THEORY FOR THE NONEQUILIBRIUM DYNAMICS OF FLEXIBLE CHAIN MOLECULES - RELAXATION TO EQUILIBRIUM OF PENTADECANE FROM AN ALL-TRANS CONFORMATION

We extend to nonequilibrium processes our recent theory for the long t ime dynamics of flexible chain molecules. While the previous theory de scribes the equilibrium motions for any bond or interatomic separation in (bio)polymers by time correlation functions, the present extension of the theory enables the prediction of the nonequilibrium relaxation that occurs in processes, such as T-jump experiments, where there are sudden transitions between, for example, different equilibrium states . As a test of the theory, we consider the ''unfolding'' of pentadecan e when it is transported from a constrained all-trans conformation to a random-coil state at thermal equilibrium. The time evolution of the mean-square end-to-end distance [R-end(2)(t)](noneq) after release of the constraint is computed both from the theory and from Brownian dyna mics (BD) simulations. The lack of time translational symmetry for non equilibrium processes requires that the BD simulations of the relaxati on of [R-end(2)(t)](noneq) be computed from an average over a huge num ber of independent trajectories, rather than over successive configura tions from a single trajectory, which may be used to generate equilibr ium time correlation functions. Adequate convergence ensues for the no nequilibrium simulations only after averaging 9000 trajectories, each of 0.8 ns duration. In contrast, the theory requires only equilibrium averages for the initial and final states, which may be readily obtain ed from a few Brownian dynamics trajectories. Therefore, the new metho d produces enormous savings in computer time. Moreover, since both the ory and simulations use identical potentials and solvent models, the t heory contains no adjustable parameters. The predictions of the theory for the relaxation of [R-end(2)(t)](noneq) agree very well with the B D simulations. This work is a starting point for the application of th e new method to nonequilibrium processes with biological importance su ch as the helix-coil transition and protein folding. (C) 1998 American Institute of Physics. [S0021-9606(98)50720-9].