Oxide-nitride heteroepitaxy for low-loss dielectrics in superconducting quantum circuits
David A. Garcia-Wetten, Mitchell J. Walker, Peter G. Lim, André Vallières, Maria G. Jimenez-Guillermo, Miguel A. Alvarado, Dominic P. Goronzy, Anna Grassellino
Abstract
Superconducting qubits show great promise for the realization of fault-tolerant quantum computing, but lossy, amorphous dielectrics limit current technology. Identifying highly crystalline and stoichiometric dielectrics with intrinsically low microwave loss is therefore a central materials challenge, yet experimentally validated platforms remain scarce. In this work, we integrate a crystalline dielectric into a heteroepitaxial TiN/-AlO/TiN trilayer grown via pulsed laser deposition. Correlative high-resolution imaging, diffraction, and spectroscopy measurements confirm the single-crystal quality and chemical integrity of all layers, with minimal defects and limited anion interdiffusion across the oxide-nitride interfaces. Using microwave lumped-element resonators with parallel-plate capacitors, we report the first direct measurement of the dielectric loss of epitaxial -AlO, for which we find a low intrinsic two-level system loss, . These results establish heteroepitaxial oxides on transition metal nitrides as an attractive materials platform for superconducting quantum circuits, particularly for integration into compact device architectures such as merged-element transmons and microwave kinetic inductance detectors.
AI Impact Assessments
(3 models)Scientific Impact Assessment
1. Core Contribution
This paper addresses a fundamental materials bottleneck in superconducting quantum circuits: the lossy, amorphous dielectrics (primarily AlOx) that host parasitic two-level systems (TLS) and limit qubit coherence. The authors demonstrate heteroepitaxial TiN/γ-Al₂O₃/TiN trilayers grown by pulsed laser deposition (PLD) on sapphire substrates, achieving single-crystal quality throughout the stack. The central result is the first direct measurement of the intrinsic TLS loss of epitaxial γ-Al₂O₃, yielding δ⁰_TLS = (2.8 ± 0.1) × 10⁻⁵ — approximately two orders of magnitude lower than conventional amorphous AlOx used in Josephson junctions.
The novelty lies in the specific combination of material system (TiN as a diffusion-resistant superconductor paired with crystalline γ-Al₂O₃), growth method (all layers deposited in the same PLD system without vacuum break), and rigorous loss quantification using purpose-designed lumped-element parallel-plate capacitor (LEPPC) resonators. While prior work explored epitaxial Al₂O₃ with Re electrodes or AlN-based all-nitride stacks, TiN offers superior superconducting properties for qubit applications and inherent resistance to oxygen interdiffusion, making this a more practically viable platform.
2. Methodological Rigor
The characterization campaign is exceptionally thorough and multi-modal. The authors employ a correlative suite including XRR, HRXRD, RHEED, ToF-SIMS, ABF-STEM, HAADF-STEM, 4D-STEM, EDS, EELS, and XPS to establish crystallinity, epitaxy, stoichiometry, and interface chemistry. Key findings are cross-validated: layer thicknesses from XRR agree with STEM; the γ-Al₂O₃ phase identification is confirmed by both XRD superlattice peaks and EELS O-K edge fine structure; oxygen content is tracked by EDS, ToF-SIMS, and XPS depth profiling.
The dielectric loss measurement methodology is carefully designed. The LEPPC geometry with filling factor >0.99 ensures that measured losses are attributable to the capacitor dielectric rather than parasitic elements. The inclusion of Al air bridges (rather than insulating spacers) eliminates extraneous dielectric loss pathways. Testing two different dielectric thicknesses (13.5 nm and 58.3 nm) is a smart control: the similar low-power loss values for both thicknesses argue that bulk γ-Al₂O₃ loss dominates over interfacial contributions, strengthening the claim that this is an intrinsic material property measurement.
One methodological limitation is the reliance on PLD, which, while excellent for achieving the required crystallinity, may face challenges in scaling to wafer-level production compared to MBE or sputtering. The authors also acknowledge that the ~1.5 nm TixOyNz interlayers at oxide-nitride interfaces, while thin, could contribute to power-independent losses observed at high drive powers. The paper is transparent about these limitations.
3. Potential Impact
The implications span multiple domains:
Superconducting qubits: A two-order-of-magnitude reduction in dielectric TLS loss directly translates to potential improvements in qubit coherence, which is the primary barrier to fault-tolerant quantum computing. This is particularly relevant for compact qubit architectures (merged-element transmons) where the dielectric participates strongly.
Device miniaturization: The LEPPC devices occupy ~2.4 × 10⁻² mm², two orders of magnitude smaller than conventional CPW geometries. This addresses the critical scalability problem of integrating thousands of qubits on a single chip.
MKID detectors: The high Qmax values (up to 6.4 × 10⁵) at high powers make this platform attractive for microwave kinetic inductance detectors used in astronomy and particle physics.
Materials platform generalizability: The demonstration that oxide-nitride heteroepitaxy can yield clean interfaces opens the door to exploring other crystalline oxide dielectrics (e.g., MgO, SrTiO₃) on nitride superconductors, potentially spawning a broader materials development effort.
4. Timeliness & Relevance
This work is highly timely. The superconducting qubit community is at an inflection point where materials quality has become the primary limiting factor for further progress. Major industry players (IBM, Google) and national laboratories are investing heavily in materials solutions. The paper directly addresses this need with a platform compatible with thin-film deposition infrastructure. Furthermore, the growing interest in compact qubit designs (mergemons, 3D integration) makes low-loss crystalline dielectrics increasingly urgent.
5. Strengths & Limitations
Strengths:
Limitations:
Overall Assessment:
This is a strong experimental paper that makes a clear and well-supported contribution to an important problem. The combination of thorough materials science and functional device validation sets a high standard for the field. While it stops short of demonstrating a working qubit, the demonstrated loss tangent and compact device geometry represent meaningful progress toward scalable, high-coherence superconducting quantum circuits.
Generated Apr 1, 2026
Comparison History (47)
Paper 1 addresses one of the most critical bottlenecks in realizing fault-tolerant quantum computing: decoherence due to dielectric loss. By experimentally validating a novel heteroepitaxial materials platform with low intrinsic loss, it offers a tangible solution that directly impacts the scalability and performance of superconducting quantum circuits. While Paper 2 presents an important security framework for QKD, the successful development of scalable quantum computers (facilitated by Paper 1's findings) would have a far broader and more transformative scientific and technological impact across multiple disciplines.
Paper 2 is more broadly impactful: it proposes a general finite-temperature quantum algorithmic framework that can apply across many Hamiltonians and domains (condensed matter, chemistry, materials, high-energy), with clear relevance to near-term quantum hardware constraints (avoids temperature-specific thermal-state prep). The approach is timely for NISQ-era computation and could influence multiple follow-on methods. Paper 1 is rigorous and valuable for superconducting-circuit materials, but its impact is more specialized to a specific materials stack and device class, whereas Paper 2’s methodological novelty and cross-field applicability suggest higher overall scientific impact.
Paper 1 addresses a critical hardware bottleneck in quantum computing—qubit coherence—by demonstrating a novel low-loss epitaxial dielectric. Advancements in foundational quantum materials have a profound, long-lasting impact, as they are a prerequisite for scalable, fault-tolerant quantum computing across all applications. While Paper 2 offers a clever near-term algorithmic workaround for data loading in the NISQ era, Paper 1's breakthrough in superconducting circuit architecture fundamentally advances the physical viability of the entire quantum computing ecosystem.
Paper 2 likely has higher impact due to direct relevance to a central bottleneck in superconducting quantum computing: dielectric loss from amorphous materials. It provides experimentally validated, quantitative microwave-loss metrics for an epitaxial dielectric integrated in a device-relevant trilayer, with strong structural/chemical characterization—high methodological rigor and clear translation to qubits, resonators, and detectors. The platform could broadly influence quantum hardware materials and scalable architectures. Paper 1 is novel and conceptually exciting, but its near-term applications and breadth beyond specialized strong-field/quantum-optics contexts are less certain.
Paper 2 addresses a critical bottleneck in the physical realization of fault-tolerant quantum computers: microwave loss in dielectrics. By demonstrating a novel materials platform (oxide-nitride heteroepitaxy) with low intrinsic two-level system loss, it offers immediate, highly relevant applications to superconducting qubits and quantum sensor architectures. While Paper 1 provides a strong theoretical computer science result for quantum query complexities, Paper 2's experimental breakthroughs possess much broader interdisciplinary impact across materials science, physics, and practical quantum engineering.
Paper 1 is likely higher impact because it delivers a rare, experimentally validated materials platform directly addressing a key bottleneck in superconducting quantum hardware: dielectric loss. It reports a first direct measurement of intrinsic TLS loss in epitaxial γ-Al2O3 and demonstrates high-quality heteroepitaxy with rigorous structural/chemical characterization, enabling near-term integration into compact qubit/resonator architectures. Paper 2 is timely and useful, but ML-based reconstruction methods are less fundamental and more field-specific, and impact depends on adoption/validation across varied experimental conditions.
Paper 1 offers experimentally validated, materials-level innovation directly addressing a key bottleneck in superconducting quantum hardware: dielectric loss. It reports a first direct loss measurement for epitaxial γ-Al2O3 with low TLS loss and demonstrates a scalable heteroepitaxial TiN/oxide/TiN platform, with clear near-term applications (qubits, MKIDs) and strong methodological rigor (correlative microscopy/diffraction/spectroscopy + microwave resonators). Paper 2 is conceptually ambitious but more speculative/theoretical, with higher dependence on modeling assumptions and less immediate, verifiable impact.
Paper 2 likely has higher scientific impact due to stronger novelty and breadth: it gives provably optimal certification of local Hamiltonians (matching a fundamental Heisenberg lower bound) and introduces sample-/time-efficient learning and certification of Gibbs states, resolving an explicit open question. These results are methodological and broadly applicable across quantum information, algorithms, and many-body physics, with timeliness for quantum verification/validation. Paper 1 is a high-quality, relevant experimental materials advance for superconducting circuits with clear application value, but its impact is narrower to a specific materials platform/device stack.
Paper 1 addresses a critical hardware bottleneck in superconducting qubits by experimentally validating low-loss crystalline dielectrics. This practical breakthrough has immediate, widespread applicability for improving coherence times and scaling fault-tolerant quantum computers. While Paper 2 offers valuable theoretical complexity results for a restricted quantum model, the tangible hardware advancements in Paper 1 demonstrate a higher potential for broad and immediate real-world impact across the quantum computing field.
Paper 2 demonstrates a scalable, practical architecture combining two frontier technologies—twin-field QKD and integrated Kerr soliton microcombs—to achieve over an order-of-magnitude improvement in secure key rates (1.57 Mbps over 201 km). This addresses a critical scalability bottleneck in quantum communications with clear real-world deployment implications. While Paper 1 makes a solid materials contribution to superconducting qubits with the first direct loss measurement of epitaxial γ-Al₂O₃, Paper 2's innovation has broader immediate impact across quantum networking, integrated photonics, and telecommunications, with a clearer path to transformative applications.