Abstract
The drive for low threshold and temperature-stable semiconductor lasers for telecommunication applications has led to a significant interest in quantum dot (QD) lasers emitting in the 1.3 mum and 1.5 mum wavelength range. The literature shows that although low threshold current densities can be achieved, this is usually at the expense of a poor temperature stability. Low-temperature and high-pressure measurements of the threshold current and its radiative component are performed on undoped and p-doped 1.3 mum InAs/GaAs and 1.5 mum InAs/InP (311)B QD lasers. The results show that despite a fairly temperature-stable radiative current around room temperature, undoped QD lasers suffer from a poor temperature stability of their threshold current. This is because there is a large contribution (70% and 90% of the threshold current at room temperature in 1.3 and 1.5 mum lasers, respectively) from a strongly temperature sensitive non-radiative Auger recombination process. Several pieces of evidence are found to explain the observed decrease of the radiative current, explained by an improvement of the carrier distribution with increasing temperature. We find that in p-doped devices the temperature dependence of the radiative component of the threshold current can be modified by the doping. In these devices the radiative current can decrease with increasing temperature around room temperature while the non-radiative current increases. This results in a small range of temperatures over which the threshold current is constant (from ~ 270 to 300 K). This effect is very sensitive to the doping concentration. If the doping concentration is carefully chosen, this can result in high T0 devices but with larger threshold currents than in comparable undoped lasers. Gain measurements reveal that the differential gain of p-doped lasers is less than that of the undoped devices because of the increased non-radiative current and the non-thermal distribution of the carriers induced by the doping. Finally, a new method is demonstrated to measure the band gap dependence of the Auger coefficient, C, using a combination of high hydrostatic pressure measurements coupled with gain calculations.