A.2 Pure component property estimation methods and software

In the area of pure component property estimation methods, we have concentrated our efforts on the development of new methods for properties not covered by us before, and, in the extension of existing methods to also cover compounds with complex molecular structures (heterocyclic-ring, multi-ring, and large molecules). The properties have been classified as primary (can be estimated only from molecular structural information), secondary (can be estimated from other known properties) and functional (depends on temperature, pressure and/or composition). New methods have been developed mainly for primary properties. For secondary properties, well known and well tested correlations have been programmed and included in the property prediction software (ProPred). For functional properties, two options have been developed. In the first option, given the data and a function, the function parameters are determined through regression. In the second option, data are generated through an equation of state or corresponding states model and then fitted to a specified function.
In the primary property estimation area, the melting point method of Constantinou and Gani is being revised and a new model for estimation of the standard heats of fusion is being developed. For primary properties, the Constantinou & Gani method as well as the Joback method has been added in ProPred. The primary property estimation methods are also being extended to polymers. In the secondary property area, a large collection of methods has been tested and added to ProPred (properties such as solubility parameter, refractive index, and many more have been added). A new atom contribution method (proposed by Wilson) is being currently added to ProPred. In the functional property area, generation and regression of temperature dependent data for vapour pressures is now possible. Other properties to be included soon are surface tension, viscosity, density, thermal conductivity and heat capacity.