Personnel
Overall Objectives
Research Program
New Software and Platforms
New Results
Partnerships and Cooperations
Dissemination
Bibliography
XML PDF e-pub
PDF e-Pub


Section: New Results

Incremental methods for long range interactions

Participants : Semeho Edorh, Stephane Redon.

Adaptively Restrained Molecular Dynamics (ARMD) were recently proposed with the purpose of speeding up molecular simulations. The main idea is to modify the Hamiltonian such that the kinetic energy is set to zero for low velocities, which allows to save computational time since particles do not move and forces need not be updated.

We continued our work on developing an extension of ARMD to electrostatic simulations. Therefore, we developed a fast method dedicated to the computation of the electrostatic potential in adaptively restrained systems. The proposed algorithm is derived from a multigrid-based alternative to the popular particle mesh methods. Our algorithm ,labeled as Incremental Mesh Continuum Method (IMCM),was implemented inside LAMMPS, a popular molecular dynamics simulation package. During ARMD simulations, IMCM scales with the number of active particles.

The performance of the new algorithm was accessed on various molecular systems. 1 showed that IMCM is able to outperform the well-established Particle Particle Particle Mesh (P3M) for adaptively restrained simulations. For an aqueous solution of sodium chloride, water molecules can be adaptively restrained. On this system, ARMD was able to reproduce static properties of sodium chloride (2). When a functionalized nanopore is placed at the center of the system, ARMD and IMCM were able to reproduce the ion selectivity property (3). For this benchmark, this positively charged nanopore acts like a sieve that blocks the flux of Sodium atoms, while promoting the crossing of Chlorine particles(4).

Figure 1. Performance depending on the percentage of active particles for different number of processes. Performance of LAMMPS P3M is shown as a reference (dotted lines, pentagram marker) — it does not depend on the percentage of active particles. In all cases electrostatics were computed at similar accuracy (10-5).
IMG/performance.png
Figure 2. Ion-water pair distribution functions using armd with the NaCl/ϵ force field at 298 K the rigid water model SPC/ϵ and an ionic concentration of 10.0 molal. Different restraining parameters (εr,εf) were tested on water molecules. Na and Cl are always active. Black line corresponds to a standard molecular dynamics simulation of the system.
IMG/edorh_rdf.png
Figure 3. Number density of chlorine (red dotted line) and sodium (blue dashed line) ions along z-axis using standard ARMD (Left) and MD (Right). Both methods show the ion selectivity of the nanopore which is located at z=0.
IMG/edorh_concentration.png
Figure 4. Nonuniform distributions of number density of chlorine (Top) and sodium (Bottom) ions driven by an external electric field (black arrow) E=1 V/A using standard MD (Left) and ARMD (Right). The gray rectangles at z=0 mark the graphene sheet. Both ions form concentration polarization layers.
IMG/edorh_contour.png