arXiv subjects

Patrick Huber

Publications and source records attributed to Patrick Huber.

At least 91 records · Page 5Linked to original sources

arXiv2021

Laser-Excited Elastic Guided Waves Reveal the Complex Mechanics of Nanoporous Silicon

Nanoporosity in silicon leads to completely new functionalities of this mainstream semiconductor. A difficult to assess mechanics has however significantly limited its application in fields ranging from nanofluidics and biosensorics to drug delivery, energy storage and photonics. Here, we present a study on laser-excited elastic guided waves detected contactless and non-destructively in dry and liquid-infused single-crystalline porous silicon. These experiments reveal that the self-organised formation of 100 billions of parallel nanopores per square centimetre cross section results in a nearly isotropic elasticity perpendicular to the pore axes and an 80% effective stiffness reduction, altogether leading to significant deviations from the cubic anisotropy observed in bulk silicon. Our thorough assessment of the wafer-scale mechanics of nanoporous silicon provides the base for predictive applications in robust on-chip devices and evidences that recent breakthroughs in laser ultrasonics open up entirely new frontiers for in-situ, non-destructive mechanical characterisation of dry and liquid-functionalised porous materials.

cond-mat.mtrl-sci↗

arXiv2021

Predicting Discourse Trees from Transformer-based Neural Summarizers

Previous work indicates that discourse information benefits summarization. In this paper, we explore whether this synergy between discourse and summarization is bidirectional, by inferring document-level discourse trees from pre-trained neural summarizers. In particular, we generate unlabeled RST-style discourse trees from the self-attention matrices of the transformer model. Experiments across models and datasets reveal that the summarizer learns both, dependency- and constituency-style discourse information, which is typically encoded in a single head, covering long- and short-distance discourse dependencies. Overall, the experimental results suggest that the learned discourse information is general and transferable inter-domain.

cs.CL↗

arXiv2021

Adsorption from Binary Liquid Solutions into Mesoporous Silica: A Capacitance Isotherm on 5CB Nematogen/Methanol Mixtures

We present a capacitance method to measure the adsorption of rod-like nematogens (4-cyano-4'-pentylbiphenyl, 5CB) from a binary liquid 5CB/methanol solution into a monolithic mesoporous silica membrane traversed by tubular pores with radii of 5.4 nm at room temperature. The resulting adsorption isotherm is reminiscent of classical type II isotherms of gas adsorption in mesoporous media. Its analysis by a model for adsorption from binary solutions, as inspired by the Brunauer-Emmett-Teller (BET) approach for gas adsorption on solid surfaces, indicates that the first adsorbed monolayer consists of flat-lying (homogeneously anchored) 5CB molecules at the pore walls. An underestimation of the adsorbed 5CB amount by the adsorption model compared to the measured isotherm for high 5CB concentrations hints towards a capillary filling transition in the mesopores similar to capillary condensation, i.e. film-growth at the pore walls is replaced by filling of the pore centers by the liquid crystal. The experimental method and thermodynamic analysis presented here can easily be adapted to other binary liquid solutions and thus allows a controlled filling of mesoporous materials with non-volatile molecular systems.

cond-mat.soft↗

arXiv2021

Dynamics of Water Confined in Mesopores with Variable Surface Interaction

We have investigated the dynamics of liquid water confined in mesostructured porous silica (MCM-41) and periodic mesoporous organosilicas (PMOs) by incoherent quasielastic neutron scattering experiments. The effect of tuning the water/surface interaction from hydrophilic to more hydrophobic on the water mobility, while keeping the pore size in the range 3.5-4.1 nm, was assessed from the comparative study of three PMOs comprising different organic bridging units and the purely siliceous MCM-41 case. An extended dynamical range was achieved by combining time-of-flight (IN5B) and backscattering (IN16B) quasielastic neutron spectrometers providing complementary energy resolutions. Liquid water was studied at regularly spaced temperatures ranging from 300 K to 243 K. In all systems, the molecular dynamics could be described consistently by the combination of two independent motions resulting from fast local motion around the average molecule position and the confined translational jump diffusion of its center of mass. All the molecules performed local relaxations, whereas the translational motion of a fraction of molecules was frozen on the experimental timescale. This study provides a comprehensive microscopic view on the dynamics of liquid water confined in mesopores, with distinct surface chemistries, in terms of non-mobile/mobile fraction, self-diffusion coefficient, residence time, confining radius, local relaxation time, and their temperature dependence. Importantly, it demonstrates that the strength of the water/surface interaction determines the long-time tail of the dynamics, which we attributed to the translational diffusion of interfacial molecules, while the water dynamics in the pore center is barely affected by the interface hydrophilicity.

physics.chem-ph↗

arXiv2021

Paraelectric KH$_2$PO$_4$ Nanocrystals in Monolithic Mesoporous Silica: Structure and Lattice Dynamics

Combining dielectric crystals with mesoporous solids allows a versatile design of functional nanomaterials, where the porous host provides a mechanical rigid scaffold structure and the molecular filling adds the functionalization. Here, we report a study of the complex lattice dynamics of a SiO$_2$:KH$_2$PO$_4$ nanocomposite consisting of a monolithic, mesoporous silica glass host with KH$_2$PO$_4$ nanocrystals embedded in its tubular channels $\sim$12 nm across. A micro-Raman investigation performed in the spectral range of 70-1600 cm$^{-1}$ reveals the complex lattice dynamics of the confined crystals. Their Raman spectrum resembles the one taken from bulk KH$_2$PO$_4$ crystals and thus, along with X-ray diffraction experiments, corroborates the successful solution-based synthesis of KH$_2$PO$_4$ nanocrystals with a structure analogous to the bulk material. We succeeded in observing not only the high-frequency internal modes ($\sim$900-1200 cm$^{-1}$), typical of internal vibrations of the PO$_4$ tetrahedra, but, more importantly, also the lowest frequency modes typical of bulk KH$_2$PO$_4$ crystals. The experimental Raman spectrum was interpreted with a group theory analysis and first-principle lattice dynamics calculations. The analysis of calculated eigen-vectors indicates the involvement of hydrogen atoms in most phonon modes corroborating the substantial significance of the hydrogen subsystem in the lattice dynamics of paraelectric bulk and of KH$_2$PO$_4$ crystals in extreme spatial confinement. A marginal redistribution of relative Raman intensities of the confined compared to unconfined crystals presumably originates in slightly changed crystal fields and interatomic interactions, in particular for the parts of the nanocrystals in close proximity to the silica pore surfaces.

cond-mat.mtrl-sci↗

arXiv2021

Anisotropic Confinement of Chromophores Induces Second-Order Nonlinear Optics in a Nanoporous Photonic Metamaterial

Second-order nonlinear optics is the base for a large variety of devices aimed at the active manipulation of light. However, physical principles restrict its occurrence to non-centrosymmetric, anisotropic matter. This significantly limits the number of base materials exhibiting nonlinear optics. Here, we show that embedding chromophores in an array of conical channels 13 nm across in monolithic silica results in mesoscopic anisotropic matter and thus in a hybrid material showing second-harmonic generation (SHG). This non-linear optics is compared to the one achieved in corona-poled polymer films containing the identical chromophores. It originates in confinement-induced orientational order of the elongated guest molecules in the nanochannels. This leads to a non-centrosymmetric dipolar order and hence to a non-linear light-matter interaction on the sub-wavelength, single-pore scale. Our study demonstrates that the advent of large-scale, self-organised nanoporosity in monolithic solids along with confinement-controllable orientational order of chromophores at the single-pore scale provides a reliable and accessible tool to design materials with a nonlinear meta-optics.

physics.optics↗

arXiv2021

Statistical interpretation of sterile neutrino oscillation searches at reactors

A considerable experimental effort is currently under way to test the persistent hints for oscillations due to an eV-scale sterile neutrino in the data of various reactor neutrino experiments. The assessment of the statistical significance of these hints is usually based on Wilks' theorem, whereby the assumption is made that the log-likelihood is $χ^2$-distributed. However, it is well known that the preconditions for the validity of Wilks' theorem are not fulfilled for neutrino oscillation experiments. In this work we derive a simple asymptotic form of the actual distribution of the log-likelihood based on reinterpreting the problem as fitting white Gaussian noise. From this formalism we show that, even in the absence of a sterile neutrino, the expectation value for the maximum likelihood estimate of the mixing angle remains non-zero with attendant large values of the log-likelihood. Our analytical results are then confirmed by numerical simulations of a toy reactor experiment. Finally, we apply this framework to the data of the Neutrino-4 experiment and show that the null hypothesis of no-oscillation is rejected at the 2.6\,$σ$ level, compared to 3.2\,$σ$ obtained under the assumption that Wilks' theorem applies.

hep-ph↗

arXiv2020

Unsupervised Learning of Discourse Structures using a Tree Autoencoder

Discourse information, as postulated by popular discourse theories, such as RST and PDTB, has been shown to improve an increasing number of downstream NLP tasks, showing positive effects and synergies of discourse with important real-world applications. While methods for incorporating discourse become more and more sophisticated, the growing need for robust and general discourse structures has not been sufficiently met by current discourse parsers, usually trained on small scale datasets in a strictly limited number of domains. This makes the prediction for arbitrary tasks noisy and unreliable. The overall resulting lack of high-quality, high-quantity discourse trees poses a severe limitation to further progress. In order the alleviate this shortcoming, we propose a new strategy to generate tree structures in a task-agnostic, unsupervised fashion by extending a latent tree induction framework with an auto-encoding objective. The proposed approach can be applied to any tree-structured objective, such as syntactic parsing, discourse parsing and others. However, due to the especially difficult annotation process to generate discourse trees, we initially develop a method to generate larger and more diverse discourse treebanks. In this paper we are inferring general tree structures of natural text in multiple domains, showing promising results on a diverse set of tasks.

cs.CL↗

arXiv2020

Do We Really Need That Many Parameters In Transformer For Extractive Summarization? Discourse Can Help !

The multi-head self-attention of popular transformer models is widely used within Natural Language Processing (NLP), including for the task of extractive summarization. With the goal of analyzing and pruning the parameter-heavy self-attention mechanism, there are multiple approaches proposing more parameter-light self-attention alternatives. In this paper, we present a novel parameter-lean self-attention mechanism using discourse priors. Our new tree self-attention is based on document-level discourse information, extending the recently proposed "Synthesizer" framework with another lightweight alternative. We show empirical results that our tree self-attention approach achieves competitive ROUGE-scores on the task of extractive summarization. When compared to the original single-head transformer model, the tree attention approach reaches similar performance on both, EDU and sentence level, despite the significant reduction of parameters in the attention component. We further significantly outperform the 8-head transformer model on sentence level when applying a more balanced hyper-parameter setting, requiring an order of magnitude less parameters.

cs.CL↗

arXiv2020

Precursor Film Spreading during Liquid Imbibition in Nanoporous Photonic Crystals

When a macroscopic droplet spreads, a thin precursor film of liquid moves ahead of the advancing liquid-solid-vapor contact line. Whereas this phenomenon has been explored extensively for planar solid substrates, its presence in nanostructured geometries has barely been studied so far, despite its importance for many natural and technological fluid transport processes. Here we use porous photonic crystals in silicon to resolve by light interferometry capillarity-driven spreading of liquid fronts in pores of few nanometers in radius. Upon spatiotemporal rescaling the fluid profiles collapse on master curves indicating that all imbibition fronts follow a square-root-of-time broadening dynamics. For the simple liquid (glycerol) a sharp front with a widening typical of Lucas-Washburn capillary-rise dynamics in a medium with pore-size distribution occurs. By contrast, for a polymer (PDMS) a precursor film moving ahead of the main menisci entirely alters the nature of the nanoscale transport, in agreement with predictions of computer simulations.

cond-mat.soft↗

arXiv2020

Unleashing the Power of Neural Discourse Parsers -- A Context and Structure Aware Approach Using Large Scale Pretraining

RST-based discourse parsing is an important NLP task with numerous downstream applications, such as summarization, machine translation and opinion mining. In this paper, we demonstrate a simple, yet highly accurate discourse parser, incorporating recent contextual language models. Our parser establishes the new state-of-the-art (SOTA) performance for predicting structure and nuclearity on two key RST datasets, RST-DT and Instr-DT. We further demonstrate that pretraining our parser on the recently available large-scale "silver-standard" discourse treebank MEGA-DT provides even larger performance benefits, suggesting a novel and promising research direction in the field of discourse analysis.

cs.CL↗

arXiv2020

MEGA RST Discourse Treebanks with Structure and Nuclearity from Scalable Distant Sentiment Supervision

The lack of large and diverse discourse treebanks hinders the application of data-driven approaches, such as deep-learning, to RST-style discourse parsing. In this work, we present a novel scalable methodology to automatically generate discourse treebanks using distant supervision from sentiment-annotated datasets, creating and publishing MEGA-DT, a new large-scale discourse-annotated corpus. Our approach generates discourse trees incorporating structure and nuclearity for documents of arbitrary length by relying on an efficient heuristic beam-search strategy, extended with a stochastic component. Experiments on multiple datasets indicate that a discourse parser trained on our MEGA-DT treebank delivers promising inter-domain performance gains when compared to parsers trained on human-annotated discourse corpora.

cs.CL↗

arXiv2020

From Sentiment Annotations to Sentiment Prediction through Discourse Augmentation

Sentiment analysis, especially for long documents, plausibly requires methods capturing complex linguistics structures. To accommodate this, we propose a novel framework to exploit task-related discourse for the task of sentiment analysis. More specifically, we are combining the large-scale, sentiment-dependent MEGA-DT treebank with a novel neural architecture for sentiment prediction, based on a hybrid TreeLSTM hierarchical attention model. Experiments show that our framework using sentiment-related discourse augmentations for sentiment prediction enhances the overall performance for long documents, even beyond previous approaches using well-established discourse parsers trained on human annotated data. We show that a simple ensemble approach can further enhance performance by selectively using discourse, depending on the document length.

cs.CL↗

arXiv2020

Sterile neutrino searches at tagged kaon beams

Tagged kaon beams are attractive neutrino sources, which would provide flavor pure $ν_e$-beams with exactly measured normalization. We point out that this also leads to an anti-tagged flavor pure $ν_μ$-beam, with equally well known normalization. Exposing a 1 kt liquid argon detector at a baseline of 1 km to this combination of unique beams allows to decisively test recent indications by IceCube and Neutrino-4 of sterile neutrino oscillations in the multi-eV range.

hep-ph↗

arXiv2020

Silicon Flexes Muscles: Giant Electrochemical Actuation in a Nanoporous Silicon-Polypyrrole Hybrid Material

The absence of piezoelectricity in silicon makes direct electro-mechanical applications of this mainstream semiconductor impossible. Integrated electrical control of the silicon mechanics, however, would open up new perspectives for on-chip actuorics. Here, we combine wafer-scale nanoporosity in single-crystalline silicon with polymerization of an artificial muscle material inside pore space to synthesize a composite that shows macroscopic electrostrain in aqueous electrolyte. The voltage-strain coupling is 3 orders of magnitude larger than the best-performing ceramics in terms of piezoelectric actuation. We trace this huge electroactuation to the concerted action of 100 billions of nanopores per square centimetre cross-section and to potential-dependent pressures of up to 150 atmospheres at the single-pore scale. The exceptionally small operation voltages (0.4-0.9 V) along with the sustainable and biocompatible base materials make this hybrid promising for bio-actuator applications.

cond-mat.mes-hall↗

arXiv2020

Telling Solar Neutrinos from Solar Axions When You Can't Shut Off the Sun

The XENON1T experiment recently reported an excess of events at low electron recoil energies, which may be due to interactions of solar neutrinos inside the detector via a large neutrino magnetic moment. We point out that a $^{51}$Cr neutrino source placed close to the detector can directly test this hypothesis.

hep-ph↗

arXiv2020

Sterile Neutrinos and the Global Reactor Antineutrino Dataset

We present results from global fits to the available reactor antineutrino dataset, as of Fall 2019, to determine the global preference for a fourth, sterile neutrino. We have separately considered experiments that measure the integrated inverse-beta decay (IBD) rate from those that measure the energy spectrum of IBD events at one or more locations. The software used is the newly developed GLoBESfit tool set which is based on the publicly available GLoBES framework and will be released as open-source software.

hep-ph↗

arXiv2020

Reactor neutrino applications and coherent elastic neutrino nucleus scattering

Potential applications of neutrino detection to nuclear security have been discussed since the 1970s. Recent years have seen great progress in detector technologies based on inverse beta decay, with the demonstration of ton-scale surface-level detectors capable of high quality neutrino spectrum measurements. At the same time coherent elastic neutrino nucleus scattering has been experimentally confirmed in 2017 with neutrinos from stopped pion decay and there is a number of experiments aimed at seeing this reaction with reactor neutrinos. The large cross section and threshold-less nature of this reaction make it plausible to consider it for applications to nuclear security and here, we present a first direct comparison of the two reaction modes.

physics.ins-det↗