1: \documentclass[12pt,preprint]{aastex}
2:
3: \shorttitle{r- and s-process Elemental Abundances}
4: \shortauthors{Bond et al.}
5:
6: \begin{document}
7:
8: \title{Beyond the Iron Peak: r- and s-process Elemental Abundances in Stars with Planets}
9:
10: \author{J. C. Bond\altaffilmark{1}, D. S. Lauretta\altaffilmark{1}, C. G. Tinney\altaffilmark{2,3}, R. P. Butler\altaffilmark{4}, G. W. Marcy\altaffilmark{5}, H. R. A. Jones\altaffilmark{6}, B. D. Carter\altaffilmark{7}, S. J. O'Toole\altaffilmark{3} $\&$ J.
11: Bailey\altaffilmark{8}}
12: \altaffiltext{1}{Lunar \& Planetary Laboratory, University of
13: Arizona, 1629 E. University Blvd, Tucson, AZ 85721-0092}
14: \altaffiltext{2}{Dept. of Astrophysics, University of NSW, NSW
15: 2052, Australia} \altaffiltext{3}{Anglo-Australian Observatory, PO
16: Box 296, Epping, NSW 1710, Australia} \altaffiltext{4}{Department
17: of Terrestrial Magnetism, Carnegie Institution of Washington, 5241
18: Broad Branch Road NW, Washington, DC 20015-1305}
19: \altaffiltext{5}{Department of Astronomy, University of
20: California, Berkeley, CA 94720} \altaffiltext{6}{Centre for
21: Astrophysics Research, University of Hertfordshire, College Lane,
22: Hatfield AL10 9AB, England} \altaffiltext{7}{Faculty of Sciences,
23: University of Southern Queensland, Toowoomba 4350, Australia}
24: \altaffiltext{8}{Physics Department, Macquarie
25: University, Sydney, NSW 2109, Australia}
26:
27: \begin{abstract}
28: We present elemental abundances of 118 stars (28 of which are
29: known extrasolar planetary host stars) observed as part of the
30: Anglo-Australian Planet Search. Abundances of O, Mg, Cr, Y, Zr,
31: Ba, Nd and Eu (along with previously published abundances for C
32: and Si) are presented. This study is one of the first to
33: specifically examine planetary host stars for the heavy elements
34: produced by neutron capture reactions. We find that the host stars
35: are chemically different to both the standard solar abundance and
36: non-host stars in all elements studied, with enrichments over
37: non-host stars ranging from 0.06 dex (for O) to 0.11 dex (for Cr and Y).
38: Such abundance trends are in agreement with other previous studies
39: of field stars and lead us to conclude that the chemical anomalies
40: observed in planetary host stars are the result of normal galactic
41: chemical evolution processes. Based on this observation, we
42: conclude that the observed chemical traits of planetary host stars
43: are primordial in origin, coming from the original nebula and not
44: from a ``pollution'' process occurring during or after formation
45: and that planet formation occurs naturally with the evolution of
46: stellar material.
47: \end{abstract}
48:
49: \keywords{stars: abundances - stars: planetary systems - stars:
50: chemical peculiar}
51:
52: \section{Introduction}
53: Extrasolar planets are known to preferentially orbit
54: metal-enhanced stars. Chemical analyses of known host stars have
55: shown that they appear to be metal enriched compared to a sample
56: of ``average'' F, G and K stars not known to harbour planets
57: \citep{g1,g2,g4,g5,g3,sb,s1,s04,g6,sm,re,fv,bond}. In addition to
58: this global metal enrichment, individual elements have also been
59: shown to exhibit similar trends, although not as statistically
60: significant or with such large host and non-host differences
61: (\citealt{g6,sb,bo,fv,bond}).
62:
63: The vast majority of abundance studies completed so far have
64: focussed on elements with atomic number (Z)$\leq$30 (i.e. those
65: located before the iron stability peak). These elements are
66: produced by a variety of processes (for example alpha particle
67: addition, the CNO cycle and stellar burning reactions) in stellar
68: interiors during main sequence evolution. Elements located beyond
69: the iron peak, however, are produced via neutron-capture
70: reactions, specifically the rapid (r-) and slow (s-) processes.
71: Here rapid and slow refers to the speed of neutron capture with
72: respect to the $\beta$-decay rate of the nuclei. In the r-process,
73: neutron capture occurs before $\beta$-decay of the unstable nuclei
74: can occur. Alternatively, in the s-process neutron capture occurs
75: less frequently, thus allowing the nuclei to undergo $\beta$-decay
76: before capturing another neutron and effectively allowing the
77: nuclide to remain within the valley of $\beta$ stability.
78:
79: Due to the different neutron fluxes required for each process
80: ($\sim$10$^{23}$ neutrons\,cm$^{-2}$\,s$^{-1}$ for the r-process
81: vs. $\sim$10$^5$ neutrons\,cm$^{-2}$\,s$^{-1}$ for the s-process
82: (\citealt{clay})), the r- and s-processes occur in different
83: stellar environments. There is some debate as to exactly where the
84: r-process occurs (see e.g. \citealt{qian}), but it is believed to
85: primarily occur in type II supernova events. The s-process is
86: thought to produce its heavier elements (such as Ba and Ce) in the
87: interior of lower mass AGB stars and its lighter elements (such as
88: Sr, Y and Zr) during the He-burning stages of stellar evolution
89: for larger mass stars (\citealt{red03}). Due to these vastly
90: different settings, the abundances of the heavy elements produced
91: by these processes can provide information on the history of the
92: material later incorporated into both the host star and planets
93: themselves, as well as testing models of galactic chemical
94: evolution (e.g. \citealt{fen} and \citealt{lan}). These elements
95: have been part of several previous spectroscopic studies of stars
96: without planets (e.g. \citealt{ed, red03,ap,ben,red06}), however
97: only Ba (\citealt{hu}) and Eu (\citealt{gonz}) have been
98: specifically examined in a small number of known extrasolar
99: planetary host stars.
100:
101: Studies of stellar abundances have become critical to our
102: understanding of planetary formation processes. However, in spite
103: of significant advances in atmospheric models, stellar interiors
104: and atomic line data in recent years, the measurement of stellar
105: abundances is still an intricate process. In particular, choices
106: made in the way an analysis is carried out can result in
107: systematic abundance differences on sizes similar to the effects
108: we would most like to understand in the stars themselves.
109: Additionally, in many cases there is no consensus on the
110: ``correct'' choice in these techniques.
111:
112: To make headway in this field, therefore, it is essential to
113: perform abundance studies in a manner immune to such systematic
114: errors, by analysing both samples of interest, and control
115: samples, in an identical manner. For the abundances of
116: exoplanetary host stars, that means both host and non-host stars
117: must be analysed in an identical manner. This is the primary goal
118: of this current study.
119:
120: We have derived abundances of five r-and s-process elements, in
121: addition to three other lighter elements, for all the
122: planet-hosting and non-planet-hosting stars in the
123: Anglo-Australian Planet Search (AAPS) data set with viable
124: template spectra, so that robust conclusions can be reached about
125: the differences in their elemental abundances. The inclusion of
126: these heavy elements begins to extend the spectroscopic studies of
127: known host stars beyond the iron peak, thus continuing the search
128: for additional chemical anomalies within these systems, while also
129: providing an independent check of nucelosynthesis models and
130: previously published abundances and trends for known host stars.
131: The lighter elements selected for study (O, Mg and Cr) are
132: included so as to complement the previous study of \citet{bond}
133: while also providing valuable information as to the cause of the
134: metal-enrichment commonly seen in known host stars.
135:
136: \section{Data}
137: \subsection{Target Stars}
138: In this study, we concentrate on the F- and G-type stars, observed
139: at the 3.9m Anglo-Australian Telescope (AAT) since January 1998 as
140: part of the AAPS program
141: \citep{b4,b5,tin1,tin3,tin2,tin05,tin06,j2,j3,j06,c4,mc,jen06,
142: w07,ot07}. As at January 2007, 31 stars present within the AAPS
143: sample were known to be planet hosts, of which 28 had spectra
144: useful for the purposes of this study. Stars known to be young
145: (age $<$ 3\,Gyr), active (logR\arcmin$_{HK}$$>$ -4.5) or with
146: other stars within 5\arcsec\ are rejected from the AAPS search.
147: For a more detailed description of the data and details of the
148: criteria applied to the AAPS target stars, the reader is referred
149: to \citet{but96} and \citet{tin05}. Of the 28 host stars
150: considered here, 26 have had some common elemental abundances
151: previously determined by other authors.
152:
153: \subsection{Spectroscopic Analysis}
154: The method utilised in this study closely follows that outlined in
155: \citet{bond} who studied Fe, C, Na, Al, Si, Ca, Ti and Ni. Spectra
156: encompassing the entire visible spectrum from 4820 to 8420\AA\
157: with a signal-to-noise ratio (S/N) between 200 and 300 per
158: spectral pixel at resolution $\lambda/\Delta\lambda\approx80000$
159: were obtained via the University College London Echelle
160: Spectrograph (UCLES) using the 31.6 line/mm echelle grating as
161: part of the AAPS program. The raw data were reduced and processed
162: so as to produce a one-dimensional spectra, suitable for spectral
163: analysis.
164:
165: We do differ slightly from \citet{bond} in the method used to
166: determine the equivalent widths of absorption lines. In
167: \citet{bond}, we obtained equivalent width estimates via direct
168: integration over the line. In this study we follow the method of
169: other similar studies (eg. \citealt{sb,g6,s1}) and make Gaussian
170: line fits to the spectra using the IRAF task \texttt{splot} in the
171: package \texttt{noao.onedspec} (due to difficulties in automating
172: the previous script). This slight difference in methodology does
173: not produce any significant difference in the final abundances
174: obtained.
175:
176: Five heavy elements were analysed for the first time as part of
177: this study and they were primarily selected based on their process
178: of production. Y, Zr and Ba are all produced primarily by the
179: s-process while Eu is primarily produced by the r-process and Nd
180: is an almost even mix between the two based on Solar System
181: abundances (\citealt{ar,sim}). The line list utilised in this
182: study is shown in Table \ref{lines} and is derived from
183: \citet{gilli} (for Mg and Cr), \citet{red03}(for Y, Zr, Ba, Eu and
184: Nd) and \citet{den} (for Nd). Additional Ba lines have been used
185: in other studies, but were neglected here as they gave
186: consistently lower abundances by approximately 0.60 dex. While the
187: use of more lines in determining an abundance should produce a
188: more robust value, we could not determine the cause of this offset
189: and elected instead to use the single stronger line at 6496.91\AA\
190: as this line produced a Ba abundance close to the reported solar
191: value for a solar spectrum. Atomic parameters for each line were
192: obtained from the NIST Atomic Spectra Database, Version 3.0 (for
193: O, Mg and Cr), \citet{pits} and \citet{han} (for Y), \citet{red03}
194: (for Ba), \citet{den} (for Nd) and the Kurucz atomic line
195: database\footnote{http://www.pmp.uni-hannover.de/cgi-bin/ssi/test/kurucz/sekur.html}
196: (for Zr and Eu). All of the atomic parameters applied here have
197: also been utilised in previous studies and produce solar elemental
198: abundances well within errors of those published elsewhere, when
199: applied to a solar spectrum, thus giving us confidence in applying
200: them here.
201:
202: Elemental abundances were obtained via standard local
203: thermodynamic equilibrium analysis, as has been done by previous
204: studies (see \citealt{sb,g6,s1,bond}). A revised version of
205: Sneden's (1973) MOOG abundance code entitled \texttt{width6}
206: \citep{r1} was once again used in conjunction with a grid of
207: \citet{kr} ATLAS9 atmospheres\footnote{http://kurucz.harvard.edu
208: or http://www.stsci.edu/hst/observatory/cdbs/k93models.html}
209: to obtain the final elemental abundances. As all of the non-host,
210: and all but 7 of the host stars, had been the subject of an
211: earlier study (\citealt{bond}), previously published stellar
212: atmospheric parameters were used. For those stars without
213: previously determined values, we followed the same method as used
214: in \citet{bond} to obtain the values and refer the reader to the
215: paper for more details. As the OI triplet lines are known to
216: suffer from non-LTE effects, the corrections of \citet{tak} for
217: $\xi_t$=1km/s and log g=4.0 cm/s$^{2}$ were applied to obtain our
218: final O abundances. When applied to a solar spectrum, this method
219: produced abundances in agreement with those of \citet{asp}.
220:
221: \section{Results}
222: The stellar elemental abundances are shown in Table \ref{allab}
223: (in the standard astronomical logarithmic form) and Table
224: \ref{allsi} (in the more cosmochemically useful form with
225: abundances normalized to 10$^6$ Si atoms). The notation of $-$ for
226: an abundance indicates that a value could not be obtained from the
227: spectrum due to noise. Additionally, for ease of comparison in
228: Section 5.1, C and Si stellar elemental abundances, along with the
229: normalized C abundances, for all target stars are presented in
230: Table \ref{oldab}. The C and Si abundances were previously
231: published in \citet{bond}.
232:
233: Other authors have previously determined abundance values for
234: several of the elements also studied here with many of these
235: abundances differing from our values. Differences between our
236: present study and that of others is not a significant issue for
237: the primary thrust of this paper, which is to compare host and
238: non-host stellar abundances which have been measured in an
239: identical fashion. However, in the interests of completeness we
240: note that Mg abundances were determined for 29 common stars (20
241: hosts) by \citet{be}, Cr in 28 stars (20 hosts) and Mg in 29 stars
242: (20 hosts) by \citet{gilli}, Cr in 25 stars (16 hosts) by
243: \citet{bo}, O in 8 stars (4 hosts) by \citet{ec06}, O and Eu in 4
244: host stars, Cr and Mg in 1 host star by \citet{gonz} and O in 1
245: host star by \citet{sb}. The mean differences between our values
246: and those previously published are shown in Table \ref{difference}
247: (for those samples having more than one common star) with the
248: difference being defined as the abundance from this study minus
249: the published abundance. Generally, our results can be seen to be
250: in agreement with those previously published for Cr and O with a
251: significantly larger mean difference occurring for Mg and and a
252: large deviation occurring for Eu. The differences between our
253: abundances and those previously published are believed to be due
254: to the use of a smaller number of lines in determining the
255: abundance (for Mg), the use of different methods (for O and Eu),
256: the use of different atomic parameters (for Eu) and the use of
257: different non-LTE corrections (for O).
258:
259: \section{Host Star Enrichment}
260: \subsection{Enrichment over Solar}
261: The mean and median abundances, standard deviation and the
262: difference between the host and non-host stars for all of our
263: target stars can be seen in Table \ref{stat} for each element. The
264: quoted uncertainties are the standard error in the mean, and the
265: median uncertainty from the algorithm of \citet{kd}\footnote{For a
266: distribution with N values, the error in the median is the range
267: in values on either side of the median which contains ($\surd$N)/2
268: values }. The mean values of this study for the known host stars
269: are all consistent to within the 1$\sigma$ value of those listed
270: by \citet{be} and \citet{gilli}. The data show that in general
271: known extrasolar planetary host stars differ only slightly from
272: the mean solar abundance patterns with all of the median
273: abundances being well within 1$\sigma$ of the solar abundance - as
274: concluded by previous studies (eg. \citealt{ec,bo}). This is
275: reassuring as the Sun is itself is obviously a planetary host star
276: with abundances enhanced over those of most other stars in the
277: solar neighbourhood (based on the abundances of \citealt{asp}). In
278: many respects, therefore, the Sun is \emph{not} a typical field
279: star, based on its abundances and its multiple planetary
280: companions. The largest enrichment over solar is seen in Nd and
281: Zr, with Cr showing a smaller enrichment and Mg showing minimal
282: enrichment. Eu showed the largest depletion relative to solar
283: values, with Y and Ba also showing mild to moderate depletions. Only O produced a mean abundance equal to the Solar abundance.
284:
285: Similarly, the non-host stars can also be seen be depleted when
286: compared to solar abundances for almost all of the elements
287: studied, with the largest depletion being $-$0.16 for Y and Eu. It is
288: also worth noting that three of the five heavy elements examined
289: show a mean depletion relative to solar for both the host and
290: non-host stars. Of these three, two are produced by the s-process
291: (Y \& Ba) with the remaining element (Eu) produced by the
292: r-process.
293:
294: \subsection{Enrichment over Non-Host Stars}
295: A more powerful comparison is obtained by comparing our host and
296: non-host populations to each other. On doing so, it can be clearly
297: seen that host stars are systematically enriched over non-host
298: stars in all elements studied. The enrichment ranges in size from
299: 0.06 (for O) to 0.11 for (for Cr and Y) (see Table \ref{stat}).
300:
301: This difference between the host and non-host populations can also
302: be seen in the results of the Kolmogorov-Smirnov (K-S) statistical
303: test. Designed to test whether two populations were drawn from the
304: same parent sample, the K-S test showed a significant difference
305: between the host and non-host samples with probabilities of a
306: different parent sample ranging from 91.2\% for O to 99.99\% for Y
307: (see Table \ref{stat}). This supports our claim that host star
308: elemental abundances are significantly different to those of
309: non-host stars and furthermore that host stars are enriched over
310: non-host stars for all elements studied.
311:
312: \section{Elemental Trends}
313: Plots of our results are shown in Figures \ref{allh} and
314: \ref{allfe} - in Figure \ref{allh} we present [X/H] versus [Fe/H]
315: as more commonly used in previous studies of planet host star
316: abundances, while in Figure \ref{allfe} we present [X/Fe] versus
317: [Fe/H] as usually analysed in cosmochemical studies. Two
318: significant outliers can be seen - one host and one non-host
319: sitting below the general trend for O, Cr and Mg. These stars are
320: HD142415 ([Fe/H]=0.02, host) and HD199288 ([Fe/H]=0.04, non-host).
321: These stars can be seen to be depleted (compared to solar
322: abundances) in the majority of elements studied here except for
323: Fe, suggesting that they have formed from Fe-rich precursor
324: material based on the assumption that it is easier to enrich one
325: element than it is to deplete seven elements. The fact that
326: HD142415 is also mildly enriched in both Ba and Y (with no Zr
327: abundance available) could also possibly indicate that the
328: material had been processed through an s-process environment,
329: either an AGB star or the He-burning stage of a larger mass star.
330:
331: \emph{Fe and Fe precursors:} The [X/H] trends in Figure \ref{allh}
332: are in agreement with the understanding we currently have about
333: the nucelosynthetic origin of the elements. All of the elements
334: located before the Fe peak (here O, Mg and Cr) increase linearly
335: with increasing [Fe/H] with Pearson product-moment correlation
336: coefficients (r) above 0.70 for both host and non-host stars. This
337: can be easily understood as the stellar evolutionary processes
338: that serve to increase the amount of Fe present in later
339: generations of stars also produce the pre-iron peak elements in
340: various amounts. Thus as the stellar Fe content increases, so too
341: would the amount of pre-iron peak elements (neglecting any unusual
342: mixing or other nebula interactions).
343:
344: \emph{s-process:} The r- and s-process elements, however, are less
345: certain. All three s-process elements (Y, Zr and Ba) still display
346: to varying degrees the same trend of increasing abundance with
347: increasing [Fe/H] as the pre-iron peak elements. One possible
348: explanation for this observed trend is that the increase in
349: s-process elemental abundances is due to the increase in the
350: number of seed nuclei (e.g. Fe atoms) available. Due to the nature
351: of the s-process, it is reliant on the sufficient availability of
352: seed nuclei to be able to proceed. Thus as the metallicity
353: increases, so too does the abundance of s-process species.
354:
355: \emph{r-process:} Unlike the s-process elements, the r-process
356: element (Eu) and the mixed source element (Nd) do not display a
357: strong correlation with increasing [Fe/H]. Observations of metal
358: poor stars have shown that the abundances of s-process elements
359: such as Y and Ba decrease faster with metallicity than the
360: abundances of r-process elements such as Eu (\citealt{spite}).
361: This has been attributed to a lack of appropriate seed nuclei
362: inhibiting the s-process significantly more than the r-process. We
363: are alternatively extending this into the metal-rich regime to
364: conclude that the r-process is not as reliant on the presence of
365: elements such as Fe, thus explaining its lack of a strong
366: dependance upon metallicity.
367:
368: From Figure \ref{allfe}, it can be seen that the overall [X/Fe]
369: trends visible here are in good agreement with those identified by
370: \citet{bo} and \citet{gilli} (which are discussed in more detail
371: below). They can also be seen as extensions into the high
372: metallicity region of those trends identified by \citet{red06}.
373:
374: \subsection{Lighter Element Trends}
375: In addition to examining the nature of the general trend of
376: increasing elemental abundances with increasing metallicity, we
377: can also make some basic determinations about the nature of the
378: various nucelosynthesis processes occurring within the precursors
379: to these systems by considering the more subtle, second order
380: trends present within the data.
381:
382: \textit{O:} Based on our data, [O/Fe] displays a weakly correlated
383: decreasing trend with increasing [Fe/H] for both the host stars
384: (slope=$-$0.16, r=$-$0.22) and non-host stars (slope=$-$0.24,
385: r=$-$0.32). Previous studies have hinted at the possibility of a
386: plateau starting at [Fe/H]$\sim$0
387: (\citealt{red03},\citealt{red06}). However, the overlap between
388: our sample and those previously published is not large enough to
389: allow us to undertake any meaningful comparison. It is also worth
390: noting that while the solar C/O ratio is 0.54 (\citealt{asp}),
391: using the [C/H] values previously published by \citet{bond} (see
392: Table \ref{oldab}), the C/O values of the host stars studied here
393: range from 0.40 to 0.89. This variation has the potential to
394: greatly impact the C chemistry of the proto-stellar disc and thus
395: also any terrestrial planets forming in the system. Further
396: studies into this subject are currently underway (see
397: \citealt{bond2}).
398:
399: \textit{Mg:} [Mg/Fe] can be seen to display a weakly correlated
400: trend of decreasing with increasing [Fe/H] values producing r
401: values of $-$0.32 (host stars) and $-$0.44 (non-host stars). The
402: distribution observed for Mg is in good agreement (for the [Fe/H]
403: regions in common) with those of previous studies who observed a
404: decrease in [Mg/Fe] with increasing [Fe/H] up to
405: [Fe/H]$\sim$$-$0.10 before both distributions plateaued
406: (\citealt{be,gilli,red06}).
407:
408: Also of interest is the Mg/Si ratio of our target stars as it has
409: the potential to greatly affect the chemical evolution of any
410: terrestrial planets forming in the system. A high Mg/Si ratio will
411: result in all of the available Si forming olivine
412: ((Fe,Mg)$_{2}$SiO$_{4}$) with excess Mg still being available,
413: while a low Mg/Si ratio will result in all of the available Mg
414: forming enstatite (MgSiO$_{3}$) with excess Si forming SiO$_{2}$.
415: Utilising the Si abundances previously published in \citet{bond}
416: (see Table \ref{oldab}), the host stars of this study were found
417: to have Mg/Si ratios ranging from 0.46 to 1.26, resulting in
418: potentially large variations in the nature of any terrestrial
419: planets forming in these systems. A separate study examining this
420: issue is currently underway (see \citealt{bond2}).
421:
422: \textit{Fe Group (Cr):} From Figure \ref{allfe}, it can be seen
423: that (with the exception of one outlier) [Cr/Fe] displays no
424: significant trend, remaining largely unchanged over all of the
425: [Fe/H] values considered here. This is in good agreement with
426: \citet{red03,ben} and \citet{gilli}. The lack of a statistically
427: significant trend with [Fe/H] is confirmed by the low r value of
428: $-$0.04 for hosts and 0.00 for non-hosts.
429:
430: \subsection{Heavy Element Trends}
431:
432: From Figure \ref{allfe}, it can be seen that all of the heavy
433: elements display varying degrees of a weakly negative to
434: non-existent correlation with [Fe/H]. The r values produced range
435: from $-$0.22 (Y and Ba) to $-$0.56 (Nd) indicating that the
436: correlation is not strong.
437:
438: These trends are in agreement with \citet{ap} and \citet{ben} over
439: the range of metallicity values in common with this study. We do
440: differ slightly from some previous studies in that we observed a
441: stronger decrease in [Y/Fe] with increasing [Fe/H] than was
442: observed by \citet{ben} and disagree with previous studies who
443: observed no significant trend with [Fe/H] for both [Ba/Fe] and
444: [Nd/Fe] (\citealt{red03}). We do observe trends in both of these
445: samples and the difference is attributed to the fact that we are
446: examining a different metallicity region to that of \citet{red03}.
447: Our population is largely concentrated in the region of
448: [Fe/H]$>$0, while the sample of \citet{red03} almost exclusively
449: has values of [Fe/H]$<$0. However, it is worth restating that the
450: general trends in the neutron capture elements identified here are
451: the same as those previously identified for other solar-type
452: stars. This implies that the host stars examined in this paper
453: follow the same trends as other field stars but with a bias
454: towards the high metallicity region. Additionally, the high degree
455: of scatter observed in these samples is also in agreement with
456: previous studies.
457:
458: Also of interest is the ratio of the heavy to light s-process
459: elements as each are thought to be produced in slightly different
460: stellar settings. The abundance of the heavy s-process element Ba
461: to the light s-process elements Y and Zr is shown in Figure
462: \ref{hl} as [heavy/light] where:
463: \begin{eqnarray}
464: \nonumber {\rm [heavy/light]} = \rm [Ba/H] - \frac{\rm
465: [Y/H]+[Zr/H]}{2}
466: \end{eqnarray}\\
467:
468: From Figure \ref{hl}, it can be clearly seen that there is no
469: dependance on metallicity, with values scattering around the solar
470: value (0.0 by definition). This is in agreement with
471: \citet{red03}. Thus we support their conclusion that the neutron
472: exposure in AGB stars is independent of the metallicity of the
473: star itself, assuming (as \citealt{red03} did) that AGB stars are
474: the primary source of both the heavy and the light s-process
475: elements.
476:
477: \subsection{Correlation with Planetary Parameters}
478: Figure \ref{orbits} shows plots of [X/H] against planetary
479: parameters (M$\sin i$, semi-major axis $a$, eccentricity and
480: planetary period) for the 5 heavy elements considered in this
481: study. HD164427 was omitted as its companion is believed to be a
482: brown dwarf, not a gas giant. As can be seen visually and by the
483: low r values (all $\leq$0.48 with most $<$0.15), no statistically
484: significant correlations exist between these abundances and any of
485: the orbital parameters. This agrees with previous studies of other
486: elements (eg. \citealt{re,s2,fv}).
487:
488: \subsection{Correlation with Stellar Parameters}
489: It is well known that the stellar atmospheric parameters
490: (specifically Teff and log g) have the potential to drastically
491: alter photospherically determined stellar abundances. As such, we
492: examined the abundances presented here and as both samples
493: produced r$^2$ correlation coefficients less than 0.5, we
494: concluded that there is no statistically significant trends
495: present with either the stellar Teff or log g values.
496:
497: \section{Discussion}
498: The host stars studied here do not significantly deviate from
499: previously established galactic chemical evolution trends.
500: Instead, they can be regarded as being extensions of many of those
501: trends into metallicities above solar. This lack of deviation from
502: previously known trends strongly suggests that while they are more
503: metal enriched than other stars not known to host planets, the
504: host stars themselves have not systematically undergone any
505: extraordinary chemical processing during their growth and
506: evolution (in agreement with \citealt{rob}). The conclusion that
507: planetary hosts stars are essentially ``normal'' may indeed
508: suggest that planetary formation is a normal result of the star
509: formation process. Of course, this does not exclude planet
510: formation at lower metallicity values (as planets have been
511: detected orbiting stars with metallicities significantly below
512: solar) nor does it guarantee planet formation at high metallicity
513: values.
514:
515: There are two primary hypotheses that have been offered to explain
516: the observed high metallicity trend. The first is the
517: ``pollution'' model which posits metal-rich material being added
518: to the photosphere as a consequence of planetary formation
519: (\citealt{la,g6,mu}). The second explanation is commonly referred
520: to as the primordial model and it suggests that the gas cloud from
521: which these systems formed was metal enriched, resulting in the
522: star itself being enriched in the same elements (\citealt{s1}).
523: Our conclusion that these host stars exhibit normal chemical
524: evolution trends and that they are simply the metal-rich members
525: of field star population lends support to the primordial model.
526: One would expect that pollution of the stellar photosphere would
527: produce deviations from the galactic evolutionary trends. To date
528: no such trends have been observed. Additionally, we also observe
529: that the abundances of the more volatile elements (such as O)
530: increase with increasing metallicity for both the host and
531: non-host stars. This would not be the case in the pollution model
532: as it is likely that only the more refractory elements (such as Fe
533: and Ni) would remain in the solid form as they migrated towards
534: the star (and thus be deposited in the stellar photosphere) while
535: the more volatile elements would be evaporated before they could
536: be incorporated into the stellar photosphere. As such, we would
537: expect to see enrichment only in the refractory elements and not
538: the volatile elements if the pollution model is accurate.
539: Furthermore, the fact that we observe no trends with metallicity
540: in the orbital parameters of the remaining planets is difficult
541: for the pollution model to explain. It is hard to imagine a
542: situation whereby almost \textit{all} planetary host stars
543: accreted a significant amount of material during planetary
544: formation without affecting the orbital parameters of the
545: remaining planets. For these reasons, we agree with previous
546: studies (such as \citealt{s1,s2,s04,s05} and \citealt{fv}) and
547: support the primordial model for explaining the metal enrichment
548: in planetary host stars.
549:
550: The abundances reported here also impact on terrestrial planet
551: formation and evolution within these systems. Those with low Mg/Si
552: ratios will have terrestrial planets dominated by enstatite (with
553: a small amount of Mg-rich olivine also present) with other
554: Si-based species also available (a composition similar to the
555: Earth's crust), while those with high Mg/Si ratios will have
556: olivine-dominated planets with other Mg-rich species also present
557: (a composition similar to the Earth's mantle). Similarly, a high
558: C/O ratio will result in planets with greatly increased C contents
559: due to solid C being incorporated into the planet itself. While
560: the detailed consequences of these examples have not yet been
561: fully examined, it is conceivable that any terrestrial planets
562: forming in these systems could differ from currently known
563: terrestrial planets in terms of their rheology (thus possibly
564: affecting the tectonics of such a planet) and the nature of
565: volcanic activity, based on the varying silica contents of the
566: magma. The full implications of such a chemical composition for
567: the evolution of the terrestrial planets themselves is the subject
568: of ongoing research.
569:
570: \section{Summary}
571: In this paper, we have presented elemental abundances for 8
572: elements, including 5 heavy elements produced by the r- and
573: s-processes, in 28 planetary host stars and 90 non-host stars from
574: the AAPS. We conclude that while the elemental abundances of the
575: planetary host stars are only slightly different from solar
576: values, the host stars are enriched over the non-hosts stars in
577: all elements studied with the mean difference varying from 0.06
578: dex to 0.11 dex.
579:
580: Additionally, we also considered the trends of the abundances
581: (both [X/H] and [X/Fe]) with [Fe/H] and found these to be largely
582: in keeping with known galactic chemical evolution trends. This
583: implies that these systems have followed normal evolutionary
584: pathways and are not significantly or unusually altered. This
585: leads us to conclude that not only are the abundance trends we are
586: observing primordial in origin and represent the initial
587: composition of the gas nebula that produced the star and its
588: planets but that planetary formation may also be a natural
589: companion to the evolution of stellar material.
590:
591: \acknowledgments
592:
593: We would like to thank Sean Ryan for his assistance with this
594: study by providing the stellar atmosphere and abundance code,
595: along with incredibly patient help. We would also like to thank
596: the anonymous reviewer for their useful suggestions and advice.
597: The Anglo-Australian Planet Search team would like to acknowledge
598: support by the partners of the Anglo-Australian Telescope
599: Agreement (CGT and HRAJ); NASA grant NAG5-8299, NSF grant AST 95-
600: 20443 (GWM), NSF grant AST 99-88087 (RPB), PPARC grant
601: PP/C000552/1 (HRAJ, CGT, SJO) and ARC grant DP0774000 (CGT). This
602: research has made use of the SIMBAD database, operated at CDS,
603: Strasbourg, France, and the NASA Astrophysics Data System.
604:
605: \begin{thebibliography}{}
606: \bibitem[\protect\citeauthoryear{Allende Prieto et al.}%
607: {2004}]{ap} Allende Prieto, C., Barklem, P. S., Lambert, D. L. \& Cunha, K., 2004,
608: A\&A, 420, 183
609: \bibitem[\protect\citeauthoryear{Arlandini et al.}%
610: {1999}]{ar} Arlandini, C., K\"{a}ppeler, F., Wisshak, K., Gallino, R., Lugaro, M., Busso, M. \& Straniero, O., 1999,
611: ApJ, 525, 886
612: \bibitem[\protect\citeauthoryear{Asplund et al.}%
613: {2005}]{asp} Asplund, M., Grevesse, N. \& Sauval, A. J., 2005,
614: ASPC, 336, 25A
615: \bibitem[\protect\citeauthoryear{Bensby et al.}%
616: {2003}]{ben03} Bensby, T., Feltzing, S. \& Lundstr\"{o}m, I., 2003,
617: A\&A, 410, 527
618: \bibitem[\protect\citeauthoryear{Bensby et al.}%
619: {2005}]{ben} Bensby, T., Feltzing, S., Lundstr\"{o}m, I. \& Ilyin, I., 2005,
620: A\&A, 433, 185
621: \bibitem[\protect\citeauthoryear{Bond et al.}%
622: {2006}]{bond} Bond, J.C., Tinney, C. G., Butler, P. R., Jones, H. R. A., Marcy, G. W.,
623: Penny, Alan J. \& Carter, B. D., 2006, MNRAS, 370, 163
624: \bibitem[\protect\citeauthoryear{Bond et al.}%
625: {2008}]{bond2} Bond, J.C., Lauretta, D. S. \& O'Brien, D. P., 2008, LPSC XXXIX, Abstract 1438
626: \bibitem[\protect\citeauthoryear{Beri\~{a}o et al.}%
627: {2005}]{be} Beri$\tilde{a}$o, P., Santos, N. C., Isralian, G. \& Mayor, M. 2005, A \&
628: A, 438, 251
629: \bibitem[\protect\citeauthoryear{Bodaghee et al.}%
630: {2003}]{bo} Bodaghee, A., Santos, N. C., Isralian, G. \& Mayor, M. 2003, A \& A, 404, 715
631: \bibitem[\protect\citeauthoryear{Butler et al.}%
632: {1996}]{but96} Butler, P. R., Marcy, G. W., Williams, E.,
633: McCarthy, C. \& Vogt, S. S., 1996, PASP, 108, 500
634: \bibitem[\protect\citeauthoryear{Butler et al.}%
635: {2001}]{b4} Butler, P. R., Tinney, C. G., Marcy, Geoffrey W., Jones,
636: Hugh R. A., Penny, Alan J. \& Apps, Kevin, 2001, ApJ, 555, 410
637: \bibitem[\protect\citeauthoryear{Butler et al.}%
638: {2002}]{b5} Butler, P. R., et al., 2002, ApJ, 578, 565
639: \bibitem[\protect\citeauthoryear{Carter et al.}%
640: {2004}]{c4} Carter, B. D., Butler, R. P., Tinney, C. G., Jones, H. R. A., Marcy, G. W.,
641: McCarthy, C., Fischer, D. A. \& Penny, A. J., 2003, ApJ, 593, L43
642: \bibitem[\protect\citeauthoryear{Clayton}{1968}]{clay}
643: Clayton, D. D., 1968, Principles of Stellar Evolution and Nucelosynthesis, 1st edn. McGraw-Hill Book Company,
644: New York, pp. 557.
645: \bibitem[\protect\citeauthoryear{Den Hartog et al.}%
646: {2003}]{den} Den Hartog, E. A., Lawler, J. E., Sneden, C. \& Cowan, J. J., 2003,
647: ApJ Supp. Ser., 148, 543
648: \bibitem[\protect\citeauthoryear{Ecuvillon et al.}%
649: {2004}]{ec} Ecuvillon, A., Isralian, G., Santos, N. C., Mayor, M., Villar, V. \& Binhain, G. 2004, A \& A,
650: 426, 619
651: \bibitem[\protect\citeauthoryear{Ecuvillon et al.}%
652: {2006}]{ec06} Ecuvillon, A., Isralian, G., Santos, N. C., Shchukina, N. G., Mayor, M. \& Rebolo, R. 2006, A \& A,
653: 445, 633
654: \bibitem[\protect\citeauthoryear{Edvardsson et al}%
655: {1993}]{ed} Edvardsson, B., Andersen, J., Gustafsson, B., Lambert, D. L.,
656: Nissen, P. E. \& Tomkin, J., 1993,
657: A \& A, 275, 101
658: \bibitem[\protect\citeauthoryear{Fenner et al.}%
659: {2006}]{fen} Fenner, Y., Gibson, B. K., Gallino, R. \& Lugaro, M., 2006,
660: ApJ, 646, 184
661: \bibitem[\protect\citeauthoryear{Fischer \& Valenti}%
662: {2005}]{fv} Fischer, D. A. \& Valenti, J., 2005,
663: ApJ, 662, 1102
664: \bibitem[\protect\citeauthoryear{Gilli et al.}%
665: {2006}]{gilli} Gilli, G., Israelian, G., Ecuvillon, A., Santos, N. C. \& Mayor, M., 2006,
666: A\&A, 449, 723
667: \bibitem[\protect\citeauthoryear{Gonzalez}%
668: {1997}]{g1} Gonzalez, G., 1997,
669: MNRAS, 285, 403
670: \bibitem[\protect\citeauthoryear{Gonzalez}%
671: {1998}]{g2} Gonzalez, G., 1998,
672: A \& A, 334, 221
673: \bibitem[\protect\citeauthoryear{Gonzalez \& Laws}%
674: {2000}]{g3} Gonzalez, G. \& Laws, C., 2000,
675: AJ, 119, 390
676: \bibitem[\protect\citeauthoryear{Gonzalez \& Laws}%
677: {2007}]{gonz} Gonzalez, G. \& Laws, C., 2007,
678: MNRAS, 378, 1141, 1152
679: \bibitem[\protect\citeauthoryear{Gonzalez et al.}%
680: {2001}]{g6} Gonzalez, G., Laws, C., Tyagi, S. \& Reddy, B. E., 2001,
681: AJ, 121, 432
682: \bibitem[\protect\citeauthoryear{Gonzalez \& Vanture}%
683: {1998}]{g4} Gonzalez, G. \& Vanture, A. D., 1998,
684: A \& A, 339, L29
685: \bibitem[\protect\citeauthoryear{Gonzalez, Wallerstein \& Saar}%
686: {1999}]{g5} Gonzalez, G., Wallerstein, G. \& Saar, S. H., 1999,
687: ApJ, 511, L111
688: \bibitem[\protect\citeauthoryear{Hannaford et al.}%
689: {1982}]{han} Hannaford, P., Lowe, R. M., Grevesse, N., Bi\'{e}mont, E. \& Whaling, W., 1982,
690: ApJ, 261, 736
691: \bibitem[\protect\citeauthoryear{Huang et al.}%
692: {2005}]{hu} Huang, C., Zhao, G., Zhang, H. W. \& Chen, Y. Q., 2005,
693: MNRAS, 363, 71
694: \bibitem[\protect\citeauthoryear{Jenkins et al.}%
695: {2006}]{jen06} Jenkins, J. S., Jones, H. R. A., Tinney, C. G., Butler, R. P.,
696: McCarthy, C., Marcy, G. W., Pinfield, D. J., Carter, B. D. \& Penny, A. J.,
697: 2006, MNRAS, 372, 163
698: \bibitem[\protect\citeauthoryear{Jones et al.}%
699: {2002}]{j2} Jones, H. R. A., Butler, R. P., Tinney, C. G., Marcy, G. W.,
700: Penny, A. J., McCarthy, C. \& Carter, B. D., 2002, MNRAS, 333, 871
701: \bibitem[\protect\citeauthoryear{Jones et al.}%
702: {2003}]{j3} Jones, H. R. A., Butler, R. P., Marcy, G. W., Tinney, C. G.,
703: Penny, A. J., McCarthy, C., Carter, B. D. \& Bailey, J., 2006,
704: MNRAS, 369, 249
705: \bibitem[\protect\citeauthoryear{Jones et al.}%
706: {2006}]{j06} Jones, H. R. A., Butler, R. P., Tinney, C. G., Marcy, G. W.,
707: Carter, B. D., Penny, A. J., McCarthy, C. \& Carter, B. D., 2002, MNRAS, 333, 871
708: \bibitem[\protect\citeauthoryear{Kendall, Stuart \& Ord}{1987}]{kd}
709: Kendall, M. G., Stuart, A. \& Ord, J. K., 1987, Kendall's
710: Advanced Theory of Statistics, 5th edn. Oxford University Press,
711: New York
712: \bibitem[\protect\citeauthoryear{Kurucz}%
713: {1993}]{kr} Kurucz. R. L., 1993,
714: CD-ROMSs, ATLAS9 A Stellar Atmospheres Program and 2\,km.s$^{-1}$
715: Grid (Cambridge: Smithsonian Astrophys. Obs.)
716: \bibitem[\protect\citeauthoryear{Lanfranchi et al.}%
717: {2008}]{lan} Lanfranchi, G. A., Matteucci, F. \& Cescutti, G., 2008,
718: A\&A, 431, 635
719: \bibitem[\protect\citeauthoryear{Laughlin}%
720: {2000}]{la} Laughlin, G., 2000,
721: ApJ, 545, 1064
722: \bibitem[\protect\citeauthoryear{Marcy \& Butler}%
723: {1996}]{mb96} Marcy, G.W. \& Butler, R.P. 1996,
724: ApJ, 464, L147
725: \bibitem[\protect\citeauthoryear{McCarthy et al.}%
726: {2004}]{mc} McCarthy, C., et al., 2004, ApJ, 617, 575
727: \bibitem[\protect\citeauthoryear{Murray \& Chaboyer}%
728: {2002}]{mu} Murray, N. \& Chaboyer, B., 2002,
729: ApJ, 566, 442
730: \bibitem[\protect\citeauthoryear{O'Toole et al.}%
731: {2007}]{ot07} O'Toole, S. J., Butler, R. P., Tinney, C. G., Jones, H. R. A., Marcy, G. W.,
732: Carter, B. D., McCarthy, C., Bailey, J., Penny, A. P., Apps, A. \& Fischer, D.,
733: 2007, ApJ, 660, 1636
734: \bibitem[\protect\citeauthoryear{Pitts \& Newsom}%
735: {1986}]{pits} Pits, R. E. \& Newsom, G. H., 1986,
736: J. Quant. Spec. Rad. Transfer., 35, 383
737: \bibitem[\protect\citeauthoryear{Qian}%
738: {2004}]{qian} Qian, Y.-Z., 2004, Prog. Part. Nuc. Phys., 150, 153
739: \bibitem[\protect\citeauthoryear{Reddy et al.}%
740: {2003}]{red03} Reddy, B. E., Tomkin, J., Lambert, D. L. \& Allende Prieto, C., 2003,
741: MNRAS, 340, 304
742: \bibitem[\protect\citeauthoryear{Reddy et al.}%
743: {2006}]{red06} Reddy, B. E., Lambert, D. L. \& Allende Prieto, C., 2006,
744: MNRAS, 367, 1329
745: \bibitem[\protect\citeauthoryear{Reid}%
746: {2002}]{re} Reid, I. N., 2002,
747: PASP, 114, 306
748: \bibitem[\protect\citeauthoryear{Robinson et al.}%
749: {2006}]{rob} Robinson, S. E., Laughlin, G., Bodenheimer, P. \& Fischer, D., 2006,
750: ApJ, 643, 484
751: \bibitem[\protect\citeauthoryear{Ryan}%
752: {2005}]{r1} Ryan, S.G., 2005,
753: Personal Communication
754: \bibitem[\protect\citeauthoryear{Santos et al.}%
755: {2000}]{sb} Santos, N. C., Israelian, G. \& Mayor, M., 2000,
756: A \& A, 363, 228
757: \bibitem[\protect\citeauthoryear{Santos et al.}%
758: {2001}]{s1} Santos, N. C., Israelian, G. \& Mayor, M., 2001,
759: A \& A, 373, 1019
760: \bibitem[\protect\citeauthoryear{Santos et al.}%
761: {2003}]{s2} Santos, N. C., Israelian, G., Mayor, M., Rebolo, R. \& Udry, S., 2003,
762: A \& A, 398, 363
763: \bibitem[\protect\citeauthoryear{Santos et al.}%
764: {2004}]{s04} Santos, N. C., Israelian, G. \& Mayor, M., 2004,
765: A \& A, 415, 1153
766: \bibitem[\protect\citeauthoryear{Santos et al.}%
767: {2005}]{s05} Santos, N. C., Israelian, G., Mayor, M., Bento, J. P., Almeida, P. C., Sousa, S. G. \& Ecuvillon, A. 2005,
768: A \& A, 437, 1127
769: \bibitem[\protect\citeauthoryear{Simmerer et al.}%
770: {2004}]{sim} Simmerer, J., Sneden, C., Cowan, J. J., Collier, J., Woolf, V. M., \& Lawler, J. E., 2004,
771: ApJ, 617, 1091
772: \bibitem[\protect\citeauthoryear{Sneden}%
773: {1973}]{sn} Sneden, C., 1973,
774: PhD Thesis, University of Texas
775: \bibitem[\protect\citeauthoryear{Str\"{o}mgen}%
776: {1966}]{st} Str\"{o}mgen, B., 1966,
777: ARA\&A, 4, 433
778: \bibitem[\protect\citeauthoryear{Smith et al.}%
779: {2001}]{sm} Smith, V., Cunha, K. \& Lazzaro, D., 2001,
780: AJ, 121, 3207
781: \bibitem[\protect\citeauthoryear{Spite \& Spite}%
782: {1978}]{spite} Spite, M. \& Spite, F., 1978,
783: A\&A, 67, 23
784: \bibitem[\protect\citeauthoryear{Takeda}%
785: {2003}]{tak} Takeda, Y., 2003, A\&A, 4402, 343
786: \bibitem[\protect\citeauthoryear{Tinney et al.}%
787: {2001}]{tin1} Tinney, C. G., Butler, R. P., Marcy, G. W., Jones, H. R. A., Penny, A. J., Vogt, S. S.,
788: Apps, K. \& Henry, G. W., 2001, ApJ, 551, 507
789: \bibitem[\protect\citeauthoryear{Tinney et al.}%
790: {2002}]{tin3} Tinney, C. G., Butler, R. P., Marcy, G. W., Jones, H. R. A., Penny, A. J., McCarthy, C.
791: \& Carter, B. D., 2002, ApJ, 571, 528
792: \bibitem[\protect\citeauthoryear{Tinney et al.}%
793: {2003}]{tin2} Tinney, C. G., Butler, R. P., Marcy, G. W., Jones, H. R. A., Penny, A. J., McCarthy, C.,
794: Carter, B. D. \& Bond, J., 2003, ApJ, 587, 423
795: \bibitem[\protect\citeauthoryear{Tinney et al.}%
796: {2005}]{tin05} Tinney, C. G., Butler, R. P., Marcy, G. W., Jones, H. R. A., Penny, A. J., McCarthy, C.,
797: Carter, B. D. \& Fischer, D. A., 2005, ApJ, 623, 1171
798: \bibitem[\protect\citeauthoryear{Tinney et al.}%
799: {2006}]{tin06} Tinney, C. G., Butler, R. P., Marcy, G. W., Jones, H. R. A., Penny, A. J., Laughlin, G.,
800: Carter, B. D., Bailey, J. A. \& O'Toole, S., 2006, ApJ, 647, 594
801: \bibitem[\protect\citeauthoryear{Valenti \& Fischer}%
802: {2005}]{vf} Valenti, J. \& Fischer, D. A., 2005,
803: ApJ Supp. Ser., 159, 141
804: \bibitem[\protect\citeauthoryear{Wright et al.}%
805: {2007}]{w07} Wright, J. T., Marcy, G. W., Fischer, D. A., Butler, R. P., Vogt, S. S., Tinney, C. G.,
806: Jones, H. R. A., Carter, B. D., Johnson, J. A.,McCarthy, C. \& Apps, K., 2007, ApJ, 657, 533
807: \end{thebibliography}
808:
809: \clearpage
810:
811: \input{tab1}
812:
813: \clearpage
814: %Table 2 is for online publication
815: \input{tab2}
816:
817: \clearpage
818: %Table 3 is for online publication
819: \input{tab3}
820:
821: \clearpage
822: %Table 4 is for online publication
823: \input{tab4}
824:
825: \clearpage
826:
827: \input{tab5}
828:
829: \clearpage
830:
831: \input{tab6}
832:
833: \clearpage
834:
835: \begin{figure}
836: \plotone{f1.eps} \caption{[X/H] vs. [Fe/H] plots for all elements
837: studied. Open squares represent non-host stars and filled squares
838: represent host stars. Typical error bars are shown in the upper
839: left of each panel. \emph{Left Column:} O, Mg and Cr. \emph{Center
840: Column:} Y, Zr and Ba. \emph{Right Column:} Eu and Nd.}
841: \label{allh}
842: \end{figure}
843:
844: \clearpage
845:
846: \begin{figure}
847: \plotone{f2.eps} \caption{[X/Fe] vs. [Fe/H] plots for all elements
848: studied. Open squares represent non-host stars and filled squares
849: represent host stars. Typical error bars are shown in the upper
850: left of each panel. \emph{Left Column:} O, Mg and Cr. \emph{Center
851: Column:} Y, Zr and Ba. \emph{Right Column:} Eu and Nd.}
852: \label{allfe}
853: \end{figure}
854:
855: \clearpage
856:
857: \begin{figure}
858: \plotone{f3.eps} \caption{[heavy/light] vs. [Fe/H] plots for the
859: five heavy elements examined as part of this study. For the
860: definition of [heavy/light], please see the text. Open squares
861: indicate non-host stars, filled squares indicate host stars and
862: dashed lines indicate solar values.} \label{hl}
863: \end{figure}
864:
865: \clearpage
866:
867: \begin{figure}
868: \plotone{f4.eps} \caption{Orbital properties of extrasolar
869: planetary systems vs. abundance of the heavy elements. HD164427
870: has been omitted. {\em Left}: s-process elements Y (triangles), Zr
871: (squares) and Ba (crosses). {\em Right}: r- and mixed process
872: elements Eu (triangles) and Nd (squares). {\em From top to
873: bottom}: Msin{\em i} of the planet, orbital semi-major axis of the
874: planet, eccentricity of the orbit of the planet and period of the
875: planet.} \label{orbits}
876: \end{figure}
877:
878:
879:
880: \end{document}
881: