PRISM theory molecular dynamics simulations

Close examination of the radii of ration from the molecular dynamirs simulations reveals that the chains are not completely ideal Overall the chains exhibit nearly ideal scaling behavior for whidi R oc Locally, however, the chains are found to be expanded relative to a freely jointed chain of the same length. This local expansion is a result of local intramolecular excluded volume which has not been completdy screened out in the melt. Thus in order to predict accurately the intermolecular structure one needs to correct for local deviations from Flory s ideality hypothesis. We were able to make this correction by employing the ra(k) directly computed from the molecular dynamics simulation. When the PRISM theory is then used to calculate g(r) from the actual, simulated d>(kX excellent agreement is seen in Fig. 3 between the theory and the simulation... 

An example of a comparison by Honnell et al. of PRISM theory with molecular dynamics simulations are shown in Figure 3. Details of the model are given elsewhere. Briefly, a meltlike density was studied for /V = 50-150 unit chains. The linear polymers were modelled as freely jointed beads with a purely repulsive, shifted Lennard-Jones interaction between all segment pairs. The corresponding chain aspect ratio is F-1.4. PRISM theory with the PY closure (plus a standard correction... 

An important question is whether PRISM theory can predict the packing in athermal blends with the same good accuracy found for one-component melts. To address this question Stevenson and co-workers performed molecular dynamics simulations on binary, repulsive force blends of 50 unit chains at a liquidlike packing fraction of -17 = 0.465. The monomeric interactions were very similar to earlier one-component melt simulations which served as benchmark tests of melt PRISM theory. Nonbonded pairs of sites (both on the same and different chains) were taken to interact via shifted, purely repulsive Lennard-Jones potentials. These repulsive potentials were adjusted so that the effective hard site diameters, obtained from Eq. (3.12), were 1-015 and = 1.215 for the chains of type A or B, respectively. Chain connectivity was maintained using an intramolecular FENE potential between bonded sites on the same chain. The resulting chain model has nearly constant bond lengths that are nearly equal to the effective hard-core site diameter. 

---