| Summary: | Optical communication was developed to allow high-speed and long-distance data transmission and is currently a £6bn market. This has also led to the adoption of optical technologies in other areas including the CD, DVD and medical imaging systems. Standardisation of components means that these systems require light sources that operate near the 1310 and 1550 nm telecommunications windows but existing lasers here are expensive due to their high temperature sensitivity. The exploitation of quantum con¯nement has led to the development of \quan- tum dot" (QD) laser material because of predictions of huge gains in performance. Emission wavelengths of InAs/GaAs QD lasers have been extended to the telecom- munications window near 1300 nm by various growth technologies and the first commercial devices have recently been brought to the market. However, progress to longer wavelengths has been stalled for several years as well as the speed and tem- perature sensitivity of these devices falling short of the predictions; partly because QDs are grown by self-assembly resulting in a random distribution of sizes, compo- sitions and strain-states, leading to inhomogeneous broadening which is a departure from the ideal \atom-like" system. This work details the growth, design and development of QD bilayer laser devices, which o®er a unique approach to fixing these shortcomings. When two QD layers are grown close together; the first layer provides a template that allows larger, more uniform QDs to be grown in the second layer, giving greater uniformity and deeper confinement. This has the potential to increase the efficiency and to achieve emission wavelengths out towards the more-commonly used telecommunications window at 1550 nm directly on GaAs substrates. Multiple bilayer laser diodes with inhomge- neous broadening of less than 30meV, lasing at up to 1430 nm and room-temperature photoluminescence at 1515 nm are shown. Despite the vastly reduced inhomogeneous broadening of QD bilayers, it is still found to be a relevant factor due to the change from de-localised geometries of quantum wells to an ensemble of separate QDs. It will be shown that understanding this is essential for describing the observed optical and electrical behaviour of the laser diodes.
|
|---|
