Research From Google Quantum AI et al, Shows Superconducting Dissipation Scales with Superfluid Density, Limiting Quantum Coherence

The quest for stable quantum computing relies on minimising energy loss in superconducting circuits, but even the most perfect conductors exhibit unwanted dissipation when carrying microwave currents. Researchers Thibault Charpentier, Anton Khvalyuk, and their colleagues from the Université Grenoble Alpes, CNRS, and Google Quantum AI have now demonstrated a universal relationship between this microwave dissipation and the superfluid density of the superconducting material itself. This discovery, spanning a diverse range of materials and device designs, reveals a fundamental source of energy loss originating from trapped, non-equilibrium particles created by imperfections within the material. By identifying this intrinsic limit to coherence, the team provides a crucial framework for both material selection and the design of future quantum circuits with significantly improved performance.
Superconducting circuits are becoming a leading platform for building quantum computers, but achieving stable and reliable computation requires maintaining the delicate quantum states of qubits for as long as possible – a property known as coherence. Researchers have consistently observed energy dissipation within superconducting materials that limits coherence times and hinders performance. Recent investigations have revealed a fundamental mechanism driving this dissipation, offering a pathway towards significant improvement.

The research demonstrates a universal relationship between microwave dissipation – the loss of energy within the superconducting material – and the superfluid density, a property reflecting the concentration of charge carriers and the level of material disorder. This connection was established through a comprehensive analysis of a diverse range of superconducting materials, from highly disordered films to immaculate systems, and across various device designs, including resonators and complex 3D cavities.

The findings indicate that a key source of dissipation isn’t related to surface imperfections, but originates from an intrinsic material property – a population of nonequilibrium quasiparticles created by disorder within the superconductor. These quasiparticles – essentially broken Cooper pairs – absorb energy from microwave currents, contributing to the observed dissipation. The team’s analysis shows that the level of dissipation scales predictably with the superfluid density, providing a framework for predicting and optimising coherence times in future devices.

This discovery is significant because it identifies a fundamental limit to coherence dictated by the material’s inherent properties. Previously, researchers focused heavily on minimising surface losses, but this work highlights the importance of controlling the density of these quasiparticles within the bulk of the superconducting material. The research defines a clear material-dependent upper limit on coherence, with niobium identified as the optimal material due to its high superfluid density, which is capable of supporting significantly longer coherence times than commonly used materials, such as aluminium or tantalum.

While improvements in filtering, gap engineering, and packaging can mitigate some losses, reaching the theoretical coherence limit of approximately 0.1 to 1 second will be particularly challenging when scaling up to multi-qubit processors. Addressing substrate dissipation, a further limiting factor in planar circuits, may require advancements in material quality or the development of radically new circuit architectures, such as suspended circuits or 3D designs. The team has made their data and code publicly available, establishing a benchmark for the superconducting quantum computing industry and providing a foundation for future materials selection and circuit design.