Optimizing superconducting quantum gate operations

Abstract

Frequency overcrowding remains a significant barrier to the scalability of superconducting transmon qubits in quantum computing architectures. When resonance frequencies of neighboring qubits are too close, crosstalk and coherence loss degrade overall gate fidelity. To address this, post-fabrication frequency calibration methods such as Alternating-Bias Assisted Annealing (ABAA) have been developed, which leverage controlled voltage pulses to fine-tune the resistance of the Josephson junctions that define the qubit’s frequency. This study investigates resistance-based tuning using ABAA with a focus on the effects of pulse parameters—amplitude, width, and polarity—on resistance adjustment in Al/AlOx/Al tunnel junctions. A comprehensive experimental setup was implemented using a source-measure unit (SMU), probe station, and oscilloscope to control and monitor electrical pulses and measure resistance changes over time. Both unipolar and alternating bias pulse profiles were applied to study their effects on resistance modulation. The setup also incorporates heating control, static shielding, and real-time automation via Python. We demonstrate consistent resistance changes exceeding 70%, with observable trends linked to pulse characteristics and temperature. Resistance measurements and pulse profiles were visualised in real time to assess signal integrity and electrical noise, offering insight into measurement artifacts and procedural limitations. Our findings reinforce ABAA’s potential as a scalable post-fabrication tuning method to mitigate frequency overcrowding and improve coherence in quantum processors.