Tantalum makes excellent superconducting qubits. The alpha phase (α-Ta) — body-centered cubic, the thermodynamically stable form — has been the material of choice for high-coherence transmon qubits. The beta phase (β-Ta) — tetragonal, metastable, historically considered a nuisance that forms during thin-film deposition — was avoided. Wrong crystal structure, wrong properties, wrong phase.
Kopas et al. (arXiv:2603.13174) fabricated transmon qubits from β-Ta on sapphire and measured quality factors approaching 10 million. The mean time-averaged quality factor was (5.6 ± 2.3) × 10⁶, with the best qubit reaching (10.1 ± 1.3) × 10⁶. These are competitive with the best α-Ta qubits.
The surprise: β-Ta has roughly twice the surface loss of α-Ta. Surface two-level systems — the primary decoherence mechanism — are worse. But β-Ta has significant kinetic inductance, with a magnetic penetration depth of ~1.78 μm. This kinetic inductance enables compact resonators, kinetic inductance detectors, and quasiparticle traps — components that α-Ta can't efficiently provide.
The tradeoff inverts the usual logic. Higher surface loss should mean worse qubits. But the kinetic inductance opens design possibilities that α-Ta forecloses: smaller footprints, integrated detection of quasiparticle poisoning events, and natural filtering of environmental noise. A qubit that's individually slightly worse but that enables a better system architecture is the better choice at scale.
The metastable phase — the one that wasn't supposed to be there — turns out to be a viable platform precisely because of properties that the stable phase lacks. The “wrong” crystal structure has the right ancillary features. Material selection for quantum processors isn't about minimizing loss in isolation; it's about maximizing the set of functions a single material platform can serve.