Single-fiber-laser-based multimodal coherent Raman system

Abstract

Coherent Raman scattering (CRS) is an appealing technique for spectroscopy and microscopy due to its molecular specificity and the ability for 3D sectioning. However, the system usually has to use two laser sources with exactly the same repetition rate but different frequencies, which makes the setup expensive and the tuning procedure complicated. As presented in this thesis, my PhD research attempts to simplify the CRS system, and extend its capability into different applications with multiple modalities. Specifically, we have designed and developed a single-fiber-laser-based CRS spectroscopy system. Instead of using two separate lasers to provide the pump and the Stokes, we split the output of a single laser into two parts; one serves as the pump, and the other passing through a photonic crystal fiber (PCF) serves as the Stokes. The Stokes frequency shift is generated by soliton self-frequency shift (SSFS) within the PFC. By using a single fiber-laser as the source, the CRS spectroscopy system can automatically maintain pulse repetition rate synchronization between the pump and the Stokes beams, which dramatically simplifies the configuration. The impact of pulse chirping on the spectral resolution and signal power reduction of CRS has been investigated. Spectra of C-H stretches of cyclohexane induced by coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS) were measured simultaneously and compared. The simulation results are verified through spectroscopy experiments. With minor modification, the CRS spectroscopy system can be extended into a multimodal microscopy system with the capability of performing CARS, SRS and photothermal microscopy, since they are all based on the same optical pump-probe configuration. CARS and SRS are ideal for detecting molecular vibrational mode without labeling, and photothermal microscopy is a sensitive technique most suitable for detection of light absorption by molecules that do not fluoresce. By combining these techniques into one microscopy system, not only different measurement modalities can be compared, but also offering complementary information of the sample. Distribution of hemoglobin in human red blood cells, and lipids distribution in sliced mouse brain have been imaged. Modulation frequency and power dependency of the photothermal signals are discussed in detail. The Raman gain or loss introduced through SRS is usually very weak, on the order of 10e-ð€€€5, and direct detection of this small perturbation is challenging. A commonly used technique for SRS microscopy is modulating the intensity of the pump or the Stokes beam, and the stimulated Raman gain (SRG) or stimulated Raman loss (SRL) can be detected by using a lock-in amplifier synchronized with the modulating waveform. In this way, the resonant SRS signal is often accompanied by photothermal signal created through thermal lensing in the sample. The SRS signal can be easily overwhelmed by photothermal signal, especially when the modulation frequency is lower than 100KHz. We proposed a polarization stimulated Raman scattering (P-SRS) method to suppress the unwanted photothermal signal. Instead of using intensity modulated pump, we modulate the state of polarization (SOP) of pump, and the polarization-dependent SRG on the Stokes is measured. With different SOP settings, this allows the detection of different elements of the third-order susceptibility matrix. On the other hand, the photothermal signal is independent of the polarization of pump so that its impact can be eliminated. We have imaged both red blood cells and sliced mouse brain samples to demonstrate the capability of suppressing the photothermal signal from the resonant SRS signal.