U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Application of Laser-Induced Breakdown Spectroscopy to Forensic Science: Analysis of Paint Samples

NCJ Number
237839
Author(s)
Michael E. Sigman, Ph.D.; Erin M. McIntee, M.S.; Candice Bridge, Ph.D.
Date Published
January 2012
Length
84 pages
Annotation
This research examined the application of laser induced breakdown spectroscopy (LIBS) as an inexpensive and rapid analytical method for the characterization of the elemental composition of automotive paint samples in order to discriminate between two samples at a known level of statistical significance.
Abstract
The study determined that LIBS had a discrimination power of 90 percent (10 percent Type II errors) at a verified 5-percent Type I error rate. Discrimination was found to be slightly lower (86.6 percent) among the white color group. Variations in the LIBS signal over time led to same-sample discriminations and an artificially high Type I error rate, which was overcome through attention to the sampling protocol and confining spectra-collection on samples that were to be compared to a narrow time window. Laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) gave the best sample discrimination (100 percent); X-ray fluorescence (XRF) and scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM/EDS) gave the lowest discriminations (85 percent and 73 percent total discrimination for each method, respectively). Study results suggest that LIBS may provide an important screening tool in the analysis of automotive paint samples; however, careful attention should be given to sampling protocols and the statistical comparison of samples. In cases where two samples cannot be distinguished, a more accurate comparison should be used. All discriminations were performed by hypothesis-testing at the alpha = 0.05 significance level, using both parametric and nonparametric statistical tests. Discrimination was tested across all paint samples, irrespective of paint color or other features. In a more forensically relevant fashion, discrimination was determined for only those samples of the same color group, number of paint layers, and the presence or absence of effect pigments in the paints. A total of 200 paint samples were examined by the different analytical methods. 35 tables, 7 figures, 73 references, and a listing of publications and presentations of this research