Experimental results on tensile-strained Al/sub 0.3/Ga/sub 0.66/As-GaAs/sub 0.78/P/sub 0.22/ separate-confinement-heterostructure single-quantum-well (SCH-SQW) laser diodes are reported. Threshold current density as low as 260 A/cm 2 for broad-stripe lasers with a cavity length of 1500 μm has been observed. Broad-stripe devices have operated cw at room temperature with output power as high as 620 mW/facet. Ridge-waveguide lasers have exhibited cw threshold currents as low as 13.5 mA and output power of 90 mW. The output of the tensile-strained lasers is TM polarized.
AlGaAs-based superlattice microcoolers are demonstrated. Maximum cooling temperatures of 0.8 °C and 2 °C were obtained at 25 °C and 100 °C, respectively, from 60 μm×60 μm devices with 100 period -type superlattice thermal barriers. These devices may be useful for in situ cooling of GaAs-based microelectronic and optoelectronic devices.
A generalized form of Landauer’s bound on the dissipative cost of classical information processing in quantum-mechanical systems is proved using a new approach. This approach sidesteps some prominent objections to standard proofs of Landauer’s bound—broadly interpreted here as a nonzero lower bound on the amount of energy that is irreversibly transferred from a physical system to its environment for each bit of information that is lost from the system—while establishing a far more general result. Specializations of our generalized Landauer bound for ideal and non-ideal information processing operations, including but not limited to the simplified forms for erasure and logical operations most familiar from the literature, are presented and discussed. These bounds, taken together, enable reconsideration of the links between logical reversibility, physical reversibility, and conditioning of operations in contexts that include but are far more general than the thermodynamic model systems that are most widely invoked in discussions of Landauer’s Principle. Because of the strategy used to prove the generalized bounds and these specializations, this work may help to illuminate and resolve some longstanding controversies related to dissipation in computation.