Extending correlation functions of molecular dynamics simulation by Kovalenko–Hirata and Kobryn–Gusarov–Kovalenko closures for monatomic Lennard-Jones solvent and its application to a calculation of solvation

Abstract

We have tried to extend the short-ranged radial distribution function of the molecular dynamics (MD) simulation for a monatomic Lennard-Jones solvent by applying the Ornstein–Zernike theory with two hybrid closures. One was the hybrid with the Kovalenko–Hirata or the KH closure (hybrid MD–KH closure), whereas the other was the hybrid with the Kobryn–Gusarov–Kovalenko or the KGK closure (hybrid MD–KGK closure). As long as the switching distance between the MD and the KH (or the KGK) is chosen appropriately, the direct correlation functions obtained by these hybrid closures were almost identical to each other, which also agree with those of the hybrid closure with the hypernetted chain or the HNC (hybrid MD–HNC closure). The calculations on the solute–solvent correlation function and the solvation free energy also supported the practical equivalence among the hybrid MD–KH, hybrid MD–KGK, and hybrid MD–HNC closures.

A study on the extension of correlation functions obtained from molecular dynamics simulations by the Ornstein–Zernike theory for modeled molten salts

“…Therefore, it is necessary to optimize the r 0 value. In the previous studies on monatomic LJ fluids, 15,16 this optimization was mainly judged in terms of the r 0 -dependence of the isothermal compressibility κ T . In this study, we add two more criteria: the r 0 -dependence of the excess internal energy u ex and that of the pressure p.…”

“…Similar coincidences among these three hybrid closures were also obtained for monatomic LJ fluids. 16 Therefore, only the MD-KGK closure is employed for the hybrid closure method hereafter in this study.…”

“…The calculation of the solvation is quantitatively affected not only by the accuracy of the approximation employed for the IET calculation between the solute and solvent but also by the quality of the solvent susceptibility response function. 15,16 As a matter of course, using an accurate solvent susceptibility response function leads to an accurate calculation of the solvation. This study focuses on a method to accurately obtain a solvent susceptibility response function.…”

“…Our group has also studied the hybrid closure method for LJ fluids, taking the application of the method to the solvation in mind. 15,16 When the IETs are applied to systems consisting of complicated polyatomic molecules such as proteins, a versatile potential model is often used [for example, Assisted Model Building with Energy Refinement (AMBER), 31 Chemistry at Harvard Molecular Mechanics (CHARMM), 32 Optimized Potentials for Liquid Simulations (OPLS), 33 and Groningen Molecular Simulation (GROMOS) 34 ]. These versatile potential models are expressed by a superposition of the LJ and Coulomb potentials.…”

“…This article discusses the hybrid closure method, which is a method to extend the RDF of modeled molten salt obtained by an MD simulation to a long-ranged space by incorporating it into the closure relation of the IET. While the hybrid closure method has been studied for single-component systems, 15,16,[26][27][28][29][30] this study considers a two-component system. Furthermore, we also discuss the accuracy of the hypernetted chain (HNC), Kovalenko-Hirata (KH), and Kobryn-Gusarov-Kovalenko (KGK) closures, which are existing closure approximations, when they are applied to the modeled molten salt.…”