Stark shift and field ionization of arsenic donors in 28Si-silicon-on-insulator structures

Abstract

We develop an efficient back gate for silicon-on-insulator (SOI) devices operating at cryogenic temperatures, and measure the quadratic hyperfine Stark shift parameter of arsenic donors in isotopically purified 28 Si-SOI layers using such structures. The back gate is implemented using MeV ion implantation through the SOI layer forming a metallic electrode in the handle wafer, enabling large and uniform electric fields up to ∼ 2 V/µm to be applied across the SOI layer. Utilizing this structure we measure the Stark shift parameters of arsenic donors embedded in the 28 Si SOI layer and find a contact hyperfine Stark parameter of η a = −1.9 ± 0.2 × 10 −3 µm 2 /V 2 . We also demonstrate electric-field driven dopant ionization in the SOI device layer, measured by electron spin resonance.

Hyperfine Stark effect of shallow donors in silicon

“…While specific measurements of these thresholds are still lacking, very recent experimental work [25] reports that Si:As donor electrons are ionized at E ∼ 2 V/µm, in full agreement with our prediction. Predictions of the size of the electric field required to ionize each donor species, and the corresponding maximum absolute hyperfine shift ∆A max that can be achieved before the electron tunnels away from the nucleus.…”

“…The best attempt so far within effective mass theory (EMT) was proposed by Friesen [15]: For Si:P he correctly predicted a quadratic hyperfine shift, but one which is one order of magnitude larger than the value expected from the two measurements performed so far (Si:Sb [23] and Si:As [25]). More recently, other theories such as tight binding (TB) and Band Minima Basis (BMB) [21,24] have been applied to the same Si:P problem, leading to closer agreement with experiment.…”

“…For these reasons, the Stark effect in doped silicon has been broadly studied in literature, either theoretically [14][15][16][17][18][19][20][21][22] or experimentally [23,[25][26][27][28]. More generally, the ability to theoretically describe the donor electron wave function accurately in a wide range of electrostatic environments is beneficial for determining the values of control parameters which provide best performance, and in the best case for estimating a priori the feasibility of quantum algorithms and error correction codes [20].…”

Donor hyperfine Stark shift and the role of central-cell corrections in tight-binding theory

“…in Si [24], and more recent experimental measurements [25] of the hyperfine Stark shift for As donor in Si provide excellent opportunities to directly bench-mark TB theory against the experiment data sets. This work for the first time evaluates the role of central-cell corrections in atomistic TB theory through a direct comparison against the experimental data of the hyperfine interaction for a single As donor in Si.…”

Multiple-Quantum Transitions and Charge-Induced Decoherence of Donor Nuclear Spins in Silicon

“…Due to different interaction strengths with their surroundings, they form a powerful combination of a fast, but more volatile electron spin and a slower, but very coherent nuclear spin qubit [9][10][11]. This nuclear spin is I = 1/2 for phosphorus, but systems with a higher nuclear spin can be realized by simply replacing phosphorus by the other hydrogenic donors As (I = 3/2) [12,13], Sb (5/2 and 7/2) [14,15], and Bi (9/2) [16][17][18]. Several advantages of the d-dimensional Hilbert spaces of such systems, sometimes called qudits, have been proposed, such as the realization of simpler and more efficient gates [19,20] or more secure quantum cryptography [21].…”