Effects of low metallicity on the evolution and spectra of massive stars

Jose Groh (Geneva Observatory, Switzerland) Image credits: NASA/ESA/J. Hester & A. Loll, Arizona State U. (Crab Nebula) Effects of low metallicity on the evolution and spectra of massive stars

Take Away: effects of low metallicity Reduced stellar wind mass and angular momentum losses Stars are more compact, hotter, and more mixed Spectral lines are weaker and less numerous; more ionizing flux

Background on massive stars and stellar evolution Effects of low metallicity on stellar evolution Effects of low metallicity on stellar spectra Scary things Jose Groh - The surprising look of massive stars before death Image credits: NASA/ESA/J. Hester & A. Loll, Arizona State U. (Crab Nebula) Outline

Outline Background on massive stars and stellar evolution Effects of low metallicity on stellar evolution Effects of low metallicity on stellar spectra Scary things Jose Groh - The surprising look of massive stars before death

Star formation Star formationregions, chemical evolution Photoionized Supernova, GRBs, Black Holes, Neutron Stars evolution Chemical Universe first Stars stars, cosmology) Supernova, Black(starbursts, Holes, Neutron Distant Cosmology interstellar, circumstellar media Intergalactic, particle physics,... Intergalactic,physics, interstellar, circumstellar media High-energy evolution High-energy physics, particle physics,... Stellar Jose Groh - The surprising look of massive stars before death Image credits: NASA/ESA/J. Hester & A. Loll, Arizona State U. (Crab Nebula) Massive stars bridge many fields of (astro)physics

Luminosity Massive star evolution is challenging OB-type at ZAMS 4.8 Effective 4.4 Temperature 4.0 3.6 3.2

Massive star evolution is challenging OB-type at ZAMS Luminosity Red Super- Giant 4.8 Effective 4.4 Temperature 4.0 3.6 3.2

Massive star evolution is challenging Luminous Blue Variable Luminosity OB-type at ZAMS Yellow Hyper- Giant Red Super- Giant 4.8 Effective Temperature 4.4 4.0 3.6 3.2

Massive star evolution is challenging Luminosity Wolf-Rayet OB-type at ZAMS Luminous Blue Variable Yellow Hyper- Giant Red Super- Giant 4.8 Effective Temperature 4.4 4.0 3.6 3.2

Massive star evolution is challenging Luminosity Wolf-Rayet OB-type at ZAMS Luminous Blue Variable Yellow Hyper- Giant Red Super- Giant Science we want to do Evolutionary channels Initial mass ranges SN type + progenitor Single x binary evol. Metallicity effects,... 4.8 Effective Temperature 4.4 4.0 3.6 3.2

Single stellar evolution models binaries: de Mink talk Five flavors

Single stellar evolution models binaries: de Mink talk non-rotating Eldridge talk Five flavors

Single stellar evolution models binaries: de Mink talk non-rotating Eldridge talk Five flavors differential rotation (NO MAGNETIC FIELDS)

Single stellar evolution models binaries: de Mink talk Five flavors non-rotating Eldridge talk differential rotation (NO MAGNETIC FIELDS) Leitherer talk; Levesque talk

Single stellar evolution models binaries: de Mink talk non-rotating Eldridge talk Five flavors differential rotation (NO MAGNETIC FIELDS) Leitherer talk; Levesque talk solid-body rotation (WITH MAGNETIC FIELDS)Yoon talk differential rotation with surface magnetic braking solid-body rotation with surface magnetic braking Observational constraints are key asteroseismology: differential x solid-body rotation spectroscopy+polarimetry: rotation and magnetic fields Metallicity dependence?

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 60 40 32 25 20 15 12 9 (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 60 40 32 25 20 15 12 9 ~8 to 17 M (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 60 40 32 25 20 15 12 9 OB-type RSG SN IIP ~8 to 17 M (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 60 40 32 25 20 15 12 9 OB-type RSG SN IIP ~17 to 30 M ~8 to 17 M (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 60 40 32 OB-type RSG YHG/LBV SN ~17 IIL/b to 30 M 25 20 15 12 9 OB-type RSG SN IIP ~8 to 17 M (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 60 > ~30 M 40 32 OB-type RSG YHG/LBV SN ~17 IIL/b to 30 M 25 20 15 12 9 OB-type RSG SN IIP ~8 to 17 M (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Rotating massive star evolution models at solar metallicity Z=0.014 85 120 OB-type LBV WR SN > Ibc ~30 M 60 40 32 OB-type RSG YHG/LBV SN ~17 IIL/b to 30 M 25 20 15 12 9 OB-type RSG SN IIP ~8 to 17 M (Groh+ 13b after Ekstrom+12, Georgy+ 12)

Outline Background on massive stars and stellar evolution Effects of low metallicity on stellar evolution Effects of low metallicity on stellar spectra Scary things Image credits: NASA/ESA/J. Hester & A. Loll, Arizona State U. (Crab Nebula) Jose Groh - The surprising look of massive stars before death

A change of metallicity affects: 1. Opacities 2. Equation of state (via mean molecular weight) 3. Nuclear reaction rates 4. Fuel available during MS (X = 1 - Y - Z) 5. Mass and angular momentum losses from stellar winds 6. Rotational effects and angular momentum transport

Rotating massive star evolution models at SMC metallicity Z=0.002 (i.e. 1/7 Zsun) 120 85 60 40 32 25 20 15 12 9 >50 M 8 to 50 M (Groh+ 15 in prep after tracks from Georgy+ 13)

Rotating massive star evolution models at SMC metallicity Z=0.002 (i.e. 1/7 Zsun) 120 85 60 40 32 25 20 OB-type RSG SN IIP 15 12 9 >50 M 8 to 50 M (Groh+ 15 in prep after tracks from Georgy+ 13)

Rotating massive star evolution models at SMC metallicity Z=0.002 (i.e. 1/7 Zsun) OB-type YHG/LBV SN IIP or IIn 120 85 60 40 32 25 20 OB-type RSG SN IIP 15 12 9 >50 M 8 to 50 M (Groh+ 15 in prep after tracks from Georgy+ 13)

Rotating massive star evolution models at ~ IZw18 metallicity Z=0.0004 (i.e. 1/50 Zsun) 120 85 60 40 32 25 20 15 12 9 >50 M 8 to 50 M (Groh+ 2015 in prep)

Rotating massive star evolution models at ~ IZw18 metallicity Z=0.0004 (i.e. 1/50 Zsun) 120 85 60 40 32 25 20 OB-type RSG/BSG SN 8 IIP to 50 M 15 12 9 >50 M (Groh+ 2015 in prep)

Rotating massive star evolution models at ~ IZw18 metallicity 120 85 60 40 32 25 Z=0.0004 (i.e. 1/50 Zsun) OB-type YHG/LBV SN IIP or IIn 20 OB-type RSG/BSG SN 8 IIP to 50 M 15 12 9 >50 M (Groh+ 2015 in prep)

Metallicity trends in stellar evolution

Low metallicity: mass loss diminishes Groh+ 2015, in prep

Low metallicity: stars are more compact and hotter Groh+ 2015, in prep

Low metallicity: faster surface rotation For high mass stars: Reduced mass and angular momentum losses from stellar winds Georgy+ 2013

Low metallicity: chemical mixing is favored At low Z: stars are more compact; transport of angular momentum is less efficient. Relative Nitrogen enhancement blue= initial angular velocity Ω/Ωcrit = 0.10 purple= initial angular velocity Ω/Ωcrit = 0.95 Georgy+ 2013

How to compare observations and stellar evolution models? Issue: massive stars develop winds that become denser as the star evolves, hiding progressively more and more of the stellar surface.

How to compare observations and stellar evolution models? Issue: massive stars develop winds that become denser as the star evolves, hiding progressively more and more of the stellar surface. low-mass stars (e.g Sun) massive stars: low Mdot massive stars: high Mdot (evolved) rhydr = rphot Evol. code rhydr Evol. code rphot rphot rhydr Stellar Wind + Atmosphere Evol. code Observations Observations Observations Consequence: challenging to compare interior models of massive stars and observations.

Solution: merge with an atmospheric code Stellar Wind + Atmosphere Mass Luminosity Temperature Surface abundances Magnitudes Colors SED Spectrum rphot rhydr Geneva evol. code CMFGEN code Observations

Unified stellar evolution and atmospheric modelling non-rotating 60 M at solar Z Groh+ 2014, A&A 564, 30

Spectral evolution of a non-rotating 60 Msun star Groh+ 2014, A&A 564, 30

Spectral evolution of a non-rotating 60 Msun star Optical Groh+ 2014, A&A 564, 30

Unified stellar evolution and atmospheric modeling See also Schaerer+96, Schaerer & de Koter 96 Groh+ 2014 merge T and density structures from stellar evolution and atmospheric codes Temperature (K) 10 7 10 6 10 5 (a) O I model stage 3 10 5 11.0 11.5 12.0 12.5 + post-processing on a large grid of models + flexibility to study the effects of mass loss + effects of line blanketing, wind clumping 10 4 10 0 Geneva CMFGEN final CMFGEN 1 iter. (b) O I model stage 3 10-7 10-8 Density (g/cm^3) 10-5 10-10 10-9 10-10 10-11 10-12 11.0 11.5 12.0 12.5 10-15 1 10 100 1000 Radius (Rsun) (Groh+ 14)

Unified stellar evolution and atmospheric modeling See also Schaerer+96, Schaerer & de Koter 96 Groh+ 2014 merge T and density structures from stellar evolution and atmospheric codes + post-processing on a large grid of models + flexibility to study the effects of mass loss + effects of line blanketing, wind clumping Temperature (K) Temperature (K) Density (g/cm^3) Density (g/cm^3) 10 7 10 8 10 7 10 6 10 5 10-5 10 6 10 5 10 4 10 4 10 0 10 0 10-5 10-10 Geneva CMFGEN final CMFGEN 1 iter. (a)(c) O I WC model model stage stage 3 48 10 5 11.0 11.5 1.25 12.0 1.30 12.5 1.35 (b)(d) O I WC model model stage stage 3 48 10-7 10-8 10-9 10-10 10-11 10-12 10 6 10 5 10-5 10-6 10-7 10-8 10-9 10-10 11.0 1.0 11.5 1.2 12.0 1.4 1.6 12.5 1.8 2.0 10-15 10-10 1 0.1 10 1.0 100 10.0 1000 100.0 Radius (Rsun) (Groh+ 14)

Unified stellar evolution and atmospheric modeling See also Schaerer+96, Schaerer & de Koter 96 Groh+ 2014 merge T and density structures from stellar evolution and atmospheric codes + post-processing on a large grid of models + flexibility to study the effects of mass loss + effects of line blanketing, wind clumping - no feedback effects from the wind onto the envelope Temperature (K) Temperature (K) Density (g/cm^3) Density (g/cm^3) 10 7 10 8 10 7 10 6 10 5 10-5 10 6 10 5 10 4 10 4 10 0 10 0 10-5 10-10 Geneva CMFGEN final CMFGEN 1 iter. (a)(c) O I WC model model stage stage 3 48 10 5 11.0 11.5 1.25 12.0 1.30 12.5 1.35 (b)(d) O I WC model model stage stage 3 48 10-7 10-8 10-9 10-10 10-11 10-12 10 6 10 5 10-5 10-6 10-7 10-8 10-9 10-10 11.0 1.0 11.5 1.2 12.0 1.4 1.6 12.5 1.8 2.0 10-15 10-10 1 0.1 10 1.0 100 10.0 1000 100.0 Radius (Rsun) (Groh+ 14)

Unified stellar evolution and atmospheric modeling See also Schaerer+96, Schaerer & de Koter 96 Groh+ 2014 merge T and density structures from stellar evolution and atmospheric codes + post-processing on a large grid of models + flexibility to study the effects of mass loss + effects of line blanketing, wind clumping - no feedback effects from the wind onto the envelope Temperature (K) Temperature (K) Density (g/cm^3) Density (g/cm^3) 10 7 10 8 10 7 10 6 10 5 10-5 10 6 10 5 10 4 10 4 10 0 10 0 10-5 10-10 Geneva CMFGEN final CMFGEN 1 iter. (a)(c) O I WC model model stage stage 3 48 10 5 11.0 11.5 1.25 12.0 1.30 12.5 1.35 (b)(d) O I WC model model stage stage 3 48 10-7 10-8 10-9 10-10 10-11 10-12 10 6 10 5 10-5 10-6 10-7 10-8 10-9 10-10 11.0 1.0 11.5 1.2 12.0 1.4 1.6 12.5 1.8 2.0 10-15 10-10 Error of computing spectra assuming Teff(stellar evolution): 200-500K (ZAMS) 1500-4000 K (O supergiants) 10000s K (WR stars and LBVs) 1 0.1 10 1.0 100 10.0 1000 100.0 Radius (Rsun) (Groh+ 14)

Metallicity effects on stellar spectra: unified models Normalized Flux 1.5 1.0 0.5 N IV Hdelta He II Hgamma mid-h burning non-rotating 60 M CNO lines He II C III/ NIII He II at Z=0.014 (O4 I; Teff= 43kK) 0.0 4000 4200 4400 4600 4800 Wavelength [Angstrom] (Groh+ 15 in prep)

Metallicity effects on stellar spectra: unified models Normalized Flux 1.5 1.0 0.5 N IV Hdelta He II Hgamma mid-h burning non-rotating 60 M CNO lines at Z=0.014 (O4 I; Teff= 43kK) mid-h burning non-rotating 60 M at Z=0.0004 (O3 V; Teff=47kK) 0.0 4000 4200 4400 4600 4800 Wavelength [Angstrom] He II C III/ NIII He II (Groh+ 15 in prep)

Metallicity effects on stellar spectra: unified models He II 4686 line is weaker at low Z = Mdot effect (20x lower) Spectral lines are less numerous (CNO) = abundance effect (50x lower) Normalized Flux 1.5 1.0 0.5 N IV Hdelta He II Hgamma mid-h burning non-rotating 60 M CNO lines at Z=0.014 (O4 I; Teff= 43kK) mid-h burning non-rotating 60 M at Z=0.0004 (O3 V; Teff=47kK) 0.0 4000 4200 4400 4600 4800 Wavelength [Angstrom] He II C III/ NIII He II (Groh+ 15 in prep)

Metallicity effects on stellar spectra: unified models 2.0 Normalized Flux 1.5 1.0 0.5 N V O IV O V Fe V forest C IV He II 0.0 mid-h burning non-rotating 60 M at Z=0.014 (O4 I; Teff= 43kK) 1200 1300 1400 1500 1600 1700 Wavelength [Angstrom] (Groh+ 15 in prep)

Metallicity effects on stellar spectra: unified models Normalized Flux 2.0 1.5 1.0 0.5 mid-h burning non-rotating 60 M at Z=0.0004 (O3 V; Teff=47kK) N V O IV O V Fe V forest C IV He II 0.0 mid-h burning non-rotating 60 M at Z=0.014 (O4 I; Teff= 43kK) 1200 1300 1400 1500 1600 1700 Wavelength [Angstrom] (Groh+ 15 in prep)

Metallicity effects on stellar spectra: unified models He II lines are weaker at low Z = Mdot effect (20x lower) Spectral lines are less numerous (Fe, CNO) = abundance effect (50x lower) Normalized Flux 2.0 1.5 1.0 0.5 mid-h burning non-rotating 60 M at Z=0.0004 (O3 V; Teff=47kK) N V O IV O V Fe V forest C IV He II 0.0 mid-h burning non-rotating 60 M at Z=0.014 (O4 I; Teff= 43kK) 1200 1300 1400 1500 1600 1700 Wavelength [Angstrom] (Groh+ 15 in prep)

Metallicity effects on stellar spectra: ionizing fluxes 10-5 Flux (erg/s/cm^2/angstrom] 10-10 10-15 10-20 10-25 mid-h burning non-rotating 60 M at Z=0.014 (O4 I; Teff= 43kK) 100 1000 10000 Wavelength [Angstrom] (Groh+ 15 in prep)

Metallicity effects on stellar spectra: ionizing fluxes increase in ionizing fluxes (abundance/line blanketing effect) Flux (erg/s/cm^2/angstrom] 10-5 10-10 10-15 10-20 10-25 mid-h burning non-rotating 60 M at Z=0.0004 (O3 V; Teff=47kK) mid-h burning non-rotating 60 M at Z=0.014 (O4 I; Teff= 43kK) 100 1000 10000 Wavelength [Angstrom] (Groh+ 15 in prep)

Scary things: how do they change with metallicity? Mass loss and eruptions Wind structure (clumping) Blackbody vs model atmosp. Binarity Magnetic Fields Rotation Convection Eddington limit and inflation

Wind clumping affects line profiles Teff=52kK Z=0.002 Mdot 2.5e-6 Msun/yr MPG 355 (O2 III) in NGC 346 (Bouret et al. 2003)

Wind clumping affects line profiles Teff=40kK Z=0.002 Mdot 1e-7 Msun/yr MPG 368 (O4-5 V) in NGC 346 (Bouret et al. 2003) atmospheric models with clumping not included in Stellar Pop. Synthesis

Clumping and Mdot effects? Starburst99)2014.Models,.with.and.w/o.Rota;on. Z=0.15.Z! Z=0.15.Z!. Z=0.60.Z!.. (courtesy C. Steidel)

Blackbody vs. Model atmospheres (or why I don t like blackbodies) Flux (erg/s/cm^2/angstrom] 10-6 10-8 10-10 10-12 10-14 10-16 10-18 10-20 Black=CMFGEN model, Teff=47kK, Z=0.0004 Red=Blackbody T=47kK Red=Blackbody T=60kK 100 1000 10000 Wavelength [Angstrom]