Ferroelectric switching of interfacial dipoles in $α$-RuCl$_3$/graphene heterostructure

TL;DR

The paper addresses the challenge of achieving electrically switchable ferroelectricity in atomically thin van der Waals heterostructures. It introduces a graphene/hBN/α-RuCl3 stack with an ultrathin spacer to modulate interfacial charge transfer while maintaining strong electrostatic coupling. A robust, ferroelectric-like hysteresis in transport emerges near 30 K and is controllable via top and back gates, with long-term retention and magnetic-field independence, indicating an electrostatic interfacial dipole mechanism. These results establish a route to gate-tunable ferroelectric phenomena in van der Waals heterostructures and open avenues to study interfacial polarization and temperature-tuned barrier crossing at the atomic scale.

Abstract

We demonstrate electrically switchable, non-volatile dipoles in graphene/thin hBN/$α$-RuCl$_3$ heterostructures, stabilized purely by interfacial charge transfer across an atomically thin dielectric barrier. This mechanism requires no sliding or twisting to explicitly break inversion symmetry and produces robust ferroelectric-like hysteresis loops that emerge prominently near 30~K. Systematic measurements under strong in-plane and out-of-plane magnetic fields reveal negligible effects on the hysteresis characteristics, confirming that the primary mechanism driving the dipole switching is electrostatic. Our findings establish a distinct and robust route to electrically tunable ferroelectric phenomena in van der Waals heterostructures, opening opportunities to explore the interplay between interfacial charge transfer and temperature-tuned barrier crossing of dipole states at the atomic scale.

Figures (4)

• Figure 1: Device structure and transport characteristics before and after top-gate fabrication.(a) Optical image and schematic of a device (D1) without a top gate, used for measurements in panel (c). (b) Optical image and schematic of the same device but with a top gate, used for measurements in panels (d) and (e). Scale bars at panel (a) and (b) for device images are 10 $\mu$m.(c) Four-terminal longitudinal resistance, $R$, measured before top gate fabrication, shown as a function of back gate voltage $V_\mathrm{BG}$ at $B = 0$ T, for temperatures ranging from 1.5 K to 50 K. Solid lines indicate forward sweeps, dashed lines backward sweeps. (d) After top gate fabrication, four-terminal resistance measured as a function of top gate voltage $V_\mathrm{TG}$ at $V_\mathrm{BG} = 0$ V. (e) Resistance as a function of back gate voltage $V_\mathrm{BG}$ measured at a fixed top gate voltage $V_\mathrm{TG} = 0$ V in the same device after top gate fabrication. Grey lines show additional measurements performed after five-month storage under floating top gate conditions. The color coding for temperature is consistent across panels (c)--(e).
• Figure 2: Spacer-thickness dependence of gate-induced hysteresis in longitudinal resistance. Dependence of longitudinal resistance on top gate voltage (upper panels) and back gate voltage (lower panels) for (a) device D0 (without thin hBN spacer), (b) device D1 (bilayer hBN), (c) device D2 (bilayer hBN), (d) device D3 (trilayer hBN), (e) device D4 (seven-layer hBN), and (f) device D5 (ten-layer hBN). Device numbers and corresponding hBN thickness are indicated in each upper panel. Red, yellow, and blue curves correspond to measurements at 50 K, 30 K, and 1.5 K, respectively.Device D0, without hBN, shows no measurable hysteresis at any temperature, confirming the critical role of spacer thickness.
• Figure 3: Insensitivity of dipole switching to magnetic field orientation.(a) Four-terminal longitudinal resistance as a function of top gate voltage $V_\mathrm{TG}$ at an in-plane magnetic field $B_{\parallel} = 9$ T and $V_\mathrm{BG} = 0$ V, measured at four temperatures: 50 K, 30 K, 20 K, and 1.5 K for D2. Solid lines represent forward sweeps, and dashed lines represent backward sweeps. (b) Identical measurement as in panel (a), now performed under a perpendicular magnetic field $B_{\perp} = 9$ T.
• Figure 4: emperature-dependent formation and locking of interfacial dipoles. Schematic representation of temperature-dependent dipole formation at the graphene/hBN/$\alpha$-RuCl$_3$ heterointerface. The grey, blue, and violet layers represent graphene, hBN, and $\alpha$-RuCl$_3$, respectively. The horizontal bar indicates three characteristic temperature regimes. (a) At high temperature ($T \sim 50$ K), thermal fluctuations suppress robust dipole formation, resulting in low dipole density. (b) At intermediate temperature ($T \sim 30$ K), dipoles form and align under an applied gate field due to thermally assisted activation. (c) At low temperature ($T \sim 10$ K), dipoles are locked and cannot be reversed by the available gate voltage.