Revisiting Tracer Liu-Silva-Macedo model for binary diffusion coefficient using the largest database of liquid and supercritical systems

Abstract

This work compiles the largest database of tracer diffusion coefficients (D12) containing 6180 experimental points from 331 non-polar and weakly polar liquid and supercritical systems. Then, the Tracer Liu − Silva − Macedo (TLSM) model and its 1-parameter correlations (TLSMd and TLSMen) are evaluated using this database, taking into account the importance of phenomenological and reliable equations for D12 estimation. The TLSM model achieves good results with absolute average relative deviations (AARD) of 16.84 % while TLSMd and TLSMen show better performance with AARD of 4.53 % and 4.55 %, respectively. All properties and parameters needed for D12 estimation are compiled in Appendix. For comparison, the models of Wilke-Chang and Reddy-Doraiswamy, and the correlations of Magalhães et al. (LJ-1) and Dymond-Hildebrand-Batschinsky (DHB) are also assessed.