are suitable for measuring the
blended equivalent width.
The abundance of oxygen in globular clusters is important for determining cluster ages, and for understanding the early chemical enrichment and star formation history of the Galaxy. Oxygen is also critical for understanding the source of carbon, nitrogen, and oxygen abundance changes (and related CN and CH strong and weak stars) in low mass stars--particularly those in globular clusters.
Stars in the galactic halo have about ~3x excess oxygen compared to
iron than for standard solar composition. In globular cluster giants,
oxygen is also enhanced relative to iron, but some clusters also
contain a few bright giants deficient in oxygen. This deficiency of
oxygen in luminous giants is attributed due to deep mixing and ON-cycle
burning. Observations of the abundance of oxygen in less evolved stars
in globular clusters would thus be helpful in testing this conclusion.
Caty Pilachowski and Taft Armandroff (NOAO) investigated the abundance
of oxygen in low luminosity giants in the globular cluster M13 (Figure
1) using the high excitation, permitted O I triplet at 7770 . Because
the lines are closely spaced, they become blended at low resolution,
and resolutions of a few times are suitable for measuring the
blended equivalent width.
Spectra of 40 M13 stars with similar temperatures and magnitudes near the base of the giant branch were obtained during three nights on the KPNO 4-m telescope using the Hydra fiber positioner. Spectra of metal poor field stars of similar temperature and luminosity were observed both with Hydra and with high spectral resolution with the Coudé Feed Telescope for comparison. The individual spectra were combined into a single spectrum for analysis, which has a signal-to-noise ratio (per pixel) of 300 (Figure 2). For the 40 stars included in the average spectrum, the average V magnitude is 16.65 and the average B-V color is 0.66. The B-V color gives an effective temperature of 5300K. The distance modulus, V magnitude, and assumed mass yield a surface gravity of log(g) = 3.0.
Adopting a 1
upper limit of 9 mÅ to the equivalent width
of the blended
O I triplet, Pilachowski and Armandroff concluded that the average
oxygen abundance among relatively unevolved stars in M ~ 13 must
have [O/Fe] 
0.1. An average oxygen abundance of [O/Fe] = 
0.3 is
excluded at the 4
level. If [O/Fe] = 
0.3, the
equivalent width of the
OI blend would be about 40 mÅ.
The luminosity of the composite M13 "star" is M
= 1.56, which places
it in the vicinity of the first dredgeup. The first dredgeup is likely
to affect the carbon isotope ratio and the C/N ratio, but is predicted
not to change the oxygen abundance. Stars of this magnitude should
retain the original surface oxygen abundance with which they formed.
That the average oxygen abundance is low suggests either that
nucleosynthesis and mixing change the surface abundance of oxygen in
M13 giants at an earlier evolutionary phase than previously
anticipated, or that M13 does not share in the general oxygen
enhancement found in the halo population.
It may not be appropriate to assume uniform [O/Fe] or [CNO/Fe] ratios in the halo for the derivation of globular cluster ages. Such assumptions could be particularly troublesome for age comparisons of second-parameter pairs of clusters. For M13, age estimates may need to be revised to slightly older ages.
A detailed report of this work appeared in the March 1996 issue of the Astronomical Journal.