Au/Pb Interface Allows the Methane Formation Pathway in Carbon Dioxide Electroreduction

: The electrochemical conversion of carbon dioxide (CO 2 ) to high-value chemicals is an attractive approach to create an arti ﬁ cial carbon cycle. Tuning the activity and product selectivity while maintaining long-term stability, however, remains a signi ﬁ cant challenge. Here, we study a series of Au − Pb bimetallic electrocatalysts with di ﬀ erent Au/Pb interfaces, generating carbon monoxide (CO), formic acid (HCOOH), and methane (CH 4 ) as CO 2 reduction products. The formation of CH 4 is signi ﬁ cant because it has only been observed on very few Cu-free electrodes. The maximum CH 4 formation rate of 0.33 mA cm − 2 was achieved when the most Au/Pb interfaces were present. In situ Raman spectroelectrochemical studies con ﬁ rmed the stability of the Pb native substoichiometric oxide under the reduction conditions on the Au − Pb catalyst, which seems to be a major contributor to CH 4 formation. Density functional theory simulations showed that without Au, the reaction would get stuck on the COOH intermediate, and without O, the reaction would not evolve further than the CHOH intermediate. In addition, they con ﬁ rmed that the Au/Pb bimetallic interface (together with the subsurface oxygen in the model) possesses a moderate binding strength for the key intermediates, which is indeed necessary for the CH 4 pathway. Overall, this study demonstrates how bimetallic nanoparticles can be employed to overcome scaling relations in the CO 2 reduction reaction.


The chemical state of Au-Pb NPs
To determine the peak shape of metallic Pb 0 and Pb-oxide species for XPS fitting, we measured the XP spectra of the as-received Pb foil and the same foil after Ar + sputtering. As seen in Figure   S4A, the XPS survey scan shows only Pb, O, and C peaks, which confirms a clean surface and no contamination from other metals within the detection limits of XPS. The high-resolution XPS scan ( Figure S4B) shows that the as-received Pb foil exhibits peaks corresponding to Pb 0 (136.62 eV) and Pb 4+/2+ (138.52 eV). The latter predominant peak is likely due to the native oxide layer, which forms rapidly on Pb upon exposure to air. 1 Table S1. Bulk and surface composition of Au-Pb bimetallic NPs. * *XPS shows the ratio between Au and Pb-oxide species on the surface.
The bulk atomic ratio of Au and Pb is close to the stoichiometric ratio of the constituent metals.
However, the surface is slightly Pb rich, which might be caused by the self-nucleation of Au at higher concentrations.        Ten hours of electrolysis was conducted at -1.07 V vs. RHE to assess the stability of Au 50 Pb 50 catalyst and verify the continuous production of CH 4 ( Figure S16). The total current stabilized at -13 mA cm -2 after 1 h and remained constant. The FE CH 4 varied within 2.8 -2.1%. There was a drop in FE CO to 17 % after 5 h. This might be due to the decrease in the reactant (CO 2 ) concentration, where the gas chromatography curve showed that the areas of CO and CO 2 became comparable. We used a gas-tight two-compartment H-type cell for electrolysis. The electrolyte was purged with CO 2 for 30 min, and then the flow was stopped before electrolysis.

Isotopic Labeling Experiment
In the experiment, the two carbon sources (CO 2 gas and KHCO 3 ) were labeled. 35 mL of 0.5 M KH 13 CO 3 was purged at first with Ar for 45 min to remove all the dissolved gases, and then

S14
We performed a CO electrolysis experiment, and CH 4 was produced with a 4.9 % FE, which is comparable to that in CO 2 reduction, but it was detected at an earlier time (the first 30 min) ( Figure S19). As reported previously that the reduction of CO 2 to CO is the first step in the production of hydrocarbons on Cu 4,5 and Cu-free surfaces. 6 This observation suggest that CO is an intermediate in the pathway of CO 2 R to CH 4 on the Au-Pb catalysts. The FE HCOOH decreased to 13 % compared to 27% in CO 2 reduction, this could be explained by the less amount of CO 2 that was derived from the equilibrium of HCO 3 near the electrode surface.

Computational Details
The Computational Hydrogen Electrode (CHE) 4,12,13 was used for modelling the Gibbs free energy for each path. For each reaction, the Gibbs energy is given by Where H is the enthalpy given by DFT calculation, ZPE is the zero-point energy, T is the temperature, k b is the Boltzmann constant, e is the electron electric charge and U(ext) is the eapplied voltage.
The initial bulk structures for pure Au and pure Pb were taken from the Materials Project. 14

Surface Structures
Here are the most stable surfaces where the reactions path were computed:     The difference between the structures in Figures S26 and S27 is that the O is lower in the subsurface of Pb-Au3%Od (111) and is more stable by 0.1 eV compared to Pb-Au3%O(111).
All input and output files can be accessed at the ioChem-BD database 15  The obtained energy barriers between steps were at most 0.7 eV lower than the applied voltage of 1.07 eV. The following HCOOH paths were investigated: Path 1a