Introduction
Photoelectrodes of semiconductor oxides have been widely investigated for photoelectrochemical (PEC) water splitting for hydrogen production [1–7]. Recently, great attentions have been attracted for the n-type WO3 photoanode, which has favorable valence band edge (3.0 eV versus normal hydrogen electrode), moderate hole diffusion length (~150 nm) and high electron mobility (~12 cm2·V−1·s−1) [8]. However, its PEC property is always limited by the relatively narrow absorption band (Eg = 2.5−2.8 eV), rapid recombination of photo-generated electron-hole pairs and tardy kinetics of holes [9]. WO3 photoanode is not suitable for efficient PEC water oxidation unless credible methods can be developed to improve its visible light absorption capacity and promote the separation of photo-generated electron-hole pairs [10].
Various strategies have been employed to solve the above mentioned problems, including doping [11], nanostructure engineering, combination of semiconductors with small band gaps and so on [12–14]. For the nanostructured electrodes such as one dimensional nanorods, nanotubes and two dimensional nanoflakes, direct electrical pathways could be obtained for the photo-generated charges, and the diffusion lengths of the minority carriers could also be reduced, which lead to superior charge separation and transportation and thus excellent PEC properties [15]. For example, Amano et al. reported a 3.6 mm-thick WO3 film consisting of perpendicularly oriented crystalline flakes fabricated with hydrothermal method, which exhibited a much larger photocurrent density of ~ 2 mA·cm−2 than the nanocrysalline WO3 thin film at 1.2 V vs. Ag/AgCl electrode in 0.1 mol·L−1 Na2SO4 electrolyte [16]. Although the construction of nanostructured electrodes could enhance the separation of photogenerated carriers, the light absorption range of bare WO3 was limited due to its inherent nature. The design of semiconductor composite was the most widely approach to develop photoelectrocatalytic materials in the past few decades. Among the numerous semiconductor composites, the p-n heterostructured systems with a staggered (Type II) band alignment have drawn much attention due to their enhanced efficient charge separation. In the composite system, materials with small band gaps could be used to broaden its light absorption range [10,17]. BiOX (X= Cl, Br, I) is a promising candidate semiconductor, due to the strong internal electric fields from a layered structure with [Bi2O2] slabs interleaved with double halogen atom slabs along the [0 0 1] direction, leading to increase effective separation of the electron-hole pairs [18]. Among the BiOX semiconductors, BiOI has the smallest band-gap (1.8 eV) which could be motivated under most of the visible light range, with a BiOI modification can effectively enhance the PEC properties of the semiconductor photoelectrodes, such as BiOI/ZnO [19], BiOI/TiO2 [20], BiOI/BiVO4 [21]. However, there are few works focusing on the preparation and property of BiOI/WO3 electrode for PEC water splitting.
In our work, we fabricated WO3 nanoflakes by a seed-mediated hydrothermal method, which were uniformly distributed on the F doped SnO2 substrate. Subsequently, BiOI were deposited through an electrodeposition method, and crossed network structure for the BiOI/WO3 electrode was thus obtained. The as-prepared BiOI/WO3 nanocomposite electrode exhibited a significantly improved photocurrent density in Na2SO4 solution under AM 1.5G illumination. The enhanced activity was mainly ascribed to the decoration of BiOI, which expanded light absorption range and facilitated separation of photo-generated carriers for the BiOI/WO3 electrode.
Experimental section
Preparation of WO3 photoanode
The WO3 nanoflake photoanode was synthesized by hydrothermal reaction with seed-mediating according to our previously reported process [22]. The seed layer was deposited onto F doped SnO2 (FTO) substrate by spin coating followed by annealing. The H2WO4 solution for the spin-coating process was prepared by dissolving H2WO4 (2.5 g) and poly(vinyl alcohol) (1 g) into of H2O2 (30 wt. %, 34 mL). Another H2WO4 solution was prepared by adding H2WO4 (2.5 g) and H2O2 (30 wt. %, 34 mL) into deionized water (50 mL), which was heated at 95°C to form a clear solution and then diluted to a concentration of 0.05 mol·L−1 for the solvothermal process. The WO3 nanoflake film was grown by adding urea (0.02 g), oxalic acid (0.02 g), HCl (6 mol·L−1, 0.5 mL) and H2WO4 solution (0.05 mol·L−1, 3 mL) into acetonitrile (12.5 mL). The mixture was then transferred into a Teflonlined autoclave containing the seeded FTO substrate and then maintained at 180°C for 2 h. The resulting substrate after solvothermal process was washed with deionized water and dried with nitrogen flow, and then annealing at 500°C for 2 h in air to obtain the WO3 photoanode.
Preparation of BiOI/WO3 photoanode
The BiOI/WO3 photoanode was synthesized by an electrodeposition method [23]. In a typical process, a solution containing Bi(NO3)3·5H2O (0.04 mol·L−1) and KI (0.4 mol·L−1) was adjusted to pH 1.7 by adding HNO3, and then 20 mL of absolute ethanol containing p-benzoquinone (0.23 mol·L−1) was added into the above solution with vigorously stirring for a few minutes at room temperature. The WO3 electrode, Ag/AgCl electrode and Pt wire were served as the working electrode, reference electrode and counter electrode, respectively. Then BiOI was electrodeposited at − 0.1 V vsAg/AgCl for different deposition times (5−25 s). The optimal deposition time was 15 s, and the corresponding photocurrent density was 1.21 mA·cm−2 at 1.23 V vs. reversible hydrogen electrode (RHE).
Sample characterization
X-ray diffraction (XRD) patterns of the electrodes were collected on an X-ray diffractometer (X' Pert PRO, PANalytical B.V.) using Cu Ka source radiation (1.540598 Å). Scanning electron microscopy (SEM, Nova NanoSEM 450, FEI) was used to investigate the surface morphologies of the samples. X-Ray photoelectron spectroscopy (XPS, AXIS-ULTRA DLD-600W, Shimdzu) was obtained to investigate chemical states of the samples. Raman spectra were recorded on a Raman spectrometer (LabRAM HR800) using a 532 nm laser as excitation source. UV-visible absorption spectra were obtained using a spectrophotometer (UV-3600, Shimdzu).
Photoelectrochemical measurements
PEC properties were investigated in a standard three-electrode configuration using CHI 630D electrochemical workstation (Chenhua, Shanghai, China). The as-synthesized samples acted as the working electrode, Ag/AgCl electrode and Pt wire as the reference electrode and counter electrode, respectively. 0.5 mol·L−1 of Na2SO4 solution was used as the electrolyte. All potentials were converted to RHE according to the Nernst equation: ERHE = EAg/AgCl + EΘAg/AgCl+ (0.059 × pH), where EΘAg/AgCl is 0.197 V at 25°C. A gas chromatography ((Trustworthy Instrument, GC-2020N, China) with argon as a carrier gas was used for the analysis of the evolved gas. Electrochemical impedance spectroscopy (EIS) measurement was performed using an electrochemical impedance analyzer (CHI-920C) over a frequency range between 100 kHz and 0.1 Hz with an AC voltage magnitude of 5 mV amplitude at a bias potential of 1.23 V (vs. RHE).
Results and discussion
XRD measurements were carried out to determine the crystal phase of the samples. Figure 1(a) shows the XRD patterns of the WO3, BiOI and BiOI/WO3 electrodes. After subtracting the diffraction peaks of FTO substrate, all other peaks in the XRD pattern of BiOI/WO3 electrode can be indexed into tetragonal BiOI (JCPDS No. 00-010-0445) and monoclinic WO3 (JCPDS No. 00-043-1035) [24,25]. The Raman spectra of BiOI, WO3 and BiOI/WO3 were presented in Fig. 1(b). The strong peaks at 85 and 147 cm−1 are assigned to Bi-I vibration of BiOI [21]. The peaks at 715 and 806 cm−1 are ascribed to W-O bending (d) and W-O stretching (n) modes respectively. The peaks at 272 and 326 cm−1 correspond to the bending d (O-W-O) vibrations, and the peak at 135 cm−1 are assigned to the lattice vibrations of WO3 [26].
The SEM images of WO3 and BiOI/WO3 electrodes were depicted in Fig. 2. Figure 2(a) suggests that the pristine WO3 sample displays uniform and vertically aligned nanoflakes. The SEM images in Fig. 2(b) clearly revealed the successful deposition of BiOI sheets, which attached onto the surfaces of WO3 nanoflakes and formed crossed networks. The coating of BiOI caused an obvious color change from pale yellow to orange (digital photographs shown in the inset of Fig. 2), identifying the formation of BiOI/WO3 composite electrode.
The chemical states of the samples are investigated by XPS. Figure 3 shows the high-resolution XPS spectra of BiOI/WO3 electrode. As shown in Fig. 3(a), the two peaks with binding energy of 38.0 and 35.9 eV correspond well with the characteristic W 4f5/2 and W 4f7/2 of W6+ in WO3, respectively [27]. Two strong peaks located at 164.4 and 159.0 eV in Fig. 3(b) are attributed to Bi 4f5/2 and Bi 4f7/2 of Bi3+ in BiOI, which are in agreement with the reported values [19]. The peaks centered at 630.4 and 618.9 eV in Fig. 3(c) are assigned to I 3d3/2 and I 3d5/2, respectively [18]. The O 1s spectrum in Fig. 3(d) could be deconvoluted into three peaks, which are related to Bi-O bonds (530.4 eV) of BiOI, W-O bonds (530.9 eV) of WO3 and O-H bonds (531.9 eV) of the surface-adsorbed water, respectively [28–30].
Figure 4(a) shows the UV-visible absorption spectra of the WO3, BiOI and BiOI/WO3 electrodes. Bare WO3 nanoflakes exhibit intrinsic absorption in the wavelength of 300−450 nm. While, a rather broad absorption range of 300−630 nm was observed for the BiOI/WO3 composite, which was due to the good visible light response of BiOI. Furthermore, the indirect bandgaps of BiOI and WO3 were determined by plotting the square root of the absorption energies against the photon energies. By extrapolating the linear portions to zero (Fig. 4(b)), the band gaps were estimated to be 1.95 and 2.72 eV for BiOI and WO3, respectively, which are consistent with previous reports [31].
Linear sweep voltammogram (LSV) measurements were conducted in 0.5 mol·L−1 Na2SO4 electrolyte to evaluate the PEC response of the samples. As shown in Fig. 5(a), both of the samples show negligible dark current densities. Upon illumination, the BiOI/WO3 composite electrode achieved a photocurrent density of 1.21 mA·cm−2 at 1.23 V vs. RHE, which was 1.81 times as high as that of the pristine WO3 nanoflakes (0.67 mA·cm−2). Figure 5(b) depicted the incident photon to current conversion efficiency (IPCE) spectra of the pristine WO3 and as-prepared BiOI/WO3 electrodes at 1.23 V vs. RHE. As can be seen, the WO3 electrode exhibited significant photocurrent response in the wavelength range of 350−450 nm, which was due to its intrinsic absorption. An obvious red-shift up to approximately 600 nm of the IPCE spectrum was observed for the BiOI/WO3 composite electrode, indicating the visible-light response of BiOI. Furthermore, the IPCE values for the BiOI/WO3 composite electrode were higher than those of bare WO3. In particular, the IPCE values at 405 nm for the WO3 and BiOI/WO3 electrodes reached 7.3% and 22.2%, respectively. These results indicated that the modification of BiOI improved the PEC activity of WO3.
Figure 6(a) shows the photocurrent densities of the obtained electrodes measured at 1.23 V vs. RHE in 0.5 mol·L−1 Na2SO4 electrolyte. The photocurrent of the bare WO3 electrode attenuated quickly at the initial time, and lost 53% of the pristine value within 2000 s. However, the BiOI/WO3 electrode exhibited a smaller attenuation of 28% as compared with that of the bare WO3. Furthermore, the gaseous products of the WO3 and BiOI/WO3 electrodes were measured as a function of time at 1.23 V vs. RHE in 0.5 mol·L−1 Na2SO4 electrolyte. As can be seen in Fig. 6(b), the amounts of H2 and O2 were close to the theoretical amounts determined by integrating the measured photocurrent over time. What’s more, the proportion of evolved H2 and O2 was 2:1, indicating that the observed photocurrent is real water splitting photocurrent, rather than due to photodegradation.
Electrochemical impedance spectroscopy (EIS) measurements were implemented to study the interface charge transport properties of the electrodes. The charge-transfer resistance at the electrode-electrolyte interface can be reflected from the semicircle in the Nyquist plots [32]. As shown in Fig. 7(a), the BiOI/WO3 electrode exhibited a smaller semicircle as compared with that of WO3, and it indicated that the introduction of BiOI led to an efficient separation of the electron-hole pairs for the composite electrode, in accordance with its improved PEC activity. Furthermore, the band energies of BiOI and WO3 were also investigated. The conduction band (CB) and valence band (VB) edge at the point of zero charge for a semiconductor can be determined from the empirical equation [21]:
where EVB is the valence band edge potential, and Ee is the energy of free electrons on the hydrogen scale (~4.5 eV). X and Eg are the electronegativity and band-gap energy of the semiconductor, respectively. The X values of BiOI and WO3 are 5.94 and 6.59 eV. As shown in Fig. 7(b), pristine BiOI and WO3 owned their respective band energies before contact, which were obtained from the above UV-visible absorption spectra. When these two semiconductors are in contact, the photo-generated electrons and holes will be generated both in BiOI and WO3 as the solar light illuminate on the BiOI/WO3. The holes in WO3 diffuse to BiOI, and caused efficient separation for the photo-generated electrons and holes. The recombination process is thus remarkably suppressed, leading to improved PEC water splitting activity.
Conclusion
In summary, we have successfully fabricated a WO3 nanoflake electrode by a seed-mediated solvothermal method, followed by loading with BiOI through electrodepositing. The novel BiOI/WO3 nanocomposite electrode achieved an enhanced PEC activity for water splitting, which may be attributed to the expansion of light absorption range as well as the facilitated separation of photo-generated carriers. These findings prove that the BiOI/WO3 photoanode is a promising candidate for PEC water oxidation.