(521g) Versatile Hybrid Particle-Field Approach for Simulating Inhomogeneous Polymeric Systems
AIChE Annual Meeting
Wednesday, October 31, 2018 - 2:15pm to 2:30pm
Inhomogeneous polymeric systems are characterized by presence of multiple chemically distinct components. In such mixtures, there are often situations when one of the components is dominant in composition, examples including dilute polymer solutions, ion-doped block copolymers, and associating polymers, etc. Albeit their small fractions, minor components play a significant role in altering thermodynamic as well as dynamical properties of the system through interacting with major components. Computer simulations of such systems using conventional particle-based approach are often plagued by low computational efficiency. This is because particle-based methods evaluate molecular interactions in the pair-wise manner. Due to disparity in number fractions, computation time is taken up mostly by resolving interactions among molecules of the major component. For example, in simulating dilute polymer solutions, the implicit solvent model is often adopted in order to improve computational efficiency. However, with solvent molecules not being explicitly considered, this strategy is not able to capture solvation structures, and consequentially is not capable of accounting for complex solvation effects such as competition between hydrogen-bond forming and hydrophobic interactions. At the other end of the spectrum, the field-theoretical approach had been extensively developed for studying polymeric systems. A distinct signature of the approach is that pair-wise interactions are replaced by interactions between effective fields. Computational cost can thus be greatly reduced by assuming the effective fields being static, i.e., the mean-field approximation. While the mean-field treatment works well with dense systems, its validness deteriorates for systems with the minor components due to increased compositional fluctuations. Situation becomes even worse if minor components carry strongly interacting sites, such as electric charges and associating groups, when correlation effects need to be accounted for accurately. Toward addressing this challenge, in this work we develop a coarse grained hybrid particle-field approach for simulating inhomogeneous polymeric systems with composition disparities. In the hybrid approach, interactions involving molecules of the minor components are treated explicitly in pair-wise manner, and the ``quasi-instantaneous'' fields are employed to account for interactions among molecules of the major component. By illustrating its applications using several examples, we show that the hybrid approach is versatile in algorithm design, can achieve accuracy comparable to full particle-based simulations at a fraction of computational cost. Comparing to the full field-based method, the hybrid approach resolve microscopic local structures faithfully, and can be readily extended to incorporate more complex interaction models such as hydrogen-bond and direction-specific interactions, etc.