United States

Abstract Body: A parametric analysis of growth conditions on the structural properties of N-Polar InGaN films was studied. Three series of N-polar InGaN films were grown on bulk GaN substrates by plasma-assisted molecular beam epitaxy (PAMBE) to study the effect of substrate temperature, Ga, and In fluxes. For all series, the surface morphology was measured by atomic force microscopy (AFM), and the composition and strain were determined by high resolution X-ray diffraction (HRXRD) and XRD reciprocal space mapping (RSM). All the samples in this study were 350 nm thick. First, a series of samples was grown at temperatures varying from 620 oC to 690 oC. The Ga flux for all samples in this series was kept at 3.8E-7 Torr and the In flux was increased by temperature to ensure metal-rich growth conditions. The surface morphology improved significantly by increasing the temperature, and the indium content was reduced as expected. The samples grown at 670 oC and 690 oC showed surprisingly smooth surface morphologies with surface roughness &lt;0.7 nm which has never been reported before on such thick InGaN films. An In content of 16.5% and 13.6% was measured on samples grown at 670 oC and 690 oC, respectively. These temperatures are significantly higher than growth temperatures (&lt;600 oC) typically used for the growth of InGaN films on Ga-polar GaN. In the next step, samples were grown with In fluxes of 3E-7 Torr and 4E-7 Torr. The Ga flux was kept at 3.8E-7 Torr and both samples were grown at a substrate temperature of 670 oC. Both samples showed step flow growth and low surface roughness despite the large difference in In flux. Nonetheless, increasing the In flux from 3E-7 Torr to 4E-7 Torr resulted in an improved surface morphology, and increased In composition and strain relaxation. These results indicate that the In flux can be tuned within an ample window to obtain specific sample properties. For the last series, three samples were grown at 660 oC and with In flux of 4E-7 Torr, while the Ga flux was varied from 3.2E-7 Torr to 3.6E-7 Torr. The surface morphology degraded with decreasing Ga flux. The Indium composition simultaneously increased due to the limited availability of Ga during growth. This indicates that the strain caused by increased In incorporation is relaxed by generating dislocations which in turn results in further increase in In incorporation as relaxation proceeds (the In pulling effect).

EMC 2021

Growth of High Quality 350nm-Thick InGaN Films on N-Polar GaN Substrates by Plasma-Assisted Molecular Beam Epitaxy

pambe

ingan

n-polar

The Electronic Materials Conference (EMC) is the premier annual forum on the preparation, characterization and use of electronic materials. This year’s event will feature a plenary session, topical sessions, a poster session and more.

EMC is coordinated with the Device Research Conference (DRC), which will be held during the same week, from June 20-23. This recognizes the strong interaction between device and electronic materials research and provides fruitful exchange of information between attendees of both Conferences.

![](https://assets.underline.io/uploads/markdown_image/1/image/eb26c24a18d5d1030616afa4f993cf89.png)
	
<a href="https://www.mrs.org/docs/default-source/meetings-events/mrs-conference-services/2021/emc-2021/2021-emc-program-at-a-glance-6-16-21.pdf?sfvrsn=e39f950c_4">Latest Program-at-a-Glance</a>

<a href="https://drive.google.com/file/d/1eiS1ao4p83VNy8xgHs3Rls7gTQjx7Rc4/view?usp=sharing">Full Program with Abstracts</a>

Abstract Body: A parametric analysis of growth conditions on the structural properties of N-Polar InGaN films was studied. Three series of N-polar InGaN films were grown on bulk GaN substrates by plasma-assisted molecular beam epitaxy (PAMBE) to study the effect of substrate temperature, Ga, and In fluxes. For all series, the surface morphology was measured by atomic force microscopy (AFM), and the composition and strain were determined by high resolution X-ray diffraction (HRXRD) and XRD reciprocal space mapping (RSM). All the samples in this study were 350 nm thick. First, a series of samples was grown at temperatures varying from 620 oC to 690 oC. The Ga flux for all samples in this series was kept at 3.8E-7 Torr and the In flux was increased by temperature to ensure metal-rich growth conditions. The surface morphology improved significantly by increasing the temperature, and the indium content was reduced as expected. The samples grown at 670 oC and 690 oC showed surprisingly smooth surface morphologies with surface roughness <0.7 nm which has never been reported before on such thick InGaN films. An In content of 16.5% and 13.6% was measured on samples grown at 670 oC and 690 oC, respectively. These temperatures are significantly higher than growth temperatures (<600 oC) typically used for the growth of InGaN films on Ga-polar GaN. In the next step, samples were grown with In fluxes of 3E-7 Torr and 4E-7 Torr. The Ga flux was kept at 3.8E-7 Torr and both samples were grown at a substrate temperature of 670 oC. Both samples showed step flow growth and low surface roughness despite the large difference in In flux. Nonetheless, increasing the In flux from 3E-7 Torr to 4E-7 Torr resulted in an improved surface morphology, and increased In composition and strain relaxation. These results indicate that the In flux can be tuned within an ample window to obtain specific sample properties. For the last series, three samples were grown at 660 oC and with In flux of 4E-7 Torr, while the Ga flux was varied from 3.2E-7 Torr to 3.6E-7 Torr. The surface morphology degraded with decreasing Ga flux. The Indium composition simultaneously increased due to the limited availability of Ga during growth. This indicates that the strain caused by increased In incorporation is relaxed by generating dislocations which in turn results in further increase in In incorporation as relaxation proceeds (the In pulling effect).

technical paper

Abstract Body: The ultraviolet-C (UVC) light with wavelength from 180 nm to 280 nm has been recognized as a preferable avenue to disinfect virus because it can effectively destroy the chemical bonds inside the DNA or RNA of the microorganisms. III-nitride based UVC light-emitting diodes (LEDs) have been recognized as one of the ideal candidates to replace the toxic mercury-based UV-lamps for the application of air/water purification, disinfection/sterilization and biochemical environmental sensing. However, the optical performance of the UVC LEDs is still limited by the poor external quantum efficiency (EQE) [1, 2]. The size and geometry of the active emitting elements are extremely important considerations in optimizing LEDs. As the InGaN-based LED topic, it has been found that the micro-LED approach can offer the potential to improve overall LED efficiency and efficiency droop. However, a systematic investigation on size-dependent AlGaN-based UVC LED on device performance and detailed analysis of the underlying changing mechanisms have not been studied as much as InGaN based visible LEDs, which have longer history. In this work, the size-dependence of the optical and electrical properties of 265 nm wavelength AlGaN based UVC LED grown on c-plane sapphire is systematically investigated. The LED devices with diameters ranging from 50 μm to 300 μm have been well fabricated. It is shown that the series resistance of the UVC LED increased with the diameter of the device decreased. Moreover, smaller UVC LEDs can deliver higher power densities despite the relatively lower absolute output powers and sustain higher current densities. Our results show that at low injection current densities, the efficiency of smaller LEDs reduces significantly via the non-radiative recombination induced by sidewall defects. At higher injection current densities, the current crowding effect dominates and the efficiency of smaller LEDs becomes higher. Reference: [1] H. Yu, Z. Ren, H. Zhang, J. Dai, C. Chen, S. Long, and H. Sun, "Advantages of AlGaN-based deep-ultraviolet light-emitting diodes with an Al-composition graded quantum barrier," Optics express, 27 (20), A1544-A1553, 2019. [2] Z. Ren, H. Yu, Z. Liu, D. Wang, C. Xing, H. Zhang, C. Huang, S. Long, and H. Sun, "Band engineering of III-nitride-based deep-ultraviolet light-emitting diodes: A review," Journal of Physics D: Applied Physics, 53 (7), 073002, 2019.

Micro-Scale AlGaN-Based Deep Ultraviolet Light-Emitting Diodes Emitting at 265 nm

Abstract Body: Electronic structure calculations of 4H-SiC doped with various transition metals reveal dilute magnetic semiconductor (DMS) behavior in a material suitable for high-power and high-frequency semiconductor devices. Our results are consistent with prior work on V, Cr, and Mn doping and explore additional metals: Fe, Co, and Ni. Charge-state calculations show that the latter maintain amphoteric semi-insulating properties while offering a higher stable spin polarization and also greater asymmetry in the spin density of states than previously studied dopants. This indicates possible enhanced half-metal properties. We then use these discoveries to inform the first calculations of the phonon-assisted optical properties of doped 4H-SiC systems, using the approach first published by Noffsinger et al. [1] based on Maximally Localized Wannier Functions to assist overcoming the numerical challenge. [1]: Phonon-Assisted Optical Absorption in Silicon from First Principles Jesse Noffsinger, Emmanouil Kioupakis, Chris G. Van de Walle, Steven G. Louie, and Marvin L. Cohen Phys. Rev. Lett. 108, 167402 – Published 20 April 2012

Transition Metal Defects 4H-SiC—Optical and Electronic Levels from First Principles

Abstract Body: As a wide bandgap semiconductor, silicon carbide (SiC) possesses unique properties that make it an advantageous replacement for silicon in high power, high frequency, and high temperature electronic devices. Ion implantation followed by high temperature annealing (>1600oC) is often used to introduce and activate dopants in the material. However,the high concentration of point defects created from implantation evolve into extended defects upon annealing, resulting in device failure. In order to address this issue, elevated temperature implants have been used extensively to prevent amorphization but little is known about higher temperature implant effects on defect evolution and how this correlates with dopant activation efficiency. In this study, 60 keV Al+ was implanted into single crystal n-type 4H-SiC wafers to a fluence of either 5x1013 or 1x1015 cm-2. The implant temperature was varied between 300°C and 700°C. High temperature furnace annealing was performed with a carbon cap from 1300°C to 1680°C for 30 minutes. Structural characterization using cross-sectional and plan-view BF-TEM and DF-TEM was performed for extended defect analysis. Ohmic contacts were created from sputter deposition of a Ti/Al/Ni metal stack followed by alloying at 1000oC for 1 minute. Electrical characterization using variable-temperature Hall Effect was performed to study dopant activation. As implanted cross-sectional TEM analysis showed amorphization is avoided for the low fluence (5x1013 cm-2) implants at all implant temperatures. For the high fluence case (1x1015 cm-2), amorphization is avoided above 400°C due to the increasing effects of dynamic annealing. The reduction of lattice damage with increasing implant temperature can be seen in UV/Vis/NIR measurements for the unannealed samples. The absorption peaks corresponding to lattice damage show a consistent reduction as the implant temperature is increased up to 700°C. A carbon cap was introduced to reduce surface decomposition during the high temperature annealing process with a variety of capping methods studied. Graphitized photoresist was shown by AFM measurements to result in the least amount of surface roughening upon annealing at around 1 nm. Extended defects were observed by TEM for the high fluence implants at all implant temperatures after annealing but were absent for the low fluence case. The loops are seen to coarsen with implant temperature as expected but are not eliminated by the anneal. TEM results suggest the concentration of trapped interstitials in the extended defect decreases with increasing implant temperature. Further analysis on the effect of implant temperature on the evolution of these extended defects will be presented. In addition to structural characterization, electrical measurements were carried out. Ohmic contacts were successfully created from the combination of an effective annealing cap and from the optimal metal thickness ratios in the metal stack. However, initial Hall effect measurements at room temperature were not successful in determining the dopant activation values even after 1680°C annealing. This is thought to be due to the relatively deep acceptor level associated with Al in SiC. Subsequently, variable field measurements at 400°C showed a dopant type change from the n-type substrate to the p-type implant enabling dopant activation values to be determined. These electrical results will be compared with structural images to highlight correlations between extended defect evolution and dopant activation efficiency.

Investigation of Implant Temperature on Dopant Activation Efficiency for Single Fluence Al+ Implanted 4H-SiC

Abstract Body: Superior material qualities of 4H-SiC such as wide bandgap, high thermal conductivity, and large breakdown electric field make devices capable of operating with high power at high temperature. In this work, the behavior of p-channel 4H-SiC metal-oxide-semiconductor field-effect transistor (MOSFET), where holes are the channel charge carriers; has been investigated. The experimental results confirm that post oxidation Nitric Oxide (NO) annealing reduces SiO2/4H-SiC interface trap densities (Dit) in the upper half of the 4H-SiC bandgap as widely reported. In addition, Dit is also significantly reduced in the lower half, which is critical for p-channel MOSFETs. This result is corroborated by capacitance-voltage (CV) measurements conducted on p- and n-type 4H-SiC MOS capacitors from room temperature to 300°C, as well as under UV illumination, which allows estimation of the Dit profile throughout the large 4H-SiC bandgap. Thereafter, p-channel 4H-SiC MOSFETs were fabricated and measured from room temperature through 300°C. The MOSFETs without NO annealing turn on at very large negative voltages with high dielectric leakage currents, making it difficult to get reliable high temperature data. In contrast, the NO annealed MOSFETs turn on at a much lower voltage, with low leakage, consistent with the reduction of Dit due to NO annealing. With the rise of temperature, the threshold voltage drops as the number of occupied interface traps reduce with the increase of temperature. However, the temperature dependence of the hole channel mobility is not obvious from these measurements. The weak temperature dependence can be explained considering the behavior of the phonon and Coulomb scattering components. The phonon scattering limited hole mobility reduces from ~120 cm2/Vs at room temperature to less than 25 cm2/Vs at 300°C in lightly doped bulk p-type 4H-SiC. On the other hand, the Coulomb scattering limited surface hole mobility increases weakly with temperature. The competition between the two mechanisms increases at higher temperatures and the overall field-effect mobility shows an almost temperature-independent behavior in the 25°C - 300°C range. While NO annealing realizes high temperature operation, the negative bias temperature instability (NBTI) is a major shortcoming. This is a result of hole injection and trapping in the oxide under high temperature and negative bias stress. In this presentation, the results from the NBTI experiments will be reported and the main mechanisms related to high temperature NBTI will be discussed. The authors gratefully acknowledge the funding support from US Army Research Laboratory grant ARMY-W911NF-18-2-0160 and National Renewable Energy Laboratory/ US Department of Energy grant NREL-AHL-9-92362-01.

High Temperature Characteristics of NO Annealed P-Channel 4H-SiC MOSFET

Abstract Body: We demonstrate a technique to spatially map and quickly build the frequency response of optoelectronic devices using optical probes. The transfer function of a linear system is the Fourier transform of its impulse response, which may be obtained from transient photocurrent measurements of devices such as photodetectors and solar cells. We introduce and apply Fourier Transform Impedance Spectroscopy (FTIS) to a PbS colloidal quantum dot (QD)/SiC heterojunction photodiode and corroborate results using intensity-modulated photocurrent spectroscopy. The cutoff frequencies of the QD/SiC devices were as high as ~10kHz, demonstrating their utility in advanced flexible and thin film electronics. The practical frequencies for FTIS lie in the mHz-kHz range, ideal for composite or novel materials such as QD films that are dominated by interfacial trap states. These can lead to characteristic lengths for charge collection ~20-500 μm dominated by transmission line effects, rather than intrinsic diffusion and drift length scales, enabling extraction of interfacial capacitances and series/parallel resistances.

Spatially Resolved Fourier Transform Impedance Spectroscopy—A Technique to Rapidly Characterize Composite Interfaces and a Study of Quantum Dot/Epitaxial Graphene/SiC Optoelectronic Devices

Abstract Body: III-Nitride (AlInGaN) avalanche photodiodes (APDs) have merits such as high breakdown field, high electron drift velocity, high thermal conductivity, and low dark current. The III-nitride material bandgap is tunable from 0.7 (1772nm) to 6.2eV (200nm), offering good UV sensitivity in both visible blind and solar blind UV regions. The Geiger-mode operation of III-Nitride APDs have numerous applications in UV single-photon detection, such as quantum-key distribution, optical time-domain reflectometry, positron-emission tomography, biomedical research, and UV communications. In this work, we report low-temperature Geiger mode measurements of GaN p-i-n APDs by using a thermoelectric cooling system. The GaN p-i-n APD structures were grown on c-plane hydride-vapor-phase-epitaxy (HVPE) free-standing (FS) GaN or ammonothermal bulk GaN. The epitaxially grown structure consists of a 0.6 mm unintentionally doped GaN, followed by a 2.3mm thick Si-doped GaN n-layer with a free-electron concentration of 4×1018 cm-3. The i-layer of 280nm thickness is grown on top of the n-layer, with a background Si doping of 2×1016 cm-3, targeting breakdown voltage of 96V. The p-layer stack consists of two layers of 100nm Mg-doped GaN and 15nm of highly Mg-doped GaN as the contact layer. The Mg-doped p-layer has a free-hole concentration of 1×1018 cm-3, and the highly Mg-doped p-layer has Mg concentration of 1×1020 cm-3. The fabrication of APD was performed in the following sequence: 1) mesa etching, 2) n-type metal deposition, 3) p-type metal deposition, 4) passivation and via-hole, and 5) metal interconnect deposition. A gated quenching circuit was used in the Geiger-mode measurement. In detail, a noninverting amplifier with gain of 2 and a transimpedance amplifier are inserted before and after the GaN APD in order to isolate the GaN APD and prevent reflection. An oscilloscope measures and collects the avalanche signals. A computer is utilized to perform signal data processing. For spectral response evaluation, the APD device is illuminated with the UV light from a UV LED with the peak wavelength of 375nm. The UV light is split by a bifurcated fiber: one end is fed to GaN APD, and the other end goes to an attenuator before being fed to a silicon photon counter for reference. For the low-temperature control, the thermoelectric cooler is installed directly under the GaN APD. The thermoelectric cooler is installed in an aluminum cage which serves as water-cooled heat sink. To prevent icing, the aluminum cage is sealed and filled with nitrogen. The Geiger-mode setup is installed in a optically isolated dark box. The method of measurement employed is the Δt method based on Poisson statistics. From the collected avalanche signal, the computer determines time interval between the rising edge of each pulse and its corresponding avalanche signal for 10,000 pulse events, then processes these time interval data into one histogram. Finally, by fitting the exponential function to the histogram, the dark count rate R is obtained. With this low-temperature Geiger-mode measurement system, the breakdown voltage was measured at various temperature from -10°C to 10°C with an increment step of 5°C. The breakdown voltage is proportional to the temperature, monotonically varying from 96.9V to 97.25V when the temperature increases from -10°C to 10°C. From this data, we obtained a temperature coefficient of the breakdown voltage of 0.0177±0.0005V/K. Based on the breakdown voltage vs. temperature, the dark count rate of various temperatures at 1V overvoltage above breakdown was measured. The dark count rate decreases as the temperature decreases from 19MHz to 11MHz when temperature decreases from 10°C to -10°C. By fitting Arrhenius equation, the obtained activation energy is . More detailed description will be presented in the conference. This work is funded by Support from ARO, DoE, NASA, Georgia Tech IEN, and the NSF National Nanotechnology Coordinated Infrastructure (NNCI)

Low Temperature Geiger-Mode Measurement of Gallium Nitride p-i-n Avalanche Photodiode

Abstract Body: Energy-saving photodetectors have recently attracted intensive research interest to meet the great demands of future photonic systems. Particularly, according to the large photocurrent, high sensitivity, environment-sensitivity, and lithography-free features, self-powered photoelectrochemical-type photodetectors (PEC PDs) have attracted tremendous attention in the optoelectronics community. However, owing to their poor crystal quality and low carrier separation efficiency of chemically synthesized semiconductor materials, PEC PDs based on such materials demonstrate slow response and low responsivity. In our work, we construct PEC PD using defect-free AlGaN nanowires photoelectrode which synthesized through molecular beam epitaxy (MBE)[1][2]. After decorating AlGaN nanowires with noble metal species (Pt or Ru), our PEC PDs feature the combined benefits of effective carrier separation and fast interfacial kinetics, resulting in superior performance compared to previous reported solar-blind PEC PDs (Responsivity of 45 mA/W and response/recovery time of 47/20 ms for Pt decorated p-type AlGaN nanowires; Responsivity of 48.8 mA/W and response/recovery time of 83/19 ms for Ru decorated n-type AlGaN nanowires;). Besides, according to the attractive surrounding sensitivity of PEC PDs, their photoresponse behaviors can be effectively modulated by varying the applied bias potential, light intensity, and the concentration of the electrolyte further promotes its adaptability in different working conditions. Our work suggests that embracement of defect-free wide-bandgap semiconductor nanowires with appropriate surface decoration offers an unprecedented opportunity for designing future energy-efficient and large-scale optoelectronic systems on the silicon platform. Reference: [1] D. Wang, X. Liu, S. Fang, C. Huang, Y. Kang, H. Yu, Z. Liu, H. Zhang, R. Long, Y. Xiong, Y. Lin, Y. Yue, B. Ge, T. K. Ng, B. S. Ooi, Z.Mi, J. H. He, and H. Sun “Pt/AlGaN Nanoarchitecture: Toward High Responsivity, Self-Powered Ultraviolet-Sensitive Photodetection,” Nano Lett., 2021, 21, 1, 120–129. [2] D. Wang, C. Huang, X. Liu, H. Zhang, H. Yu, S. Fang, B. S. Ooi, Z. Mi, J. H. He, and H. Sun, “Highly Uniform, Self-Assembled AlGaN Nanowires for Self-Powered Solar-Blind Photodetector with Fast-Response Speed and High Responsivity,” Adv. Optical Mater., 2020:202000893.

AlGaN-Based Nanowires for Self-Powered Ultraviolet Photodetectors with High Responsivity

Abstract Body: Gallium nitride (GaN) with bandgap energy of 3.4 eV with a high breakdown electric field (~3.3 MV/cm) is useful for ultraviolet (UV) optical detectors. Since GaN avalanche photodiodes (APDs) have a high sensitivity and capability of operation in the visible-blind region with high optical gain with low dark current, UV-APD arrays are considerably well-suited for many applications in UV imaging systems, astronomy, military systems, and bio-aerosol detection to improve the collection efficiency and sensitivity for low-level light detection. Achieving the high crystalline quality of epitaxial layer structures with reduced-defect densities as well as optimizing the fabrication processes will significantly improve GaN APD arrays to achieve a high pixel yield and uniform breakdown voltage with low leakage current. Evaluation of these 6x6 arrays will help to confirm the uniformity of the epitaxial materials and device processing as well. In this work, visible-blind front-illuminated GaN p-i-n APDs were grown by metalorganic chemical vapor deposition (MOCVD) on 25 mm dia. bulk Ammono® n-GaN substrates having a low etch pit density (EPD) less than 5 × 104 [cm-2] and processed into 6x6 APD arrays. The MOCVD growth was performed in a close-coupled-showerhead AIXTRON 6x2” MOCVD system and employed trimethylgallium and NH3 as primary precursors and SiH4, and Cp2Mg as dopants. X-ray diffraction (XRD) was used for the characterization of the crystallographic quality of the APD structure. The full widths at half maximum of the rocking curve derived from (0.02) and (1.02) diffraction exhibit values of 73 and 41 arc-sec, respectively, primarily determined by the properties of the GaN substrate. Hall-effect characterization was used for the measurement of the electrical properties of the doped layers including the free-carrier concentration, mobility, and resistivity. The epitaxial layers grown on the GaN substrates include a 1.0 μm thick n-GaN:Si layer (n~7×1018 cm-3), 0.2 μm thick u-GaN undoped layer, 0.1 μm thick p-GaN:Mg layer (p~6x1017 cm-3) and 0.02 μm thick p++-GaN:Mg layer for a target breakdown voltage of 80 V. The surface morphology of the GaN p-i-n UV-APD epitaxial structures were studied by atomic force microscopy (AFM). The corresponding root-mean-square (RMS) values for scanning area of 5×5 μm2 for GaN/sapphire u-GaN template and n-GaN substrate are 0.15 nm and 0.07 nm respectively for the typical case. To confirm the doping profile of the layers, secondary ion mass spectrometry (SIMS) measurements were performed. The as-grown wafers were processed using ion-implantation isolation into 6x6 APD arrays having 60 µm-dia. pixels. The reverse-biased J-V characteristics of the GaN APD arrays under dark and 355 nm illumination conditions have been measured at 300K. The dark current density is 10-9 A/cm2 at reverse bias up to -20 V and breakdown voltage of 80 ± 1 V for all 36 devices which confirms a high uniformity of growth and fabrication processes. The maximum avalanche gain reached 3 × 105 at the breakdown. The details of growth, characterization, and fabrication processes as well as further device data will be discussed in the presentation. Work supported in part by US Army Research Office under W911NF-15-1-0489, administered by M. Gerhold, by the US Department of Energy under DESC0019133, and NASA under 80NSSC18C0093. This work was performed in part in the facilities of the Georgia Tech Institute for Electronics and Nanotechnology, a member of the National Nanotechnology Coordinated Infrastructure, which is supported by the National Science Foundation (Grant ECCS-154217). The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either express or implied, of DoE, ARO, or the U.S. Government.

Growth and Characterization of p-i-n GaN UV 6 × 6 Avalanche Photodiodes Arrays with Large Detection Area

Abstract Body: RHEED spectra taking during molecular beam epitaxy of GaN and related alloys on sapphire substrates were analyzed using a convolutional neural network. An existing image-classifying convolution neural network (CNN)— AlexNet— was employed for the analysis. The RHEED images during various stages of growth were manually categorized to train the CNN. The CNN only works on images, so video frames of RHEED during growth had to first be extracted from the video. These images were then separated into a training set and a validation set for use in training the network and in validating it. Mobility, deposition temperature, X-ray diffraction, atomic force microscopy, and Hall measurements were taken on the resulting films. After training on the manually categorized images and videos, the CNN achieved a proper categorization rate of 97% on the validation data and employing mobility and various growth stages as categorical outputs. Other machine learning techniques were applied as well to analyze RHEED spectra, such as principal components analysis and k-means clustering. Initial results from growth of ternary (AlxIn1-xN) and quaternary (AlGaInN) alloys will also be presented at the conference.

Machine Learning for Evaluation of RHEED Spectra on III-Nitride Films Grown Using Molecular Beam Epitaxy

Abstract Body: III-Nitride materials have long been promising candidates for the development of infrared optoelectronic devices. However, design of multiple quantum well (MQW) structures on conventional c-plane (0001)-oriented GaN substrates is hindered by the spontaneous and piezoelectric polarization fields associated with lack of inversion symmetry. Low-defect density free-standing m-plane (10-10) GaN substrates have enabled the growth of non-polar MQW structures without strong built-in polarization fields. We have previously demonstrated the potential of m-plane Al0.16Ga0.84N/GaN heterostructures grown by plasma-assisted molecular-beam epitaxy (PAMBE) for terahertz devices [1]. To support higher energy near-infrared optical transitions, though, a larger conduction band offset (CBO) is required. This can be achieved through higher Al composition in the barriers. However, m-plane AlGaN layers with aluminum fractions above 0.2 grown under conventional gallium-rich PAMBE conditions exhibit severe material inhomogeneities [2]. Recently, we demonstrated that growth of m-plane AlGaN by indium surfactant assisted epitaxy (ISAE) at low temperature significantly reduces composition inhomogeneities [3]. Here, we present an in-depth comparison of the PAMBE growth of m-plane AlGaN with and without indium surfactant at 565°C. We investigated the compositions and growth rates of low temperature m-plane AlGaN layers grown by conventional gallium-rich epitaxy and by ISAE across a range of aluminum fluxes. It is expected that under nitrogen-limited growth conditions the growth rate would be independent of metal flux. We find the opposite to be true: as aluminum flux increases, the growth rate of AlGaN decreases rapidly. We attribute this phenomenon to nitrogen loss during PAMBE of m-plane AlGaN that increases super-linearly with increasing aluminum flux. These films also show a super-linear increase in aluminum content with increasing aluminum flux. Transmission Electron Microscopy (TEM) images indicate that these effects are correlated with high aluminum containing defects within the material. Our results indicate that ISAE mitigates this effect almost entirely. Films grown by ISAE with identical aluminum fluxes to the gallium-rich layers do not exhibit a significant decrease in growth rate with increasing aluminum flux, and their aluminum compositions increase linearly with aluminum flux. TEM images show the ISAE layers are thicker, and exhibit a higher degree of material uniformity up to 0.30 aluminum metal fraction. X-ray diffraction (XRD) measurements suggest minimal indium incorporation in the ISAE AlGaN layers. The CBO of MQWs can be further increased by using InGaN as the quantum well (QW) material. We have recently demonstrated the growth of high-quality m-plane InGaN with up to 0.16 indium metal fraction [4], enabling the growth of In0.055Al0.19Ga0.755N/In0.16Ga0.84N MQW structures using the ISAE technique [3]. We will also present results of structural and optical characterization of m-plane (In)Al0.3Ga0.7N/In0.16Ga0.84N MQW structures. Due to the anisotropic nature of strain in m-plane structures, AlGaN/InGaN structures cannot be perfectly strain-balanced. We will discuss the method used to control strain accumulation along the (0001) and (11-20) directions below the level for defect formation in both the InGaN QWs and AlGaN barriers. [1] C. Edmunds et al., Appl. Phys. Lett. 105, 021109 (2014). [2] M. Shirazi-HD et al., J. Appl. Phys. 123, 161581 (2018). [3] B. Dzuba et al., J. Appl. Phys. 128, 115701 (2020). [4] A. Senichev et al., APL Mater. 7, 121109 (2019).

Growth of High Quality 350nm-Thick InGaN Films on N-Polar GaN Substrates by Plasma-Assisted Molecular Beam Epitaxy

Downloads

Next from EMC 2021

Micro-Scale AlGaN-Based Deep Ultraviolet Light-Emitting Diodes Emitting at 265 nm

Similar lecture

Anomalous Super-Linear Growth Rate Reduction with Increasing Aluminum Flux in M-Plane AlGaN Grown by PAMBE

Growth of High Quality 350nm-Thick InGaN Films on N-Polar GaN Substrates by Plasma-Assisted Molecular Beam Epitaxy

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from EMC 2021

Micro-Scale AlGaN-Based Deep Ultraviolet Light-Emitting Diodes Emitting at 265 nm

Similar lecture

Anomalous Super-Linear Growth Rate Reduction with Increasing Aluminum Flux in M-Plane AlGaN Grown by PAMBE

Downloads