Previous Article
Next Article
Table of Contents

Lifting the Veil on Young Massive Stars (1Dec94)
(from NOAO HIGHLIGHTS!, NOAO Newsletter No. 40, 1 December 1994)

Massive stars spend the beginning of their main sequence lifetimes buried
behind 10 to 20 magnitudes of visible extinction in the molecular clouds
from which they formed. The true nature of the stellar nurseries for massive
stars has, for the most part, been obscured, as spectroscopic studies in the
optical regions have been precluded. Under the normal galactic extinction
law, interstellar absorption at 2 um (K-band) is 1/12th that at 4500 �. If a
consistent system of photospheric lines in hot stars could be identified at
2 um, they could be used to identify, classify and thereby quantify the
spectra of massive stars otherwise shrouded by dust.

Margaret Hanson and Peter Conti (Colorado) set out to create a 2 um
spectral classification system using nearby, unobscured normal OB stars.
Using a combination of hydrogen, He I and He II lines, they found a well
behaved variation of stellar features that would allow the temperature of
thestar to be estimated fairly accurately (Figure 1).

                           [Figure not included]

        Figure 1. High resolution (R ~1000) 2 um spectra showing the
        spectral sequence in dwarf O and early-B stars. The typical
        integration times were from 30 seconds (for K=4) to 5 minutes
        (K=7) on the 1.5-m. Signal-to-noise achieved is ~70-80.

With a classification system in hand, they set out to identify the stellar
content of M17, a well known, heavily shrouded, H II region.

The stars responsible for the ionization of the M17 nebula have never been
clearly identified. Radio recombination line surveys at 6 cm indicate that the
production rate of Lyman continuum photons is 5 X 10^50 s^-1. That
equates to 50 O7V type stars! Deep, high spatial resolution images of M17
at the near infrared bands of J (1.25 şm), H (1.65 şm) and K (2.2 şm)
obtained by Ian Gatley and Michael Merrill (NOAO) show tens of stars and
strong extended nebulosity not visible in the optical. The extinction toward
these stars is on the order of 15 magnitudes at V!

Hanson and Conti obtained 2 um spectra of the brightest heavily reddened
stars in the cluster in June 1994, using the OSIRIS instrument on the 4-m at
CTIO. Gatley and Merrill obtained follow up spectra in September 1994
using the newly upgraded IR Cryogenic Spectrometer (CRSP) on the KPNO
1.3-m telescope. The initial low resolution spectroscopic survey identifies 5
new O-type stars in the M17 region (Figure 2). The locations of the hot stars
found in M17 are shown in a K-band image of the region in Figure 3.

                           [Figures not included]

        Figure 2. 2 um spectra of the new hot stars found buried in the M17
        cluster. Spectra of known O stars are also shown for direct
        comparison.

        Figure 3. An image of the young star forming region M17, taken at
        2 şm. The newly found hot stars are indicated. Image provided by
        Ian Gatley and Michael Merrill.

Somewhat unexpectedly, Hanson and Conti found a second class of stars. A
large number of the brightest 2 um stellar sources show excess 2 um
emission relative to their J and H magnitudes when compared to normal
reddened stellar atmospheres. Spectra taken of these stars show no
photospheric absorption lines at all, unlike normal stars. Instead, these stars
appear to be completely featureless (with strong upper limits on the strength
of their Brackett ş absorption) or they show pronounced CO bandheads in
emission (Figure 4).

Hanson and Conti believe that these objects are still enshrouded in a dense
pre-main sequence (PMS) disk or shell, which is being picked up in the
K-band as excess emission, causing the underlying star to be completely
obscured. In some cases the disk or shell is warm and dense enough (ş
~10^10 cm^-3) to excite the CO overtone bandheads into emission. Hanson
and Conti plan to obtain additional spectra to better determine the spectral
classes of the new found O stars and to search deeper in the cluster for an
underlying B-star population that is expected to be present. A lack of normal
main sequence stars could explain the large number of very luminous PMS
stars found.

                            [Figure not included]

        Figure 4. The CO overtone emission line stars found in M17 within
        the giant molecular cloud.

Previous Article
Next Article
Table of Contents