337
0195-928X/04/0300-0337/0 © 2004 Plenum Publishing Corporation
International Journal of Thermophysics, Vol. 25, No. 2, March 2004 (© 2004)
Use of the DIPPR Database for the Development of
QSPR Correlations: Solid Vapor Pressure and Heat
of Sublimation of Organic Compounds
1
1
Paper presented at the Fifteenth Symposium on Thermophysical Properties, June 22–27,
2003, Boulder, Colorado, U.S.A.
B. T. Goodman,
2
W. V. Wilding,
2
J. L. Oscarson,
2
and R. L. Rowley
2, 3
2
Department of Chemical Engineering, 350 CB, Brigham Young University, Provo, Utah
84602, U.S.A.
3
To whom correspondence should be addressed. E-mail: rowley@byu.edu
A group-contribution method has been developed to estimate the heat of
sublimation (DH
sub
) at the triple point for organic solids. The correlation was
developed using DH
sub
values from a training set of 218 compounds, and it was
tested by comparing predicted and solid vapor pressures for 87 compounds
(1103 data points). The predicted solid vapor pressures were obtained from the
DH
sub
correlation using the Clausius–Clapeyron equation. The absolute average
deviation in the logarithm of vapor pressure was 0.371. The new method com-
pares favorably with Bondi’s method for prediction of DH
sub
and represents an
improvement over other available methods for predicting solid vapor pressures.
KEY WORDS: group contributions; heat of sublimation; organic compounds;
prediction; QSPR; solid vapor pressure.
1. INTRODUCTION
Evaluated databases such as the DIPPR 801 database [1] are useful tools
in developing new property prediction methods. Because the data have
been evaluated and an uncertainty assigned, training sets of specific
accuracy can be used to develop correlations and/or weight the regression
of parameters or constants in the correlating equations. Data in the data-
base not used in developing the correlation can be used to test the appli-
cability of the correlation to additional compounds. Additionally, a data-
base serves as an important source of fundamental and molecular-based