The performance of ZnO-based dye-sensitised solar cells (DSSCs) remains significantly limited by poor dye uptake, dye aggregation, and rapid recombination. Here, we demonstrate a solvation-assisted strategy using dodecylbenzene sulfonic acid (DBSA) to regulate the nucleation, dispersion and surface chemistry of ZnO during co-precipitation. Three DBSA-functionalised ZnO photoanodes (DZO1, DZO2, DZO3) were synthesised by varying the thermal treatment (RT, 150 °C, and 400 °C). XRD confirmed the wurtzite ZnO phase, with crystallite size increasing from 18 nm (DZO1) to 32 nm (DZO3). FTIR verified DBSA anchoring via sulfonate–Zn interactions and the modulation of surface hydroxyl groups. Optical analysis revealed a direct band gap reduction from 3.20 eV (DZO1) to 3.13 eV (DZO3), consistent with thermally induced crystallite growth. Dye-sensitised films showed a controlled bathochromic shift (527–533 nm) attributed to varying degrees of J-aggregation. DZO1 exhibited the lowest dye aggregation and moderate dye loading (0.86 nmol/mg), whereas DZO3 showed the highest dye loading (1.30 nmol/mg) but stronger aggregation. The DSSC device based on DZO1 delivered the highest experimental power conversion efficiency of 1.78
Fetal abnormality detection plays a critical role in prenatal care, as early identification of potential risks enables timely intervention and informed decision-making. However, existing approaches are constrained by scarce labeled data for rare conditions, anatomical variability across gestational stages, and inconsistent imaging quality in resource-limited settings. Conventional segmentation models often struggle with accurate multi-region delineation, fine boundary precision, and adaptability to diverse clinical environments. The Federated Biometric UNet++ (FB-SeUNet++) is introduced as a federated learning framework specifically designed for multistructure fetal abnormality detection. Unlike centralized approaches, FB-SeUNet++ enables collaborative training across hospitals and clinics without exposing sensitive patient data. Its core methodology integrates an enhanced U-Net++ architecture, where the encoder employs Diverse Scale Depthwise Convolution (DSDW) and wavelet pooling for multi-scale feature extraction, while the decoder incorporates Vision Eagle Attention to refine segmentation. Segmentation masks of fetal regions such as the head, abdomen, brain, thorax, and femur are subsequently processed using the Curvature-Based Least Squares and Zhang-Suen Thinning algorithms to obtain precise biometric measurements. These measurements are combined with a threshold-based classification strategy to differentiate between normal and abnormal conditions. Experimental evaluation on benchmark datasets demonstrates that FB-SeUNet++ achieves a Dice coefficient of 0.99 and accuracy of 99.4 %, outperforming state-of-the-art methods by 4-6 % across multiple metrics. By uniting multi-structure segmentation, accurate biometric measurement, and federated learning, FB-SeUNet++ provides a robust, generalizable, and privacy-preserving framework that enhances automated prenatal diagnostics and supports smaller clinics with limited data resources.
A mobile ad hoc network (MANET) is a group of wireless mobile nodes that create a network without the assistance of a standard support provider or central administrator. Packet transmission in a MANET is enabled by collecting intermediary equipment between the sender and the recipient. As a result of efficient packet transmission, quality of service is improved, including faster throughput, reduced latency, and assured delivery. However, challenges such as frequent topology changes and unpredictable traffic patterns caused by dynamic node mobility and uneven data distribution can significantly degrade network performance. To overcome these problems, an Enhanced QoS-aware Routing approach using an Optimized Deep Learning framework (ERODE) has been proposed for effective routing and data transmission in MANETs. The ERODE approach integrates Adaptive Neuro-Fuzzy Inference System (ANFIS) for retransmission control and Tyrannosaurus Optimization Algorithm (TOA) for energy-efficient routing. By optimizing parameters including throughput, end-to-end (E2E) delay, Network Lifetime (NL), and packet delivery ratio (PDR), the ERODE method improves network performance and reliability under dynamic conditions. The proposed method achieves an EC of 31.56 J, while the existing AFB-GPSR, FLSTMTLAR, and OFC-TR systems attain 35 J, 40 J, and 45 J, respectively. In terms of PDR, the ERODE technique outperforms AFB-GPSR, FLSTMTLAR and OFC-TR by 5.01
In this study, high-performance hybrid composites were fabricated by incorporating MXene, carbon nanotubes (CNTs), and hexagonal boron nitride (h-BN) into a poly(butylene succinate) (PBS) matrix. The synergistic interaction between two-dimensional MXene sheets, one-dimensional CNT networks, and insulating h-BN platelets enabled the formation of a controlled percolative structure with enhanced interfacial polarization and charge transport. Among all compositions, the PBS/MXene/CNT/h-BN hybrid composite (S10, total filler fraction = 0.20) exhibited the optimum multifunctional performance. As a result, the dielectric constant increased from 3.2 (pure PBS) to 110 at 1 kHz, while maintaining a dielectric loss of 0.31. The AC conductivity improved significantly, reaching 1.6 & times; 10-3 S/m, indicating efficient charge transport pathways. The composites exhibited a maximum energy density of 0.04868 J/cm & sup3; with an efficiency of 85%, demonstrating improved energy storage capability. Notably, the EMI shielding effectiveness increased from 1.5 to 52 dB, with absorption-dominated shielding (96.8%), highlighting superior electromagnetic attenuation performance. Despite increased conductivity, the incorporation of h-BN preserved dielectric stability and improved breakdown strength up to 17 kV/mm. Percolation analysis confirmed a three-dimensional conductive network with a critical exponent of t = 1.9 and a percolation threshold of pc = 0.12, confirming the establishment of an interconnected conductive structure.
A facile and effective hydrothermal method incorporating a Gemini like surfactant with different molar concentrations was developed, and its photocatalytic activity toward Direct Green 6 (DG6) degradation was systematically evaluated for the first time. The crystal phase, optical behaviour, surface morphology, Microstructural analysis, Surface area Porosity analysis, Composition Oxidation states and absorption features were systematically examined using X-ray diffraction (XRD), UV-diffused reflectance spectra (UV–DRS), scanning electron microscopy (SEM), Transmission Electron Microscope (TEM), Brunauer-Emmett-Teller (BET), X-ray Photoelectron Spectroscopy (XPS) and Fourier transform infrared (FT-IR), respectively. The 0.05 M GS-assisted Mn5O8 (Gemini surfactant stabilized Mn5O8) photocatalyst exhibited superior performance compared to the other photocatalysts. The optimized GS-assisted Mn₅O₈ photocatalyst exhibited superior degradation efficiency of 97.6
