Application of total X-Ray fluorescence to gunshot residue determination

Currently the great majority of the criminal acts have involved the use of firearms, for these reasons the evidences generate from these are one of the fundamental pillars of a forensic investigation. The firearm leaves evidence known as gunshot residue (GSR), which is principally composed of burnt and unburnt particles from the detonation, as well as fragments of the bullet, cartridge case, and the firearm. Gunshot residue (GSR) is produced when a firearm is discharged and large quantities of it can be transferred to an individual who has fired. SEM-EDX is the common technique used in the forensic laboratories, the analysis consists in detecting the particles and its elements. In this work we propose the use of X-ray Spectrometry by Total Reflection (TXRF) for the analysis of metals present in related samples in ballistic cases. The analysis was focused in the relationship of three elements present in GSR. A series of experiments with different persons firing gun of 9 mm was performed in a shooting range. Analytical XRFS signals corresponding to K line of Copper and L lines Barium and Lead were employed as the best discriminating variables. Machine Learning techniques, such as discriminant analysis, supported vector machines and partial least squares - discriminant analysis, enable the correct classification of all samples analyzed. A hundred samples were analyzed so far, this method has demonstrated a very high classification performance for detecting gunpowder residues in the skin.