Measurement of D0 production in pp

1
Measurement of D0 production in pp

• Andrea Dainese
• University of Padova

2
Outline

• One step back study of feed-down from beauty
• (b ? B ? D0), important also for systematic
  error
• Estimation of statistical error on dN(D0)/dpT
• Estimation of systematic errors
• contribution from beauty
• uncertainty on D0 ct
• MC corrections uncertainty on B.R. to Kp
• Combination of errors and error on total cross
  section

PHYSICS PERFORMANCE SENSITIVITY TO
3
Feed-down from Beauty (1)

• At LHC energies
• Fraction of D0 from chain b ? B0/B ? D0 not
  negligible
• Standard NLO pQCD Pythia fragmentation PDG
• Important to use the correct ratio D from b /
  D from c
• Selection may change the ratio
• Results have to be eventually corrected for
  feed-down
• estimate systematic error
  from b cross section

0.64
0.0017
0.0196
4
Feed-down from Beauty (2)

• How D0 signal was generated
• generate standard pp minimum bias with Pythia
• keep only those events that have a D0
• weight D0s according to their pT in order to
  reproduce distribution given by NLO pQCD
• In that sample
  factor 4 too high!
• All candidates have been re-weighted
• c ? D to reproduce pT distr. by NLO
• b ? D to have correct ratio (5.5) AND reproduce
  THEIR pT distr. by NLO (harder than for c ? D !)

5
Feed-down from Beauty (3)
After event reconstruction (tracking
vertexing) D0 from b / D0 from c
5.8 Different pT
slopes After selection (impact parameter
pointing angle) D0 from b / D0
from c 16
6
Feed-down from Beauty (4)
Selection on pointing angle expected to suppress
D0 from b (they point to the decay vertex of the
B and not to the primary vertex)
2ltpTlt3 GeV/c
7
Feed-down from Beauty (5)
Also selection based on impact parameter
enhances beauty
8
Feed-down from Beauty (6)
impact parameter distribution after selection
new cut d0lt 500 mm can suppress beauty
9
Results

• Essentially unchanged w.r.t. to last shown (March)

10
Binning Smoothing in dN/dpT

• Optimize binning in the range 0-14 GeV/c bin
  size increasing with pT
• good precision on measurement of pT slope
• small statistical error up to high pT
• Fit pT distributions to remove fluctuations due
  to reduced statistics of simulation at high pT
  (especially for Bkg)
• Use fitted function to obtain smooth dN/dpT for
  S and B

11
Statistical Error (1)

• Rel. Statisitical error on S expected to be
  1/Significance
• S (SB)observed Bestimated
• error on (SB)observed sqrt(SB)
• error on Bestimated negligible (uses info
  from wide range mass distr.)
• error on S error on (SB)observed
  sqrt(SB)
• relative error on S sqrt(SB)/S
  1/Significance
• This has been checked by generating and fitting a
  large number of invariant mass distributions,
  using as input parameters the results of the
  simulation, namely
• statistics for S and B VS pT
• width of D mass distribution VS pT
• slope of exponential bkg VS pT

12
Statistical Error (2)

• Fit with expo gaus
• with 4 free parameters
• expo slope
• gaus integral (S)
• gaus mean
• gaus sigma
• (total integral fixed)
• Many (10,000) iterations
• study resolution on S
• (statistical error)
• study robustness of fit

13
Systematic Errors

• When the data are corrected for efficiency,
  acceptance, etc systematic errors are
  introduced
• Main sources of systematic errors have been
  considered

CORRECTION ERRORS INTRODUCED
subtraction of contribution from b uncertainty on b cross section
selected ? reconstructed (MC correction) error on D0 ct
correction for PID efficiency errors from MC description of ALICE (TOF eff. contaminations)
extrapolation from D0 reconstructed to D0 produced in ylt1 errors from MC description of ALICE (tracking eff. resolutions)
from D0 ? Kp to total D0 error on B.R. D0 ? Kp
14
Systematic from b subtraction

• Fraction of D0 from b will be estimated from MC
  and subtracted to obtain D0 from c
• Input to MC
• Beauty cross section in pp will be measured by
  all 4 expts
• Now pessimistic case ? use theoretical
  uncertainty (80)
• Error on N(c ? D0) 0.8 ? N(b ? D0)
• Relative error 0.8 ? N(b ? D0)/N(c ? D0) ?
  0.8?10 8

presentth. uncertainty
15
Systematic from ct (123.4 ? 0.8)mm

• From selected to reconstructed D0
• For each D, calculated equivalent true value of
  ct L?M/p

16
Errors combined
dN(D0)/dy for y lt 1 and pT gt 0.5 GeV/c (91
of pT gt 0) statistical error 3
systematic error 13
from b 8
MC correction 10 B.R.
2.4
Note Error from MC may be lower than 10 and
dep. on pT Error from b may be much lower after
direct measure Error on BR doesnt alter pT shape
17
Comparison with CDF-Run II

• CDF Run II preliminary measurement of charm
  cross section using the Secondary Vertex Trigger
• sigma(D0, pT gt 5.5 GeV/c) (13.3 0.2 1.5)
  µb
• 
  1.5 11
• We could measure the charm cross section in pp _at_
  LHC
• from pT gt 0.5 with the same
  precision/accuracy

18
Conclusions

• Feed-down from b studied important for
  systematic error
• Statistical errors on pT distribution estimated
• Main systematic errors considered
• Physics performance on charm X section
  measurement in pp
• statistical error 3
• systematic error 13 (only indicative value)
• Systematic error from MC corrections will
  dominate
• Next steps
• try D0 impact parameter to measure (b ? D0) /
  (c ? D0) a la CDF
• sensitivity for comparison with NLO pQCD and/or
  Pythia
• error analysis for Pb-Pb
• play a bit with comparisons (RAA (pT) .)