Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 1

Leading Baryonsat HERA

• Introduction• Diffractive structure function measured in events with a leading proton (See also M. Kapishin’s talk)• D* photoproduction with a leading neutron• Dijet with a leading neutron• Summary

32nd International Conference on High Energy PhysicsBeijing, China, Aug. 16-22, 2004

Heuijin Lim(DESY) on behalf of the H1 and ZEUS Collaborations

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 2

IntroductionIntroduction

, IP

Standard fragmentation :• p or n appears in fragmentation of target jet

p, n

p, n

• A significant fraction of ep scattering events contains a leading baryon produced at small t (soft process).

In exchange models :• p from exchange of neutral iso-scalar or iso-vector (, IR, IP) • n from exchange of charged iso-vector (+, +, …)

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 3

DDiffractive structure function measured with a leading proton

22 )kk(Q 2)qp(W

22X )ppkk(M

22

2X

2

IPWQ

MQx

IP22

X

2

x

x

QM

Q

2)pp(t

Kinematics of ep eXp

IPpZL x1Epx

)t,x,Q,(F)R1(2

yy1

Q

4

dtdxdQd

dIP

2)4(D2D

2

4

2

IP2

eXpep

Diffractive structure function of proton

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 4

t dependence from LPS t dependence from LPS

• Fit t distribution to

dd|t| shows steep fall-off with t as in elastic hadron-hadron scattering.

)tbexp(tdd 29.0

5.0 GeV.)syst(.)stat(5.09.7b

• H1 b-value considerably below ZEUS result. • Regge phenomenology predicts “shrinkage” of the diffractive peak:

Additional dependence expected in models.Data not yet precise enough to show presence/absence of IP’ term.

IPIP02

X

2

0x

1ln2b

M

Wln2bb

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 5

QQ22 dependence of cross section with LPS dependence of cross section with LPS

• At high Q2, ddMX falls with Q2.• As Q2 0, d/dMX depends weakly on Q2.

• Data are compared with the color dipole model (BEKW parametrization).

• At high Q2 in high MX bins,, Data above BEKW. Expect this region (xIP>0.01) to be dominated by Reggeon exchange.

(Bartels, Ellis, Kowalski and Wüsthoff)

)1(FTqq

)1(FTgqq

For medium ,

For small ,

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 6

Diffractive structure function of the proton Diffractive structure function of the proton xxIPIPFF22

D(3)D(3) (x (xIPIP,,,Q,Q22) with LPS) with LPS

• Regge fit (xIP < 0.01) with common Pomeron flux factor

)Q,(F)x(fF 2IP2IPIP

)3(D2

dtx

e)x(f

1)t(2IP

tb

IPIPIP

IP0

t)0()t( IPIPIP

with

.)syst(02.0.)stat(02.016.1)0(IP

• xIP>0.01 xIPF2

D(3) increases as xIP1 Reggeon contributions• xIP<0.01 xIPF2

D(3) increases as xIP0 Parton evolution as xIP0

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 7

Diffractive hard scattering factorization Diffractive hard scattering factorization

)Q,z()t,x,Q,z(f)Xpp( 2

iIP2D

pi*D

*

fi/pD(z,Q2,xIP,t) : Probability to find in a proton, with a probe of resolution Q2, parton i

with momentum fraction z, under the condition that the proton remains intact and emerges with small energy loss, xIP, and momentum transfer, t.

[Collins (1998); Trentadue, Veneziano (1994); Berera, Soper (1996)…]

Universal partonic cross sectionDiffractive parton distribution function

NLO QCD fits to the ZEUS dataNLO QCD fits to the ZEUS data • Perform fit to the results from LPS and from diffractive charm in DIS.xIP < 0.01 and Q2>2 GeV2

• Regge factorisation assumption, • Donnachie and Landshoff Pomeron flux• Parametrise PDFs(quark flavour singlet and gluon) using zf(z) = a1+a2z+a3z2)(1-z)a4

For light quark distribution, assuming Charm quarks were treated in Thorne-Roberts variable flavor number (TRVFN) scheme with mc=1.45 GeV.

• NLO evolution : QCDNUM

sdusdu

)Q,(F)x(f)x,Q,(F 2IP2IPIPIP

2)3(D2

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 8

NLO QCD fit on LPS+charm data NLO QCD fit on LPS+charm data

• QCD fit describes data with 2/ndf = 37.8/36

• Fraction of the t-channel momentum carried by gluons

ZEUS at Q2=2 GeV2

H1(2002) at Q2=10 GeV2.

consistent with H1.

.)%syst(.)stat(882 516

%1575

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 9

Comparison of LPS and H1-FPS Comparison of LPS and H1-FPS

• Good agreement between the two experiments.

• H1 QCD fit (from large rapidity gap data, see M. Kapishin’s talk) describes the results of ZEUS(LPS) and H1(FPS).

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 10

Forward neutron calorimeter Forward neutron calorimeter

Z = 106 m

Leading Proton SpectrometerLeading Proton Spectrometer

FNCFNC P beam hole

Window for n-acceptance n < 0.8 mrad

ZEUS

H1

• ZEUS FNC consists of 10 I Pb-scintillator sandwich calorimeter.•

nnn E65.0E)E(

107

• H1 FNC consists of Pb-scintillator fibers.

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 11

Factorization hypothesis Factorization hypothesis

)s()t,x(fdtdx

dLp/

L

Xnp

pion form-factor

with s’=s(1-xL)

n

Independentvertices

parton distribution in pion

• Study relationship between soft and hard interactions : Hard scale : mc

2 for charm production ET for dijet events Q2 for DIS. Soft scale : pT of the neutron

• Study vertex factorization hypothesis.

• Ratios (d/dY neutron tagged)/(d/dY inclusive) are flat with Y kinematic variables of the event.

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 12

DD** photoproduction with a leading neutron photoproduction with a leading neutron

n

X s

0* DD s)K(

nXeDeXnep *

nb.)R.B(05.0.)syst(.)stat(22.008.2 12.018.0

DLN

*

tagged in FNC : 0.2 < xL < 1 n < 0.8 mrad

Q2 < 1 GeV2

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 13

Differential cross sections of D* production Differential cross sections of D* production with a leading neutronwith a leading neutron

• Standard fragmentation (HERWIG, PYTHIA, RAPGAP(inc)) and RAPGAP(OPE) describe W, pT and dist. Only OPE describes xL dist.

• Data is not sensitive to the different variants of the pion structure function.• Light-cone(f1) and exponential(f3) pion form factors are compatible with data in shape.

Different pion PDFs

Different pion form factors

Q2 < 1 GeV2

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 14

Ratio if D* yield with and without a neutronRatio if D* yield with and without a neutron

• Ratio is flat with respect to W, pT(D*) and (D*).

.)%syst(.)stat(93.085.8 48.061.0

)/(r *Dinc

*DLN

*D

Agreement with factorization hypothesis.

Q2 < 1 GeV2

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 15

Dijet with a leading neutronDijet with a leading neutron

• fLN

flat with Etjet

Support factorization hypothesis rising with x

jet

Jet

Jet Dijet cross sections compared to pion exchange MC (POMPYT) with different pionPDFs. Not sensitive.

)/(r jjinc

jjLN

jj

Heuijin Lim ICHEP04, Beijing, 16-22 Aug. 16

SummarySummary

• Diffractive structure function measured with a leading proton in DIS Recent data from ZEUS with improved precision and extended kinematic range. Slope of d/d|t| is compatible with soft interaction at the proton vertex. Data can be described by color dipole model (e.g. BEKW parameterisation). Data described by a NLO QCD fit

• H1 Very Forward Proton Spectrometer (VFPS) Tag the scattered proton at HERA II for high rate studies of diffraction Large acceptance for low xIP and 0<|t|<0.5 GeV2

Expect physics data soon.

• D* and dijet with a leading neutron Data described by one-pion-exchange model. Data agree with the factorization hypothesis.

Acceptance :Red : 90-100 %, Blue:0-10%

FPS-HVFPSFPS-V