1: %\documentclass[preprint2]{$DATA/latexstuff/apjl2004/aastex52/aastex}
2: %\documentclass[12pt,preprint]{aastex}
3: \documentclass{emulateapj}
4: \usepackage{apjfonts}
5:
6: % last modified 19 Apr 05 by GCS
7:
8: \newcommand{\vdag}{(v)^\dagger}
9: \newcommand{\mum}{\ifmmode{\rm \mu m}\else{$\mu$m}\fi}
10: \newcommand{\iso}{{\em ISO}}
11: \newcommand{\iras}{{\em IRAS}}
12:
13: \begin{document}
14:
15: \title{The detection of silicate emission from quasars at 10 and 18 microns}
16:
17: \author{
18: Lei~Hao\altaffilmark{1},
19: H.~W.~W.~Spoon\altaffilmark{1,2},
20: G.~C.~ Sloan\altaffilmark{1},
21: J.~A.~Marshall\altaffilmark{1},
22: L.~Armus\altaffilmark{3},
23: A.~G.~G.~M.~Tielens\altaffilmark{4},
24: B.~Sargent\altaffilmark{5},
25: I.~M.~van~Bemmel\altaffilmark{6},
26: V.~Charmandaris\altaffilmark{1,7,8},
27: D.~W.~Weedman\altaffilmark{1},
28: J.~R.~Houck\altaffilmark{1}
29: \email{haol@isc.astro.cornell.edu}
30: }
31: \altaffiltext{1}{Cornell University, Astronomy Department,
32: Ithaca, NY 14853-6801}
33: \altaffiltext{2}{Spitzer Fellow}
34: \altaffiltext{3}{Caltech, Spitzer Science Center, MS 220-6, Pasadena, CA 91125}
35: \altaffiltext{4}{SRON National Institute for Space Research and
36: Kapteyn Institute, P.O. Box 800, 9700 AV Groningen, The Netherlands}
37: \altaffiltext{5}{University of Rochester, Department of Physics and
38: Astronomy, Rochester, NY 14627}
39: \altaffiltext{6}{Space Telescope Science Institute, 3700 San Martin Drive,
40: Baltimore, MD 21218}
41: \altaffiltext{7}{University of Crete, Department of Physics,
42: P.~O. Box 2208, GR-71003, Heraklion, Greece}
43: \altaffiltext{8}{Chercheur Associ\'{e} Obs. de Paris, 61 Ave. de
44: l'Observatoire, F-75014 Paris, France}
45:
46: \slugcomment{Submitted to Astrophysical Journal Letters, 2005 March 24,
47: accepted 2005 April 19}
48:
49: \begin{abstract}
50:
51: We report the spectroscopic detection of silicate emission at 10 and
52: 18\,\mum\ in five PG quasars, the first detection of these two features in
53: galaxies outside the Local Group. This finding is consistent with the
54: unification model for Active Galactic Nuclei (AGNs), which predicts
55: that an AGN torus seen pole-on should show a silicate emission feature
56: in the mid-infrared. The strengths of the detected silicate emission
57: features range from 0.12 to 1.25 times the continuum at 10\,\mum\ and
58: from 0.20 to 0.79 times the continuum at 18\,\mum. The silicate grain
59: temperatures inferred from the ratio of 18-to-10\,$\mu$m silicate
60: features under the assumption of optically thin emission range from
61: 140 to 220\,K.
62:
63: \end{abstract}
64:
65: \keywords{galaxies: active --- quasars: emission lines --- galaxies: ISM ---
66: infrared: galaxies}
67:
68: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
69:
70: \section{Introduction}
71:
72: Active galactic nuclei (AGNs) are broadly classified in two
73: types. Type 1 AGNs display broad hydrogen emission lines in
74: the optical, while Type 2 AGNs do not. The AGN unification
75: model \citep[e.g.][]{ant93} ties these two types together. An
76: optically and geometrically thick dusty torus surrounds a
77: central black hole, accretion disk and broad-line emission
78: region. Sources viewed face-on are recognized as Type 1,
79: while edge-on sources are Type 2. Observations at many
80: wavelengths support this scenario \citep[e.g.][]{am85}, but
81: until now there was a lack of constraining data in the
82: mid-infrared.
83:
84: Silicate dust is a major component of the interstellar medium
85: in the Milky Way and other galaxies. It produces two spectral
86: features in the infrared, a ``10\,\mum\ feature'' which arises
87: from a Si$-$O stretching mode, and an ``18\,\mum\ feature'' from
88: an O$-$Si$-$O bending mode \citep[e.g.][]{kt73}. Emission or
89: absorption from silicate dust is a dominant feature of the
90: infrared spectrum produced by most mass-losing evolved stars,
91: interstellar clouds, H{\sc ii} regions, and circumstellar disks
92: around young stellar objects (YSOs). Silicate absorption is
93: also detected in dusty starburst galaxies \citep{rig99},
94: ultraluminous infrared galaxies \citep[ULIRGs;][]{gen98,
95: tra01, spo04, arm04}, and a sample of mid-infrared detected,
96: optically invisible, high-luminosity galaxies with redshifts
97: of 1.7 $<$ z $<$ 2.8 \citep{hou05}.
98:
99: Type 2 AGNs show silicate dust in absorption as expected, but the
100: mid-infrared spectroscopic data for Type 1 AGNs have been scarce.
101: Recently, \cite{sie05} have presented spectra of two
102: quasars with just the 10\,\mum\ feature in emission. While radiative
103: transfer models clearly predict silicates to be in emission for Type 1
104: sources \citep{ld93, pk92, pk93, gd94, err95, nie00, nie01, nie02,
105: vbd03, dvb05}, the absence of the silicate emission features at 10 and
106: 18\,\mum\ has led to models without strong silicate emission.
107: \cite{ld93} and \cite{vbd03} have suggested that larger grains
108: dominate the grain-size distribution. \cite{nie02}, instead,
109: proposed that if the torus is clumpy, the 10\,\mum\ silicate emission
110: can be sufficiently suppressed, although \cite{dvb05} have challenged
111: this suggestion.
112:
113: Here, we report the first spectroscopic detection of both
114: silicate features at high signal to noise in five PG quasars.
115: Quasars are the high luminosity counterparts of Seyfert 1 galaxies,
116: showing the same broad emission lines that define Type 1 AGNs.
117:
118:
119: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
120:
121: \section{Observations and data reduction}
122:
123: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
124:
125: The five sources presented in this letter (PG 0804+761, PG 1211+143,
126: PG 1351+640, I Zw 1 [=PG 0050+124] and 3C~273 [=PG 1226+023]) were
127: selected from a sample of 12 AGNs based
128: upon their prominant silicate emission features at both 10 and
129: 18\,\mum. The observations were part of the Guaranteed Time
130: Observation (GTO) program of the Infrared Spectrograph
131: (IRS)\footnote{The IRS was a collaborative venture between Cornell
132: University and Ball Aerospace Corporation funded by NASA through the
133: Jet Propulsion Laboratory and the Ames Research Center}\citep{hou04}
134: on the Spitzer Space Telescope \citep{wer04}. Table \ref{tab1} lists
135: the basic properties of these targets, along with their observation
136: dates and on-source integration times. All of them have broad
137: H$\beta$ lines \citep{bg92} and therefore are classified as Type 1
138: AGNs. Two of them, 3C~273 and PG 1211 are radio-loud quasars and the
139: rest are radio-quiet \citep{san89}. All five quasars are very
140: luminous, as can be seen from the bolometric luminosities in
141: Table\,\ref{tab1} \citep[from][]{san89}.
142:
143: \begin{table} % Table 1
144: \caption{Properties of Sources\label{tab1}}
145: \begin{tabular}{lccccc}
146: \colrule
147: \colrule
148: & PG 1351 & I Zw 1 & 3C 273 & PG 1211 & PG 0804 \\
149: \colrule
150: AOR key & 3760640 & 3761920 & 4978176 & 3760896 & 9074944 \\
151: Date observed & 2004/4/17 & 2004/1/7 & 2004/1/6 & 2004/1/7 & 2004/3/1\\
152: Integration time (s) & 408 & 224 & 336 & 408 & 408\\
153: Redshift & 0.0881 & 0.0611 & 0.158 & 0.0808 & 0.0999 \\
154: Lum. distance (Mpc)\tablenotemark{a} & 398 & 271 & 749 & 363 & 455 \\
155: log($L_{\rm bol}/L_{\odot}$) & 12.14 & 12.20 & 13.44 & 12.44 & 12.56\\
156: F$_{15\,\mum}$ (Jy)\tablenotemark{b} & 0.21 & 0.62 & 0.36 & 0.21 & 0.13\\
157: \colrule
158: \end{tabular}
159: \tablenotetext{a}{assuming H$_0$=71\,km\,s$^{-1}$\,Mpc$^{-1}$,
160: $\Omega_M$=0.27, $\Omega_{\Lambda}$=0.73, $\Omega_K$=0}
161: \tablenotetext{b}{rest frame}
162: \end{table}
163:
164: The observations were made with the Short-Low (SL) and Long-Low (LL)
165: modules of the {\em IRS}. The spectra were extracted from the
166: flatfielded images
167: provided by the Spitzer Science Center (pipeline version S11.0.2). The
168: images were background-subtracted by differencing the two SL apertures
169: and for LL, by differencing the two nod positions. Spectra were then
170: extracted and calibrated using the {\em IRS} standard star HR 6348 for
171: SL and the stars HR 6348, HD 166780, and HD 173511 for LL
172: \citep{slo05}.
173:
174: After extraction the orders were stitched, requiring order-to-order
175: scaling adjustments of less than 5\%. Finally, the stitched spectra
176: were scaled to match the observed {\em IRAS} fluxes at 12 and
177: 25\,\mum\ or {\em IRS} blue peak-up flux at 16~\mum. For 3C 273, we
178: did not attempt any scaling, since the source varies by up to 0.4
179: magnitudes at 10.6\,\mum\ on time scales of up to several years
180: \citep{neu99}. The scaling factors we applied average 6\%, with
181: the largest scaling factor of 16\%. The observed sources are point
182: sources for both IRAS and IRS, therefore scaling to IRAS flux is
183: justified. Figure \ref{fig1} presents the calibrated spectra. Table
184: \ref{tab1} lists the 15\,\mum\ rest frame fluxes.
185:
186: \begin{figure} % Fig. 1
187: \begin{center}
188: \resizebox{\hsize}{!}{\includegraphics{f1.ps}}
189: \caption{The 5--35\,\mum\ IRS low resolution spectra of five PG
190: quasars showing silicates in emission. The pixel-to-pixel
191: uncertainty in the data is less than 3\%. Fringing effects with
192: amplitudes of up to 5\% are visible in LL order 1 (see text). For
193: each spectrum, three choices for the silicate-graphite continuum
194: are indicated by the dashed line and grey shaded area.
195: The spectra have been scaled for plotting purposes, with scaling
196: factors of: 8.0 for PG\,1351, 1.0 for I\,Zw1, 0.77 for 3C\,273
197: and PG\,1211 and 0.50 for PG\,0804.
198: \label{fig1}}
199: \end{center}
200: \end{figure}
201:
202: In some spectra, most notably in IZw1, residual fringing, which
203: are due to minor pointing errors, appears between 21 and 30\,\mum\ (the
204: 1st order of the LL module). We have not defringed the spectra.
205: Residual artifacts also appear in some spectra in SL order 1
206: (7.5--14\,\mum). These take the form of a roughly sinusoidal
207: deviation from the actual spectrum and are most obvious in the
208: spectrum of 3C 273 at 8--10\,\mum, where they have a maximum strength
209: of $\sim$3\% of the total flux. Because 3C 273 has the weakest
210: silicate emission feature in our sample, this artifact is even more
211: pronounced in figures showing the continuum-subtracted spectra. We
212: estimate uncertainties by comparing the spectra in the two nod
213: positions, and these artifacts generally manifest themselves as a
214: disagreement between the nod positions. The uncertainties only become
215: apparant in Figures \ref{fig2} and \ref{fig3} shortward of 11\,\mum\
216: (dark shaded area). None of these artifacts change the conclusions
217: reported below.
218:
219: \begin{figure} % Fig. 2
220: \begin{center}
221: \resizebox{\hsize}{!}{\includegraphics{f2.ps}}
222: \caption{Continuum-subtracted silicate emission spectra of the
223: five PG quasars. The spectra have been normalized and offset
224: from each other for plotting purposes. Fringing effects are
225: visible in LL order 1 (see text). The shaded areas indicate the
226: uncertainty in the silicate emission spectra as introduced by the
227: uncertainty in the observed spectrum and choice of continuum.
228: Vertical dotted lines indicate the rest wavelengths of emission
229: features discussed in the text.
230: \label{fig2}}
231: \end{center}
232: \end{figure}
233:
234: \begin{figure} % Fig. 3
235: \begin{center}
236: \resizebox{\hsize}{!}{\includegraphics{f3.ps}}
237: \caption{
238: The opacity spectra for the five PG quasars compared to the
239: scaled synthetic silicate opacity curve of \cite{wein01} and
240: \cite{li01}. The opacity spectra from the quasars are derived
241: using the adopted silicate grain temperature and scaled and offset
242: for plotting purposes. The shaded areas indicate the uncertainty
243: in the silicate profile as introduced by the uncertainty in the
244: observed spectrum and choice of continuum.
245: \label{fig3}}
246: \end{center}
247: \end{figure}
248:
249: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
250:
251: \section{Analysis}
252:
253: The spectra in Figure \ref{fig1} are dominated by two broad emission
254: features, which we identify as the Si--O stretching mode and the
255: O--Si--O bending mode
256: in silicate dust grains. The strength of the features
257: is best appreciated by comparing the peak fluxes to the underlying
258: silicate-graphite ``continuum''. For PG 1351, I Zw 1 and PG 1211, we
259: define the continuum using a spline interpolation between the
260: 5--8\,\mum\ spectrum and a point at the extreme red end of the IRS
261: spectral range, where a change in slope indicates the end of the
262: 18\,\mum\ silicate feature. For 3C 273 and PG 0804, the spectral
263: structure at the red end of the IRS spectral range does not show a
264: similar change in slope. This may indicate that dust emission is
265: still contributing at these wavelengths. Therefore, we have assumed
266: that the continuum at the red end runs somewhat below the observed
267: spectrum. Figure \ref{fig1} plots the resulting spline-interpolated
268: continua (dashed lines), while Figure \ref{fig2} shows the resulting
269: continuum-subtracted spectra.
270: %The effect of slightly different choices of continua is
271: %illustrated by the grey shaded areas in both Figures.
272:
273: Comparison of the silicate emission spectra in Figure \ref{fig2}
274: reveals that the ratio of the 18 and 10\,\mum\ features varies
275: significantly within the sample. It is highest for 3C 273 and lowest
276: for PG 0804 (see Table \ref{tab2}). Also the onset and center of the
277: 10\,\mum\ features shift significantly from one source to the next, 3C
278: 273 and PG 1211 seem to appear at longer wavelengths. Interestingly,
279: these two sources happen to be the only radio loud quasars in our
280: sample. In order to test whether these differences can be attributed
281: to our choice of continuum, for each source we compared the silicate
282: emission spectra for two other choices of underlying silicate-graphite
283: continuum: one above and one below the adopted continuum, enclosing
284: the shaded areas in Figure \ref{fig1}. As illustrated by the shaded
285: areas in Figure \ref{fig2}, this results in an uncertainty of up to
286: 30\% in the peak flux of the 18\,$\mu$m feature and in the ratio of
287: the 18 to 10\,\mum\ features. In contrast, the onset and center of the
288: 10\,\mum\ features are not affected by different choices of continuum.
289:
290: \begin{table} % Table 2
291: \caption{Measured quantities\label{tab2}}
292: \begin{tabular}{lccccc}
293: \colrule
294: \colrule
295: & PG 1351 & I Zw 1 & 3C 273 & PG 1211 & PG 0804 \\
296: \colrule
297: 10$\mu$m-feature-to-cont. ratio\tablenotemark{a} & 1.25 & 0.38 & 0.12 & 0.55 & 0.60 \\
298: 18$\mu$m-feature-to-cont. ratio\tablenotemark{a} & 0.79 & 0.20 & 0.36 & 0.56 & 0.56 \\
299: 18$\mu$m\,/\,10$\mu$m feature ratio\tablenotemark{a} & 1.75 & 1.31 & 3.34 & 1.89 & 1.08 \\
300: F(30$\mu$m)/F(5.5$\mu$m) & 11.0 & 6.86 & 2.53 & 2.99 & 1.95 \\
301: L$_{\rm sil}$/L$_{\rm bol}$ & 0.087 & 0.040 & 0.009 & 0.023 & 0.021\\
302: 10/18$\mu$m cont. color temp (K)\tablenotemark{b} & 250 & 270 & 320 & 310 & 340 \\
303: %Silicate temperature (K) & 200 & 230 & 150 & 180 & 220 \\
304: Silicate temperature (K) & 200 & 220 & 140 & 180 & 210 \\
305: \colrule
306: \end{tabular}
307: \tablenotetext{a}{evaluated at the peak of the feature}
308: \tablenotetext{b}{the color temperature is computed from the adopted continuum at 10 and 18\,\mum\ }
309: \end{table}
310:
311: The presence of the silicate features at 10 and 18\,\mum\
312: allows us to estimate the mean tempeature of the emitting grains by
313: performing an unweighted $\chi^2$ fit to the continuum-subtracted
314: spectrum. This temperature is only physically meaningful under the
315: assumption that the feature emitting region is optically thin. Using
316: the continuum-subtracted silicate opacity (see top panel in Figure
317: \ref{fig3}) derived from \cite{wein01} and \cite{li01}, we force the
318: product of a blackbody and the silicate feature opacities to fit the
319: observed continuum-subtracted silicate features. This reduces the
320: problem to a single free parameter: the grain temperature, and avoids
321: concerns related to a possible nonthermal contribution to the continuum.
322: The resulting silicate grain temperatures (T$_{\rm sil}$) range between
323: 140\,K for 3C 273 and 220\,K for I Zw 1, see Table \ref{tab2}. These
324: temperatures will change by as much as 15\,K depending on the choice
325: for the underlying silicate-graphite continuum. Uncertainties in the
326: fitting process may amount to as much as another 10\,K.
327:
328: Using the derived silicate grain temperature (T$_{\rm sil}$),
329: it is possible to compute the silicate opacity spectrum by
330: dividing the observed silicate emission spectrum by the
331: blackbody spectrum with temperature T$_{\rm sil}$.
332: Figure \ref{fig3} presents the resulting opacity spectra
333: for our sample.
334:
335: The top panel in Figure \ref{fig3} shows the synthetic silicate
336: opacity spectrum after subtraction of the silicate-graphite continuum
337: component. The lower panels show the opacity spectra for the five
338: quasars overlaid on the synthetic profile. The quasar profiles
339: generally show good agreement at wavelengths beyond 10\,\mum. Below
340: 10\,\mum, only PG 1351 and I Zw 1 match the synthetic profile well.
341: In the other three spectra, the blue wing of the 10\,\mum\ appears
342: weak, especially in 3C 273 and PG 1211. Apart from the artifacts in
343: the 8--10\,\mum\ region of the spectrum of 3C 273 (indicated by the
344: large error bars in Figure \ref{fig2}; see \S 2) and the presence of
345: atomic lines and PAH emission in some spectra, the 10\,\mum\ band has
346: a smooth appearance, showing no sign of a departure from amorphous
347: grain structure. Evidence for some crystalline grains do appear at
348: longer wavelengths, most notably in the spectrum of PG 1351 at
349: 23\,\mum.
350:
351: The weakness of the blue wing of the 10\,$\mu$m silicate opacity
352: profile in three of our sources appears to be significant and cannot
353: be explained by the uncertainties in the observed spectrum (see the
354: error bars in Figure \ref{fig3}). Grain size and composition,
355: geometry, optically thick radiative transfer or a combination of these
356: factors may explain the observed weakness. If our optically thin
357: assumption is invalid, then the temperatures we have inferred cannot
358: be interpreted as real dust temperatures. \cite{jaf04} reported a
359: similar deficiency in the blue wing of the 10um absorption feature in
360: NGC1068, which they attribute to a different silicate grain
361: composition.
362:
363:
364: A close inspection of the spectra in Figures \ref{fig1} and \ref{fig2}
365: reveals numerous other features. Those visible in at least one
366: spectrum in the sample include the PAH emission features at 6.2, 7.7
367: and 11.2\,\mum, and emission lines at 10.5\,\mum\ ([S{\sc iv}]),
368: 12.8\,\mum\ ([Ne{\sc ii}]), 14.3\,\mum\ ([Ne{\sc v}]), 15.6\,\mum\
369: ([Ne{\sc iii}]), 18.7\,\mum\ ([S{\sc iii}]), and 25.9\,\mum\ ([O{\sc
370: iv}]).
371:
372: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
373:
374: \section{Comparison with Models}
375:
376: Direct comparisons of our observations to the predictions of published
377: AGN radiative transfer models are difficult given the limited resolution
378: of model spectra in the 5--40\,\mum\ range of the IRS. Additionally,
379: the model output is not tuned to be compared to the observed
380: properties presented in Table\,\ref{tab2}. However, we can measure the
381: feature-to-continuum ratio at 10\,\mum\ from the figures presented by
382: \cite{ld93, gd94, nie02, dvb05}. For the optically thin dust models
383: of \cite{ld93} the values range from 0.1 for models where the
384: grain-size distribution is dominated by larger grains to 1.5 for
385: models with a standard grain-size distribution. In the face-on cases
386: modeled by \cite{gd94}, the values range from near zero for very thick and
387: compact configurations to 1.0 for more extended torus models. These
388: results are similar to those by \cite{nie02} and \cite{dvb05}, but
389: their highest 10\,\mum\ feature-to-continuum ratios are around 0.7 and
390: 0.4, respectively. Given the observed range of 0.12 to 1.25 within
391: our limited sample (Table \ref{tab2}), the model predictions for the
392: 10\,\mum\ feature-to-continuum ratio are in good agreement. A similar
393: comparison for the 18\,\mum\ feature-to-continuum ratio could not be
394: undertaken, since the features are not as apparent as the 10\,\mum\
395: features in the AGN models.
396:
397:
398: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
399:
400: \section{Discussion and Conclusion}
401:
402: We have detected the 10 and 18\,\mum\ silicate emission
403: features in five PG quasars --- the first detection of both
404: silicate features in emission in galaxies outside the Local
405: Group and in AGNs in particular. Given the existence of
406: published mid-infrared photometric and spectroscopic data for
407: these quasars, our finding may appear surprising at first.
408: \cite{roc91} obtained 8--13\,\mum\ spectra of 3C 273 and
409: I Zw 1, but the low contrast of the silicate feature to the
410: strong continnum and the lack of coverage outside the
411: N-band window prevented them from identifying silicate emission.
412: The {\it PHT-S} spectrometer on \iso\ \citep{lem96} obtained
413: spectra of 3C 273, I Zw 1 and PG 0804 \citep{rig99},
414: but the limited wavelength coverage of {\it PHT-S} (only to
415: 11.6\,\mum) and the redshifts of the quasars allowed only the
416: blue side of the 10~\mum\ feature to be visible. Because of
417: these difficulties, silicate emission had remained unidentified
418: in these five sources until now.
419:
420: Our detection of silicate emission in Type 1 AGNs provides support for
421: the unified AGN model \citep{ant93}. An AGN torus viewed pole-on
422: should offer an unobstructed line of sight to the torus dust, the
423: surface of which is heated to temperatures of several hundred to a
424: thousand Kelvin by the radiation from the central engine. Emission
425: from this hot surface should produce a silicate feature in emission
426: under most circumstances.
427:
428: Assuming a simple optically thin geometry, we find that the silicate
429: grain temperature in our sources ranges from 140 to 220\,K, which is
430: well below the dust sublimation temperature. Detailed comparison of
431: the derived opacity profiles for our quasars indicates that our simple
432: model cannot satisfactorally explain the absence of a blue wing in the
433: 10\,\mum\ feature for three of our quasars. This likely points to a
434: more complex geometry, involving radiative transfer within a
435: (partially) optically thick environment -- probably an AGN torus. It
436: may also point to differences in the grain size or composition. Our
437: derived silicate grain temperatures should, therefore, only be used as
438: a characterization of the observed spectra, not as a means to infer
439: the actual temperature and geometry of the AGN.
440:
441: From our limited sample of PG quasars (three in addition to the sample
442: discussed in this Letter) and a few Seyfert 1 galaxies, it is clear
443: that 10\,\mum\ silicate emission is not a ubiquitous feature in
444: mid-infared spectra of Type 1 AGNs. Only 5 out of 12
445: continuum-dominated AGNs show this feature prominently. Interestingly,
446: the quasars which show a 10\,\mum\ silicate emission feature all have
447: bolometric luminosities higher than 10$^{12}$ L$_{\odot}$, while the
448: other Type 1 AGNs are less luminous. Models which can
449: explain the presence of the 10\,\mum\ silicate emission feature in
450: some type 1 AGNs will also need to explain their absence in others.
451:
452: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
453:
454: \acknowledgments
455:
456: The authors wish to thank Bill Forrest, Elise Furlan and Terry Herter
457: for their useful discussions. Support for this work was provided by
458: NASA through Contract Number 1257184 issued by the Jet Propulsion
459: Laboratory, California Institute of Technology under NASA contract
460: 1407. HWWS was supported under this contract through the
461: Spitzer Space Telescope Fellowship Program.
462:
463:
464: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
465:
466:
467: \begin{thebibliography}{}
468:
469: \bibitem[Antonucci \& Miller(1985)]{am85} Antonucci, R.~R.~J. \&
470: Miller, J.~S. 1985, \apj, 297, 621
471: \bibitem[Antonucci(1993)]{ant93} Antonucci, R. 1993, \araa, 31, 473
472: \bibitem[Armus et al.(2004)]{arm04} Armus, L., et al. 2004, \apj, 154, 178
473: %\bibitem[Chiar et al.(2000)]{chi00} Chiar, J.~E., et al. 2000, \apj,
474: % 537, 749
475: \bibitem[Boroson \& Green(1992)]{bg92} Boroson, T.~A. \& Green,
476: R.~F. 1992, \apjs, 80, 109
477: \bibitem[Chihara et al.(2002)]{chi02} Chihara, H., Koike, C.,
478: Tsuchiyama, A., Tachibana, S., \& Sakamoto, D. 2002, \aap, 391, 267
479: \bibitem[Dullemond \& van Bemmel(2005)]{dvb05} Dullemond, C.~P. \&
480: van Bemmel, I.~M. 2005, \aap, in press, astro-ph/0501570
481: \bibitem[Granato \& Danese(1994)]{gd94} Granato, G.~L. \& Danese, L.
482: 1994, \mnras, 268, 235
483: \bibitem[Green et al.(1986)]{gsl86} Green, R.~F., Schmidt, M. \&
484: Liebert, J. 1986, \apjs, 61, 305
485: \bibitem[Houck et al.(2004)]{hou04} Houck, J.~R., et al. 2004, \apjs,
486: 154, 18
487: \bibitem[Houck et al.(2005)]{hou05} Houck, J.~R., et al. 2005, \apj,
488: 622, L105
489: \bibitem[Efstathiou \& Rowan-Robinson (1995)]{err95} Efstathiou, A.,
490: Rowan-Robinson, M. 1995, \mnras, 273, 649
491: \bibitem[Genzel et al.(1998)]{gen98} Genzel, R., Lutz, D., Sturm, E.,
492: et al., 1998, \apj, 498, 579
493: \bibitem[Jaffe et al.(2004)]{jaf04} Jaffe et al. 2004, Nature, 429, 47
494: \bibitem[Knacke \& Thomson(1973)]{kt73} Knacke, R.~F., \& Thomson,
495: R.~K. 1973, \pasp, 85, 341
496: \bibitem[Koike et al.(2003)]{ko03} Koike, C., Chihara, H., Tsuchiyama, A.,
497: Suto, H., Sogawa, H., \& Okuda, H. 2003, \aap, 399, 110
498: \bibitem[Laor \& Draine(1993)]{ld93} Laor, A. \& Draine, B.~T. 1993,
499: \apj, 401, 441
500: \bibitem [Lemke et al.(1996)]{lem96} Lemke, D., Klaas, U., Abolins, J.,
501: et al. 1996, \aap, 315, L64
502: \bibitem[Li \& Draine(2001)]{li01} Li, A., Draine, B.~T., 2001, \apj,
503: 554, 778
504: \bibitem[Maiolino et al.(2001)]{mai01} Maiolino, R., Marconi, A.,
505: Salvati, M., Risaliti, G., Severgnini, P., Oliva, E., La Franca, F.,
506: \& Vanzi, L. 2001, \aap, 365, 28
507: \bibitem[Nenkova et al.(2000)]{nie00} Nenkova, M., Ivezi\'{c}, \v{Z}. \&
508: Elitzur, M. 2000, ASPC, 196, 77
509: \bibitem[Nenkova et al.(2001)]{nie01} Nenkova, M., Ivezi\'{c}, \v{Z}. \&
510: Elitzur, M. 2001, ASPC, 247, 283
511: \bibitem[Nenkova et al.(2002)]{nie02} Nenkova, M., Ivezi\'{c}, \v{Z}. \&
512: Elitzur, M. 2002, \apj, 570, L9
513: \bibitem[Neugebauer et al.(1999)]{neu99} Neugebauer, G.; Matthews, K.
514: 1999, \aj, 118, 35
515: \bibitem[Pier \& Krolik(1992)]{pk92} Pier, E.~A., Krolik, J.~H. 1992,
516: \apj, 401, 99
517: \bibitem[Pier \& Krolik(1993)]{pk93} Pier, E.~A., Krolik, J.~H. 1993,
518: \apj, 418, 673
519: \bibitem[Rigopoulou et al.(1999)]{rig99} Rigopoulou, D., Spoon, H.~W..~W.,
520: Genzel, R., Lutz, D., Moorwood, A.~F.~M., Tran, Q.~D., 1999,
521: \aj, 118, 2625
522: \bibitem[Roche et al.(1991)]{roc91} Roche, P.~F., Aitken, D.~K., Smith,
523: C.~H. \& Ward, M.~J. 1991, \mnras, 248, 606
524: \bibitem[Sanders et al.(1989)]{san89} Sanders, D.~B., Phinney, E.~S.,
525: Neugebauer, G., Soifer, B.~T. \& Matthews, K. 1989, \apj, 347 29
526: \bibitem[Siebenmorgen et al.(2005)]{sie05} Siebenmorgen, R., Haas, M.,
527: Kr\"{u}gel, E., Schulz, B., astro-ph/0504263
528: \bibitem[Sloan et al.(2005)]{slo05} Sloan et al. 2005, in preparation
529: \bibitem[Spoon et al.(2004)]{spo04} Spoon, H.~W.~W., Armus, L., Cami, J.,
530: et al., 2004, \apjs, 154, 184
531: \bibitem[Tran et al.(2001)]{tra01} Tran, Q.~D., Lutz, D., Genzel, R.,
532: et al., 2001, \apj, 552, 527
533: \bibitem[van Bemmel \& Dullemond(2003)]{vbd03} van Bemmel, I.~M. \&
534: Dullemond, C.~P. 2003, \aap, 404, 1
535: \bibitem[Weingartner \& Draine(2001)]{wein01} Weingartner, J.~C.,
536: Draine, B.~T., 2001, \aj, 548, 296
537: \bibitem[Werner et al.(2004)]{wer04} Werner, M.~W., et al. 2004, \apjs,
538: 154, 1
539: \end{thebibliography}
540:
541:
542: \end{document}
543: