Physics

Electric-field-induced phase switching in vanadium dioxide (VO 2 ) is central to its integration into adaptive electronic systems. However, the role of oxygen defects in governing structural stability and switching behavior remains insufficiently understood. Here, we engineer oxygen-deficient VO 2 (VO 2-δ ) films to investigate the structure of VO 2-δ and the electrical switching behavior systematically. The preferential occupation of oxygen vacancies was identified, which induces a 3-fold superstructure of (011) R and stacking faults with a certain degree of long-range ordering along (110) R and (020) R . Even so, the VO 2-δ maintains the rutile-like coordination structure as VO 2 . Notably, the VO 2-δ films exhibit high conductivity at room temperature. An irreversible metal–insulator transition can be triggered at 20 V for the VO 2-δ film, enabled by localized Joule-heating-assisted reorganization of the oxygen-vacancy defect structure. The resulting abrupt resistance jump highlights the potential of VO 2-δ as an ultrathin active medium for fast electronic overload protection. This work establishes vacancy ordering as a powerful lever for tailoring phase behavior in VO 2, offering a defect-engineering pathway toward reconfigurable oxide electronic devices.

Intramolecular noncovalent interactions (Intra-NIs) have emerged as key determinants of structure–property relationships in molecular systems, exerting profound influences on material performances of organic optoelectronic materials. In the development of organic light-emitting diode (OLED) materials, Intra-NIs have recently been recognized as an effective strategy for enhancing bond dissociation energies (BDEs), thereby improving molecular stability and device lifetimes. Accurate and rapid prediction of BDEs of molecules featuring pronounced Intra-NIs is therefore essential for accelerating the development of robust OLED materials. However, existing machine learning approaches for predicting BDEs primarily focus on short-range local properties, while failing to adequately capture long-range Intra-NI effects. Herein, we report a dual-attention graph neural network model for the accurate and rapid prediction of BDEs affected by Intra-NIs. By integrating through-bond and through-space attention mechanisms and incorporating fresh molecular dispersion matrix as an additional input to encode geometric information, the model achieves high accuracy for C–N BDEs, with a mean absolute error below 0.05 eV, while reducing computational cost by 4 orders of magnitude relative to density functional theory calculations. Furthermore, the model exhibits strong transferability to C–P and C–S bonds via transfer learning, demonstrating its potential applicability to other fragile bonds in diverse organic optoelectronic materials. This work establishes a robust paradigm for incorporating physically meaningful Intra-NI information into data-driven molecular property predictions, providing a powerful tool for the iterative optimization and high-throughput virtual screening of next-generation high-performance optoelectronic materials across a wide range of applications.

We report a deuterium kinetic isotope effect (KIE) in the mechanically driven formation and polymorphic transformation of a hydrogen-bonded organic salt, monitored in situ by synchrotron powder X-ray diffraction and Raman spectroscopy. Surprisingly, no measurable KIE is observed in the salt formation step (which necessarily involves H transfer), whereas a pronounced KIE emerges in the subsequent polymorphic transformation (involving only supramolecular rearrangement). Deuteration delays the onset of the polymorphic transformation and slows the transformation rate, with the fully deuterated system showing an approximately 3-fold decrease in rate and partially deuterated mixtures exhibiting intermediate kinetics. An impact-based kinetic model captures the isotope-dependent trends, while solid-state DFT including vibrational contributions does not indicate an isotope-dependent change in relative phase stability, supporting a primarily kinetic origin of the KIE. These results identify partial deuteration as a practical handle for tuning the solid-state reactivity.

The performance of photoelectrodes for water splitting depends on multiple factors, including their stability against photocorrosion. Protective layers are widely used to prevent corrosion, yet the design and optimization of the interface between the photoelectrode and their protective layer remain poorly understood. Here we focus on bismuth vanadate (BiVO 4 ), a promising photoanode material, and we investigate its interface with titania (TiO 2 ) protective layers, specifically the role of oxygen vacancies in controlling charge transfer and band alignment. Using first-principles calculations, we show that the role of oxygen vacancies in hole transfer depends on the thickness of the TiO 2 layer: the absence of vacancies is more favorable for hole transfer in thin TiO 2 layers, whereas the presence of vacancies is more favorable for thick TiO 2 layers. Hence our results provide guidelines for defect engineering depending on the thickness of the protection layer. In addition, our findings demonstrate that the impact of interfacial oxygen vacancies cannot be solely inferred from bulk defect properties, highlighting the crucial role of atomistic interface modeling in the design of heterojunction photoelectrodes.

Phase Change Memory is a leading emerging nonvolatile memory technology based on the reversible switching between amorphous and crystalline states in Ge‐Sb‐Te alloys. Repeated phase transitions induce elemental redistribution, generating local stoichiometry variations that critically influence device performance. Accessing different compositions within a single specimen is therefore essential for understanding phase change memory (PCM) behavior. We present a deposition methodology enabling the simultaneous formation of phase‐change alloys with nanoscale compositional gradients. The approach uses three Knudsen cells supplying Ge, Sb, and Te, each oriented at 45° relative to the substrate normal. The substrate includes cylindrical nanopillar arrays produced by electron beam lithography, which generate controlled shadowing effects and spatial modulation of the elemental flux. Local composition was characterized using scanning transmission electron microscopy and electron energy loss spectroscopy. The elemental profiles were compared with a ballistic transport model describing trajectories around nanopillars. The experimental and simulated data show agreement. Simulations performed with an alternative configuration, where the three cells are spaced 120°, predict predefined Ge, Sb, and Te‐deficient regions around the pillars. This methodology enables controlled nanoscale stoichiometry engineering and provides a powerful platform for studying composition‐dependent phase‐change mechanisms, supporting the optimization of PCM device performance.

Atomic layer deposition (ALD) shows promise for purposefully modifying the surface chemistry of powders because it has excellent control over the thickness of the applied coatings. ALD can precisely deposit coatings with a resolution approaching 0.1 nm, providing intricate control over surface properties. However, few studies have investigated the potential to systematically vary the surface potential of powders using this precision in ALD’s deposition process. In this work, titanium oxide thin films are deposited onto silicon oxide nanoparticles via ALD to create silica–titania core-shell nanoparticles. The change in the isoelectric points of the particles is investigated as a function of the thickness of the ALD deposited TiO2 layer. For films deposited at 1 nm thickness and below, the isoelectric point changes approximately linearly with coating thickness. For films thicker than 1 nm, the isoelectric point no longer changes, indicating that the surface potential is no longer influenced by the underlying silicon oxide core. We investigate the conformality of the coating with electron microscopy, x-ray photoelectron spectroscopy, and energy-dispersive x-ray (EDX) analysis. Additionally, we introduce a model to estimate layer thickness from the EDX measurements based on simple geometric and stoichiometry assumptions.

We demonstrate experimentally how the nucleation of skyrmions in an <a:math xmlns:a="http://www.w3.org/1998/Math/MathML"> <a:mi>Ir</a:mi> <a:mo>,</a:mo> <a:mo> </a:mo> <a:mi>Co</a:mi> </a:math> , and <b:math xmlns:b="http://www.w3.org/1998/Math/MathML"> <b:mi>Pt</b:mi> </b:math> based magnetic multilayer is affected by introducing a layer dependent sign for the Dzyaloshinskii-Moriya interaction (DMI). In one stack, the bottom half of the stack is given a positive DMI and the top half a negative DMI, and as a result, the in-plane component of the dipolar field is aligned parallel to the effective field of the DMI in every layer, enhancing the effective DMI. We show that this enhanced DMI facilitates the nucleation and stability of skyrmions using both current-driven and laser-induced skyrmion nucleation. In the devices with an enhanced effective DMI, the density of nucleated skyrmions is greater by up to a factor <c:math xmlns:c="http://www.w3.org/1998/Math/MathML"> <c:mrow> <c:mo>∼</c:mo> <c:mn>20</c:mn> </c:mrow> </c:math> and skyrmions can be observed in stronger magnetic fields—suggesting that their stability is also improved. These results show that skyrmion nucleation depends strongly on the magnitude of the effective DMI in a magnetic multilayer and that the dipolar field within such a multilayer presents an effective route towards controlling the effective DMI, and thereby, the nucleation of chiral magnetic textures.

Superionic materials have attracted considerable interest due to their unique electronic and thermal transporting properties, with far-reaching implications for solid-state electrolyte batteries and planetary science. Despite extensive studies, the physicochemical origin of superionic behavior remains insufficiently understood. In this work, we develop a generalized machine-learning descriptor for superionic materials by combining structural, elemental, and density-functional theory primary features with the Sure Independence Screening and Sparsifying Operator (sisso) framework. We identify four key factors to recognize superionic materials: electronegativity, chemical hardness, density, and bond strength. The developed machine-learning descriptor applies well beyond conventional lithium- and sodium-based superionic systems and can be extended to high-pressure conditions. Based on our machine-learning descriptor, we conducted a high-throughput screening of 13 019 materials from the materials project database, and 353 potential superionic candidates were identified. This study not only sheds light on the underlying physical origin of the superionic state but also provides guidance and insights for the design and experimental synthesis of future superionic materials.

The recently investigated nitrogen (N) incorporation into TiNbMoTaW high entropy alloy films leads to a domelike superconducting transition temperature <a:math xmlns:a="http://www.w3.org/1998/Math/MathML"> <a:msub> <a:mi>T</a:mi> <a:mi>C</a:mi> </a:msub> </a:math> vs nitrogen concentration <b:math xmlns:b="http://www.w3.org/1998/Math/MathML"> <b:mi>x</b:mi> </b:math> (with <c:math xmlns:c="http://www.w3.org/1998/Math/MathML"> <c:mrow> <c:mi>x</c:mi> <c:mo>=</c:mo> <c:mtext>N</c:mtext> <c:mo>/</c:mo> <c:mi>M</c:mi> </c:mrow> </c:math> representing the ratio between atomic fractions of nitrogen and metal atoms, <d:math xmlns:d="http://www.w3.org/1998/Math/MathML"> <d:mrow> <d:mi>M</d:mi> <d:mo>=</d:mo> <d:mi>Ti</d:mi> <d:mo>+</d:mo> <d:mi>Nb</d:mi> <d:mo>+</d:mo> <d:mi>Mo</d:mi> <d:mo>+</d:mo> <d:mi>Ta</d:mi> <d:mo>+</d:mo> <d:mi mathvariant="normal">W</d:mi> </d:mrow> </d:math> ) in the respective high entropy nitride <f:math xmlns:f="http://www.w3.org/1998/Math/MathML"> <f:mrow> <f:msub> <f:mrow> <f:mo>(</f:mo> <f:mi>TiNbMoTaW</f:mi> <f:mo>)</f:mo> </f:mrow> <f:mrow> <f:mn>1.0</f:mn> </f:mrow> </f:msub> <f:msub> <f:mi mathvariant="normal">N</f:mi> <f:mi>x</f:mi> </f:msub> </f:mrow> </f:math> , with a large <h:math xmlns:h="http://www.w3.org/1998/Math/MathML"> <h:msub> <h:mi>T</h:mi> <h:mi>C</h:mi> </h:msub> </h:math> enhancement. In this study, we report on the effect of N insertion into NbTaHfTiZr films, which show a lower valence electron concentration and a lower configuration entropy than those of TiNbMoTaW. The <i:math xmlns:i="http://www.w3.org/1998/Math/MathML"> <i:msub> <i:mi>T</i:mi> <i:mi>C</i:mi> </i:msub> </i:math> vs <j:math xmlns:j="http://www.w3.org/1998/Math/MathML"> <j:mi>x</j:mi> </j:math> dependence in <k:math xmlns:k="http://www.w3.org/1998/Math/MathML"> <k:mrow> <k:msub> <k:mrow> <k:mo>(</k:mo> <k:mi>NbTaHfTiZr</k:mi> <k:mo>)</k:mo> </k:mrow> <k:mrow> <k:mn>1.0</k:mn> </k:mrow> </k:msub> <k:msub> <k:mi mathvariant="normal">N</k:mi> <k:mi>x</k:mi> </k:msub> </k:mrow> </k:math> exhibits a different course, in which <m:math xmlns:m="http://www.w3.org/1998/Math/MathML"> <m:msub> <m:mi>T</m:mi> <m:mi>C</m:mi> </m:msub> </m:math> first decreases from <n:math xmlns:n="http://www.w3.org/1998/Math/MathML"> <n:mrow> <n:mo>∼</n:mo> <n:mn>6.5</n:mn> <n:mspace width="0.28em"/> <n:mi mathvariant="normal">K</n:mi> </n:mrow> </n:math> at <q:math xmlns:q="http://www.w3.org/1998/Math/MathML"> <q:mrow> <q:mi>x</q:mi> <q:mo>=</q:mo> <q:mn>0</q:mn> </q:mrow> </q:math> to <r:math xmlns:r="http://www.w3.org/1998/Math/MathML"> <r:mrow> <r:msub> <r:mi>T</r:mi> <r:mi>C</r:mi> </r:msub> <r:mo>≈</r:mo> <r:mn>1</r:mn> <r:mspace width="0.28em"/> <r:mi mathvariant="normal">K</r:mi> </r:mrow> </r:math> at <u:math xmlns:u="http://www.w3.org/1998/Math/MathML"> <u:mrow> <u:mi>x</u:mi> <u:mo>=</u:mo> <u:mn>0.25</u:mn> </u:mrow> </u:math> and <v:math xmlns:v="http://www.w3.org/1998/Math/MathML"> <v:mrow> <v:mi>x</v:mi> <v:mo>=</v:mo> <v:mn>0.42</v:mn> </v:mrow> </v:math> , and then starts to increase and reaches a maximum of <w:math xmlns:w="http://www.w3.org/1998/Math/MathML"> <w:mrow> <w:mo>∼</w:mo> <w:mn>8.6</w:mn> <w:mspace width="0.28em"/> <w:mi mathvariant="normal">K</w:mi> </w:mrow> </w:math> at <z:math xmlns:z="http://www.w3.org/1998/Math/MathML"> <z:mrow> <z:mi>x</z:mi> <z:mo>=</z:mo> <z:mn>0.81</z:mn> </z:mrow> </z:math> followed by a decline to <ab:math xmlns:ab="http://www.w3.org/1998/Math/MathML"> <ab:mrow> <ab:mo>∼</ab:mo> <ab:mn>6.7</ab:mn> <ab:mspace width="0.28em"/> <ab:mi mathvariant="normal">K</ab:mi> </ab:mrow> </ab:math> when <db:math xmlns:db="http://www.w3.org/1998/Math/MathML"> <db:mi>x</db:mi> </db:math> approaches the stoichiometric value 1. In this case, the transition of the initial high entropy alloy bcc structure to the fcc structure of high entropy nitrides proceeds slowly and leads to a wide <eb:math xmlns:eb="http://www.w3.org/1998/Math/MathML"> <eb:mi>x</eb:mi> </eb:math> range in which <fb:math xmlns:fb="http://www.w3.org/1998/Math/MathML"> <fb:msub> <fb:mi>T</fb:mi> <fb:mi>C</fb:mi> </fb:msub> </fb:math> is suppressed. The analysis shows that the <gb:math xmlns:gb="http://www.w3.org/1998/Math/MathML"> <gb:msub> <gb:mi>T</gb:mi> <gb:mi>C</gb:mi> </gb:msub> </gb:math> vs <hb:math xmlns:hb="http://www.w3.org/1998/Math/MathML"> <hb:mi>x</hb:mi> </hb:math> course and the <ib:math xmlns:ib="http://www.w3.org/1998/Math/MathML"> <ib:msub> <ib:mi>T</ib:mi> <ib:mi>C</ib:mi> </ib:msub> </ib:math> enhancement in high entropy nitrides depend on the valence electron concentration of the initial high entropy alloy, as higher values accelerate the tendency to form a fcc structure and are less sensitive to the localization of otherwise mobile electrons in metallic high entropy alloys due to their bonding to N atoms. However, other parameters, such as the atomic size difference of the metallic constituents and their affinity towards nitrogen, also have to be considered.

Our understanding of the origin of heavy elements beyond iron relies on the rapid neutron capture process (r-process), which accounts for roughly half of their cosmic abundance. However, the extreme neutron-rich conditions required for the r-process involve many nuclei that remain experimentally inaccessible, making theoretical predictions essential. We explore the impact of nuclear masses calculated with the ab initio valence-space in-medium similarity renormalization group, focusing on the region around the N = 82 shell closure. We show for the first time that such ab initio mass calculations can be used to refine r-process predictions compared to global, but more phenomenological mass models. With the ab initio masses, the waiting point of the second r-process peak is strengthened, which leads to an overall slower nucleosynthesis flow, lower abundances of nuclei beyond the peak, and a stronger shift of the third r-process peak.