The applicability of Satava's integral method for the kinetic description of nonisothermal heterogeneous processes is first discussed. The method is found insensitive as regards distinction between different types of nucleation-growth models. This disadvantage is countered by using additional criteria, such as the interval of straightline-fitting, the standard deviation and the value of the preexponential factor. The algorithm for the necessary mathematical operations is explained. A graphical printout in the form of a numerical plot available for any computer is described. The program is tested on different processes. It proves suitable for the preliminary classification of processes and for detection of changes in the reaction mechanism by separate analysis of three consecutive parts of the process.Data comparatively easily obtained from dynamic thermal measurements have led to a considerable rise in the number of papers dealing with the kinetics of thermally activated processes [1]. The characteristic accumulation of recorded chart strips of otherwise stored data is challenged by the use of computers [1,2]. The purpose of the present communication is to present and discuss a possible algorithm for kinetic data evaluation, based on our previous experience with computer programming directed to the elucidation of the kinetics and mechanisms of heterogeneous processes [3][4][5][6].
FormulationThe mathematical procedure is based on the integral method of kinetic data evaluation [1 ], employing a most simple kinetic equation to describe the time-temperature behaviour of heterogeneous processes [7]. This assumes direct proportionality [1, 7] between two functions, g (c~) and P(T), depending respectively on the instantaneous state of the system (represented here by a single parameter, the fractional conversion c~) and on the temperature T(controlled from the surrounding system as a linear function of time, and uniform throughout the system investigated). The main advantage is a simple set of c~ vs. T input data, with no need for measured or computed derivatives.The establishment of the function P(T) is mathematically complicated, but is based on the integration of known exponential form of the Arrhenius rate con-7
