NNuPhys2018-KozhuharovApril 22, 2019
Search for heavy neutrinos at CERN SPS
Venelin Kozhuharov INFN - Laboratori Nazionali di Frascati, Via E. Fermi 40, 00044 Frascati, ITALY
The phenomenology in the neutrino sector requires physics beyond theStandard Model. One possibility is the existence of new massive leptonicstates which could be probed at the high intensity machines. The presentresults on heavy neutral leptons from the study of kaon decays in flightwith the NA48/2 and NA62 experiments are presented and the futureprospects for such searches at CERN SPS are discussed. On behalf of the NA62 Collaboration: R. Aliberti, F. Ambrosino, R. Ammendola, B. Angelucci,A. Antonelli, G. Anzivino, R. Arcidiacono, T. Bache, M. Barbanera, J. Bernhard, A. Biagioni, L.Bician, C. Biino, A. Bizzeti, T. Blazek, B. Bloch-Devaux, V. Bonaiuto, M. Boretto, M. Bragadi-reanu, D. Britton, F. Brizioli, M.B. Brunetti, D. Bryman, F. Bucci, T. Capussela, J. Carmignani,A. Ceccucci, P. Cenci, V. Cerny, C. Cerri, B. Checcucci, A. Conovaloff, P. Cooper, E. Cortina Gil,M. Corvino, F. Costantini, A. Cotta Ramusino, D. Coward, G. DAgostini, J. Dainton, P. Dalpiaz,H. Danielsson, N. De Simone, D. Di Filippo, L. Di Lella, N. Doble, B. Dobrich, F. Duval, V. Duk,J. Engelfried, T. Enik, N. Estrada-Tristan, V. Falaleev, R. Fantechi, V. Fascianelli, L. Federici, S.Fedotov, A. Filippi, M. Fiorini, J. Fry, J. Fu, A. Fucci, L. Fulton, E. Gamberini, L. Gatignon, G.Georgiev, S. Ghinescu, A. Gianoli, M. Giorgi, S. Giudici, F. Gonnella, E. Goudzovski, C. Graham,R. Guida, E. Gushchin, F. Hahn, H. Heath, E.B. Holzer, T. Husek, O. Hutanu, D. Hutchcroft, L. Ia-cobuzio, E. Iacopini, E. Imbergamo, B. Jenninger, J. Jerhot, R.W. Jones, K. Kampf, V. Kekelidze, S.Kholodenko, G. Khoriauli, A. Khotyantsev, A. Kleimenova, A. Korotkova, M. Koval, V. Kozhuharov,Z. Kucerova, Y. Kudenko, J. Kunze, V. Kurochka, V. Kurshetsov, G. Lanfranchi, G. Lamanna, E.Lari, G. Latino, P. Laycock, C. Lazzeroni, M. Lenti, G. Lehmann Miotto, E. Leonardi, P. Lichard,L. Litov, R. Lollini, D. Lomidze, A. Lonardo, P. Lubrano, M. Lupi, N. Lurkin, D. Madigozhin, I.Mannelli, G. Mannocchi, A. Mapelli, F. Marchetto, R. Marchevski, S. Martellotti, P. Massarotti,K. Massri, E. Maurice, M. Medvedeva, A. Mefodev, E. Menichetti, E. Migliore, E. Minucci, M.Mirra, M. Misheva, N. Molokanova, M. Moulson, S. Movchan, M. Napolitano, I. Neri, F. Newson,A. Norton, M. Noy, T. Numao, V. Obraztsov, A. Ostankov, S. Padolski, R. Page, V. Palladino, A.Parenti, C. Parkinson, E. Pedreschi, M. Pepe, M. Perrin-Terrin, L. Peruzzo, P. Petrov, Y. Petrov,F. Petrucci, R. Piandani, M. Piccini, J. Pinzino, I. Polenkevich, L. Pontisso, Yu. Potrebenikov, D.Protopopescu, M. Raggi, A. Romano, P. Rubin, G. Ruggiero, V. Ryjov, A. Salamon, C. Santoni,G. Saracino, F. Sargeni, S. Schuchmann, V. Semenov, A. Sergi, A. Shaikhiev, S. Shkarovskiy, D.Soldi, V. Sugonyaev, M. Sozzi, T. Spadaro, F. Spinella, A. Sturgess, J. Swallow, S. Trilov, P. Va-lente, B. Velghe, S. Venditti, P. Vicini, R. Volpe, M. Vormstein, H. Wahl, R. Wanke, B. Wrona, O.Yushchenko, M. Zamkovsky, A. Zinchenko. Present address: Faculty of Physics, Sofia University, 5 J. Bourchier Blvd., 1164 Sofia, Bulgaria a r X i v : . [ h e p - e x ] A p r RESENTED AT
NuPhys2018, Prospects in Neutrino Physics CavendishConference Centre, London, UK, December 19–21, 2018 Introduction
A possible explanation of the neutrino oscillations is the extension of the StandardModel with three sterile Majorana neutrinos N i - the so called Neutrino MinimalStandard Model ( ν MSM)[1]. The mass of the lightest neutrino could be of the orderof 1 keV/c , making it a suitable Dark matter candidate, while the mass of the otherscould be in the range from 100 MeV/c to few GeV/c . The observed small massesof the left handed neutral leptons are provided through the see-saw mechanism. The( ν MSM) could also be extended by a scalar field χ to account for the inflation [2].In meson decays, the heavy neutrinos are produced through mixing with the or-dinary ones. For example, the rate for the decay K + → (cid:96) + N is given by [3]Γ( K + → (cid:96) + N ) = Γ( K + → (cid:96) + ν ) × ρ (cid:96) ( m N ) × | U (cid:96) | , (1)where m N is the heavy neutral lepton (HNL) mass, | U (cid:96) | is the mixing parameterbetween the HNL and the neutrino corresponding to the lepton (cid:96) , and ρ (cid:96) ( m N ) isa kinematic factor which includes also the helicity suppression in the electron case. ρ (cid:96) ( m N ) is O(1) for most of the accessible m N range.The decay width of the HNL is proportional to | U (cid:96) | × m N . Depending on itsmass and mixing, different scenarios are possible: • HNL decays within the fiducial region of the experiment. For m N <
500 MeV/c the possible final states are N → π ν , N → π ± µ ∓ , N → π ± e ∓ , N → ννν . • If | U (cid:96) | < − then γcτ N >
10 km and N could be considered invisible to theexperimental apparatus.The presence of Majorana mass term could manifest itself in the Lepton NumberViolating decay (LNV) K ± → π ∓ µ ± µ ± . In addition, the existence of a heavy neutrinowith 2 m µ < m N < m K ± − m π ± might appear as a resonance in the M πµ mass spectrumof the Lepton Number Conserving (LNC) decay K ± → π ± µ ± µ ∓ . A search for the decay N → π ± µ ∓ of a HNL produced in K + → (cid:96) + N has beenperformed by both NA48/2 and NA62 experiments.NA48/2 operated in 2003 and 2004 with simultaneous K + and K − beams with(60 ± σ ( p ) /p =(1.0 ⊕ p [GeV/c])%. The timing and a fast trigger condition were providedby a scintillator hodoscope with time resolution of 150 ps. The energy of photons1nd electrons was measured by a quasi-homogeneous liquid krypton electromagneticcalorimeter, providing resolution σ ( E ) /E = 3 . / √ E ⊕ /E ⊕ .
42% (energy is inGeV). The charged particle identification was based on the ratio E/p.Figure 1: Upper limit on the mixing parameter | U µ | as a function of the HNL massobtained from the study of the πµ mass distribution in K ± → π ± µ ± µ ∓ events (left)and from the search for LNV decay K ± → π ∓ µ ± µ ± (right).The K ± → π ± µ ± µ ∓ and K ± → π ∓ µ ± µ ± events were selected by requiring athree track vertex, one charged pion and two muons, same sign in case of K ± → π ∓ µ ± µ ± and oposite sign in the case of K ± → π ± µ ± µ ∓ . This limits the searchto short lived HNL. The signal was defined as | M πµµ − M K ± | < K ± → π ± µ ± µ ∓ candidates were selected with a background contributionof (0.36 ± K ± → π ∓ µ ± µ ± sample,with expected background 1.16 ± K ± → π ± π + π − events. No significant signal was observed in the LNV mode, leadingto the Br ( K ± → π ∓ µ ± µ ± ) < . × − , at 90% confidence level. The invariantmass distributions M π ± µ ∓ in the LNC sample and M π ∓ µ ± in the LNV sample wascompared with the Monte Carlo simulation. No signal consistent with the existenceof a resonance was observed. This allowed to put an upper limit on the branchingfractions Br ( K ± → µ ± N ) × Br ( N → π ± µ ∓ ), Br ( K ± → µ ± N ) × Br ( N → π ∓ µ ± )and on the HNL mixing parameter | U µ | , as shown in fig. 1 [5].The NA62 experiment [6], schematically shown in fig. 2, is primarily devoted tothe measurement of the Br ( K + → π + νν ) with 10 % precision [7]. The primary protonbeam from SPS interacts on a Be target to produce high intensity positive beam ofmomentum 75 GeV/c ± K + content. The secondary beam passesthrough a nitrogen filled threshold Cherenkov counter which is used for positive kaonidentification. Three stations of thin silicon pixel detectors provide a measurement2igure 2: Schematics of the NA62 experiment at CERN SPS.of the kaon momentum, flight direction, and time. A set of scintillating counters,CHANTI, provide a veto against interactions of the beam particles. The beam entersa 75 m long evacuated fiducial volume, followed by a spectrometer with a minimalmaterial budget. It is made of four chambers of straw tubes operated in vacuum andseparated by the MNP33 dipole magnet. A fast plastic scintillator charged hodoscopeis used in the trigger. The NA48 liquid krypton calorimeter with renewed readoutelectronics measures the energy deposited by the particles and also serves as a photonveto for photons with angles from 1.5 to 8.5 mrad with inefficiency less than 10 − forphotons with energy above 10 GeV. Twelve rings of lead glass counters surroundingthe decay region act as photon vetos for angles of the photons higher than 8.5 mradwith respect to the kaon flight direction. They are accompanied by two shashlyktype detectors covering photon angles down to zero. The π/µ separation is based onthe information from a neon filled ring imaging Cherenkov detector, measuring thevelocity of the charged particles, and three stations of muon detectors. Both KTAGand Gigatracker are exposed to the full 750 MHz hadron beam while the particle rateseen by the downstream detectors is at most 10 MHz. The high kaon flux combinedwith precise kinematics measurement, particle identification, and hermeticity makethe NA62 detector extremely powerful for the study of rare processes with kaons.Using a partial dataset, a search for the LNV modes K + → π − (cid:96) + (cid:96) + , both for (cid:96) = µ, e , was performed within NA62. The obtained invariant mass spectrum for theLNC modes K + → π + (cid:96) + (cid:96) − , which are used for normalization, is shown in fig. 3. The K + → π + µ + µ − sample is the world largest one, while the K + → π + e + e − analysisallowed the first observation of the decay in the mass range m ee <
140 MeV/c .The NA62 detector was fully operational during the 2017 and 2018 data taking andmajor improvements in the exclusive search for HNL is expected. The single eventsensitivity for the LNV modes is estimated to be SES ∼ − both for the electronand muon mode due to the negligible expected background.3igure 3: Invariant mass spectrum for the reconstructed π + (cid:96) + (cid:96) − candidates for themuon (left) and electron (right) channels. Only a partial NA62 dataset was used. Inclusive search for HNL can be performed by looking for “bumps” in the missingmass spectrum of the K ± → (cid:96) ± X decays, | m miss | = ( P K − P (cid:96) ) , with the chargedlepton being the only reconstructed particle in the final state.During the early stage of NA62, a large data sample of K + decays was col-lected in 2007 with a minimum bias trigger devoted to the measurement of R K =Γ( Ke / Γ( Kµ
2) and the test of the lepton universality [8]. The nominal kaon mo-mentum was 75 GeV/c and the MNP33 current was increased to provide a p T kick of265 MeV/c, leadint to momentum resolution of σ ( p ) /p = (0.048 ⊕ p [GeV/c])%.The nominal P K was calculated from K π events. The search for peaks was per-formed in the muon channel, in the mass range 300 MeV/c < m miss <
375 MeV/c with a step of 1 MeV/c . The dominant background was from K + → π µ + ν and K + → µ + ν µ ( γ ) decays and from the muon halo. No signal exceeding 3 σ above theexpected background was observed. The Rolke-Lopez method was applied and anupper limit on Br ( K + → µ + N ) was obtained [9].Using 5 days of data, recorded in 2015 with beam intensity corresponding to 1%of the nominal and a minimum bias trigger, the NA62 experiment performed a searchfor HNL in the missing mass spectrum both for the electron and the muon mode[10]. The already developed technique for the analysis of 2007 data was applied. Thecharged lepton momentum had to be between 5 GeV/c and 70 GeV/c. The searchregion for HNL was chosen to be 170 (250) MeV/c < m miss <
448 (373) MeV/c forthe electron (muon) channel. MC simulation was used to obtain the resolution σ ( m N )and to calculate the acceptance for K + → (cid:96) + N events as a function of the HNL mass.4igure 4: Upper limits on the branching fraction for the K + → (cid:96) + N (left) and theobtained upper limits on the mixing parameter | U (cid:96) | (right) as a function of theheavy neutrino mass. Data from previous experiments is also shown for comparison.The maximum value of the local signal significance z = ( N obs − N exp ) / (cid:113) N obs + δN exp was 2.2, obtained for m N = 283 MeV/c in the electron channel. The normalizationwas based on the reconstruction of the corresponding K + → (cid:96) + ν (cid:96) decay and was usedto obtain an upper limit on the branching fraction for the K + → (cid:96) + N decays, shownin fig. 4-left. These upper limits on BR were translated into upper limits on themixing parameters | U e | and | U µ | , shown in fig. 4-right.As in the exclusive search case, a major improvements on the presented resultscould be expected with the present NA62 data. The sensitivity of the presented searches at CERN SPS is limited by the statistics ofthe produced mesons, which afterwards may decay to final states with HNLs. Thislimitation could be overcome by entirely absorbing the primary proton beam andmost of its interaction products in a thick target, a technique known as beam-dump.The outgoing beam from the target consists of long lived neutral particles which entera decay region.The number of the expected HNL in the detector is a product of the producedHNLs and the probability for their observation [11]. The possible source of HNLs arethe (semi)leptonic decays of π , K , D , and B mesons and the produced number ofHNLs depends on the production of a given quark flavour in the target, its probabilityto hadronize to a certain meson and to decay to final states with HNL. The detection5robability is given by P det = (cid:20) e − liniγcτ − e − ldγcτ (cid:21) × BR ( N → visible ) × (cid:15) det (2)and depends on the length of the setup l ini before the decay region, the length ofthe decay region l d , the lifetime τ of the HNL, the visible branching fraction and thecorresponding acceptance (cid:15) det .The description of two selected facilities, a short term one (within 5 years) anda long term (10-15 years) one, discussed also within the Physics Beyond Colidersinitiative [12], follows. The NA62 beryllium target is followed by two ∼
11 nuclear interaction length watercooled copper-steel collimators (TAX) to stop the residual proton beam from SPS.The dump mode operation could be provided by closing the first TAX. This allowsto exploit the particle identification, tracking, and hermeticity of NA62 to search forlong lived neutral particles, including HNL. The goal is to collect O(10 ) POT inRUN3 [13].Figure 5: Expected sensitivity of NA62 in beam dump mode (left) [12] and SHiP(right) [11] to the mixing parameter | U µ | as a function of the HNL mass. Filled areais excluded by theory or previous experiments.The projected sensitivity of NA62 in beam dump mode (90% CL upper limit) forthe case of dominant mixing with the second generation ( | U e | : | U µ | : | U τ | = 0 :1 : 0) is shown in fig. 5-left. It was obtained assuming zero background and thepossibility to detect all two-track final states. The geometrical acceptance and thetrigger efficiency were taken into account.About 3 × POT have been collected by NA62 in a dedicated beam dumpmode. No background was achieved for fully reconstructed final states. Additionalsamples collected in the presence of kaon beam are currently being used for dimuonbackground studies. 6 .2 SHiP
The Search for Hidden Particles (SHiP) experiment, shown schematically in fig. 6,aims to collect 2 × protons on target in 5 years of data taking. It exploits ageneral purpose beam dump facility to complement the existing LHC programme inthe search for New physics.The primary proton beam from SPS interacts in a very dense target (11 λ I )followed by an hadron absorber. This assures abundant production of heavy flavourstates and also effectively stops the secondary protons and kaons. The target regionis followed by an active muon shield to deflect the muons. The neutron interactionsare reduced by evacuating the 60 m long decay region. The decay products aremeasured by a downstream magnetic spectrometer with large acceptance. A particleidentification system is also foreseen.Figure 6: SHiP experimental setup.The estimated sensitivity (90 % confidence region) to HNL are obtained underthe zero background assumption. The production fraction of B c mesons is unknownat SPS energies [14] and is part of the systematic uncertainty. For the case of mixingwith the second generation only the SHiP sensitivity is shown in fig. 5-right. A regionup to m N ∼ O(6 GeV/c ) and | U µ | down to 10 − could be covered, complementingthe possible reach with a Future Circular Colider (FCC) in e + e − mode.The activities on the design and the construction of the detectors for the SHiPexperiment are ongoing and almost every component has achieved at least first stageof prototyping [15]. A diverse heavy neutral leptons search programme is being executed at CERN withthe NA62 experiment, covering both exclusive and inclusive events reconstruction.7o far no signature of new states have been observed in the (semi)leptonic decaysof K ± . The obtained results are improving the existing limits on the HNL mixingparameters in the considered mass range.A possible increase of the sensitivity is related to the application of the beamdump technique. Beyond the K + → πνν phase of NA62 experiment, NA62++ couldcollect 10 POT in few months of operation during RUN3 (2021-2023). A dedicatedbeam dump facility with the SHiP experiment in the SPS North Area could furtherincrease the statistics to 2 × POT in 5 years of operation, starting during RUN4(after 2027).Both NA62++ and SHiP are part of the PBC working group and provide inputto the European Strategy for Particle Physics.
References [1] T. Asaka and M. Shaposhnikov,
Phys. Lett. B , 17 (2005).[2] M. Shaposhnikov and I. Tkachev, Phys. Lett. B , 414 (2006).[3] R. E. Shrock, Phys. Lett. , 159 (1980).[4] V. Fanti et al. [NA48 Collaboration], Nucl. Instrum. Meth. A , 433 (2007).[5] J. R. Batley et al. [NA48/2 Collaboration], Phys. Lett. B , 67 (2017).[6] E. Cortina Gil et al. [NA62 Collaboration], JINST , no. 05, P05025 (2017).[7] E. Cortina Gil et al. [NA62 Collaboration], Phys. Lett. B , 156 (2019).[8] C. Lazzeroni et al. [NA62 Collaboration], Phys. Lett. B , 326 (2013).[9] C. Lazzeroni et al. [NA62 Collaboration], Phys. Lett. B , 712 (2017).[10] E. Cortina Gil et al. [NA62 Collaboration], Phys. Lett. B , 137 (2018).[11] C. Ahdida et al. [SHiP Collaboration], arXiv:1811.00930 [hep-ph].[12] J. Beacham et al. , arXiv:1901.09966 [hep-ex].[13] T. Spadaro, NA62++, talk at PBC Annual Workshop, 16 - 17 Jan 2019.[14] K. Bondarenko et al. , JHEP1811