Extremely Dense Cores Associated with Chandra Sources in Ophiuchus A: Forming Brown Dwarfs Unveiled?

SM1-A was observed with the SMA¹⁹ on 2007 July 29 in its compact-north configuration over the hour angle coverage of −14 to 42. Details of the SMA are described by Ho et al. (2004). These SMA data were originally taken for polarimetric measurements (S. P. Lai 2018, private communication). The continuum data were first published in our previous paper (Nakamura et al. 2012), which describes the details of the observing parameters. Seven out of the eight SMA antennas were used, providing projected baseline lengths from 6.8 to 125.5 m. The atmospheric transparency was good, with the 225 GHz opacity ranging from ∼0.05 to 0.09 measured at the nearby Caltech Submillimeter Observatory. The double sideband receivers were tuned with a local oscillator frequency of 340.8 GHz. The IF frequency is 5 GHz, and in each sideband the correlator covers the 2 GHz bandwidth. Observations of NRAO 530 were interleaved with the target for gain calibration, whose absolute flux density at 340 GHz was measured to be 1.4 Jy by bootstrapping from observations of Uranus. The absolute flux accuracy is ∼15%. Strong quasars 3C273 and 3C454.3 were adopted as the passband calibrators. The raw visibility data were calibrated with an IDL-based reduction package, MIR (Scoville et al. 1993), and the calibrated visibility data were Fourier-transformed and CLEANed with MIRIAD (Sault et al. 1995).

After the normal calibration and imaging processes, a very bright compact source was identified, which enabled us to perform self-calibration. The phase-only self-calibration improved the signal-to-noise ratio of the continuum image by ∼30% and sharpened the image.

SM1-A was also observed at 41 GHz (λ = 7.3 mm) with the Karl G. Jansky VLA (JVLA), which consists of the 27 25 m antennas. The observations were conducted on 2012 September 3 in the B configuration, covering projected baseline lengths from 80 m to 7.4 km. The correlator was configured to have 16 spectral windows with 128 MHz bandwidth each, providing in total ∼2 GHz bandwidth from 39.998–0.064 GHz to 41.884+0.064 GHz. J1256-0547, J1625-2527, and 3C286 were used for bandpass, complex gain, and flux scale calibrations, respectively. In the observing sequence, the fast switching mode between the target and the gain calibrator J1625-2527 separated at ∼1° was adopted. The total on-source integration time was 2880 s. The 7.3 mm continuum image was made with natural weighting to maximize the signal-to-noise ratio, yielding a synthesized beam size of 032 × 015 (P.A. = 114). The FOV of the JVLA 25 m antennas at 41 GHz is ∼73'', and hence VLA 1623, separated by ∼35'' from SM1-A, was observed simultaneously. VLA 1623 and knot-a (or VLA 1623B) were clearly detected in our observations together with SM1-A, and knot-b (or VLA 1623W) was also detected above a 3σ level.

2.3.1. Band-7 Observations in Cycle-0

2.3.2. Band-6 Observations in Cycle-2

2.3.3. Cycle-2 Mosaic Observations

We obtained ALMA Cycle-2 observations at 226 GHz using the 12 m Array with 150 pointings and ACA with 58 pointings to cover a 2' × 3' region of Oph A. We successfully obtained the combined images for the continuum and three isotopic CO (J = 2 − 1) lines: ¹²CO(J = 2 − 1), ¹³CO (J =2 − 1), and C¹⁸O (J = 2 − 1). The details of the observed continuum and molecular lines are summarized in Table 2. The observation parameters are the same as those of the Oph B2 observations (Kamazaki et al. 2018). The reference position of the 12 m Array and the 7 m Array was set to (α_J2000.0, δ_J2000.0) = (16^h27^m26507, −24°31'2863), which is the same as that of Kamazaki et al. (2018). The uv ranges sampled in the 12 m Array and 7 m Array data were 12.5–348 kλ and 8.1–48 kλ, respectively. The minimum uv distance of the combined data corresponds to 25''. We used quasars, J1517-2422 and J1733-1304, for the bandpass calibrators adopted for the 12 m Array and 7 m Array observations, respectively. We observed a quasar J1625-2527 as the phase calibrator for both arrays. We used Titan and Mars for the flux calibrations of the 12 m Array and the 7 m Array observations, respectively. For the absolute flux scale of the solar system objects (i.e., flux calibrators), we used the Butler-JPL-Horizons 2012 model.

Table 2.  ALMA Cycle-2 Mosaic Observations

Continuum and Line	Rest Frequency	ΔV	Beam Size (PA)	Noise Levela
	(GHz)	(km s−1)	arcsec × arcsec (°)
Continuum	226b	⋯	1.4 × 0.92 (−81.1)	0.024 (mJy beam−1)
12CO (J = 2 − 1)	230.538000	0.096	1.39 × 0.91 (89.0)	0.02 (Jy beam−1 km s−1)
13CO (J = 2 − 1)	220.398684	0.096	1.45 × 0.94 (88.0)	0.02 (Jy beam−1 km s−1)
C18O (J = 2 − 1)	219.560358	0.096	1.45 × 0.94 (88.0)	0.015 (Jy beam−1 km s−1)

Notes.

^aNoise levels are measured for a 1 km s⁻¹ channel. ^bHere are averaged frequencies of the center frequencies of two basebands assigned for continuum observations (218 and 234 GHz). This corresponds to 1.3 mm in wavelength.

Download table as:  ASCIITypeset image

The 12 m Array and 7 m Array data were calibrated and imaged using the CASA pipeline version 4.2.2 and version 4.5.3, respectively. We modified the scripts prepared by the ALMA observatory in which the shadowing criterion for the bandpass data was reduced from 7 to 6 m to recover some data flagged by the original scripts. Then we conducted the calibration by ourselves. In this paper, we will describe briefly a combined continuum map at 226 GHz and ¹²CO (J = 2 − 1) detection toward two sources of interest for comparison purposes. The synthesized beams and the sensitivities of the dust continuum and ¹²CO data are summarized in Table 2. In Figure 1, we present the combined 12 m Array and 7 m Array 1.3 mm continuum image of Oph A with contours of blueshifted and redshifted ¹²CO (J = 2 − 1) emission. We also indicate the FOVs of the SMA, ALMA Cycle-0, ALMA 219 GHz Cycle-2, and JVLA observations with circles in Figure 1.

Zoom In Zoom Out Reset image size

The Chandra X-ray observations were performed in May 2000 (Obs ID = 637) and in 2014 December (Obs ID = 17249), both of which include the Oph A region. The comprehensive reports for the former data set have been given by Imanishi et al. (2003) and Gagné et al. (2004). Both data sets have exposure of ∼100 ks. The Oph A region was observed with the ACIS-I imaging array, which is composed of four charge-coupled devices (CCDs) that are front-side illuminated (I0, I1, I2, and I3).

After downloading both of the data sets from the Chandra X-ray Center, we did the data processing and analysis using the Chandra Interactive Analysis of Observations (CIAO) software, developed by the Chandra Science Center. We started the analysis of the latter data set (Obs ID = 17249) simply using the Level 2 events list provided by the Chandra Science Center, which was filtered on the good time intervals, cosmic ray rejection, and position transformation to celestial coordinates (R.A., decl.) from the more primitive Level 1 outputs (see more detail in the Chandra Analysis Guide: http://cxc.harvard.edu/ciao/guides/). As for the former data set (Obs ID = 637), we generated a new Level 2 events list by reprocessing the Level 1 output by ourselves, with updated calibration data. We merged the data taken in 2000 and 2014 to improve the signal-to-noise ratios, using the CIAO software, "reproject_obs."

2.5.1. Spitzer IRAC/Multiband Imaging Photometer for Spitzer (MIPS) Data and Herschel Data

2.5.2. Analysis of J, H, and Ks Band Data

We measured the flux densities of SM1-A and Source-X using the interferometric data available, Spitzer and Herschel data, and have summarized them in Table 3 together with the number of counts by Chandra. The flux densities are measured in the ∼2'' × 2'' regions of spatially resolved cases to exclude extended emission, which can be seen in the SMA visibility distribution (see also Section 3.4). Peak flux densities in units of mJy beam⁻¹ are given for unresolved cases. Upper limits of 3σ in mJy beam⁻¹ are given for non-detections.

SM1-A was not detected at 70 μm in either Spitzer or Herschel data. It was marginally detected at 4.5 and 5.8 μm. SM1-A is also associated with X-ray emission with Chandra, which is already pointed out by Friesen et al. (2014). Source-X was detected only at 219 GHz, 3 mm (Kirk et al. 2017), and X-ray with Chandra. The 3σ upper limits at K_s, IRAC, MIPS, and Herschel bands are almost identical to those obtained for SM1-A.

Below we summarize the observational results for SM1-A and Source-X. The parameters obtained here are summarized in Tables 4 (cores) and 5 (associated CO outflows).

In Figure 1, we present the combined 12 m Array and 7 m Array 1.3 mm continuum image of Oph A obtained in ALMA Cycle-2. In the image, we detected two very compact and bright 1.3 mm cores, SM1-A and Source-X, which are located between VLA 1623 and SM1-A, together with a number of less bright sources. The positions of the two cores and some other young stellar objects are also shown in Figure 1. We will describe the details of this image in a separate paper (C. Hara et al. in preparation). In the present paper, we focus on the two compact, bright cores. The JVLA 41 GHz images of the individual cores are shown in Figure 2, with contours of the ALMA 359 GHz continuum emission. We detect 41 GHz continuum emission toward SM1-A, whereas we do not detect 41 GHz continuum emission above a 3σ level from Source-X.

Zoom In Zoom Out Reset image size

Detection of SM1-A with ALMA has already been reported by Friesen et al. (2014). SM1-A and Source-X were also identified from the 3 mm ALMA Cycle-2 observations (Kirk et al. 2017). However, both SM1-A and Source-X are more prominent in our 1.3 mm image observed with the higher angular resolution.

The Chandra data merged for the data set taken in 2000 and that in 2014 revealed that both SM1-A and Source-X are significantly detected via X-ray; the significance levels of SM1-A and Source-X are 13σ and 19σ, respectively, in the 2–10 keV band, as shown in Figure 3. The positions of Chandra X-ray sources are shown as red crosses in Figure 2. Both of the X-ray detections agree well with the ALMA sources within the Chandra positional accuracy of ∼05–1''. Analysis of available NIR-FIR/Chandra data unveiled the following nature of both sources:

• (1)  
  Both sources are very dim in NIR to FIR. SM1-A is marginally detected only at IRAC 4.5/5.8 μm at a ∼3σ level. Source-X is not detected in all NIR-FIR bands. Although various NIR to FIR data toward these two sources have been available, clear signatures of emission from the central objects have not been previously obtained. In Figure 4, we present 4.5 μm Spitzer images of the region including SM1-A and Source-X.
• (2)  
  Both sources are detected in the Chandra merged map above a 13σ level in the 2–10 keV band. As Imanishi et al. (2003) has already reported, Source-X had an X-ray flare once with a timescale of a few hours. This clearly indicates the magnetic activity on low-mass stellar/substellar core, or a (proto-)BD.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Together with the above IR behavior, these facts indicate "stellar" cores (i.e., R_⊙ size central sources which have already experienced the second core collapse) have already formed at the centers of both sources, but the central objects are still extremely young and embedded. Hereafter, we call such young objects "extremely young Class 0 objects."

Another important piece of evidence for protostellar sources to distinguish between prestellar and protostellar cores is the presence of molecular outflows and jets. In the following, we search for high-velocity components of molecular outflows from SM1-A and Source-X using the ALMA ¹²CO data.

3.3.1. SM1-A

Very compact ¹²CO (J = 3 − 2) and ¹²CO (J = 2 − 1) emission was detected in four channel maps with central velocities of V_lsr = −0.7 km s⁻¹ to 1.6 km s⁻¹ (ΔV = 0.7 km s⁻¹) and −1 km s⁻¹ to 2 km s⁻¹ (ΔV = 1.0 km s⁻¹), respectively. The total integrated intensity maps are shown in Figure 5, together with those in ¹³CO (J = 2 − 1) and C¹⁸O(J = 2 − 1), which were integrated over V_lsr = ∼0.5 km s⁻¹ to 2.0 km s⁻¹. The emission is point-like and the peaks are very close to the continuum source position. For example, both ¹²CO peaks are located 07 north of SM1-A. Since the systemic velocity of the SM1 core is known to be ∼3.8 km s⁻¹ from observations of optically thin molecular lines such as N₂H⁺ (J = 1 − 0) (e.g., Di Francesco et al. 2004), the CO emission is blueshifted by 2–4 km s⁻¹. There is no other significant ¹²CO emission in the ALMA and SMA maps in the 20''–30'' region around SM1-A, although redshifted CO emission located ∼20'' NE of SM1-A is seen in the SMA maps (near the edge of the FOV). This emission likely corresponds to the redshifted part of an outflow from a NIR source, GY 30, found by a past CO outflow search (Kamazaki et al. 2003). The total integrated intensities of the blueshifted ¹²CO (J = 3 − 2) and ¹²CO (J = 2 − 1) emission are 26 ± 0.55 and 7.34 ± 0.28 Jy km s⁻¹, as summarized in Table 5. The beam-deconvolved size and the brightness temperature were measured to be ∼12 × 06 and ∼60–100 K, respectively. The simple interpretation of the blueshifted CO emission is a very compact and low-velocity outflow from the stellar core. The lack of a redshifted counterpart is puzzling and might be due to the asymmetry in velocity structure or distribution of the molecular outflow itself. In addition, surrounding/ambient molecular material potentially could have some influence (absorption) on the outflow, especially around the systemic velocity. The mass of the outflow was estimated to be between 1.3 × 10⁻⁴ M_⊙ and 1.3 × 10⁻⁵ M_⊙, where the lower and upper boundaries are for the optically thin and thick wing emission cases, respectively. (For the wing, we obtained τ ∼ 10 based on unpublished single-dish ¹²CO (J = 3 − 2) and ¹³CO (J = 3 − 2) data.) Here, we assumed LTE condition, the CO fractional abundance of 10⁻⁴, and the excitation temperature of 20 K. The dynamical time is roughly ∼430 yr assuming the outflow length is roughly ∼270/cos i au and the velocity is $3/\sin i$ km s⁻¹ (i is measured from the plane of sky) and i = 45°. Note that even if i = 75°, the time is still as short as ∼550 year. This outflow would be one of the most compact molecular outflows known. The inferred mass loss rate, 10⁻⁶–10⁻⁷ M_⊙ yr⁻¹, is consistent with that expected from a typical low-mass protostar (e.g., Tomisaka 2002; Hara et al. 2013).

Zoom In Zoom Out Reset image size

Table 5.  Outflow Properties of SM1-A and Source-X

Property	Unit	SM1-A	Source-X	BD Modela
12CO(2-1) intensity	Jy km s−1	7.34 ± 0.28	1.14 ± 0.097	⋯
13CO(2-1) intensity	Jy km s−1	2.7 ± 0.14	<0.1	⋯
C18O(2-1) intensity	Jy km s−1	0.26 ± 0.02	<0.15	⋯
12CO(3-2) intensity	Jy km s−1	26 ± 0.55	⋯	⋯
12CO(2-1) Tb peak	K	62 ± 1.6	11.5 ± 1.1	⋯
12CO(3-2) Tb peak	K	104 ± 8.1	⋯	⋯
Outflow velocity	km s−1	∼3	∼(5–10)b	∼2 km s−1
Size	au	274	571	∼350 auc
Dynamical time	year	434	∼(285–628)b	500c
Outflow mass	M⊙	(1.3–3.9) × 10−4	(1.6–2.5) × 10−5d	∼2 × 10−3
Mass loss rate	M⊙ yr−1	(0.3–0.7) × 10−6	(0.2–0.7) × 10−7	∼4 × 10−6

Notes.¹³CO image around Source-X is rather affected by the strong emission in the vicinity of VLA 1623 and SM1, especially at V = 2 km s⁻¹, and the upper limit to the ¹³CO intensity of the outflow component seen in ¹²CO is a tentative value obtained from the rms noise at V_lsr = 7 km s⁻¹.

^aQuantities listed here are taken from Machida et al. (2009), in which a proto–brown dwarf with ∼50 M_jup has an age of <500 years. ^bObtained for i = ±(15–30) degree for both blue and redshifted components. ^cWe assumed the physical quantities at the age of 500 years. ^dObtained using ¹²CO (J = 2 − 1) intensity assuming T_ex = 30–60 K, and the opacity is ∼10.

Download table as:  ASCIITypeset image

3.3.2. Source-X

The ¹²CO (J = 2 − 1) integrated intensity contours superimposed on the 219 GHz continuum image are shown in Figure 6. We detected faint blueshifted and redshifted emission in ¹²CO (J = 2 − 1) to the ∼4'' northeast of Source-X, roughly perpendicular to the elongation of dust emission at 219 GHz. Although there is no southwestern counterpart in the ¹²CO (J = 2 − 1) emission components, a possible interpretation of these components is that they are the high-velocity outflow components from Source-X, and the outflow axis is almost parallel to the plane of sky. If this emission is indeed a low-velocity outflow, then we could compute the outflow parameters of Source-X, which are listed in Table 5. The total integrated intensity of the ¹²CO (J = 2 − 1) emission is estimated to be 1.14 ± 0.097 Jy km s⁻¹, which is smaller than that of SM1-A. The outflow mass is only 10⁻⁵ M_⊙, and the mass loss rate is evaluated to be about 10⁻⁸ M_⊙ yr⁻¹.

Zoom In Zoom Out Reset image size

Figure 7 shows the SEDs of SM1-A and Source-X. There are several detections of cm to sub-mm emission toward SM1-A, but only two detections at 219 and 226 GHz toward Source-X. The 3σ upper limits to NIR to FIR flux densities in Source-X are the same as those in SM1-A. The tentative detections in 4.5 and 5.8 μm are only 3σ upper limits in SM1-A.

Zoom In Zoom Out Reset image size

We compare the SED of SM1-A with the theoretical model of the first core in Figure 7. The SED of SM1-A looks similar to that of the first core, indicating that SM1-A is in the extremely early phase close to the first core formation phase.

The source size can be derived from the beam-deconvolved size of the millimeter and submillimeter continuum emission. For SM1-A and Source-X, the beam-deconvolved sizes derived from the 359/219 GHz images (Figure 2) are 03 × 03 and 06 × 03, respectively. The beam-deconvolved size is also derived to be 02 × 02 from the 41 GHz image for SM1-A. Another method to determine size is visibility fitting, assuming that the source intensity has an axisymmetric Gaussian distribution. For the assumption, we adopt the following relation:

$$\begin{eqnarray}\begin{array}{rcl}\left(\displaystyle \frac{{{UV}}_{1/2}}{100\,k\lambda }\right)\cdot \left(\displaystyle \frac{{{\rm{\Theta }}}_{\mathrm{FWHM}}}{1\,\mathrm{arcsec}}\right) & = & \displaystyle \frac{4\ \mathrm{ln}2}{2\pi }\cdot \displaystyle \frac{360\cdot 3600}{2\pi \cdot {10}^{5}}\\ & = & 0.9111,\end{array}\end{eqnarray} \tag{ 1 }$$

where UV_1/2 is the uv distance where the visibility amplitude is a half of the peak, and Θ_FWHM is the source size (FWHM) in the image plane. In Figure 8, we show the visibility plots of SM1-A. We also compared the 359 and 41 GHz visibility distributions with those expected for the first core model (Tomida et al. 2010). The 3D radiative hydrodynamics (RHD) calculation started from critical Bonnor-Ebert (BE)–like spheres as an initial condition, which have T = 10 K, and densities increased by a factor of 1.6 to make them unstable. The calculations were done for initial clouds with masses of M = 0.1 and 1.0 M_⊙. The SEDs and visibility distributions were calculated for inclination angles of 60°, and those for the 0.1 M_⊙ model giving the better fits to the observations are shown in Figures 7 and 8.

Zoom In Zoom Out Reset image size

The obtained source diameters from the visibility amplitude fitting of SM1-A are 03 (= 41 au) and 017 (= 23 au) for 359 GHz and 41 GHz, respectively. The source diameters derived from the visibility analysis agree well with the beam-deconvolved sizes in the image domain at the same frequencies. There is a large discrepancy between the diameters derived from 359 and 41 GHz, which is likely to originate from the different optical depths (The 359 GHz emission of SM1-A is likely to be optically thick, whereas the 41 GHz emission is optically thin. See Section 4.1.1 for more detail.) The source size is a half that of Friesen et al. (2014), who derived the effective radius of 42 au, where we corrected the difference in the assumed distance.

For Source-X, we do not have enough visibility data to do the same analysis as was done for SM1-A. We adopt the geometric mean of the beam-deconvoled size, 04 (=55 au), as the source size of Source-X. Kirk et al. (2017) measured the deconvolved size of 108 × 036 for Source-X, which is slightly larger than our estimate.

From the merged data, we made averaged source spectra, the response files, and the background files for SM1-A and Source-X, using the CIAO software "combine_spectra." Both of the source regions are a circle with radius of 3 arcsec, centered at the respective sources. A common background region was used for both sources, taken from a source-free region in a circle with radius of 37 arcsec, centered at (α_J2000.0, δ_J2000.0) = (16^h26^m29390, −24°24'5294). The background region is located at the same CCD chip, which both sources are on.

We fitted the spectra using the thin thermal plasma model (the Astrophysical Plasma Emission Code; Smith et al. 2001), along with the photoelectric absorption model (WABS; Balucinska-Church & McGammon 1992). In the photoelectric absorption model, we adopted the Wisconsin cross-sections (Morrison & McCammon 1983) and the Anders & Ebihara (1982) relative abundances. The metal abundances and the plasma temperatures were fixed to 0.3 solar and 5 keV, respectively, based on the previously derived values for young stellar objects (e.g., Imanishi et al. 2001). The best-fit values of the absorbing columns and the absorption-corrected X-ray luminosities in the 0.5–10 keV band, with the errors for them in 90% confidence level, are $({N}_{H},{L}_{X})=({3.0}_{-1.2}^{+2.2}\times {10}^{23}$ cm⁻², ${0.75}_{-0.5}^{+1.0}\times {10}^{29}$ erg s⁻¹) and $({3.4}_{-2.1}^{+3.7}$ × 10²³ cm⁻², ${0.78}_{-0.6}^{+2.1}\,\times {10}^{29}$ erg s⁻¹) for SM1-A and Source-X, respectively.

4.1.1. SM1-A

The most remarkable new finding is that SM1-A is extremely dense and has an optically thick surface at 359 GHz with a radius of r ∼21 au, as derived below. First, the SED fit with a uniform dust temperature (T_d) source model with a radius 015 (=21 au) was taken to derive T_d, the optical depth at 7.3 mm (41 GHz) τ₄₁, and the dust opacity index β. The 7.3 mm continuum emission is assumed to arise from the thermal dust emission, without free–free contamination. These parameters allow us to infer the luminosity, mass, and density of the core. We fit the observed SED using the following equations to estimate T_d,

$$\begin{eqnarray}&&{S}_{\nu }=({B}_{\nu }({T}_{{\rm{d}}})-{B}_{\nu }({T}_{\mathrm{bg}}))[1-\exp (-{\tau }_{\nu })]{\rm{\Omega }},\end{eqnarray} \tag{ 2 }$$

where T_bg is the temperature of the cosmic background radiation and set to 2.7 K. Ω is the source solid angle for SM1-A, which is assumed to be 2.4 × 10⁻¹² sr (corresponding to a source diameter of 41 au) at both 41 and 359 GHz for simplicity. The function B_ν (T) is the Planck function, and the opacity κ_ν (Hilderbrand 1983) is expressed as

$$\begin{eqnarray}&&{\kappa }_{\nu }=0.1{\left(\displaystyle \frac{\nu }{{10}^{12}\mathrm{Hz}}\right)}^{\beta }\,{\mathrm{cm}}^{2}\,{{\rm{g}}}^{-1}.\end{eqnarray} \tag{ 3 }$$

The fit to its SED—that is, the obtained flux densities at 7.3 mm and 835 μm (359 GHz) together with the 24 μm and 70 μm 3σ upper limits—gives us T_d = 40 K, β = 1.5–2, and τ₄₁ = 0.18. The optical depth at 835 μm is expressed as τ₃₅₉ = (7.3/0.835)^β × τ₄₁, hence τ₃₅₉ = 5–14. This value indicates that SM1-A has an optically thick surface at submillimeter wavelengths—that is, a submillimeter photosphere—at T_d = 40 K. On the other hand, the optical depth at 41 GHz is much smaller than unity (i.e., SM1-A is optically thin). The object presumably has centrally condensed density distribution. The optically thin 41 GHz emission can trace the inner denser region, which cannot be seen at the optically thick 359 GHz. This would be the reason why the estimated source size at 41 GHz is smaller than that at 359 GHz.

Second, we estimate the luminosity and density of the core, where we assume a uniform-density spherical core with the opaque surface at r = 21 au and a temperature of 40 K. The luminosity of SM1-A calculated as such is 0.035 L_⊙, directly using

$$\begin{eqnarray}&&L=4\pi {r}^{2}{\sigma }_{\mathrm{SB}}{T}_{{\rm{d}}}^{4},\end{eqnarray} \tag{ 4 }$$

where σ_SB is the Stefan–Boltzmann constant. The opacity-corrected mass is given as

$$\begin{eqnarray}&&M=\displaystyle \frac{{\tau }_{\nu }}{1-\exp (-{\tau }_{\nu })}\ \displaystyle \frac{{S}_{\nu }{D}^{2}}{{\kappa }_{\nu }{B}_{\nu }({T}_{{\rm{d}}})},\end{eqnarray} \tag{ 5 }$$

and we estimate this to be 0.054–0.27 M_⊙ for T_d = 40 K, β = 1.5–2, and D = 137 pc. This is in good agreement with that of Friesen et al. (2014). Assuming spherical symmetry, the mean number density and volume density are estimated to be n = (2.2–8.4) × 10¹¹ cm⁻³ and ρ = (5.2–33) × 10⁻¹³ g cm⁻³, respectively, which meet those expected for a first core (Masunaga et al. 1998; Bate et al. 2002; Saigo et al. 2008; Tomida et al. 2010, 2013). In addition, the luminosity, 0.035 L_⊙, is mostly in the range of that predicted for a first core (L_int < 0.06 L_⊙). For comparison, we list the physical quantities of the first core predicted from numerical simulations in Table 4 (Larson 1969; Masunaga et al. 1998; Saigo et al. 2008). This similarity in physical quantities between SM1-A and the first core models indicates that SM1-A is extremely young object. Taking into account the fact that molecular outflows and X-ray emission are detected, SM1-A is likely to be an extremely young proto-brown dwarf or protostar that already has a second core inside (i.e., an extremely young Class 0 object).

4.1.2. Source-X

The absorbing column densities (N_H) and absorption-corrected X-ray luminosities (L_X) for SM1-A and Source-X are plotted in Figure 9. The values for YSOs in Oph reported by Imanishi et al. (2003) are also plotted, after conversion of the X-ray luminosity using an updated distance of 137 pc. The errors for the X-ray luminosities of the YSOs are not indicated, because Imanishi et al. (2003) did not provide the errors of the X-ray luminosities. Interestingly, the derived X-ray luminosities of SM1-A and Source-X are almost the same level of those for Class II type BDs, although they are not inconsistent with the other types (e.g., Class II non-BDs).

Zoom In Zoom Out Reset image size

The column density of absorbing material derived from the X-ray emission is ∼3 × 10²³ cm⁻², the highest among Class-I Chandra sources in the Oph region. This is more evidence that the sources are extremely young Class 0 objects. On the other hand, column densities are estimated to be an order of 10^25–26 cm⁻² from the obtained masses or number densities, assuming the cores are spherically symmetric and uniform in density (e.g., a core density of ∼10^11–12 cm⁻³ and a core radius of ∼20 au [3 × 10¹⁴ cm] produce the column density to the central stellar core, ∼3 × 10^25–26 cm⁻²). This discrepancy between X-ray and mm/sub-mm estimates will be reconciled if the cores have not spherically symmetric but disk-like structures where column densities decrease rapidly as viewing angles changing from edge-on to pole-on. In fact, Source-X seems to have a disk-like structure perpendicular to the outflow since the beam deconvolved size of the 219 GHz image is 06 × 03 with PA = 124°, as noted in Section 3.5. According to theoretical studies (e.g., Saigo et al. 2008; Bate 2010), a centrifugally supported disk as a remnant of the first core remains even after the stellar core formation in a rotating core. The extremely dense cores in SM1-A and Source-X would be such remnants of the first cores surviving in extremely young Class 0, which can be detected in X-ray due to preferable viewing angles with less absorbing column densities.

As we shown previously, we discovered two extremely young Class 0 objects with substellar masses. The fates of these low-mass objects remain uncertain. Taking into account the masses derived from the dust continuum emission, they can potentially evolve into brown dwarfs unless they gain significant masses from the surroundings. There are at least two widely discussed scenarios for brown dwarf formation (Machida et al. 2009; Basu & Vorobyov 2012). One is the gravitational contraction of a very-low-mass core. This scenario considers that brown dwarfs form similarly to low-mass stars. Starting from a spherical magnetized core with 0.22 M_⊙, Machida et al. (2009) demonstrated that brown dwarfs can form from such a low-mass dense core. Another scenario is the dynamical ejection from multiple systems (e.g., Reipurth & Clarke 2001) or circumstellar disks (e.g., Basu & Vorobyov 2012). Here, we briefly discuss the possibility that these objects evolve into brown dwarfs on the basis of the two scenarios.

4.3.1. Comparison with Machida et al. (2009)

According to the dynamical time of the molecular outflow detected, the ages of the two sources are likely to be 500–1000 years or less after the central stellar core formation. Such very young sources should be very rare even in the Oph star-forming region, hence it will be very unusual that two are independently formed and located closely in the small ∼5000 au region. It should be noted that similar types of sources are not detected via similar ALMA 149–pointing imaging in the Oph B2 region (Kamazaki et al. 2018). One possible common origin of the two sources could be ejections from the VLA 1623 region.

Reipurth & Clarke (2001) proposed that a possible observational test of the ejection scenario is to search for brown dwarfs the vicinity of Class 0 sources with an age of ∼10^4–5 years. If stellar embryos or first core–like dense gas cores are ejected from the Class 0 object in the phase of disk/envelope fragmentation, one might expect to detect one or more (proto) brown dwarfs around the Class 0 object. The apparent separations from the Class 0 object VLA1623A to SM1-A and Source-X are ∼5000 au and 2600 au, respectively. The two sources, SM1-A and Source-X, are located at PA = 30°–50°, measured from VLA1623A, and mostly aligned with the orbital plane of the disk-like envelope seen in C¹⁸O (PA = 32°; Murillo et al. 2013). If we assume their ejection velocity to be roughly 3 km s⁻¹, a factor of ∼1.5 higher than the rotating velocity of the innermost in the envelope, 2 km s⁻¹, and also assume the travels are along the plane of sky, the dynamical times to travel to the current locations are estimated to be 8000 years and 4000 years for SM1-A and Source-X, respectively. If the two objects are moving at 60° to the plane of sky, the timescale is doubled but still roughly consistent with the age of the Class 0 object.

Unexpectedly, VLA1623A was resolved to two sources separated by 02 (∼25 au), with the recent ALMA observations (Haris et al. 2018) and our JVLA 7 mm observations (see Figure 10) suggesting an equal-mass (∼0.05–0.1 M_⊙) protostellar binary. The disk-like structure in C¹⁸O is likely to be a common envelope of the binary. The binary would be evidence for disk fragmentation. Each mass of the binary seems to be larger than SM1-A and Source-X; hence the binary could eject the third less mass object in the system (e.g., Reipurth & Clarke 2001). These observations seem to be consistent with the ejection scenario, especially the hybrid scenario (Basu & Vorobyov 2012); first core–like dense clumps are ejected, and the ejected clumps form very-low-mass stars afterward. It is noted that the discrepancy between the traveling times of SM1-A and Source-X and the dynamical timescales of the outflow would be reconciled with the hybrid scenario—that recent (∼1000 years) second collapses in ejected first core–like clumps triggered outflow and X-ray activity. Furthermore, Tomida et al. (2010) showed with radiation hydrostatic simulations that first cores formed in very-low-mass cores live longer than 10⁴ years. If ejected cores evolve similar to such first cores, they can travel  6000 au at v = 3 km s⁻¹ in 10⁴ years.

Zoom In Zoom Out Reset image size

However, further observations are required to pursue the ejection scenario more quantitatively and carefully. Some clumps may be tidally disrupted during ejection and disperse according to the simulation (e.g., Basu & Vorobyov 2012; Stamatellos & Whitworth 2009). Such failed cores may be found around VLA1623A, and tidally formed arms as fossil of interaction may also be detected, even around SM1-A and Source-X in the more sensitive observations. Since the eastern area of VLA 1623 has no significant dense gas, the other ejected objects might be found from future high-angular resolution, high-sensitivity observations.