Physical Structure of the Planetary Nebula NGC 3242 from the Hot Bubble to the Nebular Envelope
Nieves Ruiz, Martin A. Guerrero, You-Hua Chu, Robert A. Gruendl
aa r X i v : . [ a s t r o - ph . H E ] J u l Physical Structure of the Planetary Nebula NGC 3242 from theHot Bubble to the Nebular Envelope Nieves Ruiz , Mart´ın A. Guerrero Instituto de Astrof´ısica de Andaluc´ıa, CSIC, Granada 18008, Spain [email protected], [email protected] and
You-Hua Chu , Robert A. Gruendl Astronomy Department, University of Illinois at Urbana-Champaign, Urbana, IL 61801 [email protected], [email protected]
ABSTRACT
One key feature of the interacting stellar winds model of the formation of planetary nebulae(PNe) is the presence of shock-heated stellar wind confined in the central cavities of PNe. Thisso-called hot bubble should be detectable in X-rays. Here we present
XMM-Newton observationsof NGC 3242, a multiple-shell PN whose shell morphology is consistent with the interacting stellarwinds model. Diffuse X-ray emission is detected within its inner shell with a plasma temperature ∼ × K and an intrinsic X-ray luminosity ∼ × ergs s − at the adopted distance of 0.55kpc. The observed X-ray temperature and luminosity are in agreement with “ad-hoc” predictionsof models including heat conduction. However, the chemical abundances of the X-ray-emittingplasma seem to imply little evaporation of cold material into the hot bubble, whereas the thermalpressure of the hot gas is unlikely to drive the nebular expansion as it is lower than that of theinner shell rim. These inconsistencies are compounded by the apparent large filling factor of thehot gas within the central cavity of NGC 3242. Subject headings: planetary nebulae: individual (NGC 3242)
1. Introduction
Planetary Nebulae (PNe) consist of stellar ma-terial ejected by low- and intermediate-mass stars(0.8–1.0 M ⊙ ≤ M i ≤ ⊙ ). Towards theend of the Asymptotic Giant Branch (AGB), thesestars experience copious mass loss and eject mostof their stellar envelope through a slow, denseAGB wind. The ejected material is subsequentlyionized by the central star and becomes a PN. PNeeventually disperse into the interstellar mediumas they expand, whereas the stellar cores, mainlycomposed of carbon and oxygen, will evolve to- Based on observations obtained with
XMM-Newton ,an ESA science mission with instruments and contributionsdirectly funded by ESA Member States and NASA. ward the white dwarf stage.Near the time when the hot stellar core is ex-posed, the slow AGB wind, with terminal veloc-ities 5–30 km s − (Eder et al. 1988), is super-seded by a fast stellar wind with terminal veloci-ties 1,000–4,000 km s − (Cerruti-Sola & Perinotto1985; Guerrero et al. 2010). This fast stellar windsweeps up the slower AGB wind to form a PN(Kwok 1983). In this interacting stellar winds(ISW) model, the physical structure of a PN wouldbe similar to that of a wind-blown bubble, as mod-eled by Weaver et al. (1977), comprising a centralcavity filled with shocked fast wind (the so-calledhot bubble), a dense shell of swept-up AGB windat 10 K, and an outer envelope of unperturbedexpanding AGB wind. In a simplistic model, the1emperature of the shocked stellar wind inside thehot bubble would be 10 -10 K, but turbulentmixing (e.g., Mellema & Frank 1995) or heat con-duction (Zhekov & Perinotto 1998; Steffen et al.2008) lowers the temperature of the hot gas to10 -10 K and raises its density to produce op-timal conditions for the emission of soft X-rays.Therefore, X-ray observations of shock-heated hotgas in PNe provide us a direct means to exam-ine the interaction of the fast stellar wind withthe AGB wind and to investigate the transfer ofenergy and momentum to the PN envelope.
ROSAT observations showed hints of diffuse X-ray emission in a few PNe (Guerrero et al. 2000).However, it was not until the advent of
Chandra and
XMM-Newton , with their unprecedented reso-lution and sensitivity, that we were finally able tounambiguously detect hot gas in PNe.
Chandra and
XMM-Newton have resolved the diffuse X-rayemission in a handful of PNe (e.g., Kastner et al.2000, 2001; Chu et al. 2001; Guerrero et al. 2002,2005) and revealed unexpected, hard X-ray emis-sion from the central stars of several PNe that mayoriginate from the coronal emission of unseen faintbinary companions or shocks within the fast stellarwinds (Guerrero et al. 2001; Kastner et al. 2003;Montez et al. 2010). Observations of diffuse X-rayemission in PNe demonstrate that hot gas in ellip-tical PNe is confined within the innermost nebularshell and that the high pressure of the hot gas mayindeed drive the nebular expansion as expected inbubble models.The observed values of L X and T X are in agree-ment with predictions of the time-dependent mod-els developed by Steffen et al. (2008) that includeheat-conduction for PNe with central stars of nor-mal, hydrogen-rich surface composition. Thereis, however, an unsolved puzzle: the analyses ofthe chemical composition of the X-ray-emittingplasma in PNe suggest that it is mainly composedof stellar wind, with little contamination of mate-rial from the cold nebular shell (e.g., NGC 2392and NGC 6543, Guerrero et al. 2005; Chu et al.2001). This discrepancy has been further illus-trated by the analysis of the high-spectral reso-lution Chandra
LETG spectrum of BD+30 ◦ ′′ × ′′ inner ellipsoidal shell and ansae sur-rounded by a fainter, 46 ′′ × ′′ moderately el-liptical envelope. These two shells are furtherenclosed by arcs and a giant broken halo re-vealed by deep images (Corradi et al. 2003, 2004).The double-shell morphology of the main neb-ula of NGC 3242 is highly suggestive of interac-tions between the fast stellar wind of its centralstar ( v ∞ =2,400 km s − , Pauldrach et al. 2004)and the previous slow AGB wind. The AGBwind (the nebular envelope) has been swept bythe fast stellar wind to form a thin ionized shellwith a central cavity that can be expected tobe filled with shocked fast wind. This shock-heated gas should emit X-rays, and the diffuse X-ray emission from NGC 3242 is likely detectablebecause of its proximity (distance = 0.55 ± XMM-Newton obser-vations of NGC 3242 that have detected diffuse X-ray emission within its innermost nebular shell.The observations are described in §
2, the resultsare presented in §
3, the physical structure of theoptical nebula is investigated in §
4, and the ef-fects of the shocked stellar wind in the nebula arediscussed in §
2. Observations2.1. XMM-Newton X-Ray Observations
NGC 3242 was observed with the
XMM-Newton
Observatory in Revolution 730 on 2003 December4 using the EPIC-MOS1, EPIC-MOS2, and EPIC-pn CCD cameras (OBSID = 0200240401). Thetwo EPIC-MOS cameras were operated in the Full-Frame Mode for a total exposure time of 19.1 ks,while the EPIC-pn camera was operated in the Ex-tended Full Frame Mode for a total exposure timeof 15.7 ks. The Medium filter was used for allobservations. The
XMM-Newton products were2rocessed using the
XMM-Newton Science Analy-sis Software (SAS version 10.0.0) and the calibra-tion files from the Calibration Access Layer avail-able on 2010 September 16. The event files werescreened to eliminate events due to charged parti-cles or associated with periods of high background.For the EPIC-MOS observations, only events withCCD patterns 0–12 were selected; for the EPIC-pn observation, only events with CCD pattern 0(single pixel events) were selected. Time intervalsof high background, when the count rate in thebackground dominated 10–12 keV energy range is ≥ − for EPIC-MOS and ≥ − for EPIC-pn, were discarded. The resulting netexposure times are 18.7 ks, 18.7 ks, and 13.6 ksfor the EPIC-MOS1, EPIC-MOS2, and EPIC-pnobservations, respectively.The XMM-Newton
EPIC observations detect asource of diffuse X-ray emission at the location ofNGC 3242. An inspection of EPIC-pn and EPIC-MOS images at different energy ranges revealsthat this source is soft, with most emission below1.0 keV and very little emission at higher ener-gies. The EPIC-pn background-subtracted countrate in the 0.38-2.0 keV energy range is 31.3 ± − for a total of 422 ±
22 counts. The EPIC-pn net count rates in the energy ranges 0.38-1.0keV and 1.0-2.0 keV are 30.3 ± − , and1.1 ± − , respectively. The EPIC-MOSbackground-subtracted count rates in the 0.38-2.0keV energy band are 4.4 ± − for MOS1and 4.8 ± − for MOS2, with a total of83 ±
10 cnts for MOS1 and 90 ±
10 cnts for MOS2.
In order to examine the spatial correlation be-tween the diffuse X-ray emission and the opticalnebula, narrow-band WFPC2 images of NGC 3242in the H α , He ii λ ii ] λ iii ] λ HST archive (Proposal ID 7501 and 8773, PI: ArsenHajian, and Proposal ID 6117, PI: Bruce Balick).The images used in this work are listed in Ta-ble 1 with their integration times and filters. In allcases, the innermost shell of the nebula was regis-tered on the WFPC2-PC1 CCD, while a fractionof the outer envelope was missed by the detec-tor. These images were calibrated via the pipelineprocedure and cosmic rays were removed by com-bining different exposures obtained with the same filter using standard IRAF routines. The finalimages have total exposure times of 100 s for H α ,160 s for He ii , 1260 s for [O iii ], and 2840 s for[N ii ]. Medium-dispersion long-slit spectroscopic ob-servations of NGC 3242, obtained using the Gold-Cam CCD Spectrograph (GCCAM) on the 2.1mtelescope at the Kitt Peak National Observatory(KPNO) on 1996 December 7, were kindly pro-vided to us by Dr. K. Kwitter (Henry et al. 2000).The observations, covering the spectral region3600 − ′′ × ′′ slitoriented along the East-West direction and cen-tered at 8 ′′ south of the central star. The spatialscale of the observations is 0 . ′′
78 pix − . The grat-ing 240 was used with the GG-345 blocking filterto cover the blue spectral region 3650 − − , whereas thered spectral region 5650 − − using grating58 with the OG-530 blocking filter. Two expo-sures of 60 s each were acquired in the blue spec-tral range, while exposures of 60 s, 30 s, and twoof 20 s were acquired in the red spectral range.The original spectra were reduced and analyzedusing standard IRAF routines. For the flux cali-bration we used several observations of the spec-trophotometric standard stars BD+28 ◦
3. XMM-Newton Data Analysis3.1. Spatial Properties of the X-ray Emis-sion from NGC 3242
To study the spatial distribution of the X-rayemission from NGC 3242, we have produced EPICimages of NGC 3242 in the energy band 0.25-2.5keV by extracting the individual EPIC-pn, EPIC-MOS1, and EPIC-MOS2 images, mosaicing themtogether, applying the exposure map correction,and smoothing it. The raw image is shown in the IRAF, the Image Reduction and Analysis Facility, is dis-tributed by the National Optical Astronomy Observatory,which is operated by the Association of Universities for Re-search in Astronomy (AURA) under cooperative agreementwith the National Science Foundation.
HST
WFPC2 [O iii ] image of NGC 3242(Fig. 1- right ). The comparison between the X-ray contours and the optical image suggests thatthe X-ray-emitting gas is confined within the in-nermost shell of NGC 3242, as the location ofthe lowest intensity level X-ray contours outsidethe 28 ′′ × ′′ inner shell of NGC 3242 is mostlikely caused by the point-spread function (PSF)of XMM-Newton that is known to have a half en-ergy width (HEW) of 15 . ′′ . ′′ . ′′ XMM-Newton images. For these simu-lated observations we modeled the emission fromNGC 3242 as though it arose from a constant-density ellipsoidal shell of X-ray-emitting gas inte-rior to the innermost nebular shell. We assumed aprolate ellipsoid with semi-major and semi-minoraxes of 11 . ′′ × . ′′ r/r ) of 0.1; ModelB assumed a similar shell with a fractional widthof 0.2; and Model C assumed the X-ray-emittinggas filled the ellipsoid with constant density.To simulate yhe XMM-Newton observations,we generated random simulated X-rays based onthe model distribution of the X-ray emitting gaswith further randomization in the plane of the sky consistent with the
XMM-Newton
PSF . We alsoadded random X-ray events to the simulated im-age to mimic the background emission. The totalnumber of simulated source and noise counts wereset to match those of the actual XMM-Newton ob-servations. Each model was used to make ten sim-ulations to explore the variations caused by thesmall number statistics of these Monte Carlo sim-ulations. We then adaptively smoothed the simu-lated observations with the same parameters usedfor the actual observations. Figure 3 shows a typ-ical realization for each of the three models de-scribed. In the realizations of models A and B (theshell models) a central deficit in X-ray emission isalways apparent. In the realizations of Model Cthe X-ray emission peaks on or near the center;the example shown in Fig. 3 even exhibited an off-set peak matching the actual observation. Thesesimulated observations demonstrate that the dif-fuse X-ray emission from NGC 3242 is more consis-tent with that from a central cavity filled with X-ray-emitting gas than from a thin ellipsoidal shell.They also suggest that the asymmetric distribu-tion of the X-ray emission may be spurius due tothe low count number.
To study the spectral properties of the X-ray emission from NGC 3242, we have extractedits EPIC-pn, EPIC-MOS1, and EPIC-MOS2background-subtracted spectra (Figure 4). Ourdescription of the spectral properties of NGC 3242will focus on the EPIC-pn spectrum as the num-ber of counts in this spectrum is ∼ vii triplet at 0.57 keV.We shall note that the shape of the EPIC-pn spectrum at energies below 0.55 keV is dif-ficult to explain. The unbinned spectrum re-veals noticeable oscillations in the count rates withsmall number of counts in energy bins at ∼ ∼
500 eV, and large number of counts Based on
XMM-Newton observations of the bright, softpoint source Nova LMC1995 (Orio et al. 2003). left ) and smoothed ( center ) XMM-Newton
EPIC images of NGC 3242 in the 0.25-2.5 keVenergy band, and ( right ) HST
WFPC2-PC1 [O iii ] image of NGC 3242. The [O iii ] image is overplottedwith the X-ray contours derived from the smoothed EPIC X-ray image. Contours correspond to 10 σ , 20 σ ,50 σ , 75 σ , 100 σ , and 150 σ above the background level.Fig. 2.— Normalized intensity of the EPIC-MOS X-ray surface brightness profile of NGC 3242 (solid line)and a point-source in the field of view of the instrument (dotted line) along the minor ( left ) and major ( right )axes of the nebula.in energy bins between these two energies andat ∼
400 eV. Since these “spectral features” havewidths smaller than the EPIC-pn spectral reso-lution at this energy range (FWHM ∼
90 eV), itis unlikely that these spikes are associated withreal emission lines. It has been reported thatoptical loading can produce the observed effects,but this mechanism can be ruled out because theMedium Filter used in our observations prevents the optical contamination from point sources asbright as m V =6–9 mag. (XMM-SOC-CAL-TN-0051), while the central star of NGC 3242 has m V =10.3 mag (van Altena et al. 1995). We canconclude that these oscillations are most likelycaused by stochastic effects. To mitigate this issue,the EPIC-pn spectrum has been binned to haveat least 25 counts per channel for further spectralanalysis. Similarly, the EPIC-MOS spectra have5ig. 3.— Realizations of three Monte Carlo simulations of the XMM-Newton images of NGC 3242 for anX-ray-emitting ellipsoidal shell of constant density with a shell thickness 10% (model A) and 20% (modelC), and a filled shell (model C). Contours correspond to 20 σ , 40 σ , 80 σ , and 120 σ above the backgroundlevel. I [ c n t s s − k e V − ] R e s i dua l s Energy [keV] kT [keV]N/N
Sun
Fig. 4.— ( top-panel ) EPIC-pn (black),EPIC-MOS1 (red), and EPIC-MOS2 (green)background-subtracted spectra of the diffuseemission of NGC 3242. The best-fit joint model isshown as a histogram in the corresponding color.The inset shows the nitrogen vs. temperature χ grid plot of the spectral fit where the black, red,and green curves represent the 68%, 90%, and99% confidence levels. ( bottom-panel ) Residualsof the best-fit joint model to the EPIC-pn (black),EPIC-MOS1 (red), and EPIC-MOS2 (green)spectra of NGC 3242 shown in the top panel. been binned to have at least 15 counts per channelfor spectral analyses. For the spectral analysis, we have adoptedthe nebular chemical abundances (He=0.94 He ⊙ ,C=0.78 C ⊙ , N=1.61 N ⊙ , O=0.83 O ⊙ , Ne=0.75 Ne ⊙ ,S=0.20 S ⊙ , and Ar=0.40 Ar ⊙ ), and foregroundhydrogen column density ( N H =5 × cm − ) de-rived by Pottasch & Bernard-Salas (2008). Wehave then modeled the observed EPIC spectrausing an absorbed APEC optically thin plasmaemission model and adopting the absorption cross-sections from Morrison & McCammon (1983).This model provides a reasonable fit to theEPIC spectra of NGC 3242’s diffuse emission witha reduced χ of 1.60 (=38.4/24) for kT =0.190 ± ∼ × K), although the best-fit model islower than the observed spectrum at energies < N H tovary and we find that the best-fit values of kT and N H appear to be correlated (Figure 5) as kT = 0 . − . × N H , where kT is given in keVand N H in units of 10 cm − .Alternatively, we may allow the chemical abun-dances of nitrogen and carbon to vary in thefits, as these two elements have spectral linesthat can contribute in the 0.4-0.5 keV energyband. Variations of the carbon abundance donot produce noticeable changes in the quality of6 . . . N H [10 cm −2 ] k T [ k e V ] + Fig. 5.— Temperature vs. column density χ gridplot of the spectral fit with nebular abundancesand free column density, where the three curvesrepresent the 68%, 90%, and 99% confidence lev-els, and the cross marks the best fit for the valueof 5 × cm − adopted for N H .the spectral fit, and thus the carbon abundancecannot be well constrained. On the other hand,changes to the nitrogen abundances produce a sig-nificant improvement of the spectral fit, and thebest-fit model (Figure 4) has kT =0.202 +0 . − . keV( ∼ × K), N=5.3 +2 . − . N ⊙ , and a reduced χ of 1.07 (=24.56/23). The limits to the nitrogenabundance set by this fit are better illustrated bythe nitrogen versus temperature χ grid plot ofthe spectral fit in Figure 4, which shows that therange of nitrogen abundances exclude the nebularabundance of 1.61 N ⊙ . The best-fit value for thenitrogen abundance implies a N/O ratio of the X-ray-emitting gas to be (N/O) X ∼ ∼ neb ∼ +0 . − . ) × − ergs cm − s − and an intrinsic X-ray luminosity of (7.3 +0 . − . ) × d ergs s − in the0.4-2.0 keV energy band, where d is the distancein kpc. The volume emission measure ( EM = N V , where N e is the electron density and V isthe emitting volume) of this model is ∼ × d cm − . The emitting volume of the inner shell ofNGC 3242 is ∼ × ǫ d cm , where ǫ is the vol-ume filling factor of the X-ray-emitting gas. Thus, the electron density of the X-ray-emitting gas is ∼ ǫ − / d / cm − , and the thermal pressure is1.6 × − ǫ − / d / dyne cm − . At a distance of0.55 kpc, the X-ray luminosity of NGC 3242 is(2.2 +0 . − . ) × ergs s − , the electron density ofthe X-ray-emitting gas is 4 ǫ − / cm − , and itsthermal pressure is 1.2 × − ǫ − / dyne cm − .
4. Physical Structure of the Optical Shell4.1. Bulk Physical Conditions
The electron temperature and density of the op-tical shell of NGC 3242 can be derived using stan-dard techniques (e.g., Guerrero et al. 1996) fromthe long-slit intermediate-dispersion spectroscopypresented in § splot .Table 2 presents line fluxes normalized to an H β flux of 100. The observed fluxes, F , have beendereddened using the IRAF task redcorr to derivethe intrinsic intensity of the line, I : I = F × − c Hβ × f λ (1)where c Hβ is the logarithmic H β extinction con-stant computed by comparing the observed valueof the H α to H β ratio to the expected theo-retical value of 2.87 for case B recombination(Osterbrock & Ferland 2006). The observed fluxesare subsequently corrected using the values of f λ corresponding to the interstellar extinction lawof Savage & Mathis (1979). The H α /H β ratiomeasured in the medium dispersion spectra im-plies an extinction coefficient c Hβ =0.01 for the in-ner shell and 0.05 for the outer shell. We notethat these values of c Hβ are smaller than the val-ues 0.10-0.15 derived by Balick et al. (1993) andPottasch & Bernard-Salas (2008), but consistentwith the values derived by Henry et al. (2000) us-ing this same dataset.The physical conditions for the inner andouter shells of NGC 3242 have been derived us-ing temperature-sensitive line ratios of [O iii ],[N ii ], [O ii ], and [S iii ], and density-sensitiveline ratios of [S ii ] and [Ar iv ]. The ions andlines used, their ratios, and the values of electrontemperature, T e , and density, N e , are listed inTable 3. The results suggest that the inner shell7ensity is ∼ − , in excellent agreementwith the density derived from the [S ii ] lines byPottasch & Bernard-Salas (2008), and the outershell density is ∼
370 cm − . Among the four tem-peratures derived from different lines, we adoptthat from the [O iii ] lines, 11,900 K for the innershell and 10,400 K for the outer shell, as the [O iii ]temperature diagnostic is the least affected by theextinction law and the lines are the brightest.This radial decline in temperature is consistentwith the results presented by Balick et al. (1993).Using these values, we derive thermal pressures of3.6 × − dyne cm − and 5 × − dyne cm − forthe inner and outer shells, respectively. The narrow-band
HST images and ratio mapsof NGC 3242 in different emission lines (Figure 6)display a simple shell morphology consisting of anelliptical inner shell and an attached outer shell.This shell morphology is only moderately compli-cated by ansae and knots best seen in the [N ii ] im-ages (the so-called Fast Low-Ionization EmissionRegions, FLIERS, Balick et al. 1998). A spatio-kinematic study of NGC 3242 by Balick et al.(1987) confirms this simple shell structure: theinner shell is a nearly round bubble expandingat 25–30 km s − , while the outer shell is a co-expanding envelope filled with material. Thus,the H α image of NGC 3242 can be used to ob-tain density and pressure profiles of the nebula forcomparison with those of the X-ray-emitting gasin the central cavity.The density profile of NGC 3242 has been pre-viously determined by Soker et al. (1992) using aground-based H α image. We will use the high-resolution HST
WFPC2 H α image to determinethe electron density, N e , and thermal pressure, P th , of the inner shell and the envelope as a func-tion of nebular radius. First, we examined thisimage to select a direction that would provide uswith a clean surface brightness profile. The cutalong PA=230 ◦ is orthogonal to the inner shellrim and covers the full extent of the outer shell.Thus, we extracted the surface brightness profilesin the H α , He ii λ iii ] λ ii ] λ α ,[O iii ], and [N ii ] lines, but the He ii emission is Fig. 7.— H α , He ii λ iii ] λ ii ] λ ◦ with ∆PA= 10 ◦ . The inset expands this peak to better il-lustrate the spatial differences at this location ofthe emission in the different lines.diminished most likely because the inner shell isoptically thick to He + ionizing photons. As forthe inner shell, all profiles show the bright shellrim, but peak at different radial distances: ∼ . ′′ ii , ∼ . ′′ α , and ∼ . ′′ iii ]. Theemission inside this rim drops steeply, but the fila-ments projected in the central cavity produce sec-ondary peaks in the H α , [O iii ], and [N ii ] lines,but not in the He ii line.The outer shell has been modeled assuming amean electron density ∼
370 cm − and a temper-ature of 10,400 K. Figure 8- left shows the fit tothe H α surface brightness profile of the outer shellof NGC 3242 assuming four different radial depen-dences of the density: r − , r − , r − / , and r − / .The model surface brightnesses for the r − and r − density profiles decrease outwards too rapidlycompared to the observed surface brightness pro-file, and can thus be excluded. The model surfacebrightness for the r − / and r − / density profilesmore closely match the observation.Using the r − / density profiles in the outershell, we proceed to model the H α and He ii sur-face brightness profiles at the rim of the innershell by varying the thickness of the shell and ad-8ig. 6.— HST
WFPC2 narrow-band images and ratio maps of the central region of NGC 3242. The imagesare displayed on a squared-root scale, while the ratio maps are displayed on a linear scale.justing the ionization fraction of He ++ . A shellwith a constant density 2,200 cm − , temperatureof 11,900 K, shell thickness ∼ ii profile. The fit to the rim of the shell in theH α profile is also reasonable, but the model profiledoes not include the contributions from filamentsto the H α profile (Figure 8- right ).The density profile deduced from the modelcomparisons is shown in Figure 9- left . The ra-dial profile of the thermal pressure (Figure 9- right )has been constructed assuming a constant electrontemperature of 11,900 K for the inner shell, and10,400 K for the outer shell, as derived from tem-perature sensitive optical line ratios. This figureshows that the thermal pressure of the inner shellis much higher than that of the outer shell, i.e.,the inner shell is expanding into the envelope.We can also compare the thermal pressureof the nebular shell to that of the hot interior. Assuming that the hot gas is distributed in aconstant-density shell interior to but in contactwith the inner nebular shell, the filling factor canbe expressed as ǫ = 1 − ( r i /r o ) , where r o and r i are the outer and inner radii of the hot gas shell ( r o is also the inner radius of the inner nebular shell).As the hot gas density is ∼ ǫ − / cm − ( d = 0 . r i in Figure 9- left and the thermal pressure in Figure 9- right . Thehot gas shell’s inner radii corresponding to ǫ of0.15, 0.5, and 0.8 are marked. It is apparent thatthe hot gas pressure exceeds that of the nebularshell only for small filling factor values, ǫ < .
5. Discussion
The properties of the X-ray emission fromNGC 3242, based on a preliminary analysis (Ruiz et al.9ig. 8.— ( left ) H α surface brightness profile of NGC 3242 along PA=230 ◦ (thick line) and synthetic H α surface brightness profiles of the outer shell ( r > . ′′
9) for different radial dependences of N e . As for theinner shell, a thickness of 15% in radius has been assumed (see below). ( right ) H α and He ii λ ◦ (thick lines) and synthetic surface brightness profiles of theinner shell for different values of the shell thickness: 15%, 30%, and 50%. As for the outer shell, a decay of N e ∝ r − / has been assumed (see above).Fig. 9.— ( left ) Electron density, N e , and ( right ) thermal pressure, P th , radial profiles of the central cavityand the inner shell and envelope of NGC 3242. The density and thermal pressure of the X-ray-emitting gaswithin the central cavity (radius . ′′ ) are shown as a function of the filling factor, ǫ (red curves) . Thevalues of these physical conditions for filling factors, ǫ , of 0.15, 0.5, and 0.8 are further labeled on the plotat the inner radii of the corresponding hypothetical hot gas shell.2006), were compared to 1D hydrodynamical mod-els of PNe that included the effects of stellar windevolution and heat conduction by Steffen et al.(2008). They remarked that the round shape of the rim of the inner shell of NGC 3242 makes ita suitable PN for the comparison with their 1Dsimulations. The revised values of different X-ray properties of NGC 3242 presented here need10o be discussed in the framework of Steffen et al.(2008)’s model. This revision mostly affects thevalues of the thermal pressure of the hot bubbleand inner shell rim, whereas the X-ray luminosity(scaled to the distance of 0.55 kpc) and plasmatemperature of NGC 3242 are basically the sameas those used by Steffen et al. (2008).We note that the models selected by Steffen et al.(2008) have a thermal pressure of the hot bubblethat exceeds that of the rim, while our revised val-ues indicate the opposite, i.e., the thermal pressureof the rim ( P th =3.6 × − dyne cm − ) is higherthan that of the bubble ( P th =1.6 × − ǫ − / d / dyne cm − ). One possible outcome to this issuewould be a distance for NGC 3242 much fartherthan 0.55 kpc (Terzian 1997; Mellema 2004); ata distance of 3 kpc, the thermal pressures of thehot bubble and rim would be the same, but thiswould make NGC 3242 too large ( r = 0 .
12 pc) andevolved ( τ = 6 ,
000 yrs) and we consider it to beunlikely.The other possible solution is to consider a lowvalue for the filling factor of the X-ray-emittinggas in the bubble; a value ǫ =0.15 would make bothpressures the same. Such a low filling factor is ex-pected to produce a noticeable limb-brighteningmorphology that is not observed (Figures 1 and2). As shown in Fig. 3, hints of a limb-brightenedmorphology are still apparent for a hot gas shellwith a fractional thickness of 0.2 (i.e., ǫ ∼ . XMM-Newton limited spatial resolu-tion ( ∼ ′′ ). It is arguable, however, that ex-tinction has reduced the center-to-limb contrast ofthe X-ray emission, thus shifting the emission in-wards, as this is an effect expected to be noticeable(Steffen et al. 2008) even for the small extinction( N H =5 × cm − ) towards the nebula.Alternatively, we must consider the dynami-cal effects of the photo-ionization of the nebu-lar material, which may have overcome those ofthe currently diminished stellar wind of NGC 3242(Kudritzki et al. 1997; Tinkler & Lamers 2002).In this case, the nebular expansion is driven bythe thermal pressure increase produced in the rimby photo-ionization. This may explain the rela-tively large thickness of NGC 3242 inner shell rim,as compared to that of NGC 6543. We suggestthat the rim thickness can be used as an indica-tor of the relative thermal pressure of hot gas andphoto-ionized nebular material; a sharp rim would imply that the thermal pressure of the hot gas islarger than this of the nebula, and thus that hotgas itself is present (and detectable), while a thickinner shell rim would imply a lower pressure forthe hot gas, if present.The low plasma temperature is puzzling inview of the enhanced N/O ratio of the X-ray-emitting plasma, about three times higher thanthe N/O ratio of the nebular material. Suchdifferences in the chemical abundances seem toimply that the X-ray-emitting plasma mostlyconsists of shocked stellar wind, with little con-tamination of material from the nebula. Thelow plasma temperature, however, needs largeamounts of material from the cold nebular shellto have been incorporated into the hot bubble.This problem, similar to that found in other PNesuch as BD+30 ◦ ◦
6. Summary
We have obtained
XMM-Newton
X-ray obser-vations of NGC 3242, a multiple-shell PN consist-ing of an inner shell with a bright, round rim, andan outer envelope. The observations have detecteddiffuse, soft X-ray emission confined within theinnermost shell of NGC 3242. The relatively low11emperature of the hot gas, T X =2.35 × K, com-pared to the expected adiabatic post-shock tem-perature of a stellar wind with a velocity of 2,400km s − (Pauldrach et al. 2004) suggests that heatconduction has taken place. Indeed, models in-cluding heat conduction provide a reasonable de-scription of the X-ray temperature and luminosityof NGC 3242 (Steffen et al. 2008). However, thechemical abundances of the X-ray-emitting plasmaare closer to the stellar values, suggesting littleevaporation of cold nebular material into the hotbubble and thus contradicting the expectation ofheat conduction.We have compared the physical properties ( N e , T e , P th ) of the gas in the hot bubble with those ofthe gas at the inner shell rim. The inner shell canbe described as a thin shell with a constant densityof 2,200 cm − and a thickness 15% its radius, whilethe envelope is best described by a shell whosedensity declines ∝ r − / to r − / . The gas in thehot bubble has lower thermal pressure than thegas in the shell rim, unless the X-ray-emitting gasis mostly confined within a thin shell, which isnot supported by the X-ray morphology observed.Comparisons with simulations favor a large fillingfactor for the X-ray-emitting gas.We also note the asymmetric distribution of theX-ray emission within the inner shell of NGC 3242and found it unlikely to be a result of nonuniformabsorption of the X-ray emission within the opti-cal shell. X-ray observations at higher spatial res-olution and signal-to-noise ratio (Montez et al., inpreparation) are needed to provide a sharper viewof the distribution of the X-ray-emitting plasmawithin NGC 3242.We thank the anonymous referee for commentsthat helped us to improve this paper. N.R. andM.A.G. are partially funded by grant AYA 2008-01934 of the Spanish MICINN (Ministerio de Cien-cia e Innovaci´on). Y.-H.C. and R.A.G acknowl-edge the support of NASA grant NNG04GE63Gfor this research project. We are also very grate-ful to Dr. Karen Kwitter for providing us with theoptical long-slit spectra of NGC 3242. Some of thedata presented in this paper were obtained fromthe Multimission Archive at the Space TelescopeScience Institute (MAST). STScI is operated bythe Association of Universities for Research in As-tronomy, Inc., under NASA contract NAS5-26555. Facilities: HST(WFPC), XMM, KPNO(2.1mtelescope). REFERENCES
Balick, B., Alexander, J., Hajian, A. R., Terzian,Y., Perinotto, M., & Patriarchi, P. 1998, AJ,116, 360Balick, B., Preston, H. L., & Icke, V. 1987, AJ,94, 1641Balick, B., Rugers, M., Terzian, Y., & Chengalur,J. N. 1993, ApJ, 411, 778Cerruti-Sola, M., & Perinotto, M. 1985, ApJ, 291,237Chu, Y.-H., Guerrero, M. A., Gruendl, R. A.,Williams, R. M., & Kaler, J. B. 2001, ApJ, 553,L69Corradi, R. L. M., Sch¨onberner, D., Steffen, M.,& Perinotto, M. 2003, MNRAS, 340, 417Corradi, R. L. M., S´anchez-Bl´azquez, P., Mellema,G., Giammanco, C., & Schwarz, H. E. 2004,A&A, 417, 637Eder, J., Lewis, B. M., & Terzian, Y. 1988, ApJS,66, 183Guerrero, M. A., Chu, Y.-H., Gruendl, R. A., &Meixner, M. 2005, A&A, 430, L69Guerrero, M. A., Gruendl, R. A., & Chu, Y.-H.2002, A&A, 387, L1Guerrero, M. A., Chu, Y.-H., Gruendl, R. A.,Williams, R. M., & Kaler, J. B. 2001, ApJ, 553,L55Guerrero, M. A., Chu, Y.-H., & Gruendl, R. A.2000, ApJS, 129, 295Guerrero, M. A., Manchado, A., & Serra-Ricart,M. 1996, ApJ, 456, 651Guerrero, M. A., Ramos-Larios, G., & Massa, D.2010, PASA, 27, 210Henry, R. B. C., Kwitter, K. B., & Bates, J. A.2000, ApJ, 531, 928Kastner, J. H., Balick, B., Blackman, E. G.,Frank, A., Soker, N., Vrt´ılek, S. D., & Li, J.2003, ApJ, 591, L3712astner, J. H., Soker, N., Vrt´ılek, S. D., & Dgani,R. 2000, ApJ, 545, L57Kastner, J. H., Vrt´ılek, S. D., & Soker, N. 2001,ApJ, 550, L189Kudritzki, R. P., M´endez, R. H., Puls, J., & Mc-Carthy, J. K. 1997, Planetary Nebulae, 180, 64Kwok, S. 1983, Planetary Nebulae, 103, 293Mellema, G., & Frank, A. 1995, MNRAS, 273, 401Mellema, G. 2004, A&A, 416, 623Montez, R., De Marco, O., Kastner, J. H., & Chu,Y.-H. 2010, ApJ, 721, 1820Morrison, R., & McCammon, D. 1983, ApJ, 270,119Orio, M., Hartmann, W., Still, M., & Greiner, J.2003, ApJ, 594, 435Osterbrock, D. E., & Ferland, G. J. 2006, Astro-physics of gaseous nebulae and active galac-tic nuclei, 2nd. ed. by D.E. Osterbrock andG.J. Ferland. Sausalito, CA: University ScienceBooks, 2006Pauldrach, A. W. A., Hoffmann, T. L., & M´endez,R. H. 2004, A&A, 419, 1111Pottasch, S. R., & Bernard-Salas, J. 2008, A&A,490, 715Ruiz, N., Guerrero, M. A., Chu, Y.-H., Gruendl,R. A., Kwitter, K. B., & Meixner, M. 2006,Planetary Nebulae in our Galaxy and Beyond,234, 497Savage, B. D., & Mathis, J. S. 1979, ARA&A, 17,73Soker, N., & Kastner, J. H. 2003, ApJ, 583, 368Soker, N., Rahin, R., Behar, E., & Kastner, J. H.2010, ApJ, 725, 1910Soker, N., Zucker, D. B., & Balick, B. 1992, AJ,104, 2151Steffen, M., Sch¨onberner, D., & Warmuth, A.2008, A&A, 489, 173Terzian, Y. 1997, Planetary Nebulae, 180, 29 Tinkler, C. M., & Lamers, H. J. G. L. M. 2002,A&A, 384, 987van Altena, W. F., Lee, J. T., & Hoffleit, D. 1995,VizieR Online Data Catalog, 1174, 0Weaver, R., McCray, R., Castor, J., Shapiro, P.,& Moore, R. 1977, ApJ, 218, 377Yu, Y. S., Nordon, R., Kastner, J. H., Houck, J.,Behar, E., & Soker, N. 2009, ApJ, 690, 440Zhekov, S. A., & Perinotto, M. 1998, A&A, 334,239
This 2-column preprint was prepared with the AAS L A TEXmacros v5.2. able 1HST WFPC2 Observations of NGC 3242 Emission Lines Number of images t exp Program ID[s]H α ii ii ] 3 400 61172 300 7501, 87734 260 7501, 8773[O iii ] 1 260 75015 200 7501, 8773 Table 2Line Strengths for NGC 3242
Inner Shell Outer ShellLine ID Wavelength f λ F I F I (˚A)[O ii ] 3726 0.26 10.8 11.7 3.7 4.0[O ii ] 3729 0.26 9.7 10.5 0.3 0.3[S ii ] 4068 0.20 0.5 0.5 . . . . . . [S ii ] 4076 0.20 0.5 0.5 . . . . . . H δ γ iii ] 4363 0.13 10.9 11.3 10.1 10.5[Ar iv ] 4711 0.04 4.8 4.9 4.0 4.0[Ar iv ] 4740 0.04 4.0 4.0 2.9 2.9H β iii ] 4959 –0.02 339.1 336.8 362.2 359.7[O iii ] 5007 –0.03 1018.1 1007.8 1523.3 1507.9[Cl iii ] 5517 –0.15 0.2 0.2 0.3 0.3[N ii ] 5755 –0.21 0.07 0.07 . . . . . . [S iii ] 6312 –0.30 0.7 0.6 0.4 0.4[N ii ] 6548 –0.34 0.7 0.6 . . . . . . H α ii ] 6583 –0.34 2.1 1.9 14.6 13.4[S ii ] 6716 –0.36 0.3 0.3 1.1 1.0[S ii ] 6731 –0.36 0.4 0.4 1.0 0.9[O ii ] 7320 –0.43 0.5 0.4 0.05 0.04[O ii ] 7330 –0.43 0.5 0.4 0.05 0.04[S iii ] 9069 –0.64 3.7 3.1 22.6 19.2[S iii ] 9532 –0.65 18.0 15.2 18.1 15.2F(H β ) (ergs cm − s − ) 3.1 × − × − Table 3Physical Conditions in NGC 3242
Observed Ratios ValuePhysical Parameter Ion Line Ratios Inner Shell Outer Shell Inner Shell Outer Shell T e [N ii ] (6548+6583/5755 41.6 . . . . . .T e [S iii ] (9069+9532)/6312 32.0 106.6 14,040 K 8,070 K T e [O iii ] (4959+5007)/4363 124.9 189.0 11,880 K 10,400 K T e [O ii ] (3726+3729)/(7320+7330) 19.5 43.3 10,100 K 9,860 K N e [Ar iv ] 4711/4740 1.18 1.37 2,250 cm −
400 cm − N e [S ii ] 6716/6731 0.71 1.13 2,200 cm −
340 cm −3