🔬 mesoscale physics

Design of magnonic waveguides using surface anisotropy-induced Bragg mirrors

This paper proposes and theoretically analyzes a novel magnonic waveguide design in a Co $_{20}$ Fe $_{60}$ B $_{20}$ layer that utilizes surface anisotropy-induced Bragg mirrors to effectively confine and guide high-frequency, high-velocity spin waves while overcoming limitations associated with non-uniform demagnetizing fields.

Original authors: Grzegorz Centała, Jarosław W. Kłos

Published 2026-02-03

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Grzegorz Centała, Jarosław W. Kłos

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to send a message across a crowded room using ripples in a pond. In the world of tiny computer chips, these "ripples" are called spin waves (or magnons), and they carry information instead of electricity. To make these chips work, we need to build "roads" or waveguides to direct these ripples exactly where we want them to go.

For a long time, building these roads has been tricky. Here is the problem the authors faced and the clever solution they designed.

The Problem: The "Traffic Jam" and the "Leaky Road"

Think of a traditional spin-wave road like a narrow strip of metal.

The Slowdown: If you try to send a wave down this strip, it often gets stuck or moves very slowly, like a car stuck in heavy traffic.
The Leaks: If you try to speed it up by pushing it with a strong magnetic force, the road develops "potholes" at the edges. The waves start leaking out of the road or getting stuck at the edges, creating noise and confusion.
The Frequency Limit: Most of these old roads can only handle low-frequency waves. If you try to send a high-speed, high-frequency message, the road simply doesn't work; the waves can't get through.

The Solution: The "Magic Mirror" Road

The authors, Grzegorz Centała and Jarosław W. Kłos, proposed a new way to build this road. Instead of cutting a physical strip out of metal, they kept the metal layer smooth and uniform (like a calm, flat lake).

Then, they used a special trick: Surface Anisotropy.

The Analogy: Imagine you have a flat, smooth floor. You can't change the floor itself, but you can place invisible "speed bumps" or "magnetic fences" in specific patterns on the surface.
The Bragg Mirrors: They placed these magnetic fences in a repeating pattern (like a fence with alternating tall and short posts) on both sides of a central path. These fences act like Bragg mirrors.

In physics, a Bragg mirror is like a wall that reflects specific types of waves perfectly. By creating two of these mirrors facing each other, they trapped the spin waves in the middle, creating a safe, high-speed highway.

How It Works in Everyday Terms

The Highway: The center of the road is a wide, smooth strip where the waves can zoom along.
The Barriers: On the left and right, there are repeating patterns of magnetic "fences." These fences are so good at reflecting waves that the waves bounce back and forth inside the center strip, unable to escape.
The Superpower: Because the road is made of a single, uniform piece of metal (not a cut-out strip), there are no "edge potholes" to cause leaks. The waves can travel at high speeds and carry high-frequency messages (up to 45 GHz), which is much faster than what older designs could handle.

The Trade-off: Speed vs. Staying in the Lane

The paper highlights a balancing act, much like driving a race car:

Speed (Group Velocity): The waves move very fast, which is great for sending data quickly.
Confinement (Localization): The waves stay tightly packed in the center lane, so they don't crash into neighboring roads (crosstalk).

The authors found that if the metal layer is too thick, the waves go fast but might wander off the road. If it's too thin, they stay on the road but move slower. They calculated the "Goldilocks" thickness (6 nanometers) that gives the best mix of speed and safety.

Why This Matters (According to the Paper)

The paper claims this design solves three major headaches:

No Leaks: It avoids the "edge modes" (waves getting stuck at the sides) that plague traditional strip designs.
High Speed: It allows waves to travel fast without needing a massive external magnetic field that causes other problems.
High Frequency: It is one of the few designs that can guide waves at frequencies higher than the natural limit of the material, opening the door to faster data processing.

In short, the authors built a "magnetic highway" using invisible fences rather than physical walls, allowing information to zip through at high speeds without leaking out or getting stuck.

Technical Summary: Design of Magnonic Waveguides using Surface Anisotropy-Induced Bragg Mirrors

Problem Statement
Conventional magnonic waveguides, typically realized as flat ferromagnetic strips, face significant limitations in signal transmission efficiency and reconfigurability. Their performance is often constrained by non-uniform demagnetizing fields and reduced group velocities. Specifically, aligning the magnetization along the waveguide axis (to avoid external fields) drastically reduces group velocity, while applying a strong perpendicular field to increase velocity induces static demagnetizing fields at the edges, leading to unwanted edge modes. Furthermore, alternative approaches involving spatial modification of material composition (e.g., varying saturation magnetization) or surface anisotropy are generally limited to guiding spin waves only at frequencies below the ferromagnetic resonance (FMR) of the pristine layer. As the operating frequency approaches the FMR, the confinement strength diminishes, restricting the usable frequency range.

Methodology
The authors propose a waveguide design that overcomes these limitations by utilizing a uniform, low-damping ferromagnetic layer ( $Co_{20}Fe_{60}B_{20}$ ) where confinement is achieved not by geometric patterning of the magnetic material itself, but by spatially modulating the surface anisotropy ( $K_s$ ).

Structure: The system consists of a 6-nm-thick $CoFeB$ layer sandwiched between non-magnetic regions. Two semi-infinite one-dimensional magnonic crystals act as Bragg mirrors, created by periodically alternating regions of high surface anisotropy (50 nm strips) with pristine regions (50 nm strips). A central defect, defined by a wider 150-nm strip of surface anisotropy, serves as the waveguide channel.
Theoretical Framework: Magnetization dynamics are described using the Landau-Lifshitz equation, linearized for a system saturated by an in-plane external magnetic field ( $\mu_0 H_0 = 500$ mT) applied perpendicular to the waveguide axis (Damon-Eshbach geometry). The surface anisotropy is incorporated via specific boundary conditions on the magnetization at the film interfaces.
Simulation: The governing equations, including the exchange field, dipolar field, and magnetostatic potential, were solved using the finite element method (FEM) in COMSOL Multiphysics. The study analyzed both the unit cell of the infinite magnonic crystal (to determine frequency gaps) and the full waveguide structure (two 21-period mirrors with a central defect) to calculate dispersion relations, group velocities, and spatial mode profiles.

Key Contributions and Results

High-Frequency Waveguiding: The proposed design successfully guides spin waves at frequencies significantly higher than the FMR of the uniform ferromagnetic layer. Unlike conventional anisotropy-induced waveguides, which fail above the FMR, this system utilizes the frequency gaps of the Bragg mirrors to confine spin waves. In these gaps, the wave vector in the mirror regions becomes complex, ensuring exponential decay and strong confinement even at high frequencies (20–45 GHz).
Dispersion and Group Velocity: The study calculates the dispersion relations for multiple waveguide modes. The results show that the group velocity ( $v_g$ ) increases with the wave vector ( $k_z$ ). For the chosen parameters, the group velocities are sufficiently high for practical signal transmission in low-damping materials like CoFeB. The design avoids the velocity reduction associated with magnetization aligned along the waveguide axis.
Localization and Crosstalk Suppression: The localization of modes is quantified using the Inverse Participation Ratio (IPR). The analysis reveals that lower-frequency modes (below the uniform layer's FMR) exhibit strong localization due to tunneling through the Bragg mirrors, while higher-frequency modes rely entirely on the Bragg gap mechanism. The study determines the isolation distance ( $L_i$ ) required to prevent crosstalk between adjacent waveguides, finding that for specific modes, isolation can be achieved within a few defect widths.
Absence of Edge Modes: By maintaining a uniform magnetic layer and avoiding geometric edges typical of strip waveguides, the design eliminates the static demagnetizing fields that usually induce edge modes, thereby preserving the integrity of the guided signal.

Significance and Claims
The paper claims that this design offers a robust solution for integrated magnonic systems by combining high group velocities with strong spatial confinement. The primary significance lies in the ability to operate in frequency ranges previously inaccessible to simple anisotropy-modulated waveguides. The authors emphasize that the trade-off between high group velocity and strong confinement can be managed by optimizing the ferromagnetic layer thickness and the magnitude of the surface anisotropy. The proposed architecture is presented as a feasible route for creating reconfigurable, high-performance magnonic circuits that do not suffer from the limitations of static demagnetizing fields or restricted frequency bands. The work suggests that such waveguides could be fabricated by modulating the thickness of an MgO capping layer on a CoFeB film, providing a pathway for experimental realization.