Black Hole Powered Nebulae and a Case Study of the Ultraluminous X-ray Source IC342 X-1

0.8 MB·English

by  D. Cseh

Accepted version
PreprinttypesetusingLATEXstyleemulateapjv.5/25/10
BLACK HOLE POWERED NEBULAE AND A CASE STUDY OF THE ULTRALUMINOUS X-RAY SOURCE
IC342 X-1
Da´vid Cseh1, St´ephane Corbel1, Philip Kaaret2, Cornelia Lang2, Fabien Gris´e2,
Zsolt Paragi3,4, Anastasios Tzioumis5, Valeriu Tudose6 and Hua Feng7
Accepted version
ABSTRACT
2 Wepresentnewradio,optical,andX-rayobservationsofthreeUltraluminousX-raysources(ULXs)
1 that are associated with large-scale nebulae. We report the discovery of a radio nebula associated
0
withtheULXIC342X-1usingtheVeryLargeArray(VLA).ComplementaryVLAobservationsofthe
2
nebulaaroundHolmbergIIX-1,andhigh-frequencyAustraliaTelescopeCompactArray(ATCA)and

n VeryLargeTelescope(VLT)spectroscopicobservationsofNGC5408X-1arealsopresented. Westudy
a the morphology, ionization processes, and the energetics of the optical/radio nebulae of IC342 X-1,
J HolmbergII X-1andNGC5408X-1. The energeticsofthe opticalnebula ofIC342X-1 isdiscussedin
1 the framework of standard bubble theory. The total energy content of the optical nebula is 6×1052
2 erg. The minimumenergyneededtosupply the associatedradionebula is 9.2×1050 erg. Inaddition,
wedetectedanunresolvedradiosourceatthelocationofIC342X-1atVLAscales. However,ourVery

] Long Baseline Interferometry (VLBI) observations using the European VLBI Network likely rule out
E
the presence of any compactradio source at milli-arcsecond(mas) scales. Using a simultaneous Swift
H X-ray Telescope measurement, we estimate an upper limit on the mass of the black hole in IC342
h. X-1 using the ”fundamental plane” of accreting black holes and obtain MBH ≤(1.0±0.3)×103 M⊙.
Arguing that the nebula of IC342 X-1 is possibly inflated by a jet, we estimate accretion rates and
p
efficiencies for the jet of IC342 X-1 and compare with sources like S26, SS433, IC10 X-1.
-
o Subject headings: black hole physics — accretion, accretion disks — ISM: bubbles, jets and outflows
r — X-rays: binaries
t
s
a
[ 1. INTRODUCTION tal intrinsic power of the ULX (Pakull & Mirioni 2002,

2003; Russell et al. 2011). In general, the nebu-
1 Ultraluminous X-ray sources are variable non-nuclear
lae around ULXs are either photoionized due to the
v X-ray sources in external galaxies with luminosities
high X-ray and UV luminosity of the compact object
3 greatly exceeding the Eddington luminosity of a stellar-
(Pakull & Mirioni 2002, 2003; Kaaret, Ward & Zezas
7 mass compact object, assuming isotropic emission
2004; Kaaret & Corbel 2009; Moon et al. 2011), or
4 (Colbert & Mushotzky1999;Kaaret et al.2001). Their-
4 regular variability, observed on time scales from seconds shock-ionized driven by jets, outflows, and/or disk
1. toyears,suggeststhatULXsarebinarysystemscontain- winds (Pakull & Mirioni 2002; Roberts et al. 2003;
Abolmasov et al.2007). Inseveralcases,two-component
0 ing a compact object that is either a stellar-mass black
optical line profiles are present, indicating a mix-
2 holewithbeamed(King et al.2001;K¨ording et al.2002)
ture of the two mechanisms or alternatively the nar-
1 or super-Eddington emission (Begelman 2002); or an
row line could be due to the shock precursor. An-
: intermediate-massblackhole(IMBH).IMBHshavebeen
v other common feature is the presence of the high-
invoked in contexts ranging from the remnants of Pop-
i ionization He ii emission line. It can have vari-
X ulation III (Madau & Rees 2001) stars to the formation
ous origins: the nebula, the donor, the accretion
of supermassive black holes (SMBHs) (Ptak & Griffiths
r disk, or a disk wind. The ionizing photon rate
a 1999);SMBHsmayformthroughthehierarchicalmerger
needed to produce the narrow high-excitation He ii
of lower mass black holes (eg. Ebisuzaki, et al. 2001).
line of the nebulae in photoionized sources indicates
Several ULXs show emission-line optical nebulae,
that their X-ray emission is at most mildly beamed
which can be used as a calorimeter to infer the to-
(Pakull & Mirioni 2002, 2003; Kaaret, Ward & Zezas
[email protected] 2004; Kaaret & Corbel 2009; Moon et al. 2011).
1Laboratoire Astrophysique des Interactions Multi-echelles Only a handful of radio detections of ULXs have
(UMR7158), CEA/DSM-CNRS-UniversiteParisDiderot,CEA
beenmadesofar,includingNGC5408X-1(Kaaret et al.
Saclay, F-91191GifsurYvette, France
2Department of Physics and Astronomy, University of Iowa, 2003; Soria et al. 2006a; Lang et al. 2007), Ho II X-1
VanAllenHall,IowaCity,IA52242,USA (Miller et al. 2005), and MF16 (van Dyk et al. 1994;
3Joint Institute for VLBI in Europe, Postbus 2, 7990 AA Lacey, Duric & Goss 1997), if not considered a super-
Dwingeloo,TheNetherlands
4MTAResearchGroupforPhysicalGeodesyandGeodynam- nova remnant (Matonick & Fesen 1997). Most of these
ics,POBox91,1521Budapest, Hungary sources show large nebulae (>∼50 pc) that are likely
5Australia Telescope National Facility, CSIRO, PO Box 76, powered by continuous energy input from the ULX.
Ep6pAinSgT,RNOSWN,1O71u0d,eAHusotorgaelivaeensedijk 4, 7991 PD Dwingeloo, Shock-dominatedonesareprobablypoweredinthesame
manner as the W50 nebula is powered by the Galactic
TheNetherlands
7Department of Engineering Physics and Center for Astro- binary SS433 (Dubner et al. 1998). However, the ULX
physics,TsinghuaUniversity,Beijing100084, China radio nebulae require greater total energy content than
2 Cseh et al.
W50. A similarly powerful nebula, S26, was found by weighted the radio image with a robust parameter of -2
Pakull, Soria & Motch (2010) in optical and Soria et al. in order to resolve out the diffuse nebular emission and
(2010) in radio. For other possible radio associations show the fine scale structure. The strongest radio emis-
with ULXs, we refer the reader to Sa´nchez-Sutil et al. sion appears towards the NE of the ULX and is coinci-
(2006); Soria et al. (2006b); Mezcua & Lobanov (2010). dentwiththestrongestHαemission. Theradioemission
The X-ray spectra of ULXs share some similar extends farther to the NE in regions of little or no Hα
properties with the canonical black hole states of emission.
Galactic black hole binaries (GBHBs). A num- Figure2 also revealsanunresolvedradiosourceat the
ber of ULXs show state transitions (Kubota et al. location of IC342 X-1. This unresolved source is de-
2001; Feng & Kaaret 2006, 2009; Godet et al. 2009; tectedwithafluxdensityof∼96.3µJyatthe6.5-σlevel.
Gris´e et al. 2010; Servillat et al. 2011). In GBHBs The corresponding luminosity is 8.8×1033 erg s−1. The
during the X-ray hard state, the sources are associ- HST position of the ULX is RA=03h45m55.61s, Decl.=
atedwithself-absorbedcompactjets(Corbel et al.2004; +68◦04’55.3”(J2000.0)withthe90%positionalerrorsof
Fender, Belloni & Gallo2004). Giventhesimilaritiesbe- 0.2 arcsec (Feng & Kaaret 2008). We obtained the po-
tween ULXs and GBHBs, it is interesting to investigate sition of the radio peak of RA=03h45m55.54s, Decl.=
the presence of such compact jets for ULXs with hard +68◦04’55.18”using the maxfit task of AIPS. The dis-
X-ray spectra. tancebetweenthetwopositionsis∼0.4arcsec. Usingthe
In this paper, we present new radio and optical obser- AIPS task jmfit, we estimate a radio positional uncer-
vations of two ULXs that are associatedwith large-scale taintyof∼0.1arcsecatonesigma. Thereforetheoptical
nebulae: HolmbergII X-1(HoII X-1)andNGC5408. In and radio positions are consistent using 90% positional
addition, we present discovery of a radio nebula associ- errors.
ated with IC342 X-1. In Section 2, we describe obser-
2.1.2. Multi-frequency radio observations of Ho II X-1
vations, data analysis and results. The energetics of the
opticalandradionebulaeandthejetpropertiesofIC342 The radio nebula of Ho II X-1 was first detected at
X-1areinvestigatedinSection3. InSection4,webriefly 1.4 and 5 GHz by Miller et al. (2005). Here, we have
summarize our results. conducted observations of Ho II X-1 at 5 and 8 GHz
in order to constrain the shape and spectrum of the
2. OBSERVATIONS, ANALYSIS, AND RESULTS radio nebula. Figure 3. shows the VLA images of
Ho II X-1 overlaid on HST He ii and Hβ images from
2.1. VLA observations of IC342 X-1 and Ho II X-1
Kaaret, Ward & Zezas (2004). On the left, the un-
Observations of IC342 X-1 and Ho II X-1 were car- weighted (Robust=0) 5 GHz B-array image is shown.
ried out using the C- and B-array configurations of the The right panel shows a Robust=−1 image made at
Very Large Array (VLA) of the National Radio Astron- 8 GHz using the C-arrayconfiguration. The asymmetric
omy Observatory(NRAO). The observationswere made morphology of the nebula might indicate some outflows
at 4.8 and 8.5 GHz (VLA program code: AL711) and or ambient density gradient to the West.
the details are summarizedin Table 1. Data calibration, Previously, the radio spectrum was not well con-
combination in the (u,v) plane and imaging were car- strained; Miller et al. (2005) had high uncertainties on
ried out using the NRAO Astronomical Image Process- the flux at 5 GHz of Ho II X-1. However, we now have
ing System (AIPS: e.g., Greisen (2003)). We adjusted flux density measurements at three frequencies (1.4, 8
the Robust weighting parameter between 0 and -2 to and 5 GHz, using the image from Miller et al. (2005) at
bring out the fine-scale radio emission and parameters 1.4 GHz) and Figure 4. illustrates the fitted three-point
are identified in image captions. spectral index of α=−0.53±0.07,(S ∼να). The radio
spectrum is consistent with optically thin, synchrotron
2.1.1. Radio detection of IC342 X-1
emission, further constraining the nebular nature of the
Figure 1 shows the 5 GHz VLA B- and C-array com- radio counterpart of this ULX nebula. The integrated
binedimageofIC342X-1withRobust=0weightingover- fluxdensity ofthe nebulais 1.05±0.10mJy,613±61µJy
laidontheHαHSTimageofFeng & Kaaret(2008). Ex- and 395±40 µJy at 1.4, 5 and 8 GHz.
tendedradioemissionispresentsurroundingtheposition
2.2. Swift/XRT and EVN observations of IC342 X-1
of IC342 X-1. Diffuse emission is detected at up to the
10-σ level with peak intensity of ∼120 µJy beam−1 and 2.2.1. VLBI observation
theestimatedtotalfluxdensityinthenebulais∼2mJy. We conducted simultaneous X-ray and radio measure-
The corresponding luminosity at a distance of 3.9 Mpc ments of the VLA core (see Sec. 2.2.1.) to test its
(Tikhonov & Galazutdinova 2010) is Lneb = 1.8×1035 compactness in order to estimate the mass of the BH
erg s−1; with L=νL . We find that the size ofthe neb- (see Sec. 3.4) with the minimum uncertainty. To check
ν
ula is 16” × 8”, which corresponds to 300 pc × 150 pc. whether the VLA core is consistent with a compact jet
Weestimatethevolumeofthenebulabytakingasphere at 10 mas scale, we carried out European VLBI Net-
withadiameterof12”. Comparingthe5GHzradiomap work (EVN) observations (EVN program code: EC032)
to the Hα image, the optical and the radio nebulae are at 1.6 GHz. The 12-hour observations were accommo-
bothelongatedintheNE–SWdirection,possiblyexhibit- datedonJune 15,2011from03:30to 15:30(UT), simul-
ing a shock-front. The size of the radio nebula is consis- taneouslywiththeSwift/XRTobservations. Thepartici-
tentwiththesizeoftheopticalnebula(Pakull & Mirioni patingVLBIstationswereEffelsberg(Germany),Jodrell
2002; Gris´e et al. 2006; Feng & Kaaret 2008). Bank Lovell Telescope, Cambridge (United Kingdom),
Figure 2 illustrates a more uniformly weighted image Medicina, Noto (Italy), Torun´ (Poland), Onsala (Swe-
ofthe 5 GHz radio emissionsurroundingIC342X-1. We den), Urumqi (P.R. China), Svetloe, Zelenchukskaya,
Black hole powered nebulae and the case of IC342 X-1 3
TABLE 1
SummaryofObservations
Source Instrument Config. Central On-source Observation Flux
Frequency Time Date
IC342X-1 VLA B 4.8GHz 3.2h 6Dec07 2.0±0.1mJy
IC342X-1 VLA C 4.8GHz 3.2h 25Apr08 *
HoIIX-1 VLA B 4.8GHz 3.2h 8Dec07 613±61µJy
HoIIX-1 VLA C 8.5GHz 3.2h 21Apr08 395±40µJy
NGC5408X-1 ATCA 6D 5.5GHz 12h 23Aug09 226±33µJy
NGC5408X-1 ATCA 6D 9GHz 12h 23Aug09 137±36µJy
NGC5408X-1 ATCA 6D 17.9GHz 12h 23Aug09 76±20µJy
NGC5408X-1 ESOVLT - 575-731nm 0.7h 12Apr10 -
IC342X-1 EVN - 1.6GHz 12h 15Jun11 ≤21µJy**
IC342X-1 Swift/XRT PC 0.3-10keV 9.4ks 15Jun11 2.67***
Note. —*BandCconfigurationdatawerecombined.**Three-sigmaupperlimit.***Unabsorbedfluxinunitsof10−12 ergcm−2 s−1.
Fig. 1.—VLA5GHz imageof IC342 X-1overlaidonthe HαHST image. Contours represent radioemissionandaredrawnat 3, 4,5,
6, 7, 8, 9, and10 times the rmsnoiselevel of11 µJy beam−1. The peak brightness is 122.4 µJybeam−1. Theresolution of the imageis
2′.′3×1′.′6atPA=−13◦ andtheimagewasmadewithRobust=0weighting. Thesign’><’markstheX-raypositionoftheULX.
TABLE 2
Hubbleimagesandsource parameters
Name NarrowBandFilter Centeredon Ref. Distance Optical Radio
diameter diameter
IC342X-1 F658N Hα 1 3.9Mpc 190pc 225pc
HoIIX-1 FR462N Heii 2 3.39Mpc 45pc 60pc
HoIIX-1 FR505N Hβ 2 - 101pc -
NGC5408X-1 F502N [Oiii]λ5007 3 4.8Mpc 60pc 40pc
Note. —References: 1-Feng&Kaaret(2008),2-Kaaret,Ward&Zezas (2004),3-Gris´eetal. (2012)
4 Cseh et al.
Fig. 2.— VLA 5 GHz image of IC342 X-1 overlaid on the Hα HST image shown in Figure 1. This image was made with Robust=-2
weighting and therefore the largest extended features are missing. The resolution is 1′.′6×1′.′1 at PA=−19◦. Contours represent radio
emissionandaredrawnat3,4,5,and6timesthermsnoiselevel of15µJybeam−1. Thepeakbrightness is96.3µJybeam−1. Thesign
’><’markstheX-raypositionoftheULX.
Fig.3.— Left: The 5 GHz VLA B-arrayimage of Ho II X-1 overlaid on the He ii HST image. Contours represent radio emissionat
levelsof3,4,5,6and7timesthermsnoiselevelof14µJybeam−1. Thepeakbrightnessis114µJybeam−1. Theimagewasmadeusing
Robust=0weightingandtheresolutionis1′.′5×1′.′0atPA=34◦. Right: The8.5GHzVLAC-arrayimageofHo II X-1overlaidonthe
Hβ HST image. Contours represent radio emission at levels of 3, 4, 5, 6 and 7 times the rms noise level of 15 µJy beam−1. The peak
brightness is 145 µJy beam−1. The image was made using Robust=-1 weighting and the resolution is 2′.′36×1′.′75 at PA=3◦. The sign
’><’markstheX-raypositionoftheULX.TheNorthdirectionisupinbothimages.
Black hole powered nebulae and the case of IC342 X-1 5
responding to 90% of the PSF at 1.5 keV); background
wasestimatedfromanearbycircularregionwitharadius
of 40 pixels and subtracted. We fitted the X-ray spec-
trum using the XSPEC (Arnaud 1996) spectral fitting
toolandthe swxpc0to12s6 20070901v011.rmf9 response
matrix.
Fitting the spectrum in the 0.3-10 keV band, with an
absorbed power-law, leads to a good fit with χ2/DoF=
14.7/25 with a photon index of Γ = 1.57+0.29 and
−0.26
an equivalent hydrogen absorption column density of
N = 4.5+2.0 × 1021 cm−2. The absorbed flux in the
H −1.5
0.3-10 keV band was 1.97×10−12 erg cm−2 s−1. The
source clearly has a hard X-ray spectrum with a flux
of 2.5 × 10−12 erg cm−2 s−1, slightly lower than any
of the previous XMM observations. The column den-
sityis consistentwithin the errorswith the XMMvalues
fromFeng & Kaaret(2009). Theunabsorbedfluxis2.67
×10−12 erg cm−2 s−1, corresponding to an unabsorbed
Fig.4.—Theradiospectrum ofthenebulaofHo II X-1. The luminosity of 4.86×1039 erg s−1 in the 0.3-10keV band
bestfitspectralindexisα=−0.53±0.07forafluxdensity,S∝να.
at a distance of 3.9 Mpc.
Badary(Russia)andthephasedarrayoftheWesterbork 2.3. ATCA observations of NGC5408 X-1
Synthesis Radio Telescope (WSRT; The Netherlands).
The radio nebula of NGC5408 X-1 was the first de-
Theaggregatebitrateperstationwas1024Mbps. There
tected radio counterpart of a ULX (Kaaret et al. 2003).
were eight 8 MHz intermediate frequency channels (IFs)
Later, it was confirmed that it is an extended source
in both left and right circular polarisations.
(Lang et al. 2007). We obtained deep high-frequency
The source was observed in phase-reference mode.
ATCA CABB (Compact Array Broadband Backend)
Thisallowedustoincreasethecoherentintegrationtime
(Wilson et al. 2011) observationsto better constrainthe
spent on the target source and thus to improve the
morphology and the high-frequency part of the radio
sensitivity of the observations. Phase-referencing in-
spectra of the nebula at 5.5, 9, and 18 GHz.
volves regularly interleaving observations between the
We observed NGC5408 X-1 with the CABB-upgraded
target source and a nearby, bright, and compact refer-
ATCA in configuration 6D (baselines up to 6 km) on
ence source. The delay, delay rate, and phase solutions
2009 Aug 23 (program code: C1159). The data were
derived for the phase-reference calibrator (J0344+6827)
obtained simultaneously at 5.5 & 9 GHz and at 17 &
was interpolated and applied to the target within the
19GHzwith12hon-sourceintegrationtimeforbothsets
target-reference cycle time of 5 min. The target source
(Table 1.). We observed in phase-reference mode; the
was observed for 3.5-min intervals in each cycle.
phasecalibratorwas1424-418andtheprimarycalibrator
AIPSwasusedfortheVLBIdatacalibrationfollowing
was PKS 1934-638. The data reduction was performed
standardprocedures(Diamond1995). Thevisibilityam-
using the miriad software package (Sault et al. 1995).
plitudes were calibrated using system temperatures and
We combined the 17 GHz and 19 GHz images in order
antenna gains measured at the antennas. Fringe-fitting
to enhance sensitivity.
was performed for the calibrator sourcesusing 3-min so-
Fig.5showstheATCAimagesofthenebulasurround-
lution intervals. The calibrated visibility data were ex-
ing NGC5408 X-1 at 5.5, 9 and 18 GHz. The 5.5-GHz
ported to the Caltech Difmap program (Sheperd et al.
image was uniformly weighted, and the 9 and 18-GHz
1994) used to make a naturally weighted VLBI image.
maps were naturally weighted in order to achieve the
The achieved1-σ rms noise levelwas 7 µJy beam−1 and
best sensitivity. Plotting the geometric mean beam size
we did not detect any source above the 3-σ noise level
vs. the flux density, Lang et al. (2007) found that the
in a field of view of 700 × 700 mas. Therefore,the VLA
turnover indicates that the radio emission is resolved
core is likely to be a clump of emission from the nebula.
with an angular size of 1.5–2.0 arcsec. Our new radio
Our EVN observation places a 3-σ upper limit on the
imagesshowthatthesizeofthenebulaisconsistentwith
flux density ofF ≤ 21µJyandonthe luminosity ofthe
ν the previously estimated angular size. At 9 and 18 GHz
putative compact jet of ≤1.9×1033 erg s−1.
the source extent is only slightly larger than the corre-
spondingbeamsizes,butthe elongatedcontourssuggest
2.2.2. Swift observation
that the nebula is resolved. This is typical for a weak,
The Swift X-ray Telescope (XRT) obtained 9394 sec- optically thin, steep-spectrum source, i.e. going towards
ondsofgoodexposureinitsphoton-counting(PC)mode higherfrequencies,onegainsresolutionwhiletherelative
on June 15, 2011, from 02:57:31 to 17:47:47 (UT). We sensitivity decreases.
retrieved level two event files and used the default data We used all available measurements to fit the ra-
screeningmethodsasdescribedintheXRTuser’sguide8. dio spectrum. Fig. 6 shows previous measurements
We extracted an X-ray spectrum for the source using a (Lang et al. 2007) and our new ones; combined they
circularextractionregionwith aradiusof20pixels(cor- cover the 1.4 – 18 GHz frequency range. Our new flux
8 http://heasarc.nasa.gov/docs/swift/analysis/ 9http://heasarc.nasa.gov/docs/swift/proposals/swift responses.html
6 Cseh et al.
Fig. 7.—The18GHzATCA6D-array,naturally-weightedimage
Fig.5.—The5.5(red),9(black)and18(blue)GHzATCA6D- of NGC5408 X-1 overlaid on the [O iii] HST image. The sign
arrayimageofNGC5408X-1. Contoursrepresentradioemission ’><’ marks the X-ray position of the ULX. The North direction
atlevelsof3,4,5,6and7timesthermsnoiselevelof23,17and isup.
13µJybeam−1,respectively. Thepeakbrightnessesare208,110,
66µJybeam−1. Theimagewasmadeusinguniformweightingat
5.5GHzandnaturalweightingat9and18GHzandtheresolutions tical nebula, however it is possibly an effect of the steep
are 1′.′5×1′.′0 at PA=−9◦, 2′.′6×1′.′5 at PA=−28◦, 1′.′0×0′.′8 at radio spectrum.
PA=−1◦ at5.5,9and18GHz,respectively.
2.4. ESO VLT observations of NGC5408 X-1
As one can gain information about the ionization pro-
cess from the line flux ratios of Hα vs the forbidden
sulphur lines, we conducted VLT observations. VLT
FORS-2observationsof NGC5408X-1 were obtainedon
12 April 2010 using the GRIS 1200R grism with a slit
width of 1.0′′ covering the spectral range 5750−7310 ˚A
with dispersion 0.38 ˚A pixel−1 and spectral resolution
λ/∆λ = 2140 at the central wavelengths, respectively.
The observation block (OB) consisted of three 849 s ex-
posures with a 12 pixel offset along the spatial axis be-
tween successive exposures. CCD pixels were binned for
readoutby2inboththespatialandspectraldimensions.
Theaverageseeingforournewobservationswas0.62arc-
second. DatareductionstepscanbefoundinCseh et al.
(2011). Hereweuseatracewidthof8pixelscorrespond-
ingto2′′inordertoacceptalargefractionofthenebular
emission.
Fig. 8 shows the optical spectrum of NGC5408 X-1.
Fig.6.—Theradiospectrumofthe nebulaofNGC5408 X-1. Thisportionoftheopticalspectrumshowstheforbidden
The best fit spectral index is α = −0.8±0.1 for a flux density, sulphurandnitrogenlinesandtheHαline. Thelinesare
S∝να.
at wavelengths of 6558.9 ([N ii]), 6573.7 (Hα), 6594.4
([N ii]), 6727.6 ([S ii]), and 6742.0 ([S ii]) in units of
densities are 226±33, 137±36, and 76±20 µJy at 5.5, ˚A. The corresponding fluxes are 0.46±0.02, 33.1±0.1,
9, and 18 GHz, respectively. The fitted radio spectral 1.07±0.02,1.57±0.01,1.15±0.01intheunitsof10−16erg
index of NGC5408 X-1 is α = −0.8±0.1. This is the s−1 cm−2. The Gaussian FWHMs of the lines represent
best constrained spectrum of a ULX nebula to date and the instrumental resolution of ∼2.4 ˚A.
is consistent with previous results suggesting optically
thinsynchrotronemission(Soria et al.2006a;Lang et al. 3. DISCUSSION
2007).
3.1. The optical nebula around IC342 X-1
Fig. 7 shows the ATCA image at 18 GHz, overlaid on
the HST imageof the opticalnebula ofNGC5408X-1 of In this section we use models developed for expand-
Gris´e et al. (2012). The HST filter was centered on the ing bubbles, to calculate parameters of the nebula
forbidden [O iii] nebular emission line (Table 2). The based on the observed optical properties. We follow
optical image shows a one-sided shell-like structure dis- the formalism developed by Weaver et al. (1977) and
placedfromthe ULXwhichmightbe due to geometrical Ostriker & McKee (1988) describing the hydrodynamic
effectslikelimbbrighteningtowardstheNE.Wefindthat structureofabubble. Theseformulaearevalidforshock-
the radio emissionat 18 GHz originates ”inside” the op- dominated sources, i.e. jet or wind inflated bubbles, but
Black hole powered nebulae and the case of IC342 X-1 7
thatthecolor-magnitudediagramsuggeststhatthemini-
mumstellarageintheenvironmentoftheULXis10Myr.
The characteristicageof the nebula is muchshorterand
might suggest that the nebula formation is not related
to the formationofthe centralBHorBH progenitor,in-
stead it might represent the actively accreting phase of
the binary.
Recalling the scaling of the total radiative flux and
the flux in the Balmer lines (Dopita & Sutherland 1996;
Abolmasov et al. 2007)10:
v −0.59
L =6.53×10−3 exp L (2)
Hβ 100kms−1 rad
(cid:16) (cid:17)
So, the total shock power represented as radiative losses
is L = 1.2 × 1039 erg s−1; using the Hβ flux of
rad
Abolmasov et al. (2007) of 4.3 × 10−15erg s−1 cm−2.
As a consequence, the total mechanical luminosity is
L = 77L =3.4×1039 ergs−1, andthe totalkinetic
tot 27 rad
Fig.8.—VLTopticalspectrumofthenebulaaroundNGC5408 powercarriedbytheswept-upmassisL =6.6×1038erg
k
X-1. Thedereddenedfluxisplottedvs. wavelengthandnoredshift s−1. Similarly, the internal (thermal) luminosity is
correctionwasapplied.
L = 1.5×1039 erg s−1. The energy, we see at t=τ
th
is E = L τ = 6.0×1052 erg. From Eq. 1 one can
tot tot
notforphotoionizedbubbles. AsIC342X-1isconsidered derive that n=1.0 cm−3 by substituting R,τ,L . The
tot
to be a shock-dominatedsource (Pakull & Mirioni 2003; opticalswept-upmassisthenM =µnm V =2.4×1038g
p
Roberts et al. 2003; Abolmasov et al. 2007), we can ap- or M =1.2×105 M⊙.
ply the well-known self-similar expansion law as a func-
IC342 X-1 is sometimes considered as a SN remnant
tion of time, t:
(Roberts et al. 2003, eg.). Here we intend to show that
1/5 1/5 thetotalenergycontentdoesnotdependsignificantlyon
125 L
R= tot t3/5 (1) the interpretation of the origin of the bubble. A SNR in
(cid:18)154π(cid:19) (cid:18) ρ0 (cid:19) the pressure-drivensnowplowstage– ie. radiativedomi-
nant phase following the adiabatic phase – has an ini-
where L is the mechanical luminosity (corresponding
tot tial explosion energy, E (Cioffi, McKee & Bertschinger
toajetorawindoraninitialexplosionenergy),Risthe i
1988):
radius of the bubble expanding with velocity v = R˙
exp
sintatontthaendISρM0.=Tµhmepmna,swshdeernesiµty=ρ10.3is8aissstuhmeemdetaonbaetocmonic- Ei =6.8×1043(cid:18)pRc(cid:19)3.16 kmvexsp−1 1.35 cmn−3 1.16erg
weight, m is the proton mass and n is the hydrogen (cid:16) (cid:17) (cid:16) (cid:17)
p (3)
number density. The characteristic age of the bubble is
We have treated the metallicity correctionfactor, ζ0.161,
τ =3R/5v . The kinetic energy carried by the swept- m
exp as 1 for clarity (Cioffi, McKee & Bertschinger 1988;
upmassintheexpandingshellisE = 15L t,whilethe
k 77 tot Thornton et al. 1998). Substituting R, v – and taking
energyemitted (the cooling)by the fully radiativeshock n=1.0cm−3 (seeabove),weobtainE =6.0×1052 erg.
expanding into the ISM is Erad = 2777Ltott. The ther- We note that a similar formula of Cheivalier (1974) pro-
malenergy of the gas betweenthe reverseshock and the vides E = 5.0 × 1052 erg. This initial energy is re-
swept-up shell is E = 5 L t. Thus, the total energy i
th 11 tot markably high, although somewhat expected as a sin-
is E =E +E +E =L t.
tot k rad th tot gle/simple SNR will not remain visible once it has ex-
Feng & Kaaret(2008)foundthatthebrightmainbody
panded beyond R ≃ 100 pc with a canonical E ≃
of the nebula has an angular diameter of about 6′′ in 1051 erg (Matonick & Fesen 1997; Roberts et al. 20i03).
the Hα image. Pakull & Mirioni (2003); Gris´e et al.
Theestimatedenergyisingoodagreementwiththetotal
(2006); Feng & Kaaret(2008) report an additional elon-
energy content of the bubble, E , within model uncer-
gated structure to the South-West. Considering the en- tot
tainties. We note that an SNR nature for the nebula
tire structure of the bubble, we estimate the volume of
around IC 342 X-1 is strongly challenged by its high
the optical nebula by taking a sphere with a diameter of
shock velocity coupled with large size (Pakull & Gris´e
∼ 10′′, which corresponds to ∼190 pc at a distance of
2008).
3.9 Mpc Tikhonov & Galazutdinova (2010).
The high flux ratio between the forbidden [S ii] lines
3.2. The radio nebula around IC342 X-1
at 6,716 ˚A and 6,732 ˚A and the Balmer Hα line indi-
When radiation is via synchrotron emission, one can
catesthepresenceofshock-ionizedgas. Abolmasov et al.
(2007) infer a shock velocity of v ≃20−100 km s−1, assume equipartition between energy of relativistic par-
exp
however, line ratios of a standard library of radiative
10 We note, that there is a typo in Eq. 3.3 of
shock models, e.g. He ii vs. Hβ is 0.036±0.015, suggest
Dopita&Sutherland (1996) (and consequently in Eq. 1 of
a shock velocity of ≃ 100 km s−1 (Allen et al. 2008). Abolmasovetal.(2007)),pointedouttousbyM.W.Pakull. The
Using this velocity, we find the characteristic age of the numericalfactorshouldread1.14×10−3 ratherthan2.28×10−3
bubble is τ =5.6×105 yr. Feng & Kaaret (2008) found asthemaximumradiativefluxis 21ρv3 andnotρv3.
8 Cseh et al.
ticles and the magnetic field. We calculate the mini- 3.3. The radio and optical nebulae of Ho II X-1 and
mum total energy of the radio nebula that corresponds NGC5408 X-1
to equipartition (Longair 1994): 3.3.1. Ho II X-1
V 3/7 ν 2/7 L 4/7 Our new, higher resolution VLA radio observations
E =3×1013η4/7 ν erg clearly reveal that the morphology of the radio neb-
min (cid:18)m3(cid:19) (cid:16)Hz(cid:17) (cid:18)WHz−1(cid:19) ula follows the structure of the optical one (Fig. 3).
(4) The radio nebula is resolved with a size of 5.5” ×
where η − 1 is the ratio of energy in protons to rela- 2.7”, corresponding to ∼ 81 pc × 40 pc. One could
tivistic electrons, V is the volume, ν is the observing argue that this morphology reflects either a jet activ-
frequency and Lν is the synchrotron luminosity. As is ity or an outflow. However, Pakull & Mirioni (2002);
customary, we do not account for relativistic protons, Kaaret, Ward & Zezas(2004)showed,thatthenebulaof
therefore η−1 = 0 (eg. Fender et al. 1999). Substitut- Ho II X-1 is consistent with photoionization by the cen-
ing the corresponding values of V ≃ 1.79 × 1056 m3, tral compact source. On the other hand, a complex ve-
ν = 5×109 Hz, L = 3.64×1018 W Hz−1 and a fill- locitystructureinthe innerregionsoftheopticalnebula
ν
ing factor of unity, we find the energy required to power indicatestheimpactofthecentralobjectalsointheform
the nebula is E =9.2×1050 erg. This suggests that of winds or jets (Lehmann et al. 2005). As the [S ii] vs.
min
the radio-emitting material carries a fraction ∼ 10−2 of Hα ratio is ≪ 1 (Abolmasov et al. 2007), i.e. collisional
the initialenergy. Forcomparison,the energycarriedby excitation of the nebula is negligible, the morphology
mildlyrelativisticmaterialinnormalTypeIcsupernovae probablyreflectsanoutflowratherthanawell-collimated
has been suggested to be at most 10−4 (Paragi et al. jet. This outflow is either relatively weaker than a jet –
2010), while for the jet inflated bubble around the ex- thuspreventingshock-ionization–ortheoutflowismore
tragalactic microquasar S26, this fraction is a few times isotropic than collimated. We note that weakly colli-
10−3 (Soria et al. 2010). These fractions suggest that mated outflows, in addition to jets, have been directly
most of the energy is stored in protons, nuclei, and non- observed in SS433 with VLBI (Paragiet al. 1999). In
relativistic bulk motion. addition, the asymmetry of the outflow is probably due
Wecalculatethemagneticfieldstrengthcorresponding tothefactthatthenebulaisdensityboundedtotheEast
to the minimum energy condition (Longair 1994) and South of the central object (Pakull & Mirioni 2002;
Kaaret, Ward & Zezas 2004).
ηL 2/7 Furthermore, the minimum energy required to power
Bmin =1.8×1010 ν ν1/7 µG (5) theradionebulaofHoIIX-1is2.6×1049ergwithamag-
(cid:18) V (cid:19)
neticfiledstrengthof13µGandthe synchrotronlifetime
isτ =25Myr(Miller et al.2005). Thisenergyrequire-
where the units and inputs are the same as above. We sy
ment is a factor of 35 less than needed for IC342 X-1,
obtain B =7.4µG.
min a shock-dominated source, which might also support a
When energy loss is due to synchrotron radiation, the
weak or uncollimated outflow.
lifetime of an electron is (Longair 1994; Tudose et al.
Interestingly, Gris´e et al. (2010) studied an X-ray
2006)
statetransitionofHoIIX-1andfoundthatitisdifficult
to interpret the thermal component of the X-ray spec-
τ =2.693×1013 ν −1/2 Bmin −3/2 yr (6) tra as disk emissionor thermalComptonization. On the
sy (cid:16)Hz(cid:17) (cid:18) µG (cid:19) other hand, this thermal component might be linked to
thecomplexvelocitystructureofthenebula,ie. itmight
Substituting ν =5×109 Hz andB =7.4µG,we find be due to a disk wind that results in a complex veloc-
min
τ = 18.8 Myr. So, consistently, the cooling time-scale ity structure of the nearby environment. We note that
sy
is ∼34 times longer than the age of the bubble and ∼2 disk wind signatures have been found for GRS1915+105
times higher than the minimum age of the ULX stellar (Neilsen & Lee 2009) and SS433 (Fabrika 2004), i.e.
environment. for sources likely accreting above their Eddington-limit.
We note that the value of η, the energy ratio of rel- Searching for relativistic disk lines and/or Fe K absorp-
ativistic protons to electrons, can be estimated from tionline variationin the X-rayspectrum of ULXs might
the high-energy part of SNR spectra (Abdo et al. 2010; helpdisentanglewhethersomeoftheseULXsaresimilar
Acciari et al. 2011). Assuming IC342 X-1 has a ratio to high-accretion rate Galactic binaries.
similar to that found for a scenario in which the high-
energy gamma-rays from the Tycho SNR are produced 3.3.2. NGC5408 X-1
by leptons (Acciari et al. 2011), then η −1 = 102 and Fig. 7. shows a shell-like optical morphology of
the minimum energy obtained from Eq. 4 is increased NGC5408 X-1 and a filled structure of the radio neb-
by a factor of η4/7 = 1004/7 = 13.9. This would mean ulawithamaximumintensityneartheULX.Theoptical
that Emin/Etot = 0.2. Instead, if the gamma-ray emis- nebulaisresolved,withadiameterof∼2.5”,correspond-
sion from Tycho is from hadrons, then η−1=2.5×103 ingto∼60pc(Gris´e et al. 2012),whichismuchsmaller
andE /E wouldbe ≃1.3. Thus, if the high-energy thantheopticalnebulaofIC342X-1. Probably,theneb-
min tot
gamma-rayemissionof shock inflated bubbles originates ula is powered only by photoionization without any jet
from the same electron distribution that produces the oroutflowactivity. Ouropticalspectraalsosuggestthis,
radio,then via the η parameter,the value ofE could as the [S ii] vs. Hα line ratio is ≪ 1 (Fig. 8.). Fur-
min
increase, but interestingly, does not violate the total en- thermore, the energy needed to power the radio nebula
ergy obtained from the optical. of NGC5408 X-1 is 3.6×1049 erg with an equipartition
Black hole powered nebulae and the case of IC342 X-1 9
fieldof 16µGanda synchrotroncooling time ofτ =20 tended emission in optical and/or radio. Table 3 shows
sy
Myr (Lang et al. 2007), which is, similar to Ho II X-1, themaincharacteristicsofthe specificsources: the aver-
a factor of 25 less than the energy needed to power the ageX-rayluminosity(L ),thejetpowerestimatedfrom
X
radio nebula of IC342 X-1. theenvironment(Q ),themassoftheblackhole(M ),
j BH
thetotalenergycontent(E ),thelifetimeofthebubble
tot
3.4. Upper limit on the mass of the BH in IC342 X-1
(τ), and the minimum energy estimated from the radio
Self-absorbed compact jets are ubiquitous in the counterpart of the nebula (Emin).
X-ray hard states of GBHBs (Corbel et al. 2004; The nebula around IC10 X-1 is sometimes con-
Fender, Belloni & Gallo 2004). IC342 X-1 was found sidered to be the result of a hypernova event
to have a hard X-ray spectrum in all available XMM- (Lozinskaya & Moiseev 2007). However, its size is much
Newton and Chandra observations covering the period larger than allowed by the surface brightness versus
from2001to2006(Feng & Kaaret2009,2008)aswellas size (Σ-D) relation for explosive events, i.e. supernovae
in our recent Swift/XRT measurement (see Sec. 2.4.1.), (Yang & Skillman 1993), suggesting that it is powered,
thus the presence ofcompact jets is expected if the hard instead, by the central BH (Bauer & Brandt 2004). S26
X-rayspectrumequivalenttothe canonicalhardstateof is another outlier source in the Σ-D relation and it
GBHBs. has been revealed that the bubble is powered by a jet
In addition, the morphology of the radio nebula is (Pakull, Soria & Motch 2010; Soria et al. 2010). There-
somewhat similar to the system SS433/W50 and may fore, we will consider IC10 X-1 as a black hole powered
suggest a jet orientation along an axis slightly East of nebula. Wenotethat“blackholepowerednebula”means
North (Feng & Kaaret2008). As we pointed out earlier, thereisasignificantandcontinuingoutflowfromthecen-
an unresolved radio source is detected at the location of tral black hole powering the nebula and does not argue
the ULX. This morphology might argue against the no- againsttheformationoftheblackholeinaSNexplosion.
tionthatthecompactemissionisthehotspotattheend Considering S26, the X-ray photon index is Γ = 1.4
of the jet pointing close to our line of sight. To test the (Soria et al. 2010), consistent with a source being in the
compactness of this emission, we conducted VLBI mea- hardstate. However,noradiocorewasdetectedwitha3-
surementsusingtheEVN(Sec2.2)andwedidnotdetect σ upperlimit of0.03mJy,whichisonethirdofthe peak
any source above the 3-σ noise level, thus it is likely to intensity of IC342 X-1; and the total jet power of S26 is
be consistent with a clump of emission from the nebula anorder higher than for IC342X-1 (Table 3). Using the
andmightbeaneffectofasteepradiospectrum. Wecan fundamental plane (Eq. 7), we obtain an upper limit on
set an upper limit on the flux of the putative compact the mass ofM ≃(1.0±0.4)×107 M⊙. Giventhat Qj ∝
jet of 1.9×1033 erg s−1, using our 3-σ noise level of 21 M˙ in the hard state (K¨ording, Fender & Migliari2006b;
µJy. K¨ording, Jester & Fender2008)andassumingsimilarac-
Whenblackholesareinthehardstate,i.e.theiraccre-
cretion efficiencies for IC342 X-1 and S26; then one can
tionisradiativelyinefficientandperhapsadvectiondom-
speculate on the basis of the ratios of the jet powers
inated or jet dominated, a relationship holds between
that the averagemass accretionrate of S26 is ∼15 times
X-ray luminosity, radio luminosity and black hole mass
higher than that of IC342 X-1.
(Merloni et al. 2003; Falcke,K¨ording & Markoff 2004).
IC10 X-1 has an X-ray photon index of Γ = 1.83, po-
This relationship, the fundamental plane of black holes,
tentiallybeinginthehardstate(Bauer & Brandt2004).
has been studied on a wide mass range from black
Using the fundamental plane and adopting the average
hole X-ray binaries to low luminosity active galactic nu-
clei. Using the correlation with the minimum scatter X-rayluminosityof1.5×1038ergs−1 andamassof∼30
(K¨ording et al. 2006a): M⊙ (Table 3), the expected core radio flux is then ∼10
µJy at a distance of 0.7 Mpc. Thus, the source could be
log M =1.02−1 1.59log LR −log LX −10.15 . detectable with e-MERLIN or the EVLA.
(cid:18)M⊙(cid:19) (cid:18) (cid:18)erg s−1(cid:19) (cid:18)erg s−1(cid:19) (cid:19)
(7)
andourradioandX-raymeasurements,onecanestimate 3.5.1. Jet characteristics, accretion rates and efficiencies
the upper limit of the mass of the black hole in IC342
In this and the following sections we investigate the
X-1. However,wemustconsidertheintrinsicrmsscatter
possibility that the nebula aroundIC342 X-1 is powered
in the measured fundamental plane relation of 0.12 dex
by a jet and the consequences regarding the jet proper-
within one σ. Substituting the upper limit of radio lu-
ties.
minosity of the putative compactjet (L=1.9×1033 erg
Theelongatedmorphologyofthenebulaanditsshock-
s−1) and the simultaneously measured X-ray luminosity ionized nature are indicative of jet inflation. If the neb-
LX = 4.86×1039 erg s−1, we estimate the mass of the ula is inflated by a jet, then the total power available
black hole to be MBH ≤ (1.0 ± 0.3) × 103 M⊙. This in the bubble has to be equal with the time-averaged
limitisvalidonlyifIC342X-1entersthecanonicalhard total jet power, ie. Q =L , derived from the optical
j tot
state. Wenotethatfurtherobservationscouldhelpplace bubble (Pakull, Soria & Motch 2010; Soria et al. 2010).
tighter constraints on the BH mass or help test if ULXs We note that calculating the jet power using the min-
exhibiting hard X-ray spectra are, indeed, in the radia- imum energy (E ) derived from the radio bubble
min
tively inefficient, hard X-ray state (see section 3.5.1). would lead to an underestimation. This could be either
due to a mild deviation from equipartition, resulting in
3.5. Comparison of IC342 X-1 with S26, IC10 X-1 E /E =10−100(eg.Paragiet al.2010)ordue
and SS433 radio,tot min
to the fact that the kinetic power associatedto the bulk
InthissectionwecompareIC342X-1tosourceswhose motion is transferred to thermal ions also (Soria et al.
jetpowerandaccretionratehasbeenestimatedfromex- 2010).
10 Cseh et al.
TABLE 3
Comparisonwith X-raysources embedded in nebula
Name LX (erg/s) Qj (erg/s) MBH (M⊙) Ltotτ =Etot (erg) τ (yr) Emin (erg)
IC342X-1 1.6×1040 3.4×1039 .1.0×103∗ 6.0×1052 5.6×105 9.2×1050
S26 6.2×1036 5×1040 n/a 3.16×1053 2×105 1050
IC10X-1 1.5×1038 1.27×1039 23-34 2×1052 5×105 n/a
SS433 ∼1036 2×1038 16 2×1051 2×105 1049∗∗
Note. — The Table shows the average X-ray luminosity (LX), the time-averaged jet power estimated from the environment (Qj),
the mass of the black hole (MBH), the total energy content(Etot), the lifetime of the bubble (τ), and the minimum energy esti-
mated from the radio counterpart of the nebulae (Emin). SS433 (Kirshner&Chevalier 1980; Begelmanetal. 1980; Dubneretal. 1998;
Blundell,Bowler&Schmidtobreick2008;Perez&Blundell2009;Fabrika2004). S26(Pakull,Soria&Motch2010;Soriaetal.2010).IC10
X-1(Lozinskaya&Moiseev2007;Bauer&Brandt2004). *Estimatedusingthefundamentalplane.**ThisvaluewascalculatedusingEq.4
usingtheparametersfoundbyDubneretal.(1998): aradiusof∼30pc,adistanceof3kpc,aradiofluxdensityof71Jyat1.4GHz,and
takingaspectral indexof -0.48between 85 MHzand5GHz. We notethat this valueistwoordersof magnitude largerthanthat quoted
byDubneretal.(1998),however consistentwiththevalueofBegelmanetal.(1980).
In general, the total jet power is a constant frac- g s−1 (1.3×10−6 M⊙ yr−1). This rate is close to the
tion (f) of the available accretion power (however see Eddingtonratefora100M⊙compactobject. Comparing
Coriat et al. (2011) for possible variation of f). Thus, theaccretionratetothe jetmasslossrateofM˙ ≃3.8×
j
we can write 1018 g s−1 leads to the conclusion that the jet efficiency
Q =fQ =fM˙ c2, f <1 (8) is f ≥ 0.046. This would be a typical value for a jet
j acc acc
efficiencyandcouldbeconsistentwiththeassumptionof
where f is typically in the range of 10−3 to 10−1 K¨ording, Jester & Fender(2008)off=10−1forobtaining
(Falcke & Biermann 1995, 1999). the relationship of Eq. 11.
Taking a constant rate of energy input that is charac- Let us compare IC342 X-1 to the peculiar sources
terized by the power of the jet, the rest-mass transport SS433 and GRS1915+105. The jet mass flow rate in
along the jet, M˙j, is (Kaiser & Alexander 1997): SS433 is M˙j = 5×10−7 M⊙ yr−1 and the available ac-
Q cretion rate is M˙ = 10−4 M⊙ yr−1 (Perez & Blundell
M˙ = j (9) 2009;Fabrika2004), sothe jetefficiency isf =5×10−3,
j (γ−1)c2
whichis alsowithin the typicalrangefor blackhole jets.
Taking a minimum Lorentz factor of γ = 2 Comparingthe accretionrateofIC342X-1 to SS433,we
(Mirabel & Rodr´ıguez1999;Fender2003),wefindM˙j ≃ find M˙IC342X−1/M˙SS433 ≥ 10−2. In contrast, the accre-
6.0×10−8 M⊙ yr−1 (3.8×1018 g s−1). thiiognherrattehaonf ItChe34r2atXe-o1fmGiRghSt1b9e15a+n1o0r5deorfo∼f m10a1g9ngitus−d1e
K¨ording, Jester & Fender (2008) found that the total
(eg. Rushton et al. 2010). It is interesting to note that
jetpowercouldbeestimatedfromthefluxofthecompact
if one takes the lifetime of the surrounding bubble, and
jet as:
assumes a constant accretion rate (Begelman 2002), the
Qj ≃7.2×1036(cid:18)10L3j0ete,rrgadsi−o1(cid:19)12/17 ergs−1 (10) tchoemlpifaecttimobejoecftthineSnSe4b3u3lac,otuhledcaocmcrpetaect∼o2b0jeMct⊙i.nDIuCr3in4g2
X-1 could accrete only .1 M⊙ . This comparison of
This relationship was found for FR I and FR II radio accretion rates suggests that there is no need to invoke
galaxies and scaled to Cyg X-1. (See Gallo et al. (2005) super-Eddington accretion for IC342 X-1 if the mass of
and Russell et al. (2007) for the jet power estimation of black hole is above ≃100 M⊙.
CygX-1). Substitutingthetotaljetpowerobtainedfrom
theopticalnebula(Qj =3.4×1039ergs−1),wefindthat 4. CONCLUSIONS
the expected average luminosity of the putative radio
We studied the radio and optical nebulae of three
corewouldbe6.1×1033ergs−1. Notdetectingacompact
ULXs. One of these ULXs, IC342 X-1, is a newly dis-
jet with an upper limit of 1.9×1033 erg s−1 could mean
coveredradio association.
that either its flux is variable, and currently below our
detection limit but with an average above, or the hard
X-ray spectrum does not represent the canonical hard • Weconfirmedthattheradiospectraofthenebulae
state of GBHBs. surrounding Ho II X-1 and NGC5408 X-1 are con-
In addition, it is possible to estimate the accretion sistent with optically thin synchrotron emission.
rate, for hard state objects, based on the luminosity
of the compact radio jet (K¨ording, Fender & Migliari • We estimated the energy needed to supply both
(2006b); K¨ording, Jester & Fender (2008); and refer- the optical and radio nebulae around IC342 X-1.
ences therein): The energy needed, 6×1052 erg, is ∼ 2 orders of
magnitude higher than the explosion energy of a
12/17
L SN. By comparing the age of the bubble to the
M˙ ≃4×1017 jet,radio gs−1 (11)
(cid:18)1030ergs−1(cid:19) stellar environment, we found that the nebula is
much younger, indicating that the nebula forma-
ForIC342X-1,we obtainanupper limit, due to the fact tion is not necessarily related to the formation of
that we do not detect a compact jet, of M˙ ≤ 8.2×1019 the black hole progenitor.