Scattering of monochromatic X-rays at different incident radiation angles provides quantitative information on physical and chemical properties of plastics

X-ray fluorescence spectroscopy (XRF) is a powerful tool of elemental analysis; however in most of the experimental set-ups it does not allow quantification of the light elements (with atomic number below 11). The use of scattering X-ray radiation as a source of useful analytical information is getting more and more popular in X-ray studies. The common trend in this field is in using the standard XRF instrumentation, where polychromatic incident X-ray radiation and fixed geometry are employed. In this study we explore the opportunity of obtaining valuable physical and chemical information on plastic samples using monochromatic radiation and varying incident radiation angles. The use of machine learning techniques for multivariate regression modeling of scattering radiation spectra allows quantification of certain physical and chemical properties in commercial plastic samples which are normally not available from standard XRF measurements.

Automated summary

X-ray fluorescence spectroscopy is a powerful tool for elemental analysis, but it traditionally cannot quantify light elements. The use of scattering X-ray radiation is becoming popular in X-ray studies, with the trend being towards standard XRF instrumentation. This study explores the use of monochromatic radiation and varying incident angles to obtain physical and chemical information on plastic samples, utilizing machine learning techniques for regression modeling.