Form Factors and Absolute BRs

1
Form Factors and Absolute BRs for D0 ? p ?n / K?n
Belle in a nutshell
q²

• located at KEK / Japan
• KEKB Collider
• B-Factory at ?(4s) resonance
• peak luminosity 16.270 1/nb/s
• integrated luminosity 600 1/fb(as of June 2006
  280 1/fb used in this analysis)
• main physics goal observation of CPV in B meson
  Decays

Laurenz Widhalm HEPHY Vienna Belle Collaboration
KEK ?????
2
Why are semileptonic decays interesting?
q²

• single form factor fD(q2)
• calcuable in LQCD, but
• needs checking from data
• D-System ideal for experimental input
• results can be applied in B-physics (extraction
  of CKM parameters)

3
Method of Reconstruction (Event Topology)
?
additional primary mesons
IP
e
e-
3.5 GeV
8 GeV
D
g
p
mass-constrained vertex fits
D

• note
• all possible combinations tried in parallel
• cuts after complete reconstruction
• equal weight for remaining combinations
• ? no event loss due to particle exchanges!

Ktag
p
signal side
?
tag side
?
4
D0 Signal and Background
from decays without a D0 , or combinatorial
background

• cuts
• all mass-constr. fits CL gt0.1 (released on D0
  fit for righthand plot)
• same charge Ktag/pslow

control region
signal region
same sign Ktag/pslow
s0.0006 GeV!
signal D0 invariant mass
result (282 fb-1 of BELLE data) yield
after cuts 95250
background 38789
signal charge correlation signal subtraction stats bkg sample 56461 309stat 776syst 233syst 194syst
opposite sign Ktag/pslow
note data used for bkg subtraction, MC shown
only for comparison
data (normalized)
B0
charm (D)
MC
MC
charm, no D
uds
MC
MC
B
wrong sign D
MC
MC
5
D0 ? K/pln Signal and Background

• additional cuts
• same charge pslow / lepton
• extra g energy lt 700 MeV
• no excess charge
• En gt 100 MeV

recoil neutrino mass

D0 ? Kln

• Background sources
• fake D0
• other semileptonic channels
• hadronic channels

mn² / GeV²
D0 ? pln
signal region
note high resolution s(m²n)0.016 GeV²
6
D0 ? K/pln Signal and Backgrounds (for pmn)
semileptonic background
control region for K/r bkg measurement

signal region
recoil neutrino mass for D0 ? pln
note data used for bkg subtraction, crosschecked
by MC
bkg from misidentified kaons
fake-D0 bkg
data
data
bkg from Kmn
bkg from misidentified pions
data
data
smaller background for pen and Kln handled
likewise
bkg from K/rmn
MC
data
7
Summary of Signal / Background Decomposition
D0 ? Ken
D0 ? pen
remaining signal
data
fake-D0 bkg
data
D0 ? Kmn
D0 ? pmn
hadronic bkg
data
Kln bkg
data
K/rln bkg
MC
mn² / GeV²
Results (282 fb-1 of BELLE data) Ken Kmn pen pmn
signal events 1318 37stat 7syst 1249 37stat 25syst 126 12stat 3syst 106 12stat 6syst
fake D0 bkg 12.6 2.2 12.2 4.8 12.3 2.2 12.5 4.5
semileptonic bkg 6.7 2.6 10.0 2.5 11.7 1.2 12.6 1.9
hadronic bkg 11.9 5.6 62.1 23.9 1.8 0.7 9.7 3.7
error dominated by MC stats error
dominated by fit errors bias special bkg sample
8
Absolute Branching Ratios

• ratio to total number of recoil D0 tags
• efficiency correction
• corrected for bias due to differences data/MC
• (1.93.9)

BRs () this analysis PDG (2005) CLEO-c (hep-ex/0505035)
K-en 3.45 0.10stat 0.19syst 3.62 0.16 3.44 0.10stat 0.10syst
K-mn 3.45 0.10stat 0.21syst 3.20 0.17
p-en 0.279 0.027stat 0.016syst 0.311 0.030 0.262 0.025stat 0.008syst
p-mn 0.231 0.026stat 0.019syst 0.24 0.04
9
Form Factors q² distribution
D0 ? Ken
D0 ? pen
signal
q²
non-D bkg
hadronic bkg
semileptonic bkg
s(q²) 0.0145 GeV²/c² (width of red line) ? no
unfolding necessary!
D0 ? Kmn
D0 ? pmn
background shapes from data
10
Form Factors - Theory

• in principle, two form factors f(q²) and f-(q²)
• kinematically only f(q²) relevant, f-(q²)
  suppressed by ml²
• three different models that are frequently
  discussed in literature

simple pole
f (q²) (1-q²/m²) (1-aq²/m²)
modified pole
? 0.50 (Kln)
?
atheor.
G. Armoros, S. Noguera, J. Portoles, Eur. Ph. J.
C27, 243 (2003)
? 0.44 (pln)
ISGW2
f (q²) (1-a(q²-q²max))²
?
N. Isgur and D. Scora, Phys. Lett. B 592 1(2004)
11
Form Factors Comparison with Models
modified pole model
D0 ? Kln
lattice calculation
ISGW2 model
fit results
simple pole
pole mass (GeV)
Kln 1.82 0.04stat 0.03syst
pln 1.97 0.08stat 0.04syst
D0 ? pln
modified pole (poles fixed at theo. values)
f(0)
Kln 0.695 0.007stat 0.022syst
pln 0.624 0.020stat 0.030syst
a
Kln 0.52 0.08stat 0.06syst
pln 0.10 0.21stat 0.10syst
12
Form Factors and Absolute BRs for D0 ? p ?n / K?n
Summary Conclusion

• events searched in ee- ? D()DcX (Xnp/K)
• new full-reconstruction-recoil method 56k D0 in
  282 fb-1 of BELLE data
• high resolution neutrino s(m²n)0.016 GeV²
• background lt5(lt27) for K/p
• absolute BRs of better accuracy than previous
  experiments, in good agreement with recent CLEO
  measurements
• good agreement with relative measurements done
  by BES and FOCUS
• high q² resolution, no unfolding necessary
• absolute multi-bin measurement of f(q²)
• measured form factor in good agreement with
  theoretical predictions and other experiments
• competitative with recent CLEO-c measurements

Laurenz.Widhalm_at_oeaw.ac.at
preprint hep-ex/0604049, submitted to PRL
13
Spares
14
Method of Reconstruction (Event Topology)

• tag side
• reconstruction fit of D0, ? Kp, K2p, K3p
• reconstruction fit of D0, ? Dp, Dg
• use either D or D as primary meson
• signal side
• reconstruction fit of inclusive D0, via
  recoil from ee- ? D() Dnp/K
• reconstruction fit of inclusive D0 via recoil
  from D ? Dp
• reconstruction fit of neutrino via recoil from
  D ? mpn

additional primary mesons
e
e-
( )
D
D
D
p
p
g
D
n
p
K
tag
signal
15
Method of Reconstruction (Event Topology)
16

• stable particle selection
• gammas
• p gt 40 MeV
• charged tracks (general)
• p gt 100 MeV
• trk_fit.nhits(3) gt 0
• dr lt 2 cm, dz lt 4 cm
• electron
• p gt 500 MeV
• eid.prob(3,-1,5) gt 0.9
• muon
• p gt 500 MeV
• prerejection ! 1
• Muon_likelihood gt 0.9
• kaon / pion
• atc_pid (3,1,5,3,2)
• prob(1-prob_e-prob_mu) gt 0.5
• for meson in hlnu gt 0.9

• unstable particle selection
• pi0
• PDG mass 10 MeV
• fit CL gt 0.1
• K0
• only via decay pipi-
• PDG mass 25 MeV
• D_tag
• channels Knpi, n1-3
• PDG mass 20 MeV
• D_tag
• channel Dpi, Dg
• PDG mass 5 MeV
• mass/vertex fit CL gt 0
• D_signal
• via recoil from D_tagn pi/K, n0-5
• mass/vertex fit CL gt 0.001
• D_signal
• via recoil from D_signal ? Dpi

List of Cuts

• additional Klnu / pilnu cuts
• E_leftover lt 700 MeV, no leftover charge
• E_nu gt 100 MeV
• right charges of slow pions lepton

17
Bias by mass-constrained Fits on Background?
no real D0
?
?
with real D0
after fit of D
before fit of D
?
D0 invariant mass

• very sharp mass peak after fit
• no bias on background

?
18
D0 Signal and Background
from decays without a D0 , or combinatorial
background

• cuts
• confidence level of all mass-constrained vertex
  fits gt0.1 (released on D0 fit for righthand
  plot)
• right charge correlation between slow pion and
  tag side kaon (right sign, RS)

control region
same sign Ktag/pslow
signal D0 invariant mass

• procedure to measure background
• select wrong charge correlation data (WS) to get
  shape of background
• correct for small WS signal component
• normalize to RS data in region 1.84-1.85 GeV

signal region
opposite sign Ktag/pslow
B0 MC
charm MC
data (normalized)
charm, no D
uds MC
B MC
wrong sign true D
result (282 fb-1 of BELLE data) yield
selected D0 events 95250
subtracted background 38789
signal 56461 309stat 830syst
systematics breakdown events
charge correlation RS/WS 776
WS signal subtraction 233
statistics of WS sample 194
19
Measurement of Semileptonic Background (for pln)

• procedure to measure background
• 1. crosstalk from Kln
• prepare special background sample, with K
  intentionally misidentified as p
• normalize to standard Kln sample
• then reweight the sample using known
  efficiencies / fake rates (in p,?)
• 2. background from vector mesons
• get shapes for Kln and rln from MC (simulated
  ratio K/r from PDG)
• normalize to data in region m²n gt 0.3 GeV²

recoil neutrino mass
control region for K/r bkg measurement

D0 ? pln
signal region
data
non-D bkg (measured as described previously)
background for Kln is very small, and is
handled the same way measured independently in
data
measured bkg from Kln
measured bkg from Kln
measured bkg from rln
20
Measurement of Hadronic Background (for pmn)

• procedure to measure background
• prepare special background samples, with K(p)
  intentionally misidentified as m(subtract fake
  D0 background in these samples with the method
  described above)
• separate into same sign (SS) and opposite sign
  (OS) samples, with respect to the charges of the
  lepton and the slow pion
• semileptonic channels are highly suppressed in
  OS ? clean sample of hadronic background
• perform a 2-parameter fit in the standard OS
  sample, using the shapes from the OS background
  samples for K and p, to measure the effective
  fake rates
• then apply these fake rates in the background SS
  sample to obtain the backgrounds in the signal
  sample

same sign SS signal D ? D0p
? p-mn
SS
both signs D ? D0p ?
p-pp0/K0 p-m n
opposite sign OS D ? D0p
? K-pp0 pm-n
SS
OS
OS
D ? D0p ? K-pp0/K0 p-m n
significant background only for this channel
other channels are handled likewise
21
Fit of Hadronic Background (for pmn)
D0 ? pmn
comparison with MC
MC true composition
SS
green particle seen in recoil mass
D0 ? p-pp0
D0 ? K0p-p
D0 ? K-pp0
OS
signal region
bkg from misidentified kaons
bkg from Kmn
bkg from misidentified pions
bkg from K/rmn
background for pen and Kln are much smaller
remaining events in signal region
fake-D0 bkg
22
Measured Absolute Form Factors as function of q²
D0 ? pen
D0 ? Ken
D0 ? pmn
D0 ? Kmn

• extracted by dividing q² distribution by
  kinematical factor
• no unfolding necessary due to very good q²
  resolution

23
Form Factor Theory

• in principle, two form factors f(q²) and f-(q²)
• kinematically only f(q²) relevant, f-(q²)
  suppressed by ml²
• applying certain boundary conditions, theory
  suggests model-independently a pole-structure for
  the form factor

G. Amoros et al., hep-ph/0109169