Molecular Simulation and Macroscopic Modeling of Thermodynamic and Transport Properties of Silicon-Containing Rubbery Polymer – Solvent Mixtures
Advancing the chemical engineering fundamentals
Thermodyanmics: Molecular Simulation & Related Approaches (T2-1d)
Keywords: Molecular Simulation, Molecular Thermodynamics, Polymers, Diffusion, Solubility
The accurate knowledge of thermodynamic and transport properties of polymer – solvent mixtures is highly important for novel material design, i.e. polymer membranes, and optimum process design, i.e. polymerization processes. Reliable experimental data over a wide range of conditions are often limited and expensive. Calculation of physical properties based on suitable theoretical models is, therefore, desirable. In this work, a family of relatively unexplored silicon-containing rubbery polymers, namely poly(sila-alkanes) are examined. These polymers are rooted to poly(dimethylsiloxane) by substituting carbon for oxygen. These polymers have very low Tg and were shown to exhibit promising membrane material properties for hydrocarbon separation.
Molecular simulation using an atomistic force-field that accounts explicitly for bonded and non-bonded intra- and inter-molecular interactions developed based on ab initio calculations for model oligomers is used for the elucidation of microscopic polymer melt structure. Accurate calculation of molecular mechanisms and microscopic structure permits reliable prediction of macroscopic physical properties. Molecular Dynamics calculations are shown to be in excellent agreement with experimental data for three homopolymes, including PDMS, and a co-polymer. Subsequently, solubility of various organic molecules (n-alkanes and n-perfluoroalkanes) and gases (noble gases, oxygen and nitrogen) is calculated using Widom’s test particle method. In parallel, long Molecular Dynamics runs, on the order of several hundred ns, are used for the reliable estimation of solute diffusion coefficients at ambient and high temperatures. In all cases, simulation predictions agree well with limited experimental data available.
Despite its accuracy, molecular simulation substantial computational needs make it “too heavy” for engineering calculations and call for less detailed but equally accurate macroscopic models. In this respect, calculations using the group-contribution Perturbed Chain-Statistical Associating Fluid Theory are presented for polymer – solvent isotherms. A single binary interaction parameter is used to tune the model to experimental data. Model correlations are in good agreement with experimental data, thus confirming that the model can be used reliably for such calculations.
Preferred type of presentation: Oral
Presented Tuesday 18, 12:00 to 12:20, in session Thermodyanmics: Molecular Simulation & Related Approaches (T2-1d).
