Multicomponent analysis and prediction with a cantilever array based gas sensor

We present micromechanical cantilever sensor arrays for quantification of individual components in a gas mixture. Using selectively coated cantilevers as mass sensitive transducers, their response to the analyte molecules is obtained by measuring their resonant frequency shift due to the additional mass loading. The responses were used in principal component regression (partial least squares (PLS)) for the quantitative prediction of vapors being analyzed. The results confirm the functionality of our sensor system and motivate for further optimization. (C) 2001 Elsevier Science B.V. All rights reserved.