Short Distance Neutrino with BoreXino · • Gallex and SAGE in the 90’s has made a calibration of their detector with an artiﬁcial neutrino source •Strong enough to produce

Short Distance Neutrino Oscillations with BoreXino

HeidelbergJul. 10th, 2013

Marco Pallavicini on behalf of the Borexino Collaboration

Dipartimento di Fisica - Università di Genova & INFN Sezione di Genova

Wednesday, July 10, 13

Heidelberg - July. 10th, 2013 M. Pallavicini - Dipartimento di Fisica - Università di Genova & INFN

Borexino experiment

• Mainly, a solar neutrino experiment• ν + e- → ν + e- in organic liquid scintillator• Very low background obtained with

selection, shielding e purifications• Low energy threshold, good energy resolution,

spatial reconstruction, pulse shape α/β identification• but also• Very good anti-neutrino detection (e.g. geo-neutrinos)

• sub-MeV νe detection: proved by 7Be and pep• sensitivity: as low as a few cpd/100 t• pep: 3.1 ± 0.6(stat) ± 0.3(sys) cpd/100 t

2

Phys. Rev. Lett. 107, 141302 (2011) Phys. Rev. Lett. 108, 051302 (2012)

• νe detection: proved by geo-neutrinos• total background:• << 1 events / year in the whole volume



The detector

3

Tank3300 m3 of water

210 PMTs Cherenkov

Principle:“graded shielding”.

Pure and pure materials toward the center of the

detector

Scintillator270 t PC-PPO

Nylon Vesselsinternal: R=4.25 mexternal: R=5.50 m

PIT ~ 1 m3 available at

8.25 cm from the center

Stainless Steel Sphere~1300 m3 of liquid

support for 2214 PMTs

18 m

16.9 m

13.7 m

8.25 m



Neutrino detection in Borexino

• Scintillation light detected by PMTs• # of photons → energy• time of flight → position• pulse shape → α/β β+/β-

4

57Co139Ce

203Hg85Sr

54Mn 65Zn 40K 60Co

ENERGY RESOLUTION10% @ 200 keV8% @ 400 keV6% @ 1 MeV

SPATIAL RESOLUTION35 cm @ 200 keV16 cm @ 500 keV

data

Calibration with γ sources

Montecarlo

α / β separation ( 214Bi - 214Po )

Cherenkov(small)

Scintillation(dominant)

~ 3-5 mm



Rate 7Be - I

• Experiment’s historical goal• Last paper improvements• Accurate energy calibration• Precise fiducial volume• Big effort on Monte Carlo tuning

5

Calibration source inside Borexino

calibration points


• Two quasi independent methods to check systematics

• Consistent results. Small difference includedin systematic error.• Final rate (100 t target): • 46.0 ± 1.5 (stat) ± 1.5 (sys) c d-1


Rate 7Be - II

6

Monte Carlo fit to the spectru, without α/β subtraction of the 210Po peak

Analytical fit of the spectrum afterα/β subtraction of 210Po peak

Source %

Trigger efficiency and stability < 0.1

Live time 0.04

Scintillator density 0.05

Fiducial volume +0.5 -1.3

Fit method 2

Energy response 2.7

Cuts efficiency 0.1

Total +3.4 -3.6

Phys. Lett. B658:101-108, 2008Phys. Rev. Lett. 101, 091302, 2008Phys. Rev. Lett. 107, 141302, 2011



Day-night modulation of 7Be rate

7

• Lack of modulation selects MSW-LMA

Adn = 2RN �RD

RN +RD= 0.001± 0.012± 0.007

Adn = 2RN �RD

RN +RD= 0.001± 0.012± 0.007

MeV0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Cou

nts

/5 k

eV 1

30 to

ns

210

310

410

Day spectrum

Night spectrum

MeV0.55 0.6 0.65 0.7 0.75

210

310

Day--night spectra

excluded by DN asymmetry

Phys. Lett. B707:22–26, 2012

No modulationobserved

MeV0.4 0.6 0.8 1 1.2 1.4 1.6

even

ts /

5 ke

V

-1500

-1000

-500

0

0.9 cpd/100t±Po = -21.8 210

0.57 cpd/100t±Be = 0.04 7

Be = 12 cpd/100t7

Night - Day spectrum

MeV0.55 0.6 0.65 0.7 0.75

even

ts /

5 ke

V

-50

0

50

100

150 LOW LOW



First pep detection and CNO limit

•We have got the first direct evidence of pep neutrinos and set a strong upper limit on CNO• Thanks to the low background and 11C rejection techniques• Tagging of 11C with triple coincidence• β+ - β separation exploiting positronium formation• Global multivariate analysis

• Triple concidence (TFC)

8

μ + 12C → n + 11C+ μ

n + p → D + γ (2.2 MeV)

11C→11B + e+ + νe

29.4 min

236 μsNo convection11C does not move

Fast neutron thermalization and capture

PHYSICAL REVIEW C 74, 045805 (2006)

Energy spectrum in FV

data no TFC cut

data after TFC cut

200 400 600 800 1000 1200 p.e.



PEP: β+ tagging with positronium

• Orto-positronium ~ 50% (in PC)• Signal is delayed by ~ 3 ns• Different pulse shape!

• Parameters measured ina dedicated setup

• Final selection base onBoosted Decision Tree (BDT)

9

“Typical emission time”

-20 0 20 40 60 80 ns

β+

scintillation light decay

Annihilation γs

β- da 214Bi-214Po

β+ da 11C (TFC)

Boosted Decision Tree

β+ β-



PEP - CNO: final result

• Rate: 3.1 ± 0.6(stat) ± 0.3(sys) cpd/100 t

•No oscillationsexcluded at97% c.l.

•No ν pepexcluded at98%

• Assuming MSW-LMA: • Φpep = 1.6 ± 0.3 108 cm-2 s-1

• CNO limit assuming pep @ SSM

• CNO rate < 7.1 cpd/100 t (95% c.l.)

10

Δχ2 profile with free pep and CNO

Borexino limit

Δχ2 profile for ν pep

SSM+MSW-LMA

SSMNo Osc.

PRL 108, 051302 (2012)



Phase I impact

11

2006 - Before Borexino Borexino 2012

7Be pep

8B

excluded by DN asymmetry

pp - all solar (w.o. BX)8B - all solar (Rad. + Cher. w.o. BX)HomestakeMSW prediction



Borexino background today

• A significant purification effort done in 2010/2011 to improve purity further

•Very effective on 85Kr, good on 210Bi, excellent for 238U and 232Th

12

210Bi rate = 16±4 cpd/100tons 85Kr rate = 7±5 cpd/100tons

p.e.

about 3 months of data early 2012

arbi

trar

y un

its • 85Kr• < 8.8 cpd / 100 t

• 2007-2010: 31.2 ± 5

• 210Bi• 18 ± 4 cpd / 100 t

• 2007-2010: 41.0 ± 2.8

• 238U• < 9.7 10-19 g/g

• 232Th• < 2.9 10-18 g/g

pp νregion

85Kr veryreduced

sharp 7Beshoulder

210Po peakdecays in 3 y

days

cpd/

100

tons



SOX: Short distance νe Oscillations with BoreXino

• Science•Motivations• Search for sterile neutrinos or other

short distance effects on Pee

• Measurement of ϑW at low energy (~ 1 MeV)• Measurement of neutrino magnetic moment• Check of gV e gA at low energy

• Technology•Neutrino source: 51Cr• Anti-neutrino source: 144Ce

• Project• SOX-A - 51Cr external• SOX-B - 144Ce external• SOX-C - 144Ce internal

13

ERC Ideas approved



A long standing idea

• The idea to deploy a source in Borexino dates back to the beginning of the project• Successfully implemented by Gallex (LNGS) and SAGE (Russia)• Recently, revised and re-proposed by many authors to search for sterile neutrinos

• N.G. Basov, V. B. Rozanov, JETP 42 (1985)Borexino proposal, 1991 (Sr90) J.N.Bahcall,P.I.Krastev,E.Lisi, Phys.Lett.B348:121-123,1995N.Ferrari,G.Fiorentini,B.Ricci, Phys. Lett B 387, 1996 (Cr51) I.R.Barabanov et al., Astrop. Phys. 8 (1997)Gallex coll. PL B 420 (1998) 114 Done (Cr51)A.Ianni,D.Montanino, Astrop. Phys. 10, 1999 (Cr51 and Sr90) A.Ianni,D.Montanino,G.Scioscia, Eur. Phys. J C8, 1999 (Cr51 and Sr90)SAGE coll. PRC 59 (1999) 2246 Done (Cr51 and Ar37)SAGE coll. PRC 73 (2006) 045805C.Grieb,J.Link,R.S.Raghavan, Phys.Rev.D75:093006,2007V.N.Gravrin et al., arXiv: nucl-ex:1006.2103C.Giunti,M.Laveder, Phys.Rev.D82:113009,2010C.Giunti,M.Laveder, arXiv:1012.4356SOX proposal - ERC 320873 - Feb. 2012 - approved Oct. 2012

14

a very incomplete list!

See White Paper and references therein:arxiv:1204.5379



The Science case - I

• A few well known experimental results do not match the standard three-flavors scenario. In particular:

• LSND (Los Alamos) in 2001 measured a νe excess using νμ beam• Apparently, a clear effect: 87.9 ± 22.4 ± 6.0 (3.8 σ)• L/E NOT compatible with “solari” oscillations• LSND region recently reduced by Icarus data, but not excluded

15

Venice 2013



The Science case - II

• Gallex and SAGE in the 90’s has made a calibration of their detector with an artificial neutrino source• Strong enough to produce a detectable neutrino flux (about the Sun at 10 m)• A portable Sun!

• Both experiments show a deficit w.r.t. expectations νe + 71Ga → 71Ge + e-

16

C. Giunti et al. arxiv:1210.5715 (hep-ph)

<R>=0.85 ±0.05

~ 3 σ effect



The Science case - III

• Reactor anomaly

•Many experiments at small L/E from reactors• Supposedly better calculations of reactor neutrio fluxes released recently• With these new calculations, neutrino deficit at small L/E is observed

17

Credit: T. Lasserre



Comments on the Science case

• In my opinion, taken individually, each anomaly is weak:• popular arguments, e.g.• LSND region not clearly confirmed by Miniboone, allowed region shrinked

significantly by Icarus• Gallex and SAGE calibrated their detector with sources. Can we trust the efficiency so

much to believe the anomaly ?• Can we trust the supposedly better reactor fluxes ? Were previous measurements

biased by older calculations?

• BUT

• All anomalies point consistently in the same direction, i.e. deficit at small L/E• If any of them is true, new physics is mandatory• High risk, high gain

•Methodologically, the only way to discard a wrong measurement is to do a better one• We can’t dismiss data based on theoretical prejudice

18Wednesday, July 10, 13


SOX: Three Phases

• Mission: test the existence of low L/E νe and/or νe anomalies by placing well known artificial sources close to or inside Borexino

• SOX-A• 51Cr source in pit beneath detector • 8.25 m from center [2015/2016]

• SOX-B• 144Ce-144Pr source in W.T.• PPO everywhere to enhance sensitivity • 7.15 m from center [2015/2016 ?]

• SOX-C• 144Ce-144Pr source in the center• Only after the end of solar program• More effort and more time

[>2016]

19

pittunnel

144CeSOX-C

144CeSOX-B

51CrSOX-A



Artificial neutrino sources

20

Source Production τ(days)

Decaymode

Energy[MeV]

Mass[kg/MCi]

Heat[W/kCi]

51Crνe

Neutron irradiation of 50Cr in reactor

Φn ≳ 5. 1014 cm-2 s-140

ECγ 320 keV (10%)

0.746 0.011 0.19

144Ce-144Prνe

Chemical extraction from spent nuclear fuel 411 β- <2.9975 0.314 7.6

144Ce-144Pr

Detectionthreshold

144Ce144Pr

51Cr



The tunnel beneath the detector

21

100 cm

Steel Floor of the Water Tank



Data analysis: two techniques

• Total counts: standard “disappearance” experiment• Total number of events depends on θ14 and (weakly) from Δm214

• Sensitivity depends on:• Statistics (source activity)• Error on activity (in particular) and on efficiency

• The relatively short life-time of 51Cr yield useful time-events correlation• The background is constant while the signal is not

• Spatial waves [C.. Grieb et al., Phys. Rev. D75: 093006 (2007)]

•With expected Δm2 e and ~ 1 MeV energy, the wavelength is smaller than detector size (~11 m max) and bigger than resolution (~ 15 cm)• The distribution of events as a function of distance to source shows waves• Direct measurement of Δm142 and θ14

• Very powerful and independent. Does not depend on knowledge of source activity.

• The two techniques can be combined in a single counts-waves fit



Geometry with external source

• Volume:

• Flux and decay

• Oscillations (one sterile)

• The number of νe-e- events at distance l from the source, with detection threshold T1 and maximum recoil energy T2:

23

Pee = 1.� sin2(2✓s) · sin2

✓1.27 �m2 l

E

◆

N0(l,T1,T2) = ne �(l) V(l) Pee(l,E)

Z T2

T1

d�e(E,T)

dTdT

R

d

l r

source

V (l) = 2⇡l2✓1� d2 �R2 + l2

2 d l

◆

�(l) =I04⇡l2

⌧e�tD⌧

⇣1� e�

t⌧

⌘

distance from external source (cm)300 400 500 600 700 800 900 1000 1100

arbi

trar

y un

its0

0.2

0.4

0.6

0.8

1

V(l) = 2⇡l2✓1� d2 �R2 + l2

2 d l

◆�(l) =I0

4⇡l2

V(l) ·�(l)

N.B.: The distribution of events is not uniform even without oscillations


Example for SOX-A


•Waves may be detected in the distribution of events as a function of the distance from source

• With waves, both parameters can be measured

24

Distance from the source [m]2 3 4 5 6 7 8 9 10 11 120

100

200

300

400

500

600

700

800Unoscillated overall spectrum

Oscillated overall spectrum

νCr 51

νBe 7

Po210

Other bg


100

200

300

400

500

600

700



MC Data

νCr 51

νBe 7

Po210

Other bg

Ideal curvesBorexino Background - No fluctuations

Full Geant4 simulation - exampleBorexino Background

14θ22sin0 0.05 0.1 0.15 0.2 0.25 0.3

142 m

Δ

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

Reactor anomaly central value

1 σ3 σ

sin2 2✓14

�m

2 14



Waves with νe and space-energy correlation

25

distance from center (cm)0 100 200 300 400 500 600

even

ts

2500

3000

3500

4000

4500

Δm2 = 1.0 eV2 sin2(2ϑs) = 0.11 year

1 m fiducial cut

positron en

ergy (M

eV)

positron en

ergy (M

eV)

distance from the source (m)

distance from the source (m)

•Space - Energy correlation

• With the 144Ce-144Pr source (both external SOX-B and internal SOX-C) global fit exploiting correlation between reconstructed event position and positron energy

distance from center (cm)0 100 200 300 400 500 600

even

ts

2500

3000

3500

4000

4500

Δm2 = 2.0 eV2 sin2(2ϑs) = 0.051 year


)14θ(22sin-210 -110 1

2 14m

Δ

-110

1

10

RA: 95% C.L.

RA: 99% C.L.

Cr: 95% C.L.51

Cr: 99% C.L.51

Solar: 95% C.L.

Solar: 99% C.L.


SOX-A sensitivity

26

Reactor+Ga anomaly region

• SOX-A:

• 51Cr source at 8.25 m from the center

• 10 MCi

• 1% precision in source activity

• 1% in FV determination

• Phase I can happen any time during next solar neutrino phase• 2015 is a realistic scenario



SOX-B sensitivity

27

)14θ(22sin-210 -110 1

2 14m

Δ

-110

1

10

RA: 95% C.L.

RA: 99% C.L.

Ce (water): 95% C.L.144

Ce (water): 99% C.L.144

Solar: 95% C.L.

Solar: 99% C.L.

• SOX-B

• 144Ce-144Pr source at 7.15 m from the center

• 75 kCi

• 1.5% precision in source activity

• 2% bin-to-bin errorto include all effects

• SOX-B can happen any time during next solar neutrino phase• 2015 is a realistic scenario - 1 y of data taking




SOX-C sensitivity

28

)14θ(22sin-210 -110 1

2 14m

Δ

-110

1

10

RA: 95% C.L.

RA: 99% C.L.

Ce (center): 95% C.L.144


Solar: 95% C.L.

Solar: 99% C.L.

• SOX-C:

• 144Ce-144Pr source in the center

• ~50 kCi

• 1.5% precision in source activity

• 2% bin-to-bin error toinclude other systematics

• SOX-C can happen only after the end of solar neutrino phase• 2016-2017 is a realistic scenario• desicison to be taken after SOX-A and/or SOX-B results




Other low energy neutrino physics

29

10 MCi

5 MCi

Magnetic moment

Reactorsμ < 5.6 10-11μB (90% CL)

Borexino (solar)μ < 3.6 10-11μB (90% CL)

Weinberg angle: δ(sin2ϑW)=2.6%

•With both sources (SOX-A and B or C)• Independent measurement of gv e ga • Test of SM EW running at very low energy• Standard Model• gV = -1/2 + 2 sin2ϑW = -0.038• ga = -1/2 = 0.5

CHARM II (1994)νμ ES su e- E ~ 10 GeV

SOX A+C


• Concept is the same as in Gallex 1994• ~36 kg, 50Cr enriched at 38% irradiated in a

high neutron flux reactor (we may use more material)• Candidate reactors: Russia (best), USA, Europe• 190 W/MCi from photons• ~few μSv/h on surface (required < 100)

•BUT: careful thermal design to handle 10 MCi (2 kW)• Preliminary studies are encouraging


Technology: 51Cr source

30

Gallex1994

External T must be acceptableCurrent value: T=90°C

Internal T must be below syntherization (750°C)Current value: T=260 °C

51Cr

W



The neutrino generator



Internal design



Final assembly

33

without cooling fins with cooling fins



Thermal studies

34

Bulk temperatures Surface temperatures



Technology: location for 51Cr source



Logistics at the Lab

• The neutrino generator will enter undergrounddirectly• It will stay in Hall C 4-6 months

36

rail

icarus pit entrance


• The neutrino generator will actually stay within a calorimeter for precise measurement of the activity


calorimetry

37

Tout%

Tin%

W,#Ni,#Fe#Alloy#

thermal#insula4on#

steel#

copper##



SOX-C: 144Ce source inside detector

• Very massive source• ~ 4 t of shielding• Source: spent nuclear fuel from Russia

• DENSIMET (W) shielding plus ultra-pure copper layer to reducebackground•W is very dirty for Borexino• γ background is a problem if rate too high• random coincidences make background

• Source deployment to be studied• Either from the top or from the bottom• PPO everywhere in the SSS to enlarge

active volume (active radius up to 5.5 m)• New anti-neutrino trigger • Trigger on singles would be too hard, but this is not a problem

• > 2016. No schedule yet.



Summary

•We plan to perform an extensive search of sterile neutrinos with neutrino and anti-neutrino sources

•SOX-A• 51Cr neutrino source (external)• Tentative schedule: 2015/2016

•SOX-B• 144Ce anti-neutrino source (external)• Tentative schedule:

2015-2016 (TBD)

•SOX-C• 144Ce anti-neutrino source (internal)• No schedule (>2016)

39

)14θ(22sin-210 -110 1

2 14m

Δ

-110

1

10

RA: 95% C.L.RA: 99% C.L.

Cr: 95% C.L.51

Cr: 99% C.L.51

Ce (water): 95% C.L.144

Ce (water): 99% C.L.144



Solar: 95% C.L.Solar: 99% C.L.


100

200

300

400

500

600

700



MC Data

νCr 51

νBe 51

Po210

Other bg

SOX-A SOX-C

positron en

ergy (M

eV)

positron en

ergy (M

eV)

distance from the source (m)distance from the source (m)


Documents

Short Distance Neutrino with BoreXino · • Gallex and SAGE in the 90’s has made a calibration of their detector with an artiﬁcial neutrino source •Strong enough to produce