1: \documentclass[12pt,preprint]{aastex}

2: %\usepackage[T1]{fontenc}

3: %\usepackage[latin1]{inputenc}

4: %\setcounter{tocdepth}{3}

5: %\usepackage{graphicx}

6: 

7: %\makeatletter

8: % User specified LaTeX commands.

9: %\documentclass[12pt,preprint]{aastex}

10: % \documentclass[manuscript]{aastex}

11: 

12: 

13: % p2ms10.tex 24 March 2008: revisions requested by referee

14: % 

15: 

16: \newcommand{\iso}{{\em ISO}}

17: \newcommand{\iras}{{\em IRAS}}

18: \newcommand{\mum}{\ifmmode{\rm \mu m}\else{$\mu$m}\fi}

19: \newcommand{\er}{\ifmmode{\pm}\else{$\pm$}\fi}

20: 

21: % The following are for graphics conversion in TeXShop (Max OSX)

22:  

23: %\usepackage{epstopdf}

24: \usepackage{graphicx}

25: 

26: \slugcomment{Accepted for publication in the Astrophysical Journal, 2008}

27: \shorttitle{PAHs in Herbig AeBe Stars}

28: \shortauthors{Keller et al.}

29: 

30: %\usepackage{babel}

31: %\makeatother

32: \begin{document}

33: 

34: \title{PAH emission from Herbig AeBe stars}

35: 

36: \author{Luke~D.~Keller\altaffilmark{1}, G.~C.~Sloan\altaffilmark{2}, W.~J.~Forrest\altaffilmark{3},  S.~Ayala\altaffilmark{4}, P.~D'Alessio\altaffilmark{4}, S.~Shah\altaffilmark{1}, N.~Calvet\altaffilmark{5},  J.~Najita\altaffilmark{6},

37: A.~Li\altaffilmark{7}, L.~Hartmann\altaffilmark{5}, B.~Sargent\altaffilmark{3}, D.~M.~Watson\altaffilmark{3}, \& C.~H.~Chen\altaffilmark{8}}

38: 

39: %E.~Furlan\altaffilmark{6}

40: %N.~Chitrakar\altaffilmark{1},

41: %B.~R.~Brandl\altaffilmark{9},

42: %J.~D.~Green\altaffilmark{3},

43: %T.~F.~Herter\altaffilmark{2},

44: %F.~Markwick-Kemper\altaffilmark{6}

45: %D.~J.~Barry\altaffilmark{2}

46: %P.~Hall\altaffilmark{2}

47: %P.~C.~Meyers\altaffilmark{8} \&

48: %J.~R.~Houck\altaffilmark{2}

49: 

50: \altaffiltext{1}{Department of Physics, Ithaca College, Ithaca,

51: NY 14850} 

52: \altaffiltext{2}{Cornell University, Astronomy Department,

53: Ithaca, NY 14853-6801} 

54: \altaffiltext{3}{Department

55: of Physics and Astronomy, University of Rochester, Rochester, NY 14627-0171}

56: \altaffiltext{4}{Centro de Radioastronomia y Astrofisica, UNAM, 

57: Apartado Postal 3-72 (Xangari), 58089 Morelia, Michoacan, Mexico}

58: \altaffiltext{5}{Department

59: of Astronomy, University of Michigan, Ann Arbor, MI 48109}

60: \altaffiltext{6}{National

61: Optical Astronomy Observatory, 950 North Cherry Avenue, Tucson, AZ

62: 85719}

63: \altaffiltext{7}{Department of Physics \& Astronomy, University

64: of Missouri, Columbia, MO 65211} 

65: \altaffiltext{8}{Space Telescope Science Institute, 3700 San Martin Drive

66: Baltimore, MD 21218}

67: 

68: 

69: 

70: %%\altaffiltext{6}{Astronomy Department, University of California Los Angeles}

71: %\altaffiltext{8}{NASA {\it Herschel} Science Center, IPAC/Caltech, 

72: %  MS 100-22, Pasadena, CA  91125}

73: %\altaffiltext{9}{Sterrewacht Leiden, P.O. Box 9513, 2300 RA Leiden, 

74: %  The Netherlands}

75: 

76: \begin{abstract}

77: We present spectra of a sample of Herbig Ae and Be (HAeBe) stars obtained

78: with the Infrared Spectrograph on the \textit{Spitzer Space Telescope}. All but one of the Herbig

79: stars show emission from polycyclic aromatic hydrocarbons (PAHs) and seven of

80: the spectra show PAH emission, but no silicate emission at 10~\mum. The central wavelengths of the 6.2, 7.7--8.2, and 11.3~\mum\ emission features decrease with stellar temperature,

81: indicating that the PAHs are less photo-processed in cooler radiation fields.

82: The apparent low level of photo processing in HAeBe stars, relative to other PAH emission sources, implies that the PAHs are newly exposed to the UV-optical radiation fields from

83: their host stars. HAeBe stars show a variety of PAH emission intensities and ionization fractions, but a narrow range of PAH spectral classifications based on positions of major

84: PAH feature centers. This may indicate that, regardless of their locations relative

85: to the stars, the PAH molecules are altered by the same physical processes in the proto-planetary disks of intermediate-mass stars. Analysis of the mid-IR spectral energy distributions indicates that our sample likely includes both radially flared and more flattened/settled disk systems, but we do not see the expected correlation of overall PAH emission with disk geometry.

86: We suggest that the strength of PAH emission from HAeBe stars may depend not only on the degree of radial flaring, but also on the abundance of PAHs in illuminated regions of the disks and possibly on the vertical structure of the inner disk as well.

87: \end{abstract}

88: 

89: \keywords{stars: Herbig Ae --- infrared: stars, proto-planetary disks}

90: 

91: 

92: \section{INTRODUCTION}

93: 

94: Circumstellar accretion disks that contain dust and gas are common

95: around intermediate mass stars. Studying the material in these disks

96: and their evolution can aid us in understanding the very latest stages

97: of star formation and the early stages of planet formation. Herbig Ae/Be 

98: (HAeBe) stars are intermediate-mass (M$\sim$2-10 M$_{\sun}$) 

99: pre-main-sequence stars of spectral class A, B, or early F 

100: \citep{her60, str72, the94, mal98}. Though often associated with nebulosity,

101: HAeBe photospheres are directly observable at visible wavelengths.

102: Unlike normal A and B stars, HAeBe stars have strong emission

103: lines (e.g. H$\alpha$ and Br$\gamma$) in their spectra, they can be highly variable

104: at visible wavelengths, and their infrared fluxes are strongly in

105: excess of purely photospheric emission. Current theory predicts that

106: intermediate-mass stars should form more quickly than lower-mass (T

107: Tauri) stars, but they will not reach their stellar birth lines heavily

108: enshrouded in dense circumstellar envelopes as high-mass stars do.

109: This means that we can directly view the stellar photospheres and

110: residual accretion disks of HAeBe stars, with a relatively unobscured

111: view of both. Unlike high-mass stars, while some HAeBe stars are associated

112: with star forming regions they are not common in the extreme environments

113: of HII regions near OB associations. Thus the stars themselves, and

114: not their environments, drive their disk structure, chemistry, and evolution.

115: 

116: Chemical and dynamical studies of Galactic HAeBe stars have shown

117: holes and gaps in their inner disks indicating the possible dynamical influence of newly

118: formed planets \citep[e.g.][]{gra05,bri07}, as well as evidence of different disk geometries

119: from those that flare

120: with increasing radius and have dust with a large abundance of small

121: grains, to flattened disks composed of larger dust grains \citep[e.g.][]{mee01,aa04}. These

122: characteristics are similar to those of the lower-mass T Tauri systems. 

123: 

124: Solid-state features and thermal radiation from warm dust, as well

125: as emission from atomic and molecular gas all produce

126: measurable signatures in the mid-IR spectra of young stars with

127: circumstellar disks. Mid-IR imaging and spectroscopy, much of it

128: from space telescopes, have

129: therefore played a major role in measuring and characterizing 

130: the disks orbiting T Tauri and HAeBe stars. Unlike T

131: Tauri stars, of which only $\sim$8\% show emission from

132: polycyclic aromatic hydrocarbons

133: \citep[PAHs,][]{gee06}, nearly 50\% of HAeBe stars have strong PAH emission

134: \citep[e.g.][]{mee01,aa04}, which requires a direct line of sight

135: from the emitting material---presumably located on the surfaces and/or

136: inner rims of disks---to the photospheric optical and UV radiation

137: fields. Furthermore, spectra from the Infrared Space Observatory (\iso)

138: and recent \textit{Spitzer} observations have revealed that many HAeBe

139: stars in the Galaxy have large abundances of crystalline silicate

140: grains in their disks indicating significant thermal dust processing

141: subsequent to the formation of the star-disk system. These latter

142: characteristics are not as common in T Tauri disks. Moreover, an evolutionary

143: link from HAeBe proto-planetary disks to Vega-like debris disks is

144: becoming evident in studies of large samples of HAeBe stars 

145: \citep[e.g.][]{her06,gra07}, but the details have yet

146: to be completely worked out. Though details of time scale and evolution

147: remain elusive, recent modeling efforts have succeeded

148: in predicting spectral energy distributions from varying physical characteristics

149: in the disks like dust grain size and composition, properties of the stellar radiation field, 

150: disk heating and cooling, and disk geometry \citep[e.g.][]{dul04,dal06,dul07,rob07}.

151: 

152: It is remarkable that such a large fraction of HAeBe stars are

153: PAH emission sources. \cite{mee01} suggested that PAH emission

154: may change measurably with the structure and physical properties of

155: the disks and perhaps with disk evolution. PAHs have been observed in many different

156: astrophysical contexts: the Galactic and extra-galactic ISM, dust envelopes 

157: surrounding post-AGB (asymptotic giant branch) stars, planetary nebulae (PNe),

158: and young stellar objects (YSOs). The molecular structure of PAHs can differ significantly

159: depending on the physical characteristics of the objects or environments

160: where they are found. The bottom line is that PAHs are abundant in

161: interstellar and circumstellar environments, and they can be important

162: constituents in the energy balance of those environments as sources of 

163: photoelectrons that heat the gas component \citep{kam04}. 

164: 

165: Emission from PAH molecules may

166: also trace proto-planetary disk geometry \citep{mee01,aa04,gra05}. Exactly what the

167: PAH emission can tell us about the disk environments will depend upon

168: a thorough and detailed understanding of how the molecules are energized and

169: processed by the stellar radiation fields and how they

170: are distributed radially and vertically in the disks. Throughout this paper when

171: we refer to "processing" of PAHs, we mean photo-processing by which 

172: the optical and UV stellar radiation break molecular bonds producing smaller

173: PAH molecules or completely destroying them.

174: We have begun a detailed analysis of PAH emission from HAeBe stars

175: using very high S/N spectra that allow much higher precision in analyzing

176: PAH feature strengths and wavelengths, and thus the molecular structure

177: that produces them, than has been possible before.

178: 

179: In a previous paper (Sloan et. al. 2005, hereafter Paper 1), we analyzed

180: four HAeBe stars in the 5-14~\mum\ range that have strong PAH emission

181: but no silicate peak at 10~\mum. We identified a trend towards decreasing

182: PAH molecule size with increasing PAH ionization fraction and noted

183: that the center wavelengths of the PAH features shift systematically

184: with increasing UV field strength from the stars. We suggested that

185: HAeBe stars may have distinctive PAH spectra relative to other

186: PAH sources (e.g. interstellar photon-dominated regions, PNe, etc).

187: 

188: We present 5-36~\mum\ spectra of 18 HAeBe stars, two intermediate-mass

189: T Tauri stars, HD 97300, and 51 Oph for a total of 22 sources observed with \textit{Spitzer}. 

190: This extends our original sample of HAeBe stars in Paper 1 by more than a factor of four, and we now

191: analyze HAeBe stars that do have the 10~\mum\ silicate emission feature. 

192: %Their spectral energy distributions (SEDs) cover a range

193: %dusty disks (broad NIR-FIR excesses), to disks with separate warm and

194: %cool components (peaks near 2~\mum\ and 15-17~\mum), and a few with single

195: %cool components (peaks at $\sim$60~\mum). 

196: We present a detailed analysis of the PAH feature strengths and their ratios

197: in an attempt to infer their physical environments and

198: to test for trends in PAH emission when compared with other diagnostics of

199: proto-planetary disk evolution that have recently been helpful in understanding

200: the evolution of T Tauri disks \citep[e.g.][]{fur06}.

201: 

202: The major trends suggested in

203: Paper 1 persist in the present, larger sample. \cite{aa04}

204: noted in their study of 46 HAeBe stars that those with strong mid-IR 

205: excesses relative to near-IR excesses have stronger PAH emission

206: than those with weaker mid-IR excesses. This result supports a relation

207: between the detailed structure of PAH emission and the evolution of the disks. 

208: Recent studies by \cite{gra05} suggest that the strength of (UV-optical) photon-dominated

209: spectral features should correlate with overall disk geometry. Disks

210: that are flat and whose outer disks are therefore shadowed or illuminated at

211: grazing incidence should

212: have weaker H$_{2}$ line and PAH features relative to disks that are flared

213: and therefore well-illuminated by the photospheres of their host stars. These results

214: imply a strong correlation between PAH strength and disk structure,

215: which we can now test with our data.

216: 

217: \section{OBSERVATIONS AND ANALYSIS}  % 2

218: 

219: \subsection{The Sample}

220: 

221: Table 1 summarizes the observations of our sample. SU Aur and HD 281789 

222: are intermediate-mass T Tauri

223: stars, which range in spectral class from K to late F (1.5-5.0 M$_\sun$). 

224: They are the evolutionary predecessors 

225: of HAeBe stars so their inclusion broadens our study to more primitive systems \citep{cal04}.

226: Both SU Aur and HD 281789 were included in the Taurus study of \cite{fur06},

227: in which the two stars appeared as clear outliers from the Class II \& III low-mass

228: YSOs that they focused on.

229: 51 Oph is almost certainly not an HAeBe star \citep{thi05,ber07},

230: but we include it here because its mid-IR spectrum indicates an optically

231: thick, gas- {\it and} dust-rich disk with no measurable PAH emission and

232: because it appears in many previous HAeBe studies. We argue below that HD 97300 is probably

233: not a HAeBe star either. The remaining 18 targets are HAeBe stars.

234: 

235: \subsection{Observing Methods and Data Processing}

236: 

237: We observed all of these sources with the Infrared 

238: Spectrograph \citep[IRS,][]{hou04}\footnote{The IRS was a 

239: collaborative venture between Cornell University and

240: Ball Aerospace Corporation funded by NASA through the Jet Propulsion

241: Laboratory and the Ames Research Center.} on the {\it Spitzer Space 

242: Telescope}.  All but two of the sources were observed using 

243: the Short-Low (SL) module from 5 to 14~\mum, the Short-High 

244: (SH) module from 10 to 19~\mum, and the Long-High (LH) module 

245: from 19 to 36~\mum.  The remaining two were observed using SL 

246: and Long-Low (LL; 15-36~\mum).  SL and LL have resolving powers of 

247: R$\sim$90, while SH and LH have R$\sim$600.

248: 

249: Three of the spectra were obtained in simple staring mode,

250: while the rest were obtained as part of clustered 

251: observations, with three slit positions stepped

252: across the nominal source positions.  For these observations, 

253: we extracted a spectrum from 

254: the center position, unless the source was bright enough to 

255: saturate the detector array in SL, in which case we reconstructed 

256: the SL spectrum from the off-center positions.  

257: For sources saturated in SL, we used an average of

258: the two adjacent map positions and then scaled them with a multiplicative offset

259: until they matched-up with the short-high, long-high, or long-low spectra 

260: for the same source processed as

261: described below. Because the stars in question are point sources at the spatial resolution of the 

262: IRS, the line and continuum fluxes should scale together, but this method calls 

263: into question the absolute flux calibration of the 

264: short-wavelength portions of spectra that were saturated. 

265: However this scaling does not affect {\it ratios} of spectral

266: feature integrated fluxes that we derived from SL spectra

267: for use throughout this paper.

268: Readers using these data in other ways should be aware that

269: the photometry may be more uncertain than the signal-to-noise implies. Table 1 

270: summarizes how each source was observed and clearly identifies those

271: for which SL was saturated in the central map position. We present two of our

272: sources, HD 104237 and HD 100546, in SH and LH only since their

273: 5--10~\mum\ continuum levels are above the saturation limits of the SL

274: module.

275: 

276: We started with the S14 pipeline output from the {\it Spitzer}

277: Science Center (SSC) and followed the standard calibration

278: method used at Cornell University, briefly summarized here. 

279: To remove the background emission 

280: from the SH and LH images, we subtracted images of the ``sky'' 

281: obtained by integrating on nearby offset fields.  For the SL

282: images, we subtracted images with the science target in the

283: other aperture, so that images with the target in SL order 2

284: served as sky images when the image was in SL order 1, and 

285: vice versa (i.e. aperture differences).  For LL, we 

286: differenced images with the source in the two nod positions in 

287: a given order (nod differences).  In addition to removing the 

288: background, this step also removes the majority of rogue 

289: pixels.  These pixels have temporarily high or low dark 

290: currents and are a significant source of non-gaussian noise. 

291: To correct for remaining rogue pixels and bad pixels 

292: identified in the SSC-provided mask images, we used 

293: imclean.pro\footnote{Available in IDL from the SSC as irsclean.pro.}

294: 

295: We extracted spectra from individual images using the profile,

296: ridge, and extract routines available as part of the {\it 

297: Spitzer} IRS Custom Extraction (SPICE) package.  We used 

298: standard stars to generate spectral corrections to calibrate 

299: the extracted spectra in flux-density units. For SL, HR 6348 

300: (K0 III) was the standard; for LL we used HR 6348, HD 166780 (K4 III), and 

301: HD 173511 (K5 III) as calibration standards.  For SH and LH, we 

302: used $\xi$ Dra (K2 III).  When combining the spectra from 

303: the two nod positions, we used the standard deviation at each 

304: wavelength to estimate the uncertainty.

305: 

306: The final ``stitch-and-trim'' step generates one continuous 

307: spectrum from the various segments.  We applied scalar 

308: multiplicative corrections to eliminate discontinuities, which 

309: result primarily from the fact that the target is better 

310: centered in the slit in some exposures than in others.  We scaled upwards to 

311: the best-centered aperture, which was usually LL order 2 or 

312: LH.  We combined bonus-order data in SL and LL with 

313: overlapping data in the other orders, then trimmed each spectral segment to

314: remove invalid data at the ends of the segments. Figure 1

315: presents the resulting spectra. 

316: 

317: Several sources have a choppy appearance in LH, which we

318: have called "corrugation". The effect is produced by a residual 

319: gradient signal on the IRS detector array (post flat field). The IRS team

320: is working to identify the physical source of this residual, which

321: generally appears at low light levels so usually only affects relatively faint spectra. 

322: We have removed 

323: the particularly bad corrugation in the spectrum of HD 97300 by fitting 

324: the residual signal between orders and subtracting it prior to extracting

325: the spectrum. The rest of the spectra 

326: are not as profoundly affected

327: and we have chosen to leave them alone since the corrugation effect does not

328: appear in the SL or SH modules where the PAH features that we are

329: studying appear.

330: 

331: \subsection{Overall Spectral Properties and Classification}

332: 

333: Figure 1 presents all of our spectra on the same axes with the flux scaled

334: arbitrarily to allow easy comparison. Figure 2 presents the data again with

335: each spectrum flux-calibrated and plotted on its own axes, $\lambda$ vs. F$_{\nu}$.

336: Blackbody emission from the central star and thermal emission from the circumstellar

337: disk dominate the broad-band shapes of the spectra while the detailed

338: structure comes from combinations of spectral features from PAHs and silicate dust grains.

339: In Figure 1 we have organized the spectra in groups according to the classification

340: scheme of \cite{kra02}: `SE' indicates the presence of silicate emission and `U' indicates

341: the presence of PAH emission (historically Unidentified InfraRed emission, 

342: hence the U). A lower-case

343: `u' indicates that the PAH emission feature contrast, relative to the continuum as measured

344: qualitatively by eye, is relatively weak.

345: 

346: The presence of high-contrast PAH emission in most of our sources

347: is striking particularly in the seven that have no silicate feature in the

348: 10~\mum\ region (see Figure 1). The PAH feature centered at 7.7~\mum\ 

349: consists of two primary components, one

350: at 7.65~\mum\ and the other at 7.85~\mum\ \citep{coh89, 

351: bre89, pee02}. The 7.65~\mum\ feature tends to dominate the PAH spectra of reflection

352: nebulae and H II regions where the PAHs seem to have been heavily

353: processed. The 7.85~\mum\ feature dominates in many planetary nebulae

354: and objects evolving away from the asymptotic giant branch (AGB) where

355: the PAHs seem to be relatively fresh and unprocessed. We use

356: the PAH spectral classifications of \cite{pee02} who

357: studied 57 spectra of PAH sources obtained with the \iso\ SWS. 

358: Of those spectra, 42 belong to Class A,

359: which shows the PAH emission features at 6.2 and 7.7~\mum. 

360: They found that in spectra where the 7.85~\mum\ component dominated,

361: the 6.2~\mum\ PAH feature was shifted to 6.3~\mum. They used

362: these two characteristics as the basis for what they called Class

363: B PAH spectra, which are characteristic for 12 of their objects. Class C, 

364: where the peak emission in the

365: 7--9~\mum\ range is shifted to the red of 8.0~\mum, accounts

366: for only 2 sources. Throughout the rest of this paper we refer to this feature complex as the 

367: 7.7--8.2~\mum\ feature. One source in their sample had characteristics of both

368: Class A and B. We have applied this classification system to our larger HAeBe sample,

369: and we see a strong tendency for PAH emission from HAeBe to

370: to lie between Class B and C (Table 3). 

371: 

372: We have detected a strong 17~\mum\ feature in HD 97300. \cite{bei96} first 

373: identified the 17~\mum\ feature in \iso\ SWS 

374: spectra of Galactic PNe, attributing the feature to a PAH

375: in-plane deformation mode. Recently \cite{smi07} reported this feature

376: in \textit{Spitzer} IRS spectra of galaxies in the SINGS survey. They 

377: noted a strong correlation between the strengths of the 17~\mum\ and 11.3~\mum\ features,

378: concluding that the 17~\mum\ feature is associated with

379: PAHs in galaxies. We cannot verify such a correlation in our sample, since we have many

380: sources that excite the 11.3~\mum\ PAH features, but only one that excites the

381: 17~\mum\ feature.  PAH emission observed in

382: other galaxies is dominated by interstellar photon-dominated regions, which

383: produce class A PAH spectra \citep{pee02}, while in our sample, HD 97300 is the only

384: Class A PAH spectrum.

385: 

386: In addition to the PAH spectral classification, it is useful to look for trends in the overall

387: shapes of the SEDs and the solid state features that are often prominent in HAeBe spectra.

388: \cite{mee01} and \cite{aa04} introduced a classification of the overall

389: shape of their \iso\ mid-IR SEDs for 46 HAeBe stars, including several

390: of those presented here. According to their classification, Group I SEDs

391: rise towards longer wavelengths when plotted $\lambda$ vs. $\lambda F_{\lambda}$. They suggested 

392: that Group I disks 

393: have a larger abundance of small dust grains heated in a radially flared disk. Group II

394: systems have bluer SEDs. Their conclusion is that Group II disks are probably more flattened due to 

395: settling of dust to the disk mid-plane. \cite{mee01} append the letter 'a' to indicate

396: the presence of solid state bands, most prominently the 10~\mum\ silicate

397: feature, or 'b' to indicate no solid state bands. We have included Meeus et al. group 

398: classifications with the individual spectra in Figure 2, and we have applied the Meeus et al. criterion

399: for stars in our sample that do not appear in their sample.

400: 

401: \subsection{Extraction of the PAH Emission Features}

402: 

403: We use a two-step algorithm to measure the integrated flux and

404: positions of the PAH features, following 

405: \cite{slo07}.  First, we fit a cubic spline to approximate and 

406: remove the underlying continuum from star, disk, and silicate 

407: dust grains.  This step removes the curvature under the broad 

408: PAH emission complex extending from 7.5 to 9.0~\mum.  We use two sets of 

409: spline anchor points, for spectra with and without silicate emission.  

410: Figure 3 illustrates this procedure

411: for spectra in our data set with different relative strengths of PAH and 

412: silicate emission.

413: 

414: The cubic spline fit leaves small residuals, which can result in slightly 

415: different continuum levels on the short- and long-wavelength sides of the

416: PAH features (Figure 4). When present these asymmetries

417: shift the apparent centers of the PAH features so we fit and subtract a line 

418: segment to each feature after fitting and removing the spline. 

419: Table 2 gives the 

420: wavelengths used to fit these segments.  We fit the continuum 

421: between the blue and red pairs, and integrate the feature 

422: between them to obtain our integrated fluxes.  Figure 4 illustrates this process of measuring

423: the PAH features in the spectrum of HD 135344.\footnote{SIMBAD now

424: refers to this source as HD 135344B. In any case it is still SAO

425: 206462.} For each PAH feature, we measure the integrated flux above the fitted line segment

426: and we measure the central wavelength.  

427: 

428: To determine the central wavelengths of the PAH features from the continuum-subtratcted spectra

429: we first measure the feature integrated flux, F. We then integrate the flux in each feature starting from the

430: continuum on the short-wavelength side and stopping at the wavelength corresponding to 0.5F. We

431: integrate the flux in the same feature from the continuum on the long-wavelength side of the feature and

432: stopping at the wavelength corresponding to 0.5F. We define the central wavelength as the average of 

433: these two wavelengths. The wavelength corresponding to the integrated flux median 

434: of a feature is not necessarily the same as its center wavelength 

435: because there is some uncertainty in locating the wavelengths at

436: which the flux begins to rise above the noise on either side of the feature. 

437: We estimate the uncertainties in the central 

438: wavelengths by taking the wavelengths corresponding to 

439: $F_{\nu}+\sigma_{F_{\nu}}$ and $F_{\nu}-\sigma_{F_{\nu}}$, subtracting them,

440: and dividing by 2. Here $F_{\nu}$ is the flux density at the central wavelength and

441: $\sigma_{F_{\nu}}$ is the uncertainty in $F_{\nu}$. Figure 5 presents the fully continuum-subtracted

442: spectra for all sources in our sample organized roughly in order of

443: decreasing total PAH integrated flux.

444: 

445: As \cite{slo07} demonstrated,

446: the wavelengths of PAH features shift as the temperature

447: of the radiation field varies.  The broad PAH features are 

448: generally thought to be combinations of narrow features

449: \citep[e.g.][]{bt05}, and the central wavelength is more

450: sensitive to variations between the strengths of various

451: components than the peak wavelength, which will remain at

452: the peak of the strongest feature despite variations in the

453: strengths of any secondary components. We present central

454: wavelengths in Table 3 and feature integrated fluxes in Table

455: 4. Note that we list upper limits equal to $\sigma_{F_{\nu}}$ 

456: in cases where the signal-to-noise

457: ratio is S/N$<$1.

458: 

459: For spectra with a PAH emission feature at 6.0~\mum\ measured

460: to have a S/N of two or more, we remove this feature before 

461: determining the strength and position of the 6.2~\mum\ 

462: feature.  Similarly, we remove the 8.6~\mum\ feature before 

463: measuring the 7.7--8.2~\mum\ feature.

464: 

465: The weakest continuum-subtracted PAH spectra 

466: (HD 32509, 51 Oph, HD 152404, HD 144432, and HD 31648)  are for stars with

467: strong silicate dust features at 10~\mum. Because the PAH features for

468: these sources are both intrinsically weak and superposed on a continuum

469: that is steep and varying in slope (e.g. the 10~\mum\ feature itself), we have low

470: confidence in the accuracy of our extracted PAH center wavelengths.  

471: We have therefore omitted these five sources from the feature wavelength 

472: analysis below (Figures 6-9). We use their integrated PAH flux in our comparison of PAH

473: emission to SED shape as inferred from continuum color 

474: indices (see further discussion in Section 3.2.1), 

475: since that analysis relies on a more qualitative treatment of overall PAH emission

476: and does not include PAH feature wavelengths. The five sources 

477: in question are identified with asterisks

478: in Figure 5. For HD 104237 and HD 100546, 

479: we cannot extract PAH features because we do not

480: have usable SL spectra for those two stars; both are too bright for the 

481: IRS SL module. We present their

482: SH and LH spectra in Figure 2 and their continuum-subtracted SH spectra

483: in Figure 5, but do not include them in the analysis of PAH feature positions

484: and integrated fluxes below.

485: 

486: \subsection{PAH Flux Ratios}

487: 

488: The 6.2 and 7.7--8.2~\mum\ features both arise from C$-$C 

489: bonds \citep{atb85}, which are enhanced in ionized PAHs, while the 

490: 11.3~\mum\ feature, which arises from a C$-$H bending mode, is stronger 

491: in neutral PAHs  \citep{ahs99,dra03,dra07}.  Plotting the ratio of either of the C$-$C 

492: modes over the 11.3~\mum\ integrated flux therefore indicates the degree of 

493: ionization in the PAHs.  In Paper 1 we examined 

494: these flux ratios in a sample of only four Herbig AeBe stars 

495: and found that while both ratios are supposed to measure 

496: ionization, we could not identify an obvious trend with a sample

497: of only four stars.

498: 

499: The 11.3 and 12.7~\mum\ features are both C-H out-of-plane

500: bending modes, but the 11.3~\mum\ feature arises from carbon 

501: rings with only one adjacent hydrogen (the solo mode), while

502: the 12.7~\mum\ feature arises from rings with three adjacent

503: hydrogens (the trio mode).  \cite{hon01} investigated the

504: ratio of these two features and found that it correlated with

505: the PAH ionization ratio (as measured from the ratio of the 

506: 7.7--8.2 and 11.3~\mum\ features).  They noted that PAHs with

507: long, straight edges would have more emission at 11.3~\mum\

508: than at 12.7~\mum, while PAHs whose edges had been broken up

509: by radiative processing would have higher 12.7/11.3~\mum\

510: ratios.  We found a tentative 

511: correlation in Paper 1 for our small HAeBe sample that supported the

512: conclusions  of \cite{hon01}.

513: 

514: We have plotted these integrated flux ratios for our larger sample of

515: Herbig AeBe stars in Figure 6, along with the class C PAH sources and 

516: the class A and B prototypes discussed by \cite{slo07}. The Class A prototype is

517: NGC 1333 SVS 3 (a reflection nebula), the Class B prototype is HD 44179 

518: (a post-AGB source). The Class C sources are: SU Aur (an intermediate-mass T Tau

519: star in our sample), IRAS 13416-6243 and AFGL 2688 (both post-AGB), and two red

520: giant stars HD 233517 and HD 100764.

521: The top panel plots the 6.2/11.3~\mum\ ratio vs. the 

522: 7.7--8.2/11.3~\mum\ ratio. The data show a 

523: rough increase in one flux ratio as the other increases,

524: but with substantial scatter. The variation in the relative

525: strengths of the 6.2 and 7.7--8.2~\mum\ features is enough to

526: affect estimates of the ionization ratio using only one of

527: these two bands.  Including the point for HD 141569 makes it

528: tempting to report a correlation between the two, but we

529: hesitate to draw any conclusions without data to fill the

530: gap between it and the rest of the sample. The bottom panel of 

531: Figure 6 plots the relative strength of

532: the trio and solo C$-$H modes vs. ionization ratio, F$_{12.7}$/F$_{11.3}$.  

533: A trend of increasing F$_{12.7}$/F$_{11.3}$ with increasing ionization is 

534: evident.

535: 

536: Figure 7 illustrates that the 12.7/11.3~\mum\ integrated flux ratio 

537: varies relatively smoothly with the temperature of the 

538: central source.  This figure (and the bottom panel of 

539: Figure 6) exclude a point at F$_{12.7}$/F$_{11.3}$ = 1.4 from

540: AFGL 2688, one of the class C PAH sources.  This spectrum and

541: the spectrum of IRAS 13416$-$6243 have PAH features with very

542: low contrast with respect to the strong red continuum, making

543: our measured feature sensitive to any artifacts introduced 

544: during the spline-fitting and extraction \citep[see][for more details]{slo07}.  

545: IRAS 13416 is one of the two remaining

546: outliers (at T=5440 K), for the same reason.  The remaining

547: outlier is HD 141569, which is highly ionized for its 

548: temperature (Figure 6) and appears to be more heavily processed

549: than otherwise similar objects.

550: 

551: \section{DISCUSSION} % 3

552: 

553: Since PAH emission is present in so many HAeBe stars and, when present,

554: may play a significant role in the energy balance of the circumstellar

555: material \citep{kam04}, we should ask both what the PAH emission tells us about

556: circumstellar disks orbiting intermediate-mass stars and what the

557: environment around these stars can help us learn about PAHs as

558: diagnostics in other astrophysical environments. In Paper

559: 1 we concentrated primarily on the latter analysis because of the

560: small size of our HAeBe sample at the time. In the present study we expand

561: the detailed analysis of mid-IR PAH emission from HAeBe stars that

562: we started in Paper 1, including comparisons of the spectra to preliminary

563: models of Ae star-disk systems. We find that the larger sample reveals some

564: trends that will help with both questions. A detailed discussion of the 

565: physical nature of circumstellar disks necessarily 

566: depends on accurate measurements or derivations of the physical characteristics

567: of the host stars, which we summarize in Table 5.

568: 

569: \subsection{Physical Characteristics of the PAH Molecules and Their Environments}

570: 

571: In Paper 1 we noted that the four HAeBe stars we studied had PAH spectra

572: with their 6.2~\mum\ and 7.7--8.2~\mum\ features shifted to the red \citep[e.g.][Class

573: B or Class B/C]{pee02}

574: relative to PAH spectra from HII regions. Adding more stars to the sample increases our

575: confidence in this conclusion. The Peeters et al. sample

576: included only two isolated HAeBe stars (HD 100546 and HD 179218),

577: but their classifications indicated that the degree of processing

578: of the PAH molecules is less in HAeBe disks than in

579: interstellar PDRs. That trend is much more obvious in the present HAeBe

580: sample.

581: 

582: Photo-processing of PAHs by the stellar radiation field is evident in our analysis

583: of the spectra of HAeBe stars. Figures 8 and 9 show that all of the PAH 

584: features in the 6--13~\mum\ range have their centers

585: shifted to redder wavelengths as the stellar effective temperature decreases.

586: Note that the vertical scatter in the data is real, not statistical, in the sense that earlier episodes

587: of PAH processing (e.g. prior to the current stellar radiation field) have already altered the

588: emission feature centers to varying degrees \citep{slo07}. Nevertheless a clear trend is evident in the data.

589: This trend strengthens the tentative conclusion in Paper 1 that the shifts in PAH 

590: feature wavelengths indicate the degree to which the molecules have been exposed to the

591: stellar radiation fields and therefore how processed they are (i.e. ionized, fragmented, and/or

592: dissociated) for a given star-disk system. The Peeters et al. classifications 

593: appear to be tracing the extent to which

594: PAH molecules have been processed by their environments, particularly

595: the stellar radiation field, with Class A representing the most processed and

596: Class C the least processed PAHs. Isolated HAeBe stars occupy a narrow range

597: spanning Classes B and C.

598: 

599: The HAeBe PAH spectra resemble those of post-AGB objects

600: in which the PAH molecules are relatively \textit{un}processed (i.e.

601: newly formed). In the case of HAeBe stars, since there is no clear

602: formation process for PAHs in the stars or the disks, it is more likely that the

603: PAHs appear less processed because they are newly released into the

604: stellar radiation field from some primordial reservoir. The source of 

605: PAHs may be evaporation of the icy

606: mantles on dust grains formed earlier in the proto-{\it stellar} material and

607: containing pristine PAH molecules (e.g. preserved from the pre-stellar ISM).

608: This implies that PAHs in HAeBe disks are constantly

609: being destroyed \textit{and} replenished from the reservoir by the stellar UV-optical radiation

610: field. It would also explain the strong PAH sources

611: that completely lack 10~\mum\ features commonly associated with a

612: large component of warm, relatively small silicate dust grains in disks.

613: Icy mantles on larger dust grains can evaporate, releasing PAHs 

614: in the absence of small silicate grains. We suggest that disks with

615: this "PAH no-silicate" spectrum (in our sample, HD 34282, HD 97048,

616: HD 97300, HD 100453, HD 135344, HD 141569, and HD 169142) have

617: a large inner cavities that are cleared of dust out to radii of at least a few AU, i.e. disks with inner

618: holes. All of the stars in our sample that have this distinctive spectrum, except

619: HD 97300, share an overall SED structure resembling the transitional disks recently

620: identified and studied in T Tauri systems \citep[e.g.][]{uch04,dal05,cal05,esp07,bro07}.

621: Since PAH molecules are stochastically heated by individual optical-UV photons

622: they can emit despite being located far from the star while the dust in the transitional disks is too far, therefore too cold, to emit the 10~\mum\ feature. With the exception of HD 141569, the "PAH no-silicate" systems have strong near-infrared excesses. The NIR excess in these systems may be a combination of

623: strong NIR line emission (e.g. warm molecular gas emission from the inner cavity and 

624: recombination emission from gas accretion flows) and black-body emission from small

625: bodies (e.g. very large dust particles, planetesimals, asteroids) orbiting in the inner disk region.

626: 

627: %\subsubsection{PAH Ionization Fraction and Molecule Size}

628: 

629: %Differences in PAH ionization fraction, molecular shape, and 

630: %molecular size also produce

631: %measurable differences in the PAH spectra. As detailed in Paper 1, we infer the level

632: %of PAH ionization from the $F_{7.9}/F_{11.3}$ feature strength ratio

633: %since ionized PAHs emit more strongly in the 6--9~\mum\ C-C modes than

634: %they do in the 11.3~\mum\ out-of-plane C-H bending modes.

635: %We also suggested in Paper 1 that the $F_{12.7}/F_{11.3}$

636: %ratio should increase as PAH molecule size decreases. The increasing ionization

637: %fraction should scale with  $F_{12.7}/F_{11.3}$ due photo-destruction of the PAHs

638: %as they are ionized. Figure 6 (bottom panel) plots $F_{12.7}/F_{11.3}$ vs. $F_{7.9}/F_{11.3}$.

639: %Thus, in Figure 6 the sources with more highly photo-processed PAHs are located

640: %more towards the upper right.

641: 

642: \subsection{Disk Geometry and the Locations of the Emitting PAH Molecules}

643: 

644: \cite{mee01} suggested that HAeBe stars with redder (Group I) SEDs

645: have a long-wavelength component indicating smaller dust grains

646: that are located in a relatively warm, radially flared region of the disks. Their

647: Group II contains SEDs that lack the warm, small-grain component. They interpreted 

648: Group II SEDs as indicating a more flattened disk containing mostly larger dust grains

649: that have settled to the disk mid-plane, whereas Group I SEDs represent

650: disks with a substantial radial flare. Subsequent modeling by \cite{dul03} 

651: have supported these conclusions. \cite{mee01} and \cite{aa04} also noticed a tendency

652: for the Group I sources to be among the $~50\%$ in which they detected PAH features

653: in \iso\ SWS spectra. For the sources with stronger PAH emission, our results are

654: consistent with \cite{aa04}; the strongest PAH sources are in fact all Group I. However the higher 

655: sensitivity of the {\it Spitzer} IRS has revealed that some Group I sources have weak

656: PAH emission and that many Group II sources that were tentative ISO SWS detections in only the

657: 6.2~\mum\ feature are PAH emitters in all of the mid-IR PAH features (e.g. HD 141569, HD 142666,

658: HD 35187, HD 31648, HD 144432). Thus, although it does seem that the PAH emission comes 

659: from the illuminated surfaces of flared disks, we suggest that such emission is {\it not} an 

660: unambiguous tracer of flared geometry in HAeBe disks.

661: 

662: \cite{dou07} recently detected the flared disk around HD 97048 in a mid-IR direct imaging study.

663: The flaring disk is evident in the image, and the source is more extended in the

664: 11.3~\mum\ PAH feature than in the adjacent continuum. \cite{gra05} reported that mid-IR PAH

665: emission, in particular the 6.2~\mum\ feature, correlates with optical scattered light

666: visibility of the inner disks (r$\sim$50-70 AU) of some HAeBe stars. Since

667: the PAH features and H$_{2}$ line emission are often strong where the optical-UV radiation

668: field is strong (e.g. interstellar photon-dominated regions), they should trace material 

669: that is far enough above the disk mid-plane to be directly illuminated by the star. \cite{gra05}

670: therefore suggest that both the optical visibility of the disk and the

671: mid-IR PAH emission should correlate with flaring and anti-correlate with

672: dust settling to the disk mid-planes. However, four of the STIS scattered light {\it non-detections} 

673: from \cite{gra05} are in our sample (HD 31648, HD 104237, HD 135344, HD 142666). 

674: All but HD 104237 clearly have PAH emission in our spectra. These apparently optically dark 

675: disks have optically visible stars so their

676: disk surfaces cannot appear dim to us simply because we are seeing them

677: nearly edge-on. In fact their inclination angles range from 38\arcdeg\ to 

678: 55\arcdeg\ (Table 5). Apparently disks with relatively little scattered optical light are still illuminated

679: enough to excite PAH emission, or the PAH emission comes from the inner disk and {\it not} from

680: the atmospheres of the outer disks.

681: 

682: Although it is tempting to use the PAH emission strength as a 

683: preliminary diagnostic of overall disk geometry, especially for those sources whose disks

684: are unresolved or marginally resolved in visible, near-IR, and mid-IR images, we 

685: suggest that some caution

686: is indicated. The question remains as to whether large-scale radial flaring dominates

687: or whether shadowing of the outer disk by 

688: structures in the inner disk may play a role. The former possibility implies

689: that flared disks should be strong PAH emitters unless we happen to view 

690: them edge-on. The latter possibility implies that illumination of even highly flared disks can

691: nevertheless be partially and/or temporarily attenuated by structures in the inner disk--within

692: a few AU of the star--and therefore produce weak or non-PAH emitters as well as 

693: weak or non-detections in optical/NIR scattered light observations.

694: 

695: \subsubsection{SED Shape and PAH Emission}

696: 

697: The geometry of the disks alters their SED significantly and there have been

698: many efforts to quantify this relationship for circumstellar accretion disks,

699: including those around HAeBe stars, both empirically 

700: (e.g. \cite{mee01}) and with detailed models \citep[e.g.][]{dul03,dul04,dal01,dal05,dal06}.

701: Recently \cite{fur06} and \cite{wat07} have combined these approaches

702: to study disks around low-mass young stellar objects in Taurus. They used continuum color indices 

703: derived from {\it Spitzer} IRS spectra to characterize the SED shape for comparison to 

704: grids of accretion disk models by \cite{dal06}. They defined the color 

705: indices by differencing fluxes in two narrow wavelength bins located between the prominent solid state features,

706: essentially providing a measure of the slope of the SED between the two wavelength

707: bins. The differences in flux ($\lambda F_{\lambda}$) from one bin to the next (e.g. from a to b), 

708: expressed as ratios of their base-ten logarithms, form the color index, $n_{a-b}$:

709: 

710: %\begin{center}

711: %\begin{equation}

712: 

713: $n_{a--b}=\log(\lambda_{a} F_{\lambda a}/\lambda_{b} F_{\lambda b})/ \log(\lambda_{a}/\lambda_{b})$.

714: 

715: 

716: %\end{equation}

717: %\end{ceneter}

718: 

719: We have chosen spectral bins for our indices that avoid both the solid state features

720: and the PAH features that are common in HAeBe spectra: 

721: 5.4--6.0~\mum, 13.0--14.0~\mum, and 29.75--31.75~\mum,

722: labeled 6, 13, and 30~\mum\ respectively. Following \cite{fur06}, we have analyzed

723: our sample stars using $n_{6--13}$ vs. $n_{13--30}$ color-color diagrams.

724: 

725: \cite{dal06} found that model star-disk systems changed location on such color-color 

726: diagrams extracted from grids of model SEDs in which degrees of dust settling 

727: and disk inclination were allowed to vary. 

728: \cite{fur06} and \cite{wat07} found that plotting their samples of low-mass YSOs 

729: on the same color-color diagrams indicated a wide range of dust grain sizes, which

730: they interpreted as degrees of dust settling to the disk mid-plane. D'Alessio and

731: collaborators have begun generating similar models for Herbig Ae stars and we can

732: now begin a similar analysis for intermediate-mass star-disk systems. Since these models

733: are still in a preliminary stage of development, in this paper we will use them only to infer schematic trends

734: in disk physical characteristics indicated in continuum color-color diagrams. More detailed 

735: predictions and estimates of disk properties for individual HAeBe 

736: systems will be the focus of a forthcoming study.

737: 

738: \subsubsection{Model Physics}

739: 

740: The fundamental physics of the disk model we use is described in \cite{dal98,dal99}. 

741: We assume a steady

742: state, axisymmetric and geometrically thin disk. Its mass accretion rate

743: ($\dot{M}$) is taken to be uniform (spatially constant). We decouple

744: the equations for the disk vertical and radial structure since

745: the disk is geometrically thin. In the vertical direction the disk is in

746: hydrostatic equilibrium with the energy being transported by

747: radiation, turbulent conduction, and convection; the latter mechanism is

748: active only where the convective instability criterion is

749: satisfied. In the radial direction, the disk is assumed to be in

750: centrifugal balance in the potential well of the star, i.e., a

751: Keplerian disk. The primary disk heating sources are viscous

752: dissipation, with the viscosity coefficient given by the $\alpha$

753: prescription \citep{sha73}, and stellar irradiation. In

754: order to calculate the transfer of disk and stellar radiation, we use

755: mean opacities in which the average of the monochromatic

756: opacity is calculated using, as weighting

757: functions, either the Planck function (or its derivative with respect to

758: temperature) evaluated at the local kinetic temperature or the star

759: radiation temperature, respectively.  For simulating the effect of

760: dust settling, we assume that the disk has two populations of dust grains,

761: with different grain-size and spatial distributions. In the disk atmosphere

762: there are small grains, depleted with respect to the standard dust to

763: gas mass ratio \citep{dal06}. Closer to the mid-plane,

764: there are larger dust grains, with an enhanced dust to gas mass ratio,

765: which accounts for the mass in grains that have disappeared from the

766: upper layers.  For simplicity, we assume that the settling is only in the

767: vertical direction.  The boundary conditions for the transfer

768: equations are given in the "irradiation" surface, defined as the

769: surface where the mean radial optical depth to the stellar radiation

770: is unity. The larger the depletion in the upper layers, the smaller

771: the height of the irradiation surface, i.e., the smaller the degree of

772: flaring of the disk. The flatter the disk, the smaller the fraction of stellar

773: radiative flux intercepted by its surface.

774: 

775: The disk inner boundary is assumed to be the radius where optically

776: thin dust, heated by the star, reaches its sublimation

777: temperature. Assuming that silicates comprise the dust component with

778: largest abundance, sublimated at the highest

779: temperature, we adopt a dust sublimation temperature of 1400 K.  The disk inner wall

780: model is described by \cite{muz03} and \cite{dal05}. It is assumed to 

781: be a plane-parallel atmosphere, with

782: constant temperature in the vertical direction, heated by the

783: star in the radial direction. The  temperature of this atmosphere

784: is calculated as a function of the  mean optical depth in a direction parallel

785: to the disk mid-plane (radial, in cylindrical coordinates).

786: The vertical height of the wall is roughly estimated as the height

787: where the mean radial optical depth to the stellar radiation is unity,

788: which depends on the opacity and density at the wall. We assume the

789: wall is in vertical hydrostatic equilibrium with gas  in LTE and

790: the same dust properties as in

791: the disk atmosphere. The SED of the wall and the SED of the disk are calculated 

792: separately.

793: 

794: All of our disk models are flared to varying degrees. 

795: Being 1+1D (i.e., with the vertical and radial structure

796: calculated separately), these models do not account for the possibility

797: of shadowing. However, the scale height of the wall and the scale

798: height of the flared disk at the same radius are very similar, mostly

799: because the mid-plane of the disk is heated by viscous dissipation,

800: which is not included in our wall model. This heating mechanism seems

801: to compensate for the difference in the incident stellar flux between

802: the wall and the disk. Thus, our disk models are not puffed-up in the

803: way described by \cite{dul01}.  When the wall has dust

804: settled towards its mid-plane, its effective height (i.e.,

805: the size of its emitting area) is smaller than the the wall with no

806: settling. Thus, the contribution to the SED of the former is

807: smaller than the contribution of the later. The SEDs of both kinds of

808: models are added to the same SED of the rest of the disk, from which

809: the contribution to the emission of its inner boundary has been

810: removed.

811: 

812: \subsubsection{The distribution of PAH emission with color index}

813: 

814: In Figure 10 we present n$_{6-13}$ vs. n$_{13-30}$

815: color-color diagrams derived from a grid of model disk systems for stellar spectral types

816: A2 and A6, as well as varying inclination angle (40\arcdeg - 70\arcdeg), mass accretion rates 

817: ($10^{-8}$ and $10^{-9} M_{\sun}/yr$), age (1 Myr - 10 Myr), and 

818: degree of atmospheric dust grain depletion ($\epsilon$=0.001 - 1.0). The depletion factor, $\epsilon$, 

819: varies from 1.0 to 0.001 and represents the dust-to-gas mass ratio for disk atmospheric silicate

820: dust grains relative to the solar dust-to-gas mass ratio \citep{dal06}. Thus $\epsilon$ 

821: is a measure of the depletion of atmospheric grains as they settle to the disk midplane. 

822: Decreasing $\epsilon$ is an indicator

823: of overall disk settling from flared to more flattened geometry. The more flattened 

824: disks will tend to occupy

825: the bluer (lower left) region of the color-color diagram in Figure 10, while more flared

826: disks will be redder and located towards the upper right. Disk inclination increases right to

827: left while stellar age increases from lower right to upper left. 

828: 

829: Figure 10a contains models for mass accretion rates of $10^{-8} M_{\sun}/yr$ (top panel)

830: and $10^{-9} M_{\sun}/yr$ (bottom panel)

831: in which the dust in the inner wall of the disk has settled. Figure 10b, with the same 

832: accretion rates in the top and bottom panels,

833: contains models for which dust in the inner wall has not settled. The inner wall is located at the inner

834: radius of the disk.

835: We have over-plotted our observed color indices for comparison with the models.

836: The models and observed colors overlap, but the observed HAeBe stars 

837: cover a larger range of parameter space than the current grid of models.

838: 

839: We plot our star sample on an identical color-color diagram in Figure 11 with the

840: PAH flux (Figures 11a and 11b) indicated in the figure by the size of the hexagonal symbols that are

841: centered on the star colors. We calculated the total PAH flux for each star by summing the integrated

842: fluxes of the 6.2, 7.7--8.2~\mum\, 11.3, and 12.7~\mum\ features and dividing by the square of the

843: stellar distance normalized to 100 pc so that the symbol sizes in Figure 11a represent the

844: intrinsic total PAH emission for each star. Because the 11.3~\mum\ PAH feature coincides with the

845: 11.1~\mum\ feature of crystalline silicates (olivine) and since both can be strong in 

846: HAeBe stars (e.g HD 144432, Figure 2), we have also computed luminosities for the 6.2~\mum\

847: feature alone (Figure 11b). The 6.2~\mum\ feature is not near other

848: PAH or silicate features so it serves as an independent indicator of PAH emission strength.

849: A comparison of Figures 11a and 11b shows that they are not identical, but the 

850: feature that we extracted at 11.3~\mum\ seems to be dominated by the PAH emission. Note that our continuous color indices are consistent with the

851: discrete Group designations of \cite{mee01}. Group I disks are flared and Group II disks are more flattened.

852: Nevertheless, Group I  disks have a wide range of PAH integrated fluxes.

853: 

854: \subsubsection{Notes on HAeBe stars that do not overlap the models}

855: 

856: In our analysis of color indices in the previous section we noted that our data have a larger

857: range of colors than the present grid of models. Twelve stars have colors outside the range

858: of the model grid: HD 32509, HD 34282, HD 281789, HD 152404, HD 169142,

859: HD 97300, HD 135344, HD 139619, HD 141569, HD 169142, 51 Oph, and Elias 3-1. Two of these, 51 Oph and 

860: HD 97300 are probably not HAeBe stars. The remaining

861: ten stars include HD 32509, which has NO measurable PAH emission, and five Group I stars with relatively weak PAH emission. Furthermore, {\it all} of the seven sources in our sample that have 

862: PAH emission, but no 10~\mum\ silicate emission feature, are in this "outlier" sub-sample. 

863: 

864: HD 141569 is a more evolved system with substantial clearing of material in its disk \citep{mer04,

865: aug99, wei99} . If the inner disk structure allows stellar radiation farther into the disk where PAHs

866: may be released from icy mantles on large grains, this could explain why HD 141569 is by far the most ionized PAH 

867: spectrum in our sample. HD 169142 appears to be in an epoch of central clearing of the disk and may resemble 

868: Vega in an earlier epoch of its disk evolution \citep{gra07}. The disk orbiting HD 100453  seems to be dominated 

869: by large dust grains \citep{vdb04}.

870: 

871: HD 97300 is a B9 star located in the Chamelion I cloud that is close to the zero-age main sequence and has no near-IR excess emission or H$\alpha$ emission \citep{sie00}. \cite{sie98} report that the PAH 

872: spectrum is a mixture of a small amount of emission from a circumstellar disk and a large amount from the very 

873: bright, ring-shaped reflection nebula that HD 97300 illuminates. Although HD 97300 has been included in past 

874: studies of Herbig AeBe stars, we suggest that its Herbig Be classification may be in error since it is the only Class A PAH spectrum in our sample and the only source in our sample with emission in the 17~\mum\ feature. Certainly the fact that it lacks both H$\alpha$ emission and any near-IR excess suggests  a more evolved system that has ceased active accretion. The PAH spectrum of HD 97300 is probably dominated by 

875: interstellar PAHs rather than PAHs in a circumstellar disk.

876: 

877: \subsection{What does PAH emission tell us about HAeBe disks?}

878: 

879: At first glance a weak correlation of PAH luminosity with color index is evident in Figure 11a.

880: Although there are no strong HAeBe PAH emitters that have colors indicating highly settled disks, 

881: there are six relatively weak PAH sources that have colors indicating that they should have flared 

882: disks: HD 34282, HD 100453, HD 35344, HD 139614, HD 141569, and SU Aur. The Meeus Group classifications and the spectral indices are consistent in their

883: segregation of the stars in Figure 11, which we illustrate schematically with a dashed line

884: that roughly separates the Group I and Group II sources in our sample. We use Figure 12 to illustrate the

885: frequency of Groups I and II as a function of the n$_{13-30}$ spectral index (the vertical axis

886: in Figure 11). The Meeus Groups form a bimodal distribution in our sample that is consistent

887: with the interpretation that those Groups distinguish between flared and flattened (or non-settled and

888: settled) disk geometries.

889: Nevertheless, we see overlap in Figures 11 and 12; there are relatively weak PAH emitters that should

890: have radially flared disk geometries according to both discrete (Meeus Group) and continuous

891: (spectral index) SED classifications. Why do some apparently flared disks have weak PAH

892: emission?

893: 

894: We suggest several possible explanations for the absence of strong PAH emission in 

895: radially flared disks, which should have a large surface area directly illuminated by the star: (1) the 

896: abundance of PAHs in the disk atmospheres of those systems (either free-floating or locked-up in icy dust grain mantles) is intrinsically low; (2) PAHs are abundant in the disk atmospheres, but a strong stellar UV flux has destroyed most of the PAHs in the emitting region of the disk; (3) we are viewing those 

897: systems edge-on or nearly so; or (4) the inner walls of the disks are attenuating or softening the stellar illumination of a flared outer disk as suggested by \cite{aa04} and explored in detail more recently by \cite{ise06}. We now explore these hypotheses in turn.

898: 

899: {\it PAH Abundance.} If, as we suggested earlier, most of the emitting PAHs in HAeBe disks are released from icy mantles and if the PAHs are then destroyed by

900: the photo-processing that makes them emit, it is possible that the icy reservoir might eventually

901: be exhausted. In this scenario the radially flared disks with relatively weak PAH emission may simply have

902: evaporated their PAH reservoirs more thoroughly than the stronger PAH emitters. With the exception of

903: SU Aur, an intermediate-mass T Tauri star, the flared disks with weak PAH emission all have SEDs that

904: resemble transitional disks. They may be the most evolved disk systems in our sample.

905: 

906: {\it Destruction of PAHs.} We cannot blame a strong UV field destroying PAHs since some of the hottest stars in our sample excite very strong PAH emission (e.g. Elias 3-1, AB Aur, \& HD 97300).

907: We illustrate this in Figure 11c, where  the symbol sizes are proportional to the total PAH flux 

908: density ($\Sigma F_{PAH}$) divided by the stellar bolometric luminosity ($L_*$) for each source. In Figure 

909: 11d, symbol sizes are proportional only to the stellar bolometric luminosity. We calculated 

910: bolometric luminosities using published values of the apparent visual magnitudes, distances, and 

911: extinctions for our sample stars (Table 5). Bolometric corrections are for luminosity class

912: V \citep{kh95}. There is no indication in Figures 11c-d that stellar luminosity correlates inversely

913: with PAH intensity. There is some indication that the opposite is true, but not uniformly in our 

914: sample (Figure 11d). Finally, as a sub-sample the "weak" PAH Group I sources are not uniformly the most 

915: processed nor are they all the most ionized sources in our sample. We therefore conclude that 

916: differences in the stellar radiation field from system to system alone cannot account for the variations in PAH emission.

917: 

918: {\it Disk Inclination.} For sources with measured inclinations we see no indication that

919: more inclined disks (large $i$, see Table 5) have weaker PAH emission. None of the stars have silicate 

920: absorption features at 10~\mum\ and those with

921: the weakest silicate emission have very strong contrast in the PAH features.

922: Thus inclination of the disks along our line of sight cannot  explain the variation in PAH luminosity among 

923: the redder disks. Table 5 also lists measured extinctions towards most of our sample stars, which

924: show no trend towards decreased PAH emission for more highly extinguished sources. In fact, none of

925: our sources, not even Elias 3-1 with A$_v$=4.05, has enough extinction to significantly alter its mid-IR SED. 

926: The bottom line is that there are highly flared disks nearly facing us that are nevertheless weak PAH 

927: emitters. The models summarized in Figure 10 indicate that inclination should affect

928: the mid-IR colors of our sources and though we see a range of colors, we see no trend for PAH

929: emission to weaken as colors change due to disk inclination (e.g. from lower right to upper left in Figures

930: 10a and 10b).

931: 

932: {\it Attenuated Radiation or Shadowing by the Inner Disk.} The spatial distribution of PAH molecules in proto-planetary disk systems is a subject of current discussion. Since there is no clear formation path for PAHs in proto-planetary disks, the PAHs are likely primordial in the sense that they were present in the ISM 

933: before the star and disk formed \citep{ll03}.  \cite{aa04} concluded that the PAH emission must originate 

934: from the denser environment of a disk rather than from a large diffuse halo or envelope around the 

935: star. More recently \cite{vis07} have modeled the chemistry and infrared emission from

936: PAHs around HAeBe stars. After comparing their models to observational studies of

937: extended PAH emission around HAeBe stars \citep[e.g.][]{gee05,gee06,hab05,vdb04} they conclude

938: that PAH molecules with sizes of $\sim$100 carbon atoms are responsible for most of the 

939: observed emission. PAH molecules that big can survive the stellar radiation fields

940: of intermediate-mass stars down to a minimum radius of $\sim$5 AU. Thus, \cite{vis07}

941: predict that about 60\% of the PAH emission in the 6--13~\mum\ range originates within

942: 100 AU of the star.  

943: 

944: For disks that are no longer accreting material onto their stars, a puffed-up inner wall in the flared Group I disks 

945: may attenuate or soften the stellar radiation field illuminating some or all of the outer disk.

946: This may be a relatively short-lived stage in what appear to be 

947: disks in transition from flared to flattened. If planet formation clears an inner hole, then the gas in the inner wall could be vertically inflated relatively quickly (and to varying degrees depending

948: on the system) by photoelectrons ejected by UV photons from PAHs and small dust grains. Thus for a short time, prior to the cooling and settling of the outer disk, there can be both a high inner wall and a

949: radially flared outer disk. As the grain growth and disk settling proceed, the flare of the disk may decrease 

950: until the PAH-rich parts of the outer disk settle into the shadow of the inner disk wall. Thus redder disks that 

951: are weak PAH emitters may be in an intermediate or transitional stage of disk evolution. We note that this suggested

952: scenario is beyond the scope of our current grid of disk models to predict since those models contain no treatment of shadowing in the disks.

953: 

954: \section{SUMMARY AND CONCLUSIONS}  % 4

955: 

956: We have presented \textit{Spitzer} IRS spectroscopy of a sample

957: of 22 targets including 18 Herbig Ae/Be stars and two intermediate-mass T Tauri stars

958: whose spectra show strong PAH emission. The

959: absence of silicate emission in seven of the sources has enabled a

960: detailed analysis---independent of mineralogy estimates or specific

961: disk models--of the relative strengths and positions of PAH emission

962: features in the 5--14~\mum\ region. For the remaining stars

963: that do have solid state emission features, we have run the same PAH

964: spectral analysis by fitting and subtracting dust continuum and dust

965: emission features with a model-independent method, isolating the 

966: PAH spectrum. This relatively

967: large sample and the detailed analysis afforded by high signal-to-noise

968: IRS spectra led us to the following conclusions:

969: 

970: \begin{itemize}

971: 

972: \item PAHs exposed to stronger

973: stellar radiation fields may become shattered and develop more jagged

974: edges so that the 11.3~\mum\ mono C-H bending mode is attenuated with respect

975: to the 12.7~\mum\ trio C--H bending mode. PAH ionization fraction, inferred

976: from the PAH F$_{7.7-8.2}$/F$_{11.3}$ ratio, is related to the structure (and

977: possibly the size) of the PAH molecules. 

978: 

979: \item The center wavelengths of the 6.2, 7.7--8.2, 8.6, and 11.3~\mum\ emission features 

980: change as the stellar effective temperature changes. We confirm 

981: our previous finding that the center wavelengths of the

982: PAH emission features increase with decreasing effective

983: temperature of the host star, now with a much larger sample of stars. We conclude

984: that hotter stars photo-process the PAHs more than cooler stars do and that

985: the degree of processing alters the PAH spectrum measurably.

986: 

987: \item Our sample shows spectral characteristics generally consistent with

988: the boundary between Class B or C PAH emission sources defined by \cite{pee02}, viz.

989: a 6.2~\mum\ feature shifted to 6.3~\mum\ and 7.7--8.2~\mum\ emission

990: dominated by the component at 7.9~\mum. However, in some of our

991: sources, the 7.7--8.2~\mum\ feature center is shifted closer to 8.0~\mum, as

992: in the spectrum of HD 100546, a HAeBe star observed by the SWS on

993: \iso\ and included in our sample. It is remarkable that HAeBe stars

994: as a class have such a variety of PAH luminosities and ionization

995: fractions, but such a narrow range of PAH spectral classifications. This may

996: indicate that the PAH molecules are altered by the same physical processes

997: in the disks, perhaps unique to HAeBe stars.

998: In any case we propose that the HAeBe stars comprise a distinct PAH spectral

999: group of their own. HAeBe PAH spectra indicate very little processing of the

1000: PAH molecules by the stellar radiation field, relative to other types of

1001: PAH source (e.g. HII regions, PDRs, and PNe).

1002: 

1003: \item The other Peeters Class B objects are post-AGB and

1004: that implies that the PAH are less processed and therefore relatively new to the scene in 

1005: Herbig Ae/Be disks. This is consistent with the suggestion that the PAHs are introduced

1006: at a stage of disk evolution when icy mantles evaporate from large dust grain

1007: surfaces. If this is the case, then the PAHs near HAeBe stars are being

1008: constantly destroyed and replenished by interactions with the stellar radiation

1009: field since we do not observe HAeBe systems with the (apparently more processed) 

1010: Class A spectra.

1011: 

1012: \item We have detected the 17~\mum\ PAH complex in one of our sample

1013: HAeBe stars, HD 97300, which stands apart form the rest of our sample as the

1014: only class A (e.g. highly photo-processed) PAH spectrum and is likely not a Herbig

1015: Ae/Be star. The 17~\mum\ feature 

1016: is commonly observed in the Galactic ISM and in other galaxies where strong UV fields excite extensive 

1017: PDRs. HD 97300 is also a Meeus Group IIb source having no warm small grain component in its 

1018: SED and no silicate features at 10~\mum\ or longward of 20~\mum. We suggest that, whatever

1019: the classification of HD 97300, it's spectrum is dominated by PAHs in the surrounding interstellar

1020: gas and not in a circumstellar disk.

1021: 

1022: \item Our sample includes both flared (Meeus et al. 2001, Group I) and flattened/settled (Group

1023: II) systems, but the overall PAH emission does not decrease 

1024: monotonically with indicators of depletion of small dust grains and disk settling, as

1025: one would expect if the PAH emission only originates in an optically

1026: and geometrically thin surface layer of a radially flared disk.

1027: 

1028: \item Taking into account estimates of

1029: disk inclination for the sources in our sample does not

1030: uncover a direct correlation of weak PAH emission with indicators of nearly edge-on disk

1031: geometry. Nor can we explain the anomaly by invoking strong stellar radiation fields

1032: to destroy or process the PAHs so that they emit less light in some disks.

1033: 

1034: \item Differences in PAH abundance in the illuminated parts of HAeBe disks and/or evolution of the inner

1035: disk scale height appear to be possibilities for explaining highly flared disks that have relatively

1036: weak PAH emission. These sources may be intermediate-mass versions of the transitional T Tauri

1037: disks.

1038: 

1039: \end{itemize}

1040: 

1041: A larger sample of HAeBe stars will allow us to better determine what a "normal" HAeBe star

1042: looks like. We have begun the next step, a thorough modeling effort including

1043: the addition of detailed PAH spectral synthesis and careful application of

1044: mineralogical models for an even larger sample of stars.

1045: 

1046: \acknowledgments

1047: 

1048: We thank the anonymous referee for comments and suggestions that significantly

1049: improved the manuscript. Many thanks to Nirbhik

1050: Chitrakar and Jordan Hyatt for assistance with data analysis and graphics.

1051: L. Keller is grateful to Research Corporation for support of this work. A. Li is supported in part 

1052: by grants from the {\it Spitzer} and HST Theory Programs. Support for this

1053: work was also provided by NASA through contract number 1257184 issued

1054: by JPL/Caltech. This work is based on observations made with the \textit{Spitzer Space

1055: Telescope}, which is operated by the Jet Propulsion Laboratory, California

1056: Institute of Technology under NASA contract 1407. This research has made use of the SIMBAD database

1057: operated at the Centre de Donn\'{e}es astronomiques de Strasbourg

1058: and the NASA Astronomical Data System (ADS).

1059: 

1060: \begin{thebibliography}

1061: 

1062: \bibitem[Acke \& van den Ancker(2004)]{aa04} Acke, B. \& van den

1063: Ancker, M.~E. 2004, \aap, 426, 151 

1064: 

1065: \bibitem[Allamandola et al.(1999)]{ahs99} Allamandola, L.~J., Hudgins,

1066: D.~M., \& Sandford, S.~A. 1999, \apj, 511, L115 

1067: 

1068: \bibitem[Allamandola et al.(1985)]{atb85} Allamandola, L.~J., Tielens,

1069: A.~G.~G.~M., \& Barker, J.~R. 1985, \apj, 290, L25 

1070: 

1071: %\bibitem[Allamandola et al.(1989)]{atb89} Allamandola, L.~J., Tielens,

1072: %A.~G.~G.~M., \& Barker, J.~R. 1989, \apjs, 71, 733 

1073: 

1074: \bibitem[Augereau et al.(1999)]{aug99} Augereau, J.~C., Lagrange,

1075: A.~M., Mouillet, \& Menard, F. 1999a, \aap, 350, L51

1076: 

1077: \bibitem[Berthoud et al.(2007)]{ber07} Berthoud, M.~G., 

1078: Keller, L.~D., Herter, T.~L., Richter, M.~J., \& Whelan, D.~G.\ 2007, \apj, 

1079: 660, 461

1080: 

1081: \bibitem[Beintima et al.(1996)]{bei96} Beintima et al. 1996, \aap,

1082: 315, L369 

1083: 

1084: %\bibitem[Boccatelli et al.(2003)]{boc03} Boccatelli, A., Augereau,

1085: %J.~C., Marchis, F., \& Hahn, J. 2003, \apj, 585, 494 

1086: 

1087: %\bibitem[Boissel et al.(2001)]{bjp01} Boissel, P., Joblin, C., \&

1088: %Pernot, P., 2001, \aap, 373, L5 

1089: 

1090: %\bibitem[Bouwman et al.(2001)]{bou01} Bouwman, J., Meeus, G., de

1091: %Koter, A., Hony, S., Dominik, C., \& Waters, L.~B.~F.~M. 

1092: 

1093: \bibitem[Bregman(1989)]{bre89} Bregman, J.~D. 1989, in Interstellar

1094: Dust, Proc. IAU Symp 135, ed. L.~J. Allamandola \& A.~G.~G.~M.

1095: Tielens, 109 

1096: 

1097: %\bibitem[Bregman et al.(1989)]{bat89} Bregman, J.~D., Allamandola,

1098: %L.~J., Tielens, A.~G.~G.~M., Geballe, T., \& Witteborn, F.~C.

1099: %1989, \apj, 344, 791 

1100: 

1101: \bibitem[Bregman \& Temi(2005)]{bt05} Bregman, J.~D. \& Temi, P.

1102: 2005, \apj, 621, 831

1103: 

1104: \bibitem[Brittain et al.(2007)]{bri07} Brittain, S.~D., 

1105: Simon, T., Najita, J.~R., \& Rettig, T.~W.\ 2007, \apj, 659, 685

1106: 

1107: \bibitem[Brown et al.(2007)]{bro07} Brown, J.~M., et al.\ 

1108: 2007, \apjl, 664, L107

1109: 

1110: \bibitem[Calvet et al. (2004)]{cal04} Calvet, N., Muzerolle, J.; Brice\~{n}o, C., Hern\'{a}ndez, J., 

1111:   Hartmann, L., Saucedo, J.~L., \& Gordon, K.~D.  2004, \aj, 128, 1294

1112:   

1113: \bibitem[Calvet et al.(2005)]{cal05} Calvet, N., et al.\ 

1114: 2005, \apjl, 630, L185

1115: 

1116: %\bibitem[Chiar et al.(2000)]{chi00} Chiar, J.~E., Tielens, A.~G.~G.~M.,

1117: %Whittet, D.~C.~B., Schutte, W.~A., Boogert, A.~C.~A., Lutz, D.,

1118: %van Dishoeck, E.~F., \& Bernstein, M.~P. 2000, \apj, 537, 749 

1119: 

1120: %\bibitem[Cohen et al.(2003)]{coh03} Cohen, M., Megeath, T.~G., Hammersley,

1121: %P.~L., Martin-Luis, F., \& Stauffer, J. 2003, \aj, 125, 2645 

1122: 

1123: \bibitem[Cohen et al.(1989)]{coh89} Cohen, M., Tielens, A.~G.~G.~M.,

1124: Bregman, J., Witteborn, F.~C., Rank, D.~M., Allamandola, L.~J.,

1125: Wooden, D., Jourdain de Muizon, M. 1989, \apj, 341, 246 

1126: 

1127: %\bibitem[Coulson \& Walther(1995)]{cw95} Coulson, I.~M. \& Walther,

1128: %D.~M. 1995, \mnras, 274, 977

1129: 

1130: \bibitem[D'Alessio et al.(1998)]{dal98} D'Alessio, P., Canto, 

1131: J., Calvet, N., \& Lizano, S.\ 1998, \apj, 500, 411

1132: 

1133: \bibitem[D'Alessio et al.(1999)]{dal99} D'Alessio, P., 

1134: Calvet, N., Hartmann, L., Lizano, S., \& Cant{\'o}, J.\ 1999, \apj, 527, 893

1135: 

1136: \bibitem[D'Alessio et al.(2001)]{dal01} D'Alessio, P., Calvet, N., \& Hartmann, 

1137: L.\ 2001, \apj, 553, 321 

1138: 

1139: \bibitem[D'Alessio et al.(2005)]{dal05} D'Alessio, P., 

1140: Calvet, N., \& Woolum, D.~S.\ 2005, Chondrites and the Protoplanetary Disk, 

1141: 341, 353

1142: 

1143: \bibitem[D'Alessio et al.(2006)]{dal06} D'Alessio, P., 

1144: Calvet, N., Hartmann, L., Franco-Hern{\'a}ndez, R., \& Serv{\'{\i}}n, H.\ 

1145: 2006, \apj, 638, 314

1146: 

1147: \bibitem[Dominik et al.(2003)]{dul03} Dominik, C., Dullemond, 

1148: C.~P., Waters, L.~B.~F.~M., \& Walch, S.\ 2003, \aap, 398, 607

1149: 

1150: \bibitem[Doucet et al. (2007)]{dou07} Doucet, C., Habart, E., 

1151: Pantin, E., Dullemond, C., Lagage, P.~O., Pinte, C., Duch{\^e}ne, G., \& 

1152: M{\'e}nard, F.\ 2007, \aap, 470, 625

1153: 

1154: \bibitem[Dullemond et al.(2001)]{dul01} Dullemond, C.~P., 

1155: Dominik, C., \& Natta, A.\ 2001, \apj, 560, 957

1156: 

1157: %\bibitem[Dullemond(2002)]{dul02} Dullemond, C.~P.\ 2002, 

1158: %\aap, 395, 853

1159: 

1160: \bibitem[Dullemond \& Dominik (2004)]{dul04} Dullemond \& Dominik 2004, \aap, 417, 159

1161: 

1162: \bibitem[Dullemond et al.(2007)]{dul07} Dullemond, C.~P., Henning, T., Visser, R., Geers, V.~C., van Dishoeck, E.~F., \& Pontoppidan, K.~M.\ 2007, \aap, 473, 457 

1163: 

1164: \bibitem[Draine (2003)]{dra03} Draine, B.~T. 2003, \araa, 41, 241

1165: 

1166: %\bibitem[Draine \& Li(2001)]{dl01} Draine, B.~T. \& Li, A. 2001,

1167: %\apj, 551, 807 

1168: 

1169: \bibitem[Draine \& Li(2007)]{dra07} Draine, B.~T. \& Li, A. 2007,

1170: \apj, 657, 810 

1171: 

1172: \bibitem[Dunkin et al. (1997)]{dmr97} Dunkin, S.~K., Barlow, M.~J.,

1173: \& Ryan, S.~G. 1997, \mnras, 286, 604

1174: 

1175: \bibitem[Dunkin et al. (1998)]{dun98} Dunkin, S.~K. \& Crawford, I.~A. 1998, \mnras, 298, 275

1176: 

1177: \bibitem[Espaillat et al.(2007)]{esp07} Espaillat, C., 

1178: Calvet, N., D'Alessio, P., Hern{\'a}ndez, J., Qi, C., Hartmann, L., Furlan, 

1179: E., \& Watson, D.~M.\ 2007, \apjl, 670, L135

1180: 

1181: \bibitem[Furlan et al.(2006)]{fur06} Furlan, E., et al.\ 

1182: 2006, \apjs, 165, 568

1183: 

1184: %\bibitem[Furton et al.(1999)]{fur99} Furton, D.~G., Laiho, J.~W.,

1185: %\& Witt, A.~N. 1999, \apj, 526, 752 

1186: 

1187: %\bibitem[Geballe \& van der Ween (1990)]{gvw90} Geballe, T.~R.,

1188: %\& van der Veen, W.~E.~C.~J. 1990, \aap, 235, L9 

1189: 

1190: %\bibitem[Geballe et al.(1992)]{geb92} Geballe, T.~R., Tielens, A.~G.~G.~M.,

1191: %Kwok, S., \& Hrivnak, B.~J. 1992, \apj, 287, L89 

1192: 

1193: %\bibitem[Geers et al.(2007)]{gee07} Geers, V.~C., van Dishoeck, E.~F., Visser, R., 

1194: %Pontoppidan, K.~M., Augereau, J.-C., Habart, E., \& Lagrange, A.~M.\ 2007, \aap, 476, 279

1195: 

1196: \bibitem[Geers et al. (2006)]{gee06} Geers, V.~C., et al. 2006, \aap, 459, 545

1197: 

1198: \bibitem[Geers et al.(2005)]{gee05} Geers, V.~C., Augereau, 

1199: J.-C., Pontoppidan, K.~M., K{\"a}ufl, H.-U., Lagrange, A.-M., Chauvin, G., 

1200: \& van Dishoeck, E.~F.\ 2005, High Resolution Infrared Spectroscopy in 

1201: Astronomy, 239

1202: 

1203: \bibitem[Grady et al.(2007)]{gra07} Grady, C.~A., et al.\ 

1204: 2007, \apj, 665, 1391

1205: 

1206: \bibitem[Grady et al.(2005)]{gra05} Grady, C. et al. 2005, \apj,

1207: 630, 958 

1208: 

1209: \bibitem[Gray \& Corbally (1998)]{gra98} Gray, R.~O. \& Corbally, C.~J. 1998, \aj, 116, 2530

1210: 

1211: %\bibitem[Greenberg et al.(2000)]{gre00} Greenberg, J.M., et al. 2000,

1212: %\apj, 531, 71 

1213: 

1214: %\bibitem[Habart et al.(2004)]{hab04} Habart, E., Natta, A., \& Krugel,

1215: %E. 2004, \aap, 427, 179 

1216: 

1217: %\bibitem[Habart et al.(2005)]{hab05} Habart, E., Natta, A., Testi,

1218: %L., Carbillet, M. 2005, \aap, in press

1219: 

1220: \bibitem[Habart et al.(2006)]{hab05} Habart, E., Natta, A., 

1221: Testi, L., \& Carbillet, M.\ 2006, \aap, 449, 1067 

1222: 

1223: %\bibitem[Habing (1968)]{ha68} Habing, H.~J. 1968, Bull. Astron.

1224: %Inst. Netherlands, 19, 421 

1225: 

1226: \bibitem[Herbig(1960)]{her60} Herbig, G.~H. 1960, \apjs, 4, 337 

1227: 

1228: %\bibitem[Herbst(1991)]{her91} Herbst, E. 1991, \apj, 366, 133

1229: 

1230: \bibitem[Hern{\'a}ndez et al.(2006)]{her06} Hern{\'a}ndez, 

1231: J., Brice{\~n}o, C., Calvet, N., Hartmann, L., Muzerolle, J., \& Quintero, 

1232: A.\ 2006, \apj, 652, 472  

1233: 

1234: \bibitem[Hony et al.(2001)]{hon01} Hony, S., Van Kerckhoven, C.,

1235: Peeters, E., Tielens, A.~G.~G.~M., \& Allamandola, L.~J. 2001,

1236: \aap, 370, 1030

1237: 

1238: \bibitem[Houk \& Crowley (1975)] {hou75} Houk, N., \& Cowley, A.~P. 1975, 

1239: Michigan Catalogue of Two-Dimensional 

1240:   Spectral Types for the HD Stars (Ann Arbor:  University of Michigan, 

1241:   Department of Astronomy)

1242:   

1243:  \bibitem[Houk (1978)] {hou78} Houk, N. 1978, Catalogue of Two-Dimensional Spectral Types for the HD 

1244:   Stars, Vol. 2 (Ann Arbor:  Univ. of Michigan, Department of Astronomy)

1245: 

1246: %\bibitem[Houk(1982)]{mss82} Houk, N. 1982, Michigan Catalogue of

1247: %Two-Dimensional Spectral Types for the HD Stars, Vol. 3 (Ann Arbor:

1248: %Univ. of Michigan) 

1249: 

1250: \bibitem[Houk \& Smith-Moore (1988)] {hou88} Houk, N. \& Smith-Moore, M. 1988, Michigan Catalogue of Two-Dimensional 

1251:   Spectral Types for the HD stars, Vol. 4 (Ann Arbor:  Univ. of Michigan, 

1252:   Department of Astronomy)

1253: 

1254: %\bibitem[Hudgins \& Allamandola(1995a)]{ha95a} Hudgins, D.~M. \&

1255: %Allamandola, L.~J. 1995a, J. Phys. Chem., 99, 3033 

1256: 

1257: %\bibitem[Hudgins \& Allamandola(1995b)]{ha95b} Hudgins, D.~M. \&

1258: %Allamandola, L.~J. 1995b, J. Phys. Chem., 99, 8978 

1259: 

1260: %\bibitem[Hudgins \& Allamandola(1997)]{ha97} Hudgins, D.~M. \& Allamandola,

1261: %L.~J. 1997, J. Phys. Chem., 101, 3472 

1262: 

1263: \bibitem[Hudgins \& Allamandola(1999)]{ha99} Hudgins, D.~M. \& Allamandola,

1264: L.~J. 1999, \apj, 516, L41 

1265: 

1266: %\bibitem[Hudgins et al.(1994)]{hud94} Hudgins, D.~M., Sandford,

1267: %S.~A., \& Allamandola, L.~J. 1994, J. Phys. Chem., 98, 4243 

1268: 

1269: %\bibitem[Hudgins et al.(1999a)]{hud99a} Hudgins, D.~M.,\& Allamandola,

1270: %L.~J. 1999a, \apj, 516, L41 

1271: 

1272: %\bibitem[Hudgins et al.(1999b)]{hud99b} Hudgins, D.~M.,\& Allamandola,

1273: %L.~J. 1999a, \apj, 513, L69 

1274: 

1275: %\bibitem[Higdon et al.(2004)]{hig04} Higdon, S.~J.~U.~H., et al. 2004,

1276: % \pasp, 116, 975

1277: 

1278: \bibitem[Houck et al.(2004)]{hou04} Houck, J.~R., et al. 2004, \apjs,

1279: 154, 18 

1280: 

1281: \bibitem[Isella et 

1282: al.(2006)]{ise06} Isella, A., Testi, L., \& Natta, A.\ 2006, \aap, 451, 951

1283: 

1284: %\bibitem[Jones et al.(1996)]{jon96} Jones, A.~P., Tielens, A.~G.~G.~M.,

1285: %\& Hollenbach, D.~J. 1996, \apj, 469, 740

1286: 

1287: \bibitem[Kenyon \& Hartmann(1995)]{kh95} Kenyon, S.~J., \& 

1288: Hartmann, L.\ 2004, \apjs, 101, 117

1289: 

1290: \bibitem[Kamp \& Dullemond(2004)]{kam04} Kamp, I., \& 

1291: Dullemond, C.~P.\ 2004, \apj, 615, 991

1292: 

1293: \bibitem[Kraemer et al.(2002)]{kra02} Kraemer, K.~E., Sloan, 

1294: G.~C., Price, S.~D., \& Walker, H.~J.\ 2002, \apjs, 140, 389

1295: 

1296: %\bibitem[Latter(1991)]{lat91} Latter, W.~B. 1991, \apj, 377, 187 

1297: 

1298: %\bibitem[Li \& Draine(2001)]{ld01} Li, A. \& Draine, B.~T. 2001,

1299: %\apj, 554, 778 

1300: 

1301: %\bibitem[Li \& Draine(2002)]{ld02} Li, A. \& Draine, B.~T. 2002,

1302: %\apj, 572, 762 

1303: 

1304: \bibitem[Li \& Lunine(2003)]{ll03} Li, A. \& Lunine, J.~I. 2003,

1305: \apj, 594, 987 

1306: 

1307: \bibitem[Malfait et al.(1998)]{mal98} Malfait, K., Bogaert, E, \&

1308: Waelkens, C. 1998, \aap, 331, 211 

1309: 

1310: %\bibitem[Mathis,  et al.(1983)]{mm83} Mathis, J.~S., Mezger, P.~G.,

1311: %\& Panagia, N. 1983, \aa, 128, 212 

1312: 

1313: \bibitem[Meeus et al.(2001)]{mee01} Meeus, G., Waters, L.~B.~F.~M.,

1314: Bouwman, J., van den Ancker, M.~E., Waelkens, C., \& Malfait, K.

1315: 2001, \aap, 365, 476 

1316: 

1317: \bibitem[Merin et al.(2004)]{mer04} Merin, B., et al. 2004, \aap,

1318: 419, 301

1319: 

1320: %\bibitem[Mouillet et al.(2001)]{mou01}Mouillet, D., Lagrange, A.~M.,

1321: %Augereau, J.~C., \& Menard, F. 2001, \aap, 372, L61

1322: 

1323: \bibitem[Muzerolle et al.(2003)]{muz03} Muzerolle, J., 

1324: Calvet, N., Hartmann, L., \& D'Alessio, P.\ 2003, \apjl, 597, L149

1325: 

1326: \bibitem[Nesterov et al. (1995)]{nes95} Nesterov, V.~V., Kuzmin, A.~V., Ashimbaeva, N.~T., Volchkov, A.~A., 

1327:   Roeser, S., Bastian, U. 1995, \aaps, 110, 367

1328: 

1329: %\bibitem[Oudmaijer et al.(1992)]{oud92} Oudmaijer, R.~D., van der

1330: %Veen, W.~E.~C.~J., Waters, L.~B.~F.~M., Trams, N.~R., Waelkens,

1331: %C., Engelsman, E. 1992, \aaps, 96, 625 

1332: 

1333: \bibitem[Peeters et al.(2002)]{pee02} Peeters, E., Hony, S., Van

1334: Kerckhoven, C., Tielens, A.~G.~G.~M., Allamandola, L.~J.; Hudgins,

1335: D.~M., \& Bauschlicher, C.~W. 2002 \aap, 390, 1089 

1336: 

1337: %\bibitem[Pietu et al.(2003)]{pie03} Pietu, V., Dutrey, A., \& Kahane,

1338: %C. 2003, \aap, 398, 565 

1339: 

1340: \bibitem[Raciine (1968)] {rac68} Racine, R. 1968, \aj, 73, 588

1341: 

1342: \bibitem[Robitaille et al.(2007)]{rob07} Robitaille, T.~P., 

1343: Whitney, B.~A., Indebetouw, R., \& Wood, K.\ 2007, \apjs, 169, 328

1344: 

1345: %\bibitem[Sandford et al.(1991)]{san91} Sandford, S.~A., Allamandola,

1346: %L.~J., Tielens, A.~G.~G.~M., Sellgren, K., Tapia, M., \& Pendleton,

1347: %Y. 1991, \apj, 371, 607

1348: 

1349: \bibitem[Shakura \& Syunyaev(1973)]{sha73} Shakura, N.~I., \& Syunyaev, R.~A.\ 1973, 

1350: \aap, 24, 337

1351: 

1352: %\bibitem[Shu et al.(1987)]{shu87} Shu, F.~H., Adams, F.~C., \&

1353: %Lizano, S. 1987, ARA\&A, 25, 23 

1354: 

1355: %\bibitem[Schutte et al.(1990)]{shu90} Schutte, W.~A., Tielens, A.~G.~G.~M., 

1356: % Allamandola, L.~J., Wooden, D.~H., Cohen, M. 1990, \apj, 360, 577

1357: 

1358: \bibitem[Siebenmorgen et al. (1998)]{sie98} Siebenmorgen, R., Natta, A., Kruegel, E., \&

1359: Prusti, T. 1998, \aap, 339, 134

1360: 

1361: \bibitem[Siebenmorgen et al. (2000)]{sie00} Siebenmorgen, R., Prusti, T., Natta, A., \& 

1362: M{\"u}ller, T.~G. 1998, \aap, 361, 258

1363: 

1364: %\bibitem[Sloan et al.(1999)]{slo99} Sloan, G.~C., Hayward, T.~L.,

1365: %Allamandola, L.~J., Bregman, J.~D., DeVito, B., \& Hudgins, D.~M.,

1366: %1999, \apj, 513, L65 

1367: 

1368: %\bibitem[Sloan et al.(2003)]{slo03} Sloan, G.~C., Kraemer, K.~E.,

1369: %\& Price, S.~D. 2003, \apjs, 147, 379 

1370: 

1371: %\bibitem[Sloan et al.(2005a)]{irs05} Sloan, G.~C., Spoon, H.~W.~W.,

1372: %\& Bernard-Salas, J., 2005, IRS Technical Report 05002, Low-resolution wavelength

1373: %calibration of the IRS (Ithaca, NY: Cornell, available at http://isc.astro.cornell.edu/tech/tr/) 

1374: 

1375: \bibitem[Sloan et al.(2005b)]{slo05} Sloan, G.~C., et al.\  2005, \apj, 632, 956

1376: 

1377: \bibitem[Sloan et al. (2007)]{slo07} Sloan, G.~C., et al. 2007, \apj, 664, 1144

1378: 

1379: \bibitem[Smith et al. (2007)]{smi07} Smith, J.~D., et al. 2007, \apj, 656, 770

1380: 

1381: %\bibitem[Spoon et al.(2004)]{spo04} Spoon, H.~W.~W., et al. 2004,

1382: %\apjs, 154, 184 

1383: 

1384: \bibitem[Strom et al.(1972)]{str72} Strom, S.~E, Strom, K.~M.,

1385: Yost, J., Carrasco, L., Grasdalen, G. 1972, \apj, 173, 353 

1386: 

1387: \bibitem[Strom \& Strom (1994)]{str94} Strom, K.~M. \& Strom, S.~E. 1994, \apj, 424, 237

1388: 

1389: \bibitem[Th{\'e} et al.(1994)]{the94} The, P.~S., de Winter, D., \&

1390: Perez, M.~R. 1994, \aaps, 104, 315

1391: 

1392: \bibitem[Thi et al.(2005)]{thi05} Thi, W.-F., van Dalen, B., 

1393: Bik, A., \& Waters, L.~B.~F.~M.\ 2005, \aap, 430, L61

1394: 

1395: \bibitem[Uchida et al.(2004)]{uch04} Uchida, K.~I., et al.\ 

1396: 2004, \apjs, 154, 439

1397: 

1398: %\bibitem[van den Ancker et al.(2001)]{vda01} van den Ancker, M.~E.,

1399: %Meeus, G., Cami, J., Waters, L.~B.~F.~M., \& Waelkens, C., 2001,

1400: %\aap, 369, L17 

1401: 

1402: %\bibitem[van den Ancker(2000)]{vda00} van den Ancker, M.E. 2000,

1403: %in Disks, Planetesimals, \& Planets, ASP Conf. Series 219, ed. F.

1404: %Garz\'{o}n, C. Eiroa, D. de Winter, \& T.~J. Mahoney, 242 

1405: 

1406: \bibitem[Vandenbussche et 

1407: al.(2004)]{vdb04} Vandenbussche, B., Dominik, C., Min, M., van Boekel, R., Waters, L.~B.~F.~M., Meeus, G., \& de Koter, A.\ 2004, \aap, 427, 519 

1408: 

1409: \bibitem[Visser et al.(2007)]{vis07} Visser, R., Geers, 

1410: V.~C., Dullemond, C.~P., Augereau, J.-C., Pontoppidan, K.~M., \& van 

1411: Dishoeck, E.~F.\ 2007, \aap, 466, 229

1412: 

1413: %\bibitem[Waters \& Waelkens(1998)]{ww98} Waters, L.~B.~F.~M \&

1414: %Waelkens, C. 1998, ARA\&A, 36, 233

1415: 

1416: \bibitem[Watson et al.(2007)]{wat07} Watson, D.~M., et al.\ 

1417: 2007, ArXiv e-prints, 704, arXiv:0704.1518 

1418: 

1419: \bibitem[Weinberger et al.(1999)]{wei99} Weinberger, A.~J., 

1420: Becklin, E.~E., Schneider, G., Smith, B.~A., Lowrance, P.~J., Silverstone, 

1421: M.~D., Zuckerman, B., \& Terrile, R.~J.\ 1999, \apjl, 525, L53

1422: 

1423: %\bibitem[Werner et al.(2004)]{wer04} Werner, M.~W., et al. 2004,

1424: %\apjs, 154, 1 

1425: 

1426: %\bibitem[Zuckerman et al.(1995)]{zuc95} Zuckerman, B., Forveille,

1427: %T., \& Kastner, J.~H. 1995, \nat, 373, 494

1428: 

1429: \end{thebibliography}

1430: 

1431: 

1432: 

1433: \clearpage

1434: 

1435: \begin{deluxetable}{lclccl}

1436: %\rotate

1437: \tablecolumns{6}

1438: \tabletypesize{\small}

1439: \tablewidth{0in}

1440: \tablecaption{Observation Log}

1441: 

1442: \tablehead{ \colhead{Target\tablenotemark{a}} & \colhead{IRS Campaign} & \colhead{Date} & \colhead{AOR ID}

1443: & \colhead{IRS Module(s)} & \colhead{Comments\tablenotemark{b}}}

1444: 

1445: \startdata 

1446: 

1447: HD 32509 & 4 & 2004 Feb.~27 & 3577344 & SL, LL& \\

1448: 51 Oph (HD 158643) & 5 & 2004 Mar.~22 & 3582464 & SL, LH & SL saturated \\

1449: HD 152404 (AK Sco) & 20 & 2005 Apr.~14 & 12700160 & SL, LL & \\

1450: HD 141569 & 4 & 2004 Mar.~3 & 3560960 & SL, LH & \\

1451: HD 281789 & 4 & 2004 Feb.~27 & 3529216 & SL, LH & \\

1452: HD 135344 (SAO 206462) & 11 & 2004 Aug.~07 & 3580672 & SL, LH & \\

1453: SU Aur (HD 282624) & 4 & 2004 Feb.~27 & 3533824 & SL, LH & \\

1454: HD 139614 & 11 & 2004 Aug.~07 & 3580928 & SL, LH & \\

1455: HD 142666 & 11 & 2004 Aug.~07 & 3586816 & SL, LH & \\

1456: HD 35187 & 4 & 2004 Feb.~27 & 3578112 & SL, LH & \\

1457: HD 31648 (MWC 480) & 4 & 2004 Feb.~27 & 3577088 & SL, LH & SL saturated \\

1458: HD 145718 & 4 & 2004 Feb.~27 & 3581440 & SL, LH & \\

1459: HD 100453 & 8 & 2004 Jun.~05 & 3578880 & SL, LH & \\

1460: HD 34282 & 13 & 2004 Sep.~26 & 3577856 & SL, LH & \\

1461: HD 144432 & 11 & 2004 Aug.~07 & 3587072 & SL, LH & SL saturated \\

1462: HD 97300 & 29 & 2006 Mar.~06 & 12697088 & SL, LH & \\

1463: HD 169142 & 5 & 2004 Mar.~26 & 3587584 & SL, LH & \\

1464: AB Aur (HD 31293) & 4 & 2004 Feb.~27 & 3533824 & SL, LH & SL saturated \\

1465: Elias 3-1 (V 892 Tau) & 4 & 2004 Feb.~27 & 3529984 & SL, LH & SL saturated \\

1466: HD 97048 & 29 & 2006 Mar.~06 & 12697088 & SL, LH & SL saturated \\

1467: HD 100546 & 30 & 2006 Apr.~06 & 3579136 & SH, LH & No SL \\

1468: HD 104237 (DX Cha) & 21 & 2005 May~05 & 12677632 & SH, LH & No SL\\

1469: 

1470: \enddata 

1471: \tablenotetext{a}{Alternate target names in parenthesis.}

1472: \tablenotetext{b}{'SL saturated' means the central map position had SL peaking higher than 10 Jy.

1473: In these cases a weighted average of adjacent map positions in SL is substituted. HD 100546 and HD 104237 are both too bright to observe in SL. See text for further details.}

1474: 

1475: \end{deluxetable} 

1476: 

1477: 

1478: 

1479: \begin{deluxetable}{ccc} % Table 2

1480: \tablecolumns{3}

1481: \tablewidth{0pt}

1482: \tablecaption{Wavelength intervals used to extract spectral features}

1483: \tablehead{

1484:   \colhead{Feature (\mum)} &

1485:   \colhead{$\lambda_{blue}$ (\mum)} & \colhead{$\lambda_{red}$ (\mum)} }

1486: \startdata

1487:  6.0    &  5.89--5.95  &  6.07--6.13   \\

1488:  6.2    &  5.80--5.95  &  6.54--6.69   \\

1489:  7.7--8.2    &  7.10--7.40  &  8.92--9.17   \\

1490:  8.6    &  8.32--8.44  &  8.86--8.98   \\

1491: 11.3    & 10.68--10.86 &  11.82--11.95 \\

1492: 12.0    & 11.72--11.90 &  12.13--12.25 \\

1493: 12.7    & 12.07--12.19 &  12.98--13.10 \\

1494: \enddata

1495: \end{deluxetable}

1496: 

1497: \clearpage

1498: \thispagestyle{empty}

1499: 

1500: 

1501: \begin{deluxetable}{lccccccc} % Table 3

1502: \rotate

1503: \tablecolumns{8}

1504: \tabletypesize{\small}

1505: \tablewidth{0pt}

1506: \tablecaption{Central wavelengths of the PAH features\tablenotemark{1}}

1507: \tablehead{

1508:   \colhead{ } & \multicolumn{6}{c}{$\lambda_{c}$ (\mum)} \\

1509:   \colhead{Target} & \colhead{PAH Classification\tablenotemark{2}} & \colhead{6.2~\mum\ PAH} & \colhead{7.7--8.2~\mum\ PAH} &

1510:   \colhead{8.6~\mum\ PAH} & \colhead{11.3~\mum\ PAH} &

1511:   \colhead{12.0~\mum\ PAH} & \colhead{12.7~\mum\ PAH} }

1512: 

1513: \startdata

1514: 

1515: 

1516: HD 32509  & \nodata & 6.2$\pm$0.1 & 8.2$\pm$0.3 & 8.5$\pm$0.2\tablenotemark{a}

1517:                                                                 & 11.3$\pm$0.1 & \nodata          & 12.3$\pm$0.1\tablenotemark{b}\\

1518: 51 Oph    & C & \nodata         & 8.5$\pm$0.1 & \nodata         & \nodata          & \nodata          & \nodata         \\

1519: HD 152404 & C & 6.27$\pm$0.03 & 8.34$\pm$0.05\tablenotemark{a}

1520:                                               & 8.67$\pm$0.05 & 11.27$\pm$0.03 & \nodata          & 12.7$\pm$0.2\\

1521: HD 141569 & B/C & 6.26$\pm$0.01 & 7.96$\pm$0.01 & 8.67$\pm$0.03 & 11.28$\pm$0.02 & \nodata          & 12.49$\pm$0.07\\

1522: HD 281789 & B/C & 6.25$\pm$0.01 & 7.92$\pm$0.02 & 8.65$\pm$0.02 & 11.25$\pm$0.04 & \nodata          & 12.64$\pm$0.07\\

1523: HD 135344 & B/C & 6.28$\pm$0.02 & 8.05$\pm$0.04 & 8.64$\pm$0.07 & 11.34$\pm$0.01 & 12.07$\pm$0.07\tablenotemark{b}

1524:                                                                                                       & 12.69$\pm$0.06\\

1525: SU Aur  &  C & 6.27$\pm$0.01 & 8.18$\pm$0.01 & \nodata         & 11.32$\pm$0.01 & \nodata          & 12.74$\pm$0.08\\

1526: HD 139614 & B/C & 6.277$\pm$0.004 & 8.04$\pm$0.01 & 8.68$\pm$0.02 & 11.32$\pm$0.01 & \nodata          & 12.66$\pm$0.11\\

1527: HD 142666 & B/C & 6.25$\pm$0.01 & 8.06$\pm$0.03 & 8.65$\pm$0.03 & 11.28$\pm$0.06 & \nodata          & 12.65$\pm$0.03\\

1528: HD 35187  & B/C & 6.27$\pm$0.01 & 8.04$\pm$0.01 & 8.66$\pm$0.02 & 11.30$\pm$0.04 & \nodata          & 12.66$\pm$0.05\\

1529: HD 31648  & C & 6.28$\pm$0.01 & 8.3$\pm$0.1 & \nodata         & 11.25$\pm$0.05 & \nodata          & 12.5$\pm$0.1\\

1530: HD 145718 & B/C & 6.258$\pm$0.004 & 8.05$\pm$0.02 & 8.65$\pm$0.03 & 11.28$\pm$0.01 & 12.0$\pm$0.1\tablenotemark{a}  & 12.63$\pm$0.06\\

1531: HD 100453 & B/C & 6.28$\pm$0.01 & 8.08$\pm$0.01 & 8.70$\pm$0.03 & 11.33$\pm$0.05 & 12.05$\pm$0.08 & 12.74$\pm$0.06\\

1532: HD 34282  & B/C & 6.261$\pm$0.003 & 7.93$\pm$0.02 & 8.65$\pm$0.01 & 11.27$\pm$0.01 & 11.96$\pm$0.08\tablenotemark{a}  & 12.63$\pm$0.02\\                                                                                                    

1533: HD 144432 & C & 6.27$\pm$0.02 & 8.39$\pm$0.05 & 8.66$\pm$0.07 & 11.26$\pm$0.03 & \nodata          & 12.59$\pm$0.07\\

1534: HD 97300  & A & 6.233$\pm$0.004 & 7.81$\pm$0.04 & 8.63$\pm$0.02 & 11.25$\pm$0.01 & 11.98$\pm$0.03\tablenotemark{b}  & 12.64$\pm$0.01\\

1535: HD 169142 & B/C & 6.261$\pm$0.003 & 7.99$\pm$0.01 & 8.67$\pm$0.01 & 11.30$\pm$0.01 & 12.03$\pm$0.02 & 12.70$\pm$0.02\\

1536: AB Aur    & B/C & 6.25$\pm$0.01 & 7.95$\pm$0.04 & 8.64$\pm$0.07 & 11.26$\pm$0.03 & \nodata          & 12.50$\pm$0.09\\

1537: Elias 3-1 & B/C & 6.245$\pm$0.002 & 7.81$\pm$0.04 & 8.64$\pm$0.03 & 11.23$\pm$0.06 & \nodata          & 12.63$\pm$0.09\\

1538: HD 97048  & B/C & 6.234$\pm$0.002 & 7.87$\pm$0.03 & 8.63$\pm$0.01 & 11.26$\pm$0.01 & 12.00$\pm$0.04 & 12.68$\pm$0.01\\

1539: HD 100546 & B/C & no SL         & no SL        & no SL  & 11.23$\pm$0.01 & 11.94$\pm$0.03 & 12.63$\pm$0.05\\

1540: HD 104237 & B/C & no SL         & no SL        & no SL  & 11.18$\pm$0.03 & 12.0$\pm$0.1\tablenotemark{a}

1541:                                                                                                       & 12.4$\pm$0.3\tablenotemark{a}\\

1542: 

1543: \enddata

1544: \tablenotetext{1}{Missing data indicated by "\nodata" are features for which the S/N in the spectrum was $<$1.}

1545: \tablenotetext{2}{PAH spectral classification according to \cite{pee02}.}

1546: \tablenotetext{a}{S/N of extracted feature $<$ 2.}

1547: \tablenotetext{b}{S/N of extracted feature $<$ 3.}

1548: \end{deluxetable}

1549: 

1550: \clearpage

1551: 

1552: 

1553: 

1554: \begin{deluxetable}{lrrrrrr} % Table 4

1555: %\rotate

1556: \tablecolumns{7}

1557: \tablewidth{0pt}

1558: \tabletypesize{\small}

1559: \tablecaption{Extracted strengths of the PAH features}

1560: \tablehead{

1561:   \colhead{ } & \multicolumn{6}{c}{$F$ (10$^{-15}$ W/m$^2$)} \\

1562:   \colhead{Target} & \colhead{6.2~\mum} & \colhead{7.7--8.2~\mum} &

1563:   \colhead{8.6~~\mum} & \colhead{11.3~\mum} & \colhead{12.0~\mum} &

1564:   \colhead{12.7~\mum}}

1565: \startdata

1566: 

1567: HD 32509  &    0.5$\pm$0.1 &   0.4$\pm$0.1 &  0.1$\pm$0.1 &   0.3$\pm$0.1 & $<$0.01 &    0.09$\pm$0.03\\

1568: 51 Oph    & $<$1 &   9$\pm$3 &  $<$2 &   $<$8 &    $<$0.3 & $<$0.7\\

1569: HD 152404 &    1.1$\pm$0.1 &   4$\pm$3 &  0.5$\pm$0.1 &   2.5$\pm$0.1 &    $<$0.03 &    0.2$\pm$0.1\\

1570: HD 141569 &    8.7$\pm$0.4 &  19.2$\pm$0.4 &  1.1$\pm$0.1 &   0.7$\pm$0.1 &    $<$0.1 &    0.33$\pm$0.04\\

1571: HD 281789 &    4.3$\pm$0.1 &   7.8$\pm$0.2 &  1.1$\pm$0.1 &   1.8$\pm$0.2 &    $<$0.1 &    0.4$\pm$0.1\\

1572: HD 135344 &    5.2$\pm$0.6 &  11.6$\pm$0.6 &  0.4$\pm$0.1 &   2.5$\pm$0.1 &    0.05$\pm$0.02 &    0.4$\pm$0.1\\

1573: SU Aur    &    5.9$\pm$0.3 &   8.4$\pm$0.1 &  $<$0.1 &   4.3$\pm$0.1 &    $<$0.4 &    0.4$\pm$0.1\\

1574: HD 139614 &    7.0$\pm$0.2 &  17.1$\pm$0.3 &  1.4$\pm$0.1 &   3.9$\pm$0.1 &    $<$0.1 &    0.6$\pm$0.1\\

1575: HD 142666 &   10.9$\pm$0.3 &  16.1$\pm$0.4 &  1.8$\pm$0.2 &   5.3$\pm$0.8 &    $<$0.1 &    1.0$\pm$0.1\\

1576: HD 35187  &    8.3$\pm$0.3 &  20.6$\pm$0.2 &  1.8$\pm$0.1 &   3.4$\pm$0.4 &    $<$0.2 &    0.6$\pm$0.1\\

1577: HD 31648  &    9.5$\pm$0.7 &  21$\pm$4 &  $<$3 &  16$\pm$2 & $<$0.2 &    2.1$\pm$0.4\\

1578: HD 145718 &   16.1$\pm$0.3 &  27.2$\pm$0.5 &  2.5$\pm$0.3 &   9.1$\pm$0.2 &    0.1$\pm$0.1 &    1.0$\pm$0.1\\

1579: HD 100453 &   26$\pm$1 &  37$\pm$1 &  2.8$\pm$0.3 &  12$\pm$2 &    0.24$\pm$0.01 &    1.5$\pm$0.3\\

1580: HD 34282  &   12.4$\pm$0.2 &  39$\pm$1 &  2.6$\pm$0.1 &   3.8$\pm$0.1 &    0.02$\pm$0.01 &    0.66$\pm$0.03\\

1581: HD 144432 &    2.2$\pm$0.2 &   10$\pm$1 &  0.9$\pm$0.3 &  11$\pm$1 &    $<$0.1 &    0.7$\pm$0.1\\

1582: HD 97300  &   13.4$\pm$0.2 &  40$\pm$2 &  4.5$\pm$0.3 &   7.6$\pm$0.2 &    0.2$\pm$0.1 &    2.22$\pm$0.04\\

1583: HD 169142 &   31.3$\pm$0.5 &  74$\pm$1 &  5.2$\pm$0.2 &  12.8$\pm$0.2 &    0.23$\pm$0.03 &    2.1$\pm$0.1\\

1584: AB Aur    &   66$\pm$2 & 125$\pm$5 & 12$\pm$3 &  23$\pm$2 & $<$0.3 &    5$\pm$1\\

1585: Elias 3-1 &   58$\pm$1 & 112$\pm$9 & 24$\pm$3 &  25$\pm$4 &    $<$2 &    5$\pm$1\\

1586: HD 97048  &   58$\pm$1 & 196$\pm$8 & 19$\pm$1 &  45$\pm$1 &    0.6$\pm$0.1 &   11.5$\pm$0.2\\

1587: HD 100546  &  no SL         & no SL         & no SL        & 132$\pm$6 &    5$\pm$2 &   12$\pm$2\\

1588: HD 104237 &  no SL         & no SL        & no SL        &  16$\pm$1 &    0.7$\pm$0.5 &    1.8$\pm$0.9\\

1589: 

1590: \enddata

1591: 

1592: \end{deluxetable}

1593: 

1594: 

1595: 

1596: 

1597: \begin{deluxetable}{lclcccccl} 

1598:   \rotate \tablecolumns{9} \tablewidth{0in} \tabletypesize{\small}

1599:   \tablecaption{Stellar Physical Characteristics} \tablehead{

1600:     \colhead{Star Name} & \colhead{$T_{e}(K)$} & \colhead{$Sp_{type}$}

1601:     & \colhead{$m_V$} & \colhead{d (pc)} & \colhead{SED

1602:       Group\tablenotemark{a}} & \colhead{$A_V$\tablenotemark{b}} &

1603:     \colhead{$i$ (deg)} & \colhead{Sp. Type Ref.}}

1604:  

1605: \startdata

1606: HD 32509 & 8970 & A2e & 7.51 &  151 & IIb\tablenotemark{1} & \nodata & 51 & \cite{mal98} \\

1607: 51 Oph & 9520 & A0Ve & 4.78  & 131 & IIa & 0.11 & 35 & \cite{mal98} \\

1608: HD 152404 & 6440 & F5 Ve & 9.14 & 145 & IIa & 0.62 & \nodata & \cite{mal98} \\ 

1609: HD 141569 & 9520 & A0 Ve & 7.09 & 99 & IIb & 0.37 & 51 & \cite{dmr97} \\ 

1610: HD 281789 & 9520 & A0 & 10.01 & 350 & IIa\tablenotemark{1} & \nodata & \nodata & \cite{nes95} \\

1611: HD 135344 & 6590 & F4 Ve & 8.61 & 140 & Ib & 0.31 & 45 & \cite{dmr97} \\

1612: SU Aur & 5945 & G1 & 9.42 &  152 & IIa\tablenotemark{1} & \nodata & \nodata & \cite{cal04} \\

1613: HD 139614 & 7850 & A7 Ve & 8.26 & 140 & Ia & 0.09 & 20 & \cite{hou78} \\

1614: HD 142666 & 7580 & A8 Ve & 8.81 & 145 & IIa & 0.93 & 55 & \cite{hou88} \\

1615: HD 35187 & 8970 & A2 Ve & 7.78 &  150 & IIa & 0.71 & \nodata & \cite{dun98} \\

1616: HD 31648 & 8970 & A2 pshe & 7.72  & 131 & IIa & 0.25 & 38 & \cite{mal98}\\

1617: HD 145718 & 7580 & A8 III/IVe & 8.98 &  131 & IIa\tablenotemark{1} & \nodata &  \nodata & \cite{hou88} \\

1618: HD 100453 & 7390 & A9 Ve & 7.79 & 112 & Ib & 0.02 & \nodata & \cite{hou75} \\

1619: HD 34282 & 9333 & A0.5 Vbe & 10.11 & 160 & Ib & 0.28 & 56 & \cite{gra98}\\

1620: HD 144432 & 7390 & A9 Ve & 8.15 & 253 & IIa & 0.17 & 45 & \cite{mal98} \\

1621: HD 97300 & 10500 & B9 V & 9.00 &  188 & IIb\tablenotemark{1} & \nodata & \nodata & \cite{hou75} \\

1622: HD 169142 & 8200 & A5 Ve & 8.11  & 145 & Ib & 0.43 & 8 & \cite{mal98} \\

1623: AB Aur & 9520 & A0 pe & 7.05 & 144 & Ia & 0.50 & 65 & \cite{rac68} \\

1624: Elias 3-1 &  10500 & B9 & 15.3 & 160 &  Ia & 4.05 & $<$ 45 &  \cite{str94} \\

1625: HD 97048 & 9520 & A0 pshe & 8.46 &175 & Ib & 1.26 & \nodata & \cite{hou75} \\

1626: HD 100546 & 10500 & B9 Vne & 6.70 & 103 & Ia & 0.26 & \nodata & \cite{mal98} \\

1627: HD 104237 & 7580 & A8 pe & 6.81 & 116 & IIa & 0.29 & 18 & \cite{mal98} \\

1628: 

1629: \enddata 

1630: 

1631: \tablenotetext{a}{SED Group classifications after \cite{mee01} and quoted from \cite{aa04}.}

1632: \tablenotetext{b}{Visual extinction values quoted from \cite{aa04}, except for 51 Oph \citep{mal98}.}

1633: \tablenotetext{1}{SED Group for SU Aur, HD 281789, HD 32509, HD 97300 determined using the criterion of \cite{mee01}. }

1634: \tablecomments{V892 Tau: distance assumed to be 160 pc (Taurus) \\

1635:  T$_e$from Kenyon and Hartman (1995) Table A5 for all luminosity class

1636: V, other luminosity class data from C. W. Allen (2000, Astrophysical Quantities, ed. A. N. Cox, Springer)}

1637: 

1638: \end{deluxetable}

1639: 

1640: 

1641: \clearpage

1642: 

1643: 

1644: 

1645: % Fig. 1

1646: %

1647: \begin{figure}

1648: \includegraphics[width=6in]{f1} 

1649: \caption{Spectra of all of our program stars plotted in

1650: $\lambda$  vs. F$_{\nu}$ units, but arbitrarily scaled to fit on the same axes for easy comparison of the general shape

1651: of the SEDs.  See Figure 2 for individual, flux-calibrated spectra. The designations 'U' (Stonger PAHs)

1652: and 'u' (Weaker PAHs) refer to the contrast of the PAH features relative to the continuum.}

1653: \end{figure}

1654: 

1655: \clearpage

1656: 

1657: % Fig. 2a

1658: %

1659: \begin{figure}

1660: \includegraphics[width=5in]{f2a}

1661: \caption{All spectra flux calibrated and displayed on individual

1662: axes in $\lambda$ vs. $F_{\nu}$ units. Spectra are labeled with star name and Meeus et all. (2001)

1663: SED group classification (I, Ia/b, II, IIa/b).}

1664: %An asterisk after the SED classification indicates a source not in their list, for which we applied 

1665: %their classification criterion.}

1666: \end{figure}

1667: \clearpage

1668: \begin{center}

1669: \includegraphics[width=5in]{f2b}\\

1670: {Fig. 2. --- Continued.}

1671: \end{center}

1672: \clearpage

1673: \begin{center}

1674: \includegraphics[width=5in]{f2c}\\

1675: {Fig. 2. --- Continued.}

1676: \end{center}

1677: 

1678: \clearpage

1679: 

1680: % Fig. 3

1681: %

1682: \begin{figure}

1683: \includegraphics[width=5in]{f3}

1684: \caption{Several examples of  spectra with dust continuum and dust features fit with spline

1685: curves (left, with anchor wavelengths

1686: of the spline fits indicated by open diamonds) and 

1687: continuum-subtracted (right), leaving the PAH spectra for

1688: our analysis. Examples range from 'weak or no PAH' (e.g. 51 Oph and HD 32509) to 'strong PAH'.}

1689: \end{figure}

1690: 

1691: \clearpage

1692: 

1693: % Fig. 4

1694: %

1695: \begin{figure}

1696: \includegraphics[width=6in]{f4}

1697: \caption{The continuum-subtracted spectrum of HD 135344, a relatively weak PAH emitter

1698: among our sample stars, and the linear fits used to define center wavelengths and extract the integrated fluxes

1699: of the PAH features after the spline fit is subtracted. We present the wavelengths used for this extraction

1700: method in Table 2.}

1701: \end{figure}

1702: 

1703: \clearpage

1704: 

1705: % Fig. 5

1706: %

1707: \begin{figure}

1708: \includegraphics[width=6in]{f5a} 

1709: \caption{Continuum-subtracted PAH spectra for all sources. Dotted vertical lines indicate the

1710: nominal wavelengths for the 6.2, 11.3, and 12.7~\mum\ features as well as for the nominal 

1711: peak of the 7.7--8.2~\mum\ complex. Sources arranged roughly in order of decreasing PAH

1712: integrated flux (top to bottom). Asterisks next to star names indicate sources that are not

1713: included in Figures 6-9 because their PAH feature wavelengths are uncertain.}

1714: \end{figure}

1715: \clearpage

1716: \begin{center}

1717: \includegraphics[width=6in]{f5b}\\

1718: {Fig. 5. --- Continued.}

1719: \end{center}

1720: 

1721: \clearpage

1722: 

1723: % Fig. 6

1724: %

1725: \begin{figure}

1726: \includegraphics[width=5in]{f6}

1727: \caption{Flux ratios from the different PAH features plotted vs. the ratio

1728: of the 7.9 and 11.3~\mum\ PAH features on a log-log scale. The ratio $F_{7.7-8.2}/F_{11.3}$

1729: traces the ionization fraction of the PAHs. The most ionized PAH spectrum is

1730: from HD 141569 (upper right data point in both plots). In the top panel, $F_{6.2}/F_{11.3}$

1731: increases with ionization of the PAHs (Pearson correlation coefficient, r=0.94). $F_{12.7}/F_{11.3}$ also increases with ionization fraction, suggesting that the size of the PAHs decreases with ionization fraction (correlation coefficient, r=0.69). The Class A and B prototypes and the Class C sources are from \cite{slo07} and

1732: we list them in Section 2.5. The A, B, C spectral classification system is described in Section 2 of the text.}

1733: \end{figure}

1734: 

1735: \clearpage

1736: 

1737: % Fig. 7

1738: %

1739: \begin{figure}

1740: \includegraphics[width=6in]{f7} 

1741: \caption{PAH $F_{12.7}/F_{11.3}$ vs. effective temperature. If the ratio of the 

1742: solo C-H bend (11.3~\mum) to the trio C-H bend (12.7~\mum) increases when PAH 

1743: molecules size is smaller, then these data indicate that PAH molecule size decreases with

1744: increasing stellar effective temperature (Pearson correlation coefficient, r=0.85). Class A, B, 

1745: and C prototypes are included to more

1746: clearly illustrate that HAeBe stars clump in the Class B-C range. The outlying source at

1747: top center is HD 141569.}

1748: \end{figure}

1749: 

1750: \clearpage

1751: 

1752: % Fig. 8

1753: %

1754: 

1755: \begin{figure}

1756: \includegraphics[width=6in]{f8}

1757: \caption{Wavelengths of the 7.7--8.2~\mum\ and 11.3~\mum\ PAH features plotted vs. effective temperature

1758: of the host star. Pearson correlation coefficients are r=0.95 and r=0.79 for the 7.7--8.2~\mum\ and 11.3~\mum\ features

1759: respectively.} 

1760: \end{figure}

1761: 

1762: \clearpage

1763: 

1764: % Fig. 9

1765: %

1766: 

1767: \begin{figure}

1768: \includegraphics[width=6in]{f9}

1769: \caption{Wavelengths of the 6.2~\mum\ and 8.6~\mum\ PAH features plotted vs. effective temperature

1770: of the host star. Pearson correlation coefficients are r=0.83 and r=0.81 for the 6.2~\mum\ and 8.6~\mum\ features

1771: respectively.} 

1772: \end{figure}

1773: 

1774: \clearpage

1775: 

1776: % Fig. 10

1777: %

1778: 

1779: \begin{figure}

1780: \includegraphics[width=6in]{f10a}

1781: \caption{Continuum spectral indices obtained from model spectral energy distributions: (a) with 

1782: dust settling in the inner wall of the disk and accretion rates $\dot{M}=10^{-8} M_\odot/yr$; (b) with dust settling in the inner wall and  $\dot{M}=10^{-9} M_\odot/yr$; (c) with NO dust settling in the inner wall and $\dot{M}=10^{-8} M_\odot/yr$; (d) with NO dust settling in the inner wall and $\dot{M}=10^{-9} M_\odot/yr$. Symbol sizes increase with dust settling (e.g. increasing $\epsilon$), thin symbols are for spectral type A2, bold symbols are for spectral type A6.} 

1783: \end{figure}

1784: \clearpage

1785: \begin{center}

1786: \includegraphics[width=6in]{f10b}\\

1787: {Fig. 10. --- Continued.}

1788: \end{center}

1789: 

1790: \clearpage

1791: 

1792: % Fig. 11

1793: %

1794: \begin{figure}

1795: \includegraphics[scale=0.8,angle=90]{f11a} 

1796: \caption{Continuum spectral indices of our sample stars. The dashed line indicates 

1797: a boundary in our sample, below which the disks should have a more flattened geometry 

1798: according to their Meeus et al. (2001) Group II classifications. Above the dashed line the disks 

1799: are mostly Group I (with the exceptions of HD 141569 and SU Aur), which should have more

1800: radially flared disks. All else being equal, increasing disk inclination should move sources towards the right in this diagram. Hydrocarbon ring (hexagon) symbol sizes scale with: (a) the {\it total} (integrated) 5-13~\mum\ PAH luminosity of the sources normalized to a distance of 100 pc from the observer; (b) the 6.2~\mum\ PAH feature luminosity {\it alone} normalized to a source distance of 100 pc from the observer; (c) the total PAH feature luminosity normalized to $L_{star}$, i.e symbol size is proportional to $[\Sigma F_{PAH}]/L_{star}$; and (d) $L_{star}$ only.}

1801: \end{figure}

1802: \clearpage

1803: \begin{center}

1804: \includegraphics[scale=0.8,angle=90]{f11b}\\

1805: {Fig. 11. --- Continued.}

1806: \end{center}

1807: \clearpage

1808: \begin{center}

1809: \includegraphics[scale=0.8,angle=90]{f11c}\\

1810: {Fig. 11. --- Continued.}

1811: \end{center}

1812: \clearpage

1813: \begin{center}

1814: \includegraphics[scale=0.8,angle=90]{f11d}\\

1815: {Fig. 11. --- Continued.}

1816: \end{center}

1817: 

1818: \clearpage

1819: 

1820: % Fig. 12

1821: %

1822: 

1823: \begin{figure}

1824: \figurenum{12} 

1825: \includegraphics[angle=0, scale=0.8]{f12}

1826: 

1827: \caption{Distribution of continuum color indices among Meeus Groups I \& II for our sample of HAeBe stars.} 

1828: 

1829: \end{figure}

1830: 

1831: 

1832: \end{document}

1833: 

1834: