Development of an automatic estimation system for both the partition coefficient and aqueous solubility.

Area of Science:

• Computational chemistry
• Environmental science
• Chemical engineering

Background:

• Accurate estimation of physicochemical properties like partition coefficient and aqueous solubility is crucial for predicting chemical behavior.
• Existing computational methods require refinement for broader applicability and improved accuracy.

Purpose of the Study:

• To develop an extended computer program for estimating 1-octanol/water partition coefficients and aqueous solubility from chemical structures.
• To enhance the accuracy and reliability of computational predictions for these key chemical properties.

Main Methods:

• The program utilizes an extended version of the CHEMICALC program for partition coefficient estimation.
• Aqueous solubility is predicted via two pathways: a linear relationship with partition coefficients and a combined group contribution method.
• Revisions and extensions were applied to existing methods for aqueous solubility estimation.

Main Results:

• The study developed a computational tool for estimating both partition coefficients and aqueous solubility.
• The program integrates multiple estimation strategies, including linear regression and group contribution methods.
• The accuracy of the aqueous solubility predictions for a set of 497 compounds was evaluated.

Conclusions:

• The developed program offers a robust approach for predicting essential chemical properties from structural formulas.
• The combined methods enhance the predictive power for aqueous solubility across diverse compound types.
• This tool can aid in environmental risk assessment and chemical process design.