Distinguishing binary mixtures of VOCs by pattern recognition of SAW array responses
- Sandia National Labs., Albuquerque, NM (United States); and others
Surface acoustic wave (SAW) devices coated with self-assembled monolayer and plasma-grafted polymer thin films can detect a variety of volatile organic compounds (VOCs). Distinguishing individual VOCs and VOC mixtures with arrays of these devices requires: robust pattern recognition (PR) of the nonlinear and nonadditive array responses; optimizing the choice of SAW films for VOC PR. We describe a cluster-based PR method and a selected set of six SAW device coatings which distinguish water, 6 VOCs (acetone, DIMP, i-octane, n-propanol, TCE, toluene) and their 21 binary combinations over wide concentration ranges. Recognition robustness is assessed by artificially imposing up to 16% reductions (mimicking drifts) in sensitivities of the SAW responses. Correct chemical recognition occurred for over 95% of the measured and artificial-drift data.
- OSTI ID:
- 126260
- Report Number(s):
- CONF-950402--
- Country of Publication:
- United States
- Language:
- English
Similar Records
Clustering-based pattern recognition applied to chemical recognition using SAW array signals
SAW chemical sensor arrays using new thin-film materials