Search

In recent years, the increasing complexity in scientific simulations and emerging demands for training heavy artificial intelligence models require massive and fast data accesses, which urges high-performance computing (HPC) platforms to equip with more advanced storage infrastructures such as solid-state disks (SSDs). While SSDs offer high-performance I/O, the reliability challenges faced by the HPC applications under the SSD-related failures remains unclear, in particular for failures resulting in data corruptions. The goal of this paper is to understand the impact of SSD-related faults on the behaviors of complex HPC applications. To this end, we propose FFIS, a FUSE-based fault injection framework that systematically introduces storage faults into the application layer to model the errors originated from SSDs. FFIS is able to plant different I/O related faults into the data returned from underlying file systems, which enables the investigation on the error resilience characteristics of the scientific file format. We demonstrate the use of FFIS with three representative real HPC applications, showing how each application reacts to the data corruptions, and provide insights on the error resilience of the widely adopted HDF5 file format for the HPC applications., Comment: 12 pages, 9 figures, 4 tables, accepted by IEEE Cluster'21

Heterogeneous integration strategies are increasingly being employed to achieve more compact and capable electronics systems for multiple applications including space, electric vehicles, and wearable and medical devices. To enable new integration strategies, the growth and transfer of thin electronic films and devices, including III-nitrides, metal oxides, and two-dimensional (2D) materials, using 2D boron nitride (BN)-on-sapphire templates is demonstrated. The van der Waals BN layer, in this case, acts as a preferred mechanical release layer for precise separation at the substrate-film interface and leaves a smooth surface suitable for van der Waals bonding. A tensilely-stressed Ni layer sputtered on top of the film induces controlled spalling fracture which propagates at the BN/sapphire interface. By incorporating controlled spalling, the process yield and sensitivity is greatly improved, owed to the greater fracture energy provided by the stressed metal layer relative to a soft tape or rubber stamp. With stress playing a critical role in this process, the influence of residual stress on detrimental cracking and bowing is investigated. Additionally, a selected area lift-off technique is developed which allows for isolation and transfer of individual devices while maximizing wafer area use and minimizing extra alignment steps in the integration process., Comment: 11 pages, 4 figures

This paper presents vertical (001) oriented $\beta$-Ga$_2$O$_3$ field plated (FP) Schottky barrier diode (SBD) with a novel extreme permittivity dielectric field oxide. A thin drift layer of 1.7 $\mu m$ was used to enable a punch-through (PT) field profile and very low differential specific on-resistance (R$_{on-sp}$) of 0.32 m$\Omega$-cm$^{2}$. The extreme permittivity field plate oxide facilitated the lateral spread of the electric field profile beyond the field plate edge and enabled a breakdown voltage ($V_{br}$) of 687 V. The edge termination efficiency increases from 13.5 $\%$ for non-field plated structure to 63 $\%$ for high permittivity field plate structure. The surface breakdown electric field was extracted to be 5.45 MV/cm at the center of the anode region using TCAD simulations. The high permittivity field plated SBD demonstrated a record high Baliga figure of merit (BFOM) of 1.47 GW/cm$^{2}$ showing the potential of Ga$_2$O$_3$ power devices for multi-kilovolt class applications.

The mapper construction is a powerful tool from topological data analysis that is designed for the analysis and visualization of multivariate data. In this paper, we investigate a method for stitching a pair of univariate mappers together into a bivariate mapper, and study topological notions of information gains, referred to as topological gains, during such a process. We further provide implementations that visualize such topological gains for mapper graphs.

This paper proposes a multivariable extremum seeking scheme using Fast Fourier Transform (FFT) for a network of subsystems working towards optimizing the sum of their local objectives, where the overall objective is the only available measurement. Here, the different inputs are perturbed with different dither frequencies, and the power spectrum of the overall output signal obtained using FFT is used to estimate the steady-state cost gradient w.r.t. each input. The inputs for the subsystems are then updated using integral control in order to drive the respective gradients to zero. This paper provides analytical rules for designing the FFT-based gradient estimation algorithm and analyzes the stability properties of the resulting extremum seeking scheme for the static map setting. The effectiveness of the proposed FFT-based multivariable extremum seeking scheme is demonstrated using two examples, namely, wind farm power optimization problem, and a heat exchanger network for industrial waste-to-heat recovery.

In this work $\textit{n}$-transfer and incomplete fusion cross sections for $^{9}$Be + $^{197}$Au system are reported over a wide energy range, E$_{c.m.}$ $\approx$ 29-45 MeV. The experiment was carried out using activation technique and off-line gamma counting. The transfer process is found to be the dominant mode as compared to all other reaction channels. Detailed coupled reaction channel (CRC) calculations have been performed for $\textit{n}$-transfer stripping and pickup cross sections. The measured 1$\textit{n}$-stripping cross sections are explained with CRC calculations by including the ground state and the 2$^{+}$ resonance state (E = 3.03 MeV) of $^{8}$Be. The calculations for 1$\textit{n}$-pickup, including only the ground state of $^{10}$Be agree reasonably well with the measured cross sections, while it overpredicts the data at subbarrier energies. For a better insight into the role of projectile structure in the transfer process, a comprehensive analysis of 1$\textit{n}$-stripping reaction has been carried out for various weakly bound projectiles on $^{197}$Au target nucleus. The transfer cross sections scaled with the square of total radius of interacting nuclei show the expected Q-value dependence of 1$\textit{n}$-stripping channel for weakly bound stable projectiles.

This paper tackles the problem of the heavy dependence of clean speech data required by deep learning based audio-denoising methods by showing that it is possible to train deep speech denoising networks using only noisy speech samples. Conventional wisdom dictates that in order to achieve good speech denoising performance, there is a requirement for a large quantity of both noisy speech samples and perfectly clean speech samples, resulting in a need for expensive audio recording equipment and extremely controlled soundproof recording studios. These requirements pose significant challenges in data collection, especially in economically disadvantaged regions and for low resource languages. This work shows that speech denoising deep neural networks can be successfully trained utilizing only noisy training audio. Furthermore it is revealed that such training regimes achieve superior denoising performance over conventional training regimes utilizing clean training audio targets, in cases involving complex noise distributions and low Signal-to-Noise ratios (high noise environments). This is demonstrated through experiments studying the efficacy of our proposed approach over both real-world noises and synthetic noises using the 20 layered Deep Complex U-Net architecture., Comment: Published in Interspeech 2021 ( See https://www.isca-speech.org/archive/interspeech_2021/kashyap21_interspeech.html ). 5 pages, 2 figures, 1 table

High-quality dielectric-semiconductor interfaces are critical for reliable high-performance transistors. We report the in-situ metalorganic chemical vapor deposition (MOCVD) of Al$_2$O$_3$ on $\beta$-Ga$_2$O$_3$ as a potentially better alternative to the most commonly used atomic layer deposition (ALD). The growth of Al$_2$O$_3$ is performed in the same reactor as Ga$_2$O$_3$ using trimethylaluminum and O$_2$ as precursors without breaking the vacuum at a growth temperature of 600 $^0$C. The fast and slow near interface traps at the Al$_2$O$_3$/ $\beta$-Ga$_2$O$_3$ interface are identified and quantified using stressed capacitance-voltage (CV) measurements on metal oxide semiconductor capacitor (MOSCAP) structures. The density of shallow and deep level initially filled traps (D$_{it}$) are measured using ultra-violet (UV) assisted CV technique. The average D$_{it}$ for the MOSCAP is determined to be 7.8 $\times$ 10$^{11}$ cm$^{-2}$eV$^{-1}$. The conduction band offset of the Al$_2$O$_3$/ Ga$_2$O$_3$ interface is also determined from CV measurements and found out to be 1.7 eV which is in close agreement with the existing literature reports of ALD Al$_2$O$_3$/ Ga$_2$O$_3$ interface. The current-voltage characteristics are also analyzed and the average breakdown field is extracted to be approximately 5.8 MV/cm. This in-situ Al$_2$O$_3$ dielectric on $\beta$-Ga$_2$O$_3$ with improved dielectric properties can enable Ga$_2$O$_3$-based high performance devices.

Apprentice is a tool developed for event generator tuning. It contains a range of conceptual improvements and extensions over the tuning tool Professor. Its core functionality remains the construction of a multivariate analytic surrogate model to computationally expensive Monte-Carlo event generator predictions. The surrogate model is used for numerical optimization in chi-square minimization and likelihood evaluation. Apprentice also introduces algorithms to automate the selection of observable weights to minimize the effect of mis-modeling in the event generators. We illustrate our improvements for the task of MC-generator tuning and limit setting., Comment: 9 pages, 2 figures, submitted to the 25th International Conference on Computing in High-Energy and Nuclear Physics

The parameters in Monte Carlo (MC) event generators are tuned on experimental measurements by evaluating the goodness of fit between the data and the MC predictions. The relative importance of each measurement is adjusted manually in an often time-consuming, iterative process to meet different experimental needs. In this work, we introduce several optimization formulations and algorithms with new decision criteria for streamlining and automating this process. These algorithms are designed for two formulations: bilevel optimization and robust optimization. Both formulations are applied to the datasets used in the ATLAS A14 tune and to the dedicated hadronization datasets generated by the sherpa generator, respectively. The corresponding tuned generator parameters are compared using three metrics. We compare the quality of our automatic tunes to the published ATLAS A14 tune. Moreover, we analyze the impact of a pre-processing step that excludes data that cannot be described by the physics models used in the MC event generators., Comment: 87 pages, Submission to SciPost

We report on the growth and characterization of Ge-doped \b{eta}-Ga2O3 thin films using a solid germanium source. \b{eta}-Ga2O3 thin films were grown using a low-pressure chemical vapor deposition (LPCVD) reactor with either an oxygen or gallium delivery tube. Films were grown on 6 degree offcut sapphire and (010) \b{eta}-Ga2O3 substrates with growth rates between 0.5 - 22 {\mu}m/hr. By controlling the germanium vapor pressure, a wide range of Hall carrier concentrations between 10^17 - 10^19 cm-3 were achieved. Low-temperature Hall data revealed a difference in donor incorporation depending on the reactor configuration. At low growth rates, germanium occupied a single donor energy level between 8 - 10 meV. At higher growth rates, germanium doping predominantly results in a deeper donor energy level at 85 meV. This work shows the effect of reactor design and growth regime on the kinetics of impurity incorporation. Studying donor incorporation in \b{eta}-Ga2O3 is important for the design of high-power electronic devices., Comment: 16 pages,4 figures

We report on the demonstration of metalorganic vapor phase epitaxy-regrown (MOVPE) ohmic contacts in an all MOVPE-grown $\beta$-Ga$_2$O$_3$ metal-semiconductor field effect transistor (MESFET). The low-temperature (600$^{\circ}$C) heavy (n$^{+}$) Si-doped regrown layers exhibit extremely high conductivity with sheet resistance of 73 $\Omega$/$\square$ and record low metal/n$^{+}$-Ga$_2$O$_3$ contact resistance of 80 m$\Omega$.mm and specific contact resistivity of 8.3$\times$10$^{-7}$ $\Omega$.cm$^{2}$ were achieved. The fabricated MESFETs exhibit a maximum drain-to-source current of 130 mA/mm, a high I$_{ON}$/I$_{OFF}$ of $>$10$^{10}$ with a high power FOM of 25 MW/cm$^{2}$ were achieved without any field plates. Nanoparticle-assisted Raman thermometry, thermal modeling, and infrared thermography were performed to assess the device self-heating under the high current and power conditions. This demonstration shows the promise of MOVPE technique for the realization of high-performance lateral $\beta$-Ga$_2$O$_3$ devices and also highlights the need for device-level thermal management.

The computational power increases over the past decades havegreatly enhanced the ability to simulate chemical reactions andunderstand ever more complex transformations. Tensor contractions are the fundamental computational building block of these simulations. These simulations have often been tied to one platform and restricted in generality by the interface provided to the user. The expanding prevalence of accelerators and researcher demands necessitate a more general approach which is not tied to specific hardware or requires contortion of algorithms to specific hardware platforms. In this paper we present COMET, a domain-specific programming language and compiler infrastructure for tensor contractions targeting heterogeneous accelerators. We present a system of progressive lowering through multiple layers of abstraction and optimization that achieves up to 1.98X speedup for 30 tensor contractions commonly used in computational chemistry and beyond., Comment: Proceeding of the 33rd the Workshop on Languages and Compilers for Parallel Computing (LCPC), October 2020

When solving hard multicommodity network flow problems using an LP-based approach, the number of commodities is a driving factor in the speed at which the LP can be solved, as it is linear in the number of constraints and variables. The conventional approach to improve the solve time of the LP relaxation of a Mixed Integer Programming (MIP) model that encodes such an instance is to aggregate all commodities that have the same origin or the same destination. However, the bound of the resulting LP relaxation can significantly worsen, which tempers the efficiency of aggregating techniques. In this paper, we introduce the concept of partial aggregation of commodities that aggregates commodities over a subset of the network instead of the conventional aggregation over the entire underlying network. This offers a high level of control on the trade-off between size of the aggregated MIP model and quality of its LP bound. We apply the concept of partial aggregation to two different MIP models for the multicommodity network design problem. Our computational study on benchmark instances confirms that the trade-off between solve time and LP bound can be controlled by the level of aggregation, and that choosing a good trade-off can allow us to solve the original large-scale problems faster than without aggregation or with full aggregation., Comment: 27 pages, 12 figures

In this paper, we consider the precoder design for an under-loaded or critically loaded downlink multi-user multiple-input multiple-output (MU-MIMO) communication system. We propose novel precoding and decoding schemes which enhance system performance based on rate splitting at the transmitter and single-stage successive interference cancellation at the receivers. The proposed successive null-space (SNS) precoding scheme utilizes linear combinations of the null-space basis vectors of the successively augmented MIMO channel matrices of the users as precoding vectors to adjust the inter-user-interference experienced by the receivers. We formulate a non-convex weighted sum rate (WSR) optimization problem, and solve it via successive convex approximation to obtain a suboptimal solution for the precoding vectors and the associated power allocation. Our simulation results reveal that the proposed SNS precoders outperform block diagonalization based linear and rate splitting designs, and in many cases, have a relatively small gap to the maximum sum rate achieved by dirty paper coding., Comment: 6pp, 4 figures, accepted to the Workshop on Rate-Splitting (Multiple Access) for 6G at the IEEE Intl. Commun. Conf. (ICC) 2021

Electromagnetic hot-spots at ultra-narrow plasmonic nanogaps carry immense potential to drive detection limits down to few molecules in sensors based on surface enhanced Raman or Fluorescence spectroscopies. However, leveraging the EM hot-spots requires access to the gaps, which in turn depends on the size of the analyte in relation to gap distances. Herein we leverage a well-calibrated process based on self-assembly of block copolymer colloids on full-wafer level to produce high density plasmonic nanopillar arrays exhibiting large number (> 10^10 /cm^2) of uniform inter-pillar EM hot-spots. The approach allows convenient handles to systematically vary the inter-pillar gap distances down to sub-10 nm regime. The results show compelling trends of the impact of analyte dimensions in relation to the gap distances towards their leverage over inter-pillar hot-spots, and the resulting sensitivity in SERS based molecular assays. Comparing the detection of labelled proteins in surface-enhanced Raman and metal-enhanced Fluorescence configurations further reveal the relative advantage of Fluorescence over Raman detection while encountering the spatial limitations imposed by the gaps. Quantitative assays with limits of detection down to picomolar concentrations is realized for both the small organic molecules and the proteins. The well-defined geometries delivered by nanofabrication approach is critical to arriving at realistic geometric models to establish meaningful correlation between structure, optical properties and sensitivity of nanopillar arrays in plasmonic assays. The findings emphasize the need for the rational design of EM hot-spots that take into account the analyte dimensions to drive ultra-high sensitivity in plasmon-enhanced spectroscopies., Comment: 28 pages, 7 figures +1 toc figure

The increased energy and power density required in modern electronics poses a challenge for designing new dielectric polymer materials with high energy density while maintaining low loss at high applied electric fields. Recently, an advanced computational screening method coupled with hierarchical modelling has accelerated the identification of promising high energy density materials. It is well known that the dielectric response of polymeric materials is largely influenced by their phases and local heterogeneous structures as well as operational temperature. Such inputs are crucial to accelerate the design and discovery of potential polymer candidates. However, an efficient computational framework to probe temperature dependence of the dielectric properties of polymers, while incorporating effects controlled by their morphology is still lacking. In this paper, we propose a scalable computational framework based on reactive molecular dynamics with a valence-state aware polarizable charge model, which is capable of handling practically relevant polymer morphologies and simultaneously provide near-quantum accuracy in estimating dielectric properties of various polymer systems. We demonstrate the predictive power of our framework on high energy density polymer systems recently identified through rational experimental-theoretical co-design. Our scalable and automated framework may be used for high-throughput theoretical screenings of combinatorial large design space to identify next-generation high energy density polymer materials.

Performing deep level transient spectroscopy (DLTS) on Schottky diodes, we investigated defect levels below the conduction band minima (Ec) in Czochralski (CZ) grown unintentionally-doped (UID) and vertical gradient freeze (VGF)-grown Zr-doped beta-Ga2O3 crystals. In UID crystals with an electron concentration of 10^17 cm-3, we observe levels at 0.18 eV and 0.46 eV in addition to the previously reported 0.86 (E2) and 1.03 eV (E3) levels. For 10^18 cm-3 Zr-doped Ga2O3, signatures at 0.30 eV (E15) and 0.71 eV (E16) are present. For the highest Zr doping of 5*10^18 cm-3, we observe only one signature at 0.59 eV. Electric field-enhanced emission rates are demonstrated via increasing the reverse bias during measurement. The 0.86 eV signature in the UID sample displays phonon-assisted tunneling enhanced thermal emission and is consistent with the widely reported E2 (FeGa) defect. The 0.71 eV (E16) signature in the lower-Zr-doped crystal also exhibits phonon-assisted tunneling emission enhancement. Taking into account that the high doping in the Zr-doped diodes also increases the electric field, we propose that the 0.59 eV signature in the highest Zr-doped sample likely corresponds to the 0.71 eV signature in lower-doped samples. Our analysis highlights the importance of testing for and reporting on field-enhanced emission especially the electric field present during DLTS and other characterization experiments on beta-Ga2O3 along with the standard emission energy, cross-section, and lambda-corrected trap density. This is important because of the intended use of beta-Ga2O3 in high-field devices and the many orders of magnitude of possible doping., Comment: 18 pages, 3 figures

We link the periodicity of Hitchin's uniformizing Higgs bundle with the arithmetic geometry of its underlying curve. Some new relations are discovered. We also speculate on the whole class of periodic Higgs bundles., Comment: 32 pages. This is the second version. Comments are very welcome!

In this article, we introduce the notion of periodic de Rham bundles over smooth complex curves. We prove that motivic de Rham bundles over smooth complex curves are periodic. We conjecture that irreducible periodic de Rham bundles over smooth complex curves are motivic. We show that the conjecture holds for rank one objects and rigid objects., Comment: 42 pages. This is the second version. Comments are very welcome!

Encoding the complex features of an energy landscape is a challenging task, and often chemists pursue the most salient features (minima and barriers) along a highly reduced space, i.e. 2- or 3-dimensions. Even though disconnectivity graphs or merge trees summarize the connectivity of the local minima of an energy landscape via the lowest-barrier pathways, there is more information to be gained by also considering the topology of each connected component at different energy thresholds (or sublevelsets). We propose sublevelset persistent homology as an appropriate tool for this purpose. Our computations on the configuration phase space of n-alkanes from butane to octane allow us to conjecture, and then prove, a complete characterization of the sublevelset persistent homology of the alkane $C_m H_{2m+2}$ potential energy landscapes, for all $m$, and in all homological dimensions. We further compare both the analytical configurational potential energy landscapes and sampled data from molecular dynamics simulation, using the united and all-atom descriptions of the intramolecular interactions. In turn, this supports the application of distance metrics to quantify sampling fidelity and lays the foundation for future work regarding new metrics that quantify differences between the topological features of high-dimensional energy landscapes., Comment: Minor revisions to address reviewers' comments; final version published in JCP

Coupled cluster Green's function (GFCC) calculation has drawn much attention in the recent years for targeting the molecular and material electronic structure problems from a many body perspective in a systematically improvable way. However, GFCC calculations on scientific computing clusters usually suffer from expensive higher dimensional tensor contractions in the complex space, expensive interprocess communication, and severe load imbalance, which limits it's routine use for tackling electronic structure problems. Here we present a numerical library prototype that is specifically designed for large scale GFCC calculations. The design of the library is focused on a systematically optimal computing strategy to improve its scalability and efficiency. The performance of the library is demonstrated by the relevant profiling analysis of running GFCC calculations on remote giant computing clusters. The capability of the library is highlighted by computing a wide near valence band of a fullerene C60 molecule for the first time at the GFCCSD level that shows excellent agreement with the experimental spectrum.

The quasi-static anisotropic permittivity parameters of electrically insulating gallium oxide (beta-Ga2O3) were determined by terahertz spectroscopy. Polarization-resolved frequency domain spectroscopy in the spectral range from 200 GHz to 1 THz was carried out on bulk crystals along different orientations. Principal directions for permittivity were determined along crystallographic axes c, and b, and reciprocal lattice direction a*. No significant frequency dispersion in the real part of dielectric permittivity was observed in the measured spectral range. Our results are in excellent agreement with recent radio-frequency capacitance measurements as well as with extrapolations from recent infrared measurements of phonon mode and high frequency contributions, and close the knowledge gap for these parameters in the terahertz spectral range. Our results are important for applications of beta-Ga2O3 in high-frequency electronic devices, Comment: Main Manuscript is 6 pages long with 3 figures and 1 table. The submission also contains a Supplementary Info section

By combining temperature-dependent resistivity and Hall effect measurements, we investigate donor state energy in Si-doped \b{eta}-Ga2O3 films grown using metal-organic vapor phase epitaxy (MOVPE). High magnetic field Hall effect measurements (H = +/-90 kOe) showed non-linear Hall resistance for T < 150 K revealing two-band conduction. Further analyses revealed carrier freeze-out characteristics in both bands yielding donor state energies of ~ 33.7 and ~ 45.6 meV. The former is consistent with the donor energy of Si in \b{eta}-Ga2O3 whereas the latter suggests a residual donor state, likely associated with a DX center. This study provides a critical insight into the impurity band conduction and the defect energy states in \b{eta}-Ga2O3 using high-field magnetotransport measurements., Comment: 14 pages

In this work, we use photoluminescence spectroscopy (PL) to monitor changes in the UV, UV', blue, and green emission bands from n-type (010) Ga2O3 films grown by metalorganic vapor phase epitaxy (MOVPE) induced by annealing at different temperatures under O2 ambient. Annealing at successively higher temperatures decreases the overall PL yield and UV intensity at nearly the same rates, indicating the increase in formation of at least one non-radiative defect type. Simultaneously, the PL yield ratios of blue/UV and green/UV increase, suggesting that defects associated with these emissions increase in concentration with O2 annealing. Utilizing the different absorption coefficients of 240 and 266 nm polarization-dependent excitation, we find an overall activation energy for the generation of non-radiative defects of 0.69 eV in the bulk but 1.55 eV near the surface. We also deduce activation energies for the green emission-related defects of 0.60 eV near the surface and 0.89-0.92 eV through the films, whereas the blue-related defects have activation energy in the range 0.43-0.62 eV for all depths. Lastly, we observe hillock surface morphologies and Cr diffusion from the substrate into the film for temperatures above 1050 oC. These observations are consistent with the formation and diffusion of VGa and its complexes as a dominant process during O2 annealing, but further work will be necessary to determine which defects and complexes provide radiative and non-radiative recombination channels and the detailed kinetic processes occurring at surfaces and in bulk amongst defect populations., Comment: 14 pages, 3 figures

Engineering thermal transport in two dimensional materials, alloys and heterostructures is critical for the design of next-generation flexible optoelectronic and energy harvesting devices. Direct experimental characterization of lattice thermal conductivity in these ultra-thin systems is challenging and the impact of dopant atoms and hetero-phase interfaces, introduced unintentionally during synthesis or as part of deliberate material design, on thermal transport properties is not understood. Here, we use non-equilibrium molecular dynamics simulations to calculate lattice thermal conductivity of (Mo|W)Se$_2$ monolayer crystals including Mo$_{1-x}$W$_x$Se$_2$ alloys with substitutional point defects, periodic MoSe$_2$|WSe$_2$ heterostructures with characteristic length scales and scale-free fractal MoSe$_2$|WSe$_2$ heterostructures. Each of these features has a distinct effect on phonon propagation in the crystal, which can be used to design fractal and periodic alloy structures with highly tunable thermal conductivities. This control over lattice thermal conductivity will enable applications ranging from thermal barriers to thermoelectrics.

Parameterization of interatomic forcefields is a necessary first step in performing molecular dynamics simulations. This is a non-trivial global optimization problem involving quantification of multiple empirical variables against one or more properties. We present EZFF, a lightweight Python library for parameterization of several types of interatomic forcefields implemented in several molecular dynamics engines against multiple objectives using genetic-algorithm-based global optimization methods. The EZFF scheme provides unique functionality such as the parameterization of hybrid forcefields composed of multiple forcefield interactions as well as built-in quantification of uncertainty in forcefield parameters and can be easily extended to other forcefield functional forms as well as MD engines.

We report on the growth and characterization of metalorganic vapor-phase epitaxy-grown \b{eta}-(AlxGa1-x)2O3/\b{eta}-Ga2O3 modulation-doped heterostructures. Electron channel is realized in the heterostructure by utilizing a delta-doped \b{eta}-(AlxGa1-x)2O3 barrier. Electron channel characteristics are studied using transfer length method, capacitance-voltage and Hall measurements. Hall sheet charge density of 1.06 x 1013 cm-2 and mobility of 111 cm2/Vs is measured at room temperature. Fabricated transistor showed peak current of 22 mA/mm and on-off ratio of 8 x 106. Sheet resistance of 5.3 k{\Omega}/Square is measured at room temperature, which includes contribution from a parallel channel in \b{eta}-(AlxGa1-x)2O3., Comment: 22 pages, 4 figures

Approximating model predictive control (MPC) policy using expert-based supervised learning techniques requires labeled training data sets sampled from the MPC policy. This is typically obtained by sampling the feasible state-space and evaluating the control law by solving the numerical optimization problem offline for each sample. Although the resulting approximate policy can be cheaply evaluated online, generating large training samples to learn the MPC policy can be time consuming and prohibitively expensive. This is one of the fundamental bottlenecks that limit the design and implementation of MPC policy approximation. This technical note aims to address this challenge, and proposes a novel sensitivity-based data augmentation scheme for direct policy approximation. The proposed approach is based on exploiting the parametric sensitivities to cheaply generate additional training samples in the neighborhood of the existing samples., Comment: Accepted for publication at IEEE Transactions on Automatic Control

Predictive materials synthesis is the primary bottleneck in realizing new functional and quantum materials. Strategies for synthesis of promising materials are currently identified by time-consuming trial and error approaches and there are no known predictive schemes to design synthesis parameters for new materials. We use reinforcement learning to predict optimal synthesis schedules, i.e. a time-sequence of reaction conditions like temperatures and reactant concentrations, for the synthesis of a prototypical quantum material, semiconducting monolayer MoS$_{2}$, using chemical vapor deposition. The predictive reinforcement leaning agent is coupled to a deep generative model to capture the crystallinity and phase-composition of synthesized MoS$_{2}$ during CVD synthesis as a function of time-dependent synthesis conditions. This model, trained on 10000 computational synthesis simulations, successfully learned threshold temperatures and chemical potentials for the onset of chemical reactions and predicted new synthesis schedules for producing well-sulfidized crystalline and phase-pure MoS$_{2}$, which were validated by computational synthesis simulations. The model can be extended to predict profiles for synthesis of complex structures including multi-phase heterostructures and can also predict long-time behavior of reacting systems, far beyond the domain of the MD simulations used to train the model, making these predictions directly relevant to experimental synthesis.