The ESO Large Programme First Stars
P. Bonifacio, J. Andersen, S. Andrievsky, B. Barbuy, T. C. Beers, E. Caffau, R. Cayrel, E. Depagne, P. Francois, J. I. Gonzalez Hernandez, C. J. Hansen, F. Herwig, V. Hill, S. A. Korotin, H.-G. Ludwig, P. Molaro, B. Nordstrom, B. Plez, F. Primas, T. Sivarani, F. Spite, M. Spite
aa r X i v : . [ a s t r o - ph ] J a n The ESO Large Programme “First Stars”
P. Bonifacio , , , J. Andersen , , S.M. Andrievsky , B. Barbuy ,T.C. Beers , E. Caffau , R. Cayrel , E. Depagne , P. Fran¸cois ,J.I. Gonz´alez Hern´andez , , C.J. Hansen , F. Herwig , V. Hill ,S.A. Korotin , H.-G. Ludwig , , P. Molaro , B. Nordstr¨om , B. Plez ,F. Primas , T. Sivarani , F. Spite , and M. Spite CIFIST Marie Curie Excellence Team GEPI, Observatoire de Paris, CNRS, Universit´e Paris Diderot; Place JulesJanssen 92190 Meudon, France Istituto Nazionale di Astrofisica, Osservatorio Astronomico di Trieste, ViaTiepolo 11, I-34143 Trieste, Italy The Niels Bohr Institute, Astronomy, Juliane Maries Vej 30, DK-2100Copenhagen, Denmark Nordic Optical Telescope, Apartado 474, E-38700 Santa Cruz de La Palma,Spain Department of Astronomy and Astronomical Observatory, Odessa NationalUniversity, Shevchenko Park, 65014 Odessa, Ukraine Universidade de Sao Paulo, Departamento de Astronomia, Rua do Matao 1226,05508-900 Sao Paulo, Brazil Department of Physics & Astronomy and JINA: Joint Institute for NuclearAstrophysics, Michigan State University, East Lansing, MI 48824, USA Las Cumbres Observatory, Santa Barbara, California, USA European Southern Observatory (ESO), Karl-Schwarschild-Str. 2, D-85749Garching b. M¨unchen, Germany Keele Astrophysics Group, School of Physical and Geographical Sciences, KeeleUniversity, Staffordshire ST5 5BG, UK GRAAL, Universit´e de Montpellier II, F-34095 Montpellier Cedex 05, France
In ESO period 65 (April-September 2000) the large programme 165.N-0276,led by Roger Cayrel, began making use of UVES at the Kueyen VLT telescope.Known within the Team and outside as “First Stars”, it was aimed at obtain-ing high resolution, high signal-to-noise ratio spectra in the range 320 nm –1000 nm for a large sample of extremely metal-poor (EMP) stars identifiedfrom the HK objective prism survey [3, 4]. The goal was to use these spectrato determine accurate atmospheric parameters and chemical composition ofthese stars which are among the oldest objects amenable to our detailed study.Although these stars are not the first generation of stars they must be very
Bonifacio et al. close descendants of the first generation. One may hope to gain insight on thenature of the progenitors from detailed information on the descendants.The extremely metal-poor stars are very rare objects and finding themin large numbers requires specially designed surveys. All of the proponentsof the large programme had been actively working on the medium-resolutionfollow-up of the HK survey (results still to be published), from either ESO LaSilla, Kitt Peak or Roque de los Muchachos. Such a follow-up is mandatoryin order to obtain a good list of candidates on which one can invest the timeof an 8 m telescope.The programme was allocated a total of 39 nights between periods 65and 68, these were split into 8 observational runs of unequal length. Theobservations were carried out in visitor mode because UVES was used innon-standard settings. The settings selected were Dic1 396+573 and Dic1396+850, typically with a 1 ′′ slit for a resolution R ∼ ii i
471 nm in the red and Li i The first surprise came quite early in the programme, Vanessa Hill was con-ducting the observations in August 2000 when she realised, from the quicklook data, that the giant CS 31082-001 had an exceptional spectrum, charac-terised by extremely low metallicity and a large enhancement of the r-processelements. She was in fact able to identify immediately the Th ii he ESO Large Programme “First Stars” 3 Ever since Monique and Fran¸cois Spite discovered that warm metal-poor starsshare the same Li abundance (the
Spite plateau )[17, 18], there has been anactive research on this field. What we wish to understand is if this plateau indicates the primordial Li abundance, as initially proposed[17, 18], or not.
Fig. 1.
The Spite plateau at the lowest metallicities as portrayed by the “FirstStars” data. The stars whose names are labelled are binaries, for CS 22876-32 anorbital solution is available and the analysis has been done taking properly intoaccount the veiling and Li in both components has been measured, for CS22957-15 this has not been possible, due to the lack of the necessary data, however thecorrection for the veiling is likely not very large.
The “First Stars” large programme allowed to explore the
Spite plateau at the lowest known metallicities. There are no known dwarf stars with ametallicity (meant as Z , total metallicity, not [Fe/H]) lower than the starsshown in Fig.1. The data are those of [5] and [11], the picture which emergesis that the plateau seems to continue at the lowest metallicities. It is possiblethat there is a larger scatter, however the impact on this picture of stars inwhich lithium may have been partially depleted is yet unclear. The differencein lithium content between the two components of the binary system CS22876-32 has no clear explanation. The cooler component (star B) has aneffective temperature of 5900 K and should not display Li depletion accordingto standard models.From the cosmological point of view there is a tension between the valueof the Spite plateau , A(Li) ∼ . Bonifacio et al.
A(Li)= 2 .
64. Several ways to explain this discrepancy have been suggested,and generally they go in two possible directions: a) the
Spite plateau does notrepresent the primordial abundance or b) primordial nucleosynthesis did notproceed as assumed in the “standard” model. At present both solutions arepossible and further observations of EMP stars, to understand if there is anexcess scatter of Li at the lowest metallicities, could give useful indications.
When we started the large programme, several of us, were expecting thatat the lowest metallicities we would begin to see the effects of the pollutionof very few supernovae(SNe), possibly a single supernova. As a consequencewe were expecting considerable scatter in the abundance ratios, which wouldbe the signature of the different masses of the polluting SNe and incompletemixing of the gas in the early Galaxy. To the surprise of several of us we foundinstead that the majority of elements C to Zn display a remarkable uniformity,with well defined trends with metallicity[7]. The scatter in these trends can betotally explained by observational error. One explanation of this low scatteris an efficient mixing of the early Galaxy. Alternatively one could argue infavour of a narrow range of masses of SNe actually contributing to chemicalenrichment.The exceptions, among lighter elements, were Na and Al, that displayeda star to star scatter larger than observed for other elements. This excessscatter also made the definition of trends somewhat ambiguous. A reanalysisof both elements using full NLTE line formation was in fact able to solve theproblem[1, 2]. Sodium appears to be constant with metallicity among EMPstars, with [Na/Fe]= 0 . ± .
13, and the same is true for aluminium with[Al/Fe] = 0 . ± . second r-process is the main production channel at [Ba/H] < − . he ESO Large Programme “First Stars” 5 rather diverse among the stars, suggesting nucleosynthesis taking place underdifferent physical conditions.The giant CS 22949-037 is one of the most extraordinary found in thecourse of the large programme[8]. With [Fe/H] ∼ − . ∼ +2, [C/Fe] ∼ +1 .
2, [N/Fe] ∼ +2 .
6) make its global metallicity Z notso extreme as that of the four giants with [Fe/H] ∼ − Z known. There is no totally satisfactory model toexplain the abundance pattern in CS 22949-037, however it is clear that somespecial kind of SN is needed to explain such an extraordinary pattern. One would like to extend the work done so far, with high resolution, high S/Nratio spectra of stars of even lower metallicities. Such stars should be foundby on-going and projected surveys (SEGUE, LAMOST, SkyMapper...). Mostof these are however expected to be around 18th magnitude or fainter, UVEScan work at these faint magnitudes, but...slowly. The proposed high resolutionspectrograph ESPRESSO (see Pasquini this meeting) in the mode combiningthe 4UTs, would be ideal for these targets. According to the preliminaryestimates ESPRESSO 4UTs, at a resolution of R ∼ References
1. Andrievsky, S. M., et al.: A&A , 1081 (2007)2. Andrievsky, S. M., et al.: A&A accepted
3. Beers, T. C., Preston, G. W., Shectman, S. A.: AJ , 2089 (1985)4. Beers, T. C., Preston, G. W., Shectman, S. A.: AJ , 1987 (1992)5. Bonifacio, P., et al.: A&A , 851 (2007) [Paper VII]6. Cayrel, R., et al.: Natur , 691 (2001)7. Cayrel, R., et al.: A&A , 1117 (2004) [Paper V]8. Depagne, E., et al.: A&A , 187 (2002) [Paper II]9. Fran¸cois, P., et al.: A&A , 1105 (2003) [Paper III]10. Fran¸cois, P., et al.: A&A , 935 (2007) [Paper VIII]11. Gonz´alez Hern´andez J.I, et al.: A&A accepted , arXiv 0712.2949 [Paper XI]12. Hill, V., et al.: A&A , 560 (2002) [Paper I]13. Plez, B., et al.: A&A , L9 (2004)14. Sivarani, T., et al.: A&A , 125 (2006) [Paper X]15. Sivarani, T., et al.: A&A , 1073 (2004) [Paper IV]16. Spergel, D. N., et al.: ApJS , 377 (2007)17. Spite, M., Spite, F.: Natur , 483 (1982)18. Spite, F., Spite, M.: A&A , 357 (1982)19. Spite, M., et al.: A&A , 655 (2005) [Paper VI]20. Spite, M., et al.: A&A455