Cavity-enhanced measurements of defect spins in silicon carbide
CCavity-enhanced measurements of defect spins in silicon carbide
Greg Calusine, Alberto Politi,
1, 2 and David D. Awschalom
1, 3, ∗ Department of Physics, University of California, Santa Barbara, California 93106, USA School of Physics and Astronomy, University of Southampton, Southampton SO17 1BJ, United Kingdom. Institute for Molecular Engineering, University of Chicago, Chicago, Illinois 60637, USA
The identification of new solid-state defect qubit candidates in widely used semi-conductors has the potential to enable the use of nanofabricated devices for enhancedqubit measurement and control operations. In particular, the recent discovery of opti-cally active spin states in silicon carbide thin films offers a scalable route for incorpo-rating defect qubits into on-chip photonic devices. Here we demonstrate the use of 3Csilicon carbide photonic crystal cavities for enhanced excitation of color center defectspin ensembles in order to increase measured photoluminescence signal count rates,optically detected magnetic resonance signal intensities, and optical spin initializationrates. We observe up to a factor of 30 increase in the photoluminescence and ODMRsignals from Ky5 color centers excited by cavity resonant excitation and increase therate of ground-state spin initialization by approximately a factor of two. Furthermore,we show that the small excitation mode volume and enhanced excitation and collectionefficiencies provided by the structures can be used to study inhomogeneous broadeningin defect qubit ensembles. These results highlight some of the benefits that nanofab-ricated devices offer for engineering the local photonic environment of color centerdefect qubits to enable applications in quantum information and sensing.
I. INTRODUCTION
Electronic spins associated with color center de-fects in silicon carbide (SiC) show promise as a poten-tial component for solid-state quantum technologies dueto their combination of long coherence times,[1] room-temperature operation,[2] and a host material for whichmature growth[3] and fabrication protocols exist.[4] Theability to engineer the local photonic environment of op-tically active solid-state qubits through the use of micro-fabrication techniques is crucial for their scalable appli-cation to the field of quantum information.[5] In contrastto the most widely studied forms of SiC (4H and 6H), theavailability of cubic 3C-SiC as a single crystal heteroepi-taxial layer on silicon opens up the possibility of combin-ing the favorable properties of SiC defect qubits with thefabrication capabilities available in III-V and Silicon-on-Insulator semiconductor systems. This may enable on-chip, scalable architectures for generating,[6] routing,[7]manipulating,[8] and detecting[9] single photon emissionfrom defect qubits in SiC. These capabilities can be com-bined with small mode volume optical cavities [10] [11][12] [13] to utilize strong Purcell enhancements[7] [14] forefficient single photon generation or to use cavity QEDprotocols to realize a spin-photon interface [15] as a nodein an optically connected ‘quantum network’.[16] [17]To date, only modest quality factors (‘Q’) of upto 1,550 have been demonstrated for small mode vol-ume (‘ V m ’) cavities ( V m ∼ ( λ/n ) ) operating at wave-lengths relevant for coupling to known defect qubit op-tical transitions in 3C-SiC.[18] Nevertheless, low- andmodest-Q microcavities can be used to improve defectqubit optical readout and control schemes. More specif-ically, microcavities can be used to compensate for on-chip power coupling limitations [19] to facilitate opti- cal coherent control schemes,[20] generate on-chip fre-quency conversion,[21] and improve rates of absorptionand fluorescence.[22] [23] [24] In this work, we utilizecavity resonant excitation to increase the photolumines-cence (PL) count rate, optically detected magnetic res-onance (ODMR) signal intensity, and rate of spin stateinitialization for Ky5 defect spin ensembles incorporatedinto 3C-SiC photonic crystal cavities. Furthermore, weuse these signal improvements and techniques to studyinhomogeneity in the Ky5 defect spin and optical prop-erties and extract estimates of the defects’ sublevel tran-sition rates.The Ky5 point defect is a color center in 3C-SiCthat has been previously demonstrated to exhibit spinand optical properties that are similar to the negativelycharged nitrogen vacancy (NV) center in diamond. [25]It produces an optical emission band consisting of a zerophonon line (ZPL) around 1118 nm and a red-shiftedphonon sideband extending out to approximately 1300nm and, like the NV center, its S = 1 spin ground statecan be initialized and read out optically. Spin coher-ence times ( T ) in excess of 20 µ s have been demon-strated for the ground-state spin sublevels of Ky5 en-sembles. These states can be manipulated with resonantmicrowave pulses on fast time scales (tens of nanosec-onds) even up to room temperature, making Ky5 centersa viable candidate for a defect-based spin qubit. TheKy5 center can be controllably generated within the 3C-SiC crystal lattice through a combination of radiationdamage and subsequent annealing and has been tenta-tively identified as a neutral divacancy.[26] Due to theavailability of 3C-SiC as a heteroepitaxial film on sili-con, photonic structures with three-dimensional opticalconfinement can be fabricated by removing the underly-ing substrate through conventional silicon wet and dry a r X i v : . [ c ond - m a t . m e s - h a ll ] O c t FIG. 1. (a) Simulation of the L3 cavity electric field intensity enhancement relative to an incident Gaussian beam (1 µ m beamwaist indicated by the dashed red lines) with unity electric field amplitude maximum. Note the change in scale above and belowthe dashed green line. The dashed white lines delineate the cross sections of the photonic crystal cavity holes. (b) Diagramdepicting the CEPLE scheme overlaid on top of the off-resonantly excited PL spectrum of an L3 cavity with incorporated Ky5centers at 20K. The wavelength of the fundamental mode of the cavity is near the peak of the inhomogeneously broadenedZPL. etch processes. II. CAVITY-ENHANCEDEXCITATION
The use of on-chip photonic cavities to enhancedefect qubit excitation is analogous to the use ofmacroscopic enhancement cavities to generate largeintracavity optical fields for applications such as highsensitivity absorption spectroscopy in atomic gases[27]or efficient non-linear optical frequency conversion.In general, for a fixed incident excitation power, thelocal electric field intensity inside a microcavity scalesas | E | ∝ ηQV m where η is the input power couplingefficiency. More specifically, temporal coupled-modetheory can be used to calculate the response of aphotonic crystal cavity subject to far field excitationby an externally incident Gaussian beam, yielding theexpression: | E c | | E | = Qηλn o w o n c V m (1)where | E c | is the cavity electric field intensity maxi-mum, | E | is the incident beam electric field intensitymaximum, λ is the wavelength of the cavity mode, n o is the refractive index of the medium surrounding the cav-ity, w o is the incident beam waist, and n c is the cavityindex of refraction ( ∼ ∼ µ m. The resulting simulated elec-tric field intensity enhancement of 193.48 agrees within1% of the value of calculated from Eq. 1 using the in-put parameters extracted from simulations for the L3cavity design. See [28] for a comparison of the localfield enhancements and cavity parameters for differentstructure designs and a more detailed discussion of thesimulations.To observe these cavity enhanced field intensities,we measured photonic crystal cavities that were designedto exhibit fundamental modes tuned to the inhomoge-neously broadened ZPL of the Ky5 centers incorporatedinto the cavity [around 1118 nm, see Fig. 1(b)]. The cav-ities consisted of L3 and H1 structures fabricated in 300nm thick 3C-SiC films that were implanted with carbonions at an energy of 110 KeV and annealed at 750 o C for30 minutes in order to generate ensembles of Ky5 defects.Details of the sample design, fabrication, and character-
FIG. 2. (a) SEM image of the 3C-SiC L3 cavity structure. (b) and (c) 15 µ m by 15 µ m spatial scans of the sideband PLintensity signal for the excitation wavelengths depicted in (d)((b): λ =1115.1 nm, (c): λ =1117.1 nm). The white dashedlines indicate the extent of the photonic crystal. The count rate measured on the cavity in (c) was ∼
48 kCts/s. (d) SidebandPL count rate vs. excitation wavelength measured at the position of the cavity. (e) Linecut along the dashed red line in (c). ization are presented in Ref. [18]. We primarily focusedon L3 designs to observe cavity-enhanced optical excita-tion due to their greater degree of coupling to far fieldGaussian modes as compared to the H1 cavities, whichwere better suited for improved narrowband collection ofoff-resonantly excited defect PL (10-20 times improve-ment over unpatterned thin films). The measured L3structures exhibited Q ∼
900 with simulated mode vol-umes of .68 ( λ/n ) and far field coupling efficiencies toexternal, free space Gaussian modes of 9.6%.[28]All measurements were performed in a home-builtscanning confocal microscope with an integrated he-lium flow cryostat at 20K. A 1064 nm diode laser wasused for off-resonant excitation and a tunable (1090-1180nm), narrow linewidth ( <
300 kHz) Littman-Metcalfdiode laser was used for resonant photoluminescenceexcitation and cross-polarized resonant scattering spec-troscopy. Samples were mounted with the cavity axis at45 degrees with respect to the incident laser polarizationto allow for the use of the latter technique for controlmeasurements. Collected PL and reflected laser lightpassed back through the objective and were detected us-ing an InGaAs CCD array, a low noise femtowatt pho-toreceiver, or a superconducting nanowire single photondetector (SNSPD). See [28] for a detailed description ofthe experimental apparatus and control measurements.We determined the wavelength of the cavity reso-nances by measuring the PL spectrum of emitters local-ized within the cavity under off-resonant 1064 nm exci-tation as shown in Fig. 1(b). Figure 1(b) also depictsthe excitation and PL collection wavelength ranges of the cavity-enhanced photoluminescence excitation (CE-PLE) spectroscopy measurement. In order to observeenhanced excitation of the Ky5 defects using the cavitymode, we tuned the excitation wavelength to the cav-ity resonance peak where the light is absorbed by ex-citing the ZPL transitions and collect red-shifted side-band PL. To determine the total PL signal count rateincrease with respect to the unpatterned, released thinfilm, we performed a series of spatial scans over the pho-tonic crystal cavity area and compared the overall cavityPL count rate to that of the surrounding thin film. Fig-ure 2(a) shows a scanning electron microscope (SEM)image of the L3 cavity and Fig. 2 (b) and (c) show apair of scanning confocal PL images of a 12 µ m by 12 µ mL3 photonic crystal cavity with excitation wavelengthsas designated in the Fig. 2(d). The photonic crystalextent is delineated by the white dashed lines and thebright spot in its center is fluorescence originating fromthe cavity. Figure 2(d) shows the excitation wavelengthdependence of the PL count rate originating from thecavity location. The excitation wavelength-dependentcount rate matched the cavity mode spectrum and thepeak exhibited an approximately factor of 5 increase overoff-resonant excitation. Figure 2(e) shows a line cut ofthe PL map corresponding to the red dashed line in Fig.2(c). At the excitation wavelength corresponding to thecavity resonance, the PL count rate was approximatelyΓ ≈
30 times higher than the PL count rate from the un-patterned thin film, where we have defined Γ as the ratioof the resonantly excited cavity PL count rate to the thin
FIG. 3. (a) Optical image of a photonic crystal cavity array with a 50 µ m wide 10 nm/300nm Ti/Au microstrip positioned 50 µ m away from the cavities. (b) ODMR signal at 1.319 GHz measured on the photonic crystal cavity (black line and dots) andon the 3C-SiC thin film (red line and dots) for excitation at the same power at the cavity resonance wavelength. (c) PulsedODMR signal from a Ky5 ensemble subject to cavity-enhanced excitation (red line and dots) as compared to an ensemblewithin the thin film (black line and dots). The blue line and dots show the same measurement for defects within the thin filmfor a higher optical excitation power (offset for clarity). film PL count rate at the same excitation wavelength andpower. For H1 designs, we observed a lower maximumΓ ≈
13 due to a smaller degree of input coupling for anincident Gaussian beam. Scattered laser light added anegligible contribution to the measured signal (see Ref.[28]). While we excited cavity modes tuned to the Ky5defect ZPL wavelength range, this same approach can beapplied to applications that require efficient off-resonantexcitation[29] and would be particularly beneficial forexcitation wavelengths that overlap weakly with the de-fects’ absorption spectrum.Aside from increasing the overall luminescencecount rate, cavity resonant excitation can also be used toimprove measurements on single emitters or ensemblesby exciting a significantly smaller sample volume thana standard objective lens configuration. The excitationvolume of a thin film is greatly reduced as compared tothat of bulk material because the excitation volume di-mension along the optical axis is set by the thickness ofthe thin film rather than the diffraction limit ( ∼ µ m fora .7 NA objective at λ =1.1 µ m). Furthermore, for ouroptical configuration, the cavity provides a further 12.4-fold reduction in the sample excitation volume as com-pared to the thin film, resulting in an overall 705-fold re-duction as compared to bulk material.[28] This reduction in the excitation of localized states in the proximity ofisolated single emitters within the surrounding materialcan improve the performance of single photon sources byreducing background fluorescence[30] or charge-inducedspectral diffusion.[31] III. OPTICALLY DETECTEDMAGNETIC RESONANCE SIGNALENHANCEMENTS
The observed signal improvements provide ameans to greatly increase the ODMR signal amplitudein order to probe the Ky5 center’s spin-dependent elec-tronic structure or for sub-diffraction limit, on-chip sens-ing applications. In order to perform spin-dependentmeasurements on defects within the cavity structure, weperformed an additional fabrication step that adds a 10nm/300 nm Ti/Au metallization layer to the sample sur-face for applying intense local microwave fields to thesample. Figure 3 (a) shows an optical image of an arrayof released 3C-SiC films patterned with photonic crys-tal cavities next to a 50 µ m wide microstrip positioned50 µ m away from the structures. Due to the robust-ness of the approximately 40 µ m by 40 µ m freestandingfilms, the metallization can be applied prior to or afterthe membrane release step without the need for criti-cal point drying. We used on-chip microstrips in orderto apply sufficiently intense microwave fields to achievecoherent spin manipulation (Rabi oscillations) on timescales faster than the Ky5 defects’ T ∗ of ∼
50 ns.[25]Figure 3(b) compares the Ky5 ODMR signal (∆
P L )with the excitation beam incident on the released 3C-SiCthin film (black line and dots) and the photonic crystalcavity (red line and dots) for the same optical power atzero magnetic field. The overall ODMR signal increasematches the PL count rate increase observed for thiscavity (Γ ≈ m s = 0 and m s = ± ions cm − implantation dose) like thosemeasured in the Fig. 3, we observed variations in thedefect ground-state zero field splitting (D ≈ ions cm − dose), the 10-20 times improve-ment of collection efficiency [18] for defects emitting intothe cavity mode reveals PL lines as narrow as 25 GHzwithin the inhomogeneously broadened ZPL (28.2 nmFWHM ≈ µ W vs. 3.2 mW) and higher PLcount rates ( ∼ ∼
200 cts/s) as comparedto narrowband PL collection of the emission line usinga scanning monochromator.[28] Exciting and collectingPL from a reduced mode volume facilitates the isolationof a smaller number of spectral lines within a given wave-length range because this spectral fine structure and itspolarization can vary throughout the thin film, likely dueto variations in the local crystalline environment. [33]
IV. ENHANCED SPININITIALIZATION RATES
In addition to providing increased signal intensi-ties, the excitation enhancement provided by the reso-nant cavity mode can also be used to increase the rateof optically-induced spin polarization in the Ky5 defects’ground state. To observe this process, we measuredthe defects’ time-dependent PL intensity in response tovariable-length laser pulses with interleaved microwave π pulses driving the m s = 0 to m s = ± π pulseto invert the ground-state spin polarization, which wasthen re-pumped to the steady state polarization by thesubsequent optical excitation. We measured this differ-ence in PL (PL | (cid:105) − PL |± (cid:105) ) between optical pulses thatimmediately followed a π pulse and those that did not toobtain a measure of the ground-state spin polarization.The optical excitation was turned off for at least 500 nsprior to microwave manipulation and readout to allowfor population within the ISC levels to fully relax to theground state. See [28] for pulse sequences and furtherdetails.The same model used to explain the PL dynamicsof the NV center in diamond can be applied to model theexpected dynamics.[34] Figure 4(a) shows the results ofnumerical simulations of the time-dependent differencein PL between an NV center that has been initially pre-pared in the m s = 0 state and the m s = ± FIG. 4. (a) Simulated time-dependent difference in PL for an NV center initialized into the m s = 0 vs. m s = ± excited within the thin film with those excited using thephotonic crystal cavity mode, as shown in Fig. 4(b).Figure 4(b) also shows the results of numerical simu-lations of this difference signal for varying laser pulselengths. The observed saturation behavior matches thepredicted dynamics for intersystem crossing rates thatdiffer from the NV center, as determined from a nonlin-ear least squares fit to a numerical model that includesthe expected excitation enhancement ( | E c | / | E | ∼ τ . The time constant as-sociated with this approximately exponential saturationof the PL difference signal decreases from 500 ±
55 ns to280 ±
20 ns. This approximately factor of 2 differencein the time constant for optical pumping is relativelysmall as compared to the overall PL count rate increase( ∼ π pulse and observed no difference signal.[28] These measurements provide the first direct obser-vation of spin-dependent PL dynamics in SiC and sup-port previous assumptions that these defects exhibit op-tical dynamics similar to the NV center in diamond.Furthermore, we corroborated the variable length laserpulse results using TCSPC methods and observed a spin-dependent pulsed ODMR contrast of 2.75%. By combin-ing this value with the numerical modeling, we estimate an intrinsic spin-dependent pulsed ODMR contrast of4.9% for ideal measurement parameters (for details, seeRef. [28]). While this value is lower than typically ob-served for single NV centers, it is similar to what hasbeen previously observed for high density NV center en-sembles and single defects in SiC. [35] [1] V. CONCLUSIONS AND OUTLOOK
In conclusion, we have demonstrated resonantexcitation of 3C-SiC photonic crystal cavities withintegrated defect spins for large PL and ODMR signalenhancements and increased spin initialization rates.Our analysis shows that our present cavity designs arecapable of achieving localized optical field intensitiesthat can be enhanced by a factor of almost 200 relativeto an incident Gaussian beam. This value could befurther improved with concurrent optimization of cavityQ and coupling to the Gaussian excitation mode.[36]For measurements of ensembles of defects, these smallmode volumes and excitation intensity enhancementsmay facilitate on-chip applications that are limitedby inefficient optical coupling or poor spectral overlapof the excitation source and defect absorption bands.These applications include spin ensemble-based sensingtechniques using on-chip optical traps,[37] enhancedabsorption for hole-burning experiments,[38] or for en-hancing the signal of spectrally distinct sub-ensemblesin order to study inhomogeneous broadening.[32]For applications involving single defects, cavity res-onant excitation can provide enhanced optical starkshifts,[20] compact, on-chip single photon frequencyconversion,[21] or reduced excitation of backgroundimpurities that are detrimental to single photon sourceperformance.[30] [39] Additionally, we have providedevidence that defects in 3C-SiC exhibit photodynamicssimilar to those of the NV center in diamond. Theseresults underscore the benefits of fabricating deviceswith integrated defect spin states in heteroepitaxial3C-SiC as a means to incorporate defect qubits intoscalable device architectures for applications in the fieldof quantum information and sensing.We thank David Christle, Bob Buckley, andJoerg Bochmann for helpful discussions. This work wassupported by the AFOSR QuMPASS MURI FA9550-12-1-004 and NSF DMR-1306300. A portion of this workwas done in the UC Santa Barbara nanofabricationfacility, part of the NSF funded NNIN network. Weacknowledge support from the Center for ScientificComputing from the CNSI, MRL: an NSF MRSEC(DMR-1121053) and NSF CNS-0960316. ∗ [email protected][1] David J. Christle, Abram L. Falk, Paolo Andrich,Paul V. Klimov, Jawad Ul Hassan, Nguyen T. Son, ErikJanzen, Takeshi Ohshima, and David D. Awschalom,“Isolated electron spins in silicon carbide with millisec-ond coherence times,” Nat Mater , 160–163 (2014).[2] William F. Koehl, Bob B. Buckley, F. Joseph Heremans,Greg Calusine, and David D. Awschalom, “Room tem-perature coherent control of defect spin qubits in siliconcarbide,” Nature , 84–87 (2011).[3] A. Powell, “Growth of sic substrates,” Int. J. High SpeedElectron. Syst. , 751–777 (2006).[4] C.M. Zetterling and Institution of Electrical Engineers, Process Technology for Silicon Carbide Devices , EMISprocessing series (INSPEC, 2002).[5] Marko Lonˇcar and Andrei Faraon, “Quantum pho-tonic networks in diamond,” MRS Bulletin , 144–148(2013).[6] I. J. Luxmoore, N. A. Wasley, A. J. Ramsay, A. C. T.Thijssen, R. Oulton, M. Hugues, S. Kasture, V. G.Achanta, A. M. Fox, and M. S. Skolnick, “Interfac-ing spins in an ingaas quantum dot to a semiconductorwaveguide circuit using emitted photons,” Phys. Rev.Lett. , 037402 (2013).[7] Andrei Faraon, Arka Majumdar, Dirk Englund, ErikKim, Michal Bajcsy, and Jelena Vuˇckovi´c, “Integratedquantum optical networks based on quantum dots andphotonic crystals,” New Journal of Physics , 055025(2011).[8] J. E. Kennard, J. P. Hadden, L. Marseglia,I. Aharonovich, S. Castelletto, B. R. Patton, A. Politi,J. C. F. Matthews, A. G. Sinclair, B. C. Gibson,S. Prawer, J. G. Rarity, and J. L. O’Brien, “On-chipmanipulation of single photons from a diamond defect,”Phys. Rev. Lett. , 213603 (2013).[9] G. Reithmaier, S. Lichtmannecker, T. Reichert,P. Hasch, K. Muller, M. Bichler, R. Gross, and J. J.Finley, “On-chip time resolved detection of quantum dot emission using integrated superconducting single photondetectors,” Sci. Rep. , 1901 (2013).[10] Shota Yamada, Bong-Shik Song, Seungwoo Jeon,Jeremy Upham, Yoshinori Tanaka, Takashi Asano, andSusumu Noda, “Second-harmonic generation in a silicon-carbide-based photonic crystal nanocavity,” Opt. Lett. , 1768–1771 (2014).[11] Marina Radulaski, Thomas M. Babinec, Sonia Buck-ley, Armand Rundquist, J Provine, Kassem AlAssaad,Gabriel Ferro, and Jelena Vuˇckovi´c, “Photonic crystalcavities in cubic (3c) polytype silicon carbide films,” Opt.Express , 32623–32629 (2013).[12] Jonathan Y. Lee, Xiyuan Lu, and Qiang Lin, “High-q silicon carbide photonic-crystal cavities,” AppliedPhysics Letters , 041106 (2015).[13] David O. Bracher and Evelyn L. Hu, “Fabrication ofhigh-q nanobeam photonic crystals in epitaxially grown4h-sic,” Nano Letters , Nano Lett. , 6202–6207 (2015).[14] Andrei Faraon, Charles Santori, Zhihong Huang, Vic-tor M. Acosta, and Raymond G. Beausoleil, “Cou-pling of nitrogen-vacancy centers to photonic crystal cav-ities in monocrystalline diamond,” Phys. Rev. Lett. ,033604 (2012).[15] Luozhou Li, Tim Schrder, Edward H. Chen, MichaelWalsh, Igal Bayn, Jordan Goldstein, Ophir Gaathon,Matthew E. Trusheim, Ming Lu, Jacob Mower, MirceaCotlet, Matthew L. Markham, Daniel J. Twitchen, andDirk Englund, “Coherent spin control of a nanocavity-enhanced qubit in diamond,” Nat Commun , 6173(2015).[16] H. J. Kimble, “The quantum internet,” Nature ,1023–1030 (2008).[17] W. Pfaff, B. J. Hensen, H. Bernien, S. B. van Dam,M. S. Blok, T. H. Taminiau, M. J. Tiggelman, R. N.Schouten, M. Markham, D. J. Twitchen, and R. Hanson,“Unconditional quantum teleportation between distantsolid-state quantum bits,” Science , 532–535 (2014).[18] Greg Calusine, Alberto Politi, and David D.Awschalom, “Silicon carbide photonic crystal cavitieswith integrated color centers,” Applied Physics Letters , 011123 (2014).[19] Hyun-Joo Chang, Se-Heon Kim, Yong-Hee Lee, Emil P.Kartalov, and Axel Scherer, “A photonic-crystal opti-cal antenna for extremely large local-field enhancement,”Opt. Express , 24163–24177 (2010).[20] R. Bose, D. Sridharan, G. S. Solomon, and E. Waks,“Large optical stark shifts in semiconductor quantumdots coupled to photonic crystal cavities,” AppliedPhysics Letters , 121109 (2011).[21] Murray W. McCutcheon, Darrick E. Chang, YinanZhang, Mikhail D. Lukin, and Marko Loncar, “Broad-band frequency conversion and shaping of single pho-tons emitted from a nonlinear cavity,” Opt. Express ,22689–22703 (2009).[22] Ren-Jye Shiue, Xuetao Gan, Yuanda Gao, LuozhouLi, Xinwen Yao, Attila Szep, Dennis Walker, JamesHone, and Dirk Englund, “Enhanced photodetectionin graphene-integrated photonic crystal cavity,” AppliedPhysics Letters , 241109 (2013).[23] K. Jensen, N. Leefer, A. Jarmola, Y. Dumeige, V. M.Acosta, P. Kehayias, B. Patton, and D. Budker,“Cavity-enhanced room-temperature magnetometry us-ing absorption by nitrogen-vacancy centers in diamond,”Phys. Rev. Lett. , 160802 (2014).[24] X. Liu, T. Shimada, R. Miura, S. Iwamoto, Y. Arakawa, and Y. K. Kato , “Localized guided-mode and cavity-mode double resonance in photonic crystal nanocavi-ties,” Phys. Rev. Applied , 014006 (2015).[25] Abram L. Falk, Bob B. Buckley, Greg Calusine,William F. Koehl, Viatcheslav V. Dobrovitski, AlbertoPoliti, Christian A. Zorman, Philip X.-L. Feng, andDavid D. Awschalom, “Polytype control of spin qubitsin silicon carbide,” Nat Commun , 1819 (2013).[26] V.Ya. Bratus, R.S. Melnik, S.M. Okulov, V.N. Rodionov,B.D. Shanina, and M.I. Smoliy, “A new spin one defectin cubic sic,” Physica B: Condensed Matter , 4739 –4741 (2009).[27] Jun Ye and Theresa W. Lynn, “Applications of opti-cal cavities in modern atomic, molecular, and opticalphysics,” (Academic Press, 2003) pp. 1 – 83.[28] See Supplemental Material at [URL will be inserted bypublisher] for further details.,.[29] Hannah Clevenson, Matthew E. Trusheim, Carson Teale,Tim Schroder, Danielle Braje, and Dirk Englund,“Broadband magnetometry and temperature sensingwith a light-trapping diamond waveguide,” Nat Phys ,393–397 (2015).[30] Masahiro Nomura, Satoshi Iwamoto, Toshihiro Nakaoka,Satomi Ishida, and Yasuhiko Arakawa, “Localized ex-citation of ingaas quantum dots by utilizing a photoniccrystal nanocavity,” Applied Physics Letters , 141108(2006).[31] F. Jelezko, I. Popa, A. Gruber, C. Tietz, J. Wrachtrup,A. Nizovtsev, and S. Kilin, “Single spin states in a defectcenter resolved by optical spectroscopy,” Applied PhysicsLetters , 2160–2162 (2002).[32] Eric van Oort and Max Glasbeek, “Frequency-dependentdephasing of n-v centers in diamond,” Journal of Lumi-nescence , 88 – 91 (1992). [33] Matteo Bosi, Giovanni Attolini, Marco Negri, CesareFrigeri, Elisa Buffagni, Claudio Ferrari, Tiziano Rimoldi,Luigi Cristofolini, Lucrezia Aversa, Roberta Tatti, andRoberto Verucchi, “Optimization of a buffer layer for cu-bic silicon carbide growth on silicon substrates,” Journalof Crystal Growth , 84 – 94 (2013).[34] N. B. Manson, J. P. Harrison, and M. J. Sellars,“Nitrogen-vacancy center in diamond: Model of the elec-tronic structure and associated dynamics,” Phys. Rev. B , 104303 (2006).[35] B. J. Maertz, A. P. Wijnheijmer, G. D. Fuchs, M. E.Nowakowski, and D. D. Awschalom, “Vector magneticfield microscopy using nitrogen vacancy centers in dia-mond,” Applied Physics Letters , 092504 (2010).[36] Simone L. Portalupi, Matteo Galli, Christopher Rear-don, Thomas Krauss, Liam O’Faolain, Lucio C. An-dreani, and Dario Gerace, “Planar photonic crystalcavities with far-field optimization for high coupling ef-ficiency and quality factor,” Opt. Express , 16064–16073 (2010).[37] Thijs van Leest and Jacob Caro, “Cavity-enhanced opti-cal trapping of bacteria using a silicon photonic crystal,”Lab Chip , 4358–4365 (2013).[38] R T Harley, M J Henderson, and R M Macfarlane,“Persistent spectral hole burning of colour centres in di-amond,” Journal of Physics C: Solid State Physics ,L233 (1984).[39] L. C. Bassett, F. J. Heremans, C. G. Yale, B. B. Buck-ley, and D. D. Awschalom, “Electrical tuning of sin-gle nitrogen-vacancy center optical transitions enhancedby photoinduced fields,” Phys. Rev. Lett.107