Modern automotive paints have a thin color coat layer, which on a microscopic fragment is usually too thin to obtain accurate chemical information. Fortunately, adhesion between paint layers is very strong and both primer layers will be transferred during a collision if the topcoat layer is also transferred.  Applying data fusion techniques where infrared (IR) spectra from multiple sources (e.g., IR spectra of top coat and undercoat automotive paint layers) are combined and class membership information is extracted, a hierarchical classification scheme has been developed where search prefilters for IR library searching are utilized to identify the assembly plant of an automobile from which an unknown paint sample originated. The classification scheme, which performed better than one developed only using the clear coat layer, was successfully validated using an external prediction set.  Even in challenging trials where the top coat and undercoat paint layers evaluated were all from the same make (General Motors) with a limited production year range (2000-2006), the respective manufacturing plants could be correctly identified using chemical information from the top coat and undercoat paint layers. The development of search prefilters which exploit multiple sources of IR data to facilitate the searching of the PDQ database is necessary to extract investigative lead information from top coat and primer layer paint smears that are often left at the crime scene.

Search prefilters for infrared library searching in the PDQ database to enhance investigative lead information from automotive paints

247th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.