Bio-oil obtained from biomass pyrolysis needs further hydrodeoxygenation (HDO) for its suitability in fuel applications. However, catalysts studied to date suffer from coke formation and low yields. In this study, Cu-based catalysts, prepared using the wet impregnation method, were characterized, and their performance in the twostep HDO (mild stabilization followed by severe treatment in a batch reactor) was investigated. HDO in the presence of a Cu/Al2O3 catalyst achieved a maximum 50.5 wt.% treated oil yield and 67.6% carbon conversion compared to Co-Mo/Al2O3 and Co/Al2O3 catalysts. The measured coke formation was less than 3% for the Cubased catalyst, which was attributed to controlled deoxygenation and hydrogenation. The high heating value of the treated bio-oil ranged from 31.4 to 34.6 MJ kg-1. Additionally, gas chromatography-mass spectrometry showed that the treated bio-oils were rich in alkyl phenols and hydrocarbons, with the concentration ranging from 4.6 to 55.3 and 15.9 to 32.0 wt.% of total, respectively. Direct deoxygenation, hydrogenation, and demethoxylation were the significant reactions leading to the formation of hydrocarbons. Based on the findings, Cu-based catalysts can be used for HDO of pyrolysis bio-oil to obtain valuable chemicals and fuels.
Abnormalities in cardiac wall motion are strong predictors of cardiovascular risk, making their accurate detection essential for early diagnosis and effective clinical management. Traditional imaging modalities such as echocardiography, magnetic resonance imaging (MRI), and computed tomography (CT) provide valuable insights but face limitations related to accessibility, cost, and the complexity of spatiotemporal analysis. Recent advances in machine learning (ML), particularly deep learning (DL), have enabled automated extraction of spatial and temporal features from medical imaging. They improved accuracy in segmentation, motion estimation, and detection of regional wall motion abnormalities. This paper reviews state-of-the-art methods for predicting cardiac wall motion, with emphasis on DL applications across echocardiography, 4D CT, and cine MRI datasets. Representative studies demonstrate the potential of convolutional neural networks, recurrent neural networks, and transformers to achieve performance comparable to expert clinicians, while also highlighting challenges such as data scarcity, model interpretability, and limited external validation. Addressing these issues will be critical for translating ML-based approaches into routine practice, and integration of advanced imaging with robust ML frameworks helps in developing a reliable cardiac wall motion simulators for personalized treatment planning and improved cardiovascular care.
Tool selection is a key component of LLM agents. A popular approach follows a two-step process - retrieval and selection - to pick the most appropriate tool from a tool library for a given task. In this work, we introduce ToolHijacker, a novel prompt injection attack targeting tool selection in no-box scenarios. ToolHijacker injects a malicious tool document into the tool library to manipulate the LLM agent's tool selection process, compelling it to consistently choose the attacker's malicious tool for an attacker-chosen target task. Specifically, we formulate the crafting of such tool documents as an optimization problem and propose a two-phase optimization strategy to solve it. Our extensive experimental evaluation shows that ToolHijacker is highly effective, significantly outperforming existing manual-based and automated prompt injection attacks when applied to tool selection. Moreover, we explore various defenses, including prevention-based defenses (StruQ and SecAlign) and detection-based defenses (known-answer detection, DataSentinel, perplexity detection, and perplexity windowed detection). Our experimental results indicate that these defenses are insufficient, highlighting the urgent need for developing new defense strategies.
Abstract Among closely related species, host phylogenetic relationships are typically a stronger predictor of gut microbiome composition than environmental variation. However, the relative impact of genetic admixture caused by hybridization versus environmental variation on gut microbiome communities is poorly understood. To explore this knowledge gap, we used fecal metabarcoding to characterize chickadee gut microbiomes along a hybrid zone transect in natural environments and after transfer to a common, controlled environment. We collected fecal samples from nestling black‐capped (Poecile atricapillus), Carolina (Poecile carolinensis), and hybrid chickadees immediately after removal from their nests and twice after entering captivity and experiencing common diet and environmental conditions. To characterize gut microbiome communities, we extracted fecal DNA and sequenced the V3–V4 region of the 16S rRNA gene using Illumina MiSeq. Overall, Firmicutes and Proteobacteria were the major microbial phyla present across host species groups. Our analysis of alpha diversity showed that the transition to common environmental conditions significantly impacted host gut microbiome richness. In contrast, the change in host environment did not impact the community composition (i.e., beta diversity) of the gut microbiomes. Although not statistically significant, host ancestry may influence the microbiome composition more than host environment. Additional analyses suggest chloroplast 16S rRNA sequences accurately characterized host diets in captivity. In certain cases, 16S rRNA sequences can provide reliable characterization of habitat and dietary variation in wild birds. Although environment more strongly shapes microbiome richness and evenness, host ancestry may have a greater influence on the specific microbes present in the microbiome.
Maraging steels, particularly 18Ni300, have long been valued for their exceptional strength, toughness, and dimensional stability, making them essential in aerospace, tooling, and high-performance applications. The emergence of laser powder bed fusion (PBF-LB) as a manufacturing route has enabled unprecedented design flexibility but introduced new complexities in controlling microstructure and mechanical properties. This review provides a comprehensive examination of the current understanding of 18Ni300 steel processed by PBF-LB, with an emphasis on the relationships between processing parameters, microstructural evolution, and the resulting mechanical behavior. Part I critically synthesizes recent advances in the characterization of precipitation phenomena, austenite reversion, and defect formation in PBF-LB maraging steels, comparing them to conventionally produced counterparts. The discussion spans the effects of powder quality, oxygen and nitrogen contamination, and laser processing parameters on solidification behavior, porosity, and anisotropy. Mechanical performance is analyzed through tensile, fatigue, and fracture toughness data, highlighting the contributions of aging treatments, retained austenite, and transformation-induced plasticity (TRIP) to strength and ductility. Part II presents a bibliometric and statistical analysis of nearly 600 publications retrieved from Scopus, mapping the evolution of research themes, collaborations, and methodological trends in the field. The analysis reveals a strong focus on process optimization and mechanical testing, yet a relative scarcity of studies that integrate advanced multiscale characterization and modeling approaches to establish quantitative correlations between processing, microstructure, and properties. The review concludes by identifying key knowledge gaps and future research opportunities, including the need for standardized testing protocols, in situ process monitoring, integrated computational materials engineering (ICME) approaches, and long-term performance assessments. Together, these insights provide a roadmap for advancing the fundamental understanding and technological maturity of PBF-LB maraging steels, supporting their broader adoption in critical structural applications.
