🔬 mesoscale physics

Loss Mechanisms in High-coherence Multimode Mechanical Resonators Coupled to Superconducting Circuits

This paper demonstrates that optimizing the defect density and interfaces of piezoelectric films in high-overtone bulk acoustic-wave resonators enables phonon lifetimes up to 400 μs and a record hybrid quantum coherence cooperativity of $1.1\times10^5$ , establishing a new milestone for circuit quantum acoustodynamics devices.

Original authors: Raquel Garcia Belles, Alexander Anferov, Lukas F. Deeg, Loris Colicchio, Arianne Brooks, Tom Schatteburg, Maxwell Drimmer, Ines C. Rodrigues, Rodrigo Benevides, Marco Liffredo, Jyotish Patidar, Oleksa

Published 2026-02-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Raquel Garcia Belles, Alexander Anferov, Lukas F. Deeg, Loris Colicchio, Arianne Brooks, Tom Schatteburg, Maxwell Drimmer, Ines C. Rodrigues, Rodrigo Benevides, Marco Liffredo, Jyotish Patidar, Oleksandr Pshyk, Matteo Fadel, Luis Guillermo Villanueva, Sebastian Siol, Gerhard Kirchmair, Yiwen Chu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to build a super-precise musical instrument, but instead of strings, it's made of sound waves trapped inside a tiny crystal. This instrument is so sensitive that it can hold a single "note" (a single particle of sound called a phonon) for a long time without the note fading away.

In the world of quantum computing, these sound waves are like memory banks. They can store information for computers that are faster and more powerful than anything we have today. But there's a catch: for these memory banks to work, the sound waves must be incredibly pure and quiet. If the sound gets "scratched" or "muffled" by the material it travels through, the memory is lost.

This paper is a report card on how the researchers built the best possible "sound crystal" they could, and why some of their attempts worked better than others.

The Main Characters

The Crystal (Sapphire): Think of this as a giant, perfect, clear bell. It's the main body where the sound travels. It's very high quality, like a diamond.
The Translator (Aluminum Nitride Film): To talk to the quantum computer (which uses electricity), the sound needs a translator. The researchers glued a very thin, special film onto the crystal. This film turns electrical signals into sound and vice versa.
The Quantum Computer (Superconducting Qubit): This is the brain. It needs to "listen" to the sound crystal to read the memory.

The Problem: The "Translator" Ruined the Music

The researchers found that while the giant crystal bell was perfect, the thin film they glued on top was the problem. It was like gluing a piece of sandpaper onto a violin string.

The "Sandpaper" Effect: The film wasn't perfectly smooth. It had tiny bumps and defects. When the sound wave hit these bumps, it scattered, like a ball bouncing off a rocky wall instead of a smooth floor. This caused the sound to lose energy (dissipation).
The "Ghost" Defects: Inside the film and where it meets the crystal, there were tiny "ghosts" called Two-Level Systems (TLS). Imagine these as tiny, invisible bugs that get excited by the sound, steal a little bit of energy, and then release it randomly. This makes the sound wobble and lose its clarity (dephasing).

The Experiment: Testing Different "Glues"

The team tried three different ways to grow this thin film, like testing three different brands of glue:

The "Spray Paint" Method (Sputtering): They sprayed the film on. It was okay, but the film was a bit grainy (like sandpaper). The sound got a bit muddy.
The "Crystal Garden" Method (HVPE): They grew the film like a crystal garden. This created a much smoother, higher-quality film. However, the process of growing it damaged the surface of the giant crystal underneath. It was like having a perfect garden, but the soil underneath was churned up and full of rocks. The sound got lost in those rocks at the interface.
The "High-Power" Method (HiPIMS): This was a new, high-tech way of spraying. It was almost as good as the crystal garden method but didn't damage the soil underneath as much.

The Big Breakthrough

By carefully analyzing the sound waves, the researchers realized that the interface (the boundary where the film meets the crystal) was the biggest culprit.

The Discovery: They found that if they could make the film grow perfectly without damaging the crystal underneath, the sound waves could travel for an incredibly long time.
The Result: They managed to trap a single sound note for 400 microseconds. In the quantum world, that's an eternity! It's like holding a single note on a piano for a full second without it fading, when usually it would vanish in a blink.

Why This Matters

This is a huge deal for the future of technology.

The "Cooperativity" Score: The researchers calculated a score called "cooperativity." Think of this as how well the brain (qubit) and the memory (sound crystal) can talk to each other without getting distracted by noise. Their score was 100,000 times better than previous attempts.
The Future: This proves that we can build quantum computers that use sound to store information. It's like upgrading from a scratchy cassette tape to a pristine, high-definition digital file.

The Takeaway

The paper teaches us that in the quantum world, smoothness is everything. Even a microscopic scratch or a tiny defect at the boundary between two materials can ruin the performance of a super-advanced computer. By figuring out how to grow these materials perfectly, the researchers have cleared the path for building quantum computers that are stable, fast, and capable of solving problems we can't even imagine yet.

In short: They built a perfect sound chamber, figured out why the walls were leaking sound, fixed the leaks, and now they have a room where a whisper can echo for a very, very long time.

1. Problem Statement

Circuit Quantum Acoustodynamics (cQAD) systems, which couple superconducting qubits to mechanical resonators, are promising platforms for quantum memory, transduction, and fundamental physics tests. However, achieving strong coupling and long coherence times in the quantum regime (single-phonon level) is hindered by various dissipation channels.

The Challenge: While High-Overtone Bulk Acoustic Wave Resonators (HBARs) offer high quality factors ( $Q$ ) and stable frequencies, coupling them to superconducting circuits requires piezoelectric films (e.g., Aluminum Nitride, AlN). The introduction of these films and their interfaces with the bulk substrate often introduces additional loss mechanisms, such as surface scattering, mechanical absorption, and Two-Level Systems (TLS).
The Gap: There was a lack of systematic understanding regarding how the specific growth method of the piezoelectric film and the quality of the film-substrate interface limit the quantum coherence of HBARs, particularly in the low-temperature, low-phonon-number regime.

2. Methodology

The authors investigated composite AlN-sapphire HBARs using a combination of fabrication, spectroscopy, and quantum measurement techniques:

Device Fabrication:
- Substrate: High-quality sapphire crystals ( $\sim$ 450 $\mu$ m thick).
- Piezoelectric Layer: AlN films ( $\sim$ $\sim$ 1 $\mu$ $μ$ m) grown via three different techniques to vary crystallinity and defect density:
  1. Pulsed DC sputtering (Sample A).
  2. Hydride Vapor-Phase Epitaxy (HVPE) (Samples B, C, D).
  3. High-Power Impulse Magnetron Sputtering (HiPIMS) (Sample E).
- Geometry: A "dome-shaped" transducer was created via lithography and dry etching to support Laguerre-Gaussian (LG) acoustic modes, suppressing diffraction and clamping losses.
- Coupling: Devices were flip-chip bonded to either a flux-tunable superconducting transmon qubit or a 50 $\Omega$ coplanar waveguide (CPW) antenna.
Measurement Techniques:
- Microwave Spectroscopy: Used a Vector Network Analyzer (VNA) to drive modes at high phonon occupancies ( $\bar{n}_p \sim 10^6 - 10^{10}$ ) to measure internal quality factors ( $Q_i$ ) and identify frequency-dependent loss trends.
- Quantum Measurements: Used a superconducting qubit to swap single excitations into the HBAR modes to measure energy relaxation times ( $T_1$ ) and coherence times ( $T_2^*$ ) in the single-phonon regime.
- Material Characterization: Employed X-ray Diffraction (XRD), Scanning Transmission Electron Microscopy (STEM), and Atomic Force Microscopy (AFM) to analyze crystallinity, interface defects, and surface roughness.
Theoretical Modeling:
- Developed a Multilayer Participation Model treating the HBAR as a stack of layers (Piezo, Defect, Bulk).
- Calculated energy participation ratios ( $p_X$ ) for kinetic and potential energy in each layer.
- Modeled total dissipation as a weighted sum of mechanical absorption, surface scattering, and TLS losses.

3. Key Contributions

Identification of Limiting Factors: The study definitively identifies that the defect density of the piezoelectric material and the quality of the piezo-bulk interface are the primary limiting factors for coherence, rather than the bulk sapphire itself.
Interface Engineering: Revealed that HVPE-grown AlN, while having superior bulk crystallinity, suffers from a damaged interface layer ( $\sim$ 10 nm deep) caused by surface treatment, which introduces significant scattering and TLS loss.
TLS Characterization: Distinguished between reactive TLS loss (measured via frequency shifts) and dissipative TLS loss (measured via power-dependent $Q$ saturation), showing that potential energy participation is the correct metric for modeling TLS coupling in these multimode systems.
Record Performance: Demonstrated that optimizing the growth method and interface can yield HBARs with coherence times approaching the lifetime limit, significantly outperforming previous cQAD platforms.

4. Key Results

Quality Factors and Lifetimes:
- Measured single-phonon lifetimes up to $T_1 = 396 \pm 11$ $\mu$ s.
- Achieved lifetime-limited coherence times of $T_2^* = 806 \pm 24$ $\mu$ s (approaching $2T_1$ ), indicating negligible dephasing.
- Internal quality factors reached $Q_i \approx 1.3 \times 10^7$ in the quantum regime.
Loss Mechanism Analysis:
- Sputtered AlN (Sample A): Loss dominated by mechanical absorption within the polycrystalline AlN layer ( $Q_{mech} \sim 2 \times 10^4$ ), with a smooth interface.
- HVPE AlN (Samples B/C): Despite higher bulk crystallinity, these samples suffered from a high-loss interface layer ( $Q_{D,mech} \sim 10^2$ ) and strong surface scattering, causing frequency-dependent oscillations in $Q_i$ .
- HiPIMS AlN (Sample E): Showed performance comparable to sputtered AlN, with low TLS density.
Cooperativity:
- Achieved a single-quantum coherence cooperativity of $C_{T2} = 1.13 \times 10^5$ .
- This value is more than one order of magnitude higher than previous cQAD records, establishing HBARs as the leading platform for high-coherence hybrid quantum systems.
Frequency Stability: The HBAR modes exhibited exceptional frequency stability with negligible dephasing, attributed to the large mode volume averaging out microscopic fluctuators and the shielding of the bulk mode from surface noise.

5. Significance

Milestone for cQAD: This work sets a new benchmark for the performance of circuit quantum acoustodynamic devices, proving that HBARs can maintain quantum coherence for nearly a millisecond.
Pathway to Scalability: By demonstrating that the limiting factors are controllable (via growth techniques and interface engineering), the paper provides a concrete roadmap for improving quantum acoustic devices.
Material Science Insights: The study offers a robust framework (the participation model) for assessing the quality of thin films and interfaces in acoustic devices, which is critical for the development of quantum transducers and memories.
Future Applications: The high coherence times and strong coupling enable new applications in quantum networking (transduction), precision sensing, and tests of macroscopic quantum mechanics. The authors suggest that further optimization (e.g., lower frequencies to reduce surface scattering) could push phonon lifetimes beyond 1 ms.