First-principles based micro-kinetic modeling for catalysts design

Abstract

Efficient and selective catalysis lies at the heart of many chemical reactions, enabling the synthesis of chemicals and fuels with enormous societal and technological impact. A fundamental understanding of intrinsic catalyst properties for effective manipulation of the reactivity and selectivity of industrial catalysts is essential to select proper catalysts to catalyze the reactions we want and hinder the reactions we do not want. The progress in density functional theory (DFT) makes it possible to describe interfacial catalytic reactions and predict catalytic activities from one catalyst to another. In this study, water-gas shift reaction (WGSR) was used as a model reaction. First-principles based micro-kinetic modeling has been performed to deeply understand interactions between competing reaction mechanisms, and the relationship with various factors such as catalyst materials, structures, promoters, and interactions between intermediates (e.g., CO self-interaction) that govern the observed catalytic behaviors. Overall, in this thesis, all relevant reaction mechanisms in the model reaction on well-defined active sites were developed with first-principles calculations. With the established mechanism, the promotional effect of K adatom on Ni(111) on WGSR compared to the competing methanation was understood. Moreover, the WGSR kinetic trend, with the hydrogen production rate decreasing with increasing Ni particle diameters (due to the decreasing fractions of low-coordinated surface Ni site), was reproduced conveniently from micro-kinetic modeling techniques. Empirical correlations such as Brønsted-Evans-Polanyi (BEP) relationship for O-H, and C-O bond formation or cleavage on Ni(111), Ni(100), and Ni(211) were incorporated to accelerate computational analysis and generate trends on other transition metals (e.g., Cu, Au, Pt). To improve the numerical quality of micro-kinetic modeling, later interactions of main surface reaction intermediates were proven to be critical and incorporated successfully into the kinetic models. Finally, evidence of support playing a role in the enhancement of catalyst activity and the impact on future modeling will be discussed. DFT will be a powerful tool for understanding and even predicting catalyst performance and is shaping our approach to catalysis research. Such molecular-level information obtained from computational methods will undoubtedly guide the design of new catalyst materials with high precision.