Author(s): Yan T, Wang Y, Knox C, Yan T, Wang Y, Knox C, Yan T, Wang Y, Knox C
Abstract Share this page
Abstract An electronically polarizable model, based on the AMBER nonpolarizable model, has been developed for the ionic liquid (IL) 1-ethyl-3-methyl-imidazolium nitrate (EMIM(+)/NO(3)(-)). Molecular dynamics simulation studies were then performed with both the polarizable and nonpolarizable models. These studies suggest EMIM(+) cations have a strong tendency to pack with their neighboring imidazolium rings nearly parallel to each other, bridged by hydrogen bonds to NO(3)(-) anions. Polarization has two key effects, (1) additional charge-dipole and dipole-dipole interactions enhance short-range electrostatic interactions and (2) screening reduces long-range electrostatic interactions. As a result, the polarizable model exhibited enhanced hydrogen bonding compared to the nonpolarizable model, while the latter retained more ordered long-range spatial correlations than the former. Though EMIM(+) has a very short nonpolar ethyl tail group, spatial heterogeneity, previously observed with long-chain ILs, was observed in this system and has been quantified using the heterogeneity order parameter. The polarizable model was slightly more heterogeneous than the nonpolarizable model. The enhanced spatial heterogeneity of the polarizable model is again attributed to the stronger short-range electrostatic interactions, which "push" the nonpolar tails away from the polar heads, leading to more aggregation and a strongly altered ionic packing pattern around NO(3)(-) as observed by a different anion-anion center-of-mass partial radial distribution function g(--) (r). Interestingly, both models seemed to "remember" the crystal structure even at temperatures significantly higher (approximately 90 K higher) than the melting point (311 K). Along with the results on the dynamical properties reported in the accompanying paper, the current study demonstrates that electronic polarizability is significant in ionic liquid systems.
This article was published in J Phys Chem B
and referenced in Journal of Data Mining in Genomics & Proteomics