|dc.description.abstract||Polarimetry is an useful tool in chemical identification and characterisation both inside and outside the lab. It is a modality where the effects of chiral molecules, water impurities and cancerous cells are noticeable through changes in light polarisation as it passes through these samples. In the case of chiral molecules, like glucose, this change in polarisation manifests as a rotation in polarisation. The amount of rotation that occurs is governed by Boit’s Law; however, without large optical path lengths, or high concentrations of solution, these rotations are typically very small, requiring elaborate, large and costly apparatus.
These devices ensure accuracy by performing complex optical procedures or time-averaging point measurements. This ensures that any intensity variation in the measurements is a result of rotation and not from inherent error in the system or from noise, such as sensor or shot noise. Time Averaging is a lengthy process and does not utilizes all the information from the incident light beam. To this end, we propose, design and build a novel inexpensive, compact computational polarimetric measurement system. This system computationally enhances polarimetric measurements by utilising the full spot size observed on a spatial detector array. This allows the system to need only a single acquisition to enhance the polarimetric measurements by recognizing that the full CMOS detector is a spatial array of photosensitive areas. By using the full spot, it can mimic temporal acquisitions, but through space rather than time. To ensure accuracy in computational enhancement, detector noise and system error are characterized for their effect on both pixel intensity and polarimetric measurement accuracy, given the entire acquisition system. All the while, this device achieves a spatial footprint of less than 245 cm<sup>3</sup> and costs 68\% less than the state-of-the-art lab polarimeters.
Two experiments are performed to validate this system. The first experiment is a synthetic experiment meant to demonstrate the performance and robustness of this device in determining the concentration of chiral molecules in a solution. The second experiment validates the real life capability of this system by determining the concentration of Maltodextrin in water. Through these experiments, it is shown that the spatial enhancement methods are capable of improving the estimate of chiral molecule concentrations. We also demonstrate that through the use of a model based spatial enhancement method, we can estimate with more accuracy and precision than the state-of-the-art enhancement method of Temporal Averaging.
This system demonstrates that an inexpensive, compact polarimetry device can report accurate measurements of chiral molecule concentrations through <em>a priori</em> knowledge of the imperfections in the optical elements used and through computational enhancement methods that utilise the entire light beam's spot incident on a spatial detector array. This system acts as a proof of concept that the polarimetry modality can be taken outside the lab and into the field for chiral molecule concentration identification. It also demonstrates that spatial enhancement methods can be used for polarimetric measurement enhancement, potentially reducing overall acquisition time which is advantageous in dynamic conditions.||en