Optical switching of defect charge states in 4H-SiC
OOptical switching of defect charge states in 4 H -SiC D. Andrew Golter and Chih Wei Lai ∗ U.S. Army Research Laboratory, 2800 Powder Mill Rd, Adelphi, MD 20783 (Dated: November 13, 2018)We demonstrate optically induced switching between bright and dark charged divacancy defects in4 H -SiC. Photoluminescence excitation and time-resolved photoluminescence measurements revealthe excitation conditions for such charge conversion. For an energy below 1.3 eV (above 950 nm),the PL is suppressed by more than two orders of magnitude. The PL is recovered in the presence ofa higher energy repump laser with a time-averaged intensity less than 0.1% that of the excitationfield. Under a repump of 2.33 eV (532 nm), the PL increases rapidly, with a time constant 30 µ s.By contrast, when the repump is switched off, the PL decreases first within 100-200 µ s, followedby a much slower decay of a few seconds. We attribute these effects to the conversion between twodifferent charge states. Under an excitation at energy levels below 1.3 eV, V Si V are converted intoa dark charge state. A repump laser with an energy above 1.3 eV can excite this charged state andrecover the bright neutral state. This optically induced charge switching can lead to charge-statefluctuations but can be exploited for long-term data storage or nuclear-spin-based quantum memory. Optically active point defects (color centers) in wide-band-gap semiconductors can possess long electron spincoherence times ( > – ) in diamond, in which the elec-tron spin state can be initialized with non-resonant op-tical excitation and detected via photoluminescence con-trast at ambient conditions. Recently, similar defectshave been identified in silicon carbide (SiC) for use inwafer-scale quantum technologies . SiC crystals form inthree main polytypes – 4 H , 6 H (hexagonal), and 3 C (cu-bic) – offering a broad range of defects that can actas potential spin qubits, among which are the carbonantisite-vacancy pair (C Si V C ) , the nitrogen vacancy(N C V Si ) , the silicon monovacancy V –Si 10–13 , and theneutral divacancy (V Si V ) . FIG. 1. PLE spectra under an excitation energy ( E ex ) from1.24 to 1.77 eV at T = 10 K. The four major zero-phonon-line(ZPL) emissions are labeled as PL1( hh ), PL2( kk ), PL3( hk ),and PL4( kh ). The broad background emissions are from thephonon side bands of V Si V C and V Si . The arrow indicatesone of the Raman LO and TO peaks. In this work, we focus on the 4 H − SiC polytype, inwhich the two adjacent carbon and silicon vacancies ofV Si V organize in either axial ( hh , kk ) or basal ( hk , kh )configurations as a result of varying lattice sites and ori-entations. We identify the optimal pump laser energy byusing photoluminescence excitation (PLE) measurements(see Methods). A unique zero phonon line (ZPL) is as-sociated with each configuration and is labeled as PL1–PL4 with emission energies (wavelengths) E ZP L ≈ λ ZP L ≈ I ZP L ) are proportional to the phonon-side-band (PSB) absorption and peak under an excitationenergy E ex ∼ E ex istuned toward E ZP L , I ZP L is expected to decrease gradu-ally with slowly decreasing PSB absorption. Surprisingly, I ZP L decreases precipitously by more than two orders ofmagnitude for E ex (cid:46) I ZP L occurs at distinct E ex fordefects at inequivalent lattice sites.To better determine the energy at which the ZPLis suppressed precipitously, we plot the spectrally-integrated I ZP L for individual ZPL in Fig. 2. We alsofind that that this PL suppression at lower energies isreversed by the addition of a 2.33 eV (532 nm) repumpfield with a time-averaged intensity about 0.1% that ofthe excitation field, as shown by the black curves for PL2and PL4 in Fig. 2a. The four ZPLs vary in I ZP L , likelydue to a combination of difference in defect density, radia-tive recombination efficiency, and polarization- or dipole-orientation-dependent optical collection efficiency. Thus,we plot the ratio of I ZP L with and without repump tobetter display the transitions for all four defect typesPL1–PL4 on the same plot (Fig. 2b). The excitationenergy at which ZPL is dramatically suppressed rangesfrom approximately 1.28 to 1.32 eV.Next, we determine the dynamics of the repump andsuppression processes with time-resolved PL measure-ments (Fig. 3). The sample is excited continuously by a a r X i v : . [ c ond - m a t . m t r l - s c i ] O c t FIG. 2. (a) Spectrally integrated ZPL intensities ( I ZPL ) ofPL2( kk ) axial and PL4( kh ) basal defects with (black) andwithout (red) the 532 nm repump laser as a function of thepump (excitation) energy. Pump laser power is maintained at20 mW, while the repumping 532-nm laser power is 0.2 mW.(b) Ratios of I ZPL with and without the repump (ZPL ratio)for four types of divacancies as labeled in Fig. 1.FIG. 3. Time-resolved measurement of the total PL fromV Si V during the repump and suppression processes. Thepump laser is on continuously while the repump laser ispulsed. Insets show the fast initial growth in PL when therepump laser is turned on and the suppression when the re-pump is turned off. Results are shown for a repump at 532nm (black) as well as at 910 nm (red). The 910 nm curveis scaled (x20) and offset. The rise time of the repump pro-cess is shorter for the higher energy repump. After an initialrise over 10-100 µ s, PL reaches a plateau with a much slowergrowth (100’s of ms), likely due to defects near the peripheryof the excitation spot which see a significantly lower excita-tion intensity. The decay after the 910 nm repump is identicalto that for 532 nm. laser at 1.27 eV (976 nm). To switch the defects betweenthe bright and dark states, we use a pulsed repump laserwith a time-averaged intensity < ≈ µ s. In con-trast, when the repump is switched off, the PL decreasesfirst within ≈ µ s, followed by a much slower de-cay of a few seconds. The non-exponential rise and fallof the PL become more evident under a repump of 1.362eV (910 nm). In this case, a rapid sub-ms surge in PLis followed by a gradual increase over a few ms, whilethe PL decay remains similar to that observed under the 532-nm excitation.The fast initial rise times depend on the repump power.The repump power that was chosen for the plots in Fig.3 was close to saturation for this process. For instance,we found that decreasing the 532nm repump power byalmost an order of magnitude increased the rise time byabout a factor of two. From our model we would expectthe repump process to depend on both repump energyand intensity. However accurate determination of suchenergy and power dependence is not feasible from ourensemble measurements, owing to the fact that the localintensity experienced by each defect varies for differentregions of the excitation spot.We attribute these effects to the conversion betweentwo different charge states, as observed in NV /NV – centers in diamond . Under an excitation at energylevels below 1.3 eV, V Si V are converted into a darkcharge state, where the system becomes trapped. A re-pump laser with an energy above 1.3 eV can excite thischarged state and recover the bright neutral state. Thisconversion is most effective with a repump above 1.4 eV. FIG. 4. Schematics of the charge conversion process for theneutral and positive states of the V Si V C defect. (a) VV toVV + conversion involves two photons and an Auger processthat release sufficient energy to ionize an electron from thedefect. (b) The a orbital of VV + lies in the valence band.When an electron is excited from the continuum states in thevalence band (or a orbital, not shown) to the e orbital, thehole migrates away from the defect, converting the V Si V C center back to the neutral state. We consider the conversion between the neutral andpositive charge (V Si V +C ) states (Fig. 4). In the V Si V C divacancy, there are two a and two e states, which areformed from by the six dangling bonds. The position ofthese defect energy levels varies with the occupation ofthe states, as wells as with the relative crystalline positionof the divacancy pair. In the neutral charge state, fourelectrons occupy the two a states, and two electrons oc-cupy the lower, degenerate e state . The upper a andtwo e levels are within the gap (Fig. 4a), resulting in theneutral divancancies undergoing atomic-like transitions.A charge neutral defect can be ionized and converted intoa positive charge state through two-photon absorption,followed by the Auger process (Fig. 4a). In V Si V C , thedefect energy levels of the non-zero charge states remainlargely unknown, though the formation energies of sta-ble charged states, namely, +, 0, − , and −
2, have beencalculated . For the positive charge state, we envisionthat the upper a state lies about 0.1–0.15 eV below thevalence band maximum (VBM), while the lower e stateis likely about 1.3 eV above the VBM (Fig. 4b).Under an excitation energy below 1.3 eV but above theVV ZPL transition, the conversion from VV to VV + remains effective; however, the defect becomes ’trapped’in the VV + because the optical excitation of an elec-tron out of the valence band is suppressed. Under anexcitation of energy above 1.3 eV, an optically excitedelectron occupies one of the lower e states, while the holerapidly migrates away from the defect (i.e., an electronis captured) (Fig. 4b). As a result, the defect reverts tothe neutral charge state. This charge reversion involvesonly one-photon absorption and is expected to occur at ahigher rate than ionization, which is consistent with ourmeasurements (insets in Fig. 3). When this hole migra-tion rate is much higher than that of the radiative recom-bination, the positive charge state is optically dark, as isobserved experimentally. The aforementioned electron-capture rate increases with increasing optical excitationenergy from approximately 1.28 to 1.4 eV, and becomesnearly constant for excitation energies exceeding the a - e transition energy, about 1.4 eV. The hypothesis of anincreasing electron-capture rate with increasing repumpenergies is supported by the distinct switching dynamicsunder a 532-nm or 910-nm repump (Fig. 3).In this model, we expect the energy difference betweenthe upper a orbital and VBM to vary for defects in in-equivalent lattice sites, resulting in distinct ’threshold’excitation energies as shown in Fig. 2. By contrast,the energy gap between the upper and lower a orbitalsshould be insignificant as suggested by the experimentalobservation that PL1–PL4 all peak around E ex ≈ . They model thisconversion based on switching between neutral and neg-ative charge (V Si V –C ) states. The cycling between neu-tral and negative charge states requires the inclusion ofother shallow donors surrounding the divancancies. Ourexperimental results do not preclude such a scenario. Todetermine whether the dark state is positively or nega-tively charged, it is necessary to examine optically in- duced switching of charge states of single defects andcompare with accurate DFT calculations of the electronicstructures of these charge states.In conclusion, we demonstrate that V Si V C divacanciescan become charged via optical excitation. In diamonds,both NV and NV – are optically active with identifi-able ZPLs. By contrast, the PLE and PL measurementsin V Si V C in 4 H -SiC suggest that the charged V Si V –C or V Si V +C states are optically dark. This optically in-duced charge switching can lead to charge-state fluctua-tions but can also be exploited for long-term data storageor nuclear-spin-based quantum memory, as shown for NVcenters in diamonds . METHODS
Sample.
The sample is a high-purity semi-insulating(HPSI) 4 H -silicon carbide substrate purchased fromNorstel. The V Si V C divacancies are naturally formedwithout additional electron/proton irradiation or anneal-ing. The density of V Si V C defects is estimated to beabout 10 to 10 cm − . Similar optical induced chargeswitching effects are also observed in HPSI SiC substratepurchased from Cree, Inc. Setup.
Excitation and repump lasers are focused toa ∼ µ m area on the sample via a microscope objectivewith NA = 0.75. A tunable single-frequency laser (MSquared Lasers SolsTiS) is used for the excitation with aconstant time-averaged power of 20 mW ( ± Si V divacancies toa superconducting nanowire single photon detector (Sin-gle Quantum Eos). The repump laser pulses are createdusing an acousto-optic modulator. ACKNOWLEDGEMENTS
This work was supported by Office of Secretary of De-fense, Quantum Science and Engineering Program. Theauthors thank D. E. Taylor, A. Beste, B. VanMill, and D.D. Awschalom for discussions and providing SiC samples. ∗ [email protected] Koehl, W. F., Seo, H., Galli, G. & Awschalom, D. D. De-signing defect spins for wafer-scale quantum technologies.
MRS Bull. , 1146–1153 (2015). Iwamoto, N. & Svensson, B. G.
Point Defects in SiliconCarbide , chap. 10, 369–407. Semiconduct. Semimet. (Else-vier Inc., 2015). Umeda, T. et al.
Identification of the carbon antisite-vacancy pair in 4 H -SiC. Phys. Rev. Lett. , 145501 (2006). Steeds, J. W. Photoluminescence study of the carbonantisite-vacancy pair in 4 H - and 6 H -SiC. Phys. Rev. B , 245202 (2009). Castelletto, S. et al.
A silicon carbide room-temperaturesingle-photon source.
Nat. Mater. , 151 (2013). Sz´asz, K. et al.
Spin and photophysics of carbon-antisitevacancy defect in 4 H silicon carbide: A potential quantumbit. Phys. Rev. B , 121201(R) (2015). von Bardeleben, H. J., Cantin, J. L., Rauls, E. & Gerst-mann, U. Identification and magneto-optical properties ofthe NV center in 4 H -SiC. Phys. Rev. B , 064104 (2015). von Bardeleben, H. J. et al. NV centers in 3 C , 4 H , and 6 H silicon carbide: A variable platform for solid-state qubitsand nanosensors. Phys. Rev. B , 121202(R) (2016). Zargaleh, S. A. et al.
Evidence for near-infrared photolu-minescence of nitrogen vacancy centers in 4 H -SiC. Phys.Rev. B , 060102(R) (2016). Baranov, P. G. et al.
Silicon vacancy in SiC as a promisingquantum system for single-defect and single-photon spec-troscopy.
Phys. Rev. B , 125203 (2011). Widmann, M. et al.
Coherent control of single spins insilicon carbide at room temperature.
Nat. Mater. , 164(2014). Simin, D. et al.
High-precision angle-resolved magnetome-try with uniaxial quantum centers in silicon carbide.
Phys.Rev. Appl. , 014009 (2015). Simin, D. et al.
All-optical dc nanotesla magnetometryusing silicon vacancy fine structure in isotopically purified silicon carbide.
Phys. Rev. X , 031014 (2016). Son, N. T. et al.
Divacancy in 4 H -SiC. Phys. Rev. Lett. , 055501 (2006). Koehl, W. F., Buckley, B. B., Heremans, F. J., Calusine,G. & Awschalom, D. D. Room temperature coherent con-trol of defect spin qubits in silicon carbide.
Nature ,84 (2011). Christle, D. J. et al.
Isolated electron spins in silicon car-bide with millisecond coherence times.
Nat. Mater. ,160–163 (2015). Beha, K., Batalov, A., Manson, N. B., Bratschitsch, R. &Leitenstorfer, A. Optimum photoluminescence excitationand recharging cycle of single nitrogen-vacancy centers inultrapure diamond.
Phys. Rev. Lett. , 097404 (2012). Siyushev, P. et al.
Optically controlled switching of thecharge state of a single nitrogen-vacancy center in diamondat cryogenic temperatures.
Phys. Rev. Lett. , 167402(2013). Gordon, L., Janotti, A. & Van de Walle, C. G. Defectsas qubits in 3 C - and 4 H -SiC. Phys. Rev. B , 045208(2015). Iwata, J.-I., Shinei, C. & Oshiyama, A. Density-functionalstudy of atomic and electronic structures of multivacanciesin silicon carbide.
Phys. Rev. B , 125202 (2016). Wolfowicz, G. et al.
Optical charge state control of spindefects in 4 H -SiC. arXiv:1705.09714 (2017). Dhomkar, S., Henshaw, J., Jayakumar, H. & Meriles, C. A.Long-term data storage in diamond.
Sci. Adv.2