High-throughput identification of spin-photon interfaces in silicon

Abstract

Color centers in host semiconductors are prime candidates as spin-photon interfaces for quantum applications. Finding an optimal spin-photon interface in silicon would move quantum information technologies toward a mature semiconducting host. However, the space of possible charged defects is vast, making the identification of candidates from experiments alone extremely challenging. Here, we use high-throughput first-principles computational screening to identify spin-photon interfaces among more than 1000 charged defects in silicon. The use of a single-shot hybrid functional approach is critical in enabling the screening of many quantum defects with a reasonable accuracy. We identify three promising spin-photon interfaces as potential bright emitters in the telecom band: [Formula: see text], [Formula: see text], and [Formula: see text]. These candidates are excited through defect-bound excitons, stressing the importance of such defects in silicon for telecom band operations. Our work paves the way to further large-scale computational screening for quantum defects in semiconductors.

Fig. 1.. Workflow for screening quantum defects in silicon for spin-photon interface.

Three types of simple defects were considered in this work, including substitutional, tetrahedral interstitial (T_d), and hexagonal interstitial (D_3d).

Fig. 2.. Overview of the quantum defect database with a series of candidates highlighted.

(A) TDM versus single-particle excitation energy at the single-shot HSE (HSE₀) level for the stable defects with nonsinglet spin multiplicity. The color of the circles indicates the stability of the defect in comparison to the defect configuration with the lowest formation energy, while the size represents the stability window of the charged defect within the bandgap. The isolines mark the radiative lifetime in microseconds. For comparison, we highlighted the T center (star) and ${\mathrm{Se}}_{\mathrm{Si}}^{+}$ (diamond) that have been computed with the same level of theory. ${\mathrm{Co}}_{\mathrm{Si}}^{2-}$ is removed from the figure due to nonphysically large shift of the KS levels from DFT to HSE₀. (B) Single-particle excitation energies at the HSE₀ level of theory for a series of candidates: ${\mathrm{K}}_{\mathrm{Si}}^0$, ${\mathrm{Nb}}_{\mathrm{Si}}^{-}$, and ${\mathrm{Fe}}_{\mathrm{i}}^0$. The defect levels are shown in red.

Fig. 3.. The defect formation energies, single-particle defect level diagrams, and the charge density for the relevant bands for the three defect candidates.

For simplicity, we only plot the defect levels in the gap for (A) ${\mathrm{Ti}}_{\mathrm{i}}^{+}$, (B) ${\mathrm{Fe}}_{\mathrm{i}}^0$, and (C) ${\mathrm{Ru}}_{\mathrm{i}}^0$. The localization of the defect levels is represented by the inverse participation ratio (see the Supplementary Materials for details). The charge density for the localized orbitals are shown with an isocontour value of 0.0005 Å⁻³, and the delocalized orbitals are shown with a value of 0.00008 Å⁻³.

Fig. 4.. Histogram showing the landscape of TDMs of the excitations for all the stable, nonsinglet defects.

We classified the bound-excitonic and intra-defect transitions using the inverse participation ratio.

Fig. 5.. Single-particle levels that are computed at the single-shot HSE (HSE₀ level for 3d transition metal series in the tetrahedral interstitial configuration for the neutral charge state.

The defect levels are represented by the t₂ and e symmetry due to the T_d point group.

Fig. 6.. Single-particle levels that are computed at the single-shot HSE (HSE₀) level for 4d transition metal series in the tetrahedral interstitial configuration with neutral charges.