Date of Award
Doctor of Philosophy (PhD)
Electrical and Computer Engineering
Fabien J. Josse
Edwin E. Yaz
Susan C. Schneider, Chung Hoon Lee, Richard W. Cernosek
Organophosphate pesticides (OPs) have been found as contaminants in surface and ground waters, soil, and agricultural products. Because OPs are toxic compounds, rapid detection/monitoring of OPs in groundwater is necessary to allow for real-time remediation. Detection of OPs in water has already been demonstrated using poly(epichlorohydrin) [PECH] and polyurethane as the sensing layers. However, the response times were relatively long, hindering real-time monitoring. In this work, a hybrid organic/inorganic chemically sensitive layer [bisphenol A-hexamethyltrisiloxane (BPA-HMTS)] that shows a high degree of partial selectivity for OPs is synthesized, characterized (in terms of the glass transition temperature, Tg, water stability, sensitivity, selectivity, detection limit, and absorption/response time) for the rapid detection of organophosphate pesticides. Direct chemical sensing in aqueous solutions is performed using guided shear horizontal surface acoustic wave sensor platforms on 36° rotated Y-cut LiTaO3 and 42.75° rotated Y-cut Quartz, respectively. It is shown that, for the same coating thickness, a 60% reduction in sensor response time is achieved without reduction in sensitivity compared to PECH. Considering the Tg, for the polymers, it is seen that the faster response shown by BPA-HMTS is due to the porous siloxane backbone, HMTS. Kinetic studies for the absorption of OPs (parathion-methyl, parathion, and paraoxon) from aqueous solutions into the BPA-HMTS coating are conducted. The data are analyzed within the context of two absorption models: penetration-limited and diffusion-limited absorptions. It is shown that the absorption process is rate limited by penetration with a concentration independent absorption time constant or mass transfer coefficient. The absorption time constants for parathion-methyl, parathion, and paraoxon are calculated. A limit of detection of 60, 20 and 100 μg/L (ppb) for parathion-methyl, parathion, and paraoxon, respectively, is calculated for the present non-optimized sensor. Concentrations as low as 500 μg/L (ppb) parathion are actually measured. This is much lower than the typical concentrations found on agricultural produce (≥ 10 ppm).
Furthermore, sensor signal analysis in the form of the extended Kalman filter (EKF) is employed on-line during the detection process. The sensor response was first represented by a state-space model which includes all relevant contributions to the polymer-coated device response. This allows for the steady-state response and absorption time constant to be extracted on-line well before the steady-state is reached, thus further reducing the time for analyte identification and quantification. It is noted the absorption time constant, often unique to a class of analyte-coating pairs, can be used to improve analyte recognition.