STELLAR BLACK HOLES AT THE DAWN OF THE UNIVERSE

Félix Mirabel CEA-Saclay-France & CONICET-Argentina

A subject at the crossroads of massive stellar evolution, High Energy & black hole Astrophysics, & Cosmology

HOW WAS THE IGM RE-IONIZED?  

• WHAT COULD HAVE BEEN THE ROLE OF mQSOs, HMXBs?

Based on recent results from Stellar evolution and high energy astrophysics:

• I propose that in galaxies at z>10 a large fraction of Pop III-II stars end as Stellar BHs in HMXBs X-rays & jets

• X-rays & jets overtake the HII regions produced by UVs, heat and partially ionize the IGM over large volumes of space.

•The IGM was fully ionized by the UV from the first stars (Pop III & II) HII regions expanding at < 100 Km/s.

THE « SWISS CHEESE » MODEL for the re-ionization of the IGM:

Gunn-Peterson effect

Dark age

ASTROPHYSICAL GROUNDS FOR THE STELLAR BLACK HOLE HYPOTHESIS

THEORETICAL GROUNDS

• MOST POP III & II STARS WERE FORMED AS MULTIPLE SYSTEMS Turk+Science 2009; Krumholz+ Science 2009; Clark+ Science 2011; Stacy+…etc.

• STARS OF LOW Z WITH M > 20 M END AS BHs DIRECTLY Fryer,1999;Heger+2003;Georgy+2009;Woosley+2008;Nomoto+2010; Linden,Kalogera+2011

• NO ENERGETIC SNe STELLAR BHs REMAIN BOUND TO DONORS Mirabel & Rodrigues, Science 2003; Mirabel+ Nature 2008

OBSERVATIONAL GROUNDS

• MOST ULXs & LGRBs ARE HOSTED IN LOW Z-HIGH-SSFR GALAXIES Feng & Soria,2011;LeFloc’h,Duc,Mirabel;2003;Fruchter+ Nature, 2006; Perley+ 2014

• Lx/SFR EVOLUTION WITH z IS DRIVEN BY Z EVOLUTION IN HMXBs Fragos+2012; Basu-Zych+2012

• IN LOW Z GALAXIES Lx/SFR IS LARGER THAN IN MAIN-S GALAXIES Thuan+ 2004; Kaaret+ 2014; Douna, Pellizza, Mirabel, 2014 (in progress)

EVOLUTION OF MASSIVE STARSPOP III & II STARS WERE FORMED IN MULTIPLE SYSTEMS

Turk, Abel & O’Shea (Science 2009) Krumholz et al. (Science 2009)

Stacy, Greif & Bromm (ApJ 2010) Fragmentation: Clark+ (Science 2011)

THEORY

• Pop III stars were multiple systems dominated by binaries with 10-100 M OBSERVATIONS

• In the MW >70% of OB type stars are binaries (Chini+2011;Sana+ 2012-14)

DID A LARGE FRACTION OF Pop III & II STARS END AS BH-HMXBs ?

STARS OF LOW Z & M>20 M END AS BHs DIRECTLY

with no rotation (Heger+ 2003)

• Low metal progenitors form BHs by IMPLOSION (e.g. Fryer, 1999, Heger+2003…)

with rotation (Georgy+ 2009)

solar metallicity

Mass of progenitor star

Z

THE FORMATION OF HMXBs IS STRONGLY METALLICITY DEPENDENT (Linden, Kalogera+ 2011)

Number of HMXBs per starburst of 106 M

AT LOW Z • the number of HMXBs increases

• the orbital period decreases

• progenitors of > 20 M end as BHs

What observations could test these theoretical predictions?

THE MASSES OF BHs IN HMXBs ARE LARGER IN HOSTS OF LOW METALLICITY

BH masses determined dynamically (Crowther et al. 2010)

The stellar BHs in the low metallicity galaxies M33 X-7, NGC300 X-1, IC10 X-1 have MBH>15 M whereas in the Galaxy and M31 all known stellar BHs have MBH<14 M

If the mass lost in low metallicity progenitors is relativelly small, the collapsing cores should be more massive, leading to more massive compact remnants

BH-HMXBs

EVIDENCE FOR BH FORMATION WITH NO KICKS APPROACH BASED ON

KINEMATICS (Mirabel & Irapuan Rodrigues 2001-2009)

Mirabel-Rodríguez 1994

VLA

l3.6 cml3.6 cm

Mirabel + 1992

COMPACT JETS

STEADY JETS TRANSIENTJETS

Dhawan, Mirabel, Rodríguez (2007)

In low hard state compact jets have Sizes ~ 100 AU & Same PA as large scale jets.

With VLBI get sub-miliarc sec precision. Proper motions with VLBI of compact jets and/or optical images with HST

Credit S&T

RESULTS FOR FIVE BH-XRBs

Cygnus X-1: Vp< 9+/-2 km/s < 1 M ejected in a SN

also

GRS 1915+105 & V404 Cyg suggest that

BHs WITH > 10 M

FORM IN THE DARK

Mirabel & Rodrigues (Science, 2003)

XTE J1118+480; GRO J1655-40

WITH BHs OF < 8 M

are RUNAWAY BLACK HOLES

Mirabel et al. (Nature 2001)GALACTOCENTRIC ORBIT (230 Myrs)

Yellow: Sun White: binary BH

THE OCCURRENCE RATE OF ULXs PER UNIT GALAXY MASS INCREASES WITH THE SFR AND 1/Z e.g. Zampieri & Roberts (2009)

X-ray (Chandra) Antennae

Fabbiano et al.

ULTRALUMINOUS X-RAY SOURCES (ULXs)

ULXs IN LOW METALLICITY REGIONS OF GALAXIES

The integrated ULXs luminosity of ~1042 erg s-1 rivals that of AGN

(Gao+ 2003)

BH-HMXBs ARE MOSTLY FOUND IN STAR-FORMING GALAXIES OF LOW METALLICITY & ARE POWERFUL ENERGY SOURCES

Dwarf galaxies of small Z are the best local templates of primeval galaxies

Massive outflows

shock & photo-ionized bubbles of >100 pc in size

Ha images

E > 1053 erg

BH-HMXB

Shape of X-ray spectra are critical

Review by Feng & Soria, 2011

THE HOSTS OF LGRBs WITH OPTICAL AFTERGLOWS ARE SMALL IRREGULAR GALAXIES OF LOW Z

Le Floc´h,Duc,Mirabel with VLT (2003); Fruchter+ with HST (Nature 2006); Vergani+ (2014)

Levesque et al. (2010)

From 53000 SDD galaxies

From 940 line emission galaxies

However, dark LGRBs may pose some caveats & controversy

Graham & Fruchter, (2012)

FROM STUDIES OF HMXBs IN LOCAL UNIVERSE IT IS INFERRED

THE COSMIC EVOLUTION OF METALLICITY A COSMIC EVOLUTION OF BH-HMXBs

. At low metallicities (Z<Z-5) there should be an increase of:

• The mass of stellar BHs because the progenitor cores are more massive

• The number of BH-HMXBs since massive stars form BHs by direct collapse

• The X-ray luminosity of BH-HMXBs…an issue being investigated… Mirabel. Invited Review. Proceedings of IAU Symp. 275 (2011) (arXiV:1012.4944v1 [astro-ph.CO] 22 Dec 2010)

Ionizing power of mQSOs versus ionizing power of massive stars

Mirabel, Dijkstra, Laurent, Loeb, Pritchard (A&A 2011) N&V in Nature (2011) Counting photons

N,BH

N,*

0.6N phot

64000

1MBH

M*

fedd0.1

tacc20Myr

E

keV

1

fesc,*0.1

1fesc,BH1.0

,

fedd = fraction of Eddington luminosity for a time tacc

Nphot = number of ionizing photons emitted per atom of H nucleus

<E>g = mean photon energy emitted by the accreting BH

fesc,* (fesq,BH) = fraction of ionizing photons that escape

For fiducially values of the model parameters:AN ACCRETING STELLAR BLACK HOLE EMITS A NUMBER OF X-RAY PHOTONS COMPARABLE TO THE UV PHOTONS FROM ITS PROGENITOR STAR

• But in a fully neutral medium Nsec* = 25 (E g /1 keV), where Eg is the photon energy

However, not all stars will be massive and lead to the formation of BH-HMXBs…

HMXBs & HI TOMOGRAPHY DURING THE EoR

Gas Temperature

l21cm line Mirabel+ (2011)

Mirabel+ (2011)

Fialkov+ Nature, 2014

l21cm lineX-ray hard spectrum

fx at z > 6 must increase as Z < 10-5 Z

BH-HMXBs HEATED THE IGM TO ~104 K OVER LARGE VOLUMESAt z=8.9 the IGM was already heated and ionized (Paciga+ 2014)

CONCLUSIONX-RAYS HAVE LONGER MEAN FREE PATH THAN UVs. THEY HEAT THE GAS FAR FROM THE GALAXIES AND PARTIALLY IONIZE THE BULK OF THE IGM.

BH-HMXBs LIMITED THE MASS OF DWARF GALAXIES

Mmin ~ 109 ( /100r rc)-1/2 ( /0.6)m -3/2 [ ( )/10T K 4]3/2 [(1+z)/10]-3/2 M

rc = critical mass density for a flat universe, r = mass density in the galaxy

m = mean molecular weight, z = redshift, T = temperature of the IGM

• THE THERMAL HISTORY OF THE IGM DETERMINED BY STELLAR BLACK HOLES LIMITED THE MASS & NUMBER OF DWARF GALAXIES PREDICED BY THE lCDM

OPEN QUESTIONS• I) Will l21cm tomography of HI with LOFAR, SKA show a

smoother end to the dark ages?

• II) Could the X-rays from BH-HMXBs contribute to the 10-20% unresolved hard X-ray background?

• III) Do mQSO jets penetrate deeper in the IGM? (in progress)

• IV) Will BH-BH stellar binaries may be frequently detected as sources of gravitational waves? (Belczynski+2011;Ziosi+2014)

• V) Does feedback from stellar BHs reduce the lCDM predicted number of dwarf galaxies?

• VI) Are there naked dark matter haloes with M < 109 M?

• VII) Are HMXBs at z>6 be the source of the intriguing radio background discovered with ARCADE 2?

(Fixsen+2011; Seiffer+ 2011; Condon+ 2013)

OBSERVATIONS RELATED TO HIGH z HMXBs?

• COSMIC RADIO BACKGROUND WITH ARCADE 2

• He II IONIZATION AT z > 3.5 (Worseck+; Compostella+ 2014)

Absorption spectroscopy of the He II La line lrest=303.78°A A radiation field

with photons hn> 54.4 eV that can not be provided by luminous QSQs

Excess emission ARCADE2

discrete sources

2.725 K blackbody

VLA follow up by Condon+ (2012): The ARCADE 2 excess is smooth A new unexpected discrete-source population too numerous to be associated with bright galaxies

(Fixsen+; Seiffer+ 2011)

ARE THE RADIO BACKGROUND & PHOTONS WITH

hn>54.4 eV FROM A POPULATION OF HMXBs AT z>6 OR FROM SUPERMASSIVE BHs? (Biermann+2014)

CONCLUSION I) l21cm tomography of HI with LOFAR, SKA, etc., may reveal a smoother end to the dark ages

X-rays=0 large fluctuations. X-rays=50% lower fluctuations but at shorter ls

Slice of 170×170×0.66 Mpc3

(Visbal & Loeb, 2011; Fialkov+ Nature 2014)

Pacucci et al., 20014)

Hard X-ray spectra

Softer X-ray spectra

hard (right) and soft (left ) X-ray SED

HMXBs + soft X-rays Larger fluctuations

Fialkov+ Nature 2014

X-ray spectra of galaxies

Slice 1.5 Mpc at zpeak = 16

UVs vs UVs+X-rays. HMXBs with hard spectra

Prestwich+2013 using Nustar

NULX(SFR) for individual SINGS galaxies, NGC 922 and the

Cartwheel. This plot uses the KK04 metallicity calibration. (Prestwich+2013)

Figure 2. NULX(SFR) for individual SINGS galaxies, intermediate metallicity galaxies and the combined metal poor and XMPG. This plot uses the PT05 metallicitycalibration. . (Prestwich+2013)

Basu-Zych

NO ENERGETIC NATAL SNe stBHs REMAIN BOUND

CORE COLLAPSE MODELS:

Massive stellar black holes (M>10 M) should form with no energetic kicks

(Fryer & Kalogera; Woosley & Heger; Nomoto et al.)

THIS HYPOTHESIS IS BEEN TESTED USING THE KINEMATICS OF mQSOs

Mirabel et al. in Nature, Science & A&A (2001-2012 )

e.g. Mirabel & Rodrigues (Science, 2003)

Cygnus X-1 was form in the Dark

With no energetic kick Now confirmed by Gou, McClintock+ (2011) & Reid+ (2012)

STELLAR BHs ARE OBSERVED IN BINARIES Credit: S&T

PROPER MOTIONS USING COMPACT JETS Mirabel & Rodríguez 1994

VLA

l3.6 cml3.6 cm

Mirabel, Rodriguez+ 1992

COMPACT JETS

STEADY

JETS

TRANSIENT

JETS

Dhawan, Mirabel, Rodríguez (2007)

In low hard state. Size ~ 100 AU. Same PA

USED TO DETERMINE PROPER MOTIONS

(with VLBI to get sub-miliarc sec precision)

Boera+ 2014

Panucci

Panucci

OPEN QUESTIONS• LIFE TIME OF BLACK HOLE HMXBs• INCORPORATION OF NEUTRON STAR HMXBs • ACCRETION RATE-FEEDBACK AS A FUNCTION OF METALLICITY• PHYSICAL MECHANISMS OF COOLING, IONIZATION & HEATING• X-RAYS FROM SNe BY COMPARISON WITH X-RAYS FROM HMXBs • INCLUSION OF RELATIVISTIC JETS OF PARTICLES• BINARY stBHs THE FIRST SOURCES OF GRAVIATIONAL WAVES TO DETECT?• WILL THE l21cm SIGNALS FROM HI AT HIGH z (LOWFAR, SKA, EDGES), AS

PREDICTED INCLUDING HMXBs, HAVE LOWER AMPLITUDES, BE MORE UNIFORM RATHER THAN HII REGION DOMINATED WITH PATCHY “SWISS CHESS TOPOLOGY , AND THE REIONIZATION HAVE TAKEN LONGER TIME, SHIFTING THE SIGNALS TO HIGHER FREQUENCIES?

l21cm tomography (Visbal & Loeb, 2011)

X-rays = 0

large fluctuations

X-rays=50% lower fluctuations due to more uniform heatingSignals up to lower z’s

Gunn-Peterson effect EVIDENCE FOR THE DARK AGES

• No light (continuum radiation) is observed at wavelengths less than L a because it is fully absorbed by atomic hydrogen.

The Gunn–Peterson trough is a feature of the spectra of quasars due to the presence of neutral hydrogen in the IGM. The trough is characterized by suppression of electromagnetic emission from the quasar at wavelengths less than that of the La line at the redshift of the emitted light. This effect was originally predicted in 1965 by James E. Gunn and Bruce Peterson

STELLAR EVOLUTION & HIGH ENERGY ASTROPHYSICS:

CORE COLLAPSE MODELS:

Massive stellar black holes (M>10 M) should form with no energetic kicks

(Fryer & Kalogera; Woosley & Heger; Nomoto et al.)

THIS HYPOTESIS IS BEEN TESTED USING THE KINEMATICS OF mQSOs

Mirabel et al. in Nature, Science & A&A (2001-2012 )

e.g. Mirabel & Rodrigues (Science, 2003)

Cygnus X-1 was form in the Dark

With no energetic kick Now confirmed by Gou, McClintock+ (2011) & Reid+ (2012)

STELLAR BHs ARE OBSERVED IN BINARIES

DO StBHs RECEIVE NATAL SN KICKS?

CORE COLLAPSE MODELS:

Massive stellar black holes (M>10 M) should form with no energetic kicks

(Fryer & Kalogera; Woosley & Heger; Nomoto et al.)

• IS THERE ANY EVIDENCE OF StBH FORMATION BY DIRECT COLLAPSE?

• IS THERE A RELATION BETWEEN KICK VELOCITIES AND BH MASS?

THE CORE COLLAPSE MODELS CAN BE TESTED USING THE KINEMATICS OF BH mQSOs

Mirabel et al. (2001-2009 )

FORM WITH OR WITHOUT ENERGETIC NATAL SNe?

RESULTS FOR FIVE BH-XRBsCygnus X-1 Vp< 9+/-2 km/s < 1 M ejected in a SN together with

GRS 1915+105 & V404 Cyg suggest that

BHs WITH >10 M MAY FORM IN THE DARK

Mirabel & Rodrigues (Science, 2003)

XTE J1118+480; GRO J1655-40

BHs OF < 8 M RUNAWAY BLACK

HOLES

Mirabel et al.(Nature 2001)GALACTOCENTRIC ORBIT (230 Myrs)

Yellow: Sun White: binary BH

BLACK HOLES WITH > 10 M FORM BY IMPLOSION STELLAR BLACK HOLES REMAIN BOUND TO DONORS

BHs OF >10 M FORM BY DIRECT COLLAPSE

Cyg X-1 remained in its birth place

Vp < 9 +/- 2 km/s

< 1 M in SN

THE BH IN Cyg X-1 WAS FORMED IN THE DARK Mirabel & Rodrigues (Science, 2003) Gou, McClintock et al. (2012) confirmed this result

Other BHs with >10 M form by direct collapse:

GRS 1915:W=7+/-3 km/s (Dhawan,Mirabel,Rodríguez,01) V404 Cyg:W=0.2+/-3 km/s (Miller-Jones et al. 2009)

The BHs remain bound to donorsSky & Telescope

APPROACH BASED ON THE KINEMATICS (Mirabel & Irapuan Rodrigues 2001-2009)

DO LOCAL TEMPLATES OF HIGH z GALAXIES HAVE ENHANCED Lx/SFR?

Douna, Pellizza, Mirabel (2014, in progress)

• Lx/SFR EVOLUTION WITH z IS DRIVEN BY Z EVOLUTION IN HMXBs Based on large surveys: Fragos+2012; Basu-Zych+2012

• IN LOW Z GALAXIES Lx/SFR IS LARGER THAN IN MAIN-S GALAXIES Thuan+ 2004; Kaaret+ 2014; Douna, Pellizza, Mirabel (2014, in progress)