Over the past two decades, various machine learning (ML) methods have been broadly used to predict properties from structure of materials, which is valuable in accelerating the discovery and development of new materials for different applications. Convolutional neural networks (CNN), however, have not been widely used in materials science, although they have been demonstrated to possess the state-of-the-art learning capability in the field of computer vision and pattern recognition. Recently, it was shown that CNN are powerful in predicting orientation-dependent properties when using a new structure representation—spatial density map (SDM). However, it is unclear whether the deep learning (DL) framework combining CNN and SDM is useful for predicting rotation-invariant properties because of the lack of rotational invariance of SDMs. Here, we demonstrate that this DL framework can also enable high-fidelity prediction of rotation-invariant properties through two examples, i.e., predicting the flexibility volume of metallic glasses and the grain-boundary segregation energy of polycrystals. This is realized through training CNN models via data augmentation and defining a common protocol to establish a unique local coordinate system to compute SDMs when applying CNN models. The advantages of the DL framework arise from the completeness of the SDM and the excellent learning capability of CNN models. Thanks to the high fidelity of CNN models, additional physical insights into structure–property relationships are obtained. The DL method is expected to be useful for a broad range of topics in materials science.

In ab initio studies of carrier-capture processes in defective semiconductor materials, the single-effective-mode formalism and the static-coupling approximation have become the predominant theoretical approaches for determining carrier-capture coefficients. The single-mode formalism relies on accurate nonequilibrium defect energies obtained from density-functional theory (DFT), where required inputs are a series of configurationally displaced, defect-containing supercells obtained using an interpolative ansatz, and where the DFT outputs are corresponding total energies that have traditionally been postprocessed using a long-established ground-state formulation of finite-size corrections and defect-formation energies. This formulation remains commonly used even though the defects that form a configuration-coordinate (CC) diagram typically exist as structures that are displaced from the ground state. To remedy this inconsistency, Kumagai has recently proposed novel methods for implementing finite-size corrections specifically intended for DFT calculations of the defect energies used to construct CC diagrams and implement the single-mode formalism [Y. Kumagai, Phys. Rev. B 107, L220101 (2023)]. Kumagai's approach builds on the latest finite-size-correction methods introduced to describe vertical charge-state transitions for charge-localizing point defects in semiconductors and insulators [T. Gake et al., Phys. Rev. B 101, 020102 (2020); S. Falletta et al., Phys. Rev. B 102, 041115 (2020)]. The newly identified finite-size artifact treated in these studies is the polarization charge induced on a configurationally frozen defect and its subsequent interaction with a vertical transition in charge state. In this work, we evaluate Kumagai's proposed methodology by applying it in a high-precision DFT study of carrier capture by substitutional ${\mathrm{C}}_{\mathrm{N}}$ in GaN, a well-characterized and technologically relevant defect and material. We have rigorously calculated ${\mathrm{C}}_{\mathrm{N}}$ defect energies across various supercell sizes for each defect configuration and charge state on the hole-capture CC diagram of ${\mathrm{C}}_{\mathrm{N}}$($q=-1$), enabling a direct comparison of the slopes of the defect energies versus inverse cell size with those predicted by Kumagai. The most consequential prediction of Kumagai's method is that these slopes distinctly vary as the square of the linear-interpolation parameter used to construct the nonequilibrium defect configurations. Our results quantitatively support this prediction. Moreover, with these new finite-size corrections and multiple-cell-size DFT calculations in place, we find that the classical energy barrier for hole capture by ${\mathrm{C}}_{\mathrm{N}}$($q=-1$) in GaN decreases to 0.092–0.127 eV. This finding confirms the recent ≈ 0.1 eV prediction of Reshchikov based on the weak temperature dependence for hole capture observed in photoluminescence experiments [M. A. Reshchikov, J. Appl. Phys. 129, 121101 (2021)]. These results stand in stark contrast to previously calculated barriers of 0.486 and 0.73 eV, which also used the single-mode formalism but were obtained by instead using ground-state-based finite-size corrections. Our reduced classical barrier for capture increases the temperature-dependent hole-capture coefficient of a ${\mathrm{C}}_{\mathrm{N}}$($q=-1$) defect by more than two to four orders of magnitude for temperatures of 100–600 K, compared to the previous 0.486 eV results. While other defects may not be as dramatically affected as here, we suggest that incorporating proper finite-size corrections for the vertical-transition-like states embedded within CC diagrams is an essential, yet previously unrecognized, component of accurate modeling of carrier-capture when using the single-effective-mode formalism.

We have investigated the vanadium-based Kagome metal ${\mathrm{YbV}}_3{\mathrm{Sb}}_4$ using density functional theory (DFT) combined with the Wannier function analysis. We explore the electronic properties, de Haas-van Alphen (dHvA) effect, and Fermi surface. Our calculations reveal the metallic characteristic in which the majority of the states around the Fermi energy ${\mathrm{E}}_F$ is contributed by the V-$3d$ orbitals, while the localized Yb-$4f$ states are positioned below it. The inclusion of spin-orbit coupling (SOC) induces the splitting of $\text{Yb-}4f$ states, while its impact on the $\text{V-}3d$ states is moderate. Furthermore, we have incorporated $\mathrm{U}+\mathrm{SOC}$, where the Hubbard parameter, which drastically changes the $\text{Yb-}4f$ states, creates additional splitting, leading to three distinct peaks in the density of states (DOS). Meanwhile, the $\text{V-}3d$ atoms with the Kagome lattice contribute the maximum to the transport properties, exhibit flat bands near the ${\mathrm{E}}_F$, while being protected under SOC and $\mathrm{U}+\mathrm{SOC}$. Herein, we report the vulnerability of the $\text{Yb-}4f$ states under SOC and $\mathrm{U}+\mathrm{SOC}$. Furthermore, the Fermi surface is found to comprise quasi-2D cylindrical sheets centered at the $\mathrm{\Gamma }$ point, along with smaller pockets near the Brillouin zone boundaries, which, under combined $\mathrm{U}+\mathrm{SOC}$, a small spherical pocket emerges, and the cylindrical sheet exhibits slight deformations. The dHvA frequencies reach as high as 70 kilotesla, which increases with tilt angle, exhibiting a nearly parabolic trend as expected for cylindrical orbits, while a low-frequency branch remains below 1 kilotesla. Only the $\mathrm{U}+\mathrm{SOC}$ case shows noticeable modification in both the Fermi surface and the dHvA oscillation. Crucially, the $Z_2$ invariant calculation identifies ${\mathrm{YbV}}_3{\mathrm{Sb}}_4$ as a strong topological metal (${\nu }_0=1$). These findings not only advance our understanding of the underlying quantum phenomena in rare-earth Kagome systems, but also establish ${\mathrm{YbV}}_3{\mathrm{Sb}}_4$ as a compelling and promising platform for exploring intertwined topology and electron correlations in Kagome lattices, thereby offering valuable insights for engineering quantum phases in layered materials.

We investigated the thickness dependences of the interfacial Dzyaloshinskii-Moriya interaction (DMI) and the perpendicular magnetic anisotropy (PMA) in ${\left(\mathrm{Pt}/\mathrm{Co}/\mathrm{W}\right)}_5$ multilayers by systematically varying the Co (from 1.3 to 1.9 nm) and W (from 0 to 3 nm) layer thicknesses. Vibrating sample magnetometer has been used to measure the magnetization at saturation from the Co-thickness dependence of the surface magnetic moment at saturation. Increasing the W spacer thickness induces an exponential decay of the magnetization, emphasizing the strong influence of the top interface on overall magnetic coupling. Both surface and volume PMA constants were quantitatively extracted from microstrip line ferromagnetic resonance (FMR) and Brillouin light scattering (BLS) as well as the magnetic damping and the interfacial DMI. A clear evidence of interfacial effects and magnetic inhomogeneities across the multilayers was provided by the appearance of two distinct resonance modes in the FMR spectra. By combining FMR and BLS, the origin of these modes, attributed to the different magnetic properties of the internal and the external Co magnetic layers in the multilayer, was confirmed. The criterion of skyrmions stability, applied for our samples, revealed the possibility of the existence of such stable chiral structures in some samples.

Semiconductor crosshatch patterns form as a result of strain relaxation and dislocation pileups during growth of lattice mismatched materials. Due to their connection with the internal misfit dislocation network, these patterns are a complex fingerprint of internal strain relaxation and growth anisotropy. Therefore, these patterns not only describe the residual strain state at the surface, but could also provide an indicator of electron transport quality through the material. Here, we present a method utilizing computer thresholding and machine learning to analyze antiferromagnetic (AFM) crosshatch patterns that exhibits this correlation. Our analysis reveals optimized electron transport for moderate values of $\lambda$ (crosshatch wavelength) and $\epsilon$ (crosshatch height), roughly 1 $\mu \mathrm{m}$ and 4 nm, respectively, that define the average wave form of the pattern. Simulated two-dimensional AFM crosshatch patterns are used to train a machine learning model to correlate the crosshatch patterns to dislocation density. This model is used to evaluate the experimental AFM images and predict a dislocation density based on the crosshatch wave form. Predicted dislocation density, experimental AFM crosshatch data, and transport measurements are used to train a final model to predict electron mean free path. Our model shows electron scattering is dominated by elastic effects (e.g., dislocations) below 150 nm ${\overline{\lambda }}_{\mathrm{MFP}}$.

The chemical flexibility of the $R{M}_6{X}_6$ stoichiometry, where an $f$-block element is intercalated in the CoSn structure type, allows for the tuning of flatbands associated with kagome lattices to the Fermi level and for emergent phenomena due to interactions between the $f$- and $d$-electron lattices. Yet, $5f$ members of the “166” compounds are underrepresented compared with $4f$ members. Here, we report single-crystal growth of ${\mathrm{UCr}}_6{\mathrm{Ge}}_6$, which crystallizes in a monoclinically distorted ${\mathrm{Y}}_{0.5}{\mathrm{Co}}_3{\mathrm{Ge}}_3$-type structure. The real-space character of the modulation, which is unique within the $R{M}_6{X}_6$ family, is approximated by a $3\times 1\times 2$ supercell of the average monoclinic cell. The compound has kagome-lattice flatbands near the Fermi level and a moderately enhanced electronic heat capacity, as evidenced by its low-temperature Sommerfeld coefficient ($\gamma =86.5$ mJ ${\mathrm{mol}}^{-1} {\mathrm{K}}^{-2}$) paired with band structure calculations. The small, isotropic magnetization and featureless resistivity of ${\mathrm{UCr}}_6{\mathrm{Ge}}_6$ suggest itinerant uranium $5f$ electrons and Pauli paramagnetism. Angle-resolved photoemission spectroscopy results provide evidence for uranium $5f$ weight at the Fermi level and for a flatband near the Fermi level associated with the chromium $3d$ kagome lattice. The isotropic magnetic behavior of the uranium $5f$ electrons starkly contrasts with localized behavior in other uranium 166 compounds, highlighting the high tunability of the magnetic ground state across the material family.

Solid-state spin defects in wide-band-gap semiconductors are leading candidates for quantum information processing, but systematic identification of suitable host materials remains limited by the cost of first-principles screening across vast chemical spaces. We address this with a composition-only machine learning framework built on heterogeneous Rashomon set ensembles: by contrasting the feature attributions of seven diverse classifiers, we extract consensus design rules that no single model identifies alone—filled valence $s$-, $d$-, and $f$ shells, low chemical heterogeneity, and enrichment in C, S, Si, and O favor quantum compatibility. Screening $\approx 45000$ thermodynamically stable compounds, we identify 122 high-confidence candidates (confidence $>0.95$), recovering most experimentally verified hosts (C, $\mathrm{SiC}, \mathrm{ZnO}$, and $\mathrm{ZnS}$) and predicting unexplored materials including ${\mathrm{TiO}}_2, {\mathrm{PbWO}}_4$, and layered chalcogenides (${\mathrm{HfS}}_2, {\mathrm{ZrS}}_2$). Density-functional perturbation theory calculations on 12 representative materials validate dielectric screening as a coherence proxy ($R^2=0.89$ against experimental $T_2$), and vacancy calculations for ${\mathrm{TiO}}_2$ reveal deep, isolated midgap states favorable for spin-defect hosting. The framework provides transferable, physically grounded design principles for rational quantum materials discovery beyond traditional carbide and nitride hosts.

Sulfur-doped black phosphorus (b-P) has recently emerged as a promising candidate for next-generation electronic and optoelectronic technologies owing to its enhanced environmental stability and tunable electronic properties. In this work, we systematically investigate the effects of sulfur substitution on the elastic, mechanical, and electronic properties of b-P, with a particular focus on its auxetic behavior (i.e., negative Poisson's ratio), using first-principles density functional theory calculations. Our results unveil the fundamental origin of the intrinsic auxetic response in pristine b-P and elucidate how sulfur incorporation alters this behavior. We find that sulfur atoms distort the characteristic bowtie structural motif responsible for the negative Poisson's ratio in b-P, thereby suppressing the in-plane auxeticity. Moreover, the resulting charge redistribution also effectively quenches the out-of-plane auxetic response of b-P. With increasing sulfur content, the bulk modulus and Poisson's ratio increase, whereas the Young's modulus, shear modulus, and Debye temperature decrease. Additionally, sulfur substitution suppresses the semiconducting properties of b-P, giving rise to metallicity. These findings highlight that, although sulfur substitution enhances the environmental stability of b-P, it also substantially modifies its elastic and mechanical properties, particularly the auxetic behavior, which is an important consideration in the design of nanoscale electronic devices.

By combining Density functional theory-based first-principles simulations with a random distribution model for estimating the relative percentage of distinct types of S complexes in S-hyperdoped silicon, we predict a tunability of the insulator-to-metal transition at the critical dopant concentration that is significantly larger than previously reported for other chalcogen-hyperdoped Si systems, driven by the relative abundance of the different complexes. By assuming the possibility to tune the latter during the sample synthesis, we predict that optical absorption spectra of S-hyperdoped Si can be engineered from the short wavelength infrared to far infrared region.

A Moment Tensor Potential (MTP) is developed for equiatomic CoCrNi and CoCrFeNi alloys to enable large-scale simulations of chemical short-range order (CSRO) and defect energetics with near density functional theory (DFT) accuracy. The training database includes unary to quaternary configurations as well as DFT derived short-range ordered structures, ensuring accurate representation of thermodynamically relevant chemical environments. The MTP reproduces DFT energies, forces, stresses, and elastic constants across diverse compositions and phases. Hybrid Monte Carlo/molecular dynamics simulations recover key CSRO trends, including Cr–Cr and Fe–Fe repulsion and Ni–Cr ordering, with correct sign structure across temperature. Intrinsic stacking fault (ISF) energies are predicted to be $54\mathrm{mJ}/{\mathrm{m}}^2$ for CoCrNi and $36\mathrm{mJ}/{\mathrm{m}}^2$ for CoCrFeNi, consistent with DFT trends and sensitive to local chemical fluctuations. The potential further enables simulations of solute segregation at grain boundaries while preserving structural units. The developed MTP enables transferable, physically consistent simulations of thermodynamic stability and defect mediated mechanical response in fcc medium entropy alloys across extended length and timescales.

Microstructural engineering represents a promising avenue toward controlling the macroscopic response of high-performance magnetic materials, yet the physical origins linking microstructure to magnetic properties remain to be fully established. In this contribution, we establish magnetoplastic control over macroscopic magnetic properties by inducing high densities of extended lattice defects within the prototypical $L{2}_1$-ordered intermetallic ${\mathrm{MnCu}}_2\mathrm{Al}$, which serves as an ideal model system. We demonstrate a nearly 95% decrease in the initial net saturation magnetization following a high degree of plastic deformation, with effects that are reversible through annealing. Synchrotron X-ray diffraction and scanning electron nanodiffraction permit microscopic correlation of the changes in magnetic behavior with increasing defect content. Microstructural characterization at the single defect level, coupled with detailed first-principles modeling, suggests the presence of a local antiferromagnetic coupling within the extended defects that is at the origin of the dramatic magnetoplastic effect in these magnetic intermetallics. We ascribe these effects to the planar dissociation of dislocations hosting intervening antiphase boundaries, altering the atomic environment and, in turn, the magnetic coupling in the vicinity of dislocations.

Er-substituted yttrium iron garnet (Er:YIG) holds the potential of combining the low magnetic damping of YIG with the telecom-band optical transitions of ${\text{Er}}^{3+}$ ions, making it a suitable material for hybrid optomagnonic devices and microwave-to-optical quantum transduction. We report the epitaxial growth of ${\text{Er}}_x{\text{Y}}_{3-x}{\text{Fe}}_5{\text{O}}_{12}$ films with $x=0.008–0.2$ on (111)-oriented gadolinium gallium garnet (GGG) substrates using pulsed laser deposition. X-ray diffraction, reciprocal space mapping, and scanning transmission electron microscopy confirm single-phase, fully coherent growth with atomically sharp interfaces across the entire substitution range. Magnetometry reveals a gradual decrease in saturation magnetization with increasing Er content, consistent with antiparallel coupling between ${\mathrm{Er}}^{3+}$ spins and the net ${\mathrm{Fe}}^{3+}$ moments, along with the emergence of an in-plane uniaxial magnetic anisotropy. The ferromagnetic resonance broadens with Er content due to increased Gilbert damping and inhomogeneous linewidth broadening. Er:YIG films with $x=0.008$, most relevant for optomagnonic applications, retain nearly isotropic magnetization and exhibit a damping parameter only slightly higher than that of undoped YIG. These results identify growth and substitution conditions that preserve YIG's low-loss magnetic properties while introducing optical functionality, establishing Er:YIG as a viable platform for hybrid quantum magnonics and microwave-to-optical transduction.

We report on the metamagnetic transition observed in oxides ${\mathrm{Ba}}_2{Ln\mathrm{FeO}}_5$ ($\mathit{Ln}$ = Tb, Dy, Ho, Er, Yb), which contain two magnetic sublattices: ${\mathrm{Fe}}^{3+}$ and ${Ln}^{3+}$. The magnetic ordering in these materials results from a competition between several types of interaction (Fe-Fe, Fe-$\mathit{Ln}$, and $\mathit{Ln}-\mathit{Ln}$), combined in most cases with a strong easy-axis anisotropy of the ${Ln}^{3+}$ magnetic moments induced by the crystal electric field (CEF). A specific mechanism is proposed to account for the peculiar metamagnetism in this class of materials, and its variation along the $\mathit{Ln}$ series. This study is based on a set of results including magnetization, x-ray magnetic circular dichroism, neutron powder diffraction, and calculations of the CEF.

Diamond is renowned for supreme hardness and supersized electronic bandgap, but knowledge is surprisingly incomplete on its prominent mechanical and electronic benchmarks under large loads. Here, using first-principles calculations, we create global mapping of extremal stress (GMES) and related strain and modulus profiles to bring out fresh perspectives on material behaviors at different scales (small versus large) and varieties (tension, compression or shear) of loads, unveiling patterns and trends of mechanical properties and record-shattering maximal stress. Moreover, we construct global mapping of electronic bandgap (GMEB) profiles to set benchmarks for rational modulation of electronic properties of diamond. This distinct protocol offers a fresh expansive view of key benchmarks beyond the prevailing paradigm, highlighting global emergent behaviors of extreme mechanics and electronics of diamond under ultimate elastic strain engineering.

${\mathrm{CuLa}}_2{\mathrm{Ge}}_2{\mathrm{O}}_8$ forms a distorted triangular lattice of quantum spin-1/2 ${\mathrm{Cu}}^{2+}$ ions. A crystal growth method was developed using the traveling-solvent floating zone technique, resulting in the synthesis of a large single crystal ($4\mathrm{mm}\times 4\mathrm{mm}\times 10$ mm). The crystal was characterized with regard to phase purity and crystallinity using powder x-ray diffraction, energy dispersive x-ray analysis, and Laue diffraction, and found to be of excellent quality. The magnetic properties were characterized using dc-susceptibility, magnetization, and heat capacity measurements, which revealed weak magnetic frustration with long-range magnetic order occurring below $T_N=1.14\left(1\right)$ K. The magnetic structure determined using neutron powder diffraction is a commensurate, noncollinear antiferromagnetic structure, different from the ${120}^{\circ }$ order of an equilateral triangular antiferromagnet. The ordered moments lie in the bc-plane, with components $m_b=0.50\left(3\right) {\mu }_B$ and $m_c=0.73\left(5\right) {\mu }_B$ along the b- and c-axes respectively, giving a total ordered moment of $M_{\mathrm{total}}=0.89\left(6\right){\mu }_B/{\mathrm{Cu}}^{2+}$ at 20 mK.