Pulsed Blue and Ultraviolet Laser System for Fluorescence Diagnostics based on Nonlinear Frequency Conversion

Abstract

The motivation for the current thesis work is to build a compact, efficient, pulsed, diode-pumped solid-state (DPSS) laser at 340 nm to be used for autofluorescence imaging and related cancer diagnostic experiments. By exciting endogenous fluorophores in the UV spectrum, autofluorescence imaging eliminates the need for prior preparation on the patient, and further reduces the cost of fluorescence imaging, and potentially, cancer diagnostic. By taking advantage of the relatively short wavelength of the 946 nm transition in a quasi-three-level Nd:YAG laser, the target 340 nm wavelength could be reached through sum-frequency generation (SFG) with a frequency- doubled 532 nm Nd:YAG laser. However, since the quasi-three-level transition suffer from reabsorption loss and a ten-fold reduction in stimulated emission cross-section compared to the four-level 1064 nm transition, optimization of the 946 nm laser is non-trivial. Detailed investigation into pump beam optimization has been carried out for an end-pumped 946 nm CW laser. Using an innovative external cavity tapered diode laser as pump source, a record 800 mW of output power was obtained using a single-emitter diode laser pump source. The spatial and spectral properties of the pump source were also investigated individually, and it was concluded that a broad spectrum tapered diode pump source may be most stable and cost-effective. To generate high peak power pulsed output, Q-switched lasers were considered. In particular, synchronized Q-switching between a 946 nm and 1064 nm Nd:YAG laser was achieved in a passive approach. To the author’s knowledge, stable, passive synchronization between a quasi-three-level and a four-level laser was achieved for the first time over a wide range of pump powers. The minimum delay between the two pulses was 64 ns, which translates into a 79% temporal overlap when compared to the zero-delay scenario. The minimum timing jitter between the two pulses was 9 ns, which is one-standard deviation of the delay measurements. This is comparable to previously published results for an actively synchronized system if either a four-sigma or six-sigma definition was used. Detailed investigation into the relative timing jitter between the two synchronized pulses was also carried out, where it was found that the lower limit on the relative jitter, determined by pump power fluctuations and amplified spontaneous emission, was 6 ns. Comparing this to the 9 ns relative jitter achieved in the passive system shows the performance penalty incurred in using the passive approach. Lastly, practical applications of compact semiconductor and DPSS lasers in the blue and UV spectral region are presented. A CW tapered diode at 808 nm was directly frequency-doubled to 404 nm using an external cavity, and was used in an animal experiment for a novel approach in estimating photosensitizer concentration using fluorescence imaging. Secondly, a frequency-tripled, 355 nm, Q-switched, DPSS laser was used in a preliminary clinical investigation in autofluorescence diagnostic of skin cancer. While the preliminary results are promising, the system would benefit from a 340 nm light source that tunes into the absorption peaks of endogenous fluorophores. The imaging system would also benefit from a high-peak power light source that would increase the signal-to-noise ratio. Based on the clinical results, there is a clear need for a high peak-power, 340 nm pulsed laser source for autofluorescence experiments. While the current results on passively synchronized Q-switching seem very promising for SFG generation into such wavelength, the technique could be equally applied to other wavelengths; specifically, those in the blue and UV spectral region. Using the passive synchronization technique and the optimization procedure reported for quasi-three-level lasers, a new generation of high peak power, pulsed, blue and UV laser light sources could be realized.