Letter www.acsami.org
Photoanode with Enhanced Performance Achieved by Coating BiVO4 onto ZnO-Templated Sb-Doped SnO2 Nanotube Scaffold Lite Zhou,†,‡ Yang Yang,†,‡ Jing Zhang,†,‡ and Pratap M. Rao*,†,‡ †
Department of Mechanical Engineering and ‡Materials Science and Engineering Graduate Program, Worcester Polytechnic Institute, Worcester, Massachusetts 01609, United States S Supporting Information *
ABSTRACT: The performance of BiVO4 photoanodes, especially under front-side illumination, is limited by the modest charge transport properties of BiVO4. Core/shell nanostructures consisting of BiVO4 coated onto a conductive scaffold are a promising route to improving the performance of BiVO4-based photoanodes. Here, we investigate photoanodes composed of thin and uniform layers of BiVO4 particles coated onto Sb-doped SnO2 (Sb:SnO2) nanotube arrays that were synthesized using a sacrificial ZnO template with controllable length and packing density. We demonstrate a new record for the product of light absorption and charge separation efficiencies (ηabs × ηsep) of ∼57.3 and 58.5% under front- and back-side illumination, respectively, at 0.6 VRHE. Moreover, both of these high ηabs × ηsep efficiencies are achieved without any extra treatment or intentional doping in BiVO4. These results indicate that integration of Sb:SnO2 nanotube cores with other successful strategies such as doping and hydrogen treatment can increase the performance of BiVO4 and related semiconductors closer to their theoretical potential. KEYWORDS: BiVO4, Sb:SnO2, photoelectrochemical, photoanode, core/shell nanotube
P
low product of light absorption and charge separation efficiencies (ηabs × ηsep).11 The main bottleneck is the relatively short electron diffusion length, reported to range from 10 to 70 nm, which is significantly shorter than the light absorption depth of ∼250 nm, whereas the hole diffusion length of ∼100 nm is much less of a problem.12−14 Consequently, high efficiencies can usually only be achieved under back-side (transparent conducting current collector-side) illumination, because under front illumination, most of the photoexcited electrons are generated near the BiVO4 surface and need to travel across the entire BiVO4 film to reach the current collector, resulting in large bulk recombination. Although BiVO4 photoanodes with back-side illumination can still be used in tandem PEC systems by using an external wire to connect the photoanode and photocathode, the performance of back-side illuminated BiVO4 photoanodes is limited due to the loss of blue and UV photons (λ < 350 nm) absorbed in the transparent conducting oxide current collector. Front-side illuminated BiVO4 photoanodes have the potential for higher efficiency, and enable wireless tandem PEC cells with photoanode and photocathode in a back-to-back configuration, which offers more system design choices. Many efforts have been made to improve ηabs × ηsep in BiVO4.12−24 For instance, a relatively high value of ηabs × ηsep
hotoelectrochemical (PEC) water splitting for clean and renewable hydrogen production is considered to be a promising means by which to satisfy increasing energy demands and reduce CO2 emissions.1−3 A two-electrode tandem system in which a photoanode and a photocathode are connected in series and separately carry out the oxygen evolution reaction and hydrogen evolution reaction promises to be more efficient than a single photoelectrode system due to better utilization of a larger portion of the solar light.2 Recently, bismuth vanadate (BiVO4) is reported as the top-performing photoanode because of its moderate band gap and good stability in near-neutral aqueous environment.4−6 To successfully construct a tandem PEC system, we must achieve high photocurrent at low voltages by the BiVO4 photoanode because a series-connected tandem system would operate at a voltage between the onset voltages of the photoanode and photocathode currents, at which the current generated by the photoanode is equal to that generated by the photocathode.2 Semiconductors being studied as photocathodes have onset voltages for photocathodic current at around 0.6−0.7 VRHE.7−10 Considering that the photocathodes have smaller band gaps than BiVO4 and are capable of larger photocurrents, 0.6 VRHE can be expected as the operating voltage for a BiVO4 photoanode integrated with a photocathode to create a tandem PEC system for water splitting without external bias. However, the performance of BiVO4 photoanodes at low voltage is still limited by poor carrier separation across the light absorption depth, which results in © XXXX American Chemical Society
Received: January 31, 2017 Accepted: March 22, 2017
A
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Letter
ACS Applied Materials & Interfaces
Figure 1. Synthesis procedure and scanning electron microscopy cross-section images of previous Sb:SnO2/BiVO4 nanorod arrays on FTO substrate (top) and new Sb:SnO2/BiVO4 nanotube arrays on FTO substrate (bottom), obtained using 10 h grown ZnO nanowires as a templates.
(46.8%) was achieved at 0.6 VRHE by electrodeposited nanoporous BiVO4, in which photoexcited holes can easily transfer to the semiconductor-electrolyte interface and drive the oxidation reaction.15 However, the relatively high efficiency was only achieved under back-side (transparent conducting current collector-side) illumination. In our previous study, we investigated photoanodes composed of thin layers of BiVO4 coated onto Sb-doped SnO2 (Sb:SnO2) nanorod-arrays (NRAs). The Sb:SnO2/BiVO4 NRA photoanodes achieved ηabs × ηsep of 51% at 0.6 VRHE under back illumination.19 In this core/shell-structured Sb:SnO2/BiVO4 photoanode, high light absorption efficiency was achieved by large optical path length in the BiVO4 layer along the Sb:SnO2 nanorod axial direction, and separation efficiency was improved because the very thin BiVO4 layer allows facile transport of photoexited electrons to the Sb:SnO2 electron collector core material. Moreover, the high conductivity core material Sb:SnO2 decreases the series resistance of the photoanode compared with undoped SnO2, as we have previously demonstrated.19 In addition, compared with other core materials such as WO3, use of Sb:SnO2 also prevents the uncontrolled doping of cations from the nanorods into the BiVO4, which otherwise decreases the charge separation efficiency in BiVO4 by acting as recombination centers for photoexited electrons.12,25,26 Nonetheless, the high performance was still only achieved under back-side illumination because the synthesis of Sb:SnO2 NRAs had several limitations. Figure 1 (top) shows the SEM cross-section image of previous Sb:SnO2/BiVO4 NRA photoanodes. The packing density of the NRs was too high, so that as the growth time was increased to obtain longer NRs, the space between the NRs became too small. This led to bridging of the BiVO4 coating between nanorods and extra BiVO4 coating accumulated on the top of the nanorods. This nonideal coating ultimately resulted in poor photocurrent for front-side illumination. However, a recent study reported that higher photocurrent under front-side illumination compared to back-side illumination can be achieved by “nanoworm” BiVO4. This is a type of nanoporous structure synthesized by a modified electrodeposition method on ITO substrates, which are much flatter than FTO substrates.6 Another recent study reported that higher front-
side illumination can also achieved by combining Mo doping and H2 treatment on drop-casted BiVO4 films.5 Here, we show that high photocurrent from front illumination can alternatively be achieved using the original BiVO4 drop-casting method (without doping or extra treatment) onto Sb:SnO2 nanotube-arrays on economical FTO substrates. Long Sb:SnO2 nanotube arrays with relatively low packing density were synthesized to provide much larger space for BiVO4 coating compared to the previously grown Sb:SnO2 nanorod arrays. Long length and low packing density of Sb:SnO2 nanotube arrays were achieved by coating Sb:SnO2 onto controllable ZnO nanowire-array templates, followed by dissolving the ZnO. Compared to our previously demonstrated Sb:SnO2 nanorod/BiVO4 photoanodes, the front-side illumination performance of Sb:SnO2/BiVO4 nanotube arrays is highly improved due to the uniform coating of BiVO4 on the nanotubes. The back-side illumination performance of Sb:SnO2/BiVO4 nanotube arrays is also slightly improved due to the minimal bridging of the BiVO4 coating between the nanostructures. These advantages together enable the new Sb:SnO2/BiVO4 nanotube photoanodes to achieve ηabs × ηsep of ∼58.5% under back illumination and ∼57.3% under front illumination at 0.6 VRHE for AM 1.5 G sunlight. To the best of our knowledge, these are the highest reported ηabs × ηsep efficiencies to date at this voltage for any BiVO4 photoanodes. The morphologies and crystal structures of the core/shell Sb:SnO2/BiVO4 nanotubes on FTO substrates were characterized by scanning electron microscopy (SEM, JEOL 7000F, 5 kV) and X-ray diffraction (XRD, PANalytical Empyrean, Cukα, 45 kV, 40 mA). Monoclinic BiVO4 and tetragonal SnO2 are the only phases detected by XRD for the Sb:SnO2/BiVO4 nanotubes grown on FTO substrates (Figure 2). The absence of ZnO patterns in Sb:SnO2 nanotubes/FTO indicates that the ZnO nanowires were dissolved in acetic acid during the etching process and almost no ZnO was present in the Sb:SnO2/BiVO4 nanotube photoanodes. According to SEM images (Figure S1a, d), the ZnO nanowire templates are quasi-aligned to the FTO substrates. After 10 h of growth, the length of ZnO nanowires ranges from 2.5 to 4.5 μm, with an average of 3.5 μm. The top and bottom diameters of the tapered nanowires are about 70 and 140 nm, B
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Letter
ACS Applied Materials & Interfaces
increased from 1.0 to 3.5 μm. Figure 3e shows the structure schematic and energy band diagram of the new Sb:SnO2/ BiVO4 nanotube arrays. The low packing density and long length nanotubes provided much larger space for the BiVO4 coating which prevented the accumulation of BiVO4 coating on the top of nanotubes and prevented the coating from bridging between nanotubes, as it did in the previous Sb:SnO2/BiVO4 nanorod photoanodes. The water photooxidation current of the Sb:SnO2/BiVO4 nanotubes can be expressed as JH2O = Jmax × ηabs × ηsep × ηtrans. Jmax is determined by first converting the spectral irradiance to photon flux, and then integrating the photon flux over all wavelengths below 520 nm, which is the band edge of BiVO4, and multiplying by electron charge. ηabs is determined by integrating the product of photon flux and the light harvesting efficiency (LHE, the fraction of absorbed light) at each wavelength over all wavelengths below 520 nm, multiplying by electron charge, and then dividing by Jmax. The spectral irradiance of the Xe lamp solar simulator measured at the photoanode location, as well as the spectral irradiance of AM 1.5G standard illumination are plotted in Supporting Information Figure S2a, and are measured after the light has passed through the glass wall of the beaker in which the photoelectrochemical measurement is performed. Integration of the spectral irradiance of the solar simulator over the range 300−520 nm gives Jmax of 7.57 mA/cm2, which is 1.7% lower than the Jmax given by the standard AM 1.5G spectrum (7.70 mA/cm2). The wavelength-dependent LHE of the photoanode was measured using an integrating sphere with white-light illumination from the solar simulator. The front- and back-side LHE of the Sb:SnO2/BiVO4 nanotubes and those of a photoanode made by drop-casting the same mass per area of BiVO4 directly onto the FTO current collector to form a porous BiVO4 film (FTO/BiVO4) are plotted in Figure 4a. The ηabs of Sb:SnO2/BiVO4 nanotubes under front- and back-side illumination are 72.1% and 68.9%, and the ηabs of FTO/BiVO4 under front- and back-side illumination are 65.6 and 66.2%. The ηabs of Sb:SnO2/BiVO4 nanotubes under front-side illumination is the highest among these four ηabs because the cone shaped nanotubes act as an antireflective layer.27 The J−V curves of Sb:SnO2/BiVO4 nanotubes for sulfite oxidation under front- and back-side illumination are plotted in Figure 4b. The Sb:SnO2/BiVO4 nanotube photoanodes achieved a photocurrent density for sulfite oxidation (Jsulfite) of 4.11 mA/cm2 under front-side illumination and 4.26 mA/ cm2 under back-side illumination at 0.6 VRHE. The J−V curves of FTO/BiVO4 are also plotted in Figure 4 for comparison. A very high Jfront/Jback ratio (0.96) was achieved by the Sb:SnO2/ BiVO4 nanotube photoanodes compared to the FTO/BiVO4 photoanode (0.30) and previously reported BiVO4 photoanodes (0.54 for nanoporous BiVO4,15 0.62 for nanoporous Mo:BiVO4 with H2 treatment,18 0.69 for nanoporous BiVO4 with N2 treatment,24 and 0.73 for Sb:SnO2/BiVO4 nanorods19). The surface charge transfer efficiency of the photoanode (ηtrans) can be considered as 100% due to the fast oxidation kinetics of sulfite. Therefore, the product of ηabs × ηsep can be calculated from ηabs × ηsep ≈ Jsulfite/Jmax. The ηabs × ηsep−V curve of Sb:SnO2/BiVO4 nanotubes and FTO/BiVO4 under front- and back-side illumination are plotted in Figure 4c. At 0.6 VRHE, the Sb:SnO2/BiVO4 nanotubes achieved ηabs × ηsep products of 54.3% and 56.3% under front- and back-side illumination, which are 9.05 and 2.83 times as high as those of the FTO/ BiVO 4 photoanode. At 1.23 V RHE , the Sb:SnO 2 /BiVO 4
Figure 2. X-ray diffraction of standard ZnO pattern (gray), standard SnO2 pattern (yellow), standard BiVO4 pattern (brown), FTO (black), 10 h grown ZnO nanowires (purple), Sb:SnO2 coated on ZnO nanowires (green), Sb:SnO2 coated on ZnO nanowires after acetic acid etching of the ZnO, forming hollow Sb:SnO2 nanotubes (red), and BiVO4 coated onto the Sb:SnO2 nanotubes to form Sb:SnO2/BiVO4 core/shell nanotubes (blue).
respectively, and the average inter-rod spacing is about 500 nm. After two times drop-casting of 5 layers of Sb:SnO2, annealing, acetic acid etching and DI water washing, the ZnO nanowires were fully dissolved in the acetic acid solution and were washed away from the substrates. However, the Sb:SnO2 coating kept its original shape with slightly increased top (80 nm) and bottom (160 nm) diameters and nanowire-shaped void space inside the nanotubes (Figure S1b and S1e). After 4 layers BiVO4 drop casting and annealing, hemispherical BiVO4 particles are uniformly coated on the Sb:SnO2 nanotubes with an average diameter of about 50 nm and an average thickness of about 25 nm (Figure 3a-3d). This 25 nm thick
Figure 3. Scanning electron microscopy (SEM) top images of Sb:SnO2/BiVO4 nanotube arrays on FTO substrate at (a) low and (b) high magnification. SEM cross-section images of Sb:SnO2/BiVO4 nanotube arrays on FTO substrate at (c) low and (d) high magnification. (e) Structure schematic and energy band diagram of the Sb:SnO2/BiVO4 nanotube arrays. These Sb:SnO2/BiVO4 nanotube arrays were obtained by using 10 h grown ZnO nanowire arrays as templates.
BiVO4 coating is thin enough to allow most of the photoexcited electrons to transfer to the Sb:SnO2 nanotube electron collectors without recombination. Figure 1 shows the synthesis methods and the cross-section SEM images of the difference between our new Sb:SnO2/BiVO4 nanotubes and previous Sb:SnO2/BiVO4 nanorods. By controlling the ZnO seed layer concentration and growth time of ZnO nanowire templates, the intertube spacing of new Sb:SnO2 nanotubes is increased from 170 to 500 nm and the length of the new Sb:SnO2 nanotubes is C
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Letter
ACS Applied Materials & Interfaces
illumination. Figure S3b shows the comparison of ηabs × ηsep product for Sb:SnO2/BiVO4 nanotube photoanodes and undoped SnO2/BiVO4 nanotube photoanodes. The undoped SnO2/BiVO4 nanotube photoanode has lower photocurrent onset voltage, whereas the Sb:SnO2/BiVO4 nanotube photoanode achieved higher photocurrent density at high bias. We believe that the slightly higher photocurrent onset voltage in Sb:SnO2/BiVO4 nanotube photoanode was due to slightly increased recombination of photoexcited charges at the Sb:SnO2/BiVO4 interface. The slightly higher photocurrent density at high bias in Sb:SnO2/BiVO4 nanotube photoanode was because the Sb:SnO2 nanotubes have higher electron conductivity compared to undoped SnO2 nanotubes, which decreased the series resistance in the photoanode.19 The incident photon-to-current efficiency (IPCE) of the Sb:SnO2/BiVO4 nanotubes and FTO/BiVO4 for sulfite oxidation at 0.6 VRHE (Figure 5) was calculated from IPCE =
Figure 4. Optical and photoelectrochemical performance of Sb:SnO2 nanotubes (obtained using 10 h grown ZnO nanowires as templates) coated with 4 layers of BiVO4 (Sb:SnO2/BiVO4 nanotubes), and same-mass porous BiVO4 film deposited directly onto FTO substrate (FTO/BiVO4). For a−d, Sb:SnO2/BiVO4 nanotubes with back-side and front-side illumination are shown in purple and red, respectively, and FTO/BiVO4 with back-side and front-side illumination are shown in blue and green, respectively. (a) Light harvesting efficiency (LHE) and inset showing integrated light absorption efficiency (ηabs). (b) Photocurrent for sulfite oxidation (Jsulfite) measured under simulated AM 1.5G illumination using a 3-electrode configuration in aqueous phosphate buffer (pH 7) with 1 M Na2SO3. Dark currents for Sb:SnO2/BiVO4 nanotubes and FTO/BiVO4 are shown as black and orange dashed lines, respectively. (c) Product of light absorption and charge separation efficiency (ηabs × ηsep) versus potential, with dark currents subtracted. (d) Charge separation efficiency (ηsep) with dark current subtracted.
Figure 5. Incident photon-to-current efficiency (IPCE) of Sb:SnO2/ BiVO4 nanotubes (obtained using 10 h grown ZnO nanowires as templates) and FTO/BiVO4 measured at 0.6 VRHE in pH 7 buffer with 1 M Na2SO3. Sb:SnO2/BiVO4 nanotubes with back-side and front-side illumination are shown in purple and red, respectively, whereas FTO/ BiVO4 with back-side and front-side illumination are shown in blue and green, respectively.
nanotubes achieved ηabs × ηsep products of 65.2% and 68.9% under front- and back-side illumination, which are 7.33 and 2.26 times as high as those of the FTO/BiVO4 photoanode. These large improvements are results of the combination of high absorption efficiency provided by the long absorption path in the BiVO4 coating along the nanotube axial direction and high separation efficiency provided by the thin and uniform BiVO4 particle coating. The ηsep−V curves of Sb:SnO2/BiVO4 nanotubes and FTO/BiVO4 under front- and back-side illumination were calculated by dividing the ηabs × ηsep product by the ηabs value, and are plotted in Figure 4d. The ηsep of Sb:SnO2/BiVO4 nanotube photoanodes reached 75.4 and 81.4% under front- and back-side illumination, which is 49.3 and 16.5% higher than the 50.5 and 69.9% ηsep achieved by the previous Sb:SnO2/BiVO4 nanorod photoanodes at 0.6 VRHE. These improvements can be explained by the larger intertube distance in the Sb:SnO2 nanotubes compared to the inter-rod distance in the previous Sb:SnO2 nanorods, which minimized the bridging of the BiVO4 coating across multiple tubes and prevented the accumulation of the BiVO4 coating on the tops of the nanotubes. For the sake of comparison, the J−V curves of undoped SnO2/BiVO4 nanotubes (in which the SnO2 nanotubes are not doped with Sb) for sulfite oxidation under front- and back-side illumination are plotted in Figure S3a. At 0.6 VRHE, undoped SnO2/BiVO4 nanotube photoanodes achieved 4.01 mA/cm2 under front-side illumination and 4.17 mA/cm2 under back-side
[Jsulfite (mA/cm2) × 1240 (volt × nm)]/[Pmono (mW/cm2) × λ (nm)], where Pmono is the intensity and λ the wavelength of the incident monochromatic light (Figure 5). The spectral irradiance of the monochromatic light measured at the photoanode location is plotted in Figure S2b, and is measured after the light has passed through the glass wall of the beaker in which the photoelectrochemical measurement is performed. The IPCE peak value of Sb:SnO2/BiVO4 nanotubes under front- and back-side illumination is 79.8 and 78.7%, respectively. The photocurrent at 0.6 VRHE obtained by integrating the product of the measured IPCE and the spectral irradiance of the solar simulator under front- and back-side illumination is 4.16 and 4.24 mA/cm2. The photocurrent at 0.6 VRHE obtained by integrating the product of the measured IPCE and the standard AM 1.5 G spectral irradiance under front- and back-side illumination is 4.07 mA/cm2 and 4.23 mA/ cm2. Both these values match closely with the measured frontand back-side illumination photocurrent of 4.11 and 4.26 mA/ cm2 obtained from the J−V curve. The IPCE for the Sb:SnO2/ BiVO4 nanotubes under back-illumination decreases from 400 to 300 nm, but this is mainly due to absorption of light by the FTO/glass substrate. For instance, at 300 nm, only 36.4% of the incident light is transmitted to the Sb:SnO2/BiVO4 nanotubes through the FTO/glass substrate (Figure S4a, b, and d). The IPCE calculated by instead excluding the light absorbed in FTO/glass when determining Pmono, is plotted in Figure S4c, and shows a high value from 300−400 nm, with a D
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Letter
ACS Applied Materials & Interfaces value of 93.5% at 300 nm. The IPCE for the Sb:SnO2/BiVO4 nanotubes under front-illumination also decreases from 400 to 300 nm. However, this is expected since shorter wavelength light will be absorbed closer to the BiVO4 electrolyte interface and further from the Sb:SnO2/BiVO4 interface, resulting in a longer electron transport path and therefore more bulk recombination. We additionally performed experiments to study the influence of Sb:SnO2 nanotube length on the PEC performance, while keeping the total quantity of BiVO4 fixed. Whereas the Sb:SnO2 nanotubes in the Sb:SnO2/BiVO4 nanotube photoanodes mentioned until now were synthesized using ZnO nanowire templates that were grown for 10 h, we additionally synthesized Sb:SnO2/BiVO4 nanotube photoanodes based on ZnO nanowires that were grown for 6, 8, and 12 h (Figure S5) which had average lengths of 1.8, 2.4, and 4.1 μm, respectively, compared to the 3.5 μm average length of the 10 h grown ZnO nanowires. At 0.6 VRHE, we found that the photoanodes obtained using the 6 h grown ZnO nanowires had the highest back-illumination photocurrent of 4.43 mA/cm2, the lowest front-illumination photocurrent of 3.77 mA/cm2, and a ratio of front/back photocurrent of 0.85, which is less than one. On the other hand, the photoanodes obtained using the 12 h grown ZnO nanowires had the lowest backillumination photocurrent of 4.24 mA/cm2, the highest frontillumination photocurrent of 4.34 mA/cm2, and a ratio of front/back photocurrent of 1.02, which is greater than one. To the best of our knowledge, both the back-illuminated photocurrent of 4.43 mA/cm2 and the front-illuminated photocurrent of 4.34 mA/cm2, both obtained under conditions of closely calibrated simulated AM 1.5 G illumination, are records for BiVO4 to date, and represent ηabs × ηsep products of 58.5 and 57.3% based on Jmax of 7.57 mA/cm2 obtained by integrating the spectral irradiance of the solar simulator over the range 300−520 nm. The trend of back-illumination photocurrent decreasing with increasing length of the Sb:SnO2 nanotubes can be explained by increasing recombination at the Sb:SnO2/BiVO4 interface, since the interfacial surface area increases with the increasing length of the nanotubes and since photoexcited charges are generated close to this interface under back-illumination. Conversely, the trend of front-illumination photocurrent increasing with increasing length of the Sb:SnO2 nanotubes can be explained by decreasing bulk recombination, since the thickness of the BiVO4 layer is decreasing. Because we are using ZnO nanowire templates to obtain the photoanodes in this study, we also synthesized photoanodes by depositing and annealing BiVO4 directly onto the ZnO nanowires themselves, without first creating Sb:SnO2 nanotubes (Figure S6), for comparison. The solvent we used to prepare the BiVO4 drop-casting solution is acetic acid, which dissolves ZnO. Therefore, for this experiment, we changed the solvent to 2-methoxy ethanol, which does not damage ZnO. The resulting photoanode had good nanowire and BiVO4 coating morphology but generated a sulfite oxidation photocurrent that was more than an order-of-magnitude smaller than those generated by the Sb:SnO2 nanotube/BiVO4 photoanodes. We expect that this decrease is due to detrimental and uncontrolled doping of Zn into BiVO4 during annealing. We also performed experiments to illustrate the need for etching away the ZnO cores within the Sb:SnO2 nanotubes before deposition of BiVO4. We first prepared ZnO core/Sb:SnO2 shell nanowires, then deposited an acetic acid solution of BiVO4
onto these, and then annealed to obtain crystalline BiVO4 (Figure S7). However, the acetic acid partially dissolved the ZnO, resulting in damaged structure and Zn incorporation into BiVO4. The resulting photocurrents for sulfite oxidation were below 2.7 mA/cm2 at 0.6 VRHE. In addition, we also experimented with depositing a 2-methoxy ethanol solution of BiVO4 onto the ZnO core/Sb:SnO2 shell nanowires, so that the ZnO nanowires would not be damaged by the solvent, followed by annealing (Figure S8). The resulting photoanode had good nanowires and BiVO4 coating morphology but generated photocurrents of less than 3.3 mA/cm2 at 0.6 VRHE. Again, these lower photocurrents compared to those of the Sb:SnO2 nanotube/BiVO4 photoanodes is likely due to Zn doping into the BiVO4 during annealing, although it is not as severe as if the Sb:SnO2 shell was not present. In contrast, the champion Sb:SnO2 nanotube/BiVO4 photoanodes synthesized here were obtained by drop-casting of 5 layers of Sb:SnO2 onto the ZnO nanowires, annealing to crystallize the Sb:SnO2, etching in acetic acid and washing in DI water to remove the ZnO nanowires and form Sb:SnO2 nanotubes (which may contain some Zn due to diffusion from the ZnO nanowires during annealing), drop-casting and annealing an additional 5 layers of Sb:SnO2 (which likely contains very low Zn content), followed by acetic acid and DI water wash, and deposition and annealing of BiVO4, which then has negligible Zn doping. Therefore, removing the ZnO nanowires early in the fabrication process leads to the highest-performing BiVO4 photoanodes. Finally, after coating a layer of NiFe-(oxy)hydroxide/borate (NiFeOx-B) OER catalyst onto the surface of the Sb:SnO2/ BiVO4 nanotubes (obtained using 10 h grown ZnO nanowires) according to a published procedure,6 the Sb:SnO2/BiVO4/ NiFeOx-B photoanode was tested in potassium borate buffer solution at pH 9 in a three-electrode configuration and the photocurrent density was plotted in Figure S9a. At 0.6 VRHE, the Sb:SnO2/BiVO4/NiFeOx-B photoanode achieved 2.33 mA/ cm2 and 2.87 mA/cm2 under front- and back-side illumination, which correspond to 56.7 and 67.2% ηtrans for water photooxidation. At 1.23 VRHE. the Sb:SnO2/BiVO4/NiFe-B photoanode achieved 4.16 mA/cm2 and 5.15 mA/cm2 under frontand back-side illumination which correspond to 83.2 and 98.5% ηtrans for water photo-oxidation. The relatively low photocurrent under front-side illumination is caused in part by the light absorption of NiFeOx-B. The stability test of the Sb:SnO2/ BiVO4/NiFe-B photoanode is plotted in Figure S9b, which shows a stable photocurrent at 0.6 VRHE. In conclusion, by optimizing the length and the packing density of nanotubes, these new Sb:SnO2/BiVO4 nanotubes provide higher charge separation efficiency under both front and back-side illumination. The new Sb:SnO2/BiVO4 nanotube photoanodes achieve ηabs × ηsep of ∼58.5% under back illumination and ∼57.3% under front illumination at 0.6 VRHE for simulated AM 1.5 G sunlight, which are the highest reported values for BiVO4-based photoanodes to date. Moreover, both of these high ηabs × ηsep efficiencies are achieved without any extra treatment or intentional doping in BiVO4, which can improve BiVO4 performance by modifying the BiVO4 band gap and increasing the bulk electrical conductivity.18,24,28−30 Therefore, the performance of these Sb:SnO2/BiVO4 nanotube photoanodes can be further improved in the future by these methods. The Sb:SnO2 nanotubes are promising electron collectors, and can possibly be widely used to improve the PEC performance of other E
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Letter
ACS Applied Materials & Interfaces
conversion efficiency using photon recycling. Nat. Commun. 2016, 7, 11943. (5) Kim, J. H.; Jang, J.-W.; Jo, Y. H.; Abdi, F. F.; Lee, Y. H.; van de Krol, R.; Lee, J. S. Hetero-Type Dual Photoanodes for Unbiased Solar Water Splitting with Extended Light Harvesting. Nat. Commun. 2016, 7, 13380. (6) Kuang, Y.; Jia, Q.; Nishiyama, H.; Yamada, T.; Kudo, A.; Domen, K. A Front-Illuminated Nanostructured Transparent BiVO4 Photoanode for > 2% Efficient Water Splitting. Adv. Energy Mater. 2016, 6, 1501645. (7) Ji, L.; McDaniel, M. D.; Wang, S.; Posadas, A. B.; Li, X.; Huang, H.; Lee, J. C.; Demkov, A. A.; Bard, A. J.; Ekerdt, J. G.; Yu, E. T. A Silicon-based Photocathode for Water Reduction with an Epitaxial SrTiO3 Protection Layer and a Nanostructured Catalyst. Nat. Nanotechnol. 2015, 10, 84−90. (8) Lee, M. H.; Takei, K.; Zhang, J.; Kapadia, R.; Zheng, M.; Chen, Y.-Z.; Nah, J.; Matthews, T. S.; Chueh, Y.-L.; Ager, J. W.; Javey, A. pType InP Nanopillar Photocathodes for Efficient Solar-Driven Hydrogen Production. Angew. Chem., Int. Ed. 2012, 51, 10760−10764. (9) Malizia, M.; Seger, B.; Chorkendorff, I.; Vesborg, P. C. K. Formation of a p-n Heterojunction on GaP Photocathodes for H2 Production Providing an Open-Circuit Voltage of 710 mV. J. Mater. Chem. A 2014, 2, 6847−6853. (10) Bornoz, P.; Abdi, F. F.; Tilley, S. D.; Dam, B.; van de Krol, R.; Graetzel, M.; Sivula, K. A Bismuth Vanadate−Cuprous Oxide Tandem Cell for Overall Solar Water Splitting. J. Phys. Chem. C 2014, 118, 16959−16966. (11) 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. (12) Abdi, F. F.; Savenije, T. J.; May, M. M.; Dam, B.; van de Krol, R. The Origin of Slow Carrier Transport in BiVO4 Thin Film Photoanodes: A Time-Resolved Microwave Conductivity Study. J. Phys. Chem. Lett. 2013, 4, 2752−2757. (13) Seabold, J. A.; Zhu, K.; Neale, N. R. Efficient Solar Photoelectrolysis by Nanoporous Mo:BiVO4 Through Controlled Electron Transport. Phys. Chem. Chem. Phys. 2014, 16, 1121−1131. (14) Zhong, D. K.; Choi, S.; Gamelin, D. R. Near-Complete Suppression of Surface Recombination in Solar Photoelectrolysis by “Co-Pi” Catalyst-Modified W:BiVO4. J. Am. Chem. Soc. 2011, 133, 18370−18377. (15) Kim, T. W.; Choi, K.-S. Nanoporous BiVO4 Photoanodes with Dual-Layer Oxygen Evolution Catalysts for Solar Water Splitting. Science 2014, 343, 990−994. (16) Shi, X.; Choi, I. Y.; Zhang, K.; Kwon, J.; Kim, D. Y.; Lee, J. K.; Oh, S. H.; Kim, J. K.; Park, J. H. Efficient Photoelectrochemical Hydrogen Production from Bismuth Vanadate-Decorated Tungsten Trioxide Helix Nanostructures. Nat. Commun. 2014, 5, 4775. (17) Trzesniewski, B. J.; Smith, W. A. Photocharged BiVO4 Photoanodes for Improved Solar Water Splitting. J. Mater. Chem. A 2016, 4, 2919−2926. (18) Kim, J. H.; Jo, Y.; Kim, J. H.; Jang, J. W.; Kang, H. J.; Lee, Y. H.; Kim, D. S.; Jun, Y.; Lee, J. S. Wireless Solar Water Splitting Device with Robust Cobalt-Catalyzed, Dual-Doped BiVO4 Photoanode and Perovskite Solar Cell in Tandem: A Dual Absorber Artificial Leaf. ACS Nano 2015, 9, 11820−11829. (19) Zhou, L.; Zhao, C.; Giri, B.; Allen, P.; Xu, X.; Joshi, H.; Fan, Y.; Titova, L. V.; Rao, P. M. High Light Absorption and Charge Separation Efficiency at Low Applied Voltage from Sb-Doped SnO2/ BiVO4 Core/Shell Nanorod-Array Photoanodes. Nano Lett. 2016, 16, 3463−3474. (20) Chatchai, P.; Murakami, Y.; Kishioka, S.-y.; Nosaka, A. Y.; Nosaka, Y. FTO/SnO2/BiVO4 Composite Photoelectrode for Water Oxidation under Visible Light Irradiation. Electrochem. Solid-State Lett. 2008, 11, H160−H163.
semiconductors which have similar charge transport limitations as BiVO4.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b01538. SEM top and cross-section images; spectral irradiance of Xe lamp solar simulator and standard AM 1.5G solar spectrum; photoelectrochemical performance of undoped SnO2/BiVO4 nanotubes; transmission of FTO/ glass substrate and estimation of IPCE for backillumination excluding absorption in FTO/glass; photoelectrochemical performance and SEM images of Sb:SnO2/BiVO4 nanotubes as a function of nanotube length; photoelectrochemical performance and SEM images of photoanodes obtained without etching away ZnO nanowires, with and without Sb:SnO2 shells; measurement of the photoelectrochemical stability and photocurrent for water oxidation by Sb:SnO2/BiVO4 nanotubes with the addition of oxygen evolution catalyst; experimental methods; optical images (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Lite Zhou: 0000-0002-9771-4653 Yang Yang: 0000-0003-0221-7046 Jing Zhang: 0000-0001-9122-2669 Pratap M. Rao: 0000-0003-1324-498X Author Contributions
L.Z., Y.Y., and J.Z. carried out the photoanode synthesis. L.Z. performed all the measurements and characterization of the photoanode. L.Z. and P.M.R. conceived the study and wrote the manuscript. All authors discussed the results and commented on the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge support from Worcester Polytechnic Institute (WPI), the Mechanical Engineering Department at WPI, and the Materials Science and Engineering Graduate Program at WPI. This material is based upon work supported by the National Science Foundation under Grant DMR1609538, “SusChEM: Collaborative Research: Novel Nanostructured Metal Oxides for Efficient Solar Energy Conversion−Theory, Synthesis, and Interfacial Carrier Dynamics”
■
REFERENCES
(1) Bensaid, S.; Centi, G.; Garrone, E.; Perathoner, S.; Saracco, G. Towards Artificial Leaves for Solar Hydrogen and Fuels from Carbon Dioxide. ChemSusChem 2012, 5, 500−521. (2) Prévot, M. S.; Sivula, K. Photoelectrochemical Tandem Cells for Solar Water Splitting. J. Phys. Chem. C 2013, 117, 17879−17893. (3) Fujishima, A.; Honda, K. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 1972, 238, 37−38. (4) Shi, X.; Jeong, H.; Oh, S. J.; Ma, M.; Zhang, K.; Kwon, J.; Choi, I. T.; Choi, I. Y.; Kim, H. K.; Kim, J. K.; Park, J. H. Unassisted photoelectrochemical water splitting exceeding 7% solar-to-hydrogen F
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Letter
ACS Applied Materials & Interfaces (21) Liang, Y.; Tsubota, T.; Mooij, L. P. A.; van de Krol, R. Highly Improved Quantum Efficiencies for Thin Film BiVO4 Photoanodes. J. Phys. Chem. C 2011, 115, 17594−17598. (22) Saito, R.; Miseki, Y.; Sayama, K. Highly Efficient Photoelectrochemical Water Splitting Using a Thin Film Photoanode of BiVO4/SnO2/WO3 Multi-Composite in a Carbonate Electrolyte. Chem. Commun. 2012, 48, 3833−3835. (23) Saito, R.; Miseki, Y.; Sayama, K. Photoanode Characteristics of Multi-Layer Composite BiVO4 Thin Film in a Concentrated Carbonate Electrolyte Solution for Water Splitting. J. Photochem. Photobiol., A 2013, 258, 51−60. (24) Kim, T. W.; Ping, Y.; Galli, G. A.; Choi, K.-S. Simultaneous Enhancements in Photon Absorption and Charge Transport of Bismuth Vanadate Photoanodes for Solar Water Splitting. Nat. Commun. 2015, 6, 8769. (25) Luo, W.; Wang, J.; Zhao, X.; Zhao, Z.; Li, Z.; Zou, Z. Formation Energy and Photoelectrochemical Properties of BiVO4 After Doping at Bi3+ or V5+ Sites with Higher Valence Metal Ions. Phys. Chem. Chem. Phys. 2013, 15, 1006−1013. (26) Terrier, C.; Chatelon, J. P.; Roger, J. A. Electrical and Optical Properties of Sb:SnO2 Thin Films Obtained by The Sol-Gel Method. Thin Solid Films 1997, 295, 95−100. (27) Lee, Y.-J.; Ruby, D. S.; Peters, D. W.; McKenzie, B. B.; Hsu, J. W. P. ZnO Nanostructures as Efficient Antireflection Layers in Solar Cells. Nano Lett. 2008, 8, 1501−1505. (28) Cooper, J. K.; Scott, S. B.; Ling, Y.; Yang, J.; Hao, S.; Li, Y.; Toma, F. M.; Stutzmann, M.; Lakshmi, K. V.; Sharp, I. D. Role of Hydrogen in Defining the n-Type Character of BiVO4 Photoanodes. Chem. Mater. 2016, 28, 5761−5771. (29) Chemseddine, A.; Ullrich, K.; Mete, T.; Abdi, F. F.; van de Krol, R. Solution-processed Multilayered BiVO4 Photoanodes: Influence of Intermediate Heat Treatments on the Photoactivity. J. Mater. Chem. A 2016, 4, 1723−1728. (30) Chen, L.; Toma, F. M.; Cooper, J. K.; Lyon, A.; Lin, Y.; Sharp, I. D.; Ager, J. W. Mo-Doped BiVO4 Photoanodes Synthesized by Reactive Sputtering. ChemSusChem 2015, 8, 1066−1071.
G
DOI: 10.1021/acsami.7b01538 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
