astro-ph0503665/ms.tex
1: %\documentstyle[12pt,aasms4,lscape]{article}
2: 
3: 
4: %*****************APJ STYLE**********************
5: %\documentclass{article}
6: %\usepackage{emulateapj}
7: %\usepackage{apjfonts}
8: %\usepackage{lscape}
9: %\setlength{\topmargin}{0.5in}
10: %\renewcommand{\topfraction}{1.0}
11: %\renewcommand{\bottomfraction}{1.0}
12: %\renewcommand{\textfraction}{0.00}
13: %\raggedbottom
14: %*****************APJ STYLE**********************
15: 
16: 
17: 
18: %*****************APJ STYLE**********************
19: \documentclass[12pt,preprint]{aastex}
20: %\usepackage{emulateapj}
21: %\usepackage{apjfonts}
22: \usepackage{lscape}
23: %\setlength{\topmargin}{0.5in}
24: %\renewcommand{\topfraction}{1.0}
25: %\renewcommand{\bottomfraction}{1.0}
26: %\renewcommand{\textfraction}{0.00}
27: %\raggedbottom
28: %*****************APJ STYLE**********************
29: 
30: 
31: %\documentstyle[12pt,aasms4,flushrt]{article}
32: %\documentstyle[11pt,aaspp4,flushrt]{article}
33: % \received{} \accepted{~~}
34: \newcommand{\ltsima}{$\; \buildrel < \over \sim \;$}
35: \newcommand{\simlt}{\lower.5ex\hbox{\ltsima}} % < over ~
36: \newcommand{\gtsima}{$\; \buildrel > \over \sim \;$}
37: \newcommand{\simgt}{\lower.5ex\hbox{\gtsima}} % > over ~
38: 
39: \newcommand{\pedix}[2]{$#1_{\,\mbox{\scriptsize #2}}$}
40: 
41: \newcommand{\feka}{\mbox{Fe I K$\alpha$}}
42: \newcommand{\fekb}{\mbox{Fe I K$\beta$}}
43: \newcommand{\feLa}{\mbox{Fe XXVI Ly$\alpha$}}
44: \newcommand{\nika}{\mbox{Ni K$\alpha$}}
45: \newcommand{\xmm}{{XMM-\emph{Newton} }}
46: \newcommand{\bfxmm}{{XMM-{\it \textbf{Newton}} }}
47: \newcommand{\asca}{{\emph{ASCA} }}
48: \newcommand{\lum}{erg~s$^{-1}$}
49: \newcommand{\flux}{{erg~cm$^{-2}$~s$^{-1}$ }}
50: \newcommand{\nh}{cm$^{-2}$}
51: \newcommand{\nhsym}{N_{\mbox{\scriptsize H}}}
52: \renewcommand{\arcsec}{\mbox{$^{\prime\prime}$} }
53: \newcommand{\kev}{\,\mbox{\scriptsize keV}}
54: \newcommand{\mum}{\:\mu\mbox{\scriptsize m}}
55: \newcommand{\norm}{photons~keV$^{-1}$~cm$^{-2}$~s$^{-1}$}
56: 
57: \newcommand{\sorg}{AXJ0447-0627}
58: \newcommand{\sorgs}{AXJ0447-0627 }
59: \newcommand{\chandra}{{\emph{Chandra} }}
60: 
61: 
62: \newcommand{\cl}{\centerline}
63: 
64: 
65: \received{}
66: \begin{document}
67: 
68: 
69: \title{The \bfxmm view of the relativistic spectral features in \sorg}
70: 
71: 
72: \author{R. Della Ceca$^1$, L. Ballo$^2$, V. Braito$^1$, and T. Maccacaro$^1$}
73: 
74: \affil{$^1$ INAF - Osservatorio Astronomico di Brera, via Brera 28, 20121 Milan, Italy 
75: (rdc@brera.mi.astro.it, braito@brera.mi.astro.it, tommaso@brera.mi.astro.it)}
76: \affil{$^2$ SISSA/ISAS, International School for Advanced Studies, 
77:             via Beirut 4, 34014 Trieste, Italy (ballo@sissa.it)}
78: 
79: 
80: \begin{abstract}
81: 
82: The \xmm observation of the optically Type 1 AGN \sorgs ($z=0.214$)
83: unambiguously reveals a complex, bright and prominent set of lines  in the 
84: $4-8\,\:$keV rest frame energy range. 
85: Although, from a phenomenological point of view, the observed properties can be 
86: %described by a simple power law model plus 5 narrow gaussian lines 
87: %(at rest frame energies of 4.49$^{+0.13}_{-0.17}$, 5.55$^{+0.06}_{-0.05}$,
88: %6.39$^{+0.07}_{-0.06}$, 7.02$^{+0.29}_{-0.12}$ and 7.85$^{+0.73}_{-0.76}$),  
89: described by a simple power law model plus 5 narrow Gaussian lines 
90: (at rest frame energies of $\sim 4.49$, $\sim 5.55$, $\sim 6.39$, $\sim 7.02$ 
91: and $\sim 7.85\:\,$keV),  
92: we find that a model comprising a power law ($\Gamma \sim 2.2$),  a
93: reflected relativistic continuum,  a narrow {\feka} line from neutral material
94: as well as a broad Fe  K$\alpha$ 
95: relativistic line from a ionized accretion disk  
96: represents a good physical description of the data.   
97: The ``double horned'' profile  of the relativistic line implies  an inclination
98: of the accretion disk of $\sim 45^{\circ}$, and an origin in a narrow 
99: region of the disk, from \pedix{R}{in}$\sim 19\:\,GM/c^{2}$ to 
100: \pedix{R}{out}$\sim 30\:\,GM/c^{2}$. The  narrow {\feka} line from neutral material 
101: is probably produced far from the central black hole, most likely in the 
102: putative molecular torus. Although some of these
103: properties have been already found in other 
104: Type 1 AGN and discussed in the literature, at odd with the objects
105: reported so far we measure high  equivalent widths (EWs) of the observed lines: 
106: $\sim 1.4\,\:$keV for the  ``double horned'' relativistic line and $\sim
107: 0.4\,\:$keV  for the narrow  line.
108: 
109: \end{abstract}
110: 
111: \keywords{galaxies: active -- galaxies: individual (\sorg) -- galaxies: Seyfert -- 
112: X-rays: galaxies}
113: 
114: 
115: \section{Introduction}
116: 
117: Recent \xmm  and \chandra observations of a number of Active Galactic Nuclei have 
118: discovered  complex spectral features (broad and/or narrow) red-ward
119: of the well known  fluorescent Fe emission line(s) at $6.4 - 7\:\,$keV
120: (\citealt{turner02}; \citealt{guainazzi03}; \citealt{yaqoob03}; \citealt{turner3516}; 
121: %\citealt{comastri04};  
122: \citealt{porquet04}; \citealt{bianchi04};
123: \citealt{mckernan04}; \citealt{gallo04}).
124: These features are probably  produced  in (or very close to) the supposed
125: accretion disk and, as such,  their study  provides primary information 
126: about the dynamics and  physical processes which are taking  place in the 
127: innermost part of the AGN (see \citealt{fabian00} and \citealt{reynolds03} 
128: for a review). 
129: Among the models invoked to explain these features there are:
130: {\it i)} localized hot spot on the accretion disk surface due to 
131: illumination by local flares (e.g. \citealt{dovciak04});
132: {\it ii)} reprocessed emission from narrow annuli on the surface of the 
133: accretion disk \citep{gallo04};
134: {\it iii)} inflow or outflow of material (e.g. the ejected blob model 
135: proposed by \citealt{turner766} to explain the properties of the narrow lines 
136: observed in MKN~766); 
137: and 
138: {\it iv)} destruction (spallation) by  energetic protons on the accretion disk surface 
139: of Fe into lower Z elements (mainly Cr and Mn), implying an enhancement 
140: of the line emission expected from elements of lower abundance \citep{skibo97}.
141:   
142: In this paper we use \xmm data to discuss the observed line properties of \sorg, a
143: broad line AGN at $z=0.214$, discovered during the optical identification
144: process of the X-ray sources of the \asca ($2-10\:\,$keV) Hard Serendipitous
145: Survey (\citealt{cagnoni98}; \citealt{rdc99}). 
146: We assume $H_0=70$~km~s$^{-1}$~Mpc$^{-1}$, 
147: $\Omega_{\lambda} = 0.7$ and $\Omega_{M} = 0.3$. 
148: 
149: \section{Observations and data reduction}
150: 
151: \subsection{X-ray observations}
152: 
153: \sorgs was detected with a $S/N=5.40$ in the \asca field pointed at NGC~1667
154: (\asca Sequence~ID$=71032000$) at the nominal \asca position of
155: $\mbox{RA}$ = 04:47:48.6,  $\mbox{Dec}$ = $-$06:27:50.8 (J2000.0). The net \asca
156: GIS2 counts from the source are $54 \pm  10$ ($2-10\:\,$keV), corresponding to
157: a $f_{2-10\:\,\kev} \sim 3 \times 10^{-13}\:\,$\flux (assuming a 
158: power-law photon
159: index $\Gamma=1.7$).
160: The source attracted our attention because of its position in the hardness
161: ratio diagram (cf. \citealt{rdc99}), indicative of a
162: hard,  presumably obscured, X-ray source ($HR1=-0.40\pm 0.15$; $HR2=0.32\pm
163: 0.17$).
164: 
165: We thus observed \sorgs with \xmm on Sept~8,~2002 for a total of about
166: $29\,$ks. The three EPIC cameras (MOS1, MOS2 and pn) 
167: were operating in full frame mode and with the thin
168: filter applied. The \xmm data have been cleaned and processed using the 
169: Science Analysis Software (SAS version~5.4), and analyzed using standard
170: software packages (FTOOLS version~4.2, XSPEC version~11.2). 
171: %The latest calibration files released by the EPIC team have been used.
172: Event files have been thus filtered for high-background time intervals and only
173: events corresponding to patterns $0 - 12$  (MOS 1\&2) and $0 - 4$ (pn) have been
174: used (see \citealt{xmmhb}); the net exposure times at
175: the source position after data cleaning are $\sim 21.4\:\,$ks (MOS1, MOS2)
176: and $\sim 17.5\:\,$ks (pn).
177: 
178: The \xmm MOS1, MOS2 and pn images in the $0.5-10\:\,$keV energy range reveal a
179: high signal-to-noise ratio ($S/N \sim 30$ and $\sim 50$ in the MOS and the pn,
180: respectively) point-like source within the \asca 90\% positional 
181: error circle ($\sim 2^{\prime}$
182: radius) of \sorg. This is the
183: only detected and visible X-ray source in the \asca error circle; the X-ray
184: position derived using the \xmm data is $\mbox{RA}$ = 04:47:48.62, $\mbox{Dec}$ =
185: $-$06:28:12 ($\sim 21\arcsec$ away from the nominal \asca position).
186: 
187: Source counts were extracted from a circular region of radius $22.5\arcsec$ for
188: the MOS and $17.5\arcsec$ for the pn (this smaller radius for the pn is due to
189: the proximity of a CCDs gap). Background counts were extracted from a nearby
190: source-free circular region of $\sim 42\arcsec - 50\arcsec$ radius. The net
191: count rates ($0.5 - 10\:\,$keV energy range) are $0.051\pm
192: 0.002\:\,$counts~s$^{-1}$, $0.054\pm 0.002\:\,$counts~s$^{-1}$ and $0.171\pm
193: 0.004\:\,$counts~s$^{-1}$ for MOS1, MOS2 and pn, respectively; 
194: the source counts
195: represents
196: about  85\% of the total counts in the source extraction region.  No
197: statistically significant source variability has been detected during the \xmm
198: observation.  To improve statistics, the MOS1 and MOS2 data have been combined
199: together, and the MOS and pn spectra have been fitted simultaneously, keeping
200: the relative normalization free. Source counts were binned so as to have at 
201: least $20$ counts in each energy bin.  We have also generated our own spectral
202: response matrices at the source position using the SAS tasks \emph{arfgen} and
203: \emph{rmfgen}. All the models discussed here have been filtered through the
204: Galactic absorption column density along the line of sight (\pedix{N}{H,
205: Gal}$=5.6\times10^{20}\:\,$\nh; \citealt{nh}).
206: %all the errors are at the 90\% confidence level for one interesting parameter 
207: %($\Delta \chi^2=2.71$). 
208: Unless  otherwise stated, fit parameters are quoted in the
209: rest-frame of \sorgs ($z=0.214$, see below), 
210: while the figures and the EWs are in the observer frame.
211: 
212: \subsection{Source identification and optical spectroscopy}
213: 
214: A  bright  ($\mbox{RA}$ = 04:47:48.5, $\mbox{Dec}$ = $-$06:28:13; APM red
215: magnitude = 17.7)  optical source lies about $2\arcsec$ from  the X-ray position
216: derived using the \xmm data. This object was observed spectroscopically at the
217: TNG  on  October 5th,~2002. The optical spectrum (not reported here) covers the
218: wavelength range $\sim 3500-8000\:\,$\AA ~(dispersion of $2.8\:\,$\AA/pixel)
219: and  clearly shows broad (FWHM $> 6000\:\,$km~s$^{-1}$) MgII, H$\beta$ and
220: H$\alpha$ lines  as well as narrow (FWHM $< 1000\:\,$km~s$^{-1}$)
221: [OIII]4959,5007~\AA $\ $ lines.  The optical line properties
222: and position allow us to classify \sorgs as a  classical broad line AGN  at
223: $z=0.214 \pm 0.001$. \sorgs also belongs to the \asca Medium Sensitivity Survey,
224:  so an
225: independent confirmation of the redshift and of the optical spectral
226: classification comes from the work presented in \citet{idOpt}.  
227: 
228: \section{X-ray spectral analysis}
229: 
230: A single absorbed power law (PL) model is not a good description of the overall
231: ($0.5 - 10\:\,$keV) spectrum of \sorgs ($\Gamma=2.18 \pm 0.05$; $\nhsym$
232: consistent with zero; $\chi^2/\mbox{d.o.f}=280.3/250$), since a very large 
233: discrepancy is present above $4\,\:$keV. 
234: 
235: In Fig.~\ref{fig:ratiopl} we show the ratio between the best fit PL model
236: (obtained considering only the line-free region from $\sim 0.8$ to $\sim
237: 3\:\,$keV) and the data in the  $2-10\,\:$keV energy  range.  The residuals
238: show the presence of a  possible  ``line-like'' feature at  $E \sim
239: 3.5\,\:$keV (observer frame), and of a complex structure (several different
240: lines?) in the  energy range between $\sim 4.5$ and $\sim 7\,\:$keV (observer
241: frame).  Splitting the total observation into two intervals of similar
242: exposure times, we do not find convincing evidence of temporal variability of
243: such structures.
244: 
245: We have been unable to reproduce the complex lines  structure with  a pure
246: reflected continuum, either normal (PEXRAV model in XSPEC; see \citealt{pex}) 
247: or relativistic (REFSCH model; see \citealt{pex}; \citealt{disk}).  A  pure
248: reflected continuum with associated emission lines from Ca, Cr, Fe and Ni (see
249: e.g. the modeling of the Seyfert~2 galaxy NGC~6552 by \citealt{reynolds94}) is
250: also unable to reproduce the observed structure since the relative abundances
251: of the  above elements are different from the expected ones. We also note that
252: a pure reflected continuum, more typical of optical Type~2 AGNs, is at
253: odd with the optical spectral classification of \sorg.
254: 
255: To investigate the presence of line-like features and to locate their  energy
256: centroid we slid a narrow ($\sigma=0.1\,\:$keV) Gaussian template  across the
257: data between  $4$ and $8\,\:$keV (rest frame), looking for an improvement  to
258: the fit with respect to the simple PL model.  The results are shown in the
259: inset of  Fig.~\ref{fig:ratiopl}, where we report the change in fit  statistics
260: ($\Delta \chi^2$) as a function of the centroid  energy position of the narrow
261: Gaussian line.     This analysis points out the presence of a number of
262: possible narrow lines, with rest frame 
263: energy centroids at about $4.5$, $5.6$, $6.4$, $7.0$
264: %and $\sim 7.9\,\:$keV (in the last case, the line centroid is not well
265: and $7.9\,\:$keV (in the last case, the line centroid is not well
266: constrained).    Note that in the resulting fit statistics a $\Delta \chi^2$
267: greater than $10.30$ ($5.25$) represents a feature that is significant at more 
268: than 99\% (90\%) confidence level, so the lines at energies $\sim 4.5$ keV and at 
269: $\sim 7.9$ keV
270: are of lower statistical significance with respect to those at 
271: $\sim 5.6\,\:$keV, $\sim 6.4\,\:$keV and $\sim 7.0\,\:$keV.  Following these indications, we
272: tried to reproduce the observed spectrum with a power law and five narrow
273: Gaussian lines.  The best fit spectral parameters are reported in
274: Table~\ref{tab:gauss} while the ratio between the data and this possible  best 
275: fit model is shown in  Fig.~\ref{fig:ratiocfr} (panel~a).  
276: 
277: The  line at $6.39^{+0.07}_{-0.06}\,\:$keV  (EW$\,\:\sim  700\,\:$eV) 
278: is positionally consistent with the {\feka} emission line, while 
279: the line at $7.02^{+0.29}_{-0.12}\,\:$keV (EW$\,\:\sim  600\,\:$eV)
280: is positionally consistent
281: both with the {\fekb} emission line (rest frame energy 
282: $E=7.058\,\:$keV)
283: or with the {\feLa} emission line (rest frame energy $E=6.96\,\:$keV).
284: However the association with the  {\fekb}  line is 
285: unlikely since the measured flux from this line 
286: should be at a fixed ratio ($\sim 0.11$) 
287: with the  {\feka} emission flux, clearly in disagreement 
288: with the measured EWs.
289: The association with {\feLa} seems to be  more plausible given that this line 
290: could be very prominent (and sometimes with an EW comparable with the 
291: narrow {\feka} line) in Type 1 AGN (see e.g. the case of the Seyfert 1 galaxy 
292: NGC 7314 discussed in \citealt{yaqoob03}).
293: %The  lines at $6.39^{+0.07}_{-0.06}\,\:$keV  and at 
294: %$7.02^{+0.29}_{-0.12}\,\:$keV are positionally consistent with the {\feka} and
295: %{\fekb} fluorescent emission lines, although their intensity (EW$\,\:\sim  700$
296: %and $\sim 600\,\:$eV, respectively) and their relative ratio is at odd  with
297: %those observed in local Seyfert~1 objects \citep{nandra97}. 
298: The line at 
299: $7.85^{+0.73}_{-0.76}\,\:$keV is positionally consistent with the \nika, while
300: for the remaining two lines ($E=4.49^{+0.13}_{-0.17}\:\,$keV and
301: $5.55^{+0.06}_{-0.05}\:\,$keV) there are no clear associations
302: with well known and expected elements. The strongest expected lines in the
303: spallation model are the Cr~K$\alpha$ at 5.4 keV and the Mn~K$\alpha$ at 5.9
304: keV \citep{skibo97}. Both these lines are ruled out by the mismatch with the
305: measured energy lines centroid. So, unless an energy shift occurs (but we do
306: not observe any energy shift for the {\feka} line), the spallation model is an
307: unlikely explanation of the \xmm data. 
308: %We have also evaluated the upper limits for Fe~XXV~(f) at E$\sim$ 6.64 keV
309: We have also evaluated the upper limits for Fe~XXV~(f) at $E \sim 6.64\:\,$keV
310: and for Fe~XXV~(r) at $E \sim 6.70\:\,$keV since the strenght of these lines, 
311: when combined with the strenght of other ionized 
312: Fe lines, can constrain emission 
313: models (cfr. \citealt{yaqoob04} and reference therein). 
314: These two lines are not required by the current data set and the 
315: 90\% upper limit on their EW is $\sim 400\:\,$eV.
316: 
317: %Using the best fit power-law model plus the 5 narrow gaussian 
318: %lines detailed  in Table  1 we find that the 90\% upper limit of these 
319: %last two lines is $\sim 400$ keV.
320: 
321: The complex structure detected in the spectrum of \sorgs could suggest
322: a profile of a Fe line produced by an accretion disk. 
323: We explored this interesting possibility using 
324: the DISKLINE model \citep{disk}, which assumes 
325: a non-rotating Schwarzschild black hole\footnote{We 
326: have also tried the LAOR model \citep{laor}, 
327: in which the black hole is maximally 
328: rotating, but because of the limited source
329: count  statistics we  could not discriminate between the
330: DISKLINE and LAOR model. 
331: Since the black hole spin parameter is clearly an over-parameterization 
332: of the present data set we report here only the results obtained applying the DISKLINE
333: model. We note however that similar results have been obtained using the 
334: LAOR model.}.
335: The relativistic effects have been introduced also in the
336: description of the reflected continuum, replacing the simple power law model 
337: with the XSPEC model REFSCH\footnote{
338: This model is the sum of an e-folded power law primary spectrum 
339: plus its reflected  component from a ionized relativistic accretion disk.}.
340: 
341: We started the analysis using a model composed of the REFSCH  model 
342: plus a {\feka} disk line fixing its energy position 
343: to $6.4\:\,$keV rest-frame; 
344: since in \sorgs the observed 
345: lines seems to have a 
346: significantly larger 
347: EW than usual we have also added the corresponding 
348: {\fekb} disk line emission fixing the {\fekb}/{\feka}
349: ratio to that expected form the theory ($\sim 0.11$).
350: The best fit spectral
351: parameters are reported in Table~\ref{tab:relat} while the ratio between the data
352: and the best fit model is shown in  Fig.~\ref{fig:ratiocfr} (panel~b).
353: Although the overall fit is statistically acceptable ($\chi_{\nu}^2 = 1.00$), 
354: the ratio in Fig.~\ref{fig:ratiocfr} (panel~b) shows a 
355: ``line-like'' residual at an 
356: observed  energy of 
357: $\sim 6\,\:$keV  ($\sim 7.3\,\:$keV rest frame) that we were unable to reduce. 
358: We tried to
359: consider also disk lines associated to
360: {\nika} at $\sim 7.5\,\:$keV or to {\feLa} at $\sim 6.96\,\:$keV
361: emission, and/or
362: allowed the abundance of {\fekb}, {\nika} and {\feLa}
363: to be a free parameter of the fit, but
364: we were  unable to take into account such ``line-like'' structure. 
365: 
366: Thus we tried the same  model first used by \citet{weaver97} 
367: to describe the spectral  properties of  MCG-5-23-16:  a narrow
368: {\feka} component (\pedix{E}{line} fixed at  $6.4\,\:$keV)
369: plus  a  broad relativistic line  component (DISKLINE model) along with the
370: underlying continuum (REFSCH model). The best fit spectral parameters   are
371: reported in Table~\ref{tab:relat}, the ratio between the data  and the best fit
372: model is shown in  Fig.~\ref{fig:ratiocfr} (panel~c), while the folded model 
373: and the model itself are reported in  Fig.~\ref{fig:foldedrel}. The observed
374: line position ($E=6.61\pm 0.11\:\,$keV) of the broad relativistic line  is  
375: inconsistent with the {\feka}
376: from neutral material but  strongly suggests that it is due to  
377: Fe K$\alpha$ 
378: emission from  ionized He-like material. Overall this modeling provide a good
379: description  of the broad band spectral properties of \sorg. 
380: We have also tried to add a relativistic emission line from 
381: {\feLa} to the best 
382: fit model reported in Table 2; such line can be accommodated 
383: within the present data set 
384: (with an EW of $\sim 15\:\,$eV) but it is not statistically required. 
385: The observed flux
386: and the intrinsic luminosity in the $0.5-10\,\;$keV energy range are $(6.5\pm
387: 0.4) \times10^{-13}\,\:$\flux and  $(8.9\pm 0.5) \times10^{43}\,\:$\lum, the
388: errors reflecting the uncertainties  on the best fit model. 
389: We note that the $2-10\,\:$keV flux measured with  \xmm ($\sim 3.6 \times
390: 10^{-13}\,\:$erg~cm$^{-2}$~s$^{-1}$) is in very good agreement with the $2-10\,\:$keV flux
391: measured with  \emph{ASCA}.
392: %, thus excluding  flux variability on years time scales.
393: 
394: \section{Discussion and Conclusions}
395: 
396: Using \xmm data we have revealed that the optically Type 1 AGN \sorgs at
397: $z=0.214$  is characterized  by a complex of bright and prominent set of lines
398: in the  $4.5 - 8.0\:\,$keV energy range (rest frame).
399: 
400: We have shown that these lines can be reasonably well  reproduced by a physical 
401: model comprising a power law, a reflected relativistic continuum, a narrow
402: {\feka} line from neutral material and a  Fe He-like 
403: K$\alpha$  relativistic line  from a
404: ionized accretion disk.   Although not well constrained, the  best fit ionization
405: parameter ($\xi$ in  Table~\ref{tab:relat}) is consistent with the  Fe
406: ionization state, as deduced from the best fit line position (cf.
407: \citealt{matt93}). 
408: %ionization state, as deduced from the best fit line position (cfr.
409: %\citealt{matt93}). 
410: A similar modeling of the Fe line properties (a narrow plus a broad  relativistic
411: component) has been found to describe, for example,  the spectral properties of
412: the Seyfert~1.9 MCG-5-23-16 \citep{weaver97}, of the  radio-quiet quasar
413: MKN~205  \citep{reeves01} and of the Seyfert~1 NGC~3516 \citep{turner02};  the
414: presence of the Fe relativistic line  from a high ionized accretion
415: disk has been unambiguously reported  in the case of MKN~205. As already discussed by
416: the above authors, the most likely origin of the  narrow $6.4\:\,$keV  component is from
417: neutral matter distant from the black hole, e.g.  the putative molecular torus.
418: 
419: The resulting best fit of the relativistic ``double horned''  profile implies
420: an inclination of the accretion disk of $\sim 45^{\circ}$, as well as that the
421: observed lines should be produced in a narrow region of the disk from 
422: \pedix{R}{in}$\sim 19\:\,GM/c^{2}$ to \pedix{R}{out}$\sim 30\:\,GM/c^{2}$.  As in the
423: case of e.g. ESO~198-G24 \citep{guainazzi03} and  NAB~0205+024
424: \citep{gallo04},  a few alternative possibilities can be conjectured to explain
425: why the inner radius is larger than the last stable orbit.   The first
426: possibility is that the disk is highly ionized in the inner part, so most of 
427: the Fe is completely stripped off and the production of the Fe lines is
428: suppressed. Second, the accretion disk in \sorgs could be  truncated at 
429: $10 - 15\ GM/c^{2}$ (see \citealt{muller04}). 
430: Finally the relativistic line can be
431: produced by a localized hot spot on the accretion disk surface 
432: (e.g. \citealt{dovciak04}). 
433: %following its illumination by flares (e.g. \citealt{dovciak04}). 
434: 
435: All the properties discussed above have been also observed in other AGNs  and
436: seem to be in agreement with the expectation from accretion disk theory.
437: However  there is an observed property that is very unusual in \sorgs  and
438: makes this object unique: the very  large  EW observed, which
439: is at least a factor 5  greater than that usually measured in other Type 1 AGNs 
440: (see e.g. \citealt{yaqoob04} and reference therein) or
441: expected from an accretion disk around a Schwarzschild black hole 
442: (\citealt{matt92}; \citealt{matt93}).
443: According to the modelling reported  in \cite{martocchia96} and 
444: \cite{miniutti04}, a high EW could be explained if the primary 
445: X-ray source (illuminating both the observer and the accretion disk)
446: is located very close to a central and maximally rotating Kerr black hole. 
447: However such a combination should also imply a very high value of the 
448: reflection parameter $R$ and a reflection dominated source, probably 
449: in disagreement  with the best fit found here.
450: 
451: A way to solve part of these problems is to assume that the emission features appear 
452: much stronger than normal because the continuum is strongly absorbed.
453: We have tested this possibility 
454:  by adding a partial covering absorption model in front to the 
455: underlying continuum. Unless the primary AGN emission
456: is heavly absorbed ($\nhsym \sim 10^{25}\:\,$cm$^{-2}$) and therefore with signatures
457: falling outside the \xmm bandpass (but this is clearly at odd 
458: with the optical spectral classification of \sorg), the best fit absorbing 
459: $\nhsym$ and covering fraction ($\sim 9\times 10^{21}\:\,$cm$^{-2}$ and $\sim 0.2$, 
460: respectively) imply that absorption effects can not take into account 
461: the strong emission features observed.
462: 
463: This said, and with the caveat that the model proposed here could be not
464: fully  appropriate 
465: (e.g. we have already pointed out that a phenomenological description of the 
466: data can be also obtained by a simple  power law model 
467: plus 5 narrow Gaussian lines at rest frame energies of 
468: %plus 5 narrow gaussian lines at rest frame energies of 
469: 4.49$^{+0.13}_{-0.17}\:\,$keV, 
470: 5.55$^{+0.06}_{-0.05}\:\,$keV,
471: 6.39$^{+0.07}_{-0.06}\:\,$keV,
472: 7.02$^{+0.29}_{-0.12}\:\,$keV and
473: 7.85$^{+0.73}_{-0.76}\:\,$keV), 
474: we would like to note that the emission lines  in \sorgs
475: have a total EW of $\sim 2\,\:$keV.  This is an observational
476: result and, as such, is model independent. These lines deserve further
477: attention and a deeper investigation since any model proposed to describe the
478: X-ray spectral  properties of \sorgs should be able to explain their large 
479: EW.
480: 
481: Finally we would like to note that \sorgs was selected as a target  for an \xmm
482: observation  because of its observed \asca hardness ratios, indicative of a
483: hard, presumably  obscured, X-ray source (cf. \citealt{rdc99}). The \xmm
484: observations reported here have  shown that its hard X-ray colors are due
485: to the strong line complex in the observed $4.5-8.0\,\:$keV energy range rather
486: than to absorption effects.  There have been many claims in recent years 
487: about a substantial fraction ($\sim 10$\%) of X-ray absorbed optically classified
488: Type~1 AGN (with strong implications for AGN unification models and
489: synthesis  of the cosmic X-ray background) based mainly on poor quality X-ray
490: data (e.g. hardness ratios; see \citealt{willott04}). The result presented here indicates that a number
491: of these sources, {\it thought} to be X-ray absorbed Type~1 AGNs on the basis
492: of their hardness ratios, could instead be X-ray un-absorbed AGNs with
493: substantial and complex X-ray line emission (see also 
494: \citealt{macca04}).
495: 
496: \acknowledgements
497: 
498: Based on observations obtained with XMM-\emph{Newton} (an ESA science mission
499: with instruments and contributions directly funded by ESA Member States and the
500: USA, NASA) and with the Telescopio Nazionale Galileo (operated in the island of
501: La Palma by the Centro Galileo Galilei of the INAF in the Spanish Observatorio
502: del Roque de los Muchacos of the Instituto de Astrofisica de Canarias). 
503: We would like to thank the anonymous referee for her/his useful comments that 
504: have contributed improving this paper. 
505: We
506: thank M. Cappi, M.  Dadina, A. Caccianiga, L. Maraschi, G. Matt and P. Severgnini 
507: for useful discussions. 
508: This work has received partial financial support from the Italian
509: Space Agency (ASI grant I/R/062/02) and from the MIUR
510: (Cofin-03-02-23).
511: %RDC thanks the FFSS 
512: 
513: 
514: \begin{thebibliography}{ }
515: 
516: \bibitem[Akiyama et al.(2003)]{idOpt} Akiyama, M., Ueda, Y., Ohta, K., Takahashi, T., \& Yamada, T.\ 2003, \apjs, 148, 275 
517: 
518: \bibitem[Bianchi et al.(2004)]{bianchi04} Bianchi, S., Matt, G., Balestra, I., Guainazzi, M., \& Perola, G.~C.\ 2004, \aap, 422, 65 
519: 
520: \bibitem[Cagnoni, Della Ceca, \& Maccacaro(1998)]{cagnoni98} Cagnoni, I., Della Ceca, R., \& Maccacaro, T.\ 1998, \apj, 493, 54 
521: 
522: %\bibitem[Comastri, Brusa, \& Civano(2004)]{comastri04} Comastri, A., Brusa, M., \& Civano, F.\ 2004, \mnras, 351, L9 
523: 
524: \bibitem[Della Ceca et al.(1999)]{rdc99} Della Ceca, R., Castelli, G., Braito, V., Cagnoni, I., \& Maccacaro, T.\ 1999, \apj, 524, 674 
525: 
526: \bibitem[Dickey \& Lockman(1990)]{nh} Dickey, J.~M.~\& Lockman, F.~J.\ 1990, \araa, 28, 215
527: 
528: \bibitem[Dov{\v c}iak et al.(2004)]{dovciak04} Dov{\v c}iak, M., Bianchi, S., Guainazzi, M., Karas, V., \& Matt, G.\ 2004, \mnras, 350, 745 
529: 
530: \bibitem[Ehle et al.(2001)]{xmmhb} Ehle, M.~et al.\ 2001,{\it \xmm Users' Handbook} \\
531: (http://xmm.vilspa.esa.es/external/xmm\_user\_support/documentation/\\uhb\_2.0/index.html)
532: 
533: \bibitem[Fabian, Rees, Stella, \& White(1989)]{disk} Fabian, A.~C., Rees, M.~J., Stella, L., \& White, N.~E.\ 1989, \mnras, 238, 729 
534: 
535: \bibitem[Fabian, Iwasawa, Reynolds, \& Young(2000)]{fabian00} Fabian, A.~C., Iwasawa, K., Reynolds, C.~S., \& Young, A.~J.\ 2000, \pasp, 112, 1145 
536: 
537: \bibitem[Gallo et al.(2004)]{gallo04} Gallo, L.~C., Boller, T., Brandt, W.~N., Fabian, A.~C., \& Vaughan, S.\ 2004, \mnras, 355, 330 
538: 
539: \bibitem[Guainazzi(2003)]{guainazzi03} Guainazzi, M.\ 2003, \aap, 401, 903 
540: 
541: %\bibitem[Jansen et al.(2001)]{xmm} Jansen, F., et al.\ 2001, \aap, 365, L1 
542: 
543: \bibitem[Laor(1991)]{laor} Laor, A.\ 1991, \apj, 376, 90 
544: 
545: \bibitem[Maccacaro et al.(2004)]{macca04} Maccacaro, T., Braito, V., Della Ceca, R., Severgnini, P., \& Caccianiga, A.\ 2004, \apjl, 
546: 617, L33 
547: 
548: \bibitem[Magdziarz \& Zdziarski(1995)]{pex} Magdziarz, P.~\& Zdziarski, A.~A.\ 1995, \mnras, 273, 837 
549: 
550: \bibitem[Martocchia \& Matt(1996)]{martocchia96} Martocchia, A., \& Matt, G.\ 1996, \mnras, 282, L53 
551: 
552: \bibitem[Matt et al.(1992)]{matt92} Matt, G., Perola, G.~C., 
553: Piro, L., \& Stella, L.\ 1992, \aap, 263, 453 
554: 
555: \bibitem[Matt, Fabian, \& Ross(1993)]{matt93} Matt, G., Fabian, A.~C., \& Ross, R.~R.\ 1993, \mnras, 262, 179 
556: 
557: \bibitem[McKernan \& Yaqoob(2004)]{mckernan04} McKernan, B.~\& Yaqoob, T.\ 2004, \apj, 608, 157 
558: 
559: \bibitem[Miniutti \& Fabian(2004)]{miniutti04} Miniutti, G., \& 
560: Fabian, A.~C.\ 2004, \mnras, 349, 1435 
561: 
562: %\bibitem[Molendi, Bianchi, \& Matt(2003)]{molendi03} Molendi, S., Bianchi, S., \& Matt, G.\ 2003, \mnras, 343, L1 
563: 
564: \bibitem[M{\" u}ller \& Camenzind(2004)]{muller04} M{\" u}ller, A.~\& Camenzind, M.\ 2004, \aap, 413, 861 
565: 
566: %\bibitem[Nandra et al.(1997)]{nandra97}  Nandra, K., George, I.~M., Mushotzky, R.~F., Turner, T.~J., \& Yaqoob, T.\ 1997, \apj, 477, 602 
567: 
568: \bibitem[Porquet et al.(2004)]{porquet04} Porquet, D., Reeves, 
569: J.~N., Uttley, P., \& Turner, T.~J.\ 2004, \aap, 427, 101 
570: 
571: \bibitem[Reeves et al.(2001)]{reeves01} Reeves, J.~N., Turner, 
572: M.~J.~L., Pounds, K.~A., O'Brien, P.~T., Boller, T., Ferrando, P., 
573: Kendziorra, E., \& Vercellone, S.\ 2001, \aap, 365, L134 
574: 
575: \bibitem[Reynolds et al.(1994)]{reynolds94} Reynolds, C.~S., 
576: Fabian, A.~C., Makishima, K., Fukazawa, Y., \& Tamura, T.\ 1994, \mnras, 
577: 268, L55 
578: 
579: \bibitem[Reynolds \& Nowak(2003)]{reynolds03} Reynolds, C.~S.~\& Nowak, M.~A.\ 2003, \physrep, 377, 389 
580: 
581: \bibitem[Skibo(1997)]{skibo97} Skibo, J.~G.\ 1997, \apj, 478, 522 
582: 
583: %\bibitem[Str{\" u}der et al.(2001)]{pn} Str{\" u}der, L., et al.\ 2001, \aap, 365, L18 
584: 
585: %\bibitem[Turner et al.(2001)]{mos} Turner, M.~J.~L., et al.\ 2001, \aap, 365, L27 
586: 
587: \bibitem[Turner et al.(2002)]{turner02} Turner, T.~J., et al.\ 2002, \apjl, 574, L123 
588: 
589: \bibitem[Turner, Kraemer, \& Reeves(2004)]{turner766} Turner, T.~J., Kraemer, S.~B., \& Reeves, J.~N.\ 2004, \apj, 603, 62 
590: 
591: \bibitem[Turner et al.(2004)]{turner3516} Turner, T.~J., Kraemer, S.~B., George, I.~M., Reeves, J.~N., \& Bottorff, M.~C.\ 2004, \apj, 618, 155. 
592: 
593: \bibitem[Weaver et al.(1997)]{weaver97} Weaver, K.~A., Yaqoob, 
594: T., Mushotzky, R.~F., Nousek, J., Hayashi, I., \& Koyama, K.\ 1997, \apj, 
595: 474, 675 
596: 
597: \bibitem[Willott et al.(2004)]{willott04} Willott, C.~J., et al.\ 
598: 2004, \apj, 610, 140 
599: 
600: \bibitem[Yaqoob et al.(2003)]{yaqoob03} Yaqoob, T., George, I.~M., Kallman, T.~R., Padmanabhan, U., Weaver, K.~A., \& Turner, T.~J.\ 2003, \apj, 596, 85 
601: 
602: \bibitem[Yaqoob \& Padmanabhan(2004)]{yaqoob04} Yaqoob, T., \& 
603: Padmanabhan, U.\ 2004, \apj, 604, 63 
604: 
605: 
606: 
607: 
608: \end{thebibliography}
609: 
610: 
611: \clearpage
612: 
613:  \begin{table*}
614:   \begin{center}
615: %  \caption{Results of the spectral analysis ($0.5 - 10.0\,\:$keV in the observed
616: %  frame) - Power law plus 5~narrow gaussian lines.}\label{tab:gauss}
617:   \caption{Results of the spectral analysis ($0.5 - 10.0\,\:$keV in the observed
618:   frame) - Power law plus 5~narrow Gaussian lines.}\label{tab:gauss}
619:   \vspace{1cm}
620:    \begin{tabular}{ccccccc}
621:     \hline
622:     \hline
623:     \multicolumn{2}{c}{Power Law} & & \multicolumn{3}{c}{Lines} & \\
624:     \cline{1-2} \cline{4-6}
625:      $\Gamma$ & Norm        & & \pedix{E}{r.f.} & Norm      & EW & $\chi^2/$d.o.f.\\
626:      & & & (keV) & & (eV) \\
627: %    \vspace{0.2cm}\\
628:      (1) & (2) & & (3) & (4) & (5) & (6) \\
629:     \hline
630:     \multicolumn{7}{c}{}\\
631:      2.24$^{+0.06}_{-0.05}$ & 2.38$^{+0.13}_{-0.06}$ & &
632:                            4.49$^{+0.13}_{-0.17}$ & 0.92$^{+0.71}_{u}$   &  93$^{+72}_{u}$   & 226.3/240\\
633:     \multicolumn{3}{c}{} & 5.55$^{+0.06}_{-0.05}$ & 2.68$^{+0.94}_{-1.01}$ & 436$^{+152}_{-164}$ & \\
634:     \multicolumn{3}{c}{} & 6.39$^{+0.07}_{-0.06}$ & 3.13$^{+0.82}_{-1.26}$ & 700$^{+182}_{-290}$ & \\
635:     \multicolumn{3}{c}{} & 7.02$^{+0.29}_{-0.12}$ & 2.18$^{+0.91}_{-1.08}$ & 602$^{+251}_{-299}$ & \\
636:     \multicolumn{3}{c}{} & 7.85$^{+0.73}_{-0.76}$ & 0.74$^{+0.94}_{u}$   & 263$^{+331}_{u}$  &
637:     \vspace{0.2cm}\\
638:     \hline
639:     \multicolumn{7}{c}{}\\
640:     \multicolumn{7}{l}{\footnotesize  Columns are as follow: }\\
641:     \multicolumn{7}{l}{\footnotesize  Column 1: power law photon index; Column 2: normalization of the power law in  units of }\\
642:     \multicolumn{7}{l}{\footnotesize $10^{-4}\,$\norm~@~$1\,$keV; Column 3: rest-frame energy centroid of the narrow}\\
643: %    \multicolumn{8}{l}{\footnotesize  Column 2: normalization of the power law in  units of $10^{-4}\,$\norm~@~$1\,$keV;}\\
644: %    \multicolumn{8}{l}{\footnotesize  Column 3: rest-frame energy centroid of the narrow Gaussian line;} \\
645:     \multicolumn{7}{l}{\footnotesize  Gaussian line; Column 4: normalization in units of $10^{-6}\,$photons~keV$^{-1}$~cm$^{-2}$~s$^{-1}$ in the line;} \\
646: %    \multicolumn{8}{l}{\footnotesize  Column 4: normalization of the line in units of $10^{-6}\,$photons~keV$^{-1}$~cm$^{-2}$~s$^{-1}$ in the line;} \\
647: %    \multicolumn{7}{l}{\footnotesize  Column 5: equivalent width of the line; Column 6: possible line identification;} \\
648:     \multicolumn{7}{l}{\footnotesize  Column 5: equivalent width of the line; Column 6:  $\chi^2$ and number of degrees of freedom.}
649:     \vspace{0.2cm} \\
650: %    \multicolumn{8}{l}{\footnotesize  Column 6: possible line identification;} \\
651: %    \multicolumn{7}{l}{\footnotesize  Column 7: $\chi^2$ and degree of freedom.} \\
652:     \multicolumn{7}{l}{\footnotesize  NOTE: Errors are quoted at the 90\% confidence level for 1 parameter of interest ($\Delta \chi ^2=2.71$); }\\
653: %    \multicolumn{7}{l}{\footnotesize  The observed flux and the intrinsic luminosity in the $0.5-10\,\;$keV energy range are}\\
654: %    \multicolumn{7}{l}{\footnotesize  $6.31\times10^{-13}\,\:$\flux and $8.18\times10^{43}\,\:$\lum. The \xmm measured $2-10\,\:$keV flux  is}\\
655: %    \multicolumn{7}{l}{\footnotesize  $\sim 3.4 \times 10^{-13}\,\:$\flux, in good agreement with the \asca $2-10\,\:$keV flux.}\\
656:     \multicolumn{7}{l}{\footnotesize  $u$: unconstrained parameter.}\\
657:   \end{tabular}
658:   \end{center}
659:  \end{table*}
660: 
661: \clearpage
662: 
663: 
664: \clearpage
665: 
666: \begin{landscape}
667:  \begin{table*}
668:  \scriptsize
669:   \begin{center}
670: %  \caption{Results of the spectral analysis ($0.5 - 10.0\,\:$keV in the observed frame) - REFSCH model +~Fe K$\alpha$ relativistic line +~Fe (relativistic or gaussian) line.}\label{tab:relat}
671:   \caption{Results of the spectral analysis ($0.5 - 10.0\,\:$keV in the observed frame) - REFSCH model +~Fe K$\alpha$ relativistic line +~Fe (relativistic or Gaussian) line.}\label{tab:relat}
672:   \vspace{1cm}
673:    \begin{tabular}{cccccccccccc}
674:     \hline
675:     \hline
676:     \multicolumn{7}{c}{REFSCH$^{(a)}$} & & \multicolumn{3}{c}{DISKLINE} & \\
677:     \multicolumn{7}{c}{} & & \multicolumn{3}{c}{DISKLINE or GAUSSIAN} & \\
678:     \cline{1-7} \cline{9-11}    
679:      $\Gamma$ & $i$ & \pedix{R}{in} & \pedix{R}{out} & $\xi$ & $R$ & Norm & & \pedix{E}{r.f.} & Norm & EW$^{(b)}$ & $\chi^2$\\
680:      & & & & & & & & (keV) & & (keV) & [d.o.f.]\\
681:      (1) & (2) & (3) & (4) & (5) & (6) & (7) & & (8) & (9) & (10) & (11) 
682:     \vspace{0.2cm}\\
683:     \hline
684:     \hline
685: %    \multicolumn{12}{c}{}\\
686: %feKa+feKb
687:     2.26$^{+0.04}_{-0.07}$ & 25.0$^{+4.4}_{-3.0}{(c)}$ & 16.2$^{u}_{-3.9}{(d)}$ & 25.2$^{+13.5}_{u}{(d)}$ & 1$f$  & $0.44-2.00$ & 1.55$\pm 0.05$         & & 6.4$f$  & 5.3$^{+1.6}_{-1.2}{(c)}$ & 1.3$^{+0.4}_{-0.3}$ & 235.94 \\
688:                            &                                &                               &                                &         \multicolumn{4}{r}{\it relativistic \fekb:}     & 7.06$f$ & 0.6$^{+0.2}_{-0.1}{(c)}$ & 0.18$^{+0.05}_{-0.04}$ & [246] 
689:     \vspace{0.2cm}\\
690:     \hline
691: %feKa+gaussiana
692:     2.21$^{+0.09}_{-0.15}$ & 45.1$^{+6.8}_{-7.4}$~${(c)}$ & 19.5$^{u}_{-6.0}$~${(d)}$ & 29.5$^{+102.1}_{u}$ & 576.4$^{+956.6}_{u}$~${(c)}$ & 1.00$^{+1.00}_{-0.95}$ &  1.53$^{+0.06}_{-0.09}$   & & 6.61$^{+0.11}_{-0.12}$~${(c)}$ & 6.2$^{+2.1}_{-2.5}$ & $1.4^{+0.7}_{-0.4}$ & 223.2 \\
693:                            &                                 &                             &                      &                                  \multicolumn{4}{r}{\it narrow \feka:}              & 6.4$f$                            & 1.7$^{+1.1}_{-1.0}$ & $0.4^{+0.2}_{-0.2}$ & [243]
694:     \vspace{0.2cm}\\
695:     \hline
696:     \multicolumn{12}{c}{}\\
697:     \multicolumn{12}{l}{\footnotesize  Columns are as follow: }\\
698:     \multicolumn{12}{l}{\footnotesize  Column 1: power law photon index; Column 2: inclination angle (degrees); Column 3: inner disk radius in units of $GM/c^{2}$;}\\
699: %    \multicolumn{13}{l}{\footnotesize  Column 2: inclination angle (degrees);}\\
700: %    \multicolumn{13}{l}{\footnotesize  Column 3: inner disk radii in units of $GM/c^{2}$;} \\
701: %    \multicolumn{12}{l}{\footnotesize  Column 4: outer disk radii in units of $GM/c^{2}$; Column 5: disk ionization parameter in units of erg~cm~s$^{-1}$; Column 6: reflection scaling factor;} \\
702:     \multicolumn{12}{l}{\footnotesize  Column 4: outer disk radius in units of $GM/c^{2}$; Column 5: disk ionization parameter in units of erg~cm~s$^{-1}$;} \\
703:     \multicolumn{12}{l}{\footnotesize  Column 6: reflection scaling factor; Column 7: photon flux~@~$1\:$keV of the cutoff broken power law only (no reflection) in the observed } \\
704: %    \multicolumn{13}{l}{\footnotesize  Column 5: disk ionization parameter in units of erg~cm~s$^{-1}$;} \\
705: %    \multicolumn{12}{l}{\footnotesize  Column 7: photon flux~@~$1\:$keV of the cutoff broken power law only (no reflection) in the observed frame in units of $10^{-4}$;} \\
706:     \multicolumn{12}{l}{\footnotesize  frame in units of $10^{-4}\,$\norm; Column 8: rest-frame energy centroid of the line; Column 9: normalization in units } \\
707:     \multicolumn{12}{l}{\footnotesize  of $10^{-6}\,$photons~cm$^{-2}$~s$^{-1}$ in the line; Column 10: equivalent width of the line; Column 11: $\chi^2$ and number of degrees of freedom.}
708: %    \multicolumn{12}{l}{\footnotesize  Column 8: rest-frame energy centroid of the line; Column 9: normalization in units of $10^{-6}\,$photons~cm$^{-2}$~s$^{-1}$ in the line;} \\
709: %    \multicolumn{13}{l}{\footnotesize  Column 9: in units of $10^{-6}\,$photons~cm$^{-2}$~s$^{-1}$ in the line;} \\
710: %    \multicolumn{12}{l}{\footnotesize  Column 10: equivalent width of the line; Column 11: $\chi^2$ and degree of freedom.} \\
711: %    \multicolumn{13}{l}{\footnotesize  Column 11: observed flux in the 0.5$-$10 keV energy range;} \\
712: %    \multicolumn{11}{l}{\footnotesize  Column 12: intrinsic luminosity in the 0.5$-$10 keV energy range; Column 13: $\chi^2$ and degree of freedom.} \\
713: %    \multicolumn{13}{l}{\footnotesize  Column 13: $\chi^2$ and degree of freedom.} \\
714:     \vspace{0.2cm}\\
715:     \multicolumn{12}{l}{\footnotesize NOTE: Errors are quoted at the 90\% confidence level for 1 parameter of interest ($\Delta \chi ^2=2.71$). During the fit,}\\ 
716:     \multicolumn{12}{l}{\footnotesize the disk parameters of the different components were tied together; $f$: fixed parameter; $u$: unconstrained parameter. }\\
717: %    The observed flux and the intrinsic luminosity}\\
718: %    \multicolumn{11}{l}{\footnotesize  in the $0.5-10\,\;$keV energy range are $6.21\times10^{-13}\,\:$\flux and $8.39\times10^{43}\,\:$\lum (only relativistic line),}\\
719: %    \multicolumn{11}{l}{\footnotesize  $6.21\times10^{-13}\,\:$\flux and $8.39\times10^{43}\,\:$\lum (only relativistic line),}\\
720: %    \multicolumn{11}{l}{\footnotesize  and $6.53\times10^{-13}\,\:$\flux and $8.86\times10^{43}\,\:$\lum (relativistic line$+$narrow Gaussian line).}\\
721:     \multicolumn{12}{l}{\footnotesize$^{(a)}$ The following parameters have been fixed during all the fits: cutoff energy \pedix{E}{c}$=100\,\:$keV; disk temperature $T=3\times10^{4}\,\:$K; } \\
722:     \multicolumn{12}{l}{\footnotesize power law index for reflection emissivity $\beta=-3$.} \\
723:     \multicolumn{12}{l}{\footnotesize$^{(b)}$ Equivalent widths are computed with respect to the REFSCH underlying continuum.} \\
724: %%    \multicolumn{11}{l}{\footnotesize$^{(c)}$ Errors were calculated fixing at the best fit values \pedix{R}{in} and \pedix{R}{out} ($i$); $\Gamma$ (relativistic line normalization); \pedix{R}{in}, \pedix{R}{out} and $\Gamma$ (relativistic line energy). } \\
725:     \multicolumn{12}{l}{\footnotesize$^{(c)}$ Errors were calculated fixing at the best fit values \pedix{R}{in} and \pedix{R}{out} (inclination of the accretion disk and relativistic line energy); } \\
726:     \multicolumn{12}{l}{\footnotesize \pedix{R}{in}, \pedix{R}{out} and $\Gamma$ (relativistic line normalization); $R$ and Gaussian line normalization (disk ionization parameter). } \\
727: %    \multicolumn{12}{l}{\footnotesize$^{(c)}$ Errors were calculated fixing at the best fit values \pedix{R}{in} and \pedix{R}{out} (inclination of the accretion disk and relativistic line energy); } \\
728: %    \multicolumn{12}{l}{\footnotesize \pedix{R}{in}, \pedix{R}{out} and $\Gamma$ (relativistic line normalization); $R$ and gaussian line normalization (disk ionization parameter). } \\
729: %    \multicolumn{12}{l}{\footnotesize$^{(c)}$ Errors were calculated fixing at the best fit values \pedix{R}{in} and \pedix{R}{out} (inclination of the accretion disk); $\Gamma$ (relativistic line normalization); } \\
730: %    \multicolumn{12}{l}{\footnotesize \pedix{R}{in}, \pedix{R}{out} and $\Gamma$ (relativistic line energy). } \\
731:     \multicolumn{12}{l}{\footnotesize$^{(d)}$ Errors have been evaluated performing a fit while stepping the value of \pedix{R}{in} (\pedix{R}{out}) through the range $6.0 - 25.0$ ($20.0 - 200.0$). } \\
732:    \end{tabular}
733:   \end{center}
734:  \end{table*}
735: \end{landscape}
736: 
737: \clearpage
738: 
739: \centerline{ \bf Figure Captions}
740: 
741: \figcaption[f1.eps]{Ratio between the $2 - 10\:\,$keV MOS (filled squares) and pn (open circles) data
742: and the best fit power law model (limited to the $\sim 0.8$ to $\sim 3\:\,$keV energy range:
743: $\Gamma=2.24^{+0.21}_{-0.08}$).  In the inset we report the change
744: in fit  statistic ($\Delta \chi^2$) as a function of the centroid  energy
745: position of a narrow Gaussian line model that was stepped across the data; 
746: the comparison model is the  underlying power law continuum.
747: \label{fig:ratiopl}}
748: 
749: \figcaption[f2.eps]{
750: {\it Panel~a}: ratio between the $2 - 10\:\,$keV MOS (filled squares) and pn (open circles) 
751: data and the best fit spectral model composed by a 
752: power law ($\Gamma=2.24^{+0.06}_{-0.05}$) plus five Gaussian lines as detailed 
753: in Table~\ref{tab:gauss}.
754: {\it Panel~b}: ratio between the data and the best fit spectral model composed by 
755: the REFSCH model  
756: (an e-folded power law primary spectrum 
757: plus its reflected  component from a ionized relativistic accretion disk)
758: plus relativistic {\feka} and {\fekb} lines as detailed in Table~\ref{tab:relat}. 
759: {\it Panel~c}: ratio between the data and the best fit spectral model composed by
760: the REFSCH model 
761: plus a narrow {\feka} line from neutral material and a broad
762: Fe relativistic line 
763: from a ionized accretion disk, as 
764: detailed in Table~\ref{tab:relat}.
765: \label{fig:ratiocfr}}
766: 
767: \figcaption[f3.eps]{
768: MOS (filled squares) and pn (open circles) folded spectra fitted with 
769: the REFSCH model plus a
770: narrow {\feka} line from neutral material and a  broad Fe relativistic line 
771: from a ionized accretion disk.
772: In the inset we show the best fit model as detailed in 
773: Table~\ref{tab:relat}.
774: \label{fig:foldedrel}}
775: 
776: %%%UCP%%%
777: \newpage
778: \plotone{f1.eps}
779: \newpage
780: \plotone{f2.eps}
781: \newpage
782: \plotone{f3.eps}
783: 
784: 
785: 
786: \end{document}
787: 
788: