1. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 1 A Super-Neutrino Beam From BNL to Homestake Steve Kahn http://pubweb.bnl.gov/people/kahn/talks/bnl2homestake.pdf

2. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 2 Staging to a Neutrino Factory Two feasibility studies for a Neutrino Factory have been concluded. –These studies indicate a cost of 2-2.5 B$. This does not include contingency and overhead. This kind of money may not be available in the current climate –They indicate an optimistic turn-on date of 2012. We might like to do some physics before that. A staged approach to building a Neutrino Factory maybe desirable. –First Phase: Upgrade AGS to create a 1 MW Proton Driver and target station. –Second Phase: Build phase rotation and part of cooling system. –Third Phase: Build a pre-acceleration Linac to raise beam momentum to 2.5 GeV/c –Fourth Phase: Complete the Neutrino Factory. –Fifth Phase: Upgrade to entry-level Higgs Factory Muon Collider. Each phase can support a physics program.

3. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 3 First Phase Super Neutrino Beam Upgrade AGS to 1MW Proton Driver: –Both BNL and JHF have eventual plans for their proton drivers to be upgraded to 4 MW. Build Solenoid Capture System: –20 T Magnet surrounding target. Solenoid field falls off to 1.6 T in 20 m. –This magnet focuses both  + and  . Beam will have both and –A solenoid is more robust than a horn magnet in a high radiation. A horn may not function in the 4 MW environment. A solenoid will have a longer lifetime since it is not pulsed.

4. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 4 Types of Capture/Focus Systems Considered Traditional Horn Focus System –Uses toroidal magnetic field. –Focuses efficiently B   p  –Conductor necessary along access. Concern for radiation damage. Cannot be superconducting. –Pulsed horn may have trouble surviving ~10 9 cycles that a 1-4 MW system might require. Solenoid Capture System similar to that used by Neutrino Factory Solenoid Horn System

5. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 5 Simulations to Calculate Fluxes Model Solenoid/Horn Magnet in GEANT. –Use Geant/Fluka option for the particle production model. –Use 30 cm Hg target ( 2 interaction lengths.) No target inclination. –We want the high momentum component of the pions. –Re-absorption of the pions is not a problem. –Solenoid Field profile on axis is B(z)=B max /(1+a z) Independent parameters are B max, B min and the solenoid length, L. –Horn Field is assumed to be a toroid. –Pions and Kaons are tracked through the field and allowed to decay. –Fluxes are tallied at detector positions. The following plots show  flux and e /  flux ratios.

6. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 6 Solenoid Capture Sketch of solenoid arrangement for Neutrino Factory If only and not is desired, then a dipole magnet could be inserted between adjacent solenoids above. Inserting a dipole also gives control over the mean energy of the neutrino beam. Since and events can be separated with a modest magnetic field in the detector, it will be desirable to collect both signs of at the same time.

7. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 7 Captured Pion Distributions P T distribution of   P T =225 MeV/c corresponding to 7.5 cm radius of solenoid 66% of  are lost since they have P T >225 MeV/c P T, GeV/c Decay Length of Pions =7 m L, cm P  > 2 GeV/c = 50 m A 15 cm radius of the solenoid would capture 67% of the  +

8. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 8 Rate and e /  as a function of Decay Tunnel Length

9. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 9 Comparison of Horn and Solenoid Focused Beams The Figure shows the spectra at 0º at 1 km from the target. –Solenoid Focused Beam. –Two Horned Focused Beam designed for E889. –So-called Perfect Focused beam where every particle leaving the target goes in the forward direction. The perfect beam is not attainable. It is used to evaluate efficiencies. A solenoid focused beam selects a lower energy neutrino spectrum than the horn beam. –This may be preferable for CP violation physics

10. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 10 Horn and Solenoid Comparison (cont.) This figure shows a similar comparison of the 1 km spectra at 1.25º off axis. –The off axis beam is narrower and lower energy. Also a curve with the flux plus 1/3 the anti- flux is shown in red. –Both signs of are focused by a solenoid capture magnet. A detector with a magnetic field will be able to separate the charge current and anti-.

11. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 11 Flux Seen at Off-Axis Angles We desire to have Low Energy beam. We also desire to have a narrow band beam. I have chosen 1.5º off-axis for the calculations. Angle Solenoid  QE evtsSolenoid  QE EventsHorn  QE evtsHorn  evts 0 4.21  10 6 9.86  10 5 1.38  10 7 1.20  10 5 ¼ 4.11  10 6 9.56  10 5 1.32  10 7 1.06  10 5 ½ 4.10  10 6 9.46  10 5 1.18  10 7 1.05  10 5 1 3.80  10 6 8.83  10 5 8.69  10 6 8.27  10 4 1.5 3.36  10 6 7.89  10 5 5.98  10 6 7.53  10 4 2 2.88  10 6 6.80  10 5 4.01  10 6 4.76  10 4 3 1.94  10 6 4.64  10 5 1.93  10 6 3.31  10 4 4 1.31  10 6 3.20  10 5 1.02  10 6 2.35  10 4

12. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 12 e /  Ratio The figure shows the e flux spectrum for the solenoid focused and horn beams. The horn focused beam has a higher energy e spectrum that is dominated by K  o e e The solenoid channel is effective in capturing and holding  and . –The e spectrum from the solenoid system has a large contribution at low energy from   e e. –The allowed decay path can be varied to reduce the e /  ratio at the cost of reducing the  rate. We expect the e /  ratio to be ~1%

13. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 13 Running the AGS with 12 GeV Protons We could run the AGS with a lower energy proton beam. If we keep the same machine power level we would run at a 5 Hz repetition rate. –This would work for a conventional beam since we are not concerned with merging bunches. Figure shows Perfect Beam for 12 and 24 GeV incident protons. –12 GeV profile is multiplied by 2 for the higher repetition rate. 24 GeV protons 12 GeV Protons Perfect Beam

14. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 14 12 GeV Protons (cont.) On Axis 1.25 degrees off axis

15. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 15 Detector Choices The far detector would be placed 350 km from BNL (near Ithica, NY). –There are salt mines in this area. One could go deep underground if necessary. If a massive detector were built at say 2540 km from BNL (at Homestake), this would permit the determination of the CP violation sign using mass effect. Two possible detector technologies that can be considered are Liquid Ar and Water Cherenkov. –We are considering Liquid Ar TPC similar to Icarus. The far detector would have 50 ktons fiducial volume (65 ktons total.) Provides good electron and  o detection. The detector will sit between dipole coils to provide a field to determine the lepton charge. This technology is expensive and may not be practical.

16. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 16 Detector Choices (cont.) –Water Cherenkov technology similar to Super-K may be the only reasonable way to achieve a Megaton detector. Charge determination using a magnetic field may not be possible with this type of detector. The neutrino source must sign select the. A close-in 1 kton detectors at 1 km and/or 3 km would be needed. –1 km detector gives beam alignment and high statistics for detector performance. –3 km detector is far enough away that source is a point.

17. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 17 Detectors Are Placed 1.5 o Off Beam Axis Placing detectors at a fixed angle off axis provides a similar E profile at all distances. It also provides a lower E distribution than on axis.  from  decays are captured by long solenoid channel. They provide low E enhancement. Integrated flux at each detector: –Units are /m 2 /POT

18. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 18 Neutrino Oscillation Physics The experiment would look at the following channels: –  disappearance -- primarily    oscillations. Sensitive to  m 23 2 and  23 Examine ratio of n  p (QE) at 350 km detector to 3 km detector as a function of E. –  N  o N events These events are insensitive to oscillation state of Can be used for normalization. – e appearance (continued on next transparency) Ratio of QE D350/D3 0.01 0.1 1 10 01234 Enu, GeV

19. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 19 There are several contributions to P(   e ): –Solar Term: P solar =sin 2 2  12 cos 2  13 cos 2  23 sin 2 (  m 2 sol L/4E) This term is very small. –Tau Term: P  =sin 2 2  13 sin 2  23 sin 2 (  m 2 atm L/4E) This is the dominant term. This term is sensitive to  13 and would allow us to measure it with the 1 MW proton driver. –Terms involving the CP phase  : There are both CP conserving and violating terms involving . The CP violating term can be measured as This asymmetry is larger at lower E. This could be ~25% of the total appearance signal at the optimum E The 4 MW proton driver would be necessary for this asymmetry e Appearance Channel

20. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 20 Event Estimates Without Oscillations Below is shown event estimates expected from a solenoid capture system –The near detectors are 1 kton and the far detector is 50 kton. –The source is a 1 MW proton driver. –The experiment is run for 5 Snowmass years. This is the running period used in the JHF-Kamioka neutrino proposal. –These are obtained by integrating the flux with the appropriate cross sections. Estimates with a 4 MW proton driver source would be four times larger.

21. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 21 Determination of  m 2 23 Consider a scenario where –  m 2 12 =5  10  5 eV 2 –  23 =  /4 –  m 2 31 =0.0035 eV 2 (unknown) –Sin 2 2  13 =0.01 (unknown) –This is the Barger, Marfatia, and Whisnant point Ib. =0.8 GeV is not optimum since I don’t know the true value in advance. I can determine  m 2 23 from 1.27  m 2 23 L/E 0 =  /2 Where E 0 is the corresponding null point Note that these figures ignore the effect of Fermi motion in the target nuclei. –This would smear the distinct 3  /2 minimum.  /2

22. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 22  m 23 2 with Errors The near detector at 3 km and the far detector is at 350 km. The plot is made comparing quasi-elastic events only. –E is well measured for these events. No corrections are necessary. This should produce a solid measurement of  m 23 2. Same plot as previously shown.

23. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 23 Barger, Marfatia and Whisnant Table

24. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 24 Oscillation Signal For comparison we have 28% of the flux used in Barger et al. We do not use a necessarily optimum L/E fixed configuration for all cases since the true oscillation parameters are not known in advance. We use the actual flux distribution, not a monochromatic beam (as used in Barger et al.). The following transparencies will show Quasi-Elastic event numbers for Solenoid and Horn capture systems. They assume: 1 MW Proton Driver 50 kton detector at 350 km with charge determination (Liquid Ar) 5  10 7 second running period. The size of the e appearance signal will give a  13 measurement since  m 13 2   m 23 2 is measured independently by the  disappearance.

25. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 25 Going to Homestake Most of the transparencies shown are based on Snowmass calculations for a far detector placed near Cornell. We can scale the number of events from these calculations to estimate signals that would be seen at Homestake. –Scale with detector mass –Scale with 1/r 2. Increasing the Proton Driver Power to 4 MW would be very advantageous to a detector at Homestake. With the eventual upgrade to a neutrino factory, the Homestake detector would have a significant event rate. 0.38 if 1 MW

26. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 26 Table 1: Oscillation Signal:  Consider  m 2 12 =5  10 -5 eV 2,  23 =  /4 and sin 2 2  13 =0.01 · Using a 1 MW proton driver and a 50 kton detector 350 kilometers away. · Experiment running for 5  10 7 seconds. · Solenoid capture system with e /  flux ratio=1.9 %  m 2 13 eV 2  e signal e backgroundAnti  Anti e signalAnti e BG No Oscillation15539 4553455 150 0.002506576455109618.5150 0.0035528470455128316.2150 0.005772255455176213.1150 Solenoid Capture System with 230 m Decay Tunnel  e Signal e BG  e signal e BG Ignores e BG oscillations Significance: e signal: 3.3 s.d. e signal: 1.3 s.d.

27. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 27 Table 1: Oscillation Signal:  Consider  m 2 12 =5  10 -5 eV 2,  23 =  /4 and sin 2 2  13 =0.01 · Using a 1 MW proton driver and a 50 kton detector 350 kilometers away. · Experiment running for 5  10 7 seconds. · Solenoid capture system with e /  flux ratio=1.1 %  m 2 13 eV 2  e signal e backgroundAnti  Anti e signalAnti e BG No Oscillation10582 2492560 47 0.00236005824987814.447 0.0035428250249109012.347 0.005528343249130310.647  e signal e BG  e signal e BG Ignores e BG oscillation Significance: e signal: 3.2 s.d. e signal: 1.8 s.d. Solenoid Capture System with 100 m Decay Tunnel

28. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 28 Table 1: Oscillation Signal:  Consider  m 2 12 =5  10 -5 eV 2,  23 =  /4 and sin 2 2  13 =0.01 · Using a 1 MW proton driver and a 50 kton detector 350 kilometers away. · Experiment running for 5  10 7 seconds. · Horn capture system with e /  flux ratio=1.08 %  m 2 13 eV 2  e signal e backgroundAnti  Anti e signalAnti e BG No Oscillation21645 272228 5.4 0.00283178327211515.4 0.00355165952728415.4 0.0059966692729015.4 Horn Beam 200 m Decay Tunnel  e Signal e BG  e signal e BG Ignores e BG oscillations Significance: e signal: 5.8 s.d. E889 Horn Design

29. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 29 Table 1: Oscillation Signal:  Consider  m 2 12 =5  10 -5 eV 2,  23 =  /4 and sin 2 2  13 =0.01 · Using a 1 MW proton driver and a 50 kton detector 350 kilometers away. · Experiment running for 5  10 7 seconds. · Horn capture system with e /  flux ratio=1.04 %  m 2 13 eV 2  e signal e backgroundAnti  Anti e signalAnti e BG No Oscillation691 194354 65 0.002506419157619.765 0.00353054.719101817.865 0.0053314.519207413.965 Anti Horn Beam 200 m Decay Tunnel  e Signal e BG  e signal e BG Ignores e BG oscillations Significance: e signal: 2.2 s.d. E889 Horn Design

30. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 30 Cosmic Ray Background This table shows the cosmic ray rates for a detector placed on the surface. –The rate reduction factors come from the E889 proposal. –The events shown are scaled to the 350 km detector mass and 5 Snowmass year running period. –The neutron background could be significantly reduced by going 50-100 m underground if it is a problem. Placing the detector deep below ground in a mine would be more advantageous for proton decay experiments. –The residual cosmic ray background could be reduced to ~0.002 events at ~600 m below ground.

31. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 31 Backgrounds to e Appearance Signal The largest backgrounds to the   e signal are expected to be: – e contamination in the beam. This was ~1% e /  flux ratio in the capture configuration that was used in this study. This yields a ~2% in the event ratio. –Neutral Current  o N events where the  o are misidentified as an electron. If a  from the  o converts close to the vertex (Dalitz decay) and is asymmetric. The magnetic field and dE/dx will be helpful in reducing this background. Simulation study is necessary. I estimate (guess) that this background is ~0.001 of the  o N signal.

32. February 19, 2001Neutrino Beams from BNL to Homestake Stephen Kahn Page 32 Conclusions A high intensity neutrino super beam maybe an extremely effective way to study neutrino oscillations. –In particular the 4 MW version of the super beam may be the only way to observe CP violation in neutrino oscillations without a Muon Ring Neutrino Factory. This experiment is directly competitive with the JHF- Kamioka neutrino project. –Do we need two such projects? I will not answer that!