This thesis addresses the problem of chemical vapor detection by using an integrated optical-holographic interferometer for precision interrogation of a chemical transducer. A monolithic-prism interferometer, incorporating dynamic holography in conjunction with a \textquotedblleftsniff-locked-loop\textquotedblright sampling scheme, achieves high sensitivity at high speed while mitigating the deleterious effects of environmental noise and reducing costs associated with the replacement of expendable and poisonable transducers. The \textquotedblleftsniff-locked-loop\textquotedblright rapidly alternates vapor sampling (at 5 Hz) between a reference and test sample, enabling synchronous detection and precision differential measurements.
The prism geometry is designed to minimize interferometer sensitivity to laser wavelength variations. To mitigate other sources of optical and electronic noise, the relative phase between the beams is modulated, mixing the low-frequency signal to a sideband of 40 kHz, using a piezoelectric-driven grating modulator. The prism system exhibits a displacement sensitivity of 180 fm/Hz1/2, a signal-to-noise ratio of unity in a 1 Hz bandwidth. Displacement sensitivity translates into substance sensitivity depending on the transducer materials. One benchmark uses poly(N-vinyl pyrrolidone) as an ethanol sensor, with which we achieve a 3σ limit of detection (LOD) of 6 ppm (parts per million) in a 5s measurement time, or 1.8 ppm/Hz1/2.
A second portable chemical-vapor sensor system is presented. It is designed for spatially independent signal processing of a linear array in a portable path-length-balanced prism interferometer. A volume-hologram-stabilized 660 nm laser enables portability but requires further support in the form of a thermoelectric-cooling system. While losing nearly an order of magnitude of sensitivity, attributable almost entirely to the linewidth stability, this prototype successfully demonstrates the proof of concept for various feature additions.
In addition to system details, this thesis examines various noise sources and the solutions implemented to mitigate the noise contributions. Supporting equipment, including the gas valve, detection electronics, modulator components, and other designs, are presented where relevant. Presentations of several of the hurdles involving vapor delivery and concentration verification are also presented. Finally, a few general concepts necessary for an intimate understanding of the signal detection methodology are discussed. A summary of estimated noise sources is also given, followed by a performance comparison to several other transduction-based technologies.