Application of molecular dynamics calculations in the prediction of dynamical molecular properties

Adri C.T. Van Duin, Steve R. Larter

Research output: Contribution to journalConference articlepeer-review

16 Scopus citations


Knowledge of dynamical molecular properties like partition coefficients and diffusion constants are of vital importance for a reliable description of the distribution of organic material in the subsurface. However, such data, and especially their temperature and pressure dependence, are sometimes difficult to obtain by experimental means under subsurface conditions. Molecular dynamics, a computational technique aiming to describe the time-dependent movement of molecules, may provide an interesting alternative method to reliably estimate dynamical properties. To test its applicability to a geochemical problem molecular dynamics simulations were used to predict the oil/water partition coefficients (K(o/w)) of several phenol and carbazole compounds. A limited molecular dynamics simulation indeed managed to properly predict the qualitative differences in oil/water partition coefficients between these compounds (K(o/w)(phenol) < K(o/w)(o- and m-cresol) < K(o/w)(carbazole) < K(o/w)(benzo[a] and benzo[c]carbazole). An extended molecular dynamics simulation was performed to predict the relative K(o/w) of benzo[a] and benzo[c]carbazole. The results from this simulation suggests that benzo[c]carbazole has a somewhat higher affinity for the hydrocarbon phase than its benzo[a]isomer, confirming measured solubilities of these compounds in hydrocarbon solvents.

Original languageEnglish (US)
Pages (from-to)1043-1050
Number of pages8
JournalOrganic Geochemistry
Issue number5-7 -7 pt 2
StatePublished - Nov 1998
EventProceedings of the 1997 18th International Meeting on Organic Geochemistry. Part 2 (of 2) - Maastricht, Neth
Duration: Sep 22 1997Sep 26 1997

All Science Journal Classification (ASJC) codes

  • Geochemistry and Petrology


Dive into the research topics of 'Application of molecular dynamics calculations in the prediction of dynamical molecular properties'. Together they form a unique fingerprint.

Cite this