| Popis: |
We have developed a new Fourier transform infrared chemical imaging technique1 which, when coupled with powerful multivariate data processing methods, allows the visualization of intrinsic chemical distributions in biological samples and other composite materials. Integrating spectroscopy with sample visualization and digital image processing is a potent combination of what have traditionally been two distinct methods for studying the chemistry and morphology of a sample. This synergy has been referred to as chemical imaging or hyperspectral imaging and has wide ranging implications for material characterization. In the infrared spectral region the technique relies on the use of infrared focal-plane array detectors composed of either indium antimonide (InSB), mercury cadmium telluride (MCT) or arsenic doped silicon (Si:As). These arrays, which were originally developed for defense related applications, constitute an emerging commercial technology. The arrays can be used in conjunction with standard Cassegrainian infrared optics and step-scan Michelson infrared interferometers to construct imaging systems capable of collecting tens of thousands of spatially resolved infrared spectra and images with less than 1 minute of data acquisition time.2 The data sets contain both spatial and spectral information and typically consist of hundreds of images resolved in frequency space (wavenumbers, cm-1), with each image containing many tens of thousands of pixels. Using a microscope, each pixel can sample a region as small as 2 μms2 of a sample surface. |
