Dark Current Suppression, Optical Performance Improvement and High Frequency Operation of InAs/GaSb and InAs/InAsSb Type-II Superlattices-based Infrared Devices

Chevallier, Romain Francois

doi:https://doi.org/10.21985/n2-bk57-3707

Work

Dark Current Suppression, Optical Performance Improvement and High Frequency Operation of InAs/GaSb and InAs/InAsSb Type-II Superlattices-based Infrared Devices

Public

Download PDF

Download All Files (.zip)

Every object emits radiation depending of their temperature. Objects between 300K to 100K emit radiation in the infrared range (1-12µm). These radiations cannot be seen by the human eyes. Infrared detectors find applications in many aspects of life, from night vision and target tracking for homeland security and defense, to non-destructive failure detection in industry, chemical sensing in medicine, and free-space communication. Currently, the dominant technologies of photodetectors based upon HgCdTe (MCT) and Quantum Well Photoconductors are experiencing many limitations. Under this circumstance, the Type-II superlattices which have been intensively studied recently appear to be an excellent candidate to give breakthroughs in the infrared technology. GaSb-based type-II superlattice photodetectors have the potential to compete with MCT detectors. GaSb-based superlattices are composed of either InAs/(AlSb)/GaSb or InAs/(AlAsSb)/InAsSb. In both cases, superlattice layers thicknesses and compositions can be modulated to change many properties of the materials. First, it can mainly tune the bandgap from about 1.0 µm to semi-metals. The thicknesses and composition also determine the energy bands, which can be set to suppress Auger recombination, one of the main limitation for MCT photodetectors. This possibility theoretically gives GaSb-based type-II superlattice the potential to overcome the performances of MCT photodetectors. To improve photodetectors’ performance, two possibilities come in: to reduce the dark current density of the sample and to improve the optical response. To decrease the dark current, it is first necessary to know what the origin and limitations of the dark current density are. LWIR InAs/(AlSb)/GaSb superlattice photodetectors have been found to be limited by generation-recombination (G-R) because of Schockley-Read-Hall Ga-related native defects in GaSb creating recombination centers and reducing the carrier lifetime, and limited by surface leakage current. InAs/InAsSb superlattice with high Sb concentration has been found to have much longer carrier lifetime than InAs/GaSb superlattice, which suppresses the G-R current. G-R current being proportional to the volume of the depletion region, fabricating a microjunction structure reduces the G-R current without decreasing the optical performance. This increases the overall performance of the photodetector. In addition, the surface of the photodetectors mesa needs to be etched, cleaned and treated properly to avoid creating a surface channel which could shorten the device and reduce significantly the device’s performance. This is all the more important when dealing with very small size mesas to have small pitch FPA. To improve the optical performance, several possibilities can be explored. In the case of SWIR photodetectors, the detectors tend to be in a photon starving mode. Indeed, the amount of photon in the SWIR region is low because it corresponds to objects around 1000oC. Usually, SWIR photodetectors are grown with a PiN structure on a GaSb wafer, with a double etch stop layer, InAsSb and GaSb. However, this limits the sample absorption range between 1.7µm and 2.5µm, below the bandgap energy of GaSb. To increase the spectral range to improve the number of photons absorbed by the photodetectors, the GaSb etch stop layer is replaced by an AlAsSb based etched stop layer, which has a larger bandgap than GaSb and increases the absorption up to 0.8 µm, which allows to have better image quality and room temperature imaging. Another way to improve performance is to use gain-based devices such as avalanche photodiodes and phototransistors. These types of devices are especially important in the short wavelength infrared region as there are very few photons in that wavelength range or for applications that require very low integration times, but the principle works for all cut-off wavelengths. The principle is to enhance the signal with a gain mechanism. This gain will also increase the dark current as well, generally in the same order of magnitude as the signal, which will permit to increase the detectivity of the device. This is a common way that is used in telecommunications and for light detection and ranging (LIDAR) systems. The state of the art technology for such system is based on InGaAs devices grown on InP. They have very good performance but, without significant lattice mismatch related growth issues, they are limited to a single cut-off wavelength, which limits the wavelength multiplexing that can be done. Using type-II superlattice has advantages as a large range of cut-off wavelength can be achieved and wavelength multiplexing can therefore be expanded to a larger range. Short wavelength infrared phototransistors were demonstrated with gain of several hundred at room temperature, yielding an increase of detectivity compared to traditional photodiodes of about 20 times. As such devices are intended to be used at high frequency for telecommunications, data transmission and LIDAR applications, I designed, developed and set-up device layout designs, processing steps and fabrication methods, as well as the measurement systems to be able to characterize such devices. InAs/GaSb/AlSb type-II superlattice infrared phototransistors have demonstrated to have maximum detection frequency beyond 20 GHz. However, from around 30 MHz to reaching noise level around 20 GHz, the signal decays significantly. There can be mainly reasons for that, including gain mechanism, photocarrier extraction time, applied bias, device size among others. To eliminate at least one parameter, the gain mechanism), a study was performed on a photodetector with similar absorption region design but without gain. -3dB cut-off frequency of 12 GHz was demonstrated with relatively flat response from 30 MHz to 5 GHz, with potential to reach better performance using larger electric fields (larger applied bias and thinner devices). As the fabrication design is identical as for the phototransistors, it is concluded the gain mechanism is the key limiting factor in the high frequency performance. The next steps in this project are to continue to study photodetectors to improve the base performance of the device layout design and to improve the gain mechanism high frequency performance.

Creator