🔬 materials science

Exchange striction determines how fast antiferromagnetic insulators demagnetize

This study identifies exchange striction—the sensitivity of exchange coupling to atomic displacements—as the key material parameter determining the ultrafast demagnetization speed of antiferromagnetic insulators, enabling the prediction and screening of candidate materials for high-speed memory applications.

Original authors: Aleksandr Buzdakov, Ravi Kaushik, Nikolai Khokhlov, Sergey Artyukhin, Alexey Kimel

Published 2026-09-02

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Aleksandr Buzdakov, Ravi Kaushik, Nikolai Khokhlov, Sergey Artyukhin, Alexey Kimel

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

In the world of magnets, there is a fundamental rule that governs how quickly a material can be turned off. For decades, scientists have known that when you hit a magnet with a laser pulse, its internal order can vanish in the blink of an eye. This process, called demagnetization, is the key to making faster computers and more efficient data storage. However, the speed at which this happens depends heavily on what the magnet is made of. In metals, the process is incredibly fast because the electrons can easily carry away the energy and the "spin" of the atoms. But in insulators—materials that do not conduct electricity, like many ceramics and crystals—the electrons are stuck in place. They cannot carry the energy away, so the heat must be transferred through the rigid structure of the material itself. This creates a bottleneck. For a long time, researchers have watched different insulating magnets melt their magnetic order at wildly different speeds, ranging from a few trillionths of a second to a few billionths of a second, without understanding why some are so much faster than others.

A team of researchers has now solved this mystery by looking at two specific crystals: chromium oxide and iron borate. Both are insulating magnets that look very similar under a microscope, sharing the same basic crystal structure and the same type of magnetic ordering. Yet, when hit with an ultrafast laser pulse, they behave completely differently. The chromium oxide loses its magnetic order in less than two trillionths of a second. The iron borate, sitting right next to it in the periodic table of materials, takes more than a hundred times longer to do the same thing. The researchers discovered that the secret lies in how tightly the atoms are packed together and how the magnetic forces between them react when the atoms jiggle.

The team used a technique called second-harmonic generation, which acts like a high-speed camera for magnetic fields, to watch the chromium oxide crystal as it was heated by a laser pulse. They observed that once the crystal lattice was heated past a specific temperature, the magnetic order collapsed almost instantly. To understand why this happened so fast, they turned to computer simulations that modeled the movement of every single atom and spin in the material. These simulations revealed that the speed of demagnetization is controlled by a mechanism called exchange striction. This is a fancy way of describing how the magnetic force between two atoms changes when the distance between them shifts even slightly.

In the chromium oxide crystal, the magnetic atoms are packed very closely together. Because they are so near each other, the magnetic force between them is extremely sensitive to tiny movements. When the laser heats the crystal, the atoms vibrate, and because the magnetic force is so sensitive to these vibrations, the energy flows rapidly from the vibrating atoms into the magnetic spins, scrambling the order. In the iron borate crystal, the magnetic atoms are separated by other atoms, making the connection more like a long, loose chain. The magnetic force there is much less sensitive to movement, so the energy flows much more slowly. The researchers found that the chromium oxide's magnetic force is about ten times more sensitive to atomic movement than that of the iron borate. This difference, combined with the fact that the vibrations in chromium oxide can more easily break down into pairs of magnetic waves, creates a perfect storm that allows the material to lose its magnetism at record speeds.

The study confirms that the speed limit for these insulating magnets is not a fixed property of the material type, but a calculable feature based on how the atoms are arranged. By measuring how much the magnetic force changes when the atoms move, scientists can now predict how fast a new material will demagnetize before they even build it. This discovery opens a practical path for designing the next generation of ultrafast memory devices. Instead of guessing which materials might work, engineers can now look at the atomic structure and calculate the speed, selecting only those with the tightest, most sensitive atomic connections to build devices that operate at the very edge of physical possibility.

Technical Summary: Exchange Striction Determines Demagnetization Speed in Antiferromagnetic Insulators

Problem Statement
Ultrafast demagnetization in antiferromagnetic insulators exhibits a vast spread in timescales, ranging from picoseconds to nanoseconds across different compounds. While the speed limit for metallic ferromagnets is understood via Elliott-Yafet spin-flip scattering, the mechanism governing the speed limit in insulating antiferromagnets remains unresolved. In compensated antiferromagnets, where no net angular momentum must be transferred to the lattice, the demagnetization rate is theoretically limited by the energy flow from the laser-heated lattice into the spin subsystem. However, no material parameter has been identified that predicts or explains the magnitude of this spread or the specific speed at which a new compound will demagnetize.

Methodology
The authors employed a combined experimental and theoretical approach to investigate ultrafast demagnetization in chromium oxide ( $\text{Cr}_2\text{O}_3$ ) and compare it with iron borate ( $\text{FeBO}_3$ ), a structurally similar compound sharing the $\bar{R}3c$ space group.

Experimental Approach:
- Technique: Time-resolved second-harmonic generation (SHG) was used to probe magnetic order dynamics. Circularly polarized probe pulses (1200 nm) generated SHG signals (600 nm) sensitive to the Néel vector orientation.
- Protocol: A 400 nm pump pulse heated the lattice subsystem, while the SHG contrast between magnetic domains tracked the decay of the order parameter $|L|$ . Experiments were conducted at various pump fluences to achieve final temperatures both below and above the Néel temperature ( $T_N \approx 307.3$ K).
- Metric: The characteristic demagnetization time $\tau$ was defined as the delay at which the SHG contrast decayed to $1/e$ of its total drop.
Theoretical Approach:
- Atomistic Spin Dynamics (ASD): Simulations modeled the response of spins to a sudden lattice temperature jump using a phenomenological Gilbert damping.
- First-Principles Calculations: Density Functional Theory with Hubbard correction (DFT+U) was used to calculate Heisenberg exchange constants ( $J_{ij}$ ) and their derivatives with respect to ionic displacements ( $\partial J/\partial u$ ), known as exchange striction.
- Coupled Spin-Lattice Dynamics (SLD): Non-perturbative simulations integrated the Landau-Lifshitz-Gilbert equation for spins and Newton's equations for ions, using exchange tensors and force constants derived entirely from DFT+U without adjustable parameters.
- Kinematic Analysis: The study calculated the phase space for one-phonon-to-two-magnon decay processes, comparing the overlap between phonon dispersions and the two-magnon continuum in both materials.

Key Results

Experimental Disparity: $\text{Cr}_2\text{O}_3$ demagnetizes within 2 ps when driven above $T_N$ , a rate two orders of magnitude faster than $\text{FeBO}_3$ . While $\text{FeBO}_3$ exhibits critical slowing down near $T_N$ , $\text{Cr}_2\text{O}_3$ bypasses this region to reveal an intrinsic energy-transfer rate that is exceptionally fast.
Limitations of Phenomenological Models: Standard ASD simulations with constant damping captured the critical slowing down near $T_N$ but failed to reproduce the steep drop in $\tau$ observed above $T_N$ in experiments. This indicated a missing temperature-dependent spin-lattice coupling mechanism.
Role of Exchange Striction: First-principles calculations identified exchange striction (the dependence of exchange coupling on atomic displacement) as the governing factor.
- Structural Origin: $\text{Cr}_2\text{O}_3$ features short Cr-Cr contacts (2.65 Å and 2.89 Å) where magnetic interaction is dominated by direct $t_{2g}$ - $t_{2g}$ overlap. This results in exchange-striction constants ( $|\partial J/\partial u|$ ) of $\sim 100$ meV/Å. In contrast, $\text{FeBO}_3$ relies on oxygen-mediated superexchange with no short cation-cation contacts, yielding $|\partial J/\partial u| \lesssim 10$ meV/Å.
- Phase Space: The strong exchange in $\text{Cr}_2\text{O}_3$ creates a wide magnon band, causing the two-magnon continuum to overlap almost the entire phonon spectrum (up to 25 THz). In $\text{FeBO}_3$ , the continuum is confined below 12 THz, leaving most phonons without decay channels.
Quantitative Agreement: The product of the phase space factor (approx. 3x larger in $\text{Cr}_2\text{O}_3$ ) and the matrix element factor (approx. $10^2$ larger due to striction) predicts a rate ratio of $\sim 300$ . This aligns with the experimentally observed ratio of $\sim 100$ .
Parameter-Free Simulation: Coupled SLD simulations, using only ab initio parameters, reproduced the order of magnitude of the measured ratio ( $\tau_{\text{Cr}_2\text{O}_3} \approx 10$ ps vs. $\tau_{\text{FeBO}_3} \approx 500$ ps), confirming that the displacement dependence of exchange integrals is the primary driver.

Significance and Claims
The paper establishes that the demagnetization speed of an insulating antiferromagnet is not an empirical mystery but a computable material property determined by three specific ingredients:

The exchange-striction tensor ( $\partial J/\partial u$ ).
The magnitude of thermal ionic displacements.
The overlap of the two-magnon continuum with the phonon spectrum.

The authors claim that this framework allows for the prediction of demagnetization speeds in candidate materials prior to synthesis. Specifically, materials with short cation-cation bonds (enhancing direct exchange and striction) will demagnetize faster, while those relying solely on oxygen-mediated superexchange (like calcite-structure $\text{MnCO}_3$ ) will be slower. This provides a practical pathway for screening materials for thermally assisted antiferromagnetic memory and advances the understanding of ultrafast control in insulating magnetic systems.

Technical Summary: Exchange Striction Determines Demagnetization Speed in Antiferromagnetic Insulators

More like this