0801.0752/ms.tex
1: %% This is a sample manuscript marked up using the
2: %% AASTeX v5.x LaTeX 2e macros.
3: 
4: %\documentclass[manuscript]{aastex}
5: %\documentclass[12pt,preprint]{aastex}
6: \documentclass[preprint2]{aastex}
7: %% \documentclass[preprint2,longabstract]{aastex}
8: 
9: %% If you want to create your own macros, you can do so
10: %% using \newcommand. Your macros should appear before
11: %% the \begin{document} command.
12: 
13: \newcommand{\vdag}{(v)^\dagger}
14: \newcommand{\myemail}{kvenn@uvic.ca}
15: \newcommand{\etal}{{\it et al.\,}}
16: \newcommand{\teff}{T$_{\rm eff}$\,}
17: \newcommand{\logg}{log~g\,}
18: 
19: %% You can insert a short comment on the title page using the command below.
20: 
21: %%\slugcomment{}
22: 
23: \shorttitle{UMPS and dust-gas separation}
24: \shortauthors{Venn \& Lambert}
25: 
26: %% This is the end of the preamble.  Indicate the beginning of the
27: %% paper itself with \begin{document}.
28: 
29: \begin{document}
30: 
31: %% LaTeX will automatically break titles if they run longer than
32: %% one line. However, you may use \\ to force a line break if
33: %% you desire.
34: 
35: \title{Could the Ultra Metal-poor Stars be Chemically Peculiar
36:   and Not Related to the First Stars?}
37: 
38: %% Use \author, \affil, and the \and command to format
39: %% author and affiliation information.
40: %% Note that \email has replaced the old \authoremail command
41: %% from AASTeX v4.0. You can use \email to mark an email address
42: %% anywhere in the paper, not just in the front matter.
43: %% As in the title, use \\ to force line breaks.
44: 
45: \author{K. A. Venn}
46: \affil{Department of Physics \& Astronomy, University of Victoria,
47:        Elliott Bldg, 3800 Finnerty Road, Victoria, BC, V8P 5C2, Canada}
48: \email{kvenn@uvic.ca}
49: \and
50: \author{David L. Lambert}
51: \affil{The W.J. McDonald Observatory, The University of Texas at Austin,
52:        RLM 15.308, Austin, TX, 78712}
53: \email{dll@astro.as.utexas.edu}
54: 
55: 
56: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
57: %% style, abstract will output a Received/Accepted line after the
58: %% title and affiliation information. No date will appear since the author
59: %% does not have this information. The dates will be filled in by the
60: %% editorial office after submission.
61: 
62: \begin{abstract}
63: 
64: Chemically peculiar stars define a class of stars that show
65: unusual elemental abundances due to stellar photospheric effects
66: and not due to natal variations. 
67: In this paper, we compare the elemental abundance patterns of the
68: ultra metal-poor stars with metallicities [Fe/H] $\sim -5 $ to 
69: those of a subclass of chemically peculiar stars.   These
70: include post-AGB stars, RV Tauri variable stars, and the Lambda Bootis 
71: stars, which range in mass, age, binarity, and evolutionary status, yet 
72: can have iron abundance determinations as low as [Fe/H] $\sim -5$.
73: These chemical peculiarities are interpreted as due to the separation 
74: of gas and dust beyond the stellar surface, followed by the accretion 
75: of dust depleted-gas.  
76: Contrary to this,
77: the elemental abundances in the ultra metal-poor stars are thought  
78: to represent yields of the most metal-poor supernova
79: and, therefore, observationally constrain the earliest stages of
80: chemical evolution in the Universe.  
81: Detailed chemical abundances are now available for HE1327-2326 
82: and HE0107-5240, the two extreme ultra metal-poor stars
83: in our Galaxy, and for HE0557-4840, another ultra metal-poor star 
84: found by the Hamberg/ESO survey.  
85: There are interesting similarities in their abundance ratios 
86: to those of the chemically peculiar stars, 
87: e.g., the abundance of the elements in their photospheres are
88: related to the condensation temperature of that element. 
89: If HE1327-2326 and HE0107-5240 are ultra metal-poor due 
90: to the preferential removal of metals by dust grain formation or
91: dilution through the accretion of metal-poor interstellar gas,
92: then their CNO abundances suggest true metallicities of [$X$/H] $\sim -2$
93: rather than their present metallicities of [Fe/H] $\leq -5$,
94: and, thus their status as truly ultra metal-poor stars is called into question.
95: The initial abundance for HE0557-4840 would be [$X$/H] $\ge -4$.  
96: It is important to establish the nature of these stars since they are
97: used as tests of the early chemical evolution of the Galaxy, yet if 
98: they are chemically peculiar, then those tests should be focused on 
99: stars in the metal-poor tail of the Galactic halo distribution.
100: 
101: Many, but not all, chemically peculiar stars show a mid-infrared 
102: excess from the circumstellar dust.   We examine the JHK fluxes for 
103: the ultra metal-poor stars but find no excesses.   
104: A more important test of the stars' status as
105: chemically peculiar is provided by the elemental abundances 
106: of sulphur and/or zinc. These two elements have low condensation 
107: temperatures and do not form dust grains easily; furthermore, 
108: the chemically peculiar stars universally show sulphur and zinc 
109: to be undepleted or nearly so. 
110: We show that near-infrared lines of S\,{\sc i} offer a promising test of 
111: the possibility that HE1327-2326 may be chemically peculiar.
112: Although there are some parallels between the compositions of the
113: ultra metal-poor stars and chemically peculiar stars, a definitive 
114: ruling on whether the former are chemically peculiar requires additional
115: information.
116: 
117: 
118: \end{abstract}
119: 
120: \keywords{stars: abundances, chemically peculiar, individual (HE1327-2326);
121:           cosmology: early universe }
122: 
123: %% From the front matter, we move on to the body of the paper.
124: %% In the first two sections, notice the use of the natbib \citep
125: %% and \citet commands to identify citations.  The citations are
126: %% tied to the reference list via symbolic KEYs. The KEY corresponds
127: %% to the KEY in the \bibitem in the reference list below. We have
128: %% chosen the first three characters of the first author's name plus
129: %% the last two numeral of the year of publication as our KEY for
130: %% each reference.
131: 
132: 
133: %% Authors who wish to have the most important objects in their paper
134: %% linked in the electronic edition to a data center may do so by tagging
135: %% their objects with \objectname{} or \object{}.  Each macro takes the
136: %% object name as its required argument. The optional, square-bracket 
137: %% argument should be used in cases where the data center identification
138: %% differs from what is to be printed in the paper.  The text appearing 
139: %% in curly braces is what will appear in print in the published paper. 
140: %% If the object name is recognized by the data centers, it will be linked
141: %% in the electronic edition to the object data available at the data centers  
142: 
143: \section{Introduction}
144: 
145: A fossil record of the earliest episodes of stellar nucleosynthesis 
146: in the Universe and Galaxy should be revealed by the compositions of 
147: the most metal-poor Galactic stars
148: (e.g., Tumlinson 2007a, 2007b; Tominaga \etal 2007; Umeda \& Nomoto 2003). 
149: The lure of this revelation
150: has driven the search to find and analyse such Rosetta stones. 
151: A great leap forward  was achieved recently by the 
152: discovery of two stars with  iron abundances [Fe/H] $< -5.3$
153: (Christlieb \etal 2002; Frebel \etal 2005),
154: a limit about 1.5 dex below the abundance of the previously known 
155: most metal-poor star. A third star with [Fe/H] $\simeq -4.8$ also
156: beats the previous lower bound (Norris \etal 2007). 
157: Prior to these remarkable discoveries, the most Fe-poor stars known  
158: were HR~4049 and HD~52961 with [Fe/H] $\simeq -4.8$ 
159: (e.g., Waelkens \etal 1991).   However, these and slightly more 
160: iron-rich examples were dismissed - correctly - as irrelevant 
161: to the issue of early stellar nucleosynthesis because they are 
162: `chemically peculiar', i.e., their present surface compositions 
163: are far removed from their initial compositions.
164: In particular, their compositions reflect that of gas from which
165: refractory elements have been removed to varying degrees by a process
166: dubbed `dust-gas separation'.
167: The existence of HR~4049 and HD~52961 has led us to reexamine the 
168: question of whether the recently discovered ultra metal-poor
169: stars may themselves be chemically peculiar. 
170: 
171: Two of the three ultra metal-poor stars in question are 
172: HE1327-2326 and HE0107-5240.
173: HE1327-2326 was discovered by Frebel \etal (2005) and abundance analyses
174: have been described by Aoki \etal (2006), Frebel \etal (2006), 
175: Collet \etal (2006), and most recently by Frebel \& Christlieb (2007). 
176: The latter analysis based on
177: the highest SNR spectra yields abundances for 11 elements and
178: upper limits for an additional nine elements: the new iron abundance
179: at [Fe/H] $=-5.9$
180: is even lower than the previous determination.
181: %%
182: HE0107-5240 was discovered by Christlieb \etal (2002) and abundance
183: analyses have been reported by Christlieb \etal (2004), Bessell \etal 
184: (2004), and Collet \etal (2006) with the latter suggesting
185: [Fe/H]$\simeq -5.6$. 
186: %%
187: The third star HE0557-4840 was discovered and analysed by 
188: Norris \etal (2007): [Fe/H] $= -4.8$ places it between 
189: the two ultra metal-poor stars (HE1327-2326 and HE0107-5240) 
190: and the lower boundary of the metal-poor tail of Galactic stars. 
191: 
192: A marked characteristic of these three discoveries is that some
193: abundance ratios are uncharacteristic  of metal-poor stars
194: of higher Fe abundance. Notably, the stars are C-rich for their
195: Fe abundance, i.e., [C/Fe] $= 3.7$ for HE1327-2326, and also
196: [N/Fe] $= 4.1$ and [O/Fe] $= 3.4$ (Frebel \& Christlieb 2007). This
197: property of unusual abundance ratios is shared in a qualitative sense with 
198: HR~4049 and HD~52961. 
199: 
200: In the following sections, we briefly review the classes of known 
201: chemically peculiar stars affected by dust-gas separation. 
202: Then, we discuss if the three stars HE1327-2326,  HE0107-5240, 
203: and HE0557-4840 are chemically peculiar rather than true 
204: ultra metal-poor stars. 
205: In the final sections, we discuss possible tests of the hypothesis 
206: that ultra metal-poor stars may be chemically peculiar.
207: 
208: %The idea that chemical peculiarities may have
209: %influenced their compositions has been previously aired by
210: %Christlieb et al. (2002, 2004), however the discussion was
211: %limited to a few specific differences between HE0107-5240 and 
212: %those of HR~4049 and HD~52961 which could have other explanations. 
213: %%
214: %Diffuse interstellar clouds also offer an example of gas deficient
215: %in refractory elements (Savage \& Sembach 1996). If accreted 
216: %interstellar gas were to overwhelm a  star's natal composition,
217: %the star would assume aspects of a chemically peculiar star.
218: %Christlieb et al., Aoki et al. (2006), and Shigeyama (2003) among
219: %others discuss this possibility. 
220: 
221:  
222: 
223: \section {Separation of gas and dust and chemically peculiar stars}
224: 
225: The chemically peculiar stars in question are those whose
226: atmospheres betray the operation of the dust-gas separation process. 
227: In gas of a sufficiently low temperature, dust condenses out and 
228: the gas is depleted in those elements that form the dust. 
229: The local interstellar gas, for example, displays such depletion 
230: (Savage \& Sembach 1996). 
231: The composition of the gas in a dust-gas mixture depends on several 
232: primary factors including the initial composition of the gas, the total
233: pressure, and the  history of the gas-dust mixture. 
234: If a star were to then accrete gas, preferentially over dust,
235: and the accreted gas were to comprise a major fraction of the
236: stellar photosphere,  a star with striking abundance anomalies 
237: results.   This scenario is one version of how a star could develop
238: chemical peculiarities from dust-gas separation (or ``dust-gas winnowing'').
239: Anomalies plausibly attributable to dust-gas separation have now been
240: reported for three kinds of stars: Lambda Bootis stars,
241: post-AGB A-type spectroscopic binaries, and RV Tauri variables. 
242: 
243: The Lambda Bootis stars are main sequence stars, ranging from A-
244: to mid F-types, and found at all evolutionary phases from very young 
245: (e.g., they are found in young open clusters, Gray \& Corbally 1998), 
246: to the end of the main-sequence (Kamp \& Paunzen 2002, 
247: Andrievsky \etal 2002). 
248: They are expected to have a solar-like composition, but have 
249: long been known for their significant metal-deficiencies 
250: (Morgan, Keenan, \& Kellman 1943; Burbidge \& Burbidge 1956; 
251: Baschek \& Searle 1969).  These stars have effective temperatures 
252: of from about 6500 to 9500~K and surface gravities \logg $\simeq 4$. 
253: Our (Venn \& Lambert 1990) abundance analysis of three stars, 
254: including the eponym, showed that the compositions of the stars 
255: could be attributed to accretion of circumstellar gas, without dust.
256: The iron deficiency in the case of $\lambda$~Boo was [Fe/H] $\simeq -2$, 
257: and slightly less severe for 29~Cyg, with normal abundances of 
258: C, N, O and S for both stars, in agreement with the expectations
259: for an atmosphere contaminated with  dust-free gas (see Figure~1).  
260: Even Vega, a `standard' A0 star, was shown to be a mild Lambda Bootis 
261: star (Venn \& Lambert 1990, Lemke \& Venn 1996).   Vega also shows an
262: infrared excess due to a dusty circumstellar disk (Su \etal 2005;
263: Aumann \etal 1984).
264:   
265: The shallow convective envelope of early A-type stars is 
266: a key factor in the creation and maintenance of the abundance 
267: anomalies in the {\it diffusion/accretion model} 
268: (Turcotte \& Charbonneau 1993).
269: Abundance anomalies can persist ($\sim$10$^6$ yr), 
270: even after dispersal of the circumstellar
271: dust and gas and, hence, removal of the infrared excess contributed by 
272: the dust.   Thus, not all Lambda Bootis stars show an infrared excess; 
273: Paunzen \etal (2003) estimate that 23\% of {\it bona fide} Lambda Bootis 
274: stars show evidence for circumstellar matter.
275: On the other hand, the circumstellar matter may be of interstellar 
276: origin (Kamp \& Paunzen 2002, G\'{a}sp\'{a}r \etal 2007).   
277: Movement of a star through the denser parts of the interstellar medium 
278: can create a bow shock which heats the interstellar dust causing an 
279: infrared excess.  Meanwhile, radiation pressure from the star repels the 
280: grains, while gas is accreted onto the stellar surface.
281: With return of the star to passage through low density gas, the 
282: infrared excess dissipates and accretion ceases. 
283: This alternate theory for the origin of the Lambda Bootis stars 
284: implies the chemical anomalies are transient, but could help to
285: explain why the phenomenon is seen in such a wide range of 
286: main-sequence stars.
287: 
288: Abundance anomalies should not survive the transition from the 
289: main sequence to the giant branch though.  The deep convective 
290: envelope of giant stars will surely dilute abundance anomalies 
291: beyond recognition.   
292: Yet, a metal deficiency of even greater severity can be found among 
293: post-AGB stars in spectroscopic binaries (Waelkens \etal 1991;
294: Van Winckel 2003). These stars, like HR~4049 and HD~52961 discussed 
295: above, are supergiants with \teff in the range of 6000 to 7600~K and 
296: surface gravities \logg $\simeq 1$. One must suppose that a new
297: episode of dust-gas separation led to these abundance anomalies.
298: %
299: 
300: 
301: The original quartet of post-AGB binaries (Van Winckel \etal 1995)
302: comprised HR~4049, HD~44179, HD~52961, and BD~+39~4926.  Their [Fe/H] 
303: values range from $-3.0$ to $-4.8$, but they show an abundance pattern 
304: reminiscent of interstellar gas, i.e., quasi-solar abundances 
305: of C, N, O, S, and Zn, but severe underabundances of, for example, 
306: Si, Ca, and Fe (see Figure~1). 
307: The pattern shows that it is the former set of elements that define the
308: initial composition of these stars and not the latter set.
309: Gas is thought to be accreted onto the star from a circumbinary
310: disk, while radiation pressure exerted by the star on the dust grains
311: inhibits the accretion of dust by the star and may also promote a 
312: separation of dust and gas in the disk.   The dusty disk is betrayed 
313: by an infrared excess:
314: HD~44179, also known as the Red Rectangle, is a rather special 
315: proto-planetary nebula with a striking infrared excess.  On the other hand, 
316: BD~+39~4926 lacks an infrared excess.   Shallow convective envelopes in 
317: these extended stars are presumably a key factor in the appearance 
318: of their huge abundance anomalies.  
319: 
320: Chemical peculiarities of the post-AGB stars are presumably developed 
321: from less extreme peculiarities seen in their immediate progenitors, 
322: the RV Tauri variables.  Discovery of this third category of star 
323: displaying the marks of dust-gas winnowing began with the analysis 
324: of the RV Tauri variable IW~Car (Giridhar \etal 1994). 
325: A RV Tauri star is a post-AGB star with an infrared excess 
326: (first noted by Gehrz 1972).  
327: Subsequent analyses (Giridhar \etal 2005) showed that the effects of 
328: dust-gas separation among RV Tauri stars appear limited to the warmer 
329: stars (\teff $> 4500$ K) and to the stars with intrinsic metallicities 
330: [Fe/H] $\geq -1$.   Affected stars have \teff $\simeq 4500 - 6500$~K 
331: (hotter stars fall outside the instability strip and appear as 
332: non-variable post-AGB stars) and \logg $\simeq 0$ to 1. 
333: Gonzalez \& Lambert (1997) also discussed the potential importance of
334: the composition of the photosphere (e.g., C/O ratio), and environment
335: (e.g., field vs globular cluster stars).
336: 
337:  The reasons for
338: these effective temperature and metallicity
339:  boundaries are not entirely clear, but the cooler stars possess more
340: extensive convective envelopes that dilute the accreted gas.  
341: Dust-gas winnowing may be impaired in low metallicity gas where the dust 
342: to gas ratio is necessarily lower. 
343: The winnowing site is again presumed to be a circumbinary disk; 
344: there is increasing evidence that the affected stars are spectroscopic 
345: binaries (Van Winckel 2007).  How the gas is captured onto the star 
346: from the circumbinary disk in the presence of a stellar wind remains 
347: an unsolved problem, as it does for the post-AGB stars in binaries.
348: 
349: A signature of a star that is affected seriously by dust-gas winnowing 
350: is a correlation between an element's abundance and the predicted
351: condensation temperature $T_{cond}$ (or the abundance in interstellar gas).
352: Estimates of $T_{cond}$ depend on the  initial composition and pressure
353: in a gas and the assumption that cooling of the gas and condensation
354: of grains occurs under equilibrium conditions. Lodders (2003) provides
355: a comprehensive discussion of $T_{cond}$ estimates for all elements
356: for the solar (O-rich) composition. In the post-AGB stars, and in some 
357: Lambda Bootis and RV Tauri stars, the
358: abundance [$X$/H] is smoothly correlated with $T_{cond}$ (see Figure~1).
359: 
360: %In Figure 1, we show the abundances [$X$/H] with condensation 
361: %temperature $T_{cond}$ for a sample of chemically peculiar stars.  These 
362: %include; $\lambda$ Boo (Venn \& Lambert 1990),
363: %UY CMa (Giridhar \etal 2005), a typical RV Tauri variable, 
364: %and HR~4049 and HD~52961 (Waelkens \etal 1991, van Winckel \etal
365: %1992, Takeda \etal 2002), samples of the A-type post-AGB stars.
366: %Condensation temperatures are from calculations by Lodders (2003) 
367: %assuming a solar system composition.
368: %
369: %{\bf I suggest that Figure 1 not include an ultra metal-poor star
370: %but be limited to the stars mention in the above paragraph.} 
371: 
372: 
373:  
374: In some Lambda Bootis and RV Tauri stars, 
375: [$X$/H]  shows no obvious trend with
376: the  $T_{cond}$ (e.g., EQ~Cas, Giridhar \etal 2005).
377: Rao \& Reddy (2005) brought order to chaos by noting instead 
378: that the [$X$/H] for EQ~Cas were correlated with the ionization potential 
379: of the neutral atom (`the first ionization potential' or FIP). 
380: This raises the intriguing possibility that an alternative
381: or additional process is operating in this one star. 
382: The FIP effect is a well known  phenomena in the solar corona 
383: thought to reflect the greater ease with which 
384: ions (low FIP elements) rather than neutral atoms (high FIP elements) 
385: are fed from the cool chromosphere into the corona.
386: Perhaps, EQ~Cas is a star where there is a selective
387: feeding of the stellar wind. Thus,  
388: in single stars, the {\sl stellar wind} may control the abundance 
389: anomalies but where the variable is in a  binary the {\sl circumbinary
390: disk} may exert control. 
391: 
392: Another possibility for affecting the [$X$/H] vs $T_{cond}$ correlation 
393: (at high T$_{cond}$) is a competition between accretion of metal free
394: gas and chemical separation due to, e.g., diffusion, gravitational
395: settling, radiative acceleration, or rotational mixing above the
396: convective zone.   The diffusion/accretion model for Lambda Bootis stars
397: by Turcotte \& Charbonneau (1993) suggests timescales for chemical
398: separation can vary by element, thus complicating the abundance
399: pattern established earlier by accretion.
400: %%
401: The conclusion must be that the dust-gas separation mechanism
402: does not have the same effects on the elemental abundances in
403: the various stars where it appears to operate.  
404: Either the mechanism itself, and/or the re-accretion phase of 
405: the dust-depleted gas, and/or additional processes that affect
406: chemical separation impact the observed chemical abundance pattern.
407: 
408: In summary, 
409: stars with abundance anomalies attributable to dust-gas separation
410: are seen in several parts of the HR-diagram. 
411: Associated properties of affected stars, such as age, mass, 
412: binarity, or evolutionary status,
413: do not appear to be uniform across the cases.  
414: Details in the observed abundance patterns seem to vary
415: between the cases, particularly for elements with
416: high T$_{cond}$, possibly due to the mechanism
417: itself or additional processes.
418: An accompanying infrared excess may be a warning bell, but it is 
419: not a required signature of dust-gas separation.
420: However, to date,  the inferred intrinsic metallicities 
421: of affected stars are one-tenth solar or greater. 
422: 
423: \section{Examination of dust-gas separation in the ultra metal-poor stars}
424: 
425: As the discussion turns to the ultra metal-poor stars, we  emphasize
426: the importance of the assumptions made in the determination of the $T_{cond}$ 
427: values.    Throughout our discussion, the $T_{cond}$ values are for a
428: solar composition gas.    Thus, the values are determined for a gas that 
429: is O-rich.   
430: One of the ultra metal-poor stars (HE0107-5240)
431: is presently severely C-rich, a second (HE1327-2326) has C/O $\simeq 1$,
432: and the C/O ratio is unknown for the third (HE0557-4840). 
433: Condensation of grains from C-rich gas provides C-rich solids 
434: (e.g., graphite and carbides) and a different set of condensation
435: temperatures. 
436: 
437: Calculations by Lodders \& Fegley (1995, 1999) have discussed the
438: condensation of grains in C-rich gas. 
439: For a gas with C/O $>$4, 
440: at a pressure  representative of a circumstellar region in
441: which dust forms via equilibrium chemistry (P$\sim 10^{-7}$ bar), 
442: the condensation temperatures $T_{cond}$ for graphite, TiC, and SiC 
443: (the first three condensates) 
444: are approximately 2000, 1560, and 1390 K, respectively. 
445: To lower temperatures, the condensation sequence is Fe, AlN, and CaS.
446: 
447: However, the $T_{cond}$'s are not the whole story. While Ti is 
448: effectively removed from the gas at temperatures below its $T_{cond}$, Si's
449: removal is constrained  by the fact that gaseous SiS is
450: resistant to condensation. The $T_{cond}$ for graphite but not for
451: TiC and SiC declines as C/O is decreased to near unity.
452: Condensation temperatures for Al, Mg,  Ca,  and
453: Fe are all several hundred degrees less than that of Ti. 
454: Thus, if the effective
455: $T_{cond}$ in the grain forming region is 1500~K or so, dust-gas separation
456: in a C-rich environment
457: should provide a pronounced Ti deficiency and only for much lower
458: temperatures will deficiencies for other elements be appreciable.
459: Finally, it is possible that dust-gas separation may remove sufficient 
460: carbon as graphite  
461: to lower the C/O ratio towards unity
462:  (Lodders \& Fegley 1999).    
463: 
464: In our figures, we adopt the T$_{cond}$ values determined for a 
465: solar (O-rich) composition gas, but discuss possible deviations due to the 
466: C/O ratio. 
467: 
468: 
469: \subsection{Has dust-gas separation modified the composition of
470: HE1327-2326?}
471: 
472: Our discussion of HE1327-2326's composition is based on the 
473: abundance analysis by Frebel \& Christlieb (2007), which is drawn from new, 
474: high signal-to-noise VLT/UVES spectra, but confirms and extends previous
475: detailed analyses (Aoki \etal 2006; Frebel \etal 2006).  The abundances
476: are plotted as [$X$/H] versus $T_{cond}$ in Figures~2 \& 3, 
477: and assume the star is a subgiant
478: (\teff = 6180~K, \logg = 3.7; though our conclusions would be the same,
479: and the following discussion negligibly affected, were we to adopt the abundances 
480: assuming the star to be a dwarf).
481: While a classical model atmospheres LTE analysis  is performed,
482: predicted corrections are included for 
483: effects of stellar granulation.  We also show abundances 
484: [$X$/H] for two extreme post-AGB stars (HR~4049 and HD~52961), the 
485: RV Tauri variables (HP~Lyr and UY~CMa), 
486: and two Lambda Bootis  stars (29 Cyg and HD 106223). 
487: 
488: There is  a resemblance between the compositions of
489: HE1327-2326 and the chemically peculiar stars; 
490: as T$_{cond}$ increases, the elemental depletion increases
491: relative to solar.
492: However, HE1327-2326 is unique because of its extreme [Mg/Fe] 
493: ratio (see Figure 2) which is not seen in any of the chemically
494: peculiar stars.  Iron appears to be an outlier, since it
495: is the most underabundant element in this star, which is rarely 
496: the case for the RV Tauri or Lambda Bootis stars. 
497: %%
498: Notably, S and Zn, important elements of low condensation temperature,
499:  have not been detected in  HE1327-2326.  
500: The S and Zn upper-limits provide no useful constraints
501: on HE1327-2326, although 
502: an abundance estimate for S may be possible in the future 
503: (see below).
504: 
505: Were we to insist that the abundance pattern for the RV Tauri 
506: star HP~Lyr, with its smooth trend in [$X$/H] vs T$_{cond}$, 
507: be a fair template for testing the suggestion that dust-gas 
508: separation has affected HE1327-2326, we would be bound to 
509: note the scatter in Figure~2 for $T_{cond} \geq
510: 1200$ K. In particular, the low abundance of Fe relative to
511: Al and Ti in HE1327-2326 is the opposite in HP~Lyr.   
512: It seems unlikely that a change in the applied 3D corrections 
513: for the (as yet) untested models of stellar granulation can 
514: reverse this trend. The 3D
515: corrections are small for these elements and all of the same sign 
516: with an element-to-element scatter of less than 0.2 dex. 
517: Estimates of non-LTE corrections
518: assembled by Aoki \etal (2006) for the 1D model 
519: also aggravate the situation in that the corrections raise 
520: the Al abundance by 0.4 dex relative to Fe. 
521: The non-LTE corrections applicable to the 3D model are unknown.   
522: 
523: HP Lyr (and the other RV Tauri comparison star, UY CMa,
524: in Figure~2) may be an imperfect template for HE1327-2326.
525: %%
526: Perhaps, {\it crucially}, the $T_{cond}$
527: estimates are based on a solar composition and, in particular,
528: on the fact that the Sun is O-rich (i.e., O/C $>1$). The measured
529: abundances of C and O  show that HE1327-2326 is nominally
530: C-rich  now as a subgiant and most likely the C/O ratio has been
531: depressed in evolution to the subgiant branch. (The C abundance is a mere
532: 0.06 dex greater than the O abundance, a difference less than the
533: errors of measurement.) 
534: %%
535: However, an alternative scenario, dust-gas separation in a C-rich environment 
536: also appears to fail to explain the observed composition.
537: As noted above, a signature of such a separation should be an
538: appreciable underabundance of Ti, which is not seen in Figure~2
539: relative to the other elements.
540: 
541: A comparison to the Lambda Bootis stars is more intriguing.   
542: Other than the large [Mg/Fe] ratio already noted as peculiar to
543: HE1327-2326, the relative abundances of Mg, Ca, and Ti to each
544: other (all high T$_{cond}$ elements) are similar to the Lambda Bootis 
545:  star HD106223 (see Figure~3).
546: %%
547: If the mechanism for dust-gas 
548: separation is more similar between HE1327-2326 and the Lambda Bootis
549: stars, possibly complicated by diffusion or another process (see Section 2),
550: then these stars may be a better comparison template.
551: In Figure~3, both of the Lambda Bootis stars shown have similar 
552: depletions of Ca, Ti, and Fe.  Unfortunately, Al is not available
553: for either Lambda Bootis star.
554: Al, with its very high T$_{cond}$ is 
555: severely depleted in the RV Tauri variables, but less so in 
556: HE1327-2326.
557: 
558: %Similar depletions of Ca, Al, and Fe have also been seen in
559: %other post-AGB stars with dusty disks; Maas \etal (2005)
560: %find 10 objects with an elemental abundance pattern showing 
561: %dust-gas separation, although these objects show
562: %Ti as significantly more depleted.   
563: 
564: If HE1327-2326 has undergone dust-gas separation, we can
565: estimate the star's intrinsic metallicity from the C, N, and O
566: abundances:
567: an initial metallicity of about $-2.0$ by the 3D-corrected C, N, and O
568: abundances but about $-1.3$ without these corrections.
569: %%
570: The 3D corrections are large ($\simeq-0.7$ dex) for C, N, and O because
571: their abundances are derived from molecules (CH, NH, and OH) whose
572: formation is greatly enhanced by the presence of cooler regions
573: in 3D models. At an initial metallicity of about $-1.3$, the star resembles
574: some post-AGB  and RV Tauri stars.
575: 
576: If the initial metallicity of HE1327-2326
577:  were in fact either [Fe/H] = $-2.0$ or $-1.3$,
578: this has a small impact on our expectation of the initial abundance 
579: ratios for the other elements.  For example, [Ca/Fe] $\simeq +0.3$ for 
580: {\it normal} metal-poor stars with [Fe/H] = $-2$, and this is
581: approximately seen in Figure~2.   
582: A similar ratio  can be expected of Mg and Ti.  
583: While [Ti/Fe] is slightly larger than expected, the uniquely high 
584: [Mg/Fe] ($\simeq +2$)
585:  ratio is quite out of line with expectations 
586: for normal metal-poor stars.
587: Lastly, in normal metal-poor stars, [Al/Fe] $<0$,
588: thus the high Al point is also peculiar to HE1327-2326.
589: 
590: 
591: Inspection of the [$X$/H] for HE1327-2326 (and the other two stars) shows
592: no evidence for a relationship between [$X$/H] and the FIP.
593: 
594: \subsection{Has dust-gas separation modified the composition of
595: HE0107-5240?}
596: 
597: The composition of HE0107-5240 was determined by Christlieb \etal
598: (2004) and Bessell \etal (2002) who found the star to be
599: a giant with $T_{\rm eff} = 5100$ K and \logg =2.2. Their
600: analyses adopted a classical (1D) atmosphere and LTE for the
601: atmosphere and line formation. The derived abundances  corrected
602: to those for a 3D atmosphere by Collet \etal (2006) with retention
603: of the  assumption of LTE  are plotted in Figures~4 \& 5.  
604: 
605: The abundance pattern for HE0107-5240 is more similar to the chemically
606: peculiar stars than was HE1327-2326 discussed above.  The [Mg/Fe] ratio
607: is similar to that of the chemically peculiar stars, as well as normal
608: metal-poor stars.   Ca and Ti also show depletions with respect to Fe
609: that are in fair agreement with those of the RV Tauri stars.   
610: The upper-limits to Zn and S
611: provide no useful constraint for examining the dust-gas scenario, 
612: however the very low upper-limit for Al is consistent with the
613: predictions.
614: 
615: 
616: 
617: This star is presently C-rich; C/O (by number) 
618: is about 20 for the 1D analysis and about 6 for the 3D analysis. 
619: An initial metallicity of $\simeq -4$ is crudely indicated by the 3D 
620: abundances of N and O, but $\simeq -2$  by the 1D C abundance.   
621: The result of the 1D to 3D corrections is, as indicated above, a large
622: reduction in the C, N, and O abundances, all derived from
623: diatomic molecules (CH, C$_2$, NH, CN, and OH). In 1D, the
624: abundances are larger than those plotted by about 1.2, 1.1, and 0.7
625: dex for C, N, and O, respectively.  It is an interesting point, possibly
626: one of concern, that NH and CN (assuming the C abundance
627: from CH and C$_2$) give a consistent N abundance using the 1D
628: atmosphere but inconsistent abundances (difference is 0.9 dex)
629: using the 3D atmosphere. 
630: The [C/O] ratio in HE0107-5240 
631: is similar to the carbon-enhanced metal-poor stars
632: which also range in metallicity from [Fe/H] = $-2$ to $-4$
633: (Sivarani \etal 2006). 
634: 
635: 
636: The appropriateness of using $T_{cond}$ in Figures~4 \& 5 is decidedly
637: questionable due to the star being very C-rich. 
638: If the star was and is a single star and its evolution  from the
639: main sequence to its present status as a giant even approximately
640: follows expectation, the star must have begun life even more C-rich. 
641: Then, if dust-gas separation occurs in either a protostellar, 
642: circumstellar or wind environment, it is the $T_{cond}$'s for 
643: C-rich gas that  are the relevant quantities. 
644: As noted above, a significant deficiency of Ti is
645: expected for the star accreting substantial amounts of gas.  
646: On the other hand, if the star has accreted significant
647: amounts of material from the interstellar medium of the established Galaxy,
648: including the present diffuse interstellar medium, the dust-gas separation
649: was likely to have occured or is occuring in an O-rich medium and,
650: then Figures~4 \& 5 use approximately the appropriate $T_{cond}$ values. 
651: Accretion of interstellar gas that is O-rich requires then an explanation
652: of how HE0107-5240 is so C-rich and metal-poor.
653: 
654: 
655: 
656: \subsection{Has dust-gas separation modified the composition of
657: HE0557-4840?}
658: 
659: The composition is taken from Norris \etal (2007) who determine
660: that the star is a giant and offer  spectroscopic and photometric
661: estimates of the effective temperature of 5100 K and 4900 K,
662: respectively. The surface gravity is given as \logg = 2.2. 
663: The mean of their LTE abundances for a 1D atmosphere are plotted
664: in Figure~6. Corrections for adoption of a  3D model atmosphere
665: are small, except for C (and the upper-limits on N and O), 
666: and are plotted in Figure~7. 
667: The C/O ratio is undetermined for this star.
668: 
669: 
670: If the initial composition were that suggested by the C abundance 
671: and N upper-limit for a 1D atmosphere, the initial [Fe/H] $\simeq -3$.  
672: Taking the 3D corrections into account,  
673: the initial metallicity is likely $\simeq -4$.
674: This lower initial abundance also reduces the scatter in the
675: [$X$/H] values, with a hint that [$X$/H] is independent of $T_{cond}$.  
676: 
677: 
678: There is no pattern in the elemental abundances for HE0557-4840 that
679: would clearly suggest dust-gas separation.   The [$X$/H] values for nearly
680: all elements are flat, ranging from $-4.0 > [$X$/H] > -5.5$, and
681: therefore showing similar depletions from a solar composition gas
682: in nearly all elements.  
683: This might be unexpected since the normal metal-poor stars 
684: with [Fe/H] $\sim -4$ show higher $\alpha$-element abundance ratios, 
685: and usually an extremely wide range in the $r$-process elements
686: (McWilliam 1997).   However, the [$\alpha$/Fe]
687: abundances in HE0557-4840 do not fall outside  the range of normal
688: metal-poor star abundances (e.g., [Mg/Fe] and [Ti/Fe] are
689: within expectations).
690: 
691: 
692: %%%%% Why does this next line matter if this is not a C-rich star?
693: %
694: %An exceptional Ti deficiency,
695: %an expected mark of dust-gas separation in C-rich gas, is
696: %also not evident from Figure 4.
697: %
698: %%%%%% What is the purpose of this sentence in here?
699: %
700: %The C/O ratio is unknown;
701: %the upper limit on the O abundance from the O\,{\sc i} 7770 \AA\ lines 
702: %is only slightly less than solar. Spectroscopic coverage needs to be
703: %extended into the ultraviolet to cover the OH bands.
704: 
705: \subsection{Other Metal-Poor Stars with [Fe/H] $\le -3.5$ }
706: 
707: It is natural to ask how ubiquitous the dust-gas separation pattern
708: is amongst the extremely metal-poor stars with detailed elemental
709: abundance determinations.    An examination of the literature for
710: stars with $-3.5 \ge$ [Fe/H] $\ge -4.0$ shows that none of them 
711: shows the dust-gas separation signature.   Some examples:
712: CS 29498-043 (Aoki \etal 2004) shows an enhancement in Mg similar
713: to those for CNO, and the Zn upper-limit is significantly lower 
714: than CNO which is not predicted in the dust-gas separation pattern. 
715: CS22949-037 (Depagne \etal 2002) also shows an enhancement in Mg
716: (and Na) that is close to that for CNO, and the Zn determination 
717: is signficantly lower whereas dust-gas separation would predict a
718: very similar Zn enhancement to those other elements. 
719: Finally, for HE1300+0157 (Frebel \etal 2007) and HE1424-0241 
720: (Cohen \etal 2007), several elements ($\alpha$ elements and iron
721: group elements) have the same enhancements (or nearly so) as the 
722: CNO abundances.
723: Of course, each of these stars has interesting and unusual 
724: abundance pattern(s), and while not similar to the predictions
725: for dust-gas separation, they are interesting  in terms of the 
726: progenitor mass or explosion characteristics 
727: of their (small number of) contributing supernovae.
728: 
729: 
730: \section{Infrared excesses}
731: 
732: An association of circumstellar dust, i.e., an infrared excess, with
733: a star that is a candidate for exposure to dust-gas separation
734: strengthens the explanation for the abundance anomalies.
735: The absence of an infrared excess is not fatal to the star's
736: candidacy, as shown by many of the Lambda Bootis stars, and the
737: post-AGB star BD+39 4926.
738: 
739: 
740: The broad-band colors and spectral energy distributions from 
741: model atmospheres are shown for the three of the ultra metal-poor
742: stars in Figures~8 to 10.   JHK magnitudes are from the 2MASS catalogue; 
743: UBVRI magnitudes are from Aoki \etal (2006) for HE1327-2326, 
744: Christlieb \etal (2002) for HE0107-5240, 
745: and Beers \etal (2007) for HE0557-4840.  UBVRI
746: magnitudes have been converted to fluxes using the Vega zero 
747: points by Colina \etal (1996) and JHK magnitudes converted
748: using the 2MASS Explanatory Supplement (Sect VI.4a).
749: Reddening has been taken from the reddening maps by 
750: Schlegel \etal (1998); E(B-V) = 0.08, 0.013, and 0.04
751: for HE1327-2326, HE0107-5240, and HE0557-4840, respectively. 
752: These were converted 
753: to A$_\lambda$ based on conversions in Cardelli \etal (1989).
754: 
755: The observed fluxes are compared to model flux distributions from 
756: the online grid of available MARCS (Gustafsson \etal 1975; 
757: Asplund \etal 1997) and Kurucz models (Castelli \& Kurucz 2003).
758: The closest model parameters to those adopted for each star
759: and available in both grids were 
760: \teff = 6250 K, \logg = 4.0, and [Fe/H]=$-5.0$ for HE1327-2326,
761: and  \teff = 5000 K, \logg = 2.0, and [Fe/H]=$-5.0$ for both
762: HE0107-5240 and HE0557-4840.  
763: There are small differences in the models that account for
764: their slight differences in spectral energy distributions,
765: e.g., there are differences in the opacities used, and 
766: MARCS models vary stellar mass and [$\alpha$/Fe] ratios. 
767: The observed fluxes were scaled
768: to overlay the model fluxes best at R, I, and J; the scale factor
769: is estimated for each star, 
770: but would be equivalent to (distance/radius)$^2$.
771: 
772: No infrared excess is seen in the colours of any of the ultra metal-poor
773: stars compared to the model flux distributions.    
774: While this is an important test
775: for the presence of a dusty disk, it is not definitive.   Emission
776: from a disk depends on composition, age, distance from the star,
777: and dust temperature.   For many post-AGB stars, like HR 4049,
778: the infrared excess is  strong in the K-band 
779: and accompanied by a suppression in the UV continuum 
780: (Figure~1 of Dominik \etal 2003).  
781: On the other hand, the circumstellar disk around the post-AGB 
782: star HD56126 (Van Winckel \& Reyniers 2000) is detected  only
783: beyond 4 microns, with no excess observed in the K-band or below 
784: (see Figure 1 in the review paper by van Winckel 2003).
785: Furthermore, not all Lambda Bootis stars show evidence of their
786: dusty disks either.  Paunzen \etal (2003)
787: estimated 23\% of {\it bona fide} Lambda Bootis stars show evidence for
788: circumstellar matter via infrared excesses or Infrared Space
789: Observatory (ISO) and sub-mm
790: CO(2-1) line emission.   Also, the presence of dust
791: around other A-type stars does not necessarily imply the presence
792: of the Lambda Bootis abundance pattern (Kamp \etal 2002, Dunkin
793: \etal 1997); in the study by Acke \& Waelkens (2004) only one
794: in 24 targets showed a clear Lambda Bootis abundance pattern.
795: 
796: Obviously, the search for an infrared excess should, if possible, be
797: extended to longer wavelengths.  
798: Although an infrared excess would clearly identify the presence of
799: a dusty disk, the absence of infrared emission does not preclude the
800: dust-gas separation scenario. 
801: 
802: 
803: \subsection{Are these stars spectroscopic binaries?}
804: 
805: The presence of a companion seems a prerequisite for efficient
806: dust-gas separation in a post-AGB A-type star and quite probably
807: in the RV Tauri stars. (It is not seemingly a  prerequisite
808: for the Lambda Bootis stars.) For this and other reasons, 
809: it would be valuable to show whether all three stars are binaries. 
810: Presently, the available data are limited on this point.
811: 
812: For HE0107-5240, Bessell \etal (2004) find no radial velocity 
813: variations larger than 0.5 km~s$^{-1}$ over $\sim$100 
814: (or 373 days when including two spectra taken nearly a year earlier).
815: Similarly for HE1327-2326, Frebel \etal (2006) find no variations
816: larger than their measurement error or 0.7 to 1.0 km/s over 
817: $\sim$100 days (383 if 3 spectra taken nearly a year earlier
818: are included).  For HE0557-4840, Norris \etal (2007) find no velocity 
819: variations greater than $\pm0.4$ km s$^{-1}$ from observations spanning 
820: 40 days.  These data do not find evidence for binarity, however
821: The gold standard for mass transfer across a binary creating 
822: abundance anomalies is McClure's (1985) study of velocity 
823: variations for Barium stars, which included a 5 year campaign,
824: with accuracies of 0.5 km s$^{-1}$. 
825: An intensive observing campaign may be needed to test convincingly 
826: the hypothesis that these ultra metal-poor stars are or are not
827: spectroscopic binaries.
828: 
829: 
830: 
831: \subsection{The sulphur abundance}
832: 
833: For HE1327-2326, but not for the cooler stars HE0107-5240 and HE0557-5240,
834: it may be possible to detect S\,{\sc i} lines and determine 
835: or at least set an interesting limit on the S abundance.
836: Observations of Lambda Bootis stars, RV Tauri variables, and A-type
837: post-AGB stars  show that S (and Zn) do not share the depletions of
838: iron-group elements (Figure 1). 
839: The upper-limit on the Zn abundance for HE1327-2326 
840: is consistent with a smooth interpolation of the
841: C, N, O, and Na abundance.
842: 
843: The strongest S~I lines of three different multiplets are 
844: observed  around 10455, 9212, and 8695 \AA\,
845: (Caffau \etal 2005, 2007). 
846: Given the atmospheric parameters of HE1327-2326,  
847: we have performed a spectrum synthesis using MOOG
848: (Sneden 1973) and OSMARCS models (Plez \etal 2000).
849: If the star has undergone dust-gas separation, then
850: the intrinsic metallicity as suggested by the C, N, and O
851: abundance is [$X$/H] $\simeq -2$ according to the 3D models but
852: $-1.3$ for 1D models.
853: At [$X$/H]=$-2$, the strongest S\,{\sc i} lines of the
854: 10455, 9212, and 8695 \AA\ multiplets, 
855: are estimated to have LTE equivalent widths of 
856: 15, 25, and 1 m\AA.   NLTE corrections are thought to strengthen
857: the line from these predictions (Takeda \etal 2005).
858: Thus, the 10455 \AA\, and 9212 \AA\, lines should be  detectable,
859: however the 9212 \AA\, lines are  in a region of strong 
860: telluric lines.
861: 
862: Detection of the 10455 \AA\ triplet was reported by
863: Nissen \etal (2007). These authors used CRIRES on the VLT to detect
864: with ease the S\,{\sc i} lines in G29-23, a subdwarf with [Fe/H] $= -1.7$.
865: Its atmospheric parameters ($T_{\rm eff} = 6200$ K, and \logg = 4.0)
866: are similar to those of HE1327-2326. Its [Fe/H] is within the range
867: supposed for HE1327-2326 were the C, N, and O abundances indicative
868: of the initial composition. The challenge will be to
869: extend the CRIRES observation of G29-23 at V = 10.2 to HE1327-2326
870: at V = 13.5, a decrease of about a factor of 20 in flux.  
871: 
872: 
873: 
874: \section{Concluding remarks} 
875: 
876: In this paper, we have outlined the existing evicence for and against
877: the three ultra metal-poor stars being affected by dust-gas separation.
878: The evidence for dust-gas separation having occured at the stellar 
879: photosphere is primarily the abundance pattern, which resembles that
880: of known chemically peculiar stars globally where dust-gas separation
881: has been supported by other observations.   The other observations 
882: usually include one or more of the following for at least one object
883: within a sample; infrared excesses (usually beyond the K band at 2.2 microns) 
884: or other evidence of circumstellar material, evidence of binarity, or
885: the observation that sulphur or zinc do not have the same depletions 
886: as the rest of the metals.   The latter is a critical test because
887: non-depleted abundances of these two elements is more consistent with 
888: dust formation, otherwise requiring random and puzzling variations in 
889: nucleosynthesic yields.    Most of these observational tests do not
890: yet exist for the ultra metal-poor stars, or the results are currently
891: inconclusive regarding dust-gas separation.   
892: 
893: Searchs for radial velocity variations are ongoing in the ultra metal-poor
894: stars to determine if they are in binary systems.  Currently, no variations
895: are found, which suggests these stars are not binaries, but this evidence
896: cannot be conclusive because of orbit inclinations and/or potentially small
897: amplitude variations.     Sulphur and zinc upper-limits exist for all three
898: ultra metal-poor stars, however these upper-limits are presently too high 
899: to distinguish between nucleosynthetic origins or dust-gas separation.  
900: K band photometry exists for all three ultra metal-poor stars, and none of
901: them show an infrared excess, however many RV Tau, post-AGB, and Lambda Boo
902: chemically peculiar stars only show infrared excesses beyond the K band, if at
903: all (many have no detectable infrared excesses).    
904: In addition, known chemically
905: peculiar stars suggest that neither dust not binarity are
906: necessary conditions; the Lambda Bootis stars  exist as single
907: stars, often without dust; the peculiar post-AGB are binaries but
908: one lacks dust; the affected RV Tauri stars have an infrared
909: excess and, although binarity may be a necessary condition, it is
910: as yet unknown observationally that all are binaries.
911: Nevertheless, these observational are possible for the ultra metal-poor 
912: stars, 
913: e.g.  Spitzer observations at mid-IR wavelengths, and examination 
914: of the S\,{\sc I} lines in near-IR spectroscopy,
915: but have yet to be performed.
916: 
917: One remaining concern is that two of the ultra metal-poor stars do not 
918: occupy similar atmospheric parameter ranges as any of the known
919: chemically peculiar stars. 
920: HE0107-5240 and HE0557-4840 have temperatures like the RV~Tauri
921: variables, yet higher gravities and lower intrinsic metallicities
922: (if CNO are the appropriate proxies). 
923: Physically, these stars are expected to have atmospheres  with deep
924: convective envelopes, and thus, accreted gas from a stellar wind, 
925: a circumstellar shell, or a circumbinary disk is should be diluted 
926: beyond detection.    
927: Existence of these envelopes makes it difficult to understand how 
928: dust-gas separation effects can be created and sustained in these giants.
929: Indeed, the stars must have received several tenths of a solar mass of 
930: separated gas in order that the anomalies be detectable in these giants.
931: Of course, this is a concern for the RV Tauri and post-AGB stars as well.
932: %
933: These difficulties are  less severe for HE1327-2326, which is 
934: significantly hotter and is expected to have a more shallow
935: convective envelope.   HE1327-2326 occupies (nearly) the same 
936: stellar parameter range as the Lambda Bootis stars.  Fortunately, 
937: detection and analysis of the S\,{\sc i} lines in this star offers 
938: a critical test for any chemical peculiarities due to dust-gas separation.
939: 
940: 
941: Stellar astronomers are conditioned to think about stars with
942: very peculiar abundance anomalies - real or imagined. Commonsense
943: is often a reliable guide to the boundary between real and imagined.
944: In this regard, if the three stars under consideration here
945: are truly stars of a higher intrinsic metallicity, one
946: might ask why examples have not been seen in well studied
947: samples of metal-poor stars. The globular clusters spring to
948: mind. Str\"{o}mgren photometry should be able to pick out those
949: few stars
950: with a reduced metallicity from the rest of the stars showing
951: the monometallicity that is a mark of a globular cluster.
952: Of course, if these stars are due to the random effects of
953: passing through dense interstellar clouds, and the effects on
954: the abundances are short lived, then one does not expect to find
955: similar stars in globular clusters, and from objective prism surveys
956: such stars would not stand out as peculiar if the resultant metallicity
957: is greater than $\sim -3$. 
958: 
959:   It is only because ultra metal-poor stars
960: are so important as tests of early chemical evolution in the Galaxy
961: that these stars were picked out and studied in detail from high
962: resolution spectroscopy.   It is 
963: important that the suspicion be laid to rest that
964: they are not truly ultra metal-poor but
965: chemically peculiar stars  of a more common, if low,  metallicity. 
966: Then, the focus may be placed exclusively on finding an explanation
967: in terms of stellar nucleosynthesis and the chemical evolution of
968: the young Galaxy (e.g., Iwamoto \etal 2005).
969: 
970: 
971: \acknowledgments
972: We are grateful to P. Bonifacio for valuable discussions on
973: metal-poor stars and on observations and model calculations 
974: of the S\,{\sc I} lines.
975: %
976: We thank Anna Frebel for providing abundances in advance of
977: publication and Katharina Lodders for helpful comments on dust formation
978: in C-rich material. 
979: Thanks to Ian Roederer, Inga Kamp, and the anonymous referee for 
980: many helpful comments on this manuscript.
981: DLL's contributions have been supported by the
982: Robert A. Welch Foundation of Houston, Texas.   
983: KAV would like to thank NSERC for support through a Discovery grant. 
984: 
985: 
986: %% After the acknowledgments section, use the following syntax and the
987: %% \facility{} macro to list the keywords of facilities used in the research
988: %% for the paper.  Each keyword will be checked against the master list during
989: %% copy editing.  Individual instruments or configurations can be provided 
990: %% in parentheses, after the keyword, but they will not be verified.
991: 
992: %{\it Facilities:} \facility{Subaru}, \facility{HST (STIS)}}.
993: 
994: %% The reference list follows the main body and any appendices.
995: %% Use LaTeX's thebibliography environment to mark up your reference list.
996: %% Note \begin{thebibliography} is followed by an empty set of
997: %% curly braces.  If you forget this, LaTeX will generate the error
998: %% "Perhaps a missing \item?".
999: %%
1000: %% thebibliography produces citations in the text using \bibitem-\cite
1001: %% cross-referencing. Each reference is preceded by a
1002: %% \bibitem command that defines in curly braces the KEY that corresponds
1003: %% to the KEY in the \cite commands (see the first section above).
1004: %% Make sure that you provide a unique KEY for every \bibitem or else the
1005: %% paper will not LaTeX. The square brackets should contain
1006: %% the citation text that LaTeX will insert in
1007: %% place of the \cite commands.
1008: 
1009: %% We have used macros to produce journal name abbreviations.
1010: %% AASTeX provides a number of these for the more frequently-cited journals.
1011: %% See the Author Guide for a list of them.
1012: 
1013: %% Note that the style of the \bibitem labels (in []) is slightly
1014: %% different from previous examples.  The natbib system solves a host
1015: %% of citation expression problems, but it is necessary to clearly
1016: %% delimit the year from the author name used in the citation.
1017: %% See the natbib documentation for more details and options.
1018: 
1019: \begin{thebibliography}{}
1020: %\bibitem[]{955} 
1021: \bibitem[AW04]{AW04} Acke B., Waelkens C., 2004, \aap, 427, 1009
1022: \bibitem[Aok04]{Aok04}  Aoki W., Norris J.E., Ryan E.G., Beers T.C.,
1023: Christlieb N., Tsangarides S., Ando H., 2004, \apj, 608, 971
1024: \bibitem[Aok06]{Aok06}  Aoki W., Frebel A., Christlieb N., \etal,
1025:  2006, \apj, 639, 897
1026: \bibitem[And02]{And02} Andrievsky S.M., \etal, 2002, \aap, 396 641
1027: %\bibitem[And07]{And07} Andrievsky S.M., Spite M., Korotin S.A., 
1028: %  Spite F., Bonifacio P., Cayrel R., Hill V., Francois P., 2007, 
1029: %  \aap, in press. 
1030: \bibitem[Asp97]{Asp97}  Asplund M., Gustafsson B., Kiselman D.,
1031:   Eriksson K., 1997, \aap, 318, 521
1032: \bibitem[A84]{A84} Aumann H.H., Beichman C.A., Gillett F.C., \
1033:   de Jong T., Houck J.R., Low F.J., Neugebauer G., Walker R.G.,
1034:   Wesselius P.R., 1984, \apj, 278, 23
1035: \bibitem[BS69]{BS69} Baschek B., Searle L., 1969, \apj, 155, 537
1036: \bibitem[Baui97]{Bau97} Baum{\"u}ller D., Gehren T., 1997, \aap, 
1037:   325, 1088
1038: \bibitem[Beers07]{Beers07}  Beers T., \etal, 2007, \apjs, 168, 128
1039: \bibitem[BCG04]{BCG04} Bessell M.S., Christlieb N., Gustafsson B.,
1040:   2004, \apj, 612, L61
1041: \bibitem[BB56]{BB56} Burbidge E.M., Burbidge G.R., 1956, \apj, 124, 116
1042: \bibitem[C07]{C07} Caffau E., \etal 2007, \aap, 470, 699
1043: \bibitem[C05]{C05} Caffau E., \etal 2005, \aap, 441, 533
1044: \bibitem[CCM]{CCM} Cardelli J., \etal 1989, \apj, 345, 245
1045: \bibitem[CK03]{CK03} Castelli R., Kurucz R.L., 2003,
1046:   Proc. of IAU Symp. 210, Modelling of Stellar Asmospheres, 
1047:   eds. N. Piskunov et al., poster A20 on the enclosed CD-ROM 
1048:   (astro-ph/0405087) 
1049: \bibitem[Chr02]{Chr02} Christlieb N., \etal 2002, Nature, 419, 904
1050: \bibitem[Chr04]{Chr04} Christlieb N., \etal 2004, \apj, 603, 708 
1051: \bibitem[Coh07]{Coh07}  Cohen J., McWilliam A., Christlieb N.,
1052:  Shectman S., Thompson I., Melendez J., Wisotzki L., Reimers D.,
1053:  2007, \apj, 659, L161
1054: \bibitem[Col96]{Col96}  Colina L., Bohlin R., Castelli F., 1996,
1055:   STScI Instrument Science Report CAL/SCS-008
1056: \bibitem[Collet]{Collet} Collet R., Asplund M., Trampedach R., 
1057:   2006, \apj, 644, 121
1058: %\bibitem[]{988} 
1059: \bibitem[Dep02]{Dep02} Depagne E., \etal, 2007, \aap, 390, 187
1060: \bibitem[Dom03]{Dom03} Dominik C., Dullemond C.P., Cami J., 
1061:    van Winckel H., 2003, \aap, 397, 595
1062: \bibitem[Dun97]{Dun97} Dunkin S.K., Barlow M.J., Ryan S.G. 1997,
1063:    MNRAS, 286, 604
1064: \bibitem[Fre05]{Fre05} Frebel A., \etal, 2005, Nature, 434, 871
1065: \bibitem[Fre06]{Fre06} Frebel A., \etal, 2006, \apj, 638, L17
1066: \bibitem[Fre07]{Fre07} Frebel A., \etal, 2007, \apj, 658, 534
1067: \bibitem[FC07]{FC07} Frebel A., \& Christlieb N., 2007, priv. comm.
1068: \bibitem[GAS]{GAS} G\'{a}sp\'{a}r, A., Su K.Y.L., Rieke G.H., 
1069:   Balog Z., Kamp I., Mart\'{i}nez-Galarza J.R., Stapelfeldt K.,
1070:   2007, \apj, in press (astro-ph/0709.4247)
1071: \bibitem[Ge72]{Ge72} Gehrz R.D., 1972, \apj, 178, 715
1072: %\bibitem[]{1000} 
1073: \bibitem[G94]{G94} Giridhar S., Rao N.K., Lambert D.L., 1994,
1074:  \apj, 437, 476
1075: \bibitem[GL05]{GL05} Giridhar S., Lambert D.L., Reddy B.E.,
1076:   Gonzalez G., Yong D., 2005, \apj, 627, 432
1077: \bibitem[GL97]{GL97} Gonzalez G., Lambert D.L., 1997, \aj,
1078:   114, 341
1079: \bibitem[GR99]{GR99} Gratton R., Carretta E., Eriksson K., 
1080:   Gustafsson B., 1999, \aap, 350, 955
1081: \bibitem[GC98]{GC98} Gray R.O., Corbally C.J., 1998, \aj, 116, 2530
1082: \bibitem[Gus75]{Gus75} Gustafsson B., Bell R.A., Eriksson K.,
1083:   Nordlund A., 1975, \apj, 42, 407
1084: \bibitem[KP02]{KP02} Kamp I., Paunzen E., 2002, MNRAS, 335, 45
1085: \bibitem[Kam02]{Kam02} Kamp I., Hempel M., Holweger H., 2002,
1086:   \aap, 388, 978
1087: %\bibitem[Kur70]{Kur70} Kurucz R.L., 1970, SAO Special Report, No. 308
1088: %\bibitem[Kur79]{Kur79} Kurucz R.L., 1979, \apjs, 40, 1
1089: \bibitem[LHL88]{LHL88} Lambert D.L., Hinkle K.H., Luck R.E.,
1090:   1988, \apj, 333, 917
1091: \bibitem[LV96]{LV96} Lemke M., Venn K.A., 1996, \aap, 309, 558
1092: \bibitem[LF95]{LF95} Lodders K., Fegley B., 1995, Meteoritics, 
1093:   30, 661
1094: \bibitem[LF99]{LF99} Lodders K., Fegley B., 1999, in IAU Symp. 191, 
1095:   Asymptotic Giant Branch Stars, ed. T. Le Bertre, A. L\'{e}bre, \&
1096:   C. Waelkens (San Francisco: ASP), 279 
1097: \bibitem[L03]{L03} Lodders K., 2003, \apj, 591, 1220
1098: %\bibitem[Maas05]{Maas05} Maas \etal (2005)
1099: \bibitem[McC]{McC} McClure R.D., 1985, in Cool Stars with Excesses of
1100:   Heavy Elements, Proceedings of the Strasbourg Observatory Colloquium, 
1101:   (Dordrecht: Reidel), 327
1102: \bibitem[McW97]{McW97} McWilliam A., 1997, ARAA, 35, 503
1103: \bibitem[MKK]{MKK} Morgan W.W., Keenan P.C., Kellman E., 1943,
1104:   An Atlas of Stellar Spectra with an Outline of Spectral Classification
1105:   (Chicago: Univ. Chicago Press) 
1106: \bibitem[Norris]{Norris} Norris J.E., Christlieb N., Korn A.J., 
1107:   Eriksson K., Bessel M.S., Beers T.C. Wisotzki L., Reimers D., 2007
1108:   \apj, in press (astro-ph/0707.2657v2)
1109: \bibitem[Pau03]{Pau03} Paunzen E., Kamp I., Weiss W.W., 
1110:   Wiesemeyer H., 2003, \aap, 404, 579 
1111: \bibitem[Plez]{Plez} Plez B., 2000, in The Carbon Star Phenomenon,
1112:   IAU 177 proceedings, p. 71 (Kluwer: Dordrecht) Ed. R.F. Wing.
1113: \bibitem[RR05]{RR05} Rao N.K., Reddy B.E., 2005, MNRAS, 357, 235
1114: %\bibitem[R98]{R98}  Russell S.C., 1998, Publ. Astro. Soc.
1115: %  Aust., 15, 189 
1116: \bibitem[SS96]{SS96} Savage B., Sembach K., 1996, \apj, 470, 893
1117: \bibitem[Sch]{Sch} Schlegel D.J., Finkbeiner D.P., Davis M., 1998,
1118:   \apj, 500, 525
1119: \bibitem[Sivarani]{Sivarani} Sivarani T., \etal, 2006, \aap, 459, 125 
1120: \bibitem[Sne]{Sne} Sneden C., 1973, \apj, 184, 839
1121: \bibitem[Spi05]{Spi05} Spite M., \etal 2005, \aap, 430, 655
1122: \bibitem[Su05]{Su05} Su K.Y.L., \etal 2005, \apj, 628, 487
1123: \bibitem[Suda04]{Suda04} Suda  T., Aikawa M., Machida M.N., Fujimoto M.Y.,
1124:   Iben I. Jr., 2004, \apj, 611, 476
1125: \bibitem[Tak02]{Tak02} Takeda Y., \etal 2002, PASJ, 54, 765
1126: \bibitem[Tak05]{Tak05} Takeda Y., Hashimoto O., Taguchi H., Yoshioka K., 
1127:   Takada-Hidai M., Saito Y., Honda S., 2005, PASJ, 57, 751
1128: %\bibitem[]{1056} 
1129: \bibitem[TUN]{TUN} Tominaga N., Umeda H., Nomoto K., 2007, \apj, 660, 516
1130: \bibitem[Tum07a]{Tum07a} Tumlinson J., 2007a, \apjl, 664, 63
1131: \bibitem[Tum07b]{Tum07b} Tumlinson J., 2007b, \apj, 665, 1361
1132: \bibitem[TC93]{TC93} Turcotte S., Charbonneau P., 1993, \apj, 413, 376
1133: \bibitem[UN03]{UN03} Umeda H, Nomoto K., 2003, Nature, 422, 871
1134: \bibitem[vW95]{vW95} van Winckel H., Waelkens C., Waters L.B.F.M.,
1135:   1995, \aap, 293, L25
1136: \bibitem[vW92]{vW92} van Winckel H., Mathis J.S., Waelkens C., 1992,
1137:   Nature, 356, 500
1138: \bibitem[vWR]{vWR} van Winckel H., Reyniers M., 2000, \aap, 354, 135
1139: \bibitem[vW03]{vW03} van Winckel H., 2003, ARAA, 41, 391 
1140: \bibitem[vW07]{vW07} van Winckel H., 2007, in Baltic Astronomy, 16, 112
1141: \bibitem[vW06]{vW06} van Winckel H., Lloyd Evans T., Reyniers M., Deroo P., 
1142:   Gielen C., 2006, MmSAI, 77, 943
1143: %\bibitem[]{1070} 
1144: \bibitem[VL90]{VL90} Venn K.A., Lambert D.L., 1990, \apj, 363, 234
1145: \bibitem[W91]{W91} Waelkens C., van Winckel H., Bogaert E., 
1146:   Trams N.R., 1991, \aap, 251, 495
1147: \end{thebibliography}
1148: 
1149: 
1150: 
1151: %% Use the figure environment and \plotone or \plottwo to include
1152: %% figures and captions in your electronic submission.
1153: %% To embed the sample graphics in
1154: %% the file, uncomment the \plotone, \plottwo, and
1155: %% \includegraphics commands
1156: %%
1157: %% If you need a layout that cannot be achieved with \plotone or
1158: %% \plottwo, you can invoke the graphicx package directly with the
1159: %% \includegraphics command or use \plotfiddle. For more information,
1160: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
1161: %% documentation section at the AASTeX Web site,
1162: %% http://www.journals.uchicago.edu/AAS/AASTeX.
1163: %%
1164: %% The examples below also include sample markup for submission of
1165: %% supplemental electronic materials. As always, be sure to check
1166: %% the instructions to authors for the journal you are submitting to
1167: %% for specific submissions guidelines as they vary from
1168: %% journal to journal.
1169: 
1170: %% This example uses \plotone to include an EPS file scaled to
1171: %% 80% of its natural size with \epsscale. Its caption
1172: %% has been written to indicate that additional figure parts will be
1173: %% available in the electronic journal.
1174: 
1175: \begin{figure}
1176: \plotone{f1.eps}
1177: \caption{ 
1178: Elemental abundances of chemically peculiar stars relative to solar; 
1179: including the Lambda Bootis star, 29~Cyg (red; Venn \& Lambert 1990), 
1180: two RV Tauri stars 
1181: UY~CMa and HP~Lyr (Giridhar \etal 2005), and two A-type post-AGB stars,
1182: HR~4049 and HD~52961 (Waelkens \etal 1991, with Zn from 
1183: Takeda \etal 2002 and van Winckel \etal 1992).    
1184: Elemental abundances are plotted 
1185: against the solar condensation temperatures from Lodders (2003).
1186: Typical errors are $\le0.2$~dex, shown by the errorbar in the upper right.
1187: }
1188: \end{figure}
1189: 
1190: 
1191: \begin{figure}
1192: \plotone{f2.eps}
1193: \caption{ 
1194: [$X$/H] abundances in HE1327-2326 from Frebel \& Christlieb (2007)
1195: against the solar condensation temperatures from Lodders (2003).
1196: LTE abundances (solid points) and the upper limit for Zn are shown, 
1197: all with (3D) corrections for stellar granulation.  
1198: This abundance pattern is compared to the post-AGB and RV Tauri
1199: stars (references in Figure~1).   While the abundance pattern
1200: for HE1327-2326 resembles those of the chemically peculiar stars,
1201: the high [Mg/Fe] ratio and Al values are striking. 
1202: Typical errors are $\le0.20$~dex shown by the errorbar in the upper right;
1203: Aoki \etal (2006) suggest slightly larger errors for C and N in HE1327-2326,
1204: of 0.24 and 0.30 dex, respectively. 
1205: }
1206: \end{figure}
1207: 
1208: \begin{figure}
1209: \plotone{f3.eps}
1210: \caption{ 
1211: [$X$/H] abundances in HE1327-2326 from Frebel \& Christlieb (2007)
1212: as in Figure~2 with additional upper-limits noted.
1213: The abundance pattern is compared to the post-AGB stars 
1214: and two Lambda Bootis stars (29 Cyg from Venn \& Lambert 1990, 
1215: and HD~106223 from Andrievsky \etal 2002).   
1216: }
1217: \end{figure}
1218: 
1219: \clearpage
1220: 
1221: 
1222: \begin{figure}
1223: \plotone{f4.eps}
1224: \caption{ 
1225: [$X$/H] abundances in HE0107-5240 from Collet \etal (2006;
1226: originally from Christlieb \etal 2004 and Bessell \etal 2004) 
1227: are shown against the solar condensation temperatures from 
1228: Lodders (2003).
1229: LTE abundances (solid points) and the Zn upper-limit are shown, 
1230: all with (3D) corrections for stellar granulation.
1231: This abundance pattern is compared to the post-AGB stars 
1232: and RV Tauri variables as in Figure~2.
1233: Typical errors for HE0107-5240 are $\le0.24$~dex, 
1234: shown by the errorbar in the upper right;
1235: only the uncertainly on C could be higher, 0.34~dex when 
1236: determined from the CH {\it A-X} band (Christlieb \etal 2004). 
1237: }
1238: \end{figure}
1239: 
1240: \begin{figure}
1241: \plotone{f5.eps}
1242: \caption{ 
1243: [$X$/H] abundances in HE0107-5420 from Collet \etal (2006)
1244: as in Figure~4 with additional upper-limits shown. 
1245: The abundance pattern is compared to the post-AGB stars 
1246: and Lambda Bootis stars as in Figure~3.
1247: }
1248: \end{figure}
1249: 
1250: 
1251: \clearpage
1252: 
1253: \begin{figure}
1254: \plotone{f6.eps}
1255: \caption{ 
1256: [$X$/H] abundances in HE0557-4840 from Norris \etal (2007) are shown
1257: against the solar condensation temperatures from Lodders (2003).
1258: LTE abundances (solid points) are shown with {\it no corrections
1259: for 3D effects}. 
1260: This abundance pattern is compared to the post-AGB stars 
1261: and RV Tauri variables as in Figure~2.
1262: Typical errors are $\le0.20$~dex, shown by the errorbar in the upper right.
1263: }
1264: \end{figure}
1265: 
1266: 
1267: \begin{figure}
1268: \plotone{f7.eps}
1269: \caption{ 
1270: [$X$/H] abundances in HE0557-4840 from Norris \etal (2007) are shown
1271: against the solar condensation temperatures from Lodders (2003).
1272: LTE abundances (solid points) that have been corrected for
1273: solar granulation (3D effects) are shown.  The 3D corrections 
1274: have a very small effect on most elements, except carbon.
1275: Typical errors are $\le0.20$~dex, shown by the errorbar in the upper right.
1276: The abundance pattern is compared to one of the post AGB
1277: stars and the Lambda Bootis stars as in Figure~3.
1278: }
1279: \end{figure}
1280: 
1281: 
1282: 
1283: \begin{figure}
1284: \plotone{f8.eps}
1285: \caption{
1286: The spectral energy distribution for HE1327-2326 from broad band
1287: UBVRI (Aoki \etal 2006) and JHK (2MASS) fluxes.   
1288: These are compared to model atmosphere 
1289: flux distributions from both MARCS and Kurucz with
1290: T$_{\rm eff}$=6250\,K, \logg=4.0, and [Fe/H]=$-5.0$.   
1291: Observed magnitudes are scaled to the absolute flux units
1292: in order to best fit the RIJ magnitudes.
1293: Errors in the fluxes are smaller than the symbol sizes.
1294: }
1295: \end{figure}
1296: 
1297: 
1298: \begin{figure}
1299: \plotone{f9.eps}
1300: \caption{
1301: The spectral energy distribution for HE0107-5240.
1302: Broad band BVRI colors from Christleib \etal (2002) and 
1303: JHK (2MASS) colors.   The model atmosphere 
1304: flux distributions are from both MARCS and Kurucz with
1305: T$_{\rm eff}$=5000\,K, \logg=2.0, and [Fe/H]=$-5.0$.
1306: Observed magnitudes are scaled to the absolute flux units
1307: in order to best fit the RIJ magnitudes.
1308: Errors in the fluxes are smaller than the symbol sizes.
1309: }
1310: \end{figure}
1311: 
1312: 
1313: 
1314: \begin{figure}
1315: \plotone{f10.eps}
1316: \caption{
1317: The spectral energy distribution for HE0557-4840 from broad band
1318: BVR (Beers \etal 2007) and JHK (2MASS) fluxes.   
1319: These are compared to model atmosphere 
1320: flux distributions from both MARCS and Kurucz with
1321: T$_{\rm eff}$=5000\,K, \logg=2.0, and [Fe/H]=$-5.0$.   
1322: Observed magnitudes are scaled to the absolute flux units
1323: in order to best fit the R and J magnitudes.
1324: Errors in the fluxes are smaller than the symbol sizes.
1325: }
1326: \end{figure}
1327: 
1328: 
1329: 
1330: %% Tables should be submitted one per page, so put a \clearpage before
1331: %% each one.
1332: 
1333: \end{document}
1334: