Search

This article reviews recent progress in the computer simulation of liquid crystals at the molecular level. It covers the use of simple rigid-body models of the constituent molecules, and more detailed modelling via atomistic force fields. Bulk mesophases, inhomogeneous systems, and interfaces are discussed. Recent progress in calculating elastic properties and dynamics is summarised. As well as presenting an overview, some specific topics of recent interest are highlighted: the biaxial nematic phase, chiral phases, ionic liquid crystals, and charge-transfer systems. [ABSTRACT FROM AUTHOR]

In this paper, we present molecular dynamics calculations of the Frank elastic constants, and associated time correlation functions, in nematic liquid crystals. We study two variants of the Gay–Berne potential, and use system sizes of half a million molecules, significantly larger than in previous studies of elastic behaviour. Equilibrium orientational fluctuations in reciprocal (k-) space were calculated, to determine the elastic constants by fitting at low |k|; our results indicate that small system size may be a source of inaccuracy in previous work. Furthermore, the dynamics of the Gay–Berne nematic were studied by calculating time correlation functions of components of the order tensor, together with associated components of the velocity field, for a set of wave vectorsk. Confirming our earlier work, we found exponential decay for splay and twist correlations, and oscillatory exponential decay for the bend correlation. In this work, we confirm similar behaviour for the corresponding velocity components. In all cases, the decay rates, and oscillation frequencies, were found to be accurately proportional tok2for smallk, as predicted by the equations of nematodynamics. However, the observation of oscillatory bend fluctuations, and corresponding oscillatory shear flow decay, is in contradiction to the usual assumptions appearing in the literature, and in standard texts. We discuss the advantages and drawbacks of using large systems in these calculations. [ABSTRACT FROM PUBLISHER]

We highlight some links between molecular dynamics and Monte Carlo algorithms used to simulate condensed matter systems. Special attention is paid to the question of sampling the desired statistical ensemble. [ABSTRACT FROM AUTHOR]

We have calculated the direct correlation function, c(1, 2), via a high-order virial expansion, for systems of hard spheres and spheroids, in both the isotropic and nematic phases. For hard spheres, we find that truncation at sixth order in density gives good agreement with simulation data. Close to freezing, the virial series still appears to converge to the simulation results, but there are significant discrepancies, particularly at very small separations. In the non-overlap region, the virial theory begins to capture the features found from simulation. We also calculate the pair distribution function, g(1, 2), from our estimates of c(1, 2), and find good agreement with simulation data at all densities up to freezing. For hard, prolate spheroids of aspect ratio 3 : 1, we calculate c(1, 2) and g(1, 2) in the isotropic phase, again finding good agreement with simulation data at moderate densities. Finally, we present the result of our calculations on c(1, 2) for 3 : 1 spheroids in the nematic phase. [ABSTRACT FROM AUTHOR]

Expressions for intermolecular forces and torques, derived from pair potentials between rigid non-spherical units, are presented. The aim is to give compact and clear expressions, which are easily generalized, and which minimize the risk of error in writing molecular dynamics simulation programs. It is anticipated that these expressions will be useful in the simulation of liquid crystalline systems, and in coarse-grained modelling of macromolecules. [ABSTRACT FROM AUTHOR]