Search for beta plus/EC double beta decay of 120Te

0.21 MB·English

by  E. Andreotti

Search for β+/EC double beta decay of 120Te
E.Andreottia,C.Arnaboldib,c,F.T.AvignoneIIId,M.Balatae,I.Bandacd,M.Baruccif,J.W.Beemang,F.Bellinih,C.Brofferiob,c,
A.Bryantg,i,C.Buccie,L.Canonicaj,k,S.Capellib,c,L.Carbonec,M.Carrettonib,c,M.Clemenzab,c,O.Cremonesic,
R.J.Creswickd,S.DiDomizioj,k,M.J.Dolinskil,i,L.Ejzakm,R.Faccinih,H.A.Farachd,E.Ferrib,c,E.Fiorinib,c,L.Foggettaa,
A.Giacheroc,L.Gironib,c,A.Giuliania,P.Gorlae,E.Guardincerrie,k,T.D.Gutierrezn,E.E.Hallerg,o,K.Kazkazl,S.Kraftb,c,
L.Koglerg,i,C.Maianob,c,R.H.Maruyamam,C.Martinezd,M.Martinezc,S.Newmand,S.Nisie,C.Nonesa,E.B.Normanl,p,
A.Nucciottib,c,F.Orioh,M.Pallavicinij,k,V.Palmieriq,L.Pattavinab,c,M.Pavanb,c,M.Pedrettil,G.Pessinac,S.Pirroc,
E.Previtalic,L.Risegarif,C.Rosenfeldd,C.Rusconia,C.Salvionia,S.Sangiorgiom,D.Schaefferb,c,N.D.Scielzol,M.Sistib,c,
A.R.Smithg,C.Tomeih,G.Venturaf,M.Vignatih
aDip.diFisicaeMatematicadell’Univ.dell’InsubriaandSez.INFNdiMilano,ComoI-22100-Italy
bDip.diFisicadell’Universita` diMilano-BicoccaI-20126-Italy
1 cSez.INFNdiMi-Bicocca,MilanoI-20126-Italy
1 dDept.ofPhys.andAstron.,Univ.ofSouthCarolina,Columbia,SC29208-USA
0 eLaboratoriNazionalidelGranSasso,I-67010,Assergi(L’Aquila)-Italy
2 fDip.diFisicadell’Universita` diFirenzeandSez.INFNdiFirenze,FirenzeI-50125-Italy
gLawrenceBerkeleyNationalLaboratory,Berkeley,CA94720-USA
n hDip.diFisicadell’Universita` diRomaLaSapienzaandSez.INFNdiRoma,RomaI-00185-Italy
a iDept.ofPhysics,Univ.ofCalifornia,Berkeley,CA94720-USA
J jDip.diFisicadell’Universita` diGenova-Italy
1 kSez.INFNdiGenova,GenovaI-16146-Italy
1 lLawrenceLivermoreNationalLaboratory,Livermore,CA94550-USA
mUniv.ofWisconsin,Madison,WI-USA
] nCaliforniaPolytechnicStateUniv.,SanLuisObispo,CA93407-USA
x oDept.ofMaterialsSc.andEngin.,Univ.ofCalifornia,Berkeley,CA94720-USA
e pDept.ofNuclearEngineering,Univ.ofCalifornia,Berkeley,CA94720-USA
- qLaboratoriNazionalidiLegnaro,I-35020Legnaro(Padova)-Italy
l
c
u
n
[

2 Abstract
v
Wepresentasearchforβ+/ECdoublebetadecayof120TeperformedwiththeCUORICINOexperiment,anarrayofTeO cryogenic
1 2
1bolometers.Aftercollecting0.0573kg·yof120Te,weseenoevidenceofasignalandthereforesetthefollowinglimitsonthehalf-
8life:T0ν >1.9·1021yat90%C.L.forthe0νmodeandT2ν >7.6·1019yat90%C.L.forthe2νmode.Theseresultsimprovethe
1/2 1/2
4existinglimitsbyalmostthreeordersofmagnitude(fourinthecaseof0νmode).
.
1
1
0 The discovery of neutrino oscillations [1] proved that neu- Sasso NationalLaboratories. The 120Te isotope hasbeen only
1trinos are massive, but there are fundamental issues that os- minimallyinvestigatedfromatheoreticalpointofview.Nocal-
v:cillation experiments cannot address: measuring the absolute culationsofthehalf-lifeof0νβ+/ECdecayof120Teareavailable
ineutrinomassanddeterminingwhethertheneutrinoisthean- for comparison with our result. The predictions for the same
X
tiparticle of itself, thus being of Majorana nature, or not. To decaymechanisminothernuclei,assumingtheeffectiveMajo-
r
aanswer these questionsit is necessary to look for neutrinoless ranamasshmνi=1eV,rangebetween1026–1027 y[3,4]. The
doublebeta, 0νββ, decays, which would bringconclusiveevi- largervalueismostlyduetotheneedtohavesimultaneouslya
denceoftheMajorananatureoftheneutrinoandwhosedecay decay anda captureand to the reducedphase space available.
ratesconstraintheabsoluteneutrinomass[2]. Theonlyreferenceforthetwoneutrinomode[5]yieldsatheo-
Double beta decays can occur by either emitting two elec- reticalvalueforthehalf-lifeof4.4·1026y.
trons or two positrons. In the latter case, either of the
Inrecentyears,experimentallimitsonthe120Tedecayshave
positronemissionscanbereplacedbyanelectroncapture(EC).
beensetusinganarrayofCdZnTedetectorslocatedattheGran
Whileβ−β− decayshavethelargestexpectedrates,β+/EC and
Sasso UndergroundLaboratory [6, 7] and a HPGe detector at
β+β+decaysprovideclearsignaturesfromthe511-keVannihi-
the ModaneUndergroundLaboratory[8, 9]. The presentbest
lationgammarays. Energyandmomentumconservationinthe limitsintheliteratureareT0ν > 4.1·1017 y[7]andT(0ν+2ν) >
EC/EC decayrequiresanextraradiativeprocess, reducingthe 1/2 1/2
1.9·1017y[8].
ratebyseveralordersofmagnitude.
Thispaperreportsonasearchforβ+/ECdecays120Te →120 The search strategy is the following: the emitted positron
Sn+e+ and 120Te →120 Sn+e+ +2ν with the CUORICINO carriesakineticenergyuptoK =Q-2m c2 -E ,whereE
max e b b
experiment,anarrayofTeO cryogenicbolometersattheGran isthebindingenergyofthecapturedelectronwithintheatomic
2
PreprintsubmittedtoElsevier January12,2011
capacity 2.3·10−9J/K.Tomeasuretemperaturevariationscor-
respondingtofewkeV(∆T ∼0.1µK/keV)heavilydopedhigh-
resistance germanium thermistors (NTD, Neutron Transmuta-
tionDoped)aregluedtoeachcrystal.
The CUORICINO detector consists of 62 TeO crystals ar-
2
rangedin13planes.Eachoftheupper10planesandthelowest
oneconsistsoffour5×5×5cm3TeO crystals,whilethe11th
2
and 12th planes have nine, 3×3× 6 cm3 crystals. All crys-
talshavenaturalisotopicabundancesexceptfourofthesmaller
crystals, two ofwhichareenrichedto82.3%in 128Te andtwo
to75%in130Te.
Thenaturalabundanceof120Teis0.096%[13],sothe39.4kg
of the CUORICINO experiment(enriched crystals are not in-
cluded)containN =1.43·1023nucleiof120Te.
ββ
Theexperimentisshieldedwithtwolayersofleadof10cm
minimumthicknesseach. Theouterlayerismadeofcommon
lowradioactivitylead,whiletheinnerlayerismadeofspecial
Figure1: AsketchoftheCUORICINOassemblyshowingthetowerhanging leadwithalowactivityof210Pb. Theelectrolyticcopperofthe
fromthemixingchamber,thevariousheatshieldsandtheexternalshielding.
refrigeratorthermalshieldsprovidesanadditionalshieldwitha
minimumthicknessof2cm.Anexternal10cmlayerofborated
shellandQ= (1714.8±1.3)keV[10]isthedifferencein120Te polyethylenewasinstalledtoreducethebackgroundduetoen-
and 120Sn atomic masses. The electron capture is most likely vironmentalneutrons.Thedetectoritselfisshieldedagainstthe
to occur from the K shell, whose binding energy is 30.5 keV. intrinsicradioactivecontaminationofthedilutionunitmaterials
TheratioofL-capturetoK-captureformostelementsisaround by an internallayer of 10 cm of Roman lead [14], located in-
10%(12%for120Sb→120SnECdecay)[11]. Inthefollowing sidethecryostatimmediatelyabovethetower.Thebackground
wewillalwaysassumeacapturefromtheKshell. fromtheactivityinthelateralthermalshieldsofthedilutionre-
Thebolometerwherethedecayoccurswillseethedeposition frigeratorisreducedbyalateralinternal1.4cmthickshieldof
ofboththebindingenergyandthekineticenergyofthepositron Romanlead. Another8cmleadshieldislocatedatthebottom
(maximumenergy:E = K +E = Q−2m c2 =692.8keV, ofthetower. TherefrigeratorissurroundedbyaPlexiglasanti-
0 max b e
independentofEb).1 Onceatrest,thepositronannihilateswith radonboxflushedwithclean N2 fromaliquidnitrogenevapo-
anelectronandtwophotonswithanenergyE =511.0keVare ratorand is also enclosedin a Faradaycage to eliminate elec-
γ
emitted.Thesephotonscaninteractwiththesamebolometeror tromagneticinterference.Asketchoftheassemblyisshownin
byanearbyone,ortheycanescapeundetected. Fig.1.
Theanalysispresentedheresearchesforthesignatureswith
thebestsignal-to-noiseratio. Thisisnotthecasefortheevents
Signature[energiesinkeV] µ ε[%]
wheretheentireenergyisdepositedinsidethedetector,dueto
(30.5–692.8) 1 3.00±0.02
the low detection efficiency (see Table 1). For the 0ν mode,
(30.5–692.8)+511 2 3.40±0.02
where the positron is monochromatic, this means the coinci-
(30.5–692.8)+511+511 3 0.45±0.01
dence between a bolometer with an energydeposition consis-
(541.5–1203.8) 1 16.28±0.04
tent with E and one with either E or E + E and the co-
γ 0 0 γ
(541.5–1203.8)+511 2 6.23±0.03
incidence of one bolometerwith a signal of E and two other
0
(1052.5–1714.8) 1 10.04±0.03
bolometers with a signal of E . For the 2ν mode, where the
γ
positron is emitted with a continuum of kinetic energies be-
tween 0 and K , this means the triple coincidence of one Table1: Signatures of120Teβ+/ECdecayinanarrayofTeO2 detectors and
max their corresponding multiplicity (µ), that is the number of detectors with an
bolometerwithasignalbetweenEbandE0andtwootherswith energydepositionabovethreshold. Thedetection efficiencyforthe0νmode
asignalofEγ. inCUORICINOisreportedinthelastcolumn(ε). Wedenotewiththe+sign
thecoincidenceofenergiesreleasedindifferentdetectors.Forthe0νmodethe
energy released inthedetector wherethedecay occurred corresponds tothe
1. TheCUORICINOdetector upperboundoftheinterval.Theerrorsarestatisticalonly.
The CUORICINO experiment is detailed in Ref. [12].
Briefly, it is an array of TeO crystals acting as cryogenic
2
Forthepresentanalysis,thefullCUORICINOstatistics(data
bolometersataworkingtemperatureof8–10mKandwithheat
collected between May 2004 and May 2008)for a totalexpo-
sureof0.0573kg·yof120Teisused. Thetotalenergyspectrum
1HereandinthefollowingweassumethattheX-raysfollowingtheECdo ofalldetectorsisshowninFig.2. Severalpeaksofradioactive
notescapefromthecrystalwherethedecayoccurs. isotopesareclearlyvisible,themostprominentarelabeled.
2
Figure2: TotalenergyspectrumofallCUORICINOdetectors. Themostprominentpeaksarelabeledandcomefromknownradioactivesourcessuchas: e+e−
annihilation(1),214Bi(2),40K(3),208Tl(4),60Co(5)and228Ac(6).
Figure3:ResolutionfitsatE=511keV(left)andE=1238keV(right).ThefitfunctionisaGaussianpluslinearbackground.
2. Dataacquisitionandanalysis voltage pulse to the Si resistors. Their Joule dissipation pro-
duces heat pulses in the crystal with a shape which is almost
identicaltothecalibrationγ-rays.
CUORICINO dataaredividedindatasets,eachonebeinga
collection of about a month of daily measurements. Routine
calibrations are performed at the beginning and at the end of
3. Searchstrategy
each datasetusingtwo wiresofthoriatedtungsteninsertedin-
side the external lead shield. The signals coming from each As detailed in the introduction, a β+/EC decay of 120Te in
bolometerareamplifiedandfilteredwithasix-poleBessellow- an array of TeO detectors releases an energy up to E =
2 0
passfilterandfedtoa16-bitADC.Thesignalisdigitizedwith 692.8keVinthebolometerwherethedecayoccursand,ifthe
asamplingtimeof8ms. Witheachtriggeredpulse,asetof512 annihilationgammasdonotescapeundetected,oneortwoad-
samplesisrecordedtodisk. Thetypicalbandwidthisapproxi- ditionalenergydepositsof E = 511.0keVineitherthesame
γ
mately10Hz,withsignalriseanddecaytimesoforder30and or a nearby bolometer. There are therefore several distinctive
500ms,respectively.Moredetailsofthedesignandfeaturesof signaturesaslistedinTab.1.
theelectronicssystemarefoundinRef.[15]. For the 0ν mode the energy released in the detector where
Each bolometer has a different trigger threshold, optimized thedecayoccurredcorrespondstotheupperboundoftheinter-
accordingtothebolometer’stypicalnoiseandpulseshape.The val. Thedetectionefficienciesforthe0νmodewereestimated
trigger rate is time and channeldependent, with a mean value bymeansofaGEANT4simulationoftheCUORICINOsetup
of about 1 mHz. The amplitude of the pulses is estimated by (see last column of Tab. 1) where the decays are located uni-
means of an Optimal Filter technique [16]. The gain of each formlywithinallnon-enricheddetectorsandthedecayproducts
bolometerismonitoredbymeansofaSiresistorof50–100kΩ are emitted isotropically. We always assume that the binding
attachedtoitthatactsasaheater. Heatpulsesareperiodically energyof the capturedelectronis released within the detector
supplied by an ultra-stable pulser [17] that sends a calibrated where the decay occurs. The efficiency estimate includes the
3
Figure4:Energyspectraofcoincidenceswithasingle511keVevent(left)andofcoincidenceswithtwo511keVphotons(right). Themostprominentpeaksare
labeledandcomefromthesingleescapeandthedoubleescapeofknownlinesfromthefollowingisotopes: 208Tl(1),214Bi(2)and40K(3). Inthespectrumof
doublecoincidences(left)the511keVpeakisalsovisible(4).
dead time evaluated separately for each detector. To account 4. Resultsofthe0νβ+/ECdecaysearch
for the different energy resolution of CUORICINO detectors
weassignedtoeachdetectorinthesimulationaFWHMgiven In the spectra of Fig. 4 we search for a signal with mean
bytheweightedmeanoftheFWHMscalculatedinallCUORI- energy 692.8 keV (spectra of double and triple coincidences)
CINO calibrationruns. Thehighestdetectionefficienciescor- or 1203.8 keV (spectrum of double coincidences only). The
respondtothecaseswhereoneorbothoftheelectron-positron expectedresolutionat1203.8keVisestimatedonthe214Biline
annihilation photons are fully absorbed in the detector where at1238keVandfoundtobeσ=1.67±0.05keV(seeFig.3),in
the β+/EC decay occurs. Thisis consistent with the relatively agreementwiththevalueof1.7keVfoundforthe511keVline
largesizeoftheCUORICINOdetectorsandthemeanfreepath and consistent with being constant over the energy region of
ofa511keVphotoninTeO (1.9cm).Thesesignaturesarecal- interest.
2
culatedtoaccountfornearlyhalfofalldecays. Theremainder Fig. 5 shows the energy regions of interest of the double
involve only partial energydeposition of the 511 keV gamma and triple coincidences spectra. The observed lines, as ex-
rayenergyandthereforearemoredifficulttodistinguishfrom plainedabove,correspondtosingle-anddouble-escapepeaks.
background. Due to the extremely short range of positrons Specifically,themeasuredspectrashowthesingleescapelines
in CUORICINO crystals, the efficiencies estimated for the 0ν of the 1173.2 keV transition from 60Co at 662.2 keV and of
mode are also validfor the 2νmode, apartfromsmall correc- the 1729.6 keV transition from 214Bi at 1218.6 keV plus the
tions (see below). Consideringalso the expectedbackground, double escape lines of the 2204.06 keV transition from 214Bi
we limited ouranalysisto the signaturesthatfeature a coinci- at 1182.06 keV, of the 1764.5 keV transition from 214Bi at
denceoftwoorthreeeventsforthe0νmodeandofthreeevents 742.5 keV and again of the 1729.6 keV transition from 214Bi
forthe2νmode. at 707.6 keV. The observed SE and DE peaks produce a neg-
ligiblecontributiontothebackgroundintheenergyregionsof
We consequently search only for double or triple coinci- interest.
denceswhereoneortwooftheenergydepositionsareinthein- Thesearchforthe0νβ+/ECdecayof120Teonthespectrain
terval±2.5σofEγ =511keV,whereσ=1.7keV,asestimated Fig.5isperformedwithunbinnedmaximumlikelihoodfits. In
intheinclusiveenergyspectrum(seeFig.3). Thecoincidence thelikelihoodfitthebackgroundisthereforeparametrizedwith
window is 100 ms. The probabilityof accidentalcoincidence thesumofaflatcomponentandaGaussianforeachoftheex-
isestimatedfromthemeasurementofthesinglecrystalrate(∼ pectedescapelines,withmeanvaluesfixedtothecorrespond-
1.6mHz)to bearound0.7%fordoublecoincidencesbetween ingknownenergiesandthewidthfixedtoσ=1.7keV.Thesig-
detectors in the same plane or in adjacent planes. For triple nalisalsoparametrizedwithaGaussianwithmeanvaluefixed
coincidencesitisevenless. totheexpectations,thatis692.8keVor1203.8keV,depending
onwhetherornotoneoftheannihilationphotonsisabsorbedin
Theenergyspectraoftheeventsincoincidencewithoneor thesamecrystalwhereitisemitted.Thewidthofthesignalline
two 511 keV photons are shown in Fig. 4. The structures in isfixedto2.2keVwhichisobtainedbysumminginquadrature
both spectra have been identified and correspond to single or the sigma (1.7 keV) , the error on the Q-value (1.3 keV), and
double escape lines from known radioactive lines present in theuncertaintyontheenergyscale(0.4keV)asestimatedfrom
theCUORICINObackgroundspectrum. Thecontinuumback- the calibration fit2. In summary for each spectrum the fitted
groundisduebothto accidentalcoincidencesandtruecoinci-
dencesinwhichtheenergydepositionoftheeventthataccom-
paniesthe511keVgamma(s)isnotcomplete. 2Fromabayesian pointofview (i.e. ourapproach), youshouldsumthe
4
we calculate with the same procedurethe pdfs for the param-
eter N /ǫ , the number of signal events divided by the cor-
sig tot
responding efficiency. The efficiency is calculated as ǫ =
tot
ǫ ×(ǫ ǫ )µ, where ǫ is the efficiency tabulated in Tab. 1,
noise heat
µ=2,3isthemultiplicityofthegivensignaturefromthesame
table,ǫ =99.1%accountsforthelossofsignalduetonoise
noise
as estimated on heater pulses, and ǫ = 97.7%accountsfor
heat
the dead time induced by the presence of heater. This proce-
dure, togetherwith the inclusionof the uncertaintyon the en-
ergyscaleinthefitsofthespectra,accountsforthesystematic
errors,whichhaveanegligibleimpactontheresult.
From the combinedpdf, we extracta 90% C.L. upperlimit
on the total number of double beta decays regardless of their
detection,n = 100. Wecanthensetalimitonthehalf-lifeof
B
0νβ+/ECdecayof120Te
T
T0ν >ln2N =1.9·1021y (1)
1/2 ββn
B
whereN isthenumberof120TenucleiandT isthelivetime.
ββ
5. Resultsofthe2νβ+/ECdecaysearch
The energyspectrum of the eventsin coincidencewith two
511 keV photons in the region below 692.8 keV contains 15
events (see Fig. 6). We can exclude from our analysis the re-
gions±3σaroundtheenergieswhereweexpecttheDEpeaks
from known radioactive gamma lines in the CUORICINO in-
clusivespectrum. Theseenergiesareindicatedwithredarrows
in Fig. 6 and listed in the caption. We have listed only the
gamma lines that are expected to contribute with at least one
event. This expectation is based on the comparison with the
40Klinethatproduces5eventsat438.8keVintheexperimen-
tal spectrum of triple coincidences(see Fig. 6, line labeled as
1). Afterthesubtraction,only8eventsremaininthespectrum
fromthresholdto end-point. These eventscan be due to acci-
dental coincidences or to true coincidences (double escape of
Figure5:Energyspectra,intheregionsofinterest,ofcoincidenceswithasingle
511 keV event [two instances] (top) and ofcoincidences with two 511 keV alreadycomptonscatteredgammas).Toestimatethefirstcom-
photons(bottom). Thearrowsindicatetheenergyoftheexpectedsignal. Fit ponentwelookedatthespectraofeventsintriplecoincidence
results,asexplainedinthetext,areoverlaid.
with the side-band of 511 keV (left side-band: from 470 keV
to502.5keV,rightside-band:from519.5keVto560keV)cor-
quantitiesarethenumberofeventsinthesignal,theflatback- rectlynormalizedtotheexperimentalspectrumofFig.6. This
groundandtheescapelines. Thecurvesresultingfromthefits accountsfor4.3±0.5events.Theothercomponentoftheback-
areoverlaidinFig.5andthecorrespondingfittedquantitiesare groundcannotbereliablyestimatedduetolimitedstatisticsand
reportedinTab.2. we cannotdiscriminate against it. Therefore, we set an upper
The upper limit on the numberof observedsignal eventsis limitassumingconservativelythattheremainingeventsmaybe
extracted by means of a Bayesian procedure. By integrating signal.
withaMonteCarlotechniquethelikelihoodoverthenuisance The upperlimit on the numberof observed signalevents is
parameters,wecalculate,foreachsignature,theposteriorprob- extractedby meansof a Bayesian procedure. From the likeli-
abilitydensityfunction(pdf)fortheparameterN anddefine hoodfunctionforPoissondistributeddatawithunknownmean
sig
our90%C.L.upperlimitasthevaluewhereitsintegralreaches s and known backgroundb and using as prior pdf a flat prior
the90%ofthetotalarea(seeTab.2). differentfromzeroonlyforpositivevaluesof s,weextractour
To combine the results coming from the three signatures, 90%C.L.upperlimitonthenumberofobservedeventsasthe
valuewheretheintegraloftheposteriorpdfreachesthe90%of
the total area [18], n = 9.04. We can then set a limit on the
B
likelihoodsfordifferentQ-valuesweightingthembytheprobabilityofthatQ- half-lifeof2νβ+/ECdecayof120Te
valuebeingcorrect. Thisisequivalenttoconvolutingthepeakpositionwitha
gaussian,andthereforetoagaussianwhosewidthisthesuminquadratureof T2ν >ln2N ǫtotT =0.76·1020y (2)
theuncertaintyontheQ-valueandtheresolution. 1/2 ββ n
B
5
spectrum N N N N N 90%U.L.
sig flat L1 L2 H
I 1.7±3.5 214±16 17.5±5.3 14±5 34±6 8.5
II 0.3±3.1 78±10 26±6 N/A 9.5±4.3 7.25
III 0.0±0.5 1±1 N/A N/A 8±3 2.62
Table2: Resultsoftheunbinnedlikelihoodfitstotheenergyspectrumofthedoublecoincidences aroundE = 692.8keV(”I”,showninthetopplotofFig.5),
aroundE =1209.8keV(”II”,showninthemiddleplotofFig.5),andthetriplecoincidencesaroundE =692.8keV(”III”,showninthebottomplotofFig.5).
HereNsig estimatesthenumberofsignalevents,Nflat thetotalnumberofflatbackgroundeventsinthefitinterval, NL1andNL2arethenumberofeventsinthe
escapepeaksoflowerenergiesandNHisthenumberofeventsintheescapepeakofhigherenergies.Finally,the90%C.L.upperlimitsonthenumberofobserved
eventsarereported.
withrespecttotheconstraintsfrompreviousexperiments. The
limitsobtainedwithCUORICINOcouldbeimprovedinthefu-
turebyanadditional∼2ordersofmagnitudewiththeCUORE
experiment[19] because of increased mass, higher coincident
detectionefficiency,andlowerbackgrounds.
References
[1] Y.Fukudaetal.[Super-KamiokandeCollaboration],Phys.Rev.Lett.81
(1998)1562,Y.Fukudaetal.[Super-KamiokandeCollaboration], Phys.
Rev.Lett.82(1999)2430,S.Fukudaetal.[Super-KamiokandeCollabo-
ration],Phys.Rev.Lett.86(2001)5651,Q.R.Ahmadetal.[SNOCol-
laboration],Phys.Rev.Lett.87(2001)071301,Q.R.Ahmadetal.[SNO
Collaboration],Phys.Rev.Lett.89(2002)011301,K.Eguchietal.[Kam-
LANDCollaboration],Phys.Rev.Lett.90(2003)021802,T.Arakietal.
Figure6:Energyspectrumoftriplecoincidenceswithtwo511keVphotonsin [KamLANDCollaboration],Phys.Rev.Lett.94(2005)081801.
theenergyregionwhereweexpectasignalfromthe2νβ+/EC.Theredarrows [2] ExamplesofrecentreviewsareS.ElliottandP.Vogel,Ann.Rev.Nucl.
indicate theenergieswhereweexpecttheDEpeaksfromknownradioactive Part.Sci.52(2002)115,A.MoralesandJ.Morales,Nucl.Phys.B(Proc.
gammalinesintheCUORICINOinclusivespectrum.Foreachoftheseenergies Suppl.) 114 (2003) 141, F. T. Avignone III, S. R. Elliott and J. Engel
wesubtractanenergyregionof±3σ. Thelinesthatcangiveadouble-escape (2006),Rev.Mod.Phys.80(2008)481andK.Zuber,ActaPhys.Polon.
peakbetween50keVand692.8keVare1120.3keVand1238.1keVof214Bi B37(2006)1905.
(3),1173.2keVand1332.5keVof60Co(2)and1460.8keVof40K(1). [3] V.I. Tretyak and Yu.G. Zdesenko, ATOMIC DATA AND NUCLEAR
DATATABLES61,43(1995).
[4] V.I. Tretyak and Yu.G. Zdesenko, ATOMIC DATA AND NUCLEAR
whereN isthenumberof120TenucleiandT isthelivetime. DATATABLES80,83(2002).
ββ
[5] J.Abadetal.,J.dePhysique45C3(1984).
Theefficiencyiscalculatedasǫ =ǫ×(ǫ ǫ )3×(1−ǫ ),
tot noise heat corr [6] T.Bloxhametal.,Phys.Rev.C76,(2007)025501.
where ǫ, ǫnoise and ǫheat have been defined above and ǫcorr = [7] J.V.Dawsonetal.,Phys.Rev.C80,(2009)025502.
12.5±0.5%isthecorrectiontobeappliedtothedetectioneffi- [8] A.S.Barabashetal.,J.Phys.G34(2007)1721-1728.
ciencyofthe0νmodeandaccountsforthefactthatwearenot [9] A.S.Barabashetal.,J.Phys.Conf.Ser.120(2008)052057.
[10] N.D.Scielzoetal.,Phys.Rev.C80(2009)025501.
sensitive to the portionof the positronspectrumbetween30.5
[11] W.Bambyneketal.,Rev.Mod.Phys.Vol.49,No.1(1997).
and50keV(threshold)andthatwesubtractfive10keVenergy [12] C.Arnaboldietal.,Phys.Rev.C78(2008)035502.
regions from the experimental spectrum, as explained above. [13] http://nucleardata.nuclear.lu.se/nucleardata/toi/.
[14] A.Alessandrelloetal.,Nucl.Instrum.andMeth.B142(1998)163.
Since we do not know the exact shape of the positron spec-
[15] C.Arnaboldietal.,IEEETrans.Nucl.Sci.49(2002)2440.
trum from the 2νβ+/EC decay of 120Te the estimation of ǫ
corr [16] C.Arnaboldietal.,Nucl.Instr.Meth.A518(2004)775.
hasbeenperformedonastandardβspectrumwithend-pointat [17] C.Arnaboldietal.,IEEETrans.Nucl.Sci.50(2003)979.
692.8 keV and the error is estimated from the comparison of [18] C.Amsleretal.(ParticleDataGroup),PhysicsLettersB667,1(2008).
[19] R.Arditoetal.(CUORECollaboration),arXiv:hep-ex/0501010.
ǫ estimatedasexplainedaboveandǫ estimatedonaflat
corr corr
spectrumfrom0toendpoint.Systematicerrorsontheefficien-
cieshaveanegligibleimpactontheresult.
6. Conclusions
We searched for double beta decays β+/EC of 120Te in the
TeO cryogenicbolometersofCUORICINO,using0.0573kg·y
2
of120Teofdata.Weseenoevidenceofasignalandthereforeset
newlimitsonthehalf-lifefor0νand2νdecayT0ν >1.9·1021y
1/2
andT2ν > 0.76·1020 y extendingtheexclusionregionbyal-
1/2
most three orders of magnitude (four in the case of 0ν mode)
6