Optimizing LaNiO₃ surface structure for an efficient oxygen reduction reaction

Abstract

Perovskite oxides, with their flexible electronic structure and low cost, are highly attractive alternatives to noble metal catalysts for oxygen redox reactions (ORRs). Herein, we report LaNiO₃ perovskite as an efficient ORR catalyst, where oxygen vacancies and oxygen-containing functional groups work synergistically. LaNiO₃ was subjected to thermal shock in different media to introduce defects on the structural surface. Experimental results indicate that thermal shock in air creates oxygen vacancies on the catalyst surface, while thermal shock in aqueous solution increases the content of oxygen-containing functional groups (–OH) on the surface. The samples subjected to thermal shock exhibit superior ORR catalytic performance, with the limiting current density increasing from 4.2 mA cm⁻² to 5.4 mA cm⁻². This work provides a convenient and straightforward approach for constructing a series of materials enriched with surface defects.

---

Wen-Bin Luo

Wen-Bin Luo is a professor at the School of Metallurgy, Northeastern University. He obtained his PhD from the University of Wollongong, Australia, in 2015. Professor Luo has been recognized as a national-level distinguished young talent, a young top-tier talent under Liaoning Revitalization Talents Program, and a high-level leading talent in Shenyang City. Currently, he serves as the Director of the Institute for Energy Electrochemistry and Urban Mines Metallurgy at Northeastern University. His research interests focus on the development and application of materials for secondary ion batteries, metal-based batteries, and catalytic materials. He has published extensively in prestigious journals such as Journal of Materials Chemistry A, Advanced Materials, ACS Nano, Advanced Energy Materials, Advanced Functional Materials, eScience, ACS Catalysis, Energy Storage Materials, Nano Energy, and Journal of Energy Chemistry. Additionally, he has led numerous national, provincial, and ministerial-level projects, as well as university–industry collaborative initiatives.

---

Introduction

The finite nature of fossil fuel reserves and the depletion concerns associated with their excessive extraction to meet global energy demands have prompted the international community to seek alternative energy resources.^1–6 Moreover, the adverse environmental impacts of non-renewable energy sources, such as air pollution, have further driven the research community to explore clean and sustainable energy technologies. Among various energy storage technologies, air batteries have garnered considerable attention over the past few decades due to their high theoretical capacity compared to ion batteries. In these batteries, metals serve as anodes, while intercalation cathode materials are replaced by air (oxygen).^7–9 Despite the fact that platinum-based catalysts exhibit wonderful electrocatalytic activity in the oxygen reduction reaction (ORR), their high cost, poor durability, and susceptibility to methanol poisoning severely hinder their large-scale commercialization.¹⁰ Consequently, the development of non-precious metal catalysts that can replace Pt/C and exhibit superior ORR catalytic performance is of paramount importance. Perovskite-type oxides (ABO₃) with abundant electronic structures have been considered ideal ORR catalysts and have played a significant role in solid oxide fuel cells and metal–oxygen batteries.^11,12 However, the sluggish kinetics of perovskite oxides have also severely limited their ORR catalytic performance.¹³ In perovskite oxides, the A site is occupied by alkaline earth metals or rare earth metals, while the B site belongs to transition metals, particularly those with incomplete 3d shells.^14,15 The flexible crystal structure of perovskite oxides facilitates the incorporation of dopant cations, making it a common and straightforward strategy for optimization. Doping at the A site has been shown to alter the geometric and electronic structures of the catalyst or affect the oxygen content, thereby enhancing catalytic performance.^16,17 Transition metals have a flexible electronic structure and are therefore also a major factor in ORR performance.^13,18,19

ORR is a complex process involving multiple steps in aqueous solution, such as dissociation, adsorption, and electrochemical reactions, leading to the formation of oxygen-containing species like OH⁻, O₂⁻, and HO₂⁻.^20,21 Extensive studies have demonstrated that the effective adsorption of oxygen on the electrode surface, the dissociation of the O–O bond, and the extraction of oxide species significantly influence the overall kinetics of the ORR.²² The ORR can be regarded as an ensemble of oxygen exchange processes occurring on the catalyst surface. According to the Sabatier principle, catalytic activity can be modulated by altering the bond strength between the catalyst surface and the reactants/products.²³ However, this is merely a qualitative guideline, as the adsorption strength of oxygen on the catalyst is largely dependent on the various methods employed for catalyst preparation. From the perspective of molecular orbital theory, the hybridization of the d orbitals of the B-site metal and O 2p orbitals forms the antibonding σ* (e_g) and π* (t_2g) orbitals. The literature indicates that the electron filling of the e_g orbital of the B-site cations is directly related to the catalytic activity of ABO₃ perovskite oxides.^1,12,24 It is worth noting that, compared with cation doping, the introduction of oxygen vacancies or surface engineering has been considered a simpler and more effective method to enhance the activity of perovskite oxides recently.^21,25,26 This approach not only adjusts the hybridization between the transition metal 3d and O 2p orbitals but also, with appropriate surface engineering, enhances the conductivity and activates the adsorbed O₂.

Previous studies have identified LaNiO₃ as one of the most promising catalysts for the ORR;^27–29 however, its reaction kinetics remain severely limited. Herein, we report a strategy to significantly enhance the ORR electrocatalytic activity of LaNiO₃ (LNO) through thermal shock in different media, without any additional modifications. Among LNO samples subjected to thermal shock in various media, LNO-W (quenched in deionized water) exhibited the best ORR electrocatalytic performance. Compared with LNO-F (cooled in furnace naturally), quenching in air and solution increases the content of surface functional groups and oxygen vacancies. We further explored the mechanisms underlying the enhanced ORR catalytic activity.

Results and discussion

The crystal structures of the samples LNO-F, LNO-A-700 (naturally cooled to 700 °C in tubular furnaces, heat shock in the air), LNO-A (heat shock in the air at 800 °C), LNO-W, and LNO-N (heat shock in ammonium hydroxide at 800 °C) were confirmed by X-ray diffraction (XRD), as shown in Fig. 1a–d and S1.† All samples exhibited a single rhombohedral perovskite structure with the R3̄c space group (PDF#33-0711, (#167)), and no detectable impurity peaks. This indicates that thermal shock only alters the surface properties of the samples without affecting the perovskite structure seriously. The main diffraction peaks at 32.9, 47.3, and 58.7° of samples can be assigned to (110), (202), and (122) facets, respectively. The result contrasted with the report by Zhou's group, who synthesized cubic LaNiO₃ using similar materials and temperatures.³⁰ The discrepancy may arise from their heating of LaNiO₃ at different temperatures rather than directly subjecting it to thermal shock at 800 °C. Compared with LNO-F, the diffraction peaks of the samples subjected to thermal shock (LNO-A, LNO-W and LNO-A-700) shift toward lower diffraction angles (Fig. S1†), whereas LNO-A exhibits more negative shifts compared to LNO-W and LNO-A-700. This can be attributed to the formation of a higher concentration of oxygen vacancies on the sample surface induced by thermal shock. The loss of lattice oxygen leads to increased repulsion between cations, resulting in lattice expansion of the unit cell. The finding is further supported by the Rietveld refinement of XRD data for samples, and the results are presented in Tables S1–S5.† As shown in Fig. 1e, LNO-A exhibits significant lattice expansion (from a = b = 5.43 Å, c = 13.44 Å in LNO-F to a = b = 5.45 Å, c = 13.47 Å in LNO-A). The variation in lattice parameters mirrors alterations in the bonding characteristics of the elements. The Ni–O–Ni bond angle decrease and Ni–O bond stretching occurred in LNO-A and LNO-A-700. Conversely, catalysts subjected to thermal shock in solution exhibited completely opposite effects. This suggests the presence of two counteracting factors influencing catalyst structure. In this work, thermal shock regulates the overlap of the Ni 3d and O 2p electron clouds, thereby affecting the hybridization of the Ni–O bonds, by modulating the number of oxygen vacancies.^23,25

Fig. 1 (a–d) Rietveld refinement of XRD patterns, (e) lattice parameter for LNO-F, LNO-A, LNO-F, and LNO-N. The inset shows the crystal model of LaNiO₃.

The morphology of LNO samples was analysed by SEM and TEM. As shown in Fig. 2, all of LNO-F, LNO-A, LNO-W, and LNO-N exhibit irregular block-like structures, which are composed of nanosized particles with diameters ranging from approximately 40 to 160 nm. Notably, the particles of the samples subjected to thermal shock exhibited varying degrees of surface cracking, which promote O₂ transport in samples. Furthermore, Fig. 2b, d, f and h show locally magnified SEM images for LNO-F, LNO-A, LNO-W, and LNO-N, which show that the particles are more dispersed after thermal shock and more surfaces are exposed, especially in solution. The histograms presented in the inset show the particle size distributions obtained by counting at least 100 particles in SEM images. The result shows a significant decrease of particle size after thermal shock. Notably, the catalyst particles exhibited greater interparticle distances when subjected to thermal shock in solution, which may be associated with the etching effect of OH⁻ ions. This is expected to increase the specific surface area, and facilitate exposure of more active sites.^5,29,31

Fig. 2 (a, c, e and g) SEM micromorphology images, and (b, d, f and h) partial enlarged details, inset shows grain size distribution of LNO-F, LNO-A, LNO-W, and LNO-N.

Further insights into the microstructure of nanosized particles were investigated through TEM. Apparently, when subjected to thermal shock in different media, the particle morphology undergoes continuous changes. LNO-F and LNO-A particles exhibit regular hexagonal shapes, while the particles of LNO-W and LNO-N gradually become smooth (Fig. S2†). Notably, thermal shock not only induced cracks on macroscopic particles but also created fissures within nanoparticles, which is attributed to the refinement of particle size.

As shown in Fig. 3, the LNO-F particles exhibit good crystallinity both on the surface and in the bulk. In contrast, the particles of the samples subjected to thermal shock consist of well-developed crystals and a disordered surface layer. The result probably originated from the successful introduction of oxygen vacancies leading to the disintegration and discontinuous surface lattice structure.³² Moreover, the grains of LNO-A, LNO-W, and LNO-N contain a large number of structural defects, such as dislocations and stacking faults, as shown in Fig. 3c, f and i. The Geometric Phase Analysis (GPA) of the lattice reveals evident stress concentration at the locations of these structural defects. Stress concentration is a significant factor leading to grain refinement and changes in crystal orientation. Therefore, the particle diameter of the samples after thermal shock is significantly reduced, and the particles of LNO-W and LNO-N evolve into those composed of elongated subgrains (Fig. S2† and 3g). As shown in Fig. 3b and d, the lattice spacing corresponding to the (101) plane in LNO-A (0.398 nm) is significantly larger than that in LNO-W (0.386 nm), which is consistent with the Rietveld refinement of XRD results. The formation of oxygen vacancies increases the repulsion between cations, leading to cell expansion. However, on quenching in deionized water, some of the oxygen vacancies are occupied by oxygen-containing functional groups, which in turn mitigates the lattice expansion.²¹

Fig. 3 HRTEM micrographs of (a) LNO-F, (b and c) LNO-A, (d and f) LNO-W, and (g–i) LNO-N, and corresponding strain maps. The inset images show the corresponding fast Fourier transform (FFT) of LNO-N (I and II region).

The ORR involves a series of processes in which oxygen species are exchanged on the catalyst surface. To further investigate the characteristics of surface elemental composition and electronic structure of LaNiO₃ perovskite catalysts subjected to thermal shock in different media, XPS analysis is carried out. The survey spectra, shown in Fig. S3a,† indicate that all samples have the same elemental composition with no detectable impurities on the surface. To determine whether N element is present in the LNO-N catalyst, we performed XPS analysis on the N 1 s spectrum of LNO-N (Fig. S3b†). The results revealed negligible presence of N element on the sample surface. Therefore, we focused solely on the influence of the hydroxyl concentration in the ammonia solution on the experiments. The O 1s XPS spectra consist of four distinct peaks at 528.3 eV, 529.8 eV, 531.1 eV, and 533.1 eV, corresponding to LO, oxygen vacancy (VO), adsorbed oxygen, and H₂O, respectively (Fig. 4a and S4a†).^9,11,33,34 A significant increase in the content of oxygen vacancies was observed in LNO-A, indicating that a certain amount of oxygen defects had been effectively incorporated into the surface of LNO-A nanoparticles. Moreover, reducing the quenching temperature (LNO-A-700) results in a marked decrease in the proportion of oxygen vacancies on the sample surface. This suggests that oxygen vacancies are generated on the sample surface at high temperatures, and quenching in air does not provide sufficient time for these vacancies to recombine, leading to their retention.³⁵ Some studies have shown that the adsorption of HO₂⁻ by the oxygen vacancies and transition metal can reduce the dissociation of HO₂⁻, which is conducive to the further reaction to form OH⁻ and complete the 4 e⁻ reaction.²¹ Notably, the proportion of adsorbed oxygen species (e.g., hydroxylation) on the sample surface increased in LNO-W, while the proportion of oxygen vacancies decreased (Fig. S4b†). This may be due to H₂O molecules or OH⁻ in the solution occupying the positions of oxygen vacancies during the quenching process. It is noteworthy that for samples subjected to thermal shock in solution, the peak corresponding to adsorbed oxygen shifts towards higher binding energies, indicating an enhanced adsorption capacity for oxygen on the sample surface.³⁶ According to studies on the mechanism of the ORR, adsorbed oxygen is considered to be the most active site for the ORR.^10,34,37 Therefore, increasing the concentration of hydroxyl groups can effectively modulate the catalyst surface's adsorption capacity for intermediate products.

Fig. 4 (a) O 1s, (b) valence band, (c) Ni 3p XPS spectra of LNO-F, LNO-A, LNO-W, and LNO-N, (d) proposed mechanisms for the ORR in an alkaline medium.

The XPS spectra of Ni 2p and La 3d are shown in Fig. S5(a and b),† respectively. The analysis of the XPS results is quite complex because the intensity of Ni 2p is difficult to distinguish from other spectral features. The peaks of La 3p_3/2 and Ni 2p_3/2 overlap, as well as the Ni 2p_1/2 signal loss feature belonging to La 3p_3/2.^18,38 Therefore, we conducted a qualitative analysis of the La and Ni elements in this part. According to the literature, the peaks at 833.7 eV and 835.5 eV in the La 3d_5/2 spectrum are attributed to La₂O₃ (peak d) and La(OH)₃ (peak c), respectively.^27,38 The significant increase in the intensity of peak c in LNO-W and LNO-N indicates an enhanced degree of hydroxylation. To further analyse the variation for Ni, the XPS spectrum of Ni 3p is presented in Fig. 4c. Compared to LNO-F, the Ni 3p peak of samples subjected to thermal shock shifts towards higher binding energies. Particularly in LNO-W and LNO-N, the Ni 3p peak undergoes a significant leftward shift, consistent with the trend observed for the Ni 2p peak. This indicates an increased proportion of Ni³⁺ in the samples. The presence of more empty orbitals in Ni³⁺ effectively facilitates electron transfer, thereby accelerating the ORR process. Moreover, the higher electronegativity of Ni³⁺ enables efficient adsorption of intermediates, enhancing the ORR performance.³⁹

The electronic structure near the top of the valence band is significant for ORR performance. The valence band spectra of the perovskite samples cooled in different media are shown in Fig. 4b. Generally, valence band spectra can be divided into four parts, and associated with the contributions of hybridized states, including Ni 3d–O 2p, O 2p non-bonding states, Ni t_2g states, and Ni 3d e_g states.^18,23 The ORR process involves the transformation of various intermediates, with the rate-limiting step being the displacement of surface hydroxyl groups (highlighted in red in Fig. 4d). The kinetics of this step are determined by the driving force of electron transfer from the e_g orbitals to the O–O antibonding orbitals. It is observed that an increase in the proportion of oxygen vacancies leads to a shift of the O 2p non-bonding states towards the Fermi level in LNO-A, thereby reducing the charge transfer gap in the catalytic process. Moreover, the rise in the proportion of oxygen vacancies results in a decrease in the electron density of O 2p and Ni t_2g, while enhancing the hybridization degree of Ni 3d–O 2p orbitals. This, in turn, causes a reduction in the Ni–O–Ni bond angle, which is consistent with the results of Rietveld refinement of XRD.^31,40 Notably, the changes in valence electrons in LNO-N are exactly opposite to those in LNO-A, indicating that oxygen vacancies and hydroxyl functional groups play different roles in ORR catalysis.

Electrochemical ORR properties of LNO-F, LNO-A-700, LNO-A, LNO-W, and LNO-N were characterized. Fig. S6† shows the ORR activity of the pristine nanoparticles and nanoparticles subjected to thermal shock at different rotation speeds. Based on previous experimental conclusions, acetylene black (AB) only serves to improve the electrical conductivity of the active material and has minimal impact on the ORR activity. We used linear sweep voltammetry (LSV) based on the rotating ring-disk electrode (RRDE) to represent ORR activity and evaluated ORR performance using limiting current density (j_m), onset potential (E_onset), and Tafel slope. The E_onset was defined as the potential at which the ORR current density reaches 0.1 mA cm⁻². At a rotation speed of 1600 rpm and a scan rate of 10 mV s⁻¹, the ORR LSV curves (Fig. 5a) were measured between −0.7 and 0 V vs. Ag|AgCl (3 M KCl). It clearly demonstrates that the ORR activity of the LaNiO₃ powders subjected to thermal shock was significantly enhanced, as evidenced by their more positive E_onset and increased j_m (Fig. 5d). Among the particles subjected to thermal shock, LNO-W showed the best ORR activity, with the half-wave potential increasing from 0.62 V to 0.72 V vs. RHE. In particular, the onset potentials of pristine LNO-F, LNO-A-700, LNO-A, LNO-W, and LNO-N were 0.787 V, 0.793 V, 0.799 V, 0.823 V, and 0.815 V vs. RHE, respectively, with LNO-W exhibiting the highest j_m (5.42 mA cm⁻²). To gain further insights into the observed electrochemical responses, we analysed the ORR process using Tafel slopes, which provide a kinetic perspective. For LNO-F, the ORR Tafel slope decreased from 93.1 mV dec⁻¹ to 82.2 mV dec⁻¹, indicating superior kinetics for LNO-W compared to LNO-A and LNO-F. The Koutecky–Levich equation is used to evaluate the number of electrons transferred during the ORR process.^10,32 The electron transfer number of pristine LNO-F is approximately 3.63, while the electron transfer number of LNO-W increased to 3.98, which is very close to that of commercial Pt/C.³² This suggests that LaNiO₃ subjected to thermal shock tends to experience a four-electron transfer reaction, reducing oxygen to an oxygen-containing material. The significant increase in the electron transfer number after thermal shock indicates that oxygen vacancies in conjunction with hydroxylation can facilitate a four-electron reaction.

Fig. 5 ORR performance of LNO-F, LNO-A, LNO-W, and LNO-N. (a) LSV curves, (b) Tafel slope, (c) Koutechy–Levich (K–L) plots, (d) onset potential and electron transfer number.

The results clearly demonstrate that thermal shock treatment significantly enhances the ORR catalytic activity of LaNiO₃ nanosized powders, representing an effective approach to enhancing the electrocatalytic oxygen reduction efficiency of LaNiO₃. We attribute this enhancement primarily to the introduction of oxygen vacancies and oxygen-containing functional groups on the particle surfaces, which subsequently enhance intrinsic properties such as oxygen exchange kinetics.^17,41 Notably, the overall ORR electrocatalytic performance of LNO-W surpasses that of LNO-A and LNO-N, indicating that the enhanced electrochemical performance of LNO-W results from a delicate balance between oxygen vacancies and oxygen-containing functional groups (hydroxylation). The increase in oxygen vacancy concentration leads to a shift of the O 2p non-bonding states towards the Fermi level, effectively reducing the charge transfer energy gap and enhancing the hybridization degree of Ni 3d–O 2p orbitals, thereby improving ORR performance. Conversely, an increase in the number of hydroxyl functional groups on the perovskite surface can effectively raise the concentration of adsorbed oxygen, which typically acts as the active species in the ORR process. However, hydroxyl functional groups tend to decrease the hybridization degree of Ni 3d–O 2p orbitals and enlarge the charge transfer energy gap. Therefore, there exists a synergistic effect between oxygen vacancies and oxygen-containing functional groups, which collectively optimize the ORR process. And there is an optimal surface oxygen vacancy and oxygen-containing functional group concentration that enhance oxygen exchange kinetics. These defects have a significant impact on the electronic structure of LaNiO₃ surface, leading to an increased number of active sites. Moreover, the incorporation of defects induces changes in the oxidation states of cations, which increases the electrical conductivity of the LaNiO₃ catalyst.

Conclusions

In summary, in this work, LaNiO₃ perovskite oxides were synthesized via the sol–gel method and subjected to thermal shock in different media to obtain LNO-F, LNO-W, LNO-A, and LNO-N. Their morphology and structure were characterized using XRD, XPS, TEM, SEM, and other techniques. Compared with LNO-F, LNO-W exhibited better ORR activity for three reasons: (i) the presence of abundant oxygen vacancy on the surface, which can act as an active site and facilitate the adsorption and nucleophilic reaction of O₂, (ii) the synergistic effect of hydroxylation and oxygen vacancy can improve the reaction efficiency and increase the 4 e⁻ reaction and improve the ORR reaction kinetics, and (iii) the uniform distribution and particle refinement of nanoparticles, which increases the specific surface area. Therefore, the thermal shock method can be used to achieve controllable surface structural engineering in LaNiO₃ nanosized powders. This strategy introduces defects into the surface of LaNiO₃ nanoparticles without affecting the bulk structure, thereby achieving higher catalytic activity and better stability in the ORR. Moreover, the thermal shock method is characterized by its high efficiency, high controllability, strong functionality, and ease of operation. Additionally, this method can be used to construct a series of materials with surface-rich defects, which can be applied in the fields of biology, environmental protection, and energy conversion and storage.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and its ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 52272194). The authors extend their gratitude to Shiyanjia Lab (http://www.shiyanjia.com) for providing invaluable assistance with the XPS analysis. This manuscript was written through the contributions of all the authors. All authors have given approval to the final version of the manuscript.

References

1. D. A. Agyeman, Y. Zheng, T.-H. Lee, M. Park, W. Tamakloe, G.-H. Lee, H. W. Jang, K. Cho and Y.-M. Kang, ACS Catal., 2020, 11, 424–434.
2. N. Bonmassar, M. F. Bekheet, L. Schlicker, A. Gili, A. Gurlo, A. Doran, Y. Gao, M. Heggen, J. Bernardi, B. Klötzer and S. Penner, ACS Catal., 2019, 10, 1102–1112.
3. P. D. Lund, B. Zhu, Y. Li, S. Yun, A. G. Nasibulin, R. Raza, M. Leskelä, M. Ni, Y. Wu, G. Chen, L. Fan, J.-S. Kim, S. Basu, T. Kallio and I. Pamuk, ACS Energy Lett., 2017, 2, 2752–2755.
4. B. Hua, Y. Q. Zhang, N. Yan, M. Li, Y. F. Sun, J. Chen, J. Li and J. L. Luo, Adv. Funct. Mater., 2016, 26, 4106–4112.
5. O. Celikbilek, C.-A. Thieu, F. Agnese, E. Calì, C. Lenser, N. H. Menzler, J.-W. Son, S. J. Skinner and E. Djurado, J. Mater. Chem. A, 2019, 7, 25102–25111.
6. D. Kim, J. W. Park, M. S. Chae, I. Jeong, J. H. Park, K. J. Kim, J. J. Lee, C. Jung, C.-W. Lee, S.-T. Hong and K. T. Lee, J. Mater. Chem. A, 2021, 9, 5507–5521.
7. L. Li, H. Yang, Z. Gao, Y. Zhang, F. Dong, G. Yang, M. Ni and Z. Lin, J. Mater. Chem. A, 2019, 7, 12343–12349.
8. W. G. Hardin, D. A. Slanac, X. Wang, S. Dai, K. P. Johnston and K. J. Stevenson, J. Phys. Chem. Lett., 2013, 4, 1254–1259.
9. L. Gao, Q. Li, L. Sun, T. Xia, L. Huo, H. Zhao and J.-C. Grenier, J. Mater. Chem. A, 2018, 6, 15221–15229.
10. J. Sun, L. Du, B. Sun, G. Han, Y. Ma, J. Wang, H. Huo, C. Du and G. Yin, ACS Appl. Mater. Interfaces, 2020, 12, 24717–24725.
11. P. Sennu, V. Aravindan, K. S. Nahm and Y.-S. Lee, J. Mater. Chem. A, 2017, 5, 18029–18037.
12. R. Majee, T. Das, S. Chakraborty and S. Bhattacharyya, ACS Appl. Mater. Interfaces, 2020, 12, 40355–40363.
13. M. Retuerto, F. Calle-Vallejo, L. Pascual, G. Lumbeeck, M. T. Fernandez-Diaz, M. Croft, J. Gopalakrishnan, M. A. Peña, J. Hadermann, M. Greenblatt and S. Rojas, ACS Appl. Mater. Interfaces, 2019, 11, 21454–21464.
14. M. I. Hossain, S. A. Monim, Z. M. Moushumy, N. H. Khansur, I. Mahmud, M. Rahaman, A. Aldalbahi, S. H. Firoz and M. A. Hasnat, ACS Appl. Energy Mater., 2025, 8, 949–963.
15. C. Ozgur, T. Erdil, U. Geyikci, I. Yildiz, E. Lokcu and C. Toparli, ChemPhysChem, 2024, 25, e202400531.
16. W. Zhang, N. Zhang, W. Zhu, L. Zhao, Y. Gong, J. Jin, R. Wang, H. Wang and B. He, Small, 2025, 2500272.
17. D. Shang, B. Zhang, L. Zhang, J. Li, Y. Zhao, Y. Huang, K. Tang and C. Xia, Chem. Eng. J., 2025, 506, 125516.
18. H.-S. Lee, C.-S. Park and H.-H. Park, Appl. Phys. Lett., 2014, 104, 191604.
19. M. Li, W. Zhou, X. Xu and Z. Zhu, J. Mater. Chem. A, 2013, 1, 118034.
20. Y. Dai, J. Yu, C. Cheng, P. Tan and M. Ni, Chem. Eng. J., 2020, 397, 125516.
21. Q. Ji, L. Bi, J. Zhang, H. Cao and X. S. Zhao, Energy Environ. Sci., 2020, 13, 1408–1428.
22. M. Han, I. Gómez-Recio, D. G. Martín, N. Ortiz Peña, M. L. Ruiz-González, M. Selmane, J. M. González-Calbet, O. Ersen, A. Zitolo, B. Lassalle-Kaiser, D. Portehault and C. Laberty-Robert, ACS Catal., 2023, 13, 5733–5743.
23. Y. Sun, Z. Liu, W. Zhang, X. Chu, Y. Cong, K. Huang and S. Feng, Small, 2018, 15, 1803513.
24. H. Zhang, Y. Xu, M. Lu, X. Xie and L. Huang, Adv. Funct. Mater., 2021, 31, 2101872.
25. C. Dong, Z. W. Liu, J. Y. Liu, W. C. Wang, L. Cui, R. C. Luo, H. L. Guo, X. L. Zheng, S. Z. Qiao, X. W. Du and J. Yang, Small, 2017, 13, 1803513.
26. G. Nandikes, P. Pathak, M. Karthikeyan, A. A. M. Abahussain and L. Singh, Int. J. Hydrogen Energy, 2024, 52, 627–641.
27. H. Zhang, Z. Luo, Y. Liu and Y. Jiang, Appl. Catal., B, 2020, 277, 119166.
28. J. R. Petrie, V. R. Cooper, J. W. Freeland, T. L. Meyer, Z. Zhang, D. A. Lutterman and H. N. Lee, J. Am. Chem. Soc., 2016, 138, 2488–2491.
29. M. Retuerto, A. G. Pereira, F. J. Pérez-Alonso, M. A. Peña, J. L. G. Fierro, J. A. Alonso, M. T. Fernández-Díaz, L. Pascual and S. Rojas, Appl. Catal., B, 2017, 203, 363–371.
30. W. Zhou and J. Sunarso, J. Phys. Chem. Lett., 2013, 4, 2982–2988.
31. R. Li, J. Cao, H. Chen, Y. Qi and Y. Xie, Int. J. Hydrogen Energy, 2023, 48, 28061–28069.
32. G. Ou, C. Yang, Y. Liang, N. Hussain, B. Ge, K. Huang, Y. Xu, H. Wei, R. Zhang and H. Wu, Small Methods, 2018, 3, 1800279.
33. Q. Wang, J. Hou, Y. Fan, X.-A. Xi, J. Li, Y. Lu, G. Huo, L. Shao, X.-Z. Fu and J.-L. Luo, J. Mater. Chem. A, 2020, 8, 7704–7712.
34. A. Tyagi, Y. K. Penke, P. Sinha, I. Malik, K. K. Kar, J. Ramkumar and H. Yokoi, Int. J. Hydrogen Energy, 2021, 46, 22434–22445.
35. B. Yao, Y. Dou, T. Xia, T. Song, Q. Li, H. Zhao and L. Huo, Mater. Sci. Eng., B, 2025, 314, 118034.
36. Y.-Q. Lyu and F. Ciucci, ACS Appl. Mater. Interfaces, 2017, 9, 35829–35836.
37. K. Li, J. Gao, X. Han, Q. Shao and Z. Lü, J. Energy Chem., 2024, 96, 669–678.
38. R. C. Rabelo-Neto, H. B. E. Sales, C. V. M. Inocêncio, E. Varga, A. Oszko, A. Erdohelyi, F. B. Noronha and L. V. Mattos, Appl. Catal., B, 2018, 221, 349–361.
39. G. Fu, R. Hou, L. Sun, H. Liu, Y. Wei, R. Wei, X. Meng, S. Zhang and B. Yang, ACS Appl. Mater. Interfaces, 2025, 17, 7860–7869.
40. L. Sun, T. Sun, X. Li and Y. Wang, ACS Appl. Mater. Interfaces, 2025, 17, 7716–7727.
41. W. Biao, Q. Xu, G. Wanbing, L. Zhijun, W. Fang and Z. Qiuju, Int. J. Hydrogen Energy, 2023, 48, 18744–18752.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta01346a

‡ These authors contributed equally to this work.

---