Glassy behavior of molecular crystals: A comparison between results from MD-simulation and mode coupling theory

TL;DR

This study compares molecular dynamics simulations and mode coupling theory in analyzing the glassy behavior of chloroadamantane molecular crystals, revealing insights into relaxation processes and the glass transition temperature.

Contribution

It provides a detailed comparison between MD simulation results and mode coupling theory predictions for molecular crystals, highlighting discrepancies and the role of anisotropic crystal fields.

Findings

MD simulation estimates $T_c$ at ~217K

Mode coupling theory predicts $T_c$ at ~267K

Mode coupling theory reproduces qualitative relaxation features

Abstract

We have investigated the glassy behavior of a molecular crystal built up with chloroadamantane molecules. For a simple model of this molecule and a rigid fcc lattice a MD simulation was performed from which we obtained the dynamical orientational correlators $S_{\lambda \lambda '}({\bf{q}},t)$ and the ``self'' correlators $S_{\lambda \lambda '}^{(s)}(t)$, with $\lambda = (\ell, m)$, $\lambda' = (\ell', m')$. Our investigations are for the diagonal correlators $\lambda = \lambda'$. Since the lattice constant decreases with decreasing temperature which leads to an increase of the steric hindrance of the molecules, we find a strong slowing down of the relaxation. It has a high sensitivity on $\lambda$, $\lambda '$. For most $(\ell,m)$, there is a two-step relaxation process, but practically not for $(\ell,m) = (2,1)$, $(3,2)$, $(4,1)$ and $(4,3)$. Our results are consistent with the $\alpha$-relaxation scaling laws predicted by mode coupling theory from which we deduce the glass transition temperature $T_c^{MD} \cong 217K$. From a first principle solution of the mode coupling equations we find $T_c^{MCT} \cong 267K$. Furthermore mode coupling theory reproduces the absence of a two-step relaxation process for $(\ell,m)=(2,1)$, $(3,2)$, $(4,1)$ and $(4,3)$, but underestimates the critical nonergodicity parameters by about 50 per cent for all other $(\ell,m)$. It is suggested that this underestimation originates from the anisotropic crystal field which is not accounted for by mode coupling theory. Our results also imply that phonons have no essential influence on the long time relaxation.