Electrostatically stabilized surface flat bands in rhombohedral graphite at zero displacement field
Kryštof Kolář, Andrea F. Young, Cyprian Lewandowski · 2026
Read Electrostatically stabilized surface flat bands in rhombohedral graphite at zero displacement field by Kryštof Kolář, Andrea F. Young, Cyprian Lewandowski (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-24080.txt.
Introduction
Kolář, Young, and Lewandowski show that surface flat bands in rhombohedral graphite need no displacement field. Self-consistent, nonlinear electrostatics flattens the surface-band dispersion through a potential, raising the density of states, even at zero field. In the strong-coupling limit, electrostatics fills each momentum uniformly, yielding a surface band that is asymptotically flat without gating. At realistic strengths, flatness is tuned by the proximal gate, peaking when the band is empty at hole doping, and the regime maps onto samples of six to fifteen layers.
Catalog record
| CREATOR | Kryštof Kolář, Andrea F. Young, Cyprian Lewandowski |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52544 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.24080
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Electrostatically stabilized surface flat bands in rhombohedral graphite at zero displacement field Kryštof Kolář, Andrea F. Young, Cyprian Lewandowski Abstract Rhombohedral (ABC-stacked) multilayer graphene hosts interaction-driven phases enabled by surface flat bands at large displacement fields. In thick flakes, however, strong screening suppresses internal electric fields, raising the question of whether a flat-band regime is accessible within the same experimental paradigm. Here, we show that self-consistent, nonlinear electrostatics provides a robust alternative mechanism: even in the absence of a displacement field, a nonuniform near-surface potential flattens the surface-band dispersion and enhances the density of states. In the strong-coupling limit, electrostatics drives the system toward uniform half-filling at each momentum, yielding an asymptotically flat surface band without any gating. At realistic interaction strengths, surface-band flatness is tuned by the proximal gate, with maximal flatness achieved at hole doping when the band is empty. Combining analytic arguments with fully self-consistent calculations in a realistic model, we map the resulting low-field regime and connect to finite N ∼ 6 − 15 N\!\sim\!6-15 layered samples, providing a framework for analyzing the symmetry-broken phases observed in these systems. Our results motivate future experiments in large- N N devices and establish a low-field regime for exploring electrostatically induced flat-band physics. Rhombohedral (ABC-stacked) N -layer graphene (RNG) hosts surface-polarized bands with an exceptionally large density of states and has become a prime venue for exploring interaction-driven phases Shi et al. 2020 ; Slizovskiy et al. 2019 ; Zhang et al. 2024 ; Zhang et al. 2025 ; Otani et al. 2010 ; Henck et al. 2018 ; Zhou et al. 2021a ; Zhou et al. 2021b ; Liu et al. 2023 ; Auerbach et al. 2025 ; Han et al. 2023 ; Guo et al. 2025 ; Kumar et al. 2025 ; Deng et al. 2025 ; Qin et al. 2025 ; Zhou et al. 2024a ; Lu et al. 2025 ; Myhro et al. 2018 ; Hagymási et al. 2022 ; Han et al. 2024 ; Waters et al. 2025 ; Choi et al. 2025 ; Aronson et al. 2025 ; Han et al. 2025 ; Seo et al. 2025 ; Xu et al. 2026 . Most works have focused on few-layer devices at large displacement fields, where a single surface band is both isolated and strongly flattened, enabling, amongst others, the observation of fractional Chern physics and chiral superconductivity Lu et al. 2024 ; Waters et al. 2025 ; Choi et al. 2025 ; Aronson et al. 2025 ; Han et al. 2025 ; Seo et al. 2025 ; Xu et al. 2026 . More recently, experiments have also begun mapping a lower-displacement field in thicker samples and have found a semimetallic coexistence regime in the (n,D) plane Guo et al. 2025 ; Kumar et al. 2025 ; Deng et al. 2025 ; Zhou et al. 2024b ; Seifert et al. 2024 . This regime shows as a characteristic “diamond” (Fig. 1 a) in a double-gated geometry (Fig. 1 b) where key experimental features follow lines of constant single-gate voltage - indicating the simultaneous involvement of carriers on both surfaces Kolář et al. 2026 ; Guo et al. 2025 ; Kumar et al. 2025 ; Deng et al. 2025 . Motivated by this phenomenology and emerging experimental interest in thicker RNG samples, we study the complementary limit of large N and small (or zero) D , where both surface bands can remain relevant but become increasingly electrostatically decoupled . Figure 1: (a) Schematic phase diagram of large N RNG in the n - D plane. (b) Device schematics, showing the layer and gate electron densities. The gate electric field is denoted as E^{\text{bg}}_{\perp} . (c) Band structure for an idealized N=50 RNG model at charge neutrality. (d) Band structure and density of states (DOS) for a realistic model at N=50 for zero layer potential, U_{l}=0 . (e-g) Upper panel shows the self-consistent band structure and DOS at D=0 and: (e) negative n\approx 2n_{-} , (f) n=0 , and (g) close to full filling, n\approx 2n_{+} . Lower panel shows the corresponding layer potentials U_{l} . A central open question we would like to address is what becomes of the “flat-band” physics in the N\!\to\!\infty limit. Idealized chiral models predict a surface dispersion proportional to |\mathbf{k}|^{N} , leading to increasing “flatness” with N Guinea et al. 2006 ; Koshino and McCann 2009 ; Zhang et al. 2010 ; Heikkilä and Volovik 2011 ; Kopnin et al. 2013 ; Min and MacDonald 2008 . In realistic band structures, remote hoppings qualitatively change the long-wavelength behavior Koshino and McCann 2009 ; Zhang et al. 2010 ; McCann and Koshino 2013 ; Park et al. 2023 . Such terms generate a conventional quadratic surface-band dispersion with positive effective mass even as N\!\to\!\infty . This outcome challenges the naive expectation of an asymptotically flat band (see Fig. 1 c,d for the bands for N=50 ) Koshino and McCann 2009 ; Zhang et al. 2010 . For small N{\sim}4\!-\!6 , a high displacement field D , typically modeled by a linear potential drop between layers, restores an effective flat band Zhang et al. 2010 ; Ghazaryan et al. 2023 ; McCann and Koshino 2013 . However, for large N , a linear potential is not an adequate model for the internal potential profile Koshino 2010 ; Avetisyan et al. 2009 ; Avetisyan et al. 2010 ; Kolář et al. 2026 ; Slizovskiy et al. 2021 ; Otani et al. 2010 and generally does not produce an isolated flat band. In this work, we show that, in the large- N limit ( N\gtrsim 6 ), accurately accounting for nonlinear electrostatics can instead yield a flat surface band at D=0 (Fig. 1 e,f). Our main result is a simple electrostatic mechanism that flattens the quadratically-dispersing (Fig. 1 d) surface band in rhombohedral graphite. At charge neutrality and when the out-of-plane interaction is strong ( 1/\epsilon_{\perp}\!\to\!\infty ), the electrostatic energy is lowest if every layer stays neutral. This happens when each surface-band momentum state \mathbf{k} is half-filled, which can only occur if the mean-field particle dispersion is flat. Although this argument assumes in-plane isotropy, a similar principle works for realistic, trigonally warped dispersions in RNG. While the proposed flattening mechanism requires no gating and holds at the charge neutrality point, electrostatic gating still provides a useful tuning knob in large- N devices. The key distinction from thin stacks is, however, the screening: in a double-gated geometry (Fig. 1 b), the distant gate is almost fully screened from a given surface band Kolář et al. 2026 . Accordingly, instead of the usual experimental variables n=-n_{t}-n_{b} and D=-(n_{t}-n_{b})/\epsilon_{0} , the bottom surface, for example, is controlled primarily by n_{b} with n_{t} being largely irrelevant Kolář et al. 2026 . This observation lets us follow the bottom-surface (or top-surface) evolution along essentially any trajectory in the (n,D) plane as the corresponding n_{t} (or n_{b} ) coordinate is a spurious parameter; to specifically highlight the contrast with few-layer devices, we thus focus on the D=0 line. From this viewpoint, the bottom gate sets both the surface filling and the inner electric field. As a result, the self-consistent potential is highly non-linear: the potential drops decay from the surface