THE 15–20 μm EMISSION IN THE REFLECTION NEBULA NGC 2023

The observations were taken with the IRS (Houck et al. 2004) on board the Spitzer Space Telescope (Werner et al. 2004a) and were part of the Open Time Observations PIDs 20097 and 50511.

We obtained spectral maps for two positions in the reflection nebula NGC 2023 (see Figure 1): toward the dense shell 78'' south of the exciting star HD 37903 corresponding to the H₂ emission peak, and toward a region to the north (+33'', +105'') of the exciting star (Burton et al. 1998). The north position is characterized by a much lower density of ∼10⁴ cm⁻³ than the density at the south position, which is >10⁵ cm⁻³ (Burton et al. 1998; Sheffer et al. 2011). Likewise, the north position has a slightly lower UV radiation field between 6 and 13.6 eV, i.e., ∼500 G₀,⁶ as compared to ∼10⁴ G₀ at the southern H₂ peak (Burton et al. 1998; Sheffer et al. 2011).

The spectral maps are made with the short-high (SH) mode. The SH mode covers a wavelength range from 10 to 20 μm at a resolution of ∼600 and has a pixel size of 23. We obtained off-source background observations in SH for the north position. Table 1 gives a detailed overview of the observations. We also obtained spectral maps made with the short-low (SL) mode; the combined SL and SH data set will be presented in E. Peeters et al. (in preparation).

Table 1. Observation Log

	North Position	South Position
PID	50511	20097
AOR	26337024	14033920
Cycles × Ramp time	2 × 30 s	1 × 30 s
Pointings ∥	7	12
Step size ∥	5''	565
Pointings ⊥	15	12
Step size ⊥	23	47
Backgrounda	5:42:1.00, −2:6:54.5

Note. ^aα, δ (J2000); units of α are hours, minutes, and seconds, and units of δ are degrees, arc minutes, and arc seconds.

Download table as:  ASCIITypeset image

The raw data were processed with the S18.7 pipeline version by the Spitzer Science Center. The resulting bcd-products are further processed using cubism (Smith et al. 2007a) and irsclean available from the SSC Web site.⁷

The SH off-source background observation exhibits a rising dust continuum and an 11.2 μm PAH band. Since we do not have off-source background observations for the south position, we investigated the importance of the background flux and, in particular, of the background PAH emission with respect to the on-source (PAH) flux. The 11.2 PAH flux in the SH background for the northern position is 8.5 × 10⁻⁸ W m⁻² sr⁻¹. This amounts to 3%–11% of the on-source 11.2 PAH flux across the north map and to <1%–10% of the on-source 11.2 PAH flux across the south map excluding source D from Sellgren (1983). Source D is located on the edge of the south map and has a background contribution to the PAH flux of up to 45%. These results are consistent with the IRAC [8 μm] image which is dominated by PAH emission and indicates a background flux of ∼6% the minimum on-source flux. Hence, aside from source D, we conclude that the background has only a small contribution to the on-source PAH flux and therefore we did not apply a background subtraction. For the remainder of this paper, we therefore excluded source D from the PAH analysis.

Since no background subtraction was applied, we used the tool irsclean to correct for the presence of rogue pixels. These rogue pixels are particularly problematic for the high-resolution module SH. In addition, we applied cubism's automatic bad pixel generation with σ_TRIM = 7 and Minbad-fraction = 0.50 and 0.75 for the global bad pixels and record bad pixels, respectively. Remaining bad pixels were subsequently removed manually.

Spectra were extracted from the spectral maps by moving, in one pixel steps, a spectral aperture of 2 × 2 pixels in both directions of the maps. This results in overlapping extraction apertures. No jumps in flux level were seen between the different orders within the SH module (SH11 to SH20) and hence no scaling was necessary.

A typical 15–20 μm spectrum of NGC 2023 reveals a rising dust continuum, a sharp H₂ emission line (17 μm), an emission band at 19 μm attributed to C₆₀ (Cami et al. 2010; Sellgren et al. 2010), a feature at 17.4 μm which is a blend of a PAH band with another C₆₀ band, and a plethora of PAH emission bands at 15.8, 16.4, 17.4, and 17.8 μm on top of a broad underlying emission component (Figure 2).

Zoom In Zoom Out Reset image size

NGC 2023 contains a cluster of young stars (Sellgren 1983). The young stellar object (YSO) source D is located just outside the field of view (FOV) of the south map (see Figure 1). Nevertheless, the spectra of regions close to source D show PAH emission features as well as characteristics typical of YSOs, i.e., a strong dust continuum (see the continuum map at 19.29 μm in Figure 3) and a (strong) CO₂ ice feature near 15 μm. Furthermore, the background contribution to the 11.2 PAH flux is significant for a large fraction of these (ice-)spectra. We therefore excluded these spectra in the PAH and the 14.9 μm continuum analysis. The 15–20 μm spectrum of the YSO source C has the same spectral characteristics as the spectra across NGC 2023 but has an enhanced surface brightness (see pixels ∼(21, 6) in the continuum maps at 14.9 and 19.29 μm in Figure 3). However, we did not exclude these spectra in the analysis.

Zoom In Zoom Out Reset image size

We determined the fluxes of the PAH features in the 15–20 μm region by subtracting a local spline continuum shown as a green line in the top frame in Figure 2. This continuum is determined by using anchor points at 14.91, 15.20, 15.50, 16.14, 16.69, 16.89, 17.16/17.19 (for the south and north maps, respectively), 17.58, 18.14/18.17 (for the north and south maps, respectively), 18.50, 19.29, and 19.36 μm. The fluxes are then estimated by fitting a Gaussian profile to the individual emission bands. A second global continuum (red line in the top frame) is determined by using anchor points at 14.91, 18.50, and 19.36 μm. In this way, a plateau underneath the individual bands is defined by the difference of both continua and is further referred to as the 15–18 μm plateau. This plateau is distinct from the 15–20 μm plateau observed toward H ii regions (Van Kerckhoven et al. 2000) which exhibits emission over the entire 15–20 μm range. This decomposition method is further referred to as the spline decomposition.

The applied decomposition of the 15–20 μm region is not unique. For example, Sellgren et al. (2007) fit this region by a combination of a continuum fit (similar to our global continuum) and eight Gaussians (further referred to as the Gaussian decomposition). Similarly, a combination of Drude profiles is regularly used (PAHFIT; Smith et al. 2007b). For comparison, we applied both methods to our spectra (Figure 2). To provide sufficient wavelength coverage for PAHFIT, we obtained the SL spectrum for the same extraction aperture as the SH spectrum. However, we did not correct for the different point-spread functions at these wavelengths. The PAHFIT decomposition in the 15–20 μm region results in components representing the dust continuum emission (which is a combination of modified blackbodies; see Smith et al. 2007b), the H₂ emission, and the PAH dust features including a broad component centered at 17.0 μm. We also applied a similar decomposition as Sellgren et al. (2007, their Figure 8 and Table 1) to our spectra by using their Gaussian parameters as a starting value. This method results in two possible decompositions of the 15–20 μm region: one dominated by a broad component peaking around ∼17.2 μm (Figure 2, third panel) or the other in which this 17.2 μm component is much narrower resulting in broader and stronger components near 16.6 and 17.4 μm (Figure 2, fourth panel). Hence, in order to study the PAH bands in a consistent way, one needs to restrict the decomposition to a broad or narrow 17.2 μm component. But choosing one over the other will clearly influence the derived fluxes of the Gaussian components.

The Gaussian decomposition with the broad 17.2 μm component results in very similar fluxes for the features as those obtained with the spline decomposition discussed above, except for the 15.8 and 16.4 μm bands (Figure 2, bottom panel). In contrast, different PAH band intensities are obtained with the Gaussian decomposition with the narrow 17.2 μm component and the PAHFIT decomposition. In particular, the broad component centered around 17 μm depends significantly on the decomposition method. Here we will adopt the spline decomposition (Figure 2, top panel).

Maps. Figures 3 and 4 show the spatial distribution of the various emission components in the 15–20 μm region. In these figures, in addition to the color-coded maps for the different features, we show contours for the 11.2 and 12.7 μm PAH bands for reference. These were derived from the SH data and are discussed in E. Peeters et al. (in preparation). The north map is characterized by lower flux levels and therefore larger scatter is present in the map tracing the 15.8 and 17.8 μm features and, to a lesser extent, the 18.9 μm band. Hence, when discussing the spatial distribution of the 15.8 and 17.8 μm bands, we restrict ourselves to the south map.

Zoom In Zoom Out Reset image size

Feature correlations. To investigate possible variations within the PAH spectrum itself, we need to exclude the influence of PAH abundance and column density. This is achieved by normalizing the fluxes to another band in this region. Here we start with the 11.2 μm PAH band, which is attributed to solo CH out-of-plane bending vibration of large, neutral PAHs (Hony et al. 2001; Bauschlicher et al. 2008, 2009). To look for possible correlations with the 11.2 μm band, we also apply a normalization to the 12.7 μm PAH band which is attributed to duo and trio CH out-of-plane bending vibration of large PAHs (Hony et al. 2001; Bauschlicher et al. 2008, 2009). Both the 11.2 and 12.7 μm PAH fluxes are obtained from the SH data using a spline continuum (see, e.g., Figure 1 of Hony et al. 2001) and are discussed in E. Peeters et al. (in preparation). A correlation is often found between the normalized intensity ratios of PAH bands in this region as these bands arise from a family of PAH molecules which all have the same basic structural unit. However, variations in charge states, excitation level, edge structures, etc., exist within the PAH family which can tighten as well as loosen correlations between the normalized intensity ratios of specific PAH bands. For example, the well-known tight correlation between the 6.2 and 7.7 μm PAH bands is attributed to the common charge state of these bands (e.g., Hony et al. 2001; Galliano et al. 2008). Figures 5 and 6 show some of the intensity correlations found in NGC 2023. The correlation coefficients and the fit parameters are given in Table 2. While maps smooth out the detailed, fine grained variations, allowing one to see the big picture, these correlations highlight subtle differences.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Results. The following trends are derived from the south spectral maps shown in Figure 3. The 11.2 μm feature, the 15.8 μm feature, and the 15–18 μm plateau show very similar spatial morphology with distinct peaks at the S and SSE positions in the IRAC map. In contrast, the distributions of the 12.7, 16.4, 17.4, and 17.8 μm features are displaced toward the star. The 12.7 and 16.4 μm bands have very similar spatial behavior; both peak at the SSE position and show a broad, diffuse plateau NW of the line connecting the S and SSE ridges. The 17.4 μm feature also presents a strong, diffuse plateau NW of that line, but does not show a peak at the SSE position as does the 12.7 and 16.4 μm bands. In addition, the 17.4 μm feature has a separate peak in the upper right of the field, coinciding with the peak in the 18.9 μm band. The 17.8 μm band peaks primarily on the SSE position, overlapping with the peaks of both the 11.2 and 12.7 μm bands but also shows some enhanced emission slightly offset from the 11.2 μm peak near the S ridge, though not as much as the 12.7 and 16.4 μm bands. The 18.9 μm band is clearly in a class by itself, peaking closer to the illuminating star than any of the other features considered here. Interestingly, a second peak is evident in the 18.9 μm spatial distribution at source C. Note also that the 18.9 μm band does show (weak) emission tracing the S and SSE ridges. The spatial behavior of the 14.9 and 19.29 μm continuum intensities align closely with the H₂ emission distribution, all lying along the SE portion of the 11.2 μm contours, while clearly avoiding regions with strong 12.7 μm emission. The H₂ map shows, however, additional emission toward the west.

The similarity in distribution of the 11.2 μm and 15–18 μm plateau is also evident in the north map (Figure 4): both maps peak on the NW ridge and extend to include the N ridge. There is too much scatter in the weak 15.8 μm PAH band to warrant comparison. In addition, the 16.4 μm PAH band has a spatial distribution very similar to that of the 12.7 μm band. However, unlike the situation at the SSE ridge where the 11.2, 12.7, and 16.4 μm bands all peak in the same place, the 12.7 and 16.4 μm bands are clearly displaced toward the south with respect to the 11.2 μm emission in both the peak emission and the N ridge extension. The 17.4 and 18.9 μm bands exhibit more scatter in the north map. Nevertheless, the 17.4 μm band seems to peak toward the NW ridge and further west of this ridge. In contrast, the 18.9 μm band seems to trace the N ridge and not the NW ridge while also showing secondary emission in the south. The spatial behavior of the 14.9 and 19.29 μm continuum emission and the H₂ emissions are quite distinct from the PAH emission in the north map as they do not track the 11.2 μm contours as closely and only trace the N ridge and not the NW ridge.

This morphology suggests that the features can be viewed as falling into the following groups: (1) the 11.2, 15.8 μm bands and the 15–18 μm plateau; (2) the 12.7 and 16.4 μm bands; (3) the 17.4 μm diffuse component; (4) the 17.8 μm band; (5) the 18.9 and (some of) the 17.4 μm features; (6) the underlying continuum producing the emission measured here at 14.9 and 19.2 μm; and (7) the H₂ emission.

The correlation plots shown in Figure 5 support grouping the 11.2, 15.8 μm bands and the 15–18 μm plateau into a single group. Indeed, a very tight correlation is found in our data set between the 11.2/12.7 PAH intensity ratio and the 15–18 μm plateau/12.7 PAH intensity ratio. Weaker trends, although still very tight, are seen between the 11.2/12.7 and 15.8/12.7 PAH intensity ratios and the 15.8/12.7 and 15–18 μm plateau/12.7 PAH intensity ratios. In addition, the normalized 16.4 μm PAH intensity clearly correlates with that of the 12.7 μm PAH band; their correlation coefficient is the highest found in our data, i.e., 0.889.

As shown in Figure 6, the 17.4 and 18.9 μm bands do not exhibit an overriding linear correlation with each other as do the bands in Figure 5. Instead, the 18.9 μm band does not correlate with the 17.4 μm band for low 18.9/11.2 intensity ratios (<0.02) and shows a linear correlation with the 17.4 μm band for strong band ratios. This is consistent with the assignment of the 18.9 μm band to C₆₀ and the 17.4 μm feature to being a blend of a PAH emission band with a C₆₀ band (Cami et al. 2010; Sellgren et al. 2010). Therefore, the 17.4 μm band was decomposed into these two components to estimate the spatial distribution of the PAH contribution to this band. This is referred to as the diffuse 17.4 μm band. We calculated the average ratio of the 18.9/17.4 band intensity in the upper right corner of the south map (where the 18.9 μm band peaks) and used this value to subtract the C₆₀ contribution to the 17.4 μm band to obtain the "diffuse 17.4 μm band" for both the south and north maps (see Figures 3 and 4). Note that this method assumes a constant ratio between the 17.4 and 18.9 μm C₆₀ emission (of 0.6, similar to what is observed, e.g., in Tc1; Cami et al. 2010) and only provides a first-order approximation to the diffuse 17.4 μm band distribution. Nevertheless, this lends support to our earlier suggestion that the PAH component in the 17.4 μm feature does not spatially coincide with the 11.2 or 12.7 μm PAH emission.

We investigate the band profiles in the spectrum from the total FOV of the south map to minimize the noise and apply the same continuum determination as discussed in Section 3.4. Within the uncertainty, the 15.8, 17.4, 17.8, and 18.9 μm bands are well fitted with a Gaussian. Hence, no shift is found in the central wavelength of the 17.4 μm feature as the 17.4/18.9 intensity ratio varies. In contrast, the 16.4 μm PAH feature is slightly asymmetric to the red (Figure 7). To assess the influence of the continuum on the degree of asymmetry, three different continua were constructed and the resulting 16.4 μm bands compared. First, the continuum point near 16.69 μm may be considered to be part of the red wing of the 16.4 μm band. Hence, it is omitted and a spline is fitted to the remaining continuum points. The other two ways involve fitting a straight line to continuum points at 16.14 μm and 16.69 or 16.89 μm, respectively. There is little difference between the spline fit and the straight line but the omission of the continuum point at 16.69 μm clearly amplifies the red wing of the band and thus the asymmetry (Figure 7). Comparison of the 16.4 μm band profile with the well-known asymmetric 6.2 and 11.2 μm PAH bands shows that its asymmetry is less pronounced (Figure 7).

Zoom In Zoom Out Reset image size

In order to investigate possible variations in the profile of the 15–18 μm plateau, we compared the 15–18 μm plateau of each pixel with the average 15–18 μm plateau (Figure 8). Overall, we conclude that the profile of the 15–18 μm plateau does not vary, despite a possible change in intensity of the PAH bands located on top of this plateau. Only in a handful of cases do we see slight deviation from the average. Hence, the 15–18 μm plateau profile does not depend on the individual PAH bands as determined in this paper.

Zoom In Zoom Out Reset image size

In a previous detailed study of the PAH emission features in the 15–20 μm region, Boersma et al. (2010) focused on the spectra from very different astronomical environments, not the spatial behavior in an extended object as considered here. Analyzing the spectra from some 10 objects, Boersma et al. (2010) showed that the 16.4 μm band correlated with the mid-IR 6.2 and 7.6/7.8 μm bands and that the total 15–20 μm PAH emission correlated with the 11.2 μm band. No further connections were found between the 15.8, 16.4, 17.4, 17.8, and 18.9 μm bands or between the spectral characteristics of the 15–20 μm emission and the mid-IR PAH modes. These authors then conclude that these bands must be carried by independent molecular species or classes of species. The results presented in the previous section both support and amplify these conclusions. The strong correlation of the 16.4 and the 6.2 μm bands found by Boersma et al. (2010) is confirmed by our results as both the 16.4 and the 12.7 μm PAH bands strongly correlate with each other and it is well known that the 12.7 μm correlates well with the 6.2 μm band (Hony et al. 2001; E. Peeters et al., in preparation). In contrast with Boersma et al. (2010), we find a correlation between the 15.8 and 11.2 μm bands. The 15.8 μm band is very weak and only detected in a few sources in their sample. These sources follow the correlation found in this paper.

As discussed in Section 3.4, different spectral decompositions have been applied in the literature. This is in particularly true for the treatment of the broad underlying plateaus. They are either considered to be part of the PAH bands themselves which are then fitted by Lorentzians or Drude profiles (e.g., Boulanger et al. 1998; Smith et al. 2007b), or they are considered to be independent of these PAH bands (e.g., Hony et al. 2001; Peeters et al. 2002; this paper). The spatial maps and the band correlations clearly indicate that there are significant differences in the spatial distribution between the 15–18 μm plateau and most PAH bands. These spatial differences suggest a different carrier for the plateau and the individual PAH bands (with the exclusion of the 15.8 μm band), a finding consistent with earlier suggestions for the plateaus underlying the prominent mid-IR bands between 5 and 13 μm (Bregman et al. 1989; Roche et al. 1989).

The emission features between 15 and 20 μm can be collected into seven groups as discussed in Section 4.1. Here we discuss these in terms of the specific characteristics that set each group apart and make them particularly sensitive to the local physical conditions.

Group 1: the 11.2, 15.8 μm bands and the 15–18 μm plateau. Since the 11.2 μm band is associated with the solo CH out-of-plane vibration of large, neutral PAHs (e.g., Allamandola et al. 1999; Hony et al. 2001; Galliano et al. 2008; Bauschlicher et al. 2008, 2009), the 15.8 μm band and the 15–18 μm plateau must arise from species that are favored by the same conditions that favor this subset of the PAH population. In contrast with expectations, the 15–20 μm region shows little systematic dependence on PAH class or molecular structure as present in the NASA Ames PAH IR spectroscopic database (Bauschlicher et al. 2010; Boersma et al. 2010; Ricca et al. 2010), suggesting that molecular structure is not the main origin for the observed correlations. Keep in mind that, to date, the database is biased to smaller PAHs and the astronomical emission at these longer wavelengths is dominated by larger PAHs. Thus, as knowledge of larger PAH spectroscopy increases, this suggestion should be revisited. In addition, ionization only has a moderate influence on the PAH emission in the 15–20 μm region (Van Kerckhoven et al. 2000; Boersma et al. 2010): it slightly influences the relative strength of the emission bands but not their peak position. Based on theoretical and experimental PAH spectra, one can therefore not assign a particular charge state to these individual bands, unlike for the mid-IR bands. Therefore, we suggest that large, neutral PAHs are also responsible for the 15.8 μm band and 15–18 μm plateau. PAH clusters, VSGs (very small grains) or small hydrogenated amorphous carbon (HAC) particles may also contribute to the 15–18 μm plateau. However, the spread of the correlations in Figure 5 shows there is not a one-to-one correspondence between the carriers of these different emission features. These differences reflect minor changes in the relative populations of the different band carriers, perhaps arising from slightly different molecular structures, excitation levels, and so on.

Group 2: the 12.7 and 16.4 μm bands. Similarly, since the 16.4 and 12.7 μm bands correlate with the 6.2 and 7.7 μm PAH cation bands (Hony et al. 2001; Boersma et al. 2010; E. Peeters et al., in preparation), the 12.7 and 16.4 μm bands must arise from species that are cospatial with the cationic portion of the emitting PAH population. Again, as with Group 1, the spread of the correlations in Figure 5 shows there is not a one-to-one correspondence between these different carriers. However, in contrast with Group 1, we cannot propose molecular structure or charge as the main origin for these bands. Indeed, laboratory and theoretical PAH spectra indicate that the 12.7 and 16.4 μm bands are sensitive to the molecular edge structure: the former being due to duo and trio CH groups (Hony et al. 2001; Bauschlicher et al. 2008, 2009) and the latter being systematically present in PAHs with pendent rings (Moutou et al. 1999; Van Kerckhoven et al. 2000; Peeters et al. 2004; Boersma et al. 2010) and large PAHs with pointy edges (Ricca et al. 2010). Laboratory and theoretical PAH spectra also indicate that both neutral and ionized PAHs can contribute to the 12.7 and 16.4 μm emission (Van Kerckhoven et al. 2000; Hony et al. 2001; Bauschlicher et al. 2008, 2009; Boersma et al. 2010) and hence do not allow, by themselves, the assignment of these bands to a single charge state. Similarly, blind signal separation of the PAH emission across PDRs cannot assign a single charge state to the 12.7 μm band (the 16.4 μm band is not analyzed by this method). Indeed, this method revealed three spatially different components, referred to as PAH⁰, PAH⁺, and VSG, which are attributed to, respectively, neutral PAHs, ionized PAHs, and PAH clusters (e.g., Rapacioli et al. 2005; Berné et al. 2007; Rosenberg et al. 2011). The 12.7 μm band is present in both the PAH⁰ and PAH⁺ components. However, the relative contribution of the 12.7 μm band in the PAH⁰ and PAH⁺ components seems to vary for different decomposition methods and sources (Rapacioli et al. 2005; Berné et al. 2007; Joblin et al. 2008). Hence, further investigation is warranted.

Group 3: the 17.4 μm diffuse component. This paper confirms the results by Sellgren et al. (2010) that the 17.4 μm emission band is composed of emission due to C₆₀ and PAHs (i.e., the diffuse 17.4 μm component). In addition, Sellgren et al. (2010) report a similar spatial distribution between the diffuse 17.4 μm emission and the 16.4 μm PAH band in the reflection nebula NGC 7023. In contrast, in NGC 2023, the spatial distribution of the diffuse 17.4 μm emission is quite distinct from that of the 16.4 μm band. In particular, in the north map, the diffuse 17.4 μm emission is strong close to the peak 16.4 μm emission as well as west of the NW ridge. In the south map, it is located closer toward the illuminating star compared to the 11.2, 12.7, and 16.4 μm PAH emission and does not peak in the SSE ridge where the 11.2, 12.7, and 16.4 μm PAH bands peak. Perusal of the NASA Ames PAH IR spectroscopic database (Bauschlicher et al. 2010) reveals that large compact PAHs systematically show emission near 17.4 μm and that PAH charge only has moderate influence on the relative intensities and no influence on the peak position of the bands in the 15–20 μm region (Boersma et al. 2010). We therefore conclude that the diffuse 17.4 μm band is due to doubly ionized PAHs and/or a subset of PAH cations that favor more harsh conditions compared to those responsible for the 6.2 μm PAH bands, such as, for example, dehydrogenated cations. This possible assignment will be explored further by comparison to the PAH emission at shorter wavelengths in a forthcoming publication (E. Peeters et al., in preparation).

Group 4: the 17.8 μm band. The 17.8 μm emission exhibits spatial characteristics between those of the 11.2 and 12.7 μm bands. With the former spatial distribution being attributed to neutral PAHs and the latter being cospatial with that of the PAH cations, the 17.8 μm band likely originates from both neutrals and cations (with both having similar intrinsic strengths for this band).

Group 5: the 18.9 and (some of) the 17.4 μm feature. These two bands have recently been assigned to C₆₀ (Cami et al. 2010; Sellgren et al. 2010). The C₆₀ bands show a distinct spatial distribution. In the south, they reach their peak emission intensities in regions closer to the exciting star than are the PAH bands, similar to what is observed toward NGC7023 (Sellgren et al. 2010). In addition, secondary peaks in emission are found near source C, a YSO, and where the PAH emission peaks. In contrast, in the north, the peak in the C₆₀ distribution seems to be farther away from the illuminating source (no other UV sources are present near this position) and linked to the H/H₂ transition zone. This different behavior in the C₆₀ emission may well provide important clues to the formation and evolution of C₆₀ in the ISM. This will be further explored when the full PAH spectra are analyzed (E. Peeters et al., in preparation).

Group 6: the continuum producing the emission at 14.9 and 19.2 μm. This emission component peaks further into the filament with respect to the PAH emission. This is consistent with the results of the blind signal separation discussed above: the VSG component, attributed to PAH clusters, carries the mid-IR continuum (e.g., Rapacioli et al. 2005; Berné et al. 2007). As it is located deeper in the PDR compared to the PAH emission, these authors suggested the destruction of PAH clusters by UV photons and subsequent formation of PAHs. As mentioned above, PAH clusters may also contribute to the 15–18 μm plateau, which is slightly offset compared to the dust continuum emission. This suggests that the carriers of these emission components differ slightly.

Group 7: the H₂ emission. For the analysis of the H₂ emission present in this data set, we refer to Fleming et al. (2010) and Sheffer et al. (2011).