Optimal path length identification for accurate glucose sensing with photoacoustic derived optical rotation

Non-invasive glucose monitoring is crucial for diabetes management. This study explores the use of photoacoustic (PA) signals based on optical rotation estimation at multiple depths for detection of glucose concentrations. Experiments were performed with glucose samples mixed in bovine serum albumin with different polarization incidences-vertical (V), 45° linear (P), and right circular (R) polarization. Polarized Monte Carlo (PMC) simulations were performed to understand the depth-dependent behavior between optical and photoacoustic detection of optical rotation, which allows the estimate of glucose concentration. Notably, a specific depth range exhibited both maximum rotation and a better linear relationship with concentration, which are ideal for sensing. Both experimental and simulation studies indicated significant depolarization beyond a depth of 4 mm. Additionally, the change in rotation with respect to depth (Δα) was higher for larger concentration differences compared to smaller concentration differences. Our study identified that the optimal depth for accurate glucose sensing (based on Clarke's error grid (CEG)) was found to be around 3-3.2 mm for the different polarized incidences. These findings showcase the potential of our approach for non-invasive glucose sensing and a calibration procedure to pinpoint optimal sensing depths, extendable to other chiral molecules.