Long range proximity effects in planar structures involving the halfmetal ferromagnet La0.7Sr0.3MnO3 and Pt interlayers

This study investigates long-range triplet supercurrent transport in planar La0.7Sr0.3MnO3 Josephson junctions, revealing that while critical current systematics are hindered by fabrication inconsistencies, the introduction of a Pt interlayer successfully enables zero-resistance states at electrode distances up to 2 μm, suggesting the feasibility of even longer-range transport.

The Gist

Imagine electricity usually flowing like a chaotic crowd of people, where everyone is moving in different directions and spinning randomly. But in a special material called a superconductor, electricity flows like a perfectly synchronized dance troupe. Every dancer holds hands with a partner, moving in perfect unison without any friction or resistance. These dancing pairs are called "Cooper pairs."

Usually, these pairs are made of two dancers with opposite spins (one spinning left, one spinning right). However, if you try to send this dance troupe through a magnet (which acts like a strict bouncer that only lets in dancers spinning one specific way), the pairs break up, and the dance stops.

The Problem: The "Half-Metal" Bouncer

The scientists in this paper were working with a special type of magnet called a half-metal (specifically a material called LSMO). Think of this half-metal as a bouncer who is extremely picky: it only allows dancers spinning "up" to enter. It completely blocks dancers spinning "down."

If you try to send the standard superconducting dance troupe (mixed up/down spins) into this half-metal, the "down" dancers get kicked out immediately, and the whole dance collapses. The supercurrent stops.

The Goal: Teaching the Dancers to Spin Together

The researchers wanted to see if they could trick the system. They wanted to convert the standard pairs into a new type of pair where both dancers spin the same way (both "up"). If they could do this, the half-metal bouncer would let them both in, and the supercurrent could travel a long distance through the magnet. This is called the "long-range proximity effect."

They built tiny bridges (nanostrips) of this half-metal and tried to connect them with superconducting contacts.

Experiment 1: The Rough Bridge (LSMO/NbTi)

First, they tried building these bridges by placing the superconductor (NbTi) directly on top of the half-metal (LSMO).

• The Result: It worked! They saw strong supercurrents traveling across the bridge, even when the bridge was quite long (up to 1.6 micrometers, which is huge for this scale).
• The Problem: The results were inconsistent. Sometimes the current was huge; sometimes it was tiny. It was like trying to build a bridge where the quality of the cement changed randomly every time they mixed a batch. They suspected the "glue" (the interface) between the two materials was messy and unpredictable, creating the necessary "spin-mixing" by accident rather than by design.

Experiment 2: The Smooth Interlayer (Adding Platinum)

To fix the inconsistency, they decided to insert a buffer layer between the superconductor and the half-metal. They chose Platinum (Pt).

• The Analogy: Imagine the half-metal is a rough, uneven floor. The superconductor is a delicate glass table. If you put the table directly on the floor, it wobbles and breaks. But if you put a perfectly smooth, flat sheet of plywood (Platinum) in between, the table sits perfectly steady.
• The Science: They found that Platinum spreads out perfectly flat over the half-metal (it "wets" the surface), unlike their previous attempt with Silver, which formed bumpy islands.

The Big Discovery

When they built these new "sandwich" structures (Half-Metal / Platinum / Superconductor) and placed the contacts on top of a full sheet of the half-metal:

1. Superconductivity returned: They saw the supercurrent flow again.
2. Long Distance: They successfully sent the supercurrent across a gap of 2 micrometers. This is a significant distance for this type of physics.
3. The Mechanism: The fact that it worked even without the messy direct contact between the superconductor and the half-metal suggests that the Platinum layer itself helps create the special "same-spin" pairs. The scientists suspect this is due to a quantum effect called Spin-Orbit Coupling (a fancy way of saying the electrons interact with the heavy Platinum atoms in a way that flips their spins just right).

The Conclusion

The paper concludes that while the direct contact between the superconductor and the half-metal can work, it is messy and hard to control. However, inserting a thin layer of Platinum creates a clean, smooth interface that reliably generates these special supercurrents.

In simple terms: The researchers found a way to build a reliable "highway" for supercurrents through a magnetic material by adding a smooth "platinum lane" that helps the electrons change their spin and keep dancing together, even over long distances. This proves that we can control these quantum effects better than before, though the paper stops short of saying exactly how this will be used in real-world technology yet.

Technical Summary

1. Problem Statement

The research addresses the challenge of generating and transporting spin-polarized supercurrents over long distances in Josephson junctions containing a half-metallic ferromagnet (HMF) as the weak link.

• The Physics: Conventional superconductors carry spinless Cooper pairs (singlets). To transport supercurrents through a ferromagnet, these must convert into equal-spin triplet pairs, which are immune to pair-breaking by the exchange field of the ferromagnet.
• The Material: The study focuses on La0.7Sr0.3MnO3 (LSMO), a half-metallic ferromagnet where only one spin band exists.
• The Issue: Previous work by the authors demonstrated long-range proximity (LRP) effects in planar LSMO/NbTi junctions with critical current densities ($J_c$) up to $1.1 \times 10^9$ A/m². However, the fabrication process lacked reproducibility. The critical current did not show a simple, monotonic dependence on the junction length ($L$), suggesting uncontrolled variables at the LSMO/NbTi interface, where the triplet generation mechanism (likely magnetic disorder) resides.

2. Methodology

The authors extended their research in two distinct directions to isolate the triplet generation mechanism:

A. Long Junction (LJ) Devices (LSMO/NbTi)

• Structure: Planar nanostrips of LSMO (40 nm) connected by NbTi (60 nm) contacts.
• Fabrication: Ion-etching using a Pt hard mask.
• Variables: Junction lengths ($L$) ranging from 0.47 µm to 1.58 µm and widths of 2 µm and 5 µm.
• Goal: To map the dependence of $J_c$ on $L$ and verify if the lack of systematics in previous batches was due to interface disorder.

B. Full Film (FF) Devices (LSMO/Pt/NbTi)

• Structure: A trilayer stack of LSMO (40 nm) / Pt (7 nm) / NbTi (70 nm) fabricated on a full LSMO film.
• Rationale: To separate the superconductor (NbTi) from the HMF (LSMO) using a non-reactive normal metal (Pt). This eliminates the direct LSMO/NbTi interface to test if triplet generation can occur without it.
• Fabrication: Ar-ion etching to define contact structures on the full film.
• Characterization:
  • Transport: Resistance vs. Temperature ($R(T)$), Current-Voltage ($I-V$) characteristics, and Superconducting Quantum Interference (SQI) patterns under Out-of-Plane (OOP) and In-Plane (IP) magnetic fields.
  • Microscopy: Atomic Force Microscopy (AFM) to check surface wetting/roughness; Aberration-corrected Scanning Transmission Electron Microscopy (STEM) and Electron Energy Loss Spectroscopy (EELS) to analyze the atomic structure and chemical sharpness of the LSMO/Pt interface.

3. Key Results

A. LJ Devices (LSMO/NbTi)

• Supercurrents: Strong supercurrents were observed, with $J_c$ at 2 K reaching $\sim 10^9$ A/m² even for lengths up to 1.58 µm.
• Lack of Systematics: No simple monotonic relationship between $J_c$ and $L$ was found. For example, a shorter junction (LJ7, $L=1.1$ µm) had half the $J_c$ of a longer one (LJ12, $L=1.58$ µm).
• Conclusion: The triplet generation at the direct LSMO/NbTi interface is highly sensitive to fabrication variations (likely magnetic disorder or Mn valence fluctuations), making it difficult to control.

B. FF Devices (LSMO/Pt/NbTi)

• Long-Range Transport: Devices with electrode separations up to 2.1 µm (FF11) achieved a zero-resistance state.
• Transition Behavior: The devices showed sharp superconducting transitions. While the resistance did not drop immediately to zero upon contact superconductivity (showing a plateau), it eventually reached zero between 4 K and 5 K.
• SQI Patterns: The devices exhibited sharp, non-Gaussian interference patterns, indicating that phase coherence is maintained over the long distance and that dephasing is not fully random.
• Interface Quality:
  • AFM: The LSMO/Pt interface was atomically flat (roughness < 2 nm) and fully wetting, unlike previous attempts with Ag which formed islands.
  • STEM-EELS: Confirmed a chemically sharp, abrupt transition between the perovskite LSMO and polycrystalline Pt, with no interfacial mixing or secondary phases.

C. Mechanism Insight

• The fact that FF devices (with a Pt spacer) still generate long-range triplets suggests the direct LSMO/NbTi interface is not the sole or necessary source of triplet generation.
• The authors propose that Spin-Orbit Coupling (SOC), specifically Rashba-type SOC at the Pt/LSMO interface, facilitates the singlet-to-triplet conversion. This aligns with theoretical predictions that SOC at heavy metal/ferromagnet interfaces can generate long-range triplets in planar geometries.

4. Key Contributions

1. Demonstration of Reproducible Long-Range Transport: Successfully demonstrated zero-resistance states in planar HMF structures over distances exceeding 2 µm using a Pt interlayer.
2. Interface Engineering: Identified that inserting a Pt interlayer solves the wetting and chemical mixing issues associated with direct NbTi/LSMO contacts, providing a more controlled platform for device fabrication.
3. Mechanism Hypothesis: Provided strong evidence that triplet generation in these systems can be driven by SOC at the heavy metal/oxide interface (Pt/LSMO) rather than solely relying on magnetic disorder at the superconductor/ferromagnet interface.
4. Material Characterization: Provided high-resolution structural and chemical proof (STEM-EELS) of the sharp, coherent nature of the LSMO/Pt interface.

5. Significance

• Spintronics: This work advances the feasibility of spin-polarized supercurrents for superconducting spintronics, a field aiming to combine superconductivity with magnetic memory and logic.
• Device Control: By moving away from the uncontrolled LSMO/NbTi interface to the stable LSMO/Pt/NbTi stack, the authors have established a more reliable fabrication route for long-range Josephson junctions.
• Fundamental Physics: The results support the theoretical model that Rashba spin-orbit coupling at the interface of a heavy metal (Pt) and a ferromagnet can generate long-range triplet correlations, offering a new pathway to engineer triplet superconductivity without relying on magnetic disorder.

In conclusion, the paper establishes that Pt interlayers are highly effective for interfacing oxide ferromagnets with conventional superconductors, enabling the controlled fabrication of planar Josephson junctions capable of transporting fully spin-polarized supercurrents over micrometer scales.