Strategies for Enhancing the Analytical Performance and Biocompatibility of Intravascular Amperometric Glucose Sensors

Ensuring accurate analytical performance that can overcome the body's response to foreign objects is integral to developing improved in vivo biosensors. Blood contact from intravenous sensor placement causes surface clotting, and blood-born electroactive species obscure desired signal response. In this thesis, amperometric glucose sensor selectivity improvements are combined with surface hemocompatibility augmentation strategies to enhance in vivo analytical performance.;Annealing thin films of Nafion causes morphological changes to hydrophilic regions, altering transport properties. A novel annealed Nafion layer on the electrode cavity of needle-type glucose sensors enhances selectivity (<3 % of signal response for 5 mM glucose at typical levels of ascorbic and uric acid in blood) for at least 8 days.;The transmembrane CD47 protein is ubiquitously expressed as a "self" marker. A "handshake" interaction with its cognate receptor on inflammatory response cells (SIRPalpha) suppresses platelet activation. Outer polyurethane coatings of glucose sensors were functionalized with 85.0 ng/cm2 of recombinant human CD47. In vitro calibrations confirm that immobilization has no adverse impact on sensor glucose response properties.;Nitric oxide (NO) is a potent antithrombotic molecule, endogenously released by endothelial cells to prevent platelet activation and clot formation. Layers of poly(lactic acid) containing diazeniumdiolated-N,N'-dibutyl-1,6-hexanediamine (DBHD/N2O2) grant thromboresistance to glucose sensor surfaces by mimicking endogenous NO release at a localized flux >0.5x10 -10 mol cm-2 min-1 for at least 7 d. Intravascular amperometric glucose sensors prepared with NO release coatings accurately trace modulations of in vivo glucose concentration within rabbit veins for 7 h. Preliminary in vivo glucose measurements for 20 h porcine veins are also examined, and extensive in vitro investigations conducted to identify sources of significant sensor calibration drift in the porcine model.;Finally, a commercial glucose sensor system is configured for transient NO release. Preliminary data and diffusion modeling suggest that silicone rubber catheter sensor housing can be loaded with S-nitroso- N-acetylpenicillamine (SNAP). NO released from the catheter partitions uniformly into the thin outer polyurethane sensor coating at a projected concentration of 6.77x10-5 M during a 5 min charging phase, enabling transient NO release >0.5x10-10 mol cm-2 min -1 for >5 min after sensor removal from the NO source, simulating in vivo blood contact.