Results on neutrinoless double beta decay of 130Te from

CUORICINO

Adam BryantUC Berkeley & LBNL

on behalf of the CUORICINO Collaboration

Japan-US Seminar on Double Beta Decay and NeutrinosWaikoloa, HI

October 12, 2009

Adam Bryant October 12, 2009/19

CUORICINO collaboration

2

M. Barucci, L. Risegari and G. VenturaDipartimento di Fisica dell' Università di Firenze e Sezione di Firenze dell' INFN, Firenze I-50125, Italy

L. Canonica1,2, S. Di Domizio1,2, E. Guardincerri2,3 and M. Pallavicini1,2

1 Dipartimento di Fisica dell'Universita' di Genova, Italy2 Sezione di Genova dell'INFN, Genova I-16146, Italy3 Laboratori Nazionali del Gran Sasso, I-67010, Assergi (L'Aquila), Italy

M. Balata, C. Bucci, P. Gorla, S. Nisi and C. TomeiLaboratori Nazionali del Gran Sasso, I-67010, Assergi (L'Aquila), Italy

E. Andreotti1,2, L. Foggetta1,2, A. Giuliani1,2, C. Nones1,2, C. Rusconi1,2 and C. Salvioni1,2

1 Dipartimento di Fisica e Matematica dell'Università dell'Insubria , Como I-22100, Italy2 Sezione di Milano Bicocca dell' INFN, Milano I-20126, Italy

V. PalmieriLaboratori Nazionali di Legnaro, Via Romea 4, I-35020 Legnaro (Padova) Italy ,‏

C.Arnaboldi1,2, C.Brofferio1,2, S.Capelli1,2, L.Carbone2, M.Carrettoni1,2, M.Clemenza1,2, O.Cremonesi2, E.Ferri1,2, E.Fiorini1,2, A. Giachero2, L.Gironi1,2, S.Kraft1,2, C.Maiano1,2, M. Martinez2, A.Nucciotti1,2, L. Pattavina1,2, M.Pavan1,2, G.Pessina2, S.Pirro2, E.Previtali2, D.Schaeffer2 and M.Sisti1,2 1 Dipartimento di Fisica dell'Università di Milano-Bicocca, Milano I-20126, Italy2 Sezione di Milano Bicocca dell'INFN, Milano I-20126, Italy

F.Bellini, R.Faccini, F.Orio and M.VignatiDipartimento di Fisica dell'Universita' di Roma La Sapienza e Sezione di Roma dell'INFN, Roma I-00185, Italy

J.W.Beeman1, A.Bryant2,4, E.E.Haller1,3, L.Kogler2,4 and A.R.Smith2 1 Materials Science Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720 USA2 Nuclear Science Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720 USA3 Department of Materials Science and Engineering, University of California, Berkeley, CA 94720, USA 4 Department of Physics, University of California, Berkeley, CA 94720, USA

M.J. Dolinski1,3, K.Kazkaz1, E.B. Norman1,2 M.Pedretti1 and N.D.Scielzo1

1 Lawrence Livermore National Laboratory, Livermore, CA 94550 USA2 Department of Nuclear Engineering, University of California, Berkeley, CA 94720 USA3 Department of Physics, University of California, Berkeley, CA 94720, USA

T.D. GutierrezCalifornia Polytechnic State University, San Luis Obispo, CA 93407 USA

F.T.Avignone III, I.Bandac, R.J.Creswick, H.A.Farach, C.Martinez, L.Mizouni and C.RosenfeldDepartment of Physics and Astronomy, University of South Carolina, Columbia SC 29208 USA

L.Ejzak, R.H.Maruyama and S.SangiorgioUniversity of Wisconsin, Madison, WI 53706 USA

66 collaborators15 institutions in Italy and U.S.

Adam Bryant October 12, 2009/19

Gran Sasso National Lab

3

Hall A

Hall BHall C

CUORICINOCUORE

CUORE R&D

‣ National laboratory of the Istituto Nazionale di Fisica Nucleare (INFN) of Italy

‣ Built adjacent to a highway tunnel through the Gran Sasso mountain range

‣ 1400 m of rock overburden (about 3500 meters water equivalent)

‣ Cosmic ray muon flux attenuated by about 106 to about 0.7 muons / m2 / h

Adam Bryant October 12, 2009/19

130Te as a 0νββ decay candidate

4

Natural isotopic abundance of 33.80%-no enrichment

Q-value of 2530 keV-favorable phase space-above most naturally occurring γ-rays

Nuclear Matrix Elements

0.0

1.0

2.0

3.0

4.0

5.0

<M

’0!!

RQRPA (gA

=1.25)

RQRPA (gA

=1.0)

QRPA (gA

=1.25)

QRPA (gA

=1.0)

76Ge

82Se

96Zr

100Mo

116Cd

128Te

130Te

136Xe

Rodin et al., arXiv:0706.4304v1

Adam Bryant October 12, 2009/19

The CUORICINO detector

5

5× 5× 5 cm3

3× 3× 6 cm3

‣ Predecessor to CUORE‣ 62 TeO2 crystals serve as source and detector‣ 40.7 kg of TeO2

‣ 2 small crystals were enriched to 75% in 130Te‣ 2 small crystals were enriched to 82.3% in 128Te‣ 11.6 kg of 130Te‣ Operated 2003–2008 at Gran Sasso National

Laboratory (LNGS) in Italy

790 g

330 g

Adam Bryant October 12, 2009/19

Bolometric technique

6

Dilution refrigerator cools crystals to about 8 mK.

Heat capacity of dielectric and diamagnetic crystals follows the Debye law at low temperatures:

C = β

�T

Td

�3

C ≈ 1 MeV/0.1 mK

The energy deposited by a single particle results in a measurable temperature rise.

Adam Bryant October 12, 2009/19

NTD thermistors

7

Neutron-transmutation-doped (NTD) Ge thermistors function as sensitive thermometers to measure the small temperature change, ΔT = E/C.

R(T ) = R0 exp�

T0

T

�1/2

Nuclear processes creating dopants

NTD thermistor resistance:

Adam Bryant October 12, 2009/19

Temperature pulses

8

NTD Ge thermistor

Time (s)0 0.5 1 1.5 2 2.5 3 3.5 4

Vo

lta

ge

(m

V)

2500

3000

3500

4000

4500

5000

5500

6000

6500

Thermistor voltage, after amplification

Slow signals, ok for a low background experiment

Excellent energy resolution:average resolution of CUORICINO big crystals was about 7 keV FWHM at 2615 keV ≈ 0.27%(CUORE goal is 5 keV FWHM at 2615 keV ≈ 0.19%).

However, no other information such as event location within crystal or particle identification

Adam Bryant October 12, 2009/19

Resolution of CUORICINO bolometers

9

∆E = ξ�

kBCT 2

Thermodynamic limit on energy resolution:

due to exchange of phonons with heat sink

ΔE ~ tens of eV for kg size crystals

(independent of energy)

connected to mixing chamber of dilution refrigerator (heat sink)

Extrinsic noise sources, mainly mechanical vibrations, determine observed resolutions.

Adam Bryant October 12, 2009/19

Calibration of CUORICINO

10

Energy (keV)500 1000 1500 2000 2500 3000

co

un

ts /

1 k

eV

0

1000

2000

3000

4000

5000

Calibration Spectrum

‣ Th source calibration every 1–2 months lasting 2–3 days

‣ Calibration data used to derive relationship between pulse amplitudes and energies and for obtaining resolutions at 2615 keV

511 keVe+e-

583 keV208Tl

911 keV228Ac

965 + 969 keV 228Ac

1588 keV (228Ac)+ 1593 keVdouble escapefrom 2615 keV

2104 keVsingle escapefrom 2615 keV

2615 keV208Tl

Adam Bryant October 12, 2009/19

Resolutions at 2615 keV

11

FWHM at 2615 keV (keV)4 6 8 10 12 14 16 18 20

Num

ber /

keV

0

2

4

6

8

10

12

14

16

18

20

22 crystals3 5 cm! 5 !CUORICINO 5 crystals3 5 cm! 5 !CUORICINO 5

FWHM at 2615 keV (keV)4 6 8 10 12 14 16 18 20

Num

ber /

keV

0

1

2

3

4

5

crystals3 6 cm! 3 !CUORICINO 3 crystals3 6 cm! 3 !CUORICINO 3

CUORICINO preliminary

CUORICINO preliminary

Adam Bryant October 12, 2009/19

Energy (keV)1000 2000 3000 4000 5000 6000

Even

ts /

2 ke

V

0

1000

2000

3000

4000

5000

6000

7000

8000

CUORICINO background spectrumCUORICINO background spectrum

Energy (keV)2480 2490 2500 2510 2520 2530 2540 2550 2560 2570 2580

Even

ts /

2 ke

V

0

10

20

30

40

50

60

CUORICINO background spectrumCUORICINO background spectrum

CUORICINO background spectrum

12

60Co

130Te Q-value

Crystal type background counts / (keV ⋅ kg ⋅ yr)

5 × 5 × 5 cm3 0.18 ± 0.01

3 × 3 × 6 cm3 0.20 ± 0.04

anti-coincidence cut applied

CUORICINO preliminary CUORICINO preliminary

Adam Bryant October 12, 2009/19

Anti-coincidence

13

The electrons from a 0νββ-decay event stop in one crystal 86.3% of the time (84.5% for the small crystals).

Many backgrounds, such as muons, alpha decays near crystal surfaces, and Compton-scattered gammas, deposit energy in multiple crystals in coincidence.

Adam Bryant October 12, 2009/19

Background sources

14

Energy (keV)2500 2600 2700 2800 2900 3000

Even

ts /

2 ke

V

0

10

20

30

40

50

60

CUORICINO background spectrumCUORICINO background spectrum

60Co 208Tl

Sum of 1173 keV and 1333 keV gamma-ray cascade in beta decay of 60Co. Most likely due to cosmic ray activation of copper.

from 232Th contamination in cryostat

Contribution Source

(50 ± 20)% degraded alpha particles from 238U and 232Th contaminations on copper surfaces

(10 ± 5)% degraded alpha particles from 238U and 232Th contaminations on crystal surfaces

(40 ± 10)% multiple Compton events from 2615 keV gamma of 208Tl

degraded alphas

0νββ decay

CUORICINO preliminary

Adam Bryant October 12, 2009/19

New 130Te Q-value measurements

15

DOUBLE-β-DECAY Q VALUES OF 130Te, . . . PHYSICAL REVIEW C 80, 025501 (2009)

TABLE I. Summary of frequency ratios ω2/ω1 and the resultingmass differences #M measured using two-pulse Ramsey methodtimes of 3000 and 1000 ms. The uncertainties include both statisticsand systematics, combined in quadrature.

Isotopes Time (ms) ω2/ω1 #M (keV)

130Te-130Xe 3000 1.0000208837(26) 2527.01(32)120Te-120Sn 3000 1.0000153639(142) 1715.96(159)132Xe-129Xe 3000 1.0232682365(25) 2793899.12(30)130Xe-129Xe 3000 1.0077478330(26) 930309.60(32)120Te-120Sn 1000 1.0000153436(141) 1713.69(157)128Te-128Xe 1000 1.0000072676(110) 865.87(131)132Xe-129Xe 1000 1.0232682457(81) 2793900.22(97)

This difference was found to be 2793899.12 ± 0.30 and2793900.22 ± 0.97 keV for the 3000- and 1000-ms mea-surements, respectively. These results agree with the value2793899.180 ± 0.070 keV from Ref. [33] using 931494.028 ±0.023 keV/c2 per u [34].

Most systematic effects cancel in the frequency ratiosbecause the measurements were performed for isotopes withnearly identical masses under identical experimental condi-tions. The results of the 132Xe-129Xe measurements confirmedthat mass-dependent systematic effects were <∼0.3 keV/u. Thisis consistent with previous measurements using the CPT massspectrometer [35]. Mass-dependent systematic effects weretherefore negligible because the ββ-decay mass differenceswere ×103 smaller.

The time and ion-number dependence of our results wereinvestigated in detail. Parent and daughter measurements wereinterleaved in time to minimize effects such as B-field drift.Over the course of the experiment, the measured sidebandfrequencies showed no evidence of drifts. Upper limits on thedrifts were 2.5 ppb/day for the 130Te-130Xe pair and 7 ppb/dayfor the other isotopes. These limits are consistent with ourprevious studies [21]. We assign uncertainties of 0.08 and0.2 keV for the 3000- and 1000-ms measurements, respec-tively, to account for the average time difference betweenmeasurements.

The ion detection rate was intentionally kept low (typically<10 ions detected per bunch) to limit frequency shifts thatcan arise from the trapped-ion space charge. The data wassorted into different spectra according to the number ofions detected per trap ejection. Sideband frequencies weredetermined for each of these binned spectra. Any shift causedby the trapped-ion populations was expected to be identicalfor parent and daughter measurements and this was verifiedin the analysis. For the 3000-ms measurements, linear shiftswere consistent with zero with a value of −0.05 ± 0.05 keV(−0.4 ± 0.4 ppb) per detected ion observed. For the 1000-msmeasurements, resolved linear shifts were observed. For eachisotope, the measured slope was consistent with 1.5 ± 0.2 keV(13 ± 2 ppb) per detected ion. These shifts canceled in thefrequency ratios and the Q value obtained for each ion-numberspectrum was identical within statistics. To be conservative, weused only data with !5 detected ions per bunch (resulting inan average of ≈2 detected ions per bunch in each case) todetermine the frequency ratios and Q values. Because the

Te130

Q v

alu

e (k

eV)

2524

2525

2526

2527

2528

2529

2530

2531

2532

2533(a)

Manitoba IImeasured

Manitoba IIw/auxiliary

AME2003FSU

this work

Te128

Q v

alu

e (k

eV)

864

865

866

867

868

869

870(b)

Manitoba IImeasured Manitoba II

w/auxiliary

AME2003

this work

Te120

Q v

alu

e (k

eV)

1690

1695

1700

1705

1710

1715

1720(c)

AME2003this work

2527.50

2527.52

2527.54

FIG. 2. Q values for (a) 130Te, (b) 128Te, and (c) 120Te determinedfrom recent measurements and evaluations. The values shown arefrom the Manitoba II deflection-type mass spectrometer (Manitoba IImeasured) [12], the Manitoba II result supplemented with auxiliarymass data (Manitoba II w/auxiliary) [12], the 2003 Atomic MassEvaluation (AME2003) [11], the Florida State University PenningTrap (FSU) [15], and the Canadian Penning Trap (this work). Notethe different scale in the inset displaying the FSU result.

MCP detection efficiency was <100%, we have imperfectknowledge of the number of ions held in the trap and thereforeeach ion-number bin contains data from a distribution of trappopulations. We assign systematic uncertainties of 0.1 and0.4 keV for 3000- and 1000-ms measurements, respectively.These values are based on both the consistency of thefrequency shifts and on Monte Carlo estimates of the impactfrom extreme differences in trap-population distributions. Ineach case, the total uncertainty was dominated by statistics.The results of these measurements are discussed below anddisplayed in Table I and Fig. 2.

After the completion of this work, we became aware ofadditional Te and Xe mass measurements performed using

025501-3

Figure from N. D. Scielzo et al., Phys. Rev. C 80, 025501 (2009)

Two new measurements of the ββ decay Q-value of 130Te were published in 2009.

Q-value (keV) Reference

2530.3(2.0) 2003 Atomic Mass Evaluation recommended valueG. Audi, A. H. Wapstra, and C. Thibault, Nucl. Phys. A729, 337 (2003)

2527.01(32) N. D. Scielzo et al., Phys. Rev. C 80, 025501 (2009)

2527.518(13) M. Redshaw et al., Phys. Rev. Lett. 102, 212502 (2009)

Adam Bryant October 12, 2009/19

Results from CUORICINO

16

T 0ν1/2(

130Te) > 2.9× 1024 y (90% C.L.)

(preliminary)

18.0 y ⋅ kg 130Te

Energy

Counts/keV

11.5 y ⋅ kg 130Te

Q-value = 2530.3 keV Q-value = 2527.5 keV

T 0ν1/2(

130Te) > 3.0× 1024 y (90% C.L.)

Phys. Rev. C 78, 035502 (2008)

�mββ� < 0.19 – 0.68 eV �mββ� < 0.20 – 0.69 eV

Limit aided by downward fluctuation of background in signal region

CUORICINO preliminary

60Co

60Co

Adam Bryant October 12, 2009/19

Fitting technique

17

Simultaneous fit to three spectra

Energy (keV)2480 2490 2500 2510 2520 2530 2540 2550 2560 2570 2580

Even

ts /

keV

0

5

10

15

20

25

30

35

CUORICINO big crystalsCUORICINO big crystals

Energy (keV)2480 2490 2500 2510 2520 2530 2540 2550 2560 2570 2580

Even

ts /

keV

0

0.5

1

1.5

2

2.5

3

CUORICINO small enriched crystalsCUORICINO small enriched crystals

Energy (keV)2480 2490 2500 2510 2520 2530 2540 2550 2560 2570 2580

Even

ts /

keV

0

1

2

3

4

5

CUORICINO small natural crystalsCUORICINO small natural crystals

response function =N�

i=1

gaus(Q, σi)× exposure(i)

Background includes a flat component and a Gaussian component for 60Co.

Likelihood function is used to obtain the 90% C.L. limit, following the Bayesian method with a flat prior in the physical region.

60Co CUORICINO preliminary CUORICINO preliminary

CUORICINO preliminary

Adam Bryant October 12, 2009/19

Finalizing the analysis

‣ Analysis of complete CUORICINO data set is being performed with new software developed for CUORE – opportunity to debug, improve, and validate new software with existing data.

‣ Thorough checks of data quality are in the final stages.

‣ Potential improvements being tested: treating each detector separately in the analysis and doing a simultaneous fit to all independent detectors, each with its own resolution. We are also trying taking into account variation in the resolution over time.

18

Adam Bryant October 12, 2009/19

Conclusions

‣ CUORICINO has set the strongest limit on the 0νββ decay rate of 130Te:

‣ CUORICINO has set one of the most stringent limits on the effective neutrino mass:

‣ The analysis of the complete CUORICINO data set is being finalized.

‣ CUORICINO demonstrated the technology for CUORE in a large scale bolometric experiment that operated for several years with good energy resolution and low background.

19

T 0ν1/2(

130Te) > 2.9× 1024 y (90% C.L.)

�mββ� < 0.20 – 0.69 eV

(preliminary)

(preliminary)