1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{amsmath,amsthm,amssymb}
3: \shorttitle{Light element production in SNe Ic}
4: \shortauthors{Nakamura et al.}
5:
6: \def\Msun{~M_{\odot} }
7: \def\Rsun{~R_{\odot} }
8:
9: \begin{document}
10:
11: \title{Light Element Production in the Circumstellar Matter of Energetic Type Ic Supernovae}
12:
13: \author{Ko Nakamura$^{1,2}$}
14: \author{Susumu Inoue$^3$}
15: \author{Shinya Wanajo$^1$}
16: \and
17: \author{Toshikazu Shigeyama$^1$}
18: \affil{$^1$Research Center for the Early Universe, Graduate School of Science,
19: $^2$Department of Astronomy, Graduate School of Science,
20: University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan\\
21: $^3$National Astronomical Observatory of Japan, Mitaka, Tokyo 181-8588, Japan
22: }
23:
24: \begin{abstract}
25: We investigate energetic type Ic supernovae as production sites for $^6$Li and Be
26: in the early stages of the Milky Way.
27: Recent observations have revealed that some very metal-poor stars with [Fe/H]$<-2.5$ possess unexpectedly high abundances of $^6$Li. Some also exbihit enhanced abundances of Be as well as N.
28: From a theoretical point of view, recent studies of the evolution of metal-poor massive stars
29: show that rotation-induced mixing can enrich the outer H and He layers with C, N, and O (CNO) elements, particularly N, and at the same time cause intense mass loss of these layers.
30: Here we consider energetic supernova explosions occurring after the progeniter star has lost
31: all but a small fraction of the He layer.
32: The fastest portion of the supernova ejecta can interact directly with the circumstellar matter (CSM),
33: both composed of He and CNO, and induce light element production through spallation and He-He fusion reactions.
34: The CSM should be sufficiently thick to energetic particles so that the interactions terminate
35: within its innermost regions.
36: We calculate the resulting $^6$Li/O and $^9$Be/O ratios in the ejecta$+$CSM material
37: out of which the very metal-poor stars may form.
38: We find that they are consistent with the observed values
39: if the mass of the He layer remaining on the pre-explosion core is $\sim 0.01-0.1 \Msun$,
40: and the mass fraction of N mixed in the He layer is $\sim 0.01$.
41: Further observations of $^6$Li, Be and N at low metallicity should provide critical tests
42: of this production scenario.
43: \end{abstract}
44:
45: \keywords{nuclear reactions, nucleosynthesis, abundances --- supernovae: general --- stars: abundances}
46:
47: \section{Introduction}
48: Among the light elements Li, Be and B (LiBeB),
49: $^7$Li is thought to arise from a variety of processes, including
50: big bang nucleosynthesis \citep{Spite82}, asymptotic giant branch stars, novae \citep{DAntona91},
51: and the $\nu$-process in type II supernovae \citep{Woosley90}; the latter may also contribute to $^{11}$B.
52: On the other hand, the main production channel for the rest,
53: in particular for $^6$Li and Be, is believed to be cosmic-ray induced nuclear reactions.
54: The most widely discussed models of LiBeB production are based on cosmic rays accelerated in supernova shocks
55: \citep{Meneguzzi71,Vangioni00}.
56: Observations of metal-poor stars in the Galactic halo
57: show a primary relation between [Fe/H] and [Be/H] or [B/H],
58: which is consistent with spallation by cosmic rays enriched with C, N, and O (CNO) from fresh SN ejecta
59: impinging on interstellar H or He \citep{Duncan92}.
60:
61: An alternative possibility was proposed by \citet{Fields96, Fields02},
62: who considered explosions of Type Ic supernovae (SNe Ic) as a site for primary LiBeB production.
63: Since it is expected that a fraction of C and O in the surface layers of the ejecta
64: are accelerated to energies above the threshold for spallation reactions when the shock passes through the stellar surface,
65: LiBeB production can occur through the direct interaction of the ejecta with the ambient material,
66: without the need for shock acceleration of cosmic rays.
67: This was explored in greater depth by \citet{Nakamura04},
68: who used more realistic stellar models and equations of state,
69: together with a 1-dimensional relativistic hydrodynamic code to investigate in detail
70: how much of the ejecta mass acquires sufficient energies for the LiBeB production.
71: All of these studies used stellar models which completely lost their H and He envelopes
72: and assumed the target to be interstellar matter (ISM) consisting of 90\% H and 10\% He,
73: ignoring any circumstellar matter (CSM).
74: Therefore, the only reactions under consideration were ${\rm C, O} + {\rm H, He} \rightarrow {\rm LiBeB}$.
75:
76: Recent observations by VLT/UVES \citep{Asplund05} and Subaru/HDS (Inoue et al. 2005, Aoki et al, in prep.)
77: have revealed that some very metal-poor stars possess surprisingly high abundances of $^6$Li,
78: much higher than expected from standard supernova cosmic ray models \citep{Ramaty00,Suzuki01,Prantzos06}.
79: The measured values are also higher than can be accommodated in the SN Ic production scenario discussed above,
80: even in the case of an energetic explosion similar to SN 1998bw, as can be seen from the results of \citet{Nakamura04}.
81: However, besides spallation of C and O, the fusion reaction ${\rm He + He} \rightarrow {\rm ^6Li}$
82: may also be potentially important.
83: This reaction will be significant in the conceivable case that
84: a small fraction of He is still remaining in the surface layer of the core at the time of the explosion,
85: while most of the He has been transported to the CSM through mass loss.
86: \citet{Meynet02} and \citet{Meynet06} have recently calculated the evolution of metal-poor massive stars
87: taking into account the effects of rotation-induced mixing, and shown that extensive mass loss occurs
88: even in extremely metal-poor cases,
89: in addition to significant enrichment of the N abundance near the surface.
90: If this N is accelerated at the shock breakout of the SN explosion,
91: a significant amount of Be can be produced through the reaction N + He $\rightarrow$ $^9$Be,
92: because of its low threshold $E_{\rm th}$ ($\sim 6$ MeV/A) and high cross section at peak $\sigma_{\rm p}$ ($\sim 24$ mb)
93: compared with other $^9$Be-producing reactions
94: (eg. $E_{\rm th} \sim 8$ MeV/A and $\sigma_{\rm p} \sim 8$ mb for O + He $\rightarrow$ $^9$Be, see Fig. \ref{fig-Ecs}).
95: Indeed, recent observations indicate that the abundances of both Be and N may be enhanced
96: in some metal-poor stars in which $^6$Li is detected \citep{Primas00a, Primas00b,Israelian04}.
97: In this Letter,
98: we focus on the interactions ${\rm He,CNO} + {\rm He, N} \rightarrow {\rm LiBeB}$
99: occurring between the ejecta accelerated by the energetic SN explosion and the CSM,
100: in order to account for the abundances in such very metal-poor stars.
101:
102: We summarize recent observations of the relevant elements in very metal-poor stars in \S \ref{sec-obs}.
103: In \S \ref{sec-SNIC} we discuss SN explosions of stars with stripped He layers embedded in the CSM.
104: Our calculations and results are shown in \S 4 with an appropriate parameter range.
105: Conclusions are presented in \S \ref{sec-disc}.
106:
107:
108: \section{Summary of Observations} \label{sec-obs}
109: Observational determination of the abundances of $^6$Li in metal-poor stars is extremely challenging
110: due to its deficiency and the proximity of its absorption lines to the much stronger ones of the $^7$Li isotope.
111: Owing to the advent of large telescopes and the improved capabilities of observational instruments,
112: reliable measurements have become possible only in the last several years \citep[for earlier observations, see][]{Smith98,Hobbs99,Cayrel99}.
113: \citet{Asplund05} observed 24 metal-poor halo stars with VLT/UVES and positively detected $^6$Li
114: in nine of them at the $\geq 2 \sigma$ significance level.
115: The subdwarf LP 815-43 is the most metal-poor object (${\rm [Fe/H]} = -2.74$) with $^6$Li in their sample.
116: Inoue et al. (2005) reported a tentative detection with Subaru/HDS
117: in the even more metal-poor star G 64-12 ([Fe/H]$=-3.17$),
118: although a detailed analysis is still ongoing (Aoki et al., in prep.).
119: With VLT/UVES, \citet{Primas00a, Primas00b} have measured the Be abundances of these two stars
120: to be $\log \varepsilon ({\rm Be}) = -1.09 \pm 0.20$ for LP 815-43 and $-1.10 \pm 0.15$ for G 64-12.
121: The Be abundance in G 64-12 is considerably higher than that expected from
122: previous measurements of the Be-Fe relation in stars with similar metallicities,
123: and this may be the case for LP 815-43 as well.
124: \citet{Israelian04} analyzed nitrogen abundances in 31 metal-poor stars and found that both LP 815-43 and G 64-12
125: are more N rich than average.
126: The abundances of the relevant elements in these two stars are listed in Table \ref{tbl-obs}.
127: We adopt the values obtained by 1-D LTE analyses to ensure consistency.
128:
129: \section{Metal-Poor SN Ic Embedded in CSM}\label{sec-SNIC}
130: \citet{Meynet06} simulated the evolution of metal-poor ([Fe/H]=$-6.6$, $-3.6$) massive ($60 \Msun$) stars
131: with rapid rotation (800 km/s) and found that C is enhanced in the outer layers by rotation-induced mixing,
132: promoting intense stellar winds and significant loss of their envelopes in spite of their initial low metallicities.
133: At the same time, the mixing results in enrichment of N in the He layers.
134: If a fraction of the He layer is still remaining when the SN explodes,
135: it is expected that the N within is accelerated at the shock breakout,
136: and a significant amount of Be will be produced through the reaction ${\rm N + He} \rightarrow {\rm ^9Be}$,
137: thanks to its low threshold and high cross section at peak.
138: Thus, we consider stars that have lost all of their H layer and most of their He/N layer before the explosion.
139: Here we use an explosion model of a $\sim 15 \Msun$ core, originating from a main-sequence star
140: with mass $M_{\rm ms} \sim 40 \Msun$ \citep{Nakamura01}.
141: The explosion energy is assumed to be $3 \times 10^{52}$ ergs, corresponding to an energetic explosion similar to SN 1998bw.
142: The mass of the ejecta becomes $13 \Msun$, containing $10 \Msun$ of oxygen.
143: The accelerated ejecta consisting of He and CNO will collide with the circumstellar He and N stripped from the progenitor star
144: and produce LiBeB through the reactions ${\rm He,CNO} + {\rm He, N} \rightarrow {\rm LiBeB}$.
145: The energy distribution of the C/O ejecta before interaction is that calculated in \citet{Nakamura04},
146: shown in Figure \ref{fig-Ecs} together with the cross sections for selected reactions \citep{Read84, Mercer01}.
147: Rather than recalculating the explosion hydrodynamics with the added He/N layer,
148: we approximate by changing the composition of the accelerated outermost ejecta from C/O to He/N.
149: This will not lead to considerable errors as long as the replaced mass is small.
150:
151: The ``thick target" approximation is used,
152: that is, the circumstellar He is so thick that light element production occurs entirely within the CSM
153: while the ejecta particles lose energy mainly by Coulomb collisions with free electrons.
154: This assumption is valid when the mass loss rate $\dot{M}$ is greater than $10^{-6}\,M_\odot$ yr$^{-1}$
155: for the typical wind velocity of $v_{\rm w} \sim$1,000 km s$^{-1}$,
156: as can be seen by comparing the windblown material's mass column density $\sigma$ with the stopping range $R$.
157: We obtain for an $\alpha$-particle with initial energy $\varepsilon$
158: \begin{equation}
159: \frac{R}{\sigma} \sim 0.096 \left( \frac{\varepsilon}{10{\rm MeV/A}} \right)^2 \left(\frac{\dot{M}}{10^{-6}\,M_\odot {\rm /yr}}\right)^{-1} \left(\frac{v_{\rm w}}{1,000\,{\rm km\,s}^{-1}}\right) \left(\frac{r}{R_\odot}\right),
160: \end{equation}
161: which is well below unity.
162: Here $R$ is defined as the column depth through which a particle loses all of its energy.
163: Even if the matter in the wind is not ionized, the energy loss rates due to ionization will be similar.
164: Other processes such as escape from the system can be ignored.
165: The LiBeB yields are calculated using the cross sections given by \citet{Read84} and \citet{Mercer01}.
166: We also assume that
167: the composition of the innermost CSM (i.e. the last wind material) concerned with the reactions is
168: the same as that of the progenitor's outermost layers.
169:
170: As a result of the localized production of the light elements,
171: their abundance ratios with respect to heavy elements averaged over the CSM and SN ejecta
172: are likely to be inherited by stars of the next generation, as pointed out by \citet{Shigeyama98}.
173: Thus in the next section we compare the abundance ratios calculated from the above model
174: with the metal-poor stars discussed in the preceding section.
175:
176: \section{Origin of Li and Be in Metal-poor Stars}
177: We focus on the abundances of the very metal-poor star LP 815-43
178: from which $^6$Li, Be and N have been detected, all at apparently enhanced levels.
179: As seen in Table \ref{tbl-obs}, $X_{\rm ^6Li}/X_{\rm O} \sim 6.88^{+3.08}_{-3.22} \times 10^{-7}$
180: and $X_{\rm ^9Be}/X_{\rm O} \sim 1.32^{+0.77}_{-0.49} \times 10^{-8}$ for this star.
181: The mass $M_{\rm He, \, N}$ of the He/N layer on the pre-explosion core
182: and its mass fraction $X_{\rm N}$ of N are the main parameters.
183: Figure \ref{fig-cal} shows our results.
184: The yields of $^6$Li increase until $M_{\rm He, \, N} \sim 0.01 \Msun$, which corresponds to the threshold energy
185: of the ${\rm He + He} \rightarrow {\rm ^6Li} $ reaction ($\sim 11$ MeV/A, see Fig. \ref{fig-Ecs}),
186: and saturates for larger $M_{\rm He, \, N}$.
187: They depend only slightly ($\sim 0.4$\%) on $X_{\rm N}$ (as long as $X_{\rm N}$ is not very large),
188: because most of the $^6$Li is produced through the He + He reaction.
189: Only the line corresponding to the case of $X_{\rm N} = 0.005$ is shown in the top panel of Figure \ref{fig-cal}.
190: On the other hand, the yields of $^9$Be, which is mainly produced through the ${\rm C, O + He} \rightarrow {\rm ^9Be}$ reaction when the He/N layer is deficient, depend strongly on $X_{\rm N}$.
191: Without N ($X_{\rm N} = 0$), the Be yield decreases rapidly with $M_{\rm He, \, N}$,
192: since the ${\rm He + He}$ reaction does not produce Be.
193: For small $X_{\rm N}$, Be yields first decrease with $M_{\rm He, \, N}$ for the above reason,
194: and then turn to increase since the cross section of the ${\rm N + He} \rightarrow {\rm ^9Be}$ reaction
195: peaks around $\sim 13$ MeV/A, corresponding to $\sim 0.013\Msun$ integrated from outside
196: for SN 1998bw model (see Fig. \ref{fig-Ecs}).
197: It should be noted that if the He layer is sufficiently large, for instance $M_{\rm He} \gtrsim 1 \Msun$,
198: our results here tend to overestimate the light element yields because a core with such a large He layer
199: might not be so compact in reality and effective acceleration at the shock breakout is not expected.
200: To specify how much He can be accelerated above the production threshold in such cases,
201: detailed calculations for the structures of mass-losing stars are required.
202:
203: Both ${\rm X_{^6Li}/X_O}$ and ${\rm X_{^9Be}/X_O}$ show good agreement with the observational data of LP 815-43
204: when $M_{\rm He, \, N} \sim 0.01-0.1 \Msun$ and $X_{\rm N} \sim 0.005-0.01$.
205: These are consistent with recent simulations of the evolution of metal-poor massive stars with rotation:
206: $X_{\rm N} \sim 0.008$ for $M_{\rm ms} = 40 \Msun$ (Hirschi et al., in preparation)
207: and $X_{\rm N} \sim 0.01$ for $M_{\rm ms} = 60 \Msun$ \citep{Meynet06}.
208: The observed ${\rm X_{^9Be}/X_O}$ for G64-12 can also be reproduced
209: with parameters analogous to those for LP 815-43.
210: The isotopic ratio ${\rm ^6Li/^7Li}$ in our models lies between 0.27 and 1.4,
211: which is much larger than the observed value $0.046 \pm 0.022$ for LP 815-43 \citep{Asplund05}
212: so that the scenario discussed here is not expected to contribute significantly to $^7$Li.
213:
214: \section{Conclusions and Discussion}\label{sec-disc}
215: In this Letter we have proposed a new mechanism to produce the light elements,
216: especially $^6$Li and Be recently observed in metal-poor stars.
217: This is based on recent theoretical findings that rotating, metal-poor massive stars
218: can lose substantial amounts of their envelopes,
219: and end up with SNe Ic with a small amount of He and N in the outermost layers.
220: The outer layers composed of He and CNO are accelerated at the shock breakout
221: and undergo spallation and He-He fusion reactions to produce mainly $^6$Li and $^9$Be.
222: Our calculations show that if $\sim 0.01-0.1\,M_\odot$ of the He layer is remaining on the core
223: in a SN explosion of a massive star with $M_{\rm ms}>30\,M_\odot$,
224: then the observed abundance ratios $^6$Li/O can be reproduced.
225: If the He layer contains a small amount ($\sim 0.5 - 1$\%) of N, we can also reproduce $^9$Be/O.
226:
227: Compared with the standard picture of cosmic ray shock acceleration by normal supernovae,
228: our high yield of $^6$Li results from three crucial differences.
229: First, we consider a large explosion energy, of order $10^{52}$ erg, as observed in some SNe such as SN 1998bw.
230: Second, our mechanism considers the direct interaction of fast ejecta with the CSM,
231: and does not involve an efficiency factor for cosmic ray acceleration
232: and is not affected by losses or escape during ISM propagation.
233: Third, the energy distribution of the accelerated particles is a very steeply falling power-law,
234: with spectral index $\sim 4.6$ as opposed to $2$ for shock accelerated particles.
235: This means that most of the energy is contained in the lowest energy portions,
236: where the cross sections for both ${\rm He + He} \rightarrow {\rm ^6Li}$ and ${\rm N + He} \rightarrow {\rm ^9Be}$ peak
237: (see Fig. \ref{fig-Ecs}).
238: In fact, the total energy of particles above 10 MeV/nucleon in our fiducial model
239: is $8\times10^{50}$ erg, much higher than is achievable in the standard picture, 10-30\% of $10^{51}$ erg.
240:
241: One may claim that $^6$Li, and probably also $^9$Be \citep{GPerez05},
242: in metal-poor stars may have been depleted from their initial values,
243: since the observed abundances of $^7$Li/H are a few times smaller than that predicted by standard big-bang nucleosynthesis,
244: and $^6$Li is more fragile than $^7$Li \citep{Asplund05}.
245: The actual survival fraction of $^6$Li is difficult to evaluate reliably
246: because the pre-main-sequence Li destruction is sensitive to convection,
247: which cannot be modeled without free parameters.
248: However, depletion factors as large as 10 may still be compatible with our picture.
249: The estimates in the preceding section is based on an explosion energy of the order of $10^{52}$ ergs.
250: The mass of ejecta with energy per nucleon above a certain value scales very strongly with the explosion energy,
251: $E_{\rm ex}$ as $E_{\rm ex}^{3.4}$ \citep{Nakamura04}.
252: Thus a SN with the explosion energy 2 times higher than that considered in the preceding section
253: will be able to produce $\sim 10$ times larger amounts of light elements depending on the distribution of He and N.
254: Nevertheless, if such very energetic explosions turn out to be rare,
255: the mechanism proposed here may play a rather limited role
256: compared to other potential LiBeB production processes.
257:
258: Both the degree of mass loss and the amount of N enrichment in the progenitor star
259: are expected to be sensitive to its initial rotation speed \citep{Meynet02, Meynet06}.
260: The actual rotation speed is presumably distributed over a wide range,
261: as is the explosion energy and the extent of mass loss at the time of the explosion.
262: Therefore, dispersions in the $^6$Li and Be abundances are expected,
263: and current observations suggest that this may indeed be the case
264: for $^6$Li \citep{Aoki04,Asplund05,Inoue05} as well as Be \citep{Boesgaard05}.
265: Our scenario predicts a close relation between Be and N, since Be arises directly as a consequence of N spallation.
266: A looser correlation between $^6$Li and N is also expected,
267: as effective $^6$Li production requires sufficient mass loss, which in turn implies significant N enrichment.
268: Further observations of $^6$Li, Be and N for a large sample of metal-poor stars should provide definitive tests.
269: Note that such correlations are not expected for other scenarios which involve mainly $^6$Li production only,
270: such as structure formation cosmic rays \citep{Suzuki02,Rollinde05},
271: active galactic nuclei outflows \citep{Prantzos06,Nath06},
272: and production processes in the early universe \citep{Jedamzik05}.
273: It is also mentioned that mass loss onto companion stars in binary systems may represent
274: an additional pathway to our scenario, provided that
275: sufficiently thick CSM is remaining at the time of the explosion.
276: We reiterate that more observational data is necessary to elucidate
277: what fraction of the LiBeB abundances seen in halo stars of different metallicity
278: can be explained by the mechanism proposed here.
279:
280: We are grateful to Georges Meynet, Raphael Hirschi, Takeru K. Suzuki and Sean Ryan for valuable discussions.
281: We also acknowledge the contributions of an anonymous referee,
282: which led to clarification of a number of points in our original manuscript.
283: This work has been partially supported by the grant in aid (16540213, 17740108) of the Ministry of Education, Science, Culture, and Sports in Japan.
284:
285:
286: \begin{thebibliography}{}
287: \bibitem[Akerman et al.(2004)]{Akerman04} Akerman, C.~J., Carigi,
288: L., Nissen, P.~E., Pettini, M., \& Asplund, M.\ 2004, \aap, 414, 931
289: \bibitem[Aoki et al.(2004)]{Aoki04} Aoki, W., Inoue, S.,
290: Kawanomoto, S., Ryan, S.~G., Smith, I.~M., Suzuki, T.~K., \& Takada-Hidai,
291: M.\ 2004, \aap, 428, 579
292: \bibitem[Asplund et al.(2005)]{Asplund05} Asplund, M., Lambert,
293: D.~L., Nissen, P.~E., Primas, F., \& Smith, V.~V.\ 2005, preprint (astro-ph/0510636)
294: \bibitem[Boesgaard \& Novicki(2005)]{Boesgaard05} Boesgaard, A.~M.,
295: \& Novicki, M.~C.\ 2005, \apj, in press (astro-ph/0512317)
296: \bibitem[Cayrel et al.(1999)]{Cayrel99} Cayrel, R., Spite, M.,
297: Spite, F., Vangioni-Flam, E., Cass{\'e}, M., \& Audouze, J.\ 1999, \aap,
298: 343, 923
299: \bibitem[D'Antona \& Matteucci(1991)]{DAntona91} D'Antona, F., \&
300: Matteucci, F.\ 1991, \aap, 248, 62
301: \bibitem[Duncan, Lambert, \& Lemke(1992)]{Duncan92}
302: Duncan, D.~K., Lambert, D.~L., \& Lemke, M.\ 1992, \apj, 401, 584
303: \bibitem[Fields et al.(1996)]{Fields96} Fields, B.~D., Casse,
304: M., Vangioni-Flam, E., \& Nomoto, K.\ 1996, \apj, 462, 276
305: \bibitem[Fields et al.(2002)]{Fields02} Fields, B.~D., Daigne,
306: F., Cass{\' e}, M., \& Vangioni-Flam, E.\ 2002, \apj, 581, 389
307: \bibitem[Garc{\'{\i}}a-P{\'e}rez \& Primas(2005)]{GPerez05}
308: Garc{\'{\i}}a-P{\'e}rez, A. E., \& Primas, F.\ 2005, \aap, in press (astro-ph/0512293)
309: \bibitem[Hobbs et al.(1999)]{Hobbs99} Hobbs, L.~M., Thorburn,
310: J.~A., \& Rebull, L.~M.\ 1999, \apj, 523, 797
311: \bibitem[Israelian et al.(2004)]{Israelian04} Israelian, G.,
312: Ecuvillon, A., Rebolo, R., Garc{\'{\i}}a-L{\'o}pez, R., Bonifacio, P., \&
313: Molaro, P.\ 2004, \aap, 421, 649
314: \bibitem[Inoue et al. (2005)]{Inoue05}
315: Inoue, S., Aoki, W., Suzuki, T. K., Kawanomoto, S., Garc{\'{\i}}a-P\'erez, A. E., Ryan, S. G. \& Chiba, M. 2005,
316: From Lithium to Uranium: Elemental Tracers of Early Cosmic Evolution, IAU Symp. 228,
317: ed. V. Hill et al., Cambridge University Press (2005) p.59
318: \bibitem[Jedamzik et al.(2005)]{Jedamzik05}
319: Jedamzik, K., Choi, K.-Y., Roszkowski, L., de Austri, R. R., astro-ph/0512044
320: \bibitem[Meneguzzi et al.(1971)]{Meneguzzi71}
321: Meneguzzi, M., Audouze, J. \& Reeves, H. 1971, \aap, 15, 337
322: \bibitem[Mercer et al.(2001)]{Mercer01} Mercer, D.~J., et al.\
323: 2001, \prc, 63, 065805
324: \bibitem[Meynet \& Maeder(2002)]{Meynet02} Meynet, G., \&
325: Maeder, A.\ 2002, \aap, 390, 561
326: \bibitem[Meynet et al.(2006)]{Meynet06} Meynet, G., Ekstr{\"o}m,
327: S., \& Maeder, A.\ 2006, \aap, 447, 623
328: \bibitem[Nakamura \& Shigeyama(2004)]{Nakamura04} Nakamura, K.~\&
329: Shigeyama, T.\ 2004, \apj, 610, 888
330: \bibitem[Nakamura et al.(2001)]{Nakamura01} Nakamura, T., Mazzali,
331: P.~A., Nomoto, K., \& Iwamoto, K.\ 2001, \apj, 550, 991
332: \bibitem[Nath et al.(2006)]{Nath06} Nath, B.~B., Madau, P., \&
333: Silk, J.\ 2006, \mnras, 366, L35
334: \bibitem[Prantzos(2006)]{Prantzos06} Prantzos, N.\ 2006, \aap,
335: 448, 665
336: \bibitem[Primas et al.(2000a)]{Primas00a} Primas, F., Molaro, P.,
337: Bonifacio, P., \& Hill, V.\ 2000a, \aap, 362, 666
338: \bibitem[Primas et al.(2000b)]{Primas00b} Primas, F., Asplund, M.,
339: Nissen, P.~E., \& Hill, V.\ 2000b, \aap, 364, L42
340: \bibitem[Ramaty et al.(2000)]{Ramaty00}
341: Ramaty, R., Scully, S., Lingenfelter, R. E. and Kozlovsky, B. 2000, \apj, 534, 747
342: \bibitem[Read \& Viola(1984)]{Read84} Read, S.~M., \& Viola,
343: V.~E.\ 1984, Atomic Data and Nuclear Data Tables, 31, 359
344: \bibitem[Rollinde et al.(2005)]{Rollinde05}
345: Rollinde, E., Vangioni, E. \& Olive, K. 2005, \apj, 627, 666
346: \bibitem[Shigeyama \& Tsujimoto(1998)]{Shigeyama98} Shigeyama, T.,
347: \& Tsujimoto, T.\ 1998, \apjl, 507, L135
348: \bibitem[Smith et al.(1998)]{Smith98} Smith, V.~V., Lambert,
349: D.~L., \& Nissen, P.~E.\ 1998, \apj, 506, 405
350: \bibitem[Spite \& Spite(1982)]{Spite82} Spite, F., \& Spite,
351: M.\ 1982, \aap, 115, 357
352: \bibitem[Suzuki \& Yoshii(2001)]{Suzuki01} Suzuki, T.~K., \&
353: Yoshii, Y.\ 2001, \apj, 549, 303
354: \bibitem[Suzuki \& Inoue(2002)]{Suzuki02} Suzuki, T.~K., \&
355: Inoue, S.\ 2002, \apj, 573, 168
356: \bibitem[Vangioni-Flam et al.(2000)]{Vangioni00}
357: Vangioni-Flam, E., Cass\'e, M. \& Audouze, J. 2000, \physrep, 333, 365
358: \bibitem[Woosley et al.(1990)]{Woosley90} Woosley, S.~E.,
359: Hartmann, D.~H., Hoffman, R.~D., \& Haxton, W.~C.\ 1990, \apj, 356, 272
360:
361:
362:
363: \end{thebibliography}
364:
365:
366: \clearpage
367: \begin{deluxetable}{lcccccc}
368: \tabletypesize{\scriptsize}
369: \tablecaption{Abundances from the literature. \label{tbl-obs}}
370: \tablewidth{0pt}
371: \tablehead{
372: \colhead{Star} & \colhead{[Fe/H]} &
373: \colhead{$\log \varepsilon ({\rm ^7Li})$} & \colhead{${\rm ^6Li/^7Li}$} &
374: \colhead{$\log \varepsilon ({\rm Be})$} &\colhead{$\log \varepsilon ({\rm N})$} &
375: \colhead{$\log \varepsilon ({\rm O})$}
376: }
377: \startdata
378: LP 815-43 & $-$2.74\tablenotemark{a} & 2.16\tablenotemark{b} &
379: 0.046$\pm$0.022\tablenotemark{b} & $-$1.09$\pm$0.20\tablenotemark{c} &
380: 5.61\tablenotemark{a} & 6.54\tablenotemark{c}\\
381: G 64-12 & $-$3.17\tablenotemark{d} & 2.30\tablenotemark{e} &
382: - & $-$1.10$\pm$0.15\tablenotemark{e} &
383: 6.15\tablenotemark{a} & 6.34\tablenotemark{d}\\
384:
385: \enddata
386:
387: \tablenotetext{a}{\citet{Israelian04}}
388: \tablenotetext{b}{\citet{Asplund05}}
389: \tablenotetext{c}{\citet{Primas00a}}
390: \tablenotetext{d}{\citet{Akerman04}}
391: \tablenotetext{e}{\citet{Primas00b}}
392:
393: \end{deluxetable}
394:
395:
396: \clearpage
397: \begin{figure}
398: \begin{center}
399: \epsscale{.80}
400: \plotone{f1.eps}
401: \caption{The energy distribution of ejecta and cross sections of reactions as functions of energy per nucleon $\varepsilon$. The solid line represents the result of a numerical calculation \citep{Nakamura04} where $M(>\varepsilon)$ denotes the mass of ejecta that have particle energy per nucleon greater than $\varepsilon$. The dotted, dashed, and dash-dotted lines show the cross sections $\sigma$ of reactions ${\rm He + He} \rightarrow {\rm ^6Li}$, ${\rm N + He} \rightarrow {\rm ^9Be}$, and ${\rm O + He} \rightarrow {\rm ^9Be}$, respectively.}
402: \label{fig-Ecs}
403: \end{center}
404: \end{figure}
405:
406:
407:
408: \clearpage
409: \begin{figure}
410: \begin{center}
411: \epsscale{.80}
412: \plotone{f2.eps}
413: \caption{Mass ratios of $^6$Li (top panel) and $^9$Be (bottom) to O as functions of $M_{\rm He,\, N}$. Shown are the cases with mass fractions of N $X_{\rm N} = 0.005$ (solid line), $X_{\rm N} = 0.01$ (dotted), $X_{\rm N} = 0.02$ (dash-dotted), and without N ($X_{\rm N}=0$; dashed). Shaded regions represent the observed ratios for the very metal-poor star LP 815-43 \citep{Primas00a, Asplund05} including the error.}
414: \label{fig-cal}
415: \end{center}
416: \end{figure}
417: \clearpage
418:
419: \end{document}
420:
421:
422: