College of Business and Technology

The Atlantic Meridional Overturning Circulation (AMOC) plays an important role in regulating the global climate. The AMOC change in response to global warming has important environmental and, potentially, societal impacts but remains an issue with large uncertainty. Here we use a series of coupled climate model experiments to reveal the overlooked role of Atlantic subtropical salinification, a robust consequence of an intensified hydrological cycle, in inhibiting AMOC weakening under global warming. Without subtropical salinification, the AMOC weakening more than doubles in response to a doubling of CO2, primarily driven by a reduced zonal salinity gradient that weakens the geostrophic component of AMOC through the thermal wind relation. This larger AMOC weakening reduces surface warming in the Northern Hemisphere by as much as 1–3 K at northern high latitudes when subtropical salinification is inhibited.

We have investigated the electronic, optical, and mechanical properties of six structures belonging to the Tubulanes-cross-linked carbon nanotube family, using Density Functional Theory (DFT) with the Generalized Gradient Approximation (GGA) and the Perdew–Burke–Ernzerhof (PBE) functional. Our results highlight the remarkable anisotropic mechanical behavior of these materials, distinguishing them from isotropic structures, such as diamond. Notably, the 8-tetra-22 structure has a higher Young’s modulus (YM) along the z-direction compared to diamond. Unlike diamonds, the mechanical properties of Tubulanes are direction-dependent, exhibiting significant variations in Young’s Modulus (2.3 times). Additionally, the Poisson’s ratio is highly anisotropic, with at least one direction exhibiting a value close to zero. The inherent anisotropy of these materials enables tunable mechanical properties that depend on the direction of applied stress. Regarding their electronic properties, all Tubulane structures studied possess indirect electronic band gaps, dominated by 2p orbitals. The band dispersion is relatively high, with band gaps ranging from 0.46 to 2.74 eV, all of which are smaller than that of diamond. Notably, the 16-tetra-22 structure exhibits the smallest bandgap (0.46 eV), making it particularly interesting for electronic applications. Additionally, these structures exhibit porosity, which offers advantages over denser materials, such as diamond. Given recent advances in the synthesis of 3D carbon-based materials, the synthesis of tubulane-like structures is within current technological capabilities.

The rapid development of educational technologies has fostered the integration of adaptive digital platforms into the teaching and learning of English as a foreign language. Tools such as Duolingo, Moodle and Pearson stand out for their use of learning personalisation systems, automated feedback and continuous monitoring of learners’ progress. Within this context, the B1 level of the Common European Framework of Reference for Languages (CEFR) assumes particular relevance, as it represents a critical stage in the consolidation of autonomous communicative competence. The aim of this study was to examine and synthesise the available scientific evidence on the impact of these platforms on learners’ progression towards the B1 level of English through a documentary review of previous studies. The research adopted a qualitative, exploratory and documentary approach, focused on the critical and systematic analysis of peer-reviewed academic literature selected for its theoretical and methodological relevance. The findings indicate that adaptive platforms exert a differentiated influence on language learning, demonstrating greater effectiveness in the development of receptive skills and in the consolidation of basic and intermediate linguistic structures, while progress in productive skills appears less consistent. These outcomes suggest that, although adaptive platforms constitute a relevant pedagogical resource for intermediate-level English learning, their effectiveness depends on their alignment with CEFR descriptors and on their integration with complementary pedagogical strategies.

GreenBid is a web-based agricultural marketplace that connects farmers and buyers through transparent, real-time online auctions. It eliminates middlemen, reduces paperwork, and promotes fair, market-driven pricing. Farmers can list products with images and set reserve prices, while buyers join live auctions from anywhere with instant updates and results. The platform provides role-based access for Admin, Farmer, and Buyer, ensuring security and efficiency. Built with Node.js/Express and a React frontend using MVC + RESTful API architecture, GreenBid employs JWT authentication, bcrypt hashing, and WebSockets for live bid updates. Data is managed in MongoDB, and interactive dashboards display bid history and price trends—enhancing transparency, user experience, and informed decision-making for all participants.

Artificial intelligence (AI) is poised to reshape every sector of the economy. In July 2025, the U.S. government released the “America’s AI Action Plan,” a policy blueprint designed to accelerate innovation, invest in national AI infrastructure, and assert U.S. leadership in international diplomacy and security. Yet as generative and predictive models diffuse into commerce, concerns about fairness, transparency, privacy, and workforce impacts intensify. This paper synthesizes peer-reviewed scholarship, industry surveys, and policy documents to develop an integrative framework for evaluating the responsible use of AI in corporate contexts. With examples from manufacturing, healthcare, and finance, this study illustrates how organizations can align AI initiatives with the AI Action Plan while embedding responsible practices. The proposed framework includes a stepwise algorithm, tables summarizing empirical evidence, and a set of research directions to guide scholars and practitioners in navigating AI-driven transformation.

In light of the pressing concerns regarding the inadequacy of transaction security, efficiency, and transparency within the financial system, this study endeavors to enhance the security of digital economy transactions for small and medium-sized enterprises engaged in cross-border e-commerce through the application of blockchain network topology algorithms. Specifically, the research introduces an innovative approach to classifying abnormal nodes, leveraging a dynamic update algorithm rooted in blockchain network topology. Additionally, it proposes a method for detecting security in digital economy transactions, also grounded in blockchain network topology algorithms. Under the conditions of a total of 60,000 records of real transactions in Bitcoin and Ethereum and a node scale of 100 to 1,000, the experiment uses a combination of cosine and Euclidean distance to calculate the transaction frequency, amount and time series characteristics of nodes and complete clustering. Subsequently, a sliding time window is used to dynamically update the node similarity threshold to identify anomalies. Compared with the three benchmark methods of density clustering, graph convolutional network and autoencoder, the proposed blockchain network topology algorithm has a root mean square error of 0.09, a mean absolute error of 0.09, an anomaly detection accuracy of 8.6%, and a transaction success rate of 1.1%, which is jointly determined by a 1.8-millisecond delay and a throughput of 13.2 transactions per second. All indicators are superior to the benchmark methods. The blockchain network topology algorithm can significantly improve transaction security and system stability, which is of great significance for promoting sustainable economic growth and social stability.

The promise of faster, simpler restorations has driven the rise of bulk-fill composites, but do speed and simplicity come at the cost of longevity and marginal integrity? To compare and rank the clinical performance of bulk-fill and conventional incremental layering restoration techniques in Class II posterior restorations through a network meta-analysis (NMA). Following PRISMA 2020 guidelines, a comprehensive electronic search was conducted across five databases. Twelve randomized clinical trials met the eligibility criteria. Risk of bias was assessed using RoB2, and confidence in evidence was evaluated through CINeMA. Network meta-analyses at 6, 12, and 24 months estimated relative risks for marginal adaptation, fractures, and retention using random-effects models. Across all follow-ups, no statistically significant differences were observed between bulk-fill and incremental techniques. Risk ratios clustered around unity, and P-scores indicated minimal ranking variability, confirming comparable outcomes among bulk-fill subtypes regardless of viscosity or curing mode. Most evidence was rated as low risk of bias with acceptable indirectness. CINeMA indicated predominantly low concerns for indirectness and risk of bias, supporting the robustness and transitivity of the network. The ormocer-based and overlay bulk-fill composites showed the highest, though not statistically superior, ranking probabilities at 12 and 24 months. Across 6-, 12-, and 24-month follow-ups, no statistically significant differences were detected between bulk-fill techniques and incremental placement for the primary outcomes (marginal adaptation, fractures, and retention). Treatment rankings (P-scores) showed minimal separation overall; ormocer-based and overlay bulk-fill configurations tended to occupy higher ranks at 12-24 months, but without statistically confirmed superiority. This systematic review and network meta-analysis provides robust, evidence-based guidance to support clinical decision-making in the restoration of posterior Class II cavities. The findings indicate that bulk-fill resin composites regardless of viscosity or formulation demonstrate clinical performance comparable to conventional incremental layering techniques, while markedly simplifying placement procedures and enhancing operative efficiency. This review was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number CRD420251170292).

Cyber attacks are becoming more sophisticated and happening more often. Traditional honeypot systems like rule-based methods depend on manually defined patterns requiring significant human effort for analysis and struggle to detect novel attacks thus frequently falling short in identifying sophisticated intrusion patterns and ensuring timely responses. To address these limitations, An Al-powered honeypot system is proposed to proactively detect, classify, and analyze cyber-attacks using machine learning techniques and Long Short-Term Memory (LSTM) network algorithms. By simulating common attack vectors such as DDoS, brute-force login attempts, and malware probes, the system collects real-time data which is then analyzed using autoencoder-based feature extraction and Convolutional Neural Networks (CNNs) for multi-class classification of intrusion types. Furthermore, the system automatically sends email notifications in real time to alert relevant cybersecurity personnel upon the detection of malicious activity. All captured attack data is securely stored in a computer database for further analysis and forensic investigation. The Al-driven approach significantly enhances early threat detection, enables faster and more accurate incident response, and provides a valuable repository of attack information for modern network environments.

Background: Functional limitations are common among U.S. adults and are established predictors of cardiovascular disease (CVD). However, it is unclear whether exposure to toxic metals modifies the association between functional limitation and CVD risk. Methods: We analyzed data on 6,337 U.S. adults (≥18 years) from the National Health and Nutrition Examination Survey 2021–2023. Functional status was assessed in three domains: cognition, mobility, and seeing. Toxic metal exposures (blood lead, cadmium, mercury, manganese, selenium) were measured as "high" (≥90th percentile) vs "low." The outcome was CVD, defined as any history of heart failure, coronary heart disease, angina, myocardial infarction, or stroke. Survey-weighted descriptive statistics, cross-tabulations, and survey-weighted, multiple-imputed logistic regression models (adjusted for age, sex, race, education, income, BMI, diet, physical activity, sedentary time, alcohol, and smoking) were used to estimate main and interaction effects. Results: Functional limitations were prevalent: cognitive—47.8%, mobility—32.8%, and seeing—40.3%. CVD prevalence was 12.8%. Each metal was "high" in ~10% of participants in the study population. Across all domains and metals, CVD rates were lowest among those with no functional limitation and low metal exposure (cognition × lead: 8.6%) and highest among those with both limitation and high metal (cognition × high lead: 23.8%; mobility × high lead: 32.8%; seeing × high lead: 26.3%). In multivariable regression, any difficulty in cognition was 1.52 times (aOR for CVD: 1.52, 95% CI: 1.28–1.80), mobility 2.22 times (aOR 2.22, 1.09–4.09), and seeing 1.59 times (aOR 1.59, 1.07–2.44) were independently associated with higher CVD risk, regardless of toxic metal exposure. No significant interaction was observed between high metal exposure and functional limitation. Conclusions: Functional limitation in cognition, mobility, and seeing is a strong, independent predictor of CVD in U.S. adults. Although CVD prevalence was highest among those with high toxic metals exposure and functional limitation, toxic metals exposure did not significantly modify the strength of this association in multivariable analysis in the study population.

In recent years, optical super-resolution microscopy has made significant progress in lateral resolution through the use of fluorescence techniques and optimization of optical hardware systems, approaching theoretical limits. However, its axial resolution is still constrained by physical diffraction limits. Some experimental studies have shown that introducing interferometric techniques can effectively enhance axial resolution. This paper aims to construct a comprehensive theoretical model for interferometric microscopy, analyzing the impact of different near-field reflection substrate parameters on axial resolution. The model will take into account interference effects, light wave propagation characteristics, and the interaction between the sample and the substrate. Through simulation, the mechanism of enhanced axial resolution in interferometric microscopy assisted by near-field reflection planes will be revealed. Additionally, the paper will explore how to optimize near-field parameters to maximize axial resolution, providing a theoretical foundation for the development of optical super-resolution microscopy.