Science1 distinct publisher3 min readUpdated
Rice researchers report that uniaxial strain shrinks both the residual moment and the anomalous Hall voltage in hexagonal FeS, while the underlying magnetic order sits still.
The Scientist · Science desk

Compiled by The ScientistSomething wrong?How this is made
A team at Rice University reports that gently compressing a crystal of hexagonal iron sulfide along one direction reduces two things at once: the material's very small leftover magnetic moment and the sideways voltage it generates when current flows without any applied field [1][2][4][5][7]. That matters because it turns a property that has mostly been argued about in classification terms into something an experimenter can move with a clamp [7][13].
FeS belongs to the altermagnets, a recently recognized group in which internal moments largely cancel, so the material produces no strong external field, yet it can still act on moving electrons in ways that ordinary antiferromagnets do not [3]. In FeS that action shows up as an anomalous Hall effect, a transverse voltage with no external magnetic field present [5].
The experimental point is the simultaneity. Corresponding author Pengcheng Dai said the value of this material is that the tiny magnetic signal and the electrical signal can be watched at the same time, and that squeezing the crystal in one direction makes both smaller together, which indicates the two are closely connected [6][16]. Neutron measurements at Oak Ridge National Laboratory showed that the basic magnetic structure does not change under the squeeze; what changes is which magnetic orientations dominate inside the crystal [8]. The much larger underlying magnetic order was essentially unchanged [7].
That combination narrows the interpretation. Because the crystal has several nearly equivalent directions for its moments to point, separated by very small energies, modest pressure is enough to favor some over others [9]. Put together, the reported response looks like a redistribution among near-degenerate orientation states rather than a change in the size of the antiferromagnetic order parameter [17]. The team is not claiming the standard electronic explanation is wrong. First author Weiliang Yao said the measurements show the small moment and the anomalous Hall effect are strongly linked in this material, and that understanding why is now the open question [11]; the Rice work does not rule out the usual account based on Berry curvature in the electronic structure [10].
For operators, the interesting part is the knob, not the taxonomy. Mechanical strain is cheap, local, and already used in semiconductor manufacturing, and the Rice authors frame it as a route to control in spintronics, where the aim is to use electron magnetism for storage or processing with less stray magnetic interference and lower energy use than some conventional approaches [12]. Dai's framing is that control is essential for applications, and that a relatively simple mechanical squeeze changes important magnetic and electrical properties together [13].
The caveat is sizing. The public account gives no strain magnitude, no size for the residual moment or the Hall response, and no measurement temperature, so there is no way yet to judge whether the effect is usable at device scale or only visible in a laboratory press [18]. The paper is in Advanced Materials, and the work combined crystal growth, transport, magnetization and neutron scattering with the groups of Qimiao Si and Emilia Morosan at the Rice Laboratory for Emergent Magnetic Materials [14][15].
Watch for the numbers in the published version: strain per unit change in Hall signal, the temperature range over which the coupling holds, and whether the switching is reversible on repeated cycling. Watch also for whether the same trick works in altermagnet candidates with larger Hall signals, since FeS was chosen partly because its orientation states are unusually close in energy [9].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Researchers at Rice University found that gently squeezing a crystal of iron sulfide can change two of its unusual properties at the same time: its tiny magnetic signal and the way electricity moves through it.
The Rice team studied a hexagonal form of iron sulfide (FeS).
Altermagnets combine features of two familiar magnet types: like antiferromagnets their internal magnetic moments mostly cancel, so they do not produce the strong external field of an ordinary magnet, yet they can still affect moving electrons in ways that could be useful for future electronic devices.
Pengcheng Dai said: "The interesting thing about this material is that we can watch the tiny magnetic signal and the electrical signal at the same time... When we squeeze the crystal in one direction, both become smaller together. That tells us the two effects are closely connected."
The researchers built a device that gently compresses the crystal from one direction; as the pressure increased, the small magnetic moment became weaker and so did the sideways voltage, while the much larger underlying magnetic order remained essentially unchanged.
The team used neutron beams at Oak Ridge National Laboratory; the neutron measurements showed the basic magnetic structure stays the same, but the squeeze changes which magnetic orientations are most common inside the crystal.
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 multi-technique result, single institutional account
The claim rests on a paper in Advanced Materials with a DOI, combining crystal growth, transport, magnetization and neutron scattering at Oak Ridge, and the central observation (moment and anomalous Hall voltage falling together under uniaxial strain while the underlying order is unchanged) is internally corroborated by the neutron data. Evidence is capped well below strong because everything reaching this cluster comes from one publisher relaying one institution, no numeric values are reported for strain, moment, Hall magnitude or temperature, and no independent replication or outside comment is cited.
No adoption signal in supplied sources
The supplied material reports a laboratory measurement and a journal publication only. There is no release, deployment, benchmark, pricing, licensing, or third-party usage disclosure of any kind, so adoption cannot be measured without inventing facts the sources do not contain.
Slightly overstated: application framing outruns disclosed data
The account is mostly careful: it hedges the spintronics angle ('could eventually'), explicitly declines to declare the Berry-curvature explanation wrong, and describes the underlying order as unchanged. The modest positive gap comes from pairing energy-efficiency and interference-immunity benefits with a result whose strain, moment, Hall and temperature values are not disclosed, and from an application narrative with zero adoption evidence behind it.
Single-channel institutional announcement
The only source is an institutional research announcement relayed by an aggregator: every quote comes from the authoring team, the framing highlights novelty ('newly recognized class', 'simple way to control'), and the closing section markets a downstream spintronics use. Universities and labs have a visible interest in publicity for published work and beam-time-backed results. The score is mid-range rather than high because the paper is peer reviewed with a DOI, there is no commercial product or fundraising ask disclosed, and the authors voluntarily concede the competing explanation.
Moderate-low: one publisher, one institution, no numbers
Confidence is limited by cluster structure rather than by any contradiction: a single publisher, a single institution, no independent corroboration, and no quantitative parameters to check. It is held above low because the underlying work is peer reviewed, uses two independent measurement modalities, and is reported with consistent internal detail and no contested claims.
science
Rice team flips an altermagnet's Hall signal with strain, not a magnetic field1 distinct publisher
product
Nuclea buys Moltex's entire patent shelf out of administration, in cash1 distinct publisher
build
Nuclear AI program adds an AI security vendor, and the $60M is not the company's1 distinct publisher
science
ORNL says the lipid bilayer itself remembers, putting membranes on the neuromorphic parts list1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 16, 2026