| The linear solvation energy relationships (LSER) model proposed by Kamlet, Abraham, and Taft allows the prediction of a variety of solubility interactions based on a set of descriptors found in this equation:; logSP=c+r1R2 +s1p*H2+ a1SaH2+ b1SbH2+ I1logL16 The terms on the right side of the equation represent different interactions of the stationary phase (subscript 1) with the solute (subscript 2). GC stationary phases can be characterized by determining these coefficients to quantify the ability of the solvent to engage in specific interactions. Once these coefficients have been determined, a correlation between the structure of the stationary phase and the coefficient values can be established. Information obtained can be used to predict solubility relationships for stationary phases that have never been characterized. The ultimate goal of this project is to develop predictive models for solubility of materials based on structural descriptors such as molecular connectivity indexes.; Stationary phases with functional groups containing third quantum-level elements, such as sulfur and phosphorus, are underrepresented in current databases. Several of these compounds were analyzed by inverse gas-liquid chromatography (GLC) to find their LSER coefficients. Both experimentally obtained and previously determined coefficients were correlated with calculated polarizability, dipole moment, HOMO/LUMO energy levels, and hydrogen bond acceptor capability. Quantitative structure-solubility relationships (QSSR), correlating molecular connectivity or combinations of connectivity with other structural descriptors to LSER coefficients, yielded promising relationships. Alternative methods for calculation of phosphorus connectivity indices were developed to better describe the contributions of inner shell and valence electrons to chemical and physical properties. Modified connectivity values gave improved correlation with some LSER coefficients and were used to develop similar QSSRs. |
