Mon. Not. R. Astron. Soc. 341, 707–716 (2003)

Shocked gas around Cepheus A: evidence for multiple outflowsfrom H2S and SO2 observations

C. Codella,1� R. Bachiller,2 M. Benedettini3 and P. Caselli41Istituto di Radioastronomia, CNR, Sezione di Firenze, Largo E. Fermi 5, 50125 Firenze, Italy2Observatorio Astronomico Nacional (IGN), Apartado 1143, E-28800, Alcala de Henares, Madrid, Spain3Istituto di Fisica dello Spazio Interplanetario, CNR, Area di Ricerca Tor Vergata, Via Fosso del Cavaliere 100, 00133 Roma, Italy4INAF, Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125 Firenze, Italy

Accepted 2003 January 21. Received 2003 January 17; in original form 2002 August 28

ABSTRACTThe Cepheus A star-forming region has been investigated through a multiline H2S and SO2

survey at millimetre wavelengths. Large-scale maps and high-resolution line profiles reveal theoccurrence of several outflows. Cep A East is associated with multiple mass-loss processes:in particular, we detect a 0.6-pc jet-like structure which shows for the first time that the Cep AEast young stellar objects are driving a collimated outflow moving towards the south.

The observed outflows show different clumps associated with definitely different H2S/SO2

integrated emission ratios, indicating that the gas chemistry in Cepheus A has been altered bythe passage of shocks. H2S appears to be more abundant than SO2 in high-velocity clumps, inagreement with chemical models. However, we also find quite small H2S linewidths, suggestiveof regions where the evaporated H2S molecules had enough time to slow down but not to freezeout on to dust grains. Finally, comparison between the line profiles indicates that the excitationconditions increase with the velocity, as expected for a propagation of collimated bow shocks.

Key words: ISM: clouds – ISM: individual: Cep A – ISM: jets and outflows – ISM: molecules– radio lines: ISM.

1 I N T RO D U C T I O N

Sulphur is relatively abundant ([S]/[H] � 10−5: e.g. Wilson & Rood1994) and is able to form a wide variety of volatile compounds whichhave been observed in molecular clouds and, in particular, in star-forming regions (SFRs). Even if gas-phase ion–molecule chemicalmodels (see e.g. Lucas & Liszt 2002, and references therein) haveshown that the abundance of S-bearing compounds such as SO andSO2 can be enhanced in low-ionization phases (Le Bourlot et al.1995) or where the C/O ratio is low (Swade 1989; Gerin et al.1997; Nilsson et al. 2000), sulphur chemistry can be seriously af-fected by grain surface reactions. S-bearing molecules are thereforethought to play a major role in the chemistry of the high-densitygas associated with the star-forming process. In particular, H2S isexpected to be formed on grains in molecular outflows driven byyoung stellar objects (YSOs). H2S molecules are injected into thegas phase by the occurrence of shock waves leading to a fast pro-duction of SO and SO2 (Pineau des Forets et al. 1993; Charnley1997).

The importance of sulphuretted molecules for the investigationof molecular outflows has been recently confirmed by millimetre-wavelength observations provided by high spatial resolution surveys

�E-mail: codella@arcetri.astro.it

of the CB3 (Codella & Bachiller 1999) and L1157 (Bachiller et al.2001) SFRs where the H2S and SO2 abundances definitely increasein the outflow lobes. In particular, comparison with the emissionarising from SiO, a standard tracer of the high-velocity jet-like innerpart of the outflow, reveals that the S-bearing species trace differenthigh-density outflow components, pointing out more extended re-gions produced by the interaction of the outflow with the ambientmedium. In other words, there is an indication that the H2S and SO2

species give a different picture of molecular outflows, playing anintermediate role between SiO and CO. Thus S-bearing moleculescan be useful tools to investigate the chemistry and kinematics ofcomplex SFRs, deeply embedded in dense molecular clouds andhosting multiple star formation.

Cepheus A is a good example of this kind of SFR: it is locatedat 725 pc near the Cepheus OB3 stellar association (Sargent 1977),it is associated with a far-infrared 2.5 × 104 L� source, clusters ofOH and H2O masers, shock-excited H2 emission and dense molec-ular clouds (see Bally & Lane 1982; Cohen, Rowland & Blair 1984;Lenzen 1988; Torrelles et al. 1993; Narayanan & Walker 1996;Bergin, Snell & Goldsmith 1996; Bergin et al. 1997; Goetz et al.1998, and references therein), and represents one of the best labo-ratories in which to study how the high-mass star-forming processinteracts with the ambient medium.

Cepheus A is associated with two regions where the star forma-tion is active: Cep A East and Cep A West. Cep A East is associated

C© 2003 RAS

708 C. Codella et al.

with 15 ionized sources (Hughes & Wouterloot 1984; Garay et al.1996) aligned in three strings (HW1, HW4–5–6 and HW7) proba-bly produced by shock waves driven by the YSOs HW2 and HW3.The East component displays CO outflows and in particular threevelocity components: (i) a low-velocity, bipolar, poorly collimatedcomponent, quite extended (over a 4 × 4 arcmin2 area), with ageneral north-east–south-west direction and an ill-defined struc-ture; (ii) a high-velocity (v − vLSR � 5–10 km s−1) flow whichis bipolar, symmetrically located with respect to the HW2 posi-tion and oriented east–west (e.g. Hayashi, Hasegawa & Kaifu 1988;Bergin et al. 1997); and (iii) an extremely high-velocity component(�20 km s−1) which also lies in the north-east–south-west direction(Choi, Evans & Jaffe 1993; Narayanan & Walker 1996). Possibleexplanations of the presence of different outflow axes have beenproposed by Torrelles et al. (1993), who found evidence that NH3

high-density clumps located around the HW2 coordinates are redi-recting an outflow into a quadrupolar structure; and by Narayanan& Walker (1996), who supported the occurrence of generation ofmultiple outflows. In conclusion, CO data reveal intense outflow ac-tivity roughly located along the east–west direction but they are notable to give a clear picture of the kinematics. It is worth noting thatsuch a complex region has not been unravelled by using SiO emis-sion: millimetre observations (Codella, Bachiller & Reipurth 1999)showed high-velocity SiO emission confined in a sort of clumpyfilament located in the north-east–south-west direction, and did nottrace all the velocity components observed using CO.

Cep A West, located ∼1.5 arcmin west of Cep A East, is asso-ciated with three centimetre sources (e.g. Garay et al. 1996) andHerbig–Haro (HH) object GGD 37 (Gyulbudaghian, Glushkov &Denisyuk 1978; Hartigan et al. 1986; Hughes & Moriarty-Schieven1990). CO observations revealed a molecular outflow with twooverlapping blue- and redshifted lobes (e.g. Bally & Lane 1991;Narayanan & Walker 1996). There is no consensus on the drivingsource of the activity in this region. Lenzen (1988) suggested thatCep A West is shock-excited by the wind of one of the radio sourceslocated at the centre of the Cep A East cluster. On the other hand,Hartigan & Lada (1985), Garay et al. (1996) and Raines et al. (2000)claimed that the West region is an independent centre of activity.Also in this case, SiO observations do not lead to a clear detectionof the structure of the outflow motion (Codella et al. 1999).

We report here large-scale maps (�5 × 4 arcmin2) in severaltransitions of H2S and SO2 of Cepheus A. The goal of the presenthigh angular and velocity resolution observations is twofold: (i) todisentangle the puzzling kinematics of Cepheus A, giving a pictureof the molecular component interacting with the mass-loss processof the YSOs; and (ii) to investigate further the association of H2Sand SO2 with molecular outflows and to study how the occurrenceof such outflows alters the surrounding quiescent gas. Moreover, themaps will be used to locate shocked regions to be investigated infuture multiline observations aimed at putting constraints on chem-

Table 1. List of molecular species, transitions and observing parameters.

Molecules Transition Eu Rest frequency HPBW T sys t int dv(AC) dv(1 MHz)(K) (MHz) (arcsec) (K) (s) (km s−1) (km s−1)

SO2 J K−K+ = 313–202 8 104 029.414 24 120 480 0.23 2.88SO2 J K−K+ = 16214–15313 138 104 033.582 24 120 480 0.23 2.88H2S J K−K+ = 110–101 28 168 762.781 14 440 480 0.14 1.78H2S J K−K+ = 220–211 84 216 710.437 11 550 480 0.11 1.38SO2 J K−K+ = 524–413 24 241 615.797 10 670 480 0.10 1.24

ical models of shocked regions associated with the star-formingprocess.

2 O B S E RVAT I O N S

The observations were carried out with the IRAM 30-m telescopeat Pico Veleta (Granada, Spain) in 2000 October and 2001 June.Table 1 summarizes the observed molecular species, the transitions,their upper level energies and rest frequencies and some observingparameters, such as the half-power beamwidth, the typical systemtemperature (T sys) and the integration time (t int; on- plus off-source).The main-beam efficiency varies from about 0.8 (at 104 GHz) to 0.5(at 242 GHz). The observations were made by position-switching.The pointing was checked every hour by observing nearby planetsor continuum sources, and it was found to be accurate to within4 arcsec. As spectrometers, an autocorrelator (AC) split into fourparts was used to allow simultaneous observations of four differenttransitions. Moreover, a 1-MHz filter bank, split into four parts of256 channels, was simultaneously used. The velocity resolutionsprovided by both backends, AC and 1-MHz, are shown in Table 1.When necessary, the AC spectra were smoothed to a lower velocityresolution (up to 0.90 km s−1). It is worth noting that the 104-GHzobservations have allowed us to detect both the J K−K+ = 313–202

and 16214–15313 SO2 lines, which are blended at the given spectralresolution. The spectra were calibrated with the standard chopperwheel method and are reported here in units of main-beam brightnesstemperature (T MB).

3 R E S U LT S

3.1 Overall H2S and SO2 distributions

Fig. 1 shows the maps of the integrated emission arising, for eachmolecule, from the lowest excitation transition among those ob-served, i.e. H2S 110–101 (upper panel) and SO2 313–202 (lowerpanel). The maps are centred (as well as the other ones presented inthis paper) at the HW2 coordinates: α2000 = 22h56m17.s9, δ2000 =+62◦01′49.′′7. Since the SO2 313–202 line is blended with the 16214–15313 one, we performed two Gaussian fits and chose the velocity atwhich the two profiles intersect to disentangle the observed spectraand to plot the spatial distributions of both lines. Fig. 2 reports, asan example, the molecular line profiles observed towards the Eastclump, at the (0 arcsec, 0 arcsec) position.

The H2S map brings to light an extended and irregular cloudaround Cep A East with size 0.5 × 0.7 pc2, and a 0.2-pc H2S cloudtowards Cep A West. Besides the well-known East and West ob-jects, there are several maxima suggesting clumpy structures, whichwill be kinematically analysed in Section 3.2. On the other hand,the SO2 structure is definitely less extended with respect to theH2S one. A 0.4 × 0.1 pc2 elongated SO2 cloud located along the

C© 2003 RAS, MNRAS 341, 707–716

Shocked gas around Cepheus A 709

Figure 1. Contour maps of the integrated H2S J K−K+ = 110–101 (upperpanel) and SO2 J K−K+ = 313–202 (lower panel) emission towards Cep A.The velocity integration intervals are −25, +5 and −16, +5 km s−1 for theupper and lower panels, respectively (see text). The empty circles show theIRAM beam (HPBW), while the small crosses mark the observed positions.The triangles denote the VLA 2-cm continuum components which trace inthe East region two strings of sources arising in shocks (Garay et al. 1996).The labels indicate the East and West components and the four H2S and SO2

clumps (A, B, C and D; see also the channel maps of Figs 3 and 4). Thecontours range from 1.4 to 22.4 K km s−1 (upper panel) and from 0.3 to2.7 K km s−1 (lower panel). The first contours and the steps correspond toabout 5σ (where σ is the rms of the map).

north-east–south-west direction showing a peak centred on the HW2coordinates has been observed. In addition to this East component,we also note emission associated with Cep A West and towards thesouthern direction. Finally, Fig. 1 suggests an additional SO2 struc-ture elongating from Cep A East towards the north-west. In sum-mary, the comparison between the H2S and SO2 spatial distributionsof Fig. 1 suggests that these species trace different gas componentsassociated with different physical and/or chemical conditions.

Fig. 3 shows the spatial distribution of the emission arising fromthe transitions at higher excitation (see Table 1) with respect thoseof Fig. 1. The contour maps of the integrated H2S 220–211 (upperpanel), SO2 524–413 (middle panel) and 16214–15313 (lower panel)emissions are reported. For the sake of clarity, only a close-up ofthe central region of Cepheus A has been plotted since, althoughsuch lines have been searched for towards the same region and with

Figure 2. Examples of H2S and SO2 line profiles observed towards theCep A East component (see text). Molecular species, transitions (see Table 1)and angular offset are indicated.

the same sampling as Fig. 1, we detected emission only towardsthe clump associated with Cep A East. At the H2S 220–211 and SO2

16214–15313 frequencies the clump is unresolved and thus only anupper limit on its size can be given: 0.04 and 0.08 pc, respectively.On the other hand, the SO2 524–413 emission allows us to observea roundish structure with a beam-deconvolved size of ∼14 arcsec,corresponding to about 0.05 pc. From Fig. 3 it is possible to see thatthe unresolved spatial distributions of the high-excitation transitionsH2S 220–211(Eu = 84 K) and SO2 16214–15313(Eu = 138 K) coincidewith the HW2–HW3 objects and are not associated with the threestrings tracing the confined jets of ionized gas. This result indicatesthat only towards the central YSO cluster of Cep A East can wefind the high-excitation conditions needed to observe such lines.

C© 2003 RAS, MNRAS 341, 707–716

710 C. Codella et al.

Figure 3. Close-up of the contour maps of the integrated H2S J K−K+ =220–211 (upper panel), SO2 J K−K+ = 524–413 (middle panel) and 16214–15313 (lower panel) emission of the central region of Cep A, where suchemissions have been detected. The velocity integration intervals are −20,+1, −20, 0 and −18, −4 km s−1 for the upper, middle and lower panels,respectively (see text). Symbols are drawn as in Fig. 1. The contours rangefrom 1.7 to 3.4 K km s−1 (upper panel), from 1.7 to 11.9 K km s−1 (middlepanel) and from 0.3 to 1.2 K km s−1 (lower panel). The first contours andthe steps correspond to about 5σ .

This picture is in agreement with the results of Bergin et al. (1994,1996), who mapped the region around Cep A East in HC3N andCH3C2H and found that in Cep A East both the kinetic temperatureand the hydrogen density distributions increase towards the centralposition, where the ionized Very Large Array (VLA) sources arelocated.

3.2 Kinematics

3.2.1 The H2S and SO2 outflows

In order to study the gas kinematics, we present in Fig. 4 the veloc-ity channel maps of the H2S 110–101 emission. The ambient localstandard of rest (LSR) velocity is −10.6 km s−1 according to C18O

J = 2–1 observations that we have performed at 219.56 GHz usingthe 30-m IRAM antenna. From Fig. 4 it is possible to see a verycomplex structure around Cep A East: (i) redshifted emission, upto −5 km s−1, is coming from a well-defined jet-like structure ex-tending 0.6 pc towards the southern direction; (ii) a less collimatedblueshifted emission until about −13 km s−1 is detected towards thesouth; (iii) high-velocity (up to −18 and −4 km s−1) blue- and red-shifted overlapping emissions are associated with a 0.3-pc elongatedcloud lying in the south-easterly direction. The −15 and −7 km s−1

panels of Fig. 4 well represent what is described above. We concludethat Cep A East is associated with at least two molecular outflowsdriven by the YSOs of the central cluster: that pointing towards thesouth, hereafter called the southern outflow, and that flowing fromthe central YSO cluster towards the south-east, hereafter called theeastern outflow.

Moreover, the low-velocity panels, between −12 and −9 km s−1,show blue- and redshifted emission coming from a bipolar structuresymmetrically located with respect to the HW2–HW3 coordinatesalong the north-east–south-west direction (see also Fig. 6, later).Although these velocities are quite close to the ambient one, takinginto account that such H2S emission (i) comes from an elongatedstructure which lies along the direction traced by the ionized con-tinuum sources and (ii) reflects the low-velocity bipolar CO outflowtraced by Hayashi et al. (1988), it is tempting to conclude that weare observing a third outflow, hereafter called the central outflow.

Fig. 4 shows also that red- and blueshifted high-velocity emis-sion, up to −15 and −3 km s−1 respectively, is detected towardsCep A West, indicating a molecular outflow, hereafter called thewestern outflow. The present data do not show any connection ofsuch emission with the Cep A East cloud, and cannot confirm if itis a fourth outflow associated with Cep A East or if it is driven by aYSO belonging to Cep A West.

Fig. 5 reports the channel maps of the SO2 313–202 emission. Thevelocities more negative than −14 km s−1 are not plotted because ofthe blend with the 16214–15313 line at the central position and becauseno SO2 313–202 emission at these velocities is observed outside CepA East. From Fig. 5, we can argue a situation basically consistentwith that given by H2S: a structure elongated towards the south, an-other one located along the north-west–south-east direction, and twoclumps associated respectively with Cep A East and Cep A West areobserved. However, the SO2 picture differs from the H2S one in threeaspects. First, the high-velocity gas flowing from the HW2–HW3coordinates towards the south-east is missing. Secondly, slightlyredshifted emission pointing towards the north-west is clearly de-tected in SO2, whereas it is ill-defined using H2S (see the −9 kms−1 panels of Figs 4 and 5). Finally, the H2S and SO2 emissionsassociated with Cep A West peak at definitely different positions:(−130 arcsec, +20 arcsec) and (−90 arcsec, −20 arcsec), respec-tively.

In conclusion, the comparison of the H2S and SO2 channel mapsconfirms what is suggested by the overall structures and indicatesthat these molecules can trace different components of gas associ-ated with the mass-loss process.

3.2.2 High-velocity clumps

The H2S and SO2 channel maps reveal the occurrence of differentclumps at different velocities. Along the eastern outflow the H2Sallows us to detect a high-velocity red- and blueshifted clump,hereafter called A, located at (+40 arcsec, −20 arcsec). Moreover,a blue H2S clump, called B, is located at (+80 arcsec, −80 arcsec).

C© 2003 RAS, MNRAS 341, 707–716

Shocked gas around Cepheus A 711

Figure 4. Channel map of the H2S J K−K+ = 110–101 emission towards Cep A. Each panel shows the emission integrated over a velocity interval of 1 km s−1

centred at the value given in the upper left corner. Symbols are drawn as in Fig. 1, while the labels denote the different H2S clumps (see text). The ambientvelocity emission is −10.6 km s−1 according to C18O J = 2–1 measurements (see text). The contours range from 0.25 (∼5σ ) to 3.75 K km s−1 in steps of0.25 K km s−1.

Figure 5. Channel map of the SO2 J K−K+ = 313–202 emission towards Cep A. Each panel shows the emission integrated over a velocity interval of 1 km s−1

centred at the value given in the upper left corner. Symbols are drawn as in Fig. 1, while the labels denote the different SO2 clumps (see text). The ambientvelocity emission is −10.6 km s−1 according to C18O J = 2–1 measurements (see text). The contours range from 0.08 (∼5σ ) to 0.56 K km s−1 in steps of0.08 K km s−1.

C© 2003 RAS, MNRAS 341, 707–716

712 C. Codella et al.

Given the complex morphology of Cep A East, it is not clear if clumpB is in fact a signature of the eastern outflow or if it traces anotheroutflow activity. However, if we consider the highest velocities,plotted in the −15 km s−1 panel of Fig. 4, it is possible to see thatHW2–HW3, A and B are aligned along the same direction, whichis coincident with that traced by the ionized sources reported byGaray et al. (1996). This supports the association of clump B withthe eastern outflow.

The H2S channel maps of the southern outflow suggest the occur-rence of more than one clump along its axis. The redshifted emissionplotted in the −5 km s−1 panel shows three peaks at (0, −10 arcsec),(0, −80 arcsec) and (−20 arcsec, −140 arcsec). In particular, thelatter emission comes from a well-defined clump, hereafter calledC, which is located at the end of the main outflow axis and thuscould trace the interaction between the outflowing gas and the am-bient medium. It is worth noting that the H2S clump C has an SO2

counterpart, which is located at the same position but emits mainlyat the ambient LSR velocity.

In summary, we find at least three compact structures (high-lighted in the channel maps) along the eastern and southern out-flows: A (+40 arcsec, −20 arcsec), B (+80 arcsec, −80 arcsec) andC (−20 arcsec, −140 arcsec).

Finally, Fig. 5 shows the occurrence of a SO2 clump, hereaftercalled D, at (−20 arcsec, +40 arcsec). Since this clump is associatedwith emission quite close to the ambient velocity, between −10 and−8 km s−1, the present data do not allow us to claim that this is a sig-nature of an outflow. Actually, clump D is spatially coincident withthe high-density NH3 clump called Cep A-2 detected at the samevelocities with the VLA interferometer by Torrelles et al. (1993).It is worth noting that Hayashi et al. (1988), using CO, observed acompact structure emitting at −8 and −7 km s−1 (see their fig. 3)where clump D has been detected, whereas at velocities closer tothe ambient one the CO emission is found to be strongly absorbedand thus does not trace the structure of the cloud.

4 D I S C U S S I O N

4.1 Multiple outflows driven by the Cep A East YSOs

The results show the presence of several outflows associated withCepheus A and, in particular, with Cep A East. Trying to summarizethe complex morphologies, in Fig. 6 we have focused our attentionon the most significant channel maps of Fig. 4 and compared the H2Semission with the ionized VLA jets observed by Garay et al. (1996).Shown are (i) the low-velocity central outflow along the north-east–south-west direction (the two upper panels), (ii) the high-velocityeastern outflow from the central YSO cluster to the south-east (themiddle panel) and (iii) the high-velocity southern outflow south ofHW2–HW3 (the two lower panels). The central outflow had alreadybeen detected using CO emission (Hayashi et al. 1988; Bergin et al.1997) and is well in agreement with the north-east–south-west VLAjets.

The eastern outflow reveals the molecular millimetre counterpartof the mass-loss process detected using H II continuum emission.It is worth noting that Goetz et al. (1998), using the H2 and [Fe II]emission respectively at 2.12 and 1.64 µm, identified a J-type shocklocated at the head of the ionized jet flowing towards the south-east,e.g. where the intense H2S clump A is located (see their fig. 7). Thisstrongly supports the theoretical models where the H2S moleculesare injected from grains into the gas phase by shocks.

The southern outflow allows us to map high-velocity gas flow-ing from the HW2–HW3 coordinates towards the south. Previous

Figure 6. Summary of the directions of the multiple outflows driven bythe Cep A East YSOs as traced by the H2S J K−K+ = 110–101 emission (seetext). Symbols are drawn as in Fig. 1. Each panel is a close-up of the wholechannel map reported in Fig. 4. At least three axes are detected: (i) the −12and −9 km s−1 panels display the north-east–south-west direction; (ii) the−15 km s−1 panel displays a south-easterly flow; while (iii) the −7 and−3 km s−1 panels clearly indicate the occurrence of an outflow towards thesouth.

CO observations had indicated the occurrence of redshifted gas inthe southern region (Hayashi et al. 1988; Bergin et al. 1997), butthe present H2S channel maps clearly show for the first time thatCep A East is associated with an intense and collimated mass-loss process directed towards the south. It is worth noting that,by analysing the CH3OH J K = 20–10 A+ and SO J K = 32–21

lower spatial and spectral resolution maps performed by Berginet al. (1997) using the 14-m Five College Radio Astronomy Ob-servatory (FCRAO) antenna, it is possible to see a structure indi-cating the southern direction. However, their data did not allow the

C© 2003 RAS, MNRAS 341, 707–716

Shocked gas around Cepheus A 713

authors to note indications of outflows or changes in the chemistryarising from the interaction between outflowing gas and quiescentmaterial.

In conclusion, the multiple mass-loss processes traced in thepresent maps confirm that the CO outflows detected in recent yearsalong the west–east and north-east–south-west directions are due toseparate outflows and are not generated by a single YSO. The H2Sand SO2 pictures suggest that material has been ejected by the CepA East YSOs over a wide opening angle. This result resembles thatreported by Allen & Burton (1993), who investigated the Orion SFRusing H2 and [Fe II] infrared emission and found that the OMC-1outflow contains several clumps and filaments which are thoughtto have been ejected in a sort of explosion by the BN–IRc2 region(see also the H2 images reported by Salas et al. 1999). In the Cep AEast case, the opening angle is more than 180◦: e.g. there are out-flows in the region south of the north-east–south-west and east–westdirections centred at the HW2–HW3 coordinates. One cannot ex-clude the possibility that also in Cepheus A the observed complexoutflow structures might be due to a seemingly explosive event,although such a speculation cannot be proved on the basis of thecurrent findings. An alternative explanation is that we are observingthe cumulative effects of quasi-continuous jets and winds driven byseparate YSOs over a considerable amount of time (∼105 yr; seebelow) Finally, it is still less clear what is happening in the northerncounterpart and, in particular, if the Cep A West outflow is due tothe East YSOs.

The occurrence of clumps at different velocities suggests thatepisodic increases of the mass-loss process from the central objectshave occurred. A rough estimate of the time elapsed between succes-sive ejections by the central YSOs can be obtained by comparing thepositions and the velocities of different knots located along the maindirection of the southern outflow (see Section 3.2.2) and highlightedby the −3 and −7 km s−1 (i.e. clump C) panels of Fig. 6. From its0.6-pc jet-like redshifted structure, we can derive a small inclinationto the plane of the sky: an angle between 10◦ and 30◦ is assumed.

Figure 7. Examples of H2S J K−K+ = 110–101 and SO2 313–202 line profiles observed towards the Cep A West component and the A, B, C and D clumps (seetext). Molecular species, transitions and angular offsets are indicated.

Taking into account the inclination, a time interval of about 104 yris measured. Finally, clump C, which is the farthest one, allows usto derive an estimate of the age of the outflow activity traced withH2S: 3–9 × 104 yr.

4.2 Properties of the high-velocity clumps

Whereas Fig. 2 reports the spectra of the region where all five H2Sand SO2 emission lines have been detected, i.e. Cep A East, Fig. 7 re-ports the spectra of the other five clumps (West, A, B, C and D) whichhave been observed through the low-excitation H2S and SO2 transi-tions. The six positions are characterized by different line intensitiesand the ratio between the integrated emissions (

∫TMB dv; K km s−1)

of the H2S 110–101 and SO2 313–202 lines, hereafter called R, variessignificantly. In Cep A East R is ∼2–6, but for this position wehave to be cautious because of the H2S line absorption. TowardsCep A West R is 8.5 and >16.4 at the SO2 and H2S peaks (seeSection 3.2.1), respectively. The R-values associated with the otherclumps are >26.1 (A), >9.4 (B), 16.6 (C) and 5.2 (D). The differentR-values as well as the association of H2S and SO2 with outflowsclearly show that the gas chemistry in Cepheus A has been def-initely altered by the occurrence of shocks. This is apparently incontrast with the results of Bergin et al. (1997), who, although theyobserved Cepheus A using tracers of shocked gas such as CH3OHand SO, did not find evidence of outflow chemistry. However, thesolution to this apparent disagreement can be found in the highersensitivity, spectral resolution and angular resolution of the presentobservations, which allow us to detect high-velocity emissionsclearly.

Additional information about the gas traced by H2S can be foundby analysing the line profiles. Figs 2 and 7 show that a variety ofH2S profile shapes exist: from quite narrow lines at velocities closeto the ambient one, like for clumps B (FWHM = 2.3 km s−1 for the110–101 line) and D (3.5 km s−1), to broad profiles at much highervelocities, like for the East and West emissions (�10 km s−1). If we

C© 2003 RAS, MNRAS 341, 707–716

714 C. Codella et al.

take into account the common belief that H2S is formed on grains andthen is injected into the gas phase as a consequence of shocks, thenthe detection of narrow H2S components appears to be somewhatcontradictory, since we would in principle expect the shocked gas toexhibit high velocities. An possible solution is to consider the narrowH2S emission as a signature of the evolution of the shocked gas:the H2S gas produced in high-velocity shocked regions is sloweddown because of the interaction with the ambient gas. This situationseems to resemble that found for a standard shocked gas tracer likeSiO by Lefloch et al. (1998) and Codella et al. (1999). The authorsobserved narrow and broad lines towards molecular outflows drivenby YSOs and associated narrow SiO profiles with evolved shockedgas. In this scenario the narrower H2S lines should reflect either oldershocked gas or higher density regions where the interaction withthe ambient medium is stronger. In any case, the set of H2S profilespresented here calls for further high angular and spectral resolutionobservations which are necessary to analyse the line profiles fully,and to check whether the idea of the slowed shocked gas used forSiO can be adapted also to H2S or whether we have to evoke analternative hypothesis regarding gas-phase chemical reactions orslow shocks which directly inject slowly moving H2S moleculesinto the gas component.

From the point of view of chemistry, the large H2S/SO2 ratioin high-velocity clumps can be explained with an inefficient con-version of H2S into SO2 via reactions with OH (Charnley 1997;Keane et al. 2001). This is possible if the region where the shockpropagates is heated above about 200 K, so that the majority of OHmolecules quickly produce H2O through reactions with molecularhydrogen (e.g. Hartquist, Oppenheimer & Dalgarno 1980). In thepost-shock region, the evaporated H2S molecules will have enoughtime to slow down at ambient velocities but not to freeze out on to

Figure 8. Distribution with velocity of the ratio between the brightness temperatures of the J K−K+ = 220–211 and 110–101 H2S lines (upper panel) and of theJ K−K+ = 16214–15313 and 313–202 SO2 lines (lower panel) as detected towards Cep A East. The dot-dashed line shows the ambient LSR velocity accordingto the C18O J = 2–1 emission (see text).

dust grains, thus explaining the possibility of detecting narrow H2Sprofiles which can be interpreted as a fossil signature of a shockpassage.

4.3 H2S and SO2 excitation as a functionof the emission velocity

Towards Cep A East, where the emissions arising from the fiveselected transitions have been detected, it is possible to comparethe line profiles corresponding to different excitations of the samemolecule.

Fig. 8 reports in the upper panel the distribution with velocityof the ratio between the brightness temperatures of the 220–211 and110–101 H2S lines, associated with an upper level energy of 84 and28 K, respectively. The two measurements for the velocities close tothe ambient value (marked by the dot–dashed line) are not reportedbecause of the 110–101 absorption. Even if the errors associated withthe measurements at the highest velocities do not allow us to drawdefinite conclusions, Fig. 8 suggests an increase of the excitation inthe wings.

These findings are supported by the three SO2 spectra, whichshow clear differences between the profiles. In particular, the redwing clearly becomes more prominent with increasing transitionenergy, and thus suggests that the higher the velocity the higher theSO2 excitation. This result is shown in the lower panel of Fig. 8,where the distribution with velocity of the T MB ratio of the 16214–15313 and 313–202 SO2 lines is reported. These lines are associatedwith very different excitation since they have an upper level energyof respectively 138 and 8 K. Similarly to H2S, the plotted ratio isabout 0.3 at the ambient velocity and increases by a factor of 3–4 atthe highest velocities.

C© 2003 RAS, MNRAS 341, 707–716

Shocked gas around Cepheus A 715

These results suggest that the strongest interaction between thematerial associated with the mass loss and the ambient mediumhappens at the highest velocities. Thus the present profiles sup-port those outflow models where there are collimated bow shockswhich drive the outflow propagation (e.g. Wilkin 1996; Zhang &Zheng 1997), the shock velocity being directed forward rather thanperpendicular to its line of propagation. These models have beencalculated to reproduce the basic observed feature of the outflows,since the previous general models (see Lada & Fich 1996, and refer-ences therein) behaved similarly to an isotropic wind, creating lessforward momentum than is observed and no bipolarity.

4.4 H2S and SO2 column densities

In order to have an estimate of the H2S excitation temperature (T ex)and column densities, optically thin emission and a local thermo-dynamic equilibrium (LTE) population have been assumed. By us-ing the integrated intensities of the two H2S lines detected towardsCep A East, we obtain a temperature of 27 K and a total columndensity N H2S = 3 × 1014 cm−2. However, the obtained temperaturehas to be considered as an upper limit because of the absorptionwhich affects the 110–101 profile. For the same reason the derivedcolumn density is expected to be underestimated.

With the same assumptions, we have calculated the temperatureand column density of SO2 by using the three lines detected towardsCep A East and the standard rotation diagram method. Since theSO2 profiles show intense wing emission, we have assumed twoSO2 components: one associated with the ambient velocities andanother one associated with high velocities. To disentangle theseemissions, the ambient component has been assumed to emit in thevelocity range defined by the C18O J = 2–1 line, the linewidthof which is about 3.7 km s−1. In this way, excitation temperaturesaround 75 and 130 K have been derived for the ambient and thewing components, respectively. The total column densities N SO2

are about 1015 cm−2.For the clumps detected in just one line of H2S and/or SO2,

we have estimated the total column densities assuming opticallythin emission and temperature ranges of 20–30 K for H2S and 20–130 K for SO2. The N H2S values obtained are �2–4 × 1013 cm−2

for clumps West, B, C and D, whereas in A N H2S � 8–9 × 1014

cm−2. The SO2 total column densities vary between ×1013 and 2 ×1014 cm−2 for clumps West and D, while N SO2 � 1–8 × 1013 cm−2

for clump C.

5 C O N C L U S I O N S

The Cepheus A star-forming region has been investigated througha multiline survey at millimetre wavelengths of H2S and SO2 emis-sions at different excitations. Large-scale maps (�5 × 4 arcmin2)have been obtained and the main results are the following.

(1) The lowest excitation transitions among those observed, e.g.the H2S 110–101 and SO2 313–202 lines, allow us to detect extendedand elongated clouds associated with Cep A East with sizes of about0.5 × 0.7 pc2 for H2S and 0.4 × 0.1 pc2 for SO2. Moreover, a0.2-pc cloud has been observed towards Cep A West in both lines.On the other hand, by using higher excitation transitions (up to anupper level energy of 138 K), we detected emission only towards anunresolved clump (�0.04–0.08 pc) associated with Cep A East. Inparticular, such emission is observed towards the HW2–HW3 co-ordinates, whereas it is not associated with the three strings tracingthe confined jets of ionized gas observed by Hughes & Wouterloot

(1984) and Garay et al. (1996). This shows that the excitation con-ditions are higher towards HW2 and HW3, i.e. toward the YSOsthought to drive the mass-loss process in Cep A East.

(2) Cep A East is associated with at least three molecular out-flows driven by the YSOs of the central cluster: (i) the low-velocitycentral outflow along the north-east–south-west direction, (ii) thehigh-velocity eastern outflow pointing to the south-east; and (iii)the high-velocity southern outflow moving towards the south. Inparticular, the latter outflow is associated with a well-defined jet-like structure of 0.6 pc which clearly shows for the first time thatCep A East is associated with an intense and collimated mass-lossprocess directed towards the south. On the other hand, a fourth out-flow has been detected towards Cep A West, but the present data donot allow us to confirm if this is associated with Cep A East or if itis driven by a YSO located in Cep A West.

(3) The observed outflows reveal different clumps along theiraxes, suggesting that episodic increases of the mass-loss processhave occurred. Different clumps are characterized by differentH2S/SO2 integrated emission ratios, showing that the gas chem-istry in Cepheus A has been definitely altered by the occurrence ofshocks. The comparison between the line profiles detected towardsCep A East indicates that the excitation increases with the veloc-ity. The observed lines show that the maximum interaction betweenthe mass-loss process and the surrounding medium happens at thehighest velocities, supporting the propagation of collimated bowshocks.

(4) H2S appears to be more abundant than SO2 in high-velocityclumps, in agreement with chemical models. However, quite nar-row H2S lines have also been detected, suggesting the presence ofregions where the evaporated H2S molecules had enough time toslow down at ambient velocities but not to freeze out on to grains.We speculate that these regions may have experienced large temper-atures (�200 K) so that all the OH molecules needed to transformH2S efficiently into SO2 have been converted into H2O and the H2Snarrow profiles that we observe today are a fossil signature of ashock passage.

AC K N OW L E D G M E N T S

We thank R. Cesaroni, B. Nisini and M. Walmsley for helpful dis-cussions and suggestions.

R E F E R E N C E S

Allen D. A., Burton M. G., 1993, Nat, 363, 54Bachiller R., Perez Gutierrez M., Kumar M. S. N., Tafalla M., 2001, A&A,

372, 899Bally J., Lane A. P., 1982, ApJ, 257, 612Bally J., Lane A. P., 1991, in Elston R., ed., ASP Conf. Ser. Vol. 14,

Astrophysics with Infrared Arrays. Astron. Soc. Pac., San Francisco,p. 273

Bergin E. A., Goldsmith P. F., Snell R. L., Ungerechts H., 1994, ApJ, 431,674

Bergin E. A., Snell R. L., Goldsmith P. F., 1996, ApJ, 460, 343Bergin E. A., Ungerechts H., Goldsmith P. F., Snell R. L., Irvine W. M.,

Schloerb F. P., 1997, ApJ, 482, 267Charnley S. B., 1997, ApJ, 481, 386Choi M., Evans N. J., II, Jaffe D. T., 1993, ApJ, 417, 624Codella C., Bachiller R., 1999, A&A, 350, 659Codella C., Bachiller R., Reipurth B., 1999, A&A, 343, 585Cohen R. J., Rowland P. R., Blair M. M., 1984, MNRAS, 210, 425Garay G., Ramırez S., Rodrıguez L. F., Curiel S., Torrelles J. M., 1996, ApJ,

459, 193

C© 2003 RAS, MNRAS 341, 707–716

716 C. Codella et al.

Gerin M., Falgarone E., Joulain K., Kopp M., Le Bourlot J., Pineau desForets G., Roueff E., Schilke P., 1997, A&A, 318, 579

Goetz J. A. et al., 1998, ApJ, 504, 359Gyulbudaghian A. L., Glushkov Yu. I., Denisyuk E. K., 1978, ApJ, 224,

L137Hartigan P., Lada C. J., 1985, ApJS, 59, 383Hartigan P., Lada C. J., Stocke J., Tapia S., 1986, AJ, 92, 1155Hartquist T. W., Oppenheimer M., Dalgarno A., 1980, ApJ, 236, 182Hayashi S. S., Hasegawa T., Kaifu N., 1988, ApJ, 332, 354Hughes V. A., Moriarty-Schieven G., 1990, ApJ, 360, 215Hughes V. A., Wouterloot J. G. A., 1984, ApJ, 276, 204Keane J. V., Boonman A. M. S., Tielens A. G. G. M., van Dishoeck E. F.,

2001, A&A, 376, L5Lada C. J., Fich M., 1996, ApJ, 459, 638Le Bourlot J., Pineau des Forets G., Roueff E., 1995, A&A, 297, 251Lefloch B., Castets A., Cernicharo J., Loinard L., 1998, ApJ, 504,

L109

Lenzen R., 1988, A&A, 190, 269Lucas R., Liszt H. S., 2002, A&A, 384, 1054Narayanan G., Walker C. K., 1996, ApJ, 466, 844Nilsson A., HjalmarsonA., Bergman P., Millar T. J., 2000, A&A, 358, 257Pineau des Forets G., Roueff E., Schilke P., Flower D. R., 1993, MNRAS,

262, 915Raines S. N. et al., 2000, ApJ, 528, L115Salas L. et al., 1999, ApJ, 511, 822Sargent A. I., 1977, ApJ, 466, 844Swade D. A., 1989, ApJ, 345, 828Torrelles J. M., Verdes-Montenegro L., Ho P. T. P., Rodrıguez L. F., Canto

J., 1993, ApJ, 410, 202Wilkin F. P., 1996, ApJ, 459, L31Wilson T. L., Rood R. T., 1994, ARA&A, 32, 191Zhang Q., Zheng X., 1997, ApJ, 474, 719

This paper has been typeset from a TEX/LATEX file prepared by the author.

C© 2003 RAS, MNRAS 341, 707–716