1
Iron Star AS 325 An Unusual Emission Line
Eclipsing BinaryJill Gerke (NOAO, University of
Arizona), Steve B. Howell (WIYN, NOAO), Sebastian
Otero (CEA), Arne Henden (AAVSO)
AS 325 is a complex eclipsing binary system
consisting of a BVe primary and a variable
(K2III-M4III) red giant companion. The optical
spectrum contains many emission lines, especially
those of Fe II. This forest of Fe II emission
lead to AS 325s identification as an analog for
XX Ophiuchi, the original and only other iron
star. The mean brightness of AS 325 has been
steadily increasing over the past 5 years,
coinciding with an increase in the depth of the
eclipse. New spectroscopic and photometric
observations of AS 325 were obtained over the
2005 eclipse. The spectra show a generally flat
continuum during the eclipse and a variable,
slightly red, continuum outside of eclipse. The
strength of the H emission lines (from the Be
star) increase slightly during the eclipse,
suggesting a decreased continuum level from the
red star. This fact, along with the increased
variability in the red portion of the continuum
out of eclipse, suggests that the red star is
eclipsed. Cool et al. (2005) used UV, optical,
and IR spectra to propose that AS 325 consisted
of a Be star and a K2.5 III star. Our new
optical spectroscopy indicates that the red star
now has a spectral type of M4III, suggesting that
it is a variable giant and may account for the
general system brightening. One interesting
spectral feature is the Ca II H and K absorption
lines. These lines are observed to disappear
during eclipse and reappear afterwards. A model
combination of a B4V and M4III reveals that the
observed Ca II H and K absorption does not
originate in either star and its presence and
variability are unexplained. Using the
established orbital period (513 days) and
estimated masses for a B4V and an M4III star, we
find the binary separation is 4.8 au and the
binary inclination must be at least 88 degrees.
Due to the relative sizes of the two stars, the
M4III will not be fully eclipsed by the 8 times
smaller B4Ve star. Multi-color photometric
observations of the eclipse, obtained by the
AAVSO, help confirm our assumptions that the B
star passes in front during the eclipse.
I. The complex optical spectra of AS 325 has lead
to much ambiguity in its understanding and
classification. The confusion results from the
many emission lines and essentially no stellar
absorption features found in the optical
spectrum. Optical spectra reveal strong hydrogen
emission lines as well as emission from ionized
metals such as Fe II, Ti II, and Cr II. Bopp and
Howell (1989) cited these unusual emission lines
to denote its similarity to XX Ophiuchi,
Merrills iron star (1924). Neither star is
really overly rich in Fe, and the low ionization
metal emission lines seen in the optical are a
result of wind material fluorescence. Bopp and
Howell (1989) suggested the possibility that AS
325 was a binary system and Cool et al. (2005)
proposed that AS 325 consists of a Be star and a
K2.5 III star.
V. In addition to being photometrically variable,
we also find AS 325 to be spectroscopically
variable. For example, spectra presented in Cool
et al. (2005) differ from those obtained during
our 2005 observations. Cool et al. proposed that
AS 325 consists of a Be star and a K2.5 III star.
The identification of a K2.5 III star was based
on a K-band spectrum and showed a consistent
match to AS 325 in the optical. On the left, we
repeat the Cool et al. plot comparing the red
spectrum of AS 325 to that of the single K2 III
star HD 140573. However, our new optical
spectroscopy (see VI) shows features in the (red)
optical not present in the Cool et. al (2005)
data. These features include TiO bands from the
continuum of a red giant star. We conclude that
the red star is now an M4III.
II. The figure at left shows the long term
(ASAS-3) V light curve of AS 325. The 513 day
periodic eclipses are clearly visible, confirming
AS 325 is indeed a binary. The light curve shows
variability in both the mean magnitude of the
system and the depth of the eclipse. The
increasing mean magnitude (by 0.6 mag) and depth
of eclipse imply that one of the stellar
components has increased in brightness. We find
observational evidence that this brightness
increase is due to variability of the red giant
secondary star changing from a K2.5III to a
0.5-0.8 mag brighter M4III over the past few
years.
VI. To see if the relatively flat, featureless
optical spectrum of AS 325 could be produced by a
combination of the two proposed stellar
components, we ignored the non-photospheric
emission lines and co-added a normal B4V and an
M4 III star. Normalizing each star at 5000A, we
scaled the two stellar spectra to their correct
relative brightness as viewed at the same
distance form Earth.
Our model (right) provides the correct
low-level bumps seen in AS 325. These features
include the Balmer jump of the B star and the TiO
band heads of the M giant. While this model
supports AS 325 as a binary containing a BV
MIII star, it can not explain the Ca II H and K
absorption lines. These lines disappeared during
eclipse (see III) and, when out of eclipse, they
decrease in strength as the continuum becomes
red. Our model composite spectrum shows that the
observed Ca II HK absorption lines are not from
the photosphere of either star and the fact that
the absorption goes away during eclipse seems to
imply a formation site between the two stars.
III. The most recent eclipse of AS 325 occurred
from 5 June 2005 to 30 August 2005. Three
representative optical spectra are plotted at the
right near mid-eclipse (top) and 13 and 15 days
after the eclipse ended. They are shown offset by
a constant flux amount. Spectra taken during
eclipse showed little variability and are
relatively flat over the optical range. Spectra
obtained out of eclipse show variability in the
appearance of the continuum changing from flat to
an increasing red slope. Note the absence of Ca
HK absorption during eclipse. The emission lines
are always present but variable and P-Cyg wind
features are seen at times (Cool et al. 2005).
V. If the red star gets transited by the blue
star, we would expect the system to become
slightly less red during eclipse. Using AAVSO
eclipse observations (left), we see that the B-V
color of AS 325 is composite at all times, being
F8 at mid-eclipse and G5 out of eclipse,
i.e., slightly less red in eclipse. The R-I
color, however, is nearly equal in/out of eclipse
at 0.6. This suggests, as we have seen, that the
M giant dominates the red optical continuum at
all times while the blue optical changes during
eclipse.
IV. The cartoon to the left shows the relative
sizes of the components in AS 325. We see that
the Be star will occult only a small fraction of
the secondary during eclipse. The red star now
appears to be an M4 giant, larger and brighter
than the K2 giant star found by Cool et al.
(2005). This large change in spectral type over a
period of a few years is not unheard of for a
Mira or SR post-AGB type variable. Using typical
masses and the orbital period, the two stars are
4.8 au apart. These orbital parameters lead to
a lower limit for the binary inclination of 88
degrees.
References Cool, R.J. et al., 2005 (PASP, 117,
462) Bopp, B.W., Howell, S.B. 1989 (PASP, 101,
981) Merrill 1924 (PASP, 36, 225)
We wish to thank