With the advancement of the Materials Genome Initiative and the establishment and improvement of the glass database, data-driven machine-learning methods penetrate the field of glass materials, and remarkable progress is made in the modeling of the physical, mechanical, and optical properties of glass materials. In recent years, oxide glasses have found increasingly widespread applications in the field of electronic applications, and their electrical properties as critical indicators have attracted much attention. However, modeling on the electrical properties of oxide glasses remains a challenge, and insufficient data suppress the model's performance. To address the issue of limited data availability on the electrical properties of oxide glasses, this work was to propose a data augmentation framework named WGAN-GP-CP tailored for glass materials.

Based on the Wasserstein generative adversarial network with gradient penalty (WGAN-GP) model, an additional composition penalty term (CP) was added to the generator's loss function, and a method for evaluating sample quality was proposed, which could assess the quality of the generated samples from three aspects, i.e., diversity, accuracy, and uniqueness. Moreover, the XGBoost algorithm was employed for model training, and the model was interpreted through SHAP single-feature analysis and feature interaction analysis.

The trained generators can generate high-quality synthetic samples, expanding the existing datasets effectively. The results of comparative studies demonstrate that the data-augmented models exhibit mitigated overfitting and significantly enhanced generalization capability. The generalization performances of the room-temperature bulk resistivity, relative dielectric constant, and dielectric loss models increase from 0.784 to 0.838, 0.862 to 0.897, and 0.801 to 0.861, respectively. The results obtained from SHAP analysis are generally consistent with the classical physicochemical understanding of the electrical properties of oxide glasses. In addition, some interesting mixing effects and interactive relationships are also revealed.

The data augmentation framework proposed in this work offered a strategy for developing high-performance data-driven models under data scarcity conditions. The insights from SHAP single-feature analysis and feature interaction analysis could provide an important guidance for a deeper understanding of the relationship between the composition and electrical properties of oxide glasses. This research shifted the optimization of oxide glass electrical properties from empirically driven trial-and-error methods to data-driven scientific design frameworks, laying a foundation for the development of high-performance glass materials in the field of electronic applications.

As phase-engineered composites integrating amorphous and crystalline domains, glass-ceramics synergistically retain a high transparency of vitreous materials, while achieving ceramic-like mechanical robustness. Transparent glass-ceramics include systems such as ZnO-MgO-Al₂O₃-SiO₂, Li₂O-Al₂O₃-SiO₂, etc., in which the precipitation of spinel crystalline phase can be controlled in ZnO-MgO-Al₂O₃-SiO₂ (ZMAS) system. The spinel crystal phase has a high Mohs hardness as well as a high modulus of elasticity, so the ZMAS system glass-ceramics with a spinel crystal phase has superior mechanical properties. The introduction of Li₂O into ZMAS glass-ceramics can refine the grain size and also promote the precipitation of crystal phases such as β-quartz solid solution with a low coefficient of thermal expansion, which can adjust the overall coefficient of thermal expansion of the glass-ceramics and improve the thermal shock resistance of the glass-ceramics. β-quartz solid solution crystals are close to the refractive index of the glass, which can significantly reduce the scattering of light and improve the transparency of the glass-ceramics. In this work, Li₂O-ZnO-MgO-Al₂O₃-SiO₂ glass system was selected as a research subject to investigate the effect of Li₂O content on the crystallization behavior and properties of the glass. The study revealed that the introduction of Li₂O exerted a modulating effect on the crystallization behavior, optical properties, and mechanical performance of the glass-ceramics.

The parent glasses were prepared by a melting method, and the components were designed as (54-x) SiO₂-17 Al₂O₃-x Li₂O-3 Na₂O-14 ZnO-7 MgO-3 ZrO₂-2 SnO₂ (%, x=0, 2, 4, 6). The glass was melted at 1600 ℃ for 3 h. The glass solution was poured into a copper abrasive tool that was preheated at 590 ℃ for moulding. The formed glass blocks were annealed in an annealing furnace at 590 ℃ for 5 h to remove the internal stresses in the glass, and the glass was cut into blocks with the thickness of 1.5 mm for subsequent testing. The chemical state of the elements in the base glass was characterized by a model ESCALAB250Xi X-ray photoelectron spectrometer (XPS, Thermo Fisher Co., USA). A glass powder was used for testing, corrected using the electron binding energy of C1s. The DSC curve of glass was determined by differential scanning calorimetry (DSC).. The crystalline phase of the prepared samples was examined by an X-ray diffractometer. The base glass and microcrystalline glass microstructures were analyzed by a model Ultra Plus field emission scanning electron microscope (FESEM, Zeiss Co., Germany) with an additional Ultim Max 65 energy spectrometer (JSM-IT800). The optical properties of the glass and microcrystalline glass samples after double-sided optical grade polishing were characterized by a model UV3600 UV-visible spectrophotometer (UV-vis, Shimadzu Co., Japan). The Vickers hardness of the polished glass samples was measured by a model HV-1000A microhardness tester (Huayin Co., China) with a load of 1.96 N and a dwell time of 10 s. Each glass specimen underwent five replicate measurements to minimize experimental errors. The coefficient of thermal expansion (CTE) was determined by a model DIL 402C thermal dilatometer (Netzsch Co., Germany). The glass was cut into some rectangular bars with the sizes of 4 mm × 4 mm × 25 mm for dilatometric analysis.

The free oxygen content in the glass network increases with the increase of Li₂O content from 0 to 6% (in mole), leading to a reduction in the degree of network polymerization. This structural change can reduce the glass transition temperature from 745 ℃ (i.e., at 0 mol% Li₂O) to 664 ℃ (i.e., at 6% Li₂O). The β-quartz solid solutions form as the Li₂O content increases to 2%. The temperature required for the formation of β-spodumene crystals decreases as the Li₂O content further increases, leading to an increase in the β-spodumene crystalline phase content and suppression of spinel crystal precipitation. The transmittance also undergoes a sharp decline due to the influence of large-sized β-spodumene crystals. The glass-ceramics with 2% Li₂O content, after crystallization at 850 ℃, exhibit a superior performance (i.e., a Vickers hardness of 8.62 GPa, a transmittance of 84%, and a flexural strength of 199 MPa).

Increasing Li₂O content from 0 to 6% could promote the depolymerization of the glass network, leading to a reduction in the glass transition temperature. At 2% Li₂O, this composition effectively suppressed grain agglomeration and promoted the formation of β-quartz solid solution crystalline phases, thereby enhancing the transparency of the glass-ceramics. The coexistence of spinel phases and β-quartz solid solution improved the comprehensive properties of the glass-ceramics (i.e., a Vickers hardness of~8.62 GPa and a flexural strength of~199 MPa).

Nd³⁺-doped phosphate glasses, utilizing the ⁴F_3/2 → ⁴I_11/2 energy-level transition of Nd³⁺ ions, achieve 1.06 μm luminescence and have critical applications in optical communications, biomedical engineering, and defense technologies. However, the development of laser glasses depends on empirical trial-and-error approaches, requiring extensive experimentation to generate limited data. This paradigm suffers from inefficiency, prolonged cycles, high costs, and a lack of theoretical guidance. Moreover, the luminescent properties of rare-earth-doped glasses are closely tied to the local structure of the rare-earth ions. The intricate relationship between structure and performance renders most computational models for glass properties ineffective, severely hindering the rapid advancement of laser glasses. It is thus necessary to develop cross-scale computational methods that combine physical interpretability and extrapolation capabilities to enable accurate prediction and inverse design of luminescent properties for broad-composition Nd³⁺-doped phosphate glasses.

In this study, Nd³⁺-doped Li₂O-MgO-Al₂O₃-P₂O₅ glasses were prepared. The local microstructure of Nd³⁺ in the glass as a statistical ensemble of its local environments in neighboring glassy compounds (NGCs) was described. Based on this theoretical framework, an NGCs model was proposed to quantitatively predict the luminescent properties of Nd³⁺-doped phosphate laser glasses, positing that the glass's luminescent properties could be equal to the statistical average of the corresponding properties of its NGCs. This work could provide a theoretical foundation for quantitatively calculating the local structure and luminescent properties of rare-earth-doped laser glasses.

In rare-earth-doped laser glasses, the luminescent properties were determined via the analysis of the local structure of the rare-earth ions. The local structure of rare-earth ions in multicomponent laser glasses as a statistical ensemble of their configurations in NGCs was simulated based on the broken ergodicity theory. The NGCs in the compositional space (i.e., Li₂O·Al₂O₃·4P₂O₅, 9Li₂O·3Al₂O₃·8P₂O₅, P₂O₅, Li₂O·P₂O₅, Al₂O₃·3P₂O₅, Li₂O·2MgO·P₂O₅, MgO·P₂O₅, Li₂O·6MgO·3P₂O₅, Al₂O₃·P₂O₅, MgO·2P₂O₅ and 2MgO·Al₂O_3·P₂O₅) were identified via querying the Materials Project database for the quaternary phosphate system (i.e., Li₂O-MgO-Al₂O₃-P₂O₅). The local microstructure and luminescent properties of all glass compositions in this system were derived from these 11 NGCs by the NGCs model.

The molecular dynamics (MD) simulations were performed by the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS). The Morse potential, suitable for modeling short-range interactions in rare-earth-doped glasses, was employed. The long-range interactions were treated by the Ewald summation method (cutoff radius: 12.0 Å), while the short-range interactions were truncated at 8.0 Å. The glass quenching process was simulated at a fixed timestep of 2 fs. The initial structure was thermalized at 5000 K for 1.0 ns to eliminate memory effects. The system was then cooled to 3000 K in the canonical (NVT) ensemble at 5 K/ps. To simulate the experimental conditions, the system was equilibrated in the isothermal-isobaric (NPT) ensemble at 3000 K for 1.0 ns, allowing simultaneous relaxation of the simulation box dimensions and atomic positions. The further cooling to 300 K was conducted in the NPT ensemble at 0.5 K/ps, with intermediate equilibration steps at 2500, 2000, 1500, 1000 K, and 300 K (1 ns each), respectively. Finally, a 1 ns equilibration in the microcanonical (NVE) ensemble was performed, with atomic coordinates recorded every 5000 steps during the last 500 ps.

The radial distribution function g(r) and structure factor S(q), representing short- and medium-range structural orders, were calculated by both the NGCs model and MD simulations. The results from these methods have remarkable consistency, indicating the NGCs model's capability to quantitatively determine the local structural features of Nd³⁺ in Li₂O-MgO-Al₂O₃-P₂O₅ glasses. Since the luminescent properties of rare-earth-doped glasses are governed by the local coordination environment, this agreement validates the NGCs model for predicting luminescent performance.

The NGCs model is further applied to calculate key luminescent properties, i.e., the emission cross-section (σ_e) and fluorescence lifetime (τ) of the ⁴F_3/2→⁴I_11/2 transition. The σ_e values increase monotonically with Li₂O content, showing a maximum relative deviation of 1.72% from experimental data. For τ that is influenced by non-structural factors such as hydroxyl quenching and Nd³⁺ clustering, the maximum relative error is 6.19%. These results underscore the model's accuracy in predicting luminescent properties. A robust composition-property mapping is established via leveraging this model, enabling efficient screening of high-performance Nd³⁺-doped phosphate glass compositions.

A comprehensive analysis of 11,355 valid glass compositions within the glass-forming region reveals an inverse correlation between σ_e and τ, i.e., compositions with a higher σ_e typically exhibit a shorter τ. However, the model identifies a subset of optimized compositions with both high σ_e and extended τ, corresponding to Al₂O₃ (0-7%), MgO (10%-18%), and Li₂O (1%-7%).

This study applied the NGCs model to analyze the local structure and luminescent properties of Nd³⁺-doped Li₂O-MgO-Al₂O₃-P₂O₅ glasses. The calculated g_Re (r) and S_Re(q), reflecting short- and medium-range orders, aligned closely with the MD simulation results, indicating the model's precision in characterizing Nd³⁺ local structures. The model achieved exceptional accuracy in predicting luminescent properties, with maximum relative errors of 1.72% for σ_e and 6.19% for τ, compared to the experimental data. The systematic screening of the glass-forming region identified optimal compositional ranges (i.e., Al₂O₃: 0-7%, MgO: 10%-18%, Li₂O: 1%-7%) that could simultaneously maximize σ_e and τ.

The NGCs model could provide a groundbreaking approach for designing advanced Nd³⁺-doped phosphate laser glasses via establishing a quantitative link between atomic-scale structures and macroscopic performance. This methodology could accelerate material discovery and reduce reliance on costly experimental iterations, paving an efficient way for the development of next-generation laser materials.

In recent years, glass-ceramics have become a preferred material for new-generation glass covers due to their superior mechanical properties, excellent optical characteristics, and stable chemical performance. However, their inherent brittleness makes them prone to catastrophic fracture during crack propagation. To solve this problem, ion exchange technology as a strengthening method designed for aluminosilicate glass systems is widely adopted. This process significantly enhances mechanical properties such as hardness and flexural strength, while maintaining the substrate optical transparency. However, this process introduces a large amount of sodium and potassium ions onto the glass surface, which simultaneously instigates the degradation of chemical stability. This degradation primarily arises from water-induced corrosion, which occurs due to two sequential stages, i.e., leaching, where alkali metal ions (e.g., K⁺, Na⁺) undergo ion exchange with H₃O⁺/H⁺ in aqueous environments, generating an alkaline surface; and network dissolution, where hydroxyl ions attack and depolymerize the silicate network. Crucially, condensation droplets concentrate OH^- ions while atmospheric CO₂ forms highly corrosive mixed alkali carbonates under humid conditions, dramatically accelerating degradation. This fundamentally restricts the applicability of ion-exchanged glass-ceramics as an advanced display material, which a sustained surface integrity under environmental exposure is essential. It is thus necessary to improve the surface chemical stability of chemically strengthened microcrystalline glass in industry. To address this critical industrial challenge, this study was to introduce an immersion treatment using zinc chloride (ZnCl₂) solution designed to enhance chemical stability and impact resistance in lithium aluminosilicate (LAS) glass-ceramics simultaneously.

LAS glass-ceramic specimens (composition: 69.43% SiO₂, 5.62% Al₂O₃, 12.13% Li₂O, 6.21% ZrO₂, 5.27% P₂O₅, 1.34% Na₂O; primary crystalline phases: lithium disilicate and β-spodumene) were processed through sequential cutting, grinding, and polishing procedures to obtain glass samples with the sizes of 147.0 mm×72.0 mm×0.6 mm. The specimens underwent sequential chemical strengthening, i.e., immersion in molten NaNO₃ at 440 ℃ for 4 h, followed by molten KNO₃ at 410 ℃ for 0.5 h. Afterwards, an immersion treatment was performed in both distilled water and ZnCl₂ solutions at different contents (i.e., 0.05, 0.10, 0.20 mol/L) at 98 ℃ for 3.5 h, respectively. The treated and untreated specimens underwent accelerated weathering at 85 ℃/85% RH for 1, 4, and 6 d. The tests evaluated post-treatment efficacy for enhancing surface chemical stability and analyzed the underlying mechanisms. The elemental distributions (i.e., Na, K) on the surfaces and cross-sections of samples before and after treatment under different conditions were characterized by scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS). The surface roughness of samples was characterized by atomic force microscopy (AFM) across 30 μm×30 μm scan areas. The impact resistance of samples before and after treatment under different conditions was evaluated by drop-ball testing (132 g steel ball, maximum drop height of 1700 mm), and the results represented the average values from 10 samples per group.

Immersion in ZnCl₂ solutions significantly modifies the surface composition and properties of the chemically strengthened glass-ceramic samples. The SEM-EDS analyses indicates that this immersion process reduces the concentration of K⁺ on the surface, while leading to an increase in Na⁺ content. Particularly, ZnCl₂ solutions with higher concentrations enhance the outward migration of K⁺ from the subsurface region. The reason is since a weak acidity of ZnCl₂ solution promotes H⁺/OH^- neutralization, thereby enhancing the ion exchange process. In parallel, Na⁺ migration appears in the samples during immersed progress, particularly in distilled water. This is due to the reduced cross-linking density is caused by the formation of Si-OH and the fracture of Si-O-Si bonds after K⁺ ions on the surface are displaced. This induces a relaxation in the surface network structure, forming diffusion channels. Consequently, the weakly bound Na⁺ migrates outward under the synergistic effects of concentration gradients and structural defects. However, Na⁺ migrates in the samples treated with ZnCl₂ solutions, as the smaller, higher-charge Zn²⁺ ions fill in the surface pores.

For immersion, the surface roughness of the samples increases. The most severe damage, measured at 3.2 nm occus in samples treated with distilled water due to the sustained corrosion of the glass network. In contrast, ZnCl₂ solutions cause less severe, and concentration-dependence increases in the roughness (i.e., 1.5-2.2 nm). This is because H⁺ neutralization reduces the availability of alkaline media, thereby inhibiting silicate regeneration and limiting network degradation.

The results of accelerated weathering tests conducted at 85 ℃ and 85% relative humidity (RH) for 1, 4, and 6 ddemonstrate an enhancement in the chemical stability of samples treated with ZnCl₂. The untreated samples exhibit dense white spots after just 1 d and develop severe haze at 6 d. The samples immersed in distilled water show minor spots rich in Na⁺, while those treated with ZnCl₂ exhibit a minimal visible degradation. This improvement is attributed to two mechanisms, i.e., 1) a reduction in surface K⁺ concentration, which suppresses ion exchange driven by leaching, and 2) the diffusion of Zn²⁺, which blocks further alkali ion leaching and H⁺ adsorption.

The SEM-EDS analysis shows some distinct corrosion patterns. The untreated samples display severe dendritic K₂CO₃ aggregates and needle-like corrosion zones. The samples treated with distilled water show fine Na⁺-rich surface spots with limited corrosion. The samples treated with 0.05 mol/L ZnCl₂ have sparse deposits without significant elemental anomalies, indicating the blocking effect of Zn²⁺ against alkali migration.

The samples treated with 0.05 mol/L ZnCl₂ exhibit a maximum average impact height of approximately 1600 mm, which is 1.5 times greater than that of the untreated samples. This impact resistance improvement is attributed to the blunting of microcracks on the glass surface during mild corrosion.

This study investigated the effect of ZnCl₂ immersion on the surface chemical stability of chemically strengthened glass-ceramics. The results revealed that ZnCl₂ immersion significantly reduced surface alkali metal ion concentrations and suppressed their outward diffusion, substantially improving the chemical stability. The treated specimens exhibited superior corrosion resistance and surface integrity under accelerated aging (i.e., 85 ℃/85% RH). Also, ZnCl₂ treatment simultaneously enhanced the impact resistance. The samples immersed in 0.05 mol/L ZnCl₂ at 98 ℃ for 3.5 h achieved a maximum average impact height of 1600 mm (i.e., 132 g steel ball), which was 1.5 times greater than that of the untreated samples. These results indicated that ZnCl₂ immersion treatment could improve surface chemical stability and mechanical durability, having an effective surface-modification technology for glass-ceramics.

Silicate glass is a versatile material widely used in architectural windows, electronics, optics, and nuclear waste storage. As science and technology advance, demand grows for high-performance glass materials. The mechanical properties, which directly impact service life, safety, and functionality, have attracted much attention. Conventional glass development relies on empirical experience and trial-and-error methods, which are slow and limited in scope. In addition, the amorphous nature and broad compositional range of glass materials make designing tailored materials challenging, requiring composition optimization, cost reduction, and process assurance. Recent advancements on computational modeling, machine learning (ML), and materials informatics offer promising solutions. ML models can predict glass properties and accelerate the discovery of high-performance compositions via leveraging high-dimensional compositional data and integrating computational and experimental insights. However, ML effectiveness depends on dataset quality—availability, completeness, consistency, accuracy, and representativeness. In this work, three ML-based models were constructed from three distinct sources, and datasets were sourced from SciGlass (density), molecular dynamics simulations (Young's modulus), and a QSPR model (hardness). This approach could provide an innovative dataset construction strategy for advancing glass composition design.

In this study, ML models were developed to predict the density, Young's modulus, and hardness of silicate glasses. The dataset comprised 425 density values and 43 hardness values selected from the SciGlass database, 425 Young's modulus values obtained through MD simulations, and 50 hardness values predicted using a QSPR model. In the MD simulations, the Teter potential was employed, while the QSPR modeling incorporated descriptors was related to electronegativity. This work focused on six-component borosilicate glasses composed of SiO₂-B₂O₃-Al₂O₃-MgO-CaO-Na₂O. For initial data collection, a data preprocessing step was conducted, which included outlier removal, elimination of duplicate entries, and normalization of the compositional variables. A multilayer perceptron (MLP) neural network was used to model the relationship between glass composition and properties. The neural network architecture consisted of an input layer, several hidden layers, and an output layer, with weighted connections between neurons. The input features were the molar percentages of the oxide components, while the outputs were the corresponding physical properties. Non-linearity was introduced through the use of the rectified linear unit (ReLU) activation function in the hidden layers, enabling the model to capture complex nonlinear relationships. The output layer used a linear activation function to provide direct predictions of glass properties. To train and validate the models, the dataset was split into training and test sets in a ratio of 9∶1. This partitioning ensured sufficient training data, while reserving a subset for independent evaluation of model generalization. Model performance was assessed using mean squared error (MSE) and the coefficient of determination (R²). The trained models were subsequently used to predict the density, Young's modulus, and hardness of silicate glasses. Finally, nine formulations and corresponding properties were predicted from the models and experimentally verified.

The compositions of the dataset are widely distributed. The density of the glass varies between 2.3 g/cm³ and 2.8 g/cm³, Young's modulus is primarily distributed between 82 GPa and 115 GPa, and hardness values range from 5.0 GPa to 7.0 GPa. The QSPR model related to hardness demonstrates a good linear correlation between hardness and the F_net descriptor, with the R² value of greater than 0.80, indicating good model fitting and predictive performance. In this study, three ML models are developed to predict the density, Young's modulus, and hardness of silicate glasses. The R² values of the models show clear differences. The density-related model shows a relatively low predictive accuracy, with an R² of only 0.30 on the test set, the model for Young's modulus achieves an R² of approximately 0.70 and the hardness-related model exhibits the maximum performance, with R² of > 0.95. However, the predictive accuracy of all three models was satisfactory. The maximum relative error for density prediction is only 2.67%. The predictions of Young's modulus are overestimated by approximately 10 GPa, mainly due to the limitations associated with empirical interatomic potentials. The prediction error for hardness ranges from approximately 0.05 GPa to 0.40 GPa. The SHAP analysis is further used to investigate the contributions of individual oxide components to the properties. The results reveal that SiO₂ and Al₂O₃ contents positively affect both Young's modulus and hardness, while Na₂O negatively affects the glass density. As a result, the use of MD simulations and QSPR method represents a promising alternative in the absence of abundant high-quality experimental data for model training.

This study presented a comprehensive, multi-source modeling framework for predicting the mechanical properties of silicate glasses by integrating experimental data, MD simulations, and QSPR-based data augmentation. A total of 425 experimental density values, 425 MD-derived Young's modulus values, and 93 hardness values (43 experimental, 50 QSPR-predicted) were used to train three distinct ML models. Nine formulations were predicted and experimentally verified for three properties (i.e., density, Young's modulus and hardness) based on the three models above. The experimental results exhibited small relative errors, indicating a reasonable agreement with model predictions. The results of Young's modulus took into account the effect of errors introduced by simulation. The data quality could become a critical factor affecting the accuracy of predictive models as using ML method. This work incorporated homologous glass property data obtained from MD simulations and employed QSPR method for rapid data augmentation based on experimentally acquired data. This strategy could address the limitations commonly encountered in conventional database construction, including inconsistent data quality and slow acquisition of experimental measurements. It could provide a novel approach toward establishing a data-driven paradigm for glass materials research.

In recent years, microwave-absorbing composites that are both cost-effective and high-performance are developed to address the electromagnetic radiation issues like interference, pollution, and health risks. Spinel ferrite glass-ceramics as all-inorganic composites have attracted wide attention. They combine the magnetic properties of ferrite crystals with the advantages of glass, i.e., high temperature resistance, oxidation resistance, ease of formability, cost-effectiveness, and adjustable dielectric constant. Fe₃O₄ is widely used as an electromagnetic wave absorber due to its high saturation magnetization, strong spin-polarization, half-metallic properties, and low cost. Fe₃O₄ can also utilize industrial solid wastes like iron tailings as raw materials to achieve resource recycling. In this paper, silicate glass-ceramics containing flower-like Fe₃O₄ nanocrystals were prepared based on SiO₂-Al₂O₃-K₂O-Na₂O-P₂O₅-B₂O₃-Fe₂O₃ system via appropriate heat treatment. The special structure could enable the glass-ceramics to exhibit a higher reflection loss. A thorough examination of the mechanism of reflection loss was conducted, elucidating the key factors affecting dielectric loss and impedance matching on the microwave-absorbing performance of Fe₃O₄ glass-ceramics.

Raw materials (Sinopharm Chemical Reagent Co., Ltd. China) were weighed in a nominal composition of the glasses with approximately 30 g of each. Then, they were thoroughly mixed to ensure uniformity, and melted in a corundum crucible in an electric furnace at 1 580 ℃ for 2 h. Afterwards, the melt was poured into a mold for shaping. The as-quenched glasses were annealed at 500 ℃ for 1 h. Finally, the glasses were heat-treated at 620-740 ℃ with a heating rate of 10 ℃/min for 2 h. The samples were groundinto powders or cut and polished into small pieces for subsequent characterizations.

The crystalline phases of the samples were determined by X-ray diffraction (XRD). The crystallinity was determined via calculating the I_a/I_b value, where I_a is the crystalline phase intensity and I_b is the total intensity of the sample. The microstructures of the glass-ceramics were characterized by scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HR-TEM) with energy dispersive spectroscopy (EDS). The coercive field (H_c), initial magnetic permeability (µ_i), and magnetic saturation (M_s) of the samples at room temperature were determined by vibrating sample magnetometry (VSM). The glass powder was heated and thoroughly mixed with paraffin at a ratio of 4∶1, then pressed into a ring-shape with an outer diameter of 7 mm, an inner diameter of 3.04 mm, and a thickness of 2 mm in a mold. In the frequency range of 2-18 GHz, the electromagnetic parameters were measured by a vector network analyzer (VNA), and the reflection loss values of the samples were obtained based on the transmission line theory.

The results of scanning electron microscopy (SEM), vibrating sample magnetometry (VSM), and vector network analysis (VNA) reveal that an increase in temperature results in a gradual increase in the number of Fe₃O₄ grains, while the size of the flower-like structures formed by their aggregation initially increases and subsequently decreases. Also, the saturation magnetization (M_s) increases from 5.2 to 9.5 emu·g^-1 although the reflection loss exhibits an initial increase and a subsequent decline. When heat-treated at 700 ℃, the samples exhibit the most significant enhancement in Fe₃O₄ flower-like structures and the most optimal microwave absorption performance. The interfaces of the polyphase structure in the glass-ceramics generate polarization and relaxation, which enhance the dielectric loss. The mutual migration of electrons between Fe²⁺ and Fe³⁺ ions at octahedral sites induces a loss of conductivity. Meanwhile, electron hopping triggers a resonance, leading to an enhanced dielectric loss. The enhancement of crystallinity in the glass matrix promotes the magnetism, thereby leading to an improvement in magnetic loss. As the value of dielectric loss is close to that of the magnetic loss, the glass attains an effective impedance matching, thereby enhancing its reflection loss. This allows for enhanced penetration of electromagnetic waves into the material, leading to optimal reflection loss and effective absorption bandwidth at a thickness of 7 mm.

Silicate glass-ceramics containing flower-like Fe₃O₄ were prepared. The size of the flower-like structure was adjusted via heat-treatment temperature. At 700 ℃, the size of the flower-like structure was the largest. For a thickness of 7 mm, the reflectivity reached -18 dB, and the effective absorption bandwidth was approximately 2 GHz. Based on the analysis of the Debye equation and Cole-Cole plots, the larger flower-like nanostructures enhanced the reflection and scattering of electromagnetic waves during their transmission in the glass-ceramics. This led to an increase in dielectric loss and the formation of better impedance matching, revealing a multiple conduction-reflection-electromagnetic wave absorption mechanism.

Quartz glass is a type of inorganic amorphous material prepared by ultra-clean high-temperature melting process with high-purity natural quartz or synthetic silicon compounds as raw materials, having an outstanding physical and chemical property. This material possesses superior optical performance, extreme thermal stability, ultra-low thermal conductivity, and superior dielectric properties. In the development of optoelectronic technology industry in China, quartz glass with its unique synergy of physical and chemical parameters becomes a strategic basic material in the field of optoelectronic devices. The feature size of devices is in a hundred-micron scale based on the iterative evolution of microelectromechanical systems and wafer-level packaging technologies, further demanding a deeper analysis of the surface structure of quartz glass.

The glass was firstly prepared in a deposition furnace, and then heated in a high-temperature homogenization furnace at 1800 ℃ for 2 h. Afterwards, the glass was cooled to room temperature and discharged. The glass before and after homogenization was cut and polished to obtain glass samples with the sizes of 200 mm×200 mm×5 mm. Starting from the first point position in the upper left corner, some points were taken at intervals of 20 mm in all the directions (i.e., up, down, left, and right) for the analysis of Raman spectroscopy. In addition, the PCA algorithm was utilized for data dimensionality reduction and visualization, and the results of the changes in peak center and half-peak width were also analyzed.

Before homogenization, the standard deviations of LO_FWHM and D2_FWHM are 15.43 cm^-1 and 11.73 cm^-1 respectively, which are relatively larger than those of other indicators. The fluctuation range of the two sets of data is also larger, with the mean values of 114.13 cm^-1 and 47.30 cm^-1, respectively. The standard deviation of D1_center is only 0.18 cm^-1, indicating that the data for this indicator are concentrated, with small differences between each data point. The difference between the maximum and minimum values is also similar. For instance, the maximum value of LO_FWHM is 178.79 cm^-1 and the minimum value is 100.28 cm^-1; the maximum value of D2_FWHM is 115.58 cm^-1 and the minimum value is 42.77 cm^-1. This further indicates that the dispersion before homogenization of the two sets of data is large.

After homogenization, the standard deviation of D2_center is 0.39 cm^-1, which is decreased by 79%, compared to that before homogenization. The standard deviation of D2_FWHM is 0.82 cm^-1, which is decreased by 93%, compared to that before homogenization. The standard deviation of LO_FWHM is 8.03 cm^-1, which is decreased by 48%, compared to that before homogenization. The data fluctuations between different positions significantly reduce, and the standard deviation statistics for each indicator also show a certain degree of reduction. Among them, the influence of D2 peak is dominant, and the influence of D1 peak is subordinate. The D1 peak and D2 peak are respectively related to the tetra-siloxane ring and tri-siloxane ring in the structure. This indicates that on the same sample surface, the homogenization process has a dominant impact on the tri-siloxane ring in the silicon-oxygen structure. The tri-siloxane ring is expected in quartz glass because their Si—O—Si angle is less than the lowest energy (most likely) angle in the glass. The estimation of the energy required to generate such rings is more likely to reach the distribution of the D2 tri-siloxane ring. Also, the silicon—oxygen bond angles of all rings tend to be averaged, and the overall is more normalized.

The Raman spectroscopy could precisely characterize the microscopic structural differences of quartz glass before and after homogenization via detecting the vibration modes of atoms in the silicon-oxygen network (i.e., Si—O—Si bond angles and ring structure defects). For the analysis of structural uniformity, some parameters such as peak center dispersion and half-width were used to quantify the structural uniformity of the glass surface and bulk phase, and to identify high-strain areas (i.e., the regions with concentrated trimer ring defects). In the study of defect mechanisms, the D1 peak (tetramer ring) and D2 peak (trimer ring) defect characteristic peaks were utilized to reveal the migration and recombination laws of defects during homogenization, and to clarify the influence of thermal stress release on structural relaxation. Finally, combined with the data mining technology of the PCA algorithm, a correlation between the Raman parameters and homogenization processes was established, thus providing a theoretical support for precisely regulating the microscopic structure of quartz glass.

Low-temperature co-fired ceramic (LTCC) serves as a critical material for high-performance electronic packaging. Reducing the dielectric loss of LTCC materials can minimize signal attenuation in high-frequency transmission and enhance circuit integration. Based on material composition, LTCC materials can be classified into three categories, i.e., glass-ceramic systems, ceramic systems, and glass/ceramic composite systems. Glass-ceramic systems typically exhibit a low dielectric loss but suffer from inferior mechanical properties and corrosion resistance. Ceramic systems are often employed in medium- and high-dielectric applications, such as dielectric resonators. Reducing the sintering temperature is a key challenge for this system. Glass/ceramic composite systems offer a balanced performance. Their properties can be tailored via adjusting the ratio of the two phases. The La₂O₃-B₂O₃ glass/Al₂O₃ ceramic system serves as one of the representative systems. The sintering and crystallization behavior can be controlled via modifying the glass composition. The La₂O₃-B₂O₃-CaO-P₂O₅ (LBCP) glass/Al₂O₃ composite system demonstrates a significant application potential due to its ability to crystallize low-loss phases such as LaBO₃, LaPO₄, and CaAl₂(BO₃)₂O during sintering. However, the dielectric loss (tanδ > 1×10^-3) of the existing LBCP/Al₂O₃ composites requires a further improvement. In this paper, alkali metal oxides (i.e., Na₂O and Li₂O) were introduced to control the glass crystallization behavior, and the sintering process was refined to enhance the formation of beneficial crystalline phases (i.e., CaAl₂(BO₃)₂O) to develop low-loss LTCC substrates.

La₂O₃-B₂O₃-CaO-P₂O₅-R₂O glasses were prepared by a high-temperature melting method. Analytical-grade raw materials were weighed according to the designed composition and thoroughly mixed. The mixture was melted at 1400 ℃ for 1 h, followed by rapid quenching in deionized water to obtain crushed glass. The crushed glass was ground in a ball mill to obtain glass powders with a median particle size (D₅₀) of 2-3 μm. The glass powder was mixed with solvent and binder to make the slurry, and then the slurry was casted into green tapes. The glass powder was mixed with Al₂O₃ powder in a mass ratio of 45∶55 and casted into LTCC green tapes. Glass and LTCC green tapes both were cut, stacked, and laminated to form green blocks, which were subsequently sintered into dense ceramics.

The structure of the base glass was characterized by solid-state nuclear magnetic resonance spectroscopy (NMR) and Raman spectroscopy. The thermal behavior of the samples was detremined by differential scanning calorimetry (DSC). The phase composition of the sintered samples was identified by X-ray diffraction (XRD). The microstructure of the samples was determined by scanning electron microscopy (SEM). The porosity was quantified by processing the microstructural images via a software named ImageJ. The dielectric constant and dielectric loss of the samples at 5 GHz were measured by a vector network analyzer.

The LBCP glass network consists of low-polymerization borate and phosphate groups. The introduction of alkali metal oxides further disrupts the glass network, increasing the proportion of small depolymerized borate and phosphate groups, while reducing the overall network polymerization degree. The LBCP glass exhibits a great crystallization due to the weakened integrity of the glass network and the enhanced short-range ordering induced by high-field-strength La³⁺ ions. In sintering at 850 ℃, the LBCP glass crystallizes abundant crystalline phases LaBO₃, CaB₂O₄, and LaPO₄. The introduction of alkali metal oxides further enhances crystallization, reducing the activation energy for crystallization and lowering both crystallization and polymorphic transition temperatures, without altering the types of crystalline phases. Note that Li₂O demonstrates a greater effect on promoting glass crystallization, compared to Na₂O. This great crystallization promotes more complete growth of crystals LaBO₃ and LaPO₄, and the dielectric constant and dielectric loss decrease. The great crystallization behavior results in an increased porosity in glass sintering.

The LBCP glass/Al₂O₃ composite precipitates multiple crystalline phases LaBO₃, LaPO₄, and CaAl₂(BO₃)₂O after sintering. During sintering, CaB₂O₄ reacts with Al₂O₃ to form CaAl₂(BO₃)₂O, resulting in the disappearance of CaB₂O₄. The Al₂O₃ particles increase the resistance to densification shrinkage, leading to a higher porosity. The sintered composite samples exhibit superior dielectric properties due to the high content of Al₂O₃ and CaAl₂(BO₃)₂O. The introduction of Na₂O enhances mass transport, promoting CaAl₂(BO₃)₂O precipitation and further reducing both the dielectric constant and loss. In contrast, Li₂O addition causes a premature crystallization, increasing glass viscosity and hindering CaAl₂(BO₃)₂O formation, thereby deteriorating dielectric performance. The optimization of sintering conditions indicates that CaAl₂(BO₃)₂O precipitation occurs at a heating rate of 2 ℃/min, achieving optimal dielectric properties (i.e., ε_r of 6.70 and tanδ of 0.80×10^-3).

The glass network of LBCP glass consisted of low-polymerized borate and phosphate groups, exhibiting a great crystallization. LBCP glass crystallizes LaBO₃, LaPO₄ and CaB₂O₄ when sintered at 850 ℃. Introducing alkali metal oxides further depolymerized the glass network, enhancing a crystallization behavior with an improved grain growth, reducing both dielectric constant and dielectric loss. However, this greater crystallization also increased the porosity of the sintered glass. After sintering, the LBCP glass/Al₂O₃ composite material precipitates a variety of crystal phases, mainly including LaBO₃, LaPO₄ and CaAl₂(BO₃)₂O. High contents of Al₂O₃ fillers and CaAl₂(BO₃)₂O phase could improve the dielectric properties of the sintered composites. The addition of Na₂O promoted CaAl₂(BO₃)₂O formation, thereby reducing both the dielectric constant and dielectric loss of sintered composites. In contrast, Li₂O addition induced a premature crystallization, increasing glass viscosity of the glass phase and inhibiting CaAl₂(BO₃)₂O formation, thereby elevating dielectric loss. Optimizing the heating rate to 2 ℃/min could facilitate complete sintering and maximize CaAl₂(BO₃)₂O content, yielding LTCC materials with optimal dielectric properties (i.e., ε_r of 6.70 and tanδ of 0.80×10^-3).

With the rapid iteration of smart displays, especially smartphones, cover plate glass as one of the key materials has made recent progress, becoming an important component in the field of electronic information display applications. Aluminosilicate glass with its unique properties (i.e., good impact resistance, scratch resistance, and high hardness, as well as excellent chemical and thermal stability) is an ideal candidate material for plate cover glass. There is relatively in-depth research on the influence of composition changes in aluminosilicate glass on its structure and properties, but most of the existing research focuses on the short-range structure (<5 Å) of the aluminosilicate glass network structure, while there is less research on its medium-range structure (5-20 Å). In this work, the influence of multi-layer network structure of aluminosilicate glass on its properties was investigated by molecular dynamics simulation and experiments.

The relationship between the composition-structure-properties of glass was investigated by experimental methods and molecular dynamics simulations. The glass samples were prepared by a conventional high-temperature melting method. The raw materials were weighed and mixed uniformly according to the molar ratio of oxides, put into a platinum-rhodium crucible (90% Pt and 10% Rh), and then heated in a high-temperature lifting furnace at 1640 ℃ for 2 h. After melting, the crucible containing the molten glass was removed and poured onto a preheated copper mold. Once it solidified, it was transferred to an annealing furnace, followed by natural cooling to room temperature to eliminate internal stresses in the glass, resulting in uniform, transparent bulk glass. The glass was then cut and mechanically polished to obtain the samples with various sizes.

The glass samples were simulated, and the simulated structure was a cubic box composed of approximately 10000 atoms. The initial dimensions of the simulation box were calculated based on predicted density and atomic mass. The initial structure was optimized by a soft potential to ensure structural integrity and prevent atomic loss. The potential function used was the Buckingham potential, and the Ewald summation method was used to handle long-range Coulomb interactions. The cutoff radius was set to 10 Å, with an integration time step of 0.001 ps. The initial structure was randomly generated by an open-source software named ATOMSK.

The statistical results of the short-range structural characteristics of silicate glass indicate that there are no significant changes in the position and intensity of the O—O peak, and the intensity of the Al—O peak near 1.7 Å gradually increases as Al₂O₃ content increases. The number of Al—O structural units in the glass network structure increases, having a clear symmetrical distribution with aluminum primarily in a four-coordinate environment. The strength of the Si—O bond decreases with the reduction of SiO₂. The number of [SiO₄] tetrahedral groups decreases, and some alkaline earth metal ions in the glass network structure transform into ions that serve to compensate for the charge of the [AlO₄] tetrahedral groups.

The statistical analysis of the medium-range structural characteristics of aluminosilicate glass indicates that as Al₂O₃ content increases, the Q value of the first sharp diffraction peak (FSDP) shifts towards higher values, the ring size distribution shows a certain sharpening degree with the decreasing proportions of small rings (i.e., 3-membered and 4-membered rings) and large rings (i.e., ≥7-membered rings), while increasing the proportions of medium-sized rings (i.e., 5-membered and 6-membered rings). The smaller rings, such as the three-membered and four-membered rings, typically possess a higher energy and are more prone to structural rearrangement or breakage, whereas the larger rings, such as the five-membered and six-membered rings, are relatively stable. This indicates a reduction in the medium-range disorder of the glass network structure, resulting in a denser network structure.

The mechanical properties of the glass (i.e., density, elastic modulus, and Vickers hardness) gradually increase with the increase of Al₂O₃ content. The mechanical characteristics of the glass are positively correlated to the density of its network structure. The increase in the content of the [AlO₄] network structure makes the network structure denser, while the combined effect of [AlO₄] and [SiO₄] gradually enhances the integrity of the glass network. This results in a decrease in the average coefficient of thermal expansion, an increase in the strain point and softening point, and an increase in high-temperature viscosity.

The relationship between the composition-structure-properties of aluminosilicate glass was elucidated through experiments and molecular dynamics simulations. The substitution of Al³⁺ for Si⁴⁺ induced changes in the short-range and medium-range network structures of the glass, leading to an increase in the bridging oxygen content and a sharpening of the ring size distribution. The structural factor could shift towards higher values at the position of the first sharp diffraction peak (FSDP), resulting in an increase in the order of the short-range and medium-range network structures, and making the network structure denser. Consequently, the density, elastic modulus, Vickers hardness, strain point, softening point, and viscosity of the glass melt all increased, while the average coefficient of thermal expansion decreased.

Electrochromic device (ECD) is comprised of five superimposed layers, i.e., transparent conducting layer, electrochromic layer, ion conductor layer, ion storage layer and another transparent conducting layer, respectively. It can modulate the amount of solar radiation transmitting through it under applied voltage, often used as smart windows or facades to adjust the light and heat entering buildings proactively, which can sufficiently reduce the energy consumption. The existing ECDs are mainly fabricated by a magnetron sputtering method, which has a low productivity and high cost, especially for the ion conductor layer. Sol-gel method is a simple and relatively cheap method for preparing thin films, and massive researches are conducted on sol-gel electrochromic thin films such as tungsten oxide, nickel oxide etc.. In this work, an ion conductor layer was prepared by a sol-gel method, and its application in electrochromic device and their performance were also investigated.

Lithium tantalite sol was synthesized with lithium acetate (LiAc) and tantalum pentaethoxide (TaEt) as precursors and 1,2-Propanediol as a solvent. The molar ratio of LiAc and TaEt was 1∶1. Lithium tantalite thin films were prepared by a spin coating method and heat-treated in air at 400 ℃ for 2 h. Some of the sol was dried and the subsequently formed powder was heated at 600-700 ℃. The phase compositions of the films and powders were characterized by X-ray diffraction (XRD), and the electrochemical properties of thin films were analyzed.

The electrochromic devices with a structure of ITO/NiO/LiTaO₃/WO₃/ITO were prepared by PVD method and subsequent sol-gel method. The tungsten oxide or nickel oxide thin film layer was firstly sputtered on the ITO glass and then lithium tantalite thin films were spin-coated onto the as prepared films. After heating process at 400 ℃ for 2 h, another layer of the tungsten oxide or nickel oxide films were sputtered, and another layer of ITO thin film was finally sputtered as a conducting electrode. The electrochromic devices were noted as EC-N when lithium tantalite was spin-coated onto nickel oxide films and as EC-W when lithium tantalite was spin-coated onto tungsten oxide films. The transmittance changes were measured by UV-vis-IR spectrometer.

The XRD patterns show that powders dried and heated from the sol are a polycrystalline lithium tantalite, and thin films are amorphous. The analysis of electrochemical impedance spectroscopy indicates that the sol-gel lithium tantalite thin film has an ionic conductivity of 7.44×10^-7 S/cm at room temperature, which is comparable to that of sputtered film. As for the devices, the XRD patterns show that nickel oxide thin film layers are both polycrystalline in EC-N and EC-W, while tungsten oxides is polycrystalline in EC-W and amorphous in EC-N. The XRD patterns and XPS spectra also show the existence of WO_3-x phase in polycrystalline tungsten oxide films.

For electrochromic properties of the two groups of devices, transmittance regulations are measured at a fixed wavelength of 550 nm. For both groups, the differences of transmittance between colored and bleached state increase with increasing voltage. The maximum of transmittance change for EC-N is 26% at 5 V, with transmittance of around 72% at bleached state and 46% at colored state. Its coloring time is 50 s and bleaching time is 56 s. The results of cyclic test show that the differences of colored and bleached states keep stable, no obvious degeneration occurs after 60 cycles. The transmittance changes for EC-W have a lower value when applied the same voltages as in EC-N, and the switching speed is much slower. Its transmittance change is 14% at 3 V, while the coloring and bleaching time are both larger than 2 min, exceeding the setting range. It can be concluded that EC-N has a better property for regulation of visible light, and crystallinity can slow the switching process.

The transmittance spectra of colored states in vis-NIR range show that EC-W has a significant modulating performance for NIR, which can be attributed to the existence of WO_3-x components. As reported, oxygen deficient components can generate a local surface plasmon resonance effect that can result in absorption in the spectral range of NIR.

Electrochromic devices fabricated by a process that lithium tantalite was spin-coated on nickel oxide had better electrochromic properties in a visible spectral range, with larger optical modulations and a higher switching speed less than 1 min. The devices with lithium tantalite spin-coated on the tungsten oxide films could have a slower coloring and bleaching rate, while showing a better near-infrared modulation effect. These results indicated a routine for the realization of electrochromic devices that could modulate light and heat simultaneously and independently.