Including image charge effects in the molecular dynamics simulations of molecules on metal surfaces

Abstract

Combinatorial bio-techniques have demonstrated that proteins can be good and even selective binders for several inorganic surfaces, including metals. However, the understanding of the basic physical mechanisms that govern such interactions did not keep up with the success in these experiments. The comprehension of such mechanisms would greatly benefit from the computational investigation of the problem. Because of the complexity of the system, classical molecular dynamics simulations based on an atomistic description appear to be the best compromise between reliability and feasibility. For proteins interacting with metal surfaces, however, methodological improvements with respect to standard Molecular Dynamics (MD) of proteins are needed, since the polarization of the metal induced by the protein (and the surrounding water) is not generally negligible. In this article, we present a simple approach to introduce metal polarization effects (often termed image effects) in MD simulations by exploiting standard features of bio-oriented MD codes such as the widely used GROMACS and NAMD. Tests to show the reliability of the proposed methods are presented, and the results for a model application showing the importance of image effects are also discussed.

Applying electric field to charged and polar particles between metallic plates: Extension of the Ewald method

“…Inclusion of the image effect in molecular dynamics simulations of molecules near a conducting or polarizable surface is still challenging in a variety of important systems including proteins 22 , polyelectrolytes 23 , and colloidal particles 24 . Moreover, the effects of applied electric field remain largely unexplored on the microscopic level, while the continuum electrostatics is well established 16,25 .…”

Molecular dynamics simulation of a DOPA/ST monolayer on the Au(111) surface

“…22 The computed binding energy for DOPA is À237.5 kJ mol À1 and À224.7 kJ mol À1 for ST. Because of the lack of experimental results, it is hard to compare these energies with experimental observations. Upon comparison with the results of other organic molecules, 21,47,48 we note that these binding energies are of the expected magnitude.…”

Computational chemistry for graphene-based energy applications: progress and challenges

“…To overcome this limitation a number of FFs have incorporated a description of atomic polarisation. Simple descriptions of polarisability include the rigid-rod dipole 54 and core-shell 55 (or Drude oscillator) models. A more rigourous description of polarisation is that used in the AMEOBAPRO FF [56][57][58][59] ; this model of polarisation was partnered with a high-quality description of the electrostatics via the distributed multipole approximation.…”