Newtonian limit of the singular f(R) gravity in the Palatini formalism

Recently D. Vollick [Phys. Rev. D 68, 063510 (2003)] has shown that the inclusion of the 1/R curvature terms in the gravitational action and the use of the Palatini formalism offer an alternative explanation for cosmological acceleration. In this work we show not only that this model of Vollick does not have a good Newtonian limit, but also that any f(R) theory with a pole of order n in R=0 and (d2f/d2R)(R0)≠0, where R0 is the scalar curvature of background, does not have a good Newtonian limit