Misaligned Twin Molecular Outflows From Class-0 Proto-stellar Binary System VLA 1623A Unveiled by ALMA

A large fraction of low-mass stars are born in binary systems (Leinert et al., 1993; Raghavan et al., 2010). Therefore, investigating how binary systems form is crucial to understanding low-mass star formation. Several theoretical scenarios of binary formation have been suggested including the fragmentation of a gravitationally unstable disk (disk fragmentation model; Adams et al. 1989; Bonnell & Bate 1994) and the fragmentation of a core (the model without turbulence; Matsumoto & Hanawa 2003, turbulent fragmentation model; Offner et al. 2010, and formation of binaries in a cluster ; Bate 2012). Although several observational studies have investigated which mechanism is dominant (e.g., Tobin et al., 2018), the debate is not yet settled.

Previous theoretical studies have indicated that the rotational axes of protostellar disks in a binary system can provide a key to understanding the formation process of binaries. For example, the classical numerical simulations of disk fragmentation predict that the rotational axes of the protostellar disks should be parallel to that of the circum-binary disk (Adams et al., 1989; Bonnell & Bate, 1994). In contrast, other models such as turbulent fragmentation and capture suggest that the rotational axes can be misaligned (Offner et al., 2010; Bate et al., 2010; Bate, 2012). Bate et al. (2000) also posited that misaligned disks should be brought into rough alignment by tidal torques in 20 binary orbital periods. Hence, we need to study young proto-binary systems.

Several observational studies have attempted to investigate the alignments of disks from jet/outflow orientations since jets are considered to be launched perpendicular to disks. These studies, however, have been limited to relatively wide binary systems whose separations involve several thousand au (Lee et al., 2016). Observations of molecular outflows from binary systems can therefore help us to reveal the alignment of the disks in binaries and probe the mechanism of their formation.

In recent studies, an aligned binary was discovered in the L1448 IRS 3 system making it consistent with the classical disk fragmentation scenario (Tobin et al., 2016), or the alignment by tidal torques. In contrast, the IRS 43 and L1551 NE systems observed by Brinch et al. (2016) and Takakuwa et al. (2017), respectively, are found to be misaligned. Thus, the dominant binary formation mechanism continues to be a matter of debate. In thie paper, we examine at high resolution, another binary, VLA 1623A.

VLA 16234-2417 (hereafter VLA 1623) is the prototypical Class 0 object (Andre et al., 1990). It produces the most energetic outflow in the Oph A core (Andre et al., 1990; Dent et al., 1995; Nakamura et al., 2011; White et al., 2014), which itself is located in the Ophiuchus star-forming region at a distance of 120–140 pc (Knude & Hog, 1998; Loinard et al., 2008; Lombardi et al., 2008; Mamajek, 2008). In this study, we adopt a distance to VLA 1623 of 137.3$\pm$1.2 pc (Ortiz-León et al., 2017), since it is one of the most recent values obtained with Very Long Baseline Array (VLBA) observations and it falls into the range of the former measurements, 120–140 pc.

Previous observations of VLA 1623 have revealed a highly complex structure in the protostellar system, as shown in Figure 1, and its physical characteristics are still controversial. From VLA observations, Bontemps & Andre (1997) identified four compact cm sources : VLA 1623A, VLA 1623B (Knot A), VLA 1623W (Knot B), and Knot C. Later, Looney et al. (2000), based on the Berkeley Illinois Maryland Association (BIMA) observations, proposed that the VLA 1623A/B is a binary system. More recently, Maury et al. (2012) suggested that VLA 1623B/W are cloudlets created by outflow shocks. Murillo et al. (2013), however, supported Looney et al. (2000)’s interpretation and claimed that VLA 1623B/W are Class-0 and Class-I protostars based on their spectral energy distributions (SEDs). Interestingly, the other source, Knot C, was detected on the northwest of VLA 1623W in deep VLA observations (Bontemps & Andre, 1997), but was not detected in the Gould’s Belt Very Large Array Survey (Dzib et al., 2013).

In more recent observations, VLA 1623A itself was revealed to be consisting of VLA 1623Aa and VLA 1623Ab (Harris et al., 2018; Kawabe et al., 2018). VLA 1623Aab are surrounded by the Keplerian circum-binary disk (hereafter CB disk) with a radius of 150–180 au(Murillo et al., 2013; Hsieh et al., 2020). Thus, VLA 1623 is an excellent target to study the initial state of the binary system, e.g., by measuring the relative angle of the outflow axes.

The molecular outflow from VLA 1623A has blueshifted and redshifted lobes on both the northwest and southeast of VLA 1623 (Andre et al., 1990; Dent et al., 1995). Andre et al. (1990) concluded that a single biconical flow is being ejected from VLA 1623A/B. Dent et al. (1995), however, suggested that two distinct molecular outflows are being ejected from VLA 1623A/B since two shocked H₂ spots are located in parallel on its northwest.

In this study, we present our 1$\arcsec$ resolution (corresponding to 137 au) ALMA observations in ¹²CO, ¹³CO, and C¹⁸O(2–1) lines toward VLA 1623. First, we describe the details of our ALMA observations in Section 2. In Section 3, we show the results of ALMA observations with which we successfully resolved the two outflows associated with VLA 1623A, and we derive their physical properties. In Section 4, we discuss the origin of the two outflows and attempt to understand the formation of the binary system. Finally, we summarize our main conclusions in Section 5.

The obtained ¹²CO(2–1) results are presented in Figures 2 and 3. As shown in Figure 2, blueshifted and redshifted components are seen in both the northwest and southeast of VLA 1623A/B. Although the overall structure is consistent with previous studies (Andre et al., 1990; Dent et al., 1995; Yu & Chernin, 1997), the outflow has been resolved into twin molecular outflows from the vicinity of VLA 1623A/B as shown in Figure 3.

The top panels of Figure 3 show the southeastern blue lobe and the northwestern red lobe pair, while the bottom panels show the southeastern red lobe and the northeastern blue lobe pair. We assume the velocity range of the molecular cloud is from $V_{\rm LSR}=$-0.5 to $+$6.5 km s^-1 since the systemic velocity of VLA 1623A is $V_{\rm LSR}=$4 km s^-1 from the fitting of the CB disk (Murillo et al., 2013). The axes of the outflow pairs presented in the top and bottom panels are different by roughly 10$\arcdeg$, which are measured by eye. Therefore, we interpret that there are two pairs of outflows being ejected from the vicinity of VLA 1623A/B, as suggested previously by Dent et al. (1995), rather than a single outflow in an edge-on configuration (Andre et al., 1990). As described later in Section 3.2, each of the four lobes is associated with its own jet-like structure, supporting the existence of two molecular outflows. Indeed, the measured inclination angle of the CB disk and circumstellar disks (hereafter, CS disks), 45°–55$\arcdeg$ (Murillo et al., 2013; Harris et al., 2018), is incompatible with an edge-on outflow, where the CB disk or CS disks needs to be $>$80$\arcdeg$ (Andre et al., 1990).

Hereafter, we refer to the outflow shown in the top panel of Figure 3 “outflow-1”, and that in the bottom panel “outflow-2”. We also refer to the blueshifted and redshifted outflow shells (lobes) in outflow-1 as SB1 and SR1, and those in outflow-2 as SB2 and SR2, respectively. The position angles of the four outflow shell axes, which are roughly measured by eye, are listed in Table 2.

In this subsection, we analyze the details of outflow-1 and outflow-2 in Figures 4 (channel maps) and 5 (close-up of channel maps). We find collimated jet-like structures associated with all four outflow whells (SB1, SR1, SB2, and SR2), as shown in Figure 6. Hereafter, we refer to the collimated jet-like structures inside the outflow lobes as jets.

The main features in Figures 4 and 5 are summarized below;

1. a)

   (SB1 and SR1) As already shown in Section 3.1, the blueshifted outflow shell, which corresponds to SB1, is observed in the channels corresponding to $V_{\rm LSR}=$-18.0 to -3.0 km s^-1. In contrast, SR1 is detected in the channels corresponding to $V_{\rm LSR}=+$12.0 to $+$27.0 km s^-1.

2. b)

   (JB1) There is a blueshifted jet (hereafter JB1) in the velocity channels corresponding to $V_{\rm LSR}=$-33.0 to -13.0 km s^-1. JB1 consists of two features in the $V_{\rm LSR}=$-18.0 km s^-1 channel. An integrated intensity map of JB1 is shown in Figure 6 (a). The position angle (PA) of JB1 is approximately 120°. The axis of JB1 is almost the same as that of SB1 (outflow-1). Therefore, JB1 is considered to be associated with outflow-1.

3. c)

   (JR1) In the opposite direction of JB1, a redshifted jet-like structure, JR1, is observed in the velocity channels between $V_{\rm LSR}=+$17.0 km s^-1 and $V_{\rm LSR}=+$32.0 km s^-1. JR1 is considered to be associated with SR1, which is observed in the velocity channel range of $V_{\rm LSR}=+$7.0 to $+$27.0 km s^-1, since the directions are the same. The integrated intensity map of JR1 is shown in Figure 6 (c).

4. d)

   (SB2 and SR2) The blueshifted outflow shell associated with outflow-2 (SB2) is observed in the velocity channel range of $V_{\rm LSR}=$-8.0 to -3.0 km s^-1. In the opposite direction of SB2, SR2 is in the velocity channel range of $V_{\rm LSR}=+$12.0 to $+$22.0 km s^-1. In the magnified view (Figure 5), the redshifted outflow shell on the southeastern side of VLA 1623 in the velocity channel range of $V_{\rm LSR}=+$12.0 to $+$22.0 km s^-1, which is considered to be SR2, is elongated in a direction different from that of JR1.

5. e)

   (JB2) There is a jet-like structure in the SB2 outflow shell in the velocity channel range of $V_{\rm LSR}=$-3.0 to -8.0 km s^-1. We call this jet-like structure JB2. JB2 is weakly detected in the channel maps but is clearer to see in the integrated intensity map, as shown in Figure 6 (d).

6. f)

   (JR2) A compact condensation inside SR2 is found in the velocity channel of $V_{\rm LSR}=+$12.0 km s^-1. This feature seems to be a part of jet-like structure in the $V_{\rm LSR}=+7.0$ km s^-1 channel and the integrated intensity map of Figure 6 (b). Hereafter, we refer to this jet-like structure as JR2. JR2 is found to be composed of four condensations, as shown in Figure 6 (b).

7. g)

   (JB1-2 and JR1-2) Taking a closer look at the center of VLA 1623A/B (Figure 5), the blueshifted structure in the $V_{\rm LSR}=$-8.0 km s^-1 channel is elongated from VLA 1623A in the same direction as JB1. This structure is denoted as JB1-2 in Figure 5. In contrast, a redshifted structure in the channels corresponding to $V_{\rm LSR}=+$17.0 to $+$22.0 km s^-1 in Figure 5 is elongated from the VLA 1623A/B region in the same direction as JR1. We label this structure as JR1-2 in Figure 5. JB1-2 and JR1-2 are considered to be also associated with outflow-1 since their axes are similar to JB1/SB1 and JR1/SR1, respectively.

8. h)

   (compact feature on VLA 1623B) A compact blueshifted component is associated with the northwestern side of VLA 1623B in the velocity channel range of $V_{\rm LSR}=$-3.0 to $+$2.0 km s^-1 (see Figure 5).

The features are also described in Appendix A. JB1 and JR1 are considered to be associated with outflow-1, while JB2 and JR2 are considered to be associated with outflow-2. The velocities and position angles of the jets and shells are summarized in Table 2. Table 2 also lists the opening angles of each outflow lobe, as measured by eye. We cannot calculate the dynamical time of the outflow shells, since the map does not span their entire extents. We do, however, calculate the dynamical time of jets. Since the jets are possibly elongated beyond the map, these dynamical times of jets marks the lower limits. Moreover, we note that the axes of JR1 and JB1, which are associated with outflow-1, are perpendicular to the CB disk PA of 30°–35$\arcdeg$(Murillo et al., 2013; Hsieh et al., 2020). Therefore, outflow-1 is considered to be driven by VLA 1623A.

In this subsection, we investigate the internal structures of the jets and outflow shells using the position-velocity (PV) diagrams in Figures 7 and 8, and the integrated intensity maps in Figure 9. Figure 7 (a) and Figure 8 (a) show PV diagrams cut along the jets associated with outflow-1 and outflow-2, respectively. In Figure 7 (a), several velocity components exhibit mostly linear increases of velocity along the spatial axis, indicating that the corresponding jets and shells are accelerated.

In the redshifted side, the highest velocity component is JR1. It is wiggling along the velocity axis, as shown in the magnified view of JR1 in Figure 10 (b). Also in the redshifted side of Figure 7 (a), two further components are accelerated from VLA 1623A/B and a position 56$\arcsec$ away from VLA 1623A/B, i.e., SR1 and SR1-2. For the latter, we create the integrated intensity map shown in Figure 9 (d). A shell larger than SR1 exists across SR1 on its northwestern side. In the blueshifted side, JB1 is accelerated from a position 45$\arcsec$ away from VLA 1623A/B, and SB1 is accelerated from VLA 1623A/B themselves. At the starting point of JB1, a compact feature is found with a large velocity width. This feature does not appear to be associated with JB1, since that feature is elongated more toward the east compared to JB1 in Figure 6 (a).

The magnified view of Figure 7 (a) is shown in Figure 7 (b). There is both redshifted and blueshifted compact structures at the northwestern side of VLA 1623A. The compact jet-like structure, JR1-2, is accelerated more rapidly compared to SR1 at the northwest of VLA 1623A. At the southeast of VLA 1623A, blueshifted gas is accelerated from VLA 1623A and a position 4$\arcsec$ away from VLA 1623A. The component from VLA 1623A corresponds to JR1-2, as described in Section 3.2.

In Figure 8 (a), the low-velocity jet-like structures JR2 and JB2 are marginally detected in the velocity ranges between $V_{\rm LSR}=+$6.5 and $+$13.5 km s^-1, and $V_{\rm LSR}=$-7.5 and $+$0.5 km s^-1, respectively.

Figures 8 (b) and (c) illustrate PV diagrams cut along the northern edge of SB1 and SR1. In Figure 8 (b), we see that the blueshifted lobe is accelerated two times. We refer to the component accelerated at a position 43$\arcsec$ away from VLA 1623A as SB1-2. The integrated intensity map of SB1-2 is shown in Figure 9 (b). SB1-2 is located across the southeastern side of SB1 . In Figure 8 (c), the redshifted component is accelerated three times, at VLA 1623A/B and at position 23$\arcsec$ and 64$\arcsec$ away from VLA 1623A/B. From its position, the latter correspond to SR1-2. We refer to the component 23 $\arcsec$ from VLA 1623A/B as SR1-3. It is rapidly accelerated compared to SR1. As shown in Figure 9 (e), SR1-3 is marginally detected inside SR1.

The numerous outflow features described here were likely produced by episodic mass ejections. Such episodic mass ejections have been reported in other Class-0/I protostars (e.g., HH 47 IRS; Arce et al. (2013), CARMA-7; Plunkett et al. (2015)). In those previous studies, episodic mass ejections were believed to be linked to episodic mass accretions onto the central protostars. The timescale of the mass ejection events in outflow-1 is calculated to be 750-1300 years for the redshifted outflow lobe (SR1), assuming $V_{\rm LSR}$=18 km s^-1, $V_{\rm sys}=$4 km s^-1, and an inclination of 55$\arcdeg$. The timescale of episodic mass ejections of the blueshifted outflow (SB1) is calculated to be 1100 years, assuming $V_{\rm LSR}=$-14 km s^-1 and the same assumptions made for SR1.

Figure 10 (a) illustrates a closer look at JR1, revealing its wiggling structure in position. A wiggle-like structure in velocity is also observable in Figures 10 (c). The images in Figures 10 (a) and (c) are created from the maps with a 0.2 km s^-1 grid to observe the wiggle velocity structures in more detail. In Figure 10 (a), the spatial wiggle extends at least three cycles over 29$\arcsec$ (corresponding to 4000 au). The spatial periods of the wiggles in the integrated intensity map (Figure 10 (a)) and PV diagram (Figure 10 (c)) are roughly the same.

In Figure 6 (a), JB1 also seems to be wiggling. This is clearest in the $V_{\rm LSR}=$-18.0 km s^-1 channel of Figure 4. The wiggle in JB1 is, however, not as obvious compared to JR1. Therefore, we mainly discuss the wiggle in JR1 in this study.

We derived physical properties of each outflow feature assuming an excitation temperature of 30 K for ¹²CO(2–1) from the past measurement by White et al. (2014) for VLA 1623 outflow. A disk inclination angle of $i_{\rm disk}=$55°   is used according to Murillo et al. (2013) as well as Section 3.4. We calculate the outflow mass ($M_{\rm out}$), momentum ($P_{\rm out}$), and energy ($E_{\rm out}$) assuming local thermal equilibrium (LTE) and using the following equation from Hara et al. (2013);

$M_{j}$ is the mass included in each channel ($j$), $X_{\rm{}^{12}CO}$ is abundance ratio of ¹²CO, $d$ is the distance to the star, $S_{\nu,i,j}$ is the flux density at each pixel ($I$) and each channel ($j$), $T_{\rm ex}$ is the excitation temperature, $v_{j}$ is the line-of-sight velocity difference from the systemic velocity of 4 km s^-1, and $i_{\rm disk}$ is the disk inclination angle.

The calculated mass, momentum, and kinetic energy of the outflows are summarized in Table 2. Notably, our map does not cover the whole area of these outflows. Hence, all values are the lower limits. Since it is difficult to discriminate jets and outflow lobes, we calculate the physical properties of both.

As shown in Table 2, the momentum and the energy of outflow-1 are about 14 times and 7 times larger than those of outflow-2, respectively. This difference implies that the accretion rate of the outflow-1 driving source is larger compared to that of the outflow-2 driving source.

Figures 11 (a), (b), (c), and (d) show the total integrated intensity maps of ¹³CO and C¹⁸O. High-velocity outflow components, whose velocity ranges are $V_{\rm LSR}=$-34.5 to -0.5 (blueshifted) and $+$6.5 to $+$31.5 km s^-1 (redshifted), are overlaid. A compact ($\sim 1\farcs 4$ corresponding to 200 au) core in ¹³CO and C¹⁸O emission is seen at the position of VLA 1623W. To the northwest of VLA 1623A, ¹³CO emission is strongly detected along the northern edge of blueshifted outflow lobe, SB2. In the C¹⁸O maps, cavity-like structures are elongated from VLA 1623A/B to the northwest and southeast. These cavity-like structures are extended along the blueshifted outflow lobes, SB1 and SB2. Hence, these structures likely trace the outflow cavity walls created by the blueshifted molecular outflows. The fact that the blueshifted emission is stronger than the redshifted emission in the cavity wall (discussed later) also support this interpretation. At the position of VLA 1623W, the blueshifted outflow lobe is distorted. We note that C¹⁸O condensation is also seen on the northwestern side of VLA 1623W, the location of which is close to the position of Knot C identified in Bontemps & Andre (1997).

Figure 12 shows the PV diagram cut along the positions of VLA 1623B, VLA 1623W, and Knot C detected in Bontemps & Andre (1997). Significant blueshifted emission is seen along the cavity-wall. The highest velocity component is located at the position of VLA 1623B in ¹³CO and C¹⁸O emission, and the velocity widths (FWHM) at the position of VLA 1623B are 5.9 km s^-1 and 4.2 km s^-1, respectively. The next high-velocity component in blueshifted emission is associated with VLA 1623W, and the velocity widths (FWHM) there are 4.3 km s^-1 and is 3.9 km s^-1, respectively. At the position of the Knot C, no high-velocity component exists, however the velocity width (FWHM) is 2.5 km s^-1 in C¹⁸O. This width is 2.5 times larger than the average velocity dispersion in Oph-A core of 0.43 km s^-1 (corresponding to FWHM of 1.0 km s^-1) calculated from the results of single-dish observations of the C¹⁸O(3–2) line (White et al., 2014).

In Figures 13 (a) and (b), we plot the peak temperature map of ¹²CO on the C¹⁸O total integrated intensity map. At the position of VLA 1623B, ¹²CO emission exhibits the highest temperature in the map at 89.3 K $\pm$0.4 K. This value is higher than $\sim$60 K found by Maury et al. (2012) as the peak temperature of ¹²CO(2–1) obtained by SMA observations with a spectral resolution of 0.5 km s^-1 and the very extended configuration, giving the 0$\farcs$4 beam. The good $uv$ coverage of our observations likely resolve out less emission than those of (Maury et al., 2012). Assuming that ¹²CO is optically thick, the peak temperature corresponds to the physical temperature, or is a lower limit to it in the presence of beam dilution. The positions of VLA 1623W and Knot C show higher temperatures of 35–40 K compared to the area inside the outflow lobes of $\sim$30 K. At the positions of both VLA 1623B/W, the high-temperature areas are elongated along the PAs of the dust continuum sources identified by Harris et al. (2018), as shown in Figures 13 (b).

We estimate the excitation temperatures considering the optical depths calculated from ¹²CO , ¹³CO and C¹⁸O spectra, as these will provide more precise values. The spectra at the positions of VLA 1623 A/B/W, and Knot C are shown in Figure 14. The optical depths and excitation temperatures are calculated using the following equations, assuming the abundance ratios between ¹²CO and ¹³CO, and ¹³CO and C¹⁸O are 71 and $\sim$6, respectively (Frerking et al., 1987);

where $T^{*}_{\rm R}({\rm{}^{12}CO})$, $T^{*}_{\rm R}({\rm{}^{13}CO})$, and $T^{*}_{\rm R}({\rm C^{18}O})$ are the radiation temperatures and $\tau({\rm{}^{12}CO})$ and $\tau({\rm{}^{13}CO})$ are the optical depths of each line. We calculated optical depths within the velocity ranges shown in Figure 14, i.e., where the emission of both lines is detected above $5\sigma$. We avoid the velocity ranges where spectra exhibit self-absorptions. The excitation temperatures of ¹²CO and ¹³CO, i.e., $T_{\rm ex}({\rm{}^{12}CO})$ and $T_{\rm ex}({\rm{}^{13}CO})$, are calculated from the following equations;

where $T_{\rm bg}=2.7$ K is the temperature of the cosmic microwave background radiation. The resulting values are summarized in Table 3. The excitation temperatures of ¹²CO calculated at the positions of VLA 1623W and Knot C ($T_{\rm ex}\sim$35 K) are higher than $\sim 26$ K, the averaged value integrated in SB2. All values shown in Table 3 are lower limits since all spectra exhibit self-absorption or the effects of resolved-out emission at the systemic velocity of $V_{\rm LSR}$=4 km s^-1.

From our observations, we find twin outflows launched from the VLA 1623A/B region. The driving source of these outflows could be two of the three dust continuum sources, i.e., VLA 1623Aab and VLA 1623B. The latter source, however, is dynamically interacting with outflow-2 from the evidences listed below. Thus, we conclude that the both outflows are ejected from VLA 1623Aab as discussed later.

• •

  VLA 1623W is located on the blueshifted outflow cavity wall traced by C¹⁸O emission, and a core-like structure is associated with VLA 1623W. The blueshifted outflow lobe is distorted at the position of VLA 1623W, indicating that VLA 1623W is affected by outflow-2.

• •

  The velocity widths of blueshifted ¹³CO and C¹⁸O emission exhibit the largest values at the positions of VLA 1623B/W, i.e., 4.3–5.9 km s^-1 in ¹³CO and 3.9–4.2 km s^-1 in C¹⁸O compared to the typical FWHM of C¹⁸O emission in Oph-A core, 1.0 km s^-1 (White et al., 2014).

• •

  The brightness temperature of ¹²CO emission is 89.3$\pm$0.4 K at the position of VLA 1623B, a value higher than that obtained in a previous study (Maury et al., 2012). At the positions of VLA 1623W and Knot C, the ¹²CO brightness temperature of 35–40 K is higher than that obtained inside the outflow lobes.

The sudden increase of the temperature and the velocity width at the position of VLA 1623B/W likely reflects gas heating due to outflow shocks (Bachiller, 1996; Hirano & Taniguchi, 2001). Hence, both VLA 1623B/W are dynamically interacting with outflow-2. Knot C likewise has a nature similar to that of VLA 1623W, i.e., a local increase of brightness temperature, association with a C¹⁸O condensation, and location inside the outflow cavity-wall. This similarity suggests that VLA 1623W is a shocked cloudlet. VLA 1623B/W have the same nature, i.e., the increasing of velocity widths and the brightness temperature. Therefore, we conclude that VLA 1623B is also a cloudlet shocked by outflow-2.

Although some previous studies have claimed that VLA 1623B/W are Class-0 and Class-1 protostellar candidates (Looney et al., 2000; Murillo & Lai, 2013), our results support the scenario that VLA 1623B/W are shocked cloudlets, as suggested by Bontemps & Andre (1997) and Maury et al. (2012). Hsieh et al. (2020) proposed that the increase of velocity width around VLA 1623B is due to the shock of the accretion flows. Careful inspection is, however, required since our results show that the blueshifted C¹⁸O emission is strongly affected from the molecular outflow and the origin of the blueshifted C¹⁸O is not necessarily accretion flows. Murillo & Lai (2013) performed SED fitting and estimated that the dust temperature is 5 K. Our result, which shows ¹²CO brightness temperature of $\sim$90 K, invalidates the existence of a cold core associated with VLA 1623B.

If VLA 1623B is indeed a shocked cloudlet and not protostellar in nature, both outflow-1 and outflow-2 must be ejected from VLA 1623A, i.e., VLA 1623Aa and VLA 1623Ab. It may yet be possible that VLA 1623B/W are protostars surrounded by envelopes heated from the outflow-2 shock. Even so, both outflow-1 and outflow-2 are still likely launched from VLA 1623Aab. Outflow-1 is likely ejected from VLA 1623A since the PA of outflow-1 is perpendicular to the CB disk and the root of outflow-1 is located at VLA 1623A (see Figure 5 and Section 3.1). VLA 1623A is considered to be also the driving source of outflow-2, since outflow-2 interacts with VLA 1623B and cannot be driven by it. It is beyond the scope of this study to probe whether or not a protostar exists inside the shock heated envelope around VLA 1623B.

As discussed in the previous section, two molecular outflows are likely being ejected from VLA 1623Aa and VLA 1623Ab. The blue- and redshifted lobes of outflow-1 and outflow-2 are located on opposite sides of VLA 1623A, i.e., the northwestern redshifted lobe and southeastern blueshifted lobe of outflow-1 and the northwestern blueshifted lobe and southeastern redshifted lobe of outflow-2. These outflows are likely being ejected in opposite directions across the plane of sky. Assuming that the molecular outflows are launched perpendicular to the disks, the direction of the disk rotational axes should be in the opposite direction across the plane of the sky. Harris et al. (2018) suggested that the inclination angles of the CB disk and CS disks around VLA 1623Aa and VLA 1623Ab are well aligned on the basis of their inclinations, $\sim 55\arcdeg$, which were measured from the major- to minor-axis ratios of the dust continuum sources at $\sim 0\farcs 1$ resolution. The disk inclinations derived from the aspect ratios, however, provide us with two possibilities, i.e., the disk’s northwest side corresponds to either its near or the far side. This uncertainty can be resolved with the observed outflow orientations. Here, we introduce a disk inclination angle taking a value of 0 to 180 degrees. Such definition of the inclination provides us with two possible cases $i_{\rm disk}$ and (180$\arcdeg-i_{\rm disk}$) for the CS disks with the same major- to minor-axis ratios. The CB disk and the CS disk around outflow-1 driving source (hereafter, disk-1) is considered to be aligned, since the PA of outflow-1, 120°, is almost perpendicular to the CB disk PA of 25$\arcdeg\pm 4\arcdeg$ (Harris et al., 2018). Assuming that disk-1 is aligned with CB disk, the disk-1 inclination angle is estimated to be 55$\arcdeg$. The inclination of the secondary (the CS disk hosting driving source of outflow-2; hereafter, disk-2) is calculated to be $180\arcdeg-55\arcdeg=125\arcdeg$, since all the CB disk and CS disks in VLA 1623A have similar major- to minor-axis ratio (Harris et al., 2018). Hence, the difference in the inclination between disk-1 and disk-2 is 70°. The overall schematic view of the VLA1623Aab system is shown in Figure 15.

VLA 1623A has been posited to be an equal-mass binary system (Harris et al., 2018; Kawabe et al., 2018) with an intermediate separation of 36 au (see below). We assume that the binary orbit is circular and parallel to the CB disk. Therefore, the true separation can be calculated geometrically from the apparent binary separation using the following equation, $a=a_{\rm app}[\cos^{2}{\phi}\sin^{2}{i}+\sin^{2}{\phi}]^{-1/2}$. Here, $a$ is the binary separation, $a_{\rm app}$ is the apparent binary separation, $\phi$ is the orbit phase angle, and $i$ is the inclination of the binary orbit. $\phi$ is calculated from the apparent angle between the two protostellar disks by the equation, $\tan{\phi}=\tan{(\phi_{\rm app}\cos{i})}$. Using $a_{\rm app}=0\farcs 23$ (corresponding to 32 au), $\phi_{\rm app}=142\arcdeg$ (corresponding to $\phi=155\arcdeg$) from the results of Harris et al. (2018), and the inclination of 55$\arcdeg$ (Murillo et al., 2013), the true separation is calculated to be 36 au. If inclination has the error of $\pm 25\arcdeg$, the true separation is expected to be 31–44 au. The uncertainty of the inclination is calculated from the half-opening angle of the outflow-1. If the inclination is larger than 80$\arcdeg$, outflow-1 should display blue- and red-shifted emission on both sides of VLA 1623A given outflow-1’s half-opening angle of $10\arcdeg$ (Table 2). The separation is roughly consistent with the expected binary separation of 30 au which is calculated from the equation $r=(1.8$–$2.6)a$, which is predicted by Artymowicz & Lubow (1994) from the theoretical models. Here, $r$ is the radius of the inner edge of the CB disk and was measured to be 0$\farcs$4 by Harris et al. (2018).

As described above, disk-1 is considered to be aligned with the CB disk, since the PA of outflow-1’s axis is perpendicular to the PA of the CB disk. Outflow-1 is more prominent in terms of ¹²CO intensity compared to outflow-2 and exhibits a pair of significant jets. This fact supports the idea that disk-1 is aligned with the CB disk, since mass accretion expected to be more efficient onto whichever protostellar disk is better aligned with the CB disk. In this case, disk-2 is misaligned with disk-1 and CB disk, which is in parallel to the binary orbit. We note that it is uncertain which disk, the disk-1 or -2 is associated to which source in the protostellar binary VLA1623 Aab.

Recently, Brinch et al. (2016) found that the disks are misaligned in the Class-I binary system with the intermediate separation, IRS 43 (binary separation is 75$\pm 4$ au). VLA 1623A is similar to IRS 43 in this regard. Since VLA 1623A has been identified as Class-0 source and so had an age younger than 10⁴ yrs (Evans et al., 2009), our results demonstrate that disks can be already misaligned in the earliest phases of proto-binary formation.

There are several possible models to explain the misalignment of the VLA 1623A disks. First, the fragmentation of the parent core before the protostars formed may explain the misalignment (Offner et al., 2010; Bate, 2012), since such models predict disk misalignment. Second, recent theoretical studies have suggested that the misaligned disks result from disk fragmentation. For example, a third companion could cause the misalignment by tidal interaction (Matsumoto et al., 2015). Alternatively, a difference between the large-scale and small-scale angular momentum due to turbulence could also create misalignment, since the local angular momentum determines the direction of the circumstellar disks (Tsukamoto & Machida, 2013). We note, however, that the classical models assuming single unstable disk fragmentation (Adams et al., 1989; Bonnell & Bate, 1994) cannot explain the VLA 1623A system since these models predict disk alignment. More complicated scenarios, including turbulence or third companion, are required.

In this section, we discuss the possible origin of the wiggle and compare the wiggle pattern to models. We only discuss JR1 since its wiggle is clearer compared to that of JB1.