Discipline: Chemistry and Chemical Sciences
Subcategory: Chemistry (not Biochemistry)
Kaitlin Jones - University of Central Florida
Co-Author(s): Mark Maric, National Center for Forensic Science, Orlando, FL; Candice Bridge, National Center for Forensic Science, University of Central Florida, Orlando, FL.
Currently, there are three common analytical techniques used for the analysis of automotive paint: microscopy, infrared (IR) spectroscopy, and pyrolysis-Gas Chromatography Mass Spectrometry (Py-GC/MS). Py-GC/MS is the golden standard when it comes to automotive paint analysis due to its ability to differentiate between paint samples that are indistinguishable by IR spectroscopy. The reason for the high discriminatory capability of Py-GC/MS is the technique’s sensitivity to binders, additives, and cross-linking agents. However, a disadvantage of Py-GC/MS is that the technique is time consuming and destructive. Direct analysis in real time-high resolution mass spectrometry (DART-HRMS) was evaluated to combat one of the disadvantages of Py-GC/MS. The goal was to determine if DART-HRMS compares to the discriminatory power of Py-GC/MS while also having the advantage of analyzing samples in a fraction of the time. Both techniques were compared to assess how the data obtained from DART-HRMS related to that of py-GC/MS. DART-HRMS is a rapid screening technique that uses soft ionization. The technique requires little sample preparation and rapidly analyzes samples in four minutes under ambient conditions. Moreover, DART-HRMS has the capability to measure the mass of large and high weight molecules such as polymers commonly found in automotive paints. A cross section was cut with a scalpel from 100 samples that were obtained from automotive body shops around the Orlando, FL area and from a junkyard in Pembroke Pines, FL. A VHX 6000 Keyence digital microscope was utilized to analyze the cross sections and determine layer systems of each sample. The clear coat and base coat of the 100 samples were analyzed using DART-HRMS in positive ionization mode. A few samples were analyzed, in triplicate, in negative ionization mode on the DART-HRMS, however, no additional information was obtained. The clear coats of the set of 100 samples were also analyzed in duplicate with Py-GC/MS for comparative purposes. Principal component analysis (PCA), hierarchical cluster analysis (HCA), and linear discriminant analysis (LDA), were performed on the data to determine the classification potential of each instrumental technique. HCA was performed to identify unsupervised clusters within the data. Then, PCA was utilized to reduce the dimensionality of the data set to make it easier to visualize the patterns in each dataset. Lastly, LDA was performed to determine the accuracy of the models. The discriminatory ability that DART-HRMS exhibited was as powerful as Py-GC/MS for automotive paint samples. Due to the rapid analytical time of DART-HRMS, this method could be a good screening or analytical technique that provides similar information to current methods for the forensic paint analysis community.
Funder Acknowledgement(s): This research was supported by the 2017 Forensic Science Foundation's Lucas Grant and the State of Florida
Faculty Advisor: Dr. Candice Bridge, firstname.lastname@example.org
Role: I collected the samples and analyzed them with the Keyence digital microscope, DART-HRMS, and Py-GC/MS. I also performed the chemometrics on the data obtained.