Spontaneously Sn Doped Bi/BiO x Core–shell Nanowires Toward High-Performance CO 2 Electroreduction to Formate

dealloying method. The resultant Bi/Bi(Sn)O x NWs yielded excellent eletrochemical CO 2 reduction activity with FE over 92% at a wide potential range (-0.5 ~ -0.9 V vs. RHE) and a high current density ( ＞ 300 mA cm -2 at -1.0 V) in the gas diffusion cell. Moreover, Bi/Bi(Sn)O x NWs showed stable current density and FE HCOOH without significant decay during chronoamperometry test over 10 h. In-situ Raman and DFT calculations suggested that the incorporation of Sn atoms into BiO x surfaces that stabilizes the *OCHO intermediate and suppresses the HER. The present work opens a new avenue to develop new core/shell catalysts with the desired control on reaction pathways and target products for highly efficient CO 2 RR, but offers a facile and economic strategy to create new atom doped oxide catalysts reconstructed from bimetallic surfaces for boosting other chemical reactions.

to convert CO2 into formate due to their strong adsorption of CO2 intermediates and poor activity for HER [10][11][12][13] . Unfortunately, various Bi-based oxides usually suffer from undesirable activity in a low and wide cathodic potential window owing to their relatively low intrinsic electrical conductivity and low selectivity toward formate production [14][15][16] . One feasible strategy for promoting electrical conductivity is to prepare core-shell nanostructured catalysts with a highly conductive metal core and thin metal oxide shell, which can potentially modulate the electronic properties and tune the binding strength of the intermediates [17][18][19] . To further promote the activity of electrocatalysts, engineering new Bi-based materials by incorporating other components has emerged as an efficient strategy to modulate the reactant adsorption, activation and product desorption, such as Bi2O3 nanosheets modified by nitrogendoped graphene quantum dots or multi-channel carbon matrix 13,20 , Sn-Bi eutectic alloys [21][22][23] , hybrids of Sn nanosheets and Bi nanoparticles 24 , Bi@Sn core-shell nanoparticles 17 , which are promising candidates for the electrochemical reduction of CO2 to formate. Obviously, the problems of low formate selectivity and high competition with hydrogen evolution at low overpotentials remain unsolved even by using flow cell reactors.
In this work, we present novel Sn doped Bi/BiOx nanowires with high electrical conductivity Bi metallic core and amorphous Sn-doped BiOx shell through electrochemical dealloying. The Bi/Bi(Sn)Ox NWs exhibits an excellent catalytic activity and selectivity for electrochemical electrochemical CO2 reduction to formate over a wide negative potential range, with a Faradaic efficiency (FE) > 92% from -0.5 to -0.9 V versus reversible hydrogen electrode (RHE), and a maximum FE value of 98 ± 2 % at −0.7 V under gas diffusion cell configuration. The impressive performance is attributable to the incorporation of Sn atoms into BiOx species, which can modulate the electron states of Bi, allowing the *OCHO intermediate to favorably adsorb onto the Bi(Sn)Ox surface while promotes formate production by suppressing the competitive hydrogen evolution reaction.

Results
Material synthesis and characterization. A schematic illustration of the fabrication process toward novel Bi/Bi(Sn)Ox NWs is shown in Figure 1a. By utilizing the disparity in electrochemical stability of the Bi and Sn phases, the Sn atoms were priority dissolved from the arc-melt spinning Bi1Sn99 (at%) alloy ribbons by galvanic corrosion in 0.25 M H2SO4 solution at the applied potential of -0.3 V versus Ag/AgCl (Supplementary Fig. 1) 25,26 . During the dealloying of the Bi-Sn alloys, as the Sn atoms are continuously dissolved into the electrolyte from the Sn phase, the Bi atoms would be released and reorganized into wellarranged single-phase Bi-based nanowires along the direction of etching ( Fig.   1b) 27,28 . Scanning electron microscopy (SEM) image (Fig. 1c) displays the continuous nanowire structure of the etched ribbons. The corresponding energy dispersive X-ray spectroscopy (EDS) analysis provides a chemical composition of Bi/Sn atom ratio is ~95:5 (Supplementary Fig. 2). Aberration-corrected highangle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image clearly shows that crystalline range of Bi/Bi(Sn)Ox NWs with diameter of ~30 nm is surrounded by thick amorphous-like shells (Fig. 1d). The selected-area electron diffraction (SAED) exhibits a single diffraction pattern, which could be indexed as a rhombohedral Bi structure (JCPDS No. 44-1246) with a growth direction of [101] (Fig.1e). The lattice distance measured from fast Fourier transformation (FFT) of HAADF-STEM image (Fig. 1f) (Fig. 2b). The surface adsorbed oxide (531.8 eV) was also removed after the Ar + etch (Fig. 2c). Specifically, the surface compositions obtained from XPS agree with the SEM-EDS results (Supplementary Table 1). The Bi L3-edge X-ray absorption near edge structure (XANES) spectra of the Bi/Bi(Sn)Ox NWs in Figure 2d reveal visible similarities with that of the Bi and Bi2O3 reference. These similarities are to be expected, given that Sn atoms in both Bi2O3 and BiOx lattices are expected to be octahedrally coordinated by oxygen atoms, leading to similarities in their structural and electronic properties. Meanwhile, the XANES spectra apparently reveal that Bi L3-edge position of the Bi/Bi(Sn)Ox NWs is located between the Bi foil and Bi2O3, indicating coexistence of metallic Bi and oxidation state Bi species in the core-shell structure. The corresponding Fourier-transformed extended x-ray absorption fine structure (FT-EXAFS) spectra and inversed FT-EXAFS are presented in Figure 2e, f, respectively.
The similar positions of the metallic Bi peak positions suggest that these bonds lengths in the Bi/Bi(Sn)Ox NWs are similar to those in the Bi reference, strongly indicating the metallic Bi-core structure. In stark contrast, the observed Bi-O peaks appear at distinctly positions in the Bi/Bi(Sn)Ox NWs spectra, revealing a distinct structural difference relative to the pure Bi reference. Interestingly, the formation of Bi2O3, SnO2 can be predicted from the Bi-Sn-O ternary phase diagram at low Sn content (region iii, Supplementary Fig. 3) 23 , which is consistent with our XPS and XAS results.
According to the above-mentioned surface and bulk sensitive techniques, we can conclude that the prepared Bi-based nanowires display a core-shell structure with metallic Bi cores and amorphous Bi(Sn)Ox shells. In fact, the synergetic effect of the metallic core and the amorphous oxide shell in providing both good bulk conductivity and surface activity is beneficial to high performance CO2 electroreduction. formate, with above 92% Faradaic efficiency in a wide potential range (from −0.5 to −0.9 V) and a maximum FE of nearly 100% at -0.7 V versus RHE (Fig. 3c). In contrast, Bi, Bi2O3, and SnO2 catalysts show lower selectivity towards formate, while CO or H2 dominates in the low potential range (Supplementary Fig. 9). In addition, the partial current density of formate on Bi/Bi(Sn)Ox NWs is significantly higher than that of commercial Bi, SnO2, and Bi2O3 nanopowders (Fig. 3d), showing a desirable formate current density of ~100 mA cm −2 at a maximum formate FE of −0.7 V. Compared with reported Bi-based catalysts for electrochemical CO2 reduction, Bi/Bi(Sn)Ox NWs maintain the highest formate selectivity over a wide potential range (Supplementary Table 3). The energy efficiency (EE) of the Bi/Bi(Sn)Ox NWs is further evaluated, which shows a high EE over 60% in a wide potential range (Supplementary Fig. 10).
Tafel analysis is indispensable for understanding the kinetics of CO2 electroreduction.
The Bi/Bi(Sn)Ox NWs exhibit a much lower Tafel slope of 124 mV dec −1 , suggesting that the chemical rate-determining step (RDS) is the first electron transfer step to generate surface adsorbed CO2 • − species (Supplementary Fig. 11) [29][30][31] . Furthermore, the Bi/Bi(Sn)Ox NWs displays remarkably higher CO2 adsorption capability compared with commercial Bi and Bi2O3 (Supplementary Fig. 12). This could lead to CO2 enrichment on the local working electrode surface, which is beneficial for CO2 activation and reduction 32,33 . Electrochemical impedance spectroscopy (EIS) analysis further verifies that the Bi/Bi(Sn)Ox NWs catalyst with metallic core-shell structure could generate small internal resistance and rapid charge transfer behavior for a low onset potential and fast CO2 reduction reaction (CO2RR) kinetics (Supplementary Fig.   13). To evaluate the intrinsic activity of the catalysts, we measured the electrochemically active surface area (ECSA) for Bi/Bi(Sn)Ox NWs, Bi, Bi2O3, and SnO2 (Supplementary Fig. 14) and evaluated their ECSA-normalized partial current densities for formate production, hydrogen evolution, and CO production ( Supplementary Fig. 15). Strikingly, Bi/Bi(Sn)Ox NWs show lower HER and CO production rates, which results in it having higher selectivity for formate production than pure Bi, Bi2O3, and SnO2. These results confirm that the presence of metallic Bi core in the hybrid catalysts could decrease the energy barrier and facilitate the chargetransfer process from the electrode to the adsorbed CO2, contributing to excellent selectivity and ultrahigh current density. In addition, a long-term stability testing on the Bi/Bi(Sn)Ox NWs catalyst shows a negligible current decay and a steady FE of formate over 10 h at an operated potential of -0.7 V. Furthermore, no detectable compositions and morphology changes were observed on Bi/Bi(Sn)Ox NWs after the long-term operation, implying the superior structural stability (Supplementary Fig. 16, 17). To gain further insights into the CO2RR activity and selectivity toward formate production on the reconstructed Bi/Bi(Sn)Ox NWs, density functional theory (DFT)

Active sites identification and mechanisms of CO
calculations were carried out to analyze the thermodynamic reaction process. The structural models for the DFT calculations were built based on a Bi2O3 layer deposited on the Bi surface according to the above structural identifications (Supplementary Fig.   19). The binding strength of *OCHO intermediates could be inferred from the position of the highest pelectronic states (Ep) and the charge density distribution. In Figure 4b, the Ep of Bi/Bi(Sn)Ox moves more closely to the Fermi level (Ef) and displays a stronger electronic interaction between *OCHO, which also consistent with the stronger electronic interaction between *OCHO and the substrate (Fig. 4b, inset). We also explored the reaction pathways for CO2 reduction to HCOOH and HER. In Figure 4c. accelerate the production of formate (Fig. 4d).

Conclusions
In  For clarity, all LSV data were not corrected for a negligible ohmic drop. The chronoamperometry tests were conducted at each potential for 40 min. The long-term stability measurement was operated at the potential of -0.7 V versus RHE for 10 h.
The current densities were normalized to the geometric surface area. CO2 gas was delivered into the cathodic compartment at a rate of 10 ml/min and was routed into a gas chromatograph (Shimadu, GC-2010 Plus). The catalyst was sprayed on a gas diffusion layer of Sigracet 29 BC with a surface area of 1 cm 2 , the eventually loading mass is ~1 mg cm −2 .
The gaseous products from CO2 electroreduction were quantified by a gas chromatograph equipped with a flame ionization detector (FID) for CO and a thermal conductivity detector (TCD) for H2 quantification. Before the FID, a methanizer is equipped for the detection of CO. High-purity Argon (99.999%) was used as the carrier gas. The Faradaic efficiency of each gas product was calculated by the equation: where e is the number of electrons transferred, F is Faraday constant, n is the mole fraction of the product, and Q is the total charge.
Liquid products were quantified on Bruker 400 MHz NMR spectrometer.
Typically, 500 μl of collected catholyte exiting the reactor was mixed with 100 μl D2O containing 20 ppm (m/m) dimethyl sulphoxide (≥ 99.9%; Alfa Aesar) as the internal standard. The faradic efficiency for HCOOH generation was calculated as follows: where n (mol) is the generated amount of formate; e (= 2) is the number of transferred electrons for 1 mol of formate generation; and t (s) is the electrolysis time.
The cathode energy efficiency was calculated to define the energy utilization toward the HCOOH product using the following formula: where EHCOOH = -0.2 V vs. RHE is the thermodynamic potential for HCOOH formation, 1.23 V is the thermodynamic potential for the oxygen evolution reaction, and Eapplied is the applied potential vs. RHE 14 .
DFT Calculations. The calculations were performed with the VASP package 36