RESUMEN
The application of infrared chemical imaging to the analysis of bicomponent fibers was evaluated. Eleven nominally bicomponent fibers were examined either side-on or in cross-section. In six of the 11 samples, infrared chemical imaging was able to spatially resolve two spectroscopically distinct regions when the fibers were examined side-on. As well as yielding characteristic infrared spectra of each component, the technique also provided images that clearly illustrated the side-by-side configuration of these components in the fiber. In one case it was possible to prepare and image a cross-section of the fiber, but in general the preparation of fiber cross-sections proved very difficult. In five of the 11 samples, the infrared spectra could be used to identify the overall chemical composition of the fibers, according to a published classification scheme, but the fiber components could not be spatially resolved. Difficulties that are inherent to conventional "single-point" infrared spectroscopy, such as interference fringing and sloping baselines, particularly when analyzing acrylic type fibers, were also encountered in the infrared chemical image analysis of bicomponent fibers. A number of infrared sampling techniques were investigated to overcome these problems, and recommendations for the best sampling technique are given. Chemical imaging results were compared with those obtained using conventional fiber microscopy techniques.
RESUMEN
This paper examines the potential of infrared chemical (hyperspectral) imaging as a technique for the forensic analysis of automotive paint chips in particular, and multicomponent (e.g., layered) samples in general. Improved sample preparation procedures for the infrared analysis of paint chips are detailed, with the recommendation that where mounting resins are chemically incompatible with the sample, it is better to mount and section the sample in a soft wax from which the sections can be removed and pressed into a KBr disk for transmission analysis. Infrared chemical images of multilayered paint chips have been successfully obtained, with the chief advantage over conventional infrared analysis being that thousands of infrared spectra are collected in a few minutes across the whole sample, at a spatial resolution of around 5 microm. As with conventional infrared spectroscopy, chemical species can be identified from their spectra, but the wealth of information available can be also extracted in a number of different ways that make multicomponent spectral (and hence chemical) comparisons between two samples easy to visualize and understand. In one approach, the infrared chemical images of two paint chips being compared side-by-side can be viewed as a "movie," in which each frame is an intensity map of the two samples at a given wavenumber (frequency) value. In another approach, the spectra (pixels) in the image files are classified into chemically similar groups, resulting in a "cluster" image that makes it possible to simultaneously compare all of the layers in two paint chips. These methods are applicable to other multicomponent samples, and also to other chemical imaging techniques.