Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 325, 725-744 (1997)

Previous Section Next Section Title Page Table of Contents

3. Results

Here we present the results of the line and continuum observations towards IRAS 20126+4104. We shall illustrate the single dish and interferometer results in separate sections. In Appendix 7 we give the results of gaussian fits to the lines seen with the 30-m telescope and PdBI and the parameters of the NIR continuum and H2 line emission measured with the TIRGO and NOT telescopes.

3.1. Molecular line data from the 30-m telescope

In Figs. 1 and 2 we show the spectra towards the centre position of our maps (i.e. the position of the H2 O maser) for the most relevant transitions observed. As illustrated below, this is also the position of the maximum emission in the maps. All lines peak at [FORMULA]  km s-1, but the full width at half maximum (FWHM) of the CH3 CN, CH3 OH, and C34 S(5-4) lines is about twice as much as that of the other transitions. Furthermore, the lower density tracers like 13 CO or HCO [FORMULA] present broad wings extending up to [FORMULA]  km s-1 from the peak. Even the higher density tracer C34 S, despite the much poorer signal-to-noise ratio (S/N), shows strong wings, although only up to [FORMULA]  km s-1 from the peak. On this basis, we have tentatively identified five velocity ranges (indicated by the dashed lines in Fig. 1):

[FIGURE] Fig. 1. Spectra of some rotational transitions observed with the 30-m telescope towards the H2 O maser position (i.e. the map centre [FORMULA], [FORMULA]) in IRAS 20126+4104. The main beam brightness temperature ([FORMULA]) is plotted against the local standard of rest (LSR) velocity. The conversion factor from [FORMULA] to flux density is 4.7 Jy K-1. The dotted horizontal and vertical lines correspond respectively to [FORMULA] =0 K and [FORMULA] =-3.5 km s-1. The dashed vertical lines indicate the values of [FORMULA] separating the bulk emission from the inner wings and the inner wings from the outer wings. The thick vertical lines in the bottom panel indicate the F =0-1 (left), 2-1 (centre), and 1-1 (right) hyperfine components of the HCN(1-0) transition

[FIGURE] Fig. 2. Same as Fig. 1, but with [FORMULA] versus frequency for an LSR velocity of -3.5 km s-1. The dotted lines indicate the CH3 CN (lines above the spectrum) and C [FORMULA] CN (from below), and the CH3 OH transitions; the corresponding quantum numbers are also shown. Also the CH3 CH3 O 717 -606 line at 147024.8 MHz, the HNCO 1019 -918 line at 220584.766 MHz, and the C3 H2 312 -221 line at 145089.63 MHz can be seen in the spectra

  • the bulk emission (from -5 to -2 km s-1), which corresponds to the narrow component of the C34 S(3-2) line;
  • the inner wings (from -9 to -5 km s-1 and from -2 to 2 km s-1), also clearly visible in the C34 S(3-2) line;
  • the outer wings (from -28 to -9 km s-1 and from 2 to 21 km s-1), seen in the HCO [FORMULA] (1-0), 13 CO(2-1), HCN(1-0), and CS(3-2) lines. Note that in the case of HCN the identification of the line wings is difficult because of the hyperfine components (shown in Fig. 1).

We now investigate the morphology of the regions emitting in the different velocity ranges.

In Table 3 we give the measured full width at half power (FWHP) of the maps obtained by integrating the line emission in the velocity range corresponding to the bulk emission, i.e. from -5 to -2 km s-1. The FWHP has been obtained as the diameter of the circle having area equal to that inside the 50% contour. The real angular diameter of the source, [FORMULA], computed by gaussian deconvolution from the HPBW is also listed.


Table 3. Measured (FWHP) and deconvolved ([FORMULA]) angular diameters derived from the integrated emission in the velocity range from -5 to -2 km s-1 for different lines. Also the HPBW at the frequency of the observed line is listed. In all cases the integrated emission peaks at the map centre, i.e. at the position of the H2 O masers

The bulk emission is best studied through the high density tracers. We thus show in Fig. 3 maps in the CH3 OH(3-2), CH3 CN(8-7), and C34 S(3-2) transitions; note that the 2 mm transitions have been preferred to the 1.3 mm ones because of the better S/N, even though the corresponding angular resolution is worse ([FORMULA] instead of [FORMULA]). In all cases the emission originates from the same region with size [FORMULA] (see Table 3), centred on the H2 O maser spots: this is a clear indication of the existence of a molecular clump in which the IRAS source and the H2 O masers are embedded; also, the fact that different lines trace the same clump demonstrates that the clump is real and not an artifact of chemical differentiation in the molecular cloud.

[FIGURE] Fig. 3. Pico Veleta maps of the integrated bulk emission in the CH3 OH(3-2) (top panel), CH3 CN(8-7) (middle), and C34 S(3-2) (bottom) transitions. Thick contours correspond to 50% of the maximum in each map. Other contours range from 3 to 24 by 3 K km s-1 for CH3 OH, from 5 to 13 by 2 K km s-1 for CH3 CN, and from 1.1 to 3.9 by 0.7 K km s-1 for C34 S. The HPBW is [FORMULA]. The filled triangles indicate the positions of the H2 O maser spots

In order to map the line wings, one has to use the lines with the best S/N. In the top panels of Fig. 4 we show an overlay of the CS(3-2) and HCO [FORMULA] (1-0) emission in the range from -9 to 2 km s-1 (i.e. the integrated inner wings and bulk emission) and the integrals under the outer wings of the line. In the bottom panels the same comparison is done between the bulk emission and the integral under the inner wings. The outer wings seem to trace a NW-SE velocity shift, with the blue- and red-shifted emission coming respectively from NW and SE: The surprising result is that the blue- and red-shifted emission in the inner wings shows an orientation completely reversed with respect to the NW-SE axis above: although the direction is the same, the velocity increases from SE to NW.

[FIGURE] Fig. 4. Pico Veleta maps of the CS(3-2) (left panels) and HCO [FORMULA] (1-0) (right panels) lines. Top panels: overlay of the integrated emission (grey scale) in the range from -9 to 2 km s-1 and of the blue- (continuous contours) and red-shifted (dashed contours) emission in the outer wings, i.e. respectively from -28 to -9 km s-1 and from 2 to 21 km s-1. Contour levels range from 2 to 11 by 1.5 K km s-1. Bottom panels: same as top panels, for the bulk and inner wings emission; the emission has been integrated in the ranges from -9 to -5 km s-1 (continuous contours), from -5 to -2 km s-1 (grey scale), and from -2 to 2 km s-1 (dashed contours). The white triangles indicate the positions of the H2 O maser spots

This result differs from the conclusions reached by WBM, who mapped a larger region (but with lower angular resolution) around IRAS 20126+4104 in the [FORMULA] lines of 12 CO and 13 CO. They found a powerful outflow the orientation of which is N-S, consistent with our more limited 13 CO(2-1) map. We shall discuss the different velocity patterns in detail in Sect. 4.1.

3.2. Molecular line data from PdBI

In the previous section we have seen that a radical difference does exists between the kinematics of the high and low velocity gas. A better insight is given by the high angular resolution ([FORMULA]) maps made with the PdBI in the same line. In Figs. 5 and 6 we show the spectra in the transitions detected with the PdBI, obtained by integrating over a circle of [FORMULA] radius. Note that the PdBI in the configuration used by us cannot image regions more extended than [FORMULA]: we thus miss the extended structures visible e.g. in the 13 CO(2-1) line with the 30-m telescope. We detect the CH3 OH(153 -144[FORMULA] transition at 324 K above ground, the ([FORMULA])=(1,1), (2,1), and (0,1) lines of CH3 CN(5-4) [FORMULA] =1 at [FORMULA]  K, and the K =0 and 1 transitions of C [FORMULA] CN. Thus, we clearly detect a component of hot molecular gas.

[FIGURE] Fig. 5. Spectra observed with the PdBI. The HCO [FORMULA] spectra have been obtained by integrating over a circle with [FORMULA] radius centred at the H2 O maser position (thick histogram) and over the whole region where emission is imaged by the PdBI (thin histogram). The CH3 OH and CH3 CN spectra have been obtained by integration over the same circle as above. The conversion factor from flux to brightness temperature is 2.25 K Jy-1 for all spectra but the HCO [FORMULA] spectrum integrated over the whole emitting region, for which one must use 0.3 K Jy-1: note that what we refer to is the mean brightness temperature over the region used for obtaining the integrated spectra. The dotted and dashed lines have the same meaning as in Fig. 1

[FIGURE] Fig. 6. Same as Fig. 1, for the ground state (top panel) and vibrationally excited (bottom) transitions of CH3 CN(5-4), but with flux versus frequency. The spectra have been obtained by integrating over the same circle as in Fig. 1. The conversion factor from flux to brightness temperature is 2.1 K Jy-1. The continuous and dashed vertical lines indicate respectively the transitions of CH3 CN and 13 CH3 CN; the corresponding quantum numbers are also shown

In the following we shall describe separately the maps of HCO [FORMULA], on one hand, and of CH3 CN and CH3 OH, on the other.

3.2.1. HCO [FORMULA] maps

As in Sect. 3.1 for the 30-m maps, in Fig. 7 we compare the blue- and red-shifted emission in the HCO [FORMULA] (1-0) line with the bulk emission, both for the outer and inner line wings. Fig. 7 definitely confirms the conclusions derived from Fig 4, namely the existence of a SE-NW symmetry and the reversal in the position of the blue- and red-shifted gas, going from the outer (top panel) to the inner (bottom panel) wings. This is a very surprising result and will be discussed later in Sect. 4.1. Here we note that the axis defined by the blue- and red-shifted gas coincides very well with the distribution of both the 2.2 µm H2 line emission and the H2 O maser spots. This is shown in Fig. 8, where the HCO [FORMULA] blue- and red-shifted lobes corresponding to the inner wings of the line are compared with a map of the H2 line integrated intensity from the NOT.

[FIGURE] Fig. 7. PdBI maps of blue- (continuous contours) and red-shifted (dashed) HCO [FORMULA]. Top panel shows outer wings compared with sum of bulk emission and inner wings (grey scale); bottom panel shows inner wings compared with bulk emission (grey scale). Contour levels range from 0.05 to 0.26 by 0.03 Jy/beam in the top panel and from 0.2 to 1.6 by 0.2 Jy/beam in the bottom panel. The white triangles indicate the positions of the H2 O maser spots
[FIGURE] Fig. 8. Comparison between the H2 line emission observed with the NOT (grey scale) and the blue- and red-shifted emission in the inner wings of the HCO [FORMULA] (1-0) line. Contour levels range from 0.2 to 1.6 by 0.2 Jy/beam. The black triangles indicate the positions of the H2 O maser spots

The bulk emission (represented by the grey scale in the bottom panel of Fig. 7) although extended on a [FORMULA] region, seems to originate mostly from a compact core centred on the H2 O masers. Such core is the same seen in the high density tracers, as shown in Sect. 3.2.2.

3.2.2. CH3 CN and CH3 OH maps

The contour maps of the integrated intensity of the CH3 CN(5-4) K =0+1 and CH3 OH(153 -144[FORMULA] lines are shown in Fig. 9, while Table 4 gives the peak positions of such maps, obtained with a 2-D gaussian fit, the angular diameter of the source, [FORMULA], and the peak brightness temperatures in the synthesised beam, [FORMULA]. The CH3 CN emission looks pointlike, originating from a compact core coincident with the H2 O masers; similarly, the CH3 OH line originates in an unresolved core at the same position as that seen in CH3 CN. Thus, the methyl cyanide and methanol emission originates in a hot compact source of angular size less than 16.

[FIGURE] Fig. 9. Overlay of 3.3 mm continuum map (grey scale) with contour maps in the CH3 OH(153 -144[FORMULA] (top) and CH3 CN(5-4) K =0+1 lines. Contour levels range from 0.02 to 0.12 by 0.02 Jy/beam for CH3 OH, and from 0.03 to 0.31 by 0.04 Jy/beam for CH3 CN


Table 4. Peak position, obtained with a 2-D gaussian fit, angular diameter, and peak brightness temperatures in the synthesised beam for different tracers

In order to get some information on the kinematics of the gas in the core, we have applied the method used by Cesaroni et al. (1994b) to the CH3 CN(5-4) and CH3 OH(153 -144) A [FORMULA] transitions. This consists in producing channel maps in the lines and fitting a 2-D gaussian to the map in each channel. The peak positions of the gaussians are plotted in Fig. 10, bottom: they fall within an elongated structure, [FORMULA] long and almost perpendicular to the orientation of the HCO [FORMULA] bipolar pattern. Since each peak corresponds to a velocity, one can plot this against the offset along the line (obtained from a linear fit to the points in the right ascension-declination plane) described by the distribution of the peaks. The result is shown in the top panel of Fig. 10: a good velocity-position trend is visible (correlation coefficient -0.77), with [FORMULA] increasing from NE to SW. Note that a similar plot for the offset along a direction perpendicular to that of the peaks shows no correlation at all.

[FIGURE] Fig. 10. Bottom panel: peak positions measured in different velocity channels for the CH3 CN(5-4) and CH3 OH(153 -144) A [FORMULA] transitions. Top panel: channel velocities against angular offsets along the symmetry axis denoted by the line in the bottom panel: note that the angular offset is computed from an arbitrary origin and increases from SW to NE, as shown in the figure; the vertical line indicates the offset of the continuum peak. Filled and empty circles indicate CH3 CN and CH3 OH respectively, triangles H2 O maser spots, and the asterisk the 3.3 mm continuum. The crosses in the top right and left corners of the panels show typical error bars

3.3. Continuum data from PdBI

As shown in Table 1, our continuum map at 3.3 mm has been obtained by integrating two 160 MHz bandwidths in the lower and upper side bands, where no line emission was detected. We measure a 3.3 mm continuum flux of [FORMULA]  Jy, which arises from an unresolved ([FORMULA] or [FORMULA]  pc) region coincident with the H2 O masers and the CH3 CN and CH3 OH core, as shown in Fig. 9 ; the peak position resulting from a 2-D gaussian fit and the peak [FORMULA] are given in Table 4. To our knowledge, this is the first measurement of the millimeter continuum in IRAS 20126+4104 with resolution of a few seconds of arc.

A spectrum of the continuum emission from IRAS 20126+4104, over a large frequency interval using our and literature data, is shown in Fig. 11. The different fluxes have been taken from the following references (going from low to high frequency): 6 cm and 3.6 cm from Marti & Rodriguez (in prep); 2 cm from Wilking et al. (1989); 1.3 cm from TFTH; 3.3 mm, from this paper; 2.7 mm from Wilking et al. (1989); 1.3 mm from Walker et al. (1990); 100, 60, 25, and 12 µm from the IRAS Point Source Catalogue; K, H, and J band obtained with the TIRGO telescope from this paper. It is worth noting that the [FORMULA]  mJy source detected by Marti & Rodriguez (in prep) at 3.6 cm looks unresolved ([FORMULA]) and peaks at [FORMULA] and [FORMULA], namely at the same position as the 3.3 mm continuum.

[FIGURE] Fig. 11. Spectrum of the continuum emission towards IRAS 20126+4104. The filled circle represents the PdBI measurement at 3.3 mm, while the TIRGO measurements, which may refer to scattered radiation, are shown by filled triangles: these are the sum of the NIR emission to SE with that to NW of the H2 O masers. Arrows indicate upper limits. The dashed and dotted lines correspond respectively to the spectrum of optically thick free-free emission ([FORMULA]) and optically thin dust emission ([FORMULA])

When examining Fig. 11, one must remember that the data have been taken with different angular resolutions ranging from a few minutes of arc, for IRAS, to [FORMULA], for the Very Large Array at 1.3 cm. Nevertheless, it is clear that all measurements at [FORMULA]  mm satisfy the spectral slope ([FORMULA]) typical of optically thick free-free emission (dashed line in Fig. 11). The spectral index between 3.6 cm and 3.3 mm is [FORMULA]. However, we think it more probable that the dust contribution to the 3.3 mm continuum must be dominant, because the 3.3 mm, 2.7 mm, and 1.3 mm measurements can be fitted by assuming optically thin dust emission ([FORMULA] ; dotted line in Fig. 11): it is thus hard to believe that the free-free contribution is relevant at 3.3 mm. More likely, the free-free continuum detected at 3.6 cm is optically thin, which strengthens the previous conclusion. In fact, Tofani et al. (1995) using the VLA in the A-array configuration do not detect any continuum emission at 3.6 cm above a [FORMULA] level of 0.246 mJy/beam, or 58 K in a [FORMULA] beam. This sets a lower limit of [FORMULA] on the source angular diameter, which in turn implies that the measured brightness temperature of 58 K is close to the true brightness temperature, too small for an optically thick HII region.

3.4. Continuum and H2 line data from TIRGO and NOT

In Fig. 12, the continuum map at 3.3 mm (grey scale) is compared with the K-band continuum and H2 line NIR maps obtained with the NOT. The 3.3 mm emitting region clearly coincides with the H2 O masers and the nominal position of the IRAS source. The K-band continuum and H2 line emission, instead, are extended and offset from the 3.3 mm continuum. We note that the K-band continuum arises from a region slightly closer to the masers than the two H2 line lobes. We hypothesise the existence of a luminous object, embedded in the compact, dense, dusty core where the H2 O masers are located: the star cannot be directly detected in the NIR, due to high extinction in the core, but its existence is witnessed by the two nebulosities (most probably reflected starlight) seen at K-band. Such a scenario is confirmed by the fact that high density tracers like CH3 CN and CH3 OH arise from an unresolved region coincident with the 3.3 mm continuum, as shown in Fig. 9.

[FIGURE] Fig. 12. Overlay of 3.3 mm continuum (grey scale), K-band (continuous contours), and H2 (dotted contours) maps. Also shown are the H2 O maser spots (triangles) and the IRAS point source position (plus sign) and related uncertainty (dashed ellipse)

Another evidence in favour of the existence of a dense core at the H2 O maser position is that 28% of the NIR sources lying within [FORMULA] from the H2 O masers have IR excess. In Fig. 13 we plot the sources detected at J, H, and K inside a [FORMULA] region around the H2 O masers (bottom panel) and the position of the corresponding points in the colour-colour diagram obtained from the TIRGO measurements at J, H, and K (top panel). All sources with strong IR excess form a small cluster around the H2 O masers, but no NIR source is detected exactly at such position: this implies a large column density at the location of the star which powers the masers. We shall come back to this point in Sect. 5.

[FIGURE] Fig. 13. Top: colour-colour diagram of NIR sources detected at all badwidths observed with TIRGO inside a [FORMULA] region around the nominal position of IRAS 20126+4104. Bottom: spatial distribution of the same sources as in the top panel. The ellipse indicates the 1 [FORMULA] position uncertainty of IRAS 20126+4104. The continuous line corresponds to the position of main sequence stars; the two dashed lines indicate the location of reddened main sequence stars. Filled circles indicate objects with strong NIR excess, i.e. below the lower dashed line and with H-K [FORMULA] 1: these objects cluster around the position of the IRAS source
Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: April 28, 1998