Strain effects in [001] textured Co80Ir20 thin films with negative magnetocrystalline anisotropy

This study demonstrates that magnetoelastic effects significantly influence the magnetic anisotropy of [001]-textured Co80Ir20 thin films, where the choice of underlayer (Ta vs. Pt) induces varying degrees of strain that alter the effective anisotropy field, necessitating the consideration of stress effects when estimating magnetocrystalline contributions.

The Gist

The Big Picture: The "Stressed-Out" Magnet

Imagine you have a thin sheet of metal (a magnetic film) that you want to use in a high-tech device, like a super-fast hard drive or a microwave sensor. You want the magnetic "compass needles" inside this metal to all point in the same direction, lying flat on the surface, so they can switch directions quickly and efficiently.

The scientists in this paper studied a specific alloy called Cobalt-Iridium (CoIr). This material is special because, naturally, it hates pointing up or down; it really wants to lie flat. This is called "negative magnetocrystalline anisotropy." Think of it like a flat pancake that naturally wants to stay on the table rather than standing on its edge.

The Problem:
The researchers noticed that when they made these films, the magnetic behavior changed depending on what they put underneath the film. Some films were "stiff" and hard to flip, while others were "loose" and easy to flip.

They suspected that the secret wasn't just the material itself, but stress (or strain). Imagine stretching a rubber band. If you stretch it, it changes shape and behaves differently. The scientists wanted to prove that the "rubber band" effect (stress) was the hidden culprit changing how the magnets worked.

---

The Experiment: Building Different "Sandwiches"

To test this, the team built four different "sandwiches." The filling was always the same: a 24-nanometer-thick layer of the Cobalt-Iridium alloy. But the bread (the layers underneath and on top) changed:

1. Sandwich A & C: The filling sat directly on a Tantalum (Ta) bottom layer.
2. Sandwich B & D: The filling sat on a Platinum (Pt) bottom layer.
3. They also varied the top layer (some had Ta, some had Pt) to see if the "top bun" mattered.

The Analogy:
Think of the Cobalt-Iridium layer as a group of dancers trying to line up perfectly in a row.

• Tantalum (Ta) is like a bumpy, uneven dance floor. It doesn't hold the dancers' feet very well, so they get a bit jumbled and stretched out.
• Platinum (Pt) is like a smooth, perfectly polished floor. It helps the dancers line up in a very neat, tight formation.

---

What They Found: The "Stretch" Changes the Dance

Using X-rays (like a super-powerful camera that can see the atomic structure), they looked at how the atoms were spaced out.

1. The "Stretch" (Strain):
They found that the Cobalt-Iridium layer was being squeezed and stretched differently depending on the bottom layer.

• On Tantalum, the atoms were pulled apart more (high tensile strain). It was like the dancers were being pulled by ropes from the sides, stretching their formation.
• On Platinum, the atoms were much more relaxed.

2. The Magnetic Result:
This stretching changed the magnetic rules.

• The Ta Samples (Stretched): These films acted like they had a huge extra force pushing them to stay flat. The magnetic field required to flip them was very strong.
• The Pt Samples (Relaxed): These films behaved almost exactly as theory predicted they should, with just the natural "flatness" of the material.

The "Aha!" Moment:
The scientists realized that the "extra force" in the Ta samples wasn't coming from the material's natural personality (magnetocrystalline anisotropy). It was coming from the stress of being stretched.

They calculated that this "stress-induced" force was almost as strong as the material's natural magnetic personality. In fact, if you ignored the stress, your math would be completely wrong.

The Takeaway: Don't Ignore the Rubber Band

Why does this matter?
In the past, scientists studying these magnetic films often assumed that the magnetic behavior was 100% due to the material's chemistry and crystal structure. They ignored the "rubber band" effect (stress).

This paper says: "Stop ignoring the stress!"

If you are designing a new magnetic device, you can't just pick a material and hope for the best. You have to carefully choose what you build it on top of.

• If you want a super-strong, stable magnetic layer, you might want to use a bottom layer that stretches the film (like Tantalum).
• If you want a layer that behaves exactly as the textbook says, you might use a smooth layer (like Platinum).

Summary in One Sentence

The researchers discovered that the "bottom layer" of a magnetic film acts like a pair of hands stretching the film; this stretching creates a hidden magnetic force that is just as important as the material's own natural properties, meaning engineers must control this stress to build better magnetic devices.

Technical Summary

1. Problem Statement

Ferromagnetic thin films with negative magnetocrystalline anisotropy (MCA), such as Co-rich Co$_{100-x}$Ir$_x$ alloys, are critical for advanced magnetic applications (e.g., microwave devices, spin-torque oscillators, and perpendicular magnetic recording underlayers). In these systems, the negative MCA reinforces the shape anisotropy, favoring strong in-plane magnetization alignment in [001]-textured films.

However, a significant gap exists in current literature: magnetoelastic effects (strain) are frequently overlooked when analyzing the effective magnetic anisotropy. Previous studies often attribute the total anisotropy solely to shape and magnetocrystalline contributions, potentially misinterpreting the magnitude of the intrinsic MCA. This paper investigates whether stress-induced anisotropy significantly alters the magnetic properties of Co$_{80}$Ir$_{20}$ films grown on Si/SiO$_2$ substrates with different underlayers (Ta vs. Pt).

2. Methodology

The authors synthesized and characterized four distinct multilayer structures using magnetron sputtering at room temperature on Si/SiO$_2$ substrates. All samples contained a nominal 20 nm Co$_{80}$Ir$_{20}$ (CoIr) layer but varied in seed and capping layers to induce different strain states and textures:

• CoIr-1: Ta(1.5)/CoIr(20)/Pt(5)
• CoIr-2: Ta(1.5)/Pt(5)/CoIr(20)/Ta(1.5)
• CoIr-3: Ta(1.5)/CoIr(20)/Ta(1.5)
• CoIr-4: Ta(1.5)/Pt(5)/CoIr(20)/Pt(5)
  (Thicknesses in nm)

Characterization Techniques:

• Structural Analysis: X-ray Diffraction (XRD) for lattice parameters and phase identification; Rocking curves and $\omega$-scans for texture quality; Williamson-Hall analysis to separate grain size from microstrain; X-ray Reflectivity (XRR) for layer thickness.
• Strain Analysis: $\sin^2\omega$ (or $\sin^2\psi$) scans around the CoIr (002) peak to determine out-of-plane strain ($\varepsilon_{33}$) and in-plane stress.
• Magnetic Characterization:
  • DC Magnetization: Vibrating Sample Magnetometer (VSM) and Magneto-Optic Kerr Effect (MOKE) for saturation magnetization ($M_s$), coercivity ($H_c$), and hysteresis loops at various temperatures (4–300 K).
  • Ferromagnetic Resonance (FMR): X-band (9.5 GHz) and broadband (1–20 GHz) FMR to determine effective anisotropy fields ($H_{eff}$), damping parameters ($\alpha$), and $g$-factors.

3. Key Results

A. Structural and Strain Properties

• Lattice Parameter Shift: XRD revealed a distinct shift in the CoIr (002) peak position. Samples with a Ta underlayer (CoIr-1 and CoIr-3) exhibited a smaller interplanar distance ($d_{002} \approx 0.2070$ nm) compared to those with a Pt underlayer (CoIr-2 and CoIr-4, $d_{002} \approx 0.2075$ nm).
• Strain State: $\sin^2\omega$ scans indicated a negative out-of-plane strain ($\varepsilon_{33} < 0$) for all samples, implying tensile in-plane stress. The magnitude of this strain was significantly larger for Ta-underlayer samples ($\varepsilon_{33} \approx -2.9 \times 10^{-3}$) compared to Pt-underlayer samples ($\varepsilon_{33} \approx -0.3 \times 10^{-3}$ to $-1.3 \times 10^{-3}$).
• Texture and Grain Size: Despite the strain differences, the average grain size remained constant (~20 nm) across all samples. However, CoIr-1 (Ta seed, Pt cap) showed a significantly broader rocking curve (11.4° vs. ~5°), indicating a more disordered [001] texture compared to the others.

B. Magnetic Properties

• Saturation Magnetization: All samples showed similar $M_s$ values (~910 emu/cm$^3$) at room temperature.
• Anisotropy Fields (FMR):
  • Pt Underlayers (CoIr-2, 4): The effective anisotropy field ($H_{eff}$) was close to the theoretical shape anisotropy ($4\pi M_s \approx 11.4$ kOe). This suggests that magnetocrystalline and magnetoelastic contributions largely cancel out or are minimal.
  • Ta Underlayers (CoIr-1, 3): $H_{eff}$ was significantly higher (18.2 kOe and 20.3 kOe, respectively), indicating a strong additional in-plane anisotropy.
• Damping: Samples with Pt interfaces (CoIr-1, 2, 4) exhibited higher damping parameters ($\alpha \approx 0.085$) compared to the Ta-only sample (CoIr-3, $\alpha \approx 0.058$), attributed to stronger spin-mixing conductance at the Pt/FM interface.

C. Correlation of Strain and Anisotropy

The study established a direct correlation between the underlayer material, the induced strain, and the magnetic anisotropy:

1. Ta Underlayers induce high tensile in-plane strain.
2. This strain generates a magnetoelastic anisotropy field ($H_{ME}$) that adds to the shape anisotropy.
3. Using a simple model based on pure Cobalt parameters (due to lack of specific CoIr constants), the authors estimated that the stress-induced anisotropy field is on the order of 6–9 kOe.
4. This calculated value matches the experimental excess anisotropy observed in Ta-underlayer samples, confirming that strain is a dominant factor.

4. Key Contributions

• Quantification of Magnetoelastic Effects: The paper provides experimental evidence that strain effects in Co$_{80}$Ir$_{20}$ films are not negligible and can contribute anisotropy fields comparable to the intrinsic magnetocrystalline anisotropy.
• Underlayer Engineering: It demonstrates that the choice of seed layer (Ta vs. Pt) is a critical tool for tuning the strain state and, consequently, the magnetic anisotropy of CoIr films.
• Correction of Previous Models: The authors argue that previous studies estimating the intrinsic MCA of CoIr may have overestimated its magnitude by failing to subtract the magnetoelastic contribution.
• Microstructural-Magnetic Link: The study clarifies that while grain size and texture are similar across samples, the strain state is the primary differentiator for magnetic behavior, except in cases where texture degradation (as in CoIr-1) further modifies the response.

5. Significance

This work is significant for the design of high-frequency magnetic devices and spintronic applications.

• Device Optimization: Understanding the role of strain allows engineers to intentionally tune the resonance frequency and anisotropy of CoIr films by selecting specific underlayers, rather than relying solely on composition or thickness.
• Accurate Material Characterization: It highlights the necessity of separating strain-induced anisotropy from intrinsic magnetocrystalline anisotropy to obtain accurate material parameters for theoretical modeling.
• Strain Engineering: The findings support the use of strain engineering as a viable method to control magnetic properties in thin films with negative MCA, potentially leading to more efficient microwave oscillators and recording media.

In conclusion, the authors demonstrate that stress effects cannot be disregarded when analyzing CoIr magnetic data. The effective anisotropy is a complex sum of shape, magnetocrystalline, and magnetoelastic contributions, where the latter is heavily dependent on the substrate/underlayer interface.