Widespread CCH and c-C₃H₂ in the Helix Nebula: Unraveling the Chemical History of Hydrocarbons

The presence of polyatomic molecules in diffuse clouds remains unexplained. Given their low densities of 10–100 cm⁻³, models have predicted that such clouds could only foster and sustain diatomic species (e.g., van Dishoeck & Black 1986). Recently, however, a variety of polyatomic molecules has been detected in these objects, including HCN, HNC, CCH, HCO⁺, H₂S, and c-C₃H₂ (e.g., Liszt et al. 2006; Gerin et al. 2011). Even more remarkable, there is strong evidence for the presence of both C₆₀ and ${{{\rm{C}}}_{60}}^{+}$ in these low-density sources, as carriers of a few of the diffuse interstellar bands (DIBs; Campbell et al. 2015; Berné et al. 2017). The in situ formation of such complex molecules in diffuse clouds is extremely difficult (Snow & McCall 2006), given their low densities, suggesting an alternative, external source.

One possible origin of the complex molecular material in diffuse clouds are planetary nebulae (PNe). There is a natural connection between these objects, as PNe supply most of the matter to the diffuse interstellar medium (ISM; Dorschner & Henning 1995). Furthermore, some PNe are known to contain C₆₀ (e.g., García-Hernández et al. 2012), as well as a substantial number of other polyatomic molecules, including HCN, HNC, CCH, HCO⁺, H₂CO, N₂H⁺, and SO₂ (e.g., Zhang et al. 2008; Zack & Ziurys 2013; Edwards & Ziurys 2014; Schmidt & Ziurys 2016, 2017a, 2017b). The PN phase follows the asymptotic giant branch (AGB) and is the evolutionary path for most (low- and intermediate-mass) stars. PNe are therefore quite common. Extreme ultraviolet (UV) radiation on the PN track, however, is emitted from the central star as it evolves into a white dwarf—∼10⁵ times that of the general ISM. It is predicted that such radiation in time would destroy all molecular material by the end of the PN phase (e.g., Redman et al. 2003).

Recent observations have shown otherwise. Several older nebulae contain polyatomic molecules with considerable abundances (e.g., Zeigler et al. 2013; Schmidt & Ziurys 2017a). In fact, such abundances appear to remain fairly constant with the age of the nebulae across their life span of ∼10,000 years (Schmidt & Ziurys 2016, 2017a, 2017b). The molecular material could be "fossil" species of the former AGB shell and/or be created in the nebulae, possibly by gas-phase reactions or from fragmentation of larger molecules such as C₆₀ or even hydrogenated amorphous carbon (HACs; e.g., Duley et al. 2015). The molecules in PNe also appear to be present in dense clumps, mixed with dust, which are self-shielding—consistent with slow dispersion of molecular material from these nebulae into the diffuse ISM (e.g., Zack & Ziurys 2013; Edwards et al. 2014).

The very evolved Helix Nebula (NGC 7293), with an estimated age of ∼11,000 years (Meaburn et al. 2008), offers an excellent opportunity to study the fate of molecules in the final stages of PNe. Because of its close proximity to Earth (distance of ∼219 pc; Harris et al. 2007), it has a large angular extent of over 1000''. Its unusual geometry has been interpreted as a helix, an equatorial ring with arcs and filaments, and two disks tilted with respect to each other (e.g., O'Dell et al. 2004). Recent mapping in HCO⁺ has shown that Helix is actually a bipolar nebula oriented nearly along the line of sight (Zeigler et al. 2013). At one position near the western edge of the nebula, (−372'', 0'') from the central star, various molecules have been detected, including HCN, CN, HNC, HCO⁺, CCH, c-C₃H₂, and H₂CO (Bachiller et al. 1997a; Tenenbaum et al. 2009). Further observations by Zack & Ziurys (2013) and Schmidt & Ziurys (2017a) revealed that HCO⁺, H₂CO, HCN, and HNC are present at eight other representative positions across the nebula, demonstrating the widespread distribution of polyatomic molecules. Zeigler et al. (2013) additionally showed that HCO⁺ emission traces the entire optical, R-band image, as do CO and H₂ (e.g., Young et al. 1999). Furthermore, the molecular abundances of these molecules are comparable to those in much younger nebulae, such as NGC 7027 or NGC 6537 (Zhang et al. 2008; Edwards & Ziurys 2013; Schmidt & Ziurys 2016, 2017a, 2017b). These measurements indicate that significant amounts of molecular material are likely being ejected into the diffuse ISM by the Helix Nebula.

Although polycyclic aromatic hydrocarbons have yet to be detected in the Helix, Young et al. (1997) suggest that the nebula is carbon-rich. Therefore, the presence and distribution of simple hydrocarbons such as CCH and c-C₃H₂ are of interest, as they may tracer larger C-rich species such as fullerenes and HACs. Both CCH and c-C₃H₂ had been previously identified at one position, (−372'', 0''); however, this location may not be representative of the Helix as a whole. More extensive observations are essential to formulate any general conclusions on general hydrocarbon chemistry and larger carbonaceous structures. We have consequently carried out observations of CCH and c-C₃H₂ at the eight additional positions in the Helix previously sampled in H₂CO, HCN, and HNC. We have detected these molecules at all positions, as well as CN, establishing their ubiquity throughout this aging nebula. Outside of the Helix, c-C₃H₂ has been identified in only one other PN: NGC 7027. In this Letter, we present our observations and analysis, and discuss the implications of this work for diffuse cloud chemistry.

Measurements of the N = 1 → 0 transitions of CN and CCH at 113.491 GHz and 87.317 GHz, respectively, and the J = 2_1,2 → 1_0,1 transition of c-C₃H₂ at 85.338 GHz were conducted using the 12 m Atacama Large Millimeter/submillimeter Array (ALMA) prototype antenna of the Arizona Radio Observatory (ARO) on Kitt Peak, Arizona between 2014 December and 2018 February. The observations were performed using the dual-polarization, 3 mm receiver employing sideband-separating mixers, with a typical rejection, intrinsic to the mixer architecture, of 16 dB. The temperature scale, ${T}_{A}^{* }$, was determined by the chopper wheel method; the main beam brightness temperature (T_R) is then ${T}_{R}={T}_{A}^{* }/{\eta }_{b}$, where ${\eta }_{b}$ is the main beam efficiency. Two of three filter banks with 250 kHz, 500 kHz, and 1 MHz resolutions were employed as backends, depending on the molecule, operated in parallel mode for the two receiver polarizations. The beam size ranged from θ_b ∼ 55''–74''.

The observed positions are those studied by Zack & Ziurys (2013) and Schmidt & Ziurys (2017a), which represent the nebula as a whole, and are given in arcseconds from the central star (J2000.0 α = 22^h29^m386, δ = −20°50'18''). All observations were conducted in position-switching mode with an azimuth offset of +30'. Local oscillator shifts were conducted to test image contamination, and pointing was checked regularly using continuum sources.

CN, CCH, and c-C₃H₂ were detected at all eight positions in the Helix, as shown in Figures 1 and 2. Figure 1 displays the CCH (upper) and c-C₃H₂ (lower) spectra observed at each position. The hyperfine structure is indicated below the data in black for the (−15'', 270'') position, which has only a single velocity component. Both the F = 2 → 1 and F = 1 → 0 hyperfine components of the N = 1 → 0, J = 3/2 → 1/2 transition of CCH were observed. Some positions exhibit more than one velocity component, as seen in H₂CO, HCN, and HCO⁺ (Zack & Ziurys 2013; Schmidt & Ziurys 2017a).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 2 presents the N = 1 → 0, J = 3/2 → 1/2 spectra of CN for each position. The hyperfine structure (F = 3/2 → 1/2, F = 5/2 → 3/2, F = 1/2 → 1/2, F = 3/2 → 3/2, and F = 1/2 → 3/2 components) for this radical is displayed under the data for the (−240'', −100'') position, which also has a single velocity feature. The CN line profiles for this source are particularly complex due to its extensive hyperfine structure. CN emission is typically more intense than that of CCH or c-C₃H₂.

Spectral modeling was carried out to deconvolve the velocity structure from the hyperfine components in CCH and CN, assuming the optically thin, LTE limit, as is consistent with the observed line profiles. Relative intensities and frequencies of the hyperfine lines were taken from Ziurys et al. (1982) and Skatrud et al. (1983). Velocity features at each position were well established from our previous works (Zack & Ziurys 2013). Line intensities, LSR velocities, and line widths were adjusted to match the observed features. Examples of the modeled spectra are shown in Figure 3. Here, the observed CN (upper) and CCH (lower) spectra for the (130'', −180'') position are shown in blue (or black), while the fits are overlaid in green (or gray scale). For c-C₃H₂, only the velocity structure had to be considered.

Zoom In Zoom Out Reset image size

The resultant line parameters ${T}_{A}^{* }$ and V_LSR and line widths ΔV_1/2 from the analysis for all molecules are summarized in Table 1. Only the strongest hyperfine lines for CCH and CN are presented in the table. Line intensities are ∼0.002–0.120 K. The LSR velocities and line widths are comparable to those measured for CO, HCO⁺, and H₂CO by Zack & Ziurys (2013).

Table 1.  Observations of CN, CCH, and c-C₃H₂ toward the Helix Nebula^a

Offset	Molecule	Transition	${T}_{A}^{* }$	VLSR	ΔV1/2
(Δα, Δδ)			(mK)	(km s−1)	(km s−1)
(130, −180)	CN	F = 5/2 → 3/2b	62 ± 5	−47.3 ± 1.3	3.2 ± 1.3
			75 ± 5	−9.7 ± 1.3	3.4 ± 1.3
	CCH	F = 2 → 1c	10 ± 3	−47.0 ± 1.7	5.9 ± 1.7
			15 ± 3	−11.0 ± 1.7	5.4 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1d	7.2 ± 1	−48.0 ± 1.7	2.5 ± 1.7
			8 ± 1	−9.0 ± 1.7	6.0 ± 3.4
(390, −30)	CN	F = 5/2 → 3/2b	66 ± 7	−33.5 ± 1.3	6.0 ± 1.3
	CCH	F = 2 → 1c	9 ± 1	−34.0 ± 1.7	9.0 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1	4 ± 1	−34.3 ± 1.7	6.5 ± 1.7
(−240, −100)	CN	F = 5/2 → 3/2b	48 ± 6	−39.2 ± 1.3	5.7 ± 1.3
	CCH	F = 2 → 1c	5.3 ± 2	−39.0 ± 1.7	7.0 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1d	2.7 ± 1	−39.0 ± 0.9	2.5 ± 1.7
(−15, 270)	CN	F = 5/2 → 3/2b	72 ± 15	−35.7 ± 1.3	3.5 ± 1.3
			100 ± 15	−28.7 ± 1.3	3.7 ± 1.3
	CCH	F = 2 → 1c	9 ± 3	−30.5 ± 1.7	7.0 ± 1.7
			∼3	∼−35	∼6.0
	c-C3H2	JKa,Kc = 21,2 → 10,1	3 ± 1e	−34.6 ± 1.7	3.2 ± 1.7
			6 ± 1e	−29.9 ± 0.8	3.2 ± 1.7
(−372, 0)	CN	F = 5/2 → 3/2b	120 ± 15	−15.3 ± 1.3	4.4 ± 1.3
(125, 185)	CN	F = 5/2 → 3/2b	11 ± 6e	−24.5 ± 1.3	2.5 ± 1.3
			23 ± 6e	−21.3 ± 1.3	3.0 ± 1.3
			41 ± 6	−12.7 ± 1.3	2.9 ± 1.3
	CCH	F = 2 → 1c,f	1.5 ± 0.5e	−21.0 ± 1.7	5.9 ± 1.7
			2.8 ± 0.5e	−13.0 ± 1.7	4.0 ± 1.7
			2.1 ± 0.5	−1.0 ± 1.7	4.7 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1	3.5 ± 0.5	−12.3 ± 1.7	2.8 ± 1.7
(−435, 75)	CN	F = 5/2 → 3/2b	30 ± 5e	−30.3 ± 1.3	2.5 ± 1.3
			60 ± 3e	−21.4 ± 1.3	6.2 ± 1.3
			36 ± 4e	−13.6 ± 1.3	6.2 ± 3.9
	CCH	F = 2 → 1c	11 ± 3e	−21.0 ± 3.4	9.7 ± 1.7
			4 ± 1e	−11.5 ± 6.8	9.4 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1	1.5 ± 0.5	−29.0 ± 1.7	3.3 ± 1.7
			1.5 ± 0.5	−21.0 ± 1.7	4.8 ± 1.7
			1.5 ± 0.5	−13.0 ± 1.7	6.0 ± 1.7
(−300, −200)	CN	F = 5/2 → 3/2b	11 ± 2	−27.6 ± 1.3	2.5 ± 1.3
			19 ± 2e	−20.3 ± 1.3	2.7 ± 1.3
			31 ± 2e	−17.1 ± 1.3	2.9 ± 1.3
	CCH	F = 2 → 1c,d	6 ± 2e	−24.5 ± 1.7	4.5 ± 1.7
			8 ± 2e	−16.0 ± 1.7	4.7 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1	4 ± 1	−17.8 ± 1.7	3.3 ± 1.7
(−120, 240)	CN	F = 5/2 → 3/2b	64 ± 5	−34.2 ± 1.3	5.6 ± 1.3
			38 ± 5	3.1 ± 1.3	4.0 ± 1.3
	CCH	F = 2 → 1c	10 ± 1	−34.0 ± 1.7	5.0 ± 1.7
			8.4 ± 1	3.0 ± 1.7	3.0 ± 1.7
	c-C3H2	JKa,Kc = 21,2 → 10,1	3 ± 1	−32.5 ± 3.4	3.8 ± 1.7
			5 ± 1	2.5 ± 1.7	3.8 ± 1.7

Notes.

^aMeasured with 500 kHz resolution unless noted; only the strongest hyperfine components are listed. ^bBlend of F = 5/2 → 3/2 and 3/2 → 1/2 hyperfine components (see the text). ^cF = 1 → 0 hyperfine component also measured (see the text). ^dMeasured with 250 kHz resolution. ^ePartially blended velocity components. ^fMeasured with 1 MHz resolution.

Download table as:  ASCIITypeset image

CN, CCH, and c-C₃H₂ column densities were estimated using the non-LTE radiative transfer code RADEX (van der Tak et al. 2007). Collisional rates for both CCH and c-C₃H₂ are given in the Leiden Atomic and Molecular Database (LAMDA; Schöier et al. 2005), as well as those for the hyperfine levels in CN with para-H₂. Collisional rates for ortho-H₂ were determined from the fine structure rates of Kalugina et al. (2013) using the infinite-order-sudden (IOS) approximation. Inelastic hyperfine structure rate coefficients k_{N,J,F→N'J'F'} for ²Σ molecules may be obtained from the N = 0, J = 1/2 fine structure coefficients ${k}_{0,\tfrac{1}{2}\to L,L+\tfrac{1}{2}}^{\mathrm{IOS}}(T)$ using the following equation (Alexander 1982):

$$\begin{eqnarray}{k}_{{NJF}\to {N}^{{\prime} }{J}^{{\prime} }{F}^{{\prime} }}^{\mathrm{IOS}}(T) & = & (2J+1)(2{J}^{{\prime} }+1)(2{F}^{{\prime} }+1){\displaystyle \sum }_{L}\tfrac{2L+1}{L+1}\\ & & \times \,{\left(\begin{array}{ccc}{J}^{{\prime} } & L & J\\ -1/2 & 0 & 1/2\end{array}\right)}^{2}{\left\{\begin{array}{ccc}J & {J}^{{\prime} } & L\\ {F}^{{\prime} } & F & I\end{array}\right\}}^{2}\\ & & \times \,\displaystyle \frac{1}{2}[1-\varepsilon {(-1)}^{J+{J}^{{\prime} }+L}]{k}_{0,\tfrac{1}{2}\to L,\ L+\tfrac{1}{2}}^{\mathrm{IOS}}(T).\end{eqnarray} \tag{ 1 }$$

The collisional rates were then scaled to account for the IOS approximation, which assumes that rotational level spacing is insignificantly small (Neufeld & Green 1994). Further details can be found in Faure & Lique (2012).

RADEX has three input parameters: column density, H₂ density, and gas kinetic temperature T_K, which are used to model the main beam brightness temperatures T_R. As only one rotational transition was observed for each molecule, T_K and the H₂ densities for each position and velocity component were set to those found by Zack & Ziurys (2013), based on CO, HCO⁺, and H₂CO. These values are T_K ∼ 15–40 K and n(H₂) ∼ (0.1–5) × 10⁵ cm⁻³. CN, CCH, and c-C₃H₂ possess dipole moments of 1.45 D (Thomson & Dalby 1968), 0.77 D (Woon 1995), and 3.27 D (Lovas et al. 1992), respectively—a similar range to those of HCO⁺, H₂CO, and CO (3.9 D, 2.33 D, and 0.1 D). An H₂ and c-C₃H₂ ortho-para ratio of 3:1 was assumed. The fractional abundances were derived using the average H₂ column density determined by Zack & Ziurys (2013).

Results from this analysis are given in Table 2, which presents column densities N_tot and fractional abundances f for CN, CCH, and c-C₃H₂ in the Helix. Values determined fall in the ranges N_tot(CN)  ∼  (0.7–7.4)  ×  10¹² cm⁻², N_tot(CCH) ∼ (0.8–6.6) × 10¹² cm⁻², and N_tot(c-C₃H₂) ∼ (0.3–5.9) × 10¹¹ cm⁻². Fractional abundances were estimated to be f(CN) ∼ (0.9–9.9) × 10⁻⁷, f(CCH) ∼ (1.1–8.8) × 10⁻⁷, and f(c-C₃H₂) ∼ (0.4–7.9) × 10⁻⁸. These values are similar to those determined for the position (−372'', 0'') by Tenenbaum et al. (2009), who found f(CCH) ∼ 10⁻⁶ and f(c-C₃H₂) ∼ 3.0 × 10⁻⁸.

4.1. Widespread Distribution of Basic Hydrocarbons in the Helix

4.2. Molecular Synthesis of CCH and c-C₃H₂ in the Helix

4.3. Populating Diffuse Clouds

This research was supported by NSF grant AST-1515568.