1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8:
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12:
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: \documentclass[9pt,preprint2]{aastex}
19:
20: %% manuscript produces a one-column, double-spaced document:
21:
22: %\documentclass[manuscript]{aastex}
23:
24: %% preprint2 produces a double-column, single-spaced document:
25:
26: %\documentclass[preprint2]{aastex}
27:
28: %% Sometimes a paper's abstract is too long to fit on the
29: %% title page in preprint2 mode. When that is the case,
30: %% use the longabstract style option.
31:
32: %% \documentclass[preprint2,longabstract]{aastex}
33:
34: %% If you want to create your own macros, you can do so
35: %% using \newcommand. Your macros should appear before
36: %% the \begin{document} command.
37: %%
38: %% If you are submitting to a journal that translates manuscripts
39: %% into SGML, you need to follow certain guidelines when preparing
40: %% your macros. See the AASTeX v5.x Author Guide
41: %% for information.
42:
43: \newcommand{\vdag}{(v)^\dagger}
44: \newcommand{\myemail}{GuoliangLv@gmail.com}
45:
46: %% You can insert a short comment on the title page using the command below.
47:
48: %\slugcomment{Not to appear in Nonlearned J., 45.}
49:
50: %% If you wish, you may supply running head information, although
51: %% this information may be modified by the editorial offices.
52: %% The left head contains a list of authors,
53: %% usually a maximum of three (otherwise use et al.). The right
54: %% head is a modified title of up to roughly 44 characters.
55: %% Running heads will not print in the manuscript style.
56: %%This manuscript number is 73946 for ApJ
57: \shorttitle{Chemical Abundances in Symbiotic Stars}
58: \shortauthors{L\"{u} et al.}
59:
60: %% This is the end of the preamble. Indicate the beginning of the
61: %% paper itself with \begin{document}.
62:
63: \begin{document}
64:
65: %% LaTeX will automatically break titles if they run longer than
66: %% one line. However, you may use \\ to force a line break if
67: %% you desire.
68:
69: \title{Chemical Abundances in Symbiotic Stars }
70: \author{Guoliang L\"{u}\altaffilmark{1}$^\dagger$, Chunhua Zhu\altaffilmark{1,3},
71: Zhanwen Han\altaffilmark{2}, Zhaojun Wang\altaffilmark{1,3}}
72: \email{$^\dagger$GuoliangLv@gmail.com}
73: %% Notice that each of these authors has alternate affiliations, which
74: %% are identified by the \altaffilmark after each name. Specify alternate
75: %% affiliation information with \altaffiltext, with one command per each
76: %% affiliation.
77: \altaffiltext{1}{Department of Physics, Xinjiang University, Urumqi,
78: 830046, China.} \altaffiltext{2}{National Astronomical Observatories
79: / Yunnan Observatory, the Chinese Academy of Sciences, P.O.Box 110,
80: Kunming, 650011, China}\altaffiltext{3}{School of Science, Xi'an
81: Jiaotong University, Xi'an, 710049, China}
82:
83:
84: %\date{}
85:
86: %\pagerange{\pageref{firstpage}--\pageref{lastpage}} \pubyear{2007}
87:
88: %\maketitle
89:
90: %\label{firstpage}
91:
92: \begin{abstract}
93:
94: We have carried out a study of the chemical abundances of $^1$H,
95: $^4$He, $^{12}$C, $^{13}$C, $^{14}$N, $^{15}$N, $^{16}$O, $^{17}$O,
96: $^{20}$Ne and $^{22}$Ne in symbiotic stars (SSs) by means of a
97: population synthesis code. We find that the ratios of the number of
98: O-rich SSs to that of C-rich SSs in our simulations are between 3.4
99: and 24.1, depending on the third dredge-up efficiency $\lambda$ and
100: the terminal velocity of the stellar wind $v(\infty)$. The fraction
101: of SSs with $extrinsic$ C-rich cool giants in C-rich cool giants
102: ranges from 2.1\% to 22.7\%, depending on $\lambda$, the common
103: envelope algorithm and the mass-loss rate. Compared with the
104: observations, the distributions of the relative abundances of
105: $^{12}$C/$^{13}$C vs. [C/H] of the cool giants in SSs suggest that
106: the thermohaline mixing in low-mass stars may exist. The
107: distributions of the relative abundances of C/N vs. O/N, Ne/O vs.
108: N/O and He/H vs. N/O in the symbiotic nebulae indicate that it is
109: quite common that the nebular chemical abundances in SSs are
110: modified by the ejected materials from the hot components. Helium
111: overabundance in some symbiotic nebulae may be relevant to a helium
112: layer on the surfaces of white dwarf accretors.
113: \end{abstract}
114:
115: \keywords{ binaries: symbiotic--- accretion --- stars: AGB---
116: Galaxy: stellar content}
117:
118: \section{Introduction}
119: Symbiotic stars (SSs) are usually interacting binaries, composed of
120: a cool star, a hot component and a nebula. The cool component is a
121: red giant which is a first giant branch (FGB) or an asymptotic giant
122: branch (AGB) star. The hot component is a white dwarf (WD), a
123: subdwarf, an accreting low-mass main-sequence star, or a neutron
124: star \citep{kw84,m91,y95,it96}. In general, SSs are low-mass
125: binaries with an evolved giant transferring materials to a hot WD
126: companion which burns the accreted materials more or less steadily
127: \citep{m07}. According to the observed characteristics of their
128: specular spectra and photometries, SSs may stay in either the
129: quiescent phase or the outburst phase. During the quiescent phase,
130: SSs are undergoing a stable hydrogen burning on the surfaces of the
131: WD accretors. The outbursts may be due to either the thermonuclear
132: outbursts on the surfaces of accreting WDs which are called as
133: symbiotic novae \citep{ty76,pr80}, or accretion-disk instabilities
134: around a nearly steady burning WDs which are called as multiple
135: outbursts \citep{m05,m07}. Recent reviews of the properties of SSs
136: can be found in \citet{ms99} and \citet{m03,m07}.
137:
138: %A very important and attractive peculiarity of SSs is their specular
139: %spectral and photometric variability in time. The variability of SSs
140: %may be due to the nuclear burning on the surface of accreting WD
141: %\citep{pr80} or the variations in the accretion rate onto the WD
142: %\citep{m05}. In the majority of SSs, the hot component is, most
143: %probably, a WD and accretes the materials of the cool component via
144: %stellar wind. In general, these SSs are powered by hydrogen burning
145: %on the surface of WDs. According to \cite{ty76} and \cite{pr80},
146: %they can be divided into two subgroups: `ordinary' SSs which undergo
147: %a stable hydrogen burning, and the symbiotic novae which experience
148: %thermonuclear runaways in their surface hydrogen layers. Recent
149: %reviews of the properties of SSs can be found in \citet{ms99} and
150: %\citet{m03,m07}.
151: %The nova-like eruptions of SSs were reviewed by \citet{mk92}.
152:
153: The cool components usually have a high mass-loss rate in the SSs
154: with WD accretors. Their stellar winds have the chemical abundances
155: of the red giants. Parts of the stellar winds are ionized by the hot
156: components. The abundance determinations with nebular diagnostic
157: tools are possible \citep{s92}. SSs provide a good opportunity for
158: measuring the abundances of the red giants. The hot components may
159: support an additional high-velocity wind during the symbiotic nova
160: outbursts \citep{kw84}. High-velocity outflows have been observed in
161: very broad emission lines in essentially all the symbiotic novae
162: \citep{kt05}. \cite{lt94} analyzed classical nova abundances and
163: concluded that enhanced concentrations of heavy elements are
164: significant. Therefore, the abundances of the ejected materials from
165: the hot components in SSs may be different from those of the cool
166: giant stellar winds. However they are similar to those of the
167: classical novae. In general, the hot components have high luminosity
168: and effective temperature (They are similar to the central stars of
169: planetary nebulae. See Figure 5 in \citealt{m91}) so that they can
170: ionize the nebulae. In some eruptive SSs, the nebulae can also be
171: ionized by the region where the winds from the hot and cold
172: components collide \citep{w84}. It is difficult to determine the
173: origin of the nebulae in SSs. Based on emission line fluxes from
174: C$_{\rm III}$, C$_{\rm IV}$, N$_{\rm III}$, N$_{\rm IV}$ and O$_{\rm
175: III}$, \cite{n88} found that the CNO abundance ratios of the nebulae
176: in SSs are in the transition region from giants to supergiants and
177: concluded that the nebula is mainly due to mass lost by the red
178: giant. On the other hand, \cite{nv89} suggested that the flux ratios
179: of the emission line depend on the relative abundances of the two
180: winds in their study on Z And. \cite{vn92} found that CNO abundance
181: ratios of PU Vul are of the characteristic of novae.
182:
183: In short, SSs offer an exciting laboratory for studying novae, the
184: red giants and the interaction of the two winds. The chemical
185: abundances of the ejected materials from the hot components, the
186: stellar wind of the cool components and the symbiotic nebulae are
187: the key factor to understand them. Up to now, a series of
188: observational data of the chemical abundances in SSs have been
189: published. \citet{n88}, \citet{dc92}, \citet{cd94} and \citet{lc05}
190: gave the chemical abundances of some symbiotic nebulae. \citet{s92}
191: and \citet{s06} showed the chemical abundances of the cool giants in
192: several SSs. However, there are few theoretical studies about it.
193: Recently, \citet{kp97} and \citet{jh98} carried out detailed
194: numerical simulations for the chemical abundances of the novae with
195: CO WD or ONe WD accretors. This makes it possible to simulate the
196: chemical abundances of the symbiotic novae. \cite{gj93},
197: \cite{WG98}, \cite{k02}, \cite{i04}, \cite{I04} and \cite{mg07}
198: developed thermally pulsing asymptotic giant branch (TP-AGB)
199: synthesis by which the chemical abundances of the cool components
200: can be computed. \citet{lyh06} (hereafter Paper I) constructed the
201: models for the SSs' populations. According to the above results or
202: models, it is possible to set up a preliminary model for theoretical
203: study on the chemical abundances of SSs.
204:
205: In the present paper, the study of the chemical abundances of SSs
206: are carried out by means of a population synthesis code. In $\S$ 2
207: we present our assumptions and describe some details of the modeling
208: algorithm. In $\S$ 3 we discuss the main results and the effects of
209: different parameters. $\S$ 4 contains the main conclusions.
210:
211: \section{Model}
212: For the binary evolution, we use a rapid binary star evolution (BSE)
213: code of \cite{h02}. Below we describe our algorithm from several
214: aspects.
215: %Symbiotic nebulae may consist of the
216: %ejected mass from WDs and the stellar wind of cool giants. The
217: %chemical abundances of symbiotic nebulae can be affected by the
218: %chemical abundance of the cool giants and the chemical abundance of
219: %symbiotic nova. Below we describe our algorithm from several
220: %aspects.
221:
222: \subsection{Symbiotic Stars}
223: Paper I carried out a detailed study of SSs. Comparing the observed
224: distributions with the those predicted of the orbital periods and
225: the hot component masses in SSs, it is worth to mention that there
226: is remarkable disagreement between the observations and predictions.
227:
228: \cite{m07} showed the distributions of the measured orbital period
229: in about 70 SSs. The vast majority ($\sim 75\%$) of these systems
230: have orbital periods shorter than 1000 days while Paper I predicted
231: the distribution of longer orbital periods. The main reasons results
232: from two aspects:
233: \begin{enumerate}
234: \item It is hard to measure long orbital periods. All the SSs
235: with measured orbital periods in \cite{b00} and \cite{m03} are
236: S-type and about half of them are eclipsing binaries. Considering
237: the amplitude of radial velocity changes or eclipses, it is easier
238: to measure the short orbital periods. About 15 percent of 27 SSs in
239: \cite{m03} have longer orbital periods than 1000 days while it is
240: about 25 percent in \cite{m07}. There would be more SSs with long
241: orbital periods with further detailed observations. \item In Paper
242: I, SSs are detached binary systems and the process of mass transfer
243: results from the stellar wind of cool giants. To our knowledge, if
244: the mass ratio of the components ($q=M_{\rm donor}/M_{\rm
245: accretor}$) at onset of Roche lobe overflow (RLOF) is larger than a
246: certain critical value $q_{\rm c}$, the mass transfer is dynamically
247: unstable and results in the formation of a common envelop in short
248: time scale (about 100 years). The issue of the criterion for
249: dynamically unstable RLOF $q_{\rm c}$ is still open.
250: \citet{han01,han02} showed that $q_{\rm c}$ depends heavily on the
251: assumed mass-transfer efficiency. Paper I adopted $q_{\rm c}$ after
252: \cite{h02} in which the mass-transfer efficiency is 1. In Paper I,
253: $q_{\rm c}$ is usually smaller than 1.0. The observed $M_{\rm
254: giant}/M_{\rm WD}$s in SSs are between 2.0 and 4.0 and the predicted
255: those in Paper I are from 1.0 to 4.0, which means that there should
256: not be stable RLOF in SSs. However, \cite{m07} suggested that RLOF
257: can be quite common in symbiotic binaries with orbital periods
258: shorter than 1000 days. If this were true, the theoretical model of
259: mass transfer should be advanced. But this is out of the scope of
260: this paper.
261: \end{enumerate} The disagreement between the observed and the
262: predicted distributions of orbital periods in Paper I resulted from
263: observational biases and poor knowledge of the mass transfer
264: mechanism.
265:
266: \cite{m07} showed that most of SSs have WD masses less than 0.6
267: $M_\odot$, while Paper I predicted the distribution of higher WD
268: masses. The disagreement is due to the following:
269: \begin{enumerate}
270: \item For mass estimates, an orbital inclination of $i=90^0$
271: or a limit to $i$ (see table 2 in \citet{m03}) is typically assumed,
272: hence the estimates are lower limits. \item Paper I used the
273: mass-loss law suggested by \cite{vw93} for AGB stars. However,
274: \cite{m07} suggested that both the symbiotic giants and Miras have
275: higher mass-loss rates than single giants or field Miras,
276: respectively. Therefore, Paper I may have overestimated the hot
277: component masses. \end{enumerate}
278:
279: In short, due to poor knowledge of the mass loss from the giants and
280: the mass transfer mechanism in SSs, Paper I can only crudely predict
281: some observed characteristics. In this work, we accept still all the
282: criterions and the concepts on SSs in it. Following Paper I, we
283: assume that binary systems are considered as SSs if they satisfy the
284: following conditions:
285: \begin{itemize}
286: \item The systems are detached. \item The luminosity of the hot
287: component is higher than 10$L_\odot$ which is the `threshold'
288: luminosity for the hot component of SSs as inferred by \citet{m91}
289: and \citet{mk92}. This may be due to the thermonuclear burning
290: (including novae outbursts, stationary burning and post-eruption
291: burning). \item The hot component is a WD and the cool component is
292: a FGB or an AGB star.
293: \end{itemize} The liberation of gravitational energy by the accreted
294: matter may make the luminosity of the component larger than
295: 10$L_\odot$. A detailed accretion model of SSs was discussed in
296: Paper I. Here, we do not model it.
297:
298: All symbiotic phenomena in our work are produced by the hydrogen
299: burning on the WD surface. All SSs should stay in one of the
300: following three phases:
301: \begin{itemize}
302: \item the stable hydrogen burning phase;
303: \item the thermonuclear outburst phase (symbiotic nova);
304: \item the declining phase after a thermonuclear outburst.
305: \end{itemize} Paper I and the present paper do not
306: simulate the accretion disk in SSs. As mentioned in the
307: Introduction, the multiple outbursts due to the accretion-disk
308: instabilities usually occur around nearly steady burning WDs
309: \citep{m07}. In this work, they can be included in the quiescent SSs
310: which are undergoing the stable hydrogen burning.
311:
312: %All progenitors of SSs pass through one of the following three
313: %channels\citep{y95}:\\
314: %(i) unstable Roche lobe overflow (RLOF) with formation of a common
315: %envelope;\\ (ii) stable RLOF;\\
316: %(iii) formation of a WD+giant pair without RLOF.\\
317: %The detailed description on the evolutionary channels of SSs can be
318: %seen in \citet{y95}, \citet{it96} and Paper I.
319:
320: SSs are complex binary systems. There are many uncertain physical
321: parameters which can affect the population of SSs. Paper I showed
322: that the numbers of SSs and the occurrences of symbiotic novae are
323: greatly affected by the algorithm of common envelope evolution, the
324: terminal velocity of stellar wind $v(\infty)$ and the critical
325: ignition mass $\Delta M_{\rm crit}^{\rm WD}$ which is necessary mass
326: accreted by WDs for a thermonuclear runaway. The structure factor of
327: the stellar wind velocity $\alpha_{\rm W}$ and an optically thick
328: wind give a small uncertainty. In this work, we use the models of
329: SSs in Paper I and discuss the effects of the algorithm of common
330: envelope evolution, $v(\infty)$ and $\Delta M_{\rm crit}^{\rm WD}$
331: on the chemical abundances of SSs. Other
332: subordinate parameters are the same as in the standard model of Paper I.\\
333:
334: Common Envelope:\ For the common envelope evolution, it is generally
335: assumed that the orbital energy of the binary is used to expel the
336: envelope of the donor with an efficiency $\alpha_{\rm ce}$:
337: \begin{equation}
338: E_{\rm bind}=\alpha_{\rm ce}\Delta E_{\rm orb},
339: \label{eq:alpha}
340: \end{equation}
341: where $E_{\rm bind}$ is the total binding energy of the envelope and
342: $\Delta E_{\rm orb}$ is the orbital energy released in the
343: spiral-in. \citet{n20} suggested to describe the variation of the
344: separation of components in the common envelopes by an algorithm.
345: This algorithm is founded on the equation for the system orbital
346: angular momentum balance which implicitly assumes the conservation
347: of energy:
348: \begin{equation}
349: \frac{\Delta J}{J}=\gamma\frac{M_{\rm e}}{M+m}, \label{eq:gamma}
350: \end{equation}
351: where $J$ is the total angular momentum and $\Delta J$ is the change
352: of the total angular momentum during common envelope phase. In the
353: above formula, $M$ and $M_{\rm e}$ are respectively the mass and the
354: envelope mass of the donor, and $m$ is the companion mass. Following
355: \cite{nt05} and Paper I, we call the formalism of Eq.
356: (\ref{eq:alpha}) $\alpha$-algorithm and that of Eq. (\ref{eq:gamma})
357: $\gamma$-algorithm, which are respectively simulated in different
358: cases (see Table \ref{tab:case}). We take the `combined' parameter
359: $\alpha_{\rm ce}\lambda_{\rm ce}$ as 0.5 for $\alpha$-algorithm.
360: $\lambda_{\rm ce}$ is a structure parameter that depends on the
361: evolutionary stage of the donor.
362: For $\gamma$-algorithm, $\gamma=1.75$. \\
363:
364: $v({\infty})$:\ It is difficult to determine the terminal velocity
365: of stellar wind $v({\infty})$. \cite{w03} fitted the relation
366: between the mass-loss rates and the terminal wind velocities derived
367: from their CO(2-1) observation by
368: \begin{equation}
369: \log_{10} (\dot{M}/M_\odot{\rm yr}^{-1})=-7.40+\frac{4}{3}\log_{10}
370: (v({\infty})/{\rm km \, s^{-1}}). \label{eq:winters}
371: \end{equation}
372: The mass-loss rate is given by the formulation of \cite{vw93} or
373: \cite{b95} and $v(\infty)$ can be obtained by Eq.
374: (\ref{eq:winters}). However, Eq. (\ref{eq:winters}) is valid for
375: $\dot{M}$ close to $ 10^{-6}M_\odot$ yr$^{-1}$. For a mass-loss rate
376: higher than $ 10^{-6}M_\odot$ yr$^{-1}$, Eq. (\ref{eq:winters})
377: gives too high $v(\infty)$. Based on the models of \citet{win00}, we
378: assume $v(\infty)=30 {\rm km\, s^{-1}}$ if $v(\infty)\geq30 {\rm
379: km\, s^{-1}}$. In the standard model of Paper I,
380: $v({\infty})=\frac{1}{2}v_{\rm esc}$ where $v_{\rm esc}$ is the
381: escape velocity. In the present paper, we also carry
382: out various $v(\infty)$s simulations.\\
383:
384: $\Delta M_{\rm crit}^{\rm WD}$:\ The critical ignition mass of the
385: novae depends mainly on the mass of accreting WD, its temperature
386: and material accreting rate. \cite{y95} gave the `constant pressure'
387: expression for $\Delta M_{\rm crit}^{\rm WD}$ as
388: \begin{equation}
389: \frac{\Delta M_{{\rm crit}}^{{\rm WD}}}{M_\odot}=2\times
390: 10^{-6}\left(\frac{M_{{\rm WD}}}{R^4_{{\rm WD}}}\right)^{-0.8},
391: \label{eq:yunm}
392: \end{equation}
393: where $M_{\rm WD}$ is the mass of WD accretor and $R_{{\rm WD}}$ is
394: the radius of zero-temperature degenerate objects \citep{n72},
395: \begin{equation}
396: R_{{\rm WD}}=0.0112R_\odot[(M_{{\rm WD}}/M_{{\rm
397: ch}})^{-2/3}-(M_{{\rm WD}}/M_{{\rm ch}})^{2/3}]^{1/2},
398: \end{equation}
399: where $M_{{\rm ch}}=1.433M_{\odot}$ and $R_{\odot}=7\times 10^{10}$
400: cm. \citet{n04} gave numerical fits to the critical ignition masses
401: for novae models calculated by \citet{pk95}. In most simulations, we
402: adopt Eq. (\ref{eq:yunm}) for $\Delta M_{\rm crit}^{\rm WD}$.
403: However, in order to investigate the influences of the $\Delta
404: M_{\rm crit}^{\rm WD}$ on our results, we also carry out a
405: simulation for Eq. (A1) of \citet{n04} in which $\Delta M_{\rm
406: crit}^{\rm WD}$ depends on the mass accretion rates and masses of WD
407: accretors.
408: % We make a run of the
409: %code using the fitting formula for relatively cold ($T\sim 10^7$K)
410: %WD [see Eq. (A1) of \citet{n04} in which $\Delta M_{\rm crit}^{\rm
411: %WD}$ depends on the mass accretion rates and masses of WD
412: %accretors]. For the above two $\Delta M_{\rm crit}^{\rm WD}$s, Paper
413: %I gave a detailed comparison. In this paper, they are simulated in
414: %different cases (See Table \ref{tab:case}), respectively.
415:
416: %In fact, SSs are complex binary systems. There are many other
417: %uncertain physical parameters which can affect the population of
418: %SSs. We take them the same as the standard model of Paper I.
419:
420: \subsection{Chemical Abundances on the Surface of the Giant Stars}
421: \label{sec:abun} For a single star, three dredge-up processes and
422: hot bottom burning in a star with initial mass higher than
423: $4M_\odot$ may change the chemical abundances of the stellar
424: surface. We accept the prescriptions of \S 3.1.2 and 3.1.3 in
425: \cite{i04} for the first dredge-up during the first giant branch and
426: the second dredge-up during early asymptotic giant branch (E-AGB).
427: For the third dredge-up (TDU) and the hot bottom burning during the
428: TP-AGB phase, we use the TP-AGB synthesis in \cite{gj93},
429: \cite{k02}, \cite{i04}, \cite{I04} and \cite{mg07}. All details can
430: be found in Appendix A. In the present paper, we give the chemical
431: evolutions of $^1$H, $^4$He, $^{12}$C, $^{13}$C, $^{14}$N, $^{15}$N,
432: $^{16}$O, $^{17}$O, $^{20}$Ne and $^{22}$Ne on the stellar surface.
433:
434: %In \citet{z07}, the variation of the chemical abundances of the
435: %stellar surface are greatly affected by the minimum core mass
436: %($M^{\rm min}_{\rm c}$) for the third drudge-up and the third
437: %drudge-up efficiency $\lambda$.
438:
439: %{$M^{\rm min}_{\rm c}$:}\ The third drudge-up can only occur for the
440: %stars above a certain core mass $M^{\rm min}_{\rm c}$. \cite{k02}
441: %found that $M^{\rm min}_{\rm c}$ depends on stellar mass and
442: %metallicity and gave a fitting formula by
443: %5\begin{equation} M^{\rm min}_{\rm c}=a_1+a_2M_0+a_3M_0^2+a_4M_0^3
444: %\label{eq:mcmin}
445: %\end{equation}
446: %where coefficients are shown in Table 7 of \cite{k02} and $M_0$ is
447: %initial mass in solar unit. According to the observed the carbon
448: %luminosity function in the Magellanic clouds, \cite{mg07} consider
449: %that the $M^{\rm min}_{\rm c}$ predicted by \cite{k02} is higher.
450: %In our simulations, we take $M^{\rm min}_{\rm c} ={\rm
451: %Eq.({\ref{eq:mcmin}})}$ or $M^{\rm min}_{\rm c} =0.58 M_\odot$ in
452: %different cases.
453:
454: %{$\lambda$:}\ The third drudge-up efficiency is defined by
455: %\Delta M_{\rm dred}$ is the mass brought up to the stellar surface
456: %during a thermal pulse. $\lambda$ is a very uncertain parameter.
457: %\cite{k02} show a relation of $\lambda$:
458: %\begin{equation}
459: %\lambda(N)=\lambda_{\rm max}[1-\exp(-N/N_{\rm r})] \label{eq:lamb}
460: %\end{equation}
461: %where $\lambda$ gradually increases towards an asymptotic
462: %$\lambda_{\rm max}$ with $N$ (the progressive number of thermal
463: %pulsation) increasing. $N_{\rm r}$ is taken from Eq. (6) in
464: %\cite{mg07} which reproduces the results for $N_{\rm r}$ in Table 5
465: %of \cite{k02}. $\lambda_{\rm max}$ is given by (See Eq.(6) in
466: %\cite{k02}):
467: %\begin{equation}
468: %\lambda_{\rm max}=\frac{b_1+b_2M_0+b_3M^3_0}{1+b_4M_0^3}
469: %\label{eq:lambmax}
470: %\end{equation}
471: %where coefficients are shown in Table 8 of \cite{k02}. We also take
472: %$\lambda=0.5$ for a wide simulation.
473:
474: SSs in our work are binary systems. The binary mass transfer can
475: change the chemical abundances of the stellar surface. In binary
476: systems, there are two ways to transfer mass: (i)accretion from the
477: stellar wind material of a companion; (ii)Roche lobe overflow. BSE
478: model contains a standard Bondi-Hoyle type wind accretion
479: \citep{bh44} and a conservative mass transfer during the stable
480: Roche lobe overflow. After a star obtains $\Delta M$ from its
481: companion, the chemical abundances of the stellar surface $X_2$ is
482: given by
483: \begin{equation}
484: X^{\rm new}_2=\frac{X^{\rm old}_2\times M^{\rm
485: env}_2+X_1\times\Delta M}{M^{\rm env}_2+\Delta M},
486: \end{equation}
487: where $M^{\rm env}$ is the envelope mass and $X_1$ is the chemical
488: abundances of its companion.
489: \subsection{Chemical Abundances of the Ejected Materials from the Hot WD Accretors}
490: %Based on the typical characteristics of the observed systems, nova
491: %outbursts are usually divided into symbiotic novae and classical
492: %novae. The observational differences between the symbiotic novae and
493: %the classical novae may stem from the properties of the binary
494: %systems. \cite{it96} considered that the most significant
495: %differences in the temporal behavior of the classical novae and the
496: %symbiotic novae are probably entirely a result of the different
497: %sizes of their Roche lobes and differences in the abundances of the
498: %heavy elements in their envelopes. The physical nature of the
499: %classical novae and the symbiotic novae is identical. Therefore, we
500: %assume that the nova outbursts occurring in our binary systems
501: %selected for SSs are the symbiotic novae.
502: The hot WD accretors eject materials from their surfaces.
503: \cite{kp97} and \cite{jh98} showed detailed study on their chemical
504: abundances during the thermonuclear outbursts. There is no direct
505: observational evidence that SSs in the stable hydrogen burning phase
506: and the declining phase lead to winds from the hot components.
507: Theoretically, the hot companions have high luminosity and
508: temperature during the above two phases. Some parts of the materials
509: on the surface of the accreting WDs may be blown off due to high
510: luminosity \citep{it96}. To our knowledge, no literature refers to
511: them in detail yet. However, their chemical abundances are possibly
512: between those of the ejected materials during the thermonuclear
513: outbursts and the stellar winds from the giant stars. In the present
514: paper, we only give the chemical abundances of the ejected materials
515: during the thermonuclear outbursts.
516:
517: Based on the typical characteristics of the observed systems, nova
518: outbursts are usually divided into symbiotic novae and classical
519: novae. The classical novae can only last several days or weeks and
520: the variation of their visual magnitudes are between -5 and -10. The
521: symbiotic novae usually last several decades and the variation of
522: their visual magnitudes are between -3 and -8. Thermonuclear
523: runaways appear to be the most promising mechanism for classical
524: novae \citep{k86}. \cite{ty76} suggested that the above mechanism is
525: applicable to symbiotic novae. The observational differences between
526: the symbiotic novae and the classical novae may stem from the
527: properties of the binary systems \citep{it96}. The physical nature
528: of the classical novae and the symbiotic novae is identical.
529: Therefore, we assume that the thermonuclear outbursts occurring in
530: our binary systems selected for SSs are the symbiotic novae.
531:
532: Numerical studies of the nova outbursts have been carried out by
533: \citet{spk93,kp94,pk95,y05}. \cite{kp97} published detailed
534: multicycle calculations of the nova outbursts for CO WDs with mass
535: ranging from 0.65 to 1.4 $M_\odot$. In their simulations, the
536: element abundances of nova ejecta are affected by the four basic and
537: independent parameters: C/O ratio in the accreting WD, its core
538: temperature $T_{\rm WD}$, the mass accreting rate $\dot {M}_{\rm
539: WD}$ and the mass $M_{\rm WD}$. In order to test the effect of the
540: WD composition on the abundances of nova ejecta, \cite{kp97}
541: calculated the models with the accreting WDs composed of pure-C,
542: pure-O and C/O=1, respectively. They found that the WD composition
543: is not reflected in the abundances of ejecta. In an extensive study
544: of close binary evolution, \cite{it85} showed that the AGB phase may
545: be suppressed in a close binary and a WD formed in such a system
546: should have a ratio very close to unity. In our work, we assume that
547: the CO WD composition is C/O=1. A WD temperature of $10^7$ K
548: corresponds to an age of $10^9$ yr \citep{it84}. In fact, for most
549: of SSs, the time interval between the formation of the WD and the
550: beginning of the symbiotic stage may be longer than it, up to
551: $10^{10}$ yr \citep{y95}. Neglecting the effect of the nova
552: outbursts on the temperature of WD accretor $T_{\rm WD}$, we assume
553: that it is $10^7$ K. We select 10 models in \cite{kp97} in which
554: $T_{\rm WD}=10^7$ K and C/O ratio of the accreting WD is 1. Their
555: element abundances are determined by the mass accreting rate $\dot
556: {M}_{\rm WD}$ and $M_{\rm WD}$. By a bilinear interpolation
557: \citep{p92} of 10 models in \cite{kp97}, the abundances of $^1$H,
558: $^4$He, $^{12}$C, $^{13}$C, $^{14}$N, $^{15}$N, $^{16}$O and
559: $^{17}$O in the nova ejecta are calculated. If $\dot {M}_{\rm WD}$
560: or $M_{\rm WD}$ in SSs are not in the range of the bilinear
561: interpolation, they are taken as the most vicinal those in the 10
562: models.
563:
564: \cite{kp97} did not give the abundances of $^{20}$Ne and $^{22}$Ne.
565: \cite{jh98} gave the nucleosynthesis in the nova outbursts with CO
566: and ONe WDs. In their calculations, the nova nucleosynthesis are
567: affected by $M_{\rm WD}$, $\dot {M}_{\rm WD}$, the initial
568: luminosity (or $T_{\rm WD}$) and the degree of mixing between core
569: and envelope. To test the effect of the WD mass, they carried out a
570: number of simulations involving both CO WD ($M_{\rm WD}$=0.8, 1.0
571: and 1.15 $M_\odot$) and ONe ones ($M_{\rm WD}$=1.0, 1.15, 1.25 and
572: 1.35$M_\odot$). $\dot {M}_{\rm WD}$ is $2\times10^{-10} M_\odot$
573: yr$^{-1}$ and their initial luminosity is $10^{-2} L_\odot$. The
574: degree of mixing between the core and the envelope is a very
575: uncertain parameter. \cite{jh98} modeled three different mixing
576: levels: 25\%, 50\% and 75\%. Their results showed that higher mixing
577: degree favors the synthesis of higher metal nuclei in ONe WDs.
578: Following \cite{s98}, we adopt a 50\% degree of the mixing. For the
579: simulations of CO WDs in \cite{jh98}, we select the three nova
580: models (The degree of the mixing is 50\%) to calculate the
581: abundances of $^{20}$Ne and $^{22}$Ne by the fitting formulae:
582: \begin{equation}
583: \begin{array}{ll}
584: \log X({^{20}\rm Ne})=&-3.206+0.14476M_{\rm WD},\\
585: \log X({^{22}\rm Ne})=&-2.57118+0.60784M_{\rm WD}-0.33769M_{\rm
586: WD}^2,\\
587: \end{array}
588: \label{eq:kp97}
589: \end{equation}
590: where $M_{\rm WD}$ is in solar unit and the formulae agree with the
591: numerical results to within a factor of 1.1. In the above fits, the
592: abundances of $^{20}$Ne and $^{22}$Ne depend weakly on $M_{\rm WD}$.
593: The range of CO WD mass $M_{\rm WD}$ is from $\sim$ 0.5 to 1.4
594: $M_\odot$. The abundances of $^{20}$Ne calculated by Eq.
595: (\ref{eq:kp97}) are between $\sim$ $10^{-3.13}$ and $10^{-3.00}$,
596: and the abundances of $^{22}$Ne are between $\sim$ $10^{-3.35}$ and
597: $10^{-3.38}$. Therefore, we use Eq. (\ref{eq:kp97}) to calculate
598: $X({^{20}\rm Ne})$ and $X({^{22}\rm Ne})$ of all symbiotic novae
599: with the CO WDs. For the nova of ONe WD, we fit data of Table 3 in
600: \cite{jh98} for a 50\% degree of mixing by:
601: \begin{equation}
602: \begin{array}{ll}
603: \log X(^1{\rm H})=&-1.486+1.982M_{\rm WD}-0.992M_{\rm WD}^2,\\
604: \log X(^4{\rm He})=&-0.839-0.103M_{\rm WD}+0.197M_{\rm WD}^2,\\
605: \log X(^{12}{\rm C})=&-10.664+14.798M_{\rm WD}-6.025M_{\rm WD}^2,\\
606: \log X(^{13}{\rm C})=&-14.691+22.517M_{\rm WD}-9.607M_{\rm WD}^2,\\
607: \log X(^{14}{\rm N})=&4.504-11.174M_{\rm WD}+5.079M_{\rm WD}^2,\\
608: \log X(^{15}{\rm N})=&-2.196-2.313M_{\rm WD}+2.404M_{\rm WD}^2,\\
609: \log X(^{16}{\rm O})=&-9.835+17.609M_{\rm WD}-8.550M_{\rm WD}^2,\\
610: \log X(^{17}{\rm O})=&-14.492+22.095M_{\rm WD}-9.366M_{\rm WD}^2,\\
611: \log X(^{20}{\rm Ne})=&-0.536-0.192M_{\rm WD},\\
612: \log X(^{22}{\rm Ne})=&-21.20356+34.46454M_{\rm WD}-15.9763M_{\rm
613: WD}^2,\\
614: \end{array}
615: \label{eq:jh98}
616: \end{equation}
617: which agree with the numerical results to within a factor of 1.3.
618:
619: We neglect other elements because their abundances are much smaller
620: than the above elements or their isotopes. All abundances in the
621: nova ejecta are renormalized so that their sum is 1.0.
622:
623:
624: \subsection{Symbiotic Nebulae}
625: \label{sec:syne}
626:
627: In SSs, there may be two winds. One comes from the cool giant and
628: its chemical abundances are similar to those of the red giant
629: envelope and its velocity is between $\sim$ 5 and 30 km s$^{-1}$.
630: The other may come from the hot components. The components during
631: the symbiotic outbursts have the winds with a high velocity ($\sim$
632: 1000 km s$^{-1}$) although there is not a detailed description on
633: the wind from the hot components during the quiescent phase and the
634: declining phase \citep{k86}. Wind collision is inevitable in the
635: symbiotic outbursts. Recently, \cite{kt05,kt07} carried out a
636: detailed study for the colliding winds in SSs. According to their
637: simulations, the structure of the colliding winds (including
638: temperature and density) is very complicated. Symbiotic nebulae
639: should have a similar structure with the colliding winds at least
640: during the symbiotic outbursts.
641:
642: %According to \cite{p86}, the full cycle of a classical nova model
643: %has four processes: accretion, outburst, mass loss and decline. In
644: %Paper I, the symbiotic nova are simply cut into two stages: the
645: %`plateau' stage in which the WD has high luminosity; the decline
646: %stage in which the WD cools to the temperature so that its
647: %luminosity becomes lower than 10$L_\odot$. The lasting time of the
648: %`plateau' stage $t_{\rm on}$ and the time span of the decline stage
649: %$t_{\rm cool}$ are given by Eq. (30) and Eq. (32) of Paper I,
650: %respectively.
651: %In order to simulate the
652: %mass loss from the hot component, we divide roughly the symbiotic
653: %nova into three stages: mass loss stage; 'plateau' stage; decline
654: %stage. The lasting time of the 'plateau' stage $t_{\rm on}$ and the
655: %time span of the decline stage $t_{\rm cool}$ are given by Eq. (30)
656: %and Eq. (32) of Paper I, respectively. Based to \cite{pk95} and
657: %\cite{y05}, the timescale of WD ejecting the accreted mass during
658: %nova outbursts ($t_{\rm ml}$) is from about several days to hundreds
659: %of days. By a bilinear interpolation of Table 3 in \cite{y05},
660: %$t_{\rm ml}$ of the models in which $T_{\rm WD}$=10$^7$ K can be
661: %calculated. Compared with the sum of $t_{\rm on}$ (between about 1
662: %to several thousand year, see Figure 13 in Paper I) and $t_{\rm
663: %cool}$ ($\sim$ 1000 yr), $t_{\rm ml}$ is much short.
664: In order to avoid the difficulty of modeling the real symbiotic
665: nebulae and study the nebular chemical abundances by population
666: synthesis method, we assume roughly that the compositions of the
667: symbiotic nebulae undergo two independent phases:
668: \begin{itemize}
669: \item The symbiotic nebula is mainly composed of the ejected materials
670: from the hot WD when the thermonuclear runaway occurs and this phase
671: last for $t_{\rm on}$. The lasting time scale $t_{\rm on}$ on which
672: the hot components are in a `plateau' phase with high luminosity is
673: given by Eq. (30) of Paper I. At this phase, the symbiotic nebula
674: embodies the chemical characteristics of a nova. \item After $t_{\rm
675: on}$, the symbiotic nebula is mainly composed of the stellar wind
676: materials from the cool giants until the next symbiotic nova occurs.
677: During this phase, the symbiotic nebula embodies the chemical
678: characteristics of the cool giants in our models. For stable
679: hydrogen burning SSs, the symbiotic nebula is always composed of the
680: stellar wind materials from the cool giants.
681: \end{itemize}
682:
683:
684: However, some symbiotic novae have helium WD accretors. To our
685: knowledge, no literature refers to their chemical abundance.
686: According to Paper I, the occurrence rate of the symbiotic novae
687: with helium WDs is at most about $\frac{1}{75}$ of that with CO and
688: ONe WDs. We do not consider SSs with helim WD accretors in this
689: paper.
690:
691: \subsection{Basic Parameters of the Monte Carlo Simulation }
692: We carry out binary population synthesis via Monte Carlo simulation
693: technique in order to obtain the properties of SSs' population. For
694: the population synthesis of binary stars, the main input model
695: parameters are: (i) the initial mass function (IMF) of the
696: primaries; (ii) the mass-ratio distribution of the binaries; (iii)
697: the distribution of orbital separations; (iv) the eccentricity
698: distribution; (v) the metallicity $Z$ of the binary systems.
699:
700: We use a simple approximation to the IMF of \citet{ms79}. The
701: primary mass is generated using the formula suggested by
702: \citet{e89}
703: \begin{equation}
704: M_1=\frac{0.19X}{(1-X)^{0.75}+0.032(1-X)^{0.25}},
705: \end{equation}
706: where $X$ is a random variable uniformly distributed in the range
707: [0,1], and $M_1$ is the primary mass from $0.8M_\odot$ to
708: $8M_\odot$.
709:
710: For the mass-ratio distribution of binary systems, we consider only
711: a constant distribution \citep{m92,gm94},
712: \begin{equation}
713: n(q)=1,~~ 0< q \leq 1,
714: \end{equation}
715: where $q=M_2/M_1$.
716:
717:
718: The distribution of separations is given by
719: \begin{equation}
720: \log a =5X+1,
721: \end{equation}
722: where $X$ is a random variable uniformly distributed in the range
723: [0,1] and $a$ is in $R_\odot$.
724:
725: In our work, the metallicity $Z$=0.02 is adopted. We assume that all
726: binaries have initially circular orbits, and we follow the evolution
727: of both components by BSE code, including the effect of tides on
728: binary evolution \citep{h02}. We take $2\times10^5$ initial binary
729: systems for each simulation. Since we present, for every simulation,
730: the results of one run of the code, the numbers given are subject to
731: Poisson noise. For simulations with $2\times10^5$ binaries, the
732: relative errors of the numbers of the symbiotic systems in different
733: simulations are lower than 5\%. Thus, $2\times10^5$ initial binaries
734: appear to be an acceptable sample for our study.
735:
736: We assume that one binary with $M_1\geq 0.8 M_\odot $ is formed
737: annually in the Galaxy to calculate the birthrate of SSs
738: \citep{y94,h95a,h95b}.
739:
740: \begin{table*}
741: \begin{minipage}{170mm}
742: \caption{Parameters of the models of the population of SSs. The first column
743: gives the model number. Columns 2, 3, 4 and 5 show the mass-loss rate $\dot{M}$,
744: TDU efficiency $\lambda$, the minimum core mass for TDU
745: and the inter-shell abundance, respectively.
746: The detailed descriptions on the above parameters are given in Appendix A.
747: The 6th 7th and 8th columns show the algorithm of the common envelope,
748: the terminal velocity and the critical ignition mass, respectively.
749: N04 in column 8 means \citet{n04}.
750: }
751: \tabcolsep0.8mm
752: \begin{tabular}{llllllll}
753: \hline
754: Cases & $\dot{M}$&$\lambda$& $M_{\rm c}^{\rm min}$&Inter-Shell Abundances&Common Envelope&$v(\infty)$&$\Delta M_{\rm crit}^{\rm WD}$ \\
755: case 1& Eq.(\ref{eq:vwml})&Eq.(\ref{eq:lamb})&Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
756: case 2& Eq.(\ref{eq:bml}) &Eq.(\ref{eq:lamb})&Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
757: case 3& Eq.(\ref{eq:vwml})&0.5 &Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
758: case 4& Eq.(\ref{eq:vwml})&0.75 &Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
759: case 5& Eq.(\ref{eq:vwml})&Eq.(\ref{eq:lamb})&0.58$M_\odot$ & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
760: case 6& Eq.(\ref{eq:vwml})&Eq.(\ref{eq:lamb})&Eq.(\ref{eq:mcmin}) & \cite{i04} &$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
761: case 7& Eq.(\ref{eq:vwml})&Eq.(\ref{eq:lamb})&Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\gamma=1.75$ &Eq.(\ref{eq:winters}) &Eq.(\ref{eq:yunm})\\
762: case 8& Eq.(\ref{eq:vwml})&Eq.(\ref{eq:lamb})&Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&$\frac{1}{2}v_{\rm esc}$&Eq.(\ref{eq:yunm})\\
763: case 9& Eq.(\ref{eq:vwml})&Eq.(\ref{eq:lamb})&Eq.(\ref{eq:mcmin}) & \cite{mg07}&$\alpha_{\rm ce}\lambda_{\rm ce}=0.5$&Eq.(\ref{eq:winters}) &Eq.(A1) of N04\\
764: \hline
765: \label{tab:case}
766: \end{tabular}
767: \end{minipage}
768: \end{table*}
769: \section{Results}
770: We construct a set of models in which we vary different input
771: parameters relevant to the chemical abundances of SSs. Table
772: \ref{tab:case} lists all cases considered in this work. Many
773: observational evidences showed that the terminal velocity of stellar
774: wind $v(\infty)$ increases when a star ascends along the AGB
775: \citep{o02,w03,b05}. In this work, we take $v(\infty)$ calculated by
776: Eq. (\ref{eq:winters}) as the standard terminal velocity of stellar
777: wind. Case 1 is the standard model in this work. The results of SSs'
778: population are shown in Table \ref{tab:result}.
779: \begin{table*}
780: \begin{minipage}{170mm}
781: \caption{Different models of the SSs' population. The first column
782: gives the model number according to Table \ref{tab:case}.
783: Column 2 shows the birthrate of SSs in
784: the Galaxy. From columns 3 to 5, the number of O-rich SSs,
785: C-rich SSs and all SSs are given,
786: respectively. In the 3rd and 4th columns, the numbers in
787: parentheses mean the numbers of O-rich and C-rich SSs in
788: which we neglect the effects of the mass transfer in
789: binary systems on the chemical abundances on their
790: surfaces, respectively.
791: Columns 6 and 7 show the ratios of the number of SSs in the cooling
792: phase and the stable hydrogen burning phase to their total number, respectively.
793: Columns 8, 9 and 10 give the
794: occurrence rates of SyNe (symbiotic novae) with the accreting CO WDs, ONe WDs and
795: total rates. The number of SyNe with accreting CO
796: and ONe WDs are shown in columns 11 and 12, respectively.
797: }
798: \tabcolsep0.1mm
799: \begin{tabular}{ccccccccccccc}
800: \hline
801: \multicolumn{1}{c}{Cases}&\multicolumn{1}{c}{Birthrate}&\multicolumn{3}{c}{Number
802: of SSs}&\multicolumn{1}{c}{$N_{\rm
803: tcool}$}&\multicolumn{1}{c}{$N_{\rm
804: stable}$}&\multicolumn{3}{c}{Occurrence rate of
805: SyNe (yr$^{-1})$}&\multicolumn{2}{c}{Number of SyNe}\\
806: & of SSs (yr$^{-1}$)&O-rich&C-rich&Total&$\overline{N_{\rm total}}$&$\overline{N_{\rm total}}$&CO & ONe &Total&CO&ONe\\
807: 1&2&3&4&5&6&7&8&9&10&11&12\\
808: case 1& 0.119&4800 (4860)&870 (810)&5670& 0.57&0.25&3.8&0.2&4.0&1000&4\\
809: case 2& 0.118&5850 (6070)&1240(1020)&7100& 0.42&0.45&3.4&0.2&3.6&930 &3\\
810: case 3& 0.119&5340 (5390)&220 (170)&5560& 0.59&0.24&3.7&0.2&3.9&940 &3\\
811: case 4& 0.119&4930 (4970)&840 (800)&5770& 0.58&0.24&3.9&0.2&4.1&1000&4\\
812: case 5& 0.111&4330 (4410)&1220(1140)&5550& 0.56&0.26&3.8&0.2&4.0&1000&4\\
813: case 6& 0.119&4740 (4830&930 (840)&5660& 0.57&0.25&3.8&0.2&4.0&1000&4\\
814: case 7& 0.160&7900 (7930)&1440(1410)&9340& 0.49&0.33&5.3&0.4&5.7&1700&12\\
815: case 8& 0.072&2370 (2450)&700 (620)&3070& 0.28&0.60&1.1&0.1&1.2&370 &4\\
816: case 9& 0.115&7830 (7900)&1790(1720)&9620& 0.75&0.15&9.8&0.4&10.2&1000&4\\
817: \hline
818: \label{tab:result}
819: \end{tabular}
820: \end{minipage}
821: \end{table*}
822:
823:
824: \begin{figure}
825: \includegraphics[totalheight=3.in,width=2.8in,angle=-90]{f1.ps}
826: \caption{Initial-final mass relations. Thin and thick solid lines
827: are the core masses at the first thermal pulse in \citet{h00} and
828: in our work (See Eq. (\ref{eq:mc1tp})), respectively. Thin and thick dashed
829: lines are the final masses in \citet{h00} and case 1 in this work, respectively.
830: The dotted lines (from top to bottom) represent the relations from \citet{g00},
831: \citet{h95} and \citet{w00}, respectively. The dot-dashed line represent the relation from \citet{h94}.}
832: \label{fig:mc}
833: \end{figure}
834:
835: \subsection{Galactic Birthrate and the Number of SSs}
836: First, we discuss the gross properties of the modeled population of
837: SSs as those in Paper I. As Table \ref{tab:result} shows, the
838: Galactic birthrate of SSs may range from 0.072 (case 8) to 0.160
839: (case 7) yr$^{-1}$ and their number is from about 3070 (case 8) to
840: 9620 (case 9). The occurrence rate of the symbiotic novae is between
841: 1.2 (case 8) and 10.2 (case 9) yr$^{-1}$ and the number of the
842: symbiotic novae is from about 370 (case 8) to 1700 (case 7). The
843: contribution of the symbiotic novae with ONe WD accretors to the
844: occurrence rate is negligible in all cases, which is greatly
845: different from that in Paper I. The main reason is the different
846: stellar evolutions adopted during the TP-AGB phases.
847:
848: All assumptions in cases 1 and 8 are respectively the same as those
849: in case 7 and the standard model of Paper I except the stellar
850: evolutions during the TP-AGB phase. The greatest difference of the
851: TP-AGB phase between this work and Paper I is the initial-final mass
852: relations. It is well known that the relation is still very
853: uncertain. Fig. \ref{fig:mc} shows the initial-final mass relations
854: in different literatures. Using the mass-loss rate of \cite{vw93}
855: and assuming the classical core mass luminosity relation, \cite{w00}
856: gave an initial-final mass relation which is plotted in Fig.
857: \ref{fig:mc}. The relation for the more massive stars in \cite{g00}
858: is steep because TDU and hot bottom burning were not included. By
859: applying new observational data (including NGC2168, NGC2516, NGC2287
860: and NGC3532) and improved theory (stellar evolution calculations
861: including new opacities and main sequence overshoot; the extensive
862: mass grid of WD cooling sequences including models with thick and
863: thin He- and H-layer), \citet{h95} gave an initial-final mass
864: relation which has a less slope than that in \cite{g00} for the more
865: massive stars. Based on new stellar evolution calculations
866: (including TDU and hot bottom burning) and new observational data,
867: \cite{w00} showed an initial-final mass relation which has lower
868: final mass than that of \citet{h95} in the more massive stars. Using
869: a criterion for envelope ejection in AGB or FGB stars, \cite{h94}
870: obtained an initial-final mass relation which is very similar to
871: that in \cite{w00}. Paper I accepted the description of \citet{h00}
872: on TP-AGB evolution. The thin solid line and the dashed line in Fig.
873: \ref{fig:mc} represent the core mass at the first thermal pulse
874: $M_{\rm c, 1tp}$ and the initial-final mass relations in
875: \citet{h00}, respectively. In this paper, we obtain the final mass
876: by the synthesis TP-AGB evolutions (Detailed descriptions can be
877: seen in Appendix A ). The $M_{\rm c, 1tp}$ (See Eq. (8) in
878: \citealt{k02}) and the final mass are shown by the thick solid and
879: dashed lines in Fig. \ref{fig:mc}, respectively. The relation of
880: \citet{h00} is steeper than ours for $M_{\rm initial}> 4M_\odot$
881: (especially $M_{\rm initial}> 6M_\odot$) and gives a higher WD's
882: mass than ours. In Paper I, we assumed that the ONe WD originates
883: from a star with initial mass being between 6.1 $M_\odot$ and about
884: 8 $ M_\odot$\citep{p98}. In this work, we still keep the above
885: assumption. The masses of ONe WDs in this work are much lower than
886: those in Paper I. A small mass of the accreting WD is unfavorable
887: for producing the symbiotic phenomenon. Compared with case 7 and the
888: standard model in Paper I, the birthrate, the number of SSs and the
889: symbiotic novae decrease in cases 1 and 8. Especially, the
890: occurrence rate of the symbiotic novae with accreting ONe WDs in
891: case 1 is $\sim$ 3\% of that in case 7 of Paper I, and in case 8 it
892: is $\sim$ 5\% of that in the standard model of Paper I. Therefore,
893: the initial-final mass relation has great effects on the SSs'
894: population and especially on the symbiotic novae.
895:
896: Using the observational data in the last 25 years, \cite{it96}
897: estimated that the occurrence rate of symbiotic novae is $\sim$ 1
898: yr$^{-1}$ within a factor of 3. As Table. \ref{tab:result} shows,
899: the occurrence rate in case 1 is $\sim$ 4 and close to the
900: occurrence rate estimated by \cite{it96}. Under the same conditions,
901: case 7 in Paper I gave a higher occurrence rate of $\sim$ 11.9
902: yr$^{-1}$. Obviously, an steep initial-final mass relation like that
903: in \cite{h00} overestimates the occurrence rate of symbiotic novae.
904: This indicates that the initial-final mass relations from
905: \cite{h94}, \cite{w00} and the present paper are more reliable than
906: steep ones. However, due to the complicity of SSs, the above point
907: of view needs supports from further theoretical study and more
908: observational evidence.
909:
910: %\cite{r91} estimated that the frequency of occurrence rate of
911: %classical novae with ONe WDs in the classical novae observed is
912: %between 25\% and 57\%. However, \cite{lt94} adopted a more cautious
913: %approach to the identification of novae with ONe WDs. They
914: %considered that novae with ONe WDs should have a high abundance
915: %ratio of helium to hydrogen and, neon and heavy element enrichment.
916: %\cite{lt94} found the frequency of novae with ONe WDs in all
917: %classical novae observed is $\sim$ 11\%-17\%. If three other novae
918: %which exhibit significant neon enrichment relative to their total
919: %heavy-element concentration are included, the frequency in
920: %\cite{lt94} is $\sim$ 33\%. Having taken into account the
921: %observational selection effects introduced by the different
922: %recurrence times of the classical novae and by the spatial
923: %distribution of novae, \cite{g03} found that the frequency is $\sim$
924: %1/3. Unfortunately, to our knowledge, no literature refers to
925: %whether the WD accretors in the symbiotic novae observed are CO or
926: %ONe WDs. In Paper I, the frequency of occurrence rate of the
927: %symbiotic novae with ONe WDs in all symbiotic novae is between 35\%
928: %and 74\%, and it is 59\% in the standard model and 54\% in case 7.
929: %In this work, the frequency is between 3.9\% (case 9) and 8\% (case
930: %8), and it is 5\% in case 1. We consider a selection effect: more
931: %massive WD produce more luminous outbursts, thus favoring their
932: %detection as well (See Eq. (16) in Paper I). Stellar evolution
933: %theory clearly predicts that ONe WDs are more massive than CO WDs at
934: %birth. Taking into account the above selection effect, the
935: %frequencies estimated in the standard model and case 7 of Paper I
936: %should be respectively higher than 59\% and 54\%. If we assume that
937: %the frequency of occurrence rate of the symbiotic novae with ONe WDs
938: %in all symbiotic novae observed is compatible with that in classical
939: %novae which is between $\sim$ 11\% and 33\%, the frequency estimated
940: %in Paper I is too high. The frequency estimated in this work is
941: %acceptable. This indicates that the initial-final mass relations
942: %from \cite{h94}, \cite{w00} and the present paper are more reliable
943: %than others. However, this point of view needs further and more
944: %observational evidences. }
945: %For the star ($M_{\rm initial}\leq 2M_\odot$), $M_{\rm c, 1tp}$ and
946: %the final mass in \cite{p98} and \cite{h00} are lower than those in
947: %our work, but higher than in our work for the star ($M_{\rm
948: %initial}> 2M_\odot$). Especially, for the massive stars ($M_{\rm
949: %initial}> 4M_\odot$), the results in \cite{p98} and \cite{h00} are
950: %much higher than our results. In this work and Paper I, we assume
951: %that the ONe WD origins from the star whose initial mass is between
952: %6.1 $M_\odot$ and about 8 $M_\odot$ \citep{p98}. As Fig.
953: %\ref{fig:mc} shows, the masses of ONe WDs of this paper are lower
954: %about 0.1$M_\odot$ and up to 0.4$M_\odot$ than that in Paper I. We
955: %know that the final mass (it is WD mass) has a great effect on SSs'
956: %population (Paper I). Comparing the initial-final mass relations in
957: %Fig. \ref{fig:mc}, we find that the relations of \cite{p98} and
958: %\cite{g00} are steep than others and our relation is similar to that
959: %in \citet{w00}.
960:
961:
962:
963:
964: \subsection{The Effects of Input Parameters}
965: \label{sec:Parameters} In this section, we discuss the effects of
966: the input physical parameters on the SSs' population and the ratio
967: of the number of C-rich SSs in which the cool companions are C-rich
968: giants to that of O-rich SSs in which the cool companions are O-rich
969: giants. In Paper I, we analyzed the effects of the physical
970: parameters ( the common envelope algorithm, the terminal velocity of
971: the stellar wind $v(\infty)$ and the critical ignition mass of the
972: novae $\Delta M_{\rm crit}^{\rm WD}$ ) on the SSs' population. They
973: have a great effect on the SSs' population in our simulations, which
974: is similar to those in Paper I. Table \ref{tab:result} shows, the
975: {\it common envelope algorithm} and $\Delta M_{\rm crit}^{\rm WD}$
976: have weak effects on the ratio of the number of O-rich to that of
977: C-rich SSs while the $v(\infty)$ in case 8 takes an uncertainty
978: within a factor of about 1.6. With stars ascending along the AGB,
979: the $v(\infty)$ in case 8 decreases while more C-rich giants are
980: formed. A small $v(\infty)$ is favor for the formation of SSs.
981: Therefore, the ratio of O-rich to C-rich SSs in case 8 decreases.
982: The effects of other physical parameters
983: are following. \\
984: {\it Mass-loss rate}:
985: %Accorded to \cite{w07}, the average mass
986: %loss rate of \citet{vw93} during AGB phase is higher than that of
987: %\citet{b95}. The average lifetime of AGB in the case 2 is
988: Compared with \cite{vw93}(see Eq. (\ref{eq:vwml})), the mass loss
989: rate of \cite{b95}(see Eq. (\ref{eq:bml})) greatly depends on
990: stellar luminosity. When the cool giants ascend along the AGB, the
991: luminosities gradually increase. Therefore, case 2 gives a higher
992: mass-loss rate in the later phase during the AGB, which can increase
993: the SSs' number of the stable hydrogen burning while decrease the
994: SSs' number of the thermonuclear runaways. In general, C-rich giants
995: are formed at a later phase of the AGB by the third dredge-up. The
996: mass-loss rate in \cite{b95}(See Eq. (\ref{eq:bml})) enhances the
997: number of SSs and the number ratio of C-rich to O-rich SSs but
998: decreases the occurrence rate of the symbiotic novae and their
999: number. However, the effect is
1000: weak.\\
1001: {\it TDU efficiency $\lambda$ }:\ Comparing cases 1, 3 and 4, we
1002: find that $\lambda$ has a great effect on the ratio of O-rich to
1003: C-rich SSs' population. The effect of $\lambda$ on stellar evolution
1004: has mainly two aspects, {\it i.e.} the core mass evolution and the
1005: changes of the chemical abundances in stellar envelope. The average
1006: dredge-up efficiency in case 1 is higher than that in case 3
1007: ($\lambda=0.5$) and approximated to that in case 4 ($\lambda=0.75$).
1008: In cases 1, 3 and 4, the SSs' populations are almost the same.
1009: However, the number ratio of O-rich to C-rich SSs in case 3 is 4
1010: times more than the ratio in cases 1 or 4.
1011: \\
1012: {\it The minimum core mass for TDU $M_{\rm c}^{\rm min}$}: \ For the
1013: same initial masses, the $M_{\rm c}^{\rm min}$ of \cite{k02}(See Eq.
1014: (\ref{eq:mcmin})) is higher than 0.58$M_\odot$. In case 1, a star
1015: with an initial mass higher than $\sim$ 2.0$M_\odot$ can undergo the
1016: third dredge-up, but only 1.5$M_\odot$ in case 5. Therefore, C-rich
1017: SSs are more easily produced in case 5. Its effect is within a
1018: factor about 1.5.
1019: \\
1020: {\it The inter-shell abundances}: \ As cases 1 and 6 in Table
1021: \ref{tab:result} show, the two different inter-shell abundances have
1022: no effect on the SSs' populations. In \cite{i04}, $^{16}$O abundance
1023: of the inter-shell region is lower than that in \cite{mg07}. The
1024: number of C-rich SSs in case 6 increases. However, the effect of the
1025: inter-shell abundances on ratio of O-rich to C-rich SSs is very
1026: weak.
1027:
1028: In short, the mass-loss rate and the inter-shell abundances have
1029: very weak effect on the SSs' population and the ratio of O-rich to
1030: C-rich SSs. The TDU efficiency $\lambda$ has a very weak effect on
1031: the SSs' population, while the number ratio of O-rich to C-rich SSs
1032: have a great dependence on $\lambda$.
1033:
1034: %The physical parameters (mass-loss rate, $\lambda$, $M_{\rm c}^{\rm
1035: %min}$ and intershell abundances ) have a very weak effect on the
1036: %SSs' populations, but a great effect on the ratio of O-rich to
1037: %C-rich SSs up to a factor about 4 (case 3). Other physical
1038: %parameters (mass-loss rate, the common envelope algorithm, the
1039: %terminal velocity of stellar wind $v(\infty)$and the critical
1040: %ignition mass of the novae $\Delta M_{\rm crit}^{\rm WD}$ ) have
1041: %weak effects on the ratio of O-rich to C-rich SSs within a factor of
1042: %about 1.6 (case 7) but have great effects on the SSs' population up
1043: %to a factor of about 2, which can be also seen in Paper I.
1044:
1045: \subsection{C-rich SSs}
1046: %\citet{ms99} gave the spectral classification of the cool giants in
1047: %167 Galactic SSs. They found that there are 5 C-rich SSs and the
1048: %number ratio of O-rich to C-rich SSs is about 32.
1049: % According to the relation between the metallicity ([Fe/H]) and
1050: %$\log$ N(C)/N(M0+) of local group galaxies in \cite{ch03}, a small
1051: %fraction of the observed C-rich giants in SSs may indicate that the
1052: %symbiotic giants may be related to the metal-rich M giants, which is
1053: %supported by \cite{wm92}. They studied near infrared colors of a
1054: %large sample of SSs and suggested that their giants are similar to
1055: %those of the Bulge-like stars. However, \cite{s92} and \cite{s06}
1056: %analyzed the first-overtone CO absorption features in K-band spectra
1057: %of several SSs and believed that the surveyed symbiotic giants are
1058: %identical to local normal M giants. In order to definitely settle
1059: %this question, a direct high resolution spectroscopic determination
1060: %of the metallicities of the symbiotic giants is needed \citep{j03}.
1061: %In our simulations, the metallicity $Z$=0.02. From cases 1 to 9, the
1062: %ratios of O-rich to C-rich SSs are about 5.5, 4.9, 24.1, 5.8, 3.6,
1063: %5.2, 5.6, 3.4 and 4.3, respectively. In the solar neighborhood, the
1064: %ratio of the number of M-type giants to C-type giants N(M)/N(C) is
1065: %about 5 in \cite{g02}. The relation in \cite{ch03} gave N(M0+)/N(C)
1066: %$\sim$ 10 for the galaxies with [Fe/H]=0. Our results are
1067: %approximated to the above observational ratios of the giants except
1068: %case 3 with a low $\lambda$=0.5. The result of case 3 is more close
1069: %to the observational ratio of \citet{ms99}. However, according to
1070: %the typical TP-AGB synthesis \citep{gj93,k02,i04,mg07},
1071: %$\lambda$=0.5 may be too low for the solar-like stars.
1072:
1073:
1074:
1075: %We consider that the small fraction of the observed C-rich giants
1076: %in SSs may result from the terminal velocity of the stellar wind
1077: %$v(\infty)$. As discussion in \S \ref{sec:Parameters}, the
1078: %$v(\infty)$ has great effect on the ratio of O-rich to C-rich SSs.
1079: %For Eq. (\ref{eq:winters}), we assume artificially that the maximum
1080: %value of the $v(\infty)$ is 30 km s$^{-1}$. In order to check its
1081: %effects, we run a simulations in which $v(\infty)=\min(60 {\rm km\,
1082: %s^{-1}}, v(\infty))$. The ratio increase to 6.1. Using the fit of
1083: %$v(\infty)$ in \cite{w03}, we should note that the sample of
1084: %\cite{w03} has 65 sources but only two of them are C-rich giants.
1085: %Eq. (\ref{eq:winters}) may be unsuitable to the C-rich giants. Our
1086: %simulation may overestimate the C-rich SSs. In addition, some
1087: %C-stars with a high mass-loss rate may have thick dust shells so
1088: %that they are hard to identify, which may result in a small fraction
1089: %of the observed C-rich giants in SSs.
1090:
1091: \cite{b00} presented a catalogue of SSs which includes 188 SSs as
1092: well as 28 objects suspected of being SSs. According to the spectral
1093: types of the cool components listed by the catalogue, one can find
1094: out 5 C-rich SSs in 176 Galactic SSs and the number ratio of O-rich
1095: to C-rich SSs is about 35. In our simulations, from cases 1 to 9,
1096: the number ratios of O-rich to C-rich SSs are about 5.5, 4.9, 24.1,
1097: 5.8, 3.6, 5.2, 5.6, 3.4 and 4.3, respectively. Our results are lower
1098: than the above observational ratio. In all cases, the ratio of case
1099: 3 with a low $\lambda$=0.5 is the closest to the ratio observed in
1100: \cite{b00}. However, according to the typical TP-AGB synthesis
1101: \citep{gj93,k02,i04,mg07}, $\lambda$=0.5 may be too low for the
1102: solar-like stars. In the solar neighborhood, the number ratio of
1103: M-type giants to C-type giants is $\sim$ 5 in \cite{g02}, which is
1104: close to our results except for case 3. We consider that the small
1105: fraction of the observed C-rich giants in SSs may result from two
1106: respects: (i)The C-rich giants may have high terminal velocity of
1107: the stellar wind $v(\infty)$. When we use the fit of $v(\infty)$ in
1108: \cite{w03}, we should note that the sample of \cite{w03} has 65
1109: sources but only two of them are C-rich giants. Eq.
1110: (\ref{eq:winters}) may be unsuitable for the C-rich giants.
1111: Comparing cases 1 and 8, one can find that $v(\infty)$ has a great
1112: effect on the number ratio of O-rich to C-rich SSs. We may have
1113: underestimated $v(\infty)$ of C-rich giants. (ii)The cool giants in
1114: SSs have high mass-loss rates. In cases 1 and 2, we use the
1115: mass-loss rates of \cite{vw93} and \cite{b95}, respectively. The
1116: average mass-loss rate of the former is higher than that of the
1117: later during TP-AGB when the TDU occurs. The higher the mass-loss
1118: rate is, the more quickly the envelope mass decreases. According to
1119: Eq. (\ref{eq:tip}) in the present paper, the interpulse period
1120: $\tau_{\rm ip}$ increases with the envelope mass decreasing. A long
1121: $\tau_{\rm ip}$ reduces the TDU progressive number and the TDU
1122: efficiency $\lambda$. The large mass-loss rate is unfavorable for
1123: the formation of the carbon stars, which can be seen by comparing
1124: the number ratio of O-rich to C-rich SSs in case 1 with that in case
1125: 2. \cite{m07} suggested that both the symbiotic giants and Miras
1126: have higher mass-loss rates than single giants or field Miras,
1127: respectively. Therefore, the predicted number ratio of O-rich to
1128: C-rich SSs lower than the observational value may result from our
1129: underestimating the $v(\infty)$ of C-rich giants and the mass-loss
1130: rates of the cool giants in SSs.
1131: %We have not
1132: %consider a selection effects. Some C-stars with a high mass loss
1133: %rate may have thick dust shells so that they are hard to identify,
1134: %which may result in a small fraction of the observed C-rich giants
1135: %in SSs.
1136:
1137: Based on the formation channels, C-rich stars are classed into two
1138: types. The $intrinsic$ C-rich stars mean that the carbon observed in
1139: the atmosphere of the TP-AGB stars results from the third drudge-up.
1140: The $extrinsic$ C-rich stars are the giants polluted by carbon-rich
1141: matter from the former TP-AGB companion. In this work, we consider
1142: the effects of the mass transfer in binary systems on the chemical
1143: abundances on their stellar surfaces (See \S \ref{sec:abun}). In
1144: order to check the above effects, we also carry out the simulations
1145: in which the mass transfer between the two components can not vary
1146: the abundances on their surfaces. The results are shown in the
1147: parentheses of columns 3 and 4 of Table. \ref{tab:result}. The
1148: ratios of SSs with the $extrinsic$ C-rich giants to the total C-rich
1149: SSs from case 1 to case 9 are 6.8\%, 17.7\%, 22.7\%, 4.7\%, 6.6\%,
1150: 9.7\%, 2.08\%, 11.4\% and 3.9\%, respectively. The ratio is
1151: sensitive to the efficiency of TDU $\lambda$, the common envelope
1152: algorithm and the mass-loss rate. According to the discussions in \S
1153: \ref{sec:Parameters}, we know that a high mass-loss rate occurs in
1154: case 2 when carbon abundance is increased by the TDU. The
1155: $extrinsic$ C-rich giants in case 2 are formed more easy than those
1156: in case 1. A low $\lambda$ in case 3 can prevent the formation of
1157: C-rich stars including $intrinsic$ and $extrinsic$ C-rich stars. The
1158: high ratio in case 3 results from a small number of C-rich stars.
1159: Having undergone the $\alpha$-algorithm common envelope evolution,
1160: the orbital period of binary system becomes to be several percent of
1161: that before common envelope evolution. With $\gamma$-algorithm in
1162: case 7, post-common-envelope systems are wider than with the
1163: $\alpha$-algorithm in case 1 and this facilitates a symbiotic
1164: phenomenon, allowing more stars to evolve further along the FGB or
1165: AGB before the second RLOF (Paper I). In general,
1166: post-common-envelope systems hardly form the $extrinsic$ C-rich
1167: stars but can form $intrinsic$ C-rich stars. Therefore, the ratio of
1168: SSs with the $extrinsic$ C-rich giants to the total C-rich SSs in
1169: case 7 is very low. In most cases, the fraction of $extrinsic$
1170: C-rich SSs in total C-rich SSs is lower than 10\%. This indicates
1171: that the pollution of the carbon rich materials from the former
1172: TP-AGB companions on the cool components is weak so that most of
1173: C-rich giants in SSs are $intrinsic$ C-rich stars. \cite{m07} showed
1174: that all C-rich giants observed in SSs are $intrinsic$ C-rich stars,
1175: which is in agreement with our results.
1176: %The scarcity of the observed C-rich giants in SSs results possibly
1177: %from the following reasons: (i) Some C-stars with a high mass loss
1178: %rate may have thick dust shells so that they are harder to identify.
1179: %
1180: %Furthermore, \cite{ms99} showed the cool giants in SSs have a very
1181: %late spectral type.
1182:
1183: %\subsection{Chemical Evolution of Symbiotic Nebulae}
1184: %\begin{figure*}
1185: %\begin{tabular}{c@{\hspace{3pc}}c}
1186: %\includegraphics[totalheight=3in,width=2.5in,angle=-90]{cohis.ps}&
1187: %\includegraphics[totalheight=3in,width=2.5in,angle=-90]{onehis.ps}\\
1188: %Fig. \ref{fig:history}a &Fig. \ref{fig:history}b \\
1189: %\end{tabular} \caption{---Chemical abundances ( by number )
1190: %evolution of symbiotic nebulae in case 1. Fig. \ref{fig:history}a
1191: %represents the chemical evolution for SSs with the accreting CO WDs,
1192: %in which the initial masses of primary and secondary are
1193: %respectively 4$M_\odot$ and 3$M_\odot$ and their initial separation
1194: %is 45000$R_\odot$. Fig. \ref{fig:history}b is for SSs in which the
1195: %accreting ONe WDs with the initial masses of primary and secondary
1196: %are respectively 7$M_\odot$ and 3$M_\odot$ and a 45000$R_\odot$
1197: %initial separation.}
1198: % \label{fig:history}
1199: %\end{figure*}
1200: %Before discuss the chemical abundances of symbiotic nebulae, we
1201: %present an interesting evolutionary 'history' of two binary systems
1202: %that become SSs. Figs. \ref{fig:history}a and \ref{fig:history}b
1203: %show the chemical evolution of the symbiotic nebulae in SSs with the
1204: %accreting CO WD and ONe WD in case 1, respectively.
1205:
1206: %In Fig. \ref{fig:history}a, the initial conditions of binary system
1207: %are that primary and secondary masses are respectively 4$M_\odot$
1208: %and 3$M_\odot$ and their separation is 45000$R_\odot$. When the
1209: %binary system becomes SSs by a strong symbiotic nova, the primary is
1210: %a CO WD with 0.84$M_\odot$ and the secondary is in TP-AGB phase. For
1211: %an initial 3$M_\odot$ star, it usually undergoes the third dredge-up
1212: %during TP-AGB phase but no hot bottom burning. The rich chemical
1213: %abundances ( $^{4}$He, $^{12}$C) in the intershell region are
1214: %dredged up to the stellar surface. Therefore, the abundances of
1215: %$^{4}$He and $^{12}$C in the stellar wind obviously increase. During
1216: %the mass loss phase of the symbiotic novae, the chemical ingredients
1217: %of the symbiotic nebular embody the abundances of the novae. From
1218: %Fig. \ref{fig:history}a, we find that the chemical ingredients of
1219: %the novae are greatly different from those of the giants.
1220:
1221: %Fig
1222: %\ref{fig:history} shows their abundances enhancing and $^{1}$H
1223: %abundance descends. In general, due to the diffusion convection
1224: %mechanism, the $^{12}$C abundance during a nova of the accreting CO
1225: %WD is higher than that of an ordinary giant. However, the deep third
1226: %dredge-up can change it like shown Fig \ref{fig:history}.
1227:
1228: %Fig. \ref{fig:history}b represents the chemical abundances evolution
1229: %of the symbiotic novae, in which initial primary and secondary
1230: %masses are respectively 7$M_\odot$ and 3$M_\odot$ and their
1231: %separation is 45000$R_\odot$. When the binary system becomes SSs by
1232: %a strong symbiotic nova, the primary is a ONe WD with 1.02$M_\odot$
1233: %and the secondary is in TP-AGB phase. The greatest difference from
1234: %Figs. \ref{fig:history}a to b is the abundance of $^{20}$Ne during
1235: %the symbiotic novae, which results from the ingredient of the WDs.
1236: %Comparing \ref{fig:history}a and b, one should note that all
1237: %abundances of the accreting CO WDs are different from those of the
1238: %ONe WDs except $^{12}$C.
1239:
1240: \subsection{$^{12}$C/$^{13}$C vs. [C/H] of the Cool Giants in SSs}
1241: \begin{figure*}
1242: \begin{tabular}{c@{\hspace{3pc}}c}
1243: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f2a.ps}&
1244: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f2b.ps}\\
1245: %Fig. \ref{fig:cc}a &Fig. \ref{fig:cc}b \\
1246: \end{tabular}
1247: \caption{Gray-scale maps of the chemical abundance ratios of
1248: $\log$ $^{12}$C/$^{13}$C versus [C/H]. The gradations of gray-scale
1249: correspond to the regions where the number density of systems is,
1250: respectively, within 1 -- 1/2,
1251: 1/2 -- 1/4, 1/4 -- 1/8, 1/8 -- 0 of the maximum of
1252: ${{{\partial^2{N}}\over{\partial {\log {\rm ^{12}C/^{13}C}}}{\partial {\log
1253: {\rm [C/H]}}}}}$, and blank regions do not contain any stars.
1254: Solid lines mean $\log$ $^{12}$C/$^{13}$C=$\log$ [C/H]. The left panel of Fig.
1255: \ref{fig:cc} shows the chemical abundance ratios of the cool giants
1256: in SSs. The right panel of Fig. \ref{fig:cc} gives those of the cool giants
1257: in SSs, where $^{12}$C abundance of all giants with initial masses
1258: $\leq$ 2$M_\odot$ is artificially decreased by a factor
1259: of 2.5 after the first dredge-up.
1260: In the panel of observations, the filled big-stars, the empty squares and triangle represent
1261: the observational values in \cite{s92}, those in \cite{sm03} and those in \cite{s06}, respectively.
1262: }
1263: \label{fig:cc}
1264: \end{figure*}
1265:
1266: Using the infrared spectra of SSs, \cite{s92} and \cite{sm03} gave
1267: the C abundance and $^{12}$C/$^{13}$C isotopic ratio of the cool
1268: giants in 13 SSs with $\pm$0.3 errors. Using high-resolution
1269: near-infrared spectra and the method of the standard local thermal
1270: equilibrium analysis and atmospheric models, \cite{s06} calculated
1271: the abundance of the symbiotic star CH Cyg with the errors less than
1272: 0.3 dex for all the elements. Their observational data are shown in
1273: panels (10) of Fig. \ref{fig:cc}, where [C/H] means the relative
1274: abundances to the solar, that is, $[{\rm C/H}]={\rm \log C/H-\log
1275: C_\odot/H_\odot}$. The left panel of Fig. \ref{fig:cc} shows that
1276: $^{12}$C/$^{13}$C and [C/H] of the cool giants in SSs are much
1277: higher than those from the observations. \cite{s92} used the
1278: synthetic spectra calculated for M giants by \cite{l91} in which
1279: they analyzed M giants and obtained a low mean [C/H]=$-0.64\pm0.29$.
1280: \cite{l91} suggested that the standard mixing named as the first
1281: dredge-up (See \citealt{ir83}) is insufficient to explain
1282: atmospheric abundances in M giants. In our work, we use the standard
1283: mixing which only includes the first dredge-up. $^{12}$C abundance
1284: on stellar surface is reduced by approximately 30\% after the first
1285: dredge-up during the FGB \citep{ir83}.
1286:
1287: \cite{gs00} determined Li, C, N, O, Na and Fe abundances and
1288: $^{12}$C/$^{13}$C isotopic ratios of 62 metal-poor field giants.
1289: They suggested that there are two distinct mixing stages along the
1290: red giant branch evolution of small mass field stars: (i) The first
1291: dredge-up, that is, the standard mixing; (ii) The second mixing
1292: episode which is also called the thermohaline mixing in \cite{cz07}
1293: when the giants have a brighter luminosity. The thermohaline mixing
1294: can decrease $^{12}$C abundances. \citet{cz07} calculated several
1295: evolution models of a 0.9$M_\odot$ star and obtained the abundances
1296: which are consistent with the observational data in \cite{gs00}.
1297: Recently, \cite{e06,e07} confirmed a possible mechanism for such
1298: non-canonical mixing. Using 3D-modeling of a low-mass star at the
1299: tip of red giant branch, they found that the molecular inversion can
1300: lead to an efficient mixing and Pop I stars between 0.8 and 2.0
1301: $M_\odot$ develop $^{12}$C/$^{13}$C ratios of 14.5$\pm$1.5. The
1302: average of [C/H] in \cite{s92} is about -0.54 and the average
1303: $^{12}$C/$^{13}$C is about 16. It range in our simulations is
1304: between about -0.05 ( case 3) and 0.15 (case 4) for [C/H] and
1305: between about 23.4 (case 3) and 37.2 (case 5) for $^{12}$C/$^{13}$C,
1306: respectively. The main reason of the above disagreement is that the
1307: thermohaline mixing is not included in our simulation. In order to
1308: show the effects of the thermohaline mixing, we artificially
1309: decrease $^{12}$C abundance of all giants with initial masses $\leq$
1310: 2$M_\odot$ by a factor of 2.5 after the first dredge-up. The results
1311: are given in the right panel of Fig. \ref{fig:cc}. The averages of
1312: [C/H] are between -0.31 (case 3) and -0.102 (case 5) and the
1313: averages of $^{12}$C/$^{13}$C are between 12.3 (case 3) and 19.5
1314: (case 5). Our results are then close to the observations of
1315: \cite{s92}.
1316:
1317: We believe that the thermohaline mixing can occur and give
1318: significantly different results although the observational samples
1319: of $^{12}$C/$^{13}$C vs. [C/H] of the cool giants in SSs is small.
1320: However, due to poor knowledge on the thermohaline mixing, we can
1321: not carry out a detailed simulation. In the following sections, we
1322: do not discuss its effects any more.
1323: %The average [C/H] of the cool giants in our simulations is between
1324: %about -0.09 ( case 4 ) and 0.27 (case 7). \cite{k82} gave the C, N
1325: %and O abundance in G8-K3 giants and their [C/H] is about -0.28. The
1326: %[C/H] in the planetary nebulae of \citet{pb06} is about 0.12.
1327: \subsection{Chemical Abundance of Symbiotic Nebulae}
1328: In this subsection, we discuss the chemical abundances of the
1329: symbiotic novae. In order to clarify the nature of the symbiotic
1330: nebulae, we need the observational abundances of novae and planetary
1331: nebulae. In this paper, the nova abundances come from \cite{lt94} in
1332: which they analyzed the abundances of 18 classical novae with a
1333: factor of $\sim$ 2-4 uncertainty. The planetary nebulae abundances
1334: are obtained from \cite{pb06} in which they gave the chemical
1335: abundance of 26 planetary nebulae within a 30\% uncertainty in the
1336: abundance determination.
1337: %and an additional uncertainty in the galactocentric
1338: %distance.
1339:
1340:
1341:
1342: \subsubsection{\rm C/N vs. O/N}
1343: \begin{figure*}
1344: \begin{tabular}{c@{\hspace{3pc}}c}
1345: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f3a.ps}&
1346: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f3b.ps}\\
1347: %Fig. \ref{fig:co}a &Fig. \ref{fig:co}b \\
1348: \end{tabular}
1349: \caption{Gray-scale maps of the chemical abundance ratios of
1350: $\log$ C/N versus $\log$ O/N. The gradations of gray-scale
1351: correspond to the regions where the number density of systems is,
1352: respectively, within 1 -- 1/2,
1353: 1/2 -- 1/4, 1/4 -- 1/8, 1/8 -- 0 of the maximum of
1354: ${{{\partial^2{N}}\over{\partial {\log {\rm C/N}}}{\partial {\log
1355: {\rm O/N}}}}}$, and blank regions do not contain any stars.
1356: Solid lines mean $\log$ C/N=$\log$ O/N. The left and right panels of Fig.
1357: \ref{fig:co} show the chemical abundances of the giants during the symbiotic stage
1358: and the ejected materials from the hot companions during the symbiotic novae, respectively.
1359: In the panels of observations, the filled big-stars, the
1360: filled triangles and the squares represent the observational values of the symbiotic nebulae in \citet{n88},
1361: those of PU Vul in \citet{vn92} (including C$_{\rm IV}$ and N$_{\rm IV}$ or excluding
1362: them) and those of Cyg, HM Sge and HBV 475 in
1363: \cite{ss90}, respectively. The circles in the left and right panels
1364: (10) of Fig. \ref{fig:co} represent the observational values
1365: of the planetary nebulae in \citet{pb06} and the novae in \citet{lt94}, respectively. }
1366: \label{fig:co}
1367: \end{figure*}
1368:
1369: Based on UV data, \cite{n88} presented C/N and O/N abundance ratios
1370: of 24 symbiotic nebulae. The errors in the logarithmic (basis of 10)
1371: abundance ratios remain within 0.18. In their sample, there are at
1372: least three symbiotic novae V1016 Cyg, HM Sge and HBV 475. V1016 Cyg
1373: and HBV 475 had their outbursts around 1965 while HM Sge brightened
1374: by 5 mag in 1975. \cite{ss90} calculated their abundance ratios
1375: within an error of 30\%. Their C/N and O/N abundance ratios are
1376: different from those in \cite{n88} within a factor of 30\%.
1377: According to $IUE$ observations, HM Sge is still evolving towards
1378: higher excitation whereas V1016 Cyg and HBV 475 seem to have reached
1379: their maximum excitation \citep{nv90}. The outburst of PU Vul began
1380: late in 1977. \cite{vn92} gave the C/N and O/N abundances ratios in
1381: the early nebular phase. The observational results and our simulated
1382: results are shown in Fig. \ref{fig:co}.
1383:
1384:
1385: The left and right panels (1)--- (9) of Fig. \ref{fig:co} give the
1386: distribution of the relative CNO abundances of the cool companions
1387: and the ejected materials from the hot components during symbiotic
1388: novae, respectively. Panels (10) in Fig. \ref{fig:co} show the
1389: observational ratios of C/N vs. O/N of symbiotic nebulae, novae and
1390: planetary nebulae.
1391:
1392: The left panel of Fig. \ref{fig:co} shows that we can cut the
1393: abundance ratios of C/N vs. O/N of the cool companions in SSs into
1394: three regions. i) The region of C/N $>$ O/N represents that the cool
1395: companions have undergone the deep TDU and have not undergone the
1396: hot bottom burning. The initial masses of the cool companions should
1397: be between about 2.0 or 1.5(case 5) $M_\odot$ and 4.0 $M_\odot$.
1398: Their abundances are similar to those of C-stars and planetary
1399: nebulae; ii) The top region of C/N $<$ O/N represents that the cool
1400: companions have undergone inefficient TDU or have not undergone TDU.
1401: These stars have initial masses lower than 2.0 or 1.5 (case 5)
1402: $M_\odot$. Their abundances are similar to those of M-stars; iii)
1403: The bottom region of C/N $<$ O/N represents that the cool companions
1404: undergo TDU and the hot bottom burning which turns $^{12}$C to
1405: $^{14}$N. Their initial masses are higher than 4.0$M_\odot$.
1406:
1407: In the right panel of Fig. \ref{fig:co}, the distribution of C/N vs.
1408: O/N of the ejected materials from the hot companions during
1409: symbiotic novae is composed of two regions. The top region
1410: represents symbiotic novae with accreting ONe WDs. The bottom region
1411: represents those with accreting CO WDs. Compared with the
1412: observations, our results agree with those in \cite{lt94}.
1413: %A
1414: %main reason is as followings: For the symbiotic novae with CO WDs,
1415: %$^{14}$N is higher than that with ONe WDs and their $^{16}$O and
1416: %$^{13}$C are much lower than those with ONe WDs.
1417:
1418: From the left panel of Fig. \ref{fig:co}, we find that the relative
1419: CNO abundances of the cool companions in our simulations are close
1420: to those of the observational symbiotic nebulae. However, comparing
1421: more carefully, one can find that many observed SSs lie in the
1422: transition from normal M giants with initial mass lower than
1423: 4$M_\odot$ (no hot bottom burning) to the super giants with higher
1424: initial mass. This agrees with \cite{n88}. If we compare Fig.
1425: \ref{fig:co} with the observations, we find that the abundance
1426: ratios of most of observational symbiotic nebulae are between those
1427: of the cool giants and the ejected materials from the hot companions
1428: during symbiotic novae. This suggests that the compositions of the
1429: symbiotic nebulae may be modified by the ejected materials from the
1430: hot companions.
1431: %In our simulations, the ranges of the average C/N
1432: %of the cool companions and the symbiotic novae are from 1.51 (case
1433: %3) to 2.97 (case 5) and from 0.29 (case 2) to 0.31 (case 8),
1434: %respectively. The average value of the observed C/N in \cite{n88} is
1435: %about 0.76. For O/N, their ranges of the cool companions and the
1436: %symbiotic novae in our simulations are from 3.01 (case 8) to 3.45
1437: %(case 3) and from 0.92 (case 9) to 0.99 (case 8), respectively. The
1438: %average value of the observed O/N in \cite{n88} is about 2.58.
1439:
1440: As the panels (10) of Fig. \ref{fig:co} show, the abundance ratios
1441: of the symbiotic novae Pul Vul and HM Sge are like those of the
1442: novae while the ratios of V1016 Cyg and HBV 475 are similar to those
1443: of the giants. Pul Vul and HM Sge are still staying on the early
1444: phase of the nova outbursts whereas V1016 Cyg and HBV 475 may be
1445: staying on the decline phase. \cite{vn92} predicted that the
1446: chemical abundances of Pul Vul would become closer to those of red
1447: giants in the future. Therefore, we believe that the compositions of
1448: the symbiotic nebulae have been modified by the ejected materials
1449: from the hot companions during symbiotic novae. Can the modification
1450: only last $t_{\rm on}$ like our assumption in \S \ref{sec:syne}?
1451: According to Table \ref{tab:result}, most of SSs are in the cooling
1452: phase of the symbiotic novae and the stable hydrogen burning phase.
1453: The symbiotic novae are only 18.3\% (case 7) and 10\% (case 9) of
1454: total SSs. However, most of the observed symbiotic nebulae have
1455: mixing CNO abundances, which means that the above modification
1456: should be common in all SSs. We showed the C/N and O/N abundance
1457: ratios of several typical SSs which stay in different phases in
1458: order to support the above the view of point.
1459:
1460: %Obviously, the modification should last the decline stage of the
1461: %symbiotic nova. Symbiotic star AG Pegasi is a good example for the
1462: %above point of view.
1463:
1464: {\bf Declining phase:} AG Pegasi is currently in decline from a
1465: symbiotic nova outburst which began in about 1850. Based on
1466: multi-shell radio emission from AG Pegasi in \cite{k91}, the nebula
1467: can be divided into three parts. The outer nebula is thought to be
1468: the remnant of the 1850 outburst. The intermediate nebula may
1469: represent the pre-eruption mass loss from the cool component. Using
1470: a new colliding winds models including the effects of orbital
1471: motion, \cite{kt07} suggested that the inner nebula originates from
1472: the colliding winds and has an expansion velocity of about 50 km
1473: s$^{-1}$. It takes about 1000 yr for the inner nebula to reach the
1474: region of the outer nebula. In general, the timescale of the cooling
1475: phases $t_{\rm cool}$ is several hundred years. Therefore, during
1476: the cooling phase after the thermonuclear runaways, the ejected
1477: materials from the hot components have an effect on the chemical
1478: compositions of the symbiotic nebulae.
1479:
1480: {\bf Multiple outbursts:} \cite{so06} discussed the outburst of
1481: symbiotic system Z And occurring in 2000-2002. They believed that
1482: the outburst in 2000-2002 results from the enhanced shell burning
1483: triggered by a disk instability. During the burst occurring in
1484: 2000-2002, a shell of material blown from the surface of the WD is
1485: found. The blown material may result from an optically thick wind.
1486: Both Z And and CH Cyg are SSs which have undergone multiple
1487: outbursts. Since 1964, CH Cyg has displayed a number of outbursts.
1488: \cite{s06} showed both the photospheric features of the cool giant
1489: and the nebular emission lines in CH Cyg. The average C/N and O/N
1490: ratios of the cool giants in CH Cyg are respectively 1.6 and 4.0,
1491: which are typical abundance ratios of M giants. The average ratios
1492: of the symbiotic nebulae in CH Cyg are respectively 0.57 and 2.0,
1493: which are similar to those of super giant stars and are between
1494: those of giants and novae. The significant discrepancy of C/N and
1495: O/N ratios derived from these two components suggest that the
1496: chemical abundances of the nebulae in CH Cyg are modified by the
1497: ejected materials from the hot components.
1498:
1499: {\bf Quiescent phase:} Sy Mus lacks any outburst activity and it is
1500: in the stable hydrogen burning phase \citep{m07}. \cite{n88} showed
1501: that the C/N and O/N ratios in its nebulae are 0.97 and 1.9,
1502: respectively. These ratios are similar to those of super giant stars
1503: and are between those of giants and novae. This means that the
1504: chemical abundances of the nebulae of Sy Mus in the quiescent phase
1505: are modified by the ejected materials from the hot components.
1506:
1507: Therefore, we suggest that it is quite common in SSs that the
1508: chemical abundances of the symbiotic nebulae are modified by the
1509: ejected materials from the hot WD surfaces.
1510: %As Table \ref{tab:result} shows, in most of our
1511: %simulations except for cases 2 and 8, the ratio of the number of SSs
1512: %in the decline stages and the symbiotic novae to their total number
1513: %is $\sim$ 75\%. The abundances of most of the symbiotic nebula
1514: %should lie between those of the red giants and the ejected materials
1515: %from the hot companions during symbiotic novae, which agrees with
1516: %the observations.
1517:
1518:
1519: %There is no direct observational evidence that SSs in the stable
1520: %hydrogen burning stage lead to winds from the hot components.
1521: %Theoretically, the hot companions have high luminosity and
1522: %temperature during the stable hydrogen burning stage. Some part of
1523: %the materials on the surface of the accreting WDs may be blown off
1524: %by the high luminosity \citep{it96}. However, no literature refers
1525: %to them in detail yet.
1526: %As Table \ref{tab:result} shows, the ratio of the number of SSs in
1527: %stable hydrogen burning stage to their total number is between 60\%
1528: %(case 8) and 15\% (case 9).
1529:
1530:
1531:
1532:
1533:
1534:
1535: %In present paper, we assume that the hot companions steadily burn
1536: %hydrogen and do not lose mass when its mass-accreting rate is higher
1537: %than a critical mass-accreting rate (See Eq.(19) in Paper I). In
1538: %this stage, the abundances of the stellar wind of the cool
1539: %companions should represent those of the symbiotic nebulae.
1540: %In fact, the hot companions have high luminosity and temperature
1541: %during stable hydrogen burning stage. The some part of the materials
1542: %on the surface of the accreting WDs may be blown off by the high
1543: %luminosity \citep{i96}.
1544:
1545:
1546: %In Fig \ref{fig:co}a for the relative abundances of the cool giants
1547: %in SSs, there are two obvious peaks in cases 1, 4, 5, 6 and 7. The
1548: %right-up peak origins from the binary systems whose secondary
1549: %initial masses are lower than 4$M_\odot$ and does not undergo the
1550: %hot bottom burning, The left-down peak is for the binary systems in
1551: %which the secondary initial masses are higher than 4$M_\odot$ and
1552: %undergo the hot bottom burning which turns $^{12}$C into $^{14}$N.
1553: %In case 2, the mass-loss rate deeply depends on the stellar
1554: %luminosity during TP-AGB. The hot bottom burning can afford an high
1555: %luminosity and enhance the mass-loss rate which increases the
1556: %stellar wind terminal velocity and decreases the duration TP-AGB and
1557: %the yield of $^{14}$N. In case 3, the efficiency of the third
1558: %dredge-up is lower than other cases, which affords less $^{12}$C for
1559: %burning into $^{14}$N by the hot bottom burning. Therefore, there is
1560: %not obvious left-down peak in cases 2 and 3. In all cases but case
1561: %4, there are about 30\% SSs whose C/N $>$ O/N (See Table
1562: %\ref{tab:result}), whose reason is given in \S \ref{sec:Parameters}.
1563: %In (h) panel of Fig \ref{fig:co}a, the observational relative
1564: %abundances $\log$ C/N vs. $\log$ O/N of planetary nebulae, SSs and
1565: %novae are shown. Comparing our simulating results with the
1566: %observational data, we find that the result in case 4 is best, which
1567: %means that the $M_{\rm c}^{\rm min}$ of \cite{k02} is more suitable
1568: %than that in other cases.
1569:
1570: %However, a careful comparison can find that the observations of SSs
1571: %lie between the right-up and the left-down peaks, that is, the
1572: %observations are between the low initial-mass giants (no hot bottom
1573: %burning) and the high initial-mass giants (undergo hot bottom
1574: %burning). This result agrees with that of Figs 2 and 3 in
1575: %\cite{n88}. Fig \ref{fig:co}b shows the relative abundances of the
1576: %symbiotic novae. Due to a rough assumptions that the chemical
1577: %abundances of novae only depend on the mass of the accreting WD (
1578: %See Eqs. (\ref{eq:kp97}) and (\ref{eq:jh98}) ), our results have a
1579: %less scatter than the observational data in \citet{lt94}. The
1580: %distribution of $\log $C/N vs. $\log $ O/N are divided into two
1581: %regions: the thick region represents the relative abundances of SSs
1582: %with accreting CO WDs, the thin region for SSs with accreting ONe
1583: %WDs.
1584:
1585: %Comparing Fig. \ref{fig:co}a and b, we suggest that the chemical
1586: %abundances of the symbiotic nebulae should be the mix of the cool
1587: %giants and symbiotic novae.
1588:
1589:
1590:
1591: \subsubsection{\rm Ne/O vs. N/O}
1592: \begin{figure*}
1593: \begin{tabular}{c@{\hspace{3pc}}c}
1594: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f4a.ps}&
1595: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f4b.ps}\\
1596: %Fig. \ref{fig:neo}a &Fig. \ref{fig:neo}b \\
1597: \end{tabular}
1598: \caption{Gray-scale maps of the chemical abundance ratios of
1599: $\log$ Ne/O versus $\log$ N/O.
1600: Solid lines mean $\log$ Ne/O=$\log$ N/O. The left and right panels of Fig.
1601: \ref{fig:neo} show the chemical abundances of the giants during the symbiotic stage
1602: and the ejected materials from the hot companions during the symbiotic novae, respectively.
1603: In the panels of the observations, the filled big-stars, the
1604: filled triangles and the squares represent the observational values of the symbiotic nebulae in \citet{lc05},
1605: those of 5 symbiotic nebulae in \cite{dc92} and those of V1016 Cyg and HM Sge in
1606: \cite{ss90}, respectively. The circles in the left and right panels
1607: (10) of Fig. \ref{fig:neo} represent the observational values
1608: of the planetary nebulae in \citet{pb06} and the novae in \citet{lt94}, respectively.
1609: }
1610: \label{fig:neo}
1611: \end{figure*}
1612:
1613: According to optical data, \cite{lc05} calculated He abundance and
1614: the relative abundances N/O, Ne/O and Ar/O of 43 symbiotic nebulae.
1615: We first discuss Ne/O vs. N/O . The mean errors in the logarithmic
1616: (basis of 10) abundance ratios are $\sim \pm$ 0.1. \cite{dc92}
1617: showed the ratios of N/O and Ne/O in 5 symbiotic nebulae within an
1618: error of 0.15. Using the observational data, we note that the
1619: observational results for same SSs in different literature are quite
1620: different. The ratios of N/O and Ne/O of V1016 Cyg in \cite{dc92}
1621: differ by a factor of 4 from those in \cite{ss90}.
1622:
1623: Fig. \ref{fig:neo} gives the distributions of the abundance ratios
1624: of Ne/O vs. N/O in all cases and observations. The left and right
1625: panels of Fig. \ref{fig:neo} show Ne/O vs. N/O of the cool giants in
1626: SSs and the ejected materials from the hot components during
1627: symbiotic novae, respectively. The right panel of Fig. \ref{fig:neo}
1628: shows that our results of the ejected materials from the hot
1629: companions during the symbiotic novae are less scattered than the
1630: observations of the novae in \cite{lt94}. There are mainly two
1631: reasons. One reason is too small numerical samples in \cite{kp97}
1632: and \cite{jh98}. Another reason is a high uncertainty in
1633: observations with a factor of 2-4. In the right panel of Fig.
1634: \ref{fig:neo}, the distributions of Ne/O and N/O are obviously cut
1635: into two regions. The left-up region of higher Ne/O originates from
1636: the ejected materials from the hot companions during the symbiotic
1637: novae with accreting ONe WDs and the middle-down region of lower
1638: Ne/O corresponds to the ejected materials from the hot companions
1639: during the symbiotic novae with accreting CO WDs.
1640:
1641: Panels (10) of Fig. \ref{fig:neo} show that the abundance ratios of
1642: Ne/O vs. N/O in symbiotic nebulae are basically between those in the
1643: planetary nebulae and the novae. Our simulations also support this
1644: result. Therefore, the distributions of Ne/O vs. N/O reconfirm that
1645: the symbiotic nebulae are modified by the ejected materials from the
1646: hot components.
1647: %For the cool companions and
1648: %the symbiotic novae in this present, the ranges of the average N/O
1649: %are from 0.30 (case 3) to 0.43 (case 8) and from 1.19 (case 4) to
1650: %1.36 (case 7), respectively. The average Ne/O of cool companions in
1651: %all cases are about 0.24 and the range of average Ne/O of the
1652: %symbiotic novae is from 0.07 (case 3) to 0.09 (case 7). The average
1653: %observational values of N/O and Ne/O in \cite{lc05} are about 0.89
1654: %and 0.16, respectively. They lie in those of the cool companions and
1655: %the symbiotic novae.
1656:
1657: \subsubsection{\rm He/H vs. N/O}
1658: \begin{figure*}
1659: \begin{tabular}{c@{\hspace{3pc}}c}
1660: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f5a.ps}&
1661: \includegraphics[totalheight=3.2in,width=4in,angle=-90]{f5b.ps}\\
1662: %Fig. \ref{fig:he}a &Fig. \ref{fig:he}b \\
1663: \end{tabular}
1664: \caption{Gray-scale maps of the chemical abundance ratios of He/H
1665: versus $\log$ N/O.
1666: The left and right panels of Fig. \ref{fig:he} show the chemical
1667: abundances of the giants during symbiotic stage
1668: and the ejected materials from the hot companions during the symbiotic novae, respectively.
1669: In the panels of the observations, the filled big-stars, the squares and the
1670: triangles represent the observational values of the symbiotic nebulae in \citet{lc05},
1671: those of 5 symbiotic nebulae in \cite{cd94} and those of V1016 Cyg and HM Sge in
1672: \cite{ss90}, respectively. The circles in the left and right panels
1673: (10) of Fig. \ref{fig:he} represent the observational values
1674: of the planetary nebulae in \citet{pb06} and the novae in \citet{lt94}, respectively.
1675: }
1676: \label{fig:he}
1677: \end{figure*}
1678:
1679: \cite{lc05} showed the distributions of the observational $\log$ O/N
1680: vs. He/H for 43 symbiotic nebulae. The mean error of $\log$ O/N is
1681: about 0.1 and the mean error of He/H is within $\sim$ 0.03.
1682: \cite{cd94} gave those of 5 symbiotic nebulae within an error of
1683: 0.038.
1684:
1685: As the panels (10) of Fig. \ref{fig:he} show, there are many
1686: symbiotic nebulae whose helium abundances are much higher than those
1687: of the observational planetary nebulae in \cite{pb06}. The
1688: enhancement of the helium abundance in the stellar surface mainly
1689: depends on the third dredge-up. However, the left panels (1)--(9) of
1690: Fig. \ref{fig:he} show that the He/H ratios in the stellar surface
1691: of the cool giants in all SSs are basically lower than 0.2 in our
1692: simulations. The enhancement of the helium abundance in the
1693: symbiotic nebulae should originate from the ejected materials during
1694: the symbiotic novae. The observational He/H ratios of the classical
1695: novae in \cite{lt94} range from 0.07 to 0.40. The observational
1696: distribution of He/H vs. $\log$ N/O also support that the symbiotic
1697: nebulae are modified by the ejected materials from the hot
1698: components. Unfortunately, as the right panels (1)--(9) of Fig.
1699: \ref{fig:he} show, our simulations do not offer high He/H ratios for
1700: the ejected materials from the hot components. A possible
1701: explanation is: if the mass-accretion rates of WD accretors are
1702: higher than the critical value $\dot{M}_{\rm WD}$ (See Eq. (24) in
1703: Paper I), the nova outbursts are weak and a helium layer may be left
1704: on the surface of WD accretors. We call these novae as weak
1705: symbiotic novae (Paper I). If the mass-accretion rates of WD
1706: accretors are lower than $\dot{M}_{\rm WD}$, the nova outbursts are
1707: so strong that the WDs are eroded. Therefore, no helium layer is
1708: left. These novae are called as strong symbiotic novae. A typical
1709: symbiotic star can undergo dozens of nova outbursts (Paper I). For
1710: the weak symbiotic novae, a helium layer may be left on the surface
1711: of WD accretors after each outburst. When the next outburst occurs,
1712: the remnant helium can be dredged up and ejected \citep{i92}. In
1713: addition, most WDs have been traditionally found to belong to one of
1714: the following categories: those with a hydrogen-rich atmosphere (the
1715: DAs) and those with a helium-rich atmosphere (the DBs). DA white
1716: dwarfs constitute about the 80\% of all observed WDs. They have a
1717: thin hydrogen layers whose mass is lower than about $10^{-4}M_{\rm
1718: WD}$. Also, there is a helium layer (their mass is about
1719: $10^{-2}M_{\rm WD}$) under the hydrogen layer \citep{ab98}.
1720: \cite{it85} calculated the distributions of helium, carbon and
1721: oxygen through the surfaces layers of a 1.05 $M_\odot$ DB WD coming
1722: from a case B mass-transfer event. They found the amount of helium
1723: near the surface is only about $10^{-3}M_\odot$. \cite{pm07} tested
1724: the evolution of the DB white dwarf GD 358. They suggested that
1725: binary evolution describes better GD 358 and obtained the helium
1726: layer with a mass $10^{-5.66}M_{\rm WD}$. Therefore, when the first
1727: nova outbursts occurs, the hydrogen shell flash will penetrate
1728: inward to the region where the abundance of H is $\sim$ 0.01 and a
1729: certain amount of the helium is dredged up \citep{i92}. In the above
1730: situations, helium is overabundant. Our calculations may
1731: underestimate helium abundances of the ejected materials from the
1732: hot components during the symbiotic novae.
1733:
1734: The helium of the symbiotic novae V1016 Cyg and HM Sge in
1735: \cite{ss90} is not overabundant. They may have undergone several
1736: strong symbiotic novae so that the WD accretors have no the helium
1737: layer. Their helium should be similar to our simulated result.
1738:
1739: We suggest that the helium enhancement of the symbiotic nebulae
1740: stems from the modification of the ejected materials from the WDs
1741: with a helium layer. In our simulations, the ratios of the SSs being
1742: in the weak symbiotic novae and their decline phases to total SSs
1743: are between 25\% (case 2) and 38\% (case 9). The helium element of
1744: these symbiotic nebulae should be overabundant. The ratio of the
1745: symbiotic nebulae whose He/H is higher than 0.2 to the total of the
1746: symbiotic nebulae in \cite{lc05} is 33.3\%, which are in agreement
1747: with our results.
1748:
1749:
1750:
1751:
1752: %\subsection{Chemical Yields of Symbiotic Novae}
1753: %\begin{table*}
1754: % \begin{minipage}{170mm}
1755: % \caption{The chemical yields of the symbiotic novae. The first
1756: % column gives the model number according to Table 1. From
1757: % columns 2 to 11, the average chemical yields ($10^{-4}M_\odot$yr$^{-1}$) from $^{1}$H to
1758: % $^{22}$Ne are showed. The last column gives the average ejected
1759: % materials' mass ($10^{-4}M_\odot$yr$^{-1}$) from the symbiotic novae.
1760: % }
1761: % \begin{tabular}{ccccccccccccc}
1762: % \hline
1763: %Cases&$^{1}$H&$^{4}$He&$^{12}$C&$^{13}$C&$^{14}$N&$^{15}$N&$^{16}$O&$^{17}$O&$^{20}$Ne&$^{22}$Ne&$M_{\rm ejc}$\\
1764: %case 1& 3.041 &0.3319&0.0054 &0.0024& 0.0315&0.0002&0.0276&0.0034&0.0018&0.0012&5.47\\
1765: %case 2& 2.621 &0.2900&0.0047 &0.0021& 0.0266&0.0002&0.0228&0.0029&0.0015&0.0011&4.71\\
1766: %case 3& 2.727 &0.3200&0.0053 &0.0024& 0.0303&0.0002&0.0263&0.0033&0.0015&0.0012&5.26\\
1767: %case 4& 3.041 &0.3323&0.0054 &0.0025& 0.0315&0.0002&0.0277&0.0034&0.0017&0.0012&5.47\\
1768: %case 5& 3.001 &0.3278&0.0054 &0.0024& 0.0311&0.0002&0.0277&0.0033&0.0019&0.0012&5.41\\
1769: %case 6& 3.041 &0.3319&0.0054 &0.0024& 0.0315&0.0002&0.0276&0.0034&0.0018&0.0012&5.47\\
1770: %case 7& 4.106 &0.4517&0.0111 &0.0068& 0.0456&0.0034&0.0402&0.0072&0.0050&0.0037&7.77\\
1771: %case 8& 0.7002&0.0754&0.0013 &0.0006& 0.0075&0.0001&0.0067&0.0008&0.0006&0.0003&1.27\\
1772: %case 9& 3.039 &0.3314&0.0053 &0.0024& 0.0315&0.0002&0.0276&0.0033&0.0017&0.0012&5.46\\
1773: %\hline
1774: % \label{tab:yield}
1775: %\end{tabular}
1776: %\end{minipage}
1777: %\end{table*}
1778:
1779: %The nova outbursts have been identified as a promising channel for
1780: %the synthesis of $^{13}$C, $^{14}$N and $^{17}$O, especially for
1781: %most of the Galactic abundances $^{13}$C and $^{17}$O (e. g.
1782: %\cite{jh98}). The estimated Galactic occurrence rate of classical
1783: %novae is 41$\pm20$ yr$^{-1}$\citep{h97}. In our simulations, the
1784: %occurrence rate of the symbiotic novae is between 1.2 and 10.2
1785: %yr$^{-1}$, which is compared to the classical novae occurrence rate.
1786: %On studying the effects of nova nucleosynthesis on the Galactic
1787: %chemical evolution, the contributions of the symbiotic novae should
1788: %be considered.
1789:
1790: %Due to using a constant stellar formation rate in our work, Table
1791: %\ref{tab:yield} only shows the average chemical yields every year in
1792: %the Galactic disk. The common envelope algorithm and $v(\infty)$
1793: %have great effects on the chemical yields. The $\gamma$-algorithm
1794: %common envelope affords wider binary separations for the post common
1795: %envelope binary than $\alpha$-algorithm. Then, the more post common
1796: %envelope have enough separations to produce symbiotic phenomenon. In
1797: %case 7, there are higher occurrence rate of the symbiotic novae, the
1798: %ejected mass and the chemical yields. Because the $v(\infty)$ in
1799: %case 8 is unfavorable for the formation of SSs, the occurrence rate
1800: %of the symbiotic novae, the ejected mass and the chemical yields
1801: %decrease. Case 9 is interesting. Although $\Delta M_{\rm crit}^{\rm
1802: %WD}$ in case 9 greatly enhances the occurrence rate of the symbiotic
1803: %novae, it hardly changes the ejected mass. Therefore, it hardly
1804: %changes the chemical yields of the symbiotic novae.
1805: \section{Conclusion}
1806: We have performed a detailed study of the chemical abundances in SSs
1807: and drew several important conclusions.
1808:
1809: 1. The initial-final mass relation has great effects on SSs'
1810: population and especially on the symbiotic novae with massive WD
1811: accretors. A steep initial-final mass relation result in an
1812: overestimated occurrence rate of symbiotic novae.
1813:
1814: 2. The number ratios of O-rich SSs to C-rich SSs in our simulations
1815: are between 3.4 and 24.1, and they are sensitive to TDU efficiency
1816: $\lambda$ and the terminal velocity of stellar wind $v(\infty)$. Our
1817: simulation may have underestimated the terminal velocity of stellar
1818: wind $v(\infty)$ of C-rich giants and the mass-loss rate of the cool
1819: giants in SSs. The number ratio of SSs with $extrinsic$ C-rich cool
1820: giants to all SSs with C-rich cool giants is between 2.1\% and
1821: 22.7\%. The TDU efficiency $\lambda$, the common envelope algorithm
1822: and the mass-loss rate have a great effect on it.
1823:
1824: 3. Comparing our $^{12}$C/$^{13}$C vs. [C/H] of the cool giants in
1825: SSs with those of the observations, we infer that thermohaline
1826: mixing in low-mass stars should exist. Its effect on the chemical
1827: abundances is very significant.
1828:
1829: 4. The distributions of O/N vs. C/N, Ne/O vs. N/O and He/H vs. N/O
1830: of the symbiotic nebulae indicate that it is quite common in SSs
1831: that the nebular chemical abundances are modified by the ejected
1832: materials from the hot components.
1833:
1834: 5. Helium abundances in the symbiotic nebulae during the symbiotic
1835: novae are determined by whether the WD accretors have a helium layer
1836: or not in their surfaces. If they have a helium layer, helium is
1837: overabundant. If the WD accretors have only undergone strong
1838: symbiotic nova outbursts, they have no helium layer. The helium in
1839: the symbiotic nebulae ( like V1016 Cyg and HM Sge ) are not
1840: overabundant.
1841:
1842:
1843: \section*{Acknowledgments}
1844: We thank an anonymous referee for his/her comments which helped to
1845: improve the paper. We are grateful to Dr. Izzard for providing his
1846: Doctor's thesis and J. Miko{\l}ajewska for providing us a
1847: compilation of orbital periods of SSs. LGL thanks Ms. Miranda
1848: Beckham for correcting English language of the manuscript. This work
1849: was supported by the Chinese National Science Foundation under
1850: Grants Nos. 10647003, 10521001 and 10763001, the Doctor Foundation
1851: of Xinjiang University (BS060109) and the Foundation of Xinjiang
1852: University (070195).
1853: \bibliographystyle{mn2e}
1854: \bibliography{zhongmuads}
1855:
1856: \begin{thebibliography}{99}
1857: \bibitem[\protect\citeauthoryear{Althaus \& Benvenuto}{1998}]{ab98}Althaus L. G., Benvenuto O. G., 1998, MNRAS,
1858: 296, 206
1859: \bibitem[\protect\citeauthoryear{Anders \& Grevesse}{1989}]{AG89}Anders E., Grevesse N., 1989,
1860: Grochim. Cosmochim. Acta, 53, 197
1861: \bibitem[\protect\citeauthoryear{Bedijn}{1988}]{b88}Bedijn P. J., 1988, A\&A, 205, 105
1862: \bibitem[\protect\citeauthoryear{Belczy\'{n}ski et
1863: al.}{2000}]{b00}Belczy\'{n}ski K., Miko\l ajewska J., Munari U.,
1864: Ivison R. J., Friedjung M., 2000, A\&AS, 146, 407
1865: \bibitem[\protect\citeauthoryear{Bergeat \& Chevallier}{2005}]{b05}Bergeat J., Chevallier
1866: L., 2005, A\&A, 429, 235
1867: \bibitem[\protect\citeauthoryear{Bl\"{o}cker}{1995}]{b95}Bl\"{o}cker T., 1995, A\&A, 297, 727
1868: \bibitem[\protect\citeauthoryear{Bondi \& Hoyle}{1944}]{bh44}Bondi H., Hoyle F., 1944, MNRAS, 104, 273
1869: \bibitem[\protect\citeauthoryear{Boothroyd \& Sackmann}{1988}]{bs88}Boothroyd A. I., Sackmann
1870: I. J., 1988, ApJ, 328, 653
1871: \bibitem[\protect\citeauthoryear{Caughlan \& Fowler}{1988}]{cf88}Caughlan G. R., Fowler W. A.,
1872: 1988, {\it Atomic Data and Nuclear Data Tables} 40, 283-334.
1873: \bibitem[\protect\citeauthoryear{Charbonnel \&
1874: Zahn}{2007}]{cz07}Charbonnel C., Zahn J. P., 2007, A\&A, 467, L15
1875: %\bibitem[\protect\citeauthoryear{Cioni \&
1876: %Habing}{2003}]{ch03}Cioni M.-R. L., Habing H. J., 2003, A\&A, 402,
1877: %133
1878: \bibitem[\protect\citeauthoryear{Clayton}{1983}]{c83}Clayton D. D., 1983, Principles of
1879: Stellar EVolution and Nucleosynthesis. Chicago
1880: \bibitem[\protect\citeauthoryear{Costa \&
1881: de Freitas Pacheco}{1994}]{cd94}Costa R. D. D., de Freitas Pacheco
1882: J. A., 1994, A\&A, 285, 998
1883: \bibitem[\protect\citeauthoryear{De Freitas Pacheco \&
1884: Costa}{1992}]{dc92}de Freitas Pacheco J. A., Costa R. D. D., 1992,
1885: A\&A, 257, 619
1886: \bibitem[\protect\citeauthoryear{Eggleton et al.}{1989}]{e89}Eggleton P. P.,
1887: Fitechett M. J., Tout C. A., 1989, ApJ, 347, 998
1888: \bibitem[\protect\citeauthoryear{Eggleton et al.}{2006}]{e06}Eggleton P.
1889: P., Dearborn D. S. P., Lattanzio J. C., 2006, Science, 314, 1580
1890: \bibitem[\protect\citeauthoryear{Eggleton et al.}{2007}]{e07}Eggleton P.
1891: P., Dearborn D. S. P., Lattanzio J. C., 2007, astro-ph/07062710
1892: %\bibitem[\protect\citeauthoryear{Gil-Pons et al.}{2003}]{g03}
1893: %Gil-Pons P., Garc\'{\i}a-Berro E., Jos\'{e} J., Hernanz M., Truran
1894: %T. W., 2003, A\&A, 407, 1021
1895: \bibitem[\protect\citeauthoryear{Girardi et al.}{2000}]{g00}Girardi
1896: L., Bressan A., Bertelli G., Chiosi C., 2000, A\&AS, 141, 371
1897: \bibitem[\protect\citeauthoryear{Goldberg \& Mazeh}{1994}]{gm94}Goldberg D.,
1898: Mazeh T., 1994, A$\&$A, 282, 801
1899: \bibitem[\protect\citeauthoryear{Gonz$\acute{\rm a }$lez P$\acute{\rm e}$rez \& Metcalfe}{2007}]
1900: {pm07}Gonz$\acute{\rm a }$lez P$\acute{\rm e}$rez J. M., Metcalfe
1901: T., astro-ph/07083703
1902: \bibitem[\protect\citeauthoryear{Gratton et al.}{2000}]{gs00}Gratton
1903: R. G., Sneden C., Carretta E., Bragaglia A., 2000, A\&A, 354, 169
1904: \bibitem[\protect\citeauthoryear{Groenewegen \& de Jong}{1993}]{gj93}Groenewegen M. A. T., de
1905: Jong T., 1993, A\&A, 267, 410
1906: \bibitem[\protect\citeauthoryear{Groenewegen}{2002}]{g02}Groenewegen M. A. T., 2002,
1907: Ringberg Castle Workshop (Astro-ph/0208449)
1908: \bibitem[\protect\citeauthoryear{Han et al.}{1994}]{h94}Han Z.,
1909: Podsiadlowski P., Eggleton P. P., 1994, MNRAS, 270, 121
1910: \bibitem[\protect\citeauthoryear{Han et al.}{1995a}]{h95a}Han Z.,
1911: Eggleton P. P., Podsiadlowski P., Tout C. A., 1995a, MNRAS, 277,
1912: 1443
1913: \bibitem[\protect\citeauthoryear{Han et al.}{1995b}]{h95b}Han Z.,
1914: Podsiadlowski P., Eggleton P. P., 1995b, MNRAS, 272, 800
1915: \bibitem[\protect\citeauthoryear{Han et al.}{2001}]{han01}Han Z.,
1916: Eggleton P. P., Podsiadlowski Ph., Tout C. A., Webbink R. F.,
1917: 2001, ASP Conf. Ser., 229, 205
1918: \bibitem[\protect\citeauthoryear{Han et al.}{2002}]{han02}Han Z.,
1919: Podsiadlowski, Ph., Maxted P. F. L., Marsh T. R., Ivanova N.,
1920: 2002, MNRAS, 336, 449
1921: \bibitem[\protect\citeauthoryear{Hatano et
1922: al.}{1997}]{h97}Hatano K., Branch D., Fisher A., Starrfield S.,
1923: 1997, MNRAS, 290, 113
1924: \bibitem[\protect\citeauthoryear{Herwig}{1995}]{h95}Herwig F., 1995,
1925: In: Noels A., Fraipont-Caro D., Gabriel M., Grevesse N., Demarque
1926: P.(eds.) Proc. 32. Liege Int. Astrophys. Coll., P. 441
1927: \bibitem[\protect\citeauthoryear{Hurley et al.}{2000}]{h00}Hurley J. R.,
1928: Pols O. R., Tout C. A., 2000, MNRAS, 315, 543
1929: \bibitem[\protect\citeauthoryear{Hurley et al.}{2002}]{h02}Hurley J. R., Tout C. A., Pols R.,
1930: 2002, MNRAS, 329, 897
1931: \bibitem[\protect\citeauthoryear{Iben \& Renzini}{1983}]{ir83} Iben I. Jr., Renzini A., 1983,
1932: ARA\&A, 21,271
1933: \bibitem[\protect\citeauthoryear{Iben \& Tutukov}{1984}]{it84}Iben I. Jr.,
1934: Tutukov A. V., 1984, ApJ, 282, 615
1935: \bibitem[\protect\citeauthoryear{Iben \& Tutukov}{1985}]{it85}Iben I. Jr.,
1936: Tutukov A. V., 1985, ApJS, 58, 661
1937: \bibitem[\protect\citeauthoryear{Iben et al.}{1992}]{i92}Iben I.
1938: Jr., Fujimoto M. Y., Macdonald J., 1992, ApJ, 388, 521
1939: \bibitem[\protect\citeauthoryear{Iben \& Tutukov}{1996}]{it96}Iben I. Jr.,
1940: Tutukov A. V., 1996, ApJS, 105, 145
1941: \bibitem[\protect\citeauthoryear{Izzard et al.}{2004}]{i04}Izzard R. G., Tout C. A., Karakas
1942: A. I., Pols O. R., 2004, MNRAS, 350, 407
1943: \bibitem[\protect\citeauthoryear{Izzard}{2004}]{I04}Izzard R. G., 2004, Doctor thesis,
1944: Combrige
1945: \bibitem[\protect\citeauthoryear{Jorissen}{2003}]{j03}Jorissen A., 2003, ASPC, 303, 25
1946: \bibitem[\protect\citeauthoryear{Jos$\acute{e}$ \& Hernanz}{1998}]{jh98}
1947: Jos$\acute{e}$ J., Hernanz M., 1998, ApJ, 494, 680
1948: \bibitem[\protect\citeauthoryear{Karakas, Lattanzio \& Pols}{2002}]{k02}Karakas A. I., Lattanzio
1949: J. C., Pols O. R., 2002, PASA, 19, 515
1950: \bibitem[\protect\citeauthoryear{Kenny, Taylor \& Seaquist
1951: }{1991}]{k91}Kenny H. T., Taylor A. R., Seaquist E. R., 1991, ApJ,
1952: 366, 549
1953: \bibitem[\protect\citeauthoryear{Kenny \&
1954: Taylor}{2005}]{kt05}Kenny H. T., Taylor A. R., 2005, ApJ, 619, 527
1955: \bibitem[\protect\citeauthoryear{Kenny \&
1956: Taylor}{2007}]{kt07}Kenny H. T., Taylor A. R., 2007, ApJ, 662, 1231
1957: \bibitem[\protect\citeauthoryear{Kenyon \& Webbink}{1984}]{kw84}Kenyon S. J.,
1958: Webbink R. F., 1984, ApJ, 279, 252
1959: \bibitem[\protect\citeauthoryear{Kenyon}{1986}]{k86}Kenyon S. J.,
1960: 1986, The Symbiotic Stars(Cambridge: Cambridge Univ. Press)
1961: \bibitem[\protect\citeauthoryear{Kj{\ae}rgaard et
1962: al.}{1982}]{k82}Kj{\ae}rgaard P., Gustafsson B., Walker G. A. H.,
1963: Hultqvist L., 1982, A\&A, 115, 145
1964: \bibitem[\protect\citeauthoryear{Kovetz \&
1965: Prialnik}{1994}]{kp94}Kovetz A., Prialnik D., 1994, ApJ, 424, 319
1966: \bibitem[\protect\citeauthoryear{Kovetz \&
1967: Prialnik}{1997}]{kp97}Kovetz A., Prialnik D., 1997, ApJ, 477, 356
1968: \bibitem[\protect\citeauthoryear{Lazaro et al.}{1991}]{l91}Lazaro
1969: C., Lynas-Gray A. E., Clegg R. E. S., Mountain C. M., Zadrozny A.,
1970: 1991, A\&A, 249, 62
1971: \bibitem[\protect\citeauthoryear{Livio \& Truran}{1994}]{lt94}Livio
1972: M., Truran J. W., 1994, ApJ, 425, 797
1973: \bibitem[\protect\citeauthoryear{Luna \&
1974: Costa}{2005}]{lc05}Luna G. J. M., Costa R. D. D., 2005, A\&A, 435,
1975: 1087
1976: \bibitem[\protect\citeauthoryear{L\"{u}, Yungelson \&
1977: Han}{2006}]{lyh06}L\"{u} G. L., Yungelson L., Han Z., 2006, MNRAS,
1978: 327, 1389
1979: \bibitem[\protect\citeauthoryear{Marigo}{2001}]{m01}Marigo P., 2001, A\&A, 370,
1980: 194
1981: \bibitem[\protect\citeauthoryear{Marigo \& Girardi}{2007}]{mg07}Marigo P., Girardi L., 2007, A\&A, 469,
1982: 239
1983:
1984: \bibitem[\protect\citeauthoryear{Mazeh et al.}{1992}]{m92}Mazeh T.,
1985: Goldberg D., Duquennoy A., Mayor M., 1992, ApJ, 401, 265
1986: \bibitem[\protect\citeauthoryear{Miller \& Scalo}{1979}]{ms79}Miller G. E.,
1987: Scalo J. M., 1979, ApJS, 41, 513
1988: \bibitem[\protect\citeauthoryear{Miko\l ajewska \& Kenyon}{1992}]{mk92}Miko\l ajewska J.,
1989: Kenyon S. J., 1992, MNRAS, 256, 177
1990: \bibitem[\protect\citeauthoryear{Miko\l ajewska}{2003}]{m03}Miko{\l}ajewska,
1991: J., 2003, Astronomical Society of the Pacific Conference Series,
1992: 303, 9
1993: \bibitem[\protect\citeauthoryear{Miko\l ajewska}{2007}]{m07}Miko{\l}ajewska,
1994: J., 2007, Baltic Astronomy, Vol. 16, p. 1-9
1995: \bibitem[\protect\citeauthoryear{Mitsumoto et al.}{2005}]{m05}Mitsumoto
1996: M., Jahanara B., Matasuda T., et al., 2005, ARep., 49, 884
1997: \bibitem[\protect\citeauthoryear{M\"{u}rset et al.}{1991}]{m91} M\"{u}rset U.,
1998: Nussbaumer H., Schmid H. M., Vogel M., 1991, A$\&$A, 248, 458
1999: %\bibitem[\protect\citeauthoryear{M\"{u}rset et al.}{1996}]{m96} M\"{u}rset U.,
2000: %Schild H., Vogel M., 1996, A$\&$A, 307, 516
2001: \bibitem[\protect\citeauthoryear{M\"{u}rset \& Schmid}{1999}]{ms99} M\"{u}rset U.,
2002: Schmid H. M., 1999, A$\&$A, 137, 473
2003: \bibitem[\protect\citeauthoryear{Nauenberg }{1972}]{n72}Nauenberg M., 1972, ApJ, 175, 417
2004: \bibitem[\protect\citeauthoryear{Nelemans et al.}{2000}]{n20}Nelemans G.,
2005: Verbunt F., Yungelson L. R., Portegies Zwart S. F., 2000, A$\&$A,
2006: 360, 1011
2007: \bibitem[\protect\citeauthoryear{Nelemans \& Tout}{2005}]{nt05}Nelemans G.,
2008: Tout C. A., 2005, MNRAS, 356, 753
2009: \bibitem[\protect\citeauthoryear{Nelson et al.}{2004}]{n04}Nelson L. A.,
2010: MacCannell K. A., Dubeau E., 2004, ApJ, 602, 938
2011: \bibitem[\protect\citeauthoryear{Nussbaumer et
2012: al.}{1988}]{n88}Nussbaumer H., Schild H., Schmid H. M., Vogel M.,
2013: 1988, A\&A, 198, 179
2014: \bibitem[\protect\citeauthoryear{Nussbaumer \& Vogel}{1989}]{nv89}Nussbaumer H., Vogel M., 1989,
2015: A\&A, 213, 137
2016: \bibitem[\protect\citeauthoryear{Nussbaumer \& Vogel}{1990}]{nv90}Nussbaumer H., Vogel M., 1990,
2017: A\&A, 236, 117
2018: \bibitem[\protect\citeauthoryear{Olofsson et
2019: al.}{2002}]{o02}Olofsson H., Gonz\'{a}lez Delgado D., Kerschbaum F.,
2020: Sch\"{o}ier F. L., 2002, A$\&$A, 391, 1053
2021: \bibitem[\protect\citeauthoryear{Paczy{\rm\'{n}}ski \& Rudak}{1980}]{pr80}Paczy$\acute{n}$ski B.,
2022: Rudak B., 1980, A$\&$A, 82, 349
2023: \bibitem[\protect\citeauthoryear{Pols et al.}{1998}]{p98}Pols O. R.,
2024: Schr\"{o}der K. P., Hurley J. R., Tout C. A., Eggleton P. P., 1998,
2025: MNRAS, 298, 525
2026: \bibitem[\protect\citeauthoryear{Pottasch \&
2027: Bernard-Salas}{2006}]{pb06}Pottasch S. R., Bernard-Salas J., 2006,
2028: A\&A, 457, 189
2029: \bibitem[\protect\citeauthoryear{Press et al.
2030: }{1992}]{p92}Press W. H., Teukolsky S. A., Vetterling W. T.,
2031: Flannery B. P., 1992, Numerical Recipes In Fortran 77. The Art of
2032: Scientific Computing Second Edition. Cambridge Univ. Press,
2033: Combridge
2034: \bibitem[\protect\citeauthoryear{Prialnik \& Kovetz
2035: }{1995}]{pk95}Prialnik D., Kovetz A., 1995, ApJ, 445, 789
2036: %\bibitem[\protect\citeauthoryear{Prialnik}{1986}]{p86}Prialnik D., 1986, ApJ, 310,
2037: %222
2038: \bibitem[\protect\citeauthoryear{Reimers}{1975}]{r75}Reimers D., 1975, Mem. Soc. R. Sci. Liege,
2039: 8, 369
2040: %\bibitem[\protect\citeauthoryear{Ritter et al.}{1991}]{r91}Ritter H., Politano M., Livio M., Webbink
2041: %R. F., 1991, ApJ, 376, 177
2042: \bibitem[\protect\citeauthoryear{ Schmid \& Schild
2043: }{1990}]{ss90} Schmid H. M., Schild H., 1990, MNRAS, 246, 84
2044: \bibitem[\protect\citeauthoryear{Schild, Boyle \& Schmid
2045: }{1992}]{s92}Schild H., Boyle S. J., Schmid H. M., 1992, MNRAS, 258,
2046: 95
2047: \bibitem[\protect\citeauthoryear{Schmidt \& Miko{\l}ajewska}{2003}]{sm03}Schmidt
2048: M. R., Miko{\l}ajewska J., 2003, ASPC, 303, 163
2049: \bibitem[\protect\citeauthoryear{Schmidt et al.}{2006}]{s06}Schmidt
2050: M. R., Za$\breve{\rm c}$s L., Miko{\l}ajewska J., Hinkle K. H.,
2051: 2006, A\&A, 446, 603
2052: \bibitem[\protect\citeauthoryear{Shara, Prialnik \&
2053: Kovetz}{1993}]{spk93}Shara M. M., Prialnik D., Kovetz A., 1993, ApJ,
2054: 406, 220
2055: \bibitem[\protect\citeauthoryear{Sokoloski et
2056: al.}{2006}]{so06}Sokoloski J. L., Kenyon S. J., Espey B. R., et al.,
2057: 2006, ApJ, 636, 1002
2058: \bibitem[\protect\citeauthoryear{Stancliffe \& Jeffery}{2007}]{sj07}Stancliffe R. J., Jeffery
2059: C. S., 2007, MNRAS, 375, 1280
2060: \bibitem[\protect\citeauthoryear{Starrfield et al.}{1998}]{s98}Starrfield S., Truran J.
2061: W., Wiescher M. C., Sparks W. M., 1998, MNRAS, 296, 502
2062: \bibitem[\protect\citeauthoryear{Tutukov \& Yungelson}{1976}]{ty76}Tutukov A. V.,
2063: Yungelson L. R., 1976, Astrophysics, 12, 321
2064: \bibitem[\protect\citeauthoryear{van den Hoek \& Groenewegen}{1997}]{vg97}
2065: van den Hoek L. B., Groenewegen M. A. T., 1997, A\&AS, 123, 305
2066: \bibitem[\protect\citeauthoryear{Vassiliadis \& Wood}{1993}]{vw93}Vassiliadis E.,
2067: Wood P. R., 1993, ApJ, 413, 641
2068: \bibitem[\protect\citeauthoryear{Vogel \&
2069: Nussbaumer}{1992}]{vn92}Vogel M., Nussbaumer H., 1992, A\&A, 259,
2070: 525
2071: \bibitem[\protect\citeauthoryear{Wagenhuber \& Groenewegen}{1998}]{WG98}Wagenhuber J., Groenewegen M. A. T.,
2072: 1998, A\&A, 340, 183
2073: \bibitem[\protect\citeauthoryear{Whitelock \& Munari}{1992}]{wm92}Whitelock P. A., Munari
2074: U., 1992, A\&A, 255, 171
2075: \bibitem[\protect\citeauthoryear{Willson et al.}{1984}]{w84}Willson L. A., Wallerstein G.,
2076: Brugel E. W., Stencel R. E., 1984, A\&A, 133, 254
2077: \bibitem[\protect\citeauthoryear{Willson}{2007}]{w07}Willson L. A., 2007, astro-ph/07043589
2078: \bibitem[\protect\citeauthoryear{Winters et al.}{2000}]{win00}Winters J. M., Le Bertre T., Jeong K.
2079: S., Helling Ch., Sedlmayr E., 2000, A$\&$A, 361, 641
2080: \bibitem[\protect\citeauthoryear{Winters et al.}{2003}]{w03}Winters J. M., Le Bertre T., Jeong K.
2081: S., Nyman L.-\AA., Epchtein N., 2003, A$\&$A, 409, 715
2082: \bibitem[\protect\citeauthoryear{Weidemann}{2000}]{w00}Weidemann V., 2000, A$\&$A, 363,
2083: 647
2084: \bibitem[\protect\citeauthoryear{Yaron et al.}{2005}]{y05}Yaron O.,
2085: Prialnik D., Shara M. M., Kovetz A., 2005, ApJ, 623, 398
2086: \bibitem[\protect\citeauthoryear{Yungelson et al.}{1994}]{y94}Yungelson L.,
2087: Livio M., Tutukov A. V., Saffer R. A., 1994, ApJ, 420, 336
2088: \bibitem[\protect\citeauthoryear{Yungelson et al.}{1995}]{y95}Yungelson L.,
2089: Livio M., Tutukov A. V., Kenyon S. J., 1995, ApJ, 477, 656
2090: %\bibitem[\protect\citeauthoryear{Zhu et. al.}{2007}]{z07}Zhu C. H.,
2091: %Zhang J., Wang Z. J., Yu S. H., L\"{u} G. L., 2007,
2092: \end{thebibliography}
2093:
2094:
2095: \appendix
2096: \label{sec:appa}
2097: \section[]{Synthesis TP-AGB Evolution} Stellar evolution from the zero
2098: age main sequence up to the first thermal pulse is dealt with in the
2099: rapid evolution code of \citet{h00}. The changes of the chemical
2100: abundances on the stellar surface during the giant branch phase (the
2101: first dredge up) and early asymptotic giant branch phase (the second
2102: dredge up) can be represented by simple fitting formulae in
2103: \cite{i04} and \cite{I04}. After the first thermal pulse, we use a
2104: synthetic model for TP-AGB.
2105:
2106: \subsection{The initial abundances} The
2107: initial abundances ({\it i.e.} of zero age main sequence stars) are
2108: taken from \citet{AG89} for $Z=0.02$. The
2109: following shows the initial abundances by mass fractions:\\
2110: %\begin{array}{llll}
2111: $^1$H=0.68720; \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ $^4$He=0.29280 ;\\
2112: $^{12}$C=2.92293$\times10^{-3}$; \ \ \ \ \ \ \ \ $^{13}$C=4.10800$\times10^{-5}$;\\
2113: $^{14}$N=8.97864$\times10^{-4}$; \ \ \ \ \ \ \ \
2114: $^{15}$N=4.14000$\times10^{-6}$;\\
2115: $^{16}$O=8.15085$\times10^{-3}$; \ \ \ \ \ \ \ \ $^{17}$O=3.87600$\times10^{-6}$;\\
2116: $^{20}$Ne=2.29390$\times10^{-3}$; \ \ \ \ \ \ \ $^{22}$Ne=1.45200$\times10^{-4}$.\\
2117: %\end{array}
2118: \subsection{Core mass at the first thermal pulse} Using the rapid evolution code of
2119: \citet{h00}, we can obtain the stellar mass ($M_{\rm 1TP}$) at the
2120: first thermal pulse. The core mass at the first thermal pulse,
2121: $M_{\rm c, 1TP}$ is taken from \cite{k02}:
2122: \begin{equation}
2123: M_{\rm c, 1TP}=[-p_1(M_0-p_2)+p_3]f+(p_4M_0+p_5)(1-f),
2124: \label{eq:mc1tp}
2125: \end{equation}
2126: where $f=(1+e^{(\frac{M_0-p_6}{p_7})})^{-1}$, $M_0$ is the initial
2127: mass in solar unit and the coefficients $p_1$, $p_2$, $p_3$, $p_4$,
2128: $p_5$, $p_6$ and $p_7$ are in Table 6 of \cite{k02}.
2129: \subsection{Luminosity, radius, interpulse period, and evolution of core mass}
2130: We use the prescriptions of \cite{i04}. The luminosity is taken as
2131: the value calculated from Eq.(29) in \cite{i04}. We define the
2132: radius $R$ as $L=4\pi \sigma R^2T^4_{\rm eff}$, where $\sigma$ is
2133: the Stefan-Boltzmann constant and $T_{\rm eff}$ is the effective
2134: temperature of the star. The radius is taken the value calculated
2135: from Eq.(35) in \cite{i04}. The interpulse period $\tau_{\rm ip}$
2136: is:
2137: \begin{equation}
2138: {\rm log}_{10}(\tau_{\rm ip}/\rm yr)=a_{28}(M_{\rm
2139: c}/M_\odot-b_{28})-10^{c_{28}}-10^{d_{28}}+0.15\lambda^2,
2140: \label{eq:tip}
2141: \end{equation}
2142: where the third dredged-up (TDU) efficiency $\lambda$ is defined in
2143: \S \ref{sec:lamb}, and
2144: the coefficients are:\\
2145: $a_{28}=-3.821,$\\
2146: $b_{28}=1.8926,$\\
2147: $c_{28}=-2.080-0.353Z+0.200(M_{\rm env}/M_\odot+\alpha-1.5),$\\
2148: $d_{28}=-0.626-70.30(M_{\rm c,1TP}/M_\odot-\zeta)(\Delta M_{\rm
2149: c}/M_\odot),$\\where $M_{\rm env}$ represents the envelope mass,
2150: $\alpha$ is the mixing length parameter and equals to 1.75,
2151: $\zeta={\rm log}(Z/0.02)$, and $\Delta M_{\rm c}$ is the change in
2152: core mass defined as $\Delta M_{\rm c}=M_{\rm c}-M_{\rm c, 1TP}$.
2153:
2154: \subsection{The minimum core mass for TDU and the TDU efficiency}
2155: \label{sec:lamb} TDU can occur only for stars above a certain core
2156: mass $M^{\rm min}_{\rm c}$. \cite{gj93} took $M^{\rm min}_{\rm c}$
2157: as a constant 0.58. \cite{k02} found that $M^{\rm min}_{\rm c}$
2158: depends on stellar mass and metallicity and gave a fitting formula
2159: by
2160: \begin{equation}
2161: M^{\rm min}_{\rm c}=a_1+a_2M_0+a_3M_0^2+a_4M_0^3 \label{eq:mcmin},
2162: \end{equation}
2163: where coefficients $a_1$, $a_2$, $a_3$ and $a_4$ are shown in Table
2164: 7 of \cite{k02} and $M_0$ is the initial mass in solar unit.
2165: According to the observed carbon luminosity function in the
2166: Magellanic clouds, \cite{mg07} considered that the $M^{\rm min}_{\rm
2167: c}$ predicted by \cite{k02} is high. In this work, we carry out
2168: various numerical simulations (See Table
2169: \ref{tab:case}).\\
2170: (i)Like \cite{gj93}, $M^{\rm min}_{\rm c}$=0.58$M_\odot$;\\
2171: (ii)Following \cite{k02}, we take Eq. (\ref{eq:mcmin}) as $M^{\rm min}_{\rm c}$;\\
2172: It should be recalled that $M^{\rm min}_{\rm c}$=$M_{\rm c, 1TP}$ if
2173: stellar initial mass $M_{\rm initial}\geq 4M_\odot$ \citep{k02,i04}
2174: or $M^{\rm min}_{\rm c}<M_{\rm c, 1TP}$.
2175:
2176: The TDU efficiency is defined by $\lambda=\frac{\Delta M_{\rm
2177: dred}}{\Delta M_{\rm c}}$, where $ \Delta M_{\rm dred}$ is the mass
2178: brought up to the stellar surface during a thermal pulse. $\lambda$
2179: is a very uncertain parameter. \cite{k02} showed a relation of
2180: $\lambda$:
2181: \begin{equation}
2182: \lambda(N)=\lambda_{\rm max}[1-\exp(-N/N_{\rm r})], \label{eq:lamb}
2183: \end{equation}
2184: where $\lambda$ gradually increases towards an asymptotic
2185: $\lambda_{\rm max}$ with $N$ (the progressive number of thermal
2186: pulsation) increasing. $N_{\rm r}$ is taken from Eq. (49) in
2187: \cite{i04} which reproduces the results for $N_{\rm r}$ in Table 5
2188: of \cite{k02}. $\lambda_{\rm max}$ is given by (See Eq.(6) in
2189: \cite{k02}):
2190: \begin{equation}
2191: \lambda_{\rm max}=\frac{b_1+b_2M_0+b_3M^3_0}{1+b_4M_0^3},
2192: \label{eq:lambmax}
2193: \end{equation}
2194: where coefficients are shown in Table 8 of \cite{k02}. \cite{gj93}
2195: took $\lambda$ as a constant 0.75. In order to test the influences
2196: of $\lambda$ on our results, we carry out numerical simulations with
2197: various $\lambda$ (See Table \ref{tab:case}).\\
2198: (i)Following \cite{k02}, $\lambda$=Eq. (\ref{eq:lamb});\\
2199: (ii)Following \cite{gj93}, $\lambda$=0.75;\\
2200: (iii)Simulating a small TDU efficiency, $\lambda$=0.5.
2201:
2202: \subsection{Inter-shell abundances}
2203: During every thermal pulse, the dredged mass $\Delta M_{\rm dred}$
2204: mixes into stellar envelope. Based on the nucleosynthesis
2205: calculations by \citet{bs88}, \cite{mg07} gave the fit of the
2206: abundances of $^4$He,
2207: $^{12}$C and $^{16}$O in the inter-shell region:\\
2208: For $\Delta M_{\rm c}\leq$0.025$M_\odot$:\\
2209: $^{\rm inter}X(^{4}{\rm He})=0.95+400(\Delta M_{\rm c})^2-20.0\Delta M_{\rm c}$\\
2210: $^{\rm inter}X(^{12}{\rm C})=0.03-352(\Delta M_{\rm c})^2+17.6\Delta M_{\rm c}$\\
2211: $^{\rm inter}X(^{16}{\rm O})=-32(\Delta M_{\rm c})^2+1.6\Delta M_{\rm c}$.\\
2212: For $\Delta M_{\rm c}>$0.025$M_\odot$:\\
2213: $^{\rm inter}X(^{4}{\rm He})=0.70+0.65(\Delta M_{\rm c}-0.025)$\\
2214: $^{\rm inter}X(^{12}{\rm C})=0.25-0.65(\Delta M_{\rm c}-0.025)$\\
2215: $^{\rm inter}X(^{16}{\rm O})=0.02-0.065(\Delta M_{\rm c}-0.025)$.\\
2216:
2217: However, there is some debate of the exact composition in the
2218: inter-shell region. \cite{i04} gave some fits of inter-shell
2219: abundance, in which inter-shell elements include $^{4}$He, $^{12}$C,
2220: $^{16}$O and $^{22}$Ne. Details can be seen from \S 3.3.3 in
2221: \cite{i04}. Their models did not obtain high values of inter-shell
2222: $^{16}$O such as 2\% reported by \citet{bs88}. The typical
2223: inter-shell abundances (5 $M_\odot$, $Z=0.02$) are $^{4}$He=0.74,
2224: $^{12}$C=0.23, $^{16}$O=0.005 and $^{22}$Ne=0.02.
2225:
2226: In our work, we carry out simulations with various inter-shell
2227: abundances.
2228:
2229: \subsection{The hot bottom burning}
2230: \label{sec:hbb}
2231:
2232: If the hydrogen envelope of an AGB star is sufficiently massive, the
2233: hydrogen burning shell can extend into the bottom of the convective
2234: region. This process is called as hot bottom burning(HBB). For HBB,
2235: we use a treatment similar to that in \cite{gj93}. For the model of
2236: \citet{ir83}, \cite{gj93} gave the most suitable parameter for the
2237: fraction of newly dredged up matter exposed to the high temperatures
2238: at the bottom of envelope $f_{\rm HBB}=0.94$, the fraction of
2239: envelope matter mixed down to the bottom of the envelope $f_{\rm
2240: bur}=3\times10^{-4}$, and the exposure time of matter in the region
2241: of HBB $t_{\rm HBB}=0.0014\tau_{\rm ip}$. The temperature at the
2242: bottom of convective envelope $T_{\rm bce}$ is given by Eq. (37) of
2243: \cite{i04}.
2244: %\begin{equation}
2245: %T_{\rm B}=T^0_{\rm B}+127.6(M_{\rm c}-0.8)
2246: %\end{equation}
2247: %where the zeropoint is initial mass dependent:$T^0_{\rm
2248: %B}=-25.45+16.41M_{\rm initial}$.
2249:
2250: For the drudged up mass, the amounts of material added to the
2251: envelope are:
2252: \begin{equation}
2253: \begin{array}{ll}
2254: \Delta^4{\rm He}=&^{\rm inter}X_4\Delta M_{\rm dred}\\
2255: \Delta^{12}{\rm C}=&[(1-f_{\rm HBB})^{\rm
2256: inter}X_{12}\\
2257: &+\frac{f_{\rm HBB}}{t_{\rm HBB}}
2258: \int^{t_{\rm HBB}}_{0}X^{\rm HBB}_{12}(t){\rm d}t]\Delta M_{\rm dred}\\
2259: \Delta^{13}{\rm C}=&[\frac{f_{\rm HBB}}{t_{\rm HBB}}
2260: \int^{t_{\rm HBB}}_{0}X^{\rm HBB}_{13}(t){\rm d}t]\Delta M_{\rm dred}\\
2261: \Delta^{14}{\rm N}=&[\frac{f_{\rm HBB}}{t_{\rm HBB}}
2262: \int^{t_{\rm HBB}}_{0}X^{\rm HBB}_{14}(t){\rm d}t]\Delta M_{\rm dred}\\
2263: \Delta^{16}{\rm O}=&[(1-f_{\rm HBB})^{\rm
2264: inter}X_{16}\\
2265: &+\frac{f_{\rm HBB}}{t_{\rm HBB}} \int^{t_{\rm HBB}}_{0}X^{\rm
2266: HBB}_{16}(t){\rm
2267: d}t]\Delta M_{\rm dred} ,\\
2268: \end{array}
2269: \label{eq:hbb}
2270: \end{equation}
2271: where $X^{\rm HBB}(t)$ are calculated by the way of Clayton's CNO
2272: bicycle\citep{c83}. All details of Clayton's CNO bicycle can be seen
2273: from \citet{c83,gj93,i04}. The CNO bicycle can be splited into CN
2274: cycle and ON cycle. The timescales in ON cycle are many thousands of
2275: times longer than those required to bring the CN cycle into
2276: equilibrium. Even in the most massive AGB stars undergoing vigorous
2277: HBB, the ON cycle never approaches equilibrium. Therefore, the
2278: effects of HBB mainly turn $^{12}$C into $^{14}$N and the abundance
2279: of $^{16}$O is not changed much. In calculating Clayton's CNO
2280: bicycle, the density at the base of the convective envelope is given
2281: by Eq. (42) of \cite{i04} and the analytic expressions for the
2282: nuclear reaction rates in Clayton's CNO bicycle are taken from
2283: \citet{cf88}. The initial conditions of $X^{\rm HBB}(t)$ are $X^{\rm
2284: HBB}(t=0)=^{\rm inter}X$ for $^{12}{\rm C}$ and $^{16}{\rm O}$,
2285: while $X^{\rm HBB}(t=0)=0$ for $^{13}{\rm C}$ and $^{14}{\rm N}$.
2286:
2287: After every thermal pulse, the chemical abundances of stellar
2288: envelope $X^{\rm new}$ are
2289: \begin{equation}
2290: X^{\rm new}=\frac{X^{\rm old}M_{\rm env}(1-f_{\rm bur})+\Delta
2291: X+\frac{f_{\rm bur}M_{\rm env}}{t_{\rm HBB}}\int_0^{t_{\rm
2292: HBB}}X(t){\rm d}t} {M_{\rm env}+\Delta M_{\rm dred}},
2293: \end{equation}
2294: where $\Delta X$ is given by Eq. (\ref{eq:hbb}) and the initial
2295: conditions of $X(t)$ are $X(t=0){\rm }=X^{\rm old}{\rm }$.
2296:
2297:
2298: \subsection{Mass loss}
2299: \label{sec:ml} The mass-loss rate of the cool giant during AGB phase
2300: has a great effect on the population of SSs and the chemical
2301: evolution of the stellar surface. We consider two
2302: laws of the mass-loss rate in this work. \\
2303: (i)A mass-loss relation suggested by \citet{vw93} based on the
2304: observations, given as
2305: \begin{equation}
2306: \begin{array}{l}
2307: \log \dot{M}= -11.4+0.0123(P-100\max(M/M_\odot-2.5, 0.0)),
2308: \end{array}
2309: \label{eq:vwml}
2310: \end{equation}
2311: where $P$ is the Mira pulsation period in days given by
2312: \begin{equation}
2313: \begin{array}{l}
2314: \log P = -2.07+1.94 \log(R/R_\odot)-0.90\log(M/M_\odot).
2315: \end{array}
2316: \end{equation}
2317: When $P \geq 500$ days, the steady super-wind phase is modeled by
2318: the law
2319: \begin{equation}
2320: \dot{M}(M_\odot {\rm yr}^{-1})=2.06\times
2321: 10^{-8}\frac{L/L_\odot}{v_\infty},
2322: \end{equation}
2323: where $v_\infty$ is the terminal velocity of the super-wind in km
2324: s$^{-1}$. We use $v_\infty$=15 km s$^{-1}$ in this paper.\\
2325: (ii)Based on the simulations of shock-driven winds in the
2326: atmospheres of Mira-like stars in \citet{b88}, \citet{b95} gave a
2327: mass-loss rate similar to Reimers' formula:
2328: \begin{equation}
2329: \dot{M}=4.83\times10^{-9}M^{-2.1}L^{2.7}\dot{M}_{\rm Reimers},
2330: \label{eq:bml}
2331: \end{equation}
2332: where $\dot{M}_{\rm Reimers}$ is given by Eq.(\ref{eq:rml}) but
2333: $\eta=0.02$ \citep{sj07}.
2334:
2335: On other stellar evolutionary phase, the mass-loss rates are given
2336: by Reimers' formula \citep{r75}:
2337: \begin{equation}
2338: \dot{M}=-4.0\times10^{-13}\eta\frac{LR}{M}{\rm M_\odot yr^{-1}},
2339: \label{eq:rml}
2340: \end{equation}
2341: where $L$, $R$, $M$ are the stellar luminosity, radius and mass in
2342: solar units and we use $\eta$=0.5.
2343:
2344: \subsection{Comparison with previously published yields by other models }
2345: In order to test our synthetic model of asymptotic giant branch
2346: stars, we compare the stellar yields in our model with previously
2347: published yields of \cite{vg97}, \cite{m01} and \cite{i04}.
2348: According to the definition of stellar yield in \cite{i04}, we write
2349: the total yield of isotope $j$ by
2350: \begin{equation}
2351: Y_{j}=\int^{\tau(M_{\rm i})}_{0}(X_{j}-X^0_{j})\frac{{\rm d}M}{{\rm
2352: d}t}{\rm d}t ,\label{eq:yield}
2353: \end{equation}
2354: where $\tau(M_{\rm i})$ is the entire lifetime of a star with
2355: initial mass $M_{\rm i}$, $\frac{{\rm d}M}{{\rm d}t}$ is the current
2356: mass-loss rate, $X_{j}$ and $X^0_{j}$ refer to the current and
2357: initial surface abundance of the isotope $j$, respectively.
2358:
2359: Figure \ref{fig:yield} shows the stellar yields of \cite{vg97} [for
2360: the models of $Z=0.02$, $\eta_{\rm AGB}=4.0$ and $m_{\rm HBB}=0.8$
2361: (See Table 17 of \citealt{vg97})], \cite{m01}(for the models of
2362: $Z=0.019$ and the mixing length parameter $\alpha=1.68$),
2363: \cite{i04}[for the models of $Z=0.02$ (See Table D2 of
2364: \citealt{i04})] and ours. In our models, all physical parameters are
2365: the same as those in case 1 except that we change binary systems
2366: into single star. As Figure \ref{fig:yield} shows, the stellar
2367: yields of our models in general are in agreement with those in
2368: \cite{i04} or lie between those of \cite{vg97} and \cite{i04}
2369: although our models produce more $^{15}$N, and less $^{17}$O at
2370: massive stars.
2371: \begin{figure}
2372: \includegraphics[totalheight=3.in,width=2.8in,angle=-90]{f6.ps}
2373: \caption{Total stellar yields vs. the initial stellar masses.
2374: Triangles, circles and squares represent the models of \cite{vg97},
2375: \cite{m01} and \cite{i04}, respectively. Pluses are our results. }
2376: \label{fig:yield}
2377: \end{figure}
2378:
2379:
2380:
2381:
2382: \label{lastpage}
2383:
2384: \end{document}
2385: