Pith. sign in

REVIEW 4 major objections 6 minor 2 cited by

Mixed valence Mott insulator and composite excitation in twisted bilayer graphene

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The ν=-2 insulator in twisted bilayer graphene is a mixed-valence Mott insulator: the f orbital is self-doped with roughly one third of AA sites in an f³⁺ valence, and the low-energy hole side is carried by a composite excitation…

desk verdict Clean parton construction with an interesting mixed-valence Mott idea, but the central claim sits outside the regime where their own validation applies and no numerical check is reported. read the letter →

arxiv 2507.00139 v1 pith:OGV53QWQ submitted 2025-06-30 cond-mat.str-el

Jing-Yu Zhao , Boran Zhou , Ya-Hui Zhang This is my paper
classification cond-mat.str-el
keywords twistedbilayergraphenetopologicalheavyfermionmodelmixedvalenceMottinsulatorcompositeexcitationpartonmean-fieldtheoryancillasemimetalmomentum-selectivegap
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a concrete ground-state wavefunction for the symmetric correlated insulator at $\nu=-2$ in magic-angle twisted bilayer graphene, built on the topological heavy fermion model, in which flat f orbitals on AA sites hybridize with dispersive c bands. It claims that the f orbital is not frozen in the $f^{2+}$ valence of a trivial Mott insulator: roughly one third of AA sites are self-doped, with holes entering the c orbitals away from AA sites, so the f orbital sits in a superposition of $f^{2+}$ and $f^{3+}$ valences with an average valence near $f^{2.3+}$. The resulting mixed-valence Mott insulator still has a full insulating gap and a large hybridization $\langle c^\dagger f\rangle \neq 0$, but its low-energy hole excitations are dominated by a composite $\psi$ operator orthogonal to the microscopic electron operator. A sympathetic reader should care because this gives a microscopic identity to the ancilla fermion of earlier work, distinguishes the phase from both heavy Fermi liquid and Kondo breakdown, and reframes the parent state from which hole-doped superconductivity emerges.

What carries the argument

The machinery is a restricted-Hilbert-space parton description of the f orbital: at each AA site only the singlon state $|s_\alpha\rangle = f_\alpha|d\rangle$, the spin-singlet doublon $|d\rangle$, and the triplon state $|t_\alpha\rangle = (2/\sqrt{3})f^\dagger_\alpha|d\rangle$ are kept, with $|d\rangle$ as the vacuum and the Gutzwiller constraint $n_s+n_t\le1$ enforcing that at most one of these excitations is present. The physical $f^\dagger$ operator becomes $(1/2)s + (\sqrt{3}/2)t^\dagger$, and the orthogonal combination $\psi^\dagger = -(\sqrt{3}/2)s + (1/2)t^\dagger$ defines the composite excitation. A renormalized mean-field treatment multiplies the s/t–c hybridization by $g_\gamma = \sqrt{1 - \langle n_s\rangle - \langle n_t\rangle}$, converting the Mott-localized product state into a Slater determinant of s and t fermions hybridizing with c; this mean-field theory is shown equivalent to the ancilla theory in the small-$\gamma/U$ limit.

What would settle it

A direct check is momentum-resolved photoemission on the hole-doped side of the $\nu=-2$ insulator: the paper's spectrum places the top of the lower band at $\Gamma$ almost entirely in the $\psi$ (composite) channel, which has vanishing single-particle weight, with both $A_{c1}$ and $A_f$ small at the working point $\kappa\approx0.8$, so a sharp, strongly f-derived quasiparticle peak at $\Gamma$ would contradict the claim; a complementary check is a measurement of the average f valence, which the mean field puts near $n_f \approx 2.3$ (with triplon density $n_t \approx 0.3$ to $0.4$) rather than 2.

Watch

Extended reading notes

Core claim

The central claim is that at $\nu=-2$ the correlated ground state is a mixed-valence Mott insulator: instead of every AA site carrying exactly two f electrons, the f orbital is self-doped by holes created in the c bands around the $\Gamma$ point, so about one third of AA sites acquire $f^{3+}$ (triplon) character and the f valence is a superposition of $f^{2+}$ and $f^{3+}$ with average $n_f \approx 2.3$. Because the f-derived part of the electron operator is $f \sim (1/2)s + (\sqrt{3}/2)t^\dagger$, the same hybridization $\gamma$ that would naively make a doped state metallic instead opens a gap: away from $\Gamma$ the gap is the Hubbard $U$ with fixed $f^{2+}$ valence, while near $\Gamma$ there is a much smaller charge-transfer gap whose hole side is dominated by the orthogonal composite $\psi = -(\sqrt{3}/2)s + (1/2)t^\dagger$. This $\psi$ is a microscopic trion-type operator built from the anticommutator $\{f^\dagger, n_f\}$ minus a piece of $f^\dagger$, and it has vanishing weight in ordinary single-particle probes. At $\nu=0$ the same construction yields a Mott semimetal with a quadratic band touching dominated by $\psi$. The paper identifies the phase as distinct from both the heavy Fermi liquid (hybridization present) and the Kondo breakdown (hybridization zero).

Load-bearing premise

The central calculation trusts the truncation of each AA-site f orbital to only the $f^{2+}$, $f^{1+}$, and $f^{3+}$ valences (with $f^{2+}$ a spin singlet) plus a renormalized mean-field treatment of the Gutzwiller constraint; the paper notes this is strictly justified only for very large anti-Hund coupling $J_A$ and otherwise rests on a conjecture that the physics survives at moderate $J_A$.

Editorial extensions

If this is right

  • If the phase is a mixed-valence Mott insulator, the $\nu=-2$ parent state cannot be described as a lattice of well-formed local moments, so thermodynamic and spectroscopic interpretations based on a fixed $f^{2+}$ valence need revision.
  • The low-energy hole-doped side of the $\nu=-2$ insulator is dominated by the composite $\psi$ fermion, implying that the normal state upon hole doping inherits spectral weight that is largely invisible to single-particle photoemission.
  • The microscopic mapping identifies the ancilla fermion of the earlier theory with the composite $\psi$ operator, grounding the momentum-selective Mott gap in a concrete wavefunction rather than a phenomenological hybridization.
  • At $\nu=0$ the same framework gives a Mott semimetal with quadratic band touching dominated by $\psi$, and at $\nu=-2$ for $\kappa>\kappa_c\approx1$ it gives a semimetal, so the insulator-semimetal boundary is part of the same description.
  • The paper conjectures that the composite excitation and momentum-selective Mott gap are general features of Anderson models with large hybridization, beyond the familiar Kondo and heavy-fermion regimes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • I would extend the paper's picture to the hole-doped side: the low-energy Fermi surface of $\nu=-2-x$ should be made of composite $\psi$ fermions rather than bare f quasiparticles, and the cheap on-site pair $|s\rangle\langle t|$ that avoids the Hubbard cost is a natural candidate pairing channel for the superconducting dome.
  • A quantitative prediction that follows from the mean-field data but is not flagged in the paper: the $\Gamma$-point charge-transfer gap grows with the hybridization $\gamma$ and shrinks as $\kappa$ approaches 1, so tuning $\kappa$ through screening or displacement field should drive the insulator into the semimetal, which is directly measurable.
  • The same restricted-Hilbert-space construction should carry over to $\nu=+2$ using $f^{5+}/f^{4+}/f^{3+}$ valences, and to odd fillings once the local-moment sector is replaced by a spin liquid, which the paper notes as beyond its current method.
  • More speculatively, the momentum-space extent of the self-doped region near $\Gamma$ is set by the Gaussian momentum dependence of the THFM hybridization, so systems with sharper momentum-dependent hybridization should show a smaller anomalous region; this could be tested by varying twist angle.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes a parton mean-field description of the symmetric correlated insulator at filling ν = -2 (and the semimetal at ν = 0) in twisted bilayer graphene, formulated in the topological heavy fermion model. Each AA-site f orbital is truncated to the f1+, f2+, f3+ valences with a single spin-singlet doublon, and the physical f operator is expressed as a combination of two slave-fermion fields s and t (Eq. 4). An orthogonal composite fermion ψ is defined in Eq. 6. Using a renormalized mean-field theory that replaces the local constraint by a global Gutzwiller factor gγ, the authors find that at ν = -2 about 1/3 of AA sites are in the f3+ (triplon) valence, with holes entering the c bands near Γ, so the f orbital is in a superposition of f2+ and f3+ valences rather than a simple local moment. This 'mixed valence Mott insulator' has a large hybridization ⟨c†f⟩ ≠ 0, a Mott gap of order U away from Γ, a smaller charge-transfer gap at Γ, and a low-energy hole-doped excitation dominated by the composite ψ. At ν = 0 the same construction gives a Mott semimetal. The paper also claims equivalence of this renormalized mean-field theory to the ancilla theory of Ref. [34] in the limit γ(k) ≪ U, and uses this equivalence to set the phenomenological parameter κ ≈ 0.8.

Significance. If correct, this work would provide a concrete real-space wavefunction for the symmetric correlated insulator in TBG, going beyond the Green's function approach of earlier ancilla formulations and offering a potentially universal picture for Anderson models with large hybridization. The parton construction is explicit and elegant: the operator mappings in Eqs. (4) and (6) are well defined, the wavefunction ansatz |Ψ_c⟩ = P_G|Slater⟩ is clearly stated, and the calculations are reproducible from the declared parameters and Hamiltonians. The result that the f orbital is not a simple local moment at ν = -2 is a physically interesting and falsifiable claim. However, the significance is contingent on the reliability of the renormalized mean-field treatment at the numerically employed parameters (U = 20 meV, γ = -38 meV) and on the representativeness of the heavily truncated f-orbital Hilbert space; neither is currently established by an unbiased calculation.

major comments (4)
  1. [Equivalence to the ancilla theory; Appendix F, Eq. (F7); Fig. 2] The equivalence between the renormalized mean-field theory and the ancilla theory is proved only for γ(k) = 0 or γ(k) ≪ U, with gγ ≈ 1. The headline calculation in Fig. 2 uses U = 20 meV and γ = -38 meV, giving |γ|/U ≈ 1.9, and the text states n_t ≈ 0.3–0.4, which implies gγ = √(1 - n_s - n_t) ≈ 0.8. Both conditions of the proof are therefore violated at the working point. The agreement with the ancilla spectrum used to validate the calculation and to choose κ ≈ 0.8 is thus not supported by the presented derivation. Please provide a benchmark against exact diagonalization or variational Monte Carlo at these parameters, or restrict the quantitative claims (1/3 self-doping, ψ-dominated spectrum) to the controlled regime.
  2. [Effective model with restricted Hilbert space; Discussion] The truncation to f1+, f2+, f3+ with a single spin-singlet doublon is acknowledged in the Discussion to be 'strictly speaking, justified only when JA is very large' and is otherwise a conjecture. The central quantitative claim—that around 1/3 of AA sites are self-doped into the f3+ valence—is computed entirely inside this truncated space. Appendix D extends the calculation to two additional d-wave doublon states and finds the same spectrum, but this does not cover other doublon channels or the splitting of triplon states by J_A, which the text neglects. The manuscript should either argue that the quoted parameters correspond to the large-J_A limit, or provide numerical evidence that the self-doping fraction is robust when the Hilbert space is enlarged.
  3. [Model, Eq. (3); Fig. 3] The parameter κ is introduced as a free parameter and fixed at κ ≈ 0.8 by requiring the new mean-field theory to match the authors' own ancilla theory. Since the equivalence to that ancilla theory is derived in the same paper and, as noted above, does not hold at the working parameters, this fitting is a self-consistency condition rather than an independent validation. To avoid circularity, κ should be fixed by a microscopic estimate of the c-f repulsion that motivates the term -κν, or by comparison with an unbiased calculation of a small cluster of the topological heavy fermion model.
  4. [Renormalized mean field theory, Eq. (7)] Equation (7) replaces the local constraint n_{i;s} + n_{i;t} ≤ 1 by a single global renormalization factor gγ = √(1 - n_s - n_t). At the self-consistent densities reported (n_t ≈ 0.3–0.4), the constraint is strongly fluctuating, and the factor √(1 - n_s - n_t) may not accurately capture the effect of the projection on the hybridization and on the density distribution. The paper does not compare this global treatment with a slave-boson mean field that enforces the constraint locally through Lagrange multipliers, nor with a fully projected Monte Carlo evaluation of the ansatz. A demonstration that the self-doping and the ψ-dominated spectrum survive a more faithful treatment of the constraint is needed.
minor comments (6)
  1. [Effective model with restricted Hilbert space] There is a typo: 'angular moemntum' should read 'angular momentum'.
  2. [Fig. 2 caption] The label 'νc(k) = nc(k)' is confusing because ν is already used for the filling; consider writing 'δn_c(k)' or 'hole density deviation'.
  3. [Appendix F, section 2] The sentence 'In a more physical relevant regime with γ(k) ≠ 0, γ(k) ≪ 0' should read 'γ(k) ≪ U'.
  4. [References] Reference [10] lists 'Physical Review Letter' instead of 'Physical Review Letters'.
  5. [Fig. 3(a)] The axis label '∆□' should be 'ΔΓ' to denote the Γ-point gap.
  6. [Abstract] The abstract says 'we construct the ground state wavefunction'; the paper actually constructs a variational Gutzwiller-projected Slater determinant solved approximately by RMFT, so a phrase such as 'variational ground state' would be more precise.

Circularity Check

2 steps flagged · score 4.0 of 10

Partial circularity: the quantitative self-doping fraction is calibrated by a tunable κ chosen to match the authors' own ancilla theory, and the advertised orthogonality of ψ to f is definitional; the central mixed-valence mechanism still has independent content.

  1. fitted input called prediction [Model, Eq. (3) and the paragraph following it]
    "We leave a microscopic derivation of this term to future work and treat it as a tunable parameter here. In agreement with Ref. [29], we find κ ≈ 0.8 gives reasonable results, in the sense that the new approach matches the previous ancilla theory[34] in the spectrum of single electron excitation."

    The headline quantitative claim, roughly 1/3 of AA sites self-doped with n_f ≈ 2.3, is controlled by the parameter κ through the on-site potential U/2(n_f − 4 − κν)^2 in Eq. (3). κ is introduced as a free, tunable input and is fixed to κ ≈ 0.8 by requiring that the present theory reproduce the spectrum of the authors' own ancilla theory [34]. The claimed mixed-valence fraction is therefore in large part an output of this calibration rather than an independent first-principles prediction. Because [34] is prior work by the same authors, agreement with it is not an external confirmation. The qualitative mixed-valence state is not entirely manufactured, since Figs.

  2. self definitional [Eqs. (4)-(6) and the following paragraph]
    "It is also useful to define another fermionic operator from an orthogonal linear combination: ψ†_{i;α} = −√3/2 s_{i;¯α} + 1/2 t†_{i;α}. (6) ... Microscopically ψα corresponds to a composite operator ... but it is easier to just view it as a different linear combination of s and t† operator within our subspace."

    The abstract advertises that the composite excitation has a sign structure 'such that it is orthogonal to the microscopic f operator.' That orthogonality is not a derived result: Eq. (6) is, by construction, the orthogonal linear combination to the projected f operator in Eq. (4). Thus the identification of ψ as the operator orthogonal to f is definitional. The nontrivial physical content is the computed mean-field result that ψ dominates the low-energy hole-doped side near the Γ point, and that content is not itself circular; only the advertised orthogonality property reduces to the choice of the definition.

full rationale

The paper is largely self-contained: it constructs a restricted Hilbert space of f^{1+}, f^{2+}, f^{3+} valences, derives the projected electron operator, defines ψ as the orthogonal combination, and solves a renormalized mean-field Hamiltonian. The claimed equivalence to the ancilla theory is supported by an explicit perturbative calculation in Appendix F rather than by bare citation, which gives independent mathematical content for small γ(k). However, two steps are partially circular. First, κ is a tunable parameter whose value κ ≈ 0.8 is justified by matching the authors' own ancilla theory [34], and the quantitative self-doping fraction n_f − 2 ≈ 0.3 is sensitive to κ (Fig. 3c); the headline 'around 1/3 of AA sites are self doped' is thus a calibrated output rather than an independent prediction. Second, the statement that ψ is orthogonal to f is true by construction through Eq. (6), although the more substantive claim that ψ dominates the low-energy spectrum is a computed RMFT result. The central qualitative phenomenon is not purely circular, because self-doping also occurs at κ = 0, and the wavefunction mapping between the slave-fermion and ancilla representations is explicitly derived. A separate correctness risk, not counted as circularity, is that the equivalence to the ancilla theory is proven only for γ(k) ≪ U and gγ ≈ 1, while the headline calculations use |γ| ≈ 2U and gγ ≈ 0.8; this affects whether the calibrated regime transfers, but it is a validity concern rather than a definitional reduction. Overall, the paper has partial circularity through parameter calibration and definitional orthogonalization, but retains independent content in the mean-field construction and the qualitative mixed-valence picture.

Assumptions & free parameters 4 free parameters · 5 assumptions · 2 invented entities

The central claim rests on the THFM, a restricted valence Hilbert space, an uncontrolled RMFT treatment, and a tuned phenomenological parameter κ. The composite ψ is defined by construction rather than derived from dynamics. These are the main costs.

free parameters (4)
  • kappa (κ) = 0.8
    Phenomenological attractive potential -κν between c and f electrons (Eq. 3). Tuned to match the authors' prior ancilla theory (Ref. [34]) and Lau-Coleman (Ref. [29]); no microscopic derivation is provided.
  • Hubbard U = 20 meV
    On-site interaction strength used in all band structure figures. Taken from TBG literature, but a model input; central gaps scale with U.
  • hybridization gamma = -38 meV
    Local hybridization strength at Γ; together with v'_* = -1.702 eV and λ = 0.3375, it is fitted to reproduce the BM band structure. Treated as an input from the THFM literature.
  • anti-Hund couplings JA, J'_A = assumed large (no number given)
    Required to select the spin-singlet doublon as the only doubly occupied state; the paper concedes the Hilbert space truncation is strictly valid only for very large JA.
assumptions (5)
  • domain assumption The topological heavy fermion model (THFM) accurately describes the low-energy physics of magic-angle TBG.
    Used as the starting model (Eq. 1). The paper states it uses the model 'simply for convenience' but the conclusions are meant to generalize.
  • ad hoc to paper The f-orbital Hilbert space can be truncated to f1+, f2+, f3+ valences with a single spin-singlet doublon.
    Introduced in 'Effective model with restricted Hilbert space'; justified only in the limit of very large JA, and the paper conjectures robustness.
  • domain assumption Renormalized mean-field theory with slave-boson condensation (gγ = sqrt(1-ns-nt)) correctly captures the Gutzwiller projection.
    Used in 'Renormalized mean field theory'; no controlled small parameter except γ/U in the ancilla equivalence.
  • ad hoc to paper The phenomenological κ potential is present and has magnitude ~0.8 at ν=-2.
    Introduced in Eq. (3); tuned, not derived, and affects the quantitative mixed-valence fraction.
  • domain assumption The ground state is symmetric and does not break any symmetry at ν=-2.
    The paper constructs only a symmetric ansatz, while Hartree-Fock studies (Refs. [14-17]) find isospin-polarized or Kekulé-spiral states at ν=-2.
invented entities (2)
  • s and t slave fermions (singlon and triplon partons)
    purpose: Express the physical f-electron operator and the ψ composite within the restricted Hilbert space and enable a fermionic mean-field theory.
    They are auxiliary partons with no direct physical observable; their validity relies on the Hilbert space truncation.
  • Composite fermion ψ (ancilla fermion)
    purpose: Describes the low-energy hole-doped excitation at Γ and is identified as the microscopic content of the ancilla fermion in Ref. [34].
    ψ is defined by construction as the combination orthogonal to f (Eq. 6). It is stated to have vanishing spectral weight in ARPES, so it is not directly observable; its existence is model-dependent.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mixed valence Mott insulator and composite excitation in twisted bilayer graphene." pith.science (2026). https://pith.science/paper/OGV53QWQ

@misc{pith2026250700139,
  author       = {Pith},
  title        = {Pith review of: Mixed valence Mott insulator and composite excitation in twisted bilayer graphene},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OGV53QWQ}},
  note         = {Machine review of arXiv:2507.00139}
}
abstract

Interplay of strong correlation and flat topological band has been a central problem in moir\'e systems such as the magic angle twisted bilayer graphene (TBG). Recent studies show that Mott-like states may still be possible in TBG despite the Wannier obstruction. However, the nature of such unconventional states is still not well understood. In this work we construct the ground state wavefunction and exotic excitations of a symmetric correlated semimetal or insulator at even integer filling using a parton mean field theory of the topological heavy fermion model. We label the valence of the $f$ orbital based on its occupation $n_f$. At $\nu=-2$, we show that the $f$ orbital is not in the simple $f^{2+}$ valence expected from a trivial Mott localization. Instead, around $1/3$ of AA sites are self doped, with holes entering the $c$ orbitals away from AA sites. As a result, the $f$ orbital is in a superposition of $f^{2+}$ and $f^{3+}$ valences and should not be viewed as local moment. We dub the phase as \textit{mixed valence Mott insulator}. This unconventional insulator has a large hybridization $\langle c^\dagger f \rangle\neq 0$ and is sharply distinct from the usual `kondo breakdown' picture. In most of the momentum space away from the $\Gamma$ point, there is a Mott gap equal to the Hubbard $U$. At the $\Gamma$ point, we have a `charge transfer gap' much smaller than $U$. In particular, the top of the lower band is dominated by a composite excitation, which is a linear combination of $|f^{1+}\rangle\langle f^{2+}|$ and $|f^{2+}\rangle\langle f^{3+}|$ with a sign structure such that it is orthogonal to the microscopic $f$ operator. At $\nu=0$, similar approach leads to a Mott semimetal. We hope this work will inspire more explorations of the Anderson models with a large hybridization, a regime which may host new physics beyond the familiar Kondo or heavy fermion systems.

Figures

Figures reproduced from arXiv: 2507.00139 by the authors.

Figure 1
Figure 1. FIG. 1. Illustration of our mean field description of the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Typical mean field band structure at [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. (a) Γ point gap ∆ [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Band structure for semi-metals at charge neutrality [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Typical mean field band structure calculated at [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 9
Figure 9. Figure 9: FIG. 9. Band structure as a function of [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Mean field band structure calculated at [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Trion Excitations in Twisted Bilayer Graphene: A Quantum Monte Carlo Study

    cond-mat.str-el 2026-07 conditional novelty 7.0 of 10

    Finite-temperature QMC of twisted bilayer graphene finds gapless 'Dirac trion' three-particle excitations in the normal state, exactly orthogonal to electrons at the Γ point.

  2. Monte Carlo Studies of Twisted Bilayer Graphene: Strain and Thermal Fluctuations

    cond-mat.str-el 2026-07 conditional novelty 7.0 of 10

    Sign-problem-free quantum Monte Carlo shows that strained twisted bilayer graphene at charge neutrality hosts a KIVC insulator bounded by Dirac and anisotropic semimetals, with a 15–40 K entropy plateau from a Mott-li...

Reference graph

Works this paper leans on

53 extracted references · 30 canonical work pages · cited by 2 Pith papers

  1. [34]

    P. J. Ledwith, J. Dong, A. Vishwanath, and E. Khalaf, Nonlocal Moments in the Chern Bands of Twisted Bilayer Graphene, arXiv preprint , 1 (2024), arXiv:2408.16761

    arXiv 2024

  2. [1]

    Y. Cao, V. Fatemi, S. Fang, K. Watanabe, T. Taniguchi, E. Kaxiras, and P. Jarillo-Herrero, Unconventional super- conductivity in magic-angle graphene superlattices, Na- ture 556, 43 (2018)

    2018

  3. [2]

    Y. Cao, V. Fatemi, A. Demir, S. Fang, S. L. Tomarken, J. Y. Luo, J. D. Sanchez-Yamagishi, K. Watanabe, T. Taniguchi, E. Kaxiras, R. C. Ashoori, and P. Jarillo- Herrero, Correlated insulator behaviour at half-filling in magic-angle graphene superlattices, Nature 556, 80 (2018), arXiv:1802.00553

    arXiv 2018

  4. [3]

    It corresponds to a microscopic trion operator[34, 38], but it is easier to just view it as a dif- ferent linear combination of s and t† operator within our subspace. Renormalized mean field theory We can perform a mean field calculation by relaxing the Gutzwiller con- straint PG, with a cost of including a renormalization factor [42, 43] for the γ term. ...

    work page

  5. [4]

    Yankowitz, S

    M. Yankowitz, S. Chen, H. Polshyn, Y. Zhang, K. Watan- abe, T. Taniguchi, D. Graf, A. F. Young, and C. R. Dean, Tuning superconductivity in twisted bilayer graphene, Science 363, 1059 (2019)

    2019

  6. [5]

    X. Lu, P. Stepanov, W. Yang, M. Xie, M. A. Aamir, I. Das, C. Urgell, K. Watanabe, T. Taniguchi, G. Zhang, A. Bachtold, A. H. MacDonald, and D. K. Efetov, Su- perconductors, orbital magnets and correlated states in magic-angle bilayer graphene, Nature 574, 653 (2019), arXiv:1903.06513

    arXiv 2019

  7. [6]

    Stepanov, I

    P. Stepanov, I. Das, X. Lu, A. Fahimniya, K. Watanabe, T. Taniguchi, F. H. L. Koppens, J. Lischner, L. Levitov, and D. K. Efetov, Untying the insulating and supercon- ducting orders in magic-angle graphene, Nature 583, 375 (2020), arXiv:1911.09198

    arXiv 2020

  8. [7]

    Y. Cao, D. Rodan-Legrain, J. M. Park, N. F. Q. Yuan, K. Watanabe, T. Taniguchi, R. M. Fernandes, L. Fu, and P. Jarillo-Herrero, Nematicity and competing orders in superconducting magic-angle graphene, Science 372, 264 (2021), arXiv:2004.04148

    work page Pith review arXiv 2021

Show all 53 references
  1. [8]

    X. Liu, Z. Wang, K. Watanabe, T. Taniguchi, O. Vafek, and J. I. A. Li, Tuning electron correlation in magic- angle twisted bilayer graphene using Coulomb screening, Science 371, 1261 (2021), arXiv:2003.11072

    2021 arXiv

  2. [9]

    H. S. Arora, R. Polski, Y. Zhang, A. Thomson, Y. Choi, H. Kim, Z. Lin, I. Z. Wilson, X. Xu, J.-H. Chu, et al., Superconductivity in metallic twisted bilayer graphene stabilized by wse2, Nature 583, 379 (2020)

    2020

  3. [10]

    H. C. Po, L. Zou, A. Vishwanath, and T. Senthil, Origin of Mott Insulating Behavior and Superconductivity in Twisted Bilayer Graphene, Physical Review X 8, 031089 (2018), arXiv:1803.09742

    2018 arXiv

  4. [11]

    Tarnopolsky, A

    G. Tarnopolsky, A. J. Kruchkov, and A. Vishwanath, Ori- gin of Magic Angles in Twisted Bilayer Graphene, Physi- cal Review Letter 122, 106405 (2019), arXiv:2111.10018

    2019 arXiv

  5. [12]

    J. Ahn, S. Park, and B.-j. Yang, Failure of Nielsen- Ninomiya Theorem and Fragile Topology in Two- Dimensional Systems with Space-Time Inversion Symme- 6 try: Application to Twisted Bilayer Graphene at Magic Angle, Physical Review X 9, 021013 (2019)

    2019

  6. [13]

    P. J. Ledwith, E. Khalaf, and A. Vishwanath, Strong coupling theory of magic-angle graphene: A pedagogi- cal introduction, Annals of Physics 435, 168646 (2021), arXiv:2105.08858

    2021 arXiv

  7. [14]

    Z.-d. Song, B. Lian, N. Regnault, and B. A. Bernevig, Twisted bilayer graphene. II. Stable symmetry anomaly, Physical Review B 103, 205412 (2021)

    2021

  8. [15]

    Bultinck, E

    N. Bultinck, E. Khalaf, S. Liu, S. Chatterjee, A. Vish- wanath, and M. P. Zaletel, Ground state and hidden symmetry of magic-angle graphene at even integer fill- ing, Physical Review X 10, 031034 (2020)

    2020

  9. [16]

    Y. H. Kwan, G. Wagner, T. Soejima, M. P. Zaletel, S. H. Simon, S. A. Parameswaran, and N. Bultinck, Kekul´ e spi- ral order at all nonzero integer fillings in twisted bilayer graphene, Physical Review X 11, 041063 (2021)

    2021

  10. [17]

    D. E. Parker, T. Soejima, J. Hauschild, M. P. Zaletel, and N. Bultinck, Strain-induced quantum phase tran- sitions in magic-angle graphene, Physical review letters 127, 027601 (2021)

    2021

  11. [18]

    Wagner, Y

    G. Wagner, Y. H. Kwan, N. Bultinck, S. H. Simon, and S. Parameswaran, Global phase diagram of the normal state of twisted bilayer graphene, Physical review letters 128, 156401 (2022)

    2022

  12. [19]

    Rozen, J

    A. Rozen, J. M. Park, U. Zondiner, Y. Cao, D. Rodan- Legrain, T. Taniguchi, K. Watanabe, Y. Oreg, A. Stern, E. Berg, et al., Entropic evidence for a pomeranchuk ef- fect in magic-angle graphene, Nature 592, 214 (2021)

    2021

  13. [20]

    Saito, F

    Y. Saito, F. Yang, J. Ge, X. Liu, T. Taniguchi, K. Watan- abe, J. Li, E. Berg, and A. F. Young, Isospin pomer- anchuk effect in twisted bilayer graphene, Nature 592, 220 (2021)

    2021

  14. [21]

    Song and B

    Z.-D. Song and B. A. Bernevig, Magic-Angle Twisted Bilayer Graphene as a Topological Heavy Fermion Problem, Physical Review Letter 129, 047601 (2022), arXiv:2111.05865

    2022 arXiv

  15. [22]

    C˘ alug˘ aru, M

    D. C˘ alug˘ aru, M. Borovkov, L. L. H. Lau, P. Coleman, Z.- d. Song, and B. A. Bernevig, Twisted bilayer graphene as topological heavy fermion: II. Analytical approximations of the model parameters, Low Temperature Physics 49, 640 (2023), arXiv:2303.03429

    2023 arXiv

  16. [23]

    J. Yu, M. Xie, B. A. Bernevig, and S. Das Sarma, Magic- angle twisted symmetric trilayer graphene as a topo- logical heavy-fermion problem, Physical Review B 108, 035129 (2023)

    2023

  17. [24]

    H. Hu, B. A. Bernevig, and A. M. Tsvelik, Kondo Lat- tice Model of Magic-Angle Twisted-Bilayer Graphene: Hund’s Rule, Local-Moment Fluctuations, and Low- Energy Effective Theory, Physical Review Letters 131, 026502 (2023)

    2023

  18. [25]

    Herzog-Arbeitman, J

    J. Herzog-Arbeitman, J. Yu, D. C˘ alug˘ aru, H. Hu, N. Regnault, C. Liu, O. Vafek, P. Coleman, A. Tsve- lik, Z.-d. Song, and B. A. Bernevig, Topological Heavy Fermion Principle For Flat (Narrow) Bands With Con- centrated Quantum Geometry, arXiv preprint , 1 (2024), arXiv:2404.07253

    2024 arXiv

  19. [26]

    Zhou, Y.-J

    G.-D. Zhou, Y.-J. Wang, N. Tong, and Z.-D. Song, Kondo phase in twisted bilayer graphene, Physical Review B 109, 045419 (2024)

    2024

  20. [27]

    H. Hu, G. Rai, L. Crippa, J. Herzog-Arbeitman, D. C˘ alug˘ aru, T. Wehling, G. Sangiovanni, R. Va- lent ´ ı, A. M. Tsvelik, and B. A. Bernevig, Symmetric Kondo Lattice States in Doped Strained Twisted Bilayer Graphene, Physical Review Letters 131, 166501 (2023), arXiv:2301.04673

    2023 arXiv

  21. [28]

    Chou and S

    Y.-Z. Chou and S. Das Sarma, Kondo Lattice Model in Magic-Angle Twisted Bilayer Graphene, Physical Review Letters 131, 026501 (2023), arXiv:2211.15682

    2023 arXiv

  22. [29]

    Wang, G.-D

    Y.-J. Wang, G.-D. Zhou, B. Lian, and Z.-D. Song, Elec- tron phonon coupling in the topological heavy fermion model of twisted bilayer graphene, arXiv preprint , 1 (2024), arXiv:2407.11116

    2024 arXiv

  23. [30]

    L. L. H. Lau and P. Coleman, Topological Mixed Valence Model for Twisted Bilayer Graphene, Physical Review X 15, 021028 (2025), arXiv:2303.02670

    2025 arXiv

  24. [31]

    G. Rai, L. Crippa, D. C˘ alug˘ aru, H. Hu, F. Paoletti, L. de’ Medici, A. Georges, B. A. Bernevig, R. Valent ´ ı, G. San- giovanni, and T. Wehling, Dynamical Correlations and Order in Magic-Angle Twisted Bilayer Graphene, Phys- ical Review X 14, 031045 (2024)

    2024

  25. [32]

    S. Youn, B. Goh, G.-D. Zhou, Z.-D. Song, and S.- S. B. Lee, Hundness in twisted bilayer graphene: cor- related gaps and pairing, arXiv preprint 1, 1 (2024), arXiv:2412.03108

    2024 arXiv

  26. [33]

    J. G. Checkelsky, B. A. Bernevig, P. Coleman, Q. Si, and S. Paschen, Flat bands, strange metals and the kondo effect, Nature Reviews Materials 9, 509 (2024)

    2024

  27. [35]

    J.-Y. Zhao, B. Zhou, and Y.-H. Zhang, Ancilla the- ory of twisted bilayer graphene I: topological Mott semimetal and symmetric pseudogap metal (2025), arXiv:2502.17430

    2025 arXiv

  28. [36]

    Zhang and S

    Y.-H. Zhang and S. Sachdev, From the pseudogap metal to the Fermi liquid using ancilla qubits, Physical Review Research 2, 023172 (2020), arXiv:2001.09159

    2020 arXiv

  29. [37]

    Zhou, H.-k

    B. Zhou, H.-k. Jin, and Y.-h. Zhang, Variational wave- function for Mott insulator at finite U using ancilla qubits, arXiv preprint , 1 (2024), arXiv:2409.07512

    2024 arXiv

  30. [38]

    Hu, Z.-D

    H. Hu, Z.-D. Song, and B. A. Bernevig, Projected and Solvable Topological Heavy Fermion Model of Twisted Bilayer Graphene, arXiv preprint , 225 (2025), arXiv:2502.14039

    2025

  31. [39]

    P. J. Ledwith, A. Vishwanath, and E. Khalaf, Exotic car- riers from concentrated topology: Dirac trions as the origin of the missing spectral weight in twisted bilayer graphene, arXiv preprint arXiv:2505.08779 (2025)

    2025 arXiv

  32. [40]

    Coleman, Introduction to many-body physics (Cam- bridge University Press, 2015)

    P. Coleman, Introduction to many-body physics (Cam- bridge University Press, 2015)

    2015

  33. [41]

    Coleman, New approach to the mixed-valence prob- lem, Physical Review B 29, 3035 (1984)

    P. Coleman, New approach to the mixed-valence prob- lem, Physical Review B 29, 3035 (1984)

    1984

  34. [42]

    Bistritzer and A

    R. Bistritzer and A. H. MacDonald, Moir´ e bands in twisted double-layer graphene, Proc. Natl. Acad. Sci. 108, 12233 (2011), arXiv:1009.4203

    2011 arXiv

  35. [43]

    F. C. Zhang, C. Gros, T. M. Rice, and H. Shiba, A renormalised Hamiltonian approach to a resonant valence bond wavefunction, Superconductor Science and Tech- nology 1, 36 (1988)

    1988

  36. [44]

    Vollhardt, Normal 3He : an almost localized Fermi liquid, Reviews of Modern Physics 56, 99 (1984)

    D. Vollhardt, Normal 3He : an almost localized Fermi liquid, Reviews of Modern Physics 56, 99 (1984)

    1984

  37. [45]

    H. Yang, H. Oh, and Y.-H. Zhang, Strong pairing from a small fermi surface beyond weak coupling: Application to La3Ni2O7, Physical Review B 110, 104517 (2024)

    2024

  38. [46]

    H. Oh, H. Yang, and Y.-H. Zhang, High-temperature superconductivity from kinetic energy, arXiv preprint arXiv:2411.07292 (2024). 7 Appendix A: THFM with anti-Hund’s coupling We now present the full expression of the THFM with the anti-Hund spin coupling as introduced in Eq. (1...

    2024 arXiv

  39. [47]

    Relevant f valences at ν = −2 For the ν = −2 insulator, we are mostly interested in the f 1+, f 2+ and f 3+ valences at each AA site. Specif- ically, when the anti-Hund’s spin interaction is strong and JA > 0, J′ A > 0, the doubly occupied f 2+-orbital will be restricted to a ...

  40. [48]

    Renormalized Mean-Field theory To solve the restricted HamiltonianPGH0PG with con- straints ni;s + ni;t ≤ 1, one way is to introduce a boson d as slave particle for the vacuum, similar to the familiar slave boson theory of the t-J model. Instead, here we try to use a simpler r...

  41. [49]

    5 and 6, we further show the mean-field spec- trum calculated at smaller effective attractive potential κ = 0.0 and κ = 0.4

    Spectrum for smaller κ In Figs. 5 and 6, we further show the mean-field spec- trum calculated at smaller effective attractive potential κ = 0.0 and κ = 0.4. As a result of smaller κ, there are more holes doped into the c band than into the f band. The average valence f orbital...

  42. [50]

    Spectra for different hybridization strength. In Figs. 7 and 8, we show the spectra at different hy- bridization strengths γ for the ν = −2 insulator. The overall band structures remain similar to that in Fig. 2 of the main text. The primary difference appears near the Γ point...

  43. [51]

    Doping the correlated insulator ν = −2 We now consider lightly doping the correlated insulator at ν = −2 into a metallic state. In Fig. 10, we present the spectral function for parameters U = 20 meV, γ = 0 25 50 75 −γ 0 5∆ □ meV (a) 0 25 50 75 −γ 0 1n (b) ns nt νc nf − 2 FIG. ...

  44. [52]

    γ(k) = 0 We first consider the case with γ(k) = 0. In the ancilla approach, f strongly hybridizes with the first ancilla ψ on each site i: |Slater[c, f, ψ]⟩0 = |Slater[c1, c2]⟩0 ⊗ |Slater[f, ψ]⟩0 , |Slater[f, ψ]⟩0 = NsiteY i=1 Y α 1 2 f † i;α − √ 3 2 ψ† i;α ! |0⟩ . (F9) 13 Her...

  45. [53]

    We need to rotate the wavefunction Eq

    γ(k) ̸= 0 but γ(k) ≪ U In a more physical relevant regime with γ(k) ̸= 0, γ(k) ≪ 0, the ground state of Hancilla is still a Slater determinant, but c and f, ψare no more decou- pled. We need to rotate the wavefunction Eq. F9 at each momentum k to be the ground state of Hancill...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.