The recently introduced continuum-configurational-bias method for Monte Carlo simulations is employed for the generation of large samples of many- chain n-alkane systems with chain lengths of 11, 24 and 71 carbon atoms. The simulations are used to investigate the adequacy of representing methylene groups as united-atom Lennard-Jones interaction sites, and to test the configurational-bias approach against traditional random moves and reptation moves with respect to the computational efficiency and numerical stability of the calculated ensemble averages. The results of simulations with constant pressure, temperature, and number of molecules demonstrate that, with an appropriate mixture of different types of Monte Carlo moves, an efficient and stable strategy can be obtained. Adjustment of the Lennard-Jones parameters leads to results that are in good agreement with experimental data for the density of liquid alkanes over a large temperature interval.
Bibliographical noteFunding Information:
This research was supported in part by the Bundesministerium fiir Forschung und Technologie Grant No. 03M4028. We are grateful to the Centro Svizzero di Calcolo Scientifico for generous allocations of supercomputer time on its NEC-SX/3. We are also indebted to L. Dodd and D. Theodorou and the Department of Polymer Physics of Bayer AG for many valuable and stimulating discussions.