Creating an electron-rich region on ultrafine Bi₂O₃ nanoparticles to boost the electrochemical carbon dioxide reduction to formate

Abstract

Various metal–support composites have been reported for the electrochemical CO₂ reduction reaction (ECRR), while the effect of synergy between the size effect and interface regulation on ECRR performance is still rarely explored. Herein, we report ultrafine Bi₂O₃ nanoparticles (∼1.9 nm) decorated on a nitrogen-vacancy-rich graphitic carbon nitride (Bi₂O₃/NV-C₃N₄) electrocatalyst for highly efficient ECRR to formate. The Bi₂O₃/NV-C₃N₄ electrocatalyst can achieve 95% formate selectivity at −0.9 V vs. RHE and 24 h long-term stability. Moreover, substantial formate selectivity (>80%) was gained over a wide potential range of more than 700 mV. The excellent performance was attributed to the electron transfer from NVs to Bi₂O₃, forming an electron-rich region on Bi₂O₃ and accelerating the electron transfer from Bi₂O₃ to the adsorbed intermediates. This work provides a strategy for enhancing the catalytic activity of bismuth oxides to achieve a high efficiency for CO₂ reduction to generate liquid products.

---

1 Introduction

Electrochemical reduction of CO₂ into high value-added products driven by renewable electricity is an effective way to facilitate CO₂ recycling.¹ Formate is an important liquid product that is easy to separate during preparation and can be used as a crucial chemical intermediate for industrial applications.² Over the past few years, p-block metal Bi-based electrocatalysts have exhibited preference for formate production, but high Faraday efficiency (FE) can only be obtained at a certain potential or in a confined potential field. Once the applied potentials are more negative than the optimum value (for achieving the largest FE), FE for formate (FE_formate) rapidly drops due to the significant enhancement of the competing hydrogen evolution reaction (HER).

To maintain a high FE_formate over a broad potential range, great efforts have been devoted to modify the structure of Bi-based electrocatalysts, considering that nanostructuring is an effective strategy to improve the performance of metal-based materials. A series of metallic bismuth (Bi⁰) materials with delicate nanostructures, such as Bi dendrites, Bi nanosheets, and Bi nanoflakes have been prepared. While these metallic Bi-based composites inevitably contain oxygen species, their effects on the electrochemical CO₂ reduction reaction (ECRR) are not well understood.

Since it is hard to explain the role of oxygen species on the enhancement activity of CO₂, Xia³ and co-workers strengthened the oxygen species in Bi-based materials in turn and found that the Bi–O structure in Bi₂O₃ is favorable to improve the efficiency of CO₂ reduction to formate. Meanwhile, various bismuth oxide composites have also been reported. However, those catalysts usually have a large particle size, which may reach up to several micrometers. Theoretical and experimental investigations have revealed that ultrafine nanoparticles (NPs) containing numerous coordinately unsaturated sites favor the stabilization of substrates and adsorption of intermediates on the catalyst surface and therefore show great application potential in the ECRR.⁴ Unfortunately, there have been limited studies about bismuth oxide nanoparticles smaller than 5 nm employed for the ECRR due to the lack of an effective synthesis method. Apart from precisely tuning the size of catalysts, there is another problem in the design of catalysts: ultrafine NPs tend to agglomerate in a non-controlled manner resulting in a rapid decay in the cyclability under electrochemical conditions.

The structure design of supported catalysts offers a good reference for the enhancement of stability of small sized NPs. Graphitic carbon nitride (C₃N₄) has been attracting much attention and widely explored as a support for a couple of reactions.⁵ The higher electron density around N atoms can induce strong interactions with metal precursor ions to promote oxide growth on the support forming a high-quality oxide-support interface.⁶ In addition, the formed interface is proposed to redistribute electrons of the support and catalyst and therefore affects the performance of the ECRR. However, the effect of electronic state modulation of supports on ECRR performance is still rarely explored for C₃N₄-supported catalysts.

Herein, we report ultrafine Bi₂O₃ nanoparticles (∼1.9 nm) decorated on a nitrogen-vacancy-rich graphitic carbon nitride (Bi₂O₃/NV-C₃N₄) electrocatalyst for highly efficient electrochemical reduction of CO₂ to formate. The Bi₂O₃/NV-C₃N₄ electrocatalyst can achieve 95% formate selectivity at −0.9 V vs. RHE and 24 h long-term stability. Moreover, substantial formate selectivity (>80%) was gained over a wide potential scope of more than 700 mV. The excellent catalytic performance was attributed to the following reasons: (i) nitrogen vacancies in C₃N₄ effectively lower the energy barriers for *CO₂ formation and (ii) the formed metal-carbon heterojunction structure effectively accelerates electron transfer from NV-C₃N₄ to Bi₂O₃ active sites, and therefore facilitates intermediate *COO generation. This work may pave the way for bismuth oxide catalysts to be applied in different renewable energy-conversion devices.

2 Experimental

2.1 Synthesis of NV-C₃N₄

The synthesis of bulk C₃N₄ was performed by calcining melamine and urea with a mass ratio of 1 : 3 at 550 °C for 2 h in an air atmosphere. The C₃N₄ nanosheets were obtained by liquid exfoliation of bulk C₃N₄ in absolute ethanol for 3 h. NV-C₃N₄ was prepared by calcining the as-prepared C₃N₄ nanosheets in a tube furnace under an Ar atmosphere at 650 °C for 4 h.

2.2 Synthesis of Bi₂O₃/NV-C₃N₄

Bi₂O₃/NV-C₃N₄ was synthesized by the solvothermal method with NV-C₃N₄ as the substrate.⁷ Specifically, 680 mg Bi(NO₃)₃·5H₂O was dissolved in a mixture of 24 mL ethanol, 4 mL deionized water and 12 mL ethylene glycol named solution A; 20 mg NV-C₃N₄ was dissolved in 20 mL ethanol named solution B. Solution A was slowly added to solution B, and the mixture was stirred in a water bath at 25 °C for 15 min and then sonicated for 15 min. The mixed solution was placed in a 100 mL Teflon-lined stainless-steel autoclave and kept at 160 °C for 5 h. After cooling to room temperature, the product was centrifuged and washed four times with deionized water and ethanol, respectively. Finally, the samples were dried in a vacuum drying oven at 50 °C for 10 h. The synthesis method of Bi₂O₃/C₃N₄ is the same as that of Bi₂O₃/NV-C₃N₄, except that NV-C₃N₄ is replaced by C₃N₄ nanosheets. The Bi₂O₃ nanosheets were synthesized without substrates.

2.3 Electrochemical analysis

The as-prepared sample (4 mg) was evenly dispersed in a mixed solution (950 μL) with 475 μL ethanol and 475 μL deionized water. Subsequently, 50 μL Nafion (5 wt%) was added followed by ultrasonic treatment for 0.5 h forming a homogeneous ink. Finally, 50 μL ink was dropped onto carbon paper (1 × 0.5 cm²) and dried at room temperature. The electrochemical measurement was performed in a gas-tight H-type cell containing two chambers separated by a proton exchange membrane (Nafion 117) between the cathodic and anodic compartments, each chamber containing 35 mL of electrolyte and 35 mL space, and tested using a Chenhua workstation CHI 760 and an equipped three-electrode system.

2.4 Product analysis

The liquid phase product was detected by ion chromatography (CIC-D120). The standard curve of formate is shown in Fig. S1.† The faradaic efficiency (FE) of formate is obtained using eqn (1):96 500 is the Faraday constant, c is the concentration of formate and Q is the number of accumulated charges.

2.5 DFT calculations

All density functional theory (DFT) calculations in this work were conducted using the Vienna ab initio simulation package (VASP).⁸ The ECRR calculations were performed on the catalytic materials based on the projector augmented wave (PAW) method.⁹ The free energy of each intermediate was calculated with the Perdew–Burke–Ernzerhof (PBE)¹⁰ functional of the generalized gradient approximation method (GGA). Convergence conditions: cutoff energy of 400 eV, convergence energy threshold of 10⁻⁵ eV, remaining forces of less than 0.01 eV Å⁻¹. The Brillouin zone uses 3 × 1 × 1 Monkhorst–Pack k-points for sampling, and the supercell is 3 × 3 × 1. The model is separated by a 15 Å vacuum layer. All were corrected using van der Waals (vdW) forces.

3 Results and discussion

The standard synthetic route of Bi₂O₃/NV-C₃N₄ is depicted in Fig. 1a. C₃N₄ nanosheets were first obtained by calcining a mixture of urea and melamine in air, and then by liquid exfoliation. Nitrogen-vacancy-rich C₃N₄ (NV-C₃N₄) was then obtained by annealing the as-prepared C₃N₄ nanosheets in Ar. Finally, Bi₂O₃ nanoparticles were loaded onto C₃N₄ and NV-C₃N₄ nanosheets by a simple hydrothermal method. The obtained samples were denoted as Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄, respectively. For comparison, Bi₂O₃ nanosheets were also synthesized without substrates.

Fig. 1 (a) Schematic diagram of the synthetic process of Bi₂O₃/NV-C₃N₄; (b) atomic structure of NV-C₃N₄; (c and d) TEM images of Bi₂O₃/NV-C₃N₄; Statistical analysis based on 100 Bi₂O₃ nanoparticles indicates that the average size of the ultrafine Bi₂O₃ nanoparticles is around 1.9 nm with a narrow distribution; (e) XRD patterns of Bi₂O₃, Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄; (f) Bi 4f spectra of Bi₂O₃, Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄; (g) N 1s spectra of Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄.

Electron paramagnetic resonance (EPR) and X-ray photoelectron spectroscopy (XPS) measurements were first conducted to disclose the NV in C₃N₄. In EPR spectra (Fig. S2†), the intensity signal at g = 2.0042 of NV-C₃N₄ was significantly stronger than that of C₃N₄, indicating the successful introduction of NVs. The remarkable decrease in the N/C ratio from 1.34 (C₃N₄) to 1.22 (NV-C₃N₄), further demonstrated that significant NVs were generated on NV-C₃N₄ (Fig. S3a and Table S1†). The C/N atomic ratio in C₃N₄ is 0.743 close to 0.75 (Fig. S3b†), indicating that we have synthesized C₃N₄ rather than porous carbon.^7,11 Deconvolution of the N 1s spectrum shows three high-resolution peaks of C–N=C (398.4 eV), N(C)₃ (400.0 eV) and C–N–H (401.5 eV). The quantitative evaluation demonstrates that the percent ratio of reduces from 2.14 (C₃N₄) to 2.00 (NV-C₃N₄), revealing that N–C₂ (two-coordinated-N) vacancies are mainly formed on NV-C₃N₄ rather than N–C₃ (three-coordinated-N) vacancies, as illustrated in Fig. 1b. For both C₃N₄ and NV-C₃N₄ nanosheets, a dominant (002) X-ray diffraction (XRD) peak appears at approximately 27.5°, corresponding to the interlayer reflection of the graphitic structure (Fig. S4a†). Compared with that of C₃N₄, the (002) peak of NV-C₃N₄ is relatively broadened, indicating that a disordered structure was built in the atomic arrangement of NV-C₃N₄.^7,12 In addition, N₂ desorption and adsorption isotherms (Fig. S4b†) indicate that NV-C₃N₄ (202.3 m² g⁻¹) has a larger specific surface area than C₃N₄ (77.3 m² g⁻¹) nanosheets, probably due to the introduction of NVs to generate more wrinkles. High specific surface area is generally beneficial for enhanced electrocatalytic activity, facilitating electrolyte/reactant contact by exposing more sites of geometric activity.

Transmission electron microscopy (TEM) images of Bi₂O₃/NV-C₃N₄ disclose that ultrafine metallic NPs with a narrow size distribution and a small average size of ∼1.9 nm are homogeneously dispersed on the porous C₃N₄ support (Fig. 1c and d). The elemental mapping results showed that Bi, C, N, and O elements were uniformly distributed throughout the catalyst (Fig. S5†). In addition, Bi₂O₃/C₃N₄ shares a similar morphology with Bi₂O₃/NV-C₃N₄ (Fig. S6†). XRD patterns of Bi₂O₃/NV-C₃N₄, Bi₂O₃/C₃N₄ and Bi₂O₃ samples reveal two main diffraction peaks at ∼28° and 43°, which can be indexed to the (201) and (002) planes of the Bi₂O₃ crystal phase (JCPDS No. 27-0052), respectively (Fig. 1e). These results indicate that high-density ultrafine Bi₂O₃ nanoparticles were successfully decorated on C₃N₄ and NV-C₃N₄ supports and both samples show similar morphology, particle size and chemical structure.

The electronic interactions between Bi₂O₃ and the C₃N₄ support were first investigated by X-ray photoelectron spectroscopy (XPS). As shown in Fig. 1f, the gradual shifts of typical Bi 4f_5/2 and 4f_7/2 peaks to lower energies indicate the gradually richer electron density of Bi₂O₃ in Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄ samples. The larger shift implies relatively strong electron donating properties of NV-C₃N₄. For N 1s spectra, the energy peak positions of pyridine N in Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄ shift to higher energies compared with that in C₃N₄, indicating the gradually depressed electron density of N in the C₃N₄ support after Bi₂O₃ decoration (Fig. 1g). We performed UV photoelectron spectroscopy (UPS, Fig. S7†) and the results showed that the work function (Φ) value of Bi₂O₃/NV-C₃N₄ is 6.67. This value is lower than that of Bi₂O₃/C₃N₄ (6.86), which means that Bi₂O₃/NV-C₃N₄ is more efficient than Bi₂O₃/C₃N₄ for electron donation. As shown in Fig. S8,† the charge distribution at the valence band maximum (VBM) and conduction band minimum (CBM) was calculated. It is obvious that the charge density on Bi₂O₃/NV-C₃N₄ is significantly enhanced by the introduction of N vacancies. Those results mean a spontaneous electron donation from the NV-C₃N₄ and C₃N₄ supports to the Bi₂O₃ NPs at their coupling interface, which can potentially improve the ECRR performance.

To evaluate the ECRR performance of Bi₂O₃, Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄, electrochemical measurements were conducted using a three-electrode system in a tight H-type cell. The linear cyclic voltammetry (LSV) tests show that the current density of Bi₂O₃/NV-C₃N₄ is significantly higher than that of Bi₂O₃ and Bi₂O₃/C₃N₄, indicating that Bi₂O₃/NV-C₃N₄ has excellent CO₂ reduction activity (Fig. 2a). For instance, Bi₂O₃/NV-C₃N₄ displays a current density of 18.8 mA cm⁻² at −1.2 V, which was roughly 1.5 and 2.3 times larger than that of Bi₂O₃/C₃N₄ and Bi₂O₃, respectively. In particular, the current densities of Bi₂O₃/NV-C₃N₄ electrodes under CO₂ start to accelerate at −0.8 V, which is more positive than that of the Bi₂O₃/C₃N₄ electrode (−1.0 V), indicating an improvement in the electrocatalytic activities of CO₂ reduction by introducing NVs into the substrate. The formate faradaic efficiency (FE_formate) analysis (Fig. 2b and S9†) indicates that all supported catalysts present an improved FE_formate compared with Bi₂O₃ throughout the potential range. The FE_formate order is Bi₂O₃/NV-C₃N₄ > Bi₂O₃/C₃N₄ > Bi₂O₃, which can be attributed to the interfacial effect. Bi₂O₃/NV-C₃N₄ shows the highest FE_formate of 95% at −1.2 V, which is better than those of Bi₂O₃/C₃N₄ (88.2%) and Bi₂O₃ (81.2%). More importantly, the FE_formate of Bi₂O₃/NV-C₃N₄ shows high selectivity of more than 80% over a large potential window from −1.0 to −1.7 V. The operation potential window for the high formate formation efficiency of Bi₂O₃/NV-C₃N₄ is outstanding in 0.1 M KHCO₃ among the reported electrocatalysts (Table S2† and Fig. 2c).¹³ The FE of H₂ is opposite to that of formate (Fig. S10a†), indicating that the hydrogen evolution reaction (HER) is suppressed. The FE of CO is considerably small, below 10% at all potentials (Fig. S10b†). In addition, no other liquid product was detected except for formate (Fig. S11†). The partial current densities for formate formation (j_formate) of all the samples were extracted from total current densities based on FE_formate (Fig. S9†). Among the three samples, Bi₂O₃/NV-C₃N₄ shows the highest current density for formate formation at the measured potential scope. For example, the j_formate of Bi₂O₃/NV-C₃N₄ reaches −18 mA cm⁻² at −1.2 V, which is 1.61 and 2.64 fold higher than that of Bi₂O₃/C₃N₄ and Bi₂O₃, respectively.

Fig. 2 (a) LSV curves with CO₂ and Ar-saturated 0.1 M KHCO₃ electrolyte; (b) FEs of formate formation; (c) comparison of stability and potential windows with those of other reported electrocatalysts. (d) Partial current densities of formate normalized by ECSA; (e) Tafel slope for Bi₂O₃, Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄; (f) the stability of Bi₂O₃/NV-C₃N₄ at −1.2 V.

To evaluate the intrinsic ECRR activities of the prepared catalysts, the electrochemical active surface areas (ECSAs) of Bi₂O₃, Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄ were evaluated (Fig. S12†). The capacitances of Bi₂O₃/NV-C₃N₄, Bi₂O₃/C₃N₄ and Bi₂O₃ are 0.69 mF cm⁻², 3.01 mF cm⁻² and 2.31 mF cm⁻², respectively. When the current density of formate was renormalized by ECSA, the current density of Bi₂O₃/NV-C₃N₄ remained the largest at various potentials, indicating that Bi₂O₃/NV-C₃N₄ has the highest intrinsic catalytic activity (Fig. 2d). The kinetic process of the ECRR to formate transition is further explored via Tafel plots (Fig. 2e). The Tafel slope of Bi₂O₃/NV-C₃N₄ is 253 mV dec⁻¹, much smaller than those of Bi₂O₃/C₃N₄ (303 mV dec⁻¹) and Bi₂O₃ (313 mV dec⁻¹), suggesting a more favorable reaction kinetics for the ECRR to formate on Bi₂O₃/NV-C₃N₄. The evaluation of catalytic performance suggests that the electron-rich region around Bi₂O₃ induced by electron transfer from supports to Bi₂O₃ accelerates ECRR kinetics resulting in the excellent performance of Bi₂O₃/NV-C₃N₄.

In addition to the electrocatalytic activity, stability is also a significant metric for evaluating advanced catalysts. For durable electrolysis of Bi₂O₃/NV-C₃N₄, the electrolyte was changed every four hours to test the FE_formate. As shown in Fig. 2f, Bi₂O₃/NV-C₃N₄ was tested at −1.2 V for 24 h and the FE_formate remains above 90%. In addition, the catalysts after electrolysis were characterized. TEM after electrolysis revealed that Bi₂O₃ nanoparticles exhibited slight agglomeration (Fig. S13a†), which may be the main factor for the slight performance degradation. Furthermore, XRD after electrolysis still preserves the well-preserved Bi₂O₃ crystal phase (Fig. S13b†), indicating the good stability of the catalyst.

To further study the ECRR mechanism and confirm the proposed reasons for the enhanced catalytic performance, we modeled three materials Bi₂O₃, Bi₂O₃/C₃N₄ and Bi₂O₃/NV-C₃N₄ (Fig. S14a, b† and 3a) and performed density functional theory (DFT) calculations (Fig. 3b). On the whole, the ΔG_COO of the ECRR for the three catalysts is as follows: Bi₂O₃ > Bi₂O₃/C₃N₄ > Bi₂O₃/NV-C₃N₄. For Bi₂O₃, the ΔG_COO of CO₂ adsorption can reach 2.97 eV. After Bi₂O₃ was decorated on C₃N₄, the ΔG_COO was reduced by 1.70 eV, but subsequent *HCOOH formation processes cannot proceed spontaneously. The introduction of NVs on C₃N₄ not only lowers the ΔG_COO of *COO adsorption, but also allows the formation process of *HCOOH to proceed spontaneously. For Bi₂O₃/NV-C₃N₄, both free energies for *COO and *HCOOH adsorption were reduced, and the formation of *OCHO turns into a decisive speed step, and only 0.60 eV was needed for this step. In addition, the free energy of the HER was calculated (Fig. 3c). According to the level of the energy barriers, the degree of difficulty of spontaneity is obtained as follows: Bi₂O₃/NV-C₃N₄ > Bi₂O₃/C₃N₄ > Bi₂O₃, which is consistent with the whole free energy for the ECRR and the experimental results.

Fig. 3 (a) Geometric structure models of three intermediates *CO₂, *OCHO and *HCOOH adsorbed on Bi₂O₃/NV-C₃N₄; (b) free energy graph for CO₂ conversion; (c) free energy diagrams for the HER.

Conclusions

In conclusion, it has been demonstrated that the ultrafine Bi₂O₃ nanoparticles (∼1.9 nm) supported on graphitic carbon nitride containing nitrogen vacancies (Bi₂O₃/NV-C₃N₄) exhibit excellent selectivity and stability for the ECRR to formate. Experimental and theoretical analyses show that metal–support interfacial interactions can enrich the electron density of Bi₂O₃ effectively facilitating *COO generation, which is beneficial to the generation of formate. Meanwhile, the metal–support interfacial interactions can be extended to other catalyst systems to improve their catalytic performance.

Author contributions

Jiayu Zhan: investigation, methodology, formal analysis, writing the original draft and DFT calculations. Lu-Hua Zhang: conceptualization, writing the original draft, writing review & editing, supervision, and funding acquisition. Xueli Wang: investigation, methodology, formal analysis, and writing the original draft. Yuqi Hu: writing review & editing. Yi Jiang: conceptualization and writing review & editing. Fengshou Yu: conceptualization, writing review & editing, supervision, and funding acquisition.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 22008048, 21905073, 22278108, and 22008050), Hundred Talents Project of Hebei Province (No. E2019050015), Natural Science Foundation for Outstanding Youth Scholars of Hebei Province (No. B2021202061), Natural Science Foundation of Hebei Province (No. B2021202010 and B2020202066), State Key Laboratory of Fine Chemicals (KF 2108), and Open Research Fund of CNMGE Platform & NSCC-TJ.

Notes and references

1. (a) O. S. Bushuyev, P. De Luna, C. T. Dinh, L. Tao, G. Saur, J. van de Lagemaat, S. O. Kelley and E. H. Sargent, Joule, 2018, 2, 825–832; (b) M. Schreier, F. Héroguel, L. Steier, S. Ahmad, J. S. Luterbacher, M. T. Mayer, J. Luo and M. Grätzel, Nat. Energy, 2017, 2; (c) P. Ding, H. Zhao, T. Li, Y. Luo, G. Fan, G. Chen, S. Gao, X. Shi, S. Lu and X. Sun, J. Mater. Chem. A, 2020, 8, 21947–21960; (d) D. Chen, L. H. Zhang, J. Du, H. Wang, J. Guo, J. Zhan, F. Li and F. Yu, Angew. Chem., Int. Ed., 2021, 60, 24022–24027; (e) X.-H. Zhao, Q.-S. Chen, D.-H. Zhuo, J. Lu, Z.-N. Xu, C.-M. Wang, J.-X. Tang, S.-G. Sun and G.-C. Guo, Electrochim. Acta, 2021, 367; (f) H. Guo, D.-H. Si, H.-J. Zhu, Q.-X. Li, Y.-B. Huang and R. Cao, Science, 2022, 2, 295–303.
2. (a) N. Zouaoui, B. D. Ossonon, M. Fan, D. Mayilukila, S. Garbarino, G. de Silveira, G. A. Botton, D. Guay and A. C. Tavares, J. Mater. Chem. A, 2019, 7, 11272–11281; (b) Z. Yang, F. E. Oropeza and K. H. L. Zhang, APL Mater., 2020, 8; (c) K. Van Daele, B. De Mot, M. Pupo, N. Daems, D. Pant, R. Kortlever and T. Breugelmans, ACS Energy Lett., 2021, 6, 4317–4327; (d) M. F. Philips, G.-J. M. Gruter, M. T. M. Koper and K. J. P. Schouten, ACS Sustainable Chem. Eng., 2020, 8, 15430–15444; (e) N. Han, P. Ding, L. He, Y. Li and Y. Li, Adv. Energy Mater., 2019, 10, 1902338; (f) J. Guo, W. Zhang, L. H. Zhang, D. Chen, J. Zhan, X. Wang, N. R. Shiju and F. Yu, Adv. Sci., 2021, 8, e2102884.
3. P. Deng, H. Wang, R. Qi, J. Zhu, S. Chen, F. Yang, L. Zhou, K. Qi, H. Liu and B. Y. Xia, ACS Catal., 2019, 10, 743–750.
4. (a) R. Reske, H. Mistry, F. Behafarid, B. Roldan Cuenya and P. Strasser, J. Am. Chem. Soc., 2014, 136, 6978–6986; (b) S. Nitopi, E. Bertheussen, S. B. Scott, X. Liu, A. K. Engstfeld, S. Horch, B. Seger, I. E. L. Stephens, K. Chan, C. Hahn, J. K. Norskov, T. F. Jaramillo and I. Chorkendorff, Chem. Rev., 2019, 119, 7610–7672; (c) X. Deng, D. Alfonso, T.-D. Nguyen-Phan and D. R. Kauffman, ACS Catal., 2022, 5921–5929; (d) X. Wang, L. H. Zhang, D. Chen, J. Zhan, J. Guo, Z. Zhang and F. Yu, ChemCatChem, 2022, 14; (e) Y. Peng and S. Chen, Green Energy Environ., 2018, 3, 335–351; (f) W. Shang, W. Liu, X. Cai, J. Hu, J. Guo, C. Xin, Y. Li, N. Zhang, N. Wang, C. Hao and Y. Shi, Adv. Powder Mater., 2022, 100094.
5. (a) L. Zhang, F. Mao, L. R. Zheng, H. F. Wang, X. H. Yang and H. G. Yang, ACS Catal., 2018, 8, 11035–11041; (b) X. Zeng, Y. Liu, Y. Kang, Q. Li, Y. Xia, Y. Zhu, H. Hou, M. H. Uddin, T. R. Gengenbach, D. Xia, C. Sun, D. T. McCarthy, A. Deletic, J. Yu and X. Zhang, ACS Catal., 2020, 10, 3697–3706; (c) B. Ma, G. Chen, C. Fave, L. Chen, R. Kuriki, K. Maeda, O. Ishitani, T. C. Lau, J. Bonin and M. Robert, J. Am. Chem. Soc., 2020, 142, 6188–6195; (d) Y.-N. Gong, B.-Z. Shao, J.-H. Mei, W. Yang, D.-C. Zhong and T.-B. Lu, Nano Res., 2021, 15, 551–556.
6. (a) Y. Zheng, Y. Jiao, Y. Zhu, Q. Cai, A. Vasileff, L. H. Li, Y. Han, Y. Chen and S. Z. Qiao, J. Am. Chem. Soc., 2017, 139, 3336–3339; (b) Y. Jiao, Y. Zheng, P. Chen, M. Jaroniec and S. Z. Qiao, J. Am. Chem. Soc., 2017, 139, 18093–18100.
7. K. Chu, Q.-q. Li, Y.-p. Liu, J. Wang and Y.-h. Cheng, Appl. Catal., B, 2020, 267.
8. J. Hafner, J. Comput. Chem., 2008, 29, 2044–2078.
9. P. E. Blochl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979.
10. A. V. Krukau, O. A. Vydrov, A. F. Izmaylov and G. E. Scuseria, J. Chem. Phys., 2006, 125, 224106.
11. (a) B. Zhu, P. Xia, W. Ho and J. Yu, Appl. Surf. Sci., 2015, 344, 188–195; (b) Q. Liang, Z. Li, Y. Bai, Z.-H. Huang, F. Kang and Q.-H. Yang, Sci. China Mater., 2016, 60, 109–118.
12. (a) Y. Kang, Y. Yang, L. C. Yin, X. Kang, G. Liu and H. M. Cheng, Adv. Mater., 2015, 27, 4572–4577; (b) C. Lv, Y. Qian, C. Yan, Y. Ding, Y. Liu, G. Chen and G. Yu, Angew. Chem., Int. Ed., 2018, 57, 10246–10250.
13. (a) Y. X. Duan, K. H. Liu, Q. Zhang, J. M. Yan and Q. Jiang, Small Methods, 2020, 4; (b) D. Zhang, Z. Tao, F. Feng, B. He, W. Zhou, J. Sun, J. Xu, Q. Wang and L. Zhao, Electrochim. Acta, 2020, 334; (c) S. Kim, W. J. Dong, S. Gim, W. Sohn, J. Y. Park, C. J. Yoo, H. W. Jang and J.-L. Lee, Nano Energy, 2017, 39, 44–52; (d) X. Chen, H. Chen, W. Zhou, Q. Zhang, Z. Yang, Z. Li, F. Yang, D. Wang, J. Ye and L. Liu, Small, 2021, 17, 2101128; (e) W. Zhang, Y. Hu, L. Ma, G. Zhu, P. Zhao, X. Xue, R. Chen, S. Yang, J. Ma, J. Liu and Z. Jin, Nano Energy, 2018, 53, 808–816; (f) K. Fan, Y. Jia, Y. Ji, P. Kuang, B. Zhu, X. Liu and J. Yu, ACS Catal., 2019, 10, 358–364; (g) D. Wu, J. Liu, Y. Liang, K. Xiang, X. Z. Fu and J. L. Luo, ChemSusChem, 2019, 12, 4700–4707; (h) X. Zhang, J. Fu, Y. Liu, X.-D. Zhou and J. Qiao, ACS Sustainable Chem. Eng., 2020, 8, 4871–4876.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2se01274j

---