Scalable, nanoscale positioning of highly coherent color centers in prefabricated diamond nanostructures

TL;DR

The paper addresses scalable, nanoscale placement of highly coherent nitrogen-vacancy (NV) centers in prefabricated diamond nanostructures. It combines nitrogen delta-doping during chemical vapor deposition with localized delta-electron irradiation and post-irradiation annealing to form NV centers aligned to the centers of nanopillars, with tunable NV numbers. It reports a depth confinement of about $\sim 4\,\mathrm{nm}$ and lateral confinement of $\sigma_{loc}^{pillar} = 46(1)\,\mathrm{nm}$ in 280 nm pillars and $72(1)\,\mathrm{nm}$ in 480 nm pillars (versus $\sim 102(2)\,\mathrm{nm}$ in unpatterned diamond), along with average $T_2^{Hahn} = 98(37)\,\mathrm{\mu s}$ and spin-dependent PL contrast $C_{Rabi} = 18(4)$; there is a $1.8\times$ PL enhancement for NVs localized in pillars. The approach yields about a $3\times$ higher yield of NV centers with single-electron-spin sensitivity than conventional implantation, and diffusion-capture Monte Carlo modeling gives $D_V \approx 17(4)\,\mathrm{nm^2/s}$ (bulk-like) and $D_V \approx 21(2)\,\mathrm{nm^2/s}$ for pillar-confined localization, consistent with vacancy diffusion during annealing. Overall, the method enables scalable, high-yield, high-coherence NV-based sensing and quantum information processing in integrated diamond nanophotonic devices.

Abstract

Nanophotonic devices in color center-containing hosts provide efficient readout, control, and entanglement of the embedded emitters. Yet control over color center formation - in number, position, and coherence - in nanophotonic devices remains a challenge to scalability. Here, we report a controlled creation of highly coherent diamond nitrogen-vacancy (NV) centers with nanoscale three-dimensional localization in prefabricated nanostructures with high yield. Combining nitrogen $δ$-doping during chemical vapor deposition diamond growth and localized electron irradiation, we form shallow NVs registered to the center of diamond nanopillars with wide tunability over NV number. We report positioning precision of ~ 4 nm in depth and 46(1) nm laterally in pillars (102(2) nm in bulk diamond). We reliably form single NV centers with long spin coherence times (average $T_2^{Hahn}$ = 98 $μs$) and 1.8x higher average photoluminescence compared to NV centers randomly positioned in pillars. We achieve a 3x improved yield of NV centers with single electron-spin sensitivity over conventional implantation-based methods. Our high-yield defect creation method will enable scalable production of solid-state defect sensors and processors.

Figures (5)

• Figure 1: Targeted formation of color centers aligned to prefabricated diamond nanostructures. (a) Schematic of our formation method showing a nanopillar containing a near-surface nitrogen $\delta$-doped layer (blue disk). In $\delta$-electron irradiation, an electron beam of 20 spot size (yellow line) irradiates the center of a pillar, creating vacancies along its trajectory. (b) Upon annealing, monovacancies diffuse to form a vacancy-rich region (dark shaded region) in which they can be captured by nitrogen atoms to form NV centers (inset). (c) A scanning electron micrograph of a unit block of etched diamond pillars and mesas framed by alignment marks. Each $50\times$ 50 unit block consists of two unetched mesas (top left and bottom right), and two regions of nanopillars with a diameter of $\qty{280}{\nano \meter}$ (top right) and $\qty{480}{\nano \meter}$ (bottom left). The overlaid yellow circles denote the target positions of the electron beam (scale bar: 20). The insets show confocal $PL$ images of a single $\qty{480}{\nano \meter}$ pillar before (top) and after (bottom) 4.8e21^-$\delta$-electron irradiation and annealing (scale bar: 1).
• Figure 2: Electron dose control over NV creation. Plotted are the average number of created NVs per spot in 280 diameter pillars (purple circles), in 480 diameter pillars (teal circles), and in the mesas (red circles). The error bars denote 95 confidence interval of the NV number estimation. The results from MC simulations (diamonds) are plotted in corresponding colors and show good agreement with the measurements. The fitted curves for the MC simulation results (dotted lines) are shown as a guide to the eye. The insets show confocal micrographs of a unit block after irradiation and annealing with a dose of 1.6e19^- (top left) and 4.8e21^- (bottom right). (scale bar: 20). The overlaid boxes indicate the locations of the 280 pillars (purple box), 480 pillars (teal box), and mesas (red boxes).
• Figure 3: Quantifying spatial confinement of formed NVs. (a) Schematic of pixel-wise averaging method for estimating $\sigma_{loc}$. NVs are positioned at the target position (black star) with a lateral precision $\sigma_{loc}$ (black solid line). Red dashed line indicates the point spread function $\sigma_{PSF}$ of our confocal microscope. N confocal images of unpatterned mesas are pixel-wise averaged to give a $\sigma_{tot}$ from which $\sigma_{loc}$ is extracted as discussed in the main text. Residual optical aberrations are indicated by $\sigma_{sys}$. (b) Data points show the radial $PL$ profiles (averaged over angle) of the pixel-wise averaged images (bin size: 40). Error bars show standard error of the averaging. Colored solid lines are 2D Gaussian fits, from which $\sigma_{tot}$ is extracted. For comparison, red dashed lines show the radial cuts of the 2D Gaussian functions with a peak width of $\sqrt{\sigma_{PSF}^2+\sigma_{sys}^2}$. Plots are offset for clarity, each with a relative $\Delta PL$ of 100 kcps. Insets show the averaged images with $PL$ scaling inversely proportional to the dose (scale bar: 1). (c) Measured $\sigma_{loc}$ in the mesas (red circles, error bar: 95 confidence interval) is plotted as a function of dose. (the 1.6e19^- dose data point is omitted because of large errorbars $> \qty{1}{\micro\meter}$). The red dotted line indicates the average $\sigma_{loc}$ of 102(2) measured in the mesas. Solid lines are MC simulations of $\sigma_{loc}^{pillar}$ in $\delta$-e$^-$-irradiated pillars, which are lower than the analytically calculated $\sigma_{loc}^{pillar}$ for NVs created without localization methods (dashed lines).
• Figure 4: Spin coherence time $T_2^{Hahn}$ and $PL$ properties of single NVs formed in nanopillars. (a) Histogram of $T_2^{Hahn}$ for NVs $\delta$-electron irradiated with 1.6e20^-. (Inset) Average $T_2^{Hahn}$ as a function of irradiation dose. (b) Histogram of Rabi contrast $C_{Rabi}$ at 1.6e20^-. (Inset) Average $C_{Rabi}$ as a function of $\delta$-electron irradiation dose. (c-d) Histogram of $PL_{sat}$ in (c) 280 and (d) 480 pillars. Each plot shows both non-irradiated (gray) and 1.6e20^--$\delta$-electron irradiated (yellow) pillars. The solid curves are the Gaussian fit for the histograms. (e) Dashed lines indicate mean photon collection efficiency calculated from FDTD simulations for a given lateral distribution $\sigma_{loc}^{pillar}$. Data points are the experimentally measured sample mean of the $PL_{sat}$ distribution for $\delta$-electron irradiated pillars (triangles) and non-irradiated pillars (circles) for 480 (teal) and 280 (purple) diameter pillars. The error bars denote the standard error of the estimation of the population mean. The limits of the secondary y-axis are chosen so that the simulated collection efficiency and measured $PL_{sat}$ for the nonirradiated pillars line up.
• Figure 5: High-yield fabrication of highly sensitive magnetic field sensors. (a) Histogram and (b) cumulative density function (CDF) of the estimated AC magnetic field sensitivity $\eta$ of single NVs in 480 pillars. The $\eta$ distribution is estimated for $\delta$-doped, $\delta$-electron irradiated (yellow) and $\delta$-doped, non-irradiated (gray) pillars using measured $PL_{sat}$, $T_{2}^{Hahn}$ and $C_{Rabi}$ distributions. The $\eta$ distribution of conventionally implanted pillars (cyan) is generated using $PL_{sat}$ and $C_{Rabi}$ measurements on our non-irradiated pillars with reported $T_2^{Hahn}$ distribution for 30keV implantation jakobi_efficient_2016. The distribution for $\delta$-doped, $\delta$-electron irradiated pillar with better sidewall taper angle of $\qty{70}{\degree}$ (green) is also estimated using $T_{2}^{Hahn}$ and $C_{Rabi}$ measurements on our $\delta$-electron irradiated pillars with the estimated $PL$ improvement from FDTD simulations. A secondary x-axis shows the minimum averaging time for a 53-deep NV to detect a single electron spin located on the diamond surface.