INCLUSIVE CHARGED HADRON AND K0 PRODUCTION IN 2-PHOTON INTERACTIONS

Abstract The inclusive transverse momentum distributions of charged hadrons and K 0 's produced in tagged photon-photon collisions, are measured and compared to model calculations up to a p T of 5 GeV/ c . The relative abundance of K 0 's favor the inclusion of charm.

A process analogous to quark pair production in e÷e -annihilation is the hard scattering quark exchange process in 77 interactions. This hard scattering process is expected to dominate [ 1 ] the cross section at high transverse momenta of the produced quarks. Experimentally, one has looked for this process in two-jet [ 2 ] as well as inclusive particle [ 3,4 ] production and has found a relatively low Pr onset of hard scattering in 77 interactions. The ambiguities and variations in jet definitions at low energies make it preferable to directly measure single particle inclusive cross sections. This report examines the issue in more detail (high statistics) for charged particle inclusive production and checks if the heavier quark flavors display a similar behavior in inclusive K ° production.
The results are based on an integrated luminosity of 220 pb-i obtained with the Mark II detector at PEP operating at ECM = 29 GeV. The major features of the Mark II detector have been described else-This work was supported in part by the Department of Energy, contracts DE-AC03-76SF00515 and DE-AC03-76SF00098.
where [ 5 ]. The combined tracking information from the central drift chamber and the vertex drift chamber in a 2.3 kG magnetic field provided a momentum resolution of [a(pT)/lyr]2= (0.025)2+ (0.011pr) 2 (PT in GeV/c). The small angle tagging system (SAT) measures electrons scattered between 21 and 83 mrad from the beam axis. It consists of three layers of drift chambers with a spatial resolution of 300 lira in the xy plane, followed by three layers of acceptance defining scintillator and 18 alternating layers of -~ inch lead and ½ inch scintillator. The energ~resolution achieved for electrons is a/E= 15.5%/x/E.
We detect some 60 000 events tagged by an electron in the SAT corresponding to a Q2 interval between 0.075 and 1.00 GeV 2. Since we do not observe the entire final state, the transverse momentum of hadrons, PT, is calculated in the usual manner, with respect to the e + e-beam axis. To eliminate the large QED background from e+e-~e+e-e+e-and e+e-/l+/t -we accept only events with 3 or more charged tracks observed (in addition to the tagging electron). The Liquid Argon Calorimeter and the muon chambers are used to remove events with identified leptons. To minimize the beam-gas interaction Elsevier Science Publishers B.V. 341 background we also eliminate events with protons or deuterons identified by their time of flight. The residual background from beam-gas interactions is simulated with events produced along the beam-line outside the interaction region. The background from e + e-~ e + e-~ ÷ z-subtracted using the Monte Carlo simulation of this process. Both these backgrounds vary between 5% and 10% with/rr and are subtracted from the data. Taking advantage of the precision vertex chamber, a vertex finding program (VFINDP) is used to find K°'s which decay at least 2.5 mm from the primary vertex. After several quality cuts, tracks are combined into charge zero pairs and the points of intersection are calculated. Of the two possible crossing points in the x-y plane, the program chooses the one which gives a positive decay length and is nearest to the interaction point. The track momenta are corrected for energy loss in the material traversed and the tracks are constrained to pass through a common point in space. The n+n -pair mass distribution shows a clear K ° peak with a resolution of a= 7 MeV and an estimated background of less than 8%. We select events with masses in an interval + 20 MeV about the actual K ° mass. A Two-Photon Monte Carlo program [ 6 ] is used to simulate the hard scattering quark exchange process e+e----,e+e-q(1, or for short ~--,q#. It incorporates transverse-transverse as well as transverse-longitudinal 77 luminosity functions [ 7 ], a summation over the four quark flavors u, d, s, c, and the subsequent fragmentation of the quarks according to the LUND scheme [8 ]. For this analysis we have ignored the Vector Dominance Model (VDM) contributions. The rather steep fall-offwith PT makes this a negligible contribution above Pa-= 2 GeV/c [ 9 ]. The quark fragmentation parameters are taken as the set of values which best describe the e +e-annihilation data at 29 GeV, and no attempt has been made to optimize the parameters for the lower two-photon energies. Since the quark flavors enter with the fourth power of their charges, the most important contributions come from the u and c quarks once the available two-photon energy exceeds the respective quark pair thresholds. These Monte Carlo calculations describe the gross features of the data and are used to obtain the detection efflciencies for the various particle species. The calculations include the trigger simulation and a threshold factor [ 7 ] ln(W/2mq + x/(l,V/2mq)2--1) 2 which correctly takes into account the quark masses mq. In the case of charm production the quark mass can be of the same order of magnitude as the two-photon CMS energy W. The ef-ficienc~es rise with PT and reach about 17% for charged hadrons and 3.3% for K°'s at/rr= 2 GeV/c. As first pointed out by TASSO [ 3 ] the PT distribution of single charged hadrons does not follow a simple exponential fall-off but flattens out for Pr values above I GeV/c. Similar to the TASSO procedure, the Mark II analysis uses single tag data with Q2 values around 0.5 GeV 2, but with its higher statistics Mark II can extend the sensitive range in Pr from 3 to 5 GeV/c. In fig. 1 the inclusive charged hadron cross section multiplied by p~ is plotted as a function of/ft. After the initial steep drop the data points are consistent with being constant for pa-values above 3 GeV/c. This flat part of the distribution translates into a p¥4 behavior for da/dp 2 and is an indication of the underlying structure ofa scale-invariant pointlike process, the simplest manifestation of which is the hard scattering quark exchange process ~q#. Several authors have calculated dtr/dp 2 for this process in the quark parton model, and it has been shown earlier [ 10 ] that any ,or dependence due to quark hadronization is small compared to the p~4 power law as long as one stays away from kinematic boundaries. Therefore, we can expect for our analysis a somewhat modified p~4 behavior for the inclusive particle spectrum, if the scaled transverse momen-   turn XT =Pr//~ is small compared to 1. For the data in fig. 1 XT is in the range between 0.07 and 0.34.
Nevertheless, the fragmentation introduces some uncertainty in the comparison of our data with the underlying hard scattering process. It is therefore more appropriate to compare the data with a Monte Carlo simulation of the process 77-, q#, for which the fragmentation parameters have been adjusted in the clean environment of e+e -annihilation as mentioned above. The ratio of the measured and calculated inclusive cross sections is shown in fig. 2. The enhancement below 2 GeV/c is mainly due to vector dominance processes, which are not included in the Monte Carlo. The remarkable feature is that the ratio is approximately constant and compatible with 1 for PT values above about 2 GeV/c. This means that the hard scattering process accounts for most of the high PT hadron production. Contributions from higher order hard scattering processes with 3 or 4 quarks and/ or gluons can be important for moderate PT values and for Q2= 0. However, it has been shown experimentally [ 11 ] that any excess over the 77-*qq process decreases dramatically as Q2 increases. In addition, most detectors are less sensitive to these higher order processes, since the additional spectator quarks and/or gluons are produced close to the beam direction.
With its ability to identify K°'s over a wide momentum range the Mark II detector can be used to look for the onset of charm production, because due to its charge, strange-quark production is expected to be suppressed by a factor of 16 with respect to charm quarks. With the same absolute normalization as above, the ratio of inclusive K°'s from data and Monte Carlo is shown in fig. 3 for PT values up to 3 GeV/c. Although the errors are much larger here, the data are consistent with the same general features observed for charged hadron production. The ratio of K°'s to charged hadrons increases to about 30% at the highest PT values as shown in fig. 4. From the Monte Carlo simulations with and without charm one can ascertain that the charm quarks are responsible for about half the K ° production (difference between solid and dashed lines) at high/~r. Combining the three highest Pr bins we find R (Pr > 1.6 GeV/c) = 0.30 + 0.08. This can be compared with Monte Carlo values of 0.32 and   In an attempt to resolve the discrepancy between the Born term calculations and the untagged TASSO inclusive particle data [4 ], Aurenche et al. have performed model calculations [ 9 ] including higher order QCD terms and generalized vector dominance model contributions (motivated by photo-production data). The authors have subsequently expanded their calculations to include charm production and to extrapolate to the Mark II tagging range [ 12 ]. We have folded their single charged hadron inclusive spectrum with the Pr resolution expected for the Mark II detector. As seen in fig. 5 the data points lie systematically above the predictions of Aurenche et al.
for Pr values larger than 1.5 GeV/c. However, the single-tag TASSO data [3] (in a similar Q2 range as Mark II) show an even higher hadronic cross section.
To be sure that the high PT tail is not caused by the second final state electron scattering at large angles, we removed all events with electron candidates in the end caps or central detector. From Monte Carlo simulations we expect at most 5 electrons with PT above 2 GeV/c to go undetected in regions not properly covered by shower counters. The discrepancy between the model calculations and the Mark II data may be partially due to various effects: The tagging efficiency is only taken into account for an average  The solid line is an analytic calculation by Aurenche et al. [9,12]. The dashed line is a best fit to the data from TASSO [ 3 ] and the dash-dotted lines an approximate 1 ~ contour around the data.
Q2 whereas there exists a moderate Q2 dependence. The fragmentation may lead to uncertainties and is probably best handled with Monte Carlo methods. Analytic model calculations usually contain a number of approximations which may very well describe the overall features of a process but may lead to discrepancies in kinematic areas where the relative contributions have dropped by several orders of magnitude -as is the case for the high PT tail discussed above. The model calculations have also not included multijet processes, for which there is recent experimental evidence at TRISTAN [ 13 ] confirming earlier indications at PEP [ 14] and PETRA [ 1 l, 15 ]. However, the model calculations discussed here and our simple minded Monte Carlo simulations seem to describe different aspects (kinematic regions) of the data quite well.
In summary, we extend the investigation of inclusive charged hadron production to 5 GeV/c in/~r and find evidence for the hard scattering process 77--' qq in the PT range from 3 to 5 GeV/c. With the first observation of inclusive K ° production up to 3 GeV/c in PT, we see the enhanced production expected from charm at high PT values.