Origins of blue straggler stars in open clusters, globular clusters and dwarf spheroidal galaxies 散開星団、球状星団、矮小銀河における青い彷徨い星の起源の解明
Origins of blue straggler stars in open clusters, globular clusters and dwarf spheroidal galaxies
Blue straggler stars (BSSs) lie above the current main sequence turn offon the color-magnitude diagram of star cluster. Their positions on thecolor-magnitude diagram imply that they are more massive than currentmain sequence turn-off stars in the same cluster. Thus, their presence ofsuch an upper main sequence was a challenge for standard stellar evolutiontheory, as normal stars with the same mass have already evolved into whitedwarfs. To explain their extraordinary long lifespan on the main sequence,various mechanisms have been suggested, all involving in additional hydrogenmaterial input to normal stars. Two of the most popular mechanismsare merger by direct collisions in single stars and coalescence or mass transferin close binary systems. As products of dynamical processes in the hostcluster and stellar evolution itself, studying BSSs can shed a light on theprocess of merger and/or mass transfer, and be a tracer of the dynamicalevolution of the host cluster.The first discovery of BSSs in a globular cluster M3 dates back to sixdecades ago. Since then, many observational ndings suggest that BSSsare present in globular clusters, open clusters, as well as dwarf spheroidalgalaxies (dSphs), and even in the elds and the bulge of the Milky Wayand imply that dierent formation mechanisms may operate in differentdynamical environments. Though we are still facing with problems in understandingthese objects clearly and completely. The aim of this thesis isto derive a comprehensive interpretation for origins of BSS by investigatingthe BSS properties in dierent star systems, such as open clusters, globularclusters and dSphs based on observational data of the highest quality up tothe present day.We have collected homogeneous deep photometric data for 9 dSphstaken from Subaru/ Suprime-Cam. The sample includes 3 classical dSphs,Draco, Ursa Minor (UMi) and Sextans and 6 ultra faint dwarf galaxies. Forthese classical dSphs, we have analyzed the BSSs' radial distributions andluminosity functions. BSSs radial distribution shows a lack of strong centralpeak and luminosity function is independent of radius. These results implythat in these dSphs, the blue plume stars are most likely to be real BSSsformed mainly by mass transfer in primordial binaries. The frequency ofBSSs in the 9 galaxies does not correlate with galaxy parameters, such ashalf light radius rh, metallicity [Fe/H], central surface brightness μV;C, concentrationparameter c, central density pc, velocity dispersion δV , half massrelaxation time log trh and collisional rate Γ , as well as integrated absolutemagnitude, MV. This result suggests that the evolution of primordial binaries,which would form BSSs, was not affected significantly by dynamicalinteractions in such sparse environments. Thus, the observed BSS frequencyis almost the same in these galaxies under the simple hypothesis that theprimordial binary fraction was the same in all galaxy.We have also collected homogeneous deep photometric data for 51 GCobtained by ACS/HST. 8 GCs in our sample show bimodal BSSs frequencyradial distributions. Among them, the bimodal distribution is reported forthe first time in NGC 5286, NGC 5986, NGC 6362, NGC 6535, NGC6541and NGC7099, in total 6 GCs. It has raised the number of this kind ofGCs by twice the number known before. Moreover, the significant anticorrelationbetween BSS frequency and MV is confimed with our sample.BSS frequency is also significantly anti-correlated to central density, centralvelocity dispersion and collision rate. Furthermore, luminosity functions ofBSSs in bright (MV < -8.8) and faint (MV > -8.8) clusters do not showsignificant differences. One possible interpretation for all these properties isthat BSSs formation from binaries is the dominant mechanism in globularclusters. However, The effect of collision on BSS formation is still considerablebecause collisions destroy primordial binaries and/or accelerateevolution of binary systems.Finally, we have observed 30 BSSs and 9 main sequence turn-off candidatesin three old Galactic open clusters, NGC 2682, NGC 2112 and NGC6819. Among them, radial velocities (Vr) of all target stars in NGC 2682and NGC 6819, and 5 BSSs and all TOs in NGC 2112 are consistent withmean Vr of member stars in the clusters. Rotational velocity (vsini ) ofBSSs spans a wide range from few kms-1 to hundreds km_s-1, while thoseof turn-off stars vary in small range from few kms-1 to 25 kms-1. Thehigh rotation velocity of current BSS is consistent with predictions bothfrom mass transfer and collision formation mechanism. Some turn-off starsand a few of BSSs show normal Li abundances. For the other BSSs, onlythe upper limit of Li abundance is estimated. Interestingly, BS1371 in NGC2112 and BS 10010 in NGC 6819, show very high Li abundance, A(Li) 3.94dex and A(Li) 3.52 dex. Two BSSs show depletion in oxygen abundance,while the most of BSSs show similar oxygen abundance to turn-off stars.Some BSSs show depletion in carbon abundance, while other BSSs showthe normal carbon abundance comparing to turn-off stars. These abundanceanomalies in light elements found in some BSSs suggest that masstransfer from such highly evolved stars is the origin of these BSSs. On theother hand, the progenitor of other BSSs with normal carbon and oxygenabundances and low lithium abundances might be normal red giant stars.