Abstract
Recent developments in dispersion corrected and nonlocal density functionals are aimed at accurately capturing dispersion interactions, a key shortcoming of local and semilocal approximations of density functional theory. These functionals have shown significant promise for dimers and small clusters of molecules as well as crystalline materials. However, their efficacy for predicting vapor liquid equilibria is largely unexplored. In this work, we examine the accuracy of dispersion-corrected and nonlocal van der Waals functionals by computing the vapor liquid coexistence curves (VLCCs) of hydrofluoromethanes. Our results indicate that the PBE-D3 functional performs significantly better in predicting saturated liquid densities than the rVV10 functional. With the PBE-D3 functional, we also find that as the number of fluorine atoms increase in the molecule, the accuracy of saturated liquid density prediction improves as well. All the functionals significantly underpredict the saturated vapor densities, which also result in an underprediction of saturated vapor pressure of all compounds. Despite the differences in the bulk liquid densities, the local microstructures of the liquid CFH3 and CF2H2 are relatively insensitive to the density functional employed. For CF3H, however, rVV10 predicts slightly more structured liquid than the PBE-D3 functional.
| Original language | English |
|---|---|
| Pages (from-to) | 3295-3304 |
| Number of pages | 10 |
| Journal | Journal of Chemical Theory and Computation |
| Volume | 12 |
| Issue number | 7 |
| DOIs | |
| State | Published - Jul 12 2016 |
| Externally published | Yes |
Funding
This work is funded by the National Science Foundation (NSF) under grant number CHE - 1265872. This work used the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation grant number OCI-1053575. XSEDE provided the computing support from the Texas Advanced Computing Center (TACC) at The University of Texas at Austin and University of Tennessee and Oak Ridge National Laboratory's Joint Institute for Computational Sciences. This research also used resources of the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility supported by the Office of Science of the U.S. Department of Energy under contract number No. DE-AC02-05CH11231.