NC State forensic sciences: dye database
Crime scene investigators have typically used microscopes and relatively basic analytical techniques to discern the differences between fibers. But just saying that two fibers—one found on a victim’s body and one from a suspect’s car, for instance—are both blue polyester has limited value in the courtroom.
At NC State’s College of Textiles, researchers are constructing the first-ever database that will let investigators identify dyed fibers according to the type of dye used, allowing them to more definitively link fibers that could potentially be from the same source, such as linking fibers from a car to a crime scene. The project, funded by a $600,000 grant from the National Institute for Justice, was presented at a national conference on trace evidence in September. David Hinks, Cone Mills Professor of Textile Chemistry who is working with colleagues Keith Beck and Dieter Griffis, said the response was positive. “They loved how it could change fiber-based trace evidence analysis,” Hinks says.
Here’s how it works: Hinks is using a set of books containing fiber samples from nearly every model car sold in the U.S. since 1955. Researchers take a fiber from the sample and use a solvent to extract the dye. Then they put the dye through a mass spectrometer, which determines the molecular structure of the dye. Researchers catalog the molecular structure and the type of car the fiber came from to build the database.
But it doesn’t stop there. The College of Engineering’s Analytical Instrumentation Facility is working on ways to capture information about the fiber without destroying the evidence. Griffis’ work involves embedding a portion of the fiber in resin, using an ultra-sharp knife to slice
a very thin cross-section between 50 and 100 nanometers (for reference: a strand of hair is 15,000 nanometers). The slice is subjected to a different kind of mass spectrometry. “All we need is the surface of a fiber of itself,” Griffis says. “It doesn’t require extraction of the dye.”
- Categories: