Caracterização de vidro para fins forenses utilizando fluorescência de raios X por reflexão total e técnicas espectroscópicas vibracionais

Abstract

In this work, simple and fast analytical methods were developed and/or evaluated, using minimal sample preparation for the characterization of glass for forensic purposes, using total reflection X-ray fluorescence (TXRF), attenuated total reflectance spectroscopy in the infrared average with Fourier transform (ATR-FTIR) and Raman spectroscopy in association with design of experiments, principal component analysis (PCA) and/or partial least squares discriminant analysis (PLSDA). The TXRF method was developed for the determination of 16 elements, using suspension as sample preparation. The Limit of Detection (LD) was between 4 and 10 mg kg-¹ for K, Ca, and Ba, while for Na, Mg and Al, the values were 2418, 752 and 313 mg kg-¹, respectively. For the other elements, the LD was less than 2.5 mg kg-¹. The relative standard deviation ranged from 3.1 to 12.0% under repeatability conditions. The trueness varied between 65 and 116% for Cr, Mn, Ni, Cu, Rb, Sr and Pb, while for Ti and Ba, this value was equal to 51.0 and 21.7%, respectively. The method was applied to 37 samples (smartphone screens, screen protector and automotive windshield). The PCA discriminated the original smartphone screens from the counterfeit samples. Furthermore, the windshields were distinguishable from the other samples. By conjoint ATR-FTIR spectroscopy and PLS-DA, it was possible to discriminate in a simple, fast, low-cost, and non-destructive way, the original smartphone screens from the counterfeit ones, with 100% sensitivity, specificity, and accuracy. Finally, a preliminary assessment was made of the performance of Raman spectroscopy for glass analysis, demonstrating the potential of portable instruments for this analysis. Using PCA, it was possible to identify natural clusters between the samples, detect outliers and obtain a breakdown of automotive and smartphone samples. Raman spectroscopy performed better than ATR-FTIR spectroscopy to distinguish between automotive and smartphone glasses. On the other hand, the distinction between genuine and counterfeit smartphone screens was less clear. However, the developed methods are attractive, with potential for forensic analysis.