«Submitted to The Astronomical Journal ABSTRACT We present a chemical composition analysis of 36 giants in the nearby mildy metal-poor ([Fe/H] = ...»
Star-to-Star Abundance Variations among Bright Giants in the Mildly
Metal-Poor Globular Cluster M4
Inese I. Ivans1, Christopher Sneden1, Robert P. Kraft2, Nicholas B. Suntzeﬀ3, Verne V. Smith4,5,
G. Edward Langer6,7, Jon P. Fulbright2
Submitted to The Astronomical Journal
We present a chemical composition analysis of 36 giants in the nearby mildy
metal-poor ([Fe/H] = –1.18) “CN-bimodal” globular cluster M4. The stars were observed at the Lick & McDonald Observatories using high resolution ´chelle e spectrographs and at CTIO using the multi-object spectrometer. Confronted with a cluster having interstellar extinction that is large and variable across the cluster face, we combined traditional spectroscopic abundance methods with modiﬁcations to the line-depth ratio technique pioneered by Gray (1994) to determine the atmospheric parameters of our stars. We derive a total-to-selective extinction ratio of 3.4 ± 0.4 and an average E(B–V) reddening of 0.33 ± 0.01 which is signiﬁcantly lower than that estimated by using the dust maps made by Schlegel et al. (1998).
We determine abundance ratios typical of halo ﬁeld and cluster stars for scandium, titanium, vanadium, nickel, and europium with star-to-star variations in these elements of ±0.1. Silicon, aluminum, barium, and lanthanum are overabundant with respect to what is seen in other globular clusters of similar metallicity. These overabundances 1 Department of Astronomy and McDonald Observatory, University of Texas, Austin, TX 78712;
email@example.com, firstname.lastname@example.org 2 UCO/Lick Observatory, Board of Studies in Astronomy and Astrophysics, University of California, Santa Cruz, CA 95064; email@example.com 3 Cerro Tololo Inter-American Observatory, National Optical Astronomy Observatories, which is operated by the Association of Universities for Research in Astronomy, Inc., (AURA), under cooperative agreement with the National Science Foundation. La Serena, Chile; nsuntzeﬀ@noao.edu 4 Department of Physics, University of Texas at El Paso, 500 West University, El Paso, TX 79968-0515;
firstname.lastname@example.org 5 Visiting Astronomer, Cerro Tololo Inter-American Observatory, National Optical Astronomy Observatories, which is operated by the Association of Universities for Research in Astronomy, Inc., (AURA), under cooperative agreement with the National Science Foundation.
6 Physics Department, Colorado College, Colorado Springs, CO 80903; email@example.com 7 Deceased 1999 February 16.
–2– conﬁrm the results of an earlier study based on a much smaller sample of M4 giants (Brown & Wallerstein 1992).
Superimposed on the primordial abundance distribution is evidence for the existence of proton-capture synthesis of carbon, oxygen, neon, and magnesium. We recover some of the C, N, O, Na, Mg, and Al abundance swings and correlations found in other more metal-poor globular clusters but the range of variation is muted. In the case of Mg and Al, this is compatible with the idea that the Al enhancements are derived from the destruction of 25,26 Mg, not 24 Mg. We determine that the C+N+O abundance sum is constant to within the observational errors, and agrees with the C+N+O total that might be expected for M4 stars at birth.
The AGB stars in M4 have C,N,O abundances that show less evidence for proton-capture nucleosynthesis than is found in the less-evolved stars of the RGB.
Deeply-mixed stars of the RGB, subsequent to the helium core ﬂash, might take up residence on the blue end of the HB, and thus fail to evolve back to the AGB but reasons for skepticism concerning this scenario are noted.
Subject headings: globular clusters: individual (NGC 6121) — globular clusters:
general — stars: abundances — nucleosynthesis — stars: fundamental parameters — dust, extinction
Very large star-to-star abundance variations in the light elements C, N, O, Na, Mg and Al occur among the bright giants of a number of globular clusters. In clusters where giant star samples have been suﬃciently large, the abundances of O and Na are anticorrelated, as are those of Mg and Al. Prime examples are found in M13 (Pilachowski et al. 1996; Shetrone 1996a,b, Kraft et al. 1997), ω Cen (Paltoglou & Norris 1989; Norris & Da Costa 1995a,b), NGC 6752 (Cottrell & Da Costa 1981), M15 (Sneden et al. 1997), and most recently NGC 3201 (Gonzalez & Wallerstein 1998). Recent major reviews of cluster abundance trends have been published by Suntzeﬀ (1993), Briley et al. (1994), Kraft (1994), and Wallerstein et al. (1997).
Most investigators agree that the abundance anticorrelations arise from proton-capture chains that convert C and O into N, Ne into Na, and Mg into Al in the hydrogen-burning layers of evolved cluster stars. Controversy has arisen, however, between proponents of what are termed “evolutionary” vs. “primordial” scenarios. For a brief review of these alternatives, we refer the reader to the introductory section of Sneden et al. (1997). In the former picture one supposes that, in the low-mass red giants we presently observe, the products of internal proton-capture synthesis are brought to the surface by deep-mixing of the stellar envelope through the hydrogen-burning shell. Therefore, the anomalous behavior is expected to increase (on average) with advancing evolutionary state. In the latter scenario, one supposes that the proton-capture synthesis took –3– place in a prior generation of more massive stars. As a result of mass loss, these massive stars produced the required abundance redistribution in the primordial material out of which the presently observed low-mass stars were formed. Several lines of observational evidence now suggest that both scenarios may play important roles: superimposed on a primordial spread are additional variations resulting from deep-mixing in the stars we presently observe (e.g. Briley et al. 1994).
Stellar evolution theory (Sweigart & Mengel 1979) predicts that deep-mixing should become less eﬃcient and possibly cut oﬀ as metallicity increases. Observational support for this hypothesis was found by Sneden et al. (1994), who showed that the C, N, O and Na abundances have little variation among giants in M71 ([Fe/H] = –0.8),8 whereas Sneden et al. (1992) did ﬁnd variations in these elements in M5 ([Fe/H] = –1.2) similar to those in more metal-poor clusters such as M13, M15, and NGC 3201. However, the prediction fails to explain the diﬀerences between the CNO abundances of NGC 288 and NGC 362 (Dickens et al. 1991): these two clusters have very similar metallicities ([Fe/H] = –1.40 and –1.27, respectively; Zinn & West 1984), yet one shows evidence for deep-mixing (redistribution of C, N and O abundances) whereas the other exhibits little or no variation in these elements. However, a new study by Shetrone (1999) indicates that both clusters, in fact, show wide variations among the C, N, and O elements, and that the earlier conclusions in declaring no variations in NCG 288 resulted from a limited sample size.
Variations in the cyanogen band strengths have also been studied extensively in the giant stars of old globular clusters, as well as nearby dwarf spheroidal galaxies. CN variations have been found in NGC 362, M5, M10, NGC 6352, NGC 7006, M92, M15, M71, M22, ω Cen, NGC 3201, NGC 6752, 47 Tuc and M4 (see eg. Smith & Norris 1982 and references therein) as well as in the Sculptor (Smith & Dopita 1983) and Draco (Smith 1984) dwarf galaxies. In clusters where a range in CN strengths is observed, there appears to be a correlation with elements such as oxygen, sodium, and aluminum. Examples include 47 Tuc and NGC 6752 (Cottrell & Da Costa
1981) and M22 (Smith & Wirth 1991), clusters spanning a metallicity range that encompasses M4. And, in some of the systems (NGC 6752, M22, ω Cen, and the dwarf galaxies), variations in calcium positively correlate with variations in cyanogen band strengths. Norris & Bessell (1978) were among the ﬁrst to recognize the resemblance in chemical signatures between giants in M22, ω Cen, and the Sculptor and Ursa Minor dwarf galaxies. Extensive abundance studies of the three most interesting globular cluster systems of this kind have been done by Norris et al. (1981, NGC 6752), Twarog et al. (1995, M22) and Norris et al. (1996, ω Cen).
Given the importance of decoupling the evolutionary eﬀects from primordial enrichments, these puzzles led us to consider an abundance study of a large sample of bright giants in the mildly metal-poor globular cluster M4. M4 (NGC 6121) is possibly the nearest globular cluster (d ∼ 1.7–2.1 kpc; see Dixon & Longmore 1993 for a review). Despite the fact that M4 lies behind the outer portion of the Scorpius-Ophiucus dust cloud complex and thus suﬀers relatively high
reddening and extinction, its most luminous giants have relatively small apparent magnitudes and are readily accessible to detailed analysis. M4’s metallicity on the Zinn & West (1984) scale is [Fe/H] = –1.3. High-resolution spectroscopic estimates range from [Fe/H] = –1.3 (Brown et al.
1990; Gratton et al. 1986) through –1.2 (Brown & Wallerstein 1992) to –1.05 (Drake et al. 1992).
Liu & Janes (1990) discuss M4 reddening and metallicity issues at length, and Drake et al. (1994) provide a useful summary table of the many metallicity estimates for this cluster. They compute a mean M4 metallicity, excluding their own result, of [Fe/H] = –1.17. This metallicity is therefore near the anticipated cutoﬀ in the deep-mixing process that leads to large light element abundance variations in red giant branch (RGB) stars of other clusters.
Norris (1981) discovered CN bimodality in M4: stars of similar luminosity exhibit a largely bimodal distribution of cyanogen strengths. Norris suggested that the distribution of CN band strengths among cluster giants could be related to the color distribution of stars on the horizontal branch (HB), i.e., the so-called “second parameter” problem (e.g. Smith & Norris 1993). It is not clear, however, whether CN bimodality results from deep-mixing (variations in C→N conversion among giants) or reﬂects instead some built-in primordial diﬀerence among cluster stars. Norris argued for a range in core rotational velocities inducing variable amounts of envelope mixing among the former main sequence stars that we now see as the M4 giants.
Previous abundance studies of the brighter giants in M4 present a rather complex picture.
Brown & Wallerstein (1989) showed that C was depleted and N enhanced in four stars. They and Smith & Suntzeﬀ (1989) found that M4 giants all had very low carbon isotope ratios, nearly the equilibrium ratio of 12 C/13 C 3.5. This value is much lower than the ratio of 20 to 30 expected from the “ﬁrst dredge-up” mixing episodes of Pop I giants (Iben 1964), and suggests the existence of a mixing mechanism more eﬃcient than classical convection. Additionally, the Drake et al.
(1992, 1994) analysis of two pairs of CN-strong and CN-weak giants found that the CN-weak pair had “normal” low abundances of Na and Al (i.e., normal in comparison to ﬁeld halo giants of comparable metallicity), whereas the CN-strong pair had signiﬁcantly enhanced abundances of Na and Al. The CN-strong pair also had slightly (∼ 0.2 dex) lower [O/Fe] ratios than the CN-weak pair. These results are compatible with the deep-mixing hypothesis described above (Langer et al.
1993, Langer & Hoﬀman 1995, Langer et al. 1997).
On the other hand, Brown et al. (1990) found that M4 RGB stars have a relatively small (possibly insigniﬁcant) range of oxygen abundances compared with other clusters. Thus if deep-mixing is responsible for the changes in 12 C, 13 C and N, it may be that the mixing is not deep enough to penetrate those layers of the hydrogen-burning shell in which signiﬁcant burning of O→N takes place. Similarly, Brown & Wallerstein (1992) noted that their sample of M4 giants all had signiﬁcant overabundances of Na (see also Gratton et al. 1986 and Gratton 1987) and Al, but not the corresponding depletions of O and Mg that might be expected if mixing had been deep enough to bring up the products of ON, NeNa and MgAl cycling.
To summarize, M4 RGB stars have been subjected to high resolution spectral analysis in –5– three independent investigations: Brown & Wallerstein (1992 and references therein; seven stars);
Drake et al. (1992, 1994; four stars); and Gratton et al. (1986; three stars). However, there is little agreement among these investigations about the behavior of light element abundances.
Unfortunately, the only overlap among these samples is M4 L3624 (Wallerstein, Leep, & Oke 1987;
Drake et al. 1992, 1994). Moreover, the data sets and analysis techniques vary substantially among the investigations, so possible systematic eﬀects in the abundances cannot easily be explored. The total sample size for M4 remains small.
We therefore have used the ´chelle spectrographs of McDonald and Lick Observatories to e gather high resolution, large wavelength coverage, high signal-to-noise spectra of 25 M4 giant stars. Our spectrum analysis methodology yields values of Teﬀ independent of the photometric colors. This provides an independent measure of the reddening to each program star as well as the total-to-selective extinction ratio in the part of the dark cloud that obscures the cluster, a quantity believed to be anomalous (e.g. the review by Dixon & Longmore 1993). We have derived N and O abundances and 12 C/13 C ratios, abundances of the light elements Na, Mg, Al, Si, Ca, and Ti, the Fe-peak elements Sc, V, Fe and Ni, and the neutron-capture elements Ba and Eu. Additionally, we have used a multi-object spectrometer at the Cerro Tololo Interamerican Observatory to obtain medium-resolution spectra of 24 M4 giants. Thirteen of the stars in the medium resolution survey are also part of our high resolution sample, and the other eleven are generally giants of lower luminosity. These spectra have limited wavelength coverage, centered near λ6300 ˚, from which A we derived [O/Fe] ratios. With this large sample of M4 giants we explore the distribution of several key light and heavy element abundances in M4.
2. Observations, Reductions, and Equivalent Width Measurements