High Purity Pion Beam at TRIUMF

0.14 MB·

by  A. Aguilar-Arevalo

High Purity Pion Beam at TRIUMF
A.Aguilar-Arevaloa,M.Blecherb,D.A.Brymanc,J.Comfortd,J.Doornbosa,L.Doriaa,A.Husseine,N.Itof,S.
Kettellg,L.Kurchaninova,C.Malbrunotc,G.M.Marshalla,T.Numaoa,∗,R.Poutissoua,A.Shera,B.Walkerc,K.
Yamadaf
aTRIUMF,4004WesbrookMall,Vancouver,B.C.V6T2A3,Canada
bPhysicsDepartment,VirginiaTech.,Blacksburg,VA24061,USA
cDepartmentofPhysicsandAstronomy,UniversityofBritishColumbia,Vancouver,B.C.V6T1Z1,Canada
0
dArizonaStateUniversity,Tempe,AZ85287,USA
1 eUniversityofNorthernBritishColumbia,PrinceGeorge,B.C.V2N4Z9,Canada
0 fPhysicsDepartment,OsakaUniversity,Toyonaka,Osaka,560-0043,Japan
2 gBrookhavenNationalLaboratory,Upton,NY11973-5000,USA

n
a
J

8
1 Abstract

] AnextensionoftheTRIUMFM13low-energypionchanneldesignedtosuppresspositronsbasedonanenergy-loss
h techniqueisdescribed.Asourceofbeamchannelmomentumcalibrationfromthedecayπ+ →e+νisalsodescribed.
p
-
Keywords: Beamchannel,Particleseparation,PionDecay
c
c PACS:41.85.Ja,13.20.Cz,29.27.Eg
a
.
s
c
1. Motivation
i
s
y
h
p
[

1
v
1
2
1 The branching ratio of pion decays [1, 2],
3 R=Γ(π → eν + eνγ)/Γ(π → µν + µνγ), has pro-
. vided the best test of the hypothesis of electron-muon
1
Figure1:M13channelwiththeextension.
0 universality in weak interactions. The new TRIUMF
0 PIENUexperiment[3]aimingtoimprovetheprecision
1 of the branching ratio measurement by a factor of 2. M13ExtensionDesign
v: five or more measures positrons from the π+ → e+ν
Xi decay (Ee+ = 69.8 MeV) and the π+ → µ+ → e+ 2.1. ExistingM13channel
decaychain(π+ → µ+νdecayfollowedbyµ+ → e+νν
r The layoutof the existingM13 channel[4] together
a decay,Ee+ < 52.8MeV).In orderto obtainmaximum with the new beam line extension is shown in Fig.1.
acceptance with minimum uncertainties arising from
Thebeamline takesofffromtheprimaryprotonbeam
positron energy-dependentcross sections, the detector line (BL1A) at an angleof 135◦ froma 1-cmthick Be
system involving a large NaI crystal is placed on the
productiontarget(T1). TheM13channelwithamaxi-
beam axis. Positrons in the beam (1/4 of the rate of
mumangularacceptanceof29msrisalow-momentum
pions) severely increase detector rates, trigger rates achromatic channel with −60◦ (B1 magnet) and +60◦
and background in the π+ → e+ν spectrum. The
TRIUMFM13channel[4]hasthereforebeenupgraded
tosuppressthepositroncontaminationinthepionbeam. ∗Correspondingauthor:E-mail:[email protected],Tel:[phone]-
7345
PreprintsubmittedtoNuclearInstrumentandMethod January19,2010
(B2 magnet) bends, a quadrupole doublet (Q1–Q2) initial pion momentum of 76.8 MeV/c, which was
between the production target and B1 for collecting degraded to 73.6 MeV/c by a 2.0 mm thick Lucite
pions, a quadrupole triplet (Q3–Q5) between the two absorber. Separation of 53 mm between the pion and
bends,andaquadrupoledoublet(Q6–Q7)downstream positron images was expected at F3. The positron
of B2 for the final focusing. There are three foci: image size was expected to be 16 mm (FWHM) with
F1 between B1 and Q3, F2 between Q5 and B2, and a small tail at the low momentumside, while the pion
F3 after Q7. Beam acceptance-defining slits SL0 are sizewas18mm(FWHM)witha10%tail. Twothirds
located just upstream of the first bending magnet B1, of the pions are lost due to decay-in-flight, scattering,
andtherearemomentum-definingslitsSL1andSL2at and collimation, resulting in a pion to positron ratio
F1 and F2, respectively. About 10 cm downstream of of 100 expected at F4. Third order aberrations might
SL1, there are two absorber/slitwheels each with four cause some broadeningof the beam spot but were not
mountingpositionsthatallow4×4combinationsofslit expectedtoimpactthesuppressionfactor.
and/orabsorbersettings.Themaximum75-MeV/cpion
yield at F3 is 0.8 M/s for 500 MeV, 100 µA TRIUMF
cyclotronoperation. 2.3. TheExtension
ThenewextensionstartsatF3(0.9mdownstreamof
Q7),andconsistsofa–70◦ bendingmagnet(B3)at1.5
2.2. DesignPrinciple m downstream of Q7, and a 30-cm diameter aperture
Energy-loss-basedparticleseparatorshavebeenused quadrupole triplet (Q8-Q10) after B3. A 5 cm thick
sincetheearlydaysofparticlephysicsexperiments[5]. lead collimator with a 3 cm square hole placed at F3
Thedifferentialenergyloss at75MeV/c forpionsand stops the displaced positrons and redefines the pion
positronsis large enough for clean particle separation, image. ThenewfocusF4is1.5mdownstreamofQ10.
and this concept can be applied simply to the existing ThetotallengthoftheextensionbetweenF3andF4is
M13 structure. When a thin foil is inserted at the ab- 4.5m. Tocauseamomentumspreadbetweenpositrons
sorberwheelnearF1,amomentumspreadbetweenpi- andpions,1.45mmand2.0mmthickLucitedegraders
ons and positronsresults due to the energy loss differ- are mounted on the absorber wheel near F1. Here the
ence.FollowingsubsequentmomentumanalysisbyB2, momentumwidthoftheincomingbeamisrestrictedto
pionsand positronsseparate into two distinct horizon- 1.5%(FWHM)byclosingtheSL1horizontalslitto1.5
taldistributionsatF3. Acollimatorcanbeplacedsuch cm. ThebeamafterF1istunedforthepionmomentum
thatitstopsonlypositrons. A smallmomentumtailof correspondingtothepionenergyaftertheenergyloss.
thepositronbeamduetobremsstrahlungresultsinsome
positronsat the pion spot, giving < 1 % positron con-
tamination. Dueto the energy-lossvariationin the ab-
3. Measurements
sorber,thepionbeamhasasignificantlow-momentum
tail, which results in degradation of the image at F3. 3.1. Detectorandmeasurements
Showers from stopped positrons may be a source of Tests were carried out in two stages; the first stage
backgroundwitha23MHzradiofrequency(RF)struc- wasdoneatF3beforetheinstallationoftheextension,
tureoftheprotonbeamfromtheTRIUMFcyclotronif and the second at F4 after installation. The incoming
thedetectorsystemwerelocatednearF3. beam was measured at F3 or F4 with a telescope
The extension of the M13 channel provides con- consisting of two plastic scintillator beam counters
trolledpionbeamquality,althoughthereisapioninten- (0.6 cm × 15.2 cm × 20.3 cm and 0.3 cm × 3 cm
sitylossduetodecay-in-flight. Theblurredpionbeam × 4 cm), 6 layers of 10-cm diameter wire chambers
imageduetoacombinationofmomentumspreadbythe arranged in the wire orientation of X-U-V-X-U-V (0◦
absorberanddispersionbytheB2magnetcanberede- and±60◦withrespecttotheverticalaxis),anda48cm
finedwitha collimatoratF3 foranimprovedimageat diameter48cmlongNaI(Tℓ)crystal[7]surroundedby
anewdownstreamfocusF4. Theextensionisolatesthe two cylindrical layers of 8.5 cm thick, 2×25 cm long
showersourcefromthedetector,allowingbettershield- pure CsI crystals [8]. The time of arrival of particles
ingforγ-raysfromthestoppedpositrons. at the beam counter with respect to the cyclotron
Simulation was done using a Monte Carlo beam RF time provided particle identification based on the
transportprogram, REVMOC [6], which calculates up time-of-flight(TOF)togetherwith theenergylossesin
tosecond-orderopticsincludingtheeffectsofmultiple thetelescopecounters.
Coulomb scattering. The calculation was done at the
2
3.2. MeasurementswithouttheExtension 3.3. MeasurementswiththeExtension
The pionand positronrates at 75 MeV/c were mea-
After initial tuning of the entire beam channel at
sured with the width of the momentum-defining hori- 77 MeV/c with a 5 cm thick lead collimator with a
zontalslitSL1setat1.5cmtobe0.2M/sand0.05M/s,
horizontalopeningof 3 cm placed at F3, the 1.45 mm
respectively,fora100µAprotonbeamonthe1cmthick
thick absorber was inserted and the downstream beam
Betarget. TheverticalslitsatSL1andSL2weresetat
momentum (only B2 and B3) was scaled to measure
2.5cmand4cm,respectively,restrictingthebeamrates
thepionandpositronyieldsatF4.
to1/4ofthemaximumrate. Thehorizontalandvertical
beam spot sizes at F3 were measured to be 4.5 mm ×
2.2 mm (rms), respectively, for positrons (10 % wider
forpions). 40
Inordertosuppresspositronsbasedonthedisplace-
mentatF3,thetailinthebeamprofileneedstobemin-
30 Pion
imized. Effects of slits on the beam profile were stud- z)
H
ied. Acceptance-limitingjawsSL0upstreamoftheB1 k
magnetcausedabroadtailupto1/3ofthetotalpositron ate ( 20
intensitywhentheywereclosedtonarrow(1cm),while R
notailwasobservedwhenSL0waswideopen(12cm). 10 Positron
ThemomentumdefiningslitsSL1atF1didnothavean
impactonthetail. Theseobservationsindicatedthatthe
0
jawsSL0 producedan additionalsourceimagethatal- 73 74 75 76
Momentum (MeV/c)
loweddifferentmomentumcomponentstopassthrough
SL1.
Whena 1.45mmthickLuciteabsorberwasinserted Figure3:Pion(closedcircles)andpositron(opencircles)yieldsatF4
forvariousdownstreammomenta.
into the beam near F1, the positron beam position at
F3 was displaced by 46 mm with respect to the pion
Fig.3 shows a plot of pion and positron rates at F4
beam as shown in Fig.2. In order to display pions
vs. the downstream channel momentum. The rates
and positrons (solid histograms) in the same plot, the
are normalized to a 100 µA proton current. After
upstream momentum was raised only by 1 %, and
further tuning of beam focusing at F1 and F3, the e+
the measured position and intensity might slightly be
beam contamination with respect to the pion rate was
biaseddue to the geometryof the wire chambers. The
reducedto1/60. Thehorizontalandverticalbeamspot
horizontal spot sizes of positrons and pions increased
sizes were 12 mm and 9 mm (rms), respectively. The
to7.0and8.9mm(rms),respectively,withatailinthe
observed pion rate of 40 k/s, being consistent with
low momentum side. For a thicker absorber (2.0 mm
the prediction, can be increased without significantly
thickLucite),theseparationincreasedto50mm.
affectingthe verticalimageat F4 to morethan 100k/s
byopeningthe verticalSL1 andSL2 slits, whichwere
setat2.5cmand4cm,respectively.
1200 π+
1000
S 3.4. BeamChannelCalibrationSource
T800
N Due to uncertaintiesin thefringefieldsof the bend-
OU600 µ+ e+ ing magnets, it is usually difficult to obtain an abso-
C
400 lute beam momentum calibration with accuracy. Two
sources commonly used for calibration are the high-
200
momentum edges of positrons from µ+ → e+νν and
0
−60 −40 −20 0 20 40 60 surface muonsfrom π+ → µ+ν. The decay π+ → e+ν
X position (mm)
fromstoppedπ+intheproductiontargetcouldprovidea
highenergysourcewithadefinitivepeakat69.8MeV/c.
Figure2: Pion,muonandpositronpositiondistributions atF3(his-
Unlikethesurfacemuon,themeasurementofthedecay
tograms). Theheavylines arefittedGaussiancurves forpions and
positrons. π+ → e+ν from the productiontarget does not require
specialequipment.
3
Themajorsourceofbeampositronsabove55MeV/c
produced at 500 MeV is from the decay π0 → γγ 1
promptly followed by γ-ray conversion to electron-
positronpairs in the targetmaterial. The positron mo- µ+ → e+ ν ν
mentum distribution is nearly flat as shown in Ref.[4]. 10-1
These positrons are prompt with respect to the proton
beamburst. atio π+ → e+ ν
R10-2
We searched for delayed positrons coming from the
beamlinebyvaryingthebeamchannelmomentumbe-
tween 50 MeV/c (below the µ+ → e+νν edge) and 80
10-3
MeV/c. Tightcutsonenergylossinthebeamcounters
suppressedthepionandmuoncontaminationstoaneg-
ligiblelevel. Byselectingeventswiththebeamenergy 50 60 70 80
intheNaIdetector,pionsandmuonswerefurthersup- Momentum (MeV/c)

pressedaswellasthebackgroundfromπ+ →µ+ →e+ 105
decayscomingfromstoppedpionsnearthedetector.At
this point, events originated from π0 were dominant.
104
The π+ → e+ν component was enhanced by select-
ingdelayedeventsusingthe TOF. Fig.4a showsyields S
of delayed positrons normalized to the total positrons UNT103 τ = 24.6 ± 2.8 ns
withthebeamchannelmomentum.Theedgearound52 O
C
MeV/cisfromµ+ → e+ννdecaysfromtheproduction 102
target. The rate of delayed π+ → e+ν events was 0.6
% of the total positronsat the same beam momentum. 10
The peak momentum shift of 1.3 MeV was consistent
withthe energylossof positronsinthe productiontar- 0 10 20 30 40
Time (ns)
get (0.5–1MeV) and the uncertaintyin the calibration
of1%. Fig.4bshowsa time spectrumfor68.5MeV/c
Figure4: (Top:a)Fraction ofdelayed positrons withthebeammo-
positrons with respect to the proton burst (the RF sig-
mentumand(Bottom:b)timespectrumof68.5MeV/cpositronswith
nal). The delayed component has a decay constant of respecttotheprotonburst. Thefitofthedelayedcomponenttoan
24.6±2.8ns, whichis consistentwiththe pionlifetime exponentialcurveisshownintheboldline.
[9]. Itisworthwhiletomentionthatthesepositronsare
expectedtobe∼100%circular-polarized.
the requirements of the PIENU experiment, including
By flipping the beam channel polarity to negative,
a pion rate of 100 k/s and positron suppression. A
we also searched for delayed 50 MeV/c electrons
from µ− → e−νν decays in the production target as new calibration source from π+ → e+ν decays in the
targetwasidentifiedforlow-momentumbeamchannels.
a potential high-intensity source of stopped negative
muons. Theratioofdelayedandpromptelectronswas
measuredto be (3.5±0.4)×10−3, which is consistent
with an estimate of 6 × 10−3 based on the product of
the ratio of delayed and prompt positrons (2.9), the 5. Acknowledgment
π−/π+ productionratioin this energyregion(1/5),and
thefractionofdecay-in-flightofpionsinwhichmuons
stopinthetarget(1%)[10]. The authors wish to thank C. Ballard, N. Khan, R.
Kokke,D.Evans,K.Reiniger,andthebeamlinegroup
for the design and installation work. This work was
supported by the Natural Science and Engineering
4. Conclusion Research Council and the National Research Coun-
cil of Canada. One of the authors (MB) has been
TheTRIUMFM13channelwasmodifiedtoachieve supported by US National Science Foundation grant
apion/positronratioof>50usingadifferentialenergy- Phy-0553611.
lossmethod.Theperformanceoftheextensionsatisfies
4
References
[1] D.I.Brittonetal.,Phys.Rev.Lett.68(1992)3000andPhys.Rev.
D49(1994)28.
[2] G.Czapeketal.,Phys.Rev.Lett.70(1993)17.
[3] TRIUMFproposalS1072,2005.
[4] C.J.Orametal.,Nucl.Instr.Meth.179(1981)95.
[5] D.I.Meyer,M.L.PerlandD.A.Glaster,Phys.Rev.107(1957)
279.
[6] C.KostandP.Reeve,TRIUMFreport,TR-DN-82-28,1982.
[7] G.Blanpiedetal.,Phys.Rev.Lett.76(1996)1023.
[8] I-H.Chiangetal.,IEEENS-42(1995)394.
[9] T.Numaoetal.,Phys.Rev.D52(1995)4855andV.P.Koptev
etal.,JETPLett.61(1995)877.
[10] T.Numaoetal.,Phys.Rev.D73(2006)092004.
5