Prof. John Morton,UCL, London Centre for Nanotechnology, UK
Donor spins in silicon get down to bismuth
The electron and nuclear spins of (Group V) donors in silicon have been proposed as potential quantum bits, or qubits . The past few years have seen several breakthroughs towards this goal, including the measurement of the spins of single donor atoms, with high fidelity, in nanoelectronic devices . Recent studies have moved down Group V from more commonly studied donors, such as phosphorus, to bismuth which has been found to have several interesting properties, deriving from the large (I = 9/2) nuclear spin of 209Bi and large (A = 1.47 GHz) hyperfine coupling between the electron and nuclear spins of the donor.
First, Bi donors in silicon possess so called ‘clock’ or ‘ZEFOZ’ transitions whose frequencies have zero first-order dependence on the magnetic field, making them robust to many sources of decoherence. At these points, coherence times up to 3 seconds have been measured for the electron spin in isotopically-enriched 28Si. For Bi donor spins in natural silicon, dynamical decoupling methods can be used to observed the changing nature of the 29Si nuclear spin bath around as the clock transition is approached, while electron spin coherence times approaching 1 second can be measured .
Second, the large nuclear spin and hyperfine coupling yield allowed transition frequencies of approximately 7.3 GHz at zero applied magnetic field, making Bi donors in silicon of interest for coupling to superconducting aluminium circuits such as resonators and qubits. This has enabled studies of relatively small spin ensembles (~10,000 donors) coupled to micron-scale superconducting microwave resonators with Q-factors of approximately 100,000 . In this regime, spin-relaxation by spontaneous emission can be enhanced by the cavity (by the Purcell effect ) to the point where it becomes the dominant relaxation mechanism . We expect such cavity-induced spin relaxation will be important in future studies of ESR at mK temperatures, while the ability to accelerate relaxation of particular transitions may also have applications in dynamical nuclear polarisation.
 B.E. Kane. Nature 393, 133 (1998)
 J.J. Pla et al., Nature 489 541-545 (2012); J.J. Pla et al., Nature 496 334 (2013)
 G. Wolfowicz et al., Nature Nano. 8 561 (2013); W.-L. Ma et al., Phys Rev B 92 161403 (2015)
 A. Bienfait et al., Nature Nano. 11 253 (2015)
 E. M. Purcell, Phys. Rev. 69, 681 (1946)
 A. Bienfait et al., Nature 531 74 (2016)