Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~wms/sProcessAbunds.pdf Äàòà èçìåíåíèÿ: Wed May 6 23:27:04 2009 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 00:06:13 2012 Êîäèðîâêà:

The Astrophysical Journal, 659:1265 Y 1290, 2007 April 20
# 2007. The American Astronomical Society. All rights reserved. Printed in U.S.A.

A

s-PROCESS ABUNDANCES IN PLANETARY NEBULAE
Brian Sharpee,1 Yong Zhang,2, 3, 4 Robert Williams,2 Eric Pellegrini,5 Kenneth Cavagnolo, Jack A. Baldwin,5 Mark Phillips,6 and Xiao-Wei Liu 3
Received 2005 May 6; accepted 2006 December 4
5

ABSTRACT The s-process should occur in all but the lower mass progenitor stars of planetary nebulae, and this should be reflected in the chemical composition of the gas that is expelled to create the current planetary nebula shell. Weak forbidden emission lines are expected from several s-process elements in these shells and have been searched for and in some cases detected in previous investigations. Here we extend these studies by combining very high signal-to-noise ratio echelle spectra of a sample of PNe with a critical analysis of the identification of the emission lines of Z > 30 ions. Emission lines of Br, Kr, Xe, Rb, Ba, and Pb are detected with a reasonable degree of certainty in at least some of the objects studied here, and we also tentatively identify lines from Te and I, each in one object. The strengths of these lines indicate enhancement of s-process elements in the central star progenitors, and we determine the abundances of Br, Kr, and Xe, elements for which atomic data relevant for abundance determination have recently become available. As representative elements of the ``light'' and ``heavy'' s-process peaks, Kr and Xe exhibit similar enhancements over solar values, suggesting that PN progenitors experience substantial neutron exposure. Subject headings: ISM: abundances -- nuclear reactions, nucleosynthesis, abundances -- g planetary nebulae: general Online material: machine-readable table, tar file 1. INTRODUCTION As remnants of stars that have evolved through the asymptotic giant branch (AGB) phase, most planetary nebulae ( PNe) are believed to consist of material that has undergone nuclear processing in the precursor star via the s-process. The analysis of nebular emission from elements that have experienced nucleosynthesis in the parent star provides valuable information for stellar models. However, the detection of emission lines from ions enhanced by the s-process has been hampered by their weakness and by uncertainties in the atomic data needed for the analysis, including line wavelengths. An initial attack on this problem was made a decade ago by ´ Pequignot & Baluteau (1994, hereafter PB94), who obtained a deep optical spectrum of the high-ionization PN NGC 7027. Using the best atomic data available for the energy levels of the more prominent ionization stages of elements in the fourth and fifth rows of the periodic table, they identified a number of post Y Fe peak emission lines in the nebula. From the observed line intensities they concluded that the elements Kr and Xe, the latter normally a predominantly r-process element in stars of solar metallicity, were enhanced in NGC 7027 by factors of $20 relative to their initial formation abundances, presumably roughly solar. They also detected Ba ii and [ Br iii] emission at intensities indicating that Ba could be enhanced whereas Br might be depleted relative to solar values, subject to uncertainty due to poorly known excitation cross sections. Dinerstein and collaborators have subsequently pursued the study of s-process abundances in PNe through surveys to detect
SRI International, Menlo Park, CA 94025. Space Telescope Science Institute, Baltimore, MD 21218. Department of Astronomy, Peking University, Beijing 100871, China. 4 Current address: Department of Physics, University of Hong Kong, Hong Kong, China. 5 Department of Physics and Astronomy, Michigan State University, East Lansing, MI 48824. 6 Las Campanas Observatory, Carnegie Observatories, Casilla 601, La Serena, Chile.
2 3 1

IR fine-structure nebular emission from post Y Fe peak ions ( Dinerstein 2001; Sterling & Dinerstein 2004) and from far-UV resonance-line absorption by s-process elements in the intervening nebular shell seen in Far Ultraviolet Spectroscopic Explorer (FUSE ) spectra of their central stars (Sterling el al. 2002; Sterling & Dinerstein 2003). They detected IR emission from Se and Kr in roughly 50% of their sample PNe from which abundances of up to 10 times solar were derived for these elements. The FUSE spectra revealed Ge iii absorption in five PNe for which Ge enhancements of 3 Y 10 times solar were deduced, clear evidence that central star AGB progenitors are major sites of s-process element production. As part of our ongoing program to detect and identify the weakest lines in PNe at visible wavelengths down to levels significantly below that of the continuum, viz., <10þ5 the intensity of H, we have obtained very high signal-to-noise ratio (S/ N ) spectra at high spectral resolution of some of the higher surface brightness nebulae. Some fraction of the weaker lines observed are likely to originate from post Y Fe peak elements; therefore, we have used the automatic line identification routine EMILI (Sharpee et al. 2003), supplemented by recent energy level data for heavier elements, to assist in the search for such lines. Since the publication of PB94, new calculations have been made for sponta´ neous emission coefficients ( Biemont et al. 1995) and collision strengths (Schoning 1997; Schoning & Butler 1998) for fourth¨ ¨ and fifth-row elemental ion transitions. In this study we use these newer atomic parameters to compute ionic and overall elemental abundances for Br, Kr, and Xe in order to make comparisons with their abundances in H ii regions and the Sun and to derive s-process enrichment factors relevant to the study of s-process nucleosynthesis in the progenitor stars. 2. OBSERVATIONS AND DATA REDUCTION Our present sample of objects consists of four PNe ( IC 2501, IC 4191, NGC 2440, and NGC 7027 ) that satisfy the most important criteria for the detection of weak emission lines in having (1) high surface brightnesses, (2) low expansion velocities, and 1265


1266

SHARPEE ET AL.
TABLE 1 Journal of Observations Dates (UT ) 2002 2003 2003 2003 2003 2003 Jun Feb Feb Feb Feb Feb 19 Y 22 12 Y 13 25 Y 26 12 Y 13 25 Y 26 12 Y 13 Integration Times (s) 34 ; 1200, 9 ; 300, 1 ; 120, 4 ; 60, 6 ; 30 6 ; 1800, 1 ; 300, 4 ; 60 15 ; 1800, 2 ; 60 5 ; 1800, 1 ; 1200, 1 ; 300, 1 ; 60 10 ; 1800, 1 ; 300, 3 ; 60 11 ; 1800, 1 ; 300, 1 ; 60

Vol. 659

Object NGC 7027 ........................ IC 2501 ............................ IC 2501 ............................ IC 4191 ............................ IC 4191 ............................ NGC 2440 ........................

Telescope KPNO 4 LCO 6.5 LCO 6.5 LCO 6.5 LCO 6.5 LCO 6.5 m m m m m m

Slit Center 3.500 W, 200 200 200 200 1400 0.500 N E E E E E

(3) except for NGC 7027, relatively small internal dust extinction. The latter two criteria produce sharper lines and higher peak intensities relative to the continuum. We obtained spectra of the PNe IC 2501, IC 4191, and NGC 2440 during two observing runs of two nights each in early 2003 using the Las Campanas Observatory ( LCO) Baade 6.5 m telescope with the MIKE echelle spectrograph. Similar instrumentation setups were used during the two runs. MIKE is a dual-beam spectrograph that simultaneously measures separate red and blue spectra, giving useful data over the continuous wavelength ranges 3280 Y 4700 and 4590 Y 7580 8 at resolutions (k /àk) of 28,000 and 22,000, respectively, for the 100 slit width we used. A series of long and short exposures were taken of each nebula on one or more nights, typically adding up to a period of 2 Y 3hr at a time spent observing each object. Because MIKE is used without an image rotator at a Nasmyth focus, the orientation of the slit with respect to each nebula rotated by a large amount during these series of exposures. The central star was placed along a line perpendicular to the spectrograph slit so that the slit center was roughly midway between the central star and outer edges of the nebula, and then the telescope was guided to keep the star in that same position as seen on the acquisition /guide TV. The result was that during the course of the observations the slit swept out an arc about the central star, so that the final spectrum integrates over an area of the nebula that is much larger than the 1 00 ; 5 00 slit. The total integration times in the combined long exposures were 630 minutes for IC 2501, 450 minutes for IC 4191, and 330 minutes for NGC 2440. A journal of the observations is given in Table 1 and includes the approximate position of the center of the slit with respect to each of the central stars at the start of the series of exposures. The optics of the MIKE spectrograph introduce strong distortion into the image formed on the detector, so the projected emission lines have significantly different tilts as a function of their position on it. When extracting spectra of objects such as these PNe that fill the slit, the tilts introduce unacceptable smearing in the wavelength direction (typically 3 pixels over the 40 pixel slit length) unless they are corrected for during the extraction of the one-dimensional spectra from the two-dimensional echelle image. We wrote our own set of auxiliary FORTRAN programs to calibrate the tilts and perform the proper extraction. This procedure was tested on the emission lines in the comparison lamp spectrum. Compared to the emission line from a single pixel at the slit center, lines summed over the full slit length came out at the same pixel location in the dispersion direction and were broadened by 2% on average. We are thus confident that the effects of these tilts are negligible in the extracted PN spectra. These tiltcorrected spectra were then fed into the same suite of IRAF-based reduction programs used with the NGC 7027 data as described next.

NGC 7027 was observed on four nights in 2002 June with the Mayall 4 m Telescope at Kitt Peak National Observatory ( KPNO) using the Cassegrain echelle spectrograph. Because we were interested in detecting faint 40 Y 50 km sþ1 FWHM emission lines over the widest possible wavelength range, rather than detailed measurements of the line profiles, we used the short UV camera with the 79.1 groove mmþ1 echelle grating, crossdispersing grating 226-1 in first order, and a GG-475 order separating filter. This gave full wavelength coverage from 4600 to 9200 8 at $20 km sþ1 FWHM (k /àk ¼ 15;000) resolution with our 200 slit width, with partial coverage out to $9900 8.On each night we offset to the same position with the slit at P:A: ¼ 145 and centered 0.500 north and 3.500 west of the central star, which is the brightest part of the PN shell in H emission. Most of the observing time was spent obtaining sequences of 1200 s exposures that added up to a total of 680 minutes of integration time. We also took a number of shorter exposures 30, 60, 120, an d 3 00 s i n l e n g t h t o m ea s u re b ri g ht e m i ss i o n l i ne s t h a t w er e saturated on the longer exposures. The data for all four PNe were reduced using standard procedures with IRAF-based reduction packages in the same manner that has been described in detail in our discussion of comparable spectra of the PN IC 418 that we obtained previously with the Cerro Tololo Inter-American Observatory 4 m echelle spectrograph (Sharpee et al. 2004). A correction was made for the presence of Rowland ghosts near strong lines ( Baldwin et al. 2000). To correct for flexure and temperature drift effects in the spectrographs, comparison lamp spectra were taken at roughly 1 hr intervals, and the wavelength calibration for each PN spectrum was made using the comparison lamp taken nearest to it in time. The wavelength fits indicate a 1 uncertainty in the wavelength scale that varies over each echelle order, and it is of the order 4 Y 6km sþ1 for all of our spectra. The spectra were flux-calibrated using observations of several spectrophotometric standard stars from Hamuy et al. (1994). The final extracted flux- and wavelength-calibrated spectra of all the PNe were used for our line identification and analysis and are available in FITS format as an electronic supplement to the present article. 3. EMISSION-LINE SELECTION The line selection procedure is initiated by fitting the continuum of each echelle order of the final spectrum with a smooth function that is pegged to the observed continuum at approximately 10 wavelength intervals distributed over each order. We developed an automated procedure that selects what should be reliable continuum points for the fitting by selecting wavelength regions with a paucity of lines. Because bad pixels, artifacts caused by scattered light in the spectrograph due to strong emission lines, and line blends can cause poor fits, the fits were reviewed manually for each order and adjusted as necessary to ensure that the


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES
TABLE 2 Nebula r P arameters Parameter V( geo) ( km sþ1) ........ c (H )......................... NGC 2440 +66.7 0.55 IC 2501 +23.2 0.55 IC 4191 þ40.1 0.77

1267

NGC 7027 þ7.1 1.17

Fig. 1.-- Portion of the IC 2501 spectrum illustrating the automated continuum fitting procedure. The vertical lines designate emission features defined by the ED fitting algorithm.

continuum representation was valid. An example of the output of the automatic continuum fitting algorithm is shown in Figure 1, where a typical fit prior to manual adjustment is displayed. Once the proper continuum level is established, each order is sampled pixel by pixel to detect emission features, which are defined as regions where the observed flux exceeds the continuum flux by 7 or more over a wavelength interval equal to or greater than that of the resolution of the spectrograph. All putative emission features were examined individually by eye and compared with their appearance on the original twodimensional echelle images to establish their reality. We have found that all features with fluxes greater than 12 of that of the continuum, i.e., with S / N > 12, are clearly visible on the echelle images. Since scattered light features usually trail across multiple orders, one can generally distinguish such artifacts from real lines by visually examining the images. The most uncertain aspect of the line detection process is distinguishing multiple line blends in real emission features. The expansion velocities of PNe are such that intrinsic line widths do vary by factors of 2 Y 3with the level of ionization, so it can be difficult to discriminate between one broad line and two or more closely spaced narrow lines. Figure 1 gives an example of the features within a selected wavelength region in one of our PN spectra that have been designated by our software as emission lines. For detection of the weakest nebular lines, where the distinction between a continuum noise spike and a real feature can be difficult to establish with certainty, we have intercompared the spectra of those PNe that have similar levels of ionization but different radial velocities. Noise spikes do not generally produce features having the width of the instrumental resolution, and instrument artifacts tend to occur on the same place on the detector, which is at different rest wavelengths in the PN spectra because of their differing radial velocities. Since spectra were obtained on different instruments ( MIKE and the 4 m KPNO echelle), all significant scattered light ghosts could be detected through comparisons between the PN spectra obtained with MIKE and the NGC 7027 spectrum. Our final line lists do contain a few weak features having S / N < 7 that are present in at least two of the objects. The telluric nightglow emission spectrum was also sampled by these spectra, as sky subtraction in these extended objects was deemed impractical, particularly in regard to the subtraction adding significant additional noise. However, most nightglow lines were distinguishable on the two-dimensional spectra by their uniform intensity and characteristic shape and size, namely, those of the imaging slit. The positions and intensities of prospective nightglow lines were compared to those listed in the telluric feature

atlases of Osterbrock et al. (1996) and Hanuschik (2003) and likely matches removed from the nebular list except in those cases where their blending with stronger nebular lines was deemed only a minor contaminant. For lines considered as candidate s-process ion transitions, nightglow sources were given additional scrutiny through a comparison with the original spectra of the Hanuschik (2003) atlas, the likely identifications of the atlas lines (Cosby et al. 2006), time-averaged observed intensities of their constituent systems (Cosby & Slanger 2007 ), and in some cases model spectra of those systems constructed with the molecular simulation code DIATOM.7 Following the selection of sets of likely nebular lines in each of the PNe, their wavelengths and intensities were determined by single or multiple (in the case of blended features) Gaussian fitting of their profiles and immediate underlying continua. Simple summing was used for the most irregular and ill-defined line profiles with intensity-weighted centroids utilized as wavelengths. Emission lines appearing in multiple orders were then collated and their attributes averaged together. Emission-line intensities and wavelengths from short- and long-duration exposure spectra of each PN were then normalized to a particular fiducial frame. Line wavelengths were shifted to the nebular rest frame through a comparison with either the Balmer ( MIKE PNe) or Paschen ( NGC 7027 ) sequence laboratory wavelengths. Line intensities were dereddened with the Galactic extinction law of Howarth (1983) utilizing an iterative process involving the magnitude of either the Balmer or Paschen jump, as well as the Balmer or He ii 5 Y n sequence decrements to establish electron temperatures and c (H ) logarithmic extinction at H values. The c(H ) values for each PN are in good agreement with those listed in Cahn et al. (1992). Errors in the final line intensities in all PN spectra, as determined from the formal errors to the profile and continua fits, are similar to those found for NGC 7027: 41% for I < 10þ5 I (H ), 20% for I ¼ 10þ5 I (H ) to 10þ4I(H ), 11% for I ¼ 10þ4 to 10þ3, and <6% for I > 10þ3 I (H ) on average. This is independent of any errors arising from the reddening correction. For weak lines, where the greatest contributor to uncertainty is the indeterminate level of the true continuum, the formal errors in the fit probably understate the actual uncertainty in the intensity. From random inspection of several lines at the 10þ5 H intensity level, it is likely that some lines may have uncertainties ranging upward to 100% in regions where the continuum level is rapidly undulating, complicated by scattered light artifacts from adjacent orders, or affected by strong telluric absorption. Final geocentric offsets and c(H ) values are presented in Table 2. Figure 2 presents a histogram showing the fraction of lines in each of the nebulae observed with MIKE for different intensity levels relative to H that we have determined to be real and that also appear in at least one of the other two PNe (the NGC 7027 spectrum is omitted here due to its different wavelength coverage and spectral resolution). The strongest emission lines are all detected in each of the nebulae. Even at intensities down to 10þ5 H , where S / N $ 10 typically, roughly 50% of the weak features identified in the individual PNe also appear in one of the other
7

See http://www-mpl.sri.com /software / DIATOM / DIATOM.html.


1268

SHARPEE ET AL.

Vol. 659

Fig. 2.-- Fraction of lines present in the spectrum of each PN that are also detected in another of the PNe. The numbers give the total number of detected lines in each flux interval.

Fig. 3.-- Cumulative number of observed lines exceeding a given flux level for recently published nebular spectra. We consider only those lines within the wavelength range 3510 Y 7470 8, which is covered by all the spectra.

nebulae, suggesting that they are true nebular lines. The present spectra are among the deepest emission spectra taken of nebulae, revealing some of the weakest lines yet observed. This can be seen in Figure 3, where the cumulative number of lines exceeding a given flux level relative to H is shown for several of the most extensive PN spectral studies published in the recent literature. 4. PLASMA DIAGNOSTICS AND ABUNDANCES Electron temperatures and densities are presented in Table 3. The IRAF NEBULAR package task temden (Shaw & Dufour 1995) was used to equate relative intensities of collisionally excited diagnostic lines to densities and temperatures by matching a diagnostic for each from ions of similar ionization energy and solving for self-consistent values. Each ion was modeled by a five or greater level atom for this purpose, with spontaneous emission coefficients and collision strengths for the five lowest levels drawn primarily from the compilation of Mendoza (1983). Although these are not the default values currently distributed with NEBULAR, this atomic data set yielded good agreement in temperature and density among diagnostics in our previous analysis of IC 418 and was utilized in the original NEBULAR release. Spontaneous emission coefficients from Froese Fischer & Tachiev (2004) and collision strengths from Wilson & Bell (2002) were used for N i and Cl ii, respectively. For the remaining energy levels and for the ions Cl iv and K v, t he at o m i c da t a us ed i n t he m o st r ec en t re l ea se of N E B U LA R were utilized. The departure of the [ K v] density diagnostic from other density diagnostic values, particularly in NGC 2440 and IC 4191, may indicate errors in these atomic data. However, s-process elemental abundances derived later using this diagnostic ( Kr+4/H+ and Xe+5/H+) were completely insensitive to density. Errors in diagnostic values were determined by selecting the extrema values from computations at all combinations of diagnostic line ratios plus and minus their (1 ) uncertainties, including an error estimate for the reddening correction. Indeterminate error limits, such as occurred when a ratio value exceeded the asymptotic limit or where the paired diagnostics failed to converge at a particular ratio value, are listed without a value. Balmer ( MIKE PNe) and Paschen ( NGC 7027) jump temperatures were calculated in the manner described by Zhang et al. (2004), and as reported by them for many PNe, they are lower

than those temperatures determined from the collisionally excited lines. Ionic abundances derived from both collisionally excited and nominal radiative recombination lines are presented in Table 4. The IRAF NEBULAR package tasks ionic and abundance were used to make computations from the collisionally excited lines listed in Table 3 using temperatures and densities derived from the diagnostics with the closest ionization potential to each ion. Where no diagnostics where clearly appropriate, averaged values for temperature and density were used. Uncertainties were computed in the same manner as the diagnostic values, by calculation of the abundance at every combination of temperature, density, and line intensity ratio plus or minus their (1 ) uncertainties (where available) and selecting the extrema values. Uncertainties were not calculated for abundance computed with averaged diagnostics values. For the recombination lines, effective recombination coefficients were combined with line intensities to make abundance determinations, in the manner described by Sharpee et al. (2004) with temperature and densities again drawn from diagnostics nearest in ionization potential to each ion. For He+/H+, the coefficients of Smits (1996) or Benjamin et al. (1999) were used to determine an average abundance from the k4923, k5876, and k6678 lines, with corrections for collisional excitation (case A for triplets, case B for singlets) taken from Kingdon & Ferland (1995), while coefficients from Storey & Hummer (1995) were used to calculate average He+2/H + abundances from various He ii lines. To determine C +2,N +2,O+2,and Ne+2 abundances relative to H +, recent effective recombination coefficients (Storey 1994; Kisielius et al. 1998; Davey et al. 2000; Kisielius & Storey 2002) for the strongest observable multiplets were used to calculate abundances. As seen in Table 4, O+2/H + and Ne+2/H + abundances deduced from the collisionally excited lines are systematically lower than those derived from radiative recombination lines, as is generally the case in PNe ( Liu 2006; Robertson-Tessi & Garnett 2005). 5. EMISSION-LINE IDENTIFICATION The emission-line identification code EMILI (Sharpee et al. 2003) was used to make the majority of emission-line identifications. EMILI creates models of the ionization energy Y dependent velocity field and ionization level of a PN or H ii region from user-supplied empirical data. These models are used by EMILI to select from a large atomic transition database all transitions within 5 times an observed line's wavelength measurement error


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES
TABLE 3 Elec tr on Temperatur es and Densities Diagnostic NGC 2440 Density (cmþ3) [N i] k5198/k5200 ......................................... [S ii] k6716/k6731 ......................................... [O ii] k3726/k3729a ....................................... [O ii] k3726/k3729b ....................................... [Cl iii] k5517/k5537 ....................................... [Ar iv] k4711/k4741c ..................................... [Ar iv] k4711/k4741d ..................................... [Ar iv] k4711/k4741e ..................................... [K v] k4123/k4163 ........................................ 3300×2700 þ1300 3100×1200 þ700 3300×1100 þ800 3700 4700×1100 þ900 6200×1600 þ1300 5900×1600 þ1300 6300×1600 þ1300 43000×19000 þ13000
: 21000:þ:15000 11000 11000×9000 þ4000 11000 8500×2100 þ1600 9300×2100 þ1600 7400×2200 þ1500 9900×2200 þ1700 .. . : 12000:þ:7000 ... 9000×6000 þ3000 ... 12000×3000 þ2000 11800×2300 þ1900 11000×2400 þ1800 13000×3000 þ2000 41000×17000 þ11000 : 15000:þ:8000 ... ... ... 47400×1900 þ1800 ... 49200×1300 þ1200 ... ...

1269

IC 2501

IC 4191

NGC 7027

Temperature ( K ) [O i] (k6300+k6464)/k5577........................... [S ii] (k6716+k6731)/(k4068+k4076) ........... [O ii] (k3726+k3729)/(k7320+k7330) ........... [N ii] (k6548+k6583)/k5755 ......................... [Ar iii] (k7136+k7751)/k5192 ....................... [O iii] (k4959+k5007 )/k4363 ........................ [Cl iv] (k7531+k8046 )/k5323 ....................... [Ne iii] (k3869+k3968)/k3343 ...................... [Ar v] (k6435+k7005 )/k4625 ........................ Balmer/ Paschen ( NGC 7027 ) jump .............
a b c d e

9400 ô 300 15000×6000 þ4000 16000 ×700 12600þ600 13100×700 þ600 14700×600 þ500 12900×1000 þ700 15500×500 þ400 16400×1500 þ1100 11000

6900×300 þ200 12000 13000 10800×900 þ1100 9400×600 þ500 9500×300 þ200 6100×900 þ500 11000 ô 200 .. . 7000

7900×500 þ300 ... ... 12500 ô 1100 13000×800 þ600 9900 ô 300 8600×500 þ400 11600×300 þ200 11600×1400 þ900 8000

11300×300 þ200 ... ... ... 12900 ô 200 ... 13700 ô 200 ... ... 8000

Vs. Vs. Vs. Vs. Vs.

[O ii] temperature diagnostic. [ N ii] temperature diagnostic. [O iii] temperature diagnostic. [Cl iv] temperature diagnostic. [ Ne iii] temperature diagnostic.

TABLE 4 Ionic Abundances X+i/H
+

NGC 2440

IC 2501

IC 4191

NGC 7027

Collisionally Excited Lines N0/H+ ............ N+/H+ ............ O0/H+ ............ O+/H+ ............ O+2/H+ .......... Ne+2/H+ ......... S+/H+............. S+2/H+ ........... Cl+/H+ ........... Cl+2/H+.......... Cl+3/H+.......... Ar+2/H+ ......... Ar+3/H+ ......... Ar+4/H+ ......... K+4/H+ .......... 7:25×0::21 þ0 16 7:85×0::05 þ0 06 7:76×0::06 þ0 05 7.24 8:26×0::04 þ0 05 7:41×0::03 þ0 04 5.5 ô 0.3 6:23×0::07 þ0 08 .. . 4.70 ô 0.06 4:69×0::06 þ0 07 6.13 ô 0.04 5.70 ô 0.03 5:24×0::06 þ0 08 4:59×0::09 þ0 11 6:9:þ:0::5 ×0 16 7:02þ0::10 7:68:þ:0::14 7.30 ×0 04 8:59þ0::05 7.58 ô 0.03 5.93 ×0 11 6:58þ0::12 3.43 ×0 09 4:92þ0::10 ×0 15 4:66þ0::21 6.27 ô 0.07 4.75 ô 0.03 ... ... 6:4:þ:0::4 ×0 12 6:74þ0::09 7:34:þ:0::11 7:5×0::3 þ0 2 8.79 ô 0.05 ×0 03 7:79þ0::04 5.67 ×0 07 6:08þ0::09 3.18 ×0 07 4:56þ0::08 ×0 06 5:23þ0::07 ×0 05 5:70þ0::06 ×0 03 5:74þ0::04 ×0 09 4:57þ0::12 ×0 14 3:77þ0::17 6:1:þ:0::3 7.19 7:28:þ:0::04 7.68 8.44 .. . 5.80 6.32 ô 0.02 3.94 4.76 ô 0.02 4.72 ô 0.02 6.09 ô 0.02 5.77 ô 0.02 5.60 .. .

Recombination Lines He+/H+ .......... He+2/H+ ......... C+2/H+ ........... N+2/H+ ........... O+2/H+ .......... Ne+2/H+ ......... 10.86 10.72 8.65 8.43 8.51 8.08 11.04 10.61 8.91 8.04 8.69 8.26 11.00 10.08 8.46 8.28 9.08 8.76 10.81 10.60 8.84 7.89 8.46 .. .

and to compute relative intensities for emission lines corresponding to those transitions. Those transitions predicted to produce emission-line intensities within 3 orders of magnitude of the highest value among all transitions initially selected are then subjected to a test for the presence of lines from the same LS-coupled multiplets ( but not all lines from the same upper level ) at expected wavelengths and relative intensities. Potential identifications are then assigned a numeric likelihood parameter based on their wavelength agreement with the observed line, strength of the predicted emission line, and results of the multiplet check, and they are ranked and presented. Many emission lines from Z > 30 elemental ions were observed in NGC 7027 by PB94. To place these lines on an equal footing for identification purposes with those arising from more abundant lighter elements, EMILI used the Atomic Line List version 2.05 of P. van Hoof 8 as its reference database, which extends to Z ¼ 36 ( Kr). The latest experimental determinations available in the literature for ground electron configuration energy levels of Z > 36 ions were then added to this database. Transition wavelengths for the mostly optically forbidden transitions among these levels were constructed by differencing the level energies. Sources for all Z > 30 ion energy levels and associated atomic data used in subsequent analysis are provided in Table 5. All ions with certain or probable line identifications in NGC 7027, as rated by PB94, most of those with more tenuous identifications, and a handful of other ions they suggested to be worthy of future consideration at a higher spectral resolution were incorporated into the database. However, ions with level uncertainties perceived or explicitly stated to be greater than 1.0 cmþ1 in their source literature were excluded.
8

Note.-- In units of 12 × log ( X / H ).

See http://www.pa.uky.edu /~ peter/newpage /.


1270

SHARPEE ET AL.
TABLE 5 Atomic Data for Z > 30 Ions Ion Br iii ........................... Br iv ........................... Kr iii ........................... Kr iv ........................... Kr v ............................ Rb iv........................... Rb v............................ Sr ii ............................. Sr iv ............................ Sr v ............................. Sr vi ............................ Y v ............................. Y vi ............................ Zr vii ........................... Te iii ........................... I iii .............................. I v ............................... Xe iii ........................... Xe iv ........................... Xe v............................ Xe vi ........................... Cs v ............................ Cs vi ........................... Ba ii ............................ Ba iv ........................... Ba v ............................ Ba vii .......................... Ba viii ......................... Pb ii ............................
a

Levels ´ ALL 2.05a, Biemont & Hansen (1986a) (2, 3) ALL 2.05 ALL 2.05 ALL 2.05 ALL 2.05 Persson & Wahlstrom (1985 ) Persson & Petterson (1984) Moore (1952) Hansen & Persson (1976 ) Hansen & Persson (1974) Persson & Petterson (1984) Reader & Epstein (1972) Persson & Reader (1986 ) Reader & Acquista (1976) Joshi et al. (1992) Tauheed & Joshi (1993a) Kaufman et al. (1988) Persson et al. (1988) Tauheed et al. (1993) Gallard et al. (1999) Churilov & Joshi (2000) Tauheed & Joshi (1993b) Tauheed et al. (1991) ´ Karlsson & Litzen (1999) Sansonetti et al. (1993) Reader (1983) Tauheed & Joshi (1992) Churilov et al. (2002) Moore (1958)

Transition Probabilities ´ Biemont & Hansen (1986a) ´ Biemont & Hansen (1986a) ´ Biemont & Hansen (1986b) ´ Biemont & Hansen (1986a) ´ Biemont & Hansen (1986a) ´ Biemont & Hansen (1986b) ´ Biemont & Hansen (1986a) Barge et al. (1998) (5 Y 1, 5 Y 2) ´ Biemont et al. (1988) ´ Biemont & Hansen (1986b) ´ Biemont & Hansen (1986a) ´ Biemont et al. (1988) ´ Biemont & Hansen (1986b) ´ Biemont & Hansen (1986b) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) Klose et al. (2002) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) ´ Biemont et al. (1995) Safronova et al. (2005)

Collision Strengths ... ... (1997 ) (1997 ) (1997 ) ... ... ... ... ... ... ... ... ... ... ... ... & Butler & Butler ... & Butler ... ... & Butler & Butler ... ... ... ...

Schoning ¨ Schoning ¨ Schoning ¨

Schoning ¨ Schoning ¨ Schoning ¨

(1998) (1998) (1998)

Schoning ¨ Schoning ¨

(1998) (1998)

Atomic Line List version 2.05.

EMILI was run against each nebula's set of observed emissionline wavelengths and intensities. Electron temperature and density values were provided to EMILI by the results of standard plasma diagnostics, derived from strong collisionally excited lines with certain identifications. Construction of the empirical kinetic and ionization models were also drawn from those same

lines. The EMILI reference elemental abundances were set to the solar values of Lodders (2003). EMILI was also run against the IC 418 line list of Sharpee et al. (2003) to determine if some of its remaining unidentified lines could be identified with an expanded set of transitions and to also act as a foil for the mostly higher excitation PNe considered in the present sample. The line

TABLE 6 Emissi on-Line Identi fi cations and Inte nsities kcorr obs (8) FWHM (km sþ1)

F (H ¼ 100)

I (H ¼ 100) IC 2501

S/ N

Line Identification

Notes

a

3315.23....................... 3323.75....................... 3327.17....................... 3328.69....................... 3334.83....................... 3340.81....................... 3342.64....................... 3353.22....................... 7507.48....................... 7509.97....................... 7512.69....................... 7514.36....................... 7515.77....................... 7530.42....................... 7534.70.......................

23.73 19.16 27.98 19.13 18.57 13.75 45.56 33.40 22.85 57.84 52.14 36.15 22.81 29.59 30.21

0.0088 0.0116 0.0087 0.0056 0.0529 0.0188 0.0527 0.0071 0.0048 0.0033 0.0044 0.0038 0.0173 0.5710 0.0022

0.0200 0.0260 0.0190 0.0130 0.1200 0.0420 0.1200 0.0160 0.0027 0.0019 0.0025 0.0022 0.0099 0.3300 0.0013

7.2 13.9 10.4 8.4 50.0 13.3 37.0 11.5 13.6 7.1 7.1 10.7 32.2 627.6 7.1

[ Mn iii] k3315.01 Ne ii k3323.73 Ne ii k3327.15 N ii k3328.72 Ne ii k3334.84 N ii k3340.87 [ Ne iii] k3342.54 [Cl iii] k3353.21 N i k7507.61 .. . He i k7513.33 C iii k7514.30 O iii k7715.99 [Cl iv] k7530.80 [ Xe iv] k7535.44

:?

:

Notes.--Table 6 is published in its entirety in the electronic edition of the Astrophysical Journal. A portion is shown here for guidance regarding its form and content. a : = uncertain identification; ? = uncertain feature; bl = line blend; ns = blend with night-sky line.


TABLE 7 Possible Krypton L ine Identifications NGC 2440
a

IC 2501
d

IC 4191 V (km sþ1) Mult IDI þ13.6 þ76.0 þ65.5 þ4.8 þ17.6 ... ... 2.2 19.6 10.1 þ89.2 4.5 6.2 þ10.7 4.6 2.6 21.0 þ8.2 þ60.8 44.0 7.7 þ14.7 10.8 þ16.7 þ22.6 ... .. . . .. 12.0 þ41.7 þ10.1 þ1.9 þ2.4 þ7.7 3.4 þ6.7 þ7.2 ... ... .. . ... ... ... 0/0 ... 0/0 4/1 4/0 ... ... 1/1 2/0 5/0 8/0 0/0 0/0 3/2 1/1 ... 6/0 ... 0/0 5/2 ... ... 3/0 ... ... ... ... ... 0/0 2/0 2/0 1/0 5/0 4/0 4/0 0/0 1/0 ... ... ... ... ... ... 6B: ... .. . 3A 9D ... ... 3Aö 9 6 .. . 4B 5D 4B 4Aö ... >9 ... .. . 4A ... ... 4A: ... ... ... ... .. . 5D 8 5Dö 4A 4A 5D 4A 5D 5D ... ... ... ... ... ...

Transition [Kr iii] 4p4 3P2 Y 4p4 1D2 k6826.70 ............................................ o C i 3p 1P1 Y 4d 1D2 ( V21) k6828.12 ..................................... He i 3s 3S1 Y 16p 3P o k6827.88 .............................................. [Fe iv] 3d 5 4P1/2 Y 3d 5 2D53/2 k6826.50 ................................ o Fe ii] c4F7/2 Y y2G9/2 k6826.79............................................... SKY: OH 7 Y 2 R1(3.5) k6827.46........................................... SKY: OH 7 Y 2 R1(4.5) k6828.47........................................... o o [Kr iv] 4p3 4S3/2 Y 4p3 2D5/2 k5346.02 ....................................... o S ii 4s0 2D3/2 Y 4p0 2F5/2 ( V38) k5345.71 ............................... o C iii 2s4d 3D2 Y 2p3d 3P1 ( V13.01) k5345.88 ...................... [Fe ii] a4F3/2 Y b4P3/2 k5347.65 ............................................. Mg i 4s 3S1 Y 9p 3P o k5345.98............................................... Fe ii f4D7/2 Y 4f 2½3o/2 k5345.95 ............................................ 7 o Si ii 6p 2P3/2 Y 4p0 2P3/2 k5346.25 .......................................... o o [Kr iv] 4p3 4S3/2 Y 4p3 2D3/2 k5867.74 ....................................... Al ii 4d 3D Y 6f 3Fo (V41) k5867.64, k5867.78, k5867.89 ....... o Si ii 4s0 4P3/2 Y 4p0 4P1/2 ( V48) k5867.42 .............................. Ni ii 4f 2½4o/2 Y 7d 4H9/2 k5867.99 ........................................ 7 He ii 5 Y 29 k5869.02 ............................................................. [Cr iii] 3d 4 5D3 Y 3d 4 3H5 k5866.97....................................... Ni ii 4f 2½5o /2 Y 7d 2G9/2 k5867.68 ....................................... 11 o o [Kr iv] 4p3 2D3/2 Y 4p3 2P3/2 k6107.8 ......................................... [Fe ii] a6D9/2 Y a2G7/2 k6107.28 ............................................ o Ca i 4s7d 3D2 Y 3d20p 3P1 k6107.84 ..................................... Cr ii c4F5/2 Y 4p 4F1o/2 k6107.96 ............................................. o o [Kr iv] 4p3 2D5/2 Y 4p3 2P3/2 k6798.4 ......................................... o C ii 3s0 4P3/2 Y 3p0 4D1/2 ( V14) k6798.10 .............................. o Ca i 3d5s 1D2 Y 3d13p 1P1 k6798.28 ..................................... [Kr v] 4p2 3P1 Y 4p2 1D2 k6256.06............................................. o C ii 4p 2P1/2 Y 5d 2D3/2 ( V10.03) k6257.18 ........................... o C ii 3d0 2D3/2 Y 4p0 2P1/2 ( V38.03) k6256.52 ....................... 4o S ii 5p S3/2 Y 7s 4P5/2 k6256.35 ............................................. o Fe i a3H4 Y z3G4 ( V169) k6256.36 ....................................... 03 03 o O i 3p F4 Y 4d G3 ( V50) k6256.47................................... o C ii 4p0 4D5/2 Y 5d0 4D7/2 k6256.24......................................... 1 o Ca i 4s9s S0 Y 3d25p 1P1 k6256.45 ...................................... [Co ii] a3D2 Y a1F3 k6256.46................................................. SKY: OH 9 Y 3 Q2(0.5) k6256.94 .......................................... SKY: OH 9 Y 3 Q1(1.5) k6257.96 .......................................... [Kr v] 4p2 3P2 Y 4p2 1D2 k8243.39............................................. o N i 3s 4P5/2 Y 3p 4P3/2 ( V2) k8242.39 ................................... H i 3 Y 43 k8243.69 ................................................................ O iii 5g G 2[9/2]o Y 6h H 2[11/2] k8244.10 ............................

ko, I/I

H

V b (km sþ1) Multc IDI 1.8 þ60.6 þ50.1 10.5 þ2.2 . .. þ9.2 1.1 18.5 9.0 þ90.3 3.4 5.1 þ11.8 þ3.1 2.0 16.4 þ15.8 þ68.5 36.3 0.0 þ5.9 19.6 þ7.9 þ13.8 þ17.2 þ4.0 þ11.9 12.0 þ41.7 þ10.1 þ1.9 þ2.4 þ7.7 3.4 þ6.7 þ7.2 . .. þ12.4 ... . .. . .. . .. NGC 7027 V (km sþ1) Mult 9.2 þ53.2 þ42.6 18.0 5.3 12.7 . .. þ1.7 15.7 6.2 þ93.2 0.6 2.2 þ14.6 .. . .. . 0/0 4/0 4/0 .. . ... 1/1 2/0 5/0 .. . 0/0 0/0 3/0 0/0 .. . 0/0 4/0 4/0 ... .. . 1/1 2/0 5/0 8/0 0/0 0/0 3/0 1/1 .. . 6/0 .. . .. . 5/0 .. . .. . 3/0 .. . .. . 2/0 7/2 0/0 0/0 2/0 2/0 1/0 5/0 4/0 4/0 0/0 1/0 ... .. . . .. ... ... ...

ko, I/IH



V (km sþ1) Mult IDI 5.3 þ57.1 þ46.6 14.1 14.1 þ4.8 ... 2.8 20.2 10.7 þ88.7 5.1 6.7 þ10.1 4.1 2.0 20.5 þ8.7 þ61.4 43.5 7.2 ... ... ... ... þ11.0 2.2 þ5.7 24.5 þ29.2 2.4 10.5 10.1 4.8 15.8 5.8 5.3 5.3 ... ... ... ... ... IC 418 V (km sþ1) Mult IDI 7.5 þ54.9 þ44.4 16.3 3.5 ... þ1.8 þ4.5 12.9 3.4 þ96.0 þ2.2 þ0.6 þ17.4 0/0 .. . 0/0 4/0 4/0 .. . ... 1/1 2/0 5/0 8/1 0/0 0/0 3/1 1/1 .. . 6/0 .. . 0/0 5/1 .. . ... .. . .. . .. . 2/0 7/5 .. . 0/0 2/0 2/1 1/0 5/0 4/0 4/0 .. . 1/0 .. . ... .. . ... ... ... 5Abl . .. .. . 5A 6C . ..bl ... 3Aö 9 6D .. . 4B 5C 6D 4Aö . .. .. . . .. .. . 6B . .. ... . .. . .. . .. 7D 1Aö . .. 7bl 8 2Abl 6 5D 5D 7 . .. 4B . ..bl ... . .. ... ... ...

ko, I/IH



6826.74 1.0(þ5) .. . .. . .. . .. . .. . 5346.04 3.5(þ4) .. . .. . .. . .. . .. . 5867.68 4.6(þ4) .. . .. . .. . .. . .. . 6107.68? 5.6(þ6) .. . .. . 6798.01? 2.8(þ5) 1.3(þ5)y 6256.31 4.4(þ5) 1.5(þ5)y .. . .. . .. . .. . .. . .. . .. . .. . OUT .. . .. . .. .

5A . .. .. . 5A 6C ... . ..ö 3Aö 9 7 .. . 4B 4B 7 3Aö . .. 9 . .. . .. 7B . .. . ..: 6A: . .. . .. 8bl 1Abl 6 5Dbl 8 5D 4A 4A 4A 5D 5D 5D ... . ..bl .. . ... ... ...

6826.82 4.6(þ5) 2.5(þ5)y ... ... ... ... 5346.07 6.1(þ5) ... ... ... ... ... 5867.82 8.5(þ5) ... ... ... ... ... ... [4.3(þ6)] ... ... 6798.15 1.4(þ5) ... 6256.57 5.2(þ5) 8.0(þ6)y ... ... ... ... ... ... ... ... OUT ... ... ...

6826.39? 1.5(þ5) ... .. . .. . .. . .. . 5346.06 2.4(þ4) .. . ... .. . .. . .. . 5867.83 1.9(þ4) ... ... ... .. . ... 6107.50 1.9(þ5) ... ... .. . [8.1(þ6)] ... 6256.31 1.5(þ5) .. . .. . .. . .. . .. . ... .. . .. . .. . OUT .. . .. . .. .

Transition [Kr iii] 4p4 3P2 Y 4p4 1D2 k6826.70 ............................................ o C i 3p 1P1 Y 4d 1D2 ( V21) k6828.12 ..................................... 3 3o He i 3s S1 Y 16p P k6827.88 .............................................. [Fe iv] 3d 5 4P1/2 Y 3d 5 2D53/2 k6826.50 ................................ o Fe ii] c4F7/2 Y y2G9/2 k6826.79............................................... SKY: OH 7 Y 2 R1(3.5) k6827.46........................................... SKY: OH 7 Y 2 R1(4.5) k6828.47........................................... o o [Kr iv] 4p3 4S3/2 Y 4p3 2D5/2 k5346.02 ....................................... 02 02 o S ii 4s D3/2 Y 4p F5/2 ( V38) k5345.71 ............................... o C iii 2s4d 3D2 Y 2p3d 3P1 ( V13.01) k5345.88 ...................... [Fe ii] a4F3/2 Y b4P3/2 k5347.65 ............................................. Mg i 4s 3S1 Y 9p 3P o k5345.98............................................... Fe ii f4D7/2 Y 4f 2½3o/2 k5345.95 ............................................ 7 o Si ii 6p 2P3/2 Y 4p0 2P3/2 k5346.25 ..........................................

ko, I/I

H

IDI . ..bl . .. .. .bl 7A 7A . ..bl ... 1Aö 9 6 . .. 2B 3C 8

ko, I/IH



Ref.

Notes
e,f

6826.91 5.7(þ4) 4.6(þ4)y .. . .. . .. . .. . 5345.99 1.9(þ3) .. . .. . .. . .. . .. .

6826.87 3.3(þ4) 3.2(þ4)y ... ... ... ... 5345.94 3.5(þ5) ... ... ... ... ...

0/0 5Abl .. . . .. B95, Z05 .. . . .. B95 4/0 7C 4/0 6B . . . . . .bl .. . . .. 1/1 3Aö 2/0 9 H95, P04 5/0 5 Z05 .. . . .. ZL02 0/0 5 0/0 3A 3/0 8

f


1272

SHARPEE ET AL.
TABL E 7-- Continued NGC 7027 V (km sþ1) Mult IDI þ1.5 3.6 14.8 þ14.3 þ67.0 37.8 1.5 1.5 27.0 þ0.5 þ6.4 þ1.8 11.5 3.5 15.3 þ38.4 þ6.7 1.4 1.0 þ4.3 6.7 þ3.3 þ3.8 þ34.0 . .. 34.2 70.9 23.3 8.4 IC 418 V (km sþ1) þ0.5 4.6 þ13.3 þ13.3 þ66.0 38.9 2.6 . .. .. . . .. . .. . .. .. . . .. . .. .. . . .. . .. . .. . .. . .. . .. . .. . .. . .. 11.3 47.7 0.4 þ14.6

Transition
o o [Kr iv] 4p3 4S3/2 Y 4p3 2D3/2 k5867.74 ..................................... Al ii 4d 3D Y 6f 3F o ( V41) k5867.64, k5867.78, k5867.89..... o Si ii 4s0 4P3/2 Y 4p0 4P1/2 ( V48) k5867.42 ............................ Ni ii 4f 2½4o/2 Y 7d 4H9/2 k5867.99 ...................................... 7 He ii 5 Y 29 k5869.02 ........................................................... 45 [Cr iii] 3d D3 Y 3d 4 3H5 k5866.97..................................... Ni ii 4f 2½5o /2 Y 7d 2G9/2 k5867.68 ..................................... 11 o o [Kr iv] 4p3 2D3/2 Y 4p3 2P3/2 k6107.8 ...................................... [Fe ii] a6D9/2 Y a2G7/2 k6107.28 .......................................... o Ca i 4s7d 3D2 Y 3d20p 3P1 k6107.84................................... 4o Cr ii c4F5/2 Y 4p F1/2 k6107.96 ........................................... o o [Kr iv] 4p3 2D5/2 Y 4p3 2P3/2 k6798.4 ...................................... 04 o 04 C ii 3s P3/2 Y 3p D1/2 ( V14) k6798.10 ............................ o Ca i 3d5s 1D2 Y 3d13p 1P1 k6798.28................................... 23 21 [Kr v] 4p P1 Y 4p D2 k6256.06 .......................................... o C ii 4p 2P1/2 Y 5d 2D3/2 ( V10.03) k6257.18 ......................... 02 o C ii 3d D3/2 Y 4p0 2P1/2 ( V38.03) k6256.52 ..................... o S ii 5p 4S3/2 Y 7s 4P5/2 k6256.35 ........................................... Fe i a3H4 Y z3Go ( V169) k6256.36 ..................................... 4 O i 3p0 3F4 Y 4d 0 3Go ( V50) k6256.47 ................................ 3 o C ii 4p0 4D5/2 Y 5d 0 4D7/2 k6256.24 ...................................... 1 o Ca i 4s9s S0 Y 3d25p 1P1 k6256.45 .................................... [Co ii] a3D2 Y a1F3 k6256.46............................................... SKY: OH 9 Y 3 Q2(0.5) k6256.94 ........................................ SKY: OH 9 Y 3 Q1(1.5) k6257.96 ........................................ [Kr v] 4p2 3P2 Y 4p2 1D2 k8243.39 .......................................... o N i 3s 4P5/2 Y 3p 4P3/2 ( V2) k8242.39 ................................. H i 3 Y 43 k8243.69 .............................................................. O iii 5g G 2[9/2]o Y 6h H 2[11/2] k8244.10 ..........................

ko, I/I

H

ko, I/IH



Mult 1/1 4/0 6/0 0/0 .. . 5/0 0/0 .. . . .. .. . .. . .. . . .. .. . .. . . .. .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . 0/0 .. .

IDI 2Aö 6 8 7 . .. .. . 3B . .. .. . . .. . .. . .. .. . . .. . .. .. . . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. 2Aö . ..

Ref.

Notes

5867.71 2.6(þ3) .. . .. . .. . .. . .. . 6107.83 7.2(þ5) .. . .. . 6798.36 4.1(þ5) 1.6(þ5)y 6256.38 1.7(þ4) 1.2(þ4)y .. . .. . .. . .. . .. . .. . .. . .. . 8244.33? 1.4(þ4) .. . .. .

1/1 1Aö 5867.73 5/0 5C 3.5(þ5) 6/0 8 .. . 0/0 7 .. . .. . ... ... 5/0 ... ... 0/0 4B .. . 3/1 5Aö ... 3/0 ... [1.4(þ5)] 1/0 5A ... 9/0 5A ... 3/1 4Abl ... 7/1 7bl [2.1(þ5)] 0/0 5B ... 1/0 7bl ... .. . ...bl [8.4(þ5)] 2/0 6bl ... 1/0 5D ... 5/0 4A ... 4/0 5D ... 4/0 7 ... 0/0 4A ... 1/0 4A ... .. . ...bl ... .. . ... ... .. . ...: 8243.70 .. . ... 4.2(þ4) 0/0 5A .. . 0/0 7B .. .

B00, S03 H01 E04 Z05

g

B95

S03 Z05

h

B95 S03 B95

a Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ``OUT'' means not in observed range; (3) ``?'' denotes an uncertain feature. Intensity: (1) numbers in parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features. b Observed (ko ) þ transition wavelength ( km sþ1). c EMILI multiplet check statistics: number expected /number observed. d EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ``bl'' ( blend). Definition of IDI given in x 6. e Unidentified line in IC 418 ( Baldwin et al. 2000). f Unidentified line in Orion Nebula (Sharpee et al. 2003). g NGC 2440, IC 4191, NGC 7027: identified as a separate line. h NGC 7027: affected strongly by telluric absorption. References.--( B95) PN NGC 7027 ( Baluteau et al. 1995); ( B00) Orion Nebula ( Baldwin et al. 2000); ( E04) Orion Nebula ( Esteban et al. 2004); ( H95) PN NGC 6886 ( Hyung et al. 1995); ( H01) PN IC 5217 ( Hyung et al. 2001); ( P04) PN NGC 5315 ( Peimbert et al. 2004); (S03) PN IC 418 (Sharpee et al. 2003); ( Z05) PN NGC 7027 ( Zhang et al. 2005); ( ZL02) PN Mz 3 ( Zhang & Liu 2002).

identification lists produced by EMILI were then visually inspected, and the EMILI-preferred assignments were compared to identifications in the literature for those same lines in PNe and H ii region spectra of similar depth, spectral resolution, and level of ionization. The entries in our final line list are in many cases taken from among the highest ranked EMILI-suggested identifications. A segment of a final line list is given in Table 6; the full line list is available in the electronic version of this manuscript. This table lists all observed lines with their characteristics, viz., observed wavelength in the nebula rest frame, identifications, FWHMs, S/ N, observed and reddening-corrected intensities relativetoH , and their most likely identifications. Lines with uncertain identifications are denoted by a colon, likely blends are marked with a ``bl'' or ``ns'' if blended with a telluric emission feature, and uncertain features are marked with a question mark. Only the perceived strongest component of a blend is listed in the table. Lines without obvious identifications are listed here without an identification.

6. Z > 30 LINE IDENTIFICATIONS PB94 detected 25 emission lines from several Z > 30 ions in their optical spectra of NGC 7027, including Se, Br, Kr, Rb, Sr, and Y from the fourth row of the periodic table, Xe and Ba from the fifth row, and Pb from the sixth. Eighteen of these detections were considered ``certain'' or ``probable'' and seven considered ``possible.'' Lines from 13 additional Z > 30 ions were also either tentatively identified or proposed as future targets for spectra with greater spectral resolution and better S/ N. Given the depth and high resolution of the spectra considered here, confirmation of the PB94 identifications in multiple PNe was sought, as were additional lines belonging to other Z > 30 ions. The use of EMILI allows prospective Z > 30 transitions and weaker transitions of more abundant lighter elements to be treated equally for purposes of emission-line identification. Collision strength and spontaneous emission coefficient calculations for these transitions allow accurate predictions to be made of the relative intensities of lines arising from the same


TABLE 8 Possible Xenon Line Iden tificati on s NGC 2440
a

IC 2501 IDId ... 5Bö 7C 3A ... 5ö ... 4C 3A 3A 7ö 2A 6 7 5 3B ... ... 5A . .. . .. . .. . .. . ..ö . ..: ... 4A . .. V (km sþ1) Mult IDI þ6.7 þ1.0 þ34.9 þ7.2 2.1 þ12.1 þ94.6 0.0 3.2 þ3.7 þ27.5 þ4.4 þ15.5 þ42.2 þ3.6 þ1.6 ... ... ... ... ... ... ... ... ... ... ... ... IC 418 V (km sþ1) Mult IDI þ3.6 2.1 þ31.8 þ4.1 5.1 þ15.2 þ97.7 þ3.2 0.0 þ6.8 þ19.9 3.2 þ8.0 þ34.6 4.0 6.0 ... ... ... ... ... ... ... 0/0 0/0 3/0 0/0 0/0 1/1 .. . 7/0 0/0 7/0 1/1 5/0 0/0 .. . 8/0 0/0 ... .. . .. . ... ... ... ... 5 3A .. . 4C 3A 7ö . .. 5B 4A 8 6ö 4A 5 . .. 4A 4A ... . .. . .. ... ... ... ...
ö

IC 4191 V (km sþ1) Mult 0.0 5.6 þ28.2 þ0.5 8.7 þ11.6 þ94.1 0.5 3.7 þ3.2 þ26.7 þ3.6 þ14.7 þ41.4 þ2.8 þ0.8 5.9 19.5 þ49.2 þ46.6 1.7 þ3.4 þ5.5 . .. 2.8 120.8 21.5 12.2 .. . 0/0 3/0 0/0 .. . 1/0 .. . 7/0 0/0 7/0 1/0 1/0 0/0 .. . 8/0 0/0 0/0 6/0 0/0 .. . 0/0 0/0 0/0 .. . .. . .. . 1/0 .. .

Transition [Xe iii] 5p P2 Y 5p D2 k5846.77 .............................. He ii 5 Y 31 k5846.66 ................................................ [Fe ii] a2G7/2 Y c2G9/2 k5847.32 ............................... Fe ii 4d e4F5/2 Y 4f 2½3o/2 k5846.78.......................... 5 o Fe i x5P1 Y 6d 2[3/2]1 k5846.60 ............................... o o [Xe iv] 5p3 4S3/2 Y 5p3 2D5/2 k5709.20 ......................... o N ii 3s 3P2 Y 3p 3D2 ( V3) k5710.77 ......................... [Fe i] a5D3 Y a5P1 k5708.97 ..................................... Si i 4p 3D2 Y 18s (3/2, 1/2)o k5708.91 ...................... 1 o Fe ii] y4P3/2 Y e6D1/2 k5709.04.................................. 34 o 32 o [Xe iv] 5p S3/2 Y 5p D3/2 k7535.4 ........................... o Fe ii] b2F5/2 Y z4F5/2 ( V87) k7534.82 ....................... o N ii 5f.G 2[7/2]4 Y 10d 1F3 k7535.10 ....................... Ne i 3p 2[1/2]1 Y 3d 2[1/2]o ( V8.01) k7535.77 ........ 1 [Cr ii] b4D5/2 Y c4D5/2 k7534.80 ................................ Ne ii 3d0 2P3/2 Y 6f 2½3o/2 k7534.75 .......................... 5 [Xe v] 5p2 3P0 Y 5p2 3P2 k7076.8 ................................. o C i 3p 3D2 Y 4d 3D2 ( V26.01) k7076.48 .................. [Fe iii] 3d 6 3P41 Y 3d 6 1S40 k7078.10 ...................... [Ni ii] 4s 4F3/2 Y 4s 4P3/2 k7078.04 ........................... Ni ii 5d 4F7/2 Y 4f 2½3o/2 k7076.90 ............................ 7 o Ca i 4s15s 3S1 Y 3d22p 3P1 k7077.02 ....................... 1 o Ca i 4s14s S0 Y 3d17f 1P1 k7077.08 ........................ SKY: 3 Y 2 O2 b 1ô× Y X 3ôþ P12(11) k7078.58 ...... g g o o [Xe vi] 5p 2P1/2 Y 5p 2P3/2 k6408.9 .............................. He ii 5 Y 15 k6406.38 ................................................ [Fe iii] 3d 6 1S40 Y 3d 6 3P22 k6408.50 ...................... C iv 9 Y 17 k6408.70 .................................................
43 41

ko , I/I

H

V (km sþ1) þ2.1 3.6 þ30.3 þ2.6 6.7 þ13.1 þ95.6 þ1.1 2.1 þ4.7 þ27.9 þ4.8 þ15.9 þ42.6 þ4.0 þ2.0 9.3 22.9 þ45.8 þ20.3 28.0 22.9 20.8 0.4 þ9.8 108.1 8.9 þ0.5

b

Multc .. . 0/0 3/0 0/0 .. . 0/0 .. . 7/0 0/0 7/1 0/0 0/0 0/0 0/0 4/0 0/0 .. . .. . 0/0 ... ... ... ... .. . .. . .. . 1/0 ...

ko , I/IH



ko , I/I

H

IDI . .. 4Aö 7C 4A . .. 7ö . .. 4B 3A 5D 8ö 3B 5 . .. 4C 2A 4Aö 8 .. . . .. 4A 5C 5C . .. . ..: . .. 6A . ..

5846.73 3.2(þ4) .. . .. . .. . 5708.95? 1.2(þ5) .. . .. . .. . 7534.70 1.3(þ5) .. . .. . .. . .. . 7077.02 1.5(þ5) .. . .. . .. . .. . .. . .. . 6408.69? 2.8(þ5) .. . .. .

5846.64? 7.7(þ6) .. . .. . .. . 5708.97? 6.5(þ6) .. . .. . .. . 7534.71 1.6(þ5) .. . .. . .. . .. . .. . [5.5(þ5)] ... ... ... ... ... .. . .. . [2.4(þ6)] ... ...

.. . 0/0 3/0 0/0 .. . 1/0 .. . 7/1 0/0 7/0 .. . 1/0 0/0 .. . 8/0 0/0 ... .. . .. . .. . .. . .. . .. . ... ... .. . .. . .. .

. ..: 5846.77 3A 5.2(þ5) 7C .. . 4B .. . . .. .. . 5: 5708.98? . .. 1.2(þ5) 3A: .. . 3A .. . 5 .. . . ..ö 7534.73 4A 3.6(þ5) 6C .. . . .. .. . 6C .. . 5B .. . . . . 7076.94? . .. 3.9(þ6) . .. .. . . .. .. . . .. .. . . .. .. . . .. .. . ... ... . . . 6408.96? . .. 1.6(þ6) . .. .. . . .. .. .

NGC 7027 V (km sþ1) þ6.2 þ0.5 þ34.4 þ6.7 2.6 þ16.2 þ98.8 þ4.2 þ1.1 þ7.9 þ18.3 4.8 þ6.4 þ33.0 5.6 7.6 8.5 22.0 ... þ44.1 4.2 þ0.8 þ3.0

Transition [Xe iii] 5p P2 Y 5p D2 k5846.77 .............................. He ii 5 Y 31 k5846.66 ................................................ [Fe ii] a2G7/2 Y c2G9/2 k5847.32 ............................... Fe ii 4d e4F5/2 Y 4f 2½3o/2 k5846.78.......................... 5 o Fe i x5P1 Y 6d 2[3/2]1 k5846.60 ............................... o o [Xe iv] 5p3 4S3/2 Y 5p3 2D5/2 k5709.20 ......................... o N ii 3s 3P2 Y 3p 3D2 ( V3) k5710.77 ......................... [Fe i] a5D3 Y a5P1 k5708.97 ......................................... Si i 4p 3D2 Y 18s (3/2, 1/2)o k5708.91 ...................... 1 o Fe ii] y4P3/2 Y e6D1/2 k5709.04.................................. o o [Xe iv] 5p3 4S3/2 Y 5p3 2D3/2 k7535.4 ........................... o Fe ii] b2F5/2 Y z4F5/2 ( V87) k7534.82 ....................... o N ii 5f.G 2[7/2]4 Y 10d 1F3 k7535.10 ....................... Ne i 3p 2[1/2]1 Y 3d 2[1/2]o ( V8.01) k7535.77 ........ 1 [Cr ii] b4D5/2 Y c4D5/2 k7534.80 ................................ o Ne ii 3d0 2P3/2 Y 6f 2½35/2 k7534.75 .......................... [Xe v] 5p2 3P0 Y 5p2 3P2 k7076.8 ................................. o C i 3p 3D2 Y 4d 3D2 ( V26.01) k7076.48 .................. [Fe iii] 3d 6 3P41 Y 3d 6 1S40 k7078.10 ...................... [Ni ii] 4s 4F3/2 Y 4s 4P3/2 k7078.04 ........................... Ni ii 5d 4F7/2 Y 4f 2½3o/2 k7076.90 ............................ 7 o Ca i 4s15s 3S1 Y 3d22p 3P1 k7077.02 ....................... o Ca i 4s14s 1S0 Y 3d17f 1P1 k7077.08 ........................
43 41

ko , I/IH 5846.65 4.2(þ4) 1.4(þ4)y .. . .. . 5708.89 1.1(þ4) .. . .. . .. . 7534.94 1.7(þ4) .. . .. . .. . .. . 7077.00 2.0(þ5) ... .. . .. . .. . .. .

Mult

IDI

ko , I/IH



Ref.

Notes
e

.. . . . .bl 5846.70 0/0 1Abl 1.3(þ4) 3/0 .. . ... 0/0 4B .. . .. . ... .. . 1/1 7ö 5708.91? .. . ... 1.9(þ5) 7/0 5C .. . 0/0 3A .. . 7/1 6D .. . 1/1 6ö 7534.90 5/0 6 1.6(þ5) 0/0 3A .. . .. . ... .. . 8/0 6 .. . 0/0 5D .. . 0/0 6: .. . 6/0 8 [2.8(þ5)] þ46.6 0/0 ... .. . ... .. . 0/0 5C .. . 0/0 4A .. . 0/0 4A .. .

Z05

e

Many

f

g

H01 E04, P04

h

B95 .. . ZL02

ZL02


1274

SHARPEE ET AL.
TABL E 8-- Continued NGC 7027 V (km sþ1) Mult IDI ... þ13.6 104.4 5.1 þ4.2 ... ... ... 1/0 0/0 IC 418 V (km sþ1) Mult IDI . . . . . . . . . . . . . . . . . . . . .. .. .. .. .. .. .. .. .. .. . . . . .

Vol. 659

Transition SKY: 3 Y 2 O2 b 1ô× Y X 3ôþ P12(11) k7078.58 ...... g g o o [Xe vi] 5p 2P1/2 Y 5p 2P3/2 k6408.9 ............................ He ii 5 Y 15 k6406.38 .............................................. [Fe iii] 3d 6 1S40 Y 3d 6 3P22 k6408.50 .................... C iv 9 Y 17 k6408.70 ...............................................
a

ko , I/IH



ko , I/IH

Ref.

Notes

.. . 6408.61 2.0(þ4) .. . .. .

.. . ... .. .ö ... .. . [9.3(þ6)] 4A ... 5B ...

i

Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ``OUT'' means not in observed range; (3) ``?'' denotes an uncertain feature. Intensity: (1) numbers in parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features. b Observed (ko ) þ transition wavelength ( km sþ1). c EMILI multiplet check statistics: number expected /number observed. d EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ``bl'' ( blend ). Definition of IDI given in x 6. e Unidentified line in IC 418 ( Sharpee et al. 2003). f Identified as separate line in all spectra. g On edge of NGC 2440, IC 2501, IC 4191 spectra, where poorest wavelength calibration is expected. h Listed by Hyung et al. (2001) as ``unlikely ID.'' i IC 4191 and NGC 7027: identified as separate line. References.--( E04) Orion Nebula ( Esteban et al. 2004); ( H01) PN IC 5217 ( Hyung et al. 2001); ( P04) PN NGC 5315 ( Peimbert et al. 2004); ( Z05) PN NGC 7027 ( Zhang et al. 2005); ( ZL02) PN Mz 3 ( Zhang & Liu 2002).

ion, allowing identifications of their lines to be made with more confidence. The line lists for the present PN sample and IC 418 (Sharpee et al. 2003) were searched for the strongest expected Z > 30 ion transitions within their observed bandpasses. These primarily were the 3P1,2 Y 1D2 nebular transitions of ions with 4p2,4 and o o 5p2,4 valance electrons, the 4 S3/2 Y 2 D3/2;5/2 nebular transitions 3 3 o of ions with 4p and 5p valence electrons, and the 2 P1=2;3=2 Y 5 5 2o P1=2;3=2 fine-structure transitions for 4p, 4p , 5p, 5p , and 6p valance electron ions. Fine-structure 3P Y 3P transitions, when accessible in the visible, were also included. For the cases of o o Kr iv,Xe iv, and Br iv, auroral 2 D3/2;5/2Y 2 P1/2;3/2 and transauroral 4o 2o S3/2 Y P1/2;3/2 transitions were also considered. The permitted resonance lines of Ba ii and Sr ii were also searched for. Initially, all wavelength coincidences of 1 8 or less between an observed line and a transition wavelength were considered possible Z > 30 lines, regardless of the identifications recommended by EMILI for those lines. Tables 7 Y 10 present excerpts from the EMILI output for observed lines believed to represent the best cases for a Z > 30 transition as the actual identification for an observed line in at least one of the PNe in which that line appeared. For each observed line in each PN, the identifications of highest rank, as specified by the ``Identification Index'' or IDI, the EMILI figure of merit for quality of an identification (5 or less is considered a quality identification), and their multiplet search statistics (numbers expected / observed, hereafter ``multiplet statistics'') are included. The IDI value is followed by a letter A ! D if among the top four highest ranked transitions, with ``A'' being the highest. Additional identifications drawn from the literature and from the terrestrial nightglow ( prefaced with ``SKY'') are also included. Identifications that did not yield predicted line intensities within 3 orders of magnitude of the strongest value among all identifications, or that were outside the 5 search radius, do not have a calculated IDI value. The IDI value/rank is sometimes followed by a symbol, an asterisk indicating the most likely single identification, ``bl'' indicating components of a likely blend, and a colon indicating indeterminate alternate identifications. The reddening-corrected intensity of the feature attributable to the putative s-process identification, if corrected for any blending as described in succeeding sections (due to other identifications denoted with a ``bl''), is

presented with a dagger below its originally observed value. The limiting intensity, determined from the local minimum flux of line detection considered certain (S / N ¼ 7) or from imposition of artificial lines of S / N ¼ 7 at that wavelength, is presented in brackets for the case of s-process transitions without corresponding observed features. An ``OUT'' label indicates features residing outside the observed bandpass of a spectrum. Finally, features of dubious reality are denoted by a question mark following the observed wavelength. Figure 4 depicts continuum-subtracted spectra, with wavelengths shifted to the nebular rest frame, in the vicinity of those Kr, Xe, and Br lines for which a positive identification was made in at least one PN, and which were potentially observable in all four PNe of the present sample and in IC 418. Comments on identifications pertaining to individual Z > 30 elemental ion lines follow. 6.1. Kr Line Identifications PB94 noted that [ Kr iii] 4p4 3P2 Y 4p4 1D2 k6826.70 has long been observed in various novae and PNe but has seldom been identified as such. They identified [ Kr iii] as a strong line in NGC 7027, blended with weak C i (V21) k6828.12 and He i 3s 3 S Y 16p 3P o k6827.88 lines. As seen in Table 7 and Figure 5, the higher resolution of our present PN spectra cleanly separates both the He i and C i line from the putative [ Kr iii] line, except for NGC 7027, where He i contributes minimally to its red shoulder. However, Figure 5 shows that the R-branch head of the telluric nightglow OH Meinel 7 Y 3 band is a serious contaminant in this region. In NGC 7027, a broad O vi k1032 Raman scattering line at 6829.16 8, seen previously in NGC 7027 by Zhang et al. (2005), is also observed. The OH contribution was represented by a model of the band normalized in intensity to the uncontaminated R1(1.5) line at 6834.01 8. The He i 3s 3S Y 15p 3P o k6855.91 line profile was shifted and scaled by a factor of 0.83 to represent He i k6827.88 (Smits 1991; case B, for the most appropriate grid point: Te ¼ 10 4 K, ne ¼ 10 4 cmþ3). The profile of the companion O vi Raman scattering line observed at 7088 8, scaled upward by a factor of 4 to yield the best fit with the red wing of the putative [ Kr iv] feature, represented the k6829.16 line. These line profiles and the telluric model were subtracted from the continuum-normalized spectra to yield the residual features shown in Figure 5.


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES
TABLE 9 Possi ble Bromine Line Identifications NGC 2440
a

1275

IC 2501
d

IC 4191 V (km sþ1) Mult IDI þ32.3 2.0 þ2.9 þ24.2 þ3.2 1.4 þ8.7 þ5.5 þ8.2 þ10.5 ... ... ... ... ... ... 1/1 0/0 1/0 1/1 0/0 0/0 0/0 0/0 0/0 0/0 . .. . .. . .. . .. . .. . .. 6ö 2A 5D 6ö 2A 2A 3C 3C 3C 4 .. . .. . .. . .. . .. . .. .

Transition

ko , I/IH

V b (km sþ1) Multc IDI . .. .. . . .. þ19.2 1.8 6.4 þ3.7 þ0.5 þ3.2 þ5.5 . .. .. . . .. . .. .. . . .. NGC 7027 ... ... ... 1/0 0/0 0/0 0/0 0/0 0/0 0/0 ... ... ... ... ... ...

ko , I/IH



V (km sþ1) Mult IDI þ14.7 19.6 14.7 þ22.0 þ0.9 3.7 þ6.4 þ3.2 þ5.9 þ8.2 ... .. . ... ... .. . ... IC 418 V (km sþ1) Mult IDI þ25.9 8.3 3.4 . .. .. . . .. . .. . .. . .. . .. . .. .. . . .. . .. . .. .. . ... 0/0 1/0 ... ... ... ... ... ... ... ... ... ... ... ... ... ...ö 2A 4C ... .. . ... ... ... ... ... ... .. . ... ... ... .. . 1/1 0/0 1/0 1/1 0/0 0/0 0/0 0/0 0/0 0/0 ... ... ... ... ... ...

ko , I/IH

o o .. . [Br iii] 4p3 4S3/2 Y 4p3 2D5/2 k6131.0 .................................. C iii 7h 1,3H o Y 16g 1,3G k6130.30 .................................. [1.7(þ5)] .. . [Ni vi] 3d 5 4D5/2 Y 3d 5 2F17/2 k6130.40 ......................... o o [Br iii] 4p3 4S3/2 Y 4p3 2D3/2 k6556.4 .................................. 6555.98? Fe ii 4d 4P3/2 Y 4f 2½2o/2 k6555.94 .................................. 4.2(þ5) 5 O ii 4f F 2½4o/2 Y 6g 2[5]11/2 k6555.84 ............................ .. . 9 o .. . N ii 4f F 2[5/2]3 Y 6d 3D2 k6556.06 ................................ O ii 4f F 2½4o/2 Y 6g 2[5]9/2 k6555.99 .............................. .. . 9 O ii 4f 0 H 2½5o/2;11/2 Y 6g 0 2[6] kk6556.05, 6556.08....... .. . 9 O ii 4f F 2½4o/2 Y 6g 2[4]9/2 k6556.10 .............................. .. . 9 .. . [Br iv] 4p2 3P1 Y 4p2 1D2 k7368.1 ...................................... o C ii 3p0 2D5/2 Y 3d 0 2P3/2 k7370.00.................................. [2.6(þ5)] o O ii 4p 2S1/2 Y 5s 2P3/2 k7367.68 ..................................... .. . C v 7p 3Po Y 8d 3D 7367.60 ............................................ .. . OUT [Br iv] 4p2 3P2 Y 4p2 1D2 k9450.5 ...................................... o Fe i v3D2 Y 6d 2[5/2]3 k9450.95 ..................................... .. .

. .. .. . . .. 7: 3 3 2A 2A 2A 2A . .. .. . . .. . .. .. . . ..

6130.70 1.7(þ5) .. . 6555.92? 7.3(þ5) .. . .. . .. . .. . .. . .. . [4.4(þ5)] .. . .. . OUT .. .

4Aö 6130.34 5B 3.8(þ5) 7 .. . 6ö 6555.87? 2A 8.5(þ5) 2A .. . 2A .. . 3D .. . 3D .. . 4 .. . .. . ... .. . [3.5(þ5)] .. . ... .. . ... .. . OUT .. . ...

Transition
o o [Br iii] 4p3 4S3/2 Y 4p3 2D5/2 k6131.0 .................................. C iii 7h 1,3Ho Y 16g 1,3G k6130.30................................... [Ni vi] 3d 5 4D5/2 Y 3d 5 2F17/2 k6130.40 ......................... o o [Br iii] 4p3 4S3/2 Y 4p3 2D3/2 k6556.4 .................................. Fe ii 4d 4P3/2 Y 4f 2½2o/2 k6555.94 .................................. 5 O ii 4f F 2½4o/2 Y 6g 2[5]11/2 k6555.84 ............................ 9 o N ii 4f F 2[5/2]3 Y 6d 3D2 k6556.06 ................................ O ii 4f F 2½4o/2 Y 6g 2[5]9/2 k6555.99 .............................. 9 O ii 4f 0 H 2½5o/2;11/2 Y 6g 0 2[6] kk6556.05, 6556.08....... 9 O ii 4f F 2½4o/2 Y 6g 2[4]9/2 k6556.10 .............................. 9 [Br iv] 4p2 3P1 Y 4p2 1D2 k7368.1 ...................................... o C ii 3p0 2D5/2 Y 3d 0 2P3/2 k7370.00.................................. o O ii 4p 2S1/2 Y 5s 2P3/2 k7367.68 ..................................... C v 7p 3Po Y 8d 3D 7367.60 ............................................ [Br iv] 4p2 3P2 Y 4p2 1D2 k9450.5 ...................................... o Fe i v3D2 Y 6d 2[5/2]3 k9450.95 .....................................
a

V ko , I/IH (km sþ1) Mult 6130.32 1.1(þ4) 7.7(þ5)y 6555.93? 4.3(þ4) .. . .. . .. . .. . .. . 7367.62? 4.4(þ5) .. . .. . 9450.85? 7.5(þ5) þ33.3 1.0 þ3.9 þ21.5 þ0.5 4.1 þ5.9 þ2.7 þ5.5 þ7.8 þ19.5 þ96.9 þ2.4 0.8 11.1 þ3.2 .. . 0/0 1/0 1/0 0/0 0/0 0/0 0/0 0/0 0/0 1/0 .. . 1/0 0/0 1/0 0/0

IDI

ko , I/IH



Ref.

Notes
e,f

. . .bl 6130.47 2Abl 2.8(þ5) 4C .. . 8ö ... 2A [2.1(þ4)] 3B ... 3B ... 3B ... 4 ... 5 ... 9: ... . . . [4.9(þ5)] 3A ... 4B ... 8: ... 3A [5.3(þ5)]

e

B95

Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ``OUT'' means not in observed range; (3) ``?'' denotes an uncertain feature. Intensity: (1) numbers in parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features. b Observed (ko ) þ transition wavelength ( km sþ1). c EMILI multiplet check statistics: number expected /number observed. d EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ``bl'' ( blend). Definition of IDI given in x 6. e ´ Computed using energy levels from Biemont & Hansen (1986a); see text. f Unidentified line in IC 418 (Sharpee et al. 2003). References.--( B95) PN NGC 7027 ( Baluteau et al. 1995).

For all PNe except NGC 2440, where the observed line appears to be entirely comprised of the OH line, there appears to be a substantial residual feature that is coincident but, except for IC 4191, slightly to the red of the predicted wavelength for [ Kr iii] k6826.70. The EMILI results for the corresponding observed lines suggest that besides the C i and He i transitions mentioned earlier, the [ Fe iv] k6826.50 feature is probably the only other sensible alternative identification. However, given the absence of other multiplet members that should be observed with this line and the relative observed strengths of the likely strongest [ Fe iv] lines ( Rodriguez 2003), this is not a likely identification ex´ cept perhaps for IC 4191. Unfortunately, the companion [ Kr iii] 3P Y 1D line at 9902.3 8, predicted to be $18 times weaker by 1 2 ´ Biemont & Hansen (1986b), is only potentially observable in

NGC 7027, where a line listed at 9903.55 8 is much too strong and most likely associated with C ii 4f 2F o Y 5g 2G k9903.67. Nevertheless, we conclude that the [ Kr iii] k6826.70 identification is relatively certain in IC 418, IC 2501, and NGC 7027, uncertain in IC 4191, and that line is not observed in NGC 2440. Examination of Table 7 shows that both transitions of the [Kr iv] 4S o Y 2D o nebular multiplet kk5346.02, 5867.74 are clearly identified in every spectrum, appearing as the primary EMILI ranked identifications. For k5346.02, the [ Fe ii] k5347.65 identification is clearly too far away and the S ii ( V38) k5345.71 dielectronic transition is unlikely given the absence of other multiplet members. The Al ii k5867.8 identification for k5867.74, given by Baldwin et al. (2000) and Sharpee et al. (2003), is clearly superseded. With a theoretical intensity ratio I(k5346.02)/I(k5867.74)


TABLE 10 Po ssible Iden tifications Fr om Other Z > 30 Ions NGC 2440
a

IC 2501
d

IC 4191 V (km sþ1) 3.1 þ6.8 16.1 20.1 34.1 8.9 10.6 .. . .. . .. . .. . þ5.9 þ7.8 5.4 þ11.7 .. . .. . .. . .. . .. . .. . .. . .. .

Transition [Rb iv] 4p4 3P2 Y 4p4 1D2 k5759.55...................... He ii 5 Y 47 k5759.74 ........................................ [Fe ii] a2H9/2 Y c2D5/2 k5759.30 ....................... o o [Rb v] 4p3 4S3/2 Y 4p3 2D3/2 k5363.6 .................... [Ni iv] 3d 7 4F7/2 Y 3d 7 2G9/2 k5363.35 ............. O ii 4f.F 2½4o/2 Y 4d 0 2F7/2 k5363.80................. 7 [Cr ii] a4D3/2 Y c4D7/2 k5363.77 ........................ [Te iii] 5p2 3P1 Y 5p2 1D2 k7933.3......................... He i 3p 3P o Y 28s 3S kk7932.36, 7932.41......... o o [I iii] 5p2 4S3/2 Y 5p2 2D3/2 k8536.5 ....................... o Cr ii 5p 6D3/2 Y 6s 6D5/2 k8536.68 ..................... o Ba ii 5d 2D5/2 Y 6p 2P3/2 k6141.71 ........................ o O i 18d 3D2 Y 4p0 3D3 k6141.75 ....................... o Ne iii 4p 5P3 Y 4d 5D4 k6141.48 ....................... [Ni iii] 4s 3F2 Y 4s 3P1 k6141.83 ....................... o o [Ba iv] 5p5 2P3/2 Y 5p5 2P1/2 k5696.6.................... o C iii 2s 1P1 Y 3d 1D2 ( V2) k5695.92 ................ S ii 4d 0 2G7/2 Y 5f 0 2[5]o k5696.14 .................... o o [Pb ii] 6p 2P1/2 Y 6p 2P3/2 k7099.8 ........................ Si i 4p 1S0 Y 18d (3/2, 5/2)o k7099.70 .............. 1 Ar i 3d 2½1/2o Y 9f 2[3/2] k7099.82 ................. 0 3 3o Ca i 4s16d D Y 3d51p D k7099.82 ............... o Ca i 4s17d 1D2 Y 3d47p 1P1 k7099.78..............

ko , I/IH



V (km sþ1) 22.4 12.5 35.4 . .. .. . . .. . .. .. . ... .. . ... þ4.9 þ6.8 6.4 þ10.7 ... .. . ... ... .. . ... ... ...

b

Multc ... 0/0 0/0 .. . . .. .. . .. . . .. ... . .. ... .. . 4/0 ... 5/1 ... . .. ... ... . .. ... ... ...

IDI

ko , I/I

H

V (km sþ1) Mult . .. .. . . .. 11.7 25.7 0.6 2.2 .. . . .. .. . . .. . .. .. . . .. . .. . .. .. . . .. . .. .. . . .. . .. . .. IC 418 V (km sþ1) Mult 12.0 2.1 25.0 . .. .. . . .. . .. þ12.1 21.6 . .. .. . . .. .. . . .. . .. . .. .. . . .. 5.5 9.7 4.6 4.6 6.3 .. . 0/0 0/0 ... ... ... ... .. . 0/0 ... ... ... ... ... ... ... ... ... 0/0 0/0 0/0 0/0 0/0 ... ... ... .. . 4/0 0/0 6/1 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

IDI . .. .. . . .. . . .: 7 3A 4B .. . . .. .. . . .. . .. .. . . .. . .. . .. .. . . .. . .. .. . . .. . .. . ..

ko , I/IH 5759.61 2.4(þ5) 1.5(þ5)y 5363.96? 5.4(þ6) ... ... OUT ... OUT ... 6141.59 2.7(þ5) ... ... ... [6.2(þ6)] ... ... [6.9(þ6)] ... ... ...

Mult .. . 0/0 0/0 .. . 4/0 0/0 6/0 . .. . .. . .. . .. .. . 4/0 .. . 5/0 . .. . .. . .. . .. . .. . .. . .. . ..

IDI . ..bl 4Bbl 3A . ..: 8C 5A 7B .. . .. . .. . .. . . ..ö 6A . .. 7B .. . .. . .. . .. . .. . .. . .. . .. .

5759.98 5.1(þ4) .. . .. . [3.4(þ5)] .. . .. . OUT ... OUT ... 6141.61 8.4(þ5) .. . .. . ... [2.1(þ5)] ... ... [2.4(þ5)] ... ... ...

. .. 6Bö 5A ... ... ... ... ... . .. ... . .. . . .: 5A . .. 5A: . .. ... . .. . .. ... . .. . .. . ..

.. . [2.5(þ6)] .. . 5363.81 6.7(þ6) ... ... OUT .. . OUT .. . .. . [4.0(þ6)] .. . .. . .. . [2.5(þ6)] .. . .. . [4.5(þ6)] .. . .. . .. .

NGC 7027 V (km sþ1) 2.1 þ7.8 15.1 1.1 15.1 þ10.1 þ8.4 . .. .. . 5.6 þ0.7 þ9.8 þ11.7 1.5 þ15.6 þ21.6 14.2 2.6 þ0.8 3.4 þ1.7 þ1.7 0.0

Transition [Rb iv] 4p4 3P2 Y 4p4 1D2 k5759.55...................... He ii 5 Y 47 k5759.74 ........................................ [Fe ii] a2H9/2 Y c2D5/2 k5759.30 ....................... o o [Rb v] 4p3 4S3/2 Y 4p3 2D3/2 k5363.6 .................... 74 72 [Ni iv] 3d F7/2 Y 3d G9/2 k5363.35 ............. O ii 4f.F 2½4o/2 Y 4d 0 2F7/2 k5363.80................. 7 [Cr ii] a4D3/2 Y c4D7/2 k5363.77 ........................ 23 [Te iii] 5p P1 Y 5p2 1D2 k7933.3......................... He i 3p 3Po Y 28s 3S kk7932.36, 7932.41 ......... o o [I iii] 5p2 4S3/2 Y 5p2 2D3/2 k8536.5 ....................... 6o 6 Cr ii 5p D3/2 Y 6s D5/2 k8536.68 ..................... o Ba ii 5d 2D5/2 Y 6p 2P3/2 k6141.71 ........................ o O i 18d 3D2 Y 4p0 3D3 k6141.75 ....................... o Ne iii 4p 5P3 Y 4d 5D4 k6141.48 ....................... [Ni iii] 4s 3F2 Y 4s 3P1 k6141.83 ....................... o o [Ba iv] 5p5 2P3/2 Y 5p5 2P1/2 k5696.6.................... o C iii 2s 1P1 Y 3d 1D2 ( V2) k5695.92 ................ S ii 4d 0 2G7/2 Y 5f 0 2[5]o k5696.14 .................... o o [Pb ii] 6p 2P1/2 Y 6p 2P3/2 k7099.8 ........................ Si i 4p 1S0 Y 18d (3/2, 5/2)o k7099.70 .............. 1 Ar i 3d 2½1/2o Y 9f 2[3/2] k7099.82 ................. 0 Ca i 4s16d 3D Y 3d51p 3D o k7099.82 ............... o Ca i 4s17d 1D2 Y 3d47p 1P1 k7099.78..............

ko , I/IH



Mult .. . 0/0 0/0 .. . 4/0 0/0 6/0 .. . . .. 1/0 9/1 .. . 4/0 2/1 5/0 0/0 0/0 0/0 .. . 0/0 0/0 0/0 0/0

IDI . . .bl 5Abl 6B . . .ö 7B 6A 8C ... ... 6B: 4A . . .: 8 2A: 8 8bl 6Cbl 4A . . .ö 3A 4B 4B 4B

ko , I/I

H

IDI . . .: 3A 5B . .. .. . . .. . .. . ..ö 6A . .. .. . . .. .. . . .. . .. . .. .. . . .. 4Aö 4A 4A 4A 4A

Ref.

Notes
e,f

5759.59 3.1(þ4) 2.0(þ4)y 5363.62 9.0(þ5) .. . .. . .. . [2.0(þ5)] 8536.66? 1.7(þ5) 6141.51 7.3(þ5) .. . .. . 5696.19 5.9(þ5) 2.1(þ5)y 7099.78 4.6(þ5) .. . .. . .. .

5759.78? 2.0(þ5) .. . .. . [2.3(þ5)] .. . .. . 7932.98 1.7(þ5) .. . [2.3(þ5)] .. . [1.8(þ5)] .. . .. . .. . [1.5(þ5)] .. . 7099.93 2.4(þ5) .. . .. . .. .

Z05

Many

e

Z05
e

a Wavelength: (1) ko are nebular rest frame wavelength in 8; (2) ``OUT'' means not in observed range; (3) ``?'' denotes an uncertain feature. Intensity: (1) numbers in parentheses are exponents; (2) daggers denote corrected intensities attributable to the s-process transition; (3) bracketed values as upper limits for unobserved features. b Observed (ko ) þ transition wavelength ( km sþ1). c EMILI multiplet check statistics: number expected /number observed. d EMILI IDI value/rank followed by asterisk (certain ID), colon (uncertain ID), or ``bl'' ( blend). Definition of IDI given in x 6. e Unidentified line in IC 418 (Sharpee et al. 2003). f NGC 2440, IC 4191: contaminated by flare or ghost. References.--( Z05) PN NGC 7027 ( Zhang et al. 2005).


NEBULAR s-PROCESS ABUNDANCES

1277

Fig. 4.-- Continuum-subtracted spectra in the immediate vicinity of several prominent Kr, Xe, and Br lines, shifted to each nebula's rest frame. Predicted wavelengths are indicated by dashed lines. Labels ``NS'' indicate telluric nightglow lines mentioned in the text. The regions around [ Kr iii] k6826.70, [ Kr v] k6256.06, and [ Xe iii] k5846.77 are expanded in subsequent figures and discussed in the text.

of 0.65 (Schoning 1997), assuming electron densities well below ¨ the critical density values of $1:8 ; 107 cmþ3 and $1:3 ; 106 cmþ3 o o for the 2D3=2 and 2D5=2 levels, respectively, as justified from the values in Table 3, the observed k5346.02 intensity appears to be somewhat too high for IC 4191 and IC 418. This suggests the possible blending of [ Kr iii] k5346.02 with another transition, perhaps C iii ( V13.01) k5345.85, although poor multiplet statistics and the weakness of the strongest C iii lines in IC 418 cast doubt on this possibility for that PN. However, the other three PNe have ratios of 0.76 ( NGC 2440), 0.72 ( IC 2501), and 0.73 ( NGC 7027), which are slightly higher but consistent with the expected value within the combined measurement errors of the line intensities. o o The auroral [ Kr iv] 2 D3=2; 5=2 Y 2 P3=2 kk6107.8, 6798.4 identifications are less straightforward, except for NGC 7027 where both are considered certain according to both PB94 and the present EMILI results. In NGC 2440, the observed line corresponding to the k6798.4 transition, which should be 2.5 times weaker ´ than k6107.8 ( Biemont & Hansen 1986b), is actually stronger, while in IC 2501 the stronger k6107.8 line is not present at all. As suggested by both PB94 and the EMILI results, C ii ( V14) k6798.10 may be responsible for either the total intensity in IC 2501 or excess intensity in NGC 2440. While PB94 note that C ii ( V14) is of dielectronic recombination origin and its multiplet

components should not have relative intensities expected due to LS coupling rules, an examination of the intensities of all lines from this multiplet appearing in each PN suggests that the intensities do appear to roughly follow those rules with the exception of what would be the strongest line at 6783.91 8. Therefore, as an approximation, an LS-coupled relative strength of k6798.10 to k6791.47 of 0.16, instead of the one-to-one ratio with k6812.28 used by PB94, was utilized to correct the putative [ Kr iv] k6798.4 feature in both NGC 2440 and NGC 7027, as k6812.28 itself appears in only one of the PN spectra (it would be the weakest feature in the multiplet if LS coupling rules held). Subsequent abundance analysis also suggests that the [ Kr iv] k6107.8 line itself may be too strong relative to its nebular counterparts in NGC 2440 and IC 4191, admitting [ Fe ii] k6107.28 as an alternate identification that appears more likely in the latter PN. In summary, both auroral transitions can only be identified comfortably in NGC 7027. The third magnetic dipole transition of the [ Kr iv] auroral o multiplet 2 Do=2 Y 2 P1=2 k7131.3 is not definitively detected in any 3 spectrum including NGC 7027. However, in the NGC 7027 spectrum, a feature is observed at 7131.65 8 with intensity 7:7 ; 10þ5 I (H ), comparable to k6107.8. It originally was assumed to be an undercorrected Rowland ghost arising from the nearby


1278

SHARPEE ET AL.

Vol. 659

Fig. 4-- Continued

saturated [Ar iii] 3P2 Y 1D2 k7135.80 line, but it might instead be associated with k7131.3. The electric quadrupole line o o [Kr iv] 2 D5=2 Y 2 P1=2 k8091.0 is 40 times weaker and therefore undetectable. PB94 claim a detection of [ Kr v] 3P1 Y 1D2 k6256.06 as a blend with C ii ( V10.03) k6257.18 at their instrumental resolution. In our spectra, as is shown in Figure 5, C ii k6257.18 is resolvable from the putative [ Kr v] feature in all spectra except NGC 7027, where it slightly contributes to its red wing. However, in IC 2501 and NGC 7027, as also noted by Zhang et al. (2005), another C ii line, dielectronic C ii ( V38.03) k6256.52, is believed to be a significant contaminant in some PNe based on the strength of the nearby k6250.76 line from the same multiplet and on favorable EMILI assessments. The profiles of both C ii lines were approximated and subtracted from all the PN spectra in Figure 5, where the C ii ( V10.03) k6259.56 line profile scaled downward by a factor of 0.56 represents k6257.18, and the C ii k6250.76 profile, also scaled downward by a factor of 0.56 according to LS coupling statistics, represents k6256.52. The contribution of the interloping telluric OH 9 Y 3bandwas accountedfor throughsubtraction of a model of the band normalized in intensity to the nearby Q1(2.5) k6265.21 line. The resulting residual plots show a distinct line just to the red of the predicted wavelength of [ Kr v] k6256.06 in NGC 2440, IC 2501, and NGC 7027, which we believe can be definitely

identified as such in all three PNe. For IC 4191, the residual flux, 3:0 ; 10þ6 I (H ), is probably too low to be an actual line, and the entirety of the original profile is ascribed here to C ii ( V38.03) k6256.52, particularly since both the original line and k6250.76 share the same broad profile in the spectrum. In IC 418, the putative [ Kr v] profile is completely removed by the combination of the nightglow and C ii ( V38.03) k6256.52 model profiles, as is appropriate given the nebula's low excitation. The other nebular line, [ Kr v] 3P2 Y 1D2 k8243.39, accessible only in the bandpasses of the NGC 7027 and IC 418 spectra, sits amid the head of the Paschen series, rendering it difficult to disentangle at the resolution of our NGC 7027 spectrum. In the IC 418 spectrum, a comparable observed line is indistinguishable from other nearby lines in the Paschen series and is clearly identifiable as H i 3 Y 43 k8243.39, with the Kr v identification unlikely due to the PN's low ionization level. Returning to NGC 7027, PB94 identified this transition at 8242.7 8 as a blend with N i ( V2) k8242.39 and O iii 5g G 2[9/2]o Y 6hH 2[11/2] k8244.10, compromised by telluric absorption. In our NGC 7027 spectrum, [ Kr v] k8243.39 is tentatively identified as the blue peak at 8244.33 8 of a resolved two-line blend with H i 3 Y 42 k8245.64, which appears significantly affected by telluric absorption. The alternate identification for the line, N i k8242.39, is clearly resolved here as a separate feature. Because this observed line has a measured intensity ratio with k6256.06 roughly comparable to


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES

1279

Fig. 4-- Continued

what would be expected if both lines were due to [ Kr v], ´ I (k6256:06) /I (k8243:39) ¼ 1:1( Biemont & Hansen 1986a), it is believed that O iii k8244.10 contributes only a minor amount to the line, although the presence of the telluric absorption complicates matters. To summarize, both [ Kr v] lines appear to be present with about the expected intensity ratio in the spectrum of NGC 7027, but both are absent in the spectrum of IC 418. For the cases of IC 2501 and NGC 2440, [ Kr v] k6256.06 is probably present. The spectra were also examined for evidence of other auroral and transauroral transitions of Kr ions. Only the auroral line [Kr v] 1D2 Y 1S0 k5131.78 satisfied the initial ô1 8 screening criterion and also appeared in the EMILI output as a possible identification for an observed line at $5131.0 8 in all PNe. The case for this identification is lessened by the expected weakness of this line compared to other [ Kr v] lines that were not definitively identified, and particularly because of its high observed intensity of 1:3 ; 10þ3 I (H) relative to the $1 ; 10þ4 I (H ) for kk6256.06, 8243.39 in NGC 7027, the nebula that might be expected to exhibit the best evidence for this line given the strength of the other confirmed Kr lines. Instead, either O i 3p 3P Y 8d 3D o k5131.25, which was the highest ranked line in all but one nebula ( IC 418) and has a strength comparable to other members of the same sequence, or C iii 5g 3G Y 7h 3H o k5130.83 appears the more likely identification.

6.2. Xe Line Identifications A number of Xe ion transition identifications are also considered probable in our spectra. The EMILI statistics for Xe identifications are affected by the low solar Xe abundance, which is an order of magnitude lower than Kr ( Lodders 2003). This contributes to low predicted relative emission-line intensities and consequently higher or absent IDI values for its identifications as compared to those computed for weak transitions from more abundant elements. As such, every appearance of a Xe ion transition as a candidate identification for an observed line in an EMILI list, regardless of its IDI value, was given serious consideration as it signaled that alternative identifications from more abundant elements did not predominate despite the advantage arising from greater abundances. Table 8 lists the EMILI statistics for lines judged most likely to correspond to Xe ion transitions, while Figure 4 depicts the regions around the most likely observed transitions. The identification of [ Xe iii] 3P2 Y 1D2 k5846.77 in NGC 7027 was given by PB94 to an excess intensity in the He ii 5 Y 31 k5846.66 line. This excess was searched for in the present PN sample through the subtraction of the He ii k5837.06 profiles, shifted to the k5846.66 line position and scaled assuming I (k5837:06) /I (k5846:66) ¼ 0:92 (Storey & Hummer 1995; case B, for the most appropriate grid point: Te ¼ 10 4 K,


1280

SHARPEE ET AL.

Vol. 659

Fig. 4-- Continued

ne ¼ 10 4 cmþ3), as is shown in Figure 6. The intensity of He ii k5837.06 was not corrected for the presence of C iii 7h 3H o Y 18i 3I k5836.70 because the line used for this correction by PB94, C iii 7i 3I Y 18k 3K o k5841.2, was either not present or an identification of low rank ( IDI ¼ 6, ranked fifth) for a corresponding line in NGC 7027. As seen in Figure 6, distinct residual profiles are present after subtraction of He ii k5846.66 for both NGC 7027 and IC 418 (the latter having negligible He ii)and at the correct wavelength for [ Xe iii] k5846.77. A bizarrely shaped profile is seen for IC 2501, and no profiles are seen for either NGC 2440 or IC 4191 after subtraction. While the corresponding observed lines are probably He ii k5846.66 in the latter two PNe, IC 2501 has very weak He ii lines, suggesting that the residual profile is attributable to something else, tentatively [ Xe iii] k5846.77. The [ Xe iii] k5846.77 identification, while not appearing in the EMILI lists for NGC 7027 or IC 2501, is considered a better choice than the other top ranked identification, [ Fe ii] k5847.32, which has poorer wavelength agreement and multiplet statistics in all cases. While the [ Xe iii] 3P1 Y 1D2 transition at 1.37 m is unavailable for confirmation, we believe that the [Xe iii] k5846.77 is definitely present in NGC 7027 and IC 418 but is only tentatively identified in IC 2501. The reality of the observed features potentially correspondo o ing to the [ Xe iv] 4 S 5=2; 3=2 Y 2 D3=2 kk5709.2, 7535.4 transitions is marginal except in NGC 7027. However, their identifications as

the Xe lines, should they be actual emission lines, are more certain. For k5709.2, the widely observed N ii ( V3) k5710.77 line is detected as a separate line in all of the spectra. Alternative identifications such as [ Fe i] a5D3 Y a5P1 k5708.97 either have too many missing multiplet members, except for IC 2501 where the ratio of potentially observed to total number of multiplet lines expected is marginally better, or, as in the case of the Fe ii] k5709.04 intercombination line, are unlikely given that permitted Fe ii lines were not anywhere definitively identified. For k7535.2, the reality of the observed lines is less uncertain, including IC 418 where a previously unreported line has been uncovered by more thorough examination of its spectra. The Fe ii] ( V87 ) k7534.82 identification proposed by Hyung et al. (2001) is unlikely for the same reasons as Fe ii] k5709.04 for [ Xe iv] k5709.2. The N ii o 5fG 2[7/2]4 Y 10d 1F3 k7535.10 identification, selected by Esteban et al. (2004) and Peimbert et al. (2004), also appears unlikely given the relative strengths of other known permitted N ii lines in the spectra. It should be noted that the EMILI statistics for the k7535.10 identifications in the MIKE PN sample would probably have been better if the line did not appear near the end of the last spectral order where the wavelength calibration is the poorest. Assuming an electron density well below the critical regime ($107 cmþ3), the expected intensity ratio is I (k5709:2) /I (k7535:4) ¼ 0:64 (Schoning & Butler 1998). The ¨ observed range of values that we observe, 0.33 Y 1.19, suggests


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES

1281

Fig. 4-- Continued

measurement errors on par with the average values for such weak lines. The strongest examples in NGC 7027 do yield the bestagreeing ratio (0.82). As such, we conclude that both [ Xe iv]lines may be present in all PN spectra with varying degrees of certainty. One of the stated goals for future spectroscopy of NGC 7027 and other PNe given by PB94 was the detection of [ Xe v] transitions, none of which they detected with any certainty. In the present spectra, no observed line passed the ô1 8 initial selection criterion for the 3P1, 2 Y 1D2 kk5228.8, 6998.7 lines, nor did these transitions appear as possible EMILI IDs for any observed line. This is despite the improved sensitivity and resolution that should have enhanced the chances of detecting k5228.8 and that should have easily separated k6998.7 from O i ( V21) k7002.10, the contaminant noted by PB94. Nevertheless, the detection of these lines in our spectra would remain problematic since for k5228.8 extensive flaring from [O iii] k5006.84 in an adjacent order leads to numerous ghosts in its vicinity, while for k6998.7 telluric absorption in the tail end of the Fraunhofer A band complicates its observability. EMILI did suggest the fine-structure transition [ Xe v] 3P0 Y 3P2 k7076.8 as a possible identification for observed lines in two PNe, IC 4191 and NGC 7027, PNe with appropriate excitation levels for the appearance of a line of this ion. This is the primary EMILI identification in IC 4191. A third occurrence of the corresponding observed line in NGC 2440 is clearly associated with

a nightglow line from the O2 b 1 ô× Y X 3 ôþ 3 Y 2 band on comg g parison with a simulation of that band. The competing identification C i ( V26.01) k7076.48 has poor multiplet statistics, while [Fe iii] k7078.10 and [ Ni ii] k7078.04 are feasible but have poor wavelength agreement. The Ca i identifications are also unlikely given that they arise from energy levels of large energy, where lines that would follow from cascades from these levels are not observed. Two remaining obstacles to an identification with [Xe v] k7076.8 are the lack of the nebular 3P Y 1D lines and the low branching ratio for this electric quadrupole transition ($0.008 with respect to 3P1 Y 3P2 at 2.07 m). Nevertheless, despite this transition's expected weakness, the lack of viable alternate identifications suggests that the corresponding observed line may be at least tentatively identified in NGC 7027 and probably identified in IC 4191, and both occurrences yield reasonable abundance values in subsequent analysis. o o The fine-structure transition [ Xe vi] 2 P1=2 Y 2 P3=2 k6408.9 was identified with certainty in NGC 7027 by PB94. In the present NGC 7027 spectrum, k6408.9 is probably associated with a weak but clearly separable observed line on the red wing of He ii 5 Y 15 k6406.38. The alternate primary EMILI identification of [ Fe iii] k6408.50 was not considered likely, as an inspection of the NGC 7027 spectrum for other lines from all low-energy [ Fe iii] multiplets did not show a significant number of matches to warrant strong consideration as a likely identification. [ Fe iii] k6408.50


1282

SHARPEE ET AL.

Vol. 659

Fig. 4-- Continued

also has a comparatively high excitation energy (6.25 eV ). A C iv 9 Y 17 k6408.70 identification is similarly downgraded by its high excitation. The reality of corresponding observed lines in NGC 2440 and IC 4191 is questionable, although for lack of suitable alternate identifications, uncertain identifications of [Xe vi] k6408.90 are retained for both. As with Kr, an inspection for auroral and transauroral transitions of various Xe ions was also undertaken. There were coincidences within ô1 8 between observed lines and the [ Xe iii] 3 o P1 Y 1S0 k3799.96, [ Xe iii] 1D2 Y 1S0 k5260.53, [ Xe iv] 4 S3=2 Y o 2o o o P1=2 k3565.8, [ Xe iv] 2 D3=2 Y 2 P3=2 k4466.5, [ Xe iv] 2 D5=2 Y 2 P o k5511.5, [ Xe iv] 2 D o Y 2 P o k6768.9, and [ Xe v] 1D2 Y 1S 0 3=2 3=2 1=2 k6225.3 transitions. However, except for two instances, the identifications did not appear in the EMILI list for those particular lines, and most had more reasonable and higher ranked identifications: C ii (V30) kk5259.66, 5259.76, for example, instead of the auroral [ Xe iv] k5260.53 line. In many cases the line only appeared in one of the PN spectra analyzed here and could be rejected due to the strength or absence of the nebular transitions from the same ions. 6.3. Br Line Identifications
2 o In PB94 two transitions of Br were identified: [ Br iii] 4 S3=2 Y o 2o 2 P o k7385.1 D5=2 k6131.0 with certainty, and [ Br iv] D3=2 Y 3=2 o o as possible. A second nebular transition, [ Br iii] 4 S3=2 Y 2 D5=2

k6556.4, was lost in a blend amid the [ N ii] k6548.04+H+[ N ii] k6583.46 complex. The source for the Br iii levels in the Atomic Line List version 2.05 is Moore (1958), while the source used by PB94 is stated ´ to be the experimental levels listed in Table IV of Biemont & Hansen (1986a), from unpublished work of Y. N. Joshi and Th. A. M. van Kleef, and provided by private communication with van Kleef. However, a comparison of the Ritz-determined wavelengths from both sets of levels with those listed by PB94 shows that the Joshi and van Kleef 2Do term energy levels were used for the nebular transition wavelengths, while the Moore (1958) levels were used for auroral transition wavelengths. The substantial o difference in the 2D3=2 level energy, 15,042 cmþ1 for Moore (1958) and 15,248 cmþ1 for Joshi and van Kleef, leads to a difo o ference in the 4 S3=2 Y 2 D3=2 transition air wavelength of 6646.3 8 versus 6556.4 8, respectively, with a lesser difference for the other transition, 6132.9 8 versus 6131.0 8. Since the Moore (1958) levels used by the Atomic Line List version 2.05 and therefore by EMILI have stated uncertainties of 0.63 cmþ1, while those for ´ Joshi and van Kleef in Table IV of Biemont & Hansen (1986a) are listed to a precision of 1 cmþ1, nearly the same level of uncertainty, it is difficult to know which levels are more accurate. Therefore, EMILI was run using both sets of energy levels, and the 4S o Y 2D o lines were searched for at both sets of resultant wavelengths.


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES

1283

Fig. 4-- Continued

For the transition wavelengths generated from the Moore (1958) energy levels, kk6132.9 and 6646.3, no observed line was detected in any spectrum meeting the initial ô1 8 selection crio o terion. However, for the 4 S3=2 Y 2 D5=2 transition wavelength generated from those energy levels used by PB94, k6131.0, there do appear to be observed lines in four of the five PN spectra near the k6130.4 line that PB94 identified with [ Br iii]. Although [ Br iii] k6131.0 is the highest ranked EMILI identification in only one PN, as seen in Table 9, other identifications are not compelling. PB94 cite their observed line as being a blend with C iii 7h 1,3H o Y 16g 1,3G k6130.30, which is the highest EMILI-ranked transition in each spectrum in which the putative Br iii line is observed, except for IC 2501. However, the companion line C iii 7h 1,3H o Y 16i 1,3I k6126.30, claimed by PB94 to have an equal intensity to k6130.30, is not present in either the IC 4191 or IC 418 spectra. Well-known lower excitation C iii lines are of negligible intensity in IC 418, while in IC 2501 C iii k6126.30 has poor wavelength agreement with its corresponding observed line and is not the primary EMILI identification for that line. Only in NGC 7027 is C iii k6126.30 a primary ID, suggesting that only in its spectrum is the equally intense C iii k6130.30 likely to be present and accountable for at least a portion of the putative [ Br iii] k6131.0 line. As such for NGC 7027, the k6126.30 line intensity is subtracted from the observed line at 6130.32 8, while in the remaining spectra the corresponding observed lines

are ascribed wholly to [ Br iii] k6131.0. Possible contamination due to flaring from an adjacent order in the NGC 7027 spectrum was not accounted for. The [ Ni vi] k6130.40 identification, the second ranked identification in many of the spectra, is of high excitation (8.3 eV upper level), too high an ionization for IC 418, and appears only to be ranked highly due to a favorable coincidence in wavelength. Spectroscopic confirmation for this identification is sought in o o the possible presence of the [ Br iii] 4 S3=2 Y 2 D3=2 k6556.4 transition, in the PN spectra at expected intensities similar to that of k6131.0. As mentioned, detection of this line is difficult given its proximity to the saturated H line and its associated ghosts in its immediate proximity. Nevertheless, inspection of the line lists and spectra ( Fig. 4) does show distinct co-aligned features in the IC 2501, IC 4191, and NGC 7027 spectra near this wavelength. While initially assumed to be ghosts, the fact that these features can be found in spectra from two different instruments and appear invariant to the H intensity suggests that they may be unrelated to H. That real lines can be observed between [ N ii] k6548 and H is demonstrated by the presence of the strong telluric OH P1(3.5) 6 Y 1 k6553.62 line just to the blue of the suspected [ Br iii] feature. EMILI lists the [ Br iii] k6556.4 identification among possible choices in all PNe in which the line appears, with the k6131.0 identification satisfying the multiplet search in IC 2501 and IC 4191. Alternate EMILI-favored identifications


1284

SHARPEE ET AL.

Vol. 659

Fig. 5.--Left : Spectra in the vicinity of [ Kr iii] k6826.70 (black solid line) shifted to nebular rest wavelengths, with similarly shifted and superimposed co-added modeled spectra of telluric emission and observed interloping nebular features ( gray dot-dashed line). Plus signs indicate telluric OH Meinel 7 Y 2 R1(3.5) k6827.46, R1(4.5) k6828.47, and R1(2.5) k6829.49 lines (left to right), nebular He i k6827.88 is marked with an arrow, and an O vi k1032 Raman feature at 6829.16 8 ( NGC 7027) is marked with a lower arrow. Right : Same as the left panel, but for [ Kr v] k6256.06, with OH Meinel 9 Y 3 Q2(0.5) k6256.94 and Q1(1.5) k6257.96 marked as plus signs (left to right) and nebular C ii kk6256.52, 6257.18, and 6259.65 lines marked with arrows (left to right). Panels to the right show residuals after subtraction in the vicinity of predicted wavelengths of the target line (vertical dashed line in all panels).

for these features correspond to O ii and N ii permitted and coreexcited transitions between levels of primary quantum number 4 Y 6 and are doubtful as they are of comparatively high excitation with respect to better known 3 Y 3or3 Y 4 transitions in these PNe that appear as lines of equal or lesser intensities. They appear here only due to their wavelength agreement with the observed lines. The highest ranked Fe ii k6555.94 is discounted once again by the lack of other permitted Fe transitions in any spectra. The observed intensity ratios of the two putative [ Br iii] lines, I(k6131.0)/I(k6556.4), ranging from 0.18 to 0.45, do conflict with theoretical expectation as follows. While specific collision strengths are unavailable at present for the Br iii levels, the ratio of the strengths relevant to these transitions should still be roughly proportional to their respective upper level statistical weights ´ (Pequignot & Baluteau 1994; Osterbrock & Ferland 2006), weighted by a Boltzmann factor respecting the difference in level energies, and nonnegligible fine-structure emission between the 2o D levels, at subcritical electron densities. Employing the IRAF NEBULAR package task ionic to solve for the relative populations of the 2D o levels at temperature derived from the diagnostic thought to be the most appropriate, from [Ar iii], and using appropriate Kr iv collision strengths as a proxy, scaled as discussed in x 7, the I(k6131.0)/I(k6556.4) ratio is expected to be 0.72 Y 0.83. This is close to the 0.67 value expected from the ratio of the collision strengths alone and 2 Y 3 times larger than the ratio observed in the spectra. At these levels the [ Br iii] k6131.0 line should have been observable in NGC 2440, as the predicted intensity ex-

ceeds the likely detection limit for features in its vicinity. Therefore, we consider [ Br iii] to have been only tentatively detected in NGC 2440. However, at least one [ Br iii] line does appear to be present in the remaining four PNe. They all have lines detected at the same wavelengths, although shifted 20 Y 30 km sþ1 to the blue of the wavelengths we have adopted for the [ Br iii] lines, at approximately the same relative intensities, and they lack satisfying alternate identifications. The ratios of observed intensities (or in the case of IC 418 the upper limit), while not exact, are within a factor of 2 Y 3 of those expected. The identifications of the nebular [ Br iii] lines suggest that the level energies of Moore (1958) for at least the 2D o term levels may be in error. PB94 also identified an observed line at 7384.3 8 o o as possibly auroral 2 D3=2 Y 2 P3=2 k7385.2, but no such line appears in any of the PN spectra examined here. However, since PB94 appears to have used 2P o Y 2D o transition wavelengths derived from the Moore (1958) energy levels, this might not be a surprise, although no lines in any PN were discovered o o at the corresponding wavelength for the 2 D3=2 Y 2 P3=2 transition ´ (7483.1 8) computed from the levels listed by Biemont & Hansen (1986a) either. The nebular [ Br iv] 3P1 Y 1D2 k7368.0 line was detected by PB94 at 7366.0 8, affected by telluric absorption, and blended o with C iv 10 Y 21 k7363.9 and O iii 4s 3P1 Y 4p 3P1 k7365.35. In our NGC 7027 spectrum, a line appearing at 7367.62 8 as a small protrusion above the local continuum level interpolated between telluric absorption features in both spectral orders covering that


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES

1285

may be spurious. No auroral or transauroral features of [ Br iv] were identified in any spectrum. 6.4. Other Z > 30 Line Identifications The spectra were searched for lines originating from other Z > 30 ions. Among the numerous coincidences between observed lines and transition wavelengths at the ô1 8 level, those listed in Table 10 were judged to be the most likely to correspond to real lines describable by a Z > 30 identification in at least one of the PNe. Given the low solar abundances of these ions, only transitions among the lowest lying levels were expected to be observable. PB94 declared as certain the identification of a line at 5758.7 8 in NGC 7027 as [ Rb iv] 3P2 Y 1D2 k5759.44. Corresponding lines are detected in four of five PNe here. However, at even the highest instrumental resolution, this transition would blend with He ii 5 Y 47 k5759.74. In NGC 2440 and IC 4191 there is contamination from a flare or ghost at the position of He ii 5 Y 47 k5759.74 in one spectral order. Using only the uncontaminated order, and inspecting the nearby He ii 5 Y n sequence, suggests that in IC 4191 the He ii k5759.74 line shows significant excess, while in NGC 2440 there is better agreement with recombination theory and the line is likely He ii. The NGC 7027 spectrum, taken with a different instrument, does not show any contamination but do show a similar excess in He ii 5 Y 47 k5759.74. Thus, the line intensities in IC 4191 and NGC 7027 were corrected by I (5 Y 47) /I (5 Y 46) ¼ 0:95 derived from the ratio of their emissivities (Storey & Hummer 1995). IC 418 has no detectable He ii lines, enhancing the [ Rb iv] k5759.44 identification, but the weakness and irregularity of its profile cast doubt on its reality as a line, and it is only tentatively identified here. The alternate EMILI recommended identification, [ Fe ii] k5759.30, was not considered likely given its high excitation energy. Since the expected intensity ratio of 3P2 Y 1 ´ D2 k5759.44 to 3P1 Y 1D2 k9008.75 is 17 ( Biemont & Hansen 1986b), the apparition of k9008.75 in IC 418 and NGC 7027 is probably likely due to He i 3d 3D Y 10p 3P o kk9009.23, 9009.26 or is unreal. In summary, the [ Rb iv] k5759.55 line is identified in IC 4191, NGC 7027, and tentatively in IC 418. Another Rb line detection considered probable by PB94 in o o NGC 7027 is [ Rb v] 4 S3=2 Y 2 D3=2 k5363.6, which they identified at 5364.2 8. Although it does not appear in the EMILI list for the corresponding lines in our spectra, we believe that its identification in at least NGC 7027 is viable given the poor multiplet statistics of [ Ni iv] k5363.35, where none of the four other multiplet lines are clearly present. The O ii 4f F 2½4o=2 Y 4d 0 2F7/2 7 k5363.80 identification is an interesting alternative. Well-known O ii dielectronic doublet lines of multiplets V15, V16, and V36, o such as 3s 0 2D5/2 Y 3p0 2F7=2 ( V15) k4590.97, are all clearly present in all of our PN spectra, except NGC 7027 where they are either outside the bandpass or not optimally placed for detection, at intensities $10þ4I(H ). These lines are present at similar intensities in the NGC 7027 spectrum of Zhang et al. (2005). Lines from 3d Y 4f transitions, particularly 3d 2D5/2 Y 4fF 2½4o=2 ( V92a) 7 k4609.4, arising from the lower level of the O ii k5363.80 transition are also favorably identified at roughly the same intensity. Therefore, it is not out of the realm of possibility that this line is evidence of the partial feeding of the 2½4o=2 level. However, 7 while EMILI did not perform a multiplet check on this line (due to a non Y LS-coupled lower level), two other transitions at 5361.74 and 5375.57 8 from the same upper term and ending on this level are not present in any of the spectra. Without spontaneous emission coefficients it is difficult to judge a potential branching ratio for these transitions. Therefore, it is believed that [ Rb iv] k5363.60

Fig. 6.--Spectra in the vicinity of [ Xe iii] k5846.77 (black solid line) shifted to nebular rest wavelengths, with the superimposed profile of the He ii 5 Y 32 k5837.06 shifted to the rest wavelength of the He ii 5 Y 31 line and scaled as described in the text ( gray dot-dashed line). Panels to the right show residual spectra after subtraction near the rest wavelength of the [ Xe iii] k5846.77 line (vertical dashed line in all panels).

wavelength is tentatively identified as [ Br iv] k7368.0. The alternate candidate listed by PB94, dielectronic C ii 3p0 2D5/2 Y 3d 0 2o P3=2 k7377.00, is too far away, while an O ii k7367.68 identification is doubtful given its high upper level energy. The C v 7p 3P o Y 8d 3D k7367.60 transition is a possible alternate identification, but only one other C v transition in the spectrum, C v 6gh 1,3G, H o Y 7hi 1,3H o, I k4944.50 with an IDI value of 3, is a top ranked EMILI IDI; others are of lesser rank or do not appear in the EMILI lists for the corresponding observed lines. The other nebular transition, [ Br iv] 3P2 Y 1D2 k9450.5, was not detected by PB94, even though the line should be of the same ´ intensity ( Biemont & Hansen 1986a). Comparison between the IC 418 and NGC 7027 spectra shows that something is filling in the telluric absorption feature in the latter at 9450.85 8, with an estimated intensity close to [ Br iv] k7368.10, although a ghost feature at this wavelength arising from scattered light within the spectrograph cannot be ruled out based on the proximity of other similar features. If the feature is real, however, the alternate identification of Fe i k9450.95 is not compelling. Therefore, both transitions are identified in NGC 7027, although tentatively since they appear in only one PN spectra, they may be attributable to misinterpretation of the local continuum level complicated by telluric absorption of varying degree, and the latter observed line


1286

SHARPEE ET AL.

Vol. 659

might be tentatively identified in IC 2501 and IC 4191 as well. o Unfortunately, the matching 4 S3=2 Y 2 Do=2 k4742.40 line is dis5 guised by a flare or ghost in the MIKE spectra and is not optimally placed in NGC 7027 to allow a spectroscopic confirmation, nor observed in IC 2501 and IC 4191. Numerous lines of various Sr ions were searched for, includo o ing the strong [I /I (H ) ¼ 3 ; 10þ4 ][Sr iv] 2 P3=2 Y 2 P1=2 k10276.9 fine-structure line named as a possible identification by PB94 o o o in NGC 7027, and the [Sr vi] 4 S3=2 Y 2 D5=2 k4249.2 and 2 D3/2Y 2o P1=2 k5434.3 lines identified as possible and tentative, respectively, by PB94 and also identified by Zhang et al. (2005). The Sr ii kk4077.71, 4215.52 resonance lines and the [Sr ii] 2 S = Y 2 D = = kk6738.39, 6868.17 lines observed in Carinae 12 5 2;3 2 ( Zethson et al. 2001) were also sought. While there were some instances of wavelength coincidences with observed lines, the magnitude of the wavelength differences and existence of plausible alternate identifications, such as [ Fe ii] 4249.08 for [Sr vi] k4249.2, did not warrant a claim of identification. The same was o o true for Y and Zr, where the [ Y v] 2 P3/2 Y 2 P1=2 k8023.6 and [Zr vii] 3 3 P2 Y P0 k7961.4 fine-structure lines, listed as possible and tentative identifications by PB94, respectively, did not have any wavelength coincidence with an observed line in any PNe within the initial ô1 8 screening criteria. Turning to the fifth row of the periodic table, we believe that an observed and previously unidentified line in IC 418 may correspond to [ Te iii] 3P1 Y 1D2 k7933.3, which would be the first visual identification of a Te iii line in a PN. The He i 3p 3P oY ns 3S sequence of lines, of which the alternate identifications He i kk7932.36, 7932.41 are members, does not become clearly evident in this spectrum until 3p Y 10s at 8632.76 and 8632.83 8. The wavelength coverage of our spectra does not extend out to the possibly stronger 3P2 Y 1D2 k10876.0 transition, so this line cannot be used to check the Te iii identification. Similarly, the deo o tection of a possible [ I iii] 4 S3/2 Y 2 D3/2 k8536.6 line in NGC 7027, with a less than compelling Cr ii permitted line as an alternate primary EMILI identification, cannot be confirmed through obo o servation of its companion 4 S3/2 Y 2 D5/2 k6708.7 transition. The putative k6708.7 line appears to be better explained by a combination of [ Mn ii] k6709.93 and possibly [Cr v] k6709.8, although the latter line has poor multiplet statistics (2/0). Nevertheless, the [I iii] k8536.6 transition has a definite IDI that is second ranked for its corresponding observed line, so we count this as a tentative detection for NGC 7027. PB94 claims the detection of four transitions belonging to permitted multiplets of Ba ii, with three detections considered certain and one considered possible. Some of these same transitions have also been identified in emission by Zhang et al. (2005) in NGC 7027 and by Hobbs et al. (2004) in the compact H ii region within the Red Rectangle. In the present sample it is believed that one of the transitions considered certain by PB94, o Ba ii 5d 2D5/2 Y 6p 2P3/2 k6141.71, may correspond to an observed line in three of five PNe. Given the low solar abundance and high condensation temperature (1455 K; Lodders 2003) of Ba, the detection of these transitions might initially be considered unlikely. However, as originally proposed by PB94, if sufficient gas-phase Ba+ is available for collisional excitation, emission, and moderate self-absorption, it is estimated that o optical depths of 0.5 and 2.7 in the 6s 2S1/2 Y 6p 2P3/2 k4554.03 resonance transition are sufficient to account for both its apparent nondetection and the observed intensity of the putative k6141.71 lines in NGC 2440 and IC 4191, respectively. PB94 estimated an optical depth of 3 for the k4554.03 transition in NGC 7027. Under these scenarios the k6141.71 line is either among the strongest or is the strongest observable Ba ii line, with its relative

intensity with respect to other Ba ii lines increasing with a larger optical depth in the k4554.03 line. The EMILI results do suggest some potentially viable alternative identifications other than Ba ii for individual PNe, but none that can satisfactorily account for the observed line in all three. O i 18d 3Do Y 4p0 3D3 k6141.75 arises from an autoionizing level, 2 is too strong with respect to other permitted O i transitions further down the cascade chain, and is not accompanied by any other multiplet members in any PNe. Under the abundance and ionization model created by EMILI for NGC 2440 and IC 4191, Ne iii o 4p 5P3 Y 4d 5D4 k6141.48 did not produce an emission line with an intensity within 3 orders of magnitude of the strongest predicted intensity among all putative identifications. This was not the case in NGC 7027, where the line was among the strongest predicted lines and is expected to be among the strongest in the multiplet, and where the identification is enhanced by favorable multiplet statistics. The [ Ni iii] 4s 3F2 Y 4s 3P1 k6141.83 arises from a level of high-excitation energy (9.9 eV ) and is probably not the strongest member of its multiplet as it does not originate from the level of the multiplet with the highest statistical weight, and other potentially stronger multiplet members are not observed in IC 2501 and NGC 7027. Yet in NGC 2440, the multiplet statistics are somewhat better, and the large intensity of the corresponding observed line and the Ba abundance derived from it is at odds with lower abundances derived for Kr and Xe in subsequent abundance analysis. The excellent agreement between the observed wavelengths in all three PNe and the Ba ii k6141.71 wavelength, the expectation that k6141.71 is the strongest Ba ii line, and the lack of a good alternate identification for IC 4191 suggest that Ba ii k6141.71 warrants serious consideration as the correct identification in all cases. o o The fine-structure transition [ Ba iv] 2 P3/2 Y 2 P1/2 k5696.6, another certain detection from PB94, is verified by its detection in our NGC 7027 spectrum after a correction for C iii ( V2) k5695.92 is made. The intensity attributable to k5696.6 was determined using the effective recombination coefficient for C iii k5695.92 ( ¼ 3:1 ; 10þ15 cm3 sþ1) specified by PB94, those ´ from Nussbaumer & Storey (1984) and Pequignot et al. (1991) for C iii 5g 1,3G Y 6h 1,3H o k8196.61, and the k8196.61 line's observed intensity. Subtraction yielded a line amounting to 36% of the originally observed intensity. The only sixth-row elemental transition sought in these speco o tra was the fine-structure [ Pb ii] 2 P1/2 Y 2 P3/2 k7099.80 line, which was identified with certainty by PB94. The transition appears in the EMILI list for IC 418 tied for the highest ranked transition for a previously unidentified line. For NGC 7027, the transition is not ranked, but the competing identifications in its EMILI list are not convincing given their high excitation energies. Excellent wavelength agreement is seen in both cases, and both identifications are considered certain. In summary, of the 18 Z > 30 elemental ion transitions considered certain or probable by PB94 in NGC 7027, 15 are believed to be detected to various degrees of certainty in the present set of spectra. The final intensities of all Z > 30 lines detected with any certainty within at least one PN are presented in Table 11, with certain identification shown with an asterisk and tentative identifications without. Conspicuous among those missing from PB94 is [Se iii] 3P1 Y 1D2 k8854.2. The nearby He i 3d 3D Y 11p 3P o k8854.14 transition does appear to show a large excess relative to that expected from other confirmed lines in the same series in both IC 418 and IC 7027. However, the energy levels in the Atomic Line List version 2.05, derived from Moore (1952) and also utilized by PB94, have a listed uncertainty of 6.3 cmþ1 that exceeds the maximum 1 cmþ1 tolerance allowed for inclusion of


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES
TABLE 11 Summary of Poss ible Z > 30 Ion Line Inten sities PNe NGC 7027

1287

H ii Regions

Transition [Br iii] k6131.0 ..................... [Br iii] k6556.4 ..................... [Br iv] k7368.1 ..................... [Br iv] k9450.5 ..................... [Kr iii] k6826.70 ................... [Kr iv] k5346.02 ................... [Kr iv] k5867.74 ................... [Kr iv] k6107.8 ..................... [Kr iv] k6798.4 ..................... [Kr v] k6256.06.................... [Kr v] k8243.39.................... [Rb iv] k5759.55 .................. [Rb v] k5363.6 ..................... [Te iii] k7933.3 ..................... [I iii] k8536.5 ........................ [Xe iii] k5846.77 .................. [Xe iv] k5709.2 .................... [Xe iv] k7535.4 .................... [Xe v] k7076.8 ..................... [Xe vi] k6408.9 .................... Ba ii k6141.71....................... [Ba iv] k5696.6 .................... [Pb ii] k7099.8 ......................

NGC 2440 <1.7(þ5) 4.2(þ5) <2.6(þ5) ... ... 3.5(þ4)ö 4.6(þ4)ö 5.6(þ6) 1.3(þ5)ö 1.5(þ5)ö ... ... <3.4(þ5) ... ... ... 1.2(þ5)ö 1.3(þ5)ö ... 2.8(þ5) 8.4(þ5) <2.1(þ5) <2.4(þ5)

IC 2501 1.7(þ5)ö 7.3(þ5)ö <4.4(þ5) .. . 2.5(þ5)ö 6.1(þ5)ö 8.5(þ5)ö <4.3(þ6) .. . 8.0(þ6)ö .. . <2.5(þ6) 6.7(þ6) .. . .. . 7.7(þ6) 6.5(þ6) 1.6(þ5)ö <5.5(þ5) <2.4(þ6) <4.0(þ6) <2.5(þ6) <4.5(þ6)

IC 4191 3.8(þ5)ö 8.5(þ5)ö <3.5(þ5) ... 1.5(þ5) 2.4(þ4)ö 1.9(þ4)ö ... <8.1(þ6) ... ... 1.5(þ5)ö 5.4(þ6) ... ... ... 1.2(þ5)ö 3.6(þ5)ö 3.9(þ6)ö 1.6(þ6) 2.7(þ5)ö <6.2(þ6) <6.9(þ6)

(a) 7.7(þ5)ö 4.3(þ4)ö 4.4(þ5) 7.5(þ5) 4.6(þ4)ö 1.9(þ3)ö 2.6(þ3)ö 7.2(þ5)ö 1.6(þ5)ö 1.2(þ4)ö 1.4(þ4) 2.0(þ4)ö 9.0(þ5)ö <2.0(þ5) 1.7(þ5) 1.4(þ4)ö 1.1(þ4)ö 1.7(þ4)ö 2.0(þ5) 2.0(þ4)ö 7.3(þ5) 2.1(þ5)ö 4.6(þ5)ö

(b) 7.6(þ5)ö ... 2.0(þ5) <7(þ6) 4.1(þ4)ö 1.9(þ3)ö 2.6(þ3)ö 6.3(þ5)ö 2(þ5) 1.5(þ4)ö 2.0(þ4)ö 1.7(þ4)ö 8.3(þ5) <1.1(þ5) 2.3(þ5) 5.3(þ5)ö 1.1(þ4)ö 1.9(þ6)ö .. . 1.3(þ4)ö 7.5(þ5)ö 7.3(þ5)ö 3.5(þ5)ö

(c) <1.1(þ4) . .. .. . .. . 4.4(þ4) 1.5(þ3) 2.3(þ3) 5(þ5) .. . 1.8(þ4) .. . 2.3(þ4) .. . .. . .. . 1.2(þ4) 9(þ5) 1.4(þ4) ... .. . 7(þ5) 8(þ5) 1.0(þ4)

IC 418 2.8(þ5)ö .. . <4.9(þ5) <5.3(þ5) 3.2(þ4)ö 3.5(þ5)ö 3.5(þ5)ö <1.4(þ5) <2.1(þ5) <8.4(þ5) . .. 2.0(þ5) <2.3(þ5) 1.7(þ5)ö <2.3(þ5) 1.3(þ4)ö 1.9(þ5)ö 1.6(þ5)ö <2.8(þ5) <9.3(þ6) <1.8(þ5) <1.5(þ5) 2.4(þ5)ö

Orion .. . ... .. . .. . 7.0(þ5) .. . 7.8(þ5) .. . .. . .. . ... .. . 5.6(þ6) .. . .. . <7.8(þ5) <7.8(þ5) .. . .. . .. . .. . .. . .. .

NGC 3576 ... .. . ... ... <3.0(þ5) ... <5.0(þ5) ... ... ... . .. ... .. . ... ... .. . <5.0(þ5) <2.0(þ5) ... ... ... ... ...

´ Notes.--In units of I (H) ¼ 1; certain identifications shown with an asterisk. NGC 7027: (a) present measurement; (b)Pequignot & Baluteau (1994); (c) Zhang et al. ´ (2005); Orion: Baldwin et al. (2000); NGC 3576: Garcia-Rojas et al. (2004).

supplemental Z > 36 ions into EMILI. This leads to an error of up to 10 8 in the transition wavelength. Thus, the identification cannot be confirmed under the present degree of energy level uncertainty. Also missing are the Ba ii transitions 6s 2S1/2 Y o o 6p 2P3/2 k4554.03 and 5d 2D3/2 Y 6p 2P3/2;1/2 kk5853.67, 6496.90. As discussed previously, all would be weaker in our spectra than the potentially observed k6141.71 under the assumption of a moderate optical depth in the resonance transition k4554.03 (and o 6s 2S1/2 Y 6p 2P1/2 k4934.08, not observed by PB94). Zhang et al. (2005) have claimed the identification of five additional transitions in their NGC 7027 spectra belonging to Z > 30 elements that were not observed in our spectra. Two additional permitted Ba ii lines, Ba ii k4554.03 and Ba ii k4934.08, are detected, both weaker than the k6141.71 line as would be expected for substantial optical depths of the resonance lines. They also identify auroral [ Br iii] k7385.2 and [ Rb v] k5080.2, although the nebular [ Br iii] k6131.0 and [ Rb v] kk4742.4, 5363.6 lines, which might be expected to be stronger than the auroral lines, are not identified. Zhang et al. (2005) also identify [Sr vi] k4249.2, which might be better attributable to [ Fe ii] k4249.08. It should be noted that the present spectrum of NGC 7027 and that of Zhang et al. (2005) were obtained at different spectral resolutions, signal-to-noise ratios, and locations in the PN, resulting in some differences in the lines detected and identified in each. 6.5. Z > 30 Line Identifications in H ii Regions In comparing the spectra of the PNe with H ii regions, we note that no post Y Fe peak lines were identified by Garcia-Rojas et al. ´ (2004) in their deep VLT UVES echelle spectrum of the H ii region NGC 3576. In Table 11 are included estimates of the upper limits to undetected Kr and Xe line intensities for this H ii region

from the faintest detectable lines that were identified at neighboring wavelengths or from artificial lines inserted at their wavelengths that meet minimal detection S/ N statistics. In the spectrum of the Orion Nebula, Baldwin et al. (2000) reported weak (I % 2 ; 10þ5 H ) unidentified features at 5867.8 and 6826.9 8 that were originally unidentified but can now be positively identified as o o [Kr iv] 4 S3/2 Y 2 D3/2 k5867.74 and [ Kr iii] 3P2 Y 1D2 k6826.70, respectively. However, there are no other close coincidences between any lines listed in Tables 7 Y 10, except for k5363.34, which o o would correspond with [ Rb v] 4 S3/2 Y 2 D3/2 k5363.60, an unlikely identification given the unrealistically high degree of ionization (40 eV ) for an H ii region. The observed intensities and upper limits are also included in Table 11. 7. Z > 30 ABUNDANCES The line intensities given in Table 11 were used to compute abundances for Ba, Kr, Xe, and Br ions. The atomic data listed in Table 3 were formatted for inclusion in five-level atom models for abundance analysis with the IRAF NEBULAR task abundance, or for the cases of the fine-structure lines [ Xe vi] k6408.9 and [ Ba iv] k5696.6, using a two-level atomic solution code. For the [ Br iii] and [ Br iv] lines the relevant collision strengths have not yet been calculated. However, since these ions are isoelectronic with [ Kr iv] and [ Kr v], and because collision strengths for the same levels along an isoelectronic sequence tend to vary with effective nuclear charge (Seaton 1958), with some exceptions, the collision strengths of [ Br iii] and [ Br iv] were assumed to be 25% smaller than those for Kr. Because collision strengths were not calculated by Schoning & Butler (1998) for [ Xe v] transi¨ tions, the [ Kr v] collision strengths for the same transitions were utilized in the Xe+4 analysis. Collision strengths for excitation to


1288

SHARPEE ET AL.
TABLE 12 s -Process Ionic/ Elementa l Abundances PNe

Vol. 659

H ii Regions NGC 7027 2: 3: 2: 2: 2: 2: 3: 3: 3.77 ô 0.02 3:99×0::05 þ0 06 3:8×0::3 þ0 6 3.79 ô 0.02 3.00 3.10 3:05×0::05 þ0 06 2:2×0::2 þ0 4 2.58 ô 0.04 2:63 ô 0:06 2.59 ô 0.03 2:38×0::08 þ0 09 2.28 2:41:þ:0::05 1:7×0::3 þ1 2 6.09 ô 0.02 5.77 ô 0.02 5.60 þ3:22×0::16 þ0 22 þ2.36 ô 0.04 þ3:35×0::08 þ0 09 0:73×0::16 þ0 22 0.90 ô 0.04 0:92×0::08 þ0 09 52× þ 14× þ 57× þ 53× þ 74× þ 58× þ 39× þ 78× þ
0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 0: 09 10 10 12 08 09 09 11 11 14 08 09 07 08 02 03

X+i/H + Transition Br /H k6131.0 ............ Br+2/H + k6556.4 ............ Adopted .......................... Br+3/H + k7368.1 ............ Br+3/H + k9450.5 ............ Br+3/H + adopted ............ Kr+2/H + k6826.70 ......... Kr+3/H + k5346.02 ......... Kr+3/H + k5867.74 ......... Kr+3/H + k6107.8 ........... Kr+3/H + k6798.4 ........... Kr+3/H + adopted ............ Kr+4/H + k6256.06 ......... Kr+4/H + k8243.39 ......... Kr+4/H + adopted ............ Xe+2/H + k5846.77 ......... Xe+3/H + k5709.2 ........... Xe+3/H + k7535.4 ........... Xe+3/H + adopted ........... Xe+4/H + k7076.8 ........... Xe+5/H + k6408.9 ........... Ba+/H + k6141.71........... Ba+3/H + k5696.6 ........... Ar+2/H + .......................... Ar+2/H + .......................... Ar+4/H + .......................... Br/Ar .............................. Kr/Ar .............................. Xe/Ar .............................. [ Br/Ar] ........................... [ Kr/Ar] ........................... [ Xe/Ar]...........................
+2 +

NGC 2440 <1.77 2:13×0::19 þ0 28 2:13×0::19 þ0 28 <2.21 .. . <2.21 .. . 2:96×0::09 þ0 10 2:96×0::09 þ0 10 2:8×0::2 þ0 3 3:6×0::3 þ0 7 2.94 ô 0.07 1:7×0::5 þ1 7 .. . 1:7×0::5 þ1 7 .. . 1:53×0::12 þ0 15 1:46×0::12 þ0 15 1:49×0::09 þ0 10 .. . 1:17×0::16 þ0 19 2:73×0::14 þ0 16 <1.66 6.13 ô 0.04 5.70 ô 0.03 ×0 06 5:24þ0::08 .. . þ3.33 ô 0.11 þ4:64×0::12 þ0 14 .. . þ0.07 ô 0.11 þ0:37×0::12 þ0 14

IC 2501 2: 2: 2:
× 21þ × 70þ × 27þ 0:16 0:19 0:18 0:22 0:14 0:15

IC 4191 2: 2: 2:
× 18þ × 44þ × 24þ 0:14 0:17 0:16 0:21 0:11 0:13

IC 418 2:41 <3.07 2:41×0::17 þ0 22 <3.00 <3.02 <3.00 3:53×0::13 þ0 16 2:60×0::17 þ0 18 2:40×0::16 þ0 18 <4.07 <4.65 2.46 ô 0.12 <3.15 .. . <3.15 2:53×0::13 þ0 16 2:32×0::16 þ0 18 2:02×0::16 þ0 18 2.10 ô 0.12 <2.83 <1.24 <1.97 <1.90 6:01×0::06 þ0 07 2:96×0::12 þ0 06 .. . .. . ×0 14 þ2:44þ0::18 ×0 14 þ3:34þ0::17 .. . 0:82×0::14 þ0 18 0:93×0::14 þ0 17
×0:17 þ0:22

Orion ... ... ... ... ... ... 3.01 ... 2.83 ... ... 2.83 ... ... ... <2.47 <3.02 ... <3.02 ... ... ... ... 6.42 4.37 ... ... þ3.19 <þ3.29 ... +0.07 <0.98

NGC 3576 ... ... ... ... ... ... <2.47 ... <2.60 ... ... <2.60 ... .. . ... .. . <2.80 <2.15 <2.15 ... ... ... ... 6.34 ô 0.04 4.20 ô 0.07 .. . .. . <þ3.50 .. . .. . <þ0.24 .. .

<2.89 . .. <2.89 2:4×0::3 þ0 5 ×0 12 2:75þ0::15 ×0 12 2:69þ0::15 <3.44 .. . ×0 09 2:72þ0::10 2:1×0::6 þ2 1 . .. 2:1×0::6 þ2 1 1:3×0::2 þ0 3 ×0 18 1:76þ0::28 ×0 12 1:94þ0::15 ×0 11 1:87þ0::13 <3.01 <0.68 <1.85 <1.15 6.27 ô 0.07 4.75 ô 0.03 .. . . .. þ3.3 ô 0.3 þ4:31×0::15 þ0 19 . .. þ0.1 ô 0.3 þ0:04×0::15 þ0 19

<2.74 ... <2.74 ×0 13 1:89þ0::17 3:28 ô 0:10 2.99 ô 0.10 .. . <4.03 3.07 ô 0.07 .. . ... .. . .. . ×0 13 1:97þ0::15 ×0 12 2:25þ0::15 ×0 10 2:05þ0::11 ×0 17 1:82þ0::27 ×0:3 0:3þ0:4 2:5:þ:0::2 <1.13 ×0 05 5:70þ0::06 ×0 03 5:74þ0::04 ×0 09 4:57þ0::12 ... þ2:94×0::09 þ0 10 þ3:78×0::13 þ0 17 ... ×0 09 0:32þ0::10 ×0 13 0:49þ0::17

a a

Note.--In units of 12 × log ( N/ H ). a Ar abundances using t 2 ¼ 0:00.

the 3P2 parent level of the tentatively observed [ Xe v] 5p2 3P0 Y 3 P2 k7076.8 line in other np2 ions such as Ne v ( Lennon & Burke 1994), Ar v (Galavis et al. 1995), and Kr v (Schoning 1997) ¨ depart by 15% at most from their Kr v values at 10,000 K. The temperatures and densities used for these analyses corresponded to those from diagnostics with the closest ionization potential. The [Ar iii] (27.6 eV ) temperature and [Cl iii] (23.8 eV ) density diagnostic values were used for [ Br iii] (21.8 eV ), [ Kr iii] (24.4 eV ), [ Xe iii] (21.1 eV ), and [ Ba iv] (20.0 eV ); [O iii] (35.1 eV ) temperature and [Ar iv] (40.1 eV ) density for [ Br iv] (36.0 eV ), [ Kr iv] (37.0 eV ), and [ Xe iv] (32.1 eV ); [ Ne iii] (41.0 eV ) temperature and [Ar iv] (40.1 eV ) density for [ Xe iv] (46.0 eV ); and [Ar v] (59.8 eV ) temperature and [ K v] (60.9 eV ) density for [ Kr v (52.5 eV )] and [ Xe vi] (57.0 eV ). The [O i] temperature and [ N i] density were used for Ba ii assuming collision excitation of Ba+ as the source of Ba ii k6141.71. Averaged diagnostic values were used when one of the diagnostics was unavailable for a particular ion. Uncertainties were computed in the same manner as the lighter elements, through permutation of intensity measurement and diagnostic value errors, where available, and selection of extrema values. For lines that were corrected for blending, 25% of the value of the correction was added to the measurement uncertainty. Table 12 presents the results of the abundance determinations for individual lines of Ba, Br, Kr, and Xe ions. To compute over-

all elemental abundances for the latter three ions, argon has been selected as a benchmark element in addition to hydrogen. This was done because the ionization potentials of the noble gases Ar, Kr, and Xe, as well as Br, are very similar for their first three stages of ionization and therefore ionization corrections should not be large when making abundance comparisons among these elements. In addition, the noble gases are almost completely nonreactive and have very low condensation temperatures, so corrections for gas-phase abundances depleted by grain formation are insignificant for these elements. Only ionic abundances relative to hydrogen were computed for Ba. We convert from ionic to elemental abundances by making use of the similarity in ionization potentials of the noble gases, so that Kr /Ar ¼ (Kr×2 × Kr×3 )/(Ar×2 × Ar×3 ) and (Kr×2 × Kr×3 × Kr×4 )/(Ar×2 × Ar×3 × Ar×4 ) and Xe /Ar ¼ (Xe×2 × Xe×3 )/(Ar×2 × Ar×3 ) and ( Xe×2 × Xe×3 × Xe×4 × Xe×5 )/(Ar×2 × Ar×3 × Ar×4 ) for H ii regions and PNe judged to be of low excitation ( IC 2501 and IC 418) and for highexcitation PNe ( NGC 2440, IC 4191, NGC 7027), respectively. The Kr+4/H+ abundance was included in the IC 2501 Kr /Ar ratio determination. For NGC 7027, the Br/Ar ratio was calculated the same way as the Xe/Ar ratio, with corrections for unobservable Br+4 and Br+5 made assuming Br×4 ; Br×5 /Br ¼ Xe×4 ; Xe×5 / Xe, appropriate because of their very similar ionization potentials.


No. 2, 2007

NEBULAR s-PROCESS ABUNDANCES

1289

The resulting abundances are listed in the bottom rows of Table 12 relative to their solar values from Lodders (2003). Three particular results are noteworthy. First, three out of five PNe show significant enhancements in Kr and Xe relative to solar values, while two others show similar, solar-like values. It is also interesting to note that IC 418 and NGC 7027, both considered young PNe, have the largest overabundances, although they are of greatly different ionization classes. Meanwhile, the H ii regions show only solar Kr abundances, indicative of unprocessed interstellar medium ( ISM ) gas. Secondly, the Kr and Xe abundances in all PNe all show enhancements of similar magnitude, as was seen in NGC 7027 by PB94. Finally, for NGC 7027, the Br abundance also shows a level of enhancement similar to that of Kr and Xe. Uncertainties remain regarding the adaptability of Kr collision strengths, modified as was done here, for the Br abundance calculations. However, to reduce the Br abundance to solar, the collision strengths for Br+2 and Br+3 would have to both be 4 times greater than their counterparts in Kr, and this is much larger than the difference seen for analogous transitions for other p 2 and p3 ion pairs ( N and O, Cl and Ar). In any event, the prevalence of probable Br line identifications suggests, independent of the actual atomic data, that significant Br does exist in most of the PNe comprising our sample. It is of interest to compare the s-process abundances we derive for PNe with those obtained for evolved stars from analyses of their absorption spectra. With the exception of Rb, Ba, and Pb, the elements observed in emission in PNe are different from those normally observed in the spectra of late-type stars, so a direct comparison is not feasible. In fact, lines of the same s-process elements are not necessarily even observed in similar-type stars. For this reason Luck & Bond (1991) defined two parameters, [ ls] and [ hs], that represent the mean abundance of elements associated with Sr and Ba, respectively, in what are called the ``light'' and ``heavy'' s-process peaks. They define the abundance indices [ ls] and [hs] for an object as the mean logarithmic abundances relative to iron of ( Y, Zr, and Sr) and ( Ba, Nd, La, and Sm), respectively, compared to their solar mean abundances. All of these elements are produced by the s-process, although several of them have predominantly r-process contributions for solar-type stars (Arlandini et al. 1999). Of the post Y Fe peak elements that are observed in PNe, Kr belongs to the light s-process peak near Sr, and Xe is near Ba in the heavy s-process peak. Both are produced by the s-process, although Xe is predominantly an r-process element for stars of solar metallicity. Using models to determine the appropriate correction factors, Kr and Xe can be incorporated into the [ ls] and [ hs] indices. However, in making comparisons of nebular abundances with those derived from stellar spectra, Fe should not be used as the fiducial abundance for nebulae because its consistently strong depletion from the gas phase due to grain formation causes its true abundance in nebulae to be indeterminate (Shields 1975; Perinotto et al. 1999). Argon is a good surrogate for Fe because it does not suffer depletion in nebulae, and it is well ob-

served in nebulae in multiple ionization stages and the relevant excitation cross sections are known. The extent to which s-process nucleosynthesis occurs in stars is determined by the total neutron exposure after the third dredge-up phase. Calculations show that neutron exposure is the primary factor that determines the ratio of the enhancements of the heavy to light s-process elements, i.e., [ hs/ ls], the ``neutron exposure Y related parameter.'' Low neutron exposures result in element production confined to a crowded region around the Sr peak, producing ½hs/ ls < 0, whereas higher neutron exposures significantly populate the Ba peak, leading to ½hs/ ls > 0 ( Busso et al. 1995, 1999). The fact that ½Xe/ Kr % 1 for PNe, in spite of Xe not being produced by the s-process as much as Kr, is strongly suggestive that the progenitor AGB stars of PNe experience significant neutron exposure. Of equal significance, the fact that forbidden lines of post Y Fe peak elements not observed in stars are detectable in nebulae causes H ii regions and PNe to be of potentially great value in studying the nucleosynthesis of these elements. 8. SUMMARY In summary, very high S/ N spectra of PNe do reveal lines from elements that are enhanced by the s-process, as was also found by other investigators. Because many post Y Fe peak elements are refractory, their gas-phase abundances are not reliable indicators of the nucleosynthetic processes that have occurred in the progenitor stars. Fortunately, the noble gases Ar, Kr, and Xe are largely unaffected by molecular and grain formation, and the latter two elements are situated in the light and heavy s-process peaks, respectively. They also have easily excitable and observable lines in multiple ionization stages for which atomic cross sections are now available and whose analyses should be straightforward. We find for a sample of five PNe that Kr and Xe abundances are enhanced over solar values by up to an order of magnitude, from both an analysis of their intensities relative to those of the fiducial element Ar and also a relative comparison of their line strengths in PNe versus H ii regions such as the Orion Nebula, where the lines are very weak in gas that is representative of the ISM composition. The similar enhancements of Kr and Xe in PNe are suggestive of large neutron exposures in the progenitor central stars. Further spectroscopy of PNe should reveal additional post Y Fe peak emission lines whose analyses will contribute to a more complete picture of post Y main-sequence stellar evolution, while deep spectra of H ii regions should lead to improved values of the ISM abundances of these elements.

The work of Y. Z. and X. W. L. was partially supported by Chinese NSFC grant 10325312, and Y. Z. gratefully acknowledges the award of an Institute Fellowship from STScI, where his work was carried out. E. P., K. C., and J. A. B. gratefully acknowledge support for this work from NSF grant AST 03-05833 and HST grant GO09736.02-A.

REFERENCES ´ Arlandini, C., Kappeler, F., Wisshak, K., Gallino, R., Lugaro, M., Busso, M., & Biemont, E., & Hansen, J. E. 1986b, Phys. Scr., 34, 116 ------ ´ Straniero, O. 1999, ApJ, 525, 886 Biemont, E., Hansen, J. E., Quinet, P., & Zeippen, C. J. 1995, A&AS, 111, 333 Baldwin, J. A., Verner, E. M., Verner, D. A., Ferland, G. J., Martin, P. G., Busso, M., Gallino, R., & Wasserburg, G. J. 1999, ARA&A, 37, 239 Korista, K. T., & Rubin, R. H. 2000, ApJS, 129, 229 Busso, M., Lambert, D. L., Beglio, L., Gallino, R., Raiteri, C. M., & Smith, V. V. ´ Baluteau, J.-P., Zavagno, A., Morisset, C., & Pequignot, D. 1995, A&A, 303, 175 1995, ApJ, 446, 775 Barge, T., Wahlgren, G. M., Johansson, S. G., Leckrone, D. S., & Proffitt, C. R. Cahn, J. H., Kaler, J. B., & Stanghellini, L. 1992, A&AS, 94, 399 1998, ApJ, 496, 1051 Churilov, S. S., & Joshi, Y. N. 2000, Phys. Scr., 62, 358 Benjamin, R. A., Skillman, E. D., & Smits, D. P. 1999, ApJ, 514, 307 Churilov, S. S., Joshi, Y. N., & Gayasov, R. 2002, J. Opt. Soc. Am. B, 18, 113 ´ Biemont, E., Cowan, R. D., & Hansen, J. E. 1988, Phys. Scr., 37, 850 Cosby, P. C., Sharpee, B. D., Slanger, T. G., Huestis, D. L., & Hanuschik, R. W. ´ Biemont, E., & Hansen, J. E. 1986a, Phys. Scr., 33, 117 2006, J. Geophys. Res., 111, A12307


1290

SHARPEE ET AL.
´ Pequignot, D., Petitjean, P., & Boisson, C. 1991, A&A, 251, 680 Perinotto, M., Bencini, C. G., Pasquali, A., Manchado, A., Rodriguez Espinosa, ´ J. M., & Stanga, R. 1999, A&A, 347, 967 Persson, W., & Petterson, S.-G. 1984, Phys. Scr., 29, 308 Persson, W., & Reader, J. 1986, J. Opt. Soc. Am. B, 3, 959 Persson, W., & Wahlstrom, C.-G. 1985, Phys. Scr., 31, 487 Persson, W., Wahlstrom, W.-G., Bertuccelli, G., DiRocco, H. O., Reyna Almandos, J. G., & Gallardo, M. 1988, Phys. Scr., 38, 347 Reader, J. 1983, J. Opt. Soc. Am., 73, 349 Reader, J., & Acquista, N. 1976, J. Opt. Soc. Am., 66, 896 Reader, J., & Epstein, G. L. 1972, J. Opt. Soc. Am., 62, 619 Robertson-Tessi, M., & Garnett, D. R. 2005, ApJS, 157, 371 Rodriguez, M. 2003, ApJ, 590, 296 ´ Safronova, U. I., Safronova, M. S., & Johnson, W. R. 2005, Phys. Rev. A, 71, 052506 Sansonetti, C. J., Reader, J., Tauheed, A., & Joshi, Y. N. 1993, J. Opt. Soc. Am. B, 10, 7 Schoning, T. 1997, A&AS, 122, 277 ¨ Schoning, T., & Butler, K. 1998, A&AS, 128, 581 ¨ Seaton, M. J. 1958, Rev. Mod. Phys., 30, 979 Sharpee, B., Baldwin, J. A., & Williams, R. 2004, ApJ, 615, 323 Sharpee, B., Williams, R., Baldwin, J. A., & van Hoof, P. A. M. 2003, ApJS, 149, 157 Shaw, R. A., & Dufour, R. J. 1995, PASP, 107, 896 Shields, G. A. 1975, ApJ, 195, 475 Smits, D. P. 1991, MNRAS, 251, 316 ------. 1996, MNRAS, 278, 683 Sterling, N. C., & Dinerstein, H. L. 2003, Rev. Mex. AA, 18, 133 ------. 2004, in ASP Conf. Ser. 313, Asymmetric Planetary Nebulae III, ed. M. Meixner et al. (San Francisco: ASP), 410 Sterling, N. C., Dinerstein, H. L., & Bowers, C. W. 2002, ApJ, 578, L55 Storey, P. J. 1994, A&A, 282, 999 Storey, P. J., & Hummer, D. G. 1995, MNRAS, 272, 41 Tauheed, A., & Joshi, Y. N. 1992, Phys. Scr., 46, 403 ------. 1993a, Phys. Rev. A, 47, 3092 ------. 1993b, Phys. Scr., 47, 550 Tauheed, A., Joshi, Y. N., & Kaufman, V. 1991, Phys. Scr., 44, 579 Tauheed, A., Joshi, Y. N., & Pinnington, E. H. 1993, Phys. Scr., 47, 555 Wilson, N. J., & Bell, K. L. 2002, MNRAS, 331, 389 Zethson, T., Gull, T. R., Hartman, H., Johansson, S., Davidson, K., & Ishibashi, K. 2001, AJ, 122, 322 Zhang, Y., & Liu, X.-W. 2002, MNRAS, 337, 499 ´ Zhang, Y., Liu, X.-W., Luo, S.-G., Pequignot, D., & Barlow, M. J. 2005, A&A, 442, 249 Zhang, Y., Liu, X.-W., Wesson, R., Storey, P. J., Liu, Y., & Danziger, I. J. 2004, MNRAS, 351, 935

Cosby, P. C., & Slanger, T. G. 2007, Canadian J. Phys., in press Davey, A. R., Storey, P. J., & Kisielius, R. 2000, A&AS, 142, 85 Dinerstein, H. L. 2001, ApJ, 550, L223 Esteban, C., Peimbert, M., Garcia-Rojas, J., Ruiz, M. T., Peimbert, A., & ´ Rodriguez, M. 2004, MNRAS, 355, 229 ´ Froese Fischer, C., & Tachiev, G. 2004, At. Data Nucl. Data Tables, 87, 1 Galavis, M. E., Mendoza, C., & Zeippen, C. J. 1995, A&AS, 111, 347 Gallard, M., Raineri, M., Reyna Almandos, J. G., Sobral, H., & Callegari, F. 1999, J. Quant. Spectrosc. Radiat. Transfer, 61, 319 Garcia-Rojas, J., Esteban, C., Peimbert, M., Rodriguez, M., Ruiz, M. T., & ´ ´ Peimbert, A. 2004, ApJS, 153, 501 Hamuy, M., Suntzeff, N. B., Heathcote, S. R., Walker, A. R., Gigoux, P., & Phillips, M. M. 1994, PASP, 106, 566 Hansen, J. E., & Persson, W. 1974, J. Opt. Soc. Am., 64, 696 ------. 1976, Phys. Scr., 13, 166 Hanuschik, R. W. 2003, A&A, 407, 1157 Hobbs, L. M., Thorburn, J. A., Oka, T., Barentine, J., Snow, T. P., & York, D. G. 2004, ApJ, 615, 947 Howarth, I. D. 1983, MNRAS, 203, 301 Hyung, S., Aller, L. H., Feibelman, W. A., & Lee, S.-J. 2001, ApJ, 563, 889 Hyung, S., Keyes, C. D., & Aller, L. H. 1995, MNRAS, 272, 49 Joshi, Y. N., Tauheed, A., & Davison, I. G. 1992, Canadian J. Phys., 70, 740 ´ Karlsson, H., & Litzen, U. 1999, Phys. Scr., 60, 321 Kaufman, V., Sugar, J., & Joshi, Y. N. 1988, J. Opt. Soc. Am. B, 5, 619 Kingdon, J., & Ferland, G. J. 1995, ApJ, 442, 714 Kisielius, R., & Storey, P. J. 2002, A&A, 387, 1135 Kisielius, R., Storey, P. J., Davey, A. R., & Neale, L. T. 1998, A&AS, 133, 257 Klose, J. Z., Fuhr, J. R., & Weise, W. L. 2002, J. Phys. Chem. Ref. Data, 31, 217 Lennon, D. J., & Burke, V. M. 1994, A&AS, 103, 273 Liu, X.-W. 2006, in Planetary Nebulae beyond the Milky Way, ed. J. Walsh, L. Stanghellini, & N. Douglas ( Berlin: Springer), 169 Lodders, K. 2003, ApJ, 591, 1220 Luck, R. E., & Bond, H. E. 1991, ApJS, 77, 515 Mendoza, C. 1983, in IAU Symp. 103, Planetary Nebulae, ed. D. R. Flower ( Dordrecht: Reidel), 143 Moore, C. E. 1952, Atomic Energy Levels, Vol. II, NBS Circ. 467 ( Washington, DC: NBS) ------. 1958, Atomic Energy Levels, Vol. III, NBS Circ. 467 ( Washington, DC: NBS) Nussbaumer, H., & Storey, P. J. 1984, A&AS, 56, 293 Osterbrock, D. E., & Ferland, G. J. 2006, Astrophysics of Gaseous Nebulae and Active Galactic Nuclei (2nd ed.; Sausalito: Univ. Science Books) Osterbrock, D. E., Fulbright, J. P., Martel, A. R., Keane, M. J., & Trager, S. C. 1996, PASP, 108, 277 Peimbert, M., Peimbert, A., Ruiz, M. T., & Esteban, C. 2004, ApJS, 150, 431 ´ Pequignot, D., & Baluteau, J. P. 1994, A&A, 283, 593 ( PB94)