Abstract
The calculation of the characteristics of the selectivity in the membrane separation of mixtures, based on basic physicochemical characteristics of the solution, is a difficult task due to the large number of influencing factors. To solve this problem, for a theoretical evaluation of the selectivity of ultrafiltration membranes, modern distributed algorithms are used for a probabilistic approach, in which the mechanism of the separation process differs significantly from the separation mechanism in nanomembranes. As a result of the proposed models and modern distributed algorithms, estimates of the selectivity of nanofiltration membranes for individual ions were obtained. It was found that an important feature of mixing with the mixture flowing through the membrane is the dependence of the effective diffusion coefficient on time. Also, as a result of calculations by the proposed model, it was found that this feature is modeled by the coefficient of anomalous fractal diffusion in time, as well as by the displacement of the effective separation zone in the membrane.