a deficit of ultraluminous x-ray sources in luminous ...• majority found in spiral galaxies. •...

A deficit of ultraluminous X-ray sources in luminous infrared galaxies

Wasutep Luangtip Durham University

with Tim Roberts (Durham)

Stefano Mineo (CfA and Durham)

Bret Lehmer (Johns Hopkins)

David Alexander (Durham)

et al…

Ultraluminous X-ray Sources (ULXs)

• Extragalactic X-ray sources with Lx > 1039 erg s-1 • Non nuclear point-like sources; cannot be SMBHs

30″ 30″ X-ray Optical

NGC 1365

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Possibilities for ULXs

• Intermediate mass black holes that accrete material at sub-Eddington rates

• Stellar mass black holes that accrete material at super-Eddington state

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

ULX population

• Generally found in all types of galaxy.

• Majority found in spiral galaxies.

• Colliding or interacting galaxies have largest individual populations.

• Tends to be found in star-forming regions.

Gao et al. (2003)

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

ULX population

Swartz et al. (2004)

Elliptical

Spiral

Expectation that one will detect many ULXs in high star formation rate galaxies such as U/LIRGs.

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Linden et al. (2010) Prestwich et al. (2013)

Effect of metallicity

ULX formation is more efficient in low metallicity galaxies.

Frac

tion

of H

MXB

s

High Z

low Z ULXs

Work outline

ULXs in 17 LIRGs (SFR > 7 M yr-1 , Z ~ Z)

Population study

X-ray luminosity functions

Number of ULXs per unit luminosity

Spectral analysis

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Detected ULXs • From 17 LIRGs with NH ≤ 5×1020 cm-2

and 14 Mpc ≤ DL ≤ 60 Mpc • 53 ULXs were detected • <9 sources might be background contamination. • Statistically complete sample (<10% of ULXs are missing).

Arp 299

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Detected ULXs • From 17 LIRGs with NH ≤ 5×1020 cm-2

and 14 Mpc ≤ DL ≤ 60 Mpc • 53 ULXs were detected • <9 sources might be background contamination. • Statistically complete sample (<10% of ULXs are missing).

Arp 299

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Swartz et al. (2011) sample has 2 ULXs per unit SFR. ~500 ULXs should be detected in the LIRG sample

but only ~1/10 detected!

X-ray luminosity function

Construct and fit with pure power-law model (dN/dL) /SFR = A x L39

γ

Slope = -1.75 ± 0.25 A = 0.11 ± 0.02

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Differential form

X-ray luminosity function

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Compare with literatures •Slope is consistent with that of HMXBs. • But normalisation is lower. • Support a deficit of ULXs in the sample.

Differential XLF

X-ray luminosity function N

(>L x)

Lx (1039 erg s-1)

Pure power-law

Slope = -1.2 ± 0.1 Stat. / d.o.f. = 16.4 / 48

Construct and fit with pure power-law model

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Cumulative XLF

X-ray luminosity function N

(>L x)

Lx (1039 erg s-1)

Broken power-law

Slope 1 = -0.9 ± 0.2 Slope 2 = -2.0 ± 0.5 Break Lx = 2.3 ± 0.5 keV Stat. / d.o.f. = 1.2 / 46

Might be difficult to form massive stellar

black holes in regions of ~ solar metallicity.

Eddington limit of ~20M BHs

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Cumulative XLF

Metallicity effect?

• Prestwich et al. (2013) show that the number of ULXs formed in high metallicity environments is only suppressed by a factor ~2 compared to lower metallicity regions.

• This effect alone cannot explain the deficit of ULXs by factor of ~10.

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Stellar population is too young?

Linden et al. (2010)

Need 5-10 Myr for ULX population to form

No.

of U

LXs

(per

106 M

)

• Bracket γ and FIR correlation showed that the star formation age of LIRGs is > 10 Myr (Goldader et al. 1996). • SED estimates of the star formation age of U/LIRGs between 5 -100 Myr (Vega et al. 2008).

it seems unlikely that the LIRGs are too young switch on ULXs.

Number of ULX per unit luminosity

Nea

rby

field

sp

iral

mer

ging

LI

RG (a

ll)

LIRG

(AG

N)

LIRG

(no

AGN

) NU

LX /

L FIR(

10-4

7 (er

g s-1

)-1)

spira

l m

ergi

ng

LIRG

(all)

LIRG

(AG

N)

LIRG

(no

AGN

)

ULX no. per blue luminosity ULX no. per FIR luminosity

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs N

earb

y fie

ld N

ULX

/ L bl

ue (1

0-46 (

erg

s-1)-1

)

High level of dust in LIRGs • Suppress the blue light of the galaxies via

absorption/extinction high number of ULX per unit blue luminosity

• Also substantially increasing the FIR luminosity. low number of ULX per unit FIR luminosity

• Obscure most of ULXs from our view. • Support by optical surveys of core-collapse

supernovae: 17% are detected! (Horiuchi et al. 2011; Mattila et al. 2012)

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Spectral analysis

53 ULXs

Low luminosity bin (39≤logLx≤39.33)

25 ULXs

Medium luminosity bin (39.33≤logLx≤39.67)

21 ULXs

High luminosity bin (39.67≤logLx≤40)

6 ULXs

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Spectral analysis EF

E

(keV

2 Pho

ton

cm-2

s-1 k

eV-1

)

10 1 10-6

10-5

Energy (keV)

Low luminosity Medium luminosity High luminosity

Fit with absorbed power-law model

We are seeing a change in accretion

stage from disc-dominated to super-Eddington stage (see Sutton et al. 2013).

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs

Conclusion

• We see a large deficit of ULXs in LIRGs. • Metallicity and star formation age may have some

influence on ULX numbers. • The main deficit could be explained by the high

levels of dust in LIRGs, obscuring the bulk of their ULX population.

• A stacked spectra support the accretion state changing from ~Eddington rate disc-like spectra to super-Eddington ‘ultraluminous’ state above ~2x1039 erg s-1.

17 June 2014 W. Luangtip - A deficit of ULXs in LIRGs