Leading Neutron Production at ZEUS W. Schmidke Max Planck Institute for Physics Munich, Germany ZEUS results on leading neutron production in ep scattering are presented. Production in deep inelastic scattering and photoproduction are compared, giving indications of absorption. The data are compared to Monte Carlo, meson exchange and absorption models. 1 Introduction Events with a neutron carrying a large fraction of the proton energy have been observed in ep scattering at HERA [1]. The dynamical mechanisms for their production are not completely understood. They may be the result of hadronization of the proton remnant, conserving baryon number in the final state. Exchange of virtual isovector mesons is also expected to contribute, predominantly the exchange of low mass π+ mesons [2]. In this picture the proton fluctuates into a virtual n-π+ state. The virtual π+ scatters with the projectile lepton, leaving the fast forward neutron in the final state. Depending on the virtuality of the exchanged photon, which is a measure of how pointlike the photon is, the neutron may also rescatter with it and migrate to a region outside of the detector acceptance, leading to a depletion of neutrons in some kinematic regions [3]. The ZEUS experiment at HERA had a for- ward neutron calorimeter (FNC) in the pro- ZEUS L ton beam direction. It measured the frac- -1 2 2 2 ZEUS 6 pb pT < 0.476 xL GeV 2 -4 2 tion of the beam energy carried by the neu- /dx 0.2 〈Q 〉 = 4 × 10 GeV -1 LN ZEUS 40 pb σ 2 2 0.175 〈Q 〉 = 2.7 GeV tron, xL, and the transverse momentum trans- 2 2 )d 〈Q 〉 = 8.9 GeV 2 2 〈Q 〉 = 40 GeV ferred to the neutron, pT . HERA machine el- inc 0.15 ements along the neutron flight path limited σ (1/ 0.125 neutron scattering angles to θ < 0.75 mrad, n 0.1 restricting measurement to the kinematic re- 2 2 2 0.075 gion pT < 0.476xL GeV− . Here we report re- sults on leading neutron production, in both 0.05 photoproduction, where the photon virtual- 0.025 Systematic uncertainty 2 ity Q is nearly zero, and in deep inelastic 0 scattering (DIS), where Q2 is greater than a 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 x few GeV2. Comparisons are made to Monte L Carlo models, one of which incorporate the Figure 1: Neutron x distributions for pho- pion exchange mechanism. Comparisons are L toproduction and three Q2 bins of DIS. also made to models of pion exchange incor- porating rescattering, and a model including exchanges of additional mesons. DIS 2007 691 2 Results Leading neutron xL distributions for photo- ZEUS ρ 1.4 2 production and three Q bins of DIS are shown ZEUS 6 pb-1 2 2 2 pT < 0.476 xL GeV in Fig. 1. The distributions are normalized 1.2 Systematic uncertainty D' Alesio and Pirner by σ , the inclusive cross section without the -0.13 inc D' Alesio and Pirner × (1-x ) 1 L neutron requirement. The distributions all rise NSZ NSZ × (1-x ) -0.13 2 L from lowest xL due to the increase in pT space, 0.8 reach a maximum near xL = 0.7, and fall to zero at the kinematic limit xL = 1. There is a 0.6 clear increase of the relative neutron yield with 0.4 2 2 2 Q . This is consistent with absorption models, (1/σinc) dσLN/dxL(Q <0.02 GeV ) 2 ρ = where at lower Q the larger photon size can 0.2 2 2 (1/σinc) dσLN/dxL(Q >2 GeV ) result in rescattering of the neutron, where the 0 neutron can migrate to a region outside of the 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 detector acceptance and is lost. xL For further studies the full Q2 range of the DIS sample is taken together. The ratio of Figure 2: Ratio of neutron xL distributions the photoproduction xL distribution to that in photoproduction and DIS. of the full DIS sample is shown in Fig. 2. The depletion of neutrons in the photoproduction sample increases with decreasing xL. In pion + exchange models the size of the virtual n-π ZEUS system is smaller at lower xL, with increased 0.50<xL<0.58 0.58<xL<0.66 probability of rescattering. Thus the depletion 1 1 0.9 0.9 0.8 0.7 of neutrons in photoproduction is consistent 0.8 2 2 (arb. units) 0.6 Q > 2 GeV 2 T 0.7 with pion exchange models including absorp- 2 2 -1 0.5 Q < 0.02 GeV 0.6 ZEUS 6 pb 2 2 /dp Fit dσ/dp ∝ exp(-bp ) tion. 0.4 T T LN 0.5 2 σ d 0 0.025 0.05 0.075 0.1 0 0.05 0.1 0.15 The pT distributions of neutrons in photo- production and DIS are shown for several xL 0.66<x <0.74 1 0.74<x <0.82 1 L L 0.9 bins in Fig. 3. They are normalized to unity 0.8 2 0.7 at pT = 0. The lines are fits to exponentials in 0.6 2 0.5 pT , which give a good description of the data. 0.4 The photoproduction distributions are steeper 0.3 0 0.05 0.1 0.15 0 0.05 0.1 0.15 0.2 in the range 0.6 < xL < 0.9, with relatively 2 1 fewer neutrons at high p . This is qualitatively 1 T 0.82<xL<0.90 0.90<xL<1.00 consistent with absorption, where rescattering is more likely for the small n-π+ separations 2 -1 corresponding to higher pT . 10 -1 10 0 0.1 0.2 0.3 0 0.1 0.2 0.3 2 2 3 Comparison to models pT (GeV ) 2 The normalized neutron distributions in DIS Figure 3: Neutron pT distributions for pho- were fit to the form toproduction and DIS, normalized to unity 2 at pT = 0. 1 dσLN 2 bpT 2 = ae− . σinc dxLdpT 692 DIS 2007 Figure 4 shows the xL distributions for θn < 0.75 mrad and the intercepts a and slopes b as a function of xL. Also on the plots are the expectations of the Lepto [4] and Rapgap [5] Monte Carlo models. Both were run in the configuration where neutron were produced from the fragmentation of the proton remnant [6]. Lepto was also run implementing soft color interactions [?]; Rapgap was also run including neutron production via pion exchange. Both models with only fragmentation of the proton remnant do not describe the data. They predict too few neutrons, peaked at lower xL than the data. They do not have the broad plateau in intercepts a that the data exhibit for medium values of xL, and they predict slopes b smaller than and without the steep xL dependence of the data. Lepto including soft color interactions gives an xL distribution which better describes the peak in the data and has a slight enhancement of a at medium xL value. Rapgap with pion exchange gives a good description of the shapes of all distributions, although it predicts more neutrons with larger intercepts and slopes than the data. The ratio of xL distributions photoproduc- tion over DIS in Fig. 2 also shows the predic- ZEUS L 0.25 tions of two models of absorption [3]. The ZEUS 40 pb-1 (a) 2 2 /dx 0.2 Q > 2 GeV model of D’ Alesio and Pirner describes the LN Systematic σ uncertainty shape of the data, with increasing absorption )d 0.15 inc at decreasing values of xL, but it predicts a σ 0.1 lower value of the ratio than the data. γp in- (1/ 0.05 2 2 2 teractions have a power law dependence on the pT < 0.476 xL GeV λ ) 0 photon-proton center-of-mass energy, σ W , -2 2 pT = 0 (b) with different values of λ for DIS and∝ pho- 2 toproduction. Assuming that γπ interactions a (GeV have the same dependence, and recalling that 1 √ RAPGAP-π Wγπ = 1 xLWγp, the ratio of photopro- RAPGAP std. frag. − LEPTO-SCI duction and DIS cross sections is proportional LEPTO std. frag. ∆λ ) 0 to (1 x ) . After correcting the absorption -2 L p2 < 0.476 x2 GeV2 − T L (c) factor of D’ Alesio and Pirner for this differing 10 W dependence the model give a good descrip- b (GeV 7.5 tion of the data. 5 Figure 5 shows the slopes b for DIS, and the difference between slopes in photoproduc- 2.5 tion and DIS, ∆b. Also shown are the pre- 0 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 dictions of a model of pion exchange incorpo- xL rating neutron absorption and migration [8], and a model including in addition exchanges Figure 4: Leading neutron xL distributions of ρ and a2 mesons [9]. The model with only and intercepts and slope for DIS, compared pion exchange predicts too high a value of the to Monte Carlo models. slopes, but does predicts the correct magni- tude of ∆b. The model including also ρ and a2 exchanges gives a good description of the slopes, including the turnover at highest xL. It also gives a fair prediction of the value of ∆b. DIS 2007 693 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 4 Summary Leading neutron energy and pT distributions were measured in DIS and photoproduction.