Skip to content

Quantum tomography of an entangled three-qubit state in silicon

MetadataDetails
Publication Date2021-06-07
JournalNature Nanotechnology
AuthorsKenta Takeda, Akito Noiri, Takashi Nakajima, Jun Yoneda, Takashi Kobayashi
InstitutionsRIKEN Center for Emergent Matter Science
Citations101
AnalysisFull AI Review Included

Quantum Tomography of an Entangled Three-Spin State in Silicon: Engineering Analysis

Section titled “Quantum Tomography of an Entangled Three-Spin State in Silicon: Engineering Analysis”
  • Core Achievement: Successful generation and characterization of a three-qubit Greenberger-Horne-Zeilinger (GHZ) entangled state using electron spins confined in a silicon/silicon-germanium (Si/SiGe) triple quantum dot array.
  • High Fidelity: The GHZ state achieved a state fidelity of 0.880 ± 0.007 (88.0%) via quantum state tomography, comparable to initial demonstrations in superconducting platforms.
  • Genuine Entanglement Witnessed: The measured state violates the Mermin-Bell inequality (M = 3.47 ± 0.05), confirming genuine GHZ-class multipartite entanglement that is not biseparable.
  • High Single-Qubit Performance: Average single-qubit control fidelities were benchmarked at 99.43% (Q1), 99.57% (Q2), and 99.91% (Q3).
  • Noise Mitigation Strategy: A decoupled CZ gate sequence, incorporating π pulses (Hahn echo), was implemented to effectively decouple the qubits from quasi-static low-frequency noise (nuclear magnetic and charge noise) during entanglement operations.
  • Platform Validation: This result demonstrates the potential of the silicon quantum dot platform for scalable multi-qubit quantum algorithms and is a crucial step toward implementing quantum error correction (QEC).
ParameterValueUnitContext
GHZ State Fidelity (FGHZ)0.880 ± 0.007DimensionlessMeasured via quantum state tomography.
Mermin-Bell Witness (M)3.47 ± 0.05DimensionlessViolates the biseparable limit (M < 2) by > 28 standard deviations.
Single-Qubit Fidelity (Q1)99.43%Clifford-based randomized benchmarking.
Single-Qubit Fidelity (Q2)99.57%Clifford-based randomized benchmarking.
Single-Qubit Fidelity (Q3)99.91%Clifford-based randomized benchmarking.
Two-Qubit Bell State Fidelity94.1%Average fidelity for Q2 and Q3 Bell states.
T1 Relaxation Time (Q1)4.30msSpin relaxation time.
T1 Relaxation Time (Q3)1.31msSpin relaxation time (shortest observed).
T2* Dephasing Time (Q1)1.82”sInhomogeneous dephasing time.
T2 Echo Time (Q3)45.8”sHahn echo extended coherence time (longest observed).
External Magnetic Field (Bext)0.5275TApplied in-plane field.
Zeeman Splitting~18GHzEnergy separation between spin states.
Base Electron Temperature40mKOperating temperature in dilution refrigerator.
Single-Qubit Rabi Frequency (fRabi)6MHzMaximum single-qubit drive speed.
Exchange Coupling (J12)2.8MHzNominal coupling between Q1 and Q2.
Exchange Coupling (J23)12.5MHzNominal coupling between Q2 and Q3.
Qubit Frequency Separation (ÎŽE12)435.4MHzSeparation due to micro-magnet field gradient.
Low-Frequency Charge Noise (S(f=1 Hz))0.2”V/sqrt(Hz)Measured effective charge noise in the device.
Micro-magnet StackTi/Co/Al10/250/20 nmFilm thicknesses used to create field gradient.
  1. Substrate: Isotopically natural, undoped Si/SiGe heterostructure wafer.
  2. Quantum Dot Formation: Nanofabricated overlapping aluminum gates (three layers) define the triple quantum dot confinement potential.
  3. Qubit Encoding: Each dot hosts one electron, using its spin-up and spin-down states as the qubit basis (Q1, Q2, Q3).
  4. Magnetic Field Gradient: A cobalt micro-magnet placed on top of the array generates a local Zeeman gradient, enabling fast, addressable Electric-Dipole Spin Resonance (EDSR) control.
  5. Cooling: Sample cooled in a dry dilution refrigerator to a base electron temperature of 40 mK.
  • Initialization/Readout (Q1, Q3): Performed via energy-selective tunneling with neighboring electron reservoirs.
  • Initialization/Readout (Q2): Achieved indirectly using a combination of resonant SWAP gates (with Q1 or Q3) and energy-selective tunneling.
  • Single-Qubit Control: Implemented using EDSR, leveraging the micro-magnet field gradient for frequency addressability (separations > 400 MHz).
  • Two-Qubit Control (CZ Gates): Controlled phase (CZ) gates utilize the exchange coupling (Jij) between neighboring spins, controlled by fast gate voltage pulses (virtual gate technique) to maintain the symmetric operation point and minimize charge-noise-induced dephasing.
  • Entanglement Protocol: The GHZ state was generated using a sequence of single-qubit rotations (Y/2) and decoupled CZ gates. The decoupled sequence separates the CZ operation into two √CZ gates with π pulses inserted, acting as a Hahn echo to suppress quasi-static phase noise.
  • Readout Error Correction: Measured spin-up and spin-down readout fidelities (F↑i, F↓i) were used to correct the raw measured probabilities (PM).
  • Quantum State Tomography (QST): The density matrix (ρ) was reconstructed using maximum likelihood estimation based on 27 combinations of single-qubit pre-rotations (I, X/2, Y/2)⊗3.
  • Error Analysis: Errors for QST results were obtained using a Monte-Carlo method assuming multinomial distributions for single-shot probabilities.

The successful demonstration of high-fidelity, multipartite entanglement in a silicon platform is critical for the development and scaling of fault-tolerant quantum systems.

  • Fault-Tolerant Quantum Computing: The GHZ state is essential for implementing Quantum Error Correction (QEC) protocols, which are necessary to manage error accumulation in large-scale quantum processors.
  • Scalable Quantum Processors: Silicon quantum dots are highly compatible with standard CMOS manufacturing techniques, offering a promising route for high-density integration and mass production of qubits.
  • Quantum Simulation: Multi-qubit arrays enable the simulation of complex quantum phenomena, such as many-body physics and material properties, which are intractable for classical computers.
  • Quantum Sensing: The high coherence times (T2 echo up to 45.8 ”s) and precise spin control are foundational for advanced quantum sensors that utilize entangled states for enhanced sensitivity.
  • High-Fidelity Gate Development: The techniques developed for noise mitigation (decoupled CZ gates) are directly applicable to improving gate fidelity in other solid-state qubit architectures.