The X-ray quiescence of Swift J195509.6+261406 (GRB 070610): an optical bursting X-ray binary?

0.31 MB·English

by  N. Rea

Draft version January 19, 2011
PreprinttypesetusingLATEXstyleemulateapjv.05/04/06
THE X-RAY QUIESCENCE OF SWIFTJ195509.6+261406 (GRB070610):
AN OPTICAL BURSTING X-RAY BINARY?
N. Rea1, P. G. Jonker2,3,4, G. Nelemans4, J. A. Pons5, M. M. Kasliwal6, S. R. Kulkarni6, R. Wijnands7
Draft version January 19, 2011
ABSTRACT
Wereportona∼63ksChandra observationoftheX-raytransientSwiftJ195509.6+261406 discov-
eredastheafterglowofwhatwasfirstbelievedtobealongdurationGamma-RayBurst(GRB070610).
The outburst of this source was characterized by unique optical flares on timescales of second or less,
1
morphologically similar to the short X-ray bursts usually observed from magnetars. Our Chandra
1
observationwasperformed∼2yearsafterthediscoveryoftheopticalandX-rayflaringactivityofthis
0
source,catchingitinitsquiescentstate. Wederivestringentupperlimitsonthequiescentemissionof
2
SwiftJ195509.6+261406 which argues against the possibility of this object being a typical magnetar.

n Our limits show that the most viable interpretation on the nature of this peculiar bursting source, is
a a binary system hosting a black hole or a neutron star with a low mass companion star (<0.12M ),
J (cid:12)
and with an orbital period smaller than a few hours.

8
Subjectheadings: X-ray:individual(SwiftJ195509.6+261406)—stars: magneticfields—X-rays: stars
1

]
A 1. INTRODUCTION promptlyobservedthepositionofSwiftJ1955 duringthe
outburst. A highly variable optical and infrared coun-
G On 2007 June 10 the Swift Burst Alert Telescope
terpart was observed, showing large flares for about 11
(BAT)triggeredonGRB070610,atypicallong-duration
. daysaftertheGRB-likeevent,whenitwentbacktoqui-
h GRB (see Gehrels et al. 2007 for a recent review),
escence. These large flares were characterized by a very
p with a ∼4.6s high-energy prompt emission (Pagani et
short timescale: during the largest flare the source in-
- al. 2007; Tueller et al. 2007). Follow-up soft X-ray
o creased its optical flux by more than a factor of 200 in
observations with the Swift X-ray Telescope (XRT)
r less than 4s. Furthermore, a broad quasi periodic os-
t started soon after the event, discovering only one
s cillation was observed in the optical band at ∼0.16Hz
variable X-ray source within the BAT error circle:
a
(Stefanescu et al. 2008).
[ namely SwiftJ195509.6+261406 (Kasliwal et al. 2008;
SwiftJ1955 hereafter). This transient X-ray source was The source distance was constrained by several differ-
1 verydifferentfromwhatexpectedfortheX–rayafterglow ent methods to be within 3.7–10kpc (mainly red clump
v study, and detailed measurements of the absorption col-
of a long GRB: it was decreasing in flux rather slowly,
3 umn in the mm waveband; Castro-Tirado et al. 2008).
and it showed a strong X-ray flaring activity.
8 Furthermore, the stringent optical and IR limits derived
ThesourcebecameundetectablebySwift-XRTon2007
4 June 29, ranging from a 0.5–10keV flux of ∼ 10−9 to inthequiescentlevel(H>23;R>26.0andi(cid:48) >24.5;Kasli-
3 < 10−12ergs−1cm−2 in 19 days. While in outburst, wal et al. 2008; Castro-Tirado et al. 2008) constrain the
.
1 SwiftJ1955 had an X-ray spectrum that could be de- type of any companion star to either a main-sequence
0 star with spectral type later than M5V (which means a
scribedbyaratherhardpower-lawcorrectedforthepho-
1 toelectric absorption (N = 7×1021cm−2 and Γ=1.7). mass<0.12M(cid:12)),ortoasemi-degeneratehydrogenpoor
1 H star (Castro-Tirado et al. 2008).
Duetospatialandtemporalcoincidence(itwastheonly
: The large variations of its optical and infrared coun-
v transient source in the BAT error circle), GRB070610
terpart during the decay to quiescence, its distance and
i and the X–ray transient SwiftJ1955 have been associ-
X Galacticnature, setthistransientapartfromthetypical
ated with high probability (Kasliwal et al. 2008).
r Themostinterestingandpeculiarfeaturesofthistran- optical afterglows of long-duration GRBs (see Liang et
a al. 2007 for a recent review).
sient source came from optical and infrared observa-
The resemblance of the optical bursts of SwiftJ1955
tions. Manytelescopes,triggeredbytheGRB-likeevent,
with the short X-ray bursts from magnetars (see
1Institut de Ciencies de l’Espai (CSIC–IEEC), Campus UAB, Mereghetti 2008 for a recent review) led to the idea of a
FacultatdeCiencies,TorreC5-parell,2aplanta,08193,Bellaterra newkindofX-rayandopticaltransienteventinaGalac-
(Barcelona),Spain;Email: [email protected] tic magnetar (Castro-Tirado et al. 2008; Stefanescu et
2SRON-NetherlandsInstituteforSpaceResearch,Sorbonnelaan al. 2008). On the other hand, its X-ray flaring activity
2,3584CAUtrecht,theNetherlands
was also proposed to resemble the emission of the fast
3Harvard-Smithsonian Center for Astrophysics, 60 Garden
Street,Cambridge,MA02138,USA X-ray nova V4641 Sgr (Markwardt et al. 2008; Kasliwal
4Radboud University Nijmegen, Department of Astrophysics, et al. 2008), an unusual 9 M black hole in orbit with
(cid:12)
IMAPP,POBox9010,6500GL,Nijmegen,TheNetherlands a 5–8 M B9 III companion star (in’t Zand et al. 2000;
5Departament de Fisica Aplicada, Universitat d’Alacant, Ap. (cid:12)
Orosz et al. 2001)
Correus99,03080Alacant,Spain
6CaliforniaInstituteofTechnology,DepartmentofAstronomy, In this Letter we present the results of a ∼63ks Chan-
MailStop105-24,Pasadena,CA91125,USA dra observationofSwiftJ1955 (see§2)aimedatunveil-
7UniversityofAmsterdam,AstronomicalInstitute“AntonPan- ing its X-ray properties during quiescence (see §3), and
nekoek”,Postbus94249,1090GE,Amsterdam,TheNetherlands
compare them with the current quiescent levels of the
2 REA ET AL.
magnetar and X-ray binary populations (see §4). 1021cm−2,half of which is accounted for by a molecu-
lar cloud at 3.7kpc (considered as a lower limit on the
2. OBSERVATIONANDDATAANALYSIS distance of SwiftJ1955). Comparing the Galactic N
H
TheChandra X-rayObservatoryobservedSwiftJ1955 with the one derived from fitting the X-ray spectrum
for ∼63ks with the Advanced CCD Imaging Spectrome- during outburst (NH = 7.2+−32 ×1021cm−2; Kasliwal et
ter (ACIS) instrument (ObsID10042) from 2009 August al. 2008)9, SwiftJ1955 is expected to be located at ∼4–
03 16:09:57 to August 04 09:55:59 (Terrestrial Time) in 5kpc. Similar results have been derived from the red-
VERY FAINT (VF) timed exposure imaging mode. The clump method, from which a distance of ∼4kpc could
source was positioned on the back-illuminated ACIS-S3 be inferred (Castro-Tirado et al. 2008). Unfortunately
CCD at the nominal target position (RA: 19 55 09.653, the upper limit on the source distance is not very well
Dec: +26 14 05.84 ±0(cid:48).(cid:48)27; J2000), and we used a sub- constrained, although all methods used would place the
array of 1/8 leading to a time resolution of 1.14s. Stan- source ∼4–5kpc. The main problem to assess a dis-
dard processing of the data was performed by the Chan- tanceerrorbaristhattheGalacticplaneinthedirection
dra X-rayCentertoLevel1andLevel2(processingsoft- of SwiftJ1955 extends only until ∼5kpc, behind which
wareDS8.0). ThedatawerereprocessedusingtheCIAO thereisthebulge,withitsintrinsicallylowercolumnden-
software (version 4.1.2). We used the latest ACIS gain sity. This makes it extremely hard to define an upper
map, and applied the time-dependent gain and charge limittothedistance,sinceinthebulgeacolumndensity
transfer inefficiency corrections. The data were then fil- versusdistancerelationisnotwelldefined. Hereafter,we
tered for bad event grades and only good time intervals willdiscussourresultsassumingadistancerangeof3.7–
were used. No high background events were detected, 10kpc, considering the farthest limit in the Milky Way
resulting in a final on-time exposure of 62.732ks. as a distance upper limit.
The 95% upper limit on the quiescent X-ray lumi-
3. RESULTS nosity of SwiftJ1955 is 2.8×1030d2 ergs−1 or 1.9×
5kpc
We did not detect any X-ray source at the position 1030d2 ergs−1, for the power-law or blackbody mod-
of the optical counterpart to SwiftJ1955 (Kasliwal et 5kpc
els, respectively (see also §3). Considering the whole
al. 2008; Castro-Tirado et al. 2008). In particular
3.7–10kpc range (see above), the derived quiescent X-
we detected no 0.3–10keV photon within a 1(cid:48)(cid:48) circle
rayluminosityrangeisbetween1.5–11.4×1030ergs−1 or
centred on the optical position. We took a 95% up-
1.0–7.9×1030ergs−1, again assuming a power-law or a
per limit of 3 photons (Gehrels 1986) and inferred a
blackbody, respectively.
4.78 × 10−5 counts s−1 upper limit on the X-ray qui-
escent count-rate of SwiftJ1955. 4.1. On the magnetar interpretation
Using PIMMS8 we estimated the 95% upper limit on
The discovery of fast optical bursts in SwiftJ1955 led
the source flux assuming: a) an absorbed power-law
afewauthorstoclaimthemagnetarnatureofthissource
spectrum similar to the outburst spectral energy dis-
(Castro-Tirado et al. 2008; Stefanescu et al. 2008),
tribution (N = 7 × 1021cm−2 and Γ=1.7; Kasliwal
H basedonthesimilarityoftheopticallight-curvewiththe
et al. 2008), and b) a quiescent thermal spectrum
typicalshortX-rayburstsfrommagnetars. However,nei-
with the same N = 7 × 1021cm−2 and kT=0.3keV,
H ther a complete understanding of the physical processes
typical of a magnetar in quiescence (see i.e., Muno et
involved in these optical bursts nor other magnetar-like
al. 2008; Bernardini et al. 2009). We derived a
features (slow spin period, presence of magnetar-like X-
95% upper limit on the observed 0.3–10keV absorbed
ray bursts, etc.) are helping characterizing this source
(unabsorbed) flux of 6.2(9.6) × 10−16ergs−1cm−2 and
as a high magnetic field neutron stars. Furthermore, the
2.1(6.6) × 10−16ergs−1cm−2, under the power-law and
GRB-likeeventemittedbythissourceisatvariancewith
blackbody spectral assumption, respectively. Note that
any other flaring activity detected thus far from magne-
these spectral decompositions, and the derived flux lim-
tars (see Mereghetti 2008 for a review).
its, comprise also the typical values for quiescent X-ray
For an isolated neutron star, a luminosity limit of
binaries.
∼ 1030 −1031 ergs−1 would necessarily imply a source
older than 5×105 years, for any cooling model or equa-
4. DISCUSSION
tion of state (Lattimer & Prakash 2001; Yakovlev &
We present in this Letter a deep X-ray observation of
Pethick 2004). This age limit would make SwiftJ1955
SwiftJ1955 during the quiescent state. Before entering
two order of magnitudes older than the bulk of magne-
in a detailed discussion of our X-ray limit on the quies-
tars (Mereghetti 2008), in line only with the old low-B
cent emission of this optical bursting transient, we first
field SGR0418+5729 (Rea et al. 2010). Furthermore,
need to discuss the current estimates on the source dis-
the former value can be considered a lower limit on the
tance.
age, since for the cooling models for high magnetic field
Castro-Tirado et al. (2008) studied in detail the dis-
neutron stars (Pons, Miralles & Geppert 2009; Aguilera
tance issue, first taking 12CO and H I spectra, and then
et al. 2008, 2009), and the heat released by the recent
deriving the extinction versus distance distribution in
outburst, these will go in the direction of predicting a
the SwiftJ1955 line of sight based on the red clump
larger age. In particular, for a typical cooling neutron
method (see also Lopez-Corredoira et al. 2002; Durant
star, thepresenceofahighmagneticfieldcausesamuch
& van Kerkwijk 2006). From the mm and cm spectra,
brighter source at the same age.
they derived a Galactic column density in the direction
of this source of NH = NHI + 2NHII = 14.1 ± 2.0 × 9Notethatthisisderivedfromanabsorbedpower-lawfit(Kasli-
wal et al. 2008), hence it might be an overestimate on the NH of
8 http://heasarc.gsfc.nasa.gov/Tools/w3pimms.html thesource.
The quiescence of SwiftJ195509.6+261406 (GRB070610) 3
1e+36 1e+15
’M1.40_B1e13.daBt
’= u
1 011:32
G
 ’M1.40_B1e13.dat’ u 1:(B$
=4
*110e131
G2
)
’M1.40_B1e14.daBt
’= u
1 011:42
G
 ’M1.40_B1e14.dat’ u 1:(B$
=4
*110e141
G2
)
’M1.40_B1e15.daBt
’= u
1 011:52
G
 ’M1.40_B1e15.dat’ u 1:(B$
=4
*110e151
G2
)
1e+35 s)

/s)
 aus
g 1e+34 G 1e+14
y
(er eld
(
osit 1e+33 c
Fi
n &
mi 1e+32 ne 1e+13
u g
L a
M
1e+31
1e+30 1e+12
0 200000 400000 600000 800000 1e+06 1.2e+06 1.4e+06 1.6e+06 1.8e+06 0 200000 400000 600000 800000 1e+06 1.2e+06 1.4e+06 1.6e+06 1.8e+06
Time
(years)
 Time
(years)

Fig. 1.—Luminositydecay(leftpanel)andmagneticfielddecay(rightpanel)ofa1.4M(cid:12)neutronstarwiththreeinitial”magnetar-like”
magneticfields(Pons,Miralles&Geppert2009). Theorangeregionintheleftpanelcorrespondstotheluminositylimitswederivedasa
functionofthedistance(seetextfordetails),withthemaximumluminosityupperlimitmarkedasareddashedline(11.4×1030ergs−1).
Vertical red-dashed lines in the right panel report on the age at which the corresponding cooling magnetar would reach this maximum
luminosityupperlimit.
Assuming a neutron star with 1.4 M , we have used thecompactobjectshouldbe(closeto)fillingitsRoche-
(cid:12)
the cooling code of Pons, Miralles & Geppert (2009) Lobe. Assuming it fills its Roche-Lobe, and it is a main
to simulate the cooling decay of a magnetar with three sequence star, this gives a unique relation between the
different initial magnetic field values (see Fig.1 left orbital period of the system and the mass of the com-
panel). To reach our upper limit on the luminosity of panion star (see eq. 4.11 in Frank, King & Raine 2002).
L ∼ 1031ergs−1 for a magnetar-like magnetic field of Given the limits on the companion star, the system
x
B > 1014G at birth, the source should be now older orbital period is constrained to be shorter than 1.2hr
than 120 Myrs (see dashed line in Fig.1 left panel), and if the star is on the main sequence. Obviously, shorter
havingnowamagneticfieldofafew×1013, hencebelow periodsareallowedalsointheultra-compactbinarycase
the magnetar regime. On the other hand, the magnetic with an H-poor white dwarf. Another viable possibility
field and the strong internal helicity, supposed to pro- is a hot brown dwarf companion star, in which case the
duce short bursts and outburst activity in magnetars, orbital period of the system is constrained to be shorter
should have been largely dissipated at these old times than a few hours (Bildsten & Chakrabarty 2001).
and low field (see Fig.1 right panel). Taking at face X-ray observations of low mass X-ray binaries during
value the luminosity currently measured for typical qui- quiescencehaveempiricallyshownthatneutronstarand
escentmagnetars,theluminositywederiveisfainterthan black hole binaries, with the same orbital periods, show
the faintest magnetar in quiescence (SGR0418+5729: different quiescent luminosities (see e.g. Lasota 2008,
∼ 6 × 1031ergs−1; Rea et al. 2010). The possibility andFig.2). Thisobservationalevidencemighthaveafew
of SwiftJ1955 being a case similar to the low-B field interpretations. Onepossibilityisthatneutronstars’hot
SGR0418+5729 is intriguing, however the large GRB- surface makes them always brighter during quiescence
like flare detected from the former would be hardly ex- than a black hole, where supposedly an event horizon is
plainable within the scenario of an old magnetar releas- instead in place (Narayan, Garcia & McClintock 1997).
ing its last bit of internal magnetic energy through weak Another possibility might instead be the different accre-
sporadic bursts, as for SGR0418+5729. tionenergyreleasemechanism,withblackholesreleasing
Furthermore, unless the source is in the Galactic halo more energy through their radio jets rather than in the
at a distance of >10kpc, it is also dimmer than the lu- X-ray band (Fender, Gallo & Jonker 2003). In Fig.2 we
minosityof anyX-ray Dim IsolatedNeutronStarknown plot the quiescent X-ray luminosity of all binary neu-
to date (∼ 1031ergs−1; Turolla 2009), making the pos- tron stars and black holes for which this has been mea-
sible association of this object with any of those classes sured, and an orbital period or a limit on it could be
rather unlikely. derived (see also Garcia et al. 2001; Kong et al. 2002;
Jonker et al. 2006; Lasota 2008). The orange region in
Fig.2 is the quiescent luminosity space limit we derived
4.2. An X-ray binary system forSwiftJ1955 consideringthe3.7–10kpcdistancerange
and two different spectral models (see above). We also
A more plausible scenario is the X-ray binary nature
plot the luminosity limits considering the most plausible
of SwiftJ1955. In the binary case, optical and infrared
distance of 5kpc, and for the larger distance of 10kpc
observations during quiescence could put a limit on the
(dashed and dot-dashed red and grey horizontal lines,
companionmassof<0.12M (withaspectraltypelater
(cid:12) respectively).
than M5V), or being a semi-degenerate hydrogen poor
In the rest of the discussion we attempt to distinguish
star (Castro-Tirado et al. 2008). The occurrence of the
between the neutron star and black hole hypotheses.
X-ray outburst, implies that the low mass star orbiting
4 REA ET AL.
34
33
)
s
g/
r
e
min32
L
(
g
o
l
PL
@
10
kpc

31
BB
@
10
kpc

PL
@
5kpc

30 BB
@
5kpc

1 10 100
P (hr)
orb
Fig. 2.— Quiescent 0.5–10keV luminosity versus P for neutron star (stars) and black hole (circles) X-ray binaries (adapted from
orb
Lasota2008;Garciaetal.2001;Jonkeretal.2007;Degenaaretal. 2010;plustheadditionofarchivalX-rayobservations). Verticalarrows
represent upper limits on the X-ray quiescent emission. Horizontal dashed and dot-dashed lines represent our 95% upper limits on the
X-rayquiescentluminosityofSwiftJ1955 forapower-lawandablackbodyspectralmodels,respectively,andfora5and10kpcdistance
(seetextfordetails). Theorangeshadowedregionistheallowedluminosityspaceconsideringthewhole3.7–10kpcdistancerange,again
forthetwoassumedspectralmodels. Theverticallineat1.2hristheupperlimitontheorbitalperiodofSwiftJ1955 foramain-sequence
companionstar(seetextmoredetails).
4.2.1. A neutron star system 4.2.2. A black hole system
Large X-ray and optical flares, on several timescales,
The quiescent luminosity of a neutron star after ∼2 are an ubiquitous characteristic of black hole binaries.
yearsfromanoutburstisstronglydependentontheout- In particular, transient low-mass X-ray binaries hosting
bursthistory. Thelongertheoutburstactivitythelonger ablackholecandidateundergoverydramaticX-rayand
the cooling will take to reach the pre-outburst luminos- optical outbursts, and have long periods (even decades)
itylevel(Brown,Bildsten&Rutledge1998). Inourcase, of quiescence. However, sub-second timescale optical
there are no previous outbursts or bursts recorded from bursts such as in SwiftJ1955 (Stefanescu et al. 2008)
SwiftJ1955 inthepastyears,hencetheheatingdumped wereneverobservedbeforeinanyblackholebinary(nor
on the putative neutron star surface is very little. This any other astronomical source either). Optical flares on
small heating can explain the fast decrease in luminos- severaltimescalesdowntominuteswerereportedi.e. for
ity (by ∼7 orders of magnitudes) from the GRB070610 A0620–00 (Hynes et al. 2003a), XTEJ1118+480 (Hynes
eventtillAugust2009whenourdeepX-rayupperlimits et al. 2003b), GRS 1124–684 and Cen X-4 (Shahbaz et
are derived. al. 2010). A somewhat similar case might be GX339-
It is evident from Fig.2 that if at ∼5kpc, SwiftJ1955 4, the typical black hole candidate, which showed fast
wouldbeanultra-compactbinarysystem,beingtoofaint optical variability shorter than a second during a recent
in quiescence for a neutron star accreting from a main- outburst (Gandhi et al. 2010). Similarities can be found
sequence star or a brown dwarf. In this scenario this also with the X-ray nova V4641 Sgr (as suggested by
system might be similar to the ultra-compact binary Kasliwaletal. 2008),althoughinthiscasetherelatively
H1905+00 (Jonker et al. 2006, 2007; namely the up- massive companion star can introduce a somehow differ-
per limit reported on the bottom-left of the Fig.2). On ent physical process than in SwiftJ1955.
the other hand, if the putative neutron star is instead at Apart from the peculiar optical behavior, all the other
about 10kpc, then it might still be in orbit with a main observational characteristics observed from SwiftJ1955
sequence star or a brown dwarf, and the binary should areinlinewithwhatalreadyobservedfromblackholebi-
have an orbital period shorter than a few hours (Frank, naries: energeticX-rayflares, fastdecayintoquiescence,
King & Raine 2002; Bildsten & Chakrabarty 2001). optical QPOs at 0.16Hz, and extremely faint X-ray qui-
Thatsaid,itisimportanttonotethatneutronstarac- escence luminosities (see Fig.2).
cretingsystemshavenotbeenseenshowinglargeoptical Very interesting is the possibility of SwiftJ1955 being
flaresbesidetheopticalcounterparttoTypeIorIIbursts a black hole in an ultra-compact binary system, the first
which are orders of magnitude fainter and with longer ever discovered. Although very speculative, it might be
timescales than those observed in SwiftJ1955 (Kasliwal possiblethattheuniqueopticalbehaviorofthissourceis
et al. 2008; Stefanescu et al. 2008; Castro-Tirado et al. indeed reflecting the first of such systems, the emission
2008). of which is still largely unknown.
The quiescence of SwiftJ195509.6+261406 (GRB070610) 5
5. CONCLUSIONS flares, a peculiarity that SwiftJ1955 does not share yet
with any other source.
We derived deep upper limits with Chandra on the
X-ray quiescent emission of the optical bursting tran-
sient SwiftJ1955. We showed that a magnetar scenario
is very unlikely: the source is too faint in quiescence for NR is supported by a Ram´on y Cajal fellowship
any realistic scenario of magnetar cooling. We suggest through Consejo Superior de Investigaciones Cientf´ıcas,
that SwiftJ1955 is most likely an X-ray binary, host- bygrantsAYA2009-07391andSGR2009-811,andthanks
ing a black hole or a neutron star with an orbital period P. Casella for useful discussions on the optical variabil-
fasterthanafewhours,possiblyinanultra-compactsys- ity in black hole binaries. PGJ and GN acknowledges
tem. High-time resolution optical observations of X-ray support from a VIDI grant from the Netherlands Orga-
binaries during outburst might reveal energetic optical nization for Scientific Research.
REFERENCES
Aguilera,D.N.,Pons,J.A.,Miralles,J.A.2008,A&A486,255 Kong,A.etal.2002,ApJ570,277
Aguilera,D.N.,etal.2009,Ph.Rv.L.102,1101 Lattimer,J.M.&Prakash,M.,ApJ550,426
Bernardini,F.,etal.2009,A&A498,195 LasotaJ-P.2008,NewAR51,752
Bildstein,L.&Chakrabarty,D.2001,ApJ557,292 Liang,E.etal.2007,ApJ670,565
Brown,E.F.,Bildsten,L.,&Rutledge,R.E.1998,ApJ,504,L95 Lopez-Corredoira, M., Cabrera-Lavers, A., Garzon, F.,
Castro-Tirado,A.J.,etal.2008,Nature,455,506 Hammersley,P.L.2002,A&A394,883
Degenaar,N.,etal.2009,MNRAS396,L26 Markwardt,C.B.,etal.2008,ATel1102
Durant,M.&vanKerkwij,M.K.2006,ApJ650,1070 Mereghetti,S.,2008,A&AR15,225
Fender,R.,Gallo,E.&JonkerP.G.2003,MNRAS343,L99 Muno,M.P.,etal.2008,MNRAS,378,L44
Frank, King & Raine 2002, Cambridge University Press, Narayan,R.,Garcia,M.R.&McClintock,J.1997,ApJ478,L79
p.398,2002 Orosz,J.A.,etal.2001,ApJ555,489
Gandhi,P.etal.2010,MNRASinpress(arXiv:1005.4685) Pagani,C.etal.2007,GCN6489
Garciaetal.2001,ApJ,553,L47 Pons,J.A.,Miralles,J.&Geppert,U.2009,A&A,496
Gehrels,N.1986,ApJ303,336 Rea,N.etal2010,Science,330,944
Gehrels,N.,etal.2007,NJP,9,37 Shahbaz,T.,etal.2010,MNRAS403,2167
Hynes,R.I.,etal.2003a,MNRAS,340,447 Stefanescu,A.,etal.2008,Nature455,503
Hynes,R.I.,etal.2003b,MNRAS,345,292 Tueller,M.etal.2007,GCN6491
in’tZand,J.J.M.etal.2000,A&A357,520 Turolla,R.2009,ASSL357,141
Jonker,P.G.etal.2006,MNRAS,368,1803 Yakovlev,D.G.&Pethick,C.J.2004,AR&A42,169
Jonker,P.G.etal.2007,ApJ665,L147
Kasliwal,M.M.,etal.2008,ApJ,678,1127