SEARCH FOR CIRCUMSTELLAR DISKS AND RADIO JETS IN THE MASSIVE STAR-FORMATION REGION IRAS 23033+5951

Molecular outflows seem to be a common phenomenon among high-mass stars, e.g., Arce et al. (2007). However, there is a deficit in the detection of circumstellar disks and jets associated with massive YSOs and only a few cases have been reported in the literature (e.g., Cepheus A HW2: Rodriguez et al. 1994; Curiel et al. 2006; IRAS 20126+4104: Cesaroni et al. 1999; Trinidad et al. 2005; Sridharan et al. 2005; AFGL 490: Schreyer et al 2006; G24.78+0.08: Beltrán et al. 2005). Hence, it is not completely clear whether massive YSOs are formed by the same process as low-mass stars. Therefore, studies of individual high-mass YSOs are very important to address these issues.

IRAS 23033+5951, with a bolometric luminosity of 10⁴ L_☉, has been classified as a high-mass star formation region, which is embedded in the Cepheus molecular cloud and located at a distance of 3.5 kpc (Sridharan et al. 2002). IRAS 23033+5951 has been detected at 3.6 cm by Wouterloot & Walmsley (1986), but was undetected at 6 cm (Becker et al. 1994). Beuther et al. (2002a) detected a continuum peak at 1.2 mm and two continuum peaks at 2.6 mm toward IRAS 23033+5951. Reid & Matthews (2008) detected at least three continuum peaks at 3 mm with two of them having the necessary mass to carry out the massive star formation, however, only one of them shows evidence of star formation. On the other hand, Williams et al. (2004) detected a single source at 450 and 850 μm toward IRAS 23033+5951, but they estimated a lower mass (∼100–200 M_☉) than that estimated (2000 M_☉) by Beuther et al. (2002a).

Molecular outflows of CO (2-1), HCO⁺ (1-0), CH₃OH (2-1), and SiO (2-1) have been observed in the region. In particular, Beuther et al. (2004) detected a CO (2-1) molecular outflow, which has a mass of 119 M_☉ and a derived outflow rate of ≈6 × 10⁻⁴ M_☉ yr⁻¹. Though there is some high-velocity gas toward the center of the outflow, the spatial distribution of the redshifted and blueshifted components suggests that the high-velocity gas may be due to a second outflow just barely resolved. In addition, H₂O (22 GHz), OH (1665 and 1667 MHz), and CH₃OH (at 95 GHz) masers have been detected toward IRAS 23033+5951 (Wouterloot & Walmsley 1986; Sridharan et al. 2002; Beuther et al. 2002c; Hoglund & Gordon 1973; Braz et al. 1990; Edris et al. 2007; Schnee & Carpenter 2009). The H₂O and OH masers are distributed between 1'' and 4'' from the millimeter, centimeter, and mid-infrared sources in the region and there is no correlation with the direction of the outflow.

In this paper, we analyze continuum and water maser observations carried with the VLA–EVLA (in the transition mode) toward the high-mass star formation region IRAS 23033+5951. In order to search for circumstellar disks and/or molecular outflows in IRAS 23033+5951, we present a study of the nature of the continuum sources detected in the region, as well as the kinematics of the water masers. We detected two continuum sources that could be associated with independent ultracompact H ii regions, or alternatively, tracing a radio jet. In addition, we model the spatio-kinematical distribution of the water masers, finding a rotating and contracting circumstellar disk around a young source of about 18 M_☉.

This paper is organized as follows: observations are reported in Section 2, while observational results are shown in Section 3. Discussion is given in Section 4 and main conclusions in Section 5.

The high-mass star formation region IRAS 23033+5951 was observed with the VLA–EVLA of the NRAO³ (in the transition phase) in the A configuration during 2007 June 27. Water maser and 1.3 cm continuum emission were simultaneously observed. The line emission was observed with a bandwidth of 3.125 MHz centered at the rest frequency of the H₂O 6₁₆ − 5₂₃ transition 22235.080 MHz (shifted by a V_LSR = −55 km s⁻¹), and divided into 63 channels, while the 1.3 cm continuum emission used a bandwidth of 25 MHz and 7 channels. The right and left circular polarizations were sampled at both frequencies. The absolute amplitude calibrator was 1331+305 with a flux density of 2.53 Jy, while the phase calibrator was 2322+509 with a bootstrapped flux of 1.24 ± 0.09 Jy. Reduction and calibration was performed using AIPS with the standard high frequency method and applying the corrections to the observed data with the VLA–EVLA in transition mode. The water maser emission was self-calibrated using the strongest maser observed in the region, and then the phase and amplitude corrections were applied to the 1.3 cm continuum bandwidth (cross-calibration; see Reid & Menten 1990 for details).

The 3.6 cm continuum emission was also observed toward IRAS 23033+5951. A bandwidth of 50 MHz was used and the right and left circular polarizations were sampled. The amplitude and phase calibrators were also 1331+305 and 2322+509, respectively. The flux density of 1331+305 was 5.20 Jy, while the bootstrapped flux of 2322+509 was 1.23 ± 0.03 Jy. Reduction and calibration were made using the standard techniques of AIPS. Given the large amount of closed errors reported during the calibration, we corrected the positions of the antennas (similar to the 1.3 cm data) and used the 3C286 model for the amplitude calibrator. Then, the calibration was performed, first in phase, followed by phase and amplitude.

A contour map of IRAS 23033+5951 at 3.6 cm is shown in Figure 1. Two continuum peaks are clearly detected in the region, which appear partially resolved and are labeled as VLA 1 and VLA 2. In addition, there is marginal evidence of a third continuum peak detected at a level of 3σ, which is labeled as VLA 3. In order to obtain the highest sensitivity, this contour map is made by using natural weighting, however, the angular resolution is slightly reduced (beam size 031 × 025). All sources detected at 3.6 cm are aligned in the northwest–southeast direction and VLA 1 and VLA 2 are separated by about 02. The source VLA 1 is the strongest one, with a flux density of 0.51 mJy, while the source VLA 2 has a flux density of 0.18 mJy. The continuum peak VLA 3, detected at 3σ, has a peak flux of 0.12 mJy beam⁻¹. No continuum emission is detected in the region at 1.3 cm at a level of 3σ (σ = 0.12 mJy beam⁻¹). The main parameter of the sources at 3.6 cm are given in Table 1.

Zoom In Zoom Out Reset image size

IRAS 23033+5951 had been previously detected by Beuther et al. (2002c) at 3.6 cm with lower angular resolution (∼07) than reported in this paper (025 × 018). They detected only a single source elongated in the northwest–southeast direction, while we detect two, or even three, continuum peaks aligned in the same direction. In order to compare previous results with ours, we made a radio continuum map at 3.6 cm (Figure 2) where we show the combined low and high-angular resolution data from both observations. We note that the three continuum peaks detected with high angular resolution are embedded in the extended emission detected at lower angular resolution. The strongest continuum peak (the only peak in the low-resolution map) of the elongated source is almost located between the continuum sources VLA 1 and VLA 2 observed with high angular resolution. Moreover, all continuum peaks are aligned along the elongation of the extended source. Considering CARMA's resolution and positional errors associated to these sources, the elongated source is coincident with the northernmost 3 mm continuum source detected by Schnee & Carpenter (2009), whose emission has a fraction of ∼40% from dust and ∼60% from free–free emission.

Zoom In Zoom Out Reset image size

In addition, nine water maser spots are detected in the region (see Figure 1). Seven masers are clustered and two others appear isolated. However, no water maser emission is spatially associated with the continuum sources, instead, they are located about ∼2'' to the south of VLA 1 and VLA 2. One of the three 3 mm continuum sources (Schnee & Carpenter 2009) is located about 025 to the northeast of the masers (Figure 1), whose emission is only produced by dust. The clustered water masers have velocities from −66.8 km s⁻¹ to −37.2 km s⁻¹ and the strongest water maser observed in the region is located in this group, which has a flux density of 32.23 Jy and a radial velocity of −66.8 km s⁻¹. The main physical parameters of the water maser are given in Table 2.

4.1. Radio Continuum Sources

4.2. Water Masers

Almost all water masers detected toward the high-mass star-forming region IRAS 23033+5951 are distributed in a group or clump with seven masers (labeled M1). However, they are not spatially associated with any centimeter source detected in the region. There is only a millimeter source (Schnee & Carpenter 2009) about ∼025 to the northeast of the water masers.

The water maser spectrum of M1 shows a structure of three peaks, one of them with velocity similar to that of the molecular cloud (−53.1 km s⁻¹) and the other two with blueshifted and redshifted velocities, respectively (see Figure 1). Similar spectra have been reported for several maser sources (NGC 7538: Genzel et al. 1978; S255: Cesaroni 1990; S140: Lekht et al. 1993; and IRAS 20126+4104: Cesaroni et al. 1997) and interpreted as tracers of circumstellar disks. Based on these results, we suggest that the water maser emission in M1 is tracing a circumstellar disk. Moreover, we also find that the masers in M1 are distributed, mainly, in two groups. Water masers with redshifted velocity are located to the northeast, while those with blueshifted velocity are to the southwest (see Figure 1). In addition, the spatial distribution of the masers, oriented in the northeast–southwest direction, is almost perpendicular to the CO massive molecular outflows (northwest–southeast direction; Beuther et al. 2002b), the H₂ jet (Kumar et al. 2002), and the HCO⁺ and SiO molecular outflows (Reid & Matthews 2008), which support that the water maser group M1 is tracing a circumstellar disk. Under this scenario, we are finding a disk-YSO-outflow system toward a high-mass star-forming region, where the YSO is not detected at centimeter wavelengths. There are several reasons why the central source is not detected, e.g., it is not massive enough or is too young to have developed an H ii region. Another explanation for the lack of detection is the very short Kelvin–Helmholtz timescale of the pre-main-sequence stars (Hosokawa & Omukai 2009).

In order to confirm the nature of the water masers in M1, we have built a simple geometrical and kinematic model. The spatial position of the water masers is fitted to the conical equation using the least-squares technique (see Figure 4), which is done with the function LEASTSQ in the package OPTIMIZE of the PYTHON software. This function minimizes the sum of the squares of the fitted equation using the modified Levenberg–Marquardt algorithm. The physical parameters of the fit (see Table 4) indicate that the water masers are tracing an ellipse, which can be interpreted as part of a circumstellar disk (projected on the plane of the sky; see Figure 3) of 003, corresponding to a linear radius of about 110 AU (assuming a distance of 3.5 kpc), with a position angle of 65°. The size of the disk is small, but it is in agreement with the size of the solar system and with the estimated size of the disk in AFGL5142: inner disk of 30 AU and size of 800 AU (Goddi & Moscadelli 2006). We also note that the maser disk and the H¹³CO⁺ rotating toroid (Reid & Matthews 2008) are oriented in a similar direction (65° and 35°, respectively), which could suggest that both structures are associated with the same physical phenomenon to small (110 AU) and large (40000 AU) scale, respectively.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 4. Physical Parameters from the Model for the Circumstellar Disk Traced by the Water Masers

Position Center		Major Axis	Minor Axis	P.A.	Inclination
α(J2000.0)	δ(J2000.0)	(AU)	(AU)	(°)	(°)
23h05m24s949	60°08'1401	25.2 ± 0.2	217.0 ± 3.5	65 ± 1	83 ± 1

Download table as:  ASCIITypeset image

Then, to estimate the mass of the central object, we have assumed that the water masers are located on the border of the circumstellar disk that is gravitationally bound (Keplerian velocity). Under these assumptions, a mass (M = (R³/G(seni)²)(dv_r/dx)²) of about 40 M_☉ (Figure 5(a)) is estimated for the central YSO, which is higher than that expected for a luminosity of 10⁴ L_☉ emitted by the whole region (Sridharan et al. 2002). We therefore suggest that the motion of the water masers also shows, in addition to the keplerian rotation, a motion component in expansion or contraction. Then, assuming that the circumstellar disk is rotating (v_rot) and expanding/contracting (v_exp), the central mass can be calculated as in Uscanga et al. (2008):

$$M \simeq \frac{R\big(v_{{\rm rot}}^{2}+v_{{\rm exp}}^{2}\big)}{2G}.$$

In order to calculate the contribution of the rotation and expansion/contraction velocities to the total motion of the masers, we assume that the observed LSR velocity (V_LSR) of each of the maser spots on the border of the circumstellar disk can be expressed as (Uscanga et al. 2008):

$$V_{{\rm LSR}} = v_{s} + \frac{x^{\prime }}{a}v_{{\rm rot}}\sin i + \frac{y^{\prime }}{a}v_{{\rm exp}}\tan i.$$

Zoom In Zoom Out Reset image size

Then, based on a least-squares fit to the radial velocities (V_LSR) of the maser spots (Figure 5(b)), we find that the water maser, tracing the circumstellar disk, are rotating and contracting with velocities of 18 ± 1 and −1.6 ± 1.1 km s⁻¹, respectively. We propose the disk is contracting because for such an embedded protostar it must be in an early evolutionary stage, so either it is not emitting sufficient UV photons to ionize the region or is very close to completing the rapid accretion stage. Most of the emitted photons can be absorbed by the dust and the accreted material, creating a very dense and optically thick region in the cm range, but which is detectable in the mm range (Churchwell 2002). Using these velocities, a mass of 19 M_☉ is calculated for the central object, which is consistent with the luminosity of the region. This mass is higher than that estimated for the continuum sources VLA 1 and VLA 2 detected in the region (assuming that they are UC H ii regions), but lower than that estimated for the millimeter source reported by Schnee & Carpenter (2009). The main uncertainty in the determination of the velocities comes from the redshifted maser located in the southwest, where blueshifted velocities are expected. However, its velocity (53.0 km s⁻¹) is very similar to the ambient cloud velocity (53.1 km s⁻¹). In order to reduce the error in the determination of the rotation and contraction velocities, this maser is not used in the fit. Finally, as mentioned earlier, if the circumstellar disk, traced by the water masers, is almost perpendicular to the CO molecular outflow observed in the region, then we have found a disk-YSO-outflow system associated with a very young massive object, similar to that found toward low-mass YSOs.

Given that the water masers are not spatially associated with any radio continuum source, we speculate that the masers are associated with a young massive protostar, which is in its last accretion phase, similar to that found in other star formation regions (e.g., AFGL5142; Goddi et al. 2004). Then, some of the UV photons emitted by the central object could be absorbed by the infalling material and observed as a hypercompact H ii at millimeter wavelengths (Churchwell 2002; Keto 2003), but this HC H ii is not powerful enough to be detected at centimeter wavelengths. Therefore, it could be the youngest source observed in the region.

We present observations made with the VLA–EVLA toward the high-mass star-forming region IRAS 23033+5951. Three radio continuum sources were detected in the region at 3.6 cm, which are aligned in the northwest–southeast direction. In order to study the nature of the continuum sources, we analyzed two scenarios: an H ii region and a radio thermal jet. Under the first scenario, we find that VLA 1 and VLA 2 could be consistent with UC H ii regions associated with ZAMS spectral type stars B2 and B2.5. On the other hand, under the radio jet scenario, we suggest that VLA 2 is the driving source, while VLA 1 and VLA 3 are condensations ejected by VLA 2. However, more-sensitive 3.6 cm continuum observations and proper motion measurements of VLA 1 and VLA 3 may be able to determine the nature of the continuum emission.

Almost all water maser spots detected in the region are clustered in a clump located about 2'' to the south of the radio continuum sources. Modeling the spatio-kinematical distribution of the clustered water maser, we find that they are tracing a rather small, rotating and possibly contracting circumstellar disk of about 110 AU in radius, with a central object of about 19 M_☉. Moreover, we suggest that the central massive object, the circumstellar disk, and the CO outflow observed in the region are forming a disk-YSO-outflow, similar to that found in low-mass stars. Finally, we speculate that the central object could be associated with an HC H ii region, which does not have enough ionizing photons to be detected at centimeter wavelengths.

We thank the referee for very useful comments and suggestions on the manuscript. M.A.T. acknowledges support from CONACyT grant 82543. T.R. acknowledges support from CONACyT, CONCyTEG and UG.