High temporal stability of niobium superconducting resonators by surface passivation with organophosphonate self-assembled monolayers

This study demonstrates that coating niobium superconducting resonators with alkyl-phosphonate self-assembled monolayers effectively suppresses native oxide regrowth, thereby maintaining high quality factors and temporal stability over six days of air exposure while significantly reducing two-level system losses.

The Gist

Imagine you are trying to build a super-precise, ultra-fast clock that runs on electricity. This clock is so sensitive that if even a single speck of dust lands on it, the clock starts to lose time. In the world of quantum computing, these "clocks" are called superconducting resonators, and they are the heart of quantum computers.

The problem is that these clocks are made of a metal called Niobium. When you take a fresh piece of Niobium out of a vacuum and expose it to the air, it immediately starts to "rust" (form an oxide layer). Think of this like a shiny new apple turning brown the moment you cut it.

This "rust" is full of tiny, invisible defects called Two-Level Systems (TLS). You can think of these TLSs as tiny, mischievous gremlins living in the rust. When the quantum clock tries to tick, these gremlins steal energy, causing the clock to lose its rhythm (coherence) and stop working properly.

The Problem: The "Rust" Returns

Scientists tried to fix this by washing the Niobium with a special acid (called BOE) to strip away the rust. This worked great for a few hours. But as soon as the clean metal touched the air again, the rust started growing back, and the gremlins returned. Within a few days, the clock was back to being slow and unreliable.

The Solution: A "Molecular Raincoat"

In this paper, the researchers from the Technical University of Munich came up with a clever idea: instead of just cleaning the metal, let's put a molecular raincoat on it to stop the rust from ever coming back.

They used a special type of molecule called an organophosphonate. Imagine these molecules as tiny umbrellas.

1. The Handle: One end of the molecule is a "handle" that grabs onto the Niobium metal very tightly.
2. The Canopy: The other end is a long, waxy tail that points outward.

When they dip the cleaned Niobium into a solution of these molecules, the handles grab the metal, and the tails stand up in a neat, organized row, forming a Self-Assembled Monolayer (SAM). It's like a microscopic forest of umbrellas standing shoulder-to-shoulder.

How It Works (The Analogy)

• The Unpassivated Clock (No Raincoat): Without the raincoat, the air (specifically oxygen and water) touches the metal. The metal oxidizes (rusts), and the "gremlins" (TLSs) multiply. The clock's performance gets worse every day.
• The Passivated Clock (With Raincoat): The SAM raincoat is hydrophobic (water-repelling). It's like a super-slick surface that water and oxygen just slide right off of. Because the air can't touch the metal, the rust can't grow. The gremlins can't get in.

The Results: A Clock That Keeps Time

The researchers tested their clocks over six days:

• The Unpassivated Clocks: Their performance dropped by about 80%. The "gremlins" took over, and the clock became very noisy.
• The Passivated Clocks: Their performance stayed exactly the same. The "raincoat" worked perfectly. The clock remained stable, quiet, and precise, even after sitting in the air for days.

Why This Matters

This is a big deal for the future of quantum computers.

1. Stability: Quantum computers need to stay stable for a long time to do complex calculations. If the components degrade in a few days, you can't build a reliable machine.
2. Scalability: This method uses a simple chemical dip (like dipping a cookie in chocolate) rather than expensive, complex machinery. This means we could potentially coat thousands of quantum chips at once, making it easier to build large-scale quantum computers.

The "Gremlin" Count

The scientists also did some detective work to count exactly how many "gremlins" were in the new raincoat. They found that the raincoat itself (the SAM) has very few gremlins—so few that it's almost as good as the best materials we already use. This proves that the raincoat doesn't just stop the rust; it doesn't introduce new problems of its own.

In short: The researchers found a way to give superconducting quantum circuits a "molecular raincoat" that stops them from rusting in the air. This keeps the quantum computers running smoothly and stably, paving the way for bigger, better, and more reliable quantum machines in the future.

Technical Summary

1. Problem Statement

Superconducting qubits, essential for large-scale quantum computing, suffer from limited coherence times primarily due to energy losses caused by Two-Level Systems (TLS). A major source of these TLS defects is the native oxide (specifically $NbO_x$) that forms at the metal-air interface of niobium (Nb) superconducting circuits.

• Current Limitations: While native oxides can be removed via wet chemical etching (e.g., Buffered Oxide Etch, BOE), the Nb surface is highly reactive. Upon exposure to air, the oxide rapidly regrows, leading to a degradation of the internal quality factor ($Q_i$) and resonant frequency stability over time.
• The Gap: Existing passivation methods (e.g., capping with Ta or Au) or previous organic coating attempts (e.g., silanes) have limitations. Silanes often form less defined interfaces due to lateral cross-linking, potentially leaving sites prone to early re-oxidation. There is a need for a robust, scalable passivation strategy that prevents oxide regrowth while introducing minimal additional dielectric loss.

2. Methodology

The authors developed a surface passivation protocol using organophosphonate Self-Assembled Monolayers (SAMs) and evaluated their performance through a combination of surface characterization and cryogenic microwave measurements.

• Sample Preparation:
  • Substrate: 150 nm Nb thin films on high-resistivity Si(100) substrates.
  • Cleaning: BOE etching (20 min) to remove native oxides.
  • Passivation: Immersion in a 1 mM solution of decylphosphonic acid ($C_{10}H_{23}O_3P$) in anhydrous toluene for 48 hours to grow the SAM.
• Surface Characterization:
  • Contact Angle (CA): Measured wettability to assess hydrophobicity and oxide regrowth resistance.
  • Ellipsometry & AFM: Determined SAM thickness (~1.2 nm) and surface roughness.
  • FTIR: Verified molecular ordering via C-H stretching vibrations.
  • X-ray Photoelectron Spectroscopy (XPS): Analyzed chemical composition, oxidation states ($Nb^0, Nb^{2+}, Nb^{4+}, Nb^{5+}$), and oxide thickness after aging.
• Cryogenic Measurements:
  • Setup: Coplanar waveguide (CPW) resonators measured at ~10 mK in a dilution refrigerator.
  • Protocol: $Q_i$ and resonant frequency ($f_r$) were measured over 6 days of air exposure for both un-passivated (BOE etched only) and SAM-passivated samples.
  • Analysis: Data was fitted using a two-component TLS loss model to distinguish between different loss channels based on their critical photon numbers ($n_c$) and saturation behaviors.

3. Key Contributions

• Novel Passivation Material: The study introduces organophosphonates (specifically decylphosphonic acid) as a superior alternative to silanes for Nb passivation, citing better surface binding and reduced lateral cross-linking issues.
• Quantification of SAM Loss: For the first time, the authors quantified the intrinsic dielectric loss of a phosphonate SAM in the GHz regime, estimating it to be $\delta_{SAM} \approx 5 \times 10^{-7}$.
• Mechanistic Insight: By employing a multi-TLS fitting model, the study successfully deconvoluted distinct loss channels:
  • Channel 1 ($n \sim 100$): Intrinsic losses (substrate, patterning, metal-substrate interface).
  • Channel 2 ($n \sim 10^5$): Regrown $Nb_2O_5$ (dominant in un-passivated samples).
  • Channel 3 ($n \sim 10^7$): The SAM layer itself (dominant in passivated samples).

4. Key Results

A. Surface Stability

• Hydrophobicity: Un-passivated Nb had a contact angle of ~42° (hydrophilic), while SAM-passivated Nb reached ~103° (hydrophobic). The SAM-passivated surface maintained this high contact angle for 14 days, indicating effective suppression of water-induced oxidation.
• Oxide Thickness (XPS): After 6 days of aging:
  • Un-passivated: Total oxide thickness grew to ~3.8 nm.
  • Passivated: Total oxide thickness remained at ~1.2 nm (comparable to the SAM thickness), confirming the SAM effectively blocked further oxide regrowth.

B. Microwave Performance (Temporal Stability)

• Quality Factor ($Q_i$) Degradation:
  • Un-passivated: $Q_i$ degraded significantly, with a ~80% increase in loss at single-photon power levels after 6 days. The loss associated with the regrown oxide channel ($\delta_{Nb_2O_5}$) increased by 135%.
  • Passivated: $Q_i$ remained temporally stable. The loss associated with the SAM channel ($\delta_{SAM}$) showed no significant change over 6 days.
• Resonant Frequency Stability:
  • Un-passivated resonators exhibited a large frequency shift (~400 kHz) after 6 days due to the changing dielectric environment from oxide growth.
  • Passivated resonators showed a significantly smaller shift (~300 kHz), which was comparable to the shift seen in un-passivated samples after only 2 days.

C. TLS Model Analysis

• The un-passivated samples showed a loss channel with a critical photon number $n_2 \approx 10^5$, attributed to regrown $Nb_2O_5$.
• The passivated samples replaced this channel with a new one having $n_3 \approx 10^7$, attributed to the SAM.
• The intrinsic loss channel ($n_1 \approx 100$) remained relatively stable in both, though slightly higher in passivated samples (possibly due to solvent residues or processing), but this was outweighed by the elimination of the rapidly degrading oxide channel.

5. Significance and Conclusion

This work demonstrates that organophosphonate SAMs provide a highly effective barrier against the regrowth of native niobium oxides, a primary source of decoherence in superconducting circuits.

• Industrial Relevance: The method offers a scalable, solution-based passivation route suitable for industrial-scale qubit fabrication, addressing the critical need for long-term device stability.
• Scientific Impact: The study provides the first experimental quantification of the dielectric loss of a specific SAM in the microwave regime, proving that the SAM's intrinsic loss is low enough to be a viable alternative to oxide regrowth.
• Future Outlook: While the absolute $Q_i$ of passivated resonators was slightly lower than fresh un-passivated ones (likely due to processing steps), the temporal stability is vastly superior. Future work will focus on optimizing SAM growth to minimize intrinsic losses and completely eliminate interfacial oxides.

In summary, the paper establishes that suppressing oxide regrowth via phosphonate SAMs is a critical strategy for maintaining the high coherence times required for large-scale, reliable quantum computing architectures.