Assessing the Quality of Molecular Simulations for Vapor-Liquid Equilibria: An Analysis of the TraPPE Database

Becky L. Eggimann, Yangzesheng Sun, Robert F. Dejaco, Ramanish Singh, Muhammad Ahsan, Tyler R. Josephson, J. Ilja Siepmann

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


As molecular modeling and simulation techniques become increasingly important sources of thermophysical property and phase equilibrium data, the ability to assess the robustness of that data becomes more critical. Recently, the use of the compressibility factor (Z) has been suggested as a metric for testing the quality of simulation data for vapor-liquid equilibria (VLE). Here, we analyze predicted VLE data from the transferable potentials for phase equilibria (TraPPE) database and show that, apart from data entry or typographical errors, Z will always be well-behaved in Gibbs ensemble Monte Carlo (GEMC) simulations even when the simulations are not sufficiently equilibrated. However, this is not true for grand canonical Monte Carlo simulations. When the pressure is calculated from the internal forces, then pressure and density are strongly correlated for the vapor phase and, for GEMC simulations, it is recommended to treat Z as an instantaneous mechanical property. From analysis of the TraPPE VLE data, we propose a complementary metric based on the predicted vapor pressures at three neighboring temperatures and their deviation from a local Clausius-Clapeyron fit.

Original languageEnglish (US)
Pages (from-to)1330-1344
Number of pages15
JournalJournal of Chemical and Engineering Data
Issue number3
StatePublished - Mar 12 2020

Bibliographical note

Funding Information:
This research is primarily supported through the National Science Foundation (CBET-1835067) and Wheaton College. The work on correlations between properties and histogram analysis is supported by the Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences and Biosciences, under Award DE-FG02-17ER16362. The development of the TraPPE database and website was supported by the National Science Foundation (RUI-1159731). Computational resources from the Minnesota Supercomputing Institute are also gratefully acknowledged. Several validation simulations were performed by Siepmann group members not included as authors (and not previously published): C. Bunner (ethanal, dimethyl sulfide, and n -perfluoropentane), J. Chen (2-butanethiol), E. Fetisov (2-methylpropene), D. Harwood (oxygen and 2,2-dimethylpropane), B. Liang (2-butanol), M. Shah (carbon dioxide), K. Struksheats (acetone, 2-methoxy-2-methylpropane, naphthalene, and 2-methyl-1,3-butadiene), and B. Xue (data for Table 4).

Fingerprint Dive into the research topics of 'Assessing the Quality of Molecular Simulations for Vapor-Liquid Equilibria: An Analysis of the TraPPE Database'. Together they form a unique fingerprint.

Cite this