Multiscale modeling of polymer solvent mixtures
Final Report Abstract
While this project did not solve all the problems we set out to solve, we think that it still produced scientifically interesting results. On the particle-based side (Mainz-Halle group), we established a welldefined and consistent coarse-graining procedure for 1,4-polybutadiene, clarifying especially the role on non-bonded interactions in the mapping from the atomistic to the coarse-grained scale as well as developing a way to obtain them from equation of state information for the polymer melt. Concerning the study of semi-dilute solutions of 1,4-polybutadiene in its own monomer, only first feasibility studies could be carried out, before the PhD student left the project prematurely. These, however, revealed that it is a non-trivial task to obtain the solution phase diagram for reasonable chain lengths even for a coarse-grained modeling. Furthermore, we have explored a coarse-grained modeling approach that has become very popular in the last years, the representation of Gaussian chains in a polymer melt (or random coils in a semi-dilute solution) by soft spheres or soft ellipsoids. Such a strongly coarse-grained picture is very helpful when one tries to model nanocomposites because embedded nanoparticles and polymer chains can then be modeled on the same length scales. We have shown how an effective potential of an apolar polymer interacting with a solid surface can look like and what its inherent length and energy scales would be. On the field-based side (Bielefeld-Mainz group), we explored different ways of including thermal fluctuations in field-based simulations of polymer solutions, in particular the use of the Complex Langevin simulation method. Unfortunately, we found that this method is restricted to a subclass of models - those where the local equation of state is only quadratic in the density - and that it tends to become unstable for polymer solutions with arbitrary equation of state. This prompted us to develop a hybrid particle-field method which might solve the problem. First results are promising. While our individual groups have thus clearly made progress, we have not succeeded in bringing the different levels of coarse-grained modeling together and to establish a joint multiscale picture. Even after the end of the funding, this goal is still far ahead of us.
Publications
- Phys. Chem. Chem. Phys. 11, 1942 (2009)
T. Strauch, L. Yelash, W. Paul
