Periodic Reporting for period 4 - iTPX (In-cavity thermophotonic cooling)
Okres sprawozdawczy: 2020-04-01 do 2021-03-31
The possibility that a light emitting diode (LED) cools down as a result of light emission has been acknowledged already half a century ago. This effect is known as electroluminescent cooling (ELC), but it has never been observed close to normal operating conditions of LEDs. This is despite the fact that the material parameters reported for common III-V semiconductors are sufficient to suggest that presently available materials are suitable for ELC. Harnessing ELC would allow not only fabricating extremely efficient LEDs, but also provide a promising solid state optical heat pumping mechanism that could ideally e.g. overcome some of the efficiency limitations of Peltier coolers. This project aims at observing and developing ELC in one of the most favourable experimental setups designed for the purposes of this project: an intracavity LED structure where an LED is enclosed within the same semiconductor epistructure as a photodetector (PD) used to electrically measure the LED performance. This approach allows eliminating some of the key challenges of conventional LEDs in extracting light from a high refractive index semiconductor as well as simple means to measure the performance of the LED.
The main results from the project corroborate the expectation that present technologies already allow reaching ELC at practical operating conditions. The devices fabricated in iTXP indicated that local cooling in an LED is already achievable in the device prototypes studied in the project. Since these devices still circumvented some of the challenges met in real light emitters and cooling devices, the results suggest that the remaining bottlenecks in developing new optical cooling technologies are more technical than fundamental in nature, and require 'simply' more efficient designs to extract light and to reduce the resistance of the devices. Both of these are deemed to be feasible using the know-how and fabrication capabilities gained in the project.
The main results from the project corroborate the expectation that present technologies already allow reaching ELC at practical operating conditions. The devices fabricated in iTXP indicated that local cooling in an LED is already achievable in the device prototypes studied in the project. Since these devices still circumvented some of the challenges met in real light emitters and cooling devices, the results suggest that the remaining bottlenecks in developing new optical cooling technologies are more technical than fundamental in nature, and require 'simply' more efficient designs to extract light and to reduce the resistance of the devices. Both of these are deemed to be feasible using the know-how and fabrication capabilities gained in the project.
The project started by designing, fabricating and characterizing the so called double diode structures (DDS), which are intracavity devices consisting of a double heterojunction LED and a homojunction photodetector enclosed within the same crystal structure fabricated on a GaAs substrate. The characterization of the first generation of DDS confirmed that the DDS itself is a promising platform for studying ELC, due to its small internal losses and other similar unidealities. Since the first reported DDS devices we have improved the design and fabrication methods of the device, which has allowed to increase the coupling quantum efficiencies of the device from the initial ~50% to about 70-80%. The general expectation is that the threshold for ELC can be reached when the quantum efficiency of light emission is around 80-90%. Our results suggest that in our best DDS structures the LED component itself already reached the ELC threshold, but the unoptimized spectral overlap between the emitter and photodetector of the DDS prevented the direct measurement of the highest quantum efficiencies for now. With respect to reaching the cooling regime, however, the spectral overlap is non-consequential, and only affects the overall system efficiency.
The project mainly advanced beyond the state of the art by introducing the DDS that enabled more direct study of local cooling effects and also produced LEDs that at the time had the highest directly measured quantum efficiencies. Analysis of the devices has suggested that the fabricated LEDs already pass the ELC threshold, while direct confirmation of this conclusion still requires further development to allow observing a measurable temperature difference. During the project we have also developed simulation tools for more detailed analysis of the results, allowing more accurate studies of the limits and requirements of reaching the ELC regime. Overall, the project has generated a substantial amount of new information on the possibility and requirements of ELC.