A New Group-Contribution Approach for Ideal Gas Heat Capacity, Critical Temperature and Normal Boiling Point

We propose two new group-contribution models for the
ideal gas
heat capacity cpig(T), one for
the critical temperature Tc, and the normal
boiling point Tb. For all properties the
model is based on a decomposition into base and second-order groups,
which are simultaneously parametrized to experimental data. For the
critical temperature and normal boiling point, each group is assigned
one or two parameters based on data availability. For the critical
temperature, 1454 substances were used for optimization, resulting
in a mean absolute relative deviation (MARD) from experimental data
of 2.46%. For the normal boiling point 5591 substances were used for
optimization with a MARD of 2.26%. For the ideal gas heat capacity
we propose two new group-contribution models based on the Aly-Lee
equation and a reparameterization of the Joback and Reid model. The
models are parametrized for the temperature range 50–3000 K
for 1009 substances, where the MARD to the experimental data for the
best model is 1.73%. In addition, we individually parametrize 1162
substances with a deviation to the experimental data of 0.54%. As
Supporting Information, we provide Python code to estimate the critical
temperature, the normal boiling point, and the ideal gas heat capacity
based on the proposed models.