The Optical Gravitational Lensing Experiment identified over 1,800 carbon-rich Mira and semi-regular variables in the Small Magellanic Cloud. Multi-epoch infrared photometry reveals that the semi-regulars and Miras follow different sequences in color
-color space when using colors sensitive to molecular absorption bands. The dustiest Miras have the strongest pulsation amplitudes and longest periods. Efforts to determine bolometric magnitudes reveal possible systematic errors with published bolometric corrections.
We present spectra obtained with the Infrared Spectrograph (IRS) on the Spitzer Space Telescope of 33 K giants and 20 A dwarfs to assess their suitability as spectrophotometric standard stars. The K giants confirm previous findings that the strength
of the SiO absorption band at 8 um increases for both later optical spectral classes and redder (B-V)_0 colors, but with considerable scatter. For K giants, the synthetic spectra underpredict the strengths of the molecular bands from SiO and OH. For these reasons, the assumed true spectra for K giants should be based on neither the assumption that molecular band strengths in the infrared can be predicted accurately from optical spectral class or color nor synthetric spectra. The OH bands in K giants grow stronger with cooler stellar temperatures, and they are stronger than predicted by synthetic spectra. As a group, A dwarfs are better behaved and more predictable than the K giants, but they are more likely to show red excesses from debris disks. No suitable A dwarfs were located in parts of the sky continuously observable from Spitzer, and with previous means of estimating the true spectra of K giants ruled out, it was necessary to use models of A dwarfs to calibrate spectra of K giants from observed spectral ratios of the two groups and then use the calibrated K giants as standards for the full database of infrared spectra from Spitzer. We also describe a lingering artifact that affects the spectra of faint blue sources at 24 um.
Infrared spectra of carbon-rich objects which have evolved off the asymptotic giant branch reveal a range of dust properties, including fullerenes, polycyclic aromatic hydrocarbons (PAHs), aliphatic hydrocarbons, and several unidentified features, in
cluding the 21 um emission feature. To test for the presence of fullerenes, we used the position and width of the feature at 18.7-18.9 um and examined other features at 17.4 and 6-9 um. This method adds three new fullerene sources to the known sample, but it also calls into question three previous identifications. We confirm that the strong 11 um features seen in some sources arise primarily from SiC, which may exist as a coating around carbonaceous cores and result from photo-processing. Spectra showing the 21 um feature usually show the newly defined Class D PAH profile at 7-9 um. These spectra exhibit unusual PAH profiles at 11-14 um, with weak contributions at 12.7 um, which we define as Class D1, or show features shifted to ~11.4, 12.4, and 13.2 um, which we define as Class D2. Alkyne hydrocarbons match the 15.8 um feature associated with 21 um emission. Sources showing fullerene emission but no PAHs have blue colors in the optical, suggesting a clear line of sight to the central source. Spectra with 21 um features and Class D2 PAH emission also show photometric evidence for a relatively clear line of sight to the central source. The multiple associations of the 21 um feature to aliphatic hydrocarbons suggest that the carrier is related to this material in some way.
This report describes the generation of fully calibrated spectra of the K giants HR 6348 and HD 173511 from data obtained with the low-resolution modules of the Infrared Spectrograph (IRS), with an emphasis on the spectra from the Long-Low (LL) modul
e. The spectra were calibrated using Kurucz models and IRS observations of the A dwarfs alpha Lac and delta UMi. The calibration process required mitigation for fringing in the first-order LL spectrum and a faint red excess in alpha Lac which may arise from a low-contrast debris disk. The final calibrated spectrum of HR 6348 has a spectroscopic fidelity of 0.5% or better below 29 um, with an uncertainty increasing to ~1% at 33-37 um. The final calibrated spectrum of HD 173511 has a spectroscopic fidelity of ~0.5% at all wavelengths below 35.8 um.
This report describes in detail the generation of a truth spectrum of HR 6348, using observations with the Short-Low (SL) module of the Infrared Spectrograph of HR 6348, and the A dwarfs alpha Lac and delta UMi. Using spectral ratios, we can propagat
e Kurucz models of the A dwarfs to the K giant HR 6348, which can then serve to calibrate the remaining database of SL spectra. Mitigation in the vicinity of the Pfund-a line is necessary to reduce residual artifacts at 7.45 um. In general, the new SL spectrum of HR 6348 has a spectroscopic fidelity of ~0.5% or better. Artifacts from the hydrogen recombination lines in the A dwarfs will generally be smaller than this limit, although the residual artifact from the blend of lines near Pfund-alpha exceeds the limit at ~0.7%.
We investigate how the shape of a spectrum in the Short-Low module on the IRS varies with its overall throughput, which depends on how well centered a source is in the spectroscopic slit. Using flux ratios to quantify the overall slope or color of th
e spectrum and plotting them vs. the overall throughput reveals a double-valued function, which arises from asymmetries in the point spread function. We use this plot as a means of determining which individual spectra are valid for calibrating the IRS.
We have observed a sample of 19 carbon stars in the Sculptor, Carina, Fornax, and Leo I dwarf spheroidal galaxies with the Infrared Spectrograph on the Spitzer Space Telescope. The spectra show significant quantities of dust around the carbon stars i
n Sculptor, Fornax, and Leo I, but little in Carina. Previous comparisons of carbon stars with similar pulsation properties in the Galaxy and the Magellanic Clouds revealed no evidence that metallicity affected the production of dust by carbon stars. However, the more metal-poor stars in the current sample appear to be generating less dust. These data extend two known trends to lower metallicities. In more metal-poor samples, the SiC dust emission weakens, while the acetylene absorption strengthens. The bolometric magnitudes and infrared spectral properties of the carbon stars in Fornax are consistent with metallicities more similar to carbon stars in the Magellanic Clouds than in the other dwarf spheroidals in our sample. A study of the carbon budget in these stars reinforces previous considerations that the dredge-up of sufficient quantities of carbon from the stellar cores may trigger the final superwind phase, ending a stars lifetime on the asymptotic giant branch.
We present a calibration of the acquisition data obtained by the Red Peak-Up (PU) sub-array on the Infrared Spectrograph on Spitzer, based on repeated observations of three K giants. This calibration is tied directly to the most current infrared cali
bration based on data from Multiband Imaging Photometer for Spitzer. An analysis of the responsivity of the Red PU sub-array reveals no detectable deviations from linearity in the most recent pipeline version, but older pipeli
We have observed a sample of 35 long-period variables and four Cepheid variables in the vicinity of 23 Galactic globular clusters using the Infrared Spectrograph on the Spitzer Space Telescope. The long-period variables in the sample cover a range of
metallicities from near solar to about 1/40th solar. The dust mass-loss rate from the stars increases with pulsation period and bolometric luminosity. Higher mass-loss rates are associated with greater contributions from silicate grains. The dust mass-loss rate also depends on metallicity. The dependence is most clear when segregating the sample by dust composition, less clear when segregating by bolometric magnitude, and absent when segregating by period. The spectra are rich in solid-state and molecular features. Emission from alumina dust is apparent across the range of metallicities. Spectra with a 13-um dust emission feature, as well as an associated feature at 20 um, also appear at most metallicities. Molecular features in the spectra include H_2O bands at 6.4-6.8 um, seen in both emission and absorption, SO_2 absorption at 7.3-7.5 um, and narrow emission bands from CO_2 from 13.5 to 16.8 um. The star Lynga 7 V1 has an infrared spectrum revealing it to be a carbon star, adding to the small number of carbon stars associated with Galactic globular clusters.
We observed a sample of evolved stars in the Large and Small Magellanic Clouds (LMC and SMC) with the Infrared Spectrograph on the Spitzer Space Telescope. Comparing samples from the SMC, LMC, and the Galaxy reveals that the dust-production rate depe
nds on metallicity for oxygen-rich stars, but carbon stars with similar pulsation properties produce similar quantities of dust, regardless of their initial metallicity. Other properties of the oxygen-rich stars also depend on metallicity. As the metallicity decreases, the fraction of naked (i.e. dust-free) stars increases, and among the naked stars, the strength of the 8 um absorption band from SiO decreases. Our sample includes several massive stars in the LMC with long pulsation periods which produce significant amounts of dust, probably because they are young and relatively metal rich. Little alumina dust is seen in circumstellar shells in the SMC and LMC, unlike in Galactic samples. Three oxygen-rich sources also show emission from magnesium-rich crystalline silicates. Many also show an emission feature at 14 um. The one S star in our sample shows a newly detected emission feature centered at 13.5 um. At lower metallicity, carbon stars with similar amounts of amorphous carbon in their shells have stronger absorption from molecular acetylene (C_2H_2) and weaker emission from SiC and MgS dust, as discovered in previous studies.
