Abstract: As superconducting quantum processors grow in size and complexity, so must the peripheral
hardware required for the control and readout of such processors. One singular piece of
hardware common to superconducting quantum processors setups is the microwave isolator.
Current microwave isolator technology can be generally understood in the context of timereversal
symmetry breaking via the use of ferrite materials. While generally exhibiting wide (>
GHz) bandwidths and large (>20 dB) directionality, these ferrite-based devices are physically
large with volumes exceeding 15 cm^3. These devices can also introduce uncontrolled magnetic
fields at or near the quantum processor resulting in deleterious effects such as frequency shifts,
excess flux noise, or flux vortex formation. For quantum processors at scale to achieve quantum
advantage, a replacement must be found.
In this talk, I will describe work towards the realization of a superconducting broadband
microwave isolator utilizing DC-SQUIDs. I will detail how, with appropriate application of
microwave flux drives, the non-linear inductance of the SQUIDs allows for power at the signal
frequency travelling in the forward direction to be three-wave mixed and back resulting in
constructive interference and near unity transmission. I will also show how, in the reverse
direction, the same mixing process results in destructive interference and thus suppression of
the signal frequency. Data will be presented on a variety of nanofabricated devices. The data
show excellent model-hardware correlation where directionality greater than 15 dB at
bandwidths approaching 700 MHz with minimal added insertion loss is achieved. Finally, further
extensions of the work will be discussed on how to achieve commercial levels of isolation and
the realization of a fully superconducting replacement of commercial ferrite isolators.