Helium-enhancements in globular cluster stars from AGB pollution

1
Helium-enhancements in globular cluster stars
from AGB pollution

• Amanda Karakas1, Yeshe Fenner2, Alison Sills1,
• Simon Campbell3 John Lattanzio3
• 1 Department of Physics Astronomy, McMaster
  University Hamilton ON Canada
• 2 CFA, Harvard University, Cambridge MA, USA
• 3 Centre for Stellar Planetary Astrophysics,
  Monash University, Clayton VIC Australia

2
Outline

1. Motivation
2. Evidence for enhancement
3. Helium production in AGB stars
4. The chemical evolution model
5. Results
6. Discussion

3
Motivation

• Unusual horizontal branch morphology of NGC 2808,
  M3, M13
• Stars in the extended blue tails of this cluster
  have an enhanced amount of helium, Y 0.32
  (DAntona Caloi 2004) compared to the
  primordial (0.24)
• Recent results suggest a peculiar main-sequence
  for 2808 too (DAntona et al. 2005) now
  suggesting Y up to 0.40
• Omega Centauri has a clearly defined double main
  sequence (Bedin et al. 2004 Piotto et al. 2005)
• Norris (2004) used isochrones with Y 0.38 to
  fit the bluest stars on the MS of ? Centauri.

4
? Centauris main sequence
from Norris (2004)
5
Peculiar main-sequence NGC 2808
Isochrones with age 13 Gyr and Y 0.24, 0.30,
0.40
Images from DAntona et al. (2005)
6
Extended blue HB stars NGC 2808
Horizontal Branch using data from Bedin et al.
(2000)
from DAntona Caloi (2004)
7
The self-pollution scenario

• Attributes a previous generation of more massive
  stars as being responsible for the abundance
  anomalies we observe today
• Hot bottom burning (HBB) provides an ideal
  environment (at least qualitatively) to convert C
  and O to N, Ne to Na, Mg to Al and H to Helium
• Helium suggested to have come from
  intermediate-mass AGB
• Massive AGB models can result in final surface Y
  0.36 (Karakas 2003, PhD thesis) it is unclear
  if will result in Y gt 0.30 after dilution

8
The self-pollution scenario Our approach

• We approach this problem from a global
  perspective
• Use a globular cluster chemical evolution model
  to follow the evolution of the intracluster gas
• Model previously to follow the evolution of Na,
  Mg and Al in NGC 6752
• We follow helium, C, N and O and heavy elements
  (in this case barium)
• This time we use two independent sets of AGB
  yields
• From Simons models (Campbell et al. 2004, used
  in Fenner et al. 2004)
• Ventura, DAntona Mazzitelli (2002)

9
Helium production in AGB stars

• Helium mixed to the surface by the first and
  second dredge-up as a result of the convective
  envelope moving into a region of partial (or
  complete) H-burning
• The third dredge-up (TDU) and hot bottom burning
  (HBB) further increase Y in the envelope
• The amount of 4He expelled into the IMS from
  Simons models and Ventura et al. (2002) agree to
  within 30!
• The net result of hydrogen fusion is the
  production of 4He hence the yields are fairly
  robust

10
Helium production in AGB stars

• Z 0.004 models (Fe/H -0.7)

11
The chemical evolution model

• Original study Fenner et al. (2004, MNRAS)
• Prompt initial enrichment to get the cluster gas
  to Fe/H -1.4 using Chieffi Limongi Pop. III
  SN yields
• Second stage we form AGB stars out of this gas,
  we then follow the evolution of the gas as these
  AGB stars pollute the cluster
• Besides using a different set of AGB yields and
  changing the IMF, all other parameters the same
  as original study
• Note that Ventura et al. (2002) yields are scaled
  solar whereas our models have O/Fe 0.4
  initially

12
The initial mass function

• One of the most uncertain parameters in the
  chemical evolution model
• Determines how many stars of a given mass
  contribute to the chemical enrichment of the
  cluster
• We test varying the IMF
• Salpeter with slope 1.31 (our standard)
• Using a flat Salpeter with slope 0.3
• Intermediate-mass star bias

13
The initial mass functions used
14
Evidence (or lack of) for a top heavy IMF?

• Evidence for
• DAntona Caloi (2004) need factor of 10 more 4
  to 7 Msun stars to produce He enhancements in GC
  stars
• To produce the observed number of C, s-element
  rich metal-poor stars Lucatello et al. (2005)
  need more 1 to 5 AGB stars in the early galaxy
• Evidence against
• Bekki Norris (2005) find a top-heavy IMF would
  likely result in the disintegration of the
  cluster (applicable to helium coming from massive
  OR AGB stars)
• n-body simulations by Downing Sills suggest a
  top heavy IMF is not supported in GCs for
  dynamical reasons
• Tilley Pudritz (2005) studied the IMF that
  results from 3D simulations with MHD turbulence,
  conclude IMF likely to be universal (except in
  Z0 gas)

15
Results standard IMF
Y 0.29
Y 0.26
Simons yields
Ventura et al. (2002) yields
16
IMS-biased IMF Our yields
Y 0.35
17
IMS-biased IMF Ventura et al. yields
Y 0.29
18
Discussion

• Using our standard IMF, the helium abundance in
  the gas did not exceed Y 0.30
• Require the IMS-biased IMF to produce Y 0.35
  but then note the large enhancements in CNO,
  barium
• Result for Y largely independent of AGB yields
  used
• Results NOT supportive of AGB stars producing the
  large helium enhancements
• Given the difficulties in obtaining a
  quantitative match between AGB models and GC
  stars without much fine-tuning suggests that AGB
  stars are not the solution
• Too pessimistic? There are many model
  uncertainties and unknowns

19
Uncertainties

• Uncertainties concerning convective model
• Rotational mixing?
• Efficiency of third dredge-up could be less than
  we predict
• Mass loss behaviour at low Z unknown
• Super-AGB stars? Far from clear how they
  contribute
• Binary interactions how will it affect the
  yields?
• Massive AGB stars sink toward the centre of GCs
• Probability of binary interactions higher in
  centre
• Primordial binary fraction in GCs? (Ivanova et
  al. suggests it was high, 100)

20
Summary

• We have followed the chemical evolution of
  helium, CNO using two independent sets of AGB
  yields
• Results not supportive of an AGB solution
• AGB stars may have produced some helium but
  current models cannot account for the largest
  enhancements (Y ? 0.30)
• At least, not without assuming a top-heavy IMF
• This also leads to large enhancements of CNO,
  s-process elements
• Evidence for such an IMF not overwhelming
• Perhaps Bekki Norriss idea of pollution from
  outside the cluster also application to other GCs
  besides ? Centauri?