No description available.

Cavity enhanced spectroscopies (CES) are currently amongst the most sensitive spectroscopic techniques available for probing gas-phase samples, however their application to the liquid-phase has been more limited. Sensitive analysis of submicrolitre liquid samples is highly desirable, as miniaturisation allows for the reaction and analysis of scarce or expensive reagents, produces less waste, and can increase the speed of separations and reactions, whilst having a small footprint and high throughput. Absorption spectroscopy is a particularly desirable technique due to its universal, label-free nature, however its application to small volume liquid samples is hampered by the associated short absorption pathlengths, which limit sensitivity. CES improve sensitivity by trapping light within a confined region, increasing the effective pathlength through the sample. Three distinct types of optical cavity were constructed and evaluated for the purposes of making optical absorption measurements on liquid samples. The first incorporated a high optical quality flow cell into a "macrocavity" formed from two dielectric mirrors separated by 51.3 cm. Cavity losses were minimised by positioning the flow cell at Brewster's angle to the optical axis, and the setup was used to perform a single-wavelength cavity ringdown spectroscopy experiment to detect and quantify nitrite within aqueous samples. The detection limit was determined to be 8.83 nM nitrite in an illuminated volume of only 74.6 nL. Scattering and reflective losses from the flow cell surfaces were found to be the largest barrier to increased sensitivity, leading us to focus on the integration of cavity mirrors within a microfluidic flow system in the work that followed. In the second set of experiments, cavity enhanced absorption spectroscopy (CEAS) measurements were performed on Thymol Blue using custom-made microfluidic chips with integrated cavity mirrors. Unfortunately, due to the plane-parallel configuration of the mirrors and the corresponding difficulty in sustaining stable cavity modes, the results were underwhelming, with a maximum cavity enhancement factor (CEF) of only 2.68. At this point, attention was focussed toward a more well-defined cavity geometry: open-access plano-concave microcavities. The microcavities consist of an array of micron-scale concave mirrors opposed by a planar mirror, with a pathlength that is tunable to sub-nanometer precision using piezoelectric actuators. In contrast to the other experimental setups described, themicrocavities allow for optical measurements to be performed in which we monitor the change of wavelength and/or amplitude of a single well-defined cavity mode in response to a liquid sample introduced between the mirrors. In the first microcavity experiment, we used 10 μm diameter mirrors with cavity lengths from 2.238 μm to 10.318 μm to demonstrate refractive index sensing in glucose solutions with a limit of detection of 3.5 x 10^-4 RIU. The total volume of detection in our setup was 54 fL. Thus, at the limit of detection, the setup can detect the change of refractive index that results from the introduction of 900 zeptomoles (500,000 molecules) of glucose into the device. The microcavity sensor was then adapted to enable broadband absorption measurements of methylene blue via CEAS. By recording data simultaneously from multiple cavities of differing lengths, absorption data is obtained at a number of wavelengths. Using 10 μm diameter mirrors with cavity pathlengths from 476 nm to 728 nm, a limit of detection, expressed as minimum detectable absorption per unit pathlength, of 1.71 cm^-1 was achieved within a volume of 580 attolitres, corresponding to less than 2000 molecules within the mode volume of the cavity. Finally, a new prototype was developed with improved cavity finesse, a much more intense and stable light source, and improved flow design. Using a single plano-concave microcavity within the array with a cavity pathlength of 839.7 nm, and 4 μm radius of curvature mirror, absorption measurements were performed on Methylene Blue. Analysis of this data indicated a CEF of around 9270, and a limit of detection based on the measured signal-to-noise ratio of 0.0146 cm^-1. This corresponds to a minimum detectable concentration of 104 nM Methylene Blue, which given the mode volume of 219 aL, suggests a theoretical minimum detectable number of molecules of 14.