PROPERTIES OF THE MOLECULAR CLUMP AND THE ASSOCIATED ULTRACOMPACT H II REGION IN THE GAS SHELL OF THE EXPANDING H II REGION SH 2-104

Abstract

We study the physical and chemical properties of the molecular clump hosting a young stellar cluster, IRAS 20160+3636, which is believed to have formed via the "collect and collapse" process. Physical parameters of the UC H II region associated with the embedded cluster are measured from the radio continuum observations. This source is found to be a typical Galactic UC H II region, with a B0.5 type exciting star, if it is ionized by a single star. We derive a CN/HCN abundance ratio larger than 1 over this region, which may suggest that this clump is being affected by the UV radiation from the H II region.

1. INTRODUCTION

Young stars sometimes form at the periphery of H II regions. In some cases the expansion of H II regions is thought to have triggered the star formation in the fragmented clumps in various ways as summarized by Deharveng et al. (2005). Among these processes, the “collect and collapse” method is believed to have formed the embedded cluster, IRAS 20160+3636, in the molecular ring around the H II region Sharpless 104 (Sh 2-104: hereafter Sh104, Deharveng et al. 2003). In this process the expanding H II region sweeps up the surrounding neutral gas and generates the dense, fragmented shocked layer, which leads to massive star formation (Elmegreen & Lada 1977; Whitworth et al. 1994). Deharveng et al. (2003) have analyzed the Sh104 region, especially on the molecular ring around the H II region, as a typical example of the collect and collapse process. Here we report the properties of the molecular clump embedding IRAS 20160+3636 and the associated ultracompact (UC) H II region at the eastern periphery of the ring-like shell around Sh104.

Sh104 is an optically visible Galactic H II region located in the Cygnus region. This diffuse HII region, excited by an O6V star (Israël 1977; Crampton et al. 1978; Lahulla 1985), is surrounded by a circularly shaped molecular gas ring, probably formed by the expansion of the H II region. The distance to this source was first de-termined kinematically to be about 5.3 kpc by measuring the velocity of the Hα line (Georgelin et al. 1973). The distance to the diffuse HII region was estimated to be 5 kpc by Israël (1977) from the distance modulus of the exciting star, and 4.8 kpc by Lahulla (1985) from UBVRI photometry of stars in this region. Deharveng et al. (2003) used 4 kpc as the distance to this source. We also applied the 4 kpc as the distance to this source.

Using the data taken at 1 GHz with GMRT (Giant Metrewave Radio Telescope), and at 3 mm and 1 mm with ARO (Arizona Radio Observatory) telescopes, we derived physical properties of the associated UC H II region and the molecular clump embedding this stellar cluster. The observations and data used are included in Section 2, observational results and discussions in Section 3, and the summary in Section 4.

 

2. OBSERVATIONAL DATA

2.1. ARO Observations

The millimeter-wave molecular line data were taken in 2006 May using the 12 m telescope at Kitt Peak and the Submillimeter Telescope (SMT)1 at Mt. Graham, Arizona. The receivers used were dual-channel, cooled SIS mixers operated in single-sideband dual polarization mode with image rejection of at least 16 dB. The backends used for the observations were 256 channel filter banks of 500 kHz and 1 MHz resolution operating in parallel mode. The different polarization data were averaged together to improve the spectral rms noise. Pointing was calibrated by observations of planets, Saturn and Jupiter. The field center of all observations was (α, δ)J2000 = (20h17m45s, 36◦46′00”) with the systemic velocity of this source vlsr = 0 km s−1 (cf., Israël 1977). The maps were made in two different grids: (1) for the 12CO (1 − 0), C18O (1 − 0), HCN (1 − 0), and CN (1 − 0) transitions, in the offset range +1′ ≤ Δα ≤ +5′ and −3′ ≤ Δδ ≤ +1′ from the center position by 1′ spacings, and (2) for the HCN (3 − 2) and CN (2 − 1) transitions, in the offset range +0.5′ ≤ Δα ≤ +3.5′ and −2.5′ ≤ Δδ ≤ +0.5′ from the center position by 0.5′ spacings. The data were taken in position-switching mode with the off position (α, δ)J2000 = (20h18m10s, 36◦44′20”). The beam sizes are about 65” and 32” at 100 GHz (12m) and 240 GHz (SMT), respectively. For the data reduction and map images, the GILDAS (Grenoble Image and Line Data Analysis System) software package2 was used.

The observed molecular transitions and system parameters are listed in Table 1. 12CO is known to be sensitive to the kinetic temperature of the cloud because of its low dipole moment (0.1 Debye) and high optical depth, but its rare isotope variant C18O is thought to trace gas density in the region of about nH2 103 cm−3. HCN is commonly used as a dense gas tracer for densities nH2 104 cm−3 and CN is also another tracer of dense gas, with a lower (by a factor of 5) critical density than HCN (cf., Pérez-Beaupuits et al. 2007). The temperature scale was calibrated to the radiation temperature for the 12m data and the antenna temperature scale, for the SMT data. Assuming the source fills the main beam, the main-beam temperature is estimated to be TR = /ηc and TR = /ηmb for the 12m and SMT data, respectively, where the coupling efficiencies are taken to be ηc ∼ 0.9 at 100 GHz and ηmb ∼ 0.7 at around 240 GHz.3

Table 1aFrom NIST Standard Reference Database (F. J. Lovas 2003, http://physics.nist.gov/PhysRefData). bObserved system tempera- tures. cObserved spectral rms (1σ) noise.

2.2. GMRT Observations

The ionized gas toward Sh104 has been mapped at 1.06 GHz with a bandwidth of 8 MHz in 2003 January using the Giant Metrewave Radio Telescope (GMRT) array in Pune, India. The on-source integration time is about 70−80 minutes. The GMRT consists of 30 fully steerable dishes, of 45 meter diameter with a maximum baseline of ∼ 25 km (Swarup et al. 1991). The aperture efficiency of the dishes is ∼ 40% in the 1 GHz band, which implies an effective collecting area of ∼ 19000 m2. The receiver is a prime focus uncooled receiver with a characteristic system temperature of ∼ 70 K. The observations were made in the standard spectral line mode with a spectral resolution of 125 kHz. The sources 3C 48 and 2355+498 were used as the primary flux density and the secondary phase calibrators, respectively. The visibility data were converted to FITS and analyzed using classic aips4 in the standard way. In addition to normal editing of the data, we identified and flagged time ranges affected by scintillation and channels affected by radio frequency interference, after which the central channels were averaged using the AIPS task ‘SPLAT’ to reduce the data volume. We used the AIPS task ‘IMAGR’ to map the full field at this frequency. To obtain the high-resolution image that is also sensitive to extended structure, we employed the SDI CLEANing algorithm (Steer et al. 1984). We used uniform weighting and the 3D option for w-term correction throughout our analysis. The presence of a large number of point sources in the field allowed us to do phase self-calibration to improve the image. At each iteration of self-calibration, the Fourier transform of the image and the visibilities were compared to check for improvement in the source model.

Figure 1 shows the observed radio continuum image of Sh104 at low- and high-resolution with GMRT, as described in the figure caption. The high resolution map used the full UV data (corresponds to ∼ 0.16 − 100 kilo-lambda; giving a resolution of ∼ 2 arcsec), whereas the low-resolution map was made after tapering (∼ 0.16 − 10 kilo-lambda; giving a resolution of ∼ 21 arcsec) the full UV-data. The primary field of view at 1.06 GHz is about 0.4◦. This image has an rms of ∼ 0.13 mJy beam−1. The flux density uncertainly is less than a 2 percent, which is smaller than the rms in the image. The properties of the extended diffuse HII region have been well investigated by previous studies, (e.g., Reifenstein et al. 1970; Georgelin et al. 1973; Israël 1977; Lahulla 1985). Here we focus on the UC HII region and the dense molecular clump associated with IRAS 20160+3636.

Figure 1.(Left) The 1.06 GHz continuum intensity map obtained toward Sh104 with GMRT. The (0′, 0′) position of this map is (α, δ)J2000 = (20h17m45s, 36◦46′00′′). The low-resolution (∼ 31′′ × 21′′ ; P.A.=82◦) result is in grey scale and the high-resolution (∼ 5′′ × 3′′; P.A.=33◦) in contours. Both grey and contour scales start at 3 × 10−4 Jy beam−1, and increase by a factor of 2 for next step. The peak intensities are 3.24 × 10−2 and 1.06 × 10−2 Jy beam−1 for low- and highresolution results, respectively. The continuum source near the offset (∼ 2.3′, −0.3′) coincides with the IRAS point source IRAS 20160+3636. The larger square box shows the region where molecular lines are observed using the ARO telescopes. The smaller square box is the region expanded in the right side: for the high-resolution (right upper) and the low-resolution (right lower). Contour levels are same as those in the left panel.

 

3. RESULTS AND DISCUSSION

3.1. The Ultracompact HII Region

An UC H II region has been found in association with a young stellar cluster, IRAS 20160+3636, deeply embedded in the dense molecular gas (Condon et al. 1998). The UC HII region is marginally resolved in our highresolution (5′′ × 3′′; P.A.=33◦) map and looks spherical (Figure 1), but unresolved in our low-resolution (31′′ × 21′′; P.A.=82◦) map. The size at half maximum was determined by two-dimensional Gaussian fitting for the high resolution result. The deconvolved size 7.0′′ × 6.0′′ corresponds to 0.14 pc × 0.12 pc, at a distance of 4 kpc to this source. Kim & Koo (2001) found that the vast majority of UC H II regions have diffuse extended envelopes and suggested that they are not real ionization-bounded UC H II regions, but compact cores of more extended H II regions (see also Kurtz et al. 1999; Ellingsen et al. 2005). In this context, the UC HII region associated with IRAS 20160+3636 seems to be a bona-fide UC HII region and younger than the UC HII regions with extended envelopes. Wood & Churchwell (1989) identified 1646 IRAS point sources as UC H II region candidates using IRAS color criteria, and argued that 10%−20% of massive stars are in the UC H II region phase. This implies that their ages may be 10%−20% of the typical main-sequence lifetime of massive stars, 1 × 105 yr. Later many of them were found to have extended envelopes with sizes >1′ (Kurtz et al. 1999; Kim & Koo 2001). For comparison, the UC H II region of IRAS 20160+3636 is at a dynamic age of ∼1×104 yr, assuming an expansion velocity of 10 km s−1.

From the 1.06 GHz radio continuum data, we derived the physical parameters of the UC H II region using the formulae of Mezger & Henderson (1967) and listed in Table 2. We assumed that the HII region is spherically symmetric, homogeneous, optically thin, dust-free, and ionization-bounded, and that the electron temperature is 104 K. The derived physical parameters are similar to those of other UC H II regions except electron density and emission measure, which are signficantly smaller than the typical values, ne > 104 cm−3 and EM> 106 pc cm−6. This may originate from the assumption that the observed 1.06 GHz continuum emission is optically thin. UC H II regions can easily get optically thick around 1 GHz and so the previous ra-dio continuum observations of them have been usually carried out at ≥ 5 GHz (Wood & Churchwell 1989; Kurtz et al. 1994). If the UC H II region is optically thick at < 5 GHz as many other UC H II regions are, e.g., G5.89−0.39 (Gomez et al. 1991), the flux density would be proportional to the frequency squared. In that case, the 5 GHz flux density is about 25 times larger than the 1.06 GHz flux density and one can obtain electron density of ∼ 1 × 104 cm−3 and emission measure of ∼ 1 × 106 pc cm−6. We also evaluated the Lyman continuum photon flux under the assumption that the UC H II region is produced by a single ionizing star (Rubin 1968). If a stellar cluster is responsible for the ionization, the most massive star in the cluster may be 2 subclasses later-type than the single-star spectral type equivalent (see Table 7 of Kurtz et al. 1994).

Table 2aDerived by assuming optically thin emission and the distance of 4 kpc.bDiameter at half-maximum determined by two- dimensional Gaussian fitting.

3.2. The Molecular Cloud Associated with the UC H II Region

Figures 2 and 3 show integrated intensity maps of the observed transitions in Table 1 obtained toward the region indicated as a larger square box of the left panel in Figure 1. The molecular cloud hosting IRAS 20160+3636 has a well defined boundary with a simple geometry, isolated from the background gas components. The 12CO (1 − 0) line emission is a little more extended than those of the CN (1 − 0) and HCN (1 − 0) lines, but they all show similar simple morphologies elongated in the north-south direction. We found a dense molecular core near the UC H II region within this cloud, traced by the CN (2 − 1) and HCN (3 − 2) lines (Figure 3). The dense gas is mainly associated with the 1 GHz continuum emission feature, but extended structures exist toward the south direction. As shown in Figure 4, the observed CO spectra clearly show velocity wing components around vlsr ∼ −3 and +3 km s−1. We included the high-velocity component maps in Figure 2 together with the 12CO (1 − 0) map; they appear to be located along the north-south direction: the blue-component around Δδ = 0 and the red-component at Δδ = −2′. The aligned direction of the high-velocity component coincides with the general elongation of the molecular cloud. It is not clear, however, whether this component results from the embedded stellar cluster or the interaction with the expanding H II region. We found no significant peak velocity gradient over the whole cloud.

Figure 2.Velocity integrated intensity (∫ dv) map of the 12CO (1−0) line (left, in grey) and the C18O (1−0) line (right, in solid contours). Crosses are observed positions. (Left) Contour levels of the 12CO (1 − 0) line increase by 5 K km s−1 from 5 K km s−1. High velocity components are shown with white and black solid contours, integrated for v = −4 ∼ −2 km s−1 and 2 ∼ 4 km s−1, respectively. Contour levels for both high velocity components increase by 0.5 K km s−1 from 4 K km s−1. (Right) Contour levels of the C18O (1−0) line increase by 0.4 K km s−1 from 0.2 K km s−1. The 1 GHz continuum intensity in Figure 1 is shown in grey scale together with dotted contours.

Figure 3.Velocity integrated intensity (∫ dv) map of the observed HCN and CN transitions as indicated in each panel. Crosses are observed positions. The 1 GHz continuum intensity in Figure 1 is shown in grey scales. (1) HCN (1 − 0): Contour levels increase by 1 K km s−1 from 0.5 K km s−1. (2) HCN (3 − 2): Contour levels increase by 0.2 K km s−1 from 0.2 K km s−1. (3) CN (1− 0): Contour levels increase by 0.5 K km s−1 from 0.3 K km s−1. (4) CN (2− 1): Contour levels increase by 0.4 K km s−1 from 0.2 K km s−1.

Figure 4.Sample spectra of 12CO (1 − 0), HCN (1 − 0) and CN (1 − 0) taken in declination at an offset Δα = 3′ from a center position of (α, δ)J2000 = (20h17m45s, 36◦46′0′′). The HCN and CN spectra were expanded by factor of 5 for better view.

Sizes of the cloud and the dense core were estimated to be ∼ 4.0 × 2.6 pc (HPW) and ∼ 1.6 × 1.2 pc (HPW), using the HCN (1 − 0) and (3 − 2) lines, respectively, assuming a distance of 4 kpc. We derived the total C18O column density, by assuming optically thin emission (τC18O(1−0) ≈ 0.03 − 0.05 by comparing the intensity with the CO 1 − 0 line at peak, assuming the 16O/18O = 500) and an LTE condition, NC18O = 1.6 ± 0.3 × 1015 cm−2 toward the offset (+2, −1) position. The uncertainty of the column density is from the applied rotational temperature range, Trot = 15−20 K (Deharveng et al. 2003), together with a typical spectral rms noise (1σ) of ∼ 0.06 − 0.07 K of the observed spectra (Table 1). We used the relation NH2 ≈ 0.5 − 1.0 × 107 · NC18O (see discussions and references in Harjunpää et al. 2004; Burgh et al. 2007) to derive the total hydrogen column density. From the estimated total hydrogen column densities we expected visual extinctions of AV ≈ 1 − 3 mag at the periphery of the cloud (AV = NH2/0.94 × 1021, Frerking et al. 1982), and 7 − 20 mag near the peak position, and the averaged AV of about 10 mag over the whole cloud.

Using the equation, Mgas = 2 × NH2 × μmH × Area(inside of HPW), where μ is 1.4 amu per H nuclei and accounts for the mass of He and other elements and the ‘Area’ is estimated from the HPW (4.0 × 2.6 pc), we derived the total gas mass of the molecular cloud, Mtotal ∼ (1 − 3) × 103 M⊙.

3.3. CN and HCN Abundances

Along the molecular ring around Sh104, an enhanced emission at ∼8 μm has been detected in the MSX survey (cf. Deharveng et al. 2003). This band contains emission attributed to large molecules such as polycyclic aromatic hydrocarbons (PAHs), which can be excited in the photodissociation region (PDR) (Leger & Puget 1984). This emission suggests that this region is experiencing significant turbulence caused by the interaction with the ionized gas. CN is chemically linked to HCN and can also be formed by the direct photodissociation of HCN (Huggins et al. 1984). The CN/HCN abundance ratio has been suggested as a probe of PDRs in various objects; it is observed to be typically ≲ 1 in the shielded regions of the interstellar medium, but > 1 in PDRs with enhanced UV field (Fuente et al. 1993, 1995, 2005; Sternberg & Dalgarno 1995; Thi et al. 2004; Jørgensen 2004). We searched for a possible enhancement of the CN/HCN abundance ratio toward the molecular cloud embedding IRAS 20160+3636.

We derived total column densities of CN and HCN by assuming optically thin emission with an LTE condition, and results are included in Table 3. The abundance ratios are also included in the last column of the Table, which may suggest that this molecular cloud is being affected by UV photons from the HII region as shown in the 8 μm data (Deharveng et al. 2003). However, the optical depths of the observed lines appear to be about 7.0 and 6.5 for the CN F=5/2-3/2 and HCN F=2-1 lines toward the dense core by applying the hyperfine ratios (F=5/2-3/2 & 3/2-1/2 for CN and F=2-1 & 0-1 for HCN) to the gaussian fit of the observed lines. Therefore the assumption of optically thin emission can lead to a significant underestimate of the abundances of these species. However, we derived similar abundance ratios all over the observed region, where the optical depth is much less than that at the dense core. Observations of the rare isotope species of CN and HCN with higher spatial resolution are certainly necessary for further discussion on this ratio. We think that the CN/HCN abundance ratio, if obtained with proper accuracy, will play as an important probe in understanding the collect and collapse process better.

Table 3Errors in this table are for the observed 1σ rms values, and a(b) indicates a × 10b. aOffsets in arcminutes from the (0′, 0′) position, (α, δ)J2000 = (20h17m45s, 36◦46′0′′). bExcitation temperature derived from the observed two transitions of each species. cTotal column density ratio of CN and HCN.

 

4. SUMMARY

We have investigated the physical parameters of the fragmented molecular cloud and the UC H II region associated with Sh104. This source is located in the eastern periphery of the gas ring shaped by the expansion of the diffuse H II region, which may have resulted from the “collect and collapse” process.

Comparison between our high-resolution and lowresolution radio continuum images shows that the UC H II region associated with IRAS 20160+3636 is not physically related to the diffuse extended radio continuum emission. This suggests that it might be a bona-fide UC H II region and younger than most other UC H II regions, which usually have extended envelopes. Most derived parameters are similar to the typical values, except for a significantly smaller electron density and emission, which may be an optical depth effect.

The molecular cloud in the eastern periphery, containing IRAS 20160+3636, is well isolated from the background or ambient emission with a roughly circular shape. The total gas mass was found to be about (1 −3)×103 M⊙ from the molecular line observations. From the CN and HCN line observations with the ARO telescopes (12m and SMT), we derived abundance ra-tios, CN/HCN, greater than 1 over the whole cloud, which may suggest that the whole cloud is affected by UV photons as shown in the MSX 8.3 μm observations (cf., Deharveng et al. 2003). But our data suffer large optical depths, and higher spatial resolution observations for rare isotope species for CN and HCN are required to further discuss on this ratio. If the CN/HCN abundance ratio is estimated with proper accuracy, it may be an important probe to understand the collect and collapse process