EDP Sciences
Free Access
Issue
A&A
Volume 593, September 2016
Article Number A96
Number of page(s) 12
Section Stellar structure and evolution
DOI https://doi.org/10.1051/0004-6361/201628808
Published online 29 September 2016

© ESO, 2016

1. Introduction

Supergiant fast X-ray transients (SFXTs) are high mass X-ray binaries (HMXBs) hosting most likely a neutron star (NS) and an OB supergiant companion (Sguera et al. 2005; Negueruela et al. 2006). Unlike normal supergiant HMXBs, which display a fairly constant average luminosity with typical variations by a factor of 1050 on time scales of a few hundred to thousands of seconds, SFXTs are characterised by hard X-ray flares reaching, for a few hours, 1036–1037 erg s-1 (see Romano et al. 2014b, for a catalogue of hard X-ray flares). SFXTs have also been found to be significantly subluminous with respect to classical supergiant HMXBs like Vela X-1 (Lutovinov et al. 2013; Bozzo et al. 2015), and show a soft X-ray dynamical range of up to six orders of magnitude (Sguera et al. 2005; Romano et al. 2015), since their luminosities can be as low as ~ 1032 erg s-1 during quiescence (e.g. in’t Zand 2005; Bozzo et al. 2010). The origin of this different behaviour is still unknown (see, e.g. Bozzo et al. 2013, 2015) and thus the different models proposed to explain the behaviour for these sources are still being debated. The models include a combination of more pronounced dense inhomogeneities (clumps) in the winds of the SFXT supergiant companions compared to those of classical systems (in’t Zand 2005; Walter & Zurita Heras 2007; Negueruela et al. 2008), magnetic/centrifugal gates generated by the slower rotational velocities and higher magnetic fields of the NSs hosted in SFXTs (Grebenev & Sunyaev 2007; Bozzo et al. 2008, 2016), or a subsonic settling accretion regime combined with magnetic reconnections between the NS and the supergiant stellar field transported by its wind (Shakura et al. 2014, and references therein).

Most SFXTs were first discovered, or classified as such, based on their hard X-ray properties (e.g. Sguera et al. 2005; Negueruela et al. 2006; Walter et al. 2015) as observed by INTEGRAL/IBIS (Ubertini et al. 2003). Subsequently, their long-term behaviour has also been extensively investigated (see Romano et al. 2014b) with other coded-mask large field-of-view instruments, such as the Swift/Burst Alert Telescope (BAT, Barthelmy et al. 2005). These hard X-ray monitors, however, share similiar sensitivity limits, which enable them to catch only the brightest portion of any transient event. Owing to their transient nature, SFXTs are indeed particularly difficult to find unless they experience frequent and relatively bright flares. A legitimate question is, therefore, whether we have already discovered the majority of Galactic SFXTs.

In a recent work, Ducci et al. (2014) address the issue of how common the SFXT phenomenon is, and conclude that, since detection of the entire population is hindered by the SFXT’s peculiar transient properties, a fraction of the population is probably yet to be identified, and that we are likely missing SFXTs with low outburst rates or large distances. Ducci et al. (2014), considered two datasets, the 100-month Swift/BAT catalogue (Romano et al. 2014b) and the first nine years of INTEGRAL/ISGRI data (Paizis & Sidoli 2014), applied two distinct statistical approaches to derive the expected number of SFXTs emitting bright flares in the Milky Way, . This value not only agrees with the expected number of HMXBs in the Galaxy derived from high-mass binary evolution studies (van den Heuvel 2012; Dalton & Sarazin 1995, and references therein), but also suggests that SFXTs constitute a sizeable fraction of X-ray binaries with supergiant companions. The SFXT class currently includes a mere dozen confirmed individuals, that is, X-ray binaries for which optical/IR spectroscopy has firmly established the presence of a supergiant primary of O or B spectral type (e.g. Romano et al. 2014b; Romano 2015). About as many candidate SFXTs are known, for which no optical spectroscopy has been obtained until now, but which have a reported history of bright, large dynamic range hard X-ray flaring. Since we expect a larger number of SFXTs in the Galaxy, it is worthwhile to increase the sample of these sources through new and archival multifrequency data studies of SFXT candidates and other promising unclassified transients. A larger sample of SFXTs would ultimately allow us to gain more information to understand the accretion mechanisms responsible for their enigmatic behaviour.

The Swift SFXT Project (Romano 2015) has been investigating the X-ray properties of the SFXTs since 2007, exploiting the unique observing capabilities of Swift (Gehrels et al. 2004). In particular, we have been performing follow-up observations of several tens of SFXT outbursts caught by the BAT (Romano et al. 2011b; Farinelli et al. 2012; Romano et al. 2015, and references therein), with the X-ray Telescope (XRT, Burrows et al. 2005) and the UV/Optical Telescope (UVOT, Roming et al. 2005), and have carried out long-term monitorings of virtually all SFXTs, also including five classical systems for comparison purposes. Swift broadband data of bright flares are particularly useful for increasing the SFXT sample, since they enable us to make a solid connection between the hard X-ray transient and its soft X-ray counterpart. As the XRT positional accuracy is as good as a few arcseconds, this type of association allows us to identify, in most cases, the optical/IR source associated with the X-ray transient and, subsequently, schedule dedicated optical spectroscopic campaigns to unveil its nature.

In this paper, we present the newly collected XRT and UVOT monitoring data of the three sources 2XMM J185114.3000004, IGR J174072808, and IGR J181752419, which showed an X-ray activity reminiscent of what is typically observed from the SFXTs during either the BAT triggers or the follow-up observations in the soft X-rays. Our main goal is to use the new data to investigate the associations of these systems with the SFXT class. We also supplement our data set by including: (i) serendipitous archival XMM-Newton observations of IGR J174072808; (ii) serendipitous archival XMM-Newton and ESO Very Large Telescope (VLT) observations of 2XMM J185114.3000004; (iii) archival INTEGRAL observations of IGR J181752419 carried out with the IBIS/ISGRI instrument.

2. The sample

2XMM J185114.3000004 is a source in the XMM-Newton XMMSSC catalogue (Watson et al. 2009; Lin et al. 2012) that triggered the BAT on 2012 June 17 (Barthelmy et al. 2012). At the time of discovery, the source showed an increase in the X-ray flux that was at least a factor of 40 compared to previous detections. Recently, Bamba et al. (2016) have analysed an ~ 100 ks Suzaku observation of the supernova remnant (SNR) G32.80.1 that serendipitously included 2XMM J185114.3000004 and found evidence of high time variability with flares on timescales of a few hundred seconds superimposed on a general decaying X-ray flux during the observation. No pulsations were found, but the flares were noticed as being spaced apart from one another by ~ 7000 s. The 310 keV spectrum was characterised by a high absorption (NH ~ 1023 cm-2) and a photon index Γ ~ 1.6, with a 210 keV flux of ~ 10-11 erg cm-2 s-1.

IGR J174072808 was discovered with INTEGRAL as a new transient on 2004 Oct. 9 (Götz et al. 2004; Kretschmar et al. 2004) and associated with either SBM2001 10 (Sidoli et al. 2001) or the ROSAT source 2RXP J174040.9280852. Based on its hard X-ray behaviour, Sguera et al. (2006) proposed it as a candidate SFXT since they noted peculiarly quick strong flares (2060 keV, peak flux 800 mCrab) lasting a couple of minutes. Heinke et al. (2009) found a likely association with the soft X-ray source CXOU J174042.0280724 (Tomsick et al. 2008). Following a BAT trigger on 2011 October 15, Romano et al. (2011a) identified CXOU J174042.0280724 as the soft X-ray counterpart of IGR J174072808. This in turn led to the identification of the candidate IR counterpart (Greiss et al. 2011; Kaur et al. 2011). If confirmed, this could disprove the SFXT hypothesis, since the most likely optical companion seems at present to be a late type-F dwarf.

The transient IGR J181752419 was serendipitously discovered (Grebenev 2013) in the IBIS/ISGRI data of the INTEGRAL observations performed in the direction of the X-ray nova SWIFT J174510.82624 on 2012 September 26. The characteristic short (1 h) flare displayed by the source at discovery, combined with its spectrum described by a hard power law (Γ ~ 2.1) and a possible exponential cut-off above 80 keV, suggested that IGR J181752419 could be a newly discovered SFXT source. No further detections of the source have been reported to date.

Table 1

Log of X-ray observations of 2XMM J185114.3000004 and spectral fit results.

Table 2

Log of X-ray observations of IGR J174072808 and spectral fit results.

3. Data reduction

The Swift and XMM-Newton observing logs for 2XMM J185114.3000004, IGR J174072808, and IGR J181752419 are reported in Table 13, respectively. The VLT observation logs are in Table 4.

3.1. Swift

The Swift data were uniformly processed and analysed using the standard software (FTOOLS1 v6.18), calibration (CALDB2 20160113), and methods. In particular, background-subtracted Swift/BAT light curves were created in the standard energy bands and mask-weighted spectra were extracted during the first orbit of the first automated target (AT) observation. We applied an energy-dependent systematic error vector to the BAT data. The Swift/XRT data were processed and filtered with the task xrtpipeline (v0.13.2). Pileup was corrected, when required, by adopting standard procedures (Vaughan et al. 2006; Romano et al. 2006). In these cases, the size of the point spread function (PSF) core affected by pile-up was determined by comparing the observed versus the nominal PSF, and excluding from the analysis all the events that fell within that region. In the case of 2XMM J185114.3000004 (Table 1), we extracted source events from annuli with inner/outer radii of 4/20 pixels (we note that for XRT, one pixel corresponds to 2.36′′) during the first observation. In all other cases, a circle with a radius of 20 pixels was adopted. The background events were extracted from an annular region with an inner radius of 80 pixels and an external radius of 120 pixels centered at the source position. For IGR J174072808 (Table 2), the source events were extracted from annuli with inner/outer radii of 5/20 pixels in the first observation, and a circular region with a radius of 20 pixels in all other cases. Background events were extracted from annuli with inner/outer radii of 60/120 pixels centred at the source position. The XRT light curves were corrected for PSF losses and vignetting by using the xrtlccorr tool and were background subtracted. In all observations, where no detection was achieved, the corresponding 3σ upper limit on the X-ray count rate was estimated by using the tasks sosta and uplimit within XIMAGE (with the background calculated in the neighbourhood of the source position) and the Bayesian method for low-count experiments adapted from Kraft et al. (1991). For our spectral analysis, we extracted events in the same regions as those adopted to create the light curve; ancillary response files were generated with the task xrtmkarf to account for different extraction regions, vignetting, and PSF corrections.

The Swift/UVOT observed the targets simultaneously with the XRT. It used all filters during AT observations and with the “Filter of the Day”, i.e. the filter chosen for all observations to be carried out during a specific day to minimize the filter-wheel usage, during all other observations. The data analysis was performed using the uvotimsum and uvotsource tasks included in FTOOLS. The uvotsource task calculates the magnitude of the source through aperture photometry within a circular region centered on the best source position and applies the required corrections related to the specific detector characteristics. We adopted a circular region with a radius of 5′′ for the photometry of the different sources. The background was evaluated in all cases by using circular regions with a radius of 10′′.

3.2. XMM-Newton

The XMM-Newton EPIC-pn (Strüder et al. 2001) and EPIC-MOS (Turner et al. 2001) observations of 2XMM J185114.3000004 and IGR J174072808 were processed by using the XMM-Newton Science Analysis Software (SAS, v. 15.0)3.

2XMM J185114.3000004 was serendipitously observed by XMM-Newton three times (see Table 1) on 2003 October 5 (ObsID 0017740401), on 2003 October 20 (ObsID 0017740501), and on 2012 March 18 (ObsID 0671510101). In all cases, the source was located at the very rim of the three EPIC cameras field of view (FOV). XMM-Newton observation data files (ODFs) for 2XMM J185114.3000004 were processed to produce calibrated event lists using the standard XMM-Newton SAS. We used the epproc and emproc tasks to produce cleaned event files from the EPIC-pn and MOS cameras, respectively. EPIC-pn and EPIC-MOS event files were extracted in the 0.510 keV energy range and filtered to exclude high background time intervals. The obs. 0671510101 was moderately affected by flaring background episodes. The cleaned effective exposure time was of 40.8 ks. The obs. 0017740401 was not affected by a high background, and thus we retained for the following analysis the entire exposure time available (21.4 ks for the EPIC-pn and 27.4 ks for the two MOS). In obs. 0017740501, cleaning for the high-level background resulted in an effective exposure time of 17.9 ks for the EPIC-pn and 21.2 ks for the two MOS cameras. Source and background spectra were extracted by using regions in the same CCD.

Table 3

Log of X-ray observations of IGR J18175241.

IGR J174072808 was also observed serendipitously with XMM-Newton (see Table 2) on 2016 March 6 (ObsID: 0764191301, PI. G. Ponti) during a Galactic centre lobe scan performed as an extension of the XMM-Newton scan (Ponti et al. 2015a,b). The source was located at an off-axis angle of about 7.5 arcmin from the aim point of all EPIC cameras, which were operating in full-frame mode using the medium filter. We removed an interval of increased background flaring activity at the end of the observation, yielding an effective exposure of 34.9, 36.4, and 36.3 ks for the pn, MOS1, and MOS2, respectively. Unfortunately, the source was located right at the edge of CCD1 in both MOS cameras and only a small fraction of the point spread function was covered, leading to great uncertainties in the flux reconstruction. For the MOS2, the uncertainty was too large to be useful for any scientific analysis, and thus we discarded these data. The EPIC-pn and MOS1 energy spectra and time series of the source and the background were extracted from circular regions. The radii of these regions were chosen to maximize the signal-to-noise ratio (S/N) by using the SAS tasks eregionanalyse and especget. We used single- and double-pixel events for the EPIC-pn camera and single- to quadruple-pixel events for the MOS1. For the energy spectra, events with FLAG 0 were discarded before binning the data to have S/N ≥ 5 in each bin. To produce a background-subtracted X-ray light curve of IGR J174072808, we selected single- and double-pixel events from the EPIC-pn camera in the 0.210.0 keV energy band and used a time binning of 200 s.

Table 4

Optical data on 2XMM J185114.3000004: VLT/NACO and Swift/UVOT observations.

3.3. ESO VLT

We observed the field of 2XMM J185114.3000004 with the ESO VLT equipped with NAos COnica (NACO), the adaptive optics (AO) NIR imager and spectrometer mounted at the VLT UT4 telescope, in the J and Ks-bands. Observations were carried out on 2012 July 12 starting at 07:36:01.548 UT (see Table 4). We used the S27 camera, which has a pixel size of 0.027″ and a FOV of 28″ × 28″. The visual dichroic element and wavefront sensor were used. Image reduction was carried out using the NACO pipeline data reduction, part of the ECLIPSE4 package. Unfortunately, the observations were affected by natural seeing in excess of 1″, with a resulting poor resolution. Astrometry was carried out by using the 2MASS5 catalogues as reference. Aperture photometry was performed with the PHOTOM task of the STARLINK package6. The photometric calibration was made against the 2MASS catalogue.

3.4. Other data

INTEGRAL data were only used for the source IGR J181752419 and the corresponding results are described in Sect. 4.3.

4. Results

4.1. 2XMM J185114.3000004

The source 2XMM J185114.3000004 triggered the BAT on 2012 June 17 at T0 = 15:46:55 UT (64 s image trigger = 524542, Barthelmy et al. 2012), resulting in a 7σ detection. Swift performed an immediate slew to the target and XRT started observing at T0 +172 s. The AT (sequence 00524542000-001) ran for seven orbits until T0 + 17.5 ks. Follow-up observations were obtained daily (sequences 00524542002010). Additional target of opportunity (TOO) observations were first performed when the source rebrightened a few days later (PI P. Romano, sequences 00524542011017), and then also on August 831 of the same year (PI P. Romano, sequences 00032512001005). The Swift data therefore cover the first 18 days after the beginning of the outburst, and then about three weeks more, later that year (see Table 1). The source is only detected in five observations (005245420001 and 9, 00032512001 and 5). We also found an archival Swift serendipitous observation performed on 2012 November 20 (00044565001), which also resulted in a non-detection of the source.

We used 4 ks of the XRT/PC mode data collected during the outburst in 2012 June 17 and the simultaneous Swift/UVOT images to obtain an astrometrically corrected source position (see Evans et al. 2009; Goad et al. 2007) at: RA(J, Dec(J (90% confidence level, c.l., uncertainty of 1.̋7). This position is from the catalogued position of 2XMM J185114.3000004. It is also from the Two Micron All Sky Survey source 2MASS J185114470000036 (Skrutskie et al. 2006).

thumbnail Fig. 1

Light curves of the outburst in 2012 June 17 of 2XMM J185114.3000004 (first Swift orbit data). a) BAT light curve in the 1450 keV with a time binning of 20 s. b) XRT light curve in the 0.210keV, rebinned to have at least 10 counts bin-1. Note the different x-axis scales.

Open with DEXTER

thumbnail Fig. 2

Swift/BAT and XRT light curve obtained from the entire XRT dataset on 2XMM J185114.3000004. Grey downward-pointing arrows correspond to the 3σ upper limits.

Open with DEXTER

Figure 1 shows the Swift BAT (1450 keV) and XRT (0.210 keV) light curves of the source extracted from the first orbit of data collected at the beginning of the 2012 June outburst, while Fig. 2 shows the light curve derived from the whole XRT dataset. The XRT light curve reached a maximum of 4.1 counts s-1. This corresponds to an approximate flux of 7.9 × 10-10 erg cm-2 s-1 (we used the conversion factor derived from the fit to the XRT data in observation 00524542000 PC mode, see Table 1). The lowest X-ray flux from the source was recorded at 3.8 × 10-3 counts s-1 during the observation 00524542001, thus resulting in an XRT dynamical range ≳ 103.

The BAT spectrum of the source extracted from observation 00524542000 was fit in the 1470 keV energy range with a simple power law (see Table 1). The XRT spectrum extracted from observation 00524542000 was fit in the 0.310 keV energy range with an absorbed power law (see Table 1). We then considered the total BAT spectrum and the nearly simultaneous XRT spectrum for a broadband fit. Factors were included in the fitting to allow for the different exposures of the two spectra, normalisation uncertainties between the two instruments (generally constrained within ~ 10%), and a likely spectral variation throughout the exposure.

Several models typically used to describe the X-ray emission from accreting pulsars in HMXBs were adopted (e.g. White et al. 1983; Coburn et al. 2002; Walter et al. 2015, and references therein), including: (i) an absorbed power law (hereafter POW); (ii) an absorbed power law with high-energy exponential roll-off at an e-folding energy Ef (PHABS*CUTOFFPL in XSPEC, hereon CPL); (iii) and an absorbed power law with a high-energy cut-off at an energy Ec, and an e-folding energy Ef (PHABS*POWER*HIGHECUT, hereon HCT). The results for 2XMM J185114.3000004 are reported in Table 5. The PL model clearly yielded unacceptable results. The CPL model produced significantly better results and gave NH = (13 ± 4) × 1022 cm-2, , and keV (see Fig. 3). These values are compatible with those usually expected for highly magnetised accreting NSs, SFXTs in particular (e.g. Romano et al. 2011c). The HCT model is, on the other hand, unable to constrain the cut-off energy (see Table 5); since the addition of one free parameter does not improve the statistics significantly (F-test probability of 0.269 with respect to the CPL model), we favour the CPL model.

The UVOT data obtained simultaneously with the XRT ones only yield 3σ upper limits in all filters (see Table 4). This is not surprising, given that the reddening along the line of sight (LOS) is E(BV) ~ 19, implying an extinction of AV ~ 60 mag.

thumbnail Fig. 3

Spectroscopy of the 2012 June 17 outburst of 2XMM J185114.3000004. The top panel shows simultaneous XRT/PC (filled red circles) and BAT data (empty blue circles) fit with a PHABS*CUTOFFPL model. The residuals from the best fit are shown in the bottom panel (in units of standard deviations).

Open with DEXTER

Table 5

Spectral fits of the simultaneous XRT and BAT data of 2XMM J185114.3000004 during the outburst on 2012 June 17.

thumbnail Fig. 4

XMM-Newton FOV of the observations 0017740401 and 0017740501, detector background subtracted, vignetting corrected, combining EPIC-pn and MOS. Red, green, and blue correspond to 0.52.0 keV, 2.04.5 keV, and 4.512 keV. The diffuse emission around 2XMM J185114.3000004 is due to SNR G32.80.1.

Open with DEXTER

thumbnail Fig. 5

Merged XMM-Newton EPIC-pn+MOS1+MOS2 spectra extracted from the observations 0017740401 (black) and 0017740501 (red). The best-fit model is obtained with an absorbed power law. The residuals from the fit are shown in the bottom panel.

Open with DEXTER

By using our ESO VLT observations, inside the XRT error circle we detect a single source at the following position: RA(J, Dec(J (90% c.l. associated uncertainty ). At the epoch of our observation, we measure for this source J = 15.8 ± 0.1 mag and Ks = 11.7 ± 0.1 mag (Vega system; not corrected for Galactic extinction, see Table 4). This source is present in the 2MASS catalogue (2MASS J185114470000036), with magnitudes J> 15.6 mag, H = 13.23 ± 0.07 mag, and K = 11.80 ± 0.04 mag, in agreement with our measurements.

2XMM J185114.3000004 was detected by the EPIC-pn at an average count-rate (0.510 keV) of (9.4 ± 0.7) × 10-3 count s-1 in obs. 0017740401 and (3.5 ± 0.5) × 10-3 count s-1 in obs. 0017740501. The corresponding EPIC-pn spectrum could be well fit in both cases, with an absorbed power-law model (see Table 1).

Given the relatively low count-rate of the source, we combined in each of the two observations the EPIC-pn and MOS spectra by following the online SAS data analysis threads7. The resulting spectra (shown in Fig. 5) provided values for the best-fit parameters in agreement with those estimated above by using only the EPIC-pn data (to within the uncertainties).

In obs. 0017740401, where the statistics was better, we also inspected the source light curve and event file, searching for timing features. However, the statistics was far too poor to provide a meaningful timing analysis.

The source was not detected in obs. 0671510101. From the EPIC-pn data we estimated a 3σ upper limit on the source count-rate of 8 × 10-4 in the 0.510 keV energy range. Assuming the same spectral shape as in obs. 0017740501, the count-rate upper limit would translate into a flux of F0.5−10 keV< 4 × 10-15 erg cm2 s-1.

thumbnail Fig. 6

Light curves of the 2011 October 15 outburst of IGR J174072808 (first Swift orbit data). a) BAT light curve in the 1450 keV energy band and a binning of 10 s. The horizontal lines mark the time intervals used for the spectral extraction (peak1 and peak2). b) The XRT light curve in the 0.210 keV energy band. An adaptive binning has been used to achieve in each point a signal-to-noise ratio of S/N = 3.

Open with DEXTER

4.2. IGR J174072808

The source IGR J174072808 triggered the BAT on 2011 October 15 at T0 = 01:12:40 UT (image trigger = 505516, Romano et al. 2011a). Swift immediately slewed to the target and XRT started observing at T0 +131 s. The AT (sequence 00505516000) ran for two orbits until T0 + 6.6 ks. However, owing to a loss of lock of the star tracker, only the first orbit has a stable attitude and can be used for scientific analysis. Follow-up pointings toward the source were obtained daily as TOO observations (PI P. Romano, sequences 00036122001007, see Table 2). The source was only detected during the AT observation (00505516000), but never in the following monitoring campaigns and in the serendipitous observations found in the Swift archive (see Table 2).

We used 772 s of XRT/PC mode data and the simultaneously collected UVOT images to obtain the best astrometrically-corrected source position at: RA(J, Dec(J. The associated uncertainty at 90% c.l. is 2.̋4. This position is from CXOU J174042.0 280724, so we can confirm the association between the two sources, as preliminarily reported by Romano et al. (2011a).

Figure 6 shows the BAT (1450 keV) and XRT (0.210 keV) light curves of the first orbit of data while, in Fig. 7, we plot the light curve derived from the whole XRT dataset. Around T0 + 900 s, the source displayed a sharp rise in count-rate increasing the level of its soft X-ray emission by a factor of 125 in ~ 266 s. A peak count-rate of about 14 counts s-1 is recorded in the light curve binned at S/N = 3. This corresponds to a flux of ~ 2.3 × 10-9 erg cm-2 s-1 when using the conversion factor derived from the fit to the XRT data in observation 00505516000 (see Table 2). Since the lowest point (obtained from observation 00036122007) was a 3σ upper limit at 9.5 × 10-3 counts s-1, the overall dynamical range revealed by XRT is in excess of ~ 1400.

thumbnail Fig. 7

Light curve of the XRT dataset of IGR J174072808. Grey downward-pointing arrows correspond to the 3σ upper limits calculated for the source non-detections. Points after 107 s correspond to serendipitous observations.

Open with DEXTER

We extracted two distinct BAT spectra that cover the two peaks in the BAT light curve (see Fig. 6a), i.e. from T0−20 to 80 s (peak1) and from T0 + 858 to 965 s (peak2). We fit them with a power law and the results are reported in Table 2).

The XRT spectrum extracted from obs. 00505516000 was fit in the 0.310 keV energy range using Cash (1979) statistics with an absorbed power law. We measured an absorption column density cm-2 consistent with the expected Galactic value in the direction of the source ( cm-2, Kalberla et al. 2005), and a power-law photon index , as reported in Table 2.

Since no spectral variations could be detected in either the BAT data or in the XRT data owing to the low signal, we fit together the BAT peak2 spectrum and the XRT spectrum extracted by using all exposure time available. Factors were included in the fitting to allow for normalisation uncertainties between the two instruments, the different exposures of the two spectra, and a likely spectral variation throughout the exposure.

The spectra were fit with the models that typically describe either HMXBs (see Sect. 4.1), or low mass X-ray binaries (LMXB, Done et al. 2007; Paizis et al. 2006; Del Santo et al. 2010; Wijnands et al. 2015). The results are presented in Table 6. The POW fit was not acceptable because of large residuals. The CPL fit yields and  keV (see Fig. 8). In this fit, we fixed the absorption column density to the value determined from the XRT data alone, i.e. NH = 0.84 × 1022 cm-2. The HCT model cannot properly constrain the cut-off energy; since the addition of one free parameter does not significantly improve the fit (F-test probability of 0.423), we favour the CPL model. A fit with an absorbed black body (with NH fixed at 0.84 × 1022 cm-2) also yielded acceptable results. In this case, the estimated black-body temperature of about 8 keV would be roughly consistent with that reported previously by Sguera et al. (2006). A fit with an absorbed bremsstrahlung model yields a very high temperature and suffers from large systematics in the residuals, so we consider it unacceptable. Fits with more sophisticated models, such as a Comptonisation model (PHABS*(COMPTT)) and an accretion disk model with multiple black-body components (PHABS*(DISKBB+POWER)) could not significantly improve the fits. Furthermore, the majority of the spectral fit parameters of these models turned out to be largely unconstrained owing to the limited statistics of the data.

The source was not detected in any UVOT data obtained simultaneously with XRT. The corresponding 3σ upper limits in all filters were of u = 20.23, b = 18.61, v = 17.56, m2 = 20.90, w1 = 20.64, and w2 = 20.33 mag (Vega system, not corrected for Galactic extinction).

IGR J174072808 was also observed serendipitously during an XMM-Newton observation performed on 2016 March 6. The 0.210.0 keV energy band light curve, at a binning of 200 s, reported in Fig. 9 shows three moderately bright flares, reaching between 0.1 and 0.3 counts s-1. The XMM-Newton spectra of the source could be well described by using an absorbed power-law model (Fig. 10), and an cm-2, consistent with the Galactic value, and with the results obtained with the XRT data. The absorbed 0.510 keV flux is 5.4 × 10-13 erg cm-2 s-1, the lowest X-ray flux measured for this source, enhancing its previously estimated dynamic range (with XRT) up to > 4000.

thumbnail Fig. 8

Spectroscopy of the 2011 October 15 outburst of IGR J174072808. Top panel: simultaneous XRT/PC data (filled red circles) and BAT data (empty blue circles) fit with a PHABS*CUTOFFPL model. Bottom panel: the ratio between the data and the best-fit model.

Open with DEXTER

Table 6

Spectral fits of the simultaneous XRT and BAT data of IGR J174072808 during the outburst on 2011 October 15.

4.3. IGR J181752419

IGR J181752419 was observed by Swift as part of our ongoing effort (Romano 2015) to study SFXTs, candidate SFXTs, and classical supergiant HMXBs through long-term monitoring programs with the XRT (see Romano et al. 2014a, for recent results). Our monitoring campaign (see Table 3) was performed from 2016 March 6 to 14 for 1 ks a day. As the source was poorly known, the XRT was set in AUTO mode to best exploit the automatic mode-switching of the instrument in response to changes in the observed fluxes (Hill et al. 2004). We collected a total of 9 Swift observations for a total net XRT exposure time of ~ 7 ks. No source was detected within the previously reported INTEGRAL error circle (Grebenev 2013) in either any of the 1 ks snapshots or in the combined total exposure. We estimated a 3σ upper limit on the source X-ray count rate of 1.5−3 × 10-3 counts s-1, which corresponds to 0.8−1.7 × 10-13 erg cm-2 s-1 (when using, within PIMMS, a typical spectral model for SFXTs, comprising a power law with photon index Γ = 1.5 and an absorption column density corresponding to the Galactic value in the direction of the source, i.e. cm-2).

thumbnail Fig. 9

XMM-Newton EPIC-pn light curve of IGR J174072808 extracted from the observation 0764191301, with a 200 s binning.

Open with DEXTER

thumbnail Fig. 10

XMM-Newton EPIC-pn (black) and MOS1 (red) spectra extracted from observation 0764191301. The best fit is obtained by using an absorbed power-law model. The residuals from the fits are shown in the bottom panel.

Open with DEXTER

thumbnail Fig. 11

Left: IBIS/ISGRI mosaic extracted with the OSA software from the SCW 51 in revolution 1215 (2080  keV energy band). We indicated (green squares) on the mosaic the position of 4 sources detected with a sufficiently high significance (SWIFT J174510.82624, GRS 1758258, GX 1+4, IGR J174643213) and that previously reported for IGR J181752419. The latter is indicated by using a yellow circle with a radius of 24 centred on the source best position provided by Grebenev (2013). Right: significance map obtained with the BATIMAGER software (Segreto et al. 2010a) in the 2080  keV energy band for the SCW 51. Green circles (24 radius) shows the positions of all sources significantly detected in the field, as well as the position of IGR J181752419, as reported by Grebenev (2013). In either mosaic, independently built with different software, we do not detect any significant emission from IGR J181752419. The bars at the bottom of each mosaic indicate the colour codes for the detection significances in units of standard deviations.

Open with DEXTER

As no significant emission was detected with Swift /XRT at the best known position of IGR J181752419 down to a luminosity that is usually fainter than that of SFXTs in quiescence, we reanalysed the INTEGRAL data in which the source was discovered. INTEGRAL observations are divided into so-called science windows (SCWs), i.e. pointings with typical durations of ~23 ks. The only source detection is reported in Grebenev (2013), where the IBIS/ISGRI data collected in the direction of the X-ray nova SWIFT J174510.82624 on 2012 September 26 at 14:57 (UT) were analysed using the software developed at the Space Research Institute of the Russian Academy of Sciences (Revnivtsev et al. 2004; Krivonos et al. 2010). The detection of the source with the highest significance was obtained during the SCW 51 in the satellite revolution 1215. The source was also reported to be visible during the first 10 min of the SCW 52 in the same revolution, albeit with a lower significance. We analysed these two SCWs by using version 10.2 of the Off-line Scientific Analysis software (OSA) distributed by the ISDC (Courvoisier et al. 2003).

Following Grebenev (2013), we first extracted the IBIS/ISGRI mosaic of the SCW 51 in the 2080 keV energy band and searched for the source previously reported at the best determined position of RA(J, Dec(J (the associated uncertainty is 4). The mosaic is shown in Fig. 11 (left). We did not detect any significant emission around the position of the source, which lies at an off-axis angle of about 14 deg from the center of the instrument FOV. At these large off-axis positions, the calibration of the instrument becomes gradually more uncertain and it is difficult to estimate a reliable upper-limit on the X-ray flux. By using the mosaic_spec tool, the derived 3σ upper limit on any source count-rate at this position is of 8.4 counts s-1 in the 2080 keV energy band (effective exposure time 3281 s), corresponding to roughly 40 mCrab8 (i.e., 6 × 10-10 erg cm-2 s-1). This is a factor of 2.5 lower than the flux of the source reported by Grebenev (2013).

For completeness, we also extracted the IBIS/ISGRI mosaic of the combined SCW 50, 51, and 52. This matches the time interval of the light curve shown in Fig. 2 of Grebenev (2013). Also in this case, no significant emission from the position of IGR J181752419 is detected. The estimated 3σ upper limit is 15 mCrab (i.e., 2 × 10-10 erg cm-2 s-1) in the 2080 keV energy band (effective exposure time 9890 s).

To further check the source detection, we used an independently developed software, the BATIMAGER, designed to generate sky maps for generic coded mask detectors and optimised, in particular, for the processing of Swift/BAT and INTEGRAL/ISGRI data. Detailed description and performance of the software when applied on the BAT survey data is given in Segreto et al. (2010a), while its imaging performance, when applied to the IBIS/ISGRI data, are reported in Segreto et al. (2008, 2010b). Figure 11 (right) shows the significance map of the region around IGR J181752419 obtained with the BATIMAGER software in the 2080 keV energy band by using the data in SCW 51. The significantly detected sources in the image are indicated with green circles, confirming that no source is present at the previously reported position of IGR J181752419 (Grebenev 2013).

Given the above results, we suggest that the source IGR J181752419 was erroneously reported. This might have occurred owing to some mosaic reconstruction problem at the large off-axis angle where IGR J181752419 should have been located. We thus do not discuss this source any further in this paper.

5. Discussion and conclusions

5.1. 2XMM J185114.3000004

The BAT image trigger on the transient 2XMM J185114.3 000004 was a long (64 s) and strong (7σ) one. The XRT arcsecond position we provide for this source (refinement of Barthelmy et al. 2012), with an uncertainty of 1.̋7 at 90% c.l., is consistent with the catalogued position of 2XMM J185114.3000004 and is only from 2MASS J185114470000036 (J> 15.6, H = 13.23 ± 0.07, K = 11.80 ± 0.04). This IR source, which we observed with NACO at VLT obtaining J = 15.8 ± 0.1 mag and Ks = 11.7 ± 0.1 mag, can therefore be safely considered the IR counterpart of 2XMM J185114.3000004 (finally removing the need for the cautionary statements on this association raised by Bamba et al. 2016). Even though optical spectroscopic data for this source are not yet available, and the presence of a supergiant companion cannot be firmly established, we show below that all measured properties favour the association of 2XMM J185114.3000004 with the SFXT class.

The BAT light curve of the source (Fig. 1) starts with a bright flare, lasting about 100 s (FWHM, when fit with a Gaussian), that reached about ~ 10-9 erg cm-2 s-1 (1550 keV). The overall duration of the outburst is much longer than this (Fig. 2), since XRT caught several flares in the monitoring observations following the main event. The brightest of these flares occurred at T ~ T0 + 1050 s and reached ~ 8 × 10-10 erg cm-2 s-1 (0.210 keV). After the first slew, the XRT data show a steep decay lasting until T + 23 ks, which is typical of the SFXT population (see Fig. 4, in Romano 2015), with a behaviour strongly reminiscent of, for example, the 2005 August 30 flare of IGR J164794514. The dynamical range of the initial flare is about three orders of magnitude, but the overall soft X-ray dynamical range reaches 4000 when including the archival XMM-Newton data. We note that, when using a distance of 12 kpc (see below), the observed luminosities range between L ~ 5 × 1033 and 2.6 × 1037 erg s-1. This is the typical range of an SFXT source.

The broadband spectrum of 2XMM J185114.3000004 (Fig. 3), presented here for the first time, can be well described by an absorbed cut-off power law model, with and keV, values which are well within the distribution of parameters measured for HMXB hosting relatively young NSs (see Romano 2015). The measured absorption (NH = (13 ± 4) × 1022 cm-2) is much larger than the expected Galactic value in the direction of the source ( cm-2, Kalberla et al. 2005), as expected in SFXTs owing to the presence of a dense stellar wind. We note (see Fig 4) that there is diffuse emission along the LOS to 2XMM J185114.3000004 that may contribute to the reddening. This emission is uncorrelated to 2XMM J185114.3000004 and due to the nearby supernova remnant SNR G32.80.1 as discussed in Bamba et al. (2016).

Assuming therefore a blue supergiant nature for the donor star of the 2XMM J185114.3000004 system, we can determine the reddening toward the source by considering its intrinsic NIR color as per Wegner (1994). In the present case, we obtain JK = 4.1, whereas the intrinsic value of this colour is (JK)0 ~ 0 for early-type supergiants, which implies a color excess of E(JK) ~ 4.1. Using the Milky Way extinction law of Cardelli et al. (1989), this implies a reddening AV ≈ 20 mag, or AK ≈ 2 mag. This absorption amount is consistent with the non-detection of the object at ultraviolet and optical bands, as found from the UVOT data analysis.

This quantity of reddening, when using the formula of Predehl & Schmitt (1995), implies a column density NH ~ 3.6 × 1022 cm-2, which is lower than both that of the Galaxy along the LOS of the source (see Sect. 4.1), and that measured in X-rays. This indicates that (i) the object lies within the Galaxy and (ii) additional X-ray extinction is present around the accretor, likely produced by the accreting material as observed in many other HMXBs and SFXTs.

As regards point (i), we can try to estimate the distance of the system under the assumption that it hosts an OB supergiant. Using AV ~ 20 mag and the tabulated absolute magnitudes (Lang 1992) and colours (Wegner 1994) for this type of stars, we find a distance of ~12 kpc. This would place the source on the Galactic plane beyond the Sagittarius-Carina arm tangent and close to (or possibly within) the Perseus arm, according to the Galaxy map of Leitch & Vasisht (1998).

5.2. IGR J174072808

The nature of the hard X-ray transient IGR J174072808 has been quite controversial since its discovery. Sguera et al. (2006) reported three short bright flares reaching 10-8 erg cm-2 s-1 in the 2060 keV band and proposed the association of this object with the SFXT class. These authors could not exclude different possibilities, as that of a new “burst-only” source (Cocchi et al. 2001; Cornelisse et al. 2002). The search for a soft X-ray counterpart for IGR J174072808 led Heinke et al. (2009) to propose CXOU J174042.0280724 as a likely candidate because of the observed flaring variability common to both systems. Their Chandra observations revealed CXOU J174042.0280724 as a fast transient source, varying around a level of 10-13−10-12 erg cm-2 s-1 and thus suggesting the presence of an accreting black hole or neutron star. Heinke et al. (2009) ruled out the presence of a supergiant companion for any distance larger than 10 kpc, and showed that an LMXB or a Be X-ray binary beyond 10 kpc were more likely possibilities. Based on our preliminary position (Romano et al. 2011a), Greiss et al. (2011) found a candidate NIR counterpart from the Chandra position in archival VVV survey data. They find moderate reddening along its LOS and de-reddened optical and NIR magnitudes and colours consistent with a late type-F dwarf (at a distance of ~3.8 kpc). This would thus also disfavour the SFXT hypothesis, unless the F star is a foreground object. Kaur et al. (2011) confirmed the values of the previously reported optical magnitudes and found that the candidate counterpart was about one magnitude brighter four days after the detected flares observed by Swift. This is something expected in outbursting LMXBs owing to the irradiation of the optical star by the X-rays emitted from the compact object.

Thanks to the fact that IGR J174072808 triggered the BAT, we obtained simultaneous soft X-ray coverage of the source while it was rapidly decaying, and we can now unequivocally establish that the soft X-ray counterpart of IGR J174072808 is indeed CXOU J174042.0280724 (see our preliminary results in Romano et al. 2011a). The Swift/BAT light curve (Fig. 6a) shows at least two bright flares reaching a few 10-9 erg cm-2 s-1 (1550 keV), whose profiles are symmetrical and narrow, lasting about 1520 s (FWHM, when fitted with a Gaussian shape). These values are about a factor of 10 shorter than has been measured for typical SFXTs and for 2XMM J185114.3000004. The second peak (peak2) is also clearly correlated with a soft X-ray flare (Fig. 6b) that reached ~ 2 × 10-9 erg cm-2 s-1 (0.510 keV) and that lasted at least 10 s, as the flare is truncated as a result of the satellite slew, so only a lower limit on its duration is available. When the source was once again within the XRT FOV, it was already below detection (Fig. 7), and all subsequent XRT observations of ≲ 1 ks exposures, never revealed the source again, yielding individual 3σ upper limits at the level of 10-12 erg cm-2 s-1 and a combined 3σ upper limit of 1.1 × 10-12 erg cm-2 s-1. The serendipitous XMM-Newton observations that we analysed show, consistent with what has been seen in the Chandra data (Heinke et al. 2009), a relatively steady flux of a few 10-13 erg cm-2 s-1 with three equally symmetrical and narrow flares reaching 10-12 erg cm-2 s-1 (0.510 keV) and lasting between 190 and 385 s (FWHM). We note that such flares would go undetected in XRT observations of ~1 ks. All the results thus confirmed that this is most likely the truly quiescent behaviour of IGR J174072808.

The source is therefore characterised by a relatively low, steady flux and infrequent episodes of more pronounced activity, with several short bright flares closely spaced in time (~ 103 s). These flares often reach an X-ray flux that is only a few times higher than the persistent level but more rarely can achieve an X-ray dynamic range as high as 4 orders of magnitude. As the distance is unknown, the range of observed fluxes would correspond to luminosities of L ~ 1033−1037 erg s-1 at 3.8 kpc (see before), or L ~ 1034−1038 erg s-1 at 13 kpc.

The Swift AT data also provide the first simultaneous broad-band spectroscopy for this transient. We obtained satisfactory fits with either an absorbed power law with a high energy roll-over or an absorbed black body. The first model provided results that are reminiscent of those measured from SFXTs (see, e.g. Romano et al. 2013; Romano 2015, for a review of the Swift spectra). The negative photon index remains, however, puzzling. The black-body model provides a quite unlikely high value of the temperature for an accreting object hosted in either an HMXB or LMXB. In both fits with the power law with a high-energy roll-over and the black body, the absorption needed to be fixed to that obtained from the fit to the XRT data alone and was consistent with being as low as cm-2. This is comparable with the expected Galactic extinction in the direction of the source ( cm-2, Kalberla et al. 2005). In principle, this is much lower than the absorption column density expected for an HMXB or an SFXT, thus favouring the LMXB hypothesis. However, we also note that the SFXT IGR J084084503 usually displays an absorption column density 1022 cm-2 (Bozzo et al. 2010; Sidoli et al. 2010). On the other hand, the NH obtained from the XRT data only, as well as from the fit with an absorbed power law with a high energy cut-off could still be consistent with the absorption column density expected for an HMXB or an SFXT.

The information collected so far on IGR J174072808 is therefore difficult to interpret. In the framework of the HMXB/SFXT nature of IGR J174072808, pro factors are the light curve flaring, which is characterized by at leat two orders of magnitude during the flare and the overall dynamic range of about four orders of magnitude. The spectral properties are also reminiscent of what is usually observed from these systems, even though the broadband properties of these systems are not always a distinctive feature (Romano et al. 2014a). Against the HMXB/SFXT hypothesis is the low LOS absorption, making it unlikely to hide a supergiant behind the F star, even when the F star is only a foreground source. The X-ray flares displayed by IGR J174072808 are also quite short compared to those of the SFXTs, which typically last a few thousands of seconds. The broadband spectral properties and the optical counterpart could be roughly consistent with those of a LMXB but the short flares displayed by IGR J174072808 and the fast variability are not commonly observed in LMXBs. The latter typically undergo weeks- to months-long outbursts regulated by the accretion disk instability (Lasota 2001) and might show additional flaring activity on top of this (see, for example, McClintock & Remillard 2006, for a review). The variability recorded by Swift and shown in Figs. 6 and 7 is thus not closely reminiscent of that displayed by either black holes or NS LMXBs in outburst. We note that the profiles and durations of the short flares emitted from IGR J174072808 do not resemble those of type-I X-ray bursts, thus also excluding the association between this object and the so-called burst only sources (Cornelisse et al. 2004). Given the relatively high luminosities recorded by Swift (L ~ 1033−1037 erg s-1) for any reasonable estimate of the source distance 3.8 kpc, we consider it an even less likely possibility that IGR J174072808 is a white-dwarf binary (Sazonov et al. 2006). Similarly unlikely is the possibility that IGR J174072808 is a a very faint X-ray transient (VFXT, King & Wijnands 2006; Wijnands et al. 2006), since these objects show a photon index of about 1.52.2 in this luminosity range (see Degenaar & Wijnands 2010; Del Santo et al. 2007) and similar behaviour, even when showing hybrid outbursts (faint and bright; see Del Santo et al. 2010).

We conclude that IGR J174072808 is most likely an LMXB hosting an accreting compact object. However, the detailed nature of this source remains concealed and spectroscopic follow-up of the candidate optical counterpart are encouraged to achieve a more precise classification.


8

The conversion from count-rate to mCrab was carried out by using the most recent observations of the Crab (at the time of writing) in spacecraft revolution 1597. From these data, we measured for the Crab a count-rate of 214.6 ± 0.3 counts s-1 from the IBIS/ISGRI mosaics in the 2080  keV energy band.

Acknowledgments

We thank the Swift team duty scientists and science planners for their courteous efficiency, and M. Capalbi, M. de Pasquale, P. A. Evans for helpful discussions. We also thank the referee for comments that helped improve the paper. P.R., B.S., and S.C. acknowledge contract ASI-INAF I/004/11/0. P.E. acknowledges funding in the framework of the NWO Vidi award A.2320.0076 (PI: N. Rea). L.D. acknowledges support by the Bundesministerium für Wirtschaft und Technologie and the Deutsches Zentrum für Luft und Raumfahrt through the grant FKZ 50 OG 1602. Based on observations made with ESO Telescopes at the Paranal Observatory under programme ID 089.D-0245(A). XMM-Newton is an ESA science mission with instruments and contributions directly funded by ESA Member States and NASA. The XMM-Newton project is supported by the Bundesministerium für Wirtschaft und Technologie/Deutsches Zentrum für Luft- und Raumfahrt (BMWI/DLR, FKZ 50 OR 1408) and the Max-Planck Society.

References

All Tables

Table 1

Log of X-ray observations of 2XMM J185114.3000004 and spectral fit results.

Table 2

Log of X-ray observations of IGR J174072808 and spectral fit results.

Table 3

Log of X-ray observations of IGR J18175241.

Table 4

Optical data on 2XMM J185114.3000004: VLT/NACO and Swift/UVOT observations.

Table 5

Spectral fits of the simultaneous XRT and BAT data of 2XMM J185114.3000004 during the outburst on 2012 June 17.

Table 6

Spectral fits of the simultaneous XRT and BAT data of IGR J174072808 during the outburst on 2011 October 15.

All Figures

thumbnail Fig. 1

Light curves of the outburst in 2012 June 17 of 2XMM J185114.3000004 (first Swift orbit data). a) BAT light curve in the 1450 keV with a time binning of 20 s. b) XRT light curve in the 0.210keV, rebinned to have at least 10 counts bin-1. Note the different x-axis scales.

Open with DEXTER
In the text
thumbnail Fig. 2

Swift/BAT and XRT light curve obtained from the entire XRT dataset on 2XMM J185114.3000004. Grey downward-pointing arrows correspond to the 3σ upper limits.

Open with DEXTER
In the text
thumbnail Fig. 3

Spectroscopy of the 2012 June 17 outburst of 2XMM J185114.3000004. The top panel shows simultaneous XRT/PC (filled red circles) and BAT data (empty blue circles) fit with a PHABS*CUTOFFPL model. The residuals from the best fit are shown in the bottom panel (in units of standard deviations).

Open with DEXTER
In the text
thumbnail Fig. 4

XMM-Newton FOV of the observations 0017740401 and 0017740501, detector background subtracted, vignetting corrected, combining EPIC-pn and MOS. Red, green, and blue correspond to 0.52.0 keV, 2.04.5 keV, and 4.512 keV. The diffuse emission around 2XMM J185114.3000004 is due to SNR G32.80.1.

Open with DEXTER
In the text
thumbnail Fig. 5

Merged XMM-Newton EPIC-pn+MOS1+MOS2 spectra extracted from the observations 0017740401 (black) and 0017740501 (red). The best-fit model is obtained with an absorbed power law. The residuals from the fit are shown in the bottom panel.

Open with DEXTER
In the text
thumbnail Fig. 6

Light curves of the 2011 October 15 outburst of IGR J174072808 (first Swift orbit data). a) BAT light curve in the 1450 keV energy band and a binning of 10 s. The horizontal lines mark the time intervals used for the spectral extraction (peak1 and peak2). b) The XRT light curve in the 0.210 keV energy band. An adaptive binning has been used to achieve in each point a signal-to-noise ratio of S/N = 3.

Open with DEXTER
In the text
thumbnail Fig. 7

Light curve of the XRT dataset of IGR J174072808. Grey downward-pointing arrows correspond to the 3σ upper limits calculated for the source non-detections. Points after 107 s correspond to serendipitous observations.

Open with DEXTER
In the text
thumbnail Fig. 8

Spectroscopy of the 2011 October 15 outburst of IGR J174072808. Top panel: simultaneous XRT/PC data (filled red circles) and BAT data (empty blue circles) fit with a PHABS*CUTOFFPL model. Bottom panel: the ratio between the data and the best-fit model.

Open with DEXTER
In the text
thumbnail Fig. 9

XMM-Newton EPIC-pn light curve of IGR J174072808 extracted from the observation 0764191301, with a 200 s binning.

Open with DEXTER
In the text
thumbnail Fig. 10

XMM-Newton EPIC-pn (black) and MOS1 (red) spectra extracted from observation 0764191301. The best fit is obtained by using an absorbed power-law model. The residuals from the fits are shown in the bottom panel.

Open with DEXTER
In the text
thumbnail Fig. 11

Left: IBIS/ISGRI mosaic extracted with the OSA software from the SCW 51 in revolution 1215 (2080  keV energy band). We indicated (green squares) on the mosaic the position of 4 sources detected with a sufficiently high significance (SWIFT J174510.82624, GRS 1758258, GX 1+4, IGR J174643213) and that previously reported for IGR J181752419. The latter is indicated by using a yellow circle with a radius of 24 centred on the source best position provided by Grebenev (2013). Right: significance map obtained with the BATIMAGER software (Segreto et al. 2010a) in the 2080  keV energy band for the SCW 51. Green circles (24 radius) shows the positions of all sources significantly detected in the field, as well as the position of IGR J181752419, as reported by Grebenev (2013). In either mosaic, independently built with different software, we do not detect any significant emission from IGR J181752419. The bars at the bottom of each mosaic indicate the colour codes for the detection significances in units of standard deviations.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.