Narrow magneto-optical transitions in Erbium implanted silicon carbide-on-insulator

Abstract

Solid state spin photon interfaces operating in the near telecom and telecom bands are a key resource for long distance quantum communication and scalable quantum networks. However, their optical transitions often suffer from spectral diffusion that hampers the generation of coherent spin photon entanglement. Here we demonstrate narrow magneto-optical transitions of erbium dopants implanted into thin film silicon carbide (SiC)-on-insulator, a viable platform for industrially scalable quantum networks. Using high-resolution resonant spectroscopy and spectral hole burning at cryogenic temperatures, we reveal sub megahertz homogeneous linewidths and identify two lattice sites that best stabilise the emitters. We further characterise their optical lifetimes and magneto-optical response, establishing erbium doped SiC-on-insulator as a robust and scalable platform for on-chip quantum networks.

Figures (4)

• Figure 1: Er$^{3+}$-site identification utilizing PL- and PLE- measurements within 4H-SiCOI: a schematic of the experimental setup configuration where modulated tunable excitation (green) is applied onto a 4H-SiCOI-sample via a fiber ferrule (black), aiming to address the Er$^{3+}$-defects (red) embedded in the centre of a trenched 4H-SiC-layer; bPLE- and PL-spectra obtained from the Er$^{3+}$ implanted sample; c identified narrowest optical transitions ($\alpha_{1}$, $\alpha_{2}$, $\alpha_{7}$) from Er$^{3+}$-site $\alpha$; d identified optical transitions from Er$^{3+}$-site $\beta$. The solid lines in c and d are Gaussian fits to the data; e Overview of observed inhomogeneous linewidths from identified $\alpha$ and $\beta$ resonances in relation to frequency where error-bars denote uncertainties from Gaussian fits.
• Figure 2: Spectral hole burning and homogeneous broadening: a Spectra of $\alpha_{1}$ optical transition measured with a frequency comb (black) and single laser configuration (red); b Spectra of $\alpha_{2}$ optical transition measured with a frequency comb (black) and single laser configuration (red); c Spectra of $\alpha_{3}$ optical transition measured with a frequency comb (black) and single laser configuration (red); d Spectra of $\alpha_{7}$ optical transition measured with a frequency comb (black) and single laser configuration (red); e Spectral hole burned in $\alpha_{1}$ fitted with a single Lorentzian fit (red) with $\text{P} = 0.19\ \mu\text{W}/\text{line}$. The inset illustrates the excitation power dependency $\sqrt{P_{line}}$ on the spectral hole (homogeneous) linewidth ranging between (0.88 $\pm$ 0.23) MHz to (6.97 $\pm$ 1.91) MHz ((0.44 $\pm$ 0.12) MHz to (3.49 $\pm$ 0.96) MHz)); f Spectral hole burned in $\alpha_{2}$ fitted with a single Lorentzian fit (red) with an inset illustrating the excitation power dependency $\sqrt{P_{line}}$ onto the spectral hole (homogeneous) linewidth ranging between (1.99 $\pm$ 0.42) MHz to (8.87 $\pm$ 1.81) MHz ((1 $\pm$ 0.21) MHz to (4.44$\pm$ 0.9) MHz)).
• Figure 3: Zeeman splitting versus the applied magnetic field: a resonance $l_{1}$ at 194968 GHz exhibiting a $\Delta g$= (69.09 $\pm$ 1.08) GHz/T; b resonance $\alpha_{1}$ at 195347 GHz exhibiting a $\Delta g$ = (93.42 $\pm$ 3.12) GHz/T; c resonance $\alpha_{2}$ at 195427 GHz exhibiting a $\Delta g$ = (184.83 $\pm$ 9.54) GHz/T; d resonance $\alpha_{3}$ at 196088 GHz exhibiting a $\Delta g$ = (111.48 $\pm$ 2.69) GHz/T; e resonance $\beta_{2}$ at 197191 GHz exhibiting a $\Delta g$ = (116.56 $\pm$ 1.69) GHz/T; f resonance $\alpha_{7}$ at 197527 GHz exhibiting a $\Delta g$= (158.86 $\pm$ 3.73) GHz/T.
• Figure 4: Optical lifetime: a Measured decay-transients over 6 ms in logarithmic scale for observable resonances $\alpha_{2}$, $\alpha_{3}$, $\alpha_{5}$, $\alpha_{7}$ and $\beta_{2}$ (dotted) and single exponential fits (solid lines) where optical lifetimes of (2.08 $\pm$ 0.3) ms, (3.41 $\pm$ 0.8) ms, (3.31 $\pm$ 1.1) ms, (2.52 $\pm$ 0.1) ms and (4.36 $\pm$ 1.1) ms were derived; b decay overview in dependence of observable resonance frequency where dots illustrate determined lifetimes. The error bars are from the fit uncertainties.