Miniaturized high-energy radiation drain filters for quantum computing applications

Abstract

Quantum Processing Units (QPUs) using superconducting qubits are known for their sensitivity to various types of radiation and many different aspects are being worked on in order to increase their resilience against noise. Superconducting qubits are especially sensitive frequencies with energy exceeding twice the superconducting energy gap of the superconductor. Above this frequency, Copper-pairs start to break, which partially disrupts superconductivity and degrades the performance of the QPU. In order to tackle this problem, a new type of low pass filtering technique called High Energy Radiation Drain (HERD) has been developed at at Chalmers University of Technology. Unlike previously employed filters relying on absorptive materials and resonant circuits to block high-frequency photons, this novel filtering technique overcomes the trade-off between low losses in the passband and high attenuation in the stopband. However, the filter is relatively large compared other filtering techniques, which makes it less suitable for high qubit density systems. In this thesis, we focus on the miniaturization of the HERD filtering technology and present two devices which have a reduced size of 32-57% and 47% respectively compared to their predecessor. The first device, implemented in printed circuit board technology, is manufactured and characterized with a resulting insertion loss of more than 40 dB above 80 GHz and an insertion loss of less than 0.29 dB below 8 GHz, measured at 77 K. The results show good agreement between measurements and simulations. In addition, a software is developed for performing eigenmode field decomposition of the filtering structure, which is used to better understand how the field couples to the filtering structure inside the prototype. The insights obtained from simulations and the field decomposition is then used to design the second device, a miniaturized coaxial HERD filter. The results shows that the HERD filtering technique can be made suitable for high qubit density systems where further miniaturization beyond this thesis should be possible.