Analysis of charmonium production at fixedtarget experiments

1
Analysis of charmonium production at fixed-target
experiments

• J. Spengler, DESY, Hamburg
• This analysis was performed by
• F. Maltoni several members of the HERA-B
  collaboration
• Published in Phys. Lett. B 638 (2006) 202

2

• Outline
• Motivation
• J/?? cross section measured at HERA-B
• Compilation of all pA J/? and ?(2S) cross
  sections
• The NRQCD approach
• Fit results?

3

• 1. Motivation
• Expt. view HERA-B uses a dilepton trigger
• B ? J/?, ?c ? J/?, ?(2S) ? J/? are
• normalized to J/? ? dileptons to reduce sys.
• errors.
• Thus a precise J/? cross section is required
• Measure J/? in minimum bias data
• Global fit to all J/? cross section measurements
• Theo. view test universality of non-perturbative
• matrix elements of quarkonia

4
2. J/?? cross section at HERA-B

• Interactions of 920 GeV protons on C, Ti and W
  targets (180M events)
• Minimum bias trigger with gt97 efficiency
• pA ? J/? X with J/? ? ee- or ??-
• Visible fraction of total cross section 63
• Expect low sys. error result dominated by stat.
  error despite huge statistics

5
5713 J/? ? ee-
10012 J/? ? ??-
6

• Combine both final states for each target
• Determine pA cross sections for A C, Ti, W
• Determine pN cross section by fitting
• using ? 0.960.01 at xF 0 (E866 result).
  Most
• precise measurement available.
• Final result ?pN 6637446 nb/nucleon
• Published in Phys. Lett. B, in press

7

• 3. Compilation of J/? and ?(2S) cross sect.
• Updates applied to all data
• Assume the same target mass dependence A?
• Assume symmetric xF dependence
• Update for PDG04 branching ratios
• Combine different targets of same experiment
• Assume that sys. error quoted includes all
  contri-butions (Br, A-dependence)

8
Mid-rapidity and total cross sections in pN
reactions
Several measurements are not compatible. Detailed
tables can be found in the paper.
9

• 4. The NRQCD approach
• NRQCD expresses the cross section for a
  quarkonium
• state in pp interactions in NLO as product of
• short-distance coefficient long-distance matrix
  element.
• Here we need the direct production ?D of J/?, ?cJ
  and
• ?(2S).

10

• Choices to be made
• Scale parameter ?02mc with mc1.5 GeV
• ?0 ?F(factorization) ?R(renormalization)
• PDF sets MRST2002nlo and CTEQ6m
• Color-singlet matrix elements from potential
  model calculations
• Color-octet matrix elements extracted from CDF
  data (based on LO calculations only)
• Introduce free scaling parameters for S-wave
  color-octet ME
• ?J/? and ??(2S)
• fraction of color-octet contribution required
  relative to CDF
• results

11
Matrix elements used for fit
?O8?(3PJ)? (2J 1) ?O8?(3P0)? ?O8?cJ(3S1)?
(2J 1) ?O8?c0 (3S1)? ?O8?cJ(3PJ)? (2J 1)
?O8?c0 (3P0)?
12

• 5. Fit results
• Data available 21 results on J/?, 3 on ?(2S)
• and 5 on R?.
• 1. step fit scale factor x common to all
  states
• ?F?Rx??o for both PDF sets
• 2. step fit ?J/? and ??(2S) for both PDF sets

13

• Comments
• Large ?2/dof ? error on ? scaled (PDG)
• Differences in ? due to different scaling factors
• Stability of fits was tested by
• Excluding results with bad partial ?2
• Changing the cms-energy range
• Excluding targets with Agt14
• Fit results very stable. Only significant
  changes in
• ? due to PDF and scaling factors
• Select MRST2002nlo as
• baseline (smaller ?2)

14
Open symbols not used in fit
15
NRQCD prediction of d?/dy at y0 12 (out of
23) data points partially correlated with
previous fits
16

• Conclusions
• For both PDFs and all fits we obtain
• 0.01 lt ?J/? lt 0.31 and 0.02 lt ??(2S) lt 0.14
• Required color-octet contribution about 10 of
  that for
• Tevatron data. Supports HERA ep data which do not
  need
• large octet contributions.
• Difference too large to be due to LO calc. for
  Tevatron analysis
• Missing NLO calc. of short-distance coeff. for
  color-singlet production can decrease octet
  contri. as observed in photoproduction