Rational Design of 3D Hierarchical Ternary SnO₂/TiO₂/BiVO₄ Arrays Photoanode toward Efficient Photoelectrochemical Performance

Abstract

BiVO₄ as a promising semiconductor absorber is widely investigated as photoanode in photoelectrochemical water splitting. Herein, the rational design of 3D hierarchical ternary SnO₂/TiO₂/BiVO₄ arrays is reported as photoanode for photoelectrochemical application, in which the SnO₂ hierarchically hollow microspheres core/nanosheets shell arrays act as conductive skeletons, while the sandwiched TiO₂ and surface BiVO₄ are working as hole blocking layer and light absorber, respectively. Arising to the hierarchically ordered structure and synergistic effect between each component in the composite, the ternary SnO₂/TiO₂/BiVO₄ photoanode enables high light harvesting efficiency as well as enhanced charge transport and separation efficiency, yielding a maximum photocurrent density of ≈5.03 mA cm⁻² for sulfite oxidation and ≈3.1 mA cm⁻² for water oxidation, respectively, measured at 1.23 V versus reversible hydrogen electrode under simulated air mass (AM) 1.5 solar light illumination. The results reveal that electrode design and interface engineering play important roles on the overall PEC performance.

A novel 3D hierarchical ternary SnO₂/TiO₂/BiVO₄ arrays photoanode is designed with excellent photoelectrochemical performance.

1. Introduction

Photoelectrochemical (PEC) water splitting that converts the solar light into hydrogen represents a green and sustainable way to address the energy crisis and environmental pollution issues simultaneously.1, 2, 3, 4 The overall PEC conversion efficiency is determined by light harvesting efficiency, charge separation efficiency, and surface reaction efficiency. As such, development of a suitable and efficient photoanode with high light absorption and charge separation efficiency is a key step to boost the PEC performance. In the past decades, metal oxides such as TiO₂,5, 6 Fe₂O₃,7, 8, 9 and BiVO₄ 10, 11, 12 have been widely investigated as photoanodes for PEC water splitting due to the characteristics of excellent photoactivity, low cost, and good stability. Among them, BiVO₄ semiconductor is particularly attractive because it has a narrow band gap of ≈2.4 eV, suitable conduction band (0 V vs reversible hydrogen electrode, RHE) and valence band edges, high water splitting efficiency, and good chemical stability.13, 14, 15, 16, 17, 18 Nevertheless, the poor charge transport ability, short carrier diffusion lengths (≈70 nm), and high charge recombination rates of BiVO₄ greatly limit its practical PEC performance. To address these problems, various strategies such as morphology design, element doping, host–guest heterojunctions and surface modifications are developed to enhance the performance.19, 20, 21, 22, 23, 24 Nanostructured electrode design is one effective route to enhance the light absorption and charge transport ability. For instance, 1D nanorods,25 2D nanosheets,26 3D inverse opals,27 3D hierarchical structures,28, 29, 30 and 3D brochosomes‐like arrays31 have been reported to enhance the light harvesting and charge transport performance. Besides, the host–guest heterojunction electrode using two or more dissimilar materials to take the task of charge transport and light absorption, respectively, is another popular strategy to promote the charge transport and suppress the recombination of electron–hole pairs.32 For instance, BiVO₄ as a light absorber has been combined with various other semiconductors to form a type II configuration such as TiO₂/BiVO₄,33, 34, 35, 36, 37, 38 WO₃/BiVO₄,39, 40, 41, 42 SnO₂/BiVO₄,43, 44 and Fe₂O₃/BiVO₄.45 Among them, TiO₂ is very attractive due to its relatively negative flat band potential and good chemical stability. Nevertheless, the intrinsically low mobility of TiO₂ still greatly limits the overall performance of TiO₂/BiVO₄ heterojunction photoanode.

Herein, we report the rational design and fabrication of 3D hierarchical ternary SnO₂/TiO₂/BiVO₄ arrays as photoanode for photoelectrochemical water splitting by combining multiple routes of colloid microspheres template, hydrothermal, atomic layer deposition (ALD), and electrodeposition. In this electrode design, the hierarchically hollow SnO₂ microspheres@nanosheets arrays act as skeletons to support the TiO₂ and BiVO₄ layer ensuring fast and direct charge transport, the medium ALD conformal TiO₂ layer works as a hole blocking layer to form a type II heterojunction with BiVO₄, which is beneficial to reduce the charge recombination and improve the charge separation efficiency. Moreover, the 3D hierarchically ordered structures with large specific area and rich voids can further enhance the light harvesting and charge transport. As such, the hierarchical ternary SnO₂/TiO₂/BiVO₄ arrays photoanode yields excellent PEC performance with a maximum photocurrent density of ≈5.03 mA cm⁻² at 1.23 V versus RHE under air mass 1.5 light illumination, which is much superior to the counterparts of SnO₂/TiO₂/BiVO₄ microspheres arrays and nanosheets arrays electrodes.

2. Results and Discussion

The 3D SnO₂/TiO₂/BiVO₄ hierarchical nanosheets@hollow microspheres (H‐NSs@HMs) arrays are fabricated by a combination of colloid spheres template, atomic layer deposition, hydrothermal growth, and electro deposition routes, as schematically illustrated in Figure 1 a. First, periodically ordered SnO₂ hollow microsphere arrays were prepared on F‐doped SnO₂ (FTO) substrate by colloid microsphere templates, ALD, and calcination processes. Noting that a thin ZnO layer was predeposited to protect the hexagonal closed‐packed polymer microspheres before the SnO₂ deposition, as shown in Figure S1 in the Supporting Information. Then SnO₂ nanosheets were prepared on the surfaces of each SnO₂ microsphere by hydrothermal growth of SnS₂ nanosheets and a subsequent thermal treatment in air to form hierarchical SnO₂ hollow spheres core/nanosheets shell arrays (SnO₂ H‐NSs/HMs). After that, one thin TiO₂ layer was conformally deposited on the SnO₂ surface by ALD. Finally, BiVO₄ was deposited by an electrodeposition route, resulting in 3D SnO₂/TiO₂/BiVO₄ H‐NSs@HMs.

Figure 1.

a) Scheme of the preparation processes of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. SEM images of SnO₂ H‐NSs@HMs arrays: b) top view and c) cross‐section view. SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays: d,e) top view and f) cross‐section view.

Figure 1b shows top view of the SnO₂ H‐NSs@HMs arrays. It can be seen that a number of ultrathin SnO₂ nanosheets are uniformly assembled on each hollow microsphere, forming a hierarchical hollow nanosheets@microspheres core–shell structure with diameters of ≈4.5 µm. Noting that the hierarchically hollow microspheres are periodically distributed on the substrate, which would be beneficial to the light harvesting and charge transport. From the cross‐section view in Figure 1c, it can be observed that the thickness of individual nanosheet is ranging from 5 to 30 nm, and the whole nanosheets layer thickness is ≈667 nm. After the deposition of TiO₂ and BiVO₄, the ordered structure is still well preserved, as displayed in Figure 1d, indicating the conformal thin films deposition by ALD and electrodeposition, respectively. From a closer observation in Figure 1e,f, the nanosheet surfaces become more rough after the deposition of BiVO₄ and the nanosheet thickness increased obviously. The BiVO₄ layer thickness can be determined to be ≈20 nm. It is noteworthy that the BiVO₄ films are consisted of lots of large grains with diameters from 30 to 50 nm, which are advantageous for further improving the light absorption and increasing the specific surface area. The mapping results and energy dispersive X‐ray spectrum (EDX) in Figures S2 and S3 in the Supporting Information recorded from the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs prove the existence and uniform elements distributions of Sn, Ti, Bi, V, and O. For comparison, SnO₂/TiO₂/BiVO₄ nanosheets arrays, SnO₂/TiO₂/BiVO₄ hollow microspheres arrays, and SnO₂/TiO₂/BiVO₄ hierarchical hollow nanosheets@microspheres arrays with different diameters are also prepared, the corresponding scanning electron microscopy (SEM) images are provided in Figures S4–S6 in the Supporting Information, respectively.

The transmission electron microscopy (TEM) image of the SnO₂/TiO₂ nanosheet in Figure 2 a confirms that TiO₂ are uniformly covered on the SnO₂ nanosheet, suggesting the uniform and conformal film deposition by ALD. It can be seen that the nanosheet is consisted of nanoparticles with lots of mesopores. The mesoporous structures would be beneficial to increase the interface area and facilitate the electrolyte infiltration. The measured spacing of 0.26 and 0.35 nm in Figure 2b can be assigned to tetragonal phase of SnO₂ and anatase phase of TiO₂, respectively. After the BiVO₄ deposition, a number of nanoplates are uniformly distributed on the SnO₂/TiO₂ nanosheet, as shown in Figure 2c, which is in agreement of the SEM observations. The lattice spacings of 0.47, 0.29, 0.26, and 0.35 nm in Figure 2d correspond to the (110), (040) planes of monoclinic BiVO₄, (101) planes of tetragonal SnO₂, and (101) planes of anatase TiO₂, respectively.

Figure 2.

TEM image and high‐resolution TEM images of a,b) SnO₂/TiO₂ H‐NSs@HMs arrays and c,d) SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays.

The crystal structure of SnO₂ H‐NSs@HMs and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs were analyzed by X‐ray diffraction (XRD). As displayed in Figure 3 a, the peaks of 26.41°, 34.06°, 37.88°, 51.67°, 61.87°, and 65.94° correspond to (110), (101), (200), (211), (310), and (301) planes of tetragonal phase of SnO₂ (Joint Committee on Powder Diffraction Standards (JCPDS) No. 41‐1445), while the other peaks can be assigned to monoclinic BiVO₄ phase (JCPDS No. 14‐0688). The chemical composition and surface states of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs was characterized by X‐ray photoelectron spectrum (XPS). The full survey spectrum in Figure S7 in the Supporting Information records the V, Sn, Ti, Bi, and O elements. The high‐resolution Sn element XPS spectra in Figure 3b show two peaks at 486.7 and 495.1 eV respectively, corresponding to Sn⁴⁺ in SnO₂. The two satellite peaks of the high‐resolution Ti 2p_3/2 and Ti 2p_1/2 in Figure 3c centered at binding energies 458.5 and 464.6 eV are typical characteristic Ti⁴⁺ binding in TiO₂. The signal at Ti 2p_1/2 is asymmetric and has shoulders on the lower energy side, which might be related to the combination of BiVO₄ and SnO₂.46, 47, 48 The high‐resolution Bi, V, and O spectra can help identify the built‐in potential in the interface. The binding energy located at 159.1 and 164.5 eV in Figure 3d correspond to the Bi 4f_7/2 orbit and Bi 4f_5/2 orbit, respectively. The splitting peaks at 512.0 and 516.7 eV in Figure 3e correspond to V 2p_3/2 orbit and V 2p_1/2 orbit, respectively.49, 50, 51, 52 The high‐resolution O 1s XPS spectra in Figure 3f show two peaks, one has the binding energy at 530.0 eV is attributed to the lattice oxygen of the Sn—O, Ti—O, Bi—O, and V—O, and the other peak at 532.0 eV can be assigned to the hydroxide in O—H and adsorbed oxygen.53, 54 The high‐resolution Fe 2p spectra and Ni 2p spectra results in Figure S8 in the Supporting Information indicated that FeOOH/NiOOH catalyst were successfully deposited on the SnO₂/TiO₂/BiVO₄ hierarchical structures.24

Figure 3.

a) X‐ray diffraction (XRD) pattern of SnO₂ and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. XPS core elemental spectrum recorded from SnO₂/TiO₂/BiVO₄ H‐NSs@HMs: b) Sn 3d, c) Ti 2p, d) Bi 4f, e) V 2p, and f) O 1s.

The light absorption characteristics of as fabricated samples are characterized by diffuse reflectance spectrum and transmittance spectrum. Figure 4 a shows the diffuse reflectance spectrum of SnO₂, SnO₂/TiO₂, SnO₂/TiO₂/BiVO₄, and SnO₂/BiVO₄ H‐NSs@HMs arrays. For the SnO₂ and SnO₂/TiO₂ H‐NSs@HMs arrays, the observed absorption edges of ≈310 and ≈350 nm are corresponding to the bandgap of SnO₂ and TiO₂, respectively. After the deposition of BiVO₄, the absorption edge redshifts to ≈450 nm, which is in agreement of the bandgap absorption of BiVO₄. It is worth noting that the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays display the lowest light reflectance and light transmittance intensity in the measured wavelength of 300–700 nm among all the referenced samples, as shown in Figure 4a and Figure S9 in the Supporting Information, indicating that the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays possess the highest light capturing ability, due to the ordered hierarchy structure as well as synergistic effect between each component in the composite. To demonstrate the omnidirectional light harvesting ability of the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays, the specular reflection spectra at different light incidence angles are measured. As shown in Figure 4b, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays exhibits excellent omnidirectional antireflection characteristics with light reflection intensity below 1% in a wide wavelength range of 300 to 1100 nm. The excellent omnidirectional light harvesting ability could be due to the periodically ordered hierarchy structures that result in multiple light scattering, increased optical paths, and reduced light reflectance.29, 55

Figure 4.

a) UV–vis diffuse reflectance spectra of SnO₂, SnO₂/TiO₂, SnO₂/BiVO₄, SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. b) Specular reflectance spectra of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays at incident angles of 10°, 20°, 30°, 40°, and 50°.

The PEC performance of as‐fabricated samples is evaluated by linear sweep voltammetric (LSV) measurement under AM1.5 simulated light illumination (100 mW cm⁻²). The measurement was conducted in 0.5 m Na₂SO₄ electrolyte with 0.1 m Na₂SO₃ as a hole scavenger due to the fast kinetics of the sulfite oxidation. Figure 5 a shows the comparison LSV curves of three electrodes with different morphology measured in the dark and light conditions of SnO₂/TiO₂/BiVO₄ hollow microspheres (HMs) arrays, SnO₂/TiO₂/BiVO₄ nanosheets (NSs) arrays, and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. It can be seen that all the electrodes exhibit negligible photocurrent density under dark condition. While under light illumination, obvious photocurrent density is observed. The SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays photoanode shows the highest PEC performance with a maximum photocurrent density of 5.03 mA.cm⁻² at 1.23 V versus RHE, which is much higher than that of SnO₂/TiO₂/BiVO₄ hollow microspheres arrays and SnO₂/TiO₂/BiVO₄ nanosheets arrays. Our results are superior to most of the recent reports on BiVO₄ based photoanodes, as listed in Table S1 in the Supporting Information. The significantly improved PEC performance could be due to the hierarchically ordered structure that provides large specific surface area and efficient interface contact, resulting in improved light absorption and charge separation efficiency. To further illustrate the important role of TiO₂ layer on the PEC performance, SnO₂/TiO₂/BiVO₄ H‐NSs@HMs samples with different TiO₂ layer thickness were prepared. As shown in Figure 5b, all the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays photoanodes exhibit much better PEC performance in contrast to that of SnO₂/BiVO₄ H‐NSs@HMs arrays. The greatly improved performance might be due to the important role of TiO₂ acting as a hole blocking layer to improve the charge separation efficiency and reduce the recombination rates. Noting that the optimized TiO₂ layer thickness is ≈5 nm because a thicker TiO₂ layer would affect the charge transport and reduce the PEC performance. Moreover, the effect of microspheres sizes and BiVO₄ layer thickness on the PEC performance of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays photoanode was also investigated. As shown in Figure S10 in the Supporting Information, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays electrode with diameters of ≈4.5 µm and moderate BiVO₄ layer thickness display the optimized performance. Figure 5c shows the LSV curves measured under light on/off cycles, all the photoanodes show fast photoresponse and recovery time. The photoelectrochemical sulfite oxidation efficiency can be expressed as Equation (1)

$$J_{\mathrm{sulfite}}=J_{\max }\times {\eta }_{\mathrm{abs}}\times {\eta }_{\mathrm{sep}}\times {\eta }_{\mathrm{trans}}$$ (1)

where J _sulftie is defined as the photocurrent density of sulfite oxidation, J _max is defined the maximum photocurrent density, η_abs is the light absorption efficiency, η_sep is the charge separation efficiency, and η_{tran s} is the charge transfer efficiency. Assuming that η_trans is ≈100% at 1.23 V versus RHE in the presence of hole scavengers, the η_abs × η_sep  = J _sulftie/J _max. In our system, the J _max is calculated to be ≈6.67 mA.cm⁻² by using photon energy and solar photon flux at 300–515 nm,14, 56 the calculation details are provided in the Supporting information. The η_abs is determined by the light harvesting efficiency (LHE), i.e., LHE (%) = 100% − R(λ) (%) − T(λ) (%), as plotted in Figure 5d. Hence, the η_sep can be determined. As shown in Figure 5e, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays electrode shows the highest charge separation efficiency of 84.9% in contrast to that of SnO₂/BiVO₄ H‐NSs@HMs arrays, SnO₂/TiO₂/BiVO₄ HMs arrays and SnO₂/TiO₂/BiVO₄ NSs arrays. Moreover, the PEC water splitting performance SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays with and without FeOOH/NiOOH catalyst were measured in 0.5 m Na₂SO₄ electrolyte. As shown in Figure 5f, the FeOOH/NiOOH‐modified SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays photoanode shows significantly enhanced performance, yielding a maximum photocurrent density value of ≈3.1 mA cm⁻² at 1.23 V versus RHE. This might be due to the role of the FeOOH and NiOOH dual water oxidation catalyst, which could facilitate the hole transfer efficiency and improve the water oxidation kinetics.

Figure 5.

LSV curves measured under simulated light and dark conditions in 0.5 m Na₂SO₄ solution with 0.1 m Na₂SO₃ a) SnO₂/TiO₂/BiVO₄ HMs, SnO₂/TiO₂/BiVO₄ NSs, and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. b) Comparison of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays with different TiO₂ layer thickness. c) Chopped photocurrent‐potential curves of SnO₂/TiO₂/BiVO₄ HMs, SnO₂/TiO₂/BiVO₄ NSs, and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. d,e) Light harvesting efficiency (LHE) and charge separation efficiency (η_sep) of SnO₂/TiO₂/BiVO₄ HMs, SnO₂/TiO₂/BiVO₄ NSs, SnO₂/BiVO₄ H‐NSs@HMs arrays, and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. f) PEC water splitting performance of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays with and without FeOOH/NiOOH water oxidation catalyst.

In order to study the wavelength‐dependent photoactivity of as‐fabricated photoanodes, the incident photoelectric conversion efficiency (IPCE) tests were conducted in a three‐electrode system at 1.23 V versus RHE. The IPCE can be calculated by the following formula (2)

$$\mathrm{IPCE}=\frac{1024I}{\lambda J_{\mathrm{light}}}$$ (2)

where I is the photocurrent density, λ is the wavelength of the incident light, and J _light represents the incident light power intensity. As displayed in Figure 6 a, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays electrode displays the highest IPCE values in the wavelength of 325–475 nm in contrast to that of the other two reference samples and a maximum IPCE value of ≈85% was reached at 350 nm. The improved IPCE results of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays could be due to its high light absorption and charge separation efficiency.57, 58, 59 To exclude the light absorption effect in the three different electrodes, the absorption photoelectric conversion efficiency (APCE) is used to determine quantum efficiency based on the IPCE and UV–vis absorption spectra results.31 As shown in Figure 6b, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs also yield the highest APCE efficiency in the wavelength range of 300–500 nm. This would be attributed to the ordered structure that provides large interface contact area, facilitating the charge transport, separation, and collection process.

Figure 6.

a) IPCE spectra and b) spectra APCE of SnO₂/TiO₂/BiVO₄ hollow microspheres arrays, SnO₂/TiO₂/BiVO₄ nanosheets arrays and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays collected at 1.23 V versus RHE.

In order to understand the charge transfer process of SnO₂/TiO₂/BiVO₄ electrodes, electrochemical impedance spectroscopy (EIS) measurements on the three different structures were conducted under open circuit with AM1.5 simulated light illumination. As shown in the EIS spectra in Figure 7 a, the semicircle radius in the middle frequency range represents the charge transfer resistance at the photoelectrode/electrolyte interface, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs photoanode shows the lowest charge transfer resistance in comparison to the counterparts, suggesting that the hierarchically ordered electrode has improved charge transport and separation efficiency. Besides, the SnO₂/TiO₂/BiVO₄ electrode shows lower charge transfer resistance in contrast to that of SnO₂/TiO₂ and SnO₂/BiVO₄ samples (as shown in Figure S12, Supporting Information), indicating that the TiO₂ layer is beneficial to facilitate the charge transfer process.

Figure 7.

a) EIS spectra of SnO₂/TiO₂/BiVO₄ hollow microspheres arrays, SnO₂/TiO₂/BiVO₄ nanosheets arrays, and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays measured under open circuit with light illumination. b,c) Open circuit voltage decay (OCVD) and electron lifetime of SnO₂/TiO₂/BiVO₄ HMs, SnO₂/TiO₂/BiVO₄ NSs, and SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays. Mott–Schottky plots measured in the dark at frequency of 500 Hz: d) SnO₂, e) planar TiO₂ films, and f) BiVO₄ films.

The minority carrier lifetime (τ) is an important parameter affecting the overall conversion efficiency of semiconductor electrodes. The recombination kinetics of as‐fabricated SnO₂/TiO₂/BiVO₄ samples were studied by open circuit voltage decay (OCVD) route, in which the photovoltage decay rate is directly related to the electron lifetime, and the lifetime of the injected electron can be obtained according to Equation (3), 60

$${\tau }_{\mathrm{n}}=\frac{k_{\mathrm{B}}T}{e}{\left(\frac{\mathrm{d}{V}_{\mathrm{oc}}}{\mathrm{d}t}\right)}^{-1}$$ (3)

where k _B is the Boltzmann constant, T is the temperature, e is the electron charge, d _Voc is the amount of change in photovoltage, and dt is the amount of change in time. The k _B T/e is ≈0.026 V at room temperature. Therefore, the lifetime can be determined based on the OCVD curves in Figure 7c. As shown in Figure 7d, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays photoanode show prolonged carrier life time in comparison with the counterparts of SnO₂/TiO₂/BiVO₄ HMs and SnO₂/TiO₂/BiVO₄ NSs, resulting in the reduced recombination rates and improved charge separation efficiency. Moreover, the SnO₂/TiO₂/BiVO₄ H‐NSs@HMs also have increased carrier life time in comparison to that of SnO₂/TiO₂ and SnO₂/BiVO₄ H‐NSs@HMs arrays, further indicating that the medium TiO₂ layer can facilitate the charge separation and reduce the charge recombination, leading to enhanced PEC performance, as shown in Figure S13 in the Supporting Information.

The flat band potential (V _fb) is measured by the open circuit potential of the photocurrent. The flat band potential V _fb of SnO₂, TiO₂, and BiVO₄ in the composite can be determined by the Mott–Schottky curve as Equation (4)

$$1\mathrm{/}{c}^2=\left(\frac{2}{e\epsilon {\epsilon }_0{N}_{\mathrm{d}}}\right)\left(\frac{V_{\mathrm{a}}-V_{\mathrm{fb}}-K_{\mathrm{B}}T}{e}\right)$$ (4)

where C is the capacitance of the space charge layer, e is the electron charge, ε is the dielectric constant, ε₀ is the dielectric constant of the vacuum, N _d is the density of the electron donor, V _a is the applied potential, V _fb is the flat potential, K _B is the Boltzmann constant, and T is temperature. The V _fb is determined by a linear fit of the x intercept as function of 1/c². As shown in Figure 7d–f, the V _fb of SnO₂, TiO₂, and BiVO₄ are determined to be 0.27, 0.14, and 0.12 V (vs RHE), respectively. Assuming that the gap between the conduction band bottom edge and V _fb can be ignored, the difference in the V _fb values confirms the feasible electron transfer from BiVO₄ to TiO_2.

To further determine the valence band positions, ultraviolet photoelectron spectroscopy (UPS) measurements on SnO₂, TiO_2, and BiVO₄ were conducted. The valence band edges were calculated using the photon energy of He I (21.22 eV) and the low kinetic energy cutoff of the spectrum and the valence band maximum with respect to the Fermi level (E _F).61 As shown in Figure S14a in the Supporting Information, the valence band positions of SnO₂, TiO₂, and BiVO₄ are found to be 7.93, 7.58, and 6.93 eV versus the E _vac or 3.49, 3.14, and 2.49 eV on the RHE.

The electronic transfer mechanism can be illustrated in Figure 8 . The band alignment was confirmed by the optical bandgap from the UV–vis absorption spectra (Figure S14b, Supporting Information) and the valence band edges measured by UPS. In the SnO₂/TiO₂/BiVO₄ heterojunctions system, the TiO₂/BiVO₄ and TiO₂/SnO₂ build cascade structured band alignment, which significantly improve the charge separation and reduce the charge recombination at the interface.61, 62, 63 Under light excitation, both BiVO₄ and TiO₂ can absorb photons and generate electron–hole pairs. The photogenerated electrons in BiVO₄ facilitated to transfer to the conduction band of TiO₂ due to the type II configuration and then transported by the conductive SnO₂ skeletons.64 The significantly enhanced PEC performance can be due to the following reasons. First, the hierarchically ordered hollow microspheres/nanosheets arrays structure can provide large specific area and rich porosity, as result in improved light harvesting efficiency. Second, the type II heterojunction and excellent contact between the TiO₂ and BiVO₄ facilitate the charge separation efficiency. Third, the enhanced charge injection and charge transfer process in the electrode/electrolyte interface of ordered SnO₂/TiO₂/BiVO₄ also contribute the improved charge transport and separation. The last but not the least, the synergistic effect of each component in the ternary composite leads to the improved overall performance.

Figure 8.

The electronic transfer mechanism of SnO₂/TiO₂/BiVO₄ H‐NSs@HMs arrays.

3. Conclusions

In summary, we have designed and fabricated a 3D hierarchical ternary SnO₂/TiO₂/BiVO₄ arrays photoanode for photoelectrochemical application. The unique hierarchically ordered electrode design and synergistic effect between each component in the SnO₂/TiO₂/BiVO₄ photoanode led to improved charge transport efficiency, light harvesting efficiency, and charge separation efficiency, as a result in significantly improved PEC performance. Our results demonstrate that electrode design and interface engineering play important roles to boost the overall PEC performance and would open new opportunities for design of various heterojunction photoelectrodes for solar energy conversion.

4. Experimental Section

Preparation of SnO₂/TiO₂/BiVO₄ Hierarchical Nanosheets@Hollow Microspheres (H‐NSs@HMs) Arrays: The preparation processes SnO₂/TiO₂/BiVO₄ hierarchical nanosheets@hollow microspheres (H‐NSs@HMs) arrays can be divided into three steps. First, SnO₂ H‐NSs@HMs arrays on FTO substrate were prepared by colloid spheres template, ALD, and hydrothermal growth process. Second, TiO₂ films with different thicknesses were deposited on SnO₂ H‐NSs@HMs by ALD. The sample was placed in the ALD apparatus, and TiCl₄ and H₂O were used as Ti and O source, respectively. The temperature was 80 °C and the pressure was 11 × 10³ Pa. Finally, on layer of BiVO₄ was deposited on the surface by electrochemical deposition route35 and the thickness was controlled by deposition times. For comparison, SnO₂/TiO₂/BiVO₄ hollow microspheres array and SnO₂/TiO₂/BiVO₄ nanosheets arrays are also prepared. The experiment details are described in the Supporting information.

Characterizations: The surface morphology of the samples was observed using field electron and ion company (FEI) Sirion 200 field emission SEM. The microstructure of the sample was characterized using JEM 2014F TEM. The phase analysis was carried out by XRD (D8‐Advance, Bruker Miller). The content of each component was determined by energy dispersive X‐ray Spectroscopy (EDX) of an X‐max 80 (Oxford) model. X‐ray photoelectron spectroscopy (Kratos Axis Ultra delay line detector (DLD)) characterizes elemental composition and valence state. Ultraviolet photoelectron spectroscopy measurements were conducted out on an ThermoFisher ESCALAB 250 XI spectrometer with a He discharge lamp (He I line, 21.22 eV) as the excitation source and a sample bias of −5 V was applied. The diffuse reflectance spectra were obtained on ultraviolet—see spectrophotometer (Hatachi, U3900H) and Specular reflectance spectra were recorded on Lambda 950 spectrophotometer.

PEC Measurements: The PEC performance of the samples was measured by electrochemical workstation (CS350, CS instrument) with simulated AM1.5G (100 mW cm⁻²) light illumination. The samples, Ag/AgCl (3M KCl), and Pt foil were employed as working electrodes, reference electrode, and counter electrode, respectively. The electrolyte was 0.5 m Na₂SO₄ mixed solution with 0.1 m Na₂SO₃ as hole scavenger. The measured potentials was converted from Ag/AgCl to the RHE by the following formula (5)

$$E_{\mathrm{RHE}}=E_{\mathrm{Ag/AgCl}}+0.059\mathrm{pH}+0.197$$ (5)

E _RHE is the potential of the RHE, E _Ag/AgCl is the potential of the reference electrode Ag/AgCl, and pH is the pH value of the electrolyte. Simulated sunlight with 100 mW cm⁻² was offered by a solar simulator (Zolix SS150). EIS spectra were measured in the frequency range of 10⁻¹ to 10⁵ Hz under a 10 mV amplitude. IPCE values were obtained by IPCE system (Zolix Solar cell Scan100) in the wavelength range of 300–550 nm at the bias of 1.23 V versus RHE.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Pan Q., Li A. S., Zhang Y. L., Yang Y. P., Cheng C. W., Rational Design of 3D Hierarchical Ternary SnO₂/TiO₂/BiVO₄ Arrays Photoanode toward Efficient Photoelectrochemical Performance. Adv. Sci. 2020, 7, 1902235 10.1002/advs.201902235