Physics with Very Intense Muon Beams

1. Physics with Very Intense Muon Beams W. Molzon U.S. Neutrino Factory and Muon Collider Collaboration Meeting UCLA January 29 2007 Outline: General considerations Physics motivation for lepton flavor violation experiments Experimental status of LFV Prospects for improvement Concept of m-N→e-N experiment Physics of the conversion process Background sources Detector requirements MECO muon beam and detector concept Possible m-N→e-N experiment at Fermilab Possible m-N→e-N experiment at JPARC

2. General Considerations • Physics goals • Lepton flavor violation (m-N  e-N, m+ e+g, m+ e+e+e-) • Precision electroweak (g-2, decay spectrum, lifetime, capture on p, edm) • Very intense beams are useful for experiments in which the majority of the decays do not cause rate in detectors and no accidental physics backgrounds • Cooling is generally useful, usually a tradeoff with available intensity, cost, etc. • Muon beam properties • Charge – negative for conversion (use Coulomb bound muons) positive for others. Positive muons produced using stopping pions (surface muon beams). • Energy – many stopping experiments use low energy beams, some experiments require specific energies • Energy spread – generally, reduced energy spread is beneficial • Time structure – depends on experiment, highly pulsed or DC generally preferred • Intensity – in most cases limited by instrumental or background effects William Molzon, UC Irvine Particle Physics with Intense Muon Beams

3. Existing and Possible Muon Sources • Available muon beams at TRIUMF(500MeV, 0.3MW) and PSI (590MeV, 1MW) • DC beams with intensity of order 108 per second, higher for m+ • Limitations of low energy machines • Muons per watt of beam power low due to low pion production cross section at low energy • Fewer negative muons • Fewer options to make pulsed beam • More difficulty with beam power on target (wrt higher energy accelerators) • Potential muon beam sources • BNL proton synchrotron • few to 24 GeV proton beam • low frequency RF for acceleration – relatively easy pulsing • available slow extraction • beam power 20-50 kW at 8 GeV • JPARC • Few to 40 (50) GeV proton beam • Relatively low frequency RF • Slow extraction being developed • Beam power 1 MW at 50 GeV • Fermilab • 8 to 120 GeV • No existing slow spill • Few 10s of kW at 8 GeV, few 100 kW at 120 GeV William Molzon, UC Irvine Particle Physics with Intense Muon Beams

4. Limitations: Detector Rates, Rate Induced Physics Backgrounds For LFV experiments, stop muons and look at decay or conversion processes • Detector rates from Michel decays (menn) decays • Detector rates from other beam particles, muon capture processes, etc • For most processes, physics backgrounds dominated by accidentals at useful rates • Example: m+ e+g sensitivity goal 10-14 running time 107 s detection efficiency 0.1 macro duty cycle 1 stop rate 108 background dominated by accidental coincidences of Michel positron, photon from radiative decay or positron annihilation in flight • Without some care, detector rates are too high • Example: muon conversion on a nucleus m-N  e-N sensitivity goal 10-17running time 107 s detection efficiency 0.2 macro duty cycle 0.5 stop rate 1011decay rate of 1011 Hz, instantaneous intensity higher with pulsed beameven higher fluxes from neutrons, photons from muon capture • Muon conversion experiment is unique in ability to use very high stopping rates William Molzon, UC Irvine Particle Physics with Intense Muon Beams

5. What Will Observation of m→eg or-Ne-N Teach Us?  e  W m- d L d e- lmd led Discovery of -N  e-Nor a similar charged lepton flavor violating (LFV) process will be unambiguous evidence for physics beyond the Standard Model. This process is completely free of background from Standard Model processes. • For non-degenerate neutrino masses, n oscillations can occur. Discovery of neutrino oscillations required changing the Standard Model to include massive . • Charged LFV processes occur through intermediate states with n mixing. Small n mass differences and mixing angles  expected rate is well below what is experimentally accessible:rate is proportional to [D(Mn)2/(MW)2]2 • Current limits on LFV in charged sector provides severe constraints on models for physics beyond the Standard Model • Charged LFV processes occurin nearly all scenarios for physics beyond the SM, in many scenarios at a level that current generation experiments could detect. • Effective mass reach of sensitivesearches is enormous, well beyondthat accessible with direct searches. Z,N -N  e-Nmediated by leptoquarks  William Molzon, UC Irvine Particle Physics with Intense Muon Beams

6. Sensitivity to Different Physics Processes Supersymmetry Compositeness Predictions at 10-15 Second Higgs doublet Heavy Neutrinos Heavy Z’, Anomalous Z coupling Leptoquarks After W. Marciano William Molzon, UC Irvine Particle Physics with Intense Muon Beams

7. Supersymmetry Predictions for LFV Processes 10 -11 10 -10 10 -13 10 -12 10 -15 10 -14 10 -17 10 -16 10 -18 10 -19 10 -21 10 -20 • From Hall and Barbieri Large t quark Yukawa couplingsimply observable levels of LFV insupersymmetric grand unified models • Extent of lepton flavor violation in grand unified supersymmetry related to quark mixing • Original ideas extended by Hisano, et al. Current MEGA bound Current SINDRUM2 bound B(  e g) Re PSI m→eg single event sensitivity mN→eN single event sensitivity goal 100 200 300 100 200 300 William Molzon, UC Irvine Particle Physics with Intense Muon Beams

8. History of Lepton Flavor Violation Searches - N  e-N +  e+ +  e+ e+ e- K0 +e-K+ + +e- 1 10-2 10-4 10-6 Branching Fraction Upper Limit MEGA 10-8 10-10 SINDRUM2 10-12 MEG goal 10-14 m-N→e-N goal 10-16 1940 1950 1960 1970 1980 1990 2000 2010 William Molzon, UC Irvine Particle Physics with Intense Muon Beams

9. PSI-MEG m+e+g Experiment Search for m+e+gwith sensitivity of 1 event forB(meg) = 10-13 • Italy • INFN and University of Genoa • INFN and University of Lecce • INFN and University Pavia • INFN Pisa • INFN and University of Roma • Japan • ICEPP, University of Tokyo • KEK • Waseda University • Russia • JINR, Dubna • BINP • Novosibirsk • Switzerland • Paul Scherrer Institute • United States • University of California, Irvine William Molzon, UC Irvine Particle Physics with Intense Muon Beams

10. The MEG Experiment at PSI 0.90 0.95 1.00 Eg / Emax 0.90 0.95 1.00 Ee / Emax • Experiment limited by accidental backgrounds:e+ from Michel decay, g from radiative decay or annihilation in flight. S/N proportional to 1/Rate. • DEe : 0.8% (FWHM) DEg : 4.5% (FWHM) • Dqeg : 18 mrad (FWHM) Dteg : 141ps (FWHM) • MEG uses the PSI cyclotron (1.8 mA at ~600 MeV) to produce 108m+ per second (surface muon beam) • Partial engineering runs autumn 2006, spring 2007 • First physics run 2007 • Sensitivity of 10-13 with 2 years running (c.f. MEGA 1.2x10-11) • Possible improvements to reach 10-14 in later 2 year run • Cooled beam would reduce backgrounds William Molzon, UC Irvine Particle Physics with Intense Muon Beams

11. Coherent Conversion of Muon to Electrons (-Ne-N) (-Ne-N) Re  (-NN(Z-1)) • Muons stop in matter and form a muonic atom. • They cascade down to the 1S state in less than 10-16 s. • They coherently interact with a nucleus (leaving the nucleus in its ground state) and convert to an electron, without emitting neutrinos Ee = Mm - ENR- EB.Coherence gives extra factor of Z with respect to capture process, reduced for large Z by nuclear form factor. • Experimental signature is an electron with Ee=105.1 MeV emerging from stopping target, with no incoming particle near in time: background/signal independent of rate. • More often, they are captured on the nucleus: m-(N,Z)→nm(N,Z-1) or decay in the Coulomb bound orbit: m-(N,Z)→nm(N,Z)ne( = 2.2 s in vacuum, ~0.9 s in Al) • Rate is normalized to the kinematically similar weak capture process: Goal of new experiment is to detect -Ne-N if Re isat least 2 X 10-17with one event providing compelling evidence of a discovery. William Molzon, UC Irvine Particle Physics with Intense Muon Beams

12. What Drives the Design of the Next-Generation Conversion Experiment? Considerations of potential sources of fake backgrounds specify much of the design of the beam and experimental apparatus. Prompt background Cosmic raybackground Expected signal SINDRUM2 currently has thebest limit on this process: Muon decay Experimental signature is105 MeV e-originating in a thin stopping target. William Molzon, UC Irvine Particle Physics with Intense Muon Beams

13. Potential Sources of Background dN/dE 10-10 10-5 1 0 50 100 E [MeV] • Muon decay in orbit – • Emax = Econversion when neutrinos have zero energy • dN/dEe (Emax – Ee)5 • Energy resolution of ~200 keV required • Radiative muon capture: - N   N(Z-1)  • For Al, Egmax = 102.5 MeV/c2, P(Eg> 100.5 MeV/c2) = 4  10-9 • P(g e+e-, Ee > 100.5 MeV/c2) = 2.5  10-5 • Restricts choice of stopping targets:Mz-1 > Mz • Radiative pion capture: p- N  N(Z-1)  • Branching fraction ~ 1.2% for Eg > 105 MeV/c2 • P(g e+e-, 103.5 < Ee< 100.5 MeV/c2) = 3.5  10-5 • Limits allowed pion contamination in beam during detection time Other sources: beam electrons, muon decay, cosmic rays, anti-proton interactions William Molzon, UC Irvine Particle Physics with Intense Muon Beams

14. Features of a New m-N→e-N Experiment • 1000 fold increase in muon intensity using an idea from MELC at MMF • High Z target for improved pion production • Graded solenoidal field to maximize pion capture • Produce 10-2 m-/p at 8 GeV (SINDRUM2 10-8, MELC 10-4, Muon Collider 0.3) • Muon transport in curved solenoid suppressing high momentum negatives and all positives and neutrals • Pulsed beam to eliminate prompt backgrounds following PSI method (A. Badertscher, et al. 1981) • Beam pulse duration << tm • Pulse separation  tm • Large duty cycle (50%) • Extinction between pulses < 10-9 • Improved detector resolution and rate capability • Detector in graded solenoid field for improved acceptance, rate handling, background rejection following MELC concept • Spectrometer with nearly axial components and very high resolution William Molzon, UC Irvine Particle Physics with Intense Muon Beams

15. Pulsed Proton Beam Requirement for m-N→e-N Experiment • Subsequent discussion focuses on accelerator operating 8 GeV with 41013 protons per second and 50% duty cycle – 50 kW instantaneous beam powerat 8 GeV • Pulsed proton beam generated using RF structure of appropriate accelerator or storage ring • To eliminate prompt backgrounds, we require< 10-9 protons between bunches for each proton in bunch. We call this the beam extinction. • Gap between proton pulse and start of detection time largely set by pion lifetime (~25 tp) William Molzon, UC Irvine Particle Physics with Intense Muon Beams

16. Secondary Extinction Device in Proton Beam • Improved extinction performance close in time to the filled buckets (particularly just before the filled bucket is important) • A means of continuously measuring secondary extinction at the target is essential • Conceptual idea for a secondary extinction device • Time-modulated magnetic deflection synchronized with filled buckets to deflect protons between buckets out of beamline • Field integral required depends on beam optics – ~500 G-m • Ideal time structure is rectangular wave – magnet pulsed at 731 kHz with rise and fall time < 50 ns • Less ideal solution is a time structure approximating a rectangular wave using multiple harmonics • Conceptual design of technical implementation completed • Stripline magnets with 7x7 cm2 bore, 5 m long, with peak field ~75 Gauss • Provides ~1.5 mrad deflection • Low-loss ferrite return yoke decrease power and radiated energy • Resonantly driven magnets, with one or more harmonics • Q of 100 reduces the required delivered power by ~100 • Shape (and complexity) can be adjusted for conditions • Modular design suggested: peak currents of order kA and peak voltages of order kV • Commercial high-voltage, high current capacitors found • Commercial ferrite with low losses at high frequency found • 1st two harmonics are in AM radio band, where commercial power amplifiers in needed power range (10 kW) are available Magnet cross-section William Molzon, UC Irvine Particle Physics with Intense Muon Beams

17. Monitoring Extinction production target Production solenoid bore incident proton beam magnet calorimeter collimators secondary proton TOF counters Measure time of originating in the muon production target • Dual magnetic spectrometer to measure 1-2 GeV secondary protons • Target, production solenoid fringe field, first collimator gives passive momentum selection • Two collimators with magnet between them gives second passive momentum selection • TOF counters reject soft background giving accidental coincidences • Calorimeter gives energy measurement • Trigger between bunches to get out-of-time beam rate • Periodically trigger in time with bunches to get normalization – also a good monitor of beam macrostructure Beam dump William Molzon, UC Irvine Particle Physics with Intense Muon Beams

18. A Model Muon Beam and Conversion Experiment Straw Tracker Muon Stopping Target Muon Beam Stop Superconducting Transport Solenoid (2.5 T – 2.1 T) Crystal Calorimeter Superconducting Detector Solenoid (2.0 T – 1.0 T) Superconducting Production Solenoid (5.0 T – 2.5 T) Collimators William Molzon, UC Irvine Particle Physics with Intense Muon Beams

19. MIT PSFC – MECO Design of Magnet System for m-N→e-N Experiment 5 T 2.5 T Very detailed CDR completed (300+ pages) Complete 3D drawing package prepared TS and SOW for commercial procurement developed Industrial studies contracts let and completed 2 T 1 T 1 T 150 MJ stored energy 5T maximum field Uses surplus SSC cable Can be built in industry William Molzon, UC Irvine Particle Physics with Intense Muon Beams

20. Muons Production and Capture in Graded Magnetic Field Pions produced in a target located in an axially graded magnetic field: • 50 kW beam incident on gold target • Charged particles are trapped in 5 – 2.5 T, axial magnetic field • Pions and muons moving away from the experiment are reflected • Superconducting magnet is protected byCu and W heat and radiation shield 150 W load on cold mass15 W/g in superconductor20 Mrad integrated dose Superconducting coil mW/gm in coil 2.5T 5T Azimuthal position Productiontarget Heat Shield Axial position William Molzon, UC Irvine Particle Physics with Intense Muon Beams

21. Production Target for Large Muon Yield Production target region designed for high yield of low energy muons: • High Z target material • Little extraneous material in bore to absorb p/m • Diameter 0.6 - 0.8 mm, length 160 mm • ~5 kW of deposited energy Water cooling in 0.3 mm cylindrical shellsurrounding target • Simulated with 2D and 3D thermal and turbulent fluid flow finite element analysis • Target temperature well below 100 C • Pressure drop is acceptable ( ~10 Atm) • Prototype built, tested for pressure and flow Fully developed turbulent flow in 300 mm water channel Preliminary cooling testsusing induction heating completed William Molzon, UC Irvine Particle Physics with Intense Muon Beams

22. Muon Beam Transport with Curved Solenoid Goals: • Transport low energy m-to the detector solenoid • Minimize transport of positive particles and high energy particles • Minimize transport of neutral particles • Absorb anti-protons in a thin window • Minimize long transittime trajectories Features: • Curved sections eliminate line of site transport of photons and neutrons. • Toroidal sections causes particles to drift out of plane;used to sign and momentum select beam. • dB/dS < 0 to avoid reflections 2.5T 2.4T 2.4T 2.1T 2.1T 2.0T William Molzon, UC Irvine Particle Physics with Intense Muon Beams

23. Sign and Momentum Selection in the Curved Transport Solenoid Transport in a torus results in charge and momentum selection: positive particles and low momentum particles absorbed in collimators. 3-15 MeV Detection Time Relative particle flux Relative particle rate in mbunch William Molzon, UC Irvine Particle Physics with Intense Muon Beams

24. Muon Beam Studies • Muon flux estimated with Monte Carlo calculation including models of p- production and simulation of decays, interactions and magnetic transport. • Estimates scaled to measured pion production on similar targets at similar energy • Expected yield is about 0.0025 -stops perproton • Cooling would increase stopping fraction Stopping Flux Total flux at stopping target Relative yield Data Model 0 0.5 1.0 0 50 100 Pion Kinetic Energy [GeV] Muon Momentum [MeV/c] William Molzon, UC Irvine Particle Physics with Intense Muon Beams

25. Stopping Target and Experiment in Detector Solenoid • Graded field in front section to increase acceptance and reduce cosmic ray background • Uniform field in spectrometer region to minimize corrections in momentum analysis • Tracking detector downstream of target to reduce rates • Polyethylene with lithium/boron to absorb neutrons • Thin absorber to absorb protons 1T Electron Calorimeter 1T Tracking Detector 2T Stopping Target: 17 layers of 0.2 mm Al William Molzon, UC Irvine Particle Physics with Intense Muon Beams

26. Magnetic Spectrometer to Measure Electron Momentum Measures electron momentum with precision of about 0.3% (RMS) – essential to eliminate muon decay in orbit background Electron starts upstream, reflects in field gradient • Must operate in vacuum and at high rates – 500 kHz rates in individual detector elements • Energy resolution dominated by multiple scattering • Implemented in straw tube detectors – • Vanes and octants, each nearly axial, and each with 3 close-packed layers of straws • 2800 detectors, 2.6-3.0 m long, 5 mm diameter, 0.025 mm wall thickness –issues with straightness, wire supports, low mass end manifolds, mounting system • r-f position resolution of 0.2 mm from drift time • axial resolution of 1.5 mm from induced charge on cathode pads – requires resistive straws, typically carbon loaded polyester film • High resistivity to maximize induced signal • Low resistivity to carry cathode current in high rates • Alternate implementation in straw tubes perpendicular to magnet axis has comparable performance William Molzon, UC Irvine Particle Physics with Intense Muon Beams

27. Spectrometer Performance Calculations Performance calculated using Monte Carlo simulation of all physical effects Resolution dominated by multiple scattering in tracker and energy loss in target Resolution function of spectrometer convolved with theoretical calculation of muon decay in orbit to get expected background. Geometrical acceptance ~50% (60-120) Cooling would reduce energy loss in target, improving resolution ( ~factor of two on width, much reduced tails) – offset by need for thicker proton absorber  FWHM ~900 keV William Molzon, UC Irvine Particle Physics with Intense Muon Beams

28. Expected Signal and Background with 4x1020 Protons Sources of background will be determined directly from data. Background calculated for 107 s running time at intensity giving 5 signal event for Rme = 10-16. 5 signal events with0.5 background events in 107 s running if Rme = 10-16 William Molzon, UC Irvine Particle Physics with Intense Muon Beams

29. Implementing a MECO Experiment at Fermilab • Complex of accelerators underused after Tevatron run finished • Proton beam could be shared with neutrino program – “15-20% loss of protons to neutrino experiments not a problem for a second high quality physics program” (P. Odone) • Encouraged by Fermilab director, especially in context of less than most optimistic NLC rampup • Implementation builds on mostly available accelerators, concepts of the MECO experiment • Momentum stack 3-4 booster bunches in the accumulator • Transfer to the debuncher • Rebunch the beam with RF (bunch spacing 1.6 ms) • Slow extract to new area • Very good macro duty cycle • Proton beamline, target station,muon beamline, experiment following MECO William Molzon, UC Irvine Particle Physics with Intense Muon Beams

30. Required Fermilab Beam Modifications AP4 Line A-D Line AP5 Line • Increased booster throughput (approved) • Transfer line from booster→accumulator symbiotic with neutrino program (unapproved) • Transfer line accumulator→debuncher • RF for rebunching • Slow extraction • Secondary beamline, experimental hall … • Concerns • Impact on protons for neutrino program • Radiation impact (losses, shielding) • Resource availability (NLC) Conservative intensity projection William Molzon, UC Irvine Particle Physics with Intense Muon Beams

31. LOI to Implement a MECO Like Experiment at JPARC • Uses JPARC high energy synchrotron running at 8 GeV • 4x1013 protons per second • 2x107seconds running time • 7x10-4 muon stops per proton • Pulsed, slow extracted proton beam with good extinction, 1.17 ms pulse spacing William Molzon, UC Irvine Particle Physics with Intense Muon Beams

32. JPARC Muon Beamline and Detector Schematic • Backward pion collection, production solenoid much smaller than MECO version • Transport with constant direction bend, drift compensated with dipole field • Converstion electrons transported to detectors in curved solenoid • Suppresses transport of low energy electrons • Transverse drift compensated by superimposed dipole field • Non-monotonic field gradient in transport region – allows for gradient in electron transport to improve acceptance • Expected sensitivity of 1 event for Rme 4x10-17 for 2x107 s running with 0.34 expected background events William Molzon, UC Irvine Particle Physics with Intense Muon Beams

33. END William Molzon, UC Irvine Particle Physics with Intense Muon Beams

34. Summary • A muon-to-electron conversion experiment at sensitivity below 10-16 has excellent capabilities to search for evidence of new physics and to study the flavor structure of new physics if it is discovered elsewhere first. • A well studied, costed, and reviewed experimental design exists that could be the starting point for a new effort at Fermilab. • A group of physicists is interested in exploring the possibility of doing this experiment at Fermilab and is eager to attract more interested physicists at the beginning of the effort. • This experiment would complement the neutrino program at Fermilab in the decade following the end of the Tevatron program and the beginning of a major new program. • An appropriate proton beam can probably be built and operated for such an experiment at Fermilab with net positive impact on the planned neutrino program: Dave McGinnis will discuss the beam and operational issues. William Molzon, UC Irvine Particle Physics with Intense Muon Beams

35. MECO Experiment Design as a Template for Fermilab Experiment Boston University I. Logashenko, J. Miller, B. L. Roberts Brookhaven National Laboratory K. Brown, M. Brennan, W. Marciano, W. Morse, P. Pile, Y. Semertzidis, P. Yamin University of California, Berkeley Y. Kolomensky University of California, Irvine M. Bachman, C. Chen, M. Hebert, T. J. Liu, W. Molzon, J. Popp, V. Tumakov University of Houston Y. Cui, E. V. Hungerford, N. Elkhayari, N. Klantarians, K. A. Lan University of Massachusetts, Amherst K. Kumar MECO collaborators at various stages in the experiment Institute for Nuclear Research, Moscow V. M. Lobashev, V. Matushka New York University R. M. Djilkibaev, A. Mincer, P. Nemethy, J. Sculli, A.N. Toropin Osaka University M. Aoki, Y. Kuno, A. Sato Syracuse University R. Holmes, P. Souder University of Virginia C. Dukes, K. Nelson, A. Norman College of William and Mary M. Eckhause, J. Kane, R. Welsh William Molzon, UC Irvine Particle Physics with Intense Muon Beams

36. Other Potential Sources of Backgrounds • Muon decay in flight + e- scattering in stopping target • Beam e- scattering in stopping target • Limits allowed electron flux in beam • Antiproton induced e- • Annihilation in stopping target or beamline • Requires thin absorber to stop antiprotons in transport line • Motivates proton energy not much above anti-proton production threshold • Cosmic ray induced e- – seen in earlier experiments • Primarily muon decay and interactions • Scales with running time, not beam luminosity • Requires the addition of active and passive shielding William Molzon, UC Irvine Particle Physics with Intense Muon Beams