Underwater superhydrophobic surfaces (SHSs) are promising for drag reduction but are limited by the metastability of underwater air plastrons. Inspired by the replenishment mechanism of the water spider’s air plastron, a recoverable SHS with a sparse micro-nano hierarchical structure was fabricated by obtaining a microcone array structure through laser etching and mold replication, followed by the chained nanoparticle deposition and fluorination treatment. The surface exhibited exceptional durability, sustaining 100 abrasion cycles (2 kPa) and 5.5 h water jetting (25 kPa). This is attributed to the microcone array, which protects the superhydrophobic nanoparticles under mechanical action. The hierarchical structure exhibited underwater aerophilicity, which enabled 34 cycles from fully wetting to superhydrophobicity. Rotational rheometric analysis revealed a drag reduction rate of 13.6%. Fluid dynamics simulations showed that the microstructure reduced wall shear stress, achieving a drag reduction rate of 14.2% at a flow velocity of 1 m/s and a wall-slip velocity of 0.15 m/s. This study provides design directions and theoretical foundations for the application of SHS in sustainable drag reduction.

In order to reduce dependence on petroleum resources, research on the application of polysaccharide-based materials has become one of the mainstream trends. The chitin nanofibril (ChNF) is a promising building block for fabricating chitin films from aqueous dispersions. In this study, tannic acid (TA) was introduced into the deacidified ChNF film as a cross-linking agent and functional additive. The structure and performance of the deacidified ChNF film were evaluated and its potentials in fruit packaging were explored. The cross-sectional morphology of the film was characterized by a scanning electron microscope, which revealed that the film was densified by deacidification, thereby improving the water resistance. The addition of TA further improved the film’s wet-state mechanical property, antioxidant property, UV-shielding property and antibacterial activity, while the film preserved good optical properties. Specifically, the tensile strength of the film with TA reached (145.1 ± 9.3) MPa, and its free-radical scavenging rates were over 95%. Moreover, preliminary experiments demonstrated the potential of the film with TA for strawberry preservation.

The performance of anion exchange membrane fuel cells (AEMFCs) is severely constrained by the low OH^-conductivity of anion-conductive polymers. Although increasing the ion exchange capacity of these polymers through microstructural design effectively improves the OH^-conductivity, it often compromises the mechanical strength. To address this issue, we report enhanced microphase-separated structures in poly(styrene-b-(ethylene-co-butylene)-b-styrene) (SEBS)-based anion-conductive polymers, achieved through the synergy of hydrophilic quaternary ammonium (QA) groups and hydrophobic fluorinated side chains. Specifically, by precisely tuning the fluorine grafting degree of the polymer side chains, highly interconnected nanoscale ion-conducting domains are created, forming a three-dimensional (3D) pathway for efficient ion transport in anion exchange membranes (AEMs). Additionally, the mechanical stability of AEMs is strengthened by minimizing swelling. As a result, the QA- and fluorine-grafted AEM with a molar proportion of 4-fluorophenethylamine-modified blocks to styrene blocks of 30% (denoted as QSEBS-FPh 30) achieves a high OH^-conductivity of 100.86 mS/cm at 80 ℃ and a moderate tensile strength of 19.89 MPa in a fully hydrated state. The AEMFC utilizing QSEBS-FPh 30 exhibits a peak power density of 204.31 mW/cm²at a current density of 737.29 mA/cm²and 80 ℃, which is 1.4 times that of QA-grafted SEBS (QSEBS). These findings underscore the significant role of microphase separation coupled with maximized ionic domain connectivity in enhancing the OH^-conductivity of anion-conductive polymers, offering valuable insights for the rational design of high-performance AEMs.

Silicon oxycarbide (SiOC) has emerged as a promising candidate for anode materials in lithium-ion batteries (LIBs) due to its unique properties. However, the low electrical conductivity of SiOC limits its practical applications. Herein, a novel necklace-like nanostructure was fabricated through the sol-gel method by embedding SiOC nanospheres in carbon nanotubes (CNTs) to form a stable electronic pathway within the structure. Controlling the hexadecyl trimethyl ammonium bromide (CTAB) content through CTAB removal and centrifugation to separate SiOC and CNTs @ phenylene-bridged mesoporous organosilica (CNTs @ PBMO) before the calcination is important to achieve the necklace-like CNTs @ SiOC nanostructure. The fabricated porous necklace-like CNTs@SiOC nanostructure not only improved the electrical conductivity but also ensured full utilization of SiOC active sites during discharging/ charging along with accelerating ion penetration. As a result, it provided a remarkably long cycling life of up to 1 400 cycles, retaining a specific capacity of 178 mA · h/ g.

A heterogeneous diimine-Ni catalyst (diimine-Ni@PCN-701) is synthesized through sequential ligand insertion into PCN-700, a robust Zr-based metal-organic framework (MOF), followed by chelation of Ni centers. This site-specific assembly affords a robust active center for propylene dimerization with superior activity, which arises from the electron-withdrawing effect of Zr₆O₄(OH)₄ clusters on the Ni active sites. Propylene is selectively converted into a series of dimers with 4-methyl-2-pentene (4M2P) as the main product, which may be due to the MOF structure promoting this reaction through both electronic effects and steric hindrance. This work highlights a general strategy for developing heterogeneous olefin oligomerization catalysts with high activity and shape selectivity through the precise construction of active sites.

Medium-entropy transition metal carbide ceramics and their composites are important materials for extreme environments. However, they face challenges in densification difficulties and low fracture toughness. To address this issue, a medium-entropy ceramic powder with a nominal composition of (Zr_1/3Hf_1/3W_1/3)C (MEC) was synthesized by carbothermal reduction. Then, silicon (Si) was added, and the reactive spark plasma sintering (RSPS) technology was used to prepare MEC-based composite ceramics. The results show that MEC ceramics without Si addition can only form a single-phase solid solution when sintered at 1 900 ℃, with a relative density of 96.7% and an average grain size of (3.8 ± 1.3) μm. By adding 5% Si by mass, the MEC-based composite ceramic with a relative density of 98.8% can be obtained by sintering at a relatively low temperature of 1 700 ℃. It is composed of an MEC solid solution phase and an in-situ-formed silicon carbide (SiC) secondary phase, and the average grain size of the MEC matrix phase is only (1.4 ± 0.3) μm. Compared with the single-phase MEC ceramic, the fracture toughness of the MEC-based composite ceramic prepared by adding Si increases by 24.8% to (3.42 ± 0.24) MPa·m ^1/2. It is considered that Si, with a low melting point, can form a liquid phase during sintering, promoting densification. Additionally, the reaction between Si and carbide results in the formation of carbon vacancies in the latter, which can accelerate atomic diffusion and promote the formation of the MEC solid solution phase. In addition, the in-situ-formed SiC not only pins the grain boundaries of the matrix phase to hinder its growth, but also causes crack deflection to improve the fracture toughness of materials. The research results can provide references for the design and preparation of novel medium-entropy carbide-based ceramics for extreme environments.

To enhance the interfacial properties of carbon fiber (CF)-reinforced polyamide 6 (CF/PA6) composites, a novel blocked isocyanate (BI) / polyurethane (PU) sizing agent was prepared by dispersing BI in an aqueous PU emulsion. It was then applied to the CF surfaces for modification via a room-temperature sizing method. The results indicated that the BI/PU sizing layer significantly enhanced CF surface wettability, facilitating impregnation of PA6 resin onto fibers. With the increase in BI content, the interfacial shear strength (IFSS) of CF/PA6 composites was enhanced. The IFSS of the desized CF/PA6 composites was 46.8 MPa. When the sizing agent with a BI mass fraction of 3.0% was applied, the IFSS of the CF/PA6 composite reached 62.3 MPa, which represented an increase of 33.1% compared with that of the desized CF/PA6 composite. The proposed method offers an efficient and scalable surface modification strategy for high-performance thermoplastic composites.

Wool fabric is prone to felting under hot and humid conditions due to the scale layer structure, impairing its appearance and durability. A novel Fenton-like oxidation system for anti-felting treatment of wool fabric was proposed in this study. The composite catalyst, prepared by loading Fe₃O₄ onto chitosan (CS) and hydroxyapatite (HA), denoted as CS/HA/Fe₃O₄, was used to catalyze H₂O₂ to generate free radicals, thereby reducing felting shrinkage by destroying the scale structure of wool fibers. The optimal reaction conditions were determined as CS/HA/Fe₃O₄ 4.5% on mass of fabric (omf), H₂O₂ 30.0% omf, 1-hydroxyethylidene-1,1-diphosphonic acid (HEDP) 18.0% omf, and Na₂CO₃ 5.0% omf. Fourier transform infrared (FTIR), scanning electron microscopy (SEM), and water contact angle (WCA) characterizations confirmed that the Fenton-like treatment effectively cleaved wool disulfide bonds, stripped surface scales, and reduced hydrophobicity. Preliminary reusability tests of the CS/HA/Fe₃O₄ composite catalyst showed that after three cycles, the felting shrinkage rate of wool fabric slightly increased from 4.0% to 4.7%, which still met the international standard (≤5.0%).

Oxygen solubility in perfluorocarbon (PFC) emulsions is a crucial performance metric in fields such as artificial blood, oxygen transportation and diagnostic imaging agents. However, there is a lack of simple and accurate detection methods. In this study, the perfluorooctyl bromide (PFOB) emulsion was prepared by using perfluorohexylethyl polyoxyethylene ether (FEO) as the emulsifier, and a modified enzymatic method was developed by using a more stable Orange G as the color indicator to determine the oxygen solubility in the PFOB emulsion at 298.15 K and atmospheric pressure. For 0.2 mL emulsion, the oxygen solubility measured by the modified enzymatic method was close to that measured by gas-liquid chromatography as reported in the literature. The effects of mass fractions of PFC and FEO in the emulsions on the oxygen solubility were studied based on the modified enzymatic method. The results indicate that the oxygen solubility in the emulsion increases with the increasing mass fraction of PFC, independent of the mass fraction of the emulsifier.

A semi-supervised learning framework integrating rotational invariance, contrastive learning, and adaptive hybrid thresholds, named rotational contrastive network (RoCoNet), is proposed to enhance the applicability of semi-supervised learning for medical cell datasets. Due to the unique sampling approach of cell datasets, input images often contain uncertain rotation angles, which render traditional convolution kernels ineffective in existing semi-supervised detectors. To address this challenge, rotational attention convolution is introduced, offering robustness to rotational transformations. Additionally, cross-feature contrastive loss is proposed to improve upon the contrastive loss used in supervised learning, tackling issues of poor classification performance caused by cell overlap and clustering. An adaptive hybrid threshold is also introduced to stabilize pseudo-label generation during early training. A global threshold, computed by using Gaussian mixture models (GMMs), is applied to refine the local threshold, which helps balance the quantity and quality of pseudo-labels. Experiments on the ThinPrep cytology test (TCT) dataset for cervical cytopathology show that RoCoNet achieves a mean average precision (mAP) of 31.6% with only 10% labeled data, outperforming the baseline method by 8.4% in mAP.

Video colorization is an important technique to breathe life back into old movies. While current colorization methods work well on still images and low-motion video data, they often struggle with complex dynamic scenes. To address this problem, this study proposes ColorAlignNet, a reference-based video colorization network with temporal aggregation. The network uses source-reference attention to propagate color information from reference frames to grayscale frames, guaranteeing color accuracy, and uses deformable convolution to align features of adjacent frames to enhance temporal consistency. Finally, we use the cyclic transformer module to reconstruct the final prediction results. Extensive experimental results demonstrate that ColorAlignNet achieves excellent performance on the DAVIS and Videvo datasets, outperforming other state-of-the-art methods on both the learned perceptual image patch similarity (LPIPS) and color distribution consistency (CDC) metrics.

Image colorization has attracted considerable research interest over the past few decades. However, current methodologies frequently struggle with limited local colorization flexibility and produce unnatural color outputs, primarily due to the absence of comprehensive understanding of color perception. In this work, we propose an expressive diffusion network (EDN) that leverages a robust diffusion network to significantly enhance both colorization accuracy and diversity. The EDN consists of two main components: a pre-trained latent diffusion model and a perceptual luminance model based on VQ-Diffusion. These components work together to generate rich and vibrant colors while maintaining high fidelity to the structural features of the original grayscale image. The EDN incorporates controllable creative diffusion (CCD) to direct the color generation process toward more realistic outcomes. Extensive experiments demonstrate that the EDN outperforms existing methods in perceptual quality, offering notable improvements in visual realism and vibrancy across various scenes. The proposed EDN showcases significant improvements over ChromaGAN and InstColor, confirming its robustness in both simple and complex scenarios.

To improve the accuracy of detecting prohibited items in X-ray images, this study proposes a wavelet-guided multi-feature fusion module (WaveMFFM), an easy-to-integrate, plug-and-play module that can be seamlessly incorporated into existing detectors. WaveMFFM innovatively introduces the wavelet transform and pioneers the de-occlusion wavelet convolution (DOWC) structure, which dynamically integrates low-frequency global contour information and high-frequency detailed texture features through a frequency-domain decoupling mechanism. This approach effectively resolves the feature confusion issue inherent in conventional convolutional operations under occlusion scenarios, achieving a groundbreaking synergistic enhancement between edge features and region-specific deep features. Consequently, the proposed method significantly improves the discriminative power of detection features. Extensive experiments on YOLOv8, ViT, and SSD detectors demonstrate that WaveMFFM effectively mitigates occlusion problems, thus improving the prohibited item detection performance of these representative methods.

To accurately and quickly segment the line laser stripes under interferences of strong noise and strong reflection, a lightweight weld line laser stripe segmentation network based on the DeepLabv3+ network, named WLS-Net, is proposed. To improve the segmentation speed of the network, the shallow residual network ResNet-18 is selected as the backbone network. Combining the multi-Dconv head transposed attention (MDTA) and the convolutional block attention module (CBAM), the multi-dimensional CBAM (MD-CBAM) is proposed, and the dynamic upsampling method (DySample) is chosen to replace the traditional bilinear interpolation to improve the segmentation accuracy. To address the foreground-background class imbalance in the weld line laser stripe image, the sum function of the Dice loss function (Dice Loss) and the pixel-wise cross-entropy loss function is chosen as the loss function of the model. The experimental results show that compared with the original DeepLabv3+ network, the WLS-Net achieves absolute improvements of 6.00% in IoU, 5.67% in precision, and 3.76% in F1 score on the weld line laser dataset, and the inference time of a single image is reduced by 18 ms. Compared with other semantic segmentation networks, the WLS-Net also achieves better segmentation effect and higher segmentation rate.

In relational database normalization theory, identifying all keys and prime attributes is essential yet highly challenging. It has been shown that the prime attribute problem is nondeterministic polynomial-time complete (NP-complete). Additionally, the maximum number of keys in a relation scheme is exponential in the number of attributes. These conclusions hold when the normal form is unknown, and when the normal form is known they may change. For instance, when a relation scheme is in Boyce-Codd normal form (BCNF), the prime attribute problem falls within the polynomial-time complexity class (P-class), and listing all keys becomes less cumbersome. In the realm of normalization, second normal form (2NF) serves as a foundational stage, and any scheme that is in BCNF or third normal form (3NF) inherently satisfies the requirements of 2NF. Therefore, this paper focuses on the problems related to the prime attribute and all keys for 2NF. First, we present a necessary condition and a sufficient condition for a relation scheme to be in 2NF. Then, we demonstrate that the maximum number of keys is exponential in the number of attributes and functional dependencies of a relation scheme in 2NF. Furthermore, we propose an algorithm for finding all keys of a relation scheme in 2NF. Finally, we demonstrate that the prime attribute problem remains NP-complete in 2NF. This study deepens the theoretical understanding of the computational complexity inherent in 2NF normalization and provides practical references for database scheme design.