Unravelling Photocarrier Dynamics beyond the ... - ACS Publications

Article pubs.acs.org/cm

Unravelling Photocarrier Dynamics beyond the Space Charge Region for Photoelectrochemical Water Splitting Wenrui Zhang,†,§ Danhua Yan,†,‡,§ Kannatassen Appavoo,† Jiajie Cen,‡ Qiyuan Wu,‡ Alexander Orlov,‡ Matthew Y. Sfeir,† and Mingzhao Liu*,† †

Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, New York 11973, United States Department of Materials Science and Engineering, Stony Brook University, Stony Brook, New York 11794, United States

‡

S Supporting Information *

ABSTRACT: Semiconductor photoelectrodes for photoelectrochemical (PEC) water splitting require efficient carrier generation, separation, and transport at and beyond the space charge region (SCR) formed at the aqueous interface. The trade-off between photon collection and minority carrier delivery governs the photoelectrode design and implies maximum water splitting efficiency at an electrode thickness equivalent to the light absorption depth. Here, using planar ZnO thin films as a model system, we identify the photocarriers beyond the SCR as another significant source to substantially enhance the PEC performance. The high-quality ZnO films synthesized by pulsed laser deposition feature very few deep trap states and support a long photocarrier lifetime. Combined with photoelectrochemical characterization, ultrafast spectroscopy, and numerical calculations, it is revealed that engineering the exciton concentration gradient by film thickness facilitates the inward diffusion of photocarriers from the neighboring illuminated region to the SCR and, therefore, achieves a record high quantum efficiency over 80% at a thickness far beyond its light absorption depth and the SCR width. These results elucidate the important role of the photocarriers beyond SCR for the PEC process and provide new insight into exploring the full potential for efficient photoelectrode materials with large exciton diffusivity. density of 1 for incident photon.7,8 The popular solution of siteselective or graded doping has been widely exploited either to improve the incident photon collection in wide band gap semiconductors such as TiO29−11 and SrTiO312 or to facilitate the carrier transport in semiconductors such as BiVO413−15 and Fe2O3.16,17 However, it is also noticed that chemical doping may inevitably compromise electrical or catalytic properties of the host semiconductor and degrade the photocatalytic performance.18−20 On the other hand, the transport behavior of photocarriers beyond the SCR and its effect on the overall device performance have received very limited attention. In the conventional Gartner’s model, it is simply assumed that the photoholes generated beyond SCR diffuse to SCR in a fixed

1. INTRODUCTION Since Fujishima and Honda’s pioneering work on water photolysis over TiO2 electrodes, 1 photoelectrochemical (PEC) water splitting has attracted significant research interests as a promising method for solar energy conversion.2−4 Efficient water splitting requires fast charge carrier generation, separation, and delivery in the semiconductor photoelectrodes.5 The carrier separation efficiency largely relies on the space charge region (SCR) formed at the semiconductor−electrolyte interface, which is also known as the depletion region due to the depletion of majority carriers.6 To maximize the device efficiency, it is a conventional wisdom to have the majority of photocarriers generated in the depletion region, where carrier relaxation is minimized and most minority carriers can be delivered to the reactive interface. As such, a number of studies have been focused on broadening the SCR, which is typically a few tens of nanometers in width, to match the effective photon attenuation depth dp, i.e., the depth that achieves an optical © 2017 American Chemical Society

Received: February 16, 2017 Revised: April 18, 2017 Published: April 18, 2017 4036

DOI: 10.1021/acs.chemmater.7b00672 Chem. Mater. 2017, 29, 4036−4043

Article

Chemistry of Materials

Figure 1. (a) Schematic of ZnO-based photoanodes for PEC water splitting. Top-view SEM images of 400 nm-thick ZnO films on (b) SRO/STO and (d) ITO glass substrates. (c) Cross-sectional STEM and (e) SEM image of ZnO films on SRO/STO and ITO glass substrates. electron microscopy (TEM, JEOL 2100). The surface morphology was studied with scanning electron microscopy (SEM, Hitachi S-4800) and atomic force microscopy (AFM, Park NX20). UV−vis absorption spectra were obtained with a UV−vis/NIR spectrophotometer (PerkinElmer, Lambda 950). Photoluminescence (PL) spectra of ZnO films were measured in an ISS PC1/K2 spectrofluorometer equipped with a xenon lamp for optical excitation and a photon multiplier tube for detection. An excitation wavelength of 280 nm was selected by grating and passed through a band-pass filter. On the detection side, the scattered excitation light was removed by a 320 nm long-pass filter. Photoelectrochemistry and Ultrafast Spectroscopy Measurements. PEC measurements were performed with a potentiostat (VersaStat4, PAR) in a three-electrode configuration with ZnO films as the working electrode, a platinum wire as the counter electrode, an Ag|AgCl|3 M KCl electrode as the reference electrode, and a 0.1 M KOH solution (pH = 13) as the electrolyte. An active illumination area of 0.18 cm2 was obtained from a 150 W solar simulator with an AM 1.5G filter (100 mW/cm2) calibrated by a thermopile optical detector. Incident photon-to-current efficiency measurements were conducted with a 300 W xenon arc lamp and a grating monochromator (Newport CS310) equipped with band-pass filters to remove high-order diffractions. The light power for each wavelength was measured by an optical power meter (Newport 1918-C) equipped with a UVenhanced Si photodiode sensor. The time-integrated emission measurement was carried carrier using a commercial Ti:sapphire laser (800 nm, 100 fs, 3.5 mJ, 1 kHz) with an optical parametric amplifier as the excitation source tuned to 280 nm. A portion of the fundamental beam was focused onto the benzene (with a temporal resolution of ∼500 fs) as the Kerr material for gating the ZnO emission.

probability that decays exponentially over the diffusion range. In recent years, it has been reported that the carrier dynamics beyond SCR can be nontrivial. For example, the formation of a “dead layer” at the metallic back contact is known to severely undermine charge carrier separation for hematite photoanodes.21,22 Here we use ZnO thin films grown by pulsed laser deposition as a model system to investigate the photocarrier transport beyond SCR and its effect on overall PEC performance. ZnO was selected for being naturally n-type doped,23 as well as its high carrier mobilities.24 In this study, we find a strong thickness-dependent PEC behavior of ZnO film photoanodes (25−800 nm). A maximum photocurrent density of 0.77 mA/cm2 and a record high quantum efficiency over 80% are was obtained at a thickness of 400 nm, far beyond the photon attenuation depth (dp) of about 140 nm. The thickness dependence is found to deviate from the classical Gärtner’s model that predicts the saturation of photocurrent when the film thickness reaches dp.25 Through numerical simulation of carrier transport, we find that the deviation originates from the long-range, backward diffusion of excitons that is not accounted for in Gartner’s model.

2. EXPERIMENTAL SECTION Preparation of ZnO Photoelectrodes. ZnO target was prepared by a conventional ceramic sintering method. ZnO films were grown on either SrRuO3 (SRO, 20−30 nm in thickness) buffered single crystal SrTiO3 (STO) (001) substrates or indium tin oxide (ITO) coated glass substrates by pulsed laser deposition using a KrF excimer laser (λ = 248 nm) with a laser fluence of 1.5 J/cm2 and a repetition rate of 5 Hz. For films grown on STO substrates, the SRO bottom electrode was first deposited at 750 °C in 200 mTorr of oxygen, and ZnO films were subsequently deposited under the same growth condition. For films grown on ITO coated glass substrates, the growth conditions were kept the same except that the substrate temperature was lowered to 500 °C to avoid glass substrate melting. The film thickness was controlled by adjusting the total number of laser pulses. Microstructure, Optical, and Photoluminescence Characterization. The film microstructure and crystallinity were characterized by X-ray diffraction (XRD, Rigaku Ultima III) and transmission

3. RESULTS AND DISCUSSION ZnO thin film photoanodes are fabricated by pulsed laser deposition (PLD) on conductive substrates, which are either SrRuO3 (SRO, 20−30 nm in thickness) buffered SrTiO3 single crystals (STO, ⟨001⟩ cut) or indium tin oxide (ITO) coated glass substrates (Figure 1a). Surface morphology and microstructure of the ZnO thin film photoanodes are studied by scanning electron microscopy (SEM), atomic force microscopy 4037


Article


Figure 2. θ−2θ XRD scans of ZnO films on (a) SRO/STO (001) substrate and (b) ITO glass substrate. (c) Selected area electron diffraction pattern of a 400 nm-thick ZnO film on SRO/STO. (g) High resolution TEM image of the ZnO/SRO interface showing the coexistence of (110)and (002)-oriented ZnO domains.

A strong thickness dependence on the PEC activity is found from the photocurrent density−potential (J−V) measurements for ZnO thin film photoelectrodes deposited over SRO/STO (Figure 3a). Under the simulated AM1.5G solar illumination, steady anodic photocurrents are observed for all ZnO/SRO/ STO films. The dark current, on the other hand, is largely negligible in the potential range of interest (Figure 3a, black line). In this study, the ZnO film thicknesses is varied between 40 and 750 nm, where the lower limit is comparable to the depletion width and the upper limit far exceeds the effective photon attenuation depth. As the film thickness increases from 40 to 140 nm, we observe over 2-fold increase in the photocurrent density, from 0.13 mA/cm2 to 0.33 mA/cm2 at 1.23 VRHE, as well as a 0.15 V cathodic shift of the photocurrent onset potential. This can be understood by the fact that thicker films have more efficient photon collection and deliver more photogenerated carriers to the aqueous interface. Optical absorption measurement determines an optical bandgap of 3.27 eV for as-prepared ZnO thin films (Figure 4a). The absorption coefficient is corrected with the background intensity and calculated to be in the range of 3.85−4.35 × 105 cm−1 for λ < 370 nm, which is consistent with literature reported values.26 This suggests that a 140 nm-thick film already absorbs over 99% of the incident light with λ < 370 nm. As such, one would expect, based on Gärtner’s model, that any further increase of the film thickness should not improve the photon collection or the photocurrent significantly.25,27 Contrary to this expectation, it is observed that the photocurrent improves continuously as the film thickness increases beyond 140 nm, until it reaches 400 nm. The ZnO/SRO/STO photoelectrode of this thickness exhibits the highest photocurrent density of 0.77 mA/cm2 at 1.23 VRHE, with an excellent

(AFM), and transmission electron microscopy (TEM). A columnar morphology consisting of granular domains emerges from the nucleation-assisted island growth, as evidenced by topview SEM images of a representative 400 nm-thick ZnO film on SrRuO3/SrTiO3 (SRO/STO) (Figure 1b). The average domain size increases from 54 to 180 nm as the film thickness increases from 40 to 140 nm but remains steady for even thicker films (Figure S1). Here the domain size is represented by the domain width considering the irregular shape of ZnO domains. The columnar growth extends across the entire film thickness, according to the cross-sectional scanning TEM (STEM) image of a 750 nm-thick sample (Figure 1c). The surface roughness varies between 8 and 12 nm at a film thickness below 400 nm and increases substantially for thicker films (Figure S2). The films grown on ITO substrates exhibit a similar columnar morphology with an average domain size of 130 nm (Figure 1d,e). Orientation and crystallinity of the ZnO films are investigated by X-ray diffraction (XRD). According to the θ− 2θ XRD scans, ZnO films grown on SRO/STO substrates have two major orientations respectively along the (001) and (110) axes, establishing local epitaxy along these two orientations with SRO (Figure 2a). As the film thickness increases, diffraction peaks from other orientations do appear, but with much lower intensity. The formation of (001) and (110) oriented domains is further confirmed by selective area electron diffraction (SAED, Figure 2c) and high resolution TEM (Figure 2d), reflecting the large lattice mismatch between ZnO and SRO/ STO. For comparison, ZnO films grown the ITO glass substrate are highly polycrystalline with no preferred orientations (Figure 2b). 4038


Article


Figure 3. (a) Photocurrent density versus potential curves of ZnO films on SRO/STO with different film thicknesses. The dark current density curve of a 400 nm-thick film is provided as a comparison. (b) Photocurrent density (left axis) and energy conversion efficiency (right axis) for ZnO films on SRO/STO and ITO glass as a function of film thickness. (c) IPCE as a function of excitation wavelength and electrode potential of a 400 nmthick ZnO film on SRO/STO. (d) Wavelength-dependent IPCE at 1.23 VRHE for ZnO films on SRO/STO with different film thickness.

Figure 4. (a) Tauc plot of a 70 nm-thick ZnO film determining an optical band gap of 3.27 eV. Inset is the corresponding optical film transmittance versus incident photon energy. (b) Photoluminescence spectra of ZnO films (70, 140, 400 nm) on SRO/STO and a 400 nm-thick ZnO film on ITO. (c) Time-resolved broadband emission spectrum of a 750 nm-thick ZnO film on SRO/STO. The kinetic trace in the inset determines a lifetime of 38 ps for the photoelectrons.

films grown on ITO glass substrates exhibit very similar thickness dependence despite their lower film crystallinity, with the photocurrent also maximized for a 400 nm-thick film (Figure 3b), featuring a photocurrent density of 0.75 mA/cm2 at 1.23 VRHE, an onset potential at 0.32 VRHE, and a peak energy conversion efficiency of 0.21% at 0.77 VRHE. The J−V curves of ZnO/ITO photoelectrodes with other film thicknesses are presented in Figure S3. Regardless of the back contact being

filling factor and a low onset potential at 0.35 VRHE (Figure 3a). The solar energy conversion efficiency (η) of this photoelectrode reaches a maximum value of 0.25% at 0.73 VRHE, which is calculated using η = J

E ° (O2 | H 2O) − E I0

× 100%, where J

is the photocurrent density, (E°(O2|H2O) − E) the electrode underpotential versus E°(O2|H2O), and I0 = 100 mW/cm2 the illumination intensity of the AM 1.5G solar spectrum.25 ZnO 4039


Article


Figure 5. (a) Schematic of three regions emerging in a ZnO thin film photoelectrode. Region I is the depletion region that dictates the electron− hole separation. Region II is the illumination quasi-neutral region which determines the maximum photon attenuation depth. Region III is the dark quasi-neutral region that connects Region II and the conductive substrate. The bottom panel shows the corresponding electron density profile as a function of depth x, i.e., the distance away from the aqueous interface. (b) Hole density profile and (c) hole current density as a function of depth x across the three regions.

Doping levels in the ZnO films are obtained through electrochemical impedance spectroscopy (EIS) measurements performed in the same electrolyte (0.1 M KOH) as the photocurrent measurements. The donor density (ND) is determined by the Mott−Schottky (MS) equation based on a planar semiconductor/electrolyte interface:

SRO or ITO, the photocurrent decreases continuously for films thicker than 400 nm, with anodic shift of photocurrent onset potential, suggesting the degradation of charge carrier separation and delivery. The photocurrent onset potential is between 0.3 and 0.5 VRHE for all measured ZnO films with a sample-to-sample variation of ±0.05 V. In general, the cathodic shift is associated with larger photocurrent (Figure 3a), although domain size variation and surface morphology may also cause the variation of the photocurrent onset potential as a result of carrier recombination.28 The incident photon-to-current conversion efficiencies (IPCE) of the 400 nm-thick ZnO/SRO/STO photoelectrode are measured at a set of electrode potentials between 0.56 and 1.46 VRHE (Figure 3c). For each tested potential, the IPCE remains steady for direct band edge excitation and drops rapidly at λ > 370 nm (Figure S4). In general, the IPCE improves toward more anodic potential, owing to the improved minority carrier delivery to the aqueous interface as the SCR widens. In the case of excitation at 370 nm, the IPCE improves from 41% at 0.56 VRHE to 86% at 1.23 VRHE. The thickness dependence of their IPCE values at 1.23 VRHE largely follows the same trend as their photocurrent densities under simulated solar radiation (Figure 3d). For λ < 370 nm, the IPCE increases continuously with increasing film thickness from 40 nm onward and maximizes above 80% for the 400 nm-thick film. The IPCE drops for even thicker films and reaches 66% at 750 nm-thick, which may be due to increased photogenerated carrier recombination from an extended photocarrier transport distance. It is also noted that even thicker films feature much larger surface roughness and more complicated orientations, which may further compromise the charge carrier transport. The observed high IPCEs reflect the high quality of ZnO films, which is further confirmed by their photoluminescence (PL) spectra (Figure 4b). Regardless of the film thickness or the substrate, the ZnO films exhibit only band gap emission at ∼3.3 eV with no sign of any deep-trap emission at lower energy. This facilitates a fairly long lifetime of photogenerated carriers without being captured by deep trap states.29,30 Using an ultrafast transient emission spectroscopy technique,31,32 we monitor exciton recombination kinetics from a 750 nm-thick ZnO film on SRO/STO and determine an exciton lifetime of 38 ps (Figure 4c).

1 2 = (E − E FB − kBT /e) 2 eεε0ND Csc

(1)

where Csc is the space charge capacitance per unit area, e the magnitude of the elementary charge, ε the semiconductor relative permittivity (∼10 for ZnO), ε0 the vacuum permittivity, E the electrode potential, EFB the flat band potential, and kBT the thermal energy. Regardless of their thicknesses, all the ZnO thin film electrodes show positive slopes in the MS (1/Csc2 vs E) plots, which is characteristic of an n-type semiconductor (Figure S5). The depletion width (dsc) is obtained from dsc = εε0/Csc and is about 20−30 nm for all ZnO films on SRO/ STO. The calculated values of ND are on the order of 1 × 1018 cm−3, and EFB values vary between 0.40 and 0.49 VRHE (Supporting Table S1). It is found that the PEC activity of ZnO photoelectrodes is not very sensitive to the magnitude of ND. For example, ND of the 400 nm-thick ZnO film on SRO/ STO (3.58 × 1017 cm−3) is much smaller than that of the film of same thickness on ITO glass (1.43 × 1019 cm−3), but both films show very similar photocurrent density (0.76 ± 0.01 mA/ cm2) and photocurrent onset potential (0.33 ± 0.02 VRHE). The ND variation measured from MS plots may be due to different surface roughnesses of ZnO films on SRO/STO and ITO glass. AFM measurements determine a root-mean-square roughness of 16.335 nm for a 400 nm-thick ZnO film on ITO, which is nearly twice that (8.211 nm) for the ZnO film on SRO/STO (Figure S6). The varied surface roughness makes it challenging to derive accurate ND from MS analysis.33 The most striking feature observed here is that the photocurrent does not saturate when the film thickness reaches the photon attenuation depth dp (∼140 nm) but continuously rises for thicker films. As illustrated in the top panel of Figure 5a, there are typically three regions emerging in a semiconductor photoanode upon front illumination. Immediately next to the aqueous interface is the depletion region (Region I). Further into the bulk is the quasi-neutral region, in which the transport of photoholes is dominated by diffusion. The region 4040


Article


Figure 6. (a) Back/front photocurrent density ratio of ZnO films on ITO as a function of film thickness. Inset shows the front-illuminated and backilluminated J−V curves of a 400 nm-thick ZnO/ITO photoanode. (b) Wavelength-dependent IPCE at 1.23 VRHE for the 400 nm-thick ZnO/ITO photoanode under front and back illumination.

may be divided into two, i.e., a top part (Region II) where incident photons can reach and a bottom part (Region III) which always remains dark. In the conventional Gärtner’s model, a photohole generated in Region II is assumed to reach the aqueous interface at a constant probability of e−x/dh, where dh is the hole diffusion length and x the photohole’s initial distance from Region I.25 As such, the photocurrent should saturate once the film thickness reaches dp. The apparent contradiction to our observation likely reflects the oversimplification of thermal diffusion in Gärtner’s model. In reality, thermal diffusion is collectively driven by concentration gradient and is more complicated in the problem of interest. The distribution and transport of charge carriers in a ZnO thin film photoanode are numerically calculated by the finite element method, using the drift-diffusion model.34 In the calculation, a band bending of 1 V is set across a ZnO thin film with ND = 1018 cm−3. Literature values are adapted for material properties such as the effective masses and mobilities of electrons and holes, as well as the relative permittivity and optical constants of ZnO.24,35 Because of the large exciton binding energy of ZnO (∼60 meV),24 most photoholes are bound with electrons in the quasi-neutral region as excitons, which are treated separately in the calculation. The exciton lifetime is 38 ps according to our time-resolved photoluminescence measurement and its diffusion constant is given by Jeong et al. as 5 cm2 s−1.36 The electron density profile given by the calculation clearly indicates the formation of a depletion region of about 30 nm thick (Figure 5a, bottom panel), consistent with the expectation from Mott−Schottky analysis. On the other hand, the steady state density of photoholes exhibit a much more complicated profile (Figure 5b). In general, the photoholes predominantly exist in the form of bounded excitons outside the depletion region (Regions II and III) and are almost completely dissociated to free carriers inside (Region I). The total density of photoholes is maximized in the depletion region because of the higher exciton generation rate and the attraction from upward band bending. Intriguingly, a local maximum of hole density emerges near the center of Region II, around which the hole density spreads to a very wide range and deeply into the dark Region III. The hole density remains substantial up to a depth of 400 nm, far beyond the photon attenuation depth of 140 nm.

The relatively high density of photoholes in the dark Region III attributes to the backward diffusion of photoholes. As shown in Figure 5c, the steady state hole current (Jh) is dominated by drift current in the depletion region but evolves to diffusiondominated toward the quasi-neutral region. The diffusion current is toward the aqueous interface in the front portion of Region II but turns backward at deeper regions, which diminishes only at a depth beyond 400 nm. Therefore, for ZnO films thinner than 400 nm, a significant portion of photoholes can reach the metallic back contact for recombination with free electrons, which leads to lower overall photocurrent and reproduces the trend observed experimentally. In other words, the dark Region III provides a diffusion barrier that prevents photoholes from reaching the back contact for recombination. The steady state electron current (Je) is presented in Supporting Information (Figure S7). The photocarrier density in ZnO films can be further tuned by adjusting the incident illumination intensity using neutral density filters. Maximum photocurrent is seen for the 400 nmthick ZnO film regardless of the incident illumination intensity (Figure S8), verifying the significant role of exciton diffusion in the enhanced PEC activity. The role of exciton diffusion and recombination at the back contact is further demonstrated by comparing J−V measurements under front (Jfront) and back (Jback) illumination. ZnO/ ITO glass photoelectrodes are used for this measurement since ITO glass has low optical absorption in the spectral range of 320−400 nm. Nevertheless, to achieve a fair comparison, a bare ITO glass substrate is inserted into the excitation beam path for the front illumination configuration to compensate the ITO light absorption in the back illumination configuration. We find that Jback is smaller than Jfront for all the samples. For instance, the Jback/Jfront ratio is only 64.7% at 1.23 VRHE for the 25 nmthick sample (Figure 6a). Given that the film is completely depleted of majority carriers, the lower performance for back illumination can only be explained by elevated carrier recombination at the ZnO/ITO interface, near which more carriers are generated at this configuration. As the film thickness increases, the Jback/Jfront ratio decreases further, implying that the photoholes generated near the ZnO/ITO interface are getting lower probability to reach the aqueous interface by diffusion. On the other hand, the Jback/Jfront ratio remains ∼10% even for the 800 nm-thick sample, again reflecting the 4041


Article

Chemistry of Materials ORCID

extraordinarily large exciton mobility in ZnO. IPCE measurements for the 400 nm-thick film indicate that the IPCE ratio of back illuminations is about 2% for above band gap excitation (Figure 6b). It is quite intriguing that the ratio rises significantly toward lower energy and surpasses 10% at 3.26 eV (380 nm). This can be explained by the lower optical absorption near the band edge, so that back illuminated light can penetrate deeper into the ZnO film and lead to more carrier generation near the aqueous interface, thus improving the carrier delivery efficiency.

Wenrui Zhang: 0000-0002-0223-1924 Danhua Yan: 0000-0003-4112-3018 Qiyuan Wu: 0000-0002-8751-2003 Matthew Y. Sfeir: 0000-0001-5619-5722 Author Contributions §

W. Z. and D. Y. contributed equally to this work.

Notes

The authors declare no competing financial interest.

4. CONCLUSION High-quality ZnO thin film photoelectrodes are fabricated by pulsed laser deposition to investigate the photocarrier dynamics beyond the SCR for PEC water splitting. Photoluminescence measurements reveal very few deep trap states and fairly long charge carrier lifetime in as-prepared ZnO films regardless of the underlying substrate and the film crystalline orientation. Engineering the film thickness controls the photocarrier concentration profile and directs long-range diffusion of excitons from the illuminated quasi-neutral region to the SCR, thus leading to dramatically enhanced PEC activity and very high IPCEs. A maximum photocurrent density of 0.77 mA/cm2 and a record high quantum efficiency over 80% are obtained at a thickness of 400 nm, far beyond the photon attenuation depth and the SCR width. These results emphasize the importance of charge carriers beyond the SCR that is usually neglected in the classical Gartner’s model in improving the photoelectrode performance. Semiconductors with large exciton binding energies, such as ZnO (∼60 meV) and GaN (∼23 meV),37 may feature enhanced PEC performance through inward diffusion of photocarriers from the neighboring illuminated region to the space charge region. For semiconductor materials such as α-Fe2O3, TiO2, and SrTiO3, the exciton binding energy is no more than a few meV and the exciton diffusion is very limited, in which the photoelectrode design is dominated by the canonical Gartner’s model. This study sheds light on the design of photoelectrode materials with large exciton diffusivity for efficient PEC water splitting.

■

ACKNOWLEDGMENTS

■

REFERENCES

This research used resources of the Center for Functional Nanomaterials, which is a U.S. DOE Office of Science User Facility, at Brookhaven National Laboratory under Contract No. DE-SC0012704. J.C., Q.W., and A.O. acknowledge funding support from the National Science Foundation (#1254600).

(1) Fujishima, A.; Honda, K. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 1972, 238, 37−38. (2) Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Solar Water Splitting Cells. Chem. Rev. 2010, 110, 6446−6473. (3) Kang, D.; Kim, T. W.; Kubota, S. R.; Cardiel, A. C.; Cha, H. G.; Choi, K.-S. Electrochemical Synthesis of Photoelectrodes and Catalysts for Use in Solar Water Splitting. Chem. Rev. 2015, 115, 12839−12887. (4) Moniz, S. J. A.; Shevlin, S. A.; Martin, D. J.; Guo, Z.-X.; Tang, J. Visible-light Driven Heterojunction Photocatalysts for Water Splitting - a Critical Review. Energy Environ. Sci. 2015, 8, 731−759. (5) Tachibana, Y.; Vayssieres, L.; Durrant, J. R. Artificial Photosynthesis for Solar Water-Splitting. Nat. Photonics 2012, 6, 511−518. (6) Nozik, A. J. Photoelectrochemistry: Applications to Solar Energy Conversion. Annu. Rev. Phys. Chem. 1978, 29, 189−222. (7) Hisatomi, T.; Kubota, J.; Domen, K. Recent Advances in Semiconductors for Photocatalytic and Photoelectrochemical Water Splitting. Chem. Soc. Rev. 2014, 43, 7520−7535. (8) Zhang, P.; Wang, T.; Chang, X.; Gong, J. Effective Charge Carrier Utilization in Photocatalytic Conversions. Acc. Chem. Res. 2016, 49, 911−921. (9) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. VisibleLight Photocatalysis in Nitrogen-Doped Titanium Oxides. Science 2001, 293, 269−271. (10) Park, J. H.; Kim, S.; Bard, A. J. Novel Carbon-Doped TiO2 Nanotube Arrays with High Aspect Ratios for Efficient Solar Water Splitting. Nano Lett. 2006, 6, 24−28. (11) Wang, G.; Wang, H.; Ling, Y.; Tang, Y.; Yang, X.; Fitzmorris, R. C.; Wang, C.; Zhang, J. Z.; Li, Y. Hydrogen-Treated TiO2 Nanowire Arrays for Photoelectrochemical Water Splitting. Nano Lett. 2011, 11, 3026−3033. (12) Liu, M.; Lyons, J. L.; Yan, D.; Hybertsen, M. S. SemiconductorBased Photoelectrochemical Water Splitting at the Limit of Very Wide Depletion Region. Adv. Funct. Mater. 2016, 26, 219−225. (13) Kim, T. W.; Choi, K.-S. Nanoporous BiVO4 Photoanodes with Dual-Layer Oxygen Evolution Catalysts for Solar Water Splitting. Science 2014, 343, 990−994. (14) Rao, P. M.; Cai, L.; Liu, C.; Cho, I. S.; Lee, C. H.; Weisse, J. M.; Yang, P.; Zheng, X. Simultaneously Efficient Light Absorption and Charge Separation in WO3/BiVO4 Core/Shell Nanowire Photoanode for Photoelectrochemical Water Oxidation. Nano Lett. 2014, 14, 1099−1105. (15) Abdi, F. F.; Han, L.; Smets, A. H. M.; Zeman, M.; Dam, B.; van de Krol, R. Efficient Solar Water Splitting by Enhanced Charge Separation in a Bismuth Vanadate-silicon Tandem Photoelectrode. Nat. Commun. 2013, 4, 2195.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.chemmater.7b00672. SEM and AFM images of ZnO films on SRO/STO substrates with different film thicknesses. Photocurrent density versus potential curves of ZnO films on ITO glass substrate. IPCE as a function of excitation wavelength and electrode potential of ZnO films on SRO/STO with different thicknesses. Mott−Schottky plots for ZnO films on SRO/STO and ITO glass substrates. AFM images and height profiles of 400 nmthick ZnO films on SRO/STO and ITO glass substrates. Photocurrent density under different incident illumination intensity of ZnO films. Electron current density as a function of depth x across the three regions. Electric properties of ZnO films on SRO/STO and ITO substrates with different film thicknesses (PDF)

■

■

AUTHOR INFORMATION

Corresponding Author

*(M.L.) E-mail: [email protected]. 4042


Article

Chemistry of Materials (16) Sivula, K.; Le Formal, F.; Grätzel, M. Solar Water Splitting: Progress Using Hematite (α-Fe2O3) Photoelectrodes. ChemSusChem 2011, 4, 432−449. (17) Yan, D.; Tao, J.; Kisslinger, K.; Cen, J.; Wu, Q.; Orlov, A.; Liu, M. The Role of the Domain Size and Titanium Dopant in Nanocrystalline Hematite Thin Films for Water Photolysis. Nanoscale 2015, 7, 18515−18523. (18) Choi, W.; Termin, A.; Hoffmann, M. R. The Role of Metal Ion Dopants in Quantum-Sized TiO2: Correlation between Photoreactivity and Charge Carrier Recombination Dynamics. J. Phys. Chem. 1994, 98, 13669−13679. (19) Pattengale, B.; Ludwig, J.; Huang, J. Atomic Insight into the WDoping Effect on Carrier Dynamics and Photoelectrochemical Properties of BiVO4 Photoanodes. J. Phys. Chem. C 2016, 120, 1421−1427. (20) Park, H. S.; Kweon, K. E.; Ye, H.; Paek, E.; Hwang, G. S.; Bard, A. J. Factors in the Metal Doping of BiVO4 for Improved Photoelectrocatalytic Activity as Studied by Scanning Electrochemical Microscopy and First-Principles Density-Functional Calculation. J. Phys. Chem. C 2011, 115, 17870−17879. (21) Zandi, O.; Klahr, B. M.; Hamann, T. W. Highly Photoactive Tidoped α-Fe2O3 Thin Film Electrodes: Resurrection of the Dead Layer. Energy Environ. Sci. 2013, 6, 634−642. (22) Hisatomi, T.; Dotan, H.; Stefik, M.; Sivula, K.; Rothschild, A.; Grätzel, M.; Mathews, N. Enhancement in the Performance of Ultrathin Hematite Photoanode for Water Splitting by an Oxide Underlayer. Adv. Mater. 2012, 24, 2699−2702. (23) Hofmann, D. M.; Hofstaetter, A.; Leiter, F.; Zhou, H.; Henecker, F.; Meyer, B. K.; Orlinskii, S. B.; Schmidt, J.; Baranov, P. G. Hydrogen: A Relevant Shallow Donor in Zinc Oxide. Phys. Rev. Lett. 2002, 88, 045504. (24) Ö zgür, Ü .; Alivov, Y. I.; Liu, C.; Teke, A.; Reshchikov, M. A.; Doğan, S.; Avrutin, V.; Cho, S.-J.; Morkoç, H. A Comprehensive Review of ZnO Materials and Devices. J. Appl. Phys. 2005, 98, 041301. (25) Gärtner, W. W. Depletion-Layer Photoeffects in Semiconductors. Phys. Rev. 1959, 116, 84−87. (26) Muth, J. F.; Kolbas, R. M.; Sharma, A. K.; Oktyabrsky, S.; Narayan, J. Excitonic Structure and Absorption Coefficient Measurements of ZnO Single Crystal Epitaxial Films Deposited by Pulsed Laser Deposition. J. Appl. Phys. 1999, 85, 7884−7887. (27) Sivula, K. Metal Oxide Photoelectrodes for Solar Fuel Production, Surface Traps, and Catalysis. J. Phys. Chem. Lett. 2013, 4, 1624−1633. (28) McShane, C. M.; Choi, K.-S. Photocurrent Enhancement of nType Cu2O Electrodes Achieved by Controlling Dendritic Branching Growth. J. Am. Chem. Soc. 2009, 131, 2561−2569. (29) Liu, M.; Nam, C.-Y.; Black, C. T.; Kamcev, J.; Zhang, L. Enhancing Water Splitting Activity and Chemical Stability of Zinc Oxide Nanowire Photoanodes with Ultrathin Titania Shells. J. Phys. Chem. C 2013, 117, 13396−13402. (30) Law, M.; Greene, L. E.; Radenovic, A.; Kuykendall, T.; Liphardt, J.; Yang, P. ZnO−Al2O3 and ZnO−TiO2 Core−Shell Nanowire DyeSensitized Solar Cells. J. Phys. Chem. B 2006, 110, 22652−22663. (31) Appavoo, K.; Liu, M.; Sfeir, M. Y. Role of Size and Defects in Ultrafast Broadband Emission Dynamics of ZnO Nanostructures. Appl. Phys. Lett. 2014, 104, 133101. (32) Appavoo, K.; Sfeir, M. Y. Enhanced Broadband Ultrafast Detection of Ultraviolet Emission Using Optical Kerr Gating. Rev. Sci. Instrum. 2014, 85, 055114. (33) Fabregat-Santiago, F.; Garcia-Belmonte, G.; Mora-Sero, I.; Bisquert, J. Characterization of Nanostructured Hybrid and Organic Solar Cells by Impedance Spectroscopy. Phys. Chem. Chem. Phys. 2011, 13, 9083−9118. (34) Hossain, F. M.; Nishii, J.; Takagi, S.; Ohtomo, A.; Fukumura, T.; Fujioka, H.; Ohno, H.; Koinuma, H.; Kawasaki, M. Modeling and Simulation of Polycrystalline ZnO Thin-film Transistors. J. Appl. Phys. 2003, 94, 7768−7777. (35) Janotti, A.; Van de Walle, C. G. Fundamentals of Zinc Oxide as a Semiconductor. Rep. Prog. Phys. 2009, 72, 126501.

(36) Jeong, H.; Min, K.; Byun, S.; Stanton, C. J.; Reitze, D. H.; Yoo, J. K.; Yi, G. C.; Jho, Y. D. Excitonic Diffusion Dynamics in ZnO. Appl. Phys. Lett. 2012, 100, 092106. (37) Muth, J. F.; Lee, J. H.; Shmagin, I. K.; Kolbas, R. M.; Casey, H. C., Jr.; Keller, B. P.; Mishra, U. K.; DenBaars, S. P. Absorption Coefficient, Energy Gap, Exciton Binding Energy, and Recombination Lifetime of GaN Obtained from Transmission Measurements. Appl. Phys. Lett. 1997, 71, 2572−2574.

4043


Unravelling Photocarrier Dynamics beyond the ... - ACS Publications

Recommend Documents