Free Access
Issue
A&A
Volume 521, October 2010
Article Number A57
Number of page(s) 35
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/200913440
Published online 20 October 2010

Online Material

Appendix A: List of tables

Table A.1:   List of type 1 and type 2 sources with relative  XMM-Newton observations.

Table A.2:   Best-fit parameters of the baseline models and absorption lines for the type 1 and type 2 sources.

Table A.3:   Estimates of the mean Fe K absorption line parameters for the sample sources among the  XMM-Newton observations in which they have been detected.

Appendix B: Notes on single sources

In this section we discuss and compare our work with the results already published in the literature. We payed particular attention to the works reporting a spectral analysis of the Fe K band of each sample source, especially if performed with the  XMM-Newton EPIC pn instrument.

NGC 4151: Piro et al. (2005) reported the detection of an absorption feature around $E\simeq 8.5{-}9$ keV with a statistical significance of 99.96% in one out of five Beppo-SAX observations of this source. Due to the low energy resolution of that instrument, the feature could be fitted either with an absorption edge due to highly ionized iron at rest or with an absorption line due to Fe XXV/XXVI with blue-shifted velocity of $\sim $0.1-0.2c. The presence of a complex and possibly multi-phase ionized absorber in NGC 4151 has also been reported by other authors, such as Schurch et al. (2003) using  XMM-Newton data and Kraemer et al. (2005) using simultaneous HST and Chandra observations. In the contour plots of the third  XMM-Newton observation of the source (obs. 0112830201) we find evidence for a possible narrow absorption line at $E\simeq7.8$ keV (see Fig. C.1). This line might be identified with blue-shifted Fe XXVI Ly$\alpha $. However, we did not include this line in Table A.2 because the relative Monte Carlo confidence level was lower than 95%. Instead, we clearly detected a narrow absorption line at rest-frame energy $E\simeq7.74$ keV in the last observation (obs. 0402660201). We identified the line with Fe XXVI Ly$\alpha $ with blue-shifted velocity $\simeq $0.1c (see Table A.2). It is important to note that in the contour plots of both these observations we found also evidence for two possible narrow absorption lines at energies lower than 6.4 keV. The parameters of these lines are $E= \rm 4.26^{+0.03}_{-0.06}$ keV, $\sigma =100$ eV and $EW = -15\pm4$ eV, in the former, and $E= 4.84\pm0.03$ keV, $\sigma=10$ eV and $EW = -13\pm4$ eV, in the latter. In both cases the F-test detection confidence levels are $\ga$99.9%. If identified with blue-shifted absorption lines from Ca XIX He$\alpha $ ($E\simeq3.$9 keV) and Ca XX Ly$\alpha $ ( $E\simeq4.11$ keV), their velocity is consistent with that inferred from the Fe K lines of $\simeq $0.1c.

IC4329A: the detection of a blue-shifted narrow absorption feature at $E\simeq7.7$ keV ascribable to Fe XXVI Ly$\alpha $ in the  XMM-Newton spectrum of IC4329A has already been reported by Markowitz et al. (2006). We confirm their results. Moreover, from the contour plot in Appendix C (see Fig. C.1), a broad absorption trough can be observed at an energy greater than $\simeq $9 keV. We tried to model it with an ionized Fe K edge (zedge in XSPEC). However, we obtained a better fit with a model composed of a blend of further three unresolved Fe XXVI Lyman series lines (Ly$\beta $ at E= 8.250 keV, Ly$\gamma$ at E= 8.700 keV and Ly$\delta$ at E= 8.909 keV) with a blue-shift consistent with that of the Fe XXVI Ly$\alpha $. In particular, we performed a fit adding to the baseline model (see Table A.2) four additional narrow absorption lines with energies fixed to the expected values for the Fe XXVI Lyman series and left their common energy shift as a free parameter. This series of lines provide a very good simultaneous modeling of all the absorption features, with a global $\Delta\chi^2=30$ for five additional parameters. The probability of having these four absorption lines at these exact energies simply from random fluctuations is low, about 10-4. Interestingly enough, their common blue-shifted velocity is $+0.097\pm0.005$c, completely consistent with that estimated for the Fe XXVI Ly$\alpha $ line alone (see Table A.2). The resultant EWs of the four Fe XXVI Lymann series lines are: $EW = -11\pm6$ eV for the Ly$\alpha $, $EW = -11\pm8$ eV for the Ly$\beta $, $EW = -10\pm6$ eV for the Ly$\gamma$ and $EW = -15\pm7$ eV for the Ly$\delta$. Their ratios are close to unity, which suggests possible saturation effects. A physically self-consistent modeling of these lines with the photo-ionization code Xstar is presented in a companion paper.

NGC 3783: a strong absorption line at the rest-frame energy of $E\simeq6$.7 keV has been detected by Reeves et al. (2004) in the  XMM-Newton spectrum of this source. The authors pointed out that the line energy is consistent with rest-frame resonant absorption from a blend of different highly ionized iron ions, such as Fe XXIII ($E\simeq6$.62 keV), Fe XXIV ($E\simeq6$.66 keV) and Fe XXV ($E\simeq6$.70 keV) (see Sect. 4.1). We confirm their detection and the line parameters here derived are in agreement with theirs. Furthermore, we note that in the contour plot of one observation (0112210201) there is evidence for a narrow absorption line at E<6.4 keV. From a fit with an inverted Gaussian we derived $E=4.30\pm0.04$ keV, $\sigma=10$ eV and $EW=-6\pm3$ eV. The F-test confidence level of the line is $\sim $99%, but it slightly decreases if all the remaining emission features redwards the neutral Fe K$\alpha $ line are modeled with Gaussians, indicating a possible slight model dependency. The identification of this line is ambiguous. However, as already discussed in Sect. 4.5, the line could be possibly associated with blue-shifted Ca XX Ly$\alpha $ with $v\sim10^4$ km s-1 or red-shifted Fe XXV/XXVI 1s-2p with $v\sim 0.5{-}0$.6c.

MCG+8-11-11: a detailed analysis of the  XMM-Newton spectrum of this source has already been published by Matt et al. (2006). We confirm their overall results. We did not detect any narrow Fe K absorption line.

NGC 5548: a detailed analysis of the  XMM-Newton spectrum of this source has already been published by Pounds et al. (2003b). Our analysis confirms their overall results. We did not detect any narrow Fe K absorption line.

NGC 3516: the detection of narrow highly ionized absorption features in the Chandra HETG and  XMM-Newton EPIC pn spectra of this source have already been published by Turner et al. (2005) and Turner et al. (2008). In particular, the authors reported the presence of H/He-line resonant absorption lines from Mg, Si, S and Fe. Our detection of several Fe XXV/XXVI K-shell absorption lines at E>6.4 keV is in agreement with their results. It should be noted that in the energy-intensity contour plot of the first observation of the source (0107460701) there is evidence for a narrow absorption feature at E<6.4 keV (see Fig. C.2) with F-test confidence level $\sim $99%. The resultant line parameters when modeled with a simple inverted Gaussian are $E= 4.75\pm0.04$ keV, $\sigma=10$ eV and $EW = -19\pm6$ eV. As discussed in Sect. 4.5, some possible identifications for this feature are with blue-shifted Ca XX Ly$\alpha $ with $v\sim0.1$c or red-shifted Fe XXV/XXVI 1s-2p with $v\sim0.4{-}0$.5c.

NGC 4593: a detailed analysis of the  XMM-Newton spectrum of this source has already been published by Reynolds et al. (2004). Our analysis confirms their overall results. We did not detect any narrow Fe K absorption line.

Mrk 509: the detection of narrow blue-shifted absorption lines ascribable to the Fe XXVI Ly$\alpha $ resonant transition in the  XMM-Newton spectra of this source have already been reported by Dadina et al. (2005) and Cappi et al. (2009). Our results are in complete agreement with theirs.

MCG-6-30-15: the Fe K band of this source is known to be complex, with the overlapping of several spectral components, such as a broad relativistic emission line, narrow emission/absorption features and warm absorption. Therefore, even if the EPIC pn has a high effective area in this energy band, its moderate energy resolution can be not sufficient to unambiguously disentangle all the spectral features. For instance, Fabian et al. (2002) and Vaughan & Fabian (2004) found hints for a narrow absorption feature at $\sim $6.7 keV probably due to Fe XXV He$\alpha $ in the EPIC pn spectrum of this source. However, they stated that it is difficult to clearly discriminate between narrow absorption and emission features due to the spectral complexity. Instead, from a deep spectral analysis of this source performed with the higher energy resolution Chandra HETG spectrometer, Young et al. (2005) have been able to clearly detect two narrow, unresolved, absorption lines ascribable to Fe XXV He$\alpha $ and Fe XXVI Ly$\alpha $. The lines have blue-shifted velocities $\simeq $0.007c and $EW \simeq -20$ eV. In our analysis we have not been able to unambiguously detect any blue-shifted narrow Fe K absorption features. This demonstrates the need to perform such systematic studies with other observatories as well, in order to exploit the different capabilities offered by each instrument. However, it should be noted that from the energy-intensity contour plots in Appendix C there is evidence for narrow absorption lines at E<6.4 keV with F-test significance $\sim $99% in three observations of the source (0029740101, 0029740701 and 0029740801). If modeled with inverted Gaussians, their parameters are: $E= 4.20\pm0.05$ keV, $\sigma =100$ eV and $EW = -12\pm7$ eV, for the first, $E= 4.49\pm0.05$ keV, $\sigma =100$ eV and $EW = -14\pm5$ eV, for the second, and $E= 4.27\pm0.05$ keV, $\sigma=10$ eV and $EW = -10\pm5$ eV, for the third. As discussed in Sect. 4.5, these lines could be possibly identified with blue-shifted Ca XX Ly$\alpha $ with velocities $\sim $0.02-0.08c or Fe XXV/XXVI 1s-2p with very large red-shifted velocities of $\sim $0.5-0.6c.

Ark 120: a detailed analysis of the  XMM-Newton spectrum of this source has already been published by Vaughan et al. (2004). The authors argued that from a broad-band X-ray analysis of the combined RGS and EPIC pn data there is no evidence for an intrinsic warm absorber and placed upper limits on the ionic column densities that are substantially lower than those of more typical, absorbed Seyfert 1s. This led them to claim that Ark 120 could actually represent a ``bare'' Seyfert 1 nucleus.

However, the reason for the lack of an X-ray warm absorber is quite unclear. It is plausible that a substantial column of ionized gas exists but it is either too highly ionized to show significant spectral features or lies out of the line of sight. These arguments are in agreement with our finding of a substantially blue-shifted Fe XXVI Ly$\alpha $ absorption line at the rest-frame energy of $\sim $9.18 keV.

Mrk 110: a deep analysis of the  XMM-Newton spectrum of this source have been reported by Boller et al. (2007). We did not detect any narrow blue-shifted Fe K absorption feature.

NGC 7469: a detailed analysis of the  XMM-Newton spectrum of this source has already been published by Blustin et al. (2003). Our analysis confirms their overall results. We did not detect any significant Fe K absorption feature.

IRAS 05078+1626: the analysis of the  XMM-Newton EPIC pn spectrum of the source has never been published. We found that a simple power-law continuum plus a neutral Fe K$\alpha $ emission line at $\sim $6.4 keV provides a good fit to the spectral data in the 4-10 keV band. We did not find evidence for narrow Fe K absorption features.

Mrk 279: we did not find any published work on the  XMM-Newton EPIC pn spectrum of the source. However, Fields et al. (2007) found evidence for the presence of different layers of highly ionized absorbing material from the spectral analysis of the soft X-ray Chandra LETG data. The associated column densities are low, of the order of $N_{\rm H} \sim 10^{20}$ cm-2, and the outflow velocities are of few $\sim $1000 km s-1. The authors also stated that the existence of a more ionized outflow component with iron ions from Fe XXIV to Fe XXVI cannot be ruled out. This is in agreement with our detection of a narrow absorption line ascribable to Fe XXV He$\alpha $ in the EPIC pn spectrum of the source, consistent with a rest-frame energy of $\simeq $6.69 keV.

NGC 526A: the analysis of the  XMM-Newton EPIC pn spectrum of the source has never been published. We found that a simple power-law continuum plus a neutral Fe K$\alpha $ emission line at $\sim $6.4 keV provides a good fit to the spectral data in the 4-10 keV band. We did not find evidence for narrow Fe K absorption features.

NGC 3227: a deep analysis of the EPIC pn spectrum of the source have been reported by Gondoin et al. (2003) and Markowitz et al. (2009). The 4-10 keV spectrum of the source is well modeled by a simple power-law continuum plus a neutral Fe K$\alpha $ emission line at $\sim $6.4 keV. However, in the first  XMM-Newton observation there is the need for a substantial neutral absorption component with column density $N_{\rm H} \sim10^{23}$ cm-2. This has been reported to be consistent with an eclipsing event by a broad line region cloud (Lamer et al. 2003). Our overall results are consistent with these conclusions and we did not find evidence for narrow Fe K absorption features.

NGC 7213: a deep analysis of the  XMM-Newton EPIC pn spectrum of the source in the Fe K band have been performed by Bianchi et al. (2003a). We refer the reader to that paper for detailed information. However, our overall results are consistent with theirs and we did not find evidence for highly ionized Fe absorption features.

ESO 511-G030: the analysis of the  XMM-Newton EPIC pn spectrum of the source has never been published. We found that a simple power-law continuum plus a neutral Fe K$\alpha $ emission line at $\sim $6.4 keV provides a good fit to the spectral data in the 4-10 keV band. We did not find evidence for narrow Fe K absorption features.

Mrk 79: (or UGC 3973) the spectral analysis of two snapshot  XMM-Newton observations of the source have been reported by Gallo et al. (2005). We did not consider those observations because of their too short exposures (<10 ks). Instead, here we report for the first time the spectral analysis of three new longer  XMM-Newton EPIC pn observation of the source. In all the cases the baseline model is constituted by a simple power-law continuum plus a neutral Fe K$\alpha $ emission line at $\sim $6.4 keV, with parameters consistent with those of Gallo et al. (2005). However, in one observation we have detected a narrow absorption feature ascribable to Fe XXVI Ly$\alpha $ at $E\simeq7.63$ keV, consistent with a blue-shifted velocity of $\sim $0.1c.

NGC 4051: a detailed analysis of the two  XMM-Newton observations of this source has been published by Pounds et al. (2004a). The overall spectral fit is consistent with ours. We confirm their detection of a narrow blue-shifted absorption line ascribable to Fe XXVI Ly$\alpha $ at the energy of $\sim $7.1 keV in the first observation. We detected a further absorption line at $\simeq $8.1 keV in the spectrum of the second observation. We interpreted this feature as a blue-shifted Fe XXVI Ly$\alpha $ absorption line.

Mrk 766: a detailed analysis of the  XMM-Newton EPIC pn spectra of the source has already been reported by Pounds et al. (2003c), Miller et al. (2007) and Turner et al. (2007). The spectral analysis of the Fe K band show the presence of both broad absorption troughs and narrow absorption line-like features at energies $\ga$7 keV. This complexity could indicate absorption from various layers of gas in different physical states. In fact, in two spectra we detected absorption features at energies greater than 7 keV that could only be modeled by narrow, unresolved, absorption lines. We interpreted such features as blue-shifted Fe XXVI Ly$\alpha $ absorption lines. The presence of the first one at a rest-frame energy of $\sim $7.3 keV has already been suggested by Miller et al. (2007) and Turner et al. (2007). The detection of the second one at $\sim $7.6 keV has instead never been reported. However, in another observation of the source (0109141301) a significant broad absorption trough between $E\simeq 8{-}9$ keV can be observed (see contour plot in Appendix B). As already reported by Pounds et al. (2003c), this feature is well modeled by a photoelectric edge, whose energy is consistent with the rest-frame Fe XXV K edge ( $E\simeq8.8$ keV). Therefore, we did not include this feature in Table A.2.

Mrk 841: a detailed analysis of the  XMM-Newton EPIC pn spectra of this source has already been published by Longinotti et al. (2004) and Petrucci et al. (2007). The authors tested different complex models in order to explain the broad-band X-ray spectral shape. However, our more simple spectral parametrization is in agreement with their results in the 4-10 keV band. Moreover, we detected a narrow absorption feature ascribable to a blue-shifted Fe XXVI Ly$\alpha $ resonance absorption line at the rest-frame energy of $\sim $7.2 keV.

Mrk 704: the  XMM-Newton EPIC pn spectral analysis of this source has never been published in the literature. We have found the 4-10 keV spectrum to be well modeled by an absorbed power-law continuum plus a narrow neutral Fe K$\alpha $ emission line. We did not find evidence for additional narrow Fe K absorption features.

Fairall 9: the  XMM-Newton EPIC pn spectral analysis of this source has already been published by Gondoin et al. (2001). We confirm their overall results. We did not detect any narrow Fe K absorption line.

ESO 323-G77: the detailed  XMM-Newton EPIC pn spectral analysis of this source has been published by Jiménez-Bailón et al. (2008). We confirm their detection of a couple of narrow absorption features at the rest-frame energy of $\sim $6.7 keV and $\sim $7 keV ascribable to Fe XXV He$\alpha $ and Fe XXVI Ly$\alpha $ resonant absorption.

1H419-577: the detailed broad-band spectral analysis of the  XMM-Newton EPIC pn observations of this source has been published by Pounds et al. (2004b) and Fabian et al. (2005). From their study, the authors argued that a good fit to the data can be provided either by an absorption dominated or by a reflection dominated model. However, these two models can not be distinguished in the narrow energy band we considered ( $E\simeq 4{-}10$ keV) and the continuum can be well approximated by a simple power-law. We detected a blue-shifted narrow absorption line ascribable to Fe XXVI Ly$\alpha $ at the rest-frame energy $E\simeq 7.23$ keV in the first  XMM-Newton observation of this source (obs. 0148000201), with F-test and Monte Carlo confidence levels of 99% and 95.8%, respectively. It should be noted that in the energy-intensity contour plot of this observation reported in Fig. C.5 (upper right panel) there is evidence for a possible further narrow absorption line at rest-frame energy $E\simeq8.4$ keV with F-test confidence level $\sim $99%. However, we did not include it in Table A.2 and we did not consider it further here because the associated $\Delta\chi^2\simeq9.3$ yielded a Monte Carlo confidence level $\simeq $93%, which is lower than the threshold value of 95%.

Mrk 335: the detection of an absorption feature at the rest-frame energy of $\sim $5.9 keV in the first  XMM-Newton observation of this source (obs. 0101040101) has been reported by Longinotti et al. (2007a). If identified with Fe XXVI Ly$\alpha $ resonant absorption, this would indicate a red-shifted velocity for the line of $\sim $0.18 c. We find evidence for such a feature in the energy-intensity contour plot in Fig. C.5. However, the measured detection confidence level is in the range $\sim $95-99% using the F-test. Moreover, also in the third observation (obs. 0510010701) we find evidence for a narrow absorption line at E<6.4 keV with F-test confidence level $\sim $99% (see Fig. C.5). If modeled as an inverted Gaussian, the rest-frame energy is $E= 4.58\pm0.06$ keV, $\sigma =100$ eV and $EW = -80\pm30$ eV. As already discussed in Sect. 4.5, there are different possible identifications for the line, such as: blue-shifted Ca XX Ly$\alpha $ with velocity $\sim $0.1c or red-shifted Fe XXV/XXVI 1s-2p with velocity $\sim $0.4-0.5c. However, from a further detailed spectral analysis, we find evidence for a possible additional absorption line at rest-frame energy $E\simeq3.9$ keV with F-test confidence level $\simeq $97%. Therefore, we could also speculate an association of both lines with rest-frame Ca XIX He$\alpha $ and Ca XIX He$\beta $, which are expected exactly at the energies $E\simeq3.9$ keV at $E\simeq4.58$ keV (Verner et al. 1996). Finally, we did not detect any significant narrow blue-shifted absorption lines at $E\ga7$ keV.

ESO 198-G024: the analysis of the  XMM-Newton EPIC pn spectrum of this source has already been reported by Guainazzi (2003) and Porquet et al. (2004). We agree with their overall results in the 4-10 keV band and we did not find evidence for highly ionized Fe K absorption lines. However, in the energy-intensity contour plot in Fig. C.5 there is evidence for a possible narrow absorption line at E<6.4 keV with F-test confidence level $\sim $99%. If modeled as an inverted Gaussian, the rest-frame energy is $E= 4.59\pm0.04$ keV, $\sigma=10$ eV and $EW = -14\pm6$ eV. As already discussed in Sect. 4.5, the line could be possibly associated with blue-shifted Ca XX Ly$\alpha $ with velocity $\sim $0.1c or red-shifted Fe XXV/XXVI 1s-2p with velocity $\sim $0.4-0.5c. However, as for the line in Mrk 335, the centroid energy of the feature is consistent with Ca XIX He$\beta $ at the rest-frame energy $E\simeq4.58$ keV. If associated with such a line, we would expect to observe also the Ca XIX He$\alpha $ at $E\simeq3.9$ keV. Instead, at this energy we detect only a narrow emission line with F-test confidence level $\simeq $98%. Therefore, even if appealing, this last identification is not straightforward.

Mrk 290: we did not find any published analysis of the  XMM-Newton EPIC pn spectrum of the source. We have detected a narrow absorption line ascribable to Fe XXVI Ly$\alpha $ in one out of four observations. The line rest-frame energy is $E\simeq8$ keV, which would suggest a blue-shifted velocity of $\sim $0.14c.

Mrk 205: the Fe K band analysis of the  XMM-Newton EPIC pn spectrum of the source has already been published by Reeves et al. (2001) and Page et al. (2003). We found that the spectrum in this band can be well modeled by a simple power-law continuum plus narrow Gaussian emission lines. However, we detected an additional narrow absorption feature at the rest-frame energy of $\sim $7.7 keV in one observation. If identified with Fe XXVI Ly$\alpha $ resonant absorption this would suggest an outflow velocity of $\sim $0.1c.

Mrk 590: the detailed analysis of the  XMM-Newton EPIC pn spectrum of the source has already been published by Longinotti et al. (2007b). We confirm their overall results. We did not detect any narrow Fe K absorption lines.

H 577-385: the detailed spectral analysis of the  XMM-Newton EPIC pn observations of the source has already been reported by Longinotti et al. (2009). We agree with their overall results. We did not find evidence for Fe K absorption features.

TON S180: we did not find any published analysis of the  XMM-Newton EPIC pn spectrum of the source. The 4-10 keV band is well modeled by a simple power-law continuum without additional spectral features.

PG 1211+143: a detailed analysis of the  XMM-Newton EPIC pn spectra of the source has already been reported by Pounds et al. (2003a), Pounds & Page (2006) and Pounds & Reeves (2009). In particular, Pounds et al. (2003a) clearly detected a narrow blue-shifted Fe K absorption line at the rest-frame energy of $\sim $7.6 keV in the pn spectrum of the first XMM observation. The identification of this line with Fe XXVI Ly$\alpha $ suggests a relativistic outflow velocity of $\sim $0.08c. Moreover, the presence of a highly ionized absorber in this source was confirmed also by the detection of several other blue-shifted narrow absorption lines due to lighter elements in the RGS as well. In a subsequent re-analysis of this observation using also the MOS data Pounds & Page (2006) have been able to confirm the presence of additional absorption lines, yielding a revised outflow velocity of $\sim $0.13-0.15c. Furthermore, the authors stated that they removed the ambiguity in the identification of the $\sim $7.6 keV absorption feature, with a preference for Fe XXV He$\alpha $. This would imply a consequently higher outflow velocity of $\sim $0.13c. We adopted this latter line identification.

Ark 564: we did not find any published analysis of the  XMM-Newton EPIC pn spectrum of the source. We found the 4-10 keV spectrum to be well modeled by a simple power-law continuum and Gaussian emission lines. We did not detect additional Fe K absorption features.

MCG-5-23-16: a detailed spectral analysis of the  XMM-Newton EPIC pn observations of this source has already been published by Dewangan et al. (2003) and Braito et al. (2007). We agree with their overall results in the 4-10 keV band. Moreover, we confirm the detection made by Braito et al. (2007) of a narrow absorption feature at the rest-frame energy of $\sim $7.8 keV in the last observation. If identified with Fe XXVI Ly$\alpha $ resonant absorption, the feature indicates a substantial blue-shifted velocity of $\sim $0.12c.

NGC 5506: the analysis of the  XMM-Newton spectra of the source have been reported by Matt et al. (2001) and Bianchi et al. (2003b). We agree with their overall results and we did not find evidence for narrow ionized Fe K absorption lines.

NGC 7172: the 4-10 keV spectra of the  XMM-Newton EPIC pn observations of the source have never been published. We found them to be well modeled by a simple absorbed power-law continuum plus a neutral Fe K$\alpha $ emission line. We did not detect additional Fe K absorption features.

NGC 7314: we did not find any reported analysis of the  XMM-Newton EPIC pn spectrum of the source. We found the 4-10 keV band to be well modeled by a simple absorbed power-law continuum plus a narrow neutral Fe K$\alpha $ emission line. We did not find evidence for narrow Fe K absorption lines.

NGC 2110: a detailed analysis of the broad-band X-ray spectrum of the source observed by  XMM-Newton has been reported by Evans et al. (2007). We agree with their results in the 4-10 keV band. We did not detect any Fe K absorption line.

NGC 4507: a detailed broad-band analysis of the  XMM-Newton EPIC pn spectrum of the source has been reported by Matt et al. (2004). We agree with their results in the 4-10 keV band. Furthermore, we have detected a narrow absorption feature at the rest-frame energy of $\sim $8.3 keV. If associated with Fe XXVI Ly$\alpha $ resonant absorption, the corresponding blue-shifted velocity is substantial, of the order of 0.18c.

NGC 7582: a broad-band spectral analysis of the  XMM-Newton observations of this source has been reported by Piconcelli et al. (2007). However, we report for the first time the detection a narrow absorption feature at the rest-frame energy of $E\simeq9$ keV in the first observation (obs. 0112310201). The description of this line has been included in Table A.2. Moreover, in the contour plot in Fig. C.7 there is evidence for a narrow absorption line at the rest-frame energy $E\sim4.5$ keV with F-test confidence contours $\ga$99%. If modeled with an inverted Gaussian, we derived $E= 4.52\pm0.03$ keV, $\sigma=10$ eV, $EW = \rm -76^{+18}_{-21}$ eV and F-test detection confidence level $\simeq $99.9% ( $\Delta\chi^2 \simeq13$ for two more model parameters). Furthermore, from a detailed spectral analysis we also found the presence of two further narrow ( $\sigma=10$ eV) absorption lines with F-test confidence levels $\simeq $95-98% at $E= 4.12\pm0.03$ keV, with $EW = \rm -63^{+18}_{-25}$ eV, and $E= 5.22\pm0.03$ keV, with $EW =\rm -41^{+19}_{-22}$ eV. If the absorption line at $E\simeq9$ keV is identified with the Fe XXV He$\alpha $ transition ($E\simeq6$.7 keV), these low energy lines turn out to be consistent with being absorption from Ar XVII He$\alpha $ ($E\simeq3$.14 keV), Ar XVIII Ly$\alpha $ ($E\simeq3$.3 keV) and Ca XIX He$\alpha $ ( $E\simeq3.9$ keV), blue-shifted by the same common velocity of $\simeq $0.255c. Moreover, we performed a test fitting the spectrum with the baseline model and adding four absorption lines with energies fixed to these expected values, letting the common blue-shift free to vary. We obtained a very good fit with a $\chi^2$ improvement of 35 for five additional parameters and the derived blue-shifted velocity is $+0.255\pm0.003$c, completely consistent with the value determined only from the Fe K line (see Table A.2). The probability for random fluctuations to give rise to this series of lines with the exact energy spacing and common blue-shift is very low, $\simeq $ $2\times 10^{-6}$.

Appendix C: Ratios and contour plots

\begin{figure}
\par\mbox{\includegraphics[width=3cm,height=4cm,angle=270]{13440f...
...cludegraphics[width=3cm,height=4cm,angle=270]{13440fc1x.ps} }
\par
\end{figure} Figure C.1:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 1 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

\begin{figure}
\par\mbox{\includegraphics[width=2.5cm,height=3.5cm,angle=270]{13...
...graphics[width=2.5cm,height=3.5cm,angle=270]{13440fc2zf.ps} }
\par
\end{figure} Figure C.2:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 1 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

\begin{figure}
\par\mbox{\includegraphics[width=2.5cm,height=3.5cm,angle=270]{13...
...graphics[width=2.5cm,height=3.5cm,angle=270]{13440fc3zf.ps} }
\par
\end{figure} Figure C.3:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 1 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

\begin{figure}
\par\mbox{\includegraphics[width=2.5cm,height=3.5cm,angle=270]{13...
...graphics[width=2.5cm,height=3.5cm,angle=270]{13440fc4zf.ps} }
\par
\end{figure} Figure C.4:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 1 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

\begin{figure}
\par\mbox{\includegraphics[width=2.5cm,height=3.5cm,angle=270]{13...
...graphics[width=2.5cm,height=3.5cm,angle=270]{13440fc5zf.ps} }
\par
\end{figure} Figure C.5:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 1 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

\begin{figure}
\par\mbox{\includegraphics[width=3cm,height=4cm,angle=270]{13440f...
...cludegraphics[width=3cm,height=4cm,angle=270]{13440fc6v.ps} }
\par
\end{figure} Figure C.6:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 1 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

\begin{figure}
\par\mbox{\includegraphics[width=2.5cm,height=3.5cm,angle=270]{13...
...graphics[width=2.5cm,height=3.5cm,angle=270]{13440fc7zb.ps} }
\par
\end{figure} Figure C.7:

Ratio against the continuum (upper panel) and contour plots with respect to the best-fit baseline model (lower panel; 68% (red), 90% (green), 99% (blue) levels) for the type 2 sources. The Fe K absorption lines are pointed by arrows. The vertical line refer to Fe I K$\alpha $.

Open with DEXTER

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.