Revealing active Cu nanograins for electrocatalytic CO2 reduction through operando studies

Copper (Cu) has been regarded as a highly efficient electrocatalyst for the conversion of CO2 into a multicarbon product. However, the catalytic mechanism and the active sites of Cu catalysts under operating conditions still remain elusive. Yang's team applied systematic operando characterization techniques to provide a quantitative analysis of the valence states and the chemical environment of Cu nanocatalysts under electrochemical reaction conditions, which clearly reveal the evolution of Cu nanocatalysts before and after the entire electrochemical CO2 reduction.

Environmental issues and energy crisis caused by excessive carbon dioxide (CO 2 ) emissions are severe challenges faced by major economies of the world, which has a serious impact on social and economic development. 13][4][5][6] As an outstanding electrocatalyst for CO 2 RR, copper (Cu) enables CO 2 -to-multicarbon product (C 2+ ), including ethylene, ethanol, and propanol. 7,80][11][12] Very recently, Yang's team 13 applied systematic operando characterization techniques to provide a comprehensive quantitative analysis of the valence states and chemical environment of the Cu nanocatalysts under electrochemical reaction conditions.Importantly, the clear and intuitive experimental evidence reveals the evolution of Cu nanocatalysts before and after the entire CO 2 RR.
In their previous work, 14,15 Yang et al. reported that small-scale spherical Cu nanocatalysts could show excellent CO 2 reduction activity and multicarbon product selectivity.However, because of the highly active properties of Cu nanocatalysts in the electrochemical reaction process, traditional ex situ characterization techniques can only observe the core-shell structure of Cu and cuprous oxide (Cu@Cu 2 O) in the initial state and the single-crystal Cu 2 O nanocubes formed after exposure to air following CO 2 RR.In this work, they further used the operando techniques to fully reveal the evolution process of Cu nanocatalysts in the reaction process by taking 7, 10, and 18 nm Cu nanoparticles (NPs) as the model catalyst.As shown in Figure 1, the evolution process includes the initial removal of ligand and reduction of the Cu 2 O oxide layer on the catalyst surface, the initial agglomeration and further structural transformation of Cu nanograins, and the final formation of highly polycrystalline metallic Cu nanograins rich in nanograin boundaries.Significantly, the formed nanograin boundaries contain a large number of active undercoordinated Cu sites, which can efficiently convert CO 2 into ethylene, ethanol, propanol, and other multicarbon products.However, when exposed to air after CO 2 RR, the structure rich in nanograin boundaries would also activate the double bond of oxygen molecules and promote the rapid dissociation and insertion of oxygen atoms in the Cu lattice, thus quickly restructured to single-crystal well-defined Cu 2 O nanocubes.The aforementioned process is an important reason why the actual reaction sites of Cu catalysts are difficult to be observed in the previous large number of ex situ experiments.
To track the morphology changes of Cu NPs in liquids during CO 2 RR, the improved operando electrochemical scanning transmission electron microscopy (EC-STEM) cell with a significantly enhanced spatial resolution was applied, which could ensure reliable electrochemical reaction conditions.In this work, the operando EC-STEM of the 7 nm Cu NP ensemble indicates that there are two types of morphology, that is, loosely connected small Cu nanograins and closely packed large Cu nanograins.Based on the traditional EC-STEM, the authors further developed operando fourdimensional STEM (4D-STEM) with high sensitivity and dynamic range.4D-STEM can obtain high-quality electron diffraction with very low electron dose, which is essential for beam-sensitive materials in liquid.It was found that at 0 V versus reversible hydrogen electrode (vs.RHE), Cu NPs as building blocks initially aggregated into loosely connected or closely overlapped metallic Cu nanograins.Then, at the actual operating potential of CO 2 RR (−0.8 V vs. RHE), Cu nanograins will further aggregate into closely packed highly polycrystalline metal Cu nanograins, which are dominant active Cu sites.After CO 2 RR, once exposed to air, the metallic Cu nanograins would rapidly oxidize into a single crystalline Cu 2 O nanocube with an edge length of about 60-120 nm.The rapid oxidization may arise from the reactivity of O 2 molecules with metallic Cu nanograins formed during  13 CO 2 RR, which could cause dissociation of O═O and spontaneous insertion of oxygen atoms in the Cu lattice.
In addition to 7 nm Cu NPs, the dynamic morphology changes of 10 and 18 nm Cu NPs under electrochemical conditions were also studied by operando EC-STEM. 10 nm Cu NPs showed similar structural evolution with 7 nm NPs, that is, substantial movement/aggregation into small Cu nanograins within several seconds and subsequent continuous growth into large Cu nanograins.In contrast, 18 nm Cu NPs underwent different structural evolution, that is, partially melting into Cu nanograins at the initial stage of linear sweep voltammetry and subsequently reconstructing into large Cu nanograins within about 4 minutes at −0.8 V versus RHE.Combined with Pb underpotential deposition and operando resonant soft X-ray scattering analysis, the above-mentioned results suggest that 7 nm Cu NP ensemble owns a higher density of nanograin boundaries and a higher density of undercoordinated sites than those of 10 and 18 nm Cu NPs.
To accurately study the chemical valence states and coordination environment of Cu NPs during the reduction and reoxidation life cycle, operando high-energyresolution fluorescence detected X-ray absorption spectroscopy (HERFD-XAS) was conducted.Particularly, operando HERFD X-ray absorption near-edge spectroscopy was used to trace the reaction valence states of 7 and 18 nm nanocatalysts.7 nm Cu NP ensemble could completely transform into active metallic Cu during the reaction process.In contrast, only ~30% of the 18 nm Cu NP ensemble, that is, the Cu 2 O oxide layer on the surface, could convert into activated Cu.The quantitative structure-activity correlation suggests that a higher fraction of metallic Cu nanograins could lead to higher C 2+ selectivity.Specifically, the C 2+ selectivity of the 7 nm Cu NP ensemble is sixfold higher than that of the 18 nm Cu NP ensemble.Furthermore, operando extended edge X-ray fine spectroscopy demonstrates that the coordination number of Cu NP ensemble after 1 h of CO 2 RR is only about 8.During the electroreduction process, due to the aggregation/coalescence of Cu nanograins, the coordination number of the Cu NP ensemble increased to 12 after 4 h of CO 2 RR. 16The XAS results confirm that the formation of polycrystalline nanograin boundaries could provide active Cu sites with low coordination numbers.
In conclusion, this study presents a milestone in the operando characterization and tracing of catalyst structures under the condition of liquid-phase electrochemical reaction.Further combined with operando electron microscopy and operando synchrotron radiation X-ray techniques, this study provides a paradigm for the comprehensive analysis of the morphology, structure, chemical valence state, and coordination environment of catalysts under real reaction conditions.It is believed that the combination of multiple operando characterization techniques would provide a powerful method for characterizing the actual active sites and their fine evolution in electrochemical reaction conditions associated with other energy applications.

F
I G U R E 1 (A) Scheme of the life cycle of Cu nanocatalysts during the CO 2 RR and on exposure to air.HAADF-STEM images of Cu nanocatalysts at (B) 0 V and (C) −0.8 V, and (D) Cu 2 O nanocubes formed on air exposure.Reproduced with permission: Copyright 2023, Springer Nature.