Production of D^*-Mesons with Dijets in Deep-Inelastic Scattering at HERA

Charm is produced at HERA predominantly in pairs of charm and anti-charm quarks by boson-gluon-fusion between a photon originating from one of the the positrons of the positron beam and a gluon originating from one of the protons of the proton beam. A full understanding of this interaction is only possible if the kinematic properties of the two outgoing quarks are completely determined. Charm quarks can transform into observable but different final states consisting of charmed and other hadrons. One experimentally advantageous transformation is into a D^* meson with its decay into D⁰π→Kππ. "Tagging" a charm quark this way results in high precision reconstruction of kinematic quantities (compared e.g. with the tagging of D^* mesons in semileptonic decay modes), but limits the available statistics of data considerably due to small decay ratios of that channel. This is the reason why an analysis requiring a double tag, i.e. two D^* mesons reconstruted in the very same event is not persued.

In this analysis the charm quarks in events in which a single reconstructed D^* meson is found, are approximated by their signals in the detector, namely by two jets of particles, of which one (that one in which the the D^* meson participates) is identified with one charm quark and the other one with the second charm quark. This ansatz gives access to a much larger data sample than it would be possible with double tagging and thus allows to study the production mechanism of charm at HERA in detail.

Being able to measure the kinematics in detail gives access to the production mechanism of charm and allows to seperate direct and resolved processes. In direct processes the complete momentum of the photon from the positron is interacting in the hard subprocess, while for resolved processes only a part of the momentum x_γ is interacting. It can be observed that the behaviour of x_γ can be described in deep inelastic scattering by two different theoretical approaches, one following the NLO-DGLAP approach (HVQDIS), the other one implementing the CCFM evolution equation using unintegrated gluon densities (CASCADE). The data indicate that an additional resolved component, described by a structure function of the photon, is not needed in DIS as it is in photoproduction.

As the boson gluon fusion process depends on a gluon from the proton charm production at HERA allows also to deduce the gluon density in the proton. The present analysis confirms but does not yet allow to distinguish between various currently available parton densities obtained from other processes for statistical reasons. Taking into account these results in global fits will however improve their precision.

Another interesting aspect of the production process of charm in DIS at HERA is the distribution of the azimuthal angle between the two jets Δφ. In leading order this distribution consists of a delta peak at 180 degrees, i.e. all contributions away from this value are due to higher order effects (and experimental ones). As in various other physical processes studied at HERA (charm production with jets in photoproduction, dijets in DIS) disagreement shows up between data and the NLO-DGLAP approach (HVQDIS) as well as the model based on the CCFM evolution equation using unintegrated gluon densities (CASCADE). The former lacks higher order contributions at small angles, while the latter overestimates the contributions in that region most probably due to a too broad unintegrated gluon density.