TY - JOUR
T1 - Improved Estimates of the Critical Point Constants for Large n-Alkanes Using Gibbs Ensemble Monte Carlo Simulations
AU - Messerly, Richard A.
AU - Knotts, Thomas A.
AU - Rowley, Richard L.
AU - Wilding, W. Vincent
N1 - Publisher Copyright:
© 2016 American Chemical Society.
PY - 2016/10/13
Y1 - 2016/10/13
N2 - In this work, we present improved estimates of the critical temperature (Tc), critical density (ρc), critical pressure (Pc), and critical compressibility factor (Zc) for n-alkanes with chain lengths as large as C48. These are obtained for several different force field models with Gibbs ensemble Monte Carlo simulations. We implement a recently proposed data analysis method designed to reduce the uncertainty in Tc, ρc, Pc, and Zc when predicted with molecular simulation. The results show a large reduction in the uncertainties compared to the simulation literature with the greatest reduction found for ρc, Pc, and Zc. Previously, even the most computationally intensive molecular simulation studies have not been able to elucidate the n-alkane Pc trend with respect to larger carbon numbers. The results of this study are significant because the uncertainty in Pc is small enough to discern between conflicting experimental data sets and prediction models for large n-alkanes. Furthermore, the results for Tc resolve a discrepancy in the simulation literature with respect to the correct Tc trend for large n-alkanes. In addition, the Zc results are reliable enough to determine the most accurate prediction trend for Zc. Finally, finite-size effects are shown to not be significant even for the relatively small system sizes required for efficient simulation of longer chain lengths.
AB - In this work, we present improved estimates of the critical temperature (Tc), critical density (ρc), critical pressure (Pc), and critical compressibility factor (Zc) for n-alkanes with chain lengths as large as C48. These are obtained for several different force field models with Gibbs ensemble Monte Carlo simulations. We implement a recently proposed data analysis method designed to reduce the uncertainty in Tc, ρc, Pc, and Zc when predicted with molecular simulation. The results show a large reduction in the uncertainties compared to the simulation literature with the greatest reduction found for ρc, Pc, and Zc. Previously, even the most computationally intensive molecular simulation studies have not been able to elucidate the n-alkane Pc trend with respect to larger carbon numbers. The results of this study are significant because the uncertainty in Pc is small enough to discern between conflicting experimental data sets and prediction models for large n-alkanes. Furthermore, the results for Tc resolve a discrepancy in the simulation literature with respect to the correct Tc trend for large n-alkanes. In addition, the Zc results are reliable enough to determine the most accurate prediction trend for Zc. Finally, finite-size effects are shown to not be significant even for the relatively small system sizes required for efficient simulation of longer chain lengths.
UR - http://www.scopus.com/inward/record.url?scp=84991644008&partnerID=8YFLogxK
U2 - 10.1021/acs.jced.6b00574
DO - 10.1021/acs.jced.6b00574
M3 - Article
AN - SCOPUS:84991644008
SN - 0021-9568
VL - 61
SP - 3640
EP - 3649
JO - Journal of Chemical and Engineering Data
JF - Journal of Chemical and Engineering Data
IS - 10
ER -