Gatemon Qubit Revisited for Improved Reliability and Stability
The development of quantum circuits based on hybrid superconductor-semiconductor Josephson junctions holds promise for exploring their mesoscopic physics and for building novel superconducting devices. The gate-tunable superconducting transmon qubit (gatemon) is the paradigmatic example of such a superconducting circuit. However, gatemons typically suffer from unstable and hysteretic qubit frequencies with respect to the applied gate voltage and reduced coherence times. Here we develop methods for characterizing these challenges in gatemons and deploy these methods to compare the impact of shunt capacitor designs on gatemon performance. Our results indicate a strong frequency- and design-dependent behavior of the qubit stability, hysteresis, and dephasing times. Moreover, we achieve highly reliable tuning of the qubit frequency with 1 MHz precision over a range of several GHz, along with improved stability in grounded gatemons compared to gatemons with a floating capacitor design.
💡 Research Summary
This paper addresses four critical limitations of gate‑tunable superconducting transmon qubits (gatemons) built from hybrid superconductor‑semiconductor Josephson junctions: (1) unreliable qubit frequency versus gate voltage, (2) temporal instability of the frequency, (3) hysteresis depending on sweep direction, and (4) reduced coherence times compared with conventional flux‑tunable transmons. Using InAs nanowires with epitaxial Al shells to form S‑Sm‑S junctions, the authors fabricate two gatemon variants that differ only in the geometry of the shunt capacitor: a grounded design where the capacitor island is directly tied to ground through the junction, and a floating design where the capacitor pads are isolated from the ground plane. Both designs are engineered to have comparable charging energies, allowing a direct comparison of the impact of grounding on performance.
Reliability is quantified by repeating a full gate‑voltage sweep ten times and extracting the qubit transition frequency (f_q(V_g)) from two‑tone spectroscopy at each step. In the high‑frequency regime (>5 GHz), the grounded device exhibits an average standard deviation of 0.68 MHz, an order of magnitude better than the floating device’s 6.08 MHz. At lower frequencies the grounded device shows occasional charge‑induced jumps (e.g., at (V_g = -1.02) V and (-0.1) V) that increase the spread, whereas the floating device displays a relatively flat but larger variance across the whole range. This suggests that the lack of a well‑defined ground reference in the floating geometry makes the qubit more susceptible to ambient charge noise.
Stability over time is examined by fixing the gate voltage at several points—both at sweet spots (where (\partial f_q/\partial V_g \approx 0)) and at points with strong voltage sensitivity—and monitoring the frequency continuously for six hours. In the grounded design, frequencies above 5 GHz remain essentially constant, with no observable jumps or drifts, while frequencies below 5 GHz can drift by up to 1 GHz over the measurement period, often triggered by discrete charge jumps. The floating design shows MHz‑scale drifts and small jumps throughout the entire measured band (<7 GHz), indicating that temporal stability is more strongly linked to the absolute transition frequency than to the local gate‑sensitivity.
Hysteresis is probed by sweeping the gate voltage downwards and upwards repeatedly across the same voltage interval and comparing the resulting (f_q) traces. In the grounded device, when the sweep terminates inside a “reliable” frequency region, the up‑ and down‑sweeps overlap within a few MHz, indicating negligible hysteresis. However, when the final point lies in an “unstable” region, the two directions diverge by tens of MHz, revealing a pronounced hysteresis that likely originates from charge rearrangements in the semiconductor channel. The floating device exhibits smaller but still measurable hysteresis across its entire range, consistent with its weaker grounding.
Coherence measurements reveal that both designs achieve long energy‑relaxation times ((T_1) ≈ 20–30 µs) and dephasing times ((T_2^) ≈ 10–15 µs) at frequencies above 5 GHz, comparable to state‑of‑the‑art flux‑tunable transmons. Below this threshold, (T_1) drops dramatically to a few microseconds, and (T_2^) follows suit, underscoring the detrimental impact of low‑frequency charge noise and gate‑induced fluctuations on decoherence.
Overall, the study demonstrates that grounding the shunt capacitor dramatically improves frequency reliability, reduces hysteresis, and stabilizes the qubit over time in the high‑frequency regime, while the floating design suffers from larger variance and drift across the board. The authors achieve sub‑MHz tuning precision over several gigahertz, establishing gatemons as a viable alternative to flux‑tunable qubits for scalable quantum processors, provided that careful engineering mitigates low‑frequency charge noise and hysteresis. Future work is suggested to incorporate advanced charge‑noise filtering, optimized dielectric environments, and systematic material improvements to extend the high‑performance window down to lower frequencies, thereby fully unlocking the potential of gate‑tunable superconducting qubits.
Comments & Academic Discussion
Loading comments...
Leave a Comment