astro-ph0409014/t.tex
1: 
2: 
3: %\documentclass[referee]{aa} % for a referee version
4: \documentclass{aa}
5: 
6: \usepackage{epsfig}			
7: %\usepackage{graphicx}           
8: 
9: \usepackage{txfonts}             
10: 
11: \begin{document}
12: 
13:    \title{An optical absorption study of the Helix  Planetary 
14:    Nebula}
15:   \subtitle{Na\,{\sc i} and Ca\,{\sc ii} lines, and a search for diffuse bands}
16:    \author{N. Mauron\inst{1} \and T.R. Kendall\inst{2} }
17:    \offprints{N.Mauron}
18: 
19:    \institute{ Groupe d'Astrophysique, UMR 5024 CNRS, Case CC72, Place Bataillon, 
20: 	     F-34095 Montpellier Cedex 05, France\\
21:              \email{mauron@graal.univ-montp2.fr}
22:       \and 
23:      Laboratoire d'Astrophysique, Observatoire de Grenoble, 
24:    Universit\'{e} Joseph Fourier, BP 53, F-38041 Grenoble Cedex 09, France
25:    \email{tkendall@obs.ujf-grenoble.fr}
26: \thanks{Based on observations carried out with the European 
27: 	     Southern Observatory VLT/UVES, program 70.C-0100}}
28: 
29:    \date{Received xxx; accepted xxx}
30: 
31: 
32:    \abstract {We present the first results of an optical absorption study 
33:  of NGC\,7293, the Helix planetary nebula (PN), designed to search definitively  
34:  for diffuse band (DIB) absorptions perhaps arising in the carbon-rich 
35:  circumstellar matter of the PN. We used the VLT-UVES spectrograph to achieve 
36:  high resolution (R=50000) spectra of 8 stars located angularly close to and 
37:  behind the Helix. These targets were selected through their photometric UBV 
38:  or 2MASS properties, permitting derivation of spectral types (mainly F-G dwarfs) 
39:  and distance that place several of them far (700--1500\,pc) beyond the Helix 
40:  (210\,pc). Through a detailed analysis of the Na\,{\sc i} and Ca\,{\sc ii} 
41:  lines to the 8 targets, we find that  two lines of sight situated close to 
42:  the nebula as mapped in CO and H\,{\sc i} exhibit very strong Na\,{\sc i} 
43:  absorption. This absorption is unlikely to arise in the foreground or background 
44:  interstellar medium which has a relatively  low column density, because the 
45:  Helix is at high galactic latitude $b = -57^{\circ}$. It is much more probable
46:  that it is due to the PN neutral or molecular material. This circumstellar origin 
47:  is reinforced by the fact that no corresponding Ca\,{\sc ii} line is observed 
48:  (as it would usually be from the ISM), which is in agreement with the very high 
49:  Ca depletion often observed  in PNs. No trace of any  DIB 
50:  features was found in these two circumstellar lines of sight,  nor to the other 
51:  targets. The two circumstellar sightlines discovered in this work open the way to search 
52: for molecular species such as C$_2$ which are not observable at radio wavelengths, 
53: and to obtain more information on the rich neutral and molecular content of the Helix.
54:              \keywords{Stars; individual: NGC\,7293 -- stars: planetary nebulae -- 
55: 	     stars: circumstellar matter -- ISM: lines and bands -- 
56:               ISM: molecules}
57:    }
58:    \titlerunning{Optical absorption line spectroscopy of the Helix PN}
59: 
60:    \maketitle
61: %
62: %________________________________________________________________
63: 
64: \section{Introduction}
65: 
66:  The identification of the carriers of the diffuse interstellar bands (DIBs)
67:  has proved one of the most difficult problems in observational molecular 
68:  astrophysics (see Herbig \cite{her95}, hereafter H95, for a review).  
69:  DIBs have been widely 
70:  observed throughout the interstellar medium (ISM) for nearly a century, but 
71:  no convincing identification of a carrier has yet been forthcoming. In addition, 
72:  the origin of the carrier(s)  remains unknown.  Numerous studies have suggested 
73:  that DIB carriers are not present in the circumstellar matter of evolved cool 
74:  O-rich or C-rich stars that inject various dust or molecular compounds into the 
75:  ISM (Snow \cite{sno73}; Snow \& Wallerstein \cite{snw72}; Zacs et al. \cite{zac03}). 
76:  However, there are a few exceptions that would certainly  deserve further 
77:  attention, e.g. the  unusual 6177\,\AA~ band seen in some evolved objects by 
78:  Le Bertre \& Lequeux (\cite{leb93}), or the case 
79:  of DIBs possibly observed against the A-type companion of the carbon star CS\,776 
80:  (Le Bertre \cite{leb90}; for this object, see also comments in H95 and in Kendall et al. 
81:  \cite{ken02}).
82:  
83: The initial findings of Scarrott et al. (\cite{sca92}) and Sarre et al. (\cite{sar95}) 
84: connecting the optical emission 
85: bands in the Red Rectangle (RR) with a subset of well-known DIBs have been qualified 
86: recently by Van Winckel et al. (\cite{vwi02}) and Glinski \& Anderson (\cite{gli02}) on 
87: the basis of e.g. a 2\,\AA~wavelength 
88: mismatch between the $\lambda$5797 DIB in the ISM and the nearby $\sim$5800\,\AA\,\,
89: RR\,\,emission band. This mismatch remains at the largest observed distances from the central 
90: star, leading both authors to conclude that the  carriers of the DIB and the RR band 
91: are not identical. However, Van Winckel et al. (\cite{vwi02}) also note five  further possible 
92: matches between DIB wavelengths and RR features, and there certainly exist in 
93: the RR molecules closely related to the likely DIB carriers.
94: 
95: A hitherto undisputed fact is that a subset of diffuse bands are also clearly seen 
96: in emission in the circumstellar layers of the unusual, carbon rich, R\,CrB-type 
97: variable V854\,Cen (Rao \& Lambert \cite{rao93}). DIB carriers perhaps exist also 
98: in the envelopes of 
99: rare, hotter C-rich post-AGB objects (Zacs et al. \cite{zac99}; Klochkova et al.
100:  \cite{klo99}; Klochkova et al. \cite{klo00}; Klochkova et al.  
101: \cite{klo01}), although this is not certain.  A clear-cut case concerning 
102: a post-AGB object is HR\,4049,
103: which has a carbon-rich envelope; no diffuse band of circumstellar origin 
104: could be found in the spectrum of the central hot star (Waters et al. \cite{wat89}). 
105: Because it is often thought that large carbon bearing molecules might be the 
106: (or one of the) DIB carriers (e.g. H95, Duley \cite{dul98}), every effort is desirable to 
107: test  this hypothesis  by seeking DIBs at carbon-rich sites of origin
108: where these molecules may be abundant.
109: 
110: In a recent paper (Kendall et al. \cite{ken02}), we have pioneered a  new 
111: observational method using optical absorption spectroscopy of background objects to 
112: probe the spatially extended circumstellar envelope (CSE) of the archetypal 
113: extreme mass-losing carbon star IRC\,+10$^{\circ}$\,216. By searching for absorptions 
114: against the relatively smooth continua of thick-disk background targets, the method 
115: alleviated confusion with photospheric features in the highly complex spectrum of 
116: the central, cool, evolved object. These observations showed that  DIB carriers are 
117: of very low abundance, or non-existent, around IRC\,+10$^{\circ}$\,216, and we 
118: suggested that the DIB carriers might be sought in compact, UV-rich regions where ionised 
119: and neutral molecular material interact. Such sites are the circumstellar shells of 
120: planetary nebulae (PN), and this paper reports our first attempts to detect DIBs in 
121: the Helix PN. 
122: 
123: In fact, DIB carriers have already been sought in a few other carbon-rich PN, most 
124: notably NGC\,7027~(Le Bertre \& Lequeux \cite{leb92}) and no circumstellar diffuse 
125: bands  have been detected.  However, it is justified to apply our method of background 
126: target spectroscopy to PN and especially to use several lines of sight.
127:  This is an important point, since the clear clumpiness and inhomogeneity
128: of typical PN shells means that DIB carriers, and other species,  may not be distributed 
129: uniformly throughout such shells. Indeed, in cases where only
130: the central star is observed  through relatively low density material, the column density 
131: may be insufficient  for observable absorptions to arise. 
132: 
133: Therefore,  in this paper, we report  the first results of a
134: similar absorption spectroscopy  experiment performed using 8 targets seen through, or
135: located angularly very close to, the Helix. As explained below in more detail (Sect. 2), this  
136: PN was selected primarily because of its proximity, its high galactic 
137: latitude, and its wealth in ionized, neutral and molecular content.
138: 
139: %%%==================================== TABLE OF TARGETS ===================================
140: \begin{table*}
141:  \caption[]{Background targets beyond the Helix Nebula, with basic and derived data. 
142:   For clarity, we will use the identification given in the second column when referring 
143:   to individual targets.  
144: The column t$_{\rm exp}$ gives for each target
145: the total exposure time; for objects 1, 3, 4, and 8, two spectra of 
146: half this time were exposed. $d$ is the estimated distance in pc.
147: The final column states whether  a target is directly behind the nebula (on-PN) 
148: or a control object (off-PN). Targets 7 and 8 are   HD\,213056 and HD\,213069, 
149: respectively, with spectral types  from Simbad.}
150: \begin{center}
151: 
152:         \begin{tabular}{crccrrrrrcr}
153:         \noalign{\smallskip}
154:         \hline
155:         \hline
156:         \noalign{\smallskip}
157: IAU  &Id. & $\alpha$\,\,  $\delta$ & $V$ & $B-V$ & $V-J$  & t$_{\rm exp}$ &Sp.& $d$ &
158:  Notes \\
159: name & & (J2000)             & (mag)&(mag)&(mag)    & (min)          &      & (pc)   &\\
160:  
161: \noalign{\smallskip}
162: \hline
163: \noalign{\smallskip}
164: USNO-B1.0 0691-0897349& 1 & 22 29 47.3 $-$20 50 31 & 13.17 & 0.51  & 1.02 & 66.2 & F8V & 680 &  on PN\\
165: USNO-B1.0 0691-0897340& 2 & 22 29 38.6 $-$20 50 14 & 13.54 &$-$0.39& -\,\,& 46.1 &-\,\,& 210 & nucleus\\
166: USNO-B1.0 0692-0882100& 3 & 22 29 33.5 $-$20 46 18 & 12.04 & 0.64  & 1.08 & 22.7 & G2V & 290 & on-PN\\
167: USNO-B1.0 0691-0897309& 4 & 22 29 19.0 $-$20 48 48 & 13.94 & -\,\, & 0.75 & 66.7 & F2V & 1500& on-PN\\
168: USNO-B1.0 0690-0897407& 5 & 22 29 46.5 $-$20 56 33 & 13.57 & -\,\, & 0.68 & 47.4 & F0V & 1500& on-PN\\
169: USNO-B1.0 0691-0897499& 6 & 22 30 28.0 $-$20 48 31 & 12.31 & 0.59  & 1.16 & 14.6 & G1V & 380 & off-PN\\
170: USNO-B1.0 0692-0882044& 7 & 22 29 09.4 $-$20 46 07 & 10.02 & 0.89  & 1.88 & 01.7 &K1III& 760 & on-PN\\
171: USNO-B1.0 0690-0897288& 8 & 22 29 13.1 $-$20 57 49 & 10.36 & 0.73  & -\,\,& 05.2 & F5V & 240 & off-PN\\
172: \noalign{\smallskip}
173: \hline
174: \end{tabular}
175: \end{center}
176: \end{table*}
177: %%%%%%%%%%%%%%%%%++=======================================================================================
178: 
179: 
180: %=========================== figure showing Helix; graph bidon pour ArXiv
181:    \begin{figure}
182:    \centering
183:    \resizebox{\hsize}{!}{
184:   {\rotatebox{-90} {\includegraphics{za-helix-bidon.ps}}   }
185:   } 
186:   \caption{Observed targets labelled on the $R$-band POSS-2 image. 
187:         The field  is 25$^{\prime}$ $\times$ 25$^{\prime}$, centered on
188:        the nucleus (\#2). North at top, East to the left.}
189:    \label{f1}
190:    \end{figure}
191: %%%%%%%%%%%%%%%%%%%%%====================================================================
192: 
193: \section{The Helix Nebula and the background stars used for absorption spectroscopy}
194: 
195: The Helix Nebula is undoubtedly the best target of its kind for this study, being 
196: close ($d$\,=\,$210\pm30$\,pc, from parallax $4.70\pm0.75$\,mas; Young et al. \cite{you99}), 
197: as well as rich in carbon, dust and being irradiated by a strong central UV source. 
198: Its high galactic latitude ($b=-57^{\circ}$)
199: minimises confusion with the foreground or background ISM: the maps of 
200: Burnstein \& Heiles (\cite{bur82}) and Schlegel et al.  (\cite{sch98})
201:  show a very low interstellar (IS) reddening of
202: $E_{\rm B-V}$\,$\sim$\,0.03\,mag. The nebula has been detected in the lines of 
203: H\,{\sc i}, H$_2$ and CO, together with other species (CN, HCN, HNC, HCO$^+$
204: and C\,{\sc i}) which strongly indicate a carbon-rich medium (Young et al. \cite{you99}; 
205: Speck et al. \cite{spe02}). 
206: While the H\,{\sc i} and CO\,(2--1) maps (Rodriguez et al. \cite{rod02}; 
207: Young et al. \cite{you99}, respectively) 
208: roughly outline the main body of the optical PN, i.e. as traced by ionised species,
209: these maps reveal strong inhomogeneity in the form of a marked ``knotty'' structure. 
210: This is fully consistent with the presence of the well known numerous 
211: dust globules seen in optical images. Also, the radio maps show that the global PN 
212: geometry is not spherical, but approximately disk-like and seen nearly pole-on. This 
213: complex structure makes obvious the necessity of considering as many  background 
214: stars as possible. Finally, because the neutral phase of the PN   contains $\sim$\,3 
215: to 6 times more mass than the molecular phase (Rodriguez et al. \cite{rod02};
216:  Young et al. \cite{you99}), it can 
217: be expected that essentially the neutral material, but perhaps also some molecular 
218: knots, may be probed by observing  background stars.
219: 
220: Optical images (Fig.\,\ref{f1}) show a large  number of potential targets within or 
221: near the PN shell which might function as sightlines to probe the nebula and the ISM 
222: in angularly nearby sightlines. Targets were first selected using homogeneous, accurate
223: $JHK$ photometry from 2MASS. 
224: For some of them, a $V$ magnitude is available in Simbad.
225: By comparing the $J$$-$$K$ and/or $V$$-$$J$ colours of our targets to 
226: template Population\,I dwarfs, an approximate spectral type could be inferred. 
227: Then, absolute magnitudes and therefore distances were derived. Where independent 
228: $U$$-$$B$, $B$$-$$V$ colours are available in Simbad, we find excellent agreement with 
229: the predictions of the $J$$-$$K$ or $V$$-$$J$ colours. 
230: Only about half of the targets which were 
231: bright enough to obtain high signal-to-noise ($S/N$) spectra could eventually be observed, 
232: owing to poor weather.
233:  
234: The derived types and distances for these 8 observed targets are given in Table\,1. 
235: The targets  have $V$\,$\sim$\,10--14 and are located directly behind, or 
236: close to, the shell.  With the exception of the nucleus, target \#2, 
237: the sightlines probed have spectral types F or G. The  uncertainties on distances are
238: of the order of 20\%. Distances would be 30\% smaller (or less) if we had adopted 
239: the luminosity of low-metallicity subdwarfs, but they would be considerably larger
240: if we had adopted the luminosity of giants. In the end, several targets are clearly much more 
241: distant ($>$\,400\,pc) than the Helix. As we shall see below, all  these targets  provide
242: a very usable continuum for absorption spectroscopy 
243: despite  their photospheric lines.  Moreover, for each target,
244:  the precise velocity of the photospheric Na\,{\sc i} D lines could be easily predicted
245: through consideration of  
246: $\sim$20 prominent unblended  Fe\,{\sc i} lines available in our spectra 
247: (vertical bars in Fig.\,\ref{f2}).
248: 
249: \section{Observations and data reduction}
250: 
251: Observations were performed on 2002 October 11--12 using the Very Large 
252: Telescope with the Ultraviolet and Visual Echelle 
253: Spectrograph (UVES) at the European Southern Observatory, Cerro Paranal, Chile. 
254: Total integration times of up to $\sim$\,one hour were achieved, depending on the 
255: target magnitude. Using the 0.8\arcsec~slit (R\,=\,50000) with both red 
256: and blue dichroics, spectra were obtained over three wavelength ranges, 
257: $\sim$\,3500--4520\,\AA, 4630--5600\,\AA~ and 5670--6650\,\AA.
258: These ranges include in particular the strongest DIBs at 5780, 5797, 6283 
259: and 6614\,\AA,  the Na\,{\sc i} D and Ca\,{\sc ii} H \& K lines. 
260: The target signal-to-noise ratio of $\sim$\,70 (per 0.017\AA\,bin) was achieved 
261: over the most part of these ranges, except in the blue region
262: comprising the Ca\,{\sc ii} lines where $S/N$ is generally poor because
263: the target is not hot enough (with the exception of the nucleus).
264: The data analysis was carried out by using the
265: reduced spectra  that are provided by the ESO/UVES reduction pipeline, and by
266: using the ESO MIDAS and the NOAO IRAF packages.
267:  
268: %%%%%%%%%%%%%%%%%%%  TABLE ON 3 FIELD HOT STARS  =================================
269: \begin{table*}
270:  \caption[]{Literature data for 3 bright stars located at 2.8--5.4\degr\, 
271:             from the Helix. $\theta$ is the angular distance of each star
272: 	    to the Helix nucleus. W$_{\lambda}$(D1) \& W$_{\lambda}$(D2) are the observed 
273: 	    equivalent widths in the IS Na\,{\sc i} D1 5895\,\AA\, and D2 5890\,\AA\, 
274: 	    lines, and W$_{\lambda}$(K) is that for the Ca\,{\sc ii} K 3933\,\AA\, line, all 
275: 	    expressed in m\AA. Relative uncertainties on 
276: 	    W$_{\lambda}$ values are $\sim$ a few percent.}
277:  \begin{center}
278:  \begin{tabular}{lcrlrrrrl}
279:  \noalign{\smallskip}
280:  \hline
281:  \hline
282:  \noalign{\smallskip}
283: 
284:  Star      & $\theta$ &  V\,\,\,\,\,   & Sp.&dist. &W$_{\lambda}$(D1)& W$_{\lambda}$(D2)&
285:  W$_{\lambda}$(K) & Notes\\
286:  name          &  (deg)   & (mag)&       & (pc) &(mA)  & (mA)  & (mA)     & \\ 
287: \noalign{\smallskip}	     
288:  \hline
289:  \noalign{\smallskip}
290:  PHL\,346   & 2.86     &11.47 & B1    &10000 &  -\,\,   &    -\,\,  &  199     & (1)(4)\\
291:  
292:  HD\,214080 & 4.70     & 6.83 & B1Ib  & 3000 & 171      &   207     &  149     &  (2)\\
293:  
294:  HD\,210191 & 5.38     & 5.81 & B2.5IV&  460 &  90      &  164      &  285     & (2)(3)\\
295: \noalign{\smallskip}
296: \hline
297: \end{tabular}
298: \end{center}
299: 
300: Notes: (1) Data from Keenan et al. (\cite{kee88});    (2) Data from Albert (\cite{alb83});
301:  (3) Also known as 35 Aqr; (4)  for PHL\,346, no Na\,{\sc i} data could be found 
302:  in the literature.
303:  \end{table*}
304: %%%====================================================================================
305: 
306: \section{The interstellar medium in the direction of the Helix Nebula}
307: 
308:   Before interpreting the spectra of our program stars, e.g. 
309:  the Na\,{\sc i} and Ca\,{\sc ii} profiles, 
310:  one has to estimate the effect of the  foreground and background 
311:  ISM, deriving, at least approximately, the expected strength of the IS lines.
312:  One  source of information is the $E_{\rm B-V}$ maps derived 
313:  by Schlegel et al. (\cite{sch98}).  By considering a zone of
314:  3\degr$\times$\,3\degr~centred on  the Helix (the nucleus is at  $l$\,=\,36\fdg2,
315:  $b$\,=\,$-57$\fdg1)   but excluding a central region of 
316:  1\degr$\times$\,1\degr~to avoid the nebula itself,
317:  we find that $E_{\rm B-V}$ would be on average $\approx$ 0.035 mag. 
318:  with $\sigma$\,=\,0.005, as derived  from  20 independent positions.
319:  One has to note that the $\sigma$ above only represents the local 
320:  scatter of $E_{\rm B-V}$ as given by the Schlegel et al. maps which have an
321:  angular resolution of 6\arcmin. This does not exclude  
322:  moderately larger deviations from this mean when one considers the very narrow 
323:  pencil-like sightlines to  background stars. 
324:  
325:  Based on this estimate of $E_{\rm B-V}$, one can derive 
326:  the expected column density of Na\,{\sc i}. From Bohlin et al. (\cite{boh78}), 
327:  the hydrogen (H\,{\sc i} + 2H$_2$) column density is obtainable with  
328:  $N$(H)/$E_{\rm B-V}$ = 5.8\,$\times$\,10$^{+21}$ cm$^{-2}$, with
329:   deviations from this mean generally less than a factor 1.5. Consequently, 
330:   we obtain $N$(H)\,=\,2.0\,$\times$\,10$^{+20}$ cm$^{-2}$. For this column 
331:  density, the correlation found by Ferlet et al. (\cite{fer85}) with $N$(Na\,{\sc i})
332:  suggest that $N$(Na\,{\sc i})\,=\,1.0\,$\times$\,10$^{+12}$  cm$^{-2}$ with  
333:  a scatter of about a factor 3 (see their Fig.\,1). This Na\,{\sc i} column density 
334:  is relatively small but is already in the non linear regime of the curve of 
335:  growth for typical IS sightlines.  It corresponds to a typical 
336:  equivalent width (W$_{\lambda}$) in the Na\,{\sc i} D2 (5890\,\AA) line 
337:  W$_{\lambda}$(D2)\,$\sim$\,200\,m\AA. Consequently, we expect that 
338:  W$_{\lambda}$(D2) observed towards our targets  be close to 200\,m\AA.
339: 
340:  
341:  Another source of information  concerning the ISM in the direction of the Helix 
342:  is obtained by considering available observations of Na\,{\sc i} or other  
343:  atoms for stars angularly close to the Helix. By interrogating the catalogue of 
344:  Garcia (\cite{gar91}), we find 3 stars with IS information available 
345:  and located  at less than 6\degr~from the Helix centre. The data, taken 
346:  from Albert (\cite{alb83}) and Keenan et al. (\cite{kee88}), are listed in Table 2, and
347:  the values of W$_{\lambda}$(D2) are found to be also  $\sim$\,200\,m\AA\, in
348:  excellent agreement with estimations made above. It is interesting also to note 
349:  that Albert (\cite{alb83}) gives for HD\,214080 $E_{\rm B-V}$\,=\,0.05  in very 
350:  good agreement with $E_{\rm B-V}$\,=\,0.047 provided by the Schegel et al. maps 
351:  in the direction of this star. In contrast, for HD\,210191, Albert (\cite{alb83}) gives 
352:  $E_{\rm B-V}$\,=\,0.07 whereas the maps of Schlegel et al. give 
353:  $E_{\rm B-V}$\,=\,0.025. It is possible that Albert (\cite{alb83}) overestimates the
354: extinction because the intrinsic colour $(B-V)_{0}$ of HD\,210191 has 
355: to be derived  from its spectral classification and may have been 
356: incorrectly taken as too blue. 
357: 
358: An equally important point to consider is the expected range of heliocentric 
359: velocities for IS lines. By considering Fig.\,2 of Albert (\cite{alb83}), it is clear that
360: the H\,{\sc i} (21\,cm) line, the Na\,{\sc i} and the Ca\,{\sc ii} lines toward 
361: HD\,214080 and HD\,210191 are all consistently 
362: found within the velocity interval $v_{\rm  LSR}$\,=\,$-$25 km\,s$^{-1}$ 
363: to +20\,km\,s$^{-1}$,
364: which corresponds to $v_{\rm helio}$\,=\,$-$28 to +17\,km\,s$^{-1}$
365: (for the direction of the Helix $v_{\rm helio}$\,=\,$v_{\rm  LSR}$\,--3, in km\,s$^{-1}$).
366: Concerning PHL\,346, Keenan et al. (\cite{kee88}) find the Ca\,{\sc ii} K IS line 
367: at $v_{\rm LSR}$\,=\,+21\,km\,s$^{-1}$, or $v_{\rm helio}$\,=\,+18\,km\,s$^{-1}$, 
368: in good agreement with the other two bright stars. 
369: 
370: To conclude, the above considerations  suggest that IS lines in our target 
371: sightlines should display typical W$_{\lambda}$ of 200\,m\AA~ for Na\,{\sc i} D2, and
372: about the same for the Ca\,{\sc ii} K line (see Table 2). We expect
373: the IS lines to have $v_{\rm helio}$ within the interval 
374: from $-$28 to +17\,km\,s$^{-1}$, although a velocity slightly outside of this expected
375: range is not excluded, because it is based on 3 sightlines which
376: are not exactly toward the Helix, but angularly shifted by a few degrees.
377: 
378: %%%%%%%%%%%%%%%%%%%%%  figure 2  showing NAI and Ca\,{\sc ii} profiles ===========
379: \begin{figure*}
380: 
381: \resizebox{\hsize}{!}{
382: {\rotatebox{-90}{\includegraphics{zzp000.ps}}}
383: {\rotatebox{-90}{\includegraphics{zzp013.ps}}} }
384: 
385: \resizebox{\hsize}{!}{
386: {\rotatebox{-90}{\includegraphics{zzp046.ps}}}
387: {\rotatebox{-90}{\includegraphics{zzp213.ps}}} }
388: 
389: \resizebox{\hsize}{!}{
390: {\rotatebox{-90}{\includegraphics{zzp166.ps}}}
391: {\rotatebox{-90}{\includegraphics{zzp085.ps}}} }
392: 
393: \resizebox{\hsize}{!}{
394: {\rotatebox{-90}{\includegraphics{zzp185.ps}}}
395: {\rotatebox{-90}{\includegraphics{zzp091.ps}}} }
396: 
397: \caption[]{Na\,{\sc i} D1 (5896\AA) and D2 (5890\AA) profiles (upper plots) 
398: and Ca\,{\sc ii} H (3968\AA) and K (3933\AA) profiles for all eight targets. 
399: For Na, D2 is the solid line and D1 the dotted. For Ca, the solid line 
400: is K, the dotted H. The velocity of photospheric Na \& Ca lines is 
401: shown by a vertical bar. In the left-hand four panels, we show targets \#1, 
402: \#3, \#6 and \#7 for which we suspect there is no circumstellar Na D component. 
403: The same is true of the nucleus (\#2, upper right) where we see 
404: a pure IS spectrum because of the lack of photospheric contamination. 
405: For \#8, the photosphere is well separated, at $\sim$\,+76\,km\,s$^{-1}$. 
406: For \#4, the component at $-$25\,km\,s$^{-1}$ is most likely nebular, as 
407: is most of the Na\,{\sc i} line for \#5 (see text).} 
408: \label{f2}
409: \end{figure*}
410: %%%%==============================================================================
411: 
412: \section{Results and analysis}
413: 
414: \subsection{Characterising the nebular absorption using Na\,{\sc i} D profiles}
415: 
416: Figure\,\ref{f2}
417: displays the Na\,{\sc i} D1 and D2 line profiles for all eight targets.
418: It is important to realize first that, for a given target, any detail seen in 
419: {\it both} D1 and D2 profiles should be the signature of the same IS (or possibly
420: circumstellar) component, and one can indeed see in these spectra many
421: components with corresponding fully consistent D1 and D2 features. 
422: This compensates for one of the difficulties of our experiment:
423: one has  to be  careful about  the fact that the stellar continua of our targets 
424: are not flat, as they would be for OB type stars. Consequently,
425: depending on the radial velocity of the targets, the Na\,{\sc i} 
426: IS (or circumstellar) lines may
427: be blended with photospheric Na\,{\sc i} lines in the target stars themselves, 
428: or, possibly, with a different photospheric feature. 
429: Finally, since Na\,{\sc i} has now been detected in a number of PNs 
430: (Dinerstein \& Sneden \cite{din88}; Dinerstein et al. \cite{din95}), 
431: one can expect some circumstellar contribution, that could well be 
432: significant. 
433: 
434:  In Fig.\,\ref{f2}, the approximate range of velocities for IS components 
435: (discussed above) is indicated by a horizontal dashed line labelled ``IS''.
436:  Concerning the velocities of possible circumstellar components, we have
437:  plotted with a horizontal solid line labelled ``PN'' the $v_{\rm helio}$ range 
438:  of the Helix gas: this range is  from   $-$55 to +2\,km\,s$^{-1}$, as derived 
439:  from CO and H\,{\sc i} line profiles (integrated over the PN). 
440:  In principle, in order to predict at which 
441:  velocity one could expect a circumstellar component, one might consider the 
442:  CO or H\,{\sc i} maps and look at the profiles at the exact position of each 
443:  target. In practice, this provides little information: none of the 8 targets are
444:  covered by the H\,{\sc i} map of Rodriguez et al. (\cite{rod02}; their Fig.~4), 
445:  and CO is detected only to  target \#3, after a careful examination of  Fig.~1 of 
446:  Young et al. (\cite{you99});
447:  the case of target \#3 is discussed below. Therefore, we adopt the view that
448:  circumstellar Na\,{\sc i} components may  occur anywhere within the CO or H\,{\sc i} 
449:  velocity range.
450: 
451:  \subsubsection{Targets 1, 3, 6 and 7 (Fig.\ref{f2}, left panels)}
452: 
453: The case of blending with photospheric Na is obviously present
454: for targets \#1, \#3, \#6 and \#7. In the light of the
455: above comments, these four targets
456: show nothing exceptional which is not explained by photospheric
457: {\it plus} IS contributions. 
458: We note one component at $v_{\rm helio} \approx +25$\,km\,s$^{-1}$ 
459: for  \#1 and \#6. This is +7\,km\,s$^{-1}$ outside 
460:  the expected velocity range of IS lines, and not near the $v_{\rm helio}$ 
461:  of the Helix itself or that of the target photospheres. It may be a local 
462:  feature of the ISM affecting only these lines of sight. 
463: 
464: There is also  a distinct component at 
465: $v_{\rm helio}$\,$\approx$\, $-10$\,km\,s$^{-1}$ 
466:  for \#3 and \#7, and this might {\it a priori} be assigned to
467:  IS matter in these lines of sight, since their W$_{\lambda}$(D2) are
468:  $\sim$\,100 and 230\,m\AA~for \#3 and \#7 respectively, and because
469:  this velocity falls in the expected range for IS absorptions. The relatively small 
470: value for \#3 (suggesting a lack of a circumstellar contribution) 
471: is surprising in view of the location of this target within the CO isophotes
472: (see Fig.~1 of Young et al. \cite{you99}).This line of sight should trace 
473: neutral/molecular nebular material, but this is not apparent in Na\,{\sc i} and 
474: it may be that the Helix is located beyond this star, which is the nearest of 
475: all the targets. 
476: In any case, the component is not at the observed CO velocity
477: in this region, between $-$30 and $-$50 \,km\,s$^{-1}$, see Fig.~3 of 
478: Young et al. (\cite{you99}). 
479: For target \#7, well off the PN and CO contours, the strength
480: of this particular component is typical of the ISM in the direction of the Helix.
481: 
482: \subsubsection{ Targets 2, 8 (Fig.\,\ref{f2}, upper right panels)}
483: 
484:   One can pursue this analysis further by considering target \#2 (the PN nucleus)
485:  and target \#8, for which no blending with photospheric lines
486:  exists. For the nucleus, this is because no photospheric Na\,{\sc i} line is expected, 
487:  given its likely very hot $T_{\rm eff}$. Target \#8 is cooler with an 
488:  estimated F5V spectral type, but its  $v_{\rm helio}$ is such 
489:  (+76\,km\,s$^{-1}$) that the Na\,{\sc i} components are unaffected by any 
490:  photospheric Na line, or indeed any other line, since the D1 and D2 
491:  profiles resemble each other accurately.
492:  The distances of these two stars are  $\sim$\,210--240 pc.
493:  For \#2, W$_{\lambda}$(D2) is found to be  $\sim$\,150\,m\AA, and we
494:   do not expect any circumstellar absorption, because of the PN geometry, 
495: or photospheric absorption, so by definition this is the ISM probed towards the nucleus 
496: itself, at its fairly well-known distance. 
497: 
498: For target \#8, well off the PN, the
499: photospheric and (presumably) IS components are well separated (as mentioned above) 
500: and we can measure the  photospheric (D2) strength as $\sim$\,250\,m\AA~(seemingly
501: typical of all our F--G type targets). The IS component is surprisingly strong at 
502: 410\,m\AA. Because the absorption velocities in both \#2 and \#8 are  
503: at $\sim$\,+15 km\,s$^{-1}$, not within the Helix $v_{\rm helio}$ range, 
504: the simplest  interpretation is that we are again tracing  the foreground ISM, although
505: one cannot entirely rule out a nebular contribution.
506: 
507: \subsubsection{ Targets 4 and 5 (Fig.\,\ref{f2}, lower right panels)}
508:  
509:  Finally, targets \#4 and \#5 are particularly interesting for our 
510:  purposes because they are distant ($\sim$\,1500\,pc), and because they 
511:  are located close to the edge or angularly slightly 
512:  outside the bulk of the ionized PN gas. They are close to 
513:  the lowest contours of the available H\,{\sc i} and CO maps. Therefore,
514:  one can expect that these line of sight probe some neutral PN gas. Examination
515:  of the data indeed suggest this is the case.
516:  
517: Given the strength of the Na\,{\sc i} absorptions towards \#5, and that 
518: their velocity range  lie largely within the PN CO velocity range, we are led to
519:  attribute at least a significant part of the absorption  to circumstellar
520: material. The case of \#4 is perhaps even more clear cut; there is a strong, 
521: well-resolved component centred very closely
522: on the $v_{\rm helio}$ of the Helix  which lies at the (negative) velocity limit 
523: of where IS absorptions are expected. This component is at 
524: $\sim$$-$25\,km\,s$^{-1}$; the equivalent width of the individual component 
525: in D2 is 350\,m\AA, exceeding the canonical ISM expectation by nearly a factor of 2. 
526: Similarly, we note that \#5 has by far the greatest total Na\,{\sc i} equivalent 
527: width,  in both D1 and D2, of all the targets, certainly where the photosphere 
528: is well removed in velocity (\#7 has comparable absorption strength
529: but includes a significant photospheric component). This in itself lends support to 
530: the thesis that we are probing nebular Na\,{\sc i} towards \#4 and \#5.
531: 
532:  The surprising strength of the Na\,{\sc i} lines to
533:  target \#5 suggests a further favourable argument; summing over the three components, 
534:  one finds  W$_{\lambda}$(D2) $\approx$\,1000\,m\AA, much stronger than the  
535:  expected level from the ISM. Moreover, the individual D1 and D2 components, 
536:  which are only partially resolved our velocity resolution (6\,km\,s$^{-1}$), 
537:  are in fact probably saturated, and the ratio W$_{\lambda}$(D2)/W$_{\lambda}$(D1) is
538:  found to be 980/740\,$\sim$\,1.3, significantly less than a factor of 2, indicating that 
539:  the lines are not formed in the linear regime of the curve of growth. A much better
540:  resolution ($\sim$\,1\,km\,s$^{-1}$) would be necessary to accurately derive
541:  column densities, and assess, e.g. the importance of blending, saturation 
542:  and wing shapes. However, one can obtain a rough estimate by comparing 
543:  the data to IS cases described in previous studies of IS Na\,{\sc i} lines. 
544:  By considering the IS Na\,{\sc i} survey of Hobbs (\cite{hob78}), a 
545:  W$_{\lambda}$(D2) value of $\approx$\,1000\,m\AA\,\, is enormous compared 
546:  to usual IS lines of sight, especially when one considers the Helix 
547:  high galactic latitude. In Hobbs' survey, the largest D2 Na\,{\sc i} widths are at 
548:  500, 520 and 530 m\AA, and the corresponding $N$(Na\,{\sc i}) are 3.1, 1.3 and 7.8 
549:  $\times$ 10$^{+13}$\,cm$^{-2}$. In view of these numbers,
550:  one can reasonably estimate that to \#5 the column density
551:  $N$(Na\,{\sc i}) is of the order of at least $\sim$\,4\,$\times$\,10$^{+13}$\,cm$^{-2}$,
552:  or $\sim$\,\,20 or 40 times stronger than the expectation from the galactic layer 
553:  ($\sim$ 1--2\,$\times$\,10$^{+12}$\,cm$^{-2}$, Sect.~4). 
554:  
555:  To conclude, we strongly favour
556:  a circumstellar origin for the Na\,{\sc i} components described above.
557:  Finding some Na\,{\sc i} in the neutral gas of the Helix is also consistent with
558:  other observations on several PNs (Dinerstein \& Sneden \cite{din88}, 
559:  Dinerstein et al. \cite{din95}).
560:  In addition, we shall see in the following that further evidence for a circumstellar 
561:  origin is carried  by  the Ca\,{\sc ii} lines analysed below.
562:   
563: 
564: \subsection{Analysis of the Ca\,{\sc ii} lines}
565: 
566: In Fig.\,\ref{f2} are displayed the Ca\,{\sc ii} regions for  all targets. Except for 
567: the nucleus (\#2) that provides a generous, flat continuum, all other targets show 
568: broad, deep Ca\,{\sc ii} photospheric lines, implying unfortunately low $S/N$ ratios. 
569: Note that with the exception of \#7, the spectra have been normalised to the highest point
570: in the displayed range so as to best display the details, since the IS or circumstellar
571: features must be sought against the local continuum  noise. For \#7, the signal is much too
572: faint and the $S/N$ too poor for showing any detection, because the velocity range falls
573: in the very deep core of the Ca\,{\sc ii} lines (\#7 a K giant). For this star, 
574: the line profile have been shifted up by 0.2 for clarity. 
575: 
576: The IS Ca\,{\sc ii} line is clearly seen to target \#2, with a strength  in
577: very good agreement with those of the neighbouring stars of Table 2. This IS component at 
578: $v_{\rm helio}$\,$\sim$\,+10 km\,s$^{-1}$ is also seen, more or less clearly, to
579: \#1, \#8, \#4 and \#5. For  \#3, \#6 and \#7, the noise is too large.
580: 
581: It is therefore quite surprising to notice that, for  \#4, absolutely no trace
582: of a Ca\,{\sc ii} line is seen for the strong, double component at $\sim -20$\,km\,s$^{-1}$.
583: Similarly, it is  even more striking that, to  \#5, 
584: the components at $v$\,$<$\,$-5$\,km\,s$^{-1}$ , being nearly saturated in Na D, have no
585: Ca\,{\sc ii} counterparts. The contrast with the case of \#2 is obvious.
586: 
587: This lack of Ca\,{\sc ii} components can be  easily understood in the case
588: of absorption by circumstellar matter. There have been already a number of studies
589: concerning abundances in the gas phase of PN that showed that Ca is
590: very depleted in the ionized phase, otherwise e.g. the [Ca\,{\sc ii}] 7291\,\AA\, and 
591: 7323\,\AA\, forbidden lines would be of great strength, whereas they are generally absent. 
592: The commonly accepted explanation is that, even in the ionized phase,  
593: Ca is almost completely 
594: locked in very robust grains that resist the effects of the hard UV radiation or
595: interaction with the PN plasma  (see for example Shields \cite{shi83}; 
596: Ferland \cite{fer93};  Volk et al. \cite{vol97}; Dopita et al. \cite{dop94};
597: Dopita et al. \cite{dop00}).
598: These grains are not easily directly observed (or/and their abundance measured),
599:  but they are necessary to explain the 
600: strong depletion of various refractories. This extreme Ca 
601: depletion in PNs is quite general and occurs in both O-rich or C-rich PNs. 
602: It seems to us natural to deduce that since Ca is depleted in PNs, 
603: e.g. {\it their ionized phase}, Ca is also very depleted in the neutral phase 
604: that we are probing with Na\,{\sc i} (Na is
605: not a refractory species and is not, or only  very lightly, depleted). 
606: Our interpretation is  the Ca atoms reside in 
607: the hard cores, condensed at high temperature, of the heterogeneous grains that 
608: form in the cool AGB winds and that eventually lose their more fragile 
609: envelope during the PN formation. 
610: 
611: \subsection{A search for circumstellar diffuse bands}
612: 
613: %%%%=====================================================================================
614: \begin{figure}
615: \resizebox{\hsize}{!}{\rotatebox{-90}{\includegraphics{zz-plotz-6614-t85.ps}}}
616: \resizebox{\hsize}{!}{\rotatebox{-90}{\includegraphics{zz-plotz-6614-t91.ps}}}
617: \resizebox{\hsize}{!}{\rotatebox{-90}{\includegraphics{zz-plotz-5780-t91.ps}}}
618: \caption[]{{\it Upper panel}: Spectrum of \#4 in the region of the 6614\,\AA\, 
619: DIB (``Data''). The overplotted dotted line is the expected 
620: 6614\,\AA\, feature  for an equivalent width of 50\,m\AA\, and no noise. 
621: The lower curve (``Simul.'') is the same DIB simulated after adding noise 
622: similar to the data noise. 
623: {\it Middle panel}: Same for \#5, for which the data $S/N$ is significantly better. 
624: If present, a 6614\,\AA\,  DIB of 50\,m\AA\, would be very clearly seen to \#5. 
625: {\it Lower panel}: Spectrum of \#5 near the 5780 DIB. 
626: A  DIB at 5780\,\AA\, of 100\,m\AA\, is shown and  would be fairly well detected.}
627: \label{f3}
628: \end{figure}
629: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%==============================================
630: 
631:  We searched for DIBs in all the targets spectra, and found none. For targets
632:  probing the ISM (see above), this lack of DIBs is fully expected, because the
633:  IS excess is only $E_{\rm B-V}$\,$\sim$\,0.03, and the $S/N$ of our spectra is limited,
634:  despite $\sim$\,1\,hr exposure times with VLT-UVES.
635:  For example, from the work of Kumar (\cite{kum86}) on low colour excess stars (see his Table 1),
636:  one finds that, for $E_{\rm B-V}$\,=\,0.10\,$\pm$\,0.01, the 6614\,\AA~ DIB equivalent 
637:  width (noted $W_{\lambda}$)
638:  is $W_{\lambda}$(6614)\,=\,20\,m\AA\, (4 stars, $\sigma$\,=\,2\,m\AA);
639:  for  $E_{\rm B-V}$\,=\,0.075\,$\pm$\,0.005, $W_{\lambda}$(6614)\,=\,5\,m\AA\, 
640:  (8\,\,stars, $\sigma$\,=\,4\,m\AA). 
641:   Alternatively, Galazutdinov et al. (\cite{gal98}) show for the $\lambda$5780 DIB that
642:  equivalent widths of the order of 30--40\,m\AA\, are observed for their lowest colour excess
643:  sightlines ($E_{\rm B-V}$\,=\,0.05\,$\pm$\,0.01, comparable to those discussed above). 
644:  The DIB strengths due to the ISM in the 
645:  direction of the Helix are thus quite small (perhaps as low  as 
646:  $\sim$\,5\,m\AA\,  but possibly up to 30\,m\AA) and 
647:  simulations presented below show that our experiments cannot detect such weak  
648:  DIB signals.  However, our experiment may only be perhaps
649:  a factor 2--3 insufficiently sensitive to detect the ISM in the strongest DIBs.  
650: 
651:  
652:  Figure \ref{f3} shows the observed spectra of targets \#4 and \#5 in the region 
653:  of the 6614\,\AA~ DIB (labelled ``Data''). For
654:  both objects, we also show  the DIB profiles that would be seen if their equivalent
655:  width were 50\,m\AA\, either with an infinite $S/N$ (overplotted dotted line), or
656:   with an added noise similar to the data (lower curve). The DIB profile
657:  was approximated by a Gaussian with FWHM\,1.0\,\AA\, (Walker et al. \cite{wal00}). 
658:  The DIB central wavelength is the rest value,
659:  6613.56\,\AA, as given by Galazutdinov et al. (\cite{gal00}), with uncertainty $\sim$ 0.1\,\AA. 
660:  No velocity shift was applied since the Earth velocity shift of +20\,km\,s$^{-1}$ 
661:  and the average heliocentric velocity of the Helix ($-28$\,km\,s$^{-1}$) almost cancel. 
662:  One can see that for \#4, a 50\,m\AA\, 6614\,\AA~  DIB would be fairly
663:  detectable, and for \#5 where the $S/N$ is better, it would be very clearly seen. 
664:  On the basis of these plots, a 3\,$\sigma$ upper limit to 
665:  $W_{\lambda}$(6614) for target \#5 is taken as $\sim$\,25\,m\AA. 
666:  
667:  The limits for other DIBs, e.g. at 5780\,\AA , are not better (see Fig.\,\ref{f3}). 
668:  The 5780\,\AA\,  DIB  is generally stronger in equivalent width 
669:  than $\lambda$6614, but it is also 
670:  twice wider; $\sim$\,2.0\,\AA\, FWHM, as observed by Galazutdinov et al. (\cite{gal98}). 
671:  In our spectra, its detection is not easier than for $\lambda$6614, 
672:  especially because of faint photospheric lines lying in this
673:  region. In the end, for $W_{\lambda}$(5780), we derive an
674:  upper limit of $\sim$\,70\,m\AA. 
675:   
676:   Instead of comparing the data with simulations, 
677:  one might derive  3\,$\sigma$ upper limits with the usual formula, as given by, 
678:   e.g.,\, Cowie \& Songaila (\cite{cow86}): for a line that falls over 
679:   $m$ resolution elements, the $3\sigma$ detectable equivalent width is 
680:  $W(3\sigma)$\,=\,3\,$\lambda$\,$m^{1/2}$\,$R^{-1}$\,$(S/N)^{-1}$, 
681:  where $S/N$ is the signal-to-noise ratio in each resolution element.
682:  For the $\lambda$6614 DIB, one finds with our data $W(3\sigma)$\,$\sim$\,10\,m\AA. 
683:  However, the above formula, while taking into account the observed $S/N$, assumes an absence of 
684:  other (photospheric) lines,
685:  which may not be the case (see Fig.\,\ref{f3}).  
686:  Therefore, we believe that the limits derived from simulations are more 
687:  appropriate.
688:  
689: \section{Discussion}
690: 
691:   The primary goal of this work was to search for circumstellar diffuse bands: none 
692:  were found, but we discovered two targets, \#4 and \#5, whose lines of sight indeed probe
693:  some neutral PN gas, with such a strong Na\,{\sc i} absorption, and strong Ca 
694:  depletion, that the two together cannot possibly  be due to the foreground or 
695:  background ISM. 
696: 
697:  A further indication that the sightline to \#5 is far from typically interstellar is obtained by
698:  plotting $W_{\lambda}$(5780) $versus$ $W_{\lambda}$(D2) (Fig.\,\ref{f4}). The data for IS
699:  lines were taken from the compilation of Herbig (\cite{her93}) and involves sightlines with
700:  $E_{\rm B-V}$ between 0 and 1.5. We see a clear increase of $W_{\lambda}$(5780)
701:  with increasing D2 equivalent width. The representative point
702:  of sightline \#5 is the full black square. Not only $W_{\lambda}$(D2) appears
703:  huge compared to other sightlines, especially when one realizes that \#5 is 
704:  at high galactic latitude, but this plot suggests that if {\em only} IS matter were
705:  probed, then we would see $W_{\lambda}$(5780)$\ga$\,400\,m\AA,
706:  which is definitely not the case. Therefore, the material seen in the line of sight to
707:  \#5 must be overwhelmingly of circumstellar origin.
708:  
709:  
710:  The fact that no DIBs have been seen is consistent with their absence in several other
711:  archetypal sources with carbon-rich circumstellar material. For example, in the case of
712:  IRC\,+10$^{\circ}$\,216 which is very rich in molecules, including carbon chains, 
713:  no DIB was detected on a sightline for which the Na\,{\sc i} absorption lines were
714:  heavily saturated, the K\,{\sc i}  lines were very strong  with 
715:  $W_{\lambda}$\,$\sim$\,500\,m\AA~ (Kendall et al. \cite{ken02}), 
716:  and the C$_2$ molecule was detected (Kendall et al.  \cite{ken04}).
717:  In the case of the post-AGB star HR\,4049, Waters et al. (\cite{wat89}) showed that the
718:  circumstellar dust is carbon-rich and displays the emission features at 7, 9 and 11.3
719:  $\mu$m attributed to PAHs, but no DIB was detected. Waters et al. (\cite{wat89})
720:   obtained a stringent 
721:  limit on the DIB abundance, with $W_{\lambda}(6614) < 4.2$ m\AA\, despite
722:  a large circumstellar reddening, $E_{\rm B-V}$ being between 0.19 and 0.29.
723:  Compared to these two objects, the Helix is a third interesting case; its nucleus is much
724:  hotter than HR\,4049, but the PN also still contains much circumstellar 
725:  molecular and dusty material left by the AGB progenitor wind. A peculiarity
726:  of the Helix is that there are no PAH mid-infrared features, interpreted by Cox et al. 
727:  (\cite{cox98})
728:  as due to a destruction of small grains by the hard radiation field of the nucleus in the
729:  early PN phase.
730:  
731:  Finding no DIBs in these three typical but quite different
732:  objects suggests that DIBs are not fabricated in AGB envelopes or PNs. However, 
733:  it will now become even more  necessary to $i$.) fully explain the origin of the
734:  emission bands in the Red Rectangle and their possible relation to DIBs; 
735:  and $ii$.) further examine the few cases
736:  of carbon stars and post-AGB objects mentioned in Sect.~1 for which evidence for 
737:  circumstellar diffuse bands have been put forward. 
738:  
739:  It will also be interesting  to perform specific observations of targets \#4 and \#5 
740:  with a higher spectral resolution and/or a better signal-to-noise ratio in order to detect 
741:  other species, e.g. K\,{\sc i}, C$_{2}$ or CN,  to establish whether 
742:  these circumstellar lines of sight probe any molecular matter, to  infer 
743:  the total hydrogen column density and obtain a more quantitative upper limit to the 
744:  DIB abundance.
745: 
746: %=========================== figure showing $W_{\lambda}$(5780) versus $W_{\lambda}$(NaI 5890)
747:    \begin{figure}
748:    \centering
749:    \resizebox{\hsize}{!}{
750:   {\rotatebox{-90} {\includegraphics{zzz-plot-na-5780.ps}}   }
751:   } 
752:   \caption{Plot of the equivalent widths of the 5780\,\AA~DIB {\it versus} the 
753:    Na\,{\sc i} D2 5890\,\AA\, equivalent widths (open circles), for interstellar 
754:    data of Herbig (\cite{her93}). The solid square at lower right
755:    represents the upper limit of $\lambda$5780 for target \#5, 
756:    which has $W_{\lambda}$(Na\,{\sc i} D2) $\sim$ 1000\,m\AA. This plot
757:    illustrates the deficiency of $\lambda$5780, 
758:    further supporting the {\it circumstellar} 
759:    nature of the matter probed by this target.}
760:    \label{f4}
761:    \end{figure}
762: %%%%%%%%%%%%%%%%%%%%%====================================================================
763: 
764: 
765:    
766:   \section{Summary and conclusions}
767:  
768:  1.-- With the goal of studying the neutral gas
769:  of the Helix PN and the presence of circumstellar
770:  diffuse bands, VLT/UVES  spectroscopy was achieved 
771:  on 8 targets lying within
772:  $\sim$\,13\arcmin~ of the PN centre, including its nucleus. 
773:  Using  available 2MASS and UBV photometry, the spectral 
774:  types of the targets could be estimated. This allowed the  
775:   derivation of  distances, which are  found to be in the
776:  range $\sim$ 200\,pc to 1500\,pc. Thus, several targets
777:  are located far beyond the Helix and can be used as background
778:  sources, angularly placed on or off the visible nebula.
779:  
780:  2.-- Of the 8 targets, four (including the nucleus)
781:   can particularly be exploited because they show little 
782:  blending between  the interstellar/circumstellar Na\,{\sc i} lines 
783:  and the photospheric lines of the background target itself, 
784:  which are either absent, favorably shifted by the stellar radial 
785:  velocity, or/and broadened by stellar rotation. 
786:   
787:  3.--  By analysing  the  extinction maps around the Helix, and  IS
788:  lines observed towards 3 bright stars in a $\sim$\,5\degr~zone around the Helix,
789:  we could estimate the expected contribution 
790:  of the foreground and background ISM, in terms of equivalent
791:  width of the Na\,{\sc i} lines (W$_{\lambda}$(D2) $\sim$\,200m\AA), 
792:  and the expected range of heliocentric velocities for IS lines.
793:  
794:  4.-- It appears that many of the Na\,{\sc i} components observed in the target
795:  spectra are probably of IS origin. However, for two targets (\# 4 and \#5), 
796: our analysis shows that {\it circumstellar} Na\,{\sc i} is observed in absorption. 
797:  For  \#5, a  very strong Na\,{\sc i} column density is derived, at least 
798:  20 times larger than  the expected IS contribution. 
799:   The circumstellar origin is also supported by the fact that no corresponding 
800:  absorption Ca\,{\sc ii} K line is detected, in contrast to IS lines. This fact
801:  is well understood: since Ca is always very depleted  in the ionized phases of
802:  planetary nebulae, it is probably  depleted as well and locked 
803:  in the circumstellar grains of the cool neutral phase, where dust has not yet been 
804:  subject to strong UV radiation and plasma interaction. 
805:  
806: 5.-- No diffuse bands were detected. The expected bands from the foreground
807: and background ISM are not strong enough ($W_{\lambda}$(6614)\,$\sim$\,5 to 20\,m\AA) to 
808: be seen in our spectra. Nor do the circumstellar lines of sight to \#4 and \#5
809: show any bands. 
810: Qualitatively, our data do not favour abundant fabrication of diffuse band 
811: carriers in the Helix.
812: 
813: 
814: \begin{acknowledgements}
815: The authors warmly thank Claudio Melo, Jonathan Smoker, and the ESO staff,
816: for their assistance during observations. Thanks  also to Christophe Mercier
817: at Montpellier for his help in making Figure 1. We also thank  the anonymous referee
818: for comments that helped to improve the paper. N.M. acknowledges support from the
819: French CNRS Program  {\it Physico-chimie du Milieu Interstellaire}. T.R.K.
820: acknowledges support from the French {\it Minist\`{e}re de la Recherche}. The publication 
821: makes use of data products from 2MASS, which is a joint project of the
822: Univ. of Massachusetts and the Infrared Analysis and Processing Centre, California 
823: Institute of Technology, funded by  NASA and NSF. 
824: This research has also made use of the ESO MIDAS data analysis 
825: software, the NOAO IRAF software and the SIMBAD database, operated
826: at CDS, Strasbourg, France. 
827:     \end{acknowledgements}
828: 
829: \begin{thebibliography}{}
830: 
831: \bibitem[1983]{alb83} Albert, C.E. 1983, ApJ, 272, 509
832: 
833: \bibitem[1978]{boh78} Bohlin, R.C., Savage, B.D., 
834: \& Drake, J.F. 1978, ApJ, 224, 132
835: 
836: \bibitem[1986]{cow86} Cowie, L.L., \& Songaila, A., 1986, 
837: ARA\&A, 24, 499
838: 
839: \bibitem[1998]{cox98} Cox, P., Boulanger, F., Huggins, P.J., et al. 1998,
840: ApJ, 495, L23
841: 
842: \bibitem[1982]{bur82} Burnstein, D., 
843: \& Heiles, C. 1982, AJ, 87, 1165 
844: 
845: \bibitem[1988]{din88} Dinerstein, H.L., 
846:  \& Sneden, C. 1988, ApJL, 335, L23
847:  
848: \bibitem[1995]{din95} Dinerstein, H.L., Sneden, C., \&
849:  Uglum, J. 1995, ApJ, 447, 262
850:  
851: \bibitem[1994]{dop94} Dopita, M.A., Vassiliadis, E., 
852: Meatheringham, S.J., Ford, H.C., et al. 1994, ApJ, 426, 150
853: 
854: \bibitem[2000]{dop00} Dopita, M.A., \& Sutherland, R.S. 
855: 2000, ApJ, 539, 742
856:  
857: \bibitem[1998]{dul98} Duley, W.W. 1998, MNRAS, 301,955
858:  
859: \bibitem[1993]{fer93} Ferland, G.J. 1993, in  Planetary Nebulae, 
860: ed. R. Weinberger \& A. Acker (Dordrecht, Reidel), 123
861: 
862: \bibitem[1985]{fer85} Ferlet, R., Vidal-Madjar, A., 
863: \& Gry, C. 1985, ApJ, 298, 838
864: 
865: \bibitem[1991]{gar91} Garcia, B. 1991, A\&AS, 89, 469
866: 
867: \bibitem[1998]{gal98} Galazutdinov, G.A., 
868: Krelowski, J., Moutou, C., \& Musaev, F.A. 1998, MNRAS, 295, 437 
869: 
870: \bibitem[2000]{gal00} Galazutdinov, G.A., 
871: Musaev, F.A., Krelowski, J., \& Walker, G.A.H. 2000, PASP, 112, 648
872: 
873: \bibitem[2002]{gli02} Glinski, R.J., \& Anderson, 
874: C.M. 2002, MNRAS, 332, L17
875: 
876: \bibitem[1993]{her93} Herbig, G.H. 1993, ApJ, 407, 142
877: 
878: \bibitem[1995]{her95} Herbig, G.H. 1995, ARA\&A, 33, 19
879: 
880: \bibitem[1978]{hob78} Hobbs, L.M. 1978, ApJSS, 38, 129
881: 
882: \bibitem[1999]{hug99} Huggins, P.J., Cox, P., Forveille, T.,
883: Bachiller, R., \& Young, K. 1999, in IAU Symposium 191, 
884: Asymptotic Giant Branch Stars, ed. T. Le Bertre, 
885: A. Lebre \& C. Waelkens (Astronomical Society of the Pacific), 425
886: 
887: \bibitem[1988]{kee88} Keenan, F.P., Conlon, E.S., 
888:   Brown, P.J.F., \& Dufton, P.L. 1988, A\&A, 192, 295
889:   
890: \bibitem[2002]{ken02} Kendall, T.R., Mauron, N., McCombie, J., 
891: \& Sarre, P.J. 2002, A\&A, 387, 624
892: 
893: \bibitem[2004]{ken04} Kendall, T.R., Mauron, N., McCombie, J., 
894: \& Sarre, P.J. 2004, Ap\&SS, 289, 203  
895: 
896: \bibitem[1999]{klo99} Klochkova, V.G., Szczerba, R., 
897: Panchuk, V.E., \& Volk, K. 1999, A\&A, 354, 905
898: 
899: \bibitem[2000]{klo00} Klochkova, V.G., Szczerba, R., 
900: \& Panchuk, V.E. 2000, Astr. Lett., 26, 88
901: 
902: \bibitem[2001]{klo01} Klochkova, V.G., Panchuk, V.E., 
903: \& Szczerba, R. 2001, in  Post-AGB objects
904: as a Phase of Stellar Evolution, ed. R. Szczerba \& S.K. Gorny 
905: (Kluwer Academic Publishers), 265
906: 
907: \bibitem[1986]{kum86} Kumar, C.K. 1986, ApJ, 306, 38
908: 
909: \bibitem[1990]{leb90} Le Bertre, T. 1990, A\&A, 236, 472
910: 
911: \bibitem[1992]{leb92} Le Bertre, T., \& Lequeux, J. 
912: 1992, A\&A, 255, 288
913: 
914: \bibitem[1993]{leb93} Le Bertre, T., \& Lequeux, J. 
915: 1993, A\&A, 274, 909
916: 
917: \bibitem[1993]{rao93} Rao, N.K., \& Lambert, D.L. 
918: 1993, MNRAS, 263, L27
919: 
920: \bibitem[2002]{rod02} Rodriguez, L.F., Goss, W.M., \& 
921:  Williams, R. 2002, ApJ, 574, 179
922: 
923: \bibitem[1995]{sar95} Sarre, P.J., Miles, J.R., 
924: \& Scarrott, S.M. 1995, Science, 269, 674
925: 
926: \bibitem[1992]{sca92} Scarrott, S.M., Watkin, S., Miles, J.R., \& Sarre, P.J. 1992,
927: MNRAS, 255, 11
928: 
929: \bibitem[1998]{sch98} Schlegel, D.J., Finkbeiner, D.P., 
930: \& Davis, M. 1998, ApJ, 500, 525. 
931: 
932: \bibitem[1983]{shi83} Shields, G.A. 1983, in  Planetary Nebulae, 
933: ed. D.R. Flower (IAU), 259
934: 
935: \bibitem[2002]{spe02} Speck, A.K., Meixner, M., Fong, D., McCullough, 
936: P.R., Moser, D.E., \& Ueta, T. 2002, AJ, 123, 346
937: 
938: \bibitem[1973]{sno73} Snow, T.P. 1973, PASP, 85, 590
939: 
940: \bibitem[1972]{snw72} Snow, T.P., Wallerstein, G. 1972, PASP, 84, 492
941: 
942: \bibitem[2002]{vwi02} Van Winckel, H., Cohen, M., 
943: \& Gull, T.R. 2002, A\&A, 390, 147
944: 
945: \bibitem[1997]{vol97} Volk, K., Dinerstein, H., \& Sneden, C.1997, in 
946: IAU Symposium 180,  Planetary Nebulae, ed. H.J. Habing \& H.J.G.L.M. 
947: Lamers (Kluwer Academic,  Publishers, Dordrecht), 284
948: 
949: \bibitem[2000]{wal00} Walker, G.A.H., Bohlender, D.A., 
950: \& Krelowski, J. 2000, ApJ, 530, 362 
951: 
952: \bibitem[1989]{wat89} Waters, L.B.F.M., Lamers, H.J.G.L.M., Snow, T.P.,
953: et al. 1989, A\&A, 211,208
954: 
955: \bibitem[1997]{you97} Young, K., Cox, P., Huggins, P.J., Forveille, Y., 
956: \& Bachiller, R. 1997, ApJ, 482, L101
957: 
958: \bibitem[1999]{you99} Young, K., Cox, P., Huggins, P.J., Forveille, T., 
959: \& Bachiller, R. 1999, ApJ, 522, 837
960: 
961: \bibitem[1999]{zac99} Zacs, L., Schmidt, M.R., \& Szczerba, R. 1999, 
962: MNRAS, 306, 903
963: 
964: \bibitem[2003]{zac03} Zacs, L., Spelmanis, R., 
965:          Musaev, F.A., \& Galazutdinov, G.A. 2003, MNRAS, 339, 460
966:   
967: \end{thebibliography}
968: \end{document}
969: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
970: 
971: 
972: 
973: 
974: 
975: 
976: 
977: