Context. Observations indicate that in young stellar clusters the binary fraction for
massive stars is higher than for solar mass stars. For the Orion Nebula Cluster
(ONC) there is a binary frequency of ~50% for solar-mass stars
compared to 70-100% for the massive O- and B-stars.
Aims. We explore the reasons for this discrepancy and come up with two possible answers:
a) a primordially higher binarity of massive stars could be inherent to the star
formation process or b) the primordial binary rate might be the same for solar-mass
and massive stars, but the higher capture cross section of the massive stars possibly
leads to the formation of additional massive binaries in the early cluster development. Here we
investigate the likelihood of the latter scenario in detail using the ONC as an example.
Method. Method.N-body simulations are performed to track the capture events in an ONC-like cluster.
Results. We find that whereas low-mass stars rarely form bound systems through capture, the dynamics
of the massive stars - especially in the first 0.5 Myrs - is dominated by a rapid
succession of "transient binary or multiple
systems".
In observations the transient nature of these systems would not be apparent, so
that they would be rated as binaries. At 1-2 Myrs, the supposed age of the ONC, the
"transient" massive systems become increasingly stable,
lasting on average several 106 yrs. Despite the ONC being so young, the observed binary
frequency for massive stars --- unlike that of solar-mass stars --- is not identical to the
primordial binary frequency but is increased by at least 10-15% through dynamical interaction
processes. This value might be increased to at least 20-25% by taking disc effects into account.
Conclusions. The primordial binary frequency
could well be the same for massive and solar mass stars because the observed difference
can be explained by capture processes alone.