Now you see it, now you don't: Material can transition between quantum states

A team at Argonne National Laboratory has demonstrated that a single material can switch between two fundamentally different electronic states — and the switch is reversible, controlled by an electrical current. The material, a potassium-intercalated nickel sulfide designated KxNi4S2, transitions between a Dirac metal and a flat-band antiferromagnet depending on how much potassium sits between the nickel-sulfide layers. The work appears in Matter (DOI: 10.1016/j.matt.2025.102418) and was posted to arXiv in September 2025.

The finding is real, the physics is interesting, and the press coverage calling it a breakthrough in \u201cquantum states\u201d is doing what press coverage does: turning a solid experimental result into a vague superlative. What Argonne actually demonstrated is band-structure switching in a layered crystal — electrons in a Dirac cone behave as massless, high-mobility quasiparticles, while electrons in a flat band are heavy and slow. Applying a current drives potassium out of the structure, collapsing the layered sandwich and flipping the material between these two electronic regimes. It\u2019s reversible. It\u2019s reproducible. And it\u2019s genuinely unusual.

\u201cI cannot name another material that can do this,\u201d said Mercouri Kanatzidis, a professor of chemistry at Northwestern University with a joint appointment at Argonne National Laboratory. Kanatzidis is not a man given to restraint — he recently received the 2026 William H. Nichols Medal from the American Chemical Society for his work on halide perovskites. When he says he cannot name another material, it\u2019s worth paying attention.

The material was first characterized in a 2021 paper in the Journal of the American Chemical Society, originally synthesized as part of a broader effort to develop new superconductors. The current work, led by the Argonne team, maps out the electronic phase diagram across potassium concentrations and shows that the transition between Dirac metal and flat-band antiferromagnet is accessible via electrochemical gating — essentially, poking it with a current. Samples were prepared at Argonne\u2019s Center for Nanoscale Materials; electronic-structure calculations were run on the Bebop high-performance computing cluster at Argonne\u2019s Laboratory Computing Resource Center.

The flat-band state has a Neel temperature — the point at which antiferromagnetic ordering disappears — of up to 10.1 Kelvin as potassium content approaches zero, which is cryogenic by any practical standard. The Dirac cone formation is driven by nickel-nickel bonding exclusive to the potassium-intercalated phase, a structural feature that vanishes when the potassium is driven out.

Here is where the story becomes speculative, and where reporters covering this need to be precise. Transistor applications and quantum sensor applications are being floated as implications of this result. They are not demonstrated in the paper. They are extrapolations from the observation that a material can switch between two electronic characterizations under electrical control. That is a reasonable thing to speculate about in a conversation with a researcher. It is not the same as a demonstrated device application. The gap between \u201cwe can switch the band structure reversibly\u201d and \u201cthis will enable better transistors\u201d is roughly the gap between discovering a material property and engineering a product. Kanatzidis\u2019s own enthusiasm for the result does not close that gap.

The paper was posted to arXiv on September 12, 2025. The press pickup in late March 2026 is not unusual for a materials science result — the field moves more slowly than quantum computing, and Matter is a paywalled journal — but it means this has been sitting in the literature for six months without broad coverage. That\u2019s either a sign that the result is incremental (it isn\u2019t, particularly) or that quantum materials doesn\u2019t generate the same algorithmic-advantage frenzy as quantum computing. Probably both.

For people building quantum systems, the relevance is indirect: this is a materials result, not a qubit result. But the ability to engineer electronic band structure reversibly — to go from massless to massive electron behavior in the same piece of material — is a tool that materials physicists will want to know about. Whether it becomes a transistor depends on whether somebody can build a device around it, and nobody has done that yet.

The paper is available on arXiv (2509.09903). The Matter publication is behind a paywall; primary source access may require institutional subscription or direct author request.