Localized Exciton Emission with Spontaneous Circular Polarization in NiPS3/WSe2 Heterostructures

This study demonstrates that few-layer NiPS3/WSe2 van der Waals heterostructures exhibit spontaneous circularly polarized exciton emission and nonlinear Zeeman splitting due to interface-induced localization and magnetic proximity effects, as confirmed by low-temperature spectroscopy and DFT calculations.

The Gist

Imagine you have two very different, ultra-thin sheets of material. One is a semiconductor (like a tiny, high-speed computer chip made of Tungsten Diselenide, or WSe₂), and the other is a magnet (an antiferromagnet made of Nickel Phosphorus Trisulfide, or NiPS₃).

In the world of 2D materials, scientists love stacking these sheets on top of each other to create "heterostructures." Think of it like making a sandwich where the bread and the filling are made of completely different ingredients, but they stick together perfectly because they are so thin.

This paper is about what happens when you stack these two specific sheets together and look at them with a super-powerful microscope in a freezing cold room. Here is the story of what they found, explained simply:

1. The "Ghost" Lights

Usually, if you shine a laser on the semiconductor sheet (WSe₂) alone, it glows a certain way. If you shine it on the magnet sheet (NiPS₃) alone, it glows differently.

But when the scientists stacked them, something magical happened. New, sharp, bright lights appeared that didn't exist in either sheet on its own.

• The Analogy: Imagine you have a red flashlight and a blue flashlight. If you tape them together, you might expect purple light. Instead, you suddenly see a bright, sharp green laser beam that neither flashlight could produce alone.
• What's happening: The interface (the sticky spot where the two sheets touch) creates tiny "traps" or "pits." Electrons get stuck in these pits, forming what scientists call "localized excitons." These trapped electrons are glowing with a very specific, sharp color.

2. The Spontaneous Spin (The Magic Compass)

Here is the most surprising part. Usually, to make light spin in a circle (called "circular polarization"), you need a giant magnet nearby to force the electrons to line up. It's like needing a strong wind to make a pinwheel spin.

But in this experiment, no external magnet was needed.

• The Analogy: Imagine a group of people (electrons) holding hands in a circle. Usually, they stand randomly. But because they are standing next to a "magnetic neighbor" (the NiPS₃ layer), the neighbor's invisible magnetic influence whispers to them, "Hey, everyone, spin clockwise!"
• The Result: Even without a magnet on the outside, the light coming out of the stack is already spinning in a specific direction. This is called spontaneous circular polarization. It suggests that the magnetic sheet is "leaking" its magnetic personality onto the semiconductor sheet, forcing the electrons to align.

3. The Non-Linear Dance

When the scientists did add an external magnet, they expected the light's energy to change in a straight, predictable line (like a car accelerating at a constant speed).

• The Reality: The energy change was weird and curved. It was like the car suddenly sped up, then slowed down, then sped up again as the magnet got stronger.
• Why? This "non-linear" behavior is the fingerprint of a strong magnetic proximity effect. It proves that the magnetic field from the NiPS₃ sheet is fighting and interacting with the external magnet in a complex dance, creating a powerful internal magnetic field right at the interface.

4. The Computer Simulation (The "Virtual Lab")

To understand why this was happening, the scientists used supercomputers to build a virtual model of the two sheets.

• The Discovery: The computer showed that the atoms on the bottom of the magnet sheet and the top of the semiconductor sheet are "holding hands" (hybridizing). They are mixing their electronic properties.
• The Spin Texture: The simulation showed that the "spin" (the tiny internal compass of the electron) gets twisted and rearranged right at the boundary. This confirms that the magnetic sheet is physically changing the rules of the game for the semiconductor electrons.

Why Does This Matter?

This isn't just a cool science trick; it's a blueprint for the future.

• Valleytronics: Think of "valleys" in the material as different lanes on a highway. Usually, cars (electrons) can go in any lane. This experiment shows we can force them to stay in a specific lane just by stacking materials, without needing big, bulky magnets.
• Chiral Light Sources: We can create tiny light bulbs that emit "spinning" light, which is crucial for new types of secure communication and quantum computers.
• Tunable Devices: By stacking these materials, we can build devices where we can turn magnetic properties on and off just by changing the layers, leading to faster, smaller, and more efficient electronics.

In a nutshell: By stacking a semiconductor and a magnetic crystal, the scientists created a new type of light source that spins on its own and reacts strangely to magnets. They proved that the "neighborhood" between these two materials creates a powerful, invisible magnetic force that can control how light and electricity behave, opening the door to a new generation of smart, magnetic electronics.

Technical Summary

1. Problem Statement

Two-dimensional (2D) van der Waals (vdW) heterostructures offer a platform to engineer electronic and optical properties via proximity effects. While transition metal dichalcogenides (TMDs) like WSe2 possess strong spin-valley coupling, their valley degeneracy typically requires external magnetic fields (Zeeman effect) to lift, resulting in small splitting (~0.2 meV/T). Conversely, magnetic 2D materials like the antiferromagnet NiPS3 exhibit rich magnetic ordering but lack the direct bandgap and strong spin-valley coupling of TMDs.

The central challenge addressed in this work is the lack of comprehensive understanding regarding the optical and magnetic properties of few-layer NiPS3/WSe2 heterostructures (HSs), particularly in the absence of strain. Previous studies focused on monolayer systems or strained interfaces, leaving a gap in knowledge regarding how uncompensated spins at the interface of an antiferromagnet and a semiconductor influence valley polarization and exciton dynamics in thicker, few-layer systems without artificial strain.

2. Methodology

The researchers employed a multi-faceted approach combining experimental characterization, magneto-optical spectroscopy, and theoretical modeling:

• Sample Fabrication: Two distinct devices (HS1 and HS2) were fabricated using a dry-transfer technique inside a glovebox.
  • HS1: Si/SiO2 / hBN / NiPS3 (multilayer) / WSe2 (35 layers).
  • HS2: Si/SiO2 / Au electrodes / NiPS3 (multilayer) / WSe2 (10 layers) / hBN capping.
• Structural Characterization:
  • AFM: Verified layer thicknesses (NiPS3: ~84–237 layers; WSe2: 10–35 layers).
  • Raman Spectroscopy: Confirmed the presence of both materials in the overlap region and verified the absence of significant strain (no peak shifts/broadening compared to bare flakes).
  • Polarized Second Harmonic Generation (P-SHG): Analyzed crystal symmetry and orientation.
• Optical Spectroscopy:
  • Micro-Photoluminescence (μ-PL): Performed at low temperatures (4–300 K) using unpolarized 480/488 nm excitation.
  • Magneto-PL: Measured emission under out-of-plane magnetic fields (up to several Tesla) with circular polarization detection ($\sigma^+$ and $\sigma^-$).
  • Polarization Analysis: Quantified Degree of Circular Polarization (DCP) and Degree of Linear Polarization (DLP).
• Theoretical Modeling:
  • Density Functional Theory (DFT): Calculations performed using the FHI-aims code to model the electronic structure of monolayer/bilayer WSe2 interfaced with monolayer NiPS3.
  • Key Parameters: Included spin-orbit coupling, many-body dispersion corrections, and external electric fields to mimic built-in junctions. Calculated band structures, momentum matrix elements (MMEs), and spin textures.

3. Key Contributions & Results

A. Emergence of Localized Intralayer Excitons

• Observation: Unlike bare few-layer WSe2 (which shows negligible PL) or bare NiPS3 (showing a single linearly polarized peak at 1.476 eV), the heterostructures exhibit multiple sharp, discrete emission peaks between 1.48 and 1.56 eV.
• Origin: These peaks are attributed to localized intralayer excitons within the WSe2 layer, confined by interface-induced potentials or local stacking registries. The absence of a moiré superlattice (due to incommensurate lattice constants) suggests confinement arises from local high-symmetry stacking configurations rather than a periodic moiré potential.
• Thermal Stability: These emission lines are robust up to ~50 K, above which they quench, consistent with localized exciton behavior.

B. Spontaneous Circular Polarization (Zero-Field)

• Discovery: The localized excitons exhibit strong spontaneous circular polarization (DCP ~50–70%) even at zero external magnetic field.
• Significance: This indicates a valley polarization induced by an effective interfacial exchange field. Since bulk NiPS3 is an antiferromagnet with compensated spins, this effect arises from uncompensated or canted spins at the NiPS3/WSe2 interface, likely due to symmetry breaking and orbital hybridization.
• Reproducibility: The effect was observed consistently across different devices (HS1–HS4) and measurement spots, with opposite helicities observed in different devices, suggesting sensitivity to the specific interfacial spin configuration.

C. Nonlinear Zeeman Splitting and Magnetic Proximity

• Magneto-PL Behavior: Under applied magnetic fields, the Zeeman splitting ($\Delta E$) of the exciton peaks deviates significantly from linearity at fields $|B| > 2$ T.
• Modeling: The data fits a model incorporating an interfacial exchange field ($B_{exc}$):
  $$\Delta E(B) = g_{eff}\mu_B [B + B_{exc} \tanh(B/B_0)]$$
• Parameters: Extracted exchange fields were substantial ($B_{exc} \approx 25–27$ T), far exceeding the external field, confirming a robust magnetic proximity effect. This behavior mirrors findings in ferromagnetic TMD heterostructures but is achieved here with an antiferromagnet.

D. Theoretical Validation (DFT)

• Intralayer Nature: Calculations of Momentum Matrix Elements (MMEs) showed that intralayer WSe2 transitions are orders of magnitude stronger than interlayer transitions, confirming the experimental assignment of intralayer excitons.
• Hybridization & Spin Texture: DFT revealed significant orbital hybridization between WSe2 and Ni 3d states at the interface. This hybridization modifies the spin texture, particularly near the $\Gamma$ point, providing a microscopic mechanism for the observed magnetic proximity effect.
• Built-in Field: The presence of a p-n junction (verified by I-V measurements) creates a built-in electric field. DFT showed this field shifts Ni 3d states toward the WSe2 conduction band, enhancing hybridization and increasing the Degree of Linear Polarization (DLP).

4. Significance and Impact

• Antiferromagnetic Proximity: This work demonstrates that antiferromagnetic 2D materials (NiPS3) can induce strong magnetic proximity effects in semiconductors (WSe2), challenging the notion that only ferromagnets are suitable for valley manipulation.
• Chiral Light Sources: The ability to generate spontaneously circularly polarized light without external magnetic fields or strain engineering opens new avenues for chiral quantum emitters and integrated photonic devices.
• Valleytronics: The system provides a tunable platform for controlling valley degrees of freedom via interface engineering, offering a pathway toward low-power, magnetically tunable optoelectronic devices.
• Interface Engineering: The study highlights that even in the absence of a moiré lattice, local interface potentials and uncompensated spins at antiferromagnetic interfaces can create robust, localized quantum states with unique magnetic properties.

In conclusion, the paper establishes NiPS3/WSe2 heterostructures as a versatile platform where interface-induced potentials and magnetic proximity effects combine to create localized, chiral excitonic emitters, bridging the gap between antiferromagnetism and valleytronics.