Via WebEx scheduled to begin at 4:45 p.m.
All-solid-state batteries are being considered as one of the most promising technologies for safer, high-energy, and long-term energy storage. However, key materials issues remain unsolved and serious barriers must be overcome for their full-scale commercialization. In this presentation I will discuss some of our approaches to understand the key challenges in solid electrolyte materials. Based on a modified electrochemical measurement and first-principles computations, we show that the electrochemical stability window of solid electrolytes was significantly overestimated from the conventional measurement. Electrochemical decompositions of solid electrolytes occur and can lead to interfacial resistances in solid state batteries. Suppressing the (electro)chemical reactions between electrode and electrolyte by engineering their interphase enables a high performance all-ceramic lithium battery. I will further show the general belief that solid electrolytes can prevent lithium dendrite formation is incorrect. Using time-revolved neutron depth profiling, we visualize the deposition of lithium dendrites directly inside the solid electrolytes, thus highlighting the important role of electronic conductivity in dendrite formation. I will conclude my presentation by outlining my future research.
The isolation of graphene using the mechanical exfoliation technique opened the possibility of obtaining other 2D crystals for the investigation of their physical properties [1,2]. Indeed, short after the first reports on the physical properties of graphene, other 2D materials were isolated and investigated . Thanks to simulations, we have an estimate of potential 2D crystals that can be obtained by exfoliation which amount to ca. 1800 , and high throughput computation has been an important tool to investigate them . In this talk, I will present some of the interesting physical properties that we have found in new 2D materials, and I will discuss why simple models, such as tight binding, are still very important for understanding new physics.
The rise of two-dimensional (2D) materials has opened up possibilities for exploring new physical phenomena that motivate the synthesis of more complex low dimensional systems. In this colloquium, we will discuss doping routes that allow the tunability of electronic properties in 2D semiconducting transition metal dichacogenides (TMDs). Zero dimensional (0D) defects such as vacancies and substitutional dopants within tungsten disulfide (WS2) monolayers, will be discussed. In particular, TMD substitutional doping with CH units can be achieved using a novel radio-frequency plasma assisted (RF-PA) approach. Electron microscopy studies confirmed the presence of CH units within the WS2 lattice of plasma treated islands, and DFT calculations confirm the stability of these CH species in sulfur mono-vacancies. Furthermore, field effect transistors fabricated using these CH-doped WS2 exhibit an ambipolar behavior, instead of the n-type transport showed by pristine WS2. The photoluminescence (PL) emission (at 77K) of defective TMD monolayers will also be presented. In particular, sulfur mono-vacancies are concentrated along the edges of triangular WS2 monocrystals. We observed the appearance of bound excitons located 300 meV below the neutral (A) exciton. DFT calculations reveal that sulfur monovacancies introduce midgap states exactly 300 meV below the edge of the conduction band. High–resolution scanning transmission electron microscopy (HR-STEM) images indicate that edges of the WS2 monolayers that exhibit bound excitons contain a very large concentration of sulfur mono-vacancies. Finally, the challenges and new directions in defect engineering of 2D materials will be introduced.
SNO+ is a multi-purpose experiment whose main purpose is to study the nature of the neutrino mass through observation of neutrino-less double beta decay. Detection of this rare process would indicate that neutrinos are elementary Majorana particles, different from the rest of the standard model family of particles. SNO+ can also measure neutrino oscillation parameters, detect geo and reactor anti-neutrinos and low energy solar neutrinos while its main goal is to search for neutrino-less double beta decay in the isotope Tellurium-130. The first of the three SNO+ phases started in May 2017, with the detector filled with ultra-pure water. SNO+ began the transition to the scintillator phase in late 2018. Later this year, the double-beta decay phase will start when the ultra-pure liquid scintillator will be loaded with 3.9 tonnes of natural tellurium, for a half-life sensitivity larger than 2x10^26 years. Previous results, current status, and the potential and prospect of SNO+ for precise solar neutrino measurements and neutrino-less double beta decay search will be presented.
Since 2004 two-dimensional (2D) materials including graphene, transition metal dichalcogenides (TMDCs) and their heterostructures have continued to draw intense research world-side due to their fascinating new fundamental science and diverse potential applications. However, making and characterization large area 2D materials remain a big challenge. In this talk, I will discuss some recent developments in this area of research including a metal organic chemical vapor deposition technique to produce large area 2D materials. I will also show how we can probe the perfection of the large area 2D materials using a combination of local probe techniques such as atomic force microscopy and transmission electron microscopy, and a unique, newly developed global characterization technique at Rensselaer called azimuthal refection high-energy electron diffraction (ARHEED).
Biography: Professor Gwo-Ching Wang is Travelstead Institute Chair of physics. Her research mainly focuses on the growth and characterization of advanced materials. She is Fellow of American Physical Society, American Vacuum Society, American Association for the Advancement of Sciences, and the Materials Research Society. She served as the Chairman of the Department of Physics, Applied Physics and Astronomy, Rensselaer Polytechnic Institute from 2000 to 2010.
The moiré superlattice formed between two-dimensional (2D) materials provides a powerful tool to engineer novel quantum phenomena. The most striking phenomena emerge in the “strong-coupling” regime, where the periodic moiré potential dominates over the relevant kinetic energy and qualitatively changes the quasiparticle behaviors in both real and momentum space. Electrons in the “strong-coupling” regime have shown intriguing phenomena. However, a similar opportunity to engineer bosonic phases has not been experimentally explored. In this talk, I will show the emergence of intra- and inter-layer moiré excitons, i.e. bosons composed of tightly-bound electron-hole pairs, in WSe2/WS2 superlattices in the “strong-coupling” regime. The strong moiré potential trapped excitons into a periodic boson lattice in the real space, exhibiting exotic behaviors. I will also discuss strongly correlated electron phases emerging in this platform.
Over the last decade, flat optical elements composed of an array of deep-subwavelength dielectric or metallic nanostructures of nanoscale thicknesses – referred to as metasurfaces – have revolutionized the field of optics and nanophotonics. Because of their ability to impart an arbitrary phase, polarization or amplitude modulation to an optical wavefront as well as perform multiple optical transformations simultaneously on the incoming light, they promise to replace the traditional bulk optics in applications requiring compactness, integration and/or multiplexing.
In this talk, we discuss the ability of metasurfaces to arbitrarily shape both the temporal and spatial evolution of optical fields, ranging from the deep-ultraviolet to the terahertz frequency range. This requires independent control over the amplitude, phase and/or polarization, achieved here by designing individual metasurface elements to act as nanoscale half-wave plates. We will discuss the various nanofabrication strategies and material constraints governing for their design for operation at these various frequency ranges and outline the advantages of the metasurface approach to light shaping over the more traditional use of spatial light modulators to do the same.
Finally, we demonstrate the versatility of spatial shaping metasurfaces to be directly integrated on integrated photonic chips for their applications as an interface to quantum or biological systems. Through spatial multiplexing of metasurfaces integrated with grating out-couplers directly on a nanophotonic chip, we show the ability to create arbitrary optical fields in the far-field for applications in cold atom traps, biosensing or LIDAR.
This talk will address two-dimensional materials properties and the use of machine learning to predict and understand dynamical phenomena. The discovery of graphene and related two-dimensional materials enables the possibility of engineering metamaterials with desired electronic and optical properties. Plasmonic nanocrystals are optical metamaterials that consist of engineered structures at the sub-wavelength scale. They exhibit optical properties, such as negative-refractive-index and epsilon-near-zero (ENZ) behavior, that are not found under normal circumstances in nature. We will describe a systematic approach for constructing graphene-based tunable metamaterials that exhibit anisotropic ENZ behavior. Subsequently, we will focus on graphene and Dirac solids that constitute two-dimensional materials where the electronic flow is ultra-relativistic. When graphene is deposited on a substrate with roughness, a local random potential develops through an inhomogeneous charge impurity distribution. This disordered potential induces a chaotic pattern in the electronic flow in the form of current branches. We will describe the physics that governs this ultra-relativistic electronic branched flow and demonstrate analytically and numerically the laws of the onset of branching. Finally we will address Machine learning (ML) methods that are currently employed for understanding physical systems as well as for designing materials. We use ML techniques in graphene and produce results that show how ML can predict the electronic branching by learning from past temporary states of the flow. In addition to the data-driven forecasting, we show how unsupervised neural networks can solve differential equations. We focus on energy-conserving equations and propose an architecture that is time invariant and guarantees the energy conservation through an embedded Hamiltonian symplectic structure.
In this talk I will describe how a radio astronomy search for more of the puzzling objects known as quasars led to the accidental discovery of some even more puzzling radio sources, or pulsars. I will briefly outline the properties of pulsars and recount some earlier instances where pulsars were nearly discovered.
Bio: Jocelyn Bell Burnell inadvertently discovered pulsars as a graduate student in radio astronomy in Cambridge, opening up a new branch of astrophysics - work recognised by the award of a Nobel Prize to her supervisor.
She has subsequently worked in many roles in many branches of astronomy, working part-time while raising a family. She is now a Visiting Academic in Oxford, and the Chancellor of the University of Dundee, Scotland. She has been President of the UK’s Royal Astronomical Society, in 2008 became the first female President of the Institute of Physics for the UK and Ireland, and in 2014 the first female President of the Royal Society of Edinburgh. She was one of the small group of women scientists that set up the Athena SWAN scheme.
She has received many honours, including a $3M Breakthrough Prize in 2018.
The public appreciation and understanding of science have always been important to her, and she is much in demand as a speaker and broadcaster. In her spare time she gardens, listens to choral music and is active in the Quakers. She has co-edited an anthology of poetry with an astronomical theme – ‘Dark Matter; Poems of Space’.
Founded in 1851, Corning Incorporated has a longstanding track record of creating life-changing innovations. Starting with railroad signal lights and the mass production of Edison's light bulb, and including the high performance window glass for every US space mission, and currently the strong glass on almost every touchscreen in the world, Corning's material innovations have had a far-reaching impact.
In this talk, I'll give a brief history of Corning's biggest innovations and the attendant science, discuss how our persistent support of R&D undergirds our success, and close with our plans to continue innovation for another 168 years.
We think of materials such as eggshells, white paint and snow as being opaque because the random arrangement of scatterers within these media frustrates the passage of light. It turns out that they are only opaque in a statistical sense. Remarkably, it is possible to ‘sculpt’ light to make these media transparent. We describe how random matrix theory sheds light on this subject and discuss recent experimental successes.
Recently, machine learning tools have been used to aid in the search for novel materials with desirable properties. Materials informatics – the combination of machine learning with materials science – is a promising area of research which opens up new avenues for materials discovery and the unearthing of physical insights. In this talk, we will use materials informatics to search for new two-dimensional (2D)magnetic materials. The recent discovery of intrinsic ferromagnetism in monolayer CrI3 and bilayer Cr2Ge2Te6 created great interest in 2D materials with intrinsic magnetic
order. How many of these materials exist? What are their properties? We use materials informatics to study the magnetic and thermodynamic properties of 2D materials. Crystal structures based on monolayer Cr2Ge2Te6, of the form A2B2X6, are studied using density functional theory (DFT) calculations and machine learning tools. Magnetic properties, such as the magnetic moment are determined. The formation energies are also calculated and used to estimate the chemical stability. We show that machine learning, combined with DFT, provides a computationally efficient means to predict properties of two-dimensional (2D) magnets. In addition, data analytics provides insights into the microscopic origins of magnetic ordering in 2D. This non-traditional approach to materials research paves the way for the rapid discovery of chemically stable 2D magnetic materials.
Topological defects in ferroic order are extensively studied as templates for unique physical phenomena and in the design of low-dimensional, reconfigurable functional elements such as high-density memory “bits.” Since such defects may offer localized non-bulk properties and low dissipative spatial controllability within a chemically homogenous nanoscale medium, the ability to noninvasively detect and probe the 3D morphology and dynamic of ferroelectric vortex-core in operando, with sub-nanometre precision remains a daunting experimental task. Here, we develop and demonstrate the applicability of X-ray Bragg coherent diffractive imaging (BCDI) to address these challenges in individual ferroelectric and multiferroic nanocrystals. Applicability of BCDI to a wider class of systems are discussed.
Emerging quantum materials, such as novel two-dimensional (2D) materials and topologically nontrivial materials, have gained increasing attention due to their unique electronic and photonic properties. The realization of the optoelectronic applications of these materials still faces several challenges. For example, it is critical to gain clear understandings of (1) the fundamental light-matter interactions, which govern many of the key material properties, and (2) the coupling with other nanostructures, which is a required structure for devices and systems. This talk introduces new discoveries and pioneer work using optical spectroscopic techniques on these critical challenges, and novel applications of 2D materials in sensing. The first part of this talk presents the essential material properties investigated using spectroscopy, including interlayer coupling of Moirè patterns of 2D materials, and anisotropic light-matter interactions of 2D materials and Weyl semimetals. The second part of this talk focuses on the interaction of 2D materials with other nanostructures and the related applications. The interactions of 2D materials and selected organic molecules revealed novel enhancement effect of Raman signals for molecules on 2D surface, which offers a new paradigm in biochemical sensing. The works presented in this talk are significant in fundamental nanoscience, and offer important guidelines for practical applications in optoelectronics, sensing, and quantum technologies. The methodologies used here also provide a framework for the future study of many new quantum materials.
High-energy gamma-ray observations are an essential probe of cosmic-ray acceleration mechanisms. The detection of the highest energy gamma rays and the shortest timescales of variability are the key to improve our understanding of the acceleration processes and the environment of the cosmic accelerators.
The High Altitude Water Cherenkov (HAWC) experiment is a large field of view, multi-TeV, gamma-ray observatory continuously operating at 14,000 ft since March, 2015. The HAWC observatory has an order of magnitude better sensitivity, angular resolution, and background rejection than the previous generation of water-Cherenkov arrays. The improved performance allows us to discover TeV sources, to detect transient events, to study the Galactic diffuse emission at TeV energies, and to measure or constrain the TeV spectra of GeV gamma-ray sources. In addition, HAWC is the only ground-based instrument capable of detecting prompt emission from gamma-ray bursts above 100 GeV.
In this colloquium I will present the most recent results using the first three years of data from the HAWC observatory. I will also briefly mention the exciting perspectives of building a next-generation gamma-ray experiment at very high altitude in the Southern Hemisphere.
Interaction of light with matter lies at the heart of a plethora of fundamental phenomena and technological applications. The strength of this interaction can be controlled by engineering the electromagnetic environment surrounding the matter. In this talk I will discuss the ongoing work in my group to explore emergent material properties that arise through the coherent interaction between material excitations and artificially engineered electromagnetic media. This work is motivated by the quest to understand the ultimate limits of controlling optical transitions, carrier transport, energy harvesting, nonlinear optical response and quantum effects. Specifically, we will discuss the regimes of weak and strong light-matter interactions and their implication on the material response. I will begin with a discussion of strong light-matter coupling realized by embedding two-dimensional materials in a photon box (a.k.a optical cavity) [Finally, I will briefly talk about defect engineering in van der Waals materials as a means to realize deterministic quantum emitters [Optica 5, 1128 (2018)] and approaches to couple their emission to high finesse optical cavities. These four example topics all have the underlying theme of controlling light-matter interaction at the nanoscale as a means for engineering matter to realize emergent optoelectronic properties., 30 (2015)] and approaches to control them optically [Nature Photon. 11, 491 (2017)] and electrically. Following this, I will present results on modifying properties such as energy transfer and luminescence of organic molecular materials through strong light-matter coupling [PNAS 116, 5214 (2019)]. In the second part of the talk, I will discuss the regime of weak light-matter coupling to enhance luminescence from low-dimensional semiconductors by designing artifical optical media based on ideas from topology [Science 336, 205 (2012); PNAS 114, 5125 (2017)].
Plasmonic nanostructures have long been appreciated for their ability to harvest photon energy and transform it into other forms, including chemical energy (through the production of hot charge carriers) and thermal energy. Hot carrier production has enabled a variety of photoelectrochemical reactions at plasmonic nanoparticle electrodes, driven by either hot electrons or hot holes. However, at the same time, thermal energy increases mass transport of reactants to the surface and can shift the standard potential of electrochemical reactions, further impacting the rate of photoelectrochemical reactions. Decoupling the relative contributions of plasmon-mediated local heating from hot carrier effects on these reactions has long been an experimental challenge, because we do not have the ability to control each plasmon decay pathway independently. Moreover, plasmon-generated hot carriers quickly lose their energy due to collisions, generating a distribution of hot carrier energies with varying oxidizing or reducing power. This talk will describe our work to isolate local heating effects from hot carrier effects as well as provide quantitative values for hot carrier energies using scanning electrochemical microscopy (SECM) on gold nanoparticles at semiconductor interfaces. We generate real temperature values as well as effective hot carrier temperatures, allowing us to understand how plasmon excitation promotes reactions on plasmon substrates.
Pearl Jam’s hit, “The Light Years,” declares “We were but stones, light made us stars.” Nanophotonic materials and methods promise to elucidate many unknown dynamic processes at the molecular and nanoscale, provided they can ‘shine’ in reactive environments. Here we present our research developing nanophotonic techniques for dynamic, in-situ imaging of photocatalysis, single cell processes, and in-vivo force transduction at the nanoscale. First, we present methods to visualize plasmon-induced chemical transformations with sub-2nm spatial resolution. Our goal is to help unravel the means by which plasmons mediate and control the local chemistry, and ultimately, use that knowledge to optimize photocatalyst performance. As a model reaction, we study the gas-phase photocatalytic dehydrogenation of Au-Pd systems, in which the Au acts as a plasmonic light absorber and Pd serves as the catalyst. Under controlled hydrogen pressures, temperatures, and illumination conditions, we study the study the kinetics of the desorption reaction triggered by the optical excitation of plasmons. We find that plasmons increase the overall rate more than ten-fold and open a new reaction pathway that is not observed without illumination. These results help elucidate the role of plasmons in light-driven phase transformations, en-route to design of site-selective and product-specific photocatalysts. Second, we combine Raman spectroscopy and deep learning to accurately classify bacteria by both species and antibiotic resistance in a single step. We design a convolutional neural network (CNN) for spectral data and train it to identify 30 of the most common bacterial strains from single-cell Raman spectra, achieving antibiotic treatment identification accuracies exceeding 99% and species identification accuracies similar to leading mass spectrometry identification techniques. Our combined Raman-CNN system represents a proof-of-concept for rapid, culture-free identification of bacteria and their antibiotic resistance. Finally, we introduce a new class of in vivo optical probes to monitor biological forces with high spatial and temporal resolution. Our design is based on upconverting nanoparticles that, when excited in the near-infrared, emit light of a different color and intensity in response to microNewton forces. The nanoparticles are sub-30nm in size, do not bleach or photoblink, and can enable deep tissue imaging with minimal tissue autofluorescence. We present the design, synthesis, and characterization of these nanoparticles both in vitro and in vivo, focusing on the forces generated by the roundworm C. elegans as it feeds and digests its bacterial food. Chronic cytotoxicity assays are used to confirm biocompatibility. Our force measurements are coupled with electrical measurements of muscle contractions in both wild-type and mutant animals, providing insight into the interplay between mechanical, electrical, and chemical signaling in vivo.
Jennifer Dionne is an associate professor of Materials Science and Engineering at Stanford, and an affiliate faculty of the Wu Tsai Neurosciences Institute, TomKat Center for Sustainable Energy, Institute for Immunity, Transplantation, and Infection, and Bio-X. She serves as director of the Department of Energy funded Photonics at Thermodynamic Limits Energy Frontier Research Center and faculty co-director of Stanford’s Photonics Research Center. Jen received her B.S. degrees in Physics and Systems Science and Mathematics from Washington University in St. Louis in 2003 and her Ph. D. in Applied Physics at the California Institute of Technology in 2009, advised by Harry Atwater. Prior to joining Stanford, she served as a postdoctoral researcher in Chemistry at Berkeley, advised by Paul Alivisatos. Jen’s research develops new materials and microscopies to observe chemical and biological processes as they unfold with nanometer scale resolution. She then uses these observations to help improve energy-relevant processes (such as photocatalysis and energy storage) and medical diagnostics. Her work has been recognized with a Moore Inventor Fellowship, the Materials Research Society Young Investigator Award, Adolph Lomb Medal, Sloan Foundation Fellowship, and the Presidential Early Career Award for Scientists and Engineers, and was featured on Oprah’s list of “50 Things that will make you say ‘Wow’!”. When not in the lab, Jen enjoys teaching both undergraduate and graduate classes (including “Waves and Diffraction,” “Materials Chemistry”, “Optoelectronics”, and “Science of the Impossible”), exploring the intersection of art and science, cycling the latest century, and reliving her childhood with her two young sons.
Material informatics is a new initiative which has attracted a lot of attention in recent scientific research. The basic strategy is to construct comprehensive data sets and use machine learning to solve a wide variety of problems in material design and discovery. In pursuit of this goal, a key element is the quality and completeness of the databases used. Recent advance in the development of crystal structure prediction algorithms has made it a complementary and more efficient approach to explore the structure/phase space in materials using computers. In this talk, we discuss the importance of the structural motifs and motif-networks in crystal structure predictions. Correspondingly, powerful methods are developed to improve the sampling of the low-energy structure landscape.
The thermodynamic power conversion efficiency limit for silicon solar cells is close to 33%, while commercially available cells have efficiencies in the 17-20% range. The world record for silicon solar cells has inched upward from 25% to 26.7%, in the past twenty years, using cell thicknesses ranging from 450 microns to 165 microns. Photonic crystal architectures enable broadband light absorption beyond the longstanding Lambertian limit and allow silicon to absorb sunlight nearly as well as a direct-bandgap semiconductor. When combined with state-of-the-art electronics, a technological paradigm shift appears imminent. In this lecture, I describe how wave-interference-based solar light-trapping in realistic photonic crystals can break longstanding barriers, enabling flexible, thin-film, silicon to achieve an unprecedented, single-junction, power conversion of 31% [1, 2].
1. "Towards 30% Power Conversion Efficiency in Thin-Silicon Photonic-Crystal Solar Cells," S. Bhattacharya, I. Baydoun, Mi Lin and Sajeev John, Physical Review Applied, 11, 014005 (2019)
2. “Beyond 30% conversion efficiency in silicon solar cells” S. Bhattacharya and Sajeev John (to be published)
Studying nature directly from fundamental degrees of freedom is often computationally limited by physical characteristics of exponentially growing configuration (Hilbert) spaces with particle number and signal-to-noise problems. This leaves many systems of interest to nuclear and particle physics intractable for known algorithms with current and foreseeable classical computational resources. By leveraging their natural capacity to describe nature, the use of quantum systems themselves to form a computational framework leads to constructions of basic quantum field theories with resource requirements that are expected to scale only polynomially with the precision and size of the system. In this talk, I will present an overview of recent progress in, and the potential for, manipulating controllable quantum devices to pursue computational access to our microscopic descriptions of nature.
Since 2004 two-dimensional (2D) materials including graphene, transition metal dichalcogenides (TMDCs) and their heterostructures have continued to draw intense research world-side due to their fascinating new fundamental science and diverse potential applications. 2D systems contain a very small amount of material and scanning probe microscopy (atomic force microscopy and scanning tunneling microscopy) and transmission electron microscopy are commonly used to study their local structural quality. At Rensselaer we have developed a unique and simple method called azimuthal refection high-energy electron diffraction (ARHEED) to measure the wafer scale structural quality of 2D materials. In this talk, I will introduce the basic principle of ARHEED and its characterization of wafer-scale 2D materials. Several examples including graphene and MoS2 will be presented to illustrate the use of AHREED to probe the large scale integrity of the samples.
Pure carbon structures contain a wealth of information and potential properties, we focus on two, long linear carbon chains and Schwarzites. Long linear carbon chains, a one dimensional sp hybridized carbon chain,
have been observed to be encapsulated by a carbon nanotube. Together this system produces a resonant Raman signal from 1770-1860 cm-1, known as the C-mode. The origin of this signal is still under scrutiny. We explore the nature of Raman activity in long linear carbon chains through the use of first principles density functional theory and identify the effect of exact exchange in calculations using hybrid functionals. With exact exchange the most intense Raman active mode, the longitudinal optical mode, converges, with respect to length, to 1831
cm-1, within the range of reported measurements of the C-mode. Also the electronic gap converges, with respect to length, to 1.8 eV, near the known resonance energy.
Schwarzites are sp2, or hexagonal, triply periodic minimal carbon surfaces with negative Gaussian curvature from the introduction of 7-, 8-, 9-, and 10-membered rings. Recently theoretical impregnation of zeolites, simulating the templating processes, has produced Schwarzites.
We present first principles and classical dynamics calculations of electronic and vibronic density of states calculations to establish a connection between the theoretical structures proposed to the experimentally obtained materials. We identify the theoretical structures as energetically and dynamically stable, graphitic in nature, and semimetallic.
Quantum computers promise to solve problems that are not practically feasible with classical computers, with applications ranging from drug development and quantum chemistry to artificial intelligence and cryptography. In this talk, I will give an overview of the current state of experimental quantum computing, specifically results with superconducting qubits. I will then highlight our work on improving the scalability of superconducting quantum devices by interfacing them with classical superconducting logic. Recent results as well as future experiments will be discussed.
Abstract 1: Adiabatic quantum computers like the D-Wave 2000Q can approximately solve the QUBO problem, which is an NP-Hard problem, and have been shown to outperform classical computers on several instances. Solving the QUBO problem literally means solving virtually any NP-Hard problem like the Traveling Salesman Problem (TSP), Airline Scheduling Problem, Protein Folding Problem, Genotype Imputation Problem etc., thereby enabling significant scientific progress, and potentially saving millions / billions of dollars in logistics, airlines, healthcare and many other industries. However, before QUBO problems are solved on quantum computers, they must be embedded (or compiled) onto the hardware of quantum computers, which in itself is a very hard problem. In this work, we propose an efficient embedding algorithm, that lets us embed QUBO problems fast, uses less qubits and gets the objective function value close to the global minimum value. We then compare the performance of our embedding algorithm to that of D-Wave's embedding algorithm, which is the current state of the art, and show that our embedding algorithm convincingly outperforms D-Wave's embedding algorithm. Our embedding approach works with perfect Chimera graphs, i.e. Chimera graphs with no missing qubits.
Abstract 2: For training unsupervised probabilistic machine learning models, matrix computation and sample generation are the two key steps. While GPUs excel at matrix computation, they use pseudo-random numbers to generate samples. Contrarily, Adiabatic Quantum Processors (AQP) use quantum mechanical systems to generate samples accurately and quickly, but are not suited for matrix computation. We present a Classical-Quantum Hybrid Approach for training unsupervised probabilistic machine learning models, leveraging GPUs for matrix computations and the D-Wave quantum sampling library for sample generation. We compare this approach to classical and quantum approaches across four performance metrics. Our results indicate that while the hybrid approach--which uses one AQP and one GPU--outperforms quantum and one of the classical approaches, it performs comparably to the GPU approach, and is outperformed by the CPU approach, which uses 56 high-end CPUs. Lastly, we compare sampling on AQP versus sampling library and show that AQP performs better.