1

LEADING BARYON PRODUCTION LEADING BARYON PRODUCTION at HERAat HERA

Lorenzo RinaldiLorenzo RinaldiOn behalf of H1 and ZEUS CollaborationsOn behalf of H1 and ZEUS Collaborations

2

MotivationsMotivations Large fraction of events with a Leading Baryon (LB)Large fraction of events with a Leading Baryon (LB) LB produced at small angle in forward direction: difficult LB produced at small angle in forward direction: difficult detectiondetection Production mechanism still not clear: soft scale, alternative Production mechanism still not clear: soft scale, alternative approach needed approach needed Interest in LB study for next experiments @ LHC Interest in LB study for next experiments @ LHC absorptive corrections related to gap survival probability absorptive corrections related to gap survival probability (diffractive Higgs, pile-up background...)(diffractive Higgs, pile-up background...)

Results discussed in this talk:Results discussed in this talk: Leading Proton (LP) spectra in DIS Leading Proton (LP) spectra in DIS NEW NEW Leading Neutron (LN) spectra in DIS and Leading Neutron (LN) spectra in DIS and pp Dijet Dijet p with a LN p with a LN NEW NEW Latest developments in theoryLatest developments in theory Comparison with modelsComparison with models

3

Leading baryon production in ep collisionsLeading baryon production in ep collisions

Standard fragmentationStandard fragmentation • LB from hadronization of p remnant• Implemented in MC models (Cluster, Lund strings...)

Virtual particle exchangeVirtual particle exchange , IR, IP, ,..LB also from p fragmentation in double dissociative diffraction

LB variables:pT

2, xL=ELB/Ep

t=(p-p’)2

,IR,IPpp, n

LB cross sections vs structure functions:LB cross sections vs structure functions:(QCD-based approach)

Lepton variablesQ2, W, x, y 22)4(

2

2

4

2

22

224

,,,2

14,,,

TLLB

TL

TL pxQxFy

yxQdpdxdxdQ

pxQxd

4

Leading baryon detectorsLeading baryon detectors

ZEUS Leading Proton Spectrometer (LPS)ZEUS Leading Proton Spectrometer (LPS) 6 stations each made by 6 Silicon-detector planes 6 stations each made by 6 Silicon-detector planes Stations inserted at 10Stations inserted at 10beambeam from the proton beam during data taking from the proton beam during data taking xxLL

< 1% < 1% ppTT22 ~ ~ few MeVfew MeV2 2 (better than p-beam spread (better than p-beam spread ~ 50 - 100 MeV~ 50 - 100 MeV))

ZEUS Forward Neutron Tracker (FNT)ZEUS Forward Neutron Tracker (FNT) Scint. hodoscope @ 1Scint. hodoscope @ 1λλint, int, σσx,yx,y=0.23cm, =0.23cm, σσθθ=22=22μμradrad

H1 Forward Neutron Calorimeter (FNC)H1 Forward Neutron Calorimeter (FNC) Lead-scintillator calorimeter @ 107m from I.P. + veto hodoscopesLead-scintillator calorimeter @ 107m from I.P. + veto hodoscopes (E)/E(E)/E≈20%, neutron detection eff. 93±5%≈20%, neutron detection eff. 93±5%

ZEUS Forward Neutron Calorimeter (FNC)ZEUS Forward Neutron Calorimeter (FNC) 1010 lead-scintillator sandwich lead-scintillator sandwich σσ/E =0.65/√E, Energy scale=2%/E =0.65/√E, Energy scale=2% Acceptance Acceptance nn<0.8 mrad, azimuthal coverage 30%<0.8 mrad, azimuthal coverage 30%

H1 Forward proton spectrometer (FPS)H1 Forward proton spectrometer (FPS) 2 stations each made by 4 scintillating fibres hodoscopes planes2 stations each made by 4 scintillating fibres hodoscopes planes xx==55radradyy=100=100rad, Energy resolution 8 GeVrad, Energy resolution 8 GeV Acceptance 500<EAcceptance 500<Ep’p’<780 GeV<780 GeV

5

Leading Proton: cross section vs xLeading Proton: cross section vs xLL

Flat below diffractive peakCross section at low pT

2

Agreement with photoproduction

NEW NEW

NEW LP results: LPS stations full set used

6

LP: cross section vs pLP: cross section vs pTT2 2 and b-slopesand b-slopes

Exponential behaviourFit to A*exp(-b*pT

2) shown with stat. error

No strong dependence of b on xL

NEW

NEW

7

),(

),(

),(

),,(),,(

22

2)2(2)2(

22

2)3(22)3(

QxF

QxFr

QxF

xQxFxQxr

LPLP

LLP

LLP

Structure function ratio

Information on LP productionas a function of DIS variables

Test of vertex factorization

LP: ratio to inclusive DISLP: ratio to inclusive DIS

NEW

8

17-18% of DIS events have a 17-18% of DIS events have a LP with 0.5<xLP with 0.5<xLL<0.92, almost <0.92, almost

independently of x and Qindependently of x and Q22

No strong dependence on x and No strong dependence on x and QQ22 when integrating over when integrating over 0.5<x0.5<xLL<0.92<0.92

No clear evidence of vertex No clear evidence of vertex factorization violationfactorization violation

NEWNEW

Ratios to inclusive DIS - 2Ratios to inclusive DIS - 2

9

LP structure functionsLP structure functions

FF22LPLP=r=rLPLP*F*F22 (ZEUS-S parametrization used) (ZEUS-S parametrization used)

FF22LPLP: same dependence on x and Q: same dependence on x and Q22 as F as F22

NEW NEW

10

Comparisons to Reggeon exchange modelComparisons to Reggeon exchange model

Szczurek et al., Phys Lett B428, 383 (1998)Pomeron ReggeonN

Predictions good in shape but:Predictions good in shape but:xxLL slighty underestimated slighty underestimated

b-slope slightly overestimatedb-slope slightly overestimated

11

O.P.E. partially explains the LN O.P.E. partially explains the LN production production

),)1((),(),,,( 22

*/

22

QWxtxfdtdx

txQWdLLp

L

LeXnep

),()1()(

),( 2)(22/ txFx

mt

ttxf L

tLLp

(t)(t) and form factor and form factor FF22(x(xLL,t),t) model dependentmodel dependent

Longitudinal momentum spectrum and pLongitudinal momentum spectrum and pTT22 slopes discriminate slopes discriminate

between different parametrizations of fluxesbetween different parametrizations of fluxes

Leading Neutron: Leading Neutron: One-Pion-Exchange modelOne-Pion-Exchange model

12

Rescattering model and absorptionRescattering model and absorption

D’Alesio and Pirner D’Alesio and Pirner (EPJ A7(2000) 109)(EPJ A7(2000) 109)

Neutron rescatters on Neutron rescatters on hadronic component. hadronic component.

Absorption enhanced when Absorption enhanced when

• -n system size larger -n system size larger low x low xLL

• size larger size larger photoproduction photoproduction

p

DIS

Nikolaev,Speth and ZakharovNikolaev,Speth and Zakharov (hep-ph/9708290)(hep-ph/9708290)

Re-scattering processes via Re-scattering processes via additional Pomeron exchanges additional Pomeron exchanges (Optical Theorem)(Optical Theorem)

Kaidalov, Khoze, Martin, Ryskin (KKMR)Kaidalov, Khoze, Martin, Ryskin (KKMR) (hep-ph/0602215, hep-ph/0606213)(hep-ph/0602215, hep-ph/0606213)

Enhanced absorptive corrections (Enhanced absorptive corrections ( exclusive Higgs @ LHC), calculation of exclusive Higgs @ LHC), calculation of migrations, include also migrations, include also and aand a22

exchange (different xexchange (different xLL & p & pTT dependences) dependences)

13

LN: longitudinal momentum spectrumLN: longitudinal momentum spectrum

• LN yield increases with xLN yield increases with xLL due due

to increase in phase space: to increase in phase space: ppTT

22 < < 0.476 0.476 xxLL22

• LN yield decreases for xLN yield decreases for xLL1 1

due to kinematic limitdue to kinematic limit

DISDIS

Phase space coverage

14

LN: ratio LN: ratio p/DISp/DIS

ppDISDIS

W dependence:W dependence:

~W~W, , ((pp) ) (( *p*p) )

WW22=(1-x=(1-xLL)W)Wpp

22 (1-x (1-xLL) ) -0.13-0.13

absorption rate rescaled absorption rate rescaled

Models in agreement with dataModels in agreement with data

• LN yield in PHP < yield in DIS LN yield in PHP < yield in DIS

factorization violationfactorization violation

Data compared to OPE with Data compared to OPE with absorption.absorption. • Qualitatively similar to D’Alesio Qualitatively similar to D’Alesio and Pirner (loss through absorption)and Pirner (loss through absorption)• Nikolaev,Speth and Zakharov Nikolaev,Speth and Zakharov model also shown: similar trend but model also shown: similar trend but weaker xweaker xLL dependence dependence

15

LN: KKMR absorption modelLN: KKMR absorption model

Kaidalov, Khoze, Martin, Ryskin:Kaidalov, Khoze, Martin, Ryskin: • Pure Pure exchange (not shown) too high exchange (not shown) too high• Absorption and migration effects Absorption and migration effects

reduce the LN yield and fit the data reduce the LN yield and fit the data betterbetter

• Additional Additional and a and a22 exchanges exchanges

enhance the LN yieldenhance the LN yield

pp

16

LN: DIS cross section vs pLN: DIS cross section vs pTT22 in x in xLL bins bins

• ppTT22 distributions well described distributions well described

by an exponentialby an exponential

• Intercept a(xIntercept a(xLL) and slopes b(x) and slopes b(xLL) )

fully characterize the xfully characterize the xLL-p-pTT22

spectraspectra

2)(2

2

)(1

TL pxbL

TL

LN

inc

exadpdx

d

17

LN: intercepts and slopes in DISLN: intercepts and slopes in DIS

b consistent with b consistent with 0 for x0 for xLL<0.3<0.3

2)(2

2

)(1

TL pxbL

TL

LN

inc

exadpdx

d

18

LN b-slopes: DIS & photoproduction comparisonLN b-slopes: DIS & photoproduction comparison

Fit to exponential

p

DISslopes different in slopes different in p and DISp and DIS in general agreement with in general agreement with expectation from absorption models expectation from absorption models

19

LN b-slopes: comparison to modelsLN b-slopes: comparison to models

OPE models:OPE models: Dominant at 0.6<xDominant at 0.6<xLL<0.9<0.9 (non-) Reggeized flux, different form (non-) Reggeized flux, different form factors with different parametersfactors with different parameters none of the models seem to decribe none of the models seem to decribe the data wellthe data well

KKMR model:KKMR model:good description of the data good description of the data considering absorption effects and considering absorption effects and ,a,a22 exchange contributions exchange contributions

20

LN spectrum: QLN spectrum: Q22 dependence dependence

• LN yield increases monotonically with LN yield increases monotonically with QQ22

• consistent with absorption (larger Qconsistent with absorption (larger Q22 smaller smaller

3 Q3 Q22 bins + bins + pp

21

Comparisons LP - LN dataComparisons LP - LN data

Very similar behaviour xVery similar behaviour xLL<0.85<0.85

LP cross section almost twice LNLP cross section almost twice LN

In particle exchange model:In particle exchange model:expected from isospin-1: LP=1/2LNexpected from isospin-1: LP=1/2LNOther exchanges needed (isoscalars)Other exchanges needed (isoscalars)

Slopes comparable 0.7<xSlopes comparable 0.7<xLL<0.8 <0.8

where where exchange dominates exchange dominates

22

LN production compared to MC predictionsLN production compared to MC predictions

• Compare LN DIS distribution to MC Compare LN DIS distribution to MC models:models:

– RAPGAP standard fragmentationRAPGAP standard fragmentation – RAPGAP OPERAPGAP OPE– LEPTO standard fragmentationLEPTO standard fragmentation– LEPTO soft color interactionLEPTO soft color interaction

• Both standard fragmentation failBoth standard fragmentation fail– Too few n, too few xToo few n, too few xLL

– b-slopes too low b-slopes too low

• RAPGAP-OPE: close to data in shape but RAPGAP-OPE: close to data in shape but not in magnitudenot in magnitude• LEPTO-SCI: reasonable description of xLEPTO-SCI: reasonable description of xLL

spectrum and intercepts, bad slopesspectrum and intercepts, bad slopes

Other models also fail (ARIADNE, Other models also fail (ARIADNE, CASCADE,PYTHIA,PHOJET)CASCADE,PYTHIA,PHOJET)

23

Dijet Dijet p with a LNp with a LNPresence of jets hard-scale

Naively, rescattering effects expected in resolved photoproduction:

hadz

,jet

ηT

obsγ )p(E

eE

x

21 low x, photon acts

like a hadron

H1 data: ratios (jj+LN)/jj reasonably well described by photoproduction MC

24

Dijet with a LNDijet with a LN

Ratios (LN+jj)/jj

Model by Klasen and Kramer based on OPE

Ratios well described by NLO predictions

xL spectrum:

Reasonable shape, NLO too high

NEW

NEW

25

Dijet production with a LNDijet production with a LN

epejjnX vs epenXSuppression observed in dijet+LN:Kinematic or absorption/rescattering?

Look at xBP=1-(E+pZ)/2Ep

Fraction of proton beam energy available for particle production in the forward beampipe

Kinematic constraint: xL<xBP

The lower xBP values in dijet-p constrain the neutrons to lower xL than DIS

NEW

NEW

26

Dijet with a LNDijet with a LN

After reweighting xBP, the xL distributions of the two processes epejjnX and epenX agree:Mainly kinematic effect, no clear evidences of absorption

b-slopes No significant difference within errors

NEW

NEW

27

SummarySummary

• LP spectra in DIS; well described by Reggeon-exchange modelLP spectra in DIS; well described by Reggeon-exchange model

• LP production as a function of DIS variables: no dependence LP production as a function of DIS variables: no dependence observed observed

• LN production measured in DIS and LN production measured in DIS and pp

• LN characterized by rescattering and absorption effects: well LN characterized by rescattering and absorption effects: well reproduced by some models reproduced by some models

• MC generators in general fail to reproduce the measured quantities MC generators in general fail to reproduce the measured quantities need to tune the generators need to tune the generators

• Dijet-Dijet-p with LN: suppression most probably due to kinematic effects. p with LN: suppression most probably due to kinematic effects.

HERA provided high precision measurements of leading baryon HERA provided high precision measurements of leading baryon production. production.

Now it’s time to work together with theory people and apply our Now it’s time to work together with theory people and apply our knowledge to the next future physics knowledge to the next future physics

28

29

Rescattering model and absorption 1Rescattering model and absorption 1

Model 1: One pion exchange in the Model 1: One pion exchange in the framework of triple-Regge formalismframework of triple-Regge formalism

Nikolaev,Speth & ZakharovNikolaev,Speth & Zakharov

Re-scattering processes via Re-scattering processes via additional pomeron exchanges additional pomeron exchanges (Optical Theorem)(Optical Theorem)

(hep-ph/9708290)(hep-ph/9708290)

((Kaidalov,) Khoze, Martin, Kaidalov,) Khoze, Martin, Ryskin (KKMR)Ryskin (KKMR)

Enhanced absorptive corrections Enhanced absorptive corrections (( exclusive Higgs @ LHC), exclusive Higgs @ LHC), calculation of calculation of migrations, include migrations, include also also and aand a22 exchange exchange

(different x(different xLL & p & pTT dependences) dependences)

(hep-ph/0602215, hep-ph/0606213)(hep-ph/0602215, hep-ph/0606213)

30

Rescattering model and absorption 2Rescattering model and absorption 2

Model 2: calculations from D’Alesio and Model 2: calculations from D’Alesio and Pirner in the framework of target Pirner in the framework of target fragmentation fragmentation (EPJ A7(2000) 109)(EPJ A7(2000) 109)

more absorption when photon size more absorption when photon size larger (small Qlarger (small Q22) ) less neutrons less neutrons detected in photoproductiondetected in photoproduction more absorption when mean more absorption when mean -n -n

system size (system size (‹‹rrnn››) smaller at low x smaller at low xLL

less neutrons detected at low x less neutrons detected at low xLL

more absorption more absorption fewer neutrons fewer neutrons detected with higher pdetected with higher pTT

22 larger b- larger b-

slope expected in photoproductionslope expected in photoproduction

p

DIS