A Rapid and Non-Destructive Discrimination and Source Correspondence of Tyres by Raman Spectroscopy Combined with Chemometrics

Identifying and comparing the tyre (rubber) threads are crucial in forensic investigations involving collision accidents and hit-and-run incidents. Focusing on improving the effectiveness of tyre classification and identification, the current study pioneered the utilisation of Raman spectroscopy combined with PLS-DA on 140 samples of light-motor vehicle (LMV) and heavy-motor vehicle (HMV) automobile tyres. All the spectral peaks were examined from the region of 400–2000 cm−1, which indicated the molecular vibrations of the tyre (rubber) composition. Meanwhile, the chemometric method of PLS-DA achieved a very highly accurate (99.4 %) classification between LMV and HMV samples. The model was further tested using 17 unknown samples (treated as blind samples). 94.1 % of these blind samples were accurately predicted into their respective classes. Following after 6 months period, the study proceeded to explore the application of the PLS-DA model in real settings using 12 different unknown samples after scratch on the road, which resulted in 83.3 % correct predictions. To the best of our knowledge, this study is the most comprehensive on tyre classifications and source correspondence as a non-destructive approach to examine tyre rubber, yielding valuable insights for related crime scene investigations.