High-Throughput Microwave Package for Precise Superconducting Device Measurement

Abstract

Cryogenic microwave measurement of superconducting quantum devices is complicated by the packaging required to connect devices to control and readout circuitry. In this work, we outline the design and experimental demonstration of a wirebond-free, PCB-free, drop-in microwave package for on-chip superconducting quantum devices. The package is composed of a superconducting aluminum cavity with a suspended tungsten transmission pin. The fundamental package cavity mode is far detuned from the 4 GHz to 8 GHz band of interest, and the pin transmission exhibits less than 3 dB of ripple across this range. We demonstrate the use of this package to extract the loss tangent of superconducting ring resonators, measuring a value of (1.10 +- 0.09) x 10^-6, which agrees with measurements of lambda/4 resonators in wirebond-based packaging. This high-throughput measurement system will allow the rapid generation of large datasets for improving superconducting qubit performance, and facilitate time-sensitive surface passivation and oxide regrowth studies.

Figures (4)

• Figure 1: Microwave package design. (a) CAD model showing tungsten pin (yellow), pogo pins (red), and installed chip (bright green). The package consists of four main parts: the cavity base (dark green), the lid (light blue), the interposer (pink), and the SMP holders (pale green). The tungsten pin feeds directly into SMP bullets (dark pink) on either side. (b)-(d) Photographs of closed package (b), cavity base and SMP holders (c), and lid and interposer (d).
• Figure 2: Experimental and simulated transmission through the microwave package. HFSS simulation of the empty package is shown (solid grey). Cryogenic measurements of empty package (blue) and associated through (dashed blue) in cooldown 1, and package with ring resonator chip installed (green), and associated through (dashed green) in cooldown 2 are shown. Comparison between package and through is facilitated by the use of cryogenic switches. The close agreement between package and through traces demonstrates a clean microwave background in the cavity package.
• Figure 3: Characterization of a ring resonator using the cavity package. (a) Measured $S_{21}$ transmission showing the mode splitting for one ring resonator. (b) Ansys HFSS simulation showing the electric field distributions of the two standing-wave modes of a single ring resonator.
• Figure 4: Microwave characterization of ring resonators in the superconducting cavity package. (a) and (b) Diameter correction method (DCM) fit of a single resonance in the low-power regime. A ring resonator measured at an estimated on-chip power of -136 dBm ($\langle n \rangle = 4$) is shown as an example. Points are normalized data, star is resonance, and line is fit to DCM. (c) and (d) Inverse internal quality factor $Q_i^{-1}$ and coupling quality factor $Q_c$ as a function of average number of photons $\langle n \rangle$ for all resonances. Error bars denote the 95 $\%$ confidence interval for each circle fit, and lines denote fits to Eq. \ref{['eq:scurve']}. (e)-(h) Extracted loss parameters for the ring resonators in the cavity package compared to $\lambda/4$ resonators measured in a standard wirebonded package. TLS loss tangent $F \delta^0_{\mathrm{TLS}}$, high power quality factor $Q_{\mathrm{HP}}$, critical photon number $n_c$, and empirical parameter describing TLS interaction $\beta$ are compared. Box shows the median value and the 25th and 75th percentiles, and points denote individual measurements.