arXiv:2609.33251 · Nature Communications
High-fidelity geometric quantum gates exceeding 99.9% in germanium quantum dots
Yu-Chen Zhou, Rong-Long Ma, Zhenzhen Kong, Ao-Ran Li, Chengxian Zhang, Xin Zhang, Yang Liu, Hao-Tian Jiang, Zhi-Tao Wu, Gui-Lei Wang, Gang Cao, Guang-Can Guo, Hai-Ou Li, Guo-Ping Guo
Abstract
Achieving high-fidelity and robust qubit manipulations is a crucial requirement for realizing faulttolerant quantum computation. Here, we demonstrate a single-hole spin qubit in a germanium quantum dot and characterize its control fidelity using gate set tomography. The maximum control fidelities reach 97.48%, 99.81%, 99.88% for the I, X/2 and Y /2 gate, respectively. These results reveal that off-resonance noise during consecutive I gates in gate set tomography sequences severely limits qubit performance. Therefore, we introduce geometric quantum computation to realize noiseresilient qubit manipulation. The geometric gate control fidelities remain above 99% across a wide range of Rabi frequencies. The maximum fidelity surpasses 99.9%. Furthermore, the fidelities of geometric X/2 and Y /2 (I) gates exceed 99% even when detuning the microwave frequency by +-2.5 MHz (+-1.2 MHz), highlighting the noise-resilient feature. These results demonstrate that geometric quantum computation is a potential method for achieving high-fidelity qubit manipulation reproducibly in semiconductor quantum computation.
AI Summary
The paper investigates whether nonadiabatic geometric control can reduce sensitivity to resonance-frequency noise in germanium hole-spin qubits. It implements a single-hole spin qubit in a strained-Ge double quantum dot, using electric-dipole spin resonance and enhanced latching readout. The qubit reaches a Rabi frequency of $19$ MHz, with $T_2^*=136$ ns and dynamical-decoupling coherence time up to $6.75\,\mu\mathrm{s}$. Conventional gates are characterized by randomized benchmarking and gate set tomography (GST). GST uses 448 sequences at depth 8 and reports maximum fidelities of $97.48\%$, $99.81\%$, and $99.88\%$ for idle, $X/2$, and $Y/2$, respectively. The substantially poorer idle fidelity is attributed to off-resonance noise accumulated during consecutive idles; randomized benchmarking gives higher idle fidelity, illustrating sequence-dependent noise sensitivity. Three-segment geometric pulse sequences are then implemented with paths targeting either drive-amplitude or off-resonance errors. For the off-resonance-targeted path, GST yields maximum fidelities of $99.98\%$, $99.80\%$, and $99.97\%$ for geometric identity, $X/2$, and $Y/2$. Fidelities remain above $99\%$ across the measured Rabi-frequency range. Deliberate microwave detuning tests show fidelities above $99\%$ out to $\pm2.5$ MHz for $X/2$ and $Y/2$, and $\pm1.2$ MHz for identity, with slower degradation than conventional gates. Noise simulations reproduce the overall trends. The paper emphasizes noise-dependent protocol selection and the trade-off between robustness and duration: the geometric gates take four or six times longer than conventional gates. The amplitude-error-targeted path performs worse in this device, as does a longer path targeting both errors. Two-qubit geometric gates remain future work.
Error found: The reported quality factor Q = 18.04 contradicts the stated definition Q = 2 T₂ᴿᵃᵇⁱ f_Rabi, which gives 36.07 for the quoted coherence time and Rabi frequency. The proposed protection against uncorrelated stochastic dephasing and depolarizing noise is broader than the demonstrated robustness to imposed frequency offsets.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Journal reference
- Nature Communications 16, 7953 (2025)
- DOI
- 10.5281/zenodo.16450274
- Comments
- 13 pages, 4 figures