Search

We use an ensemble Monte Carlo transport approach to calculate and compare the impact ionization and avalanche photodiode excess noise characteristics in three materials—a band-engineered InAlAs/InAsSb type-II superlattice, bulk InAs, and HgCdTe—all with an identical bandgap of 370 meV at 250 K. The electronic band structures and energy–momentum conservation conditions are used to calculate the impact ionization rates, carrier histories, multiplication gains, and excess noise characteristics. The calculated impact ionization coefficients and excess noise factors indicate a single carrier species multiplication in all three materials under low applied electric fields. We find the ratio of impact ionization coefficients to be k = 7 × 10 − 4 for InAs and 3 × 10 − 4 for HgCdTe under an applied field of 50 kV/cm, and the superlattice to be k < 10 − 6 at fields up to 400 kV/cm. The bulk materials experience avalanche breakdown as the applied field increases, transitioning to Geiger mode behavior at gains above 10 3 for InAs and 10 4 for HgCdTe. However, this breakdown is absent from the superlattice at the highest fields considered in this study due to hole confinement, indicating superior performance compared to the bulk materials. Our results demonstrate the role of superlattice band engineering in designing quality avalanche photodiode materials. [ABSTRACT FROM AUTHOR]

For short-wavelength infrared (SWIR) avalanche photodiodes, a separate absorption, charge, and multiplication design is widely used. AlInAsSb on an InP substrate is a potential multiplication layer with a lattice match to absorber candidates across the SWIR. Our new measurements demonstrate that AlInAsSb on InP is a promising multiplier candidate with a relatively low dark current density of 10−4 A/cm2 at a gain of 30; a high gain, measured up to 245 in this study; and a large differentiation of electron and hole ionization leading to a low excess noise, measured to be 2.5 at a gain of 30. These characteristics are all improvements over commercially available SWIR detectors incorporating InAlAs or InP as the multiplier. We measured and analyzed gain for multiple wavelengths to extract the ionization coefficients as a function of an electric field over the range 0.33–0.6 MV/cm. [ABSTRACT FROM AUTHOR]

Background doping polarity is a critical design parameter for the performance of many optoelectronic devices, including avalanche photodiodes. We have applied a technique by using capacitance–voltage (CV) measurements on double mesa structures with a p-i-n or n-i-p homojunction to determine the background polarity type of the unintentionally doped intrinsic region. Because CV measurements scale with the size of the mesa, they support design flexibility in producing variable-sized top and bottom mesa diameters. In this work, we grew, fabricated, and tested AlGaAsSb and AlInAsSb random alloy double mesa p-i-n structures and undertook CV measurements at 295, 150, and 77 K. It was found that the capacitance varied with the top mesa diameter for both material systems, and not the bottom mesa diameter, indicating that the unintentionally doped intrinsic region is n-type in nature. [ABSTRACT FROM AUTHOR]

We demonstrate low noise random alloy (RA) Al0.85Ga0.15AsSb (hereafter AlGaAsSb) avalanche photodiodes (APDs) nearly lattice-matched to InP substrates. In contrast to digital alloy (DA), RAs are manufacturable due to the ease of growth. The 910 nm-thick RA AlGaAsSb was grown at a low temperature around 450 °C to mitigate phase separation by suppressing surface mobility of adatoms. The high quality of the RA AlGaAsSb material was verified by x-ray diffraction, Nomarski, and atomic force microscope images. Capacitance–voltage measurement found that the background doping concentration was 6–7 × 1014 cm−3, indicating very low impurity density in the RA AlGaAsSb material. Current–voltage measurements were carried out under dark condition and 455 nm laser illumination at room temperature. The breakdown occurs at −58 V. The dark current density at a gain of 10 was found to be 70 μA/cm2. This value is three orders of magnitude lower than previously reported DA AlAs0.56Sb0.44 APDs [Yi et al., Nat. Photonics 13, 683 (2019)], one order of magnitude lower than DA AlGaAsSb [Lee et al., Appl. Phys. Lett. 118, 081106 (2021)], and comparable to RA AlInAsSb APDs [Kodati et al., Appl. Phys. Lett. 118, 091101 (2021)]. In addition, the measured excess noise shows a low k (the ratio of impact ionization coefficients) of 0.01. These noise characteristics make the RA AlGaAsSb multiplier suitable for commercial applications, such as optical communication and LiDAR systems. [ABSTRACT FROM AUTHOR]

Digital alloy Al0.85Ga0.15As0.56Sb0.44 and random alloy Al0.79In0.21As0.74Sb0.26 avalanche photodiodes with a 1-μm multiplication layer exhibit low excess noise under 543-nm laser illumination comparable to Si avalanche photodiodes. This has motivated a study of the optical characteristics of these materials. The absorption coefficients and complex refractive indices were extracted via variable-angle spectroscopic ellipsometry. Features of three semiconductor layer fitting approaches are compared, and a Kramers–Kronig-consistent basis spline function was chosen due to its flexibility and accuracy of approximating optical constants of quaternary materials. The external quantum efficiency has been calculated based on the extracted absorption coefficients and is shown to agree well with the measured external quantum efficiency. [ABSTRACT FROM AUTHOR]

We report the gain, noise, and dark current characteristics of random alloy Al0.79In0.21As0.74Sb0.26 (hereafter AlInAsSb)-based avalanche photodiodes (APDs) on InP substrates. We observe, at room temperature, a low excess noise corresponding to a k value (ratio of impact ionization coefficients) of 0.018 and a dark current density of 82 μA/cm2 with a gain of 15. These performance metrics represent an order of magnitude improvement of the k-value over commercially available APDs with InAlAs and InP multiplication layers grown on InP substrates. This material is also competitive with a recently reported low noise AlAsSb on InP [Yi et al., Nat. Photonics 13, 683 (2019)], with a comparable excess noise and a room temperature dark current density almost three orders of magnitude lower at the same gain. The low excess noise and dark current of AlInAsSb make it a candidate multiplication layer for integration into a separate absorption, charge, and multiplication layer avalanche photodiode for visible to short-wavelength infrared applications. [ABSTRACT FROM AUTHOR]

We report on the demonstration of Al0.85Ga0.15As0.56Sb0.44 (hereafter, AlGaAsSb) avalanche photodiodes (APDs) with a 1000 nm-thick multiplication layer. Such a thick AlGaAsSb device was grown by a digital alloy technique to avoid phase separation. The current-voltage measurements under dark and illumination conditions were performed to determine gain for the AlGaAsSb APDs. The highest gain was ∼ 42, and the avalanche initiation occurred at 21.6 V. The breakdown voltage was found to be around −53 V. The measured dark current densities of bulk and surface components were 6.0 μA/cm2 and 0.23 μA/cm, respectively. These values are about two orders of magnitude lower than those for previously reported 1550 nm-thick AlAs0.56Sb0.44 APDs [Yi et al., Nat. Photonics 13, 683 (2019)]. Excess noise measurements showed that the AlGaAsSb APD has a low k of 0.01 (the ratio of electron and hole impact ionization coefficients) compared to Si APDs. The k of the 1000-nm AlGaAsSb APD is similar to that of the thick AlAsSb APDs (k ∼ 0.005) and 5–8 times lower than that of 170 nm-thick AlGaAsSb APDs (k ∼ 0.5–0.8). Increasing the thickness of the multiplication layer over 1000 nm can also reduce k further since the difference between electron and hole impact ionization coefficients becomes significant in this material system as the thickness of the multiplication layer increases. Therefore, this thick AlGaAsSb-based APD on an InP substrate shows the potential to be a high-performance multiplier that can be used with available short-wavelength infrared (SWIR) absorption layers for a SWIR APD. [ABSTRACT FROM AUTHOR]

We present a method of determining the background doping type in semiconductors using capacitance–voltage measurements on overetched double mesa p–i–n or n–i–p structures. Unlike Hall measurements, this method is not limited by the conductivity of the substrate. By measuring the capacitance of devices with varying top and bottom mesa sizes, we were able to conclusively determine which mesa contained the p–n junction, revealing the polarity of the intrinsic layer. This method, when demonstrated on GaSb p–i–n and n–i–p structures, concluded that the material is residually doped p-type, which is well established by other sources. The method was then applied to a 10 monolayer InAs/10 monolayer AlSb superlattice, for which the doping polarity was unknown, and indicated that this material is also p-type. [ABSTRACT FROM AUTHOR]

We investigate carrier localization in the InAs/AlSb type-II superlattice (T2SL) material system using temperature- and excitation power (Iex)-dependent photoluminescence (PL). Evidence of carrier localization in T2SLs was observed by an S-shaped temperature dependence of the PL peak position. Analysis of the Iex-dependent PL at various temperatures also shows the existence of carrier localization in the T2SLs. The thermal activation energies in T2SLs were extracted to identify the nonradiative recombination mechanisms and the possible origins of localized states. We found that there are two thermal activation energies, E1 = 8.2–1.2 meV and E2 = ∼60 meV at various Iex. We interpret E1 as a thermal activation energy that comes from Anderson localization, associated with roughness due to As2 diffusion into the interfaces. This is because the extracted E1 values are comparable to the exciton binding energy of localization in various quantum structures. Carrier trapping at a state in the InSb interfacial layer (Tamm state) may account for the origin of E2. Based on previous reports, we believe that the 60 meV state might be a Tamm state if we consider thickness variations in the InSb interfacial layer for the T2SLs. [ABSTRACT FROM AUTHOR]

We report on engineering impact ionization characteristics of In0.53Ga0.47As/Al0.48In0.52As superlattice avalanche photodiodes (InGaAs/AlInAs SL APDs) on InP substrate to design and demonstrate an APD with low k-value. We design InGaAs/AlInAs SL APDs with three different SL periods (4 ML, 6 ML, and 8 ML) to achieve the same composition as Al0.4Ga0.07In0.53As quaternary random alloy (RA). The simulated results of an RA and the three SLs predict that the SLs have lower k-values than the RA because the electrons can readily reach their threshold energy for impact ionization while the holes experience the multiple valence minibands scattering. The shorter period of SL shows the lower k-value. To support the theoretical prediction, the designed 6 ML and 8 ML SLs are experimentally demonstrated. The 8 ML SL shows k-value of 0.22, which is lower than the k-value of the RA. The 6 ML SL exhibits even lower k-value than the 8 ML SL, indicating that the shorter period of the SL, the lower k-value as predicted. This work is a theoretical modeling and experimental demonstration of engineering avalanche characteristics in InGaAs/AlInAs SLs and would assist one to design the SLs with improved performance for various SWIR APD application. [ABSTRACT FROM AUTHOR]

Digital alloy Al0.85Ga0.15As0.56Sb0.44 and random alloy Al0.79In0.21As0.74Sb0.26 avalanche photodiodes with a 1-μm multiplication layer exhibit low excess noise under 543-nm laser illumination comparable to Si avalanche photodiodes. This has motivated a study of the optical characteristics of these materials. The absorption coefficients and complex refractive indices were extracted via variable-angle spectroscopic ellipsometry. Features of three semiconductor layer fitting approaches are compared, and a Kramers–Kronig-consistent basis spline function was chosen due to its flexibility and accuracy of approximating optical constants of quaternary materials. The external quantum efficiency has been calculated based on the extracted absorption coefficients and is shown to agree well with the measured external quantum efficiency. [ABSTRACT FROM AUTHOR]