🔬 mesoscale physics

Observation of an exciton crystal in a moiré excitonic insulator

This study reports the first observation of a thermodynamically stable exciton crystal in a tunable moiré excitonic insulator, evidenced by optical Umklapp scattering and transport resistance peaks at one exciton per three moiré sites, thereby establishing a versatile platform for exploring correlated bosonic and fermionic phases.

Original authors: Ruishi Qi, Qize Li, Haleem Kim, Jiahui Nie, Zuocheng Zhang, Ruichen Xia, Zhiyuan Cui, Jianghan Xiao, Takashi Taniguchi, Kenji Watanabe, Michael F. Crommie, Feng Wang

Published 2026-01-28

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Ruishi Qi, Qize Li, Haleem Kim, Jiahui Nie, Zuocheng Zhang, Ruichen Xia, Zhiyuan Cui, Jianghan Xiao, Takashi Taniguchi, Kenji Watanabe, Michael F. Crommie, Feng Wang

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 a world where tiny particles, usually zipping around like chaotic bees, suddenly decide to stop and stand in perfect, rigid rows, forming a crystal. This is what scientists call a "crystal," but usually, we think of crystals as made of atoms (like salt or diamonds).

In this new study, researchers achieved something much more elusive: they made a crystal out of excitons.

What is an Exciton?

Think of an exciton as a "cosmic couple." In a semiconductor, an electron (which has a negative charge) can get paired up with a "hole" (a missing electron that acts like a positive charge). Because opposites attract, they stick together and dance around each other. This pair is the exciton.

Usually, these couples are very shy and short-lived. They break apart quickly, making it nearly impossible to get them to organize into a crystal. It's like trying to build a house of cards while the wind is blowing and the cards keep flying away.

The Recipe for Success

To solve this, the researchers built a special "playground" using a sandwich of ultra-thin materials (like layers of graphene and other 2D crystals). Here is how they made the excitons behave:

The Trap (The Moiré Pattern): They stacked two layers of material at a slightly twisted angle. This created a giant, invisible grid pattern (called a "moiré superlattice") on the surface. Imagine a giant checkerboard painted on the floor. This grid acts like a series of tiny bowls or traps.
The Long-Lived Couple: They used a special setup where the electron and the hole are in different layers of the sandwich, separated by a tiny insulating barrier. This keeps them from crashing into each other and breaking apart. They become "dipolar excitons"—long-lived couples that repel each other slightly, like two magnets with the same pole facing up.
The Freeze: By cooling the system down to near absolute zero and adjusting the number of couples, they slowed the excitons down enough that their natural repulsion forced them to settle into the "bowls" of the grid.

The Big Discovery: The 1-in-3 Rule

The team found a magical moment when they filled the grid.

The Scenario: Imagine a grid of 30 empty parking spots (the moiré sites).
The Result: When they put exactly 10 exciton couples into those 30 spots (a "1/3 filling"), something amazing happened. The excitons didn't just park randomly. They organized themselves so that no two couples sat next to each other. They spaced themselves out perfectly, like soldiers standing in a formation.

This is the Exciton Crystal.

How Did They See It?

Since you can't see these tiny particles with a regular microscope, the researchers used two clever tricks to prove the crystal existed:

The Light Test (Optical Spectroscopy): They shined a light on the material. Usually, the light bounces off in a predictable way. But when the exciton crystal formed, the light bounced back with a new, distinct "echo" (called an Umklapp scattering peak). It's like how a guitar string sounds different when you press it at a specific fret; the crystal changed the "note" of the light.
The Traffic Test (Transport): They tried to push the excitons through the material. When the excitons were free-flowing, they moved easily. But right at that "1-in-3" moment, the traffic jammed completely. The excitons refused to move because hopping to the next spot would mean sitting too close to a neighbor, which they were programmed to avoid. This "traffic jam" proved they were stuck in a rigid crystal structure.

Why Is This Cool?

The researchers also discovered that this system is like a versatile Lego set.

If they added extra "lonely" electrons or holes (unpaired charges), they could create a mix of a crystal of charges and a crystal of excitons living together.
They found that these exciton crystals are surprisingly stable, surviving at temperatures up to 15 degrees Kelvin (which is very cold, but warm for quantum physics).

In short: The scientists built a microscopic playground where long-lived particle couples were forced to stand in perfect, rigid rows. They proved this happened by watching how light bounced off them and how they stopped moving like traffic in a jam. This is the first time a stable crystal of these "light-matter" couples has been seen in a state of thermal equilibrium.

Technical Summary: Observation of an Exciton Crystal in a Moiré Excitonic Insulator

Problem Statement
While the crystallization of electrons into Wigner lattices due to strong Coulomb interactions is a well-established phenomenon in two-dimensional systems, the realization of a bosonic analogue—a crystal of excitons—has remained elusive. The primary obstacles are the inherently short lifetimes of conventional photoexcited excitons and the weakness of their interactions compared to electrons. Previous attempts to observe excitonic ordering relied on transient, photoexcited states (pico- to nanosecond lifetimes) which are far from thermal equilibrium, making it difficult to definitively identify a thermodynamically stable exciton crystal. The goal is to engineer a system with long-lived excitons possessing strong inter-exciton interactions and sufficiently suppressed kinetic energy to allow spontaneous crystallization in the ground state.

Methodology
The authors constructed a gate-tunable electron-hole (e-h) bilayer heterostructure to create a stable excitonic insulator (EI) coupled to a moiré potential. The device architecture consists of:

Hole Layer: An angle-aligned WS₂/WSe₂ moiré superlattice (created by a ~4% lattice mismatch), providing a periodic potential with an ~8 nm period.
Electron Layer: A monolayer MoSe₂.
Barrier: An ultrathin (~2 nm) hexagonal boron nitride (hBN) spacer separating the layers to prevent interlayer tunneling while maintaining strong interlayer coupling.
Gating: Top and bottom graphite gates with hBN dielectrics allow independent control of the net charge density ( $V_G$ ) and the interlayer bias ( $V_B$ ), which tunes the electron and hole densities ( $n_e$ and $n_h$ ) and the exciton density ( $n_x$ ).

The study employed two primary characterization techniques:

Optical Spectroscopy: Reflection contrast measurements at cryogenic temperatures (down to 2 K) to monitor exciton resonances ( $X_0$ , trions) and identify spectral signatures of symmetry breaking.
Transport Measurements:
- Coulomb Drag: A closed-circuit experiment measuring the drag current in the hole layer induced by a drive current in the electron layer to confirm exciton dominance and measure exciton resistance.
- Optical-Electrical Hybrid Four-Terminal Measurement: A novel technique using electrical bias to drive exciton currents and a scanning laser probe to optically measure the local exciton potential drop, thereby extracting contact-free exciton resistance.

Key Contributions and Results

Realization of a Stable Exciton Crystal at 1/3 Filling:
The authors identified a thermodynamically stable exciton crystal phase at an exciton filling factor of $n_x/n_0 = 1/3$ (one exciton per three moiré sites).
- Spectroscopic Evidence: At 1/3 filling, a distinct satellite peak appeared on the high-energy side of the main MoSe₂ intralayer exciton ( $X_0$ ) peak. This peak, separated by ~8 meV, is attributed to Umklapp scattering resulting from the folding of the exciton dispersion due to the new periodicity of the exciton crystal ( $\sqrt{3} \times \sqrt{3}$ lattice). This peak is absent in electron-only doped samples or at other filling factors and persists up to ~15 K.
- Transport Evidence: Exciton resistance measurements revealed a pronounced peak at the same 1/3 filling. The resistance is significantly higher than contact resistance, indicating a strongly insulating phase where exciton hopping is suppressed to avoid nearest-neighbor occupation. The resistance peak follows a thermal activation behavior with an energy scale of ~17 K, consistent with the melting temperature of the crystal observed in spectroscopy.
Characterization of the Excitonic Insulator (EI) Phase:
The system exhibits a dipolar EI phase at net charge neutrality ( $n_e = n_h$ ) where electrons and holes spontaneously bind into long-lived interlayer excitons. This is confirmed by perfect Coulomb drag (drag ratio $\approx 1$ ), indicating that charge transport is mediated entirely by neutral excitons. The EI phase transitions to an electron-hole plasma (EHP) at higher biases, where the drag ratio drops to zero.
Discovery of Correlated Mixed Phases:
By doping the system away from charge neutrality, the authors observed new correlated insulating states where dipolar excitons coexist with background charge crystals:
- Exciton + Mott Insulator (MI): At integer hole filling ( $n_h/n_0 = 1$ ) with excess electrons, a phase forms where dipolar excitons exist on top of a hole Mott insulator.
- Exciton + Generalized Wigner Crystal (GWC): At fractional hole fillings ( $n_h/n_0 = 1/3, 2/3$ ), dipolar excitons coexist with hole GWCs.
  These mixed phases exhibit perfect drag but are more fragile (lower binding energy and lower melting temperatures) than the pure EI phase due to increased many-body interactions from the excess charges.

Significance
The paper claims to provide the first unambiguous observation of a thermodynamically stable exciton crystal in thermal equilibrium. By combining a moiré potential with a gate-tunable excitonic insulator, the authors successfully suppressed exciton kinetic energy enough for dipolar interactions to dominate, overcoming the limitations of transient photoexcited systems.

The work establishes moiré excitonic insulators as a versatile platform for realizing and controlling correlated crystalline phases of both bosons (excitons) and fermions (electrons/holes), as well as tunable mixtures of both. This offers a solid-state alternative to ultracold atom systems for exploring strongly correlated quantum matter, specifically enabling the study of extended Bose-Hubbard models and exotic phases like dipolar excitonic insulators coexisting with charge crystals.