Controlling Terahertz Spintronic Photocurrents in 2D-Semiconductor|Ferromagnet Heterostructures through a Functional Hybrid Interface

This study reveals that a photon-energy-dependent hybrid metallic layer formed at the MoS2/Co interface acts as a transducer to efficiently generate and control ultrafast terahertz spin and charge currents in 2D-semiconductor|ferromagnet heterostructures, regardless of whether the pump photon energy is above or below the MoS2 band gap.

The Gist

Imagine you have a high-speed data highway where information travels as tiny magnetic spins. Scientists want to build a super-fast "traffic cop" for this highway using a sandwich of two special materials: a magnetic metal (Cobalt) and a super-thin semiconductor (MoS₂, a type of 2D material).

The goal is to use a flash of light (a laser) to kick-start a burst of electrical current that can be used for ultra-fast computing. But there's a problem: nobody was sure exactly how the light was making this happen, especially when they changed the "color" (energy) of the light.

Here is the story of what this paper discovered, explained simply:

The Old Theory: The "High-Speed Jump"

Previously, scientists thought the process worked like a high-jump competition.

• The Setup: The magnetic metal (Cobalt) is on the ground, and the semiconductor (MoS₂) is on a high platform.
• The Idea: When you hit the metal with light, it creates "hot" electrons. These electrons were thought to need enough energy to jump over a wall (an energy barrier) to get from the metal into the semiconductor.
• The Expectation: If you use a weak light (low energy), the electrons can't jump, and nothing happens. If you use a strong light (high energy), they jump easily. Also, because these "hot" electrons cool down and lose energy very quickly (in a blink of an eye), scientists expected the speed of the current to change depending on how hard you hit it.

The Surprise: The "Identical Rhythm"

The researchers tested this by hitting their material sandwich with three different colors of light: low energy, medium energy, and high energy.

What they found was weird:

1. The Speed didn't change: No matter which color of light they used, the "heartbeat" of the electrical current was exactly the same. It didn't speed up or slow down. This proved the "high-jump" theory was wrong. The electrons weren't jumping over a wall; they were staying put.
2. The Volume did change: While the speed was the same, the loudness (amplitude) of the current changed drastically. High-energy light made a much louder signal than low-energy light.

The Real Solution: The "Magic Hybrid Layer"

So, if the electrons aren't jumping, what is happening?

The scientists discovered that at the exact spot where the metal and the semiconductor touch, they don't just sit next to each other—they merge.

Think of it like mixing red and blue paint.

• Red Paint: The magnetic metal (Cobalt).
• Blue Paint: The semiconductor (MoS₂).
• The Purple Paint: A new, thin "hybrid" layer that forms right at the interface.

This "Purple Paint" layer is special. It acts like a smart energy sponge.

• When you shine low-energy light (like a dim flashlight), the sponge barely absorbs it.
• When you shine high-energy light (like a bright spotlight), the sponge soaks up a massive amount of energy.

How the Current is Generated (The New Story)

Here is the step-by-step process of what is actually happening, using our analogy:

1. The Sponge Heats Up: The laser hits the "Purple Sponge" (the hybrid layer). Because of its unique properties, it absorbs the light energy and gets hot very fast.
2. Passing the Heat: This heat doesn't stay in the sponge; it instantly transfers to the nearby magnetic metal (Cobalt), making the metal's electrons jittery and excited.
3. The Magnetic Pulse: This excitement creates a burst of magnetic activity in the Cobalt.
4. The Spin-Current: This magnetic burst pushes out a stream of "spins" (tiny magnetic arrows) into the hybrid layer.
5. The Conversion: Because the hybrid layer is a mix of metal and semiconductor, it acts like a translator. It instantly converts those magnetic spins into an electrical current, which shoots out as a Terahertz (THz) signal.

Why This Matters

• It's not about jumping: The electrons don't need to jump over a barrier. The magic happens right at the surface where the materials touch.
• It's tunable: Because the "sponge" absorbs different amounts of energy depending on the light color, scientists can now control how strong the signal is just by changing the color of the laser.
• The Future: This discovery suggests that by engineering these "hybrid layers" (the purple paint), we can build ultra-fast, efficient sensors and computers that process data at the speed of light.

In a nutshell: The researchers found that the secret to generating ultra-fast electricity isn't forcing electrons to jump over a wall, but rather creating a special "hybrid zone" at the interface that acts like a smart energy converter, turning light into magnetic pulses and then into electricity with incredible efficiency.

Technical Summary

1. Problem Statement

The integration of two-dimensional (2D) semiconductors, specifically transition-metal dichalcogenides (TMDs) like MoS₂, with ferromagnetic metals (FMs) like Cobalt (Co) holds promise for ultrafast spintronic devices. A critical challenge in this field is the efficient injection of spins from the metal into the semiconductor.

• Previous Hypothesis: Prior studies (e.g., Cheng et al.) suggested that ultrafast spin-current injection occurs via highly nonthermal, spin-polarized photocarriers generated in the FM. These carriers were thought to overcome the metal-semiconductor energy barrier (Schottky barrier) only if their energy exceeded the barrier height.
• The Gap: This model predicted that the dynamics of the generated terahertz (THz) current should change significantly depending on the pump photon energy (specifically, whether it is above or below the semiconductor band gap or the Schottky barrier). However, previous experiments lacked the temporal resolution (~50 fs) to verify these dynamic changes, relying only on amplitude measurements.

2. Methodology

The authors employed ultrafast THz-emission spectroscopy with ~50 fs temporal resolution to investigate MoS₂|Co bilayers.

• Sample Structure: A monolayer of MoS₂ (~0.7 nm) deposited on a 5 nm Co film, capped with SiO₂, on an Al₂O₃ substrate.
• Reference Sample: A Co|Pt bilayer was used as a fully metallic reference. In Co|Pt, THz emission is driven by the demagnetization dynamics of Co, which are known to be independent of pump photon energy (within the studied range).
• Experimental Conditions:
  • Three distinct pump photon energies were used: 0.8 eV (hot electrons only), 1.5 eV (hot electrons and holes), and 3.0 eV (direct excitation across the MoS₂ band gap).
  • An external magnetic field was applied to reverse the in-plane magnetization ($M$) of Co.
  • The THz signal was isolated by measuring the component odd under magnetization reversal ($S(t) = [S(t, +M) - S(t, -M)]/2$), filtering out charge-transfer artifacts.
• Theoretical Support: Ab-initio calculations (Density Functional Theory) were performed to model the electronic structure of the MoS₂/Co interface and calculate optical absorptance.

3. Key Results

The study yielded two primary, somewhat contradictory observations regarding the THz emission:

1. Identical Temporal Dynamics:

   • The time-domain THz waveforms for MoS₂|Co were indistinguishable from the Co|Pt reference across all three pump energies (0.8, 1.5, and 3.0 eV).
   • This implies that the spin-current generation mechanism is governed entirely by the Co layer's excess magnetization dynamics (which relaxes on a ~100 fs scale) and is not influenced by the presence of the TMD or the specific energy of the pump photons.
   • Implication: This refutes the "nonthermal photocarrier injection" model, which would have predicted faster dynamics (decay within tens of femtoseconds) for high-energy pumps.

2. Strong Amplitude Dependence on Photon Energy:

   • While the shape of the THz pulse remained constant, the amplitude varied drastically with pump energy.
   • The signal amplitude increased by a factor of ~2 when moving from 0.8 eV to 1.5 eV, and by a factor of ~4 when moving to 3.0 eV.
   • Crucially, this amplitude scaling did not correlate with the total absorptance of the entire stack, suggesting a localized mechanism.

4. Proposed Mechanism: The Hybrid Interfacial Layer

To explain the amplitude dependence without changing the dynamics, the authors propose the formation of a functional hybrid metallic layer at the MoS₂/Co interface.

• Hybridization: Ab-initio calculations confirm that electronic wavefunctions overlap at the interface, creating a ~1.4 nm thick hybrid layer (comprising the MoS₂ monolayer and the top ~0.7 nm of Co). This layer exhibits a mix of Co (metallic, magnetic) and MoS₂ (semiconducting) properties.
• Energy Transducer Role:
  • The hybrid layer possesses a pronounced photon-energy-dependent absorptance, particularly around the optical band gap of MoS₂.
  • As pump energy increases (0.8 $\to$ 3.0 eV), the hybrid layer absorbs significantly more energy, acting as a "transducer."
  • This localized heating increases the electron temperature in the adjacent Co layer, thereby enhancing the excess magnetization generated in the Co.
• Spin-Current Generation: The enhanced magnetization in the Co drives a larger spin current into the hybrid layer. Spin-to-charge conversion (likely via the Inverse Rashba-Edelstein Effect due to the d-state character of the hybrid states) then generates the THz pulse.
• Why Dynamics Remain Unchanged: Since the driving force is the Co magnetization (which relaxes on a fixed timescale regardless of how much energy is deposited), the temporal profile of the THz pulse remains identical to the Co|Pt reference, even though the amplitude scales with the hybrid layer's absorption efficiency.

5. Significance and Contributions

• Paradigm Shift: The paper challenges the prevailing view that ultrafast spin injection in FM|TMD stacks relies on nonthermal carrier injection over energy barriers. Instead, it demonstrates that the TMD does not directly participate in the spin transport dynamics but modifies the efficiency via interfacial hybridization.
• Validation of Hybridization: It provides experimental evidence that interfacial hybridization creates a metallic, magnetic layer that fundamentally alters optoelectronic properties, a phenomenon previously only theorized.
• Design Principle for Devices: The findings suggest that THz spintronic sources can be tuned not just by changing the bulk materials, but by engineering the interfacial band structure. By controlling the hybridization, one can maximize the absorption of specific photon energies to boost spin-current generation without altering the ultrafast response time.
• Universal Applicability: The authors argue that this hybridization-mediated mechanism is likely ubiquitous in metal|semiconductor interfaces, offering a new blueprint for tailoring ultrafast spin-current generation in next-generation data processing and sensing technologies.