Search

The heating of a semiconductor laser bar imbedded between two heat spreaders has been studied theoretically and experimentally. The model included the p-cladding layer, active region, n-cladding layer, and n-substrate. Heat sources from the active region and both cladding layers were considered. An analytical relation was obtained between the temperature distribution in the laser and the bar geometry. The analytical approach revealed features that are usually missing in purely numerical modeling. It was shown that the current density across the active region is nonuniform, and that the magnitude of the nonuniformity grows when the resistance of the p-cladding layer decreases. This nonuniform current distribution can lead to higher temperatures at the edges of the laser stripe than in the middle. It was found that mutual heating of the individual lasers in the bar is controlled mainly by the overlap of the temperature fields in the heat spreaders. The theoretical results obtained without any fitting parame...

Optical gain and loss have been measured in type II InAs/GaInSb interband cascade lasers operating in the 3.4–3.6 μm wavelength range. The maximum temperature of cw operation was found to be limited by strong gain saturation due to active region overheating, while the temperature increase of the total optical loss was relatively small. In devices with a longer lasing wavelength and a thinner substrate-side cladding layer, a strong periodic modulation of the optical gain spectra was observed. This effect is consistent with resonant optical leakage into the substrate.

Continuous-wave power of 25 W at 1.47-μm was obtained from a 20-element, 1-cm-wide, one-dimensional diode laser array mounted in a microchannel water-cooled heat sink. The coolant temperature was 16 °C. A two-dimensional array comprising four laser bars achieved a quasi-cw output of 110 W at a wavelength of 1.49 μm, with an 8–9-nm full width at half-maximum spectrum width. The coolant temperature was 18 °C. We developed a theoretical model that describes array heating. Thermal resistances of 0.56, 0.4, and 0.34 K/W were experimentally and theoretically determined for arrays with fill factors of 10%, 20%, and 40%, respectively.

Gain in broad-area midinfrared diode “W” lasers (λ=3–3.1 μm) has been measured using lateral mode spatial filtering combined with the Hakki–Paoli approach. The internal optical loss of ≈19 cm−1 determined from the gain spectra was the same for devices with either ten or five period active regions and nearly constant in the temperature range between 80 and 160 K. Analysis of the differential gain and spontaneous emission spectra shows that the main contribution to the temperature dependence of the threshold current is Auger recombination, which dominates within almost the entire temperature range studied (80–160 K).

A wide optical gain spectrum with full width at half maximum Δλ≈350 nm has been demonstrated in midinfrared (2.2–2.5 μm) room-temperature-operated InGa(As)Sb/GaSb quantum-well (QW) ridge-waveguide diode lasers. A QW width of 200 A ensured a small energy separation between the first and second quantized states in the QW. Transitions from both quantized states contributed equally to the overall gain spectrum when the optical loss was optimized.

The effect of hole transport on the performance of infrared light emitting diodes (LED) was investigated. The active area of the LEDs comprised two type-II superlattices with different periods and widths connected in series. Electroluminescence spectra of the devices with different positions of long wave and mid wave superlattice sections were mostly contributed by the superlattice closest to the p-contact. The experimental results indicate that due to suppressed vertical hole transport, the recombination of electrically injected electrons and holes in a type II superlattice LED active region takes place within a few superlattice periods near p-barrier. Possible reason for the effect is reduction of hole diffusion coefficient in an active area of a superlattice LED under bias.

Metamorphic heterostructures containing bulk InAs1−xSbx layers and AlInAsSb barriers were grown on GaSb substrates. The lattice mismatch (up to 2.1%) between the GaSb substrates and the InAsSb layers was accommodated by the growth of GaInSb linearly graded buffers. The 1 μm thick InAsSb0.44 layer with an absorption edge above 9 μm exhibited an in-plane residual strain of about 0.08%. InAs1−xSbx structures with x = 0.2 and x = 0.44 operated as light emitting diodes at 80 K demonstrated output powers of 90 μW and 8 μW at 5 μm and 8 μm, respectively.

We have designed and developed dual wavelength type I quantum well light emitting diodes (LEDs) operating at 2 μm and 3–3.4 μm wavelengths with independently controlled intensities. The room temperature quasicontinuous wave output power was 2.8 mW at 2 μm and 0.14 mW at 3 μm. The design of the dual wavelength structure allows for monolithically integrating LED pixels with different wavelengths opening the way for the fabrication of multiwavelength LED arrays for multispectral and hyperspectral imaging applications.

Mid-IR (λ≈3–3.5 μm) light emitting diodes with quinternary AlInGaAsSb barriers and InGaAsSb strained quantum wells grown on GaSb substrates have been demonstrated. The devices produced a quasi-cw emission power of 0.7 mW at room temperature and 2.5 mW at T=80 K.

The temperature kinetics of InGaAsP∕InP high-power laser diode arrays with different fill factors was studied experimentally and theoretically. It was shown that except during a short initial period, the laser array heating is determined by the heat flux propagation through the heat spreader. To characterize this heating, we developed a two-dimensional analytical model. Experimentally, the temperature change in the active region was obtained by measuring the laser spectrum’s temporal evolution during a single current pulse. Three distinctive periods in the transient heating process were clearly identified—an initial temperature rise, a square-root-of-time dependence of the active-region temperature increase, and an exponential approach of the active-region temperature to its steady-state value. We demonstrated that in the initial period of time, the heat propagates within the laser bar structure, and the laser bar design (fill factor) strongly affects the active-region temperature rise. In the later perio...