CKM unitarity problem: results from NA48 experiment

1
CKM unitarity problemresults from NA48
experiment

• Evgueni Goudzovski (JINR)
• JINR Scientific Council
• January 20, 2005

2
The CKM matrix

• The Cabibbo-Kobayashi-Maskawa matrix connects the
  eigenstates d,s,b of weak interaction with the
  quark flavour eigenstates d,s,b

Vud Vus Vub
Vcd Vcs Vcb
Vtd Vts Vtb
d
s
b
d
s
b

• Conservation of probability
• CKM matrix has to be unitary VCKMVCKMI
• If unitary not fulfilled
• New Physics (e.g. 4th quark generation)

E. Goudzovski ? JINR SC ? January 20, 2005
3
CKM unitarity problem

• CKM unitarity requires for the 1st row
• Vud2 Vus2 Vub2 1
• Particle Data Group (PDG) 2004 review
• Vub(3.67 0.47)10-3 negligible
  contribution into unitarity relation, 10-5
• Vud0.9738 0.0005 well measured, e.g.
  neutron lifetime
• Vus 0.2200 0.0026 (sin?Cabibbo) old
  measurements semileptonic kaon decays
• Vud2Vus2Vub2 1?(4.31.9)10-3
• (2.2? deviation from unitarity)

E. Goudzovski ? JINR SC ? January 20, 2005
4
CKM unitarity Vus measurement

• Best way to determine Vus semileptonic
  (neutral and charged) kaon decays K??e? (Ke3)
• A few recent measurements of BR(Ke3) lead to
  Vus values significantly above PDG 2004 values
• New measurements of Vus are desirable.

E. Goudzovski ? JINR SC ? January 20, 2005
5
Determination of Vus from BR(Ke3)
E. Goudzovski ? JINR SC ? January 20, 2005
6
NA48 experiment at CERN SPS

• Main detector components
• Magnetic spectrometer (4 DCHs)
• redundancy ? high efficiency
• ?p/p 0.5 0.009p GeV/c.

LKr
HOD

• Liquid Krypton EM calorimeter (LKr)
• High granularity, quasi-homogenious
• ?E/E 3.2/vE 9/E 0.42
• electron/pion discrimination
• ? registration.

DCH

• Hodoscope
• high granularity multiplicity trigger
• precise time measurement (150ps).

E. Goudzovski ? JINR SC ? January 20, 2005
7
Vus measurements by NA48

• NA48/2 Semileptonic K decays K??0e? (Ke3)
• Data 90k events in 8 hours of low intensity run
  2003
• Loose hodoscope trigger (1 charged track)

• NA48 Semileptonic KL decays KL???e? (Ke3)
• Data 6.8 mln events in 2 days of special run
  1999
• Trigger on 2 charged particles in DCH or
  hodoscope

• NA48/1 Semileptonic ?0 decays ?0??e?
• Data whole high intensity run 2002 (6.2k
  events)
• Approach different to the one for kaon decays
• These preliminary results are not discussed in
  this talk.

E. Goudzovski ? JINR SC ? January 20, 2005
8
Br(K???0e?) measurement

• Measurement method normalize Ke3 events to
  K?????0 events (Br0.2113?0.0014)
• Signal practically background free
• Statistics selected from the minimum bias run

Decay Statistics
K??e? 59103
K????e? 33103
K???0 468103
K?????0 260103
E. Goudzovski ? JINR SC ? January 20, 2005
9
K???0e? Data/MC comparison
With radiative corrections
Without radiative corrections
Data/MC
Data/MC
E. Goudzovski ? JINR SC ? January 20, 2005
10
Br(K???0e?) result
confirms the deviation from PDG observed by BNL
E865! the most precise measurement of Br(Ke3)!

• Main systematics

?Br
Detector acceptance 0.038
Trigger efficiency 0.004
BR (K?????0) 0.034
Radiative events 0.006
MC statistics 0.022
Total systematics 0.056
Statistic uncertainty 0.017
Branching ratio
E. Goudzovski ? JINR SC ? January 20, 2005
11
Br(KL???e?) measurement
Phys.Lett. B602 (2004) 41

• Measurement method
• Use minimum bias trigger to collect KL?2-track
  events
• Normalization to Br(2-track) 1.0048 ? Br(3?0)
• Best input precision ?Br(2-track)/Br(2-track)lt0.9
  
• Exactly the same selection for signal (Ke3) and
  normalization events, but electron identified by
  energy deposit in the LKr calorimeter
• Measured quantity

E. Goudzovski ? JINR SC ? January 20, 2005
12
Data sample and selection

• Data sample
• 80 million 2-track triggers taken during 2 days
  of minimum bias run with pure KL beam
• Selection criteria
• Conditions on track geometry and kinematics
• Leave a sample of 12.6 million 2-track events
• Additional criterion for Ke3
• Electron ID E(LKr)/Pgt0.93for 1 track
• 6.8 mln candidates selected.

E. Goudzovski ? JINR SC ? January 20, 2005
13
Electron/pion separation

• Background to Ke3 sample
• K?3/K3? with ? misidentified as e
• Estimated from Ke3 data with identified e
  (E/pgt1)
• Prob(??e)5.810-3

• Inefficiency of electron ID
• Estimated from data with identified ?
  (0.3ltE/plt0.7)
• Prob(e??)4.910-3

E. Goudzovski ? JINR SC ? January 20, 2005
14
Monte Carlo simulation

• To determine acceptances, Monte Carlo simulation
  of 5 significant 2-track modes involved
  (radiative corr. included)

Decay channel BR (PDG04) Acceptance
KL??e? 38.8 0.2599
KL???? 27.2 0.2849
KL???-?0 12.6 0.0975
KL???- 2.110-3 0.5229
KL??0?0?0D 7.610-3 0.0001

• For average 2-track acceptance use ratios of BR
  averages from PDG KTeV (B?3/Be3, B3?/Be3, )
• Absolute BRs are not used!
• Acceptance(2-track events) 0.2412 ? 0.0004
• Small normalization uncertainty ?A/A0.16

E. Goudzovski ? JINR SC ? January 20, 2005
15
Errors on R?(Ke3)/?(2-track)

• Statistical errors are negligible
• Dominating systematic uncertainty due to inexact
  knowledge of beam energy spectrum
• Summary of systematic errors

?R/R,
Energy spectrum 0.67
Normalization (input BR) 0.16
E/P cut (electron ID) 0.05
Trigger efficiency 0.05
DCH overflows 0.05
Magnet polarity 0.07
E. Goudzovski ? JINR SC ? January 20, 2005
16
Data/Monte Carlo comparison
Kaon energy spectrum major uncertainty
Profiles at drift chamber
E. Goudzovski ? JINR SC ? January 20, 2005
17
Br(KL???e?) result
Phys.Lett. B602 (2004) 41

• Experimental result

• To compute Br(KL???e?) we use Br(KL?3?0)

PDG 04 0.2105?0.0023
KTeV 04 prel. 0.1945?0.0018
Br(KL?3?0)0.1992?0.0070
(inconsistent data scale factor applied to
error)

• Result on Ke3 branching ratio
• Br(KL???e?) 0.4010 ? 0.0028(exp) ? 0.0035(norm)
• 0.4010 ? 0.0045

E. Goudzovski ? JINR SC ? January 20, 2005
18
From BRs to unitarity test

• Br(K?) (5.14 ? 0.06)
• Br(KL) 0.4010 ? 0.0045
• K? f(0)Vus 0.2245 ? 0.0013
• KL f(0)Vus 0.2146 ? 0.0016
• K? Vus 0.2241 ? 0.0026
• KL Vus 0.2187 ? 0.0028
• (PDG 0.2200 ? 0.0026)

accuracy 1.1 1.1 0.6 0.7 1.3 1.2

E. Goudzovski ? JINR SC ? January 20, 2005
19
Results on Vusf(0)
(preliminary)
(2003)
PL B602 (2004) 41
E. Goudzovski ? JINR SC ? January 20, 2005
20
Conclusions

• Values of Vus obtained by experiments before
  2003 are in poor agreement with CKM unitarity
• Recent NA48 measurements of Vus
• From K?
• In agreement with new BNL result and CKM
  unitarity
• In disagreement with the old measurements
• From KL
• In agreement with new KTeV and KLOE measurements
• In better agreement with CKM unitarity than the
  old measurements.
• More precise calculations of f(0) are desirable.

E. Goudzovski ? JINR SC ? January 20, 2005