Research on concentration detection method for turbid solutions using multidimensional spectroscopy with a parabolic...

To address the difficulty of effectively utilizing scattering information in conventional optical methods for turbid solution concentration detection, this study presents a parabolic sample cell enabling rapid simultaneous acquisition of multidimensional hyperspectral data. The parabolic incident surface produces non-cylindrically symmetric transmission spots, capturing richer spectral information while the variable curvature concentrates the spot and suppresses stray light. The structure also minimizes wall-reflection stray light from oblique incidence, ensuring high purity of the diffuse reflection signal. Monte Carlo simulations identified the parabolic cell with coefficient a = 0.04 as optimal. Partial least squares regression models established between intensity distribution data and the reduced scattering coefficient show that the parabolic cell achieves superior prediction accuracy over flat and hemispherical structures under both transmission and diffuse reflection conditions. Experimental validation using 60 Intralipid-20% samples demonstrates that the two-dimensional model based on the parabolic cell improves the test set coefficient of determination (Rₚ²) by 1.11% over the hemispherical model, while the spatially resolved diffuse reflectance curve achieves a test set Rₚ² of 0.9861. Thus, the proposed method significantly enhances detection accuracy, offering a promising approach for high-precision analysis of turbid solutions.