Exploring the adsorption site coordination as a strategy to tune copper catalysts for CO₂ electro-reduction

Abstract

The electrochemical reduction of CO2 is a promising technology to reach a carbon-neutral economy. However, among other challenges, the design of active and selective catalysts still bottlenecks such advances. Herein,...

What do we talk about, when we talk about single-crystal termination-dependent selectivity of Cu electrocatalysts for CO2reduction? A data-driven retrospective

“…We recall that previous multi-scale site-counting methods successfully enabled to bridge NEAS distributions characterized by their coordination with the performance of catalysts for oxygen and CO 2 reduction reaction. [23][24][25][26][27][28] Similarly coordination based representation were suitable to predict surface energies of nanostructured surfaces. 23,24 We utilize coordination number features under the assumption that the original coordinations distribution is preserved inoperando, or that it is a variable which anyway affects the catalyst activity if the latter undergoes any structural rearrangement.…”

“…15,16 Machine learning (ML) methods have been indeed exploited to systematically assess the effect of catalyst compositions and stoichiometry, [17][18][19][20][21] as well the interplay between the catalyst structure and its activity and selectivity. [22][23][24][25][26][27][28] When ML workflows rely on a database of DFT calculations, these methods have access to a full knowledge and tunability of the electronic and geometric structure of the systems under consideration. ML-accellerated DFT screening then enables to predict and rationalize the catalytic properties of model systems as a function of their composition and geometry.…”

“…Paradigmatic examples in these area include the optimization of bimetallic and/or nanostructured catalysts for small molecule conversion, e.g., oxygen reduction, hydrogen evolution, CO 2 reduction, hydrogen peroxide formation, N 2 reduction. [17][18][19][20][21][22][23][24][25][26][27][28] Yet, experiments involve a chemistry that is often more complex than the one encoded in models. In turn, translating DFT-based suggestions to the laboratory is often non-trivial.…”

“…15,16 Machine learning (ML) methods have been indeed exploited to systematically assess the effect of catalyst compositions and stoichiometry, 17–21 as well the interplay between the catalyst structure and its activity and selectivity. 22–28…”

From Single Crystal to Single Atom Catalysts: Structural Factors Influencing the Performance of Metal Catalysts for CO2 Electroreduction

“…Also, significant restructuring under the applied potential was observed. Even though defects in Cu single crystals have both experimentally and computationally been shown to improve the CO 2 RR to hydrocarbons, − the inactivity of the clean Cu(100) and Cu(111) as well as the significant restructuring of the catalyst under operation is a nontrivial result. This raises the question of how fast and to what extent single crystals reconstruct under applied potential.…”