1

Enhanceddual-beamexcitationphotoelectricdetectionofNVmagnetic

resonancesindiamond

E.Bourgeoisa,b,E.Londeroc,K.Buczakd,Y.Balasubramaniamb,G.Wachterd,J.Stursae,K.Dobesf,F.Aumayrf,M.

Trupked,A.Galic,g,andM.Nesladeka,b

aIMOMECdivision,IMEC,Wetenschapspark1,B-3590Diepenbeek,Belgium.bInstituteforMaterialsResearch(IMO),Hasselt

University,Wetenschapspark1,B-3590Diepenbeek,Belgium.cInstituteforSolidStatePhysicsandOptics,WignerResearch

CentreforPhysics,HungarianAcademyofSciences,POBox49,H-1525Budapest,Hungary.dViennaCenterforQuantum

ScienceandTechnology,Atominstitut,TUWien,1020Vienna,Austria.eNuclearPhysicsInstitute,v.v.i.,ASCR,CZ-25068Rez,

CzechRepublic.fInstituteofAppliedPhysics,TUWien,WiednerHauptstr.8-10,1040Vienna,Austria.gDepartmentof

AtomicPhysics,BudapestUniversityofTechnologyandEconomics,Budafokiút8,H-1111Budapest,Hungary.

Date:2016-07-04

Abstract

ThecoreissuefortheimplementationofthediamondNVcentrequbitstechnologyisthesensitivereadoutof

NVspinstate.WehaverecentlydemonstratedthephotoelectricdetectionofNVmagneticresonances(PDMR),

anticipatedtobefasterandmoresensitivethanopticaldetection(ODMR).HerewereportonaPDMRcontrast

of9%-threetimesenhancedcomparedtopreviouswork-onshallowN-implanteddiamond.Basedonab-

initiomodelling,wedemonstrateanovelone-photonionizationdual-beamPDMRprotocol.Wepredictthat

thisschemeissignificantlylessvulnerabletotheinfluenceofdefectssuchassubstitutionalnitrogen.

2

Thenegativelychargednitrogen-vacancy (NV-)centre indiamondhasattractedparticularattentionasaroom

temperaturesolidstatequbit(1)thatcanberead-outbyopticaldetectionofmagneticresonances(ODMR)(2).

Numerousapplicationsinthefieldofsolid-statequantuminformationprocessing(3)andsensing(4)(5)(6)(7)(8)

are being studied, including non-perturbing nanoscale magnetometry with single NV- (9) and ultrasensitive

magnetometrywithNV-ensembles(10).

We have recently demonstrated the photoelectric detection ofNV- electron spinmagnetic resonances under

greenillumination(single-beamPDMR,ors-PDMR)(Fig.1a),basedontheelectricdetectionofchargecarriers

promotedtothediamondconductionband(CB)bytheionizationofNV-andperformeddirectlyonadiamond

chip equipped with electric contacts (11). Since the sensitivity of the magnetic resonance (MR) detection is

inversely proportional to the product between theMR contrast and the rootmean square of the number of

detectedphotonsNporelectronsNe(4)(12)(11),achievingahighphotocurrentsignalandasufficientMRcontrast

iscriticalforfurtherapplicationsofPDMRinquantumtechnology.Usings-PDMR,weachievedadetectionrateof

~107electrons.s-1perNV-(N.A.oftheobjective:0.95,greenillumination:3.4mW,electricfield:2.4104Vcm-1),

comparedto2104photons.s-1perNV-usingconfocalODMRdetection.However,theone-photonionizationof

single substitutional nitrogen (NS0) is one of the factors limiting the PDMR contrast, making a high green

illumination power necessary to achieve a sufficient contribution of NV- two-photon ionization to the total

photocurrent(11).Weobservedthatunderblueillumination,theionizationofNV-canbeachievedbyamore

effectiveone-photonprocess,enhancingtheproportionofthephotocurrentassociatedwithNV-comparedto

NS0. Based on this idea, we developed the dual-beam excitation PDMR (d-PDMR) scheme (Fig. 1a), that is

anticipated to lead to enhanced PDMR contrast in the case of sampleswith high [NS0]/[NV-] ratio and could

thereforeholdpromiseforthephotoelectricreadoutofsingleNV-spin(sincetheproportionofNS0inthedetection

volume remains substantial even in the case of single NV- centers contained in ultra-pure diamond) or for

ultrasensitivediamondmagnetometrywithNV-ensembles(forwhichirradiatedtype-Ibdiamondscontaininga

highproportionofNs0areused).

3

TodeterminethethresholdforNV-andNS0one-photonionization,wefirstmeasuredthephotocurrentspectraof

irradiatedtype-Ibdiamonds.Basedontheidentificationoftheionizationbands,wedesignedthed-PDMRscheme,

inwhichpulsedbluelight(2.75eV)directlypromoteselectronsfromNV-groundstatetotheCBandconvertsthe

resultantNV0backtoNV–.SimultaneousCWgreenillumination(2.33eV)independentlycontrolstheMRcontrast

byinducingspin-selectiveshelvingtransitionstoNV-metastablestate(13).Byperformingab-initiocalculationsof

Ns0,NV-andNV0ionizationcross-sections,wecouldexplorethephoto-physicsrelatedtotheproposedscheme,

relevantforachievinghigherMRcontrastandphotocurrentsignal.

ThesampleusedforPDMRmeasurements(sampleTP4)isanelectronicgradetype-IIadiamondimplantedwith

14N4+ions,resultingafterannealingintheformationofashallowNV-layer(density~30µm-2,depth:12±4nm).

Forphotocurrentspectroscopy,anas-receivedtype-Ibdiamondplate(sampleR,[NS0]~160ppm)wasusedasa

reference,whiletwootherswererespectivelyproton-(sampleA,[NV-]~35ppm)andelectron-irradiated(sample

E2,[NV-]~20ppm)andannealed.Coplanarelectrodeswithadistanceof100µm(samplesR,E2andA)or50µm

(sampleTP4)werepreparedonthesurfaceofthesediamondplates.Thetype-Ibsampleswerecharacterizedby

photoluminescence,FTIRandUV-visibleabsorptionspectroscopy(seesupplementaryinformation).

Torealizethed-PDMRscheme,aDCelectricfield(2.4104Vcm-1)isappliedinbetweenelectrodes.Acollimated

blue(2.75eV)laserbeam,pulsedat131Hz,isfocusedinbetweenelectrodesontothediamondsurfaceusinga

40Xairobjective(NA:0.95,lightspotdiameter~600nm).CWgreen(2.33eV)lightproducedbyalinearlypolarized

Nd:YAG laser is combined with the blue beam using the same objective. The resulting photocurrent is pre-

amplified and measured by a lock-in amplifier referenced to the blue light pulsing frequency, so that the

photocurrentinducedbyCWgreenlightdoesnotcontributetothemeasuredsignal.Thediamondchipismounted

onacircuitboardequippedwithmicrowaveantennas(14).Forphotocurrentspectroscopy,monochromaticlight

(1 to 300 µW) pulsed at 12Hz is focused onto the sample. At each photon energy, the photocurrent is pre-

4

amplified andmeasured by lock-in amplification (photocurrent detected down to 3 fA). The photocurrent is

normalizedtothefluxofincomingphotons.

To gain insight into NV-, NV0 and NS0 photo-ionization mechanisms, we apply ab-initio Kohn-Sham density

functionaltheory(DFT)calculations.Inthephoto-ionizationprocess,anelectronisexcitedfromanin-gapdefect

leveltotheCBorfromthevalenceband(VB)toanin-gapdefectlevel.Inourmeasurements,abiasvoltageis

applied to thesample,making the resultantelectronorhole instantly leave thedefects.Thephoto-ionization

probabilityisthendirectlyproportionaltotheabsorptioncross-sectionthatdependsontheimaginarypartofthe

dielectric function related to the transition between the initial ground state and the final excited state. This

processcanbewellapproximatedbythetransitionofasingleelectronfrom/tothein-gapdefectlevelto/from

thebandedges,thustheimaginarypartofthedielectricfunctioncanbecalculatedbetweenthecorresponding

Kohn-Shamlevels.Insummary,thetaskistocalculatetheexcitationenergiesandthecorrespondingimaginary

partofthedielectricfunction.

Tothisend,wecalculatethelowestexcitationenergythatcorrespondstothepureelectronictransition[zero-

phononline(ZPL)energy]bytheconstraintDFTapproach.Basedonourpreviousstudies(15)(16),weusearange-

separatedandscreenedhybriddensityfunctionalHSE06(17)(18).WeexplicitlycalculatedtheZPLenergiesonly

forthebandedges,andassumedthattheexcitationsathigherenergyfollowthecalculatedbandenergiesw.r.t.

thebandedgeenergy.Theimaginarypartofthedielectricfunctioniscalculatedatthegroundstategeometry,

followingtheFranck-Condonapproximation.Opticaltransitionstothebandsrequireaccuratecalculationofthe

electrondensityofstates.SinceHSE06calculationswithmanyk-pointsintheBrillouin-zonearecomputationally

prohibitive,andgiventhatPBEandHSE06Kohn-ShamwavefunctionsareverysimilarforNSandNVcentres,we

appliedageneralisedgradientapproximatedfunctionalPBEtocalculatetheiropticaltransitiondipolemoments

(19). The defects were modelled in a 512-atom diamond supercell. Details about ab-initio calculations are

presentedinsupplementaryinformation.

5

The proposed d-PDMR schemewas tested on shallowNV- ensembles implanted in electronic grade diamond

(sampleTP4).Asareference,s-PDMRwasmeasuredonthesamesample.Atafixedmicrowavepowerof1W,a

maximum s-PDMR contrast of 8.9 ± 0.3%was obtained (Fig. 1b), higher than the PDMR contrast previously

observedon type-Iband type-IIadiamond (11).Undergreen illumination thePDMRcontrast is limitedby the

backgroundphotocurrentresultingfromtheionizationofNs0(11).Theenhanceds-PDMRcontrastobservedon

sampleTP4canbeexplainedbythehighercontributionofthequadraticNV-two-photonionizationtothetotal

photocurrent (see supplementary Fig. 1),which is due to the confinement of the defects to thewaist of the

illumination beam – where the intensity is highest. By contrast, the illumination intensity in bulk samples

decreaseswithdepth,leadingtoahigherproportionoflinearNS0ionizationinthetotalphotocurrent.

Figure1.(a)Schematicdiagramofthes-PDMRandd-PDMRschemes(b)Comparisonbetweend-PDMR(pulsedblueexcitation:226µW,CWgreenexcitation:9.1mW)and s-PDMR (pulsedgreenexcitation:3mW) spectrameasuredonshallowNV-ensembles,intheconditionsleadingtomaximumMRcontrast(sampleTP4,microwavepower:1W).

Amaximumd-PDMRcontrastof9.0±0.4%isobtainedonsampleTP4(Fig.1b),closetothemaximums-PDMR

contrast observed on the same sample at identical microwave power. Measurements of photocurrent as a

functionofgreenandbluelightpoweronsampleTP4(supplementaryFig.1band2b)showinadditionthatat

identicalpower,blueone-photonexcitationinduceshigherphotocurrentthangreenexcitation.Forexample,the

(a)

Lock-inamplifier

+

Pulsedgreenlight

SinglecrystaldiamondSingle-beamPDMR

NV0 NV- +hNV- NV0 +e

hν =2.33eV

CWgreenlight

Singlecrystaldiamond

Pulsedbluelight

NV0 NV- +hNV- NV0 +e

hν =2.75eV

Dual-beamPDMRMW

antenna

MWantenna

Lock-inamplifier

+

2820 2840 2860 2880 2900 29200.90

0.95

1.00

1.05

1.10

Lorentzian fit

Dual-beam PDMR

Norm

alized photocurrent (a.u.)

Nor

mal

ized

pho

tocu

rrent

(a.u

.)

Microwave frequency (MHz)

(b)

Single-beam PDMR

0.85

0.90

0.95

1.00

6

photocurrentmeasuredunder226µWexcitation(conditionsleadingtomaximald-PDMRcontrast)isfivetimes

higherunderblue(800fA)thanundergreenlight(165fA).

AlthoughonshallowimplantedNV-ensemblestheMRcontrastsobtainedbyd-PDMRands-PDMRaresimilar,the

d-PDMRschemecouldpotentiallyleadtohighercontrastthans-PDMRincaseofsampleswithhigh[NS0]/[NV-]

ratio,duetothelowercontributionofNS0ionizationtothetotalphotocurrentunderblueillumination.Indeed,

considering the green light power dependence of the photocurrent measured on type-Ib sample E2

(supplementaryFig.1a),under4mWgreenexcitationthetwo-photonionizationofNV-(quadraticfractionofthe

photocurrent)representsonly~1.5%(0.6pA)ofthetotaldetectedphotocurrent,whileab-initiocalculations

indicatethatunder4mWblueillumination~30%(0.6nA)ofthetotalphotocurrentoriginatesfromthe1-photon

ionizationofNV-(seesupplementaryFig.8).

Toexplorethemechanismbehindthed-PDMRscheme,westudiedtheenergydependenceofNVandNSphoto-

ionizationcross-sectionsbyphotocurrentspectroscopy.Thesemeasurementswereperformedonirradiatedand

annealed type-Ibdiamonds, sinceahighdensityofdefects allows to reachahighdynamic rangeofdetected

photocurrent,leadingtoaprecisedeterminationofphoto-ionizationthresholds.

The photocurrent spectrum measured on a type-Ib reference diamond (sample R, [NS0] ~ 160 ppm) can be

observedinFig.2a.Thisspectrumdisplaysaphoto-ionizationbandwithathresholdionizationenergyof~2.2eV,

obtainedbyfittingexperimentaldatatoInkson’sformulaforthephoto-ionizationcross-sectionofdeepdefects

(20).ThisbandcorrespondstotheionizationofNS0toNS

+(21)(22).Thoughitscalculated(+|0)pureelectronic

chargetransitionlevelisatEC-1.7eV(EC:CBminimum)thegiantredistributionofpositionoftheNandCatoms

inthecoreofthedefectuponionizationofNS0resultsinalowionizationcross-sectionatthisenergy.Duetovery

strongelectron-phononinteraction,aphoto-ionizationbandemergesinthephononsidebandsaroundEC-2.2eV

(see(23)and(24)forfurtherdiscussion).

7

Figure 2.Measured and calculated photo-ionization bands. (a) Comparison between photocurrent spectrameasuredonreferencetype-Ibdiamond[sampleR,Ei=2.18(3)eV]andirradiatedandannealedtype-Ibdiamonds[samplesA,Ei=2.66(4)eV;sampleE2,Ei=2.69(3)eV].Ei:thresholdionizationenergyfromInkson’sfitting.(b)Fitting of photocurrent measured on sample A using calculated ionization cross-sections. In the ab-initiocalculationitwasassumedthatNV-,NV0andNS

0ionizationdominatesthespectrum,withotherparasiticdefectscontributingtoasmallextenttothespectruminthelowenergyregion(notshown).

Compared tonon-irradiateddiamond, thephotocurrent spectrameasuredonproton- andelectron-irradiated

type-Ibdiamonds(samplesAandE2,inwhich~10%ofNSdefectsareconvertedtoNV-centres)showablueshift

andtheformationofanionizationbandwiththresholdat~2.7eV(Fig.2a).Photoluminescence,FTIRandoptical

absorptionspectroscopyindicatethatthedominantdefectsinthesesamplesareNSandNV,withsomeadditional

spuriousdefects(possiblyassociatedtoNi)insampleA(seesupplementarynote3).

Wecalculatedtheionizationcross-sectionsofNSandNVdefectsasafunctionofthephoto-excitationenergy(see

supplementarynote4)andcompared the resultswith thephotocurrentmeasurements (Fig.2b for sampleA,

supplementaryFig.8forsampleE2).Inthesimulationplotswesettheexperimentalvalueof[NS0]andfit[NV0]

and[NV–]totheexperimentaldata.Usingonlythesetwofittingparametersweobtained[NV–]≈31.4ppmand

[NV0]≈1.0ppminsampleA,inexcellentagreementwiththeconcentrationsdeterminedfromphotoluminescence

measurements([NV–]≈34.0ppmand[NV0]≈1.1ppm).Ourab-initiocalculationspredicted(25)thatthe(0|–)

acceptorlevelofNVliesjustinthemiddleofthediamondgap,atEC-2.75eV.UnlikeNS,NV0andNV–presentvery

8

similar geometries, thus pure electronic transitions dominate the ionization process. Photonswith an energy

above2.7eVcanthereforeionizeNV–toNV0bypromotinganelectrontotheCB,butalsoconvertNV0backto

NV–bydirectpromotionofanelectronfromtheVB.CalculationspredictinadditionthatNV0has~10timeslarger

ionizationcross-sectionthanNV–at2.75eV,implyingalargerratefortheback-conversionthanfortheionization

ofNV-.

Figure3.d-PDMRonNV-centres. (a) Schematicdiagramof thed-PDMRmechanism(not toscale). Left:one-photon ionization of NV-. Right: Back-conversion from NV0 to NV-. RS: resonant state. (b) d-PDMR contrastmeasuredon shallowNV- ensembles as a functionof the ratioRGB between the green andblue light powers(sampleTP4).Errorbarsrepresentthestandarderrorsofthefittingparameters.

Basedontheresultsofphotocurrentspectroscopyandab-initiocalculations,weexplainthed-PDMRschemeas

follows(Fig.3a).Pulsedbluelight(2.75eV)promoteselectronsfromNV–tripletgroundstate3A2totheCBbya

one-photonprocess(transition1),andinducesalsotheone-photonback-conversionfromNV0toNV-(transition

4).SimultaneousilluminationbyCWgreenlaserlight(2.33eV)inducestransitionsfromthegroundstate3A2to

theexcitedstate3E(transition2),followedbyspin-selectivenon-radiativedecayfromthe|±1>spinmanifoldof

3Etothesingletstate1A1(transition3)(13).Fromthere,electronsfalltothemetastablestate1E(220nslifetime)

(27).Atresonantmicrowavefrequency(2.87GHz),theseshelvingtransitionsresultinatemporarydecreasein

theoccupationofNV–groundstate,andthustoadecreaseinthephotocurrentassociatedwiththeone-photon

3A2

3E

MW2E

2A1

Free electron

|0>

|0>

NV-

NV0

1A1

1E

|±1>

|±1>

12

3

(a)

RS CB

VB

Free holeRS 0 20 40 60 80 100 120 140 160 180

0

2

4

6

8

10

Fixed green light power (9.1 mW) Fixed blue light power (226 µW)

PDM

R c

ontra

st (%

)

Ratio between green and blue light power

(b)

9

ionizationofNV–.Hereweassumethatthephoto-ionizationcross-sectionfromthe1EshelvingstatetotheCBis

low,althoughthemetastablestate1Estatehasbeenrecentlyestimatedtobelocated~0.4eVaboveNV-ground

state(i.e.~2.3eVbelowtheCB)(28)andcouldthereforetheoreticallybeionizedby2.75eVphotons.However,

thenegativeresonancesobservedind-PDMR(Fig.1b)indicatethatthecontributionofthisprocesstothetotal

photocurrentissignificantlylowerthanthecontributionofdirecttransitionsfromNV-groundstatetotheCB.

Ourd-PDMRmodelindicatesthattherelativeratesbetweenthedirectionization,back-conversionandshelving

transitionstothemetastablestate,whichcanbecontrolledbyvaryingtheratioRGBbetweenthegreenandblue

lightpowers,aredominantlyresponsibleforthePDMRcontrast.Atafixmicrowavepower(1Winthepresented

experiments), the PDMR contrasts obtainedby varying the green light power at constant blue power andby

varyingthebluepoweratconstantgreenpowerpresentasimilartrend(Fig.3b),whichindicatesthatintherange

oflaserpowerconsideredhereandforRGB<40,thed-PDMRcontrastriseswithRGB.Thed-PDMRschemeallows

thusanindependentcontrolofthephotocurrentintensity(bythebluelightpower)andtheMRcontrast(byRGB).

Theincreaseinthed-PDMRcontrastwithRGB(observedbelowRGB≈40)canbeexplainedbythetransferofan

increasedproportionofelectronsinitiallyinthe|±1>spinmanifoldtothemetastablestate.AboveRGB≈40,the

contrastsaturatesandslightlydecreases.Itshouldbenotedthatthiseffectdoesnotresultfromthesaturationof

thesingletstate1E,sinceitoccurswhenthebluelightpowerisreducedatfixedgreenlightpower.

Inconclusion,wedemonstratedthataone-photonionizationschemecanbeusedforreadingoutthespinstate

ofNV-. Based on this principle,we designed a novel photoelectric scheme for the detection ofNV-magnetic

resonances,inwhichblueilluminationinducestheone-photonionizationofNV–andconvertsNV0backtoNV-,

while theMRcontrast is independentlycontrolledbyCWgreen light.AmaximalPDMRcontrastof9.0%was

obtainedonshallowNV–centresimplantedinelectronicgradediamond.Thed-PDMRschemeisexpectedtobe

lesssensitivethans-PDMRtobackgrounddefectsindiamondandtoleadthustoenhancedMRcontrastinthe

caseofsampleswithhigh[NS0]/[NV-]ratio.Thisrobustphotoelectricdetectionschemecouldthereforerepresent

10

animportantsteptowardthephotoelectricreadoutofsingleNV-spinstateandbeusedfortheconstructionof

diamondquantumopto-electronicsdeviceswithenhancedperformances.

11

Bibliography

1.Doherty,M.W.,etal.1,2013,Phys.Rep.,Vol.528,pp.1-45.

2.Jelezko,F.andWrachtrup,J.13,2006,phys.stat.sol.(a),Vol.203,pp.3207-3225.

3.Childress,L.andHanson,R.February2013,MRSBulletin,Vol.38,pp.134-138.

4.Rondin,L.,etal.5,2014,Rep.Prog.Phys.,Vol.77,p.056503.

5.Hong,S.,etal.2013,MRSBulletin,Vol.38,pp.155-161.

6.Schirhagl,R.,etal.2014,Annu.Rev.Phys.Chem.,Vol.65,pp.83-105.

7.Balasubramanian,G.,etal.2014,Curr.Opin.Chem.Biol.,Vol.20,pp.69-77.

8.Shi,F.,etal.2015,Science,Vol.347,pp.1135-1138.

9.Maletinski,P.,etal.2012,Nat.Nanotechnol.,Vol.7,pp.320-324.

10.Wolf,T.,etal.2015,Phys.Rev.X,Vol.5,p.041001.

11.Bourgeois,E.,etal.2015,Nat.Com.,Vol.6,p.doi:10.1038/ncomms9577.

12.Dréau,A.,etal.2011,Phys.Rev.B,Vol.84,p.195204.

13.Tetienne,J.-P.,etal.2012,New.J.Phys.,Vol.14,p.103033.

14.Mrozek,M.,etal.2015,AppliedPhysicsLetters,Vol.107,p.013505.

15.Gali,A.,etal.2009,Phys.Rev.Lett.,Vol.103,p.186404.

16.Deàk,P.,etal.2010,Phys.Rev.B,Vol.81,p.15320.

17.Heyd,J.,Scuseria,G.E.andErnzerhof,M.2003,J.Chem.Phys.,Vol.118,p.88207.

18.Krukau,A.V.,etal.2006,J.Chem.Phys.,Vol.125,p.224106.

19.Perdew,J.P.,Burke,K.andErnzerhof,M.1996,Phys.Rev.Lett.,Vol.77,pp.3865-3868.

20.Inkson,J.1981,Phys.C:SolidStatePhys.,Vol.14,pp.1093-1101.

21.Nesladek,M.,etal.25,1998,Appl.Phys.Lett.,Vol.72,pp.3306-3308.

22.Rosa,J.,etal.1999,Diam.Relat.Mater.,Vol.8,pp.721-724.

23.Enckevort,W.J.P.andVersteegen,E.H.1992,J.Phys.:Condens.Matter,Vol.4,p.2361.

24.Thiering,G.,etal.Tobesubmitted.

12

25.Deàk,P.,etal.2014,Phys.Rev.B,Vol.89,p.075202.

26.Siyushev,P.,etal.2013,Phys.Rev.Lett.,Vol.110,p.167402.

27.Acosta,V.M.,etal.2010,Phys.Rev.B,Vol.82,p.201202.

28.Goldman,M.L.,etal.2015,Phys.Rev.B,Vol.91,p.165201.

29.Bube,R.H.Photoconductivityofsolids.s.l.:Wiley,1960.

Acknowledgements

SupportfromEU(FP7projectDIADEMS,grantNo.611143) isacknowledged.A.G.acknowledgestheLendület

programoftheHungarianAcademyofSciences.TheauthorsthankA.JarmolaandD.BudkerfromtheDepartment

ofPhysicsoftheUniversityofCalifornia(Berkeley,California)forthepreparationoftheelectron-irradiatedtype-

Ibdiamond.

Authorcontributions

E.B.andK.B.performedtheexperiments.E.B.processedthePDMRdataandperformedtheanalysis.E.L.andA.G.

carriedouttheab-initiodevelopmentsandcalculations.Y.B.preparedelectrodesontype-Ibdiamondsamples.

G.W.designedandbuiltthebluediodelaseranddesignedtheMWantennas.J.S.preparedtheproton-irradiated

type-Ibdiamond.K.D.andF.A.performedthediamond ion implantation.M.T.proposedtheuseof implanted

defects,designedtheelectrodesandassembledthedevice.M.N.,M.T.andA.G.supervisedthework.E.B.,A.G.

andM.N.wrotethemanuscript.Allauthorsdiscussedtheresultsandcommentedonthemanuscript.

SupplementaryInformationaccompaniesthispaper.

Competingfinancialinterests:Theauthorsdeclarenocompetingfinancialinterests.