Molecular dynamics calculations on amylose fragments. I. Glass transition temperatures of maltodecaose at 1, 5, 10, and 15.8% hydration.

Molecular dynamics simulations (NPT ensembles, 1 atm) using the all atom force field AMB99C (F. A. Momany and J. L. Willett, Carbohydrate Research, Vol. 326, pp 194-209 and 210-226), are applied to a periodic cell containing ten maltodecaose fragments and TIP3P water molecules. Simulations were carried out at 25 K intervals over a range of temperatures above and below the expected glass transition temperature, T(g), for different water concentrations. The amorphous cell was constructed through successive dynamic equilibration steps at temperatures above T(g) and the temperature lowered until several points of reduced slope (1/T vs volume) were obtained. This procedure was carried out at each hydration level. Each dynamics simulation was continued until the volume remained constant without up or down drift for at least the last 100 ps. For a given temperature, most simulations required 400-600 ps to reach an equilibrium state, but longer times were necessary as the amount of water in the cell was reduced. A total of more than 30 ns of simulations were required for the complete study. The T(g) for each hydrated cell was taken as that point at which a discontinuity in slope of the volume (V), potential energy (PE), or density (rho) vs 1/T was observed. The average calculated T(g) values were 311, 337, 386, and 477 K for hydration levels of 15.8, 10, 5, and 1%, respectively, in generally good agreement with experimental values. The T(g) for anhydrous amylose is above the decomposition temperature for carbohydrates and so cannot be easily measured. However, it has also been difficult to obtain a value of T(g) for anhydrous amylose using simulation methods. Other molecular parameters such as end-to-end distances, mean square distributions, and pair distributions are discussed. Published 2002 John Wiley & Sons, Inc.