Generation of THz radiation at KEK LUCX facility K. Lekomtsev on behalf of LUCX collaboration JAI Seminar: Introduction Seminar by Recently Started Research Staff
Generation of THz radiation at KEK LUCX facility
K. Lekomtsev on behalf of LUCX collaboration
JAI Seminar: Introduction Seminar by Recently Started Research Staff
Personal Introduction
JAI Seminar, Oxford University 2
KonstantinLekomtsev
Currently:Marie– CurieresearchfellowatRoyalHollowayUniversityofLondon.
2012– 2016:PostdoctoralresearcheratHighEnergyAcceleratorResearchOrganization(KEK),Tsukuba,Japan.
• AnalyticalstudiesandSimulationsoftheradiativephenomenainelectronaccelerators(Transition,Diffraction,Cherenkov,Smith-Purcelletc.).
• Highpowerfslasersystemtuningandmaintenance,experimentalstudiesattheLaserUndulator CompactX-rayfacility(LUCX).• etc.
2009– 2012:Marie– Curieearlystageresearcher(PhDstudent)atRoyalHollowayUniversityofLondon
• AnalyticalstudiesofcoherentDiffractionradiation.• ExperimentalstudyatCLICTestFacility3atCERN.• etc.
Earlier: NationalResearchNuclearUniversity(MoscowEngineeringPhysicsInstitute)
• MasterdegreeinAppliedMathematics.
Introduction
JAI Seminar, Oxford University 3
1.OverviewoftheLUCXfacility.a.Beamparameters.b.Multi-bunchbeamgeneration.
2.MonochromaticityofcoherentTHzSmith-Purcellradiation(SPR).a.Brieftheoreticalbackground.b.ParticleInCellsimulationsofSPRspectrum.c.Discussionofexperimentaldata.
3.CherenkovSmith– Purcellradiation(ChSPR)fromcorrugatedcapillary.a.Brieftheoreticalbackgroundandcomparisonwithsimulations.b.ParticleInCellsimulationsoftheradiationfrommulti-bunchbeamandcapillarywithreflector.c.Discussionofexperimentaldata.
LUCX facility: Overview
JAI Seminar, Oxford University 4
“Femtosecondmode”n Ti:Sa lasern e-bunchRMSlength~100fsn e-bunchcharge<100pCn Singlebunchtrain,Micro-bunching4-16(4isconfirmed)n TypicalRep.rate3.13Hzn Experiments:THzprogram
“Picosecondmode”n Q-switchNd:YAG lasern e-bunchRMSlength~10psn e-bunchcharge<0.5nCn Multi-bunchtrain2- few103
n MaxRep.rate12.5Hzn Experiments:Compton,CDR
• TheLaserUndulator CompactX-rayfacility(LUCX)isamultipurposelinearacceleratorwhichwasinitiallyconstructedasanRFguntestbenchandlaterextendedtofacilitateComptonscatteringandcoherentradiationgenerationexperiments.
LUCX laser system
JAI Seminar, Oxford University 5
• FemtoseconddurationelectronbuncheswithTHzrepetitionfrequency illuminateaphotocathodeandandelectronbeamisgeneratedonasingleRFacceleratingfieldcycle.
• Titanium– Sapphire“ChirpedPulseAmplificationtechnique”lasersystemisusedtogenerateasequenceofmicrobunches.
Micro-bunch beam generation and characterization
JAI Seminar, Oxford University 6
• TheRMSelectronbunchlengthismeasuredbythezero-phasingmethod*.Timecorrelatedmomentumdeviation imposedonthebunch ifweoperateatthezerocrossingoftheacceleratingwave.
• ThebeamisthendispersedbyadipolemagnetBH1GsothatthedifferenttimeslicesoftheelectronbunchareprojectedontoascintillatingYAGscreenatdifferenthorizontalpositions, andthus beamimageonthescreenshows theintensity distributionoftheelectronbunchalongitstemporalprofile.
• Thecorrelation oftheRFphasewiththeYAGbeamimageshiftismeasured.• ThelinearcorrelationofthisapproximationgivesthescaleofthehorizontalimagesizeinRFdegrees,whichcanberecalculatedtotimescale.
*D.X.Wangetal,Phys.Rev.E57,2283 (1998).
Monochromaticity of SPR
7
Smith-Purcellradiationappearswhenchargedparticlesmoveaboveandparalleltoadiffractiongrating.Spectrallinespositionsaredefinedbythedispersionrelation*:
𝜆" =$"
%& − 𝑐𝑜𝑠𝜃 ; (1)
where 𝜆"isthewavelengthoftheresonanceorder𝑘, 𝑑 isthegratingperiod, 𝛽 istheparticleintheunitsofthespeedoflight,and 𝜃 istheobservationangle.
Thespectral-angulardistributionofthecoherentSPR** producedfromgratingwithfinitenumberofperiodsN:
$/0$1$2 =
$/03$1$2
456/ 78456/ 8 𝑁: + 𝑁: 𝑁: − 1 𝐹 ; (2)
𝜑 = 𝑑 ?1@ 𝛽A% − 𝑐𝑜𝑠𝜃 - phaseassociatedwithstripsperiodicity, $
/03$1$2 isthespectralangulardistributionfromasinglegratingperiod,
𝜈 isradiationfrequency,𝑁 isthenumberofgratingperiods,𝑁: isthebunchpopulation,and𝐹 isthebunchform-factor.
From(2),ifFWHMistakenasanabsolutespectrallinewidth,themonochromaticity isdefinedas:
∆DD =
E.GH"7 . (3)
*S.J.SmithandE.M.Purcell,Visiblelightfromlocalizedsurfacechargesmovingacrossagrating,Phys.Rev.92,1069 (1953).**A.P.Potylitsynetal.,DiffractionRadiationfromRelativisticParticles,Springer(2010).
Monochromaticity of SPR
JAI Seminar, Oxford University 8
Measurementslimitations:
ThewidthofSPRspectrallinescanbecomelargeriftheyaremeasuredbythedetectorplacedinthesocalled“pre-wavezone”.IfthegratingtodetectordistanceisL,thenthefar-fieldzone(orwavezone)condition isdeterminedby*:
𝐿 ≫ 𝐿KK = 𝑘𝑁L𝑑 1 + 𝑐𝑜𝑠𝜃 .
Monochromaticityoftheradiationgeneratedfromaninfinitegrating(𝑁 →∞) andmeasuredwithafiniteaperturedetector∆𝜃:
∆DD =
456OPQA@R4O
∆𝜃.
Assumingthatthereallineshape𝛿𝜆T andthespectrometerresolution𝛿𝜆4U canbeapproximatedbyaGaussiandistribution,theFWHMofthemeasuredline:
𝛿𝜆 = 𝛿𝜆TL + 𝛿𝜆4U
L�.
*D.V.KarlovetsandA.P.Potylitsyn,JETPLetters84,489 (2006).
Particle in Cell simulations (SPR)
JAI Seminar, Oxford University 9
Simulation parameter Value
L 500mm
D 60mm
h 0.6mm
d 4 mm
𝛼 30deg.
Bunch length 0.5ps
Bunchtransversesize 250𝜇m
Beamenergy 8MeV
• SimulationswereperformedinCSTParticleStudio,ParticleInCellSolver.• Consideredtwocalculationdomainsinordertoshowthe influenceof
thepre-wavezoneeffectforthefirstdiffractionorderofSPR.
• SPRspectrumobtainedbyrecordingelectricfieldcomponentsasfunctionsoftime andthenbyperformingFouriertransformofthetimedependenceofthedominantcomponent.
Particle in Cell simulations (SPR)
JAI Seminar, Oxford University 10
• Boththesimulationandthetheoryshowthathigherorderspectrallinesbecomemoremonochromatic.
• Whencomparingthelinewidthsforthetheoryandthesimulation,itisimportanttorememberthatthetheorywasdevelopedfor𝑁 → ∞andnottakingintoaccountrealshapeofthegrating.
ComparisonoftheSPRspectrallinewidths:
Experimental study (SPR)
JAI Seminar, Oxford University 11
• Vacuumwindow:12mmthick2deg.wedgedsapphire,witheffectiveapertureof145mm.
• 5-axismanipulatorsystemwasinstalledonthetopofthevacuumchamber.Usedforfineadjustment ofthegratingpositions in3orthogonal directionsandalsoforthecontrolforthe2rotationalangles.
• Thegratingwasalignedwithrespecttotheelectronbeamusing theforwardbremsstrahlungappearingduetodirectinteractionoftheelectronbeamwiththetargetmaterial.
Experimental study (SPR)
JAI Seminar, Oxford University 12
MeasurednormalizedTransitionRadiationspectrumcanbeusedasspectralefficiencyoftheentiremeasurementsystem,including:
ü spectraltransmission efficiencyofthevacuumwindow,ü detectorwavelengthefficiency,ü splitterefficiency,ü reflectioncharacteristicsofthemirrorsandabsorption inair.
SpectralresolutionofFourierspectrometer: 𝛿𝜆 defined asFWHMofthespectralpeakfromamonochromatic source:
YDD= 1.21 D
L[\]^,
where 𝐿56_ istheinterferometermaximumopticalpathdifferencefromzeroposition.
TransitionRadiationangularscanusing SBD320- 460
TransitionRadiationspectralmeasurements: Applying thiscriteriontotheinterferograms,thespectrometerresolution:
YDPDP
= 15%; YDbDb
= 6%
ThemeasuredpeaksFWHM:
YDPd
DP= 16%; YDb
d
Db= 6.1%
SBD60– 90GHzSBD320– 460GHz
Theory and simulation of THz radiation from corrugated channel
JAI Seminar, Oxford University 13
Spectral– angulardistribution *oftheradiationgeneratedasaresultofthepointlikeelectronpassing throughcorrugatedchannelininfinite dielectric(𝑅 → ∞):
CST Particle In Cell simulation:
𝑐𝑜𝑠 𝛳 = L?g"$
+ %& h i� ;
ThediffractionordersofCherenkov andSmith-Purcellradiationpeakssatisfy the dispersion relation:
where𝛳 ispolarangle,𝛽 istheelectronspeed intermsofthespeedoflight,𝑘 isthewavenumberindielectric,𝑑 isthecorrugationperiod,𝑚 isadiffractionorder,𝜀 𝜔 isdielectricpermittivityasafunctionoffrequency.
*A.A.Ponomarenko et.al,Terahertzradiationfromelectronsmovingthroughawaveguidewithvariableradius,basedonSmith-PurcellandCherenkovmechanisms,NIMB309,223 (2013).
Bluecurve– corrugatedchannel.Reddashedcurve– channelwithconstantradius.
PIC simulations (ChSPR)
JAI Seminar, Oxford University 14
x
y
• 3Dpowerpatternandthecorrespondingazimuthalcross-sections(𝜃 = 90 deg.)fortheradiationat300GHzfortheoff-centralbeampropagationx=0,y=1mm.
• Cherenkovradiationisreflectedbytheoutside boundariesofthecorrugatedcapillaryanddirectedatsmallangles𝜃 ≈ 10 deg.
𝑃U:q" =r i∆_ ,
∆𝑡 =0.13ns(simulationtime)
Experiment (ChSPR)
JAI Seminar, Oxford University 15
15
x
z
y
zDet.
220
mm
θ≈ 40𝑑𝑒𝑔.
beam
x
y
xDet.beam
manX
θ≈ 40𝑑𝑒𝑔.
20 mm180 mm
20 mm
180 mm
Bunchsize:200x200𝜇𝑚
Bunchcharge:1bunch 25pC
Detector:SBD320– 460GHz
Quartzvacuumwindow:100mm(effectivediameter)
manY
Experiment (ChSPR)
JAI Seminar, Oxford University 16
Cross-checkwithTransition radiation:
Positioning ofthecapillarieswithrespecttothebeam(bremsstrahlung, appearing duetodirectinteractionoftheelectronbeamwiththetargetmaterial):
Experiment (ChSPR)
JAI Seminar, Oxford University 17
Thepolardistribution oftheradiation:
• The simulated geometry was identical to the realisticone with the exception of holders, which were nottaken into account in the simulations.
• The beam parameters were chosen to be the sameas during the time of the experiment, and the beamwas moving at 0.6 mm from the corrugation to allowfor 3𝜎 beam – corrugation separation.
• The power distribution was obtained by, first,calculating the power spectrum of emitted radiationin the frequency range 240 – 360 GHz. The radiationdirectivity pattern for each frequency was calculatedas well, hence it was possible to convert powerspectrum into the power distribution at the detectorlocations on the translation stage.
• After the converted power distribution wasobtained, the beam contribution in the powerspectrum was subtracted.
• The power spectrum of the radiation emittedthrough the surface of the outside boundary A ofthe calculation domain during the simulation time∆𝑡 was calculated as:
𝑃(𝜔) = ∫ ∯𝑺�� 𝜔 ∗ 𝒏𝑑𝐴𝑑𝑡∆_E ;
where 𝑺(𝜔) is the Poynting vector, n is a unity vector in theoutward normal direction from the boundary 𝐴of thecalculation domain.
Experiment (ChSPR)
JAI Seminar, Oxford University 18
• Azimuthaldistributionoftheradiationwassimulatedusingthefar-fieldmonitorofCSTParticleStudio,whichextrapolatestheelectricfieldvaluesattheborderofthecalculationdomaintoobtaintheelectricfields inthefar-field(distances≫ 𝛾L𝜆)atasinglefrequency.
• Accordingtothedispersion relation:𝑐𝑜𝑠 𝛳 = L?g
"$+ %
& h i� the
frequencyat𝛳 = 90deg.(detZ =0)is300GHz(𝜆 = 1mm).
• Theredcurve:measurementoftheazimuthaldistribution oftheradiationat300GHz.
Theazimuthaldistributionoftheradiation:
Overview and outlook
JAI Seminar, Oxford University 19
• ThemainobjectiveoftheSPRstudywastodemonstratethefeasibilitytogenerateSPRwithmonochromaticitybetterthan1-2%,bychoosingahigherdiffractionorderandrelativelysmallnumberofperiods(intheorderof10).
• Themeasurementswerelimitedbyseveralfactors,includingtheresolutionofthespectrometer,thequalityofitsalignment,angularacceptanceofthedetector.Allofthesefactorscontributetothewideningofthemeasuredspectrallines.
• Furthermonochromaticityimprovementsareexpected ifamulti-bunchbeamisused.
• TheobjectivesoftheChSPR studyweretocross-checktheradiationwithotherwellknownradiationmechanism(TR),tostudytheeffectofcorrugationandmeasuretheradiationdistributions.
• Measured10-foldincreaseoftheradiationintensityforthecorrugatecapillaryincomparisontotheblankcapillary.• Confirmed thatthemaximumoftheradiationintensity isachievedfortheoff-centralbeampropagation.• Thecompositedesignofthecorrugatedcapillaryallowsforflexibility toeasilychangethegeometry,whichcanbeveryuseful
foravarietyofstudiesincludingradiationgeneration,beamenergyandchargemodulationetc.
Marie Sklodowska-Curie Horizon 2020 project
JAI Seminar, Oxford University 20
Researchobjectives:
• toinvestigateCherenkovandSmith- PurcellmechanismsforTHzradiationgenerationusingEMsimulationtoolsandproposethemostefficienttargetconfigurationandobservationgeometry;
• tomanufacturearadiationsourcethatprovideshighpeakpower levelsandbasedonacompactlinearacceleratortechnologywithfs-durationbeam;
• toevaluatethepossibleapplicationsoftheinvestigatedradiationmechanismsforbeampositionandbunchlengthdiagnostics;• Toexpandresearchinterestsinthedirectionswhichallowforeffectiveskillsandknowledgetransfer(dielectricwake-field
acceleration,accelerator-scalewake-fieldsimulations,evaluationofpositiveandnegativeeffectsofwake-fieldsinaccelerator).• toprovidetheknowledgeexchangebetweenthepartnerorganizationsandotherinterestedpartiesviaseminars,satelliteand
progressmeetings,conferencesandworkshops;• toimprovethepublicawarenessabouttheongoingresearchviaoutreachactivities;• tofurtherdevelopprojectmanagementskills(financial,workflow,deliverablesplanningetc.)
Acknowledgements:
Thisprojecthasreceivedfunding fromtheEuropeanUnion’sHorizon2020researchandinnovation programme undertheMarieSklodowska-Curie grantagreementNo655179.
PIC simulations (capillary with reflector)
JAI Seminar, Oxford University 21
Parameter Value
beamLorentz- factor,𝛾 16
frequency up to700 GHz
bunch length,𝜎�R6� 0.03mm
bunch transversesize,𝜎_Tq64�
0.3 mm
micro-bunch charge 0.1nC
Nofmicro-bunches 4
distancebetweenmicro-bunches
variable(0.25– 1mm)
capillarymaterial Fusedquartz
holdermaterial Copper
numberofperiods 30
cylindrical ringwidth,l 0.5mm
corrugationperiod,2l 1mm
groovedepth, r2– r1 0.2 mm
internal radius, r1 2mm
outer radius, r3 2.7mm
Simulationgeometry(generalview):
Simulationgeometry(meshviewinOyzplane):
z
y
x
Probe
• Electricfieldprobeislocatedoutsideofthecalculationdomain;• Thefieldvaluesattheprobelocationareobtainedbasedonthe
extrapolationofthefieldvaluesattheborderofthecalculationdomain*.
beam
*Yee K S, Ingham D and Shlager K 1991 Time-Domain Extrapolation to the far field based on FDTD calculations IEEE Trans. of Ant. and Prop. 39 410
PIC simulations (capillary with reflector)
JAI Seminar, Oxford University 22
The power radiated through the surface of the outsideboundary 𝐴 of the calculation domain during thesimulation time ∆𝑡 at each frequency is given by thefollowing expression:
𝑃(𝜔) = ∫ ∯𝑺�� 𝜔 ∗ 𝒏𝑑𝐴𝑑𝑡∆_E ;
where 𝑺(𝜔) is the Poynting vector, n is a unity vector inthe outward normal direction from the boundary 𝐴 of thecalculation domain.
Single bunch:relatively flat response at all frequencies.
Blank capillary:power spectral modulation due to Cherenkov radiation.
Corrugated capillary:even larger spectral modulation if the distance betweenbunches is equal to the corrugation period.
Resonantcondition:
bunchdistance=corrugationperiod
PIC simulations (capillary with reflector)
JAI Seminar, Oxford University 23
Beamposition:x=0mm,y=1mm
x
y
Form-factorsoffourbuncheswithdifferentbunchspacing,andtheformfactorofasinglebunch:
Spectralmodulationoftheradiationdependsontheperiodicityofthecorrugationaswellasthedistancebetweenbunches.
Assembly of the capillary with holders
JAI Seminar, Oxford University 24
Capillarywithholders andradiationreflectorassembledonthebaseplate:
Laserscanswereperformedontheoutside surfaceofthecapillaryinstalledintheholders:
• Laserbeamscannedalongtheoutersurfaceofcorrugatedandblankcapillaries.
• Lightreflectedfromthesurfacedetectedbyanarraydetectorandtheverticaloffsetofthelaserbeamrecorder.
• Asaresultobtainedtheverticaloffsetasafunctionofthehorizontaltravelrange.
Both corrugated and blank capillaries are constructed as sets of cylindrical rings:
Experiment (capillary with reflector)
JAI Seminar, Oxford University 25
detY
detX
beamscan
holder
Target – beam separation and the corresponding radiation yield: