| The goal of this dissertation is to describe the operation of a multiple analyte fluorescence immunosensor based on a silicon oxynitride integrated optical waveguide (IOW). The study is divided into three categories: sensor characterization, multiple analyte immunoassay, and rapid kinetic analysis. Sensor characterization involved measuring the efficiency of rectangular diffraction grating couplers, the intrinsic loss and scatter of light from the waveguide, the linearity of the transducer, and the light gathering potential of the detection system. Multiple analyte immunoassays consisted of constructing a series of calibration curves for three sets of three analytes. One set was made of three polyclonal IgG antibodies; another, monoclonal antibodies to bovine serum albumin, IgG, and avidin; and the last, monoclonal antibodies to three cardiac analytes, troponin I, creatin kinase, and myoglobin. In all cases, fluorescence, sandwich style multiple analyte endpoint assays were compared to identical assays where only a single analyte was measured at a time. Kinetics analyses were then considered as a way to decrease overall assay time, a crucial aspect of point-of-care biosensors, and as a way to deduce the affinity constants for some of the antibody-antigen reactions. An empirical model was applied to the kinetic, cardiac analyte assay data to study how the analytical sensitivity was affected by shortening incubation times from 5 to 2.5 minutes, and then 1 minute. A linearization of the bimolecular model allowed for the kinetic determination of forward and reverse rate Constants of the reactions for bovine serum albumin, IgG, and avidin and were compared with values from the static endpoint assays. A summary of results suggests that the grating coupled IOW can be used for multiple analyte sensing. The specially designed gratings demonstrated adequate coupling efficiency (≈30% incident intensity) while the composition of the thin guiding film was sufficiently loss-less (<0.5 dB/cm). The transducer reliably sensed less than a monolayer coverage (0.7%) for a fluorescently labeled protein adsorbed to the waveguide surface, however, it was not completely linear over the 0.7%--100% coverage range, most likely due to surface chemistry inhomogeneities. Of the three multiple analyte systems tested, the polyclonal antibodies for the IgG analytes showed too much cross reactivity and non-specific binding to be useful in a multiple analyte immunassay. Both monoclonal antibody systems, however, demonstrated nearly one-to-one-correlation between multiple and single analyte immunoassay formats. Results from the rapid, empirical kinetic analyses showed that shortening incubation time did not result in a consistent change in analytical sensitivity for the cardiac analytes. Limits of detection between kinetic and endpoint models were also not consistently different. Affinity constants from the kinetic analysis were on the same order of magnitude as those listed in literature but nearly an order of magnitude different from those calculated by the endpoint method. These results suggest that the biosensor capture layer, which contains the antibodies, is far from being in equilibrium with the multiple analytes. Future studies will attempt to refine the grating and waveguide designs to enhance sensor detection, study the sensor response to unknown sample concentrations (blind testing), and improve the surface chemistry and functionality of antibody capture layers through engineered immobilization, such as photodeposition. |
