Elemental Characterization of Printing Inks and Strengthening the Evaluation of Forensic Glass Evidence

Abstract

Boeglin for their time and helpful feedback. A special thank you goes out to Dr. Tatiana Trejos for her eager willingness to assist me with my research. Her passion for teaching constantly motivated me to work harder and strive to be a better scientist. I am thankful for Pupi and the Almirall research group (the "A team"), in particular the laser lab girls, who made this journey easier to bear. I am grateful to have had constant companions to laugh with and do Zumba with during challenging times. I would like to acknowledge the US Secret Service for providing ink specimens. I would also like to thank Vidia (who helped analyze glass samples) and all those who participated in the Glass Interpretation Working Group (GIWG). Last, but certainly not least, I am grateful to my family and friends for their unwavering support. In particular, I would like to acknowledge my parents Susy and Mario, my brother Isaac, my sister Elie, my best friend Kathy, my sister-in-law Mari, and my niece Evie for much needed distractions during stressful times. v

Interpretation of chemical data from glass analysis for forensic purposes

“…It is well‐known and well‐documented that the composition of float glass changes only slowly during the production process so that glass produced on the same day or on some consecutive days is expected to show essentially the same elemental composition. For example, glass produced in the same plant within days to sometimes a few weeks apart can usually not be discriminated by LA‐ICP‐MS 16,36,41 . Forensic examiners would take this into account in their consideration of glass evidence, if necessary.…”

“…In all cases, ASTM E2927‐16e1 was used to collect the multivariate data. The calculation of a calibrated log likelihood ratios (LLRs) for the work presented here was performed using an R‐script developed at Florida International University (FIU) in collaboration with Ramos from the AUDIAS research group 36,37 . The calculation of calibrated LLRs relies on the initial calculation of an LR using the multivariate kernel density (MVK) model reported by Aitken and Lucy, 38 followed by its calibration by the pool adjacent violators (PAV) algorithm.…”

“…The extension of the derivation for situations with an unbalanced number of fragments across the background sources is trivial. In this project, the background database was provided by the Bundeskriminalamt (BKA) forensic laboratory 36,37,40 …”