Characterization of polymeric surface acoustic wave sensor coatings and semiempirical models of sensor responses to organic vapors.

Responses from an array of four polymer-coated surface acoustic wave sensors exposed to a series of 39 organic vapors were used to investigate sensor response models based on vapor boiling point, solubility parameters, and solvation parameters in conjunction with linear solvation energy relationships. As part of this effort, sensor response data were used to estimate the solubility parameters and solvation parameters of the sensor coatings by adaptation of methods originally developed for use with gas-liquid chromatographic retention data. Values of these parameters were found to be consistent with the structures of the coatings though in some cases different from those determined by other methods. Discrepancies were attributed to differences in the conditions used for the determinations. Sensor responses were linear over the concentration ranges examined and could be summarized using the empirically determined partition coefficient, Ke, for each vapor-coating pair. Linear correlations were found between log Ke and vapor boiling point, and the slopes of the regressions lines were similar to those expected for ideal behavior. The strength of the correlations decreased with increasing coating polarity, and it was necessary to divide the vapors into two or three broad chemical classes in order to obtain satisfactory results. Improved correlations were found by use of Hildebrand solubility parameters in a model based on regular solution theory which attempts to account for nonideal vapor-coating interactions. The use of solvation parameters in linear solvation energy relationships, however, provided the strongest correlations, with modeled K values falling within a factor of 2 of experimental values in all cases and within +/- 25% of experimental values in 83% of the cases. Application of these models to the prediction of sensor array response patterns appears promising.