Animal studies often fail to predict human tissue responses to new drugs or newly developed therapies. Besides generating tremendous costs for clinical studies, it also raises significant ethical concerns. Therefore, novel approaches in mimicking natural human environments like vascular system growth control, are broadly developed to deliver a reproducible model to test novel drugs. Recently, researchers from the Institute of Physical Chemistry demonstrated a unique system that is based on endothelial cells coated onto the surface of microparticles that can be spatially organized into pre-designed patterns to initiate the growth of vascular systems of well-defined micro-architecture. The patterning is achieved via directed-assembly using external magnetic fields. The discovery opens up new opportunities for personalized drug testing and precision medicine. Let’s take a cool closer on this breakthrough.

Studying the ultrathin layers of molecules on surfaces—where catalysts work, batteries react, and proteins fold—is crucial for chemistry and materials science. But a major challenge persists: the inherently weak signals from interfacial molecules are often too faint to detect, and many delicate samples are prone to light damage, requiring non-invasive low-intensity illumination that further reduces signal levels. Researchers at the Dalian Institute of Chemical Physics, Chinese Academy of Sciences have developed an innovative technique that amplifies the weak molecular signals by optical amplifier while suppressing the background, enabling clearer, faster measurements of surface molecular structures.

Achieving high-throughput, precise manipulation of diverse biological particles—from nanoscale vesicles to living cells—within their native, often curved environments is a pivotal challenge for advanced biomedical research. To break the longstanding trade-offs between throughput, resolution, and substrate rigidity in optical manipulation, researchers have now developed flexible and stretchable on-chip optical tweezers (FSOT). This platform employs large-scale microlens arrays on flexible substrates to enable diffraction-unlimited trapping of hundreds of bioparticles directly on complex surfaces such as skin and tissue. Its unique bendability and stretchability further allow conformal operation on biological interfaces and precise control over inter-cellular interactions, opening new avenues for wearable diagnostics and implantable sensing platforms.

Prof. WU Kaifeng's team from the Dalian Institute of Chemical Physics of the Chinese Academy of Sciences, has addressed a missing link in understanding proton-coupledelectronic processes: triplet energy transfer coupled with proton transfer. As another key energy transduction mechanism in natural and synthetic systems, triplet energy transfer isfundamentally distinct from singlet energy transfer.

Chinese scientists proposed an anti-interference diffractive deep neural network (AI D²NN) designed for object recognition in multi-object scenarios. Unlike conventional optical neural networks built for single-object classification, AI D²NN achieves robust target recognition while effectively suppressing interference from other objects. By incorporating optical multi-dimensional multiplexing, the system enables flexible, high-capacity parallel recognition of multiple targets. This approach is expected to accelerate the practical deployment of optical neural networks in autonomous driving, medical diagnostics, and security monitoring.

Researchers at the Center for Computational Sciences, University of Tsukuba, have developed an accessible platform to overcome the limitations of conventional static docking simulations, offering new avenues for education, training, and reproducible research in molecular recognition and supramolecular chemistry. Their Distance-Guided Fully Dynamic Docking (DFDD) platform is a cloud-ready simulation framework that enables students and researchers to explore dynamic docking, visualize molecular binding in motion, and understand how host-guest crystal structures emerge from molecular interactions.

Researchers developed high-performance, lead-free piezoelectric thin films on silicon that convert everyday vibrations into electricity. By using strain engineering and a novel sputtering method, they achieved record performance and demonstrated the compatibility of efficient, battery-free MEMS energy harvesters with standard semiconductor manufacturing processes. 

This study investigates the co-pyrolysis behavior and product distribution of peanut straw and polyethylene film blends through thermogravimetric analysis and gas chromatography-mass spectrometry. Thermogravimetric analysis results revealed distinct pyrolysis temperature intervals: 247-356°C for peanut straw, 448-505°C for polyethylene film, and an extended range of 247-510°C for their mixtures. Synergistic effects, quantified through experimental-theoretical deviations, demonstrated enhanced mass conversion rates and accelerated pyrolysis kinetics in blended systems. As the mass ratio of peanut straw to polyethylene increases from 1:1 to 1:7, the bio-oil yield increased from 62.1% to 76.86%, accompanied by elevated alkane from 20.84% to 31.41% and olefin from 24.73% to 42.89%. HZSM-5 catalyst further optimized product profiles, achieving 77.08% bio-oil yield with enhanced hydrocarbon selectivity (alkanes: 35.69%; olefins: 46.16%) while suppressing oxygenates from 20.07% to 8.85%. These findings establish that co-pyrolysis with catalytic intervention effectively promotes hydrocarbon production and inhibits oxygenated compounds, providing strategic insights for agricultural plastic waste valorization.

Optical neural networks (ONNs) offer a pathway toward low-latency, energy-efficient artificial intelligence (AI); however, their scalability in terms of parameter count remains constrained. Addressing this challenge, a research team from The Chinese University of Hong Kong has developed a metasurface-based optical learning machine that integrates 41 million parameters, achieving performance comparable to state-of-the-art AI models. This approach experimentally enables highly scalable machine vision, thereby offering a practical pathway toward large-scale, high-performance optical AI computing.