- © 2013 Mineralogical Society of America
Atomistic simulations—molecular dynamics (MD) and Monte Carlo (MC) simulations, ab initio and density functional theory (DFT) calculations—have proved useful in gaining insight into the molecular basis of fundamental processes in aquatic geochemistry, such as solvation, ion pair formation, adsorption, molecular diffusion, and the energetics of mineral phases (Rotenberg et al. 2007; Bickmore et al. 2009; Hamm et al. 2010; Kerisit and Liu 2010; Hofmann et al. 2012; Stack et al. 2012; Wallace et al. 2013). Key strengths of these simulations are their ability to examine the behavior of individual atoms (where spectroscopic and other experimental methods probe the average behavior of large numbers of molecules) and to allow constraints that would be difficult or impossible to impose in the laboratory. These features make atomistic simulations powerful tools for elucidating the manner in which collective phenomena arise from molecular scale properties in geochemical systems. The range of length and time scales probed by atomistic simulations (from angstroms to tens of nanometers and from femtoseconds to microseconds, continuously expanding with advances in the availability and sophistication of computational resources) makes them ideally suited to complement several spectroscopic techniques, including X-ray, neutron, and nuclear magnetic resonance approaches.
A well known limitation of the methods described in the present chapter, particularly in the case of classical mechanical (MD and MC) simulations, is the approximate nature of the models that are used to describe interatomic forces. In the simplest of these simulations, bond lengths and angles are fixed; inter-atomic interactions are modeled as the sum of two-body interactions that depend only on the identity of the interacting atoms and the distance between them; and chemical bonds are not allowed to break or form during a simulation (Allen and Tildesley 1987; Frenkel and Smit 2001). The choice of force …