32. Eye on ArXiv: 30 Apr 2008 – Curvature perturbation from false vacuum inflation

Jinn-Ouk Gong, Misao Sasaki, “Curvature perturbation spectrum from false vacuum inflation”

Another very nice and technical paper by Misao Sasaki and his collaborator. Let us recall how inflation developed in the old scenario introduced by Alan Guth back in 1981. Inflaton field was trapped in a local (false) minimum of its potential, so the Universe was exactly de Sitter-like during inflationary stage. Since the minimum was false (and true minimum corresponded to the Minkowski spacetime, very much in the spirit of KKLT or, better to say, KKLT was written very much in the spirit of old Guth’s scenario ), inflaton has to tunnel into the true minimum quantum mechanically (along classically fobidden Coleman-de Luccia trajectory). As we know now, in this scenario it is impossible to get a Universe with homogeneity and isotropy at the level we observe today, but let us forget about this issue for a moment…

If the inflaton is trapped in a local false minimum, how to calculate the spectrum of primordial perturbations which one observes in the end of de Sitter stage (after tunneling)? Naively, one has for the curvature perturbation

,

but inflaton does not move classically and .  This is  a well known  situation in exact de Sitter space,  where  curvature perturbation    is the  pure  gauge,  since  its  effective action is proportional to the slow roll parameter and the latter is zero near the true minimum of the inflaton ponential. One may say that inflation is in the eternal regime of self-reproduction, which may end in some Hubble patches only by tunneling into Minkowski vacuum.

Let us model the end of de Sitter stage in different stage: by introducing a thermal correction to the inflaton mass: , where . At early stages is small and the effective mass is positive (inflaton is trapped near the minimum), while at late stages effective mass squared becomes negative, and the inflaton rapidly reaches the true minimum of the potential at .

The authors calculate exact QFT Green function of for such model and pair correlation function of Bardeen potential (from the corr. func. of energy-momentum tensor). Resulting power spectrum of the curvature perturbation turns out to be strongly blue: .

What is the main lesson of this paper?  I guess, it is already learn by many: you generate curvature perturbation only if you explicitly breake down de Sitter invariance (for example, by introducing a time-dependent mass for the inflaton). In this case such things as deviate from  zero  (they are small w.r.t. a parameter which controls breakdown of de Sitter symmetry)  and  one can use old well known formulae to calculate curvature perturbation.