Rapid screening for smokeless powders using DART-HRMS and thermal desorption DART-HRMS

• NCJ Number: 308545
• Journal: Forensic Chemistry Volume: 13 Dated: May 2019
• Author(s): Emily Lennert; Candice M. Bridge
• Date Published: May 2019
• Grant Number(s): 2018-R2-CX-0008
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)
• Award: Analysis and Characterization of Smokeless Powders and Smokeless Powder Residues

Annotation

This study compares two methods for rapid screening of smokeless powders:  thermal desorption – direct analysis in real time – high resolution mass spectrometry (TD-DART-HRMS) and traditional direct analysis in real time – high resolution mass spectrometry (DART-HRMS).

Abstract

In this study, TD-DART-HRMS is compared with traditional DART-HRMS for smokeless powder analysis.  Smokeless powders are a form of trace evidence that may be present at the scene of a shooting or explosive event, i.e. the explosion of an improvised explosive device (IED). Thermal desorption – direct analysis in real time – high resolution mass spectrometry (TD-DART-HRMS) is a novel technique in which an independent thermal desorption unit is coupled to a DART-HRMS. Samples were extracted to minimize sample variability and allow for simple, direct comparison between techniques. Chemometric statistical approaches were applied for the comparison of TD-DART-HRMS and DART-HRMS. Sorensen correlation was applied for direct comparison of spectra, and positive ionization mode mass spectra were found to be more highly correlated than the negative ionization mode spectra, indicating greater similarity among DART-HRMS and TD-DART-HRMS spectra in the positive ionization mode. Hierarchical cluster analysis (HCA) was utilized for the formation of groupings based on underlying features in the spectra, and classification modeling was performed using linear discriminant analysis (LDA) along with introducing an external test set to the LDA model. High potential for classification was observed in the LDA modeling with 20% withheld cross validation. External test set validation demonstrated that the TD-DART-HRMS model was a more robust model than the DART-HRMS model, with a higher classification accuracy observed. This study established the utility of TD-DART-HRMS for the analysis of smokeless powders, and demonstrated the complementary nature of TD-DART-HRMS associated with DART-HRMS. (Published Abstract Provided)