NONEQUILIBRIUM FLUCTUATION-DISSIPATION RELATIONS FOR INDEPENDENT RANDOM RATE-PROCESSES WITH DYNAMICAL DISORDER

A class of rate processes with dynamical disorder is investigated base d on the two following assumptions: (a) the system is composed of a ra ndom number of particles (or quasiparticles) which decay according to a first-order kinetic law; (b) the rate coefficient of the process is a random function of time with known stochastic properties. The formal ism of characteristic functionals is used for the direct computation o f the dynamical averages. The suggested approach is more general than the other approaches used in the literature: it is not limited to a pa rticular type of stochastic process and can be applied to any type of random evolution of the rate coefficient. We derive an infinity of exa ct fluctuation-dissipation relations which establish connections among the moments of the survival function and the moments of the number of surviving particles. The analysis of these fluctuation-dissipation re lations leads to the unexpected result that in the thermodynamic limit the fluctuations of the number of particles have an intermittent beha vior. The moments are explicitly evaluated in two particular cases: (a ) the random behavior of the rate coefficient is given by a non-Markov ian process which can be embedded in a Markovian process by increasing the number of state variables and (b) the stochastic behavior of the rate coefficient is described by a stationary Gaussian random process which is generally non-Markovian. The method of curtailed characterist ic functionals is used to recover the conventional description of dyna mical disorder in terms of the Kubo-Zwanzig stochastic Liouville equat ions as a particular case of our general approach. The fluctuation-dis sipation relations can be used for the study of fluctuations without m aking use of the whole mathematical formalism. To illustrate the effic iency of our method for the analysis of fluctuations we discuss three different physicochemical and biochemical problems. A first applicatio n is the kinetic study of the decay of positrons or positronium atoms thermalized in dense fluids: in this case the time dependence of the r ate coefficient is described by a stationary Gaussian random function with an exponentially decaying correlation coefficient. A second appli cation is an extension of Zwanzig's model of ligand-protein interactio ns described in terms of the passage through a fluctuating bottle neck ; we complete the Zwanzig's analysis by studying the concentration flu ctuations. The last example deals with jump rate processes described i n terms of two independent random frequencies; this model is of intere st in the study of dielectric or conformational relaxation in condense d matter and on the other hand gives an alternative approach to the pr oblem of protein-ligand interactions. We evaluate the average survival function in several particular cases for which the jump dynamics is d escribed by two activated processes with random energy barriers. Depen ding on the distributions of the energy barriers the average survival function is a simple exponential, a stretched exponential, or a statis tical fractal of the inverse power law type. The possible applications of the method in the field of biological population dynamics are also investigated. (C) 1996 American Institute of Physics.