Giulio Cerullo | Politecnico di Milano | Milan, Italy
Scientific Talks, Session VIII | Wednesday, September 14 | 9:00 – 9:30
Coherent Raman scattering (CRS) microscopy is a powerful nonlinear optical technique for chemical identification of (bio)-molecules based on their intrinsic vibrational spectrum and for high-speed label-free imaging of cells and tissues. Single-frequency CRS microscopy allows the detection of molecules with a specific Raman response but is not sufficient to distinguish different components within complex heterogeneous systems in which chemical species display spectrally overlapped resonances. For this reason, there is ongoing research aimed at extending CRS to broadband detection, combining the speed of coherent Raman spectroscopy with the information content of spontaneous Raman. In this talk we present two innovative approaches to broadband CRS, both in the stimulated Raman scattering (SRS) and coherent anti-Stokes Raman scattering (CARS) modalities. We first present a broadband CARS (B-CARS) system based on a high power ultrafast solid-state laser and white light continuum generation in a bulk material, acquiring high-quality spectra in a 1-ms time, limited by the spectrometer refresh rate. A convolutional neural network is used to denoise the B-CARS spectra and remove the unwanted non-resonant background. Next, we present a broadband SRS system equipped with a home-built multichannel lock-in amplifier simultaneously measuring the stimulated Raman signal over 32 frequencies within 44 µs, allowing for detailed, high spatial resolution mapping of spectrally congested samples. We demonstrate the performance of our SRS microscope by discriminating the relative concentrations of different fatty acids in cultured hepatocytes at the single lipid droplet level and by identifying fibrosarcoma tumor lesions embedded within healthy tissue.