1: \documentclass[12pt,preprint]{aastex}
2: %\usepackage{longtable}
3: %\def\refe {\par \noindent \parshape=6 0cm 12.5cm 0.5cm 12.5cm 0.5cm 12.5cm 0.5cm 12.5cm 0.5cm 12.5cm 0.5cm 12.5cm}
4:
5: \slugcomment{Accepted by the Astronomical Journal}
6:
7: \begin{document}
8:
9: \def\ang{{\rm\AA}}
10:
11: \def\n{\noindent}
12:
13: \def\ph{\phantom}
14:
15: \newcommand{\myfigure}[4]{
16: \begin{figure*}
17: \setbox100=\hbox{
18: \epsfxsize=#1 cm
19: \epsfbox{#2.ps}}
20: \centerline{\hbox{\box100}}
21: \caption{#3}
22: \label{#4}
23: \end{figure*}
24: }
25:
26: \title{The Ultraviolet Spectra of the Weak Emission Line Central
27: Stars of Planetary Nebulae}
28:
29: \author{W. L. F. Marcolino\altaffilmark{1,2}}
30: \email{wagner@on.br}
31:
32: \author{F. X. de Ara\'ujo\altaffilmark{1}}
33: \email{araujo@on.br}
34:
35: \author{H. B. M. Junior\altaffilmark{1,3}}
36: \email{helson@on.br}
37:
38: \and
39:
40: \author{E. S. Duarte\altaffilmark{3}}
41: \email{esduarte@cefeteq.br}
42:
43: \altaffiltext{1}{Observat\'orio Nacional/MCT, Rua Gen. Jos\'e Cristino 77,
44: 20921-400, Rio de Janeiro, Brazil.}
45:
46: \altaffiltext{2}{Laboratoire d'Astrophysique de Marseille, Traverse du Siphon-Bp 8, 13376 Marseille Cedex 12, France.}
47:
48: \altaffiltext{3}{Centro Federal de Educa\c c\~ao Tecnol\'ogica de Qu\'imica de
49: Nil\'opolis, 26530-060, Rio de Janeiro, Brazil.}
50:
51: %\bigskip
52:
53: \begin{abstract}
54:
55: The ultraviolet spectra of all ``weak emission line central stars of
56: planetary nebulae'' (WELS) with available {\it IUE} data are presented and discussed.
57: We performed line identifications, equivalent width and flux measurements for
58: several features in their spectra. We found that the WELS can be divided in three
59: different groups regarding their UV: i) Strong P-Cygni profiles
60: (mainly in \ion{C}{4} $\lambda 1549$); ii) Weak P-Cygni features
61: and iii) Absence of P-Cygni profiles. The last group encompasses
62: stars with a featureless UV spectrum or with intense emission lines and a
63: weak continuum, which are most likely of nebular origin. We have measured wind
64: terminal velocities for all objects presenting P-Cygni profiles in
65: \ion{N}{5} $\lambda 1238$ and/or \ion{C}{4} $\lambda 1549$.
66: The results obtained were compared to the UV data of the two prototype stars of
67: the [WC]-PG 1159 class, namely, A30 and A78. For WELS presenting P-Cygnis,
68: most of the terminal velocities fall in the range $\sim 1000-1500$ km
69: s$^{-1}$, while [WC]-PG 1159 stars possess much higher values, of
70: $\sim$3000 km s$^{-1}$. The [WC]-PG1159 stars are characterized by
71: intense, simultaneous P-Cygni emissions in the $\sim 1150-2000$\AA\,
72: interval of \ion{N}{5} $\lambda 1238$, \ion{O}{5} $\lambda 1371$ and
73: \ion{C}{4} $\lambda 1549$. In contrast, we found that
74: \ion{O}{5} $\lambda 1371$ is very weak or absent in the WELS
75: spectra. On the basis of the ultraviolet spectra alone, our findings
76: indicate that [WC]-PG 1159 stars are distinct from the WELS, contrary
77: to previous claims in the literature.
78:
79: \end{abstract}
80:
81: \keywords{planetary nebulae:general - stars:fundamental parameters -
82: stars:atmospheres - ultraviolet:stars}
83:
84: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
85:
86: \section{INTRODUCTION}
87:
88: Central stars of planetary nebulae (CSPN) that are hydrogen deficient
89: are generally divided in three main groups: [WR], PG 1159 and [WC]-PG 1159
90: stars. The first one present spectra with strong and broad emission lines
91: mainly from He, C and O, similar to the Wolf-Rayet stars (WR) of Population I.
92: Depending on the ionization stages of the elements dominating the atmosphere,
93: [WR] stars are subdivided in [WCL], [WCE] and [WO] (Crowther et al. 1998;
94: Acker \& Neiner 2003). On the other hand, PG 1159 stars are quite distinct
95: objects. They are pre-white dwarfs and show mainly absorption lines of \ion{He}{2} and
96: \ion{C}{4} in their spectra (Werner et al. 1997). Only a handful of
97: PG 1159 stars are known to possess wind features (Koesterke et al. 1998;
98: Koesterke \& Werner 1998). The [WC]-PG 1159 group present strong P-Cygni
99: lines in the ultraviolet (e.g. \ion{N}{5} $\lambda 1238$ and \ion{C}{4}
100: $\lambda 1549$) and resemble the PG 1159 stars in the optical. Three objects
101: are considered prototypes of this class: A 30, A 78 and Longmore 4 (at
102: outburst).
103:
104: The origin and evolution of hydrogen deficient CSPN have been investigated
105: from different point of views during the last decade. Evolutionary
106: models are now able to provide a reasonable match to the observed
107: chemical abundances, although it is still debated the role
108: of binarity and different thermal pulse models (De Marco \& Soker 2002; Herwig
109: 2001). Nebular analyses as well as sophisticated non LTE atmosphere models have been
110: also very useful to address several questions regarding these central stars:
111: it is generally inferred that the groups above mentioned form
112: the following evolutionary sequence: late type [WR] $\rightarrow$ early-type
113: [WR] $\rightarrow$ [WC]-PG 1159 $\rightarrow$ PG 1159 $\rightarrow$ non DA
114: white dwarfs (see e.g. Zijlstra et al. 1994; Pen\~a et al. 2001; Koesterke
115: 2001). However, this scenario has important issues unsolved, such as the C/He
116: mass ratio and the exact position of [WC]-PG 1159 stars in the HR diagram
117: (Hamann 1997; Marcolino et al. 2007). Moreover, the evolutionary status
118: of the so-called ``weak emission line stars'' and their relation to the
119: [WC]-PG 1159 stars is not at all clear. Let us discuss it now.
120:
121: In an extensive observational study, Tylenda et al. (1993) analyzed
122: spectroscopy data of 77 hydrogen deficient CSPN.
123: In their sample, 39 were classified as [WR] stars. The
124: remaining objects were called ``weak emission line stars'' (WELS or [WELS]).
125: According to these authors, the WELS show emission of \ion{C}{4} $\lambda 5805$
126: (actually \ion{C}{4} $\lambda \lambda 5801, 12$) systematically
127: weaker and narrower than in [WR] stars, and a feature in $\lambda 4650$,
128: which is possibly a blend of \ion{N}{3}, \ion{C}{3} and \ion{C}{4} emissions.
129: Moreover, \ion{C}{3} $\lambda 5696$ is very weak or absent.
130: These optical characteristics were confirmed by Marcolino \& de Ara\'ujo
131: (2003) with a homogeneous and higher resolution set of data.
132: Interestingly, by comparing a large sample of WELS, [WC]-PG 1159 and
133: PG 1159 spectra, Parthasarathy et al. (1998) claimed that
134: the WELS and the [WC]-PG 1159 stars actually constitute the same class.
135: However, this claim was based solely on comparisons of optical spectra and
136: was not confirmed by further studies. Indeed, some authors
137: argued that this assertion should be taken with caution until
138: an analysis of a larger sample of objects is performed
139: (e.g. Werner \& Herwig 2006).
140:
141: Undoubtedly, the ``weak emission line stars'' (WELS) constitute
142: the least understood class of hydrogen deficient CSPN. An important point
143: is that the WELS might not be descendants of the [WR] stars, as it is
144: explicited in the [WR] $\rightarrow$ [WC]-PG 1159 (=WELS) $\rightarrow$
145: PG 1159 evolution. Pen\~a et al. (2003) for example, derived lower average nebular
146: expansion velocities for WELS than for [WR] stars, while the
147: contrary would be expected from the long-term action of a stellar
148: wind on a planetary nebula during a [WR] $\rightarrow$ WELS transition.
149: Furthermore, based on a kinematical study of a large sample of planetary nebulae, Gesicki et al. (2006)
150: raised the interesting possibility that some WELS can be progenitors
151: of the hottest [WR] stars, i.e., the [WO] group (WELS $\rightarrow$ [WO]).
152: A major issue hinders the determination of the evolutionary status of
153: the WELS: their physical parameters (e.g. $T_{eff}$, $v_\infty$
154: and $\dot{M}$) and chemical abundances remain unknown. The properties
155: of A 30, A 78, and Longmore 4 ([WC]-PG 1159 stars) are known
156: (Koesterke 2001) but again, their identity as WELS is questionable.
157:
158: So far, most of previous studies involving WELS were done in the optical part of
159: the spectrum. In fact, the very definition of a WELS is based in the optical
160: features in $\lambda 4650$, \ion{C}{4} $\lambda 5805$ and \ion{C}{3} $\lambda 5696$
161: (Tylenda et al. 1993). A few works in the ultraviolet including some WELS can
162: be found in the literature. However, they often include stars of different
163: spectral classes (e.g. Feibelman 2000), and/or are focused in the
164: determination of nebular properties (e.g. Adams \& Seaton 1982;
165: Pottasch et al. 2005), without special attention to the WELS and their
166: evolutionary state. Motivated by this fact and considering the open questions above described,
167: we investigate in this paper the UV spectra of all WELS with available
168: {\it IUE} ({\it International Ultraviolet Explorer}) data.
169: Our main aims are to understand their main UV characteristics; identify and
170: measure the most intense spectral lines; measure wind terminal velocities from
171: the available P-Cygni profiles; and finally, to compare the results to the
172: data of the two prototype [WC]-PG 1159 stars: A30 and A78.
173:
174: The present paper is organized as follows: in Section 2 we present
175: the observational data retrieved from the Multi-mission Archive at STScI
176: (MAST); in Section 3 we discuss the main characteristics
177: of the UV spectra of the WELS, and present line identifications
178: of the most conspicuous features, as well as equivalent
179: width and line flux measurements. In Section 4
180: we empirically measure the terminal velocities for
181: all objects presenting P-Cygni profiles in \ion{N}{5} $\lambda 1238$
182: and/or \ion{C}{4} $\lambda 1549$, from low and high resolution data.
183: Finally, in Section 5, we discuss and compare the results obtained for
184: the WELS to the data of A 30 and A 78 (the two prototype [WC]-PG 1159 stars),
185: and present the main conclusions of our work.
186:
187: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
188:
189: \section{OBSERVATIONAL DATA}
190:
191: We used public data from the {\it IUE} ({\it International Ultraviolet
192: Explorer}) satellite, which are available at the Multi-mission Archive
193: at STScI (MAST)\footnote{http://archive.stsci.edu/}. The total
194: number of WELS currently known is about 50. Most of them are
195: listed in the work of Tylenda et al. (1993) and Parthasarathy
196: et al. (1998), and 15 were recently discovered in the
197: direction of the Galactic bulge by G\'orny et al. (2004). Two objects
198: considered as WELS in Parthasarathy et al. (1998) (Hen 2-86 and M 2-31)
199: were found to be actually early type [WR] stars by Marcolino \& de Ara\'ujo
200: (2003) on the basis of the equivalent width of the \ion{C}{4} $\lambda 5805$
201: line.
202:
203: We have searched for {\it IUE} data for all WELS known, and found spectra
204: for only $\sim 40\%$ of the total WELS population. The list of
205: SWP, LWP and LWR archives considered in our study, including the
206: ones for A30 and A78 (the two [WC]-PG 1159 stars), is shown in Table \ref{objects}.
207: The majority of the spectra disponible are of low resolution ($\sim 6$\AA).
208: The high resolution spectra ($\sim 0.2$\AA) which were utilized are also
209: shown in the table. Most of the observations were done using the {\it IUE}
210: large aperture mode ($10'' \times 20''$).
211:
212: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
213:
214: \section{THE ULTRAVIOLET SPECTRUM OF WELS}
215: \label{uvwels}
216:
217: The ``weak emission line'' class of central stars is defined based on
218: the optical spectrum. Tylenda et al. (1993) showed that the WELS are
219: characterized mainly by a feature in $\lambda 4650$ (possibly a sum of
220: \ion{N}{3}, \ion{C}{3} and \ion{C}{4}), and the \ion{C}{3} $\lambda 5696$ and
221: \ion{C}{4} $\lambda 5805$ transitions. Because their \ion{C}{4} $\lambda 5805$ line
222: is usually weak, a [WCL] spectral classification seems reasonable for
223: these stars at a first glance. However, the \ion{C}{3} $\lambda 5696$ line
224: is generally strong in the [WCL] class and it is weak or absent
225: in the WELS. Conversely, because the \ion{C}{3} $\lambda 5696$ is
226: weak or absent, a [WCE] classification seems plausible. However,
227: the \ion{C}{4} $\lambda 5805$ in [WCE] stars is much stronger than
228: in the WELS. Besides these characteristics, the WELS are also known
229: to present some absorptions in the optical from lines such as
230: \ion{He}{2} $\lambda 4541$ and $\lambda 5412$ (Parthasarathy et al. 1998;
231: Marcolino \& de Ara\'ujo 2003), indicating a stellar wind not
232: so dense as it is the case of the [WR] stars, where the photospheric
233: part is completely hidden.
234:
235: Although we can say that the main characteristics of the WELS
236: are known in the optical, their ultraviolet spectra were never
237: discussed before. As we have mentioned, although some works
238: in the UV involved some WELS, no attention was given to their
239: evolutionary status or relation to the [WC]-PG 1159 stars.
240: The present paper addresses directly this issue. Instead of
241: presenting a homogeneous set of features, we found that the WELS
242: can be divided in three main groups regarding their ultraviolet spectra:
243:
244: \begin{enumerate}
245:
246: \item presence of a strong P-Cygni in \ion{C}{4} $\lambda 1549$;
247:
248: \item weak \ion{C}{4} $\lambda 1549$ in P-Cygni;
249:
250: \item absence of P-Cygni features.
251:
252: \end{enumerate}
253:
254: In Fig. \ref{group1} we show stars that are representatives of the first
255: group. The total number of objects is 11: NGC 6629, NGC 6891, Hen 2-12,
256: PB 8, Hen 2-108, NGC 6543, NGC 6567, NGC 6572, Vy 2-1, Hb7, and IC 4776.
257: As can be seen, the \ion{C}{4} $\lambda 1549$ transition
258: is often accompanied by \ion{N}{5} $\lambda 1238$,
259: also in P-Cygni. Some other lines present are \ion{He}{2} $\lambda 1640$
260: (e.g. PB 8), \ion{Si}{4} $\lambda \lambda 1394, 1403$ (e.g. Hen 2-108),
261: and \ion{C}{3} $\lambda 1909$ (e.g. NGC 6891). A feature near $\sim 1720$\AA\,
262: is also exhibited by several objects, and it might be due to \ion{N}{4}
263: or \ion{Si}{4}. Line identifications, measurements of equivalent widths
264: and fluxes for all objects will be presented in the next section.
265:
266: Only two stars were assigned to the second group, namely,
267: CN 3-1 and Hen 2-131. Their spectra are shown
268: in Fig. \ref{group23} (top and 2nd panel). The \ion{C}{4}
269: $\lambda 1549$ line is clearly weaker than in group 1.
270: As will be described later in Section \ref{vinfsec}, these two stars present also
271: significantly lower terminal velocities. Despite weak,
272: the transitions \ion{Si}{4} $\lambda \lambda 1394, 1403$ can also
273: be identified. The remaining objects are: J900, IC 4997, NGC 5873, NGC 6818, NGC
274: 6803, and M 1-71. They compose the third group, and some of them are also shown
275: in Fig. \ref{group23} (3rd to bottom panel). It can be readily noted the
276: presence of strong emissions and of a weak and flat continuum. No P-Cygni
277: emissions are seen. These characteristics suggest that
278: only nebular features were detected by the {\it IUE} observations.
279: Other archives than the ones listed in Table \ref{objects} were
280: examined, but they all present similar spectra.
281: An exception is the case of IC 4997. By analyzing another {\it IUE}
282: archival spectrum (SWP08578) besides the one shown in Fig. \ref{group23} (SWP31683), we
283: could see a slightly higher continuum and a \ion{N}{5} $\lambda 1238$ line
284: possibly in P-Cygni. We chose to keep this star in group 3 since the
285: \ion{C}{4} $\lambda 1549$ line is not in P-Cygni in any other
286: spectrum, including the ones in the high resolution mode (e.g. SWP41947).
287: We have also included the star M 1-71 in group 3, although it shows only
288: a featureless UV spectrum. For this object however, there is only one {\it IUE}
289: file available, and it is possible that the central star was not
290: within the {\it IUE} aperture during the observation.
291:
292: One object does not fit in our division scheme, namely, IC 5217.
293: The \ion{C}{4} $\lambda 1549$ line in this star is intense, but not in P-Cygni.
294: On the other side, a P-Cygni is found in \ion{N}{5} $\lambda 1238$, and also in
295: \ion{O}{5} $\lambda 1371$, although somewhat weak. Furthermore, this star does
296: not present a flat continuum. We conclude that the wind spectrum is clearly
297: visible in this star, but it is highly contaminated by nebular emissions.
298:
299: We should emphasize that the three groups above were
300: introduced by analyzing the $\sim 1150-2000$\AA\, interval,
301: and that although we have shown in Figs. \ref{group1} and
302: \ref{group23} low resolution spectra, the characteristics
303: of these groups are confirmed from the available high dispersion data.
304: We chose the $\sim 1150-2000$\AA\, region for two reasons. First, it is in this wavelength range
305: that we have found practically all the P-Cygni lines. Obviously, such
306: profiles are formed in the stellar wind of the central star - which is our
307: main interest - and not in the nebula. Second, the $\sim 2000-3000$\AA\,
308: interval sometimes have no or very few features (see next section), whose
309: nature (nebular or stellar) we cannot decide from low resolution data.
310: The exception is the third group, where we have found various transitions,
311: but as we previously argued, they are most likely nebular.
312:
313: In addition to Figs. \ref{group1} and \ref{group23}, we show in Fig.
314: \ref{wcpg} the {\it IUE} spectra of the two proto-type
315: [WC]-1159 stars: A30 and A78, also at the same resolution.
316: It is clear that these two objects have a spectrum
317: that is remarkably different from the WELS
318: spectra. The three P-Cygni lines \ion{N}{5} $\lambda 1238$,
319: \ion{O}{5} $\lambda 1371$, and \ion{C}{4} $\lambda 1549$ are
320: simultaneously present and more intense than in the WELS.
321: Furthermore, the absorption part of these P-Cygni profiles
322: (mainly in \ion{C}{4} $\lambda 1549$) are considerably broader,
323: suggesting large terminal wind velocities (see Section \ref{vinfsec}).
324: Although the E(B-V) for these stars have a role in
325: determining the slope of the continuum, it is also
326: evident that the flux is higher in the blue
327: ($\sim 1150$\AA) than in the red part of the {\it IUE}
328: spectrum ($\sim 2000$\AA). This is not seen in most of the
329: WELS, and might indicate that [WC]-PG 1159 stars have higher
330: effective temperatures.
331:
332: We have assigned the stars NGC 6567, NGC 6572, and NGC 6543,
333: to the group presenting intense \ion{C}{4} $\lambda 1549$ in
334: P-Cygni. However, in addition to the \ion{C}{4} $\lambda 1549$ line
335: exhibited by the stars in group 1, these three objects also show
336: simultaneously \ion{N}{5} $\lambda 1238$ and \ion{O}{5} $\lambda 1371$ P-Cygni
337: emissions. This characteristic resembles the [WC]-PG 1159 class and is shown
338: in Fig. \ref{wcpg2}. Thus, although a comparison between spectra shown
339: in Fig. \ref{group1}, \ref{group23}, and \ref{wcpg} argue against
340: those WELS being [WC]-PG 1159 stars, we cannot
341: be sure regarding the objects shown in Fig. \ref{wcpg2}.
342: In fact, the central star of NGC 6567 for example, was already
343: considered as a [WC]-PG 1159 object by Hamann (1996).
344: This point will be addressed later in the paper.
345:
346: \subsection{Line Identifications and Measurements}
347:
348: In Tables 2, 3, 4 and 5 we present line identifications, measurements
349: of equivalent widths and line fluxes of all {\it IUE} low resolution spectra
350: considered in this work. For completeness, we have included IC 5217 in Table
351: 4, although this object does not belong to the group of WELS without P-Cygni
352: emissions (group 3). The equivalent widths were measured by the conventional
353: method of adjusting a Gaussian function to the line profile. When this
354: was not possible, we have computed the line area above the adopted
355: continuum. In the case of a P-Cygni profile, only its emission feature
356: was considered in the measurement. We emphasize that our intention is
357: not to construct an UV atlas for the WELS. Therefore, we have only
358: listed and measured the most conspicuous lines in their spectra.
359: It should also be kept in mind that throughout this paper we make
360: reference to lines such as \ion{C}{4} $\lambda 1549$ and
361: \ion{N}{5} $\lambda 1238$, but the exact atomic transitions
362: might be slightly different. In these two cases for example, the carbon and
363: nitrogen lines are actually the doublets: \ion{C}{4} $\lambda \lambda 1548, 51$
364: and \ion{N}{5} $\lambda \lambda 1238, 42$.
365:
366: Previous studies have investigated the UV spectrum of some stars seen in Tables 2-5.
367: In general, we have found a good agreement between our results and the
368: ones that also present line measurements. For Vy 2-1, NGC 6629, M 1-71, and Hen 2-108, we found
369: no reference in the literature regarding their {\it IUE} spectra from the MAST
370: database. Although we can find references for the stars CN 3-1, NGC 6891,
371: J900, Hen 2-12, NGC 5873, and NGC 6803, they generally involve
372: a large sample of objects and are not focused on line identifications,
373: measurements or the WELS evolutionary state. Details concerning
374: some individual objects are discussed below.
375:
376: {\bf NGC 6567:} Although there are four references regarding this object
377: in the MAST database, only the one of Hyung et al. (1993) have focused
378: in some detail the main features of its ultraviolet spectrum.
379: For \ion{C}{4} $\lambda 1549$, these authors found a line flux (in
380: units of $10^{-13}$ ergs s$^{-1}$ cm$^{-2}$) of 12.8, while we
381: have measured 13 $\pm$ 1. Taking into consideration the uncertainty
382: in the line measurements of their work (of $\sim 15\%$), our results
383: present a very reasonable agreement for this line and others such
384: as \ion{N}{5} $\lambda 1238$, \ion{O}{5} $\lambda 1371$,
385: \ion{C}{2} $\lambda 2325$, and \ion{Mg}{2} $\lambda 2798$.
386:
387: {\bf NGC 6572:} There are several references to the UV spectrum of this
388: object. The most important to our discussion is the one of Hyung et al.
389: (1994), which presents a detailed high resolution spectral study from
390: the UV to the optical, with several line identifications and measurements.
391: Their main aim was however to explore diagnostic lines in the rich
392: spectrum of NGC 6572 in order to determine plasma parameters, and
393: not to discuss its evolutionary status.
394:
395: {\bf NGC 6543:} The {\it IUE} spectrum of this star was studied before
396: by Bianchi et al. (1986) and Perinotto et al. (1989). Both works
397: have heavily focused only in the determination of stellar and wind
398: parameters (e.g. $T_{eff}$ and $\dot{M}$). No comparison to other
399: objects as in the context of the present paper was made.
400:
401: {\bf PB 8:} This planetary nebula was studied together with other
402: 7 objects by Feibelman (2000). Line fluxes provided in his work
403: are in good agreement with our measurements. For \ion{N}{5}
404: $\lambda 1238$ for example, we encountered $\sim 2.6$ (units of
405: 10$^{-13}$ ergs cm$^{-2}$ s$^{-1}$) while Feibelman derived
406: 2.4. Although the emission in $\sim 1724$\AA\, was identified
407: in his work as \ion{N}{4} $\lambda 1719$, it is possible that
408: it arises actually from \ion{Si}{4} $\lambda 1722$. In fact, we
409: have also identified two lines in 1394\AA\, and 1403\AA\, which
410: are likely to stem from \ion{Si}{4}.
411:
412: {\bf Hen 2-131:} Its {\it IUE} spectra were briefly discussed
413: by Adams \& Seaton (1982). However, the attention was given to the
414: determination of the C/O ratio in the nebula, and not to the central
415: star.
416:
417: {\bf IC 4776:} This object was studied by Herald \& Bianchi
418: (2004) by means of expanding atmosphere models, and its
419: physical parameters and chemical abundances were obtained.
420: As far as we are concerned, their work constitute the first
421: to analyze the stellar wind of a WELS with a state-of-art
422: non LTE code (CMFGEN; Hillier \& Miller 1998). The same main
423: observed wind emissions were identified in their and our
424: paper. The \ion{O}{5} $\lambda 1371$ is predicted by their model, but
425: it is very weak or absent in the {\it IUE} spectrum. Herald \& Bianchi
426: (2004) have also analyzed the star A78, and a [WC]-PG 1159 classification
427: was confirmed. Surprisingly, the physical parameters obtained for A78
428: and IC 4776 are quite different. A78 is almost twice hotter ($T_{eff} \sim 113kK$)
429: and present a larger terminal velocity (3200 km s$^{-1}$) than IC 4776.
430: These findings support the idea that these two stars do not belong
431: to the same spectroscopy class, i.e., that WELS are not [WC]-PG 1159 stars
432: (see Section \ref{discussion}).
433:
434: {\bf Hb7:} Gauba et al. (2001) also analyzed the low {\it IUE} resolution
435: spectrum of this star, together with another planetary nebula (Sp 3). Some
436: wind parameters (e.g. $\dot{M}$) as well as the effective temperature and
437: core mass were estimated, but no atmosphere models were used. The spectral
438: energy distribution from the far infrared to the ultraviolet was also
439: investigated. No line identifications and measurementes were presented
440: and the status of Hb7 as a WELS was not discussed.
441:
442: {\bf NGC 6818:} The spectrum of this object (nebular plus central star)
443: was extensively studied by Hyung et al. (1999). We refer the reader to
444: their work for a more detailed analyses than the one presented here.
445: The chemical composition of the planetary nebula was further explored
446: by the work of Pottasch et al. (2005).
447:
448: {\bf IC 5217:} A very similar analysis to NGC 6818 (Hyung et al. 1999)
449: was done in the case of IC 5217 by Hyung et al. (2001). Again, we refer the reader
450: to their work for detailed line identifications and measurements. We
451: highlight that in both references the identity of these two stars
452: as WELS is not discussed, and emphasis is given mainly to the
453: determination of plasma parameters such as the electronic
454: temperature, density, ionic concentrations, and chemical
455: abundances.
456:
457: \section{TERMINAL VELOCITIES}
458: \label{vinfsec}
459:
460: In this section, we derive terminal velocities ($v_\infty$) for all
461: WELS presenting P-Cygni profiles in \ion{N}{5} $\lambda 1238$ and/or \ion{C}{4} $\lambda
462: 1549$. The $v_\infty$, together with the mass-loss rate ($\dot{M}$), constitute
463: the most important physical parameters describing the stellar winds of hot
464: stars. Their empirical determination are often used to test the theory of
465: radiatively driven stellar winds, to compare stars of different spectral
466: classes, and also as an input in stellar evolution models and interstellar
467: medium studies (see for example Prinja et al. 1990; Lamers \& Cassinelli
468: 1999). The determination of $\dot{M}$ for the WELS presenting wind features
469: is beyond the scope of the present paper, since it requires sophisticated
470: non LTE expanding atmosphere models for reliable estimates. Work is under
471: way by our group to accomplish this goal.
472:
473: Because only a few high resolution spectra are available, most of the
474: terminal velocities of the WELS can only be obtained from low
475: dispersion {\it IUE} spectra. A priori, an accurate
476: determination of $v_\infty$ from a $\sim 6$\AA\, resolution spectrum
477: is a complicated task, since we generally do not see an extended
478: and saturated absorption in a P-Cygni line. Nevertheless, in order to
479: overcome this difficulty, Prinja (1994) provided a relation between
480: $\Delta \lambda$ (defined below, obtained at low resolution)
481: and $v_\infty$ (obtained at high resolution) by analyzing a large
482: sample of different hot stars (O, B, WR, and also CSPN). The resultant
483: calibration of his work is represented by:
484:
485: \[ v_\infty = a_1 + a_2[\Delta \lambda] + a_3[\Delta \lambda]^2. \]
486:
487: In this equation, $\Delta \lambda$ ($= \lambda _{peak} - \lambda _{min}$)
488: is the wavelength difference between the peak and the absorption minima
489: of a low resolution P-Cygni profile, and $a_1$, $a_2$ and $a_3$ have
490: different values for \ion{C}{4} $\lambda 1549$ and \ion{N}{5} $\lambda 1238$
491: (see Prinja 1994). This calibration is perfectly suitable for our purpose,
492: and obeys the definition of the terminal velocity from the black absorption
493: core ($v_{black}$) of a P-Cygni line (Prinja et al. 1990).
494:
495: In Table 6 we present the terminal velocities derived for the WELS,
496: and also for the [WC]-PG 1159 stars A30 and A78. We used the usual method for
497: WELS with high resolution spectra (using the Doppler effect and $v_{black}$)
498: and the equation described above for low resolution spectra. We estimate
499: the errors in the determination of $v_\infty$ to be $\sim 20\%$.
500:
501: The most important thing to be noted in Table 6 is that A30 and A78
502: have much higher terminal velocities than the WELS. While the majority of
503: WELS have values concentrated in the $\sim 1000-1500$ km s$^{-1}$ interval,
504: these two [WC]-PG 1159 stars have a terminal velocity of $\sim 3000$ km
505: s$^{-1}$ (from \ion{C}{4}). Such high differences cannot be explained
506: by the uncertainties in the measurements. Although different
507: works in the literature have estimated $v_\infty$ for A30 and A78, we
508: have also measured this quantity in their low resolution spectra to follow
509: the same method applied for the other stars (Prinja's calibration). Among
510: the previous studies regarding A30, we highlight the work of Harrington \&
511: Feibelman (1984), who obtained a terminal velocity of $\sim 4000$ km
512: s$^{-1}$. At that time however, the $v_\infty$ determination was based on
513: the largest negative velocity seen in a P-Cygni profile
514: (when the line returns to the continuum). This procedure tends to obtain
515: larger values than the $v_{black}$ method (Prinja et al. 1990). The central
516: star of A78 was recently analyzed by sophisticated non LTE atmosphere models
517: by Herald \& Bianchi (2004). These authors found that a velocity
518: of $\sim 3200$ km s$^{-1}$ could satisfactorily represent $v_\infty$, a value
519: slightly higher than ours.
520:
521: Other informations can be obtained by analyzing Table 6.
522: It is clear for example, that CN 3-1 and Hen 2-131 have
523: lower terminal velocities than most of the WELS. Indeed,
524: as we have shown in Section \ref{uvwels} (Fig. \ref{group23}),
525: their UV spectra are quite different.
526:
527: In order to further illustrate the differences between the terminal
528: velocities of the WELS and the two prototype [WC]-PG 1159 stars A30 and A78,
529: as shown in Table \ref{vinf}, and to compare the results obtained to other hydrogen deficient CSPN, we
530: present in Fig. \ref{hist} the $v_\infty$ distribution for several
531: stars belonging to the spectroscopic classes [WCL], [WCE], [WC]-PG 1159,
532: WELS, and PG 1159. Once again, it can be seen that the terminal
533: velocities measured for the WELS are mainly concentrated between $\sim 1000-1500$ km s$^{-1}$.
534: Moreover, their $v_\infty$ tends to be higher and lower than in the
535: [WCL] and [WCE] class, respectively. It is also clear that the [WC]-PG 1159
536: class (represented by the two prototype stars A30 and A78) and the
537: few PG 1159 stars that show a stellar wind, have the highest terminal
538: velocities among the hydrogen deficient CSPN.
539:
540: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
541:
542: \section{DISCUSSION AND CONCLUSIONS}
543: \label{discussion}
544:
545: From a comparison between several optical spectra of WELS, [WC]-PG 1159,
546: and PG 1159 stars, Parthasarathy et al. (1998) have proposed that
547: the WELS are actually [WC]-PG 1159 stars. This claim was not
548: confirmed by further studies and as we mentioned,
549: some authors have warned about this assertion until a more
550: comprehensive study is achieved (Werner \& Herwig 2006). After an
551: analysis of the main UV characteristics of the WELS and the two
552: prototype [WC]-PG 1159 stars A30 and A78, our next step was
553: to compare the results obtained for these two class of objects in
554: order to address this important issue.
555:
556: As we have shown, most of the WELS present a UV spectrum considerably
557: different from the [WC]-PG 1159 stars. While this last class present simultaneously
558: P-Cygni profiles in \ion{N}{5} $\lambda 1238$, \ion{O}{5} $\lambda 1371$, and
559: \ion{C}{4} $\lambda 1549$, the majority of WELS present a very weak or no
560: \ion{O}{5} $\lambda 1371$ (see Fig. \ref{group1}). The same is true at least for
561: some objects regarding the \ion{N}{5} $\lambda 1238$ line.
562: The only exceptions are the objects NGC 6543, NGC 6567, and NGC 6572.
563: Their spectra in fact resemble the ones of the [WC]-PG 1159 stars
564: (see Figs. \ref{wcpg} and \ref{wcpg2}): their \ion{O}{5} $\lambda 1371$
565: line is clearly visible as well as the other transitions mentioned.
566:
567: Besides the spectral differences found in the ultraviolet part of
568: the spectrum, we have also found that the terminal velocities of the
569: WELS are considerably lower than in [WC]-PG 1159 stars. Our
570: Table 6 shows that the bulk of the WELS have $v_\infty$
571: between $\sim 1000-1500$ km s$^{-1}$, and A30 and A78 have
572: values about $3000$ km s$^{-1}$. This difference might
573: represent different physical parameters underlying these two
574: class of stars. The theory of radiatively driven stellar winds
575: for example, predicts that the terminal velocity of a star is
576: related with the escape velocity ($v_{esc}$), which in turn
577: depend on other physical parameters (e.g. mass and radius; Abbott 1978; Lamers
578: \& Cassinelli 1999). Moreover, $v_\infty$ is also known to correlate with the
579: effective temperature ($T_{eff}$) in several class of stars
580: (see Fig. 8 of Prinja et al. 1990). The $T_{eff}$ tends
581: to be higher for stars with high terminal velocities.
582:
583: From the considerations above described, we conclude that the
584: [WC]-PG 1159 stars are distinct from the WELS, in contrast
585: with the claim made by Parthasarathy et al. (1998) on the
586: basis of optical spectroscopy. It should be noted however,
587: that the situation for the central stars NGC 6543, NGC 6567 and
588: NGC 6572 is ambiguous. From one side, they have a spectrum
589: compatible with the [WC]-PG 1159 class. On the other hand, they do
590: not present high terminal velocities as it is the case of
591: A30 and A78. From both low and high resolution data it is
592: obtained $v_\infty$ values less than $2000$ km s$^{-1}$ for
593: these three stars (see Table 6).
594:
595: If the WELS are not [WC]-PG 1159 stars, what is their
596: role in the evolutionary sequence [WR] $\rightarrow$ PG 1159 ?
597: Do they form an alternative channel of evolution ? In order
598: to elucidate these and other similar questions, and to further
599: clarify the differences between them and the [WC]-PG 1159 stars,
600: we clearly need to determine their physical parameters and
601: chemical abundances. Non LTE expanding atmosphere models are
602: being computed with this purpose by our group with
603: the CMFGEN code of Hillier \& Miller (1998).
604: In this way, their position in the HR diagram could be
605: determined, a more efficient comparison to [WR], [WC]-PG 1159
606: and PG 1159 stars could be made, and their evolutionary status
607: could be better determined.
608:
609: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
610:
611: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
612:
613: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
614:
615: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
616:
617: \begin{acknowledgements}
618:
619: W. M. acknowledges CNPq for financial support (151635-2005-6).
620: H. J. acknowledges CNPq for financial support.
621:
622: \end{acknowledgements}
623:
624: \begin{thebibliography}{}
625:
626: \bibitem[]{} Abbott, D. C., 1978, ApJ, 225, 893
627:
628: \bibitem[]{} Acker, A., \& Neiner, C., 2003, A\&A, 403, 659
629:
630: \bibitem[]{} Adams, S., \& Seaton, M. J., 1982, MNRAS, 200, 7
631:
632: \bibitem[]{} Bianchi, L., Cerrato, S., \& Grewing, M., 1986, A\&A, 169, 227
633:
634: \bibitem[]{} Crowther, P. A., De Marco, O., \& Barlow, M. J., 1998, MNRAS, 296, 367
635:
636: \bibitem[]{} De Marco, O., \& Soker, N., 2002, PASP, 114, 602
637:
638: \bibitem[]{} Feibelman, W. A., 2000, PASP, 112, 861
639:
640: \bibitem[]{} Gauba, G., Parthasarathy, M., Nakada, Y., \& Fuji, T., 2001, A\&A, 373, 572
641:
642: \bibitem[]{} Gesicki, K., Zijlstra, A. A., Acker, A., G\'orny, S. K., Gozdziewski, K., \&
643: Walsh, J. R., 2006, A\&A, 451, 925
644:
645: \bibitem[]{} G\'orny, S. K., Stasi\'nska, G., Escudero, A. V., \& Costa, R. D. D., 2004,
646: A\&A, 427, 231
647:
648: \bibitem[]{} Hamann, W. -R., 1996, Ap\&SS, 238, 31
649:
650: \bibitem[]{} Hamann, W. -R., 1997, in Proc. IAU Symp. 180, Planetary Nebulae, ed.
651: H. J. Habing \& H. J. G. L. M. Lamers (Dordrecht:Kluwer), 91
652:
653: \bibitem[]{} Harrington, J. P., \& Feibelman, W. A., 1984, ApJ, 277, 716
654:
655: \bibitem[]{} Herald, J. E., \& Bianchi, L., 2004, ApJ, 609, 378
656:
657: \bibitem[]{} Herwig, F., 2001, Ap\&SS, 275, 15
658:
659: \bibitem[]{} Hillier, D. J., \& Miller, D. L., 1998, ApJ, 496, 407
660:
661: \bibitem[]{} Hyung, S., Aller, L. H., \& Feibelman, W. A., 1993, PASP, 105, 693
662:
663: \bibitem[]{} Hyung, S., Aller, L. H., \& Feibelman, W. A., 1994, MNRAS, 269, 975
664:
665: \bibitem[]{} Hyung, S., Aller, L. H., \& Feibelman, W. A., 1999, ApJ, 514, 878
666:
667: \bibitem[]{} Hyung, S., Aller, L. H., Feibelman, W. A., \& Lee, W-B., 2001, AJ, 122, 954
668:
669: \bibitem[]{} Koesterke, L., Dreizler, S., \& Rauch, T., 1998, A\&A, 330, 1041
670:
671: \bibitem[]{} Koesterke, L., \& Werner, K., 1998, ApJL, 500, 55
672:
673: \bibitem[]{} Koesterke, L., 2001, Ap\&SS, 275, 41
674:
675: \bibitem[]{} Lamers, Henny J. G. L. M., Cassinelli, J. P., 1999, Introduction to Stellar Winds,
676: Cambridge University Press
677:
678: \bibitem[]{} Marcolino, W. L. F., \& de Ara\'ujo, F. X., 2003, AJ, 126, 887
679:
680: \bibitem[]{} Marcolino, W. L. F., Hillier, D. J., de Ara\'ujo, F. X., \& Pereira, C. B., 2007,
681: ApJ, 654, 1068
682:
683: \bibitem[]{} Parthasarathy, M., Acker, A., \& Stenholm, B., 1998, A\&A, 329, L9
684:
685: \bibitem[]{} Pen\~a, M., Stasi\'nska, G., \& Medina, S., 2001, A\&A, 367, 983
686:
687: \bibitem[]{} Pen\~a, M., Medina, S., \& Stasi\'nska, G., 2003, RevMexAA, 18, 84
688:
689: \bibitem[]{} Perinotto, M., Cerruti-Sola, M., \& Lamers, H. J. G. L. M., 1989, ApJ, 337, 382
690:
691: \bibitem[]{} Pottasch, S. R., Beintema, D. A., \& Feibelman, W. A., 2005, A\&A, 436, 953
692:
693: \bibitem[]{} Prinja, R. K., Barlow, M. J., \& Howarth, I. A., 1990, ApJ, 361, 607
694:
695: \bibitem[]{} Prinja, R. K., 1994, A\&A, 289, 221
696:
697: \bibitem[]{} Tylenda, R., Acker, A., \& Stenholm, B., 1993, A\&AS, 102, 595
698:
699: \bibitem[]{} Werner, K., Dreizler, S., Heber, U., Kappelmann, N., Kruk, J., Rauch, T., \& Wolff, B.,
700: 1997, RvMA, 10, 219
701:
702: \bibitem[]{1295} Werner, K., \& Herwig, F., 2006, PASP, 118, 183
703:
704: \bibitem[]{1297} Zijlstra, A. A., van Hoof, P. A. M., Chapman, J. M., \& Loup, C., 1994, A\&A, 290, 228
705:
706: \end{thebibliography}{}
707:
708: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
709:
710: \clearpage
711: \input{tab1.tex}
712:
713: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
714:
715: \clearpage
716: \input{tab2.tex}
717:
718: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
719:
720: \clearpage
721: \input{tab3.tex}
722:
723: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
724:
725: \clearpage
726: \input{tab4.tex}
727:
728: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
729:
730: \clearpage
731: \input{tab5.tex}
732:
733: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
734:
735: \clearpage
736: \input{tab6.tex}
737:
738: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
739:
740: \clearpage
741:
742: \begin{figure}[!ht]
743: \centering
744: %\includegraphics[width=\textwidth]{f1.eps}
745: \epsscale{.9}
746: \plotone{f1.eps}
747: \caption{WELS presenting strong P-Cygni profiles (group 1) in \ion{C}{4}
748: $\lambda 1549$. The \ion{N}{5} $\lambda 1238$ is often present and also in
749: P-Cygni. Low resolution {\it IUE} spectra ($\sim 6$\AA).}
750: \label{group1}
751: \end{figure}
752:
753: \begin{figure}[!ht]
754: \centering
755: %\includegraphics[width=\textwidth]{f2.eps}
756: \plotone{f2.eps}
757: \epsscale{1}
758: \caption{WELS presenting weak P-Cygni profiles in \ion{C}{4} $\lambda 1549$
759: (group 2): top and 2nd panels. WELS presenting absence of P-Cygnis
760: and very weak continuum (group 3): 3rd to 5th panels. Low resolution {\it IUE} spectra ($\sim 6$\AA).}
761: \label{group23}
762: \end{figure}
763:
764:
765: \begin{figure}[!ht]
766: \centering
767: \includegraphics[width=18cm, height=10cm]{f3.eps}
768: \caption{Low resolution {\it IUE} spectra of the two prototype [WC]-PG 1159 stars - A 30 (top) and A 78 (bottom).}
769: \label{wcpg}
770: \end{figure}
771:
772: \begin{figure}[!ht]
773: \centering
774: \includegraphics[width=17cm, height=13cm]{f4.eps}
775: \caption{Low resolution {\it IUE} spectra of WELS resembling [WC]-PG 1159
776: stars: NGC 6543, NGC 6567, and NGC 6572.}
777: \label{wcpg2}
778: \end{figure}
779:
780: \begin{figure}[!ht]
781: \centering
782: \includegraphics[width=15cm, height=15cm]{f5.eps}
783: \caption{Distribution of terminal velocities of [WCL], [WCE], WELS,
784: [WC]-PG 1159, and PG 1159 stars. The data for the WELS and
785: [WC]-PG 1159 stars were homogeneously obtained from low resolution
786: {\it IUE} spectra using \ion{C}{4} $\lambda 1549$ and the calibration
787: of Prinja (1994). The terminal velocities for the [WCL], [WCE] and
788: PG 1159 stars are from Koesterke (2001), and were obtained by means of non
789: LTE expanding atmosphere models. The bin width is 500 km s$^{-1}$.}
790: \label{hist}
791: \end{figure}
792:
793: \end{document}
794: