
the DM candidate is never thermalized due its extremely weak coupling to the SM, so weak
that they evade the current accelerator constraints.
In the early Universe, FIMPs can be produced from either the decay or annihilation
of states in the visible sector. When the SM temperature becomes smaller than the typical
mass scale of the interaction (i.e. the maximum of the DM and the mediator mass), the
generation process becomes suppressed, leaving a constant comoving DM number density.
Such a scenario is often referred as the freeze-in mechanism [32]. In contrast to the “WIMP-
miracle” which produces the observed relic density with near weak-scale couplings and masses,
a “FIMP-miracle” occurs when one considers renormalizable couplings of order ∼ O(10−11)
independent of the mass of the DM. If a priori such couplings seem unatural, UV versions
of the freeze-in mechanism may invoke effective couplings, suppressed by a large mass scale
above the temperature of the thermal bath. This can be achieved via non-renormalizable
operators [35], suppressed by a high mass scale, e.g., in models where the mediators between
the visible sector and the dark sectors are very massive. This is the case in unified theories
like SO(10) with a heavy Z0gauge boson [33,36,37], moduli fields [38], high scale SUSY [39–
43] or heavy spin-2 constructions [44]. In other examples, freeze-in of DM may proceed via
loops [45,46] or 4-body final states [47]. All of these scenarios are particularly interesting, as
the DM yield is sensitive to the highest temperature Tmax reached by the SM plasma [48–53],
controlled by the dynamics of the inflaton decay.
Even feebler interactions are possible when the only effective coupling at the UV scale
is gravity. Indeed, the minimal irreducible interaction that should exist between DM and
the Standard Model (SM) is mediated by graviton exchange [44,54–71] which can lead to
the observed amount of DM through the scattering of the particles in the thermal bath or
directly through the gravitational transfer of the energy stored in the inflaton condensate, as
already been discussed in detail Refs. [65–69].
DM requires an extension to the SM, but it is not the only reason why an extension
is necessary. As is well known, the visible or baryonic matter content of the Universe is
asymmetric. One interesting mechanism to produce the baryon asymmetry of the Universe
(BAU) via the lepton sector physics is known as leptogenesis [72], where, instead of creat-
ing a baryon asymmetry directly, a lepton asymmetry is generated first and subsequently
gets converted into baryon asymmetry by the (B+L)-violating electroweak sphaleron tran-
sitions [73]. In thermal leptogenesis [74–77], the decaying particles, typically right-handed
neutrinos (RHNs), are produced thermally from the SM bath. However, the lower bound
on the RHN mass in such scenarios (known as the Davidson-Ibarra bound), leads to a lower
bound on the reheating temperature TRH &1010 GeV [78] so that the RHNs can be produced
from the thermal bath. One simpler alternative is the non-thermal production of RHNs [79–
83] originating from the decay of inflaton. This interaction is necessarily model dependent
as it depends on the Yukawa interaction between the inflaton and the RHNs.
In addition to providing the DM abundance, gravitational interactions can also be the
source of baryogenesis. As shown in [84], it is possible to have a model-independent theory
of non-thermal production of RHNs from inflation, once the inflaton potential is specified.1
The abundance of RHNs is calculated in the same manner as the dark matter abundance
and can lead to observed BAU from the out-of-equilibrium CP violating decay of the RHNs,
produced during the reheating epoch.
1The simultaneous generation of gravitational DM and the baryon asymmetry was also discussed in [85].
Our results differ, as their choices of parameters are in conflict with the tensor-to-scalar ratio bound from
Planck.
– 2 –