The team then fine-tuned the resonant strength of the meta-atoms to independently adjust the group delay for each spin. At the same time, frequency tuning and local structural rotation were used to set the phase while keeping unwanted crosstalk low. The PB phase, added through global rotation, extends the available phase range toward a full 2π without significantly altering the group delay design. Together, these elements create a practical single-layer design strategy for dual-spin achromatic control.
Experimental Proof Across Multiple Frequency Bands
The researchers demonstrated their approach experimentally using two types of devices operating in the 8–12 GHz range. One class consisted of spin-unlocked achromatic beam deflectors that maintained stable, spin-dependent steering across the band. The other involved achromatic metalenses that assigned different focusing functions to RCP and LCP light while preserving strong performance over a broad frequency range.
Mikhail Lukin’s team at Harvard presented a “universal” design for neutral-atom processors with robust error-correction capabilities using just 448 qubits, alongside a 3,000-qubit processor that can run for hours.
As Lukin notes: “These are really new kinds of instruments—by some measures, they’re not even computers… What’s really exciting is that these systems are now working already at a reasonable scale and we can start experimenting with them to figure out what we can do with them.”
A string of surprising advances suggests usable quantum computers could be here in a decade.
Researchers at the Department of Energy’s Oak Ridge National Laboratory have developed a deep learning algorithm that analyzes drone, camera, and sensor data to reveal unusual vehicle patterns that may indicate illicit activity, including the movement of nuclear materials. The work is published in the journal Future Transportation.
The software monitors routine traffic over time to establish a baseline for “patterns of life,” enabling detection of deviations that could signal something out of place. For example, a surge in overnight truck traffic at a facility which is normally only visited during the day could reveal illegal shipments.
The research builds on a previous ORNL-developed technology for recognizing specific vehicles from side views. Researchers improved the structure of this software’s deep learning network to provide much broader capabilities than any existing recognition systems, said ORNL’s Sally Ghanem, lead researcher.
Proteins are the molecular machines of cells. They are produced in protein factories called ribosomes based on their blueprint—the genetic information. Here, the basic building blocks of proteins, amino acids, are assembled into long protein chains. Like the building blocks of a machine, individual proteins must have a specific three-dimensional structure to properly fulfill their functions.
To achieve this, the newly produced protein chains in human cells are folded into their stable and functional form with the help of various protein folding helper proteins, known as chaperones, such as TRiC/PFD, or HSP70/40. The protein folding helpers isolate the amino acid chains, which have different chemical properties depending on the amino acid, from the cellular environment. This prevents the newly produced protein chains from clumping together and causing disease.
F.-Ulrich Hartl, a director at the Max Planck Institute of Biochemistry, has spent decades studying the mechanisms of protein folding. Niko Dalheimer, a scientist in Hartl’s department and one of the two lead authors of a new study published in Nature, explains: Much of what we know about protein folding has been learned from studies conducted in test tubes. However, it is virtually impossible to faithfully replicate the cellular environment in vitro.
Imagine an asteroid striking Earth and wiping out most of the human population. Even if some lucky people survived the impact, Homo sapiens might still face extinction, because the social networks humans rely on would collapse.
This dynamic also plays out in the wild.
Social interactions are essential for many animals, helping them to locate food, spot predators and raise offspring. Without such connections, individuals can struggle to survive.
It is not easy to follow the interactions of large molecules with water in real time. But this can be easier to hear than to see. This is how an international team deciphered the role of water in the collapse of PNIPAM.
Some polymers react to their environment with conformational changes: one of these is the polymer PNIPAM, short for poly(N-isopropylacrylamide). It is water-soluble below around 32 degrees Celsius, but above this temperature it precipitates and becomes hydrophobic. This qualifies it for smart sensor applications. But what actually happens between PNIPAM and the solvent water?
Researchers at Ruhr University Bochum, Germany, and the University of Illinois Urbana Champaign collaborated with sound production specialists from Symbolic Sound Corporation to investigate this question. Using sound representation, they were able to decipher the interaction of water molecules with PNIPAM for the first time. They reported their findings in the journal Proceedings of the National Academy of Sciences on February 4, 2026.
A team of biochemists at the University of California, Santa Cruz, has developed a faster way to identify molecules in the lab that could lead to more effective pharmaceuticals. The discovery advances the rapidly growing field of biocatalysis, which depends on generating large, genetically diverse libraries of enzymes, and then screening those variants to find ones that perform a desired chemical task best.
This strategy has attracted major investment, particularly from drugmakers, because it promises quicker routes to complex, high-value molecules. However, traditional approaches to finding new biologically beneficial molecules often require “lots of shots on goal,” where researchers test enormous numbers of candidates through slow and inefficient workflows.
The method developed by the UC Santa Cruz team aims to significantly shorten that process by introducing smarter and faster decision-making tools that help researchers identify promising enzyme variants much earlier. The researchers detail their new approach in the journal Cell Reports Physical Science.
Muscles make up nearly 40% of the human body and power every move we make, from a child’s first steps to recovery after injury. For some, however, muscle development goes awry, leading to weakness, delayed motor milestones or lifelong disabilities. New research from the University of Georgia is shedding light on why.
UGA researchers have created a first-of-its-kind CRISPR screening platform for human muscle cells, identifying hundreds of genes critical to skeletal muscle formation and uncovering the potential cause of a rare genetic disorder. The findings come from two companion papers published in Nature Communications, one describing the large-scale screen and a second digging into a particular gene’s role in muscle development.
Together, the studies provide a comprehensive genetic map of how human muscle fibers are built and lend insights into the effects of genetic mutations on developmental muscle defects. By linking specific genes to the muscle-building process, this genetic roadmap gives clinicians a practical shortlist to more quickly pinpoint the likely genetic causes of a patient’s muscle-development disorder. It also provides researchers with clear targets to prioritize future drug or gene therapy approaches.