Abstract
One of the challenges in the field of quantum sensing and information processing is to selectively address and coherently manipulate highly homogeneous qubits subject to external perturbations. Here, we present roomtemperature coherent control of highdimensional quantum bits, the socalled qudits, associated with vacancyrelated spins in silicon carbide enriched with nuclear spinfree isotopes. In addition to the excitation of a spectrally narrow qudit mode at the pump frequency, several other modes are excited in the electron spin resonance spectra whose relative positions depend on the external magnetic field. We develop a theory of multipole spin dynamics and demonstrate selective quantum control of homogeneous spin packets with subMHz spectral resolution. Furthermore, we perform twofrequency Ramsey interferometry to demonstrate absolute dc magnetometry, which is immune to thermal noise and strain inhomogeneity.
Introduction
Quantum bit or qubit is a twolevel system, which builds the foundation for quantum computation, simulation, communication, and sensing^{1}. Quantum states of higher dimension, i.e., qutrits (D = 3) and especially qudits (D = 4 or higher), offer significant advantages. Particularly, they can provide noiseresistant quantum cryptography^{2}, simplify quantum logic^{3}, and improve quantum metrology^{4}. Flying and solidstate qudits have been implemented on the basis of photonic chips^{5} and superconducting circuits^{6}, respectively. The extension of the superconducting phase qubit to D = 5 were used to emulate the dynamics of single spins S = 1/2, 1, and 3/2, allowing a measurement of the Berry phase under 2π rotation and paving a way to complex quantum computational protocols and architectures. However, superconducting phase qudits operate in the millikelvin temperature range. Therefore, there is still a lack of roomtemperature qudits with long coherence time.
To overcome this challenge, we have considered optical active centers in solids. Up to now, only two materials have been known to host the centers with spin and optical properties allowing coherent control and manipulation of their ground state spins by means of the optical and radiofrequency quanta at ambient condition. Namely, diamond containing negatively charged nitrogenvacancy (NV^{−}) centers and silicon carbide (SiC) containing silicon–carbon divacancies or negatively charged silicon vacancies (V_{Si}) centers^{7,8,9}. Commercially produced waferscale SiC possesses ideal combination of hardness, radiation stability, and thermal conductivity of diamond, with the semiconducting properties and technological maturity of silicon^{10}. In addition, the polymorphism of SiC enables the tuning of magnetic and optical characteristics of the centers^{11,12}.
The divacancy centers posses the triplet ground state (S = 1), similarly to the NV^{−} centers in diamond. They have been recently used to demonstrate a variety of prominent examples of the quantum phenomena in solids^{8,13}, including the quantum entanglement in a macroscopic spin ensemble at ambient conditions^{14}. Despite the unique properties, the triplet ground states of both the NV^{−} centers in diamond and divacancies in SiC impose limitations on their use as qudits.
These limitations are lifted for the centers with the S = 3/2 (D = 4) ground state, i.e., for the negatively charged silicon vacancy (V_{Si}) centers in SiC. Due to stable singlephoton emission in NIR range, moderate zerofield splitting in the MHz region, zerophonon lines with high Debay–Waller factor (up to 40%), and long spin coherence times up to 20 ms at cryogenic temperatures, V_{Si} centers were proposed for various quantum applications^{8,9,15,16,17}. However, until now these centers have been treated as a canonical S = 1/2 qubit system^{9,15,16,17,18,19}.
Here, we develop a concept of operating V_{Si} centers as qudits, which can reside in multiple superpositions of four basis states in the Hilbert space with the spin projections m_{S} = ± 3/2, ± 1/2. Similar to atomic orbitals, an ensemble of such spin qudits is described by 15 linearly independent spherical multipoles, with three components being the spin dipole \({\cal{P}}\), five components being the spin quadrupole \({\cal{D}}\), and seven components being the spin octupole \({\cal{F}}\)^{20}. Examples of such multipoles and their description in terms of the spindensity matrix ρ are sketched in Table 1. When qudit multipoles are excited, they decay with three relaxation times: that of the spin dipole (T_{p}), quadrupole (T_{d}), and octupole (T_{f}).
We use this concept to experimentally excite the qudit modes at ambient conditions by means of twofrequency optically detected magnetic resonance (ODMR). Strikingly, we found their spectral width to be about an order of magnitude narrower than the inhomogeneous broadening of the corresponding spin resonance. By applying the Ramsey interferometry to these spin qudits, we achieve a spectral selectivity of 600 kHz and a spectral resolution of 30 kHz. As a practical consequence, we demonstrate absolute DC magnetometry insensitive to thermal noise and strain fluctuations.
We focus on the V_{Si} centers in two main hexagonal polytypes of SiC, namely 6 H and 4 H, with natural (^{29}Si 4.7%, ^{13}C 1.1%) and modified (^{29}Si 1%, ^{13}C 1.1%) isotope abundance (see the Methods section). Exemplary electron paramagnetic resonance (EPR) spectrum of the investigated 6H^{28}SiC sample is shown in Supplementary Fig. 1a.
Results
Visualization of the spin qudit modes
To experimentally visualize the V_{Si} spin qudit modes, we use optically detected magnetic resonance (ODMR) as described elsewhere^{11,15,19,21,22} (see also Methods and Supplementary Fig. 2). Optical pumping results in a preferential population of either the m_{S} = ± 3/2 or m_{S} = ± 1/2 states (depending on SiC polytype and the V_{Si} crystallographic site, as was shown in refs. ^{9,12,20,23}). Such a spin alignment is theoretically described by the contributions to the diagonal components of the spindensity matrix δρ_{+3/2,+3/2} = δρ_{−3/2,−3/2} = −δρ_{1/2,1/2} = −δρ_{−1/2,−1/2}. In the multipole decomposition of the spin density matrix, this corresponds to the appearance of the spin quadrupole \({\cal{D}}_0\) (see Table 1). Application of a strong microwave (MW) field at a fixed frequency v_{pump} mixes the m_{S} = ± 3/2 and m_{S} = ± 1/2 states, resulting in the excitation of other qudit modes as well. They are probed as relative difference of the spindependent photoluminescence ΔPL/PL, while sweeping the frequency of a second (weak) MW field v_{probe}.
Figure 1a shows an ODMR resonance associated with the V3(V_{Si}) center in 6H^{28}SiC^{23,24,25}. The ODMR resonance without the pump MW field (black curve) reveals the zerofield spin splitting 2D = 26.8 MHz between the Kramers doublets ±1/2 and ±3/2 states, whereas the sign of D determining the ordering of the Kramers doublets, is negative^{12}. Application of a strong, pump MW field at v_{pump} = 26.8 MHz saturates the spin transitions, that is seen as a spectral hole burning^{26,27}. To model it, we assume that the ODMR resonance is inhomogeneously broadened and agregates many homogeneous spin packets with different resonance frequencies, as schematically shown by thin lines in Fig. 1b. The pump MW field excites qudit modes in particular spin packets, which are detected as a reduction of the ODMR signal at certain frequencies. In zero magnetic field, all the excited modes are degenerated and manifest themselves as a single spectral hole. Remarkably, the spectral width of such hole can be much narrower than the inhomogeneous linewidth, and in certain cases we observe 250 kHz (Supplementary Figs. 3, 4).
The twofrequency technique enables potentially the access to the properties of spin centers with the resolution determined by the natural linewidth of individual centers. At the same time, the detected signal is much stronger and less noisy than that from a single center, which is crucial for applications.
To increase the sensitivity, we modify the detection scheme. The MW pump is now modulated on/off, and the ODMR spectrum is detected using a lockin amplifier. The pumpinduced changes of the ODMR spectrum are presented in Fig. 1c for several magnetic fields B_{z} applied along the caxis of SiC. Several qudit modes v_{mod} are now clearly detected and their spectral positions relative to v_{pump} depend on B_{z}.
To understand the spectrum of spin qudit modes, we analyze successively the most probable sources of inhomogeneous broadening: (i) magnetic fluctuations (for instance, due to nuclear fields) and (ii) variations of the zerofield splitting (for instance, due to local strain). The first mechanism sufficiently influences the inhomogeneous broadening of the NV^{−} center ODMR transitions in diamond, as shown in ref. ^{26}. Upon application of the magnetic field along the caxis, a single spin packet is characterized by four resonances shifting linearly with the magnetic filed as 2D ± γB_{z} and 2D ± 2γB_{z}^{28}. Here, γ = 28 MHz/mT is the gyromagnetic ratio. The corresponding fan chart is shown by red lines in Fig. 1d. The line thickness represents the coupling strength to the MW field. Another spin packet feels another local magnetic field δB and the fan chart is shifted along the vertical axis in Fig. 1d. The mean value \(\delta \bar B\) defines the inhomogeneous ODMR linewidth.
Assume that the pump MW field is in resonance with one of the spin transitions of a particular spin packet (the solid circle in Fig. 1d), then the other three spin resonances of the same spin packet will be effected due to the excitation and relaxation of spin multipoles, leading to the appearance of qudit modes (the open circles in Fig. 1d). Their spectral position relative to the pump frequency is symmetric in respect to the zerofield splitting 2D. In case when v_{pump} = 2D, the excited qudit modes remain degenerate even when \(B_z \, \ne \, 0\) (the square in Fig. 1d), which contradicts to the experimental observation of Fig. 1c. Thus, we conclude that magnetic fluctuations are not the main source of inhomogeneous broadening.
We now consider the other mechanism of broadening, caused by the variation of the zerofield splitting around the mean value \(2\bar D = 26.8\,{\mathrm{MHz}}\). This mechanism is natural to expect, given that we investigate isotopically purified 6H^{28}SiC with low content of the spincarrying ^{29}Si isotopes. In this case, the fan chart of the fielddependent ODMR lines associated with different spin packets are shifted along the horizontal axis in Fig. 1e. Positions of the excited qudit modes are given by v_{mod} = v_{pump} ± sγB_{z} with s = 1, 2, 3, and 4. They depend on the magnetic field strength, which is in qualitative agreement with the experiment of Fig. 1c.
Figure 2a visualizes the magnetic field evolution of the V3(V_{Si}) spin qudit modes in 6H^{28}SiC. The observed behavior is more complex compared with the qualitative consideration above. Not all qudit modes shift linearly with B_{z}, they have different strength and some of them are not even detected. To understand this behavior, we have developed a theory of qudit mode excitation in inhomogeneously broadened spin ensembles (see the Methods section).
We start from the effective spin Hamiltonian of the V_{Si} centers in a simple axial model \(H = D\left( {S_z^2  5/4} \right) + \gamma {\mathbf{S}} \cdot {\mathbf{B}}\). In small magnetic fields that we consider, \(\gamma B \ll D\), the eigenstates are given by
where B_{z} and \(B_ \bot\) are the magnetic field components parallel and perpendicular to the caxis, respectively. Due to the mixing of the spin states by \(B_ \bot\) and a low trigonal pyramidal local symmetry of the V_{Si} centers in SiC, all four transitions between the ±1/2 and ±3/2 spin states are allowed^{28}. Their spectral positions read
where s, s′ = 0, ±1. Equation (2) describes the spectral hole burning at v_{pump} (s, s′ = 0) and 8 qudit modes.
We now analyze the strength of the qudit modes coupling to the MW field and how excitations between these modes are transfered. At room temperature, when the energy of thermally excited phonons is much higher than the spin splitting, one can neglect the anisotropy of spin relaxation and consider that all spin multipoles relax independently. If spin relaxation occurs due to fluctuating magnetic fields, only the transitions with Δm_{S} = ±1 are allowed, as had also been assumed in the earlier works^{15}. Application of this constraint to the matrix of spin relaxation, see Eq. (15) in the Methods section, yields the ratio between the multipole relaxation times T_{P} = 3T_{d} = 6T_{f}. Remarkably, the relaxation rates +3/2 ↔ + 1/2 and −3/2 ↔ −1/2 are not equal to that of +1/2 ↔ −1/2, and we obtain for them \((1/2)T_d^{  1}\) and \((2/3)T_d^{  1}\), respectively.
The analytical solution of the rate equations, Eq. (12) in Methods, shows that the strength of the modes (s, s′) = (+1, +1), (+1,−1), (−1,+1), and (−1,−1) are proportional to 5T_{d} − T_{p} − 4T_{f}. Therefore, they vanish if the aforementioned ratio T_{p} = 3T_{d} = 6T_{f} holds. As a consequence, only four out of eight possible qudit modes are observed in Fig. 2a. The evolution of the qudit modes with B_{z}, calculated as described in the Methods section, is shown in Fig. 2b. Here, we assume the Gaussian distribution of the zerofield splitting, \(f(D) \propto exp[(D  \bar D)^2/(\delta D^2)]\), with the mean value \(2\bar D = 26.8\,{\mathrm{MHz}}\) corresponding to the V3(V_{Si}) center in 6HSiC. A perfect agreement with the experimental data is achieved for \(B_ \bot = 60\,\mu {\mathrm{T}}\), accounting for a uncompensated perpendicular component of the external magnetic field.
Some discrepancy between the measured and calculated images may stem from more general form of spin relaxation, not described by a single parameter, or more general form of inhomogeneous broadening.
Coherent control of spin qudit modes
To further confirm the conclusions of our model, we investigate the qudit modes in a stronger magnetic field B_{z} = 211 μT when the Zeeman splitting is larger than inhomogeneous broadening. A precise analysis of the ODMR spectra^{29,30} shows that \(B_ \bot = 73\,{\mathrm{\mu T}}\) remains almost the same. Figure 3a shows relative positions of the V_{Si} spin sublevels calculated for this field configuration. The ODMR spectrum consists of four lines [the lower curve in Fig. 3b] and the corresponding spin transitions enumerated from v_{1} to v_{4} are shown by arrows in Fig. 3a. The inner transitions v_{1,2} are stronger than the outer transitions v_{3,4}, in accordance with earlier studies^{28}.
Now, we burn a spectrally narrow hole at v_{1} (v_{pump} = 21.8 MHz), as shown in Fig. 3b. As expected from our model, excitation of qudit modes leads to the emergence of spectral holes at the v_{2}, v_{3}, and v_{4} transitions [see the upper curve in Fig. 3b]. Their spectral positions shift linearly with v_{pump}, keeping the frequency difference the same. This is particularly pronounced in another sample with much larger inhomogeneous broadening, isotopically purified 4H^{28}SiC (Supplementary Fig. 5). We hence conclude that by varying v_{pump} within the v_{1} ODMR line, different spin packets are selected. Remarkably, there is another signal at v_{5} in Fig. 3b, which has the opposite sign and appears only if the v_{1} qudit mode is excited. The same behavior is also observed in 4HSiC^{30} (see also Supplementary Fig. 6). These general properties are used to implement spectrally selective coherent control of qudit modes.
We start with driving the v_{1} transition. Figure 3c shows Rabi oscillations when the driving power is relatively low W_{1} = 11 dBm. The corresponding πpulse duration of 1.2 μs provides subMHz spectral selectivity. We then spectrally select a spin packet using a long π pulse at v_{1} and coherently drive it at the v_{5} resonance followed by a second long π pulse at v_{1} for the readout [inset of Fig. 3d]. A high driving power W_{5} = 33 dBm yields fast Rabi oscillations, as presented in Fig. 3d. For such a power, the π/2 pulse is 80 ns, corresponding to a bandwidth of ~10 MHz. This bandwidth is wide enough to encompass the v_{5} linewidth.
Twofrequency Ramsey interferometry
To highlight the advantages of qudit modes for quantum information processing and sensing, we demonstrate Ramseybased absolute magnetometry performing twofrequency experiments with the protocol shown in Fig. 4a (and Supplementary Fig. 2). When the probe frequency v_{probe} = 6.7 MHz is equal to the v_{5} qudit mode frequency, the signal represents free induction decay (Fig. 4b). We note that though the other, v_{3} resonance lies within the bandwidth, the population of the m_{S} = +3/2 and m_{S} = −1/2 is equal after the v_{1} πpump pulse (Fig. 3a), and this resonance is not driven. This is also confirmed by the absence of pronounced fringes in the free induction decay of Fig. 4b.
When v_{probe} = 11.7 MHz is detuned from the v_{5} mode, Ramsey fringes are clearly detected (Fig. 4c). A fit of these dynamics to \(cos(2\pi f_R\tau )exp(  \tau /T_2^ \ast )\) gives the coherence time of spin packets \(T_2^ \ast = 357 \pm 24\,{\mathrm{ns}}\). For comparison, we also perform standard Ramsey measurements using single MW frequency, but do not observed fringes because of inhomogeneous broadening and nonselective excitation (Supplementary Fig. 7).
Solid circles in Fig. 4d represent the fast Fourier transform (FFT) of the experimental data from Fig. 4c. The fitting of these data to a Lorentz function yields the frequency of the Ramsey fringes fR = 4.51 ± 0.03 MHz, with the corresponding spectral resolution of 30 KHz. Following Eq. (1), the effective magnetic field can be measured with high accuracy as
Here, θ is the angle between the magnetic field direction and the caxis of SiC. It is calculated from the ratio (v_{2} − v_{1})/(v_{4}− v_{3}) with resolution better than 1° following the previously described procedure^{29}, which yields θ = 19° ± 0.5°. The gyromagnetic ratio γ is the constant value. We then obtain from Eq. (3) B_{eff} = 223 ± 2 μT. To analyze the spectral selectivity of different spin packets, we marginally vary the external magnetitic field strength and the pump/probe frequencies. The results are summarized in Table 2, where measurement No. 1 corresponds to the case discussed above. We then increase B and change v_{pump} to select the same spin packet (measurement No. 2 in Table 2). The corresponding FFT of the Ramsey fringes is represented by open squares in Fig. 4d and a lorentzian fit (the dashed line) yields f_{R} = 3.92 ± 0.05 MHz and B_{eff} = 242 ± 3 μT. A comparison of measurements No. 1 and No. 2 demonstrates the spectral selectivity of spin packets to be ~600 kHz. We note that an unusually narrow ODMR linewidth of 500 kHz in SiC was observed earlier in twofrequency experiments, but its origin had not been discussed^{30}.
Discussion
Measurements Nos. 1 3 in Table 2 demonstrate that the effective magnetic field B_{eff} seen by different spin packets, which are selected by different v_{pump}, is the same within the error bars. This is a manifestation that the inhomogeneous broadening is caused mostly by local variations of the zerofield splitting 2D rather than magnetic fluctuations. Remarkably, another pair of measurements (No. 2 and No. 3) in Table 2 show that the magnetic field strength can be measured without calibration of the zerofield splitting, which can be used to implement absolute (i.e., immune to thermal noise and strain inhomogeneity) DC magnetometry^{31}.
To summarize, we demonstrated coherent manipulation of spin qudit modes in isotopically purified SiC at room temperature. We also developed a theory describing the excitation and detection of these modes in inhomgeneously broadened systems and showed that qudits are characterized by multiple relaxation times. These findings can lead to dipolecoupled networks^{11}, unconditional electronnuclear spin registers^{14}, and spectral selection of highly coherent individual spins^{15,22,32}, particularly in nanocrystals^{33}. Our results hence open new possibilities to improve the sensitivity of quantum sensors and execute nontrivial quantum protocols in dense spin ensembles.
Methods
Samples preparation
The semiinsulating (SI) 4HSiC substrate with natural isotope content (4.7% of ^{29}Si and 1.1% of ^{13}C) was purchased from Cree. The 6H and 4H^{28}SiC samples, purified from the ^{29}Si isotope, were grown by the seeded physical vapour transport method^{34} with a ^{28}Si enriched precursor material on a 6HSiC substrate with natural isotope content. The ^{29}Si isotope abundance in these samples of 1% was established by the EPR spectroscopy (Supplementary Note and Supplementary Fig. 1). To create V_{Si} centers, the 6H^{28}SiC and SI 4HSiC samples were irradiated with 2 MeV electrons at temperatures close to room temperature to a total dose of 10^{18} cm^{−2}. To create V_{Si} centers in 4H^{28}SiC, the sample was irradiated with 3–5 MeV neutrons at temperatures close to room temperature to a total dose of 10^{15} cm^{−2}.
Twofrequency optically detected magnetic resonance (ODMR)
A scheme of the experimental setup used in our ODMR/holeburning experiments is presented in Supplementary Fig. 2. A 787.7 nm diode laser (LD785SE400 from Thorlabs) was pulsed using an acoustooptic modulator (AOM A&A OptoElectronic MT250A02800). The free space laser beam was coupled by several mirrors into a 50μm optical fiber, which transfers the light through a ×10 microscope objective (Olympus LMPLN10XIR). The objective focused the beam on the sample surface with a spot diameter of approximately 10 μm. The laser power was about 10 mW at the sample surface. The PL was collected by the same ×10 objective and filtered by an 850 nm and 875 nm long pass before it was coupled into a 600 nm optical fiber. As a detector, a Si avalanche photodiode was used (APD120A from Thorlabs). The signal was processed by a lockin amplifier (Signal Recovery DSP 7230). Radiofrequency was generated by a signal generator (Stanford Research Systems SG384) and amplified by an amplifier (Vectawave VBA100018) before being supplied to the sample by a 0.5 mm copperstripline the sample was placed on. In the hole burning experiments, the second radiofrequency was supplied by an identical Stanford Research Systems SG384 generator. The radio waves were modulated using RFswitches (MiniCircuits ZASWA250DR+) and a mixed before entering the amplifier using combiner (MiniCircuits ZFSC24S+). To apply the external magnetic field, a permanent magnet below the sample and Helmholtz coil was used.
Multipole decomposition
An ensemble of spin3/2 centers is described by the 4 × 4 spindensity matrix ρ, which can be decomposed into a sum of the 16 basis matrices,
The matrices \({\cal{M}}_j\) comprise three matrices of the spin dipole,
five matrices of the spin quadrupole,
seven matrices of the spin octupole,
and the diagonal matrix \({\cal{I}}/2\), with \({\cal{I}}\) being the unit matrix. The above matrices form a complete orthogonal basis, \({\mathrm{Tr}}({\cal{M}}_i^\dagger {\cal{M}}_j) = \delta _{ij}\).
Calculation of ODMR spectra
The density matrix describing an ensemble of spin centers satisfies the equation^{20},
where H is the Hamiltonian,
\(\widetilde {\boldsymbol{B}}(t) = \mathop {\sum}\nolimits_{\alpha = 1,2} {\left( {{\boldsymbol{B}}_\alpha {\mathrm{e}}^{  {\mathrm{i}}\omega _\alpha t} + {\mathrm{c}}.c.} \right)}\) is the magnetic field of the twofrequency (pump and probe) MW radiation, and 2D is the zerofield splitting. Optical initialization of the spin centers is described by the term \(G = \eta {\cal{D}}_0I_0\), where \({\cal{D}}_0\) is the spin quadrupole given by Eq. (6), η is a material parameter (η > 0 or η < 0 depending on whether the m_{S} = ±3/2 or m_{S} = ±1/2 spin states are preferentially populated at optical excitation), and I_{0} is the light intensity.
Spin relaxation is taken into account in Eq. (8) by the term
where \(\rho _{\cal{P}} = \mathop {\sum}\nolimits_{k =  1}^1 {{\mathrm{Tr}}} ({\cal{P}}_k^\dagger \rho ){\cal{P}}_k\), \(\rho _{\cal{D}} = \mathop {\sum}\nolimits_{k =  2}^2 {{\mathrm{Tr}}} ({\cal{D}}_k^\dagger \rho ){\cal{D}}_k\), and \(\rho _{\cal{F}} = \mathop {\sum}\nolimits_{k =  3}^3 {{\mathrm{Tr}}} ({\cal{F}}_k^\dagger \rho ){\cal{F}}_k\) are the dipole, quadrupole, and octupole contributions to the density matrix, respectively, T_{p}, T_{d}, and T_{f} are the corresponding relaxation times.
The spindensity matrix ρ found from Eq. (8) is used to calculate the spindependent correction to the PL intensity^{20}
where α is a material constant. The result is then averaged over the ensemble of centers with slightly different values of the zerofiled splitting D, as described above.
Semiclassical rate equations
Equation (8) is simplified in the basis of the eigenstates of the Hamiltonian (9) without the MW field. Disregarding the offdiagonal components of the density matrix, we obtain the semiclassical rate equations for the populations ρ_{ii} of the spin sublevels
Here, W_{ij} are the rates of spin transitions by the MW fields,
δω_{α} are the halfwidths of the MW field spectra, which are supposed to have the Lorentzian shape, \(G_i = \eta \left[ {{\cal{D}}_0} \right]_{ii}I_0\), and R_{ij} = R_{ji} are the component of the matrix of relaxation rates
For Bz, the components R_{ij} have a particular simple form
The PL correction is given by \({\mathrm{\Delta PL}}/{\mathrm{PL}} = \alpha \mathop {\sum}\nolimits_i \left[ {{\cal{D}}_0} \right]_{ii}\rho _{ii}\).
Data availability
The data supporting this study are available from the corresponding author upon reasonable request.
References
Awschalom, D. D., Bassett, L. C., Dzurak, A. S., Hu, E. L. & Petta, J. R. Quantum spintronics: engineering and manipulating atomlike spins in semiconductors. Science 339, 1174–1179 (2013).
Kaszlikowski, D. et al. Quantum cryptography based on qutrit Bell inequalities. Phys. Rev. A. 67, 656 (2003).
Lanyon, B. P. et al. Simplifying quantum logic using higherdimensional Hilbert spaces. Nat. Phys. 5, 134–140 (2008).
Shlyakhov, A. R. et al. Quantum metrology with a transmon qutrit. Phys. Rev. A. 97, 022115 (2018).
Kues, M. et al. Onchip generation of highdimensional entangled quantum states and their coherent control. Nature 546, 622–626 (2017).
Neeley, M. et al. Emulation of a quantum spin with a superconducting phase qudit. Science 325, 722–725 (2009).
Atatüre, M., Englung, D., Vamivakas, N., Lee, S.Y. & Wrachtrup, J. Material platforms for spinbased photonic quantum technologies. Nat. Rev. Mater. 3, 38–51 (2018).
Awschalom, D., Hanson, R., Wrachtrup, J. & Zhou, B. B. Quantum technologies with optically interfaced solidstate spins. Nat. Photonics 12, 516–527 (2018).
Kraus, H. et al. Roomtemperature quantum microwave emitters based on spin defects in silicon carbide. Nat. Phys. 10, 157–162 (2014).
Kimoto, T. & Cooper, J. A. Fundamentals of Silicon Carbide Technology: Growth, Characterization, Devices and Applications. (John Wiley & Sons, Singapore, 2014).
Falk, A. L. et al. Polytype control of spin qubits in silicon carbide. Nat. Commun. 4, 1819 (2013).
Biktagirov, T. et al. Polytypism driven zerofield splitting of silicon vacancies in 6Hsick. Phys. Rev. B 98, 195204 (2018).
Falk, A. L. et al. Electrically and mechanically tunable electron spins in silicon carbide color centers. Phys. Rev. Lett. 112, 187601 (2014).
Klimov, P. V., Falk, A. L., Christle, D. J., Dobrovitski, V. V. & Awschalom, D. D. Quantum entanglement at ambient conditions in a macroscopic solidstate spin ensemble. Sci. Adv. 1, e1501015 (2015).
Widmann, M. et al. Coherent control of single spins in silicon carbide at room temperature. Nat. Mater. 14, 164–168 (2015).
Simin, D. et al. Locking of electron spin coherence above 20 ms in natural silicon carbide. Phys. Rev. B 95, 161201(R) (2017).
Nagy, R. et al. Quantum properties of dichrioic silicon vacancies in silicon carbide. Phys. Rev. Appl. 9, 034022 (2018).
Mizuochi, N. et al. Continuouswave and pulsed EPR study of the negatively charged silicon vacancy with S = 3/2 and C_{3v} symmetry in ntype 4HSiC. Phys. Rev. B 66, 235202 (2002).
Carter, S. G., Soykal, Ö. O., Dev, P., Economou, S. E. & Glaser, E. R. Spin coherence and echo modulation of the silicon vacancy in 4HSiC at room temperature. Phys. Rev. B 92, 161202 (2015).
Tarasenko, S. A. et al. Spin and optical properties of silicon vacancies in silicon carbide  a review. Phys. Stat. Sol. (b) 255, 1700258 (2018).
Kraus, H. et al. Magnetic field and temperature sensing with atomicscale spin defects in silicon carbide. Sci. Rep. 4, 5303 (2014).
Christle, D. J. et al. Isolated electron spins in silicon carbide with millisecond coherence times. Nat. Mater. 14, 160–163 (2015).
Soltamov, V. A., Soltamova, A. A., Baranov, P. G. & Proskuryakov, I. I. Room temperature coherent spin alignment of silicon vacancies in 4H and 6HSiC. Phys. Rev. Lett. 108, 226402 (2012).
Sörman, E. et al. Silicon vacancy related defect in 4H and 6H SiC. Phys. Rev. B 61, 2613–2620 (2000).
Riedel, D. et al. Resonant addressing and manipulation of silicon vacancy qubits in silicon carbide. Phys. Rev. Lett. 109, 226402 (2012).
Kehayias, P. et al. Microwave saturation spectroscopy of nitrogenvacancy ensembles in diamond. Phys. Rev. B 89, 245202 (2014).
Mrózek, M. et al. Coherent population oscillations with nitrogenvacancy color centers in diamond. Phys. Rev. B 94, 552 (2016).
Simin, D. et al. AllOptical dc nanotesla magnetometry using silicon vacancy fine structure in isotopically purified silicon carbide. Phys. Rev. X 6, 031014 (2016).
Simin, D. et al. Highprecision angleresolved magnetometry with uniaxial quantum centers in silicon carbide. Phys. Rev. Appl. 4, 014009 (2015).
Niethammer, M. et al. Vector magnetometry using silicon vacancies in 4HSiC under ambient conditions. Phys. Rev. Appl. 6, 034001 (2016).
Fang, K. et al. Highsensitivity magnetometry based on quantum beats in diamond nitrogenvacancy centers. Phys. Rev. Lett. 110, 130802 (2013).
Fuchs, F. et al. Engineering nearinfrared singlephoton emitters with optically active spins in ultrapure silicon carbide. Nat. Commun. 6, 7578 (2015).
Muzha, A. et al. Roomtemperature nearinfrared silicon carbide nanocrystalline emitters based on optically aligned spin defects. Appl. Phys. Lett. 105, 243112 (2014).
Tairov, Y. M. & Tsvetkov, V. F. Investigation of growth processes of ingots of silicon carbide single crystals. J. Cryst. Growth 43, 209–212 (1978).
Acknowledgements
This work was supported by the German Research Foundation (DFG) under Grants DY 18/13 and AS 310/5, by the ERA.Net RUS Plus program and the German Federal Ministry of Education and Research (BMBF) within project DIABASE. V.A.S. was supported by the Alexander von HumboldtFoundation. A.V.P. and S.A.T. acknowledge the support by the RFBR #170200383 and the “BASIS” Foundation.
Author information
Authors and Affiliations
Contributions
V.A.S., C.K., A.S., and A.N.A. performed the experimental work. A.V.P. and S.A.T. developed the theory. E.N.M. grew the samples. V.A.S., P.G.B., G.V.A., and V.D. conceived the experiments and supervised the project. All the authors contributed to analysis of the data, discussions and to the writing of the paper.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Journal peer review information: Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Soltamov, V.A., Kasper, C., Poshakinskiy, A.V. et al. Excitation and coherent control of spin qudit modes in silicon carbide at room temperature. Nat Commun 10, 1678 (2019). https://doi.org/10.1038/s4146701909429x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s4146701909429x
This article is cited by

Inverted fine structure of a 6HSiC qubit enabling robust spinphoton interface
npj Quantum Information (2022)

Spin defects in hBN as promising temperature, pressure and magnetic field quantum sensors
Nature Communications (2021)

Uncomputability and complexity of quantum control
Scientific Reports (2020)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.