MNRAS000,1–8(2016) Preprint11January2017 CompiledusingMNRASLATEXstylefilev3.0 Suzaku spectroscopy of the neutron star transient 4U 1608–52 during its outburst decay. M. Armas Padilla1,2,3(cid:63), Y. Ueda,3, T. Hori3, M. Shidatsu4 and T. Mun˜oz-Darias1,2 1Instituto de Astrof´ısica de Canarias (IAC), V´ıa La´ctea s/n, La Laguna 38205, S/C de Tenerife, Spain 2Departamento de Astrof´ısica, Universidad de La Laguna, La Laguna, E-38205, S/C de Tenerife, Spain 3Department of Astronomy, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan 4MAXI Team RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan 7 1 AcceptedXXX.ReceivedYYY;inoriginalformZZZ 0 2 n ABSTRACT a We test the proposed 3-component spectral model for neutron star low mass X-ray J binariesusingbroad-bandX-raydata.Wehaveanalysed4X-rayspectra(0.8–30keV) 0 obtained with Suzaku during the 2010 outburst of 4U 1608–52, which have allowed us 1 toperformacomprehensivespectralstudycoveringalltheclassicalspectralstates.We useathermallyComptonizedcontinuumcomponenttoaccountforthehardemission, ] E as well as two thermal components to constrain the accretion disc and neutron star H surfacecontributions.Wefindthattheproposedcombinationofmulticolordisc,single- temperature black body and Comptonization components successfully reproduces the . h data from soft to hard states. In the soft state, our study supports the neutron star p surface (or boundary layer) as the dominant source for the Comptonization seed pho- - tonsyieldingtheobservedweakhardemission,whileinthehardstatebothsolutions, o r either the disc or the neutron star surface, are equally favoured. The obtained spec- t tral parameters as well as the spectral/timing correlations are comparable to those s a observed in accreting black holes, which support the idea that black hole and neutron [ star low mass X-ray binaries undergo a similar state evolution during their accretion episodes. 1 v Key words: accretion, accretion discs – stars: individuals (4U 1608–52) – stars: 8 neutron star – X-rays: binaries 2 7 2 0 . 1 1 INTRODUCTION to distinguish between NS and BH accretors in LMXBs. A 0 definitive proof for a BH is obtained when the compact ob- Stellar-mass black holes (BHs) and numerous neutron stars 7 ject mass exceeds 3M(cid:12) (e.g. Casares & Jonker 2014), the (NSs) are revealed when they form part of low mass X-ray 1 maximumstablemassofaNSingeneralrelativity(Rhoades v: binaries(LMXBs),wheretheyareaccretingmaterialfroma & Ruffini 1974; Kalogera & Baym 1996). In absence of dy- i lowmass(< M(cid:12))companionstar.Accretiontakesplacevia namicalmeasurements,NSsareconfirmedbyeventsassoci- X an accretion disc (Shakura & Sunyaev 1973), whose inner ated with the presence of a solid surface and magnetic field regions are viscously heated up to ∼106−7 K, making these r (absent in BHs), such as thermonuclear X-ray burst and a sources the brightest in the night X-ray sky. Depending on coherent pulsations, respectively. From the X-ray point of the accretion rate, LMXBs can be found in two flavours, view, BH-LMXBs are simpler objects with two main spec- transientandpersistentsources.Theformerspendmostpart tralcomponents:(i)ahardcomponentbelievedtoarisefrom oftheirlivesinafaint,quiescentstateatX-rayluminosities inverse-Compton processes in an optically thin inner flow L ∼1030−32 erg s−1, but show occasional outburst – that X (corona) and typically modelled with a power-law, and (ii) can last several months – when brightness increases over asoft,thermalaccretiondisccomponent.Theevolvingbal- a million fold. Persistent systems are always active (i.e. in ancebetweenthemleadtothepresenceoftheso-calledhard outburst), with L >1036 erg s−1 (but see Armas Padilla X and soft accretion states, which alternate following well- et al. 2013). definedhysteresispatterns(e.g.Miyamotoetal.1992,1995) Given the rather comparable binary system parame- in the spectral and variability domains (Homan et al. 2001; tersandgravitationalwells,itisnotalwaysstraightforward Done et al. 2007; Mun˜oz-Darias et al. 2011; Belloni et al. 2011, for a review). (cid:63) e-mail:[email protected] InNS-LMXBs,theX-rayemissionfromthesolidsurface (cid:13)c 2016TheAuthors 2 M. Armas Padilla et al. and/or boundary layer (hereafter, we will refer this as NS hardstatediscsaredetectedinthesameBHswhenobserv- surfaceforsimplicity)isexpectedtocontributetotheX-ray ing with instruments providing lower energy coverage (e.g. spectrum, in addition to the aforementioned hard (comp- Plant et al. 2015; De Marco et al. 2015). tonizing corona) and soft (accretion disc) spectral compo- In this paper we take the LRH07 approach as the nents. This lead to a more complex spectral evolution and starting point to analyse 4 broad-band (0.8–30 keV) statephenomenology(Hasinger&vanderKlis1989;vander Suzaku observations covering the 2010 outburst decay of Klis 2006). Nevertheless, BH and NS-LMXBs are known to 4U 1608–52 from soft to hard states. This source, one of have very similar timing properties (e.g. Wijnands & Van thetwousedtotesttheaforementionedspectralmodel,isa Der Klis 1999). Recently, BH-like hysteresis paths (Mun˜oz- well-known transient NS-LMXB (Grindlay & Gursky 1976; Dariasetal.2014)havebeenfoundtobeacommonfeature Tananbaum et al. 1976), which recurrently undergoes into also in NS systems accreting at sub-Eddintong rates (a.k.a. outburst (e.g. Negoro et al. 2014). From radius-expansion atoll sources). type I X-ray burst the distance is estimated to be in the Since the dawn of X-ray astronomy many efforts have range 3.2–4.1 kpc (Nakamura et al. 1989; Galloway et al. been made to model the NS-LMXBs spectra. In the 80s, 2008), and the detection of rapid oscillation in some of two models using differing approaches were proposed, and these events indicates that the NS spins at 620 Hz (Muno dubbed Eastern (Mitsuda et al. 1984, 1989) and Western et al. 2001). 4U 1608–52 has been extensively studied, (White et al. 1988) models. Both include two components, with works published on its X-ray variability properties but they differ on their physical origin. On one hand, the (e.g. Yoshida et al. 1993; Altamirano et al. 2008; Barret Eastern model attributes the soft emission to the accre- 2013), burst episodes (e.g. Galloway et al. 2008; Poutanen tion disc and the hard component to Comptonization of et al. 2014), disc reflection features (e.g. Degenaar et al. seed photons emitted on the NS surface. In the Western 2015) and spectral properties (e.g. Gierlinski & Done 2002; model, on the other hand, the NS surface and the Comp- Asai et al. 2012, LRH07). The latest group, however, tonized disc are the sources of the soft and hard compo- was mostly carried out with data from instruments with nents, respectively (see Barret 2001 for a review). Along a coverage-lack at low energies, which results in strong the years, both approaches have been successfully applied. limitations when modelling the soft components; a caveat As an example, Done et al. (2002), fitted the same Cyg X- that can be accounted for by using Suzaku data. 2 data with both methods, obtaining statistically indistin- guishable results. Besides model degeneracy, it is not either clear whether any the above methods can be consistently (i.e. keeping the same physical meaning for every compo- 2 OBSERVATIONS AND DATA REDUCTION nent) describe the spectral evolution through the different stages (states) of an outburst. Moreover, given the overall 4U 1608–52 was observed with the spacial facility Suzaku similarity between LMXBs with NS and BH accretors, the (Mitsuda et al. 2007) during its 2010 outbourst. Four ob- largedatabasenowavailableforbothgroups,andthespec- servations were performed throughout the outburst decay, tral simplicity expected for the latter, a BH oriented ap- on 2010 March 11, 15, 18 and 22, for a total observing proach could potentially lead to more general conclusions. time of ∼114 ksec (see Figure 1). We used the heasoft Linetal.(2007,hereafterLRH07)testedthetwoabove v.6.18softwareandSuzaku CalibrationDatabase(CALDB) scenarios and end up proposing a hybrid, 3-component released on 2015 March 12 for the reduction and analysis model. These are, (i) a constrained broken power-law of our data. After reprocessing the data with the Suzaku (CBPL), (ii) a multi-color disc model(MCD) and (iii) a FTOOLaepipelineandremovingbursteventspresentinob- single-temperature blackbody (BB), which account for the servations404044020and404044030,wefollowedtheSuzaku Comptonization, the accretion disc and NS surface emis- ABC guide1 to obtain the final data products. sions, respectively. This solution is able to model the spec- The X-ray Imaging Spectrometer (XIS; Koyama et al. tra of two transient NS-LMXBs (see also Lin et al. 2009 2007) was operated in the 1/4 window mode and using the for an extension) at different states and luminosities, ful- burst option to avoid pile-up effects (see Table1 for the ob- filling physical and phenomenological constraints typically servation log). We used a circular region with a radius of observedinBHsystems(seeSection3.2).Likewise,bycom- 110(cid:48)(cid:48)centred on the system position to extract the source bining this modelling with the evolution of the fast tempo- events, and a circular region of the same size placed in a ralvariability,aconsistentBH-NSaccretionstatepictureis source-freepartoftheCCDstoextractthebackground.The found even for the brightest NS-LMXBs (a.k.a. Z sources). XIS response matrix and ancillary response files were cre- This includes similar accretion-outflow coupling properties atedusingxisrmfgenandxisarfgentasks.Wemadeuseof for both families (Mun˜oz-Darias et al. 2014; Fender & theaepileupcheckup.pyscriptdevelopedbyYamadaetal. Mun˜oz-Darias 2015, for a review). (2012) to evaluate the pile-up fraction in our observations. One of the limitations of the LRH07 study is the lack It returned an estimation of less than 2% at the core of the of coverage at energies <3 keV (they used RXTE data), PSFs. Nevertheless, to ensure that our spectral results are ∼ which hampered a detailed modelling of the thermal com- notaffectedbypile-up,weextractedthesourcespectrausing ponents (disc and NS surface). This is particularly trouble- several annular regions excluding photons coming from the some in the hard state, when these are weaker. As a mat- most central region. The spectra are consistent with each teroffact,LRH07foundthattheaccretiondisccomponent other confirming that the pile-up effect is negligible. We wasnotrequiredtofitthehardstateobservations,alimita- tioncommonlysufferedinBHsystemswhenusingthesame instrumentation (e.g., Mun˜oz-Darias et al. 2013). However, 1 http://heasarc.gsfc.nasa.gov/docs/suzaku/analysis/abc/ MNRAS000,1–8(2016) Broad band Suzaku spectroscopy of 4U 1608–52 3 Table 1.Suzaku observationslog. Observations Date XIS HXD-PIN (Obs) ID MJD yyyy-mm-dd Burstoption Netexposurea Netexposurea 1 (cid:70) 404044010 55266.07 2010-03-11 0.13sec 2.2ksec 28ksec 2 • 404044020 55270.69 2010-03-15 0.13sec 2.3ksec 25.1ksec 3 (cid:78) 404044030 55273.99 2010-03-18 0.5sec 7.8ksec 14.1ksec 4 (cid:4) 404044040 55277.98 2010-03-22 1.0sec 15.9ksec 14.5ksec aExposuretimepersensorafterdead-timeandbursteventssubstration. RXTE pointing of 4U 1608–52 during its ∼15 years of op- erations. Every RXTE observation is represented by a grey dot. Observations simultaneous with the Suzaku data are c s/se 1.5 highlighted using the same colour code as in Fig. 1. The nt RIDwasperformedfollowingMun˜oz-Dariasetal.(2011)by u co computing fractional root-mean-square (rms) in the 0.1–64 eV 1.0 Hzfrequencyband.IntheHID,Hardness wasdefinedasthe k 0 ratioofcountsbetweenthe10–16keVand6–10keVbands, 2 2- respectively.RXTEnetcount-rate(YaxisinFig.2)iscom- AXI 0.5 puted in the band 2–15 keV. Hard states are found at rms M >15%andhardness>0.6,whilstsoftstatesarecharacter- ∼ ∼ ized by rms <5 % and hardness <0.4 (see Mun˜oz-Darias ∼ ∼ et al. 2014, for further details). 55255 55265 55275 55285 Day [MJD] 3 ANALYSIS AND RESULTS Figure 1.The2010outburstlightcurve(2-20keV)of4U1608– 52 obtained by MAXI. The times of the 4 Suzaku observations Weusedxspec(v.12.9,Arnaud1996)toanalyseourspectra. are indicated by the coloured symbols. Observation 1, 2, 3, and We added a 1 per cent systematic error to every spectrum 4 are indicated by a red star, a red circle and a blue square, to account for calibration uncertainties (Makishima et al. respectively.Samesymbol/colourcodeisusedforallthefigures. 2008), and excluded both the 1.7-1.9 keV and 2.2-2.4 keV energy ranges to avoid the silicon K-edge and gold M-edge calibration uncertainties, respectively. In order to deal with combined spectra and response files of the two frontside- cross-calibration issues we added a constant factor (CON- illuminated cameras (FI-XISs; XIS-0 and XIS-3) with the STANT) to the spectral models with a value fixed to 1 for addascaspectasktomaximizethesignal-to-noiseratio.The XIS spectra and 1.16 for HXD spectra 2. We included the spectrafromthebackside-illuminatedcameraXIS-1wasex- photoelectric absorption component (PHABS) to account cluded from our study due to discrepancies (probably by for the interstellar absorption assuming the cross-sections inaccurate cross calibration) with the FI-XISs, which are of Balucinska-Church & McCammon (1992) and the abun- more sensitive at high energies. dances of Anders & Grevesse (1989). Finally, the 1-10 keV We extracted the spectra from PIN silicon diodes of FI-XISs and 15-30 keV HXD/PIN spectra were simultane- the Hard X-ray Detector (HXD; Takahashi et al. 2007) us- ouslyfittedbyusingtiedspectralparameters.Wenotethat ing the FTOOL hxdpinxbpi provided by the Suzaku team. HXD/PIN data above 30 keV were excluded since they do This script produces the background spectrum extracting not exceed the background level (at 3σ).3 the non-X-ray background (NXB) from the tunned back- In this paper we aim at modelling 4 Suzaku spectra ground modeled by the HXD team, and adding the contri- using the LRH07 3-component model. The broader energy butionofthecosmicX-raybackground(CXB)followingthe range(particularlyinthesoftband)ofourdataallowusto Boldt (1987) recipe. Due to the Galactic coordinates of the giveanstepforwardwithrespecttopreviousworks.Onone source (l=330.92◦, b=-0.85◦), we also account for contam- hand,anddifferentlyfromLRH07whouseda(constrained) ination caused by Galactic ridge X-ray emission following broken power-law to model the hard component, we utilize Revnivtsev et al. (2003, 2006). In our analysis we used the a more physically meaningful thermally Comptonized con- appropiate response file provided on the Suzaku CALDB. tinuum component (NTHCOMP in xspec; Zdziarski et al. ThesourcewasnotsignificantlydetectedwiththeGadolin- 1996; Zycki et al. 1999). On the other hand, we apply the ium Silicate (GSO) scintillator. very same 3-component model also to the hard state spec- trum(obs.4),contrastinglytothesimpler2-componentap- proach proposed in the hybrid model. Thus, we use the 2.1 Rossi X-ray Timing Explorer modelDISKBB+BBODYRAD+NTHCOMPinxspectofit The four Suzaku observations were (at least) partially cov- eredbysimultaneousRossiX-rayTimingExplorer (RXTE) 2 Suzaku Memo 2008-06 at http://www.astro.isas.jaxa.jp/ data. In Fig. 2 we present both the hardness-intensity suzaku/doc/suzakumemo and rms-intensity diagrams (HID and RID, respectively) 3 The40-70keVNXBhavea3σ systematicerrorof8.4-5.4%for from Mun˜oz-Darias et al. (2014).These include every (980) anexposureof10-20ksec,respectively(Fukazawaetal.2009) MNRAS000,1–8(2016) 4 M. Armas Padilla et al. Figure2.Hardnessintensitydiagram(leftpanel)andrmsintensitydiagram(rightpanel)of4U1608–52usingRXTEobservations(see text).DatasimultaneoustoourSuzaku campaignarehighlightedusingthesamecolourschemeasinFig.1.TheDashedlinerepresents the5%variabilitylevel,whichroughlydefinethesoftstatesinBHandNSLMXBs.AdaptedfromMun˜oz-Dariasetal.(2014). soft (obs. 1 and 2), intermediate (obs. 3) and hard (obs. 4) The soft-to-hard transition occurs at ∼0.02 L , which Edd state data of the same source taken during the same out- agreeswiththe0.01-0.04L rangeproposedbyMaccarone Edd burst. Results are reported in Table 2 and Fig. 3 with un- (2003) for this transition in BH and NS LMXBs. certainties given at 90 per cent confidence level. The disc and NS temperatures behave as expected for We have two possible seed-photon solutions for the the corresponding accretion rates, since they drop from 1 Comptonized emission. One assumes that the up-scattered to 0.2 keV and from 1.6 to 0.4 keV, respectively. From the photons arise from the disc and the second associates them black body normalization we infer an emission radius (R ) bb withtheNSsurface.Thus,wecoupledKT toeitherKT in the range of ∼1–6 km, which is consistent with previous seed disk or KT , respectively, and changed the seed photons shape works(e.g.,Sakuraietal.2014).ThesesmallR valuescan bb bb parameteraccordingly(im typeinNTHCOMP;seeTable2). be interpreted as resulting from emission from a boundary Both approaches yield statistically indistinguishable results layer taking the shape of an equatorial belt (LRH07 and (χ2<1.05).InTable2,wereportonlythecaseinwhichthe referencestherein;seealsoMatsuoka&Asai2013).Wenote, ν seed photons arise from the NS surface (see Section 4 for however, that it is not straightforward to obtain physically discussion).Ontheotherhand,giventhecomplexityofthe meaningfulR valuesfromthiskindofmodelling.Also,we bb modelling, uncertainties have been calculated in two steps, did not apply any correction for the photons scattered by that is, freezing the thermal components when calculating the Corona which might increase the obtained values by a the error associated with the hard tail (NTHCOMP) and factor <2 (see Kubota et al. 2004). vice versa. Finally, the (Comptonization) asymptotic power-law Wenotethat,incontrasttopreviousstudiesofthelow photon index (Γ) slightly decreases as the flux goes down, statespectrawhereareflectioncomponentwasrequired(e.g, from ∼2.3 to 2, whereas the electron corona temperature Yoshida et al. 1993 using Ginga data; Degenaar et al. 2015 (KT)increasesfrom3.3to18keV.Thesevaluescorrespond e using NuSTAR), our modelling do not require reflection re- to a decrease on the electron scattering optical depth (τ) latedfeatures.Asatest,weaddedtoourmodelsaGaussian from ∼ 8 to ∼3 4. The thermal components contribute by componentinordertosearchforthepresenceofaFe-Kline ∼77 per cent to the total luminosity in the soft state (∼54 withcentralenergyof6.4keVandawidthof0.1keV,consis- percentfromthediscand23percentfromtheNSsurface). tently with previous works (e.g. LRH07). We always found On the contrary, the thermal contribution is only ∼15 per upper limits as stringent as Fe-K equivalent width < 14eV. cent in the hard state . Therefore, the reflection features from a cold outer disk are either absent or buried by the continuum. 3.2 Some initial sanity tests 3.1 Spectral parameters LRH07 approached model degeneracy by setting a number Spectral fits yield a hydrogen column density (N ) of ∼ of physical evaluation criteria. These are: H 1×1022cm−2, which agrees with values previously reported 1. The inner disc radii have values comparable/higher than (Penninx et al. 1989; Gu¨ver et al. 2010). The inferred 0.8– the size of the NS. 30keVunabsorbedfluxgoesfrom13.8×10−9erg cm−2 s−1 in the soft state, to 0.6 ×10−9erg cm−2 s−1 in the hard state. Assuming a distance of 3.6 kpc, they translate into a lumi- 4 Photon index is related to kTe and optical depth (τ) through (cid:104) (cid:105)1/2 nositydropfrom21.5to0.9×1036erg s−1 (0.1to0.004L the relation Γ=α+1= 9+ 1 −1 (Sunyaev & assuming an Eddington luminosity of LEdd=2×1038erg s−E1d)d. Titarchuk1980;Lightman4&Z(kdTez/imaercs2k)τi(11+9τ8/37)) 2 MNRAS000,1–8(2016) Broad band Suzaku spectroscopy of 4U 1608–52 5 2. The thermal components (MCD and BB) roughly evolve (DISK+NTHCOMP) are able to reproduce the data. Even as L ∝T4. if degeneracy is obviously an issue, we note that an F-test X 3. The disc temperature is lower than the NS surface tem- indicatesthatthe3-componentmodel(χ2of1445with1400 perature(KT <KT ;Mitsudaetal.1989;Pophametal. dof.) is significantly better. In addition, the obtained spec- diskbb bb 2001); and therefore the seed photons temperature should tralparametersfollowthesametrendthanfortheprevious not exceed the NS surface temperature (KT (cid:54)KT ). 3observations,suggestingthatthismodellingprovidesare- seed bb 4. The Comptonization fraction is consistent with the alistic description of the spectral evolution of the system. power density spectrum, assuming that in NS-LMXBs the X-ray variability changes as a function of spectral hardness similarly to the BHs systems. 4 DISCUSSION In order to test the first condition we converted the Model degeneration is a fundamental problem when study- normalizationofthedisccomponent(N )intoinnerdisk ingthespectralevolutiononNS-LMXBs.Here,wetakead- diskbb radius (R ) following the relation: vantage of the broad-spectral coverage and good spectral in resolution of Suzaku to test the 3-component model pro- posed by LRH07 as part of the hybrid model. To this end, N D Rin=ξκ2(codsiski)0.510 kpc [km] (1) we have used 4 observations of the NS-LMXB 4U 1608–52 taken during its 2010 outburst decay, covering the soft-to- (Kubota et al. 1998; Gierlinski & Done 2002). D is the hard transition (two in the soft state [SS], one in the inter- source distance, i is the disc inclination angle, κ is the ra- mediate state [IS] and one in the hard state [HS]). tio of color temperature to effective temperature (Shimura Our3-componentmodelcombinesMCD,BBandNTH- & Takahara 1995), and ξ is the correction factor for the COMP, which account for the emission from the accretion inner torque-free boundary condition (Kubota et al. 1998), disc and NS surface, and the Comptonization contribution, although we note that the torque-free boundary condition respectively. Using this model we have successfully fit all is most likely not fulfilled in NS-LMXBs. Here we adopt four observations, obtaining spectral parameters which are κ=1.7,ξ=0.4andi=70◦.Theobtainedinnerdiscradiiare within typical ranges from previous studies on 4U 1608–52 in the range Rin∼15–45 km. These values become slightly (e.g.Gierlinski&Done2002;Linetal.2007;Degenaaretal. smaller (Rin ∼11–30 km) if i=40◦(Degenaar et al. 2015), 2015)andotherNSLMXBs(e.g.Doneetal.2002;Linetal. butareinanycaselargerthantheexpectedNSradius(e.g. 2009; Takahashi et al. 2011; Sakurai et al. 2012). Lattimer & Prakash 2000; Zdunik & Haensel 2012). These The evolution of each component through the differ- valuesareconsistentwiththoseobtainedbyDegenaaretal. ent accretion states could be interpreted according to the (2015),whichuseareflectionmodeltofita4U1608–52NuS- following picture: In the soft state, most of the emission TARobservationat∼0.02LEdd(i.e.similartothoseconsider is produced by the the NS surface and the accretion disc here). (which probably extends close to the NS surface) and only Condition 3 is also fulfilled (see Table 2), while Condi- a little amount of radiation (∼20%) is produced by Comp- tion 4 is discussed in Section 4 since it has important im- tonization of seed photons emitted by these components. plications on the nature of the Comptonized emission. Fi- As accretion drops, the disc temperature and its associated nally, Condition 2 only makes sense if the inner disc radius luminosity decrease, which might be a consequence of disc isconstantwithtime.Weobservethanboththeblack-body truncation. Following this drop in accretion rate, there is and the disc components roughly evolved as LX∝T4 (not a subsequent decrease on the NS surface temperature and showed),whichisconsistentwiththerathersmallvariation luminosity. Thus, the Comptonization emission starts to be measuredintheinnerradius.Nevertheless,thereadershould dominant,andeventuallybecomesthemainsourceofemis- bearinmindthatourdiscradiusmeasurementsarederived sion in the hard state. This simplistic picture is similar to from a Newtonian model (DISKBB) and should be taken what is seeing in BH-LMXBs, with the difference that for with caution when used for more complex calculations. NS systems an additional soft component is required to ac- countfortheNSsurface.Thepresenceofthisextrathermal radiation should have an important role in producing the 3.3 Hard state observation Comptonized emission. Wehavesuccessfullyfittedthehardstateobservation(Obs. 4) with the 3-component model. We tried also the simpler 4.1 The nature of the Comptonized emission 2-component approach, which includes the Comptonization component and only one of the thermal emissions (either So far we have discussed results obtained by assuming that MCD or BB). We find that the DISK+NTHCOMP model themainsourceofseedphotonsfortheComptonizationpro- fits adequately the data with χ2= 1487 and 1403 dof . The cess is the NS surface (or boundary layer; i.e. we coupled correspondingspectralparametersareshowninthelastcol- KT withKT ).However,goodfitsarealsoobtainedwhen seed bb umnofTable2andthefititself(reddashedline)inthelast assuming that the disc is the main source of seed photons panel of Fig. 3. On the other hand, the BB+NTHCOMP (fifthcolumninTable2).Indeed,theobtainedtemperatures model results in χ2of 1507 and soft residuals are observed. andphotonindicesareintherangeexpectedforNS-LMXBs. Statisticallyspeaking,theformerapproachisconsistentwith Fig.4showstheevolutionofthefractionalrmsasfunctionof thenullhypothesisat5percentsignificancelevel(p=0.057), the thermal luminosity fraction for both possible seed pho- whileitisrejectedbythelatterone.Therefore,wefindthat tonsolutions.Whencomparingthetwoapproaches,thefirst both the 3-component model and a simplified version of it striking thing is that the thermal fraction drops to ∼40 per MNRAS000,1–8(2016) 6 M. Armas Padilla et al. Table 2. Fitting results for the DISKBB+BBODYRAD+NTHCOMP model for the 4 observations assuming that seed-photons arise from the NS surface (KT tied to KT ). The fifth column shows the results for obs.4 with the same model but assuming that seed- seed bb photonsarisefromthedisc(KT tiedtoKT ).Lastcolumnshowstheresultsforobs.4usingthesimplerDISKBB+NTHCOMPmodel seed in (KT tiedtoKT ).BothspectraandmodelsareshowninFig.3.Uncertaintiesareexpressedat90percentconfidencelevel. seed in Component Obs1(SS) Obs2(SS) Obs3(IS) Obs4(HS) Obs4(HS)alternativemodels seedphotons(diskbb) (diskbb+nthcomp) NH (×1022cm−2) 1.07±0.01 1.05±0.01 1.01±0.02 0.99±0.15 1.1±0.1 1(fix) KT (keV) 1.00±0.006 0.92±0.006 0.621±0.004 0.23±0.02 0.15±0.01 0.40±0.02 in N 448±10 386±10 623±17 4160+2046 53694+70658 140+27 diskbb −1346 −25350 −13 KT (keV) 1.63±0.004 1.70±0.006 1.25±0.005 0.395±0.001 0.42±0.01 – bb N 40±1.2 21.2±0.7 14.7±0.6 277±27 143±22 – bb Γ/τ a 2.26±0.06/7.69 2.18±0.1/7.8 2.14±0.03/3.53 1.98±0.01/3.23 1.98±0.01/3.15 2.03±0.02 KTe (keV) 3.31±0.09 3.2±0.1 14.5−+26..95 18+−144 19+−154 999+−259 imp type 0 0 0 0 1 1 Nnthcomp (×10−2) 3.16±0.06 1.44±0.04 1.71±0.02 5.11±0.04 12.3±0.1 9.8±0.7 FXb (ergcm−2 s−1) 14.20±0.06 8.33±0.05 3.09±0.03 0.76±0.02 0.83±0.04 0.76±0.02 LXc (ergs−1) 2.20±0.01 1.29±0.01 0.479±0.005 0.118±0.003 0.128±0.006 0.118±0.003 MCDfraction 54 57 41 7 3 7 BBfraction 23 23 12 8 5 – NTHCOMPfraction 23 20 47 85 92 93 χ2 /dof 0.96/1859 1.02/1324 1.03/1370 1.03/1400 1.03/1400 1.06/1403 ν (cid:104) (cid:105)1/2 aTheelectronscatteringopticaldepth(τ)isobtainedfollowingtherelationΓ= 9+ 1 −1 4 (kTe/mec2)τ(1+τ/3) 2 bUnabsorbed0.8-30keVfluxinunitsof10−9ergcm−2 s−1. c0.8-30keVluminosityinunitsof1037ergs−1 assumingadistanceofD=3.6kpc. s) Obs.1 Obs.2 2m c 1 V/ e k s n 0.1 o ot h P E) ( 0.01 f( 2E o 1.1 ati 1.0 R 0.9 s) Obs.3 Obs.4 2m c 1 V/ e k s n 0.1 o ot h P E) ( 0.01 f( 2E o 1.1 ati 1.0 R 0.9 0.8 1 2 5 10 20 30 0.8 1 2 5 10 20 30 Energy [keV] Energy [keV] Figure 3. Unfolded spectra and data-to-model ratio using the DISKBB+BBODYRAD+NTHCOMP model. The full fit is shown as asolidgreyline,thedisccomponentasadot-dashedline,theblackbodycomponentasadottedlineandtheComptonizedcomponent asadashedline.Inthelastpaneltheobs.4resultsobtainedbyusingthesimplerDISKBB+NTHCOMPmodelareoverplottedinred. XISdataarere-binnedinXSPECforclarity. MNRAS000,1–8(2016) Broad band Suzaku spectroscopy of 4U 1608–52 7 Finally, we note that we do not see a strong evolution in our photon indices, which decrease from 2.3 to 2. While theformervalueisconsistentwithBHsoftstates(Γ∼2–3), %) 15.0 values Γ∼<2 are more typical for the hard state. We note, S ( however,thatourhardstateobservationisoneofthesoftest M of 4U 1608–52 in this state, as significantly larger rms and R al 10.0 hardness values have been observed (see Fig. 2). n o cti a r F 5.0 5 CONCLUSIONS Wehaveperformedadetailedstudyofthespectralevolution of the neutron star X-ray transient 4U 1608–52 during the 0.0 0.2 0.4 0.6 0.8 1.0 decay of its 2010 outburst. Our 0.8–30 keV Suzaku obser- Fraction of thermal luminosity vations covered soft, intermediate and hard states epochs. We find that the 3-component model provides an excellent Figure 4. RMS fraction versus fraction of thermal luminos- description of the outburst evolution. The inferred spectral ity (DISKBB+BB). Filled and empty symbols correspond to 3- parametersevolveinagreementwithexpectations,fulfilling component model solutions assuming that the seed photons for severalphysicalcriteria,andreportingvaluesconsistentwith theComptonizedemissionarisefromtheNSsurfaceandtheac- thoseseeninotherNSsourcesandinBHtransients,whose cretiondisc,respectively. spectralevolutionisthoughttobelesscomplex.Inthehard state, where the thermal emission is weaker and less ener- getic, the 3-component model is statistically preferred over cent in obs. 1, versus ∼80 per cent if the NS surface solu- simpler solutions. Likewise, we find that the disc compo- tion is adopted. The former value is much lower than that nent, typically observed in BHs when low energy coverage observed in BHs at the same variability levels (note that is available, is in any case required. Future works including variabilityisgenerallyobservedtotracetheComptonization moreobservations(e.g.atdifferentstagesofthehardstate) fraction; e.g. Remillard & Mcclintock 2006). Moreover, the and a variety of sources should be able to further extend valuesobtainedforthesoftstatesobservations(obs.1and2) this study. To this end, we note that high-quality coverage differ significantly from each other when the disc approach at low energies is a desired feature, especially at low fluxes. is used. This is at odds with the similar hardness and rms values measured, both quantities being model independent andtypicallyusedtotracktheevolutionoftheaccretionflow (Fig. 2). Nevertheless, we note that, as usual, the complex- ACKNOWLEDGEMENTS ity of the nature exceeds any spectral fitting and probably MAP was funded by the International Research Fellowship both NS surface and disc photons are Comptonized. Thus, program of the Japan Society for the Promotion of Sci- it is important to bear in mind that our model is only able ence (PE15024). MAP and TMD acknowledge the hospi- to favour one components as the dominant source of seed tality of the Kyoto University, where part of this work was photons.Ontheotherhand,fromFig.4wecannotruleout carried out. MAP acknowledge the hospitality on her visit thediscasthemainsourceofseedphotonsforobs.3and4. toRIKEN,andthanksthevaluablediscussionwithK.Mak- Indeed, our spectral modelling with 2-components strongly ishima, T. Mihara and the other members of the MAXI suggestthat,ifanything,thedisccomponentmightprovide team.MAP’sresearchisfundedundertheJuandelaCierva the most relevant thermal contribution in the hard state. FellowshipProgrammeoftheMinistryofScienceandInno- vation (MINECO) of Spain. MAP and TMD acknowledge support by the Spanish MINECO grant AYA2013-42627. 4.2 Black holes comparative MS acknowledges support by the Special Postdoctoral Re- One of the main advantages of the 3-component model is searchers Program at RIKEN. This research has made use the straightforward comparison that it provides with BH- of data obtained from the Suzaku satellite, a collaborative LMXBs, which show only one thermal component (i.e. the mission between the space agencies of Japan (JAXA) and disc).Oursoftstateandintermediatetemperaturesarecon- the USA (NASA). sistent with values typically observed in BH systems (e.g., Mun˜oz-Darias et al. 2013) in agreement with findings by LRH07. However, the big advantage of our Suzaku-based REFERENCES studyistheabilitytostudythethermalcomponentsinthe hardstate,whenthesearemuchweaker.Plantetal.(2015) AltamiranoD.,vanderKlisM.,M´endezM.,JonkerP.G.,Klein- WoltM.,LewinW.H.G.,2008,ApJ,685,436 modelled hard state observations of GX 339-4, a BH tran- AndersE.,GrevesseN.,1989,GeochimicaetCosmochimicaActa, sient that is seen at an intermediate orbital inclination, as 53,197 it is probably the case of 4U 1608–52. They find KT in diskbb Armas Padilla M., Degenaar N., Wijnands R., 2013, MNRAS, therange0.15–0.28inremarkableagreementwithourwork 434,1586 (KTdiskbb= 0.23 ±0.02). In the same way, and using Suzaku ArnaudK.,1996,inJacobyG.,BarnesJ.,eds,AstronomicalSo- observations, Shidatsu et al. (2014) find a similar range of ciety of the Pacific Conference Series Vol. 101, Astronomical disc temperatures in the hard state of H1743–322. DataAnalysisSoftwareandSystemsV.p.17 MNRAS000,1–8(2016) 8 M. Armas Padilla et al. AsaiK.,etal.,2012,PASJ,64,128 RevnivtsevM.,GilfanovM.,SunyaevR.,JahodaK.,Markwardt Balucinska-ChurchM.,McCammonD.,1992,ApJ,400,699 C.,2003,A&A,411,329 BarretD.,2001,Adv.Sp.Res.,28,307 Revnivtsev M., Sazonov S., Gilfanov M., Churazov E., Sunyaev BarretD.,2013,ApJ,770,9 R.,2006,A&A,452,169 Belloni T. M., Motta S. E., Mun˜oz-Darias T., 2011, Bulletin of RhoadesC.,RuffiniR.,1974,Phys.Rev.Lett.,32,324 theAstronmicalSocietyofIndia,39,409 Sakurai S., Yamada S., Torii S., Noda H., Nakazawa K., Mak- BoldtE.,1987,ObservationalCosmology,124,611 ishimaK.,TakahashiH.,2012,PASJ,64,72 CasaresJ.,JonkerP.G.,2014,SpaceSci.Rev.,183,223 SakuraiS.,etal.,2014,PASJ,66,10 DeMarcoB.,PontiG.,Mun˜oz-DariasT.,NandraK.,2015,MN- ShakuraN.I.,SunyaevR.A.,1973,A&A,24,337 RAS,454,2360 ShidatsuM.,etal.,2014,ApJ,789,100 DegenaarN.,MillerJ.M.,ChakrabartyD.,HarrisonF.A.,Kara ShimuraT.,TakaharaF.,1995,ApJ,445,780 E.,FabianA.C.,2015,MNRASLett,451,L85 SunyaevR.A.,TitarchukL.G.,1980,A&A,86,121 DoneC.,ZyckiP.T.,SmithD.A.,2002,MNRAS,331,453 TakahashiT.,etal.,2007,PASJ,59,S35 Done C., Gierlinski M., Kubota A., 2007, The Astronomy and TakahashiH.,SakuraiS.,MakishimaK.,2011,ApJ,738 AstrophysicsReview,15,1 Tananbaum H., Chaisson L. J., Forman W., Jones C., Matilsky T.A.,1976,ApJ,209,L125 FenderR.,Mun˜oz-DariasT.,2015,Lect.NotesPhys.,905,65 WhiteN.E.,StellaL.,ParmarA.N.,1988,ApJ,324,363 FukazawaY.,etal.,2009,PASJ,61,S17 WijnandsR.,VanDerKlisM.,1999,Astron.Astrophys,345,35 Galloway D. K., Muno M. P., Hartman J. M., Psaltis D., YamadaS.,etal.,2012,PASJ,64,53 ChakrabartyD.,2008,ApJS,179,360 Yoshida K., Mitsuda K., Ebisawa K., Ueda Y., Fujimoto R., GierlinskiM.,DoneC.,2002,MNRAS,337,1373 YaqoobT.,DoneC.,1993,PASJ,45,605 GrindlayJ.,GurskyH.,1976,ApJ,209,L61 Gu¨verT.,O¨zelF.,Cabrera-LaversA.,WroblewskiP.,2010,ApJ, ZdunikJ.L.,HaenselP.,2012,A&A,551 ZdziarskiA.,JohnsonW.,MagdziarzP.,1996,MNRAS,283,193 712,964 ZyckiP.T.,DoneC.,SmithD.A.,1999,MNRAS,309,561 HasingerG.,vanderKlisM.,1989,A&A,225,79 vanderKlisM.,2006,In:CompactstellarX-raysources.Edited HomanJ.,WijnandsR.,vanderKlisM.,BelloniT.,vanParadijs byWalterLewin&amp;MichielvanderKlis.CambridgeAs- J., Klein-Wolt M., Fender R., M´endez M., 2001, ApJS, 132, trophysicsSeries 377 KalogeraV.,BaymG.,1996,ApJ,470,L61 KoyamaK.,etal.,2007,PASJ,59,S23 ThispaperhasbeentypesetfromaTEX/LATEXfilepreparedby KubotaA.,TanakaY.,MakishimaK.,UedaY.,DotaniT.,Inoue theauthor. H.,YamaokaK.,1998,PASJ,50,667 KubotaA.,etal.,2004,ApJ,601,428 LattimerJ.M.,PrakashM.,2000,ApJ,550,426 LightmanA.P.,ZdziarskiA.A.,1987,ApJ,319,643 LinD.,RemillardR.A.,HomanJ.,2007,ApJ,667,1073 LinD.,RemillardR.A.,HomanJ.,2009,ApJ,696,1257 MaccaroneT.J.,2003,A&A,409,697 MakishimaK.,etal.,2008,PASJ,60,585 MatsuokaM.,AsaiK.,2013,PASJ,65,26 MitsudaK.,etal.,1984,PASJ,36,741 MitsudaK.,InoueH.,NakamuraN.,TanakaY.,1989,PASJ,41, 97 MitsudaK.,etal.,2007,PASJ,59,S1 Miyamoto S., Kitamoto S., Iga S., Negoro H., Terada K., 1992, ApJ,391,L21 MiyamotoS.,KitamotoS.,HayashidaK.,EgoshiW.,1995,ApJ, 442,L13 Muno M. P., Chakrabarty D., Galloway D. K., Savov P., 2001, ApJ,553,L157 Mun˜oz-Darias T., Motta S., Belloni T. M., 2011, MNRAS, 410, 679 Mun˜oz-DariasT.,CoriatM.,PlantD.S.,PontiG.,FenderR.P., DunnR.J.H.,2013,MNRAS,432,1330 Mun˜oz-DariasT.,FenderR.P.,MottaS.E.,BelloniT.M.,2014, MNRAS,443,3270 NakamuraN.,DotaniT.,InoueH.,MitsudaK.,TanakaY.,Mat- suokaM.,1989,PASJ,41,617 NegoroH.,etal.,2014,Astron.Telegram PenninxW.,DamenE.,vanParadijsJ.,TanJ.,LewinW.H.G., 1989,A&A,208,146 PlantD.S.,FenderR.P.,PontiG.,Mun˜oz-DariasT.,CoriatM., 2015,A&A,573,A120 PophamR.,etal.,2001,ApJ,547,355 PoutanenJ.,NattilaJ.,KajavaJ.J.E.,LatvalaO.-M.,Galloway D. K., Kuulkers E., Suleimanov V. F., 2014, MNRAS, 442, 3777 RemillardR.A.,McclintockJ.E.,2006,ARA&A,44,49 MNRAS000,1–8(2016)