First Detection of Interaction between a Magnetic Disk Wind and an Episodic Jet in a Protostellar System
Chin-Fei Lee, Benoit Tabone, Sylvie Cabrit, Claudio Codella, Linda Podio, Ferreira, Jacquemin-Ide
aa r X i v : . [ a s t r o - ph . GA ] J a n First Detection of Interaction between a Magnetic Disk Wind andan Episodic Jet in a Protostellar System
Chin-Fei Lee , , Benoit Tabone , , Sylvie Cabrit , Claudio Codella , , Linda Podio ,Ferreira, J. , Jacquemin-Ide, J. ABSTRACT
Rotating outflows from protostellar disks might trace extended magneto-hydrodynamic (MHD) disk winds (DWs), providing a solution to the angularmomentum problem in disk accretion for star formation. In the jet system HH212, a rotating outflow was detected in SO around an episodic jet detected in SiO.Here we spatially resolve this SO outflow into three components: a collimated jetaligned with the SiO jet, the wide-angle disk outflow, and an evacuated cavity inbetween created by a large jet-driven bowshock. Although it was theoreticallypredicted before, it is the first time that such a jet-DW interaction is directlyobserved and resolved, and it is crucial for the proper interpretation and mod-eling of non-resolved DW candidates. The resolved kinematics and brightnessdistribution both support the wide-angle outflow to be an extended MHD DWdominating the local angular momentum extraction out to 40 au, but with aninner launching radius truncated to ∼ > Academia Sinica Institute of Astronomy and Astrophysics, No. 1, Sec. 4, Roosevelt Road, Taipei 10617,Taiwan Graduate Institute of Astronomy and Astrophysics, National Taiwan University, No. 1, Sec. 4, RooseveltRoad, Taipei 10617, Taiwan Leiden Observatory, Leiden University, PO Box 9513, 2300 RA Leiden, The Netherlands Observatoire de Paris, PSL University, Sorbonne Universit´e, CNRS, LERMA, 61 Av. de l’Observatoire,75014 Paris, France INAF, Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125 Firenze, Italy Univ. Grenoble Alpes, CNRS, Institut de Plan´etologie et d’Astrophysique de Grenoble (IPAG), 38000Grenoble, France
Subject headings: accretion, accretion disks — stars: formation — ISM: individ-ual objects (HH 212) — ISM: jets and outflows
1. Introduction
Rotating outflows from protostellar disks are a newly discovered component in star for-mation (Launhardt et al. 2009; Greenhill et al. 2013; Zapata et al. 2015; Bjerkeli et al. 2016;Hirota et al. 2017; Tabone et al. 2017; Lee et al. 2018a,b; Zhang et al. 2018; Louvet et al.2018; deValon et al. 2020). They might trace extended magneto-hydrodynamic (MHD) diskwinds (DWs), providing a solution to the angular momentum problem in disk accretion. TheHH 212 protostellar system (Zinnecker et al. 1992) is a young system located in Orion at ∼
400 pc, harboring a nearly edge-on rotating disk optimal for detecting such an extended diskwind. The central protostar has a mass of 0.25 ± M ⊙ , deeply embedded in an infalling-rotating envelope (Lee et al. 2017b). Previous Atacama Large Millimeter/submillimeter Ar-ray (ALMA) observations have spatially resolved the disk (Lee et al. 2017a), which is rotatingwithin a centrifugal barrier at ∼
44 au (Lee et al. 2017b). A spinning jet was also detectedin SiO carrying angular momentum away from the innermost disk at a radius of ∼ − at 60 au resolution,rotating in the same sense as the disk, and consistent with an extended MHD disk windlaunched from ≃ . ∼ ∼
13 au (0 ′′ . 033), we retrieveadditional SO emission structures reconciling the above seemingly contradictory results,providing a confirmation for an extended disk wind as well as the first evidence of jet-diskwind interaction first predicted by Tabone et al. (2018). This interaction provides uniquefirst clues to the unknown magnetic field strength and distribution in young protostellardisks. 3 –
2. Observations
HH 212 was observed with ALMA in Band 7 centered at a frequency of ∼ ∼ − ∼ − per channel.The uv data was calibrated manually by the ALMA QA2 team using the CommonAstronomy Software Applications (CASA) package version 5.1.1. No self-calibration wasperformed due to insufficient signal to noise ratio of the continuum data in the long baselines.A robust factor of 0.5 was used for the visibility weighting to generate SO and SiO channelmaps with a synthesized beam of 0 ′′ . 036 × ′′ . 030 at a position angle of ∼ − ◦ and a noiselevel of ∼ − (7.0 K). We also included the SO visibility data obtained inCycle 3 (Lee et al. 2018a) and reduced the noise level slightly down to 0.67 mJy beam − (6.2K) in the SO channel maps. The velocities in the channel maps are LSR velocities.
3. Results
Figure 1 shows the SO map in comparison to the SiO map of the jet and the continuummap of the dusty disk (adopted from Lee et al. 2019) within 1400 au of the central protostarat 13 au resolution. SiO shows an episodic jet launched from the innermost disk, appearingfirst as a highly collimated chain of knots in inner 200 au and then a chain of broader bowshocks downstream at larger distances. SO also shows a jet aligned with the SiO jet.We can unveil the wide-angle outflow by separating the SO emission into two velocitycomponents. At high velocity (more than ± − away from the systemic velocity of ∼ − , Figure 1c), SO traces a collimated jet aligned with the SiO jet, but widerpossibly because the SO line has a lower critical density than SiO and thus can trace lessdense material. The critical densities (in H ) are 7 . × cm − for SO and 1 . × cm − for SiO (Schoier et al. 2005). At low velocity (within ± − of the systemic velocity,Figure 1d), thin outflow shells (marked with white brackets) are detected in SO surroundingthe jet. Only their bases were detected before (Lee et al. 2018a). They are now detectedfurther away and seen to smoothly connect to the SiO/SO bow shocks downstream at largerdistances ( ∼
600 au) (Figure 1e). Faint extended SO emission is also detected surrounding 4 –the base of the shells, within z ∼ <
150 au from the disk. This emission shows up better in anintensity-weighted velocity map (Figure 2), forming a wide-angle rotating outflow togetherwith the base of the shells, appearing as a thick X-shape fanning out from the disk, rotatingaround the jet. Away from the base, the shells are mainly blueshifted in the north andredshifted in the south, similar to the velocity sense of the bow shocks at larger distanceand thus driven by them. The inner part of the wide-angle outflow coincides with the baseof the shells and is thus perturbed by the bow shocks. The wide-angle outflow has anouter boundary outlined by the inner infalling-rotating envelope traced by the high-velocityemission of HCO + (Figure 2b), confirming that it originates from the disk. Its outer partis unperturbed by the bow shocks, providing the best opportunity to check the previouslyproposed MHD disk wind interpretation (Tabone et al. 2017, 2020).
4. MHD Disk Wind Model
Various MHD models are being developed to launch disk winds and carry away part orall of the angular momentum from accretion disks (Turner et al. 2014; Bai 2017; Zhu & Stone2018; Riols et al. 2020). The first and most simple 2D version of these models is a steady-state, axisymmetric, self-similar wind launched from a geometrically thin Keplerian disk(Blandford & Payne 1982; Ferreira 1997). These models are well suited for comparison withobservations because they allow for parameter studies. As discussed in Tabone et al. (2020),the observable structure and kinematics of the wind in these models are mainly determined bythree parameters: (1) protostellar mass M ∗ defining the Keplerian rotation v k, = p GM ∗ /r at a radius r in the disk, (2) magnetic level arm parameter λ ≃ ( r A /r ) , where r A is theAlfv´en radius along the streamline launched from a footpoint at r , determining the poloidalacceleration and the extracted specific angular momentum. In particular, the terminal windvelocity and the final specific angular momentum achieved along each streamline are v w ∼ v k, √ λ − l ∼ λ l , respectively, where l = r v k, is the value at the footpoint at r .And (3) widening factor W ≡ r max /r , where r max is the maximum radius reached by thestreamline at large distance, controlling the flow transverse size.Assuming M ∗ ∼ . M ⊙ , Tabone et al. (2017, 2020) found their disk wind Model L5W30(with λ ∼ . W ∼
30) to broadly reproduce the transverse spatio-kinematic structureof the SO rotating outflow, with r = 0.1 to 40 au. The same model is thus adopted hereto compare with the SO wide-angle outflow resolved at higher resolution and sensitivity. Asshown in Figure 2b, the wide-angle outflow shows an opening structure in good agreementwith the predicted model streamlines but with a launching radius truncated to ∼ > r ∼ −
40 au for the wide- 5 –angle outflow, as shown in Figure 3. The wind is assumed to be symmetric with respectto the disk midplane and extend out to 185 au to the north and south. The inner part ofthe wind bounded by the model streamlines launched from 4 to 8 au roughly coincides withthe base of the shells (Figure 2b) and is thus assumed to become the shell perturbed by thebow shocks, with its outflow velocity replaced by a radially expanding velocity v r = r/t s ,where t s is the dynamical age of the shell (Lee et al. 2018a). We assume a temperature of100 K in the outer unperturbed part and 200 K in the inner perturbed part (shell), based onthe temperatures derived before for the disk atmosphere and the shells (Lee et al. 2018a).The SO jet is assumed to have a launching radius of 0.10 to 0.20 au in the dust-free zone(Tabone et al. 2020). It has a temperature of 300 K, which is a mean value adopted beforeto derive the SO abundance (Podio et al. 2015). Since the observed jet is unresolved, thisjet component is only for illustrative purposes.In this self-similar model, the disk has an accretion rate varying with radius as˙ M acc ( r ) = ˙ M in ( r r in ) ξ (1)with ˙ M in being the accretion rate at the inner radius r in and ξ being the ejection efficiency(Ferreira & Pelletier 1995). Thus, the mass-loss rate in the wind between the inner radius r in and outer radius r out will be˙ M DW = ˙ M out − ˙ M in = ˙ M in [( r out r in ) ξ −
1] (2)where ˙ M out is the accretion rates at r out . For the disk wind to remove all of the accretionangular momentum, we have (Tabone et al. 2020) ξ ∼ λ − . (3)With λ ∼ . r out ∼
40 au, r in ∼ M in ∼ × − M ⊙ yr − (Lee 2020), we have˙ M DW ∼ . × − M ⊙ yr − . Along each streamline, the corresponding wind density hasbeen given in Tabone et al. (2020).This model can be compared quantitatively to the observed wide-angle outflow in termsof kinematics. A radiative transfer code (Lee et al. 2014) adding the SO line is used togenerate the position-velocity (PV) diagrams of the SO emission from the model, assumingLTE. The SO abundance (wrt molecular hydrogen) x SO is a free parameter to be derived bymatching the observed intensity. Within 100 au of the protostar, since the jet has a propermotion of ∼
64 km s − (Claussen et al. 1998) and a mean radial velocity of ∼ − − in the northern component and ∼ − in the southern component (Lee et al. 2017c), 6 –the inclination angles of the wide-angle outflow and jet are assumed to be ∼ − ◦ in thenorthern component and ∼ ◦ in the southern component.Figure 4 shows the resulting model PV diagrams on the observed ones cut across thejet axis centered at increasing distance from the protostar to near the end of the wide-angleoutflow, with x SO ∼ . × − in the extended disk wind and x SO ∼ . × − in the jet.As can be seen, with the outer unperturbed part of the wind, this model roughly reproducesthe PV structures of the faint unperturbed wide-angle outflow (marked with blue brackets),even though it predicts a rotation velocity slightly larger than observed. Note that the wide-angle outflow in the north is only detected within ∼
130 au of the protostar (see also Figure2). Moreover, with the inner part of the wind being radially expanding, the model producestilted elliptical PV structures for the shell and can roughly match the observations with thedynamical age t s ∼
37 yrs, similar to that found in Lee et al. (2018a). The resulting outflowvelocity is drawn in Figure 3. This age roughly agrees with the axial distance ( ∼
600 au)traveled by the cavity apex for a jet speed ∼
64 km s − , supporting that the shell at thebase is created by the same large jet bowshock seen downstream. However, since the modelPV structures of the shell are tilted more than observed, the rotation velocity in the shell isalso over-predicted. Note that the density in the shell is ∼ λ value,keeping the same for the other parameters. At the same time, we decrease the inner launchradius of the SO jet to 0.05 au to maintain the same maximum jet velocity. As shown inFigure 5, this very simple “modified” model with λ ∼ . λ value is favored by recent MHD simulations including stellarirradiation (Wang et al. 2019). Moreover, with this smaller λ , the wind density will be afactor of ∼ ≃ − , close to thatin the HH 212 disk (Podio et al. 2015). This result is in excellent agreement with thermo-chemical modeling of dense Class 0 MHD disk winds (Panoglou et al. 2012), which predictsthat the SO wind abundance should remain “frozen” near the disk value up to z ≃ z ∼
150 au (Figures 1 and2) is also predicted by this model, as the wind becomes transparent to photodissociatingFUV photons from the accretion shock (Panoglou et al. 2012). The disk wind is thus denseenough to remove the disk angular momentum, if its SO abundance is close to that in thedisk. 7 –
5. Conclusions
Our new observations have reconciled previous studies of SO rotating outflow in HH212, supporting the presence of an extended MHD disk wind out to the disk outer edge at 40au, removing most of the angular momentum flux required for accretion (Tabone et al. 2017,2020), but with a smaller magnetic level arm and an inner launching radius truncated to ∼ > r ∼ . − .
20 au) drives large bow shocks interacting with the extendeddisk wind and producing a cavity, with the thin SO shell forming its boundary. These dataare thus providing the first unambiguous evidence for the theoretically predicted interactionbetween a time-variable jet and an outer disk wind (Tabone et al. 2018). Resolving such aninteraction is crucial for the proper interpretation and modeling of less well resolved diskwind candidates.Furthermore, the width of the cavity provides us with the first quantitative clues tothe magnetic field strength in a disk wind. Indeed, the twisting of field lines at the base ofMHD disk winds creates a strong magnetic pressure that efficiently confines the sidewaysexpansion of jets and jet bow shocks (Meliani et al. 2006; Matsakos et al. 2009). With a jetmass-flux ∼ − M ⊙ yr − ejected sideways at a speed ∼
10 km s − (Lee et al. 2015), andthe magnetic field strength required in our wind model to drive accretion at ˙ M in ∼ × − M ⊙ yr − , equilibrium at z ∼
150 au between the lateral ram pressure and magnetic pressure isreached on the streamline launched from 4 au, as observed. Hence, the cavity width appearsconsistent with the outer disk ( r ≥ ∝ r − / ) than the wind ram pressure ( r − ),the disk wind should be much weaker inside of r ∼ ∼ ∼ > REFERENCES
Armitage, P. J. 2011, ARA&A, 49, 195Bai, X.-N. 2017, ApJ, 845, 75Balbus, S. A., & Hawley, J. F. 2006, ApJ, 652, 1020Bjerkeli, P., van der Wiel, M. H. D., Harsono, D., Ramsey, J. P., & Jørgensen, J. K. 2016,Nature, 540, 406Blandford, R.D. & Payne, D.G, 1982, MNRAS 199, 883Claussen, M.J., Marvel., K.B., Wootten, A., Wilking, B.A. 1998, ApJL, 507, L79de Valon, A., Dougados, C., Cabrit, S. et al. 2020, A&A, 634, L12Eisloeffel, J. & Mundt, R. 1992, A&A, 263, 292Ferreira, J., 1997, A&A, 319, 340Ferreira, J., Deguiran, R. 2013, High Energy Density Physics, 9, 67–74 9 –Ferreira, J., & Pelletier, G. 1995, A&A, 295, 807Greenhill, L. J.,Goddi, C., Chandler, C. J., Matthews, L. D., & Humphreys, E. M. L. 2013,ApJ, 770, L32Hirota, T., Machida, M. N., Matsushita, Y., et al. 2017, Nature Astronomy, 1, 0146Launhardt, R., Pavlyuchenkov, Ya. , Gueth, F., et al. 2009, A&A, 494, 147Lee, C.-F., Hirano, N., Zhang, Q., et al. 2014, ApJ, 786, 114Lee, C.-F., Hirano, N., Zhang, Q., et al. 2015, ApJ, 805, 186Lee, C.-F., Li, Z.-Y., Ho, P. T. P., et al. 2017a, Science Advances, 3, e1602935Lee, C.-F., Li, Z.-Y., Ho, P. T. P., et al. 2017b, ApJ, 843, 27Lee, C.-F., Ho, P. T. P., Li, Z.-Y., et al. 2017c, Nature Astronomy, 1, 0152Lee, C.-F., Li, Z.-Y., Codella, C., et al. 2018a, ApJ, 856, 14Lee, C.-F., Li, Z.-Y., Hirano, N., et al. 2018b, ApJ, 863, 94Lee, C.-F., Li, Z.-Y., Ching, T.-C., Lai, S.-P., & Yang, H. 2018c, ApJ, 854, 56Lee, C.-F., Codella, C., Li, Z.-Y., et al. 2019, ApJ, 876, 63Lee, C.-F. 2020, A&A Rev., 28, 1Louvet, F., Dougados, C., Cabrit, S., et al. 2018, A&A, 618, A120Matsakos, T., Massaglia, S., Trussoni, E. et al. 2009, A&A 502, 217Meliani, Z., Casse, F., & Sauty, C. 2006, A&A 460, 1Mori, S., Bai, X.-N., & Okuzumi, S. 2019, ApJ, 872, 98Panoglou, D., Cabrit, S., Pineau des Forˆets, G., et al. 2012, A&A 538, A2Podio, L., Codella, C., Gueth, F.,et al. 2015, A&A, 581, A85Riols, A., Lesur, G., Menard, F. 2020, A&A, 639, A95Sch¨oier, F. L., van der Tak, F. F. S., van Dishoeck, E. F., et al. 2005, A&A, 432, 369Tabone, B., Cabrit, S., Bianchi, E., et al. 2017, A&A, 607, L6 10 –Tabone, B., Cabrit, S., Pineau des Forˆets, G., et al. 2020, A&A, 640, A82. doi:10.1051/0004-6361/201834377Tabone, B., Raga, A., Cabrit, S., et al. 2018, A&A, 614, A119Turner, N.J.,Fromang, S.,Gammie, C., et al. 2014, Protostars and Planets VI, pp. 411-432Vorobyov, E. I., Elbakyan, V. G., Plunkett, A. L., et al. 2018, A&A, 613, A18.doi:10.1051/0004-6361/201732253Wang, L., Bai, X.-N., Goodman, J. 2019, ApJ 874, 90Zapata, L. A., Lizano, S., Rodr´ıguez, L. F., et al. 2015, ApJ, 798, 131Zhang, Y., Higuchi, A. E., Sakai, N., et al. 2018, ApJ, 864, 76Zinnecker, H. , Bastien, P. , Arcoragi, J. -P. & Yorke, H. W. 1992, A&A, 265, 726Zhu, Z., & Stone, J. M. 2018, ApJ, 857, 34
This preprint was prepared with the AAS L A TEX macros v5.2.
11 –Fig. 1.— SiO and SO intensity maps toward the HH 212 system within 1400 au of theprotostar, together with the 350 GHz continuum map of the disk (gray image adopted fromLee et al. 2019). The maps are all rotated by 22.5 ◦ clockwise to align the jet axis in thenorth-south direction. SO-HV indicates the SO map at high velocity more than ± − away from the systemic velocity. SO-LV indicates the SO map at low velocity within ± − of the systemic velocity. Color codes are the same as the labels. White brackets markthe shells. 12 –Fig. 2.— Intensity-weighted velocity maps of SO in the inner region at low velocity. Thegray X-shaped curve in (a) outlines the wide-angle rotating outflow detected in SO. Panel(b) zooms into the central region. Black contours show the same disk map as in Figure 1.Red and blue contours show the high-velocity (HV) HCO + emission adopted from Lee et al.(2017b), outlining the boundary of the innermost envelope. Gray curves plot the streamlinesof the disk wind in Model L5W30, with footpoints at r = 4, 8, 16, and 40 au. Dashed curvesshow the streamlines with footpoints at r = 0.2, 1, 2 au. 13 –Fig. 3.— A schematic diagram showing the extended disk wind and jet in our model, andthe interaction between them. The wind has a launching radius of 4 to 40 au, while the jethas a launching radius of 0.05 to 0.20 au. The shell (gray cross-hatched region) extends fromthe inner part of the wind to the jet-driven bow shock. The rotation velocity (color image)and outflow velocity (vectors) in the wind are derived from Model L5W30, with the outflowvelocity in the inner part replaced with a radially expanding velocity. 14 –Fig. 4.— Comparison of model PV diagrams (red contours) to the observed PV diagrams(black contours) of SO emission cut across the jet axis centered at increasing distances (asindicated in the upper left corners in au) from the protostar along the jet axis. Contoursstart at 2 σ with a step of 3 σ , where σ ∼ . λ ∼ ..