Improved hydrodynamic equations for the accurate prediction of diffusivities in supercritical carbon dioxide

•Very simple and accurate predictive models for tracer diffusivities (D12) in SC-CO2.•D12 depends on temperature, solvent viscosity, and common properties of solute.•Global absolute errors of the pure predictions lie between 7.86% and 8.56%.•Global errors of the pure predictions are negligible (between 0.47% and 0.53%).•Very large database compiled for the validation: 159 systems/4529 data points.

The tracer diffusion coefficients are fundamental quantities in simulation and design. Due to the increasing interest upon biorefinery and sustainability in general, green solvents and processes, like carbon dioxide and supercritical fluid extraction, are attracting relevance in both chemistry and chemical engineering research and development. In this work, tracer diffusion coefficients at infinite dilution are focused aiming to propose reliable models for their pure estimation. Four predictive hydrodynamic models were proposed on the basis of modifications introduced in the original expressions of Wilke–Chang, Scheibel, Lusis–Ratcliff, and Tyn–Calus. The modified equations provide reliable results with average absolute errors between 7.86% and 8.56%, and inferior dispersion around the averages. On the contrary, the original correlations taken from the literature achieve errors between 11.89% and 27.25%, along with higher scattering of results. Furthermore, the new expressions offer average errors between 0.47% and 0.53%, while the original ones provide systematic overestimations between 2.95% and 27.23%. In the whole, the new expressions proposed in this work are equally able to predict accurately tracer diffusion coefficients of any solutes in supercritical carbon dioxide.