English

• Photoelectrochemical (PEC) water splitting is one of the promising strategies to convert solar light into hydrogen to meet the increasing demand for clean energy from renewable sources [1-3]. Although hydrogen is produced from water by the hydrogen evolution reaction (HER) on the cathode, usually a Pt-based electrode [4], the oxygen evolution reaction (OER) on the photoanode is the rate-determining step in PEC water splitting, but this process is sluggish due to the multiple proton-coupled electron transfer [1,5]. Therefore, it is crucial to design and prepare efficient photoanodes with sufficient light absorption, effective charge separation, and fast reaction kinetics for OER [6-8]. Among the different materials, nanostructured bismuth vanadate (BiVO₄) has attracted much attention due to its suitable bandgap (~2.4 eV), low cost, and facile synthesis [9-14]. Nevertheless, its low carrier mobility and short hole diffusion length cause serious bulk and surface electron–hole recombination [15], which causes photo-corrosion of BiVO₄, loss of V⁵⁺, and poor long-term stability [14,16-18]. Doping [19-21], facet tailoring [22-24], and oxygen vacancies construction [15,25-27], have been proposed to overcome the transport limitation and enhance the separation of charge carriers. In addition, coatings made of oxygen evolution catalysts (OECs) and surface passivation layers have been suggested to inhibit the loss of V⁵⁺ and improve the OER activity [9,28-30].

  Among the different types of OECs, Ⅷ metal (Fe, Co, Ni)-based oxides and/or (oxy)hydroxides, especially FeOOH, have attracted interest in recent years [9,31-35]. The Bi–O–Fe interfacial bonds have been observed to be the active sites for the enhanced OER activity and the suppressed dissolution of V⁵⁺ in turn improves the PEC stability [32]. For example, Zhang et al. have prepared ultrathin and crystalline β-FeOOH on BiVO₄ photoanodes, which exhibit remarkable photocurrents because the ultrathin structure of FeOOH exposes more active sites to accelerate the reaction kinetics [33]. However, it is challenging to cover BiVO₄ crystals completely unless a sufficiently thick layer is deposited, but a thick OECs/protector causes unnecessary parasitic light absorption and may compromise the PEC performance of BiVO₄ [2]. Moreover, the surface structure changes under strong alkaline conditions at a high anodic potential consequently hurting the durability [36]. Hence, uniform and conformal deposition of FeOOH with a suitable thickness on BiVO₄ with large aspect ratios remains challenging.

  The composition of electrolytes and modulation of the catalytic environment have gained more recent attention because there is a dynamic ion exchange process between the electrocatalysts and electrolytes, especially in electrochemical cycling [11,37-39]. For example, Lee and Choi have used a V⁵⁺-saturated electrolyte to inhibit the loss of V⁵⁺ from the BiVO₄ lattice by dissolution consequently suppressing photo-corrosion during solar water splitting and increasing the photostability of BiVO₄ [11]. Nevertheless, the addition of V⁵⁺ to the electrolyte cannot boost the OER kinetics. Our recent study reveals that Fe(Ⅲ) ions can serve as the dynamic sites to promote the OER activity and long-term stability by cyclical formation of the intermediates on the electrode/electrolyte interface and in situ growth of FeOOH on the surface of the electrode [40,41].

  Herein, a convenient strategy to simultaneously overcome the sluggish reaction kinetics and poor stability of BiVO₄ is described by incorporating Fe(Ⅲ) cations into the electrolyte. During OER, the Fe(Ⅲ)-containing electrolyte shows dynamic changes between Fe(Ⅲ) and FeOOH (Fe(Ⅲ) → Fe^x+→ Fe–OH →Fe–O → FeOOH → Fe(Ⅲ)) accompanying with FeOOH co-catalyst produced in situ on the BiVO₄ photoanode (Scheme 1). The BiVO₄/FeOOH photoanode shows a current density of 2.21 mA/cm² at 1.23 V (vs. RHE) under AM 1.5G illumination (100 mW/cm²), which is considerably higher than that of the pristine BiVO₄ (1.29 mA/cm²). The excellent PEC properties stem from the in situ generated FeOOH on the surface, which, as the active sites, accelerates oxygen evolution. Furthermore, the PEC stability of the BiVO₄/FeOOH is enhanced due to the in situ formed FeOOH that prevents the corrosion of the photoanode by suppressing the dissolution of V⁵⁺. The results reveal a facile, cost-effective, and efficient strategy for efficient and stable PEC systems.

  Scheme 1

  

  Scheme 1.  Schematic illustration of the BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ).

  DownLoad: Full-Size Img PowerPoint

  The PEC water oxidation properties of the BiVO₄ photoanodes are determined under AM 1.5G illumination, and the current density curves are shown in Fig. 1. Various concentrations of Fe(Ⅲ) ions are added to the KBi electrolyte to examine the effects on the PEC activity. As shown in Fig. S1 (Supporting information), Fe(Ⅲ) ions in the electrolyte dramatically enhances the current densities. The maximum current density of 2.21 mA/cm² is achieved at an onset potential of 0.40 V by adding 1 mmol/L Fe(Ⅲ) to the electrolyte (KBi + Fe(Ⅲ)). Fig. 1a shows the instantaneous current density curves of the BiVO₄ photoanodes in the KBi and KBi + Fe(Ⅲ) electrolytes. In both electrolytes, the dark current densities are negligible, indicating that Fe(Ⅲ) ions do not take part in the redox reaction for OER at below 1.23 V. To explore the stability of the photoanodes, the i-t curves are acquired at 1.23 V. As shown in Fig. 1b, the photoanode shows a rapid decrease in the current densities in the KBi electrolyte from about 1.41 mA/cm² at the beginning to 0.1 mA/cm² after 3 h indicative of serious photo-corrosion and V⁵⁺ dissolution from the crystal lattice [1,11]. Remarkably, the photoanode displays a different trend if Fe(Ⅲ) ions are added to the solution (KBi + Fe(Ⅲ)). The current densities increase in the first 1 h perhaps due to the redox process of Fe(Ⅲ) and deposition of FeOOH on the surface of the photoanode [40-42]. After 1 h, the current density plateaus at 2.21 mA/cm² and remains steady for over 10 h, consistent with the LSV results in Fig. S2a (Supporting information). The i-t curves of FTO without BiVO₄ obtained from the KBi + Fe(Ⅲ) solution under light illumination exhibit negligible current densities (Fig. S2b in Supporting information), indicating that the Fe species are unable to generate photogenerated charge carriers in OER. Inductively coupled plasma (ICP) spectroscopy was used to analyze the electrolyte after 3 h of i-t measurement. The results show the presence of V in both the KBi and KBi + Fe(Ⅲ) electrolytes (Fig. S3 in Supporting information), indicating the dissolution of V in both systems. However, the concentration of V in KBi (78.35 µg/L) is significantly higher than that of the KBi + Fe(Ⅲ) (41.32 µg/L), suggesting the role of Fe(Ⅲ) in suppressing the dissolution of V that ultimately prevents the BiVO₄ photoanode from photo-corrosion. These results indicate that the Fe species plays the role of a co-catalyst in the BiVO₄ photoanode to improve both the PEC activity and stability. The PEC performance of our system reaches and even outperforms most recent photoelectrocatalysts based on OECs decorated BiVO₄ electrodes (Table S1 in Supporting information), while our method is much more convenient.

  Figure 1

  

  Figure 1.  (a) Chopped illumination curves. (b) Photostability of BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ). CV measurements of the BiVO₄ photoanodes in (c) KBi and (d) KBi + Fe(Ⅲ). (e) LSV and (f) i-t curves of the photoanodes for different pretreatment of Fe(Ⅲ).

  DownLoad: Full-Size Img PowerPoint

  The Fe(Ⅲ) catalytic mechanism is explored. CV is performed on the BiVO₄ photoanode in 1.0 mol/L KBi with and without Fe(Ⅲ) addition. As shown in Fig. 1c, the current densities decrease with CV cycles in KBi, implying gradual inactivation of the BiVO₄ photoanode. In contrast, the current densities increase gradually in KBi + Fe(Ⅲ) as shown in Fig. 1d. During CV, the photoanode is cycled between the PEC active and inactive states. The catalyst/electrolyte interface undergoes dynamic changes during PEC, leading to Fe incorporation into the photoanode [43]. To verify the photocurrent of BiVO₄ photocurrent has stabilized, the different CV (1^st, 50^th, 80^th, and 100^th) cycles has been carried out as shown in Fig. S4 (Supporting information). The 80^th and 100^th photocurrent density indicate almost the same, suggesting that 100 cycle is enough for supper saturation of the photocurrent. To confirm that Fe is incorporated into the photoanode rather than absorbing on the surface, we compare the PEC activity of the BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ), Fe(Ⅲ)-chemical absorbed (activated by CV cycles in KBi + Fe(Ⅲ)) in KBi, and Fe(Ⅲ)-physical absorbed (immersed in KBi + Fe(Ⅲ) for the same time as that of CV-activated sample) in KBi. As shown in Fig. 1e, the Fe(Ⅲ)-chemical absorbed and Fe(Ⅲ)-physical absorbed photoanodes display higher current densities than the fresh photoanode in KBi, indicating that Fe(Ⅲ) is vital to improving the PEC activity. Moreover, the photoanode in KBi + Fe(Ⅲ) shows higher current densities than the Fe(Ⅲ)-physical absorbed photoanode in KBi, confirming the incorporation of Fe(Ⅲ) into the photoanode during PEC rather than absorption. In addition, the Fe(Ⅲ)- chemical absorbed anode in KBi exhibited lower activity than that of the photoanode tested in KBi + Fe(Ⅲ), ascribing that Fe(Ⅲ) availability is the key to maintaining the PEC activity, which is further confirmed by the i-t curves in Fig. 1f. To verify the advantages of the proposed strategy, FeOOH is deposited on the surface of the BiVO₄ photoanode (FeOOH/BiVO₄) by photo-deposition using a method reported by Choi et al. [34] and compared its performance with the BiVO₄ photoanode. As shown in Fig. S5 (Supporting information), the current densities of FeOOH/BiVO₄ tested in KBi and BiVO₄ tested in KBi + Fe(Ⅲ) are comparable (Fig. S5a) while the FeOOH/BiVO₄ presents a poor stability (Fig. S5b).

  The applied bias potential efficiency (ABPE) is measured in KBi and KBi + Fe(Ⅲ) and the maximum ABPEs are calculated to be 0.18% (0.92) and 0.55% (0.82), respectively (Fig. 2a), confirming that the addition of Fe(Ⅲ) to the electrolyte can enhance the catalytic ability in water oxidation [44]. Fig. 2b and Fig. S6 (Supporting information) show the interfacial charge transfer (η_trans) efficiency of the water oxidation reaction. In the electrolyte without Fe(Ⅲ), the photoanode exhibits a very low efficiency of 29.1% at 1.23 V, but adding Fe(Ⅲ) increases the η_trans efficiency to 62.65%. Fig. 2c shows the EIS data to further elucidate the interfacial charge transfer and oxygen evolution kinetics. The curvature diameter of the photoanodes in KBi + Fe(Ⅲ) under illumination is the smallest among the samples, providing evidence that Fe(Ⅲ) improves the charge transfer between the photoanode and electrolyte [45].

  Figure 2

  

  Figure 2.  (a) ABPE, (b) charge transfer efficiency, (c) Nyquist plots at 0.8 V under illumination or in the dark for the BiVO₄ photoanodes with or without Fe³⁺. (d) PL spectra, (e) UV–vis absorption spectra, and (f) IPCE at 1.23 V of BiVO₄ photoanodes after the i-t test in KBi or KBi + Fe(Ⅲ).

  DownLoad: Full-Size Img PowerPoint

  Photoluminescence (PL) is monitored to determine the electron-hole recombination rates [46]. As shown in Fig. 2d, all the photoanodes show the maximum intensity at about 420 nm. The photoanode after the i-t test in KBi exhibits stronger intensity than the fresh one, demonstrating higher electron-hole recombination rates. However, after the i-t test in KBi + Fe(Ⅲ), the PL peak intensity declines, implying smaller electron-hole recombination ratios compared with other photoanodes. The results demonstrate that addition of Fe(Ⅲ) to the electrolyte improves the separation efficiency of photo-generated carriers [1,47]. As revealed by the UV–vis diffuse (Fig. 2e), the absorptance does not change significantly. All the photoanodes exhibit a light absorption edge at ~500 nm, corresponding to a bandgap of about 2.40 eV (Fig. S7a in Supporting information). After Fe incorporation, the photoanodes show a similar light absorption edge and bandgap compared to the fresh one, indicating that the Fe species do not affect the bandgap and the light absorption properties of the BiVO₄ photoanodes [48]. There is a slight shift in absorption edges of the BiVO₄ photoanodes after the i-t test in KBi + Fe(Ⅲ) (Fig. S8 in Supporting information), attributing to the formation of FeOOH on the surface of the photoanode. The PEC performance at different incident lights is studied using the IPCE curves of the different photoanodes. As shown in Fig. 2f, all the photoanodes have IPCE values of less than 500 nm, consistent with the UV–vis light absorption curves. The IPCE values of the fresh photoanode and photoanode after the i-t test in KBi and KBi + Fe(Ⅲ) at 420 nm are 11.27%, 8.98%, and 46.53%, respectively. To get more insights into the electronic properties, Mott-Schottky (MS) curves are acquired in KBi without illumination. As shown in Fig. S7b (Supporting information), all the samples exhibit the n-type semiconductor behavior. Therefore, the above experiments confirm that the incorporation of Fe(Ⅲ) improves the charge transfer kinetics rather than changing the band structure.

  The effects of Fe incorporation on the surface chemical states and electronic structures of the photoanodes are examined by XPS (Fig. 3 and Fig. S9 in Supporting information). The survey XPS spectra of the prepared BiVO₄ photoanodes reveal the existence of Bi, V, and O elements. From the high-resolution XPS spectra, the peaks at about 159.0 and 164.7 eV are attributed to Bi³⁺ of Bi 4f (Fig. 3a). The two peaks at about 516.9 and 524.5 eV are ascribed to the V 2p_3/2 and V 2p_1/2 states, respectively, indicative of the V⁵⁺ characteristics (Figs. 3a and b) [49]. The XPS peaks of O 1s can be divided into lattice oxygen (O_L), oxygen vacancy (O_V), and chemically adsorbed or dissociated oxygen species (O_C) from water molecules, respectively (Fig. 3c) [50,51]. It is worth noting that the XPS spectra of Bi 4f and V 2p after the i-t test in KBi + Fe(Ⅲ) show a shift toward lower binding energies, suggesting that the interaction between Fe species and BiVO₄ produces excellent electron transport [29]. Furthermore, the chemical shift of O_L decreases after the i-t test in KBi + Fe(Ⅲ), indicating oxygen vacancies and bigger electron densities around the Bi and V atoms. The peaks at 711.0 and 724.3 eV can be ascribed to Fe 2p_3/2 and Fe 2p_1/2, and those at 713.8 and 734.3 eV are typical satellite (Sat.) peaks (Fig. 3d). The XPS survey spectra corroborate successful incorporation of Fe species into the photoanodes.

  Figure 3

  

  Figure 3.  XPS spectra of (a) Bi 4f, (b) V 2p, and (c) O 1s of the pristine BiVO₄ photoanodes before and after the i-t test with or without Fe(Ⅲ). (d) Fe 2p of the photoanode after the i-t test with Fe(Ⅲ).

  DownLoad: Full-Size Img PowerPoint

  To better understand the roles of Fe species, SEM, XRD, TEM, and Raman scattering are performed. As shown in Fig. 4a, the SEM image exhibits irregular nanoparticles with a length of about 220 nm. The initial nanopores in the films expand and are interconnected with each other to form a three-dimensional (3D) nanoporous structure, which benefits the interactions between the photoanodes and electrolytes. The stability evaluation in KBi shows that the BiVO₄ structure is destroyed and the density of nanopores is less than that of the fresh one (Fig. 4b) reflecting dissolution of BiVO₄. However, the surface of the photoanode tested in KBi + Fe(Ⅲ) retains the original structure but becomes roughing, indicating that the photoanodes can be protected from photo-corrosion by the Fe species during PEC (Fig. 4c).

  Figure 4

  

  Figure 4.  SEM images of (a) pristine BiVO₄ photoanodes, (b) after 5-h i-t test without Fe(Ⅲ), and (c) after i-t test with Fe(Ⅲ). (d) TEM images of pristine BiVO₄ photoanode. (e, f) BiVO₄ photoanode after the i-t test with Fe(Ⅲ). (g) HAADF-STEM image and elemental maps of the BiVO₄ photoanode after the 10-h i-t test with Fe(Ⅲ). The insets in (a-d) are the corresponding high-magnification images.

  DownLoad: Full-Size Img PowerPoint

  The XRD patterns (Fig. S10 in Supporting information) of the BiVO₄ photoanodes show monoclinic BiVO₄ (JCPDS No. 14–688) and the FTO substrate [52]. No additional diffraction peaks can be detected after the i-t test in KBi + Fe(Ⅲ), maybe due to the small concentration of the Fe(Ⅲ) species [44]. As shown in Fig. S11 (Supporting information), the main Raman peaks at 327 and 369 cm⁻¹ correspond to the asymmetric and symmetric deformation modes of the VO₄³⁻ tetrahedron [53]. The peaks at 713 and 829 cm⁻¹ arise from the asymmetric and symmetric stretching modes of the V–O bond [54]. After the i-t test in KBi + Fe(Ⅲ), the photoanodes show the typical peaks at ~447 and 609 cm⁻¹ related to the Fe–O–Fe and Fe–O vibrational mode of FeOOH, respectively [55,56]. Compared to the pristine BiVO₄ photoanode, the photoanode after the i-t test in KBi + Fe(Ⅲ) shows a negative shift and broader peak for the symmetric stretching of V–O (829 cm⁻¹) implying that FeOOH weakens the lattice structure of BiVO₄ [57]. Interestingly, the peak intensity obtained after the i-t test in KBi + Fe(Ⅲ) decreases subtly in comparison with the fresh sample, while the intensity of the photoanode declines noticeably after the i-t test in KBi, which further confirms that the Fe species can prevent photoanodes from photo-corrosion. Raman scattering shows that FeOOH is loaded onto the surface of BiVO₄ photoanode during PEC. TEM conducted on the pristine photoanode shows lattice fringes of 0.308 nm corresponding to the (−121) faces of BiVO₄ (Fig. 4d), consistent with the XRD. The HR-TEM image after the i-t test in KBi + Fe(Ⅲ) reveals in situ generation of ultrathin FeOOH with a thickness of ~5 nm on the monoclinic BiVO₄ photoanode (Fig. 4e). Fig. 4f shows a lattice fringe of 0.254 nm corresponding to the (101) face of FeOOH and the EDS elemental maps disclose uniform distributions of Bi, V, and Fe (Fig. 4g).

  Based on the above data, the FeOOH/BiVO₄ heterostructure is formed after the i-t test. The current densities do not change with i-t testing time. The FeOOH/BiVO₄ heterostructure may be formed in situ even after a short time and Fe(Ⅲ) boosts the OER kinetics by cyclic formation of intermediates at the photoelectrode/electrolyte interface, so the Fe(Ⅲ)-containing atmosphere is dynamic to preserve the photo-stability of the system. The PEC stability of the BiVO₄ photoelectrodes is investigated in KBi and KBi + Fe(Ⅲ). As expected, the photoanode shows obvious decline in the current densities in KBi (Fig. 5a, red plots). After adding Fe(Ⅲ) ions, the current density increases and remains stable at 1.43 mA/cm² rather than 2.21 mA/cm², suggesting partial dissolution of BiVO₄ and that the Fe species can prevent the photoanode from further photo-corrosion. On the other hand, as shown the blue plots in Fig. 5a, the BiVO₄ photoanode is first tested for 2 h in KBi + Fe(Ⅲ) for the growth of FeOOH/BiVO₄, during which the OER activity decreases in the first half hour but increase gradually to a plateau at 1.5 h. After 2 h, the photoanode is removed from the PEC cell and cleaned with DI water to remove Fe(Ⅲ) from the surface and then tested in KBi. The current densities decrease slowly to 1.73 mA/cm² and it is better than that of pristine BiVO₄ (1.29 mA/cm²) due to the in situ formation of FeOOH on the surface of BiVO₄ (FeOOH/BiVO₄) photoanode. Owing to the dominant dissolution rate, the activity and photo-response of FeOOH/BiVO₄ decreases subsequently leading to deteriorated PEC performance [58].

  Figure 5

  

  Figure 5.  (a) i-t plots of the BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ). (b) Proposed mechanism of Fe(Ⅲ) ions enhanced PEC water splitting.

  DownLoad: Full-Size Img PowerPoint

  A convenient technique to simultaneously boost the reaction kinetics and stability of the BiVO₄ photoanode is designed by adding Fe(Ⅲ) ions into the electrolyte. The in situ generation of the FeOOH co-catalyst on the BiVO₄ (FeOOH/BiVO₄) photoanode is achieved by the dynamic changes in KBi + Fe(Ⅲ) during OER. The FeOOH overlayer on the BiVO₄ photoanode turns into a hole transfer layer on account of the hole attraction ability of Fe, which can reduce the free energy barriers and accelerate the reaction kinetics. As a result, the current density of the BiVO₄ photoanode increases to 2.21 mA/cm² at 1.23 V vs. RHE, which is much higher than that of the pristine BiVO₄ photoanode (1.29 mA/cm²). Moreover, the BiVO₄/FeOOH photoanode shows excellent stability as a result of the inhibition of photo-corrosion rendered by FeOOH. The results demonstrate that changing the microenvironment by the addition of transition metal ions is a facile and promising strategy to enhance the PEC performance and boost the stability of the BiVO₄ photoanode. This approach opens a new avenue for the multifunctional modification of photoanodes for efficient solar conversion.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Hailang Deng: Writing – original draft, Methodology, Investigation, Conceptualization. Abebe Reda Woldu: Writing – review & editing, Formal analysis. Abdul Qayum: Methodology, Investigation. Zanling Huang: Validation, Formal analysis, Data curation. Weiwei Zhu: Writing – review & editing, Visualization, Software, Funding acquisition, Formal analysis, Data curation. Xiang Peng: Visualization, Software, Formal analysis. Paul K. Chu: Writing – review & editing, Visualization, Funding acquisition. Liangsheng Hu: Writing – review & editing, Visualization, Supervision, Resources, Project administration, Investigation, Funding acquisition, Data curation, Conceptualization.

  Acknowledgments

  This work was financially supported by the Guangdong Basic and Applied Basic Research Foundation (No. 2022A1515110212), the Special Fund Project for Science and Technology Innovation Strategy of Guangdong Province (Nos. STKJ202209077, STKJ202209083, and STKJ202209021), Scientific Research Foundation of Shantou University (No. NTF21022), and City University of Hong Kong Strategic Research Grant (SRG) (No. 7005505).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109892.

  
  
• 1. [1]

     B. Zhang, S. Yu, Y. Dai, et al., Nat. Commun. 12 (2021) 6969. doi: 10.1038/s41467-021-27299-0

  2. [2]

     J.H. Kim, D. Hansora, P. Sharma, et al., Chem. Soc. Rev. 48 (2019) 1908–1971. doi: 10.1039/c8cs00699g

  3. [3]

     S. Zhao, B. Liu, G. Zhang, et al., Trans. Tianjin Univ. 29 (2023) 473–481. doi: 10.1007/s12209-023-00374-x

  4. [4]

     J. Zhu, L. Hu, P. Zhao, et al., Chem. Rev. 120 (2020) 851–918. doi: 10.1021/acs.chemrev.9b00248

  5. [5]

     H. Sun, C.W. Tung, Y. Qiu, et al., J. Am. Chem. Soc. 144 (2021) 1174–1186. doi: 10.1109/icip42928.2021.9506288

  6. [6]

     Y. Xiao, C. Feng, J. Fu, et al., Nat. Catal. 3 (2020) 932–940. doi: 10.1038/s41929-020-00522-9

  7. [7]

     R.T. Gao, N.T. Nguyen, T. Nakajima, et al., Sci. Adv. 9 (2023) eade4589. doi: 10.1126/sciadv.ade4589

  8. [8]

     R.T. Gao, L. Liu, Y. Li, et al., Proc. Natl. Acad. Sci. U. S. A. 120 (2023) e2300493120. doi: 10.1073/pnas.2300493120

  9. [9]

     T.W. Kim, K.S. Choi, Science 343 (2014) 990–994. doi: 10.1126/science.1246913

  10. [10]

      Y. Kuang, Q. Jia, G. Ma, et al., Nat. Energy 2 (2016) 16191. doi: 10.1038/nenergy.2016.191

  11. [11]

      D.K. Lee, K.S. Choi, Nat. Energy 3 (2018) 53–60.

  12. [12]

      D. Lee, W. Wang, C. Zhou, et al., Nat. Energy 6 (2021) 287–294. doi: 10.1038/s41560-021-00777-x

  13. [13]

      J. Huang, Y. Wang, K. Chen, et al., Chin. Chem. Lett. 33 (2022) 2060–2064. doi: 10.1016/j.cclet.2021.08.082

  14. [14]

      L. Wang, X. Shi, Y. Jia, et al., Chin. Chem. Lett. 32 (2021) 1869–1878. doi: 10.1016/j.cclet.2020.11.065

  15. [15]

      B. Liu, X. Wang, Y. Zhang, et al., Angew. Chem. Int. Ed. 62 (2023) e202217346. doi: 10.1002/anie.202217346

  16. [16]

      R.T. Gao, L. Wang, Angew. Chem. Int. Ed. 59 (2020) 23094–23099. doi: 10.1002/anie.202010908

  17. [17]

      F.M. Toma, J.K. Cooper, V. Kunzelmann, et al., Nat. Commun. 7 (2016) 12012. doi: 10.1038/ncomms12012

  18. [18]

      R. Lei, Y. Tang, W. Qiu, et al., Nano Lett. 23 (2023) 11785–11792. doi: 10.1021/acs.nanolett.3c03743

  19. [19]

      Y. Shi, Y. Yu, Y. Yu, et al., ACS Energy Lett. 3 (2018) 1648–1654. doi: 10.1021/acsenergylett.8b00855

  20. [20]

      Q. Shi, S. Murcia-López, P. Tang, et al., ACS Catal. 8 (2018) 3331–3342. doi: 10.1021/acscatal.7b04277

  21. [21]

      J. Eichhorn, S.E. Reyes-Lillo, S. Roychoudhury, et al., Small 16 (2020) 2001600. doi: 10.1002/smll.202001600

  22. [22]

      S. Wang, G. Liu, L. Wang, Chem. Rev. 119 (2019) 5192–5247. doi: 10.1021/acs.chemrev.8b00584

  23. [23]

      C. Zhou, S. Wang, Z. Zhao, et al., Adv. Funct. Mater. 28 (2018) 1801214. doi: 10.1002/adfm.201801214

  24. [24]

      L. Jiang, H. Du, L. Li, et al., Trans. Tianjin Univ. 29 (2023) 462–472. doi: 10.1007/s12209-023-00376-9

  25. [25]

      W. Wang, P.J. Strohbeen, D. Lee, et al., Chem. Mater. 32 (2020) 2899–2909. doi: 10.1021/acs.chemmater.9b05047

  26. [26]

      S. Wang, T. He, P. Chen, et al., Adv. Mater. 32 (2020) 2001385. doi: 10.1002/adma.202001385

  27. [27]

      D. Kong, J. Qi, D. Liu, et al., Trans. Tianjin Univ. 25 (2019) 340–347. doi: 10.1007/s12209-019-00202-1

  28. [28]

      X. Wang, Chem. Mater. 33 (2021) 6251–6268. doi: 10.1021/acs.chemmater.1c01507

  29. [29]

      Y. Zhang, L. Xu, B. Liu, et al., ACS Catal. 13 (2023) 5938–5948. doi: 10.1021/acscatal.3c00444

  30. [30]

      J.B. Pan, B.H. Wang, S. Shen, et al., Angew. Chem. Int. Ed. 62 (2023) e202307246. doi: 10.1002/anie.202307246

  31. [31]

      M.A. Gaikwad, U.V. Ghorpade, U.P. Suryawanshi, et al., ACS Appl. Mater. Interfaces 15 (2023) 21123–21133. doi: 10.1021/acsami.3c01877

  32. [32]

      B. Zhang, X. Huang, Y. Zhang, et al., Angew. Chem. Int. Ed. 59 (2020) 18990–18995. doi: 10.1002/anie.202008198

  33. [33]

      B. Zhang, L. Wang, Y. Zhang, et al., Angew. Chem. Int. Ed. 57 (2018) 2248–2252. doi: 10.1002/anie.201712499

  34. [34]

      A.M. Hilbrands, S. Zhang, C. Zhou, et al., J. Am. Chem. Soc. 145 (2023) 23639–23650. doi: 10.1021/jacs.3c07722

  35. [35]

      H. Du, J. Fan, C. Miao, et al., Trans. Tianjin Univ. 27 (2021) 24–41. doi: 10.1007/s12209-020-00277-1

  36. [36]

      W. Zhang, J. Ma, L. Xiong, et al., ACS Appl. Energy Mater. 3 (2020) 5927–5936. doi: 10.1021/acsaem.0c00834

  37. [37]

      N. Clament Sagaya Selvam, S.J. Kwak, G.H. Choi, et al., ACS Energy Lett. 6 (2021) 4345–4354. doi: 10.1021/acsenergylett.1c01983

  38. [38]

      S. Anantharaj, S. Kundu, S. Noda, Nano Energy 80 (2021) 105514. doi: 10.1016/j.nanoen.2020.105514

  39. [39]

      S. Lee, L. Bai, X. Hu, Angew. Chem. Int. Ed. 59 (2020) 8072–8077. doi: 10.1002/anie.201915803

  40. [40]

      Z. Huang, S. Zhu, Y. Duan, et al., J. Energy Chem. 89 (2024) 99–109. doi: 10.1016/j.jechem.2023.10.036

  41. [41]

      Z. Huang, A.Reda Woldu, X. Peng, et al., Chem. Eng. J. 477 (2023) 147155. doi: 10.1016/j.cej.2023.147155

  42. [42]

      F. Yang, M. Lopez Luna, F.T. Haase, et al., J. Am. Chem. Soc. 145 (2023) 21465–21474. doi: 10.1021/jacs.3c07158

  43. [43]

      C. Kuai, C. Xi, A. Hu, et al., J. Am. Chem. Soc. 143 (2021) 18519–18526. doi: 10.1021/jacs.1c07975

  44. [44]

      Q. Sun, K. Ren, L. Qi, ACS Appl. Mater. Interfaces 14 (2022) 37833–37842. doi: 10.1021/acsami.2c10741

  45. [45]

      G. Fang, Z. Liu, C. Han, et al., ACS Appl. Energy Mater. 4 (2021) 3842–3850. doi: 10.1021/acsaem.1c00247

  46. [46]

      S. Chen, B. Li, D. Huang, et al., ACS Appl. Mater. Interfaces 13 (2021) 17586–17598. doi: 10.1021/acsami.1c01998

  47. [47]

      R.T. Gao, J. Zhang, T. Nakajima, et al., Nat. Commun. 14 (2023) 2640. doi: 10.1038/s41467-023-38343-6

  48. [48]

      H. Duan, H. Wu, H. Zhong, et al., J. Phys. Chem. C 126 (2022) 7688–7695. doi: 10.1021/acs.jpcc.2c00433

  49. [49]

      H. Wu, S. Qu, Z. Xie, et al., ACS Appl. Energy Mater. 5 (2022) 8419–8427. doi: 10.1021/acsaem.2c00963

  50. [50]

      Q. Wang, L. Wu, Z. Zhang, et al., ACS Appl. Mater. Interfaces 14 (2022) 26642–26652. doi: 10.1021/acsami.2c02790

  51. [51]

      R.T. Gao, D. He, L. Wu, et al., Angew. Chem. Int. Ed. 59 (2020) 6213–6218. doi: 10.1002/anie.201915671

  52. [52]

      Z. Wang, Y. Guo, M. Liu, et al., Adv. Mater. 34 (2022) e2201594. doi: 10.1002/adma.202201594

  53. [53]

      W. Fang, L. Fu, A. Qin, et al., ACS Appl. Energy Mater. 5 (2022) 6313–6323. doi: 10.1021/acsaem.2c00477

  54. [54]

      T. Zhou, S. Chen, J. Wang, et al., Chem. Eng. J. 403 (2021) 126350. doi: 10.1016/j.cej.2020.126350

  55. [55]

      M. Pradhan, S. Maji, A.K. Sinha, et al., J. Mater. Chem. A 3 (2015) 10254–10257. doi: 10.1039/C5TA01427A

  56. [56]

      Y. Hu, J. Zhou, L. Li, et al., J. Mater. Chem. A 10 (2022) 602–610. doi: 10.1039/d1ta08938b

  57. [57]

      X. Xiong, C. Zhang, X. Zhang, et al., Electrochim. Acta 389 (2021) 138795. doi: 10.1016/j.electacta.2021.138795

  58. [58]

      G. Zhang, Z. Li, J. Zeng, et al., Appl. Catal. B: Environ. 319 (2022) 121921. doi: 10.1016/j.apcatb.2022.121921

• 

  Scheme 1  Schematic illustration of the BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ).

  下载: 全尺寸图片 幻灯片
  

  Figure 1  (a) Chopped illumination curves. (b) Photostability of BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ). CV measurements of the BiVO₄ photoanodes in (c) KBi and (d) KBi + Fe(Ⅲ). (e) LSV and (f) i-t curves of the photoanodes for different pretreatment of Fe(Ⅲ).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) ABPE, (b) charge transfer efficiency, (c) Nyquist plots at 0.8 V under illumination or in the dark for the BiVO₄ photoanodes with or without Fe³⁺. (d) PL spectra, (e) UV–vis absorption spectra, and (f) IPCE at 1.23 V of BiVO₄ photoanodes after the i-t test in KBi or KBi + Fe(Ⅲ).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  XPS spectra of (a) Bi 4f, (b) V 2p, and (c) O 1s of the pristine BiVO₄ photoanodes before and after the i-t test with or without Fe(Ⅲ). (d) Fe 2p of the photoanode after the i-t test with Fe(Ⅲ).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  SEM images of (a) pristine BiVO₄ photoanodes, (b) after 5-h i-t test without Fe(Ⅲ), and (c) after i-t test with Fe(Ⅲ). (d) TEM images of pristine BiVO₄ photoanode. (e, f) BiVO₄ photoanode after the i-t test with Fe(Ⅲ). (g) HAADF-STEM image and elemental maps of the BiVO₄ photoanode after the 10-h i-t test with Fe(Ⅲ). The insets in (a-d) are the corresponding high-magnification images.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) i-t plots of the BiVO₄ photoanodes in KBi and KBi + Fe(Ⅲ). (b) Proposed mechanism of Fe(Ⅲ) ions enhanced PEC water splitting.

  下载: 全尺寸图片 幻灯片
•