Forensic laboratories have widely adopted micro-X-ray fluorescence (μXRF) spectrometry for glass examinations and benefit from its scientific validity and standard test methods. Recently, modern silicon drift detectors (SDD) started replacing traditional silicon-doped lithium (SiLi) detectors. This interlaboratory study addresses the need to adapt protocols designed for SiLi detectors to modern systems. Eight participants received two standard reference glasses and 100 fragments from ten sets of windshield glass. Although performance varied by instrument configuration, generally, the μXRF-SDD offered improved sensitivity (2–10 times), faster acquisition (1200 Ls to 300 Ls), and allowed lower sample sizes (spot sizes from 100–300 μm to 20–30 μm) compared to older SiLi systems. Each participant conducted 45 known-to-questioned pairwise comparisons without knowing the ground truth of the samples' origin. Of these, 28 were from the same source, and 17 from different sources. Overall, 800 spectral files were compared using spectral overlay and comparison intervals of elemental ratios (3 s and modified 3 s). Spectral overlay provided correct inclusions and exclusions for all comparisons across participants. The comparison intervals did not produce false inclusions. The modified 3 s interval accounted for the improved precision, reducing false exclusions from 16.3 % to 4.7 % compared to the 3 s criteria. Increasing the number of fragments of the known sample may compensate for heterogeneities derived from analyzing smaller areas. Additionally, spectral contrast angle ratios (SCAR) and Score Likelihood Ratios (SLR) assisted with data interpretation and comparisons across laboratories, offering a complementary tool to enhance objectivity in spectral comparisons.
(Publisher abstract provided.)