A spin qubit, in which quantum information is encoded in the spin state of an electron, is one of the most promising platforms for quantum computing. Spin qubits exhibit long coherence times and are compatible with advanced semiconductor manufacturing technologies. The leading implementation of spin qubits involves confined electrons inside quantum dots, a nanoscale semiconductor architecture that behaves like a controllable artificial atom. Recent advances have enabled high-fidelity operation of single- and two-qubit gates, exceeding the threshold required for certain surface code quantum error correction techniques.
However, to achieve practical fault-tolerant quantum computing, the variability issues of spin qubit gates must be addressed. A key challenge in this context is fluctuations in the qubit resonance frequency caused by microscopic noise sources. A constant qubit resonance frequency (fq), also known as "Larmor frequency," is needed for effective qubit operation. Recent studies have shown that microwave signals used to control qubits can generate heat that shifts the fq. Specifically, the fq exhibits a sharp increase at low temperatures, followed by a gradual decrease at higher temperatures. This non-monotonic temperature dependence disrupts resonance, thus deteriorating gate fidelity. Surprisingly, previous research has shown that a higher temperature of 200 millikelvin (mK), rather than the standard temperature of 20 mK, can mitigate the effect of fq shift on gate fidelity. Despite the importance of this phenomenon, its microscopic origin has remained unclear.
In a recent study, a collaborative research team from Tokyo University of Science (TUS) and the National Institute of Advanced Industrial Science and Technology (AIST), Japan, led by Professor Takayuki Kawahara from the Department of Electrical Engineering at TUS, has finally clarified the noise mechanisms that affect silicon spin qubit performance. By combining theoretical modeling with large-scale statistical simulations of charge noise arising from two-level fluctuators (TLFs), they demonstrated how higher temperatures can improve gate fidelity.
"Several candidates have been proposed to explain the origin of the qubit or Larmor frequency shift," explains Prof. Kawahara. "Among them, the charge-noise model seems to be most promising as it can reproduce key features of fq shift. In this study, we focused on the charge noise model to elucidate the origin of the temperature dependence of fq shift and to analyze qubit fabrication approaches that can alleviate its effect on gate fidelity." Their study was published in Volume 14 of the journal IEEE Access on May 04, 2026.
The team developed a spin qubit model in which electrons were confined within a quantum dot formed in a silicon/silicon-germanium (Si/SiGe) double heterostructure. Electron spins were manipulated using microwave control under an externally applied magnetic field gradient. Using this framework, the researchers statistically simulated the effects of numerous TLFs located near the semiconductor/oxide interface.
They systematically varied a wide range of TLF parameter settings, including spatial distributions, activation-energy distributions, minimum transition times, and the temperature dependence of switching times. In total, the team evaluated 108 parameter sets, each containing 5,000 randomly generated TLF configurations.
For each parameter set, they then calculated qubit frequency shifts and analyzed the temperature dependence and the fidelity of the X quantum gate. Their analysis showed that the experimental observations were best reproduced when TLF activation energies followed an exponential distribution, minimum switching times were short, and switching rates exhibited strong temperature dependence. Under these conditions, the model successfully reproduced the experimentally observed non-monotonic temperature dependence of the qubit frequency shift. Gate fidelity simulations further showed that the fidelity improvement at 200 mK occurs when transition times are much shorter than the gate times and parameters exhibit a steep temperature transition.
Importantly, based on these findings, the researchers concluded that electronic transitions between the conduction band and trap states (which involve generation/recombination or band–edge trap processes) are the most likely origin of the relevant TLFs and associated qubit frequency shifts, rather than slower atomic-scale structural motion. This finding provides new insight into the microscopic origin of charge noise in silicon spin qubits.
"Our findings highlight the importance of controlling semiconductor/oxide interface trap states and adopting fabrication procedures that stabilize qubit frequencies in improving gate fidelities for future large-scale silicon quantum processors," remarks Prof. Kawahara. "This could contribute significantly to the development of practical large-scale quantum computers with reduced noise."
Overall, this study provides important insights for improving spin qubit gate performance, bringing us closer to realizing large-scale fault-tolerant quantum computing.
