Detection and diversity of fluorinated oil- and water-repellent coatings on apparel fibers

• NCJ Number: 304091
• Journal: Journal of Forensic Sciences Volume: 66 Issue: 4 Dated: 2021 Pages: 1285-1299
• Author(s): Michael J. Dolan Jr. MS; Wanqing Li; Kaveh Jorabchi PhD
• Date Published: 2021
• Length: 15 pages
• Grant Number(s): 2017-R2-CX-0007
• Agencies: NIJ-Sponsored
• Award: A New Analytical Technique to Identify Surface Treatments of a Single Fiber

Annotation

This study used a sensitive fluorine-selective analytical technique to detect and evaluate the diversity of fluorinated coatings in apparel.

Abstract

Fluorinated coatings, often used for oil and water repellency and stain resistance in fabrics, are potentially persistent forensic fiber markers; however, they have received limited attention because of challenges in their detection caused by the small size of a single fiber and thin nature of stain-resistant coatings. In the current study, 12 clothing items marketed as stain-resistant were tested, with 9 showing oil- and water-repellent properties. Fluorinated pyrolysis products of single fibers from all of the 9 items were detected by gas chromatography coupled to plasma-assisted reaction chemical ionization mass spectrometry (GC-PARCI-MS), indicating the prevalence of fluoropolymer coatings in stain-resistant clothing articles. Furthermore, three major classes of fluorinated coatings were identified via principal component analysis of pyrogram patterns. The classes were coating-specific and did not correlate with fiber core and color, highlighting a robust detection methodology. To evaluate the effect of fiber lifting in crime scenes, fibers from the 9 clothing items were used to develop a multinomial logistic regression model based on pyrogram principal components. The model was then tested using fibers subjected to contact with Post-it® notes. The test set fibers were sampled from the clothing items of the training set and from three additional garments of differing color but the same brands as the training set. The coating classes were predicted with 98.4-percent accuracy, confirming robust classification of fiber coatings using py-GC-PARCI-MS regardless of fiber color, core, and fiber lifting. (publisher abstract modified)