472231 Computationally Driven Discovery of Novel Materials for Separation and Catalysis

Sunday, November 13, 2016
Continental 4 & 5 (Hilton San Francisco Union Square)
Peng Bai, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN

Research Interests:  Theory and molecular modeling have seen tremendous growth in recent years; capitalizing on theoretical and algorithmic developments and advancements in computer hardware, applications of computational methods have begun to penetrate into all fields of the chemical sciences. In this poster presentation, I will highlight separation and catalysis applications using crystalline nanoporous materials, primarily zeolites, and show how molecular modeling can be used both to gain a microscopic understanding into phenomena difficult to probe experimentally and to predict accurate data at a scale beyond the reach of the traditional trial-and-error approach. I developed a new, transferable force field for all-silica zeolites, TraPPE-zeo, that is applicable to any type of sorbate molecule and framework structure. With the improved intermolecular potentials as well as better methodologies, it was then possible to study several complex systems involving articulated molecules and liquid mixtures, such as the multi-component adsorption of aqueous alcohol, polyol, glucose, and furfural solutions. These studies demonstrate the important effects of hydrogen-bonding, which render both the bulk and adsorbed phases highly non-ideal and are for certain molecules so strong that adsorption becomes entropically rather than enthalpically driven. In another application, the unusual adsorption and diffusion behaviors of a hierarchical zeolite were investigated, and our simulations clearly indicate regimes where the introduction of mesoporous “highways” impedes, rather than enhances, molecular transport in a material without structural defects, and provide insights into the nature of surface resistance and its roles in the practical application of hierarchical materials and ultrathin membranes. More recently, the effect of reactant chain lengths and architectures and the degrees of framework confinement on the reactivity and selectivity of zeolites as solid acid catalysts were carefully examined using first-principles calculations. Information obtained from this work provide the basis for an effort in which a large number of materials were screened with the purpose of finding better candidates for two energy-related applications, ethanol/water separation and hydrocarbon iso-dewaxing, and some of our predictions were subsequently verified by experiments. This strategy was recently extended to account for transition-state selectivity [1], and several advancements were made to allow for highly parallel ab-initio Monte Carlo simulations, which together bring closer the possibility of using theory and computation in a predictive mode to expedite the pace of materials discovery for complicated, realistic, and industrially relevant applications.

Teaching Interests: Thermodynamics and statistical mechanics, catalysis and chemical kinetics, heat and mass transfer, numerical methods and linear algebra, computational chemistry

[1] “A method to screen nanoporous catalysts for transition-state selectivity,” to be presented at 2016 AIChE Annual Meeting, Symposium on ‘Catalysis with Microporous and Mesoporous Materials,’ San Francisco, CA.

Extended Abstract: File Not Uploaded