1. Potentiality of a (very) high-   -Beam complex Pasquale Migliozzi INFN – Napoli

2. !!!WARNING!!! The physics potential of the BB of any  has been carefully evaluated by several groups and discussed in this talk and in the previous ones However, solid feasibility studies from the accelerator side are still missing, although some interesting ideas are on the market The BB are “in principle” a great idea, but we need more studies (regardless of the  option) to endorse their “practical” realization

3. Future neutrino oscillation exps E p (GeV) Power (MW) Beam 〈 E 〉 (GeV) L (km) M det (kt)  CC (/yr) e @peak K2K120.005WB1.325022.5~50~1% MINOS(LE)1200.4WB3.57305.4~2,5001.2% CNGS4000.3WB18732~2~5,0000.8% T2K-I500.75OA0.729522.5~3,0000.2% NO A1200.4OA~2810?50~4,6000.3% C2GT4000.3OA0.8~12001,000?~5,0000.2% T2K-II504OA0.7295~500~360,0000.2% NO A+PD1202OA~2810?50?~23,0000.3% BNL-Hs281 WB/OA ~12540~500~13,000 SPL-Frejus2.24WB0.32130~500~18,0000.4% FeHo 8/120 “4” WB/OA 1~31290~500~50,000 Running, constructing or approved experiments

4. Possible scenarios after first results of the planned experiments and implications  13 is so small (< 3°, sin 2 2  13 ≤ 0.01) that all give null result  We need a “cheap” experiment to probe sin 2 2  13 values down to O(0.001 - 0.0001)  13 is larger than 3° (sin 2 2  13 ≥ 0.01)  We need an experiment (or more than one) to ¤ Measure  13 more precisely ¤ Discover  (if not done yet) or precisely measure it ¤ Measure the sign of  m 2 13 ¤ Measure  23 (is it ≠45°?) NB Independently of the scenario the worsening of the experimental sensitivity due to the eightfold degeneracy has to be taken into account

5. Possible strategy (detector side) We think that one should figure out the best setup depending on the results of phase I experiments Null result for  13 : are we ready to risk several billions $? NO, it is better to try a cheap, although not the ultimate, approach to two important parameters like  13 and  Observation of a non vanishing  13 : are we ready to invest several billions $? YES, since there is the possibility to fully measure the PMNS mixing matrix

6. The  -beam (BB) role The BB was born in 2001 when P. Zucchelli put forward the idea to produce pure (anti-) e beam from the decay of radioactive ions Originally the BB was thought as a low (  ~100) energy neutrino beam and its performance studied in combination with a Super-Beam (SB), by assuming a 130 km baseline and 1 Mton detector located at Frejus (M. Mezzetto et al.) However, very recently (december 2003) the possibility of medium/(very) high energy BB was put forward (see hep- ph/0312068) What is the impact of the BB (low (see S. Rigolin talk for details), medium, high, very-high  ) in the future of neutrino oscillation experiments?

7. Comparison of low  BB with some of the future projects Low  BB+SB

8. Why high  BB? statistics increases linearly with E (cross section)  increase rates (very important for anti-neutrinos) longer baseline  enhance matter effects  possibility to measure the sign of  m 2 13 increase the energy  easier to measure the spectral information in the oscillation signal  important to reduce the intrinsic degeneracies Atmospheric background becomes negligible (this is a major background source in the low energy option)  the bunching of the ions is not more a crucial issue

9. Which  ’s? Use a refurbished SPS with super-conducting magnets to accelerate ions Maximum  ~600 Use the LHC to accelerate ions Up to  ~2488 for 6 He and 4158 for 18 Ne In the US (see talk of S.Geer and APS meeting @ Snowmass, 28-30 Jun 04):

10. How to exploit high  BB? Phase I exps give null result See hep-ph/0405081 for a cheap and extremely sensitive to  13 experiment Phase I discover  13 See Nucl.Phys.B695:217-240,2004 for possible setups to search for  New ideas

11. Signal: an excess of horizontal muons in coincidence with the beam spill (possible thanks to the BB flavour composition) Number of unoscillated events: increase linearly with E Range of muons: increase linearly with E as well. The effective volume of rock contributing to the statistics increase linearly with E We gain a quadratic increase of the sensitivity if we increase  and we reduce the detector cost by order of magnitudes! We loose the possibility to fully reconstruct the events F. Terranova, A. Marotta, M. Spinetti, P.M. hep-ph/0405081 The cost of the detector increase with the surface and not with the volume A proposal for a cheap experiment

12. Schematic view of the detector Rock e   Instrumented surface: 15x15 m 2 (one LNGS Hall) Thickness: at least 8 I (1.5 m) of iron for a good  /  separation Iron detector interleaved with active trackers (about 3kton)  H

13. A possible scenario: BB from CERN to Gran Sasso A cavern already exists at GS, but Too small to host 40 kton WC or LAr detectors On peak exp requires E ~ 1-2 GeV (  = 350/580)  too small to efficiently exploit iron detectors What happens if we consider  > 1000 (i.e. off-peak experiment)? The oscillation probability decreases as  -2 The flux increases as  2 The cross-section and the effective rock volume increase both as  Matter effects cancel out at leading order even if the baseline is large  We recover the quadratic increase of sensitivity but we test now CP-even terms and no matter effects

14. Beam assumptions 1.1x10 18 decays per year of 18 Ne 2.9x10 18 decays per year of 6 He Applied cuts 2 GeV energy cut in a 20° cone 100 % oscillated events/year: 9.3x10 4 ( e @  =2500) 2.0x10 4 (anti e @  =1500) 7.9x10 5 ( e @  =4158) 2.1x10 5 (anti e @  =2488) Event rate

15. test sin 2 2  13 values down to 10 -3 -10 -4 !!! Sensitivity of a “massless” detector located 730 km from a (very)high-  BB

16. Comparison of very-high  BB with some of future projects Low  BB+SB BB very high  In case of null result very difficult to build new facilities!

17. Two setups studied for the medium/high  options Medium (350) and high (1500)  for medium (730 km) and far (3000 km) baselines Water detector (UNO) like; 1 Mton mass. Includes full simulation of efficiencies and backgrounds (only statistical study for high gamma option) Running time 10 years Full analysis (including the eightfold degeneracy, all systematics on cross-sections, detector, beam, performance at small  13, etc.) still to be done

18. Results J.Burguet-Castell et al., Nucl.Phys.B695:217-240,2004 Baseline option (Frejus) 40 Kton/y WC @ 730 km  =350 ( 6 He) / 580 ( 18 Ne) 4 Mton/y WC detector @ 3000 km  =1550 ( 6 He) / 2500 ( 18 Ne) 4 Mton/y WC @ 730 km  =350 ( 6 He) / 580 ( 18 Ne) L=732 km Uno like detector L=3000 km Uno like detector L=732 km SK like detector L=130 km UNO like detector 99% CL

19. Comments The idea of medium/(very) high-  BB is very appealing Whatever  (medium, high, very-high) we consider its performance is better than the low one The medium scenario has been put forward in Nucl.Phys.B695:217- 240,2004 to measure  13,  and the sign of  m 2 13, but more studies are needed to fully exploit its potential (i.e. the  23 ambiguity) However, we think this is not the optimal solution It foresees the construction of a 1 Mton detector! There are no place in the world able to host it It is very expensive, so to risky to build if phase I exps give null results The optimal solution is the very-high  scenario In case of null result of phase I exps it allows a cheap investigation of very small values of sin 2 2  13 (see hep-ph/0405081) In case of positive result of phase I exps it allows a complete study of the PMNS matrix through different channels, see next slides for details On top of that it makes possible the usage of magnetized calorimeters which are smaller (40 kton -> about 10 4 m 3 ) than WC detectors (1 Mton -> about 10 6 m 3 )  cheaper (easier) civil engineer costs

20. Preliminary studies/ideas on how to use the very-high  BB A. Donini, PM, S. Rigolin, …

21. BB vs NuFact spectra Ne He NuFact

22. Expected rates (1ktonx1year) NuFactBB e anti-  e anti- e L=730 km296x10 3 176x10 3 --- L=730 km  =2500/1500 --- 17x10 3 5.5x10 3 L=730 km  =4158/2488 --- 77x10 3 25x10 3 L=3000 km18x10 3 11x10 3 --- L=3000 km  =2500/1500 --- 1.0x10 3 0.3x10 3 L=3000 km  =4158/2488 --- 4.6x10 3 1.5x10 3 NB There is less than a factor 10 difference in the #evts BB allows simultaneous run with and anti-, while NuFact does not

23. Potentiality of a very-high  BB Simultaneous search for e   (golden) and e   (silver) channels This combination is highly efficient in removing the intrinsic and the sign degeneracy (see A.Donini, D.Meloni, P.Migliozzi Nucl.Phys.B646:321-349,2002) Simultaneous search for and anti- channels (i.e. 1 year BB  2 years NuFact) Detectors: 40kton magnetized iron detector (MID) at 3000km; ≥5kton ECC detector at 730km The physics potential of this setup is currently under study as well as its comparison with a NuFact

24. Very preliminary results at a high-  BB with golden plus silver channels Octan clone 68%, 90%, 99% CL MID MID+ECC

25. Parameter extraction in presence of signal (II) with a low  BB plus a SB Continuous line: intrinsic degeneracy Dashed line: sign ambiguity Dot-dashed line: octant ambiguity Dotted line: mixed ambiguity

26. Conclusion Whatever  (medium, high, very-high) BB we consider its performance is better than the low one The optimal solution is the very-high  scenario In case of null result of phase I exps it allows a cheap investigation of very small values of sin 2 2  13 (see hep-ph/0405081) In case of positive result of phase I exps it allows a complete study of the PMNS matrix through different channels, see next slides for details On top of that it makes possible the usage of magnetized calorimeters which are smaller (40 kton -> 10 4 m 3 ) than WC detectors (1 Mton -> 10 6 m 3 )  cheaper (easier) civil engineer costs The potentiality of a very-high BB are under study including the eightfold degeneracy and both the golden and the silver channels. Some preliminary results look interesting MORE STUDIES FROM THE ACCELERATOR SIDE ARE NEEDED INDEPENDENTLY OF THE  OPTION

28. Is the proposed low energy setup (CERN to Frejus) the optimal one? NO! Why? In spite of the large detector mass, the performance is limited by small rates (due to small cross-sections) and by the eightfold degeneracy By using the BB or the SB alone is not possible to solve any of the degeneracies, although for large enough  13 a first estimate of the two continuous parameters  13 and  can be attempted Neither the sign of  m 2 13 nor the absolute value of  23 can be determined The combination of a BB and a SB (as proposed in the CERN scenario) is not a real synergy (i.e. NO degeneracy is solved). Indeed, it only determines an increase of statistics for both and anti- The sensitivity of a BB is comparable with the one of other proposed future projects Can higher  BB be more suitable?

29. Beam related background: Pion punch-through deep hadron plug Early  /K decays in flight energy cut (charge id) Charm backgroundonly for the highest , energy cut (charge id) Beam unrelated background: Atm. neutrinosenergy cut (beam timing) Cosmicsangular cut (huge slant depth near the horizon) Beam unrelated background is so small that we can release by two order of magnitudes the request on the bunch length of the BB An enormous technical simplification Can we handle background with such a rough detector?

30. Event rates vs  (L = 732 km)

31. Parameter extraction in presence of signal (I) with a low  BB plus a SB SB BB+SB BB Continuous line: intrinsic degeneracy Dashed line: sign ambiguity Dot-dashed line: octant ambiguity Dotted line: mixed ambiguity NB The black dots show the theoretical clone location computed following Ref. JHEP 0406:011,2004