Shear-Mode Raman Imaging of Ferroelectric Switching in Multilayer 3$R$-MoS$_2$

This study employs shear-mode Raman imaging and second-harmonic generation to reveal that ferroelectric switching in multilayer 3$R$-MoS$_2$ is a non-uniform, domain-wall-mediated process governed by pinning sites and exfoliation-created boundaries, which facilitate partial stacking transformations and distinct chiral orientations.

The Gist

Imagine a stack of playing cards. In a normal deck, the cards are perfectly aligned. But in a special type of material called 3R-MoS2 (a thin, flaky crystal), these "cards" (atomic layers) can slide past each other, like shuffling a deck. When they slide, the material becomes ferroelectric, meaning it develops an electric charge that can be flipped back and forth. This is called "sliding ferroelectricity."

The researchers in this paper wanted to see exactly how this sliding happens and what gets in the way. To do this, they used a special "camera" called Shear-Mode Raman Imaging. Think of this camera not as taking a photo of light, but as listening to the specific "hum" or vibration frequency of the layers as they rub against each other. Different ways the layers are stacked produce different "notes." By mapping these notes, the team could watch the layers move in real-time.

Here is what they discovered, explained through simple analogies:

1. The "One Big Sheet" is Actually a Patchwork Quilt

You might think a single flake of this material is one smooth, uniform piece. The researchers found that it's actually more like a patchwork quilt. Even within a single flake, there are invisible "seams" or boundaries where the material was torn or stressed during the peeling process.

• The Finding: These seams act like walls. When they applied an electric field to make the layers slide, one section of the flake would switch its charge, while the section right next to it stayed put. They acted like independent neighborhoods rather than one big city.

2. The "Staircase" vs. The "Elevator"

When you want to flip the electric charge, the layers don't all slide at once like a giant elevator dropping down. Instead, they move like people climbing a staircase, one step at a time.

• The Finding: To flip the charge, the top layer slides first, then the middle, then the bottom. However, the researchers saw that sometimes the "stairs" get skipped. In some areas, the layers moved so fast that the "middle steps" (intermediate states) were invisible to their camera. It was like a magician pulling a rabbit out of a hat so quickly you couldn't see the rabbit in the hat for a split second.
• The Pinning Effect: In other areas, the layers got "stuck" on a step. Imagine trying to slide a heavy box across a floor; sometimes it gets caught on a bump. The researchers found that tiny defects in the material act like these bumps (called pinning sites). These bumps hold the layers in place, making the "middle steps" visible and stable for a while before the layers finally jump to the next position.

3. The "Traffic Patterns" of the Boundaries

When the layers slide, they create boundaries between the old stacking order and the new one. The researchers used a laser technique (Second-Harmonic Generation) to see the direction of these boundaries.

• The Finding: They expected the boundaries to only go in two main directions (like the straight lines on a grid). Instead, they found a third, very common direction that runs diagonally, almost like a chiral (twisted) path. It's as if the material has a favorite "diagonal highway" that it prefers to use when switching, a path that wasn't predicted by previous theories.

4. The "Dead Zones"

The researchers also noticed that if the material was covered by metal electrodes (the wires used to apply electricity), the sliding stopped completely.

• The Finding: The metal acted like a shield, blocking the electric force from reaching the layers underneath. This confirmed that the sliding is driven by the electric field, but only if the field can actually reach the "cards" in the stack.

Summary

In short, this paper is like a high-speed traffic report for a microscopic city. The researchers used a special vibration-sensing camera to watch how layers of a crystal slide to flip their electric charge. They learned that:

• The material is often broken into independent zones by invisible cracks.
• The layers usually slide one by one, but sometimes get stuck on tiny defects, and sometimes move so fast we can't see the middle steps.
• There is a popular "diagonal" direction these sliding boundaries prefer to travel, which is a new discovery.

This helps scientists understand the "traffic rules" of these materials, which is essential for building future electronic devices that rely on this sliding behavior.

Technical Summary

Technical Summary: Shear-Mode Raman Imaging of Ferroelectric Switching in Multilayer 3R-MoS2

Problem Statement
Interfacial ferroelectricity in van der Waals layered crystals, such as 3R-MoS2, arises from inversion-breaking stacking that enables interlayer charge transfer and out-of-plane polarization. This polarization can be reversed via relative sliding between adjacent layers, a mechanism known as sliding ferroelectricity. While theoretical models suggest that domain wall (DW) motion, rather than coherent sliding of entire layers, drives this switching, the microscopic mechanism remains under active investigation. A critical gap exists in systematically understanding the switching dynamics in multilayer samples, particularly how the number of layers, the presence of defects, and mechanical boundaries influence the accessible stacking configurations and the stability of intermediate states. Previous studies have utilized photoluminescence and reflection-contrast spectroscopy, but a direct, structural probe capable of visualizing field-driven domain evolution on the device scale in multilayer MoS2 has been lacking.

Methodology
The authors employ shear-mode Raman spectroscopy and spatial mapping as a direct structural probe for ferroelectric switching. This technique exploits the sensitivity of interlayer shear modes to the specific stacking order (e.g., ABC vs. ABA) and layer number. By measuring these modes in a crossed-polarization configuration, the authors selectively track the in-plane layer translation associated with sliding ferroelectricity.

The experimental setup involves dual-gate field-effect transistors fabricated with few-layer 3R-MoS2 flakes as the channel material. By sweeping top- and bottom-gate voltages simultaneously, a vertical electric field ($E_\perp$) is applied to induce polarization reversal. The switching process is monitored in real-time through Raman mapping of the shear-mode frequency, allowing for the visualization of domain distribution evolution under varying electric fields. Complementary measurements include second-harmonic generation (SHG) to determine crystallographic orientations and reflection-contrast spectroscopy to resolve excitonic features.

Key Contributions and Results

1. Direct Visualization of Independent Switching Pathways:
   The study reveals that nominally single flakes are often mechanically segmented into regions that switch independently. Within a single flake, these segmented regions follow distinct switching pathways. For example, in trilayer samples, the transition between fully polarized states (ABC $\leftrightarrow$ CBA) can proceed through intermediate non-polar or partially polarized states (ABA or BAB). The authors observe that the dwell time of these intermediate states varies widely; in some regions, they are effectively bypassed (appearing as direct switching), while in others, they are stabilized by strong pinning sites.

2. Layer-Selective Stacking Transformations:
   The diverse switching pathways imply that domain walls constrain the accessible end states. In multilayer samples (e.g., tetralayers), the data indicates layer-selective stacking transformations. For instance, switching may involve the sliding of only the top or bottom layer, or sequential sliding of multiple layers, rather than a coherent shift of the entire stack. This results in partially polarized end states where domain walls reside between specific layer pairs.

3. Identification of Pinning and Boundary Effects:
   The research identifies mechanical segmentation boundaries and pinning sites as key factors governing switching dynamics.

   • Segment Boundaries: Contiguous flakes split into regions that switch independently. These boundaries often align with zigzag, armchair, or chiral directions and are attributed to strain-induced deformations formed during mechanical exfoliation. These boundaries block domain wall propagation, likely because extended DWs cannot conform to local structural discontinuities.
   • Pinning: The variation in the apparent critical switching fields and the stability of intermediate states are attributed to pinning by defects. In regions with strong pinning, DW velocity is reduced, stabilizing intermediate states long enough to be captured by Raman measurements. Conversely, in regions with weak pinning, DWs may move superlubrically at velocities up to km/s, rendering intermediate states invisible to the measurement timescale.

4. Discovery of a Prevalent Chiral Domain Wall Orientation:
   Second-harmonic generation measurements reveal three characteristic domain wall orientations: along the zigzag axis ($\Sigma_{\pi/2}$), the armchair axis ($\Sigma_0$), and a prevalent chiral direction near the zigzag-armchair bisector. This chiral orientation, corresponding to indices such as (8,3), (7,3), or (5,2) in the honeycomb lattice convention, is observed at sample edges and internal segment boundaries. This finding extends beyond existing theoretical models that primarily predict zigzag and armchair orientations.

Significance
The paper establishes shear-mode Raman mapping as a powerful, non-invasive tool for directly visualizing domain evolution in interfacial ferroelectrics. The results provide direct evidence that polarization reversal in multilayer 3R-MoS2 is governed by pre-existing domain walls and follows diverse, layer-selective pathways rather than coherent layer sliding. By identifying mechanical segmentation and pinning as critical determinants of switching dynamics and critical fields, the work offers a direct structural view of the factors limiting device performance. Furthermore, the identification of a prevalent chiral domain wall orientation challenges and expands current theoretical models of sliding ferroelectricity. These insights are crucial for understanding the fundamental physics of domain-wall-mediated switching and for the future integration of ultra-thin ferroelectric devices.