Mixed-Stacked Pentalayer Graphene

• Mixed-stacked pentalayer graphene is defined by the coexistence of Bernal and rhombohedral stacking sequences that break inversion symmetry and induce intrinsic layer polarization.
• Its electronic structure features hybridized cubic and parabolic low-energy bands, leading to asymmetric displacement-field responses and multiple Lifshitz transitions.
• The system enables unusual quantum Hall phenomena, including an ultra-low-field ν=-6 state and displacement-tunable even-denominator fractional quantum Hall effects.

Searching arXiv for papers on mixed-stacked pentalayer graphene and related mixed-stacking graphene phenomena. Mixed-stacked pentalayer graphene denotes five-layer graphene in which Bernal and rhombohedral stacking motifs coexist within a single crystal, so that the low-energy spectrum combines distinct chiral sectors rather than reducing to a single Bernal-like or rhombohedral-like hierarchy. In the experimentally studied non-centrosymmetric sequence ABCBC, the structure can be regarded as an ABC trilayer coupled to an AB bilayer, yielding a “3+2” chiral decomposition with coexisting cubic and parabolic bands that hybridize (Liu et al., 18 May 2025). More generally, mixed-stacked five-layer graphene is important because broken inversion and mirror symmetries can permit out-of-plane polarization, intrinsic layer asymmetry, and transport responses unavailable in centrosymmetric pentalayers (Garcia-Ruiz et al., 2023). Recent work has established ABCBC as a platform for intrinsic layer polarization, multiple flatbands, Lifshitz transitions, unusual Landau-level degeneracy patterns, ultra-low-field quantum Hall phenomena, and, at higher fields, a cascade of even-denominator fractional quantum Hall states (Liu et al., 18 May 2025, Sha et al., 28 Jul 2025).

1. Definition, stacking taxonomy, and symmetry

Mixed stacking in few-layer graphene means that a single flake contains both Bernal and rhombohedral sequences, so global inversion or mirror symmetry can be broken and an out-of-plane polarization is no longer symmetry-forbidden (Garcia-Ruiz et al., 2023). For pentalayer graphene, the general mixed-stacking problem can be classified either by explicit letter sequences such as ABABA, ABCAB, ABCBC, ABACA, and related permutations, or by Bernal-section decompositions separated by stacking faults (Koshino et al., 2012). Within this taxonomy, ABCBC corresponds to the decomposition $(1,3)$, while ABACA corresponds to $(2,2)$, and pure Bernal ABABA to $(4)$ (Koshino et al., 2012).

Among the six pentalayer sequences discussed in the 2025 transport study, ABCBC belongs to the “none” symmetry group, with no inversion and no mirror symmetry (Liu et al., 18 May 2025). This absence of inversion and mirror symmetry is central to its phenomenology: atomic sites in different layers experience different chemical environments, producing layer-dependent onsite energies even at zero displacement field $D$, a built-in internal electric field, and an intrinsic band gap at charge neutrality (Liu et al., 18 May 2025). By contrast, centrosymmetric multilayers such as ABABA and ABCBA obey $R_{xx}(n,D)=R_{xx}(n,-D)$, do not exhibit an intrinsic $D=0$ gap from built-in polarization, and tune gaps symmetrically under $D$ (Liu et al., 18 May 2025).

A related but distinct symmetry classification appears in the weak-ferroelectric analysis of mixed-stacking pentalayers built as $n{\rm ABA}m$ twins with $n+m=2$ (Garcia-Ruiz et al., 2023). In that framework, asymmetric five-layer twins such as 2ABA0 and 0ABA2 are polar, while the symmetric 1ABA1 remains nonpolar because a $z\to -z$ mirror or inversion symmetry survives (Garcia-Ruiz et al., 2023). This suggests that polarity in five-layer graphene is not unique to ABCBC, but rather a broader consequence of mixed stacking whenever inversion and mirror symmetry are simultaneously broken.

2. Electronic structure and effective descriptions

In ABCBC-stacked pentalayer graphene, the low-energy spectrum contains both an ABC-like cubic band and an AB-like parabolic band that hybridize (Liu et al., 18 May 2025). The constituent effective Hamiltonians near $(2,2)$0 are given in two-band form. For AB bilayer graphene,

$(2,2)$1

where $(2,2)$2, $(2,2)$3, and $(2,2)$4 (Liu et al., 18 May 2025). For ABC trilayer graphene,

$(2,2)$5

with $(2,2)$6, modified by $(2,2)$7, $(2,2)$8, and $(2,2)$9 (Liu et al., 18 May 2025).

The coupled ABCBC system is described by a $(4)$0 block Hamiltonian

$(4)$1

where $(4)$2 encodes interface hybridization across the 3|4 interface, mediated mainly by $(4)$3 and $(4)$4 (Liu et al., 18 May 2025). To leading order in $(4)$5,

$(4)$6

In the absence of hybridization, the dispersions reduce to $(4)$7 and $(4)$8; hybridization repels crossings, produces avoided gaps, and generates multiple low-energy flatbands whose extrema underpin Lifshitz transitions and large Berry curvature (Liu et al., 18 May 2025).

The broader mixed-stacking theory developed earlier used a minimal $(4)$9–$D$0 model and a decomposition into Bernal sections separated by faults (Koshino et al., 2012). In that scheme, low-energy eigenstates are mostly localized in each Bernal section, and the spectrum is approximated by a collection of spectra of independent sections (Koshino et al., 2012). That analytical construction classifies bands as linear, quadratic, or cubic according to the section content. For ABCBC $D$1, the minimal model predicts no even sections, hence no Dirac cone, no $D$2-type quadratic band, and no cubic $D$3-$D$4 pair; only a nearly flat $D$5 boundary mode may remain around zero energy (Koshino et al., 2012). The 2025 ABCBC transport study instead finds a hybridized parabolic-plus-cubic low-energy manifold once realistic Slonczewski–Weiss–McClure couplings, hBN-induced onsite terms, and interface hybridization are included (Liu et al., 18 May 2025). This contrast reflects the difference between the idealized $D$6–$D$7 section model and the full SWMcC-plus-DFT description.

The modeling in the transport experiment uses $D$8 Å, $D$9 Å, an hBN-induced on-site term of 18 meV, and a field-induced interlayer potential step $R_{xx}(n,D)=R_{xx}(n,-D)$0 referenced to the third graphene layer and shifted by 5 meV to match experiment (Liu et al., 18 May 2025). Typical graphite values quoted are $R_{xx}(n,D)=R_{xx}(n,-D)$1 eV, $R_{xx}(n,D)=R_{xx}(n,-D)$2 eV, $R_{xx}(n,D)=R_{xx}(n,-D)$3 eV, and $R_{xx}(n,D)=R_{xx}(n,-D)$4 eV (Liu et al., 18 May 2025).

3. Intrinsic layer polarization and weak ferroelectricity

The defining consequence of non-centrosymmetry in ABCBC is intrinsic layer polarization. Because layer onsite energies differ even at $R_{xx}(n,D)=R_{xx}(n,-D)$5, the trilayer and bilayer blocks experience opposite internal fields and opposite gaps (Liu et al., 18 May 2025). Experimentally, an Arrhenius analysis of $R_{xx}(n,D)=R_{xx}(n,-D)$6 yields a finite intrinsic gap $R_{xx}(n,D)=R_{xx}(n,-D)$7 meV at $R_{xx}(n,D)=R_{xx}(n,-D)$8 (Liu et al., 18 May 2025). A polarization order parameter is introduced as

$R_{xx}(n,D)=R_{xx}(n,-D)$9

with alternating $D=0$0 chosen to capture the trilayer-versus-bilayer dipoles; nonzero $D=0$1 at $D=0$2 signals spontaneous layer polarization originating from the non-centrosymmetric stacking (Liu et al., 18 May 2025).

A closely related literature frames such behavior as elemental weak ferroelectricity in mixed-stacked few-layer graphene (Garcia-Ruiz et al., 2023). There the polarization is primarily electronic, arising from registry-dependent redistribution of $D=0$3-electron charge density in the presence of symmetry-breaking interlayer couplings and an asymmetric twin boundary, with no ionic displacement (Garcia-Ruiz et al., 2023). The polarization is computed from layer densities as

$D=0$4

with

$D=0$5

and screening enters via self-consistent interlayer potential differences with effective out-of-plane dielectric permittivity $D=0$6 (Garcia-Ruiz et al., 2023). For five-layer twins, the predicted polarization is “weak,” of order $D=0$7–$D=0$8 e/$D=0$9m, corresponding to $D$0 cm$D$1 transferred between outer surfaces (Garcia-Ruiz et al., 2023).

The transport results on ABCBC do not report ferroelectric hysteresis, and the weak-ferroelectric study notes that no ferroelectric hysteresis is observed in the ABCBC experiment, likely due to single-domain samples (Liu et al., 18 May 2025). Domain walls or opposite-stacking seeds such as ABABC inclusions are suggested as possible ingredients required for switchable ferroelectricity as in mixed tetralayers (Liu et al., 18 May 2025). A plausible implication is that intrinsic layer polarization in ABCBC should be viewed as established, whereas truly switchable ferroelectric behavior in mixed-stacked pentalayers remains conditional on domain structure and sliding pathways rather than already demonstrated in the same form.

4. Displacement-field response, band alignment, and Fermi-surface reconstruction

The displacement field $D$2 acts on ABCBC through layer-dependent potentials satisfying $D$3, referenced to the middle layer (Liu et al., 18 May 2025). Because the stack lacks inversion and mirror symmetry, external $D$4 adds to or subtracts from the built-in fields, producing a strongly asymmetric response of the gap and band alignment (Liu et al., 18 May 2025).

For $D$5, the gap first grows and then decreases, giving a nonmonotonic dependence. For $D$6, the gap rapidly closes (Liu et al., 18 May 2025). This asymmetry is forbidden in centrosymmetric ABA and ABC trilayers, which exhibit $D$7 symmetry under $D$8, and therefore serves as a hallmark of ABCBC (Liu et al., 18 May 2025). The asymmetry is not merely spectroscopic: for $D$9, the AB bilayer conduction band overlaps the ABC trilayer valence band, producing a two-carrier regime with $n{\rm ABA}m$0 and $n{\rm ABA}m$1 magnetoresistance at 6 T, whereas for $n{\rm ABA}m$2 the bands remain separated and the two-carrier signature is absent (Liu et al., 18 May 2025).

The same tuning of $n{\rm ABA}m$3 and carrier density drives multiple Lifshitz transitions in the Fermi-surface topology. Landau-level maps at $n{\rm ABA}m$4 T reveal several regions with distinct degeneracies: region II has degeneracy 4, region III has degeneracy 8, and region IV on the hole side at $n{\rm ABA}m$5 has degeneracy 12; resistive ridges in region V mark Lifshitz transitions (Liu et al., 18 May 2025). Band calculations show four topology changes as the chemical potential sweeps through the hybridized flatbands under $n{\rm ABA}m$6, consistent with the measured ridges and degeneracy changes (Liu et al., 18 May 2025). Trigonal warping in both $n{\rm ABA}m$7 and $n{\rm ABA}m$8 sectors, together with their hybridization, is essential to the multi-pocket regimes (Liu et al., 18 May 2025).

More generally, trigonal warping in the ABC and AB blocks converts circular isoenergy contours into triangular or multi-pocket structures (Liu et al., 18 May 2025). This provides a direct microscopic explanation for why LL degeneracies change between 4, 8, and 12 as the Fermi level crosses different topological regimes of the hybridized multiband system (Liu et al., 18 May 2025).

5. Berry curvature, Landau quantization, and quantum Hall structure

For a generic two-band Hamiltonian $n{\rm ABA}m$9, the Berry curvature of the lower band is

$n+m=2$0

In ABCBC, broken inversion and mirror symmetry plus inter-block hybridization produce sizable Berry curvature near band edges and avoided crossings, while trigonal warping concentrates it into valley-contrasting hot spots around $n+m=2$1 and $n+m=2$2 (Liu et al., 18 May 2025). This supports anomalous or valley Hall responses and facilitates weak-field Chern insulating behavior when valley degeneracy is lifted (Liu et al., 18 May 2025).

Landau quantization inherits the coexistence of $n+m=2$3 and $n+m=2$4 chiral sectors. The general chiral scaling is

$n+m=2$5

For AB bilayer-like states, $n+m=2$6, with a twofold orbital $n+m=2$7 zero-energy degeneracy per valley and spin. For ABC trilayer-like states, $n+m=2$8, with a threefold orbital $n+m=2$9 zero-energy degeneracy per valley and spin, giving a 12-fold zero-energy Landau level after including spin and valley (Liu et al., 18 May 2025). In ABCBC, hybridization of these sectors yields LL fans with degeneracy 4, 8, and 12 depending on density and $z\to -z$0 (Liu et al., 18 May 2025).

A particularly notable observation is the $z\to -z$1 quantum Hall state at exceptionally low magnetic field. At $z\to -z$2, the first developed Hall plateau is $z\to -z$3 already by $z\to -z$4 mT; $z\to -z$5 reaches $z\to -z$6 and $z\to -z$7 drops at $z\to -z$8 cm$z\to -z$9 and $(2,2)$00 mK (Liu et al., 18 May 2025). The state survives for $(2,2)$01 between about $(2,2)$02 and $(2,2)$03 V/nm and, for $(2,2)$04 V/nm, remains the first LL on both electron and hole sides (Liu et al., 18 May 2025).

Two origin scenarios are considered compatible with the data. One is LL quantization from the ABC-like cubic band, whose 12-fold zero-energy LL can generate the earliest robust plateau, with partial hybridization or symmetry breaking reducing the first resolved sequence to $(2,2)$05 (Liu et al., 18 May 2025). The other is a weak-field Chern insulator via spontaneous valley polarization, in which built-in internal fields lift valley degeneracy even at $(2,2)$06, and a small $(2,2)$07 selects a valley to yield finite Chern number and $(2,2)$08 (Liu et al., 18 May 2025). The data are explicitly stated to be compatible with either scenario.

6. Correlated states, domain physics, and experimental realization

ABCBC domains were identified by scanning near-field optical microscopy on exfoliated pentalayer graphene; mixed domains appear in $(2,2)$09 of flakes and are smaller than pure Bernal or rhombohedral domains (Liu et al., 18 May 2025). Selected regions were isolated by AFM cutting, encapsulated by hBN, and fabricated into dual-gated Hall bars with one-dimensional edge contacts (Liu et al., 18 May 2025). Transport measurements were carried out at $(2,2)$10 K in a VTI system and down to 14 mK in a dilution refrigerator, with $(2,2)$11 and $(2,2)$12 controlled by top and back gates (Liu et al., 18 May 2025). DFT calculations used VASP with PBE-GGA, $(2,2)$13 Å, and $(2,2)$14 Å, and a SWMcC model was fitted to ab initio bands near $(2,2)$15 (Liu et al., 18 May 2025).

A weak hBN–graphene moiré of period $(2,2)$16 nm is present in one device, but its effects are reported to be weak and negligible for the transport phenomena emphasized in the ABCBC study (Liu et al., 18 May 2025). The later fractional quantum Hall work reinforces this point: one device shows Brown–Zak oscillations and Hofstadter features only near LL crossings, while a second device without moiré shows a near-identical high-field phase diagram, indicating that the half-filled phases are intrinsic to ABCBC-5LG (Sha et al., 28 Jul 2025).

At higher magnetic fields, mixed-stacked pentalayer graphene exhibits a cascade of even-denominator fractional quantum Hall states at $(2,2)$17, $(2,2)$18, $(2,2)$19, $(2,2)$20, and $(2,2)$21, interwoven with conventional odd-denominator Jain states (Sha et al., 28 Jul 2025). These states arise within a hybridized zeroth Landau level that inherits two zero-mode orbitals from AB and three from ABC, for up to 20-fold degeneracy before interactions and symmetry breaking (Sha et al., 28 Jul 2025). Tuning $(2,2)$22 shifts the active half-filled LL between two intra-ZLL branches: a more AB-like branch hosting $(2,2)$23 and $(2,2)$24 near $(2,2)$25–0.5 V/nm, and a more ABC-like branch hosting $(2,2)$26, $(2,2)$27, and $(2,2)$28 near $(2,2)$29–0.36 V/nm (Sha et al., 28 Jul 2025).

Exact-diagonalization including Coulomb interactions, LL mixing, and $(2,2)$30 finds a sixfold quasi-degenerate ground-state manifold at the relevant half fillings, described as the hallmark of Moore–Read physics on a torus (Sha et al., 28 Jul 2025). Chiral-graviton spectroscopy distinguishes Pfaffian versus anti-Pfaffian tendencies, with $(2,2)$31, $(2,2)$32, and $(2,2)$33 leaning anti-Pfaffian and $(2,2)$34, $(2,2)$35 leaning Pfaffian (Sha et al., 28 Jul 2025). This suggests that mixed-stacked pentalayer graphene is not only a multiband quantum Hall system but also a tunable platform for displacement-field control of non-Abelian candidate phases.

Spatial inhomogeneity and mixed-domain networks provide another experimental dimension. STM on few-layer graphene on mica observed triangular networks of partial dislocations separating ABA and ABC stacked domains, with stacking-specific LDOS signatures and a pronounced peak at about $(2,2)$36 eV above the Fermi level exclusively in ABA areas (Hattendorf et al., 2012). No pentalayer was measured explicitly in that work, but the ABA–ABC contrast and partial-dislocation network are described as generic to tri- and multilayer graphene and therefore expected to persist in five-layer systems including locally mixed ABA|ABC coexistence (Hattendorf et al., 2012). This suggests that mixed-stacked pentalayers may often need to be understood not only as ideal single-domain crystals but also as mesoscale networks of stacking domains, solitons, and domain-wall scattering channels.

7. Relation to other pentalayer stackings and open questions

The physical behavior of mixed-stacked pentalayer graphene depends sharply on stacking sequence. The most relevant contrasts stated in the source literature are summarized below.

Stacking class	Symmetry/property	Low-energy consequence
ABABA, ABCBA, ABCAB, ABACA	Mirror or inversion symmetric	$(2,2)$37; no intrinsic $(2,2)$38 gap from built-in polarization (Liu et al., 18 May 2025)
ABCBC	No inversion, no mirror	Intrinsic layer polarization, $(2,2)$39 meV at $(2,2)$40, asymmetric gap response, 4/8/12 LL degeneracies (Liu et al., 18 May 2025)
2ABA0 / 0ABA2 vs 1ABA1	Asymmetric twins polar; symmetric twin nonpolar	Weak electronic $(2,2)$41 for asymmetric twins; $(2,2)$42 for 1ABA1 (Garcia-Ruiz et al., 2023)

Pure rhombohedral pentalayer graphene is described as having a strongly chiral $(2,2)$43 band with ultra-flat dispersion and inversion symmetry that allows symmetric gap tuning under $(2,2)$44 (Liu et al., 18 May 2025). Centrosymmetric mixed stacks such as ABABA and ABCBA retain symmetric $(2,2)$45-dependence and lack the intrinsic $(2,2)$46 polarization gap of ABCBC (Liu et al., 18 May 2025). ABCAC, although also non-centrosymmetric, would lose its gap rapidly for both $(2,2)$47; the experimentally observed asymmetric persistence for $(2,2)$48 is therefore used to rule out ABCAC in the studied devices (Liu et al., 18 May 2025).

Several open questions are stated explicitly. Distinguishing between LL-origin and Chern-insulator-origin mechanisms for the ultra-low-field $(2,2)$49 state requires complementary probes such as nonlocal transport, edge-state measurements, STM, or optical circular dichroism (Liu et al., 18 May 2025). Further theory is needed to refine the effective coupling matrices $(2,2)$50 from SWMcC fits and quantify Berry-curvature distributions (Liu et al., 18 May 2025). For weak ferroelectricity, the role of screening, substrate dielectric environment, doping, disorder, and domain walls remains central, and the available predictions emphasize strong sensitivity to these factors (Garcia-Ruiz et al., 2023). In the high-field regime, the precise spin and valley polarization of the active subspace, the detailed pairing channel of the even-denominator states, and the impact of LL mixing and intrinsic particle–hole symmetry breaking remain unresolved (Sha et al., 28 Jul 2025).

Taken together, the current literature presents mixed-stacked pentalayer graphene as a family of five-layer graphene crystals in which stacking faults, hybridized chiral sectors, and broken spatial symmetries reorganize the electronic structure far beyond the standard Bernal-versus-rhombohedral dichotomy. In the specific ABCBC realization, intrinsic layer polarization, asymmetric displacement-field response, multiband flatband transport, anomalously low-field integer quantum Hall behavior, and displacement-tunable even-denominator fractional quantum Hall states place the system at the intersection of band topology, spontaneous symmetry breaking, and correlated quantum Hall physics (Liu et al., 18 May 2025, Sha et al., 28 Jul 2025).