Results from BELLE

1
Results from BELLE

• Sören Lange (Frankfurt/RIKEN-BNL)
• Cornell, April 26, 2002

2

• Some emphasis ondilepton, mixing, SVD
• What I cannot show you ACP Dilepton (not ready)
• What I can show you B?J/?Ks update 42/fb
  B???- and what it might mean (if true) SVD
  (still alive) mixing (still alive, too) BB,
  DD(yCP) Towards 1034 DAQ/Trigger, SVD 2.0
• Frankfurt SVD L2 trigger Dilepton CP, QM
  test(97/98 SVD DAQ for Tokyo Metropolitan Univ.)
• RIKEN QCD (fragmentation function)

3
Luminosity now

• 65 x 1033 (BABAR )
• Moriond 42/fb
• summer conferences (ICHEP) 90/fb (definitely
  maybe) re-processed

4
BELLE now (04/24)

• LER 0.95 A (2.6_at_1034), HER 0.77 A (1.1_at_1034)
• 6.792 x 10337.249 x 1033 (03/28)
• 70.125/fb on tape
• for summerre-processing of ALL data
  underwayD(sin2?1)?0.08 (0.12_at_42/fb)
• tried e (LER) continuous injection

5
Analysis - what re-processing means

• run-by-run IP profile in database (from now on
  even 1/10.000 events)from bhabha, mm or hadron
  (dx 140 mm, dy 85 mm, dz 2.8 mm)for re-fit helix
• run-by-run beam energy O(a few MeV) full
  reconstructed?B-gt Dop- shift in
  Mbc?(Ebeamcms)2-(pBcms)2)reminder typical
  resolution DMbc 10-40 MeVcould mean 1 more
  event in rare channels
• scale_momenta()solenoid runs _at_ 0.1 higher than
  nominal current (4,160 A)
• correct_ecl()EMCal low Eg non-linearity
• MC pseudo experiments uses run-by-runrandom
  trigd data different beam BG conditions

6
Analysis contd - new techniques

• J/? - mm- resolution is now smasslt10 MeV -
  ee- include e?ge Bremsstrahlung (Dqlt2.8o in ECL)
• slow pi efficiency no z -gt 2-dim tracks -gt use
  realghostnon-perfect SVD cluster1.3 more
  hadron events reconstructed
• Dalitz-plot becomes standarde.g. B?Kp-p,
  B?KKK- (works, because large BR observed
  5.5x10-5, 3.5x10-5)

7
SVD 1.4

• achieved vertexing _at_ efficiencygt91sz75 um (CP
  side) sz140 um (tag side charm)
• dose 213.7 kRad _at_ 04/18 3.1 kRad/fb-1
• VA1 gain loss -10 per 100 kRad
• 1/R2 confirmed for- occupancy (still O.K.
  lt3 _at_ inner layer, but few events buffer
  overflow)- dose (horizontal/vertical 21)-
  VA1 gain drop
• BABAR head-on -gt worse BG500 kRad/yr, 1 Mrad
  totalnew SVT 2004 -gt inner layer total 3.5 MRad

8
SVD 2.0

• 4 layers (z vertex resolution factor 1.9)
  increase of ch 81,920 -gt 110,592Ri3 cm -gt
  Ri2.2 cm (rule of thumb szmm Rmm) Ro6 cm
  -gt Ro8.8 cm (slow p track efficiency 54-gt77)
• polar angle coverage increased, now matches
  CDCladder length 22 cm -gt 46 cm (17oltqlt150o)
• re-designed IR region beampipe R1.5 cm dose
  1st layer / yr R1.5 cm 80 kRadR1.0 cm 660
  kRad ? IB decision too risky

9
SVD 2.0 VA1TA

• VA1TA- 0.35um -gt might reduce SEU
  probability (under test) -gt rad-tested up to
  5 Mrad (2.5 years) leakage
  current/noise increase only few - TA can
  generate trigger whenever any channel gt
  threshold -gt chip-global trigger on ta,tb
  (edge-trigd monostable flip-flop which gives
  fixed pulse, width externally adjustable)

10
SVD 2.0 DSSD

• before HPK S6936, now HPK 4387
• old DSSD contained integrated coupling
  capacitors? strips were AC coupled to
  pre-amplifier(prevents bad channels from
  saturating pre-amp)capacitors are now
  removedDSSD now DC opto-coupled to core
  (XILINXbuffer)requiresa.) leackage current
  compensation in VA1b.) floating ground in
  repeater system
• n-side - no double metal anymore ? changed to
  FLEX readout - atoll p-stop (same as before, but
  not Al-field-wire-stop)
• -gt NO pinhole ( AC couple double metal)
  anymore(but question to Haba-san didnt we
  have one p-side-pinhole ???)

11
SVD 2.0 FLEX

• SVD 1.x DSSD directly wirebonded to VA1SVD 2.0
  DSSD ? FLEX ? VA1TAa.) more freedom for
  mechanical arrangementb.) no circuits for z
  readout on DSSD necessary ? reduces load
  capacitance, less noisec.) thermal decoupling
  no heat VA ? silicon

12
New DAQ/Trigger (get ready for 1034)

• DAQ deadtime gt10 _at_ begin-of-spill ( CDC current
  limiter)
• new event builder 3 layers PC GB ethbut
  shows non-linear deadtime behaviour ( DMA
  interference)
• deadtime SVD now 160 us -gt 50 us(other
  subdetectors FASTBUS 80 us)
• L3 (if NO, no tape), L4 (if NO, no DST) since 2
  years
• L2 find SVD triplets (bitmask 32 strips
  ORd)calculated on DSP (3.5s gt noise)send to
  farm, decode/unpack R1.1 kHz if z vertexgt25
  cm, no EVB
• L0 TOF VA1TA (new !) Rlt50 kHz
• L1 TDC stop, as before (48 bits)
• new SVD L1.5 fast L2 (SVD lookup table )but
  (backward)-abort L1 (board under test, Princeton)

13
B-gtpi pi-

• high pT pi/K -gt PID fakes exactly BG for
  dilepton CP charge asymmetry
• special situation tree penguin -gt direct CP
  does not vanish
• BELLE mesurement is outside physical boundary
  S2C21
• toy Monte Carlo S0 and C0-gt 1.6 even further
  outside
• 3 sigma means -gt might go away w/ 3 x more
  statistics
• little wiggle _at_ dt 2 psis not cos() term, but
  qq sideband asymmtry !
• penguin pollution 0.3
• if true, 2nd appearance of direct CP in
  natureNA48 Re(e/e)(15.3?2.6)e-4 edirect

14
B-gtpi pi- BELLE vs. BABAR Phys.Rev.D
65(2002)051502

• data sample 41.8/fb 33.7/fb B-gtpipi B-gthh
  tight PID loose PID purity 36.6 purity 3
  of fit parameters 2 18 (Apipi,Spipi) (A_Kpi,
  event yields)

15
B-gtpi pi- how do we estimate A(penguin) ?

• BRsK pi gt pi pi gt K Kpenguin gt tree gt
  annihilation(no KK yet)
• T(K pi)/T(pi pi) P(pi pi)/P(K pi) Vtd/Vts
  1/4
• penguin pollution 0.3

16
B-gtpio pio

• BR 10-6
• if it stays 3sigma_at_45/fb, 4.4sigma_at_90/fb
• Grossman/Quinn bound2phi2-phi2effltacos(1-2BR(pi
  opi0)/BR(pipi0))should it have CPV ???
• BELLE BRs point to large phi3 or bbar-gtubar
  u d can change isospin by 1/2 or 3/2 DI3/2 ?
  (but this would mean problem for factorization
  based extractions of phi3)but so far
  consistent with CKM fits

17
BB Mixing

• was 1st BELLE paper Dm0.463?0.008?0.016 ps-1
  (5.6/fb)
• theory uncertainty still 10Dm mtop1.52 x
  hNLO x Bbaglattice x VtbVtd
• dilepton summer update sure 32/fb, maybe
  60/fb- syst. error (dominant) reduced by factor
  1.6 Dmgt0.500 ps-1 (now consistent with
  Dmhadronic)- ACP( - --)/( --) theory
  can not predict if positive or negative ?
  problem charge asymmetry for high pT fakes
  (p,K,p)- CPT fit, maybe DG and maybe maybe 1st
  QM test
• one side high-p lepton, one side D?p?
  (BR2.7x10-3)stat. and syst. error larger
  factor 2but see oscillation 1st step to
  Acpsin(2?1?3)confirms Dmgt0.500 ps-1

18
Test of QM in Dilepton

• Einstein-Podolsky-Rosen Correlations needs 2
  orthogonal variables B gt ? ?B gt one side
  oscillates B1 gt ? B2 gt other side does not
  oscillatebut B 1,2 gt B gt ??B gt thus
  oscillation is enough
• Decoherence could originate froma.) formfactor
  (quark substructure) QM is defined strictly
  for pointlike objects Weinberg
  DxDph(1Dp...)2.) quantum gravitation (small
  effect, but macroscopic Ds) tested w/ KK
  mixing (CPLEAR, hep-ph/9607434, Ellis et al.)
  fit sensitivity reaches 10-19 GeV goes
  M2, i.e. from KK to BB factor 100
• could occur as a dip _at_ Dz1.6 mm (Eur.Phys.J.
  C14(2000)469)-gt exactly where all dz
  distributions go to zero

19
Apex (Where are we going ?)

• Dilepton CP, CPT and probably 1st QM testfor
  sure 30/fb, maybe 60/fb
• SVD 2.0
• J/Psi K paper ready(Dsin2phi1lt1)

20
Future 2

• J/Psi 6 sigma deviation from zeroerror bar
  deltaphi1x 1/2 50/fb (1 yr)x 1/2 300/fb (5
  yr)
• BRs1034 BR10-8 1st SUSY tests1035 BR10-9
  if Super-KEKB
• nearCP in modes w/ no tree (phi K_s) or
  penguingtgttree (Ks pi0)-gt sensitive to new
  physics
• farb-gts nue nuebar
• 3

21
Ersatz
22
DD mixing search

• tau(Ds)/tau(D0) not theoretically explained yet
  (W)
• surprising FOCUS result ycp0.0342 non-zero 2
  sigma
• look for mixing effects in difference in decay
  ratebetween CP odd and CP even
• y_cp tau(D-gtK pi)/tau(D-gtK K) - 1problem
  K-pi is not an eigenstate -gt 10 shift in ycp
• BELLE ycp-0.005-0.0100.007-0.008 _at_
  23.4/fbconsistent with zero within 1
  sigma(error smaller stat. 2.5 syst. 2.0 than
  CLEO hep-ex/0104008)
• different from BB mixing (2 sides)1 side only,
  tag D flavour by D-gtD0piwrong sign can only
  occur through mixing or double cabbibo
  suppressed
• Do-gtKpi-pio will come but BR is how much lower ?

23
3 body B decays

• Dalitz-plot becomes standarde.g. B?Kp-p,
  B?KKK-hep-ex/0201007, submitted to
  PRD(works, because large BR observed 5.5x10-5,
  3.5x10-5)look for intermediate 2-body states1st
  observation of Pseudoscalar fo(980)Kalso for
  remove charm no r?? X states yet (why not ?)