Conclusions and Outlook

Engineering of $V_{\text{Si}}$ spin defects:

In this work electron irradiation was used to generate silicon vacancies in SiC. With electrons accelerated to an energy of 2 MeV, a homogeneous and statistical distribution of the defects across the crystal volume can be achieved. However, many applications, such as embedding the defects in photonic structures (Bracher et al., 2016), require much more accurate positioning of the defects. A higher spatial resolution, with positioning accuracy of 15 nm, can be achieved for example with proton irradiation (Kraus et al., 2017; Ohshima & Ed., 2017). An entirely new approach is the generation of defects using ultrashort laser pulses. Promising positioning accuracy of 40 nm have been achieved for the NV in diamond (Chen et al., 2016). Initial results of current research show that individual defects in silicon carbide can be generated with this approach. For the $V_{\text{Si}}$ in SiC, a higher positioning accuracy is to be expected, since the defect does not have to diffuse in order to form a color center (e.g. in diamond with a nitrogen). In comparison to ion implantation (Kraus et al., 2017; Wang et al., 2017), this laser pulse method is not only less technically complicated, but also enables the simultaneous detection of the created defects, as the same optical setup for generation can be used. However, in all of these methods, two more important questions need to be clarified, namely how much damage is introduced to the surrounding crystal lattice upon bombardment with ions or laser pulses, and to what extent it affects the spin coherence. Conventional crystal healing methods require temperatures of 1800°C, but the $V_{\text{Si}}$ can be annealed out above 600°C (Itoh et al., 1990; Kawasuso et al., 1996; Maier, 1993) and forms yet another promising defect; the divacancy (D. Christle et al., 2015; D. J. Christle et al., 2017). However, it was shown for the NV in diamond, that it is possible to suppress the formation of vacancy complexes during thermal annealing by doping (Fávaro de Oliveira et al., 2017), and a similar approach may be applied to SiC.

Coherence and storage of quantum information:

One of the most important results in this presented work is the finding of the initialization and readout of a single V2 electron spin ($S = 3/2$) with obtained coherence times of 200 µs at room temperature. The weak spin-orbit coupling of SiC support long spin coherence times but it is limited due to the interaction of the electron spin with the nuclear spin bath. Recently published results confirmed the theory presented in this work: the heteronuclear nuclear spin-bath (Yang et al., 2014) responsible for decoherence can be diluted using a magnetic field above 1000 G (D. Christle et al., 2015; Nagy et al., 2018b; Simin et al., 2017). Coherence times of up to 2 ms can be achieved at room temperature (Yang et al., 2014). The spin coherence time can be drastically extended, when the spin system is operated at low temperatures. Spin locking measurement of ensemble $V_{\text{Si}}$ reported 20 ms coherence times (Simin et al., 2017). In addition, a nearby nuclear spin of Si or C offers the possibility to transfer the information of the electron spin to nuclear spins via hyperfine coupling. The latter approach has already been achieved with state transfer of ensemble divacancy spins to nuclear spins (Ivá et al., 2016) in 4H-SiC. Initial results of current research in Stuttgart show that this is also possible with a single spin. Here, a single $V_{\text{Si}}$ is weakly coupled to a Si nuclear spin (Nagy et al., 2018a), which promises access to a nuclear spin memory with a long lifetime (Neumann et al., 2010). That would enable, for example, quantum error correction (Waldherr et al., 2014) and enhanced quantum sensing (Zaiser et al., 2016), as demonstrated for the NV in diamond.

Integration into electronic components:

The integration of the defects into electronic components presented in this work also shows that it is possible to gain control over charge states of the $V_{\text{Si}}$ defect. It was shown that charge states of impurities are sensitive to the environment and electric fields. The observed photo-ionization and recombination processes further modify the charge state dynamics (Wolfowicz et al., 2017), which opens the way for electric field sensing (Wolfowicz et al., 2018). The demonstration of electrical driving of hitherto unknown defects also shows that it is possible to produce non-classical light in SiC even without optical pumping, confirming published results (Lohrmann et al., 2015). A lateral p-i-n structure based on this work (Sato et al., 2018; Widmann et al., 2018a) recently showed that it is possible to excite V2 defects at room temperature electrically (Yamazaki et al., 2018). Here, the above-mentioned proton implantation was used to create small ensembles in the p-i-n layered area. However, more advances in fabrication is necessary to remove the bright defects produced at the SiC/SiO$_2$ interface, which were activated upon oxidation (Sato et al., 2018). Another promising feature of spin defects in such electronic structure is the possibility to detect their spin state electrically at room temperature (Cochrane et al., 2016). This remained elusive for the $V_{\text{Si}}$ in SiC and is a subject of current research.

Towards a spin-to-photon interface:

A closer spectral analysis at low temperatures of the $V_{\text{Si}}$ at the hexagonal lattice side (V1) has recently shown another important feature for quantum communication, namely the coherent photon emission into the zero-phonon line (Bracher et al., 2016). The ratio of photons in the zpl to the phonon sideband is known as the dbf. Here, at temperatures of 4 K, outstanding dbf values of over 40% of a single V1 center have been reported (Nagy et al., 2018b). Similar systems, like the NV and silicon vacancy center in diamond show 4% (Aharonovich et al., 2011) and 70% (Dietrich et al., 2014), respectively. In addition, optical resonant excitation with a narrow-band laser provides a spin initialization of nearly 100% in the magnetic ground state $|-1/2\rangle$ (Nagy et al., 2018a). The reported high initialization fidelity, a large spin signal and millisecond long spin coherence times (Nagy et al., 2018b, 2018a) in combination with the high dbf, spectrally stable and close to Fourier transform limited photon emission are basic prerequisites for a recently proposed spin-to-photon interface (Economou & Dev, 2016; Soykal et al., 2016). Such kind of interfaces are important building blocks for quantum repeaters (Briegel et al., 1998), which allow to distribute quantum entanglement via nodes over long distances (Bernien et al., 2013; Hensen et al., 2015; Humphreys et al., 2018; Kalb et al., 2017). Further advances in the fabrication of photonic structures (Bracher et al., 2016) offer promising opportunities to increase the dbf and hence the potential application for a spin-to-photon interface even further.

In conclusion, the silicon monovacancy in 4H-SiC offers a high potential to become a versatile and robust system (Baranov et al., 2011) in a wide variety for quantum technological applications (Awschalom et al., 2018).