Nature of the active sites for carbon dioxide reduction on metal nanoparticles: suggestions for optimizing performance

Abstract

Metal nanoparticles exhibit superior carbon dioxide (CO_2) redn. performance due to the existence of unique surface active sites. For example, gold (Au) nanoparticles (NPs) show high efficiency in reducing CO_2 to carbon monoxide (CO), and copper (Cu) NPs efficiently convert CO to C2 productions. However, due to the difficulties in operando measurements, the at. details of these actives site have never been revealed. In this work, we employed multi-scale simulation technologies to "computationally synthesized" the Au NPs and Cu NPs on carbon nanotube (CNT) support by closely simulating the chem. vapor deposition (CVD) expts. The large-scale reactive force field simulations with up to million atoms enable the direct simulation of Au NPs and Cu NPs with a size of 10 to 20 nm, which is consistent with the realistic catalysis using in the expt. condition. The simulated NPs well reproduce the expt. X-ray Powder Diffraction (XRD) patterns and transmission electron microscopy (TEM) images. These NPs all-atom models allow us to scan the active site on realistic catalysis surface. By using quantum mechanics based high-throughput screening, we have located the active surface site on Au NPs for CO_2 redn. to CO and the active sites on Cu NPs for CO redn. to C2. We find that: at Au NPs, the surface sites at twin boundary are responsible for promoting CO_2 redn . to CO; at Cu NPs, the unsatd. square site is responsible for C-C coupling. Armed with this at. information of active sites, we applied these mulita-scale high-throughput screening technologies to explore the alloy catalysis searching for better catalysis. We predict Au-Fe alloy NPs as a better CO_2 redn. catalysis than pure Au NPs, and exptl. confirm the superior performance by collaborating with expt. group. We believe the work-flow in this work will profoundly accelerate the searching for improved CO_2 redn. catalysis.

Additional details