Hyperfine dephasing of electron spin qubits in GaAs quantum dots
Final Report Abstract
A key concept that distinguishes quantum from classical physics is quantum coherence, which refers to the ability to sustain quantum superposition states. It is also an essential ingredient for quantum technologies. As a quantum system interacts with an environment, quantum coherence is gradually lost on the time scale of the so called coherence time. This decoherence process leads to the transition from quantum to classical behavior. This project investigated several aspects of decoherence for qubits based on electrostatically defined quantum dots, a candidate system for the realization of quantum computing. The specific implementation considered is based on a GaAs/AlGaAs heterostructure and uses two electron spins to encode a single qubit. On the one hand, GaAs-based qubits have the advantage of a good reproducibility. On the other hand, the hyperfine interaction of the electron with nuclear spins complicates high performance qubit operation. Improving our understanding of the hyperfine interaction was one of the main topics of this project. A common method to extend the coherence time is the so-called Hahn echo, which is based on manipulating the qubit state such that any decoherence effects arising from slow fluctuations cancel out. It is known to be very effective for our qubits. We have shown that the detailed performance of this method is substantially affected by quadrupolar coupling of the nuclear spin to electric field gradients in the host material, which are determined via the material response to electric fields. As the effect depends on the orientation of the magnetic fields, we were able to show that rotating the field direction can enhance the coherence time by more than a factor of two. At the same time, we discovered that a g-factor anisotropy of a few percent leads to collapses and revivals of the Hahn echo amplitude for the newly explored magnetic field orientation. We developed a semiclassical model for both effects that is in excellent agreement with the data. A second set of experiments exploiting this g-factor anisotropy explored whether the back-action from a single electron to the surrounding nuclear spin bath plays a role in the decoherence process. This question is of fundamental and practical interest as the laws of quantum mechanics imply that such a back-action is unavoidable, while the most widely used models for decoherence neglect any back-action. The system we investigated is an interesting model system as its environment of about 106 nuclear spins is large but still finite. All previous decoherence experiments were in good agreement with models neglecting any back action. Using a technique that correlates the outcome of subsequent measurements, we unambiguously detected the presence of a back action on the time scale of the qubit’s coherence time, thus identifying a limit of classical models. Another key requirement for qubits is their accurate manipulation. For the type of qubits considered here, it is common to manipulate them via the exchange interaction between the two electron spins. This exchange interaction is also of high relevance for two-qubit operations for spin qubits. Its strength can be controlled via gate voltages applied to the qubit. A further main result of the project relates to the detailed form of the dependence of the exchange coupling on the gate voltages over the full relevant range of the latter, leading to a variation of by three orders of magnitude. We found large deviations from well-known models, and develop an extended Hubbard model including excited orbital states that shows a much better agreement. Finally, we explored a new method to determine high frequency charge noise in the GHz range, which is also relevant for qubit coherence and accurate control, but so far has not been characterized systematically for quantum dot qubits. The idea is to measure the qubit relaxation time as a function of the level splitting, which can be varied by polarizing nuclear spins. We implemented the corresponding measurement protocols and verified by modelling that the method could produce useful results, but did not complete the experiments so far.
Publications
- “Quadrupolar and anisotropy effects on dephasing in two-electron spin qubits in GaAs”; Nat. Commun. 7, 11170 (2016)
T. Botzem, R. P. G. McNeil, J.-M. Mol, D. Schuh, D. Bougeard, and H. Bluhm
(See online at https://doi.org/10.1038/ncomms11170) - “Tuning Methods for Semiconductor Spin Qubits”, Phys. Rev. Applied 10, 054026 (2018)
Tim Botzem, Michael D. Shulman, Sandra Foletti, Shannon P. Harvey, Oliver E. Dial, Patrick Bethke, Pascal Cerfontaine, Robert P. G. McNeil, Diana Mahalu, Vladimir Umansky, Arne Ludwig, Andreas Wieck, Dieter Schuh, Dominique Bougeard, Amir Yacoby, Hendrik Bluhm
(See online at https://doi.org/10.1103/PhysRevApplied.10.054026) - A High-Fidelity Gateset for Exchange-Coupled Singlet-Triplet Qubits
Pascal Cerfontaine, René Otten, M. A. Wolfe, Patrick Bethke, Hendrik Bluhm
(See online at https://doi.org/10.1103/PhysRevB.101.155311) - Closed-loop control of a GaAs-based singlet-triplet spin qubit with 99.5% gate fidelity and low leakage
Pascal Cerfontaine, Tim Botzem, Julian Ritzmann, Simon Sebastian Humpohl, Arne Ludwig, Dieter Schuh, Dominique Bougeard, Andreas D. Wieck, Hendrik Bluhm
(See online at https://doi.org/10.1038/s41467-020-17865-3) - Coherent hyperfine back-action from single electrons on a mesoscopic nuclear spin bath
Patrick Bethke, Robert P. G. McNeil, Julian Ritzmann, Tim Botzem, Arne Ludwig, Andreas D. Wieck, Hendrik Bluhm
(See online at https://doi.org/10.1103/PhysRevLett.125.047701) - “A machine learning approach for automated fine-tuning of semiconductor spin qubits”, Appl. Phys. Lett. 114, 133102 (2019)
Julian D. Teske, Simon Sebastian Humpohl, René Otten, Patrick Bethke, Pascal Cerfontaine, Jonas Dedden, Arne Ludwig, Andreas D. Wieck, Hendrik Bluhm
(See online at https://doi.org/10.1063/1.5088412)
