Molecular simulation of transport in liquids and polymers

Computer simulations of complex multi- particle systems have attracted more and more research interest. Molecular dynamics (MD) simulations have been used intensively in various scientific fields such as molecular biology, polymer physics, nanotechnology and many others. System properties measured at a certain time can be deduced from the coordinates and velocities of classical particles. If the interatomic forces are known with a good accuracy and the initial conditions of the system can be defined properly, molecular dynamics simulation can act as a computer simulation. It means that these results can be compared to experimentally obtained values and, more importantly, some other information about the system can be accessed, which sometimes is hard or impossible to measure. After a short overview on MD methods, several MD simulations will be presented. Thermal conductivity of polymer crystals is a typical quantity that is difficult to experimentally determine. This is because samples of large-enough single crystals of polymers for thermal conductivity measurements have not yet been prepared, therefore the single crystal properties can only be determined via computer simulation. In Chapter 3 we have summarized extensive calculations of the thermal conductivity of the δ -phase of syndiotactic polystyrene (sPS). Until now, only partial theoretical data dealing with thermal conductivity of crystalline polymers was available. This available data was particularly concerned with the correlation between thermal conductivity and the polymer’s morphology and orientation [D. Hansen and G. A. Bernier, Polym. Eng. Sci. 12 (3), 204 (1972)]. In comparison with the amorphous structure of polymer a large anisotropy can be established in crystalline polymer as result of varied structural and morphological parameters in different directions. MD simulations permit us, for example, to restrict some oscillations and to set the bond length between two atoms, which can be done by addition of constraints in the system. Such artificial constraints limit the free movement of the particles which decreases the degrees of freedom of the system. In this study we investigated the sensitivity of the thermal conductivity to different numbers and locations of such constraints in different parts of IX the polymeric chains. It was found that the thermal conductivity has a tendency to decrease when the number of active degrees of freedom in the system is reduced by the introduction of stiff bonds. This dependence is, however, weaker and more erratic than previously found for molecular liquids and amorphous polymers [E. Lussetti, T. Terao, and F. Müller-Plathe, J, of Phys, Chem, B 111 (39), 11516 (2007)]. Another physical property of polymers, which has attracted a great deal of attention from researchers in the recent times, is the understanding of the dynamic and static properties of polymer chains. Many technologies such as electronics packaging, coatings, adhesion, and composite materials are based on these polymeric properties. In Chapter 4 we discussed the physical properties of short polyvinyl-alcohol (PVA) oligomers up to a chain length of ten monomers chain (H(-CH2-CH(OH)-)NCH3). The specific volume was found to depend linearly on the inverse number of repeat units N, a result that is in agreement with experimental findings for other polymers. The gyration radius was found to depend on the number of formula units via N0.65±0.03 . The exponent simulated is somewhat larger than the known N0.588 dependence for long chains in good solvents. We also discuss the orientation correlation function for different bonds in the chain. The relaxation times for these bond vectors, as obtained via the Kohlrausch- Williams-Watt expression, showed an exponential dependence on the number of repeat units. In Chapter 3 we studied the thermal conductivity of crystal polymer but under certain conditions and as a response to a temperature gradient, it was possible to correlate the separation between different chemical species. This effect is called the Soret effect or thermal diffusion effect and is quantified by the Soret coefficient (S-T). Although this effect has been studied for more than 150 years, a microscopic understanding of thermal diffusion processes in liquids is still unavailable. The precise prediction of S-T from theory and simulations and even the experimental determination for more complex systems is often a challenge. In Chapter 5, we studied the thermal diffusion behavior of an equimolar mixture of hydrocarbon chains in xylene. Hydrocarbon chains (alkanes and alkenes) with the same carbon number were considered in order to exclude the mass contribution and to investigate the influence of molecular structure on the Soret coefficient. Thermal diffusion behavior was analyzed in terms of static and dynamic properties of the mixtures and an explanation for the observed results has been supplied. Chapter 6 finally summarizes the main conclusions of the present study in the thesis and provides summary of the work