Abstract

The effect on threshold current density J_th of the use of an air-lock sample exchange mechanism, substrate preparation, and Al source purity was investigated for GaAs-Al_0.27Ga_0.73As double heterostructure (DH) lasers grown by molecular-beam epitaxy (MBE). The importance of having an airlock sample exchange can be illustrated by the improvement of threshold current density as a function of the number of runs after the growth chamber is opened to atmosphere for oven refilling. Usually the threshold current density J_th is more than four times higher than that of wafers grown by liquid phase epitaxy on the first run, and then J_th improves to about 1.5 times higher on the third and fourth runs. Finally, J_th on the fifth and later runs is comparable to that for wafers grown by LPE. Also, care must be taken in substrate selection and preparation and in the use of Al free of undesirable impurities. The detailed sample preparation and the analysis of the Al source will be discussed. Figure 1 shows the temperature-time cycles used to grow the DH lasers. Broad-area lasers with cavity lengths between 400 to 500 μm having J_th of 1.25×10³, 1.16x10³, 1.04×10³ A/cm² for active layer thickness of 0.22, 0.16 and 0.10 μm respectively have been achieved. For a cavity length of 570 μm, the wafer with 0.1 μm active layer gave J_th as low as 910 A/cm². These DH lasers have been fabricated as proton bombarded stripe-geometry lasers (12 μm × 380 μm) and have a room temperature threshold current of 80 mA. The desirable features of molecular beam epitaxy are the capability of growth over a large substrate area (4 cm x 2.5 cm) and the reproducibility of this growth technique.

© 1980 Optical Society of America