First-hand information for everyone
Researchers studying moiré materials have long focused on holes in the valence bands of twisted transition‑metal dichalcogenides (TMDs). A new transport study on twisted bilayer molybdenum disulfide (MoS2), however, turns attention to electrons in the conduction bands and reports unexpectedly robust interaction effects — including signatures that look like a Pomeranchuk‑type, thermally assisted localization — surviving to unusually high temperatures.
The team built near‑AA stacked MoS2 bilayers with twist angles between roughly 2.3° and 3.5° and tuned the electron density with gates so that they could count electrons per moiré unit cell. They measured electrical resistance from millikelvin temperatures up into the hundreds of kelvin and applied magnetic fields while sweeping carrier density. What stands out is a highly non‑monotonic phase diagram as a function of twist angle. At the smallest angles in the set, correlated gaps (an interaction‑driven suppression of conduction) endure up to about 160 K. At intermediate twist angles those gaps weaken and collapse to temperatures around 20 K, and a richer variety of interaction‑driven states appears.
At the largest twist angle studied (about 3.5°) the authors observe that resistance at a filling of one electron per moiré actually increases when the sample is warmed — a hallmark of thermally assisted localization. This counterintuitive trend is reminiscent of the Pomeranchuk effect known from other strongly correlated systems, in which heating can favor an insulating, entropy‑rich state. The authors therefore describe the observation as “Pomeranchuk‑like,” and emphasize that the mechanism is consistent with charge localization assisted by entropy rather than conventional disorder heating.
But the behavior is not a carbon copy of prior moiré Pomeranchuk reports in graphene: the magnetic‑field response in twisted MoS2 is strongly anisotropic (it depends on field direction), whereas canonical isospin Pomeranchuk phenomena in graphene tend to be more isotropic and tied to internal valley–spin degeneracies. Moreover, some weak but discernible Pomeranchuk‑like signatures survive even above 100 K near a filling of two electrons per moiré, despite only modest changes in resistance there.
The study points to several broader messages. First, semiconducting moiré systems can host rich correlated physics in their conduction bands, not only in valence bands, expanding the palette of target materials and carrier types. Second, twist angle is a powerful knob: small changes shift interaction energy scales dramatically and reshape the temperature range of correlated behavior. Third, the interplay of interactions, internal flavor (spin/valley) degrees of freedom, and possible band topology in TMD moirés can give rise to phenomenology that differs from graphene‑based moirés.
The authors and the broader community are cautious about interpretation. Transport alone cannot uniquely identify the microscopic origin of the localization: disorder, phonons, orbital effects, or more exotic correlated order could also contribute. The anisotropic magnetic response, in particular, calls for microscopic explanation and for measurements that can separate particle‑hole symmetry breaking, spin/valley polarization, and orbital magnetism. The reported high temperatures of some correlated features are striking, but twist inhomogeneity, sample‑to‑sample variability, and screening from nearby gates remain possible complicating factors.
Next steps include spatially resolved probes (STM or nanoscale optical techniques) and momentum‑resolved spectroscopy (ARPES or micro‑ARPES where feasible) to map band reconstructions directly; thermodynamic measurements (compressibility, specific heat) to test entropy arguments; and theory work that incorporates the multiorbital nature and spin–orbit coupling of MoS2 to explain the anisotropic field response. If confirmed, conduction‑band moiré correlation effects that persist toward room temperature would open new possibilities for devices that exploit interaction physics at practical temperatures.
In short, twisted bilayer MoS2 adds a new chapter to moiré materials by showing interaction‑driven localization and Pomeranchuk‑like behavior in its conduction bands, with a rich dependence on twist angle and surprisingly high characteristic temperatures. The observations are intriguing and suggestive, but follow‑up experiments and modeling will be needed to pin down the underlying mechanisms and assess how general the phenomena are.