sign in sign up
<<
Home , Kamioka Observatory, Neutrino detector, Soudan 2

PROCEEDINGS OF THE 31st ICRC, ŁOD´ Z´ 2009 1

Search for GUT monopoles at Super-Kamiokande

Koh Ueno∗ for the Super-Kamiokande collaboration

∗, ICRR, , Higashi-Mozumi, Kamioka, Hida, Gifu 506-1205, Japan

Abstract. GUT monopoles captured by the ’s process above subsequently decay into positive , 0 + + gravitation are expected to catalyze decays (ρ ,ω,η,K , · · ·) → π , νe, νµ and ν¯µ are produced via the Callan-Rubakov process. In this scenario, pro- as described below, tons, which initially decay into pions, will ultimately + + π → µ + νµ (2) produce νe, νµ and ν¯µ. After going through + + oscillation, all neutrino species appear when they µ → e + νe +¯νµ (3) arrive at the , and can be detected by a 50,000 ton Cherenkov detector, Super-Kamiokande (SK). A search for produced through such a process has been carried out with SK and is expected 0.5 to give the cross section and velocity dependent limit 0.45 −23 −2 −1 on the monopole flux, F < 1 × 10 cm s 0.4 sr−1 for β = 10−3 and σ = 1 mb at 90%C.L. This is one order of magnitude more stringent a limit 0.35 than the previous result, which was obtained by the 0.3 Kamiokande experiment. 0.25 Keywords: GUT monopole, neutrino, water 0.2 ν ν± e h

Cherenkov detector Arbitrary Unit 0.15 ν h I.INTRODUCTION 0.1

Grand Unified Theories (GUT) predict superheavy 0.05 ν± e monopoles (GUT monopoles) produced in the very 0 early . GUT monopoles are predicted to have 0 10 20 30 40 50 60 70 appeared as topological defects at the phase transition of Eν (MeV) , where GUT gauge group spontaneously broke to leave the U(1) of electromagnetism. If we take as Fig. 1. Expected neutrino spectra at SK detector. an example the temperature of the phase transition at effect is taken into account. µ, τ components are grouped together as h 15 since these two flavor neutrinos have the same cross section in which the monopoles were formed equal to 10 GeV . and assume the average production rate of about one monopole per horizon at that time [1], the density of the monopoles today exceeds critical density of the universe After going through neutrino oscillation, all neutrino by more than 14 orders of magnitude. Even though the species appear when they arrive at the Earth (Fig. 1), inflationary universe scenario [2], [3] overcomes this and such low neutrino events can be detected by problem, the monopole flux in the universe depends on a water Cherenkov detector. some parameters such as monopole mass or the reheating II. DETECTOR AND EXPECTEDSIGNAL temperature, and therefore the uncertainty remains. In fact, due to the wide variety of elementary particle We searched for monopole-induced neutrinos using models, several models coexist with the Parker limit (∼ Super-Kamiokande (SK) [10], a large water Cherenkov 10−15 cm−2 s−1 sr−1) [4], [5], [6], and a flux in that detector located at Kamioka mine in Japan. SK is a range can be relatively easily detected by underground high-performance consisting of 11,146 experiments. Arafune et al. [7] pointed out copious low 20-inch PMTs and 50,000 tons of pure water. The energy neutrinos might be emitted when monopoles fiducial volume of this search, which amounts to 22,500 accumulating inside the Sun catalyze ; tons, is defined to be more than 2 meters from the walls of the inner detector as there are p → (ρ0,ω,η,K+, · · ·)+ e+(or µ+) (1) backgrounds near the wall. Monopole-induced neutrinos along their paths with cross sections typical of strong includes all six types of neutrinos, and electron scatter- − − interactions via so called Callan-Rubakov process [8], ing (νx(¯νx)+e → νx(¯νx)+e ) in the fiducial volume [9]. But, contrary to spontaneous proton decay, direct is used for the monopole-induced neutrino search. Fig. 2 neutrino production process , such as p → ν¯ + π+, shows the recoil electron spectra in SK, where the values is forbidden [8]. When decay mesons produced by the of the reference [11] are used as neutrino oscillation 2 UENO et al. GUT MONOPOLE SEARCH AT SUPER-KAMIOKANDE parameters. Matter effects on oscillation in the Sun are this cut. The source of is from the interaction also taken into account. of atmospheric neutrinos. with E > 18 ◦ MeV have a Cherenkov angle of θC ∼ 42 . Muons that III. DATA REDUCTION still remain in the monopole-induced neutrino search In this analysis, 1496 days’ data of SK-I taken from have pµ less than ∼ 350MeV, which corresponds to θC ◦ ◦ April 1996 to July 2001 is used. The backgrounds to ∼ 38 ; thus, the events with θC < 38 were removed the monopole-induced neutrino search are mainly atmo- from the data. The Cherenkov angle cut is also used to spheric neutrinos and muon-induced products. remove events with θC > 50 degrees, which eliminates To minimize the contribution of the latter backgrounds, events without a clear Cherenkov ring pattern, such as the lower energy threshold in this analysis is 18 MeV multiple γ rays emitted by a nuclear de-excitation. in total electron energy. The higher energy threshold is Some events originating from outside of the fidu- set to be 55 MeV, which is the end-point of the recoil cial volume have the possibility of being reconstructed electron spectrum (Fig. 2). within the fiducial volume of SK. γ-rays from the mate- rials of the detector structure and surrounding rock are the source of these backgrounds. To remove such events, 0.1 the reconstructed direction is projected backwards from 0.09 the vertex position until it reaches the inner detector ν ν± ν ν± wall, and events with this distance between the vertex 0.08 e+ e+ h+ h and the wall greater than 450 cm are removed from the 0.07 data. 0.06 Finally, interactions of atmospheric νµ ν 0.05 e produce muons and decay electrons. Some low energy muons ( 160 MeV) or decay electrons survive in this 0.04 ∼

Arbitrary Unit analysis but are tagged by preceding muons. Some frac- 0.03 tion of those muons which passed the Cherenkov angle ν 0.02 h cut can be identified by the following decay electron. ν± 0.01 h In order to remove those backgrounds we eliminate ν± e 0 the events which have time-correlated events before or 0 10 20 30 40 50 60 70 after the candidate events. However, the decay electrons Ee (MeV) whose mother muons are invisible (below Cherenkov threshold), Michel electrons, cannot be dropped by this Fig. 2. Recoil electron spectrum in the SK described analytically cut; how to handle these events is described later. without energy resolution included. How to read this figure is the same of Fig. 1 IV. ANALYSIS AND RESULTS The energy spectrum after each selection criterion is shown in Fig. 3. After all cuts are applied, 163 Besides setting the fiducial volume and energy crite- monopole-induced neutrino candidates remain in the ria, some other background reductions are additionally energy range from 18 MeV to 55 MeV. Most these candi- applied to the data. After a “standard cuts” (total charge dates are due to three kinds of irreducible backgrounds. cut, outer detector trigger cut, flasher event cut, etc...), The first background is spallation events from cosmic the data are subjected to a normal spallation cut. muons which still remain past the two spallation cuts ray muons can spall oxygen and induce unstable nuclei described above. called spallation products (µ + 16O → µ + X). It The other two backgrounds are ν and ν com- is one of the most abundant backgrounds in < ∼20 e µ ponents of atmospheric neutrinos. The “ν component” MeV region, and as described above, the ability to e means atmospheric ν and ν¯ , while the “ν component” remove the background determines the lower threshold e e µ is atmospheric ν and ν¯ which produce muons below of the monopole-induced neutrino search. The spallation µ µ Cherenkov threshold whose decay electrons are then background is reduced by a likelihood method that uses observed in the detector (invisible muon). The spectrum timing and track information of the muons preceding of these “invisible muons” is a Michel spectrum. the candidate events. The same algorithm is used in the To subtract these backgrounds, angular distribution standard analysis of SK. In addition to the with respect to the direction of the Sun was used. As normal spallation cut, a tighter criterion is then applied shown in Fig. 4, the data are divided into 20 angle bins in order to enhance the rejection efficiency of spallation and the following χ2 function is minimized with respect background. The spallation products with the shortest to α for each β. half lives, such as 11Li and 12N, tend to create spallation events with the highest . So, we remove events 20 2 {Nreal(i) − αNflat(i) − βNMC(i)} which occur within 0.15 sec from cosmic ray muons. χ2(α, β) ≡ Next, a Cherenkov angle cut is applied to the re- σ2 i=1 i maining data set. The remaining muons are removed by X (4) PROCEEDINGS OF THE 31st ICRC, ŁOD´ Z´ 2009 3

10 3 2.25 after 1st reduction after spallation cut 2 after ultimate spallation cut after goodness cut 1.75 after Cherenkov angle cut 10 2 after effective wall cut 1.5 after subevent cut 1.25

1

10 0.75

0.5

Events / bin 1496 days 22.5 kton 0.25 Events / MeV 1496 days 22.5 kton

1 0 20 25 30 35 40 45 50 55 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 θ Energy (MeV) cos Sun

Fig. 3. Spectrum of data at each reduction step Fig. 5. Expected angular distribution of the signal, where the total monopole-induced flux is set equal to 90%C.L. upper limit obtained from the analysis in this paper, 3.9 × 102 cm−2s−1.

18 where Bπ+ (∼ 0.5 for several GUT models) is the 16 branching ratio of proton decay into π++ anything. The 14 rate of monopole-catalyzed proton decay in the Sun is

12 given by, 3 10 fp = nMσvrelρHNAd x decay/s (6)

8 Z where nM is the monopole number density, σ the 6 catalysis cross section, Vrel = βrelc the relative veloc-

4 ity between the monopole and the hydrogen, ρH the hydrogen weight density, and NA Avogadro’s number Events / bin 1496 days 22.5 kton 2 (6.0×1023). gives a negligible contribution [12]. 0 Assuming the monopoles are accumulated at the center -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 3 −3 θ of the Sun (ρH = 55g/cm ,βrel = 1.7 × 10 ), one cos Sun 29 −1 obtains, fp < 9.0 × 10 (0.5/Bπ+ ) s (90%C.L.). 2 The rate fp is further expressed as fp =4πd I0/3/Rνe , Fig. 4. Angular distribution with respect to the direction of the Sun where d is the distance between the Sun and the Earth of the remaining 163 events. The horizontal axis shows the cosine of 8 + + the angle between the solar direction and the reconstructed direction (1.5 × 10 km), Rνe =Bπ (1 − Aπ ) the number of νe of an event. The average value is shown by the horizontal dotted line. produced in a proton decay. Aπ+ (0.2) is the absorption probability of π+ at the center of the Sun [7]. From the above arguments and the limits on I0,

σ0 Bπ+ 18 Here, we use Nreal(i) as the number of observed events NM < 1.6 × 10 (90%C.L.) (7) for the ith bin. Because the background shape is thought 1mb 0.5     to be almost symmetric with respect to the direction The monopole flux is calculated from the following of the Sun, the number of the background events in equation, assuming monopole-antimonopole annihilation the same exposure time is set to 0.05 for any ith bin is negligible, and expressed as Nflat(i). NMC(i) means the number 2 2 βesc of expected signal events for ith bin (Fig. 5) calculated NM = πR⊙ 1+ 4πFMt⊙ (8) 2 2 2 βM by Monte Carlo simulation. σi should be σstat,i +σsys,i,   ! but this time we haven’t included the systematic error −3 where βesc is the escape velocity (2 × 10 ), t⊙ the yet, and therefore set its value to zero for every i. Finally, time elapsed after the birth of the Sun (4.5 × 109yr). we obtained events as upper limit at 90 C.L. −3 β = 11.5 % This to (βM < 10 ), These 11.5 events correspond to the following limit 2 σ0 Bπ+ −23 βM on the total monopole-induced neutrino flux, FM < 1.5 × 10 1mb 0.5 10−3 2 −2 −1     −2 −1 −1  I0 < 3.9 × 10 (Bπ+ /0.5) cm s (90%C.L.) (5) cm s sr (90%C.L.)(9) 4 UENO et al. GUT MONOPOLE SEARCH AT SUPER-KAMIOKANDE

This limit is not immediately applicable to monopoles REFERENCES 17 heavier than 10 GeV, since the probability for the Sun [1] T.W.B.Kibble, J. Phys. A9 (1976) 1387. to trap monopoles decreases as their masses become [2] A.H.Guth, Phys. Rev. D23 (1981) 347. heavier. [3] K.Sato, Phys. Lett. B 99 (1981) 66. [4] E.N.Parker, Astrophys. J. 160 (1970) 383; M.S.Turner et al., Fig. 6 shows limits on the monopole flux for various Phys. Rev. D26 (1982) 1296. masses. This is one order of magnitude more stringent [5] G.Lazarides, C.Panagiotakopoulos, and Q.Shafi, Phys. Rev. Lett. a limit than the previous result, which was obtained by 58 (1987) 1707. [6] S.Dar, Q.Shafi, and A.Sil, Phys. Rev. D74 (2006) 035013. the Kamiokande experiment [13], [14]. [7] J.Arafune and M.Fukugita, Phys. Lett. B 133 (1983) 380. The present limit may be compared with the one [8] V.A.Rubakov, Pis’ma Zh. Eksp. Theor. Fiz. 33 (1981) 658 [JETP obtained from the X-ray excess of old neutron stars, Lett. 33 (1981) 644]; Nucl. Phys. B203 (1982) 311; V.A.Rubakov and M.S.Serbryakov, Nucl. Phys. B218 (1983) 240. [9] C.G.Callan, Phys. Rev. D25 (1982) 2141; D26 (1982) 2058. σ0 −23 −2 −1 −1 FM βrel < 3r × 10 cm s sr (10) [10] S.Fukuda et al., Nucl. Instr. and Meth. A 501 (2003) 418. 1mb [11] M. Maltoni et al., arXiv:hep-ph/0405172.   −3 [12] J.Arafune and M.Fukugita, Phys. Rev. Lett. 50 (1983) 1901. where βM is 10 , r the ratio of the total (neutrino + [13] T.Kajita et al., JPSJ 54 (1985) 4065. ) intensity to the photon intensity of old neutron [14] A.Sakai, master thesis (1993). 4 stars, and takes a value [15] between 1 and 10 . βrel in [15] E.W.Kolb and M.S.Turner, Astrophys.J. 286 (1984) 624. old neutron stars is of order of 0.1 ∼ 0.3. It is noted that the present limit depends on less ambiguous assumptions than that derived from old neutrons stars.

V. CONCLUSION In conclusion, this experiment has not found any ev- idence for monopole-catalyzed proton decay. The limits for the monopole flux are shown in Fig. 6.

-12 10

-13 10 MACRO (1989) UCSD II -14 10 Soudan-2 -15 Parker Bound 10 Kolar Baksan -16 Kamioka (Track Etch) 10 ) MACRO (2003) -1 -17 sr

-1 10

s Mica -2 -18 10

-19 10

-20 10

-21 10

-22 10 Neutron Star (no condensate) -23

Flux upper limit (90% C.L.) (cm 10

Kamiokande-I (274 days) -24 10 Kamiokande-III (319 days) -25 10 Neutron Star (pion condensate) -26 10 SuperKamiokande-I (1496 days) (This work)

-27 10

-5 -4 -3 -2 -1 10 10 10 10 10 1 β M

Fig. 6. 90% C.L. upper limits on the monopole flux as a function of monopole velocity, βM. Catalysis cross section, σ0, is assumed.