The missing mass conundrum of post-common-envelope planetary nebulae
Miguel Santander-García, David Jones, Javier Alcolea, Roger Wesson, Valentín Bujarrabal
WWhy Galaxies Care About AGB Stars. A Continuing Challengethrough Cosmic TimeProceedings IAU Symposium No. xxx, 2018XXX, eds. c (cid:13) The missing mass conundrum ofpost-common-envelopeplanetary nebulae
Miguel Santander-Garc´ıa , David Jones , , Javier Alcolea , RogerWesson , and Valent´ın Bujarrabal Observatorio Astron´omico Nacional, Alfonso XII, 3, 28014, Madrid, Spainemail: [email protected] Instituto de Astrof´ısica de Canarias, E-38205 La Laguna, Tenerife, Spain Departamento de Astrof´ısica, Universidad de La Laguna, E-38206 La Laguna, Tenerife, Spain Department of Physics and Astronomy, University College London, Gower St, London, UK European Southern Observatory, Alonso de C´ordova 3107, Casilla 19001, Santiago, Chile
Abstract.
Most planetary nebulae (PNe) show beautiful, axisymmetric morphologies despitetheir progenitor stars being essentially spherical. Angular momentum provided by a close binarycompanion is widely invoked as the main agent that would help eject an axisymmetric nebula,after a brief phase of engulfment of the secondary within the envelope of the Asymptotic GiantBranch (AGB) star, known as a common envelope (CE). The evolution of the AGB would bethus interrupted abruptly, its (still quite) massive envelope fully ejected to form the PN, whichshould be more massive than a PN coming from the same star were it single. We test thishypothesis by deriving the ionised+molecular masses of a pilot sample of post-CE PNe andcomparing them to a regular PNe sample. We find the mass of post-CE PNe to be actuallylower, on average, than their regular counterparts, raising some doubts on our understanding ofthese intriguing objects.
Keywords. (ISM:) planetary nebulae: general, (stars:) binaries: close, ISM: jets and outflows
1. Introduction
Most planetary nebulae (PNe) show beautiful, aspherical morphologies with high de-grees of symmetry, despite their progenitor stars being essentially spherical. The mecha-nism behind their shaping, however, is still poorly understood (e.g. Balick & Frank 2002).Angular momentum provided by a close binary companion has been widely invoked asthe main shaping agent that would eject an axisymmetric nebula (Jones & Boffin 2017).The mechanism in close binary systems is thought to be as follows: a star undergoingthe Asymptotic Giant Branch (AGB) stage engulfs a companion via Roche-lobe overflowas it expands during the Asymptotic Giant Branch (AGB) phase. The system then un-dergoes a very brief ( ∼ et al. et al. a r X i v : . [ a s t r o - ph . S R ] O c t Miguel Santander-Garc´ıa et al. imply tapping energy from atomic recombination in the envelope (e.g. Ohlmann et al.
Single star vs. CE evolution: the total nebular mass . It can be argued that CE evolu-tion implies significant differences in the mass-loss history of the primary star.Let us consider a single AGB star on its way to produce a PN. Most of its envelope’smass is slowly lost along the AGB evolution, and gets too diluted in the InterstellarMedium (ISM) to be detected. In contrast, the mass lost by the star during the superwindphase (last ∼ ∼ M (cid:12) for a 1.5 M (cid:12) star (seereview by H¨ofner & Olofsson), will form the nebula visible during the PN stage.On the other hand, let us consider the same AGB star, but now as part of a binarysystem close enough to engulf its companion and undergo a CE stage. AGB engulfmentwill thus occur during the last few ( ∼ also be part of the PN as it is suddenly ejected.In other words, despite the large uncertainties in the mass-loss history along the AGB, PNe arising from CE events should, on average, be more massive than their single starcounterparts .This additional mass should be detectable, as it will be close to the central stars duringthe lifetime of the PN, as opposed to the single star case, where it will be long gone,diluted into the ISM. Testing this hypothesis would lead to a better understanding ofthe ejection process. Nevertheless, complete mass determinations of post-CE PNe arevirtually nonexistent. We hereby present the results of a pilot survey of this kind.
2. Sample and Observations
Our pilot sample is composed of 10 post-CE PNe, which amount roughly to 1/6 th of thetotal currently known. It covers a broad range of kinematical ages, central star effectivetemperatures and luminosities, orbital periods and morphologies. These objects are PM 1-23, Abell 41, Hen 2-428, ETHOS 1, NGC 6778, Abell 63, the Necklace, V 458 Vul,Ou 5, and NGC 2346. They lacked any attempt at detecting their molecular content bymeans of radioastronomical observations, except for NGC 6778 (undetected by Huggins& Healy 1989), and NGC 2346, already known to host a massive molecular envelope (e.g.Bachiller et al. CO and CO J =1-0 and J =2-1 emission, using EMIR inthe IRAM 30m radiotelescope. The angular size of the objects of the sample is generallywell suited to the telescope Half Power Beam Width at the observed frequencies.We complemented the mm-range data with archival H α images and optical spectra ofthe sample, from various telescopes and instruments, to derive their ionised masses.
3. Results
Molecular content . No object was detected in CO or CO down to a rms sensitivitylimit in the range 6-25 mK at 230 GHz, except for NGC 6778. This PN shows a simple,broad CO J =1-0 emission profile, as well as double-peaked emission profiles in COand CO J =2-1, whose kinematics correspond to the broken, equatorial ring investigatedby Guerrero & Miranda (2012). The peak intensity relations lead us to conclude that the he missing mass conundrum of post-CE PNe CO J =1-0 is optically thin, and the excitation temperatures relatively low. Furtheranalysis of these profiles and the excitation conditions in this nebula will be presented inSantander-Garc´ıa (in preparation).The CO J =1-0 profile of NGC 6778 allows us to derive a molecular mass of 5 × − M (cid:12) (at 1 kpc) for this PNe by assuming a representative value of the CO abundanceof 3 × − . On the other hand, the sensitivities achieved in the rest of the observationsallow us to derive conservative (3- σ ) upper limits for the molecular masses of the otherobjects in the sample. Ionised content . The ionised mass of NGC 6778, NGC 2346, Abell 41, ETHOS 1, Hen 2-428 and PM 1-23 were derived from their H β fluxes and apparent sizes extracted fromarchival data. Assumptions about the electronic temperatures were made where neces-sary, in order to produce conservative estimates of the ionised masses of these nebulae(i.e. largest T e wherever more than one was available). Ionised masses of the Necklace,Abell 63, Ou 5, and V458 Vul were obtained from Corradi et al. (2011), Corradi et al. (2015), Corradi et al. (2015), and Wesson et al. (2008), respectively. Total mass comparison at 1 kpc . Masses found in this work scale with the distance tothe nebulae squared. Distances to PNe, however, are still poorly known. Hence, in orderto do a proper comparison with PNe not undergoing CE, we must first remove this largedependance by examining the mass every PNe would have at the same distance. Figure 1shows the ionised and molecular masses of our sample of post-CE PNe at 1kpc, togetherwith the ionised and molecular masses of a large sample of 44 PNe selected by Huggins et al. (1996) in an attempt to approach a volume-limited sample, and another sample of27 PNe in the galactic disk, whose ionised/molecular masses were determined by Boffi &Stanghellini (1994) and Huggins & Healy 1989, respectively.Strikingly, except for NGC 2346, the total masses of the post-CE sample seem similar,if not lower, than those of regular PNe. The median mass at 1 kpc of the combinedcomparison samples is 0.021 M (cid:12) , whereas for the post-CE sample it is (cid:54) M (cid:12) .
4. Conclusions
This preliminary work provides a first indication that, contrary to expectations, post-CE PNe seem to be slightly less massive, on average, than their single star counterparts.This discrepancy could however be removed if the molecular gas of these nebulae were toocold (or hot) to be detected, or the ionised gas too hot to emit H α , but these possibilitiesseem rather unlikely. Some of the mass could also be in atomic, neutral form, which hasnot been investigated in this work, and will be part of a future study.On the other hand, should these results be confirmed by further observations andcareful analysis of the possible biases involved, they would present us with the followinginteresting (and so far speculative) implications. The problem of models unable to unbindsuch a large mass would be less severe. A fraction of the mass could fall back forming acircumbinary disk (as in Reichardt et al. et al. References
Bachiller, R., Planesas, P., Mart´ın-Pintado, J., et al.
A&A , 210, 366
Miguel Santander-Garc´ıa et al.
Balick, B. & Frank, A. 2002,
ARA&A , 40, 439Boffi, F. R., & Stanghellini, L. 1994,
A&A , 284, 248Corradi, R. L. M., Sabin, L., Miszalski, B., et al.
MNRAS , 410, 1349Corradi, R. L. M., Garc´ıa-Rojas, J., Jones, D., Rodr´ıguez-Gil, P. 2015,
ApJ , 803, 99Garc´ıa-Segura, G.; Ricker, P. M.; Taam, R. E. 2018,
ApJ , 860, 19Guerrero, M. A., & Miranda, L. F. 2012,
A&A , 539, 47Huarte-Espinosa, M., Frank, A., Balick, B., et al.
MNRAS , 424, 2055Huggins, P. J., & Healy, A. P. 1989,
ApJ , 346, 201Huggins, P. J., Bachiller, R., Cox, P., et al.
A&A , 315, 284Hillwig, T. C., Jones, D., de Marco, O., et al.
A&AR , 26, 1Jones, D., & Boffin, H. M. J. 2017,
Nature Astronomy , 1, 117MacLeod, M., Guillochon, J., Ramirez-Ruiz, E. 2012,
ApJ , 757, 134Ohlmann, S. T.; R¨opke, F. K.; Pakmor, R.; Springel, V.; M¨uller, E. 2016,
MNRAS , 462, L121Reichardt, T. A., De Marco, O., Iaconi, R. 2018, MNRAS, submitted ( arXiv :1809.02297)Soker, N. 2015,
ApJ , 800, 114Shiber, Sagiv; Kashi, Amit; Soker, Noam 2017,
MNRAS , 465, L54Wesson, R., Barlow, M. J., Corradi, R. L. M., et al.
ApJ , 688, L21Wesson, R., Jones, D., Garc´ıa-Rojas, J., et al.
MNRAS , 480, 4589
Figure 1.
Logarithmic ionised mass vs. logarithmic molecular mass at 1 kpc of our post-CEPNe sample (filled circles), PNe from Huggins et al. (1996) (triangles), and a combined sam-ple from Boffi et al. (1994) and Huggins et al.et al.