Photochemistry in the inner layers of clumpy circumstellar envelopes: formation of water in C-rich objects and of C-bearing molecules in O-rich objects
aa r X i v : . [ a s t r o - ph . GA ] O c t Draft version June 22, 2018
Preprint typeset using L A TEX style emulateapj v. 08/22/09
PHOTOCHEMISTRY IN THE INNER LAYERS OF CLUMPY CIRCUMSTELLAR ENVELOPES:FORMATION OF WATER IN C-RICH OBJECTS AND OF C-BEARING MOLECULES IN O-RICH OBJECTS
Marcelino Ag´undez , Jos´e Cernicharo and Michel Gu´elin Draft version June 22, 2018
ABSTRACTA mechanism based on the penetration of interstellar ultraviolet photons into the inner layers ofclumpy circumstellar envelopes around AGB stars is proposed to explain the non-equilibrium chem-istry observed in such objects. We show through a simple modelling approach that in circumstellarenvelopes with a certain degree of clumpiness or with moderately low mass loss rates (a few 10 − M ⊙ yr − ) a photochemistry can take place in the warm and dense inner layers inducing importantchanges in the chemical composition. In carbon-rich objects water vapor and ammonia would beformed with abundances of 10 − − − relative to H , while in oxygen-rich envelopes ammonia andcarbon-bearing molecules such as HCN and CS would form with abundances of 10 − − − relativeto H . The proposed mechanism would explain the recent observation of warm water vapor in thecarbon-rich envelope IRC +10216 with the Herschel Space Observatory, and predict that H O shouldbe detectable in other carbon-rich objects.
Subject headings: astrochemistry — molecular processes — stars: AGB and post-AGB — circumstellarmatter INTRODUCTION
Evolved stars in the asymptotic giant branch (AGB)phase undergo important mass loss processes that pro-duce extended circumstellar envelopes (CSEs) composedof dust and molecules. The molecular composition ofCSEs is affected by several processes during the traveltowards the interstellar medium (ISM) but it is originallyestablished in the stellar atmosphere under thermochem-ical equilibrium (TE) conditions and is to a large extentgoverned by the C/O abundance ratio. In C-rich stars(C/O >
1) most of the oxygen is in the form of CO re-sulting in a carbon-based chemistry while in O-rich stars(C/O <
1) CO locks most of the carbon leaving little forother molecules (Tsuji 1973; see also Fig. 1).Astronomical observations have in the main confirmedthis picture, although with a significant number of dis-crepancies. HCN emission is widely observed in O-rich objects (Bujarrabal et al. 1994; Bieging et al. 2000;Decin et al. 2008), in some of them coming from the innerregions of the CSE (Duari & Hatchell 2000). Water va-por in the C-rich envelope IRC +10216 is confined to thewarm near surroundings of the star, according to recentHerschel observations (Decin et al. 2010). Moreover, am-monia is observed in the inner regions of C- and O-richCSEs (Keady & Ridgway 1993; Menten & Alcolea 1995)with abundances much larger than predicted by TE.On the theoretical side, shocks induced by the stel-lar pulsation have been proposed to explain the non-equilibrium chemistry observed in the inner regions ofCSEs (Cherchneff 2006). These models explain the for-mation of HCN and CS in the inner wind of O-richCSEs, although fail to explain the presence of H O in LUTH, Observatoire de Paris-Meudon, 5 Place Jules Janssen,92190 Meudon, France; [email protected] Departamento de Astrof´ısica, Centro de Astrobiolog´ıa, CSIC-INTA, Ctra. de Torrej´on a Ajalvir km 4, Torrej´on de Ardoz, 28850Madrid, Spain Institut de Radioastronomie Millim´etrique, 300 rue de laPiscine, 38406 Saint Martin d’H´eres, France
IRC +10216 and of NH in both C- and O-rich CSEs.In this Letter we investigate an alternative mechanismof non-equilibrium chemistry, based on the penetrationof interstellar ultraviolet (UV) photons into the innerregions of CSEs with a certain degree of clumpiness. MODEL
The model is based on a central AGB star surroundedby a spherical circumstellar envelope, whose physical pa-rameters are given in Table 1, and is adopted to inves-tigate the chemistry in both C- and O-rich CSEs withmass loss rates between 10 − and 10 − M ⊙ yr − .In CSEs with an intense mass loss process strictlyisotropic and homogeneous, the inner regions arewell shielded from interstellar UV light, so that onlythe outer layers are affected by photochemistry (e.g.Charnley et al. 1995; Willacy & Millar 1997). Obser-vations, however, have shown that CSEs have usu-ally clumpy structures both at small and large scales(Chapman et al. 1994; Gu´elin et al. 1997; Weigelt et al.1998; Fong et al. 2003; Le˜ao et al. 2006), which allow fora deeper penetration of interstellar UV photons into theinner regions. To model the effects of clumpiness on thecircumstellar chemistry we adopt a simple approach inwhich the CSE consists of two components: a major onewhose inner regions are well shielded against interstellarUV light, and a minor one (which accounts for a frac-tion f m of the total circumstellar mass) for which theshielding matter located in the radial outward directionis grouped into clumpy structures leaving a fraction f Ω of the solid angle of arrival of interstellar photons free ofmatter.The gas phase chemical composition of these two com-ponents is computed as they expand from the inner-most regions ( r = 2 R ∗ ) until the end of the CSE. Theadopted abundances at the initial radius are given inTable 2. CO is also included with an abundance 30times lower than CO (Milam et al. 2009). The chem-ical network has been used in previous chemical models Ag´undez et al.
TABLE 1Model physical parameters
Star effective temperature ( T ∗ ) 2000 KStellar radius ( R ∗ ) 5 × cmExpansion velocity ( v ) a − for r < R ∗
15 km s − for r ≥ R ∗ Gas kinetic temperature ( T ) b T ∗ ( r / R ∗ ) − . Volume density of gas particles ( n ) c ˙ M /(4 πr h m g i v ) Note . — r is the radius measured from the center of the star. a Velocity field similar to that adopted in previous studies of in-ner CSEs (Keady & Ridgway 1993; Fonfr´ıa et al. 2008). b Values ofthe exponent are typically between − . − c ˙ M is the mass loss rate and h m g i is themean mass of the gas particles (typically ∼ of warm and dense UV illuminated regions (Cernicharo2004; Ag´undez et al. 2008a). Photodissociation and pho-toionization rates are evaluated as a function of the visualextinction A V (Woodall et al. 2007; van Dishoeck et al.2006), adopting the interstellar UV field of Draine (1978).For CO the photodissociation rate is evaluated accord-ing to Mamon et al. (1988), who included the effect ofself-shielding, and for CO through the expression 2 × − exp( − . A V ) s − (Woodall et al. 2007).For the major component, shielded by a smooth enve-lope, A V depends on the column density of H nuclei N H in the radial outward direction as A V = N H (cm − )/1.87 × (Bohlin et al. 1978). The UV field for this com-ponent may be expressed as:4 πJ λ = I λ Ω exp n − [ A λ /A V ]1 . A V o (1)where I λ is the unattenuated interstellar UV field at awavelength λ , Ω the solid angle of arrival of most of theUV flux (depends strongly on the radius and ranges froma small fraction of π sr in the inner regions up to almost4 π sr in the outermost layers), and [ A λ / A V ] the ratio ofthe dust extinction at λ and at visual wavelengths (3.8 for λ = 1250 ˚Aas found for the ISM by Fitzpatrick & Massa1990). For the minor component, shielded by a clumpyenvelope, the UV field may be analogously expressed as:4 πJ λ = I λ Ω h (1 − f Ω ) exp n − [ A λ /A V ]1 . A V o + f Ω i (2)For the minor component we compute an effective visualextinction A ef V by substituting A V by A ef V into Eq. (1)and equalling the right parts of Eqs. (1) and (2).Near-infrared interferometry, able to trace the dustemission at milli-arcsecond scales, has revealed a ex-tremely clumpy structure in objects such as IRC +10216,with five individual clumps in the close surroundings ofthe star, some of them with angle separations of up to20 − ◦ (Weigelt et al. 1998). Millimeter-wave interfer-ometry of molecular lines can, unlike infrared observa-tions of dust, provide information on the projected ve-locity in the line of sight and allow to build three di-mensional maps. Observations of IRC +10216 in molec-ular lines of CN, C H, C H, and HC N (Gu´elin et al.1993; Dinh-V-Trung & Lim 2008) have shown that thesespecies are distributed in a circumstellar shell with a ra-dius of 15-20 ′′ , and with two conical holes in the NNEand SSW directions having an aperture angle of about30 ◦ . This corresponds to a solid angle of π /4 sr, which TABLE 2Initial abundances relative to H in C- and O-rich CSEs Carbon-rich Oxygen-richSpecies Abundance Ref Species Abundance RefHe 0.17 He 0.17CO 8 × − (1) CO 3 × − (1)N × − (2) N × − (2)C H × − (3) H O 3 × − (4)HCN 2 × − (3) CO × − (5)SiO 1.2 × − (3) SiO 1.7 × − (6)SiS 10 − (3) SiS 2.7 × − (7)CS 5 × − (3) SO 10 − (8)SiC × − (3) H S 7 × − (9)HCP 2.5 × − (3) PO 9 × − (10) References . — (1) Teyssier et al. (2006); (2) TE abundance; (3)abundance in IRC +10216, see Ag´undez (2009); (4) Maercker et al.(2008); (5) Tsuji et al. (1997); (6) Sch¨oier et al. (2006); (7)Sch¨oier et al. (2007); (8) Bujarrabal et al. (1994); (9) Ziurys et al.(2007); (10) Tenenbaum et al. (2007). may be a good fraction of Ω in the inner circumstellarregions. Both f m and f Ω are phenomenological parame-ters, difficult to quantify in any CSE. Anyway, adopting f m = 0.1 − −
20 % of the total circumstellarmass) and f Ω = 0.2 − f m = 0.1 and f Ω = 0.25. RESULTS
Fig. 1 shows the TE abundance of various moleculescalculated under conditions typical of the stellar at-mosphere of AGB stars. The calculations have beendone with the code described in Tejero & Cernicharo(1991) and most of the thermochemical data have been
Fig. 1.—
TE abundances as a function of temperature for aC- and O-rich gas (with a C/O abundance ratio of 1.5 and 0.5respectively) with a constant particle density of 10 cm − . hotochemistry in the inner layers of clumpy circumstellar envelopes 3 Fig. 2.—
Calculated abundance of several molecules as a function of radius for carbon-rich CSEs with mass loss rates of 10 − , 10 − , and10 − M ⊙ yr − . Dashed-dotted lines correspond to the abundance of the minor UV exposed component properly corrected by the factor f m , while continuous lines correspond to the abundance weighted-averaged over the minor and major components, which can be expressedas x i ( r ) = (1 - f m ) x Mi ( r ) + f m x mi ( r ), where x Mi ( r ) and x mi ( r ) are the abundances of the species i at a radius r in the major and minorcomponents, respectively. Although in the reality the minor UV exposed and the major UV shielded components would not coexist inspace, this treatment allows to simplify taking the abundance averaged over all radial directions. Note that in those regions where thedashed-dotted and continuous lines of a given species coincide, only the minor UV exposed component contributes to the abundance. taken from Chase (1998). The main purpose of thesecalculations is to show that several molecules which areobserved in the inner regions of CSEs have very low TEabundances (e.g. H O, NH , and SiH in C-rich objects,and HCN, CS, and NH in O-rich objects).Now focusing on the models based on chemical kinetics,Fig. 2 and Fig. 3 show the calculated abundance distri-bution of some molecules in C- and O-rich CSEs withmass loss rates of 10 − , 10 − , and 10 − M ⊙ yr − .In C-rich CSEs with mass loss rates as high as 10 − M ⊙ yr − , typical of an object such as IRC +10216, wa-ter vapor would be effectively formed in the dense andwarm inner regions of the minor UV illuminated compo-nent, for which A ef V is < in excess of 10 − (see Fig. 2). In such re-gions the photodissociation of CO and SiO, the majorreservoirs of oxygen besides CO (hard to photodisso-ciate due to self-shielding effects), liberates atomic oxy-gen which is effectively converted into water through thechemical reactions O + H → OH + H (3)OH + H → H O + H (4)which despite having activation barriers are rapid enoughdue to the high temperatures attained in these innerlayers. The photodestruction of molecules in the minorUV illuminated component occurs fast, but for some ofthem (e.g. H O) the formation rate is high enough toallow them to extend up to relatively large radii, ∼ cm (see Fig.2). This mechanism would explain the ob-servation with Herschel of warm water vapor in the in-ner circumstellar regions of the carbon star IRC +10216(Decin et al. 2010). Other mechanisms proposed suchas sublimation of cometary ices (Melnick et al. 2001),Fischer-Tropsch catalysis on the surface on iron grains(Willacy 2004), or radiative association between O andH (Ag´undez & Cernicharo 2006), place water in cool re- gions located farther than 10 cm.The proposed mechanism would also have some otherinteresting consequences. Hydrides other than H O,such as NH , CH , H S, SiH , and PH , could alsobe effectively formed in the inner CSE by successivehydrogenation reactions of the heavy atom. All themare observed in IRC +10216 (Keady & Ridgway 1993;Cernicharo et al. 2000; Ag´undez et al. 2008b) with abun-dances which, except for CH , are much larger than pre-dicted by TE (see Fig. 1). Ammonia, for example, isobserved in the inner CSE of IRC +10216 with an abun-dance relative to H of 10 − − − (Keady & Ridgway1993; Hasegawa et al. 2006), and yet no efficient for-mation mechanism has been proposed, apart from thesuggestion that it could be formed on grain surfaces(Keady & Ridgway 1993). Our model predicts an effec-tive formation for NH and H S (see Fig. 2), but not forSiH and PH (likely due to the lack of chemical kinet-ics data for the relevant hydrogenation reactions). Othermolecules such as HC N increase their abundance in theinner envelope due to the penetration of interstellar UVphotons (see Fig. 2), something that has been recentlyconfirmed through observations of IRC +10216 at λ = 0.9mm with the IRAM 30-m telescope (Decin et al. 2010;Kahane et al. in preparation).Still focusing on C-rich sources (see Fig. 2), if we movetoward lower mass loss rates then the whole CSE startsto be more transparent to interstellar UV photons. Forexample, for a mass loss rate as low as 10 − M ⊙ yr − most of H O is formed in the major UV shielded com-ponent, which is no longer shielded as it has a visualextinction < − M ⊙ yr − )we should expect a relatively large H O abundance evenif the CSE is not particularly clumpy, prediction thatshould be easily tested with Herschel.In the case of O-rich CSEs the penetration of interstel-lar UV photons into the inner layers has also interestingchemical effects (see Fig. 3). Among them it is worth Ag´undez et al.
Fig. 3.—
Same as in Fig. 2 but for oxygen-rich CSEs. mentioning the formation of NH , CH , HCN, and CSin the inner envelope with abundances relative to H inthe range 10 − − − , i.e. much larger than predicted byTE calculations (see Fig. 1). The formation of C-bearingmolecules in a dense and warm UV illuminated O-rich gashas been discussed by Ag´undez et al. (2008a) in the con-text of the chemistry of protoplanetary disks. HCN, CS,and NH are observed in O-rich CSEs with abundancesof 10 − − − relative to H (Bujarrabal et al. 1994;Menten & Alcolea 1995; Bieging et al. 2000), which aresomewhat higher than predicted by us. Other mecha-nisms based on shocks induced by the stellar pulsation(Duari et al. 1999; Cherchneff 2006) predict fractionalabundances for HCN and CS of 10 − − − , which aresomewhat higher than observed, but a negligible abun-dance for NH . CONCLUSIONS
We have shown through a simple modelling approachthat in CSEs envelopes with a certain degree of clumpi- ness or with moderately low mass loss rates (a few 10 − M ⊙ yr − ) a photochemistry can take place in the warmand dense inner layers inducing important changes inthe chemical composition. This mechanism allows forthe formation of H O and NH in C-rich objects andHCN, CS, and NH in O-rich objects, with abundancesmuch higher than predicted by thermochemical equilib-rium but close to the values typically derived from astro-nomical observations. This mechanism explains the re-cent observation of warm water vapor in the carbon-richenvelope IRC +10216 with the Herschel Space Observa-tory, and predict that H O should be detectable in othercarbon-rich objects.M.A. is supported by a
Marie Curie Intra-EuropeanIndividual Fellowship within the European Community7th Framework Programme under grant agreement n ◦ REFERENCESAg´undez, M. & Cernicharo, J. 2006, ApJ, 650, 374Ag´undez, M., Cernicharo, J., & Goicoechea, J. R. 2008a, A&A,483, 831Ag´undez, M., et al. 2008b, A&A, 485, L33Ag´undez, M. 2009, PhD Thesis, Univ. Aut´onoma de MadridBieging, J. H., Shaked, S., & Gensheimer, P. D. 2000, ApJ, 543,897Bohlin, R. C., Savage, B. D., & Drake, J. F. 1978, ApJ, 224, 132Bujarrabal, V., Fuente, A., & Omont, A. 1994, A&A, 285, 247Cernicharo, J., Gu´elin, M., & Kahane, C. 2000, A&AS, 142, 181Cernicharo, J. 2004, ApJ, 608, L41Chapman, J. M., Sivagnanam, P., Cohen, R. J., & Le Squeren, A.M. 1994, MNRAS, 268, 475Charnley, S. B., Tielens, A. G. G. M., & Kress, M. E. 1995,MNRAS, 274, L53Chase, M., W. 1998, NIST-JANAF thermochemical Tables, 4thed., J. Phys. Chem. Ref. Data, Monograph n 9Cherchneff, I. 2006, A&A, 456, 1001Decin, L., et al. 2008, A&A, 480, 431Decin, L., et al. 2010, Nature, 467, 64Dinh-V-Trung & Lim, J. 2008, ApJ, 678, 303Draine, B. T. 1978, ApJS, 36, 595Duari, D., Cherchneff, I., & Willacy, K. 1999, A&A, 341, L47Duari, D. & Hatchell, J. 2000, A&A, 358, L25Fitzpatrick, E. L. & Massa, D. 1990, ApJS, 72, 163 Fonfr´ıa, J. P., Cernicharo, J., Ritcher, M. J., & Lacy, J. H. 2008,ApJ, 673, 445Fong, D., Meixner, M., & Shah, R. Y. 2003, ApJ, 582, L39Gu´elin, M., Lucas, R., & Cernicharo, J. 1993, A&A, 280, 19Gu´elin, M., Lucas, R., & Neri, R. 1997, in IAU Symp. 170, CO:Twenty-Five Years of Millimeter Wave Spectroscopy, ed. W. B.Latter, et al. (Dordrecht: Kluwer), 359Hasegawa, T. I., et al. 2006, ApJ, 637, 791Justtanont, K., Skinner, C. J., & Tielens, A. G. G. M. 1994, ApJ,435, 852Keady, J. J. & Ridgway, S. T. 1993, ApJ, 406, 199Le˜ao, I. C., et al. 2006, A&A, 455, 187Maercker, M., et al. 2008, A&A, 479, 779Mamon, G. A., Glassgold, A. E., & Huggins, P. J. 1988, ApJ, 328,797Melnick, G. J., et al. 2001, Nature, 412, 160Menten, K. M. & Alcolea, J. 1995, ApJ, 448, 416Milam, S. N., Woolf, N. J., & Ziurys, L. M. 2009, ApJ, 690, 837Sch¨oier, F. L., Olofsson, H., & Lundgren, A. A. 2006, A&A, 454,247Sch¨oier, F. L., Bast, J., Olofsson, H., & Lindqvist, M. 2007, A&A,473, 871Tejero, J., & Cernicharo, J. 1991, IGN, MadridTenenbaum, E. D., Woolf, N. J., & Ziurys, L. M. 2007, ApJ, 666,L29 hotochemistry in the inner layers of clumpy circumstellar envelopes 5hotochemistry in the inner layers of clumpy circumstellar envelopes 5