Development of transmon qubits solely from optical lithography on 300mm wafers

  1. N. Foroozani,
  2. C. Hobbs,
  3. C. C. Hung,
  4. S. Olson,
  5. D. Ashworth,
  6. E. Holland,
  7. M. Malloy,
  8. P. Kearney,
  9. B. O'Brien,
  10. B. Bunday,
  11. D. DiPaola,
  12. W. Advocate,
  13. T. Murray,
  14. P. Hansen,
  15. S. Novak,
  16. S. Bennett,
  17. M. Rodgers,
  18. B. Baker-O'Neal,
  19. B. Sapp,
  20. E. Barth,
  21. J. Hedrick,
  22. R. Goldblatt,
  23. S. S. Papa Rao,
  24. and K. D. Osborn
Qubit information processors are increasing in footprint but currently rely on e-beam lithography for patterning the required Josephson junctions (JJs). Advanced optical lithography
is an alternative patterning method, and we report on the development of transmon qubits patterned solely with optical lithography. The lithography uses 193 nm wavelength exposure and 300-mm large silicon wafers. Qubits and arrays of evaluation JJs were patterned with process control which resulted in narrow feature distributions: a standard deviation of 0:78% for a 220 nm linewidth pattern realized across over half the width of the wafers. Room temperature evaluation found a 2.8-3.6% standard deviation in JJ resistance in completed chips. The qubits used aluminum and titanium nitride films on silicon substrates without substantial silicon etching. T1 times of the qubits were extracted at 26 – 27 microseconds, indicating a low level of material-based qubit defects. This study shows that large wafer optical lithography on silicon is adequate for high-quality transmon qubits, and shows a promising path for improving many-qubit processors.

State-Signal Correlations of a Continuously Monitored Superconducting Qubit

  1. N. Foroozani,
  2. M. Naghiloo,
  3. D. Tan,
  4. K. Mølmer,
  5. and K. W. Murch
A superconducting transmon qubit undergoing driven unitary evolution is continuously monitored to observe the time evolution of its quantum state. If projective measurements are used
to herald a definite initial state, the average of many measurement records displays damped Rabi oscillations. If instead the average of many measurements is conditioned on the outcome of a final post-selection measurement, the result exhibits similar damped Rabi oscillations with the exception that the damping of the signal occurs backwards in time. Such pre- and post-selections are specific examples of qubit state and signal temporal correlations and stimulate a more general discussion of the temporal correlations in stochastic quantum trajectories associated with continuous quantum measurements.