Characterization of Novel Solvents and Absorbents for Chemical Separations

Predictive methods have been employed to characterize chemical separation mediums including solvents and absorbents. These studies included creating Abraham solvation parameter models for room-temperature ionic liquids (RTILs) utilizing novel ion-specific and group contribution methodologies, polydimethyl siloxane (PDMS) utilizing standard methodology, and the micelles cetyltrimethylammonium bromide (CTAB) and sodium dodecylsulfate (SDS) utilizing a combined experimental setup methodology with indicator variables. These predictive models allows for the characterization of both standard and new chemicals for use in chemical separations including gas chromatography (GC), solid phase microextraction (SPME), and micellar electrokinetic chromatography (MEKC). Gas-to-RTIL and water-to-RTIL predictive models were created with a standard deviation of 0.112 and 0.139 log units, respectively, for the ion-specific model and with a standard deviation of 0.155 and 0.177 log units, respectively, for the group contribution fragment method. Enthalpy of solvation for solutes dissolved into ionic liquids predictive models were created with ion-specific coefficients to within standard deviations of 1.7 kJ/mol. These models allow for the characterization of studied ionic liquids as well as prediction of solute-solvent properties of previously unstudied ionic liquids. Predictive models were created for the logarithm of solute's gas-to-fiber sorption and water-to-fiber sorption coefficient for polydimethyl siloxane for wet and dry conditions. These models were created to standard deviations of 0.198 and 0.122 logunits for gas-to-PDMS wet and dry, respectively, as well as 0.164 and 0.134 log units for water-to-PDMS wet and dry, respectively. These models are particularly useful in solid phase microextraction separations. Micelles were studied to create predictive models of the measured micelle-water partition coefficient as well as models of measured MEKC chromatographic retention factors for CTAB and SDS. The resultant predictive models were created with standard deviations of 0.190 log units for the logarithm of the mole fraction concentration of water-to-CTAB, 0.171 log units for the combined logarithms of both the mole fraction concentration of water-to-CTAB and measured MEKC chromatographic retention factors for CTAB, and 0.153 log units for the combined logarithms of both the mole fraction concentration of water-to-SDS and measured MEKC chromatographic retention factors for SDS