Can ordinary silicon host fast, accurate spin qubits?
A carefully designed tiny magnet let electron spins in ordinary (non-purified) silicon be flipped fast enough and cleanly enough to pass the error threshold for fault-tolerant quantum computing.
Source
A fault-tolerant addressable spin qubit in a natural silicon quantum dot
Study at a glance
- Design
- Other — Low-temperature device experiment on a Si/SiGe double quantum dot with a cobalt micromagnet, driven by electric dipole spin resonance
- N
- One device; most data are from the left dot, with the right dot showing similar results in the supplement
- Population
- Single electron spins in a natural-silicon Si/SiGe double quantum dot at about 120 mK
- Outcome
- Rabi frequency, Rabi decay time, dephasing time T2*, quality factor Q and single-qubit Clifford gate fidelity
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The team built a double quantum dot in a natural Si/SiGe heterostructure, each dot holding one electron, and placed a cobalt micromagnet on top to create a strong magnetic field gradient. Microwaves applied to a gate shake the electron in this gradient, which acts like an oscillating magnetic field that rotates the spin (electric dipole spin resonance). They measured resonance spectra, Rabi oscillations, Ramsey fringes, how driving strength affects speed and decay, and finally gate fidelity by randomized benchmarking.
What they found
The two dots' resonances were split by about 800 MHz, so each spin could be addressed with negligible crosstalk (about 0.02% on the idle qubit). The dephasing time T2* was 1.83 μs, the longest reported for natural-silicon dot qubits, and Rabi frequencies reached about 35 MHz at maximum drive. The best trade-off between speed and decay was near a 10 MHz Rabi frequency with a quality factor of about 140, where the average single-qubit Clifford gate fidelity was 99.6%. Stronger driving shortened the decay time, which the authors attribute mainly to microwave heating.
The limits
What it doesn't show
Only single-qubit operations are shown; a two-qubit gate, which is needed for real computation, is proposed but not demonstrated. The results come from one device, and detailed data mostly from one of the two dots because the charge sensor was more sensitive to it. The heating explanation for faster decay at high drive is inferred rather than measured directly, and nuclear-spin noise from natural silicon still limits dephasing.
Key terms
- Rabi oscillation
- The periodic flipping of a two-level system (here a spin) between states while it is driven on resonance; its frequency sets how fast a gate can be performed.
- T2* (dephasing time)
- How long a spin keeps a well-defined phase when left alone, measured with a Ramsey sequence; limited here by fluctuating nuclear spins and charge noise.
- Electric dipole spin resonance
- Rotating a spin by moving the electron back and forth in a magnetic field gradient with an electric (gate) signal, instead of applying an oscillating magnetic field directly.
- Randomized benchmarking
- Applying long random sequences of Clifford gates and measuring how fast the final-state fidelity decays, giving an average error per gate that is insensitive to preparation and readout errors.
- Quality factor Q
- The Rabi decay time divided by the time for a π rotation; roughly how many gate operations fit before the qubit decays, which bounds the fidelity.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What is the main advantage of the micromagnet approach in this study?
Common questions
Why is natural silicon harder to use than purified silicon?
Natural silicon contains a few percent of the silicon-29 isotope, whose nuclear spins create a fluctuating magnetic field that dephases the electron spin quickly. Purified silicon removes this, but it is rare and expensive.
How does a micromagnet help if it doesn't remove the noise?
It does not lengthen coherence much; instead its field gradient lets the spin be rotated much faster, so many more operations fit inside the short coherence time, raising Q and fidelity.
Why not just drive the qubit as hard as possible?
Above a certain microwave amplitude the Rabi decay time drops sharply (likely heating), so Q and fidelity fall; there is an optimum drive level.
More on Qubits and quantum devices