Search

Electroencephalography (EEG) is a prominent non-invasive neuroimaging technique providing insights into brain function. Unfortunately, EEG data exhibit a high degree of noise and variability across subjects hampering generalizable signal extraction. Therefore, a key aim in EEG analysis is to extract the underlying neural activation (content) as well as to account for the individual subject variability (style). We hypothesize that the ability to convert EEG signals between tasks and subjects requires the extraction of latent representations accounting for content and style. Inspired by recent advancements in voice conversion technologies, we propose a novel contrastive split-latent permutation autoencoder (CSLP-AE) framework that directly optimizes for EEG conversion. Importantly, the latent representations are guided using contrastive learning to promote the latent splits to explicitly represent subject (style) and task (content). We contrast CSLP-AE to conventional supervised, unsupervised (AE), and self-supervised (contrastive learning) training and find that the proposed approach provides favorable generalizable characterizations of subject and task. Importantly, the procedure also enables zero-shot conversion between unseen subjects. While the present work only considers conversion of EEG, the proposed CSLP-AE provides a general framework for signal conversion and extraction of content (task activation) and style (subject variability) components of general interest for the modeling and analysis of biological signals., Comment: Accepted for publication at the 37th Conference on Neural Information Processing Systems (NeurIPS 2023)

In this paper, we reflect on the educational challenges and research opportunities in running data visualization design activities in the context of large courses. With the increasing number and sizes of data visualization course, we need to better understand approaches to scaling our teaching efforts. We draw on experiences organizing and facilitating activities primarily based on one instance of a master's course given to about 130 students. We provide a detailed account of the course with particular focus on the purpose, structure, and outcome of six two-hour design activities. Based on this, we reflect on three aspects of the course: First, how the course scale led us to thoroughly plan, evaluate, and revise communication between students, teaching assistants, and lecturers. Second, how we designed learning scaffolds through the design activities, and the reflections we received from students on this matter. Finally, we reflect on the diversity of the students that followed the course, the visualization exercises we used, the projects they worked on, and when to key in on simple boring problems and data sets. Thus, our paper contributes with discussions about balancing topical diversity, scaling courses to many students, and problem-based learning., Comment: Copyright 2023 IEEE. This is the author's version of the article that has been published in the proceedings of IEEE Visualization conference. The final version of this record is available at: https://doi.org/10.1109/EduVis60792.2023.00015

Pourbaix diagrams have long been an essential tool for determining the phase stability of solids and their associated ionic species under electrochemical conditions. In recent years, Pourbaix diagrams have been used for applications ranging from corrosion-resistance alloy design to electrocatalysis, and the data from which they are generated has been enhanced by the availability of materials data in various online databases. However, generation of multi-element Pourbaix diagrams has a critical bottleneck which makes 3-element systems difficult to analyze quickly, and 4 and 5 element systems intractable. In this work, we present a method for constructing Pourbaix diagrams which uses a pre-processing step to circumvent the most egregious computational bottleneck and make many-element Pourbaix diagrams computationally efficient.

The combination of strong spin-orbit coupling, large $g$-factors, and the coupling to a superconductor can be used to create a topologically protected state in a semiconductor nanowire. Here we report on growth and characterization of hybrid epitaxial InAsSb/Al nanowires, with varying composition and crystal structure. We find the strongest spin-orbit interaction at intermediate compositions in zincblende InAs$_{1-x}$Sb$_{x}$ nanowires, exceeding that of both InAs and InSb materials, confirming recent theoretical studies \cite{winkler2016topological}. We show that the epitaxial InAsSb/Al interfaces allows for a hard induced superconducting gap and 2$e$ transport in Coulomb charging experiments, similar to experiments on InAs/Al and InSb/Al materials, and find measurements consistent with topological phase transitions at low magnetic fields due to large effective $g$-factors. Finally we present a method to grow pure wurtzite InAsSb nanowires which are predicted to exhibit even stronger spin-orbit coupling than the zincblende structure., Comment: 10 pages and 5 figures

Water is a life-giving source, fundamental to human existence, yet, over a billion people lack access to clean drinking water. Present techniques for water treatment such as piped, treated water rely on time and resource intensive centralized solutions. In this work, we propose a decentralized device concept that can utilize sunlight to split water into hydrogen and hydrogen peroxide. The hydrogen peroxide can oxidize organics while the hydrogen bubbles out. In enabling this device, we require an electrocatalyst that can oxidize water while suppressing the thermodynamically favored oxygen evolution and form hydrogen peroxide. Using density functional theory calculations, we show that the free energy of adsorbed OH$^*$ can be used as a descriptor to screen for selectivity trends between the 2e$^-$ water oxidation to H$_2$O$_2$ and the 4e$^-$ oxidation to O$_2$. We show that materials that bind oxygen intermediates sufficiently weakly, such as SnO$_2$, can activate hydrogen peroxide evolution. We present a rational design principle for the selectivity in electrochemical water oxidation and identify new material candidates that could perform H$_2$O$_2$ evolution selectively., Comment: 13 pages, 3 figures, 6 pages supporting information

Bacterial biofilms are usually assumed to originate from individual cells deposited on a surface. However, many biofilm-forming bacteria tend to aggregate in the planktonic phase so that it is possible that many natural and infectious biofilms originate wholly or partially from pre-formed cell aggregates. Here, we use agent-based computer simulations to investigate the role of pre-formed aggregates in biofilm development. Focusing on the initial shape the aggregate forms on the surface, we find that the degree of spreading of an aggregate on a surface can play an important role in determining its eventual fate during biofilm development. Specifically, initially spread aggregates perform better when competition with surrounding unaggregated bacterial cells is low, while initially rounded aggregates perform better when competition with surrounding unaggregated cells is high. These contrasting outcomes are governed by a trade-off between aggregate surface area and height. Our results provide new insight into biofilm formation and development, and reveal new factors that may be at play in the social evolution of biofilm communities.

We present a systematic theoretical investigation of the surface properties, stability and reactivity, of rock-salt type alkaline-earth metal oxides including MgO, CaO, SrO, and BaO. The accuracy of commonly used exchange-correlation density functionals (LDA, PBE, RPBE, PBEsol, BEEF-vdW and hybrid HSE) and random-phase approximation (RPA) is evaluated and compared to existing experimental values. Calculated surface energies of the four most stable surface facets under vacuum conditions: the (100) surface, the metal and oxygen terminated octopolar (111), and the (110) surfaces exhibit a monotonic increase in stability from MgO to BaO. On the MgO(100) surface, adsorption of CO, NO, CH4 is characterized by physisorption while H2O chemisorbs, which is in agreement with experimental findings. We further use the on-top metal adsorption of CO and NO molecules to map out the surface energetics of each alkaline-earth metal oxide surface. The considered functionals all qualitatively predict similar adsorption energy trends. The ordering between the adsorption energies on different surface facets can be attributed to differences in the local geometrical surface structure and the electronic structure of the metal constituent of the alkaline-earth metal oxide. The striking observation that CO adsorption strength is weaker than NO adsorption on the (100) terraces as the period of the alkaline-earth metal in the oxide increases, is analyzed in detail in terms of charge redistribution within the {\sigma} and {\pi} channels of adsorbates. Finally, we also present oxygen adsorption and oxygen vacancy formation energies in these oxide systems., Comment: 10 pages, submitted to PRB as regular article

In order to accommodate an increasing demand for carbon nanotubes (CNTs) with desirable characteristics one has to understand the origin of helicity of their structures. Here, through in situ microscopy we demonstrate that the nucleation of a carbon nanotube is initiated by the formation of the carbon cap. Nucleation begins with the formation of a graphene embryo that is bound between opposite step-edges on the nickel catalyst surface. The embryo grows larger as the step-edges migrate along the surface, leading to the formation of a curved carbon cap when the steps flow across the edges of adjacent facets. Further motion of the steps away from the catalyst tip with attached rims of the carbon cap generates the wall of the nanotube. Density Functional Theory calculations bring further insight into the process, showing that step flow occurs by surface self diffusion of the nickel atoms via a step-edge attachment-detachment mechanism. Since the fact that cap forms first in the sequence of stages involved in nanotube growth, we suggest that it originates the helicity of the nanotube. Therefore, the angular distribution of catalyst facets could be exploited as a new parameter for controlling the curvature of the cap and, presumably, the helicity of the nanotube., Comment: 4 figures

We use density functional theory (DFT) to investigate the electronic structure and chemical properties of gold nanoparticles. Different structural families of clusters are compared. For up to 60 atoms we optimize structures using DFT-based simulated annealing. Cluster geometries are found to distort considerably, creating large band gaps at the Fermi level. For up to 200 atoms we consider structures generated with a simple EMT potential and clusters based on cuboctahedra and icosahedra. All types of cluster geometry exhibit jellium-like electronic shell structure. We calculate adsorption energies of several atoms on the cuboctahedral clusters. Adsorption energies are found to vary abruptly at magic numbers. Using a Newns-Anderson model we find that the effect of magic numbers on adsorption energy can be understood from the location of adsorbate-induced states with respect to the cluster Fermi level., Comment: 14 pages, 18 figures

The structure of liquid water at ambient conditions is studied in ab initio molecular dynamics simulations using van der Waals (vdW) density-functional theory, i.e. using the new exchange-correlation functionals optPBE-vdW and vdW-DF2. Inclusion of the more isotropic vdW interactions counteracts highly directional hydrogen-bonds, which are enhanced by standard functionals. This brings about a softening of the microscopic structure of water, as seen from the broadening of angular distribution functions and, in particular, from the much lower and broader first peak in the oxygen-oxygen pair-correlation function (PCF), indicating loss of structure in the outer solvation shells. In combination with softer non-local correlation terms, as in the new parameterization of vdW-DF, inclusion of vdW interactions is shown to shift the balance of resulting structures from open tetrahedral to more close-packed. The resulting O-O PCF shows some resemblance with experiment for high-density water (A. K. Soper and M. A. Ricci, Phys. Rev. Lett., 84:2881, 2000), but not directly with experiment for ambient water. However, an O-O PCF consisting of a linear combination of 70% from vdW-DF2 and 30% from experiment on low-density liquid water reproduces near-quantitatively the experimental O-O PCF for ambient water, indicating consistency with a two-liquid model with fluctuations between high- and low-density regions.

The formation energies of nanostructures play an important role in determining their properties, including the catalytic activity. For the case of 15 different rutile and 8 different perovskite metal oxides, we find that the density functional theory (DFT) calculated formation energies of (2,2) nanorods, (3,3) nanotubes, and the (110) and (100) surfaces may be described semi-quantitatively by the fraction of metal--oxygen bonds broken and the bonding band centers in the bulk metal oxide.

We present a practical scheme for performing error estimates for Density Functional Theory calculations. The approach which is based on ideas from Bayesian statistics involves creating an ensemble of exchange-correlation functionals by comparing with an experimental database of binding energies for molecules and solids. Fluctuations within the ensemble can then be used to estimate errors relative to experiment on calculated quantities like binding energies, bond lengths, and vibrational frequencies. It is demonstrated that the error bars on energy differences may vary by orders of magnitude for different systems in good agreement with existing experience., Comment: 5 pages, 3 figures

Large databases that can be used in the search for new materials with specific properties remain an elusive goal in materials science. The search problem is complicated by the fact that the optimal material for a given application is usually a compromise between a number of materials properties and the price. In this letter we present a database consisting of the lattice parameters, bulk moduli, and heats of formation for over 64,000 ordered metallic alloys, which has been established by direct first-principles density-functional-theory calculations. Furthermore, we use a concept from economic theory, the Pareto-optimal set, to determine optimal alloy solutions for the compromise between low compressibility, high stability and price., Comment: 5 pages, 2 figures, To be published in Appl. Phys. Lett

We present molecular-dynamics simulations of the Si(100) surface in the temperature range 1100-1750K. To describe the total energy and forces we use the Effective-Medium Tight-Binding model. The defect-free surface is found to melt at the bulk melting point, which we determine to be 1650 K, but for a surface with dimer vacancies we find a pre-melting of the first two layers 100 K below the melting point. We show that these findings can rationalize recent experimental studies of the high temperature Si(100) surface., Comment: 13 pages, 7 figures; To appear in Surface Science

The energetics of H$_2$ interacting with the Si(100) surface is studied by means of {\em ab initio} total energy calculations within the framework of density functional theory. We find a direct desorption pathway from the mono-hydride phase which is compatible with experimental activation energies and demonstrate the importance of substrate relaxation for this process. Both the transition state configuration and barrier height depend crucially on the degree of buckling of the Si dimers on the Si(100) surface. The adsorption barrier height on the clean surface is governed by the buckling via its influence on the surface electronic structure. We discuss the consequences of this coupling for adsorption experiments and the relation between adsorption and desorption., Comment: 18 pages + 8 figures

A new method for calculating the total energy of Si systems is presented. The method is based on the effective-medium theory concept of a reference system. Instead of calculating the energy of an atom in the system of interest a reference system is introduced where the local surroundings are similar. The energy of the reference system can be calculated selfconsistently once and for all while the energy difference to the reference system can be obtained approximately. We propose to calculate it using the tight-binding LMTO scheme with the Atomic-Sphere Approximation(ASA) for the potential, and by using the ASA with charge-conserving spheres we are able to treat open system without introducing empty spheres. All steps in the calculational method is {\em ab initio} in the sense that all quantities entering are calculated from first principles without any fitting to experiment. A complete and detailed description of the method is given together with test calculations of the energies of phonons, elastic constants, different structures, surfaces and surface reconstructions. We compare the results to calculations using an empirical tight-binding scheme., Comment: 26 pages (11 uuencoded Postscript figures appended), LaTeX, CAMP-090594-1

A new scheme for constructing approximate effective electron potentials within density-functional theory is proposed. The scheme consists of calculating the effective potential for a series of reference systems, and then using these potentials to construct the potential of a general system. To make contact to the reference system the neutral-sphere radius of each atom is used. The scheme can simplify calculations with partial wave methods in the atomic-sphere or muffin-tin approximation, since potential parameters can be precalculated and then for a general system obtained through simple interpolation formulas. We have applied the scheme to construct electron potentials of phonons, surfaces, and different crystal structures of silicon and aluminum atoms, and found excellent agreement with the self-consistent effective potential. By using an approximate total electron density obtained from a superposition of atom-based densities, the energy zero of the corresponding effective potential can be found and the energy shifts in the mean potential between inequivalent atoms can therefore be directly estimated. This approach is shown to work well for surfaces and phonons of silicon., Comment: 8 pages (3 uuencoded Postscript figures appended), LaTeX, CAMP-090594-2

We simulate the melting of a 71 A diameter cluster of Cu. At low temperatures the crystal exhibits facets. With increasing temperatures the open facets pre-melt, the melted regions coalesce into a liquid envelope containing a crystalline nucleus, and the nucleus finally goes unstable to the supercooled liquid. Using critical droplet theory and experimental data for Cu, we explain the thermodynamics of the coexistence region. The width of the transition scales as (Number of particles) to the power (-1/4)., Comment: 10 pages