Short answer: Microwave loss in tantalum superconducting circuits comes mostly from two-level defects at surfaces and in the bulk. A 2023 Physical Review X paper (also posted as arXiv 2301.07848) measured resonators with internal quality factors of 5 to 15 million and found surface and bulk losses comparable, and a 2026 arXiv preprint reports that molecular monolayers on tantalum raised the quality factor to 1.8 million, about 140% above untreated devices. The second is a preprint and a different device type and conditions.
What limits tantalum circuits
The 2023 abstract says superconducting qubits suffer from coherence times limited by microwave dielectric loss, and that recently discovered tantalum-based qubits show record lifetimes above 0.3 ms. The authors measure tantalum resonators systematically to separate the loss sources. By varying temperature, photon number and device geometry, they find losses are dominated by several types of saturable two-level systems (TLSs), with evidence that both surface and bulk TLSs contribute.
2023: chemistry changes surface loss
The authors show that surface TLSs can be altered with chemical processing. With four surface conditions, they quantitatively extract the linear absorption associated with different surface TLS sources. At single-photon powers, the relevant regime for qubits, they measure resonators with internal quality factors from 5 to 15 million, which they say is comparable to the best qubits reported. In these devices the surface and bulk TLS contributions are comparable, so the authors conclude systematic improvements in materials on both fronts are needed. The Physical Review X abstract and the arXiv abstract are the same work.
2026: molecular passivation
The arXiv preprint 2604.22112 says native oxide regrowth makes interfacial TLS losses worse. The authors apply self-assembled organic monolayers to freshly etched tantalum and silicon in coplanar waveguide resonators. Contact angle, X-ray photoelectron spectroscopy, infrared spectroscopy and transmission electron microscopy confirm ordered, nanometer-thick films that suppress oxide formation.
Microwave measurements at 5 to 9 GHz show internal quality factors up to 1.8 million in the single-photon regime at 100 millikelvin, which the authors describe as about a 140% improvement over untreated devices with native oxide. Power- and temperature-dependent measurements attribute the improvement to reduced TLS loss. They conclude that molecular passivation provides a scalable route to high-coherence devices, which is their projection. The preprint has not been reviewed.
How to read the two sets of numbers
The 2023 resonators reach quality factors of 5 to 15 million, more than the 1.8 million in the 2026 preprint, but they come from different fabrication, different treatments and different measurement conditions. The 2026 number is relative to an untreated control in the same study. The abstracts do not compare the two directly.
Why both surface and bulk matter
The 2023 conclusion has a practical reading: treating only the surface leaves bulk loss in place, and the reverse. The 2026 preprint works on the surface side, using monolayers on tantalum and also on silicon, and the abstract does not say it addresses bulk loss.
Limits
Resonator quality factors are a proxy for qubit loss, and the abstracts do not report qubit coherence from the treated films.
FAQ
What is a two-level system?
A defect that absorbs microwave energy, here found at surfaces and in the bulk.
What quality factors were reported?
5 to 15 million in 2023, and up to 1.8 million in the 2026 preprint.
Does chemistry help?
Both papers report that surface treatment changes loss.
Which loss matters more?
The 2023 authors find surface and bulk losses are comparable.
Is the 2026 result reviewed?
No. It is an arXiv preprint.
All field notes