO3/BiVO4 heterojunction where they are efficiently separated. Moreover, the large surface area nanorod array structure provides electrons a direct path to the conductive substrate and a short (orthogonal) path for holes to reach the aqueous electrolyte for water oxidation. Intensity modulated photocurrent spectroscopy (IMPS) is a useful method to study semiconductor charge transport properties and is widely used in electron transport characterization of dye-sensitized solar cells.24,25 Figure 7 shows a complex plane plot of the IMPS spectrum for both planar (≈770 nm thick) and nanorod (≈1000 nm thick) WO3/BiVO4 heterojunction films. The response appears in the fourth quadrant of the complex plane and displays one semicircle, where the frequency at the apex of the semicircle can be related to the time constant of the charge transfer process. The low-frequency intercept corresponds to the additional dc photocurrent generated by the intensity increment.26 The average time photogenerated electrons need to reach the back contact, the transit time τD, can be estimated from τD = (2πfmax)1,27 where fmax is the frequency at which the minimum in the IMPS plot occurs. Figure 7a shows τD values of the planar WO3/BiVO4 heterojunction films under front and back illumination of 2.7 and 1.5 ms, respectively; electron transit under back illumination is about two times faster than that under front illumination. Under back illumination the electrons are generated closer to the back contact. Fast electron transport can improve charge-collection efficiency and thus increase photocurrent density, which explains how the low frequency

intercept of the IMPS plot for back illumination is about twice that for front illumination, a result in agreement with the IPCE measurements. τD for nanorod-array WO3/BiVO4 heterojunction films, Figure 7b, measured under front and back illumination are 35.3 and 46.8 ms, respectively; both values are much greater than their planar counterparts. The reason for the slower electron transport in the WO3 nanorod array films is still unknown, but it could be a result of the different electronic properties between hexagonal and monoclinic WO3. It is reported28 that the octahedral ordering in WO3 dominates its electronic properties, with the WO3 structure determined by the arrangements of WO6 octrahedra.29 Still, the IPCE of the nanorod WO3/BiVO4 heterojunction films was higher than that of its planar counterparts, indicating that recombination in the nanorod WO3/BiVO4 films is smaller with the nanostructure facilitating charge separation. Figure 8 shows the photocurrent time dependence, under illumination provide by a 455 nm LED at an intensity of approximately 3 mW/cm2, in 0.5 M Na2SO4, for a planar WO3/BiVO4 heterojunction film under front and back illumination, and a front illuminated BiVO4 nanowire electrode film approximately 700 nm thick. Since the photocurrent for a single layer of spin-coated BiVO4 is quite low, we instead used a BiVO4 nanowire film that we previously reported.16 During the first 300 s after the LED was first switched on, the anodic photocurrent for front and back illuminated WO3/BiVO4 heterojunction film was found to decrease by 70% and 60%, respectively, while the photocurrent of the BiVO4 nanowire film was found to decrease by 95%. Sayama and co-workers19 hypothesize that the observed decline in photocurrent is due to small changes in the BiVO4 valence state or surface structure that in turn promote charge recombination. The photocurrent of the WO3/BiVO4 heterojunction film is more stable than that of the BiVO4 electrode, with the heterojunction film structure promoting electronhole separation on the BiVO4 surface that in turn reduces unwanted charge carrier recombination. For the WO3/BiVO4 heterojunction, under prolonged irradiation the photocurrent recorded under back illumination was more stable than for front illumination. We believe the back illumination further enhances the photoconductivity of the WO3 layer, in turn promoting electron hole separation resulting in a more stable photocurrent. 1932

dx.doi.org/10.1021/nl2000743 |Nano Lett. 2011, 11, 1928–1933

Nano Letters Conclusions. WO3 nanorod array films on FTO coated glass substrates were fabricated using a solvothermal technique. The as-prepared nanorods were orthorhombic WO3 3 0.33H2O, converted to hexagonal WO3 nanorods after annealing at 500 °C for 2 h. A layer of BiVO4 was deposited by spin-coating to form WO3 nanorod/BiVO4 heterojunction films. The photocurrent was enhanced as a result of the visible response of the BiVO4 layer, and the improvement of charge separation in the WO3/BiVO4 heterojunction. In comparison to planar WO3/BiVO4 heterojunction films, nanorod array WO3/BiVO4 heterojunction films showed improved IPCE values, at 420 nm increasing from 9.3% to 31%. Intensity modulated photocurrent spectroscopy showed that electron transport in WO3 nanorod films was dominated by diffusion, a much slower process than electron drift. However faster charge separation induced by the nanostructured heterojunction reduced charge recombination thus improving the overall phototocurrent conversion efficiency. Our results show how one-dimensional nanorod arrays integrated into a heterojunction structure promotes charge-carrier separation and transfer, as well as photocorrosion stability, offering a promising strategy for improving photoelectrochemical water splitting efficiencies.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]; [email protected].

LETTER

(15) Feng, X.; Shankar, K.; Varghese, O. K.; Paulose, M.; Latempa, T. J.; Grimes, C. A. Nano Lett. 2008, 8, 3781–3786. (16) Su, J.; Guo, L.; Yoriya, S.; Grimes, C. A. Cryst. Growth Des. 2010, 10, 856–861. (17) Su, J.; Feng, X.; Sloppy, J. D.; Guo, L.; Grimes, C. A. Nano Lett. 2011, 11, 203–208. (18) Patil, P. S.; Patil, P. R. Bull. Mater. Sci. 1996, 19, 651–656. (19) Sayama, K.; Nomura, A.; Arai, T.; Sugita, T.; Abe, R.; Yanagida, M.; Oi, T.; Iwasaki, Y.; Abe, Y.; Sugihara, H. J. Phys. Chem. B 2006, 110, 11352–11360. (20) Li, M.; Zhao, L.; Guo, L. Int. J. Hydrogen Energy 2010, 35, 7127–7133. (21) Matsumoto, Y.; Omae, K.; Watanabe, I.; Sato, E. J. Electrochem. Soc. 1986, 133, 711–716. (22) Varghese, O. K.; Grimes, C. A. Sol. Energy Mater. Sol. Cells 2008, 92, 374–384. (23) Lindstr€om, H.; Rensmo, H.; S€odergren, S.; Solbrand, A.; Lindquist, S. J. Phys. Chem. 1996, 100, 3084–3088. (24) Zhu, K.; Neale, N. R.; Miedaner, A.; Frank, A. J. Nano Lett. 2007, 7, 69–74. (25) Dloczik, L.; Ileperuma, O.; Lauermann, I.; Peter, L. M.; Ponomarev, E. A.; Redmond, G.; Shaw, N. J.; Uhlendorf, I. J. Phys. Chem. B 1997, 101, 10281–10289. (26) Peter, L. M.; Ponomarev, E. A.; Franco, G.; Shaw, N. J. Electrochim. Acta 1999, 45, 549–560. (27) Kruger, J.; Plass, R.; Gratzel, M.; Cameron, P. J.; Peter, L. M. J. Phys. Chem. B 2003, 107, 7536–7539. (28) Hjelm, A.; Granqvist, C. G. Phys. Rev. B 1996, 54, 2436–2445. (29) de Wijs, G. A.; de Groot, R. A. Phys. Rev. B 1999, 60, 16463–16474.

’ ACKNOWLEDGMENT The authors wish to thank Dr. Oomman K. Varghese for helpful comments and suggestions. The authors gratefully acknowledge financial support from the National Basic Research Program of China (No. 2009CB220000) and the National Natural Science Foundation of China (No. 50821064). ’ REFERENCES (1) Fujishima, A.; Honda, K. Nature 1972, 238, 37–38. (2) Nozik, A. J. Annu. Rev. Phys. Chem. 1978, 29, 189–222. (3) Khaselev, O.; Turner, J. A. Science 1998, 280, 425–427. (4) Mor, G. K.; Shankar, K.; Paulose, M.; Varghese, O. K.; Grimes, C. A. Nano Lett. 2005, 5, 191–195. (5) Murphy, A. B.; Barnes, P. R. F.; Randeniya, L. K.; Plumb, I. C.; Grey, I. E.; Horne, M. D.; Glasscock, J. A. Int. J. Hydrogen Energy 2006, 31, 1999–2017. (6) Grimes, C. A.; Varghese, O. K.; Ranjan., S. Light, Water, Hydrogen: The Solar Genration of Hydrogen by Water Photoelectrolysis; Springer: Norwell, MA, 2007 (ISBN 978-0-387-28597-933198-0). (7) Maruska, H. P.; Ghosh, A. K. Sol. Energy Mater. 1979, 1, 411– 429. (8) Kamat, P. V. Chem. Rev. 1993, 93, 267–300. (9) Chatchai, P.; Kishioka, S.; Murakami, Y.; Nosaka, A. Y.; Nosaka, Y. Electrochim. Acta 2010, 55, 592–596. (10) Chatchai, P.; Murakami, Y.; Kishioka, S.; Nosaka, A. Y.; Nosaka, Y. Electrochim. Acta 2009, 54, 1147–1152. (11) Sivula, K.; Formal, F. L.; Gratzel, M. Chem. Mater. 2009, 21, 2862–2867. (12) Mor, G. K.; Varghese, O. K.; Wilke, R. H. T.; Sharma, S.; Shankar, K.; Latempa, T. J.; Choi, K.; Grimes, C. A. Nano Lett. 2008, 8, 1906–1911. (13) Long, M.; Cai, W.; Cai, J.; Zhou, B.; Chai, X.; Wu, Y. J. Phys. Chem. B 2006, 110, 20211–20216. (14) Seabold, J. A.; Shankar, K.; Wilke, R. H. T.; Paulose, M.; Varghese, O. K.; Grimes, C. A.; Choi, K. Chem. Mater. 2008, 20, 5266–5273. 1933

dx.doi.org/10.1021/nl2000743 |Nano Lett. 2011, 11, 1928–1933