Previous Article
Next Article
Table of Contents

Probing the Galactic Center (1Mar95)
(from NOAO HIGHLIGHTS!, NOAO Newsletter No. 41, 1 March 1995)

The center of the galaxy is a complex and interesting region.  It provides the
only opportunity to study the processes in the nucleus of a galaxy on the
scale of individual stars. The Galactic Center is, however, heavily obscured
by the intervening dust lying in the plane of our galaxy, so that at visible
wavelengths it is unobservable.  Extinction by dust is much lower in the
infrared, and, the Galactic Center is well-studied at these wavelengths.

A major outstanding question is the nature and origin of the radiation field
present in the galactic center.  There is sufficient ultraviolet radiation to
ionize gas and to be absorbed and re-radiated by warm dust, but its origin is
uncertain.  One possible explanation is that it originates in an accretion disk
around the massive black hole that is thought to lie at the center of the
galaxy.  The other possibility is that the primary source of ionizing flux is
from young, massive stars.  The presence of luminous M supergiants
supports the idea that there is active star formation at the Galactic Center,
and the detection of  infrared He I and H I emission point sources without
radio counterparts (which thus must be stellar or extremely compact) offers
plausible candidates for young, hot stars.

The strong line emission from these objects implies that they are not hot,
luminous main sequence stars; several alternatives have been suggested.  One
possibility is that these are post-main-sequence stars with mass loss,
equivalent to the Ofpe/WN9 stars in the Large Magellanic Cloud.  Another is
that they are cooler stars, whose outer envelopes are ionized by external
sources of radiation.  A third suggestion is that they are the result of stellar
collisions.

The best way to settle on the classification is by means of spectroscopy -
given the high obscuration of the Galactic Center, this has to be spectroscopy
in the near-infrared.  Bob Blum, Darren DePoy, and Kris Sellgren (Ohio
State University) have carried out a program to classify these stars, using
OSIRIS on the CTIO 4-m telescope in July and September 1993 and June
1994.  Spectra were taken in the K (1.9-2.4 um) band for eight Galactic
Center He I sources and a number of galactic and LMC comparison stars;
spectra were also taken in the H (1.45-1.85 um) band for one of the Galactic
Center sources (the so-called AF or AHH source) and for the comparison
stars.  These spectra were all at a resolution of roughly 570 (see Figures
1-3); an additional higher-resolution spectrum of one of the brighter GC
sources (IRS13) was also obtained.  Blum et al. were also able to use data
from the literature to supplement their comparison spectra.

                         [Figures not included]

         Figure 1. K band spectra of He I emission line sources in the
         Galactic Center showing the He I 2.06 um and Brg
         (2.17 um) emission lines. Several sources also show He I
         emission at 2.11 um. The lines near 2.15 um, 2.22 um, and
         2.35 um in IRS are due to [Fe III].

         Figure 2. H and K ban spectra of He I emission line sources
         in the Galaxy showing the He I 1.70 um and 2.06 um lines
         and Brackett series emission lines. Reliable detections are
         made for Br(gamma) (2.17 um), Br9 (1.82 um), Br10
         (1.74 um), Br11 (1.68 um), and Br12 (1.64 um).

         Figure 3. Same as Figure 2 but for He I Emission Line
         Sources in the LMC. A spectrum of the Luminous Blue
         Variable, S Dor, which shows no He I, is shown for
         comparison. The lines near 1.69 um, 1.74 um, and 209 um
         are due to Fe II.

They caution that massive evolved stars show a range of properties in their
spectra, based on observations of objects observable over a wide wavelength
range.  This complicates attempts to classify them based on observations in
only one wavelength region.  Nevertheless, they find from their spectra that
the GC objects show a general similarity in their spectra to the comparison
objects, in the sense that the distributions of line equivalent widths and line
ratios overlap.  They find that the AF source (the only one measured at H)
has Brackett series line ratios that imply the lines are optically thick;
they are thus almost certainly produced in an outflow and not in a more
extended region.  This is also the case for most of the comparison objects.

The He I emission for a subset of the sample (two GC stars and two galactic
comparison stars) is extremely strong, and Blum et al. argue that this is not
consistent with normal He/H abundance ratios, and that the ratio may be
greater than 1 in some cases.  This kind of He enhancement is consistent
with models for the evolution of these stars.  They also find that the He I
line widths in the GC objects are wider than for the comparison objects,
which may suggest higher effective temperatures.  One of the sources (IRS1 W)
does not have any He I emission after proper background subtraction,
which indicates that this is in fact an H II region with a more normal
embedded star.

Although all these results do not lead to a simple classification of the GC
stars, it is possible to exclude a number of possible classifications.

The infrared spectra are quite different from those of luminous blue variables
(LBV) and Wolf-Rayet stars (both types WC and WN).  The emission lines
are also much stronger than those of normal O and B stars.  Thus, while the
results strongly support the idea that the GC objects are massive, early-type,
mass-losing stars, a precise classification is not yet possible.  Calculations
of allowable luminosities and temperatures, based on the observed lines and
photometry, do show that these stars lie well above the main sequence.

Blum et al. state that the solution to the classification problem is to extend
the wavelength coverage of the classification spectra, and they will be back
at CTIO in June (this time with the upgraded IRS) to obtain additional
spectra in the 3 um region for the brighter GC objects.

Previous Article
Next Article
Table of Contents