Investigation of Superlattice Surface Characteristics

Strained-layer superlattices (SLS) are a type of meta-material which exhibit tunable properties and are expected to provide a host of benefits for infrared (IR) photodetectors, most notably an arbitrarily long cutoff wavelength. This represents an opportunity for III-V absorber materials to be extended into mid-wave infrared (MWIR) and long-wave infrared (LWIR) applications. However, noise measurements remain too high in these IR photodetectors for SLS absorbers to be competitive with current MWIR and LWIR technologies. One of the major contributors to noise is surface leakage, where electrical conduction occurs along exposed semiconductor surfaces. While surface leakage occurs even in relatively large detector devices, the detrimental effects will be exacerbated as detector size decreases, and therefore the relative surface area and conduction pathways of surface leakage increases. An investigation into the surface characteristics of SLS structures through modeling, growth, processing, and characterization was conducted. The outcome of this investigation is a greater understanding of the physical mechanism of surface currents in SLS structures, and an IR photodetector design which utilizes the unipolar barrier architecture to suppress all surface leakage currents, and therefore reduce noise to competitive levels. A SLS model was built and refined against grown structures. This model is capable of predicting band-edge energies, among other characteristics. Using this model, trends of surface carrier conductivity and drift current activation energy with constituent layer thicknesses were found. Five SLS structures, all targeting a cut-off wavelength of 4 µm at 150 K, and each with more GaSb than the last, were proposed for further investigation. A model for the calculation of equilibrium band diagrams was also developed. This was used to design three devices, each with their own predicted dominant surface dark current mechanism. The first design was then used to investigate the inherent surface characteristics of the SLS structures. The other two designs are the subject of future study. Improvements in the growth of the constituent layer materials InAs and GaSb by molecular beam epitaxy (MBE) were developed. The homoepitaxy growth techniques of InAs were refined and improvements were evidenced by the competitive performance of unipolar barrier devices (nBn’s) grown with these techniques. A growth study was conducted with the goal of finding optimal growth conditions for InAs, and the results were translated to the growth of InAs in a Ga-containing SLS structure. The homoepitaxy growth techniques of GaSb were refined, with a focus on substrate preparation, oxide removal, and buffer growth. This is important as it is the surface that all of the Ga-containing SLS structures are grown on. Enhancements in the growth of lattice-matched InAsSb was also explored. This material was then used to test novel surface characterization techniques. The growth of the Ga-containing SLS structures by MBE was explored. A useful growth technique was identified, which can easily be applied to a broad range of SLS structures, with minimal adjustment. Three key growth variables were focused on: growth temperature, interface engineering, and molecular flux ratios. Three essential tools for feedback when exploring the growth design space were used, and novel techniques were developed for each. Reflection high-energy electron diffraction (RHEED) was used to calibrate the surface temperature during growth. A short pre-growth calibration yields a reliable surface temperature indicator at the desired growth temperature. High-resolution X-ray diffraction (HRXRD) was used to determine the period thickness of the SLS as well as the strain. A novel method for estimating the amount of InSb present in the SLS structure was developed. Photoluminescence was used to measure the effective bandgap of the SLS structures at low temperatures. The Varshni parameters calculated with the model from the SLS model was used to estimate the cut-off wavelength of the SLS structures at 150 K. By altering the growth conditions, a wide array of structures can be explored through a single growth. The surface characteristics of etched sidewalls was investigated for both bulk material (InAsSb), and Ga-containing SLS. A novel method to easily determine the surface recombination velocity of an etched sidewall was presented, and the results from the analysis of two InAsSb samples with opposite bulk conductivity types were explored. The surface Fermi level of InAsSb was determined to be located above the conduction band edge, which matches the theory and experimental evidence available to date. A similar analysis was performed on the five SLS structures proposed using the SLS model, combining two existing analysis techniques. Evidence was found that the conduction- and valence-band energies on the surface, relative to the surface Fermi level, can be adjusted by changing the constituent layer thicknesses in the SLS structure. This provides an additional degree of freedom to the device engineer when designing an infrared photodetector. Employed correctly, it can be used to eliminate surface drift currents and subsurface g-r current. Such a device would deliver the true promise of SLS structures. As the material quality improves with more research, the device will eventually reach its fundamental Auger limit, which is predicted to be lower than HgCdTe absorbers. Once this limit is reached, such a detector would be the new state-of-the-art for high-performance infrared photodetectors in MWIR and LWIR, and perhaps beyond.