The puzzle begins with a simple expectation. When a material enters the superconducting state, electrical resistance should drop to exactly zero as the temperature falls. Yet in a growing family of two-dimensional superconductors — including gate-tuned graphene systems — researchers have repeatedly observed regions where the resistance falls sharply but then saturates at a finite value instead of vanishing. These states show abrupt transitions out of the superconducting-like regime when temperature, magnetic field or current exceed critical thresholds, mimicking the phenomenology of true superconductivity. But the resistance never reaches zero. Theoretical physicists proposed in 1999 that such a ‘Bose metal’ could exist as a genuine quantum ground state, and a 2019 colloquium review catalogued the anomalous metal as a ‘failed superconductor’ seen in disordered films since the late 1980s. What has been missing is a clean, well-controlled system in which the state can be studied systematically.

The new experiment delivers exactly that. The team, led by Matthew Yankowitz at the University of Washington and Joshua Folk at the University of British Columbia, built devices from rhombohedral graphene — a stacked form of carbon in which the layers are aligned in the same lateral registry rather than the staggered arrangement of ordinary graphite. This stacking order produces flat electronic bands that amplify interactions between electrons, and rhombohedral graphene has in recent years become a prolific source of unconventional superconductivity. The researchers encapsulated their graphene in hexagonal boron nitride crystals supplied by Kenji Watanabe and Takashi Taniguchi of the National Institute for Materials Science in Japan, and placed it on a tungsten diselenide substrate that imparts strong spin–orbit coupling to the carbon electrons. Dual electrostatic gates allowed them to tune both the carrier density and the perpendicular electric field across the device with exquisite precision.

What they found was striking. At zero magnetic field, mapping the resistance across the full two-dimensional gate-voltage landscape revealed isolated pockets where the resistance dropped to zero — bona fide superconducting islands — embedded in a broader region of gate space where the resistance instead saturated at a finite value. Crucially, the two kinds of pockets displayed nearly identical phenomenology. Both showed sharp, abrupt transitions to the normal state as temperature was raised, as a perpendicular magnetic field was applied, or as the driving current was increased beyond a critical value. To any casual inspection, the two regimes looked like siblings. One simply carried current perfectly; the other did not.

The most dramatic twist came when the researchers applied a small magnetic field parallel to the plane of the graphene sheet. In conventional superconductors, an in-plane field has little effect because the field lines slip between the atomically thin layers without generating vortices. Here, however, a modest parallel field of a few hundred millitesla expanded the finite-resistance pockets and merged them with their zero-resistance neighbours, without qualitatively changing their behaviour. The result was a razor-sharp boundary, at millikelvin base temperature, between states of zero and finite resistance carved directly into the gate-voltage map. The superconducting pockets shrank while the anomalous-metal territory grew, suggesting that the parallel field gently tips the balance between the two competing phases rather than destroying either one outright.

The finite-resistance state reproduced the key experimental signatures that define the anomalous metal in thin-film superconductors. Most tellingly, the team measured a non-monotonic dependence on current: the differential resistance of the state varied in a complex, non-monotonic way as the applied direct current was swept, a hallmark previously documented in disordered films and Josephson-junction arrays. The researchers also performed a Berezinskii–Kosterlitz–Thouless analysis, examining how the voltage scales with current near the transition. In the true superconducting pocket, the current–voltage curves followed the expected cubic power law over a broad range, yielding a transition temperature of about 61 millikelvin. In the anomalous-metal pocket, the same criterion was never satisfied — the state flirts with superconducting order but never locks it in.

Why does this matter so much? The anomalous metal has been observed in quench-condensed aluminium films, granular metals, amorphous molybdenum–germanium films, electrostatically doped cuprate films, ion-gated two-dimensional superconductors, and hybrid semiconductor–superconductor Josephson arrays. In every previous case, the material contained substantial disorder, and skeptics could always argue that the finite resistance arose from sample-specific imperfections, inhomogeneous current paths, or heating effects rather than a genuine quantum state of matter. Ultra-clean rhombohedral graphene, with carrier mobility far exceeding that of any thin film and gate-tunable parameters that can be swept continuously, places severe constraints on all such extrinsic explanations. The state here is not an artifact of dirty metallurgy; it is an intrinsic property of a pristine quantum material.

The theoretical implications are equally profound. Proposed explanations for the anomalous metal invoke quantum fluctuations of the superconducting phase, dissipation from normal quasiparticles, a glassy tangle of superconducting vortices, or microscopic inhomogeneity that percolates superconducting islands into a resistive network. Each model makes different predictions about how the state should respond to magnetic field, current and disorder. The rhombohedral graphene platform, in which the superconducting and anomalous-metal pockets can be relocated, expanded and merged at will by turning gate voltages and applying small in-plane fields, offers a testing ground in which those predictions can be discriminated for the first time. The fact that the two phases share nearly identical critical behaviour — the same abrupt transitions, the same response to temperature and field — hints that they may be two faces of a single underlying quantum phase transition, separated by a boundary whose sharpness at base temperature suggests a genuine quantum critical line.

There is also a broader resonance with the wider graphene superconductivity programme. Since the 2018 discovery of unconventional superconductivity in magic-angle twisted bilayer graphene, the carbon family has yielded spin-polarized superconductors, chiral superconducting states, record violations of the Pauli paramagnetic limit, and superconductivity intertwined with ferroelectric orbital magnetism and topological charge order. The new result adds a crucial missing character to this cast: the anomalous metal, long confined to the margins of thin-film physics, now sits squarely within the same gate-tunable phase diagrams that host the most celebrated unconventional superconductors. Understanding why some pockets superconduct and others fail to do so — separated by nothing more than a small change in carrier density or electric displacement field — may reveal the microscopic ingredient that tips the balance, whether it be spin polarization, isospin texture, or the delicate interplay of flat-band kinetics and Coulomb repulsion.

For now, the anomalous metal remains without an accepted theoretical explanation, but the conditions of the hunt have fundamentally changed. A state once accessible only in disordered films of variable quality can now be dialled up on demand in an atomically clean, electrostatically programmable material, measured reproducibly across multiple devices and contact pairs, and probed with the full modern toolkit of quantum transport. As the authors and their collaborators — including theorists Boris Spivak, Steven Kivelson and Aharon Kapitulnik, whose work frames the field — continue to explore this system, the failed superconductor may soon give up its secrets. If it does, the answer will reshape our understanding not only of two-dimensional superconductivity but of the quantum phase transitions that govern how collective electronic order emerges, or collapses, at the coldest temperatures physics can reach.

Subject of Research: Anomalous metal and superconducting phases in rhombohedral graphene on a WSe2 substrate

Article Title: Anomalous metal and superconducting phases in rhombohedral graphene

Article References: Okounkova, A., Sohm, A., Faehndrich, T., Kumar, M., Waleffe, D., Yan, J., Watanabe, K., Taniguchi, T., Folk, J., & Yankowitz, M. (2026). Anomalous metal and superconducting phases in rhombohedral graphene. Nature. https://doi.org/10.1038/s41586-026-11033-1

Image Credits: AI Generated

DOI: 10.1038/s41586-026-11033-1

Keywords: rhombohedral graphene, anomalous metal, superconductivity, two-dimensional materials, quantum phase transition, gate tuning, spin–orbit coupling, Berezinskii–Kosterlitz–Thouless transition, failed superconductor, WSe2, quantum transport, flat bands