.
Lead-free ferroelectric ceramics based on sodium bismuth titanate (Na0.5Bi0.5TiO3, NBT) are promising candidates for environmentally benign multifunctional devices; however, their practical application is limited by high coercive fields and restricted electromechanical response. In this study, 0.94Na0.5Bi0.5-x-yErxYbyTiO3–0.06BaTiO3 (NBEY-BT) ceramics with (x = y = 0.0 0.1) were synthesized via a conventional solid-state route to investigate the effect of Er3+/Yb3+ co-doping on the structural, microstructural, ferroelectric, and optical properties. X-ray diffraction combined with Rietveld refinement confirms a single-phase perovskite structure with coexisting rhombohedral (R3c) and tetragonal (P4mm) phases, characteristic of morphotropic phase boundary behavior. Er3+/Yb3+ co-substitution increases the tetragonal phase fraction and induces local lattice strain without forming secondary phases. Microstructural analysis reveals dense ceramics with uniform grain distribution and a reduced average grain size upon rare-earth doping. Ferroelectric measurements show well-saturated polarization–electric field hysteresis loops, with the co-doped composition exhibiting a reduced coercive field and enhanced domain switchability while maintaining high polarization. A pronounced improvement in piezoelectric performance is achieved in the doped composition, which exhibits higher d₃₃ values at lower electric fields due to facilitated non-180° domain switching and polarization rotation near the morphotropic phase boundary. Furthermore, efficient Yb3+-sensitized Er3+ upconversion photoluminescence under 980 nm excitation is observed, introducing multifunctional optical functionality absent in the undoped ceramic. Notably, only the Er3+/Yb3+-substituted ceramics display strong upconversion photoluminescence under 980 nm excitation, characterized by green (4S3/2 → 4I15/2) and red (4F9/2 → 4I15/2) emissions arising from efficient Yb3+ → Er3+ energy transfer. The coexistence of ferroelectric polarization and upconversion luminescence in NBEY-BT ceramics highlights their potential as multifunctional dipolar luminescent materials for advanced optoelectronic, sensing, and energy-harvesting applications.
Purpose This paper introduces a novel four-dimensional chaotic electronic circuit modeled using the Caputo-Fabrizio (CF) Fractional derivative (FD), which features a non-singular, exponentially decaying kernel. The existence, uniqueness, and stability of the system's solutions are rigorously established through Picard approximation, Banach's fixed point theorem, and an iterative Laplace transform (LT) technique. A customized fractional Euler method is developed to perform numerical simulations, revealing rich dynamical behavior that depends sensitively on the fractional order a ? (0, 1). The results demonstrate that the proposed fractional-order model effectively captures memory effects, offering a more realistic framework for analyzing complex nonlinear electronic systems compared to classical integer-order approaches. Design/methodology/approach The study introduces a novel four-dimensional chaotic electronic circuit model using the CF FD, which features a non-singular, exponentially decaying kernel. The authors employ Picard approximation and Banach's fixed point theorem to rigorously establish the existence, uniqueness, and stability of the system's solutions. An iterative scheme based on the LT is applied to analyze stability, while a customized fractional Euler method is developed for numerical simulations. This approach enables the exploration of the system's dynamic behavior under varying fractional orders (0 < a = 1), demonstrating how memory effects influence chaos in electronic circuits and offering a more realistic modeling framework compared to classical integer-order systems. Findings The most significant findings of this study revolve around the successful development and analysis of a novel four-dimensional chaotic electronic circuit modeled using the CF FD, which features a non-singular, exponentially decaying kernel. The authors rigorously proved the existence, uniqueness, and stability of the system's solutions using Picard approximation, Banach's fixed point theorem, and an iterative LT technique. A customized fractional Euler method was introduced for numerical simulations, revealing that the system exhibits rich chaotic dynamics highly sensitive to the fractional order a (0 < a = 1). These results demonstrate that fractional-order modeling with the CF operator captures memory effects more realistically than classical integer-order models, offering enhanced fidelity in representing real-world electronic systems with inherent memory dependence. Originality/value This work is original in proposing the first four-dimensional chaotic electronic circuit modeled with the CF FD-characterized by a non-singular, exponentially decaying kernel-offering a more physically realistic representation of memory effects than classical fractional models. Its value lies in the rigorous theoretical analysis (existence, uniqueness, and stability via Picard approximation and Banach's fixed point theorem), the development of a tailored fractional Euler method for simulation, and the demonstration that system dynamics are highly sensitive to the fractional order. These contributions advance both fractional calculus theory and its application in secure communications, circuit design, and chaos-based engineering.
Federated learning is a decentralized model training paradigm with significant potential. However, the quality of Federated Network's client updates can vary due to non-IID data distributions, leading to suboptimal global models. To address this issue, we propose a novel client selection strategy called FedPA (Performance-Based Federated Averaging). This proposed model selectively aggregates client updates based on a predefined performance threshold. Only clients whose local models achieve an F1 score of 70% or higher after training are included in the aggregation process. Clients below this threshold receive the updated global model but do not contribute their parameters. In this way, the low-performance clients are still in the process of learning and, after some rounds, will be able to contribute. If no client meets the performance threshold in a given round, the system falls back to standard FedAvg aggregation. This ensures the global model continues to improve even when most clients perform poorly. We evaluate FedPA on a subset of the MURA dataset for abnormality detection in radiographs of four bone types. Compared to baseline federated learning algorithms such as Federated Averaging (FedAvg), Federated Proximal (FedProx), Federated Stochastic Gradient Descent (FedSGD), and Federated Batch Normalization (FedBN), FedPA consistently ranks first or second across key performance metrics, particularly in accuracy, F1 score, and recall. Moreover, FedPA demonstrates notable efficiency, achieving the lowest average round time (832270 s) and minimal memory usage (83645.58 MB), all without relying on GPU resources. These results highlight FedPA's effectiveness in improving global model quality while reducing computational overhead, positioning it as a promising approach for real-world federated learning applications in resource-constrained environments.
Functionalized dibenzothiophene‐S,S‐dioxides (DBTO) have emerged as a versatile class of heteroaromatic scaffolds with remarkable optoelectronic and photophysical properties. Their rigid π‐conjugated framework, combined with the strong electron‐withdrawing nature of the sulfone group, facilitates precise tuning of frontier molecular orbitals, enhancing charge transport and luminescence. This review presents a comprehensive exploration of synthetic strategies aimed at diversifying the structural and electronic landscape of functionalized DBT‐SO2 derivatives. Key methodologies include regioselective bromination, directed lithiation, Suzuki–Miyaura and Stille cross‐coupling reactions, as well as oxidative and reductive modifications to introduce electron‐donating (or) electron‐withdrawing substituents. By fine‐tuning the substitution pattern and conjugation length, a diverse set of DBTO based molecules was synthesized with tailored optical bandgaps and charge transport properties. The resulting compounds exhibit tunable absorption and emission properties, high photostability, and strong fluorescence quantum yields, making them promising candidates for multifunctional applications. Detailed spectroscopic, electrochemical, and thermal characterizations reveal structure–property relationships critical for optimizing performance in organic light‐emitting diodes (OLEDs), organic field‐effect transistors (OFETs), and fluorescence‐based bioimaging platforms. Furthermore, select derivatives demonstrate excellent photostability and biocompatibility, enabling their use as fluorescent probes for cellular imaging. Their strong absorption in the UV–visible region, combined with deep‐blue to green emission and low cytotoxicity, underscores their promise in biomedical applications. Overall, this work provides a systematic framework for the molecular design and functional optimization of DBT‐SO2‐based materials. The results emphasize how structural engineering can be leveraged to unlock multifunctional performance, bridging the fields of organic electronics and bioimaging. These findings open new avenues for the development of high performance, tunable organic materials based on the DBTO scaffold.
Histopathology image analysis faces challenges due to magnification variability, limiting robust tumor categoriza tion. Existing deep learning models prioritize accuracy but neglect explainability, ethical biases, and real-world deployment. This study proposes xMagNet, a hybrid Transformer-Convolutional Neural Network (CNN) frame work that synergizes technical rigor, clinical transparency, and ethical fairness for multi-magnification breast cancer diagnostics. xMagNet integrates a hybrid encoder combining Vision Transformers (ViT) for global tis sue modeling at low magnifications (4x-10x) and Separable Dilation Convolutions (SDC) for localized nuclear texture extraction at high magnifications (20x-40x). Magnification-Aware Gating (MAG) dynamically balances ViT and SDC features via temperature-scaled sigmoid activation. A multi-task decoder employs Thresholded Grad-CAM (top 10% gradients) for explainable decision-making and Point-wise Reformation Blocks (PRB) for boundary preservation. Federated learning (FL) with momentum-enhanced aggregation and Sinkhorn divergence regularization ensures scanner/stain-invariant training across six institutions (Hamamatsu/Leica, H&E/IHC). Uncertainty-quantified predictions (Monte Carlo dropout) and adversarial debiasing mitigate demographic leak age. xMagNet achieves a 97.8% F1-score for tumor segmentation on Camelyon16 and a 93% Gleason AUC on PANDA, with 96.5% pathologist concordance via Grad-CAM. At 40x magnification, it detects micro-metastases with 94% sensitivity (vs. UNet++'s 89% and ResUNet's 91%). Computational efficiency includes sub-second inference (0.42 sec/slide) and 2.3x faster convergence than HoVer-Net. Ethical auditing reveals <3% fairness gaps (OCFG) and a 73% domain shift reduction (MMD: 0.12 vs. FedAvg's 0.45), validated on 15,000 whole-slide images (WSIs) from TCGA-BRCA, Camelyon16, and PANDA datasets. xMagNet bridges critical gaps in multi-magnification histopathology by harmonizing technical robustness (MAG fusion, bounded gradients) with clinical utility (HER2+/ER+ subtyping, Gleason grading) and ethical scalability. By achieving high accuracy, rapid in ference, and equitable deployment, it advances AI-driven diagnostics toward trustworthy, deployable systems for breast, prostate, and metastatic cancer imaging. Code available at: xMagNet.
