Home
Research
When stars die, they eject most of their mass back into space, until only their core is left. This process is driven by pulsations in the star's envelope, the condensation of dust in its outer layers, and radiation pressure on the dust. As the star dies, it is hidden by its own cocoon of dust, disappears from the optical sky, and emerges simultaneously as a bright infrared source. Throughout my career, I have been involved in one project or another unravelling how these stars die and how they form dust. See the page on dying stars for more background on what happens.
Understanding how stars die is important. Because the formation of dust and its acceleration from radiation pressure is essential to the process, how much dust can form and what it is made of determine how long the star will take to die. If a star lingers on the AGB, it will have more time to fuse heavy elements in its core, dredge them to its surface, and eject them along with everything else to enrich its environment. The Universe began with just hydrogen and helium. All of the heavier elements were made in stars, many of them on the AGB. So how stars die is a key part of the evolution of galaxies and the Universe.
Most of the energy emitted by dusty AGB stars is in the infrared, which means that if you want to understand the process, you need infrared telescopes, preferably infrared telescopes in space. My research on evolved stars has relied heavily on them. My work has also focused on studying the spectra of stars.
Most stars in our own Galaxy produce oxygen-rich dust, which primarily takes the form of silicates. Silicate dust grains are a lot like the grands of sand on a beach, just smaller. However, oxygen-rich stars make other kinds of dust, too. The next most common form is alumina (Al2O3). If you can somehow crystallize the alumina, and you add some impurities to it, you can make gems like sapphire and corundum. Alumina turned out to be the key to my first project in this area. It's not a coincidence that my wife's engagement ring is sapphire!
My introduction to infrared spectroscopy was as a graduate student and summer intern at the Air Force Geophysics Laboratory (AFGL) in 1987. I worked with Paul LeVan that summer on a new infrared spectrometer he was developing to use at the Wyoming Infrared Observatory (WIRO). One of his projects was to confirm the existence of a 13 µm emission feature seen in spectra from IRAS and first reported by Little-Marenin and Price (1986). Our spectra from WIRO showed that the feature was real (LeVan et al. 1989).
When I joined Paul as a post-doc at AFGL (which the Air Force had renamed as Phillips Lab. by then), I took a careful look through the IRAS database of Low-Resolution Spectra (LRS) to find out how many 13 µm sources there might be and what kind of sources they were. We discovered that nearly all spectra from semi-regular variables had a 13 µm feature, compared to a much smaller fraction of Mira variables (Sloan et al. 1996). We also showed that the most likely carrier of the feature was crystalline alumina (Al2O3). The biggest question we left unanswered was why the dust would be crystalline around semi-regulars and not Miras.
In a later paper, following up with spectra from ISO, the next major infrared space telescope, my colleagues and I pointed out that semi-regulars aren't losing mass as fast as Miras, and we hypothesized that if dust grains formed and grew slowly, they could be crystallized before they are pushed away from the star (Sloan et al. 2003). This hypothesis has not been tested, so it may or may not be correct.
In order to compare the spectra with 13 µm features with spectra from other oxygen-rich dust sources, we had to look more carefully at the larger sample. Irene Little-Marenin had already discovered that the IRAS/LRS spectra showed several different dust emission features in the infrared (Little-Marenin and Little 1988, 1990). We followed up with a new classification system which showed that all oxygen-rich dust spectra from evolved stars followed what we called the silicate dust sequence. For a quick review, see our first report, Sloan et al. (1995a).
Our first refereeed paper (Sloan and Price 1995b) focused on AGB stars. The silicate dust sequence showed a continuous progression of dust types from low-contrast emission from amorphous alumina dust (at 11-13 µm) to much stronger emission from amorphous silicates (with broad features at 10 and 18 µm). It was puzzling that we found no relation between the variability type of the star (Mira, semi-regular, or irregular) and the type of dust it produced, even though we could see that semi-regulars were more likely to have a 13 µm feature.
Sloan and Price (1998) added supergiants and S stars to the sample and found that supergiants generally produced just classic silicate dust. The S stars have C/O ratios close to one, so that CO molecules take up almost all of the carbon and oxygen, leaving hardly any to make dust. These stars generally just showed alumina dust emission, leading us to wonder if the normal AGB stars with alumina hadn't run out of oxygen before they could form silicates.
The Spitzer Spact Telescope made it posssible to study evolved stars in nearby galaxies. One of the problems with studying evolved stars inside our Galaxy is that we don't know their distances. The distances to the other galaxies in the Local Group are much better known, so that we can work out the absolute brightnesses of our targets. Another advantage is that stars in nearby galaxies have different abundances of heavier elements like carbon and oxygen, so that we can use stars in different galaxies to better understand how the abundances a star forms with influence how it dies.
Sloan et al. (2008) found that as long as stars stayed oxygen-rich, those with lower abundances of heavy elements produced less dust. This makes sense because the dominant components of dust are oxygen, silicon, and aluminum, which are not created by most stars. They only have what they're born with, and the less they start with, the less dust they make. Sloan et al. (2010), studying lower-mass stars in globular clusters, found pretty much the same thing.
The carbon stars in the Local Group sample were the real surprise. First of all, most of our evolved stars turned out to be carbon-rich. Second of all, we didn't see any evidence that carbon stars born with fewer heavy elements produced less dust. The results came out incrementally, with Sloan et al. (2006) and Zijlstra et al. (2006) publishing the initial results on the Small and Large Magellanic Clouds, respectively (SMC and LMC), Lagadec et al. (2007) and Leisenring et al. (2008) following up with more stars from the SMC and LMC (respectively), and Sloan et al. (2008) realizing that the production of dust did not appear to be a function of elemental abundances.
Our Local Group sample reached beyond the LMC and SMC, which are about 50,000 and 60,000 pc away from us (180,000 and 200,000 light years). We also observed carbon stars in the Carina, Sculptor, Fornax, and Leo I Dwarf Spheroidal galaxies, and these had heavy-element abundances down to only 1/30th of the Sun's. Amazingly, these stars still produce nearly as much dust as their cousins with more heavy elements (Sloan et al. 2012). When you think about it, it actually makes sense, because these stars are fusing carbon from helium in their cores and dredging it to their surfaces, where it can condense into dust as it's ejected from the star.
We aren't done yet! Stay tuned for more results.
Lagadec, E., Zijlstra, A.A., Sloan, G.C. et al. 2006, “Spitzer spectroscopy of carbon stars in the Small Magellanic Cloud,” MNRAS, 376, 1270.
Leisenring, J.M., Kemper, F. & Sloan, G.C. 2007, “Effects of metallicity on the chemical composition of carbon stars,” ApJ, 681, 1557.
LeVan, P.D., & Sloan, G. 1989, “Ten-micron observations of bright circumstellar shells - Spectral properties and a search for extended emission,” PASP, 101, 114.
Little-Marenin, I.R., & Little, S.J. 1988, “Emission features in IRAS low-resolution spectra of MS, S, and SC stars,” ApJ, 333, 305.
Little-Marenin, I.R., & Little, S.J. 1990, “Emission features in IRAS LRS spectra of M Mira variables,” AJ, 99, 1173.
Little-Marenin, I.R., & Price, S.D. 1986, “The shapes of the circumstellar silicate features,” in Summer School on Interstellar Processes: Abstracts of Contributed Papers, 137.
Sloan, G.C., Price, S.D., Little-Marenin, I.R., & LeVan, P.D., 1995a, “Silicate and related dust emission in stars on the asymptotic giant branch,” in Airborne Astronomy Symposium on the Galactic Ecosystem: From Gas to Stars to Dust, ed. M.R. Haas, J.A. Davidson, & E.F. Erickson (San Francisco: ASP), 425.
Sloan, G.C., & Price, S.D. 1995b, ”Silicate emission at 10 microns in variables on the asymptotic giant branch,” ApJ, 451, 758.
Sloan, G.C., LeVan, P.D., & Little-Marenin, I.R. 1996 , “Sources of the 13 micron feature associaed with oxygen-rich circumstellar dust,” ApJ, 463, 310.
Sloan, G.C., & Price, S.D. 1998, “Infrared spectral classification of oxygen-rich dust shells,” ApJS, 119, 141.
Sloan, G.C., et al. 2003, “Guilt by association: The 13 µm dust emission feature and its correlation to other gas and dust features;” ApJ, 594,, 483.
Sloan, G.C., Kraemer, K.E., Matsuura, M., et al. 2006, “Mid-infrared spectroscopy of carbon stars in the Small Magellanic Cloud,” ApJ, 645, 1118.
Sloan, G.C., Kraemer, K.E., Wood, P.R., et al. 2008, “The Magellanic zoo: Mid-infrared Spitzer spectroscopy of evolved stars and circumstellar dust in he Magellanic Clouds,” ApJ, 686, 1056.
Sloan, G.C., Matsunaga, N., Matsuura, M., et al. 2010, “Spitzer spectroscopy of mass loss and dust production in globular clusters,” ApJ, 719, 1274.
Zijlstra, A.A., Matsuura, M., Wood, P.R., et al. 2006, “A Spitzer mid-infrared spectral survey of mass-losing carbon stars in the Large Magellanic Cloud,” MNRAS, 370, 1961.
Last modified 6 July, 2015. © Gregory C. Sloan.