Physical and chemical fingerprint of protostellar disc formation
E. Artur de la Villarmois, J. K. Jørgensen, L. E. Kristensen, E. A. Bergin, D. Harsono, N. Sakai, E. F. van Dishoeck, S. Yamamoto
AAstronomy & Astrophysics manuscript no. 34877corr c (cid:13)
ESO 2019May 1, 2019
Physical and chemical fingerprint of protostellar disc formation
E. Artur de la Villarmois , J. K. Jørgensen , L. E. Kristensen , E. A. Bergin , D. Harsono , N. Sakai , E. F. vanDishoeck , , and S. Yamamoto Niels Bohr Institute & Centre for Star and Planet Formation, University of Copenhagen, Øster Voldgade 5–7, 1350 CopenhagenK., Denmarke-mail: [email protected] Department of Astronomy, University of Michigan, 311 West Hall, 1085 S. University Ave, Ann Arbor, MI 48109, USA Leiden Observatory, Leiden University, PO Box 9513, NL-2300 RA Leiden, the Netherlands The Institute of Physical and Chemical Research (RIKEN), 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan Max-Planck-Institut f¨ur extraterrestrische Physik, Giessenbachstraße 1, 85748, Garching bei M¨unchen, Germany Department of Physics, The University of Tokyo, Bunkyo-ku, Tokyo 113-0033, JapanMay 1, 2019
ABSTRACT
Context.
The structure and composition of emerging planetary systems are likely strongly influenced by their natal environmentwithin the protoplanetary disc at the time when the star is still gaining mass. It is therefore essential to identify and study the physicalprocesses at play in the gas and dust close to young protostars and investigate the chemical composition of the material that is inheritedfrom the parental cloud.
Aims.
The purpose of this paper is to explore and compare the physical and chemical structure of Class I low-mass protostellar sourceson protoplanetary disc scales.
Methods.
We present a study of the dust and gas emission towards a representative sample of 12 Class I protostars from the Ophiuchusmolecular cloud with the Atacama Large Millimeter / submillimeter Array (ALMA). The continuum at 0.87 mm and molecular tran-sitions from C O, C S, H CO + , CH OH, SO , and C H were observed at high angular resolution (0 (cid:48)(cid:48) . ∼
60 au diameter) towardseach source. The spectrally and spatially resolved maps reveal the kinematics and the spatial distribution of each species. Moreover,disc and stellar masses are estimated from the continuum flux and position-velocity diagrams, respectively.
Results.
Six of the sources show disc-like structures in C O, C S , or H CO + emission. Towards the more luminous sources,compact emission and large line widths are seen for transitions of SO that probe warm gas ( E u ∼
200 K). In contrast, C O emissionis detected towards the least evolved and less luminous systems. No emission of CH OH is detected towards any of the continuumpeaks, indicating an absence of warm CH OH gas towards these sources.
Conclusions.
A trend of increasing stellar mass is observed as the envelope mass decreases. In addition, a power-law relation is seenbetween the stellar mass and the bolometric luminosity, corresponding to a mass accretion rate of (2.4 ± × − M (cid:12) year − forthe Class I sources, with a minimum and maximum value of 7.5 × − and 7.6 × − M (cid:12) year − , respectively. This mass accretionrate is lower than the expected value if the accretion is constant in time and rather points to a scenario of accretion occurring in bursts.The di ff erentiation between C O and SO suggests that they trace di ff erent physical components: C O traces the densest and colderregions of the disc-envelope system, while SO may be associated with regions of higher temperature, such as accretion shocks. Thelack of warm CH OH emission suggests that there is no hot-core-like region around any of the sources and that the CH OH columndensity averaged over the disc is low. Finally, the combination of bolometric temperature and luminosity may indicate an evolutionarytrend of chemical composition during these early stages.
Key words.
ISM: molecules – stars: formation – protoplanetary discs – astrochemistry – Individual: Ophiuchus
1. Introduction
The formation and evolution of protoplanetary discs are funda-mental in the process of low-mass star formation and to under-stand how our own solar system formed. The chemical complex-ity of the protoplanetary disc is established either by the mate-rial that is inherited from the envelope or from processed mate-rial within the disc, or a combination of the two (e.g. Pontoppi-dan et al. 2014; Drozdovskaya et al. 2018), providing the initialchemical conditions for planet formation. However, the physi-cal and chemical processes at play on small scales ( ≤
500 au) inlow-mass protostars are still not well understood.The dynamical evolution of the system a ff ects the mass dis-tribution, where material from the inner envelope falls towardsthe disc, the disc accretes material onto the central protostar, and part of this material is ejected through outflows (Terebeyet al. 1984; Shu et al. 1993; Hartmann 1998). Class I sourcesare associated with the formation and evolution of circumstel-lar discs (e.g. Sheehan & Eisner 2017; Yoo et al. 2017) and stillhave a significant contribution from the envelope material (Ro-bitaille et al. 2006). Therefore, Class I sources act as a bridgebetween the deeply embedded Class 0 sources and the protoplan-etary discs that are associated with Class II sources. The evolu-tionary sequence, Class 0 - Class I - Class II, still has puzzlingquestions such as when and how quickly the envelope dissipates,how early discs form and how quickly they grow in mass andsize, and how material accretes from the disc onto the protostar.The presence of knots or bullets in outflows (e.g. Reipurth 1989;Arce et al. 2013) and the low luminosity of protostars comparedto models (Kenyon & Hartmann 1995; Evans et al. 2009; Dun- Article number, page 1 of 21 a r X i v : . [ a s t r o - ph . S R ] A p r & A proofs: manuscript no. 34877corr ham & Vorobyov 2012) suggests that the mass accretion ratesmay vary with time. Direct measurement of the mass accretionrate is extremely challenging for embedded protostars, but anapproximate estimate can be obtained from the accretion lumi-nosity equation (e.g. Kenyon & Hartmann 1995; White et al.2007; Dunham et al. 2014a; Mottram et al. 2017). The changesin accretion, and thus in luminosity, may have significant con-sequences in the chemistry and further evolution of the system(e.g. Jørgensen et al. 2013, 2015; Frimann et al. 2016b).The complex environment in which protoplanetary discsform and evolve is exposed to large variations in temperature(tens to hundreds of K) and density (10 − cm − ), whichleave strong chemical signatures and make molecules excellentdiagnostics of the physical conditions and processes. Observeddi ff erent molecular species that trace di ff erent physics, such asdisc tracers (such as CO and C O; e.g. Harsono et al. 2014),warm gas tracers (such as CH OH; e.g. Jørgensen et al. 2013),and shock tracers (such as SO; e.g. Sakai et al. 2014), are essen-tial in order to understand the di ff erent physical and chemicalprocesses involved at disc scales. Previous studies of molecularline emission towards Class I sources were focused on charac-terising the gas kinematics, or the chemistry of single sourcesor of a few sources (e.g. Jørgensen et al. 2009; Harsono et al.2014), which makes it di ffi cult to compare them. In addition, lit-tle is known about the evolution of the molecular content as afunction of physics in these stages.With the high sensitivity and angular resolution of the At-acama Large Millimeter / submillimeter Array (ALMA), it is be-coming possible to resolve disc scales (10 −
50 au towards nearbystar-forming regions) and study the physics and chemistry ofthese environments. Disc rotation, when detectable, is a strongtool for determining protostellar masses (e.g. Jørgensen et al.2009; Harsono et al. 2014; Yen et al. 2014, 2015). A comparisonbetween the chemistry and the physical parameters of protostarsat disc scales and in a more statistical way is therefore now be-coming possible.One of the closest low-mass star-forming regions is the Ophi-uchus molecular cloud (Wilking et al. 2008), with a distance ( d )of 139 ± ff erent evolutionary stages, which makes this star-forming region an excellent laboratory for the study of low-massstar formation and the discs of low-mass stars.We present ALMA observations of a representative sampleof 12 Class I sources in the Ophiuchus molecular cloud. The ob-servations include continuum emission at 0.87 mm and molecu-lar lines that trace di ff erent components of the star-forming en-vironment: C O, C S, H CO + , CH OH, SO , and C H. Sec-tion 2 describes the observational procedure, the data calibra-tion, the source properties, and the molecular transitions that arecovered. The results are presented in Sect. 3, where moment 0and 1 maps are shown for each molecular transition. Section 4presents an analysis of the mass evolution, with a comparisonbetween disc masses, stellar masses, envelope masses, and bolo-metric luminosities, following an estimate of the mass accretionrate. Section 5 describes the chemical evolution, focusing on thelack of warm CH OH detection, the compact and broad SO emission, and the chemical trend observed within the sources.Finally, Sect. 6 summarises our main findings.
2. Observations
In order to address the physical and chemical properties ofdiscs in their earliest stages, a sample of 12 Class I protostars in Ophiuchus was observed. The sources are all well charac-terised through large-scale mid-infrared (
Spitzer ) and submil-limetre (SCUBA) surveys (Jørgensen et al. 2008), and were se-lected in order to include protostars with bolometric tempera-tures ( T bol ) below 400 K. In addition, the sample covers a widerange of bolometric luminosities ( L bol ), from 0.03 to 18 L (cid:12) , andenvelope masses from about 0.05 M (cid:12) up to 0.3 M (cid:12) (Table 1).With this approach, the sample is a representative set for explor-ing the physical and chemical structures of Class I sources.The 12 sources were observed with ALMA on four oc-casions, between 2015 May 21 and June 5 (program code:2013.1.00955.S; PI: Jes Jørgensen). At the time of the obser-vations, 36 antennas were available in the array (37 for the June5 observations), providing baselines between 21 and 556 metres(784 metres for the June 5 observations) and a maximum angularscale of ∼ (cid:48)(cid:48) . Each of the four sessions provided an on-sourcetime of 43 minutes in total for the 12 di ff erent sources (i.e. eachsource was observed for approximately 15 minutes in total). Theobserved sources are listed in Table 1, with their physical prop-erties and other common identifiers.The observations covered five di ff erent line settings, and thechoice of species was made specifically to trace di ff erent as-pects of the structure of protostars. Lines of C O, H CO + andC S were targeted as tracers of disc kinematics, while SO andCH OH are expected to trace the warm chemistry in the innerenvelope or disc, and C H is commonly associated with outer-envelope regions. Two spectral windows consist of 960 channelseach with 122.07 kHz (0.11 km s − ) spectral resolution centredon C O J = − S J = −
6, while the other three contain1920 channels each with 244.14 kHz (0.22 km s − ) spectral reso-lution centred on H CO + J = −
3, the CH OH J k = k − k branchat 338.4 GHz, and the CH CN 14 −
13 branch at 349.1 GHz. Thelatter two settings also pick up SO and C H transitions. Theobserved molecular transitions and their parameters are sum-marised in Table 2.The calibration and imaging were done in CASA (Mc-Mullin et al. 2007): the complex gains were calibrated throughobservations of the quasars J1517-2422 and J1625-2527, pass-band calibration was based on J1924-2914, and flux calibrationwas based on Titan. A robust weighting with the Briggs param-eter set to 0.5 was applied to the visibilities, and the resultingdataset has a typical beam size of 0 (cid:48)(cid:48) . × (cid:48)(cid:48) .
32 ( ∼ ×
40 au), acontinuum rms level of 0.3 mJy beam − , and a spectral rms levelof 13 and 9 mJy beam − per 0.11 and 0.22 km s − , respectively.
3. Results
The sources that show continuum emission at 0.87 mm arelisted in Table 3, with the sub-millimetre coordinates, decon-volved sizes, and fluxes calculated by fitting two-dimensional(2D) Gaussians in the image plane. Two of the sources, IRS 43(Girart et al. 2000) and IRS 67 (McClure et al. 2010), are knownbinary systems, and their components are separated by ∼
70 au(0 (cid:48)(cid:48) .
53) and ∼
90 au (0 (cid:48)(cid:48) .
68) for IRS 43 and IRS 67, respectively.The sources associated with a single continuum peak have di-ameters from ∼
20 au (0 (cid:48)(cid:48) .
14 for ISO-Oph 203) to ∼
60 au (0 (cid:48)(cid:48) . http://casa.nrao.edu/ Article number, page 2 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation -101
GSS30-IRS1 [GY92] 30 WL 12 [GY92] 197 -101-101
Elias 29 -101
IRS 44 -101
IRAS 16253-2429 -101
ISO-Oph 203 -2-1012
Offset ["] -2-1012 O ff s e t [ " ] IRS 43 VLA 1VLA 2 -2-1012
IRS 67 BA I [ m J y b ea m ] Fig. 1.
Continuum emission of the detected sources above 5 σ ( σ = − ). The contours start at 5 σ and follow a step of 20 σ . Thetypical synthesised beam is represented by the black filled ellipse in the upper and lower left panels. The (0,0) position represents the positionfitted with a 2D Gaussian (see Table 3). The lower panels show the binary systems with the common identifiers of both components. Table 1.
Properties of the observed sources: infrared position, bolometric temperature ( T bol ), bolometric luminosity ( L bol ), envelope mass ( M env ),and other common identifiers. Source Infrared position a T bol b L bol b M env c Other common identifiersRA [J2000.0] Dec [J2000.0] [K] [L (cid:12) ] [M (cid:12) ]J162614.6 16 26 14.63 −
24 25 07.5 7 0.03 0.095 [EDJ2009] 800GSS30-IRS1 16 26 21.35 −
24 23 04.3 250 11.00 0.15 Oph-emb 8, [GY92] 6, Elias 21[GY92] 30 16 26 25.46 −
24 23 01.3 200 0.12 0.27 Oph-emb 9WL 12 16 26 44.19 −
24 34 48.4 380 1.40 0.076 VSSG 30, [GY92] 111IRAS 16238-2428 16 26 59.10 −
24 35 03.3 120 1.70 0.045 [EDJ2009] 856[GY92] 197 16 27 05.24 −
24 36 29.6 120 0.18 0.16 Oph-emb 6, LFAM 26Elias 29 16 27 09.40 −
24 37 18.6 420 18.00 0.045 Oph-emb 16, [GY92] 214IRS 43 d
16 27 26.91 −
24 40 50.7 300 3.30 0.13 Oph-emb 14, [GY92] 265, YLW 15IRS 44 16 27 27.98 −
24 39 33.4 280 7.10 0.057 Oph-emb 13, [GY92] 269, YLW 16IRAS 16253-2429 16 28 21.61 −
24 36 23.4 36 0.24 0.15 Oph-emb 1, MMS 126ISO-Oph 203 16 31 52.45 −
24 55 36.2 330 0.15 0.072 Oph-emb 15IRS 67 d
16 32 00.99 −
24 56 42.6 180 2.80 0.078 Oph-emb 10, L1689S1 3
Notes. ( a ) Position of
Spitzer sources from the c2d catalog (Evans et al. 2009). ( b ) From Dunham et al. (2015). ( c ) Envelope mass from models ofsub-millimetre and IR data following description in Jørgensen et al. (2009). ( d ) Binary system.
J162614.6 and IRAS 16238-2428 show neither continuumnor molecular line emission; the upper limit is 0.9 mJy beam − (3 σ ) for the continuum emission. J162614.6 shows faint emis-sion at 24 µ m and a low concentrated SCUBA core (Jørgensenet al. 2008), suggesting that this source is more likely related The Submillimeter Common-User Bolometer Array (SCUBA) on theJames Clerk Maxwell Telescope (JCMT) with a prestellar core than with a protostar. For IRAS 16238-2428, the separation between its
Spitzer position and SCUBAcore is 38 (cid:48)(cid:48) . ∼ (cid:48)(cid:48) ). However, the continuumemission detected in the same field of view as IRAS 16238-2428may correspond to a T Tauri source (LFAM 23) that is located ∼ (cid:48)(cid:48) o ff set from the Spitzer position that is associated with IRAS16238-2428. Table 3 presents results from a 2D Gaussian fit
Article number, page 3 of 21 & A proofs: manuscript no. 34877corr
Table 2.
Spectral setup and parameters of the detected molecular transitions.
Molecular transition Frequency a A i j a E u a n crit b [GHz] [s − ] [K] [cm − ]C O J = − × −
32 3.5 × C S J = − × −
50 1.3 × SO J K A K C = , − , × −
197 5.9 × CH OH J K = − − − E 338.34463 1.7 × −
70 2.0 × CH OH J K = − A + × −
65 2.8 × H CO + J = − × −
42 8.5 × C H N = − J = / − / F = − × −
42 2.2 × C H N = − J = / − / F = − × −
42 2.3 × Notes. ( a ) Values from the CDMS database (Müller et al. 2001). ( b ) Calculated values for a kinetic temperature of 30 K and collisional rates fromthe Leiden Atomic and Molecular Database (LAMDA; Schöier et al. 2005). The collisional rates of specific species were taken from the followingsources: C O from Yang et al. (2010), C S from Lique et al. (2006), SO from Balança et al. (2016), CH OH from Rabli & Flower (2010),H CO + from Flower (1999), and C H from Spielfiedel et al. (2012).
Table 3.
Results of 2D Gaussian fits towards the continuum peak of our Oph sources.
Source RA Dec Size a PA F . S . (0 (cid:48)(cid:48) . b [J2000.0] [J2000.0] [ (cid:48)(cid:48) ] [ ◦ ] [mJy] [mJy beam − ]GSS30-IRS1 16 26 21.358 -24 23 04.85 0.19 ± × ± ± ± ± ± × ± ± ± ± ± × ± ±
11 160.4 ± ± c
16 26 59.166 -24 34 59.07 0.09 ± × ± ±
45 17.6 ± ± ± × ± ± ± ± ± × ± ±
90 41.2 ± ± ± × ± ±
17 41.6 ± ± ± × ± ±
37 7.5 ± ± ± × ± ±
19 38.6 ± ± ± × ± ± ± ± ± × ± ±
18 11.3 ± ± ± × ± ±
17 35.7 ± ± ± × ± ± ± ± Notes. ( a ) Deconvolved size (FWHM). ( b ) Brightness at the continuum peak. ( c ) Class II T Tauri source located in the same field as IRAS 16238-2428. for LFAM 23, revealing that it is a point source. However, thissource does not show molecular line emission and is thereforenot included in the discussion in this paper.Figure 1 shows the continuum emission of the remaining tensources. For the binary systems, the continuum also traces thecircumbinary material, which is remarkably extended for IRS67 ( ∼
550 au; for a detailed analysis of IRS 67, see Artur de laVillarmois et al. 2018). It is worth mentioning that IRS 43 VLA1 and IRS 67 B are brighter than IRS 43 VLA 2 and IRS 67 Aat 0.87 mm, while the opposite situation is observed at infraredwavelengths (Haisch et al. 2002; McClure et al. 2010).
The molecular transitions trace di ff erent components and are re-lated with di ff erent morphologies. Some of them are detectedtowards the source position, while others peak o ff set from thesource, and the emission may present compact or extendedstructures. Table 4 summarises the regions where the individ-ual molecular transitions are detected, specifying if the emissionis compact or extended (with respect to the extent of the con-tinuum emission), and the source velocity ( V source ). The latter isestimated by visual inspection of C O emission, or SO emis-sion when C O is not detected. For sources where no molecular line emission is detected, V source is taken from the literature (seeTable 4). The two C H lines listed in Table 2 are both hyperfinetransitions, and the C H emission arises from a blended doublet,C H N = − J = / − / F = − H N = − J = / − / F = −
3. Emission of at least one molecular transition is detectedtowards eight of the ten sources. WL 12 (the brightest contin-uum source) and ISO-Oph 203 (the weakest continuum source)are the only sources where no line emission is detected withinthe spectral setting shown in Table 2. O, H CO + , and C S C O, H CO + , and C S are less abundant isotopologues, andthey are expected to be optically thin tracers associated with thedisc kinematics. For example, a Keplerian disc was detected to-wards the Class 0 / I source R CrA in similar C O observationswith ALMA (Lindberg et al. 2014).Figure 2 shows moment 0 and 1 maps for C O, indicatingthe outflow direction, if available, from the literature (Bontempset al. 1996; Allen et al. 2002; van der Marel et al. 2013; Zhanget al. 2013; White et al. 2015; Yen et al. 2017). Five of thesources, [GY92] 30, [GY92] 197, IRAS 16253-2429, IRS 43,and IRS 67, show on-source C O emission, and with the excep-tion of IRAS 16253-2429, they are associated with a rotational
Article number, page 4 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation -101
GSS30-IRS1 [GY92] 30 C O WL 12 ND [GY92] 197 -101-101 Elias 29 -101
IRS 44 ND -101 IRAS 16253-2429 -101
ISO-Oph 203 ND -3-2-10123 Offset ["] -3-2-10123 O ff s e t [ " ] IRS 43 -3-2-10123
IRS 67 -4 -2 0 2 4V [km s ] Fig. 2.
Moment 0 (black contours) and moment 1 (colour scale) maps for C O J = σ and integrated over a velocity range of 8 km s − .The contours start at 5 σ and follow a step of 4 σ ( σ = − km s − ). The yellow star indicates the position from a 2D Gaussian fit, and theblack stars in the lower panels represent the position of the binary components. The arrows show the outflow direction from the literature, wherethe blue and red represent blue- and red-shifted emission, respectively, and the grey arrow represents infrared observations. The typical synthesisedbeam is shown by the black filled ellipse in the upper and lower left panels. The ND label marks a non-detection. profile perpendicular to the outflow direction, consistent with adisc-like structure (Sect. 4.2). In addition, the binary systems(IRS 43 and IRS 67) show C O emission that is much moreextended than towards the other sources: 6 (cid:48)(cid:48) . (cid:48)(cid:48) . (cid:48)(cid:48) . (cid:48)(cid:48) . (cid:48)(cid:48) . < ± − ).This low-velocity component may be tracing infalling materialfrom the inner envelope (see detailed discussion in Sect. 4.2).Figures 3 and 4 show images of the H CO + and C S emis-sion. H CO + is detected towards only three of the ten sources:IRS 44 and the two binary systems, IRS 43 and IRS 67. IRS 44shows a velocity gradient with two components, one consistentwith the outflow direction, and the another rotated by ∼ ◦ . Thepeak of emission (moment 0), however, is related with more qui-escent material. The more massive of the binary systems (IRS67) shows more extended emission, and its velocity profile isperpendicular to the outflow direction, while the emission to-wards IRS 43 is concentrated in the inner regions ( ≤ (cid:48)(cid:48) ) and atentative velocity gradient is detected. In addition, C S is solelyseen towards the binary systems. IRS 43 shows a velocity gra-dient perpendicular to the infrared jet direction and the emis-sion peaks spatially o ff set from the system ( ∼ (cid:48)(cid:48) ), where the red- shifted emission stands out. A di ff erent situation is observed forIRS 67: it has no clear velocity gradient and shows isolated peaksof emission that are spatially o ff set from the system. OH and SO CH OH forms exclusively on ice-covered dust grain surfacesbecause gas-phase reactions produce negligible CH OH abun-dances (Garrod et al. 2006; Chuang et al. 2016; Walsh et al.2016). When the temperature reaches ∼
90 K, CH OH thermallydesorbes from the grain mantles, and its gas-phase abundance isenhanced close to the protostar (Brown & Bolina 2007). There-fore, CH OH transitions associated with high rotational levels,such as J = k − k , are expected to trace the warm gas close to theprotostar.The observed CH OH J = k − k branch includes transitionswith E u from 65 to 376 K, but no significant emission is de-tected towards the continuum position from a 2D Gaussian fitof any source (see Table 3). This is highlighted in Fig. 5, wherethe observed and predicted spectrum are plotted together. Thepredicted spectrum is obtained using the statistical equilibriumradiative transfer code (RADEX; van der Tak et al. 2007) and as-suming local thermodynamic equilibrium (LTE), a CH OH col-
Article number, page 5 of 21 & A proofs: manuscript no. 34877corr -101
GSS30-IRS1 ND [GY92] 30 ND H CO + WL 12 ND [GY92] 197 ND -101-101 Elias 29 ND -101 IRS 44 -101
IRAS 16253-2429 ND -101 ISO-Oph 203 ND -3-2-10123 Offset ["] -3-2-10123 O ff s e t [ " ] IRS 43 -3-2-10123
IRS 67 -4 -2 0 2 4V [km s ] Fig. 3.
Same as Fig. 2 for H CO + J = −
3, integrated over a velocity range of 10 km s − . The contours start at 5 σ and follow a step of 3 σ, withthe exception of IRS 67, which follows a step of 6 σ . -101 GSS30-IRS1 ND [GY92] 30 ND C S WL 12 ND [GY92] 197 ND -101-101 Elias 29 ND -101 IRS 44 ND -101 IRAS 16253-2429 ND -101 ISO-Oph 203 ND -3-2-10123 Offset ["] -3-2-10123 O ff s e t [ " ] IRS 43 -3-2-10123
IRS 67 -4 -2 0 2 4V [km s ] Fig. 4.
Same as Fig. 2 for C S J = −
6, integrated over a velocity range of 10 km s − . The contours start at 5 σ and follow a step of 3 σ, with theexception of IRS 43, which follows a step of 6 σ .Article number, page 6 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation Table 4.
Regions where the individual molecular transitions are detected above 3 σ towards each source and the estimated source velocity ( V source ). Source Molecular transitions V source a C O H CO + C S SO CH OH C H [km s − ]GSS30-IRS1 O ff set - - On source [C] - - 3.4[GY92] 30 On source [C] - - - O ff set O ff set 3.1WL 12 - - - - - - 4.0 b [GY92] 197 On source - - - - On source [C] 3.1Elias 29 O ff set - - On source [C] - - 3.6IRS 43 On source [E] On source On source [E] On source B [C] - O ff set 4.0IRS 44 - On source [E] - On source [C] - - 2.3IRAS 16253-2429 On source [E] - - - - On source [E] 4.0ISO-Oph 203 - - - - - - 4.5 b IRS 67 On source [E] On source [E] O ff set On source B [C] - On source [E] 4.2 Notes. ( a ) Estimated from visual inspection of the C O spectrum, or SO when C O is not detected. ( b ) Values taken from Lindberg et al. (2017).The [C] and [E] labels refer to compact or extended emission (with respect to the extension of the continuum emission). -0.020.000.020.040.060.08
GSS30-IRS1 [GY92] 30 -0.020.000.020.040.060.08
WL 12 [GY92] 197 -0.020.000.020.040.060.08 I n t e n s it y [ J y b ea m ] Elias 29 IRS 43 -0.020.000.020.040.060.08
IRS 44 IRAS 16253-2429
Frequency [GHz] -0.020.000.020.040.060.08
ISO-Oph 203
IRS 67
Fig. 5.
Observed CH OH spectra (blue) towards the 10 sources thatshow continuum emission, superimposed on the predicted CH OHspectrum (red) for a column density of 2.5 × cm − , ∼ umn density of 2.5 × cm − ( ∼ OH column density measured towards the Class 0 sourceIRAS 16293-2422; Jørgensen et al. 2016), and a kinetic temper-ature ( T kin ) of 100 K. Clearly, no CH OH lines are seen even atthis level.Nevertheless, [GY92] 30 is a peculiar case (the least lumi-nous source of the sample, L bol = L (cid:12) , and the source asso-ciated with the more massive envelope, M env = (cid:12) ), wheretwo CH OH transitions are detected o ff set from the source, be-yond the 25 σ continuum contour (Figs. A.1 and A.2 in the Ap-pendix). These transitions are associated with the lowest E u lev-els (65 and 70 K) and the emission is related with low velocities of between − − from the source velocity. Thus,the CH OH emission towards [GY92] 30 is likely related withextended envelope material. In addition, there is no clear asso-ciation between the CH OH emission and the direction of theinfrared jet.SO is commonly associated with outflows and shocked re-gions (e.g. Jørgensen et al. 2004; Persson et al. 2012; Podio et al.2015; Tabone et al. 2017). In addition, the highly excited rota-tional transitions ( E u ∼
200 K) may be tracing warm shocked gas.The observed SO transition, listed in Table 2, is associated witha high rotational level ( E u =
197 K) and is detected towards fiveof the sources, showing compact emission (Fig. 6). A rotationalprofile is seen for Elias 29 and IRS 44, associated with highvelocities (up to ±
10 km s − with respect to the source veloc-ity), and almost perpendicular to the outflow direction. Towardsthe binary systems, the SO emission is relatively weak and de-tected around one of the sources, IRS 43 VLA1 and IRS 67 B;the sources correspond to the brightest components at 0.87 mm.In addition, GSS30-IRS1 shows only a blue-shifted component,without a clear velocity profile. The SO emission towards Elias29 and IRS 44 may be consistent with a disc-like structure, butthe broad-line profile suggests that SO traces a di ff erent com-ponent that is associated with warm shocked gas. H The emission of C H has been associated with dense regionsexposed to UV radiation (e.g. Nagy et al. 2015; Murillo et al.2018). Its emission is shown in Fig. 7 and is detected towardsfive of the sources. Two of the them, [GY92] 197 and IRAS16253-2429, show on-source emission related with low veloc-ities, showing a compact morphology for [GY92] 197 and anextended structure for IRAS 16253-2429. Towards [GY92] 30,IRS 43, and IRS 67, there is no C H emission (above 5 σ ) at thepositions of the sources, but a velocity gradient is seen for thecircumbinary material, with a remarkable extended emission forIRS 67 ( ∼ (cid:48)(cid:48) ). Because IRAS 16253-2429 has one of the mostmassive envelopes ( M env = (cid:12) ) and circumbinary discs areusually related with a higher mass content and greater extensionthan circumstellar discs, C H likely traces more dense regions.
4. Mass evolution
The final mass of a protostar and the amount of material avail-able in the disc to form planets depend on the mass evolution of
Article number, page 7 of 21 & A proofs: manuscript no. 34877corr -101
GSS30-IRS1 [GY92] 30 ND SO WL 12 ND [GY92] 197 ND -101-101 Elias 29 -101
IRS 44 -101
IRAS 16253-2429 ND -101 ISO-Oph 203 ND -3-2-10123 Offset ["] -3-2-10123 O ff s e t [ " ] IRS 43 -3-2-10123
IRS 67 -10 -8 -6 -4 -2 0 2 4 6 8 10V [km s ] Fig. 6.
Same as Fig. 2 for SO J K A K C = , − , , integrated over a velocity range of 20 km s − . The contours start at 5 σ and follow a step of20 σ . -101 GSS30-IRS1 ND [GY92] 30 C H WL 12 ND [GY92] 197 -101-101 Elias 29 ND -101 IRS 44 ND -101 IRAS 16253-2429 -101
ISO-Oph 203 ND -3-2-10123 Offset ["] -3-2-10123 O ff s e t [ " ] IRS 43 -3-2-10123
IRS 67 -4 -2 0 2 4V [km s ] Fig. 7.
Same as Fig. 2 for the doublet C H N = − J = / − / the whole system (envelope, disc, protostar, and outflow). Deter-mining when and how quickly the envelope dissipates and thedisc grows in mass and size, the rate at which the protostar gainsmass and the amount of material expelled through the outflowsare all linked to each other and are crucial components in themass evolution. In this section, we estimate the disc masses fromthe continuum emission (Sect. 4.1) and the stellar masses frommolecular lines that show Keplerian profiles (Sect. 4.2). Later on,a comparison between envelope mass, disc mass, stellar mass,and bolometric luminosity is presented (Sect. 4.3). Finally, themass accretion rate is estimated from a relationship between thestellar mass and the bolometric luminosity, and this is comparedwith the bolometric temperature. The disc masses ( M disc ) were calculated from the continuumfluxes ( F . ), listed in Table 3, and M disc = S ν d κ ν B ν ( T ) , (1)where S ν is the surface brightness, d is the distance to the source, κ ν is the dust opacity, and B ν (T) is the Planck function for a sin-gle temperature. A distance of 139 ± κ . of 0.0175 cm g − (Ossenkopf & Henning 1994), com-monly used for dust in protostellar envelopes and young discs inthe millimetre regime (e.g. Shirley et al. 2011), was adopted forthe calculations (Artur de la Villarmois et al. 2018). A value of15 K was adopted for the dust temperature ( T dust ) following theanalysis by Dunham et al. (2014b). The calculated disc masses(gas + dust, assuming a gas-to-dust ratio of 100) are listed inTable 5, together with values from the literature. Errors in T dust are not considered in Table 5, but the disc masses will decreaseby a factor of ∼ ff erence seen for WL 12 appears to bedue to the choice of dust temperature (15 vs. 30 K). Assuming T dust =
30 K, a value of (10.3 ± × − M (cid:12) is found for M disc associated with WL 12. In order to investigate whether the optically thin tracers areassociated with Keplerian motions and to estimate the stellarmasses ( M (cid:63) ), position-velocity (PV) diagrams were created forthe sources that show disc-like structures in C O, H CO + orC S. The peak emission for each channel was obtained throughthe CASA task imfit, and the o ff set position was calculated byprojecting the peak emission onto the disc position angle (Ta-ble 3). Next, a Keplerian profile ( v ∝ r − . ) was employed to fitthe points with velocities above ± − . This velocity rangewas chosen in order to avoid the envelope contribution (e.g. van’t Ho ff et al. 2018). The PV diagrams and peak points are shownin Figs. 8 and 9, and the resulting Keplerian curve from the fitis overplotted. Figure 8a shows a robust Keplerian profile for[GY92] 197, while the PV diagram for IRS 44 is noisy (Fig. 8b),with an unclear Keplerian profile. IRAS 16253-2429 shows low-velocity emission ( ≤ − ) and both negative and positivevelocities for the same distance to the source (see Fig. 8c). This Table 5.
Calculated disc masses for T dust =
15 K and values from theliterature.
Source M disc [10 − M (cid:12) ]This work LiteratureGSS30-IRS1 6.2 ± ± ± a LFAM 23 3.1 ± ± ± a < b IRS 43 VLA 1 7.3 ± a IRS 43 VLA 2 1.3 ± ± ± ± ± ± Notes.
The errors in the second column do not include uncertaintiesin the assumed dust temperature ( T dust =
15 K), but the disc massesdecrease by a factor of ∼ ( a ) From Jørgensen et al. (2009) at 1.1 mm, with T dust =
30 K. ( b ) FromLommen et al. (2008) at 1.1 mm, with T dust =
30 K. is a characteristic feature of infalling material (e.g. Tobin et al.2012). Therefore, these points were fitted with an infalling pro-file under the conservation of angular momentum ( v ∝ r − ; e.g.Lin et al. 1994; Harsono et al. 2014). The resulting stellar massesare listed in Table 6, adopting an inclination ( i ) of 70 ◦ from theplane of the sky. This value is adopted from the mean value ofthe inclination of the sample, calculated from the deconvolvedcontinuum sizes (Table 3) and assuming a circular structure. Inaddition, 70 ◦ is consistent with values from the literature: 70 ◦ forIRS 43 (Brinch et al. 2016) and 60 ◦ for IRAS 16253-2429 (Yenet al. 2017). The stellar mass changes by less than 14% when theinclination varies with ± ◦ .IRS 43 shows extended emission in C O and C S, with PVdiagrams and Keplerian fits shown in Fig. 9 for both molecu-lar transitions. The blue-shifted C S emission is well fitted witha Keplerian profile, but the red-shifted emission shows an en-hancement around 2 (cid:48)(cid:48) and a vertical velocity structure, from ∼ ∼ − . PV diagrams for IRS 67 from C O and H CO + emission are presented in Artur de la Villarmois et al. (2018).Table 6 lists the stellar masses obtained from PV diagramsand values from the literature. For IRS 43, the stellar massesinferred from a Keplerian fit from C O and C S emission area factor of 2 higher than the 1.80 M (cid:12) from Brinch et al. (2016).To be consistent within the sample, the stellar mass of 4.0 M (cid:12) isused for the IRS 43 system hereafter. For IRAS 16253-2429, thestellar mass from Yen et al. (2017) is consistent with our valuefrom the fit for the infalling gas.
Figure 10 compares the envelope masses ( M env ; Table 1), discmasses ( M disc ; Table 5), and stellar masses ( M (cid:63) ; Table 6). Theuncertainties in M env are assumed to be 20%. There is no cleartrend between M disc and M (cid:63) or M env , however, all the points sat-isfy the condition M (cid:63) > M disc and M env > M disc , characteristic ofClass I stages (Robitaille et al. 2006). For M (cid:63) as a function of M env , the stellar mass increases as the mass of the envelope de-creases, and only one of the sources (IRAS 16253-2429) satisfies Article number, page 9 of 21 & A proofs: manuscript no. 34877corr -1.0 -0.5 0.0 0.5 1.0
Offset ["] -6-4-20246 V [ k m s ] [GY92] 197(a) C O -1.0 -0.5 0.0 0.5 1.0
Offset ["] -8-6-4-202468 V [ k m s ] IRS 44(b) H CO + -1.0 -0.5 0.0 0.5 1.0 Offset ["] -4-3-2-101234 V [ k m s ] IRAS 16253-2429(c) C O Fig. 8.
Position-velocity diagrams towards [GY92] 197, IRS 44 and IRAS 16253-2429. Blue and red dots represent blue- and red-shifted emissionpeaks above ± − , respectively, while green dots indicate velocities below ± − . Blue and red solid lines show the best fit for a Keplerianvelocity profile with their respective errors, shown in dashed blue and red lines. The cut taken from the image data is shown in grey contours,ranging from 3 σ ( σ =
13 mJy beam − ) to the maximum value of each transition. Each adjacent contour represents an increment of 30% of themaximum value for panels (a) and (c) , and an increment of 50% for panel (b) . The black dashed lines indicate the velocity above which theKeplerian profile was fitted. The solid black curves in panel (c) show the best fit for an infalling velocity profile. -5 -4 -3 -2 -1 0 1 2 3 4 5 Offset ["] -6-4-20246 V [ k m s ] IRS 43(a) C O -5 -4 -3 -2 -1 0 1 2 3 4 5
Offset ["] -6-4-20246 V [ k m s ] IRS 43(b) C S Fig. 9.
Position-velocity diagrams for C O and C S towards IRS 43. Blue and red dots represent blue- and red-shifted emission peaks above ± − , respectively, while green dots indicate velocities below ± − . Blue and red solid lines show the best fit for a Keplerian velocityprofile with their respective errors, shown in dashed blue and red lines. The cut taken from the image data is shown in grey contours, ranging from3 σ ( σ =
13 mJy beam − ) to the maximum value of each transition. Each adjacent contour represents an increment of 30 and 20% of the maximumvalue for C O and C S, respectively. The black dashed lines indicate the velocity above which the Keplerian profile was fitted. M [M ] M d i s k [ M ] M env [M ] M d i s k [ M ] M env [M ] M [ M ] IRAS 16253-2429
Fig. 10.
Comparison between disc mass and stellar mass ( left ), disc mass and envelope mass ( centre ), and stellar mass and envelope mass ( right ).The dashed black line is for M i = M j .Article number, page 10 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation Table 6.
Stellar masses obtained from the velocity profile fits and values from the literature.
Source This work Literature M (cid:63) [M (cid:12) ] a Molecule M (cid:63) [M (cid:12) ] Method[GY92] 197 0.23 ± OElias 29 2.5 ± b HCO + J = − ± S 1.80 ± c HCN J = − ± OIRS 44 1.2 ± CO + IRAS 16253-2429 0.03 ± O (infalling) 0.03 ± d C O J = − e ± O Notes. ( a ) The stellar masses were calculated assuming an inclination of 70 ◦ . ( b ) From Lommen et al. (2008). ( c ) From Brinch et al. (2016), for thebinary system. ( d ) From Yen et al. (2017). ( e ) A more detailed analysis of IRS 67 is presented in Artur de la Villarmois et al. (2018). M [M ] L bo l [ L ] L bol =10 * M This work - Class IThis work - Class 0From Aso et al. (2015) - Class IFrom Aso et al. (2015) - Class 0 T bol [K] M acc [ M y ea r ] Fig. 11.
Left:
Bolometric luminosity as a function of stellar mass with data points from this work and from Aso et al. (2015). The black straightline shows the best fit for all Class I sources.
Right:
Mass accretion rate as a function of the bolometric temperature. the condition M env > M (cid:63) , which is a characteristic of Class 0 pro-tostars. This is also consistent with its low bolometric tempera-ture, T bol of 36 K, and the infalling profile for the gas kinematics.The left panel of Fig. 11 shows the bolometric luminositiesas a function of stellar masses from Table 6 and values of Class 0and I sources summarised in Aso et al. (2015) (see Table B.1 inthe Appendix). The uncertainties in M (cid:63) and L bol are 20% and15%, respectively (e.g. Yen et al. 2017), and for the particularcase of the binary systems (IRS 43 and IRS 67), the M (cid:63) and L bol values were divided by 2 to account for each component. ForClass I sources, the correlation between L bol and M (cid:63) appears tobe linear, with one outlier that corresponds to the source L1551IRS 5. The straight line in Fig. 11 shows the best fit for all Class Isources, providing the following power-law relationship: L bol = . ± . M . ± . (cid:63) . (2)Assuming that L bol results from the gravitational energy thatis released by the material accreted onto the surface of the pro-tostar, the mass accretion rate ( ˙ M acc ) can be estimated from˙ M acc = L bol R (cid:63) GM (cid:63) , (3)where R (cid:63) is the protostellar radius and G the gravitational con-stant. Assuming R (cid:63) = (cid:12) (Stahler et al. 1980), the calcu-lated accretion rates are shown in the right panel of Fig. 11 as a function of T bol . Combining Eqs. 2 and 3, a value of(2.4 ± × − M (cid:12) year − is obtained for ˙ M acc for the Class Isources, with a minimum and maximum value of 7.5 × − and7.6 × − M (cid:12) year − , respectively. The mean value is consis-tent with the mass accretion rates from Yen et al. (2017), whofound ˙ M acc from ∼ × − to 4.4 × − M (cid:12) yr − for a sam-ple of Class I sources. The right panel of Fig. 11 shows ˙ M acc as a function of T bol . Myers & Ladd (1993) defined T bol as thetemperature of a blackbody that has the same mean frequency asthe observed continuum spectrum, and this is often taken as anindicator of the evolutionary stage of the source (e.g. Dunhamet al. 2014a; Frimann et al. 2016a; Fischer et al. 2017). A ten-tative decrease in ˙ M acc as the systems evolve is seen in Fig. 11.Nevertheless, the trend is not clear, and the sample of Class 0sources is still small.When the accretion rate is assumed to be constant (2.4 × − M (cid:12) year − ), a typical Class I protostar will reach 1 M (cid:12) in ∼ / I phase (from 0.13 to 0.78 Myr; Dunham et al.2015; Kristensen & Dunham 2018). This inconsistency is knowas the luminosity problem (Kenyon & Hartmann 1995), wherea constant accretion rate cannot explain the low luminosities ofembedded protostars. With interferometric facilities, the stellarmasses are better constrained from molecular line emission, andthe only way to reconcile theory with observations is to assume atime-variable accretion rate onto the protostar (Kenyon & Hart-mann 1995; White et al. 2007; Evans et al. 2009; Vorobyov &
Article number, page 11 of 21 & A proofs: manuscript no. 34877corr
Basu 2010; Dunham & Vorobyov 2012; Audard et al. 2014;Dunham et al. 2014a). The time-variable or burst model of ac-cretion operates in the embedded phase of protostellar evolu-tion, and the disc spends practically all of its time in a lowstate (low ˙ M acc ) and accretes significant mass in relatively shortbursts, which would account for the low average L bol of proto-stars (Vorobyov & Basu 2010; Dunham et al. 2014a).In Fig. 11, L1551 IRS 5 is the only Class I source with high L bol and high ˙ M acc that stands out over the others. This source hasbeen proposed to be a young star experiencing an FU Ori-likeoutburst (Hartmann & Kenyon 1996; Osorio et al. 2003). Withthe exception of L1551 IRS 5, the rest of the Class I sources inthis sample appear to be in a low state of accretion (see Table B.1in the Appendix).If the accretion onto the central star is episodic, the accretionbursts may have strong e ff ects on the chemistry. In consequence,observable chemical e ff ects (such as the C O spatial distribu-tion) can provide clues to the luminosity history (e.g. Jørgensenet al. 2013, 2015; Frimann et al. 2016b). In our sample, the lackof significantly extended C O emission contrasts with the C Oemission observed by Jørgensen et al. (2015) and Frimann et al.(2016b) towards a di ff erent sample of Class 0 / I sources, whichsuggests a relatively quiescent phase for our sources or a lowC O column density at larger distances from the source.
5. Chemical evolution bol and T bol
The di ff erences in the molecular emission signatures for thesources in the sample may be an indication that the chemistrydoes depend on the physical evolution of the sources, that is tosay, the molecular column densities vary as a function of thesource bolometric luminosities and temperatures. For the molec-ular transitions detected towards the source position, the spec-trum was extracted from the pixel that corresponds to the peakof the continuum emission (see Table 3), and a Gaussian fit wasapplied to the line profile (see Fig. C.1 in the Appendix). Theresulting values are listed in Table 7 and are estimates of the lineintensities towards the peak, and not the full integrated emissionover the maps. Most of the line profiles show a single centralcomponent centred at V source . When more than one component isobserved, the intensity listed in Table 7 is the sum of all the indi-vidual components. For the sources without on-source detection,the 3 σ upper limit per 1 km s − , that is, 0.013 Jy beam − km s − ,is used in the comparison.Figure 12 shows the line intensities from Table 7 as a func-tion of L bol and T bol . The binary systems are represented byempty bars in order to distinguish them from the single systemsbecause the former are particularly rich in molecular lines. Thebinary systems show emission of almost all the transitions pre-sented in Table 2, with the exception of CH OH. This chemicalrichness appears to be related to the mass content and extent ofthe circumbinary discs, in agreement with the results of Murilloet al. (2018).Considering only the single sources, C O is seen mostly to-wards the least luminous ones (see upper left panel of Fig. 12),while the opposite situation is the case for SO (lower left panelof Fig. 12). The sources with high bolometric luminosity areassociated with high temperatures close to the protostar. Thismeans that for the same continuum flux, the dust mass will belower for sources with high L bol , and thus, a lower column den-sity of C O is expected. On the other hand, SO traces a dif-ferent physical process than C O and may be related with more
Table 7.
Line intensity towards the source position calculated from aGaussian fit.
Source Intensity [Jy beam − km s − ]C O C H H CO + SO GSS30-IRS1 - - - 0.439[GY92] 30 0.081 - - -WL 12 - - - -[GY92] 197 0.226 0.073 - -Elias 29 - - - 1.036 b IRS 43 VLA 1 0.049 - 0.133 0.389 b IRS 43 VLA 2 0.053 - 0.051 -IRS 44 - - 0.124 1.478 b IRAS 16253-2429 0.213 0.206 - -ISO-Oph 203 - - - -IRS 67 A 0.134 - 0.187 -IRS 67 B 0.488 a a b Notes. ( a ) Line profile with three components. ( b ) Line profile with twocomponents. O [ GY ] I S O - O ph [ GY ] I R A S W L I R S B I R S A I R S V L A I R S V L A I R S G SS -I R S E li a s I n t e n s it y [ J y b ea m k m s ] C H0.00.10.20.30.4 H CO + . . . . . . . . . . . . L bol [L ] O I R A S [ GY ] I R S B I R S A [ GY ] G SS -I R S I R S I R S V L A I R S V L A I S O - O ph W l E li a s I n t e n s it y [ J y b ea m k m s ] C H0.00.10.20.30.4 H CO + T bol [K] Fig. 12.
Line intensity as a function of the bolometric luminosity (left) and bolometric temperature (right) towards the source position. Theempty bars represent the binary systems, and the grey horizontal lineindicates a value of 3 σ . energetic processes that are linked to the bolometric luminosity,and thus, the accretion history. If T bol is taken as an evolutionaryindicator, C O and C H emission are associated with the lessevolved sources. In particular, Figs. 2, 7, and 12 show a trendbetween the extent of the C O and C H emission and the evo-lutionary stage of the source. On the other hand, there is no clear
Article number, page 12 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation correlation between the sources with detected H CO + emissionand L bol or T bol .Figure 13 shows the chemical signatures in a plot of L bol as a function of T bol for the sources that were observed as partof this study with additional points from the literature: Class 0sources from Dunham et al. (2015) and Class I sources fromHarsono et al. (2014) . Three groups of sources can be identifiedbased on on-source detections. The least evolved and luminoussources show C O emission towards the source position, whilethe more evolved and luminous sources are associated with SO emission. In addition, the more evolved sources associated withlow luminosities do not show any line detection. As the systemevolves from Class 0 to Class I and later to Class II, the enve-lope mass decreases, leading to a lower gas column density andlines may be harder to detect. This is reflected in the emissionof a high-density tracer, such as C O: the extent of the emissiondecreases with the evolutionary stage of the source, followed bya non-detection towards the more evolved sources. For the lat-ter, emission of more abundant isotopologues, such as C O and CO, is expected to be seen, as in Class II sources. On the otherhand, there is a chemical di ff erentiation between C O and SO ,where the latter may be tracing a more energetic process that isprobably linked to higher accretion rates, which is mostly deter-mined by inner disc properties. Future observations of CO iso-topologues and SO transitions with di ff erent E u values towardsa larger sample of Class I sources will provide a more statisticalperspective.
100 200 300 400036912151821 T bol [K] L bo l [ L ⊙ ] Class I/IIClass 0
This work (single source) LiteratureLiterature with SO ( E u = 60K) detection NGC 1333 IRAS 4A-SENGC 1333 IRAS 4BL1448-mmNGC 1333 IRAS 2ATMC1 TMC1ATMCR1
This work (binary system) D e c r e a s i n g C O Class I C o l d e r S O C O SO (E u = 197 K) No lines on-source
Fig. 13.
Bolometric luminosity as a function of the bolometric tem-perature, highlighting the regions where specific molecular transitionsare detected towards the source position, and well-known Class 0 andI sources from Dunham et al. (2015) and Harsono et al. (2014), re-spectively. This plot only covers the lines listed in Table 2 plus SO , − , from Harsono et al. (2014). The Class 0 covers T bol ≤
70 K(Dunham et al. 2014b), and the Class I / II region includes the sourceswhere no line emission is detected. Two Class I sources in the Harsono et al. (2014) study (TMC1A andTMCR1) show SO , − , emission towards the source position,but this transition is from a lower level with E u of 60 K and does notshow a broad line profile. Therefore, this cold SO emission appears tobe tracing a di ff erent component than the warm SO seen in Fig. 6 OH emission
One of the chemical results of this study is the absence of com-pact CH OH emission towards all of the sources. At large scales,CH OH is found to be in solid form with typical abundanceswith respect to water, CH OH:H O, of about 5% (Bottinelli et al.2010; Öberg et al. 2011; Boogert et al. 2015). CH OH is ex-pected to sublimate o ff dust grains in the inner envelopes and inthe protostellar discs in regions where the temperature increasesabove ∼
90 K (Brown & Bolina 2007). Such warm CH OH hasbeen inferred for some Class 0 sources based on single-dish ob-servations (e.g. van Dishoeck et al. 1995; Schöier et al. 2002;Maret et al. 2005; Jørgensen et al. 2005b; Kristensen et al. 2010)and it has also been imaged with interferometers (e.g. Jørgensenet al. 2005a; Maury et al. 2014; Jørgensen et al. 2016; Higuchiet al. 2018).The absence of CH OH line emission provides strict upperlimits to the column densities of the warm gas-phase CH OH.The upper limit to the column density is estimated using thepredictions for a synthetic spectrum for methanol calculated un-der the assumption of LTE and adopting a kinetic temperatureof 100 K and a typical line width of 5 km s − . For the 3 σ rms noise in the spectra of 13 mJy beam − km s − , this correspondsto an upper limit for the column density of 5 × cm − (as-suming that the emission fills the beam uniformly), which ismore than four orders of magnitude below that of the Class 0protostar IRAS 16293-2422 (Jørgensen et al. 2016). As an il-lustration, Fig. 5 shows the calculated synthetic spectrum for acolumn density that is higher than this by a factor of five (i.e.2.5 × cm − ).The low upper limit for the CH OH column density impliesthat (i) there is no hot-core-like region in the inner envelope closeto the protostar if its envelope density profile can be extrapolatedto the smallest scales and (ii) the gas-phase CH OH averagedover the entire disc is low. For the warm gas in the inner enve-lope, the column density can be translated into a constraint onthe abundance by comparing to the results from Lindberg et al.(2014) for the Class 0 / I protostar R CrA IRS 7B: through lineradiative transfer modelling of ALMA detections of the sameCH OH lines, Lindberg et al. (2014) found a CH OH abundanceof 10 − in the inner region of the 2.2 M (cid:12) envelope. The aver-age envelope mass for our sample is 0.11 M (cid:12) , which is an orderof magnitude below that from R CrA IRS 7B, and because ourbeam size and distance are similar to those of R CrA IRS 7B andour upper limits comparable to the brightest lines in the spec-tra from Lindberg et al. (2014), the upper limit to the CH OHabundance for our sources should be about an order of magni-tude higher than the inferred abundance for R CrA IRS 7B, thatis, ∼ − . Conversely, by ignoring the envelope, an upper limitto the CH OH abundance for the disc can be estimated by com-paring the inferred disc mass from the ALMA dust continuummeasurements to the upper limit of the CH OH column densityassuming that both fill the beam uniformly. Adopting an averagedisc mass of 0.0085 M (cid:12) for our sample and the upper limit tothe CH OH column density of 5 × cm − , the correspondingupper limit for the CH OH abundance is 10 − , averaged overthe entire disc.These numbers may seem at odds with the results from icemeasurements mentioned above, with CH OH at least of order1% relative to water, which in turn has a typically quoted abun-dance ([H O] / [H ]) of 10 − (Boogert et al. 2015). However, inboth cases the assumptions of the physical structures are criti-cal: for example, Lindberg et al. (2014) showed for the proto-stellar envelope that the CH OH abundance would increase by
Article number, page 13 of 21 & A proofs: manuscript no. 34877corr two orders of magnitude if a constant density profile were usedat scales smaller than the disc size (corresponding to the flat-tening of the inner envelope in rotating collapse), rather thana centrally peaked power-law envelope density profile as ex-pected from free-fall. Because a number of sources in our sam-ple show evidence for resolved discs, it is reasonable to applythe same argument, that is, the upper limit to the CH OH abun-dance would be 10 − (rather than 10 − ), which is comparableto the estimates for hot cores towards Class 0 protostars (e.g.Schöier et al. 2002; Maret et al. 2005; Jørgensen et al. 2005b).Likewise, for the disc argument, Persson et al. (2016) showedthat in simple parametrised disc models, only a small fraction,as low as 1%, of the total disc mass may have temperaturesabove 100 K where water could sublimate. These are the re-gions where CH OH would also be in the gas phase, thus theupper limit for the CH OH abundance in the warm parts of thediscs would likewise be less stringent, ∼ − . Furthermore, byanalysing the H O emission towards 4 Class 0 sources, Pers-son et al. (2016) showed that the H O abundance in the warmregions of the discs could be as low as 10 − –10 − , which wouldbe consistent with the CH OH abundance limit derived abovefor typical CH OH:H O ice abundance ratios of ∼ OH is still to some degree caused bychemical e ff ects such as the suppression of methanol formationthat is due to higher temperatures in the precursor environments,as also discussed for the case of Corona Australis by Lindberget al. (2014). Future modelling e ff orts and observation of lower-excited CH OH transitions at large and small scales may providefurther insights. trace accretion shocks? The combination of compact ( r < (cid:48)(cid:48) . ±
10 km s − ; see Fig.C.1) and the veloc-ity gradient perpendicular to the outflow direction suggests thatthe warm SO emission may be tracing warm shocked material,in particular, accretion shocks. Material from the inner envelopefalls onto the circumstellar disc and produces accretion shocksat the envelope-disc interface, releasing molecules from the dustgrains and altering the chemistry. Miura et al. (2017) investi-gated the thermal desorption of molecules from the dust-grainsurface by accretion shocks and found that the enhancement ofsome species (such as SO) can be explained by the accretionshock scenario, where the main parameters are the grain size, thepre-shock gas number density, and the shock velocity. Taking ashock velocity of 10 km s − , Miura et al. (2017) predicted thatSO can be released from the dust-grain surface for a pre-shockgas number density of ∼ cm − .Another formation path for SO is by oxidation of SO in thegas phase (Charnley 1997):SO + OH → SO + H , (4)which is very e ffi cient for T between 100 and 200 K. In thisscenario, the SO abundance depends strongly on the presenceof SO and OH in the gas phase. SO can be released from dustgrains at lower velocities and densities than SO , since SO hasa lower desorption energy ( E d0 ) than SO (2600 K and 3400 K for SO and SO , respectively; Wakelam et al. 2015; Miura et al.2017). In addition, OH is seen towards Class I sources and hasbeen associated with shocked regions in the inner envelope closeto the protostar (Wampfler et al. 2013). In order to test thesetwo scenarios, oxidation of SO or desorption from dust grains,transitions from warm SO ( E u ∼
200 K) need to be observed.In addition to SO and SO , CH OH has also been relatedwith shocked regions (e.g. Avery & Chiao 1996; Jørgensen et al.2007). This raises the question why CH OH is not detected ifSO (or SO) is released from the grain surface by accretionshocks. There are two possibilities: (i) CH OH is not releasedfrom the grain surface because it has a higher desorption energy( E d0 = ∼ cm − ; Miura et al. 2017), or (ii) CH OH isreleased, but is later destroyed by the high velocities. Suutarinenet al. (2014) demonstrated that CH OH is sputtered from ices inshocks with v ≥ − , survives at moderate velocities, but islater dissociated for v ≥
10 km s − .A Keplerian disc or disc winds are less plausible scenarios toexplain the SO emission. For a Keplerian profile, high velocities( ≥ − ) would be reached at scales smaller than 0 (cid:48)(cid:48) . component seen at ∼ (cid:48)(cid:48) . ∼ are tracing a discwind and the emission is observed between ∼
50 and ∼
150 au. Inany case, neither option can be completely ruled out, and higherspatial resolution is required in order to create a PV diagram andobtain a velocity profile for SO .
6. Summary
We presented high angular resolution (0 (cid:48)(cid:48) . ∼
60 au) ALMA ob-servations of 12 Class I sources in the Ophiuchus star-formingregion. The continuum emission at 0.87 mm was analysed to-gether with C O, C S, H CO + , SO , C H, and CH OH, andthe main results are provided below.Of the 12 sources, 2 show no continuum emission nor molec-ular line emission, whereas another 2 show continuum emissionbut no line detection. Two sources are proto-binary systems withvery rich line emission, and the remaining 6 sources show con-tinuum emission plus some of the molecular transitions. C Ois seen towards the less evolved sources and the binary systems,tracing high gas column densities. Keplerian profiles are foundfor three sources, while an infalling profile is seen for one ofthem. More abundant isotopologues, such as C O and CO,may be better disc tracers for the more evolved sources.The non-detection of warm CH OH implies that there in nohot-core-like region in the inner envelope close to the protostar(which follows a power-law density profile) and that the aver-aged CH OH column density over the entire disc is low. Thissuggests that (i) the presence of a disc flattens the envelope den-sity profile at small scales and thus leads to a low column densityof warm material, (ii) only a small portion (1%) of the disc mayhave temperatures above 100 K, or (iii) chemical e ff ects maysuppress the formation of methanol. Clearly, future modellinge ff orts and observations of colder CH OH at envelope and discscales are required in order to provide stronger conclusions.Warm ( E u =
197 K) and compact ( r <
70 au) SO emis-sion is detected towards five of the sources, with particularly Article number, page 14 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation large line widths (between −
10 and 10 km s − ) towards Elias 29and IRS 44. This emission may be related with accretion shocks.The shocks would also liberate CH OH from dust mantles, butit would later be destroyed by the high velocities ( >
10 km s − ).The fact that C O is detected towards the less evolved andless luminous sources agrees with a decrease in gas column den-sity, which is a consequence of the evolution of the system,and with the low column density of material due to high tem-peratures. The envelope mass decreases as the system evolves,therefore the gas column density related to quiescent materialdecreases and lines are hardly detected. However, no similartrend is observed for SO , and instead, a chemical di ff erenti-ation between C O and SO is seen. SO is detected towardsthe most evolved sources with high L bol , and is therefore relatedwith higher accretion rates and a di ff erent physical process.The comparison between disc, envelope, and stellar massesshows a trend between M (cid:63) and M env : the most massive stars arerelated with less envelope material, as expected. In addition, L bol shows a linear dependence with M (cid:63) for Class I sources, wherethe best fit gives L bol = . ± . M (cid:63) . ± . . Assuming that L bol isa consequence of accretion onto the protostar, a mean ˙ M acc of(2.4 ± × − M (cid:12) year − is calculated, with values rangingfrom 7.5 × − to 7.6 × − M (cid:12) year − . If ˙ M acc is constant,the time required to accrete enough mass will be greater that themean lifetime of the embedded stage, supporting the scenario ofepisodic accretion bursts and a variable accretion rate. Withinthis scenario, the Class I sources discussed in this work may bein a quiescent phase, with the exception of L1551 IRS 5.This work shows the importance of a representative sam-ple for exploring the physical and chemical structure of Class Isources, by comparing not only the continuum emission, but alsothe emission of specific molecules and the protostellar massesobtained from the velocity profiles. The observation of disc andwarm gas tracers is crucial in order to interpret the physical andchemical processes and evolution at disc scales. Future observa-tions will provide more statistical results, and the study of otherspecies will contribute to a better understanding of the chemicalevolution of low-mass protostars. Acknowledgements.
We thank the anonymous referee for a number of good sug-gestions that helped us to improve this work. This paper makes use of the follow-ing ALMA data: ADS / JAO.ALMA / NRAO and NAOJ. The group of JKJ acknowledges sup-port from the European Research Council (ERC) under the European Union’sHorizon 2020 research and innovation programme (grant agreement No 646908)through ERC Consolidator Grant “S4F”. Research at the Centre for Star andPlanet Formation is funded by the Danish National Research Foundation.
References
Allen, L. E., Myers, P. C., Di Francesco, J., et al. 2002, ApJ, 566, 993Arce, H. G., Mardones, D., Corder, S. A., et al. 2013, ApJ, 774, 39Artur de la Villarmois, E., Kristensen, L. E., Jørgensen, J. K., et al. 2018, A&A,614, A26Aso, Y., Ohashi, N., Saigo, K., et al. 2015, ApJ, 812, 27Audard, M., Ábrahám, P., Dunham, M. M., et al. 2014, Protostars and PlanetsVI, 387Avery, L. W. & Chiao, M. 1996, ApJ, 463, 642Balança, C., Spielfiedel, A., & Feautrier, N. 2016, MNRAS, 460, 3766Bontemps, S., Andre, P., Terebey, S., & Cabrit, S. 1996, A&A, 311, 858Boogert, A. C. A., Gerakines, P. A., & Whittet, D. C. B. 2015, ARA&A, 53, 541Bottinelli, S., Boogert, A. C. A., Bouwman, J., et al. 2010, ApJ, 718, 1100Brinch, C. & Jørgensen, J. K. 2013, A&A, 559, A82Brinch, C., Jørgensen, J. K., Hogerheijde, M. R., Nelson, R. P., & Gressel, O.2016, ApJ, 830, L16 Brown, W. A. & Bolina, A. S. 2007, MNRAS, 374, 1006Charnley, S. B. 1997, ApJ, 481, 396Choi, M., Tatematsu, K., & Kang, M. 2010, ApJ, 723, L34Chou, T.-L., Takakuwa, S., Yen, H.-W., Ohashi, N., & Ho, P. T. P. 2014, ApJ,796, 70Chuang, K.-J., Fedoseev, G., Ioppolo, S., van Dishoeck, E. F., & Linnartz, H.2016, MNRAS, 455, 1702Drozdovskaya, M. N., van Dishoeck, E. F., Jørgensen, J. K., et al. 2018, MNRAS,476, 4949Dunham, M. M., Allen, L. E., Evans, II, N. J., et al. 2015, ApJS, 220, 11Dunham, M. M., Stutz, A. M., Allen, L. E., et al. 2014a, Protostars and PlanetsVI, 195Dunham, M. M. & Vorobyov, E. I. 2012, ApJ, 747, 52Dunham, M. M., Vorobyov, E. I., & Arce, H. G. 2014b, MNRAS, 444, 887Evans, II, N. J., Dunham, M. M., Jørgensen, J. K., et al. 2009, ApJS, 181, 321Fischer, W. J., Megeath, S. T., Furlan, E., et al. 2017, ApJ, 840, 69Flower, D. R. 1999, MNRAS, 305, 651Frimann, S., Jørgensen, J. K., & Haugbølle, T. 2016a, A&A, 587, A59Frimann, S., Jørgensen, J. K., Padoan, P., & Haugbølle, T. 2016b, A&A, 587,A60Froebrich, D. 2005, ApJS, 156, 169Garrod, R., Park, I. H., Caselli, P., & Herbst, E. 2006, Faraday Discussions, 133,51Girart, J. M., Rodríguez, L. F., & Curiel, S. 2000, ApJ, 544, L153Haisch, Jr., K. E., Barsony, M., Greene, T. P., & Ressler, M. E. 2002, AJ, 124,2841Harsono, D., Jørgensen, J. K., van Dishoeck, E. F., et al. 2014, A&A, 562, A77Hartmann, L. 1998, Cambridge Astrophysics Series, 32Hartmann, L. & Kenyon, S. J. 1996, ARA&A, 34, 207Higuchi, A. E., Sakai, N., Watanabe, Y., et al. 2018, ApJS, 236, 52Jørgensen, J. K., Bourke, T. L., Myers, P. C., et al. 2007, ApJ, 659, 479Jørgensen, J. K., Bourke, T. L., Myers, P. C., et al. 2005a, ApJ, 632, 973Jørgensen, J. K., Johnstone, D., Kirk, H., et al. 2008, ApJ, 683, 822Jørgensen, J. K., Schöier, F. L., & van Dishoeck, E. F. 2004, A&A, 416, 603Jørgensen, J. K., Schöier, F. L., & van Dishoeck, E. F. 2005b, A&A, 437, 501Jørgensen, J. K., van der Wiel, M. H. D., Coutens, A., et al. 2016, A&A, 595,A117Jørgensen, J. K., van Dishoeck, E. F., Visser, R., et al. 2009, A&A, 507, 861Jørgensen, J. K., Visser, R., Sakai, N., et al. 2013, ApJ, 779, L22Jørgensen, J. K., Visser, R., Williams, J. P., & Bergin, E. A. 2015, A&A, 579,A23Kenyon, S. J. & Hartmann, L. 1995, ApJS, 101, 117Kristensen, L. E. & Dunham, M. M. 2018, A&A, 618, A158Kristensen, L. E., van Dishoeck, E. F., Bergin, E. A., et al. 2012, A&A, 542, A8Kristensen, L. E., van Dishoeck, E. F., van Kempen, T. A., et al. 2010, A&A,516, A57Lin, D. N. C., Hayashi, M., Bell, K. R., & Ohashi, N. 1994, ApJ, 435, 821Lindberg, J. E., Charnley, S. B., Jørgensen, J. K., Cordiner, M. A., & Bjerkeli, P.2017, ApJ, 835, 3Lindberg, J. E., Jørgensen, J. K., Brinch, C., et al. 2014, A&A, 566, A74Lique, F., Spielfiedel, A., & Cernicharo, J. 2006, A&A, 451, 1125Lommen, D., Jørgensen, J. K., van Dishoeck, E. F., & Crapsi, A. 2008, A&A,481, 141Mamajek, E. E. 2008, Astronomische Nachrichten, 329, 10Maret, S., Ceccarelli, C., Tielens, A. G. G. M., et al. 2005, A&A, 442, 527Maury, A. J., Belloche, A., André, P., et al. 2014, A&A, 563, L2McClure, M. K., Furlan, E., Manoj, P., et al. 2010, ApJS, 188, 75McMullin, J. P., Waters, B., Schiebel, D., Young, W., & Golap, K. 2007, in As-tronomical Society of the Pacific Conference Series, Vol. 376, AstronomicalData Analysis Software and Systems XVI, ed. R. A. Shaw, F. Hill, & D. J.Bell, 127Miura, H., Yamamoto, T., Nomura, H., et al. 2017, ApJ, 839, 47Mottram, J. C., van Dishoeck, E. F., Kristensen, L. E., et al. 2017, A&A, 600,A99Müller, H. S. P., Thorwirth, S., Roth, D. A., & Winnewisser, G. 2001, A&A, 370,L49Murillo, N. M., Lai, S.-P., Bruderer, S., Harsono, D., & van Dishoeck, E. F. 2013,A&A, 560, A103Murillo, N. M., van Dishoeck, E. F., van der Wiel, M. H. D., et al. 2018, A&A,617, A120Myers, P. C. & Ladd, E. F. 1993, ApJl, 413, L47Nagy, Z., Ossenkopf, V., Van der Tak, F. F. S., et al. 2015, A&A, 578, A124Öberg, K. I., Boogert, A. C. A., Pontoppidan, K. M., et al. 2011, ApJ, 740, 109Ohashi, N., Saigo, K., Aso, Y., et al. 2014, ApJ, 796, 131Osorio, M., D’Alessio, P., Muzerolle, J., Calvet, N., & Hartmann, L. 2003, ApJ,586, 1148Ossenkopf, V. & Henning, T. 1994, A&A, 291, 943Panoglou, D., Cabrit, S., Pineau Des Forêts, G., et al. 2012, A&A, 538, A2Persson, M. V., Harsono, D., Tobin, J. J., et al. 2016, A&A, 590, A33Persson, M. V., Jørgensen, J. K., & van Dishoeck, E. F. 2012, A&A, 541, A39
Article number, page 15 of 21 & A proofs: manuscript no. 34877corr
Podio, L., Codella, C., Gueth, F., et al. 2015, A&A, 581, A85Pontoppidan, K. M., Salyk, C., Bergin, E. A., et al. 2014, Protostars and PlanetsVI, 363Rabli, D. & Flower, D. R. 2010, MNRAS, 406, 95Reipurth, B. 1989, Nature, 340, 42Robitaille, T. P., Whitney, B. A., Indebetouw, R., Wood, K., & Denzmore, P.2006, ApJS, 167, 256Sakai, N., Sakai, T., Hirota, T., et al. 2014, Nature, 507, 78Schöier, F. L., Jørgensen, J. K., van Dishoeck, E. F., & Blake, G. A. 2002, A&A,390, 1001Schöier, F. L., van der Tak, F. F. S., van Dishoeck, E. F., & Black, J. H. 2005,A&A, 432, 369Sheehan, P. D. & Eisner, J. A. 2017, ApJ, 851, 45Shirley, Y. L., Huard, T. L., Pontoppidan, K. M., et al. 2011, ApJ, 728, 143Shu, F., Najita, J., Galli, D., Ostriker, E., & Lizano, S. 1993, in Protostars andPlanets III, ed. E. H. Levy & J. I. Lunine, 3–45Spielfiedel, A., Feautrier, N., Najar, F., et al. 2012, MNRAS, 421, 1891Stahler, S. W., Shu, F. H., & Taam, R. E. 1980, ApJ, 241, 637Suutarinen, A. N., Kristensen, L. E., Mottram, J. C., Fraser, H. J., & vanDishoeck, E. F. 2014, MNRAS, 440, 1844Tabone, B., Cabrit, S., Bianchi, E., et al. 2017, A&A, 607, L6Takakuwa, S., Saito, M., Saigo, K., et al. 2014, ApJ, 796, 1Terebey, S., Shu, F. H., & Cassen, P. 1984, ApJ, 286, 529Tobin, J. J., Hartmann, L., Bergin, E., et al. 2012, ApJ, 748, 16van der Marel, N., Kristensen, L. E., Visser, R., et al. 2013, A&A, 556, A76van der Tak, F. F. S., Black, J. H., Schöier, F. L., Jansen, D. J., & van Dishoeck,E. F. 2007, A&A, 468, 627van Dishoeck, E. F., Blake, G. A., Jansen, D. J., & Groesbeck, T. D. 1995, ApJ,447, 760van ’t Ho ff , M. L. R., Tobin, J. J., Harsono, D., & van Dishoeck, E. F. 2018,A&A, 615, A83Vorobyov, E. I. & Basu, S. 2010, ApJ, 719, 1896Wakelam, V., Loison, J.-C., Herbst, E., et al. 2015, ApJS, 217, 20Walsh, C., Loomis, R. A., Öberg, K. I., et al. 2016, ApJ, 823, L10Wampfler, S. F., Bruderer, S., Karska, A., et al. 2013, A&A, 552, A56White, G. J., Drabek-Maunder, E., Rosolowsky, E., et al. 2015, MNRAS, 447,1996White, R. J., Greene, T. P., Doppmann, G. W., Covey, K. R., & Hillenbrand, L. A.2007, Protostars and Planets V, 117Wilking, B. A., Gagné, M., & Allen, L. E. 2008, Star Formation in the ρ OphiuchiMolecular Cloud, ed. B. Reipurth, 351Yang, B., Stancil, P. C., Balakrishnan, N., & Forrey, R. C. 2010, ApJ, 718, 1062Yen, H.-W., Koch, P. M., Takakuwa, S., et al. 2015, ApJ, 799, 193Yen, H.-W., Koch, P. M., Takakuwa, S., et al. 2017, ApJ, 834, 178Yen, H.-W., Takakuwa, S., Ohashi, N., et al. 2014, ApJ, 793, 1Yoo, H., Lee, J.-E., Mairs, S., et al. 2017, ApJ, 849, 69Zhang, M., Brandner, W., Wang, H., et al. 2013, A&A, 553, A41
Article number, page 16 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation
Appendix A: CH OH emission towards [GY92] 30
Two CH OH transitions are detected towards [GY92] 30, wherethe emission peaks beyond the 25 σ continuum contour and noemission is seen towards the source position (Fig. A.1). Thesetransitions are associated with the lowest E u (65 and 70 K) andthe emission is related with low velocities of between − − from the source velocity. In addition, they presentdi ff erent morphologies, and the spectra taken towards di ff erentpositions show a variation in the intensity of both lines. The in-tegrated spectrum towards a region of ∼ ×
600 au is shown inFig. A.2, where both CH OH transitions are detected. This emis-sion may be related with quiescent envelope material because[GY92] 30 has the more massive envelope from the sample (seeTable 1).
Appendix B: Mass accretion rates
The values of L bol , T bol , M (cid:63) , and ˙ M acc , plotted in Fig. 11, arelisted in Table B.1. L bol , T bol , and M (cid:63) are taken from this workand from the literature (see Tables 1 and 6), while ˙ M acc is calcu-lated from Eq. 3. Appendix C: Gaussian fits
The spectra extracted from the source position (see Table 3) werefitted by a Gaussian profile with one, two, or three components(see Fig. C.1). The resulting intensities from the fit are listed inTable 7 and plotted in Fig. 12. Most of the line profiles show asingle central component centred at V source , with the exception ofsource IRS 67 B and the SO transition, which show more thanone component. C O and H CO + emission towards IRS 67 Bshows three components that are associated with blue-shifted,red-shifted, and a central component associated with more quies-cent material, while the SO emission shows two components re-lated with blue-shifted and red-shifted material (with the excep-tion of GSS30-IRS1, where only blue-shifted emission is seen). Article number, page 17 of 21 & A proofs: manuscript no. 34877corr
Fig. A.1.
Left : Contour maps of CH OH 7 − − − E and CH OH 7 − A + towards [GY92] 30. The contours start at 3 σ and follow a step of 3 σ ( σ = − km s − ), representing velocities from -0.5 to 0.5 km s − . The dashed black contours show the continuum emission for valuesof 4 and 25 σ . The yellow star indicates the position of the 2D Gaussian fit, and the synthesised beam is represented by the black filled ellipse.The dotted box represents the region from which the spectrum of Fig. A.2 is integrated. Right : Spectra towards di ff erent positions marked on thecontour maps. The grey dashed horizontal line represents a value of 3 σ, and all the spectra are rebinned by a factor of 4. The dotted black verticallines indicate the rest frequency of the two CH OH transitions.Article number, page 18 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation
Table B.1.
Bolometric luminosity, bolometric temperature, and stellar masses of Class 0 and I sources from this work and from the literature, withthe calculated mass accretion rates.
Source L bol T bol M (cid:63) ˙ M acc References[L (cid:12) ] [K] [M (cid:12) ] [10 − M (cid:12) yr − ]Class 0NGC 1333 IRAS 4A2 1.9 51 0.08 2.27 Choi et al. (2010)VLA 1623A 1.1 10 0.22 0.48 Murillo et al. (2013)L1527 IRS 1.97 44 0.30 0.63 Ohashi et al. (2014)IRAS 16253-2429 0.24 36 0.03 0.76 This workClass IR CrA IRS 7B 4.6 89 2.3 0.19 Lindberg et al. (2014)L1551 NE 4.2 91 0.8 0.50 Froebrich (2005); Takakuwa et al. (2014)L1551 IRS 5 22.1 94 0.5 4.22 Kristensen et al. (2012); Chou et al. (2014)TMC1 0.9 101 0.54 0.16 Harsono et al. (2014)TMC-1A 2.7 118 0.68 0.38 Aso et al. (2015)TMR1 3.8 133 0.7 0.52 Harsono et al. (2014)L1489 IRS 3.7 238 1.6 0.22 Yen et al. (2014)L1536 0.4 270 0.4 0.09 Harsono et al. (2014)IRS 63 1.0 327 0.8 0.12 Kristensen et al. (2012); Brinch & Jørgensen (2013)[GY92] 197 0.18 120 0.23 0.07 This workElias 29 18.0 420 2.5 0.69 This workIRS 43 VLA 1 1.65 300 0.9 0.17 This workIRS 43 VLA 2 1.65 300 0.9 0.17 This workIRS 44 7.1 280 1.2 0.56 This workIRS 67 A 1.4 180 1.1 0.12 This workIRS 67 B 1.4 180 1.1 0.12 This work Article number, page 19 of 21 & A proofs: manuscript no. 34877corr
Frequency [GHz] -0.0020.0000.0020.0040.0060.0080.0100.0120.014 I n t e n s it y [ J y b ea m ] Fig. A.2.
Integrated spectrum towards the squared region highlightedin the left panels of Fig. A.1.Article number, page 20 of 21. Artur de la Villarmois et al.: Physical and chemical fingerprint of protostellar disc formation C O C H H CO + SO G SS -I R S [ GY ] [ GY ] E li a s I n t e n s it y [ J y b ea m ] I R S V L A I R S V L A I R S I R A S - I R S A -10 0 1000.040.080.12 -10 0 10 -10 0 10 V LRS V source [km s ] -20 -10 0 10 20 I R S B Fig. C.1.
Spectra towards the source position; a Gaussian fit is overplotted (see Table 7). The blue, magenta, and red curves represent blue-shifted,quiescent, and red-shifted material, respectively. The dashed grey horizontal line shows the value of 3 σσ