E-beam is certainly much easier. Any PVD evaporation of oxides will result in usually a nonstoichiometric oxygen deficient film. Some e-beam evaporators may be equipped with oxygen plasma lines to assist with oxide PVD. If you want to deposit stoichiometric films, consider using oxygen plasma assisted PVD if it is available.
I would recommend (a) sputtering of TiO2 or (b) e-beam evaporation of Ti followed by annealing in oxygen or simply in air atmosphere. However, both approaches will require some optimization with regard to your application.