arXiv subjects

Keith Butler

Publications and source records attributed to Keith Butler.

At least 19 recordsLinked to original sources

arXiv2025

Infrared spectral signatures of light r-process elements in kilonovae

A central question regarding neutron star mergers is whether they are able to produce all the r-process elements, from first to third peak. The high abundances of first-peak elements (atomic number $Z \sim 31-40$) in the solar composition means they may dominate the ejecta mass in kilonovae. We here study theoretical infrared signatures of such light elements with spectral synthesis modelling. By combining state-of-the-art NLTE physics with new radiative and collisional data for these elements, we identify several promising diagnostic lines from Ge, As, Se, Br, Kr and Zr. The models give self-consistent line luminosities and indicate specific features that probe emission volumes at early phases ($\sim$10d), the product of ion mass and electron density in late phases ($\gtrsim$75d), and in some cases direct ionic masses at intermediate phases. Emission by [Se I] 5.03 \mum\ + [Se III] 4.55 \mum\ can produce satisfactory fits to the Spitzer photometry of AT2017gfo. However, the models show consistently that with a Kr/Te and Se/Te ratio following the solar r-process pattern, Kr + Se emission is dominant over Te for the blend at 2.1 \mum\ observed in both AT2017gfo and AT2023vfi. The somewhat better line profile fit with [Te III] may suggest that both AT2017gfo and AT2023vfi had a strongly sub-solar production of the light r-process elements. An alternative scenario could be that Kr + Se in an asymmetric morphological distribution generates the feature. Further JWST spectral data, in particular covering the so far unobserved $>5$ \mum\ region, holds promise to determine the light r-process production of kilonovae, and in particular whether the light elements are made in a slow disk wind or in a fast proto-NS outflow. We identify specific needs for further atomic data on recombination rates and collision strengths for $Z=31-40$ elements.

astro-ph.SR↗

arXiv2025

Self-Optimizing Machine Learning Potential Assisted Automated Workflow for Highly Efficient Complex Systems Material Design

Machine learning interatomic potentials have revolutionized complex materials design by enabling rapid exploration of material configurational spaces via crystal structure prediction with ab initio accuracy. However, critical challenges persist in ensuring robust generalization to unknown structures and minimizing the requirement for substantial expert knowledge and time-consuming manual interventions. Here, we propose an automated crystal structure prediction framework built upon the attention-coupled neural networks potential to address these limitations. The generalizability of the potential is achieved by sampling regions across the local minima of the potential energy surface, where the self-evolving pipeline autonomously refines the potential iteratively while minimizing human intervention. The workflow is validated on Mg-Ca-H ternary and Be-P-N-O quaternary systems by exploring nearly 10 million configurations, demonstrating substantial speedup compared to first-principles calculations. These results underscore the effectiveness of our approach in accelerating the exploration and discovery of complex multi-component functional materials.

cond-mat.mtrl-sci↗

arXiv2025

Two-Dimensional Graphene-like BeO Sheet: A Promising Deep-Ultraviolet Nonlinear Optical Materials System with Strong and Highly Tunable Second Harmonic Generation

Two-dimensional (2D) materials with large band gaps and strong and tunable second-harmonic generation (SHG) coefficients play an important role in the miniaturization of deep-ultraviolet (DUV) nonlinear optical (NLO) devices. Despite the existence of numerous experimentally synthesized 2D materials, none of them have been reported to meet DUV NLO requirements. Herein, to the first time, an experimentally available graphene-like BeO monolayer only formed by NLO-active [BeO3] unit is suggested as a promising 2D DUV NLO material due to its ultrawide band gap (6.86 eV) and a strong SHG effect (\{chi}_"22" ^((2))(2D) = 6.81 {\AA}\times pm/V) based on the first-principles calculations. By applying stacking, strain, and twist engineering methods, several 2D BeO sheets have been predicted, and the flexible structural characteristics endow them with tunable NLO properties. Remarkably, the extremely stress-sensitive out-of-plane \{chi}_"15" ^((2))(2D) and \{chi}_"33" ^((2))(2D) (exceptional 30% change) and the robust in-plane \{chi}_"22" ^((2))(2D) against large strains can be achieved together in AC-, AAC-, AAE, and ACE-stacking BeO sheets under in-plane biaxial strain, exhibiting emergent phenomena uniquely not yet seen in other known 2D NLO materials. Our present results reveal that 2D BeO systems should be a new option for 2D DUV NLO materials.

cond-mat.mes-hall↗

arXiv2023

Quantitative spectroscopy of late O-type main-sequence stars with a hybrid non-LTE method

Context. Late O-type stars at luminosities $\log L/L_\odot \lesssim 5.2$ show weak winds with mass-loss rates lower than 10$^{-8} M_\odot$ yr$^{-1}$. This implies that their photospheric layers are not strongly affected by the stellar wind. Aims. A hybrid non-local thermodynamic equilibrium (non-LTE) approach is tested for analyses of late O-type stars. A sample of 20 mostly sharp-lined Galactic O stars of spectral types O8 to O9.7 and luminosity classes V and IV, previously studied in the literature using full non-LTE model atmospheres, is investigated. Methods. Hydrostatic plane-parallel atmospheric structures and synthetic spectra computed with Kurucz's Atlas12 code together with non-LTE line-formation codes Detail and Surface, which account for the effects of turbulent pressure on the atmosphere, were employed. High-resolution spectra were analysed to derive atmospheric parameters and elemental abundances. Fundamental stellar parameters were derived by considering stellar evolution tracks and Gaia EDR3 parallaxes. Interstellar reddening was characterised by fitting spectral energy distributions from the UV to the mid-IR. Results. A high precision and accuracy is achieved for all derived parameters for 16 sample stars. Turbulent pressure effects turn out have significant effects. Effective temperatures are determined to 1-3% uncertainty levels, surface gravities to 0.05 to 0.10 dex, masses to better than 8%, radii to better than 10%, and luminosities to better than 20% uncertainty typically. Abundances for C, N, O, Ne, Mg, Al, Si are derived with uncertainties of 0.05 to 0.10 dex and for helium within 0.03 to 0.05 dex (1$\sigma$ standard deviations) in general. Distances to the Lac OB1b association and to the open clusters NGC 2244, IC 1805, NGC 457, and IC 1396 are determined as a byproduct.

astro-ph.SR↗

arXiv2022

Unified Graph Neural Network Force-field for the Periodic Table

Classical force fields (FF) based on machine learning (ML) methods show great potential for large scale simulations of materials. MLFFs have hitherto largely been designed and fitted for specific systems and are not usually transferable to chemistries beyond the specific training set. We develop a unified atomisitic line graph neural network-based FF (ALIGNN-FF) that can model both structurally and chemically diverse materials with any combination of 89 elements from the periodic table. To train the ALIGNN-FF model, we use the JARVIS-DFT dataset which contains around 75000 materials and 4 million energy-force entries, out of which 307113 are used in the training. We demonstrate the applicability of this method for fast optimization of atomic structures in the crystallography open database and by predicting accurate crystal structures using genetic algorithm for alloys.

cond-mat.mtrl-sci↗

arXiv2021

Interpretable and Explainable Machine Learning for Materials Science and Chemistry

While the uptake of data-driven approaches for materials science and chemistry is at an exciting, early stage, to realise the true potential of machine learning models for successful scientific discovery, they must have qualities beyond purely predictive power. The predictions and inner workings of models should provide a certain degree of explainability by human experts, permitting the identification of potential model issues or limitations, building trust on model predictions and unveiling unexpected correlations that may lead to scientific insights. In this work, we summarize applications of interpretability and explainability techniques for materials science and chemistry and discuss how these techniques can improve the outcome of scientific studies. We discuss various challenges for interpretable machine learning in materials science and, more broadly, in scientific settings. In particular, we emphasize the risks of inferring causation or reaching generalization by purely interpreting machine learning models and the need of uncertainty estimates for model explanations. Finally, we showcase a number of exciting developments in other fields that could benefit interpretability in material science and chemistry problems.

cond-mat.mtrl-sci↗

arXiv2021

Determining the maximum information gain and optimising experimental design in neutron reflectometry using the Fisher information

An approach based on the Fisher information (FI) is developed to quantify the maximum information gain and optimal experimental design in neutron reflectometry experiments. In these experiments, the FI can be analytically calculated and used to provide sub-second predictions of parameter uncertainties. This approach can be used to influence real-time decisions about measurement angle, measurement time, contrast choice and other experimental conditions based on parameters of interest. The FI provides a lower bound on parameter estimation uncertainties and these are shown to decrease with the square root of measurement time, providing useful information for the planning and scheduling of experimental work. As the FI is computationally inexpensive to calculate, it can be computed repeatedly during the course of an experiment, saving costly beam time by signalling that sufficient data has been obtained; or saving experimental datasets by signalling that an experiment needs to continue. The approach's predictions are validated through the introduction of an experiment simulation framework that incorporates instrument-specific incident flux profiles, and through the investigation of measuring the structural properties of a phospholipid bilayer.

physics.data-an↗

arXiv2020

Semantic based model of Conceptual Work Products for formal verification of complex interactive systems

Many clinical workflows depend on interactive computer systems for highly technical, conceptual work products, such as diagnoses, treatment plans, care coordination, and case management. We describe an automatic logic reasoner to verify objective specifications for these highly technical, but abstract, work products that are essential to care. The conceptual work products specifications serve as a fundamental output requirement, which must be clearly stated, correct and solvable. There is strategic importance for such specifications because, in turn, they enable system model checking to verify that machine functions taken with user procedures are actually able to achieve these abstract products. We chose case management of Multiple Sclerosis (MS) outpatients as our use case for its challenging complexity. As a first step, we illustrate how graphical class and state diagrams from UML can be developed and critiqued with subject matter experts to serve as specifications of the conceptual work product of case management. A key feature is that the specification must be declarative and thus independent of any process or technology. Our Work Domain Ontology with tools from Semantic Web is needed to translate UML class and state diagrams for verification of solvability with automatic reasoning. The solvable model will then be ready for subsequent use with model checking on the system of human procedures and machine functions. We used the expressive rule language SPARQL Inferencing Notation (SPIN) to develop formal representations of the UML class diagram, the state machine, and their interactions. Using SPIN, we proved the consistency of the interactions of static and dynamic concepts. We discussed how the new SPIN rule engine could be incorporated in the Object Management Group (OMG) Ontology Definition Metamodel (ODM)

cs.SE↗

arXiv2019

Machine Learning and Big Scientific Data

This paper reviews some of the challenges posed by the huge growth of experimental data generated by the new generation of large-scale experiments at UK national facilities at the Rutherford Appleton Laboratory site at Harwell near Oxford. Such "Big Scientific Data" comes from the Diamond Light Source and Electron Microscopy Facilities, the ISIS Neutron and Muon Facility, and the UK's Central Laser Facility. Increasingly, scientists are now needing to use advanced machine learning and other AI technologies both to automate parts of the data pipeline and also to help find new scientific discoveries in the analysis of their data. For commercially important applications, such as object recognition, natural language processing and automatic translation, deep learning has made dramatic breakthroughs. Google's DeepMind has now also used deep learning technology to develop their AlphaFold tool to make predictions for protein folding. Remarkably, they have been able to achieve some spectacular results for this specific scientific problem. Can deep learning be similarly transformative for other scientific problems? After a brief review of some initial applications of machine learning at the Rutherford Appleton Laboratory, we focus on challenges and opportunities for AI in advancing materials science. Finally, we discuss the importance of developing some realistic machine learning benchmarks using Big Scientific Data coming from a number of different scientific domains. We conclude with some initial examples of our "SciML" benchmark suite and of the research challenges these benchmarks will enable.

physics.comp-ph↗

arXiv2017

Quantitative spectroscopy of extreme helium stars - Model atmospheres and a non-LTE abundance analysis of BD+10$^\circ$2179?

Extreme helium stars (EHe stars) are hydrogen-deficient supergiants of spectral type A and B. They are believed to result from mergers in double degenerate systems. In this paper we present a detailed quantitative non-LTE spectral analysis for BD+10$^\circ$2179, a prototype of this rare class of stars, using UVES and FEROS spectra covering the range from $\sim$3100 to 10 000 {\AA}. Atmosphere model computations were improved in two ways. First, since the UV metal line blanketing has a strong impact on the temperature-density stratification, we used the Atlas12 code. Additionally, We tested Atlas12 against the benchmark code Sterne3, and found only small differences in the temperature and density stratifications, and good agreement with the spectral energy distributions. Second, 12 chemical species were treated in non-LTE. Pronounced non-LTE effects occur in individual spectral lines but, for the majority, the effects are moderate to small. The spectroscopic parameters give $T_\mathrm{eff}$ = 17 300$\pm$300 K and $\log g$ = 2.80$\pm$0.10, and an evolutionary mass of 0.55$\pm$0.05 $M_\odot$. The star is thus slightly hotter, more compact and less massive than found in previous studies. The kinematic properties imply a thick-disk membership, which is consistent with the metallicity $[$Fe/H$]\approx-1$ and $\alpha$-enhancement. The refined light-element abundances are consistent with the white dwarf merger scenario. We further discuss the observed helium spectrum in an appendix, detecting dipole-allowed transitions from about 150 multiplets plus the most comprehensive set of known/predicted isolated forbidden components to date. Moreover, a so far unreported series of pronounced forbidden He I components is detected in the optical-UV.

astro-ph.SR↗

arXiv2014

A new method for an objective, $\chi^2$-based spectroscopic analysis of early-type stars

A precise quantitative spectral analysis - encompassing atmospheric parameter and chemical elemental abundance determination - is time consuming due to its iterative nature and the multi-parameter space to be explored, especially when done "by eye". A robust automated fitting technique that is as trustworthy as traditional methods would allow for large samples of stars to be analyzed in a consistent manner in reasonable time. We present a semi-automated quantitative spectral analysis technique for early-type stars based on the concept of $\chi^2$ minimization. The method's main features are: far less subjective than typical "by eye" methods, correction for inaccurate continuum normalization, consideration of the whole useful spectral range, simultaneous sampling of the entire multi-parameter space (effective temperature, surface gravity, microturbulence, macroturbulence, projected rotational velocity, radial velocity, elemental abundances) to find the global best solution, applicable also to composite spectra. The method is fast, robust and reliable as seen from formal tests and from a comparison with previous analyses. Consistent quantitative spectral analyses of large samples of early-type stars can be performed quickly with very high accuracy.

astro-ph.SR↗

arXiv2011

Testing common classical LTE and NLTE model atmosphere and line-formation codes for quantitative spectroscopy of early-type stars

It is generally accepted that the atmospheres of cool/lukewarm stars of spectral types A and later are described well by LTE model atmospheres, while the O-type stars require a detailed treatment of NLTE effects. Here model atmosphere structures, spectral energy distributions and synthetic spectra computed with ATLAS9/SYNTHE and TLUSTY/SYNSPEC, and results from a hybrid method combining LTE atmospheres and NLTE line-formation with DETAIL/SURFACE are compared. Their ability to reproduce observations for effective temperatures between 15000 and 35000 K are verified. Strengths and weaknesses of the different approaches are identified. Recommendations are made as to how to improve the models in order to derive unbiased stellar parameters and chemical abundances in future applications, with special emphasis on Gaia science.

astro-ph.SR↗

arXiv2008

HD 271791: An Extreme Supernova Runaway B Star Escaping from the Galaxy

Hyper-velocity stars (HVSs) were first predicted by theory to be the result of the tidal disruption of a binary system by a super-massive black hole (SMBH) that accelerates one component to beyond the Galactic escape velocity (the Hills mechanism). Because the Galactic centre hosts such a SMBH it is the suggested place of origin for HVSs. However, the SMBH paradigm has been challenged recently by the young HVS HD271791 because its kinematics point to a birthplace in the metal-poor rim of the Galactic disc. Here we report the atmosphere of HD271791 to indeed show a sub-solar iron abundance along with an enhancement of the alpha-elements, indicating capture of nucleosynthesis products from a supernova or a more energetic hypernova. This implies that HD271791 is the surviving secondary of a massive binary system disrupted in a supernova explosion. No such run-away star has ever been found to exceed the Galactic escape velocity, hence HD271791 is the first hyper-runaway star. Such a run-away scenario is an alternative to the Hills mechanism for the acceleration of some HVSs with moderate velocities. The observed chemical composition of HD271791 puts invaluable observational constraints on nucleosynthesis in a supernova from the core-collapse of a very massive star (M_ZAMS >= 55 M_Sun), which may be observed as a gamma-ray burst of the long-duration/soft-spectrum type.

astro-ph↗

arXiv2008

A cosmic abundance standard: chemical homogeneity of the solar neighbourhood and the ISM dust-phase composition

A representative sample of unevolved early B-type stars in nearby OB associations and the field is analysed to unprecedented precision using NLTE techniques. The resulting chemical composition is found to be more metal-rich and much more homogeneous than indicated by previous work. A rms scatter of ~10% in abundances is found for the six stars (and confirmed by six evolved stars), the same as reported for ISM gas-phase abundances. A cosmic abundance standard for the present-day solar neighbourhood is proposed, implying mass fractions for hydrogen, helium and metals of X=0.715, Y=0.271 and Z=0.014. Good agreement with solar photospheric abundances as reported from recent 3D radiative-hydrodynamical simulations of the solar atmosphere is obtained. As a first application we use the cosmic abundance standard as a proxy for the determination of the local ISM dust-phase composition, putting tight observational constraints on dust models.

astro-ph↗

arXiv2006

Metal-Rich A-Type Supergiants in M31

We discuss results of an exploratory non-LTE analysis of two metal-rich A-type supergiants in M31. Using comprehensive model atoms we derive accurate atmospheric parameters from multiple indicators and show that non-LTE effects on the abundance determination can be substantial (by a factor 2-3). The non-LTE analysis removes systematic trends apparent in the LTE approach and reduces statistical uncertainties. Characteristic abundance patterns of the light elements provide empirical constraints on the evolution of metal-rich massive stars.

astro-ph↗

arXiv2005

Interpreting the Hydrogen IR Lines - Impact of Improved Electron Collision Data

We evaluate the effect of variations in the electron-impact excitation cross sections on the non-LTE line formation for hydrogen in early-type stars. While the Balmer lines are basically unaffected by the choice of atomic data, the Brackett and Pfund series members allow us to discriminate between the different models. Non-LTE calculations based on the widely-used approximations of Mihalas, Heasley & Auer and of Johnson fail to simultaneously reproduce the observed optical and IR spectra over the entire parameter range. Instead, we recommend a reference model using data from ab-initio calculations up to principal quantum number n<=7 for quantitative work. This model is of general interest due to the ubiquity of the hydrogen spectrum.

astro-ph↗

arXiv2004

Non-LTE line-formation for hydrogen revisited

We discuss aspects of non-LTE line formation for hydrogen in early-type stars. We evaluate the effect of variations in the electron-impact excitation cross sections in model atoms of differing complexity by comparison with observation. While the Balmer lines are basically unaffected by the choice of atomic data, the Paschen, Brackett and Pfund series members allow us to discriminate between the different models. Non-LTE calculations based on the widely-used approximation formulae of Mihalas, Heasley & Auer and of Johnson fail to simultaneously reproduce the optical and IR spectra over the entire parameter range. The use of data from ab-initio calculations up to principal quantum number n<=7 largely solves the problem. We recommend a reference model using the available data. This model is of general interest because of the ubiquity of the hydrogen spectrum.

astro-ph↗

arXiv2004

The solar hydrogen spectrum in non-LTE

We investigate the synthesis of the Balmer and Paschen lines of the quiet Sun, using both classical semi-empirical and theoretical model atmospheres, modern line broadening theory and non-LTE line-formation. The computations alleviate long-standing discrepancies between LTE predictions and the observed lines. Theoretical and semi-empirical model atmospheres without a chromosphere on the one hand and semi-empirical models with a chromosphere on the other produce two physically disjoint solutions for the run of non-LTE level populations, including HII, throughout the model stratification. The resulting synthetic non-LTE line profiles are practically identical and reproduce the observation in either case, despite large differences in the line-formation depths, e.g., a chromospheric origin of the Halpha core (in concordance with observation) versus a photospheric origin. The findings are of much broader interest, assuming the Sun to be a prototype cool dwarf star. A consistent account for chromospheres in cool star analyses is required, due to their potential to change atmospheric structure via non-LTE effects on the ionization balance of hydrogen and thus the free electron pool. The latter in turn affects the main opacity source H-. This will in particular affect the atmospheres of metal-poor and evolved stars, in which the contribution of hydrogen to the electron pool becomes dominant.

astro-ph↗