And here;

http://arxiv.org/abs/1202.2362/

On the origin of planets at very wide orbits from re-capture of free
floating planets

Hagai B. Perets, M. B. N. Kouwenhoven
(Submitted on 10 Feb 2012)

In recent years several planets have been discovered at wide orbits
(>100 AU) around their host stars. Theoretical studies encounter
difficulties in explaining their formation and origin. Here we propose
a novel scenario for the production of planetary systems at such
orbits, through the dynamical recapture of free floating planets
(FFPs) in dispersing stellar clusters. This process is a natural
extension of the recently suggested scenario for the formation of wide
stellar binaries. We use N-body simulations of dispersing clusters
with 10-1000 and f_FFP=0.5-2 to study this process. We find that
planets are captured into wide orbits, ~100-10^6 AU, and a thermal
eccentricity distribution. Typically, 3-6x(f_FFP/1) % of all stars
capture a planetary companion (f_FFP is the number of FFP per star).
The planetary capture efficiency is comparable to that of capture-
formed stellar-binaries, and shows a similar dependence on the cluster
size and structure. The capture efficiency is almost independent of
the specific planetary mass; planets as well as sub-stellar companions
of any mass can be captured, where the capture efficiency decreases
with increasing cluster size. For a given cluster size the capture
efficiency increases with the host/primary mass. More than one planet
can be captured around the same host, and planets can be captured into
binary systems. We also expect planets to be captured into pre-
existing planetary systems as well as around compact objects, if these
formed early enough before the cluster dispersal. In particular,
stellar black holes have a high capture efficiency (>50 % and
5-10x(f_FFP/1) % for capture of stars and planetary companions,
respectively) due to their large mass. Finally, although rare, two
FFPs or brown dwarfs can become bound and form a FFP-binary system
with no stellar host through this process.

There really could be a giant planet hidden far beyond Pluto

Pluto is about forty times the distance from the Sun as Earth. But the
Solar System is over 50000 times that length across, meaning it could
be hiding some huge secrets. That's now looking like a small but real
possibility.
In recent years, astronomers have discovered a bunch of planets
located at least 100 astronomical units, or the distance from the Sun
to the Earth, away from their host stars. These planets are gas giants
- they would have to be for us to see them at all - so this is
something very different the dwarf planets like Pluto and Eris
discovered in our solar system's Kuiper Belt and beyond.

There's almost no chance that these giant planets could have formed as
part of their host star's planetary disc, considering their immense
distance away. That strongly suggests that these are former rogue
planets captured by the star's gravity. Hagai Perets of the Harvard-
Smithsonian Center for Astrophysics and Thijs Kouwenhoven at Peking
University's Kavli Institute for Astronomy and Astrophysics teamed up
to figure out just how often we can expect stars - potentially
including those like our own Sun - to capture these giant wandering
planets. ScienceNOW reports their results:

Because most stars are born with others, Perets and Kouwenhoven ran
computer simulations to see what happens when a star cluster contains
free-floating planets. If the number of free-floating planets equals
the number of stars, then 3% to 6% of the stars succeed in capturing a
planet, and some stars capture two or three. Most of the captured
planets end up hundreds or thousands of times farther from their stars
than Earth is from the sun. Furthermore, most captured planets have
orbits tilted to those of native-born planets, and half the captured
planets revolve around their stars backward.

Their work depends on having a good grasp of how many rogue planets
there really are, and we can't be sure our current estimates, which
suggest there area as many wandering planets as there are stars, is
accurate. But if these results are accurate, then our Sun, whose mass
is slightly above average, had a real chance of capturing one or more
planets eons ago.

The chances aren't huge by any stretch - probably only a few percent,
according to Perets - but if their numbers are accurate, then the
possibility of such a planet definitely exists. Coupled with some
recent, admittedly far from convincing signs pointing to such a
planet's possible existence in the Oort Cloud, the Solar System might
just have one gigantic secret waiting for us to discover.

For more, check out ScienceNOW. Read the original paper at arXiv.
Artist's conception of rogue planet via NASA/JPL.