Study on Design and Characterization of High Energy Femtosecond Laser Oscillators

Abstract

With a repetition rate of about 100 MHz, the pulse energy that can be obtained directly from a conventional femtosecond laser is only a few nJ, so laser amplifier systems are generally used in experiments that require high energy pulses. However, the amplifier systems cause increase in cost and size. Extended-cavity femtosecond lasers can produce the pulse energy of 100 nJ or more, which results from reducing the laser repetition rate and hence increasing the pulse energy. Thus, the extended-cavity femtosecond lasers can replace the laser amplifier systems in many experiments such as laser processing and frequency conversion, requiring sub-μJ pulses. In this thesis, we developed and analyzed an extended-cavity femtosecond laser for laser ablation. Since the extended-cavity femtosecond laser had to be installed in a limited space, the Herriott multipass cavity (HMPC) configuration and the telescope cavity configuration using 4 mirrors had been utilized for designing the extended cavity. The extended cavity allowed the laser repetition rate to be reduced to about 9.7 MHz, resulting in 145 nJ, 115 fs pulses. The extended-cavity femtosecond laser was installed in a laser machining system for processing silicon wafers. In the laser machining system, the laser beam focused through an optical microscope objective lens was irradiated to the silicon wafers with a fluence of 0.33 J/cm2. As a result, it was confirmed that laser processing is possible using this laser oscillator without any amplifier systems. The details are covered in Chapter 3. On one hand, the pulse energy scalability in femtosecond laser oscillators is limited by the instabilities such as cw generation and multiple pulsating, resulting from excessive nonlinear effects in a gain medium. One of well-known methods of suppressing the instabilities is to reduce the peak power below the instability threshold by stretching the pulse. In positive dispersion regime, the chirped-pulse oscillators generate heavily-chirped pulses requiring external compression and thereby offer the potential of scaling the pulse energy. On the other hand, due to the low peak power of the chirped-pulse, it may be difficult to operate the Kerr-lens mode-locking of the chirped pulse oscillators without a saturable absorber. Among saturable absorbers, the semiconductor saturable absorber mirrors (SESAMs) have been frequently used because it is easy to use as an end mirror. In such the chirped-pulse oscillators with a SESAM, the intracavity dispersion can be controlled through a pair of prisms placed near an output coupler. However, this placement results in a spatially chirped output beam because of the angular dispersion occurring in the prism pair. To eliminate the spatial chirp, additional prisms are required out of the cavity. However, this increases a system size and introduces unwanted dispersion compensation. In order to solve these problems, we designed the chirped-pulse oscillator with the SESAM located at the middle of the cavity. In the cavity of the oscillator, the prism pair and output coupler are placed in different cavity-arms. As a result, the spatial-chirp-free output beam is generated even when the intracavity GDD is compensated by the prism pair. However, the output characteristics of the laser oscillator with the SESAM inside the cavity have not been reported, so the effects of the SESAM position on pulse evolution were firstly analyzed. In Chapter 4, the analysis was done by comparing two lasers that have the different SESAM positions (The cavity end or the middle of the cavity). The two compared lasers generated about 50 nJ pulses in positive dispersion regime, and this result was analyzed via the effective loss induced at a SESAM. In negative dispersion regime, it was found that the upper energy limit of a stable single pulse in the newly designed-cavity laser is lower than that of the conventional laser. This result was analyzed via soliton dynamics. Numerical simulations confirmed that pulse contrast and the stability against cw generation is better when SESAM is inside the cavity than at the cavity end. As mentioned above, in order to demonstrate that the newly designed-cavity laser can produce the output beam with no spatial chirp, we developed the chirped-pulse oscillator with the SESAM in the middle of the cavity that has a prism pair. The developed oscillator produced 172 nJ chirped-pulses that were compressed to 98 fs after pulse compression. An additional prism pair out of the cavity is not required to compensate for the spatial chirp so the developed oscillator can be said to be a more compact laser oscillator than the conventional laser oscillators.