Despite our awareness of fluorine's unique properties, its astrophysical origin is not well understood. Like most heavy elements, fluorine is a product of nuclear reactions in the hot cores of stars during various phases of their evolution. Furthermore, fluorine is not easy to observe in the universe, and very few spectroscopic measurements of fluorine in stars exist. Until the recent Gemini work reported here, only a meager set of data for a few stars in our own Milky Way was available.

A spectroscopic program conducted at Gemini South, located on Cerro Pachón, Chile, has changed this picture. Katia Cunha (Observatório Nacional, Brazil) led a PHOENIX team of researchers on Gemini South to measure the abundance of fluorine in our own galaxy's satellite, the Large Magellanic Cloud (LMC), and in the massive galactic globular cluster Centauri. The team studied fluorine in its most accessible form, hydrofluoric acid (HF), which is detectable through vibration-rotation transitions falling in the near infrared close to the wavelength of 2.3 microns (main figure and Figure 1). These Gemini observations provide unique insight on how fluorine behaves as a function of the abundance of the other elements, specifically metals, helping to identify how fluorine is made.

To better understand how the abundance of fluorine depends on metallicity and stellar populations, Cunha and her collaborators compared their Gemini measurements with known K and M type stars of our own Milky Way. The total sample of stars with fluorine abundance determinations contains 23 red giant stars across three stellar populations including the solar neighborhood, Centauri and the LMC. It is to be noted that the abundance of fluorine, like that of other elements, does not only depend on the yield from nucleosynthesis processes. Star formation history and element dispersal mechanisms can also play important roles.

Figure 1. Observed (dots) and modeled (lines) spectra for the LMC giant star 2.3256. Three fluorine-19 abundances are presented. An abundance of A(F)=3.93 - with the Sun hydrogen abundance being 12.0, on a logarithmic base 10-scale was derived.

Figures 2 and 3. Logarithm abundance of fluorine, A(F), plotted versus oxygen, A(O), and logarithm abundance of fluorine, A(F), plotted versus iron, A(Fe). The colors of the symbols refer to the location of the red giants: the Milky Way (blue squares), the LMC (green circles) and Centauri (purple triangles). The encircled black dots correspond to the solar abundance of fluorine

Although a W-R source cannot be excluded at this stage, the best bet is the neutrino process of spallation in supernovae of massive stars (SN II). Models of neutrino nucleosynthesis predict that the [F/O] ratio declines steadily as oxygen declines. This is shown very clearly in Figure 4.

The Centauri red giants stand out from this prediction and trend. We know that Centauri had 3 or 4 major isolated episodes of star formation, which ended several billions years ago, giving the cluster its wide range of metal abundance. In contrast, the Milky and the LMC are undergoing continuous star formation.

Taking into account the different star-formation histories, the authors conclude that the dependence of [F/O] on A(O) - as measured in Galactic LMC and Centauri red giant stars is best explained by the chemical evolution models using the known neutrino spallation present in massive supernovae, and that these explosions must be the principal source of fluorine-19.

Figure 4. The ratio of fluorine to oxygen abundance versus oxygen in red giants stars located in the Milky Way, the LMC and the rich globular cluster Centauri.

The PHOENIX spectrometer (Hinkle et al. 1998, Proc. SPIE, 3354, 164) was built by NOAO.