Researchers from the California Institute of Technology (Caltech) and Yale University have developed a computational method that can more accurately predict the Kondo effect in specific metals, potentially advancing research into complex quantum materials.
The Kondo effect occurs when magnetic impurities in a metal cause its electrical resistance to increase as the material is cooled below a characteristic temperature. Although the underlying physics has been understood for decades, accurately predicting how the phenomenon behaves in individual materials has remained a challenge.
Normally, cooling a metal reduces the thermal vibrations of its atoms, allowing electrons to move more easily and lowering electrical resistance. However, when magnetic atoms are present as impurities, interactions between their localized electrons and the metal's conducting electrons can cause resistance to rise at sufficiently low temperatures.
The researchers addressed this challenge by applying computational techniques originally developed in quantum chemistry to the study of metals. Rather than relying on simplified models of the electronic structure, their approach accounts for the actual atomic and electronic properties of the material.
The team tested the method on seven different transition-metal impurities embedded in copper. The calculations reproduced key trends observed in experiments and, for most of the impurities studied, outperformed conventional model-based approaches. In some cases, the improvement in accuracy reached approximately two orders of magnitude.
The findings, published in the journal Science on July 30, 2026, also provide insight into why the Kondo effect varies among different magnetic impurities.
Although the method has so far been tested on relatively simple systems, the researchers say it could help advance computational studies of more complex materials, including high-temperature superconductors and quantum magnets.
The study demonstrates how calculations based on the fundamental properties of real materials could help scientists predict complex electronic behavior with greater accuracy, potentially reducing reliance on experimental measurements.
Sources: Original study in Science,Caltech research announcement