BINARIES AT LOW METALLICITY

Transcription

1 Sara Lucatello INAF Padova Astronomical Observatory BINARIES AT LOW METALLICITY Why we care, what we know, what we do not know and what we are doing to improve that The first stars and the evolution of the early universe INT 3-7 Jul 2006

2 Why are binaries interesting? They retain lots of information about the star formation process provide a wealth of constraints to star formation theory which must reproduce the observed statistics (P, e and mass ratios distributions etc) For double lines eclipsing binaries, can measure directly masses; Explanation for peculiar abundance patterns; So, are binary characteristics at low Z different from their solar metallicity counterpart?

3 What we know: solar neighborhood A very large fraction of stars in the Solar Neighborhood are in multiple systems. Fraction is likely dependent on spectral types Jahreiß & Wielen 2000 ~60% for F and G dwarfs Siegler et al ~25% for M dwarfs Gizis et al ~15% for L dwarfs Duquennoy & Mayor 1991 parallax limited survey targeted 181 F and G dwarfs in the Solar Neighborhood, monitoring radial velocities for about 13 years

4 Orbital parameters distributions Duquennoy & Mayor 1991

5 What we know about low Z binaries? A few old studies suggested low metallicity population deficient in binaries (Abt & Levy 1969; Abt & Willmarth 1987) More recent studies seem to disagree: Imaging surveys, combine photometric and proper motion infos. Ryan 1991, Zapatero-Osorio & Martin 2004 find no change over about 2 dex Spectroscopic surveys. Carney, Latham etc targeted more than 1300 over 20 years (-2.5<[Fe/H]<0).

6 Carney & Latham survey Carney et al 2005 Spectroscopic binary fraction= detection fraction

7 Carney & Latham survey Modeled orbits for 171 single lined and 34 double lined binaries (one eclipsing)

8 Going for EMP stars CLLA survey has only a few EMP stars, stops at about [Fe/H]=-2.5, lower than that it is unexplored territory Star formation models expect the first stars form as single, so what will binary fraction as a function of Z below [Fe/H]<-2.5 look like? Binarity likely explanation for some (or most) of the peculiar abundance patterns observed among EMP. In particular CEMP-s stars are all in binaries (Lucatello et al. 2005), and thus like CH stars arise from IMS AGB transfer. CEMP-s incidence (~16%) much higher than CH (~2%): is it because of an increase in binary fraction or in the number of IMS stars (Abia et al 2003, Lucatello et al 2005)?

9 The Lick EMP binary survey David Lai, Mike Bolte (UCSC), Debra Fischer(SFSU), Jennifer Johnson(OSU), Sara Lucatello (OAPD-INAF) Aimed at determining binary fraction and orbital properties of EMP stars Uses the same technique as extra solar planet search (Hamspec R~45,000 and iodine cell), time consuming (needs high S/N) but has a great payoff. Typical error about 50 m/s. Sample made up of fairly bright (6<V<13.5) stars with [Fe/H]<-2.5, exposure times up to 4 hours Started in 2004, so far collected multiple observations for 33 stars more are planned and/or underway.

10 The Lick EMP binary survey: results Given exceptional precision, small Vrad variation can be detected adopting the standard χ2 statistics 7 stars out of 33 are show radial velocity variations, 2 stars evolved enough to have to show vrad jitter 5 out of 31 stars 16%±8% of our sample (Carney et al sample ~17%±2%) Preliminary data agree, fraction seem to be the same as at higher ([Fe/H] -2) metallicity. Sample (and coverage) yet too small, but the survey has lots of potential. What binary fraction is our detection fraction consistent with?

11 The Lick EMP binary survey: simulations Monte Carlo simulations of detection fraction as a function of binary fraction with present observational pattern show that the results are consistent with a fairly range of values (20-50% parent population) Adopted Duquennoy & Mayor P distribution, time baseline (1.5 yr) very small with respect to typical P (120 yr)

12 C-stars at low metallicity and binaries Herwig et al 2004 HERES indicate ~20% stars at [Fe/H]<-2 have [C/Fe]>+1, ~80% of them CEMP-s, ~5% CEMP-r and ~15% CEMP-no CEMP-s abundance pattern we can account for (IMS-AGB transfer) but still many things to explore: P and e distribution can help understand mass transfer as well as shed light to the (possible) effect of a companion on AGB nucleosynthesis

13 C-rich stars and binaries CEMP-no and CEMP-r origin still a mystery, but it could very well be that binarity plays a role in these cases too, but so far we have not enough data to be able to assess whether this is the case Multi telescope program (ESO FEROS at 2.2, UVES at VLT, HDS at Subaru, SARG at TNG, UES at WH, UCLES at AAT). About 50 CEMP stars (s, r and no) monitored over about 3 yr, still ongoing Typical precision ~ km/s takes longer baseline to detect Vrad variations, hopefully first results in 1-2 years

14 Summary & Conclusions Binarity properties seem to remarkably metallicity independent over a large range -2.5<[Fe/H]<0; First results suggest that EMP binary fraction might be lower, but cannot draw any definite conclusions from the limited data; In a few years will be able to constrain EMP binary properties (as a function of metallicity): fraction, P and e distribution For CEMP stars we will be able to assess whether and how peculiar abundance patterns are connected to the presence of a companion and what effect (if any) that has on AGB nucleosynthesis