Determination of solubility of phenanthrene in binary and ternary solvent mixtures and application of the Jouyban-Acree model for solubility prediction of chemical/drug compounds in mixed solvents

Abstract

Solubility of a compound in a solvent is an important thermodynamic property in chemical and pharmaceutical sciences. Its major effects can be seen in chemical and pharmaceutical related properties and processes such as synthesis, extraction, purification, crystallization, analysis, formulation, absorption, distribution, metabolism, elimination, and toxicity. Different methods of solubilization are applied to modify this property such as altering temperature, cosolvency, micelization, and complexation. Cosolvency and temperature altering are most accessible methods to modify solubility amount. Numerical methods have been introduced during recent decades; some of these calculative methods were structured, based on only empiric data correlation, some based on empiric data and theoretical models, and some other are based on theoretical calculations. For example, the linear solvation energy relationship (LSER) quantitative structure property relationships (QSPRs) of Abraham were developed based on solvatochromic properties of solvents and solutes and are used for solubility prediction in mono-solvents. Catalan solvent parameters are another set of solvatochromic properties which have been measured for more than 200 solvents. The Jouyban-Acree model is the most accurate semi-empirical model for solubility prediction in non-aqueous binary or multi-component solvents mixtures. In an attempt, coefficients of the Jouyban-Acree model for non-aqueous binary mixtures have been described using Abraham solvent and solute parameters. Phenanthrene is a polycyclic aromatic hydrocarbon (PAH) which can be used as a substrate for synthesis of medicinal compounds and is a model solute in solubility studies. As a part of QSPR modeling of this work, we are going to define Catalan solute parameters based on solubility amounts of a solvent in more than 8 solvents with known Catalan solvent parameters, using a multiple linear regression approach. The solubility data of phenanthrene in alcohols at different temperatures is measured. Another part of this study includes measurement of experimental solubility data for phenanthrene in different non-aqueous binary and ternary solvent mixtures and checking the predictability of the Jouyban-Acree model combined with Abraham parameters. To do this, four different numerical methods have been described. Using experimental solubility data in mono-solvents, numerical method I and II have been defined employing water-to-solvent and gas-to-solvent Abraham parameters, respectively. Predicted solubility amounts using Abraham parameters in mono-solvents and numerical method III and IV have been defined employing water-to-solvent and gas-to-solvent Abraham parameters, respectively. The mean percentage deviation has been used as an error criterion. The overall mean percentage deviations are 3.4 %, 14.6 %, 11.1 %, and 93.8 % for numerical methods I to IV, respectively. In all parts of experiment, phenanthrene has been recycled and reused. The results of this study suggest that the Jouyban-Acree model can be applied for solubility prediction in non-aqueous binary and ternary solvent mixtures with acceptable prediction error. To make it easy-to-use for chemical and pharmaceutical scientists and researchers, developing computer software might be a useful method.