Classification and discrimination of automotive glass using LA-ICP-MS

Abstract

Glass is one of the most common types of trace evidence found at crime scenes and on suspects. The elemental analysis of float glass has become increasingly important as the range in values of refractive index has substantially narrowed due to increased quality control in glass manufacture. The purpose of this research was to investigate the variation of elements in automotive float glass relevant to New Zealand for the purpose of classifying and discriminating samples. Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) was used to analyse the elemental composition of 243 automotive glass samples. An intact side window (tempered pane) and an intact windscreen (laminated pane) were analysed to investigate the spatial trend of elements in automotive glass. Most elements displayed no spatial variation over the panes. Pb had the largest variation in the windscreen and was also found to have a large variation in the database. Most samples were able to be classified into the vehicles' country of origin using a multiclass classifier. However, this was not possible for all samples, due to the origin of glass differing from the origin of the vehicle in some cases. The elemental composition of Australian and Northern Hemisphere samples differed greatly making them easy to separate; however, there was little variation within the Australian samples, making it hard to discriminate between different samples. A three step method, which combined the use of elemental composition, DRI and RI, was used to discriminate database samples. The method distinguished 84% (172/204) of samples in the database. When Australian samples were removed from the analysis, the discrimination increased to 95% (148/156). The type II errors were reduced by using both elemental composition and RI measurement.

Discrimination of side‐window glass of Korean autos by laser ablation inductively coupled plasma mass spectrometry

“…(f)), and that the LREEs in glass from HAN can be differentiated according to glass thickness. In addition, the E s /E R ratios of LREEs can be used as a forensic tool to discriminate between vehicles from the USA and South Korea, in agreement with a previous study . However, our results indicate that LREEs cannot be used to clearly discriminate among samples from different car manufacturers.…”

“…Previous studies have used various statistical methods (e.g., Tukey's honestly significant difference (HSD) test, principal component analysis, and discriminant analysis) to classify and discriminate types of glass from different countries and vehicle models, based on elemental percentages. [10,11,14,23] As LDA has been used in numerous studies to classify samples (e.g., oil, food, and bird) according to origin, [24][25][26] we also applied this method. Prior to carrying out each LDA, meaningful variables were extracted through ANOVA.…”

“…Therefore, recent studies have used elemental analysis as an alternative discrimination tool. [2,[8][9][10][11][12][13] For example, windowpanes from the USA and Australia were clearly differentiated based on 15 elements measured by inductively coupled plasma mass spectrometry (ICP-MS), [13] and the origins of intact side windows and windscreens were identified by elemental composition using laser ablation (LA)-ICP-MS. [14] Many other techniques for analyzing elements in glass have been tested, including X-ray fluorescence (XRF) and scanning electron microscopy/energy dispersive spectroscopy (SEM/ EDS), [5,13,15,16] both of which has their advantages and disadvantages. [14,17] However, LA-ICP-MS is considered the most efficient technique for analyzing small fragments.…”

Characterisation and classification of automotive clear coats with Raman spectroscopy and chemometrics for forensic purposes

“…Chemometrics has been previously utilised in forensic science in order to mitigate bias and partiality in the interpretation of analytical data obtained from trace evidence. [7][8][9] A combination of principal component analysis (PCA) and linear discriminant analysis (LDA) has been applied in the discrimination and classification of forensic evidence, including fibres, [10][11][12] glass, [13,14] inks, [15,16] soil, [17,18] hair, [19] accelerants, [20][21][22] photocopy and printer toners, [23][24][25][26] paper, [27] electrical tapes [28,29] and paint. [30][31][32][33][34] Specifically, chemometrics has also been shown to be a viable technique for the interpretation of automotive paint evidence.…”

Data Mining Methods Applied to a Digital Forensics Task for Supervised Machine Learning

“…In the context of glass microtraces, Zadora et al [12] proposed a quantitative elemental analysis using a scanning electron microscope with an energy dispersive X-ray spectrometer (SEM-EDX) in order to achieve a classification scheme for samples collected in Poland. Float glass samples of relevant cars in New Zealand using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) have been analyzed in [13]. Uzkent et al [14] have developed a system for classifying the sound produced by the glass breaking.…”