Piezoelectric materials exhibit complex fracture behaviors due to their anisotropic and multi-field coupling properties. While the stress intensity factor (SIFs) and electric displacement intensity factor (EDIF) are critical for assessing propagation risks, obtaining their closed-form solutions remains challenging. Motivated by recent calls for AI-enabled mechanics and handbook-oriented knowledge organization, this study aims to transcend traditional verification by automating the generation of handbook-style closed-form entries for these multiphysics parameters. A novel framework is proposed combining dimension-based dimensionality reduction and element optimization (DDREO) with dimensional homogeneity constrained gene expression programming (DHC-GEP). Rather than simple regression, the problem is formulated as learning a dimensionally consistent, similarity-preserving mapping from a reduced set of dimensionless coordinates to normalized intensity factors. Utilizing a comprehensive database generated via interaction integral (I-integral) and extended finite element methods (XFEM), the framework applies active learning to simplify the original 12 parameters (spanning mass M, length L, time T, and electric current I) to nine parameters. This reduction effectively merges and eliminates the electric current dimension, decreasing data complexity to facilitate the discovery of physically meaningful models. Consequently, three analytical expressions are derived that comprehensively consider key geometric features and force-electric material parameters, such as crack inclination and polarization direction. After domain and stability screening, 8793 unique samples are retained; on the 793-case outer test, the three frozen expressions achieved R2=0.9837-0.9957, NRMSE normalized by the target RMS of 4.41%–6.54%, and 100% finite prediction coverage. The expressions provide reusable handbook entries for rapid parameter sweeps and engineering assessments within the registered two-dimensional central-crack family with electrically impermeable crack faces under coupled electromechanical loading. A separate analytical benchmark of an impermeable crack under general electromechanical in-plane loading recovered the exact load-angle laws for all three intensity factors on 5,940 held-out cases R2=1, supporting the portability of the entry-construction workflow to a broader loading-coordinate set.

The low reliability of CFRP composites under impact loading seriously limits its application fields. In this work, bioinspired sinusoidal structures with a gradient design mimicking mantis shrimp's dactyl club were introduced into CFRP laminates to improve the impact resistance. The effects of structural configurations and impact loading on the ballistic performance and energy absorption mechanism of biomimicking CFRP laminate were studied through transient response monitoring, non-destructive detection, interlayer fracture testing, and finite element analysis. The underlying relationship between the penetration stage and the failure mechanisms was analyzed. The results show that the ballistic limit and energy absorption rate (EAR) of CFRP laminates with sinusoidal-gradient coupling structure were significantly improved. The coupling structure increased the penetration distance of the projectile. By optimizing the wavelength and amplitude of the designed structure, the EAR of the CFRP laminates was greatly enhanced, achieving a balance and synergy between delamination and deformation. The optimized gradient structure not only increases the amount of secondary cracks, the efficiency of stress transfer, and the deflection angle of the projectile, but also increases the interlaminar fracture toughness by 156.2 %, thereby improving the EAR of the laminate by 40.4 % at an impact velocity of 191.8 m/s. Compared with spherical-nosed and conical-nosed projectiles, the CFRP laminate with gradient structure showed a higher EAR when hit by flat-nosed projectiles due to the large deflection and delamination region. The coupling structure with gradient design provides a novel way for the development of impact resistant CFRP composites.

A data-driven framework is proposed that integrates automated geometric modeling and high-throughput finite element simulations to enable both forward prediction and inverse design of lattice unit cells. By establishing an efficient workflow, a substantial numerical simulation database is generated encompassing diverse truss-based topologies, the fidelity of which is corroborated by 3D-printed prototypes and quasi-static compression tests. A specially designed network is trained to predict elastic modulus, demonstrating improved accuracy (MSE = 0.046, R > 0.994) and reduced overfitting compared to graph neural network (GNN). Building upon this predictive model, a conditional variational autoencoder (C-VAE) is introduced that learns a low-dimensional latent space simultaneously conditioned on topological features and mechanical performance. Subsequent dimensionality reduction and clustering analyses, utilizing principal component analysis (PCA) and K-means algorithms, elucidated intrinsic correlations between rod connectivity and structural stiffness. Ultimately, by coupling the C-VAE with the predictive model, a high-throughput inverse design strategy is realized, enabling the fabrication of unit cells that achieve the prescribed elastic modulus with remarkable fidelity (design error < 2 %), and with a computational design time on the order of 10 s. And the design speed is approximately 66 times faster than that of traditional topology optimization methods. This paradigm significantly accelerates the design of 3D-printed architected materials, offering a promising and data-efficient approach for exploring complex structural geometries.

Electrically conductive coordination polymers (ECCPs), particularly those incorporating benzenehexathiol (BHT) ligands, are emerging as a distinctive class of electronic materials with tunable semiconducting and metallic properties. However, the exploration of novel ECCPs with low-symmetry structures and electrical anisotropy remains under development. Here, we report the on-water surface synthesis of a novel ECCP, namely Cu5BHT, which exhibits a low-symmetry structure and unique in-plane electrical anisotropy that differs from the well-known Cu3BHT phase. Utilizing imaging and diffraction techniques, we elucidate the unit cell and crystal structure of Cu5BHT, revealing an asymmetric arrangement of the kagome resembling lattice connected by two different secondary building units: square planar CuS4 and non-planar Cu2S4. Theoretical studies indicate that Cu5BHT is metallic and exhibits in-plane electrical anisotropy due to the structure arranged in interconnected well-conducting CuS4 chains and less-conducting Cu2S4 slabs oriented along single crystal direction. Single-crystal electrical measurements confirm a metallic character characterized by the increase of conductance upon cooling. Notably, the measured conductance along different crystal directions within the ab plane unambiguously reveals a significant anisotropy, with an anisotropic factor reaching ~8. This work demonstrates a novel low-symmetry ECCP and highlights its potential for achieving in-plane electrical anisotropy.

Background Rhodococcus is an important genus of soil bacteria known for its metabolic diversity and environmental adaptability under harsh and contaminated conditions. However, few studies have reported on the selenium metabolism of Rhodococcus species. Results Here, we isolated a highly selenite-resistance strain PM1 (up to 100 mM) from a selenium-rich mine in Enshi City. This strain reduced 50 mM sodium selenite by 99 % within 72 h. SEM and XPS revealed that PM1 reduced selenite to selenium nanorods (SeNRs). Phylogenetic analysis identified PM1 as R. qingshengii. The whole genome of strain PM1 was sequenced, and a comparative genome analysis of strain PM1with 64 other genomes of Rhodococcus was performed. Whole genome sequencing identified a total of 97 heavy metal resistance genes in strain PM1. Comparative genomics revealed that Rhodococcus species possess an open pan-genome, indicating adaptability to diverse environments. Genomic analysis revealed a total of 96 putative selenite-reducing proteins in strain PM1. Four gene clusters, involved in the pentose phosphate pathway, iron-sulfur cluster assembly, sulfate reductase cluster, and sulfate transport complex, showed high conservation of sequence identity within these species. Conclusions To our knowledge, this research enhances our understanding of high selenite reduction in strain PM1 at genomic level and elucidates the biotechnological applications of selenite-reducing bacteria in environmental remediation.

Rhodococcus species are renowned for their metabolic diversity and environmental adaptability, yet their selenium metabolism remains insufficiently studied. In our previously work, we isolated a highly selenite-tolerant strain, Rhodococcus qingshengii PM1, from selenium-rich soils in Enshi, China. To reveal the reduction mechanism of sodium selenite, integrated transcriptomic and metabolomic analyses were conducted. Biochemical assays confirmed that Se exposure induced pronounced oxidative stress in strain PM1 and elicited strong induction of the antioxidant defenses. A total of 308 differential metabolites were detected, with bioactive compounds, organic acids, lipids, secondary metabolites and organoheterocyclic compounds. A total of 1,511 differentially expressed genes were identified. These changes were primarily associated with sulfite reductase complex genes (CysNDHIJ), Fe–S cluster biosynthesis genes (SufBCDSE), glutathione metabolism, lipid remodeling, redox metabolic pathways and antioxidant pathways, all contributing to the detoxification and reduction of selenite. Notably, metabolites such as prostaglandin D3 were upregulated, reflecting lipid signaling in response to selenium, while others including physangulide, enhydrin, and sebacic acid were downregulated, indicating a metabolic shift away from lipid biosynthesis and secondary metabolism. These findings elucidate the molecular mechanisms underlying microbial selenite detoxification and highlight R. qingshengii PM1 as a promising candidate for bioremediation of selenium-contaminated environments.

Dynamic intensity factors (IFs) provide an important parameter for assessing the failure risk of an interfacial crack for piezoelectric composites exposed to electromechanical impact loadings. In the present research, a new dynamic interaction integral (I-integral) is built for computing the dynamic stress IFs (SIFs) and electric displacement IF (EDIF) of an interfacial crack located in dissimilar inhomogeneous piezoelectric media. Through proper selection of auxiliary variables, the domain expression of I-integral for the inhomogeneous piezoelectric bi-materials does not need to take into account any derivative term for any piezoelectric material property. Further, after rigorous theoretical derivation, the resulting I-integral is still valid for complex models where the integration domain contains other multiple interfaces, regardless of whether the extra materials are straight or curved. Incorporating the modified extended finite element method, the accuracy is confirmed by checking the dynamic IFs extracted from the I-integral with referenced results. The domain-independence is tested by varying ranges of integration domains for inhomogeneous piezoelectric bi-materials and multi-interface piezoelectric composites. Finally, typical examples are employed to discuss the influences of the direction and magnitude of electric impact loading, combination of polarizations, inhomogeneous degree of piezoelectric bi-materials and complex distribution of diverse material properties on the dynamic IFs.

The mining industry in China plays a pivotal role in economic development but also leads to severe environmental issues, particularly heavy metal pollution in soils. Heavy metal pollution significantly impacts soil microbial communities due to its persistence and long-term residual effects. We assessed changes in microbial diversity, community structure, and assembly mechanisms in selenium-impacted soils. This study investigates the impacts of selenium (Se) and other heavy metals on soil microbial communities in selenium-rich mining areas using full-length 16S rRNA gene sequencing. Our results showed that Se and other heavy metal contamination significantly altered microbial community composition, favoring metal-tolerant phyla such as Proteobacteria, Actinobacteriota and Firmicutes, while reducing the abundance of sensitive groups like Acidobacteriota and Chloroflexi. Microbial diversity decreased as Se and other heavy metal concentrations increased. Mantel test analysis revealed that soil total potassium (TK), soil organic carbon, total nitrogen, and several other metals, including zinc, niobium, titanium (Ti), manganese, rubidium, barium, potassium, cobalt, gallium (Ga), Se, chromium (Cr), vanadium, and copper were significantly and positively correlated with microbial community composition across all soil samples. Random forest analysis showed that soil TK and multiple elements [Cr, Ti, nickel (Ni), Ga and Se] were the most important predictors of bacterial diversity, emphasizing the role of multiple elements in shaping microbial communities. Co-occurrence network analysis revealed that Se and other heavy metal contamination reduced network complexity and stability, with high Se-contaminated soils exhibiting fragmented microbial networks. Community assembly was primarily driven by drift in control soils, whereas dispersal limitation became more prominent in Se-contaminated soils due to heavy metal toxicity. These findings highlight the ecological consequences of heavy metal pollution on microbial communities and offer valuable insights for effective soil management and remediation strategies.

As large language models continue to grow in size, parameter-efficient fine-tuning (PEFT) has become increasingly crucial. While low-rank adaptation (LoRA) offers a solution through low-rank updates, its static rank allocation may yield suboptimal results. Adaptive low-rank adaptation (AdaLoRA) improves this with dynamic allocation but remains sensitive to initial and target rank configurations. We introduce AROMA, a framework that automatically constructs layer-specific updates by iteratively building up rank-one components with very few trainable parameters that gradually diminish to zero. Unlike existing methods that employ rank reduction mechanisms, AROMA introduces a dual-loop architecture for rank growth. The inner loop extracts information from each rank-one subspace, while the outer loop determines the number of rank-one subspaces, i.e., the optimal rank. We reset optimizer states to maintain subspace independence. AROMA significantly reduces parameters compared to LoRA and AdaLoRA while achieving superior performance on natural language understanding and generation, commonsense reasoning, offering new insights into adaptive PEFT.

Understanding charge transport properties of large-area single-layer 2D materials is crucial for the future development of novel optoelectronic devices. In this work, the synthesis and electrical characterization of large-area single-layers of Cu3BHT 2D conjugated coordination polymers are reported. The Cu3BHT are synthesized on the water surface by the Langmuir-Blodgett method and then transferred to SiO2/Si substrates with pre-patterned electrical contacts. Electrical measurements revealed ohmic responses across areas up to approximate to 1 cm2, with a mean resistance of approximately 53 +/- 3 k ohm at a probe separation of 50 mu m. Cooling and heating cycles show hysteresis in the electrical response, suggesting different current pathways are formed as the samples underwent structural-chemical changes during temperature sweeps. This hysteresis vanished after several cycles and the conductivity shows a stable exponential behavior as a function of temperature, suggesting that a temperature-dependent tunneling process is governing the conduction mechanism in the analyzed polycrystalline single-layer Cu3BHT samples. These results, together with density functional theory calculations and valence band X-ray photoelectron spectroscopy data suggest that the single-layer samples exhibit a semiconducting rather than a metallic behavior.

Soil microbial communities are particularly sensitive to selenium contamination, which has seriously affected the stability of soil ecological environment and function. In this study, we applied high-throughput 16S rRNA gene sequencing to examine the effects of low and high doses of sodium selenite and the selenite-degrading bacterium, Rhodococcus qingshengii PM1, on soil bacterial community composition, diversity, and assembly processes under controlled laboratory conditions. Our results indicated that sodium selenite and strain PM1 were key predictors of bacterial community structure in selenium-contaminated soils. Exposure to sodium selenite initially led to reductions in microbial diversity and a shift in dominant bacterial groups, particularly an increase in Actinobacteria and a decrease in Acidobacteria. Sodium selenite significantly reduced microbial diversity and simplified co-occurrence networks, whereas inoculation with strain PM1 partially reversed these effects by enhancing community complexity. Ecological modeling, including the normalized stochasticity ratio (NST) and Sloan’s neutral community model (NCM), suggested that stochastic processes predominated in the assembly of bacterial communities under selenium stress. Null model analysis further revealed that heterogeneous selection and drift were primary drivers of community turnover, with PM1 inoculation promoting species dispersal and buffering against the negative impacts of selenium. These findings shed light on microbial community assembly mechanisms under selenium contamination and highlight the potential of strain PM1 for the bioremediation of selenium-affected soils.

To explore and compare the failure modes, deformation behaviors, and load-bearing capacities of single-edge notched (SEN) beams strengthened with carbon fiber-reinforced polymer (CFRP) and steel bars, static and dynamic three-point bending tests on both types of concrete beams have been carried out in this study. During the static tests, the electro-hydraulic servo machine served as a loading device to apply pressure to CFRP beams and reinforced concrete (RC) beams. During the impact experiments, different impact velocities were imparted by adjusting the drop hammer's height. Thus, information regarding crack propagation, energy absorption, and deformation was obtained. The results from the static tests showed that the RC beams predominantly experienced shear failure. In contrast, the CFRP beams primarily exhibited bending-shear failure, attributed to the relatively weaker bond strength between the bars and the concrete. Impact tests were conducted at three different velocities in this study. As the impact velocity increased, both types of concrete beams transitioned from bending failure to bending-shear failure. At the lowest velocity, the difference in energy absorption between beams reinforced with different materials was insignificant during the bending process. However, at the highest velocity, CFRP beams absorbed less energy than RC beams. The study of structures' impact failure modes and their mechanical characteristics offers valuable references for the anti-collision design and protection of structures.

This study employs a data-driven methodology that embeds the principle of dimensional invariance into an artificial neural network to automatically identify dominant dimensionless quantities in the penetration of rod projectiles into semi-infinite metal targets from experimental measurements. The derived mathematical expressions of dimensionless quantities are simplified by the examination of the exponent matrix and coupling relationships between feature variables. As a physics-based dimension reduction methodology, this way reduces high-dimensional parameter spaces to descriptions involving only a few physically interpretable dimensionless quantities in penetrating cases. Then the relative importance of various dimensionless feature variables on the penetration efficiencies for four impacting conditions is evaluated through feature selection engineering. The results indicate that the selected critical dimensionless feature variables by this synergistic method, without referring to the complex theoretical equations and aiding in the detailed knowledge of penetration mechanics, are in accordance with those reported in the reference. Lastly, the determined dimensionless quantities can be efficiently applied to conduct semi-empirical analysis for the specific penetrating case, and the reliability of regression functions is validated.

M-estmators including the Welsch and Cauchy have been widely adopted for robustness against outliers, but they also down-weigh the uncontaminated data. To address this issue, we devise a framework to generate a class of nonconvex functions which only down-weigh outlier-corrupted observations. Our framework is then applied to the Welsch, Cauchy and lp-norm functions to produce the corresponding robust loss functions. Targeting on the application of robust matrix completion, efficient algorithms based on these functions are developed and their convergence is analyzed. Finally, extensive numerical results demonstrate that the proposed methods are superior to the competitors in terms of recovery accuracy and runtime.

Fatty acids (FAs) participate in extensive physiological activities such as energy metabolism, transcriptional control, and cell signaling. In bacteria, FAs are degraded and utilized through various metabolic pathways, including β-oxidation. Over the past ten years, significant progress has been made in studying FA oxidation in bacteria, particularly in E. coli, where the processes and roles of FA β-oxidation have been comprehensively elucidated. Here, we provide an update on the new research achievements in FAs β-oxidation in bacteria. Using Xanthomonas as an example, we introduce the oxidation process and regulation mechanism of the DSF-family quorum sensing signal. Based on current findings, we propose the specific enzymes required for β-oxidation of several specific FAs. Finally, we discuss the future outlook on scientific issues that remain to be addressed. This paper supplies theoretical guidance for further study of the FA β-oxidation pathway with particular emphasis on its connection to the pathogenicity mechanisms of bacteria.

Applying half-quadratic optimization to loss functions can yield the corresponding regularizers, while these regularizers are usually not sparsity-inducing regularizers (SIRs). To solve this problem, we devise a framework to generate an SIR with closed-form proximity operator. Besides, we specify our framework using several commonly-used loss functions, and produce the corresponding SIRs, which are then adopted as nonconvex rank surrogates for low-rank matrix completion. Furthermore, algorithms based on the alternating direction method of multipliers are developed. Extensive numerical results show the effectiveness of our methods in terms of recovery performance and runtime.