Beta beams Jacques Bouchez CEA Salay/ APC Paris Basics of beta beams The baseline CERN scenario Higher energy beta beams Other ideas Conclusions POFPA Cern, 18/10/2005

Oscillation maximum seen at ~ 30 km/MeV Solar oscillation: e disappearance  m 12 2 = eV 2 sin 2 2θ 12 =0.85 Atmospheric oscillation:  disappearance  m 23 2 = eV 2 sin 2 2θ 23 = 1. 0 KAMLANDSK PRESENT EVIDENCE FOR OSCILLATIONS Explains solar neutrino deficitStudied by MINOS, OPERA, ICARUS

Mixing matrix: the missing parameters 1   e          l = U l i i U is a unitary matrix: 3 angles :  12,  13,  23 plus 1 CP violating phase  3 masses m 1, m 2, m 3 SUN :   m 12 2 = eV 2,  12 ~ 35 o ATM :  m 23 2 = eV 2,  23 = 45 o Missing :  13 and the phase  both govern the    e oscillation at the atmospheric frequency We know that  13 is < 10 o (CHOOZ) we have to look for a small oscillation

Another look at the mixing matrix 12      ATM.     e ATM.  e     SOLAR. This “factorization” of effects is due to the smallness of  13 and the strong difference in oscillation frequencies

The quest for  13 ….. New reactor experiments (double-Chooz, …) Neutrino superbeams of first generation (T2K,NOvA) ….and for  CP Megaton detectors and… neutrino superbeams of second generation and/or neutrino betabeams and/or neutrino factories while solving as much as possible intrinsic, mass hierarchy and θ 23 ambiguities CERN Europe absent

L opt (km) ~ 0.5 E (MeV) (Initial design)

(September 2005)

SPL main goals: - increase the performance of the CERN high energy accelerators (PS, SPS & LHC) - address the needs of future experiments with neutrinos and radio-active ion beams The present R&D programme concentrates on low-energy (Linac4) items, wherever possible in collaboration with other laboratories. SPL current design 4 MW From R.Garoby

6o6o 2o2o 0.5 o 3o3o 1o1o SPL= 10 x T2K1 + sensitivity on  Same performance as T2K2 (megaton)

How to overcome superbeam limitations ? Main problem : SPL protons produce less negative pions, so less antineutrinos antineutrino cross-section ~ 5 times smaller than neutrinos So 10 SPL years have to be shared as ~ 2 neutrino + 8 antineutrino years The solution : Produce a e beam to study e    oscillation and run it SIMULTANEOUSLY with   beam from SPL Compare   e and e    (T asymetry, equivalent to CP asymetry) THIS WAS THE INITIAL MOTIVATION FOR A BETA BEAM

BETA BEAMS Concept proposed by Piero Zucchelli Produce radioactive ions (ISOL technique) Accelerate them in the CERN accelerator complex up to  of order 100 Store ions in a storage ring with long straight sections aimed at a far detector Advantages strongly focussed neutrino beam due to small Q value of beta decays (quality factor  /Q) very pure flavour composition (   contamination ~ ) perfectly known energy spectrum Baseline scenario first studied at CERN (Mats Lindroos and collaborators) and now part of the EURISOL TDS Strong synergy between beta beams and EURISOL

6 He production from 9 Be(n,  ) Converter technology preferred to direct irradiation (heat transfer and efficient cooling allows higher power compared to insulating BeO). 6 He production rate is ~2x10 13 ions/s (dc) for ~200 kW on target. Converter technology: (J. Nolen, NPA 701 (2002) 312c)

18 Ne production Spallation of close-by target nuclides – 24 Mg 12 (p, p 3 n 4 ) 18 Ne 10. –Converter technology cannot be used; the beam hits directly the magnesium oxide target. –Production rate for 18 Ne is ~ 1x10 12 ions/s (dc) for ~200 kW on target. – 19 Ne can be produced with one order of magnitude higher intensity but the half-life is 17 seconds!

Beta-beam base line design Strategy for the conceptual design study: –Design should be based on known technology. –Avoid large number of technology jumps, requiring major and costly R&D efforts. –Re-use wherever possible existing infrastructure (i.e. accelerators) for the “first stage” base line design. Major ingredients: –ISOL technique for production of radioactive ions. –Use CERN PS and SPS accelerators for acceleration –Selected ions: 6 He (anti e ) and 18 Ne ( e ) M.Benedikt CERN-ISS22/9/2005

Beta-beam baseline design Neutrino Source Decay Ring Ion production ISOL target & Ion source Proton Driver SPL Decay ring B  = 1500 Tm B = ~5 T C = ~7000 m L ss = ~2500 m 6 He:  = Ne:  = 100 SPS Acceleration to medium energy RCS PS Acceleration to final energy PS & SPS Beam to experiment Ion acceleration Linac Beam preparation ECR pulsed Ion productionAcceleration Neutrino source Low-energy part High-energy part M.Benedikt CERN-ISS22/9/2005

From dc to very short bunches 1 s 2x1s2x1s t B t B PS SPS 2x1s2x1s t B 1 s PS t 2x1  s to decay ring (2x4 bunches of <5 ns) PS: 1 s flat bottom with 8 (16) injections. Acceleration in ~1 s to top PS energy. Target: dc production during 1 s. 60 GHz ECR: accumulation for 1/8 (1/16) s ejection of fully stripped ~20  s pulse. 8 (16) batches during 1 s. RCS: further bunching to ~100 ns Acceleration to ~300 MeV/u. 8 (16) repetitions during 1 s. SPS: injection of bunches from PS. Acceleration to decay ring energy and ejection. Repetition time 8 s. 1 s7 s Post accelerator linac: acceleration to ~100 MeV/u. 8 (16) repetitions during 1 s. t

General collection/acceleration scheme Production PS SPS Decay ring Ramp time PS Time (s)0 8 No collection Reset time SPS Typically (for 6 He): 1 out of 8 produçable ions is collected 1 out of 4 collected ions makes it to the ring (factor 2 for decays, factor 2 for transmission losses between machines) Adapted from M.Lindroos Main losses Ramp time SPS

Simulation (in the SPS)

Goals - Status For the base line design, the aims are (cf. Bouchez et al., NuFact’03): –An annual rate of anti-neutrinos ( 6 He) along one straight section –An annual rate of neutrinos ( 18 Ne) at  =100 always for a “normalized” year of 10 7 seconds. The present status is (after 8 months of the 4-year design study): –Antineutrino rate (and 6 He figures) have reached the design values but no safety margin is yet provided. –Neutrino rate (and 18 Ne figures) are one order of magnitude below the desired performance. (*) M.Benedikt CERN-ISS 22/9/2005 (*) for a single ISOL target, BUT: multiple Eurisol targets, optimize collection time and add cooling, decrease losses due to incomplete stripping, relax duty cycle as a last resort. With present ideas, we could be off by only a factor of 2

SUPERBEAM BETABEAM  → e e →   Superbeam + betabeam 2 ways of testing CP, T and CPT : redundancy and check of systematics Furthermore: the small signal searched for with one beam is the bulk of events with the other beam: Better handle on detector response, strong added value 2 beams 1 detector

Beta beam optimization Signal = 1-ring mu-like Background = charged pions mimicking a muon (all pions except when absorbed by strong interaction)  Lower energies : less background pions, but muon Cerenkov threshold and bad energy resolution (Fermi momentum)  Higher energies : more background, but better muon efficiency and energy bins become possible  Optimization obtained through detailed detector response to signal and background Present optimization:  = 100, 100 (M. Mezzetto)

M.Mezzetto, NNN05

BETA BEAM PERFORMANCES, ALONE and WITH SUPERBEAM 0.2 o 0.5 o 1 o 2 o 4 o 1o1o 0.3 o 0.1 o M.Mezzetto TAUP05 Zaragoza For Frejus and HyperK detectors: 4.4 Mton.year (~10 year run, 440 kT fiducial mass) Frejus: full simulation of signal and backgrounds, based on SK software: results cross-checked using GLoBES others: GLoBES simulation Systematics on signal and background put at 2% on all projects for the same(10 y) running time NuFACT; decays/year, 2 x 50 kTdetectors at 3000 and 7500 km, detection threshold 4 GeV 0.5 o 0.2 o

Degeneracy/ambiguity issues(1) First, one should clearly distinguish between the discovery potential of an experiment (θ 13 # 0,  CP # 0 ) and the determination of these parameters, which might be plagued by ambiguities (a similar situation existed for many years for solar neutrinos : significant DEFICIT, many solutions : SMA, LMA, LOW, VAC…). Taking the highest value of θ 13 among different solutions as an indicator of sensitivity on θ 13, as is done for exemple by Lindner et al, is in my opinion misleading (a 2-dim sensitivity contour bringing much more information). With this method, any oscillation experiment would have no sensitivity on θ as long as  m 2 is unknown…

Degeneracy/ambiguity issues(2) T2K2 T.Schwetz, NuFact’05 Hep-ph/

 disappearance in superbeam may also help in solving degeneracies (Donini et al, hep-ph/ )

A possible schedule for a european lab. at Frejus Year Safety tunnel Excavation Lab cavity Excavation P.S Study detector PM R&DPMT production Det.preparation InstallationOutside lab. Non-acc.physics P-decay, SN Superbeam Construction Superbeam betabeam Beta beam Several modules: physics can start when the 1 st is ready Construction decision for cavity digging decision for SPL construction decision for EURISOL site

3 MeV test place ready Linac4 approval * “… in , to decide on the implementation of the Linac 4 and any increased R&D programme, depending on new funds made available and on a new HR policy” SPL approval * “ in , to review and redefine the strategy for CERN activities in the next decade in the light of the first results from LHC and of progress and results from the previous actions. “ RF tests in SM 18 of prototype structures* for Linac4 CDR 2 Planning … * Quotes from R. Aymar (Jan.2005) R. Garoby

High energy beta beams Many papers in the last 2 years have advocated using higher energy beta beams, with  ranging from 350 (refurbished SPS) to 1000 or higher (greenfield scenario). Higher energy offers several advantages, in particular to solve the mass hierarchy through matter effects. MOST STUDIES BASED ON THE HYPOTHESIS THAT THE DECAY RATE IN THE RING STAYS THE SAME, so that the event rate grows like  Caveats: phase space limitation => increase bunch length => more background: This needs to be studied carefully At higher energy, performances of water Cerenkov deteriorate (background increase). Possibility => change technology => smaller mass => less statistics (and loss for non-accelerator physics) Need a new accelerator, new site and a very big decay ring => cost issue These projects, if proven feasible and as performant as expected, will arrive much later, but would then be direct competitors to a neutrino factory. [ see P.Hernandez review talk at CERN ISS meeting ]

Cross-Sections D.Casper, NuFact’05

Signal and Backgrounds, WC 1-ring  -like sample 1-ring e-like sample D.Casper, NuFact’05

Conclusions of D.Casper on WC A very mature and powerful technology Backgrounds to low-medium energy super-beams or beta beams are fairly manageable –Depends on details of beam, baseline, etc. Energies above GeV create difficulties –May be mitigated by migration D.Casper, NuFact’05

First example (hep-ph/ ) J. Burguet-Castell et al Flux independent of  Overestimate duty cycle reduction 3 setups considered (with megaton detector) : I  = 120, L=130 km (optimal Frejus) II  =150, L=300 km (max,SPS at optimal distance) III g=350, L=730 km (S-SPS, Canfranc?)

Sensitivity to CP-Violation vs  : Sensitivity to θ 13 ≠0 vs  :

Comparison of setups From E.Couce, NuFact’05

2 nd example (Migliozzi, TAUP05)  = 350 (S-SPS), same flux as CERN- Frejus 40 kT at Gran Sasso 10 year run Performances similar to CERN-Frejus and no non-accelerator physics

3rd example (hep-ph/ ) (Terranova et al) g between 1500 and 2500 Wall detector counting muons at Gran Sasso

EC ions: monochromatic beams J.Bernabeu et al, hep-ph/ Candidates: 148 Dy T 1/2 = 3.1 mn EC/  = 96/4 Q EC = 2062 keV 150 Dy T 1/2 = 7.2 mn EC/  = 99.9/0.1 Q EC = 1397 keV 130 km Pending question: achievable flux ? Challenge: needs partially stripped ions Run at different  ’s, with or without betabeam: interesting potentialities Even at low flux, ideal beam to measure neutrino cross-sections in a near detector And also… intense monochromatic neutron beams about n/sec, useful for nuclear physics (cf NuPAC)

(very) low energy beta-beams(  ~7-14) C. Volpe, Journ. Phys. G30 (2004) L1 The idea To exploit the beta-beam concept to produce intense and pure low-energy neutrino beams boost 6 He Beta-beam B e C Design study within EURISOL Design of a small storage ring and possible construction at CERN (in old antiproton storage ring ?) Physics potential Neutrino-nucleus interaction studies for: -nuclear astrophysics -neutrino experiments (oscillations, neutrinoless double-beta decay) -nuclear structure

Conclusions(1) There are clearly 2 different strategies concerning beta-beams : A. The baseline scenario (g=100, L=130 km) with a megaton WC detector at Frejus, which makes full use of some happy coincidences: 1.CERN-Frejus is the right distance for the SPL superbeam 2.Digging a safety gallery at Frejus gives a good opprtunity to dig the new cavern ( ) 3.Beta-beam energy obtained from SPS fits also Cern-Frejus distance, and a strong synergy with Eurisol project has become evident and fruitful The betabeam + superbeam combination gives by far the best second generation project worldwide (superior to T2K2 or BNL/UNO) We can easily take the risk of a null result on oscillations, since such a detector has a very rich and fundamental program of its own (proton decay, supernova explosions) In this approach, the project is not meant to compete with neutrino factories, but to arrive sooner (~2017) and allow significant progress on oscillations….while bringing back neutrino physics to Europe

Conclusions (2) B. The higher energy betabeam, which clearly aims at competing with a neutrino factory. To achieve these goals, the baseline scenario has to be significantly improved to override present limitations. It means: 1.A new complex of accelerators (faster, higher acceptable fluxes) 2.A better (and bigger!) decay ring Such a project will anyhow need much more R&D, will arrive much later and will be also more costly. Non oscillation physics could also become a criterim to choose between betabeams and neutrino factories, in case 2 nd generation experiments have seen nothing Actually, these 2 strategies are not exclusive, we could have both in sequence. An alternative would be to skip the baseline betabeam and go directly to a high energy betabeam or a neutrino factory, but I personnally find this too risky.

Summary A megaton water Cerenkov detector,consisting of several modules, can be installed between 2010 and 2020 at Frejus It has a very rich non-accelerator program of its own: proton decay, SN explosions, atmospheric, solar, past SN neutrinos If CERN decides to build the SPL, it offers an excellent superbeam of 2 nd generation, with similar performances to the japanese project T2K2 and starting sooner (HK ready by ) If this SPL is the proton machine for EURISOL at CERN, it opens the way to beta beams, with modest added cost, offering better performances than the superbeam and a strong complementarity between the two beams Concerning neutrino physics, this project is not meant to compete with a neutrino factory, but to arrive sooner and make significant progress on the presently unknown parameters… and bring back physics to Europe. Higher energy beta beams correspond to a different strategy, as an alternative to neutrino factories. It gives a common interest to nuclear physicists and neutrino physicists to have the SPL at CERN

BACKUPS

Present acceler ator Replacement accelerator Improvement INTEREST FOR LHC upgrade physics beyond CNGS RIB beyond ISOLDE Physics with Kand  Linac2Linac4 50 → 160 MeV H + → H - +0 (if alone) PSB 2.2 GeV RCS* for HEP 1.4 → 2.2 GeV 10 → 250 kW +0 (if alone)+ 2.2 GeV/mMW RCS* 1.4 → 2.2 GeV 0.01 → 4 MW + ++ (super-beam,  -beam ?, factory) + (too short beam pulse) 0 (if alone) 2.2 GeV/50 Hz SPL* 1.4 → 2.2 GeV 0.01 → 4 MW (super-beam,  -beam, factory) +++0 (if alone) PS SC PS*/** for HEP 26 → 50 GeV Intensity x (if alone)0+ 5 Hz RCS*/** 26 → 50 GeV 0.1 → 4 MW ++ ( factory) 0+++ SPS 1 TeV SC SPS*/** 0.45 → 1 TeV Intensity x 2 +++?0 * with brightness x2** need new injector(s) HIP WG: long term alternatives (R.Garoby)

● Future neutrino facilities offer great promise for fundamental discoveries (such as CP violation) in neutrino physics, and a post-LHC construction window may exist for a facility to be sited at CERN. ● CERN should arrange a budget and personnel to enhance its participation in further developing the physics case and the technologies necessary for the realization of such facilities. This would allow CERN to play a significant role in such projects wherever they are sited. ● A high-power proton driver is a main building block of future projects, and is therefore required. ● Alone, a direct superbeam from a 2.2 GeV SPL does not appear to be the most attractive option for a future CERN neutrino experiment as it does not produce a significant advance on T2K. ● We welcome the effort, partly funded by the EU, concerned with the conceptual design of a β-beam. At the same time CERN should support the European neutrino factory initiative in its conceptual design. SPSC (Villars) recommendations CERN Context [3/6] ->

SPL & PDAC [1/3] Ion speciesH-H- Kinetic energy3.5GeV Mean current during the pulse 40 (30 ?)mA Mean beam power4MW Pulse repetition rate50Hz Pulse duration0.57 (0.76 ?)ms Bunch frequency352.2MHz Duty cycle during the pulse62 (5/8)% rms transverse emittances0.4  mm mrad Longitudinal rms emittance0.3  deg MeV SPL (CDR2) characteristics R. Garoby

[Extrapolation from PDAC based on the SPL CDR-1] Mean beam power4MW Kinetic energy3.5GeV Pulse repetition rate50Hz Pulse duration1.66 ss RF frequency44.02MHz Number of bunches (buckets)68 (73) Number of protons per pulse (per bunch) 1.43 E14 (2.1 E12) Number of turns for injection345 rms normalized transverse emittances 50  mm mrad Longitudinal emittance0.2eVs SPL & PDAC [3/3] SPL (CDR2) + PDAC characteristics R. Garoby

Linac4 Ion speciesH-H- Kinetic energy160MeV Mean current during the pulse40mA Pulse duration  0.4 ms Number of particles per pulse  Pulse repetition rate  2 Hz Beam power  5 kW Bunch frequency352.2MHz LINAC4 Characteristics - Location: PS South Hall and extension - Technical Design report in preparation (publication mid-2006) - Possible planning: - authorization: December construction: setting-up & commissioning: availability for physics (replacing Linac2): January 2011 R. Garoby

3 MeV test place (Linac4 front-end) H- source IPHI RFQ Chopper line Measurement line LEP Klystron - Ongoing project supported by HIPPI (FP6) + IPHI (CEA+IN2P3+CERN) - Location: PS South Hall extension (future location in Linac4) -Test with beam : R. Garoby

M.Zisman, NuFact’05

EC: A monochromatic neutrino beam M.Lindroos, NuFact’05

150 Dy Partly stripped ions: The loss due to stripping smaller than 5% per minute in the decay ring Possible to produce Dy atoms/second (1+) with 50 microAmps proton beam with existing technology (TRIUMF) An annual rate of decays along one straight section seems as a realistic target value for a design study Beyond EURISOL DS: Who will do the design? Is 150 Dy the best isotope? M.Lindroos, NuFact’05

Possible  + emitters ( e )

Possible  - emitters ( e )

some personal thoughts main aim in the coming years:  13,  CP, sign (  m 2 ) Europe is absent from first round of superbeams Can Europe regain strong position for 2 nd round? I think so if we are able to start between 2015 and 2020 super+betabeams superior to T2K phase2 and arrive before ( if we stick to “baseline” scenario ) so that Europe can attract a worldwide collaboration. If detector and beams wait for the other to take first step, nothing will happen The alternative is to skip second round and prepare 3 rd round (neutrino factory or higher energy beams) I personally find this risky for Europe We should try to find again a consensual road map for Europe

EXCLUDED BY CHOOZ Θ 13 degrees YEAR Minos, Icarus, Opera Double Chooz T2K1, NovA SB  B T2K 2

Contamination vs. Smearing 1-ring  -like sample 1-ring e-like sample

PMT size cost Diameter 20“ 12“ projected area cm² QE(typ) % CE % SxQExCE Cost € Cost/ cm² per useful PE U = cost /( cm² xQExCE) € → 1 20-inch PMT replaced by 2 12-inch PMTs :. same number of PE, better SE, better timing, better granularity ! Saving ( per 20-inch) = 900 € - electronics extra cost Cf Photonis, NNN05

New pump capacity needed? Delivery over 6 years 300 working days/year –20“ tube 200,000/6/300 => 112 good tubes / yield 0.7 = 160 starts/day (1 start/pump/day) => 160 pumps ( € 30M or so) –12“ tube 400,000/6/300 => 222 good tubes / yield 0.7 = 320 starts/day. A multi-array computerised pump at Photonis handles 20 starts/day => 16 pumps ( € 5M or so) Cf Photonis, NNN05