1. ROTATING MASSIVE STAR MODELS: FROM PRIMORDIAL STARS TO HIGH METALLICITY REGIONS Georges Meynet, André Maeder, Raphael Hirschi and Sylvia Ekstroem Geneva Observatory

2. Massive stars can be seen far away in the universe galaxyD[kpc]m v (B superG) LMC468.5 SMC639.8 M3172414.3 M81330017.7 M1001700026.5 VLT limiting magnitude 28.5, a 25 M sol star can be detected up to distances > 70 Mpc O- and B- stars contribute to about 2/3 of the optical light of galaxies

3. Filaments from Supernovae Effects of SN in a galaxy: GALACTIC WINDS Ionising sources, energy and momentum sources, nucleosynthetic sites

4. Von Zeipel 1924; Eddington 1925; Vogt 1925 An old topic… ROTATION…

5. Star deformation due to its fast axial rotation … but quite topical nowadays Link between Long GRB and Hypernova confirmed Dominiciano de Souza et al. 2003 Hjorth et al. 2003 Cf also van Belle et al. 2003

6. Spectrum of a Type Ic Hypernova seen in the afterglow of the GRB 030329 Link between some GRB and SN Explosion confirmed ! Hjorth et al. 2003, Nature, 423, 847 Stanek et al. 2003, ApJ, 591, L17 E kin =4 x 10 52 ergs 0.35 M sol of 56 Ni Mass of ejecta ~8 M sol Progenitor mass 25-30 M sol

7. Nine of 17 O-type stars show a surface enrichment of N up to a solar level, [N]=7.92. Heap and Lanz 2003; 2005 O-type stars in the SMC

8. PHYSICS OF ROTATION STRUCTURE Oblateness (interior, surface) New structure equations Shellular rotation MIXING Meridional circulation Shear instabilities + diffusion Horizontal turbulence Advection + diffusion of angular momentum Transport + diffusion of the chemical elements MASS LOSS Increase of mass loss by rotation Anisotropic losses of angular momentum

9. GRATTON- ÖPIK CELL Cells of meridional circulation Very important process for the transport of the angular momentum Outwards and inwards transport of angular momentum Occurs when star deformed… 20 M sol on the ZAMS

10. Meridional circulation Gradients of Shear instabilities Zahn 1992: strong horizontal turbulence, shellular rotation Transport of the chemical species Transport of the angular momentum

11. Schear diffusion coefficient Maeder 1997, Talon and Zahn 1997 Velocity of the meridional currents Maeder and Zahn 1998

12. FOR HIGH M MIXING TIME < MS TIMESCALE WHY MIXING IN MASSIVE STARS ? For 15 – 120 M O

13. « … the radiation observed to be emitted must work its way through the star, and if there were too much obstruction it would blow up the star. »

14. The Von Zeipel theorem (1924) F rad  g eff

15. iso mass loss For stellar formation also Maeder, 1999 ; cf. Owocki, 1996 STELLAR WINDS & ROTATION

16. Idem with Teff =25000 K

17. van Boekel et al. 2003

18. The present wind around Eta Carinae is elongated along a direction aligned with the Homunculus Nebula Smith et al. 2003 also indicate a latitude dependent wind velocity, with the highest velocities near the poles Support polar enhanced mass loss.

19. Different evolution of rotation The account for anisotropic mass loss favours break- up Mainly for 20 - 60 Mo, high rot. Maeder, 2002 Anisotropic mass loss 40 M sol, V ini =400 km/s

20. GLOBAL MASS LOSS RATES Maeder and Meynet 2000 Enhancement at break-up velocity Log Teff 4.35 4.30 4.00 3.90 !

21. Meynet and Maeder 2003 New grids of stellar models Also Z=0.040; 0.008, 0.004, 0.00001 +Pop III

23. N/C grows during the MS, even for early B stars (cf.Lyubimkov 1996) OK with B, A supergiants (cf. Gies & Lambert 1992; Lennon 1994; Venn 1998,…) (cf. Maeder, 1987; Langer, 1992; ….) 300 km/s 200

24. WHAT CHANGES AT VERY LOW Z FOR ROTATING MODELS ? Less angular momentum removed by stellar winds Steeper gradients of the angular velocity in the interiors MORE EFFICIENT MIXING BREAK-UP LIMIT Meridional velocities smaller MORE ANGULAR MOMENTUM IN THE CORE

25. Gradients of  steeper at lower metallicity 20 M sol, X c mass fraction of H at the centre, V ini = 300 km/s Why ? Stars more compact, transport of angular momentum less efficient Consequences ?More efficient mixing of the chemical elements

26. 9 M sol When Z Surface enrichments

27. Venn & Przybilla 2003 Max/ini N/H =40 Max/ini N/H =8 Log (N/H)+12 8.88.48.07.67.26.86.4 Number of stars

28. B/R PROBLEM Lots of RSG observed at low Z, but current models predict none. B/R ~ 50 Langer & Maeder, 1995 Models with rotation are OK with B/R = 0.5–0.8 in SMC cf. Maeder & Meynet 2001

29. with rotation With rotation: - Larger core - More He in shell - H shell less active - no intermed. conv. zone RSG Y M r /M sun

30. CONSISTENT WITH MODELS More fast More Rotators RSG

31. NUMBER RATIOS OF MASSIVE STARS IN NEARBY GALAXIES M310.0350.240.44 1.7 6-7.50.0290.210.55 -- 7.5-90.0200.1040.48 ~1 9.5- 11 0.0130.0330.33 -- M330.0130.060.52 ~4 LMC0.0060.040.20 -- 68220.0050.02 -- 8.3 SMC0.0020.0170.11 -- 16130.0020.02 GALAXY Z WR/O WC/WR RSG/WR Conti & Maeder’94; Massey ‘02

32. What are the effects of rotation on the Wolf-Rayet star formation process ? Maeder 1987 Fliegner and Langer 1995 Meynet and Maeder 2003 Hot stars  Log T eff > 4.0 Mass fraction of hydrogen At the surface below 0.4 In non-rotating models: Mass loss = the key parameter In rotating models: Rotational diffusion and mass loss

33. For a given metallicity, the minimum initial mass of single stars which become Wolf-Rayet star is decreased for higher rotation velocities 37Msol 22Msol WR lifetimes also increased for a given initial mass

34. When the velocity increases The WR lifetime increases The WNL phase increases EFFECT ON THE WOLF-RAYET LIFETIMES WNL WNE WC WN/WC Meynet and Maeder 2003

35. Mass loss rates proportional to the number of strong lines Number of strong lines proprtional to Z Wind models for hot stars show this effect ZMass loss [10 -6 M sol /y.] 0.0202.12 0.0061.35 0.0020.72 O5V Kudritzki et al 86 VERY IMPORTANT CONSEQUENCES

36. Mass loss rates depend on Z Cf Kudritzki 2002 NO MASS LOSS INITIAL MASS FINAL MASS Mass loss rates from Vink et al. 2000;2001 V ini =300 km/s Meynet and Maeder 2004

37. 20Msol 22Msol 25Msol 40Msol M min WNE Meynet and Maeder 2004

39. Observed points from Prantzos and Boissier (2003) Meynet and Maeder 2004

40. Smoother evolution of the surface abundances Presence at the surface of both H and He-burning products (Langer 91; Crowther 95,2002) During the WC phase, 22 Ne overabundance in agreement with observations (Willis 97; (Dessart et al. 2000)

41. 25 Msol: from core H-burning to Si-burning Hirschi, Meynet, Maeder, 2004 V=0 km/s V=300 km/s Heger, Langer, Woosley 2000

42. Consequences for the nature of the supernova progenitor Depending on its initial rotational velocity a star may end its evolution as a red supergiant or as blue supergiant, or even as a Wolf-Rayet star 1 2 3 With rotation 1) Tracks bluer 2) Redwards evolution favoured, more time spent at the RSG stage, more mass lost 3) Return to the blue favoured Hirschi et al. 2003

43. Hirschi et al. 2004 For M ini < M minWR ROTATION  increases 12 C and 16 O by about a factor 2 A non rotating ~30 M sol as a rotating 20 M sol For M ini > M minWR ROTATION  increases 4 He

44. YIELDS x IMF

45. Pettini et al 2002 Metal-poor dwarfs of the Solar neighborhood Carbon et al. 1987 HII regions DLA Pagel 1997 Garnett 1990 PRIMARY NITROGEN

46. A new mechanism induced by rotation S-process ? Cf. Herwig et al, 2003

47. For Z=0.004 and Z=0.020, nearly no primary N production

48. At Z= 0, stars are more compact Feijoo 1999 diploma work DELTA Log T eff ~0.3 PopIII star: radii decreased by a factor 4 Ekström 2004 diploma work

49. Z = 0.020 PopIII 60 M sol V ini = 300 km/s V ini ~ 800 km s -1

50. 14 N 16 O 12 C 4 He 1H1H 1H1H 25M sol ~30M sol M co larger in the rotating model 60 M sol Pop III Chemical composition of the radiative envelope PRIMARY 13 C and 14 N

51. 14 N 16 O 12 C 4 He 1H1H 1H1H 25M sol ~30M sol M co larger in the rotating model 60 M sol Pop III Chemical composition of the radiative envelope PRIMARY 13 C and 14 N elementNon-rotatingrotating 12 C23.7 M sol 26.5 M sol 13 C1.8E-08 M sol 1.4E-02 M sol 14 N2.4E-07 M sol 5.1E-02 M sol 16 O14.5 M sol 17.25 M sol

52. CONSEQUENCES FOR NUCLEOSYNTHESIS Z = 0 Z = 0.00001 Non rotating rotating -6.6

53. New data 2004 Spite et al. 2004 Israelian et al. 2004 NEED OF IMPORTANT PRIMARY PRODUCTION BY MASSIVE STARS

54. MIGHT THE STAR LOOSE MASS BY OTHER PROCESSES ? Mass loss rates much lower Cf Kudritzki, Hillier NO MASS LOSS INITIAL MASS FINAL MASS Mass loss rates from Vink et al. 2000;2001 V ini =300 km/s Meynet and Maeder 2004

55. WHAT CHANGES AT VERY LOW Z FOR ROTATING MODELS ? Less angular momentum removed by stellar winds Steeper gradients of the angular velocity in the interiors MORE EFFICIENT MIXING BREAK-UP LIMIT Meridional velocities smaller MORE ANGULAR MOMENTUM IN THE CORE

56. WHAT CHANGES AT LOW Z FOR ROTATING MODELS ? Less angular momentum removed by stellar winds Steeper gradients of the angular velocity in the interiors MORE EFFICIENT MIXING BREAK-UP LIMIT

57. Z= 0.020Z= 0.004 Z= 0.00001 V ini on the ZAMS = 300 km/s V surf Age [in My]

58. Mass Fraction of Hydrogen at the centre 200 M sol 85 M sol 40M sol 60M sol Pop III stellar models M ini Mass lost on MS Phase 40 M sol 1 M sol 60 M sol 2 M sol 85 M sol 4 M sol 200 M sol 16 M sol

59. Mass Fraction of Hydrogen at the centre 200 M sol 85 M sol 40M sol 60M sol Pop III stellar models

60. CONSEQUENCES FOR NUCLEOSYNHESIS Z = 0.00001 Non rotating rotating

61. Maeder, Grebel, Mermilliod 1999 Rotation seems faster at lower Z Is this a general trend ? What at Z = 0 ? From 19 clusters in Galaxy, LMC & SMC

62. CUMULATIVE DISTRIBUTION OF V sin i Keller 2004 LMC MW 100 early B-type MS stars in LMC Galactic young clusters =116 km/s MW 20% > 200 LMC 35% > 200 LMC young clusters =146 km/s Keller 2004

63. ! Could very low metallicity stars loose a lot of mass when reaching the break-up ?

64. 60 M sol, Z = 0.00001, V ini =800 km s -1 300 km s -1 800 km s -1 2/3 of the Main Sequence phase spent near the break-up limit

65. 60 Msol, [Fe/H]=-3.3 and –6.3 Non rotating Log T eff YcYc Rotating

66. 14 N 12 C 16 O Y c = 0.40 Z surf /Z ini =1 Y c = 0.12 Z surf /Z ini =64 Y c = 0.08 Y c = 0.02 Z surf /Z ini =392 Z surf /Z ini =1336 4 He [Fe/H]=-3.3 Case for [Fe/H]=-6.3 Very similar

67. COMPOSITION OF THE WIND EJECTA

68. Christlieb et al. 2004 HE0107-5240 [Fe/H]~-5.3 Depagne et al. 2002 CS 22949-037 [Fe/H]~-4 Norris et al. 2001 CS 22949-037 [Fe/H]~-4 Aoki et al. 2002 CS 29498-043 [Fe/H]~-3.75 COMPARISON WITH C-RICH EMP STARS

71. Evolution = f (M, Z, Ω, …)

72. ROTATING MODELS Surface enrichements Blue to red supergiants ratio at low metallicity Wolf-Rayet to O-type stars at various metallicities Type Ibc to type II supernovae at various metallicities At low metallicity predict higher enrichments higher velocities primary Nitrogen Pulsar rotation rates/GRB progenitors very metal poor stars

73. EFFECTS OF ROTATION AT VERY LOW METALLICITY ROTATIONAL MIXING 13 C and 14 N produced in great quantities May loose half of their initial mass through stellar winds NUCLEOSYNTHESIS Pair instability supernovae avoided ? ROTATIONAL MASS LOSS

74. Do the models reproduce the observed rotation rate of young pulsars ? Observed rotation periods of young pulsars: 2ms – 100 ms (20ms) Middletich et al. 2000; Romani & Ng 2003; Marshall et al. 2003 Pre SN For NS with P=20ms 25 – 320 X angular momentum in young pulsars In pre SN stages, more efficient angular momentum Processes ? Loss of angular momentum during the collapse ? Fryer and Warren 2004

75. COLLAPSAR: Woosley, 2002 Relation SN - GRB Precursor: Rotating WR star ? Is there enough rotation ? 1 % of all WR would be enough. Hjorth et al. 2003

76. Conditions for a collapsar Pre SN For NS with P=20mms For NS at break-up Woosley 2003   cm 2 s -1 Z sol To have lost its H-rich envelope, to be a WO star (M he > 8 M sol -> BH) WR To have sufficient angular momentum But No WO STARS !

78. Z SMC Z sol WR Candidates for Collapsars, reduced region at low Z

79. Z SMC Z sol WR Candidates for Collapsars, reduced region at low Z WO At Z=0.004 ~1% of the Core collapse supernovae are of type Ic

80. Conclusions: Evolution = f (M, Z, Ω) Evolution of rotational velocities Lifetimes, tracks Evolution properties Be, B[e], LBV, WR stars in galaxies Nebulae Cepheid properties Surface abundances in massive stars and red giants Primary N Pre – supernova stages Chemical yields and nucleosynthesis Rotation periods of pulsars Final masses Collapsars, γ- bursts, ….

81. A correct treatment of the transport of angular momentum in all phases is necessary ! If so, high final ang. momentum Hirschi, Meynet & Maeder, 2004

83. Age in Myr 300 km/s 800 km/s MASS LOST DUE TO THE APPROACH OF THE BREAK-UP LIMIT End MS 00.20.4 3000.39.3 8005.823.5 V ini Km/s Mass lost on MS in M sol Mass lost After MS Effect break-up M dot ~3 10 -6 M sol /yr Redwrads evolution and CNO enhanced at the surface

84. Umeda and Nomoto 2003

85. Marigo, Chiosi, Kudritzki 2003 M he =31; M CO =28 M sol

86. COMPOSITION OF THE WIND EJECTA C N O F Ne Na Mg Al

87. COMPOSITION OF THE WIND EJECTA C N O F Ne Na Mg Al Depagne et al. 2002 CS 22949-037 [Fe/H]~-4 Norris et al. 2001 CS 22949-037 [Fe/H]~-4 Aoki et al. 2002 CS 29498-043 [Fe/H]~-3.75 Christlieb et al. 2004 HE0107-5240 [Fe/H]~-5.3 12 C/ 13 C~4

88. Hypothesis M fer ~ 0.02 M sol WIND AND SUPERNOVA EJECTA C N O F Ne Na Mg Al

89. Angular momentum at the pre SN stage Specific angular momentum in the central part of the star which eventually will be locked into the compact remnant ZAMS End core He-burning Begin WR phase Changes of specific angular  momentum are only due to internal transport processes Decreases as a function of time A great part of the decrease occurs during the MS phase 120 M sol  ini /  fin =40 9 M sol  ini /  fin =2.3

91. Maeder & Meynet 2003 15M sol, Z = 0.02, V ini = 300 km s -1 No magnetic fields 15M sol, Z = 0.02, V ini = 300 km s -1 With magnetic fields MAGNETIC FIELDS ? According to Heger et al. 2004 magnetic braking can slow down the core by about an order of magnitude. Help but still insufficient No long surface enrichments

92. 15M sol, Z = 0.02, V ini = 300 km s -1 No magnetic fields 15M sol, Z = 0.02, V ini = 300 km s -1 With magnetic fields

93. New models with magnetic fields accounting for a local energy condition The energy for building the magnetic field is extracted from the energy in the differential rotation The amplitude of the magnetic field is determined by the energy available in the shear in the zone where it develops. In preparation Magnetic field appears mainly in the outer parts of the star (inhibiting effect of the gradient of the mean molecular weight) Evolution with magnetic field very similar to that obtained without magnetic field These results would favour the loss of angular momentum at the time of collapse for accounting the observed rotation rate of young pulsars

94. Gies & Huang 2003

95. ROTATION AND ASTEROSEISMOLOGY Xc=0.33 Xc=0.44    Hz    Hz Eggenberger 2003 Determination of  Rotation changes the values of the small separation

96. Has rotation an impact on the quantity of 26 Al released ?

97. cf Langer et al. 95Vuissoz et al., in preparation

98. Quantity of 26 Al ejected by the winds is increased by rotation Effect different from an enhancement of the mass loss rate Vini0 km/s300km/s500km/s M( 26 Al) 10 -4 Ms 1.302.18 X 1.68 2.61 X 2.00 Knödlseder et al. 2002  in Cygnus region models underestimate 26 Al production by about a factor 2

101. Pettini et al 2002 Metal-poor dwarfs of the Solar neighborhood Carbon et al. 1987 HII regions DLA Adapted by Pagel 1997 from Garnett 1990 See also Matteucci and Tosi 85 Matteucci 86

103. Increase of primary N production when rotation increases

104. Prantzos 2003 Contribution from rotation of the same order of magnitude as contribution from classical models of thermal pulse AGB stars. See also Carigi 2003; Chiappini et al. 2003

105. Pop III Ekstroem et al. in preparation

106. Z=0.004 At lower Z, more stars reach break- up velocities. PARADOXICAL !

107. SHORT EJECTION Peanut shaped nebulae Maeder,1999 cf also Owocki et al. 1996 Fast rotating hot stars have Polar winds

108. When Z decreases stars become more compact In U the Gratton Öpick scales with the Inverse of the density Less angular momentum is removed by the stellar winds

109. ? Ekstroem, Meynet, Maeder in preparation, cf Marigo, Chiosi, Kudritzki 2003 600 km/s 175 km/s 300 km/s

110. End core Si-burning phase HIRSCHI 2004 IN ROTATING MODELMore massive Si-core V = 0 M Si = 1.6 M sol V=300 M Ne = 2.2 M sol Slightly more massive Ni-core V = 0 M Ni = 1.1 M sol V=300 M Ne = 1.25 M sol

112. Christlieb et al. 2004 HE0107-5240 [Fe/H]~-5.3 Depagne et al. 2002 CS 22949-037 [Fe/H]~-4 Norris et al. 2001 CS 22949-037 [Fe/H]~-4 Aoki et al. 2002 CS 29498-043 [Fe/H]~-3.75 COMPARISON WITH C-RICH EMP STARS F Ne