Investigating spin and orbital effects via spin-torque ferromagnetic resonance

This study experimentally investigates spin and orbital torque phenomena in various normal metal/ferromagnet bilayers using spin-torque ferromagnetic resonance, successfully extracting torque components and providing compelling evidence for orbital torque driven by the orbital Hall effect, including the demonstration of an out-of-plane torque attributed to interfacial mechanisms.

The Gist

Imagine you are trying to push a heavy swing (the magnet) to make it move. In the world of tiny computer chips, we want to control these "swings" (magnets) using electricity to store and process information.

For a long time, scientists knew two main ways to push this swing:

1. The Spin Push: You shoot a stream of tiny spinning tops (electrons with "spin") at the swing to knock it over.
2. The Orbital Push: You use a stream of electrons that are spinning around their own axis (like a planet orbiting the sun), known as "orbital angular momentum."

This paper is like a detective story where the researchers built a special laboratory to figure out exactly how much of the "push" comes from the spinning tops versus the orbiting planets, and which materials are best at doing the job.

Here is the breakdown of their investigation using simple analogies:

1. The Setup: The "Spin-Torque Ferromagnetic Resonance" (ST-FMR) Machine

Imagine the researchers built a tiny, high-tech playground. They have a strip of metal with two layers:

• The Magnet Layer (The Swing): This is the part they want to move. They used two types of magnets: Permalloy (a soft, easy-to-move magnet) and Nickel (a tougher, more complex magnet).
• The Pusher Layer (The Normal Metal): This is a non-magnetic metal sitting right next to the magnet. They tested many different types of "pushers" like Platinum, Bismuth, Copper Oxide, and Silver.

They zap this strip with a high-speed radio wave (like a microwave signal). This creates a current that tries to shake the magnet. The researchers then listen to the "noise" (a voltage signal) the magnet makes as it wobbles. By analyzing this noise, they can tell exactly how hard the "pusher" is hitting the "swing."

2. The Big Discovery: The "Orbital" Secret

For years, scientists thought the "Spin Push" (Spin Hall Effect) was the main hero. But this paper suggests there's a hidden co-pilot: the Orbital Torque.

Think of it like this:

• Spin Hall Effect (The Old Way): Imagine a crowd of people running down a hallway. Because of their internal spinning (spin), they bump into the walls and push a door open.
• Orbital Hall Effect (The New Way): Imagine the same crowd, but instead of just spinning, they are also carrying heavy backpacks and swinging them around as they run. This swinging motion (orbital momentum) creates a much stronger push against the door.

The researchers found that in certain materials, this "backpack swinging" (Orbital Hall Effect) is actually doing a huge amount of the work, sometimes even more than the spinning.

3. The "Nickel" vs. "Permalloy" Showdown

The researchers tested their "pusher" metals against two different magnets:

• Permalloy (The Simple Magnet): It's like a lightweight plastic swing. It responds okay to the push, but it doesn't interact much with the "backpack swinging."
• Nickel (The Complex Magnet): It's like a heavy, iron swing with a special surface. When the "backpack swinging" electrons hit Nickel, the Nickel grabs onto that orbital momentum and converts it into a massive push.

The Result: When they used Nickel, the "push" was significantly stronger than with Permalloy. This proved that the "Orbital Push" is real and that Nickel is a super-efficient converter of this energy.

4. The "Out-of-Plane" Twist

One of the coolest findings was that the push wasn't just happening from the side (like pushing a swing forward). Sometimes, the push came from above or below (out-of-plane).

Imagine trying to push a swing, but instead of pushing it forward, you are tapping it from the top. This happens because of the messy, imperfect connection between the two metal layers (the interface). It's like a handshake that's slightly crooked, creating a weird angle of force. The researchers found this "crooked handshake" is actually a sign of these new orbital effects happening right at the surface where the metals touch.

5. Why Does This Matter?

Why should you care about tiny electrons swinging backpacks?

• Faster Computers: If we can use this "Orbital Push" to switch magnets on and off, we can build memory chips that are faster and use less energy.
• New Materials: The study showed that some materials (like Copper Oxide and Bismuth) are amazing at this "Orbital Push," while others (like Silver) are terrible at it. This gives engineers a "shopping list" of materials to build better devices.
• The Future of "Orbitronics": We are moving from an era of "Spintronics" (using electron spin) to "Orbitronics" (using electron orbits). This paper is like a map showing us that the "Orbit" highway is wide open and full of potential.

In a Nutshell

The researchers built a tiny microphone to listen to how magnets wobble when hit by electricity. They discovered that while we thought electrons were just spinning to push magnets, they are also "orbiting" (swinging around) and that this orbital motion is a powerful, previously underappreciated force. By using the right combination of metals (especially Nickel), we can harness this force to build the next generation of super-efficient, super-fast electronic devices.

Technical Summary

1. Problem Statement

The electrical control of magnetization in spintronic devices is primarily driven by Spin-Orbit Torques (SOTs), traditionally attributed to the Spin Hall Effect (SHE) in heavy metals. However, recent advances in "orbitronics" suggest that the Orbital Hall Effect (OHE) plays a critical, and sometimes dominant, role in generating torques, particularly in materials with light elements.

• The Challenge: Distinguishing between spin-mediated torques (SHE) and orbital-mediated torques (OHE) is experimentally difficult. Conventional DC switching methods suffer from low sensitivity, heating issues, and difficulty in separating damping-like (DL) and field-like (FL) components.
• The Gap: There is a need for a robust, quantitative technique to disentangle the intertwined contributions of SHE and OHE in Normal Metal/Ferromagnet (NM/FM) heterostructures, specifically understanding how the ferromagnet's intrinsic properties (like Spin-Orbit Coupling) influence the conversion of orbital currents into spin torques.

2. Methodology

The authors employed the Spin-Torque Ferromagnetic Resonance (ST-FMR) technique, which operates in the frequency domain to amplify torque responses and separate different torque components with high sensitivity.

• Sample Design: A series of bilayer systems were fabricated via DC sputtering with the structure SiO₂/FM/NM.
  • Ferromagnetic (FM) Layers: Permalloy (Py, Ni₈₁Fe₁₉) and Nickel (Ni). Py was chosen for its relatively weak Spin-Orbit Coupling (SOC), while Ni was chosen for its stronger intrinsic SOC.
  • Normal Metal (NM) Layers: A variety of materials were tested to probe different transport mechanisms, including Pt, Sb, Bi, Ag, Zr, Mo, Ta, Ir, and CuOₓ.
  • Thicknesses: FM layers were fixed at 5 nm (Py) and 10 nm (Ni); NM layers were generally 8 nm (with exceptions for Sb at 4 nm and CuOₓ at 3 nm).
• Experimental Setup:
  • An RF current ($I_{RF}$) was applied along the strip, generating an oscillating Oersted field and a transverse spin/orbital current via SHE/OHE.
  • The resulting magnetization precession modulates the resistance via Anisotropic Magnetoresistance (AMR), generating a rectified DC mixing voltage ($V_{mix}$).
• Data Analysis:
  • Signal Decomposition: $V_{mix}$ was decomposed into symmetric ($V_S$) and antisymmetric ($V_A$) Lorentzian components.
  • Torque Extraction: The authors utilized two methods:
    1. Amplitude Ratio Method: Calculating torque efficiency ($\xi_{DL}$) based on the ratio $V_S/V_A$ and varying FM thickness.
    2. Direct Angular Fit Method: Fitting the angular dependence of $V_S$ and $V_A$ against the azimuthal angle $\phi$. Crucially, they incorporated out-of-plane torque components ($\tau^\perp$) into the fitting model to account for unconventional angular dependencies observed in several samples.
  • Control Measures: RF power dependence tests confirmed operation in the linear regime, ruling out significant thermal artifacts (heating or spin-Seebeck effects).

3. Key Contributions

• Differentiation of Spin vs. Orbital Torques: By comparing Py (weak SOC) and Ni (strong SOC) bilayers with identical NM layers, the study isolates the role of the ferromagnet's SOC in converting orbital angular momentum into spin torque.
• Identification of Out-of-Plane Torques: The paper demonstrates the presence of unconventional out-of-plane torque components ($\tau^\perp$) in several systems, attributing them to interfacial symmetry breaking and spin-orbital polarized currents along the $z$-direction.
• Quantification of Orbital Hall Effect (OHE): The study provides experimental evidence that OHE is a dominant torque generation mechanism in specific NM/FM combinations, particularly when the FM has strong SOC.
• Refined Analysis Framework: The authors propose a more robust analysis framework that accounts for out-of-plane torques, which are often overlooked in standard ST-FMR analyses of polycrystalline bilayers.

4. Key Results

• Ferromagnet Dependence:
  • Ni-based bilayers consistently exhibited significantly higher damping-like torque efficiencies ($\xi_{DL}$) compared to Py-based bilayers for the same NM materials (e.g., Ni/Sb vs. Py/Sb, Ni/Pt vs. Py/Pt).
  • This enhancement in Ni is attributed to its stronger intrinsic SOC, which facilitates more efficient orbital-to-spin conversion within the ferromagnet layer.
• Material Specificity (NM Layer):
  • High Efficiency: Materials like Pt, Bi, Sb, and CuOₓ showed large positive $\xi_{DL}$ in Ni bilayers, indicating strong orbital torque contributions. Specifically, Ni/CuOₓ showed exceptionally high efficiency, linked to the orbital Rashba effect at the interface.
  • Low/Anomalous Efficiency: Materials like Ag showed reduced efficiency (weak SOC/OHE). Zr and Mo showed efficiencies lower than theoretical predictions, and Ni/Mo exhibited a negative torque sign despite positive orbital Hall conductivity predictions. This suggests complex interplay between bulk OHE, interfacial effects, and non-reciprocity in orbital transport.
• Angular Dependence & Out-of-Plane Torques:
  • Many samples (e.g., Ni/Bi, Ni/Sb) displayed angular dependencies that deviated from the standard $\sin(2\phi)\cos(\phi)$ form.
  • These deviations were successfully modeled by adding a $\sin(2\phi)$ term, corresponding to out-of-plane torques ($\tau^\perp$).
  • Control experiments with Au spacers in Ni/Bi systems suppressed these asymmetries, confirming that the out-of-plane torques originate from interfacial mechanisms (Rashba-like effects) rather than bulk properties.
• Torque Efficiency Values:
  • For SiO₂/Ni(10)/Pt(8), $\xi_{DL}$ was approximately 0.127 (derived from Table I data scaling), significantly higher than Py/Pt.
  • The study confirms that the total torque is a superposition of SHE and OHE, with the OHE contribution becoming dominant in Ni-based systems due to efficient $L \to S$ conversion.

5. Significance

• Paradigm Shift: The work moves beyond the traditional Spin Hall Effect paradigm, establishing Orbital Torque as a robust, material-dependent mechanism for magnetization control.
• Design Guidelines: It provides critical design rules for future spintronic and orbitronic devices:
  • To maximize orbital torque, one should pair NM layers with strong OHE (e.g., Bi, Sb, CuOₓ) with FM layers possessing strong intrinsic SOC (e.g., Ni) to maximize orbital-to-spin conversion.
  • Interface engineering is crucial, as interfacial symmetry breaking can generate significant out-of-plane torques.
• Methodological Advancement: The demonstration that ST-FMR can detect and quantify out-of-plane orbital torques offers a powerful tool for characterizing next-generation magnetic materials where orbital degrees of freedom are exploited.
• Future Applications: These findings pave the way for more efficient magnetization switching in memory (MRAM) and logic devices by leveraging orbital angular momentum, potentially overcoming limitations of pure spin-current devices.