Abstract

We consider asymptotically stable scalar-tensor dark energy (DE) models for which the equation of state parameter wDE tends to zero in the past. The viable models are of the phantom type today: however, this phantomness is milder than in general relativity if we take into account the varying gravitational constant when dealing with the SNIa data. We study further the growth of matter perturbations and we find a scaling behaviour on large redshifts which could provide an important constraint. In particular, the growth of matter perturbations on large redshifts in our scalar-tensor models is close to the standard behaviour , while it is substantially different for the best-fit model in general relativity for the same parametrization of the background expansion. As for the growth of matter perturbations on small redshifts, we show that in these models the parameter can take absolute values much larger than in models inside general relativity. Assuming a constant γ when γ′0 is large would lead to a poor fit of the growth function f. This provides another characteristic discriminative signature for these models.