Unveiling Orbital-mediated Ultrafast Demagnetization in Rare Earth-Transition-Metal Ferrimagnets

This study establishes that the speed of ultrafast demagnetization in Rare Earth-Transition-Metal ferrimagnets is governed by a universal orbital-mediated mechanism where the competition between 3d and 4f spin-orbit coupling channels determines whether angular momentum dissipates via rapid direct orbital-to-lattice transfer or slower multi-step pathways.

The Gist

The Big Picture: The Speed Limit of Magnetic Memory

Imagine you have a hard drive or a smartphone that stores data using tiny magnets. To write new information, you have to flip these magnets. The faster you can flip them, the faster your device works.

Scientists have known for a long time that there is a "speed limit" to how fast these magnets can flip. This speed depends on how quickly the magnets can dump their "spin energy" (angular momentum) into the material's structure (the lattice) so they can reset.

For a long time, scientists were confused about Rare-Earth-Transition-Metal (RE-TM) magnets. These are special materials made by mixing a "Rare Earth" metal (like Gadolinium or Terbium) with a "Transition Metal" (like Iron or Cobalt). Some of these mixtures flip incredibly fast (in less than a trillionth of a second), while others are much slower. The paper asks: Why do some mixtures zoom and others crawl?

The New Discovery: It's All About the "Orbital Highway"

The authors of this paper propose a new rulebook to explain this speed difference. They say the secret lies in a specific type of internal friction called Spin-Orbit Coupling (SOC).

To understand this, imagine the electrons in the magnet are like cars on a highway.

• Spin is the car's engine power.
• Orbit is the road the car is driving on.
• The Lattice is the parking lot where the cars need to stop to reset.

The paper argues that the speed of the "flip" depends on which "road" (orbital) the energy takes to get to the "parking lot."

Scenario A: The "Cobalt" Express Lane (Fast)

When the material uses Cobalt (Co) as the Transition Metal, it has a "strong" connection between the engine and the road (Strong Spin-Orbit Coupling).

• What happens: When the laser hits the magnet, the energy flows directly from the engine, onto the road, and immediately dumps into the parking lot.
• The Result: The magnet flips in a single, super-fast step. It's like taking a direct highway with no traffic lights.

Scenario B: The "Iron" Detour (Slow)

When the material uses Iron (Fe), the connection between the engine and the road is "weak."

• What happens: The energy can't get straight to the parking lot. Instead, it gets stuck in a side lane. It has to travel through the "Rare Earth" part of the material first.
• The Detour: The energy goes from the Iron engine $\rightarrow$ into the Rare Earth orbit $\rightarrow$ and then tries to get to the parking lot.
• The Result: This takes much longer. The magnet flips in two steps: a quick initial drop, followed by a slow, lingering recovery. It's like taking a scenic route with many stops.

The Role of the "Rare Earth" Passenger

The paper also explains that the specific Rare Earth metal matters, acting like a passenger in the car who can either help or hinder the trip.

• The Helpful Passenger (e.g., Terbium, Dysprosium): These passengers have their own "orbital" skills. If the Iron engine is weak, these passengers can help shuttle the energy to the parking lot, making the slow process a bit faster.
• The Unhelpful Passenger (e.g., Gadolinium): This passenger has no "orbital" skills. If the Iron engine is weak, the energy gets stuck in the passenger seat and bounces back to the driver. This causes a delay, making the whole process even slower and "jumpy."

How They Proved It

The researchers didn't just guess; they tested this theory with a "stopwatch" made of lasers.

1. The Test: They zapped different mixtures (Iron vs. Cobalt, mixed with various Rare Earths) with ultra-fast laser pulses.
2. The Observation:
   • Cobalt mixtures always flipped in one fast step, no matter which Rare Earth was added.
   • Iron mixtures always took two steps, and the speed of the second step depended entirely on which Rare Earth was added.
3. The "Tuning" Experiment: They added a tiny bit of Nickel (which is even stronger than Cobalt) to the Cobalt mixtures. As they added more Nickel, the magnets flipped even faster, confirming that strengthening the "road connection" speeds up the whole process.

The Conclusion

The paper concludes that the speed of these magnets isn't random. It is controlled by a competition between two things:

1. How strong the Transition Metal's "road" is (Cobalt is strong, Iron is weak).
2. How the Rare Earth passenger helps or hinders the trip.

If the "road" is strong (Cobalt), the energy dumps out instantly. If the "road" is weak (Iron), the energy gets stuck in a detour through the Rare Earth, slowing everything down.

This discovery gives engineers a clear recipe: If you want the fastest possible magnetic memory, you need to choose materials with strong "road connections" (like Cobalt or Nickel) to ensure the energy takes the express lane, not the scenic detour.

Technical Summary

Technical Summary: Unveiling Orbital-mediated Ultrafast Demagnetization in Rare-Earth-Transition-Metal Ferrimagnets

Problem Statement
The ultimate speed limit of magnetic recording and spintronic devices is governed by the efficiency of angular momentum (AM) transfer during ultrafast demagnetization. While the mechanism in simple transition-metal (TM) ferromagnets (e.g., Ni, Fe, Co) is well-described by the Elliott-Yafet model involving spin-orbit coupling (SOC)-mediated spin-flip scattering, the microscopic pathway in Rare-Earth-Transition-Metal (RE-TM) ferrimagnets remains debated. These materials exhibit a striking diversity in demagnetization timescales, ranging from sub-picoseconds to tens of picoseconds, depending on the specific RE and TM constituents. Despite their importance for applications like all-optical switching and terahertz emission, the interplay between 3d-SOC, 4f-SOC, and intersublattice exchange coupling in governing AM dissipation has not been unified.

Methodology
The study employs a combined experimental and theoretical approach to resolve the microscopic mechanisms of AM transfer:

• Experimental: Systematic time-resolved magneto-optical Kerr effect (TR-MOKE) measurements were conducted on a diverse set of RE-TM alloy thin films (RE = Sm, Gd, Tb, Dy, Ho; TM = Fe, Co, CoNi) prepared via DC magnetron sputtering. Measurements were performed under saturating magnetic fields with femtosecond laser excitation (800 nm, ~100 fs pulses).
• Theoretical: An Extended Four-Temperature Model (E4TM) was developed to explicitly incorporate SOC and inter-sublattice 3d-4f exchange. This model tracks energy exchange among electrons ($T_e$), the lattice ($T_l$), TM 3d spins ($T_s^{TM}$), and RE 4f spins ($T_s^{RE}$). The model utilizes material-specific constants ($G_{soc}^{3d}$ and $G_{soc}^{4f}$) to represent spin-to-lattice energy transfer efficiencies.
• Validation: Theoretical simulations were validated against experimental data, including the use of the Bloch $T^{3/2}$ law to convert simulated spin temperatures into magnetization dynamics. Additionally, terahertz (THz) emission measurements were used to independently confirm enhanced energy dissipation pathways.

Key Contributions and Results
The paper establishes an orbital-mediated framework where the competition between 3d-SOC and 4f-SOC dictates the AM dissipation channel:

1. Distinct Dynamical Regimes:

   • RE-Co Systems (Strong 3d-SOC): Exhibits single-step, sub-picosecond demagnetization. The strong 3d-SOC in Co (~90 meV) enables efficient, direct transfer of angular momentum from spin to orbital degrees of freedom, followed by rapid dissipation into the lattice.
   • RE-Fe Systems (Weak 3d-SOC): Exhibits two-step dynamics. The weaker 3d-SOC in Fe (~70 meV) suppresses the direct orbital-to-lattice channel. Instead, energy and AM are redirected into RE 4f orbitals via 3d-5d6s-4f pathways, resulting in a slower, secondary relaxation process.

2. Role of Rare-Earth Orbital Character:

   • The second-stage rate ($R_2$) in RE-Fe alloys scales with the orbital quantum number ($Q_L$) of the RE element.
   • For RE elements with $Q_L \neq 0$ (e.g., Tb, Dy), the 4f-SOC facilitates both back-transfer to the 3d subsystem and direct dissipation into the lattice, accelerating the second stage.
   • For RE elements with $Q_L = 0$ (e.g., Gd), the lack of an orbital moment blocks efficient 4f-SOC-mediated dissipation. This forces energy to flow back to the 3d subsystem via weaker exchange pathways, leading to a delayed, discontinuous two-step response.

3. Tuning via Nickel Doping:

   • To test the hypothesis that 3d-SOC strength is the dominant control parameter, Ni (with higher SOC, ~110 meV) was doped into TbCo and GdCo alloys.
   • The results showed a systematic acceleration of the demagnetization rate ($R_1$) and a preservation of single-step dynamics as Ni concentration increased, confirming that enhancing 3d-SOC reshapes the dissipation landscape.

4. Model Validation:

   • The E4TM successfully reproduced the experimental distinction between RE-Co and RE-Fe systems. Simulations demonstrated that a high $G_{soc}^{3d}$ drives rapid equilibration, while low $G_{soc}^{3d}$ forces reliance on slower, 4f-mediated channels.
   • THz emission measurements confirmed that stronger 3d-SOC (in GdCo vs. GdFe) generates more efficient orbital currents, enhancing THz emission intensity via the inverse orbital Hall effect.

Significance
The paper claims to resolve the controversy regarding the microscopic pathways of ultrafast demagnetization in RE-TM ferrimagnets by identifying the SOC-driven competition between 3d and 4f orbital channels as the universal mechanism. The study posits that the relative strength of 3d-SOC versus 4f-SOC is the decisive control parameter for energy and AM transfer. By establishing this orbital-mediated framework, the authors provide a predictive guide for the rational design of switching speeds in next-generation spintronic devices, moving beyond the previous focus solely on intersublattice exchange coupling.