Table Of Content

Local stars formed at z>10: a sample extracted from the SDSS 1 1 0 2 n a ∗ J LucaSbordone 7 Max-PlanckInstitutfürAstrophysik-Garching,Germany 1 GEPI,ObservatoiredeParis,CNRS,UniversitéParisDiderot,France ] Piercarlo Bonifacio A GEPI,ObservatoiredeParis,CNRS,UniversitéParisDiderot,France G INAF-OsservatorioAstronomicodiTrieste,Italy . h ElisabettaCaffau p - ZAH–LandessternwarteHeidelberg,Germany, o GEPI,ObservatoiredeParis,CNRS,UniversitéParisDiderot,France r t s Hans-GünterLudwig a [ ZAH–LandessternwarteHeidelberg,Germany, 2 GEPI,ObservatoiredeParis,CNRS,UniversitéParisDiderot,France v 0 As the Universe emerged from its initial hot and dense phase, its chemical composition was 1 2 extremelysimple,beinglimitedtostableHandHeisotopes,andtracesofLi. Thefirststarsthat 5 formedhadsuch initialcomposition. However,theyquicklybegantoproducea wholearrayof . 9 heavier nuclei, polluting the interstellar medium. While none among these first stars has been 0 0 detectedtodate, anincreasingsampleexistsoftheirdirectdescendants,starswith afractionof 1 heavyelementsoftheorderof1/1000ofthesolarvalue,orless. Inmostcases,suchstarsshould : v haveformedatredshiftofabout10orbeyond,andtheirchemicalcompositioncanprovidecrucial i X constraintstothenatureoftheveryfirststars. Extremelymetalpoor(EMP)starsareexceedingly r a rare. We used the low resolution spectra obtained by the Sloan Digital Sky Survey (SDSS) to search for EMP candidates: results of VLT-UVES high resolution follow-up for 16 of them is presentedhere. A newly developedautomatic abundanceanalysisand parameterdetermination code,MyGIsFOS,hasbeenemployedtoanalyzethedetailedchemicalabundancesofsuchstars. 11thSymposiumonNucleiintheCosmos 19-23July2010 Heidelberg,Germany. ∗ Speaker. (cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ Localstarsformedatz>10: asampleextractedfromtheSDSS LucaSbordone 1. Target selectionandlow-resolutionmetallicityestimate Candidate EMP stars were searched among stars with colors compatible with Halo the F- G turn-ff population, and having been observed spectroscopically by SDSS (up to data release 6). SDSS low-resolution (R=2000) spectra were analyzed by means of an automatic abundance- estimation code, comparing a number of metallic features against agrid of synthetic spectra. The stellar effectivetemperature wasestimatedbymeansofatheoretical calibration oftheT /(g-z) eff relationship. Thegravitywaskeptfixedatlogg=4inthiscandidate-selection phase. SeeLudwig etal.(2008)foramoredetailed description ofthecandidate selection phase. 2. Automated abundance analysiswithMyGIsFOS The candidate EMP stars selected were then observed by means of UVES@VLTat a typical resolution of about R=30,000, S/N∼40-50 per pixel. Three among them appeared to be carbon- enhancedmetal-poorstarsandtheirfullanalysisispresentedbyBeharaetal.(2010). Sixteenmore objects, notshowinganomalous CNOabundances, arepresented inthiscontribution. Anew,fullyautomaticparameterestimationandabundanceanalysiscodehasbeendeveloped andapplied tothefullchemical analysis ofthesestars. TheMyGIsFOScodeworksbyreplicating astandard manualabundance analysis whereT isprovided fromoutside (e.g. fromphotometry eff calibration), gravity, microturbulence and [Fe/H] are derived by measuring selected Fe I and Fe II features, and, after the atmosphere parameters are established, a full chemical analysis can be performed by measuring lines of all the other significant elements. The user selects a number of spectral regions for pseudo-continuum estimation, as well as all the needed metallic lines. Each line is then fitted against a grid of pre-calculated synthetic spectra varying in T , logg, micro- eff turbulence, [Fe/H],and[a /Fe],bya c 2 minimization scheme. Stellarparametersareiteratedfrom first-guess valuestothefinalones. AnexampleofMyGIsFOSoutput isshowninFig. 1. ParameterestimationandfullchemicalanalysiswithMyGIsFOStakesbeween15and30sec- onds per star on a mainstream PC. The code is currently available on collaboration basis, and is foreseentobereleasedtothecommunityafterdebugging, performanceassessment, anddocumen- tationarecompleted. 3. Abundance analysisresults Comparedtothesampleanalyzedinthelargeprogram“FirstStars”(Bonifacioetal.,2009,and referencestherein)thenewsampleappearsremarkablymoremetalpoor,assessingtheeffectiveness of using SDSSfortarget selection (see Fig. 2). Theselection strategy proved highly effective: all theselected targetswereconfirmedasEMPbythehighresolution analysis. In fig. 3 we present [Cr/Fe] and [Ni/Fe] plotted against [Fe/H] for the new sample along withthe“FirstStars”sample. Intheleftpanelthethree[CrII/Fe]measurements appear discrepant withrespectofthe[CrI/Fe]measurements forthesamestars,underlining theimportance ofNLTE effects for Crin EMPstars (Bonifacio et al., 2009; Bergemann & Cescutti, 2010) which were not accounted forinthepresentanalysis. Ofparticular interest is the evidence ofa sizable population ofNi-rich stars at extremely low metallicities: 8 stars out of 16 show [Ni/Fe]>0.2. While the current analysis is preliminary, it 2 Localstarsformedatz>10: asampleextractedfromtheSDSS LucaSbordone Figure 1: An exampleof a set of MyGIsFOSfinalfits overFe I (blueboxes), Si I (greenbox), andSc II (redbox)features. Darkgreyboxesoutlinepseudo-continuumregions. Thelightgrey,shadedboxisaFeI featurewhichhasbeenrejectedasbad. Figure2: ComparisonofthemetallicitydistributionofthesampleofEMPstarsextractedfromSDSS(left panel)withtheonepresentedintheESOlargeprogram“FirstStars”(rightpanel) 3 Localstarsformedatz>10: asampleextractedfromtheSDSS LucaSbordone Figure3:Intheleftpanel,[CrI/Fe]isplottedagainst[Fe/H]fortheSDSSsample(filleddots)andthe“First Stars” sample (crosses). The three red filled dots correspond to Cr II measurements in three stars of the SDSSsample.Intherightpanel,[Ni/Fe]isplottedagainst[Fe/H],withthesamesymbols. would appear that this result is indeed robust. A final analysis, including temperatures based on 3D-hydrodynamical Ha profiles,iscurrently underway. References Behara,N.T.,Bonifacio, P.,Ludwig,H.-G.,Sbordone,L.,GonzálezHernández,J.I.,&Caffau,E. 2010,A&A,513,A72 Bergemann, M.,&Cescutti,G.2010, arXiv:1006.0243 Bonifacio, P.,etal.2009,A&A,501,519 Ludwig,H.-G.,Bonifacio,P.,Caffau,E.,Behara,N.T.,GonzálezHernández, J.I.,&Sbordone,L. 2008,PhysicaScriptaVolumeT,133,014037 4