🔬 mesoscale physics

Layer-engineered quantum anomalous Hall effect in twisted rhombohedral graphene

This paper demonstrates a layer-engineered platform based on twisted rhombohedral graphene where the Chern number of the quantum anomalous Hall effect can be programmatically set by the layer configuration and dynamically tuned or switched via electrical fields, enabling the on-demand design of reconfigurable topological states.

Original authors: Zhangyuan Chen, Naitian Liu, Jiannan Hua, Hanxiao Xiang, Wenqiang Zhou, Jing Ding, Xinjie Fang, Linfeng Wu, Le Zhang, Qianmei Chen, Xuanyu Chen, Kenji Watanabe, Takashi Taniguchi, Na Xin, Wei Zhu, Shu

Published 2026-07-21

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Zhangyuan Chen, Naitian Liu, Jiannan Hua, Hanxiao Xiang, Wenqiang Zhou, Jing Ding, Xinjie Fang, Linfeng Wu, Le Zhang, Qianmei Chen, Xuanyu Chen, Kenji Watanabe, Takashi Taniguchi, Na Xin, Wei Zhu, Shuigang Xu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the world of electricity as a busy highway. Usually, cars (electrons) drive in lanes, bumping into each other and losing energy as heat, which is why your phone gets warm when you use it. But physicists have discovered a special kind of "super-highway" called the Quantum Anomalous Hall (QAH) effect. On this magical road, cars can zip along the very edge without ever crashing or losing a drop of energy, even without an external magnetic field to guide them. The number of these special, friction-free lanes is called the "Chern number." Think of it like the number of lanes on a highway: a Chern number of 1 means one super-fast lane, while a higher number means a multi-lane superhighway. For a long time, scientists could only build highways with just one lane. They wanted to design custom roads with more lanes to carry more data faster and cooler, but figuring out how to build them was like trying to stack blocks in a way that defied gravity.

Now, a team of researchers has built a new kind of "Lego set" using twisted sheets of graphene (a material made of carbon atoms arranged in a honeycomb pattern) to solve this puzzle. They discovered that by stacking different numbers of these graphene sheets on top of each other and twisting them slightly, they can program the number of friction-free lanes directly into the material. Even cooler, they found a way to change the direction of traffic or even add more lanes just by turning a dial (using electricity) on the device, without having to rebuild it. This turns the study of these quantum highways from a game of "discovery" into a game of "design," opening the door to computers that are incredibly fast and use almost no power.

The Layered Cake of Quantum Magic

In this study, the researchers, led by Zhangyuan Chen and colleagues, created a series of devices they call "twisted rhombohedral graphene." To understand what they did, imagine taking a stack of paper sheets. If you stack them perfectly flat, they act like a single thick block. But if you twist the top sheet just a tiny bit relative to the bottom ones, you create a new pattern called a "moiré superlattice." It's like holding two window screens over each other and twisting them; a new, larger pattern of dots appears.

The team built these stacks using a specific recipe: they took a single layer of graphene (or sometimes two layers) and placed it on top of a rhombohedral stack of graphene that had 3, 4, or 5 layers. They named these combinations based on their layers, like a "1+3" sandwich (one layer on top of three) or a "1+4" sandwich. By twisting these layers by a very small angle (around 1.2 to 1.4 degrees), they created a playground for electrons.

The Main Discovery: Programming the Lanes

The big breakthrough here is that the researchers found a direct link between the number of layers in the stack and the number of friction-free lanes (the Chern number, $C$ ) the electrons can use.

The 1+3 Device: When they used a stack with 1 layer on top of 3 rhombohedral layers, they measured a Chern number of 3. This means the electrons flowed in 3 distinct, dissipationless lanes.
The 1+4 Device: When they added a fourth layer to the bottom stack (making it 1+4), the number of lanes jumped to 4.
The 1+5 Device: With a 1+5 stack, they observed 5 lanes.

The paper explicitly states that the layer number $N$ directly sets the Chern number $C$ (where $C=N$ ). This wasn't just a guess; they measured the electrical resistance and saw it drop to zero while the Hall resistance (the sideways push on the electrons) locked perfectly onto specific values: $h/3e^2$ , $h/4e^2$ , and $h/5e^2$ . These numbers are the "fingerprints" proving that 3, 4, or 5 lanes were open. This confirms that the topology of the material is "layer-engineered," meaning you can design the number of lanes just by choosing how many layers to stack.

The Magic Switch: Changing Direction and Adding Lanes

But the story gets even more playful. The researchers didn't just build static highways; they built highways where you can change the rules while driving.

1. Flipping the Traffic (Switchable Chirality)
In their "1+3" device, they found a way to flip the direction of the traffic flow. Usually, the electrons flow in a specific direction (say, clockwise). By adjusting the "displacement field" (an electric field created by top and bottom gates) or by changing the number of electrons (doping), they could switch the flow to counter-clockwise.

They observed the Hall resistance jump from a positive value ( $+h/3e^2$ ) to a negative value ( $-h/3e^2$ ).
This switch is sharp and hysteretic, meaning the system "remembers" which direction it was last pushed. It's like a light switch that clicks firmly into place. The authors suggest this happens because the electric field changes the energy stability of the electron "valleys," effectively flipping the magnetic orientation of the material.

2. The Shape-Shifter (Topological Phase Transition)
The most striking result came from a different stack: a "2+4" device (2 layers of Bernal graphene on top of 4 layers of rhombohedral graphene). Here, they didn't just flip the direction; they changed the number of lanes.

By tuning the displacement field, they drove the device from a state with 3 lanes ( $C=3$ ) to a state with 4 lanes ( $C=4$ ).
This is a "topological phase transition." Imagine a highway that suddenly gains an extra lane just because you turned a knob. The paper notes that while theory predicted this might be possible, actually seeing a device switch between two different non-zero Chern numbers (3 and 4) in a single experiment had been elusive until now.

What They Didn't Find (and What They Are Sure Of)

It is important to note what this paper does not claim. While they saw hints of other strange states at non-integer fillings (like at $\nu = 3/2$ ), they explicitly state that the quantization at these points was not perfect at zero magnetic field, likely due to disorder or contact issues. They do not claim to have solved the problem of fractional quantum Hall states yet, but they suggest their platform is a promising place to look for them in the future.

Furthermore, they ruled out the idea that these high Chern numbers are just a fluke of the specific material or a result of the magnetic field they applied. The effect persists down to zero magnetic field, and the layer dependence ( $C=N$ ) holds up across multiple devices (D1, D2, D3, and D4). The authors are very confident in the layer-dependent evolution of the topology, having confirmed it with both transport measurements and theoretical calculations (Hartree-Fock).

Why This Matters

This work shifts the paradigm from "discovering" topological materials to "designing" them. Before, scientists found a material and hoped it had the right properties. Now, they can say, "I want a highway with 5 lanes," and build a "1+5" graphene stack to get it. Even better, they can tune the device on the fly to change the number of lanes or the direction of flow.

The authors suggest that this could lead to "topological transistors" where the number of dissipationless channels is controlled by a gate voltage. While they don't claim to have built a working computer chip yet, they have laid the foundation for a new kind of electronics that could be faster and use significantly less energy than today's technology. They have turned the abstract concept of a "Chern number" into a programmable knob, opening the door to a future where we can engineer quantum matter exactly how we need it.

Technical Summary: Layer-engineered quantum anomalous Hall effect in twisted rhombohedral graphene

Problem Statement
The Quantum Anomalous Hall (QAH) effect, characterized by quantized Hall resistance ( $R_{xy} = h/Ce^2$ ) and vanishing longitudinal resistance without an external magnetic field, is a landmark in topological physics. However, existing QAH platforms—such as magnetically doped topological insulators and moiré systems—typically exhibit a Chern number of $|C|=1$ . While theoretical proposals suggest that stacking decoupled $C=1$ layers could yield higher Chern numbers, experimental realization of intrinsically designed QAH states with high, tunable Chern numbers remains elusive. Furthermore, the systematic evolution of QAH topology with layer number and configuration in rhombohedral graphene systems is not fully understood, particularly given that previous rhombohedral graphene/h-BN moiré experiments predominantly reported $C=1$ regardless of layer count.

Methodology
The authors fabricated a series of van der Waals heterostructures using a standard cut-and-stack technique. The devices consist of twisted graphene homostructures where a monolayer (or bilayer) graphene is placed atop a rhombohedral multilayer graphene with a controlled small twist angle ( $\theta$ ).

Device Architecture: The stacks are encapsulated by top and bottom hexagonal boron nitride (h-BN) dielectrics and utilize dual-graphite gates to independently tune the carrier density ( $n$ ) and displacement field ( $D$ ). Crucially, the graphene is intentionally misaligned with the h-BN to avoid additional moiré potentials, isolating the effects of the twisted graphene homostructure.
Materials: The study focuses on two families:
1. Twisted monolayer-rhombohedral $N$ -layer graphene, denoted as $(1+N)$ L, with $N=3, 4, 5$ .
2. Twisted Bernal bilayer-rhombohedral tetralayer graphene, denoted as $(2+4)$ L.
Characterization: Transport measurements were conducted in cryostats (base temperatures of 0.3 K and ~50 mK). The researchers measured longitudinal resistivity ( $\rho_{xx}$ ) and Hall resistivity ( $\rho_{xy}$ ) under varying magnetic fields, gate voltages, and displacement fields. Chern numbers were determined via two methods: (i) quantized plateaus in $\rho_{xy}$ and (ii) the Středa formula applied to Landau fan diagrams derived from magnetotransport data. Raman mapping and optical contrast confirmed the rhombohedral stacking and twist angles.
Theory: Self-consistent Hartree-Fock calculations based on a continuum model were employed to predict the band structures and valley Chern numbers.

Key Contributions and Results
The paper reports the experimental realization of a layer-engineered platform where the Chern number $C$ is programmable and electrically tunable.

Layer-Dependent Chern Numbers ( $C=N$ ):
In the twisted $(1+N)$ L family ( $N=3, 4, 5$ ), the authors observed robust QAH states at moiré filling $\nu=1$ . The measured quantized Hall resistivity corresponded to $C=N$ :
- $(1+3)$ L device: $C=3$ ( $\rho_{xy} = h/3e^2$ ).
- $(1+4)$ L device: $C=4$ ( $\rho_{xy} = h/4e^2$ ).
- $(1+5)$ L device: $C=5$ ( $\rho_{xy} = h/5e^2$ ).
  This confirms that the layer number $N$ directly sets the Chern number, driven by the intrinsic, layer-dependent Berry curvature of rhombohedral graphene.
Switchable Chirality (Sign of $C$ ):
In the twisted $(1+3)$ L device, the authors demonstrated that the chirality (sign of $C$ ) of the QAH state near $\nu=3$ can be switched. By varying the carrier density (filling factor $\nu$ ) or the displacement field ( $D$ ), the system transitions between $C=3$ and $C=-3$ . This switching manifests as a hysteresis loop in $\rho_{xy}$ , indicating that the displacement field can alter the energetic stability of valley-polarized states (K vs. K' valleys), effectively reversing the total magnetization and the Chern number.
Displacement-Field-Induced Topological Phase Transition:
In the twisted $(2+4)$ L device, the authors achieved a topological phase transition between two distinct QAH states with different absolute Chern numbers ( $C=3$ and $C=4$ ) within a single device. By tuning the displacement field $D$ , the system transitions from a state quantized at $h/3e^2$ to one quantized at $h/4e^2$ . This represents a transition between two nontrivial topological phases driven solely by an electric field.
Unconventional Chern Insulators:
The study also identified signatures of unconventional Chern insulators at non-integer fillings (e.g., $\nu=3/2$ with $C=2$ ), suggesting interaction-induced doubling of the moiré unit cell and simultaneous breaking of time-reversal and translational symmetries.

Significance and Claims
The authors claim that this work establishes a "layer-engineered and electrically tunable platform" that shifts the paradigm of topological quantum matter from discovery to design.

Design Principle: The Chern number can be programmed by selecting specific layer configurations (e.g., choosing $N$ in $(1+N)$ L), rather than relying on material discovery.
Reconfigurability: The ability to electrically switch between different Chern numbers (both sign and magnitude) in a single device offers a mechanism for "topological electronics," potentially enabling topological transistors that modulate dissipationless edge channels via gate voltage.
Future Potential: The platform provides a promising route for exploring fractional Chern insulators in high- $C$ bands, which are candidates for non-Abelian excitations relevant to topological quantum computation. The authors note that while their current devices show robust quantization, residual disorder prevents perfect quantization at non-integer fillings, but the system remains a viable platform for approaching these exotic states.

The work validates the theoretical prediction that rhombohedral graphene homostructures can host high- $C$ topological bands, overcoming the $|C|=1$ limitation observed in previous rhombohedral graphene/h-BN systems.