J. Chem. Phys.

Self-diffusion of poly(propylene glycol) in nanoporous glasses studied by pulsed field gradient NMR: A study of molecular dynamics and surface interactions
A. Schonhals, F. Rittig, and J. Karger
Pulsed field gradient NMR is applied to investigate the self-diffusion of poly(proypylene glycol) in nanoporous glasses (nominal pore sizes of 2.5–7.5 nm). In general, the diffusion is slowed down by the confinement compared to the bulk. For native pore surfaces covered by hydroxyl groups the spin echo attenuation Ψ displays a bimodal behavior versus q²t (q-norm of a generalized scattering vector). This was explained assuming spatial regions of different diffusivities in a two-phase model. The slow component is assigned to segments forming a surface layer close to the pore walls in which the segments have a lower mobility than those located in the center of the pores. By variation of observation time it was concluded that time constant for the dynamic exchange of segments between these two regions is around 100 ms at room temperature. For silanized pores, the bimodal behavior in the spin echo attenuation Ψ shows a stretched exponential decay versus q²t. The estimated diffusion coefficients decrease strongly with decreasing pore size. The temperature dependence of the diffusion coefficient can be approximated by an Arrhenius law where the activation energy increases with decreasing pore size. The observed pore size dependence for the diffusion of poly(propylene glycol) in silanized nanoporous glasses can be discussed assuming interaction and confining size effects.

Dynamical effects in ab initio NMR calculations: Classical force fields fitted to quantum forces

Mark Robinson and Peter D. Haynes
NMR chemical shifts for an L-alanine molecular crystal are calculated using ab initio plane wave density functional theory. Dynamical effects including anharmonicity may be included by averaging chemical shifts over an ensemble of structural configurations generated using molecular dynamics (MD). The time scales required mean that ab initio MD is prohibitively expensive. Yet the sensitivity of chemical shifts to structural details requires that the methodologies for performing MD and calculating NMR shifts be consistent. This work resolves these previously competing requirements by fitting classical force fields to reproduce ab initio forces. This methodology is first validated by reproducing the averaged chemical shifts found using ab initio molecular dynamics. Study of a supercell of L-alanine demonstrates that finite size effects can be significant when accounting for dynamics.

s