Drell-Yan and Bethe-Heitler processes

In our present analysis, e⁺e^- and $\mu^+\mu^-$ production in ep collisions (E_e=26.7 GeV, E_p=820 GeV) was simulated by PYTHIA5.7 (for DY leptons) and LPAIR [17] (for BH leptons). This allows us easily to apply all the necessary cuts to account for detector resolution and detector geometry. The results presented here are correspond to the geometry of the ZEUS detector and are based on samples of $2\cdot10^4$ events (for DY pairs) and 10⁵ events (BH pairs).

To study the effects of different selection criteria on the background suppression, cuts on the the polar angle $\theta_l$ and energy E_l of produced leptons were defined as follows: 1L$_e\equiv 2.2^{\circ}<\theta_{e}<176.5^{\circ}$;  1L$_{\mu}\equiv 5^{\circ}<\theta_{\mu}<170^{\circ}$;  2L$_e\equiv 2.2^{\circ}<\theta_{e^+}\,,\, \theta_{e^-} <176.5^{\circ}$;  2L$_{\mu}\equiv 5^{\circ}<\theta_{\mu^+}\,,\, \theta_{\mu^-} <170^{\circ}$;  1E $\equiv E_l \gt 1$ GeV, 1E5 $\equiv E_l \gt 5$ GeV, 2E $\equiv E_{1, 2} \gt 1$ GeV. These conditions correspond to lepton acceptances in the ZEUS detector. Two types of trigger were considered: a) ``tagged", requiring the detection of the scattered electron (5<E<sub>e'</sub><25 GeV) at very small angles, thereby limiting Q<sup>2</sup> to less than 0.02 GeV<sup>2</sup>, and b) ``untagged", requiring the absence of a detected scattered electron in the main rear calorimeter (Q²< 4 GeV²).

The total ``untagged" and ``tagged" cross sections for photoproduction of DY pairs with masses M_l⁺l^->1 GeV were found to be 84 pb and 25 pb respectively.

PYTHIA PYTHIA PYTHIA

$$Q^2\, <\,4\,GeV^2 Q^2\,<\,0.02\,GeV^2 Q^2\,<\,0.02\,GeV^2$$ TRIGGERS LPAIR

$$5\,GeV\,<\,E_{e'}\,<25\,GeV$$

No Cuts 20 000 20 000 7 125 10⁵

1L 14 301 14 099 5 903 54 569

2L 12 475 12 248 3 974 44 649

1E 15 560 15 589 5 221 5 801

1E5 8 874 9 058 2 209 1 050

$$\otimes$$ 9 824 9 693 3 997 8

$$\otimes$$ 5 771 5 609 2 439

Table 1: Numbers of events with photoproduction of an electron pair (PYTHIA 5.7 for DY and LPAIR for BH) under different trigger conditions.

Table 1 presents numbers of events with electron pairs passing the different combinations of cuts defined above. From the last column of the table it can be seen that the requirement of detecting at least one of the BH electrons in the calorimeter (1L_e$\otimes$1E) dramatically reduces the number of BH events passing this cut. With simultaneous detection of both electrons (2L_e$\otimes$2E) only about 12$\%$ of the DY electron pairs survive, but the BH background is totally eliminated. The ZEUS muon detector system has a smaller angular acceptance and so the number of DY muon pairs passing the 2L$_{\mu}$$\otimes$2E cuts is 1548 events of $20\,000$.

 

Figure 3: Pseudo-rapidity distributions of (a) DY leptons for all events (open histogram), and the subsets with applied cuts 2E (shaded, light) and 2E5 (shaded, heavy), (b) BH leptons with the same cuts as above, (c) leptons passing the 2L$\otimes$2E cuts with an integrated luminosity of 10 pb^-1. The solid and dashed histograms correspond to BH and DY leptons respectively. The acceptance of the ZEUS calorimeter is indicated.

Figure 3 illustrates these results and shows the (pseudo)-rapidity distributions of DY and BH leptons (i.e. electrons + muons) under different trigger conditions. The final plot summarizes the results of our analysis; at an integrated luminosity of 10 pb^-1, with the 2L$\otimes$2E cuts in the ZEUS detector, about 40 events can be detected with DY electron pairs and 25 events with DY muon pairs. With an integrated luminosity of 1000 pb^-1 these data samples will increase by two orders of magnitude, which should just be sufficient to enable the photon structure to be investigated.

A further feature which could be used to distinguish DY from BH events is the presence of a photon remnant. This can be quantified once again by evaluating $x_\gamma$ (in terms of the two leptons instead of jets). The BH events will have $x_\gamma\approx 1$; $x_\gamma$ must be evaluated in any case for a study of the photon structure, and a cut to remove high $x_\gamma$ events should help to remove the BH background and perhaps allow other conditions to be loosened. We have not had time to investigate this question further, but it is clear that a number of possibilities exist for studying DY pairs at HERA. All require the highest luminosities that can be obtained.