1: %%\documentclass[manuscript]{aastex}
2: \documentclass[12pt,preprint]{aastex}
3: %%\documentclass[preprint2]{aastex}
4: %%\documentclass{emulateapj}
5:
6:
7: \shorttitle{}
8: \shortauthors{Newsham \& Terndrup}
9:
10:
11: \begin{document}
12:
13: \title{Observational consequences of the hypothesized helium rich
14: stellar population in $\omega$ Centauri.}
15:
16:
17: \author{G. Newsham and D. M. Terndrup}
18: \affil{Department of Astronomy, The Ohio State University,
19: Columbus, OH 43210-1173}
20: \email{newshamg@astronomy.ohio-state.edu,terndrup@astronomy.ohio-state.edu}
21:
22:
23: \begin{abstract}
24:
25: In response to the proposed high helium content stars as an
26: explanation for the double main sequence observed in $\omega$
27: Centauri, we investigated the consequences of such stars elsewhere
28: on the color-magnitude diagram. We concentrated on the horizontal branch
29: where the effects of high helium are expected to show themselves more
30: clearly. In the process, we developed a procedure for comparing
31: the mass loss suffered by differing stellar populations in a
32: physically motivated manner. High helium stars in
33: the numbers proposed seem absent from the horizontal branch of
34: $\omega$ Centauri unless their mass loss history is very different
35: from that of the majority metal-poor stars. It is possible
36: to generate a double main sequence with existing $\omega$ Centauri
37: stars via accretion of helium rich pollution consistent with
38: the latest AGB ejecta theoretical yields, and
39: such polluted stars are consistent with the observed HB
40: morphology of $\omega$ Centauri. Polluted models are
41: consistent with observed merging of the main sequences as opposed
42: to our models of helium rich stars. Using the
43: $(B-R)/(B+V+R)$ statistic, we find that the high helium bMS stars
44: require an age difference compared to the rMS stars that is too
45: great, whereas the pollution scenario stars have no such conflict
46: for inferred $\omega$ Centauri mass losses.
47:
48: \end{abstract}
49:
50:
51: \keywords{globular clusters: general --- globular clusters:
52: individual ($\omega$ Centauri)}
53:
54:
55: \section{Introduction}
56:
57: The globular cluster $\omega$ Centauri has long been known to
58: exhibit a wide range of metallicities
59: \citep[e.g.,][]{dic67,fre75,sun96}, possibly in discrete
60: subpopulations. \citet{nor96} presented an extensive survey of
61: Ca-triplet abundances, revealing a majority metal-poor component
62: at ${\rm [Ca/H]} = -1.4$, an intermediate metal-poor peak at ${\rm
63: [Ca/H]} = -0.9$, and a long tail extending up to a ${\rm [Ca/H]} =
64: -0.3$. \citet{pan00} discussed wide-field $BI$ photometry and the
65: Norris et al.\ abundances, identifying four peaks at ${\rm [Ca/H]}
66: = -1.4$, $-1.0$, $-0.5$ and $-0.1$. The [Ca/Fe] ratio is almost
67: flat as a function of [Fe/H] \citep{nor95} with an average value
68: $\langle{\rm [Ca/Fe]}\rangle \approx +0.4$, so the first two peaks
69: of the abundance distribution have ${\rm [Fe/H]} \approx -1.7$ and
70: $-1.3$, respectively.
71:
72: The peculiarities of $\omega$ Centauri were further revealed by
73: the discovery of multiple turn-offs and a double main sequence
74: (MS) in HST photometry by \citet{bed04}. As seen in their Figure
75: 1, the two main sequences are clearly separated by $\Delta (V-I)
76: \sim 0.06$ over several magnitudes in $V$, with the region
77: between them almost devoid of stars. It can be seen that towards
78: faint magnitudes the two main sequences merge together, this is even
79: more obvious in the latest data of \citet{vil07}. The blue main sequence (bMS)
80: is less populated than the red main sequence (rMS), comprising
81: about 25\% to 35\% of MS stars. As discussed in \citet{bed04}, if
82: metallicity alone were controlling the morphology of the main
83: sequence, the MS would be about 0.03 magnitude in width with a
84: concentration toward the blue edge corresponding to the ${\rm
85: [Fe/H]} = -1.7$ majority ($\approx 65\%$) population, a tail towards
86: the red of stars corresponding to the intermediate metallicity
87: population ($\approx 30\%$) and a small ($\approx 5\%$) of the
88: stars redder still.
89:
90: Of the various hypotheses put forth in \citet{bed04} to explain
91: the double main sequence, most intriguing has been the idea that
92: the bMS stars have an unusually high helium content. \citet{nor04}
93: noted that the rMS and bMS stars were present in a ratio of 2:1,
94: which was approximately like the ratio of stars with ${\rm [Fe/H]}
95: = -1.7$ to those near ${\rm [Fe/H]} = -1.3$. Taking the Revised
96: Yale Isochrones \citep{gre87} and fitting them to a synthetic
97: color magnitude diagram (CMD) of $\omega$ Centauri, Norris found
98: that the double main sequence can be reproduced if the
99: intermediate metallicity population is more helium rich by $\Delta
100: Y \approx 0.10 \-- 0.15$ than the metal-poor population.
101:
102: \citet{pio05} undertook a spectroscopic investigation of the bMS
103: and rMS stars in $\omega$ Centauri. They obtained VLT spectra of
104: 17 rMS stars and 17 bMS stars which were combined respectively to
105: form a single rMS and single bMS composite spectrum. They found
106: that the rMS stars have ${\rm [Fe/H]} = -1.57$ and the bMS stars
107: have ${\rm [Fe/H]} = -1.26$, close to the largest two peaks in the
108: distribution of abundances on the giant branch. \citet{pio05} also
109: fit new stellar isochrones to the bMS and showed that it can
110: best be modelled with 0.35 $\lesssim$ Y $\lesssim$ 0.45, the best
111: value being $Y = 0.38$. This confirms the
112: \citet{nor04} result but with metallicities directly from the MS
113: and more up-to-date stellar models. Similar conclusions were drawn
114: by \citet{lee05}, who determined a best fit to the observations
115: implied a bMS helium content of 0.38, implying $\Delta Y \approx
116: 0.15$. \citet{dan05} have reached a similar conclusion of a spread
117: of helium content for the main sequence stars of NGC 2808 using
118: HST photometry. They noted that the color distribution is not
119: Gaussian and is wider than that expected for a single metallicity
120: population. They found some 20\% of the stars are much bluer than
121: expected and from their stellar models conclude that the helium
122: mass fraction of these stars is $Y \sim 0.4$.
123:
124: Attempts to explain the high helium content of the bMS stars have
125: not been entirely successful. If massive stars in the metal-poor
126: population were responsible for both the helium and metal
127: enrichment of the intermediate metallicity stars, values of
128: $\Delta Y / \Delta Z \sim 100$ are implied \citep[e.g.,][]{nor04},
129: far in excess of the canonical value $3 - 4$ \citep[e.g.,][]{pag92}.
130: \citet{pio05} noted that massive stars should also produce a large
131: amount of carbon, nitrogen, and oxygen, but in $\omega$ Centauri
132: the total $[(\rm{C+N+O)/Fe}]$ abundances for all stars is about +0.4 dex
133: \citep{nor95}. This is also the case for enrichment by AGB star
134: winds \citep{kar06}. \citet{bek06} extensively discussed
135: enrichment by intermediate mass AGB stars, massive stars that produce
136: helium-rich winds \citep[see][]{mae06} and SNe II. They
137: concluded that for reasonable Initial Mass Function (IMF)
138: choices $\omega$ Centauri could not have survived disintegration
139: due to cluster mass loss. The most serious problem with the AGB
140: scenario is that the total mass of ejecta from rMS AGB stars is
141: too small to be consistent with the observed fraction of the bMS
142: of $\omega$ Centauri. If $\omega$ Centauri formed with a very
143: unusual IMF, producing little or no SNe II, then the observed bMS
144: fraction can be reproduced; that is, however, inconsistent with the
145: known presence of neutron stars in $\omega$ Centauri. Similarly,
146: the number of massive stars with helium-rich winds do not
147: contribute a large enough mass fraction of ejecta for allowable
148: IMF choices. In addition, there would be little difference in
149: metallicity between the rMS stars and the massive star wind
150: ejecta, which is at odds with the \citet{pio05} result on the
151: metallicity difference between the bMS and rMS populations.
152: Conversely, the problem with the SNe II scenario is that the heavy
153: element abundances would be too high to be consistent with the bMS
154: stars forming from SNe II ejecta. The conclusion was that neither
155: alone or combined could these three sources explain the bMS stars
156: as being formed exclusively from the ejecta of rMS stars. Recent work
157: by \citet{cho07} have concluded the only way to achieve such high
158: helium enrichments without a commensurate level of metal production
159: is by appealing to Population III star ejecta as the source of the
160: material that formed the bMS. However, they admit this scenario is itself
161: of an extreme nature.
162:
163: Several studies have tried to find the evolutionary descendants of
164: the He-rich population on the horizontal branch (HB). Stars with
165: high ($Y \gtrsim$ 0.35) helium content would reach the tip of the
166: RHB (TRGB) with a reduced mass since they have shorter lifetimes.
167: This would make the HB both hotter (bluer) and more luminous when
168: compared to stars of lesser helium content \citep{dan02}.
169: \citet{dcr00} \& \citet{mom04} noted that $\omega$ Centauri does
170: have a noticeably blue HB with substantial numbers of Extreme
171: Horizontal Branch (EHB) stars with an effective temperature
172: ($T_{\rm eff}$) greater than 20,000K. This has been noted as
173: qualitative evidence of the bMS stars progeny populating the HB of
174: $\omega$ Centauri \citep{nor04,pio05,lee05}.
175:
176: Recent observations of the RR Lyrae variables in $\omega$ Centauri
177: by \citet{sol06} showed spectroscopically the presence of both
178: metal-poor and intermediate metallicity stars in their sample of
179: 74 RR Lyrae variables. These intermediate metallicity variables
180: are not consistent with a high helium enhancement, in that they do
181: not have a higher luminosity and they have an ordinary
182: relationship between period and luminosity. Thus, at least some
183: intermediate metallicity stars have normal helium abundances. As
184: \citet{sol06} themselves noted, this is not necessary a strong
185: argument against the enhanced helium hypothesis, since bMS stars
186: would not appear in the RR Lyrae instability strip unless they are
187: some 4 Gyr younger than the rMS population or have undergone a
188: radically different mass loss history.
189:
190: In this paper we perform a quantitative calculation of the color
191: distribution of HB stars for various helium abundances. We also
192: investigate whether accretion of helium-rich material onto
193: existing stars can produce the observed double main sequence, and
194: follow the evolution of these stars to the HB. A description of
195: the stellar evolutionary calculations we performed is presented in
196: \S 2, which also discusses models of stars with homogeneous high
197: helium abundance. Models with a high surface abundance of helium
198: are presented in \S 3. In \S 4 we summarize our results.
199:
200: \section{Models with homogeneous enhanced helium}
201:
202: \subsection{Stellar Evolution Models}
203:
204: We computed evolutionary tracks appropriate for $\omega$ Centauri
205: using the modern version \citep{sil00} of the Yale Rotating
206: Stellar Evolution Code \citep[YREC,][]{gue92}; in the
207: computations, the rotation routines were turned off. Nuclear
208: reaction rates are taken from \citet{gru98} and the heavy element
209: mixture is that of \citet{gre93}. Gravitational settling of helium
210: and heavy elements is included in these models, as in
211: \citet{bah90} and \citet{tho94}. We also include neutrino losses
212: from photo, pair, bremmstrahlung and plasma neutrinos, following
213: \citet{ito96}.
214:
215: We use the OPAL opacities \citep{igl96} for temperatures $\log T
216: \geq 4$. For lower temperatures, we employ the molecular opacities
217: of \citet{ale94}. The alpha-enhanced low {\it T} opacities
218: are known to be in error \citep{wei06} but this does not
219: affect our calculations as we used the solar mixture opacities.
220: For regions of the star with $\log T \geq 3.7$,
221: we use the 2001 version of the OPAL equation of state
222: \citep{rog96}\footnote{Web updates at http://www-phys.llnl.gov/Research/OPAL/Download}
223: and for $\log T < 3.7$ we take the equation of state from \citet{sau95}.
224:
225: For the surface boundary conditions, we use the stellar atmosphere
226: models of \citet{all95}, which include molecular effects and are
227: therefore appropriate for low mass stars. We use the standard
228: B\"{o}hm-Vitense mixing length theory \citep{boh58,cox68} with $\alpha
229: = 2.013$, obtained by calibrating a solar model against
230: observations of the solar radius ($6.9598 \times 10^{10}$ cm) and
231: luminosity ($3.8515 \times 10^{33}$ erg s$^{-1}$) at the present
232: age (4.57 Gyr) and metal fraction ($Z = 0.01757$) of the Sun. By
233: comparison, models of the Sun excluding diffusion require $\alpha
234: \approx 1.7$ \citep[e.g.,][]{pin03}.
235:
236: The helium content of ``ordinary'' (i.e., not enhanced in helium)
237: $\omega$ Centauri stars was computed using the primordial helium
238: abundance abundance $Y_p = 0.23$ and the enrichment parameter
239: $\Delta Y / \Delta Z = 2.0$, in line with the treatment in the
240: Yale-Yonsei isochrones of \citet{yi03}. The value for $Y_p$ is
241: within, but on the low end of, the range of current estimates
242: \citep[e.g.,][and references therein]{bono02,thuan02,oli04}
243: although some values are considerably higher
244: \citep[e.g.,][]{fk06}. This would yield a helium abundance for the
245: Sun of $Y = 0.265$ in models lacking diffusion. Models compatible
246: with helioseismology which include both rotational mixing and
247: diffusion have a surface abundances $Y_{\odot,{\rm surf}} = 0.249
248: \pm 0.003$ and an initial solar composition $Y_\odot = 0.274$,
249: $Z_\odot = 0.019$ \citep{bah01}. Such values would imply $\Delta Y
250: / \Delta Z = 2.3$, close to $\Delta Y / \Delta Z = 2.1 \pm 0.4$
251: inferred for nearby field stars \citep{jim03}, but larger than
252: $\Delta Y / \Delta Z = 1.3 \pm 0.2$ from open clusters in the
253: solar neighborhood \citep{an07}. At the low metallicities of the
254: stars in $\omega$ Centauri, however, the properties of the models
255: for unenhanced stars are insensitive to the exact value of the
256: helium enrichment parameter.
257:
258: All models began with a zero-age main sequence (ZAMS) model
259: evolved from the deuterium-burning birthline. The ZAMS was
260: defined as the point at which the core hydrogen abundance had
261: dropped by 2\% from the initial value. Horizontal branch models
262: were created by rescaling the core masses, metallicity and
263: envelope masses of horizontal branch models created for the YREC
264: code provided by A.\ Sills (private communication). The core
265: masses were determined from the corresponding RGB tip (TRGB) core
266: masses from our evolved models and Zero Age Horizontal Branch
267: (ZAHB) sequences developed by rescaling the envelope masses to
268: define the ZAHB. This method of creating ZAHB sequences has been
269: found to be adequate by \citet{ser05} with deviations at the few
270: percent level being found for the hottest HB stars when compared
271: to ZAHB models calculated directly through the core helium flash.
272:
273: We calculated theoretical isochrones by interpolating the model
274: evolutionary tracks using a scheme based upon the algorithm of
275: \citet{ber92}. The tracks, all with ages above 7 Gyr, have a
276: simple topology, so we used five major evolutionary points to
277: create the isochrones. These were the zero age main sequence
278: point, the point of the exhaustion of core hydrogen, the base of
279: the red giant branch, the red giant bump and the tip of the red
280: giant branch (i.e., at the helium flash). Observational colors and
281: magnitudes were calculated from $M_{\rm bol}$, log {\it g} and $T_{\rm
282: eff}$ using the transformations of \citet{van03}.
283:
284: \subsection{Model populations}
285:
286: \citet{nor04} and \citet{lee05} both modelled the bMS of $\omega$
287: Centauri as a population with ${\rm [M/H]} = -1.27$ and a helium
288: enhancement of $\Delta Y = 0.12 - 0.18$ over the rMS.
289: \citet{nor04} associated the bMS population with the second most
290: common RGB metallicity, while \citet{lee05} used the the bMS
291: metallicity as determined spectroscopically in \citet{pio05}.
292:
293: In this present work we created isochrones for the two main
294: sequence populations using parameters shown in Table
295: \ref{parameters}, where we list the metal content, helium content,
296: metallicity, the TRGB total mass, and the TRGB helium core mass.
297: Here we are using models for the bMS that are homogeneous in helium
298: content; below we compute models in which the outer convective
299: zones are polluted with helium-rich material. Hereafter, the
300: proposed helium-rich population in all evolutionary states shall
301: be referred to as the bMS population, while the majority
302: metal-poor stars shall be called the rMS
303: population. We later discuss polluted models where we refer to a
304: population with the metallicity of the bMS stars but with a normal
305: helium content; we shall call this the intermediate metal poor
306: population (MintP). Its characteristics are also shown in Table
307: \ref{parameters}.
308:
309: In Figure \ref{fig:isochrones} we show isochrones for the bMS and
310: rMS populations in the $M_{V}, \bv$ plane; both groups have an
311: age of 13 Gyr. The two populations overlap at the turnoff (TO) but
312: separate again high on the RGB, where the bMS stars now appear on
313: the cool side of the rMS giant branch. As expected from theory,
314: the ZAHB of the bMS is more luminous (by $\sim 0.8$ magnitudes)
315: than that of the rMS. Both sequences merge on the hot tail of the
316: blue HB.
317:
318: \subsection{Mass loss methods}
319:
320: In order to populate the ZAHB sequences we need to parameterize
321: the mass loss at the TRGB. There are two approaches that are used
322: to calculate a distribution of HB masses. In the first, mass loss
323: is parameterized with mean value ($\Delta M$) and a dispersion
324: $\sigma_{M}$; the form of the mass loss distribution is often
325: taken as a Gaussian. Using this approach, \citet{lee94} and
326: \citet{cat00} studied the HB color distribution of several
327: globular clusters, and found mean HB masses of $0.63 - 0.75
328: M_{\sun}$, implying mass losses of the order $0.1 - 0.2 M_{\sun}$.
329: The presence of extremely blue HB stars in some clusters implies
330: even greater mass losses, of the order $0.3 M_{\sun}$ (cf.\ Table
331: \ref{parameters}), since these stars have a very thin ($M \leq
332: 0.002 M_\sun$) hydrogen envelope surrounding the helium core.
333:
334: Alternatively, some authors calculate mass loss during the RGB
335: evolution, employing various empirical approximations. The mass
336: loss scales with one or more physical parameters of the RGB
337: star \citep[see][for a summary]{cat00}. It has long been
338: questioned whether or not these formulae accurately estimate the
339: mass loss rates of RGB stars. \citet{ori02} showed that the
340: observed loss rates are at least an order of magnitude greater
341: than the predictions of the empirical formulae, and the rates do
342: not seem to follow the dependence on luminosity, gravity, radius,
343: or metallicity which appear in the empirical expressions. They
344: also concluded that the mass loss occurs only in the last $10^{6}$
345: years of RGB evolution (i.e., very near the TRGB) and is episodic
346: in nature; the observed time scale for mass loss is greater than a
347: few decades and less than a million years.
348:
349: In this paper, we follow the first approach. The models are run
350: to the TRGB with no mass loss. At the helium flash, we generate a
351: distribution of HB masses to populate the ZAHB. By doing so, we
352: do not need to follow the effect of mass loss on the structural
353: evolution of stars near the TRGB.
354:
355: Because the rMS and bMS stars reach the TRGB with quite different
356: masses (Table \ref{parameters}), they may have a different average
357: mass loss. Simply assuming a mean mass loss of order
358: $0.2M_{\sun}$ for the bMS would remove the entire envelope and
359: part of the helium core, yet we know from EHB star spectra that
360: hydrogen is present in these stars. We note that any mass loss
361: greater than $0.174M_{\sun}$ for the bMS stars at an age of 13 Gyr removes
362: the entire hydrogen envelope of the star.
363:
364: We consider several cases for computing the mass loss for bMS
365: stars. The first case (Case I) corresponds to the situation in
366: which the mass loss takes place on time scales shorter than
367: the Kelvin-Helmholtz time, perhaps as short as the bottom of the
368: range of time scales discussed by \citet{ori02}. Whatever the
369: mechanisms operate to produce the mass loss, they have to remove
370: material from the star by doing work against the gravitational
371: potential. In Figure 2 we plot the acceleration of gravity at an
372: interior point against $\Delta M$; the latter is the total
373: stellar mass minus the mass at the interior point. Shown are TRGB
374: models of the bMS and rMS populations at an age of 13 Gyr. The bMS
375: model is distinctly more compact in structure than the rMS star.
376: Rapid mass loss, therefore, is likely to remove a smaller amount
377: of material in helium-rich stars.
378:
379: The second case (Case II) assumes that the time scale for mass
380: loss is long, and that both populations have the same dependence of
381: loss rate on luminosity. This implies that the total mass loss
382: for the rMS and bMS stars would be in proportion to their
383: evolutionary timescales near the TRGB. The bMS stars evolve more
384: rapidly than the rMS stars, so in this case they would also have a
385: lower amount of material removed, though more than in Case I.
386:
387: We also compute (Case III) mass loss using an empirical mass loss
388: law, the Reimers prescription as generalized in \citet{cat00}.
389: Such a prescription assumes that the mass loss rate includes a
390: dependence on the luminosity, temperature and gravity of the star and that it
391: occurs over a large timescale. Thus the total mass loss can be
392: approximated as the product of the evolutionary timescale of the
393: star on the upper RGB and the mass loss rate from the Reimers
394: formula. In this case the mass loss is approximately the same in both rMS and bMS stars.
395:
396: In Table \ref{deltam}, we show various $\Delta M$ for for rMS stars
397: and the corresponding cases for mass loss of bMS stars. The
398: lowest relative mass loss is for Case I, while the highest is for
399: Case III. In the discussion following we calculate HB models
400: using Case I. We justify this choice on the basis that it represents
401: the most conservative case of mass loss, the other cases would
402: push the HB stars from the bMS population to even higher temperatures
403: producing even more of a clump in the HB tail that is not seen. We also
404: feel that in light of the observational data of \citet{ori02} that
405: Case I is the physically most likely choice for mass loss. We
406: implemented the mass loss by assuming
407: a $\Delta M$ with a small dispersion, $\sigma_{M}$, to be subtracted from
408: the TRGB mass for rMS stars. We found the mass of the immediate ZAHB star
409: progenitor from the mass of the star at the TRGB on the
410: appropriate isochrone. We took the $\Delta M$ value
411: and found the local acceleration due to gravity, $g$, in the corresponding
412: rMS TRGB stellar model. We then applied these values of $g$ to
413: the bMS model and calculated the corresponding $\Delta M$ and
414: $\sigma_{M}$ for the bMS model. The resulting ZAHB population
415: distributions were proportioned in accordance to the observed
416: numbers of rMS and bMS stars, with 70\% from the rMS and 30\% from
417: the bMS.
418:
419: \subsection{Results}
420:
421: Figure \ref{fig:hbmodels} displays an example of the relative
422: locations of the bMS and rMS stars on the HB. The plot is in the
423: observational ($M_{V}, \bv$) plane. The rMS population is shown
424: as open triangles, and the bMS population as filled circles. The
425: isochrones are also shown as solid and dashed lines, respectively. Here,
426: both populations have an age of 13 Gyr. The ZAHB is populated with
427: $\Delta M = 0.24 M_{\sun}$ and a dispersion of $\sigma_{M} = 0.015 M_{\sun}$ for
428: the rMS stars and the equivalent $\Delta M = 0.143 M_{\sun}$
429: with a dispersion of $\sigma_{M} = 0.009 M_{\sun}$ for
430: the bMS stars. The ZAHB populations stay on the hot tails of both
431: ZAHB sequences, although the rMS population resides to the cooler
432: side and is redder by about 0.15 magnitudes in $\bv$ and brighter
433: on average by some 2 magnitudes in $V$.
434:
435: In Figure 4 we see the effect of varying the mass loss
436: prescription with $\Delta M = 0.20$ and $0.28 M_{\sun}$ for
437: the rMS stars and equivalent $\Delta M$ of 0.118 and 0.168
438: $M_{\sun}$ for the bMS stars of 13 Gyr age. The stars are seen to
439: remain on the hot tail of the ZAHB sequence with $\bv$ $\lesssim
440: -0.05$. For the rMS population, stars of the lowest 0.20 M$_{\sun}$
441: mean mass loss start to appear on the horizontal region of the
442: ZAHB extending all the way up to $\bv \sim 0.2$.
443:
444: The effect of age on the ZAHB distribution is also shown in Figure
445: 4 for the bMS stars at 7 Gyr age ($\Delta M$ = 0.143 M$_{\sun}$).
446: This shows we can populate the ZAHB from the hot tail at $\bv$
447: $\lesssim -0.1$ at 13 Gyr all the way to $\bv$ = 0.6 at the
448: extreme red end for a 7 Gyr bMS population. To clearly differentiate
449: a high helium bMS population we should look for a more luminous stellar
450: population at $\bv$ $\gtrsim$ 0.0 where the bMS stars populate a
451: ZAHB some 0.8 magnitudes brighter. This effect is likely to result
452: for bMS populations 2 to 6 Gyr younger than the rMS population
453: with greater mass losses requiring a corresponding greater age
454: difference. For bMS and rMS populations at the same 13 Gyr age the
455: differences on the ZAHB is primarily one of color alone and the
456: rMS stars being on average about 0.1 magnitudes in $\bv$ redder
457: than the bMS ZAHB stars, though for equivalent mass losses the
458: rMS stars are about $1 \-- 2$ magnitudes brighter in $V$.
459:
460: \subsection{Comparison with observations}
461:
462: In Figure \ref{fig:cmd}, we plot the CMD of $\omega$ Centauri ground-based
463: data of \citet{rey04}. Also shown are the 13 Gyr isochrones and
464: ZAHBs of the bMS and rMS populations from Figure
465: \ref{fig:isochrones}. We followed Rey et al.\ in adopting $E(\bv) =
466: 0.12$ and $(m - M)_V = 14.1$. Figure \ref{fig:rrstrip} shows the
467: HB only along with the location of the RR Lyrae instability strip
468: from \citet{bon95}.
469:
470: On the CMD, the majority of the HB stars lie within $14.4 < V <
471: 16.4$ and $0.00 < \bv < 0.32$. Following \citet{nor04}, we make
472: the assumption these are mainly from the rMS population.
473: If the age of the rMS population is 13 Gyr, then the rMS stars
474: have lost a mass of $0.168 - 0.247 M_{\sun}$. For a Case I mass
475: loss, the bMS stars at 13 Gyr would have lost $0.102 -
476: 0.147M_{\sun}$. This corresponds approximately to the location of
477: the filled circles on Figure \ref{fig:altdeltam}, so we should see
478: the bMS stars between $-0.2 \leq \bv \leq -0.15$ and on average about $1.5 \-- 3.0$
479: mag fainter in $V$ than the main HB. This would place them on
480: the blue tail of the helium-rich HB, with $16.5 \leq V \leq 18.5$.
481: For mass loss amounts described by Case II or III, the bMS stars
482: would be even bluer and fainter forming a distinct clump at the
483: bottom of the HB tail. Although there are many such stars on the CMD (Fig.\
484: \ref{fig:cmd}) they are not present as expected from the relative
485: numbers on the main sequence (\citet{nor04} reached a similar conclusion)
486: in which the bMS composes about 37\% of the bMS
487: plus rMS total \citep[e.g.,][]{fer04,sol07,vil07}.
488:
489: We also checked the deeper $\omega$ Centauri HST CMD of \citet{fer04} using data
490: furnished by A. Sollima (private communication).
491: In analyzing the HST data we find that the EHB stars compose
492: some $25\% \pm 3\%$ of the HB population due to the bMS and rMS stars.
493: This should be contrasted with the $37\% \pm 2\%$ derived in the latest MS data
494: by \citet{vil07}. The 10\% of stars of higher metallicity that
495: are not attributed to the bMS and rMS populations would not appear
496: in the main clump of the HB for the inferred rMS mass loss as determined from our theoretical
497: models discussed above so we counted from the main clump and
498: blueward giving us the rMS and bMS stars only.
499: Note that since the HB lifetime for both rMS and bMS stars is similar, as
500: expected from theory and confirmed in our models, then this difference
501: in the number of bMS to rMS HB stars is not due to the effects of a
502: different HB evolutionary lifetime. Even if {\it all}
503: the stars in the blue tail of the HB are in the bMS population, they number
504: only about 25\% of the stars on the main part of the HB.
505:
506:
507:
508: It does not help to assume that the bMS stars are younger than 13
509: Gyr, and so have redder colors. If the bMS population is 2 Gyr
510: younger, our models predict that they would be found on the ZAHB
511: with colors $0.05 < \bv < 0.16$ and brighter than the rMS HB
512: by 0.8 magnitudes in $V$. When we count the stars in this
513: region of the CMD, we find that they number $\sim 1.3\%$ of the
514: number of stars main clump on the HB. Some of these stars in
515: this region of the CMD are undoubtedly rMS stars that have evolved
516: off the ZAHB, thus reducing the number of possible bMS stars in
517: that region. It thus appears that we do not see the substantial
518: number of stars in the CMD at the location we would expect the bMS
519: stars to appear. These calculations are also consistent with the
520: conclusions of \citet{sol06}, who failed to find any RR Lyrae
521: stars in $\omega$ Centauri that were helium rich, as indicated by
522: a high luminosity. Such stars would be in the instability strip
523: only if they were $2 \-- 5$ Gyr younger than the rMS population. Such
524: young ages are excluded by the analysis of the turnoff region
525: \citep{sol05,lee05}.
526:
527: Using the Rey et al.\ CMD, we also calculated the $(B-R)/(B+V+R)$
528: statistic \citep{lee94}, where $B$, $V$ and $R$ are the numbers of
529: red HB (RHB), RR Lyrae variables and blue HB (BHB) stars
530: respectively. Following \citet{rey04}, we define the RHB stars as
531: having $14.2 < V < 14.85$ and $0.55 < \bv < 0.71$. In total we
532: find 2781 HB stars in the $\omega$ Centauri CMD of which 2497 are
533: BHB stars, 149 in the RR Lyrae instability strip, and 135 are RHB
534: stars. This gives us $(B-R)/(B+V+R) = 0.85 \pm 0.01$, with the
535: error arising from counting statistics. We shall use this result
536: later as a discriminator in discussing polluted stellar models
537: versus homogeneous helium-rich bMS stars. Using the HST data of
538: \citet{fer04} we calculated $(B-R)/(B+V+R) = 0.85 \pm 0.01$ which
539: is identical to the value we determined using the \citet{rey04} data.
540:
541:
542: \section{Pollution models}
543:
544: \subsection{Construction of the models}
545:
546: As noted by \citet{bek06}, the amount of helium required to form
547: a later generation of helium rich bMS stars is far greater than
548: would be produced by massive stars in the rMS population, unless
549: the initial mass function was top-heavy. The large amount of
550: helium could have come from the ejecta of a much larger stellar
551: population, perhaps an ancient host galaxy of which the current
552: $\omega$ Centauri is the remaining nucleus; this was the scenario
553: advanced by \citet{bek06}. Alternatively the helium could be
554: the end product of an initial Population III population as advanced
555: by \citet{cho07}.
556:
557: An alternative scenario, which we admit is somewhat contrived, is
558: that helium-rich material was accreted onto pre-existing rMS or
559: MintP stars; the latter have $\rm[M/H] = -1.27$ but an ordinary
560: helium abundance. Such stars will appear on the hot side of the
561: rMS, as required by observation of the double main sequence, but
562: the amount of helium required would be far less than in the case
563: of bMS stars that are helium-rich throughout. As the
564: helium-polluted stars ascend the giant branch, the deepening
565: convection zone would erase the abundance gradient in these stars,
566: and the final product would be expected to have TRGB total masses
567: and core masses that are more like those of the rMS population.
568:
569: The question of accretion onto globular cluster stars has been addressed
570: in the past by \citet{tho02}. The mechanism
571: for this is the classical Bondi accretion \citep{bon52} in which
572: the star, undergoing steady motion, accretes material
573: in a spherically symmetric manner from an external medium.
574: As shown by the \citet{tho02} calculations, $\omega$ Centauri
575: is not considered to be a good example of an environment likely to allow much
576: accretion despite its deep gravitational well, though this may well have been
577: different in the past especially if $\omega$ Centauri was originally
578: part of a larger system as is often proposed. This pollution scenario
579: has also been considered by \citet{tsu07}.
580:
581: The starting models for the pollution scenario investigation were
582: taken to be either the MintP or the rMS populations as shown in
583: Table \ref{parameters}. We created a set of models with
584: total fractional stellar mass accreted, composition and time delay
585: from the ZAMS, with values shown in Table 3. The accreted material
586: was of metal content $Z = 0.001$ so that the
587: polluted stellar models always maintained the surface metallicity
588: ${\rm [M/H]} = -1.27$ of the observed bMS population main
589: sequence. These models ranged from $0.40 \-- 0.90 M_{\sun}$ in
590: 0.05 $M_{\sun}$ increments. To each of these stars mass was
591: accreted as a fraction of the total starting mass of the stellar
592: model at a rate $10^{-10} M_{\sun}$ yr$^{-1}$, starting at a given
593: age of the stellar model ranging from the ZAMS point (age of 0.0)
594: up to a maximum of 9 Gyr in 3 Gyr increments. We chose our accretion
595: rate based upon the calculations of \citet{tho02} which show that
596: for typical globular cluster environments the accretion rates expected
597: are of the order $10^{-10} M_{\sun}$ yr$^{-1}$.
598:
599: In Figure \ref{fig:modelseq} we show the 13 Gyr main sequence
600: isochrones for the rMS and bMS populations listed in Table
601: \ref{parameters}. The accretion of helium-rich material will drag
602: the rMS isochrone to the left and up in the $M_{\rm bol}, T_{\rm eff}$
603: plane. We also show the isochrone from a polluted rMS population
604: scenario and how it overlaps the bMS isochrone at brighter magnitudes
605: whilst merging with the rMS isochrone at fainter magnitudes. This
606: behavior is described in more detail in Section 3.2 below.
607:
608: Only certain combinations of the amount and helium fraction
609: of the accreted material will shift isochrones to overlap the bMS.
610: In Figure \ref{fig:deltat} we plot temperature shift produced by
611: various amounts of accreted mass. The several curves display the
612: shift as a function of the helium fraction ($Y$) of the accreted
613: material. The horizontal axis is defined as the temperature
614: difference between the bMS isochrone and that of the accreted
615: population after a certain mass was added. It makes little
616: difference if either rMS or MintP stars are the ones being
617: polluted, since they start with similar masses and temperatures.
618: The only allowed models are those that lie along $\Delta T = 0$.
619: Only a very limited combination of parameters can provide the
620: necessary shift of the isochrones required to produce the bMS
621: population. The amount required is insensitive to the time at
622: which the accretion takes place: later times increased
623: $\Delta T$ by only $\sim 26{\rm K}$ per 3 Gyr time increment.
624: No amount of material with less than 45\% helium would be adequate
625: based upon extrapolation (we did not explore accretion with greater than 20\% of the
626: original starting mass). If the pollution occurred more recently
627: than the ZAMS point, in the last few Gyr, then pollution with
628: $\sim 35\%$ helium would work.
629:
630: In Table 4 we show the two of our pollution scenarios that
631: produced the bMS population 13 Gyr isochrone, along with the TRGB
632: mass and helium core mass of these models. In both of these
633: scenarios the stars being polluted were rMS models. Results for
634: the pollution of MintP models were very similar, requiring only slightly
635: more mass in accreted material. The effects of
636: polluting the stars of the rMS with the parameters do not
637: appreciably affect the mass of the stars at the TRGB and that the
638: mass of the helium core at the TRGB hardly changes. A ZAHB
639: population generated with these stars as progenitors should be
640: very similar to one created by regular rMS stars or MintP stars of
641: normal helium content.
642:
643: In Figure 9 we show the ZAHBs produced by polluted rMS and
644: polluted MintP stars as well as those of the rMS, MintP and bMS
645: stars. We also show, as thicker lines, the mass loss we inferred for
646: the rMS stars from the CMD of $\omega$ Centauri, the same mass
647: loss applied to the MintP and the polluted rMS and polluted MintP
648: models as well as the equivalent mass loss of the bMS stars. The
649: polluted-rMS HB is very similar to that of the rMS HB itself with
650: the stars populating a similar range of color and only some 0.25
651: magnitudes more luminous (vs.\ 0.8 magnitudes for helium-rich
652: homogeneous bMS stars) due to a combination of the slight
653: metallicity enhancement and helium enhancement due to the
654: pollution. A similar result is seen for the MintP and polluted
655: MintP HB though with the stars being shifted in color to the red
656: compared to the rMS stars by some 0.3 magnitudes in \bv. We find
657: that these polluted stellar models when ascending the giant branches
658: develop deep outer convective zones which result in the material accreted
659: earlier being mixed with the original stellar material and thus
660: being diluted by a factor of $\sim 5$. Therefore, the metallicity
661: and helium content at the TRGB will be close to that of the
662: original unpolluted star as seen in Table 4, where we list the ZAHB
663: parameters for the polluted scenarios which can be contrasted with
664: the parameters for the parent stars listed in Table 1. Thus the
665: polluted star descendants will arrive on the ZAHB in a similar
666: location on the CMD for a given mass loss, to be contrasted with
667: the results derived earlier for bMS stars born with a homogeneous
668: high level of helium.
669:
670: Using the data in Table 4, we calculated the amount of helium
671: required to pollute the rMS stars to the required level as
672: compared to the amount of helium these stars would contain if
673: their helium levels were homogeneous as for the bMS stars. Taking
674: a typical $0.6 M_{\sun}$ main sequence star, we see that the mass
675: of helium in the accreting material varies from $0.022 \-- 0.034
676: M_{\sun}$ depending upon the helium content of the polluting
677: material. Now this same star if fully made up of helium rich
678: material ($Y = 0.38$) would contain some $0.228 M_{\sun}$ of helium.
679: We find that the polluted stars require approximately
680: $0.1 - 0.15$ times the amount of helium compared to the
681: homogeneous high helium content stars.
682:
683: We calculated the amount of helium that can be produced from the
684: rMS stars that have already ended their lives, assuming such a
685: population directly produced the observed bMS population. We assumed
686: a total stellar mass of $5 \times 10^{6}$ solar masses for
687: $\omega$ Centauri today. We took yield calculations from the
688: literature \citep{por98,van97,ven05}. Using a variety of IMF's
689: including the Salpeter IMF \citep{sal55}, the
690: Miller-Scalo IMF \citep{mil79}, the Kroupa-Tout-Gilmore IMF \citep{kro93},
691: the Scalo IMF \citep{sca98} and the Kroupa IMF \citep{kro01} we
692: calculated the mass of helium produced and ejected back into
693: the cluster environment assuming none was lost from the cluster.
694: We took account of the evolutionary lifetimes of the rMS and
695: bMS stars as represented by the different RGB tip masses at our
696: assumed age of $\omega$ Centauri of 13 Gyrs. This allowed us to calculate
697: the mass of the original rMS and bMS populations at their formation from
698: the current observed number fractions of these populations. We also calculated
699: the mass of helium required to create the bMS as a primordial helium rich
700: population. Our results are shown in Table 5. We note that \citet{tsu07}
701: perform a similar calculation for the Salpeter IMF and show results
702: similar to ours.
703:
704: We find that, depending upon the IMF chosen, that no more than 25$\%$
705: of the helium required to make a helium rich bMS population can be
706: produced from the intermediate mass AGB stars ($3 - 8$ solar masses)
707: of the rMS population. For the Kroupa-Tout-Gilmore IMF this number
708: is about 1$\%$. Even if all the stars above the current RGB tip
709: mass are included in the calculation the mass of helium is still
710: only from $20\% - 75\%$ of that required and this material would
711: be of a far lower helium fraction than that incorporated in the
712: proposed levels for the bMS stars. We thus find it implausible that
713: the bMS population can be born from the helium rich AGB ejecta of the
714: rMS population.
715:
716: We note that the pollution scenarios require
717: only some $10\% - 15\%$ of helium compared to the homogeneous
718: bMS stars. Such helium amounts are within the bounds of all
719: but the aforementioned Kroupa-Tout-Gilmore IMF choice.
720: The calculations above do not vary appreciably for alternative
721: yields in the literature for helium production. \citet{kar06}
722: point out that another problem for the rMS AGB stars to have
723: formed the bMS stars is that for AGB helium yields of the
724: necessary magnitude ($Y > 0.3$) the total C+N+O levels would be
725: increased enormously. This is inconsistent with the observed
726: constant values of the C+N+O abundances.
727:
728: \subsection{Observational consequences of pollution models}
729:
730: An interesting difference between the polluted model
731: isochrone and our theoretical bMS isochrone is shown
732: in Figure \ref{fig:modelseq}. Both sets of models
733: exhibit a gradual merging with the rMS isochrone but
734: at a significantly different bolometric luminosity. It is seen
735: in \citet{vil07} that the observed bMS and rMS sequences appear
736: to merge at magnitude of $R \sim 21.3$. We find with our models
737: that this merging corresponds to a $M_{\rm bol} \sim 7.69$
738: for bMS stars with a mass of $0.410M_{\sun}$ and
739: $M_{\rm bol} \sim 7.65$ for rMS stars with a mass of $0.495M_{\sun}$.
740: However, we find that our bMS and rMS isochrones do not merge until
741: $M_{\rm bol} \sim 8.8$ some 1.1 magnitudes fainter than is observed. In
742: contrast our pollution scenario 2 from Table 4 merges with the rMS
743: isochrone at $M_{\rm bol} \sim 7.5$. We note that the
744: luminosity at which the merging takes place with stellar
745: models of $0.45M_{\sun}$ or less is sensitive primarily to the
746: choice of the equation of state employed in the stellar model
747: calculations. \citet{sol07} showed theoretical models calculated
748: using the evolution code of \citet{str97} using a different equation
749: of state than ours and that their bMS and rMS sequences merged close
750: to the observed merge point. We also tested our models with the old Yale
751: equation of state and found that the merging did not occur until
752: $M_{\rm bol} \sim 10$.
753:
754:
755: Using our theoretical models we calculated the expected
756: $(B-R)/(B+V+R)$ statistic for a composite population of the bMS
757: ($30\%$), rMS ($65\%$) and the known metal rich population (${\rm
758: [M/H]} \sim -0.6$) numbering some $5\%$ of the stars. We
759: calculated this statistic for three mean mass losses for the rMS
760: population and for bMS stars coeval, 2 Gyr and 4 Gyr younger than
761: the majority rMS population. We repeated this for polluted rMS
762: models representing the bMS stars fraction of the composite
763: population. The results of this are plotted in Figure 10 where the
764: labeled cases (A thru F) are described in Table 5.
765: Using the observed CMD, we calculated the $\omega$ Centauri
766: $(B-R)/(B+V+R)$ statistic to be $0.85 \pm 0.01$.
767: We predict various age differences for the rMS and bMS populations
768: that both depend on the helium content and mass loss. These age
769: differences are shown in Table 6. We can see immediately that for
770: a given mass loss and $(B-R)/(B+V+R)$ statistic value that the age
771: differences for the composite population containing bMS stars
772: implies a much larger age difference than if the bMS stars are
773: replaced by polluted rMS stars.
774:
775: The constraint on the age difference between the various
776: populations of $\omega$ Centauri has been previously investigated.
777: Using the morphology of the turn-off region \citet{sol05} put a
778: maximum on the age difference between the rMS and bMS stars
779: (irrespective of helium enhancement) of $\sim$ 1.5 Gyr. In their
780: main sequence fitting of model isochrones \citet{lee05} determined
781: that the age difference between the rMS and bMS stars was $\sim$ 1
782: Gyr. We can see from the data in Table 5 that our models predict
783: that only mean mass losses in the 0.2M$_{\sun}$ range with
784: polluted rMS stars reproduce a $(B-R)/(B+V+R)$ statistic in line
785: with observation. All the scenarios with the high helium bMS stars
786: imply age differences that are much too large. Furthermore, we
787: earlier calculated the mass loss range from the $\omega$ Centauri
788: CMD using the prominent HB clump as representing the majority rMS
789: population and obtained mass losses in the range $0.168
790: - 0.247M_{\sun}$, such mass losses are within or close to the allowable
791: range based upon figure 10.
792:
793:
794:
795: \section{Summary}
796:
797: It has been hypothesized that the double main sequence of $\omega$ Centauri reveals
798: an intermediate metal-poor population that have a large enhancement of helium, contrary to expectation of standard stellar theory. This hypothesis seemed
799: even more credible after spectroscopic analysis confirmed that the blue main sequence stars were indeed more metal-rich than the majority red main sequence. The proposed helium levels in these stars
800: are extremely large compared to known stellar populations, imply an extremely large and preferential enrichment history for their formation whilst coexisting with, or forming from,
801: a majority stellar population that is not unusual in any noticeable way.
802:
803: We confirmed using our latest YREC stellar models that the proposed helium rich bMS stars do lie to the blue side of the more metal-poor rMS stars.
804: We developed a new procedure to allow mass loss comparison between stars of different
805: TRGB masses so as to predict their HB morphology and compare to the observations.
806: Under the assumption that the large clump of the $\omega$ Centauri HB is a product of the
807: majority rMS metal-poor population, we inferred that the mass loss suffered by these stars on the RGB spans a range of $0.168 - 0.247M_{\sun}$ from TRGB progenitors
808: of $0.814M_{\sun}$ at an age of 13 Gyrs. This in turn implied that the proposed high helium bMS stars would have lost a minimum of $0.102 - 0.147M_{\sun}$ from a TRGB progenitor
809: of some $0.629M_{\sun}$ at the same age. We find such helium rich stars would appear on the HB in the blue tail below the majority clump that is assumed to be from the rMS stars. If we assume all
810: such stars on the CMD of the $\omega$ Centauri HB are attributable to the bMS population then we have only $\sim 25\%$ of the bMS plus rMS stars residing there. This seems distinctly
811: at odds with the $\sim 37\%$ of bMS plus rMS star total we observe in the MS itself. We also note that if we use our Case II or Case III mass losses for these stars the HB bMS stars should
812: appear far down the blue tail in a large clump clearly separated from the observed majority clump we identify with the rMS stars. Such a disconnected clump at the bottom of the
813: blue tail of the $\omega$ Centauri HB is not seen at all.
814:
815: Unless the mass loss mechanism for the proposed high helium bMS stars operates differently from the rMS stars causing them the reside in the large observed
816: clump then these stars are absent from the HB of $\omega$ Centauri. In fact, for the bMS stars to be hiding amongst the majority HB clump then their mass loss must be
817: in the range $0.059 - 0.101M_{\sun}$, where the lower limit is determined by the fact we do not see overluminous HB stars in substantial numbers above the main HB clump.
818: Prior observations of the instability strip RR Lyrae variables have not found evidence of high helium stars
819: there, implying that the mass loss of these stars on the RGB was not far lower than what we have assumed. We noted earlier that if the bMS stars were $2 - 5$ Gyr younger they would
820: also appear in the instability strip of $\omega$ Centauri but such age differences have been ruled out by turn-off morphology in other studies.
821:
822: It is possible to create a blue main sequence theoretically by the pollution of existing rMS or MintP stars with helium rich material for limited combinations
823: of helium content and mass accreted. The helium content of such material is not far from the range of the latest intermediate mass AGB stellar yields \citep{her04}, especially
824: if the accretion is relatively recent. The total mass of the accreted material is in line with estimates of the
825: possible pollution in a globular cluster environment \citep{tho02}. We do note that the AGB helium yields we require are on the upper end (or slightly beyond)
826: of such calculations and that the mass accreted to create our blue main sequences is incompatible with the current $\omega$ Centauri environment, though possibly not that of past
827: when the cluster was not only more massive but the velocity dispersion lower; both of which would increase the amount of accretion possible onto existing stars.
828: Creating bMS stars out of homogeneous helium rich material is not the only possible mechanism for their formation, and pollution via accretion
829: of AGB ejecta in a globular cluster environment is another possibility. We see that of the observed $\omega$ Centauri RR Lyrae variables the presence
830: of both MintP stars and rMS stars but no helium-rich bMS stars, this implies that at least some MintP stars formed with normal helium enrichments. An accretion pollution
831: scenario has also been hypothesized in the past by \citet{can98} in regard to the main sequence abundance anomalies in the globular cluster 47 Tuc, where the C and N
832: abundance variations are seen on the main sequence and in red giants. We note our simple pollution scenario does not explain the abundance variations seen on both
833: the main sequence and giant branches, as they would be washed out as the stars evolve up the RGB.
834:
835: The polluted star scenario predicts that the ZAHB for such a population should not differ appreciably from that of the unpolluted parent stars. We see from this that
836: we would observe in the $\omega$ Centauri CMD a double main sequence and a HB that possesses a majority single clump, very few over luminous stars in the horizontal region
837: of the HB, no large disconnected clump of stars in the HB blue tail and no RR Lyrae variables that are over luminous. This is what is observed in contrast to the predictions of the homogeneously
838: helium rich bMS stars that have been proposed. The polluted star scenario though requiring certain combinations of pollution helium content and total mass accreted does not
839: require the very large helium amounts that the proposed bMS stars do, amounts that seem greater than that possible from ejecta of more massive rMS stars in $\omega$ Centauri.
840:
841: Another prediction from our modeling is that the polluted stars will merge with the rMS stars at approximately the observed luminosity whereas our model homogeneous bMS stars do not
842: do this until about 1.1 magnitudes fainter. The position of the this merger may turn out to be an important diagnostic in determining the composition of the bMS stars.
843:
844: We used the $(B-R)/(B+V+R)$ statistic that is usually applied to the investigation of single stellar population horizontal branches in globular clusters. When we apply
845: our calculated $(B-R)/(B+V+R)$ of $0.85 \pm 0.01$ for $\omega$ Centauri to theoretical predictions of this statistic using our models we found that the high helium bMS star models
846: predict age differences for the bMS population that are too great when compared to turn-off morphology estimates. The pollution scenario for mass losses less than $\sim 0.2M_{\sun}$ are
847: within the allowable limit of age differences. A result that is also approximately the range of estimated mass losses for the rMS population ($0.168 - 0.247M_{\sun}$)
848: we inferred from the $\omega$ Centauri CMD using our theoretical models.
849:
850: %%High helium bMS star progeny seem to be either absent from the HB region of $\omega$ Centauri
851: %%or not in the required numbers unless the mass loss history of their progenitors was quite contrived (very low) to allow them to hide in the majority HB clump with the rMS population HB stars.
852: %%Whatever mass loss a high helium bMS stars would undergo, it is constrained on the high side by the lack of a large clump in the blue tail off the HB and is constrained on the low side
853: %%by their absence in the horizontal regions of the HB by appearing overluminous or as RR Lyrae variables. We also conclude that the pollution via
854: %%accretion of helium rich material is also capable of generating an observed bMS. Such stars
855: %%would appear approximately in the same places on the HB as their unpolluted counterparts, in contrast to high helium bMS stars as proposed.
856: %%Using the $(B-R)/(B+V+R)$ statistic we conclude that the high helium bMS stars predict that their age difference from the rMS stars
857: %%is too great except for very small mass losses whereas the pollution scenario presents no such conflict.
858:
859: \acknowledgements
860: We would like to thank M. Pinsonneault and D. Weinberg for helpful suggestions and A. Sollima for kindly furnishing to us the HST data. We
861: would also like to thank the anonymous referee whose many and detailed comments have greatly improved this paper.
862:
863:
864:
865:
866:
867: \begin{thebibliography}
868:
869: \bibitem[Alexander \& Ferguson(1994)]{ale94} Alexander, D. R.,
870: \& Ferguson, J. W., 1994, \apj, 437, 879
871:
872: \bibitem[Allard \& Hauschildt(1995)]{all95} Allard, F., \&
873: Hauschildt, P. H. 1995, \apj, 445, 433
874:
875: \bibitem[An et al.(2007)]{an07} An, D., Terndrup, D. M.,
876: Pinsonneault, M. H., Paulson, D. B., Hanson, R. B., \& Stauffer,
877: J. R., 2007, \apj, 655, 233
878:
879: \bibitem[Bahcall \& Loeb(1990)]{bah90} Bahcall, J. N.,
880: \& Loeb, A., 1990, \apj, 360, 267
881:
882: \bibitem[Bahcall et al.(2001)]{bah01} Bahcall, J. N.,
883: Pinsonneault, M. H., \& Basu, S., 2001, \apj, 555, 990
884:
885: \bibitem[Bedin et al.(2004)]{bed04} Bedin L., Piotto G.,
886: Anderson J., Cassisi S., King I., Momany Y., \& Carraro G., 2004,
887: \apj, 605, L125
888:
889: \bibitem[Bekki \& Norris(2006)]{bek06} Bekki, K., \& Norris,
890: J., 2006, \apj, 637L, 109
891:
892: \bibitem[Bergbusch \& Vandenberg(1992)]{ber92} Bergbusch, P. A.,
893: \& Vandenberg D. A., 1992, \apjs, 81, 163
894:
895: \bibitem[B\"{o}hm-Vitense(1958)]{boh58} B\"{o}hm-Vitense, E. 1958,
896: Z. Astrophys., 46, 108
897:
898: \bibitem[Bondi(1952)]{bon52} Bondi, H., 1952, MNRAS, 112, 195
899:
900: \bibitem[Bono et al.(2002)]{bono02} Bono, G., Balbi, A.,
901: Cassisi, S., Vittorio, N., \& Buonanno, R. 2002, \apj, 568, 463
902:
903: \bibitem [Bono et al.(1995)]{bon95} Bono, G., Caputo, F.,
904: \& Marconi, M. 1995, \aj, 110, 2365
905:
906: \bibitem[Cannon et al. (1998)]{can98} Cannon, R. D., Croke,
907: B. F. W., Bell, R. A., Hesser, J. E., \& Stathakis, R. A., 1998,
908: \mnras, 298, 601
909:
910: \bibitem[Catelan(2000)]{cat00} Catelan, M., 2000, \apj, 531, 826
911:
912: \bibitem[Choi \& Yi(2007)]{cho07} Choi, E., \& Yi, S. K., MNRAS, 375, L1
913:
914: \bibitem[Cox et al.(1968)]{cox68} Cox, J. P., \& Guili,
915: R. T., 1968, Principles of Stellar Structure (New York: Gordon and
916: Breach)
917:
918: \bibitem[D'Antona et al.(2002)]{dan02} D'Antona, F.,
919: Caloi, V., Montalban, J., Ventura, P., \& Gratton, R., 2002, \aap,
920: 395, 69
921:
922: \bibitem[D'Antona et al.(2005)]{dan05} D'Antona, F.,
923: Bellazzini, M., Caloi, V., Fusi-Pecci, F., Galleti, S., \& Rood,
924: R. T., 2005, \apj, 631, 868
925:
926: \bibitem[D'Cruz et al.(2000)]{dcr00} D'Cruz et al., 2000,
927: \apj, 530, 352
928:
929: \bibitem[Delahaye \& Pinsonneault(2005)]{del05} Delahaye,
930: F., \& Pinsonneault, M., 2005, \apj, 625, 563
931:
932: \bibitem[Dickens \& Wooley(1967)]{dic67} Dickens, R. J.,
933: \& Wooley S. J., 1967, Royal Obs. Bull., 128, E255
934:
935: \bibitem[Ferraro et al.(2004)]{fer04} Ferraro, F. R., Sollima, A.,
936: Pancino, E., Bellazzini, M., Origlia, L., Straniero, O., \& Cool, A.,
937: 2004, \apj, 603, L81
938:
939: \bibitem[Freeman \& Rodgers(1975)]{fre75} Freeman, K. C.,
940: \& Rodgers, A. W., 1975, \apj, 201, L71
941:
942: \bibitem[Fukugita \& Kawasaki(2006)]{fk06} Fukugita, M., \&
943: Kawasaki, M. 2006, \apj, 646, 691
944:
945: \bibitem[Green et al.(1987)]{gre87} Green, E. M., Demarque, P.,
946: \& King, C. R., 1987, the Revised Yale Isochrones and Luminosity
947: Functions (New Haven: Yale Univ. Obs.)
948:
949: \bibitem[Grevesse \& Noels(1993)]{gre93} Grevesse, N., \& Noels,
950: A., 1993, in Origin and Evolution of the Elements, ed. N.
951: Prantzos, E. Vangioni-Flam, \& M. Cassé (Cambridge: Cambridge
952: Univ. Press), 15
953:
954: \bibitem[Gruzinov \& Bahcall(1998)]{gru98} Gruzinov, A., \&
955: Bahcall, J., 1998, \apj, 504, 996
956:
957: \bibitem[Guenther et al.(1992)]{gue92} Guenther, D. B.,
958: Demarque, P., Kim, Y.-C., \& Pinsonneault, M. H., 1992, \apj, 387,
959: 372
960:
961: \bibitem[Herwig (2004)]{her04} Herwig, F., 2004, \apjs, 155, 651
962:
963: \bibitem[Iglesias \& Rogers(1996)]{igl96} Iglesias, C. A.,
964: \& Rogers, F. J., 1996, \apj, 464, 943
965:
966: \bibitem[Itoh et al.(1996)]{ito96} Itoh, N., Hayashi, H., \&
967: Nishikawa, A., 1996, \apjs, 102, 411
968:
969: \bibitem[Jimenez et al.(2003)]{jim03} Jimenez, R., Flynn, C.,
970: MacDonald, J., \& Gibson, B. K., 2003, Science, 299, 1552
971:
972: \bibitem[Karakas et al.(2006)]{kar06} Karakas, A., Fenner, Y.,
973: Sills, A., Campbell, S. W., \& Lattanzio, J. C., 2006, preprint
974: (astro-ph/0608366)
975:
976: \bibitem[Kroupa et al.(1993)]{kro93} Kroupa, P., Tout, C. A., \&
977: Gilmore, G., 1993, MNRAS, 262, 545
978:
979: \bibitem[Kroupa (2001)]{kro01} Kroupa, P., 2001, MNRAS, 322, 231
980:
981: \bibitem[Lee et al.(1994)]{lee94} Lee Y.-W., Demarque
982: P., \& Zinn R., 1994, \apj, 423, 248
983:
984: \bibitem[Lee et al.(2005)]{lee05} Lee Y.-W., Joo S-J.,
985: Han S-I., Chung C., Ree C., Sohn Y-J., Kim Y-C., Yoon S-J., Yi S.
986: \& Demarque P., 2005, \apj 621, L57
987:
988: \bibitem[Maeder \& Meynet(2006)]{mae06} Maeder, A., \&
989: Meynet, G., 2006, \aap, 448, L37
990:
991: \bibitem[Miller \& Scalo(1979)]{mil79} Miller G. E., \&
992: Scalo, J. M., 1979, \apjs, 41, 513
993:
994: \bibitem[Momany et al.(2004)]{mom04} Momany, Y., Bedin, L. R.,
995: Cassisi, S., Piotto, G., Ortolani, S., Recio Blanco, A., De
996: Angeli, F., \& Castelli, F., 2004, \aap, 420, 605
997:
998: \bibitem[Norris(2004)]{nor04} Norris, J. E., 2004,
999: \apj, 612, L25
1000:
1001: \bibitem[Norris \& Da Costa(1995)]{nor95} Norris, J.,
1002: \& De Costa G. S., 1995, \apj, 447, 680
1003:
1004: \bibitem[Norris et al.(1996)]{nor96} Norris, J. E.,
1005: Freeman, K. C., \& Mighell, K. J., 1996, \apj, 462, 241
1006:
1007: \bibitem[Olive \& Skillman(2004)]{oli04} Olive, K. A., \&
1008: Skillman, E. D. 2004, \apj, 617, 29
1009:
1010: \bibitem[Origlia et al.(2002)]{ori02} Origlia, L., Ferraro,
1011: F. R., Fusi Pecci, F., \& Rood, R. T., 2002, \apj, 571, 458
1012:
1013: \bibitem[Pagel(1992)]{pag92} Pagel, B. E. J., 1992, IAU
1014: Symp. 149, Stellar Populations in Galaxies, ed. B. Barbury, \& A.
1015: Renzini, (Dordrecht: Kluwer), 133
1016:
1017: \bibitem[Pancino et al.(2000)]{pan00} Pancino E., Ferraro F.,
1018: Bellazzini M., Piotto G., \& Zoccali M., 2000, \apj, 534, L83
1019:
1020: \bibitem[Pinsonneault et al.(2003)]{pin03} Pinsonneault,
1021: M. H., Terndrup, D. M., Hanson, R. B., \& Stauffer, J. R., 2003,
1022: \apj, 598, 588
1023:
1024: \bibitem[Piotto et al.(2005)]{pio05} Piotto et al., 2005,
1025: \apj, 621, 777
1026:
1027: \bibitem[Portinari et al.(1998)]{por98} Portinari, L.,
1028: Chiosi, C., \& Bressan, A., 1998, \aap, 334, 505
1029:
1030: \bibitem[Rey et al.(2004)]{rey04} Rey, S.-C., Lee, Y.-W.,
1031: Ree, C.-H., Joo, J.-M., \& Sohn Y.-J., 2004, \aj, 127, 958
1032:
1033: \bibitem[Rogers et al.(1996)]{rog96} Rogers, F. J., Swenson,
1034: F. J., \& Iglesias, C. A., 1996, \apj, 456, 902
1035:
1036: \bibitem[Salpeter(1955)]{sal55} Salpeter, E., 1955, \apj, 121, 161
1037:
1038: \bibitem [Scalo(1998)]{sca98} Scalo, J., 1998, in ASP Conf. Ser., 142, The
1039: Stellar Initial Mass Function, 201
1040:
1041: \bibitem[Saumon et al.(1995)]{sau95} Saumon, D., Chabrier, G.,
1042: \& Van Horn, H. M., 1995, \apjs, 99, 713
1043:
1044: \bibitem[Serenelli \& Weiss(2005)]{ser05} Serenelli, A., \& Weiss,
1045: A. 2005, \aap, 442, 1041
1046:
1047: \bibitem[Sills et al.(2000)]{sil00} Sills, A., Pinsonneault,
1048: M. H., \& Terndrup, D. M., 2000, \apj, 534, 335
1049:
1050: \bibitem [Sollima et al.(2007)]{sol07} Sollima, A., Ferraro, F. R.,
1051: Bellazzini, M., Origlia, L., Straniero, O., \& Pancino, E., 2007, ApJ, 654, 915
1052:
1053: \bibitem[Sollima et al.(2006)]{sol06} Sollima, A., Borrisova,
1054: J., Catelan, M., Smith, H. A., Minniti, D., Cacciari, C., \&
1055: Ferraro, F. R., 2006, \apj, 640, L43
1056:
1057: \bibitem[Sollima et al.(2005)]{sol05} Sollima, A., Pancino,
1058: E., Ferraro, F., Bellazzini, M., Straneiro, O., \& Pasquini, L.,
1059: 2005, \apj, 634, 332
1060:
1061: \bibitem [Straniero et al.(1997)]{str97} Straniero, O.,
1062: Chieffi, A. \& Limongi, M., 1997, ApJ, 490, 425
1063:
1064: \bibitem[Suntzeff \& Kraft(1996)]{sun96} Suntzeff, N. B.,
1065: \& Kraft, R. P., 1996, \aj, 111, 1913
1066:
1067: \bibitem[Thoul et al.(1994)]{tho94} Thoul, A. A., Bahcall,
1068: J. N., \& Loeb, A., 1994, \apj, 421, 828
1069:
1070: \bibitem[Thoul et al.(2002)]{tho02} Thoul, A., Jorissen, A.,
1071: Goriely, S., Jehin, E., Magain, P., Noels, A., \& Parmantier, G.,
1072: 2002, \aap, 383, 491
1073:
1074: \bibitem[Thuan \& Izotov(2002)]{thuan02} Thuan, T. X., \&
1075: Izotov, Y. I. 2002, Space Sci.\ Rev., 100, 263
1076:
1077: \bibitem[Tsujimoto et al.(2007)]{tsu07} Tsujimoto, T., Shigeyama, T.,
1078: \& Suda, T., 2007, \apj, 654, L139
1079:
1080: \bibitem[Vandenberg \& Clem(2003)]{van03} VandenBerg, D. A.,
1081: \& Clem, J. L., 2003, \aj, 126, 778
1082:
1083: \bibitem[Van den Hoek \& Groenewegen(1997)]{van97} Van den
1084: Hoek, L. B., \& Groenewegen, M. A. T., 1997, \aaps, 123, 305
1085:
1086: \bibitem[Ventura \& D'Antona(2005)]{ven05} Ventura, P., \&
1087: D'Antona, F., 2005, \aap, 439, 1075
1088:
1089: \bibitem [Villanova et al.(2007)]{vil07} Villanova, S., Piotto, G.,
1090: King, I. R., Anderson, J., Bedin, L. R., Gratton, R. G., Cassisi, S.,
1091: Momany, Y., Bellini, A., Cool, A. M., Recio-Blanco, A. \& Renzini, A.,
1092: 2007, preprint (astro-ph/0703208)
1093:
1094: \bibitem[Weiss et al.(2006)]{wei06} Weiss, A., Salaris, M., Ferguson, J. W.,
1095: \& Alexander, D. R., 2006, preprint (astro-ph/0605666)
1096:
1097: \bibitem[Yi \& Demarque(2003)]{yi03} Yi, S., Kim, Y.-C.,
1098: \& Demarque P., 2003, \apjs, 144, 259
1099:
1100: \end{thebibliography}
1101:
1102: \clearpage
1103:
1104: \input{tab1}
1105:
1106: \clearpage
1107:
1108: \input{tab2}
1109:
1110: \clearpage
1111:
1112: \input{tab3}
1113:
1114: \clearpage
1115:
1116: \input{tab4}
1117:
1118: \clearpage
1119:
1120: \input{tab5}
1121:
1122: \clearpage
1123:
1124: \input{tab6}
1125:
1126: \clearpage
1127:
1128: \begin{figure}
1129: \epsscale{0.85}
1130: \plotone{f1.eps}
1131: \caption{The rMS and bMS 13 Gyr isochrones using the parameters in
1132: Table \ref{parameters}.\label{fig:isochrones}}
1133: \end{figure}
1134:
1135: \clearpage
1136:
1137: \begin{figure}
1138: \epsscale{0.85} \plotone{f2.eps} \caption{Stratification of the
1139: rMS and bMS models near the TRGB. The plot shows the
1140: gravitational acceleration as a function of stellar mass below the
1141: surface. \label{fig:deltam}}
1142: \end{figure}
1143: \clearpage
1144:
1145: \begin{figure}
1146: \epsscale{0.85} \plotone{f3.eps} \caption{rMS (open triangles) and
1147: bMS (filled circles) 13 Gyr ZAHBs populated for a rMS mean mass
1148: loss of 0.24$M_{\sun}$ and a bMS mean mass loss of
1149: 0.143$M_{\sun}$ using Case I mass losses.\label{fig:hbmodels}}
1150: \end{figure}
1151: \clearpage
1152:
1153: \begin{figure}
1154: \epsscale{0.85} \plotone{f4.eps} \caption{The rMS 13 Gyr ZAHBs
1155: populated for rMS mean mass losses of 0.20$M_{\sun}$ (open
1156: circles) and 0.28 $M_{\sun}$ (open triangles) and the bMS 13 Gyr
1157: population populated for bMS mean mass losses of 0.118 $M_{\sun}$
1158: (filled circles) and 0.168 $M_{\sun}$ (filled triangles) and the
1159: bMS 7 Gyr population with mean mass loss 0.143$M_{\sun}$ (filled
1160: stars). All bMS mass losses are using Case I.\label{fig:altdeltam}}
1161: \end{figure}
1162: \clearpage
1163:
1164: \begin{figure}
1165: \epsscale{0.85} \plotone{f5.eps} \caption{CMD of $\omega$ Centauri
1166: with 13 Gyr rMS and bMS isochrones. Photometry is from
1167: \citet{rey04}.\label{fig:cmd}}
1168: \end{figure}
1169: \clearpage
1170:
1171: \begin{figure}
1172: \epsscale{0.85} \plotone{f6.eps} \caption{Horizontal branch CMD of
1173: $\omega$ Centauri showing the RR Lyrae instability strip for the
1174: rMS population.\label{fig:rrstrip}}
1175: \end{figure}
1176: \clearpage
1177:
1178: \begin{figure}
1179: \epsscale{0.85} \plotone{f7.eps} \caption{Model main sequences of
1180: the rMS and bMS populations and a polluted rMS sequence.\label{fig:modelseq}}
1181: \end{figure}
1182: \clearpage
1183:
1184: \begin{figure}
1185: \epsscale{0.85} \plotone{f8.eps} \caption{Mass fraction of pollution to
1186: total stellar mass vs.\ change in $T_{\rm eff}$ ($\Delta$T) for rMS
1187: stellar models.\label{fig:deltat}}
1188: \end{figure}
1189: \clearpage
1190:
1191: \begin{figure}
1192: \epsscale{0.85} \plotone{f9.eps} \caption{Standard and pollute
1193: population ZAHB's and equivalent mass losses (Case I) based upon the
1194: inferred rMS mass loss of 0.168$M_{\sun} - 0.247M_{\sun}$}
1195: \end{figure}
1196: \clearpage
1197:
1198: \begin{figure}
1199: \epsscale{0.85} \plotone{f10.eps} \caption{$(B-R)/(B+V+R)$
1200: statistic vs. age difference of a composite $\omega$ Centauri
1201: population. See Table 6 for the case (A \-- F) details.}
1202: \end{figure}
1203:
1204:
1205: \end{document}
1206: