Unidirectionality of spin waves in Synthetic Antiferromagnets

This paper demonstrates through experiments and modeling that symmetric CoFeB/Ru/CoFeB synthetic antiferromagnets in a scissors state exhibit strong frequency non-reciprocity in their acoustical spin waves, enabling unidirectional energy transfer for wavevectors parallel to the applied magnetic field.

The Gist

The Big Idea: One-Way Streets for Magnetic Waves

Imagine you are driving a car. Usually, if you turn around and drive the opposite way, you can still move forward. But what if you discovered a magical highway where, no matter which way you point your steering wheel, your car only moves forward? You can't drive backward at all.

That is exactly what this team of scientists discovered, but instead of cars, they are dealing with spin waves.

What are spin waves?
Think of a magnet not as a solid block, but as a crowd of tiny people (atoms) all holding hands and marching in step. If you push one person, a "wave" of movement ripples through the crowd. In physics, we call these ripples spin waves. They carry energy and information, just like light waves carry images or sound waves carry music.

The Special Playground: The "Scissors" State

The scientists built a special sandwich to study these waves. It consists of two layers of magnetic material (like two sheets of metal) separated by a very thin spacer (Ruthenium).

Normally, these two layers want to face opposite directions (like two magnets repelling each other). But when they apply a magnetic field, they get stuck in a unique position called the "Scissors State."

• The Analogy: Imagine two people holding a pair of scissors. If you push the handles together, the blades open. In this state, the two magnetic layers are like the blades of the scissors: they are slightly open, pointing toward the magnetic field but not perfectly aligned.

The Magic Trick: Unidirectional Flow

The researchers found something bizarre happening in this "scissors" state, specifically with a type of wave called the Acoustical Spin Wave.

1. The Normal Rule: Usually, if a wave moves to the right, its frequency is $X$. If it moves to the left, its frequency is $Y$. The direction matters.
2. The Discovery: In their special sandwich, when the wave tries to move "backward" (against the magnetic field), something weird happens. The wave doesn't just slow down; it refuses to go backward.
   • If you try to send a wave to the left, it magically turns around and goes to the right.
   • If you try to send it to the right, it goes to the right.
   • Result: The energy flows in one direction only, regardless of how you try to push it.

The Analogy: Imagine a river that flows downstream. Usually, if you swim upstream, you can fight the current. But in this "magnetic river," the current is so strong and the rules of physics are so twisted that if you try to swim upstream, you don't just get pushed back; you are instantly teleported to the other side of the river, still swimming downstream. It is a one-way street for energy.

How They Proved It

The team used two main tools to catch these waves in action:

1. The Laser Eye (Brillouin Light Scattering): They shot a laser at the magnetic sandwich. When the light bounced off, it changed color slightly depending on how the waves were moving. This let them see the "speed limits" of the waves. They saw that for the "backward" waves, the speed was effectively zero or reversed, proving the one-way nature.
2. The Radio Antennas (Propagating Spin Wave Spectroscopy): They built a tiny device with two antennas (like a transmitter and a receiver).
   • Experiment A: They sent a signal from Antenna A to Antenna B. Result: Strong signal received!
   • Experiment B: They sent a signal from Antenna B to Antenna A. Result: Silence. The wave couldn't go that way.
   • The Switch: The coolest part? They could flip a switch (by toggling the magnetic state) and suddenly, the one-way street flipped direction. Now A-to-B was silent, and B-to-A was loud.

Why Does This Matter?

This isn't just a cool physics trick; it's a blueprint for future technology.

• Magnonic Diodes: Just like a diode in electronics lets electricity flow in only one direction (preventing short circuits), this magnetic material acts as a diode for waves.
• Better Computers: As computers get smaller, heat becomes a huge problem. Spin waves generate very little heat. If we can build circuits where information flows only one way without needing complex electronics to stop it from bouncing back, we could create faster, cooler, and more efficient computers.
• Security: Imagine a communication system where a signal can be sent from Point A to Point B, but Point B cannot accidentally send a signal back to Point A. This prevents feedback loops and eavesdropping.

Summary

The scientists discovered that by arranging magnetic layers in a specific "scissors" shape, they can force magnetic waves to travel in only one direction. It's like creating a magnetic one-way street where the traffic can never turn around. This discovery opens the door to a new generation of electronic devices that are faster, more efficient, and capable of controlling information flow in ways we've never seen before.

Technical Summary

1. Problem Statement

The research addresses the challenge of achieving unidirectional energy transport of spin waves (SWs) in magnetic materials. While frequency non-reciprocity (NR)—where the frequency $\omega$ depends on the direction of the wavevector $\vec{k}$, i.e., $\omega(\vec{k}) \neq \omega(-\vec{k})$—is a known phenomenon in multilayers, achieving true unidirectionality (where the group velocity $\vec{v}_g$ maintains the same sign regardless of the sign of $\vec{k}$) is rare and difficult to control.

The authors focus on Synthetic Antiferromagnets (SAFs), specifically symmetric CoFeB/Ru/CoFeB stacks. In these systems, two ferromagnetic layers are coupled antiparallelly by a spacer (Ru). The study investigates the behavior of SWs when the SAF is set in a "scissors state" (a configuration where the magnetizations of the two layers are canted toward an applied in-plane magnetic field but remain antiparallel to each other). The goal is to determine if the layer-to-layer dipolar interactions in this specific state can induce a dispersion relation where energy flows in only one direction for a wide range of wavevectors.

2. Methodology

The study employs a tripartite approach combining experimental characterization, analytical modeling, and numerical simulations:

• Experimental Characterization:

  • Materials: Symmetric SAF stacks (Ta/CoFeB(17nm)/Ru(0.7nm)/CoFeB(17nm)/Ru(0.4nm)/Ta) were fabricated on two substrates: Y-cut LiNbO$_3$ (for electrical measurements) and Si/SiO$_x$ (for optical measurements).
  • Brillouin Light Scattering (BLS): Used to measure the dispersion relations $\omega(\vec{k})$ with wavevector resolution. The setup allowed for varying the angle between the applied field $\vec{H}_0$ and the wavevector $\vec{k}$.
  • Propagating Spin Wave Spectroscopy (PSWS): A device with two inductive antennas was used to measure transmission parameters ($S_{21}$ and $S_{12}$). This allowed the authors to directly observe the direction of energy flow. A "toggle switching" protocol was used to switch between the two degenerate scissors states by rotating the magnetic field, thereby reversing the chirality of the system.

• Analytical Modeling:

  • The authors developed a 2-macrospin approximation model. They treated the SAF as two uniformly magnetized layers.
  • They derived analytical expressions for the group velocities ($v_g = d\omega/dk$) of both acoustical (in-phase) and optical (out-of-phase) modes under various field orientations ($\vec{H}_0 \parallel \vec{k}$, $\vec{H}_0 \perp \vec{k}$, and arbitrary angles).
  • The model utilized Taylor expansions for low wavevectors to provide simple, interpretable formulas for group velocities.

• Numerical Simulations:

  • Micromagnetic Simulations: Performed using Mumax3 software to validate the analytical models and account for effects neglected in the 2-macrospin approximation (such as intra-layer exchange stiffness gradients).
  • Excitation Methods: The simulations used specific spatial and temporal stimuli to selectively excite acoustical or optical modes and to resolve the sign of the wavevector ($+k$ vs. $-k$) in non-reciprocal regimes.

3. Key Contributions

1. Discovery of Unidirectional Acoustical Modes: The paper identifies a unique regime in SAFs where the acoustical spin wave mode exhibits a quasi-linear dispersion relation around $k=0$ when $\vec{H}_0 \parallel \vec{k}$. In this state, the group velocity is positive for both positive and negative wavevectors, meaning energy flows in a single direction regardless of the phase propagation direction.
2. Reconfigurable Unidirectionality: The authors demonstrate that this unidirectional flow is switchable. By toggling the SAF between its two degenerate scissors states (via field rotation), the direction of the unidirectional energy flow can be reversed.
3. Unified Formalism: The authors provide a simplified analytical formalism (Table I in the paper) that predicts group velocities and non-reciprocity magnitudes. This formalism bridges the gap between complex micromagnetic calculations and experimental observations, offering a design tool for non-reciprocal devices.
4. Quantitative Validation: The study quantitatively validates the magnitude of frequency non-reciprocity ($\delta f$), showing it can reach several GHz (e.g., $\sim 3.4$ GHz for acoustical modes at $k=12$ rad/$\mu$m), which is exceptionally large.

4. Key Results

• Dispersion Relations:
  • Optical Mode: Exhibits a tilted "V" shape (non-reciprocal) where the group velocity sign correlates with the wavevector sign (conventional behavior).
  • Acoustical Mode (Scissors State, $\vec{H}_0 \parallel \vec{k}$): Exhibits a quasi-linear dispersion. Crucially, the group velocity remains positive for both $k > 0$ and $k < 0$. This results in a "one-way" energy transfer.
• Experimental Verification (PSWS):
  • In the initial scissors state, strong transmission was observed from Antenna 1 to Antenna 2 (forward), but negligible transmission in the reverse direction.
  • Upon toggling the scissors state, the transmission direction reversed (strong backward, weak forward), confirming the switchable nature of the unidirectionality.
• Model vs. Simulation:
  • The analytical 2-macrospin model accurately predicts the behavior at low fields and low wavevectors.
  • Discrepancies arise at higher fields or wavevectors due to the development of magnetization gradients within the layers (breaking the 2-macrospin assumption), which are captured by the full micromagnetic simulations.
• Non-Reciprocity Magnitude: The non-reciprocity of the acoustical and optical modes is predicted to be equal in magnitude but opposite in sign ($\delta f_{ac} \approx -\delta f_{op}$), a finding confirmed by both simulations and experiments.

5. Significance

• Magnonic Diodes and Logic: The ability to achieve unidirectional propagation without external biasing (other than the static field) is a critical step toward realizing magnonic diodes and chiral logic devices. These components are essential for building low-power, high-speed information processing systems based on spin waves.
• Reconfigurability: The fact that the direction of energy flow can be switched by simply toggling the magnetic state (without changing the device geometry) makes these SAFs highly attractive for reconfigurable magnonic circuits.
• Design Guidelines: The analytical formalism provided allows researchers to predict and design SAF structures with specific non-reciprocal properties, facilitating the optimization of devices for applications like passive non-reciprocal filters and directional emitters.
• Fundamental Physics: The work deepens the understanding of how layer-to-layer dipolar interactions dominate the dispersion relations in SAFs, particularly in the scissors state, revealing a regime where the group velocity decouples from the wavevector sign.

In conclusion, this paper establishes that symmetric synthetic antiferromagnets in a scissors state are a robust platform for generating switchable, unidirectional spin waves, offering a promising pathway for next-generation magnonic technologies.