We show that the Gaussian Approximation Potential (GAP) machine-learning framework can describe complex magnetic potential energy surfaces, taking ferromagnetic iron as a paradigmatic challenging case. The training database includes total energies, forces, and stresses obtained from density-functional theory in the generalized-gradient approximation, and comprises approximately 150,000 local atomic environments, ranging from pristine and defected bulk configurations to surfaces and generalized stacking faults with different crystallographic orientations. We find the structural, vibrational, and thermodynamic properties of the GAP model to be in excellent agreement with those obtained directly from first-principles electronic-structure calculations. There is good transferability to quantities, such as Peierls energy barriers, which are determined to a large extent by atomic configurations that were not part of the training set. We observe the benefit and the need of using highly converged electronic-structure calculations to sample a target potential energy surface. The end result is a systematically improvable potential that can achieve the same accuracy of density-functional theory calculations, but at a fraction of the computational cost.

We investigated the metal-insulator transition for epitaxial thin films of the perovskite ${\mathrm{CaFeO}}_3$, a material with a significant oxygen ligand hole contribution to its electronic structure. We find that biaxial tensile and compressive strain suppress the metal-insulator transition temperature. By combining hard x-ray photoelectron spectroscopy, soft x-ray absorption spectroscopy, and density functional calculations, we resolve the element-specific changes to the electronic structure across the metal-insulator transition. We demonstrate that the Fe sites undergo no observable spectroscopic change between the metallic and insulating states, whereas the O electronic configuration undergoes significant changes. This strongly supports the bond-disproportionation model of the metal-insulator transition for ${\mathrm{CaFeO}}_3$ and highlights the importance of ligand holes in its electronic structure. By sensitively measuring the ligand hole density, however, we find that it increases by $\sim 5–10\%$ in the insulating state, which we ascribe to a further localization of electron charge on the Fe sites. These results provide detailed insight into the metal-insulator transition of negative charge transfer compounds and should prove instructive for understanding metal-insulator transitions in other late transition metal compounds such as the nickelates.

The influence of elastic strain on the lithium vacancy formation and migration in bulk ${\mathrm{LiCoO}}_2$ is evaluated by means of first-principles calculations within density functional theory (DFT). Strain dependent energies are determined directly from defective cells and also within linear elasticity theory from the elastic dipole tensor $\left(G_{ij}\right)$ for ground state and saddle point configurations. We analyze finite size effects in the calculation of $G_{ij}$, compare the predictions of the linear elastic model with those obtained from direct calculations of defective cells under strain, and discuss the differences. Based on our data, we calculate the variations in vacancy concentration and mobility due to the presence of external strain in bulk ${\mathrm{LiCoO}}_2$ cathodes. Our results reveal that elastic in-plane and out-of-plane strains can significantly change the ionic conductivity of bulk ${\mathrm{LiCoO}}_2$ by up to several orders of magnitude and thus strongly affect the performance of Li-secondary batteries.

Theoretical evidence of the existence of 12 inequivalent Dirac cones at the vicinity of the Fermi energy in monolayered ${\mathrm{ZrB}}_2$ is presented. Two-dimensional ${\mathrm{ZrB}}_2$ is a mechanically stable $d$- and $p$-orbital compound exhibiting a unique electronic structure with two Dirac cones out of high-symmetry points in the irreducible Brillouin zone with a small electron-pocket compensation. First-principles calculations demonstrate that while one of the cones is insensitive to lattice expansion, the second cone vanishes for small perturbation of the vertical Zr position. Internal symmetry breaking with external physical stimuli, along with the relativistic effect of spin-orbit coupling, is able to remove selectively the Dirac cones. A rational explanation in terms of $d$- and $p$-orbital mixing is provided to elucidate the origin of the infrequent Dirac cones in a flat structure. The versatility of transition-metal $d$ orbitals combined with the honeycomb lattice provided by the B atoms yields particular features in a two-dimensional material.

Based on first-principles calculations, we predict that the methyl-functionalized III-Bi monolayers, namely III-Bi-$\left({\mathrm{CH}}_3\right){}_2$ (III = Ga, In, Tl) films, own quantum spin hall (QSH) states with band gaps as large as 0.260, 0.304, and 0.843 eV, respectively, making them suitable for room-temperature applications. The topological characteristics are confirmed by $s-p_{x,y}$ band inversion, topological invariant $Z_2$, and topologically protected edge states. Noticeably, for GaBi/InBi-$\left({\mathrm{CH}}_3\right){}_2$, the $s-p_{x,y}$ band inversion occurred in the process of spin-orbital coupling (SOC), while for $\mathrm{TlBi}({\mathrm{CH}}_3){}_2$, the $s-p_{x,y}$ band inversion happened in the process of chemical bonding. Significantly, the nontrivial topological states in III-Bi-$\left({\mathrm{CH}}_3\right){}_2$ films are robust against the mechanical strain and various methyl coverage, making them particularly flexible to substrate choice for device applications. Besides, we find the $h$-BN is an ideal substrate for III-Bi-$\left({\mathrm{CH}}_3\right){}_2$ films to realize large-gap nontrivial topological states. These findings demonstrate that the methyl-functionalized III-Bi films may be good QSH effect platforms for topological electronic devices design and fabrication in spintronics.

We present a computational investigation on heat transport in a single polymer chain of poly-3,4-ethylenedioxythiophene (PEDOT). By applying equilibrium and nonequilibrium molecular dynamics simulations to evaluate the thermal conductivity, as well as by investigating how the polymer chain approaches equilibrium upon a local thermal excitation, we provide a robust picture assessing the anomalous superdiffusive (i.e., intermediate between ballistic and diffusive) character of its thermal transport. This assessment is provided by the present simulations showing that three scaling laws with unlike physical meaning and characterizing the thermal energy transport in one-dimensional systems indeed occur in the very same polymer chain with consistent critical exponents. In order to disentangle the effect of dimensionality, we perform a systematic comparison of transport features for a single one-dimensional (1D) PEDOT chain and a three-dimensional (3D) PEDOT crystal. Present simulations suggest that by increasing the dimensionality, the anomalous regime is completely removed as due to the occurrence of strong interchains anharmonic interactions. Finally, we prove that thermal transport in isolated single PEDOT chains belongs to a novel universality class of superdiffusion characterized by a critical exponent $\beta =1/2$.

The optical properties of ${\mathrm{MoS}}_2$ monolayers are dominated by excitons, but for spectrally broad optical transitions in monolayers exfoliated directly onto ${\mathrm{SiO}}_2$ substrates detailed information on excited exciton states is inaccessible. Encapsulation in hexagonal boron nitride (hBN) allows approaching the homogenous exciton linewidth, but interferences in the van der Waals heterostructures make direct comparison between transitions in optical spectra with different oscillator strength more challenging. Here we reveal in reflectivity and in photoluminescence excitation spectroscopy the presence of excited states of the $A$ exciton in ${\mathrm{MoS}}_2$ monolayers encapsulated in hBN layers of calibrated thickness, allowing us to extrapolate an exciton binding energy of $\approx 220$ meV. We theoretically reproduce the energy separations and oscillator strengths measured in reflectivity by combining the exciton resonances calculated for a screened two-dimensional Coulomb potential with transfer matrix calculations of the reflectivity for the van der Waals structure. Our analysis shows a very different evolution of the exciton oscillator strength with principal quantum number for the screened Coulomb potential as compared to the ideal two-dimensional hydrogen model.

Monolayer transition metal dichalcogenides (TMDs) show an optimal surface-to-volume ratio and are thus promising candidates for novel molecule sensor devices. It was recently predicted that a certain class of molecules exhibiting a large dipole moment can be detected through the activation of optically inaccessible (dark) excitonic states in absorption spectra of tungsten-based TMDs. In this paper, we investigate the molecule signatures in photoluminescence spectra in dependence of a number of different experimentally accessible quantities, such as excitation density, temperature, as well as molecular characteristics including the dipole moment and its orientation, molecule-TMD distance, molecular coverage, and distribution. We show that under certain optimal conditions even room-temperature detection of molecules can be achieved.

We propose an unfolding scheme to analyze energy spectra of complex large-scale systems which are inherently of double periodicity on the basis of the density-functional theory. Applying our method to a twisted bilayer graphene (tBLG) and a stack of monolayer ${\mathrm{MoS}}_2$ on graphene (${\mathrm{MoS}}_2$/graphene) as examples, we first show that the conventional unfolding scheme in the past using a single primitive-cell representation causes serious problems in analyses of the energy spectra. We then introduce our multispace representation scheme in the unfolding method and clarify its validity. Velocity renormalization of Dirac electrons in tBLG and mini gaps of Dirac cones in ${\mathrm{MoS}}_2$/graphene are elucidated in the present unfolding scheme.

Based on first-principles calculations and effective model analysis, a Dirac nodal-net semimetal state is recognized in ${\mathrm{AlB}}_2$-type ${\mathrm{TiB}}_2$ and ${\mathrm{ZrB}}_2$ when spin-orbit coupling (SOC) is ignored. Taking ${\mathrm{TiB}}_2$ as an example, several topological novel states have been found in this nodal-net structure including triple point, nexus, and nodal link, which are protected by the coexistence of spatial-inversion symmetry and time-reversal symmetry. In addition, linearly and quadratically dispersed two-dimensional surface Dirac points have been identified as getting on the B-terminated and Ti-terminated (001) surfaces of ${\mathrm{TiB}}_2$, respectively, which are analogous to those of monolayer and bilayer graphene.

The increasing interest in mobile and wearable technology demands the enhancement of functionality of clothing through incorporation of sophisticated architectures of multifunctional materials. Flexible electronic and photonic devices based on organic materials have made impressive progress over the past decade, but higher performance, simpler fabrication, and most importantly, compatibility with woven technology are desired. Here we report on the development of a weaved, substrateless, and polarization-sensitive photodetector based on doping-engineered fibers of highly aligned carbon nanotubes. This room-temperature-operating, self-powered detector responds to radiation in an ultrabroad spectral range, from the ultraviolet to the terahertz, through the photothermoelectric effect, with a low noise-equivalent power (a few ${\mathrm{nW}/\mathrm{Hz}}^{1/2}$) throughout the range and with a $ZT$-factor value that is twice as large as that of previously reported carbon nanotube-based photothermoelectric photodetectors. Particularly, we fabricated a $\sim 1$-m-long device consisting of tens of $p^{+}\text{−}{p}^{-}$ junctions and weaved it into a shirt. This device demonstrated a collective photoresponse of the series-connected junctions under global illumination. The performance of the device did not show any sign of deterioration through 200 bending tests with a bending radius smaller than 100 $\mu \mathrm{m}$ as well as standard washing and ironing cycles. This unconventional photodetector will find applications in wearable technology that require detection of electromagnetic radiation.

The electronic band structure of ${\mathrm{SrTiO}}_3$ is investigated in the all-electron quasiparticle self-consistent $\mathit{GW}$ ($\mathrm{QS}\mathit{GW}$) approximation. Unlike previous pseudopotential-based $\mathrm{QS}\mathit{GW}$ or single-shot $G_0{W}_0$ calculations, the gap is found to be significantly overestimated compared to experiment. After putting in a correction for the underestimate of the screening by the random phase approximation in terms of a $0.8\mathrm{\Sigma }$ approach, the gap is still overestimated. The $0.8\mathrm{\Sigma }$ approach is discussed and justified in terms of various recent literature results including electron-hole corrections. Adding a lattice polarization correction (LPC) in the $\mathbf{q}\to 0$ limit for the screening of $W$, agreement with experiment is recovered. The LPC is alternatively estimated using a polaron model. We apply our approach to the cubic and tetragonal phases as well as a hypothetical layered postperovskite structure and find that the local density approximation (LDA) to $\mathit{GW}$ gap correction is almost independent of structure.

The ground state of macroscopic samples of magnetically ordered materials is a domain state because of magnetostatic energy or entropy, yet we have limited experimental means for imaging the bulk domain structure and the magnetization process directly. The common methods available reveal the domains at the surface or in electron- or x-ray transparent lamellae, not those in the bulk. The magnetization curve just reflects the vector sum of the moments of all the domains in the sample, but magnetostriction curves are more informative. They are strongly influenced by the domain structure in the unmagnetized state and its evolution during the magnetization process in an applied field. Here we report a method of determining the bulk domain structure in a cubic magnetostrictive material by combining magneto-optic Kerr microscopy with magnetostriction and magnetization measurements on single crystals as a function of applied field. We analyze the magnetostriction of $\mathrm{F}{\mathrm{e}}_{83}\mathrm{G}{\mathrm{a}}_{17}$ crystals in terms of a domain structure that is greatly influenced by sample shape and heat treatment. Saturation magnetostriction measurements are used to determine the fraction of domains orientated along the three $\langle 100\rangle$ axes in the initial state. Domain wall motion and rotation process have characteristic signatures in the magnetostriction curves, including those associated with the $\mathrm{\Delta }E$ effect and domain rotation through a $\langle 110\rangle$ auxetic direction.

We investigate the electromagnetic response of staggered two-dimensional materials of the graphene family, including silicene, germanene, and stanene, as they are driven through various topological phase transitions using external fields. Utilizing Kubo formalism, we compute their optical conductivity tensor taking into account the frequency and wave vector of the electromagnetic excitations, and study its behavior over the full electronic phase diagram of the materials. In particular, we find that the resonant behavior of the nonlocal Hall conductivity is strongly affected by the various topological phases present in these materials. We also consider the plasmon excitations in the graphene family and find that nonlocality in the optical response can affect the plasmon dispersion spectra of the various phases. We find a regime of wave vectors for which the plasmon relations for phases with trivial topology are essentially indistinguishable, while those for phases with nontrivial topology are distinct and are redshifted as the corresponding Chern number increases. The expressions for the conductivity components are valid for the entire graphene family and can be readily used by others.

Evaluating the total energy of an extended distribution of point charges, which interact through the Coulomb potential, is central to the study of condensed matter. With near ubiquity, the summation required is carried out using Ewald's method, which splits the problem into two separately convergent sums; one in real space and the other in reciprocal space. Density functional-based electronic structure methods require the evaluation of the ion-ion repulsive energy, neutralized by a uniform background charge. Here a purely real-space approach is described. It is straightforward to implement, computationally efficient, and offers linear scaling. When applied to the evaluation of the electrostatic energy of neutral ionic crystals, it is shown to be closely related to Wolf's method.

Aberration-corrected scanning transmission electron microscopy and density-functional theory calculations have been used to investigate the atomic and electronic structure of stair-rod dislocations connected via stacking faults in GaAs nanowires. At the apexes, two distinct dislocation cores consisting of single-column pairs of either gallium or arsenic were identified. Ab initio calculations reveal an overall reduction in the energy gap with the development of two bands of filled and empty localized states at the edges of valence and conduction bands in the Ga core and in the As core, respectively. Our results suggest the behavior of stair-rod dislocations along the nanowire as one-dimensional charge channels, which could host free carriers upon appropriate doping.

We present two complementary approaches to investigate the In content in GaAs/(In,Ga)As/(GaAs) core-shell-(shell) nanowire (NW) heterostructures using synchrotron radiation. The key advantage of our methodology is that NWs are characterized in their as-grown configuration, i.e., perpendicularly standing on a substrate. First, we determine the mean In content of the (In,Ga)As shell by high-resolution x-ray diffraction (XRD) from NW ensembles. In particular, we disentangle the influence of In content and shell thickness on XRD by measuring and analyzing two reflections with diffraction vector parallel and perpendicular to the growth axis, respectively. Second, we study the In distribution within individual NWs by nano x-ray fluorescence. Both the NW (111) basal plane, that is parallel to the surface of the substrate, and the {10-1} sidewall plane were scanned with an incident nanobeam of 50 nm width. We investigate three samples with different nominal In content of the (In,Ga)As shell. In all samples, the average In content of the shell determined by XRD is in good agreement with the nominal value. For a nominal In content of 15%, the In distribution is fairly uniform between all six sidewall facets. In contrast, in NWs with nominally 25% In content, different sidewall facets of the same NW exhibit different In contents. This effect is attributed to shadowing during growth by molecular beam epitaxy. At the same time, along the NW axis the In distribution is still fairly homogeneous. In NWs with 60% nominal In content and no outer GaAs shell, the In content varies significantly both between different sidewall facets and along the NW axis. This fluctuation is explained by the formation of (In,Ga)As mounds that grow simultaneously with a thinner (In,Ga)As shell. The methodology presented here may be applied also to other core-shell NWs with a ternary shell and paves the way to correlating NW structure with functional properties that depend on the as-grown configuration of the NWs.