Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://hea.iki.rssi.ru/integral06/papers/G32_Fiocchi01-22Dec06_16:28:45.pdf Äàòà èçìåíåíèÿ: Fri Dec 22 16:28:45 2006 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 23:25:03 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: movie

HARD TAIL DETECTION IN THE LMXB 4U 1636­53 FROM INTEGRAL
Mariateresa Fiocchi, Angela Bazzano, Pietro Ubertini, and Memmo Federici Istituto di Astrofisica Spaziale e Fisica Cosmica di Roma (INAF). Via Fosso del Cavaliere 100, Roma, I-00133, Italy

ABSTRACT Recent INTEGRAL observations and archival BeppoSAX data have been analyzed to deeply investigate the hard X-ray behavior of the neutron star, atoll type, low mass X-ray Binary 4U 1636­53 . Our investigation in three different periods outline three corresponding different sates. In fact, the source was detected in both the canonical high and low state and moreover in one occasion INTEGRAL spectrum shows, for first time in this source, a hard tail dominating the emission above 30 keV. This spectrum is fitted as the sum of a Comptonized component similar to soft state and a power-law component ( = 2.76), indicating the presence of a non thermal electron distribution of velocities. Key words: accretion, accretion disks ­ gamma rays: observations ­ radiation mechanisms: non-thermal ­ stars: individual: 4U 1636­53 ­ stars: neutron ­ X-rays: binaries.

2. OBSERVATIONS AND DATA ANALYSIS Table 1 summarizes the log of INTEGRAL and BeppoSAX observations of 4U 1636­53 . BeppoSAX observed the source on three occasions: February and March 1998 and February 2000. LECS, MECS and PDS event files and spectra, available from the ASI Scientific Data Center, were generated by means of the Supervised Standard Science Analysis [14]. Both LECS and MECS spectra were accumulated in circular regions of 8' radius. The PDS spectra were obtained with the background rejection method based on fixed rise time thresholds. Publicly available matrices were used for all the instruments. Fits are performed in the following energy band: 0.5­3.5 keV for LECS, 1.5­10.0 keV for MECS and 15­70 keV for PDS. The analyzed INTEGRAL [33] data set consists of all observations in which 4U 1636-53 was within the highenergy detectors field of view. Observations are organized into uninterrupted 2000 s long science pointing, windows (scw): light curves and spectra are extracted for each individual scw. Wideband spectra (from 5 to 150 keV) of the source are obtained using data from the two high-energy instruments JEM-X [23] and IBIS [29]. Data were processed using the Off-line Scientific Analysis (OSA version 5.1) software released by the INTEGRAL Scientific Data Centre. Data from the Fully Coded field of view only for both instrument have been used.

1. INTRODUCTION 4U 1636­53 is a neutron star low mass X-ray binary (LMXB) classified as a atoll source [17], with an orbital period of 3.8 hr derived from the optical variability of its companion V801 Arae [25] and at distance of 3.7­6.5 kpc [15, 27, 2]. While the X-ray burst properties and timing signatures have been analyzed extensively (see [18], [4] and references therein) the spectral characteristics have been studied only at low energy with Einstein [6], EXOSAT [30], Temna [32] and ASCA [1]. In general, the spectrum was acceptably fitted by a Comptonization model plus a black body component. We report here a broad band spectral analysis performed on data from BeppoSAX and INTEGRAL satellites, which allowed us to better constrain the spectral parameters and to detect the presence of a high energy tail dominating the spectrum above 30 keV. A similar feature has been observed in other LMXBs namely GX 17+2 [11], GX 349+2 [12], Sco X-1 [9], 4U 1608-522 [34], XB 1254-690 [20], Cir X-1 [21] and 4U 0614+091 [26].

3. SPECTRAL ANALYSIS Several physical models have been used to fit the whole data set, keeping the number of free parameters as low as possible. Each time a new component was added to the model, a F-test was performed. We assumed that a F probability larger than 95% implies a significative improvement of the fit. The uncertainties are at 90% confidence level for one parameter of interest (2 = 2.71). When spectra are from more than one detector, we allow the relative normalization to be free with respect to the MECS and IBIS data, for BeppoSAX and INTEGRAL respectively. XSPEC v. 11.3.1. has been used.


Table 1. BeppoSAX and INTEGRAL Observations
Start Date LECS ksec 1 1 1
st st st

BeppoSAX Journal Exposure time MECS ksec PDS ksec LECS [0.4-3 keV]

Count s

-1

MECS [1.5-10 keV] 39.37 ± 0.02a 25.42 ± 0.04b 29.09 ± 0.03b

PDS [20-60 keV] 0.48 ± 0.04 0.80 ± 0.06 0.73 ± 0.03

epoch (a) epoch (b) epoch (c)

1998-02-24 1998-03-01 2000-02-15

13 6 12

39 17 14.66 ± 0.03 14 7 14.95 ± 0.05 37 19 22.08 ± 0.04 INTEGRAL Journal Exposure time JEM-X IBIS ksec ksec 36 16 594 117
b

Start Date

Count s-1 JEM-X IBIS [5-15keV] [20-150 keV] 6.3 ± 0.2 3.70 ± 0.06 10.74 ± 0.08 3.22 ± 0.04

2nd epoch 3rd epoch
a

2003-03-04 2003-03-04

MECS count rates refer to MECS2 and MECS3 units.

MECS count rates refer to only MECS2 unit.

Spectral behavior has been studied separately in three epochs consisting of the following data: · 1st epoch: all three BeppoSAX observations available from February 1998 to February 2000. During these periods the source was always in a soft/high state. · 2nd epoch: JEM-X and IBIS data available from 52644 MJD to 53644 MJD with count rate < 5 counts s-1 in the 20­40 keV energy band. For the chosen period the source was in a hard/low state. · 3rd epoch: JEM-X and IBIS data available from 52644 MJD to 53644 MJD with count rate > 5 counts s-1 in the 20­40 keV energy band. This epoch does not corrspond to either the soft or the hard state and and here we call it peculiar state as will be explained in detail later on. After veryfing there were no changes in the three observations and in order to achieve the highest signal to noise ratio, we build the BeppoSAX average spectrum arranging all three observations. We then can take advantage of the high quality of the average spectrum up to about 70 keV. The most simple model which provides a good fit to the average BeppoSAX spectrum in the energy band 0.5­70 keV consist of a thermal Comptonized component modeled in XSPEC by COMPTT (a spherical geometry was assumed) plus a soft component which we modeled with two temperature blackbody. Being the column density NH towards the source measured by the LECS and MECS instrument always close to the galactic column density (NH g alactic = 3.58 â 1021 cm-2 , estimated from the 21 cm measurement [10]), we fix this parameter at this value. Spectral fit results are given in Table 2, the average spectrum is shown in Fig. 1. The source was in the soft/high state with an un-absorbed luminosity of L0.1-200 keV 2.0 â 1037 erg s-1 , assuming a distance of 5.9 kpc (Cornelisse et al. 2003). As in the case of the Einstein observation [6], thermal Comptonization of the optically thick plasma corona with a quite low electron temperature is dominating. Figure 1. The BeppoSAX average spectrum in the soft state (epochs 1), shown together with the total model and its components: the total, two blackbody and the Comptonization components. The most simple model which provides a good fit to the INTEGRAL hard state consist of a thermal Comptonized component modeled in XSPEC by COMPTT [28] (a spherical geometry was assumed) plus a soft component which we modeled by a single temperature blackbody. Because of the very good low energy BeppoSAX coverage, we used the average value of input soft photon temperature and column density (T0 = 1.3k eV and NH = NH Galactic) measured by the BeppoSAX observations. During this period, the source has a luminosity of L0.1-200 keV 1.4 â 1037 erg s-1 , lower than the one in the soft state, in agreement with the usual ranking of the luminosity in atoll sources (e.g., [17]; [31]; [16]). The electron temperature is now substantially higher than in the soft state, Te 23 keV, and the Comptonization component extends well above 100 keV. The same model was been applied to the 2nd INTEGRAL data set but resulted in a poor fit with a 2 /d.o.f = 97/57 with clear residuals above 60 keV. Adding a power law improves the fit significantly (2 /d.o.f becomes 64/55). The disk component becomes negligible and it is not necessary to best fit the data. Spectral fit results are given in Table 2, and spectra are shown in Fig. 2. Since the spectral parameters in the peculiar state are similar to


Table 2. Results of the fit of 4U 1636­53 spectra in the energy band 0.5 - 70 keV and 5 - 150 keV for BeppoSAX and INTEGRAL observations respectively, for three different spectral states.
TBB1 keV 0.24 ± 0.01 TBB2 keV 0.58 ± 0.02 T0 keV 1.3 ± 0.1 BeppoSAX average high/soft state spectrum model = bbody + bbody + comptt Te RBB1 keV km 3.4 ± 0.3 3.8 ± 0.4 79 ± 6 INTEGRAL low/hard state spectrum model = bbody + comptt Te RBB2 keV km
+7 23 2 1.1+0.5 5+3 - -0.3 -2 INTEGRAL peculiar state spectrum model = comptt + pow er law nCOMPTT 10-2

RBB2 km 25 ± 1

n

COMPTT

2 /d.o.f

0.19 ± 0.02 2 /d.o.f

235/190

TBB2 keV 1.2
+0.2 -0.3

T0 keV 1.3 fixed

n

COMPTT -2

10

1.0

+0.3 -0.1

57/59

T0 keV 1.3 fixed

Te keV
+4 53 -

npowerlaw ph keV -1 cm 1.0
+1.4 -0.5

2 /d.o.f
-2

3±2

2.6 ± 0.1

2.8

+0.8 -0.6

72/57

those in the soft/high BeppoSAX state, a possible interpretation of this state is that 4U 1636­53 was in a similar soft/high state as during the BeppoSAX observations, but with a new overlapped component at high energies simultaneously present. The un-absorbed luminosity is L0.1-200 keV 7.4 â 1037 erg s-1 .

4. CONCLUSIONS The black-body component in the soft states could originate at both the neutron star surface and the surface of an optically-thick accretion disk. In our observations the two black body components seems to originate from two different parts of the disk, corresponding to two different temperature. The Comptonization component may arise from a corona above the disk and/or between the disk and the neutron star surface. In the hard states, accretion probably assumes the form of a truncated outer accretion disk as previously reported by [3] for LMXB 4U 1705-44. The spectral transitions are generally, but not necessarily, coupled with changes in luminosity, indicating they are driven by variability of the accretion rate or change of the geometry of the system as for Black Hole hard/soft transition at constant luminosity [4]. For 4U 1636­53 the accretion rate is lower in the hard state than in the soft state and becames very high in the peculiar state. We note this value can be influenced by the steep power law that dominates the energy spectrum at low energies. The emission is best described by Comptonization from a complex electron distribution due to a low temperature ( 5keV) thermal electron distribution together with non-thermal power-law electrons. This two-component electron distribution could be explained by the following hypothesis: · Non-thermal electron acceleration regions powered by magnetic reconnections above a disc. Lowenergy electrons cool preferentially by Coulomb collisions leading to a thermal distribution while the

Figure 2. The spectra of epochs 2 (soft state, top) and 3 (peculiar state, bottom) observed by INTEGRAL , shown together with the total model and its components.


high-energy electrons cool by Compton scattering, preserving a non-thermal distribution [7]. · The magnetic reconnection above the disc can produce a non-thermal electron distribution, while overheating of the inner disc produces the thermal Comptonization [19]. · The power-law component could be produced by Componization by synchrotron emission in the jet [5, 13] In our source, this hypothesis is strengthened by radio detection: Sydney University Molonglo Sky Survey catalog gives a flux of 7.5 mJy at 843 GHz [24]. Finally, Laurent & Titarchuk [22] suggest the detection of a power-law component at high energy to be a signature of presence black hole in an X-ray binary system. Data extended up to a few hundred keV are need to check this criterion proposed to distinguish black hole versus neutron star binaries.

[20] Iaria, R., Di Salvo, T., Burderi, L., Robba, N. R., 2001, ApJ, 548, 883 [21] Iaria, R., Di Salvo, T., Robba, N. R., Burderi, L., 2002, ApJ, 567, 503 [22] Laurent, P. & Titarchuk, L., 1999, ApJ, 511, 289 [23] Lund, N., et al., 2003, A&A, 411, L231 [24] Mauch, T. et al. 2003, MNRAS, 342, 1117 [25] Pedersen, H. van Paradijs, J. Lewin, W. H. G. 1984, IAUC, 3952, 3 [26] Piraino, S., Santangelo, A., Ford, E. C., Kaaret, P., 1999, A&A, 349, 77 [27] Smale, Alan P., & Lochner, James C., 1992, ApJ, 395, 582 [28] Titarchuk, L. 1994, ApJ, 434, 570 [29] Ubertini, R., Lebrun, F., di Cocco, G. et al. 2003, A&A, 411, 131 [30] Vacca, W. D., Sztajno, M., Lewin, W. H. G., Truemper, J., van Paradijs, J., Smith, A. 1987, A&A, 172, 143 [31] van der Klis, M., 2000, ARA&A, 38, 717 [32] Waki, I., et al., 1984, PASJ, 36, 819 [33] Winkler, C., Gehrels, N., Schnfelder, V., Roques, J.-P., Strong, A. W., Wunderer, C., Ubertini, P. , 2003, A&A, 411, 349 [34] Zhang, S. N., et al., 1996, A&A, 120, 279

REFERENCES
[1] Asai, K., Dotani, T., Mitsuda, K., Inone, H., Tanaka, Y., Lewin, W. H. G., 1998, IAUS, 188, 354 [2] Augusteijn, T., van der Hooft, F., de Jong, J. A., van Kerkwijk, M. H., van Paradijs, J., 1998, A&A 332, 561 [3] Barret, D., & Olive, J. F. 2002, ApJ, 576, 391 [4] Belloni T., Mendez M. & Homan J., 2005, A&A, 437, 209 [5] Bosch-Ramon, V., Romero, G. E. and Paredes, J. M., 2005, A&A, 429, 267 [6] Christian D. J. & Swank J. H., 1997, ApJ, 109, 177 [7] Coppi P. S., 1999, in ASP Conf. Ser. 161, ed. J. Poutanen & R. Svensson (San Francisco: ASP), 375 [8] Cornelisse, R. et al., 2003, A&A, 405, 1033 [9] D'Amico, F., Heindl, W.A., Rothschild, R.E., Gruber, D.E., 2001, ApJ, 547, 147 [10] Dickey & Lockman, 1990, Ann. Rev. Astron. Astrophys. 28, 2 [11] Di Salvo, T., et al., 2000, ApJ, 544, 119 [12] Di Salvo, T., Robba, N. R., Iaria, R., Stella, L., Burderi, L., Israel, G. L., 2001, ApJ, 554, 49 [13] Fender, R. P., 2004, eds. W.H.G. Lewin and M. van der Klis, Cambridge University. astro-ph/0303339 [14] Fiore, F., Guainazzi, M., & Grandi, (www.asdc.asi.it/bepposax/software/cookbook) P. 1999,

ACKNOWLEDGMENTS We acknowledge the ASI financial/programmatic support via contracts ASI-IR 046/04. A special thank to K. Kretschmer for making data available before becoming public.

[15] Fujimoto, M. Y., Sztajno, M., Lewin, W. H. G., van Paradijs, J., 1988, A&A, 199, 9 [16] Gierlinski, M., & Done, C., 2002, MNRAS, 337, 1373 ´ [17] Hasinger, G., & van der Klis, M. 1989, A&A, 225, 79 [18] Jonker P. G., Mendez M., van der Klis M., 2005, MNRAS, 360, 921 [19] Kubota A., Makishima K., Ebisawa K., 2001, ApJl, 560, L147