Observational Evidence For Self-Interacting Cold Dark Matter
Observational Evidence For Self-Interacting Cold Dark Matter
Observational Evidence For Self-Interacting Cold Dark Matter
24 APRIL 2000
3760
0031-90070084(17)3760(4)$15.00
21
l
225
2 mx
sXX 8.1 3 10
cm
,
(1)
GeV 1 Mpc
intriguingly similar to that of an ordinary hadron. (In this
paper, we consider the case of dark matter particles scattering only from themselves but, in a forthcoming paper, we
consider the possibility that dark matter is a stable, neutral hadron.) If the dark matter particles scatter through
strong interactions similar to low-energy neutron-neutron
scattering, then the cross section is s 4pa2 , where a is
the scattering length. For neutrons, the scattering length is
more than 100 times its Compton wavelength. Using the
estimate a 100fmx21 , we obtain
13
l
mx 4
f 23 GeV .
(2)
1 Mpc
Alternatively, the self-interaction may be weak but longer
range, as in the case of the exchange of a light intermediate vector boson of mass my , in which case the cross
section is s ay mx2 my4 . The mass of the vector boson must be large enough that there is no dissipation
when dark matter particles scatter; this requires that my .
24 APRIL 2000
450 eV mx 1 GeV y200 kms2 , where y is the typical velocity of dark matter particles in the halo. This mass
scale for my corresponds to a force that is short range compared to the dark matter interparticle spacing (about 1 cm
in the halo). Hence, we need consider only two-body interactions in our analysis. If my gmx and ay O1,
then the maximum dark matter mass is
13
l
(3)
g243 MeV .
mx , 80
1 Mpc
Beyond what is expressed in the relations above, there is no
significant constraint on how light the dark matter particles
can be.
The strong self-interaction might occur if the dark matter
consists of particles with a conserved global charge (such
as a hidden baryon number) interacting through a hidden
gauge group (e.g., hidden color). If the gauge group is
unbroken, then the particles experience strong interactions
which can be nondissipative but the particle number is conserved. M-theory and superstrings, for example, suggest
the possibility that dark matter fields reside on domain
walls with gauge fields separated from ordinary matter
by an extra (small) dimension [33,34]. Similar scenarios
can be constructed in purely four-dimensional supergravity models. Note that, if the sum of the hidden baryon
number and the ordinary sector baryon number is zero,
then Vx mx mproton Vb 0.19 mx 41 GeV (using
Vb h2 0.02 and h 0.65). The particles we suggest
include light versions of Q-balls [35].
How does the mean free path of the dark matter particles affect astrophysics? Since interactions alter only
the evolution of cold dark matter when the density inhomogeneities are large, the cosmic microwave background
and large-scale power spectrum measurements are not sensitive to the self-interactions. So long as the dark matter
is cold (Tx mx , F, where F is the depth of the gravitational potential), the dark matter will collapse to form a
bound halo regardless of its collisional properties. If the
dark matter mean path were much longer than 1 Mpc,
the typical dark matter particle would not experience any
interactions as it moves through a halo. In this regime,
the usual, triaxial cold dark matter halo with dense core
forms through gravitational collapse. On the other hand,
if the dark matter mean free path is much smaller than
1 kpc, then the dark matter behaves as a collisional gas and
this alters the halo evolution significantly. The dark matter
will shock: this will heat up the low entropy material that
would usually collapse to form a core and produce a shallower density profile. Since collisions tend to make the
dark matter velocity distribution isotropic, the halo cannot
be triaxial and will be elliptical only if flattened by significant rotation. Since dark halos form with little angular
momentum, if the dark matter is not dissipative, then all
halos will be nearly spherical. X-ray observations of clusters [36] reveal that most halos are moderately ellipsoidal.
This implies that the collision time scale for dark matter
3761
24 APRIL 2000
24 APRIL 2000
3763