Estimating the time since discharge of handgun cartridges using a novel headspace sorptive extraction (HSSE) approach

Estimating the time since discharge of handgun cartridges may be a useful piece of information in the forensic investigation of firearm-related crimes. Many approaches were reported throughout the last century. The most promising were generally based on analysing over time the decrease in the headspace concentration of dissipating compounds released during the discharge (especially, the most volatile ones, such as naphthalene). Following the good results obtained in analogous applications (i.e., the estimation of the time since last discharge of firearms and shotgun shells), headspace solid phase micro-extraction (SPME) followed by gas chromatography was recently suggested to analyse shot handgun cartridges. However, unsatisfactory results were reported, likely because of analytical limitations (reproducibility and sensitivity issues) and/or weak choices in target compounds and interpretative models. In an attempt to further study the potential of headspace analysis to estimate the time since discharge of handgun cartridges, a novel approach was developed in this PhD research. This was based on the extraction of fired cases by headspace sorptive extraction (HSSE), which is more exhaustive (and thus, more reproducible and sensitive) than SPME. Target compounds were selected on the basis of a preliminary study, and the entire analytical methodology was then optimized using a multivariate statistical approach based on the design-ofexperiments theory. Results showed good analytical performances in terms of detection limits and repeatability. Its application to real specimens revealed that the amounts of volatile GSR compounds rapidly decreased in the first hours after discharge (< 10 h), but they did not totally disappear from cartridge headspaces. Thus, through multivariate interpretation models, it was possible to provide time-since-discharge estimates up to 48 h and to differentiate recently fired cases from older ones. The best combinations of pretreatments and multivariate regression models were pairwise ratios followed by logarithm transformation (i.e., PLR) and random forests (RF) and/or partial least squares (PLS). The effects of several influential factors (i.e., cartridge position, humidity and temperature) were further studied. While all affected in some way the ageing kinetics of the single GSR compounds, they did not affect the multivariate models' dating accuracy, as long as there was no significant difference between reference and examined material. An evaluative model was also developed and allowed assigning a likelihood ratio to analytical outcomes given alternative hypotheses on the time since discharge forwarded by the parties at trial. Thus, headspace analysis of spent cartridges was found to provide helpful evidence to assist forensic practitioners in answering questions about time since discharge. Further research is needed, but results toward the development of a complete dating methodology, and its implementation in real forensic cases, are promising