(192al) Computational Discovery of New Materials and Processes for Industrial Separations

Shah, M. S., University of Minnesota
Tsapatsis, M., University of Minnesota
Siepmann, J. I., University of Minnesota
Chemicals separations are responsible for nearly half of the US industrial energy consumption. Our goal is to contribute towards developing the next-generation technologies for cost-effective and energy-efficient separations. Two important industrial separations – sour gas sweetening and ethane/ethylene separation – are discussed here.

We use Monte Carlo (MC) simulations for predicting phase and adsorption equilibria. Accurate force fields are at the heart of any reliable computational prediction. These molecular models comprise of several parameters, that may be system-specific for matching to experimental data. However, this work takes a different approach, in that, a large amount of effort is first invested in a multi-dimensional, parallel, and physically guided parameter optimization for capturing a wide range of properties for the molecules of interest. The expense involved in parametrizing transferable force fields is recouped by allowing the use of these models for diverse systems, without the need for reoptimization each time a new system is investigated. This also provides truly predictive and reliable answers to problems where experimental data may not be available.

MC simulations are used to probe adsorption in zeolites – nanoporous materials with highly specific pore sizes and shapes. Using tailored simulation protocols, a large database of zeolite structures is screened for adsorption performance and promising structures for efficient separations are discovered. Some of these zeolite frameworks are currently synthesized and assessed experimentally to validate the predictions. The new separation technologies that we are developing have the potential for far-reaching impacts on the global energy and chemical industries.