Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.arcetri.astro.it/~lt/preprints/hh8081/hh8081_final.pdf Äàòà èçìåíåíèÿ: Tue Sep 11 16:57:04 2007 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 07:55:53 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: reflection nebula

Mon. Not. R. Astron. So c. 000, 1­?? (2002)

Printed 16 Septemb er 2003

A (MN L TEX style file v2.2)

Millimeter observations of the IRAS 18162-2048 outflow: evidence for cloud disruption around an intermediate-mass protostar.
M. Benedettini1, S. Molinari1, L. Testi2, A. Noriega-Crespo 1
2 3

3

CNR­Istituto di Fisica del lo Spazio Interplanetario, Area di Ricerca di Tor Vergata, via del Fosso del Cavaliere 100, 00133, Roma, Italy INAF-Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125 Firenze, Italy SIRTF Science Center, California Institute of Technology, 220-6 Pasadena, CA 91125

Accepted . Received ; in original form

ABSTRACT

In order to study the morphology and dynamics of the molecular outflow associated with IRAS 18162-2048, a wide area of 95 arcmin2 around the source has been mapped by means of CO and 13 CO (1-0) lines, and complemented by a map of a smaller region surrounding the high-mass ob ject using the C18 O(1-0) and CH3 OH (2k -1k ) and (3k -2k ) transitions. The lines profile reveals the presence of several velocity components among which two ma jor line components at 11.9 and 12.8 km s -1 have been detected in all the tracers. Simple morphological and energetic considerations led us to interpret the observations in a relatively straightforward scenario in which the powerful jet ejected by IRAS 18162-2048 sets a big portion of the surrounding molecular cloud into motion. The energy and momentum deposited by the flow breaks the cloud apart, shifting to a blue velocity the northern region and to a red velocity the southern region, and giving rise to a giant outflow. We calculated the physical parameters of the outflow, which makes the IRAS 18162-2048 outflow, as one of the most massive (M=570 M ) and energetic (K>1046 ergs) knowns. Despite the intrinsic difficulties in giving a precise value of the age and the inclination angle of the flow, we used different methods to derive a reliable estimate. Our data show evidence in favor of a small inclination angle (< 50 ) and of a maximum outflow age of 106 yr. C18 O and CH3 OH trace the dense core surrounding IRAS 18162-2048 and show an elongated emission in the direction perpendicular to the outflow axis. Besides the peak emission associated with the IRAS source, we found another peak at the position RA(B1950)=18h16m 20.2s DEC(B1950)=-2049 18 which coincides with a red near infrared source. We provided evidence that this second peak may be surrounded by a flattened rotating structure, suggesting that the newly discovered IR source can be another site of recent star formation in this region. Our analysis suggest that the powerful wind/outflow from the luminous stars within the young cluster embedded in the GGD27 nebula are tearing apart the parental molecular cloud. The IRAS 18162-2048 appears to be in the act of clearing the surrounding material on the verge of becoming an optically revealed young stellar cluster, similar to those associated with Herbig Be stars. Key words: ISM: individual: IRAS 18162-2048 (HH80/81) - ISM: jets and outflows - stars: formation.

1

INTRODUCTION

Based on observations collected with SEST, the Swedish-ESO submillimeter telescop e at La Silla (Chile) c 2002 RAS

The early stage of the star formation process is characterized by the presence of energetic mass ejections which interact with the surrounding molecular cloud giving rise to bipolar outflows. Molecular outflows are thought to be the


2

M. Benedettini et al.
a study of the region by means of far infrared spectroscopy using the Infrared Space Observatory (ISO), showing that the shock velocities measured toward the HH ob jects are of the order of 100 km s-1 only, indicating that the shocks may arise at the interface between two fast-moving flows. They also find evidence for a Photo-Dissociation Region induced on the cavity walls by the UV radiation produced by the shocks. Near and mid infrared observations of IRAS 181622048 resolve it in a cluster of YSOs, with at least two massive stars in an early evolutionary state: namely GGD27-ILL and IRS 7 (Aspin et al. 1994; Stecklum et al. 1997). GGD27-ILL has been identify as the exciting source of the radio jet and the molecular outflow as well as the illuminating source of the reflection nebula. In this paper we present multi-line mapping observations, in CH3 OH, CO and its isotopes, over a 95 arcmin2 around IRAS 18162-2048 aimed at the study of the structure of the large scale molecular outflow and the dense core and to investigate how the mass ejection from the high mass forming ob ject affects the parent molecular cloud.

mechanism through which the accreting central ob ject loses the exceeding angular momentum, and therefore, plays an important role in stopping the gravitational collapse thus setting the final mass of a young stellar ob ject (YSO). More than two decades of investigations have revealed that molecular outflows are ubiquitously present in association with low-mass protostars. The situation for the high-mass protostars, however, is less clear because of the intrinsic difficulties in observing this fast evolving ob jects. Recent systematic studies nevertheless indicate that 80-90% of the observed massive star forming region are associated with outflows (Zhang et al. 2001; Beuther et al. 2002) and that the massive outflows show physical characteristics similar, apart from a scale factor, to those of low-mass outflows. This suggests that a similar mechanism is at work in the formation of low and high-mass stars (Beuther et al. 2002). While this may be correct statistically, for a relationship between the global flow energetics and the exciting source bolometric luminosity, it is not clear that the detailed properties of high mass flows are similar to those of the lower-mass counterparts. In particular high-mass flows generally appear to be less collimated and more turbulent than flows from low-mass ob jects. Additionally the fast moving, high excitation and highly collimated jet component usually observed in lowmass systems is much less important or absent in high-mass flows (Shepherd et al. 1997; Shepherd et al. 2003a). Important differences in the physical structure of the flows have also been observed at millimeter wavelengths (Molinari et al. 2002). All these results suggest that simple scaling of low-mass flow models may not be applicable (Richer et al. 2000; Shepherd 2003b). High angular resolution studies are crucial in this respect, although up to know only a handful of outflows from high-mass young stellar ob jects have been studied with the necessary detail: IRAS 20126+4104 (Cesaroni et al. 1997; 1999; Shepherd et al. 2000), G192 (Shepherd et al. 1997), W75N (Shepherd et al. 2003a). A relatively poorly studied massive flow is that powered by the luminous (L=2â104 L ) infrared source IRAS 181622048 (Yamashita et al. 1989) at a kinematic distance of 1.7 Kpc (Rodr´ iguez et al. 1980). IRAS 18162-2048 is associated with the infrared reflection nebula GGD 27 (Yamashita et al. 1987), an ultracompact H ii region and a number of H2 O and OH masers (Rodr´ iguez et al. 1980). A CO bipolar outflow centered on the IRAS source and an elongated disk­like structure perpendicular to the flow have been observed by Yamashita et al. (1989). They derived a mass of 460 M and an age of 105 yr for the outflow, while for the disk the density is in excess of 1.2â105 cm-3 and the mass is 560 M . IRAS 18162-2048 also drives a highly collimated radio jet composed of a multitude of knots (Mart´ et al. 1993) i which reach velocities of 600-1400 km s-1 (Mart´ et al. i 1995). The pro jected extent of the jet is 5.3 pc, one of the largest ever found. The Hergib-Haro ob jects HH80 and HH81 lie along the jet's path in the southern lobe and have been studied in detail by Heathcote et al. (1998). They found in HH80/81 exceptional high excitation conditions (indicating strong shocks), H and [S II] line widths up to 600 km s-1 and tangential velocity of 350 km s-1 . HH80N defines the northern tip of the radio jet and it is only revealed in the far infrared and radio because of the high visual extinction toward the denser cloud interior where the northern flow lobe is advancing into. Molinari et al. (2001) have carried out

2

OBSERVATIONS

We performed a multi-line observations of a region along the outflow excited by IRAS 18162-2048 (RA(B1950)=18h 16m 12.9s DEC(B1950)=-20 48 49 ). The observations were carried out with the SEST telescope at La Silla (Chile) from 27 to 31 May 2001. The SESIS 100/150 GHz receiver has been coupled with the Acousto Optical Spectrometers to observe simultaneously with high (80 kHz) and low (1.4 MHz) resolution. We obtained a map of 6.75â14.25 arcmin2 in the CO (1-0) and CH3 OH (3k -2k ) lines, while in the 13 CO (1-0), C18 O (1-0) and CH3 OH (2k -1k ) transitions a smaller region has been covered (as detailed in Table 1). The grid had a PA=-17 and a 45 spacing (about 1 beam) with the exception of the CH3 OH (2k -1k ) line map where the spacing was 22.5 arcsec. In Table 1 we give the details of the observations: the observed molecule and transition (Col. 1 and 2), the rest frequency of the transition (Col. 3), the HPBW (Col. 4), the main beam efficiency (Col. 5), the typical system temperature (Col. 6), the integration time in each map position (Col. 7), the spectral resolution of the 80 kHz (high resolution) data (Col. 8) and the mapped area (Col. 9). To improve the signal to noise ratio the spectra have been smoothed at half the velocity resolution. The intensity calibration was done with the classic chopper wheel method. The antenna temperature has been corrected for the main beam efficiency (see Table 1) and the data is being presented in units of main-beam brightness temperature.

3 3.1

RESULTS Large-scale Structures

We investigated the large scale structure mapping a wide area of 95 arcmin2 around IRAS 18162-2048 by means of CO and 13 CO (1-0) lines. The CO (1-0) integrated map is shown in Fig. 1 while Figs. 2 and 3 show the spectra of the CO and 13 CO (1-0) lines in different positions of the mapped
c 2002 RAS, MNRAS 000, 1­??


The HH80-81 region
Table 1. List of the observed transitions and observing parameters. Molecule Transition Frequency (GHz) 115.271 110.201 109.782 96.746 145.113 HPBW (arcsec) 45 47 47 52 34 Beam eff. Tsy s (K) 580 400 420 280 290 tint (s) 90 60 60 60 30 dv(80 kHz) (km s-1 ) 0.1107 0.1157 0.1162 0.1318 0.0879 mapp ed area arcminâarcmin 6.75â14.25 6.00â12.75 5.25â6.75 2.25â3.75 6.75â14.25

3

CO CO C18 O CH3 OH CH3 OH
13

1-0 1-0 1-0 2k -1 3k -2

k k

0.70 0.71 0.71 0.73 0.66

Figure 1. CO (1-0) integrated map. The contours start from 3 K km s-1 with spacing of 5 K km s-1 . The crosses indicate the p oints of the map and the triangles the p ositions of IRAS 18162-2048 and the three HH ob jects HH80, HH81 and HH80N. The three lines indicate the directions along which the sp ectra shown in the following figures are taken from. The central line corresp onds to the direction of the radio jet.

Figure 2. Sp ectra of the CO (1-0) transition along the three directions parallel to the flow axis as indicated by the straight lines of Fig. 1. The offset from the central p osition (co ordinate of IRAS 18162-2048) with resp ect to the flow axis (PA=-17 ) is given in the upp er left corner in arcsec. The two vertical lines indicate the velo city p ositions of the two ma jor outflow comp onents: vLSR =11.9 km s-1 and vLSR =12.8 km s-1 .

area, moving from south to north along three directions parallel to the radio jet axis (i.e. the three straight line drawn in Fig. 1). In the on source pointing a non­gaussian high velocity wing emission is detected with total velocity width at zero intensity of about 20 km s-1 (see Fig. 4). Thanks to the high spectral resolution of the observations, we can identify multiple components of line emission at most of the observed positions. Among these, two ma jor line components can be identified. These components show a similar intensity in the immediate surrounding of IRAS 18162-2048 but behave differently as one moves away from the source along the jet axis. In particular, in the northern part of the map the blue velocity component prevails, while the red component is predominant in the southern part of the map. The same behavior has been observed in the 13 CO and C18 O as can be seen in Fig. 5 were the spectra of the three species are overplotted for different positions along the flow axis. We anticipate that the two line components are also detected in the CH3 OH lines (see Fig. 6). We have measured the velocity of these two ma jor components using the C18 O line because
c 2002 RAS, MNRAS 000, 1­??

this tracer is less sensitive to the ambient component and is affected less for opacity effects. The blue component is centered at 11.9 km s-1 and the red one at 12.8 km s-1 . The relative proximity of the two velocity components, prevents an accurate estimate of their line parameters for most of the observed positions. Essentially it is impossible to reliably fit the overall line profile by means of two (or more) gaussian contributions, without fixing an arbitrary FWHM. The ratio between the brightness temperature of the CO and 13 CO in the core of the (1-0) line ranges from 1 to 7, indicating that the CO (1-0) line is optically thick all over the mapped area, thus a measurement of the excitation temperature can be given from the line peak. Merging all the CO spectra, we derive a mean temperature of 12 K. The ratio between the 13 CO and the C18 O is about the abundance ratio of the two isotopes so that both the lines are optically thin. The column density of the CO is deduced assuming an excitation temperature T=12 K (see Sect. 3.1) and a CO abundance relative to H2 of 2.8â10-4 (Cardelli


4

M. Benedettini et al.

Figure 3. As in Fig. 2 for the

13

CO (1-0) transition.

Figure 5. Sp ectra of the ground state transition of CO (dotted line), 13 CO (long-dashed line) and C18 O (solid line, 3 times the real value) in different p ositions along the flow axis. For each panel, the offset in arcsec, from the central p osition with resp ect to the flow axis (PA=-17 ) is given in the left upp er corner.

1

Figure 4. CO (1-0) line in the on-source p osition with sup erimp osed a two comp onents gaussian fit.

0.5

0

et al. 1996). The optical depth is calculated from the ratio between the CO and the 13 CO brightness temperatures in each velocity channel. In those cases where the 13 CO data are not available, the optical depth is deduced by a linear extrapolation from the data of the neighboring positions. Based on these measurements we obtain a total cloud mass of 570 M .

-0.5

15

20

25 v (km/s)

30

35

Figure 6. Sp ectra of the CH3 OH 2k -1k line in different p osition of the mapp ed area. The offsets from the central p osition with resp ect to the flow axis (PA=-17 ) are 0 ,0 (solid line), 22.5 ,22.5 (dotted line) and 45 ,45 (long-dashed line).

3.2

Emission Knots

Knots of CO (1-0) emission at various velocities, vLSR from 9.5 to 43 km s-1 , are clearly detected both along and aside of the radio jet axis (see Figs. 7 and 8). In Table 2 we list
c 2002 RAS, MNRAS 000, 1­??


The HH80-81 region

5

D

L
E

H C

G

A

I

F

B

Figure 7. Velo city channel map of the CO (1-0) transition. The velo city of each panel is indicated in the upp er left corner. The contour lines range from 1 to 23 K km s-1 with spacing of 1.5 K km s-1 . IRAS 18162-2048 is identify by a cross while the p ositions of HH80, HH81 and HH80N are indicated by triangles. Knots of CO emission A, D, G, and L, with velo city b etween 9 km s-1 and 11 km s-1 , are also indicated . The filled circle in the last panel shows the HPBW.

the coordinates of the emission peak of the identified knots, which we label with a letter form A to L, together with their line center velocity. We estimate an uncertainty of 20 arcsec, which corresponds to the half spatial resolution of the map. Knots B, C, F and I are spectroscopically well identified and we can estimate their total mass and momentum (see Table 2 Cols. 5 and 6, respectively); these are only lower limits for C, F and I since our observations did not map them completely. Two of the CO knots are spatially associated, within the quoted uncertainty, with radio spots i detected by Mart´ et al. (1993), namely C and H which correspond to the radio sources 9 and 33, respectively. The radio source 33 has been identified by Mart´ et al. as an i extragalactic source because of its negative spectral index (=-0.8±0.1). The radio source 9 has a positive spectral index (=0.9) but it is considered by Mart´ et al. (1993) as i a possible time-variable ob ject, in which case the derived spectral index is meaningless. We note that the emission at 18.7 km s-1 of knot C is close to the position of HH80 even if the emission of the knot C peaks more south. Knots A, D, F, G, I and L have also been detected in the 13 CO (1-0) line. 3.3 The Surroundings of IRAS 18162-2048

Figure 8. Velo city channel map of the CO (1-0) transition. The velo city of each panel is indicated in the upp er left corner. Lowest contour and contour interval is 0.4 K km s-1 . The filled circles show the HPBW, the cross identifies the p osition of IRAS 181622048 and the triangles indicate the p osition of HH80, HH81 and HH80N. Knots of CO emission B, C, E, F, H, and I, with velo city b etween 17 km s-1 and 43 km s-1 , are also indicated.

The surroundings of the massive source IRAS 18162-2048 have been mapped by means of high density tracers as C18 O and CH3 OH. We detected the C18 O (1-0) line and five transitions of the CH3 OH molecule, namely the 2-1 1-1 E at 96.73939 GHz, 20 -10 A+ at 96.74142 GHz, 20 -10 E at 96.74458 GHz, 3-1 -2-1 E at 145.09747 GHz and 30 -20 at 145.10323 GHz. Also in these lines we identify the two
c 2002 RAS, MNRAS 000, 1­??

ma jor line components, as in the CO and 13 CO tracers (see Fig. 6). In the contour map of the C18 O (1-0) line (Fig. 9) and CH3 OH 20 -10 A+ line (Fig. 10) we can see that the integrated line emission is elongate in the east direction along the plane orthogonal to the flow axis and peaks 30 arcsec north-east with respect to the IRAS source, we will discuss this point in Sect. 4.2. A second peak at 100 arcsec south-east the IRAS coordinate has been detected at RA(B1950)=18h 16m 20.2s DEC(B1950)=20 49 18 . Consulting the 2MASS quick-look images, we found a red source in the same position of the second maximum, visible only in the K band image. In Fig. 11 we show the position-velocity diagram for the (1-0) C18 O line, along the direction orthogonal to the jet axis and crossing the IRAS source. We see two maxima on both sides and at equal distance to the second peak at a velocity of 11.9 and 12.3 km s-1 , respectively. Also the position-velocity plot of the CH3 OH 2k -1k lines (see Fig. 12) shows a structure similar to that of the C18 O with two methanol maxima in opposite direction even if closer to the second peak. This feature will be discussed in Sect. 4.2. The methanol emission does not exhibit line wings, we therefore attribute it to the dense core; contrary to the expectations (e.g. Bachiller et al. 1995), we do not detect a methanol component related with the shocked outflow. Under the assumption of LTE condition, from the rotational diagram (Blake et al. 1987) we derive a Trot =7.5 K (see Fig.13), a value lower than the temperature of


6

M. Benedettini et al.
Table 2. Knots identified in the CO line map. Knots RA(B1950) hms 18 18 18 18 18 18 18 18 18 18 15 16 16 16 16 16 16 16 16 16 58 07 07 13 17 18 26 26 27 28 DEC(B1950)


v km s

-1

Mass M ­ 0.033 >0.034 ­ ­ >0.632 ­ ­ >0.208 ­

Momentum M km s-1 ­ 0.99 >0.21 ­ ­ >10.4 ­ ­ >3.22 ­

tdyn 105 years ­ 0.48 4.2 2.4 3.1 ­ 12 ­ ­ 16

A B C D E F G H I L

-20 -20 -20 -20 -20 -20 -20 -20 -20 -20

47 51 53 47 46 53 44 48 50 42

46 31 53 20 24 55 20 00 05 00

10.6 43.0 18.7 9.5 16.7 28.6 10.3 17.0 28.1 10.1

Figure 9. Contour plot of the C18 O (1-0) transition. The contours range from 0.42 (3 ) to 7 K km s-1 with spacing of 0.5 K km s-1 .

Figure 10. Contour plot of the CH3 OH 20 -10 A+ transition. The contours range from 0.4 (3 ) to 4 K km s-1 with spacing of 0.4 K km s-1 .

the molecular outflow, and a methanol column density of N(CH3 OH)=3â1014 cm-2 . To calculate the methanol fractional abundance with respect to H2 we derive the H2 column density from the C18 O data, since both molecules trace the quiescent gas of the dense core. We derive a methanol fractional abundance of X(CH3 OH)=4â10-8 . Both the rotational temperature and the fractional abundance are similar to the values found in ambient gas close to young bipolar outflows (Bachiller et al. 1995). In principle the methanol abundance is expected to be enhanced in shocked regions associated with outflows, however, our source, as well as regions like Orion Hot Core, do not show any strong enhancement. It has been suggested that the CH3 OH is not abundant in regions where extreme high temperature produces evaporation of the volatile CO molecules from the grain mantles, preventing the formation of methanol (Caselli et al. 1993).

Figure 11. Position-velo city plot of the C18 O (1-0) line in the direction orthogonal to the flow axis and crossing the central source. The contour levels range from 0.5 to 3.5 K by steps of 0.15 K.

c 2002 RAS, MNRAS 000, 1­??


The HH80-81 region

7

Figure and 2-1 axis and 0.3 to 1

12. Position-velo city plot of the CH3 OH 20 -10 A+ (left) -1-1 E (right) lines in the direction orthogonal to the flow crossing the central source. The contour levels range from K by steps of 0.05 K.

Figure 13. Rotational diagram for the CH3 OH observed lines.

4 4.1

DISCUSSION A molecular cloud breaking apart

As shown in the previous section, the wide mapped area surrounding the IRAS 18162-2048 source shows several velocity components. The two ma jor ones have been observed in all the tracers and they cover most of the mapped area with a coherent spatial distribution. They define two large-scale structures peaking at two different positions almost along the jet axis and on opposite sides with respect to IRAS 18162-2048 (see Fig. 5). We stress that these two components represent the cloud in its entirety and not just highvelocity tails of a cloud whose main body is at some intermediate systemic velocity. An interpretation of our findings in terms of cloud rotation seems then unlikely. First of all the rotation would be around an axis perpendicular to the axis of the observed radio jet. If this was the case, one might observe a systematic shifting of the line center in positions farther away from the central source and the line width might be the same in all the positions, but it is not what we observe. Furthermore, a cloud rotation axis orthogonal to the flow axis is difficult to explain in the framework of standard star formation where the contraction of a rotating cloud conserves the direction of angular momentum (e.g. Shu et al. 1987; Olmi & Testi
c 2002 RAS, MNRAS 000, 1­??

2002) and forms YSO with jet parallel to the cloud rotation axis. A more plausible interpretation is that we may be looking at two different clouds which are in relative motion. The first scenario is one in which the two clouds are approaching to each other, triggering the star formation process in the region. However, in the case of star formation stimulated by a cloud-cloud collision, it is expected that the compressed material will have a net angular momentum perpendicular to the plane containing the clouds velocity vectors. This direction is expected to be also the direction of the outflow axis of the protostar(s) generated by the collision. In the geometry derived from our observations, such axis should be pro jected onto the plane of the sky in approximately an east-west direction, i.e. orthogonally to the observed radio jet. A perhaps more intriguing possibility is suggested by the spatial correspondence between the red(blue)-shifted cloud and the red(blue)-shifted radio jet lobe, namely that the powerful jet ejected by IRAS 18162-2048 sets a big portion of the surrounding molecular cloud into motion. This interpretation is supported by simple energy balance argument. The gravitational potential energy of the overall gas material can be roughly estimated assuming that the observed mass is distributed in a sphere of uniform density and radius of 400 arcsec, equal to the extent of the CO emission each side of IRAS 18162-2048. In this case the potential energy U=(3GM2 )/(5R)5â10 45 ergs is less that the kinetic energy of the moving gas (>1046 ergs), so that the energy and momentum deposited by the flow might be sufficient to break the cloud apart. As a result the gas emission in the northern region would be predominantly blue-shifted, while the southern region would be predominantly red-shifted. If this interpretation is correct the entire cloud could be seen as a giant outflow, with the individual clouds being the two lobes. In addition to the two main components, we also identify ten separate blue and red shifted CO knots. Three of these, namely A, F, I are aside the axis of the flow. The others seven, namely B, C, D, E, G, H, L are aligned along the flow axis. In agreement with the velocity of the large scale outflow all the knots (but E) lying in the northern part of the map show blue-shifted velocity, while those lying in the southern part show red-shifted velocity. Therefore it is reasonable to assume that these clumps are associated with the outflow and to include their contributions into the calculation of the mass and energy budget of the flow. High velocity molecular clumps are often observed along the outflow of both low mass (e.g. L1448-mm (Bachiller et al. 1990)) and high mass stars (e.g. IRAS 20126+4104 (Shepherd et al. 2000)). On the other hand, infrared observations at high spatial resolution have revealed a cluster of protostellar ob jects in the region of the massive IRAS 18162-2048 source (Aspin & Geballe 1992; Stecklum et al. 1997), and the observed CO knots could also be an indication of the presence of other cores/outflows less powerful with respect to the main one, originated from the protostars of the cluster which cannot be resolved by our observations because of their limited spatial resolution. North-west of HH80N we can see the signature of the red and blue line wings of a CO outflow already discovered by Girart et al. (2001) by means of interferometric observa-


8

M. Benedettini et al.

Table 3. Physical parameters of the outflow, not corrected for the inclination angl for an inclination angle of 3 (Col. 3) and 56 (Col. 4). Parameter H2 column density (cm-2 ) CO column density (cm-2 ) Mass ( M ) Momentum (M km s-1 ) Energy (ergs) dynamical time (yr) Momentum Flux (M km s Mass Loss Rate (M yr-1 ) Luminosity (L ) i = 90


i=3



i = 56



-1

yr

-1

)

1.5â1021 4.2â1017 570 536 1â1046 1â106 5.4â10-4 5.7â10-4 0.08

1.5â1021 4.2â1017 570 10242 3.7â1048 1â106 1â10-2 5.7â10-4 30

1.5â1021 4.2â1017 570 647 1.5â1046 1â106 6.5â10-4 5.7â10-4 0.12

Figure 14. A map of the CO (1-0) line wings: the blueshifted comp onent (dashed line) is integrated from 3 to 10 km s-1 and the redshifted comp onent (solid line) is integrated from 15 to 23 km s-1 . Lowest contour is 1 K km s-1 (3 ) and the step is 3 K km s-1 .

tion. They attribute the CO outflow to a local Class 0 ob ject whose formation has been triggered or at least speed up by the outflow generated by IRAS 18162-2048. The emission arising from this outflow is negligible with respect to the main one and then we do not exclude this region for the computation of the overall gas properties. In the scenario described above where the overall observed gas is participating to the motion, we calculate the physical parameters of the IRAS 18162-2048 outflow integrating the line flux of the CO (1-0) transition over the whole spatial extent of our map (6.75â14.25 arcmin2 ) and over the entire velocity range of the line (vLSR from 3 to 23 km s-1 ) and we also add the contributions of the knots at larger velocities (see Table 2). Fig. 14 shows the morphology of the outflow, where for clarity we plot only the contribution of the line wings: the blue-shifted emission is integrated between 3 and 10 km s-1 and the red-shifted one from 15 and 23 km s-1 . The momentum computed is 536 M km s-1 with a kinetic energy of 1046 ergs. A rough estimate of the age of the outflow can be derived from: i) the outflow dynamical time tdyn =3â106 yr, calculated as the ratio between the mean size of the lobes in the CO (1-0) transition and the mean velocity weighted with the mass and from ii) the dynamical time of the most distant knot tdyn =1.6â106 yr. Both the methods give similar numbers so we assume tdyn 106 yr for the age of the outflow and we estimate a mass loss rate M =5.7â10-4 M yr-1 , a momentum flux F =5.4â10-4 M km s-1 yr-1 and a luminosity L=8â10-2 L. These estimates rely on the assumption that the gas is ejected from the forming star and flows at a constant velocity. However, if the massive outflow is generated by the fragmentation and the acceleration of the hosting cloud, as we discussed above, most of the lobe material is clearly not ejected gas from the source and has not traveled along the

full extent of the lobe. In this case the dynamical time will be an upper limit of the real age of the outflow. This also affects all the physical parameters of the outflow which depend on the time, namely the mass loss rate, the flux momentum and the luminosity. The derived outflow mass is a factor 1.24 larger than the value that Yamashita et al. (1989) quoted from a smaller CO (1-0) map (3â6 arcmin2 ) and by calculating the optical depth only from three pointings of 13 CO along the flow axis. They did not identify the two ma jor CO components and calculated the outflow parameters by integrating the blueshifted gas over a velocity range of 6-10.5 km s-1 and the red-shifted gas over 14.5 -19 km s-1 range, moreover they assumed an excitation temperature of 40 K, 3.3 times higher than our value. Only two of our CO knots are inside the area mapped by Yamashita et al., namely D and E and they can be identified also in the Fig. 2 of their paper. The momentum and the energy we derived, assuming the same inclination angle (i = 3 ), are respectively a factor 1.9 and 47 higher than the values given by Yamashita et al. (1989). The outflow parameters we derived are not corrected for the inclination of the flow axis with respect to the plane of the sky. In fact, the measurement of the inclination angle suffers of large uncertainty and very different estimates can be found in the literature: Yamashita et al. (1989) deduced i 3 from a geometrical model, while Heathcote et al. (1998) deduced i 56 based on modeled position-velocity diagrams of the optical lines of the HH ob jects. Of course the smaller the inclination angle is, the larger the correction of the flow velocity is. In the case of i = 3 , the outflow momentum would be p 104 M km s-1 the energy E =3.7â1048 ergs, the momentum flux F =1â10-2 M km s-1 yr-1 and kinetic luminosity L=30 L while in the case of i = 56 , p=647 M km s-1 , E =1.5â1046 ergs, F =6.5â10-4 M km s-1 yr-1 and L=0.12 L . The comparison of our outflow map in the CO (1-0) line (Fig. 14) with the models by Cabrit & Bertout (1986) indicates a small inclination angle (i<50 ) and a wide opening angle ( 40 ). Comparing the values of the outflow physical parameters corrected for the possible inclination angle range (see Table 3) with the estimates of the same physical quantities derived for the radio jet M =6â10-6 M yr-1 , F =6â10-3 M km s-1 yr-1 and L=500 L (Mart´ et al. i 1995), we conclude that in the case of i 3 the jet power is less to that required to put into motion the observed molecular outflow while in the case of i 56 the jet is 10 times more powerful than the outflow. If we suppose that the ionc 2002 RAS, MNRAS 000, 1­??


The HH80-81 region
ized jet is the responsible for the acceleration of the outflow, imposing the identity of the momentum flux of the jet and the momentum flux of the molecular outflow, we derive an inclination angle of 5 . Despite a definite estimate of the inclination of the outflow cannot be given, our data would seem to favor a small inclination angle for various reasons: i) a comparison of the morphology of the outflow with the models of Cabrit & Bertout (1986) suggests an inclination angle 50 ; ii) the component of the velocity of the CO (10) line along the line of sight is low (on-source FWZI=20 km s-1 ) compared to the very high proper motion velocity of the HH ob jects (<600 km s-1 ) and the radio knots (< 1400 = = km s-1 ) of the flow; iii) the momentum and the energy of the outflow corrected for the inclination angle, when compared with the typical values derived for high mass YSOs (Beuther et al. 2002; Zhang et al. 2001), seem more plausible in the case of a small inclination angle; iv) if the radio jet is the responsible for the acceleration of the outflow, the inclination angle is 5 . The properties that we derive for the IRAS 18162-2048 outflow, qualify it as one of the most massive, energetic and extended outflows; the extreme excitation condition found in the visible HH ob jects associated with it (Heathcote et al. 1998) do not contradict our results. In principle, one would expect that the molecular gas associated to the high velocity jet and HH ob jects has itself a high velocity. However we have shown that from an energetic point of view, if the flow is close to the plane of the sky, the jet could be the responsible for its acceleration and given the large mass involved, the velocity of the large scale molecular outflow could be as low as we observe. This outflow shows remarkable similarities with respect to others high mass outflows (e.g. IRAS 20126+4104 (Shepherd et al. 2000), W75 N (Shepherd et al. 2003a)) in its low collimation degree and complex velocity structure with presence of high velocity clumps. All the observed lines show an intensity that drops off in the west side of the map, indicating that IRAS 18162-2048 is at the edge of its hosting cloud and that the red lobe of the outflow, flowing in the south-west direction, has broken out of the cloud, while the blue lobe, flowing in the north-east direction, is propagating inside the cloud. This is confirmed by the fact that the two southern HH ob jects, HH80 and 81, are visible in the optical while the northern counterpart, HH80N, has not been detected at optical wavelengths. 4.2 The dense core

9

velocity diagram obtained along the ma jor axis of the detected structure and across the peal position, shows two maxima at the opposite sides of the peak, at a reciprocal distance of 60 arcsec, (corresponding to 1.02â105 AU) and at velocities of 11.9 and 12.3 km s-1 respectively (see Fig. 11). A similar behavior is shown by CH3 OH. This feature could be interpreted as the signature of a rotating disk, however the distance between the is too large to be attributed to a disk. It is more likely that they are produced by velocity gradients inside the cloud suggesting a flattened rotating structure. Assuming a rigid rotation around the peak we derived an angular velocity 2.6â10-14 s-1 , a value of the same order of magnitude of that found by Olmi & Testi (2002) for the Serpens molecular cloud. Photometric and spectroscopic millimeter observations at higher angular resolution will be necessary to study in more detail this area which is a likely site of a star formation event in a very early stage.

4.3

Dispersal of Circumstellar Matter in IRAS 18162-2048

In spite of the different critical densities, C18 O and CH3 OH trace a similar structure for the dense core. Their integrated emission (Fig. 9 and 10) is elongate in the east-west direction and peaks at 30 arcsec north-east with respect to IRAS 18162-2048. Although the offset is almost within the uncertainty set by the beam size and 45 arcsec spacing of the observations for the C18 O, it is relevant for the methanol which has been sampled at a resolution of 22.5 arcsec. However, this offset can be ascribed to the higher intensity of the northern, blue component which shift the peak of the integrated line flux to the north-east direction (see Fig. 6). A second peak at 100 arcsec south-east the IRAS source has been observed in both C18 O and CH3 OH. The 2MASS database reports a red source in the same position, only visible in the K-band image. The C18 O positionc 2002 RAS, MNRAS 000, 1­??

The case of IRAS 18162-2048 is particularly interesting as this source shares the same luminosity as the most massive of the Herbig Be systems and related young stellar clusters (Testi et al. 1998; 1999) and of many intermediate- and high-mass protostellar systems (Molinari et al. 2002). This source, however, appears in an intermediate stage between the most embedded protostellar phase and the optically revealed young stellar ob ject. This is supported by the fact that the young stellar cluster in IRAS 18162-2048 is still deeply embedded into a dense molecular cloud, as for young (proto-)stellar ob jects, while the estimated outflow dynamical age is comparable with the ages of the youngest Herbig Ae/Be stars (Testi et al. 1998). Thus the IRAS 18162-2048 system represent a very good candidate for the study of the late stages of intermediate-mass (proto-)stellar evolution and the clearing of circumstellar material. Recently, Fuente et al. (2002) have studied the history of mass dispersal around Herbig Ae/Be stars, and concluded that 90% of the circumstellar material should be dispersed by powerful outflows in the early stages of stellar evolution, before the ob jects become optically visible. Our findings are in very good agreement with their conclusion, in fact we find evidence for a very powerful outflow that at tdyn 105 - 106 yrs is dispersing the parental cloud surrounding IRAS 18162-2048. This picture is also consistent with the morphology of the inner dense core which is elongated perpendicular to the outflow axis. The situation of IRAS 18162-2048 appears very similar to that of lower mass systems where outflows appear to be the key factor in the clearing of circumstellar material (see e.g. the case of B5­ IRS1 discussed in Velusamy & Langer (1998)).

5

CONCLUSIONS

We presented a multiwavelength spectroscopic mapping of a wide area surrounding the IRAS 18162-2048 source: by means of CO and 13 CO (1-0) transitions, a large scale bipolar outflow has been observed while a dense core has been traced by the C18 O and CH3 OH lines. All the tracers we


10

M. Benedettini et al.
Beuther H., Schilke P., Sridharan T.K., et al., 2002, A&A, 383, 892 Blake G.A., Sutton Masson C.R., Phillips T.G., 1987, ApJ, 315, 621 Cabrit S., Bertout C., 1986, ApJ, 307, 313 Cardelli J.A., Meyer D.M., Jura M., Savage B.D., 1996, ApJ, 467, 334 Caselli P., Hasegawa T.I., Herbst E., 1993, ApJ, 408, 548 Cesaroni R., Felli M., Tesi L., Walmsley M., Olmi L., 1997, A&A, 325, 725 Cesaroni R., Felli M., Jenness T., et al., 1999, A&A, 345, 949 Fuente A., Mart´ in-Pintado J., Bachiller R., Ro dr´ iguez-Franco A., Palla F., 2002, A&A, 387, 977 Heathcote S., Reipurth B., Raga C., 1998, AJ, 116, 1940 Girart J.M., Estalella R., Viti S., Williams D.A., Ho P.T.P., 2001, ApJ, 562, L91 Mart´ J., Ro dr´ i iguez L.F., Reipurth B., 1993, ApJ, 416, 208 Mart´ J., Ro dr´ i iguez L.F., Reipurth B., 1995, ApJ, 449, 184 Molinari S., Noriega-Cresp o A., Spinoglio L., 2001, ApJ, 547, 292 Molinari S., Testi L., Ro dr´ iguez L.F., Zhang Q., 2002, ApJ, 570, 758 Olmi L., Testi L., 2002, A&A, 392, 1053 Richer J., et al. 2000, in "Protostars and Planets IV" V. Mannings, A. Boss and S.S. Russell eds., (Tucson: University of Arizona Press), p. 867 Ro dr´ iguez L.F., Moran J.M., Ho P.T.P., Gottlieb E.W., 1980, ApJ, 235, 845 Shepherd D.S., Churchwell E., Wilner D.J., 1997, ApJ, 482, 355 Shepherd D.S., Yu K.C., Bally J., Testi L., 2000, ApJ, 535, 833 Shepherd D.S., Stark D., Testi L., 2003a, ApJ, 584, 882 Shepherd D.S. 2003b, in "Galactic Star Formation Accross the Mass Spectrum", J. de Buizer ed., ASP Conf. Series, in press Shu F.H., Adams F., Lizano S., 1987, ARAA, 25, 23 Stecklum B., Feldt M., Richichi A., Calami G., Lagage P.O., 1997, ApJ, 479, 339 Testi L., Palla F., Natta A. 1998, A&AS, 133, 81 Testi L., Palla F., Natta A. 1999, A&A, 342, 515 Velusamy T., Langer W.D., 1998, Nature, 392, 685 Yamashita T., Sato S., Nagata T., et al., 1987, Astr. Ap., 177, 258 Yamashita T., Hiroko S., Kaifu N., et al., 1989, ApJ, 347, 894 Zhang Q., Hunter T.R., Brand J., et al., 2001, ApJ, 552, L167

have observed reveals the presence of two ma jor line components, defining two large-scale structures with a coherent spatial distribution. The blue component is centered at 11.9 km s-1 , and prevails in the northern part of the map while the red one is centered at 12.8 km s-1 and prevails in the southern part of the map. We suggest that the powerful jet ejected by IRAS 18162-2048 has broken the surrounding molecular cloud apart, giving rise to a giant outflow where the two observed structures can be considered the two lobes. The axis of the large scale molecular outflow is coincident with the radio jet ejected by IRAS 18162-2048 and if the inclination angle of the outflow is very low (i=5 ), the jet could have enough power to accelerate the flow. The total mass put into motion is 570 M . The physical parameters of the outflow we calculated (see Table 3) are affected by two ma jor sources of uncertainty. The first one derives from the age of the outflow and the second one from the inclination angle. Despite a precise value of the two quantities can not be given, we used different methods to derive a reliable estimate. Our data show few evidences in favor of a small inclination angle and of a maximum outflow age of 106 yr. C18 O and CH3 OH trace the dense core surrounding IRAS 18162-2048 and show an elongated emission in the direction perpendicular to the outflow axis. Besides the ma jor peak of the emission correspondent at the IRAS source, we found a new peak at the position RA(B1950)=18h 16m 20.2s DEC(B1950)=-20 49 18 coincident with a red near infrared source. We provided evidence that this second peak may be surrounded by a flattened rotating structure, suggesting that the newly discovered IR source can be another site of recent star formation in this region. From our data IRAS 18162-2048 appears to be at an intermediate stage of its pre Main Sequence evolution, where the powerful wind/outflow is breaking apart the parent molecular cloud and dispersing the circumstellar material. The high mass protostellar ob ject is thus on the verge of becoming an optically visible young stellar cluster, similar to those associated with the most massive Herbig Be systems.

ACKNOWLEDGMENTS MB is grateful to Riccardo Cesaroni and Felipe Mac-Auliffe for their kind assistance during the observations at SEST. ANC's research was partially supported by NASA-APD Grant NRA0001-ADP-096 and by the Jet Propulsion Laboratory, Caltech. This publication makes use of the data products from the Two Micron All Sky Survey, which is a joint pro ject of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation.

REFERENCES
Aspin C., Geballe T.R., 1992, A&A, 266, 219 Aspin C., Puxley P.J., Blanco P.R., et al., 1994, A&A, 292, L9 Bachiller R., Cernicharo J., Martin-Pintado J., Tafalla M., Lazareff B., 1990, A&A, 231, 174 Bachiller R., Liechti S., Walmsley C.M., Colomer F., 1995, A&A, 295, L51 c 2002 RAS, MNRAS 000, 1­??