Activities
Optimization of Absorption in Mid-Wave Infrared Type-II InAs/GaSb Strained Layer Superlattice (T2SL)
REU Student: Eric Frantz
Graduate Student Mentor: Lilian Acosta, Brianna Klein
Faculty Mentor: Sanjay Krishna, Elena Plis
A. Introduction
Infrared (IR) radiation is the portion of the electromagnetic spectrum between 1 to 300µm and is just outside the visible light spectrum. Interest in IR detectors has focused primarily on two regions of the IR spectrum, the mid-wave IR (MWIR, 3-5µm) and the long-wave IR (LWIR, 8-14µm) [1]. These regions are particularly useful for three reasons: most objects emit IR radiation at room temperature in this spectrum, most chemical species have spectral signatures in the IR regime, and that the atmosphere has a clear transmission in these spectral windows [2].
MWIR IR detectors are found in a wide range of applications, such as: night vision, thermal imaging, chemical spectroscopy, gas sensing, pollution monitoring, optical remote sensing, and medical diagnostics. Important properties of high-performance MWIR detector are high signal-to-noise ratio (SNR), i.e. low noise and high signal levels; high operation temperature; large format and affordable cost.
Currently HgCdTe (MCT) photodetectors are the state-of-the-art devices in MWIR spectral range [1]. Although MCT devices have large quantum efficiency (QE), they are sensitive to small changes in the alloy composition ratio during the epitaxial process, making the spatial non-uniformity a challenging problem for large format focal plane arrays (FPAs). Moreover, MCT detectors are characterized by a low electron effective mass (~0.009 mo) [3] resulting in large dark current due to tunneling especially at longer wavelengths.
Bulk InSb photodetectors are also used for MWIR applications due to their high QE, however they are limited to cryogenic operation temperatures. The necessity of cryogenic cooling increases both size and cost of IR detection systems. Therefore, there is a need for high-performance MWIR detectors operating near or at ambient temperature.
Of the alternatives we focus on photodetection using InAs/GaSb T2SL photodetectors. T2SL offer us high operating temperature, high QE, and are relatively cheap to produce [2]. In addition, the larger effective mass (~ 0.04m0) in T2SL leads to a reduction of tunneling currents compared with MCT detectors of the same bandgap. Moreover, commercial availability of low defect density substrates as well as a high degree of uniformity for III-V growth and processing over a large area also offers technological advantages for the InAs/GaSb T2SL technology
For our study we focus on InAs-GaSb T2SL for application in MWIR detection.
B. Background
B.1. Principle of Semiconductor Photodetection
IR detectors may be divided into two broad classes, thermal and photon detectors. Thermal detectors respond to the heat generated by the absorbed energy of the optical radiation whereas the principal of operation of photon detectors is based on the direct interaction of the optical radiation with the atomic lattice of material.
In thermal detectors (Si-based microbolometers, pyrometers, thermocouples, Golay cells, and superconductors) incident optical radiation increases the temperature of the detector which leads to a change in physical parameters, such as resistance or voltage. Some of the characteristic features of thermal detectors are broad spectrum, room temperature operation, low cost, and slow response. Photon detectors generate free electrical carriers through interaction of photons and bound electrons. The interaction of light and matter produces a parameter change (resistance, conductance, voltage or current) that is detected by an associated circuitry. Photon detectors for
IR are typically realized using III-V and II-VI semiconductors, and can further be divided into intersubband and interband detectors.
Examples of intersubband detectors are quantum well infrared photodetectors (QWIPs) and quantum dot infrared photodetectors (QDIPs), where intersubband transition energies in the conduction band can be tailored to detect infrared radiation. Examples of interband detectors are HgCdTe (MCT), InSb, and InAs/GaSb type-II strained layer superlattice (T2SL) systems.
The principle of operation of photon detector may be further described as follows: when a photon with energy higher than the semiconductor’s bandgap is incident upon an electron in the valence band the photon will be absorbed as energy. After photon absorption the electron is excited from the semiconductor’s valence band to its conduction band and an electron-hole pair is created, with electron in conduction band and the corresponding hole in the valence band from where the electron left. Figure 1 illustrates the absorption of a photon and consequential creation of an electron-hole pair in a pin photodiode.
The energy of a photon is dependent upon its wavelength and the wavelength past which the photon no longer has enough energy to excite an electron is called the cutoff wavelength. The wavelength to energy relation is shown in Equation 1, were h is planks constant, c is the speed of light, and is the wavelength. The condition for photon absorption is shown in Equation 2.The equation for cut-off wavelength can be derived from these two formulas and is shown as Equation 3.
/ (1)/ (2)
/ (3)
While the cutoff wavelength is the limit of photons being able to excite electrons response is also shown to decrease as the photons have lower and lower wavelengths past the cutoff wavelength [4]. Due to this phenomenon optimal light absorption occurs for light of wavelength near, but not past, the cutoff wavelength.
B.2 pin Photodiode Design
The pin photodiode design consists of three regions, a positively doped p region, a negatively doped n region, and a non-intentionally doped (n.i.d.) intrinsic i region. In this design of photodiode the p and n regions act as contacts and the i region acts as the absorption layer. A band diagram for a pin photodiode is shown in Figure 1.
Figure 1 - Heterostructure schematic of pin photodiode and pin photodiode band diagram
For application as a photodetector we operate the diode in reverse bias, with negative voltage applied to the p region and positive voltage applied to the n region. In this configuration carriers generated in the absorption layer are swept away by the electric field towards either the n or p contact and contribute to the photocurrent. Despite the simplicity of pin detector design it has several significant drawbacks. In particular, pin detectors demonstrate high noise due to the large Shockley-Read-Hall (SRH) associated with the depletion layer.
B.3 pBiBn Photodiode Design
The pBiBn photodiode, like the pin photodiode, has n and p contact layers and an n.i.d. absorption layer. Unlike the pin design the pBiBn has an additional two barrier layers. These barrier layer are located on either side of the i layer blocking minority carriers from moving from the contact layers to the absorber region. The barrier located between the p and i region blocks electrons from moving from the p to the i region and is called the electron barrier (EB), and the barrier between the n and i region blocks holes from moving from the n to the i region and is called the hole barrier (HB). The design also reduces the electric field drop across the absorber region, resulting in most of the electric field drop across the barrier layers. The reduction in electric field across the absorber layer leads to a small depletion region, this in turn reduces SRH current. Band-to-band (BTB) and trap-assisted-tunneling (TAT) are also reduced due to the significant reduction in electric field drop across the absorber region [5]. The reduction of minority carrier diffusion, SRH, BTB, and TAT currents reduces the dark current of the devices, reducing the overall noise and increasing its sensitivity [5]. A diagram of the pBiBn photodiode is shown below in Figure 2. For use as a photodetector the pBiBn is used in the same bias configuration as the pin photodiode.
Figure 2 - pBiBn Energy Band Diagram. Reprinted from [5]
B.4 Semiconducting Superlattices
Research into semiconductor superlattices began in 1969 with a proposal by Esaki and Tsu for a one-dimensional potential structure composed of alternating layers thin enough to allow for electron tunneling [6]. The proposed superlattice materials were calculated to have a negative differential resistance. However, superlattice structures remained theoretical as no fabrication techniques where able to demonstrate this quantity. It wasn’t until 1972 when Esaki et al. grew a GaAs-GaAlAs superlattice through molecular beam epitaxy (MBE) that negative resistance was observed [6]. Five years later the first InAs-GaSb SLS was proposed for IR detection by Tsu et al. (1977) [2].
Superlattices work on the principle of quantum confinement. The principle of “quantum confinement describes the discretization of the energy spectrum which occurs when a particle is restricted in one or more dimensions by a potential well whose dimensions are comparable to the Bohr radius or de Broglie wavelength in the material system” [2]. In the superlattice structure the quantum wells are created by the energy bands of the alternating semiconductor. These quantum wells cause the quantum confinement of electrons. As mentioned earlier the layers of semiconductor must be thin enough to allow for electron tunneling between quantum wells. The tunneling of electrons between quantum wells causes the formation of electron and hole minibands [2]. It’s these minibands that act as the conduction and valence bands of the superlattice structure as a whole. Figure 3 shows minibands formed in a GaSb-InAs T2SL band diagram generated by electron tunneling.
There are four different classifications of superlattices based on the layers’ band energies [6]. Type I superlattices have what is called a nested band gap, where the bandgap of one semiconductor is completely enclosed by that of the other semiconductor. Type II-staggered superlattices have the valence band of one semiconductor between the valence and conduction bands of the other, while its valence band is above that of the other semiconductor. Type II-misaligned superlattices have both the conduction and valence bands of one semiconductor above the conduction band of the other semiconductor. Type III superlattices are classified as such when one of the layers of the superlattice is metallic. In our testing we will be focusing on the type II superlattice composed of InAs-GaSb.
It has been shown that the introduction of strain into a superlattice, by having mismatched lattice constants of the two materials, causes the splitting of valence bands into light and heavy hole components. This leads to larger electron effective mass which results in reduced diode tunneling currents, reducing the dark current [7].
By using SLS material for photodetection we are able to optimize photodiode design by engineering the band structure. As mentioned previously optimal absorption occurs near the cutoff wavelength. Through band engineering SLS materials offer bandgaps that can be very close to the cutoff wavelength. Band engineering also allows us to create barriers better suited for our needs for a more optimal device.
Figure 3 - GaSb / InAs T2SL diagram
C. Research Objective
The objective of this study is to optimize the absorption in MWIR InAs/GaSb T2SL photodetectors by varying (1) thickness of detector absorber region, (2) T2SL period thickness, and (3) strain between T2SL and GaSb substrate.
By researching what effect these parameters have on absorption we hope to further optimize T2SL materials for use in MWIR photodetectors.
The drive behind this study is the potential in the T2SL material for IR MWIR detection against existing technologies. T2SL detectors offer us high operating temperatures, high quantum efficiency, spectral tuneability, low cost, high responsivity, normal incidence absorption, and multicolor operation. Comparison of T2Sl photodetectors against existing technologies is shown below in Table 1.
Characteristic / HgCdTe Photodiodes / InSb Photodiodes / Type II Strained-Layer Superlattice / Quantum Well Infrared Photodetector / Quantum Dot Infrared PhotodetectorNormal incidence absorption / yes / yes / yes / no / yes
Operation Temperature / high / low / high / low / high
Spectral Tunability / no / no / yes / yes / yes
High Quantum Efficiency / yes / yes / yes / no / no
High Responsivity and Detectivity / yes / yes / yes / no / no
Multicolor Operation / yes / no / yes / yes / yes
Homogeneity / no / yes / yes / yes / no
Material Yield / low / low / high / high / high
Table 1- Comparison of T2SL against other photodetector technologies
D. Methodology
For our study we use InAs-GaSb T2SL to study the effects of absorption layer thickness, T2SL composition, and interfaces between layers on T2SL detectors performance for use in the MWIR spectrum. The study is divided into three parts, one for each parameter:
- Part 1: Effects of absorption layer thickness on detector performance.
- Part 2: Effects of T2SL composition on detector performance.
- Part 3: Effects of strain between T2SL detector structure and GaSb substrate on detector performance.
For the first part of the study, MWIR homojunction pin and pBiBn architectures based on 8 monolayers (MLs) InAs/8 MLs GaSb T2SL were utilized. Thickness of non-intentionally doped absorber layer was 1.5 um, 2.5 um, and 3.5 um.
For the second part of the study, MWIR detectors with pBiBn designs were grown. The thickness of absorber region and the T2SL constituent InAs layer thickness were kept the same (1.5 um and 8 MLs, respectively) whereas the T2SL constituent GaSb thickness was varied as 6 MLs, 8 MLs, and 10 MLs.
In the third part of the study, pBiBn T2SL detectors with 8 MLs InAs/8MLs GaSb T2SL composition and different thickness of InSb strain compensation layer were intended to be grown. However, due to the issues with MBE reactor this part of study was not realized.
Table 2 shows a summary of devices for the first two parts of the study, and their variables. Figures 4 and 5 illustrate the heterostructre schematics of pin and pBiBn InAs-GaSb T2SL detectors, respectively.
Completion of the study requires the following steps:
1. Growth of InAs – GaSb T2SL detector samples.
2. Photoluminescence (PL) and high-resolution x-ray diffraction (HRXRD) measurements of as-grown T2SL samples to evaluate relative material quality.
3. Fabrication
a. Mesa definition (etching)
b. Contact Metal Deposition
c. Sidewalls Passivation
- Spectral Response testing
- Dark current measurements
- Responsivity Testing
7. Dark Current Testing
8. Spectral Response Testing