Substructures in the Disk-Forming Region of the Class 0 Low-Mass Protostellar Source IRAS 16293$-$2422 Source A on a 10 au Scale

Recently, rotationally-supported disks are found not only in Class I sources but also in some Class 0 sources (e.g. Yen et al., 2013, 2017; Murillo et al., 2013; Ohashi et al., 2014; Tobin et al., 2015, 2016a, 2016b; Seifried et al., 2016; Lee et al., 2017; Aso et al., 2017; Okoda et al., 2018). In spite of these extensive efforts, it is still controversial when and how a disk structure is formed around a newly born protostar. Moreover, the disk formation process has been revealed to be much more complicated for binary and multiple cases both in observations (Tokuda et al., 2014; Dutrey et al., 2014; Takakuwa et al., 2014, 2017; Tobin et al., 2016a, b; Boehler et al., 2017; Artur de la Villarmois et al., 2018; Alves et al., 2019) and numerical simulations (e.g. Bate & Bonnell, 1997; Kratter et al., 2008; Fateeva et al., 2011; Shi et al., 2012; Ragusa et al., 2017; Satsuka et al., 2017; Price et al., 2018; Matsumoto et al., 2019). For instance, circumbinary/circummultiple disk structures with a spiral structure as well as a circumstellar disk for each component are reported (e.g. Tobin et al., 2016b; Takakuwa et al., 2017; Artur de la Villarmois et al., 2018; Matsumoto et al., 2019; Alves et al., 2019). In addition, it is not clear how molecules are processed during the disk formation process and what kinds of molecules are finally inherited to protoplanetary disks and potentially to planets. Understanding these processes is crucial, as they will provide important constraints on the initial physical and chemical conditions for the planetary-system formation study. In this context, physical and chemical structures and their mutual relation for disk-forming regions of low-mass protostellar sources have been investigated with ALMA (Atacama Large Millimeter/submillimeter Array) (e.g. Sakai et al., 2014a, b; Oya et al., 2016, 2017, 2018, 2019; Imai et al., 2016, 2019; Jacobsen et al., 2019). These studies reveal that infalling envelopes and rotationally-supported disks are not smoothly connected to each other both in physical structure and chemical composition unlike previous expectations.

IRAS 16293–2422 is a well-studied low-mass protostellar source located in Ophiuchus ($d\sim$ 140 pc; Ortiz-León et al., 2017). It is a Class 0 binary system, consisting of Source A and Source B, which are separated by 5″. Both the components are famous for their richness in complex organic molecules (COMs) and are known as hot corinos (e.g. van Dishoeck et al., 1995; Cazaux et al., 2003; Schöier et al., 2002; Bottinelli et al., 2004; Kuan et al., 2004; Ceccarelli, 2004; Chandler et al., 2005; Caux et al., 2011; Jørgensen et al., 2012). Recently, the chemical characteristics of Source B has extensively been investigated by the PILS (Protostellar Interferometric Line Survey) program with ALMA (e.g. Coutens et al., 2016; Jørgensen et al., 2016; Lykke et al., 2017; Manigand et al., 2020b). As well, the characteristics of Source A has recently been reported by Manigand et al. (2020a) in the same program. Meanwhile, Source A was investigated in terms of the kinematic structures; the rotating signature of the gas surrounding the protostar was detected thanks to its nearly edge-on configuration (e.g. Pineda et al., 2012; Favre et al., 2014). Oya et al. (2016) reported that the kinematic structure of the gas in the vicinity of the protostar in Source A can be disentangled into the two major components; the infalling-rotating envelope and the rotating disk. They evaluated the protostellar mass ($M=0.75$ $M_{\odot}$) and the specific angular momentum of the infalling-rotating envelope ($j=1.3\times 10^{-3}$ km s^-1 pc) of this source under the assumption of the inclination angle of 60° (0° for a face-on configuration). The transition from the infalling-rotating envelope to the rotating (Keplerian like) disk was found to be occurring at the radius of $(40-60)$ au from the center of gravity, which corresponds to the radius of the centrifugal barrier of the infalling-rotating envelope. The chemical composition of the gas was found to change drastically in this transition zone.

The chemical evolution in protostellar sources is tightly related to the physical condition, and especially, the gas temperature distribution is of particular importance for its understanding. The temperature distribution in IRAS 16293–2422 has recently been modeled by Jacobsen et al. (2018); their 3D model of the dust and gas surrounding IRAS 16293–2422 Source A and Source B explains the observed distributions of the CO isotopologue lines and the dust emission. The gas dynamics connecting Source A and Source B has also been reported by van der Wiel et al. (2019). As for Source A, Oya et al. (2016) and van’t Hoff et al. (2020) derived the temperature structure on a $(100-200)$ au scale by analyzing the intensity ratios of $K$-structure lines of H₂CS observed at a 0$\farcs$5 resolution ($\sim$70 au). The result is discussed in relation to the drastic chemical change found on a 50 au scale mentioned above (Oya et al., 2016). Detailing a possible substructure in the transition zone is important for understanding the initial condition of the physical/chemical evolution in the disk formation process. However, the previous observations (Oya et al., 2016; van’t Hoff et al., 2020) do not have a sufficient spatial resolution ($\sim$0$\farcs$5; 70 au) for this purpose. In this study, we make use of the ALMA data observed at a high spatial resolution ($\sim$0$\farcs$1; 14 au) and have a close look at the disk/envelope structure of IRAS 16293–2422 Source A with particular emphasis on its substructure and gas-temperature distribution. We note that Maureira et al. (2020) have very recently reported a high resolution observation at 3 mm with ALMA, which can be compared with this study.

We summarize the observation in Section 2 and its overall results in Section 3. We present the kinematic structure of the gas around Source A in Section 4, and the gas temperature distribution in Section 5. Then, we discuss possible interpretations for the disk/envelope system of Source A in Section 6. The major conclusions are summarized in Section 7.

Our observations toward IRAS 16293–2422 were carried out on 21st August 2017 with ALMA during its Cycle 4 operation. Forty-four antennas were used in the observations. We employed the Band 6 receiver to observe the spectral lines of C¹⁷O and H₂CS (Table 1). The field center of the observations was set to IRAS 16293–2422 Source A, which is located at $(\alpha_{\rm ICRS},\delta_{\rm ICRS})=(16^{\rm h}32^{\rm m}22\fs 8713,-24\arcdeg 28\arcmin 36\farcs 502)$. The baseline lengths of the antennas ranged from 21.0 to 3696.9 m. The size of the field of view was 25$\farcs$15, and the typical size of the synthesized beam was $\sim$0$\farcs$1 for each image (see Table 1). The largest recoverable angular scale was 2$\farcs$27. The total on-source time was 70.6 minutes. Four spectral windows were observed. Their spectral resolution and the band width are summarized in Table 2. The bandpass calibration was performed with J1517-2422, while the phase calibration was performed with J1625-2527 every 2 minutes. J1517-2422 and J1733-1304 were observed to derive the absolute flux density scale. The absolute accuracy of the flux calibration is expected to be better than 15 % (ALMA Partnership, 2016).

The continuum and line images were obtained with the CLEAN algorithm. We employed the Briggs weighting with a robustness parameter of 0.5, unless otherwise noted. The 1.3 mm continuum image was prepared by averaging line-free channels, whose total frequency range was 0.3 GHz. The line maps were obtained after subtracting the continuum component directly from the visibility data. The line maps were resampled to make the channel width to be 0.2 km s^-1. A primary beam correction was applied to the continuum and the line maps. The root-mean-square (rms) noise level is 0.3 mJy beam^-1 for the continuum map, while it is 2.5 and 2.0 mJy beam^-1 for the C¹⁷O and H₂CS line images, respectively. The velocity channel maps of the molecular lines were obtained by smoothing the velocity width of 1 km s^-1 so that the rms noise level is 1.1 and 0.9 mJy beam^-1 for the C¹⁷O and H₂CS line images, resepctively. We tried self-calibration in various ways. However, the images are not improved or even deteriorated. Since we discuss the proper motion of the continuum peaks, we employ the images without self-calibration in this paper.

In this section, we investigate the temperature distribution using the high spatial resolution data of H₂CS. So far, the temperature structure on a $(100-200)$ au scale have been studied by Oya et al. (2016) and van’t Hoff et al. (2020) by using the $K$-structure lines of H₂CS observed at a 70 au resolution. The rotation temperature derived from these lines is a good measure of the gas kinetic temperature, because the radiation processes between the different $K_{\rm a}$ levels ($\Delta K_{\rm a}=\pm 2,\pm 4$) are very slow. Hence, they can be used for delineating the temperature structure around the protostar. Oya et al. (2016) derived the rotation temperature at the five positions with offsets of $\pm 1\arcsec$, $\pm 0\farcs 5$, and 0″ from the continuum peak position, which correspond to the envelope, the transition zone (centrifugal barrier), and the disk, respectively, by using the intensity of the $7_{0,7}$–$6_{0,6}$, $7_{2,5}$–$6_{2,4}$, and $7_{4,3}$–$6_{4,2}$/$7_{4,4}$–$6_{4,3}$ lines integrated over the velocity range corresponding to each component. Based on the result, they suggested the temperature rise at the transition zone by comparison of the temperatures at the five positions and discussed the result in relation to the enhancement of COMs near the transition zone. Later, van’t Hoff et al. (2020) observed more H₂CS lines to derive the temperature structure of this source. They use the peak fluxes instead of the integrated intensity to derive the line ratio and delineate the temperature distribution in Source A. The temperature profile along the direction of the disk/envelope system does not reveal the local temperature rise suggested by Oya et al. (2016). Thus, they interpreted the temperature profile in terms of the radiation heating from the protostar. Although these works reveal the fundamental temperature distribution of this source, the spatial resolution is too coarse to examine the temperature structure around the transition zone in detail. Since substructures in the circummultiple system of Source A are now evident, we revisit the temperature structure of this source with the high resolution H₂CS data. We here employ the H₂CS ($7_{2,5}$–$6_{2,4}$) and ($7_{4,3}$–$6_{4,2}$, $7_{4,4}$–$6_{4,3}$) lines for the analysis, because the H₂CS ($7_{0,7}$–$6_{0,6}$) line suffers from the contamination by the H₂CS ($7_{5,2}$–$6_{5,1}$, $7_{5,3}$–$6_{5,2}$) lines. As we describe in Section 4.4.4, the weak contamination for the H₂CS ($7_{2,5}$–$6_{2,4}$) line (Excess 2 in Figure 10c, d) does not matter as long as we discuss the line intensities with PV diagrams or velocity channel maps. The unresolved doublet of H₂CS ($7_{4,3}$–$6_{4,2}$, $7_{4,4}$–$6_{4,3}$) is confirmed to have no significant contamination in the molecular line database and this observation (Figure 11), as the H₂CS ($7_{2,5}$–$6_{2,4}$) line case mentioned in Section 4.4.4.

Figure 12(a) shows the distribution of the ratio of the integrated intensity of the H₂CS ($7_{4,3}$–$6_{4,2}$, $7_{4,4}$–$6_{4,3}$) line relative to that of the H₂CS ($7_{2,5}$–$6_{2,4}$) line, while Figure 12(b) shows the distribution of the rotation temperature of H₂CS calculated from the integrated intensity ratio. Figures 13(a) and 14(a) show the spatial profiles of the integrated intensity ratio and the rotation temperature along the disk mid-plane direction (a red arrow in Figure 6b), respectively. Here, we assume the LTE condition and the optically thin conditions for the two lines. The LTE condition for H₂CS is well fulfilled in this hot and dense region, as revealed with the non-LTE calculation by Oya et al. (2016). In order to verify the optically thin condition, we roughly calculate the optical depths of the H₂CS ($7_{2,5}$–$6_{2,4}$) and ($7_{4,3}$–$6_{4,2}$, $7_{4,4}$–$6_{4,3}$) lines from the observed brightness temperature and the derived rotation temperature. Here, we ignore the contribution of the dust emission, because the treatment of this effect requires detailed radiation transfer and is out of the scope of this paper. The derived optical depths are mostly lower than 0.5, and its median is 0.28 and 0.10 for the H₂CS ($7_{2,5}$–$6_{2,4}$) and ($7_{4,3}$–$6_{4,2}$, $7_{4,4}$–$6_{4,3}$) lines, respectively. The optical depth of the H₂CS ($7_{2,5}$–$6_{2,4}$) line exceeds 1 for some parts. However, such high optical depth parts are in the low temperature regions ($<$100 K). Since we discuss the temperature structure of hot components around the protostar, the optical depth effect does not affect seriously.

The rotation temperature of H₂CS derived from the $7_{2,5}$–$6_{2,4}$ and $7_{4,3}$–$6_{4,2}$/$7_{4,4}$–$6_{4,3}$ lines is confirmed to be consistent with the gas kinetic temperature derived by using RADEX code (van der Tak et al., 2007) with the H₂ density range of $10^{7-9}$ cm^-3, which is relevant to disk/envelope systems (Oya et al., 2016). The rotation temperature of H₂CS is evaluated to be as high as 300 K near A1. A high rotation temperature area is extended from A1 to southwestern part. The rotation temperature apparently decreases as the increasing distance from the protostar A1 position with a slight enhancement at the southwest part. These features are consistent with the result obtained at a lower spatial resolution by van’t Hoff et al. (2020). It is worth noting that the tongue-like feature from A1 (Section 3.2.2) shows high temperature ($>$200 K; Figure 12b). As discussed in Section 4.4, this might be related to the outflowing motion.

We have also derived the rotation temperature from the intensity ratio of the $7_{2,5}$–$6_{2,4}$ line to the $7_{0,7}$–$6_{0,6}$ line for reference. The temperature is generally lower than that derived from the $7_{2,5}$–$6_{2,4}$ and $7_{4,3}$–$6_{4,2}$/$7_{4,4}$–$6_{4,3}$ lines. For instance, the former is 100 K and the latter is 200 K at the continuum peak A3. This is probably because the $7_{0,7}$–$6_{0,6}$ line would sample colder regions and is partly contaminated by the H₂CS ($7_{5,2}$–$6_{5,1}$, $7_{5,3}$–$6_{5,2}$) line (Section 4.4.4). Hence, we use the rotation temperature derived from the $7_{2,5}$–$6_{2,4}$ and $7_{4,3}$–$6_{4,2}$/$7_{4,4}$–$6_{4,3}$ lines in the following discussions.

Figure 15 shows the velocity channel maps of the rotation temperature of H₂CS. Each map is derived based on the intensity ratio of the velocity channel maps of the H₂CS ($7_{2,5}$–$6_{2,4}$) and ($7_{4,3}$–$6_{4,2}$, $7_{4,4}$–$6_{4,3}$) lines. As in Figure 12(b), the optically thin conditions for the two lines and the LTE condition are assumed. In Figure 12, the envelope and disk components are contaminated with each other along the line of sight, because the line emission is integrated for the velocity axis. On the other hand, they are expected to be disentangled to some extent in the velocity channel maps. In Figure 15, the rotation temperature of H₂CS apparently becomes higher as approaching to the protostar A1 position. It is as high as 400 K near A1.

In the panels for the velocity of $-3$ and $+11$ km s^-1 of Figure 15, the rotation temperature is as high as $\sim$400 K at some positions apart from the protostar A1. These positions roughly correspond to the positions within which the C¹⁷O emission decreases. Figures 16(a) and (b) show the maps of the highest intensity ratio of the two H₂CS lines and the highest rotation temperature along the velocity channels, which are derived from the cube data. In other words, the maximum values along the line of sight are shown in these maps. These maps are obtained by using the immoments task of casa with the option “moments = [8]”. Figure 13(b) shows their spatial profiles along the disk mid-plane direction. The maximum intensity ratio along the line of sight steeply increases at the distance of 50 au from the protostar A1, and are almost constant within this radius. The highest rotation temperature along the line of sight shown in Figure 14(b) also steeply increases at the same position but shows a large scatter within it. It should be noted that very high temperatures obtained at some positions would be artificial. Since the intensity ratio at these positions are close to the high-temperature limit (1.47 for $T=\infty$) as shown in Figure 13(b), the derived temperature is sensitive to a small change in the ratio and suffers from a large error. Therefore, the intensity ratio itself better represents the temperature profile. Interestingly, the positions of the steep increase corresponds the transition zone from the circummultiple structure to the circumstellar structure, as mentioned above.

Thus, the local rise of the rotation temperature seems occurring near the transition zone from the circummultiple structure of Source A to the circumstellar structure of the protostar A1. A hint for such a temperature rise in the transition zone was previously suggested by Oya et al. (2016). However, this previous report would have been just fortuitous based on the results for a few positions, given the detailed analysis by van’t Hoff et al. (2020). Now, our high-resolution observation ($\sim$0$\farcs$1) indeed reveals a sharp change in the temperature. The local rise of the rotation temperature of H₂CS most likely reflects a local rise of the gas temperature. Candidate mechanisms causing a local temperature rise are discussed in Section 6.

When the radiation heating by the central protostar (A1) is considered as a heating source, the gas temperature is expected to gradually decrease as a power-law of the distance to the protostar. In the edge-on case, the maximum gas temperature along the line of sight is expected to be as well (Figure 17a). If the protostars other than A1 could contribute to the heating, the temperature profile would be more gentle than the case of the heating only by A1. This effect may contribute to flatten the temperature profile around A1 within the radius of 50 au. Nevertheless, it seems difficult to explain the local steep rise of the temperature on both the northeastern and southwestern sides of A1 by the protostellar heating alone. Thus, another heating mechanism is expected in IRAS 16293–2422 Source A.

The peak temperature profile projected onto the plane of the sky is expected to be flat, if the actual radial temperature profile is flat as well or it has a local rise in a ring-like structure. The observed temperature profile, which shows a steep rise around 50 au and then flattens, requires that the temperature is not significantly higher in the vicinity of the protostar than at the distance of 50 au from the protostar due to some mechanisms. The temperature is naturally expected to increase as approaching to the protostar by the radiation heating, and thus, a completely flat temperature profile within the radius of 50 au is unlikely. It seems more likely that the radial temperature profile has an increase at the radius of 50 au, drops just inside it, and gradually increases as approaching to the protostar.

If the local temperature rise just outside the circumstellar structure of A1 occurs due to an accretion shock by the infalling gas, it will appear in a ring-like structure surrounding the disk structure (Figure 17b). In fact, the model study by Fateeva et al. (2011) shows that bow-shocks can occur near the inner edge of a circumbinary disk. It may heat up the innermost edge of the circummultiple structure in IRAS 16293–2422 Source A. In the edge-on or nearly edge-on case, the maximum gas temperature along the ling-of-sight is expected to show a flat feature as a function of the distance from the protostar (A1). This corresponds to what can be seen in Figures 13(b) and 14(b).

Such a gas temperature distribution with a ring-like structure is also caused by a physical structure of the gas; the stagnated gas in the transition zone from the circummultiple structure to the curcumstellar structure can be piled up vertically to the mid-plane of the disk, and the gas is efficiently heated up by the radiation from the protostar without the shielding by the circumstellar disk structure (Figure 17b). In any case, the distribution of the rotation temperature in IRAS 16293–2422 Source A shown in Figure 16(b) seems consistent with the picture of the hot ring-like region.

The picture of the hot ring-like region would also be related to the local enhancement of the COM emission reported by Oya et al. (2016). The CH₃OH emission and the HCOOCH₃ emission likely come from the ring-like region with a radius of 50 au surrounding the protostar, based on their kinematic structures. According to Oya et al. (2016), the COM emission seems weaker in the circumstellar disk structure than at the transition zone. Although the spatial-resolution of their observation is insufficient, this distribution could be caused by the temperature distribution shown in Figure 17(b). COMs may be re-depleted onto dust grains because of the lower temperature of the disk mid-plane. If this picture is the case, the ‘hot corino’ would not have a simple spherical or disk-like structure in the vicinity of the protostar, but a ring-like structure surrounding the protostar. The gas and dust temperature will rise again in the closest vicinity of the protostar, and COMs can desorb from the dust grains there. Although one may think that the weak COM emission is due to the optically thick dust emission, it is not likely because the H₂CS emission is strongly detected in the same area. As described in Section 4.3.4 for the C¹⁷O and H₂CS case, this situation could be the case if H₂CS and COM distribute in the gas phase of different parts of the disk/envelope system. To understanding the actual situation, it is required to confirm whether the distribution of the COM line emission is really weak within a certain radius. Observations of COM lines at a high spatial resolution will help us in this aspect, which are in progress. Observations at a lower frequency are favorable for this purpose in order to avoid the effect of dust opacity.

We have observed the Class 0 low-mass protostellar source IRAS 16293–2422 Source A in the C¹⁷O and H₂CS lines as well as in the 1.3 mm dust continuum at a high angular-resolution of $\sim$0$\farcs$1 (14 au). The major findings are as follows:

• (1)

  The 1.3 mm continuum emission traces the nearly edge-on disk/envelope system. Moreover, the substructures on a 0$\farcs$5 ($<$100 au) scale are resolved, and 5 continuum peaks are identified. The positions of two continuum peaks (A1 and A2) are consistent with those previously observed at cm wavelength, if their proper motions are considered. They also agree with the recent report on the 3 mm continuum emission. We also detect new continuum peaks (A3, A4, and A1a). Meanwhile, A2$\alpha$ and A2$\beta$ observed at cm wavelength, which are thought to be the ejecta from A2, are missing. Thus, IRAS 16293–2422 Source A is likely a multiple system.

• (2)

  The C¹⁷O ($J$ = 2–1) line emission traces the rotating gas on a 300 au scale centered at the intensity centroid of the dust continuum emission. On the other hand, C¹⁷O is found to be deficient within the radius of 50 au from the intensity centroid of the continuum emission. Enhancement of the temperature, depletion of CO onto dust grains in the mid-plane of the disk/envelope structure, and exhaustion of carbon to non-volatile organic molecules can be considered as the origin of this feature.

• (3)

  The multiple lines of H₂CS mainly trace the disk component around the strongest continuum peak (A1). Even the high-excitation lines ($7_{5,2}$–$6_{5,1}$, $7_{5,3}$–$6_{5,2}$; $E_{\rm u}$ $=$ 375 K) are detected.

• (4)

  The kinematic structures of the C¹⁷O and H₂CS lines are examined by a Keplerian model and an infalling-rotating envelope model. The center of gravity, the inclination angle, and the systemic velocity of the circummultiple structure of Source A traced by C¹⁷O are possibly different from those of the circumstellar structure around the protostar A1 traced by H₂CS.

• (5)

  The distribution of the rotation temperature of H₂CS is derived. In an overall view, the rotation temperature increases as approaching to the protostar A1, as usually expected for the gas temperature profile determined by the radiation heating from the protostar. Interestingly, the rotation temperature shows a steep rise around the transition zone from the circummultiple structure to the circumstellar disk associated to the protostar A1. This local steep rise could be attributed to a possible accretion shock by the infalling gas and/or the thermal heating of the stagnated gas by the protostar. If the dust temperature rises there as well, it would be related to the rich COM emission, which is the chemical characteristics of IRAS 16293–2422 Source A. ‘Hot corino’ may have a ring-like structure, instead of a simple spherical or disk-like structure.

While theoretical studies have extensively progressed in the disk-formation study (e.g. Hennebelle, & Ciardi, 2009; Li et al., 2011; Machida et al., 2011; Tomida et al., 2015; Tsukamoto et al., 2017; Lam et al., 2019), observational studies tend to have been behind them due to the limited high-resolution observations. In these years, the radio observational studies have revealed the fundamental picture of the infalling-rotating envelope and the disk inside it, thanks to the unprecedented angular-resolution and the sensitivity of millimeter and submilimeter interferometers including ALMA. This work demonstrates that the transition zone from the circummultiple disk/envelope to the circumstellar disk is not smooth. Moreover, even substructures of the disk-forming region are now in the scope of the observations. Further study on these issues will provide us with deep insight into the formation process of the disk and the chemical evolution during it.