A new understanding of antibiotic resistance could transform the diagnosis and treatment of humans and animals and help combat the global threat of antibiotic resistance more effectively.

Quantum photonic chips could transform future quantum networks and quantum computers, but the best quantum emitters and photonic materials are difficult to combine without losing light. A self-aligned heterogeneous photonic integration method overcomes this barrier by joining diamond quantum emitters and spins with integrated photonics at low insertion loss, advancing scalable quantum photonic technologies.

Quantitative structure–property relationship (QSPR) modeling is a fundamental approach for property-oriented fuel design because it establishes quantitative relationships between molecular structures and physicochemical properties. Conventional QSPR models generally rely on manually engineered molecular descriptors, are usually developed to predict a single property, and often cannot readily incorporate newly available data, limiting their scalability and long-term applicability. Recent advances in large language models (LLMs) have demonstrated remarkable capabilities in text understanding and parsing knowledge in specialized fields such as chemistry. This study proposes FuelProp-LM, a QSPR framework based on instruction tuning and in-context learning that predicts multiple fuel properties with only the SMILES notation of a fuel molecule as input. Four open-source language models were fine-tuned using 100000 data entries obtained from the PubChem database. To enhance contextual reasoning, high-information prompts were constructed through MACCS molecular fingerprint similarity retrieval. The proposed framework was evaluated on 10 fuel physicochemical property datasets and compared with the widely-adopted LLM DeepSeek-V3.2 and existing QSPR approaches. FuelProp-LM consistently achieved competitive or superior predictive performance across downstream tasks. Quantitative analysis further revealed how instruction tuning and in-context learning improve molecular property prediction by enhancing chemically relevant attention patterns. The accuracy, generalizability, data scalability, and ease of use of this work demonstrate that FuelProp-LM can serve as a practical and flexible tool for fuel property prediction and play a significant role in future AI-assisted fuel design.

Oregon State University scientists have patented a better and more eco-friendly process for separating two metals that are critical in their pure forms for the semiconductor and atomic energy industries.

This study proposes a "fine molecular engineering" strategy to address the trade-off between efficiency and stability by designing non-fullerene acceptors (NFAs). Based on 7C-2F with a 1,4-diazahexane core, three NFAs  were synthesized by selectively controlling the side chain length and the degree of π-conjugation of the terminal groups. The results show that shortening the side chain (7C-BO-2F and 7C-eC9-2F) or extending the π-conjugation of the terminal groups (7C-N2F) both improve molecular crystallinity, shorten the π-π stacking distance, and suppress nonradiative energy loss. In bulk heterojunction devices, the PM6:7C-eC9-2F system achieved a PCE of 19.67% with an energy loss of only 0.539 eV. Devices based on 7C-eC9-2F exhibit excellent photostability and thermal stability: under continuous illumination, the device's T₈₀ reaches 2728 h⁻¹; at 65 °C, its T₈₀ reaches 1242 h⁻¹, significantly outperforming other control systems. Furthermore, when these acceptors are applied to quasi-planar heterojunction structures, excellent photovoltaic performance is also obtained, demonstrating good fabrication adaptability. In summary, precise molecular engineering provides a reliable pathway to obtaining high-efficiency, low-energy-loss, and intrinsically stable OPVs, laying the foundation for the practical application of next-generation organic photovoltaic devices. The article was published as an open access Research Article in CCS Chemistry, the flagship journal of the Chinese Chemical Society.

With the continuously increasing demands on imaging accuracy and on-orbit service life of space optical remote sensors, efficient thermal management under extreme thermal environments has become a critical bottleneck constraining their performance. The substantial temperature fluctuations induced by solar radiation and cold space can cause thermal deformation of optical components, elevation of focal-plane dark current, and even system failure. Traditional metallic thermal conductive materials can no longer meet the increasingly stringent requirements for heat dissipation and lightweighting. Carbon-based materials—including graphene, carbon nanotubes, and carbon/carbon composites—exhibit enormous potential for breaking through the bottlenecks of space thermal management by virtue of their ultra-high thermal conductivity, low density, and low coefficient of thermal expansion. However, systematic performance comparisons and application guidelines for these materials in the specific scenarios of space optical remote sensors remain absent, hindering the effective translation from laboratory research to engineering implementation. Therefore, a systematic review of the thermal conduction mechanisms, fabrication processes, and space applicability of carbon-based high-thermal-conductivity materials is of great significance for promoting their precise application in the thermal management of space optical remote sensors.

Researchers led by the University of Iowa have made the most detailed observations yet of Jupiter’s bow shock, a barrier that helps deflect harmful particles from the sun. They found that Jupiter uses a more complex wave structure than Earth to efficiently handle the solar wind. The findings could help scientists understand the physics associated with powerful energy released by dying stars. Results appear in the journal Nature Communications.