Layered metallic vanadium diboride as an active cocatalyst for efficient dye-sensitized photocatalytic hydrogen evolution

Abstract

Here, we report that layered metallic vanadium diboride (VB₂) can function as an active cocatalyst for efficiently catalyzing photocatalytic hydrogen evolution in a dye-sensitized system under visible light irradiation. In an Erythrosin B–triethanolamine (ErB–TEOA) molecular system, the highest turnover number (TON) of hydrogen evolution based on ErB reaches 40. In addition, VB₂ outperforms V₂O₅, VS₂, VN, and VC for photocatalytic hydrogen evolution.

---

Hydrogen (H₂) is being pursued as the most promising energy carrier in the future because it has a high-energy density and its combustion product is only water. Photocatalytic water splitting has been regarded as one of the most ideal technologies for sustainable production of H₂.^1–4 To lower the kinetic barriers of the photocatalytic H₂ evolution reaction (HER) on semiconductors or in dye-sensitized systems, the use of a cocatalyst that can reduce the overpotential and consequently increase the efficiency of the photocatalytic HER is typically essential.⁴ So far, noble metal-based materials are still proven to be the most efficient cocatalysts and widely used for the photocatalytic HER due to their low overpotential of the HER and excellent photochemical stability.^4,5 However, the practical application of noble metals in a large-scale photocatalytic HER is greatly restricted because of their high cost and scarcity. Therefore, developing highly active and stable H₂ evolution cocatalysts (HECs) based on earth-abundant elements is highly desirable and has recently been extensively pursued.

Recently, inspired by the progress in the electrocatalytic HER, a large number of electrocatalysts based on transition-metal-based compounds including sulfides,⁶ selenides,⁷ phosphides,⁸ carbides,⁹ and nitrides¹⁰ have been successfully employed as HECs for the photocatalytic HER. Among them, transition metal carbides (TMCs), nitrides (TMNs), and diborides (TMBs) are more promising as HECs as they possess high electronic conductivity and a high density of Pt-like active sites for the HER.^11–13 On the other hand, recently various vanadium-based catalysts, such as VS₂,¹⁴ VSe₂,¹⁵ VN,¹⁶ VC,¹⁷ and VB₂,¹³ have been identified as promising alternatives to noble-metal-based electrocatalysts in catalyzing the electrocatalytic HER with high performance, and this research wave has transitioned to photocatalysis. Specifically, it has been shown that VS₂ can function as an efficient HEC to greatly promote the visible light photocatalytic HER activity of g-C₃N₄.¹⁸ In addition, in our previous studies, we have also shown that VC can act as a versatile, active, and robust cocatalyst when integrated with CdS and dye-sensitized systems under visible light irradiation.^17,19 In comparison with other vanadium-based catalysts, VB₂ not only has excellent chemical stability, but also inherently possesses graphene-like boron sheet motifs (borophene subunits) in its structure (Fig. 1a),^19–21 endowing VB₂ with high electronic conductivity and a large density of active sites, which make VB₂ highly promising as a HEC for the photocatalytic HER. However, to the best of our knowledge, there is no study on employing VB₂ as a HEC for promoting the photocatalytic HER.

Fig. 1 (a) The crystal structures of VB₂. (b) XRD pattern of as-received VB₂. (c) TEM image of VB₂ and corresponding SAED pattern (inset). (d) HAADF-STEM image of VB₂ and corresponding EDX elemental maps (V, B, and O). (e) V 2p and (f) B 1s XPS spectra of VB₂.

Herein, we report that layered metallic VB₂ could function as a superior HEC for catalyzing the visible-light-driven photocatalytic HER in an organic dye-based sensitization system for the first time. In an Erythrosin B–triethanolamine (ErB–TEOA) system, VB₂ can efficiently catalyze H₂ evolution, affording a H₂ evolution turnover number (TON) of 40 based on VB₂. Moreover, VB₂ exhibits much higher activity than V₂O₅, VS₂, VN, and VC in catalyzing photocatalytic H₂ evolution.

Fig. 1b shows the XRD pattern of as-received VB₂ powder, where the main XRD peaks located at 29.19°, 34.53°, 45.82° 60.53°, 61.84°, 69.83°, 71.60° and 72.81° can be assigned to the (001), (100), (101), (002), (110), (111), (102) and (200) crystal planes of hexagonal VB₂ (JCPDS 38-1463),¹⁹ respectively. No additional peaks related to impurities such as V and B oxides can be observed, suggesting the high purity of VB₂. The transmission electron microscopy (TEM) image shown in Fig. 1c indicates that VB₂ possesses a sheet-like structure with a lateral size up to 0.5–5 μm. The corresponding selected area electron diffraction (SAED) pattern (inset of Fig. 1c) shows well-resolved diffraction spots, further manifesting that VB₂ has a single-crystal structure. From its high resolution TEM (HRTEM) image (Fig. S1, ESI†), the observed interplanar distance of the lattice fringes is 0.260 nm, which corresponds to the (100) planes of VB₂ nanoparticles. Furthermore, the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of VB₂ and the corresponding energy-dispersive X-ray (EDX) elemental maps (Fig. 1d) further demonstrate that VB₂ is mainly composed of V and B elements. The presence of O element is probably due to the surface oxidation of VB₂ since no characteristic diffraction peaks of V and B oxides can be observed in its XRD pattern. The Brunauer–Emmett–Teller (BET) specific surface areas, average pore size, and total pore volume for VB₂ were determined to be 9.3 m² g⁻¹, 5.24 nm, and 0.012 cm³ g⁻¹, respectively (Fig. S2, ESI†). In addition, VB₂ is a black powder and its UV-vis diffusion reflectance spectrum (UV-vis DRS) exhibits a strong light absorption from 300 to 800 nm but without an obvious absorption edge (Fig. S3, ESI†), which implies its metallic character.¹⁷ Survey X-ray photoelectron spectroscopy (XPS) reveals that VB₂ consists of V, B and O elements (Fig. S4, ESI†). The V 2p envelope (Fig. 1e) can be fitted into two main peaks with different chemical environments. The V 2p peak at 513.2 eV is assigned to V–B, while the high binding energy peak at 516.6 eV is attributed to surface V–O species.^19–21 Accordingly, the fitting of the B 1s XPS spectrum also shows the presence of both V–B (188.3 eV) and surface B–O_x species (192.4 eV) (Fig. 1f). The presence of V–O and B–O_x species is a result of the surface oxidation of VB₂ when exposed to air and will passivate VB₂ from further oxidation.

The catalytic activity of VB₂ for photochemical H₂ evolution was evaluated in a dye-sensitized system using xanthene dyes as the photosensitizer and triethanolamine (TEOA) as the electron donor under visible light. Control experiments indicated that no H₂ was evolved in the absence of light or any of the other components. As shown in Fig. 2a, VB₂ can efficiently catalyze photocatalytic H₂ evolution with various xanthene dyes including Erythrosin B (ErB), Eosin Y (EY), Rose Bengal (RB), Fluorescein sodium (FS), and Rhodamine B (RhB), and ErB outperforms the others in photosensitizing VB₂ for H₂ evolution. The activity of dye/VB₂ systems follows an order of ErB > EY ≈ FS > RB ≫ RhB. The best performance obtained with ErB as the photosensitizer can be attributable to the heavy-atom effect of I substituents in the xanthene ring, which can improve the quantum yield of long-lived triplet states compared to their light-atom-substituted counterparts.²² The pH value of the TEOA solution was also found to significantly influence the photocatalytic activity of the ErB/VB₂/TEOA system, and the activity of H₂ evolution reaches the highest at pH 11 (Fig. S5, ESI†). The decrease in activity at much higher pH values (>11) is likely due to the lower H⁺ concentration in solution and the fact that H₂ evolution becomes less favorable with increasing pH values.²³ In contrast, decreasing the pH value from 11 to 9 leads to much lower H₂ evolution activity because the decomposition of protonated TEOA (TEOA⁺) becomes unfavorable, which will diminish the ability of TEOA as an electron donor.

Fig. 2 (a) H₂ evolution from TEOA solution (10 vol%, 25 mL, pH 11) containing 6.0 mM VB₂ as the catalyst and various xanthene dyes (1.0 mM) as the photosensitizers. (b) Time courses of H₂ evolution from TEOA solution (10 vol%, 25 mL, pH 11) containing VB₂ (6.0 mM) alone, ErB (1.0 mM) alone, and both ErB (1.0 mM) and VB₂ (6.0 mM). (c) Comparison of the H₂ evolution activity of V-base materials (6.0 mM) in the ErB (1.0 mM)–TEOA (10 vol%, 25 mL, pH 11) system under visible light irradiation. (d) H₂ evolution from systems containing 4.0 mM VB₂ and different concentrations of ErB. (e) H₂ evolution from TEOA solution (10 vol%, 25 mL, pH 11) containing 1.0 mM ErB and different concentrations of VB₂. (f) Stability of VB₂ (6 mM) for catalyzing H₂ evolution from TEOA solution (10 vol%, 25 mL, pH 11) containing ErB (1 mM). Light source: white LED lamp (380 nm ≤ λ ≤ 780 nm).

Fig. 2b presents the time course of H₂ evolution in the ErB/VB₂ system under visible light irradiation, including those of VB₂ and ErB systems for comparison. It is found that the ErB/VB₂ system can rapidly produce H₂ gas with a high initial H₂ evolution rate of 34 μmol h⁻¹. The total amount of H₂ evolved from this system is 213.1 μmol in a 10 h reaction, and the calculated turnover numbers (TONs) of H₂ evolution are about 17 and 7 based on ErB and VB₂, respectively, confirming that the H₂ evolution reaction in the ErB/VB₂ system proceeds photocatalytically.¹⁷ In contrast, the ErB system in the absence of VB₂ produces only a trace amount of H₂ (4.3 μmol in 10 h) along with rapid degradation of ErB, while VB₂ alone also shows negligible photocatalytic H₂ evolution activity (5.1 μmol in 10 h), although it has strong absorption toward visible light. These results clearly indicate that VB₂ could function as an efficient HEC in a dye-sensitized system for catalyzing H₂ evolution under visible light, but its activity is not comparable to that of most recently reported noble-metal-free cocatalysts in both dye sensitized and semiconductor-based photocatalytic H₂ evolution systems, probably due to its large size and thus limited active sites (Table S1†). In addition, when Ru(bpy)₂Cl₂ was used as a photosensitizer, VB₂ also shows a substantial catalytic activity for H₂ evolution in the presence of ascorbic acid as the sacrificial donor (Fig. S6, ESI†) under visible light. Furthermore, it was also found that VB₂ can efficiently catalyze photocatalytic H₂ evolution when integrated with inorganic semiconductors such as CdS and TiO₂ (Fig. S7, ESI†) under visible and UV-visible light, respectively, again verifying the versatility of VB₂ as an active HEC for solar photocatalytic H₂ evolution.

Most recently, it has been reported that various V-based materials such as V₂O₅, VS₂, VN, and VC could function as efficient (co)catalysts for electrocatalytic and photocatalytic H₂ evolution.^12,16,17,24 For comparison, the catalytic H₂ evolution activity of these V-based cocatalysts (XRD patterns shown in Fig. S8, ESI†) was tested and compared with that of VB₂ under the same reaction conditions. As shown in Fig. 2c, among the V-based cocatalysts tested, VB₂ exhibits the highest efficiency for photocatalytic H₂ evolution in the ErB–TEOA system. Typically, VB₂ shows 2 and 16 times higher activity than metallic VC and VN, respectively, while V₂O₅ and VS₂ are almost inactive for photocatalytic H₂ evolution. In addition, the catalytic activity of VB₂ was also compared with that of the most commonly studied transition metal diborides such as Mo and W borides and the obtained results are provided in Fig. S9, ESI.† Among the borides tested, VB₂ outperforms Mo₂B₅ and WB in catalyzing H₂ evolution in ErB–TEOA solution under visible light.

Furthermore, the effect of the concentrations of ErB and VB₂ on the photocatalytic H₂ evolution activity of the ErB/VB₂ system was investigated. As shown in Fig. 2d, at a fixed VB₂ concentration (4.0 mM) the amount of H₂ evolved and the system lifetime increase with the concentration of ErB and reach a maximum at 1.0 mM ErB, corresponding to a TON of 14.3 based on ErB, which suggests that at this value of [ErB] the system becomes limited by the concentration of VB₂.²³ Further increasing the concentration of ErB will lead to a decrease in the activity and TON probably due to the severe light-shield effect caused by excessive ErB present in solution.²² Notably, at 0.1 mM ErB, the highest TON for H₂ evolution is found to be as high as 40.1 based on ErB (Fig. S10, ESI†). On the other hand, as [VB₂] is increased at fixed [ErB] (Fig. 2e), the initial rate of H₂ evolution has a first-order dependence on [VB₂] (Fig. S11, ESI†) and the amount of H₂ produced monotonously increases with the concentration of VB₂ in a 10 h reaction. At 2 mM VB₂, the highest TON of H₂ evolution achieved in the ErB/VC system is 5.2 after a 10 h reaction.

The catalytic stability of VB₂ for photocatalytic H₂ evolution in the ErB–TEOA system was then evaluated by performing three cycles of H₂ evolution reaction under continuous light irradiation for 36 h. As shown in Fig. 2f, during the first 12 h reaction, although the ErB/VB₂ system shows a high initial H₂ evolution activity, the activity is gradually decreased with reaction time, which might be caused by the degradation of ErB and deactivation of VB₂. In order to clarify these hypotheses, the UV-vis absorption spectra of ErB during the photocatalytic H₂ evolution reaction were first monitored, as shown in Fig. S12, ESI.† As expected, the results show that ErB is not stable and degrades over time, which partially accounts for the observed decrease in H₂ evolution. On the other hand, after the first 12 h reaction, the used VB₂ was recovered by centrifugation, washed with water, and then redispersed in a reaction solution containing fresh ErB and TEOA for the subsequent H₂ evolution cycles. It is found that the H₂ evolution activity of the ErB/VB₂ system could only be partially revived in the subsequent cyclic reaction even with fresh ErB and TEOA added. This result suggests that VB₂ is not very stable during long-term photocatalytic H₂ evolution. The used VB₂ catalyst was characterized. As shown in Fig. S13, ESI,† although all the characteristic diffraction peaks ascribed to VB₂ can be observed in the XRD pattern of used VB₂, these peaks are slightly shifted to high 2θ values and their intensities become lower. These results indicate that the bulk phase structure of VB₂ almost keeps unchanged, while the surfaces of VB₂ might undergo oxidization to form amorphous V and/or B oxides. This hypothesis was confirmed by XPS analysis (Fig. S14, ESI†), revealing that the V–B species totally vanishes and only V–O could be observed in the XPS spectrum of V 2p. In addition, TEM and HAADF-STEM analyses (Fig. S15, ESI†) indicate that the surface of VB₂ was covered by 30–40 nm thick amorphous oxides. To further exclude the possibility that the VO_x or BO_x species on the surfaces of VB₂ may act as active sites, a series of air-annealed VB₂ samples obtained by annealing VB₂ at different temperatures (XRD patterns shown in Fig S16a, ESI†) with increased surface VO_x or BO_x species were also tested for the photocatalytic HER in the ErB–TEOA system (Fig S16b, ESI†). It is found that the activities of all the a-VB₂ samples are lower than that of unannealed VB₂ and decrease with increasing annealing temperature, clearly revealing that the surface VO_x or BO_x species cannot act as the active sites for the photocatalytic HER. These results clearly indicate that the decrease in catalytic stability is due to the surface oxidation of VB₂ during the photocatalytic H₂ evolution and catalyst recovery processes.

The photoinduced electron transfer and the photocatalytic H₂ evolution mechanism in the ErB/VB₂ system were explored by photoluminescence (PL) spectroscopy. As shown in Fig. 3a and b, it is found that the PL emission of ErB (excited at 480 nm) could be either reductively quenched by TEOA or oxidatively quenched by VB₂. These results suggest that when both TEOA and VB₂ cocatalysts are present in the system, reductive and oxidative quenching may simultaneously occur and compete with each other in the electron transfer processes for H₂ evolution from the excited state of ErB.^17,22,23 However, the oxidative quenching of excited ErB by VB₂ would dominate because of the much larger k_q (9.17 × 10¹¹ M⁻¹ s⁻¹) for oxidative quenching compared to k_q (8.57 × 10⁹ M⁻¹ s⁻¹) for reductive quenching by TEOA. Therefore, it could be proposed that the photocatalytic H₂ evolution in the ErB/VB₂/TEOA system most likely proceeds via an oxidative quenching mechanism.¹⁷ Upon visible light irradiation, ErB absorbs visible light photons to form excited state ErB*, which was then oxidatively quenched by directly transferring electrons to the VB₂ cocatalyst, where H⁺/H₂O is reduced to form H₂ gas. At the same time, the oxidative ErB cation radicals are regenerated to ground state ErB by TEOA as an electron donor, thereby completing the catalytic cycles. The cocatalyst role of VB₂ is further confirmed by electrochemical measurements. As shown in the polarization curves (Fig. 3c), the observed cathodic current in the range of 0 to −0.6 V versus RHE can be ascribed to the H₂ evolution. In comparting to pristine carbon paper (CP), VB₂ loaded CP (VB₂/CP) shows a much higher cathodic current. This result indicates that VB₂ indeed is a highly active cocatalyst for catalyzing photocatalytic H₂ evolution. Moreover, VB₂/CP exhibits a much smaller semicircle diameter than CP (Fig. 3d), suggesting an obviously reduced charge transfer resistance due to the metallic character of VB₂,¹³ which is favorable for accepting electrons from photoexcited ErB, in turn enhancing the charge separation efficiency and thus the photocatalytic H₂ evolution performance of the ErB/VB₂ system under visible light irradiation.

Fig. 3 (a) PL mission quenching of ErB solution (10 μM) with TEOA and corresponding Stern–Volmer plot. (b) PL mission quenching of ErB solution (10 μM) with VB₂ and corresponding Stern–Volmer plot. (c) Catalytic HER polarization curves of carbon paper (CP) and VB₂ loaded carbon paper (VB₂/CP) recorded in Na₂SO₄ (0.5 M) containing TEOA (10 vol%, pH 11) at a scan rate of 10 mV s⁻¹. (d) EIS Nyquist plots of CP and VB₂/CP recorded in Na₂SO₄ (0.5 M) containing TEOA (10 vol%, pH 11). Light source: white LED lamp (380 nm ≤ λ ≤ 780 nm).

In summary, metallic VB₂ has been identified as an efficient cocatalyst for catalyzing photocatalytic H₂ evolution in a dye-sensitized system under visible light irradiation. VB₂ was also found to catalyze the H₂ evolution reaction more efficiently than various potential V-based cocatalysts such as V₂O₅, VS₂, VN, and VC. This work not only confirms the feasibility of using metallic and earth-abundant VB₂ as a H₂ evolution cocatalyst in molecular systems, but also could be expected to stimulate research on employing and developing V-based materials in large-scale light energy conversion devices for practical applications.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (21763001 and 21463001), the Natural Science Foundation of Ningxia Province (2018AAC02011), the West Light Foundation of the Chinese Academy of Sciences (XAB2018AW13), the Foundation of Training Program for Young and Middle-aged Talents of State Ethnic Affairs Commission of China, and the Foundation of Key Laboratory of Electrochemical Energy Conversion Technology and Application, North Minzu University.

Notes and references

1. Z. Q. Cao, H. B. Hu, M. Z. Wu, C. J. Tu, D. W. Zhang and Z. Q. Wu, Sustainable Energy Fuels, 2019, 3, 2409.
2. S. T. Wu, X. W. Shi and M. S. Zhu, Sustainable Energy Fuels, 2019, 3, 1461.
3. X. Chen, S. Shen, L. Guo and S. Mao, Chem. Rev., 2010, 110, 6503.
4. J. Yang, D. Wang, H. Han and C. Li, Acc. Chem. Res., 2013, 46, 1900.
5. S. X. Min, Y. Xue, F. Wang, Z. G. Zhang and H. T. Zhu, Chem. Commun., 2019, 55, 10631.
6. X. Zong, H. Yan, G. Wu, G. Ma, F. Wen, L. Wang and C. Li, J. Am. Chem. Soc., 2008, 130, 7176.
7. M. L. Liu, B. B. Chen, R. S. Li, C. M. Li, H. Y. Zou and C. Z. Huang, ACS Sustainable Chem. Eng., 2017, 5, 4154.
8. A. Indra, A. Acharjya, P. W. Menezes, C. Merschjann, D. Hollmann, M. Schwarze, M. Aktas, A. Friedrich, S. Lochbrunner, A. Thomas and M. Driess, Angew. Chem., Int. Ed., 2017, 56, 1653.
9. Q. S. Gao, W. B. Zhong, Z. P. Shi, L. C. Yang and Y. Tang, Adv. Mater., 2019, 31, 1802880.
10. R. Godin, Y. O. Wang, M. A. Zwijnenburg, J. W. Tang and J. R. Durrant, J. Am. Chem. Soc., 2017, 5, 4154.
11. S. Meyer, A. V. Nikiforov, I. M. Petrushina, K. Kohler, E. Christensen, J. Jensen and N. Bjerrum, Int. J. Hydrogen Energy, 2015, 40, 2905.
12. Y. Zhong, X. H. Xia, F. Shi, J. Y. Zhan, J. P. Tu and H. J. Fan, Adv. Sci., 2016, 3, 1500286.
13. F. F. Guo, Y. Y. Wu, X. Ai, H. Chen, G.-D. Li, W. Chen and X. X. Zou, Chem. Commun., 2019, 55, 8627.
14. J. Yuan, J. Wu, W. J. Hardy, P. Loya, M. Lou, Y. Yang, S. Najma ei, M. Jiang, F. Qin, K. Keyshar, H. Ji, W. Gao, J. Bao, J. Kono, D. Natelson, P. M. Ajayan and J. Lou, Adv. Mater., 2015, 27, 5605.
15. M. Yan, X. Pan, P. Wang, F. Chen, L. He, G. Jiang, J. Wang, J. Z. Liu, X. Xu, X. Liao, J. Yang and L. Mai, Nano Lett., 2017, 17, 4109.
16. B. B. Bin, G. S. Tang, H. F. Liang, Z. B. Qi, D. F. Zhang, W. S. Hu, H. Shen and Z. C. Wang, Electrochem. Commun., 2018, 93, 166.
17. L. Tian, S. X. Min, Y. G. Lei, S. S. Chen and F. Wang, Chem. Commun., 2019, 55, 6870.
18. M. M. Shao, Y. F. Shao, S. J. Ding, J. W. Wang, J. C. Xu, Y. J. Qu, X. W. Zhong, X. M. Chen, W. Fai, N. Wang, B. M. Xu, X. Q. Shi, X. E. Wang and H. Pan, Appl. Catal., B, 2018, 237, 295.
19. L. Tian, S. X. Min and F. Wang, Appl. Catal., B, 2019, 259, 118029.
20. V. Mazánek, H. Nahdi, J. Luxa, Z. Sofer and M. Pumera, Nanoscale, 2018, 10, 11544.
21. P. Wang, R. Kumar, E. M. Sankaran, X. T. Qi, X. Y. Zhang, D. Popov, A. L. Cornelius, B. S. Li, Y. S. Zhao and L. P. Wang, Inorg. Chem., 2018, 57, 1096.
22. J. H. Hou, Y. G. Lei, F. Wang, X. H. Ma, S. X. Min, Z. L. Jin and J. Xu, Int. J. Hydrogen Energy, 2017, 42, 11118.
23. Z. J. Han, W. R. McNamara, M.-S. Eum, P. L. Holland and R. Eisenberg, Angew. Chem., Int. Ed., 2012, 51, 1667.
24. J. Yuan, J. Wu, W. J. Hardy, P. Loya, M. Lou, Y. Yang, S. Najmaei, M. Jiang, F. Qin, K. Keyshar, H. Ji, W. Gao, J. Bao, J. Kono, D. Natelson, P. M. Ajayan and J. Lou, Adv. Mater., 2015, 27, 5605.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and additional data. See DOI: 10.1039/c9se00820a

---