Publications

Abstract not provided.

The landscape of power transmission and distribution is quickly evolving as more power conversion is done through power electronics and transmitted or distributed at medium voltage direct current (MVDC). As power electronics tend to store less energy and be less resilient to faults than conventional power transformers, a reliable and fast protection against faults is critical to protect power electronics (PE) based infrastructure, especially for medium- and high-voltage applications. This motivates the development of PE based protection circuits to replace slower contemporary electromechanical breakers. In this project a novel PE based MVDC circuit breaker is developed; the new design is comprised of 1) a normally-on leg made of commercially-available, cascaded SiC junction field effect transistors (JFETs) with a passive balancing network, and 2) a normally-off leg based on an optically-triggered gallium nitride (GaN) photoconductive semiconductor switch (PCSS). The normally-off leg was designed to be quickly turned on, to divert current to an auxiliary dissipative circuit, as the normally-on leg is turned off. This approach, using solid-state devices, was selected for a high-performance, fast-switching operation for the DC circuit breaker as compared to approaches that use slower mechanical switches. To be practical, the circuit breaker must have low conduction loss (low Ron) in the normally-on leg and fast coordinated triggering of the normally-off leg to avoid damage from inductive flyback, which could be considerable for long lengths of cable.

Abstract not provided.

Abstract not provided.

Abstract not provided.

K-means clustering analysis is applied to frequency-domain thermoreflectance (FDTR) hyperspectral image data to rapidly screen the spatial distribution of thermophysical properties at material interfaces. Performing FDTR while raster scanning a sample consisting of 8.6 μ m of doped-silicon (Si) bonded to a doped-Si substrate identifies spatial variation in the subsurface bond quality. Routine thermal analysis at select pixels quantifies this variation in bond quality and allows assignment of bonded, partially bonded, and unbonded regions. Performing this same routine thermal analysis across the entire map, however, becomes too computationally demanding for rapid screening of bond quality. To address this, K-means clustering was used to reduce the dimensionality of the dataset from more than 20 000 pixel spectra to just K = 3 component spectra. The three component spectra were then used to express every pixel in the image through a least-squares minimized linear combination providing continuous interpolation between the components across spatially varying features, e.g., bonded to unbonded transition regions. Fitting the component spectra to the thermal model, thermal properties for each K cluster are extracted and then distributed according to the weighting established by the regressed linear combination. Thermophysical property maps are then constructed and capture significant variation in bond quality over 25 μ m length scales. The use of K-means clustering to achieve these thermal property maps results in a 74-fold speed improvement over explicit fitting of every pixel.

Abstract not provided.

Abstract not provided.

3D integration of multiple microelectronic devices improves size, weight, and power while increasing the number of interconnections between components. One integration method involves the use of metal bump bonds to connect devices and components on a common interposer platform. Significant variations in the coefficient of thermal expansion in such systems lead to stresses that can cause thermomechanical and electrical failures. More advanced characterization and failure analysis techniques are necessary to assess the bond quality between components. Frequency domain thermoreflectance (FDTR) is a nondestructive, noncontact testing method used to determine thermal properties in a sample by fitting the phase lag between an applied heat flux and the surface temperature response. The typical use of FDTR data involves fitting for thermal properties in geometries with a high degree of symmetry. In this work, finite element method simulations are performed using high performance computing codes to facilitate the modeling of samples with arbitrary geometric complexity. A gradient-based optimization technique is also presented to determine unknown thermal properties in a discretized domain. Using experimental FDTR data from a GaN-diamond sample, thermal conductivity is then determined in an unknown layer to provide a spatial map of bond quality at various points in the sample.

Multiple scattering is a common phenomenon in acoustic media that arises from the interaction of the acoustic field with a network of scatterers. This mechanism is dominant in problems such as the design and simulation of acoustic metamaterial structures often used to achieve acoustic control for sound isolation, and remote sensing. In this study, we present a physics-informed neural network (PINN) capable of simulating the propagation of acoustic waves in an infinite domain in the presence of multiple rigid scatterers. This approach integrates a deep neural network architecture with the mathematical description of the physical problem in order to obtain predictions of the acoustic field that are consistent with both governing equations and boundary conditions. The predictions from the PINN are compared with those from a commercial finite element software model in order to assess the performance of the method.

This study presents a deep learning based methodology for both remote sensing and design of acoustic scatterers. The ability to determine the shape of a scatterer, either in the context of material design or sensing, plays a critical role in many practical engineering problems. This class of inverse problems is extremely challenging due to their high-dimensional, nonlinear, and ill-posed nature. To overcome these technical hurdles, we introduce a geometric regularization approach for deep neural networks (DNN) based on non-uniform rational B-splines (NURBS) and capable of predicting complex 2D scatterer geometries in a parsimonious dimensional representation. Then, this geometric regularization is combined with physics-embedded learning and integrated within a robust convolutional autoencoder (CAE) architecture to accurately predict the shape of 2D scatterers in the context of identification and inverse design problems. An extensive numerical study is presented in order to showcase the remarkable ability of this approach to handle complex scatterer geometries while generating physically-consistent acoustic fields. The study also assesses and contrasts the role played by the (weakly) embedded physics in the convergence of the DNN predictions to a physically consistent inverse design.

Abstract not provided.

Abstract not provided.

Radiation-hard high-voltage vertical GaN p-n diodes are being developed for use in power electronics subjected to ionizing radiation. We present a comparison of the measured and simulated photocurrent response of diodes exposed to ionizing irradiation with 70 keV and 20 MeV electrons at dose rates in the range of 1.4× 107 - 5.0× 108 rad(GaN)/s. The simulations correctly predict the trend in the measured steady-state photocurrent and agree with the experimental results within a factor of 2. Furthermore, simulations of the transient photocurrent response to dose rates with uniform and non-uniform ionization depth profiles uncover the physical processes involved that cannot be otherwise experimentally observed due to orders of magnitude larger RC time constant of the test circuit. The simulations were performed using an eXploratory Physics Development code developed at Sandia National Laboratories. The code offers the capability to include defect physics under more general conditions, not included in commercially available software packages, extending the applicability of the simulations to different types of radiation environments.

Abstract not provided.

Abstract not provided.

Abstract not provided.

In this study, we develop an end-to-end deep learning-based inverse design approach to determine the scatterer shape necessary to achieve a target acoustic field. This approach integrates non-uniform rational B-spline (NURBS) into a convolutional autoencoder (CAE) architecture while concurrently leveraging (in a weak sense) the governing physics of the acoustic problem. By utilizing prior physical knowledge and NURBS parameterization to regularize the ill-posed inverse problem, this method does not require enforcing any geometric constraint on the inverse design space, hence allowing the determination of scatterers with potentially any arbitrary shape (within the set allowed by NURBS). A numerical study is presented to showcase the ability of this approach to identify physically-consistent scatterer shapes capable of producing user-defined acoustic fields.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

A method is developed to calculate the length into a sample to which a Frequency Domain Thermoreflectance (FDTR) measurement is sensitive. Sensing depth and sensing radius are defined as limiting cases for the spherically spreading FDTR measurement. A finite element model for FDTR measurements is developed in COMSOL multiphysics and used to calculate sensing depth and sensing radius for silicon and silicon dioxide samples for a variety of frequencies and laser spot sizes. The model is compared to experimental FDTR measurements. Design recommendations for sample thickness are made for experiments where semi-infinite sample depth is desirable. For measurements using a metal transducer layer, the recommended sample thickness is three thermal penetration depths, as calculated from the lowest measurement frequency.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Advanced GaN power devices are promising for many applications in high power electronics but performance limitations due to material quality in etched-and-regrown junctions prevent their widespread use. Carrier diffusion length is a critical parameter that not only determines device performance but is also a diagnostic of material quality. Here we present the use of electron-beam induced current to measure carrier diffusion lengths in continuously grown and etched-and-regrown GaN pin diodes as models for interfaces in more complex devices. Variations in the quality of the etched-and-regrown junctions are observed and shown to be due to the degradation of the n-type material. We observe an etched-and-regrown junction with properties comparable to a continuously grown junction.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Researchers at Sandia have developed a semiconductor-based high-voltage switch, with experimental results showing potential for enhanced radiation hardness, for use in multiple power conversion applications. Gallium nitride (GaN) metal-oxide semiconductor field effect transistors (MOSFETs) were modeled using commercial and Sandia CHARON simulation software to understand their performance and for future prediction of device operation in radiation environments.

This work provides the first demonstration of vertical GaN Junction Barrier Schottky (JBS) rectifiers fabricated by etch and regrowth of p-GaN. A reverse blocking voltage near 1500 V was achieved at 1 mA reverse leakage, with a sub 1 V turn-on and a specific on-resistance of 10 mΩ-cm2. This result is compared to other reported JBS devices in the literature and our device demonstrates the lowest leakage slope at high reverse bias. A large initial leakage current is present near zero-bias which is attributed to a combination of inadequate etch-damage removal and passivation induced leakage current.

Etched-and-regrown GaN pn-diodes capable of high breakdown voltage (1610 V), low reverse current leakage (1 nA = 6 μ A /cm2 at 1250 V), excellent forward characteristics (ideality factor 1.6), and low specific on-resistance (1.1 m Ω.cm2) were realized by mitigating plasma etch-related defects at the regrown interface. Epitaxial n -GaN layers grown by metal-organic chemical vapor deposition on free-standing GaN substrates were etched using inductively coupled plasma etching (ICP), and we demonstrate that a slow reactive ion etch (RIE) prior to p -GaN regrowth dramatically increases diode electrical performance compared to wet chemical surface treatments. Etched-and-regrown diodes without a junction termination extension (JTE) were characterized to compare diode performance using the post-ICP RIE method with prior studies of other post-ICP treatments. Then, etched-and-regrown diodes using the post-ICP RIE etch steps prior to regrowth were fabricated with a multi-step JTE to demonstrate kV-class operation.

Abstract not provided.

Abstract not provided.

Steady-state photocapacitance (SSPC) was conducted on nonpolar m-plane GaN n-type Schottky diodes to evaluate the defects induced by inductively coupled plasma (ICP) dry etching in etched-and-regrown unipolar structures. An ∼10× increase in the near-midgap Ec - 1.9 eV level compared to an as-grown material was observed. Defect levels associated with regrowth without an etch were also investigated. The defects in the regrown structure (without an etch) are highly spatially localized to the regrowth interface. Subsequently, by depth profiling an etched-and-regrown sample, we show that the intensities of the defect-related SSPC features associated with dry etching depend strongly on the depth away from the regrowth interface, which is also reported previously [Nedy et al., Semicond. Sci. Technol. 30, 085019 (2015); Fang et al., Jpn. J. Appl. Phys. 42, 4207-4212 (2003); and Cao et al., IEEE Trans. Electron Devices 47, 1320-1324 (2000)]. A photoelectrochemical etching (PEC) method and a wet AZ400K treatment are also introduced to reduce the etch-induced deep levels. A significant reduction in the density of deep levels is observed in the sample that was treated with PEC etching after dry etching and prior to regrowth. An ∼2× reduction in the density of Ec - 1.9 eV level compared to a reference etched-and-regrown structure was observed upon the application of PEC etching treatment prior to the regrowth. The PEC etching method is promising for reducing defects in selective-area doping for vertical power switching structures with complex geometries [Meyers et al., J. Electron. Mater. 49, 3481-3489 (2020)].

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Proper edge termination is required to reach large blocking voltages in vertical power devices. Limitations in selective area p-type doping in GaN restrict the types of structures that can be used for this purpose. A junction termination extension (JTE) can be employed to reduce field crowding at the junction periphery where the charge in the JTE is designed to sink the critical electric field lines at breakdown. One practical way to fabricate this structure in GaN is by a step-etched single-zone or multi-zone JTE where the etch depths and doping levels are used to control the charge in the JTE. The multi-zone JTE is beneficial for increasing the process window and allowing for more variability in parameter changes while still maintaining a designed percentage of the ideal breakdown voltage. Impact ionization parameters reported in literature for GaN are compared in a simulation study to ascertain the dependence on breakdown performance. Two 3-zone JTE designs utilizing different impact ionization coefficients are compared. Simulations confirm that the choice of impact ionization parameters affects both the predicted breakdown of the device as well as the fabrication process variation tolerance for a multi-zone JTE. Regardless of the impact ionization coefficients utilized, a step-etched JTE has the potential to provide an efficient, controllable edge termination design.

Abstract not provided.

Abstract not provided.

Abstract not provided.

GaN p-n diodes were formed by selective area regrowth on freestanding GaN substrates using a dry etch, followed by post-etch surface treatment to reduce etch-induced defects, and subsequent regrowth into wells. Etched-and-regrown diodes with a 150 μm diameter achieved 840 V operation at 0.5 A/cm2 reverse current leakage and a specific on-resistance of 1.2 mΩ·cm2. Etched-and-regrown diodes were compared with planar, regrown diodes without etching on the same wafer. Both types of diodes exhibited similar forward and reverse electrical characteristics, which indicate that etch-induced defectivity of the junction was sufficiently mitigated so as not to be the primary cause for leakage. An area dependence for forward and reverse leakage current density was observed, suggesting that the mesa sidewall provided a leakage path.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Impacts of silicon, carbon, and oxygen interfacial impurities on the performance of high-voltage vertical GaN-based p–n diodes are investigated. The results indicate that moderate levels (≈5 × 10¹⁷ cm^-3) of all interfacial impurities lead to reverse blocking voltages (V_b) greater than 200 V at 1 μA cm^-2 and forward leakage of less than 1 µA cm^-2 at 1.7 V. At higher interfacial impurity levels, the performance of the diodes becomes compromised. Herein, it is concluded that each impurity has a different effect on the device performance. For example, a high carbon spike at the junction correlates with high off-state leakage current in forward bias (≈100× higher forward leakage current compared with a reference diode), whereas the reverse bias behavior is not severely affected (> 200 V at 1 μA cm^-2). High silicon and oxygen spikes at the junction strongly affect the reverse leakage currents (≈ 1–10 V at 1 μA cm^-2). Regrown diodes with impurity (silicon, oxygen, and carbon) levels below 5 × 10¹⁷ cm^-3 show comparable forward and reverse results with the reference continuously grown diodes. The effect of the regrowth interface position relative to the metallurgical junction on the diode performance is also discussed.

Abstract not provided.

Abstract not provided.

Abstract not provided.

The impact of dry-etch-induced defects on the electrical performance of regrown, c-plane, GaN p-n diodes where the p-GaN layer is formed by epitaxial regrowth using metal-organic, chemical-vapor deposition was investigated. Diode leakage increased significantly for etched-and-regrown diodes compared to continuously grown diodes, suggesting a defect-mediated leakage mechanism. Deep level optical spectroscopy (DLOS) techniques were used to identify energy levels and densities of defect states to understand etch-induced damage in regrown devices. DLOS results showed the creation of an emergent, mid-gap defect state at 1.90 eV below the conduction band edge for etched-and-regrown diodes. Reduction in both the reverse leakage and the concentration of the 1.90 eV mid-gap state was achieved using a wet chemical treatment on the etched surface before regrowth, suggesting that the 1.90 eV deep level contributes to increased leakage and premature breakdown but can be mitigated with proper post-etch treatments to achieve >600 V reverse breakdown operation.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

This project is part of a multi-lab consortium that leverages U.S. research expertise and facilities at national labs and universities to significantly advance electric drive power density and reliability, while simultaneously reducing cost. The final objective of the consortium is to develop a 100 kW traction drive system that achieves 33 kW/L, has an operational life of 300,000 miles, and a cost of less than $\$6$/kW. One element of the system is a 100 kW inverter with a power density of 100 kW/L and a cost of $\$2.7$/kW. New materials such as widebandgap semiconductors, soft magnetic materials, and ceramic dielectrics, integrated using multi-objective cooptimization design techniques, will be utilized to achieve these program goals. This project focuses on a subset of the power electronics work within the consortium, specifically the design, fabrication, and evaluation of vertical GaN power devices suitable for automotive applications.

GaN is an attractive material for high-power electronics due to its wide bandgap and large breakdown field. Verticalgeometry devices are of interest due to their high blocking voltage and small form factor. One challenge for realizing complex vertical devices is the regrowth of low-leakage-current p-n junctions within selectively defined regions of the wafer. Presently, regrown p-n junctions exhibit higher leakage current than continuously grown p-n junctions, possibly due to impurity incorporation at the regrowth interfaces, which consist of c-plane and non-basal planes. Here, we study the interfacial impurity incorporation induced by various growth interruptions and regrowth conditions on m-plane p-n junctions on free-standing GaN substrates. The following interruption types were investigated: (1) sample in the main MOCVD chamber for 10 min, (2) sample in the MOCVD load lock for 10 min, (3) sample outside the MOCVD for 10 min, and (4) sample outside the MOCVD for one week. Regrowth after the interruptions was performed on two different samples under n-GaN and p-GaN growth conditions, respectively. Secondary ion mass spectrometry (SIMS) analysis indicated interfacial silicon spikes with concentrations ranging from 5e16 cm^-3 to 2e18 cm^-3 for the n-GaN growth conditions and 2e16 cm^-3 to 5e18 cm-3 for the p-GaN growth conditions. Oxygen spikes with concentrations ~1e17 cm^-3 were observed at the regrowth interfaces. Carbon impurity levels did not spike at the regrowth interfaces under either set of growth conditions. We have correlated the effects of these interfacial impurities with the reverse leakage current and breakdown voltage of regrown m-plane p-n junctions.

Abstract not provided.

Photodetectors sensitive to the ultra-violet spectrum were demonstrated using an AlGaN high electron mobility transistor with an GaN nanodot optical floating gate. Peak responsivity of 2 x 10⁹ A/W was achieved with a gain-bandwidth product > 1 GHz at a cut-on energy of 4.10 eV. Similar devices exhibited visible-blind rejection ratios > 10⁶. The photodetection mechanism for $β$-Ga₂O₃ was also investigated. It was concluded that Schottky barrier lowering by self-trapped holes enables photodetector gain.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

GaN is an attractive material for high-power electronics due to its wide bandgap and large breakdown field. Verticalgeometry devices are of interest due to their high blocking voltage and small form factor. One challenge for realizing complex vertical devices is the regrowth of low-leakage-current p-n junctions within selectively defined regions of the wafer. Presently, regrown p-n junctions exhibit higher leakage current than continuously grown p-n junctions, possibly due to impurity incorporation at the regrowth interfaces, which consist of c-plane and non-basal planes. Here, we study the interfacial impurity incorporation induced by various growth interruptions and regrowth conditions on m-plane p-n junctions on free-standing GaN substrates. The following interruption types were investigated: (1) sample in the main MOCVD chamber for 10 min, (2) sample in the MOCVD load lock for 10 min, (3) sample outside the MOCVD for 10 min, and (4) sample outside the MOCVD for one week. Regrowth after the interruptions was performed on two different samples under n-GaN and p-GaN growth conditions, respectively. Secondary ion mass spectrometry (SIMS) analysis indicated interfacial silicon spikes with concentrations ranging from 5e16 cm-3 to 2e18 cm-3 for the n-GaN growth conditions and 2e16 cm-3 to 5e18 cm-3 for the p-GaN growth conditions. Oxygen spikes with concentrations ∼1e17 cm-3 were observed at the regrowth interfaces. Carbon impurity levels did not spike at the regrowth interfaces under either set of growth conditions. We have correlated the effects of these interfacial impurities with the reverse leakage current and breakdown voltage of regrown m-plane p-n junctions.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.