MULTIWAY ANALYSIS OF PRECONCENTRATOR-SAMPLED SURFACE-ACOUSTIC-WAVE CHEMICAL SENSOR ARRAY DATA

New data processing methods for preconcentrator-sampled surface acoust ic wave (SAW) sensor arrays are described. The preconcentrator-samplin g procedure is used to collect and concentrate analyte vapors on a por ous solid sorbent, Subsequent thermal desorption provides a crude chro matographic separation of the collected vapors prior to exposure to th e SAW array. This article describes experiments to test the effects of incorporating retention information into the pattern-recognition proc edures and to explore the feasibility of multiway classification metho ds. Linear discriminant analysis (LDA) and nearest-neighbor (NN) patte rn-recognition models are built to discriminate between SAW sensor arr ay data for four toxic organophosphorus chemical agent vapors and one agent simulant collected under a wide variety of conditions. Classific ation results are obtained for three types of patterns: (a)first-order patterns; (b) first-order patterns augmented with the time of the lar gest peak; and (c) second-order patterns with the use of the SAW frequ ency for each sensor over a broad time window. Classification models f or the second-order patterns are also developed with the use of unfold ed and multiway partial least-squares discriminants (uPLSD and mPLSD) and NN and LDA of the scores from unfolded and multiway principal-comp onent analysis (uPCA and mPCA), It is determined that classification p erformance improves when information about the desorption time is incl uded, Treating the preconcentrator-sampled SAW sensor array as a secon d-order analytical instrument and using a classification model based u pon either uPLSD, uPCA-LDA, or NN results in the correct identificatio n of 100% of the patterns in the prediction set. With the second-order patterns, the other pattern-recognition algorithms only do slightly w orse. (C) 1998 John Wiley & Sons, Inc.