(342bg) Development of Accurate Coarse-Grained Models of Polar Organic Solvents

Organic solvents are key components in a variety of hybrid material applications including polymer synthesis and self-assembly, nanoparticle film stabilization, colloid studies etc. Polar organic solvents like dimethyl acetamide (DMA) and dimethyl sulfoxide (DMSO) are majorly used for biochemical synthesis including polysaccharides and peptides. N-methyl-2-pyrrolidone (NMP) has been used in several fields of industrial applications including electronics, petroleum and pharmaceutical industries. Tetrahydrofuran (THF) is used as a solvent for surface coatings, adhesives as well as printing inks. It is also reported that solvent systems, consisting majorly of high boiling polar organic solvents like N,N-Diethylformamide (DEF) and 1,3-dimethyl-2-imidazolidinone (DMI), play an important role in MOF synthesis as well as in deciding the morphology of MOFs. Thus, in order to gain molecular insights into many of these applications using coarse-grained (CG) molecular dynamics (MD) simulations, it is necessary to have accurate solvent models, capable of replicating the process of MOF self-assembly. Here we have utilized an approach that integrates MD simulations with particle swarm optimization (PSO) to accelerate the development of polarizable and nonpolarizable CG models of six of the aforementioned polar organic solvents. Using a combination of beads previously parameterized in the group as well as a few newly defined CG beads with 2:1 or 3:1 mapping schemes, the bonded and non-bonded force-field (FF) parameters for all CG models were optimized to reproduce experimentally or atomistically obtained properties of density, enthalpy of vaporization, surface tension. The models were further tested at varying temperatures and system sizes to study their transferability. We also performed uncertainty quantification of parameters and properties using the Bayesian method.