aa r X i v : . [ h e p - e x ] M a y First observation of (anti)deuterons in DIS
S. ChekanovFor the ZEUS CollaborationDESY Laboratory, 22607, Hamburg, Germany.On leave from the HEP division, Argonne National Laboratory,9700 S.Cass Avenue, Argonne, IL 60439, USAE-mail: [email protected] observation of (anti)deuterons in deep inelastic ep scattering (DIS) measured withthe ZEUS detector at HERA is reported. The production rate of deuterons is higherthan that of antideuterons. However, no asymmetry in the production rate of protonsand antiprotons was found. The (anti)deuteron yield is approximately three ordersof magnitude smaller than that of (anti)protons, which is consistent with the worldmeasurements. Deuterons ( d ) are loosely bound states whose production in high energy collisions can hardlybe accommodated in the standard fragmentation models. The measurements of such parti-cles may provide an important information on the structure of fragmentation region [1] andon the formation of multiquark states [2].In collisions involving elementary particles, several measurements have been performed inorder to understand the production of d ( ¯ d ). The cross section of ¯ d in e + e − → q ¯ q collisions [3]is lower than that measured in hadronic Υ(1 S ) and Υ(2 S ) decays, and disagrees with thepredictions based on the LUND string model [4]. The ( ¯ d ) rate in e + e − → q ¯ q is also lowerthan that in hadronic ( pA [5], pp [6]) and in γp collisions at HERA [7], but higher than thatin nucleus-nucleus ( AA ) collisions [8].According to the coalescence model [1] developed for heavy-ion collisions, the d rate isdetermined by the overlap between the wave function of a proton, p , and a neutron, n ,with the wave function of a d . The d cross section is the product of single-particle crosssections for p and n , with a coefficient of proportionality, B , reflecting the spatial size ofthe fragmentation region emitting the particles.Unlike studies in AA , Ap and pp , all previous measurements in collisions involving ele-mentary particles have been performed for ¯ d , since the reconstruction of d requires a carefulseparation of such states from particles produced in interactions of colliding beams withresidual gas in the beam pipe and secondary interactions of particles on detector material.The deep inelastic scattering (DIS) events, which are characterised by the presence of ascattered electron in the final state, provide an ideal environment for studies of d , since thebackground from non- ep interactions is minimal. In addition, during the 1996-2000 datataking, the ZEUS detector had a small amount of inactive material between the interac-tion region and the central tracking detector (CTD). This material consisted of the centralbeam-pipe and inner wall of the CTD, with the overall thickness of 2 . d , as well as to a small contribution from secondary deuterons. DIS 2007
EUS
M (GeV) E n t r i es + K p d t -1 ZEUS 120 pb(a)M (GeV) E n t r i es - K p d t -1 ZEUS 120 pb(b)
ZEUS
DCA (cm) -4 -3 -2 -1 0 1 2 3 4 E n t r i es p (a) ZEUS -1
120 pb
DCA (cm) -4 -3 -2 -1 0 1 2 3 40204060 d (b)DCA (cm) -4 -3 -2 -1 0 1 2 3 401000020000 p (c) DCA (cm) -4 -3 -2 -1 0 1 2 3 40102030 d (d) Figure 1. Left: The mass spectra reconstructed using dE/dx for positive and negativeparticles. Right: The
DCA distributions for: (a)-(b) particles and (c)-(d) antiparticles.The data sample used in the analysis corresponds to an integrated luminosity of 120 pb − taken between 1996 and 2000 with the ZEUS detector at HERA. After a DIS selection,the average Q for the data sample was about 10 GeV . The particle identification wasperformed using the dE/dx measurement. The masses, shown in Fig. 1(left), were calculatedfrom the measured track momentum and energy loss using the Bethe-Bloch formula.To identify particles produced in ep collisions, the distance of closest approach ( DCA )of the track to the beam spot in the transverse plane was used, since particles originatingfrom the primary ep collisions feature small values of | DCA | . The DCA distribution afterthe mass cuts and an additional cut on the distance, ∆ Z , of the Z -component of the trackhelix to the primary vertex, is shown in Fig. 1 (right). The number of particles produced in ep was assessed from the DCA distribution by using a side-band background subtraction.The numbers of d and ¯ d after the side-band background subtraction were 177 ±
17 and53 ±
7, respectively. This difference was found to be unlikely related to the CTD efficiency,which usually leads to a larger number of negative tracks compared to positive ones: Forexample, the number of reconstructed p (¯ p ) in the data after the DCA side-band backgroundsubtraction was 1 . × (1 . × ). Such p − ¯ p symmetry is fully accounted for by knowndifference in the tracking efficiency for positive and negative tracks.Several sources of background processes for the d sample were considered [9]: events dueto interactions of the incoming proton (or electron) beam with residual gas in the beampipe (termed beam-gas interactions) and secondary interactions of particles on materialbetween the interaction point and the central tracking detector. Extensive checks have beenperformed to exclude the first source. In particular, a special event selection was used fornon-colliding electron and proton bunches. It was found that, after the DIS event selection,a contribution from the beam-gas events is unlikely.Even for a clean DIS sample, d can still be produced by secondary interactions of particleson inactive material. One possible source for d is the reaction N + N → d + π + , where onenucleon, N , is produced by ep collisions, while the second one comes from inactive materialin front of the CTD. Secondary d can also be produced in the pickup reaction p + n → d by primary nucleon interacting in the surrounding material. Checks for such sources of DIS 2007 ackground were either negative or not conclusive due to insufficient information on theproduction cross section of the pickup process. In particular, a check was done using theHERAII data collected with the ZEUS detector equipped with a vertex detector. For the dE/dx measurement, this additional detector increased the overall material between theinteraction point and the CTD by a factor three. As expected, the production rate of d hassignificantly increased due to a stronger contribution from spallation processes. However,the rate of ¯ d after the DCA background subtraction was the same as for HERAI.
ZEUS /M T p R a t i o -4 -3 -2 -1 ) >1GeV d/p (ZEUS, Q ) >1GeV (ZEUS, Qp/d p) g (H1, p/d (a) /M T p R a t i o ) >1GeV /p (ZEUS, Qp ) >1GeV /d (ZEUS, Qd (b) Figure 2. Left: d/p , ¯ d/ ¯ p , ¯ p/p and ¯ d/d production ratios as a function of p T /M . Right:Comparison of the B values extracted from DIS with other world measurements.The detector-corrected production ratios as a function of p T /M are shown in Fig. 2(left).For the antiparticle ratio, there is a good agreement with the H1 published data for photo-production [7], as well as with pp data [6]. The production rate of d is higher than that of ¯ d at low p T . If (anti)deuterons are produced as a result of a coalescence of two (anti)nucleons,then one should expect that the ¯ d/d ratio is approximately equal to the (¯ p/p ) ratio for thesame p T per (anti)nucleon, assuming the same source radius for particles and antiparticles.In this case, one should expect ¯ d/d ≃
1, since the measured ¯ p/p ratio is consistent withunity. Under the assumption that secondary interactions do not produce an enhancementat
DCA = 0 for the d case, the result would indicate that the relation between ¯ d/d and(¯ p/p ) expected from the coalescence model with the same B for particles and antiparti-cles does not hold in the central fragmentation region of ep DIS collisions. In terms of thecoalescence model, the d production volume in momentum space is larger than that for ¯ d .For collisions involving incoming baryon beams, there are several models [10] whichpredict p − ¯ p production asymmetry in the central rapidity region. The predicted asymmetrycan be as high as 7% due to the presence of the incoming proton. As shown in Fig. 2(left),the experimental data for p (¯ p ) are not sufficiently precise to confirm such expectations.There are no predictions for the d − ¯ d asymmetry. It is possible that theoretical expec-tations for such compound states are different than those for the p − ¯ p asymmetry, since d ( ¯ d ) are not contaminated by a large contribution from the standard baryons produced inquarks and gluon fragmentation.The production of d ( ¯ d ) was studied in terms of the coalescence model. The B valuesextracted in the region 0 . < p T /M < . DIS 2007 urement of B has larger experimental errors than those in the studies of the productionasymmetries, it is still seen that B for d tends to be higher than that for ¯ d . The valuesof B for ¯ d are in agreement with the measurements in photoproduction [7], but disagreewith the B measured in e + e − annihilation [3] at the Z resonance. In contrast to heavy-ioncollisions, where B strongly increases as a function of p T due to an expanding source, the B measured in ep does not strongly depend on p T /M [9].In summary, the first observation of (anti)deuterons in ep collisions in the DIS regime ispresented. The yield of d ( ¯ d ) is three orders of magnitude smaller than that of p (¯ p ), whichis in broad agreement with other experiments. The production rate of p is consistent withthat of ¯ p , while the production rate of d is higher than that for ¯ d for the same kinematicregion. References [1] S. Butler, C. Pearson, Phys. Rev. , 836 (1963).[2] M. Karliner, B. R. Webber, JHEP , 045 (2004).[3] ARGUS Collaboration, H. Albrecht, et al., Phys. Lett.
B157 , 326 (1985);ARGUS Collaboration, H. Albrecht, et al., Phys. Lett.
B236 , 102 (1990);OPAL Collaboration, R. Akers, et al., Z. Phys.
C67 , 203 (1995);ALEPH Collaboration, S. Schael, et al., Phys. Lett.
B639 , 16 (2006).[4] G. Gustafson, J. Hakkinen, Z. Phys.
C61 , 683 (1994).[5] IHEP-CERN Collaboration, F. Binon, et al., Phys. Lett.
B30 , 510 (1969);Y. Antipov, et al., Phys. Lett.
B34 , 164 (1971);J. Cronin, et al., Phys. Rev.
D11 , 3105 (1975).[6] B. Alper, et al., Phys. Lett.
B46 , 265 (1973);BRITISH-SCANDINAVIAN Collaboration, W. Gibson, et al., Nuovo Cim. Lett. , 189 (1978);V. Abramov, et al., Sov. J. Nucl. Phys. , 845 (1987).[7] H1 Collaboration, A. Aktas, et al., Eur. Phys. J. C36 , 413 (2004).[8] M. Aoki, et al., Phys. Rev. Lett. , 2345 (1992);NA52 (NEWMASS) Collaboration, G. Appelquist, et al., Phys. Lett. B376 , 245 (1996);STAR Collaboration, C. Alper, et al., Phys. Rev. Lett. , 262301 (2001);E802 Collaboration, L. Ahle, et al., Phys. Rev. C57 , 1416 (1998);NA44 Collaboration, I. Bearden, et al., Eur. Phys. J.
C23 , 237 (2002);PHENIX Collaboration, S. S. Adler, et al., Phys. Rev. Lett. , 122302 (2005).[9] ZEUS Collaboration, S. Chekanov, et al., (2007). DESY-07-070, hep-ex/0705.3770.[10] B. Kopeliovich, B. Povh, Z. Phys. C75 , 693 (1997);B. Kopeliovich, B. Povh, Phys. Lett.
B446 , 321 (1999);G. Garvey, B. Z. Kopeliovich, B. Povh, Comments Mod. Phys. A2 , 47 (2001);S. Chekanov, Eur. Phys. J. C44 , 367 (2005);F. Bopp, Y. Shabelski, Phys. Atom. Nucl. , 2093 (2005).4, 2093 (2005).4