astro-ph0703269/ms.tex
1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{aastex}
3: \usepackage{emulateapj5}
4: \topmargin 20pt
5: 
6: \shorttitle{A Chandra view of Naked AGN}
7: \shortauthors{Gliozzi et al.}
8: 
9:   \def\pks{PKS~0558--504}
10:   \def\oiii{[{\sc O\, iii}]}
11:   \def\oii{[{\sc O\, ii}]}
12:   \def\mgii{[{\sc Mg\, ii}]}
13:   \def\feka{Fe K$\alpha$}
14:   \def\chandra{{\it Chandra}} 
15:   \def\xmm{{\it XMM-Newton}} 
16:   \def\asca{{\it ASCA}} 
17:   \def\hst{{\it HST}} 
18:   \def\rxte{{\it RXTE}} 
19:   \def\sax{{\it BeppoSAX}} 
20:   \def\ginga{{\it Ginga}} 
21:   \def\rosat{{\it ROSAT}} 
22:   \def\ein{{\it Einstein}} 
23:   \def\integral{{\it Integral}} 
24:   \def\merlin{{\it MERLIN}} 
25:   \def\lum{erg s$^{-1}$}
26:   \def\flux{erg cm$^{-2}$ s$^{-1}$}
27:   \def\nh{cm$^{-2}$}
28:   \def\arcsec{$^{\prime\prime}$}
29: 
30:   \def\ltsima{$\; \buildrel < \over \sim \;$}
31:   \def\simlt{\lower.5ex\hbox{\ltsima}} % < over ~
32:   \def\gtsima{$\; \buildrel > \over \sim \;$}
33:   \def\simgt{\lower.5ex\hbox{\gtsima}} % > over ~
34: 
35: \begin{document}
36: \title{A Chandra view of Naked Active Galactic Nuclei}
37: 
38: \author{Mario Gliozzi}
39: \affil{George Mason University, 4400 University Drive, Fairfax, VA 22030}
40: 
41: \author{Rita M. Sambruna}
42: \affil{NASA's Goddard Space Flight Center, Code 661, Greenbelt, MD 20771}
43: 
44: \author{Luigi Foschini}
45: \affil{INAF/IASF-Bologna, via Gobetti 101, 40129 Bologna, Italy}
46: 
47: 
48: \begin{abstract}
49: We present the first X-ray observations of three sources belonging to a 
50: new AGN class: the naked AGNs. Based on optical spectroscopic studies, these 
51: sources appear as classical type 2 (obscured) AGNs, with only narrow emission 
52: lines.  However, long-term optical monitoring campaigns, carried out over more 
53: than two decades, show that the same sources are strongly variable, like 
54: type 1 (un-obscured) AGNs. Based on short Chandra observations, 
55: the sources appear to be fairly bright in the X-rays, with typical Seyfert 1s
56: values for the photon index ($\Gamma\sim$1.8) and without significant 
57: intrinsic 
58: absorption, supporting the conclusion that some bright AGNs may genuinely lack 
59: a broad line region. Future, broad-band studies as well as
60: deeper X-ray observations, probing both the 
61: spectral and the temporal properties of the naked AGNs, are crucial to shed 
62: light on the central engine of these sources, which may be representative of 
63: a large class of AGNs.
64: \end{abstract}
65: 
66: \keywords{Galaxies: active -- 
67:           Galaxies: nuclei -- 
68:           X-rays: galaxies 
69:           }
70: 
71: \section{Introduction}
72: More than two decades of multi-wavelength observations and theoretical studies
73: have led to a generally accepted unification scheme for AGNs. According to this
74: scheme, all AGNs share some basic ingredients: a supermassive black hole; 
75: an accretion disk coupled with a hot corona,
76: which radiates from the optical through X-ray energies;
77: high velocity and high density ($n_{\rm e} > 10^9~{\rm cm^{-3}}$) gas located 
78: at $\sim$ pc scales, usually referred to as the broad-line
79: region (BLR), since its radiation is dominated by broad (velocity dispersion 
80: of $\sim$ 2000--10000 km ${\rm s^{-1}}$) permitted lines; lower velocity and 
81: low density ($n_{\rm e} = 10^3-10^6~{\rm cm^{-3}}$) gas located at $\sim$ kpc 
82: scales, referred to as the narrow-line region (NLR), and emitting narrow (velocity
83: dispersion $<$ 1000  km ${\rm s^{-1}}$) forbidden and permitted lines; a 
84: gaseous and molecular torus on pc scales presumably obscuring the inner engine
85: and the BLR when viewed edge-on; and, in $\sim$ 10\% of the AGNs (the 
86: radio-loud AGNs), a relativistic jet extending up to Mpc scales, whose
87: axis is generally assumed to be parallel to the torus axis. Within
88: this scheme, several observational differences in the broad-band
89: and line spectral properties between type 1 and type 2 AGNs are explained by
90: orientation effects (Urry \& Padovani 1995; Antonucci 1993).
91: 
92: Although observations generally support orientation-based unification models
93: for AGNs, exceptions do exist. For example, only about 50\% of the brightest 
94: Seyfert 2s show the presence of  hidden broad line regions (HBLRs; Tran 2001) 
95: in their polarized optical spectra. Several studies indicate 
96: that the presence or absence of HBLRs depends on the AGN luminosity, with HBLR
97: sources having larger luminosities (e.g., Tran 2001; Gu \& Huang 2002). More 
98: specifically, Nicastro (2000) hypothesized that the absence of HBLRs 
99: corresponds to low values of accretion rate onto the central black hole. X-ray 
100: observations, which are less affected by absorption compared to optical 
101: studies,
102: indicate that, in several cases, the 
103: non-detection of a BLR cannot be ascribed to obscuration both in radio-quiet 
104: (e.g., Georgantopoulos \& Zezas 2003; Nicastro et al. 2003) and in radio-loud 
105: AGNs (e.g. Gliozzi et al. 2004). However, for column densities
106: $N_{\rm H} > 10^{24}{~\rm cm^{-2}}$, i.e. for Compton thick AGNs, also the 
107: X-rays
108: at energies $<$ 10 keV are completely absorbed. Therefore, in those cases
109:  neither \chandra\ 
110: nor \xmm\ can measure the intrinsic absorption or
111: probe the nuclear region directly (see Maiolino \& Risaliti
112: 2007 for a recent review).
113: 
114: 
115: Recently, Hawkins (2004) reported the discovery of a new class of AGNs,
116: apparently characterized by the absence of the broad line region 
117: accompanied by strong continuum emission and strong variability in the optical
118: band. This result is based on a large-scale monitoring program carried out in 
119: different optical bands over a period of 25 years, and on the spectroscopic 
120: follow-up (see Hawkins 2004 and references therein for a more detailed 
121: description of the optical observations). Of the 200,000 objects observed 1500
122: were selected as AGN candidates. The sample was further reduced to 55 
123: Seyfert-like candidates, after measuring the redshift of all emission line 
124: objects and selecting only sources with an adequate signal-to-noise ratio. 
125: 
126: An effective way to discriminate between the various classes of AGNs is based 
127: on the width of the Balmer lines and on the [O III]$\lambda$5007/H$\beta$ line
128: ratio (see Figure 5 of Hawkins 2004). Type 1 AGNs are characterized 
129: by H$\beta$ 
130: FWHM  $>2000{~\rm km~s^{-1}}$ and [O III]$\lambda$5007/H$\beta< 3$. On the 
131: other hand, type 2 AGNs typically show H$\beta$ FWHM  $<1000{~\rm km~s^{-1}}$ 
132: and large values of [O III]$\lambda$5007/H$\beta$. 
133: Interestingly, Hawkins finds that six objects with high 
134: [O III]$\lambda$5007/H$\beta$ line ratios (i.e., type 2 AGNs) show a strong 
135: variability typical of type 1 AGNs. These are the {\it naked AGNs}.
136: Note that, based on their optical properties, the naked AGNs cannot be 
137: classified as narrow line Seyfert 1 galaxies (NLS1), which are characterized 
138: by H$\beta$ FWHM  $<2000{~\rm km~s^{-1}}$,  
139: [O III]$\lambda$5007/H$\beta< 3$ and strong FeII (see Pogge 2000 for a review).
140: 
141: The existence of AGNs without BLR is apparently at odds with the standard
142: unification models. However, it must be outlined that, although the
143: unification models are able to explain most of the AGN observational
144: properties, they do not offer any insight into the physical origin of some
145: basic AGN ingredients, namely the BLR, the NLR, and the torus. Therefore, 
146: detailed investigations of AGNs without BLR have the potential of shedding
147: some light on the physical mechanisms that lead to the suppression 
148: (and possibly also to the formation) of the BLR.
149: 
150: Previous detections of AGNs without BLR were mostly based on the 
151: spectro-polarimetric observations, which require data with very high 
152: signal-to-noise (S/N) ratio
153: and rely on the existence of an appropriately placed scattering region
154: to view the obscured nucleus. Therefore, hidden BLR can be missed by this
155: method, leading to misclassified AGNs without BLR. On the other hand, the
156: detection of naked AGNs relies on the combined use of temporal and spectral 
157: properties. In principle, the former ensures that the nucleus is viewed 
158: directly (via the long-term variability typical of type 1 AGNs), 
159: whereas the latter confirms
160: the existence of a typical NLR excluding the presence of a BLR. 
161: 
162: However,
163: Hawkins (2004) does not explicitly address the important issue 
164: of the source of the observed optical variability. If the latter
165: is related to an unresolved jet component, we cannot a priori exclude that
166: the AGN nuclear region (including the BLR) is obscured and that
167: the naked AGNs are not ``pure'' type 2 objects. In addition,
168: recent studies have shown that the optical classification
169: of type 1 /type 2 objects is not trivially related to the classification
170: obtained via X-ray observations (e.g., Bauer et al. 2004; Cappi et al. 2006).
171: An investigation of the broad-band properties is clearly necessary to
172: confirm the genuine lack of BLR and hence shed some light on the nature 
173: of these enigmatic AGNs.
174: 
175: 
176: One of the most effective ways to investigate the properties of AGNs is with 
177: X-ray
178: observations, since: 1) The X-rays are produced and reprocessed in the inner, 
179: hottest nuclear regions of the source. 2) The penetrating power of (hard) 
180: X-rays allows 
181: them to carry information from the inner core regions without being 
182: substantially
183: affected by absorption, and to probe the presence of the putative torus. 3) 
184: Compared
185: to optical radiation, X-rays are far less affected by the host galaxy 
186: contamination.
187: Interestingly, the six naked AGNs, which have been extensively studied at 
188: optical wavelengths, have never been detected in the X-ray band, either 
189: because dedicated pointed X-ray observations have never been performed, or 
190: because X-ray sky surveys performed by previous satellites were not sensitive
191: enough to detect them. 
192: 
193: Here, we report the first X-ray study of three naked AGNs obtained with 
194: \chandra\ snapshot observations. The primary goal of this project is to 
195: verify whether the
196: sources are detected at X-ray energies and derive the basic information
197: (count rate and spectral slope) necessary to refine their X-ray investigation 
198: with deeper observations. In addition, we use the X-ray properties, 
199: complemented with information from the optical band, to investigate the
200: possible presence of a jet and to compare the naked AGNs with
201: other AGNs without BLR obtained with spectro-polarimetric studies.
202: 
203: 
204: The outline of the paper is as follows. In $\S~2$ we describe the observations
205: and  data reduction. The main findings are reported in $\S~3$. In $\S~4$ we
206: discuss the results, and  $\S~5$ we summarize the main conclusions.
207: Hereafter, we adopt $H_0=71{\rm~km~s^{-1}~Mpc^{-1}}$, $\Omega_\Lambda=0.73$ and
208: $\Omega_{\rm M}=0.27$ (Bennet et al. 2003)
209: 
210: \section{Observations and Data Reduction}
211: Three of the naked AGNs discovered by Hawkins,
212: Q2122-444, Q2130-431, Q2131-427, were observed with \chandra\ ACIS-S in 
213: December 2005. Since we had no X-ray 
214: information on these sources, for precaution the observations were carried 
215: out  using subarray modes (with frame time ranging between 0.4 and 0.8 s) to
216: minimize possible pile-up effects.
217: The general source properties from Hawkins (2004) are summarized in 
218: Table 1. The values for the broad-band spectral index $ \alpha_{\rm OX}$,
219: defined as $0.3838 
220: \log(l_{\rm 2keV}/l_{\rm 2500\AA})$, have been obtained from the best-fit
221: parameters of the $ \alpha_{\rm OX} - l_{\rm 2500\AA}$ relation
222: provided by Steffen et al. (2006). The B magnitudes  have been converted into 
223: fluxes at
224: 2500\AA\ assuming a typical optical slope of 0.7 ($F_\nu\propto \nu^{-0.7}$).
225:  
226: 
227: 
228: The data, provided by
229: Chandra X-Ray Center, were processed  using \verb+CIAO+ v. 3.2.2
230: and \verb+CALDB+ v. 3.2 following standard criteria. 
231: Only events with 
232: grades 0, 2--4, and 6 and in the energy range 0.3--8 keV were retained. 
233: We also checked that no flaring background events occurred during the 
234: observations. The extraction
235: of images and spectra, as well as the construction of response matrices,
236: were obtained following the standard procedure. 
237: Background spectra and light curves were extracted from source-free regions 
238: on the same chip of the source.
239: The observation log with
240: dates of the observations, net exposures, frame times, counts, and the 
241: average count rates are reported in Table 2. The counts were extracted from
242: circular regions with radius of 2\arcsec\ centered on the nominal optical
243: positions. The uncertainties
244: for the X-ray counts were
245: calculated following Gehrels (1986), as appropriate in a regime of low
246: counts.
247: 
248: The spectral analysis  was performed using the {\tt XSPEC v.12.3.0}
249: software package (Arnaud 1996; Dorman \& Arnaud 2001). 
250: Given the low number of counts in each
251: spectrum, due to the short exposure and the moderate count rate, we 
252: used the $C-$statistic (Cash 1979)
253: to fit the un-binned spectra over an energy range of 
254: 0.5--8 keV (in the observer's reference frame), where the calibration is best 
255: known (Marshall et al. 2005) and the background negligible.
256: The errors on spectral parameters are at 90\% confidence level for one 
257: interesting parameter ($\Delta C=2.71$).
258: 
259: \section{X-ray Results}
260: 
261: \subsection{Imaging}
262: Figure~\ref{figure:fig1} show the \chandra\ images of the three naked AGNs 
263: in the total energy range 0.3--8 keV. In order to fully exploit the 
264: unsurpassed spatial resolution of \chandra, for the two brightest sources
265: Q2122-444 and Q2130-431 (left and middle panels) the images were
266: extracted from ACIS-S event files after rebinning with a factor of 0.2 
267: (yielding pixel size of 0.1\arcsec) and smoothing with the subpackage 
268: \verb+fadapt+ of \verb+FTOOLS+, with a circular top hat filter of adaptive 
269: size in order to achieve a minimal number of 10 counts under the filter.
270: This procedure cannot be applied to Q2131-427 for the low number of counts.
271: As a consequence, for this source we simply extracted a
272: raw image, which is  shown in the right panel of Fig~\ref{figure:fig1}.
273: All the sources are clearly detected during the short \chandra\ exposures,
274: and no nearby objects are present within a radius of 1\arcmin\ from the central
275: AGN. 
276: 
277: \chandra, with its sub-arcsecond spatial resolution, is the best X-ray
278: satellite to investigate the presence of extended emission, which is of great
279: interest in AGN studies for the following reasons. 
280: Firstly, if the bulk of the X-rays 
281: arises from an extended region rather than from a compact nuclear source,
282: the X-ray emission is likely to be dominated by starburst activity  and not 
283: by an AGN (e.g., Dudik et al. 2005). Secondly, the presence of 
284: circum-nuclear diffuse X-ray emission
285: in low redshift AGNs, might be used to probe the AGN fuel reservoir and thus to
286: put constraints on the radiative efficiency (e.g., Gliozzi et al. 2003 and
287: references therein). Finally, and more importantly for the naked AGNs, an
288: asymmetric departure from the azimuthally symmetric point spread function (PSF)
289: model may reveal the presence of X-ray jets, which have been frequently 
290: detected in
291: longer \chandra\ observations (e.g.  Sambruna et al. 2002).
292: 
293: The visual inspection of Fig.~\ref{figure:fig1} suggests that
294: the three AGNs are point-like sources at the \chandra\ spatial resolution.
295: This conclusion is confirmed by the analysis of their radial profiles, which 
296: can be adequately fitted by a PSF plus background.
297: The PSF was created using the \chandra\ Ray Tracer (\verb+ChaRT+) simulator 
298: which takes into account the spectrum of the source and its location on the 
299: CCD. However, the short exposures hamper an accurate spatial analysis and 
300: thus the presence of a weak extended emission cannot be ruled out based on 
301: the current data.
302: 
303: 
304: 
305: \subsection{Spectral analysis} 
306: Two of the three AGNs observed with \chandra, Q2122-444 and Q2130-431, have 
307: enough counts to allow a direct fitting of the un-binned spectrum using the 
308: $C-$statistic. Both sources are reasonably well fitted with a simple 
309: power-law model absorbed by Galactic $N_{\rm H}$. The values of the photon 
310: indices  ($\Gamma=1.8\pm0.4$ for both Q2122-444 and  Q2130-431) are typical 
311: for Seyfert 1 galaxies (e.g., Nandra et al. 1997), although with substantial 
312: uncertainties due to the limited photon statistics.
313: The spectra and residuals of Q2122-444 and Q2130-431 are shown in 
314: Figure~\ref{figure:fig2}. If $N_{\rm H}$ is left free to vary in the spectral 
315: fits, the resulting spectral parameters are  poorly constrained, as outlined 
316: by  Figure~\ref{figure:fig3} that shows the confidence contours of $\Gamma$ 
317: and  $N_{\rm H}$.
318: On the one hand, these results indicate that snapshot \chandra\ observations 
319: are inadequate for a detailed spectral analysis. On the other hand, the low 
320: upper limits obtained for $N_{\rm H}$ suggest that the sources are not heavily
321: absorbed (provided that the bulk of the X-rays is indeed produced by the AGN 
322: and not by the jet).
323: 
324: The third source, Q2131-427, has less than 20 counts in the 0.5--8 keV energy 
325: band, which are not enough for a meaningful spectral fitting: The spectrum can
326: be fitted with a simple power law, but the photon index, $\Gamma=2.1\pm0.9$, 
327: is basically unconstrained. Nevertheless, the spectral properties of Q2131-427
328: can be probed by computing the hardness ratio $HR=(h-s)/(h+s)=-0.60\pm0.27$ 
329: (where $s$ refers to the 0.5--2 keV energy band and $h$ to 2--8 keV).
330: 
331: The photon indices,  obtained by fixing $N_{\rm H}$ to the Galactic value,
332: along with the hardness ratios, the observed fluxes, and the intrinsic 
333: luminosities are summarized in Table 3. Since the uncertainties are dominated 
334: by the errors on the photon index, the errors on fluxes and luminosities have 
335: been derived by varying $\Gamma$ by $\pm1\sigma$.
336: The similarity of the hardness ratio of Q2131-427 with the $HR$ values 
337: obtained for Q2122-444 and Q2130-431 (see Table 3) suggests that also the 
338: spectrum of Q2131-427  is consistent with a simple power law with a typical 
339: Seyfert-like photon index.
340: 
341: For completeness, in the last column of Table 3, we have also reported
342: the 2 keV monochromatic luminosities expected on the basis of the 
343: $\alpha_{\rm OX}$ values, which in turn  were obtained
344:  assuming the best-fit parameters of the correlation between
345: $\alpha_{\rm OX}$ and $l_{2500\AA}$  proposed by Steffen et al. (2006).
346: The measured X-ray luminosities are
347: roughly consistent with the predicted values, with the values for Q2122-444 
348: and Q2131-427 lower by a factor 1.5--2 compared to the expected values, and
349: with Q2130-431 higher by a factor of 1.5.
350: This discrepancy can be explained by the fact that the X-ray
351: and optical measurements are not simultaneous and these sources are
352: known to be highly variable. Alternatively, it
353: might indicate that not only the line properties but 
354: also the broad-band 
355: continuum of the naked AGNs is different from that of more typical AGNs.
356: 
357: 
358: 
359: \section{Discussion}
360: One of the key property of the naked AGNs is their pronounced and 
361: continuous variability observed in the optical band over a period of 25 years.
362: This might indicate that the central region of the naked AGNs is viewed 
363: directly and thus that lack of the BLR cannot be ascribed to low-quality 
364: observations or to obscuration effects. On the other hand, this variability
365: might be a signature for an unresolved jet component, casting doubts on the
366: conclusion that the naked AGNs are pure type 2 objects.
367: 
368: Although a jet dominance in the continuum properties of the naked AGNs
369: cannot be ruled out a priori, we regard this possibility as unlikely for
370: several reasons. First of all, the jet dominance is observed in blazars,
371: which represent a very small fraction of the AGN population ($<1\%$), 
372: while Hawkins found that more than 10\% of the monitored Seyfert-like objects
373: have significant variability. Secondly,
374: a direct comparison of the naked AGN light curves with those typically
375: observed in blazars reveals several differences.
376: Compared to jet dominated sources (e.g., Raiteri et al. 2006),
377: the naked AGNs seem to be characterized by smoother, lower-amplitude 
378: variability without short-term events. Furthermore, the strength of the 
379: optical lines typically observed in jet-dominated AGNs is weaker than
380: that observed in naked AGNs (cf, Sbarufatti et al. 2006). Finally,
381: the 2--10 keV luminosity measured for naked AGNs is systematically lower than
382: the  values obtained for blazars that show optical lines (Donato et al.
383: 2005). 
384: 
385: Although the above suggestive evidences seem to indicate that the
386: continuum emission in naked AGNs is not jet-dominated, only a broad-band
387: investigation including infrared and radio data will be able to 
388: definitively rule out the presence of a jet in naked AGNs. We plan to carry out
389: this kind of analysis in the near future, along with deeper X-ray 
390: observations. 
391: The latter will focus at investigating not only the time-averaged
392: spectra, but also the temporal and spectral variability properties, which 
393: might be crucial to discriminate between radio-loud and radio-quiet 
394: scenarios (see, e.g., Gliozzi et al 2006). For the rest of the discussion, 
395: we will assume that the bulk of the nuclear emission is associated with
396: the accretion process.
397: 
398: It is instructive to compare the properties of naked AGNs with those of
399: other AGNs lacking a BLR discovered with spectro-polarimetric observations.
400: For example, the median value of [O III]$\lambda$5007/H$\beta$ derived by Tran
401: (2001) for non-HBLR Seyfert 2 galaxies (the pure Seyfert 2s) -- $6.8\pm1.5$ --
402: is fairly consistent with the ratio measured for Q2122-444 and Q2131-427, but
403: much lower than the value obtained for Q2130-431 (see Table 1). The other 
404: properties discussed by Tran (2001) are related to the infrared band and thus 
405: cannot be used for a direct comparison with the naked AGNs, which only have 
406: optical and X-ray observations. 
407: 
408: On the other hand, Nicastro et al. (2003) discuss the X-ray properties
409: of sub-sample of non-HBLR Seyfert 2s obtained by Tran (2001). The values for 
410: the 2--10 keV luminosity from Nicastro et al. (2003) are significantly lower 
411: than the luminosities measured by \chandra\ for the 3 naked AGNs. However, 
412: according to Nicastro et al. (2003), the driving parameter is the accretion 
413: rate (in Eddington units), not the X-ray luminosity. Therefore,  in order to 
414: carry out a direct comparison with the findings of the above authors, we need 
415: to estimate the black hole masses of the naked AGNs. This can be done by 
416: exploiting the measurement of FWHM([O III]) (Green \& Ho 2005). However, it 
417: must be kept in mind that this is probably not a robust indicator (Netzer \& 
418: Trakhtenbrot 2007) and thus the values for the black hole 
419: mass should be simply considered as rough estimates.
420: 
421: 
422: Making use of $L_{\rm X}$ measured by \chandra\ and exploiting the information
423: provided by the ground-based optical observations, we can thus estimate the 
424: accretion rate of these AGNs in term of their Eddington ratio.
425: The results of these calculations are summarized in Table 4. Specifically, 
426: the values of the X-ray luminosity, unambiguously related to the AGN emission,
427: can be used to estimate the bolometric luminosity of the three naked AGNs, 
428: assuming a bolometric correction factor of 10 (e.g., Elvis et al. 1994). 
429: From the values of FWHM([O III])
430: we can derive an estimate for the stellar velocity dispersion (Greene \& Ho,
431: 2005) and thus the black hole mass, using the $M_{\rm BH} - \sigma_{\star}$
432: correlation (Tremaine et al. 2002). We are now able to estimate the Eddington
433: ratio $L_{\rm bol}/L_{\rm  Edd}$ for the three sources. 
434: 
435: For Q2122-444 and Q2130-431 we find Eddington ratios of the order of 
436: $1\times 10^{-2}$, which is about one order
437: of magnitude larger than the threshold proposed by Nicastro et al. (2003) to 
438: separate AGNs with HBLRs from  non-HBLR sources.
439: On the other hand, the third source, Q2131-427, is in agreement with Nicastro 
440: et al. (2003) predictions. The large Eddington ratios found for Q2122-444 and 
441: Q2130-431 are more consistent with the findings of Bian \& Gu (2007), who, 
442: using a large sample of 90 Seyfert 2
443: galaxies, found that above a given $L_{\rm bol}/L_{\rm  Edd}$ threshold both 
444: HBLR and non-HBLR 
445: sources are present, whereas below that threshold only non-HBLR objects are 
446: found.
447: 
448: In summary, the X-ray observations of naked AGNs seem to indicate that AGNs 
449: without BLR are associated not only with low-luminosity AGNs, as suggested 
450: by Nicastro (2000) and Laor (2003), but also with fairly powerful AGNs, as 
451: hypothesized by Bian \& Gu (2007). This is also supported by the findings of
452: Wolter et al. (2005), who detected with \xmm\ three QSO 2 candidates, 
453: which do not show any sign of obscuration. These sources might be somewhat 
454: related to the naked AGNs discussed here, although their [O III]/H$\beta$ 
455: line ratios are more typical of NLS1 galaxies and they are one order of 
456: magnitude brighter in the X-ray band. 
457: 
458: Our results are also in agreement with the findings of 
459: Steffen et al. (2003), who, using Chandra deep field data, showed that a 
460: sizable fraction of the local AGN population (z=0.1--1) is without broad 
461: lines and spans a large range of X-ray luminosities: 
462: $L_{\rm X}\sim 10^{42} - 10^{44}{~\rm erg~s^{-1}}$.
463: Importantly, they show that AGNs without BLRs (comprising both obscured
464: objects and pure type 2 AGNs) in fact represent the dominant class 
465: of AGNs for X-ray luminosities below $\sim 4\times 10^{43}{~\rm erg~s^{-1}}$.
466: 
467: \section{Conclusions}
468: We have used data from snapshot \chandra\ observations to investigate the 
469: basic X-ray properties of three naked AGNs, in preparation of deeper X-ray
470: observations. The main results can be summarized as follows:
471: 
472: \begin{itemize}
473: 
474: 
475: \item The three AGNs are easily detected in the 0.3--8 keV band, despite the
476: short exposures. 
477: 
478: 
479: \item The sources appear to be point-like objects at the \chandra\ ACIS-S
480: spatial resolution. However, the presence of weak extended emission cannot
481: be ruled out. No nearby sources are detected within a radius of 1\arcmin.
482: 
483: \item At first order, their X-ray spectral properties seem to be
484: consistent with those of Seyfert 1 galaxies, with typical photon 
485: indices of $\sim$1.8 and without strong intrinsic absorption.
486: 
487: \item The three sources are fairly bright in the
488: X-rays, with $L_{\rm 2-10 ~keV}$ of the order of few units in $10^{43}$ \lum, 
489: which are comparable with the X-ray luminosities observed in Seyfert 1 
490: galaxies (see Nandra et al. 1997) and larger than typical Seyfert 2 
491: luminosities (Turner et al. 1997).
492: \end{itemize}
493: 
494: 
495: The high detection rate ($>10\%$) of naked AGNs in the starting sample of 
496: Seyfert-like objects,
497: combined with the concrete possibility that they may be representative of
498: the large class of AGNs without BLRs, bolster the need 
499: to understand the nature of their central engine. A thorough investigation 
500: of the X-ray properties of the naked AGNs is crucial to reach this goal and 
501: to derive a more efficient way (compared to optical monitoring campaigns over 
502: periods of 25 years) to discover other members of this class, and shed new 
503: light on the AGN phenomenon.
504: 
505: \begin{acknowledgements} 
506: We thank the referee for his/her
507: comments and suggestions that improved the clarity of the paper.
508: MG acknowledges support by the Chandra Guest Investigator Program
509: under NASA grant 200997.  
510: \end{acknowledgements}
511: 
512: 
513: \begin{thebibliography}{}
514: \bibitem[Antonucci 1993]{anto93} Antonucci, R. 1993, ARA\&A, 31, 473
515: 
516: \bibitem[Arnaud 1996]{arn96} Arnaud, K. 1996, in ASP Conf. Ser. 101, 
517: Astronomical Data Analysis Software and Systems V, ed. G. Jacoby \& J. 
518: Barnes (San Francisco: ASP), 17
519: 
520: \bibitem[Bauer et al. 2004]{bau04} Bauer, F.E., et al. 2004, AJ, 128, 2048
521: 
522: \bibitem[Bennet et al. 2003]{ben03} Bennet, C.L. et al. 2003, ApJS, 148, 1
523: 
524: \bibitem[Bian \& Gu 2006]{bian06} Bian, W. \& Gu. Q, 2006, ApJ, 657, 159
525: 
526: \bibitem[Dorman \& Arnaud 2001]{dorm01} Dorman, B. \& Arnaud, K.A. 2001 in
527: Astronomical Data Analysis Software and Systems X, ASP Conference
528: Proceedings, Vol. 238. Edited by F. R. Harnden, Jr., Francis A. Primini,
529: and Harry E. Payne. San Francisco: Astronomical Society of the Pacific,
530: p.415
531: 
532: \bibitem[Cappi et al. 2006]{cappi06} Cappi, M., et al. 2006, A\&A, 446, 459
533: 
534: \bibitem[Cash 1979]{cash79} Cash, W. 1979, ApJ 228, 939
535: 
536: \bibitem[Donato et al. 2005]{don05} Donato, D., Sambruna, R.M, \& Gliozzi, M. 2005,
537: A\&A, 433, 1163
538: 
539: \bibitem[Dudik et al. 2005]{dudi05} Dudik, R.P., et al. 2005, ApJ, 620, 113
540: 
541: \bibitem[Elvis et al. 1994]{elv94} Elvis, M. et al. 1994, ApJS, 95, 1
542: 
543: \bibitem[Gehrels]{gehr86} Gehrels, N. 1986, ApJ, 303, 336
544: 
545: \bibitem[Georgantopoulos \& Zezas 2003]{geo03} Georgantopoulos, I. \& Zesas, 
546: A. 2003, ApJ, 594, 704
547: 
548: \bibitem[Gliozzi et al. 2003]{glio03} Gliozzi, M., Sambruna, R.M., Brandt, 
549: W.N. 2003, A\&A, 408, 949
550: 
551: \bibitem[Gliozzi et al. 2004]{glio04} Gliozzi, M., Sambruna, R.M., Brandt, 
552: W.N., Mushotzky, R., Eracleous, M. 2004, A\&A, 413, 139
553: 
554: \bibitem[Gliozzi et al. 2006]{glio06} Gliozzi, M., Papadakis, I.E. \& Brinkmann, W.P., 2007,
555: ApJ, 656, 691
556: 
557: 
558: \bibitem[Greene \& Ho 2005]{gree05} Greene, J.E. \& Ho, L.C. 2005, ApJ, 627, 721
559: 
560: \bibitem[Gu \& Huang 2002]{gu02} Gu, Q. \& Huang, J. 2002, ApJ, 579, 205
561: 
562: \bibitem[Hawkins 2004]{hawk04} Hawkins, M.R.S. 2004, A\&A, 424, 519
563: 
564: \bibitem[Laor 2003]{lao03} Laor, A. 2003, ApJ, 590, 86
565: 
566: \bibitem[Maiolino \& Risaliti 2007]{maio07} Maiolino R., \& Risaliti, G. 2007, astro-ph/0701109
567: 
568: \bibitem[Marshall et al. 2004]{marsh04} Marshall, H.L. 2005, ApJS, 156, 13
569: 
570: \bibitem[Nandra et al. 1997]{nand97} Nandra, K., George, I.M., Mushotzky, R.F., Turner, T.J., Yaqoob, T.
571: 1997, ApJ, 477, 602
572: 
573: \bibitem[Netzer \& Trakhtenbrot 2007]{netz07} Netzer, H. \& Trakhtenbrot, B. 2007,
574: ApJ, 654, 754
575: 
576: \bibitem[Nicastro 2000]{nic00} Nicastro, F. 2000, ApJ, 530, L65
577: 
578: \bibitem[Nicastro et al. 2003]{nic03} Nicastro, F., Martocchia, A.,
579: \& Matt, G. 2003, ApJ, 589, L13
580: 
581: \bibitem[Pogge 2000]{pogg00} Pogge, R.W. 2000, New Astronomy Review 44, 381
582: 
583: \bibitem[Raiteri et al. 2006]{rait06} Raiteri, C., et al. 2006, A\&A, 459, 731
584: 
585: \bibitem[Sambruna et al. 2002]{sambr02} Sambruna, R.M., et al. 2002, ApJ, 571, 206
586: 
587: \bibitem[Sbarufatti et al. 2006]{sbarr06} Sbarufatti, B., Falomo, R., Treves, A., 
588: Kotilainen, J. 2006, A\&A, 457, 35
589:  
590: \bibitem[Steffen et al. 2003]{stef03} Steffen, A.T., Barger, A.J., Cowie, L.
591: L., Mushotzky, R.F., Yang, Y. 2003, ApJ, 596, L23
592: 
593: \bibitem[Steffen et al. 2006]{stef06} Steffen, A.T., et al. 2006, AJ, 131, 2826
594: 
595: \bibitem[Tran 2001]{tran01} Tran, H.D. 2001, ApJ, 554, L19
596: 
597: \bibitem[Tremaine et al. 2002]{trem02} Tremaine, S., Gebhardt, K., Bender, R., et al. 2002, ApJ, 574, 740
598: 
599: \bibitem[Turner et al. 1997]{turn97} Turner, T.J., George, I.M., Nandra, K., \& Mushotzky, R.F., 1997,
600: ApJ, 113, 23
601: 
602: \bibitem[Urry \& Padovani 1995]{urr95} Urry, M. \& Padovani, P. 1995, PASP, 107
603:  
604: \bibitem[Wolter et al. 2005]{wolt05} Wolter A., Gioia, I.M., Henry, J.P., 
605: Mullis, C.R. 2005, A\&A 444, 165
606: \end{thebibliography}
607: 
608: %%--------------FIG1-------X-ray images-------------------------     
609: \begin{figure}     
610:  \includegraphics[bb=5 180 609 614,clip=,angle=0,width=6.cm]{f1a.eps}\includegraphics[bb=5 180 609 614,clip=,angle=0,width=6.cm]{f1b.eps}\includegraphics[bb=5 180 609 614,clip=,angle=0,width=6.cm]{f1c.eps}          
611: \caption{\chandra\ ACIS-S  images of Q2122-444, Q2130-431, Q2131-427 in the 0.3--8~keV band.
612: The left and middle images were smoothed using the sub-package {\it fadapt} of     
613: {\it FTOOLS} with a circular top hat filter of adaptive size, adjusted so as to
614: include a minimum number of 10 counts under the filter; each final     
615: pixel is 0{\farcs}1. The image of  Q2131-427 (right panel) is unsmoothed.}  
616: \label{figure:fig1}   
617: \end{figure}     
618: 
619: %%--------------FIG2-------spectrum & residuals-------------------------
620: \begin{figure}
621: \includegraphics[bb=40 2 567 704,clip=,angle=-90,width=9.cm]{f2a.eps}\includegraphics[bb=40 2 567 704,clip=,angle=-90,width=9.cm]{f2b.eps}
622: \caption{Spectra of Q2122-444 and Q2130-431 and data/model ratios to a simple power-law model modified by
623: photoelectric absorption in the Milky Way.}
624: \label{figure:fig2}
625: \end{figure}
626: 
627: %%--------------FIG3-------NH-Gamma contours-------------------------
628: \begin{figure}
629: \includegraphics[bb=120 95 545 640,clip=,angle=-90,width=9.cm]{f3a.eps}\includegraphics[bb=120 95 545 640,clip=,angle=-90,width=9.cm]{f3b.eps}
630: \caption{Confidence contours (68\%, 90\%, and 99\%) in the $\Gamma$ --
631: $N_{\rm H}$ plane for Q2122-444 and Q2130-431.}
632: \label{figure:fig3}
633: \end{figure}
634: 
635: \input{tab1}
636: \input{tab2}
637: \input{tab3}
638: \input{tab4}
639: \end{document}
640:  However, based on optical
641: observations alone, a contribution from a jet component cannot be firmly
642: ruled out.
643: 
644: 
645: