Charm Meson Pair Cross Sections

1

• Charm Meson Pair Cross Sections
• Burkard C. Reisert
• Fermilab

Picture of FNAL or Oxford

• Heavy Quark Production
• Instrumentation Tevatron, CDF
• DD- Correlation Analysis
• - Methodology for the Measurement
• - Result Charm Pair Cross sections
• Conclusion

2
Motivation

• Heavy Quark Production
• Test of Quantum Chromodynamics at the transition
  from perturbative to non-perturbative regime
  - masses of c, b quarks provide hard
  (?) scale for
• QCD calculations
• Beauty and charm cross sections in pp collisions
  are measured to be larger than expected (?next
  slides)
• Important engineering number for other
  CDF measurements and future experiments
  e.g. background to top measurements
  and
• Higgs/SUSY searches, tagger dilution
  for Bs
• mixing, production rates of beauty
  charm
• hadrons, etc

_
3
Beauty Cross Sections at the Tevatron
Integrated Cross Section for b-Quark Production
B Meson Differential Cross Section
hep-ex/0008021
hep-ph/0111359
y(B)lt1
Measurements in many channels lower than
theoretical predictions
4
Inclusive Charm Meson Cross Sections
CDF Run II (L5.8 pb-1)
CDF Run II (L5.8 pb-1)
FONLL
D0
Cacciari et al., JHEP, 0500, 1998
CDF Collab., PRL 241804

• Inclusive charm (D0, D, D, Ds) cross
  sections was one of the first CDF Run II results
• factor 2 higher than expected
• progress in theory reduced deviation
  
• measurement systematically limited

Kniehl et al., hep-ph/0508129
5
Charm Production

• Example Feynman Diagrams
• Calculations with massive quarks
• incoming charm
• charm in final state
• parton shower
• All shown example graphs can be interpreted as
  higher order flavor creation

Flavor Creation (a2s)
Flavor Creation (a2s)
Flavor Excitation (a3s)
Gluon Splitting (a3s)
Gluon Radiation (a3s)
Gluon Radiation (a3s)
Interference Term (a4s)
Interference Term (a4s)
6
Heavy Quark Production Leading Order Picture
PS Parton Shower

• Search for second charm particle in the event
• Correlations between first and second charm
  particle give access to detailed understanding of
  the underlying production mechanism

7
Leading Order Generator Level Study
MC cc Pythia
Dfltp/2 Gluon Splitting dominated region (toward
side)
Dfgtp/2 Flavor Creation dominated region (away
side)
Flavor Creation (FC)
Gluon Splitting (GS)
Flavor Excitation (FE)

• Pair selection on Generator (Pythia) level with
  realistic cuts
• Trigger-side D0 PTgt5.5 GeV/c,yD lt 1, PT (K,p)
  gt 2GeV/c, SK,pPTgt5.5GeV/c
• Probe-side D PTgt4.5 GeV/c, yD lt 1,
  PT(K,p) gt 1 GeV/c, PT(p) gt 0.4GeV/c

8
Tevatron
New world record
ISR _at_ vs62 GeV

• Run II Physics Goals
• Properties of top quark
• Precision Electroweak Physics
• CKM, Bs Mixing
• Searches for new phenomena
• Tests of QCD

9
Collider Detector at Fermilab
Muon System
New
Old
Central Calorimeter
Partially New
Solenoid
Fwd Calorimeter
Plug Calorimeter
Muon

• CDF has
• Tracking with excellent momentum and impact
  parameter resolution
• High-bandwidth trigger

Time-of-Flight
Drift Chamber
Silicon Microstrip Tracker
10
CDF Trigger DAQ System
CDF Detector
Hardware tracking for pT ?1.5 GeV
1.7 MHz crossing rate
Muon-track matching
Dedicated hardware
Electron-track matching
L1 trigger
Missing ET, sum-ET
32 kHz L1 accept
Silicon tracking
Hardware CPU
Jet finding
L2 trigger
Refined electron/photon finding
1000 Hz L2 accept
Refined Muon-track matching
Linux farm (gt200 CPUs)
L3 farm
Full event reconstruction
100 Hz L3 accept
The trigger is the key to heavy flavour physics
at hadron colliders
disk/tape
11
Triggers forBeauty Charm Physics
Two Track Trigger PT(trk)gt2 GeV IP(trk) gt100
mm Fully hadronic modes
Displaced track(s) lepton (e, m) IP(trk)gt120
mm PT(lepton) gt4GeV Semileptonic modes

• Di-Muon (J/y)
• PT(m) gt 1.5 GeV
• J/y, y(2S) modes
• Down to low PT(J/y)
• ( 0 GeV)

• Bs mixing (semileptonic)
• Tagging, lifetime

• y(2s), X(3872)? J/ypp (quarkonia)
• Bs?J/yf, Bu,d J/yKs() Lb?J/yL (masses,
  lifetimes, mixing calibration)
• Bs,d?mm (rare decays)
• Bc(lifetime B?J/ylX, mass B? J/yp)

• 2-body charmless decays (B0,Bs,Lb)
• Bs mixing (hadronic)
• Charm physics
• - inclusive cross sections
• - D0?Kp,pp,KK
• Heavy quark production

12
Charm Meson Pair Analysis Approach

• Simple experimental approach
• Find triggered charm meson D0,
  D, D, Ds,
• Look for second charm meson probe D
• Correlation variable Df gives access to
  underlying production mechanism
• Measured inclusive cross section
  provides normalization

Primary Vertex
13
Inclusive Charm Mesons D0D

• Inclusive triggered D Mesons reconstruction as in
  inclusive Charm Meson Analysis
• Trigger D0?Kp
• flight distance Lxy 500mm,
• impact parameters d0(K)d0(p)lt0
• Trigger D?(Kp)D0p
• add p to triggered D0
• Probe D?(Kp)D0p
• Keep selection simple
• Lxy(D0)gt0,
• Dm provides handle for clean up

14
Inclusive Charm Mesons D Ds
selection as in inclusive Charm Meson Analysis,
with some improvements Trigger side D?Kpp
maxDz0(K,p,p) 5cm veto Dm(Kpp)-m(Kp)lt0.18
GeV/c2 d0d0(maxDf tracks)lt0 Lxy gt1000mm
Trigger side Ds?(KK)fp maxDz0(K,K,p)
5cm m(KTrKTr)raw lt 1.06GeV/c2 (veto Ks?pp, ps
fake Ks) tight m(f) window Lxy 500mm
15
D0TrigDProbe Pairs
2dim SB-subtraction by weighting events In mass
plane of Trigger D and soft p tagged Probe D0
16
D0TrigDProbe Yields
2dim SB-subtraction by weighting events In mass
plane of Trigger D and soft p tagged Probe D0
17
DsD-
DD-
DD-
18
Charm Meson Pairs vs. Df (Raw Yields)
DD-
D0D-
DD-
DsD-
19
Challenges of the Charm Meson Correlation analysis

• Convert raw Df yields into a cross section
• Correct for physics background
• DD- pairs from B hadron decays
• Correct for detector acceptance and efficiency
  effects
• Control systematic effects over a data taking
  period of 4 years
• Technical challenge efficient analysis of huge
  inclusive D samples

20
Prompt Charm Production Cross Sections

• Inclusive production (already measured by CDF)
• Pair production cross section
• Relative pair Cross section
• uses

Ni Number of observed charm mesons in ith PT
bin Nij Number of observed charm meson pairs
with D1 in ith PT bin and D2 in jth PT bin
fPD fraction of prompt D mesons fPDD fraction
of prompt pairs ei trigger and reconstruction
efficiency eij pair trigger and reconstruction
efficiency ej probe-side reconstruction
efficiency ??dt integrated luminosity B
branching fraction e.g. D0?Kp
21
Realistic Monte Carlo

• Challenge
• Produce sufficiently large event sample in a
  realistic simulation
• Means to face this challenge
• - Use of detailed event generator (Pythia),
• - select potentially interesting events at the
  earliest possible stage,
• - force D meson decays
• Some details
• Generator Pythia
• LO, generic QCD 2?2 processes, mass less scheme
• Use parameter set, which best described the
  underlying event in CDF Run I
• (Rick Fields Tune A on CTEQ5L PDF set)
• Generator level cuts PTgt5GeV/c, hhardlt1.3
• Only keep events with charm (bottom) quark
  cc (bb) with PTgt4GeV/c,hlt1.3
• Redecayer EvtGen Package
• force D0?Kp
• Detector and Trigger Simulation, Reconstruction
  for selected events only
• Trig D0 PTgt5GeV/c, hlt1.25, Probe D0
  PTgt3GeV/c, hlt1.25

22
Data vs Monte Carlo Comparison
23
Efficiency Factorization

• ? Efficiency of D0Trig cancels if
• e(ij) e(i) e(j)
• index i PT(D0Trig), index j PT(DProbe)
• ? Factorize overall efficiency e into
• acceptance for primary event vertex avtx
• geometric and kinematic acceptance aNAcc/NGen
  and
• reconstruction (trigger) efficiency
  eNRec,Sel/NAcc
• pair e(ij) avtxa(ij)e(ij)
  avtxa(i)e(i)a(j)e(j)
• inclusive e(i) avtxa(i)e(i)
• ? For factorization of trigger and probe
  efficiencies we need to show
• 
  

Binned in PTGen(D0Trig)
PTGen(DProbe)
generator level
reconstructed events
24
Acceptances
Trig
Probe
DD
DTrig Acceptance independent of pT (Probe) and
vice versa
25
Acceptance Factorization
Acceptance factorization is valid to a level
better than 1
26
Efficiencies
Trig
Probe
DD

• Use sideband subtraction to count
• reconstructed trigger D, probe D
• and DD Pairs
• All efficiencies are flat after
• applying acceptance cuts.

27
Efficiency Factorization
Efficiency factorization is proven to work to a
level better than 2
28
Measurement of Pair Cross Section
published CDF measurements of inclusive cross
sections
develop methods to extract these numbers
branching fraction from PDG
apply methods developed for inclusive cross
section measurements to extract these numbers
obtain parameterization from simulation
29
Measurement of Pair Cross Section
published CDF measurements of inclusive cross
sections
develop methods to extract these numbers
2d sideband subtraction
branching fraction from PDG
binned log likelihood fit
apply methods developed for inclusive cross
section measurements to extract these numbers
obtain parameterization from simulation
30
Prompt vs. Secondary Charm Production

• Secondary D (i.e. D from B-Hadron decays) are a
  physics background to direct charm production

Secondary Production D has finite impact parameter
Prompt Production D points back to PV
Ip

• Use D Impact parameter (Ip) distribution to
  determine fraction of D from B decays

31
D0Trig Impact Parameter Distribution

• Apply method developed for inclusive
• cross section measurement
• True impact parameter distribution
• obtain template function from fit to
• Generator level Ip of reconstructed triggered D0
  in realistic MC
• Prompt D mesons
• impact parameter distribution
• impact parameter resolution function
• Secondary D mesons impact parameter
  distribution true impact parameter
  ? resolution
• Obtain parameters of resolution function and
  fraction of secondary Ds from fit to data Ip
  (inclusive D0)

32
Fraction of Secondary D0 in Sample of Inclusive D0
Fraction of Secondary D from B decays fb(D0Trig)
fb is correlated to fraction of Gaussian core of
Ip resolution function

• Account for fit model dependencies in systematic
  uncertainties

33
Secondary Pair Production
Single B decaying into Pair of charm mesons
Both B hadrons decay to charm mesons

• Prompt fraction extraction uses Ip of unbiased
  probe side D

34
Extraction of N(prompt pairs)
Small statistic sample develop counting method

• Count sideband subtracted candidates in central
  and tail region of Ip distrib. (N cent, N Tail)
• Extract template of true impact parameter
  distribution of trigger (probe) D of selected
  pairs
• Convolute true Ip with resolution obtained from
  inclusive D sample
• Calculate fractions of prompt and secondary D
  expected to be found in central and tail region
  (fPCent, fBTail)
• From the above inputs calculate NPr
• use Ip(D0Trig) (default) or
• Ip(DProbe) (cross check)

35
Extraction of Prompt D Production

• Extract prompt/secondary fractions from Ip of
  probe side D0 fb 34
• Cross checks Ip, Lxy of trigger/probe D and
  DD- pair

36
Measurement of Charm Meson Pair Cross Section
published CDF measurements of inclusive cross
sections
DD Pair yields and Prompt fraction
branching fraction from PDG
Inclusive Charm Meson Yields and prompt
Fractions
Efficiency for fully reconstructed Probe side
D
37
Efficiency Correction
Use realistic MC to study the Probe-side
efficiency. Probe-side efficiency eProbeae is
independent of PT(D0Trig) and Df Parameterize ePr
obe as function of PT(DProbe) Apply correction
as Weight 1/e to pair-candidates before Sideband
subtraction and Npr extraction
38
D0D- Pair Cross Section binned in PT(Trig)

• Error
• stat?sys(fb)
• Common
• Syst. Error
• 15

39
D0D- Cross Section

• Magnitude of D0D- cross section in Pythia is
  about right,
• Collinear production as important as
  back-to-back production
• Pythia under estimates collinear (over
  estimates back-to-back)

40
DD- Pair Cross Section binned in PT(Trig)

• Error
• stat?sys(fb)
• Common
• Syst. Error
• 19

41
DD- Cross Section
Significant shape discrepancy consistent for both
D0D- and DD-
42
Conclusion

• The unprecedented integrated luminosity delivered
  by the
• Tevatron in conjunction with CDFs high bandwidth
  trigger
• allows us to perform a detailed study of charm
  production
• First measurement of charm meson pair cross
  section in a hadron collider environment probes
  open charm production
• We find collinear production to be as important
  as back-to-back production,
• Same result is seen consistently in 2 different
  modes
• Expectation from Pythia has significantly
  different shape
• Charm Physics at the Tevatron is still good for
  some surprises