M. Barbi Exclusive Vector Meson Production and Inclusive K 0 S K 0 S Final State in DIS at HERA Outline: ¥ Exclusive vector meson production ¥ Summary • First observation of resonances in inclusive K S 0 K S 0 final state in DIS • Summary For the H1 and Zeus Collaboration M. Barbi McGill University Photon2003, Frascati 07-11/04

M. Barbi Exclusive Vector Meson Production VM (M VM ) At HERA: Cross section Hard ( pQCD ) Soft ( Regge + VDM ) q VM p p' ** q R e' e ** Q2Q2 p p' t WpWp p ** VM |P VM = , J/y,  (2S), etc At low |t|  Gluon from fit to F 2 scaling violation  softhard x ≈ 1/W 2 Scale Q 2, |t|, M VM 2 Transv. size of the interaction region Small size. No W dependence.

M. Barbi Fit   W   :  0.22 for  0, ,   0.8 for J/  e' e p p' ** P VM M 2 VM  high M VM 2 sets hard scale Elastic VM in photoproduction (Q 2 = 0) Exclusive Vector Meson Production A first look:

M. Barbi J/  Photoprod. in bins of W and comparison to LLA calculations Exclusive Vector Meson Production Leading Log. Approx. (LLA) pQCD with different gluon parametrisations Models sensitive to input parametrisation of the gluon density Looking at high M VM (consistent with hard regime  previous slide) Models with in reasonable agreement with data Indeed, M VM is a hard scale

M. Barbi Q 2 (GeV 2 ) 00 J/  b (GeV -2 ) Steepness of  0 and J/  cross-section R is the transverse size of the interaction region  R decreases with Q 2 for  0  R already small for J/  at small Q 2 Exclusive Vector Meson Production b ≈ 0,5 GeV -2  proton size And the power-law W  dependence?  Can Q2 provide a hard scale?

M. Barbi 00 transition soft  hard Q 2 also provides a hard scale. W-dependence of elastic  0 and J/  in bins of Q 2 (PhP and DIS) Exclusive Vector Meson Production Can Q 2 provide a hard scale? J/  already steep at Q 2 = 0  Data fitted with W   pQCD models consistent with data  tot (  p  J/  p)[nb] W [GeV] J/  Q 2 [GeV 2 ]  Flat 

M. Barbi Exclusive Vector Meson Production  p   Y  p   Y  p  J/  Y Photoproduction of proton-dissoc. VM at high |t| p ** VM Y Use BFKL LLA approach to fit data (Forshaw and Poludniowski) Typical elastic VM production has |t|<1GeV 2. Use proton dissociation to reach higher |t| BFKL LLA approach is in agreement with data   |t| sets a hard scale What about |t|?

M. Barbi  Q 2, M VM 2 and |t| set a hard scale.  Perturbative QCD predictions agree with data Exclusive Vector Meson Production Summary

M. Barbi Inclusive K S 0 K S 0 final state in DIS K 2 (1430) f 2 (8)f 2 (1) a 2 (1320) K 2 (1430) - K 0 (1430) f 0 (8)f 0 (1) a 0 (????) K 0 (1430) - f 0 (1370) ~ uu+dd f 0 (1500) ~ ?? f 0 (1710) ~ ss f 0 (8) f 0 (1) Tensor J CP =2 ++ Nonet Scalar J CP =0 ++ Nonet f 2 (1270) ~ uu+dd f 2 (1525) ~ ss f 2 (8) f 2 (1) f 0 (1710) is a glueball candidate 3 candidates for 2 spots QCD predicts the existence of hadrons made up by gluons (glueballs). From Lattice QCD calculations, the lightest glueball has J CP =0 ++ with a mass 1730±100 MeV. K S 0 K S 0 couples to meson states with J CP =(even) ++

M. Barbi  A total luminosity of 121 pb -1 was used  Only events with at least 2 K S 0 were selected  Clean K S 0 sample Inclusive K S 0 K S 0 final state in DIS

M. Barbi Inclusive K S 0 K S 0 final state in DIS First observation of J CP =(even) ++ in DIS. Two states are observed • a state consistent with f 2 ’ (1525) • X(1726) (is this the f 0 (1710) ?) A third state is observed in the (problematic) 1300 MeV mass region, consistent with the f2(1270)/a2(1320) interference Several states have been observed in the 2GeV region (see PDG02) M. Barbi

P z (KsKs) 0 (93%) Increasing x p More gluons 78% of the KsKs pairs have x p >1 Current region in DIS is equivalent to an e + e - hemisphere  K S 0 K S 0 in the Breit-frame Inclusive K S 0 K S 0 final state in DIS

M. Barbi  First observation of resonances in K S 0 K S 0 final state in DIS was reported  Two states are observed in the 1300 MeV, 1500 MeV mass region consistent with f 2 (1270)/a 2 (1320) and f 2 ’(1525)  Another state X(1726) is observed, probably the f 0 (1710) (a glueball candidate)  States are produced in a gluon rich environment Inclusive K S 0 K S 0 final state in DIS Summary

M. Barbi Extra Slides

M. Barbi Vector Meson Production e' e p p' ** VM (M VM ) e' e p Y (M Y ) ** VM Q2Q2 t WpWp ElasticProton Dissociative Hard regime (pQCD) Q 2 =  * virtuality ; Q 2 < 100 GeV 2 W  p =  * p CMS energy; 20 < W  p < 290 GeV t = 4-mom. transf. squared ; |t| < 20 GeV 2 VM =  J/  '  R = 1/[ z(1-z) Q 2 + m q 2 ] 1/2 ; z = E q / E  * |P pQCD q VM p p' ** q R

M. Barbi Vector Meson Production Soft regime ( Regge + VMD ) VM eM VM 2 /g V ** p p' ( Donnachie-Landshoff )   rising weakly with W: At low |t|, |t| < 1.5 GeV 2    steep exponential t dependence (shrinkage): Photon transversally polarized ; low Q 2 or |t| ( Shrinkage )

M. Barbi Vector Meson Production Hard regime ( pQCD ) Photon mainly longitudinally polarized Large Q 2, M VM or |t| ¯ qq system is small, probes the proton p p' ** q R q Two-gluon-exchange approach at LO p p' ** VM Gluon ladder (LLA approach) Gluon from F 2 scaling violations For ex.: Q 2 eff = ¼ · (Q 2 + M VM 2 + |t|) (Ryskin Model) At which scale Q eff 2 should xg be evaluated ?  rising steeper than expected from Regge+VMD

M. Barbi -t (GeV 2 )  *p  J/  Y  *p   Y  *p    Y d  /dt (nb/GeV 2 ) LLA BFKL two gluon Forshaw and Poludniowski fitted the ZEUS data for p-dissociative photoproduction of  0,  and J/  mesons (hep-ph/ ): Dependence at large |t| not exponential: > BFKL LLA approach: consistent with data  indication that large |t| may provide a hard scale to apply perturbative QCD Vector Meson Production Photoprod. of proton-dissoc. VM at high |t| J/   p Y t p p' ** VM > two-gluon-exchange approach at LO: inadequate

M. Barbi Vector Meson Production J/  cross-section and QCD models (PhP and DIS)  QCD models FKS and MRT(CTEQ5M) are in good agreement with data

M. Barbi Vector Meson Production Diffractive Photoproduction of  (2S) - pQCD  Cross-section suppressed with respect to that of J/y  Steeper W dependence pQCD q VM p p' ** q  Overall R ≈ consistent with pQCD prediction  Fit W    ≈  consistent with W dependence for  (2S )  (2S) wavefunction has a node

M. Barbi 00 transition soft  hard Fit   W  : J/  already steep at Q 2 = 0  Q 2 provides a hard scale Vector Meson Production W-dependence of elastic   and J/  in bins of Q 2 (PhP and DIS)

M. Barbi Extras….. KsKs final state

M. Barbi Scalar and Tensor Mesons qq S=1 L=1 J PC = 0 ++ J PC = 2 ++ Masses 1-2 GeV, scalar and tensor mesons much heavier than the pseudoscalars Quark Model (u,d,s) I=0 P = (-1)x(+1)x(-1) L = (-1) L+1 C = (-1) L+S J = 0 or 2 S I Scalars Tensors

M. Barbi QCD: richer than Quark Model, predicts gluon states with J PC = 0 ++,2 ++, I=0 g f 0 (1370) f 0 (1500) f 0 (1710) Observed States mix KsKs final state couples only to J PC =(even) ++ and it is CLEAN. This is the golden channel (given statistics) to look for scalar and tensor meson resonances Ks: has S=L=0, P=-1, C=+1 KsKs: has P=+1, C=+1  J=even Is this a qq state?

M. Barbi Production mechanisms in DIS Odderon Exchange VM Radiative decays In quark Jets In gluon jets Gluon rich processes Gluon poor processes

First observation of J CP =(even) ++ in DIS. Two states are observed:  a state consistent with f 2  (1525) • X(1726) (is this the f 0 (1710) ?) Several states have been observed in the 2GeV region (see PDG02) f 0 (980)/a 0 (980) gives K s pair with very small opening angle in the lab. We would like to remove collinear K s pairs and then fit the spectrum. Fit 3 Breit-Wigner and a backg. function: K S 0 K S 0 final state in DIS Threshold Enhancement due to f 0 (980)/a 0 (980) Problematic region

cosKK < 0.92 cut Observation of J=even meson resonances in ep

Fit with 4 Breit-Wigner  Attempt to include the 1980 MeV mass region K S 0 K S 0 final state in DIS

How well a smooth background describes the data: It is less than 1% likely that a smooth distribution describes the data.

e + e -    K s K s If f(1710) is a glueball it should have small coupling to 