Adaptive resolution molecular dynamics simulation through coupling to an internal particle reservoir

Abstract

For simulation studies of (macro) molecular liquids it would be of significant interest to be able to adjust or increase the level of resolution within one region of space, while allowing for the free exchange of molecules between open regions of different resolution or representation. We generalize the adaptive resolution idea and suggest an interpretation in terms of an effective generalized grand canonical approach. The method is applied to liquid water at ambient conditions.

Statistical mechanics of Hamiltonian adaptive resolution simulations

“…(47) is valid for any Hamiltonian system: notably, this intimate relation between the force density field and the density gradients is independent on the fluid compressibility. It explains the essence of many algorithms [15][16][17] designed to impose a flat density profile by adding an external force "correction" to the system (which, according to Eq. (47) has to ensure vanishing total force density field f r = 0).…”

“…This relation explains the basis of the "correction force field" used to control the density profile, not only in Ref. 15 but also in many other algorithms using domain decomposition (see, e.g., Refs. 16 and 17).…”

“…5 Density fluctuations are determined by the fluid compressibility, specified by the integral of the radial distribution function, and by finely tuning the CG potential one can match the compressilibities of the CG and AA domains. Having the same compressibility does not however ensure the same pressure equation of state and to ensure a constant density profile over the CG and AA domains, a recent work 15 proposes the imposition of a "correction force field," which is iteratively evaluated according to the idea of imposing pressure balance (and thus involving compressibilities). The existence of a Hamiltonian permits us to derive a fundamental relation between the force density and the density gradient, which turns out to be independent on the compressibility.…”

Hamiltonian Adaptive Resolution Simulation for Molecular Liquids

“…In particular, the chemical potential gradient, which is generally established across the transition region, would have to be counterbalanced [29]. This idea leads to the following form of the compensation term in Eq.…”

“…The method is nonetheless robust, since it allows us to define temperature, pressure and density everywhere, and the introduction of a thermodynamic force [28] in the transition zone paved the way to open system MD simulations [20,29]. Despite the success of the force-interpolation based AdResS method, though, the lack of a Hamiltonian description in the transition region is a drawback: it hampers a general statistical theory for the whole setup, limits the choice of the simulation ensemble and prevents Monte Carlo simulations.…”