Mixed Quantum-classical Molecular Dynamics Simulations of the Structure of Ionic Solutions

Neutron scattering offers a powerful window into the atomistic structure and dynamics of solid, liquid and amorphous materials. A key measurable quantity is the pair distribution function, which relates the distances between scattering sites in the material, containing a wealth of structural information. However, even in a somewhat simple chemical system such as salt dissolved in water, it is not always easy to extract the contributions of specific atom pairs. Classical Monte Carlo and molecular dynamics simulations are commonly employed to supplement the analysis of the experimental data, but these methods are only as good as the classical force fields they employ. Molecular interactions are more accurately framed within the quantum mechanical electronic structure problem. Such methods are far more computationally intensive than classical mechanics for performing molecular dynamics simulations. This places practical limitations on the system size and trajectory length, both of which are important for minimizing fluctuations in statistical structural and dynamical quantities. Particularly in the case of studying solvation, the system size limitations place constraints on studying concentration effects. We propose employing a mixed quantum-classical molecular dynamics paradigm for studying ion solvation in water using the recent QM/MM extension of the NAMD molecular dynamics package. Together with neutron scattering measurements taken at the Nanoscale-Ordered Materials Diffractometer at Oak Ridge National Laboratory, the aim is to use computer simulations and high-resolution measurements to better understand ion solvation in atomistic detail.