Electroluminescent Cooling in III–V Intracavity Diodes: Practical Requirements

Abstract

Recent studies of electroluminescent cooling in III-V structures demonstrate the need to understand better the factors affecting the efficiency of light emission and energy transport in light-emitting diodes (LEDs). In this work, we establish the physical and operational requirements for reaching the efficiencies needed for observing electroluminescent cooling in III-V intracavity double diode structures at high powers. The experimentally-validated modeling framework used in this work, coupling the drift-diffusion charge transport model with a photon transport model, indicates that the bulk properties of the III-V materials are already sufficient for electroluminescent cooling. Furthermore, the results suggest that the bulk power conversion efficiency of the LED in the devices, that allowed the experimentally measured record high coupling quantum efficiency of 70%, already exceeds 115%. However, as shown here, direct observation of electroluminescent cooling by electrical measurements still requires a combination of a more efficient suppression of the non-radiative surface recombination at the LED walls and the reduction of the detection losses in the photodetector of the intracavity structures.

On the temperature dependence of the efficiency of electroluminescence

“…According to simulations and as presented in previous works, the shoulder mentioned above indicates a transition from a surface(mesa edge)-recombination to a bulk-and interface-recombination dominated region 10,11,18 . The shoulder becomes significantly more visible and shifts to slightly larger currents as temperature increases (see Fig.…”

“…This result suggests that, in order to observe high CQEs in our DDSs, and thus observe ELC in the present structures, temperature ranges different from the RT must be explored. The IQE of the LEDs can depend strongly on the extent of surface recombination 11 . This is especially true for the 1-mm and the worst 0.5-mm diameter devices, which show a substantial decrease in the peak LED IQE from 250K to 300K.…”

“…In this paper, we study the temperature dependence of the light emission and detection (in the range 100 K -300 K) and its implications on high-power ELC using an optically coupled emitter and absorber system 11,17,18 . This configuration, where a double-heterojunction (DHJ) LED acts as the emitter of photons and a GaAs homojunction photodetector (PD) as the absorber has produced this far the highest directly measured quantum efficiencies for GaAs LEDs 10 , making the structure well suited for studying the fundamental properties of devices intended for ELC, partly due to photon collection within the structure leading to alleviating some of the common light-extraction difficulties.…”

Observation of local electroluminescent cooling and identifying the remaining challenges

“…Most imporatantly, (i) for the study of the spatial distribution of the LED emission the top LED contact contained holes for direct detection of the EL emission intensity, 5 and (ii) to estimate the DDS efficiency, an omnidirectional reflector (ODR) pattern was made for the top LED contact. 15 In case of the ODR a large part of the top contact p-GaAs layer was etched away and replaced with silicon nitride providing much better reflectivity. The contact to the GaAs was made through openings in the silicon nitride layer and varied from 5 to 100 % of the top LED area.…”

“…More detailed analysis, performed in section 3.3, suggests that this shoulder is produced by the saturation of the surface current that is an alternative way for non-radiative recombination of charge carriers. 14,15 One of the most obvious parameters containing information about the energy transfer between the coupled diodes that can be extracted from experimentally measured I − V -I − V curves is the ratio of PD and LED currents. We define the coupled quantum efficiency (CQE) as η CQE = I 2 /I 1 , where I 1 is LED injected current and I 2 is the photocurrent generated in the PD under short circuit conditions.…”