Search for Lepton Flavour Violation using muon beams

1
Search for Lepton Flavour Violation using muon
beams

• Fabrizio Cei - INFN University of Pisa
• on behalf of the MEG Collaboration
• International Workshop on New Neutrino Beams
• Trento, 22 October 2004

2
Outline

• Theoretical motivations for searching for LFV.
• Previous searches and experimental limits.
• The MEG Experiment at PSI
• The experimental setup
• Present status
• Sensitivity and perspectives.
• Searches for m - e conversion (m-A ? e-A)
• - MECO PRIME
• What next at a high intensity muon source ?
• Conclusions.

3
Generalities on Lepton Flavour Violation
searches
4
What is Lepton Flavour Violation (LFV) ?

• LFV processes are reactions in which the leptonic
• number is not conserved (i.e. neutrino
  oscillations).
• LFV refers to processes involving
• CHARGED LEPTONS (m,e,t)
• Examples

5
Experimental limits
LFV searches History
MEG 2007
MECO 2010
PRIME ?
6
Advantages of usingmuon beams

• Muon limits are the most sensitive
• intense muon beams
• can be obtained at
• meson factories
• muon lifetime is
• rather long (2.2 ms)

(M. Aoki, 2004)
(upgraded by Belle BaBar 3.1 0.9 x 10-7)
7
Theoretical motivations
In the Standard Model with massive Dirac
neutrinos LFV processes (as m ? eg, t ? eg, m ?
eee, mA ? eA) are predicted at immeasurably small
levels (BR 10-50) However, Super Symmetric
Theories predict such processes at much more
reasonable rates. Since the SM background is
negligible, LFV processes are clear evidences for
Super Symmetry. Problem are such rates
experimentally observable ?
8
Predictions for m ? eg
LFV processes especially sensitive to SSM grand
unified theories (SUSY-GUT).
LFV induced by finite slepton mixing through
radiative corrections.

• Some predictions
• SUSY SU(5) BR (m ? eg) ? 10-14 ? 10-13
• SUSY SO(10) BRSO(10) ? 100 BRSU(5)
• (R. Barbieri et al., Phys. Lett. B338
  (1994) 212
• R. Barbieri et al., Nucl. Phys. B445
  (1995) 215)

9
Predictions for m ? eg BR in SU(5) Models
J. Hisano et al., Phys. Lett. B391 (1997) 341
Goal of MEG Experiment
Small (lt10) tan b values are highly disfavoured
by recent combined LEP data. (ALEPH, DELPHI,
L3 OPAL Collaborations, hep-ex/0107030)
10
Connection with n-oscillations
Additional contribution to slepton mixing from
V21, matrix element responsible for solar
neutrino deficit. (J. Hisano N. Nomura, Phys.
Rev. D59 (1999) 116005)
tan(b) 30O
tan(b) 0O
Largely favoured and confirmed by Kamland
After Kamland
All solar n experiments combined
11
Predictions for m-A ? e-A conversion
Compositeness
Supersymmetry
Predictions at 10-15
Second Higgs
After W. Marciano
12
SUSY predictions for m-A ? e-A
From Barbieri,Hall, Hisano
Rme
MECO single event sensitivity
PRIME

• ? eg m-A ? e-A Branching
• Ratios are linearly correlated

Complementary measurements (discrimination
between GUT models)
13
Search for m ? eg the MEG Experiment at PSI
14
Signal and background
background
signal ? ? e g
accidental ? ? e n n ? ? e g n n ee ? g g eZ ?
eZ g
physical ? ? e g n n
qeg 180 Ee Eg 52.8 MeV Te Tg
g
15
The Paul Scherrer Institute
Experimental Hall

• The most powerful machine
• in the world
• Proton energy 590 MeV
• Power 1.1 MW
• Nominal operation current 1.8 mA

16
Required performances
The sensitivity is limited by the accidental
background BR (m ? eg) ? 10-13 allowed, but
needed
FWHM
Need of a DC beam
17
The MEG Collaboration
ICEPP, University of Tokyo Y. Hisamatsu, T.
Iwamoto, T. Mashimo, S. Mihara, T. Mori, H.
Nishiguchi, W. Ootani, K. Ozone, T. Saeki, R.
Sawada, S. Yamada, S. Yamashita
KEK, Tsukuba T. Haruyama, A. Maki, Y. Makida, A.
Yamamoto, K. Yoshimura
Osaka University Y. Kuno
Waseda University T. Doke, J. Kikuchi, S. Suzuki,
K. Terasawa, A. Yamaguchi, T. Yoshimura
INFN Genova University S. Dussoni, F. Gatti, D.
Pergolesi, R. Valle
INFN Lecce University S. Spagnolo, C. Chiri, P.
Creti, M. Panareo, G. Palama
INFN Pavia University A.de Bari, P. Cattaneo,
G. Cecchet, G. Nardo, M. Rossella
INFN Pisa University A. Baldini, C. Bemporad,
F.Cei, M.Grassi, F. Morsani, D. Nicolo, R.
Pazzi, F. Raffaelli, F. Sergiampietri, G.
Signorelli
INFN Roma I D. Zanello
PSI, Villigen J. Egger, P. Kettle, M.
Hildebrandt, S. Ritt
Budker Institute, Novosibirsk L.M. Barkov, A.A.
Grebenuk, D.G. Grigoriev, B, Khazin, N.M. Ryskulov
18
MEG Experiment Layout 1)
The BEAM PSI pe5 surface muon beam

• 29 MeV/c muons from p stop at rest (surface
  muons)
• Provides a DC beam of ? 108 m/s.

19
MEG Experiment Layout 2)

• Muon beam stopped in a 150 mm target
• Liquid Xenon calorimeter for ? detection (using
  scintillation light)
• (Thin wall) solenoid spectrometer drift
  chambers for e momentum
• Scintillation counters for e timing

The Detector
Easy signal selection for m decaying at rest
20
The COBRA Spectrometer
COnstant Bending RAdius (COBRA) spectrometer

• Constant bending radius independent of emission
  angles

• High pT positrons quickly swept out

21
Positron Tracker

• 17 chamber sectors aligned radially with 10
  intervals
• Two staggered arrays of drift cells
• Chamber gas He-C2H6 mixture
• Vernier pattern to measure z-position made of
  15 mm kapton foils.

?(X,Y) 200 mm (drift time) ?(Z) 300 mm
(charge division on vernier strips).
22
Positron Timing Counter
BC404

• Two layers of scintillator read by PMTs placed
  at right angles with each other
• Outer timing measurement
• Inner additional trigger information.
• Goal ?time 40 psec (100 ps FWHM)

23
Liquid Xenon Calorimeter

• Only scintillation light (l 175 nm)
• 800 l of Liquid Xe ? 800 PMT
• fast 4 / 22 / 45 ns
• high Light Yield ? 0.8 NaI
• short X0 2.77 cm
• transparent to its own scintillation light

poorly known in literature
24
Where are we ? 1)

• Beam
• - m/e separation by Wien filter (7 ? 11) s
• - Rm after separation (7 ? 10) x 107 m/s
• - Spot size s ? (6 ? 7) mm in V/H directions
• - Beam line commissioning before the end of
  the year.
• Cobra Spectrometer
• - Builded by Toshiba and shipped to PSI last
  summer
• - Crash tests completed
• - Successful excitation test up to 380 A
• - Field mapping going on.
• Timing Counter
• - External layer tested, FWHM ? 100 ps
  achieved
• - PMT gain vs rate measured inside COBRA
  magnet
• - Final design of internal layer close to
  completion.

25
Where are we ? 2)

• Drift chambers
• - Small prototype tested at PSI with and
  without
• magnetic field sR ? 100 mm achieved
• - Successful tests of frame mechanics
• - Full scale prototype ready in November.
• Liquid Xenon Calorimeter
• - Cryostat and PMT supporting structure in
  design stage
• - Large Prototype (100 l) built and
  successfully operated
• - Large absorption length achieved (l gt 1
  m)
• - New beam test in progress
• - PMT characterization by using a cryogenic
  test facility.

26
Xenon Calorimeter Prototype

• The Large Prototype (LP)
• 40 x 40 x 50 cm3
• 228 PMTs, 100 litres LXe
• (the largest in the World)
• Purpose
• Measure the LXe optical properties
• (absorption, Rayleigh scattering)
• Test cryogenic operations on a long term and on a
  large volume
• Check the reconstruction methods
• Measure the Energy, Position and Timing
  resolutions.
• with
• Cosmic rays
• ?-sources
• 60 MeV e from KSR storage ring
• 40 MeV ? from Compton backscattering
• 50 MeV ? from p at PSI

27
Some results with Xenon Calorimeter Prototype
1) Measurement of absorption length
March 2002
Now
Strong absorption due to contaminants (H20)
After purification
labsgt 1m _at_ 90 C.L.
28
2) Energy resolution
?- p ? ?0 n and ?0 ? ? ?
4.8 FWHM _at_ 55 MeV with R lt 1.5 cm D from wall
gt 3 cm
still to be improved with a better
characterization of PMTs (QE, gain, ...) in the
dedicated test facility
29
Summary of MEG sensitivity
30
MEG time profile
Revised document
now
LoI
Proposal
Planning
R D
Assembly
Data Taking
1998 1999 2000 2001 2002
2003 2004 2005 2006 2007
Discovery of LFV is just around the corner ?
http//meg.psi.ch http//meg.pi.infn.it http//meg
.icepp.s.u-tokyo.ac.jp
More details at
31
A Look at the Futurefor m ? eg
32
What next for m ? eg ?
It should be very interesting to explore lower
BRs Can we gain order of magnitudes in
sensitivity by using more intense muon beams
( 1010 m/s) ?
J. Hisano et al., Phys. Lett. B391 (1997) 341 and
B397 (1997) 357
33

• Not an easy task !
• Sensitivity limited by accidental
  background
• a simple increase of muon rate does
  not improve sensitivity !
• We need much better detectors to reach BR (m ?
  eg) ? 10-15.
• Most limiting factors performances of e.m.
  calorimeter.
• Some possible suggestions to reduce the
  background
• - use high resolution beta spectrometers
  (DEe/Ee 0.1 feasible)
• - reduce the target thickness to improve the
  qeg resolution
• (possible because of higher intensity of
  muon beams)
• - use a finely segmented target (it requires
  good directional
• sensitivity to distinguish adjacent
  targets)
• -

34
A correlated process m ? 3e BR(m ? 3e) a
BR(m ? eg) 10-2 BR(m ? eg)Present limit BR(m ?
3e) lt 10-12 (SINDRUM Coll., Nucl. Phys. B260
(1985) 1) Also limited by accidental
background.(a Michel positron an ee- pair
produced by Bhabha scattering in the target)
Experimental advantage no photon ? no need of
e.m. calorimeter. However expected very high
rate in tracking system. ? dead time, trigger
pattern recognition problems need of large
modularity.
35
Beam requirements

• Continuous beam to reduce the accidental
  background
• Small momentum bite (Dp/p lt 10 ) to use thin
  targets
• Use surface muons (ps 29 MeV/c) to maximize
  the pion stopping
• rate in the target (R ? p3.5 for p lt ps)
• Good m/e separation by Wien filters
• Possible solution to realize an (almost)
  continuos muon beam
• insert a thin ( 10-3 interaction
  lengths) pion production
• target inside the proton synchrotron or
  recirculating LINAC.
• Protons recirculate many times and end up
  interacting.
• If they stay in the synchrotron/LINAC for a
  long time,
• an almost continuous muon beam can (in
  principle) be obtained.
• Problems - target heating
• - radiation in the target
  area (safety requirements).
• See J. Äystö et al., Physics with low-energy
  muons at a neutrino
• factory
  complex, hep-ph/0109217

36
Conclusions about m ? eg

• The MEG experiment at PSI is in advanced state
  of building and
• testing the data taking is foreseen for
  2006.
• The expected MEG sensitivity is down to BR(m ?
  eg) ? 10-13,
• a two orders of magnitude improvement in
  respect with
• present bound. Many SUSY models predict LFV
  in the
• m ? eg channel at this level or even higher.
• A further improvement in sensitivity by using
  more intense
• muon beams is not easy because of accidental
  background
• limitations strong improvements in detector
  technologies
• are needed. Moreover, a continuous beam must
  be realized.

37
Searches for m-A ? e-A conversionMECO and
PRIME
38
Muon to Electron Conversion Mechanism

• Low energy negative muons are stopped in material
  foils (Al for MECO, Ti for PRIME), forming muonic
  atoms.
• Three possible fates for the muon
• Nuclear capture
• Three body decay in orbit
• Coherent LFV decay (extra factor of Z in the
  rates).
• Signal is a single mono-energetic electron
• Muon lifetime in Al 0.9 ?s, in Ti 0.35 ms
  (in vacuum 2.2 ?s).

39
Signal and background
main backgrounds
signal ? (A,Z) ? e (A,Z)
MIO (muon decay in orbit) ? (A,Z) ? e n n (A,Z)
RPC (radiative pion capture) ? (A,Z) ? ? (A,Z-1)
Ee mm - EB
Beam related background
N.B. No coincidence ? no accidental background
40
Reduction of beam background

• 1) Beam pulsing
• Muonic atoms have some hundred of ns
  lifetime ? use a pulsed
• beam with buckets short compared to this
  lifetime, leave
• pions decay and measure in a delayed time
  window.
• 2) Beam quality
• insert a moderator to reduce the pion
  contamination (pion
• range ? 0.5 muon range) a 106 reduction
  factor obtained by SINDRUM II. No more than 105
  pions may stop in the
• target during the full measurement (? 1
  background event)
• select a beam momentum ? 70 MeV/c (muon decaying
  in
• flight produce low energy electrons).

41
Present limit SINDRUM II

• SINDRUM II parameters
• beam intensity 3 x 107 m-/s
• m- momentum 53 MeV/c
• magnetic field 0.33T
• acceptance 7
• momentum res. 2 FWHM
• S.E.S 3.3 x 10-13
• B(m-Au ? e-Au ) ? 8 x 10-13

42
The MECO Experiment

• Expected 1011 stopping m/s (1000fold increase in
  m beam intensity) for
• a proton current of 4 x 1013 protons/s and high
  Z target.

Superconducting Solenoids
Muon Beam
1 T
1 T
Crystal Calorimeter
2 T
Straw Tracker
Stopping Target Foils
Proton Beam
2.5 T

• Curved transport selects low momentum m- (n, g
  removed)
• High rate capability electron detectors in a 1 T
  field

5 T
Pion Production Target
43
MECO proton beam
Pulsed beam from AGS (BNL) to eliminate prompt
backgrounds.

• 1.35 µs separation between pulses for
• a 2.7 µsec rotation time. AGS cycle time
• is 1 sec.
• Work to be done.
• 2 x 1013 protons/bucket is twice
• the present AGS bunch intensity.
• In preliminary tests, extinctions
• of 107 have been achieved, but
• the needed extinction is gt 109.

44
MECO expected sensitivity
Expected 5 signal events for 107 s (2800 hours)
running if Rme 10-16
45
MECO background
Expected 0.45 bck events for 107 s running
time.
46
MECO time profile and funding
RSVP is in NSF budget, beginning in FY06 MECO
represents about 60 of its capital cost.
47
PRISM/PRIME (1)
PRIsm Muon Electron conversion experiment Phase
Rotation Intense Slow Muon source To be operated
at JPARC (Japan) Fixed Field Alternating
Gradient (FFAG) synchrotron
FFAG financed, ready in 2007

• High intensity pulsed proton beam
• ( 1014 p/s)
• Pion capture solenoid
• Pion decay section.

48
PRISM/PRIME (2)

• Muon energy spread reduction
• by means of a RF field
• (phase rotation) ?
• 2?3 FWHM energy spread
• Intensity ? 1012 m/s (no pions)
• Muon momentum 68 MeV/c.
• The small energy spread is essential to
• stop enough muons in very thin
• targets, improving the momentum
• resolution. If a momentum resolution
• ? 350 keV (FWHM) is reached,
• the experiment can be sensitive to
• m ? e conversion BRs down to 10-18.

Phase rotation concept
49
A Look at the Futurefor m- ? e- Conversion
50
A rough estimate for a possible pulsed muon beam
at NUFACT or Fermilab Proton Driver

• Using MECO numbers for scaling
• MECO
  Proton Driver NUFACT
• p/s (8 GeV) 4 x 1013 1 ?
  2 x 1014 1.5 x 1015
• m/s (tungsten target) 1 x 1011 3 ? 5 x
  1011 1 x 1012
• Sensitivity 2 x 10-17
  few x 10-18 1 x 10-18
• 
  Competitive with PRIME.
• (same extinction factor assumed 109)
• Power release 10 kW many tens
  of kW 100 kW
• 
  (need of target cooling to avoid melting)
• Need precise design and estimates.
• Ep 2.2 GeV scaling by using GHEISHA.

51
Conclusions about m-A ? e-A conversion

• The m-A ? e-A conversion BR is predicted by many
  SUSY theories
• at measurable levels ( 10-15). The m-A ? e-A
  conversion and the
• m ? eg channels are complementary in
  discriminating between
• LFV theoretical models.
• Two LFV experiments working in the m-A ? e-A
  conversion channel
• are in preparation/project MECO and PRIME
  results are
• expected in some years from now. They should
  improve the
• present limit on the m-A ? e-A BR ( 10-13) by
  at least
• three orders of magnitude.
• Since they are not limited by accidental
  background, m-A ? e-A
• conversion experiments can potentially benefit
  from the muon
• flux increase expected in Neutrino Factories
  and Muon Facilities.
• The key factors are momentum resolution and pion
  extinction
• factor. A suitably tuned pulsed muon beam is
  needed.

52
General conclusions

• Muons are sensitive probes of physics beyond the
  standard
• model SUSY theories predict LFVs not far
  from present limits.
• Important results in the search for LFV
  processes are expected
• in the next ten years from experiments under
  assembly or
• proposed MEG, MECO and PRIME further
  improvements could
• be obtained by using the intense muon fluxes
  to be provided by
• new muon facilities
• A relevant RD work is necessary in many cases
  to conceive
• better experiments, improve the detector
  technologies and
• design suitable muon beams
• However, the effort is worthwhile new physics
  could be around
• the corner .

53
On Demand
54
A simulated event
52.8 MeV photon
52.8 MeV positron
55
Detector Construction
Switzerland Drift Chambers Beam Line DAQ
Russia LXe Tests Purification
Italy e counter (GePv) Trigger (Pi) LXe
Calorimeter (Pi)
Japan LXe Calorimeter Magnetic spectrometer
56
The COBRA solenoid

• Bc 1.26 T, current 359 A
• Five superconducting coils with three different
  diameters conventional compensation coils to
  suppress the stray field around the LXe detector
• High-strength aluminum superconductor ? thin
  magnet
• (1.46 cm Aluminum, 0.2 X0).

57
COBRA gradient field
58
Trigger electronics

• Uses simple quantities
• ? energy
• Positron - ? coincidence in
• time and direction
• Built on a FADC-FPGA architecture
• More complex algorithms implementable

• Beam rate 108 s-1
• Fast LXe energy sum gt 45 MeV 2 ? 103 s-1
• g interaction point (PMT of max charge)
• e hit point in timing counter
• time correlation g e 200 s-1
• angular correlation g e 20 s-1

• First prototype board
• successfully tested
• Design and simulation of final
• prototype going on.

59
Readout electronics

• Waveform digitizing for all channels
• Custom domino sampling chip designed at PSI
• Cost per DSC 1 US
• 2.5 GHz sampling speed _at_ 40 ps timing resolution
• Sampling depth 1024 bins
• Readout similar to trigger

Prototypes delivered in autumn
60
The LP from inside
FRONT face
?-sources and LEDs for PMT calibration and
monitoring.
61
3) Intrinsic Timing resolution
Threshold effect
Fit by
? (0.55 ? 0.03)
55 MeV

• increased QE (5 ? 20)
• should improve the resolution

62
4) Radioactive background

• ?-trigger with 5?106 gain
• Geometrical cuts to exclude ?-sources
• Energy scale ?-source
• 208Tl (2.59 0.06) MeV
• 40K (1.42 0.06) MeV
• Other lines ??
• uniform on the front face
• few 10 min (with non-dedicated trigger)
• nice calibration for low energy ?s.

Never seen before !
63
PMT cryogenic test facility
Measurement of PMT relevant parameters (gain,
Q.E. ..) at level of ? 1 . Use of light sources
and a. Manifactured by CINEL-Vigonza (PD),
Italy.
64
MECO at BNL
65
MECO Spectrometer Performance (MC)
55, 91, 105 MeV e- from target

• Performance calculated using Monte
• Carlo simulation of all physical effects.
• Resolution dominated by multiple
• scattering in tracker.
• Resolution function of spectrometer
• convolved with theoretical calculation
• of muon decay in orbit to get
• expected background.

66
The JPARC Facility