Chris RzepaStudent: Christopher Rzepa

Project: Towards Data-Driven Structure-Property Relations for Predicting Adsorption Entropy in Siliceous Zeolites   

View: Research Poster (PDF) | Presentation (YouTube)

Department: Chemical and Biomolecular Engineering

Advisor: Srinivas Rangarajan, Jeetain Mittal

Abstract

A thorough understanding behind shape selective zeolite catalysis and its relationship to entropy remains elusive1. Furthermore, the interaction of both the adsorbent pore structure and adsorbate molecular structure on entropic transfer quantities is largely unexplored and, consequently, significant opportunities exist for the development of systems or correlations for relating the adsorption-catalysis-thermodynamics to fundamental and more easily measured physical properties. Previous works using experimental data have demonstrated simple linear scaling relationships between the gas- and adsorbed-phase entropies2,3, allowing for sensible predictions based on more-easily obtained physical parameters. However, the resultant conclusions were based on experimental data for a relatively small subset of industrially relevant alkanes and a few alcohols.

Motivated by these results in the literature, here we computationally explore the generality of such relationships across multiple classes of molecules and frameworks. We model our adsorption systems (i.e., the adsorbates and adsorbents) by implementing TraPPE forcefields4 and quantify the adsorbate entropy by performing canonical Monte-Carlo integrations (cf. Widom test particle method)5 using the FEASST6 molecular simulation package. Our dataset was composed of thirty-seven adsorbates, across ten functional categories, adsorbed within five siliceous zeolite structures. Our results show that simple linear correlations between the gas- and adsorbed-phase entropies, to a good first approximation, continued to exist for our larger and more diverse set of molecules.7 Moreover, we have found that each correlation was largely dependent on the zeolite’s size, characterized by physical descriptors such as the largest cavity diameter and occupiable volume, indicating that the adsorbate entropy may be predicted using such metrics. To further elicit such “structure-topology-thermochemistry” relations, we have expanded our dataset to include over eight thousand combinations between molecules and zeolites, each consisting of twenty-four unique topological features for zeolites and eleven unique molecular descriptors. Ultimately, our effort is to develop an interpretable data-driven model, which, given framework and molecular descriptors of the system, may be used to predict the adsorbate entropies of novel systems.

References

  1. Smit, B., & Maesen, T. L. M. (2008). Towards a molecular understanding of shape selectivity. Nature, 451, 671.
  2. Campbell, C. T.; Sellers, J. R. V. The entropies of adsorbed molecules. J. Am. Chem. Soc. 2012, 134 (43), 18109−18115.
  3. P. J. Dauenhauer, O. A. Abdelrahman, ACS Central Science 2018 4 (9), 1235-1243
  4. M. G. Martin and; J. Ilja Siepmann, J. Physical Chemistry B 1998 102 (14), 2569-2577
  5. Widom, B. Some Topics in the Theory of Fluids.J. Chem. Phys.1963,39, 2808–2812.
  6. Hatch, H. W., Mahynski, N. A., and Shen, V. K. (2018) FEASST: Free Energy and Advanced Sampling Simulation Toolkit. J. Res. Natl Inst Stan, 123, 123004.
  7. Rzepa, C., Siderius, W. D., Hatch, H. W., Shen, V. K., Rangarajan, S., and Mittal, J., J. Physical Chemistry C 2020 124 (30), 16350-16361

About Christopher Rzepa

Christopher Rzepa is a 4th year Graduate student at Lehigh university within the department of Chemical & Bimolecular Engineering. He is advised by Professor Srinivas Rangarajan and co-advised by Professor Jeetain Mittal. He earned his B.S. in Chemical Engineering (Minor: Math) at Penn State University-University Park. Chris uses computational techniques to study two categories of problems: (i) elucidating the mechanism of catalytic reactions, and (ii) shape selectivity from the confinement effects of organic and inorganic materials. In particular, he uses a combination of density functional theory, grand canonical transition matrix Monte-Carlo, and mean field microkinetic modeling to rationally design confinement in porous catalysts.