Near-infrared spectrophotometry of carbon stars : from the IRTS to dome C

T. Le Bertre1, M. Tanaka2,3, I. Yamamura2,

H. Murakami2, D.J. MacConnell4, A. Guertin1,5

1 Paris Observatory, France2 Institute of Space and Astronautical Science, Japan3 National Astronomical Observatory, Japan4 Computer Sciences Corporation/Space Telescope Science Institute, USA 5 University of Montréal, Canada

ARENA conference, Paris, 14-16 June 2006

Introduction

Carbon stars are characterized by the presence of carbon molecules such as C2 or CN. The enrichment in carbon may occur when they evolve on one of the giant branches or may be due to mass transfer from a more evolved companion.

Inventories of cool carbon stars are of special interest because these objects can be used to trace matter at large galactocentric distances out to the Magellanic Clouds.

Carbon stars contribute to the replenishment of the ISM.

We will examine the potential for finding and characterizing carbon stars by using near-infrared spectrophotometric surveys. For that purpose, we use the data provided by the Japanese space experiment IRTS.

The IRTS

The InfraRed Telescope in Space (IRTS) is a 15 cm diameter telescope operated in space which surveyed ~ 7 % of the sky in two strips, one along the galactic plane and the other covering the high galactic latitude region (Murakami et al., 1996, PASJ 48, L41).

The 2 strips covered by the IRTS in galactic coordinates

The IRTS was equipped with four scientific instruments operating in parallel: the Near-InfraRed Spectrometer (NIRS; 1.4-4.0 µm), the Mid-Infrared Spectrometer (MIRS; 4.5-11.7 µm), the Far-Infrared Line Mapper (FILM; 145, 155, 158, 160 µm), and the Far-InfraRed Photometer (FIRP; 150, 250, 400, 700 µm). Most wavelengths observed by IRTS are blocked by the Earth's atmosphere and cannot be observed with ground-based telescopes.

The NIRS

The Near-InfraRed Spectrometer (NIRS) is a low-resolution grating spectrophotometer (Noda et al., 1994, ApJ 428, 363). A 12 x 2 channel InSb detector array covers two wavelength ranges, 1.4-2.5 and 2.8-4.0 µm, with a spectral resolution ranging from ~20, at short wavelengths, to ~40 at the long wavelength end. The entrance aperture was 8 arcmin by 8 arcmin on the sky.

More than 14 000 point sources have been detected. Their spectra can be accessed via the DARTS archive

URL: http://www.darts.isas.jaxa.jp/

or http://irsa.ipac.caltech.edu/data/IRTS/

Mass losing carbon stars can be easily identified from the deep absorption band at 3.1 µm.

NIRS spectra of

cool carbon stars show

molecular absorptions at

1.4 µm (CO + CN),

1.8 µm (C2),

2.3 µm (CO),

3.1 µm (C2H2+HCN) and

3.8 µm (C2H2).

T Lyr, CGCS 4038

Mass-losing M and C AGB stars can be easily separated with infrared

spectrophotometry.

By contrast, NIRS spectra of late-type, oxygen-rich stars show only CO and H2O bands,in particular at 1.9 µm.

increasing depth of the 3.1 µm feature

Le Bertre et al., 2005, PASP 117, 199

We have identified 139 cool carbon stars in the NIRS survey of the IRTS from the conspicuous presence of molecular absorption bands at 1.8, 3.1, and 3.8 µm (Le Bertre et al., 2005). Among them, 14 are new, bright (K ~ 4.0-7.0) carbon stars; two such cases are shown below. Some dusty carbon stars, like NIRS 07036-2221, can only be revealed by infrared surveys.

Le Bertre et al., 2003, A&A 403, 943

The IRTS data show a clear spatial separation in the Galactic Plane between mass-losing, oxygen-rich stars and mass-losing, carbon stars, with the former (dots) outnumbering the latter (diamonds) for rGC< 8 kpc, and the reverse for rGC> 10 kpc.

ISO/ SWS: 3-8 µm;

R ~200-1000

Aoki et al. 1999, IAUS 191, 175

Y CVn; KAO

Goebel et al. 1980, ApJ 235, 104

C2 bandhead @1.77 µm

CN bandhead @1.40 µm

CN C2

HCN+C2H2

C3

IRTS/ NIRS : UV Cam; CGCS 177 (R8)

CN

C2

CO

CN bandhead @1.40 µm

C2 bandhead @1.77 µm

Le Bertre et al. 2005, PASP 117, 199

• Infrared spectrophotometry in the range 1-4 µm is an important tool to identify and characterize late-type stars.

• Since the IRTS there has been no other near-infrared spectrophotometric survey, although integral-field techniques and large-scale panoramic detectors are now available in this wavelength range.

• We suffer (and may suffer for some time) from a gap in our vision of the Universe, that is general, i.e. not specific to carbon star studies.

• methane and water bands

• Pemission at 1.88 µm

• 3.3-3.5 µm emission bands

• 3.1 µm water-ice band

+ possibility of synthesizing numerically photometric bands/combs

Allard (2001)

For galactic carbon stars :

• We need to cover as continuously as possible the spectral range from 1.5 (1.2 / CN) µm to 3.6 (4.0 / C2H2) µm.

• with a spectral resolution > 50 (100).

• We need to survey large areas of the sky in different galactic directions (l and b) + MCs, etc.

Request :

An atmospheric transmission (and if possible emission during day and night) from Concordia in the range 1.2-4.1 µm, in order to determine which part of the spectrum can be accessed from dome C.

Suggestion :

An integral-field spectro-imager, R ~ 80-100, operating from

1.2 to 4.0 µm that would allow to explore an area that has

been barely touched by the IRTS.

Spectrophotometric imaging surveys:

General resource for the study of stars, stellar structures and ISM

• late-type stars / PNs in the Galaxy and nearby galaxies

• Galactic plane; Magellanic Clouds

• ISM (P , UIBs, …)

• Star forming regions (Chamaeleon, ...)

• Galactic halo (galactic pole)

• etc.

KHJL’ M

H20 = 1mm; 1 ai

Modelled atmospheric transmission from Antarctica “dome C” (Epchtein 2005)

Le Bertre et al., 2005, PASP 117, 199

For the IRTS carbon stars, we find a trend relating the 3.1 µm band strength to the K-L’ color index, which is known to correlate with mass-loss rate.

The 14 new carbon stars are represented by a ‘ ’

NIRS 07036-2221

NIRS 07037-2326

Discussion

• Winters et al. (2000, A&A 361, 641) have shown that mass loss from red giants is favored by a low stellar effective temperature, a low stellar mass, a high luminosity, and a large pulsation amplitude. Loidl et al. (1999, A&A 342, 531) have shown that the strengths of the C2H2 and HCN features follow these same trends because these molecules form preferentially at low temperature in the upper stellar atmosphere.

• However, the correlation breaks down for large mass loss rates due to the effect of dust emission filling in the 3.1 µm feature.

Conclusion

• Infrared spectrophotometry is an important tool to identify and characterize late-type stars.

• Sensitive, near-infrared spectrophotometric surveys covering the range from ~1.5 to ~4.5 µm with a spectral resolution > 40 have the potential to reveal new carbon stars in the Galaxy and in its satellites and to provide useful information on their physical properties.