5. The diffuse emission
The diffuse emission is better studied from the filter maps because of their higher sensitivity. However the CVF observations are useful in the interpretation of the filter observations.
There is no reason to doubt that far from the emission peaks which coincide with concentrations of hot stars, most of the radiation at wavelengths shorter than about 9 µm is due to AIBs and their associated continuum. This is already clear for Peak G (see Fig. 5) which is far from the main far-UV sources even if it contains two 16th-magnitude hot stars. Consequently, we believe that the best view of the distribution of the AIBs is offered by the LW2 (5.0-8.0 µm) map which encompasses the 6.2 and 7.7 µm features (Fig. 5 and Fig. 6), although there is some contribution from Very Small Grains (VSGs: Désert et al. 1990; Dwek et al. 1997) in the peaks where the radiation field is very high (see Fig. 6 and Cesarsky et al. 1996b). The stellar contribution in this filter is limited to that of a few red stars identified on Fig. 5, and perhaps to the emission of circumstellar dust around hot stars as discussed in the previous section. The LW6 (7.0-8.5 µm) and LW7 (8.5-10.7 µm) maps (Fig. 8 and Fig. 9) and the LW4 (5.5-6.5 µm) map (not shown) are very similar to each other and to the LW2 map, although the NE extension and some stars are more easily visible on the LW2 map which is more sensitive due to the broader passband of this filter.
The filter maps which include AIBs at longer wavelengths, e.g. the LW8 (10.7-12.0 µm, not shown) and LW10 (IRAS filter: 8.0-15.0 µm, Fig. 16) maps, are more difficult to interpret because they contain a contribution of both AIBs and VSGs.
A particularly interesting feature in the LW2 (5.0-8.0 µm) and LW6 (7.0-8.5 µm) maps is the emission spur that extends to the NE of N 66A. This spur is probably dominated by AIB emission. It is barely visible in filters like LW3 (12.0-18.0 µm) in which the contribution of AIBs is minor (see Fig. 10). Fig. 12 shows a superposition of the CO(2-1) line emission in the region of N 66 over the LW2 map. The CO emission coincides very well with the spur of AIB emission. As discussed above, this can be easily explained by emission from the surface of the molecular cloud bathed by a lower and softer radiation field than in the bar of N 66.
Fig. 10 is the LW3 (12.0-18.0 µm) map of the N 66 region. Although there is some contribution from the [Ne III ] 15.6 µm line and of the [Ne II ] line and AIB at 12.7 µm in the LW3 filter, our CVF spectra show that it can generally be neglected with respect to the continuum. This is shown by Fig. 11 on which the CVF image in the continuum on each side of the [Ne III ] 15.6 µm line (contours) is superimposed on the LW3 image (grey scale): the agreement is very good given the differences in field of view and sensitivity. Thus the LW3 map in our case represents adequately the emission of the Very Small Grains (VSGs). It is noteworthy that the distribution in the LW3 map is more extended around the "bar" than the LW2 map although the latter is more sensitive (compare Fig. 10 with Fig. 5). This has rarely been seen before and may indicate VSG emission in regions where the AIB carriers have been partly destroyed.
Fig. 13 presents the "color" map of the LW3(12.0-18.0 µm)/LW2 (5.0-8.0 µm) intensity ratio. For building this map, the LW3 data have been convolved with the LW2 PSF as measured on the LW2 map, and vice-versa before division; this resulted in a small loss of resolving power but produced approximately similar PSFs after convolution. Then only the part of the data with a signal to noise ratio larger than 2 after convolution has been retained in both filters.
Previous observations with ISO (e.g. Cesarsky et al. 1996b, Contursi et al. 1998) have shown that the VSGs start to emit appreciably near 15 µm when the ultraviolet radiation field is a few 103 times the LISRF. Under these conditions the VSGs temperatures are high enough for their spectrum to shift towards short wavelengths increasing the 15/6.75 µm ratio, This ratio ranges from 0.5 to 0.8 in the LISRF environnement. The FUV values obtained in N66 indicate that the ISRF intensity is well above 103 times the LISRF everywhere in the observed region except in the region of the molecular cloud. In order to study how the 15/6.75 µm color ratios relate to the UV ISRF we have evaluated the 15/6.75 µm ratio of each peak over regions of the same size (radius = 2.8 pc), and plotted them as a function of the ISRF at 1600 Å integrated over the same regions (Fig. 14). As expected, the general trend is that the higher the ISRF, the higher is the 15/6.75 µm ratio. One can see that the spur (peak G) has a typical 15/6.75 µm "cirrus" value of 1. The same effect is observed for the global IR emission properties of galaxies. The 15/6.75 µm - 60/100 µm color-color diagram shows that the global mid-IR (15/6.75 µm) colors are 1 for normal galaxies ("cirrus" value) and become significantly greater than 1 for more active galaxies (Vigroux et al. 1998, Dale et al. 2000). The same behavior is also observed inside three nearby galaxies, IC 10, NGC 1313 and NGC 6946 (Dale et al. 1999).
Surprisingly, the highest value of the the 15/6.75 µm ratio in Fig. 14 does not correspond to the highest value of the ISRF, located at the center of the star cluster (peak C). The CVF spectrum of the region with the largest value of the 15/6.75 µm ratio is shown on Fig. 15. Following the interpretation of Cesarsky et al. (1996b) and Contursi et al. (1998) we would expect a continuum towards 15 µm steeper than that observed in Peak C. Fig. 15 shows that this is not the case. The high 15/6.75 µm value observed is due to the nearly complete absence of AIB carriers and of continuum at short wavelengths (which is instead present in peak C). This dramatically lowers the flux in the LW2 filter. The LW2 and LW3 fluxes of this region are respectively 8 and 4 times smaller than those of peak C in the same filters. The contribution of the [Ne III ] emission line in the LW3 filter is only 10. We remark that this region is close to the earliest-type star of N66 (OIII(f*)) suggesting that here the ISRF is not only very strong but also very hard. This results in a complete destruction of the AIB carriers and partially also of the smallest VSGs. The VSGs might also be destroyed in the other peaks, although to a lesser degree. This might explain why even if the ISRF throughout the N 66 region is least 102 times that of the H II region N 4 in the LMC, the 15/6.75 µm ratios are similar to those found in N 4 (Contursi et al., 1998).
Finally, we show on Fig. 16 a LW10 (8.0-15.0 µm = IRAS 12 µm) filter map superimposed on the DSS image: comparison with Fig. 5 and Fig. 10 demonstrates that this image contains features of both maps at 6.75 and 15 µm although it is closer to the 6.75 µm map. While interesting for comparison with IRAS data, the LW10 image is more difficult to interpret than the images in some other filters which have been presented here.
© European Southern Observatory (ESO) 2000
Online publication: October 30, 19100