High-power, picosecond pulse generation from Q-switched diode lasers

This dissertation studies high-power picosecond optical pulse generation from Q-switched diode lasers. Multisection quantum-well lasers have been fabricated and thoroughly characterized. Various device structures and switching schemes have been experimentally investigated in order to understand the large-signal dynamics as well as to explore potential applications. A comprehensive theoretical analysis is also presented.;In this work, we have investigated the possibility of boosting the peak power of the Q-switched pulse while maintaining a high repetition rate operation. We accomplished this by combining a tapered stripe structure and fast ridge-waveguide modulation sections in the same optical cavity of a monolithic device. By this approach we were able to produce above 1 W peak power, 20-30 picosecond optical pulses with flexible repetition rate up to a few GHz. Nearly diffraction-limited far-field emission patterns were also measured. The flexibility and functionality of our devices were demonstrated by a data encoding experiment, in which NRZ electronic data were converted to RZ optical output.;To further explore potential applications of high-power Q-switched diode lasers, we have fabricated multisection InGaAsP devices emitting near 1550 nm. We have conducted CW injection-seeding experiments, aiming at generating single-mode, wavelength tunable picosecond pulses. We have achieved a tuning range as large as 33 nm, with side-mode suppression ratio larger than 10 dB over the entire range. The pulse width was shortened to 7 ps after linear fiber compression; a pulse compression factor of 6 was achieved. After spectral filtering, it may be used for generating sub-picosecond pulses through soliton-compression effect in optical fibers.;We have devised a novel, all-optical method to characterize in detail the ultrafast large-signal dynamics of high power Q-switched diode lasers. We also performed extensive numerical simulations. Useful design rules and scaling properties were derived. The theoretical model has been successfully applied to predict the device behavior and explain the experimental results.