Figure 3.
Panel (a) presents
the radial velocities
measured from the
4686 Angstrom
helium emission line
in the Gemini spectra
of M101 ULX-1.
Panel (b) illustrates
the chi-squared
value obtained
when fitting circular
orbits of different
periods to the data,
demonstrating that
the best-fitting orbital
period is 8.2 days.
Panel (c) again shows
the measured radial
velocities, now folded
over the inferred
orbital period. For
further details, please
see Liu et al., 2013.
massive star that burns brightly, fiercely, and
erractically). More detailed analysis indicated a star with spectral type WN8, with a
mass somewhere between 17.5 and 19 solar masses. For details, see the journal paper
(Liu et al., 2013).
The motion of those emission lines showed
clear radial velocity variations, indicating
that the observations had successfully detected the orbital motion of the donor star
about the black hole (see Figure 3). Assuming a circular orbit, the best-fitting orbital
period was 8.2 days (which achieved a chisquared value of 1.6). Clearly substantial uncertainties remain, given moderately large
error bars for each of the radial-velocity
measurements and imperfect phase coverage (remember that this is in a galaxy about
20 million light-years away!)
In reality, the orbit might also have non-zero
eccentricity (as often observed in the windaccreting Galactic high-mass X-ray binaries),
although the good fit to the data using a
pure sine curve suggests that any eccentric-
6
GeminiFocus
ity in this case would be small. The best-fit
model radial velocity curve indicates a minimum mass for the compact object in M101
ULX-1 of five solar masses, which confirms
that it is a black hole.
Despite those uncertainties in the precise
properties of the binary, two conclusions are
very hard to escape: the black hole in M101
ULX-1 is not an IMBH, and it accretes from
the wind of the Wolf-Rayet star.
For any binary system whose inclination is
unknown, radial velocity measurements can
only ever lead to a lower limit on the component masses — since the binary could,
in principle, be arbitrarily close to face-on
to the line of sight. However, the chance of
detecting such a system is small. For M101
ULX-1, the combination of our best-fitting
orbital period and Wolf-Rayet mass would require an orbital inclination within 5 degrees
of face-on to contain a black hole of 300 solar
masses or greater. This means we’d have a 0.3
percent probability of discovering such a system by chance. If ULXs are systems in which
January2014