Terahertz time-domain signatures of the inverse Edelstein effect in topological-insulator|ferromagnet heterostructures

This study utilizes terahertz time-domain spectroscopy to distinguish the inverse Edelstein effect from the inverse spin Hall effect in ferromagnet-topological insulator heterostructures by identifying a unique 270-fs relaxation component in the transient charge current that serves as a signature of interfacial spin-charge conversion.

The Gist

Imagine you have a high-speed highway where cars (electrons) are driving. In the world of quantum physics, some of these cars have a special "handedness" or "spin" (like a left-handed or right-handed screw). Usually, if you want to turn this "spin" into an electrical current (to power a device), it's a messy job because the cars are mixed up in a giant traffic jam.

This paper is about a special kind of highway called a Topological Insulator (specifically a material called Bi₂Te₃). Think of this material as a magical road where the "spin" of the cars is locked to their direction. If a car spins left, it must drive forward; if it spins right, it must drive backward. This makes it incredibly efficient at turning spin into electricity.

However, scientists have a problem: They know this magic road exists, but they can't easily tell if the electricity they are measuring is coming from the surface of the road (where the magic happens) or from the bulk (the dirt and rocks underneath the road, which is just a normal, messy highway).

The Experiment: The "Terahertz Camera"

The researchers built a sandwich:

1. The Top Layer: A magnetic metal (like Iron or Cobalt) that acts as a "spin factory."
2. The Bottom Layer: The Topological Insulator (the magic road).

They hit this sandwich with a super-fast laser pulse (lasting only a femtosecond, which is a quadrillionth of a second). This laser acts like a giant push, shoving a wave of "spinning" electrons from the metal into the magic road.

To see what happens, they used a Terahertz (THz) camera. When the spinning electrons turn into an electric current, they emit a tiny burst of radio waves (THz radiation). By measuring this burst, they can "see" the current in real-time.

The Discovery: Two Different "Footprints"

The big breakthrough in this paper is that they didn't just look at how strong the signal was; they looked at how fast it happened. They found two distinct "footprints" in the data:

1. The "Instant Flash" (The Bulk Effect)

• The Analogy: Imagine throwing a bucket of water onto a dry sponge. The water hits the surface and immediately soaks in. It's fast, messy, and happens everywhere at once.
• The Science: This is the Inverse Spin Hall Effect (ISHE). It happens deep inside the material (the bulk). It's an instant reaction where the spin turns into electricity immediately. It's fast, but it's not very efficient.

2. The "Slow Leak" (The Surface Effect)

• The Analogy: Now imagine the same bucket of water hitting a special sponge that has a tiny, magical reservoir on top. The water doesn't just soak in; it pools up in the reservoir for a moment (about 270 femtoseconds) before slowly leaking out as a steady stream.
• The Science: This is the Inverse Edelstein Effect (IEE). This happens only on the surface of the Topological Insulator. The spins pile up (accumulate) at the interface, creating a "spin pressure," and then convert into electricity. This process takes a tiny bit of time (the "leak").

Why This Matters

The researchers realized that by looking at the "Slow Leak" (the 270-femtosecond delay), they could finally separate the magic surface current from the messy bulk current.

• The Result: They proved that the "magic" surface effect is real and distinct.
• The Catch: They found that only a tiny fraction (less than 1%) of the spinning electrons actually make it into the magic surface states. Most of them just fall into the "dirt" (the bulk).
• The Good News: Even though the connection is weak, the "Slow Leak" signature is consistent. It doesn't matter if you use Iron or Cobalt as the top layer; the magic surface always behaves the same way.

The Takeaway

Think of this like listening to a band play.

• The Bulk is the drummer playing a fast, chaotic beat (the instant flash).
• The Surface is the violinist playing a slow, beautiful melody that lingers (the slow leak).

Before this paper, scientists were hearing a jumbled mess and couldn't tell the violin from the drums. By using a super-fast "camera" (THz spectroscopy) to watch the timing of the notes, they finally isolated the violin's melody.

Why should you care?
This is a blueprint for building faster, more efficient computers. If we can figure out how to get more electrons to use that "magic surface" (the violin) instead of the "messy bulk" (the drums), we could create spintronic devices that process information at the speed of light with almost no heat loss. This paper gives us the first clear map of how to do that.

Technical Summary

Here is a detailed technical summary of the paper "Terahertz time-domain signatures of the inverse Edelstein effect in topological-insulator|ferromagnet heterostructures."

1. Problem Statement

Three-dimensional topological insulators (TIs) possess topologically protected surface states (TSS) with spin-momentum locking, enabling highly efficient spin-charge-current interconversion (SCI) via the Inverse Edelstein Effect (IEE). However, a major experimental challenge remains: distinguishing the surface-related IEE from the Inverse Spin Hall Effect (ISHE), which occurs in the bulk of the material.

• The Conflict: Both ISHE and IEE produce transverse charge currents with identical macroscopic symmetry properties, making them difficult to separate in standard transport experiments.
• The Gap: While ultrafast THz emission spectroscopy has been used to study these systems, previous works focused primarily on signal amplitudes rather than temporal dynamics. Theoretical predictions suggest that ISHE is quasi-instantaneous (sub-50 fs), while IEE involves spin accumulation and should exhibit slower dynamics. This paper aims to exploit these temporal differences to isolate the IEE signature.

2. Methodology

The authors employed broadband Terahertz (THz) emission spectroscopy to probe ultrafast spin transport in heterostructures.

• Sample Architecture: They fabricated model stacks consisting of a ferromagnetic metal layer ($\mathcal{F}$) and a topological insulator layer ($\mathcal{N}$).
  • $\mathcal{F}$ (Ferromagnet): Co (1.5 nm) and Fe (2 nm).
  • $\mathcal{N}$ (Non-magnetic): TIs including $\text{Bi}_2\text{Te}_3$, $\text{SnBi}_2\text{Te}_4$, and $\text{Bi}_{1-x}\text{Sb}_x$.
  • Reference: $\mathcal{F}|\text{Pt}$ stacks were used as a baseline, where SCI is dominated by the bulk ISHE.
• Excitation: A femtosecond laser pulse (800 nm, ~10 fs duration) optically heats the ferromagnetic layer, inducing a transient spin voltage ($\mu_s^\mathcal{F}$). This drives an ultrafast longitudinal spin current ($j_s$) from $\mathcal{F}$ into $\mathcal{N}$.
• Detection: The resulting transverse charge current ($j_c$) emits a THz electromagnetic pulse. The electric field $E(t)$ is measured via electro-optic sampling.
• Data Analysis Strategy:
  1. Extract the charge-current sheet density $I_c(t)$ from the THz signal.
  2. Define a response function $H(t)$ by deconvolving the measured current $I_c(t)$ with the driving spin voltage dynamics (derived from the $\mathcal{F}|\text{Pt}$ reference).
  3. Analyze the temporal shape of $H(t)$ to identify distinct components (instantaneous vs. delayed).

3. Key Contributions

• Temporal Separation of Mechanisms: The study successfully separates the ISHE and IEE contributions not by material modification, but by analyzing the femtosecond time-domain dynamics of the photocurrent.
• Identification of a Distinct IEE Signature: The authors identify a specific, non-instantaneous relaxation component in the response function that is unique to TI interfaces and independent of the ferromagnetic material used.
• Quantitative Modeling: They propose a microscopic model where the total response is a superposition of an instantaneous bulk ISHE term and a delayed interfacial IEE term, allowing for the estimation of the spin injection fraction into surface states.

4. Key Results

• Dual-Component Response: The extracted response function $H(t)$ for $\mathcal{F}|\text{Bi}_2\text{Te}_3$ stacks reveals two distinct components:
  1. Component (i): A quasi-instantaneous response (limited by the system's 74 fs resolution). This is attributed to the ISHE in the bulk of the TI and the ferromagnet.
  2. Component (ii): A longer-lived exponential decay with a relaxation time of $\tau \approx 270 \pm 20$ fs.
• Material Independence: The 270 fs relaxation time is identical for both Co and Fe ferromagnetic layers. This confirms that the delayed component originates from the TI interface physics rather than the ferromagnet's properties.
• Mechanism Assignment:
  • The instantaneous component is assigned to bulk ISHE.
  • The delayed (270 fs) component is assigned to the Inverse Edelstein Effect (IEE) at the $\mathcal{F}/\text{Bi}_2\text{Te}_3$ interface. The delay corresponds to the time required for spin accumulation ($\mu_s^{IF}$) to build up and relax at the interface.
• Injection Efficiency: By comparing the amplitudes of the instantaneous and delayed components, the authors estimate that the fraction ($f$) of incident spins injected into the topological surface states is $f < 10^{-2}$ (less than 1%). This suggests a relatively weak coupling between the ferromagnet and the TSS, consistent with the long lifetime of the surface states.
• Exclusion of Alternatives: The authors rigorously ruled out alternative explanations such as:
  • Spin trapping in an intermediate layer (unlikely because the time constant is independent of the ferromagnet type).
  • Thermal (Seebeck) effects (ruled out by the linear dependence of the signal on pump fluence).
  • Bulk spin accumulation (ruled out by the short bulk scattering time of $\text{Bi}_2\text{Te}_3$).

5. Significance

• Fundamental Insight: This work provides the first clear experimental evidence distinguishing IEE from ISHE in TI heterostructures based purely on ultrafast temporal signatures. It confirms that the IEE is a distinct, slower process governed by interfacial spin accumulation.
• Methodological Blueprint: The approach of extracting a "spin-transport/SCI response function" serves as a powerful tool for disentangling complex spin dynamics in multilayer spintronic devices, applicable to other conversion processes like spin trapping.
• Device Implications: The findings highlight that while TIs offer high SCI efficiency, the coupling to ferromagnets is a critical bottleneck (low injection fraction). Understanding these dynamics is essential for optimizing THz emitters and spintronic logic devices.
• Technological Application: The ability to tailor the response of THz emitters by tuning interface properties (e.g., enhancing the IEE contribution) opens new avenues for designing broadband THz sources and detectors.

In summary, the paper demonstrates that femtosecond time-domain spectroscopy is a decisive tool for resolving the microscopic origins of spin-charge conversion, revealing that the Inverse Edelstein Effect in topological insulators manifests as a distinct 270 fs relaxation tail superimposed on the instantaneous bulk ISHE response.