American Chemical Society
Original Item ID
This work investigates a sensor system for direct groundwater monitoring, capable of aqueous-phase measurement of aromatic hydrocarbons at low concentrations (about 100 parts per billion (ppb)). The system is designed to speciate and quantify benzene, toluene, and ethylbenzene/xylenes (BTEX) in the presence of potential interferents. The system makes use of polymer-coated shear-horizontal surface acoustic wave devices and a signal processing method based on estimation theory, specifically a bank of extended Kalman filters (EKFs). This approach permits estimation of BTEX concentrations even from noisy data, well before the sensor response reaches equilibrium. To utilize estimation theory, an analytical model for the sensor response to step-changes, starting from clean water, to mixtures of multiple analytes is first formulated that makes use of both equilibrium frequency shifts and response times (for individual analyte), the latter being specific for each combination of coated device and analyte. The model is then transformed into state-space form, and the bank of EKFs is used to estimate BTEX concentrations in the presence of interferents from transient responses prior to attainment of equilibrium. Samples used in the experiments were either manually mixed in the laboratory or taken from real monitoring sites; they contained multiple chemically similar analytes with concentrations of individual BTEX compounds in the range of 10–2000 ppb. The estimated BTEX concentrations were compared to independent gas chromatography measurements and found to be in very good agreement (within about 5–10% accuracy), even when the sample contained multiple interferents such as larger aromatic compounds or aliphatic hydrocarbons.
Sothivelr, Karthick; Bender, Florian; Josse, Fabien; Ricco, Antonio J.; Yaz, Edwin E.; Mohler, Rachel E.; and Kolhatkar, Ravi, "Detection and Quantification of Aromatic Hydrocarbon Compounds in Water Using SH-SAW Sensors and Estimation-Theory-Based Signal Processing" (2016). Electrical and Computer Engineering Faculty Research and Publications. 727.
ADA Accessible Version