Due to its characteristics of high electron mobility, moderate band gap, excellent radiation resistance, and high-frequency performance, gallium arsenide (GaAs) is widely used as an important semiconductor material in microelectronics and optoelectronics. Here, we systematically investigated the crystallization mechanism of GaAs by machine learning molecular dynamics simulation. By combining unbiased molecular dynamics simulations with deep neural network potentials, the melting temperature of GaAs crystals was predicted, revealing the crystallization mechanism and phase competition rules. The results show that under undercooling conditions, the cubic zinc blende (ZB) phase has a slower crystallization rate than hexagonal wurtzite (WZ), and its excellent thermodynamic stability makes it the dominant phase in the crystallization process. These factors ultimately determine that a GaAs melt forms a mixed crystal mainly composed of the ZB phase. Ga and As atoms cooperatively complete the lattice arrangement at the solid-liquid interface, and the short-range order is the core driving force for rapid crystal growth. Thus, the microscopic mechanism of the solid-liquid phase transition of GaAs has been probed by applying atomic-scale simulation methods, providing theoretical support for the design, preparation, and performance control of semiconductor materials.

Due to the hexagonal structure, thermal stability, and wide bandgap, two-dimensional group-III nitrides (h-BN, hAlN, h-GaN and h-InN) show great promise for electronic and optoelectronic applications. Density functional theory (DFT) and classical molecular dynamics (MD) methods have advantages in calculation accuracy and scale respectively, but they are limited in the application of high-precision large-scale structure and performance research. Herein, we employ deep potential (DP) method to construct a high-precision machine learning potential (MLP) and systematically investigate the lattice dynamics, thermodynamic, mechanical, and thermal transport properties of two-dimensional Group III nitrides. The DP method can achieve DFT accuracy in energy and atomic force predictions and accurately reproduce phonon dispersion and thermodynamic functions (free energy, heat capacity, entropy) across the 0-1200 K temperature range. MD simulations of uniaxial tensiles reveal distinct mechanical behavior differences among materials. h-BN exhibits high strength but brittle fracture characteristics, while h-AlN and h-GaN demonstrate good strength and ductility. h-InN shows relatively weak overall mechanical performance. Non-equilibrium MD simulations on thermal conductivity reveal significant length-dependent effects in h-BN and h-AlN, attributed to longer phonon mean free paths. Enhanced phonon scattering in h-GaN and h-InN results in lower thermal conductivities. These findings demonstrate that the DP method combines DFT accuracy with large-scale simulation capabilities can deepen understanding of structures and properties of two-dimensional Group III nitrides and provide a computational framework and theoretical foundations for material design and device application.

Semiconductor materials are widely used in electronic, optoelectronic, and energy applications. While DFT-based phonon calculations provide highly accurate assessments for dynamical stability of structures, their prohibitive computational cost poses a significant bottleneck for large-scale materials screening. Herein, we develop DynStabNet, an E(3)-equivariant graph neural network (E3GNN) framework that learns dynamical stability from phonon-informed data, enabling rapid prediction without the need for explicit phonon calculations at the inference stage. Rather than serving as a replacement for first-principles phonon calculations, DynStabNet is designed as a fast surrogate model to accelerate the prescreening of candidate structures prior to subsequent MLIP- or DFT-based phonon validation. Specifically, a crystal structure generation model is utilized to produce diverse candidates, and a pretrained machine learning potential is employed to rapidly compute their phonon spectra to construct the training data set. The E(3)-equivariant architecture enables DynStabNet to capture the complex relationships between crystal structures and their dynamical stability. Consequently, unstable configurations can be rapidly eliminated early in the materials design pipeline. The DynStabNet model achieves an accuracy of 97% while reducing the evaluation time per structure from several hours to approximately 1 ms. By dramatically accelerating the overall screening process, this approach provides an efficient framework for large-scale materials screening.

BACKGROUND:Tetracycline (TC) is widely used in medicine and animal husbandry, but its excessive use has caused serious environmental and health concerns. Prolonged exposure can induce immune suppression, genetic mutations, and antibiotic resistance, emphasizing the need for rapid and reliable TC detection. Fluorescent (FL) and electrochemical (EL) sensing are attractive for rapid analysis, while FL-EL dual-mode strategies enable self-validation and improved accuracy. Nevertheless, achieving highly sensitive TC detection using an intrinsically integrated FL-EL dual-mode material remains rare and technically challenging. RESULTS:A rationally designed zinc-based metal-organic framework (Zn-MOF) is reported as a label-free dual-mode sensor for TC detection, synthesized through a rapid, room-temperature, one-pot strategy. The framework incorporates a fluorescent ligand and a redox-active naphthalenediimide unit, enabling concurrent fluorescence enhancement and pronounced electrochemical signal attenuation upon TC binding. The sensor achieves an ultralow detection limit of 0.6 nM, together with high selectivity and excellent tolerance toward coexisting interferents. Quantitative determination of TC in lake water, honey, fish feed and pork samples yields satisfactory recoveries, demonstrating the analytical robustness and reliability of the proposed sensing system in complex real-world matrices. SIGNIFICANCE:This study presents a rare example of an intrinsically integrated, label-free FL-EL dual-mode MOF sensor for antibiotic analysis. The key novelty lies in combining fluorescence and electrochemical signal generation within a single Zn-MOF without auxiliary labels, aptamers, or nanocomposites. Mechanistic elucidation confirms synergistic coordination and noncovalent interactions as the origin of the dual response, offering a sensitive and practical strategy for TC monitoring.

Due to the excellent electronic and optical properties, gallium phosphide (GaP) has extensive application potential in the fields of optoelectronics. Herein, an improved method of crystal structure prediction of GaP via generation adversarial network (GAN) is proposed. Through combining material chemical composition with Gaussian distributions and optimizing potential space, novel GaP structures with varying compositions were successfully generated. The stability of these structures was verified using phonon spectra. According to their electronic structures, five candidate structures with direct band gaps were identified. Further optical property analysis revealed that some of these structures exhibit excellent optical characteristics. This crystal structure and property prediction (CSPP) framework can be applied for intelligent design and high-throughput screening of semiconductor and other functional materials.

High-entropy alloys (HEAs) possess exceptional properties but face challenges in phase prediction due to data imbalance, feature redundancy and poor interpretability. Based on 2403 experimental records, we propose an explainable ensemble learning framework for five-phase classification of HEAs. A three-step feature selection strategy optimizes 26 descriptors to 14 key features. The ensemble fusion strategy effectively integrated the complementary strengths of multiple base learners, resulting in more balanced and robust phase classification. The proposed model achieved balanced overall classification performance, while the IM category remained affected by a precision-recall trade-off. SHAP, PDP/ICE, and LIME analyses consistently identify Λ and mixing entropy as influential descriptors that show strong statistical associations with the predicted phase categories. This work provides a reliable tool for data-driven design of HEAs.

CONTEXT:Lithium-based compounds, such as Li2O, Li2S, Li3N, and LiF, have important applications in the fields of semiconductors, optics and energy. By performing density functional theory calculations, the structures and properties of lithium-based compounds, including the crystalline, electronic, mechanical, and optical characteristics, were systematically investigated. The impact of oxygen doping on the structure and performance was further studied. The results revealed that doping can reduce their band gaps and elastic constants and change the magnetic properties and the optical moments. These insights can provide theoretical guidance for the design and development of novel lithium-based, sulfide, nitride and fluoride compounds. METHODS:The calculations were performed using the generalized gradient approximation (GGA) method with Perdew-Burke-Ernzerhof (PBE) functional. The Heyd-Scuseria-Ernzerhof (HSE06) hybrid functional was further used for electronic structure corrections. All computations were performed using VASP program.

Battery-type electrode materials store charge through chemical reactions, leading to low power densities and rendering it difficult to reach the performance levels associated with capacitors. This study introduces a coordination polymer electrode material for hybrid supercapacitors, specifically, poly[[tri-μ-ethanolato-di-μ3-sulfido-hexakis(μ4-2-sulfidobenzoato)dinickel(II)tetranickel(III)dipotassium(I)[potassium(I)/sodium(I)(0.60/0.40)]] quaterhydrate], {[NiIII4NiII2K2.60Na0.40(C2H5O)3(C7H4O2S)6S2]·0.25H2O}n or Ni-mba-K(Na), where H2mba is 2-mercaptobenzoic acid, (1). As a hybrid capacitor electrode material, Ni-mba-K(Na) exhibits the characteristics of battery-type electrode materials but demonstrates a capacitor-level power density and cycling stability. The microstructure, element composition, phase structure and thermodynamic stability of Ni-mba-K(Na) are characterized through techniques such as crystal structure determination and X-ray photoelectron spectroscopy. Single-crystal X-ray diffraction shows that compound (1) consists of a hexanuclear nickel cluster, two mononuclear K nodes and a mononuclear K0.60/Na0.40 shared node, resulting in the formation of a complex covalent three-dimensional network. Due to this novel structure, Ni-mba-K(Na) exhibits supercapacitor performance that combines high energy density and high power density.

Wide-band-gap semiconductor gallium oxide (Ga2O3) has significant advantages for use in high-power semiconductor devices and deep-ultraviolet-detection optoelectronic devices. Crystal structure prediction is one of the most challenging and interesting issues in condensed matter science. In this work, 11 three-dimensional (3D) Ga2O3 structures and four two-dimensional (2D) Ga2O3 structures were predicted and screened based on a multi-objective differential evolution algorithm combined with density functional theory (DFT) calculations. Two low-energy 3D structures proved to be consistent with previously characterized β-Ga2O3 (C2/m) and α-Ga2O3 (R3̄c). Their stabilities were confirmed by calculation of their structural parameters, phonon spectra, elastic constants, and elastic moduli. Their photoelectric properties were also investigated. The results showed that both the stable 3D and 2D Ga2O3 structures had wide band gaps, and some of them exhibited good optical properties. These findings are of great significance in providing theoretical guidance for the structural design of Ga2O3 materials and their application in microelectronic and photoelectric devices.

Crystal structure prediction (CSP) represents a fundamental research frontier in computational materials science and chemistry, aiming to predict thermodynamically stable periodic structures from given chemical compositions. Traditional methods often face challenges such as high computational costs and local minima trapping. Recently, artificial intelligence methods, represented by generative adversarial networks (GANs), variational autoencoders (VAEs), diffusion models, and large language models (LLMs), have revolutionized the traditional prediction paradigm. These computational frameworks efficiently extract chemical rules and structural features from crystal databases, significantly reducing computational costs while maintaining prediction accuracy. This Perspective systematically evaluates the advantages and limitations of various generative models, explores their synergies with conventional approaches, and discusses their future prospects in accelerating materials discovery and development, providing new insights for future research directions.

Due to the outstanding thermal stability, inherent high melting points, and elevated temperature strengths, refractory high-entropy alloys (RHEAs) have been widely used for extreme environments in aerospace, nuclear energy, and advanced propulsion systems. Herein, we present an integrated design and simulation framework for RHEAs, combining machine learning potentials, supervised regression models, and multiobjective optimization algorithms. Utilizing a universal neuroevolution potential version 1 (UNEP-v1), the framework significantly enhances the accuracy of atomic-scale simulation while substantially reducing computational cost. High-throughput molecular dynamics simulations generate melting points and ultimate tensile strengths at 1000 K for various alloy compositions. Supervised regression models enable a rapid performance prediction. Integrating Shapley Additive exPlanations, Partial Dependence Plots, Accumulated Local Effects, and Individual Conditional Expectation analysis can provide a comprehensive interpretability toolkit. Validation of the proposed method in the TiVCrZrMo alloy system demonstrates its efficacy in designing high-strength, high-temperature resistant alloys. We not only develop a precise and interpretable predictive modeling paradigm but also establish procedural frameworks, promoting the integration of atomic-scale simulations with data-driven approaches for RHEAs in extreme environments.

Perovskite solar cells (PSCs) have attracted wide attention for their tunable bandgap and excellent photoelectric conversion efficiency. Herein, the structures of two-dimensional molybdenum disulfide (MoS2) were searched by the inverse design of materials by multi-objective differential evolution method. The structure, electronic, optical and transport properties of the MoS2/CsPbBr3 heterostructure were investigated. The results show that the interfacial interaction has a significant effect on the structural stability, electronic and optical properties of the perovskite heterostructure. In the visible light range, the MoS2/CsPbBr3 type-II heterostructure exhibits optical properties superior to those of monolayer materials. The MoS2/perovskite interface has a strong interaction and can improve the stability and photoelectric efficiency of PSCs. These can provide theoretical support for the design and application of perovskite solar cells.

Wide-band-gap semiconductor gallium oxide (Ga2O3) has significant advantages for use in high-power semiconductor devices and deep-ultraviolet-detection optoelectronic devices. Crystal structure prediction is one of the most challenging and interesting issues in condensed matter science. In this work, 11 three-dimensional (3D) Ga2O3 structures and four two-dimensional (2D) Ga2O3 structures were predicted and screened based on a multi-objective differential evolution algorithm combined with density functional theory (DFT) calculations. Two low-energy 3D structures proved to be consistent with previously characterized β-Ga2O3 (C2/m) and α-Ga2O3 (R3̄c). Their stabilities were confirmed by calculation of their structural parameters, phonon spectra, elastic constants, and elastic moduli. Their photoelectric properties were also investigated. The results showed that both the stable 3D and 2D Ga2O3 structures had wide band gaps, and some of them exhibited good optical properties. These findings are of great significance in providing theoretical guidance for the structural design of Ga2O3 materials and their application in microelectronic and photoelectric devices.

Crystal structure prediction (CSP) is an important field of material design. Herein, we propose a novel generative adversarial network model, guided by a data-driven approach and incorporating the real physical structure of crystals, to address the complexity of high-dimensional data and improve prediction accuracy in materials science. The model, termed GAN-DDLSF, introduces a novel sampling method called data-driven latent space fusion (DDLSF), which aims to optimize the latent space of generative adversarial networks (GANs) by combining the statistical properties of real data with a standard Gaussian distribution, effectively mitigating the "mode collapse" problem prevalent in GANs. Our approach introduces a more refined generation mechanism specifically for binary crystal structures such as gallium nitride (GaN). By optimizing for the specific crystallographic features of GaN while maintaining structural rationality, we achieve higher precision and efficiency in predicting and designing structures for this particular material system. The model generates 9321 GaN binary crystal structures, with 16.59% reaching a stable state and 24.21% found to be metastable. These results can significantly enhance the accuracy of crystal structure predictions and provide valuable insights into the potential of the GAN-DDLSF approach for the discovery and design of binary, ternary, and multinary materials, offering new perspectives and methods for materials science research and applications.

We successfully synthesized a 3D porous lanthanide metal-organic framework, [Tb (obdb)][Tb1/3(H2O)7/3]NMP (Tb-MOF) with hydrated terbium ions in the pores by solvothermal method using 4 ',4 '''-oxybis[1,1 '-biphenyl]-3,5-dicarboxylic acid (H4obdb) as ligand. It is worth mentioning that Tb-MOF exhibits excellent performance in structural stability and fluorescence response. Even after soaking in various solvents and pH solutions for 3 months, Tb-MOF can maintain its structural integrity and excellent stability. The fluorescence spectra show that Tb-MOF not only exhibits the luminescence of the ligand centered at 368 nm (IL), but also emits the characteristic emissions of the metal terbium, among which the luminescence at 542 nm is the strongest. In the detection of antibiotic amphotericin B (AB), Tb-MOF showed significant detection efficiency. Upon increasing the concentration of AB, the fluorescence intensity at 542 nm gradually decreased, whereas the fluorescence at 368 nm decreased rapidly. The I542/IL value showed a very high linear correlation with AB concentration, with the slope reaching 1.29 x 108 M-1 (0-100 mu M). In the field of fluorescent sensors for AB detection, Tb-MOF is notable not only for its impressively low LOD (limt of detection) of 0.072 nM but also for pioneering the application of a self-calibrating ratio-based detection method. To better elucidate this fluorescence sensing phenomenon, we also conducted a detailed discussion of the sensing mechanism for AB.

Three isostructural transition metal-organic frameworks, [M(bta)0.5(bpt)(H2O)2]·2H2O (M = Co (1), Ni (2), Zn (3), H4bta = 1,2,4,5-benzenetetracarboxylic acid, bpt = 4-amino-3,5-bis(4-pyridyl)-1,2,4-triazole), were successfully constructed using different metal cations. These frameworks exhibit a three-dimensional network structure with multiple coordinated and lattice water molecules within the framework, contributing to high stability and a rich hydrogen-bond network. Proton conduction studies revealed that, at 333 K and 98% relative humidity, the proton conductivities (σ) of MOFs 1-3 reached 1.42 × 10-2, 1.02 × 10-2, and 6.82 × 10-3 S cm-1, respectively. Compared to the proton conductivity of the initial ligands, the σ values of the complexes increased by 2 orders of magnitude, with the activation energies decreasing from 0.36 to 0.18 eV for 1, 0.09 eV for 2, and 0.12 eV for 3. An in-depth analysis of the correlation between different metal centers and proton conduction performance indicated that the varying coordination abilities of the metal cations and the water absorption capacities of the frameworks might account for the differences in conductivity. Additionally, the potential of 1 as a supercapacitor electrode material was assessed. 1 exhibited a specific capacitance of 61.13 F g-1 at a current density of 0.5 A g-1, with a capacitance retention of 82.4% after 5000 cycles, making it a promising candidate for energy storage applications.

Antibiotic pollution in hydrophytic ecosystems has aroused widespread notice due to its adverse impact on both the ecological environment and human physical well-being. Thus, the development of effective antibiotic detection materials is not only urgently needed but is also challenging. Herein, the luminescent metal-organic framework [Mn3/2(HBCTC)5H2O] (1) was successfully synthesized based on [9,9 '-bicarbazole]-3,3 ',6,6 '-tetracarboxylic acid (H4BCTC) under solvothermal conditions. Compound 1 can be employed as a fluorescent probe to detect antibiotic difloxacin hydrochloride (DIF) in water, demonstrating a remarkable linear relationship between fluorescence intensity ratio I0/I and the concentration of DIF (0-40 mu M), with Ksv of 79,405 M-1 and LOD of 114 nM. More importantly, 1 represents the first fluorescent sensing material based on MOF for the detection of DIF, exhibiting exceptional selectivity, sensitivity, and anti-interference capability. Furthermore, the mechanism of fluorescence quenching was studied comprehensively.

A 1D coordination polymer (CP) [Cd2L2(H2O)(4)] & sdot; 3H(2)O (1) was prepared by solvothermal method using 1-(4-carboxyphenyl)-1H-pyrazole-3-carboxylic acid (H2L) as single ligand. Fluorescence sensing experiments show that 1 has dual-function specific recognition ability for moxifloxacin (MXF) and silicate (SiO32-), which can be used as the turn-off sensing material for their detection. When 1 specifically identifies MXF, the Ksv is as high as 6.46x10(5) M-1 (6-20 mu M) and the limit of detection (LOD) is as low as 14 nM. 1 is the second example of a CP-based fluorescent probe for the detection of MXF, which has confirmed that the Ksv is the largest to date for the detection of MXF, with a detection sensitivity increase of more than three times. For SiO32- ions detection, the fluorescence intensity ratio has a good linear correlation with the concentration of SiO32- ions in the concentration range of 0-100 mu M, with a slope of 1.33x10(4) M-1, and the LOD is as low as 0.68 mu M. According to the reported literature, 1 is the only example of SiO32- ions sensing by CP-based fluorescence sensor so far. In order to better understand the sensing phenomenon, we also discussed the sensing mechanism for MXF and SiO32- ions.

As a nanofabrication technology, atomic layer deposition (ALD) has been widely used in the fields of displays, microelectronics, nanotechnology, catalysis, energy and coatings. It demonstrates excellent conformality, large- area uniformity and precise control of the sub-monolayer film. Al2O3 2 O 3 ALD using trimethylaluminum (TMA) and water (H2O) 2 O) as precursors is the most ideal ALD model system. In this work, the reactions of TMA and H2O 2 O with the surface have been investigated using density functional theory (DFT) calculations in order to obtain more information on the reaction mechanism of the complicated H2O-based 2 O-based ALD of Al2O3. 2 O 3 . In the TMA reaction, the methyl ligands can be eliminated and new Al-O bonds can be formed via ligand exchange reactions. In the H2O 2 O reaction, the methyl ligand on the surface can be further eliminated and new Al-O - O bonds can be formed. Meanwhile, the coupling reactions between the surface methyl and hydroxyl groups can further form new Al-O - O bonds and release CH4 4 or H2O 2 O to densify the Al2O3 2 O 3 film. These complicated reaction mechanisms of Al2O3 2 O 3 H2O- 2 O- based ALD can provide theoretical guidance for the precursor design and ALD growth of other oxides and aluminum-based compounds.

As the unique nanofabrication technique, atomic layer deposition (ALD) can be used to deposit high-quality, uniform, and conformal nanoparticle coatings. Ru and its oxides (RuOx) are important materials in the fields of microelectronics and catalysis. In this work, the reaction of Ru precursors with different ligands on the hydroxylated surfaces were explored. The elimination of the pyrrolyl (Py) ligand is easier than that of the cyclopentadienyl (Py) ligand on the hydroxylated surface. The reaction involved in RuO2·xH2O ALD using RuPy2 and H2O as precursors was further investigated. During the RuPy2 reaction, the pyrrolyl ligand is eliminated by the substitution reaction with a surface hydroxyl group. However, the desorption of the pyrrole molecule is difficult due to the strong adsorption between the pyrrole molecule and Ru surface. During the H2O reaction, the H2O molecules further help the elimination of pyrrolyl ligands from the surface and the dissociative desorption of pyrrole molecules This unique reaction mechanism involved in the ALD of RuO2 with the assistance of H2O can be termed new cascade mechanism. These findings provide theoretical guidance for precursor design and ALD growth of metal oxides.