Close to the Dredge: Precise X-ray C and N Abundances in lambda Andromeda and its Precocious RGB Mixing Problem
J.J. Drake, B. Ball, John J. Eldridge, J.-U. Ness, Richard J. Stancliffe
aa r X i v : . [ a s t r o - ph . S R ] S e p To be submitted to the Astronomical Journal
Close to the Dredge: Precise X-ray C and N Abundances in λ Andromeda and its Precocious RGB Mixing Problem
Jeremy J. Drake , B. Ball , John J. Eldridge , J.-U. Ness , Richard J. Stancliffe Smithsonian Astrophysical Observatory, MS-3, 60 Garden Street, Cambridge, MA 02138,USA. Email: [email protected] of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA,UK. Email: [email protected] Space Agency, XMM-Newton Observatory SOC, SRE-OAX, Apartado 78, 28691Villanueva de la Ca˜nada, Madrid, Spain. Email: [email protected] for Stellar and Planetary Astrophysics, School of Mathematical Sciences, MonashUniversity, Melbourne, VIC 3800, Australia. Email: richard.stancliff[email protected]
ABSTRACT
Chandra
LETG+HRC-S and
XMM-Newton
RGS spectra of H-like C andN lines formed in the corona of the primary star of the RS CVn-type binary λ And, a mildly metal-poor G8 III-IV first ascent giant that completed dredge-up ∼
50 Myr ago, have been used to make a precise measurement of its surfaceC/N ratio. We obtain the formal result [C/N]=0 . ± .
07, which is typical of olddisk giants and in agreement with standard dredge-up theory for stars . M ⊙ .In contrast, these stars as a group, including λ And, have C/ C .
20, whichis much lower than standard model predictions. We show that the abundancesof the old disk giants are consistent with models including thermohaline mixingthat begins at the red giant branch luminosity function “bump”. Instead, λ Andindicates that the C/ C anomaly can be present immediately following dredge-up, contrary to current models of extra mixing on the red giant branch. In thecontext of other recent C and N abundance results for RS CVn-type binaries itseems likely that the anomaly is associated with either strong magnetic activity,fast rotation, or both, rather than close binarity itself.
Subject headings: stars: evolution — stars: abundances — stars: activity —stars: coronae — stars: late-type — X-rays: stars 2 –
1. Introduction
When a low-mass star (0 . − M ⊙ ) evolves off the main sequence and up the red giantbranch (RGB), its outer convective envelope extends inward, probing the CN-processedregion of the hydrogen-burning core and propagating the processed material up to the stellarsurface. Standard stellar evolution models predict this “dredge-up” to result in a decrease ofthe surface C/ C and C/ N ratios on the stellar surface (Iben 1967). While an extensivebody of observational evidence grew demonstrating that such abundance changes do indeedoccur, the changes observed were often more extreme than evolutionary model predictions(see, e.g., reviews by Iben & Renzini 1984; Charbonnel 2005; Chanam´e et al. 2005). Themore salient disagreements were for metal-poor field and globular cluster giants, for which the C/ C ratio approaches the CN-cycle equilibrium value of ∼ C/ C and C/ N ratios decrease with increasingluminosity along the RGB (e.g. Gilroy & Brown 1991; Kraft 1994; Charbonnel et al. 1998;Gratton et al. 2000; Keller et al. 2001), a variety of mechanisms to produce extra mixingbetween core and envelope on the RGB have been proposed to explain the observations(see, e.g., Angelou et al. 2011, for a recent summary). Of these, the thermohaline instabilitypointed out by Charbonnel & Zahn (2007b, see also Eggleton et al. 2006 who first noted themolecular weight inversion that causes it) appears most promising. The instability sets inbeyond the RGB “bump”—the point in the RGB luminosity function where the outwardprogress of the H-burning shell encounters the compositionally uniform layers resulting fromthe deepest extent of convection during the first dredge-up. It results from the mean molecu-lar weight of this region exceeding that of a layer just above the H-burning shell in which themean molecular weight is reduced by the reaction He( He,2p) He (Abraham & Iben 1970;Ulrich 1971).Several recent studies have discussed thermohaline mixing and its apparent success incomparisons of model predictions with observed abundances on the RGB (e.g. Charbonnel & Zahn2007b; Recio-Blanco & de Laverny 2007; Stancliffe et al. 2009; Charbonnel & Lagarde 2010;Cantiello & Langer 2010; Denissenkov 2010; Smiljanic et al. 2010; Tautvaiˇsien˙e et al. 2010b;Angelou et al. 2011). Thermohaline mixing on the RGB changes both the surface C/ Cand C/ N ratios after the end of the first dredge-up, in contrast to classical models thatdo not include extra mixing. In principle, comparison of predicted and observed values ofboth ratios should provide a more stringent test of the theory than either ratio alone. The C/ C ratio can usually be determined from CN molecular features for RGB stars with aprecision of ∼
20% or so—useful for comparison with and constraining models. Unfortu- 3 –nately, assessment of photospheric N abundances is not so precise and usually has to relyon similar CN features. The N abundance then also depends on the derived C abundancethrough molecular equilibrium, and the resulting uncertainties in the C/N ratio can be quitelarge and rarely below 0.2 dex. The use of field RGB stars for confronting model predictionsis also hampered by their uncertain masses and evolutionary phases, especially near spectraltype K0 where the evolutionary tracks of clump stars of different mass overlap with those offirst ascent stars.Here, we draw attention to nearby evolved active binary stars, whose masses and evo-lutionary phases can generally be much better constrained than for field stars, as offeringpotentially valuable laboratories for further study of dredge-up and post dredge-up mixing.Denissenkov et al. (2006) have also piqued interest in these stars through theoretical mod-elling that indicates extra mixing on the RGB might be induced by tidal spin-up. In thiscontext, the nearby (26 pc) old disk giant λ And presents an interesting case: it is a mildlymetal-poor first-ascent G8 III-IV star at an evolutionary phase in which CN-cycle productsshould have just appeared at its surface due to first dredge-up (Savanov & Berdyugina 1994;Donati et al. 1995; Ottmann et al. 1998; Tautvaiˇsien˙e et al. 2010a; further characteristicsare listed in Table 1). Other than its close binarity, it is similar to members of the sampleof mildly metal-poor ([Fe/H] ∼ − .
5) disk giants that Cottrell & Sneden (1986) found tohave unevolved (approximately solar) C/N ratios, but lower C/ C ratios of ∼ . M ⊙ thought typical of the Cottrell & Sneden sample indeed indicatelittle change in the surface C/N ratio, but also only mild reduction in the C/ C ratio to val-ues &
30. The observed low C/ C ratios suggest that additional mixing of CN-processedmaterial has occurred further to that of classical model predictions, but of a nature as notto change the surface C/N ratio. Mixing of such a characteristic was initially proposed by,e.g., Dearborn & Eggleton (1977); Sweigart & Mengel (1979); Hubbard & Dearborn (1980);Lambert & Ries (1981) in order to explain C rich stars such as Arcturus, which also has[C/N] ∼ < M ⊙ . In § X-ray measurement of the surface C/N abundance of λ And to showthat it is typical of old disk giants from the Cottrell & Sneden (1986) sample. In §
3, we 4 –combine this measurement with the C/ C assessments of Tautvaiˇsien˙e et al. (2010a) andSavanov & Berdyugina (1994) and compare the results with predictions from a state-of-the-art stellar evolutionary model including thermohaline mixing.
2. Observations and Analysis
We estimate the surface C/N abundances from emission lines of H-like ions of C and Nseen in
Chandra
Low Energy Transmission Grating (LETG) X-ray spectra of λ And. Ouranalysis follows those presented previously by Drake (2003b) and Drake & Sarna (2003), towhich we refer the reader for details; see also Schmitt & Ness (2002). Our method exploitsthe insensitivity of the relative intensities of the C VI λ . λ . Chandra
X-ray spectra of λ And were obtained on 2002 July 22 and 23 (ObsID 2558and 3722, respectively) in two separate observations of 50 ks each, using the LETG andHigh Resolution Camera spectroscopic (HRC-S) detector in its standard instrument con-figuration. Data were obtained from the Chandra Data Archive , and were reduced usingthe CIAO software package version 3.2. This latter processing included filtering of eventsbased on observed signal pulse-heights to reduce background. Spectra were analysed usingthe PINTofALE IDL software suite (Kashyap & Drake 2000). XMM-Newton observed λ And on 2001 January 26 for 32 ks. Data were processed andreduced using standard
XMM-Newton
Science Analysis System software, and RGS spectrawere extracted with the rgsproc task. Spectral line fluxes were measured by fitting lineprofiles using the cora program (Ness & Wichmann 2002). Measured lines intensities forboth
Chandra and
XMM-Newton spectra are listed in Table 2.The LETG spectrum of λ And for the wavelength range that includes both the N VIIand C VI lines is illustrated in Fig. 1, alongside spectra of the single giant β Ceti (K0 III)and the unevolved dwarf ǫ Eri (K2 V). The latter two stars were also used by Drake & Sarna(2003) as examples of spectra of evolved and unevolved comparison stars. Simple visualinspection reveals that the λ And subgiant C/N line strength ratio is similar to that of ǫ Eri,and is very unlike that of the evolved β Ceti. More formally, Drake (2003b) showed that the http://asc.harvard.edu/cda Interactive Data Language, Research Systems Inc. n ( C ) /n ( N ), is given by n ( C ) n ( N ) = 1 . × ( I C /A Ceff )( I N /A Neff ) , (1)where I C and I N are the number of counts in the C VI (24.7 ˚A) and N VII (33.7 ˚A) lines,and A Ceff and A Neff are the effective area normalising factors (in cm ) at the appropriatewavelengths of the C and N transitions, respectively. Applying this formula to the λ Andobservations gives n ( C ) /n ( N ) = 3 . ± .
75 and 4 . ± . n ( C ) /n ( N ) = 3 . ± .
6, or [C/N]= − . ± . .
10% (Drake 2003b).In light of the now well-documented chemical fractionation related to element first ion-ization potential (FIP) that occurs between the coronae and photospheres of stars (e.g. Drake2003a), it might be questioned whether the coronal C/N ratio derived here is representativeof that of the underlying photosphere. Drake (2003b) studied the coronal C/N ratio in theactive binary V711 Tau and concluded that C and N are not fractionated to any significantextent with respect to the photosphere. This is expected based on the similar high FIPsof these elements (11.3 and 14.5 eV for C and N, respectively). While we cannot rule outrelative fraction among C and N with absolute certainty, we therefore consider our derivedC/N abundance ratio to be directly applicable to the photosphere of λ And.Our C/N ratio is in good agreement with the earlier estimates of [C/N]= − .
24 and − .
25 (adjusted to the Asplund et al. (2009) scale) by Savanov & Berdyugina (1994) andTautvaiˇsien˙e et al. (2010a), bearing in mind the “0.2-0.3 dex” uncertainty they assess fortheir C/N abundance ratio.
3. Discussion
The currently favoured solar (unevolved) n ( C ) /n ( N ) abundance ratio is 3.98 (Asplundet al. 2009), with stated uncertainties of 0.05 dex for both C and N. C/N ratios for galacticfield stars are known with somewhat less precision. Galactic disk dwarfs show [C/N]=0 downto metallicities of [Fe/H]= − .
4, though with some scatter attributable to uncertainties inN abundances determined from weak N I lines (Reddy et al. 2003). Mishenina et al. (2006)find no significant trend of C/N with [Fe/H] over a similar metallicity range for a sampleof 177 disk giants, indicating that the post-dredge-up ratio also evolves similarly for solarmetallicity and mildly metal-poor stars. 6 –Our observed C/N ratio for λ And is perfectly consistent with an unevolved composi-tion, showing no signs of post dredge-up change. In contrast, Savanov & Berdyugina (1994)and Tautvaiˇsien˙e et al. (2010a) found an unambiguous signature of dredge-up in the carbonisotope ratio for which they derived C/ C= 20 ± C/ C= 17 ±
5, though our conclusions would be un-changed regardless of whether we adopted either result. This CN signature of λ And is typicalof the old disk giants analysed by Cottrell & Sneden (1986), in keeping with its mild metaldeficiency credential ([Fe/H]= − . λ And. We used a version of the STARS code (Eggleton 1971; Pols et al. 1995; Stancliffe & Eldridge2009, and references therein) with updated opacities (Eldridge & Tout 2004; Stancliffe & Glebbeek2008) and nucleosynthesis routines (Stancliffe et al. 2005; Stancliffe 2005) to generate evo-lutionary tracks from the main sequence for stars with metallicity z = 0 .
008 and massesin the range 0.7–1.7 M ⊙ . Thermohaline mixing was included in these models followingStancliffe et al. (2009), using the prescription of Kippenhahn et al. (1980), with the diffu-sion coefficient multiplied by a factor of 1000, following Charbonnel & Zahn (2007b). Theresulting evolutionary tracks are illustrated in Figure 2, together with the data point for λ And (see Table 1). Also shown is the locus describing the evolutionary stage at whichstars of different mass essentially reach the end of the first dredge-up. While dredge-up endis formally defined as the point at which the convection zone reaches its deepest penetrationin mass, we defined it here as the point at which the surface C/ C ratio reaches within5% of what would be the final RGB values in the absence of additional mixing processes;after this, canonical models predict their surface abundances remain constant until laterdredge-up phases. We use this looser definition of the end of dredge-up to illustrate that the C/ C ratio is predicted to reach close to its final value significantly below the RGB bump.Figure 2 shows that λ And lies a factor of 2-3 in luminosity below horizontal branchtracks characterising the clump, and so must be a first ascent giant (see also the extensivediscussion of Donati et al. 1995). It also lies just beyond the “end” of first dredge-up; usingthe STARS tracks, we estimate that this phase was completed ∼
50 Myr ago, in qualitativeagreement with the earlier assessment of Donati et al. (1995). We also estimate from theseevolutionary tracks a mass for λ And of 1 . ± . M ⊙ , which compares favourably withthe spectroscopic estimate based on log g and the radius (Table 1) of 1 . +1 . − . M ⊙ . This isconsistent, within experimental error, with the estimate of Donati et al. (1995); our slightlylarger value is mostly due to our adopted surface gravity being higher (see analyses bySavanov & Berdyugina 1994; Ottmann et al. 1998; Tautvaiˇsien˙e et al. 2010a). 7 –The value of the precise X-ray measurement of C/N is apparent in Figure 3, whichcompares the observed C isotope and C/N ratios for λ And with predictions from canon-ical evolution models past the first dredge-up but with no additional post-dredge-up mix-ing (Schaller et al. 1992; Vandenberg 1992; Girardi et al. 2000). In accordance with well-established dredge-up results, models predict a reduction in C/ C with increasing mass,with only very mild reductions for the lowest mass stars. Also shown are the observed ratiosfor the old disk giants from Cottrell & Sneden (1986). This kinematically-selected samplecomprises low mass stars with M . M ⊙ , very much like λ And. λ And and the old diskgiants have C/N ratios more or less in agreement with the models, but have much lower C/ C ratios. There is some selection effect in the latter group because stars with higher C/ C ratios tend to be represented only by lower limits, though this applies only to 4out of the 34 stars in the Cottrell & Sneden (1986) sample; no C/ C data are availablefor a further 8 stars. Statistically, the majority of the Cottrell & Sneden (1986) sample areexpected to be core He-burning stars, since this evolutionary phase lasts several times longerthan the first RGB ascent. The problem with the abundance ratios for these old disk gi-ants is that they exhibit much more extreme processing of C to C than the models, butessentially no processing of C to N.The STARS models with thermohaline mixing predict changes in the surface C/ Cand C/N ratios above the RGB ‘bump’. The C/ C vs. C/N loci from the main-sequenceto the RGB tip for four models corresponding to different stellar masses are illustrated inFigure 4. The most relevant of these for the sample under consideration are the 0.7 and1 . M ⊙ ones. While the Cottrell & Sneden sample has systematically higher C/N ratios onaverage than the models by about 50%, a systematic error of this magnitude in the observedvalues probably cannot be ruled out given the 0.2 dex uncertainty on each. The range ofobserved C/ C ratios is instead in very good agreement with the thermohaline mixingpredictions.At face value, the generally reasonable agreement between observed and model abun-dances for the metal-poor old disk star sample is encouraging. However, λ And does presenta problem. Its C/N ratio is significantly higher than model predictions, though for a singleobject such a discrepancy might be attributed to cosmic variance. More important is theevolutionary phase at which the thermohaline mixing begins to affect the surface C/ Cratio. This is above the RGB bump. Tautvaiˇsien˙e et al. (2010a) pointed out that λ Andinstead lies well below the bump, as demonstrated in Figure 2, and indicates that C/ Cis anomalous earlier in its evolution, and perhaps immediately following dredge-up.One possible source of the anomaly is contamination from its companion. There is nodirect observational information to assess with certainty the nature of the companion, though 8 –Donati et al. (1995) presented cogent arguments based on the asynchronicity of the rotationand orbital periods of λ And to rule out a white dwarf nature for the secondary. Sucha star might otherwise have contaminated the current primary with carbon-rich materialduring its asymptotic giant branch (AGB) phase. However, finding a companion that isable to provide material of a suitable composition is difficult. Low-mass AGB stars thatundergo the third dredge-up become rich in carbon-12, yet do not produce much in the wayof carbon-13 and nitrogen-14. For example, Karakas & Lattanzio (2007) find that a 1 . M ⊙ star of Z = 0 .
004 produces ejecta with a C/ C ratio of over 100 and a C/N ratio of about10. Stancliffe & Jeffery (2007) find similar values for the ejecta of a 1 . M ⊙ star of Z = 0 . λ And, even if one allows for the dilutionof any accreted material via mixing with the pristine material of the receiving star. Onemay wish to invoke some sort of extra mixing processes at work during the AGB in orderto lower these values. The physical nature of this extra mixing process is unknown. It isunlikely to be thermohaline mixing, as this has been shown unable to produce C/ C ratiosas low as 10 at lower metallicity than that of λ And even though it is more effective at lowmetallicity (Stancliffe 2010). There is also debate about whether this is actually required inlow-mass AGB stars (see Busso et al. 2010 and Karakas et al. 2010 for opposite sides of thisdebate). Appealing to a higher mass companion, one that underwent hot bottom burningon the AGB, is also unlikely to help: the models of Karakas & Lattanzio (2007) suggest thatthese would produce C/ N less than 1—too low to match the observations.We also note here a fundamental difference between λ And and the subgiants in the glob-ular clusters NGC 6752 and 47 Tuc that were found to have C/ C .
10 by Carretta et al.(2005): unlike λ And these stars also have depleted C and enhanced N abundances. Carretta et al.(2005) interpreted these abundances as the signature of contamination by mass loss fromintermediate-mass AGB stars.Recently, the C isotope and C/N ratios have been estimated for two additional RS CVn-type binaries, 29 Dra and 33 Psc. Bariseviˇcius et al. (2010, 2011) find similar C/N ratios([C/N] ∼ − .
25) to the Tautvaiˇsien˙e et al. (2010a) value for λ And, but find C/ C= 16for 29 Dra and 30 for 33 Psc. The latter is “normal” for a post-dredge-up star, and indeed33 Psc, like λ And, lies below the bump luminosity (Bariseviˇcius et al. 2011) and is notexpected to have experienced additional post-dredge-up mixing episodes. The isotope ratiofor 29 Dra is instead similar to that of λ And. Based on the atmospheric parameters ofBariseviˇcius et al. (2010), 29 Dra is probably cooler by ∼
100 K and more luminous by0.05 dex than λ And, and therefore lies slightly above and to the right of λ And in Figure 2.Nevertheless, it is likely still located just below the luminosity bump and would thereforeappear anomalous in terms of the C/ C ratio. 9 – λ And, and likely 29 Dra, then seem to present a challenge to the current description ofthermohaline mixing as the controlling factor in post-dredge-up surface abundance changes.While this mechanism appears generally successful in explaining post-dredge-up CNO andlight element surface abundances ( § He burning is unlikely to be thesole mechanism of extra mixing on the RGB and must be enhanced or augmented by otherprocesses. Interestingly for the magnetically-active λ And, Denissenkov & Merryfield (2011)suggested strong toroidal magnetic fields arising from differential rotation in the radiativezone might enable more growth of thicker fingers. Alternatively, Charbonnel & Zahn (2007a)note that strong magnetic fields can act to damp the thermohaline instability and suggestthis as a means of suppressing extra mixing in the RGB descendents of Ap stars.Even if the magnetic activity of λ And might provide some assistance to the process, itdoes not solve the problem that its surface abundances are changed before the predictedon-set of the thermohaline instability. Rotation-driven mixing, once thought the mainabundance-modifying mechanism on the RGB, also would not help for similar reasons, al-though Palacios et al. (2006) have essentially ruled the process out for producing significantsurface abundance modifications. One possible alternative is the magnetic mixing processproposed by Busso et al. (2007), in which buoyant flux tubes generated in a stellar dynamooperating near the hydrogen-burning shell transport processed matter upward into the con-vective envelope. Again, the challenge for such a mechanism to work for λ And would be tobring sufficient C-rich material to the surface so soon after dredge-up.We are otherwise drawn back to the discussion of the C-rich giants by Lambert & Ries(1981), who reasoned that some C must be removed from the CN-processing zone prior toconversion through proton capture to N on the main sequence (see also, e.g. Dearborn & Eggleton1977, Sweigart & Mengel 1979, and the “magnetic mixing” of Hubbard & Dearborn 1980).Thermohaline mixing seemed to solve much that was raised in that discussion, though inthe light of the conclusions of Denissenkov & Merryfield (2011), the odd exception such as λ And and 29 Dra also seems to point to other effects. Existing studies of rotationally-drivenmixing and other mechanisms do not yet generally include aspects such as close binarity andstrong magnetic fields that distinguish λ And from other stars in RGB abundance surveys.Theoretical models including tidal interaction and spin-up of giants in RS CVn-type binaries 10 –developed by Denissenkov et al. (2006) present an interesting exception.Denissenkov et al. (2006) found evolutionary models with tidal spin-up caused by aclose companion to exhibit large excursions of M V ∼ . λ And is notoriously unsynchronized,but it lies only just below the bump luminosity. Were such large bump excursions to occur,they would allow λ And and 29 Dra to be“post-bump” stars, lying on the downward lowerluminosity excursion. The extended stay at this point in its evolution would also providegreater opportunity for mixing of C-rich material into its convection zone by rotation,thermohaline or other mixing process.The question then arises why λ And and 29 Dra exhibit C isotope ratio anomalies but33 Psc does not? Bariseviˇcius et al. (2011) pointed to the low activity of 33 Psc comparedto that of λ And and 29 Dra. The orbital period of 33 Psc is 79.4 days (Harper 1926) thoughits rotation velocity is not constrained. The orbital period of 29 Dra is much longer, 904days, but it has a rotation period of about 30 days (Hall et al. 1982; Zboril & Messina 2009),which is shorter than that of λ And (54 days), and based on activity indices is likely a fasterrotator than 33 Psc. Since 29 Dra is a much wider binary than either λ And or 33 Psc, tidalinteractions would appear not to be the primary factor governing the isotope anomaly, andinstead rotation and activity appear to be the culprits.
4. Conclusions
A precise X-ray measurement of H-like C and N formed in the corona of the G8 III-IV primary of the RS CVn-type binary λ And has revealed a solar C/N abundance ratiothat is essentially unchanged since the star was formed. This is in qualitative agreementwith evolutionary calculations for such low-mass stars ( . . M ⊙ ), but is quite inconsistentwith its photospheric ratio C/ C= 14 (Savanov & Berdyugina 1994; Tautvaiˇsien˙e et al.2010a) that is much lower than predicted by theory. This abundance pattern is typical ofthe old disk giants first brought to prominence by Cottrell & Sneden (1986). λ And is a firstascent red giant that underwent dredge-up only ∼
50 Myr ago; its anomalously low C/ Chas appeared immediately after dredge-up. We echo the conclusions of Tautvaiˇsien˙e et al.(2010a) that extra mixing on the RGB in magnetically-active low-mass stars like λ And and29 Dra appears to act below the luminosity function bump, in contradiction with currentthermohaline mixing models. Magnetic flux tube buoyant mixing (Busso et al. 2007) would 11 –appear to warrant more detailed investigation. Evolutionary models of tidally-interactingbinaries by Denissenkov et al. (2006) that predict large luminosity excursions at the bumpmight also allow λ And to be a “post-bump” giant and alleviate its otherwise precociouslydiminished C isotope ratio. Such a mechanism would appear to be less promising for widerbinaries such as 29 Dra.We thank the NASA AISRP for providing financial assistance for the development of thePINTofALE package. JJD was supported by NASA contract NAS8-39073 to the
ChandraX-ray Center during the course of this research. WB was supported by
Chandra awardnumber AR4-5002X issued by the
Chandra
X-ray Center. JJD thanks H. Tananbaum andthe CXC science team for advice and support.
REFERENCES
Abraham, Z., & Iben, Jr., I. 1970, ApJ, 162, L125+Alonso, A., Arribas, S., & Mart´ınez-Roger, C. 1999, A&AS, 139, 335Angelou, G. C., Church, R. P., Stancliffe, R. J., Lattanzio, J. C., & Smith, G. H. 2011, ApJ,728, 79Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481Bariseviˇcius, G., Tautvaiˇsien˙e, G., Berdyugina, S., Chorniy, Y., & Ilyin, I. 2010, BalticAstronomy, 19, 157—. 2011, Baltic Astronomy, 20, 53Brown, J. A. 1987, ApJ, 317, 701Busso, M., Palmerini, S., Maiorca, E., Cristallo, S., Straniero, O., Abia, C., Gallino, R., &La Cognata, M. 2010, ApJ, 717, L47Busso, M., Wasserburg, G. J., Nollett, K. M., & Calandra, A. 2007, ApJ, 671, 802Cantiello, M., & Langer, N. 2010, A&A, 521, A9+Carretta, E., Gratton, R. G., Lucatello, S., Bragaglia, A., & Bonifacio, P. 2005, A&A, 433,597Carretta, E., Gratton, R. G., & Sneden, C. 2000, A&A, 356, 238 12 –Chanam´e, J., Pinsonneault, M., & Terndrup, D. M. 2005, ApJ, 631, 540Charbonnel, C. 2005, in ASP Conf. Ser. 336: Cosmic Abundances as Records of StellarEvolution and Nucleosynthesis, 119–+Charbonnel, C., Brown, J. A., & Wallerstein, G. 1998, A&A, 332, 204Charbonnel, C., & Lagarde, N. 2010, A&A, 522, A10+Charbonnel, C., & Zahn, J. 2007a, A&A, 476, L29—. 2007b, A&A, 467, L15Cottrell, P. L., & Sneden, C. 1986, A&A, 161, 314Dearborn, D. S. P., & Eggleton, P. P. 1977, ApJ, 213, 177Dempsey, R. C., Linsky, J. L., Schmitt, J. H. M. M., & Fleming, T. A. 1993, ApJ, 413, 333Denissenkov, P. A. 2010, ApJ, 723, 563Denissenkov, P. A., Chaboyer, B., & Li, K. 2006, ApJ, 641, 1087Denissenkov, P. A., & Merryfield, W. J. 2011, ApJ, 727, L8+Donati, J.-F., Henry, G. W., & Hall, D. S. 1995, A&A, 293, 107Drake, J. J. 2003a, Advances in Space Research, 32, 945—. 2003b, ApJ, 594, 496Drake, J. J., & Sarna, M. J. 2003, ApJ, 594, L55Eggleton, P. P. 1971, MNRAS, 151, 351Eggleton, P. P., Dearborn, D. S. P., & Lattanzio, J. C. 2006, Science, 314, 1580Eldridge, J. J., & Tout, C. A. 2004, MNRAS, 348, 201Gilroy, K. K., & Brown, J. A. 1991, ApJ, 371, 578Girardi, L., Bressan, A., Bertelli, G., & Chiosi, C. 2000, A&AS, 141, 371Gratton, R. G., Sneden, C., Carretta, E., & Bragaglia, A. 2000, A&A, 354, 169Grevesse, N., & Sauval, A. J. 1998, Space Science Reviews, 85, 161 13 –Hall, D. S., Henry, G. W., Louth, H., Renner, T. R., & Shore, S. N. 1982, InformationBulletin on Variable Stars, 2109, 1Harper, W. E. 1926, Publications of the Dominion Astrophysical Observatory Victoria, 3,341Hubbard, E. N., & Dearborn, D. S. P. 1980, ApJ, 239, 248Iben, I., & Renzini, A. 1984, Phys. Rep., 105, 329Iben, Jr., I. 1967, ApJ, 147, 624Karakas, A., & Lattanzio, J. C. 2007, PASA, 24, 103Karakas, A. I., Campbell, S. W., & Stancliffe, R. J. 2010, ApJ, 713, 374Kashyap, V., & Drake, J. J. 2000, Bulletin of the Astronomical Society of India, 28, 475Keller, L. D., Pilachowski, C. A., & Sneden, C. 2001, AJ, 122, 2554Kippenhahn, R., Ruschenplatt, G., & Thomas, H. 1980, A&A, 91, 175Kjaergaard, P., Gustafsson, B., Walker, G. A. H., & Hultqvist, L. 1982, A&A, 115, 145Kraft, R. P. 1994, PASP, 106, 553Lambert, D. L., & Ries, L. M. 1981, ApJ, 248, 228Landis, H. J., Lovell, L. P., Hall, D. S., Henry, G. W., & Renner, T. R. 1978, AJ, 83, 176Mishenina, T. V., Bienaym´e, O., Gorbaneva, T. I., Charbonnel, C., Soubiran, C., Korotin,S. A., & Kovtyukh, V. V. 2006, A&A, 456, 1109Ness, J.-U., & Wichmann, R. 2002, Astronomische Nachrichten, 323, 129Ottmann, R., Pfeiffer, M. J., & Gehren, T. 1998, A&A, 338, 661Palacios, A., Charbonnel, C., Talon, S., & Siess, L. 2006, A&A, 453, 261Pols, O. R., Tout, C. A., Eggleton, P. P., & Han, Z. 1995, MNRAS, 274, 964Recio-Blanco, A., & de Laverny, P. 2007, A&A, 461, L13Reddy, B. E., Tomkin, J., Lambert, D. L., & Allende Prieto, C. 2003, MNRAS, 340, 304Savanov, I. S., & Berdyugina, S. V. 1994, Pisma Astronomicheskii Zhurnal, 20, 279 14 –Schaller, G., Schaerer, D., Meynet, G., & Maeder, A. 1992, A&AS, 96, 269Schmitt, J. H. M. M., & Ness, J.-U. 2002, A&A, 388, L13Shetrone, M. D. 1996, AJ, 112, 2639Shetrone, M. D., Sneden, C., & Pilachowski, C. A. 1993, PASP, 105, 337Smiljanic, R., Pasquini, L., Charbonnel, C., & Lagarde, N. 2010, A&A, 510, A50+Soubiran, C., Bienaym´e, O., Mishenina, T. V., & Kovtyukh, V. V. 2008, A&A, 480, 91Stancliffe, R. J. 2005, PhD Thesis (Cambridge University, UK)—. 2010, MNRAS, 403, 505Stancliffe, R. J., Church, R. P., Angelou, G. C., & Lattanzio, J. C. 2009, MNRAS, 396, 2313Stancliffe, R. J., & Eldridge, J. J. 2009, MNRAS, 396, 1699Stancliffe, R. J., & Glebbeek, E. 2008, MNRAS, 389, 1828Stancliffe, R. J., & Jeffery, C. S. 2007, MNRAS, 375, 1280Stancliffe, R. J., Lugaro, M., Ugalde, C., Tout, C. A., G¨orres, J., & Wiescher, M. 2005,MNRAS, 360, 375Suntzeff, N. B., & Smith, V. V. 1991, ApJ, 381, 160Sweigart, A. V., & Mengel, J. G. 1979, ApJ, 229, 624Tautvaiˇsien˙e, G., Bariseviˇcius, G., Berdyugina, S., Chorniy, Y., & Ilyin, I. 2010a, BalticAstronomy, 19, 95Tautvaiˇsien˙e, G., Edvardsson, B., Puzeras, E., Bariseviˇcius, G., & Ilyin, I. 2010b, MNRAS,409, 1213Traxler, A., Garaud, P., & Stellmach, S. 2011, ApJ, 728, L29+Ulrich, R. K. 1971, ApJ, 168, 57van Leeuwen, F. 2007, A&A, 474, 653Vandenberg, D. A. 1992, ApJ, 391, 685Zboril, M., & Messina, S. 2009, Astronomische Nachrichten, 330, 377 15 –
This preprint was prepared with the AAS L A TEX macros v5.2.
16 –Table 1. Summary of parameters of the λ And primary
Spec. Type Dist (pc) a P rot (d) b T eff (K) c log g c [Fe/H] c M ( M ⊙ ) d R ( R ⊙ ) e log L/L ⊙ f L X (erg s − ) g G8III-IV+ 25.8 54 4800 ±
100 2 . ± . − . . +1 . − . . ± . . ± .
04 2 . × . a Hipparcos distance from van Leeuwen (2007) . b Landis et al. (1978) c Adopted here based on estimates from Savanov & Berdyugina (1994); Donati et al. (1995); Ottmann et al. (1998); Soubiran et al.(2008); Tautvaiˇsien˙e et al. (2010a); [Fe/H] is expressed relative to the solar composition
F e/H = 7 .
50 (Grevesse & Sauval 1998;Asplund et al. 2009). d Spectroscopic mass from L bol , T eff and log g . e Spectroscopic radius based on L bol and T eff . f Tautvaiˇsien˙e et al. (2010a); uncertainty estimated here based on the Alonso et al. (1999) bolometric correction change for a 100 Ktemperature uncertainty at our adopted T eff . g Dempsey et al. (1993), based on
ROSAT
All-Sky Survey observations, updated to the
Hipparcos distance of van Leeuwen (2007).
Table 2. Measured line intensities (in counts) for λ And
Chandra XMM-Newton λ (˚A) Ion ObsID 2558 ObsID 3722 A eff RGS2 A eff Transition24.779 N VII 163.1 ±
34 211.7 ±
36 15.2 136 . ±
19 46.7 (2p) 2P / → (1s) 2S / ±
27 376.1 ±
35 11.6 138 . ±
18 22.6 (2p) 2P / → (1s) 2S /
17 –Fig. 1.— The λ And LETGS spectrum in the 20-40 ˚A region, showing C and N lines,compared with β Ceti (evolved) and ǫ Eri (unevolved). The λ And spectrum has beenmultiplied by 3 for clarity. The λ And C and N line strengths are in a similar ratio to thatof the unevolved dwarf ǫ Eri and show no significant signs of the C depletion evident in thespectrum of β Ceti. 18 –Fig. 2.— Comparison of the effective temperature and luminosity of λ And with evolutionarytracks computed using the STARS program for metallicity Z = 0 .
008 (see text). “Dredge-up end” indicates the point at which the surface abundances are within 5% of their finalpost-dredge-up values. λ And finished dredge-up about 50 Myr after this point. 19 –Fig. 3.— Model predictions of the post-dredge-up surface C/N abundance ratio (by number)as a function of the C/ C ratio for stars of different mass (denoted by symbol size), fromSchaller et al. (1992); Vandenberg (1992) and Girardi et al. (2000). Also shown are the C/ C and C/N ratios estimated for old disk giants by Cottrell & Sneden (1986), togetherwith a representative 0.2 dex error bar. The X-ray λ And C/N ratio is illustrated togetherwith the C/ C ratio from Tautvaiˇsien˙e et al. (2010a). 20 –Fig. 4.— Similar to Figure 3 but illustrating STARS predictions of the post-dredge-up surface C/ C vs C/N trajectories of models including thermohaline mixing on the RGB comparedwith the ratios estimated for old disk giants by Cottrell & Sneden (1986). The X-ray λ AndC/N ratio is illustrated together with the C/13