Researchers have uncovered the initial adaptive mechanisms of Bacillus velezensis C5 when exposed to polyethylene microplastics. Using comparative transcriptomics, the study revealed a robust, polymer‑specific transcriptional program distinct from general starvation or surface‑contact responses. Exposure to polyethylene triggered massive upregulation of genes involved in biofilm formation and sporulation, while genes associated with canonical degradation were not induced. The bacterium remained viable without growth, formed dense biofilms, and increased surface hydrophilicity but caused only subtle physicochemical changes to the plastic. This survival‑first strategy represents a critical initial phase that must be overcome to enable efficient biodegradation.

Researchers have developed a hierarchical catalyst (NiNP‑BCN@C) featuring nickel nanoparticles encapsulated within a nitrogen‑doped carbon shell on a hollow‑rod carbon substrate. The hollow architecture maximizes exposure of Ni nanoparticles, while the carbon shell suppresses the competing hydrogen evolution reaction and promotes CO₂ activation. The catalyst achieves a CO Faradaic efficiency exceeding 90 % in an H‑cell and >98 % at 300 mA·cm⁻² in a flow cell, with a turnover frequency of ~93,579 h⁻¹ and stable operation over 32 h. In situ FTIR confirmed enhanced *COOH and COO⁻ intermediates, highlighting superior CO₂ adsorption and activation.

Researchers have uncovered the initial adaptive mechanisms of Bacillus velezensis C5 when exposed to polyethylene microplastics. Using comparative transcriptomics, the study revealed a robust, polymer‑specific transcriptional program distinct from general starvation or surface‑contact responses. Exposure to polyethylene triggered massive upregulation of genes involved in biofilm formation and sporulation, while genes associated with canonical degradation were not induced. The bacterium remained viable without growth, formed dense biofilms, and increased surface hydrophilicity but caused only subtle physicochemical changes to the plastic. This survival‑first strategy represents a critical initial phase that must be overcome to enable efficient biodegradation.

Researchers have developed a multi‑objective optimization framework (MO‑DIDC) that handles infeasible solutions in complex distillation process optimization through data‑driven identification and adaptive directed correction. By identifying promising infeasible solutions near Pareto‑optimal designs and correcting their most influential variables, the framework reduces unnecessary exploration of infeasible regions. Demonstrated on a side‑stream double‑column extractive distillation system and a four‑column extractive distillation system, MO‑DIDC achieves optimization time reductions of 35.3 % and 20.8 %, respectively, compared to a widely used genetic algorithm, while delivering superior or comparable process designs.

A new study in Science Bulletin shows that fructooligosaccharide, a widely used prebiotic, may alleviate hyperuricemia by reshaping the gut microbiota. The study identifies a microbial relay in which Ligilactobacillus murinus promotes the enrichment of Anaerotruncus colihominis, a bacterium capable of degrading uric acid. The findings suggest a new mechanism by which prebiotics may regulate host metabolism through interbacterial interactions.

The generation of high-intensity, ultrashort, coherent terahertz pulses in the 5-15 THz frequency range presents a formidable challenge in ultrafast science. Recently, researchers have introduced a novel scheme for generating intense, broadband THz pulses over a gapless wide spectral range based on collinear phase matched Raman-resonance-enhanced four-wave mixing (R-FWM) in strained diamond, providing a scalable route toward high-field THz sources for driving and interrogating complex dynamics in matter.

As intelligent electronics become increasingly integrated into daily life, health care, virtual interaction, and assistive systems, human–machine interfaces (HMIs) require sensing platforms that are not only wearable and self-powered but also capable of translating human signals into adaptive machine functions. Triboelectric wearable sensors are particularly attractive in this regard because they directly transduce human-generated mechanical stimuli, provide broad material and structural design freedom, and are readily adaptable to body-interfaced formats. In this review, wearability refers to body-mounted, skin-interfaced, textile-integrated, or otherwise human-attached triboelectric sensing platforms, whereas human-centric smart electronics refers to downstream electronic systems that remain functionally anchored to human-originated sensing, interpretation, feedback, or control. From this perspective, we review triboelectric wearable sensors from fundamentals to applications, covering working principles, material selection, device architectures, and fabrication strategies. We further discuss artificial intelligence-assisted signal processing, triboelectric artificial synapses, and neuromorphic computing as key bridges from self-powered sensing to intelligent HMI. Representative application spaces, including health care, gesture recognition, device control, immersive virtual interaction, wearable-to-robotic extensions, and intelligent transportation are discussed only when wearable triboelectric sensing serves as the primary human–input interface. Finally, the remaining challenges and future directions toward next-generation human-centric smart electronics are outlined.