| The success of short-wavelength (850 nm) vertical-cavity surface-emitting lasers (VCSELs) as low-cost components in fiber optic networks has created a strong demand for similar low-cost devices at longer wavelengths (1.3–1.55μm), which are even more important in telecommunications systems. Extending the success of VCSELs to these longer wavelengths, however, has been slowed by the absence of a mature technology that incorporates all of the necessary components on one substrate without sacrificing the inexpensive and manufacturable nature of VCSELs. Although InAlGaAs active regions on InP substrates have been developed extensively, the other components of vertical-cavity lasers, especially epitaxially-grown distributed Bragg reflectors (DBRs), are less mature on these substrates.; This thesis examines the materials and technologies that enable long-wavelength VCSELs to be grown in a single, epitaxial, lattice-matched step on InP substrates. The advantages and shortcomings of each material system are identified and the impact on devices examined. Additionally, processing technologies that rely on the properties of these materials are developed. From these studies, a InP-based, lattice-matched VCSEL design is presented that utilizes AlGaAsSb for high reflectivity DBRs, InAlGaAs for high quality active regions, InP for heat and current spreading, and a materials selective etch for electrical and optical confinement. In short, the design avoids the shortcomings of each material system while emphasizing the advantages. The resulting devices, showing low threshold currents, high efficiencies and powers, and high operating temperatures, not only validate this approach but demonstrate that such lattice-matched, InP-based devices may be a low-cost, manufacturable answer to this long-wavelength VCSEL demand. |
