Flow-enhanced detection of biological pathogens using piezoelectric microcantilever arrays

The piezoelectric microcantilever sensor (PEMS) is an all-electrical resonant oscillator biosensor system capable of in-situ and label-free detection. Immobilized receptors on the sensor surface enable real-time electrical measurement of the resonance frequency shift due to the binding of target antigens to the surface. With silane-based insulation methods and bifunctional linker antibody immobilization schemes, it is wellsuited for applications in sensitive, specific detection of biological pathogens with limits of detection on the order of relevant lethal infectious dosage concentrations. Initial PEMS implementation demonstrated biodetection of Bacillus anthracis (BA) spores at a concentration of just 36 total spores in 0.8 mL of liquid. While these results are exciting, concerns that cross reactivity between the antibody and closely related species of the target pathogens cast doubts on the usefulness of any antibody-based assays in terms of the specificity of pathogen detection. The goal of this dissertation is to develop the PEMS biosensor as a viable antibody-based assay for in-situ, label-free, water-borne pathogen detection with better limits of detection than current antibody-based methods as well as high sensitivity and specificity, by exploring array PEMS detection and specificity augmentation by engineered fluidics. In the detection of BA spores, controlled fluid flow experiments in an 8 mm wide flow channel at flow rates ranging from 0 to 14 mL/min led to a determination of optimal flow rates for discriminatory detection of BA spores among close cousins: B. cereus (BC), B. thuringiensis (BT) and B. subtilis (BS). It is shown that the detection signal of all such spores first increased with an increasing flow rate. The detection signals of BC, BT and BS eventually diminished with the increasing flow rate as the force of the flow overcame the interaction force of the BC, BT, and BS spores with the sensor surface. The optimal flow rate was determined to be 14 mL/min at which detection signals of BC, BT, and BS all fell to within the noise level of the sensor, while the detection BA was still nearly optimal. As a result, it was deduced that the interaction forces of BC, BT, and BS were about 100 pN. Design and implementation of array sensing systems enabled real-time simultaneous redundant biosensor assays and concurrent background determination by a reference PEMS. By virtue of this advance in PEMS technology, successful real-time detection of just 10 BA spores/mL was achieved and step-wise, single Cryptosporidium parvum (CP) oocyst detection at 0.1 oocysts/mL was accomplished with resonance frequency step-wise shifts of 290 Hz and signal to noise ratios greater than 5 per instance of oocyst detection. It was found that, in a 19 mm wide flow channel, optimal single oocyst detection efficiency was achieved at 2 mL/min, while optimal discrimination of CP from C. muris (CM) oocysts was achieved at 5 mL/min. At this flow rate the detection signal of CP was close to optimal with a signal to noise ratio of 5 per step-wise shift and that of CM was close to the noise level. The interaction forces of CP and CM oocysts with the biosensor surfaces were deduced to be 110 and 70 pN, respectively.