A study by the University of Cordoba shows that Mediterranean soils in southern Spain lose their properties when temperatures exceed 40 degrees Celsius, and proposes strategies to increase their resilience

Osaka Metropolitan University researchers found that a nonuniform particle size distribution in the solid electrolyte significantly improves the efficiency of lithium‑ion transport within the electrodes.

Researchers from The University of Osaka report the creation of step–terrace–ordered GaN surfaces using catalyst-referred etching (CARE). Mechanical testing on these surfaces demonstrates exceptional reproducibility across 100 measurements, achieving record-low stress scatter.

The LIGO–Virgo–KAGRA (LVK) detector network has a new trick up its sleeve to improve the instruments’ sensitivity to gravitational waves: it’s called Astrophysical Calibration and it plays a role similar to auto-tune in music production.

A new method could enable physicists to spot signs of dark matter in gravitational waves that are detected on Earth. This could occur if two colliding black holes spiral through a dense region of dark matter and merge, leaving an imprint in gravitational waves that are rippling across space and time.

Aston University and the Science and Technology Facilities Council’s (STFC) Hartree Centre are joining forces to accelerate the development, deployment, and industrial adoption of neuromorphic computing across the UK. They have signed an agreement to set up the Aston-Hartree Neuromorphic Centre of Competence at Aston University which will bring together academia, industry, and public sector stakeholders. It will act as a focal point for joint activities, shared expertise, and coordinated engagement across the UK and internationally.

The kingfisher’s brilliant blue feathers were once used like paint to create works of art. The technique, known as tian-tsui, was popular during China’s Qing Dynasty. And because tian-tsui uses delicate feathers, previous scientists struggled to study them using traditional analytical techniques. So, researchers reporting in ACS Omega developed new methods of investigating these featherworks without harming them. The team found that multiple bird species and layered pigments provided a one-of-a-kind palette. 

Simulating the nonlinear optical processes that shape ultrafast laser pulses is essential for modern laser and accelerator facilities, but conventional physics-based models are too slow for real-time use. Researchers from Stanford University, UCLA, and SLAC National Accelerator Laboratory report a deep learning surrogate that accelerates simulations of second‑order (χ²) nonlinear frequency conversion by orders of magnitude while maintaining high accuracy across complex pulse shapes. Focusing on noncollinear sum‑frequency generation relevant to SLAC’s LCLS‑II X‑ray free‑electron laser, the team developed a long short‑term memory (LSTM) neural network that replaces the most computationally expensive step of standard simulations. Operating entirely in a compressed frequency-domain representation, the model accurately reproduces temporal and spectral pulse features—including strong phase modulation and spectral holes—while reducing simulation times to milliseconds on GPUs. The work points toward machine‑learning‑enabled digital twins that could enable adaptive control and tighter integration between simulation and experiment in laser‑driven facilities.