Towards dislocation-driven quantum interconnects

TL;DR

The paper addresses the challenge of creating robust quantum interconnects in the solid state by proposing dislocation-driven patterns of spin qubits along dislocation cores, using NV centers in diamond as a representative system. It combines high-throughput screening with advanced first-principles methods (DFT, TDDFT, QDET, gCCE) in large diamond supercells to map ground- and excited-state properties, ISC rates, and coherence near dislocations. The authors identify NV configurations near 30° and 90° dislocations that retain or even enhance qubit-relevant traits, including favorable optical cycles, larger zero-field splitting components, clock transitions, and significantly longer coherence times under suitable dynamical decoupling, supporting the feasibility of 1D spin-defect arrays as interconnects. The work provides a theoretical foundation and concrete guidance for experimental realization of dislocation-based quantum interconnects, with potential extension to other hosts like SiC and various interfaces, marking a path toward scalable solid-state quantum networks.

Abstract

A central problem in the deployment of quantum technologies is the realization of robust architectures for quantum interconnects. We propose to engineer interconnects in semiconductors and insulators by patterning spin qubits at dislocations, thus forming quasi one-dimensional lines of entangled point defects. To gain insight into the feasibility and control of dislocation-driven interconnects, we investigate the optical cycle and coherence properties of nitrogen-vacancy (NV) centers in diamond, in proximity of dislocations, using a combination of advanced first-principles calculations. We show that one can engineer spin defects with properties similar to those of their bulk counterparts, including charge stability and a favorable optical cycle, and that NV centers close to dislocations have much improved coherence properties. Finally, we predict optically detected magnetic resonance spectra that may facilitate the experimental identification of specific defect configurations. Our results provide a theoretical foundation for the engineering of one-dimensional arrays of spin defects in the solid state.

Figures (5)

• Figure 1: Atomic sites chosen to create NV defects near dislocation cores in diamond. (a) Structure of the 30$^{\circ}$ dislocation in diamond. The {111} planes are indicated by dashed green lines near the dislocation core marked by the red $\top$ symbol. The shaded region indicates the stacking fault. The glide plane of the dislocation is denoted as the G plane; the adjacent {111} plane under compression (tension) is denoted as G$^-$ (G$^+$). (b) Same as (a), but for the 90$^{\circ}$ dislocation. (c) Atomic sites considered to create NV defects near the 30$^{\circ}$ dislocation core. NV defects are created in the G, G$^-$, and G$^+$ {111} planes, with the N atom and the nearby carbon vacancy located at one of the orange, red, or blue atomic sites. We denote an atomic site based on the {111} plane it resides in and the number assigned in the figure. For instance, G$^+$1 stands for the atomic site 1 in the G$^+$ plane. (d) Same as (c), but for the 90$^{\circ}$ dislocation. (e) Distribution of relative formation energies ($\Delta E_{\rm f}$, obtained using DFT at the PBE level of theory; see SI) of the 90 NV configurations near the 30$^{\circ}$ dislocation core compared to that in bulk diamond. (f) Same as (e), but for the 112 NV configurations near the 90$^{\circ}$ dislocation core.
• Figure 2: DFT single-particle levels computed at the DDH level of theory. (a) Bulk NV. (b) G9-G8 NV at the 30$^\circ$ dislocation core. (c) G4-G9 NV at the 90$^\circ$ dislocation core. See Fig. \ref{['fig:NV_location']} and SI for the NV labeling. Levels are shown in red (blue) for the spin-up (spin-down) channel, and next to each of them, we indicate the energy separation (in eV) relative to the valence band edge of diamond. The gray areas represent the valence and conduction bands of the system; the band edges of bulk diamond are indicated by the horizontal dashed lines in (b) and (c). In (a), the defect levels are labeled as $a_1$ and $e$ ($\bar{a}_1$ and $\bar{e}$) in the spin-up (spin-down) channel according to the irreducible representations of the $C_{3v}$ group. In (b) and (c), they are labeled as $a_1^d$, $e_x^d$, and $e_y^d$ ($\bar{a}_1^d$, $\bar{e}_x^d$, and $\bar{e}_y^d$) in the spin-up (spin-down) channel, with the superscript $d$ indicating the lowering of the $A_1$ and $E$ symmetries due to the presence of the dislocation. These labels are used here to illustrate the correspondence between the defects near dislocations and in the bulk.
• Figure 3: Many-body electronic states diagram, derived from the vertical excitation energies (VEEs) obtained using TDDFT calculations at the DDH level of theory, and inter-system crossing (ISC) rates. (a) Bulk NV. (b) G9-G8 NV at the 30$^\circ$ dislocation core. (c) G4-G9 NV at the 90$^\circ$ dislocation core. (d) G$^+$7-G$^+$3 NV at the 30$^\circ$ dislocation core. See Fig. \ref{['fig:NV_location']} and SI for the NV labeling. In (a), the many-body triplet (singlet) states are labeled as $^3A_2$ and $^3E$ ($^1E$ and $^1A_1$) according to the irreducible representations of the $C_{3v}$ group. In (b), (c), and (d), the many-body triplet (singlet) states are labeled as $^3A$(1), $^3A$(2), and $^3A$(3) [$^1A$(1), $^1A$(2) and $^1A$(3)] in order of increasing energy. The ISC rate (in MHz) between two many-body states is shown next to the arrow connecting the corresponding states.
• Figure 4: State population simulations for the initialization, optically detected magnetic resonance (ODMR), and readout processes of NV qubits. (a) Bulk NV. (b) G9-G8 NV at the 30$^\circ$ dislocation core. (c) G4-G9 NV at the 90$^\circ$ dislocation core. (d) G$^+$7-G$^+$3 NV at the 30$^\circ$ dislocation core. See Fig. \ref{['fig:NV_location']} and SI for the NV labeling. The left panels display the change of the population of different spin sub-levels of the triplet ground state (GS) and excited state (ES) and the singlet state (SS) as a function of the time ($t$) of the optical initialization process; the initial population is set to 1/3 for each of the GS spin sub-levels. The middle panels display the continuous wave ODMR simulations; the left (right) peak corresponds to the GS $\ket{0}$ to GS $\ket{-}$ (GS $\ket{0}$ to GS $\ket{+}$) transition enabled by the microwave (MW). The right panels display the photoluminescence (PL) spin readout simulations, where each line shows the change of the total population of the ES as a function of $t$, when the system is initialized into different sub-levels of the GS.
• Figure 5: Coherence times of NV centers near dislocations. (a) Clock transitions at weak magnetic field for the G$^+$7-G$^+$3 NV near the core of 30$^{\circ}$ dislocation. (b) Computed $T_2$ times at zero magnetic field as a function of the transverse zero-field splitting (ZFS) $E$, considering different axial $D$ values. (c) Computed $T^{\rm CPMG}_2$ times at zero magnetic field at different decoupling pulses $N$ for the G$^+$7-G$^+$3 NV near the core of 30$^{\circ}$ dislocation. (d) Coherence functions of the G$^+$7-G$^+$3 NV near the core of 30$^{\circ}$ dislocation at zero magnetic field for various decoupling pulses $N$.