arXiv subjects

Sherif Abdulkader Tawfik

Publications and source records attributed to Sherif Abdulkader Tawfik.

15 recordsLinked to original sources

arXiv2026

ChargeFlow: Flow-Matching Refinement of Charge-Conditioned Electron Densities

Accurate charge densities are central to electronic-structure theory, but computing charge-state-dependent densities with density functional theory remains too expensive for large-scale screening and defect workflows. We present ChargeFlow, a flow-matching refinement model that transforms a charge-conditioned superposition of atomic densities into the corresponding DFT electron density on the native periodic real-space grid using a 3D U-Net velocity field. Trained on 9,502 charged Materials Project-derived calculations and evaluated on an external 1,671-structure benchmark spanning perovskites, charged defects, diamond defects, metal-organic frameworks, and organic crystals, ChargeFlow is not uniformly best on every in-distribution class but is strongest on problems dominated by nonlocal charge redistribution and charge-state extrapolation, improving deformation-density error from 3.62% to 3.21% and charge- response cosine similarity from 0.571 to 0.655 relative to a ResNet baseline. The predicted densities remain chemically useful under downstream analysis, yielding successful Bader partitioning on all 1,671 benchmark structures and high-fidelity electrostatic potentials, which positions flow matching as a practical density-refinement strategy for charged materials.

cond-mat.mtrl-sci↗

arXiv2025

AI4X Roadmap: Artificial Intelligence for the advancement of scientific pursuit and its future directions

Artificial intelligence and machine learning are reshaping how we approach scientific discovery, not by replacing established methods but by extending what researchers can probe, predict, and design. In this roadmap we provide a forward-looking view of AI-enabled science across biology, chemistry, climate science, mathematics, materials science, physics, self-driving laboratories and unconventional computing. Several shared themes emerge: the need for diverse and trustworthy data, transferable electronic-structure and interatomic models, AI systems integrated into end-to-end scientific workflows that connect simulations to experiments and generative systems grounded in synthesisability rather than purely idealised phases. Across domains, we highlight how large foundation models, active learning and self-driving laboratories can close loops between prediction and validation while maintaining reproducibility and physical interpretability. Taken together, these perspectives outline where AI-enabled science stands today, identify bottlenecks in data, methods and infrastructure, and chart concrete directions for building AI systems that are not only more powerful but also more transparent and capable of accelerating discovery in complex real-world environments.

physics.soc-ph↗

arXiv2024

Efficient Symmetry-Aware Materials Generation via Hierarchical Generative Flow Networks

Discovering new solid-state materials requires rapidly exploring the vast space of crystal structures and locating stable regions. Generating stable materials with desired properties and compositions is extremely difficult as we search for very small isolated pockets in the exponentially many possibilities, considering elements from the periodic table and their 3D arrangements in crystal lattices. Materials discovery necessitates both optimized solution structures and diversity in the generated material structures. Existing methods struggle to explore large material spaces and generate diverse samples with desired properties and requirements. We propose the Symmetry-aware Hierarchical Architecture for Flow-based Traversal (SHAFT), a novel generative model employing a hierarchical exploration strategy to efficiently exploit the symmetry of the materials space to generate crystal structures given desired properties. In particular, our model decomposes the exponentially large materials space into a hierarchy of subspaces consisting of symmetric space groups, lattice parameters, and atoms. We demonstrate that SHAFT significantly outperforms state-of-the-art iterative generative methods, such as Generative Flow Networks (GFlowNets) and Crystal Diffusion Variational AutoEncoders (CDVAE), in crystal structure generation tasks, achieving higher validity, diversity, and stability of generated structures optimized for target properties and requirements.

cs.LG↗

arXiv2024

Embedding material graphs using the electron-ion potential: application to material fracture

At the heart of the flourishing field of machine learning potentials are graph neural networks, where deep learning is interwoven with physics-informed machine learning (PIML) architectures. Various PIML models, upon training with density functional theory (DFT) material structure-property datasets, have achieved unprecedented prediction accuracy for a range of molecular and material properties. A critical component in the learned graph representation of crystal structures in PIMLs is how the various fragments of the structure's graph are embedded in a neural network. Several of the state-of-art PIML models apply spherical harmonic functions. Such functions are based on the assumption that DFT computes the Coulomb potential of atom-atom interactions. However, DFT does not directly compute such potentials, but integrates the electron-atom potentials. We introduce the direct integration of the external potential (DIEP) methods which more faithfully reflects that actual computational workflow in DFT. DIEP integrates the external (electron-atom) potential and uses these quantities to embed the structure graph into a deep learning model. We demonstrate the enhanced accuracy of the DIEP model in predicting the energies of pristine and defective materials. By training DIEP to predict the potential energy surface, we show the ability of the model in predicting the onset of fracture of pristine and defective carbon nanotubes.

cond-mat.mtrl-sci↗

arXiv2022

Naturally-meaningful and efficient descriptors: machine learning of material properties based on robust one-shot ab initio descriptors

Establishing a data-driven pipeline for the discovery of novel materials requires the engineering of material features that can be feasibly calculated and can be applied to predict a material's target properties. Here we propose a new class of descriptors for describing crystal structures, which we term Robust One-Shot Ab initio (ROSA) descriptors. ROSA is computationally cheap and is shown to accurately predict a range of material properties. These simple and intuitive class of descriptors are generated from the energetics of a material at a low level of theory using an incomplete ab initio calculation. We demonstrate how the incorporation of ROSA descriptors in ML-based property prediction leads to accurate predictions over a wide range of crystals, amorphized crystals, metal-organic frameworks and molecules. We believe that the low computational cost and ease of use of these descriptors will significantly improve ML-based predictions.

cond-mat.mtrl-sci↗

arXiv2021

The current magnetization hypothesis as a microscopic theory of the Ørsted magnetic field induction

A wire that conducts an electric current will give rise to circular magnetic field (the Ørsted magnetic field), which can be calculated using the Maxwell-Ampere equation. For wires with diameters in the macroscopic scale, the Maxwell-Ampere equation is an established physical law that has can reproduce a range of experimental observations. A key implication of this equation is that the induction of Ørsted magnetic field is only a result of the displacement of charge. A possible microscopic origin of Ørsted magnetic induction was suggested in [J. Mag. Mag. Mat. 504, 166660 (2020)] (will be called the current magnetization hypothesis (CMH) thereupon). The present work establishes computationally, using simplified wire models, that the CMH reproduces the results of the Maxwell-Ampere equation for wires with a square cross section. I demonstrate that CMH could resolve the apparent contradiction between the observed induced magnetic field and that predicted by the Maxwell-Ampere equation in nanowires, as was reported in [Phys. Rev. B 99, 014436 (2019)]. The CMH shows that a possible reason for such contradiction is the presence of non-conductive surface layers in conductors.

cond-mat.mes-hall↗

arXiv2020

The generalized spin-orbit interaction: a microscopic origin of the Ørsted magnetic field

This work introduces a generalization of the form of the spin-orbit interaction, the generalized spin-orbit interaction (GSOI). It expresses the magnetic field induced by two charged particles moving with a non-zero relative velocity as a field defined at all points in space, and exists in the reference frames of both particles. This is in contrast to spin-orbit interaction theory, in which the generated magnetic field is defined at only one point in space, and exists in the reference frame of one of the two particles. At the macroscopic scale, it is shown that the GSOI theory implies the same form of the Ørsted magnetic field produced by a current-carrying wire. However, the theory is incompatible with the microscopic form of the Biot-Savart equation that implies that a charged particle induces a magnetic field by having a non-zero velocity. The implications of the GSOI theory on properties of the Ørsted magnetic field in current-carrying atomically thin two-dimensional materials, such as graphene, are discussed. The framework established in this paper aims at re-imagining classical physical concepts in light of an advanced microscopic understanding.

cond-mat.mes-hall↗

arXiv2018

Van der Waals forces control ferroelectric-antiferroelectric ordering in ABP2X6 laminar materials

We show how van der Waals (vdW) forces outcompete covalent and ionic forces to control ferroelectric ordering in CuInP2S6 nanoflakes as well as in CuInP2S6 and CuBiP2Se6 crystals. While the self-assembly of these 2D layered materials is clearly controlled by vdW effects, this result indicates that the internal layer structure is also similarly controlled. Using up to 14 first-principles computational methods, we predict that the bilayers of both materials should be antiferroelectric. However, antiferroelectric nanoflakes and bulk materials are shown to embody two fundamentally different types of inter-layer interactions, with vdW forces strongly favouring one and strongly disfavouring the other compared to ferroelectric ordering. Strong specific vdW interactions involving the Cu atoms control this effect. Thickness-dependent significant cancellation of these two large opposing vdW contributions results in a small net effect that interacts with weak ionic contributions to control ferroelectric ordering.

cond-mat.mtrl-sci↗

arXiv2018

van der Waals forces control the internal chemical structure of monolayers within ABP2X6 lamellar materials

Following the recent demonstration that van der Waals forces control the ferroelectric ordering of layers within nanoflakes and bulk samples of CuBiP2Se6 and CuInP2S6, it is demonstrated that they also control the internal geometrical structure of isolated monolayers of these materials. This internal structure involves large displacements of the copper atoms, either normal to the layer plane or else within the plane, that change its ligation environment. In both cases, the van der Waals dispersion force out-competes traditional bonding effects to control structure. However, we find that the aspects of the dispersion force giving rise to each effect are uncorrelated: long range effects control inter-layer ferroelectric ordering whereas short-range effects control internal layer structure. These conclusions are drawn considering predicted properties of monolayers, bilayers, and bulk materials obtained using 14 density-functional-theory based methods. While the different methods used often predict starkly different quantitative results, they concur as to the basic nature of ABP2X6 materials. Of the methods used, only the PBE-D3 and optPBEvdW methods were found to predict a wide range of observed properties without serious disparity. Finding optimal computational methods remains a significant challenge for which the unusual multi-scale nature of the van der Waals interactions in ABP2X6 materials provides demanding criteria.

cond-mat.mtrl-sci↗

arXiv2018

Superconductivity in intercalated buckled two-dimensional materials: KGe$_2$

Germanene has emerged as a novel two-dimensional material with various interesting properties and applications. Here we report the possibility of superconductivity in a stable potassium intercalated germanene compound, KGe$_2$, with a transition temperature $T_c \sim 11$ K, and an electron-phonon coupling of 1.9. Applying a 5\% tensile strain, which reduces the buckling height by 4.5\%, leads to the reduction of the electron-phonon coupling by 11\% and a slight increase in $T_c \sim 12$ K. That is, strong electron-phonon coupling results from the buckled structure of the germanene layers. Despite being an intercalated van der Waals material similar to intercalated graphite superconductors, it does not possess an occupied interlayer state.

cond-mat.mtrl-sci↗

arXiv2018

Evaluation of van der Waals density functionals for layered materials

In 2012, Bjorkman et al. posed the question "Are we van der Waals ready?" [J. Phys.: Condens. Matter, 2012, 24, 424218] about the ability of ab initio modelling to reproduce van der Waals (vdW) dispersion forces in layered materials. The answer at that time was no, however. Here we report on a new generation of vdW dispersion models and show that one, fractionally-ionic atom (FIA) theory, offers close to quantitative predictions for layered structures. Furthermore, it does so from a qualitatively correct picture of dispersion forces. Other methods, such as D3 and optB88vdW also work well, albeit with some exceptions. We thus argue that we are nearly vdW ready, and that some modern dispersion methods are accurate enough to be used for nanomaterial prediction, albeit with some caution required.

cond-mat.mtrl-sci↗

arXiv2017

First principles investigation of quantum emission from hBN defects

Hexagonal boron nitride (hBN) has recently emerged as a fascinating platform for room-temperature quantum photonics due to the discovery of robust visible light single-photon emitters. In order to utilize these emitters, it is necessary to have a clear understanding of their atomic structure and the associated excitation processes that give rise to this single photon emission. Here we perform density-functional theory (DFT) and constrained DFT calculations for a range of hBN point defects in order to identify potential emission candidates. By applying a number of criteria on the electronic structure of the ground state and the atomic structure of the excited states of the considered defects, and then calculating the Huang-Rhys (HR) factor, we find that the CBVN defect, in which a carbon atom substitutes a boron atom and the opposite nitrogen atom is removed, is a potential emission source with a HR factor of 1.66, in good agreement with the experimental HR factor. We calculate the photoluminescence (PL) line shape for this defect and find that it reproduces a number of key features in the the experimental PL lineshape.

cond-mat.mtrl-sci↗

arXiv2017

Detection of adsorbed transition-metal porphyrins by spin-dependent conductance of graphene nanoribbon

Electronic transport in a zig-zag-edge graphene nanoribbon (GNR) and its modification by adsorbed transition metal porphyrins is studied by means of density functional theory calculations. The detachment reaction of the metal centre of the porphyrin is investigated both in the gas phase and for molecules adsorbed on the GNR. As most metal porphyrins are very stable against this reaction, it is found that these molecules bind only weakly to a perfect nanoribbon. However, interaction with a single-atom vacancy in the GNR results in chemical bonding by the transition metal centre being shared between nitrogen atoms in the porphyrin ring and the carbon atoms next to the vacancy in the GNR. For both the physisorbed and the chemisorbed geometry, the inclusion of van der Waals interaction results in a significant enlargement of the binding energy and reduction of the adsorption height. Electronic transport calculations using non-equilibrium Greens functions show that the conductivity of the GNR is altered by the chemisorbed porphyrin molecules. Since the metal centers of porphyrins carry an element-specific magnetic moment, not only the net conductance, but also the spin-dependent conductance of the GNR is affected. In particular, the adsorption of Ru-porphyrin on the single-atom vacancy results in a very large spin polarization of the current of 88% at small applied source-drain voltages. Based on our results, we suggest that a spin valve constructed from a GNR with ferromagnetic contacts could be used as a sensitive detector that could discriminate between various metal porphyrins.

cond-mat.mes-hall↗

arXiv2017

Near-perfect spin filtering and negative differential resistance in an Fe(II)S complex

Density functional theory and nonequilibrium Green's function calculations have been used to explore spin-resolved transport through the high-spin state of an iron(II)sulfur single molecular magnet. Our results show that this molecule exhibits near-perfect spin filtering, where the spin-filtering efficiency is above 99%, as well as significant negative differential resistance centered at a low bias voltage. The rise in the spin-up conductivity up to the bias voltage of 0.4 V is dominated by a conductive lowest unoccupied molecular orbital, and this is accompanied by a slight increase in the magnetic moment of the Fe atom. The subsequent drop in the spin-up conductivity is because the conductive channel moves to the highest occupied molecular orbital which has a lower conductance contribution. This is accompanied by a drop in the magnetic moment of the Fe atom. These two exceptional properties, and the fact that the onset of negative differential resistance occurs at low bias voltage, suggests the potential of the molecule in nanoelectronic and nanospintronic applications.

cond-mat.mtrl-sci↗

arXiv2016

Robust Solid State Quantum System Operating at 800 K

Realization of Quantum information and communications technologies requires robust, stable solid state single photon sources. However, most existing sources cease to function above cryogenic or room temperature due to thermal ionization or strong phonon coupling which impede their emissive and quantum properties. Here we present an efficient single photon source based on a defect in a van der Waals crystal that is optically stable and operates at elevated temperatures of up to 800 K. The quantum nature of the source and the photon purity are maintained upon heating to 800 K and cooling back to room temperature. Our report of a robust high temperature solid state single photon source constitutes a significant step to-wards practical, integrated quantum technologies for real-world environments.

quant-ph↗