Study%20of%20?-%20?%20?%20-?%20-?%20 ?%200?%200%20n?

1
Study of ?- ? ? -? -? ? 0? 0 n?

• Jenny Williams
• Manchester
• for Christina Edgar, George Lafferty, JW
• BaBar Tau Physics Workshop
• 6 December 2003

2
Where we got to last time
Future plans from George Laffertys talk at July
03 BaBar Collaboration meeting
?

• Use full Run1Run2 datasets with Release 12
• Use more and larger MC samples
• Improve cuts to enhance efficiency and purity
• Investigate kinematics/dynamics in Monte Carlo
  possibly use reweighting
• Measure branching ratios for 5 p (including w p
  p, h p p ?)
• Produce a BAD by end 2003
• Spin-parity or amplitude analysis of constituent
  channels
• Include Run3 data
• Note for spring or summer conferences 2004
• Journal paper by end 2004

?
?
Some progress
?
?
Coming soon
?
Still on track
Still on track
3
Some motivation

• Measurement of branching ratio for p - p - p p
  0 p 0
• Two existing measurements (CLEO) (0.48 ? 0.04
  ? 0.04)
• 
  (ALEPH) (0.50 ? 0.07)
• Study hadronic structure of final state, and BRs
  for sub-channels
• e.g. w p- p 0, h p - p 0, f p - p 0, b1 p,
  a1 p p ...
• Contribution to studies of axial spectral
  functions
• Look for evidence for a JP 0- q q g hybrid
  state, suggested by Crystal Barrel at M ? 1500
  MeV, G ? 120 MeV
• 
  (Phys Lett B563, 140 (2003))
• Because its there

-
4
Data and Monte Carlo Samples

• Analysis uses latest TauUser ntuples at
  Manchester
• series-12 Runs 1-3 data and SP5 MC (kk2f t MC),
  analysis-14a
• Samples used

Lumi (fb-1) N events (?106)
Data 124.4
tt- 176 157
uds 143 300
cc 143 186
B0B0 459 246
BB- 412 220
m m- 83 97
5
Event Selection I

• Nano Preselection
• Stream19/TauQEDallevents
• 2-11 nano tracks
• 1-9 nano good tracks

• Micro Preselection
• 1-3, 3-3 topologies
• Thrust ? 0.85
• 0-10 AWG p0s
• 0-20 pi0AllLoose
• 0-20 CalorNeutral

Further Event Selection

• Zero net charge, ?1 on either hemisphere
• Exactly 2 AWG p0s, lt 6 pi0AllLoose
• Missing energy gt -50 MeV
• Use only 1-3 events uds backgrounds in 3-3 s
  too high
• Tag electron (eMicroTight) or muon
  (muMicroLoose)
• Or
• Cone require overlap of cones from reconstructed
  ts (need decays ???? ? hadrons)

6
Thrust lt 0.93
Event Selection II
Thrust and E(p0) cuts
Optimise ratio
S num signal events (p-p-pp0p0)
(including all subchannels) B num bkg events
(t non-t)
Pi0 energy gt 205 MeV
7
Data/MC comparison of some distributions for
selected tracks and neutrals
All these plots normalised to the same area
8
Monte Carlo truth codes for selected 3-prong t s
p -p -p p 0p 0
a1 ? p -p -p

• p -p -p p 0

h p -p 0, h ? p p -p 0
p -p -p p 0p 0p 0
a1 ? p -p 0p 0
120
51
52
15
13
8
10
Overall signal efficiency e?
1.88

• Code 10 p -p -p p 0p 0
• Code 120 h p -p 0, h ? p p -p 0

Signal
9
5p mass spectrum
Background-subtracted 5p mass
pppp0p0
hpp0, h? p p - p0
Other t
uds
ccbar
10
pp-p0 mass spectrum for M5p lt 1.8 GeV
11
w region in p p - p 0 mass spectrum for M5p lt
1.8 GeV
PEAK
WING
WING

• Fit (not shown) to w peak gives
• m 781.9 ? 0.5 MeV
• ? 11.9 ? 0.6 MeV
• Define peak and sideband regions as
• Peak m ? 2s
• Lower sideband -5s ? -3s
• Upper sideband 3s ? 5s
• Then peak minus sidebands defines bgd-subtracted
  signal

12

• w p - p 0 mass spectrum using
• peak minus sidebands
• Distribution in data is narrower than in MC
  signal (but broader than 120 MeV)
• Branching ratio in MC, based on ALEPH result, is
  higher than data

13
Spectrum of recoil p-p0

• Look at leftover p p 0 for 3 p combination in w
  signal region
• Distribution of the recoil p -p0 different in
  data and MC
• MC generated with leftover p- p 0 as phase space
• Data has strong distortion of phase space
• Suggests a preference for
• ?-? ? ? - ? t

14
Dalitz plot density variable l

• ? (p p x p p- ) 2
• For w (JP1-) density is p(l) ? l
• For phase-space p(l) ? constant
• Data looks fine, MC signal has wrong phase space
  distribution

Peak minus bgd
MC signal
Bgd region
Peak region
15

• ? (p p x p p- )2
• Define JP1- weight as
• W 3 l-1
• For p(l) ? l , average weight is 1
• For p(l) ? constant, average is 0
• p p - p 0 mass spectrum, for
• M5p lt 1.8 GeV, weighted by W
• Shows W0 in signal MC (open hist)

16
p- p 0 mass spectrum opposite w in data weights
(points) and bgd subtraction (hist)
p- p 0 mass spectrum opposite w, using weights
Excellent agreement in data (but not in MC)
between weighting and background subtraction
17
Helicity angle for w Normal to 3p decay plane in
w frame wrt to w in 5p frame
Peak region
Sideband regions
Peak - sidebands
18
Helicity angle for p p 0 (r) system p
direction in r frame with respect to r direction
in 5p frame
Peak region
Sideband regions
Peak - sidebands
19
Angle between p wrt r and w decay plane
normal wrt w
For data ratios (peak-bgd)/bgd for the
three angular distributions
Interesting non-isotropic distributions, not
reproduced by MC
20
?-?? - ? - ? ? 0 ? 0 nt branching ratio
measurement
BR(t-?p- p- p p0 p0 nt) (4.93 ? 0.06 ? 0.07)
? 10-3 (Preliminary)
BR(t-?p- p- p p0 p0 nt) (5.4?0.4) ? 10-3
(PDG) 5.43 ? 10-3 (MC)
21
?-?? - ? - ? ? 0 ? 0 nt branching ratio
measurement

• Statistical error includes statistical
  uncertainty from background subtraction
• Systematic uncertainty includes
• MC statistics 0.56
• Luminosity assumed 1
• t t - cross section assumed 1
• Other sources ???
• New BABAR result improves relative overall error
  from 7.4 in PDG to 1.8

22
?-?? - ? - ? ? 0 ? 0 nt h, w, f subchannels

• Mass resolution from fitting MRECON - MTRUE

h mass region
w mass region
f mass region
RMS9.9 MeV
RMS11.9 MeV
RMS7.9 MeV
23
? -?f ? - ?0 nt branching ratio limit
BR(t-?f p-p0nt) lt 3.5 ? 10-4 at 95 CL
(Preliminary)

• Worlds first limit on this branching ratio
• Solid curve fit to quadratic polynomial
  Gaussian
• Gaussian
• central value fixed to PDG f mass
• width from fit to MRECON - MTRUE
• Dotted curve maximum possible contribution from
  any f, at 95 CL
• Statistical error gt systematic error

24
? - ? h? -?0 nt branching ratio
BR(t-?h p-p0nt) (1.32?0.11?0.03) ? 10-3
(Preliminary)
BR(t-?h p-p0nt) (1.74?0.24) ? 10-3
(PDG)

• Good agreement
• Factor 2 reduction in error

25
? - ? w? -?0 nt branching ratio
BR(t-?w p-p0nt) (2.94?0.09?0.05) ? 10-3
(Preliminary)
BR(t-?w p-p0nt) (4.3?0.5) ? 10-3
(PDG)

• Gaussian
• Mass fixed
• Resolution allowed to float
• Reasonable agreement with MC resolution, but so
  far taking no account of w width
• Factor 5 reduction in error

26
What now?

• Efficiency/purity studies
• Including tag selection and signal-side p-id
• Include two-photon MC
• Sort out systematics for branching ratio
  calculations
• Spin-parity or amplitude analysis of 5 p
  subchannels
• Study evidence for w r
• BAD note released by end December 2003
• Note for spring or summer conferences 2004
• Journal paper by end 2004

27
Extra Stuff
28
Inclusive pp- mass spectrum
Inclusive pp0 mass spectrum
Generally good agreement, though more K0
and r in data
29
Event Selection 3-3 rejection, and need for
cones
uds bkgd in 3-3 events too high
Need 1-prong tag or cone requirement
5p mass with 3-3 topology
5p mass without 1-prong tag
uds
30
AWG Tracks

• goodTracksVeryLoose
• Exclude tracks which are daughters of entries in
  conversions list gammaConversionDefault

AWG p0s

• pi0AllLoose
• No merged p0s
• No dead/noisy channels
• Photon energy deposited in ? 2 crystals
• Lat 0.001-0.05
• Eg gt 50 MeV
• Split-off energy cut 110 MeV
• Split-off distance cut 25 cm
• Fit probability c2 lt 5.0
• Combinatorics keep best fit p0

31
Pi0 Energy Cut Studies
32
5p mass with no p0 energy cut

• More charm pairs
• Monte Carlo slightly above data in region above
  1.8 GeV

33
Missing signal-side energy - Data
DE EBEAM,CMS- EVIS

• Gaussian peaked at zero for fully reconstructed
  decays
• Negative for tau events missing neutrino
  energy

34
Missing signal-side energy - MC
35
5p spectrum for 1-3 with cone selection only

• Similar shape to lepton-tagged sample
• Lower statistics