Effect of three-body forces on the statics and dynamics of SF6–(Rg)n and (Rg)13 clusters

Abstract

Molecular dynamics and Monte Carlo simulations are used to examine the effect on the structural properties of heterogeneous SF,(Ar) n and SF6-(Kr) n clusters, and on the melting behavior of heterogeneous SF,-(Ar), and homogeneous (Ar) 13 and (Kr) 13 clusters, of including the three-body Axilrod-Teller-Muto triple-dipole dispersion energies in the total potential energy surface governing the dynamics of the system. The behavior of these systems is governed by potentials constructed from the best available two-body interactions, and from accurate constrained dipole oscillator strength values for the triple-dipole dispersion energy coefficients reported here for the first time. The structural studies show that (virtually) all isomers are destablized by inclusion of the three-body terms, with the "stacked" or "nonwetting" structures being destablized relatively more than isomeric "monolayer" or "wetting" structures. However, the qualitative trends in relative stability are unchanged; in particular, the preference for the SF6 to be fully solvated in larger clusters formed with Ar, but to lie on the surface of larger clusters formed with Kr, remains unchanged. In contrast, the melting temperatures of the stacked and monolayer isomers of the heterogeneous SF,(Ar) 12 cluster undergo substantial change on inclusion of the three-body terms, the former dropping from ca. 30 to 22 K and the latter from 10 to 6 K. The melting temperatures of the homogeneous (Ar) 13 and (Kr) 13 icosahedral isomers also decreased on inclusion of the three-body interactions, and the resulting values are quite different than those obtained using the commonly accepted "effective" two-body LJ(12,6) pair potentials for these systems.

The geometric structure of pure SF6 and mixed Ar/SF6 clusters investigated by core level photoelectron spectroscopy

“…This can be explained by that doping rates and host cluster sizes used are such that they only allow local heating of the surface. The same kind of states observed in molecular dynamics simulations discussed above [11][12][13]. If the Ar core of the cluster would also melt the whole cluster would become liquid.…”

How do quantum effects change conclusions about heterogeneous cluster behavior based on classical mechanics simulations?

“…49,[52][53][54][55] For heterogeneous SF 6 -͑Ar͒ n and SF 6 -͑Kr͒ n cluster systems, including such terms significantly affects the melting temperatures determined from classical Monte Carlo and molecular dynamics simulations. 35 However, it did not change qualitative conclusions regarding trends in structural proclivities and melting behavior yielded by simulations based only on additive ͑anisotropic͒ two-body interactions.…”

“…Unlike some previous work in this area, 63,64 it was not found necessary to impose an artificial constraining potential in order to prevent cluster evaporation during a simulation run. As in most previous studies of these systems, [27][28][29][30][31][32][33][34][35] the two-body potentials appearing in the first sum of Eq. ͑1͒ are represented by the full anisotropic Rg-SF 6 potentials of Pack et al, 42,43 while those in the second sum are the accurate Rg-Rg interaction potentials of Aziz and Slaman.…”

“…5,[15][16][17][18][19][20][21][22][23][24][25][26] The present paper extends a line of theoretical work on such systems which uses the most realistic molecular interaction potentials available and showed that the perturbed infrared spectrum of a chromophore ''solute'' molecule inside of or attached to an inert gas ''solvent'' cluster can probe the structure and dynamical state of such systems. [27][28][29][30][31][32][33][34][35][36][37][38] That work focused attention on the family of clusters SF 6 -͑Rg͒ n ͑RgϭNe, Ar or Kr͒, for which accurate Rg-Rg potential curves [39][40][41] and realistic anisotropic Rg-molecule potential surfaces 42,43 are known, and for which extensive experimental work has been done. 5,[15][16][17][18][19][20][21][22][24][25][26] It showed that even for a quasispherical octahedral molecule like SF 6 interacting with rare gas ''solvent'' particles, the anisotropy of the pairwise potential can significantly affect proclivities for solvation or ''wetting'' vs phase separation in small clusters.…”

“…14,32,34 Further work examined the effect of the three-body triple-dipole Axilrod-Teller-Muto ͑ATM͒ interactions 44,45 on conclusions regarding structural proclivities and melting behavior yielded by such simulations. 35 Barker and coworkers had shown that representing the total potential energy of a many-body system as a pairwise additive sum of the best available two-body potentials plus the ATM terms yielded predictions for the properties of bulk solid, liquid, and gaseous rare gas systems in very good agreement with experiment over a wide range of conditions. [46][47][48][49][50][51] This indicates that the triple-dipole dispersion energies provide a good effective representation of the overall nonadditive energy for such nonpolar species, albeit at least partly because of mutual cancellations of other types of nonadditive energies.…”