Browse Articles

Symmetry-breaking thermomagnetic dynamics of ferrofluid: Oscillation, spontaneous rotation, and flow bistability under laser irradiation

Fluidic oscillation and flow bistability—classical signatures of nonlinear fluid dynamics—typically occur at high Reynolds numbers, whereas the Rosensweig instability, manifested as spike formation on a ferrofluid surface under a magnetic field, reflects a static nonlinear phenomenon. Here, we report oscillation and bistable rotation of ferrofluid spikes under continuous-wave laser excitation. Enabled by Marangoni instability, local laser heating causes a single spike to vanish or to oscillate around the laser spot like a pendulum. For two or more spikes, the pattern undergoes steady clockwise or counterclockwise rotation, depending on the laser position. A brief puff of air or a gentle drag with a tip in the opposite direction can reverse the rotation direction. These behaviors arise from symmetry breaking and asymmetric thermomagnetic forces. Single spike oscillation results from the breaking of axial symmetry of the magnetic field, whereas multispike rotation occurs even in a perfectly axisymmetric field through spontaneous symmetry breaking of the coupled magnetic-field–ferrofluid–laser system. The two rotation directions constitute bistable states separated by an energy barrier, analogous to deformable mechanical systems that switch states under external perturbations. Our findings provide a simple, reconfigurable platform for exploring nonlinear fluid dynamics at low Reynolds numbers and open opportunities in optofluidics and soft robotics.

Solar PVT-assisted high-temperature CO2 heat pump with thermal storage: Design for net-zero electricity operation and ethical sustainability

To build a clean, low-carbon and efficient energy utilization system, solar photovoltaic–thermal (PVT) assisted heat pump technologies are promising for supplying high-temperature heat in industrial sectors and accelerating the green transition of the energy system. By targeting net-zero daily electricity use and reduced reliance on fossil fuels, such systems can also support ethically sustainable industrial process heat. In this study, a solar PVT-assisted high-temperature CO 2 heat pump system with a hot-water storage tank is proposed and analyzed. The PVT field simultaneously generates electricity to drive the compressor and recovers photovoltaic waste heat to charge the thermal storage, which serves as the low-temperature heat source of a transcritical CO 2 heat pump. This work presents a dynamic co-design of an integrated PVT-storage-heat-pump system to achieve net-zero daily electricity operation, with explicit consideration of ethical sustainability in industrial heat supply. A dynamic mathematical model is developed and applied to a typical meteorological day in Xi’an, China, to provide 10 kW continuous heat above 100 °C over 24 h. The effects of the number of PVT modules, storage tank volume and initial storage temperature on system behavior are investigated. Results show that at least 32 PVT modules are required to guarantee 24 h stable operation, with the storage temperature and heat pump COP (Coefficient of performance) exhibiting a decrease–increase–decrease trend throughout the day; as the initial storage temperature increases from 25 °C to 50 °C, the average COP rises from 3.59 to 4.37. For a given number of PVT modules, the net daily electrical output increases with storage volume and significantly increases with the initial storage temperature, varying from −6.06 kWh to 2.32 kWh within the investigated range, and reaching its maximum at a storage volume of 5 m³ and an initial temperature of 50 °C. A storage volume of 4.4 m³ and an initial temperature of 42 °C yield a net daily electrical output of approximately zero, i.e., the electricity generated by the PVT field fully compensates the compressor consumption. Under this self-sufficient condition, the maximum instantaneous electrical efficiency gain of the PVT modules compared with a standalone PV array is 5.03%, while the maximum hourly PVT electrical output and net electrical output reach 5.91 kWh and 3.43 kWh, respectively.

Hybrid spatial-field attention network for meteorological data downscaling

Correction for Craig et al., Lipid residue analysis reveals divergent culinary practices in Japan and Korea at the dawn of intensive agriculture

Healthcare providers’ beliefs about the health effects of nicotine and electronic cigarettes

Background Previous research has shown that a significant proportion of healthcare providers hold inaccurate beliefs about the direct health risks of nicotine. However, the extent to which these misconceptions impact beliefs about the health effects of electronic cigarettes (e-cigarettes) remains unclear. This study examines the relationship between beliefs about nicotine and e-cigarettes among healthcare providers and trainees. Methods A questionnaire was distributed via email and printed flyers to healthcare providers and medical trainees at an academic medical center. Participants rated their beliefs about the health effects of nicotine harms on a Likert scale for 3 statements: “Nicotine is the substance that... 1) makes people want to smoke, 2) causes most of the cancer related to tobacco 3) causes most cardiovascular disease (CVD) associated with smoking.” Participants were also asked, “Compared to smoking cigarettes, would you say that electronic cigarettes are...” (less harmful, just as harmful, more harmful). Means and frequencies were tabulated, and logistic regression identified variables associated with believing e-cigarettes to be less harmful than combustible cigarettes. Results Participants (n = 598) were 75.8% female with a mean age of 36.4 years (SD = 13.3). The distribution was as follows: 38.7% (n = 232) were physicians, PAs, or NPs, 27.6% (n = 165) were students, 24.4% (n = 146) were RNs, and 9.2% (n = 55) were RTs. Although 91.5% correctly identified nicotine as the chemical that makes people want to smoke, 25.9% and 42.8% incorrectly believed nicotine is the cause of cancer and CVD associated with cigarette smoking, respectively. Only 21.4% identified e-cigarettes as less harmful than cigarettes. Those believing e-cigarettes to be less harmful than combustible cigarettes were more likely to be male (OR=2.11, 95% CI 1.32–3.37), a student (OR=1.88, 95% CI 1.20–2.94), and to disagree that nicotine is the main substance that causes cancer (OR=1.39, 95% CI 1.03–1.88) or CVD (OR=1.62, 95% CI 1.23–2.13). Conclusions Inaccurate beliefs regarding nicotine harms persist among healthcare providers and are associated with beliefs about e-cigarette harms. Targeted education on the distinct risks of nicotine, e-cigarettes, and combustible tobacco products is crucial to improving understanding and to support evidence-based counseling on harm reduction strategies.

Development and characterization of a neutron calibration field at the Isfahan MNSR

Mechanistic links between coexistence, productivity, and stability in experimental grasslands

The escalating biodiversity crisis underscores the urgent need for a unified framework that links the mechanisms maintaining biodiversity to its functional consequences. However, studies of species coexistence and biodiversity effects on ecosystem functioning have largely progressed independently. Here, using long-term data from five grassland biodiversity experiments, we quantified “coexistence potential” (i.e., the degree to which niche differences exceed fitness differences) and tested its relationships with biodiversity effects on both ecosystem productivity (via complementarity and selection effects) and stability (via species asynchrony and species stability). We found that the relationships within the coexistence–productivity–stability triad were overall positive. These patterns were mechanistically explained by phylogenetic and trait composition: Phylogenetically and functionally more diverse communities supported higher coexistence potential and greater productivity, while those dominated by species with stronger root-mycorrhizal collaboration and larger seeds exhibited enhanced productivity and stability. Our work provides integrative empirical evidence linking biodiversity maintenance to ecosystem functioning, demonstrating that conserving phylogenetically and functionally diverse communities, particularly those including collaborative species, is key to sustaining biodiverse, productive, and stable ecosystems.

Tensor enhanced chest cancer classification via CNN and Vision Transformer models

Lung diseases, particularly lung cancer, remain a leading cause of mortality worldwide, accounting for approximately 1.8 million deaths annually. Early and accurate diagnosis is critical for improving patient outcomes. This study also introduces a unified platform for evaluating multiple convolutional neural network architectures and comparing them to a Vision Transformer model while utilizing a common tensor-based preprocessing pipeline for classifying lung cancer with CT/PET-CT imaging. To enhance model adaptability, all input images were initially converted into tensors prior to training, enabling implicit fine-tuning without altering the original architecture. The YOLOTransfer dataset, comprising diverse and annotated medical images, was used to benchmark model performance. Classical CNN models such as AlexNet, VGG-16, ResNet-50, DenseNet, and EfficientNet were compared against ViT in terms of accuracy, sensitivity, specificity, F1-score, and AUC-ROC. Among all models, ResNet-50 and EfficientNet achieved the highest accuracy, while the Vision Transformer showed competitive results in capturing complex global patterns. The findings highlight the complementary strengths of convolutional and transformer-based architectures for medical image analysis and demonstrate the feasibility of deep learning approaches for lung cancer detection.

6D Building information modelling (BIM) for demolition waste management of semi-submersible wind turbines

Abstract The rapid global expansion of offshore wind is matched by an emerging wave of decommissioning. By the 2030s, over 30 GW of Europe’s capacity will reach end-of-life (EoL), rising to 40 million tonnes of material waste globally by 2050, with blades alone contributing approximately 325,000 tonnes of waste per year in Europe if not effectively recovered. This shift presents a risk of waste accumulation and an opportunity to embed circular economy (CE) practices into infrastructure renewal. To address this challenge, this study develops a BIM-based framework for dismantling a semi-submersible floating wind turbine (FWT), combining 3D modelling, 4D time sequencing and component-level inventories. The analysis adopts three condition-based scenarios — intact, minor and major damage — reflecting the real uncertainty at EoL, where exposure to marine environments can produce highly variable outcomes. This structure allows engineers to assess recovery routes in a realistic manner, from high-value reuse to advanced recycling. Our new findings demonstrate that Scenario 1 enables reuse and delivers the lowest cost, time and CO₂e impacts, while enabling the highest-value CE applications. Scenario 2 introduces moderate repair burdens due to processing requirements and Scenario 3 is dominated by energy-intensive processes. By linking those results to cumulative inventory, the study provides a replicable digital catalog that captures component fate across scenarios. The creation of material passports / inventories provides traceability of resources, highlights landfill avoidance and supports market preparation for secondary materials. For engineers, the framework offers a reproducible, inspection-driven decision-support tool that links component conditions to dismantling schedules and CE pathways, informing planning and procurement in upcoming decommissioning projects of the offshore wind industry.

Parsing the functions of immediate-early proteins in the lytic–latent balance of HCMV infection

Human cytomegalovirus (HCMV) establishes lifelong latency, during which immediate-early (IE) gene expression is strongly repressed. IE1 and IE2 are considered master regulators of the lytic cycle, yet it remains unclear whether their expression levels are sufficient to determine infection outcome, defined here as the balance between lytic and latent infection, and which viral or cellular processes underpin this control. Here, we show that when viral entry is enhanced in monocytes, overexpression of either IE1 or IE2 significantly increases lytic replication, identifying their abundance as a critical barrier governing infection outcome. Mechanistic analysis reveals that these effects of IE1 and IE2 are mainly mediated through two distinct modes of host manipulation. IE1 promotes disruption of PML bodies, thereby facilitating a broad increase of viral gene expression, whereas IE2 elevates cellular dNTP pools and creates an environment permissive for viral DNA replication, consistent with two key limiting barriers to the initiation of lytic replication. Notably, induction of IE1 in latently infected cells is also sufficient to promote reactivation of viral gene expression. Together, these findings define the limiting functions of IE1 and IE2 in overcoming host-imposed barriers to lytic replication and reactivation.

Evoked temporal summation in dogs to assess pain central sensitization and modulation – A feasibility study

Central sensitization and pain endogenous controls imbalance were reported in humans affected by chronic pain. In cats, nociplastic changes, such as the wind-up phenomenon, ( i.e ., decreased tolerance to repeated stimuli) were described using a validated mechanical device inducing temporal summation of pain (TSP). The study aim was to validate this method in dogs and to describe pro- or anti-nociceptive profiles occurring with canine osteoarthritis (OA). Healthy ( N  = 4) and OA ( N  = 31) dogs were assessed. Six TSP stimulation protocols were tested for their reliability and specificity, at mid-antebrachium (0.38 Hz at 2–4Newtons [N], 0.25 Hz or 0.50 Hz at 4N) or tail-base (0.25 Hz or 0.50 Hz at 2N). Endogenous facilitatory, or inhibitory, controls were assessed using a pressure pain threshold (PPT) before and after TSP, or conditioning pain modulation (CPM) via an ischemic model, respectively. The PPT pre/post stimuli were used to calculate ratio of facilitation (< 0%) or functional inhibition (> 7%). Statistical analyses included intraclass coefficient of correlation, Spearman’s correlations, Fisher and Mann U tests, with α = 0.05. The response to mechanical TSP scores tended to be moderately reliable, while tail-base stimulation decreased inter-trials variability compared to mid-antebrachium for healthy dogs ( P  = 0.029). The response was specific, OA dogs tolerated 50% less stimulations than healthy dogs at 0.5 Hz frequency and 2N ( P  = 0.008). This protocol led to spinal hyperexcitability; among facilitated OA dogs, 45% presented functional inhibition versus 73% for the non-facilitated dogs. No correlations were found between radiographic score, orthopedic score, response to mechanical TSP or age with the facilitation or inhibition ratios ( P  > 0.173). OA dogs developed central sensitization reflected by the wind-up phenomenon. An imbalance in favor of endogenous facilitatory controls was observed suggesting inhibitory control fatigue and neuroplasticity. Characterizing endogenous pain modulation will enable better management of OA pain by preserving inhibitory control and preventing its fatigue.

An explainable machine learning approach to predict fragility fractures and the identification of important features

mRNA-laden LNP-enabled in situ CAR-macrophage alleviates liver fibrosis via inhibiting activated HSCs and modulating the immune microenvironment

Liver fibrosis, marked by an abnormal buildup of extracellular matrix (ECM), poses a major health threat. Myofibroblasts, predominantly derived from hepatic stellate cells (HSCs) and portal fibroblasts, are the primary drivers of ECM synthesis. Fibroblast activation protein (FAP), highly expressed by activated HSCs, is a pivotal player in the pathogenesis of liver fibrosis. Delineating the mechanisms underlying HSC activation and devising strategies to curb their hyperactivity are paramount for the management and prevention of liver fibrosis. In this study, we explored a pioneering therapeutic approach leveraging CD163 antibody-conjugated liposomal nanoparticles (LNPs) encapsulating FAP-specific chimeric antigen receptor macrophage (CAR-M) mRNA (αCD163/LNP-FAPCAR). These LNPs are designed to selectively transduce liver macrophages, facilitating the in situ generation of FAP-specific CAR-modified macrophages (FAPCAR-M). Our findings revealed that these LNPs efficiently transduced macrophages, augmenting their phagocytic capabilities toward target cells. This resulted in a significant reduction of ECM and a concomitant enhancement of liver fibrosis resolution. The overarching goal is to precisely target and neutralize hyperactive fibroblasts, offering a promising avenue for treating liver fibrosis.

Synergistic interfacial engineering of mesoporous magnetic metal oxide TiO2 nanocomposites for sustainable visible-light photocatalysis: Experimental insights and ML-based performance prediction

This study investigates the structural, optical, morphological, magnetic, and photocatalytic properties of Fe 3 O 4 /TiO 2 nanocomposites (FeT NCs), synthesized through a modified sol-gel method for the photodegradation of Reactive Yellow 145 (RY145). Characterization of FeT NCs (PL, XRD, FTIR, VSM, DRUV-Vis, DLS, Zeta potential, XPS, BET, SEM, TEM, TGA) revealed that Fe 3 O 4 incorporation into TiO 2 enhances charge separation, suppresses electron–hole recombination through Ti–O–Fe linkages, and improves photocatalytic efficiency. The calcined 0.025FeT3 exhibited high crystallinity with dominant anatase TiO 2 and no rutile transition. SEM and TEM revealed a core–shell morphology with Fe 3 O 4 cores encapsulated by TiO 2 , while aggregation was minimized by synthesis conditions. Optimal photocatalytic performance (84.51% % RY145 removal at neutral pH) was achieved using 1 mg mL -1 0.025FeT3 following pseudo-first-order kinetics. The Langmuir–Hinshelwood model yielded rate and equilibrium constants of 2.80 mg.L -1 min -1 and 2.42 L mg -1 , respectively. Mechanistic and scavenging experiments indicated that photogenerated holes and •OH radicals dominated the degradation process. The FeT catalyst maintained high stability over six cycles. Magnetic measurements showed soft magnetic behavior with low coercivity and remanence, favoring easy recovery. The reduced bandgap (2.62 eV) facilitated visible-light activation, while BET analysis confirmed a mesoporous structure with high surface area. XPS verified the oxidation states of Fe and Ti, and HPLC confirmed RY145 decomposition via azo bond cleavage and oxidation to carboxylic acids, demonstrating efficient and sustainable photocatalytic activity. 0.025FeT3 demonstrated efficient, stable, and magnetically retrievable photocatalytic activity under visible light, highlighting its potential for sustainable treatment of textile wastewater. To optimize the batch experimental data, a novel ML-driven predictive framework was tested to model and map the relationships between the selected optimization parameters (FeT contents, FeT dose, reaction time), to predict RY145 photodegradation efficiency, and to identify the optimal operating window for improved photocatalytic performance (using three regression measures R 2 , MAE, and RMSE). The CNN models outperformed with a predicted accuracy and R 2 value of 0.91. Based on the results, ML-based evaluation outperformed manual optimization and traditional statistical methods, delivering a more efficient and reliable way for process optimization.

Daily briefing: Bad supervisors bump early-career researchers out of academia

Bletilla striata polysaccharide modified glass ionomer cement: enhanced mechanical and biological properties

Integrated hierarchical surface restructuring of assembled electrode arrays for next-generation neural interfaces

Neurostimulation devices rely on electrode arrays to deliver targeted electrical stimulation for modulating nerve activity. Enhancing stimulation specificity, device battery and energy efficiency, and device miniaturization requires low-profile electrodes with exceptional electrochemical performance. Hierarchical Surface Restructuring (HSR™), a femtosecond laser-based electrode surface treatment technology, enables these improvements by significantly increasing the electrochemically active surface area of the electrode contacts through the formation of highly textured, multi-scale architectures. Although HSR™ offers substantial potential to enable both high-performance electrodes and further miniaturization of electrode arrays, its broader adoption in medical device manufacturing has been limited by cost considerations and the inherent complexities of integrating new surface modification steps into established production workflows. This study investigates the feasibility of applying HSR™ technology to commercially available Pt-10Ir paddle-type electrode arrays and, for the first time, demonstrates that HSR™ can be implemented as a stand-alone, post-fabrication surface modification process that is compatible with existing device geometries and material constraints. This advancement represents a significant step toward broader adoption of HSR™ by medical device manufacturers and demonstrates its overall manufacturing viability. The process developed in this study circumvents key barriers to industrial implementation by enabling HSR™ to be seamlessly integrated into existing production lines as a post-fabrication surface modification step, thereby eliminating the need for major or costly process changes. The morphology, electrochemical performance, and processing efficiency of the restructured electrodes were systematically characterized. HSR™ enhanced key electrochemical metrics—including charge storage capacity, specific capacitance, and impedance—by up to two orders of magnitude, while maintaining short processing times and full compatibility with the device’s geometry and constituent materials. These findings demonstrate the potential for HSR™ to be seamlessly integrated into existing manufacturing workflows as a post-fabrication step, providing a scalable and cost-effective approach for enhancing the electrochemical performance of neurostimulation electrode arrays. Furthermore, in-operando CO 2 -snow-assisted processing was shown to be equally compatible with established production lines, improving electrode stability and surface cleanliness without necessitating any upstream process modifications.

Association of stress urinary incontinence with stress, sleep quality, and health-related quality of life in nulligravid college women

Automatic selection of the best neural architecture for time series forecasting

The diffusion of large language models in published academic articles

Large language models (LLMs) are rapidly changing academic research, raising questions of who is adopting these tools and under what conditions. This article analyzes full texts of 7.3 million journal articles published from 2020–2025 by four major publishers (Elsevier, Frontiers, MDPI, and PLoS) to track the prevalence of LLM-associated language and identify social and institutional correlates of adoption. A corpus of 228 focal words exhibiting sharp post-2022 frequency increases consistent with LLM output was developed; articles were scored on their rate of focal word usage. By 2025, an estimated 57% of published articles exhibited evidence of LLM influence, up from 12% in 2023. Among articles exhibiting LLM-influenced text, there is substantial heterogeneity, ranging from subtle linguistic influence to articles mostly or entirely LLM-generated. Difference-in-differences models reveal that LLM-associated language varies markedly across regions, institutional ranks, publishers, disciplines, and journal tiers. Economic development and proximity to English as a primary language are key predictors of regional variation. Lower-ranked institutions exhibit higher rates than elite universities, young for-profit publishers show elevated rates vis-à-vis competitors, and academic fields differ widely in adoption. LLM adoption in academic writing is pervasive but socially stratified. As models grow more powerful and their use becomes further entrenched in academic research, understanding social dynamics of adoption will be essential for governing the evolving relationship between AI and academic knowledge production.