Due to its potential to replace fossil fuels in chemical processes by converting carbon dioxide (CO₂) to industrial base reactants, the electroreduction of CO₂ is receiving tremendous attention. Among the large number of electrocatalysts currently under investigation for this reaction, palladium (Pd) is increasingly appealing due to its ability to produce both carbon monoxide (CO) and formate at high selectivities depending on the applied potential. This unique change in selectivity between CO and formate at high vs. low overpotentials (i.e., -0.5 to -1.0 vs. -0.1 to -0.4 V vs. the reversible hydrogen electrode (V_RHE), respectively) has been described multiple times on different Pd-electrocatalysts [1-3]. Those mechanistic investigations mainly focused on understanding the ability of palladium to form hydride (PdH_x) under CO₂ electroreduction conditions. While β-PdH is usually found at high overpotentials, the formation of α-PdH is described at low overpotentials, and a mixed-phase between both compositions is generally reported at intermediate overpotentials [1,3-5]. As a result, the CO₂ reduction reaction (CO₂RR) is assumed to take place on β-PdH at the high overpotentials leading to CO production, while α-PdH or a mixed α/β-phase is believed to be the active phase at low overpotentials associated with high formate yields [1,3,4]. However, large differences between the potential of β-PdH formation, as well as the extent of the mixed hydride state (i.e., featured as a sudden transition between α- and β-phases [3,4], or extending over hundreds of millivolts [1,5]), have been reported. In parallel to this, Pd-surface poisoning with CO is known to influence palladium’s CO₂RR selectivity [2,3] while also slowing down hydride formation [6]. However, the precise interplay between surface-adsorbed CO and the formation of PdH, along with their combined effects on the mechanism of CO₂ electroreduction remains poorly understood.

In light of this, in this work, we investigated these effects using an unsupported Pd-aerogel with a web thickness of ≈ 6 nm, synthesized via a novel ethanolic synthesis approach [7]. We first analyzed the influence of the electrolyte composition (i.e., presence vs. absence of CO₂ and bicarbonate) on the formation of surface-adsorbed CO and PdH by performing potential holds followed by positive linear sweep voltammograms in a rotating disk electrode (RDE) setup. As shown in Fig. 1A, these voltammograms featured two oxidation peaks at ≈ 0.8 vs. ≈ 1.0 V_RHE that we could assign to the desorption of absorbed hydrogen and the oxidation of adsorbed CO, respectively. This in turn allowed us to track the formation of both of these species in relation to the applied potential and duration of the potential holds. Under CO₂ electroreduction conditions, a full monolayer of CO was found to form across the entire CO₂ electroreduction potential range. Moreover, large differences in the rate of PdH formation as a function of the applied potential were found in the presence of CO₂ and/or bicarbonate, while in CO₂- and bicarbonate-free electrolytes hydride formation is quasi-instantaneous at all potentials relevant for CO₂ electroreduction. Time-dependent RDE investigations at constant potentials also revealed an initial decrease in the hydride content of the PdH phase during the formation of the CO monolayer, hinting at the involvement of this hydride in a hydrogenation step previously described as part of the CO₂ electroreduction mechanism at low overpotentials [2]. Finally, these experiments were complemented with time-resolved in-situ X-ray absorption measurements that confirmed the formation of full β-PdH at all applied overpotentials under CO₂ electroreduction conditions (see Fig. 1B), however, differences in the time-dependence of PdH formation on the applied potential were revealed.

In summary, this contribution showcases the potential- and time-dependent CO monolayer formation under CO₂RR conditions and its influence on PdH-formation. Specifically, our results reveal the independence of the final PdH-stoichiometry on the applied potential, while exposing the effect of the latter variable on the rates of CO-monolayer and β-PdH formation. On this basis, the mechanisms of CO₂ electroreduction on Pd that have been proposed so far should be revisited to consider the simultaneous presence of full β-PdH and a near-full monolayer of CO at all overpotentials unveiled by our results.

References:

[1] D. Gao et al., Nano Res. 2017, 10 (6), 2181-2191.

[2] X. Min et al., J. Am. Chem. Soc. 2015, 137 (14), 4701-8.

[3] W. Zhu et al., Adv. Energy Mater. 2019, 9 (9), 1802840.

[4] W. Sheng et al., Energy Environ. Sci. 2017, 10 (5), 1180-1185.

[5] J.H. Lee et al., Chem. Commun. 2020, 56 (1), 106-108.

[6] B. Łosiewicz et al., J. Electroanal. Chem. 2007, 611 (1-2), 26-34.

[7] M. Georgi et al., Mater. Chem. Front. 2019, 3 (8), 1586-1592.