Entropy Scaling-Based Correlation for Estimating the Self-Diffusion Coefficients of Pure Fluids

A new model for correlating self-diffusion coefficients
is proposed
here. It is based on the entropy scaling concept and makes it possible
to quantitatively estimate self-diffusion coefficients in any fluid
state (subcritical gas and liquid as well as supercritical states).
A general relationship between a dimensionless form of the self-diffusion
coefficient and a function of the Tv-residual entropy is proposed.
This relationship involves thermodynamic properties that were straightforwardly
estimated using the I-PC-SAFT and tc-PR equations of state (EoSs). First, the proposed model was parameterized
using component-specific parameters. To do so, 72 pure components
and ∼2400 pieces of experimental data were considered. Second,
a chemical-family parametrization method was used. Four chemical families
were defined, and parameter sets valid for all of the compounds of
a given family were estimated. Regardless of the considered EoS, the
component-specific parametrization enables us to describe data with
an average deviation of 7.5%, while when using a chemical-family parametrization,
this deviation reaches 9.5%.