Synchrotron beams, the long- and much-appreciated workhorse of structural biologists for X-ray crystallography, small-angle X-ray scattering (SAXS) and other techniques, are perfect for these applications because of their high intensity, high polarization and low angular divergence. However, the stable and bright nature of their radiation also makes them excellent energy sources for other applications, including Fourier transform infrared (FTIR) imaging—a technique that generates images based on the infrared spectrum of biological samples.
FTIR imaging has an advantage over traditional microscopy as it does not require chemical labeling of samples with dyes, stains or tagged biomolecules. Instead, it uses mid-infrared absorption to produce an image by focal plane array (FPA) detection, with no additional processing of the sample needed. The image is determined by how areas of the sample respond to specific infrared wavelengths.
One challenge of FTIR microspectroscopy has been the lack of brightness and spatial resolution available from thermal sources of mid-infrared radiation, the traditional energy sources used in this technique. Previous work addressed this problem by adapting FTIR microspectroscopy to work with a synchrotron beam coupled to a mercury-cadmium-telluride detector (for raster scanning measurements), which led to significant improvements in the signal-to-noise ratio. The downside of the synchrotron approach was the low angular divergence of the synchrotron beam, which necessarily illuminates a small area and led to exceedingly long data collection times. Now, Hirschmugl and colleagues have addressed this issue and improved the utility of FTIR microspectroscopy.
Using a specially designed system, the authors split a fan of synchrotron beam radiation into 12 individual beams that were then bundled to illuminate a larger field of view before being collected by an FPA detector. The end result was a near 100-fold improvement in pixel size relative to thermal radiation sources, which allowed them to identify specific components of tissue samples. Without any additional chemical labeling, the authors were able to distinguish cell types, epithelial and stromal areas in tumor samples, and even high-collagen areas on the basis of their different absorption spectra.
One of the most pronounced benefits of the bundled beam approach was the time necessary for image acquisition. A 280 × 310 μm area required only 30 min for acquisition at diffraction-limited resolution, whereas the previous system using a synchrotron beam source would have required 11 d. The combination of high spatial resolution with substantially improved image acquisition time will undoubtedly make FTIR imaging using a synchrotron multi-beam source a vital technique for researchers looking for detailed images with minimal processing of their samples.