CUOREOliviero Cremonesi INFN - Sez. Milano Bicocca
CUORE October 2, 2017 - Meeting of the LNGS Scientific Committee
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
2
Outline
• Status
• First results
• Plans
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
3
CUORE(Cryogenic Underground Observatory for Rare Events)
Closely packed array of 988 TeO2 crystals arranged in 19 towers
130Te: • large transition energy: Qββ (130Te) 2527.5 keV • highest natural isotopic abundance (33.8%)
CUORE projected sensitivity (5 years, 90% C.L.):T1/2 > 9 × 1025 yr
Primary goal: search for 0νββ decay in 130Te
CUORE design parameters: • mass of TeO2: 742 kg (206 kg of 130Te ) • low background aim: 10-2 c/(keV⋅kg⋅yr)• energy resolution: 5 keV FWHM in the Region Of Interest (ROI) • high granularity • deep underground location • strict radio-purity controls on materials and assembly
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
4
The CUORE cryostat
DetectorTowers
Top LeadShield
Side LeadShield
300 K
40 K
4 K
600 mK50 mK
10 mK
Bo�om LeadShield
Plates:• Designed to cool down ~1 ton detector to ~10 mK • Mechanically decoupled for extremely low vibrations • Low background environment
• Cryogen-free cryostat • Fast Cooling System (4He gas) down to ~50K • 5 pulse tubes cryocooler down to ~4K • Dilution refrigerator down to operating temperature ~10 mK • Nominal cooling power: 3 μW @ 10mK • Cryostat total mass ~30 tons • Mass to be cooled < 4K: ~15 tons • Mass to be cooled < 50 mK: ~3 tons (Pb, Cu and TeO2)
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
5
The CUORE Collaboration
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
6
CUORE construction• June 2016: last steps of the CUORE construction process
• Installation of the front-end electronics (2nd floor) • Installation of the DAQ (2nd floor) • Detector installation preparation
• August 2016: detector installation
• Installation of cryostat interfaces • Test of readout wires • Close-out of the cryostat vessels • IVC eventually closed in October 2016
• September-November 2016:
• December 5, 2016: cool-down start
• December 2, 2016: system close-out complete
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
7
CUORE cooldown• Reached a stable base temperature of ~7
mK on Jan 27, 2017
• Cooldown to base T: ~ 3 days
• Lowest temperature reached: 6.7 mK
• 984/988 working detectors
• Excellent performance @ full load�t [d]
0 0.5 1 1.5 2 2.5 3 3.5 4
T[K
]
0.01
0.1
1
10
Still plate
HEX plate
MC plate
CUORE cool down
Start: 2017-01-23 10:00
• Stable operation
14.0
14.2
14.4
14.6
14.8
15.0
15.2
15.4
15.6
15.8
16.0
12/5/17 14/5/17 16/5/17 18/5/17 20/5/17 22/5/17
Mixing Chamber Temperature
Base
tem
pera
ture
[mK]
6
8
10
12
14
16
18
20
0 5 10 15Power on mixing chamber [µW]
HEXisconnectedtotoplead
dilu1onstarts
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
8
Pre-operationAfter the successful cool-down, we faced with the challenge of operating a thousand bolometers in a completely new system
A long list of activities • DAQ and front-end electronics optimization • Detector working points
• Select representative subset • Load curves • Temperature scan for the best operating conditions
• Noise reduction • vibration induced • electrically induced
End of March 2017: • optimization not yet complete • decision to start calibrations and science runs • selected working temperature: 15 mK
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
9
Transition to operationDetector Performance
• The key parameters expected from CUORE are: • energy resolution• backgrounds in ROI
• Prioritary to obtain informations on both• CUORE has opted for a gradual transition to operation, alternating science runs to technical
measurements aiming to achieve better performance and to demonstrate improvements towards the ultimate performance.
First Physics Result• With few weeks of data CUORE can report the world’s best limit on 0νββ from 130Te with a high
degree of confidence. • A preliminary physics result can demonstrate the successful operation of CUORE and be worth publishing
by itself.• Milestones:
• Preliminary results for TAUP 2017 • First CUORE physics paper by October 2017
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
10
Schedule for Early Physics with CUORE
2017:Feb-Mar: system optimisationApr-May: science runsJun-Jul: system optimisationAug-Sep: science runsOct-Nov: system optimisation
2018:Science runs
Firstresults:Apr-May2017(ds3018)
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
12
Science runs• Science operations started on April 14, 2017
• Dataset 1: very short (identified issue with the thermistor bias on about 1/3 of the channels)
• Reoptimization of the detector working point • Dataset 2: 3 weeks of physics data bracketed by
2 calibration periods (May 4 - June 11) • Second optimization campaign • Dataset 3: August - September 2017
Operational performance: • 984/988 operational channels• Excellent data-taking efficiency when in operations • Much improved detector stability, compared to
Cuoricino/CUORE-0 • Calibrations/physics ratio data still to be optimized to
maximize 0νββ sensitivity
65.4%
21.8%
9.4%3.3%
Physics CalibrationOther Test
May
June
2017
Acquired statistics for 0νDBD decay search: • natTeO2 exposure: 38.1 kg yr • 130Te exposure: 10.6 kg yr
Dataset 2 time breakdown
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
13
Calibration spectrum
• 232Th sources deployed inside the CUORE detector
• Energy spectrum of all the CUORE detectors
Energy [keV]500 1000 1500 2000 2500
Even
ts [c
ount
s / k
eV]
210
310
410
510Calibration spectrum
CUORE Preliminaryyr⋅Exposure: 38.1 kg
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
Energy [keV]2560 2580 2600 2620 2640 2660 2680
Even
ts [c
ount
s / 0
.5 k
eV]
0
2000
4000
6000
8000
10000
12000 CUORE Preliminaryyr⋅Exposure: 38.1 kg
14
Detector performance: energy resolution• 899 (90%) best performing channels used for initial
analysis; most discarded channels had poor line or pulse shapes, and should be recovered in future runs
• Average (“harmonic mean”) energy resolution in calibration runs: 10.6 keV FWHM @ 2615 keV
Energy [keV]2570 2580 2590 2600 2610 2620 2630 2640 2650 2660
]σ
Resid
ual [
2−1−0123
Energy [keV]2570 2580 2590 2600 2610 2620 2630 2640 2650 2660
Even
ts [c
ount
s / 2
keV
]
05
10152025303540 CUORE Preliminary
yr⋅Exposure: 38.1 kg
• Significantly better performance in physics data: (7.9 ± 0.6) keV FWHM @ 2615 keV
FWHM [keV]5 6 7 8 9 10 20 30 40 50 60 70
Chan
nels
0
20
40
60
80
100
120
Calibration resolution at 2615 keVCUORE Preliminary
yr⋅Exposure: 38.1 kg
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
15
Detector performance: line shape• Fit components:
• a flat background • a step-wise smeared background • a double gaussian for the main peak • a combination of gaussian escape lines
• A quadratic dependence of the energy resolution is determined from gamma lines in the physics spectrum
Energy [keV]2560 2580 2600 2620 2640 2660 2680
Even
ts [c
ount
s / 2
keV
]
1
10
210
310
CUORE Preliminaryyr⋅Exposure: 38.1 kg
CUORE Preliminaryyr⋅Exposure: 38.1 kg
Tower 9
Energy [keV]800 1000 1200 1400 1600 1800 2000 2200 2400 2600
Rela
tive
reso
lutio
n
0.5
0.55
0.6
0.65
0.7
0.75
0.8
Resolution relative to 2615 keV calibration line
CUORE Preliminaryyr⋅Exposure: 38.1 kg
CUORE Preliminaryyr⋅Exposure: 38.1 kg
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
16
Analysis steps
Essentially the same steps and procedures developed and used for CUORE-0 (Phys. Rev. C 93, 045503 (2016) - arXiv:1601.01334)
• Acquisition of triggered signals • Data preprocessing: estimation of raw parameters • Pulse filtering • Thermal Gain Stabilization (TGS) • Energy calibration • Particle event selection • Coincidence analysis • Energy spectrum
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
17
Background spectrum: γ regionDataset 2
Energy [keV]500 1000 1500 2000 2500
Phys
ics e
vent
rate
[cou
nts /
(keV
kg
yr)]
3−10
2−10
1−10
1
10
CalibrationPhysics
CUORE Preliminaryyr⋅Exposure: 38.1 kg
Energy [keV]2350 2400 2450 2500 2550 2600 2650 2700
Phys
ics e
vent
rate
[cou
nts /
(keV
kg
yr)]
3−10
2−10
1−10
1 CalibrationPhysics
CUORE Preliminaryyr⋅Exposure: 38.1 kg
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
18
Blinded spectrum• To blind our data we randomly move a
fraction of events from +/- 20 keV of 2615 keV to the Q-value and vice versa
• The blinding algorithm produces an artificial peak around the 0νDBD Q-value and blinds the real 0νDBD rate of 130Te.
• This method of blinding the data preserves the integrity of the possible 0νDBD events while maintaining the spectral characteristics with measured energy resolution and introducing no discontinuities in the spectrum.
Energy [keV]2400 2450 2500 2550 2600 2650 2700
Eve
nts [
coun
ts / k
eV]
02468
1012141618
Tl208
Bi214 Co60
CUORE Preliminaryyr⋅Exposure: 38.1 kg
CUORE physics spectrum (blinded)
Energy [keV]2400 2450 2500 2550 2600 2650 2700
Eve
nts [
coun
ts / k
eV]
02468
1012141618 Tl208
Bi214 Co60
CUORE Preliminaryyr⋅Exposure: 38.1 kg
CUORE physics spectrum (unblinded)
Energy [keV]2400 2450 2500 2550 2600 2650 2700
Eve
nts [
coun
ts / k
eV]
02468
1012141618 Blinded
UnblindedTl208
Bi214 Co60
CUORE Preliminaryyr⋅Exposure: 38.1 kg
• When all data analysis procedures are fixed the data are eventually unblinded
• The blinding procedure is more evident by comparing directly the two spectra
Qββ
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
19
Fit in the ROI• We determined the yield of 0νββ events by performing a simultaneous UEML fit in the energy region
2465-2575 keV • The fit has 3 components:
• a posited peak at the Q-value of 130Te • a floating peak to account for the 60Co sum gamma line (2505 keV) • a constant continuum background, attributed to multi scatter Compton events from 208Tl and surface
alpha events
Energy [keV]2480 2500 2520 2540 2560
]σ
Resid
ual [
2−1−0123
Energy [keV]2480 2500 2520 2540 2560
Even
ts [c
ount
s / 2
.5 k
eV]
0123456789 CUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgCUORE Preliminary
yr⋅Exposure: 38.1 kgQββUnblinded spectrum fit
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
20
Fit in the ROI
We have also evaluated limits according to “W. Rolke et al., Nucl. Instrum. Meth. A 551, 493-503 (2005)”: - Half-life limit (90% CL, including systematics): 6.1×1024 yr - Decay rate limit (90% CL, including systematics): 0.11×10-24 / yr - Median expected sensitivity: 3.7×1024 yr
• Best fit decay rate: (-0.03–0.04+0.07 (stat.) ± 0.01 (syst.))×10-24 / yr• Decay rate limit (90% CL, including systematics): 0.15×10-24 yr-1
• Half-life limit (90% CL, including systematics): 4.5×1024 yr• Median expected sensitivity: 3.6×1024 yr (arXiv:1705.10816)
• Profile likelihood • Integrated on the physical region
- Region of interest: 2465 to 2575 keV- ROI background index: (9.8–1.5+1.7) × 10-3 c/(keV⋅kg⋅yr)- Events in the region of interest: 50- Best fit for 60Co mean: (2504.8 ± 1.2) keV
Trigger and energy reconstruction (98.469 ± 0.009) %Anti-coincidence (99.3 ± 0.3) %Pulse shape analysis (64 ± 3) %All cuts except containment (62.6 ± 3.4) %0vββ containment (88.345 ± 0.085) %Total 0vββ efficiency (55.3 ± 3)%
Efficiencies
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
yr]-24Decay rate [100.1− 0 0.1 0.2 0.3 0.4 0.5 0.6
NLL
02468
101214161820
CuoricinoCUORE-0CUORECUORE + CUORE-0 + Cuoricino
CUORE Preliminaryyr⋅Exposure: 38.1 kg
21
Combination with previous results
• We combined the CUORE result with the existing 130Te
• 19.75 kg·yr of Cuoricino • 9.8 kg·yr of CUORE-0
• The combined 90% C.L. limit is T0ν > 6.6 × 1024 yr
Combined “Rolke” limit: 8.1×1024 yr [meV]lightestm
1−10 1 10 210
[meV
]ββ
m
1−10
1
10
210
310
(Preliminary)Cuoricino + CUORE-0 + CUORE limit (Te)
CUORE sensitivity (Te)
Normal hierarchy
Inverted hierarchy
0 0.2 0.4 0.6 0.8 1
1−10
1
10
210
310
Other isotopes
Mo
Ge
Xe
NME: Phys. Rev. C 91, 034304 (2015) Phys. Rev. C 87, 045501 (2013) Phys. Rev. C 91, 024613 (2015) Nucl. Phys. A 818, 139 (2009) Phys. Rev. Lett. 105, 252503 (2010)
Experiments: 130Te: 6.5 × 1024 yr from this analysis 76Ge: 5.3 × 1025 yr from Nature 544, 47–52 (2017) 136Xe: 1.1 × 1026 yr from Phys. Rev. Lett. 117, 082503 (2016) 100Mo: 1.1 × 1024 yr from Phys. Rev. D 89, 111101 (2014) CUORE sensitivity: 9.0 × 1025 yr
mββ < 210–590 meV
Addi1onalphysicsrun:Aug2017(ds3020)
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
23
Science runs• Total statistics doubled with respect to June 2017
• All the available data have been fully reprocessed • Adopted the usual blinding procedure • Data analysis almost complete • Expected release of results: Oct. 23,2017
• In the meantime ….
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
24
System optimization• Further optimization steps:
• Detector commissioning stopped in April to start science runs • Still room for improvements • Mainly directed to: detector performance and noise abatement
• Detector optimization campaigns: June-July and fall 2017 • Careful investigation/upgrades to the electronics grounding in
the CUORE Faraday cage • Active cancellation of the PT-induced noise • Optimization of the operating temperature and detector
working points • System upgrades • Software and analysis upgrades
just started
Oliviero Cremonesi - October 2, 2017 - Meeting of the LNGS Scientific Committee
CUORE
25
Conclusions• The cryostat is working extraordinarily well. • With 3 weeks of physics data we have accumulated higher exposure than CUORE-0/
Cuoricino and surpassed their limit. • Total exposure: 38.1 kg⋅y • Invaluable operational experience • Important information on detector performance, noise, resolutions, background levels
• Developed and debugged physics tools, stress-tested end-to-end data processing with quality appropriate for science results
• Background rates are consistent with the background model • New data taken during the summer (still embargoed) • Available data have been fully reprocessed
• Further improvements possible: • Detector optimization campaigns have been
scheduled, focused on improving the resolution through noise reduction.
• First results already available • More to come