Test of mode coupling theory for a supercooled liquid of diatomic molecules.I. Translational degrees of freedom

TL;DR

This study uses molecular dynamics simulations to test mode coupling theory predictions for a supercooled diatomic liquid, confirming key theoretical aspects like the transition temperature and scaling laws, with some limitations in the critical law validation.

Contribution

It provides a detailed comparison of MCT predictions with simulation data for translational dynamics in a supercooled diatomic liquid, highlighting the theory's validity and limitations.

Findings

Existence of a unique transition temperature T_c.

First and second scaling laws are largely confirmed.

Power-law behavior of relaxation times with weak q-dependence.

Abstract

A molecular dynamics simulation is performed for a supercooled liquid of rigid diatomic molecules. The time-dependent self and collective density correlators of the molecular centers of mass are determined and compared with the predictions of the ideal mode coupling theory (MCT) for simple liquids. This is done in real as well as in momentum space. One of the main results is the existence of a unique transition temperature T_c, where the dynamics crosses over from an ergodic to a quasi-nonergodic behavior. The value for T_c agrees with that found earlier for the orientational dynamics within the error bars. In the beta- regime of MCT the factorization of space- and time dependence is satisfactorily fulfilled for both types of correlations. The first scaling law of ideal MCT holds in the von Schweidler regime, only, since the validity of the critical law can not be confirmed, due to a strong interference with the microscopic dynamics. In this first scaling regime a consistent description within ideal MCT emerges only, if the next order correction to the asymptotic law is taken into account. This correction is almost negligible for q=q_max, the position of the main peak in the static structure factor S(q), but becomes important for q=q_min, the position of its first minimum. The second scaling law, i.e. the time-temperature superposition principle, holds reasonably well for the self and collective density correlators and different values for q. The alpha-relaxation times tau_q^(s) and tau_q follow a power law in T-T_c over 2 -- 3 decades. The corresponding exponent gamma is weakly q-dependent and is around 2.55. This value is in agreement with the one predicted by MCT from the value of the von Schweidler exponent but at variance with the corresponding exponent gamma