Solar and geoneutrino physics with Borexino

Marco G. Giammarchi n

Istituto Nazionale di Fisica Nucleare, Sezione di Milano, Italy

On behalf of the Borexino Collaboration1

a r t i c l e i n f o

Available online 6 December 2013

Keywords:Solar neutrinosGeoneutrinosScintillation detectorsLow background

a b s t r a c t

The Borexino detector is a high-radiopurity liquid scintillator for low background neutrino physics. Thedetector is located in the Hall C of the Laboratori Nazionali del Gran Sasso (central Italy). During its sevenyears of operation, Borexino has detected and measured solar neutrinos from 7Be, 8B and pep reactions inthe Sun, as well as geoneutrinos coming from the Earth.

& 2013 Elsevier B.V. All rights reserved.

1. Introduction

Solar neutrino physics began with the goal of studying the nuclearfusion reactions taking place in the core of the Sun. These reactionsproduce the solar energy and emit neutrinos that can be detected onEarth. The Davis experiment [1] was the first one to (radiochemically)detect solar neutrinos, measuring a significant deficit with respect tothe predicted flux. More experiments were performed starting fromthe end of the 1980s, both in radiochemical mode [2–4] and in real-time mode [5,6]. At the same time the widely accepted model of theSun structure and evolution evolved into what is known today as theStandard Solar Model [7,8].

Contrary to radiochemical experiments, real-time experimentshave an energy threshold of about 5 MeV, mainly due to naturalradioactivity. Because of this limitation, only � 0:001% of the solarneutrinos have been directly observed before 2007.

Borexino was specifically designed to measure neutrinos inreal-time and with a low energy threshold; this program has

required an intensive research and development, culminated withthe filling of the detector in 2007. Borexino [9] is a real timeexperiment to study sub-MeV neutrinos, using the νe-νe elasticscattering for solar neutrinos and the inverse beta decay forneutrinos coming from the Earth (geoneutrinos). The experimen-tal design threshold is of 50 keV while the analysis threshold is200 keV. The events are observed in a large mass (100 t) ofcarefully shielded liquid scintillator.

The prediction of the 7Be solar flux depends both on theStandard Solar Model and the value of the parameters of the largemixing angle (LMA) solution of neutrino oscillations [10,11]. Fig. 1shows the predicted spectrum at production, together with theenergy thresholds of past (on top) and future experiments. TheBorexino experimental program makes it possible to directly testthis prediction as well as opening up the previously unexploredterritory of real time sub-MeV solar neutrino spectroscopy.

This paper summarizes the main achievements of Borexino interms of detecting neutrinos from the Sun and the Earth duringthe years from 2007 to 2013. First of all, 7Be neutrinos from theSun were detected for the first time. Secondly, the 8B componentwas detected for the first time with an energy threshold below5 MeV, which was important to study the transition regionbetween the vacuum and the matter-dominated solar neutrinooscillations (see below). Thirdly, the pep solar neutrino flux wasmeasured for the very first time and the best limit was put on theCNO production cycle in the Sun. Finally, Borexino has measuredwith the highest confidence level to date neutrinos produced byradioactivity from the Earth itself (geoneutrinos).

2. The Borexino detector

Borexino [12] is an unsegmented scintillation detector featuring300 t of ultrapure liquid scintillator viewed by 2200 photomultipliers

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/nima

Nuclear Instruments and Methods inPhysics Research A

0168-9002/$ - see front matter & 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.nima.2013.11.088

n Tel.: þ39 0250317305.E-mail address: marco.giammarchi@mi.infn.it1 The Borexino Collaboration: G. Bellini, J. Benziger, D. Bick, G. Bonfini, D. Bravo,

M. Buizza-Avanzini, B. Caccianiga, L. Cadonati, F. Calaprice, P. Cavalcante,A. Chavarria, A. Chepurnov, D. D'Angelo, S. Davini, A. Derbin, A. Empl, A. Etenko,G. Fiorentini, K. Fomenko, D. Franco, C. Galbiati, S. Gazzana, C. Ghiano, M.Giammarchi, N. Goeger-Neff, A. Goretti, L. Grandi, C. Hagner, E. Hungerford, AldoIanni, Andrea Ianni, V.V. Kobychev, D. Korablev, G. Korga, Y. Koshio, D. Kryn, M.Laubenstein, T. Lewke, E. Litvinovich, B. Loer, P. Lombardi, L. Ludhova, G. Lukyan-chenko, I. Machulin, S. Manecki, W. Maneschg, F. Mantovani, G. Manuzio, Q. Meindl,E. Meroni, L. Miramonti, M. Misiaszek, P. Mosteiro, V. Muratova, L. Oberauer,M. Obolensky, F. Ortica, K. Otis, M. Pallavicini, L. Papp, L. Perasso, S. Perasso, A.Pocar, G. Ranucci, A. Razeto, A. Re, B. Ricci, A. Romani, N. Rossi, A. Sabelnikov, R.Saldanha, C. Salvo, S. Schoenert, H. Simgen, M. Skorokhvatov, O. Smirnov,A. Sotnikov, S. Sukhotin, Y. Suvorov, R. Tartaglia, G. Testera, D. Vignaud, R.B.Vogelaar, F. von Feilitzsch, J. Winter, M. Wojcik, A. Wright, M. Wurm, J. Xu, S.Zavatarelli, G. Zuzel.

Nuclear Instruments and Methods in Physics Research A 742 (2014) 250–253

(Fig. 2). Several layers of shielding with increasing purity are used inorder to define an inner Fiducial Volume of 100 t where the residualbackground is dominated by the intrinsic radiopurity of thescintillator.

The scintillator mixture is pseudocumene (PC) and PPO (1.5 g/l)as a fluor. It is contained in a transparent spherical vessel (NylonSphere, 100 μm thick), 8.5 m of diameter, and surrounded by1000 t of high-purity buffer liquid (PC with the addition of DMPas light quencher). The photomultipliers are supported by aStainless Steel Sphere, separating the inner part of the detectorfrom the external shielding, provided by 2400 t of ultrapure water.An additional containment vessel (Nylon film Radon barrier) isplaced between the Scintillator Nylon Sphere and the photomul-tipliers, serving the purpose of reducing Radon diffusion towardsthe inner part of the detector.

The outer water shield is equipped with 200 outward-pointingphotomultipliers; they are used to veto for penetrating muons, theonly significant remaining cosmic ray background (� 1muon m�2 h�1) at the Gran Sasso depth (about 3700 m of waterequivalent).

The 2200 inner photomultipliers are divided into a set of 1800devices equipped with light cones (so that they see light only from

the Nylon Sphere region) and a set of 400 PMTs without any lightcone, sensitive to light originated in the whole Stainless SteelSphere volume. This design increases the capability of rejectingthe background generated by muons crossing the PC buffer (andnot the scintillator).

Borexino features several external systems conceived to purifythe fluids (water, nitrogen and the scintillator itself) used by theexperiment (see e.g. Ref. [13]).

The Borexino detector was completed in 2007 and the experi-mental data taking in the final configuration began in May 2007.

The main detection reaction of Borexino, νe-νe, is sensitive toall neutrino flavors, while having a higher cross-section forelectron neutrinos. The Reines–Cowan inverse beta decayνp-eþn reaction is also used, especially for the detection ofgeoneutrinos, which are produced as electron antineutrinos.

The energy deposited in the active target produces scintillationlight which is collected by the photomultipliers. The energy of theevent can be reconstructed from the number of photoelectrons(� 500=MeV), while the position of the event is reconstructedfrom the photoelectron arrival times.

3. 7Be flux measurement

Borexino reported the first detection of 7Be solar neutrinos afew months after the start of the data taking [14]. The evidencewas based on detecting the recoil spectrum of the electron fromthe νe-νe elastic scattering. Since solar neutrinos from 7Be (thehigher-energy component) have 0.862 MeV of energy, the featurebeing searched for is a Compton-like shoulder at 664 keV.

Cuts were applied to remove muons and Rn daughters, todiscriminate against alpha particles (notably from 210Po) and 85Kr.The radiopurity of the scintillator was also essential in producingthe evidence of the 664 keV Compton-like neutrino shoulder(generated by 861 keV monochromatic neutrinos, Fig. 3). Thisconstituted the first experimental evidence of the 7Be nuclearreaction inside the Sun.

Subsequent analyses have profited from better statistics [15] anda subsequent intensive calibration campaign [19], so that thebest measurement of 7Be solar neutrinos in Borexino is now46:071:5ðstat:Þþ1:5

�1:6ðsyst:Þ counts=ðday 100 tÞ, in excellent agree-ment with the Standard Solar Model and the Mikheyev–Smirnov–Wolfenstein (MSW) LMA neutrino oscillation mechanism.

A day–night asymmetry study was also performed on the7Be solar neutrino rate [20]. The result found in Borexino wasof 2ðN�DÞ=ðNþDÞ ¼ 0:0170:012ðstat:Þ70:007ðsyst:Þ, consistentwith no significant day–night variation.

Fig. 1. Solar neutrino spectrum in the Standard Solar Model (see text).

Stainless Steel Water Tank18m∅

Stainless SteelSphere 13.7m ∅

2200 8" Thorn EMI PMTs(1800 with light collectors

WaterBuffer

100 ton fiducial volume

Borexino Design

PseudocumeneBuffer

Steel Shielding Plates8m x 8m x 10cm and 4m x 4m x 4cm

Scintillator

Nylon Sphere8.5m ∅

Holding Strings

200 outward-pointing PMTs

Muon veto:

Nylon filmRn barrier

400 without light cones)

Fig. 2. Schematics of the Borexino detector (see text).

Fig. 3. 7Be solar neutrino fit, with the signal and the various backgroundcomponents, from Ref. [19].

M.G. Giammarchi / Nuclear Instruments and Methods in Physics Research A 742 (2014) 250–253 251

4. 8B measurement

Solar neutrinos from 8B are measured in Borexino [16] bystudying the high-energy part of the spectrum, starting from3 MeV; this limit is imposed by the presence of the 208Tl contam-ination. For this analysis, muon and cosmogenic background hadto be treated with a special care. 214Bi and 208Tl removal wereperformed together with neutron rejection. Fig. 4 shows the finalspectrum obtained after the cuts. The fitted number of 8B events,0:2270:04ðstat:Þ70:01ðsyst:Þ counts/(day 100 t), is in good agree-ment with the Standard Solar Model and the MSW-LMA oscillationmechanism for neutrinos.

5. The pep first detection and the CNO limit

Observation of solar neutrinos in the 1.0–1.5 MeV energy rangeposes a special experimental challenge. First of all, 11C background,of cosmogenic origin, is a βþ emitter that is copiously producedeven at the Gran Sasso depth ð � 30 events=ðday 100 tÞÞ. Secondly,other backgrounds like 10C and 210Bi need to be addressed.

This spectral range is of great interest for two reasons. First ofall, the pep component of the solar neutrino spectra – a mono-chromatic neutrino line – can be found in this region; thiscomponent is interesting because it occurs at the very beginningof the pp production cycle in the Sun and is therefore wellconstrained by solar models. Secondly, this range offers thepossibility to look for CNO production in the Sun, which ispredicted to be � 1% of the pp cycle and has never been observedbefore.

Similar to the case of the 8B analysis, 11C background wasreduced by using the threefold coincidence (between the parentmuon, a spallation neutron and the 11C βþ decay itself). Inaddition, the pulse-shape difference between e� and eþ (from11C) was measured in organic liquid scintillators; a small differencein the time distribution of the scintillation signals arises in factfrom the finite lifetime of the ortho-Positronium (formed only bythe eþ) [21]. This effect was taken into account in the final fit.

Fig. 5 shows the spectra before and after the final backgroundsubtraction, with the first observation of the pep solar component [22]at 3:170:6ðstat:Þ70:3ðsyst:Þ counts=ðday 100 tÞ which agrees withthe solar models and the MSW-LMA solution. A pep-CNO correlatedanalysis was made to study the robustness of the pep result; a CNOflux upper limit was found at o7:9 counts=ðday 100 tÞ at 95% CL(pep fixed at the Standard Solar Model value), which is the best limitto date.

Thanks to the detection of several different solar neutrinocomponents (7Be, 8B and pep), Borexino makes a decisive con-tribution to the measurement of the survival probability ofelectron neutrinos produced in the Sun's core. The behavior ofthe survival probability as a function of energy clearly shows thetransition between the vacuum and the (MSW) matter-dominatedenergy range (Fig. 6).

6. Geoneutrinos

Geoneutrinos are electron neutrinos produced in β decays ofnaturally occurring radioactive isotopes in the interior of the Earth.They are produced in the decays of 40K and in the chains ofradioactive isotopes 238U and 232Th. The detection reaction in theBorexino scintillator, νp-eþn (having a 1.806 MeV threshold),

Fig. 4. 8B spectrum reconstructed in Borexino together with the relevant residualbackground components (from Ref. [16]).

Fig. 5. pep solar spectrum before and after the subtraction of the relevantbackground components (see Ref. [22]).

Fig. 6. Survival probability of electron neutrinos produced in the Sun andmeasured by Borexino and other experiments.

M.G. Giammarchi / Nuclear Instruments and Methods in Physics Research A 742 (2014) 250–253252

makes it possible to detect only a part of the 238U and 232Thantineutrinos (and no antineutrinos from 40K).

The positron in the final state comes rapidly to rest in thescintillator and annihilates by emitting two 511 keV γ rays, giving aprompt signal with a visible energy of EðνÞ�0:782 MeV, while thefree neutron is typically captured on protons with a mean lifetimeof � 260 μs. This results in the emission of a 2.22 MeV de-excitation γ-ray providing a delayed coincidence event. Back-ground rejection must be performed with a special emphasis onaccidental coincidences, primary muons producing a secondaryneutron and cosmogenically produced neutron emitters such as9Li and 8He. The high radiopurity of the Borexino scintillator wasalso critical in keeping the ðα;nÞ background small.

The main residual background is due to European nuclearpower reactors, producing electron antineutrinos with energiesup to 10 MeV.

Geoneutrinos in Borexino were first observed in Ref. [17], and amore precise measurement has been published that spans 1353 daysof data [18]. Fig. 7 shows the evidence of geoneutrino signal [18] abovethe reactors background. The signal corresponds to 14.374.4 geo-neutrino events to be compared with a total reactor rate of 31:2þ7:0

�6:1

events. The quality and the statistics of the data make it possible to tryto disentangle the U and Th different contributions, as depicted inFig. 7. The resulting Uranium and Thorium geoneutrino fluxes fromthe best fit to the data are ΦðUÞ ¼ ð2:171:5Þ � 106 cm2 s�1 andΦðThÞ ¼ ð2:673:1Þ � 106 cm2 s�1 respectively.

7. Conclusions

The Borexino detector is a low background detector mainlydevoted to solar neutrinos and geoneutrinos. Several componentsof the solar neutrino flux have been measured, namely 7Be, 8B andpep. Most of the solar neutrino spectrum is becoming directlydetectable for the first time thanks to Borexino, giving us a uniquetool to study the fusion reactions at the center of our star. Inaddition, geoneutrinos were identified and measured and a firsttry at separating U and Th geoneutrino contribution has beenperformed.

References

[1] R. Davis, Nobel Prize Lecture, 2002.[2] W. Hampel, et al., Physics Letters B 447 (1999) 127.[3] J.N. Abdurashitov, et al., Physical Review Letters 83 (1999) 4686.[4] M. Altmann, et al., Physics Letters B 616 (2005) 174.[5] S. Fukuda, et al., Physical Review Letters 86 (2001) 5651;

S. Fukuda, et al., Physics Letters B 539 (2002) 179.[6] Q.R. Ahmad, et al., Physical Review Letters 87 (2001) 071301.[7] J.N. Bahcall, M.H. Pinsonneault, Physical Review Letters 92 (2004) 121301.[8] A. Serenelli, et al., Astrophysical Journal Letters 705 (2009) L123.[9] G. Alimonti, et al., Astroparticle Physics 16 (2002) 205.[10] J.N. Bahcall, et al., JHEP 0408 (2004) 016.[11] G.L. Fogli, et al., Progress in Particle and Nuclear Physics 57 (2006) 742.[12] G. Alimonti, et al., Nuclear Instruments and Methods in Physics Research

Section A 600 (2009) 568.[13] G. Alimonti, et al., Nuclear Instruments and Methods in Physics Research

Section A 609 (2009) 58.[14] C. Arpesella, et al., Physics Letters B 658 (2008) 101.[15] C. Arpesella, et al., Physical Review Letters 101 (2008) 091302.[16] G. Bellini, et al., Physical Review D 82 (2010) 033006.[17] G. Bellini, et al., Physics Letters B 687 (2010) 299.[18] G. Bellini, et al., Physics Letters B 722 (2013) 295.[19] G. Bellini, et al., Physical Review Letters 107 (2011) 141302.[20] G. Bellini, et al., Physics Letters B 707 (2012) 22.[21] D. Franco, G. Consolati, D. Trezzi, Physical Review C 83 (2011) 015504.[22] G. Bellini, et al., Physical Review Letters 108 (2012) 051302.

Fig. 7. Geoneutrino components as seen by Borexino. U, Th and nuclear reactorsare shown (see text).

M.G. Giammarchi / Nuclear Instruments and Methods in Physics Research A 742 (2014) 250–253 253