Epitaxial superconducting refractory metals for quantum computing

1. Epitaxial superconducting refractory metals for quantum computing David P. Pappas NIST - Colorado University of California - Santa Barbara John M. Martinis Ken Cooper Matthias Steffen Robert McDermott Seongshik Oh Raymond Simmonds Katarina Cicak Kevin Osborn

2. Challenges in solid state qubits 1) Need longer T1 • identify dominant loss mechanisms • Substrate & insulator– SiO2? • Kevin Osborn • John Martinis • Next session 2 ) Need higher measurement fidelity • Identify, eliminate intrinsic resonances • Junction dielectric?

3. Design of tunnel junctions What we want: What we have: Spurious resonators in junctions Fluctuations in barrier No spurious resonators Stable barrier Poly - Al Crystalline barrier g-Al2O3 Interfaces: Smooth Stable No dangling bonds Amorphous tunnel barrier a -AlOx-OH Rough interfaces Unstable at room temp. Dangling bonds SC bottom electrode Poly- Al amorphous SiO2 dangling bonds at interface Low loss substrate Silicon

4. Q: Can we prepare crystalline Al2O3 on Al? 68 Metallic aluminum 10 Å AlOx on Al (300 K + anneal) 10 Å AlOx on Al (exposed at elevated temp.) AES Energy of Reacted Al (eV) Aluminum Melts Al in sapphire Al203 Annealing Temp (K) • Anneal the natural oxides • Oxidize at elevated temp. Binding energy of Al AES peak in oxide A: No

5. Chose bottom superconducting electrode to stabilize crystalline Al2O3 tunnel barrier Elements with high melting temperature

6. Elements with TC > 1K

7. Elements that lattice match sapphire (Al203)

8. Elements that form weaker bond with oxygen than Al

9. Elements that are not radioactive

10. UHV growth system • Pbase< 5x10-10 Torr • Sapphire c-axis substrates • Sputter deposit Re Load Lock LEED, RHEED, AES Re Sputtering STM

11. Morphology of Re/sapphire Room temperature growth100 nm Re 0.5x0.5 um • 3 nm RMS roughness • Mixed growth planes • c-plane • a-plane • Needs to be heated for barrier growth

12. 100 nm Re, room temperature deposition + 750 C anneal 0.5x0.5 um • 1 nm RMS roughness • Re surface begins to crystallize between 550–650C • Need higher temperature to crystallize Al2O3

13. Growth of epitaxial Re(0001) at high temperature RHEED diffraction images + 100 nm Re @ 850 C Sapphire substrate epi-Re on Sapphire

14. High temperature growth – 100 nm Re @ 850 C 500 x 500 nm • 1.5 nm RMS roughness • 2 atomic layer steps • Screw dislocations on mesas • Stranski-Krastanov growth • Initial wetting of substrate • Formation of 3-d islands • Islands fill in gradually • Evidence of step bunching => some very large steps

15. 100 nm Re, 850 C deposition – zoom in 200 x 200 nm • Step bunching on corners • Sharp dropoffs where multiple steps come together • ~100 nm wide mesas

16. 100 nm Re, 850 C deposition, 1200 C anneal 500 x 500 nm • Much large mesas ~ 200 nm diameter • Still find step bunching • Temperatures very high

17. Grow thin film at low T, anneal=> add thick film with homoepitaxy @ high T + 100 nm Re @ 850 2 nm Re, R.T. + 850 C anneal 500 x 500 nm => 200 nm terraces, comparable to 1200 C anneal

18. Conclusions • Need bottom electrodes that are stable at high T T > 700 C • Demonstrated Re growth with large terraces • Films need to be annealed to > 800 C to stabilize surface • Large mesas with wide terraces can be obtained 3 ways: • High temperature growth ~850 C => 100 nm mesas • Anneal to very high temperature, ~ 1200 C => 200 nm • Low T buffer, anneal to 850, then 850 C film => 200 nm • Need to grow epitaxial Al2O3 on these surfaces

19. Chose bottom superconducting electrode to stabilize crystalline Al2O3 tunnel barrier • Element with high melting temperature • TC > 1K • Epitaxial match to Al2O3 – hcp, 2.77 Å Re - hcp (0001) < 1% lattice mismatch • Re - smaller oxidation energy (sharp interface)