Consistent acetonitrile molecular models for both standard and computationally efficient molecular dynamics studies

Abstract

Acetonitrile is widely utilized in scientific research, presenting an ideal solvent media for a large number of organic reactions. Its use at the industrial scale ranges from the production of molecules of pharmaceutical interest to photographic films. Additionally, certain enzyme based catalytic processes show great functionality in Acetonitrile media. Furthermore, numerous enzymes continue to act as efficient biocatalyzers in acetonitrile solution, showing in some cases significant changes in their original specificity and selectivity. Consequently, the study of the behavior of such proteins in this solvent by means of potent computational methods such as Molecular Dynamics results of great interest. Many molecular models for Acetonitrile have been developed for use in Molecular Dynamic studies. Nevertheless, all acetonitrile models developed up to date are only capable of performing reasonably when used with integration time-steps no greater than 2 femtoseconds (fs). We present two molecular models for acetonitrile which perform both efficiently and reliably with integration time steps of up to 4 fs. Furthermore, the optimization procedure used has enabled to achieve this performance improvement at no cost as regards the agreement between the experimental macroscopic data for Acetonitrile and the corresponding properties evaluated for the models here presented.