First-principles calculations of polymer interactions

Sammanfattning: This thesis reports on studies of interactions in sparse matter by first-principles calculations, in particular polymeric systems. The focus is on the three unbranched simple polymers: polyethylene (PE), isotactic polypropylene (PP) and isotactic polyvinylchloride (PVC), which together represent an important class of materials that form complexes stabilized by weak but long-range dispersive interactions. The latter ones are not included in traditional density functional theory (DFT). Due to the missing nonlocal description traditional implementations of DFT predicts, e.g., the PEcrystal to be unstable, contradicting both experiments and intuition. Two schemes are applied which extend the applicability of DFT to such sparse systems: 1. A systematic correction scheme that applies to parallel well-separated polymers is proposed. From the length-averaged electron densities of the polymers and their static polarizabilities, both calculated with the traditional DFT, the dynamic response and in turn the asymptotic dispersive interaction of the polymers are modeled. Simple expressions for the orientation dependent polymerpolymer interaction energy are obtained, and even simpler expressions are found by enforcing the polymers to be cylindrically symmetric. Explicit results are given for PE, PP, PVC. 2. The nonlocal correlation energy for pairs of PE-molecules is calculated also for short and intermediate separations using the recently developed general geometry (gg) DFT scheme [Phys. Rev. Lett. 92, 246401, 2004]. The gg-scheme models the electrodynamic response on the basis of the electron density only and applies also at binding distance and out. This allows us to calculate the cohesive energy landscape for the PE crystal, showing promising agreement with experimental values for equilibrium lattice constants. For well separated PE-chains (center-to-center distance > 8 Å) the two approaches turn out to be consistent with each other.