Estimation of laser spot size in Raman micro-imaging and improved particle size estimation of small microplastics

Abstract

In this study, we developed a method to estimate laser spot size during Raman micro-imaging by integrating experimental data from standard microplastic samples with a simulation model that characterizes interactions between the laser spot and microplastic particles. The experimental data were also used to examine the relationships among correlation coefficient between sample spectra and standard spectrum (CFF), standard deviation of spectral noise (σ_ns), signal-to-noise ratio (SNR), and their spatial properties. Analysis of the micro-imaging data shows that CFF, σ_ns, and SNR are related to the presence of microplastic particles. Specifically, when a measurement point detects a particle, the values of CFF, σ_ns, and SNR are higher than those in the background filter. Further analysis of CFF reveals that its values are spatially dependent, showing a notable pattern within the effective range that is similar to, or slightly exceeds, the particle size. Additionally, CFF values decrease with increasing distance from the particle center, a trend that can be described by a logistic function. By applying this dataset, we determined the laser spot size in our setup to be 65 μm, allowing for a 24 % - 74 % improvement in particle size estimation accuracy, as measured by the root-mean-square-error. This study highlights the important role of laser spot size in Raman micro-imaging analysis and provides a robust methodology that can be adapted to other instruments and micro-imaging techniques.