The Signal
Researchers Demonstrate Optical Control of a ZnSe Donor Spin
A team reports optical initialization, readout and coherent control of a single chlorine-donor electron spin in a ZnSe quantum well. The measurements include Rabi oscillations, Ramsey interference and a charge-limited relaxation time of 18.
Quantum.world Team · Written with AI · fact-checked ·
A single chlorine donor in zinc selenide can now be optically initialized, read out and coherently controlled, according to a paper posted to arXiv on October 27. The measurements connect a donor-bound electron spin to optical transitions in a ZnSe quantum well, and show both spin rotations and Ramsey interference. The authors report a spin-relaxation time of 18.5 ± 3.8 microseconds, limited not by a measured spin flip but by the donor losing its charge.
“Here we demonstrate optical initialization, coherent control, and readout of a single donor-bound electron spin in ZnSe.”
Yuxi Jiang et al.
A chlorine donor becomes an optical spin qubit
The experiment uses a single electron bound to a chlorine donor in a ZnSe quantum well. The researchers placed the doped layer between zinc magnesium selenide barriers and patterned the structure into nanoposts with integrated solid-immersion lenses to collect light from individual emitters. Measurements ran at 3.6 kelvin, with an in-plane magnetic field separating the electron-spin states and producing distinct, spin-selective optical transitions.
Driving one transition pumps the electron into a different spin state; fluorescence on another transition then provides a readout of the population. At the shortest delay in their pump-and-probe sequence, the readout fluorescence was 1.71 times the initialization fluorescence, a contrast the authors interpret as evidence of effective spin preparation. Supplementary analysis estimates that the preparation fidelity approaches 0.95 after long pumping, with residual off-resonant excitation limiting it.
Charge loss, not spin flips, limits the measured lifetime
To measure relaxation, the team prepared the spin, waited in the dark for a variable interval and then used fluorescence to check whether it remained in the target state. The measured relaxation time was 18.5 ± 3.8 microseconds. The observed decay was primarily caused by charge tunneling: the donor’s electron leaves the state needed for the optical measurement before an intrinsic spin flip becomes visible.
The authors therefore set a 90% confidence lower bound of 0.51 milliseconds on the intrinsic spin-flip time. That is not a measured coherence time, and it does not mean the present device preserves a readable spin for that long. It indicates that suppressing charge tunneling could extend the observed relaxation toward a substantially longer spin-flip limit.
Picosecond pulses rotate the spin
For coherent control, the researchers applied off-resonant optical pulses generated from a 2.3-picosecond laser, doubled to a wavelength of 436.3 nanometers. The pulses were red-detuned by 700 gigahertz from the optical transition to reduce unwanted occupation of the excited state. As pulse power increased, the measured spin population oscillated, showing Rabi rotations; the largest fitted rotation reached 5π/2, limited by available laser power.
A pair of pulses produced Ramsey interference, the signature of spin precession between rotations. At a magnetic field of 4 tesla, the observed oscillation period was 16 ± 1 picoseconds, close to the 15.3-picosecond period expected from the measured electron g-factor. The paper also reports an inhomogeneous dephasing time, T2*, of 21.9 nanoseconds from coherent-population-trapping measurements. The Ramsey delay scan extended only to 1.2 nanoseconds, constrained by the travel range of the mechanical delay stage.
Four dark resonances expose the nuclear spin
When two spin-selective optical transitions were driven together, fluorescence fell sharply at two-photon resonance. The authors attribute the dip to coherent population trapping: the two optical fields prepare a spin superposition that no longer couples to the excited state. The dip contrast reached 0.83 ± 0.02, while the measured linewidth and a model of the system yielded the 21.9-nanosecond dephasing estimate.
At stronger optical drive, the dark resonance separated into multiple features. The authors identify four resonances associated with hyperfine interaction between the electron and the chlorine nucleus, which has spin I = 3/2. The result adds a nuclear degree of freedom to the measurements, while the current device’s practical limits remain explicit: a charge-tunneling-limited 18.5-microsecond relaxation measurement and a 21.9-nanosecond inhomogeneous dephasing time.