AN INTERFEROMETRIC SPECTRAL-LINE SURVEY OF IRC+10216 IN THE 345 GHz BAND

Understanding the formation of complex molecules and dust grains in space is a major problem in modern astrophysics. Stars on the asymptotic giant branch (AGB) efficiently produce C, N, O, and s-process elements. Mass loss leads to the formation of expanding circumstellar envelopes (CSEs) whose molecules and dust grains are major sources of replenishment of the interstellar medium (Herwig 2005; Busso et al. 1999). The radius of the C/O core, after the exhaustion of core He burning, is R_c ≈ 10⁹ cm, and the temperature is ∼10⁸ K. He burning continues in a shell around the core and the size of the stellar photosphere expands to R_* ≈ 10¹³ cm. During the phase of intermittent burning of H and He (in the so-called thermal pulse), products of nuclear burning are dredged up by convection and brought to the stellar surface. The zone within a few stellar radii is dynamically important for mass loss; in this zone molecules produced in the stellar atmosphere are moved to a region away from the star that is cool enough (∼1000 K) so that dust grains can condense. Radiation pressure on the grains drives the CSE's expansion. The outer circumstellar region is comparatively less dense by several orders of magnitude and richer in molecular gas, steadily expanding outward with velocities ∼10 km s⁻¹. The chemistry in the outermost part of the circumstellar shell is driven by the interstellar UV radiation (Glassgold 1996). Due to the clumpy nature of the shell, photochemistry may be important in the inner regions as well (Decin et al. 2010).

IRC+10216 (CW Leo) is a well-known AGB carbon star ([C] > [O] and the presence of s-process elements) with a high mass-loss rate (several ×10⁻⁵ M_☉ yr⁻¹) at a distance of 150 pc (e.g., Young et al. 1993; Crosas & Menten 1997). Owing to its closeness to the Sun, it has been possible to study the physical and chemical processes in its large CSE in great detail (e.g., Olofsson 1999). There are nearly 60 molecules observed in the circumstellar shell of IRC+10216 as a result of previous single-dish line surveys (Kawaguchi et al. 1995; Cernicharo et al. 2000; Avery et al. 1992; Groesbeck et al. 1994; He et al. 2008; Ziurys et al. 2002; Cernicharo et al. 2010; Tenenbaum et al. 2010).

Mapping the spatial distribution of molecules in the CSE of IRC+10216 is important for several reasons: (1) molecular (and isotopic) abundances can be accurately determined, since the excitation temperature can be inferred from the spatial location of the molecules in the envelope; (2) such data can be readily and quantitatively compared with chemical models predicting abundances as a function of radial distance from the star; (3) parent molecules can be distinguished from product molecules given their distribution in the envelopes; (4) molecules important for the creation of dust can be identified (e.g., the distributions of SiO, SiS, SiN, and silicon carbides appear to be centrally concentrated near the region of dust formation and probably are dominant constituents of grains forming in this region, their abundance decreases with distance from the star); and (5) multiple transitions of the same molecule allow mapping of physical conditions (e.g., temperature) in the envelope.

Interferometric maps of NaCN, SiO, SiS, CS, HC₅N, SiCC, NaCl, MgNC, CN, HNC, C₂H, C₃H, and C₄H have been presented by Guélin et al. (1996) and Dayal & Bieging (1995). Except for the SiO J = 5–4 line at 217 GHz (Schöier et al. 2006) and the CS J = 14–13 line at 685 GHz (Young et al. 2004) mapped with the Submillimeter Array⁷ (SMA), all the other maps were obtained using the IRAM Plateau de Bure Interferometer (PdBI) or the Berkeley–Illinois–Maryland Array (BIMA) at around 100 GHz.

We selected the 345 GHz band for our survey primarily because very little data exist in this frequency range, and it contains transitions from many astrochemically important molecules, including various salts, the cyanopolyyne HC₃N, and cyclic molecules such as C₃H₂. Two line surveys in the 345 GHz band have been published by Avery et al. (1992) in the frequency range of 339.6–364.6 GHz with a sensitivity of 0.3 K rms made with the 15 m diameter James Clerk-Maxwell Telescope (JCMT) and by Groesbeck et al. (1994), in the frequency range of 330.2–358.1 GHz, with the Caltech Submillimeter Observatory (CSO) 10.4 m telescope with an rms noise level of 65 mK (4.6 Jy). Our frequency range goes well beyond these surveys, including the range of 300–330 GHz, which is almost unexplored.

The SMA observations of IRC+10216 were done in two periods, each about one week long. The first was 2007 January and February and the second 2009 February. All observations were made with the array in the subcompact configuration, with baselines from 9.5 m to 69.1 m. The typical synthesized beam size was 3'' × 2''. Table 1 summarizes the observational parameters for all observations. The duration of each track was from 7 to 9 hr. The phase center was at α(2000) = 09^h47^m5738, δ(2000) = +13°16'4370 for all observations. All tracks in the first phase of observation (in 2007) were carried out in mosaiced mode, with five pointings with offsets in right ascension and declination of (0, ''0'') and (±12, '' ± 12''). The 2009 epoch observations were done with a single pointing toward IRC+10216. The u–v coverage and synthesized beam in one of the single-pointing tracks are shown in Figure 1. Titan and the quasars 0851+202 and 1055+018 were observed every 20 minutes for gain calibration. The spectral bandpass was calibrated by observations of Mars and Jupiter. Absolute flux calibration was determined by observations of Titan and Ganymede.

Zoom In Zoom Out Reset image size

Table 1. Summary of Observations

Date	LSB	USB	Synthesized	τ225 GHza	Tsysb
	Coverage	Coverage	Beam		(SSB, K)
2007 Jan 25	338.9–340.9	348.9–350.9	30 × 20, P.A. = −4°	0.08	180–270
2007 Feb 5	338.4–340.1	348.7–350.0	30 × 20, P.A. = −4°	0.08	180–270
2007 Feb 7	298.1–300.1	308.1–310.1	32 × 24, P.A. = −4°	0.08	180–270
2007 Feb 8	300.1–302.1	310.1–312.1	33 × 24, P.A. = −6°	0.09	160–230
2007 Feb 9	336.4–338.4	346.5–348.4	30 × 24, P.A. = −8°	0.08	180–350
2007 Feb 10	334.5–336.4	344.5–346.4	30 × 24, P.A. = −8°	0.08	180–350
2007 Feb 12	332.5–334.5	342.5–344.5	30 × 24, P.A. = −8°	0.08	180–350
2009 Jan 22	302.0–306.0	313.9–317.9	33 × 23, P.A. = −24°	0.05	100–200
2009 Jan 23	312.0–316.0	323.9–327.9	27 × 21, P.A. = −13°	0.05	200–500
2009 Jan 26	315.9–319.9	327.9–331.9	27 × 21, P.A. = −13°	0.08	200–400
2009 Jan 30	338.9–342.9	350.8–354.8	30 × 22, P.A. = −2°	0.06	150–300
2009 Jan 31	319.9–323.9	331.9–335.8	30 × 24, P.A. = −6°	0.08	280–700
2009 Feb 2	293.9–297.9	305.9–309.9	59 × 30, P.A. = 38°	0.06	100–240

Notes. ^aZenith optical depth measured at 225 GHz with the CSO's radiometer. ^bT_sys range is over all antennas and over the range of elevations covered during the observations of IRC+10216.

Download table as:  ASCIITypeset image

The visibility data were calibrated with the Miriad package (Sault et al. 1995). For the 2007 data, the mosaiced images were de-convolved using the Miriad task mossdi; the resulting synthesized beams are summarized in Table 1. The single-pointing 2009 data were calibrated with the MIR-IDL package⁸ and imaged in Miriad using the standard tasks invert, clean, and restor. Maps of continuum emission show the peak to have a position offset of (Δα, Δδ) ≈ (07, 02) from the phase center position. The absolute position measurements for the continuum emission are estimated to be accurate to ∼01. Taking into account the proper motion of IRC+10216 of ($\dot{\alpha },\dot{\delta })\approx (26,4)$ mas yr⁻¹ determined by Menten et al. (2006), our position is consistent with theirs. The continuum emission was unresolved at the highest angular resolution of ∼08. The integrated continuum flux density was ∼650 mJy at 300 GHz and ∼1 Jy at 350 GHz, with an uncertainty of about 15% in the absolute flux calibration. The frequency resolution was 0.812 MHz per channel.

We detected a total of 442 lines. Of these, 297 could be assigned to known molecular transitions. Table 2 summarizes all detections, with fitted parameters and molecular assignments.