FLUX AND SPECTRAL VARIATIONS OF 1E1740.7−2942 OVER THE YEARS 2003−2012 5 1 0 2 n a Manuel Castro∗, Flavio D’Amico, Francisco Jablonski, João Braga J InstitutoNacionaldePesquisasEspaciais(INPE),SãoJosédosCampos,SãoPaulo,Brazil 4 1 E-mail: [email protected] ] E Theblackholesystem1E1740.7–2942isusuallythebrightesthardX-raysource(above20keV) H near the Galactic Center, but presents some epochs of low emission (below the INTEGRAL . h detectionlimit,forexample). Inthiswork,wepresenttheresultsofstudieson1E1740.7–2942 p - over 10 years, using the instruments ISGRI/IBIS and JEM-X, both on board the INTEGRAL o observatory. WefitthespectrawithboththecompTTandcutoffplmodels. Accordingtothefits r t s andtakingthemeanvalueoverthe10years,wehaveobtainedaplasmatemperatureintherange a [ ∼20–90keV, and an average powerlaw index of 1.41 (σ=0.25). We have also made a Lomb– Scargle periodogram of the flux in the 50–20keV band and found two tentative periods at 2.90 1 v and3.99days. Wepresentherethepreliminaryresultsofthisongoingwork. 4 0 4 3 0 . 1 0 5 1 : v i X r a 10thINTEGRALWorkshop: "ASynergisticViewoftheHighEnergySky"-Integral2014, 15-19September2014 Annapolis,MD,USA ∗Speaker. (cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ 1E1740.7−2942withINTEGRAL ManuelCastro 1. Introduction The source 1E1740.7–2942 is a black hole candidate nearby the center of our Galaxy. It was discovered by the Einstein satellite [1] and it was thoroughly studied by several high energy missions like SIGMA/GRANAT, INTEGRAL, Suzaku and Chandra [2, 3, 4, 5]. It was also the subjectofarecentstudybyourgroup[6],wherewemodelledthesoft–to–hardX-rayspectrumas dueto(thermal)Comptonization. TheINTEGRALsatellite[7]hasalreadyprovidedanuniformobservationdatabaseof1E1740.7– 2942. Thisdatabaseissuitableforastudyofanypossiblelong–termvariabilityofthesource. One interestinggoalofsuchastudyisthesearchformodulationsintheX-raylightcurvetorevealany possible orbital period. Reported orbital periods for 1E1740.7–2942 [8, 9] are still tentative, and a more precise determination of this period would be very important in the search for a (longer wavelength)counterpartof1E1740.7–2942,achallengingtaskthatremainstobeaccomplished. In this work we show the results of an ongoing effort to determine the orbital period of 1E1740.7–2942 (if there is one). We have made use of the INTEGRAL database consisting of all of the 1E1740.7–2942 observations from 2003 up to 2012. We have found two marginally significanttentativeperiodsat∼2and∼4days. Inthenextsectionswedescribeourdatabase,our datareductionandanalysis,anddiscussourresults. 2. DataReductionandAnalysis From the public archive of INTEGRAL, we downloaded all of 1E1740.7–2942 observations starting at 2003 up to 2012. 1E1740.7–2942 spends most of its time in the so-called canonical low–hard state (LHS), where the fraction of the total flux is higher in the hard X-ray band than in the soft X-ray band. Accordingly, the source flux is below the sensitivity of JEM-X detectors in their energy range, being preferably monitored in the band covered by the IBIS telescope. As a result, Figure(1) shows how our retrieved 314 observations database is distributed between the 3 INTEGRALimaging(unit)telescopes. Figure1:Ourdatabaseof314retrievedobservationsof1E1740.7–2942. Inthevastmajorityofcases(212) the source was imaged by IBIS alone. In 14 occasions the source was imaged by the 2 JEM-X units plus IBIS.In54withIBISplusonlytheJEM-X1unitand34withIBISplusJEM-X2unit. Spectra were extracted for each individual revolution using an automatic extraction script basedontheOSA10.0software. InsomecasesthespectraarebuiltonlywithIBISdata,covering 2 1E1740.7−2942withINTEGRAL ManuelCastro the20–200keVenergyband. WhenthesourcewasalsoimagedbyJEM-X,aspectrumfrom5–200 keV could be extracted with the aid of our automatic script. After spectrum extraction, data were fitted also automatically with the aid of a Tcl script. For this preliminary study, we fitted the 20– 200orthe5–200keVspectra(whenpresent)withtwomodels. ThefirstoneisasimplecutoffPL, suitable,forinstance,todetermineinwhichstatethesourcewasobserved[10]. Thesecondmodel is compTT, following the procedures of our previous study of the source [6]. In 14 occasions, as is shown in Figure (1), the spectrum extraction script needs to run three times (first run is always for IBIS), since both JEM-X units provided data. After fitting the data, our Tcl script saves the fitparameters(witherrorestimates)andcalculatedfluxes. Webuiltourdatabasewiththiskindof datafiles. Inthisfirststageofourstudy,forsimplicity,errorsinthefluxmeasurementsareassumed tobewithin±5%. 3. Results The1E1740.7–2942fluxhistoryofourdatabaseisshowninFigure(2). InFigure(3)weshow thecorrelationbetweenthefluxesinthetwohighestenergybandsofthisstudy. Thehistogramdis- tributionsforthe4bandsweadoptedinthisstudyarepresentedinFigure(4). Thepowerlawindex andplasmatemperaturedataforourdatabasearepresentedinFigures(5)and(6),respectively. We have also produced a Lomb-Scargle periodogram using the R software environment [11]. ThisisshowninFigure(7). 4. Discussion Our results have confirmed, as it is widely known, that 1E1740.7–2942 spends most of its timeinitscanonicalLHS,asderivedfromthepowerlawindexofafitwithcutoffPL(seeFigure5). Spectralfittinginourdatabasefrom20to200keVor5to200keV(whenavailable)usingcompTT models,yieldsanaveragetemperatureoftheelectronicplasma,which(thermally)Comptonizesthe soft seed photons, of ∼50keV (see Figure6). The ∼300 points in our database were helpful to better constrain this value. It must be emphasized, however, that at this point of our preliminary studythedataforderivingΓandkT arebasedonverylowfluxes(seeFigure2)sotheuncertainties e aresomewhathigh. Asmentionedabove,oneofthemaingoalsinthisworkistosearchforpossiblebinaryperiods. Ourdatabaseisoneofthebestforsuchpurposes,since1E1740.7–2942spendsmostofitstimein the LHS, i.e., the flux of 1E1740.7–2942 is higher in the hard X-ray band (E>20keV). Previous searchesuptodatewherelimitedbydatabasesinsofterbands. INTEGRALandthetelescopeswe used in this study, JEM-X and IBIS, are also imaging instruments and consequently 1E1740.7– 2942 flux is not affected by source confusion. Previous studies of searches for 1E1740.7–2942 periodshavemadeuse(mostly)ofdatacollectedbynon-imaginginstrumentsandatthesoftX-ray band(E<20keV). Our results, which can be seen in (Figure7), show the presence of periods at approximately 2.90 and 3.99days. Even though relatively well above local noise level in the periodogram, the second period must be taken with caution since it is quite close to a multiple of the one-day sampling period used. In addition, the 2.90-day period produces a folded light curve that is not 3 1E1740.7−2942withINTEGRAL ManuelCastro 111111000000 111111EEEEEE 111111777777444444000000......777777−−−−−−222222999999444444222222 IIIIIInnnnnntttttteeeeeeggggggrrrrrraaaaaallllll//////JJJJJJeeeeeemmmmmm−−−−−−XXXXXX ddddddaaaaaattttttaaaaaa 999999 bbbbbblllllluuuuuueeeeee iiiiiissssss 555555−−−−−−222222000000 kkkkkkeeeeeeVVVVVV aaaaaannnnnndddddd mmmmmmaaaaaaggggggeeeeeennnnnnttttttaaaaaa iiiiiissssss 111111000000−−−−−−222222000000 kkkkkkeeeeeeVVVVVV 888888 777777 666666 555555 1111 444444 −−−− ssss 333333 2222 222222 −−−− mmmm 1177 g cg cg cg c 1155 1r1reeEEdd 11iiss77 442200 00..−−7755−−0022 99kk44ee 22VV IIaannnntteeddgg gg rraarreell//eeIIBBnnII SSiiss aa55ll00oo−−nn22ee00 dd00 aa kkttaaeeVV rrrr 1133 eeee 1111 0000 1111 99 −−−− 0000 77 1111 55 n n n n 33 i i i i xxxx 11 uuuu FlFlFlFl 11117777 11115555 1111EEEE 1111777744440000 ....7777−−−−222299994444 2222 IIIInnnntttteeeegggg rrrraaaallll ddddaaaattttaaaa ccccoooolllloooorrrr sssscccchhhheeeemmmmeeee ffffoooolllllllloooowwwwssss pppprrrreeeevvvviiiioooouuuussss pppplllloooottttssss 11113333 11111111 9999 7777 5555 3333 1111 5555 2222 5555 4444 5555 6666 5555 8888 5555 11110000 5555 11112222 5555 11114444 5555 11116666 5555 11118888 5555 22220000 5555 22222222 5555 22224444 5555 22226666 5555 22228888 5555 33330000 5555 OOOObbbbsssseeeerrrrvvvvaaaattttiiiioooonnnn nnnnuuuummmmbbbbeeeerrrr Figure2: Thefluxhistoryof1E1740.7–2942basedonourdatabase. very clean, which introduces uncertainties that we are still analysing. It must also be emphasized thatmuchmoresensitivestudiesusingRXTE[9]haveshownaperiodof12.73±days. 5. Conclusion We have analysed data from the INTEGRAL mission to build flux and spectral history of 1E1740.7–2942from2003to2012. Ourdataanalysisisautomaticanddividedintwomainfronts. FirstweuseascripttoautomaticallyreducethedatausingtasksoftheOSA10.0software. Second, a Tcl script fits the data with XSPEC–aided commands. We collected, when possible, flux in 4 energybands. Ourdatashow,basedonourfitswithcutoffPL,thatthesourceisobservedtobein its canonical LHS state, where the flux in the hard X-ray band is greater than in the soft one. Our characterization ofthe 1E1740.7–2942 spectra withcompTTshows anaverage kT of 50keV. A e Lomb–Scargleperiodogramanalysishasshowntwotentativeperiodsof2.90and3.99day. These resultsmustbetakenwithcautionduetouncertaintiesthatarestillbeingaddressed. 4 1E1740.7−2942withINTEGRAL ManuelCastro 18 16 1E 1740.7−2942 INTEGRAL/IBIS data 1 − s 14 2 − m c g 12 er 0 1 −0 10 1 ux in 8 V fl e 0 k 6 0 2 o 0 t 4 5 2 0 0 1 2 3 4 5 6 7 8 9 20 to 50 keV flux in 10−10 erg cm−2 s−1 Figure3: Thefluxesinthetwohighestenergybandsinthisstudy. Shownintheplotalso(inblue)isthe regionofaverage±1σ forbothbands. Thisshowstheregion(influx)ofpreferredfluxesforthe(canonical) highstatewhere1E1740.7–2942iscommonlyobserved. 35 mean=6.70 1E 1740.7−2942 30 median=6.60 5−20 keV flux 25 σ=1.14 102 data points by Jem−X 20 15 10 5 25 1E 1740.7−2942 mean=3.90 20 1100−2 2d0a tkae Vpo filnutxs by Jem−X mσ=e0d.i7a4n=3.80 15 10 y c 5 n e 65 u q 55 1E 1740.7−2942 mean=5.1 Fre 234555 2207−4 5d0a tkae Vpo filnutxs by IBIS mσ=e1d.i6an=5.2 15 5 55 1E 1 740.7 −294 2 27 6 dat a poi nts b y IBIS 50−200 keV flux mean=8.7 45 median=8.6 35 σ=3.4 25 15 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 −10 −2 −1 Flux (10 erg cm s ) Figure4: Histogramdistributionforfluxesinthe4energybandsconsideredinthisstudy. 5 1E1740.7−2942withINTEGRAL ManuelCastro 50 1E 1740.7−2942 mean=1.48 median=1.48 40 σ=0.22 30 y c n e qu e Fr 20 10 0 1.0 1.2 1.4 1.6 1.8 2.0 2.2 Γ Figure5: HistogramdistributionforthepowerlawindexΓderivedfromacutoffPLfitinXSPEC. 60 1E 1740.7−2942 INTEGRAL/IBIS data median=46 mean=49 σ=18 50 40 y c n e qu e 30 Fr 20 10 0 20 40 60 80 100 kT, the compTT plasma temperature (keV) Figure6: HistogramdistributionfortheplasmatemperaturekTderivedfromacompTTfitinXSPEC. 6 1E1740.7−2942withINTEGRAL ManuelCastro 2.90 days 0.5 0.4 1 − z 3.99 days H 2 n) a e 0.3 m s/ m er, (r 0.2 w o P 0.1 0 0.00 0.10 0.20 0.30 0.40 0.50 Frequency (days−1) Figure 7: Lomb-Scargle periodogram of the 50–200keV flux showing two tentative possible periods for 1E1740.7–2942(seetextfordetails). References [1] Hertz,P.,Grindlay,J.,TheEinsteingalacticplanesurvey-StatisticalanalysisofthecompleteX-ray sample,ApJ,278,137(1984) [2] Syunyaev,R.,Gilfanov,M.,Churazov,E.,etal.,TwoHardX-RaySourcesNeartheGalacticCenter- theKnownSource1E1740.7-2942andtheNewlyDiscoveredSourceGRS:1758-258,Sov.Astron. Lett+.17.50(1991) [3] Bouchet,L.,delSanto,M.,Jourdain,E.,etal.,UnveilingtheHighEnergyTailof1E1740.7-2942 WithINTEGRAL,ApJ,693,1871(2009) [4] Reynolds,M.,Miller,J.,SuzakuObservationsoftheGalacticCenterMicroquasar1E1740.7-2942, ApJ,716,1431(2010) [5] Gallo,E.,Fender,R.,Chandraimagingspectroscopyof1E1740.7-2942,Mont.Not.R.Astron.Soc., 337,869(2002) [6] Castro,M.,D’Amico,F.,Braga,J.,etal.,ConfirmingthethermalComptonizationmodelforblack holeX-rayemissioninthelow-hardstate,A&A,569,A82(2014) [7] Winkler,G.,Courvoisier,T.J.-L.,DiCocco,G,.etal.,TheINTEGRALmission,A&A,411,L1(2003) [8] Smith,D.,Heinfl,W.,Swank,J.,etal.,X-RayTimingin1E1740.7-2942andGRS1758-258,ApJ, 489,L51(1997) [9] Smith,D.,Heindl,W.,Swank,J.,OrbitalandSuperorbitalPeriodsof1E1740.7-2942andGRS 1758-258,ApJ,578,L129(2002) 7 1E1740.7−2942withINTEGRAL ManuelCastro [10] Remillard,R.,McClintock,J.,X-RayPropertiesofBlack-HoleBinaries,Annu.Rev.Astron.Astr.,44, 49(2006) [11] RCoreTeam,R:ALanguageandEnvironmentforStatisticalComputing,RFoundationforStatistical Computing,Vienna,Austria,2008 8