Disorder-induced chirality in superconductor-ferromagnet heterostructures revealed by neutron scattering and multiscale modeling

By combining neutron scattering experiments with multiscale modeling, this study reveals that intrinsic chemical disorder and compositional gradients in FePd-based superconductor-ferromagnet heterostructures, rather than just interface effects, are the microscopic origin of finite net magnetic chirality that stabilizes chiral magnetic modulations.

The Gist

The Big Picture: Finding the "Twist" in a Straight Line

Imagine you have a block of metal that is supposed to be perfectly symmetrical, like a square tile. If you look at it from the top, bottom, left, or right, it looks exactly the same. In physics, this is called being "centrosymmetric." Usually, if a material is perfectly symmetrical, its magnetic spins (the tiny internal compasses of the atoms) line up in a straight, boring line. They don't like to twist or turn.

However, this paper is about a special metal alloy called FePd (Iron and Palladium). The scientists wanted to know: Why does this metal sometimes develop a "twist" or "chirality" (a handedness, like a left-handed screw) even though it looks perfectly symmetrical on paper?

This twist is important because it helps create a "superpower" in hybrid materials where superconductors (materials that conduct electricity with zero resistance) meet magnets. This could lead to better quantum computers and faster memory.

The Mystery: The "Messy Room" Theory

For a long time, scientists thought that for a metal to have this magnetic twist, it needed a special interface with another material (like a heavy metal) to force the twist. They thought the "perfect" metal itself was too boring to do it on its own.

But this team of researchers (from Sweden, Brazil, Germany, and France) had a hunch. They suspected that the metal wasn't actually perfect. They thought the "messiness" inside the metal was the real culprit.

The Analogy: The Perfectly Organized Bookshelf vs. The Messy One

• The Perfect Bookshelf: Imagine a bookshelf where every red book is on the left and every blue book is on the right, perfectly alternating. This is the "ordered" state. In this state, the magnetic spins are straight and symmetrical.
• The Messy Bookshelf: Now, imagine someone starts shuffling the books. A red book ends up next to another red book. A blue book is stuck in the middle of red ones. This is chemical disorder or intermixing.

The scientists discovered that this "messiness" (atoms of Iron and Palladium swapping places where they shouldn't) breaks the perfect symmetry. It's like a slight tilt in the floor of the room. Because of this tilt, the magnetic spins can't stand up straight anymore; they have to lean or twist to find a comfortable position. This creates the "chirality."

How They Solved the Puzzle

The team used a "detective" approach, combining two different ways of looking at the problem:

1. The Experimental Detectives (Looking at the Crime Scene)

They built two samples:

• Sample A: Just the messy Iron-Palladium metal.
• Sample B: The messy metal sitting next to a Superconductor (Niobium).

They used a giant "neutron camera" (a technique called GISANS) to take pictures of the magnetic spins. Neutrons are like tiny, invisible bullets that bounce off the magnetic fields.

• The Result: They saw that the spins were indeed twisting. Even the sample without the superconductor had a twist! This proved that the twist didn't come from the superconductor; it came from the metal itself.
• The Clue: They also looked at the metal under a super-microscope and saw that the atoms were indeed "messy." There was a gradient, meaning the top of the metal was messier than the bottom. It was like a cake where the ingredients weren't fully mixed, creating a gradient of flavor from top to bottom.

2. The Theoretical Detectives (Simulating the Crime)

Knowing the metal was messy, they needed to prove why that mess caused a twist.

• The Old Way: Traditional computer models treat the metal like a "smoothie." They average out all the atoms, assuming the red and blue books are perfectly mixed. This model said, "No twist here!"
• The New Way (The Secret Weapon): The team used a Deep Learning AI (a Graph Neural Network). Think of this AI as a super-smart student who didn't just memorize the rules of the game but actually played millions of games to learn the nuances.
  • Instead of averaging the atoms, the AI looked at every single atom and its specific neighbors. It saw that when an Iron atom is surrounded by a specific, messy mix of Palladium atoms, it creates a tiny "push" (called the Dzyaloshinskii-Moriya interaction or DMI) that forces the spin to twist.
  • The AI calculated that this "push" is strong enough to create the twist observed in the experiments.

The "Aha!" Moment

The most exciting discovery was that the "twist" wasn't just a surface effect. It was intrinsic.

The Analogy: The Crumpled Paper
Imagine a piece of paper. If you crumple it, it becomes messy and irregular. If you try to draw a straight line on it, the line will curve because of the crumples.

• Old Belief: The curve only happens if you put the paper on a curved table (the interface).
• New Discovery: The paper is crumpled on its own (the disorder). The curve comes from the paper itself, not the table.

The paper showed that the "crumples" (atomic disorder and gradients) inside the FePd metal are strong enough to create a magnetic twist all by itself.

Why Does This Matter?

1. It Changes the Rules: We used to think we needed complex, expensive layers of different metals to create these magnetic twists. Now we know that a "messy" version of a simple metal can do it too.
2. Better Tech: These twists are crucial for Superconducting Spintronics. This is the technology behind future quantum computers and ultra-fast, low-energy memory. If we can control these twists better, we can build better devices.
3. AI in Science: This paper is a great example of how Artificial Intelligence is helping physics. The AI didn't just crunch numbers; it learned the complex "personality" of the atoms, allowing the scientists to bridge the gap between the tiny atomic world and the larger world we can see.

Summary in One Sentence

The scientists discovered that the "imperfections" and "messiness" inside a metal alloy are actually the secret recipe that causes its magnetic fields to twist, a discovery made possible by using a smart AI to simulate the chaotic dance of atoms.

Technical Summary

1. Problem Statement

Superconductor-ferromagnet (S/F) heterostructures are critical for spintronics and quantum computing, particularly regarding the generation of long-range triplet superconductivity and topological states. These phenomena rely heavily on magnetic chirality (non-collinear spin textures like skyrmions or chiral domain walls).

• The Core Issue: In nominally centrosymmetric ferromagnets (like $L1_0$-FePd), the bulk Dzyaloshinskii-Moriya interaction (DMI) should theoretically vanish due to inversion symmetry. Consequently, observed chirality is often attributed solely to interface effects (e.g., S/F or F/heavy-metal interfaces).
• The Gap: The microscopic origin of net chirality within the bulk of these nominally centrosymmetric ferromagnetic layers remains unclear. Specifically, it is unknown whether intrinsic structural disorder and compositional gradients can generate sufficient DMI to stabilize chiral textures without relying on interface proximity.

2. Methodology

The authors employed a comprehensive multiscale approach combining advanced experimental characterization with deep-learning-assisted theoretical modeling.

A. Experimental Characterization

• Samples: Two high-quality $L1_0$-FePd films with perpendicular magnetic anisotropy (PMA) were studied:
  1. i.Nb/FePd: A superconductor/ferromagnet heterostructure (Nb/FePd).
  2. ii.FePd: A single FePd film (control).
• Structural Analysis:
  • HAADF-STEM & EDX: Revealed atomic intermixing, anti-phase boundaries (APBs), and a depth-dependent compositional gradient (Fe/Pd ratio varying across the film thickness).
  • XRD & SQUID: Confirmed partial long-range order ($0 < S < 1$) and strong uniaxial anisotropy.
• Magnetic Chirality Detection:
  • Polarization-Analyzer Grazing-Incidence Small-Angle Neutron Scattering (PA-GISANS): Performed at room temperature. This technique probes magnetic correlations on the nanometer scale. By analyzing spin-flip asymmetry ($\Delta I_{SF}$), the authors quantified the net chirality and distinguished between in-plane (Bloch-type) and out-of-plane (Néel-type) contributions.

B. Theoretical Modeling

• First-Principles (DFT): Used RS-LMTO-ASA to calculate exchange ($J_{ij}$) and DMI ($D_{ij}$) interactions in chemically disordered FePd supercells.
• Deep Learning (SAGNN): Developed a Species-Aware Graph Neural Network trained on DFT data.
  • Purpose: To act as a surrogate model that captures the variability of magnetic interactions based on local atomic environments, bridging the gap between atomistic DFT scales and mesoscopic system sizes ($\sim 10^8$ interactions).
  • Advantage: Unlike shell-averaged models, SAGNN preserves local fluctuations essential for capturing disorder-induced effects.
• Mesoscopic Modeling:
  • Implemented a modified gradient model incorporating the experimentally observed compositional gradients.
  • Used a flat spin-spiral ansatz to minimize energy and determine the ground state wavevector ($Q$) and chirality measure ($C_N$).

3. Key Contributions

1. Identification of Bulk Chirality Source: Demonstrated that chemical disorder (atomic intermixing) combined with compositional gradients is a sufficient intrinsic mechanism to generate net magnetic chirality in nominally centrosymmetric FePd, independent of interface effects.
2. Development of SAGNN: Created a machine-learning framework capable of predicting magnetic interactions ($J_{ij}, D_{ij}$) with ab-initio accuracy across mesoscopic scales, specifically handling the statistical variability of disordered alloys.
3. Mechanism Elucidation: Showed that disorder breaks local inversion symmetry, inducing finite DMI. This DMI stabilizes chiral finite-$q$ magnetic modulations with a mixed Bloch-Néel character.
4. Quantitative Agreement: Theoretically predicted modulation lengths ($\sim 68\text{--}72$ nm) and chirality signatures that align qualitatively and semi-quantitatively with experimental GISANS data ($\sim 185\text{--}217$ nm), bridging the gap between atomistic defects and macroscopic observation.

4. Key Results

Experimental Findings

• Structural Disorder: Both samples exhibited partial $L1_0$ order ($S \approx 0.52\text{--}0.70$), atomic intermixing, and a clear gradient in defect density and composition from the substrate to the surface.
• Neutron Scattering (PA-GISANS):
  • Observed a finite net magnetic chirality at room temperature.
  • The spin-flip asymmetry was significantly larger for in-plane polarization ($\nu = P_y$) than out-of-plane ($\nu = P_z$), indicating dominant in-plane (Bloch-type) chiral contributions.
  • A non-zero out-of-plane asymmetry was linked to the depth-dependent magnetic inhomogeneity.
  • The i.Nb/FePd sample showed stronger chirality than ii.FePd, suggesting sensitivity to the specific defect landscape.

Theoretical Findings

• Disorder-Induced DMI: In pristine $L1_0$ FePd, DMI vanishes. However, in disordered models (intermixing), finite DMI emerges.
• Radial Decay of DMI: The disorder-induced DMI decays as $r^{-2.75}$, which is slower than the canonical $r^{-3}$ decay of RKKY interactions. This implies longer-range coherence of the chiral interaction.
• Mesoscopic Stability:
  • The gradient model predicted a stable chiral ground state with a modulation period of $\sim 71$ nm.
  • The chirality measure ($C_N \approx 0.98$) indicated a strong admixture of Néel and Bloch characters, consistent with the experimental observation of finite asymmetry in both polarization directions.
  • The model confirmed that the compositional gradient creates an effective DMI field with a finite odd component under reflection, driving the net handedness.

5. Significance

• Paradigm Shift: This work challenges the assumption that chirality in centrosymmetric S/F hybrids is exclusively an interface phenomenon. It establishes that bulk disorder and gradients are intrinsic sources of chirality.
• Methodological Advancement: The integration of SAGNN with multiscale modeling provides a robust framework for studying complex disordered magnetic systems where direct DFT is computationally prohibitive. It successfully captures the "frustration" and local environment variability necessary for emergent phenomena.
• Implications for Applications: Understanding that disorder can induce chirality is crucial for:
  • Superconducting Spintronics: Designing S/F devices where triplet correlations are controlled not just by interfaces but by bulk engineering of disorder.
  • Topological Quantum Computing: The stability of skyrmions and Majorana modes in these systems may be tunable via compositional gradients and defect engineering.
  • Material Design: Future strategies for S/F hybrids can leverage controlled disorder to stabilize chiral textures without requiring specific heavy-metal underlayers.

In conclusion, the paper provides a unified microscopic and multiscale explanation for the origin of chirality in FePd-based heterostructures, proving that structural disorder is not merely a defect to be minimized but a functional parameter that can drive complex magnetic textures.