S' Stone Syracuse Univ' July 2003

1
S. StoneSyracuse Univ.July 2003

• Experimental Results in Heavy Flavor Physics

Far too much interesting Material to include in
45 min. Apologies in advance.
Many references not properly cited here.
Apologies in advance.
2
Physics Goals

• Discover, or help interpret, New Physics found
  elsewhere using b c decays - There is New
  Physics out there Standard Model is violated by
  the Baryon Asymmetry of Universe by Dark Matter
• Measure Standard Model parameters, the
  fundamental constants revealed to us by
  studying Weak interactions
• Understand QCD necessary to interpret CKM
  measurements.

3
b c quark decays

• Complete picture requires many studies including
  rare decays CP Violation Hitoshi will cover
  CP
• I will give an overview and cover rare decays,
  and show how they can uncover new physics.
• Interpreting fundamental quark decays requires
  theories or models than relate quarks to hadrons
  in which they live and die. I will say something
  about this.
• Theoretical issues will be dealt with in more
  depth by Thomas Mannel this afternoon.
• Hitoshi will also cover future experiments

4
Some basics b Lifetimes

• Note ratio tB/tBo
• 1.0730.014, a
• 5.2s difference
• Also tL ? tBo
• According to
• proponents of the
• Heavy Quark Expansion model, there should be
  at most a 10 difference between b-bayron Bo
  lifetimes

b
5
Some basics Charm lifetimes

• New precise lifetimes from FOCUS, a challenge to
  QCD (from Pedrini)

6
The Basics Quark Mixing the CKM Matrix
d s b
mass
u
c
m a s s
t

• A, l, r and h are in the Standard Model
  fundamental constants of nature like G, or aEM
• h multiplies i and is responsible for CP
  violation
• We know l0.22 (Vus), A0.8 constraints on r h

7
The 6 CKM Triangles

• From Unitarity
• ds - indicates rows or columns used
• There are 4 independent phases b, g, c, c? (a
  can be substituted for g or b, as abgp)

c?
c
g
b
a
8
B Meson Decay Diagrams not rare

• a) is dominant
• More diagrams for baryons
• Semileptonic decays are the simplest. Lets start
  with them

9
Semileptonic B Decays average Bsl

• On Y(4S), use high momentum lepton to tag flavor
  of 1st B

BaBar
Bsl10.890.23

• LEP (most recent EW fit) 10.59?0.22?10.76 at
  Y(4S))

10
Bo B semileptonic Decays

• Accurate, separate Bsl for Bo B from Belle,
  tagged by fully reconstructing one B
  (preliminary)
• B(B?Xln)/B(B0?Xln)1.14?0.04?0.01

Gsl(B)/Gsl(Bo) 1.0630.038
11
Vcb Vub

• Theory versus Models
• Theories describe phenomena make predictions
  based on general principles. They can have one or
  two unknown parameters (e.g. coupling constants)
  and if not exact, must prescribe a convergent
  series approximation.
• Examples are Lattice QCD (unquenched) and HQET
• Models contain assumptions. It is not only that
  the models may be wrong that causes us a problem,
  just as serious is that the errors on the
  predictions are difficult to estimate

12
HQET Vcb

• Heavy Quark Effective THEORY (HQET) (Isgur
  Wise)
• QCD is flavor independent, so in the limit of
  infinitely heavy quarks qa?qb occurs with unit
  form-factor F(1)1 when the quarks are moving
  with the same invariant 4-velocity, w1.
• Example for B?D?n
• All form-factors are related to one universal
  shape that can be measured
• Corrections to F(1) due to finite quark masses
  are calculable along with QCD corrections. These
  corrections are parameterized in a series
  SnCn(1/mqi)n, n1, 2
• To find Vcb measure value of decay rate at w1,
  here D is at rest in B rest frame

13
F(1)Vcb using B?D?n
F(1)Vcb
Belle

• Fit to function shape given by Caprini et al.
• Yields value of F(1)Vcb shape, parameterized
  by r2.
• F(1)Vcb (38.8?1.1)?10-3 (HFAG)
• r21.490.15 (HFAG)

14
Theoretical calculations of F(1) Vcb

• F(1)hQEDhQCD(1d1/m2)
• Lukes theorem no d1/m corrections (would be in
  Dln)
• hQED1.007, hQCD0.9600.007 at two loops
• d1/m2 involves 1/mb2, 1/mc2, 1/mcmb
• First Lattice Gauge calculations
  (quenched-no light quark loops)
  ultimate solution
• PDG (Artuso Barberio) F(1)0.910.05
• Average Vcb(42.61.2exp2.3thy)x10-3

15
The Heavy Quark Expansion

• Predict inclusive distributions
• Total widths, i.e. lifetimes
• b??c or u semileptonic widths
• Relies on mbgtgtLQCD
• Uses Operator product expansion
• Express decay widths in series in (1/mq)n asn
• Some authors Chay, Bigi, Uratsalev

16
Problems with HQE

• Terms in 1/mb3 are multiplied by unknown
  functions hard to evaluate error due to these
  higher order terms
• Duality is assumed integrated over enough phase
  space the exclusive charm bound states the
  inclusive hadronic result will match at
  quark-level. But no way to evaluate the error
• Appears to miss Lb lifetime by 105 b-baryon
  by 18 3 however semileptonic decay may be
  easier
• Need experimental tests to evaluate errors
• Not used for charm as mc is not expected to be
  large enough, but could be done to see how much
  it diverged
• Perhaps use Vcb as a test? Then could be used for
  Vub

17
Parameters of the HQE

• The heavy quark expansion of the B-meson (B?X l
  n) decay rate is described to the order
  (LQCD/mb)2 by three parameters
• l1 (1/2?MB) ltB(v) hv (iD)2 hv B(v)gt is the
  kinetic energy of the residual motion of the
  b-quark.
• l2 (-1/2?MB) ltB(v) hv (g/2)?smn Gmn hv B(v)gt
  the Chromo-magnetic coupling of the b-quark spin
  to the gluon field. Is determined from (MB-MB)
  mass splitting as 0.12 GeV2
• Decay rate also depends on quark-masses via,
• MBmbL-(l13?l2)/(2?mb)
• MBmbL-(l1- l2)/(2?mb) (Thus defining L)

18
How to Measure l1 L

• Can determine l1 and L, and thus Vcb by measuring
  moments in semileptonic decays
• Hadronic mass moments (ex ?MX2 - MD2?, MD is
  spin-averaged D, D mass) where B?Xln
• Semileptonic moments
• Can also use b?sg decays,
• here we use the 1st moment
• of the photon energy

19
Moments (CLEO)

• Hadronic Mass Lepton Energy moments found in
  semileptonic decays detecting the neutrino
  using missing energy
• b?sg moment determination shown later
• Fitting this other data Bauer, Ligeti, Luke
  Manohar find Vcb (40.80.9)x10-3
  mb4.740.10 GeV (hep-ph/0210027)
• Within 5-10 of Vcb(42.61.2exp2.3thy)x
  10-3 from D l- n

L0.35 0.07 GeV l1 -0.240.07 GeV2 exp errors
only
20
BaBar Moments Result

• Using only BaBar hadronic moments Bsl
• Vcb(42.11.00.7)x10-3 again within 7 of Dln
• mb1S4.640.090.09 GeV
• (Mx2 as function of lepton momentum, is now
  consistent with theory)

Mx2 moments
2002
1s contours
See talk by Urs Langenegger
Doesnt include 1/mb3 errors
21
Comparison of Hadron Lepton Moments (BaBar)

• Lepton Hadron moments differ somewhat. Does
  this indicate a Duality violation?
• Difference of 0.2 GeV in mb leads to 20
  difference in Vub

9
22
Delphi 2003 Moments Results

• Vcb(42.40.60.9)x10-3, using mb1S4.560.15,
  (see Hoang similar to using b?sg)
• L0.540.07 GeV
• l1-0.350.07 GeV2
• (only experimental errors)

Moments situation is still evolving, expect more
results soon
23
Vub
u

• Use semileptonic decays. Even here, modeling
  errors will dominate - there isnt a precise
  theory our path through this discussion will be
  perilous
• One method is to use exclusive decays
• B?pln B?rln
• Currently unquenched lattice
• error 10 quenching
• error 20
• Also QCD sum rules (Ball)
  

B?pln
24
Vub from exclusivesB?pln B?rln

• Use detector hermeticity to reconstruct n
• CLEO finds rough q2 distributions, good enough to
  limit model dependence of form-factor in
  determining branching ratios

25
Vub Results from pln rln

• CLEO
• theory error averages over quenched lattice
  results for q2gt16 GeV2, light cone sum rules for
  q2lt16
• BaBar
• theory error averages over one lattice model and
  various form-factor models
• Theory errors are assigned by the experiment. Are
  they large enough?

26
Modeling Issues for Inclusive Vub Determinations

• HQE can probably get to 9 accuracy if entire
  rate was measured. Duality mb5 errors
  (mb4.740.10 GeV)
• However experimental cuts are required to reduce
  100x larger b?c rate
• Parton rates are modified by Fermi motion
  distribution must know f(k) or error is
  introduced

See DeFazio Neubert
From Bauer, Ligeti Luke
Cut Mxlt3.4
Cut q2lt12
2
Cut Elgt2.2
27
Problem due to Weak Annihilation
Gluons break helicity suppression

• An issue for all inclusive determinations and
  exclusive decays, especially B-
• Relative size of effect worsens the more severe
  the cut
• No reliable estimate of the size

28
Vub Inclusive (b?u l n) Experimental cuts
Theoretical Issues (ala Luke)

• typical cuts used to eliminate 100x larger b?c
  backgrounds
• All cuts increase theory errors
• f(k) refers to Fermi motion
• WA weak annihilation

q2 region
Lepton Energy
Mx2
Mx2
Mx2
29
Vub Using Inclusive Leptons

• ALEPH DELPHI, OPAL select samples of charm-poor
  semileptonic decays with a large number of
  selection criteria
• Mass lt MD ? b ? u
• Small signal on large bkgrd
• Vub(4.09?0.37?0.44
  ?0.34)x10-3

DELPHI
30
Vub using reconstructed tags - BABAR

• Use fully reconstructed B tags

• Vub(4.62?0.28(stat)?0.27(sys)
• ?0.40(fu) 0.26(thy)) x10-3

31
Vub from lepton endpoint

• Vub both overall rate fraction of leptons in
  signal region depends on model. Use CLEO b ?sg
  spectrum to predict shape
• CLEO Vub(4.080.34exp 0.44fu0.16OPE0.24sg)
  x10-3
• BaBar Vub(4.430.29exp 0.50fu0.25OPE0.35sg)x1
  0-3
• Bauer, Luke Mannel error due to subleading
  twist (called sg) should be 15
• Also new Belle result, see talk by Schwanda

CLEO
? Y(4S) data b?cln continuum
CLEO
? b?uln
Shape from b ?sg

• Additional error from WA up to 30

32
Vub from Belle

• Two techniques (Both Preliminary)

D()ln tag uses Mx lt 1.5 GeV
Vub 5.00 ? 0.60 ? 0.23 ? 0.05 ? 0.39 ? 0.36 ?
10-3
b?c
b?u
stat.
syst.
theor.
n reconstruction and Annealing uses Mxlt1.5 GeV,
q2gt7 GeV2
Vub 3.96 ? 0.17 ? 0.44 ? 0.34 ? 0.26 ? 0.29 ?
10-3
stat.
syst.
b?c
b?u
theor.
Annealing is a method of separating the Y(4S)
event into one semileptonic B decay and the other
B decay
33
Vub Summary

• All measurements nicely clustered. RMS 0.3x10-3
• However, there are theoretical errors that have
  not been included
• Also previous values may have influenced new
  values
• Possibly safe to say Vub(4.01.0)x10-3
• The more you learn the bigger the errors get
• Future
• More and better tagged data from B-factories
• Lattice calculations (unquenched) for exclusives
  in high q2 region

34
My Best Value for Vub

• Since we want to see if New Physics is present we
  need to be conservative in assigning errors
• For exclusives Unquenched Lattice QCD has 10
  errors to which I add a 20 quenching error (22
  total error)
• For inclusive calculations I take the
  experimental statistical systematic errors,
  modeling errors add duality error, mb error
  WA error, the sizes of which depend on the phase
  space region

35
My Value Error on Vub

• Exclusives average CLEO BaBar rln, CLEO pln
• Add in quadrature 10 (models) and 20
  (quenching) for theory error, get 3.52
  0.270.78
• Lepton endpoint Average CLEO BaBar using
  experimental errors, 4.280.22. Add in
  quadrature 0.44fu0.16OPEsg 0.64twist 1.3WA
  0.35duality0.22mb get 4.280.221.44
• Dont use results with just Mx cut due to parton
  model singularity and resulting uncertain error
• Use Belle annealing result since both Mx and q2
  cuts are made (add duality error of 7, WA
  error of 7 increase theory error for mb to 10)
  gives 3.960.470.56
• Now average these values taking into account
  common theoretical error (3.90 0.160.53)x10-3
  Very subjective!

36
Bd Mixing in the Standard Model

• Relation between B mixing CKM elements
• F is a known function, hQCD0.8
• BB and fB are currently determined only
  theoretically
• in principle, fB can be measured, but its very
  difficult, need to measure B ?ln, find fB
  Vub
• BaBar B(B?tn) lt 4.1x10-4, fB lt 390 MeV (Cartaro)
• Current best hope is Lattice QCD

37
Bs Mixing in the Standard Model

• When Bs mixing is measured then we will learn the
  ratio of Vtd/Vts which gives the same essential
  information as Bd mixing alone, but with much
  less theoretical error
• Vtd2/ Vts2(1-r)2h2 ? fBsBBs2/fBdBBd2
• Circle in (r,h) plane centered at (1,0)
• Lattice best values for

Partially unquenched JLQCD Aoki et al.

To test lattice measure fDs/fD at CLEO-c
38
Upper limit on Dms

• P(BS?BS)0.5X
• GSe-GSt1cos(DmSt)
• To add exp. it is useful to analyze the data as a
  function of a test frequency w
• g(t)0.5 GS
• e-GSt1Acos(wt)

DmS lt 14.4 ps-1 _at_ 95 cl
39
CDF measures Bs?Dsp-needed for Bs mixing
o

• For fs/fd 0.270.03, B(Bs)/B(Bd)1.60.3

40
Charm mixing Another Place New Physics Could
Appear

• A place to look for new physics (also rare decays)

41
Generic tests for New Physics Separate Checks

• Use different sets of measurements to define apex
  of triangle
• (ala Peskin)
• Also have eK (CP in KL system)

Bd mixing phase
Magnitudes
Bs mixing phase
Can also measure g via B-?DoK-
42
Current Status

• Constraints on r h from CKM Fitter using Rfit
• The theory parameters are allowed to have equal
  probability within a restricted but arbitrary
  range
• Therefore large model dependence for Vub/Vcb, eK
  and Dmd, smaller but significant for Dms/ Dmd.
  . The level of theoretical uncertainties is
  arguable

43
Another Method

• See talk by Eigen for another good method Scan.
  Fits are done spanning theoretical space. Points
  allowed if c.l. gt 5.
• New handling of Bs mixing
• Contrast with Bayesian method, adding exp.
  theory errors in quadrature

44
Rare b Decays

• A good place to find
• new physics

• New fermion like objects in addition to t, c or
  u, or new Gauge-like objects
• Inclusive Rare Decays such as inclusive b?sg,
  b?dg, b?sll-
• Exclusive Rare Decays such as B?rg, B?Kll-
  Dalitz plot polarization

g,
ll-
SUSY examples
45
Inclusive b?sg

• CLEO B(b? sg)
• (2.850.350.23)x10-4
• ALEPH, Belle Babar,
• Average(3.280.38)x10-4
• Theory Lowest Order

CLEO
CLEO
Theory NLO (3.570.30)x10-4
46
Implications of B(b? sg ) measurement

• Measurement is consistent with SM
• Limits on many non-Standard
• Models minimal supergravity,
• supersymmetry, etc
• Define ala Ali et al.
• Ri(ciSMciNP)/ciSM i7, 8
• Black points indicate various New Physics models
  (MSSM with MFV)

SM
47
B?K()ll-

• Belle discovered of Kll-
• They see Kmm-
• BaBar confirms in Kee- Kmm-
• Belle u.l. on Kll- lt1.4x10-6
• BaBar sees some evidence at 3s
• 18 events using 82 fb-1

48
B?Xsll-

• Heff f(O7, O9, O10)

• Belle finds
• B(b? sll- )
• (6.11.4 )x10-6
• Must avoid J/Y, Y
• Important for NP, but not
• nearly enough data. In agreement with SM, Ali
  et al.

1.4 -1.1
hadronic mass
dilepton mass
49
Hadronic B Decays Phase shifts can be large

• Known to be large in D decays
• Bo?Dopo observed by CLEO Belle BaBar
  construct isospin triangle with Bo?Dp-
  B-?Dop-, find strong phase shift between 16.5o -
  38.1o
• Bo?Ds-K observed by Belle BaBar at 4x10-5
  level, evidence for W exchange diagram? or is it
  rescattering from Dp- ? (phase shifts
  rescattering go hand-in-hand)
• Final state rescattering plays a role in
  interpreting the fundamental CP violating angles
  from charmless two-body decays

50
Factorization

• Simple concept amplitude is a
• product of two hadronic currents
• Compare with semileptonic decays
• a1 calculated in BBNS theory as 1.05, a1
  measured

New CLEO, rest PDG 2003 (thanks to Karl
Ecklund)
51
B?Kp pp

• Can have both tree loop diagrams in hp-
• Others

If t?s then K-
If (u, s) then K-
52
QCD factorization for B?(p or K) p (BBNS)

• Amplitude can be taken as involving b spectator
  plus part from virtual W- with corrections
  parameterized in a series S(LQCD/mb)n thus a
  theory
• Can compute amplitudes for Tree Penguin
  diagrams
• BBNS M. Beneke, G. Buchalla, M. Neubert C.T.
  Sachrajda, Nucl.Phys. B606 (2001) 245-321

53
B?Kp pp Summary
54
QCD Factorization (BBNS)

• Strong phase shifts are included
• g is limited to be between 58o-80o at 2s, unless
  there is new physics

Statistics? New Physics? Bad Theory?
55
Consistent with other Measurements

• No evidence for New Physics
• Alternative model by Hou et al explicitly
  includes final state rescattering (see talk),
  finds g90o-100o, large rescattering phases

56
Revelations about QCD

• Since QCD is so relevant to extract quark
  parameters, let us see how well its doing in
  other areas
• New narrow Ds()po states
• Double Charm Baryons Selex (no time)
• The hC(2S) and implications for Potential Models
  Belle, BaBar, CLEO (no time)
• The Upsilon D States - CLEO (no time)
• D states in B decays - Belle (no time)

57
Narrow Ds States

• Ds (c s in l1 states) predicted Jp 0, 1,
  1 2. One 1 2 previously seen, these decay
  into D()K, are relatively narrow. Others are
  also predicted to be above D()K threshold and
  have large 200 MeV widths
• BaBar Narrow peak in Dspo
• mass distribution. Mass is
• 2316.80.43.0 MeV, width
• consistent with mass resolution
• 9 MeV
• Lighter than most potential
• model predictions. Mass is
• 40 MeV below DK threshold

BABAR 91 fb-1
58
Possible Explanations

• DK molecule Barnes, Close Lipkin hep-ph/0305025
• Ordinary excited cs states Ds, narrow
  because isospin is violated in the decay (is only
  way for hadronic decay to occur since its below
  DK threshold - see Cho Wise). Use HQET chiral
  symmetry to explain. Bardeen, Eichten Hill
  hep-ph/0305049. Also Nowak, Rho Zahed
  hep-ph/0307102 (cs states are all I0, p is I1)
• Colangelo De Fazio hep-ph/0305140 Godfrey
  hep-ph/0305122 ask us to look for radiative
  decays such transitions support cs
• Van Beveran Rupp explain in terms of unitarized
  meson model hep-ph/0305049
• Bali says lattice cannot accommodate these as cs
  states hep-ph/0305049, however this claim is
  disputed, Dougall et al hep-ph/0307001
• Dai, Huang Zhu hep-ph/0306274 show that the
  masses can be obtained using QCD sum rules
• Browder et. al, mixture with 4-quark state above
  DK threshold hep-ph/0307054

59
CLEO Sees Two States
2.32 GeV
2.11 GeV
Ds p0

• Confirms the BaBar observation of Ds(2317) (13.5
  fb-1)
• s MeV
• Detector resolution 6.00.3 MeV
• 16520 events in peak

60
CLEO finds new state near 2460 MeV
2.46 GeV

• See 2nd state decaying
• into Dspo, at 2460 MeV
• s 6.11.0 MeV
• Detector res 6.60.5 MeV
• 5510 events in peak

Ds p0
61
Can these states be reflections of other states?
each other?

• No known source has been thought of to create
  these peaks
• However, since the mass differences are both 350
  MeV, they can reflect into each other!
• Which is feeding which and how much?
• CLEO has two methods
• First method MC simulation of feeddown feedup
• Second (a) Fit 2317 peak to two Gaussians,
• (b) Perform Ds sideband subtraction and fit
  2460 peak

62
Alternative Way to Estimate Dspo Signal fit to
two Gaussians(CLEO)

• Use two Gaussian functions whose means and widths
  are allowed to float.
• The fit is consistent with the existence of a
  narrow signal and a broader feed-down
  contribution.
• The feed-down not only broadens the peak, but
  also shifts the center position. Using this fit
  CLEO extracts a more precise mass difference.

• M(Dsp0) - M(Ds) 350.41.21.0 MeV
• Wide Gaussian peaks at 344.9 MeV

63
Belle Confirms Both States

Ds sidebands
M(Ds p0 )-M(Ds)
M(Ds p0 )-M(Ds)
(See Seusters talk)
64
Other Charge States Not Seen

• CLEO Ds p- or Ds p not seen at level more than
  factor of 10 lower than Ds p- speaks against
  molecule. Also

CLEO

2317
CDF
65
Belle B?DDsJ

• Sees Radiative decay

DE sidebands
Ds(2317)?poDs
Ds(2460)?poDs
Ds(2460)?gDs
DDs 1
Factorization implies similar B to DDs, thus
there are indications of not pure cs states, see
Chen Li hep-ph/0307075 Datta Odonnel
hep-ph/0307106 Cheng Hou hep-ph/0305038
66
Belle Examines JP of DsJ(2460)
Helicity angle

• Use B?DDsJ(2460),
• DsJ(2460) ?gDs
• Helicity angle consistent with 1 (angle of Ds
  in DsJ restframe wrt to DsJ direction in B rest
  frame)

2
1
67
BaBar on the Ds(2460)

• BaBar presented Dspo peak near 2460 in their
  original publication, but did not claim a real
  signal
• Recently they agreed that it was a signal. Their
  feedup rate as a fraction of the 2460 signal size
  is 50 compared with CLEO (20) Belle (30)

Fit for mass to sideband subtracted data
sideband background
DMM(Dspo)-M(Ds) (GeV)
68
Summary of Narrow Ds (Masses Widths)
0
1

• DmM(0)-M(1)2.11.4 MeV
• Width G lt 7 MeV (both states) CLEO

69
Conclusions on Dss

• BaBar discovered a narrow Dspo state near 2317
  MeV, confirmed by CLEO
• CLEO has observed a new narrow state near 2460
• Belle confirms both states, observes radiative
  transition of 2460 and sees states in B decay
  BaBar also confirms 2460.
• The mass splittings are consistent with being
  equal (2.11.4 MeV) as predicted by BEH (HQET
  Chiral Symmetry) if these are the 0 - 1 states
• Radiative Decays seen at expected levels for cs
  state
• Seen modes and u.l. are consistent with these
  assignment except 1? Dspp- is above threshold
  for decay, predicted to be 19 but is limited to
  lt8.1 _at_ 90 c. l. (CLEO)
• Other explanations, e.g. csgt D()Kgt are
  possible
• Factorization in B?DDsJ seems to be in
  contradiction with cs

70
Conclusions

• There have been lots of surprises in Heavy Quark
  Physics, including
• Long b Lifetime
• Bo Bo mixing
• Narrow Ds states
• Now finding the effects of New Physics in b c
  decays would not be a surprise - we expect to do
  it!

71
Backups Follow
72
New versus old CLEO BaBar Moments
73
New BaBar Delphi Moments

• BaBar data now agree with theoretical shape (blue
  curve) using HQE input, although CLEO using b?sg
  moment and extrapolating from 1.5 GeV/c is not in
  great agreement

74
Comparison of Bayesian Scan Method
Bayes
Scan
Constraint on sin2b included in both Scan and
Bayesian fit of M. Ciuchini et al., JHEP 0107,
13 (2001)