For the future neutrino oscillation experiment DUNE, liquid argon time projections chambers with a fiducial mass of 10 kton each are foreseen. The dual phase concept is one of the two implementations considered, wherein electrons produced by ionization in the liquid are extracted to a gaseous region above the liquid where they are amplified. For the amplification, large electron multipliers will be used. The technology was tested in various prototypes, most recently with a 3 x 1 x 1 m$^3$ large setup. An even larger prototype of 6 x 6 x 6 m$^3$ is currently being constructed and will start operation in 2019. An intensive R&D program was carried out with the focus on achieving an effective gain of at least 20. In the simulation study here presented for the first time not only the electron signal is considered but also the ion backflow and the expected production of secondary scintillation light is studied, because the latter might limit the capability of the detector to trigger on low energetic no-beam physics. It is found that the ion backflow and the light yield can be expected to be very large. The results for the effective gain show a discrepancy with experimental data, both in size and shape of the gain curve. Based on literature studies, it is argued that photon feedback contributes to the gain in detectors filled with pure noble gases, especially in the case of pure argon.