On the origin of the molecular outflows in IRAS 16293-2422
Josep M. Girart, Robert Estalella, Aina Palau, Jose M. Torrelles, Ramprasad Rao
aa r X i v : . [ a s t r o - ph . GA ] D ec Draft version October 8, 2018
Preprint typeset using L A TEX style emulateapj v. 04/20/08
ON THE ORIGIN OF THE MOLECULAR OUTFLOWS IN IRAS 16293 − Josep M. Girart , Robert Estalella , Aina Palau , Jos´e M. Torrelles , Ramprasad Rao Institut de Ci`encies de l’Espai, (CSIC-IEEC), Campus UAB, Facultat de Ci`encies, C5p 2, 08193 Bellaterra, Catalonia, Spain,[email protected] Departament d ′ Astronomia i Meteorologia, Institut de Ci`encies del Cosmos (UB-IEEC), Mart´ı i Franqu`es, Universitat de Barcelona,08028 Barcelona, Catalonia, Spain and Institute of Astronomy and Astrophysics, Academia Sinica, 645 N. Aohoku Pl., Hilo, HI 96720, USA
Draft version October 8, 2018
ABSTRACTWe present CO 3–2, SiO 8–7, C S 7–6, and 878 µ m dust continuum subarcsecond angular resolutionobservations with the Submillimeter Array (SMA) toward the IRAS 16293 − S emission traces the 878 µ m dust continuum well, and inaddition clearly shows a smooth velocity gradient along the major axis of component I16293A. COshows emission at moderate high velocities arising from two bipolar outflows, which appear to beperpendicular with respect to each other. The high sensitivity and higher angular resolution of theseobservations allows us to pinpoint well the origin of these two outflows at the center of componentI16293A. Interestingly, the most compact outflow appears to point toward I16293B. Our data showthat the previously reported monopolar blueshifted CO outflow associated with component I16293Bseems to be part of the compact outflow arising from component I16293A. In addition, the SiOemission is also tracing this compact outflow: on one hand, the SiO emission appears to have a jet-like morphology along the southern redshifted lobe; on the other hand, the SiO emission associatedwith the blueshifted northern lobe traces a well defined arc on the border of component I16293B facingI16293A. The blueshifted CO lobe of the compact outflow splits into two lobes around the positionof this SiO arc. All these results lead us to propose that the compact outflow from componentI16293A is impacting on the circumstellar gas around component I16293B, possibly being diverged asa consequence of the interaction. Subject headings:
ISM: individual objects (IRAS 16293 − INTRODUCTION
The dark cloud Lynds 1689N, located in theOphiucus star-forming region at a distance of120 pc (Knude & Hog 1998; Loinard et al. 2008;Lombardi et al. 2008), harbors IRAS 16293 − ∼ ′′ ( ∼
600 AU), first detected at cm-continuumwavelengths (Wootten 1989; Estalella et al. 1991) andlater also detected at (sub)mm wavelengths (e.g.,Chandler et al. 2005; Rodr´ıguez et al. 2005; Rao et al.2009; Pineda et al. 2012; Loinard et al. 2013). WhileI16293A shows an “hourglass” magnetic field structure,I16293B shows an ordered magnetic field (Rao et al.2009). Wootten (1989) found that I16293A splits intotwo subcomponents at cm wavelengths (usually referredas A1 and A2), separated by 0 . ′′ O maser emission is also observed toward this binary(Wilking & Claussen 1987; Wootten 1989; Imai et al.2007), tracing zones of compressed gas produced byshocks in the presence of very strong line-of-sight mag-netic fields ( ∼
113 mG, Alves et al. 2012). I16293 hastwo bipolar outflows at scales of ∼ . ∼ ′′ led Loinard et al. (2013) to suggestthat I16293B is ejecting a blueshifted bubble-like outflowwith low velocity and moderate collimation. Based onthese results, together with the small kinematic ageestimated for this outflow, these authors proposed thatI16293B is the youngest object in the region, and one ofthe youngest protostars known (Loinard et al. 2013).In this Letter, we present new SMA CO 3–2, SiO 8–7, C S 7–6, and continuum observations at 345 GHztoward I16293. Our observations are more sensitive toextended structures ( ∼ ′′ ) than those of the recentlyreported 690 GHz ALMA observations (the visibilityrange for ALMA is 62–943 k λ and for the SMA is 20-240 k λ ). Due to these observational properties, our datashow that the blueshifted outflow reported previouslywith ALMA data seems to be part of a bipolar outfloworiginating from one of the stars within I16293A, ratherthan originating from I16293B as was recently proposed.More importantly, this bipolar outflow is interacting withI16293B producing a shock structure at its SW edge seenin SiO. OBSERVATIONS
The SMA observations were taken on 2010 August28 in the extended configuration. These observationswere performed in the polarimetric mode. The resultsof the polarization data have been presented in a differ-ent paper (Rao et al. 2013). The receiver was tuned tocover the 333.5-337.5 GHz and 345.5-349.5 GHz frequen-cies in the lower side band (LSB) and upper side band,(USB) respectively. The phase center of the telescopewas RA(J2000.0)= 16 h m . s
90 and DEC(J2000.0)= − ◦ ′ . ′′
73. The correlator provided a spectral res-olution of about 0.8 MHz (i.e., 0.7 km s − at 345 GHz).The gain and bandpass calibrators were QSO J1733-130 and QSO 3C454.3, respectively. The absolute fluxscale was determined from observations of Neptune. Theflux uncertainty was estimated to be ∼ S 7–6(337.396 GHz) line to trace the circumstellar gas in theregion. In addition, the dust continuum at 878 µ m is alsopresented. Maps were obtained from the visibilities usingnatural weighting, which yielded a synthetic beam size of ≃ . ′′
8, and allowed the tracing of smaller spatial scales( ≃
100 AU) around I16293 than the scales ( ≃
300 AU)traced by Rao et al. (2009). RESULTS
Figure 1 shows the channel maps of the high velocitycomponent of the CO 3–2 emission. This figure showsthat the high velocity gas exhibits a quadrupolar mor-phology well centered in I16293A, apparently formingtwo bipolar molecular outflows, an extended E-W out-flow and a compact NW-SE outflow (the direction ofthese two outflows are delineated in Fig. 1 b ). The E-W outflow consists of blueshifted emission in the easternlobe (with some redshifted gas at low flow velocities) andredshifted emission in the western lobe. This bipolar out-flow has been previously well studied and extends to dis-tances of 0.1 pc from its powering source (Mizuno et al.1990; Hirano et al. 2001; Yeh et al. 2008), far beyond theSMA field of view.The NW-SE bipolar outflow extends only ≃ ′′ (0.005 pc, 1000 AU), so it is very compact comparedwith the other two bipolar outflows detected in the re-gion (Mizuno et al. 1990). This compact NW-SE bipolaroutflow was already reported by Rao et al. (2009), butour new SMA observations now reveal it more clearly(in particular from the SiO data; see below). At low out-flow velocities the blueshifted lobe appears to split in twoparts just before reaching I16293B (Fig. 1 c and 1 d ). Athigher velocities (Fig. 1 a and Fig. 1 b ) the blueshifted gasends at the position of I16293B. Another characteristicof the outflow is an apparent acceleration of the gas, withthe higher velocities appearing farther from the drivingsource. This is more evident in the position-velocity cut Fig. 1.—
Channel maps of the CO 3-2 for the blueshifted (bluecontours) and redshifted (red contours) emission, overlapped withthe 878 µ m dust emission (black contours and grey scale). Panels a , b and c : the CO contour’s step level and the first contour are0.47 Jy beam − . Panel d : the CO contour’s step level as well thefirst contour is 0.65 Jy beam − . In all panels the first contour is at2- σ level. The dust contours show the emission at the 5, 30, 55, and80% level of the maximum intensity, 1.88Jy beam − . The outflowvelocity (i. e., the velocity of the gas with respect to the systemicvelocity of I16293A, ≃ . − ) of each channel is indicated inthe top left corner of each panel. The offset spatial positions arewith respect to the phase center (given in Section 2). The greendashed line in panel b shows the cavity traced by the E-W CO lobeat scales of ∼ d . along the blueshifted lobe (Fig. 2 a ): the terminal out-flow velocity increases with the distance from the source.A cut across the blueshifted lobe (Fig. 2 b ) shows thatthe highest outflow velocities are spatially more compactthan the outflow component at velocities closer to thesystemic velocity of the I16293A core, suggesting that Fig. 2.—
Position–velocity plots of the CO 3–2 emissionalong (panel a ) and across (panel b ) the blueshifted lobe of theNW-SE compact outflow. The reference position (0 ′′ offset) isRA(J2000.0)= 16 h m . s
65 and DEC(J2000.0)= − ◦ ′ . ′′ a (offset position of +3 ′′ ). The vertical dashedline indicates the systemic velocity of I16293A. The red lines areshown to better indicate the velocity gradient of the blueshiftedCO emission. the outflow is more collimated at higher outflow veloci-ties.The SiO 8-7 emission (Fig. 3) appears to arise mainlyfrom two patches of emission. One of them extends fromI16293A, in a jet-like structure, toward the south-eastfollowing well the redshifted lobe of the compact NW-SE outflow seen in CO 3-2. Most of the emission in thejet-like structure appears close to the cloud velocity butslightly redshifted ( v LSR ≃ . . − ). On theother hand, the northern patch of SiO emission shows aclear partial ring surrounding I16293B, facing I16293A(Fig. 3), and appears near the position where the bluelobe of the CO compact NW-SE outflow diverges spa-tially (Fig. 1).The C S 7-6 mainly traces the circumstellar gasaround I16293A at scales of few hundreds AU andmatches the 878 µ m dust emission very well (Fig. 4).Both the circumstellar molecular gas and the dust struc-tures are elongated along the NE-SW direction ( P A ≃ ◦ ), with position velocity cuts of the CS emission alongand across the major axis suggesting that the gas is rotat-ing with a Keplerian-like pattern. Below, we describe theprocedure to fit the C S 7–6 emission toward I16293Awith a rotating geometrically thin disk.
Thin Disk Model for I16293A
Fig. 3.—
Panel a : Spectra of the SiO 8-7 (black spectrum)and CH CHO (blue spectrum) lines, averaged over an area of 9arcsec around I16293B. The CH CHO spectrum was obtained byaveraging the 18 , -17 , A and E, and 17 , -16 , A and Elines (these four lines are located in the 334.9-335.4 GHz range andhave similar excitation temperatures and line strengths). Panel b :Zoom in of the panel d from Fig. 1 (contours of the CO and dustemission are the same as those from Fig. 1), overlapped of theintegrated SiO 8-7 emission (grey image). We considered a geometrically thin disk, with an inner( r i ) and outer ( r o ) radius. The angle between the diskaxis and the plane of the sky is i ( i = 0 ◦ for an edge-ondisk). We consider a rotation velocity given by a powerlaw of the radius, v r ( r/r ) q r , where r is an arbitraryreference radius and v r is the rotation velocity at thereference radius.We computed, for each point of a regular grid in theplane of the sky, the projection of the rotation velocityof the corresponding point of the disk along the line ofsight v z . A Gaussian line profile of width ∆ v and cen-tered on v z was added to the channels associated with thegrid point. Finally, each channel map was convolved spa-tially with a Gaussian beam of width ∆ s . However, theintensity scale of the channel maps is arbitrary. A scalingfactor, the same for all channel maps, was obtained byminimizing the sum of the squared differences betweenthe data channel maps and the synthetic channel maps. Fig. 4.—
Bottom panel: color image of the first order moment(velocity field) of the C S 7–6 line toward I16293A. The color scale(in km s − ) is shown in the right side of the panel. The black thickdashed contours show the 878 µ m continuum emission. The thickdashed blue and red arrows show the direction of outflows. Toppanels: velocity–position plots of the C S 7-6 lines taken alongthe minor (panel a ) and major (panel c ) axes. Panels b and d showthe modeled data. The model depends on a total of 10 parameters, namelythe beamwidth, ∆ s ; the linewidth, ∆ v ; the disk center,( x , y ); the disk systemic velocity, v ; the disk inner andouter radii, r i and r o ; the disk rotation velocity at thereference radius, v r ; the radial dependence power-law in-dex of the rotation velocity, q r ; and the disk inclinationangle. i . Some of the parameters are known beforehand,such as ∆ s and ∆ v . Some other can be guessed based onphysical grounds (i.e. q r = − . x , y , v ), and the last 4 have with physical interest ( r i , r o , TABLE 1Parameters of the Best Fit Model
Parameter Units Value
Fixed:
Beamwidth ∆ s (arcsec) 0 . v (km s − ) 1 . q r − . Fitted:
Disk center x (arcsec) − . ± . y (arcsec) 0 . ± . v (km s − ) 3 . ± . r i (AU) 1 ± r o (AU) 140 ± v r a (km s − ) − . ± . i (deg) 44 . ± . a For a reference radius r = 0 . ′′ v r and i ), which can be estimated through model fittingto the data.The fitting procedure was the sampling of the seven-dimensional parameter space, using the same proce-dure as that described in Estalella et al. (2012) andS´anchez-Monge et al. (2013). The parameter space wassearched for the minimum value of the rms fit resid-ual. Once a minimum of the rms fit residual was found,the uncertainty in the parameters fitted was found asthe increment of each of the parameters of the fit nec-essary to increase the rms fit residual by a factor of[1 + ∆( m, α ) / ( n − m )] / , where n is the number of datapoints fitted, m is the number of parameters fitted, and∆( m, α ) is the value of χ for m degrees of freedom (thenumber of free parameters) and α is the significance level(0 < α < m = 7, and for a significance level of0.68 (equivalent to 1 σ for a Gaussian error distribu-tion), the increment in the rms fit residual is given by∆(7 , .
68) = 8 .
17 (S´anchez-Monge et al. 2013).The model was fitted to the C S 7–6 emission asso-ciated to I16293A. The rotation axis in the plane of thesky (derived from the CS velocity gradient) was foundto be at a position angle of − . ◦
8. The best fit valuesand their errors are shown in Table 1. Figure 4 showsthe comparison of the synthetic position-velocity cuts forthe best solution with those for the SMA C S 7–6 data,with the best fit values matching the data very well. Themodel used assumes that the intensity is proportional tothe geometrical depth of the disk along the line of sight(i.e. optically thin emission, and constant density andtemperature), which is reasonable for the C S emission,except near the disk center. Thus, the non-zero valueobtained from the fit for the disk inner radius can beonly interpreted in the sense that the contribution of theemission near the disk center is low. The simple kine-matical model used here cannot discard the idea thatthe inner radius is actually larger, as suggested by thebinarity of the central source, with a semi-major axis of0 . ′′
35 (Loinard et al. 2007).The mass of the protostar in source A can be esti-mated from the values derived from the disk fitting. As-suming Keplerian rotation ( M = r v /G ), the mass is M I16293A = 2 . ± . M ⊙ . This mass is similar to themass derived from the relative motions of the A1 and A2objects (Pech et al. 2010, these two objects are embed-ded in the C S structure). DISCUSSION AND CONCLUSIONS
The first single-dish observations in I16293 showedtwo bipolar molecular outflows, one extended alongthe NE (redshifted)-SW (blueshifted) direction, and an-other along the E (blueshifted)-W (redshifted) direction(Walker et al. 1988; Mizuno et al. 1990). The NE-SWoutflow has not been detected at smaller scales throughinterferometric observations, suggesting that it might bea fossil outflow (Yeh et al. 2008; Rao et al. 2009). The E-W outflow has been well detected and studied with theSMA at arcsecond angular resolution (Yeh et al. 2008;Rao et al. 2009). The CO emission associated with thewestern lobe at scales of ∼ ◦ with re-spect to the plane of the sky (see the green dashed line inFigure 1; Yeh et al. 2008). However, our maps show thatnear the protostars the CO outflow emission does not fol-low this parabolic cavity. At the position of I16293A theCO 3–2 outflow is very bright and extends roughly inthe direction of the E-W outflow. At the highest veloc-ity channel (Fig. 1 a ) the CO emission around I16293A iscompact and the blue and redshifted peaks form a posi-tion angle of ≃ ◦ , so they are possibly also associatedwith the E-W outflow.The relatively high velocity CO 3–2 emission as tracedby the SMA delineates a well defined bipolar compactoutflow of only 0.005 pc in the NW-SE direction, wellcentered around I16293A, as was already suggested byRao et al. (2009). However, this differs from what hasrecently been reported from ALMA CO 6–5 observa-tions (Loinard et al. 2013; Kristensen et al. 2013). Thereare some observational features that support our state-ment. First, the NW-SE outflow appears to be parallelto the rotation axis of the circumstellar disk-like struc-ture, traced by the C S, around I16293A (see Fig. 4).Second, the overall kinematical and morphological fea-tures of the NW blue lobe appear to be very consistentwith those of the prototypical molecular outflows associ-ated with class 0 protostars (e. g., Arce & Sargent 2006;Palau et al. 2006) if it is powered by I16293A: it has aconical-like structure starting in this source; the CO 3-2 channel maps show an apparent Hubble-like velocitystructure (higher velocities arise farther from the pow-ering protostar: see also the position-velocity cut alongthe blue lobe in Fig. 2); a cut across the blue lobe shows(Fig. 2) that the highest CO 3-2 velocities occur alongthe outflow axis (i. e., the highest velocities are more col-limated than the lowest). Third, the SiO emission isfound only along the SE-NW direction. The SiO is amolecule that traces shocks strong enough to producedust sputtering, releasing silicates from the dust mantles(e. g., Anderl et al. 2013). The morphology of the SiOemission associated with I16293B suggests that the SiOarises from the external shells of the circumstellar ma-terial of this component. Its location, facing componentI16293A and overlapping with the NW blueshifted lobe,suggests that the SiO traces the region where the NW-SE outflow (powered by I16293A) is impacting on thecircumstellar gas around I16293B. In fact, the SiO spec-trum of the emission around I16293B is much broaderthan the acetaldehyde (CH CHO) spectrum (see Fig 2 b ).Acetaldehyde is a hot core tracer and is likely tracing thequiescent (apparently unperturbed) circumstellar gas in I16293B.The terminal velocity of this compact outflow is small, ≃
13 km s − . This outflow is in projection perpendic-ular to the major axis of the disk-like structure asso-ciated with I16293A, so we can fairly assume that thisconfiguration holds in 3-D. Thus correcting for the out-flow inclination (44 ◦ , see Table 1), we find a dynami-cal timescale of ∼
400 yr. This is much smaller thanthe kinematic timescale of the extended E-W outflow(5000 yr, Mizuno et al. 1990). We also estimated theoutflow parameters of the compact NW-SE outflow fol-lowing Palau et al. (2007), and assuming the same in-clination given above, optically thin emission, and anexcitation temperature of ∼