Prediction of Critical Temperature and Critical Volume of Mixtures by a New Group Contribution Method

The vapor–liquid critical properties of mixtures
are essential
for defining the vapor–liquid phase boundary and play a crucial
role in predicting various thermophysical properties of mixtures,
which are particularly significant in applications such as supercritical
extraction and transcritical cycles. While experimental measurement
is the most efficient and direct approach to obtaining critical properties,
it is often time-consuming and labor-intensive, necessitating the
use of theoretical prediction methods. Nonetheless, existing prediction
models tend to be complex and frequently rely on the critical properties
of pure substances, which limit their applicability. In this article,
a new group contribution method for predicting the critical temperature
and critical volume of mixtures is proposed. The new group contribution
method is simple in calculation form, simple in group division, has
good accuracy, and does not need the critical temperature and critical
volume of pure substances when calculating the critical temperature
and critical volume of mixtures. This new method includes 24 groups
and can be applied to systems consisting of organic compounds made
up of C, H, O, F, Cl, Br, and I elements or CO2. The experimental
critical temperatures of 272 compounds and 368 groups of binary mixtures
(3223 data points), as well as the experimental critical volumes of
224 compounds and 68 groups of binary mixtures (400 data points),
were used to determine the group contribution values and model parameters.
The average absolute relative deviations (AARDs) for the correlation
of compounds are 1.28% for critical temperature and 4.93% for critical
volume. For binary mixtures, the AARDs are 1.62% for the critical
temperature and 7.33% for the critical volume. Additionally, the predictive
capability of the new method for critical temperature and critical
volume has been evaluated. The AARDs for critical temperature are
2.48, 1.98, and 0.94% for 25 data points of pure substances, 615 data
points of binary mixtures, and 565 data points of ternary mixtures,
respectively. For critical volumes, the AARDs are 5.52% for 26 data
points of pure substances and 7.49% for 61 data points of binary mixtures.