EDP Sciences
Free Access
Issue
A&A
Volume 559, November 2013
Article Number A86
Number of page(s) 4
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/201220869
Published online 19 November 2013

© ESO, 2013

1. Introduction

Ultraluminous X-ray sources (ULX) are extragalactic X-ray point sources located off-nucleus from their host galaxies. Their bolometric luminosities exceed ~1039 erg s-1 (Fabbiano 1989, 2006; Roberts et al. 2004). These extreme luminosities, assuming isotropic emission at (or below) the Eddington limit, would imply accreting black holes (BH) with masses of ~103–104 solar masses. The existence of such intermediate-mass black holes (IMBH) is in dispute. ESO 243-49 HLX-1 (Farrell et al. 2009; Wiersema et al. 2010; Farrell et al. 2012) is perhaps the only ULX case for which a consensus for IMBH has been reached to date. None of the other IMBH candidates, however, is widely accepted as definitive (see for instance the discussion about the mass of NGC 5408 X-1, Casella et al. 2008; Middleton et al. 2011; Dheeraj & Strohmayer 2012).

One important aspect is therefore to exclude that ULX could be background AGN, by using optical spectra (see for example the case of NGC 4168-ULX1, Masetti et al. 2003, and XMMU J124825.9+083020, Foschini et al. 2002a,b).

thumbnail Fig. 1

Upper left panel: R-band acquisition image taken with FORS2 CCD1. The two brighter galaxies are IRAS 13197-1627 (MCG-3-34-64) on the left/east and MCG-3-34-63 on the top/north. The images are all oriented in the same way. Upper right panel: HST image obtained with the W606 filter of MCG-3-34-64. The position of XMMUJ132218.3-16424 is indicated by the Chandra error circle in red (the white circle was the XMM-Newton error circle) while the dashed green lines indicate the position of the slit during the FORS spectral observation. As stated in the text, the ULX candidate, once the astrometry is calibrated using USNO-B1 catalogs, falls within the Chandra error circle. The W606 magnitude of the target has been recorded to be mF606 = 24.5. Lower panels: XMMUJ132218.3-16424 field as seen by XMM-Newton (left panel) and Chandra (right panel). In the XMM-Newton field the background extraction region for the pn is also displayed.

Open with DEXTER

In this context we present the analysis of optical and X-ray data of the ULX XMMUJ132218.3-164247. The target was observed in X-rays for the first time during an XMM-Newton pointing of the type II Seyfert galaxy IRAS 13197-1627 (also known as MCG-03-34-64). XMMUJ132218.3-164247 was the second-strongest X-ray source in the field of view with an X-ray flux of F0.5 − 10 keV ~ 10-13 erg s-1 cm-2. The source was detected within the D25 ellipse of nearby galaxy MCG-03-34-63 (see upper left panel of Fig. 1). If it is located at the distance of this galaxy, the implied luminosity of XMMUJ132218.3-164247 is L0.5 − 10 keV ~ 1041 erg s-1. Considering also the variability observed in its X-ray light curve, the source could have been detected at up to L0.5 − 10 keV ~ 1.5 × 1041 erg s-1 (Miniutti et al. 2006). The intrinsic nature of the source was tentatively investigated by Miniutti et al. (2006) using the data available at the time. In particular, the extreme Fx/Fo ~ 90 ratio suggested that the source was most probably not a background AGN (log   (FX/FO) = log FX + 0.4 R + 5.61, where Fx is the 2–10 keV flux and R is the magnitude in the R band, see Miniutti et al. 2006). Here we report on the redshift measurement of the source and with this precisely identify it as a QSO.

2. Observation log and data reduction

MCG-03-34-63 was observed with Chandra ACIS for 7 ks on July 31, 2006. We ran the source detection tool wavedetect on the Chandra data and found that XMMUJ132218.3-164247 is at RA 13:22:18.343 and Dec –16:42:46.91 (J2000). However, we also checked the accuracy of the Chandra astrometry for this particular observation. The only other clearly detected source in the vicinity of XMMUJ132218.3-164247 is the relatively bright AGN IRAS 13197-1627, about 1.8′ away from XMMUJ132218.3-164247. We have cross-correlated the Chandra position for the AGN with the USNO-B1.0 optical catalogue to find that the Chandra RA and Dec are both displaced by ~0.3′′. We therefore corrected the output of the wavedetect tool and found that the accurate X-ray position of XMMUJ132218.3-164247 is RA 13:22:18.365 and Dec –16:42:47.20 (J2000). The cross-correlation with the AGN IRAS 13197-1627 allowed to reduce the nominal Chandra error-circle from 0.6′′ in radius to a much more stringent 0.3′′. The Chandra position falls well within the XMM-Newton error-circle as reported in Miniutti et al. (2006). As reported there, the field was pointed by Hubble Space Telescope (HST) in 1994 for 500 s. After correcting the HST astrometry by using five common sources with the USNO-B1.0 catalog, we found that the source that was indicated as the counterpart (Miniutti et al. 2006) has coordinates RA 13:22:18.341 and Dec –16:42:47.00, well within the Chandra error circle (see right panel of Fig. 1).

The VLT observed XMMUJ132218.3-164247 with the FORS2 instrument for a total of 6 h 36 m. We took fifteen spectra of the source on April 11, 2007 starting at around 03:00 UT (see Table 1).

Table 1

Observation log.

The observation was performed in service mode and had to last three hours, but the pointing was longer because of the variable seeing. We summed all the data to obtain the highest possible signal-to-noise ratio. Because of the faintness of the source, we used the lowest resolution grism (150I) without order-sorting filters, affording a spectral coverage in the useable range of ~4000−10 500 Å. The slit width was 1′′ for all the observations, and this set-up secured a resulting spectral resolution of R = λλ = 250 at ~5000 Å. The spectra were extracted within the IRAF1 environment. After correcting for flat-field and bias, the spectra were background subtracted and optimally extracted (Horne 1986). Wavelength and flux calibration of the spectra were obtained by using He-Ne-Ar and Hg-Cd lamps and by observing the spectrophotometric standard star LTT6248 (Hamuy et al. 1994). The resulting spectrum is shown in Fig. 2.

The XMM-Newton data were taken on January 24, 2008 for ~41 ks. The EPIC instruments on-board XMM-Newton were operated in full window mode with the medium filter applied. The raw data were cleaned using SAS2 version 12.0 and inspected for high background level periods induced by soft protons. The overall exposure was ~41 ks (see Table 1) for EPIC pn and 41 and 46 ks for EPIC MOS cameras 1 and 2. The X-ray spectrum of XMMUJ132218.3-164247 was extracted from a circular region centered on the source and with a radius of 22′′ and 20′′ for EPIC pn and MOS respectively. The pn/MOS background has been extracted from large rectangular/circular regions in the same chips of the target (see lower left image of Fig. 1 for the EPIC/pn case). The data were then cleaned for high background periods that were present at the end of the observation.

thumbnail Fig. 2

Average optical spectrum of XMMUJ132218.3-164247. The main spectral features are labeled. The absorption features at ~6900 and ~7600 Å are telluric.

Open with DEXTER

3. Optical classification

The average spectrum taken at VLT with the FORS2 instrument is shown in Fig. 2. It displays a number of emission lines that can be readily identified with typical optical emission from an AGN. Among these we can detect the [OIII] doublet (λ = 4959 and 5007 Å restframe). Moreover, we also identify the [OII] line at 3727 Å restframe and some hints of a broad components of Hβ and Hγ. Unambiguously identified as a broad emission line is the Mg II emission feature at 2798 Å restframe. After fitting with a single Gaussian, the width of this line is σ~95 Å, corresponding to a FWHM ~ 2700 km s-1 at the redshift of the source. All these emission lines are found at a wavelength that is twice of its restframe value, which indicates a redshift z ~ 1. To exclude all, we also analyzed the spectrum of the nucleus of MCG-03-34-63 (see upper left panel of Fig. 1), finding that no similar features are present and confirming that these emission lines are not caused by high-order contamination. Averaging the value obtained from the [OIII], [OII] and Mg II lines, we finally obtained a redshift of z = 1.006 ± 0.003. This shows that XMMUJ132218.3-164247 is in fact a background type I QSO.

Following McLure & Jarvis (2002), we use the width of the MgII line and the UV flux at 3000 Å (restframe) to infer the mass of the SMBH of the QSO. When we correct the observed flux at 6000 Å for the absorption caused by our own Galaxy, we obtain that the BH has M ~ 2   ×   107 M (we used H0 = 73 km s-1 Mpc-1, Ωm = 0.27 and ΩΛ = 0.73). It is worth noting here that this must be considered only as rough estimate since it does not account for any possible intervening additional absorption component that may alter the UV flux estimation. That there is such a component is suggested by the curvature of the spectrum below ~6000 Å and may explain the observed FO/FX. The quality of the data, moreover, prevents a detailed modeling of the broad MgII emission line, and the error associated to its width is on the order of ~50%. Furthermore, the Hα and Hβ absorption lines measured at the redshift of the foreground MCG-03-34-63 (see Fig. 1) clearly indicate that part of the measured UV flux is probably contaminated by the stellar component of the galaxy itself. Finally, it has been demonstrated that the MgII estimator may easily misestimates the mass of the central SMBH by a factor of ~3 (McGill et al. 2008).

Table 2

Simple power-law models for the X-ray spectrum of XMMUJ132218.3-164247: best-fit parameters assuming different positions of the absorbing column density.

We also tried to refine the spectral classification taking advantage of the detection of the Hβ emissione line. The broad component of this line has a complex structure (see Fig. 2), and given the data quality, only a crude estimate of its flux can be obtained. After fitting with a single Gaussian, we obtained a ratio between Hβ/[OIII] ~ 2–3. This result suggests that the source may be a type 1.2 object according to the classification of Winkler (1992).

4. New X-ray spectrum

The high-energy emission of the source displays some variability (see Fig. 3). Its steadiness is rejected at more than the 99% confidence level. The variability of the source was highlighted by Miniutti et al. (2006).

thumbnail Fig. 3

0.3–10 keV light curve of XMMUJ132218.3-164247 as observed by XMM-Newton on January 24, 2008.

Open with DEXTER

thumbnail Fig. 4

0.3–10 keV spectrum of XMMUJ132218.3-164247 as observed by XMM-Newton on January 24, 2008.

Open with DEXTER

The X-ray spectrum of the source is well modeled by an absorbed power-law (see Fig. 2 and Table 2). An emission FeKα line is not required by the data (Δχ2 ≤ 1). The low energy cut-off detected below ~1 keV (observer frame, Fig. 4), however, indicates that there is an absorption column in addition to the Galactic one (NH = 4.98   ×   1020 cm-2, Dickey & Lockman 1990).

The measured NH of the extra absorber depends on the redshift at which we assume it to be located (first and second row of Table 2, respectively). MCG-03-34-63 intercepts the line of sight, which may add an additional absorption component. We therefore tested the cases in which the absorber is in MCG-03-34-63 or in XMMUJ132218.3-164247. The quality of the X-ray data prevents us from testing the case in which both objects contribute to the absorption. The fit slightly improved when the absorber is located at the redshift of XMMUJ132218.3-164247 (see Table 2 and Fig. 5), but this improvement is not statistically significant.

Finally, the X-ray spectrum of the source is flat (Γ ~ 1.6) comparing with nearby AGNs (Γ ~ 1.8 − 2; Dadina 2008) but, when the errors are considered, it is consistent with what have been found for high-z quasars (Γ = 1.6 ± 0.4 Cappi et al. 1997; Vignali et al. 2003). This value, however, is not consistent with the one previously measured by Miniutti et al. (2006) at a time when the source was weaker and displayed a steeper spectrum (F0.5 − 7 keV ~ 8 × 10-14 erg s-1 cm-2, while in 2008 F0.5 − 7 keV ~ 1.2 × 10-13 erg s-1 cm-2 and Γ ~ 1.95 in 2006).

5. Discussion and conclusions

We presented the analysis of optical and X-ray observations of the ULX candidate XMMUJ132218.3-164247.

The optical spectrum taken with the FORS2 instrument on the VLT clearly demonstrates that the source is a background QSO at z ~ 1. The new X-ray data presented here confirm the variability of the X-ray emission previously reported in Miniutti et al. (2006). Here the source was observed to be in a brighter state and with a flatter spectrum. The source is undoubtely a type I object, even though it displays a certain amount of absorption in X-rays. We speculate that the [OIII]/Hβ ratio may indicate that the source is a type 1.2 object. If so, we can expect a certain amount of nuclear absorption to be detected in X-rays (Bassani et al. 1999; Panessa et al. 2006).

thumbnail Fig. 5

Γ vs. NH contours assuming the absorber to be at z = 0.0213 (dotted contours) and at z = 1 (solid lines). Note that both values are not consistent with, and exceed at more then 99% confidence level, the Galactic column (NH = 4.98   ×   1020 cm-2, Dickey & Lockman 1992).

Open with DEXTER

We took advantage of the new optical data to obtain a rough estimate of the SMBH mass, which is M ~ 2 × 107 M. As previously stated, this value is only a very rough estimate.

We argue that the absorption explains the FX/FO ratio measured using HST data by Miniutti et al. (2006). The observed magnitude in R corresponds to an intrinsic magnitude mR = 20.4

when we assume that the column measured in X-rays also affects the optical. From this, we found a Av ~ 4.2 and, consequently, we obtained that FX/FO ~ 6, which is not as extreme as previously reported in Miniutti et al. (2006). Again, the assumptions made above are rather simplistic because it has been demonstrated that the absorption measured in X-rays and in the optical and IR are often different (Maiolino et al. 2001).


Acknowledgments

We thank the referee, Matt Malkan, for his comments that improved the manuscript. M.D. warmly thanks Eliana Palazzi and Elisabetta Maiorano for their helpful comments and their help in reducing and analyzing the optical data. This project has been funded by the INAF PRIN 2011. G.P. acknowledge support via an EU Marie Curie Intra-European fellowship under contract no. FP-PEOPLE-2012-IEF-331095.

References

All Tables

Table 1

Observation log.

Table 2

Simple power-law models for the X-ray spectrum of XMMUJ132218.3-164247: best-fit parameters assuming different positions of the absorbing column density.

All Figures

thumbnail Fig. 1

Upper left panel: R-band acquisition image taken with FORS2 CCD1. The two brighter galaxies are IRAS 13197-1627 (MCG-3-34-64) on the left/east and MCG-3-34-63 on the top/north. The images are all oriented in the same way. Upper right panel: HST image obtained with the W606 filter of MCG-3-34-64. The position of XMMUJ132218.3-16424 is indicated by the Chandra error circle in red (the white circle was the XMM-Newton error circle) while the dashed green lines indicate the position of the slit during the FORS spectral observation. As stated in the text, the ULX candidate, once the astrometry is calibrated using USNO-B1 catalogs, falls within the Chandra error circle. The W606 magnitude of the target has been recorded to be mF606 = 24.5. Lower panels: XMMUJ132218.3-16424 field as seen by XMM-Newton (left panel) and Chandra (right panel). In the XMM-Newton field the background extraction region for the pn is also displayed.

Open with DEXTER
In the text
thumbnail Fig. 2

Average optical spectrum of XMMUJ132218.3-164247. The main spectral features are labeled. The absorption features at ~6900 and ~7600 Å are telluric.

Open with DEXTER
In the text
thumbnail Fig. 3

0.3–10 keV light curve of XMMUJ132218.3-164247 as observed by XMM-Newton on January 24, 2008.

Open with DEXTER
In the text
thumbnail Fig. 4

0.3–10 keV spectrum of XMMUJ132218.3-164247 as observed by XMM-Newton on January 24, 2008.

Open with DEXTER
In the text
thumbnail Fig. 5

Γ vs. NH contours assuming the absorber to be at z = 0.0213 (dotted contours) and at z = 1 (solid lines). Note that both values are not consistent with, and exceed at more then 99% confidence level, the Galactic column (NH = 4.98   ×   1020 cm-2, Dickey & Lockman 1992).

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.