Identification of natural spring water using shear horizontal SAW based sensors

The acoustoelectric interaction between a shear horizontally polarised surface acoustic wave (SH-SAW) sensor and different commercial brands of natural still spring water is investigated. Transient frequency responses for the samples are presented where the sensing mechanism is based upon perturbation of the electrical properties of the adjacent liquid. SH-SAW devices were fabricated on 36 degrees rotated Y-cut, X propagating LiTaO3 and implemented in an oscillator configuration. Modelling and simulation of the SH-SAW sensor/liquid interface is achieved using a high frequency electronic simulation software package. Theoretical and experimental results are compared and show good agreement and support the validity of the SH-SAW sensor model. Pattern recognition using a neural network algorithm is applied for the discrimination between the four types of sample liquids, with an initial recognition probability of 90%. (C) 1998 Elsevier Science S.A. All rights reserved.