Abstract: The rate, selectivity and efficiency of plasma-based conversion processes is strongly affected by nonequilibrium phenomena. High concentrations of vibrationally excited molecules are such a plasma-induced effect. It is frequently assumed that vibrationally excited molecules are important in plasma catalysis because their presence lowers the apparent activation energy of dissociative chemisorption reactions and thus increases the conversion rate. A detailed atomic-level understanding of vibrationally stimulated catalytic reactions in the context of plasma catalysis is however lacking. Here, we couple a recently developed statistical model of a plasma-induced vibrational nonequilibrium to molecular dynamics simulations, enhanced sampling methods, and machine learning techniques. We quantify the impact of a vibrational nonequilibrium on the dissociative chemisorption barrier of H2 and CH4 on nickel catalysts over a wide range of vibrational temperatures. We investigate the effect of surface structure and compare the role of different vibrational modes of methane in the dissociation process. For low vibrational temperatures, very high vibrational efficacies are found, and energy in bend vibrations appears to dominate the dissociation of methane. The relative impact of vibrational nonequilibrium is much higher on terrace sites than on surface steps. We then show how our simulations can help to interpret recent experimental results, and suggest new paths to a better understanding of plasma catalysis.

Abstract: A number of recent computational material design studies based on density functional theory (DFT) calculations have put forward a new class of materials with electrically switchable chemical characteristics that can be exploited in the development of tunable gas storage and electrocatalytic applications. We find systematic flaws in almost every computational study of gas adsorption on polarized or charged surfaces, stemming from an improper and unreproducible treatment of periodicity, leading to very large errors of up to 3 eV in some cases. Two simple corrective procedures that lead to consistent results are proposed, constituting a crucial course correction to the research in the field.

Abstract: Optimization of catalytic materials for a given application is greatly constrained by linear scaling relations. Recently, however, it has been demonstrated that it is possible to reversibly modulate the chemisorption of molecules on nanomaterials by charging (i.e., injection or removal of electrons) and hence reversibly and selectively modify catalytic activity beyond structure−activity correlations. The fundamental physical relation between the properties of the material, the charging process, and the chemisorption energy, however, remains unclear, and a systematic exploration and optimization of charge-switchable sorbent materials is not yet possible. Using hybrid DFT calculations of CO2 chemisorption on hexagonal boron nitride nanosheets with several types of defects and dopants, we here reveal the existence of fundamental correlations between the electron affinity of a material and charge-induced chemisorption, show how defect engineering can be used to modulate the strength and efficiency of the adsorption process, and demonstrate that excess electrons stabilize many topological defects. We then show how these insights could be exploited in the development of new electrocatalytic materials and the synthesis of doped nanomaterials. Moreover, we demonstrate that calculated chemical properties of charged materials are highly sensitive to the employed computational methodology because of the self-interaction error, which underlines the theoretical challenge posed by such systems.

Abstract: Atomistic simulations can in principle provide an unbiased description of all mechanisms, intermediates, and products of complex chemical processes. However, due to the severe time scale limitation of conventional simulation techniques, unrealistically high simulation temperatures are usually applied, which are a poor approximation of most practically relevant low-temperature applications. In this work, we demonstrate the direct observation at the atomic scale of the pyrolysis and oxidation of n-dodecane at temperatures as low as 700 K through the use of a novel simulation technique, collective variable-driven hyperdynamics (CVHD). A simulated timescale of up to 39 seconds is reached. Product compositions and dominant mechanisms are found to be strongly temperature-dependent, and are consistent with experiments and kinetic models. These simulations provide a first atomic-level look at the full dynamics of the complicated fuel combustion process at industrially relevant temperatures and time scales, unattainable by conventional molecular dynamics simulations.