We have developed and validated a new force field for cations in zeolites, which explicitly distinguishes Si and Al atoms, as well as different types of oxygens in the framework. Our new force field gives excellent agreement with experimental data on cation positions, site occupancies and vibrational frequencies. Energy minimizations show that Na cations in site I are not at the centers of hexagonal prisms, but rather are in one of two symmetric S I sites displaced by about 0.6 Å along the [111] direction. Molecular dynamics (MD) simulations show that most cations are immobile in Na-X and Na-Y on the MD time scale, even at 1000 K. Only Na-X cations in site III′ exhibit diffusive motion at 1000 K, with a self-diffusivity from MD of 3.6 × 10 -10 m 2 s -1 . The MD simulations also show that cation movement is highly correlated, composed of jumps involving at least 4 cations or more.

No abstract

We have developed and applied a new force field for simultaneously modeling the dynamics of hydrofluorocarbons (HFCs) and exchangeable Na cations in faujasite-type zeolites. Our aim is to account for (i) the zeolite's capacity of separating HFC isomers, (ii) the experimentally observed unusual cation migration in Na−Y from the β-cages into the supercages upon the adsorption of HFCs, and (iii) the abnormal trans/gauche ratio in these systems. Energy minimizations and molecular dynamics simulations performed with this force field give excellent agreement with experimental data on heats of adsorption, guest−host distances, infrared spectra, and conformer ratios for different coverages of HFC-134 (CF2H−CF2H) and HFC-134a (CH2F−CF3) in Na−X (Si:Al = 1.2) and Na−Y (Si:Al = 2.4). The force field also accounts partially for the observed cation migration at intermediate loadings and fully at high loadings. The extent of migration is found to be influenced by the competition among Na−O, Na−F, and Na−Na interactions. The Na−O interaction disfavors migration because Na(I‘) sites are found to be energetically more favorable than Na(III‘) sites; the Na−F attraction obviously favors migration; and surprisingly, the Na−Na repulsion also favors migration because moving Na cations into supercages leads to better cation dispersion. This migration occurs in a two-step mechanism that involves first a concerted two-cation jump, S I‘ → S II‘/S II → (S III or S III‘), followed by a S II‘ → S II jump, leading to a net process S I‘ → (S III or S III‘). The preferred binding site in both Na−X and Na−Y involves HFCs anchored by both site II and site III‘ cations. The loading dependence of the heat of HFC adsorption in zeolite Na−X is predicted to be different from that in Na−Y, because of the energetics of cation migration in Na−Y. HFC-134 is generally more strongly bound to both zeolites, because of its ability to make close Na−F and O−H contacts with the zeolites. The binding energy for the gauche conformer of HFC-134 is larger than that for trans at low loadings, but as loading increases, the difference decreases. The highly correlated small-amplitude motion predicted for cations in bare faujasites is quenched upon adsorption of HFCs. Most of the HFCs are too strongly bound to exhibit diffusive behavior during our molecular dynamics simulations.

The unusual mixing behavior of polyisobutylene (PIB) with head-to-head (hhPP) and head-to-tail polypropylene (PP) is studied using large-scale molecular dynamics (MD). The heats of mixing and Flory chi parameters were computed from MD simulations of both blends using a united atom model. The chi parameters from the simulations were estimated from the structure factors using the random phase approximation in analogy with neutron scattering (SANS) experiments. MD simulations for syndiotactic hhPP/PIB predicted a lower critical solution temperature with a chi parameter in very good agreement with SANS experiments on the atactic hhPP/PIB blend. MD simulations also predicted that the isotactic PP/PIB blend was immiscible at high molecular weight in qualitative agreement with cloud point measurements on atactic PP/PIB.