Hard Diffraction at D

1
Hard Diffraction at DØ During Run I

• Michael Strang
• DØ Collaboration / Fermilab
• University of Texas, Arlington

• Central Rapidity Gaps (Hard Color Singlet
  Exchange)
• Forward Rapidity Gaps (Hard Single
• Diffraction)

2
DØ Detector
(nl0 tiles in L0 detector with signal 2.3 lt
h lt 4.3)
beam
L0 Detector
End Calorimeter
Central Calorimeter
EM Calorimeter
Central Drift Chamber
(ntrk charged tracks with h lt 1.0)
Hadronic Calorimeter
(ncal cal towers with energy above threshold)
Central Gaps EM Calorimeter ET gt 200 MeV h
lt 1.0 Forward Gaps EM Calorimeter E gt 150
MeV 2.0 lt ? lt 4.1 Had. Calorimeter E gt 500
MeV 3.2 lt ? lt 5.2)
3
Particle Multiplicity Distributions
4
Measuring CSE
5
Hard Color Singlet Studies
QCD color-singlet signal observed in 1
opposite-side events (p )
f
Dh
jet
jet
h
Publications DØ PRL 72, 2332(1994) CDF
PRL 74, 885 (1995) DØ PRL 76, 734 (1996)
Zeus Phys Lett B369, 55 (1996) (7) CDF
PRL 80, 1156 (1998) DØ PLB 440, 189 (1998)
CDF PRL 81, 5278 (1998)

• Newest Results
• Color-Singlet fractions at
• ?s 630 1800 GeV
• Color-Singlet Dependence on
• Dh, ET, ?s (parton-x)

6
Data Selection
?s 1800 GeV high, medium, low jet ET
triggers ?s 630 GeV low ET trigger
Four data samples collected during 1994 - 1996

• Require centered event vertex
• Cut events with spurious jets
• Require single interaction
• Leading jets with h gt 1.9
• Dh gt 4.0 (Opposite-side events)
• ET cut
• ET gt 12 GeV (low-ET 630, 7k events)
• (low-ET 1800, 48k events)
• ET gt 25 GeV (med-ET 1800, 21k events)
• ET gt 30 GeV (high-ET 1800, 72k events)

(Also same-side control samples at both CM
energies using L0 trigger to suppress SD signal
for background studies)
7
Measurement of fs
f
Count tracks and EM Calorimeter Towers in h lt
1.0
Dh
jet
jet
h
High-ET sample (ET gt 30 GeV, ?s 1800 GeV)
fS color-singlet fraction (Ndata- Nfit)/Ntotal
ET gt30 GeV
fS 1800 0.94 ? 0.04stat ? 0.12sys
(Includes correction for multiple interaction
contamination. Sys error dominated by background
fitting.)
8
630 vs. 1800 Multiplicities
Jet ET gt 12 GeV, Jet h gt 1.9, Dh gt 4.0
Opposite-Side Data
Same-Side Data
1800 GeV
ncal
ncal
ntrk
ntrk
630 Gev
ncal
ncal
ntrk
ntrk
fS 1800 (ET 19.2 GeV) 0.54 ? 0.06stat ?
0.16sys fS 630 (ET 16.4 GeV) 1.85 ?
0.09stat ? 0.37sys
630
R1800 3.4 ? 1.2
9
Color Singlet Models

• If color-singlet couples preferentially to
  quarks or gluons, fraction depends on initial
  quark/gluon densities (parton x)
• larger x ? more quarks
• Gluon preference perturbative two-gluon models
  have 9/4 color factor for gluons
• Naive Two-Gluon model (Bj)
• BFKL model LLA BFKL dynamics
• Predictions
• fS (ET) falls, fS (Dh) falls (2
  gluon) / rises (BFKL)
• Quark preference
• Soft Color model non-perturbative
  rearrangement prefers quark initiated processes
  (easier to neutralize color)
• Photon and U(1) couple only to quarks
• Predictions
• fS (ET) fS (Dh) rise

10
Model Fits to Data
Using Herwig 5.9
?s 1800 GeV
Soft Color model describes data
11
Fit Results
Apply Bayesian fitting method, calculate
likelihood relative to free-factor model
(parametrization as weighted sums of relative
fractions of quarks and gluons in pdf)
Color factors for free-factor model Cqq Cqg
Cgg 1.0 0.04 0 (coupling to quarks
dominates)
Data favor free-factor and soft-color
models single-gluon not excluded, but all other
models excluded (assuming S not dependent on ET
and Dh)
12
Survival Probability

• Assumed to be independent of parton x (ET , Dh)
• Originally weak ?s dependence
• Gotsman, Levin, Maor Phys. Lett B 309
  (1993)
• Subsequently recalculated
• GLM hep-ph/9804404
• Using free-factor and soft-color model
• 
  (uncertainty from MC stats and model difference)
• with

13
Hard Color Singlet Conclusions

• DØ has measured color-singlet fraction for
• 630 GeV and 1800 GeV same ET, h
• as a function of ET and Dh at 1800 GeV
• fS shows rising trend with
  ET, ?h
• Measured fraction rises with initial quark
  content (Assuming the Survival Probability is
  constant with ET and Dh)
• Consistent with a soft color rearrangement model
  preferring initial quark states
• ? Inconsistent with two-gluon, photon,
• or U(1) models
• ? Cannot exclude single-gluon model

14
Modifications to Theory

• BFKL Cox, Forshaw (manhep99-7) use a non-running
  as to flatten the falling ET prediction of BFKL
  (due to higher order corrections at non-zero t)

• Soft Color Gregores subsequently performed a
  more careful counting of states that produce
  color singlets to improve prediction.

15
Hard Single Diffraction Studies
Measure multiplicity here
Measure min multiplicity here
-4.0 -1.6 h 3.0
5.2
OR
-5.2 -3.0 -1. h 1. 3.0
5.2
hep-ex/9912061, submitted to PLB

• Gap fractions (central and forward) at ?s 630
  1800 GeV
• Single diffractive x distribution

16
Data Selection

• Require single interaction
• Require central vertex
• Gap Fraction
• (diffractive dijet events / all dijet
  events)
• Forward Jet Trigger
• two 12GeV Jets hgt1.6
• (_at_ 630, 28k events)
• (_at_ 1800, 50k events)
• Central Jet Trigger
• two 15(12) GeV Jets hlt1.0
• (_at_ 630(12), 48k events)
• (_at_ 1800(15), 16k events)
• SD Event Characteristics
• Single Veto Trigger
• two 15(12) GeV Jets and no hits in L0
  North of South array
• (_at_ 630(12), 64k events)
• (_at_ 1800(15), 170k events)

17
Event Characteristics
1800 Forward Jets
Solid lines show show HSD candidate events Dashed
lines show non-diffractive events

• Less jets in diffractive events
• Jets are narrower and more back-to-back
• Diffractive events have less overall radiation
• Gap fraction has little dependence on average
  jet ET

18
630 vs. 1800 Multiplicities
?s 1800 GeV
?s 630 GeV
19
Fitting Method
Example of fit for 1800 forward sample Signal is
fit with a 2D falling exponential while
background is fit with a 4 parameter polynomial
surface Shapes are in agreement with Monte
Carlo, the residual distributions are well
behaved Distributions have c2/dof lt 1.2
20
Single Diffractive Results
Measure Multiplicity here
or
-4.0 -1.6 -1.0 h 1.0
3.0 5.2
(Gap Fraction diffractive Dijet Events /
All Dijets)
Data Sample Measured Gap Fraction
1800 Forward Jets 0.65 0.04 - 0.04
1800 Central Jets 0.22 0.05 - 0.04
630 Forward Jets 1.19 0.08 - 0.08 630
Central Jets 0.90 0.06 - 0.06
Data Sample Ratio 630/1800 Forward
Jets 1.8 0.2 - 0.2 630/1800 Central
Jets 4.1 0.8 - 1.0 1800 Fwd/Cent Jets 3.0
0.7 - 0.7 630 Fwd/Cent Jets 1.3 0.1 -
0.1
Forward Jets Gap Fraction gt Central Jets Gap
Fraction 630 GeV Gap Fraction gt 1800 GeV Gap
Fraction
21
Comparison to MC
fvisible ?gap fpredicted
? ?gap Add diffractive multiplicity from MC to
background data distribution Fit to find percent
of signal events extracted
? Find predicted rate POMPYT x 2 / PYTHIA Apply
same jet ? cuts as data, jet ETgt12GeV Full
detector simulation
Model pomeron exchange in POMPYT26
(Bruni Ingelman) based on PYTHIA
define pomeron as beam particle Use
different structure functions
22
(No Transcript)
23
Pomeron Structure Fits

• Demand data fractions composed of linear
  combination of hard and soft gluons, let overall
  ? s normalization and fraction of hard and soft
  at each energy be free parameters
• If we have a ? s independent normalization then
  the data prefer
• 1800 hard gluon 0.180.05(stat)0.04-0.03(syst)
• 630 hard gluon 0.390.04(stat)0.02-0.01(syst)
• Normalization 0.430.03(stat)0.08-0.06(syst)
• with a confidence level of 56.
• If the hard to soft ratio is constrained the data
  prefer
• Hard gluon 0.300.04(stat)0.01-0.01(syst)
• Norm. 1800 0.380.03(stat)0.03-0.02(syst)
• Norm. 630 0.500.04(stat)0.02-0.02(syst)
• with a confidence level of 1.9.
• To significantly constrain quark fraction
  requires additional experimental measurements.
• CDF determined 56 hard gluon, 44 quark but this
  does not describe our data without significant
  soft gluon or 100 quark at 630

24
x Calculation
Where ? is the momentum fraction lost by the
proton Can use calorimeter only to measure
Weights particles in well-measured
region Can define for all events
Collins (hep-ph/9705393)
?true ?calc 2.2 0.3
? calculation works well not dependent on
structure function or center-of-mass energy
25
Single Diffractive x Distributions
? distribution for forward and central jets using
(0,0) bin
Dp p

? ? 0.2 for ?s 630 GeV
26
Summary

• Measurements at both CM energies with same
  detector adds critical new information for
  examining pomeron puzzle
• The higher rates at 630 were not widely predicted
  before the measurements
• Pioneering work in Central Rapidity Gaps
• Assuming x-independent survival probability, data
  support soft color rearrangement models (single
  gluon not excluded)
• Modifications to theory based on data have
  occured
• Forward gaps data
• In a partonic pomeron framework require reduced
  flux factor combined with gluonic pomeron
  containing significant hard and soft components
• Found larger fractional momentum lost by
  scattered proton than expected by traditional
  pomeron exchange
• Results imply a non-pomeron based model should be
  considered.

27
Gap Efficiencies
? Efficiency tag events with gap ?gap Add
diffractive multiplicity from MC to
background data distribution Fit to find
percent of signal events extracted Statistical
error MC energy scale error
MC Sample 1800 FWD JET 1800 CENT JET Hard
Gluon 74 ? 10 34 ? 5 Flat Gluon 66 ?
9 55 ? 7 Quark 61 ? 8 18 ? 2 Soft
Gluon 22 ? 3 2.2 ? 0.8
MC Sample 630 FWD JET 630 CENT JET Hard
Gluon 92 ? 16 48 ? 6 Flat Gluon 71 ?
14 60 ? 8 Quark 55 ? 11 26 ? 3 Soft
Gluon 23 ? 4 6.0 ? 4.8