Measurement of the Heavy Quark cross-sections at CDF

1
Measurement of the Heavy Quark cross-sections at
CDF

• Alberto Annovi - INFN Frascati
• for the CDF collaboration
• annovi_at_fnal.gov

HEP2007 July 19-25th Manchester, UK
2
Outline

• Single b cross section
• Sensitive high order QCD effects
• Correlated bb cross section
• Dominated by LO diagrams

3
Single b production

• Tev Run I b-quark production higher then
  expected from NLO
• Big theoretical effort to understand
  discrepancies
• FONLL PDF Frag. Fun. ?s -gt New prediction
  for Tevatr. in 2002 1
• Theory and measurement are now compatible
• Predictions affected by large uncertainties on
  renormalization and factorization (40)
• Tevatron Run II huge production of b quarks
• First measurement of B-quark production using
  inclusive J/y?2
• Good agreement with FONLL
• Precise measurement of the cross section in
  different modes
• Confirm agreement with FONLL?

1 M.Cacciari and P.Nason, PRL 89, 122003
(2002) 2 D. Acosta et al., PRD71 032001 (2005)
4
?B using Hb???m-D0X, D0 ???K-p

• Clean mode low uncertainties
• SVT trigger lepton displaced track
• Kinematical cuts
• pT(?D0) gt 9 GeV/c ?(?D0) lt 0.6
• Calculate acceptance with MC
• pT-y spectrum from CDF meas. NLO calc.
• Measure efficiencies with data
• Remove resonant bkg12.0 3.2
• Unfolding pT(?D0) --gt pT(Hb) with MC input
• Result

345794 D0
PRD coming soon
(D0)
(D0)
sub sample
(D)
5
?B using B???J/yK, J/y???mm-
8197239 B
PRD75, 012010 (2007)

• Fully reconstructed mode
• Reduce systematics
• No Lxy cut systematics
• Kinematical cuts
• pT(B) gt 6 GeV/c y lt 1
• pT(?) gt 2 GeV/c ?(?) lt 0.8
• pT(K) gt 1.25 GeV/c ?(K) lt 1.3
• Acc. times eff. from MC
• Correct efficiencies with data
• Dominant sys. 6 from lumi

6
Single b production CDF Run II

• 3 separate measurements
• improved precision down to 10
• allow a reliable test of theory
• consistent with FONLL 2
• no need, but still room for new physics 3
• 3 Berger et. al., PRL 86 (2001)

7
bb correlations review 4

• Run I bb xsection increasing with of muons?

Ratio to HVQJET
lt R2b gt ?(data)/?(NLO)
4 F. Happacher et. al., PRD73, 014026 (2006)
8
?bb using inclusive ?? events
2D data

• Select ?? events
• pT 3 GeV/c ? 0.7 5 lt m??lt 80 GeV/c2
• Dimuon events are produced by
• Decays of heavy flavor quark pairs (bb, cc)
• Prompt Drell Yan processes
• Charmonium, bottomonium, K and ? decays
• Misidentification of K and ? mesons
• Sample composition --gt fit 2D impact parameter
• templates from Data (Prompt)
• templates from tuned
• Herwig simulation

1D templates
PRD coming soon
2D template
9
Fit projection
BB 54583-678
10
sb?m,b ?m? sc?m,c ?m
Dimuon purity bb 0.96-0.04 cc 0.81-0.09
Remove fake muons Measure acceptance efficiency

• sb?m,b ?m?? 1549-133 pb
• sc?m,c ?m? 624-115 pb

See also D0D xsec measurement presented at
BEAUTY 2006
11
Data vs NLO
NLO MNR MRST98 Peterson EvtGen
sb?m,b ?m?? sc?m,c ?m?
Data 1549-133 pb 624-104 pb
NLO Peterson 1293-55 pb ?0.006 230-16 pb ?0.06
NLO Peterson 1543 pb ?0.002 383 pb ?0.02
NLO 15 uncertainty

• R1 data/NLO - experimental and BR uncertainties
  only
• R1b?m, b?m???0.006???? 1.20-0.11

12
bb correlations

• High di-muon cross-section NOT observed in Run II

Ratio to HVQJET
lt R2b gt ?(data)/?(NLO)
13
Summary

• Single b quark xsec. in Run I --gt higher than
  theory
• Theoretical prediction went up
• Run II improved precision
• 3 measurements agree within 10
• and compatible with FONLL
• Run I correlated bb xsec increases w/ of
  leptons ?
• Run II ?bb(pp--gt??X) compatible with NLO
• Indicates no issues with bb xsec
• We now have a clear picture of b cross sections!

14
BACKUP
15
Data (sPlots6) vs MC (herwig)
6 M. Pivk and F. R. Le Diberder, NIM A555, 356
(2005)
16
? in prompt
Prompt sPlots
All data
9952-122
17
Data (sPlots) vs MC (herwig)
18
?B using BJ/yK, J/ymm-
Acc. times eff. from MC Correct efficiencies w/
data Dominant sys. 6 from lumi
19
?B using Hbm-D0X, D0 K-p

• Data used whenever possible to measure the
  efficiency of the trigger and detectors.
• 9 independent e measurements.
• 8 using data, COT efficiency from MC
• Unfolding pT(?D0) --gt pT(B) with MC input

Results
(D0)
(D0)
(D)
20
Acceptance and Efficiencies

• calculate acceptances efficiency from MC
• determine efficiency scale factors from DATA
• for relevant acceptance efficiencies

21
Account for fake muons

• Measure fake probability with D0-gtK?
• herwig sim. to estimate of
• ?-K, ?-??pairs within acceptance
• --gt Predict number of fakes
• CMUP with loose/tight ??CMU
• --gt extract number of fakes
• with a 2 unknowns system

360000 D0-gtK?
K fakes CMUP
With fakes removal

mean - maximum RMS bb 0.96-0.04 cc
0.81-0.09
22
Data vs NLO
NLO MNR MRST98 Peterson EvtGen
sb?m,b ?m?? sc?m,c ?m?
Data 1549-133 pb 624-104 pb
NLO Peterson 1293-55 pb ?0.006 230-16 pb ?0.06
NLO Peterson 1543 pb ?0.002 383 pb ?0.02

• R1 data/NLO - experimental and BR uncertainties
  only
• R1b?m, b?m???0.006???? 1.20-0.11
• R1c?m, c?m ??0.06???? 2.71-0.49

23
Heavy Flavor Production at pp
Leading Order Diagrams Dominate bb correlated
prod.
Next to Leading Order Diagrams Essential for
single b prod.
O (as2) flavor creation
sb is inferred from the measurement of the
production rate as a function of pT of the B
hadrons or some of their decay products
Evt Gen
frag
p structure
decay
parton level calculation (NLO, FNLLO)
measured
24
Error summary
Source BB CC
Stat. 1.2 6.4
Fit likelihood 2.9 8
Fake removal 4.2 11.6
Acceptance 3.2 3.9
Luminosity 6 6
Total 8.9 16.7