Abstract

One of the physics goals of the COMPASS experiment at CERN was to measure the contribution of gluons to the nucleon spin. To achieve this, it was proposed to scatter polarized \Value{160}{GeV/c} muons on a polarized deuteron target and to detect $D$ mesons in the final state. The underlying process in this $D$ meson production is supposed to be the \Gls{PGF}, where a virtual photon emitted by the muon interacts with a gluon from the target nucleon, producing a charm-anticharm quark pair. Fragmentation of a charm (anticharm) quark leads with high probability to the creation of a $\DZ$ or $D^*$ meson, which COMPASS detects via the $D^0 \to K\pi$ and $D^{*} \to D^0 \pi \to K\pi\pi$ decay modes. From the longitudinal cross section spin asymmetries of the $D$ meson production and theoretical predictions for the \Gls{PGF} cross section, the gluon contribution to the nucleon spin has been measured by the COMPASS experiment. The results presented in the thesis are the following. Based on data from the year 2004 a total visible cross section of \TotalCrossSection, for the $D^*$ meson production, has been measured, with the error being dominated by systematic effects. It is validated that the $D$ mesons are indeed produced through the \Gls{PGF} process, by comparison of measured $D$ meson kinematic distributions to the ones predicted by a theory (AROMA generator). A good agreement was found for the distribution shapes, which confirms that \Gls{PGF} plays a major role. However, a $20\%$ difference was found in the number of produced $\DZ$ and $\aDZ$ mesons (and for the $D^{*+}$ and $D^{*-}$ mesons as well) which is significantly larger than predicted by AROMA. Kinematic distributions of $D^0$ and $D^*$ mesons were compared with the background and also with the nearby $\KTS$ resonance, using all longitudinal data taken in 2002-2006. The particle-antiparticle asymmetry has been studied as a function of several kinematic variables. The $20\%$ excess of mesons decaying into $K^+\pi^-$ over mesons decaying into $K^-\pi^+$ was observed for all three mesons. The behavior of the $\DZ/\aDZ$ (and $D^{*+}/D^{*-}$) asymmetries as a function of virtual photon energy suggests that associated production of $D^{*-}\Lambda_c^+$ or $\overline{D^0}\Lambda_c^+$ may be responsible for the observed effect. The longitudinal double spin asymmetries have been studied for the $D^0$, $D^*$ and $\KTS$ mesons separately for particle, antiparticle and for the sum of particle and antiparticle. It was found that the asymmetries extracted for $D^0$ and $D^*$ mesons are compatible with zero. A 3-sigma deviation from zero asymmetry was observed for the $\KTS$ meson. An investigation of the $\KTS$ double spin asymmetry reveals a dependence as function of the Bjorken $x_{Bj}$ variable.