The dominant technology in the current generation of quantum computers relies on superconducting circuits, tiny electrical oscillators that, when cooled sufficiently, carry current without resistance. These devices have enabled impressive demonstrations of quantum computation, but they come with a fundamental weakness. Superconducting qubits are exquisitely sensitive to electromagnetic noise and to stray electrical charges, the kind of static electricity that makes hair cling to a balloon on a dry day. Even minuscule perturbations of this sort can scramble the delicate quantum states that encode information, introducing errors that must be corrected through elaborate overhead. As engineers attempt to pack more and more qubits onto a chip, keeping these error rates under control becomes one of the central bottlenecks to scaling the machines up.

The Surrey team, drawn from the university’s Quantum Sciences Group, has proposed a radically different approach to quantum hardware. Their proposed device, named the Superfluid Helium Oscillator Quantum, or SHOQ, would store and manipulate quantum information in quantized oscillations within superfluid helium-3. Because the medium is electrically charge-neutral, the qubit is naturally immune to many of the electromagnetic disturbances and stray charges that torment conventional superconducting devices. According to the team’s theoretical analysis, this intrinsic protection could translate into error rates roughly 100 times lower than those of standard superconducting qubits, a margin that would substantially ease the burden of quantum error correction in a large-scale machine.

The concept is, the researchers note, the first reported design for a qubit based on superfluid helium. While the individual physical ingredients have long been studied in isolation, the Surrey group is the first to assemble them into a coherent microfluidic device architecture and to work out the specific parameters and specifications needed for the device to function as a qubit. Dr Priya Sharma, Daphne Jackson Fellow in Hybrid Quantum Systems at the University of Surrey’s School of Mathematics and Physics and lead author of the study, emphasized that the work is an educated design grounded in established physics rather than a speculative sketch. The mathematics, she explained, indicates that the device should work as intended, and the crucial next step is to fabricate a prototype and test the predictions experimentally.

The underlying physics is as fascinating as the engineering ambition. Helium-3, the lighter isotope of helium, becomes a superfluid at temperatures only a few thousandths of a degree above absolute zero. In this state, the liquid flows with zero viscosity and exhibits quantum behavior on a macroscopic scale, with collective oscillations whose energy levels are quantized just like those of atoms. The SHOQ proposal taps into these quantized mechanical vibrations as the carrier of quantum information. Because these oscillations involve neutral atoms rather than moving charges, they do not couple strongly to the electric fields and charge fluctuations that are ubiquitous in solid-state environments, offering what physicists call a quieter platform for preserving delicate quantum states.

An especially significant feature of the proposal is that the SHOQ device is not intended to replace existing quantum technology outright. The paper outlines how the superfluid-based qubit could be coupled with current superconducting quantum hardware, raising the possibility that the two technologies might operate side by side within a single larger quantum system. Dr Eran Ginossar, Associate Professor at the University of Surrey’s Department of Physics and Advanced Technology Institute and co-author of the study, argued that no single qubit technology needs to do everything. Combining different quantum platforms, he suggested, could allow engineers to exploit the particular strengths of each, and superfluid helium offers a fundamentally new type of quantum hardware to explore. If the predicted performance can be demonstrated in the laboratory, such devices could eventually work alongside superconducting systems as components of hybrid architectures.

One potential application highlighted by the team is quantum memory. In a future hybrid computer, a version of the SHOQ device could serve as a long-lived repository for quantum information, storing fragile states while a separate processor built from different hardware performs calculations. This division of labor mirrors the separation between memory and processing units in classical computers and could prove decisive in the quest for machines that are both powerful and reliable. The low sensitivity of charge-neutral superfluid helium to environmental noise makes it a natural candidate for the memory role, where preservation of quantum coherence over time is the paramount requirement.

The Surrey effort is not proceeding in isolation. The work was carried out in collaboration with Professor Jens Koch of Northwestern University in the United States, a physicist who was among the researchers behind the development of the transmon, the superconducting qubit design that has become the workhorse of much of today’s quantum computing industry. That pedigree gives the new proposal considerable weight, since the transmon itself succeeded by engineering away sensitivity to charge noise, and the SHOQ concept extends the same philosophy into an entirely different physical medium. The involvement of researchers with hands-on experience in bringing a qubit design from theory to widespread laboratory use may help the new idea avoid some of the pitfalls that accompany novel hardware concepts.

The team is now turning its attention to building a prototype to determine whether the theoretical predictions survive contact with reality, an effort supported by an IAA Commercialisation Fellowship awarded to Dr Sharma. The cryogenic challenge is formidable but not unprecedented: although the SHOQ device would need to operate at extremely low temperatures, conditions of exactly this kind have already been achieved experimentally in superfluid helium-3 research laboratories around the world. That existing experimental infrastructure means the path from concept to prototype does not require inventing entirely new cryogenic techniques, only adapting well-established ones to a new microfluidic device. If the prototype confirms the predicted hundredfold reduction in error rates, superfluid helium could move from the margins of low-temperature physics to the center of the conversation about how to scale quantum computers, adding a genuinely new and remarkably quiet material platform to the engineer’s toolkit.

The choice of helium-3 rather than the more common helium-4 is central to the proposal. Helium-4 atoms are bosons and form a superfluid at around two kelvin, but helium-3 atoms are fermions, which means they cannot condense directly. Instead, at temperatures a few thousandths of a degree above absolute zero, pairs of helium-3 atoms bind together in a manner analogous to the Cooper pairs of electrons in a superconductor, and it is these paired atoms that flow without viscosity. This pairing mechanism gives superfluid helium-3 a rich internal structure, including multiple distinct superfluid phases, and endows the liquid with collective modes whose quantum properties are exceptionally well characterized by decades of low-temperature research.

The quantized vibrations that the SHOQ design would exploit belong to a broader family of mechanical quantum systems that physicists have been developing for years. Researchers have previously succeeded in cooling micromechanical drums and membranes to their quantum ground states and entangling them with light, establishing that mechanical oscillators can genuinely store and process quantum information. What has been missing is a mechanical oscillator whose intrinsic noise performance rivals that of the best electronic qubits, and the Surrey team argues that a charge-neutral superfluid medium could supply exactly that, since acoustic modes in helium couple only weakly to the solid-state defects and two-level fluctuators that degrade fabricated resonators on chips.

The significance of a hundredfold reduction in error rates becomes clearer when viewed through the lens of quantum error correction. Theoretical studies of fault-tolerant computation indicate that below a critical error threshold, adding more physical qubits suppresses logical errors exponentially, but the overhead involved is enormous when physical error rates sit near the threshold. Lowering the physical error rate by two orders of magnitude would reduce the number of physical qubits needed per logical qubit by a comparable factor, potentially shrinking the machine required for useful fault-tolerant computation from millions of qubits to a far more manageable scale.

The hybrid vision also echoes patterns from other parts of the quantum technology landscape. Trapped-ion systems already combine different species of ions, using one type for memory and another for logic, while superconducting processors have been coupled to spin defects in diamond and to atomic ensembles acting as quantum memories. The SHOQ concept would extend this modular philosophy to a liquid platform, connecting a microfluidic cell through microwave circuitry to conventional superconducting control electronics. The paper’s authors suggest that such interfaces, rather than any single monolithic technology, may ultimately define how large quantum computers are assembled.

Considerable uncertainty remains, as is inevitable for a purely theoretical design. Real devices must contend with damping of acoustic modes at their boundaries, thermal excitations that must be filtered out, and the practical difficulty of coupling a liquid oscillator strongly enough to microwave circuits to allow fast quantum gates. The prototype planned under the fellowship is intended to probe precisely these questions, and the coming experimental results will determine whether the elegant mathematics translates into working hardware.

Subject of Research: A conceptual superfluid helium-3 based qubit design for fault-tolerant quantum computing

Article Title: Superfluid-based qubit design could be key to scaling up quantum computers

Article References: Superfluid-based qubit design could be key to scaling up quantum computers. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: superfluid helium, qubit, quantum computing, error rates, superconducting qubits, SHOQ device, quantum memory, hybrid quantum systems, npj Quantum Information, University of Surrey, helium-3, microfluidics

Historically, enhancing qubit coherence has largely hinged on materials engineering. Previous strides utilized tantalum as a superconducting base layer and sapphire as a substrate, yielding significant improvements in qubit quality. However, losses attributed to two-level systems (TLS) persisted, originating comparably from surface and bulk dielectrics, thus hampering further gains. This dual-source loss mechanism indicated that merely refining surfaces or bulk materials was insufficient; a holistic approach targeting both was essential.

In a bold conceptual shift, the team replaced the conventional sapphire substrates with high-resistivity silicon. This substitution markedly curtailed dielectric losses associated with the bulk substrate, which had historically imposed heavy limitations on coherence. Silicon’s intrinsic properties, including superior crystalline purity and reduced defect densities, underpin the substantial decrease in bulk dielectric loss. Consequently, this innovation has led to 2D transmon qubits with average quality factors (Q_avg) reaching heights of approximately 9.7 million across a broad ensemble of 45 qubits—a consistency that bodes well for scalability.

Drilling down to individual device performance, the researchers demonstrated a single best qubit attaining an astonishing Q_avg of 1.5 × 10^7 and peaking at a maximum Q of 2.5 × 10^7. This translates directly to a relaxation lifetime (T1) stretching up to 1.68 milliseconds, markedly surpassing historic performance benchmarks. Such extended coherence durations hold extraordinary promise for quantum error correction and complex algorithmic executions, permitting longer algorithm runtimes before decoherence events intrude.

Beyond these raw coherence metrics, the lowered material loss environment facilitated a novel observation: decoherence phenomena tied specifically to the Josephson junction itself. This component, fundamental to qubit operation, had traditionally eluded isolation due to overshadowing bulk and surface losses. Recognizing this nuanced decoherence source, the team engineered an optimized junction deposition method characterized by minimal contamination. This methodological refinement addressed previously unquantified loss channels and further elevated coherence.

Implementing the enhanced junction fabrication techniques bore immediate fruits, as evidenced by Hahn echo coherence times (T2E) surpassing even T1 lifetimes. This counterintuitive regime—where dephasing times outstrip relaxation times—signifies a breakthrough in mitigating intrinsic noise sources, highlighting a profound leap in qubit stability. Achieving T2E > T1 is a hallmark of qubits entering practically usable regimes for quantum information processing.

Crucially, all these advancements were achieved without altering the core 2D transmon qubit architecture. The tantalum-on-silicon platform involves a straightforward material stack amenable to current production frameworks, thus promising facile integration into existing fabrication pipelines. This compatibility not only simplifies the transition to wafer-scale production but also maintains the versatility needed for implementing standard quantum control gates without necessitating extensive redesigns.

In operational demonstrations, these refined qubits exhibited single-qubit gate fidelities of 99.994%, approaching the elusive fault-tolerant threshold. Such near-perfect gate operations underscore the platform’s readiness for the intense demands of practical quantum computation, where every decimal point in fidelity can exponentially enhance algorithmic success.

Beyond just engineering prowess, this work exemplifies the intrinsic interplay between materials science and quantum device physics. By astutely tailoring the substrate environment and honing junction quality, the researchers have not simply improved performance metrics but have charted a systematic route to minimizing decoherence. Their findings illuminate pathways to potentially overcome the longstanding qubit lifetime barriers in 2D architectures.

Looking forward, tantalum-on-silicon qubits offer a promising platform for scaling quantum processors. The simplicity and reproducibility of the material stack invite large-scale production, offering a pragmatic solution to expanding qubit arrays without sacrificing coherence. This aligns well with the broader industry imperative: transitioning high-coherence qubits from lab curiosities to industrial workhorses.

This remarkable advance also signals fresh opportunities to explore fundamental quantum phenomena hitherto masked by material-imposed losses. Access to ultra-long-lived qubits provides a richer playground for probing interactions and noise mechanisms at unprecedented resolution, potentially sparking innovations in quantum control and calibration techniques.

In sum, the convergence of material innovation, precise fabrication, and quantum engineering embedded in this tantalum-on-silicon platform stands to accelerate the quantum computing revolution. The leap to millisecond-scale lifetimes in 2D transmons could redefine benchmarks for error correction thresholds, gate fidelities, and ultimately the realization of fault-tolerant quantum processors. As the field races toward practical quantum advantage, such breakthroughs chart the essential path forward.

Subject of Research: Superconducting qubits, specifically 2D transmon qubits, and their coherence and lifetime improvements through advanced materials engineering.

Article Title: Millisecond lifetimes and coherence times in 2D transmon qubits.

Article References:
Bland, M.P., Bahrami, F., Martinez, J.G.C. et al. Millisecond lifetimes and coherence times in 2D transmon qubits. Nature (2025). https://doi.org/10.1038/s41586-025-09687-4

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41586-025-09687-4