Electrical and Computer Engineering ETDs

Publication Date

6-9-2016

Abstract

Infrared detection has proven to be an essential capability in a variety of fields, primarily the medical and defense fields. The current industry standard material is HgCdTe, which has been developing since its inception in the late 1950s. SLS materials are created from very thin layers, on the order of several to several tens of atomic monolayers, of two different bulk materials. These layers are alternated periodically, which results in entirely new material properties. The most common variety of SLS is the InAs-GaSb SL. This material shows great promise, but has not yet exceeded the performance of MCT detectors. This is due to the fact that the performance of the InAs-GaSb SL is limited by its Shockley-Reed-Hall (SRH) carrier lifetime. A new type of SLS, called the Ga-free SL, has demonstrated a much longer SRH lifetime, and has the potential to exceed the performance of the InAs-GaSb SL. The focus of this dissertation is the use and optimization of the Ga-free SLS in infrared detectors. This optimization is accomplished though a series of studies that focus on several aspects of growing Ga-free SL-based detectors, including 1. Initial device characterization 2. Doping level in the detector absorber region 3. As:In BEP ratio 4. MMBE-based InAs-InSb vs. ternary InAsSb growth 5. InAs-on-InAsSb interface optimization The first electrical characterizations of a Ga-free SL detector are reported from the initial device. This device was a MWIR PIN design created from a 14ML InAs - 12ML InAs_{0.65}Sb_{0.35} superlattice. The results of this showed an unexpectedly high level of dark current. The cause of this current was investigated using a curve-fitting program provided by IRNova, which showed that trap-assisted tunneling may be the cause of this result. The next study compared varied levels of doping in the absorber region of a Ga-free SL MWIR PIN detector. The designs for these devices were similar to the initial device, but the InAsSb composition was changed to be InAs_{0.81}Sb_{0.19}, which provided a better lattice match to the GaSb substrate. The absorber regions were doped at NID (reference), 5x10^{15}, 1x10^{16}, and 3x10^{16} cm^{-3} p-type. P-type doping was chosen to offset the intrinsic n-type behavior of the Ga-free material. The results showed that the highest doping level, at 3x10^{16} cm^{-3}, resulted in the lowest dark current (0.00248 vs. 0.08353 A/cm^{2} for the best device from each sample). The quantum efficiency for this device was slightly reduced from the reference sample, (37.4% vs. 42.4%), as a tradeoff for the reduced dark current levels. The dark current levels of these devices were reduced as the doping increased, thereby indicating G-R dominance. The BEP ratio study tested several As:In ratios with the purpose of optimizing the growth conditions for an MWIR Ga-free SL. This study performed electrical characterization on three samples with BEP ratios of 5:1, 6:1, and 7:1. These devices were grown using the same MWIR PIN design used in the doping study. Both quantum efficiency and dark current levels from these devices indicated that the 7:1 As:In BEP ratio was an improvement on the lower ratios. The fourth study discussed in this dissertation compared an MWIR Ga-free SL grown using an MMBE-grown InAs-InSb alloy in place of the InAsSb ternary material with an MWIR Ga-free reference sample. The purpose of this was to determine if the use of the MMBE technique during growth had adverse effects on the electrical performance of the resulting detectors. However, the dark current of the MMBE-based device was lower than the reference sample (0.0853 vs. 0.1227 A/cm^{2} at -10mV bias and 80K), and the quantum efficiency was higher (39.7% vs. 24.1% at -10mV bias and 80K). This indicated that MMBE may improve the resulting devices, rather than detract from their performance. The final study tested a series of interface layers placed at the InAs-on-InAsSb interface. This was intended to minimize the diffusion of Sb from the InAsSb layer into the following InAs layer. These samples were the same design as used in previous studies, with the exception that the InAsSb layers were all replaced with MMBE InAs-InSb (as tested in the MMBE study). The interface layers tested were a 1 second growth interrupt, a 1 second As soak, a 1 second Al flash, and a 1ML AlSb layer. A reference sample with no special interface layer was also grown. The results showed that the reference, the As soak, and the AlSb layer all showed comparable performance, whereas the Al flash and growth interrupt reduced the quality of the resulting devices. From the resulting data, further study of the use of an As soak is encouraged.

Sponsors

U.S. Government

Document Type

Dissertation

Language

English

Degree Name

Electrical Engineering

Level of Degree

Doctoral

Department Name

Electrical and Computer Engineering

First Committee Member (Chair)

Krishna, Sanjay

Second Committee Member

Balakrishnan, Ganesh

Third Committee Member

Steenbergen, Elizabeth

Fourth Committee Member

Sheik-Bahae, Mansoor

Share

COinS