The solvation parameter model is used to determine the system constants for three columns containing mixtures of poly(dimethylsiloxane) and poly(ethylene glycol) and a poly(cyanopropylphenyldimethylsiloxane) containing 6% of cyanopropylphenylsiloxane monomer at five equally spaced temperatures in the range 60–140°C. Together with literature data for a poly(dimethylsiloxane) and a poly(ethylene glycol) stationary phase the influence of temperature and composition on selectivity is studied for mixing ratios of 0 to 1 poly(ethylene glycol) for the temperature range 60–140°C. Using literature data for two poly(cyanopropylphenyldimethylsiloxane) stationary phases containing 14% and 50% of cyanopropylphenylsiloxane monomer groups the influence of temperature and replacing dimethylsiloxane monomer groups by cyanopropylphenylsiloxane groups on selectivity is studied for incorporation of 0 to 0.5 cyanopropylphenylsiloxane groups over the temperature range 60–140°C. Addition of poly(ethylene glycol) or introduction of cyanopropylphenylsiloxane monomer groups into a poly(dimethylsiloxane) influences selectivity through an increase in dipolarity/polarizability, hydrogen-bond basicity, electron lone pair interactions, and changes in cohesion. The changes in system constants as a function of temperature and composition are simply modeled as smooth quadratic response surfaces. Curvature in the response surfaces along the composition axis is significant while changes along the temperature axis are modest for both stationary phase types. Cluster analysis is used to demonstrate that the mixed poly(dimethylsiloxane)/poly(ethylene glycol) stationary phases containing 0.5 and 0.85 weight fraction of poly(ethylene glycol) have different selectivity to a database of common open-tubular column stationary phases. The mixed poly(dimethylsiloxane)/poly(ethylene glycol) stationary phase containing 0.10 weight fraction of poly(ethylene glycol) has similar selectivity to the poly(cyanopropylphenyldimethylsiloxane) containing 6% cyanopropylphenyl monomer groups, and could replace the mixed phase for all but the most critical of separations.
Abstract The solvation parameter model is used to identify contributions from intermolecular interactions responsible for non‐specific retention in gas chromatography for three dissolved β‐cyclodextrin derivatives in a poly(cyanopropylphenyldimethylsiloxane) stationary phase. The cyclodextrins are permethylated β‐cyclodextrin (Cyclodex‐B), heptakis(2,3‐di‐ O ‐methyl‐6‐ O ‐ t ‐butyldimethylsilyl)‐β‐cyclodextrin (CycloSil‐B) and heptakis(2,3‐di‐ O ‐acetoxy‐6‐ O ‐ t ‐butyldimethylsilyl)‐β‐cyclodextrin (Rt‐βDEXsa). Taking DB‐1701 as a reference phase for the poly(cyanopropylphenyldimethylsiloxane) solvent, it is shown that the dominant interactions for the cyclodextrin derivatives are associated with their hydrogen‐bond basicity and capacity for dipole‐type interactions. None of the cyclodextrin derivatives are hydrogen‐bond acids and all are weakly electron lone pair repulsive. The cohesive properties of the dissolved phases are similar to those of the solvent, except for Rt‐βDEXsa, which is significantly more cohesive. Also, Rt‐βDEXsa shows significant inclusion complexation for the compounds used to determine the system constants of the solvation parameter model resulting in poor statistical models, suitable only for qualitative interpretation. The Cyclodex‐B and CycloSil‐B columns are compared to a database of 23 open‐tubular column stationary phases possessing similar selectivity to each other but different selectivity for non‐specific interactions to the other stationary phase types.