1: \documentclass[12pt,preprint]{aastex}
2: \notetoeditor{refer to: Fiamma Capitanio. Institute: IASF-INAF,Via Fosso de
3: l Cavaliere, 100, 00133 Rome ITALY. Tel: +390649934557. E-mail: fiamma.capitanio@rm.iasf.cnr.it ; pietro.ubertini@rm.iasf.cnr.it}
4: \newcommand{\inst}{\altaffilmark}
5: \newcommand{\instM}{\altaffiltext}
6:
7: \shorttitle{Spectral states of the new X-ray binary IGR J17091--3624}
8: \shortauthors{Capitanio F. et al.}
9: \begin{document}
10: \title{Spectral states of the X-ray binary IGR J17091--3624 observed by INTEGRAL and RXTE}
11:
12: \author {F. Capitanio\inst{1,}\inst{3}, A. Bazzano\inst{1}, P. Ubertini\inst{1}, A. A. Zdziarski\inst{2}, A. J. Bird\inst{3}, G. De Cesare\inst{1}, A. J. Dean\inst{3}, J. B. Stephen\inst{4}, A. Tarana\inst{1} }
13:
14: \altaffiltext{1}{IASF/INAF, Rome, Italy, Via Fosso del Cavaliere 100, I-00133 Rome, Italy}
15: \altaffiltext{2}{Centrum Astronomiczne im.\ M. Kopernika, Bartycka 18, 00-716 Warszawa, Poland}
16: \altaffiltext{3}{School of Physics and Astronomy, University of Southampton, Highfield Southampton, SO17 1BJ, UK}
17: \altaffiltext{4}{ IASF/INAF, Bologna, Via Godetti 101, I-40129 Bologna, Italy}
18:
19: % \date{Received; accepted }
20:
21: \begin{abstract} IGR J17091--3624 was discovered in 2003 April by {\it
22: INTEGRAL}/IBIS during its Galactic Centre Deep Exposure programme. The source was initially detectable only in the 40--100 keV range, but after two days was also detected in the 15--40 keV range. Its flux had by then increased to 40 mCrab and 25 mCrab in the 15--40 keV and 40--100 keV bands respectively. {\it RXTE\/} observed the source simultaneously on 2003 April 20, with an effective exposure of 2 ksec. We report here the spectral and temporal evolution of the source, which shows a transition between the hard and soft states. We analyse in detail the {\it RXTE}/{\it INTEGRAL\/} Comptonised spectrum of the hard state as well as the JEM-X detection of a blackbody component during the source softening. Even though the source spectral behavior and time variability show a similarity with the outburst of the black-hole candidate IGR J17464--3213 (= H1743--322), observed by {\it INTEGRAL\/} in 2003, the nature of its compact object (BH vs.\ NS) remains controversial.
23: \end{abstract}
24:
25: \keywords{X-rays: binaries -- X-rays: stars -- X-rays: individual: IGR
26: J17091--3624 -- gamma rays: observations -- black hole physics}
27:
28:
29: \section{Introduction}
30:
31: A new source was discovered by {\it INTEGRAL}/IBIS during a pointed observation on 2003 April 14--15. It was designated IGR J17091--3624 based on its position of R.A.\ (2000) $=17^{\rm h}09.1^{\rm m}$, Dec.\ $=-36\degr 24.5\arcmin$, with an error box of $3\arcmin$ at 90\% confidence \cite{Atel1}.
32:
33: Initially, its flux was $\sim$20 mCrab in the 40--100 keV energy band exhibiting
34: a hard spectrum, while it was not detected in the 15--40 keV band with an upper
35: limit of $\sim$10 mCrab. A week earlier, the source did not appear in either
36: energy band, with upper limits of $\simeq$10 mCrab \cite{Atel1}. During
37: subsequent observations of the Galactic Center Deep Exposure (GCDE) on 2003
38: April 15--16, the source flux increased to $\sim$25 mCrab in the 40--100 keV
39: band. The source was also included in the {\it INTEGRAL\/} IBIS/ISGRI Galactic
40: plane survey catalogue \cite{Bird} and classified as unknown.
41: Observations with the Very Large Array revealed the presence of a
42: possible radio counterpart with a steeply falling spectrum typical of
43: synchrotron emitters \cite{Atel3}, this was also confirmed by a subsequent observation
44: with the Giant Metrewave Radio Telescope (Pandey et al.\ 2006).
45:
46: Immediately after the {\it INTEGRAL\/} discovery, an {\it RXTE\/} observation
47: was performed. The analysis of the {\it RXTE\/} data showed a relatively stable
48: source flux without features associated with quasi-periodic oscillations (QPOs)
49: in the power spectrum. Based on the energy and power spectra, Lutovinov \&
50: Revnivtsev (2003) suggested that IGR J17091--3624 is an X-ray binary in the
51: low/hard state and probably a black-hole system. IGR J17091--3624 was then
52: searched for in the X-ray catalogs and detected in the archival data of both
53: the TTM telescope on board the KVANT module of the {\it Mir\/} orbital station
54: \cite{Atel2}, and in the {\it BeppoSAX\/} WFC \cite{Atel4}. Following this,
55: Lutovinov et al.\ (2005) performed a preliminary study of the IBIS/ISGRI
56: spectrum of the source, showing that its flux changed by a factor of $\sim$2
57: between 2003 April and 2003 August--September. During 2003, the source spectrum
58: was described by a simple power law with a photon index of $\Gamma\simeq 2.2$ in
59: the 20--150 keV energy band, i.e., softer than the {\it RXTE\/} spectrum. As the
60: source flux in the hard X-rays increased in 2004, the spectrum became harder
61: with $\Gamma\simeq 1.6$. From the above investigations, IGR J17091--3624 appears
62: as a moderately bright variable (probably a transient) source with flaring
63: activity in 1994 October ({\it Mir}/KVANT/TTM), 1996 September ({\it
64: BeppoSAX}/WFC), 2001 September ({\it BeppoSAX}/WFC, in 't Zand et al.\ 2003),
65: and 2003 April ({\it INTEGRAL}/IBIS, Kuulkers et al.\ 2003).
66:
67: \subsection{Improvements over previous analysis}
68:
69: We have performed a more accurate spectral analysis of this source compared to those
70: previously reported, analysing all the {\it INTEGRAL\/} data available
71: while the source was detectable (until the end of 2004 April). The new set of
72: data allowed us to fit a joint {\it RXTE}/{\it INTEGRAL\/} spectrum of IGR
73: J17091--3624 in the low/hard state. The improved version of the {\it INTEGRAL\/}
74: data analysis software (OSA 4.2, Goldwurm et al.\ 2003) has allowed us to obtain
75: {\it INTEGRAL\/} spectra up to 150 keV. An initial test with the newly delivered OSA 5
76: software has shown a substantial agreement with the former OSA 4.2 results.
77:
78: We have fitted the data with detailed models, thus obtaining new information on
79: the geometry of the source and on the accretion disk optical depth and
80: temperature. We have also extracted a JEM-X spectrum that confirms the previous
81: hypothesis of a transition to the soft state \cite{Lut} and furthermore we have
82: found suggestions of a hysteresis-like behaviour in the source spectral
83: evolution, with the hard-to-soft state transition occurring at higher luminosity
84: than the soft-to-hard transition. We then speculate on the source nature, even
85: if the final picture is still not evident.
86:
87: \section{Data Analysis}
88:
89: In this paper, we analyse the data set consisting of all {\it INTEGRAL\/}
90: \cite{Winkl} Core Program and public observations in which IGR J17091--3624 was
91: within the IBIS field of view, and of the {\it RXTE\/} Proportional Counter
92: Array (PCA, Bradt et al.\ 1993) data collected during the 2 ks observation
93: performed on 2003 April 20. The {\it INTEGRAL\/} observations, though not
94: continuous, cover a period of one year from 2003 April to 2004 April. The {\it
95: RXTE\/} observation corresponds to a period between {\it INTEGRAL\/} rev.\ 61
96: (2003 April 14--16) and the beginning of rev.\ 63 (2003 April 21). In general,
97: all {\it INTEGRAL\/} observations are divided into uninterrupted 2000-s time
98: intervals, called science windows (SCW). Table~\ref{integral_log} summarises the
99: {\it INTEGRAL\/} observations of the source. Light curves, hardness ratio and
100: spectra are then obtained for each individual SCW. We use here {\it INTEGRAL\/}
101: data from the coded mask imager, IBIS \cite{Uber}. Then, in order to obtain
102: broad-band spectra, we add the lower energy data from the PCA, and, whenever
103: available, those from the {\it INTEGRAL\/} X-ray Monitor, JEM-X \cite{Lund}.
104:
105: Figure \ref{image} shows the 20-150 keV {\it INTEGRAL\/} image of IGR
106: J17091--3624 during the period of the {\it RXTE\/} observation. The figure shows
107: that the transient source IGR J1709.8-3628, detected $9.4\arcmin$ from IGR
108: J17091--3624 on 2005 March 24 2005 \cite{source}, was then not visible. This
109: rules out any contamination of other sources in the {\it RXTE}/PCA field of
110: view.
111:
112: The IBIS data were processed using the Off-line Scientific Analysis, OSA 4.2,
113: Goldwurm et al.\ 2003) released by the {\it INTEGRAL\/} Science Data Centre,
114: ISDC \cite{Courv}. The data set has been divided in two subsets according to the
115: source position being either within the partially coded field of view (PCFOV,
116: $19\degr\times 19\degr$) or in the narrower fully coded field of view (FCFOV,
117: $9\degr\times 9\degr$). Both data sets have been used to produce the source
118: light curves in different energy ranges while only FCFOV data have been used for
119: spectral extraction because of the better signal to noise ratio.
120:
121: Since the IBIS spectra do not show any substantial variability during 2003 April
122: 14--21, we have integrated over this observation period. The JEM-X spectra were
123: derived with the ISDC OSA 5.0, using the latest available spectral matrices and
124: extracting the source spectrum with the fixed position of the source. While IBIS
125: provides a large FOV, $>30\degr$, that of JEM-X is narrower ($>10\degr$), thus
126: providing only a partial overlap with the high energy detector. In general, the
127: effective JEM-X exposure is $\sim$15\% of that of IBIS. Therefore, a
128: combined JEM-X/IBIS spectrum was possible only for a small fraction of the data.
129:
130: We have extracted the PCA spectrum and the light curve from the {\it RXTE\/}
131: public data
132: archive\footnote{http://heasarc.gsfc.nasa.gov/docs/xte/xte\_public.html.}. When
133: spectral data are obtained in the same time interval from more than one
134: instrument, we allowed a free relative overall normalisation constant for each
135: instrument with respect to IBIS.
136:
137: \section{Results}
138:
139: \subsection{Time evolution}\label{time}
140:
141: IGR J17091--3624 is a very faint source, which is often below the detection
142: limits of the IBIS and JEM-X in a single 2000-s SCW ($\sim$5 mCrab in the
143: 20--100 keV IBIS energy range, and $\sim$23 mCrab in the 3--20 keV JEM-X energy
144: range). Furthermore, most of the JEM-X data were not in its FCFOV, which is
145: partly due to the pointing algorithm of the GCDE. As a result, the JEM-X
146: effective exposure was $\sim$10 times shorter than that of IBIS.
147:
148: We have followed the IBIS evolution of the source flux from the first {\it
149: INTEGRAL\/} detection of the source. First, a flux increase took place in the
150: first SCWs of revs.\ 61--63 (2003 April 14--21), as can be seen in Figure
151: \ref{lc_20_50}. We have also found a very faint signal in the data of rev.\ 60.
152:
153: We have analysed the data until rev. 185 (2004 April 19--20), as later observations
154: do not show any signal in the direction of the source at least until rev.\ 232
155: (2004 September), the current limit of the public data set. We have produced
156: IBIS light curves in two different energy bands, 20--40 and 40--100 keV (the
157: top and middle panel, respectively, of Figure \ref{lc_20_50}).
158: The light curves are derived from the flux and variance maps for each individual
159: SCW by extracting values directly from the derived source positions. The flux maps
160: are already corrected for off-axis effects so this correction is automatically
161: included in the light curves. Differences in the size of error the bars on
162: individual points are due to both different SCW exposures and different source locations
163: within the field of view. The bottom panel of Figure~\ref{HR} shows the hardness ratio,
164: $HR\equiv R_{40-100}/R_{20-40}$, where $R$ is the count rate. During MJD 52860--52924
165: (2003 August--October), the source was rather faint, especially in the 40--100
166: energy range, and the hardness ratio shows a weak indication of softening
167: corresponding to 25\%.
168:
169: The {\it RXTE\/} observation lasted only 2 ks, so that it has not been possible to extract information on the long-term temporal behaviour. A detailed timing analysis of the 2-ks data set is given by Lutovinov \& Revnivtsev (2003).
170:
171: For the third observation period, during MJD 53054--53097 (2004 February--April), only IBIS data were available. The light curves are characterised by an increased flux with respect to the previous periods together with a faint tendency of source hardening.
172:
173: \subsection{Spectral evolution of IGR J17091--3624}
174: \label{spectral}
175:
176: We have studied the spectral behaviour separately for 3 epochs:
177:
178: 1) revs.\ 61--63 (2003 April 15--21, $\sim$15 ks), which is the first {\it INTEGRAL\/} observation combined with the {\it RXTE}/PCA data;
179:
180: 2) revs.\ 100--119 (2003 August--October, $\sim$77 ks), during which the IBIS spectrum softened, and which includes the only JEM-X detection.
181:
182: 3) revs.\ 165--179 (2004 April, $\sim$77 ks), containing only the IBIS data.
183:
184: The data sets have been fitted with XSPEC 11.3.1. Several models have been tested for each of the data sets. In our fits, we have confirmed the constraint found by Lutovinov \& Revnivtsev (2003), $N_{\rm H}<10^{22}$ cm$^{-2}$, which is also in agreement with the Galactic column density along the direction to the source, $\simeq 8\times 10^{21}$ cm$^{-2}$ (http://heasarc.gsfc.nasa.gov).
185:
186: For revs.\ 61--63, the relatively high brightness and the lack of variability have allowed us to obtain a combined IBIS/PCA spectrum, (see Figure~\ref{spectra}). We use two thermal Comptonization models, {\sc comptt} \cite{Tit} and {\sc compps} \cite{Pou}, both assuming a slab geometry. The {\sc compps} model is an accurate iterative-scattering model, in which subsequent photon scatterings are directly followed. It has been extensively tested against Monte Carlo results, e.g., by Zdziarski, Poutanen \& Johnson (2000). On the other hand, {\sc comptt} is based on an approximate solution of the kinetic equation with some relativistic corrections, and the resulting spectra are also only approximate (see the appendix of Zdziarski, Johnson \& Magdziarz 1996). For the {\sc compps} model, we assumed a viewing angle of $60\degr$ and the source of seed photons to be at the slab bottom (the model geometry parameter was set equal to 1) with a blackbody distribution at $kT_{\rm bb}=0.1$ keV. For the {\sc comptt} model, we used Wien seed photons with the temperature as above. The free parameters of each model are the electron temperature, $kT_{\rm e}$, the Thomson optical depth, $\tau$, and the normalization, $N$. There is no evidence of a Compton reflection component, and the weak fluorescent Fe K$\alpha$ line present in the spectrum appears mostly due to the Galactic ridge emission \cite{RXTE}. Table~\ref{61_63}, the left panel of Figure~\ref{counts} and the red curve in Figure~\ref{spectra} show our results.
187:
188: The results corresponding to revs.\ 100--119 consist of the combined JEM-X (6 SCWs) and ISGRI average spectra. Given the difference in the JEM-X and ISGRI exposures, we have also compared the ISGRI count rate for the 6 SCWs for which the JEM-X spectrum was obtained, and found it is almost the same as the count rate averaged over the entire period of revs.\ 100--119. We have analysed in detail the JEM-X data in order to exclude any possible contamination from close by sources, being IGR J17091-3624 located in a crowed region near the galactic center.
189:
190: The source shows a bright component in soft X-rays, followed by a weak high-energy tail. Thus, we have used the model consisting of a disk blackbody ({\sc diskbb}, Mitsuda et al.\ 1984) and a power law spectrum. Table~\ref{100_119} summarizes our results, and the spectrum is shown in the right panel of Figure~\ref{counts} and in the blue curve in Figure~\ref{spectra}.
191:
192:
193: We clearly see a transition from the hard to the soft state from the first to the second epoch. After this, the source returns to the hard state, as shown by the IBIS data for revs.\ 164--179, green spectrum in Figure~\ref{spectra}. The epoch 3 IBIS data are fitted by {\sc comptt} with $kT_{\rm e}= 21^{+4}_{-4}$ keV and $\tau= 2.7^{+1.2}_{-0.5}$; these values are similar to that of epoch 1 (Table~\ref{61_63}).
194: %The upper limit on the 3--20 keV flux from JEM-X is $\sim$25 mCrab.
195:
196: The spectral evolution of IGR J17091--3624 shows both similarities and
197: differences with respect to that of another black-hole candidate observed by
198: {\it INTEGRAL\/}, IGR J17464--3213 \cite{Cap,JOINET}. Therefore, we report here
199: the spectra and fit results for that object as well. Table~\ref{1746} shows fit
200: results with both {\sc comptt} and {\sc compps} for the hard state and Figure
201: \ref{comparison} shows a comparison of its hard and soft states with those of
202: IGR J17091--3624.
203:
204: \section{Discussion and conclusions}
205:
206: We have followed the evolution of an outburst of IGR J17091--3624. Within the course of less than one year, the source went from the hard to the soft state and then back to the hard state.
207:
208: Although Markoff et al.\ (2005) have recently suggested that the hard state
209: emission could be due to synchrotron self-Compton emission from the base of the
210: jet, we have considered here the standard hard-state model with the dominant
211: radiative process being thermal Comptonization of disk blackbody photons. An
212: intriguing feature of our hard-state spectra is the rather low electron
213: temperature, $\sim$20 keV, while the range of temperatures typical for
214: black-hole binaries is $\sim$50--100 keV (e.g., Zdziarski \& Gierli\'nski 2004).
215: This low temperature is obtained using both the {\sc compps} and {\sc comptt} models,
216: and for both occurrences of the hard state, in the rise and decline. Such a low
217: temperature may be a characteristic of the hard state of neutron-star low-mass
218: X-ray binaries, although at present it is rather poorly constrained (e.g.,
219: Gierli\'nski \& Done 2002). On the other hand, we might have caught the source
220: during transitionary phases, with enhanced Compton cooling due to the increasing
221: flux of disk blackbody emission and the correspondingly lower electron
222: temperature.
223:
224: During the soft state in epoch 2, the peak of the $EF(E)$ emission is at several keV, as compared to the $\sim$100 keV peak in the hard state. There is a high-energy tail, but its weakness does not permit us to put constraints on its origin.
225:
226: In Section \ref{spectral}, we have also compared the evolution of IGR
227: J17091--3624 with that of another transient source observed by {\it INTEGRAL},
228: IGR J17464--3213. This is a black-hole candidate discovered by {\it HEAO 1\/} in
229: 1977, named then H1743--322. An outburst of that source was observed during the
230: GCDE in 2003 \cite{Cap}. Both sources show similar spectral evolution
231: during outbursts of similar duration. However, as shown in Figure\
232: \ref{comparison}, the hard-state spectrum of IGR J17464--3213 is substantially
233: softer than that of IGR J17091--3624, although both sources share the relatively
234: low electron temperature of $\sim$20 keV. Furthermore, the relative
235: normalisation of the observed hard and soft state spectra is different.
236: Curiously, the high-energy parts ($\ga$30 keV) of both spectra of IGR
237: J17464--3213, are virtually identical. The various amplitudes of the hard and
238: soft state spectra are often seen in X-ray transients (e.g., Zdziarski et al.\
239: 2004), and are due to the hysteretic behaviour of the accreting systems.
240:
241: IGR J17091-3624 is located $10\degr$ from the Galactic Centre region, where
242: the highest concentration of LMXRBs in the Galaxy is present \cite{White}. The
243: model unabsorbed bolometric (0.01--500 keV) luminosities in the hard (revs.\
244: 61--63) and soft state (revs.\ 100--119), assuming a distance of 8.5 kpc and
245: isotropy, are $L\simeq 1\times 10^{37}$ erg s$^{-1}$ and $L\simeq 2\times
246: 10^{37}$ erg s$^{-1}$, respectively. These values are not far from those
247: characteristic to the hard and soft states of black-hole binaries (e.g.,
248: Zdziarski \& Gierli\'nski 2004), although the neutron-star nature of the source
249: cannot be excluded. On the other hand, the {\it RXTE\/} power spectrum, without
250: QPO--like features and with band-limited noise \cite{RXTE}, commonly observed in black-hole binaries in the low/hard spectral state, as well as the probable
251: correlation with a radio source \cite{Pand,Atel3} may also argue for the
252: black-hole nature of this source.
253:
254: \begin{acknowledgements}
255:
256: We acknowledge the ASI financial/programmatic support via contracts I/R/046/04.
257: We thank the unknown referee for his/her useful comments, which helped us to
258: improve the quality of this work. We also thank Catia Spalletta for the careful
259: editing of the manuscript and Memmo Federici for supervising the {\it
260: INTEGRAL\/} data analysis. AAZ has been supported by grants 1P03D01827,
261: 1P03D01128, 4T12E04727.
262:
263: \end{acknowledgements}
264:
265: \begin{thebibliography}{}
266: \bibitem[Bird et al.\ 2004]{Bird}
267: Bird, A. J., Barlow, E. J., Bassani, L., Bazzano, A., Bodaghee, A., Capitanio, F., Cocchi, M., Del Santo, M, et al. 2004, ApJ, 607, L33
268: \bibitem[Bradt et al.\ 1993]{rxteref}
269: Bradt, H. V., Rothschild, R. E., \& Swank, J. H. 1993, A\&AS, 97, 355
270: \bibitem[Capitanio et al.\ 2005]{Cap}
271: Capitanio, F., Ubertini, P., Bazzano, A., Kretschmar, P., Zdziarski, A. A., Joinet, A., Barlow, E. J., Bird, A. J., et al. 2005, ApJ, 622, 503
272: \bibitem[Courvoisier et al.\ 2003]{Courv}
273: Courvoisier, T. J. L., Walter, R., Beckmann, V., Dean, A. J., Dubath, P., Hudec, R., Kretschmar, P., Mereghetti, S., et al. 2003, A\&A, 411, L53
274: \bibitem [Gierli{\'n}ski \& Done 2002]{gd02}
275: Gierli{\'n}ski, M., \& Done, C. 2002, MNRAS, 337, 1373
276: \bibitem[Goldwurm et al. 2003]{Gold}
277: Goldwurm, A., David, P., Foschini, L., Gros, A., Laurent, P., Sauvageon, A., Bird, A. J., Lerusse, L., et al., 2003, A\&A, 411, L223
278: \bibitem[ Grebenev et al.\ 2005]{source}
279: Grebenev, S. A., Molcov, S., V., \& Sunyaev 2005, ATel, 444
280: \bibitem[in 't Zand et al.\ 2003]{Atel4}
281: in 't Zand, J. J. M., Heise, J., Lowes, P., \& Ubertini, P. 2003, ATel, 160
282: \bibitem[Joinet et al.\ 2005]{JOINET}
283: Joinet, A., Jourdain, E., Malzac, J., Roques, J.~P., Sch{\"o}nfelder, V.,
284: Ubertini, P., \& Capitanio, F. 2005, ApJ, 629, 1008
285: \bibitem[Kuulkers et al. 2003]{Atel1}
286: Kuulkers, E., Lutovinov, A., Parmar, A., Capitanio, F., Mowlavi, N. \& Hermsen, W., 2003, ATel, 149
287: \bibitem[Lund et~al. 2003]{Lund}
288: Lund, N., Budtz-Jorgensen, C., Westergaard, N. J., Brandt, S., Rasmussen, I. L., Hornstrup, A., Oxborrow, C. A., Chenevez, J., et al. 2003, A\&A, 411, L231
289: \bibitem[Lutovinov \& Revnivtsev 2003]{RXTE}
290: Lutovinov, A. A., \& Revnivtsev, M. G. 2003, Astron.\ Lett., 29, 719
291: \bibitem[Lutovinov et al.\ 2005]{Lut}
292: Lutovinov, A., Revnivtsev, M., Molkov, S., \& Sunyaev, R. 2005, A\&A, 430, 997
293: \bibitem[Markoff et al. 2005]{Markoff}
294: Markoff, S., Nowak, M.~A., \& Wilms, J. 2005, ApJ, 635, 1203
295: \bibitem[Mitsuda et~al. 1984]{Mit}
296: Mitsuda, K., Inoue, H., Koyama, K., Makishima, K., Matsuoka, M., Ogawara, Y., Suzuki, K., Tanaka, Y. et al. 1984, PASJ, 36, 741
297: \bibitem[Pandey et~al.\ 2006]{Pand}
298: Pandey, M., Manchanda, R. K., Rao, A. P., Durouchoux, P. \& Ishwara-Chandra 2006, A\&A, in press (astroph/0509645)
299: \bibitem[Poutanen \& Svensson 1996]{Pou}
300: Poutanen, J., \& Svensson, R. 1996, ApJ, 470, 249
301: \bibitem[Revnivtsev et al.\ 2003]{Atel2}
302: Revnivtsev, M., Gilfanov, M., Churazov, E., \& Sunyaev R. 2003, ATel, 150
303: \bibitem[Rupen et al.\ 2003]{Atel3}
304: Rupen, M. P., Mioduszewski, A. J., \& Dhawan, V. 2003, ATel, 152
305: \bibitem[Titarchuk 1994]{Tit}
306: Titarchuk, L. 1994, ApJ, 434, 570
307: \bibitem[Ubertini et~al. 2003]{Uber}
308: Ubertini, P., Lebrun, F., di Cocco, G., Bazzano, A., Bird, A. J., Broenstad, K., Goldwurm, A., La Rosa, G. et al. 2003, A\&A, 411, L131
309: \bibitem[White \& van Paradijs 1996]{White}
310: White, N. E., \& van Paradijs, J. 1996, ApJ, 473, L25
311: \bibitem[Winkler et~al. 2001]{Winkl}
312: Winkler, C., Courvoisier, T. L. L., di Cocco, G., Gehrels, N., Giménez, A., Grebenev, S., Hermsen, W., Mas-Hesse, J. M et al. 2003, A\&A, 411, L1.
313: \bibitem[]{zg}
314: Zdziarski, A. A., \& Gierli{\'n}ski, M. 2004, Prog.\ Theor.\ Phys.\ Suppl., 155, 99
315: \bibitem[Zdziarski et al.\ 1996]{Zias3}
316: Zdziarski, A. A., Johnson, W. N., \& Magdziarz, P. 1996, MNRAS, 283, 193
317: \bibitem[Zdziarski et al.\ 2000]{Zias2}
318: Zdziarski, A. A., Poutanen, J., \& Johnson, W. N. 2000, ApJ, 542, 703
319: \bibitem[]{z04}
320: Zdziarski A. A., Gierli{\'n}ski, M., Miko{\l}ajewska, J., Wardzi{\'n}ski, G., Smith, D. M., Harmon, B. A., \& Kitamoto, S. 2004, MNRAS, 351, 791
321:
322: \end{thebibliography}
323:
324: \clearpage
325:
326: \begin{figure}
327: \centering
328: \includegraphics[angle=0, scale=0.4]{fig1.ps}
329: \caption{The 20--150 keV IBIS image of the field near IGR J17091--3624. Can be seen that IGR J17098-3628 (1) is not visible during the {\it RXTE}/PCA observations of IGR J17091--3624 (2), which argues against any contamination in the PCA field of view.}
330: \label{image}
331: \end{figure}
332:
333: \begin{figure}
334: \centering
335: \includegraphics[angle=0, scale=0.8]{fig2.ps}
336: \caption{The 20-40 keV and 40-100 keV light curves (the top and middle panel, respectively) and the corresponding hardness ratio (the bottom panel), with the solid line representing a value of to 1.}
337: \label{lc_20_50}
338: \label{lc_50_150}
339: \label{HR}
340: \end{figure}
341:
342: \begin{figure}
343: \centering
344: \includegraphics[angle=-90, scale=0.3]{fig3.ps}
345: \includegraphics[angle=-90, scale=0.3]{fig4.ps}
346: \caption{Left panel: the {\it RXTE}/IBIS count spectra for revs.\ 61--63; Right panel: the JEM-X/IBIS count spectra for revs.\ 100--119. }
347: \label{counts}
348: \end{figure}
349:
350: \begin{figure}
351: \centering
352: \includegraphics[angle=90, scale=0.7] {fig5.ps}
353: \caption{The unfolded spectra showing the source evolution during the {\it INTEGRAL\/} and {\it RXTE\/} observations. The red spectrum (PCA/IBIS) corresponds to revs.\ 61--63, the blue spectrum (JEM-X/IBIS) corresponds to revs.\ 100--119, and the green spectrum (IBIS) corresponds to revs.\ 164--179. The Fe $K_{\alpha}$ line, shown by the red dotted curve, is due to the Galactic ridge emission. The blue dotted curves denote the disk blackbody and power-law components of the models.}
354: \label{spectra}
355: \end{figure}
356:
357:
358: \begin{figure}
359: \centering
360: \includegraphics[angle=90, scale=0.7] {fig6.ps}
361: \caption{The spectra of IGR J17091--3624 during the rising part of the outburst (red: hard state; blue: soft state), compared with the spectra of IGR J17464--3213 also corresponding to the rising part of its outburst (dark green: soft state; light green: hard state). The blue and red dotted curves denote the same model components as in Fig.\ \ref{spectra}. The dark green dotted curves denote the blackbody and thermal-Compton components of the corresponding fit.}
362: \label{comparison}
363: \end{figure}
364:
365: \clearpage
366: %tables
367:
368: \begin{table}
369: \begin{center}
370: \caption{The log of the {\it INTEGRAL\/} observations of IGR J17091--3624 used for the spectral fits.}
371: \label{integral_log}
372: \begin{tabular}{lccc}
373: \hline
374: \hline
375: Revolution & Start Date & End Date & ISGRI exposure [ks] \\
376: \hline
377: 61--63$^{\mathrm{a}}$ & 2003-04-12 & 2003-04-21 & 15\\
378: 100--119 & 2003-08-10 & 2003-10-04 & 77\\
379: 164--185 & 2004-02-17 & 2004-04-02 & 77\\
380: \hline
381: \end{tabular}
382: \begin{list}{}{} \item[$^{\mathrm{a}}$]{Including the 2-ks {\it RXTE\/} observation on 2003-04-20.}
383: \end{list}
384: \end{center}
385: \end{table}
386:
387: \begin{table}
388: \begin{center}
389: \caption{The fit parameters to the PCA/IBIS spectrum of revs.\ 61--63 with the {\sc comtt} and {\sc compps} models (see the red spectrum in Figure \ref{spectra}).
390: \label{61_63}
391: }
392: \begin{tabular}{lccccc}
393: \hline
394: \hline
395: model$^{\mathrm{b}}$ & $kT_{\rm e}$& $ \tau$& $\chi^2$/$\nu$& d.o.f & $F$(3--100 keV)\\
396: & keV & & & & erg cm$^{-2}$ s$^{-1}$ \\
397: \hline
398: {\sc comptt} & $24^{+9}_{-4}$& $2.1^{+0.3}_{-0.5}$& 0.93& 54& $2.5\times 10^{-10}$\\
399: {\sc compps} & $31^{+12}_{-4}$ & $2.3^{+0.2}_{-0.4}$ & 0.96& 54 & $2.4\times 10^{-10}$ \\
400: \hline
401: \end{tabular}
402: \end{center}
403: \begin{list}{}{} \item[$^{\mathrm{b}}$]{$N_{H}<10^{22}$}
404: \end{list}
405: \end{table}
406:
407: \begin{table}
408: \begin{center}
409: \caption{The fit parameters to the JEM-X/IBIS spectrum of revs.\ 100--119 with the disk blackbody and a power law model (see the blue spectrum in Figure \ref{spectra}).}
410: \label{100_119}
411: \begin{tabular}{lccccccc}
412: \hline
413: \hline
414: Model & $kT_{\rm in}$ & $N_{\rm disk}$ & $\Gamma$ & $N_{\rm pl}$$^{\mathrm{c}}$ & $\chi^2$/$\nu$ & d.o.f. & $F$(3--100 keV)\\
415: &keV& & & & & & erg cm$^{-2}$ s$^{-1}$ \\
416: \hline
417: {\sc diskbb+pow} & $2.6^{+0.2}_{-0.2}$ & $1.9^{+0.7}_{-0.6}$ & $2.1^{+0.3}_{-0.4}$ & $0.2^{+0.4}_{-0.1}$ & 1.0 & 53 & $2 \times 10^{-9}$\\
418: \hline
419: \end{tabular}
420: \end{center}
421: \begin{list}{}{} \item[$^{\mathrm{c}}$]{${\rm keV}^{-1}\,{\rm cm}^{-2}\,{\rm s}^{-1}$ at 1 keV.}
422: \end{list}
423: \end{table}
424:
425: \begin{table}
426: \begin{center}
427: \caption{The fit parameters to the hard-state spectrum of the black-hole candidate IGR J17464--3213 (= H1743--322, Capitanio et al.\ 2005) with the {\sc comtt} and {\sc compps} models (see the light green spectrum in Figure \ref{comparison}).} \label{1746}
428: \begin{tabular}{lccccc}
429: \hline
430: \hline
431: model & $kT_{\rm e}$& $ \tau$& $\chi^2$/$\nu$& d.o.f & $F$(3--100 keV) \\
432: & keV & & & & erg cm$^{-2}$ s$^{-1}$ \\
433: \hline
434: {\sc comptt} & $18^{+4}_{-2}$ & $1.3^{+0.2}_{-0.4}$ &1.05&119 & $6\times 10^{-9}$\\
435: {\sc compps} & $20^{+3}_{-1}$ & $2.0^{+0.1}_{-0.2}$ & 1.04& 119 & $4 \times 10^{-9}$\\
436: \hline
437: \end{tabular}
438: \end{center}
439: %\begin{list}{}{} \item[$^{\mathrm{3}}$]{Normalization constant in XSPEC units}
440: %\end{list}
441: \end{table}
442:
443:
444:
445: \end{document}
446: