Rigorous semiclassical theory of moiré-induced potential wells in twisted bilayer graphene quantum dots

A 2020 tight-binding study of circular, edge-free quantum dots in twisted bilayer graphene (tBLG) found that moiré AA-stacked regions act as effective quantum wells that control the confined spectra, that level energies follow twist-angle-dependent power laws in the dot radius, and that a perpendicular magnetic field drives gap closure with states localized at AA/AB stacking boundaries and interlayer current textures. The study, however, left its central physical picture without a derivation, connected its results to experiment only through a qualitative scaling argument, fixed the interlayer-hopping cutoff and staggered-mass amplitude without any sensitivity analysis, and used magnetic fields far beyond laboratory values (tens to thousands of tesla), leaving the proposed scanning-tunneling-microscopy verification undemonstrated. This problem record poses the open question that the source paper itself declares: establish a controlled semiclassical (or equivalent analytical) derivation of the moiré-induced effective potential wells at the AA regions and of their quantized spectra in the confined dot geometry. No resolution of this question is known in the literature to date.

Background

Twisted bilayer graphene (tBLG) consists of two graphene layers rotated by a relative twist angle $\theta$, which produces a long-wavelength moiré superlattice: the local stacking registry cycles periodically between AA regions (carbon atoms directly on top of each other), AB/BA regions (Bernal stacking), and intermediate spiral regions, with a moiré period that grows as $\theta$ decreases. At small so-called magic angles the moiré superlattice flattens the electronic bands, producing correlated insulating and superconducting phases, which is why tBLG became a central platform of modern condensed-matter physics. Independent of the magic-angle regime, a robust single-particle fact about tBLG is that low-energy states in an extended sheet localize at the AA-stacked regions. For the one-dimensional analogue — a strain-induced moiré in AB-stacked bilayer graphene — Vogl, Pankratov, and Shallcross developed a semiclassical theory for matrix-valued Hamiltonians (Hamiltonians whose components are matrices acting on the sublattice/layer degrees of freedom, treated with a generalization of the Gutzwiller trace formula, a standard semiclassical tool that expresses quantum spectra in terms of classical periodic orbits) and showed that the AA localization is literally trapping by a semiclassical moiré potential, with a semiclassical density of states in quantitative agreement with exact quantum calculation.

A quantum dot (QD) is a nanoscale region in which carriers are confined in all directions, so that the spectrum becomes discrete. Confining single-layer graphene electrostatically is hampered by Klein tunneling (the unimpeded transmission of Dirac fermions through potential barriers), so a standard device in theory is the infinite-mass (IM) boundary condition: a staggered, sublattice-asymmetric mass potential of amplitude $M_0$ applied outside the dot region, which opens a gap in the exterior and confines Dirac fermions without physical edges and therefore without edge-specific states. Mirzakhani, Peeters, and Zarenia (Phys. Rev. B 101, 075413 (2020)) combined these ideas and computed, within a tight-binding model (a lattice Hamiltonian with intralayer nearest-neighbor hopping $V^0_{pp\pi}\approx-2.7$ eV and interlayer Slater–Koster hoppings $V_{pp\sigma}(d_{ij})=V^0_{pp\sigma}\exp[-(d_{ij}-d_0)^2/\delta_0^2]$ with $V^0_{pp\sigma}\approx0.48$ eV, $d_0=0.335$ nm, truncated at interatomic distances $d_{ij}\leq 4a_{\mathrm{cc}}$, following the parametrization of Moon and Koshino), the spectra of circular IM-defined tBLG QDs of radii $R=1.1$–$7.5$ nm at twist angles from $0^\circ$ to $60^\circ$, with and without a perpendicular magnetic field $B$ (magnetic length $l_B=\sqrt{\hbar/(eB)}$). Their findings: (1) when $l_B$ exceeds the moiré length, the low-energy states localize at the AA regions and, at small twist angles, hybridize with the confinement induced by the central mass potential; (2) the lowest confined energies obey power laws in $R$ with twist-angle-dependent exponents ($\sim 1/R$ for AA stacking, $\sim 1/R^{1.66}$ for AB stacking, $\sim 1/R^{2.11}$ at the magic angle $1.08^\circ$); (3) when $l_B$ is smaller than the moiré length, the energy gap closes and edge-like states spread along the boundaries between AA and AB regions; and (4) interlayer current flows between the layers through the AA spots and back through the AB spots. The authors proposed low-bias scanning tunneling microscopy (STM) as an experimental probe of this localization.

The physical picture — AA regions behaving as effective potential wells that quantize the confined states — was, however, supported only by the numerical evidence and by analogy with the one-dimensional strain-moiré case: the paper states explicitly that a systematic study, providing a semiclassical approach similar to the one-dimensional one, is needed to rigorously show the existence of these moiré potential wells. Moreover, because the computationally tractable dots are tiny, the magnetic fields used in the calculations are enormous (tens to thousands of tesla, far beyond laboratory fields); the connection to experimentally accessible radii and fields was made only through a qualitative scaling argument, borrowed from earlier scaling practice for graphene quantum dots (Ferry et al.), that sets the dot radius equal to the cyclotron radius at the Fermi energy, $R=l_B^2 k_F$, giving $R=(E\pm V^0_{pp\sigma})/(e v_F B)$ and, for example, $R\approx 50$ nm at $B=20$ T. Finally, the computations used a single interlayer-hopping cutoff, a single mass amplitude $M_0=2.0$ eV, and rigid (unrelaxed) commensurate lattice structures, so the robustness of all reported features to these modeling choices is unknown. These gaps — mechanism, experimental mapping, robustness, and experimental test — jointly define what is missing to turn the numerical observation into a validated, experimentally testable theory of moiré quantum confinement in tBLG dots.

Problem Statement

Provide a systematic study of the moiré-induced effective potential wells at the AA-stacked regions of twisted bilayer graphene (tBLG) that rigorously establishes the existence and properties of these wells, in the setting of circular, edge-free tBLG quantum dots defined by an infinite-mass (staggered) potential, completing the program left open by Mirzakhani, Peeters, and Zarenia (2020). The source paper declares this open question explicitly: although the existence of moiré potential wells in the AA regions of the tBLG structure had been discussed in earlier work, 'a systematic study (e.g., providing a semiclassical approach similar to the one in Ref. 69)' — the one-dimensional strain-moiré matrix-Hamiltonian theory of Vogl, Pankratov, and Shallcross — 'is needed to rigorously show this existence.'

Concretely, the objective is a systematic semiclassical treatment (or an equivalently controlled analytical derivation) of the moiré-induced effective potential wells at the AA regions, generalizing the one-dimensional theory of Vogl, Pankratov, and Shallcross to the two-dimensional twisted moiré and to the confined geometry of a circular dot defined by an infinite-mass (staggered) potential, so as to rigorously establish the existence and properties of these wells and to derive their quantized spectra for the states confined in the dot, including the localization of low-energy states at AA spots and the hybridization between the mass-induced central confinement and the moiré wells at small twist angles. The objective is to establish the mechanism of the moiré wells and of the quantized spectra of the states confined in the dot in the small-twist regime where those wells exist — the AA-spot localization, the hybridization between the mass-induced central confinement and the moiré wells, and the power law of the lowest confined energy there ($\sim 1/R^{2.11}$ at the magic angle $1.08^\circ$). The source also reports power laws for the aligned AA- and AB-stacking limits ($\sim 1/R$ and $\sim 1/R^{1.66}$), gap closure with states localized at AA/AB boundaries when the magnetic length falls below the moiré length, and interlayer current textures.

The source paper acknowledges that the magnetic fields used in its computations are too high to be achievable in experiments and offers only the qualitative scaling argument $R=l_B^2 k_F$ (giving $R=(E\pm V^0_{pp\sigma})/(e v_F B)$ and, for example, $R\approx 50$ nm at $B=20$ T); it uses a single interlayer-hopping cutoff $d_{ij}\leq 4a_{\mathrm{cc}}$, a single mass amplitude $M_0=2.0$ eV, and rigid (unrelaxed) commensurate lattices; and it proposes low-bias scanning tunneling microscopy as an experimental probe without demonstrating it. These acknowledged limitations and the suggested probe are natural follow-up directions and useful checks on a derived theory.

References

1. M. Mirzakhani, F. M. Peeters, and M. Zarenia, "Circular quantum dots in twisted bilayer graphene", Phys. Rev. B 101, 075413 (2020). DOI: 10.1103/PhysRevB.101.075413, https://doi.org/10.1103/PhysRevB.101.075413 (source paper)
2. M. Vogl, O. Pankratov, and S. Shallcross, "Semiclassics for matrix Hamiltonians: The Gutzwiller trace formula with applications to graphene-type systems", Phys. Rev. B 96, 035442 (2017). DOI: 10.1103/PhysRevB.96.035442, arXiv:1611.08879, https://arxiv.org/abs/1611.08879
3. P. Moon and M. Koshino, "Optical absorption in twisted bilayer graphene", Phys. Rev. B 87, 205404 (2013). DOI: 10.1103/PhysRevB.87.205404, https://doi.org/10.1103/PhysRevB.87.205404
4. D. K. Ferry, L. Huang, R. Yang, Y.-C. Lai, and R. Akis, "Open quantum dots in graphene: Scaling relativistic pointer states", J. Phys.: Conf. Ser. 220, 012015 (2010). DOI: 10.1088/1742-6596/220/1/012015, https://doi.org/10.1088/1742-6596/220/1/012015
5. J. Bucko and F. Herman, "Large twisting angles in bilayer graphene moiré quantum dot structures", Phys. Rev. B 103, 075116 (2021). DOI: 10.1103/PhysRevB.103.075116, arXiv:2101.11698, https://arxiv.org/abs/2101.11698
6. N. S. Bandeira, A. Chaves, L. V. de Castro, R. N. Costa Filho, M. Mirzakhani, F. M. Peeters, and D. R. da Costa, "Energy levels and Aharonov-Bohm oscillations in twisted bilayer graphene quantum dots and rings", Phys. Rev. B 111, 125409 (2025). DOI: 10.1103/PhysRevB.111.125409, arXiv:2410.12014, https://arxiv.org/abs/2410.12014
7. H. Wang, S. Ma, S. Zhang, and D. Lei, "Intrinsic superflat bands in general twisted bilayer systems", Light Sci. Appl. 11, 159 (2022). DOI: 10.1038/s41377-022-00838-0, arXiv:2201.00291, https://arxiv.org/abs/2201.00291
8. H. Wang and D. Lei, "Pressure-tuned superflat bands and electronic localization in twisted bilayer graphene-like materials", J. Appl. Phys. 136, 045107 (2024). DOI: 10.1063/5.0207883, https://doi.org/10.1063/5.0207883
9. A. L. Rakhmanov, A. V. Rozhkov, and A. O. Sboychakov, "Magic radius of an AA-stacked bilayer graphene quantum dot", Phys. Rev. B 105, 235415 (2022). DOI: 10.1103/PhysRevB.105.235415, arXiv:2202.08736, https://arxiv.org/abs/2202.08736
10. G. Giavaras, "Quantum dot energy levels in bilayer graphene: Exact and approximate study", Phys. Rev. B 112, 115408 (2025). DOI: 10.1103/9p4f-42bc, arXiv:2506.20162, https://arxiv.org/abs/2506.20162
11. X.-F. Zhou, Y.-W. Liu, H.-Y. Yan, Z.-Q. Fu, H. Liu, and L. He, "Electronic confinement in quantum dots of twisted bilayer graphene", Phys. Rev. B 104, 235417 (2021). DOI: 10.1103/PhysRevB.104.235417, arXiv:2110.10823, https://arxiv.org/abs/2110.10823
12. A. Rothstein, A. Fischer, A. Achtermann, E. Icking, K. Hecker, L. Banszerus, M. Otto, S. Trellenkamp, F. Lentz, K. Watanabe, T. Taniguchi, B. Beschoten, R. J. Dolleman, D. M. Kennes, and C. Stampfer, "Gate-Defined Single-Electron Transistors in Twisted Bilayer Graphene", Nano Lett. 25, 6429 (2025). DOI: 10.1021/acs.nanolett.4c06492, arXiv:2409.08154, https://arxiv.org/abs/2409.08154
13. Z. Huang, Y.-P. He, J. Cao, Q. Li, C. Li, G. Li, M. Bo, and C. Yao, "Single electron transistor based on twisted bilayer graphene quantum dots", J. Phys. D: Appl. Phys. 57, 115104 (2024). DOI: 10.1088/1361-6463/ad1351, https://doi.org/10.1088/1361-6463/ad1351
14. Z. Hou, Y.-Y. Hu, and G.-W. Yang, "Moiré pattern assisted commensuration resonance in disordered twisted bilayer graphene", Phys. Rev. B 109, 085412 (2024). DOI: 10.1103/PhysRevB.109.085412, arXiv:2307.09587, https://arxiv.org/abs/2307.09587
15. X. Liu, Z. Tao, W. Luo, and T. Chakraborty, "Correlated quantum phenomena in confined two-dimensional hexagonal crystals", Physica E 182, 116575 (2026). DOI: 10.1016/j.physe.2026.116575, arXiv:2603.16677, https://arxiv.org/abs/2603.16677
16. P. Ouyang, J. Yu, Q. Li, G. Jia, Y. Wang, K. Xiao, H. Zhang, Z. Hu, P. A. Pantaleón, Z. Zhan, S. Zhou, F. Guinea, Q.-K. Xue, and W. Li, "Structural and electronic signatures of strain-tunable marginally twisted bilayer graphene", Natl. Sci. Rev. 13, nwaf568 (2026). DOI: 10.1093/nsr/nwaf568, arXiv:2603.09295, https://arxiv.org/abs/2603.09295
17. X. Wang and M. Yang, "Enhanced interlayer coupling in twisted bilayer graphene quantum dots", Appl. Surf. Sci. 600, 154148 (2022). DOI: 10.1016/j.apsusc.2022.154148, https://doi.org/10.1016/j.apsusc.2022.154148