BME PhD Defense: Lisa M. Bonanno
Advancing Porous Silicon Biosensor Technology for Use in Clinical Diagnostics
Supervised by Professor Lisa DeLouise
Abstract
Inexpensive and robust analytical techniques for detecting molecular recognition events are in great demand in healthcare, food safety, and environmental monitoring. Despite vast research in this area, challenges remain to develop practical biomolecular platforms that meet the rigorous demands of real-world applications. This includes maintaining low-cost devices that are sensitive and specific in complex test specimens, are stable after storage, have short assay time, and possess minimal complexity of instrumentation for readout. Nanostructured porous silicon (PSi) material has been identified as an ideal candidate towards achieving these goals and the past decade has seen diverse proof-of-principle studies developing optical-based sensing techniques.
Work presented in this thesis developed nanostructured porous silicon (PSi) 1-D photonic crystals for use as label-free optical biosensors beyond proof-of-concept and explored challenges of real-world use. Optimization of PSi morphology, surface chemistry, and assay protocol format facilitated improved detection sensitivity of small molecular weight targets (opiates, limit of detection = 5 nM) over a relevant clinical concentration range in urine and addressed stability concerns of sensor storage. Performance of the developed opiate assay (specificity and sensitivity) was then validated in a blind clinical study that screened real patient urine samples (n=70) for opiates in collaboration with Strong Memorial Hospital Clinical Toxicology Laboratory. Here, we demonstrate for the first time, successful clinical capability of a PSi sensor to detect opiates as a model target in real-world patient samples with improved clinical specificity over comparison commercial immunoassay techniques.
The final part of this thesis explores novel sensor designs to leverage the tunable optical properties of PSi and facilitate colorimetric readout of molecular recognition events by the unaided eye. Such a design is ideal for uncomplicated diagnostic screening at point-of-care as no instrumentation is needed for result readout. The photonic PSi transducers were integrated with target analyte-responsive hydrogels (TRAP-gels) that upon exposure to a target solution would swell and dissolute, inducing material property changes that were optically detected by the incorporated PSi transducer. Successful colorimetric detection was exhibited with the hybrid sensor design for a model chemical target (reducing agent) and for opiates in buffer solution as proof-of-concept.