The Hottest Horizontal-Branch Stars in omega Centauri - Late Hot Flasher vs. Helium Enrichment
S. Moehler, S. Dreizler, T. Lanz, G. Bono, A.V. Sweigart, A. Calamida, M.Monelli, M. Nonino
aa r X i v : . [ a s t r o - ph ] S e p Astronomy&Astrophysicsmanuscript no. ocen c (cid:13)
ESO 2018October 28, 2018
The Hottest Horizontal-Branch Stars in ω Centauri – Late HotFlasher vs. Helium Enrichment ⋆ S. Moehler , S. Dreizler , T. Lanz , G. Bono , A.V. Sweigart , A. Calamida , M. Monelli , and M. Nonino European Southern Observatory, Karl-Schwarzschild-Str. 2, D 85748 Garching, Germany e-mail:[email protected] Georg-August-Universit¨at, Institut f¨ur Astrophysik, Friedrich-Hund-Platz 1, D 37077 G¨ottingen, Germany e-mail:[email protected] Department of Astronomy, University of Maryland, College Park, MD 20742-2421, USA e-mail:[email protected] INAF - Rome Astronomical Observatory, via Frascati 33, 00040 Monte Porzio Catone, Italy e-mail:bono,[email protected] NASA Goddard Space Flight Center, Code 667, Greenbelt, MD 20771, USA e-mail:[email protected] IAC - Instituto de Astrofisica de Canarias, Calle Via Lactea, E38200 La Laguna, Tenerife, Spain e-mail:[email protected] INAF - Trieste Astronomical Observatory, via G.B. Tiepolo 11, 40131 Trieste, Italy e-mail:[email protected] online version: October 28, 2018
ABSTRACT
Context.
UV observations of some massive globular clusters uncovered a significant population of very hot stars below the hot end of thehorizontal branch ( HB ), the so-called blue hook stars. This feature might be explained either as results of the late hot flasher scenario wherestars experience the helium flash while on the white dwarf cooling curve or by the progeny of the helium-enriched sub-population recentlypostulated to exist in some clusters. Previous spectroscopic analyses of blue hook stars in ω Cen and NGC 2808 support the late hot flasherscenario, but the stars contain much less helium than expected and the predicted C, N enrichment could not be verified.
Aims.
We want to compare e ff ective temperatures, surface gravities, and abundances of He, C, and N of blue hook and canonical extremehorizontal branch ( EHB ) star candidates to the predictions of the two scenarios.
Methods.
Moderately high resolution spectra of stars at the hot end of the blue HB in the globular cluster ω Cen were analysed for atmosphericparameters and abundances using LTE and Non-LTE model atmospheres.
Results.
In the temperature range 30,000 K to 50,000 K we find that 35% of our stars are helium-poor (log n He n H < − − . . log n He n H . − .
5) and 14% are helium-rich (log n He n H > − . Conclusions.
At least 14% of the hottest HB stars in ω Cen show helium abundances well above the highest predictions from the heliumenrichment scenario ( Y = n He n H ≈ − . ω Cen cannot be explained solely by the helium-enrichment scenario invoked to explain the blue main sequence.
Key words.
Stars: horizontal branch – Stars: evolution – Techniques: spectroscopic – globular clusters: individual: NGC 5139
1. Introduction
UV-Visual colour-magnitude diagrams of the two very massiveglobular clusters, ω Cen and NGC 2808, show a rather puz-zling “hook-like” feature at the hot end of their extended hori-zontal branches with stars lying below the canonical horizontalbranch (Whitney et al. 1998; D’Cruz et al. 2000; Brown et al.2001). These stars cannot be explained within the frameworkof canonical stellar evolution. Brown et al. (2001) have pro-posed a “flash-mixing” scenario to explain the blue hook stars.According to this scenario stars which lose an unusually largeamount of mass will leave the red giant branch (
RGB ) before ⋆ Based on observations with the ESO Very Large Telescope atParanal Observatory, Chile (proposal IDs 075.D-0280(A) and 077.D-0021(A)) the helium flash and will move quickly to the (helium-core)white dwarf cooling curve before igniting helium (Castellani& Castellani 1993; D’Cruz et al. 1996; Brown et al. 2001).However, the evolution of these “late hot helium flashers” dif-fers dramatically from the evolution of stars which undergothe helium flash on the RGB. Ordinarily when a star flashesat the tip of the RGB or shortly thereafter, the large entropybarrier of its strong hydrogen-burning shell prevents the prod-ucts of helium burning from being mixed to the surface. Suchcanonical stars will evolve to the zero-age horizontal branch(
ZAHB ) without any change in their hydrogen-rich envelopecomposition. In contrast, stars that ignite helium on the whitedwarf cooling curve, where the hydrogen-burning shell is muchweaker, will undergo extensive mixing between the helium-and carbon-rich core and the hydrogen envelope (Sweigart
S. Moehler et al.: The Hottest Horizontal-Branch Stars in ω Centauri – Late Hot Flasher vs. Helium Enrichment α carbon during the flash-mixing phase. For both deep and shallow mixing, the blue hookstars should be helium-rich compared to the canonical EHBstars.Alternatively, the recently observed split among the mainsequence stars of ω Cen and NGC 2808 (Piotto et al. 2005,2007) has been attributed to a sub-population of stars with he-lium abundances as large as Y ≈ Y ≈ ω Cenyield a carbon abundance of [C / M] = ω Cen have the low C / C ratios ( ≈
4) and low average carbonabundances ([C / Fe] = − ω Cen (Moehleret al. 2002) and NGC 2808 (Moehler et al. 2004) showed thatthese stars are indeed both hotter and more helium-rich than thecanonical EHB stars. However, the blue hook stars still showconsiderable amounts of hydrogen. Unfortunately due to lim-ited resolution and signal-to-noise ( S / N ) we could not derivegood abundances for C and N. Instead we could only state thatthe most helium-rich stars appear to show some evidence forC / N enrichment. Therefore we started a project to obtain higherresolution spectra of EHB and blue hook stars in ω Cen.
2. Observations, Data Reduction, and Analysis
We selected stars along the blue HB in ω Cen from the multi-band ( U , B , V , I ) photometry of Castellani et al. (2007). Thesedata were collected with the mosaic CCD camera Wide FieldImager available at the 2.2m ESO / MPI telescope. The field ofview covered by the entire mosaic is 42 ′ × ′ across the cen- Fig. 1.
Here we show the e ff ective temperatures and surfacegravities derived for our hottest target stars (formal errorsonly). Helium-poor, solar helium, and helium-rich stars aremarked by red, green, and blue squares, respectively. The solidlines mark the canonical HB locus for [M / H] = − = TAHB). The dotted line connects the series of ZAHB mod-els computed by adding a hydrogen-rich layer to the surfaceof the ZAHB model of the late hot flasher. The small dotsmark – with decreasing temperature – hydrogen layer massesof 0 , − , − , − , − M ⊙ (for details see Moehler et al.2002).ter of the cluster. These data together with multiband data fromthe Advanced Camera for Surveys on board the Hubble SpaceTelescope provided the largest sample of HB stars ( ≈ ≈ . ′′ + GIRAFFE on the UT2 Telescope of the VLT. Weused the low spectroscopic resolution mode with the spectralrange 3964Å – 4567Å (LR2, R = / EHB starcandidates and 17 empty positions for sky background.For our analysis we used the pipeline reduced data. Foreach exposure we subtracted the median of the spectra fromthe sky fibres from the extracted spectra. We corrected all spec-tra for barycentric motions. The individual spectra of each tar- . Moehler et al.: The Hottest Horizontal-Branch Stars in ω Centauri – Late Hot Flasher vs. Helium Enrichment 3 get star have been cross-correlated with appropriate templatespectra, in order to search for radial velocity variations. Sincethe few spectra per object did not permit a sophisticated pe-riod search, we determined the standard deviation of the radialvelocity measurements for each star and compared it with theS / N ratio of the spectra. As expected, the standard deviationof the radial velocity measurements decreases with increasingS / N ratio. None of our target stars deviates significantly fromthis correlation, which would be the case for close binaries.Therefore none of our target stars appears to be in a close binarysystem. After verifying that there were no radial velocity vari-ations we co-added all spectra for each star. The co-added andvelocity-corrected spectra were fitted with various model atmo-spheres: metal-free helium-rich non-LTE (Werner & Dreizler1999), metal-free helium-poor non-LTE (Napiwotzki 1997),and metal-rich helium-poor LTE (Moehler et al. 2000) as de-scribed in Moehler et al. (2004). This procedure yielded the ef-fective temperatures, surface gravities, and helium abundancesshown in Figs. 1 and 2. In this paper we concentrate only onthe hottest HB stars with T e ff >
3. Results and Discussion
The helium-poor stars in Fig. 1 basically agree with the pre-dictions of canonical evolutionary theory in that they populatethe HB up to its hot end and then also contribute some evolvedstars at higher e ff ective temperatures and lower surface grav-ities. As we move to hotter stars (T e ff & ff ective temperature andsurface gravity between a fully mixed late hot flasher and thehot edge of the canonical HB. These stars show roughly so-lar helium abundance (cf. Fig. 2). The hottest stars lying alongthe evolutionary track of a fully mixed late hot flasher showthe highest helium abundances, albeit with still some hydro-gen in their atmospheres. In the temperature range 30,000 K to50,000 K we find that 35% (15) of our stars are helium-poor(log n He n H < − − . . log n He n H . − .
5) and 14% (6) are helium-rich(log n He n H > − . e ff > ii and C iii lines were detecteddespite the higher S / N in their spectra. We have constructedadditional TLUSTY NLTE line-blanketed model atmospheres(Hubeny & Lanz 1995; Lanz & Hubeny 2003, 2007) for the at-mospheric parameters of the helium-rich stars, assuming eitherscaled-solar abundances appropriate for the dominant ω Cenmetallicity ([M / H] = − ii and C iii lines indicates that thehelium-rich stars have a photospheric C mass fraction of at least1% and up to 2–3% for the stars with the strongest lines. Thetypical line detection limit provides an upper limit of about 1%by mass for the N abundance. These C abundances represent asignificant enhancement relative to the expected C abundancein ω Cen stars ( . Fig. 2.
Here we show the e ff ective temperatures and heliumabundances for our hottest target stars (formal errors only).The dashed line marks solar helium abundance, the hashed areamarks the range for the helium-enrichment scenario. The sym-bols have the same meaning as in Fig. 1. Fig. 3.
Here we show sample spectra of stars with super-solarhelium abundance, compared to model spectra with the clus-ter carbon abundance for metal-poor stars (red) and a carbonabundance of 3% by mass (blue). The labels give the numberof the star, its e ff ective temperature, and its helium abundancelog n He n H .Any discussion of the surface abundances in hot HB starsmust consider the e ff ects of di ff usion. Fortunately the di ff usionof H, He and the CNO elements in the envelopes of stars fol-lowing deep flash mixing has been investigated by Unglaub(2005). Not surprisingly, the results depend on the assumedmass loss rate and on the residual hydrogen abundance remain-ing after the flash mixing. For the low residual hydrogen abun-dance X = S. Moehler et al.: The Hottest Horizontal-Branch Stars in ω Centauri – Late Hot Flasher vs. Helium Enrichment log n He n H ≈
0. . . 2 during most of the HB phase, in rough agree-ment with the helium-rich stars in Fig. 2. However, the residualhydrogen abundance following flash mixing is quite uncertain,since it depends on the mixing e ffi ciency (Cassisi et al. 2003)and possibly on where the helium flash occurs along the whitedwarf cooling curve. For a larger, but still low, residual hydro-gen abundance of X = ff usion. The ZAHB models with ahydrogen-rich layer in Fig. 1 show that the e ff ective tempera-ture will decrease as the amount of surface hydrogen increasesin qualitative agreement with the trend towards lower e ff ectivetemperatures between the helium-rich and solar helium abun-dance stars in Fig. 2. Unglaub (2005) also noted that the dif-fusion e ffi ciency increases substantially once the helium abun-dance approaches the solar value, leading to a rapid decreasein log n He n H and perhaps accounting for the gap between the so-lar helium and helium-poor stars in Fig. 2. Di ff usion in flash-mixed stars also leads to a decrease in the carbon and nitrogenabundances, which becomes more pronounced when the atmo-sphere is hydrogen-rich. Thus the carbon abundances derivedhere for the helium-rich stars may underestimate the initial car-bon abundances in these stars.The referee asked us to discuss the possibility that di ff u-sion may not be active in all stars above 30,000 K. Let us con-sider the extreme case that only the helium-poor stars in thistemperature range are a ff ected by di ff usion. In this case themost helium-rich stars could still be reconciled with the latehot flasher scenario, but the same does not apply to the solar-helium stars. For the latter to be considered as the progeny ofthe helium-enriched main sequence stars, however, one wouldexpect to see in Fig. 2 stars highly concentrated at log n He n H = − . . . . − .
74 (i.e. Y = ff ect becomes evident if one plots the spatialdistribution of our target stars: Dividing the sample along a linerunning at 55 ◦ counter-clockwise from east-west, the helium-poor stars are evenly distributed (28:30 for all, 8:7 for thoseabove 30,000 K), while the stars with roughly solar or super-solar helium abundance show a noticeable preference for thenorth-west section of the globular cluster (17:5 and 5:1, re-spectively). This peculiar spatial distribution appears similar tothe reddening distribution observed by Calamida et al. (2005),who found a clumpy extinction variation with less reddenedHB stars concentrated on the east side of the cluster (see theirFig. 5).
4. Conclusions
All of these results taken together o ff er strong support forthe late hot flasher scenario as the explanation for the bluehook stars while posing a significant problem for the helium-enrichment scenario. This scenario predicts helium enrichmentof up to Y = . . . . .
42, i.e. log n He n H = − − This resulttogether with the observed carbon enhancement does not ruleout the helium enhancement scenario, but it implies that addi-tional processes are required to produce the hottest HB stars in ω Cen.
Acknowledgements.
We thank the sta ff at the Paranal observatory andat ESO Garching for their excellent work, which made this paper pos-sible. We also acknowledge that without the request from the ESOOPC to look into the data we have before applying again, these resultswould not have been found so soon. We thank the anonymous refereefor his / her suggestions. References
Brown, T. M., Sweigart, A. V., Lanz, T., Landsman, W. B., & Hubeny,I. 2001, ApJ, 562, 368Cassisi, S., Schlattl, H., Salaris, M., & Weiss, A. 2003, ApJ, 582, L43Calamida, A., Stetson, P. B., Bono, G., et al. 2005, ApJ, 634, L69Castellani, M., & Castellani, V. 1993, ApJ, 407, 649Castellani, V., Calamida, A., Bono, G., et al. 2007, ApJ, 663, 1021D’Antona, F., Bellazzini, M., Fusi Pecci, F., Galleti, S., Caloi, V., &Rood, R. T. 2005, ApJ, 631, 868D’Antona, F., Ventura, P. 2007, MNRAS, 379, 1431D’Cruz, N. L., Dorman, B., & Rood, R. T. 1996, ApJ, 466, 359D’Cruz, N. L., O’Connell, R. W., Rood, R. T., et al. 2000, ApJ, 530,352Gratton, R. G., Sneden, C., Carretta, E., & Bragaglia, A. 2000, A&A,354, 169Hubeny, I., & Lanz, T. 1995, ApJ, 439, 875Kraft, R. P. 1994, PASP, 106, 553Lanz, T., Brown, T. M., Sweigart, A. V., Hubeny, I., & Landsman,W. B. 2004, ApJ, 602, 342Lanz, T., & Hubeny, I. 2003, ApJS, 146, 417Lanz, T., & Hubeny, I. 2007, ApJS, 169, 83Lee, Y.-W., Joo, S.-J., Han, S.-I., et al. 2005, ApJ, 621, L57Moehler, S., Sweigart, A. V., Landsman, W. B., & Heber, U. 2000,A&A, 360, 120Moehler, S., Sweigart, A. V., Landsman, W. B., & Dreizler, S. 2002,A&A 395, 37Moehler S., Landsman W. B., Sweigart A. V., & Grundahl, F. 2003,A&A, 405, 135Moehler, S., Sweigart, A. V., Landsman, W. B., Hammer, N. J., &Dreizler, S. 2004, A&A 415, 313Napiwotzki, R. 1997, A&A, 322, 256Newsham, G., Terndrup, D. M. 2007, ApJ, 664, 332Norris, J. E. 2004, ApJ, 612, L25Origlia, L., Ferraro, F. R., Bellazzini, M., & Pancino, E. 2003, ApJ,591, 916Piotto, G., Villanova, S., Bedin, L. G., et al. 2005, ApJ, 621, 777Piotto, G., Bedin L. R., Anderson, J., et al. 2007, ApJ, 661, L53Sweigart, A. V. 1997, The Third Conference on Faint Blue Stars, ed.A. G. D. Philip, J. Liebert & R. A. Sa ff er (Schenectady: L. DavisPress), 3Unglaub, K. 2005, The 14th European Workshop on White Dwarfs,ASP Conf. Ser. Vol. 334, eds. D. Koester & S. Moehler (ASP:San Francisco), p. 297Werner, K., & Dreizler, S. 1999, The Journal of Computational andApplied Mathematics, Vol. 109, eds. H. Ri ff ert & K. Werner,Elsevier Press, Amsterdam, p. 65Whitney, J. H., Rood, R. T., O’Connell, R. W., et al. 1998, ApJ, 495,284Zacharias, N., Urban, S. E., Zacharias, M. I., Wyco ff , G. L., Hall, D.M., Monet, D. G., Ra ffff