The dependence of the individual mean square displacement of rare gases in binary mixtures is studied by a combined experimental and theoretical approach. We show that the diffusion constant can be varied in a considerable range by changing the molar fractions of the mixtures. On the experimental side, NMR diffusion measurements are done on hyperpolarized 3He and 129Xe, mixed with several inert buffer gases, in the presence of a magnetic field gradient. The results are compared to diffusion coefficients obtained from atomistic molecular dynamics simulations based on Lennard-Jones type potentials of the corresponding gas mixtures, and to appropriate analytical expressions, yielding very good mutual agreement. This study is the first quantitative validation of the effects of the mutual interactions between gas particles on the individual diffusion properties. It is shown that the dependency of gas phase diffusion properties on the local chemical environment may not be neglected, e.g. in diffusion-controlled chemical reactions.

We present first principles calculations of the NMR solvent shift of adenine in aqueous solution. The calculations are based on snapshots sampled from a molecular dynamics simulation, which were obtained via a hybrid quantum-mechanical/mechanical modeling approach, using an all-atom force field (TIP3P). We find that the solvation via the strongly fluctuating hydrogen bond network of water leads to nontrivial changes in the NMR spectra of the solutes regarding the ordering of the resonance lines. Although there are still sizable deviations from experiment, the overall agreement is satisfactory for the 1H and 15N NMR shifts. Our work is another step toward a realistic first-principles prediction of NMR chemical shifts in complex chemical environments.

We present a capping scheme for hybrid calculations which is designed for a systematic optimization to reproduce the molecular structure, frontier bond potential, and spectroscopic properties for the quantum subsystem. Our technique is capable of reducing the perturbations of the electronic structure which are normally caused by conventional link atoms between quantum and classical regions. Specifically, we propose analytic effective core potentials with a small set of adjustable parameters, which are optimized to reproduce the full-quantum-mechanical (full-QM) properties in the direct environment of the bond cleavage. The capping potentials are conceptually simple and easy to employ in most instances without significant code modifications. They do not require any further external geometry constraints and yield also reasonable results for the potential energy surface. We benchmark these potentials for a series of chemically and biologically relevant molecules calculating NMR chemical shifts, protonation energies, and optimized geometries. Our optimized QM/mechanical modeling (MM) potentials are another step toward a realistic first-principles prediction of spectroscopic parameters in complex chemical environments using hybrid QM/MM calculations.