Quantum Hall edges beyond the plasma analogy

TL;DR

The paper challenges the widely used plasma analogy between quantum Hall (QH) droplets and 2D Coulomb gases (CGs) by showing that edge properties, unlike bulk properties, are generically incompatible. The core distinction is that QH physics is governed by incompressible, area-preserving (symplectic) deformations with edge velocity $v(\theta)$, whereas CGs are controlled by electrostatics and conformal maps, yielding edge correlations parameterized by the conformal angle $\alpha$. Through explicit constructions of anisotropic droplets (anisotropic disks, flower droplets with $k$ petals, and squares), the authors demonstrate that the edge correlators with coordinates $\theta$ and $\alpha$ typically do not match, except in fine-tuned cases such as an elliptical edge ($k=2$). They further quantify the impact on edge fluctuations and admittance, showing measurable differences in microwave absorption and providing a clear path for experimental verification in solid-state systems and quantum simulators. The work suggests careful consideration of edge geometry when employing plasma-analytic intuitions and points to extensions to fractional QH states and related edge-physics problems.

Abstract

We demonstrate that the widely used plasma analogy is unreliable at predicting edge properties of quantum Hall states. This discrepancy arises from a fundamental difference between quantum Hall droplets and plasmas (Coulomb gases): the former are incompressible liquids subject to area-preserving deformations, while the latter are governed by electrostatics and thus involve conformal transformations. Consequently, the plasma analogy fails at the edge, except in fine-tuned geometries, as it does not account for the emergent local edge velocity. We quantitatively show how the analogy's failure affects physical quantities, such as fluctuations of local observables and absorption rates in microwave spectroscopy, measurable in both solid-state experiments and quantum simulators.

Figures (6)

• Figure 1: QH droplets and CGs, shown in yellow along with a few level curves of the confining potential in the QH case. Both involve maps that send them on a disk $\mathbb{D}$, but with different geometric constraints, namely symplectic (area-preserving) in QH and conformal (angle-preserving) in CG. Given the same initial droplet $\Omega$, such maps do not generally coincide, leading to different angle coordinates governing the emergent long-range edge correlations: the coordinate is $\theta$ for QH, $\alpha$ for CG, and $\varphi$ in the isotropic case.
• Figure 2: Monte Carlo simulation of the norm of the total electric field $\mathbf{E}(\textbf{y}) = -\nabla_{\textbf{y}}\bigl(\braket{\Phi(\textbf{y})}^{\textbf{x}}-\braket{\Phi(\textbf{y})}-\log |\textbf{x}-\textbf{y}|\bigr)$ for $N=256$ particles and potential $W(x,y) = x^2 + y^2 - (x^3-3xy^2)/10$, shown as a function of $\textbf{y} =(y^1, y^2)$. (The potential is polynomial so as to give an anisotropic CG at minimal numerical cost, but it is otherwise unimportant.) Left: Fixed unit charge at $\textbf{x}\in\Omega$; its effect is screened over short distances. Right: Fixed unit charge at $\textbf{x}\notin\Omega$, in which case the electric potential is a long-range Green's function. In both plots, the electric field vanishes inside $\Omega$ and is discontinuous on $\partial \Omega$, signaling charge accumulation on the boundary.
• Figure 3: Two confining potentials, with their level curves as insets. The one in (a) is isotropic, so the norm \ref{['v_theta']} of the edge velocity is uniform in the resulting QH droplet. The one in (b) is anisotropic, but at least one of its equipotentials is a circle, so it gives rise to a QH droplet whose edge modes have an inhomogeneous velocity \ref{['v_theta']} despite the droplet's isotropic shape. QH physics distinguishes these two situations. By contrast, from the CG perspective, the two systems are equivalent since edge properties are determined by the droplet's shape alone.
• Figure 4: Differences between coordinates $\theta(\varphi)$ and $\alpha(\varphi)$ for edge-deformed droplets \ref{['eq:edge']} given by flower deformations \ref{['s1z']}, with $(k, \lambda) = (3, 0.2)$ (purple), $(k, \lambda) = (3, 0.4)$ (yellow), $(k, \lambda) = (2, 0.4)$ (blue), and a square \ref{['s15y']} (magenta). Equipotentials of the respective QH droplets are shown on the side.
• Figure 5: The total variance of $U(x, y) = x$ for flower droplets with $k=3$ petals and $\lambda$ ranging from $0$ to $0.4$. Main figure: The variance \ref{['s1w']} for QH droplets is compared with the CG result \ref{['t15w']} obtained from a numerical evaluation of the conformal map \ref{['ss15n']}, including an extrapolation based on Monte Carlo (MC) data for the CG at $\lambda = 0$, $0.1$, $0.2$, $0.3$, and $0.4$SM. The two match for $\lambda=0$ but not $\lambda>0$. Inset: Details of the extrapolation for $\lambda=0.3$. The variance is extrapolated by a fit of the form $a+b/\sqrt{N}$ for $N = 64$, $128$, and $256$ particles. The intercept with the $y$ axis yields the coefficient $a$, which is shown as an empty green circle and differs significantly from its QH counterpart. To further illustrate this, we include an MC point for $N=1024$, which has a larger error bar but remains consistent with a leading $1/\sqrt{N}$ correction.