A5055231928- Machine Learning in Materials Science
- Phase-change materials and chalcogenides
- X-ray Diffraction in Crystallography
- Chalcogenide Semiconductor Thin Films
- Crystallography and molecular interactions
- 2D Materials and Applications
- Crystallization and Solubility Studies
- Inorganic Chemistry and Materials
- Quantum Dots Synthesis And Properties
- Electron and X-Ray Spectroscopy Techniques
- Computational Drug Discovery Methods
- Advanced Chemical Physics Studies
- Solid-state spectroscopy and crystallography
- Graphene research and applications
- Crystal Structures and Properties
- Advanced Thermoelectric Materials and Devices
- Electronic and Structural Properties of Oxides
- Advanced Memory and Neural Computing
- Diamond and Carbon-based Materials Research
- Advanced Electron Microscopy Techniques and Applications
- Advanced Battery Materials and Technologies
- Metal-Organic Frameworks: Synthesis and Applications
- Catalysis and Oxidation Reactions
- Perovskite Materials and Applications
- Advancements in Battery Materials
University of Oxford
2019-2025
RWTH Aachen University
2011-2020
University of Cambridge
2016-2020
London Centre for Nanotechnology
2020
University College London
2020
Thomas Young Centre
2020
Oxfam
2020
Robert Bosch (Germany)
2020
Jülich Aachen Research Alliance
2013-2017
FH Aachen
2013-2016
The computer program LOBSTER (Local Orbital Basis Suite Towards Electronic‐Structure Reconstruction) enables chemical‐bonding analysis based on periodic plane‐wave (PAW) density‐functional theory (DFT) output and is applicable to a wide range of first‐principles simulations in solid‐state materials chemistry. incorporates analytic projection routines described previously this very journal [J. Comput. Chem. 2013 , 34 2557] offers improved functionality. It calculates, among others,...
Simple, yet predictive bonding models are essential achievements of chemistry. In the solid state, in particular, they often appear form visual indicators. Because latter require crystal orbitals to be constructed from local basis sets, application most popular density-functional theory codes (namely, those based on plane waves and pseudopotentials) appears as being ill-fitted retrieve chemical information. this paper, we describe a way re-extract Hamilton-weighted populations plane-wave...
Quantum-chemical computations of solids benefit enormously from numerically efficient plane-wave (PW) basis sets, and together with the projector augmented-wave (PAW) method, latter have risen to one predominant standards in computational solid-state sciences. Despite their advantages, plane waves lack local information, which makes interpretation densities-of-states (DOS) difficult precludes direct use atom-resolved chemical bonding indicators such as crystal orbital overlap population...
Abstract We present an update on recently developed methodology and functionality in the computer program Local Orbital Basis Suite Toward Electronic‐Structure Reconstruction (LOBSTER) for chemical‐bonding analysis periodic systems. LOBSTER is based analytic projection from projector‐augmented wave (PAW) density‐functional theory (DFT) computations (Maintz et al., J. Comput. Chem. 2013 , 34 2557), reconstructing chemical information terms of local, auxiliary atomic orbitals thereby opening...
We introduce a Gaussian approximation potential (GAP) for atomistic simulations of liquid and amorphous elemental carbon. Based on machine-learning representation the density-functional theory (DFT) potential-energy surface, such interatomic potentials enable materials with close-to DFT accuracy but at much lower computational cost. first determine maximum that any finite-range can achieve in carbon structures; then, using novel hierarchical set two-, three-, many-body structural...
Abstract While solid‐state materials are commonly classified as covalent, ionic, or metallic, there cases that defy these iconic bonding mechanisms. Phase‐change (PCMs) for data storage a prominent example: they have been claimed to show “resonant bonding,” but clear definition of this mechanism has lacking. Here, it is shown solids fundamentally different from resonant in the π‐orbital systems benzene and graphene, based on first‐principles vibrational, optical, polarizability properties....
A 2D map is created for solid-state materials based on a quantum-mechanical description of electron sharing and transfer. This intuitively identifies the fundamental nature ionic, metallic, covalent bonding in range elements binary compounds; furthermore, it highlights distinct region mechanism recently termed "metavalent" bonding. Then, shown how this can be extended third dimension by including physical properties application interest. Finally, coordinates yield new insight into Peierls...
Amorphous silicon (a-Si) is a widely studied noncrystalline material, and yet the subtle details of its atomistic structure are still unclear. Here, we show that accurate structural models a-Si can be obtained using machine-learning-based interatomic potential. Our best network by simulated cooling from melt at rate 1011 K/s (that is, on 10 ns time scale), contains less than 2% defects, agrees with experiments regarding excess energies, diffraction data, 29Si NMR chemical shifts. We this...
We present an accurate machine learning (ML) model for atomistic simulations of carbon, constructed using the Gaussian approximation potential (GAP) methodology. The potential, named GAP-20, describes properties bulk crystalline and amorphous phases, crystal surfaces, defect structures with accuracy approaching that direct ab initio simulation, but at a significantly reduced cost. combine structural databases carbon graphene, which we extend substantially by adding suitable configurations,...
The allotropes of boron continue to challenge structural elucidation and solid-state theory. Here we use machine learning combined with random structure searching (RSS) algorithms systematically construct an interatomic potential for boron. Starting from ensembles randomized atomic configurations, alternating single-point quantum-mechanical energy force computations, Gaussian approximation (GAP) fitting, GAP-driven RSS iteratively generate a representation the element's potential-energy...
Machine-learned force fields have transformed the atomistic modelling of materials by enabling simulations ab initio quality on unprecedented time and length scales. However, they are currently limited by: (i) significant computational human effort that must go into development validation potentials for each particular system interest; (ii) a general lack transferability from one chemical to next. Here, using state-of-the-art MACE architecture we introduce single general-purpose ML model,...
Machine learning (ML) approaches enable large-scale atomistic simulations with near-quantum-mechanical accuracy. With the growing availability of these methods, there arises a need for careful validation, particularly physically agnostic models-that is, potentials that extract nature atomic interactions from reference data. Here, we review basic principles behind ML and their validation atomic-scale material modeling. We discuss best practice in defining error metrics based on numerical...
Abstract Computer simulations can play a central role in the understanding of phase-change materials and development advanced memory technologies. However, direct quantum-mechanical are limited to simplified models containing few hundred or thousand atoms. Here we report machine-learning-based potential model that is trained using data be used simulate range germanium–antimony–tellurium compositions—typical materials—under realistic device conditions. The speed our enables atomistic multiple...
Abstract Silicon–oxygen compounds are among the most important ones in natural sciences, occurring as building blocks minerals and being used semiconductors catalysis. Beyond well-known silicon dioxide, there phases with different stoichiometric composition nanostructured composites. One of key challenges understanding Si–O system is therefore to accurately account for its nanoscale heterogeneity beyond length scale individual atoms. Here we show that a unified computational description full...
We study the deposition of tetrahedral amorphous carbon (ta-C) films from molecular dynamics simulations based on a machine-learned interatomic potential trained density-functional theory data. For first time, high sp^{3} fractions in excess 85% observed experimentally are reproduced by means computational simulation, and energy dependence film's characteristics is also accurately described. High confidence direct access to atomic interactions allow us infer microscopic growth mechanism this...
Abstract Despite its simple chemical constitution and unparalleled technological importance, the phase‐change material germanium telluride (GeTe) still poses fundamental questions. In particular, bonding mechanisms in amorphous GeTe have remained elusive to date, owing lack of suitable bond‐analysis tools. Herein, we introduce a indicator for structures, dubbed “bond‐weighted distribution function” (BWDF), apply this method GeTe. The results underline peculiar role homopolar GeGe bonds,...
The phase-change material, Ge2Sb2Te5, is the canonical material ingredient for next-generation storage-class memory devices used in novel computing architectures, but fundamental questions remain regarding its atomic structure and physicochemical properties. Here, we introduce a machine-learning (ML)-based interatomic potential that enables large-scale atomistic simulations of liquid, amorphous, crystalline Ge2Sb2Te5 with an unprecedented combination speed density functional theory (DFT)...
ConspectusThe visualization of data is indispensable in scientific research, from the early stages when human insight forms to final step communicating results. In computational physics, chemistry and materials science, it can be as simple making a scatter plot or straightforward looking through snapshots atomic positions manually. However, result "big data" revolution, these conventional approaches are often inadequate. The widespread adoption high-throughput computation for discovery...
Elemental phosphorus is attracting growing interest across fundamental and applied fields of research. However, atomistic simulations have remained an outstanding challenge. Here, we show that a universally applicable force field for can be created by machine learning (ML) from suitably chosen ensemble quantum-mechanical results. Our model fitted to density-functional theory plus many-body dispersion (DFT + MBD) data; its accuracy demonstrated the exfoliation black violet (yielding...
Machine-learning and DFT modelling, linked to experimental knowledge, yield new insight into the structures reactivity of carbonaceous energy materials.