Inverse proximity effect in thin-film superconductor/magnet heterostructures with metallic and insulating magnets

This paper demonstrates that while the effective homogeneous exchange field model accurately describes thin-film superconductor/insulating magnet heterostructures by producing well-defined spin splitting, it fails for superconductor/metallic magnet systems where the effect is chaotic, though the latter still support significant triplet correlations suitable for spintronics.

The Gist

Imagine you have a very special, quiet dance floor called a Superconductor. On this floor, electrons (the dancers) hold hands in perfect pairs and glide without any friction. This is the state of superconductivity.

Now, imagine you place a Magnet right next to this dance floor. The magnet is like a noisy, chaotic neighbor who wants to spin the dancers in opposite directions. This interaction is called the Proximity Effect.

For a long time, scientists thought that no matter what kind of magnet you used, the result would be simple and predictable: the magnet would just push the dancers apart in a uniform way, like a gentle, steady wind blowing across the whole dance floor. They called this the "Effective Model."

This paper says: "Not so fast!"

The authors, Bobkov and colleagues, discovered that the type of magnet matters a lot, and the old "simple wind" model only works for some magnets, not others. They looked at two main types of neighbors: Insulating Magnets (like a solid brick wall) and Metallic Magnets (like a chaotic, vibrating metal sheet).

Here is the breakdown of their discovery using simple analogies:

1. The "Brick Wall" Neighbor (Insulating Magnets)

When the superconductor touches an Insulating Magnet (like a Ferromagnetic Insulator), the interaction is clean and predictable.

• The Analogy: Imagine the magnet is a solid, silent brick wall. It doesn't let any of its own "noise" (electrons) leak into the dance floor. It just exerts a steady, uniform pressure.
• The Result: The dancers on the floor feel a smooth, consistent force. The "spin splitting" (the difference between dancers spinning left vs. right) is uniform across the whole floor.
• The Verdict: The old "Effective Model" works perfectly here. You can predict exactly what will happen.

2. The "Chaotic Metal Sheet" Neighbor (Metallic Magnets)

When the superconductor touches a Metallic Magnet (like a Ferromagnetic Metal), things get messy.

• The Analogy: Imagine the magnet is a sheet of metal that is vibrating wildly. It's not just pushing; it's leaking its own chaotic vibrations into the dance floor.
• The Result: Instead of a smooth wind, the dancers feel a chaotic, unpredictable storm. Some dancers get pushed hard to the left, their neighbors get pushed hard to the right, and the next ones get pushed gently.
• The Surprise: If you look at the average of the whole floor, it might look like nothing is happening. The "spin splitting" is so messy and chaotic that if you try to measure it with standard tools (looking at the density of states), you see a blurry mess, not a clear split.
• The Verdict: The old "Effective Model" fails completely here. You cannot describe this system as a simple, uniform wind. It is a chaotic patchwork.

3. The "Hidden Superpower" (Triplet Correlations)

Here is the most exciting part. Even though the Metallic Magnet creates a chaotic mess that looks like it's destroying the superconductivity (because the spin splitting is hidden), the dancers are actually doing something amazing.

• The Analogy: Even though the wind is chaotic, the dancers have secretly learned a new, complex dance move called the "Triplet Correlation." They are still holding hands, but now they are holding hands in a way that allows them to carry information (spin) over long distances.
• The Application: This is the holy grail for Superconducting Spintronics (using superconductors for computer memory and logic).
• The Proof: The authors showed that even in these chaotic Metallic Magnet systems, you can build a "Spin Valve" (a switch that controls the flow of spin). They built a sandwich of magnets and superconductor and showed that by flipping the magnets, they could turn the superconductivity on and off by 20%.
• The Lesson: Just because you can't see the spin splitting (because it's chaotic), doesn't mean the useful "Triplet" power isn't there. In fact, these chaotic systems are great for building new types of computer chips.

4. The "New Neighbor" (Altermagnets)

The paper also looked at a brand-new type of magnet called an Altermagnet.

• The Result: They found the exact same pattern!
  • If the Altermagnet is an Insulator, the effect is smooth and predictable (like the Brick Wall).
  • If the Altermagnet is a Metal, the effect is chaotic and unpredictable (like the Metal Sheet).

The Big Takeaway

For decades, scientists have used a simple "one-size-fits-all" model to describe how magnets affect superconductors. This paper says:

1. If your magnet is an insulator: The simple model works. You can predict everything.
2. If your magnet is a metal: The simple model is wrong. The effect is chaotic and unpredictable.
3. But don't panic! Even though the chaos makes it hard to see the effect, the useful "Triplet" superconductivity is still strong. These chaotic metal-superconductor sandwiches are actually better for building future spin-based computers than the smooth ones.

In short: Don't judge a book by its cover (or a magnet by its smoothness). The "messy" metallic magnets are hiding a powerful, chaotic superpower that could revolutionize our technology, even if we can't see it with our old measuring tools.

Technical Summary

1. Problem Statement

The paper addresses a fundamental question in superconducting spintronics: To what extent is the "effective model" applicable to thin-film superconductor/magnet (S/M) heterostructures?

• The Effective Model: Conventionally, S/M heterostructures are modeled as a homogeneous superconductor subjected to a uniform effective exchange field ($h_{eff}$). This model assumes that the proximity effect induces a clear, spatially uniform spin splitting of the electronic density of states (LDOS), analogous to a Zeeman splitting.
• The Discrepancy: While this model successfully describes S/Insulating Ferromagnet (S/FI) systems, experimental evidence for clear LDOS spin splitting in S/Metallic Ferromagnet (S/FM) systems is scarce or absent.
• The Core Question: Does the absence of observed LDOS splitting in S/FM systems imply the absence of triplet correlations (which are crucial for spintronics applications like spin valves), or does the effective model simply fail to describe the complex physics of metallic interfaces? The authors also extend this inquiry to Altermagnets (a recently discovered class of magnets with zero net magnetization but spin-split bands), comparing insulating (AI) and metallic (AM) variants.

2. Methodology

The authors employ a rigorous theoretical approach combining analytical derivations and numerical simulations:

• Model System: They study ballistic thin-film heterostructures consisting of $N_S$ layers of a superconductor (S) and $N_F$ layers of a magnetic material (Ferromagnet or Altermagnet).
• Hamiltonian: The system is described using a tight-binding Hamiltonian on a cubic lattice, incorporating:
  • Nearest-neighbor hopping ($t$) within layers and interlayer hopping ($t_{SF}$) at the interface.
  • On-site s-wave pairing ($\Delta$) in the S layer.
  • Exchange fields ($h$) in the magnetic layers. For Altermagnets, the exchange field is momentum-dependent ($h(\mathbf{k}) \propto k_x^2 - k_y^2$).
• Analytical Approach (Normal State): In the normal state ($\Delta=0$), they derive boundary conditions for wave functions to calculate the discrete momentum components ($k_S, k_F$) and the resulting effective exchange field $h_{eff}(n)$ for each spectral branch $n$.
• Numerical Approach (Superconducting State): They solve the Bogoliubov-de Gennes (BdG) equations self-consistently to determine:
  • The superconducting order parameter ($\Delta$).
  • The anomalous Green's function (to extract singlet and triplet correlations).
  • The Local Density of States (LDOS).
  • Tunneling conductance.
• Comparative Analysis: They systematically compare four cases:
  1. Superconductor/Insulating Ferromagnet (S/FI)
  2. Superconductor/Metallic Ferromagnet (S/FM)
  3. Superconductor/Insulating Altermagnet (S/AI)
  4. Superconductor/Metallic Altermagnet (S/AM)

3. Key Contributions and Results

A. Validity of the Effective Model

The study reveals a stark dichotomy between insulating and metallic magnets:

• Insulating Magnets (S/FI and S/AI):

  • The proximity effect creates a well-defined, smooth, and spatially uniform spin splitting of the electronic spectra in the S layer.
  • The effective exchange field $h_{eff}$ is directly proportional to the true exchange field of the magnet.
  • Conclusion: The effective model (homogeneous S in a uniform $h_{eff}$) is highly accurate for these systems. The LDOS shows clear Zeeman splitting of coherence peaks.

• Metallic Magnets (S/FM and S/AM):

  • The proximity effect creates a spin splitting that is chaotic, irregular, and highly sensitive to fine details (e.g., the exact number of atomic layers, interface parameters).
  • $h_{eff}(n)$ oscillates wildly between different spectral branches.
  • Conclusion: The effective model is invalid for these systems. The chaotic distribution of $h_{eff}$ smears out the spin splitting in the total LDOS, often resulting in a spectrum that looks like a standard BCS superconductor (or a gapless state) without clear peak splitting, even though significant spin splitting exists at the branch level.

B. Triplet Correlations vs. LDOS Splitting

A critical finding is the decoupling of LDOS splitting from triplet correlations:

• In S/FM and S/AM systems, the LDOS may show no observable spin splitting of coherence peaks.
• However, robust triplet correlations are still generated and are comparable in magnitude to those in S/FI systems.
• Implication: The absence of LDOS splitting is not a criterion for the absence of triplet correlations. Relying solely on LDOS splitting to detect triplet states in metallic heterostructures leads to false negatives.

C. Spin-Valve Effect in Metallic Systems

The authors demonstrate that despite the lack of LDOS splitting, S/FM heterostructures are functional for spintronics:

• They simulated an FM/S/FM trilayer and calculated the critical temperature ($T_c$) as a function of the angle $\theta$ between the magnetic moments.
• Result: A significant spin-valve effect (~20%) was observed ($T_c(\theta=\pi) > T_c(\theta=0)$).
• Mechanism: While the local $h_{eff}$ is chaotic, the spatially averaged effective field depends on the relative orientation of the ferromagnetic layers. This average field is sufficient to modulate superconductivity, proving that these systems are viable for spintronic applications even without clear LDOS signatures.

D. Role of Impurities

The paper discusses the impact of disorder:

• In ballistic metallic systems, the chaotic $h_{eff}$ prevents clear splitting.
• In the diffusive limit (strong impurity scattering), electrons scatter between different spectral branches, effectively averaging the chaotic $h_{eff}$.
• Result: Impurities can actually restore the applicability of the effective model (smoothing out the chaos) and potentially enhance superconductivity by reducing the average pair-breaking effect of the chaotic field.

4. Significance and Impact

1. Correction of Theoretical Frameworks: The paper challenges the widespread assumption that the effective homogeneous exchange field model is universally applicable to S/M heterostructures. It establishes that this model fails for metallic interfaces due to quantum interference effects at the atomic scale.
2. Experimental Guidance: It explains why experimentalists often fail to observe LDOS spin splitting in S/FM systems despite the presence of strong proximity effects. It advises against using the absence of LDOS splitting as a proxy for the absence of triplet correlations.
3. Spintronics Applications: It validates the use of metallic S/FM and S/AM heterostructures for spintronic devices (like spin valves and thermoelectric generators), confirming that triplet correlations persist and are functional even when the spectral signature is "hidden."
4. Altermagnet Physics: It extends these findings to the emerging field of altermagnets, showing that while insulating altermagnets behave predictably, metallic altermagnets exhibit the same chaotic proximity effects as metallic ferromagnets.

In summary, the authors demonstrate that metallic superconductor/magnet interfaces are inherently complex quantum systems where the "effective field" is a chaotic function of momentum and layer index. While this complexity obscures the spin splitting in the density of states, it does not hinder the generation of the triplet correlations necessary for advanced superconducting spintronics.