1: \documentclass[12pt,preprint]{aastex}
2:
3: \bibliographystyle{apj}
4: \newcommand{\teff}{$T_{\rm eff}$}
5: \newcommand{\kms}{km s$^{-1}$}
6: \begin{document}
7:
8: \title{Rubidium and lead abundances in giant stars of the globular
9: clusters M4 and M5\footnote{Based on
10: observations made with the Magellan Clay Telescope at Las Campanas
11: Observatory.}}
12:
13: \author{David Yong}
14: \affil{Research School of Astronomy and Astrophysics, Australian National
15: University, Mount Stromlo Observatory, Cotter Road, Weston Creek, ACT 2611,
16: Australia}
17: \email{yong@mso.anu.edu.au}
18:
19: \author{David L.\ Lambert}
20: \affil{The W.J. McDonald Observatory, University of Texas, Austin, TX 78712}
21: \email{dll@astro.as.utexas.edu}
22:
23: \author{Diane B.\ Paulson}
24: \affil{NASA's Goddard Space Flight Center, Code 693.0, Greenbelt MD
25: 20771}
26: \email{diane.b.paulson@gsfc.nasa.gov}
27:
28: \author{Bruce W.\ Carney}
29: \affil{Department of Physics \& Astronomy, University of North
30: Carolina, Chapel Hill, NC 27599-3255}
31: \email{bruce@physics.unc.edu}
32:
33: \begin{abstract}
34:
35: We present measurements of the neutron-capture elements Rb and Pb for
36: bright giants in the globular clusters M4
37: and M5. The clusters are of similar metallicity
38: ([Fe/H] $\simeq -1.2)$ but M4 is decidedly $s$-process enriched
39: relative to M5: [Ba/Fe] = +0.6 for M4 but 0.0 for M5. The Rb and Pb abundances
40: were derived by comparing synthetic spectra
41: with high-resolution, high signal-to-noise ratio spectra obtained with
42: MIKE on the Magellan telescope. Abundances of Y, Zr, La, and Eu
43: were also obtained. In M4, the mean abundances from 12 giants are
44: [Rb/Fe] = 0.39 $\pm$ 0.02 ($\sigma$ = 0.07),
45: [Rb/Zr] = 0.17 $\pm$ 0.03 ($\sigma$ = 0.08), and
46: [Pb/Fe] = 0.30 $\pm$ 0.02 ($\sigma$ = 0.07). In M5, the mean abundances from two giants are
47: [Rb/Fe] = 0.00 $\pm$ 0.05 ($\sigma$ = 0.06),
48: [Rb/Zr] = 0.08 $\pm$ 0.08 ($\sigma$ = 0.11), and
49: [Pb/Fe] = $-$0.35 $\pm$ 0.02 ($\sigma$ = 0.04).
50: Within the measurement uncertainties, the abundance ratios [Rb/Fe],
51: [Pb/Fe] and [Rb/X] for X = Y, Zr, La
52: are constant from star-to-star in each cluster and
53: none of these ratios are correlated with O or Na abundances. While M4 has a
54: higher Rb abundance than M5, the ratios [Rb/X] are similar in both clusters
55: indicating that
56: the nature of the $s$-products are very similar for each cluster
57: but the gas from which M4's stars formed had a higher concentration
58: of these products.
59:
60: \end{abstract}
61:
62: \keywords{Galaxy: Abundances, Galaxy: Globular Clusters: Individual: Messier
63: Number: M4, Galaxy: Globular Clusters: Individual: Messier Number: M5,
64: Stars: Abundances}
65:
66: \section{Introduction}
67: \label{sec:intro}
68:
69: Globular clusters continue to provide a source of fascination and frustration
70: to both theorists and observers. Two notable accomplishments include the
71: use of globular clusters to (a) check the age of the Universe
72: (e.g., \citealt{gratton03c})
73: and to (b) test and refine our understanding of stellar
74: evolution (e.g., \citealt{renzini88}).
75: Despite these successes, globular clusters
76: continue to present bewildering puzzles. The most persistent puzzle relates
77: to chemical composition.
78:
79: For many years, globular clusters have been known to exhibit star-to-star
80: abundance variations for the light elements C, N, O, Na, Mg, and Al
81: (e.g., see reviews by \citealt{smith87}, \citealt{kraft94},
82: and \citealt{gratton04}). While
83: the amplitude of the star-to-star abundance
84: dispersion can vary from cluster to cluster,
85: the now familiar anticorrelations between C and N, O and Na, and Mg and Al
86: reveal that the abundance variations are likely produced during
87: hydrogen burning at high temperatures via the CNO, Ne-Na, and Mg-Al
88: cycles. (The O-Na and Mg-Al anticorrelations are not seen in field stars.)
89: However, the stars responsible for the nucleosynthesis and the
90: nature of the pollution mechanism(s) remain poorly understood (see
91: \citealt{lattanzio06} for a recent summary).
92:
93: One possible explanation for the observed abundance anomalies
94: is internal mixing and nucleosynthesis
95: (e.g., \citealt{sm79,charbonnel95}) within the present cluster members,
96: the so-called evolutionary scenario.
97: The systematic variation of
98: the C and N \citep{ss91} and Li \citep{grundahl02} abundances
99: with luminosity along the red giant branch demand an evolutionary
100: component to the star-to-star abundance variations.
101: Dredge-up of CN-cycled material accounts for the C and N variations. Development
102: of a giant's convective envelope leading to mixing with highly Li-depleted
103: gas accounts for the decline of the Li abundance with increasing luminosity.
104: The proton-capture reactions causing the O, Na, Mg, and Al variations demand
105: much higher temperatures and much deeper mixing than those required for
106: CN-cycling. Such
107: mixing is not predicted by standard theoretical models of red giants
108: and the discovery of the O, Na, Mg, and Al anomalies in main sequence
109: stars (e.g., \citealt{briley96}, \citealt{gratton01}) eliminates
110: deep mixing as a viable explanation for the O-Al variations. The
111: interiors of main sequence stars are too cool to process Ne to Na or Mg to Al.
112: Therefore, the cluster gas must have been inhomogeneous when the present
113: stars were formed. This alternative explanation for the abundance anomalies
114: is the so-called primordial scenario.
115:
116: In the primordial scenario,
117: intermediate-mass asymptotic giant branch stars (IM-AGBs) from the generation
118: to which the observed stars belong have long been
119: considered candidates for synthesizing the abundance variations
120: \citep{cottrell81}. In IM-AGBs, the convective envelope can reach the top
121: of the hydrogen-burning shell, a process called hot-bottom burning. For
122: sufficiently massive and metal-poor AGBs, the temperatures at the
123: base of the convective envelope can exceed
124: 100 million degrees thereby allowing the efficient operation of the CNO,
125: Ne-Na, and Mg-Al cycles (e.g., \citealt{karakas03}).
126: That IM-AGBs do not alter the abundances of the
127: alpha or iron-peak elements (as required by observations)
128: adds to their qualitative appeal. However, quantitative tests reveal
129: problems with the IM-AGB primordial scenario. Theoretical yields
130: from IM-AGBs combined with a chemical evolution model \citep{fenner04}
131: suggest that O is not sufficiently depleted, Na is overproduced, Mg is
132: produced rather than destroyed, the isotope ratios of Mg do not match the
133: observations, and the sum of C+N+O increases substantially in contrast to
134: the observations.
135: \citet{ventura05a,ventura05b,ventura05c} find that many of the flaws noted
136: above can be alleviated when IM-AGB yields are calculated using a revised
137: treatment for convection and mass-loss. However, \citeauthor{ventura05a}
138: note that problems persist, namely with the Mg isotope ratios, and warn
139: that the predictive power of the current AGB models is limited. Recently,
140: \citet{prantzos06}, \citet{smith06}, and \citet{decressin06}
141: suggest that the winds from massive stars may be more
142: promising candidates than IM-AGBs. There is no satisfactory explanation for
143: the complex patterns for the light element abundances exhibited by every
144: well studied Galactic globular cluster. Therefore, our present
145: understanding of globular cluster chemical evolution and/or stellar
146: nucleosynthesis is incomplete.
147:
148: Determinations of the stellar abundances of the trans-iron or heavy elements
149: offer clues to the history behind the chemical evolution of
150: globular clusters. Here, we provide novel information -- the Rb and Pb abundances --
151: for giants in M4 and M5, a pair of clusters of similar metallicity but
152: with distinctly different levels of $s$-process products.
153: The quintessential $r$-process element Eu has similar abundances in the
154: two clusters and, indeed, across
155: the collection of Galactic globular
156: clusters. In sharp contrast, the $s$-process products are more evident in M4 than in
157: M5 and other clusters of similar metallicity: [Ba/Fe] is about +0.6 in M4 but
158: 0.0 in M5. The questions - Are there differences in the Rb and Pb abundances
159: between this pair of clusters? and Are the star-to-star variations in the
160: abundances of light elements (O, Na, Mg, and Al) reflected in variations among the
161: abundances of Rb and Pb? -- seem likely to probe the origins of the
162: $s$- and $r$-process products for globular clusters.
163:
164: Due to a critical
165: branching point in the $s$-process path at $^{85}$Kr, the abundance
166: of Rb relative to Sr, Y, or Zr can differ by a factor of 10 depending upon
167: the neutron density at the $s$-process site. In the case of AGB stars, the
168: neutron density in the He-shell is dependent on the stellar mass
169: (e.g., see
170: \citealt{tomkin83}, \citealt{lambert95}, \citealt{busso99}, and
171: \citealt{abia01} for further details). Since the isotopes of Pb and Bi are
172: the last stable nuclei on the $s$-process path, the $s$-process terminates at these
173: elements and overabundances of Pb and Bi will arise if seed nuclei are
174: shuffled by neutron captures down the entire $s$-process path.
175: In particular, metal-poor AGB stars may
176: produce large overabundances of Pb and Bi if the neutron supply per seed
177: exceeds a critical value (e.g., see \citealt{goriely01},
178: \citealt{travaglio01}, and \citealt{busso01} for further details).
179: The suspicion is that the star-to-star abundance variations for light
180: elements are due to contamination by IM-AGBs.
181: Some contend that IM-AGBs also synthesize $s$-process nuclides
182: and then one might expect to see star-to-star variations in the
183: Rb and Pb abundances as well as correlations with light element abundances.
184:
185: To further examine the possible role of IM-AGBs in the chemical evolution
186: of globular clusters, \citet{rbpbsubaru}
187: measured Rb and Pb in NGC 6752 and M13, the two
188: clusters that exhibit the largest amplitude for Al variations. It was found that
189: the abundance ratios [Rb/Zr] and [Pb/Fe] were constant from
190: star-to-star within the measurement uncertainties.
191: If IM-AGBs do synthesize
192: Rb and Pb, then they may not be responsible for the abundance variations.
193: On the other hand, if IM-AGBs are responsible for the abundance variations,
194: they cannot synthesize Rb or Pb.
195:
196: In this paper, we extend the measurements of Rb and Pb to the globular
197: clusters M4 and M5. While these clusters are more metal-rich than NGC 6752
198: or M13, both M4 and M5 are known to exhibit large dispersions
199: and correlations for the light element abundances [see pioneering
200: studies on CN bimodality by \citet{norris81a} and \citet{smith83} as well
201: as recent high-resolution spectroscopic studies by \citet{M4,M5},
202: \citet{ramirez03}, and references therein]. In particular, as noted above,
203: M4 is remarkably, perhaps uniquely among globular clusters, enriched in $s$-process
204: products.
205:
206: \section{Observations and data reduction}
207: \label{sec:data}
208:
209: The targets included 12 stars in M4 previously studied by \citet{M4}
210: and two stars in M5 previously studied by \citet{M5} and \citet{ramirez03}.
211: The main focus was to observe a large number of stars in M4 and though we
212: were restricted to the brightest giants, we note that the sample spans
213: a considerable range of the known star-to-star elemental abundance variations.
214: The smaller sample in M5 is due to the fact that those observations were
215: conducted during other observing programs when primary targets were
216: unavailable. In Table \ref{tab:param}, we list the program stars.
217: The stellar identifications for M4 and M5 are from \citet{lee77} and
218: \citet{arp62} respectively.
219:
220: The observations were performed with the Magellan
221: Telescope using the Magellan Inamori Kyocera
222: Echelle spectrograph (MIKE; \citealt{mike}) on 2004 June 12-13, July 16,
223: and July 18. A 0.35\arcsec\
224: slit was used providing a resolving power of
225: R $\equiv \lambda/\Delta\lambda$ =
226: 55,000 in the red and R $=$ 65,000 in
227: the blue per 4 pixel resolution element with wavelength coverage from
228: 3800~\AA~to 8500~\AA. The IRAF\footnote{IRAF (Image Reduction and Analysis
229: Facility) is distributed by the National Optical Astronomy
230: Observatory, which is operated by the Association of Universities
231: for Research in Astronomy, Inc., under contract with the National
232: Science Foundation.} package of programs
233: was used for most of the data reduction. In order
234: to correct for the fact that the lines are severely tilted with respect
235: to the orders and the tilt varies across the CCD, we used the
236: {\sc mtools}\footnote{http://www.lco.cl/lco/magellan/instruments/MIKE/reductions/mtools.html}
237: set of tasks written by Jack Baldwin to extract the spectral orders.
238: In the one-dimensional wavelength-calibrated normalized spectra, the
239: typical signal-to-noise ratio was 70 per pixel at 4050~\AA~(140 per
240: resolution element) and 400 per pixel at 7800~\AA~(800 per resolution
241: element).
242:
243: \section{Analysis}
244: \label{sec:analysis}
245:
246: \subsection{Stellar parameters and the iron abundance}
247: \label{sec:param}
248:
249: The required stellar parameters for an abundance analysis are the
250: effective temperature (\teff), the surface gravity ($\log g$), and the
251: microturbulent velocity ($\xi_t$).
252: Our analysis techniques closely follow \citet{rbpbsubaru} in which
253: we determined these values adopting a
254: traditional spectroscopic approach. Using routines in IRAF, we measured the
255: equivalent widths (EWs) for a set of Fe\,{\sc i} and Fe\,{\sc ii} lines.
256: The set of Fe lines was identical to those used by \citet{mghsubaru} for
257: which the $gf$-values were taken from \citet{lambert96}, \citet{fe2},
258: \citet{blackwell95} and references therein. Model atmospheres were taken
259: from the \citet{kurucz93} local thermodynamic equilibrium (LTE) stellar
260: atmosphere grid and we interpolated within the grid when necessary. We
261: used the LTE stellar line analysis program {\sc moog} \citep{moog} to determine
262: abundances for a given line. We adjusted \teff~until there was no trend
263: between the abundances from Fe\,{\sc i} lines
264: and the lower excitation potential.
265: We adjusted $\log g$ until the abundances from Fe\,{\sc i} and Fe\,{\sc ii}
266: lines were in agreement. We adjusted $\xi_t$ until there was no trend between
267: the abundances from Fe\,{\sc i} lines and EW. This process was iterated
268: until
269: all three parameters were simultaneously constrained. The final [Fe/H] was
270: the mean from all Fe lines assuming a solar abundance
271: log~$\epsilon$(Fe) = 7.50. The stellar parameters are given in Table
272: \ref{tab:param} and we estimate the internal errors to be \teff~$\pm$~50~K,
273: $\log g~\pm$~0.2 dex, and $\xi_t~\pm$~0.2 \kms.
274:
275: As an additional check on our surface gravities (and analysis techniques),
276: we compared our
277: $\log g$ values with the $Y^2$ isochrones \citep{y2}. We adopted the
278: 13 Gyr isochrones with [$\alpha$/Fe] = +0.3, our
279: spectroscopic \teff, and interpolated between the two closest metallicities
280: z = 0.004 and z = 0.001. We found that our surface gravities were
281: in fair agreement with the $Y^2$ isochrones,
282: $\log g_{\rm spec} - \log g_{\rm isochrone} = -0.22$ $\pm$ 0.07
283: ($\sigma$ = 0.26).
284:
285: The stellar parameters are in good agreement with those obtained by
286: \citet{M4,M5} and \citet{ramirez03}. For the 12 M4 giants, the mean
287: differences (This Study $-$ \citet{M4}) are
288: $\Delta$\teff~= 71 $\pm$ 17 ($\sigma$ = 58 K),
289: $\Delta\log g$ = 0.02 $\pm$ 0.05 ($\sigma$ = 0.17),
290: $\Delta\xi_t$ = 0.04 $\pm$ 0.04 ($\sigma$ = 0.15 \kms), and
291: $\Delta$[Fe/H] = $-$0.05 $\pm$ 0.01 ($\sigma$ = 0.05). For the 2 M5 giants,
292: the mean differences (This Study $-$ \citet{ramirez03}) are
293: $\Delta$\teff~= 20 $\pm$ 5 ($\sigma$ = 7 K),
294: $\Delta\log g$ = $-$0.15 $\pm$ 0.15 ($\sigma$ = 0.21),
295: $\Delta\xi_t$ = $-$0.16 $\pm$ 0.03 ($\sigma$ = 0.04 \kms), and
296: $\Delta$[Fe/H] = 0.05 $\pm$ 0.02 ($\sigma$ = 0.03). For M5 IV-81, our
297: \teff~is 105 K higher, $\log g$ is 0.3 dex higher, $\xi_t$ is
298: 0.02 \kms~lower, and [Fe/H] is 0.12 dex higher than the values determined
299: by \citet{M5}.
300:
301: \subsection{Rubidium abundances}
302:
303: The abundances for Rb were determined via spectrum synthesis of the
304: 7800~\AA~Rb\,{\sc i} line (see Figure \ref{fig:rbfit}).
305: For all stars in our sample, this Rb line is blended with a
306: Si\,{\sc i} line as well as with weak CN lines and therefore
307: an equivalent width analysis
308: is not possible. Even in the coolest stars, the
309: Rb line is only 10\% deep relative to the continuum such that accurate
310: abundances can be derived only from high-resolution, high
311: signal-to-noise ratio spectra. The adopted wavelengths and relative strengths
312: for the isotopic and hyperfine-structure components of Rb were identical to
313: \citet{lambert76} and \citet{tomkin99}. We assumed a solar isotope ratio
314: $^{85}$Rb/$^{87}$Rb = 3 and the macroturbulent broadening was fixed by
315: fitting the profile of the nearby 7798~\AA~Ni\,{\sc i} line. Synthetic
316: spectra were generated using {\sc Moog} and the Si and Rb abundances
317: were varied to obtain the best fit to the observed spectrum.
318: Our tests confirmed the finding by \citet{lambert76} that Rb isotope ratios
319: cannot be measured from the 7800~\AA~line due to the hyperfine structure
320: and the small isotopic shift. Not surprisingly, our tests showed that
321: the derived Rb abundances are insensitive to the assumed isotope ratio.
322: In \citet{rbpbsubaru}, we took the \citet{kurucz84}
323: solar atlas and measured an abundance log $\epsilon$(Rb) = 2.58 using a model
324: atmosphere with \teff/$\log g$/$\xi_t$ = 5770/4.44/0.85. Our solar Rb
325: abundance is in excellent agreement with the \citet{grevesse98} value,
326: log $\epsilon$(Rb) = 2.60.
327: The weaker Rb\,{\sc i} resonance line near 7947~\AA~is detected but in a region
328: affected by
329: unidentified blends, atmospheric absorption, and fringing.
330: While preliminary
331: analyses suggest that the abundances derived from the 7947~\AA~line are
332: similar to those derived from the 7800~\AA~line
333: (see \citealt{rbpbsubaru} for a more detailed comparison),
334: we restrict our analyses to the 7800~\AA~line.
335:
336: In this study, we included CN molecular lines when fitting the 7800~\AA~Rb
337: line. The CN lines were taken from \citet{kurucz06}. While the inclusion of
338: these lines resulted in an improved fit to the local continuum, we note that
339: the Rb abundances were unaffected. Even for the coolest stars in M4 for
340: which the CN line strength was at a maximum, the
341: Rb abundances increased by $\le$ 0.02 dex if the CN lines were omitted.
342: For the more metal-poor cluster NGC 6752, we re-analyzed a subset of the
343: \citet{rbpbsubaru} spectra and found that the Rb abundances were
344: unchanged when CN lines were included in the analysis.
345:
346: \subsection{Lead abundances}
347:
348: The abundances for Pb were determined via spectrum synthesis of the
349: 4058~\AA~Pb\,{\sc i} line (see Figure \ref{fig:pbfit}).
350: High-resolution, high signal-to-noise ratio
351: spectra are essential for measuring Pb abundances
352: because the region near 4058~\AA~is crowded with
353: molecular lines of CH as well as Mg, Ti, Mn, Fe, and Co atomic lines.
354: Since M4 and M5 are more metal-rich than NGC 6752 and M13, the
355: 4058~\AA~region is more crowded and the uncertainties in the derived
356: Pb abundances are greater. However, we note that our syntheses
357: provide a very good fit to the observed
358: spectra. The macroturbulent broadening was fixed by fitting the profiles
359: of nearby lines. The $gf$-value of the Pb\,{\sc i} line
360: was the same as that used by \citet{aoki02}
361: as were the hyperfine-structure and isotopic components. We assumed
362: a solar isotope ratio and our tests confirmed that the derived Pb abundances
363: do not depend upon the assumed isotope ratio. For the solar Pb
364: abundance, we adopted log~$\epsilon$(Pb) = 1.95 from \citet{grevesse98}.
365:
366: \subsection{Additional elements}
367:
368: The abundances for O, Na, Mg, Al, Y, Zr, La, and Eu were also measured
369: in the program stars using the same lines as \citet{67522}. Y and Zr were
370: selected since the
371: ratios [Rb/Y] and [Rb/Zr] are
372: sensitive to the neutron density at the site of the
373: $s$-process. O, Na, Mg, and Al were
374: measured since they are known to vary from star-to-star in these and other
375: globular clusters. Re-measuring these
376: elements in M4 and M5 provides a check to see if our abundance
377: determinations are on the same scale as other investigators. Similarly, the
378: neutron-capture elements La and Eu were measured (hyperfine and/or isotopic
379: splitting was included).
380: When deriving O abundances, we generated synthetic spectra to account
381: for possible blending from CN molecular lines. We assumed C and N
382: abundances interpolated as a function of Na with the C and N abundances
383: taken from \citet{M4,M5} and \citet{smith97}. A subset of abundance
384: determinations were conducted using spectrum synthesis to account for
385: possible blends.
386: In Table \ref{tab:abund} we present the measured elemental abundances for
387: O, Na, Mg, Al, Rb, Y, Zr, La, Eu, and Pb. The adopted solar abundances
388: were 8.69, 6.33, 7.58, 6.47, 2.60, 2.24, 2.60, 1.13, 0.52, and 1.95
389: respectively. These values are identical to those we used in previous papers
390: and were taken from \citet{grevesse98} for all elements
391: except O \citep{allendeO}, La \citep{la}, and Eu \citep{eu}.
392: Table \ref{tab:abund} also contains the Zr abundances shifted
393: onto the \citet{smith00} scale.
394:
395: For M4, the measured elemental abundances
396: are in good agreement with \citet{M4}.
397: The mean differences (This Study $-$ \citet{M4}) are
398: $\Delta$[O/Fe] = 0.34 $\pm$ 0.03 ($\sigma$ = 0.11),
399: $\Delta$[Na/Fe] = 0.14 $\pm$ 0.03 ($\sigma$ = 0.12),
400: $\Delta$[Mg/Fe] = 0.14 $\pm$ 0.02 ($\sigma$ = 0.06),
401: $\Delta$[Al/Fe] = 0.05 $\pm$ 0.03 ($\sigma$ = 0.10),
402: $\Delta$[La/Fe] = 0.01 $\pm$ 0.03 ($\sigma$ = 0.10), and
403: $\Delta$[Eu/Fe] = 0.03 $\pm$ 0.02 ($\sigma$ = 0.07). The agreement for
404: Al, La, and Eu is excellent.
405: The largest discrepancy is for O and differences in the
406: adopted solar abundances can account for 0.24 dex leaving a 0.10 dex
407: residual. While both studies use the same O lines and $gf$ values, they
408: employ spectrum synthesis which presumably involves a different set of
409: atomic and molecular lines.
410: A small offset also exists for
411: Na and Mg which could be due to the different set of lines employed.
412:
413: For M5, the elemental abundances are in fair agreement with \citet{ramirez03}.
414: The mean differences (This Study $-$ \citet{ramirez03}) are
415: $\Delta$[O/Fe] = 0.27 $\pm$ 0.14 ($\sigma$ = 0.20),
416: $\Delta$[Na/Fe] = $-$0.05 $\pm$ 0.10 ($\sigma$ = 0.14),
417: $\Delta$[Mg/Fe] = 0.11 $\pm$ 0.06 ($\sigma$ = 0.08),
418: $\Delta$[Zr/Fe] = $-$0.05 $\pm$ 0.19 ($\sigma$ = 0.27),
419: $\Delta$[La/Fe] = $-$0.03 $\pm$ 0.03 ($\sigma$ = 0.04), and
420: $\Delta$[Eu/Fe] = $-$0.06 $\pm$ 0.09 ($\sigma$ = 0.13). For O, the difference
421: can be attributed to the adopted solar abundance. For the single star also
422: studied by \citet{M5}, the measured abundances are in good agreement
423: $\Delta$[X/Fe] $\le$ 0.07 dex. The difference for La is 0.14 dex (possibly
424: due to the different set of lines) and 0.58 dex for O.
425: The adopted solar abundances can account for 0.24 dex leaving
426: a 0.34 dex residual. Dr I.\ Ivans kindly sent the synthetic and
427: observed spectra near the 6300 \AA~and 6363 \AA~used to derive
428: O abundances.
429: The difference in the measured O abundances between the two studies
430: is $\Delta\log~\epsilon$(O) = 0.14 dex. The difference in the measured
431: Fe abundance is 0.12 dex. The remaining 0.08 dex residual may be
432: attributed to differences in the spectrum synthesis analysis.
433:
434: The adopted line lists and measured equivalent widths are
435: presented in Table \ref{tab:line}. In Table \ref{tab:parvar}, the abundance
436: dependences on the model parameters are given.
437:
438: \section{Discussion}
439: \label{sec:discussion}
440:
441: \subsection{The heavy-element canvas}
442:
443: The heavy elements are synthesized by neutron-captures in the
444: $s$- and the $r$-process. (We ignore here the nuclides, all of
445: low abundance, referred to as $p$-nuclides.) In all but material
446: dominated by
447: $s$-processed products, Eu is a signature
448: element for the $r$-process. There is evidence from analyses of
449: field stars, especially those severely enriched in $r$-process
450: products that the relative abundances for the $r$-process of
451: nuclides from Ba to Eu are invariant with metallicity
452: (e.g., see \citealt{cowan06}).
453: This invariance does not extend to the ratio of heavy products (i.e., Eu)
454: to light products (i.e., Sr, Y, and Zr).
455: In a solar mix of elements, Ba is
456: taken as a measure of the $s$-process with but a slight contamination from
457: the $r$-process \citep{burris00}. The [Ba/Eu] ratio of a star is widely
458: taken to indicate the relative mix of $s$- to $r$-processed
459: material.
460:
461: In the case of ratios such as [Eu/Fe] and [Ba/Fe]
462: for globular cluster stars, three questions
463: arise: Is there an intercluster spread in the ratios? Is there an
464: intra-cluster spread, especially for clusters exhibiting a large
465: spread among light element abundances? How do the cluster ratios
466: compare with those found for the field stars, as a function of [Fe/H]?
467:
468: \subsubsection{The [Eu/Fe] ratio}
469:
470: With but the single known exception of Ruprecht 106, the [Eu/Fe]
471: ratio across
472: the collection of examined Galactic globular
473: clusters with [Fe/H] from about $-$1.0 to $-2.5$ appears to be
474: single-valued.
475: Values of [Eu/Fe] in the literature for well-sampled
476: clusters with [Fe/H] $< -1$ span a reported range of only 0.2 dex
477: from about $+0.35$ to $+0.55$ with no obvious trend with [Fe/H].
478: [Ruprecht 106 is an outlier at $0.0$
479: from analyses of two stars \citep{brown97}. M80 is probably
480: also an outlier at +1.0, but the analyses were based on less
481: than ideal spectra, R=18,000 \citep{csp04}.]
482: This range is gathered from examination of the
483: extensive literature; references cited by \citet{gratton04}
484: were read and supplemented by more recent papers. No
485: attempt has been made by us to correct published values to
486: a common standard. It would be premature to claim that the
487: 0.2 dex spread in [Eu/Fe] is real and thus not entirely attributable to
488: cumulative errors of measurement. Our results
489: [Eu/Fe] = $+0.41$ for M4 and $+0.53$ for M5 fall within this
490: range and, as noted above, are consistent with previously published
491: results.
492:
493: Although our samples are small (see also Yong et al. 2006), the stars
494: chosen for their contrasting light element abundances show no
495: evidence of a dependence on these light element abundances. This
496: decoupling of light element abundances from that of Eu is confirmed
497: by previous studies of well-sampled clusters.
498: Several authors have compared [Eu/Fe] for globular clusters with
499: results for Galactic field stars. The evidence is that the [Eu/Fe]
500: -- [Fe/H] trends (i.e., the apparent independence of [Eu/Fe] on
501: [Fe/H] for [Fe/H] $< -0.7$)
502: are the same for both samples (see, for example, \citet{gratton04},
503: \citet{james04b}, and \citet{pritzl05}).
504:
505: In summary, globular cluster and field stars appear with remarkably few
506: exceptions and to within measurement errors to have the same
507: [Eu/Fe] ratio. Our measurements of this ratio for M4 and M5 are
508: consistent with previous measurements for these clusters and the
509: common cluster-field star set of data.
510:
511: \subsubsection{The [Y/Fe], [Zr/Fe], [Ba/Fe], and [La/Fe] ratios}
512:
513: Among the few heavy elements in abundance analyses
514: of globular cluster stars from which the $s$-process contribution
515: may be assessed, Ba is the most widely reported.
516: Other elements for which abundance data are available
517: for several clusters include Y, Zr, and La.
518:
519: For most globular clusters, the [Ba/Fe] ratio is positive
520: ([Ba/Fe] $\simeq +0.2$) and independent of [Fe/H] over the interval
521: $-0.7$ to about $-1.5$. For the few investigated clusters with
522: [Fe/H] $\le -1.8$, [Ba/Fe] declines to roughly 0.0 at
523: [Fe/H] of $-2.0$ to $-2.5$. This run with [Fe/H] for the clusters
524: is matched by the field stars \citep{james04b}. The spread in
525: [Ba/Fe] at a given [Fe/H] is about $\pm0.2$ dex, a value
526: consistent with the errors of measurement, with several outliers.
527: The sole outlier with [Ba/Fe] above the mean trend is M4 with
528: an excess of about 0.4 dex in [Ba/Fe]. The outliers below the
529: mean trend are NGC 3201, Ruprecht 106,
530: and Palomar 12 \citep{brown97,gonzalez98} as well as
531: NGC 5694 \citep{lee06}.
532: Data for [La/Fe] is less extensive but suggests [La/Fe] $\simeq +0.1$
533: with M4 again as an outlier ([La/Fe] $= +0.45$) but with the Ba-outliers
534: NGC 3201 and Pal 12 as conformers to the mean value. NGC 5694 is again
535: an outlier with [La/Fe] = $-$0.26 based on analysis of a single star
536: \citep{lee06}. The [Y/Fe] ratio is
537: close to zero \citep{james04b}. Our result for M5 at +0.2 from just two
538: stars may be an (unlikely) outlier. At [Y/Fe] = +0.7, M4 is certainly
539: an outlier.
540:
541: The discussed data, as well as less extensive data on other
542: heavy elements, show that the vast majority of the
543: clusters are not
544: distinguishable from field stars by their abundances of Ba and other heavy
545: elements. For a given cluster, there is no compelling evidence for a
546: star-to-star variation correlated with the variations seen among
547: the light elements for many clusters. There are, however,
548: examples of individual cluster stars with $s$-process enrichments,
549: e.g., the CH stars in M2 and M55 \citep{smith90,briley93}, the
550: possible CH star in M22 \citep{mcclure77,vanture92},
551: a Y-rich star in M13 \citep{cohen05}, and a hint of Zr abundance variations
552: in 47 Tuc \citep{wylie06}, M5 \citep{ramirez03}, and NGC 6752 \citep{67522}.
553: While M15 exhibits
554: variations of Ba and Eu, these variations are not correlated with the light
555: element abundances and the ratio [Ba/Eu] is constant from
556: star-to-star \citep{sneden97,otsuki06}.
557:
558: \subsection{Rubidium}
559:
560: In M4 the mean Rb abundance is [Rb/Fe] = 0.39 $\pm$ 0.02 ($\sigma$ = 0.07)
561: and in M5 the mean abundance is [Rb/Fe] = 0.00 $\pm$ 0.05 ($\sigma$ = 0.06).
562: As expected, M4 has a higher abundance of Rb
563: than M5. While 50\% of the solar Rb abundance can be attributed to the
564: $r$-process \citep{burris00},
565: M4 has a slightly lower [Eu/Fe] ratio
566: than does M5 and therefore, it would
567: be difficult to attribute any Rb excess in M4 relative to M5 as
568: being due to a larger $r$-process contribution.
569: Neither cluster shows any evidence for a dispersion in Rb abundances. That is,
570: the scatter in the Rb abundances within M4 and M5 is small and can be
571: attributed entirely to the measurement uncertainties.
572: The Rb abundances are not correlated with O or Na
573: (the two elements that exhibit large star-to-star abundance variations).
574:
575: The Rb abundances of M4 and M5 are compared with the limited
576: data in the literature for other clusters and field stars.
577: In Figure \ref{fig:rb}, we compare the [Rb/Fe] abundance ratios with
578: field dwarfs and giants \citep{gratton94,tomkin99} as well as the
579: globular clusters NGC 3201 \citep{gonzalez98}, $\omega$ Cen \citep{smith00},
580: and NGC 6752 \citep{rbpbsubaru}. M5 has [Rb/Fe] ratios that are essentially
581: identical to $\omega$ Cen giants at the same metallicity. [We note that
582: $\omega$ Cen is now regarded as the nucleus of an accreted dwarf
583: spheroidal galaxy \citep{smith00}.] These Rb abundances
584: are comparable to the lowest values seen in field stars at the same
585: metallicity.
586: Given that M4 is $s$-process enriched relative to
587: other globular clusters and field stars, it is puzzling that at the
588: metallicity of M4, the majority of field stars share the same [Rb/Fe] ratio.
589: With the addition of M4 and M5,
590: the [Rb/Fe] ratios in NGC 6752 appear rather low compared to other globular
591: clusters and field stars. While the sample sizes remain small and the
592: analyses have been performed by different investigators, the globular clusters
593: cover a wider range of [Rb/Fe] values than do the field stars in the same
594: metallicity regime.
595:
596: The relative abundances of Rb, Y, and Zr are very similar for M4 and M5.
597: In Figure \ref{fig:new3}, the upper panel shows [Rb/Y] versus [Y/Fe] for
598: mean values for M4, M5, and NGC 6752. The dotted line corresponds to
599: [Rb/Y]$ = -0.25$, the mean value for the three clusters.
600: Published results for NGC 3201 and $\omega$ Cen do not fall on
601: the line but straddle it. This may indicate zero-point differences between our
602: and other analyses or a real cluster-to-cluster difference among the
603: Rb, Y, and Zr abundances.
604: Considering the individual measurements, we find
605: [Rb/Y] = $-$0.29 $\pm$ 0.02 ($\sigma$ = 0.06) for M4 and
606: [Rb/Y] = $-$0.17 $\pm$ 0.10 ($\sigma$ = 0.13) for M5.
607: For Zr,
608: we find [Rb/Zr] = 0.17 $\pm$ 0.03 ($\sigma$ = 0.08) for M4 and
609: [Rb/Zr] = 0.08 $\pm$ 0.08 ($\sigma$ = 0.11) for M5.
610: (These Zr abundances have
611: been shifted by +0.3 dex onto the \citealt{smith00} scale,
612: see \citealt{rbpbsubaru} for details.)
613: Neither the [Rb/Y] nor the
614: [Rb/Zr] ratio exhibits a
615: star-to-star dispersion and neither is
616: correlated with O or Na abundances.
617: In Figure \ref{fig:rbzr}, we compare the [Rb/Zr] abundance ratio with
618: field dwarfs and giants as well as globular cluster giants. Compared
619: to the $\omega$ Cen giants at a similar metallicity, M4 and M5 have
620: [Rb/Zr] ratios 0.4-0.5 dex higher (which may be due to systematic
621: offsets). For M4, the difference arises due to
622: $\omega$ Cen having less Rb and more Zr. For M5, the difference results from
623: $\omega$ Cen having similar Rb and considerably more Zr.
624:
625: The similarity in heavy element abundance ratios involving Rb extends to
626: [Rb/La] - see middle panel of Figure \ref{fig:new3}. The indication
627: from our analyses is that the relative abundances of Rb to Y, Zr, and La are
628: the same for M5 and NGC 6752 as well as M4,
629: the cluster clearly enriched
630: in $s$-process elements.
631: This demonstrates that the
632: source of the M4 enrichment is identical to that providing the $s$-process
633: elements for the natal material for M5 and NGC 6752.
634: The abundance of Rb relative to the other elements is slightly
635: subsolar, i.e., [Rb/Y] $= -0.25$. There is a well known difference between the
636: spectroscopically determined Rb abundance and that obtained from carbonaceous
637: meteorites: $\log\epsilon$(Rb) = 2.60 versus 2.33, respectively
638: \citep{asplund05}. If the meteoritic value is adopted [Rb/Y] and similar
639: ratios are increased by 0.27 dex and rendered indistinguishable
640: from the solar ratios. Then, $s$-process donors to the natal clouds of
641: the globular clusters and of local stars including the Sun must have
642: been similar.
643:
644: \subsection{Lead}
645:
646: In M4 the mean Pb abundance is
647: [Pb/Fe] = 0.30 $\pm$ 0.02 ($\sigma$ = 0.07)
648: and in M5 the mean abundance is
649: [Pb/Fe] = $-$0.35 $\pm$ 0.02 ($\sigma$ = 0.04).
650: We note that 20\% of the solar Pb abundance is attributable to the
651: $r$-process \citep{burris00} but the [Eu/Fe] ratio is comparable for
652: M4 and M5. Again, since M4 has a lower [Eu/Fe] ratio than M5, the
653: excess Pb in M4 relative to M5 cannot be due to an increased
654: $r$-process contribution for M4.
655: The excess likely arises
656: from more complete processing down the $s$-process chain to its
657: termination at Pb.
658: The difference in [Pb/Fe] between M4 and M5 is 0.65 dex. For
659: the other neutron-capture elements synthesized via the $s$-process,
660: the difference in [X/Fe] between M4 and M5 is typically 0.4 dex.
661: Regarding the scatter in Pb abundances in M4 and M5, we again find that
662: the small dispersion can be entirely explained by the measurement
663: uncertainties. The Pb abundances in
664: M4 and M5 are not correlated with O or Na.
665:
666: In Figure \ref{fig:pb}, we compare the [Pb/Fe] abundance ratios
667: with field stars \citep{sneden98,travaglio01} as well as the globular clusters
668: M13 and NGC 6752 \citep{rbpbsubaru}.
669: Data on Pb in the field stars is obviously very limited. The one certain
670: data point is for HD 23439A, a star enriched in $s$-process products
671: (Tomkin \& Lambert 1999), with [Pb/Fe] $= +0.6$. The other three data points not
672: indicated as upper limits in
673: Figure \ref{fig:pb} are marked with the customary `:' by
674: \citep{travaglio01} and inspection of their published spectra shows that
675: an alternative designation as an upper limit is possibly more
676: appropriate. Our Pb abundance determination for HD 141531 (Yong et al. 2006b)
677: is consistent with the sub-solar [Pb/Fe] ratios for NGC 6752, M5, and M13.
678: Although more Pb abundances are needed for field stars, we suppose that
679: the Pb abundances of the globular clusters NGC 6752, M5, and M13 and
680: normal field stars
681: are in fair agreement, as found for Rb, Y, Zr, Ba, and other heavy
682: elements.
683:
684: In the lower panel of Figure \ref{fig:new3},
685: we plot [Rb/Pb] versus [Pb/Fe] for M4, M5, and NGC 6752.
686: Interestingly, the [Rb/Pb] ratios for M4 and
687: NGC 6752 appear similar which again suggests that the source of the
688: $s$-process enrichment in M4 may be identical to the source of the
689: $s$-process elements for NGC 6752. However, in this Figure, M5 now appears
690: to be the outlier.
691:
692: Given the small comparison sample, clearly it would be of great interest
693: to expand the measurements of Pb to larger samples of field stars. The
694: synthetic spectra in this study and in \citet{rbpbsubaru} indicate that
695: measurements of Pb in cool giants in the range
696: $-$2.0 $\le$ [Fe/H] $\le$ $-$1.0 are feasible.
697:
698: \subsection{Consequences for the primordial scenario}
699:
700: Formation of the Galactic and extragalactic globular clusters
701: remains a subject with many unanswered questions \citep{brodie06}.
702: From the point of view of the quantitative
703: spectroscopist interested in the composition of stars in the
704: Galactic globular clusters, two questions seem preeminent: (i)
705: Why do the cluster and field stars follow the same
706: [X/Fe] versus [Fe/H] relations? (ii) What causes the light element
707: abundance variations first observed among cluster giant stars but now
708: found also in subgiant and main sequence stars?
709:
710: The simplest answer to question (i) would appear to be that
711: the clusters are formed during episodic mergers in the hierarchical
712: formation of the Galaxy, as discussed by \citet{bekki02} and
713: \citet{beasley02}. If the gas
714: content of the protocluster is dominated by that from the
715: Galaxy and not the infalling system, the composition of the
716: cluster stars will be very similar to that of the Galactic
717: stars formed just preceding the merger. Gas remaining after star
718: formation in the cluster will be ejected by supernovae
719: from the massive stars. Star formation will be restricted to a
720: single generation of stars, i.e., stars within a given cluster
721: will have the identical metallicity. Indeed the
722: complete composition of all stars will be the same
723: but for changes resulting from
724: stellar evolution and the agents responsible for the star-to-star
725: variations among light element abundances. These agents comprise
726: what is widely referred to as `the primordial scenario'.
727:
728: For the majority of the Galactic globular clusters subjected to abundance
729: analysis, the compositions do closely resemble field stars of the
730: cluster's metallicity. This is as expected on the merger hypothesis.
731: M4 is an apparent exception with its overabundance of heavy
732: elements (Y, Zr, Ba, La etc, but not Eu) matched by very
733: few field stars. Some of the field stars with
734: the cluster's degree of overabundance
735: of $s$-process products are binaries in which the products were
736: transferred to the present star when its companion was an AGB star.
737: There are very few known examples where mass transfer across a binary
738: is not a viable explanation. The case of HD 23439A and B \citep{tomkin99}
739: would appear to be such a case. HD 23439A shows no sign of radial
740: velocity variation during 14 years of monitoring \citep{latham02}.
741: On the other hand, HD 23439B
742: is a binary but its period and orbital eccentricity do not suggest
743: mass transfer \citep{latham02}.
744:
745: Other clusters
746: with few counterparts among field stars include Rup 106 and Pal 12.
747: In their case, the Sagittarius dwarf spheroidal galaxy now undergoing
748: destruction by the Galaxy has giants with the peculiar composition of
749: Rup 106 and Pal 12 \citep{sbordone06}
750: indicating that they belong to the dwarf spheroidal galaxy.
751: A conjecture is that the
752: cluster M4 was captured from a stellar system in which
753: chemical evolution had led to $s$-process enrichment; a change in
754: the IMF might have been immediately responsible.
755: However, consideration of space velocities reveals that M4's orbit
756: is restricted to the inner disk and bulge (apocentric radius, $R_a$ =
757: 5.9 kpc) whereas M5's orbit may be more consistent with a capture
758: event, $R_a$ = 35.4 kpc \citep{dinescu99}. Given M4's orbit, a
759: comparison of the $s$-process elements in M4 and comparable
760: metallicity stars of the inner disk and bulge would be of great interest.
761:
762: In the primordial scenario, the star-to-star abundance variations
763: apart from those attributable to dredge-up in giants arise from
764: the accretion of gas (or contamination of natal clouds)
765: to varying degrees by cluster stars.
766: The accreted gas is assumed to be mass lost from the more massive
767: and now dead stars of the cluster. Proposed candidates include
768: IM-AGBs who experience extensive nucleosynthesis
769: and lose their envelopes at low velocity so that
770: the ejecta reside within the cluster's potential \citep{cottrell81}.
771: Recently, a role for massive stars in polluting the
772: cluster's environment has been proposed
773: \citep{decressin06,smith06,prantzos06}.
774:
775:
776: Both M4 and M5 exhibit star-to-star abundance variations for the light
777: elements, as does every well studied Galactic globular cluster.
778: The variations involve O, Na, Mg, and Al, as well as C and N for
779: which there is a component related to dredge-up by giants.
780: For the majority of the analyzed clusters (M4 is an obvious exception),
781: the elemental abundances identified with the unpolluted
782: cluster stars match those of field stars of the same metallicity.
783: Two clusters have been studied for variations among the isotopic
784: ratios of Mg but
785: the normal stars in M13 and NGC 6752 (i.e., cluster stars with high O, high Mg,
786: low Na, low Al and whose O-Al abundances match field stars at the same
787: metallicity) had Mg isotope ratios that exceeded measurements in
788: comparable metallicity
789: field stars \citep{6752,mghdwarf,mghsubaru}.
790:
791: In this study, we extended the elemental abundance measurements
792: of Rb and Pb to the globular clusters M4 and M5.
793: All four clusters that we have studied
794: (M4, M5, M13, and NGC 6752)
795: show light element variations, but
796: constant Rb and Pb abundances within each cluster.
797: Thus, the accreted gas providing the light element variations
798: cannot contain significant amounts (or significant deficiencies)
799: of Rb and Pb. The accreted gas
800: must have the same proportions of Rb and Pb as the ambient material.
801:
802: The source of the gas may be IM-AGBs that at least qualitatively
803: account for the range of abundance variations.
804: A quantitative test is impossible at present
805: because the predicted yields from metal-poor IM-AGBs may be model dependent
806: \citep{busso01,ventura05a,ventura05b,ventura05c}.
807: As far as $s$-products are concerned which, as we have emphasised, do not
808: show abundance variations,
809: IM-AGBs remain a viable candidate because
810: \citet{lattanzio04}
811: and Straniero (2006, private communication) suggest that IM-AGBs will not
812: produce any $s$-process elements since the mass of the He shell and the
813: duration of the thermal pulse decreases with increasing AGB mass.
814:
815: Massive stars achieve limited synthesis of $s$-process products but are held
816: to be the sites of the weak $s$-process (and the $r$-process)
817: contributing nuclides
818: from the iron group up to about Rb, Y and Zr \citep{raiteri91}. Predicted
819: yields are such that massive stars could serve to provide the light
820: element abundance variations without leading to predicted variations
821: for Rb, Y, and Zr.
822:
823: In short, the lack of abundance variations for Rb, Y, Zr and
824: other s-process elements may not exclude either IM-AGBs or massive stars
825: as sources of the accreted pollutants.
826:
827: \section{Concluding remarks}
828: \label{sec:summary}
829:
830: In this paper we present measurements
831: of the neutron-capture elements Rb and Pb
832: in the globular clusters M4 and M5. While both clusters exhibit star-to-star
833: abundance variations for the light elements, we find that the
834: abundances of Rb and Pb are constant.
835: None of the abundance ratios [Rb/Fe], [Rb/Zr], and [Pb/Fe] are correlated
836: with O or Na abundances.
837: In the primordial scenario, the abundance variations for the light
838: elements are attributed to different levels of accretion of
839: ejecta from IM-AGBs or massive stars. The fact that the heavy
840: elements including Rb and Pb do not show abundance variations implies
841: that the accreted material has the same composition as the ambient material
842: for the heavy elements (i.e., the accreted material cannot be
843: highly underabundant or overabundant in these elements).
844: That the ratios [Rb/X] for X = Y, Zr, La are
845: similar for M4 and M5 suggests that the source of the $s$-process
846: elements are similar and that M4 had a greater concentration of
847: these products.
848:
849: There remains a need to pursue additional observational tests of the
850: primordial scenario. In particular, present data on the Rb and Pb
851: abundances in field and cluster stars are sparse. The indication
852: that the Mg isotopic ratios of unpolluted or normal cluster
853: stars differ from those of field stars of the same metallicity
854: deserves closer scrutiny by, in particular, extending the
855: measurement of these isotopic ratios to additional clusters.
856:
857: \acknowledgments
858:
859: This research has made use of the SIMBAD database,
860: operated at CDS, Strasbourg, France and
861: NASA's Astrophysics Data System.
862: We thank the anonymous referee for helpful comments and
863: Inese Ivans for sending observed and synthetic spectra.
864: DY thanks Amanda Karakas, Francesca D'Antona,
865: Inese Ivans, John Lattanzio, Oscar Straniero, Paolo Ventura,
866: and Roberto Gallino for helpful discussions and Chris Sneden for
867: providing a line list for the Pb region.
868: This research was performed while DBP held a National Research Council
869: Research Associateship Award at NASA's Goddard Space Flight Center.
870: DLL acknowledges support from the Robert A.\ Welch Foundation of Houston,
871: Texas. BWC acknowledges support from the National Science Foundation
872: through grant AST-0305431 to the University of North Carolina.
873: This research was
874: supported in part by NASA through the American Astronomical Society's Small
875: Research Grant Program.
876:
877:
878: \begin{thebibliography}{85}
879: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
880:
881: \bibitem[{{Abia} {et~al.}(2001){Abia}, {Busso}, {Gallino}, {Dom{\'{\i}}nguez},
882: {Straniero}, \& {Isern}}]{abia01}
883: {Abia}, C., {Busso}, M., {Gallino}, R., {Dom{\'{\i}}nguez}, I., {Straniero},
884: O., \& {Isern}, J. 2001, \apj, 559, 1117
885:
886: \bibitem[{{Allende Prieto} {et~al.}(2001){Allende Prieto}, {Lambert}, \&
887: {Asplund}}]{allendeO}
888: {Allende Prieto}, C., {Lambert}, D.~L., \& {Asplund}, M. 2001, \apjl, 556, L63
889:
890: \bibitem[{{Aoki} {et~al.}(2002){Aoki}, {Ryan}, {Norris}, {Beers}, {Ando}, \&
891: {Tsangarides}}]{aoki02}
892: {Aoki}, W., {Ryan}, S.~G., {Norris}, J.~E., {Beers}, T.~C., {Ando}, H., \&
893: {Tsangarides}, S. 2002, \apj, 580, 1149
894:
895: \bibitem[{{Arp}(1962)}]{arp62}
896: {Arp}, H. 1962, \apj, 135, 311
897:
898: \bibitem[{{Asplund} {et~al.}(2005){Asplund}, {Grevesse}, \&
899: {Sauval}}]{asplund05}
900: {Asplund}, M., {Grevesse}, N., \& {Sauval}, A.~J. 2005, in ASP Conf. Ser. 336:
901: Cosmic Abundances as Records of Stellar Evolution and Nucleosynthesis, ed.
902: T.~G. {Barnes}, III \& F.~N. {Bash}, 25
903:
904: \bibitem[{{Beasley} {et~al.}(2002){Beasley}, {Baugh}, {Forbes}, {Sharples}, \&
905: {Frenk}}]{beasley02}
906: {Beasley}, M.~A., {Baugh}, C.~M., {Forbes}, D.~A., {Sharples}, R.~M., \&
907: {Frenk}, C.~S. 2002, \mnras, 333, 383
908:
909: \bibitem[{{Bekki} {et~al.}(2002){Bekki}, {Forbes}, {Beasley}, \&
910: {Couch}}]{bekki02}
911: {Bekki}, K., {Forbes}, D.~A., {Beasley}, M.~A., \& {Couch}, W.~J. 2002, \mnras,
912: 335, 1176
913:
914: \bibitem[{{Bernstein} {et~al.}(2003){Bernstein}, {Shectman}, {Gunnels},
915: {Mochnacki}, \& {Athey}}]{mike}
916: {Bernstein}, R., {Shectman}, S.~A., {Gunnels}, S.~M., {Mochnacki}, S., \&
917: {Athey}, A.~E. 2003, in Instrument Design and Performance for
918: Optical/Infrared Ground-based Telescopes. Edited by Iye, Masanori; Moorwood,
919: Alan F. M. Proceedings of the SPIE, Volume 4841, pp. 1694-1704 (2003).,
920: 1694--1704
921:
922: \bibitem[{{Bi{\' e}mont} {et~al.}(1991){Bi{\' e}mont}, {Baudoux}, {Kurucz},
923: {Ansbacher}, \& {Pinnington}}]{fe2}
924: {Bi{\' e}mont}, E., {Baudoux}, M., {Kurucz}, R.~L., {Ansbacher}, W., \&
925: {Pinnington}, E.~H. 1991, \aap, 249, 539
926:
927: \bibitem[{{Blackwell} {et~al.}(1995){Blackwell}, {Lynas-Gray}, \&
928: {Smith}}]{blackwell95}
929: {Blackwell}, D.~E., {Lynas-Gray}, A.~E., \& {Smith}, G. 1995, \aap, 296, 217
930:
931: \bibitem[{{Briley} {et~al.}(1993){Briley}, {Smith}, {Hesser}, \&
932: {Bell}}]{briley93}
933: {Briley}, M.~M., {Smith}, G.~H., {Hesser}, J.~E., \& {Bell}, R.~A. 1993, \aj,
934: 106, 142
935:
936: \bibitem[{{Briley} {et~al.}(1996){Briley}, {Smith}, {Suntzeff}, {Lambert},
937: {Bell}, \& {Hesser}}]{briley96}
938: {Briley}, M.~M., {Smith}, V.~V., {Suntzeff}, N.~B., {Lambert}, D.~L., {Bell},
939: R.~A., \& {Hesser}, J.~E. 1996, \nat, 383, 604
940:
941: \bibitem[{{Brodie} \& {Strader}(2006)}]{brodie06}
942: {Brodie}, J.~P. \& {Strader}, J. 2006, \araa, 44, 193
943:
944: \bibitem[{{Brown} {et~al.}(1997){Brown}, {Wallerstein}, \& {Zucker}}]{brown97}
945: {Brown}, J.~A., {Wallerstein}, G., \& {Zucker}, D. 1997, \aj, 114, 180
946:
947: \bibitem[{{Burris} {et~al.}(2000){Burris}, {Pilachowski}, {Armandroff},
948: {Sneden}, {Cowan}, \& {Roe}}]{burris00}
949: {Burris}, D.~L., {Pilachowski}, C.~A., {Armandroff}, T.~E., {Sneden}, C.,
950: {Cowan}, J.~J., \& {Roe}, H. 2000, \apj, 544, 302
951:
952: \bibitem[{{Busso} {et~al.}(2001){Busso}, {Gallino}, {Lambert}, {Travaglio}, \&
953: {Smith}}]{busso01}
954: {Busso}, M., {Gallino}, R., {Lambert}, D.~L., {Travaglio}, C., \& {Smith},
955: V.~V. 2001, \apj, 557, 802
956:
957: \bibitem[{{Busso} {et~al.}(1999){Busso}, {Gallino}, \& {Wasserburg}}]{busso99}
958: {Busso}, M., {Gallino}, R., \& {Wasserburg}, G.~J. 1999, \araa, 37, 239
959:
960: \bibitem[{{Cavallo} {et~al.}(2004){Cavallo}, {Suntzeff}, \&
961: {Pilachowski}}]{csp04}
962: {Cavallo}, R.~M., {Suntzeff}, N.~B., \& {Pilachowski}, C.~A. 2004, \aj, 127,
963: 3411
964:
965: \bibitem[{{Charbonnel}(1995)}]{charbonnel95}
966: {Charbonnel}, C. 1995, \apjl, 453, L41
967:
968: \bibitem[{{Cohen} \& {Mel{\' e}ndez}(2005)}]{cohen05}
969: {Cohen}, J.~G. \& {Mel{\' e}ndez}, J. 2005, \aj, 129, 303
970:
971: \bibitem[{{Cottrell} \& {Da Costa}(1981)}]{cottrell81}
972: {Cottrell}, P.~L. \& {Da Costa}, G.~S. 1981, \apjl, 245, L79
973:
974: \bibitem[{{Cowan} \& {Sneden}(2006)}]{cowan06}
975: {Cowan}, J.~J. \& {Sneden}, C. 2006, \nat, 440, 1151
976:
977: \bibitem[{{Decressin} {et~al.}(2007){Decressin}, {Meynet}, {Charbonnel},
978: {Prantzos}, \& {Ekstr{\"o}m}}]{decressin06}
979: {Decressin}, T., {Meynet}, G., {Charbonnel}, C., {Prantzos}, N., \&
980: {Ekstr{\"o}m}, S. 2007, \aap, 464, 1029
981:
982: \bibitem[{{Demarque} {et~al.}(2004){Demarque}, {Woo}, {Kim}, \& {Yi}}]{y2}
983: {Demarque}, P., {Woo}, J.-H., {Kim}, Y.-C., \& {Yi}, S.~K. 2004, \apjs, 155,
984: 667
985:
986: \bibitem[{{Dinescu} {et~al.}(1999){Dinescu}, {Girard}, \& {van
987: Altena}}]{dinescu99}
988: {Dinescu}, D.~I., {Girard}, T.~M., \& {van Altena}, W.~F. 1999, \aj, 117, 1792
989:
990: \bibitem[{{Fenner} {et~al.}(2004){Fenner}, {Campbell}, {Karakas}, {Lattanzio},
991: \& {Gibson}}]{fenner04}
992: {Fenner}, Y., {Campbell}, S., {Karakas}, A.~I., {Lattanzio}, J.~C., \&
993: {Gibson}, B.~K. 2004, \mnras, 353, 789
994:
995: \bibitem[{{Gonzalez} \& {Wallerstein}(1998)}]{gonzalez98}
996: {Gonzalez}, G. \& {Wallerstein}, G. 1998, \aj, 116, 765
997:
998: \bibitem[{{Goriely} \& {Siess}(2001)}]{goriely01}
999: {Goriely}, S. \& {Siess}, L. 2001, \aap, 378, L25
1000:
1001: \bibitem[{{Gratton} {et~al.}(2004){Gratton}, {Sneden}, \&
1002: {Carretta}}]{gratton04}
1003: {Gratton}, R., {Sneden}, C., \& {Carretta}, E. 2004, \araa, 42, 385
1004:
1005: \bibitem[{{Gratton} {et~al.}(2001){Gratton}, {Bonifacio}, {Bragaglia},
1006: {Carretta}, {Castellani}, {Centurion}, {Chieffi}, {Claudi}, {Clementini},
1007: {D'Antona}, {Desidera}, {Fran{\c c}ois}, {Grundahl}, {Lucatello}, {Molaro},
1008: {Pasquini}, {Sneden}, {Spite}, \& {Straniero}}]{gratton01}
1009: {Gratton}, R.~G., {Bonifacio}, P., {Bragaglia}, A., {Carretta}, E.,
1010: {Castellani}, V., {Centurion}, M., {Chieffi}, A., {Claudi}, R., {Clementini},
1011: G., {D'Antona}, F., {Desidera}, S., {Fran{\c c}ois}, P., {Grundahl}, F.,
1012: {Lucatello}, S., {Molaro}, P., {Pasquini}, L., {Sneden}, C., {Spite}, F., \&
1013: {Straniero}, O. 2001, \aap, 369, 87
1014:
1015: \bibitem[{{Gratton} {et~al.}(2003){Gratton}, {Bragaglia}, {Carretta},
1016: {Clementini}, {Desidera}, {Grundahl}, \& {Lucatello}}]{gratton03c}
1017: {Gratton}, R.~G., {Bragaglia}, A., {Carretta}, E., {Clementini}, G.,
1018: {Desidera}, S., {Grundahl}, F., \& {Lucatello}, S. 2003, \aap, 408, 529
1019:
1020: \bibitem[{{Gratton} \& {Sneden}(1994)}]{gratton94}
1021: {Gratton}, R.~G. \& {Sneden}, C. 1994, \aap, 287, 927
1022:
1023: \bibitem[{{Grevesse} \& {Sauval}(1998)}]{grevesse98}
1024: {Grevesse}, N. \& {Sauval}, A.~J. 1998, Space Science Reviews, 85, 161
1025:
1026: \bibitem[{{Grundahl} {et~al.}(2002){Grundahl}, {Briley}, {Nissen}, \&
1027: {Feltzing}}]{grundahl02}
1028: {Grundahl}, F., {Briley}, M., {Nissen}, P.~E., \& {Feltzing}, S. 2002, \aap,
1029: 385, L14
1030:
1031: \bibitem[{{Ivans} {et~al.}(2001){Ivans}, {Kraft}, {Sneden}, {Smith}, {Rich}, \&
1032: {Shetrone}}]{M5}
1033: {Ivans}, I.~I., {Kraft}, R.~P., {Sneden}, C., {Smith}, G.~H., {Rich}, R.~M., \&
1034: {Shetrone}, M. 2001, \aj, 122, 1438
1035:
1036: \bibitem[{{Ivans} {et~al.}(1999){Ivans}, {Sneden}, {Kraft}, {Suntzeff},
1037: {Smith}, {Langer}, \& {Fulbright}}]{M4}
1038: {Ivans}, I.~I., {Sneden}, C., {Kraft}, R.~P., {Suntzeff}, N.~B., {Smith},
1039: V.~V., {Langer}, G.~E., \& {Fulbright}, J.~P. 1999, \aj, 118, 1273
1040:
1041: \bibitem[{{James} {et~al.}(2004){James}, {Fran{\c c}ois}, {Bonifacio},
1042: {Carretta}, {Gratton}, \& {Spite}}]{james04b}
1043: {James}, G., {Fran{\c c}ois}, P., {Bonifacio}, P., {Carretta}, E., {Gratton},
1044: R.~G., \& {Spite}, F. 2004, \aap, 427, 825
1045:
1046: \bibitem[{{Karakas} \& {Lattanzio}(2003)}]{karakas03}
1047: {Karakas}, A.~I. \& {Lattanzio}, J.~C. 2003, Publ.\ Astron.\ Soc.\ Australia,
1048: 20, 279
1049:
1050: \bibitem[{{Kraft}(1994)}]{kraft94}
1051: {Kraft}, R.~P. 1994, \pasp, 106, 553
1052:
1053: \bibitem[{{Kurucz}(1993)}]{kurucz93}
1054: {Kurucz}, R. 1993, ATLAS9 Stellar Atmosphere Programs and 2 km/s grid.~Kurucz
1055: CD-ROM No.~13.~ Cambridge, Mass.: Smithsonian Astrophysical Observatory,
1056: 1993., 13
1057:
1058: \bibitem[{{Kurucz}(2006)}]{kurucz06}
1059: ---. 2006, http://kurucz.harvard.edu/LINELISTS/LINESMOL/
1060:
1061: \bibitem[{{Kurucz} {et~al.}(1984){Kurucz}, {Furenlid}, \& {Brault}}]{kurucz84}
1062: {Kurucz}, R.~L., {Furenlid}, I., \& {Brault}, J. 1984, {Solar flux atlas from
1063: 296 to 1300 NM} (National Solar Observatory Atlas, Sunspot, New Mexico:
1064: National Solar Observatory, 1984)
1065:
1066: \bibitem[{{Lambert} {et~al.}(1996){Lambert}, {Heath}, {Lemke}, \&
1067: {Drake}}]{lambert96}
1068: {Lambert}, D.~L., {Heath}, J.~E., {Lemke}, M., \& {Drake}, J. 1996, \apjs, 103,
1069: 183
1070:
1071: \bibitem[{{Lambert} \& {Luck}(1976)}]{lambert76}
1072: {Lambert}, D.~L. \& {Luck}, R.~E. 1976, The Observatory, 96, 100
1073:
1074: \bibitem[{{Lambert} {et~al.}(1995){Lambert}, {Smith}, {Busso}, {Gallino}, \&
1075: {Straniero}}]{lambert95}
1076: {Lambert}, D.~L., {Smith}, V.~V., {Busso}, M., {Gallino}, R., \& {Straniero},
1077: O. 1995, \apj, 450, 302
1078:
1079: \bibitem[{{Latham} {et~al.}(2002){Latham}, {Stefanik}, {Torres}, {Davis},
1080: {Mazeh}, {Carney}, {Laird}, \& {Morse}}]{latham02}
1081: {Latham}, D.~W., {Stefanik}, R.~P., {Torres}, G., {Davis}, R.~J., {Mazeh}, T.,
1082: {Carney}, B.~W., {Laird}, J.~B., \& {Morse}, J.~A. 2002, \aj, 124, 1144
1083:
1084: \bibitem[{{Lattanzio} {et~al.}(2004){Lattanzio}, {Karakas}, {Campbell},
1085: {Elliott}, \& {Chieffi}}]{lattanzio04}
1086: {Lattanzio}, J., {Karakas}, A., {Campbell}, S., {Elliott}, L., \& {Chieffi}, A.
1087: 2004, Memorie della Societa Astronomica Italiana, 75, 322
1088:
1089: \bibitem[{{Lattanzio} \& {Tout}(2006)}]{lattanzio06}
1090: {Lattanzio}, J.~C. \& {Tout}, C.~A. 2006, in EAS Publications Series, ed.
1091: T.~{Montmerle} \& C.~{Kahane}, 189--197
1092:
1093: \bibitem[{{Lawler} {et~al.}(2001{\natexlab{a}}){Lawler}, {Bonvallet}, \&
1094: {Sneden}}]{la}
1095: {Lawler}, J.~E., {Bonvallet}, G., \& {Sneden}, C. 2001{\natexlab{a}}, \apj,
1096: 556, 452
1097:
1098: \bibitem[{{Lawler} {et~al.}(2001{\natexlab{b}}){Lawler}, {Wickliffe}, {den
1099: Hartog}, \& {Sneden}}]{eu}
1100: {Lawler}, J.~E., {Wickliffe}, M.~E., {den Hartog}, E.~A., \& {Sneden}, C.
1101: 2001{\natexlab{b}}, \apj, 563, 1075
1102:
1103: \bibitem[{{Lee} {et~al.}(2006){Lee}, {L{\'o}pez-Morales}, \& {Carney}}]{lee06}
1104: {Lee}, J.-W., {L{\'o}pez-Morales}, M., \& {Carney}, B.~W. 2006, \apjl, 646,
1105: L119
1106:
1107: \bibitem[{{Lee}(1977)}]{lee77}
1108: {Lee}, S.~W. 1977, \aaps, 27, 367
1109:
1110: \bibitem[{{McClure} \& {Norris}(1977)}]{mcclure77}
1111: {McClure}, R.~D. \& {Norris}, J. 1977, \apjl, 217, L101
1112:
1113: \bibitem[{{Norris}(1981)}]{norris81a}
1114: {Norris}, J. 1981, \apj, 248, 177
1115:
1116: \bibitem[{{Otsuki} {et~al.}(2006){Otsuki}, {Honda}, {Aoki}, {Kajino}, \&
1117: {Mathews}}]{otsuki06}
1118: {Otsuki}, K., {Honda}, S., {Aoki}, W., {Kajino}, T., \& {Mathews}, G.~J. 2006,
1119: \apjl, 641, L117
1120:
1121: \bibitem[{{Prantzos} \& {Charbonnel}(2006)}]{prantzos06}
1122: {Prantzos}, N. \& {Charbonnel}, C. 2006, \aap, 458, 135
1123:
1124: \bibitem[{{Pritzl} {et~al.}(2005){Pritzl}, {Venn}, \& {Irwin}}]{pritzl05}
1125: {Pritzl}, B.~J., {Venn}, K.~A., \& {Irwin}, M. 2005, \aj, 130, 2140
1126:
1127: \bibitem[{{Raiteri} {et~al.}(1991){Raiteri}, {Busso}, {Picchio}, \&
1128: {Gallino}}]{raiteri91}
1129: {Raiteri}, C.~M., {Busso}, M., {Picchio}, G., \& {Gallino}, R. 1991, \apj, 371,
1130: 665
1131:
1132: \bibitem[{{Ram{\'{\i}}rez} \& {Cohen}(2003)}]{ramirez03}
1133: {Ram{\'{\i}}rez}, S.~V. \& {Cohen}, J.~G. 2003, \aj, 125, 224
1134:
1135: \bibitem[{{Renzini} \& {Fusi Pecci}(1988)}]{renzini88}
1136: {Renzini}, A. \& {Fusi Pecci}, F. 1988, \araa, 26, 199
1137:
1138: \bibitem[{{Sbordone} {et~al.}(2007){Sbordone}, {Bonifacio}, {Buonanno},
1139: {Marconi}, {Monaco}, \& {Zaggia}}]{sbordone06}
1140: {Sbordone}, L., {Bonifacio}, P., {Buonanno}, R., {Marconi}, G., {Monaco}, L.,
1141: \& {Zaggia}, S. 2007, \aap, 465, 815
1142:
1143: \bibitem[{{Smith}(1987)}]{smith87}
1144: {Smith}, G.~H. 1987, \pasp, 99, 67
1145:
1146: \bibitem[{{Smith}(2006)}]{smith06}
1147: ---. 2006, \pasp, 118, 1225
1148:
1149: \bibitem[{{Smith} \& {Mateo}(1990)}]{smith90}
1150: {Smith}, G.~H. \& {Mateo}, M. 1990, \apj, 353, 533
1151:
1152: \bibitem[{{Smith} \& {Norris}(1983)}]{smith83}
1153: {Smith}, G.~H. \& {Norris}, J. 1983, \apj, 264, 215
1154:
1155: \bibitem[{{Smith} {et~al.}(1997){Smith}, {Shetrone}, {Briley}, {Churchill}, \&
1156: {Bell}}]{smith97}
1157: {Smith}, G.~H., {Shetrone}, M.~D., {Briley}, M.~M., {Churchill}, C.~W., \&
1158: {Bell}, R.~A. 1997, \pasp, 109, 236
1159:
1160: \bibitem[{{Smith} {et~al.}(2000){Smith}, {Suntzeff}, {Cunha}, {Gallino},
1161: {Busso}, {Lambert}, \& {Straniero}}]{smith00}
1162: {Smith}, V.~V., {Suntzeff}, N.~B., {Cunha}, K., {Gallino}, R., {Busso}, M.,
1163: {Lambert}, D.~L., \& {Straniero}, O. 2000, \aj, 119, 1239
1164:
1165: \bibitem[{{Sneden}(1973)}]{moog}
1166: {Sneden}, C. 1973, \apj, 184, 839
1167:
1168: \bibitem[{{Sneden} {et~al.}(1998){Sneden}, {Cowan}, {Burris}, \&
1169: {Truran}}]{sneden98}
1170: {Sneden}, C., {Cowan}, J.~J., {Burris}, D.~L., \& {Truran}, J.~W. 1998, \apj,
1171: 496, 235
1172:
1173: \bibitem[{{Sneden} {et~al.}(1997){Sneden}, {Kraft}, {Shetrone}, {Smith},
1174: {Langer}, \& {Prosser}}]{sneden97}
1175: {Sneden}, C., {Kraft}, R.~P., {Shetrone}, M.~D., {Smith}, G.~H., {Langer},
1176: G.~E., \& {Prosser}, C.~F. 1997, \aj, 114, 1964
1177:
1178: \bibitem[{{Suntzeff} \& {Smith}(1991)}]{ss91}
1179: {Suntzeff}, N.~B. \& {Smith}, V.~V. 1991, \apj, 381, 160
1180:
1181: \bibitem[{{Sweigart} \& {Mengel}(1979)}]{sm79}
1182: {Sweigart}, A.~V. \& {Mengel}, J.~G. 1979, \apj, 229, 624
1183:
1184: \bibitem[{{Tomkin} \& {Lambert}(1983)}]{tomkin83}
1185: {Tomkin}, J. \& {Lambert}, D.~L. 1983, \apj, 273, 722
1186:
1187: \bibitem[{{Tomkin} \& {Lambert}(1999)}]{tomkin99}
1188: ---. 1999, \apj, 523, 234
1189:
1190: \bibitem[{{Travaglio} {et~al.}(2001){Travaglio}, {Gallino}, {Busso}, \&
1191: {Gratton}}]{travaglio01}
1192: {Travaglio}, C., {Gallino}, R., {Busso}, M., \& {Gratton}, R. 2001, \apj, 549,
1193: 346
1194:
1195: \bibitem[{{Vanture} \& {Wallerstein}(1992)}]{vanture92}
1196: {Vanture}, A.~D. \& {Wallerstein}, G. 1992, \pasp, 104, 888
1197:
1198: \bibitem[{{Ventura} \& {D'Antona}(2005{\natexlab{a}})}]{ventura05a}
1199: {Ventura}, P. \& {D'Antona}, F. 2005{\natexlab{a}}, \aap, 431, 279
1200:
1201: \bibitem[{{Ventura} \& {D'Antona}(2005{\natexlab{b}})}]{ventura05b}
1202: ---. 2005{\natexlab{b}}, \aap, 439, 1075
1203:
1204: \bibitem[{{Ventura} \& {D'Antona}(2005{\natexlab{c}})}]{ventura05c}
1205: ---. 2005{\natexlab{c}}, \apjl, 635, L149
1206:
1207: \bibitem[{{Wylie} {et~al.}(2006){Wylie}, {Cottrell}, {Sneden}, \&
1208: {Lattanzio}}]{wylie06}
1209: {Wylie}, E.~C., {Cottrell}, P.~L., {Sneden}, C.~A., \& {Lattanzio}, J.~C. 2006,
1210: \apj, 649, 248
1211:
1212: \bibitem[{{Yong} {et~al.}(2006{\natexlab{a}}){Yong}, {Aoki}, \&
1213: {Lambert}}]{mghsubaru}
1214: {Yong}, D., {Aoki}, W., \& {Lambert}, D.~L. 2006{\natexlab{a}}, \apj, 638, 1018
1215:
1216: \bibitem[{{Yong} {et~al.}(2006{\natexlab{b}}){Yong}, {Aoki}, {Lambert}, \&
1217: {Paulson}}]{rbpbsubaru}
1218: {Yong}, D., {Aoki}, W., {Lambert}, D.~L., \& {Paulson}, D.~B.
1219: 2006{\natexlab{b}}, \apj, 639, 918
1220:
1221: \bibitem[{{Yong} {et~al.}(2003{\natexlab{a}}){Yong}, {Grundahl}, {Lambert},
1222: {Nissen}, \& {Shetrone}}]{6752}
1223: {Yong}, D., {Grundahl}, F., {Lambert}, D.~L., {Nissen}, P.~E., \& {Shetrone},
1224: M.~D. 2003{\natexlab{a}}, \aap, 402, 985
1225:
1226: \bibitem[{{Yong} {et~al.}(2005){Yong}, {Grundahl}, {Nissen}, {Jensen}, \&
1227: {Lambert}}]{67522}
1228: {Yong}, D., {Grundahl}, F., {Nissen}, P.~E., {Jensen}, H.~R., \& {Lambert},
1229: D.~L. 2005, \aap, 438, 875
1230:
1231: \bibitem[{{Yong} {et~al.}(2003{\natexlab{b}}){Yong}, {Lambert}, \&
1232: {Ivans}}]{mghdwarf}
1233: {Yong}, D., {Lambert}, D.~L., \& {Ivans}, I.~I. 2003{\natexlab{b}}, \apj, 599,
1234: 1357
1235:
1236: \end{thebibliography}
1237:
1238: \begin{figure}
1239: \epsscale{0.8}
1240: \plotone{fig1.ps}
1241: \caption{Observed spectra (filled circles) for M4 L3413 (upper panel),
1242: M4 L4611 (middle panel), and M5 IV-81 (lower panel) near the 7800~\AA~Rb
1243: feature. Synthetic spectra with different Rb abundances are shown
1244: (the solid line represents the best fit).\label{fig:rbfit}}
1245: \end{figure}
1246:
1247: \clearpage
1248:
1249: \begin{figure}
1250: \epsscale{0.8}
1251: \plotone{fig2.ps}
1252: \caption{Observed spectra (filled circles) for M4 L1411 (upper panel),
1253: M4 L3624 (middle panel), and M5 IV-82 (lower panel) near the 4058~\AA~Pb
1254: feature. Synthetic spectra with different Pb abundances are shown
1255: (the solid line represents the best fit).\label{fig:pbfit}}
1256: \end{figure}
1257:
1258: \clearpage
1259:
1260: \begin{figure}
1261: \epsscale{0.8}
1262: \plotone{fig3.ps}
1263: \caption{[Rb/Fe] vs.\ [Fe/H] for M4 (filled circles) and M5 (open circles).
1264: Green triangles show NGC 6752 \citep{rbpbsubaru}, red
1265: squares represent $\omega$ Cen \citep{smith00}, blue asterisks are NGC 3201
1266: \citep{gonzalez98}, and magenta plus signs and upper limits
1267: represent field stars from
1268: \citet{tomkin99} and \citet{gratton94}. A representative error
1269: bar is shown. The abundances have been shifted onto the \citet{smith00}
1270: scale.\label{fig:rb}}
1271: \end{figure}
1272:
1273: \clearpage
1274:
1275: \begin{figure}
1276: \epsscale{0.8}
1277: \plotone{fig4.ps}
1278: \caption{[Rb/Y] vs.\ [Y/Fe] (upper), [Rb/La] vs.\ [La/Fe] (middle), and
1279: [Rb/Pb] vs.\ [Pb/Fe] (lower) for M4, M5, and NGC 6752.
1280: In each panel, the dotted line represents the mean value.
1281: \label{fig:new3}}
1282: \end{figure}
1283:
1284: \clearpage
1285:
1286: \begin{figure}
1287: \epsscale{0.8}
1288: \plotone{fig5.ps}
1289: \caption{[Rb/Zr] vs.\ [Fe/H]. The symbols are the same as in Figure
1290: \ref{fig:rb}. The abundances have been shifted onto the \citet{smith00}
1291: scale. A representative error bar is shown.\label{fig:rbzr}}
1292: \end{figure}
1293:
1294: \clearpage
1295:
1296: \begin{figure}
1297: \epsscale{0.8}
1298: \plotone{fig6.ps}
1299: \caption{[Pb/Fe] vs.\ [Fe/H] for M4 (filled circles) and M5 (open circles).
1300: Green triangles show NGC 6752, red
1301: squares represent M13, and blue asterisks are HD 141531 \citep{rbpbsubaru}.
1302: The magenta plus signs and upper limits represent field stars from
1303: \citet{sneden98} and \citet{travaglio01}. A representative error bar is
1304: shown.\label{fig:pb}}
1305: \end{figure}
1306:
1307: \clearpage
1308:
1309: \input{tab1}
1310: \input{tab2}
1311: \input{tab3}
1312: \input{tab4}
1313:
1314: \end{document}
1315:
1316: