Low-ionization pairs of knots in planetary nebulae: physical properties and excitation
aa r X i v : . [ a s t r o - ph . S R ] J u l Mon. Not. R. Astron. Soc. , 1–11 (2002) Printed 31 October 2018 (MN LaTEX style file v2.2)
Low-ionization pairs of knots in planetary nebulae:physical properties and excitation
D. R. Gon¸calves , A. Mampaso , R. L. M. Corradi , and C. Quireza UFRJ - Observat´orio do Valongo, Ladeira Pedro Antonio 43, 20080-090, Rio de Janeiro, Brazil Instituto de Astrof´ısica de Canarias, E-38205 La Laguna, Tenerife, Spain Isaac Newton Group of Telescopes, Apartado de Correos 321, E-38700 Sta. Cruz de La Palma, Spain Observat´orio Nacional, Rua General Jos´e Cristino 77, 20921-400 Rio de Janeiro, Brazil
Released 2002 Xxxxx XX
ABSTRACT
We obtained optical long-slit spectra of four planetary nebulae (PNe) with low-ionization pair of knots, namely He 1-1, IC 2149, KjPn 8 and NGC 7662.These data allow us to derive the physical parameters and excitation of the pairsof knots, and those of higher ionization inner components of the nebulae, separately.Our results are as follows. 1) The electron temperatures of the knots are withinthe range 9 500 to 14 500 K, similar to the temperatures of the higher ionizationrims/shells. 2) Typical knots’ densities are 500 to 2 000 cm − . 3) Empirical densitiesof the inner rims/shells are higher than those of the pairs of knots, by up to a factor of10. Theoretical predictions, at variance with the empirical results, suggest that knotsshould be denser than the inner regions, by at least a factor of 10. 4) Empirical andtheoretical density contrasts can be reconciled if we assume that at least 90% of theknots’ gas is neutral (likely composed of dust and molecules). 5) By using Raga et al.(2008) shock modeling and diagnostic diagrams appropriated for spatially resolvedPNe, we suggest that high-velocity shocked knots traveling in the photoionized outerregions of PNe can explain the emission of the pairs of knots analysed in this paper. Key words: ( ISM ): Planetary nebulae: individual: He 1-1, IC 2149, KjPn 8,NGC 7662. Planetary nebulae: structure.
The small-scale low-ionization structures (LIS) of planetarynebulae were classified in terms of their morphology andkinematics by Gon¸calves, Corradi & Mampaso (2001) asknots, jets and jetlike systems, either in pairs or isolated.From this, and subsequent works (e.g. Gon¸calves 2004, andreferences therein) it has became apparent that: i) around10% of the Galactic PNe are known to possess LIS; ii) theyare indistinctly spread among all the PNe morphologicalclasses; iii) 50% of these PNe have highly collimated, high-velocity jets, and/or high-velocity pairs of knots (FLIERs;fast, low-ionization emission regions, Balick et al. 1993); andiv) most of them are mainly photoionized.However, a number of key questions remain open: arethere significant density contrasts between LIS and thehigher ionization main components (rims, attached shells These are features whose morphology resemble highly super-sonic jets, but that share the expansion of the nebular compo-nents in which they are embedded. and detached haloes) of the nebula? What are the typicalelectron densities of the different types of LIS?We are involved in a long-term project for characteriz-ing the small-scale LIS of PNe since they can tell us muchabout the formation and evolution of PNe. The main issueswe can access by studying the LIS are:(i) the collimation processes of LIS in regard to those pro-cesses responsible for the shape of the PN itself, or the mass-loss processes during the AGB and post-AGB phases;(ii) the effect of the ionization front on the fossil AGBfeatures; and(iii) the role of disks and magnetic fields in shaping thehighly collimated outflows (jets) in PNe.Our motivation for the present analysis is to go a stepfurther on the study of the different LIS analyzing their den-sities. LIS can move through their environments either withhigh or low velocity. In the first case (the so-called FLIERS),radial velocities of 24 −
200 km s − with respect to thePN main components are typically measured (Balick et al.1993), whereas in the second instance (the SLOWERs, slowmoving low-ionization emission regions; see Perinotto 2000) c (cid:13) D. R. Gon¸calves et al.
Table 1.
Log of the INT+IDS observationsPN Name and P.A. Obs. Date Exposure (s)3 × He 1-1 - 315 ◦ Aug 31 300, 2400IC 2149 - 70 ◦ Sep 01 300, 900, 1800KjPn 8 - 98 ◦ Aug 30 1200KjPn 8 - 120 ◦ Aug 30 300, 1200NGC 7662 - 175 ◦ Sep 05 60, 300NGC 7662 - 248 ◦ Sep 04 30, 60, 300 the knots do not show peculiar velocities and share the typi-cal 30 km s − (for elliptical shells) to 100 km s − (for bipolarlobes) expansion velocities of their host PNe. A straightfor-ward question arises from this: how is the density contrastbetween the LIS and rims, shells or haloes correlated withthe different velocity regimes?In this paper we analyse 4 PNe that contain pairs ofknots, either of high- or low-velocities.We present our spectroscopic data and treatment, inSection 2. The line fluxes obtained for as many as possiblelarge- and small-scale components of the four PNe, and thederived physical properties for each of the components areshown in Sections 3 to 6. Finally, Section 7, is dedicatedto the discussion of the results in terms of the match withthe theoretical model predictions, and to our concluding re-marks. A 120 sec exposure H α + [N ii ] (6568/95˚A) image ofHe 1-1 (Figure 1) was retrieved from the IPHAS database(INT/WFC Photometric H α iii ] (F502N, 100 sec) images, obtained on 1995 (pro-posal 6119), see Figure 3. The [N ii ] (F658N, 400 sec) imageof NGC 7662 (Perinotto et al. 2004; Figure 5), was also re-trieved from the HST archive, and was obtained on 1996.The image of KjPn 8 was obtained with the 2.5-m NordicOptical Telescope (NOT) at the Observatorio del Roque delos Muchachos (European Northern Observatory, La Palma,Spain), with ALFOSC camera, in June 26, 2002. The imagewe show in Figure 4 is the result of one exposure of 1 800 sthrough the narrow-band filter [N ii ]6584˚A.Spectra of the 4 PNe were obtained on August 31 andSeptember 1, 4 and 5 of 2001, at the 2.5 m Isaac New-ton Telescope (INT) at the Observatorio del Roque de losMuchachos using the Intermediate Dispersion Spectrograph(IDS). The 235 mm camera and the R300V grating wereused, providing a spectral coverage from 3650 to 7000 ˚A witha spectral reciprocal dispersion of 3.3 ˚A pixel − . The spatialscale of the instrument was 0 . ′′
70 pixel − , with the TEK5CCD. Seeing varied from 0 . ′′ . ′′
1. The slit width andlength were 1 . ′′ ′ , respectively. These data were takenwith the slits positioned through the centre of the nebulae atthe position angles (P.A.), and with exposure times given inTable 1. Each of the exposures was taken 3 times, as listedin the table. The longer exposures were used to measure the fluxes of most emission lines. However, for almost all thePNe, the [O iii ] λλ ii ] λλ α and even the H β lines were satu-rated at the brightest features in the longer exposures, thus,the shorter exposures where used in these cases. Differentlyof the way we took long-slit spectra of the other PNe, in thecase of KjPn 8 the slit was centred on the position of thecentral star, and then it was shifted in order to cover each ofthe knots. Compared to the weather conditions during theobservation of the other three PNe, that of the KjPn 8 waspoor (clouds appeared during the integration).During the night, bias frames, twilight and tungstenflat-field exposures, wavelength calibrations, and exposuresof standard stars (BD +332642, Cyg OB2 No. 9, HD 19445,and BD +254655) were obtained. Spectra were reducedfollowing the IRAF instructions for long-slit spectra, be-ing bias-subtracted, flat-fielded, combined in order to im-prove the signal-to-noise ratio (S/N) and eliminate cos-mic rays, wavelength-calibrated, and sky-subtracted. Fi-nally, they were flux-calibrated using the above mentionedstandard stars and the mean atmospheric extinction curvefor La Palma. PN G055.3+02.7 (He 1-1) is a point-symmetric PN with apair of low-ionization knots. See in Figure 1 the extensionsof its bright Rim (or a combination of a rim and a shell),the large-scale component that dominates the emission ofthe nebula, and the SE and NW strings of knots, whichcorrespond to the small-scale faint structures seen in the [N ii ] NOT image.In the uppermost part of Table 2 we list selected linefluxes of the Rim, the knots and the entire portion of thenebula covered by the slit. These include all the electrondensity and temperature diagnostic emission lines coveredby the IDS spectra. Absolute H β fluxes, F H β , integratedalong the slit for each nebular component are also givenin Table 2. The Balmer lines ratios (H α /H β , H γ /H β andH δ /H β ) were then used to derive the c H β , the logarithmicratio between observed and dereddened H β fluxes. For thederivation of c H β , we assumed T e = 10 K and the densitiesgiven by the [S ii ] ratio of each region. Theoretical Balmerline ratios from Osterbrock & Ferland (2006) and the red-dening law of Cardelli, Clayton & Mathis (1989) were used.The weighted average of c H β , per PN component, is the onegiven in the table. c H β is constant along the slit, and the c H β =1.49 ± .
16 for the entire nebula is lower than the valuepreviously published by Tylenda et al. (1992). The latterauthors found a range of values for c H β , which goes from1.9 to 2.2. Note, however, that their values are based onlyon the H α /H β ratio. Fluxes were then dereddened using thederived c H β . The dereddened fluxes, or intensities, are notshown in the table. Instead, figures in the table correspondto the observed fluxes.The middle part of Table 2 shows the errors on thefluxes. They were calculated taking into account the statis-tical errors in the measurements, as well as systematic errorsof the flux calibrations, background determination, and skysubtraction.In the lowermost part of Table 2 electron densities and c (cid:13) , 1–11 hysical properties and excitation of pairs of knots SEk NWkP.A.=315 o Rim NE
Figure 1.
He 1-1: IPHAS H α +[N ii ] image. The 3 structures un-der analysis are indicated in the image. The size of the extractionwindows in the spectra includes 4 . ′′ . ′′ ′′ , and from -4 . ′′ . ′′ . ′′ temperatures are presented. At least two different measuresof T e and N e are available for each PN component. Theyrepresent regions of low (N e [S ii ], T e [N ii ] and T e [S ii ]) andmedium excitation (N e [Cl iii ] and T e [O iii ]).The density of the nebula can be compared withSamland et al. (1992). They find N e [S ii ]=1.19 × cm − ,in moderate agreement with ours. Figures for T e and N e of the knots were previously published only in a proceedingcontribution by Ben´ıtez et al. (2002): 11 300 K for the tem-perature of the NW knot, and densities of 300 cm − and600 cm − , for the SE and NW knots respectively. These val-ues hardly agree with ours. Note that they did not derive,but assumed the 10 K of the SE knot.Figure 2 (top left panel) and Table 2 shows that bothdensity measurements in each region of the nebula are equal
Table 2.
He 1-1: Observed line fluxes, flux errors, c H β , N e and T e . Line ID Rim Knots NEBSEk NWk[S ii ] 4072.0 6.232 - - 6.464H δ γ iii ] 4363.2 11.36 16.89 14.61 13.04H β iii ] 4958.9 543.4 587.7 601.7 566.8[O iii ] 5006.86 1715. 1773. 1808. 1774.[CL iii ] 5517.7 3.321 22.56 15.57 4.152[CL iii ] 5537.9 3.041 18.23 13.04 3.640[N ii ] 5754.6 11.90 23.46 24.87 13.83[N ii ] 6548.0 319.2 480.6 413.9 335.1H α ii ] 6583.4 1034. 1436. 1321. 1075.[S ii ] 6716.5 129.3 114.4 124.4 132.3[S ii ] 6730.8 169.6 114.7 139.2 172.1F H β a H β ± ± ± ± β
35 52 60 31(0.05–0.15)H β
15 25 27 13(0.15–0.30)H β
11 20 20 11(0.30–2.0)H β
10 17 16.5 8.5(2.0–5.0)H β β > β − ) and Temperatures (K)N e [S ii ] 1600 ±
230 600 ±
140 900 ±
210 1550 ± e [Cl iii ] 1700 ±
830 800 ±
230 1000 ±
330 1350 ± e [O iii ] 12500 ± ± ± ± e [N ii ] 10800 ± ± ± ± e [S ii ] 8750 ± ± a In units of 10 − erg cm − s − within the errors, whereas N e of the Rim is higher than thatof the SE and NW knots, by factors of ∼ ∼ e at the centre could be some-what lower than at the knots, but the variation is withinthe errors.Summarising, the pair of knots in He 1-1 are substan-tially less dense than the Rim, whereas no strong evidenceexist for significant temperature variations from the Rim tothe knots. PN G166.1+10.4 has an apparent shape that does not easilycompare with the classification bins (round, elliptical, bipo-lar or quadrupolar, irregular and point-symmetric) usuallyadopted for PNe. Kinematic modelling suggests that it isa bipolar PN (V´azquez et al. 2002; Feibelman et al. 1994;Zhang & Kwok 1998). As shown in Figure 3, IC 2149 is com-posed by a bright higher ionization emission zone, the Core,and by a pair of lower excitation knots, the North-Easternone being much brighter than its counterpart to the South-West. In fact only the two former structures can be identifiedfrom Figure 3. Please refer to Fig.1 of V´azquez et al. (2002),in which not only a higher excitation image is show, but also c (cid:13) , 1–11 D. R. Gon¸calves et al.
Figure 2.
Electron densities and temperatures as a function of the distance to the centre of the nebula. The different structures underanalysis are shown as symbols that correspond to the centre of each structure, as defined in the images (Figures 1, 3, 4, and 5). Filled(N e [S ii ]) and open (N e [Cl iii ]) triangles represent densities, while filled (T e [N ii ]) and open (T e [O iii ]) circles, as well as the filled box(T e [S ii ]) represent electron temperatures. Symbols are plotted slightly displaced in distance in order to avoid overlapping. Horizontallines represent densities and temperatures averaged along the slit (NEB): continuous line for N e [S ii ], large-small dashed line for N e [Cl iii ], small dashed line for T e [O iii ], medium dashed line for T e [N ii ], and large dashed line for T e [S ii ]. TOP LEFT: He 1-1 - from left toright the NW knot, the Rim and the SE knot, respectively, as in Figure 1. BOTTOM LEFT: IC 2149 - From left to right, the NE knot,the Core and the SW knot as in Figure 3. TOP RIGHT: NGC 7662 - P.A.=175 ◦ ; five zones of the Inner rim and Outer shell, as shownin Figure 5 and Table 5. BOTTOM RIGHT:NGC 7662 - P.A.=248 ◦ ; further five zones of the Inner rim and FLIERs, following Figure 5and Table 5. the lower excitation [N ii ] and [S ii ] are presented. In their[N ii ] image the SW LIS is also clearly seen.Table 3 contains the spectroscopic results for IC 2149,obtained from the INT+IDS spectra at P.A.=70 ◦ (observedand absolute H β fluxes, flux errors, c H β , N e and T e ) for theCore and pair of LIS of the nebula.The H α /H β ratio was used to obtain the c H β of theNEk, SWk and NEB regions. Owing to the presence ofBalmer line absorption in the central zones (V´azquez et al. 2002), the c H β of the Core was adopted as the averagefrom the knots’ values. The value measured for the nebula c H β =0.12 ± .
01 is in agreement with other published values(Feibelman et al. 1994; Ciardullo et al. 1999), but it is muchsmaller than the value (0.41) obtained by V´azquez et al.(2002) for the Core and the knots.The gas physical conditions we derived to this PN (Ta-ble 3 and left bottom panel of Fig. 2) compare relatively wellwith those derived by V´azquez et al. (2002) for the North- c (cid:13) , 1–11 hysical properties and excitation of pairs of knots P.A.=70 o SWkNEk Core E N
Figure 3.
IC 2149: HST 555W (V) image. Since this filter iscentred in [O iii ], only the Core is clearly identified in this image.The low-ionization NEk and its fainter SWk counterpart are hardto see in this image, but are easily identified in the [N ii ] imagein Fig.1 of V´azquez et al. (2002). The spectroscopic sizes of theCore, NE and SW knots are, respectively: 2 . ′′ . ′′ . ′′
4; and + 2 . ′′ . ′′
3. The entire PN, NEB, is14 . ′′ Eastern knot and for the NEB. The exception is the N e [S ii ], for which they obtained values of 10 000 and 9 600 cm − while we found 5 350 ± ±
660 cm − , for theNEk and NEB, respectively. We cannot understand the rea-son of this discrepancies, since they do not depend on thevery different c H β applied to correct the fluxes. On theother hand, the T e that would be affected by our differentchoices in terms of c H β , are in reasonable agreement withtheirs (see V´azquez et al. 2002 Table 1). If only NEB is con-cerned, our results also agree with the results obtained byFeibelman et al. (1994).As [Cl III] λλ e [S ii ] could be derived for the Core. Figure 2 shows that thetwo density diagnostics for each of the knots are such thatSWk (N e [S ii ]=2 050 ± e [S ii ]=5 350 ± ± e [O iii ] and T e [N ii ] from the North-Eastern to the South-Western side of thenebula, variations are within the errors. PN G112.5-00.1 (KjPn 8) is a huge PN (size ∼ ′ × ′ ; L´opez, V´azquez & Rodr´ıguez 1995) with multiple po-lar ejections. A number of papers deal with the character-istics of this PN, including: i) narrow-band imaging of itsmany structures in the different scales, L´opez et al. (1995);ii) proper motions of its latest bipolar ejections, Meaburn(1997); iii) kinematics of the large bipolar lobes and of the Table 3.
IC 2149: Observed line fluxes, flux errors, c H β , N e and T e Line ID Core Knots NEBNEk SWk[S ii ] 4072.0 - 5.743 5.625 3.011H δ γ iii ] 4363.2 3.838 2.457 5.333 3.850H β iii ] 4958.9 180.1 124.5 140.7 150.9[O iii ] 5006.86 540.2 378.4 408.4 454.0[CL iii ] 5517.7 - 0.367 0.545 0.386[CL iii ] 5537.9 - 0.516 0.481 0.440[N ii ] 5754.6 1.140 1.633 1.227 1.299[N ii ] 6548.0 15.03 30.27 18.55 22.06H α ii ] 6583.4 43.07 91.54 51.81 65.23[S ii ] 6716.5 1.288 3.201 2.418 2.261[S ii ] 6730.8 2.312 5.607 3.385 3.860F H β a H β b ± ± β
17 24 39 10(0.05–0.15)H β
10 17 24 5.5(0.15–0.30)H β β β β >
10 H β − ) and Temperatures (K)N e [S ii ] 6050 ± ± ± ± e [Cl iii ] - 7600 ± ±
870 4000 ± e [O iii ] 10350 ± ± ± ± e [N ii ] 12300 ± ± ± ± a In units of 10 − erg cm − s − b This c H β values is the average between the knots’ values; see text. pair of knots, L´opez et al. (1997); and iv) temperature, den-sity and chemical abundances of the inner nebula, V´azquez,Kingsburgh & L´opez (1998). At much smaller scales, a ringthat seems to collimate the many structures (with its COand H counterpart) has been detected giving support tothe idea that KjPn 8 had two PNe-like events in its history(L´opez et al. (2000)). However, N e and T e of its two pairsof low-ionization knots were not properly measured so far,due to their extreme faintness.In our spectra only one of the knots of each pair could bemeasured with a good S/N. At variance with the other PNein our sample, no analysis of the internal or entire nebulawill be given, because of the huge extension of the PN. Ourresults for the two SE knots of KjPn 8 are given in Table 4.When comparing our c H β values with those in the lit-erature, particular care should be taken because they usu-ally refer to different nebular regions, and reddening mayvary along this large nebula. V´azquez et al. (1998) measuredvery different c H β values for the three knots which they la-belled A1, A3 and B1: c H β = 1.49, 0.96 and 0.56, respec-tively. Our c H β for SEk (their A1) is therefore 3 timessmaller than theirs, whereas for SEk (their knot B1) bothdeterminations agree very well, c H β =0.56 and 0.52 ± . c H β in SEk is not clear at present. As for the density, they reportedN e [S ii ]=100 cm − for both SEk and SEk , while we de-rived substantially larger values N e [S ii ]=600 ±
140 ([Cl iii ]=450 ± ±
90 for SEk and SEk , respectively. c (cid:13) , 1–11 D. R. Gon¸calves et al.
E NSEkSEk
98 120
P.A.=98 o P.A.=120 o Figure 4.
KjPn 8: NOT [N ii ] image. The two P.A. of the struc-tures under analysis are indicated in the image. The spectroscopicsizes of the two knots, SEk is 14 ′′ and that of SEk is 10 . ′′ Table 4.
KjPn8: Observed line fluxes, flux errors, c H β , N e and T e Line ID South Eastern KnotsSEk
SEk [S ii ] 4072.0 10.78 8.869H δ γ iii ] 4363.2 1.149 2.466H β iii ] 4958.9 100.5 101.8[O iii ] 5006.86 307.9 307.4[CL iii ] 5517.7 - 3.000[CL iii ] 5537.9 - 2.280[N ii ] 5754.6 9.351 12.95[N ii ] 6548.0 392.8 398.0H α ii ] 6583.4 1191. 1207.[S ii ] 6716.5 99.79 98.28[S ii ] 6730.8 100.4 98.19F H β a H β ± ± β β β β β β > β − ) and Temperatures (K)N e [S ii ] 600 ±
90 600 ± e [Cl iii ] - 450 ± e [O iii ] 9100 ± ± e [N ii ] 8500 ± ± e [S ii ] 9150 ± ± a In units of 10 − erg cm − s − The electron temperature in the knots is measured in thispaper for the first time.Summarising, the densities and temperatures of the twoknots are the same within the errors.
This bright nebula (PN G106.5-17.6) consist of a centralcavity surrounded by three concentric shells, the inner Rim,the outer shell (Guerrero, Jaxon & Chu 2004), and a halo(Corradi et al. 2004). In addition, on smaller scales it hasa number of LIS. Most of them appear, in projection, dis-tributed at the outer edge of the outer shell. All these struc-tures are easily identified in Figure 5, and are named asin Balick et al. (1998) and Perinotto et al. (2004). This fig-ure also identify the structures that we analyze in this pa-per. Balick, Preston & Icke (1987) studied the kinematics ofthe different nebular components, from which the acronymsFLIERs and SLOWERs were later introduced.Perinotto et al. (2004) studied this PN with unprece-dented spatial resolution using the STIS spectrograph onthe HST. However, their spectra were taken through twoslits (called POS1 and POS2, see their Figure 1), that interms of LIS, are restricted to only one side of the nebula.We have used much longer slits that cover all the structuresalong both sides of the nebula, in each direction at P.A. 175 ◦ and 248 ◦ . This allows us to compare the densities and tem-peratures of any given symmetrical pair of LIS. The slit atP.A. 248 ◦ contains a pair of low-ionization filaments (theserpentine-like filaments described in Balick et al. 1998).The slit at P.A. 175 ◦ does not contain any LIS, and is used tocompare the physical parameters along the line joining theFLIERs with those in a direction free of micro structures.In this way, the present work can be seen as complementaryto the analysis of Perinotto et al. (2004).In Table 5 we report the derived values of c H β that arebased on the average weighted c H β given by the H α , H γ ,and H δ to H β ratios. These values are all between 0.13 and0.21 for the 5 regions under analysis in each position angle.These c H β agree very well with those previously published:0.18 - 0.22 (Tylenda et al. 1992); 0.1 (Hyung & Aller 1997);0.2 (Perinotto et al. 2004); and 0.18 Zhang et al. (2004).This table gives the observed line fluxes, densities andtemperatures of the 10 structures we measured along P.A.175 ◦ and 248 ◦ . These parameters are also presented in Fig-ure 2. The average T e [O iii ] derived from the emission in-tegrated along both P.A. are the same within errors (T e [O iii ]=14 300 ±
700 K and 13 400 ±
680 K, respectively). In factthese values also agree with the T e [N ii ]=12 750 ± ◦ . The T e [O iii ] we obtained hereare more in agreement with the results by Barker (1986) –whose [O iii ] temperatures in different regions of the nebula(see his Table 3) varies from 11 200 K to 13 800 K – andZhang et al. (2004) (13 300 K), than with the somewhatlower values of Perinotto et al. (2004).As for the [S ii ] and [Cl iii ] densities of the integratedemission we found 2 300 ±
440 cm − and 2 550 ±
490 cm − for P.A.=175 ◦ and 2 850 ±
450 cm − and 2 850 ±
460 cm − for P.A.=248 ◦ . N e [S ii ] results very closely reproduce thosefound by a Wang et al. (2004, N e [S ii ]=2 884 cm − ). The c (cid:13) , 1–11 hysical properties and excitation of pairs of knots Inner rim Outer shellCavityP.A.=175 P.A.=248 o o
E N NE FLIER SW FLIER
Figure 5.
NGC 7662: HST archive images.
Top : [O iii ]. Bottom :[N ii ]. In both images the box corresponds to 34 ×
34 arcsec .The size of the extraction windows in the spectra are as follows. P.A.=175 ◦ (inner rim + outer FLIER) : -14 . ′′ . ′′
5, SE outershell; -8 . ′′ . ′′
8, SE inner rim; -1 . ′′
75 to +1 . ′′
75, Cavity + Star;+2 . ′′ . ′′
4, NW inner rim; +12 . ′′ . ′′
7, NW outer shell.
P.A.=248 ◦ (inner rim + outer shell) : -14 . ′′ . ′′
2, SW FLIER;-9 . ′′ . ′′
8, SW inner rim; -2 . ′′ . ′′
1, Cavity + Star; +3 . ′′ . ′′
8, NE inner rim; +10 . ′′ . ′′
4, NE FLIERS. same authors give a value of 1 862 cm − for the [Cl iii ]density, but errors are not reported in their work.Focusing on the densities and temperatures of the innerRim and outer shell with respect to those of the FLIERs(pair of filaments at P.A.=248 ◦ ), it is clear from Figure 2,that: i) temperatures T e [O iii ](where errors are smaller) areslightly higher at the inner Rim than in the outer knots(P.A.=175 ◦ ), but this is just at the 2- σ level of the adoptederrors; ii) densities of the Rim agree very well one to anotherin both position angles; iii) the Rim is roughly a factor of 2.8denser than the outer shell (P.A.=175 ◦ ); iv) the Rim (innershell) along P.A.=248 ◦ , is a factor of ∼ e =2200 cm − for the two FLIERs studied by Perinotto et al.(2004), and 2 100 cm − in our work. Perinotto et al. (2004)did find significantly different densities for FLIERs andSLOWERs, being the latter approximately 1.3 times denserthan the former.Our analysis shows that FLIERS of NGC 7662 are lessdense than the inner regions of the nebula, but they are atthe same time much denser than the PN outer shell in whichthey are located. We have estimated individual temperatures T e [O iii ] andT e [N ii ] for the different regions of the four nebulae studiedin this paper. As it is well known, temperature errors arealways important and they range in our case typically fromaround 1000 K for T e [O iii ] in the brighter zones to morethan 4000 K for T e [N ii ] in the fainter regions. Therefore, asensible comparison of temperatures for the different com-ponents of each PN is precluded, and only rather generalconclusions can be extracted: 1) temperatures are those ofnormal photoionized PNe, varying from 9 500 to 14 500 Ktypically; 2) T e [O iii ] agrees within the errors with T e [N ii ]for every PN component measured. 3) No relevant differ-ences between T e [O iii ] at the LIS and at the higher excita-tion PN structures are found. The empirical density contrasts between the knots and theinner nebular regions, as described in this paper, are suchthat knots are typically 1 to 3 times less dense than the rimsor cores of these PNe (He 1-1, IC 2149 and NGC 7662): inno one of the LIS measured here the electronic density isdefinitely larger than at the main PN structures. This un-expected result is nevertheless also found in the full sampleof PN with LIS. Table 4 of Gon¸calves et al. (2001) lists 26PNe with pairs of knots similar to those in the present work.We have searched in the literature for density measures bothat the inner rims (or shells) of each PNe and at their pairsof knots, finding eight PNe (in addition to the four neb-ulae here studied) with adequate data. In all but one PN(NGC 6826) the densities of the knots are equal or lower than those of the inner rims or shells, by factors between 1and 10.This result is not consistent with the theoretical ex-pectations. The formation of PN major structures andpairs of jets/knots has been modeled in the past usinga broad variety of processes and scenarios (see, for in-stance, Gon¸calves et al. (2001) and Balick & Frank (2002)). NGC 4634 (Guerrero, Jaxon & Chu 2004); Hb 4(Hajian et al. 1997); IC 4593 (Corradi et al. 1997); NGC 7009(Gon¸calves et al. 2003); K 3-35 (Miranda et al. 2000); NGC 6826(Balick et al. 1994); NGC 2440 (Cuesta & Phillips 2000); K 1-2(Exter, Pollacco & Bell 2003).c (cid:13) , 1–11
D. R. Gon¸calves et al.
Table 5.
NGC 7662 Observed line fluxes, flux errors, c H β , N e and T e . P.A.=175 ◦ (Rim andOuter shell) and P.A.=248 ◦ (Rim + FLIERs) P.A.=175 ◦ Line ID SE outer Shell SE inner Rim Cavity + Star NW inner Rim NW outer Shell NEBH δ γ iii ] 4363.2 17.17 15.66 19.78 18.46 19.12 17.82H β iii ] 4958.9 504.4 368.0 343.2 338.7 481.7 372.1[O iii ] 5006.86 1533. 1105. 1018. 1018. 1434. 1114.[CL iii ] 5517.7 0.761 0.317 1.756 0.345 0.398 0.424[CL iii ] 5537.9 0.652 0.344 1.669 0.378 0.475 0.423[N ii ] 5754.6 0.169 0.087 0.469 0.099 0.152 -H α ii ] 6583.4 7.496 2.981 2.502 2.540 5.191 3.115[S ii ] 6716.5 1.164 0.366 0.242 0.249 0.916 0.371[S ii ] 6730.8 1.428 0.593 0.373 0.396 1.107 0.536F H βa H β ± ± ± ± ± ± β
35 15.5 12.5 16.5 30 8.5(0.05–0.15)H β
16 10.5 10 9.5 11 4.5(0.15–0.30)H β
10 8.5 9 8.5 9 4(0.30–2.0)H β β β > β − ) and Temperatures (K)N e [S ii ] 1300 ±
735 3650 ± ±
730 3400 ± ±
560 2300 ± e [Cl iii ] 1250 ±
700 3450 ± ±
370 3550 ± ±
550 2550 ± e [O iii ] 12400 ± ± ± ± ± ± e [N ii ] 12650 ± ± ± ± ◦ Line ID SW outer SW inner Cavity + Star NE inner NE outer NEBFLIER Rim Rim FLIER[S ii ] 4068.6 1.420 0.885 0.814 0.920 1.263 1.232[S ii ] 4076.4 0.530 0.469 0.966 0.430 0.317 0.259[S ii ] 4072.0 1.923 0.954 2.127 1.131 1.381 1.804H δ γ iii ] 4363.2 21.83 17.27 16.43 16.53 18.89 16.98H β iii ] 4958.9 552.7 387.6 339.8 398.1 517.8 397.1[O iii ] 5006.86 1706. 1184. 1040. 1214. 1570. 1213.[CL iii ] 5517.7 0.637 0.395 0.408 0.389 0.606 0.410[CL iii ] 5537.9 0.585 0.431 0.390 0.437 0.570 0.422[N ii ] 5754.6 0.585 0.088 - 0.089 0.552 0.155H α ii ] 6583.4 31.14 3.551 2.734 4.289 22.88 6.321[S ii ] 6716.5 2.630 0.454 0.311 0.431 2.141 0.611[S ii ] 6730.8 3.843 0.751 0.423 0.679 3.064 0.932F H βa H β ± ± ± ± ± ± β β β
13 6.5 7.0 5.5 8 4.5(0.30–2.0)H β β β > β − ) and Temperatures (K)N e [S ii ] 2400 ± ±
650 1850 ±
550 3250 ±
550 2250 ±
800 2850 ± e [Cl iii ] 1800 ± ±
580 2150 ±
630 3800 ± ±
790 2850 ± e [O iii ] 14750 ± ±
970 14100 ± ±
820 12700 ± ± e [N ii ] 11350 ± ± ± ± ± a In units of 10 − erg cm − s − . Note that [N ii ] 6548˚A was not measured for any of the structures. Thus, T e [N ii ]were calculated assuming the theoretical relation between [N ii ] 6583˚A and [N ii ] 6548˚A. c (cid:13) , 1–11 hysical properties and excitation of pairs of knots -4-3-2-10 Data from literature FLIERs Rims and ShellsRaga et al. (2008)Models: a/b/c100 A/B/C100 A/B/C150 C70,C40-2 -1 0 1-2-101 This work:Inner regions He 1-1 IC 2149 NGC 7662Outer shells NGC 7662This work:Pairs of knots He 1-1 IC 2149 KjPn8 NGC 7662-2 -1 0 1 l og ( [ S II] ( + ) / H ) l og ( [ N II] / H ) log([O III] 5007/H ) log([O III] 5007/H ) Figure 6.
Diagnostic diagrams for spatially resolved PNe. Left panels are an adaptation of the Raga et al. (2008) results, whereas rightpanels superpose to these results the data for our sample. The Raga et al. (2008) simulations for high-velocity shocked knots that travelthrough a photoionized region of a PN involve four families of models, as indicated in the labels of the upper left panel. Coded intheir names, the numbers represent the knot’s velocity, from 40 to 150 km s − whereas letters from ‘a’ to ‘C’ indicate different stellarluminosities and temperatures which are kept constant in each model while the knot’s distance to the ionizing source decreases. Therefore,model ‘a100’ has the lowest, and ‘C100’ the highest, photoionization rate, while both assume a velocity of 100 km s − for the knots. Acompilation of empirical literature data performed by Raga et al. (2008) is also sketched in the four panels as regions filled with diagonallines. In the two right panels, different regions of the PNe of our sample are added: empty symbols correspond to the pairs of knots,filled symbols to the nebular inner regions (Rim, Core, Cavity + Star), and half-filled circles correspond to the outer shells. Recent models can be grouped into three main families.i) Magnetohydrodynamic disk wind models, based on thescenario of accretion disks formed by binary interactions(Frank & Blackman 2004, Blackman, Frank & Welch 2001).ii) Interacting AGB and post-AGB wind models, in whichthe post-AGB wind is driven exclusively by the magneticpressure of a single star (Garc´ıa-Segura, L´opez & Franco2005, Garc´ıa-D´ıaz et al. 2008). iii) Models that explore theeffect of axisymmetrical light (low density) jets that, for ashort period of time, expand into a spherical AGB wind(Akashi & Soker 2008). Some of the above models are par-ticularly tailored to reproduce the main morphological andkinematic properties of a few well observed PNe: He 3-401,M2-9 and He 2-90 (Garc´ıa-Segura et al. 2005); Mz 3 and thepre-PN M1-92 (Akashi & Soker 2008). Nevertheless, and de-spite the different assumptions of the different models, all ofthem are able to account for the formation of PNe with jets.Model predictions state that, after the eventual shut offof the jets, dense knots would form embedded in the outernebular components. Moreover, both the knots present atthe tips of the existing jets and those left over when jets dis-appear are predicted to be denser than the inner PN compo- nents by large factors ranging from 10 up to 1000 (typically10) depending on the model assumptions.Therefore, there is a clear disagreement between ob-served and modeled pairs of knots: the former are foundto be typically 10 times less dense, while the latter are ex-pected to be typically 10 times denser, than the inner rimsand shells of their parent PN.A straightforward possibility to reconcile the empiricaland theoretical density contrasts is keeping in mind thatempirical densities, as those measured here, correspond toonly the ionized fraction of the PN gas. Models, on the otherhand, usually predict the formation of knots accounting forits morphology, kinematics and total gas content, but donot detail the ionization status of each component nor theirevolution with time. The strong discrepancy between mod-els and observations found here (a typical factor of 100) cantherefore be explained if 90% or more of the low-ionizationknots’ material is neutral. It is worth mentioning here thecase of one prominent cometary globule of the Helix neb-ula –which contains the best known low-ionization knotsof PNe. The molecular matter (H + CO; Huggins et al.2002) and the dust content (Meaburn et al. 1992) of this c (cid:13) , 1–11 D. R. Gon¸calves et al. knot amount to ∼ × − M ⊙ , whereas O’Dell & Handron(1996) estimated that the typical mass of the photoionizedgas of a globule in Helix is ∼ − M ⊙ .In another well-studied system of knots, NGC 7009, pre-vious modeling by Gon¸calves et al. (2006) succeeded in re-producing the observed ionization structure and chemicalabundances under the assumption that the total gas densityof the knots is equal to the one empirically derived from the[S ii ] lines (Gon¸calves et al. 2003), therefore, not account-ing for any neutral matter. This apparently contradicts thepresent results suggesting that the knots might be mostlyneutral. However, the effect of the density assumed for theknots in the modeling by Gon¸calves et al. (2006) is partlybalanced by its geometry, i.e., by its shape and size. As theshape was fixed based on the knots’ appearance in the HSTimages, the size was allowed to vary until it fits the observa-tions (i.e. the optical line ratios and electron density). Forthat reason Gon¸calves et al. (2006) were able to reproducethe observations of NGC 7009 without invoking need forneutral matter. It would be very interesting to attempt 3Dphotoionization modeling as in Gon¸calves et al. (2006) ex-ploring high density regimes for the knots as suggested inthis paper. Raga et al. (2008) have recently simulated the evolution ofhigh-density (10 cm − ), high-velocity (100 to 150 km s − )knots that travel away from a photoionizing source crossinga uniform lower density (10 cm − ) environment. They ex-plored a range of initial conditions for the shocked knots (interms of photoinization rate), and qualitatively reproducedthe emission-line ratios of several FLIERs, rims and shellsobserved in pre-PNe and PNe. At the end of the model evo-lution, 400 yr, the number densities for the knots are 10 to10 cm − , broadly in agreement with measured N e of knotsin NGC 7009 and IC 4634. These simulations were indeedintended to reproduce the line ratios of the FLIERs of thesetwo PNe. No predictions are given for the properties of theother PN components (rims, shells and haloes). In Figure 6we show two of the diagnostic diagrams from Raga et al.(2008) comparing empirical and theoretical line ratios forFLIERs, as well as showing a compilation of empirical datafor rims and shells of some PNe.Fig. 6 shows, first, that most of the inner regions (rimsand shells) of the PNe in our sample occupy the same zoneas the rims and shells of other PNe in the literature. Sec-ond, the same is true for the zone occupied by the pairs ofknots in He 1-1 and NGC 7662 (open boxes and circles) withrespect to the zones of the FLIERs sample in Raga et al.(2008). Third, knots in the other two nebulae, IC 2149 andKjPn 8 (open triangles and stars), are displaced from theempirical zone of FLIERs in Raga et al. (2008). Their lo-cations are consistent with the results of simulations withthe highest initial velocity of the knots (150 km s − , modelA/B/C/150), and indicate that a mixing of shock and photoionization is playing a role in the measured emission-line ra-tios. In fact, the highly supersonic velocity (radial velocityof 226 km s − ; L´opez et al. 1997) measured in KjPn 8 forthe SE knot at P.A.=98 ◦ could explain this agreement. Ex-pansion velocities of IC 2149’s knots are more moderate ( ∼
47 km s − ; V´azquez et al. 2002). Raga et al. (2008) show that models with low pho-toionization rates yield line ratios typical of shock-excitedknots, whereas in models with higher photoionization rate,photoionized emission dominates the spectrum of the high-velocity, high-density knots. This strengthen the idea that itis evolution (PN age; see Gon¸calves 2004, Raga et al. 2008;Viironen et al. 2009) and photoionization rate (Dopita 1997;Raga et al. 2008), rather than the highly supersonic veloci-ties of the knots, the key parameters to separate the shock-excited from the photoionized structures in planetary nebu-lae. We have derived the physical parameters of low-ionizationpairs of knots, and of higher ionization PN main componentsfor a small sample of nebulae (He 1-1, IC 2149, KjPn 8 andNGC 7662). We have shown that electron temperatures atthe knots are comparable to the temperatures of the mainnebular components, rims and shells, whereas densities inthe latter structures are significantly higher than in the pairsof knots, by up to a factor of 2. Typical knots’ densities inthe studied PNe are 500 to 2 000 cm − .We have argued that optical line ratios are not appropri-ate to constrain the total gas density, as they only accountfor the ionized matter of the knots, whereas none of theavailable models properly explore the knots ionization frac-tion, but the total (neutral plus ionized) density structure.A direct comparison of our results, plus compiled literaturedata, with the theoretical density contrasts expected frommodel predictions, yields a clear discrepancy that amount,typically, up to a factor of 100. To reconcile models andobservations we suggest that an important fraction of thegas at the knots –90%– should be neutral. The results pre-sented here can help in constraining more realistic modelsfor jets and knots in PNe, by providing empirical physicalparameters of the ionized gas.We have analyzed the location of the knots of our fourPNe in diagnostic diagrams studied by Raga et al. (2008)concluding that the observed line emission ratios are com-patible with shocked knots traveling in the photoionizedouter regions of their host PNe. ACKNOWLEDGMENTS
We would like to thank the anonymous referee of thismanuscript, for his suggestions. D.R.G. thanks the Brazilianagency FAPERJ (E-26/110.107/2008) for its partial sup-port. A.M., and R.L.M.C. acknowledge funding from theSpanish Ministry of Science AYA2007-66804 grant. This pa-per makes use of data obtained at the 2.5 m Isaac NewtonTelescope (INT: IDS and IPHAS), operated by the IsaacNewton Group; and at the 2.5 m Nordic Optical Telescope(NOT: ALFOSC), operated by NOTSA. Both are telescopesof the European Northern Observatory on the island of LaPalma in the Spanish Observatorio del Roque de los Mucha-chos of the Instituto de Astrof´ısica de Canarias. We alsouse NASA/ESA Hubble Space Telescope data, obtained atthe Space Telescope Science Institute, which is operated byAURA for NASA under contract NAS5-26555. c (cid:13) , 1–11 hysical properties and excitation of pairs of knots REFERENCES
Akashi M., & Soker, N. 2008, MNRAS in pressBalick B., & Frank A., 2002, ARA&A, 40, 439Balick, B., Preston, H. L., Icke, V., 1987, AJ, 94, 1641Balick, B., Rugers, M., Terzian, Y., Chengalur, J. N., 1993,ApJ, 411, 778Balick B., Perinotto M., Maccioni A., Alexander J., TerzianY., Hajian A. R., 1994, ApJ, 424, 800Balick, B., Alexander, J., Hajian, A. R., Terzian, Y.,Perinotto, M., & Patriarchi, P., 1998, AJ, 116, 360Barker, T., 1986, ApJ, 308, 314Ben´ıtez G., V´azquez R., Cook R., & Olgu´ın L., 2002,RMxAC 12, 159Blackman E. G., Frank A., & Welch C., 2001, ApJ 546, 2881998, A&A 331, 361Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ,345, 245Ciardullo, R., Bond, H. E., Sipior, M. S., Fullton, L. K.,Zhang, C.-Y., Schaefer, K. G., 1999, AJ, 118, 488Corradi R. L. M., Guerrero M., Manchado A., & MampasoA., 1997, NewA 2, 461Corradi R. L. M., Sch¨onberner D., Steffen M., & PerinottoM., 2003, MNRAS 340, 417Cuesta L. & Phillips J. P., 2000, ApJ 543, 754Dopita, M. A., 1997, ApJ, 485, L41Exter K. M., Pollacco D. L., & Bell S. A., 2003, MNRAS341, 1349Frank A., & Blackman E. G., 2004, ApJ 614, 737Feibelman, W. A., Hyung, S., Aller, L. H., 1994, ApJ, 426,653Garc´ıa-D´ıaz M. T., L´opez J. A., Garc´ıa-Segura G., RicherM. G., Steffen W., 2008, ApJ 676, 402Garc´ıa-Segura G., L´opez J. A., & Franco, J., 2005, ApJ618, 919Gon¸calves D. R., Corradi R. L. M., & Mampaso A., 2001,ApJ, 547, 302Gon¸calves D. R., Corradi R. L. M., Mampaso A., PerinottoM., 2003, ApJ, 597, 975Gon¸calves D. R., 2004, ASPC, 313, 216Gonc calves D. R., Ercolano B., Carnero A., Mampaso A.,& Corradi R. L. M., 2006, MNRAS 365, 1039Guerrero M. A., Jaxon E. G. & Chu Y.-H., 2004, AJ, 128,1705Guerrero M. A., Miranda L. F., Riera A., Vel´azquez P. F.,& Olgu´ın L., 2008, ApJ 683, 272Hajian A. R., Balick B., Terzian Y., & Perinotto M., 1997,ApJ 487, 313Huggins P. J., Forveille T., Bachiller R., Cox P., AgeorgesN., & Walsh J. R., 2002, ApJ, 573, L55Hyung, S.,& Aller, L. H., 1997, ApJ, 491, 242L´opez, J. A., Vazqu´ez, R., & Rodr´ıguez, L. F., 1995, ApJ,445, L63L´opez, J. A., Meaburn, J., Bryce, M., & Rodr´ıguez, L. F.,1997, ApJ, 475, 705L´opez, J. A., Meaburn, J., Rodr´ıguez, L. F., Vazqu´ez, R.,Steffen. W., & Bryce, M., 2000, ApJ, 538, 233Meaburn J., Walsh J. R., Clegg R. E. S., Walton N. A.,Taylor D., & Berry D. S., 1992, MNRAS 255, 177Meaburn, J., 1997, MNRAS, 292, L11Miranda L. F., Fern´andez M., Alcal´a J. M., Guerrero M.A., Anglada G, et al., 2000, MNRAS 311, 748 O’Dell C. R., & Handron K. D., 1996, AJ, 111, 1630Osterbrock, D. E., & Ferland, G. J., in ”Astrophysics ofGaseous Nebulae and Active Galactic Nuclei” (2nd Edi-tion) / University Science Books, 2006Perinotto, M., 2000, Ap&SS, 274, 205Perinotto, M., Patriarchi, P., Balick, B., Corradi, R. L. M.,2004, A&A, 422, 963Raga A. C., Riera A., Mellema G., Esquivel A., &Vel´azquez P. F., 2008, A&A, 489, 1141Samland, M., Koeppen, J., Acker, A., & Stenholm, B.,1992, A&A, 264, 184Tylenda, R., Acker, A., Stenholm, B., & Koeppen, J., 1992,A&AS, 95, 337V´asquez, R., Kingsburgh R.L., & L´opez, J. A., 1998, MN-RAS, 296, 564V´asquez, R., Miranda, L. F., Torrelles, J. M., Olgu´ın, L.,Ben´ıtez, G., Rodr´ıguez, L. F., L´opez, J. A., 2002, ApJ,576, 860Viironen K., Mampaso A., Corradi R. M. L., Rodr´ıguezM., Greimel R., et al., 2009, A&A accepted,2009arXiv0904.1937VWang, W., Liu, X.-W., Zhang, Y., & Barlow, M. J., 2004,A&A, 427, 873Zhang, C. Y., & Kwok, S., 1998, ApJS, 117, 341Zhang, Y., Liu, X.-W., Wesson, R., Storey, P. J., Liu, Y.,& Danziger, I. J., 2004, MNRAS, 351, 935 c (cid:13)000