Revisiting the dynamical exponent equality $z=d$ for the dirty boson problem

Abstract

It is shown that previous arguments leading to the equality $z=d$ ($d$ being the spatial dimensionality) for the dynamical exponent describing the Bose glass to superfluid transition may break down, as apparently seen in recent simulations (Ref. \cite{Baranger}). The key observation is that the major contribution to the compressibility, which remains finite through the transition and was predicted to scale as $κ\sim |δ|^{(d-z)ν}$ (where $δ$ is the deviation from criticality and $ν$ is the correlation length exponent) comes from the analytic part, not the singular part of the free energy, and therefore is not restricted by any conventional scaling hypothesis.

Similar papers