Construction of a Ru@WO_3−x–W catalyst with hydrogen spillover effect toward ampere-level hydrogen evolution in acidic solution

Abstract

A Ru@WO_3−x–W/TM catalyst with WO_3−x(surface)–W(bulk) structure was carefully designed and prepared. Compared with W or WO₃ support, the presence of WO_3−x–W modifies the electronic structure, optimizes H adsorption/desorption and promotes hydrogen spillover with Ru. The Ru@WO_3−x–W/TM catalyst can generate an industrial HER current density of 1204 mA cm⁻² in an acidic system.

---

As a hydrogen production technology with zero carbon emission, research on electrocatalytic water splitting catalysts has been widely undertaken. Pt is undoubtedly the best catalyst for the hydrogen evolution reaction (HER), but its high cost and scarce reserves limit its large-scale application. Ru has a hydrogen adsorption energy (65 kcal mol⁻¹) similar to that of Pt (62 kcal mol⁻¹) and is considered a promising candidate as a substitute for Pt. The HER activity of Ru can be improved by optimizing its electronic structure and morphology, including alloying, engineering defects, phase transformation, strain engineering, heteroatom doping and functionalized support. At present, the design of efficient HER catalysts using the “hydrogen spillover strategy from metal to support” has attracted wide attention. In the process of hydrogen spillover, H atoms are adsorbed and desorbed at two different active sites, which can overcome the limitations of the Sabatier principle.

Some progress has been made in Ru-based HER catalysts designed with hydrogen spillover in mind. The commonly used supports are reducible metal oxides, such as WO₃, CeO₂, MoO₂ and TiO₂,^1–3 which exhibit hydrogen spillover effects with Ru. To further improve the efficiencies of hydrogen spillover, metal oxide supports rich in oxygen defects, such as WO_3−x, MoO_x, NiMoO_4−x and LaCeO_x, were studied.^4–8 In addition, various supports, such as metal phosphides,^9–12 metal carbides,^13–16 metal sulfides,^17,18 metal selenides,¹⁹ metal hydroxides,²⁰ transition metal (oxy)hydroxides,²¹ MXene,²² layered double hydroxides²³ and functional carbon supports,^24,25 have also been combined with metal Ru to induce efficient hydrogen spillover. In addition to the improvement of the support, the hydrogen spillover efficiency can be further improved by replacing metal Ru with metal Ru composites, such as RuMo alloy,⁵ Ru₂P,¹¹ Ru₁Fe₁¹² and RuS_x.¹⁷

With the deepening understanding of hydrogen spillover, it has been found that there is also a significant reverse spillover effect in HER catalysts. The reverse hydrogen spillover pathway is that the adsorbed hydrogen overflows from the support to the metal catalyst. Reverse hydrogen spillover can also greatly improve HER performance, especially in neutral and alkaline solutions. For the Ru/WO_3−x electrocatalyst,⁴ the existence of the reverse spillover effect (H overflow from WO_3−x to Ru) was demonstrated through in situ Raman spectroscopy and DFT calculations. In the Ru/W₂C heterogeneous catalyst,²⁶ H₂O dissociates on the W₂C support, and the H intermediate thus formed overflows to the adjacent interface Ru site for H–H coupling and H₂ release. In addition, significant reverse hydrogen spillover also occurs for Ru/NiMoO_4−x,⁶ LaCeO_x@NGr/Ru₁,⁷ Ru/ac-ZrO₂,⁸ Ru–Mo₂C,¹³ Ru/WC_x,¹⁴ and Ru₁–Mo₂C¹⁶ catalysts, which significantly improves their HER performances. Meanwhile, some researchers have put forward the principle of minimum work function difference^{8,12,18,19,23,27} and crystal surface matching²⁶ in the design of catalysts with obvious hydrogen spillover.

Ru-based hydrogen spillover catalysts still have the following two problems. (1) The conductivity of reducible oxide support is poor, and the preparation process of support with a defect structure is relatively complicated. (2) There has been little research into acidic systems (Table S2, SI), and the current density rarely reaches the ampere level. In this paper, a Ru@WO_3−x–W/TM catalyst with WO_3−x(surface)–W(bulk) phase structure is designed for the first time, and its preparation process is simple, as shown in Scheme 1 (detailed description is presented in the SI). The presence of the WO_3−x–W phase modifies the electronic structure of the catalyst, increases the conductivity, optimizes the H adsorption/desorption performance, and leads to an improvement of hydrogen spillover efficiency. The current density of Ru@WO_3−x–W/TM reaches 1204 mA cm⁻², which meets the requirements of industrial application.

Scheme 1 Schematic illustration of the preparation of Ru@WO_3−x–W.

The morphology of Ru@WO_3−x–W/TM was analyzed by transmission electron microscopy (TEM). In Fig. 1a and S1a (SI), it presents a thin-film structure. The particle size distribution is uniform within the range of 2–5 nm in Fig. S1b (SI). In the high-resolution TEM (HRTEM) image, W (211) and (210) planes with interplanar spacings of 0.206 and 0.230 nm were observed (Fig. 1b). In Fig. S1c and d (SI), W nanoparticles in an amorphous state can also be observed. Moreover, typical dark dots—corresponding to Ru atoms distributed as atomic clusters—can be observed on the lattice of W in the aberration-corrected high-angle annular dark-field scanning TEM (AC-HAADF-STEM) image (Fig. 1c).²⁸ The corresponding elemental mapping in Fig. 1d confirms that Ru, W and O are evenly distributed. In the XRD patterns of Ru@W/TM and Ru@WO_3−x–W/TM (Fig. S2a, SI), because the diffraction peaks of W and Ti are very close, only the diffraction peaks of the TM support were observed. However, in Fig. S2b (SI), if the TM support is replaced by an indium–tin oxide (ITO) support, the diffraction peaks of W(210) and WO_3−x(110) can be observed. In addition, the composition analysis results of Ru@WO_3−x–W are listed in Table S1 (SI).

Fig. 1 TEM (a) and HRTEM (b) images of Ru@WO_3−x–W. AC-HAADF-STEM image of Ru@WO_3−x–W (c). STEM elemental mappings of Ru@WO_3−x–W (d). The Ru 3p spectra of Ru@WO_3−x–W and Ru@W (e). The O 1s spectra of Ru@WO_3−x–W and Ru@W (f).

XPS analysis of Ru@WO_3−x–W/TM was carried out to investigate the modification of its electronic structure. For the W in Ru@W, the binding energy (B.E.) peaks at 33.8 and 31.3 eV belong to the 4f orbitals of W⁰, and the B.E. peaks at 38.0 and 35.8 eV belong to the 4f orbitals of W⁶⁺ in Fig. S3 (SI). For Ru@WO_3−x–W, the situation is different. The B.E. peaks of the W⁶⁺ 4f orbital have an obvious positive shift, which are located at 39.0 and 36.9 eV. This is the result of electrochemical oxidation. Meanwhile, a weak B.E. peak of the W 4f orbital was observed at 35.2 eV, and its valence state was between 0 and 6⁺. The above analysis proves that the WO_3−x–W phase is formed. In Fig. 1e, the B.E. peaks of 466.4 and 460.4 eV can be attributed to the spin-splitting peaks of Ru⁴⁺ and Ru⁰ 3p orbitals in Ru@W/TM. Compared with those of Ru@W/TM, the B.E. peaks of the Ru 3p orbital of Ru@WO_3−x–W/TM are negatively shifted. There are two forms of adsorbed oxygen in the O 1s peaks of Ru@W/TM, namely OH⁻ (530.7 eV) and H₂O (532.5 eV) in Fig. 1f. However, for Ru@WO_3−x–W/TM, they are located at 532.1 and 533.1 eV. The obvious positive shift of B.E peaks of O 1s indicates the formation of a RuO(OH)₂ phase on the surface of Ru@WO_3−x–W/TM. The theoretical calculation results of Karlsson et al.²⁹ had indicated that, after 1–6H atom is modified to the surface of RuO₂, the peak of O 1s for OH⁻ will move by 1.7–2.6 eV, while the peak of O1s without H will move by 0.3–0.8 eV. Detailed analysis is presented in Table S3 (SI).

The HER activity of a Ru@WO_3−x–W/TM electrode is shown in Fig. 2a. At the same time, the HER activities of Ru/TM, Ru@W/TM and commercial Pt/C electrodes are also compared. First, the titanium mesh (TM) support has a porous structure, but it has almost no HER activity. The Ru/TM electrode obtained by simply electrodepositing Ru on the TM support shows a certain HER activity, but it is not ideal. The overpotential at 10 mA cm⁻² (η₁₀) of the Ru/TM electrode is 89 mV in Fig. 2b. However, for the Ru@W/TM electrode, the situation changes greatly, and its η₁₀ decreased to 34 mV. The electrode with the best HER activity is Ru@WO_3−x–W/TM, and its η₁₀ is reduced to 16 mV, which is similar to that of Pt/C (η₁₀ = 13 mV). From the perspective of current density, Ru@WO_3−x–W/TM reaches 1204 mA cm⁻² at an overpotential of 200 mV (j₂₀₀) in Fig. 2c, reaching a level suitable for industrial production. The j₂₀₀ of Ru@WO_3−x–W/TM is not only 6.5 and 2.5 times that of Ru/TM and Ru@W/TM respectively, but also 1.8 times that of Pt/C. At present, Ru-based electrocatalysts have been extensively studied in alkaline systems and have demonstrated satisfactory HER activities (Table S4, SI). However, in acidic systems, Ru-based electrocatalysts have not been studied enough (Table S2, SI), and HER activities need to be further improved. In this paper, the Ru@WO_3−x–W/TM catalyst has achieved HER activity corresponding to ampere-level current, which is superior to that of most reported Ru-based catalysts in acidic systems (Fig. 2d). Mass activities (MAs) of Ru/TM and Ru@W/TM are 9.1 and 21.0 A mg⁻¹, respectively (Fig. 2e). The MA of Ru@WO_3−x–W/TM is 48.9 A mg⁻¹, which is 5.4 times higher than that of Ru/TM and 2.3 times than that of Ru@W/TM. This suggests that the utilization rate of Ru has been improved in Ru@WO_3−x–W/TM. In Fig. S4, the electrochemically active surface area (ECSA) of Ru@WO_3−x–W/TM is the largest, which is beneficial for improved HER activity (detailed description is shown in the SI). At the same time, the specific activity and turnover frequency of Ru@WO_3−x–W/TM are the largest in Fig. S5 (SI), which indicates that the intrinsic activity is also improved.

Fig. 2 LSV curves of Ru/TM, Ru@W/TM, Ru@WO_3−x–W/TM, Pt/C and TM (a). The values of η₁₀ (b) and j₂₀₀ (c) of Ru/TM, Ru@W/TM, Ru@WO_3−x–W/TM, Pt/C and TM. Comparison of the acidic HER activities of Ru-based hydrogen spillover catalysts in the literature (d). The MAs of Ru/TM, Ru@W/TM and Ru@WO_3−x–W/TM (e). LSV curves of Ru@W/TM, Ru@WO₃/TM and Ru@WO_3−x–W/TM (f).

The formation of WO_3−x–W on the surface plays an important role in improving HER activity. This is related to the scan rate during electro-oxidation. The slower the scan rate, the more efficient the oxidation. With the decrease of scan rate, the η₁₀ of Ru@WO_3−x–W/TM decreases from 32 to 16 mV, and the j₂₀₀ increases from 817.5 to 1204 mA cm⁻² (Fig. S6, SI). The MA of Ru@WO_3−x–W/TM increases from 37.4 to 48.9 A mg⁻¹ (Fig. S7, SI), and its ECSA increases from 134.7 to 195.5 cm² (Fig. S8, SI). In Fig. 2f and Fig. S9 (SI), the HER activity of Ru@WO₃/TM (bulk oxidation, as proved by XRD in Fig. S10, SI) is obviously lower than that of Ru@WO_3−x–W/TM (surface oxidation), even lower than that of Ru@W/TM. Meanwhile, the MA and ECSA of Ru@WO₃/TM are also lower than those of Ru@WO_3−x–W/TM (Fig. S11 and S12, SI). This indicates that bulk oxidized WO₃ is not as effective in promoting HER activity as surface oxidized WO_3−x.

The increase in HER activity of Ru@WO_3−x–W/TM may come from the change of its H adsorption and desorption properties. In Fig. S13 (SI), the H adsorption/desorption charges of Ru/TM, Ru@W/TM and Ru@WO_3−x–W/TM are obtained by integrating the hydrogen region areas. Whether it is H adsorption or desorption charge, they both exhibit the following changing trend: Ru@WO_3−x–W/TM > Ru@W/TM > Ru/TM. An increase in H adsorption charge facilitates the Volmer step in the HER process, and an increase in H desorption charge promotes the Tafel or Heyrovsky step in the HER process. In addition, H adsorption and desorption potentials can also reflect the change of electrochemical properties of the electrode. In Fig. S14a and b (SI), the potential of H_ad of Ru@WO_3−x–W/TM is the highest, which indicates that it is most beneficial to the H adsorption process. At the same time, the potential of H_des of Ru@WO_3−x–W/TM is the lowest, indicating that it is also the most beneficial to H desorption. In Fig. S14c (SI), by comparing the CVs of W/TM and WO_3−x–W/TM, it can be concluded that WO_3−x–W/TM is more effective in activating and adsorbing H (from the charge and potential of H_ad). Therefore, the change in the characteristics of Ru@WO_3−x–W/TM comes from the formation of the WO_3−x–W phase and the resulting hydrogen spillover effect between WO_3−x–W and Ru. In addition, the changes of H_ad/H_des properties of Ru@WO_3−x–W/TM with different oxidation degrees (Fig. S15, SI) and Ru@WO₃/TM (Fig. S16, SI) are further discussed in the SI.

In order to study the dynamics of the HER process, Tafel analyses are performed, as shown in Fig. 3a. The Tafel slope of Ru/TM is very high (74.37 mV dec⁻¹), indicating slow reaction kinetics. The Tafel slope of Ru@W/TM decreases to 64.31 mV dec⁻¹, indicating the acceleration of reaction kinetics. The fastest kinetics of the HER reaction is exhibited by Ru@WO_3−x–W/TM, and its Tafel slope has been reduced to 40.70 mV dec⁻¹, which is close to that of Pt/C (21.80 mV dec⁻¹). Moreover, the values of Tafel slope indicate that the mechanism of HER on a Ru@WO_3−x–W/TM electrode is consistent with the Volmer–Heyrovsky pathway, in which the Heyrovsky step is a rate-determining one. Tafel slopes of Ru@WO_3−x–W/TM electrodes with different surface oxidation degrees were analyzed, as shown in Fig. S17 (SI). In addition, the Nyquist diagrams of Ru/TM, Ru@W/TM and Ru@WO_3−x–W/TM, recorded at an overpotential of 50 mV by electrochemical impedance spectroscopy (EIS) measurements, are shown in Fig. S18. The EIS measurements of Ru@WO_3−x–W/TM electrodes with different surface oxidation degrees are shown in Fig. S19 (SI). The results of long-term stability tests of Ru@WO_3−x–W/TM are shown in Fig. S20 (SI). XPS, XRD and composition analysis results after stability tests are shown in Fig. S21, S22 and Table S1 (SI), respectively.

Fig. 3 Tafel slopes of Ru/TM, Ru@W/TM, Ru@WO_3−x–W/TM and Pt/C (a). Evidence of hydrogen spillover effect of hydrogen desorption potential vs. scan rate (b), j vs. pH (c) and R_ctvs. potential (d). DFT calculations during the hydrogen spillover process (e).

In order to further quantitatively prove the influence of hydrogen spillover of Ru@WO_3−x–W/TM on H adsorption/desorption kinetics, the following analyses were made. In the potential range of 0–0.5 V, CVs were performed at different scan rates in Fig. S23 (SI). With the increase of scan rate, the H_des potentials of Ru@W/TM and Ru@WO_3−x–W/TM do not move forward obviously in Fig. 3b, which means that their hydrogen desorption kinetics are faster. Second, the effect of [H⁺] on HER activity kinetics was investigated, as shown in Fig. S24 (SI). In Fig. 3c, the slope for Ru@WO_3−x–W/TM is 1.63, which is obviously higher than that for Ru/TM (1.02) and Ru@W/TM (0.47). The slope for Ru@WO_3−x–W/TM is close to 2, which indicates that there is obvious hydrogen spillover. EIS measurements were conducted on Ru/TM, Ru@W/TM and Ru@WO_3−x–W/TM at different overpotentials as shown in Fig. S25 (SI). The hydrogen adsorption kinetics can be quantified by plotting log R_ctvs. potential in Fig. 3d. Ru@WO_3−x–W/TM exhibits a lower Tafel slope, indicating a higher hydrogen adsorption rate, which may be due to enhanced hydrogen spillover. As shown in Fig. 3e, the H_ads Gibbs free energy (ΔG_H*) at site-A (Ru-top) is −0.54 eV, which is beneficial for the occurrence of H adsorption. ΔG_H* is 0.21 eV for site-D (O–WO_3−x support), which is beneficial for H desorption. Moreover, ΔG_H* at site-A is closer to zero, which can balance the H adsorption and desorption for enhancing HER activity. In addition, the transfer between active sites has a small migration energy barrier. ΔG_H* between site-C (Ru–O bridge) and site-D is 0.41 eV, implying a lower energy barrier for the migration of H_ads from Ru to WO_3−x.

In summary, there is a significant hydrogen spillover effect between Ru and WO_3−x–W, resulting in an industrial-level current density of 1204 mA cm⁻² in acidic systems. The η₁₀ of Ru@WO_3−x–W/TM is 16 mV, which is obviously lower than that of Ru/TM, Ru@W/TM and Ru@WO₃/TM, and close to that of Pt/C.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: experimental details and addition data. See DOI: https://doi.org/10.1039/d5cc05104e.

Notes and references

1. K. Shun, K. Mori, S. Masuda, N. Hashimoto, Y. Hinuma, H. Kobayashi and H. Yamashita, Chem. Sci., 2022, 13, 8137–8147.
2. S. Zhou, H. Jang, Q. Qin, L. Hou, M. G. Kim, S. Liu, X. Liu and J. Cho, Angew. Chem., Int. Ed., 2022, 1, e202212196.
3. Y. Chen, Y. Liu, L. Li, T. Sakthivel, Z. Guo and Z. Dai, Adv. Funct. Mater., 2024, 34, 2401452.
4. J. Chen, C. Chen, M. Qin, B. Li, B. Lin, Q. Mao, H. Yang, B. Liu and Y. Wang, Nat. Commun., 2022, 13, 5382.
5. Z. Zhao, J. Sun, X. Li, S. Qin, C. Li, Z. Zhang, Z. Li and X. Meng, Nat. Commun., 2024, 15, 7475.
6. X. Liu, X. Wang, K. Li, J. Tang, J. Zhu, J. Chi, J. Lai and L. Wang, Angew. Chem., Int. Ed., 2024, 63, e202316319.
7. V. Dao, H. Choi, S. Yadav, J. D. Jiménez, C. Kim, T. V. Nguyen, K. Chen, P. Uthirakumar, Q. V. Le, S. D. Senanayake, H. Y. Kim and I.-H. Lee, Appl. Catal., B, 2024, 343, 12345.
8. J. Niu, H. Duan, T. Sun, Z. Zhi, D. Li, X. Fan, L. Zhang and D. Yang, Small, 2025, 2410436.
9. S. Liu, Z. Li, Y. Chang, M. G. Kim, H. Jang, J. Cho, L. Hou and X. Liu, Angew. Chem., Int. Ed., 2024, 63, e202400069.
10. K. Ji, S. Wang, S. Yao, Y. Ji, J. Li, X. Wang, L. Shi, G. Wang, W. Ren, J. Wang, F. Zhang, J. Xie, Z. Yang and Y.-M. Yan, Adv. Mater., 2025, 37, 2503182.
11. Y. Hong, S. Jeong, J. H. Seol, T. Kim, S. C. Cho, T. K. Lee, C. Yang, H. Baik, H. S. Park, E. Lee, S. J. Yoo, S. U. Lee and K. Lee, Adv. Energy Mater., 2024, 14, 2401426.
12. J. Li, Y. Tan, M. Zhang, W. Gou, S. Zhang, Y. Ma, J. Hu and Y. Qu, ACS Energy Lett., 2022, 7, 1330–1337.
13. D. K. Sam, Z. Tang and Y. Cao, J. Colloid Interface Sci., 2025, 699, 138213.
14. J. Zhao, M. Kou, Q. Yuan, Y. Yuan and J. Zhao, Small, 2025, 21, 2406022.
15. J. Sun, Z. Wang, Y. Wang, Y. Song, Y. Pei, W. Yan, R. Xiong, Y. Liu, B. Lin, X. Wang, X. Zhang, J. Chen and L. Zhang, Small, 2025, 21, 2411975.
16. T. Chao, W. Xie, Y. Hu, G. Yu, T. Zhao, C. Chen, Z. Zhang, X. Hong, H. Jin, D. Wang, W. Chen, X. Li, P. Hu and Y. Li, Energy Environ. Sci., 2024, 17, 1397–1406.
17. H. Yue, Z. Guo, Z. Zhou, X. Zhang, W. Guo, S. Zhen, P. Wang, K. Wang and W. Yuan, Angew. Chem., Int. Ed., 2024, 63, e202409465.
18. J. Chen, J. Ni, H. Xu, X. Qian, Y. Liu, L. Jin, K. Wang, G. He and H. Chen, Int. J. Hydrogen Energy, 2025, 139, 621–629.
19. C. Mu, H. Xin, Q. Luo, Y. Li and F. Ma, J. Mater. Chem. A, 2023, 11, 7016–7024.
20. R. Tang, P. Yan, Y. Zhou and X.-Y. Yu, EES Catal., 2024, 2, 932–940.
21. H. Lei, W. Yang, S. Hu, L. Yi, C. Ma, C. Hu, F. Qian, M. Zhao, L. Liu, G. Yang and Q. Chen, Angew. Chem., Int. Ed., 2025, 64, e202503871.
22. J. Tang, X. Liu, X. Wang, S. Wu, J. Zhu, X. Wang, T. Wang, J. Chi, Z. Wu and L. Wang, Small, 2024, 21, 2409675.
23. J. Zhu, J. Chi, X. Wang, T. Cui, L. Guo, B. Dong, X. Liu and L. Wang, Nano Energy, 2024, 121, 109249.
24. X. Liu, F. Xie, Q. Xi, Y. Wang, Y. Wang, H. Li, R. Li, C. Fan, J. Liu and J. Wang, J. Colloid Interface Sci., 2025, 694, 137738.
25. Q. Li, F. Huang, S. Li, H. Zhang and X.-Y. Yu, Small, 2021, 18, 2104323.
26. J.-Z. Jiang, S. Liu, Z. Li, M. G. Kim, H. Jang, X. Liu and L. Hou, Adv. Energy Mater., 2025, 15, 2405546.
27. Y. Liu, H. Li, X. Liu, Y. Wang, L. Wang, T. Yang, A. R. Jadhav, J. Zhang, Y. Wang, M. Wu, J. Y. Lee, M. G. Kim and H. Lee, ACS Nano, 2024, 18, 874–884.
28. H. Liu, G. Tan, M. Li, Z. Zhang, M. G. Sendeku, Y. Li, Y. Kuang and X. Sun, Chem. Eng. J., 2023, 458, 141414.
29. R. K. B. Karlsson, A. Cornell and L. G. M. Pettersson, J. Phys. Chem. C, 2016, 120, 7094–7102.

---