Sharif University of Technology

A new series of spiro-acenaphthylene pyrrolizidine and pyrrolidine derivatives was synthesized via 1,3-dipolar cycloaddition of acenaphthoquinone with sarcosine or L-proline and various chalcone derivatives in the ionic liquid [Bmim]Br under ultrasonic irradiation at room temperature, with yields of up to 98% within 30 minutes. All compounds were fully characterized and evaluated for their α-glucosidase inhibitory activity. Most derivatives showed inhibitory activity, with IC₅₀ values ranging from 32.0 ± 0.3 to 463.4 ± 1.2 µM, showing lower IC₅₀ values than acarbose (IC₅₀ = 750.0 ± 1.5 µM). Notably, compound 4k, bearing a 4-nitro substituent, had the lowest IC₅₀ value (32.0 ± 0.3 µM), suggesting favorable interactions with the enzyme active site. These findings highlight the relevance of spiro-acenaphthylene scaffolds as non-glycosidic α-glucosidase inhibitors and demonstrate the utility of ionic liquids for rapid and sustainable heterocyclic synthesis, which may provide a basis for future in vivo and pharmacokinetic studies.

Composite nanofibers of polyvinyl alcohol/carboxymethyl cellulose/collagen (PVA/CMC/Col) were successfully fabricated using the electrospinning method, incorporating various proportions of each component. These nanofibers were meticulously characterized using advanced analytical techniques, including Fourier-transform infrared spectroscopy (FTIR), thermogravimetric analysis (TGA), X-ray diffraction (XRD), and mechanical testing, to ensure a comprehensive understanding of their structural and functional properties. The resulting nanofibers displayed a uniaxial structure with mean diameters ranging from 224 to 408 nm, showcasing precise control over the electrospinning process. The results revealed a direct correlation between material composition and performance, with the PCC5 formulation emerging as the optimal sample. This was evidenced by FTIR and TGA data, which confirmed that an increased collagen concentration led to a more integrated polymer network with enhanced intermolecular hydrogen bonding and superior thermal stability. Furthermore, mechanical testing demonstrated that the PCC5 sample exhibited the highest tensile strength, solidifying the link between chemical composition and robust physical properties. The biocompatibility of the nanofibers was rigorously evaluated through in vitro testing on MG-63 osteoblast-like cells using the MTT assay. The results revealed excellent compatibility with the cellular environment, with the PCC5 sample consistently promoting the highest rates of cell adhesion and proliferation. This study underscores the groundbreaking potential of PVA/CMC/Col nanofibers in the field of bone tissue engineering, as their optimized blend of physical and biological properties makes them a promising candidate for bone regeneration applications. This research paves the way for future advancements in regenerative medicine, offering a robust platform for further exploration and innovation.

Recent advances in Vision-Language Models (VLMs) and large language models (LLMs) have greatly enhanced visual reasoning, a key capability for embodied AI agents like robots. However, existing visual reasoning benchmarks often suffer from several limitations: they lack a clear definition of reasoning complexity, offer have no control to generate questions over varying difficulty and task customization, and fail to provide structured, step-by-step reasoning annotations (workflows). To bridge these gaps, we formalize reasoning complexity, introduce an adaptive query engine that generates customizable questions of varying complexity with detailed intermediate annotations, and extend the JRDB dataset with human-object interaction and geometric relationship annotations to create JRDB-Reasoning, a benchmark tailored for visual reasoning in human-crowded environments. Our engine and benchmark enable fine-grained evaluation of visual reasoning frameworks and dynamic assessment of visual-language models across reasoning levels.

Gas chromatography-ion mobility spectrometry (GC-IMS) has emerged as a powerful analytical platform in quality control of food, beverages, and flavor products. The technology allows for point-of-care application without the need for sample preparation, which makes it advantageous in resource-limited and equipment-hostile environments. One of these fields is the quality assessment of raw cocoa, which is fundamental for ensuring authenticity, product quality, as well as food safety and compliance. At the same time, analytical departments are facing an increasing urge to turn to a more sustainable use of resources, as well as substantial cost pressure. In the present study, a fast GC-IMS strategy was used to evaluate the provenance of cocoa in combination with machine learning. While most of the commercially available GC-IMS systems are based on nitrogen as a carrier gas, this approach was optimized and translated to a fast, hydrogen-based GC method. This was applied to a set of commercial cocoa liquor and data were evaluated by machine learning approaches, such as multivariate curve resolution-alternating least squares (MCR-ALS) and partial least squares-discriminant analysis (PLS-DA). By cutting down the analysis time by a factor of 2.5, this study demonstrates that in contrast to most conventional gas chromatography-mass spectrometry (GC-MS) systems, GC-IMS can be easily optimized towards higher throughput using the faster flow rates possible with hydrogen. Furthermore, this leads to enhanced signal quality and thus, a better basis for machine learning and finally, to an optimal tool for the classification of raw cocoa origins. Therefore, H2-based GC-IMS can be considered as a greener, resource-friendly, and efficient approach for the analysis of volatile food and beverage samples.

In energy storage and conversion systems, the working condition can have a significant influence on their performance. In this article, a phase change material (PCM) block in an enclosure containing a participating medium is studied under the combined impact of nanofluid natural convection and volumetric radiation. The object is to determine the optimal condition that enhances both energy storage and heat transfer performance. For this purpose, the lattice Boltzmann method (LBM) was implemented to solve the governing equations for flow, energy, phase change, and radiative transfer. The simulation results showed that the interaction effects of radiation and buoyancy influence the flow and temperature distribution within the system. Then, an artificial neural network (ANN) model was developed using the obtained data from the numerical simulation. This model was able to provide results within seconds, compared to the time-consuming numerical simulation. Genetic algorithm and particle swarm optimization were both used to find optimal conditions using the developed ANN model. Both optimization methods found similar optimal values, which were validated through numerical simulation. Using the obtained optimal solution can enhance the energy transfer performance by 3.34% and energy storage by 7.86%. Optimization based on LBM-ANNs provided a reliable and computationally efficient tool for the intelligent design of PCM thermal storage units operating under combined heat transfer.