1: %\documentclass[manuscript]{aastex}
2: %%
3: \documentclass[12pt,preprint]{emulateapj}
4: %\documentclass[12pt,preprint]{aastex}
5: %\documentclass[preprint]{aastex}
6: %\documentclass[preprint2]{aastex}
7: \usepackage{epsfig}
8: \shorttitle{Fluorine in PNe}
9: \shortauthors{Zhang \& Liu}
10:
11: \begin{document}
12:
13: %\pagestyle{empty}
14:
15: \title{Fluorine abundances in planetary nebulae}
16:
17:
18: \author{Y. Zhang\altaffilmark{1,2} and X.-W. Liu\altaffilmark{1}}
19:
20: \altaffiltext{1}{Department of Astronomy, Peking University, Beijing 100871,
21: P. R. China; E-mail: zhangy@bac.pku.edu.cn.}
22:
23: \altaffiltext{2}{Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD
24: 21218, USA.}
25:
26: \begin{abstract}
27: We have determined fluorine abundances from the [\ion{F}{2}] $\lambda$4789 and
28: [\ion{F}{4}] $\lambda$4060 nebular emission lines for a sample of planetary
29: nebulae (PNe). Our results show that fluorine is generally overabundant in PNe,
30: thus providing new evidence for the synthesis of fluorine in asymptotic giant
31: branch (AGB) stars. [F/O] is found to be positively correlated with the C/O
32: abundance ratio, in agreement with the predictions of theoretical models of
33: fluorine production in thermally pulsing AGB stars. A large
34: enhancement of fluorine is observed in the Wolf-Rayet PN NGC\,40, suggesting
35: that high mass-loss rates probably favor the survival of fluorine.
36: \end{abstract}
37:
38: \keywords{ ISM: abundances --- planetary nebulae: general ---
39: line: identification}
40:
41:
42: \section{Introduction}
43:
44: The astrophysical origin of fluorine is still an unresolved issue. The element
45: has only one stable, yet rather fragile, isotope $^{19}$F. In stellar
46: interiors it is readily annihilated by the most abundant elements hydrogen and
47: helium, via reactions $^{19}$F(p, $\alpha$)$^{16}$O and $^{19}$F($\alpha$,
48: p)$^{22}$Ne, respectively. In order to explain the presence of F a mechanism
49: is required that enables F to escape from the hot stellar interior after it
50: is created. Three scenarios have been proposed as the potential sources of F:
51: explosions of Type {\sc ii} supernovae (SNe), stellar winds of Wolf-Rayet (WR)
52: stars, and the third dredge-up of asymptotic giant branch (AGB) stars.
53: \citet{w88} show that inelastic scattering of neutrinos escaping from the
54: collapsing core of a Type {\sc ii} SN by nuclei in the shell, often referred to
55: as the $\nu$-process, can convert $^{20}$Ne into $^{19}$F. In WR stars,
56: $^{19}$F is probably produced during the He-burning phase and then ejected into
57: space by strong stellar winds before it is destroyed \citep{m00}. For AGB
58: stars, \citet{f92} propose that $^{19}$F is synthesized and then dredged up to
59: the surface during the He-burning thermal pulses. The reaction chain for F
60: production in WR and AGB stars is $^{14}$N($\alpha$,
61: $\gamma$)$^{18}$F($\beta^+$)$^{18}$O(p, $\alpha$)$^{15}$N($\alpha$,
62: $\gamma$)$^{19}$F, where protons are liberated through $^{13}$C($\alpha$,
63: n)$^{16}$O followed by neutron captures $^{14}$N(n, p)$^{14}$C. Based on a
64: semi-analytic model, \citet{r04} suggest that both WR and AGB stars are
65: significant sources of F. To discriminate between the three possible scenarios of F
66: production, accurate measurements of F abundances in these objects are needed.
67:
68: Observations of F outside the solar system have been scarce. A pioneering work
69: was carried out by \citet{j92}, who determined F abundances in red giants using
70: infrared HF vibration-rotation transitions and found enhanced F in C-rich
71: stars, providing evidence of F production in AGB stars. Their results were
72: supported by observations of \ion{F}{5} and \ion{F}{6} absorption lines in the
73: far-UV spectra of hot post-AGB stars \citep{w05}. On the other hand,
74: \citet{c03} measured HF lines for a sample of nine red giants (RGs) in the Large
75: Magellanic Cloud (LMC) and two RGs in the Galactic globular cluster $\omega$
76: Centauri. Very low F/O ratios were found for the two metal-poor $\omega$ Cen
77: giants. Given that the cluster must have been significantly enriched by AGB
78: evolution, as indicated by large enhancements of $s$-process elements observed
79: in its member stars, they concluded that AGB stars are probably not the
80: dominant source of F. To our knowledge, no observation of F in a WR star has
81: been reported to date. The first measurement of interstellar F was obtained by
82: \citet{f05}, who detected \ion{F}{1} $\lambda$955 interstellar absorption in
83: two sight lines towards the Cep OB2 association using the Far Ultraviolet
84: Spectroscopic Explorer ({\it FUSE}). Their measurements yield no evidence of
85: enhanced F abundances resulting from the $\nu$-process in Type {\sc ii} SNe.
86:
87: Planetary nebulae (PNe) are ejected from AGB stars and they therefore
88: serve as a direct test
89: bed for F production in AGB stars. Given its low condensation temperature,
90: depletion of F onto dust grains is unlikely to be significant in PNe.
91: Nevertheless, due to its intrinsic low abundance, F emission lines are
92: difficult to measure in PN spectra. Based on uncertain measurements of the
93: [\ion{F}{4}] $\lambda$4060 nebular line in two PNe, \citet{a83} claimed that the
94: F abundance in PNe was consistent with the solar value. A definite measurement of
95: the F abundance was presented by \citet{l98} for the high-excitation PN NGC\,4361,
96: indicating that F was overabundant in this metal-poor PN possibly belonging to
97: the halo population.
98:
99: In this paper we present F abundances for a sample of PNe and we examine the F
100: production in AGB stars. In Section 2, we survey measurements of the
101: [\ion{F}{2}] $\lambda$4789 and the [\ion{F}{4}] $\lambda$4060 lines available
102: from recently published high quality spectra. The derived abundances are
103: presented in Section 3. A discussion follows in Section 4, and section 5
104: summarizes the results.
105:
106: \begin{figure*}
107: \centering
108: \epsfig{file=f1.eps, height=5.5cm}
109: \caption{a) Spectrum of a low-excitation PN IC\,418 showing the
110: [\ion{F}{2}] $\lambda$4789 line; b) Spectrum of a high-excitation
111: PN NGC\,7027 showing the [\ion{F}{4}] $\lambda$4060 line. Note that
112: interstellar extinction has not been corrected for.}
113: \label{spe}
114: \end{figure*}
115:
116:
117: \section{Fluorine emission lines}
118:
119: Among the ionic species of fluorine, F$^+$ and F$^{3+}$ are detectable in the
120: optical via the [\ion{F}{2}] $^1$D$_2$--$^3$P$_2$ $\lambda4789.48$ and the
121: [\ion{F}{4}] $^1$D$_2$--$^3$P$_2$ $\lambda4059.94$ nebular lines, respectively.
122: We have surveyed high signal-to-noise spectra published in the literature (see
123: the last column of Table~\ref{F_PN}). Many of them are published recently by
124: our own and other research groups, obtained in spectroscopic surveys aimed at
125: detecting weak emission lines from heavy element ions.
126:
127: PNe from which the [\ion{F}{2}] $\lambda4789$ line or the [\ion{F}{4}]
128: $\lambda4060$ line have been detected are listed in Table~\ref{F_PN}. The fact
129: that all PNe where the [\ion{F}{2}] line is detectable are low-excitation PNe
130: and all PNe from which the [\ion{F}{4}] has been detected are high-excitation
131: objects strengthens the confidence in our line identifications. Fig.~\ref{spe}
132: plots the spectra of two PNe, one of low-excitation and another of
133: high-excitation, around the wavelengths of the [\ion{F}{2}] line and the
134: [\ion{F}{4}] line, respectively. The excitation class (E.C.) and the dereddened
135: intensities of the [\ion{F}{2}] $\lambda4789$ and the [\ion{F}{4}]
136: $\lambda4060$ line, relative to that of H$\beta$, are listed respectively in
137: Cols. 2, 3, and 4 of Tables~\ref{F_PN}. Typical intensities of the two fluorine
138: lines relative to H$\beta$ are of the order of $10^{-4}$.
139:
140: \citet{b00} report detection of the [\ion{F}{2}] $\lambda4789$ line from the
141: Orion Nebula, albeit with a low signal-to-noise ratio. A feature at the right
142: wavelength has also been detected by \citet{p04} in the low-excitation PN
143: NGC\,5315, but identified by them as \ion{N}{2}. Given that the feature
144: does not appear in high-excitation PNe, we regard this as a mis-identification.
145: The detection of this line is very uncertain in the low-excitation PN IC\,2501
146: because of severe contamination at this wavelength by scattered light from
147: H$\beta$.
148:
149:
150:
151:
152: \begin{deluxetable*}{lcccccccl}
153: \tabletypesize{\scriptsize}
154: \tablecaption{Measurements of fluorine in nebulae\tablenotemark{a}.
155: \label{F_PN}}
156: \tablewidth{0pt}
157: \tablehead{
158: \colhead{Source} & \colhead{E.C.$^b$} &
159: \colhead{$I$([\ion{F}{2}]$_{4789}$)$^c$} &
160: \colhead{F$^+$/H$^+$} &
161: \colhead{O$^+$/H$^+$}&
162: \colhead{F/O$^d$} & \colhead{C/O}&
163: \colhead{O/H} &\colhead{Ref.}
164: }
165: \startdata
166: %\sidehead{Low-excitation nebulae}
167: IC\,418 & 1 & 2.72$(-4)$ & 4.34$(-8)$ & 1.70$(-4)$ & 2.55$(-4)$&1.8 &2.90$(-4)$ & S04 \\
168: IC\,2501 & 3 & 1.70$(-5)$:& 3.38$(-9)$:& 8.25$(-5)$ & 4.09$(-5)$:&1.2 &5.13$(-4)$ & W05a\\
169: NGC\,40 & 1 & 6.60$(-4)$ & 2.95$(-7)$ & 4.83$(-4)$ & 6.10$(-4)$&1.4 &4.90$(-4)$ & L04 \\
170: NGC\,5315& 2 & 4.40$(-5)$ & 6.45$(-9)$ & 4.17$(-5)$ & 1.54$(-4)$&0.95 &7.41$(-4)$ & P04 \\
171: Orion &\nodata& 5.37$(-5)$ & 9.01$(-9)$ & 7.94$(-5)$ & 1.13$(-4)$&0.58 &4.47$(-4)$ & B00 \\
172: \\[-0.06in]
173: \hline
174: \hline
175: \\[-0.07in]
176: \colhead{Source} & \colhead{E.C.$^{b}$} &
177: \colhead{$I$([\ion{F}{4}]$_{4060}$)$^c$} &
178: \colhead{F$^{3+}$/H$^+$} &
179: \colhead{Ne$^{3+}$/H$^+$}&
180: \colhead{F/O$^d$} & \colhead{C/O}&
181: \colhead{O/H} &\colhead{Ref.} \\
182: \\[-0.09in]
183: \hline
184: \\[-0.07in]
185: %\sidehead{High-excitation nebulae}
186: IC\,2003 &10 & 1.69$(-3)$& 3.49$(-8)$& 4.19$(-5)$ & 1.65$(-4)$& 0.93& 2.75$(-4)$ & W05b\\
187: NGC\,2022&12 & 1.84$(-3)$& 2.31$(-8)$& 4.47$(-5)$ & 7.82$(-5)$& 0.46& 4.57$(-4)$ & T03\\
188: NGC\,2440&10 & 2.17$(-4)$& 3.28$(-9)$& 5.29$(-5)$ & 1.26$(-5)$& 0.90& 4.40$(-4)$ & W05a\\
189: NGC\,3242&9 & 6.48$(-5)$& 1.75$(-9)$& 9.18$(-6)$:&4.46$(-5)$:& 0.41& 3.31$(-4)$ & T03\\
190: NGC\,3918&9 & 3.75$(-4)$& 7.84$(-9)$& 5.95$(-5)$ & 1.69$(-5)$& 0.60& 7.24$(-4)$ & T03\\
191: NGC\,4361&12$^+$& 3.10$(-3)$& 2.57$(-8)$& 2.54$(-5)$ & 6.82$(-4)$& 2.0 & 6.61$(-5)$ & L98\\
192: NGC\,6302&10 & 6.77$(-4)$& 4.99$(-9)$& 2.37$(-5)$ & 6.35$(-5)$& 0.30& 2.51$(-4)$ & T03\\
193: NGC\,7027&11 & 9.10$(-4)$& 1.71$(-8)$& 4.19$(-5)$ & 1.04$(-4)$& 2.7 & 4.57$(-4)$ & Z05 \\
194: NGC\,7662&10 & 8.50$(-4)$& 1.47$(-8)$& 4.30$(-5)$ & 5.93$(-5)$& 1.4 & 3.31$(-4)$ & L04 \\
195: \enddata
196: \tablenotetext{a}{Here $a(-b)=a\times10^{-b}$;
197: $^{b}$From \citet{g97}; $^{c}$Corrected for reddening, and on a scale
198: where H$\beta=1$; $^{d}$The solar F/O ratio is 5.89$(-5)$ \citep{lodders}.}
199: \tablerefs{ [B00] \citet{b00}; [L98] \citet{l98}; [L04] \citet{l04}; [P04] \citet{p04}; [S04] \citet{s03};
200: [T03] \citet{t03}; [W05a] \citet{w05a}; [W05b] \citet{w05b}; [Z05] \citet{z05}.
201: }
202: \end{deluxetable*}
203:
204: An unidentified weak feature has been detected in the echelle spectrum of
205: IC\,418 at 4869\,{\AA} \citep[cf.][]{s03}. We identify it as the [\ion{F}{2}]
206: $^1$D$_2$--$^3$P$_1$ transition at 4869.02\,{\AA}. Given the fact that the
207: [\ion{F}{2}] $\lambda\lambda4789,4869$ lines arise from the same upper level,
208: the $\lambda4789/\lambda4869$ intensity ratio depends only on their spontaneous
209: transition probabilities, and not on the nebular physical conditions. For
210: IC\,418, the measurements yield a $\lambda4789/\lambda4869$ intensity ratio of
211: 5.2. In the spectrum of the bright high-excitation PN NGC\,7027, we have
212: detected the [\ion{F}{4}] $^1$D$_2$--$^3$P$_1$ transition at 3996.96\,{\AA}
213: \citep[cf.][]{z05}. The measurements yield an intensity ratio
214: $I(4060)/I(3997)=3.6$. Note that under the assumption of $LS$-coupling, the
215: intensity ratio of the $^1$D$_2$ -- $^3$P$_2$ to $^1$D$_2$ -- $^3$P$_1$
216: transitions is simply given by $3\times [\lambda$($^1$D$_2$ --
217: $^3$P$_1$)/$\lambda$($^1$D$_2$ -- $^3$P$_2$)$]^3$. This relation predicts an
218: intensity ratio of 3.2 and 2.8 for [\ion{F}{2}] $\lambda4789/\lambda4869$ and
219: [\ion{F}{4}] $\lambda4060/\lambda3997$, respectively. On the other hand, in
220: the well studied case of the much stronger [\ion{O}{3}]
221: $\lambda\lambda$4959,5007 nebular lines, based on high quality observations,
222: \citet{m99} demonstrate that $\lambda5007/\lambda4959 = 3.00\pm0.01$,
223: significantly higher than the value of 2.88 predicted by the above theoretical
224: relation under $LS$-coupling. \citet{s00} have recently recalculated intensity
225: ratios of the $^1$D$_2$ -- $^3$P$_2$ to $^1$D$_2$ -- $^3$P$_1$ transitions for
226: carbon-like and oxygen-like ions, taking into account the effects of the
227: relativistic corrections for the magnetic dipole operators. For [\ion{O}{3}],
228: they find that the $\lambda5007/\lambda4959$ ratio increases from the
229: $LS$-coupling value 2.88 to 2.98; the latter is in excellent agreement with the
230: measurement of \citet{m99}. For [\ion{F}{2}] and [\ion{F}{4}], the
231: corresponding line ratios obtained by \citet{s00} are 3.12 and 2.90,
232: respectively. Considering the uncertainties in measuring the extremely faint
233: [\ion{F}{2}] and [\ion{F}{4}] lines, especially the weaker [\ion{F}{2}]
234: $\lambda$4869 line, which falls very close to the strong H$\beta$ and suffers
235: from severe contamination by scattered light from H$\beta$, the discrepancies
236: between the observations and theory, i.e. 5.2 versus 3.12 in the case of
237: [\ion{F}{2}] in IC\,418, and 3.6 versus 2.90 in the case of [\ion{F}{4}] in
238: NGC\,7027, are probably insignificant.
239:
240: Finally, we note that in the infrared, the [\ion{F}{4}] $25.8\mu$m and
241: $44.2\mu$m fine-structure lines, analogues of the [\ion{O}{3}] $51.8\mu$m and
242: $88.3\mu$m lines, have intensities comparable to the optical [\ion{F}{4}]
243: $\lambda\lambda3996,4060$ lines, and are hence good candidates to search for
244: fluorine in PNe, especially in those that suffer from large dust extinction in the
245: optical.
246:
247: \section{Fluorine abundance determinations}
248:
249: F$^+$/H$^+$ and F$^{3+}$/H$^+$ ionic abundances have been derived from the
250: observed intensities of the [\ion{F}{2}] and [\ion{F}{4}] lines, respectively,
251: following standard procedures. Multilevel collisional-radiative models were
252: constructed for F$^+$ and F$^{3+}$, using collision strengths from \citet{b94}
253: and \citet{l94}, and radiative transition probabilities from \citet{b88} and
254: \citet{f85}, respectively. When calculating F$^+$ ionic abundances, we have
255: adopted electron densities derived from the [\ion{S}{2}]
256: $\lambda6731/\lambda6716$ ratio and temperatures deduced from the [\ion{N}{2}]
257: nebular-to-auroral line ratio. For F$^{3+}$, densities derived from the
258: [\ion{Ar}{4}] $\lambda4740/\lambda4711$ ratio and temperatures obtained from
259: the [\ion{O}{3}] nebular-to-auroral line ratio were used. Given the relatively
260: high critical densities ($>10^6$\,cm$^{-3}$) of the [\ion{F}{2}] $\lambda$4789
261: and [\ion{F}{4}] $\lambda$4060 lines, the abundances are essentially
262: independent of the adopted electron densities. For NGC\,4361, we have adopted
263: the F$^{3+}$/H$^+$ ratio of \citet{l98}, obtained using the same atomic data
264: set. The results are presented in Table~\ref{F_PN}.
265:
266: To convert ionic abundance ratios to those of total elemental abundances,
267: ionization corrections need to be made for unseen ions. Given that O$^+$ and
268: Ne$^{3+}$ have ionization potentials comparable to F$^{+}$ and F$^{3+}$,
269: respectively, we assume that ${\rm F}/{\rm O}={\rm F}^+/{\rm O}^+$ and ${\rm
270: F}/{\rm O}=({\rm F}^{3+}/{\rm Ne}^{3+})({\rm Ne}/{\rm O})$, for low- and
271: high-excitation PNe, respectively. Ionic and elemental abundances of Ne, O and
272: C adopted in our analysis were taken from the literature. Table~\ref{F_PN}
273: summarizes the abundances. Typical uncertainties in the F/O ratios are
274: about 12 per\,cent.
275:
276: \section{Discussion}
277:
278: \begin{figure}
279: \epsfig{file=f2.eps,
280: height=6cm, bbllx=100, bblly=279, bburx=528, bbury=602, clip=, angle=0}
281: \caption{[F/O] versus C/O. Open and filled circles denote low- and
282: high-excitation PNe, respectively. A correlation between the two quantities
283: is apparent. The error bar on the lower right indicates typical uncertainties
284: of [F/O].
285: \protect\label{co}}
286: \end{figure}
287:
288: \begin{figure}
289: \epsfig{file=f3.eps,
290: height=6cm, bbllx=100, bblly=279, bburx=528, bbury=602, clip=, angle=0}
291: \caption{[F/O] versus O/H. Open circles and filled circles denote low- and
292: high-excitation PNe, respectively. With the exception of NGC\,40 (represented
293: by a filled triangle), an anti-correlation between F/O and O/H is
294: observed amongst other PNe.
295: \protect\label{fo}}
296: \end{figure}
297:
298:
299: The most recent measurement of the solar F/O abundance ratio yields ${\rm F/O}
300: = 5.89\times$10$^{-5}$ \citep{lodders}. Table~\ref{F_PN} shows that most PNe
301: exhibit F overabundances. The sample yields an average value [F/O]$=$0.3. Our
302: measurements thus provide further evidence in support of \citet{j92} who
303: suggest that AGB stars can be a significant source of F production.
304:
305: According to the AGB evolution models of \citet{j92}, fluorine is synthesized
306: in the He intershell and then dredged up to the surface during the thermal
307: pulses. During this process, He is converted into $^{12}$C via partial
308: He-burning in the intershell and then transported to the stellar surface along
309: with F. The scenario thus predicts a positive correlation between the [F/O]
310: abundance enhancement and the C/O ratio. Such a correlation is indeed observed
311: in red giants \citep{j92} and in our current sample of PNe. In Fig.~\ref{co},
312: we plot observed [F/O] versus C/O ratio for the sample. It is clear that [F/O]
313: increases with increasing C/O ratio, which strongly suggests that F is indeed
314: produced in He-burning environments of thermally pulsing AGB stars.
315: Fig.~\ref{co} also shows that there is no systematic difference between the F
316: abundances deduced for low- and high-excitation PNe in our sample, suggesting
317: that the ionization correction formulae adopted in our analysis are probably
318: reliable.
319:
320: No PN in our sample shows [F/O] overabundances as much as 30 times solar, as
321: found by \citet{j92} in giant stars. The highest [F/O] value observed in our
322: sample is in the halo PN NGC\,4361, which yields a F/O ratio 12 times solar.
323: \citet{lugaro04} present new calculations of F yields in AGB stars. Their
324: models fail to match the very high F abundances found by \citet{j92}. A
325: comparison with Fig.~3 of \citet{lugaro04} shows that the [F/O] and C/O
326: abundance ratios derived here for the current sample of PNe are in good
327: agreement with the predictions of their 3\,$M_\sun$ and $Z=0.02$ model. The
328: very high F abundances observed in some red giants thus remain a puzzle.
329:
330: A large F enhancement of 10 times solar is observed in NGC\,40, a PN ionized by
331: a WC8 central star \citep{smith1969,crowther}. The large enhancement is
332: probably a consequence of the strong stellar winds from its progenitor star,
333: leading to a larger amount of the synthesized F being ejected into space. It is
334: possible that WR PNe play an important role in F production. Further
335: spectroscopy of additional WR PNe is however required to verify this
336: possibility.
337:
338: Fig.~\ref{fo} presents [F/O] versus O/H for the sample. Excluding the WR PN
339: NGC\,40, a negative correlation is seen, indicating that the dredge-up process
340: that produces F is more efficient in a metal-poor environment. The result is in
341: accord with the predictions of the dredge-up models \citep[e.g.][]{bs88} and
342: the fact that more carbon stars are observed in metal-poor galaxies
343: \citep{g04}. NGC\,40 does not follow this general trend, however
344: it is consistent with
345: the idea that strong stellar winds may have dramatically enhanced F in this
346: nebula (see above).
347:
348: It has been found that $s$-elements are generally overabundant in PNe
349: \citep{w05a}, indicating that AGB stars experience strong slow neutron
350: capture process. The
351: origin of the neutrons, however, is much debated \citep[see][for a review]{b99}.
352: Two possible sources have been proposed, $^{13}$C($\alpha$, n)$^{16}$O and
353: $^{22}$Ne($\alpha$, n)$^{25}$Mg. If the later process dominates, then F is
354: unlikely to survive in the He-intershell since the reaction rate of
355: $^{19}$F($\alpha$, n)$^{22}$Ne is much higher than that of $^{22}$Ne($\alpha$,
356: n)$^{25}$Mg \citep[see][for a detailed discussion]{j92}. Consequently, our
357: finding that F is generally enhanced in PNe rules out reaction
358: $^{22}$Ne($\alpha$, n)$^{25}$Mg as the main process of neutron production.
359:
360: In this paper, we have also presented, for the first time, the F abundance for an
361: \ion{H}{2} region, the Orion Nebula (see Table~\ref{F_PN}). The value is about
362: two times higher than those found for K--M dwarfs in the Orion Nebula Cluster
363: \citep{c05}. The significance of the finding is however questionable, given the
364: considerable uncertainties in the line measurements -- according to
365: \citet{b00}, the extremely faint [\ion{F}{2}] $\lambda4789$ nebular line has a
366: signal-to-noise of only 5.3. Further observation of improved sensitivity should
367: be helpful.
368:
369: \section{Conclusions}
370:
371: We have obtained measurements of the [\ion{F}{2}] $\lambda4789$ and
372: [\ion{F}{4}] $\lambda4060$ nebular emission lines from the recent literature
373: for a sample of PNe. The lines are detectable in low- and high-excitation
374: nebulae, respectively. From the measurements, abundances of fluorine are
375: derived. Our results show that fluorine is overabundant in PNe compared to the
376: solar value, suggesting that thermally pulsing AGB stars may play an important
377: role in the production of the element. The result is further corroborated by
378: the positive correlation observed between the derived [F/O] and C/O. Given that
379: the WR PN NGC\,40 shows a high F/O ratio, we infer that PNe with a WR central
380: star are probably significant fluorine producers. Additional observations are
381: required to verify this point.
382:
383: \acknowledgments We thank the anonymous referee for useful comments. YZ deeply
384: appreciates R. Williams' hospitality during
385: the year he worked in STScI. The work is partly supported by NNSFC grant
386: \#10325312.
387:
388:
389: \begin{thebibliography}{}
390:
391: \bibitem[Aller \& Czyzak (1983)]{a83} Aller, L. H., \& Czyzak, S. J.
392: 1983, ApJS, 51, 211
393:
394: \bibitem[Baldwin et al.(2000)]{b00} Baldwin, J. A. et al.
395: 2000, \apjs, 129, 229
396:
397: \bibitem[Baluja \& Zeippen(1988)]{b88} Baluja, K. L., \&
398: Zeippen, C. J. 1988, J. Phys. B., 21, 1455
399:
400: \bibitem[Boothroyd \& Sackmann(1988)]{bs88} Boothroyd, A. I., \&
401: Sackmann, I.-J. 1988, \apj, 328, 671
402:
403: \bibitem[Busso, Gallino \& Wasserburg (1999)]{b99} Busso, M.,
404: Gallino, R., \& Wasserburg, G. J. 1999, \araa, 37, 239
405:
406: \bibitem[Butler \& Zeippen(1994)]{b94} Butler, K., \&
407: Zeippen, C. J. 1994, \aaps, 108, 1
408:
409: \bibitem[Crowther et al.(1970)]{crowther} Crowther, P. A., De Marco, O., \& Barlow, M. J. 1998, \mnras, 296, 36
410:
411:
412: \bibitem[Cunha et al. (2003)]{c03} Cunha, K., Smith, V. V., Lambert,
413: D. L., \& Hinkle, K. H. 2003, \apj, 126, 1305
414:
415: \bibitem[Cunha \& Smith (2005)]{c05} Cunha, K., \& Smith, V. V.
416: 2005, \apj, 626, 425
417:
418: \bibitem[Federman et al. (2005)]{f05} Federman, S. R.,
419: Sheffer, Y., Lambert, D. L., \& Smith, V. V. 2005, \apj, 619, 884
420:
421: \bibitem[Fisher \& Saha (1985)]{f85} Fisher, C. F., \& Saha, H. P.
422: 1985, Phys. Scr., 32, 181
423:
424: \bibitem[Forestini et al. (1992)]{f92} Forestini, M., Goriely, S.,
425: Jorissen, A., \& Arnould, M. 1992, A\&A, 261, 157
426:
427:
428: \bibitem[Gurzadyan (1997)]{g97} Gurzadyan, G. A. 1997, The Physics and Dynamics of Planetary Nebulae (Berlin: Springer)
429:
430:
431: \bibitem[Groenewegen(2004)]{g04} Groenewegen, M. A. T. 2004, in
432: Planetary Nebulae beyond the Milky Way, eds. J. Walsh, L. Stanghellini and N. Douglas (ESO, Garching bei M\"{u}nchen), in press (astro-ph/0407282)
433:
434:
435:
436: \bibitem[Jorissen et al. (1992)]{j92} Jorissen, A., Smith, V. V.,
437: \& Lambert, D. L. 1992, A\&A, 261, 164
438:
439:
440: \bibitem[Lennon \& Burke (1994)]{l94} Lennon, D. J., \& Burke, V. M.
441: 1994, \aaps, 103, 273
442:
443: \bibitem[Liu (1998)]{l98} Liu, X.-W. 1998, \mnras, 295, 699
444:
445: \bibitem[Liu et al.(2004)]{l04} Liu, Y., Liu, X.-W., Luo, S.-G. \&
446: Barlow, M. J. 2004, MNRAS, 353, 1231
447:
448: \bibitem[Lodders(2003)]{lodders} Lodders, K. 2003, \apj, 591, 1220
449:
450: \bibitem[Lugaro et al.(2004)]{lugaro04} Lugaro, M. et al.
451: 2004, \apj, 615, 934
452:
453: \bibitem[Mathis \& Liu (1999)]{m99} Mathis, J. S., \& Liu, X.-W.
454: 1999, \apj, 521, 212
455:
456: \bibitem[Meynet \& Arnould (2000)]{m00} Meynet, G., \&
457: Arnould, M. 2000, A\&A, 355, 176
458:
459:
460: \bibitem[Peimbert et al.(2004)]{p04} Peimbert, M., Peimbert, A.,
461: Ruiz, M. T., \& Esteban, C. 2004, \apj, 150, 431
462:
463:
464: \bibitem[Renda et al. (2004)]{r04} Renda, A. et al. 2004, MNRAS, 354, 575
465:
466:
467: \bibitem[Sharpee et al.(2003)]{s03} Sharpee, B., Williams, R., Baldwin, J. A., \& van Hoof, P. A. M.
468: 2003, \apj, 149, 157
469:
470: \bibitem[Smith(1969)]{smith1969} Smith, L. F., \& Aller, L. H. 1969, \apj, 157, 1245
471:
472: \bibitem[Storey \& Zeippen(2000)]{s00} Storey, P. J., \& Zeippen, C. J.
473: 2000, \mnras, 312, 813
474:
475: \bibitem[Tsamis et al.(2003)]{t03} Tsamis, Y. G., Barlow, M. J., Liu,
476: X.-W., Danziger, I. J., \& Storey, P. 2003, \mnras, 345, 186
477:
478:
479: \bibitem[Werner et al. (2005)]{w05} Werner, K., Rauch, T., \& Kruk, J.
480: W. 2005, A\&A, 433, 641
481:
482: \bibitem[Wesson et al. (2005)]{w05b} Wesson, R. J., Liu,
483: X.-W., \& Barlow, M. J. 2005, \mnras, in press
484:
485: \bibitem[Williams et al. (2005)]{w05a} Williams, R., Zhang, Y., Pellegrini, E.,
486: Cavagnolo, K., Baldwin, J. A., Sharpee, B., Phyllips, M., \& Liu, X.-W. 2005,
487: \apj, submitted
488:
489: \bibitem[Woosley \& Haxton (1988)]{w88} Woosley, S. E., \& Haxton,
490: W. C. 1988, Nature, 334, 45
491:
492: \bibitem[Zhang et al(2005)]{z05} Zhang, Y., Liu, X.-W., Luo, S.-G.,
493: P{\' e}quignot, D., \& Barlow, M. J. 2005, \aap, in press (astro-ph/0507155)
494:
495: \end{thebibliography}
496:
497:
498: \end{document}
499: