1: %%\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: \def\Msun{M$_{\odot}$}
4:
5: \def\um{$\mu$m }
6:
7: \def\h2o{H$_2$O}
8:
9: \def\ch4{CH$_4$}
10:
11: \def\arcs{\ifmmode {''}\else $''$\fi}
12:
13:
14:
15: \received{2007 October 8}
16: \slugcomment{Accepted to ApJ}
17: \begin{document}
18:
19: \title{A SURVEY OF 3.3 $\mu$m PAH EMISSION IN PLANETARY NEBULAE}
20:
21:
22:
23:
24:
25: \author{ERIN C. D. SMITH\altaffilmark{1}, IAN S. MCLEAN\altaffilmark{2}}
26:
27:
28:
29: \altaffiltext{1}{Department of Physics and Astronomy, UCLA, Los Angeles, CA 90095-1562; erincds@astro.ucla.edu}
30:
31: \altaffiltext{2}{Department of Physics and Astronomy, UCLA, Los Angeles, CA 90095-1562; mclean@astro.ucla.edu}
32:
33:
34:
35: \begin{abstract}
36:
37:
38:
39: Results are presented from a pilot survey of 3.3 $\mu$m PAH
40: emission from planetary nebulae using FLITECAM, an instrument
41: intended for airborne astronomy with SOFIA. The observations were
42: made during ground-based commissioning of FLITECAM's spectroscopic
43: mode at the 3-m Shane telescope at Lick Observatory. Direct-ruled
44: KRS-5 grisms were used to give a resolving power (R)$\sim$1,700.
45: Targets were selected from IRAS,
46: KAO and ISO sources with previously observed PAH emission at
47: longer wavelengths. AGB stars and PN with C$/$O ratios $<$ 1 were
48: also added to the target list in order to test PAH detection
49: thresholds. In all, 20 objects were observed. PAH emission was
50: detected in 11 out of 20 observed targets.
51:
52:
53: \end{abstract}
54:
55:
56:
57: \keywords{infrared: general --- techniques: spectroscopic
58: --- planetary nebulae: general --- stars: AGB and post-AGB --- astrochemistry: PAHs}
59:
60:
61:
62: \section{Introduction}
63:
64: The broad infrared emission features observed at 3.3, 6.2, 7.7,
65: 8.6, and 11.2 $\mu$m have been the subject of astronomical
66: investigation since their first observation in the 1970s (Gillett et al. 1973; Gillett
67: et al. 1975). Through comparison to laboratory spectra and
68: additional observational data these bands have been attributed to
69: bending and stretching modes of polycyclic aromatic hydrocarbons
70: (PAHs) (Sellgren 1984; L\'{e}ger \& Puget 1984; Geballe et al. 1985). PAHs are formed
71: by aromatic structures of carbon and hydrogen which can bind
72: together into larger assemblies, forming a family of small
73: molecules such as pyrene, coronene and others. Emission features
74: associated with PAHs have been observed in a wide variety of
75: objects, including planetary nebulae, HII regions, proto-stellar
76: clouds and star-forming galaxies (Peeters et al. 2002; Hony 2002). This
77: ubiquity shows PAHs to be potentially important components of the
78: ISM, and has led to several studies aimed at correlating PAH
79: emission with region characteristics (Tokunaga et al. 1991; Roche
80: et al. 1996; Cohen \& Barlow 2005; Peeters et al. 2002; van
81: Diedenhoven et al. 2004). Recent surveys (Peeters et al. 2002, van Diedenhoven et al. 2004) using space-based telescopes have
82: shown variability in the PAH emission. Peeters et al. (2002)
83: introduced a classification scheme which grouped 6.2, 7.7 and 8.6
84: $\mu$m PAH spectra into three groups based on band FWHM, peak
85: wavelength and overall shape. HII regions, reflection nebulae, and
86: young stellar objects tended to exhibit Class A$_{6-9}$ PAH behavior,
87: while evolved stellar objects like planetary nebulae and post-AGB
88: stars exhibited class B$_{6-9}$ emission. A small number of objects,
89: tending to be very young AGB stars, exhibited class C$_{6-9}$ PAH
90: emission (Sloan et al. 2007). Van Diedenhoven et al. (2004) expanded this
91: classification to the 3.3$\mu$m and 11.2$\mu$m bands. Variations
92: in the 3.3 micron feature were found to be smaller than those in
93: the 7.7 and 6.2 micron features. This is explained by the fact
94: that the 3.3 and 11.2 $\mu$m features are due to stretching and
95: out-of-plane bending modes of the C-H bonds in the PAH molecules,
96: while the 6.2 and 7.7 $\mu$m features arise from stretching in the
97: C-C bonds (Allamandola et al. 1989). Tokunaga et al. (1991) and
98: van Diedenhoven et al. (2004) classified the 3.3 micron feature
99: into two classes: class A$_{3.3}$, or class 1, with emission centered at 3.290$\mu$m
100: and a FWHM of 0.040$\mu$m, and class B$_{3.3}$, or class 2, which have a FWHM of
101: 0.037$\mu$m with emission centered at 3.293$\mu$m (class B1$_{3.3}$) or
102: 3.297$\mu$m (class B2$_{3.3}$). Planetary nebulae and their evolutionary
103: precursors tended to be class A$_{3.3}$ objects. While other emission
104: features are sensitive to PAH charge state, comparisons with
105: laboratory data show the 3.3$\mu$m feature to arise from neutral
106: PAHs, rather than PAH cations (Allamandola et al. 1989).
107:
108: PAHs are produced in the hot outflows accompanying the last stages
109: of evolution of low and intermediate-mass stars. As a star evolves
110: off the Asymptotic Giant Branch (AGB), a period of dramatic mass
111: loss begins. Roche et al. (1996) found a threshold C$/$O ratio for
112: 3.3$\mu$m PAH emission in planetary nebulae of $\sim$0.6. A
113: study of 7 and 8$\mu$m emission by Cohen \& Barlow (2005) found a
114: roughly consistent threshold ratio of 0.56$\pm$0.3. In nebulae
115: with lower C$/$O ratios, nebular carbon is expected to be
116: incorporated in silicates and CO, while carbon-rich nebulae
117: produce hydrocarbons including PAHs (Allamandola et al. 1989). PAH
118: emission is caused by UV excitation. In planetary nebulae the
119: stellar remnant bathes the expanding gas shell with UV photons
120: which produce elemental nebular lines and also excite the
121: vibrational and stretching modes of PAH molecules. The relaxation
122: of these modes produces the broad infrared emission features which
123: have been observed in several planetary nebulae, as well as their
124: evolutionary precursors: proto-planetary nebulae and post-AGB
125: stars (Matsuura et al. 2005; Justtanont et al. 1996).
126:
127: This survey has three purposes: investigate the properties of 3.3
128: $\mu$m PAH emission in planetary nebulae by expanding the analysis
129: performed by Roche et al. (1996) and Cohen \& Barlow (2005); expand the
130: classification of 3.3 micron emission to a larger sample of
131: planetary nebulae using a higher resolution than previous surveys;
132: and demonstrate the FLITECAM instrument's spectroscopy mode in a
133: survey likely to be undertaken once aboard SOFIA.
134:
135: As part of the science verification of the FLITECAM instrument for
136: SOFIA, we proposed spectroscopic and narrow band imaging
137: observations of the 3.3 $\mu$m PAH feature, together with atomic
138: and molecular emission, in post-AGB stars, planetary nebulae, HII
139: regions and reflection nebulae. With the significantly increased
140: sensitivity of stratospheric observations, the overall goal of
141: this project is to trace the origin and evolution of PAHs from
142: their birth environment through their processing in the ISM to
143: their incorporation in proto-stellar material. The 3.3$\mu$m
144: feature is quite broad and typically resolved at R$\sim$300.
145: FLITECAM's grism mode has a higher spectral resolution that allows
146: investigation of not only the PAH feature itself, but also PAH
147: interaction with ionized and neutral gas. In principle, higher
148: resolution allows subtle changes in PAH emission shape to be
149: observed. In addition, FLITECAM's narrow-band imaging capability
150: allows one to study the spatial distribution of PAH emission in
151: spatially-resolved objects. FLITECAM offers three advantages for
152: this study: a wide field of view ($\sim$8 arc minutes diameter)
153: for imaging studies, good spectral resolution (R$\sim$1700) and,
154: when used with SOFIA, a reduced water vapor overburden and colder
155: telescope thus significantly reducing the background at this
156: wavelength.
157:
158: Prior to the start of SOFIA flight operations we decided to try a
159: pilot-study from a ground-based observatory in order to develop
160: the methodology for this project. Although the low water-vapor
161: advantage is completely lost and backgrounds are much increased
162: for a ground-based telescope, we have successfully detected the
163: 3.3 $\mu$m PAH feature, with relatively high spectral resolution,
164: in several planetary and proto-planetary nebulae. Wherever
165: possible, the strength of the emission has been quantified and
166: compared with other physical properties such as C$/$O ratio. The
167: pilot study demonstrates the potential for more sensitive PAH
168: studies when SOFIA is operational. Our results and methods,
169: including a brief description of the instrument, are described in
170: this paper.
171:
172: Target selection, the FLITECAM grism mode and the observations are
173: described in \S2, data reduction methods are explained in \S3 and
174: results are presented in \S4. Analysis and conclusions are given
175: in \S5.
176:
177: %section 2
178: \section{Observations}
179:
180: \subsection{Target Selection}
181: \input{tab1.tex}
182: Targets for our 3.3$\mu$m survey were selected from ISO, KAO and
183: IRAS observations of planetary nebulae, proto-planetary nebulae
184: and post-AGB stars. Our sample was compiled based on the reported
185: presence of PAH emission in longer wavelength bands observed from
186: space, as well as practical issues of signal-to-noise limitations
187: at a low-altitude ground-based site. Of the 20 objects in the
188: survey, 13 had previous long-wave PAH detection in at least one
189: band (Cohen et al. 1989; Jourdain de Muizon et al. 1990; Rinehart et al. 2002). Additional sources were added in order to extend the sample
190: to post-AGB stars and proto-planetary nebulae. Since PAH emission
191: is expected only in objects with a carbon-rich dust chemistry, all
192: but 3 objects have C$/$O ratios above the threshold ratio for PAH emission determined by Roche et al. and Cohen \& Barlow. One O$-$rich nebula shows a somewhat
193: mixed dust chemistry, while the remaining two O$-$rich objects
194: were used as a test of methodology (Jourdain de Muizon et al. 1990). The targets in this sample
195: represent a somewhat larger range in morphology, size and
196: brightness than would normally be selected for a survey of this
197: type. This was done in order to meet the broad goals of the pilot
198: study and to facilitate the fullest possible science verification
199: of the FLITECAM grism mode.
200:
201: All data presented here were obtained using the Shane 3-meter
202: telescope at Lick Observatory, Mt. Hamilton, CA, with the FLITECAM
203: instrument attached. Observations were collected over the course
204: of five separate observing runs from May 2004 to October 2006.
205: Table 1 lists all the targets, total observation times, and the
206: standard stars used to remove telluric absorption lines. Shorthand
207: names (such as BD30 for BD+30$^{\rm o}$3639) are used throughout the
208: text for simplicity, but full names are given in Table 1.
209:
210: %section 2.2
211: \subsection{Instrumentation}
212:
213: As mentioned above, our pilot-study of 3.3$\mu$m emission from PN
214: was carried out as part of the commissioning and performance
215: verification of a new instrument called FLITECAM (First Light Test
216: Experiment Camera). FLITECAM is a 1-5$\mu$m camera developed by
217: us at the UCLA Infrared Laboratory (P.I.: McLean) for NASA's SOFIA
218: (Stratospheric Observatory for Infrared Astronomy) project. SOFIA
219: is a modified Boeing 747-SP airplane with a 2.5-meter f$/$19.6
220: bent-Cassegrain telescope operating at altitudes up to 45,000 ft
221: and therefore above 99\% of the atmosphere's water vapor content.
222: Most of the SOFIA instruments are designed for much longer
223: wavelengths, but because FLITECAM operates in the near-infrared,
224: and because the f$/$17 optics of the Lick 3-meter telescope
225: provides a plate scale similar to that of SOFIA, FLITECAM can be
226: used from the ground without additional optics. Performance and
227: technical details of the instrument are described in McLean et al.
228: 2006. First light was obtained in October 2002 at Lick Observatory
229: and the instrument has accumulated data with the 3-meter telescope
230: on eight separate occasions. Photometric results from a narrow
231: band survey of nearby star-forming regions for methane-bearing
232: planetary mass objects have been published elsewhere (Mainzer and
233: McLean 2003; Mainzer et al. 2004; McLean et al. 2006).
234:
235: Briefly, FLITECAM utilizes large refractive optics to inscribe an
236: $\sim$8 arc minute diameter FOV on a 1024x1024 InSb ALADDIN III
237: detector with a plate scale of just under 0.5 arc seconds per
238: pixel. The instrument is cooled with 20 liters of LN$_{2}$ and 20
239: liters of LHe and can be operated in either the up-looking
240: position (Lick) or the horizontal position (SOFIA). Another
241: feature is a dual filter wheel which houses standard broadband
242: filters, several narrow band filters plus a suite of three
243: direct-ruled KRS-5 grisms and their appropriate order-sorting
244: filters. Each grism can be used in one of three orders giving a
245: total of 9 settings to cover the entire 1-5 $\mu$m region. The
246: KRS-5 grisms provide a moderate resolution spectroscopy mode in
247: conjunction with fixed slits of either 1\arcs~ or 2\arcs~ width
248: and 60\arcs~ total length to yield resolving powers of R$\sim$1700
249: and 900 respectively (Smith and McLean 2006). When airborne,
250: FLITECAM will be the only instrument providing spectral coverage
251: from 2.5-5.5$\mu$m with a low water vapor and low background
252: environment.
253:
254: %Section 2.3
255:
256: \subsection{Observations}
257: %section 3
258: FLITECAM's imaging mode was used to place each object on the
259: selected slit, and the observatory's offset CCD guide camera was
260: used to hold the source on the slit. Spectroscopic observations
261: were made in a standard ABBA nod pattern. Typically, integrations
262: of 300 s each (3 s exposure with 100 coadds) were taken with the
263: object placed at two positions, designated A and B, separated by
264: $\sim$26$\arcs$ on the $\sim$60$\arcs$ long entrance slit of
265: FLITECAM. Shorter exposure times were used for brighter objects.
266: The total number of ABBA cycles depends on the object, but all
267: targets were observed for a minimum of one cycle. Total
268: integration times per object ranged from 20 minutes to 1.5 hours.
269: Signal-to-noise ratios were typically greater than 10 (10\%) per
270: resolution element. Seeing conditions were $\sim1\farcs0-1\farcs2$
271: and a slit width of 1$\farcs$0 ($\sim$2 pixels) was used for all
272: observations, except in the case of HB~5 for which we used a
273: $2\farcs0$ (4 pixels) slit because of poorer seeing.
274:
275: The grism mode selected provides coverage from 3.0-3.4$\mu$m. This
276: spectral region exhibits significant contamination from absorption
277: lines of terrestrial water vapor, making ground-based observations
278: very difficult. In principle, these absorption features can be
279: removed by observations of standard stars. Two methods were used
280: to obtain satisfactory telluric corrections; 1) ATRAN models of
281: the atmosphere (Lord 1979); 2) observations of a bright A0V star
282: or MIII giant star at an airmass very close to that of the target
283: object. Great care was taken to ensure the telluric standard did
284: not saturate the FLITECAM detector. Arc lamp spectra, taken at the
285: end of the night, and OH night-sky lines in the observed spectra,
286: were used for wavelength and dispersion calibration. A white-light
287: spectrum and a corresponding dark frame were also obtained for
288: flat-fielding. Calibration stars are listed in Table 1.
289:
290: \section{Data Reduction Methods}
291: \subsection{Redspec}
292: Survey data were reduced using REDSPEC, an IDL-based software
293: package developed at UCLA for spectra obtained with the Keck
294: NIRSPEC instrument by S. Kim, L. Prato and I. McLean\footnote{See
295: http://www2.keck.hawaii.edu/inst/nirspec/redspec/index.html}.
296: REDSPEC was modified for use on FLITECAM data by adding routines
297: to flip the raw spectra, to automatically subtract nod pairs to
298: enable spatial rectification of thermal-IR spectra (in which the
299: traces are hard to see against the background), and to apply
300: FLITECAM wavelength solutions to the spectra.
301:
302: To extract spectra free from atmospheric background and uneven
303: detector response, the difference of an A/B image pair was formed
304: and flat-fielded. The flat-fielded difference frame was then
305: rectified using the spatial and spectral maps and the raw spectrum
306: produced by summing 5$-$10 rows from each trace in the rectified
307: image. The extracted traces (one positive, one negative) are then
308: subtracted again to produce a positive spectrum with residual
309: night-sky emission line features removed, unless a line was
310: saturated. Telluric standard star spectra were reduced in the
311: same way. After interpolating over the intrinsic hydrogen
312: absorption lines, the reduced target spectrum was divided by the reduced
313: calibrator star spectrum to remove telluric features. The true
314: slope of the target spectrum was restored by multiplication with a
315: blackbody spectrum of T$_{eff}=9500$ K for an A0 V star, and
316: T$_{eff}=3000$ K for an MIII star(Tokunaga 1991). Finally, the
317: spectra reduced from multiple A/B pairs were averaged together to
318: improve the signal-to-noise ratio.
319:
320: \subsection{ATRAN telluric correction}
321: Ground-based thermal infrared observations are dramatically
322: affected by variability in atmospheric water vapor and background
323: emission. Therefore, good telluric correction is critical to
324: accurate reduction of thermal IR spectra. In many cases,
325: appropriate telluric standards are either too faint for good
326: correction, or too widely separated for a good airmass match. In
327: addition, the infrared backgrounds at Lick observatory can change
328: during observations of target objects, making accurate telluric
329: correction using observed standards very difficult. As an
330: alternative to observed telluric standards, ATRAN, an atmospheric
331: modelling program developed by Steve Lord, was used to generate
332: artificial telluric calibrator spectra at the target's zenith
333: angle and airmass. ATRAN models allow tweaking of water vapor
334: over-pressure and other atmospheric conditions to produce
335: optimized telluric corrections for each target. Spectra of targets
336: with high S/N ratios (NGC 7027, BD+30$^{\rm o}$3639) showed no appreciable
337: differences between telluric-corrected spectra using observed
338: standard stars and ATRAN-corrected spectra. Low S/N object spectra
339: were noticeably improved using ATRAN-produced telluric standards.
340: This makes sense because low S/N targets were observed for longer
341: total time periods than high S/N targets, increasing the
342: contribution of atmospheric variability in the resulting spectrum.
343: While optimized ATRAN models can be fit individually to each
344: target AB pair in a long sequence, the single observed telluric
345: calibrator is only a good fit for one moment in time.
346:
347: ATRAN models were computed using a standard 2 layer atmosphere at
348: 4000 ft elevation using the appropriate zenith angle for each
349: target observation. The water vapor over-pressure was varied from
350: 8 to 12 mm (around the standard predicted value of 10 mm) for each
351: target object to obtain the best fit to the data. This produced a
352: normalized, flat-fielded calibrator spectrum. Each target
353: observation was reduced using the REDSPEC software package as
354: described in \S3.1, but without division by a calibrator.
355: Instead, the reduced target spectrum was divided by the
356: ATRAN-produced telluric calibrator spectrum.
357:
358: %section 4
359: \section{Results}
360:
361: \subsection{Overview}
362:
363: Those objects that yielded the best PAH spectra tended to be
364: young, high-excitation compact planetary nebulae. Objects such as
365: NGC 3242 for example, were too low in surface brightness to yield
366: a useful spectrum given the high background of the ground-based
367: observations. Very bright objects, such as AFGL 3086, also did not
368: yield a PAH detection because the central object saturated the
369: FLITECAM detector, even at the shortest possible exposure times,
370: and any extended PAH emission located far enough away from the
371: central object so as to avoid saturation was too low in surface
372: brightness to be detected. All reduced spectra show residual
373: structure from atmospheric water vapor lines that could not be
374: removed. L-band observations are difficult at Lick Observatory
375: because the site is only 4209 ft above sea level, and is located
376: in a particularly humid area. These conditions lead to highly
377: variable telluric features, often variable on the timescale of one
378: nod (5 minutes). While the telluric correction steps described
379: above removes the majority of these features, even the best
380: spectrum, that of NGC 7027, shows residual structure. It should be
381: noted that these complications exist, although to a much reduced
382: level, even at the world's best astronomical sites, such as Mauna
383: Kea. However, with the caveat on the limitations of the telluric
384: corrections, we can analyze the detected PAH emission features.
385: %Figure 1: NGC7027 spectrum
386: \begin{figure}
387: \epsscale{1.15}
388: \plotone{f1.eps}
389: \caption{Spectrum of NGC 7027 with prominent emission features labelled. The missing data from 3.31 to 3.33 $\mu$m is not plotted due to saturated atmospheric methane and water features. The dashed line shows the gaussian fit to the data.}.\label{fig1}
390: \end{figure}
391: %--------------------
392: Figure 1 shows the extracted 3.3$\mu$m spectrum for NGC 7027. Of
393: course, NGC 7027 is a particularly bright, high-excitation
394: planetary nebula with well-known PAH emission in all bands.
395: Hydrogen and helium lines are seen in emission, as is the broad
396: PAH feature. One hydrogen line, Pfund delta (3.296$\mu$m) is seen
397: in emission on top of the broad PAH feature. Also evident is the
398: effect of residual terrestrial water absorption. The blue slope of
399: the PAH feature has several unsaturated, but variable water
400: absorption features, leading to a slightly stair-stepped shape
401: even on NGC 7027. On the red slope, there is a deep atmospheric
402: methane line at 3.32$\mu$m. This line cannot be removed from the
403: spectrum, thus no data points are plotted for the
404: 3.310-3.330$\mu$m region.
405:
406: The FLITECAM spectra presented here cover from 3.0 to 3.4$\mu$m at
407: a resolution of R$\sim$1700. After reduction, all spectra
408: underwent a 5-sigma cut to remove hot pixels and residuals from
409: saturated atmospheric lines. Figures 2 and 3 show a selection of the
410: target spectra all plotted in a similar way. There are four
411: classes of resulting spectra: 1) continuum objects, 2) PAH
412: non-detections, 3) PAH detections and 4) PAH possible detections.
413: (See Table 1 again for a list of all observed PNs and their
414: classification.)
415:
416: %Figure 2: All nebula spectrum
417: \begin{figure*}
418: \epsscale{.75}
419: \plotone{f2.eps}
420: \caption{Reduced spectra from PAH survey. The missing data from 3.31 to 3.33 $\mu$m is not plotted due to saturated atmospheric methane and water features. Fainter objects show noise spikes in the 3.0 to 3.2 $\mu$m region due to terrestrial absorption lines. Detected emission lines are listed in Table 2. All detected PAH features are well-fitted by a Gaussian curve. IRAS 21282 shows some non-symmetry, but has FWHM and peak PAH wavelength emission consistent with class A$_{3.3}$ emission.}.\label{fig2}
421: \end{figure*}
422:
423: %Figure 3: All nebula spectrum
424: \begin{figure*}
425: \epsscale{.75}
426: \plotone{f3.eps}
427: \caption{Reduced spectra from PAH survey. The missing data from 3.31 to 3.33 $\mu$m is not plotted due to saturated atmospheric methane and water features. Fainter objects show noise spikes in the 3.0 to 3.2 $\mu$m region due to terrestrial absorption lines. Detected emission lines are listed in Table 2. All detected PAH features are well-fitted by a Gaussian curve.}.\label{fig3}
428: \end{figure*}
429: %--------------------
430: \subsection{Continuum objects}
431: Four members of the sample exhibit a stellar spectrum with no
432: obvious emission features. All four of these objects are post-AGB
433: or AGB stars, and are thus less evolved than the rest of the
434: sample. RAFGL 3116 does have reported PAH detections at longer
435: wavelengths from ISO observations (Jourdain de Muizon et al. 1990), but does not
436: show an obvious PAH 3.3$\mu$m band peak in our observations. The
437: feature may be present, but could be swamped by the strong, red
438: spectrum of the highly evolved AGB star. The spectrum of RAFGL
439: 3116 in Figure 2 is typical of this type of object.
440:
441: \subsection{PAH non-detections}
442: Three members of the sample (M 2-9, NGC 3234, and NGC 6210) exhibit standard PN features (i.e.
443: atomic emission lines and non-existent or very faint continuum),
444: but yet show no PAH emission at 3.3$\mu$m.All
445: three non-detections have an oxygen-rich chemistry, with
446: C$/$O$<<$1. In these objects, crystalline silicate features tend to dominate, rather than PAH emission (Allamandola 1989). M 2$-$9 in Figure 2 is an example of a oxygen-rich
447: planetary nebula with no PAH emission.
448:
449: \subsection{PAH detections and possible detections}
450: Eleven members of the sample show distinct PAH emission at
451: 3.3$\mu$m, while two more targets (WHME-1 and J900) show
452: possible PAH emission. PAH flux was quantified by fitting a
453: gaussian to the PAH band assuming a flat continuum. NGC 7027 is
454: the best example of a strong PAH detection, and can be seen either
455: in Figure 1 or Figure 2. WHME-1 and J900 are examples of
456: possible-detection of PAH emission. While PAH emission is seen at
457: longer wavelengths in these objects, the nebulae may have been too
458: faint compared to the high backgrounds at Lick for solid 3.3$\mu$m
459: detections. Figures 2 and 3 show the extracted spectra for all the PAH
460: detections. Hydrogen emission at 3.296$\mu$m is evident in many of
461: the sources with a PAH detection (HB~5, M3-35, BD+30$^{\rm o}$3639 and NGC 7027 in
462: particular). This profile was subtracted out of the extracted
463: spectra by fitting a gaussian before fitting a curve to the PAH
464: band itself in order to determine its strength and shape. Table 2
465: summarizes the derived PAH measurements for the observed nebulae. Those nebulae with non detections have upper limits on the PAH equivalent widths. The C/O ratios of the nebulae, if known, are also listed.
466:
467: \section{Discussion and Analysis}
468: \input{tab2.tex}
469: Prior to further analysis, each detected PAH feature was fitted
470: with a gaussian to determine the feature's FWHM and central
471: wavelength. The average central wavelength of the PAH emission
472: feature was 3.288$\mu$m $\pm$ 0.002$\mu$m, with a FWHM of 0.042
473: $\mu$m$\pm$ 0.002$\mu$m. This is consistent with surveys of
474: similar objects that show this feature to have a gaussian profile
475: centered at 3.289$\mu$m and a FWHM of 0.042$\mu$m (Tokunaga et al.
476: 1991).
477:
478: For more detailed analysis of the 3.3$\mu$m feature we adopt the
479: method used by Roche et al. (1996), that is, we quantify PAH
480: emission strength by determination of the emission band's
481: equivalent width (EW). This method is superior to absolute flux
482: measurements as the EW is less affected by atmospheric absorption
483: and fluctuation. However, the determination of equivalent width
484: depends on an accurate model of continuum emission for comparison
485: to the measured feature, and can thus be rendered inaccurate for
486: objects with weak continua, or in regions where emission plateaus
487: mask true continuum emission. As this survey was limited in scope
488: to only the brightest, most compact planetary nebula, most of the
489: observed objects exhibited continuum emission above the threshold
490: necessary for good equivalent width determination. The wavelength
491: regime of this study (3.0-3.4$\mu$m) is also advantageous to
492: determining continuum emission levels, because the region just
493: blue of the PAH emission is relatively free of atmospheric
494: absorption bands, and is constant over a range roughly comparable
495: to the width of the PAH feature. Three nebulae have unusable
496: equivalent width measurements due to either weak PAH emission (as
497: in WHME1 and J900), or noisy continuum spectra (as in M 4-18).
498:
499: \subsection{Correlating PAH and C/O ratio}
500:
501: The most well-studied correlation is between the nebular C$/$O
502: ratio and the PAH emission strength. Current theory predicts that
503: nebulae with high C$/$O ratios will exhibit PAH emission, while
504: more oxygen rich nebulae will exhibit silicate features, having
505: trapped most nebular C$/$O in CO molecules. Figure 4 shows the
506: derived 3.3$\mu$m emission equivalent width plotted against the
507: nebular C$/$O ratio obtained from the literature (Kholtygin 1998). Only 7 nebulae
508: in our sample have known C/O ratios, and only 5 of these nebulae
509: have usable equivalent widths. PAH emission equivalent widths are
510: uncertain up to 10 percent, and even the known C$/$O ratios
511: measurements have large uncertainties.
512:
513: Figure 4 shows the C$/$O ratio plotted against the five nebulae
514: with measured equivalent widths. Also plotted is the linear
515: regression fit to the data. This shows a PAH equivalent width
516: detection threshold at C$/$O=0.65 $\pm$ 0.28. This result is
517: within the uncertainties obtained by Roche et. al. (1996), who
518: performed a similar analysis, and with Cohen \& Barlow (2005), who
519: measured a C$/$O ratio versus 7.7$\mu$m PAH emission.
520:
521: \subsection{PAH emission classification}
522:
523: For further analysis we characterize the PAH emission according to
524: the system suggested by van Diedenhoven et al. using band shape,
525: location and FHWM. Class A$_{3.3}$ emission is characterized by symmetric
526: PAH emission centered at 3.290$\mu$m with a FWHM of 0.040 $\mu$m.
527: Class B$_{3.3}$ emission is centered at 3.293 or 3.297$\mu$m with a FWHM
528: of 0.030$\mu$m. In van Diedenhoven's survey planetary nebulae were
529: classified as class A sources. Tokunaga et al. (1991) used similar
530: criteria to categorize 3.3 $\mu$m emission, with their class 1
531: equivalent to the van Diedenhoven's class A$_{3.3}$. The results of this
532: classification are listed in Table 2. All sources with PAH
533: emission are of class A$_{3.3}$, as is consistent with both van
534: Diedenhoven and Tokunaga. The presence of strong atmospheric water
535: lines makes analysis of PAH emission line shape more difficult,
536: but all detected PAH features have FWHM and central emission wavelengths consistent with class A$_{3.3}$ emission, and appear to be largely symmetric. Other surveys of 3.3 $\mu$m emission in planetary nebulae
537: (Roche et al. 1996) were at lower resolution than the R=1700
538: spectra in this study. Our resolution allows us to fit and remove
539: the Pfund delta hydrogen emission line, which could otherwise
540: affect the fitted gaussian curve and apparent line shape.
541:
542: The 3.3 $\mu$m feature shows relatively little variation even in
543: very different emission sources (van Diedenhoven 2004, Cohen \& Barlow 2005,
544: Tokunaga 1991), while longer wavelength PAH emission shows greater
545: variation in central wavelength and FWHM (Peeters et al. 2002). In
546: longer wavelength studies planetary nebulae are classified as
547: class B$_{6-9}$ objects, while some very young post-AGB stars exhibit
548: class C$_{6-9}$ emission. This apparent evolution in PAH emission during
549: the AGB to planetary nebula transition is the primary motivation
550: for expanding the sample of evolved stars investigated at high
551: resolution in all PAH bands.
552:
553: In the current survey we did not sample multiple portions of the
554: target objects, instead sampling a 4-5 arcsec band across the
555: extent of the nebula. If there are more subtle changes in 3.3
556: $\mu$m emission due to nebular position, they would not be
557: detected by this approach.
558:
559: %Figure 4: EW vs C/O ratio
560: \begin{figure}
561: \epsscale{1.1}
562: \plotone{f4.eps}
563: \caption{PN C/O Ratio versus measured PAH EW. See Table 2 for data and references. The plotted line yields a threshold value of C$/$O=0.65 $\pm$ 0.28}.\label{fig4}
564: \end{figure}
565:
566: \section{Conclusions}
567:
568: With a resolution $\sim$1,700 we have carried out a preliminary
569: survey of 3.3 $\mu$m PAH emission using FLITECAM, an instrument
570: developed at UCLA, on the Lick Observatory 3-m telescope. This
571: survey is a pilot project for a deeper program to be carried out
572: using the same instrument on board the Stratospheric Observatory
573: for Infrared Astronomy (SOFIA).
574:
575: (1)Out of 20 objects observed, 11 showed detectable 3.3$\mu$m PAH
576: emission.
577:
578: (2)We found the 3.3$\mu$m feature to have a FWHM and central
579: wavelength consistent with that found by Tokunaga et al. (1991),
580: Roche et al. (1996) and van Diedenhoven et al. (2004).
581:
582: (3)A correlation was found between C$/$O ratio and PAH emission
583: strength, consistent with that reported by Roche et al. (1996) and
584: Cohen \& Barlow (2005). We calculate a threshold C/O ratio of 0.65 $\pm$ 0.28.
585:
586: (4)We classified the 11 objects showing PAH emission according to
587: the system suggested by van Diedenhoven.
588:
589: Our high-resolution survey of 3.3$\mu$m emission has expanded the
590: sample of PNs with characterized PAH emission at this wavelength.
591: This step is important for evolutionary studies because shifts are
592: seen in longer wavelength PAH emission between PNs and their
593: evolutionary pre-cursors. It also serves as a starting point for
594: detailed studies of spatial variations of PAH emission in
595: individual planetary nebulae, as well as studies of PAH emission
596: in young AGB stars.
597:
598: Although the sample of objects presented is relatively small,
599: these high-resolution 3.3$\mu$m PAH spectra provide a clear
600: indication of the importance of airborne astronomy to this field.
601: With SOFIA, the thermal background will be reduced by more than an
602: order of magnitude and none of the water vapor lines will
603: saturate. Higher signal-to-noise ratios will be possible, and it
604: may soon be practical to detect subtle changes in line shape due
605: to PAH ionization. Studies of the 3.3 $\mu$m band in Orion have
606: shown a shift from class A$_{3.3}$ to class B$_{3.3}$ emission with position (van Diedenhoven et
607: al. 2004), as well as a shift from symmetric to asymmetric
608: 3.3$\mu$m emission. Similar shifts may occur in planetary nebulae,
609: and would be observable with lower backgrounds and water vapor
610: contributions. Likewise, in extended objects the lower backgrounds
611: will make it possible to probe the shape of the PAH feature across
612: the spatial extent of the nebula.
613:
614: \acknowledgements
615: The authors wish to thank the staff of the Lick
616: Observatory for their outstanding support. We also thank Eric
617: Becklin (UCLA) for his support throughout the development of the
618: FLITECAM instrument. E.C.S. is supported by a NASA Graduate Student Research Program (GSRP) fellowship. The authors wish to thank the anonymous referee, Dr. Mike Jura and Dr. Mark Morris for their helpful suggestions in improving this manuscript.
619:
620: \begin{thebibliography}{}
621: \bibitem[Allamandola et al.(1989)]{1989ApJS...71..733A} Allamandola, L.~J.,
622: Tielens, G.~G.~M., \& Barker, J.~R.\ 1989, \apjs, 71, 733
623: \bibitem[Cohen \& Barlow(2005)]{2005MNRAS.362.1199C} Cohen, M., \& Barlow, M.~J.\ 2005, \mnras, 362, 1199
624: \bibitem[Cohen et al.(1989)]{1989ApJ...341..246C} Cohen, M., Tielens, A.~G.~G.~M., Bregman, J., Witteborn, F.~C., Rank, D.~M., Allamandola, L.~J., Wooden, D., \& Jourdain de Muizon, M.\ 1989, \apj, 341, 246
625: \bibitem[De Marco \& Barlow(2001)]{2001Ap&SS.275...53D} De Marco, O., \& Barlow, M.~J.\ 2001, \apss, 275, 53
626: \bibitem[Geballe et al.(1985)]{1985ApJ...292..500G} Geballe, T.~R., Lacy, J.~H., Persson, S.~E., McGregor, P.~J., \& Soifer, B.~T.\ 1985, \apj, 292, 500
627: \bibitem[Gillett et al.(1975)]{1975ApJ...200..609G} Gillett, F.~C., Forrest, W.~J., Merrill, K.~M., Soifer, B.~T., \& Capps, R.~W.\ 1975, \apj, 200, 609
628: \bibitem[Gillett et al.(1973)]{1973ApJ...183...87G} Gillett, F.~C., Forrest, W.~J., \& Merrill, K.~M.\ 1973, \apj, 183, 87
629: \bibitem[Hony(2002)]{2002PhDT..........H} Hony, S.\ 2002, Ph.D.~Thesis
630: \bibitem[Jourdain de Muizon et al.(1990)]{1990A&AS...83..337J} Jourdain de Muizon, M., Cox, P., \& Lequeux, J.\ 1990, \aaps, 83, 337
631: \bibitem[Justtanont et al.(1996)]{1996ApJ...456..337J} Justtanont, K., Skinner, C.~J., Tielens, A.~G.~G.~M., Meixner, M., \& Baas, F.\ 1996, \apj, 456, 33
632: \bibitem[Kholtygin(1998)]{1998A&A...329..691K} Kholtygin, A.~F.\ 1998, \aap, 329, 691
633: \bibitem[Leger \& Puget(1984)]{1984A&A...137L...5L} Leger, A., \& Puget, J.~L.\ 1984, \aap, 137, L5
634: \bibitem[Liu et al.(2004)]{2004MNRAS.353.1251L} Liu, Y., Liu, X.-W., Barlow, M.~J., \& Luo, S.-G.\ 2004, \mnras, 353, 1251
635: \bibitem[Liu et al.(2001)]{2001MNRAS.323..343L} Liu, X.-W., et al.\ 2001,
636: \mnras, 323, 343
637: \bibitem[Lord(1992)]{1992nstc.rept.....L} Lord, S.~D.\ 1992, NASA TM103957
638: \bibitem[Mainzer \& McLean(2003)]{2003ApJ...597..555M} Mainzer, A.~K., \& McLean, I.~S.\ 2003, \apj, 597, 555
639: \bibitem[Mainzer et al.(2004)]{2004ApJ...604..832M} Mainzer, A.~K., McLean,
640: I.~S., Sievers, J.~L., \& Young, E.~T.\ 2004, \apj, 604, 832
641: \bibitem[Matsuura et al.(2005)]{2005A&A...434..691M} Matsuura, M., et al.\ 2005, \aap, 434, 691
642: \bibitem[McLean et al.(2006)]{2006SPIE.6269E.168M} McLean, I.~S., Smith, E.~C., Aliado, T., Brims, G., Kress, E., Magnone, K., Milburn, J., Oldag, A., Silvers, T., \& Skulason, G. \ 2006, \procspie, 6269.
643: \bibitem[Middlemass(1990)]{1990MNRAS.244..294M} Middlemass, D.\ 1990, \mnras, 244, 294
644: \bibitem[Peeters et al.(2002)]{2002A&A...390.1089P} Peeters, E., Hony, S., Van Kerckhoven, C., Tielens, A.~G.~G.~M., Allamandola, L.~J., Hudgins, D.~M., \& Bauschlicher, C.~W.\ 2002, \aap, 390, 1089
645: \bibitem[Pwa et al.(1986)]{1986A&A...164..184P} Pwa, T.~H., Pottasch, S.~R., \& Mo, J.~E.\ 1986, \aap, 164, 184
646: \bibitem[Rinehart et al.(2002)]{2002MNRAS.336...66R} Rinehart, S.~A., Houck, J.~R., Smith, J.~D., \& Wilson, J.~C.\ 2002, \mnras, 336, 66
647: \bibitem[Roche et al.(1996)]{1996MNRAS.280..924R} Roche, P.~F., Lucas, P.~W., Hoare, M.~G., Aitken, D.~K., \& Smith, C.~H.\ 1996, \mnras, 280, 924
648: \bibitem[Sellgren(1984)]{1984ApJ...277..623S} Sellgren, K.\ 1984, \apj, 277, 623
649: \bibitem[Sloan et al.(2007)]{2007ApJ...664.1144S} Sloan, G.~C., et al.\ 2007, \apj, 664, 1144
650: \bibitem[Smith \& McLean(2006)]{2006SPIE.6269E..50S} Smith, E.~C., \& McLean, I.~S.\ 2006, \procspie, 6269, 50
651: \bibitem[Tokunaga et al.(1991)]{1991ApJ...380..452T} Tokunaga, A.~T., Sellgren, K., Smith, R.~G., Nagata, T., Sakata, A., \& Nakada, Y.\ 1991, \apj, 380, 452
652: \bibitem[van Diedenhoven et al.(2004)]{2004ApJ...611..928V} van Diedenhoven, B., Peeters, E., Van Kerckhoven, C., Hony, S., Hudgins, D.~M., Allamandola, L.~J., \& Tielens, A.~G.~G.~M.\ 2004, \apj, 611, 928
653:
654: \end{thebibliography}
655:
656:
657:
658:
659:
660:
661:
662:
663:
664:
665:
666: \end{document}
667: