English

• Increasing concentration of atmospheric CO₂ is a worldwide concern and continues to trigger various environmental problems [1,2]. Electrocatalytic CO₂ reduction reaction utilizing intermittent renewable electricity, driven by wind and solar, presents a promising avenue to facilitate the carbon cycle due to its sustainable nature and the valuable products yielded in this procedure [3–5]. CO₂ can be converted into a variety of reduction products such as carbon monoxide, formic acid, methanol, methane, ethanol, and ethylene [6]. Among the above carbonaceous compounds, formate (formic acid) is regarded as one of the ideal value-added products with low toxicity and flammability. Formate has a wide range of applications in the chemical industry and can also be used for direct formic acid fuel cells (DFAFC) [7]. However, the dynamic process of CO₂RR is typically slow and inefficient due to the electrochemical inert of CO₂ molecules and the competing hydrogen evolution reaction (HER) [8–10]. As a result, the development of highly efficient and selective catalysts for CO₂RR to formate has emerged as a research priority.

  At present, the metal-based catalysts (such as Sn and Bi) have been regarded as the most promising alternatives for CO₂RR to formate [11–13]. Some efforts have been reported to overcome the sluggish kinetics and enhance the efficiency for formate production, including tuning the size, shape, composition, and crystallinity of the metallic component [14–18]. Besides, the dispersion of active metal nanoparticles (NPs) on supports is also an effective strategy to improve the utilization rate of metals and thus reduce catalyst costs [19–21]. Additionally, upon interaction with the support material, the metal active centers demonstrate heightened metal-support interactions attributed to the increased prevalence of interface sites. These interactions not only bolster stability but also exert a profound influence on the electronic configuration of the metal, subsequently exerting a pivotal role in shaping its catalytic characteristics [22,23]. For example, Xu and coworkers reported an electronic metal–support interaction between Cu and a-C on Cu-based MOF, achieving high selective and stable CO₂ electroreduction to CH₄ [24]. Other reports also showed that the incorporation of various supports, such as carbon-based materials or metal compounds (including metal oxides and metal carbides), in conjunction with metal active centers, can manifest significantly enhanced electrocatalytic activity, compared to catalysts lacking such support [20,25–27]. Among these varieties of loading support, porous carbon fibers (PCF), which amalgamate the structural and functional attributes of conventional carbon fibers and porous carbon materials, have garnered significant attention in a broad spectrum of applications including energy conversion and storage, catalysis, adsorption and separation, sensing, and various other domains [28–31]. Compared with traditional porous carbon powder materials that require an insulating binder, PCF stands out as an exemplary binder-free, self-supporting, robust material characterized by its fibrous morphology and well-defined porous architecture. Therefore, the architectures, morphological characteristics, and elemental compositions of PCFs, along with the integration of active materials, are regarded as pivotal factors for enhancing their efficacy in the realms of energy and environmental applications [32,33].

  In this work, we designed a SnO₂ nanoparticles loading freestanding porous carbon nanofibers (SnO₂PCNF) catalyst via simple electrospinning technology and later hydrothermal method. As a potential loading support, porous carbon fiber gives full play to its porous local confinement effect to the dispersed SnO₂ particles. DFT calculations showed that after interaction between SnO₂ particles and carbon nanofibers, the oxidation state of Sn was enhanced to a higher Sn^δ+ peak and the d-band center of active sites Sn was shifted to be more conducive for the CO₂ to formate process. Compared with the bulk SnO₂, SnO₂ loaded general carbon fiber (SnO₂CNF), the discernible impact of the porous structure and support effects on SnO₂PCNF can be unequivocally attributed to the substantial enhancement of current density, Faraday efficiency, and mass activity in the electrocatalytic reduction of CO₂ to formate.

  The dispersed SnO₂ nanoparticles loaded on porous carbon nanofibers (denoted as SnO₂PCNF) were synthesized via electrospinning and pyrolysis followed by the hydrothermal method, as shown schematically in Fig. 1a. First, the zeolitic imidazolate framework-8 (ZIF-8) were fabricated as pore former via a facile liquid-solid-solution method and homogeneously dispersed in PAN solution for electrospinning to obtain nanofibers precursors. Subsequently, porous carbon nanofibers were fabricated as a stable substrate with the evaporation of the Zn precursor via pyrolysis in the argon atmosphere. Finally, SnO₂ nanoparticles were grown in situ on porous carbon nanofibers by hydrothermal method. For comparison, the SnO₂ nanoparticles, bare porous carbon fibers (PCNF) and SnO₂ nanoparticles loaded on carbon nanofibers (SnO₂CNF) were also synthesized to research the effects on carbon supports and SnO₂ nanoparticles. The details were shown in Supporting information.

  Figure 1

  

  Figure 1.  (a) Synthesis scheme of the SnO₂PCNF. (b) The free-standing fiber membrane of SnO₂PCNF. (c) TEM image of SnO₂PCNF. (d, e) HR-TEM images of SnO₂PCNF. (f) HAADF-STEM-EDS elemental mapping images for SnO₂PCNF.

  DownLoad: Full-Size Img PowerPoint

  The morphology evolutions from PCNF to SnO₂, SnO₂CNF, and SnO₂PCNF were characterized using multiple methods. The as-obtained SnO₂PCNF catalyst is a flexible and self-supporting carbon membrane, which can be prepared at least 240 square centimeters in one single time under existing implementation conditions (Fig. 1b). As displayed in Fig. S1 (Supporting information), the scanning electron microscope (SEM) images show that the one-dimensional nanofibers framework in PCNF, SnO₂CNF and SnO₂PCNF are all well preserved after pyrolysis and the later hydrothermal treatments. Notably, in Fig. S1c, the size of the SnO₂ nanoparticles is hard to determine since they are agglomerated. The transmission electron microscopy (TEM) image in Fig. S2 (Supporting information) further demonstrates that the PCNF has an interconnected porous skeleton, which could provide abundant channels for SnO₂ nanoparticle loading. In addition, Fig. 1c shows that the SnO₂ loaded on PCNF can be effectively dispersed, and the high-resolution TEM (HR-TEM) in Fig. 1d and Fig. S3 (Supporting information) indicate an average particle size of about 5 nm. Moreover, Fig. 1e exhibits a set of lattice fringe on SnO₂PCNF with an interplanar distance of 0.335 nm, indexed to the (110) plane of SnO₂. The elemental mapping images of SnO₂PCNF in Fig. 1f show that C, O, Sn, and N elements are homogeneously distributed in the PCNF support. As a comparison, TEM images in Fig. S4 (Supporting information) of SnO₂ and SnO₂CNF are also investigated, which further proves the importance of ZIF-8 as a pore former for SnO₂ nanoparticle loading.

  X-ray powder diffraction (XRD) was further employed to obtain the crystal structure information of these synthesized samples. In Fig. 2a, unlike the bare PCNF, all peaks of the SnO₂-contained samples are well-matched with the rutile-type tetragonal SnO₂ (PDF #88-0287). To gain more insight into the porous structure of PCNF, SnO₂CNF and SnO₂PCNF, N₂ adsorption–desorption analysis and pore size distribution were characterized (Fig. 2b and Fig. S5 in Supporting information). It is evident that PCNF exhibits the highest specific surface area, measuring at 184.16 m²/g, and encompasses a broad spectrum of pore sizes, manifesting a layered porous structure consisting of both micropores and mesoporous pores. As a comparison, the BET surface areas of SnO₂, SnO₂CNF and SnO₂PCNF are listed in Table S1 (Supporting information). When juxtaposed with the CO₂ adsorption curve depicted in Fig. 2c, it becomes apparent that the porous structure in the catalyst facilitates the dispersion and growth of SnO₂ nanoparticles, along with the efficient penetration of the electrolyte and rapid ion diffusion during the CO₂RR process. It is indicated that a subsequent hydrothermal reaction loading SnO₂ nanoparticles appears to reduce the specific surface area of the porous carbon fibers, but also optimizes the pore size distribution on the carbon fibers. Meanwhile, the FT-IR spectra of PCNF, SnO₂, SnO₂CNF and SnO₂PCNF were recorded at transmittance mode (Fig. 2d). The peak at 607 cm⁻¹ on SnO₂ is attributed to the Sn–O stretching [34]. From this spectrum, the peaks of Sn-O stretching on SnO₂CNF and SnO₂PCNF are clearly redshifted compared to the SnO₂ sample, which represents electron migration and the lattice constant decreases.

  Figure 2

  

  Figure 2.  (a) XRD patterns, (b) N₂ adsorption/desorption isotherms, (c) CO₂ adsorption isotherms, (d) FT-IR spectra of PCNF, SnO₂, SnO₂CNF and SnO₂PCNF. (e) Sn 3d XPS spectra, (f) valence band spectra of SnO₂, SnO₂CNF and SnO₂PCNF.

  DownLoad: Full-Size Img PowerPoint

  Furthermore, the chemical compositions of these four samples were characterized by X-ray photoelectron spectra (XPS). The signals of C, O, Sn, and N elements are identified in the survey spectrum. As a comparison of Sn 3d XPS spectra in Fig. 2e, the signal peaks at 487.1 eV/595.6 eV are much higher than the Sn⁴⁺ peak (486.3 eV/594.8 eV) on SnO₂, which represents an enhanced of the Sn oxidation (denoted as Sn^δ+). In Figs. S6 and S7 (Supporting information), the C 1s, and N 1s XPS spectra of these four samples, combined with the Raman spectra (Fig. S8 in Supporting information), demonstrate a stable nitrogen-containing carbon substrate and stronger binding energy of Sn=O bonds from electronic structure optimization [15,35]. Therefore, further characterization was carried out with the surface valence band photoemission spectra to explore the d-band center of SnO₂, SnO₂CNF and SnO₂PCNF, which relates to the adsorption energy of an adsorbed molecule on transition metal sites [36,37]. In Fig. 2f, compared to SnO₂ (6.84 eV), the d-band center of SnO₂CNF (7.74 eV) and SnO₂PCNF (7.88 eV) upshifts. The changes in the d-band center have been proven to reduce the adsorption of C intermediates and enhance the adsorption of oxygen-containing intermediates during the CO₂RR process. These characteristics jointly reflect the support effects between SnO₂ and carbon substrates, which is embodied in the change of electronic structure.

  To explore the CO₂RR performance, the catalytic performance of SnO₂/PCNF electrocatalysts for CO₂RR was evaluated in a gastight two-compartment H-cell with CO₂-saturated 0.1 mol/L KHCO₃ as the electrolyte. The free-standing SnO₂PCNF membrane can be tailored for testing in Fig. S9 (Supporting information). Compared with linear sweep voltammetry (LSV) curves in the N₂-saturated electrolyte, the current densities measured for these four catalysts in the CO₂-saturated catholyte show a dramatic increase, indicating that CO₂ is more favorably reduced with their intrinsic activity than the HER in the CO₂-saturated catholyte (Fig. S10 in Supporting information). The PCNF has the highest current density, which is greater than that of SnO₂PCNF, SnO₂CNF, and SnO₂ in turn (Fig. 3a). These current densities seem to be related to the specific surface area of the catalysts. Therefore, analysis of catalyst reduction products is needed to further evaluate the electrochemical properties of these catalysts. Specifically, the resultant products under different potentials were quantitatively analyzed with potentiostatic methods via online gas chromatography (GC) for gas products and ¹H nuclear magnetic resonance (¹H NMR) spectroscopy for liquid products (Figs. S11 and S12 in Supporting information). As shown in Fig. 3b, the FE of SnO₂PCNF for formate is higher than 70% over the wide potential ranges and it reached up to 86% at −1.25 V_RHE in the H-type cell, with effective suppression of the hydrogen evolution reaction and CO production. In Figs. S13a-c (Supporting information), SnO₂ had a maximum formate FE of 73.6% at −1.65 V_RHE, while SnO₂CNF displayed a maximum formate FE of 84.5% at −1.45 V_RHE. Besides, the PCNF shows a competitive HER activity with a maximum H₂ FE of 72% at −1.45_RHE. As a comparison in Figs. 3c and d and Fig. S13d (Supporting information), the SnO₂PCNF exhibits the highest partial current density for formate of 24.8 mA/cm². According to current densities and Faradaic efficiency for the formate, SnO₂PCNF exhibits the best CO₂RR performance. Among them, SnO₂ as the active site provides excellent formate production, the interaction with carbon support improves higher formate selectivity, and the porous structure increases the current density and lowers the applied potential at the highest formate faradaic efficiency.

  Figure 3

  

  Figure 3.  (a) Comparisons of LSV curves of PCNF, SnO₂, SnO₂CNF and SnO₂PCNF in CO₂-saturated 0.1 mol/L KHCO₃ aqueous. (b) FEs of CO₂RR products on SnO₂PCNF. Comparisons of (c) the partial current densities for formate and (d) C₁ products at different applied potentials on PCNF, SnO₂, SnO₂CNF and SnO₂PCNF. (e) FE and mass active current for formate on flow-cell. (f) CO₂RR to formate performance comparisons of SnO₂PCNF with other Sn-based catalysts. (g) Long-term durability of SnO₂PCNF at −0.85 V for 21 h.

  DownLoad: Full-Size Img PowerPoint

  To evaluate the electrocatalytic CO₂ conversion performance of the as-prepared SnO₂PCNF materials at current densities close to industrial relevance, a gas diffusion electrode (GDE) flow cell was used with 1 mol/L KOH solution as the electrolyte. In the GDE cell configuration, a continuous supply of gaseous CO₂ feedstock is channelled through the porous carbon support, facilitating direct access to the catalyst surface. This strategic approach effectively bypasses the constraints imposed by gas solubility limitations within the aqueous electrolyte. Concurrently, on the opposing side of the porous electrode, the electrolyte is methodically circulated over the catalyst surface using a peristaltic pump. This controlled flow of electrolyte can help to mitigate the accumulation of formate species at the catalyst surface, and act as a pH buffer to curb undue fluctuations in the electrolyte’s pH. This orchestrated design seeks to minimize mass transfer limitations and enhance the overall efficiency of the system. As shown in Fig. 3e and Fig. S14 (Supporting information), the FE of SnO₂PCNF for formate is higher than 80% over the wide potential ranges from −0.43 V_RHE to −1.23 V_RHE and it reached up to 98.1% at −1.05 V_RHE in the flow cell. Moreover, a formate partial current density of 133 mA/cm² is achieved for SnO₂PCNF at −1.23 V_RHE. The porous carbon nanofibers provide high exposure conditions for SnO₂ nanoparticles dispersed, so the wt% of Sn on PCNF is calculated as low as 3.69%. In Fig. 3f, the mass activity can reach up to 7.09 A/mg at −1.23 V_RHE on SnO₂PCNF, which is 124 times higher than that of bare SnO₂ nanoparticles (0.057 A/mg). In comparison to other previously reported Sn-based catalysts (Fig. 3f and Table S2 in Supporting information), it is evident that the SnO₂PCNF demonstrates a competitive performance in the electrochemical reduction of CO₂ to formate.

  To precisely ascertain the impact of both the inherent characteristics and quantity of active sites during the electrolysis process, we conducted an evaluation of the electrochemical surface area (ECSA) for the freshly prepared catalysts. This evaluation was performed by quantifying the double-layer capacitance (C_dl) in the non-Faradaic potential region within an H-type electrochemical cell (Figs. S15 and S16a in Supporting information). The ECSA of these catalysts is in accordance with the previously investigated surface area from the BET test. Since the catalytic activity of PCNF is mainly hydrogen evolution, the ECSA for SnO₂PCNF is the highest for CO₂RR which is 7 times and 20 times than that of SnO₂CNF and SnO₂, respectively. Electrochemical impedance spectroscopy (EIS) was carried out to gain further insight into CO₂RR kinetics. The Nyquist plots demonstrate that SnO₂PCNF shows much smaller interfacial charge-transfer resistance during the CO₂ reduction process (Fig. S16b in Supporting information), suggesting a favorable faradaic process. The long-term electrolysis tests at −0.85 V_RHE demonstrate the high stability of the SnO₂PCNF catalyst in CO₂RR procedure (Fig. 3g). Furthermore, the catalyst powder loaded on the electrode is easily detached in electrolyte using the traditional drop-coating method. SnO₂PCNF membrane owns good mechanical strength as well as the CO₂RR performance to formate. Meanwhile, the chemical composition and morphology of SnO₂PCNF were also undamaged after these long-time stability tests (Figs. S17 and S18 in Supporting information).

  To monitor the reaction process and the intermediate species of CO₂RR on SnO₂PCNF, we carried out in situ electrochemical Raman spectroscopy tests. In Fig. 4a, the peaks at 410 cm^–1 are ascribed to the Raman phonon modes of E_g and E_2g for the SnO₂ rutile phase structure [38]. Moreover, two obvious Raman peaks at 1068 cm^–1 and 1015 cm^–1 are ascribed to CO₃^2– and HCO₃^– respectively, which act as indicators to reflect the change in pH on the KHCO₃ electrolyte. Evidently, the peak intensities of CO₃^2– gradually increase as the potential negatively shifts, representing a hydrogen proton consumption and the increase in pH value [39,40]. Besides, a stronger peak at 2713 cm^–1 corresponding to HCOO^– was observed, suggesting the promoted production of formate [41]. The intensifying peak located at 2966 cm^–1 is attributed to the key intermediates OCHO^* for the formate product during the electroreduction CO₂ process [42,43].

  Figure 4

  

  Figure 4.  (a) In situ Raman spectra collected of SnO₂PCNF at different applied potentials from OCP to −1.65 V_RHE in a 0.1 mol/L KHCO₃ electrolyte. (b) Charge density difference plots of SnO₂PCNF on the SnO₂-PCNF interface (The red and yellow areas represent charge consumption and accumulation). (c) PDOS of Sn 3d orbit and (d) reaction energy change of CO₂RR to formate on SnO₂, C-SnO₂ and C-N-SnO₂ models.

  DownLoad: Full-Size Img PowerPoint

  To gain deeper insights into the superiority of SnO₂PCNF in terms of support effect and formate production, density functional theory (DFT) calculations were implemented. Firstly, the charge density difference was used to analyze the electron interaction between the SnO₂ active sites and PCNF support. The theory SnO₂PCNF models (denoted as C-N-SnO₂ model) are presented as well as the comparable C-SnO₂ models in Figs. S19-S21 (Supporting information). After optimisation, the charge density difference plots are obtained (Fig. 4b, Figs. S22 and S23 Supporting information). The charge transfer mainly occurs from the Sn atom to the C/N substrate atoms and from the Sn/C atoms to the O atoms at the contact interface between SnO₂ and the C/N or C substrate. Moreover, the charge transfer is deeper in the SnO₂PCNF models, indicating a stronger interaction. The special charge transfer from Sn to C/N/O atoms in SnO₂PCNF is also in accordance with the discussion about the Sn oxidation modulation from the Sn 3d XPS spectra result. Secondly, to further understand the detailed contributions of the electronic structure, we further demonstrate the projected partial density of state (PDOS) in Fig. 4c. Eminently, compared to bulk SnO₂, the surface on C-SnO₂ and C-N-SnO₂ has displayed significantly downshifted 3d orbitals, suggesting the electronic structure modulations between the SnO₂ and the C or C/N substrate. The optimum electron transfer efficiency for the surface shows a higher electroactivity for CO₂RR. To understand the reaction process, the CO₂ reduction energy changes are compared to these three theory models. As previously detected by Raman spectroscopy, the intermediate ^*OCHO is the key to formate formation. The optimized ^*OCHO adsorption on the related theory models are shown in Figs. S24-S26 (Supporting information). The calculation for the free energy on CO₂RR process in Fig. 4d implies a lower energy barrier from CO₂ to the key intermediate ^*OCHO of 0.86 eV on C-N-SnO₂ than that of SnO₂ (2.1 eV) and C-SnO₂ (1.27 eV). This process also represents a rate-determining step for CO₂RR, thus indicating an excellent performance of SnO₂PCNF about higher formate selectivity and lower reaction applied potentials.

  In conclusion, the SnO₂ nanoparticles have been successfully loaded on freestanding porous carbon fibers to achieve high performance of CO₂RR to formate with 98.1% selectivity. According to DFT calculations and in situ Raman spectroscopy, the interaction between SnO₂ and porous carbon fibers modulates the electron structure of Sn active centers with a stronger electron transfer to the carbon substrate. The alteration in electron state density precipitates a shift in the d-band center, thereby promoting the adsorption of pivotal intermediates involved in formate formation. In addition, the porous structure is beneficial to improve the exposure and utilization rate of the active sites. The mass activity can reach up to 7.09 A/mg on SnO₂PCNF, which is 124 times higher than that of support-free SnO₂ nanoparticles (0.057 A/mg). This research offers valuable insights for the development of freestanding electrodes with high electrocatalytic activity and metallic utilization.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 22172099, U21A20312), Guangdong Basic and Applied Basic Research Foundation (Nos. 2023A1515012776, 2022B1515120084), and the Shenzhen Science and Technology Program (No. RCYX20200714114535052). We also acknowledge the Instrumental Analysis Centre of Shenzhen University for performing TEM.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109555.

  
  
• 1. [1]

     S. Perathoner, G. Centi, ChemSusChem 7 (2014) 1274–1282. doi: 10.1002/cssc.201300926

  2. [2]

     D. McCollum, N. Bauer, K. Calvin, A. Kitous, K. Riahi, Climatic Change 123 (2014) 413–426. doi: 10.1007/s10584-013-0939-5

  3. [3]

     E.V. Kondratenko, G. Mul, J. Baltrusaitis, G.O. Larrazábal, J. Pérez-Ramírez, Energy Environ. Sci. 6 (2013) 3112–3135. doi: 10.1039/c3ee41272e

  4. [4]

     R.S. Haszeldine, Science 325 (2009) 1647–1652. doi: 10.1126/science.1172246

  5. [5]

     M. Gong, C. Cao, Q.L. Zhu, EnergyChem 5 (2023) 100111. doi: 10.1016/j.enchem.2023.100111

  6. [6]

     S. Garg, M. Li, A.Z. Weber, et al., J. Mater. Chem. A 8 (2020) 1511–1544. doi: 10.1039/C9TA13298H

  7. [7]

     S.A. Al-Tamreh, M.H. Ibrahim, M.H. El-Naas, et al., ChemElectroChem 8 (2021) 3207–3220. doi: 10.1002/celc.202100438

  8. [8]

     Y. Wang, D. Wang, Y. Li, Adv. Mater. 33 (2021) 2008151. doi: 10.1002/adma.202008151

  9. [9]

     C. Tang, Y. Zheng, M. Jaroniec, S.Z. Qiao, Angew. Chem. Int. Ed. 60 (2021) 2–21. doi: 10.1002/anie.202014556

  10. [10]

      D.D. Ma, S.G. Han, S.H. Zhou, et al., CCS Chem. 5 (2023) 1827–1840. doi: 10.31635/ccschem.022.202202294

  11. [11]

      G. Chen, D. Ye, R. Chen, et al., J. CO2 Util. 44 (2021) 101409. doi: 10.1016/j.jcou.2020.101409

  12. [12]

      S. Zhao, S. Li, T. Guo, et al., Nano-Micro Lett. 11 (2019) 1–19. doi: 10.1007/s40820-018-0235-z

  13. [13]

      F. Cheng, X. Zhang, K. Mu, et al., Energy Technol. 9 (2021) 2000799. doi: 10.1002/ente.202000799

  14. [14]

      T.D. Nguyen-Phan, J.E. Ellis, A.V. Nagarajan, et al., Appl. Catal. B: Environ. 340 (2024) 123250. doi: 10.1016/j.apcatb.2023.123250

  15. [15]

      Y. Pei, W. Gu, S. Cheng, et al., ACS Catal. 13 (2023) 12082–12091. doi: 10.1021/acscatal.3c02630

  16. [16]

      H. Wang, N. Wen, Y. Wang, et al., Adv. Funct. Mater. 32 (2023) 2303473.

  17. [17]

      Q. Zhang, M. Sun, C.Y. Yuan, et al., ACS Catal. 13 (2023) 7055–7066. doi: 10.1021/acscatal.3c00688

  18. [18]

      P.F. Yin, J. Fu, Q. Yun, et al., Adv. Mater. 34 (2022) 2201114. doi: 10.1002/adma.202201114

  19. [19]

      M. Chu, C. Chen, Y. Wu, et al., Green Energy Environ. 7 (2022) 792–798. doi: 10.1016/j.gee.2020.12.001

  20. [20]

      J.H. Kim, H. Woo, J. Choi, H.W. Jung, Y.T. Kim, ACS Catal. 7 (2017) 2101–2106. doi: 10.1021/acscatal.6b03706

  21. [21]

      T.W. van Deelen, C. Hernández Mejía, K.P. de Jong, Nat. Catal. 2 (2019) 955–970. doi: 10.1038/s41929-019-0364-x

  22. [22]

      Y. Pi, Z. Qiu, Y. Sun, et al., Adv. Sci. 10 (2023) 2206096. doi: 10.1002/advs.202206096

  23. [23]

      Y. Lou, J. Xu, Y. Zhang, et al., Mater. Today Nano 12 (2020) 100093. doi: 10.1016/j.mtnano.2020.100093

  24. [24]

      G. Liu, Q.T. Trinh, H. Wang, et al., Small 19 (2023) 2301379. doi: 10.1002/smll.202301379

  25. [25]

      X. Wang, T. Bürgi, Angew. Chem. Int. Ed. 62 (2023) e202300146. doi: 10.1002/anie.202300146

  26. [26]

      Y. Li, T. Xu, Q. Huang, et al., ACS Catal. 13 (2023) 7597–7605. doi: 10.1021/acscatal.3c01610

  27. [27]

      Z. Gao, J. Li, Z. Zhang, W. Hu, Chin. Chem. Lett. 33 (2022) 2270–2280. doi: 10.1016/j.cclet.2021.09.037

  28. [28]

      C. Liu, Q. Li, W. Kang, et al., J. Mater. Chem. A 10 (2022) 10–49. doi: 10.1039/D1TA08646D

  29. [29]

      M. Wang, J. Yang, X. You, et al., Int. J. Hydrogen Energy 47 (2021) 2114–2123.

  30. [30]

      H. Yang, X. Wang, Q. Hu, et al., Small Methods 4 (2020) 1900826. doi: 10.1002/smtd.201900826

  31. [31]

      Q. Wang, F. Liu, Z. Jin, et al., Adv. Funct. Mater. 30 (2020) 2002580. doi: 10.1002/adfm.202002580

  32. [32]

      Y. Zhang, X. Xia, X. Cao, et al., Adv. Energy Mater. 7 (2017) 1700220. doi: 10.1002/aenm.201700220

  33. [33]

      S. Ding, Q. Chen, S. Chen, Y. Tian, J. Zhang, Chin. Chem. Lett. 34 (2023) 108232. doi: 10.1016/j.cclet.2023.108232

  34. [34]

      H. Chen, L. Ding, W. Sun, et al., RSC Adv. 5 (2015) 56401–56409. doi: 10.1039/C5RA10268E

  35. [35]

      X. Jiang, X. Li, Y. Kong, et al., Small 18 (2022) 2204148. doi: 10.1002/smll.202204148

  36. [36]

      J. Yang, R. Hübner, J. Zhang, et al., Angew. Chem. Int. Ed. 60 (2021) 9590–9597. doi: 10.1002/anie.202015679

  37. [37]

      G. Jia, Y. Wang, M. Sun, et al., J. Am. Chem. Soc. 145 (2023) 14133–14142. doi: 10.1021/jacs.3c04727

  38. [38]

      M. Sharma, R.N. Aljawfi, K. Kumari, et al., J. Mater. Sci. Mater. Electron. 30 (2019) 760–770. doi: 10.1007/s10854-018-0345-x

  39. [39]

      Y. Zhao, X.G. Zhang, N. Bodappa, et al., Energy Environ. Sci. 15 (2022) 3968–3977. doi: 10.1039/D2EE01334G

  40. [40]

      X. Lu, C. Zhu, Z. Wu, et al., J. Am. Chem. Soc. 142 (2020) 15438–15444. doi: 10.1021/jacs.0c06779

  41. [41]

      A. Heyns, J. Chem. Phys. 84 (1986) 3610–3616. doi: 10.1063/1.450197

  42. [42]

      I.V. Chernyshova, P. Somasundaran, S. Ponnurangam, Proc. Natl. Acad. Sci. U. S. A. 115 (2018) E9261–E9270.

  43. [43]

      Z. Li, B. Sun, D. Xiao, et al., Angew. Chem. Int. Ed. 62 (2023) e202217569. doi: 10.1002/anie.202217569

• 

  Figure 1  (a) Synthesis scheme of the SnO₂PCNF. (b) The free-standing fiber membrane of SnO₂PCNF. (c) TEM image of SnO₂PCNF. (d, e) HR-TEM images of SnO₂PCNF. (f) HAADF-STEM-EDS elemental mapping images for SnO₂PCNF.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) XRD patterns, (b) N₂ adsorption/desorption isotherms, (c) CO₂ adsorption isotherms, (d) FT-IR spectra of PCNF, SnO₂, SnO₂CNF and SnO₂PCNF. (e) Sn 3d XPS spectra, (f) valence band spectra of SnO₂, SnO₂CNF and SnO₂PCNF.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Comparisons of LSV curves of PCNF, SnO₂, SnO₂CNF and SnO₂PCNF in CO₂-saturated 0.1 mol/L KHCO₃ aqueous. (b) FEs of CO₂RR products on SnO₂PCNF. Comparisons of (c) the partial current densities for formate and (d) C₁ products at different applied potentials on PCNF, SnO₂, SnO₂CNF and SnO₂PCNF. (e) FE and mass active current for formate on flow-cell. (f) CO₂RR to formate performance comparisons of SnO₂PCNF with other Sn-based catalysts. (g) Long-term durability of SnO₂PCNF at −0.85 V for 21 h.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) In situ Raman spectra collected of SnO₂PCNF at different applied potentials from OCP to −1.65 V_RHE in a 0.1 mol/L KHCO₃ electrolyte. (b) Charge density difference plots of SnO₂PCNF on the SnO₂-PCNF interface (The red and yellow areas represent charge consumption and accumulation). (c) PDOS of Sn 3d orbit and (d) reaction energy change of CO₂RR to formate on SnO₂, C-SnO₂ and C-N-SnO₂ models.

  下载: 全尺寸图片 幻灯片
•