Multiple stellar Multiple stellar populations and the populations and the horizontal branch of horizontal branch of

globular clustersglobular clustersRaffaele GrattonRaffaele Gratton

INAF – Osservatorio INAF – Osservatorio Astronomico di PadovaAstronomico di Padova

CollaboratorsCollaborators

Angela BragagliaAngela Bragaglia Eugenio CarrettaEugenio Carretta Valentina D’OraziValentina D’Orazi Sara LucatelloSara Lucatello Yazan MomanyYazan Momany Chris SnedenChris Sneden Antonio SollimaAntonio Sollima

Franca D’AntonaFranca D’Antona Paolo VenturaPaolo Ventura Santi CassisiSanti Cassisi Giampaolo PiottoGiampaolo Piotto Anna Fabiola MarinoAnna Fabiola Marino Antonino MiloneAntonino Milone Alessandro VillanovaAlessandro Villanova

Single stellar population Single stellar population (SSP)(SSP)

A set of stars having:A set of stars having: Same ageSame age Same chemical composition (both Y and Z)Same chemical composition (both Y and Z) Different mass (distributed according to an IMF)Different mass (distributed according to an IMF) Possibly, binaries includedPossibly, binaries included It is described by a single isochrone in the It is described by a single isochrone in the

CMDCMD Tool widely used in stellar and galactic Tool widely used in stellar and galactic

evolution contextevolution context Stellar populations in galaxies are usually assumed Stellar populations in galaxies are usually assumed

to be reproduced by suitable weighted sums of to be reproduced by suitable weighted sums of SSPsSSPs

Stellar clusters are usually considered Stellar clusters are usually considered good examples of SSPgood examples of SSP

Evidences for multiple Evidences for multiple populations in GCspopulations in GCs

SpectroscopySpectroscopy From ’70s. But for a long time attributed to From ’70s. But for a long time attributed to

peculiar evolutionpeculiar evolution From 2001: Na-O anticorrelation on the From 2001: Na-O anticorrelation on the

MSMS PhotometryPhotometry

From ’60s (HB second parameter). However, From ’60s (HB second parameter). However, for a long time not understoodfor a long time not understood

Still from ’60s: Still from ’60s: ωω Cen Cen considered peculiar considered peculiar From ’70s: NGC2808 and NGC1851 From ’70s: NGC2808 and NGC1851 still still

considered as peculiarconsidered as peculiar From 2004: multiple MSs and SGBsFrom 2004: multiple MSs and SGBs

ESO Large Program 165-L0263

The O-Na anticorrelation is present among TO-stars and subgiants in NGC6752. For the same stars, also a Mg-Al anticorrelation is observed

This clearly rules out deep mixing as explanation for the O-Na anticorrelation

The sum of C+N abundances is not constant: a substantial fraction of O is transformed into N in some NGC6752 stars

A fraction of the stars in GCs (second generation, SG) formed from the ejecta of an earlier population (first generation or primordial population)

Carretta et al. extensive Carretta et al. extensive survey (2009)survey (2009)

(Flames@VLT2)(Flames@VLT2)

All GCs have multiple All GCs have multiple populationspopulations

Red: have Na-O anticorrelationGreen: do not have Na-O anticorrelationEmpty: not yet studied

Different symbols are for GCs in the MW, LMC or DSph’s

NGC6791

It is modulated by a It is modulated by a combination of combination of

metallicity and cluster metallicity and cluster massmass

O-Na anticorrelation and O-Na anticorrelation and HBHB

Median mass of HB stars Median mass of HB stars determined mainly by determined mainly by

[Fe/H] and age[Fe/H] and age

Gratton et al. 2010

Median masses on the HB can be derived by comparison with models

If ages are known from main sequence photometry (e.g. Marin-French et al.) mass loss by stars along the RGB can be derived

This mass loss result to be roughly a simple linear function of [Fe/H]

Multiple populations in GCs: Multiple populations in GCs: NGC2808 (MNGC2808 (MVV=-9.4)=-9.4)

Piotto et al. 2007, ApJL 661, L53Piotto et al. 2002

D’Antona et al. 2005: D’Antona et al. 2005: Na-rich stars should be Na-rich stars should be richer in Hericher in He

He-rich stars evolve He-rich stars evolve fasterfaster

if same mass loss if same mass loss evolved He-rich stars evolved He-rich stars have lower mass have lower mass they they are bluer when on the are bluer when on the HBHB

Na-O Na-O anticorrelation anticorrelation

He He HB HB

ωω Cen is the largest GC in Cen is the largest GC in the MW: Multiple RGB the MW: Multiple RGB

sequencessequences

Ferraro et al. 2004, ApJ, 603, L81 Bellini et al. 2009• The distribution of stars with metallicity

is not continuous: various episodes of star formation

• There is a metal-rich population, with [Fe/H]~-0.6 (Pancino et al.): RGB-a

ωω Cen: Main sequence Cen: Main sequence splittingsplitting

Bedin et al. 2004, ApJ 605, L125 Piotto et al. 2005, ApJ 621, 777

•There are two MSs; the blue one has ¼ of the stars

• The bluest MS is more metal-rich [Fe/H]~-1.2) than the redder one ([Fe/H]~-1.6)

• This agrees with the redder one be more populous

• But this implies a higher He-content (Y~0.4 rather than 0.25)!

• Populations suggest that the He-rich MS is connected to the extreme BHB

Giraffe@VLT2HST-ACS

He in HB-stars: He in HB-stars: expectationsexpectations

Stars distribute along the HB of a GC Stars distribute along the HB of a GC according to their massaccording to their mass

TO masses of stars in a GC should depend TO masses of stars in a GC should depend on their He-contenton their He-content

Assuming similar mass loss, stars of Assuming similar mass loss, stars of different He should distribute along the HBdifferent He should distribute along the HB

Redder HB/He-poor/O-rich/Na-poorRedder HB/He-poor/O-rich/Na-poor Bluer HB/He-rich/O-poor/Na-richBluer HB/He-rich/O-poor/Na-rich

On HB, possibility to derive He-On HB, possibility to derive He-abundancesabundances

Once [Fe/H] and ages are Once [Fe/H] and ages are known, He can be derived known, He can be derived

from coloursfrom coloursThe spread in He derived from the spread in colours masses of stars along the HB is correlated with the amplitude of the O-Na and Mg-Al anticorrelations.

This is expected if He is produced with Na and Al

Villanova et al. 2009: Villanova et al. 2009: NGC6752 (UVES@VLT2)NGC6752 (UVES@VLT2)

Diff+Rad lev.

evolved

Marino et Marino et al. 2010: al. 2010:

M4M4(Flames@ (Flames@

VLT2)VLT2)

Gratton et al. 2011: Gratton et al. 2011: NGC2808NGC2808

(Flames@VLT2)(Flames@VLT2)

Gratton et al. 2012: 47 TucGratton et al. 2012: 47 Tuc(Flames@VLT2)(Flames@VLT2)

-0.50

-0.30

-0.10

0.10

0.30

0.50

0.70

0.90

1.10

1.30

0.600 0.700 0.800 0.900 1.000

B-V

[Na/O]

Faint

Bright

The main parameter driving The main parameter driving the multiple populations is the multiple populations is

the cluster massthe cluster mass

ConclusionsConclusions (All) GCs host multiple stellar populations(All) GCs host multiple stellar populations These populations differ in their abundances of These populations differ in their abundances of

He, C, N, O, Na, Mg, Al abundances (signature of He, C, N, O, Na, Mg, Al abundances (signature of H-burning at high temperature)H-burning at high temperature)

Formation mechanismFormation mechanism The location of the stars on the HB is determined The location of the stars on the HB is determined

by a number of factorsby a number of factors A mass loss proportional to metallicityA mass loss proportional to metallicity The ages of GCsThe ages of GCs The spread in He (extension of the second population) The spread in He (extension of the second population)

related to their mass related to their mass This explains most, perhaps all the second This explains most, perhaps all the second

parameter issueparameter issue

PerspectivesPerspectives

This scenario explains many observables!This scenario explains many observables! If true, it connects the formation of GCs to If true, it connects the formation of GCs to

that of the halo (which is mainly made of FG that of the halo (which is mainly made of FG lost by GCs)lost by GCs)

Main uncertainty is the nature of the stars Main uncertainty is the nature of the stars responsible for the second generation responsible for the second generation timescale of the phenomenontimescale of the phenomenon

The cosmological implications need still to be The cosmological implications need still to be fully understood (e.g. mechanisms of star fully understood (e.g. mechanisms of star formations in disk and spheroids, missing formations in disk and spheroids, missing satellite issue, reionization, etc.)satellite issue, reionization, etc.)