A predictive PC-SAFT EOS based on COSMO for pharmaceutical compounds.

The present study was conducted to develop a predictive type of PC-SAFT EOS by incorporating the COSMO computations. With the proposed model, the physical adjustable inputs to PC-SAFT EOS were determined from the suggested correlations with dependency to COSMO computation results. Afterwards, we tested the reliability of the proposed predictive PC-SAFT EOS by modeling the solubility data of certain pharmaceutical compounds in pure and mixed solvents and their octanol/water partition coefficients. The obtained RMSE based on logarithmic scale for the predictive PC-SAFT EOS was 1.435 for all of the solubility calculations. The reported values (1.435) had a lower value than RMSE for COSMO-SAC model (4.385), which is the same as that for RMSE for COSMO-RS model (1.412). The standard RMSE for octanol/water partition coefficient of the investigated pharmaceutical compounds was estimated to be 1.515.