🔬 mesoscale physics

Gatemon Qubit Revisited for Improved Reliability and Stability

This paper addresses the instability and reduced coherence of gate-tunable superconducting transmon qubits (gatemons) by developing characterization methods and demonstrating that specific shunt capacitor designs, particularly grounded configurations, enable highly reliable frequency tuning with 1 MHz precision and improved stability.

Original authors: David Feldstein-Bofill, Zhenhai Sun, Casper Wied, Shikhar Singh, Brian D. Isakov, Svend Krøjer, Jacob Hastrup, András Gyenis, Morten Kjaergaard

Published 2026-02-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: David Feldstein-Bofill, Zhenhai Sun, Casper Wied, Shikhar Singh, Brian D. Isakov, Svend Krøjer, Jacob Hastrup, András Gyenis, Morten Kjaergaard

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to tune an old-fashioned radio to find a specific song. Sometimes, the dial is sticky, or the station drifts in and out of clarity. You turn the knob, and the music is clear; you turn it back, and suddenly it's static. This is the problem scientists face with a specific type of quantum computer part called a Gatemon Qubit.

This paper is like a repair manual and a design upgrade guide for these "quantum radios." Here is the story of what they found and how they fixed it, explained simply.

The Problem: The Sticky, Drifting Radio

Quantum computers use tiny circuits called qubits to do math. To make these qubits useful, scientists need to be able to "tune" them, changing their frequency (like changing the radio station) to make them talk to each other.

For a long time, they used a method involving magnetic fields (like a compass needle) to tune them. It was reliable. But recently, scientists started trying a new method using electric voltage (like turning a volume knob) on a special wire made of semiconductor material. They called this a "Gatemon."

The idea was great, but the reality was messy. The Gatemon had four major "bad habits":

Unreliable Tuning: You turn the knob to a specific setting, but the station (frequency) isn't always the same.
Drifting: Even if you get it tuned perfectly, it slowly drifts away over time, like a clock that runs fast.
The "History" Problem (Hysteresis): If you turn the knob up to a setting, you get one station. If you turn it down to that exact same setting, you get a different station. The qubit remembers where it came from.
Fragility: The quantum information (the "song") fades away too quickly.

The Experiment: Two Different Designs

The researchers decided to test two different ways of building the Gatemon to see which one behaved better. Imagine they were building two different types of houses to see which one stayed warmer in the winter.

The "Grounded" House: In this design, the electrical "island" where the qubit lives is directly connected to the ground (the earth). It has a solid, stable reference point. Think of this as a house built on a massive, unshakeable concrete foundation.
The "Floating" House: In this design, the island is separated from the ground. It's like a house built on a raft in the middle of a lake. It's not anchored down; it floats.

The Discovery: Stability Wins

They ran thousands of tests, turning the voltage knobs back and forth and watching the qubits for hours. Here is what they found:

1. The "Grounded" Design is the Rock-Solid Winner
The "Grounded" qubit was incredibly reliable. When they turned the voltage knob, the frequency stayed exactly where they put it. They could tune it over a huge range (several billion cycles per second) with a precision of just 1 MHz. That's like tuning a radio so precisely you can hear a whisper in a hurricane.

The Floating Design: The "Floating" qubit was jittery. Its frequency jumped around wildly, and the "history problem" (hysteresis) was much worse. It was like trying to tune a radio while standing on a boat in choppy water.

2. The Drift is About Frequency, Not the Knob
They discovered that the qubit's stability didn't depend on how sensitive the knob was, but rather on what frequency the qubit was playing.

High Frequencies: When the qubit was playing a "high note," it was very stable, even in the floating design (though the grounded one was still better).
Low Frequencies: When the qubit dropped to a "low note," things got chaotic. The frequency would jump by huge amounts.
The Analogy: Imagine a tightrope walker. If they are high up (high frequency), the wind doesn't bother them much. If they are low down (low frequency), a tiny breeze (noise) knocks them off balance. The researchers realized that at low frequencies, the "channels" carrying the electricity become too few, making the system sensitive to tiny electrical noise.

3. The "Memory" Problem
They found that the "history problem" (hysteresis) could be solved by being careful. If you start your tuning in a "safe zone" and end in a "safe zone," the qubit behaves perfectly. But if you wander into the "unsafe zone" (usually at low frequencies), the qubit gets confused about where it is.

4. Coherence (How long the song lasts)
Finally, they checked how long the qubit could hold onto its information.

Relaxation (T1): Both designs held the note for about the same amount of time.
Dephasing (T2): This is where the "Grounded" design won big. The "Floating" design was very sensitive to low-frequency noise (like a faint hum in the background), causing the signal to scramble three times faster than the grounded one. The "Grounded" design blocked this noise much better.

The Conclusion: A New Standard

The paper concludes that if you want to build a reliable quantum computer using these Gatemon qubits, you should ground your design.

By connecting the qubit's island directly to the ground, you create a stable reference point that stops the qubit from drifting, jumping, or getting confused by its own history. This allows scientists to tune these quantum devices with the precision needed to build the super-computers of the future.

In short: They took a wobbly, unpredictable quantum radio, figured out that it needed a solid foundation, and built a "Grounded" version that stays perfectly in tune. This is a huge step forward for making quantum computers practical and reliable.

1. Problem Statement

Gate-tunable superconducting transmon qubits, known as gatemons, utilize hybrid superconductor-semiconductor (S-Sm-S) Josephson junctions (typically InAs nanowires with epitaxial Al). While they offer a promising alternative to flux-tunable qubits by using electrostatic gates rather than magnetic fields to tune Josephson energy, they suffer from four critical limitations that hinder scalability:

Unreliability: The qubit frequency ( $f_q$ ) vs. gate voltage ( $V_g$ ) relationship is inconsistent; repeated sweeps yield different frequencies.
Instability: The qubit frequency drifts or jumps over time.
Hysteresis: The qubit frequency depends on the sweep direction (up vs. down) of the gate voltage.
Reduced Coherence: Relaxation times ( $T_1$ ) are often shorter than in standard SIS transmons, and dephasing is significant.

The authors aim to systematically characterize these issues and determine if specific capacitor geometries can mitigate them.

2. Methodology

The study focuses on InAs nanowire gatemon qubits fabricated on a single chip using two distinct capacitor geometries:

Grounded Design: The capacitor island is directly connected to the ground plane through the junction, providing a well-defined reference potential.
Floating Design: The capacitor pads are separated from the ground plane (floating island).

Experimental Approach:

Reliability Testing: Performed 10 back-to-back gate voltage sweeps for both designs, recording the qubit frequency via two-tone spectroscopy.
Stability Testing: Monitored the qubit frequency continuously for 6 hours at various operating points (both "sweet spots" where $\partial f_q/\partial V_g \approx 0$ and sensitive slope regions).
Hysteresis Analysis: Swept gate voltages in both positive and negative directions to map frequency differences based on sweep history.
Coherence Measurements: Measured energy relaxation time ( $T_1$ ), Ramsey dephasing time ( $T_{2,R}$ ), and Hahn-Echo dephasing time ( $T_{2,E}$ ) for both designs at similar frequencies.

3. Key Contributions

Systematic Characterization: Established robust protocols to quantify reliability, stability, and hysteresis in gatemon devices, distinguishing between frequency-dependent behaviors and design-dependent behaviors.
Design Comparison: Directly compared grounded vs. floating capacitor architectures on the same chip to isolate geometric effects on performance.
Identification of Noise Sources: Linked frequency instability and dephasing to low-frequency charge noise, demonstrating that the floating design is significantly more susceptible to this noise than the grounded design.
Operational Protocols: Developed strategies to minimize hysteresis by ensuring both start and end points of gate sweeps lie within "reliable frequency zones."

4. Key Results

A. Reliability (Frequency vs. Gate Voltage)

Grounded Design: Shows significantly higher reliability at high frequencies ( $>5.1$ GHz). The standard deviation of frequency across 10 sweeps is $\approx 0.68$ MHz. However, reliability degrades at lower frequencies.
Floating Design: Exhibits poor reliability across the entire range with a standard deviation of $\approx 6.08$ MHz (an order of magnitude worse than the grounded design at high frequencies).
Charge Jumps: Both designs exhibit discrete charge jumps, but the floating design shows larger, less predictable variations.

B. Stability (Time Domain)

Grounded Design: Highly stable at high frequencies ( $>5$ GHz) with no observable jumps or drifts over 6 hours. At low frequencies ( $<5$ GHz), it suffers from large discrete jumps (up to 1 GHz change) and drifts.
Floating Design: Shows consistent frequency drift and discrete jumps across the entire frequency range, regardless of whether the qubit is at a sweet spot or a sensitive slope.
Conclusion: Stability is dominated by the absolute qubit frequency rather than the sensitivity to the gate voltage ( $\partial f_q/\partial V_g$ ). Lower frequencies (lower Josephson energy) correlate with fewer Andreev channels, making the critical current more sensitive to noise.

C. Hysteresis

Hysteresis is observed when gate sweeps end in "unstable frequency zones."
If the sweep starts and ends within a "reliable zone" (typically $>5$ GHz for the grounded design), hysteresis is minimized, and the up/down sweep standard deviation drops below 1 MHz.
The downward sweep (towards negative voltage) generally shows higher reliability than the upward sweep.

D. Coherence (T1 and T2)

Relaxation ( $T_1$ ): Both designs show similar $T_1$ times (ranging from $\approx 2.4$ $\mu$ s at 5 GHz to $\approx 8$ $\mu$ s at 3 GHz). This suggests the limiting loss mechanism is intrinsic to the S-Sm-S junction rather than surface dielectric loss (which differs between grounded and floating geometries).
Dephasing ( $T_2$ ):
- Grounded: $T_{2,R} \approx 1.4$ $\mu$ s.
- Floating: $T_{2,R} \approx 0.5$ $\mu$ s (3x shorter).
- Hahn-Echo ( $T_{2,E}$ ): Both designs show similar $T_{2,E} \approx 2$ $\mu$ s.
Interpretation: The floating design is significantly more sensitive to low-frequency noise (which Ramsey sequences detect but Hahn-Echo filters out). The grounded design effectively shields the qubit from this noise.

5. Significance and Conclusion

This work provides a critical roadmap for improving hybrid superconducting-semiconductor qubits:

Grounded Superiority: The grounded capacitor design is superior for reliability and stability, particularly at high operating frequencies, offering frequency tuning precision within 1 MHz over a multi-GHz range.
Noise Mitigation: The results confirm that grounding the capacitor island significantly reduces sensitivity to low-frequency charge noise, directly improving Ramsey dephasing times.
Scalability: By identifying "reliable zones" and optimal sweep protocols, the authors provide a method to operate gatemons with the stability required for scalable quantum information processing.
Future Outlook: The study suggests that while S-Sm-S junctions have intrinsic loss mechanisms (limiting $T_1$ ), their coherence can be maximized by optimizing device geometry (grounding) and operating at higher frequencies where the number of Andreev channels is sufficient to stabilize the critical current.

The paper concludes that with these design and operational improvements, gatemons can become a viable, stable platform for advanced hybrid quantum circuits.