Diffractive photoproduction of dijets was measured with the ZEUS detector at the ep collider HERA using an integrated luminosity of 77.2 pb-1. The measurements were made in the kinematic range Q^2 < 1 GeV^2, 0.20 < y < 0.85 and x_pom < 0.025, where Q^2 is the photon virtuality, y is the inelasticity and x_pom is the fraction of the proton momentum taken by the diffractive exchange. The two jets with the highest transverse energy, E_T^jet, were required to satisfy E_T^jet > 7.5 and 6.5 GeV, respectively, and to lie in the pseudorapidity range -1.5 < eta^jet < 1.5. Differential cross sections were compared to perturbative QCD calculations using available parameterisations of diffractive parton distributions of the proton.
Inclusive photoproduction of $\dspm$ in ep collisions at HERA has been measured with the ZEUS detector for photon-proton centre of mass energies in the range \linebreak \wrang and photon virtuality Q~2 < 4 \g2. The cross section $\sigma_{ep \to \ds X} $ integrated over the kinematic region \ptrangand \etarang is {\xsecs}. Differential cross sections as functions of $p_{\perp}~{\ds}$, $\eta~{\ds}$ and W are given. The data are compared with two next-to-leading order perturbative QCD predictions. For a calculation using a massive charm scheme the predicted cross sections are smaller than the measured ones. A recent calculation using a massless charm scheme is in agreement with the data.
The photoproduction of beauty quarks in events with two jets and a muon has been measured with the ZEUS detector at HERA using an integrated luminosity of 110 pb$^{- 1}$. The fraction of jets containing b quarks was extracted from the transverse momentum distribution of the muon relative to the closest jet. Differential cross sections for beauty production as a function of the transverse momentum and pseudorapidity of the muon, of the associated jet and of $x_{\gamma}^{jets}$, the fraction of the photon's momentum participating in the hard process, are compared with MC models and QCD predictions made at next-to-leading order. The latter give a good description of the data.