Results of on-line process control with an SAW-sensor based gas-sensor micr
osystem (SAGAS PC) are presented for industrial applications. The control o
f the solvent mixer of an HPLC-system was chosen as a typical example. All
applications dealt with the detection and subsequent qualification as well
as quantification of volatile organic compounds (VOC's), e.g. organic solve
nts or even water by investigating their vapor. Mass-sensitive transducers,
especially surface acoustic wave (SAW)-devices, have become a feasible too
l for VOC-analytics. The measurement principle is based on frequency change
s which are directly correlated with mass changes on the sensor surface and
thus provide a particularly suitable signal transduction. The data analysi
s of all applications is based on pattern recognition methods, and focused
on artificial neural networks for quantitative determination of these solve
nts and solvent-mixtures. The results obtained clearly showed the suitabili
ty of such a sensor system for on-line process-control purposes.