(193x) Incorporating Information from MD Simulations into COSMO-RS Predictions for Polymers

COSMO-RS¹ is a statistical thermodynamics model that has garnered much interest for its use of QM-derived electronic/structural information and its extremely fast calculation times after the DFT step, making it an attractive tool for large-scale compound screening. To date, COSMO-RS has been used widely for predicting the thermodynamic properties of fluids, particularly those which consist of small molecules. Though there has been some recent success in extending COSMO-RS methods to systems with larger molecules/structures^2,3, current polymer modeling approaches³ with COSMO-RStypically rely on using a linear combination of electronic information from the polymer’s monomeric subunits. This approach is successful for many applications but is limited by the neglect of connectivity information in the COSMO-RS step, which has implications for capturing macromolecular structural/steric effects.

In this work, we present a method to generate weighting terms for the surface interactions used in COSMO-RS. Short-timescale MD simulations are used to provide surface interaction distribution functions for the polymeric system of interest, and these distributions are incorporated into the statistical thermodynamics equations of COSMO-RS. In this way, we are able to capture the approximate macromolecular behavior of polymeric molecules and provide more information about the probable surface-surface interactions of such a sterically-constrained macromolecular system. We perform experiments with several variables in our MD simulations: (1) comparing different lengths of polymer chains in the approximate system, (2) using different atoms in the polymer chains to build distribution functions, and (3) altering the timescale of the simulations. We finally discuss a few cases where this method demonstrates more accurate property prediction (solubility parameter, activity coefficients, etc.) than the current approach.

1. Klamt A, Jonas V, Bürger T, Lohrenz JC. Refinement and parametrization of COSMO-RS. The Journal of Physical Chemistry A. 1998 Jun 25;102(26):5074-85.
2. Klamt A, Huniar U, Spycher S, Keldenich J. COSMOmic: a mechanistic approach to the calculation of membrane− water partition coefficients and internal distributions within membranes and micelles. The Journal of Physical Chemistry B. 2008 Aug 28;112(38):12148-57.
3. Loschen C, Klamt A. Prediction of solubilities and partition coefficients in polymers using COSMO-RS. Industrial & Engineering Chemistry Research. 2014 Jul 7;53(28):11478-87.