A Cornell team led by Assistant Professor Valla Fatemi reported a new tantalum deposition process for superconducting qubits in Nature Materials on August 19, 2026. The group replaced argon with krypton during magnetron sputtering. The change lowered the substrate temperature needed to form the desired body-centered-cubic alpha phase of tantalum from more than 400°C to 200°C.
The manufacturing problem
Tantalum attracts superconducting qubit makers because its stable surface oxide reduces dielectric loss compared with niobium. The problem is temperature. Argon sputtering needs silicon substrates above 400°C to produce alpha-phase tantalum. That thermal budget exceeds standard CMOS back-end-of-line limits and causes tantalum-silicon intermixing, which hurts qubit coherence.
Krypton ions carry more momentum than argon ions. The heavier ions transfer enough energy to ejected tantalum atoms to stabilize the alpha phase at 200°C. The Cornell team fabricated transmon qubits with compact 20-micrometer capacitor gaps and reported internal quality factors up to 16.9 million.
Why 200°C matters
Commercial semiconductor foundries already run processes near 200°C. A 200°C tantalum step opens the door to integrating high-performance superconducting qubits into existing automated CMOS lines without damaging control circuits or interconnect layers. The U.S. Department of War's Microelectronics Commons Program and the Air Force Office of Scientific Research supported the work.
The caveat
A materials paper is not a full foundry process. The Cornell team demonstrated the technique in an academic cleanroom, not at volume in a commercial fab. Quality-factor numbers are useful indicators, but the real test is whether a foundry is able to repeat the process across wafers with yield and uniformity good enough for error correction.
For background on superconducting qubits, see our T1 vs T2 explainer and our understanding quantum error correction post.