Very Massive Stars (VMS):Properties, Evolution & Fates

Raphael HIRSCHISHYNE @ Keele: I. Walkington, C. Ritter, J. den Hartogh, A. Cristini, L. Scott

in collaboration with:GVA code: G. Meynet, A. Maeder, C. Georgy, S. Ekström, P. Eggenberger and C. Chiappini (IAP, D)VMS: N. Yusof, H. Kassim (UM, KL, Malaysia), P. Crowther (Sheffield), O. Schnurr (IAP)Nucleo: F.-K. Thielemann, U. Frischknecht, T. Rauscher (Basel, CH/Herts, UK) N. NishimuraNUGRID: F. Herwig (Victoria, Canada), M. Pignatari (Hull), C. Fryer, S. Jones (LANL), Laird (York), UChicago, UFrankfurt, …MESA: B. Paxton (KITP), F. X. Timmes, (UArizona, US)SNe: K. Nomoto (IPMU, J), C. Frohlich, M. Gilmer (NCSU), A. Kozyreva (Tel Aviv,Il), T. Fischer (W.,P)HYDRO: C. Meakin, D. Arnett (UArizona), C. Georgy (GVA), M. Viallet (MPA), F. Roepke (HITS, D), P. Edelmann (Newcastle, UK)

Armagh, Aug 17

Plan

- VMS (M > 100 M⊙) properties

- Key processes affecting the evolution of VMS:

mass loss, rotation & binarity

- Evolution and fate of VMS

- Key uncertainties

- Conclusions & outlook

Mass Determination Best with binary system

but often not available

Using model atmosphere and

compare them to observations:

Crowther et al 10, MNRAS

NGC 3603Cluster in our galaxy with about 104 M

o so we can expect 1-2 stars

more massive than 150 Mo.

Results:age: 1.5+/-0.1 Myr

Initial masses:B: 160 +/-20 M

o

A1a: 148 +40-27 Mo

C: 137 +/-17 Mo

A1b: 106 +/-23 Mo

Checks: masses of A1a & A1b consistent with dyn. masses; X-ray data for bin.

Crowther et al 10, MNRAS

R136Cluster in the LMC with about 5x104 M

o so we expect a few stars

more massive than 150 Mo.

Results:age: 1.7+/-0.2 Myr

Initial masses:a1: 320 +100-40 M

o

a2: 240 +/-45 Mo

c: 220 +55-45 Mo

a3: 165 +/-30 Mo

Checks: clumped mass loss rates derived: 2-5x10-5 Mo/yr match Vink et al predictions

Crowther et al 10, MNRAS

Very Massive Stars are Very Luminous (~107 L⊙)

R136a1 (107L⊙) alone supplies 7% of the ionizing flux of the entire

30 Doradus region!

What is the shape of the

luminosity vs mass relation

in this mass range?

Textbooks: L ~ M3 for starsin the solar mass range

Above 100 Mo: L~M1.5

Yusof et al 13 MNRAS, aph1305.2099

Mass Loss: Types, Driving & RecipesMass loss driving mechanism and prescriptions for different stages:

● O-type & “LBV” stars (bi-stab.): line-driven Vink et al 2000, 2001

● WR stars (clumping effect): line-driven Nugis & Lamers 2000, Gräfener & Hamann

(2008)

● RSG: Pulsation/dust? de Jager et al 1988

● RG: Pulsation/dust? Reimers 1975,78, with =~0.5

● AGB: Super winds? Dust Bloecker et al 1995, with =~0.05

● LBV eruptions: continuous driven winds? Owocki et al

● ...

What changes at low Z?

Z(LMC)~Zo/2.3 => Mdot/1.5 – Mdot/2

Z(SMC)~Zo/7 => Mdot/2.6 - Mdot/5

● Stars are more compact: R~R(Zo)/4 (lower opacities) at Z=10-8

● Rotation at low Z: stronger shear, weaker mer. circ.

● Mass loss weaker at low Z: → faster rotation

M (Z )=M (Z o)(Z /Z o)α - α = 0.5-0.6 (Kudritzki & Puls 00, Ku02)

(Nugis & Lamers, Evans et al 05)

- α = 0.7-0.86 (Vink et al 00,01,05)

Mass loss at low Z still possible?

RSG (and LBV?): no Z-dep.; CNO? (Van Loon 05, Owocky et al)

Mechanical mass loss ← critical rotation/ Eddington limit

(e.g. Hirschi 2007, Ekstroem et al 2008, Yoon et al 2012)

Binarity

Aug 10, 2017 Raphael Hirschi University of Keele (UK)

Population synthesis for VMS Schneider et al 2014ApJ...780..117S

The Evolution of VMS

VMS: much larger convective core & mass loss!

20 M⊙

300 M⊙

H

He

VMS = Very Massive Stars for M > 100 M⊙

H

He

Mr

Log10(Time left until collapse)Age [Myr]

(Yusof et al 13 MNRAS, aph1305.2099)

The Evolution of VMS

Rotating VMS have even larger convective core and (usually) mass loss

H

He

H

He

ROTNO ROT

(Yusof et al MNRAS 2013)

Age [Myr] Age [Myr]

Evolution of VMS across HRD: role of rotation/MdotLanger et al 07 (see also Yusof et al 2013)

Fast rotation → stars stay hot. Slow rotation → stars become cool.Different mass loss driving, Z dep.?

12 Raphael Hirschi Keele University (UK)

The fate of VMS: PCSN/BH/CCSN?

PCSN rangefrom Heger &Woosley(2002)

Zsolar

: no PCSN

(Rotating) models with Z<Z(LMC) lose less mass,

and enter the PCSN instability region!

BUT mass loss uncertain!

MCO

Mini

13 Raphael Hirschi Keele University (UK)

Consistent with Langer et al (2007): PCSN for Z<Z⊙

/3

(Yusof et al 13 MNRAS, aph1305.2099)

The fate of VMS: SNII/SNIb-c?

SN type:

- NO SNIIn predicted!~ NOT ok for SN2006gy(e.g. Woosley et al 2007)

- SNIc at solar Z,

- SNIb/c at Z(SMC)~ ok for SN2007bi(Gal-Yam 2009) BUT see Dessart et al 12,13+ Panstarrs resultsJerkstrand et al 16

(Yusof et al 13 MNRAS, aph1305.2099)

NEW FLASH PCSN SIMULATIONS OF GENEC MODELS UNDERWAY!!Kozyreva et al 2017, Gilmer et al 2017, talk by Kozyreva

First Generations: Fate of Non-Rotating StarsZ~0:

M<25 Mo: SNII

25-40: weak SNII

40-140:BH, no SN

140-260: PCSN (=PISN)

260-?: BH, no SN

Heger, Fryer et al 2003

Pair Creation SN (M:140-260 Mo) (Heger and Woosley 02, Scannapieco et al 05): - Chemical signature of PCSN not observed in EMP stars

(Umeda and Nomoto 02,03,05, Chieffi and Limongi 00,02,04)

- Due to strong mass loss? Hirschi 2007, Ekström et al 2008- 2006gy might be a PCSN (Smith et al 07, Langer et al 07, Heger et al 07)

15 Raphael Hirschi Keele University (UK)

PCSN Model Grid at Z=0.001 (Kozyreva+RH+ 2017MNRAS.464.2854K, Gilmer+RH+ ApJ accepted, ArXiV170607454G)

- New GENEC progenitor models at Z=0.001 (non-rotating): - M

ini=150,175, 200, 250 M

⊙

- Exploded with FLASH in 1D, 2D and 3D + Light curves with STELLASee talk by Kozyreva

Pre-SN: H-rich, extended envelope (1267R⊙

) H-poor, compact env. (2.4R⊙

)

The fate of VMS @ Z=0Yoon et al 2012 (see also Chatzopoulos, & Wheeler 2012 and Heger &

Woosley 2012, Heger & Woosley 2002)

Z= 0:Models including Mdot, rotation & B-fields

Rotation lowers mass range for PISN

Mechanical Mdot important

Key Open Questions Concerning Mass Loss

● Mass loss in cool parts of HRD: LBV & RSG, especially at low Z

● Position in & evolution across HRD: effects of rotation-induced mixing, feedback from mass loss Yusof et al 13, Langer 07, Sanyal et al 15, Kohler et al 15...

● Mass loss near Eddington limit Graefener & Hamann 08, Vink et al 11, ...

● Importance of clumping, porosity, inflation Fullerton et al 06, Graefener et al. 12, Vink et al, ...

● Which stars may explode in the LBV phase? Smith et al 11, ... ,Vink et al, ...

● Look of WR stars: radius, spectra Graefener et al. 2012, Groh et al 2013-...

● Additional mass loss mechanisms? Critical rotation at low Z? Shell mergers in late phases of evolution? ... Hirschi 2007, Meynet et al 2006, … , Smith & Arnett 2014, ...

● ...

Evolution of Eddington FactorΓ

Edd < 1 but Γ

Edd close to 1 if mass loss is low

Yusof et al 13 MNRAS, aph1305.2099

ΓEdd

may be larger than one below surface, see Sanyal et al. (2015).

Mass Loss near the Eddington LimitVink et al A&A 531, A132 (2011)

Envelope Inflation - VMS may be very extended after MS. This sometimes leads to a

density inversion in outer layer:

Unstable numerically → use of density scale height (black curve) stabilizes models with

modest impact on radius

Conclusions

22 Raphael Hirschi Keele University (UK)

- Very massive stars found in NGC3603 & R136: M up to 320 Mo!!

- PCSNe may occur for Z< ZLMC

- Nearby PCSNe are predicted to be SNIb/c, not SNIIn! (SN2007bi ok; SN2006gy X)

- Possible SNII PCSNe near Z=0

but it is easy to lose H-envelope in VMS

- Major uncertainties: mass loss (LBV, RSG)

- Other ingredients still uncertain: convection, rotation, mass

loss, B-fields, binarity + their interplay.

- Many key open questions → A lot of work ahead of us

Lots of other interesting recent work - Massive stars and the (not always) weak s process:

Large grid of massive star models + weak s proc (Frischknecht+2016, MNRAS):

Nugrid: set 1 (Pignatari+2016, ApJ), set1extension (Ritter+in prep),

(main) s process with new convective boundary mixing (CBM): (Battino+ ApJ 2016)

- Nuclear uncertainties: MC-based sensitivity studies for gamma-process (Rauscher+2016,

MNRAS), weak s process (Nishimura+2017, MNRAS), main s process (Cescutti+in prep)

- Stellar uncertainties:

Multi-D tests of convection (Cristini+ 2017, MNRAS) and rotation (Edelmann+2017, A&A)

- Reviews/book chapters: Springer Handbook of Supernovae

“Pre-supernova Evolution and Nucleosynthesis in Massive Stars and Their Stellar Wind Contribution” (doi:10.1007/978-3-319-20794-0_82-1)

“Very Massive and Supermassive Stars: Evolution and Fate” (doi:10.1007/978-3-319-20794-0_120-1)

- ChETEC COST Action started in April 2017: see www.chetec.eu for details

ChETEC COST Action (2017-2021)

29 countries have already joined ChETEC to coordinate research efforts in Nuclear

Astrophysics: Austria, Belgium, Bulgaria, Croatia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary,

Ireland, Israel, Italy, Lithuania, Malta, The Netherlands, Norway, Poland, Portugal, Romania, Serbia, Slovakia, Slovenia, Spain,

Sweden, Switzerland, Turkey and United Kingdom

www.chetec.eu

ChETEC Objectives What is ChETEC about? (pronounced [ketek])

Main challenge: tackle key open questions and link European facilities.

www.chetec.eu

Working Groups (WG) & Management Structure (MC) - WG1: nuclear data for astrophysics: needs, coordination and dissemination

- WG2: modelling pipelines connecting nuclear processes to astronomical observables

- WG3: astronomical data coordination, analysis and interpretation

- WG4: tools, techniques, knowledge exchange and innovation

Management Committee (MC): 2 members per country (+2-3 substitutes)

CORE group/Steering Committee (each CORE group member represents a team, see “Key Info” for more details)

Action Chair: R. Hirschi

Vice Chair: M. Lugaro

WG leaders: Alessandra Guglielmetti (WG1), Georges Meynet (WG2), Andreas Korn (WG3), Daniel Bemmerer (WG4)

Gender coordinator: Maria Lugaro

Pan-European coordinator: Sevdalina Dimitrova

Inter-sectoral (bi-direction Knowledge Transfer) coordinator: Daniel Bemmerer

STSM manager: Neven Soic

Dissemination coordinator: Jordi Jose

How to Get Involved?COST Actions are open and inclusive

Everyone can participate … but budget is limited given scale of network

(Most countries already have management committee members)

1) Join coordination effort at WG level or Action level

2) Sign up to ChETEC mailing list (to be set up soon)

3) Contribute to the “knowledge hubs”: including at least one directory of

datasets per WG

4) “Young” scientists are encouraged to attend the training schools

5) Propose, co-organise COST events

http://www.cost.eu/participate/join_action

www.chetec.eu

Activities Planned in 2017-2018 (Year 1)

1) Training schools:

- An experiment of Nuclear Physics for Astrophysics using direct methods

(main contact: Livius Trache): April 2018 @ IFIN-HH (ELI-NP), Bucharest, Romania

- R-matrix calculations for nuclear astrophysics

(main contact: Fairouz Hammache): 13-15 September 2017 @ IPN, Orsay, France

2) Main Action workshop involving all WGs: October 9-11, Keele

University, UK (main contact R. Hirschi)

www.chetec.eu

COST Acknowledgements www.chetec.eu

The ChETEC Action (CA16117) is supported by COST (www.cost.eu). COST (European Cooperation in Science and Technology) is a funding agency for research and innovation networks. Our Actions help connect research initiatives across Europe and enable scientists to grow their ideas by sharing them with their peers. This boosts their research, career and innovation.

Funded by the Horizon 2020 Framework Programme of the European Union

Key Open Questions Concerning Rotation

● Uncertainties in strength of rotation-induced mixing Hunter et al 07/08, Maeder et al 07, …

● Importance/impact of diff. prescriptions & their implementations (advective vs diffusive) Meynet et al LNP, 13, Meynet/Maeder et al ..., Chieffi & Limongi et al 13, Heger et al 2000, Paxton et al 13 (MESA), Martins & Palacios, 13

● Interaction between magnetic fields and rotation: Solid body rotation? More or less mixing? Spruit 02, Heger et al 05-..., Yoon et al 06-... Maeder et al 2005-..., Potter et al 12, ...

● Impact of binary interactions on distribution of rotation velocities Talk by Norbert Langer, de Mink et al 2013, ...

● Additional transport mechanism for Ω needed ←asteroseismology Cantiello et al. 14, Eggenberger 15; Spada et al. 16, Eggenberger et al 16 in prep

● ...