Pressure regulated CO₂ electrolysis on two-dimensional Bi₂O₂Se

Abstract

The electrochemical reduction of carbon dioxide (CO₂RR) offers potential for sustainable production and greenhouse gas mitigation, particularly with renewable energy integration. However, its widespread application is hindered by expensive catalysts, low selectivity, and limited current density. This study addresses these challenges by developing a low-mass-loading two-dimensional (2D) Bi₂O₂Se catalyst via chemical vapor deposition (CVD). The catalyst achieves a formate faradaic efficiency (FE) of 47.1% with a high current density of 4649 mA mg⁻¹ at −1.15 V (vs. RHE), significantly outperforming bulk Bi₂O₂Se. Pressurizing CO₂, a condition commonly encountered in industrial processes, further enhances formate selectivity and current density, increasing from 2189 mA mg⁻¹ at ambient pressure (1.01 bar) to 7457 mA mg⁻¹ at 40 bar. In situ Raman spectroscopy and DFT calculations reveal the intermediates and pathways involved, underscoring the critical role of pressure in regulating CO₂RR pathways. These findings highlight the potential of 2D catalysts for sustainable and industrially relevant CO₂ conversion under high pressure.

---

Production of value-added chemicals via renewable-driven electrochemical reduction of CO₂ (CO₂RR) holds promise to deal with the pressing global warming and energy crisis.^1–3 To date, the predominant focus of CO₂RR research lies in examining its performance under ambient pressure conditions, while industrial CO₂ is typically pressurized during capture, transport, and storage. Coincidentally, CO₂RR benefits from high pressure in aqueous solutions because low CO₂ solubility under ambient pressure usually leads to the formation of unfavorable active carbon species, and subsequently causes diminished current density and reduced selectivity.^4–6 We have reported that higher CO₂ pressure can significantly improve the formate selectivity during aqueous-based CO₂RR over commonly used catalysts such as Cu, Au, Ag and Sn.⁷ Cu₂O@Cu catalysts with a hollow sphere morphology were also found to produce ethanol when operating under pressure.⁸ These studies pointed out the important role of high pressure in regulating the CO₂RR pathways.

Bismuth-based catalysts have garnered considerable interest in the CO₂RR field because of their high selectivity toward formate, cost effectiveness and low toxicity, rendering them a capable material for large-scale applications.^3,9 Two-dimensional (2D) bismuth materials have aroused increasing attention in light of their high specific surface area, large atomic exposure rate and tunable electronic states.¹⁰ Han et al. showcased the effectiveness of ultrathin BiNS catalysts, achieving superior formate selectivity (>90% FE), current density (24 mA cm⁻²), and durability.¹¹ Peng et al. synthesized ultrathin bismuth nanosheets (1.02 nm thick) with enhanced intrinsic activity and abundant active sites.^12,13 Recently, 2D Bi₂O₂Se has emerged as a robust semiconducting material with ultra-high electron mobility and quantum oscillations,¹⁴ and we believe that 2D Bi₂O₂Se may hold the potential to catalyze the CO₂RR effectively. Moreover, 2D Bi₂O₂Se may be highly cost-effective: on the one hand, Bi is a kind of non-noble metal and on the other hand, 2D materials exhibit low mass loading.

Here we report the pressure-regulated CO₂RR to formate as catalyzed by 2D Bi₂O₂Se. The 2D Bi₂O₂Se catalyst with a layered “2D Zipper” structure is synthesized by a chemical vapor deposition (CVD) method on glassy carbon. While the catalyst loading is as low as 5.2 μg cm⁻², our 2D Bi₂O₂Se manifests a CO₂RR current density as high as 4649 mA mg⁻¹ with a formate FE of 47%, surpassing reported Bi based 2D catalysts and the Bi₂O₂Se bulk counterparts. By pressurizing CO₂ from ambient pressure (1.01 bar) to 40 bar, we steer the CO₂RR pathway toward formate, achieving a record-high current density of 7084 mA mg⁻¹ with a formate FE of 78%. In situ Raman spectroscopy and density functional theory (DFT) calculations reveal the mechanism of CO₂RR to formate under high pressure, evidencing a stronger CO₂ adsorption, enhanced *OCHO intermediate formation and more favorable pathway.

The 2D Bi₂O₂Se was synthesized through a two-step process: (i) preparation of bismuth oxide on glassy carbon via E-beam evaporation and (ii) selenization by the CVD method (Fig. 1a). Using glassy carbon as the substrate allowed the in situ synthesis of 2D Bi₂O₂Se, enabling its direct use as a working electrode in H-cell setups without the need for 2D material transfer. Fig. 1b illustrates the atomic structure of Bi₂O₂Se, which comprises alternating layers of bismuth atoms in hexagonal lattices, oxygen atoms positioned between these layers, and selenium atoms filling the interlayer gaps.¹⁵ As documented in the literature, the Raman peak at the A_1g mode (159 cm⁻¹) indicates the symmetric stretching vibration mode of the Bi–O bonds in the Bi₂O₂ layer (Fig. S1, ESI†).¹⁶ The lattice structure of Bi₂O₂Se was confirmed using high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), which aligns with the proposed “2D Zipper” model (Fig. S2, ESI†).¹⁷ This model suggests that the chemical bonds near the surface of Bi₂O₂Se are strengthened, enhancing the stability of the few-layer structures. Additionally, electron energy loss spectroscopy (EELS) mapping of Bi₂O₂Se revealed a uniform distribution of Bi, O, and Se elements across the entire area (Fig. S3, ESI†). Atomic force microscopy (AFM) characterization on Bi₂O₂Se confirmed an average thickness of 10.49 nm (Fig. S4, ESI†), revealing 17 layers of 2D monolayer.¹⁴ This thickness suggests a low catalyst mass loading, which was further validated by inductively coupled plasma mass spectrometry (ICP-MS), revealing a catalyst loading of 5.2 μg cm⁻². X-ray photoelectron spectroscopy (XPS) was used to verify the presence of Bi, O, and Se (Fig. S5 and S6, ESI†). The spectra revealed two oxidation states of the Bi atom, represented by Bi 4f_7/2 and Bi 4f_5/2 at binding energies of 159 and 164 eV, respectively, as well as O 1s states at 530 eV and Se 3d states at 53 eV, consistent with previous reports.¹⁸ These observations authenticated the successful synthesis of 2D Bi₂O₂Se and motivated us to conduct the subsequent electrochemical experiments.

Fig. 1 Schematic of (a) 2D Bi₂O₂Se synthesis by CVD and (b) the atomic structure.

In contrast to wet chemical reduction,¹⁹ galvanic replacement reaction,²⁰ hydrothermal reaction,²¹ and electrochemical conversion²² methods, the direct CVD growth of 2D Bi-based catalysts on glassy carbon substrates offers low mass loading by depositing only a few layers of catalyst (Fig. S4, ESI†). This may also enhance electron transport from the conductive substrate to the 2D catalyst. The CO₂RR activity of 2D Bi₂O₂Se was firstly assessed in 0.5 M KHCO₃ electrolyte by linear sweep voltammetry (LSV) at ambient pressure (1.01 bar), as shown in Fig. S7 (ESI†). In a CO₂ saturated electrolyte, the 2D Bi₂O₂Se demonstrates a higher current density compared to an N₂ saturated electrolyte at −0.7 V vs. reversible hydrogen electrode (RHE), indicating a greater favorability for CO₂RR over the hydrogen evolution reaction (HER).

To assess the activity in terms of current density per unit mass (mA mg⁻¹) and faradaic efficiency (FE), we compared the synthesized 2D Bi₂O₂Se with Bi₂O₂Se bulk prepared by the spin coating method with catalyst loading of 0.5 mg cm⁻² (Fig. 2a). As shown in Fig. S8 (ESI†), 1H nuclear magnetic resonance (NMR) analysis identified formate as the only liquid product from the CO₂RR, while gas chromatography (GC) detected H₂ and CH₄ as the gas products. It was observed that 2D Bi₂O₂Se exhibits significantly higher current density per unit mass, reaching 4649 mA mg⁻¹ at −1.15 V (vs. RHE), which is 2 orders of magnitude greater than that of the bulk counterpart. However, the FE toward formate for both 2D Bi₂O₂Se and Bi₂O₂Se bulk remained at the same level and decreased with increased potential. This was attributed to the stronger dependence of the Volmer step potential on the overall potential compared to CO₂ adsorption, leading to a higher preference for the HER at more negative potentials (Fig. S8, ESI†).²³ To understand the specific surface area properties of 2D Bi₂O₂Se and Bi₂O₂Se bulk, electrochemically active surface area (ECSA) calculation (Fig. S9 and S10, ESI†) and non-mass-normalized LSV (Fig. S7, ESI†) were performed. Both 2D Bi₂O₂Se and Bi₂O₂Se bulk exhibit similar surface area and non-mass-normalized current density, indicating minimal effect of ECSA when comparing the mass-normalized CO₂RR current density. More than that, we found that both Bi₂O₂Se and Bi₂O₃ were firstly reduced to the Bi metal state and it is Bi (012) that acts as the main facet during the CO₂RR (Fig. S11, ESI†). These findings suggested that 2D Bi₂O₂Se growth on glassy carbon enabled more effective utilization of the catalyst in light of the low mass loading, while maintaining comparable CO₂RR activity compared to bulk Bi₂O₂Se. A similar trend can be found when comparing 2D Bi₂O₃ and Bi₂O₃Se bulk (Fig. S12, ESI†). Furthermore, the CVD grown 2D Bi₂O₂Se catalyst greatly outperforms the benchmarks in the literature, including mesoporous Bi nanosheets,²⁴ free-standing 2D bismuth nanosheets,²⁵ layered Bi nanosheets,²⁶ electron-rich Bi nanosheets,²⁷ 3D network of interconnected 2D bismuthene arrays,²⁸ atomically thin bismuthene with rich defects,²⁹ heterostructured bismuth-based catalysts,³⁰ Bi₂O₂CO₃ nanosheets³¹ and Bi-MOF^32,33 (Fig. 2b).

Fig. 2 Effective utilization of the 2D catalyst. (a) CO₂RR to HCOO⁻ performance of 2D Bi₂O₂Se and Bi₂O₂Se bulk at 1.01 bar. (b) The comparison of some typical Bi-based 2D catalysts for CO₂RR to HCOO⁻ with 2D Bi₂O₂Se.

Pressurization of CO₂ has been reported to steer the selectivity^34,35 and boost the CO₂RR current density.^4,36 To further enhance the CO₂RR performance of our 2D Bi₂O₂Se catalyst, we subjected the system to elevated CO₂ pressures ranging from 1.01 to 40 bar. In general, the FE towards formate increased as the CO₂ pressure increased, rising from 79.5% at 1.01 bar to 95% at 40 bar at −0.85 V (vs. RHE) (Fig. S13, ESI†). This demonstrates the capability of pressurized CO₂ to regulate the selectivity towards formate on 2D Bi₂O₂Se, consistent with previous study.⁷ The corresponding total and formate partial current densities at various CO₂ pressures are plotted in Fig. S14 (ESI†) and Fig. 3. In particular, the formate partial current density was improved from 2146 to 7308 mA mg⁻¹ when the pressure increased from 1.01 to 40 bar at −1.15 V, indicating a greatly promoted production rate of formate. To compare the 2D Bi₂O₂Se with Bi₂O₂Se bulk under high CO₂ pressure, we plotted current densities with corresponding formate FEs (Fig. 2a and Fig. S15, ESI†). The results align with prior research, underscoring the efficacy of our 2D Bi₂O₂Se catalyst with minimal mass loading in sustaining high current densities. As depicted in Fig. S16 (ESI†), Nyquist plots for CO₂RR over the 2D Bi₂O₂Se catalyst showed a decrease in charge transfer resistance with increased CO₂ pressures. This indicates that a significantly accelerated electron transfer with higher CO₂ pressure as enhanced CO₂ solubility in electrolytes promoted reactant transportation to the electrodes, that means, despite HER being more favorable under more negative potentials according to previous reported works,^37,38 the elevated CO₂ pressure can still suppress the HER (Fig. S8 and S17, ESI†).

Fig. 3 Partial current density of pressure regulated CO₂RR to formate on 2D Bi₂O₂Se.

These findings motivated us to investigate the mechanism of the CO₂RR to formate under high CO₂. In general, the CO₂RR to formate pathway follows 4 steps:

CO₂ + * → *CO₂ (1)

*CO₂ + H⁺ + e⁻ → *OCHO (2)

*OCHO + H⁺ + e⁻ → *HCOOH (3)

*HCOOH → * + HCOOH (4)

In order to track reaction intermediates during CO₂RR over 2D Bi₂O₂Se, in situ Raman spectroscopy measurements were performed in ambient CO₂ pressure (1.01 bar) and pressurized CO₂ (20 bar) (Fig. 4a and Fig. S18, ESI†). In the scan of applied potential from −0.75 to −1.2 V vs. RHE over the 2D Bi₂O₂Se catalyst, two intrinsic Raman peaks were detected. The Raman peaks at 1052 cm⁻¹ were attributed to HCO₃⁻, respectively, indicating a CO₂ saturated environment during the CO₂RR.^39,40 In addition, the peak at 2898 cm⁻¹ was ascribed to the C–H stretching of *OCHO radicals.⁴¹ This revealed the favored *CO₂⁻ formation to activate CO₂ molecules and enhanced adsorption strength of *OCHO intermediates, leading to superior activity and selectivity toward formate. Due to the low catalyst loading and strong Raman signal from glassy carbon, the D band and G band of glassy carbon were observed at 1346 cm⁻¹ and 1572 cm⁻¹. To gain a deep understanding on high pressure CO₂RR, the DFT and Ab initio molecular dynamic (AIMD) calculations were performed in the Vienna Ab initio Simulation Package (VASP). In light of the pressure dependent CO₂ solubility, we sought to explore the influence of CO₂ coverage from 1/8 monolayer (ML) to 3/8 ML on the Bi (012) facet, the dominant facet of the prepared catalyst, confirmed by X-ray diffraction (XRD) measurement, as shown in Fig. S11 (ESI†).

Fig. 4 Mechanism study on high pressure CO₂RR to formate over 2D Bi₂O₂Se. (a) In situ Raman spectroscopy measurements at 20 bar. (b) Free energy diagram of adsorbed CO₂RR intermediates with the CO₂ coverages of 1/8 ML and 3/8 ML.

The optimized structures and free energy diagram of adsorbed CO₂RR to formate intermediates are depicted in Fig. S19 (ESI†) and Fig. 4b. In the initial step (CO₂ + * → *CO₂), a smaller free energy difference for 3/8 ML CO₂ coverage (0.1 eV) indicated a stronger CO₂ adsorption when compared with that of 1/8 ML CO₂ coverage (0.32 eV). The first proton-coupled electron transfer process (*CO₂ + H⁺ + e⁻ → *OCHO) was defined as the rate-determine-step (RDS) of the CO₂RR to formate with 1/8 ML CO₂ coverage due to the large uphill energy. Moreover, in this step, the required free energy for 3/8 ML coverage is 0.1 eV, notably lower than that for 1/8 ML coverage (0.25 eV). That is, CO₂RR under high pressure exhibits higher activity for *OCHO formation due to the lower reaction energy barrier. For the CO₂RR to formate with 3/8 ML CO₂ coverage, the second proton-coupled electron transfer process (*OCHO + H⁺ + e⁻ → *HCOOH) was defined as the RDS. This proton-coupled electron transfer process (*OCHO + H⁺ + e⁻ → *HCOOH) and the desorption process (*HCOOH → * + HCOOH) contributed to the final formate production. The free energy difference of these two steps with 3/8 ML CO₂ coverage is lower, revealing a more favorable pathway. DFT calculations on the HER under pressurized CO₂ were also conducted, as shown in Fig. S20 (ESI†). HER exhibited decreased energy barrier for RDS from 1 V to 0.92 V with enhanced CO₂ coverage from 1/8 ML to 3/8 ML, whereas the CO₂RR has an RDS energy barrier of only 0.13 V under high pressure. As a result, CO₂RR is preferred compared to HER as the pressure increases, aligning with the experimental data (Fig. S8 and S17, ESI†).

In conclusion, we devise a low-mass-loading 2D Bi₂O₂Se catalyst and leverage the pressure to enhance CO₂RR for formate production. The 2D catalyst exhibited a significantly higher current density of 4649 mA mg⁻¹ with 47.1% FE to formate at −1.15 V (vs. RHE) compared to the bulk phase, indicating its cost-effectiveness. High CO₂ pressure, commonly encountered in industrial processes involving CO₂ capture, transport, and storage, significantly enhances formate selectivity and current density during CO₂RR. Under high pressure, the partial current density increases from 2146 mA mg⁻¹ at 1.01 bar to 7308 mA mg⁻¹ at 40 bar at −1.15 V. Moreover, in situ Raman measurements and DFT calculations under pressurized CO₂RR revealed the intermediates and pathways involved. These findings highlight the potential of 2D catalysts for sustainable CO₂ conversion at high pressure with industrial relevance and underscore the importance of pressure in regulating the CO₂RR pathways.

This work was financially supported by the Baseline Fund (BAS/1/1413-01-01) to X. L. from King Abdullah University of Science and Technology (KAUST).

Data availability

The data supporting the findings of this study are available within the article and its ESI.† Additional raw data and experimental details can be provided by the corresponding author upon reasonable request.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. 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.
2. S. Yoshida, M. Sampei, N. Todoroki, E. Hisamura, K. Nakao, K. Albrecht and T. Wadayama, Chem. Commun., 2023, 59, 3459–3462.
3. B.-Q. Miao, W.-S. Fang, B. Sun, F.-M. Li, X.-C. Wang, B.-Y. Xia and Y. Chen, Chin. J. Struct. Chem., 2023, 42, 100095.
4. J. Li, Y. Kuang, Y. Meng, X. Tian, W.-H. Hung, X. Zhang, A. Li, M. Xu, W. Zhou and C.-S. Ku, J. Am. Chem. Soc., 2020, 142, 7276–7282.
5. J. Li, J. Guo and H. Dai, Sci. Adv., 2022, 8, eabo0399.
6. S. Ruan, B. Zhang, J. Zou, W. Zhong, X. He, J. Lu, Q. Zhang, Y. Wang and S. Xie, Chin. J. Catal., 2022, 43, 3161–3169.
7. L. Huang, G. Gao, C. Yang, X.-Y. Li, R. K. Miao, Y. Xue, K. Xie, P. Ou, C. T. Yavuz and Y. Han, Nat. Commun., 2023, 14, 2958.
8. R. Qiu, J. Jia, L. Peng, R. Li, S. Yan, J. Li, J. Zhang, D. T. Sun, Z. Lan and T. Xue, Green Chem., 2023, 25, 684–691.
9. B. Ávila-Bolívar, V. Montiel and J. Solla-Gullón, ChemElectroChem, 2022, 9, e202200272.
10. S. Lu, F. Lou and Z. Yu, Catalysts, 2022, 12, 228.
11. N. Han, Y. Wang, H. Yang, J. Deng, J. Wu, Y. Li and Y. Li, Nat. Commun., 2018, 9, 1320.
12. C. J. Peng, X. T. Wu, G. Zeng and Q. L. Zhu, Chem. – Asian J., 2021, 16, 1539–1544.
13. S. Liu, Y. Fan, Y. Wang, S. Jin, M. Hou, W. Zeng, K. Li, T. Jiang, L. Qin and Z. Yan, Nano Lett., 2022, 22, 9107–9114.
14. J. Wu, H. Yuan, M. Meng, C. Chen, Y. Sun, Z. Chen, W. Dang, C. Tan, Y. Liu and J. Yin, Nat. Nanotechnol., 2017, 12, 530–534.
15. H. Zhou, X. Zhu, W. Liu, S. Liu, Y. Ding, Q. Zhang, Z. Zhang and R. Zhang, Appl. Surf. Sci., 2024, 672, 160851.
16. W. Chen, U. Khan, S. Feng, B. Ding, X. Xu and B. Liu, Adv. Funct. Mater., 2020, 30, 2004960.
17. Q. Wei, R. Li, C. Lin, A. Han, A. Nie, Y. Li, L.-J. Li, Y. Cheng and W. Huang, ACS Nano, 2019, 13, 13439–13444.
18. M. T. Hossain and P. Giri, J. Appl. Phys., 2021, 129, 17.
19. H. Xie, T. Zhang, R. Xie, Z. Hou, X. Ji, Y. Pang, S. Chen, M. M. Titirici, H. Weng and G. Chai, Adv. Mater., 2021, 33, 2008373.
20. J. Fan, X. Zhao, X. Mao, J. Xu, N. Han, H. Yang, B. Pan, Y. Li, L. Wang and Y. Li, Adv. Mater., 2021, 33, 2100910.
21. J. Ma, J. Yan, J. Xu, J. Ni, R. Li, L. Li and L. Lu, Chem. – Eur. J., 2022, 28, e202201747.
22. P. Liu, H. Liu, S. Zhang, J. Wang and C. Wang, J. Colloid Interface Sci., 2021, 602, 740–747.
23. P. D. Pedersen, M. M. Melander, T. Bligaard, T. Vegge, K. Honkala and H. A. Hansen, J. Phys. Chem. C, 2023, 127, 18855–18864.
24. H. Yang, N. Han, J. Deng, J. Wu, Y. Wang, Y. Hu, P. Ding, Y. Li, Y. Li and J. Lu, Adv. Energy Mater., 2018, 8, 1801536.
25. F. Yang, A. O. Elnabawy, R. Schimmenti, P. Song, J. Wang, Z. Peng, S. Yao, R. Deng, S. Song and Y. Lin, Nat. Commun., 2020, 11, 1088.
26. Y. Qiao, W. Lai, K. Huang, T. Yu, Q. Wang, L. Gao, Z. Yang, Z. Ma, T. Sun and M. Liu, ACS Catal., 2022, 12, 2357–2364.
27. Z. Li, B. Sun, D. Xiao, Z. Wang, Y. Liu, Z. Zheng, P. Wang, Y. Dai, H. Cheng and B. Huang, Angew. Chem., Int. Ed., 2023, 62, e202217569.
28. Y. C. He, D. D. Ma, S. H. Zhou, M. Zhang, J. J. Tian and Q. L. Zhu, Small, 2022, 18, 2105246.
29. M. Zhang, W. Wei, S. Zhou, D.-D. Ma, A. Cao, X.-T. Wu and Q.-L. Zhu, Energy Environ. Sci., 2021, 14, 4998–5008.
30. P. F. Sui, C. Xu, M. N. Zhu, S. Liu, Q. Liu and J. L. Luo, Small, 2022, 18, 2105682.
31. T. Fan, W. Ma, M. Xie, H. Liu, J. Zhang, S. Yang, P. Huang, Y. Dong, Z. Chen and X. Yi, Cell Rep. Phys. Sci., 2021, 2, 3.
32. Z. W. Yang, J. M. Chen, Z. L. Liang, W. J. Xie, B. Zhao and L. N. He, ChemCatChem, 2023, 15, e202201321.
33. Q. Huang, X. Sha, R. Yang, H. Li and J. Peng, ACS Appl. Mater. Interfaces, 2024, 16, 13882–13892.
34. C. M. Gabardo, A. Seifitokaldani, J. P. Edwards, C.-T. Dinh, T. Burdyny, M. G. Kibria, C. P. O’Brien, E. H. Sargent and D. Sinton, Energy Environ. Sci., 2018, 11, 2531–2539.
35. K. Hara, A. Kudo and T. Sakata, J. Electroanal. Chem., 1995, 391, 141–147.
36. E. J. Dufek, T. E. Lister, S. G. Stone and M. E. McIlwain, J. Electrochem. Soc., 2012, 159, F514.
37. X.-Z. Wang, S. Liu, Q. Liu and J.-L. Luo, Electrochem. Commun., 2019, 107, 106531.
38. R. Sun, J. Zhao and X. Lu, J. Mater. Chem. A, 2024, 12, 8429–8437.
39. A. Dutta, I. n Zelocualtecatl Montiel, K. Kiran, A. Rieder, V. Grozovski, L. Gut and P. Broekmann, ACS Catal., 2021, 11, 4988–5003.
40. H. Li, P. Wei, D. Gao and G. Wang, Curr. Opin. Green Sustainable Chem., 2022, 34, 100589.
41. S. Jiang, K. Klingan, C. Pasquini and H. Dau, J. Chem. Phys., 2019, 150, 4.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cc05357e

‡ These authors contributed equally.

---