Loading today’s stories
Science1 distinct publisher2 min readPublished
The benchmark problem of quantum many-body physics has been solved for specific materials using their actual structure. The senior author calls the systems studied a prototype for harder cases.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
Kenneth Wilson and others settled the general physics of the Kondo effect in the 1970s [12], and general physics is not the same product as a number for a named sample. The older understanding explains why a metal carrying a magnetic impurity stops getting more conductive as it cools and then gets worse [9]. It does not tell you the temperature at which that turnaround happens in a particular alloy, or how sharp it is. The second question is the one that decides whether a calculation can stand in for a measurement.
The mechanism is why the distinction bites. The impurity's unpaired electrons give the atom a magnetic moment, and as the metal cools, conduction electrons flip their own spins against it and eventually assemble into a cloud that cancels the moment outright [11]. The extra scattering along the way is what lifts resistance back up [9]. Everything quantitative about that depends on how strongly the impurity's orbitals couple to the conduction states of the host, which depends on which atom sits in which metal. A simplified model buries all of it in a few parameters [2]. Working from the actual atomic and electronic structure [3] removes the step where those parameters are chosen, and that is what Garnet Chan's claim about predicting properties "without referring to experiment" [6] rests on.
The announcement carries none of the resulting numbers [17]. Iron or manganese in copper appears only as an illustration of the class of problem [8]; no Kondo temperature is quoted, and no comparison against a measured resistance curve. Until those figures are in view, this is a claim about method rather than a demonstrated agreement, and the resistance minimum is precisely the observable that will settle it, because it is measurable in the same material the calculation names.
Chan's own framing is the honest boundary. A single magnetic atom embedded in a metal is the simplest strongly correlated system there is [8], and he describes the materials studied as a baby step toward high-temperature superconductors and quantum magnets [7]. Those are dense lattices in which the sensitivity of one electron's motion to all the others applies at every site and cannot be averaged away [16]. Whether an impurity solver built on real structure carries into that regime at tolerable cost is not something this result answers. What it does answer is narrower and worth having: for a class of materials where the interactions cannot be ignored the way they can in silicon [15], the chemistry can now be put in at the front end instead of fitted at the back.
Ranked by verification strength, evidence, and original report placement.
A team from Caltech and Yale University has shown for the first time how to accurately quantify the Kondo effect for specific real materials.
Previous approaches have for decades relied on simplified models to qualitatively describe the Kondo effect.
The new work uses the actual atomic and electronic structures of materials to solve the problem directly.
Chan: "It is now possible to predict the properties of some complicated materials purely through computation without referring to experiment."
Chan says the first materials studied "are like a baby step, or a prototype problem, along the way to more complex phenomena such as high-temperature superconductors and quantum magnets."
Physicists in the 1970s, including Caltech alumnus Kenneth Wilson (Ph.D. '61), established a theoretical understanding of the general physics around the Kondo effect.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Peer-reviewed result, single-source and uninspected
The underlying artifact is a Science paper with a named senior author and lead authors, and the coverage reports a concrete study scope (seven transition-metal impurities in copper) and a relative accuracy figure. Against that, the cluster contains exactly one publisher relaying an institutional write-up: no independent analysis, no computed Kondo temperatures, no measured-versus-computed comparison, and no reproducibility artifacts are present in the supplied material.
No adoption signal beyond publication
The supplied material describes a research publication only. There is no code, package, service, third-party usage, benchmark leaderboard entry or deployment disclosure that would let adoption be measured, and inferring uptake from a paper alone would be a guess.
Mildly overstated framing, author-hedged
Headline framing ('for the first time', 'predict the properties of some complicated materials purely through computation without referring to experiment') runs ahead of what the coverage substantiates, since no computed Kondo temperature or experimental comparison is given and the accuracy figure is relative to earlier models rather than to measurement. The gap is small rather than large because the researchers themselves hedge - calling the studied systems a baby step or prototype toward superconductors and quantum magnets - and the study scope is stated concretely.
Institutional announcement pipeline
The single item is a university-origin research write-up carried by an aggregator, quoting the senior author and lead author with credentialing detail (Bren Professor, center director, Simons Investigator) and no external or dissenting voice. Authors and institution have a direct reputational interest in first-of-its-kind framing, and the cluster contains no adversarial or independent commentary to offset it.
Moderate: consistent but unreplicated single source
Internal consistency is good and the underlying venue is a top peer-reviewed journal, which supports moderate confidence in the descriptive claims about the Kondo effect and the study's scope. Confidence is held down by the single-publisher cluster, the absence of independent verification, and one derived ledger claim that the source text partially contradicts.
science
Graz group measures the geometry a single-molecule collision has to hit1 distinct publisher
science
One thorium site in four: the crystal detail that makes a solid-state nuclear clock buildable1 distinct publisher
science
Flatiron's tensor network matches D-Wave's annealer on the same spin-glass dynamics1 distinct publisher
product
A 21km air gap in New York's quantum testbed, and the wavelengths it buys1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
phys.org
1 article · August 25, 2026