Development of a broadband coherent Raman microscope

The Coherent Raman Scattering Microscopy (CRS-Microscopy) techniques are rapidly gaining recognition in biomedical optics thanks to theirs non-invasive, label-free imaging capabilities of tissues and cells. CRS is a third-order non-linear process and in particular, two synchronized laser pulses of angular frequency !p and !s are used to interact with the sample, when the frequency difference between the two pulses matches the vibrational resonance of the molecules in the sample, all together coherently vibrate and the Raman signal is enhanced by many order of magnitude. In this thesis work, has been developed a complete vibrational spectroscopy system, starting from the two beam-lines to the realized multi-photon microscope and using the most employed CRS techniques, the Stimulated Raman Scattering (SRS) and the Coherent Anti-Stokes Raman Scattering (CARS). In detail, the broadband vibrational spectrum of solvents, with vibrational features in the accessible spectral region, such as methanol has been acquired and analyzed using software simulation. Then, the multiphoton microscope has been tested acquiring CARS images, that are bi-dimensional maps in which each pixel contains the information of the related vibrational spectrum. These images have been used to identify and distinguish among different materials in the specimen using multivariate analysis and therefore, the information contained in the whole spectrum and not only on specific vibrational peaks, like in single color techniques. XI Exploiting the fact that the signal is generated only in the focal volume, it has been possible to acquire hyper-volumetric images in which an entire volume of the sample was imaged. This four-dimensional images (three in space and one in the spectral domain) are of fundamental importance in analyzing biological sample where the entire tissue structure has to be taken into account and not only a section and was made on home-made sample with known features to evaluate the system performances.