When a Moiré Grid Learns to Rearrange: Tunable Charge States in Twisted Semiconductors
11 May 2026, Yanjiang
Electrons transfer between twisted MoSe₂ and WS₂ layers under a vertical electric field, revealing a clean ladder of charge-transfer states up to integer filling.
There is something quietly compelling about the moiré superlattice. Take two atomically thin sheets of semiconductor, stack them at a slight twist angle, and a new periodic pattern emerges — a grid within a grid, like the overlapping mesh of two window screens. This structural artistry has already produced a rich catalogue of surprises: correlated insulators, superconductivity, and topological states, all arising not from new materials but from the geometry of the interface alone. The system is both simple and inexhaustible.
Now, a preprint (arXiv:2605.05571) from Zheyu Lu and a team of eighteen researchers reports a careful study of how electrons move between two such layers — MoSe₂ and WS₂ — when both the doping level and the vertical electric field are tuned. The work reveals a surprisingly clean progression of interlayer charge-transfer states, from n/n₀ = 1 all the way to n/n₀ = 4, and shows that these transitions can be controlled with remarkable precision. The result is a platform where the Fermi-Hubbard model — a cornerstone of condensed matter theory — can be realized with a tunable charge-transfer gap on an effective honeycomb lattice.
How moiré patterns confine electrons
The basic physics is straightforward but worth stating clearly. When MoSe₂ and WS₂ are aligned at a 60° twist (the H-type stacking configuration), their bands form a Type-I alignment: the conduction band minimum of MoSe₂ sits below that of WS₂, so added electrons prefer to stay in the MoSe₂ layer. The moiré potential — the periodic landscape created by the lattice mismatch — further corrals these electrons into a triangular array of sites, each site holding at most one electron at integer filling.
But this static picture tells only half the story. By applying a vertical electric field across the device, the team can tilt the band alignment from Type-I to Type-II, switching which layer becomes energetically favourable for the electrons. At a critical field strength, the electrons begin to transfer from the MoSe₂ layer to the WS₂ layer — a controlled migration across the interface, monitored optically through the emergence and disappearance of specific excitonic signatures.
The team observed this directly. The low-energy moiré trion (LET) — a bound state of two electrons and one hole — dominated the optical spectrum at negative fields, when electrons stayed in MoSe₂. As the field was swept to positive values, this signal faded, and a new moiré exciton (EX) peak rose in its place, marking the arrival of electrons in the WS₂ layer. The transition was clean and reversible, a spectroscopic fingerprint of interlayer charge transfer.
From one to four: a ladder of charge-transfer states
The real surprise came when the team tuned the electron doping. At low doping (n/n0 < 1), the charge-transfer transition occurred at a single well-defined field. But as electrons were added — one per moiré site, then two, then three — the story became both richer and more orderly.
At n/n₀ = 1 (one electron per moiré site), the LET intensity dropped sharply around the critical field, and the EX intensity rose in its place. At n/n₀ = 2, the same pattern repeated. At n/n₀ = 3 and n/n₀ = 4, the charge-transfer signatures remained well-resolved, though the peak structures evolved. This is not trivial. In many moiré systems, higher fillings produce complicated multi-particle states that are difficult to interpret. Here, the charge-transfer physics remained clean up to at least n/n₀ = 4 — a ladder of states that the team could climb step by step.
The key to this clarity lies in the moiré potential itself. The first-principles calculations show that the lowest moiré conduction band in MoSe₂ is strongly localized at the BSe/S stacking sites. Each site acts as an approximately independent quantum dot, and at integer fillings, the added electrons simply fill these dots one by one. The charge-transfer process — moving an electron from one layer to the other — is therefore a single-particle process for each site, not a collective rearrangement. This simplicity is what makes the system such a powerful testbed.
A tunable Fermi-Hubbard model
The Fermi-Hubbard model describes electrons hopping on a lattice, with on-site repulsion when two electrons occupy the same site. It is the simplest model that captures the competition between kinetic energy and interaction energy, and it is believed to contain the essential physics of high-temperature superconductivity. But solving it exactly is notoriously difficult, and experimental realizations are precious.
What Lu and colleagues have built is a version of this model with an additional knob: the charge-transfer energy between the two layers. By tuning the vertical field, they change the energy cost for an electron to occupy the WS₂ layer relative to the MoSe₂ layer. This effectively adjusts the charge-transfer gap — the energy separation between the two bands — without changing the hopping or the on-site interaction. The honeycomb lattice geometry arises from the moiré pattern itself: the interlayer charge-transfer sites in WS₂ form an effective honeycomb lattice, a consequence of the stacking symmetry.
Monte Carlo simulations of the doping-dependent electric field susceptibility predict that multiple correlated charge-ordered states should appear — not only at integer fillings, but at fractional ones as well. These predictions remain to be tested experimentally, but they point toward a rich phase diagram waiting to be explored.
Why this matters
The condensed matter community has spent the past decade learning to engineer moiré systems with exquisite control. The ability to tune band alignment, doping, and superlattice geometry has unlocked phenomena that were once confined to theoretical models. But the gap between discovery and understanding remains large: many moiré phenomena are observed before they are explained, and the complexity of the systems often obscures the underlying physics.
This work does not claim to have solved that problem. But it offers something almost as valuable: a clean, optically accessible system where the charge-transfer physics is both tunable and predictable. The LET and EX states serve as sensitive probes of the electron distribution, and the field susceptibility measurement provides a direct experimental observable that can be compared to theoretical calculations. The team’s approach complements the transport measurements that have dominated the field, adding an optical dimension that can resolve charge states with spectral clarity.
The road ahead is reasonably clear. The Monte Carlo predictions for fractional filling states are a natural target for future experiments. The team also expects that the same approach can be extended to other TMD heterobilayers with different band offsets, potentially creating a family of tunable charge-transfer systems. The holistic understanding promised in the title — covering both the emergent optical excitations and the correlated charge-transfer states — is not an exaggeration, but a goal that this work brings significantly closer.
There is a particular satisfaction in watching a system this complex yield such orderly behaviour. The moiré grid, once a curiosity of twisted geometry, has become a reliable laboratory for quantum many-body physics.
Yanjiang is an online editor of LoomSci
References
- Zheyu Lu et al., Tunable Interlayer Charge-transfer States in MoSe₂/WS₂ Moiré Superlattices, arXiv:2605.05571