NEW DETECTIONS OF HNC IN PLANETARY NEBULAE: EVOLUTION OF THE [HCN]/[HNC] RATIO

Hb 5: The $J=1\to 0$ transition of HNC was detected with a central velocity of −4.6 km s⁻¹ in this PN, close to the systemic velocity of the star, which is −7 km s⁻¹ (López et al. 2012). As for HCN and HCO⁺, no obvious Galactic contamination was present in the spectrum. The line profile appears asymmetric, and the line width of ∼33 km s⁻¹ is comparable to that of the $J=3\to 2$ transition of HCN (∼40 km s⁻¹). López et al. (2012) calculated a kinematic age of 1500 years for this source, based on the radial expansion velocity.

K3-17: HNC was detected in the $J=3\to 2$ transition of this source, with V_LSR ∼ 32 km s⁻¹ and ΔV_1/2 ∼ 50 km s⁻¹, consistent with the $J=3\to 2$ lines of both HCN and HCO⁺. The nebula is small (8'') and multi-polar with several secondary loops. Multiple narrow velocity components at ∼20, 30, 39, and 57 km s⁻¹ appear to be visible in the spectra of HNC, HCN, and HCO⁺, indicative of this complex geometry. An age of ∼1000 years was derived for K3-17 from the expansion velocity inferred from CO (Schmidt & Ziurys 2016).

K3-45: Only the $J=3\to 2$ transition of HNC was detected in this small (68: Acker et al. 1992) bipolar object. The observed V_LSR of ∼34 km s⁻¹ and ΔV_1/2 of ∼34 km s⁻¹ are comparable to those seen in HCN. The HNC line profile is clearly double-peaked, with maxima at ∼18 km s⁻¹ and ∼45 km s⁻¹, as also seen in the HCN $J=3\to 2$ transition. Schmidt & Ziurys (2016) calculated an age estimate of ∼800 years for this nebula.

K3-58: Both the $J=1\to 0$ and 3 → 2 transitions of HNC were observed in this small (17'') bipolar nebula. The LSR velocities of ∼20 km s⁻¹ and line widths of ∼30 km s⁻¹ are consistent with the HCN and HCO⁺ spectra. An age of ∼11,000 years was estimated for this source by Dobrinčić et al. (2008) from their spatio-kinematical study.

M1-7: The $J=1\to 0$ and 3 → 2 transitions of HNC were both detected in this small, older, compact (∼11'') bipolar PN. The line widths and LSR velocities are consistent with what is seen for HCN and HCO⁺ within the uncertainties. There may be additional velocity structure in the spectra near −29 km s⁻¹ and −7 km s⁻¹. Schmidt & Ziurys (2016) derived an age of ∼6000 years for M1-7.

M3-28: HNC was detected in both transitions in this small (∼17'') quadrupolar nebula. As was the case for both HCN and HCO⁺, the $J=1\to 0$ transition of HNC was contaminated by Galactic emission toward redshifted velocities. Thus, the V_LSR and ΔV_1/2 of the $J=1\to 0$ transition (∼33 km s⁻¹ and ∼35 km s⁻¹ respectively) are estimates. The line parameters of the $J=3\to 2$ transition match reasonably well with those observed for the HCN and HCO⁺ transitions (V_LSR ∼ 33 km s⁻¹ and ΔV_1/2 ∼ 36–42 km s⁻¹). An adopted age of ∼7700 years was obtained by Schmidt & Ziurys (2016).

M3-55: While both transitions were detected in this small (∼95) bipolar PN, the $J=1\to 0$ spectrum was partially obscured by Galactic contamination, making estimates of the line parameters less reliable. The V_LSR and ΔV_1/2 of the $J=3\to 2$ transition, ∼30 km s⁻¹ and ∼23 km s⁻¹ respectively, are comparable to those seen for the $J=3\to 2$ transition of both HCN and HCO⁺. An age of ∼5000 years is given by Dobrinčić et al. (2008).

M4-14: In this 15'' quadrupolar nebula, both transitions of HNC were detected. The V_LSR of ∼49–53 km s⁻¹ and ΔV_1/2 of ∼26–27 km s⁻¹ are consistent with those seen in the $J=3\to 2$ spectra of HCN and HCO⁺ , although the HNC lines are slightly narrower. From spatio-kinematic modeling, Dobrinčić et al. (2008) assign an age of ∼5500 years to this source.

NGC 2440: A detection of the $J=3\to 2$ transition was made toward this source (age ∼3000 years: Górny et al. 1997). The line appears to have a peak temperature of ∼7 mK and a width of ∼50 km s⁻¹, and is centered at ∼41 km s⁻¹. This is consistent with what was seen in HCN and HCO⁺. Several sharp, narrow features may be seen near ∼27, 41, and 57 km s⁻¹, which are also present in HCN and HCO⁺.

NGC 6772 a/b: The $J=1\to 0$ transition of HNC was detected in this large (756: Tylenda et al. 2003) nebula at the central position (19^h 14^m 373: −2°42' 198 J2000.0) provided by Milne (1973), hereafter referred to as "NGC 6772a." The observed position differs from that used by Schmidt & Ziurys (2016) for HCN and HCO⁺ (19^h 14^m 364 in R.A., −2°42' 25'' J2000.0 from Kerber et al. (2003), hereafter known as "NGC 6772b"). This position was also observed, but no positive detections were made. The line intensity for NGC 6772a is ${T}_{A}^{* }\sim 3$ mK; the line width of ∼31 km s⁻¹ and LSR velocity of ∼24 km s⁻¹ are comparable to what is seen in HCN and CO. Ali et al. (2012) calculated a dynamical age of ∼11,000 years.

K4-47: This young bipolar source contains two diametrically opposed bullets or jets of gas, which have been observed in low-ionization atomic lines. The $J=1\to 0$ and 3 → 2 HNC lines were both detected in K4-47, with the V_LSR of ∼−28 km s⁻¹ closely matching that seen in HCN and HCO⁺, although the line widths may be slightly narrower. As in the case of other molecules (e.g., Edwards et al. 2014), the $J=3\to 2$ spectrum of HNC appears to contain both a central and a plateau component, with line widths near 22 km s⁻¹ and 37 km s⁻¹, respectively. As described in Schmidt & Ziurys (2016), the plateau feature may be the result of an outflow associated with bullets of gas ejected from the central star. An age of ∼900 years was estimated by Corradi et al. (2000).

(130'', −180''): Two distinct velocity components are clearly visible in both the HNC and HCN $J=1\to 0$ spectra, one at −10 km s⁻¹ and the other at −47 km s⁻¹, as also seen in CO, HCO⁺, and H₂CO. The HNC and HCN line widths of 2.9–3.5 km s⁻¹ are consistent with those of these other molecules. The HCN line profiles appear noticeably wider because of the nitrogen hyperfine structure.

(390'', −30''): One velocity component can be seen in both the HCN and the HNC spectra, with a central velocity of about −33 km s⁻¹. The line widths for both molecules are ∼6 km s⁻¹, thus blending the hyperfine components in HCN.

(−240'', −100''): A single velocity component, centered at ∼−38 km s⁻¹, is visible in both the HNC and HCN spectra. This feature is particularly narrow, with a ΔV_1/2 of ∼2–3 km s⁻¹.

(−15'', 270''): Two closely spaced velocity components can clearly be distinguished in the HNC spectrum, one centered at −35 km s⁻¹ and one at −29 km s⁻¹. The widths of each component are ∼3 km s⁻¹. These velocity components are also present in the HCN spectrum, but are interwoven with the hyperfine structure, creating a line profile that at first glance appears to have non-LTE intensities. This profile is more accurately explained by the two velocity components, as shown in Figure 3.

(−120'', 240''): Two widely spaced velocity components are present in both the HCN and HNC spectra, at ∼−34 and 3 km s⁻¹, respectively. The ΔV_1/2 is ∼5–6 km s⁻¹ for the −33 km s⁻¹ component, and around 2 km s⁻¹ for the 3 km s⁻¹ component. Hyperfine splitting in HCN is discernable in the 3 km s⁻¹ component but the lines are completely blended in the −34 km s⁻¹ feature.

(125'', 185''): Three velocity components are apparent in the HNC spectrum, at −12.5, −20, and −25 km s⁻¹, respectively, with line widths ∼3 km s⁻¹. Due to their relative proximity, the −20 and −25 km s⁻¹ features are blended. These components also appear to be present in the HCN spectrum, but the hyperfine structure complicates the line profile and the individual components are difficult to distinguish. The widths in the HCN spectrum appear to be around 3 km s⁻¹ as well.

(−435'', 75''): Three velocity components are present in both the HCN and HNC spectra, at ∼−14, −21, and −29 km s⁻¹, respectively. These are partially blended together, creating one broad feature. The line widths range around 4–5 km s⁻¹ for each component. The complex appearance of the HCN line profile is readily accounted for by these three components superimposed on the hyperfine structure; see Figure 3.

(−300'', −200''): A component near 16–17 km s⁻¹ is evident in both the HNC and HCN spectra, consistent with what is seen in CO, HCO⁺, and H₂CO by Zack & Ziurys (2013). However, to adequately model the HCN spectrum, an additional velocity component at ∼21 km s⁻¹ was included. There is a tentative detection of this component in the HNC spectrum as well. In addition, there may be a third component in HCN near −27 km s⁻¹, as also observed in CO. The line widths are typically ∼3 km s⁻¹ for each component, as best can be established, though the ∼−17 km s⁻¹ component in HNC appears to have a broader line width of ∼7 km s⁻¹.

The detection of HNC in all observed targets, which represent a somewhat random sampling of the PN population, is striking, particularly considering that, just as in the case with HCN and HCO⁺, the abundance is not negligible. The average HNC fractional abundance is ∼4.5 × 10⁻⁸, including both the survey sources (assuming clumping) and the Helix positions. Furthermore, the HNC and HCN abundances across the Helix vary only by a factor of about 23 and 15, respectively. For the (−372'', 0'') position in the Helix, the data from Bachiller et al. (1997) suggest f(HNC/H₂) ∼ 3.2 × 10⁻⁸ and f(HCN/H₂) ∼ 5.4 × 10⁻⁸—in line with those seen at our positions. Their measurements also suggest HNC abundances with respect to H₂ of 3.5 × 10⁻⁸, 1.1 × 10⁻⁸, and 2.2 × 10⁻⁸ for NGC 6781, M4-9, and NGC 6720, scaling by their ¹²CO/¹³CO ratios. In NGC 6072 and IC 4406, HNC abundances of 1.9 × 10⁻⁸ and 1.4 × 10⁻⁸ were determined from the data of Cox et al. (1992). Previous values for the HNC abundances are comparable to those established here.

Of the Schmidt & Ziurys (2016) survey sources, the abundance of HNC appears to be highest in K4-47 and lowest in M3-28 (assuming clumping). Similarly, K4-47 possesses a high abundance of HCN of ∼7.6 × 10⁻⁷, while that in M3-28 is considerably lower at 9.8 × 10⁻⁹. In the Helix, the V_LSR = −17 km s⁻¹ component at position (−300'', −200'') has the highest HNC and HCN abundances, with f(HNC) ∼ 3.2 × 10⁻⁷ and f(HCN) ∼ 2.2 × 10⁻⁷. Thus, the abundances of HCN and HNC generally appear to correlate.

The common precursor to both HCN and HNC in PNe is thought to be HCNH⁺ (e.g., Kimura et al. 2012), as mentioned previously. This species is created by different routes, triggered by ions generated by photons or by cosmic rays. The correlation of HCN with HNC in PNe thus supports this hypothesis. HCNH⁺ has yet to be detected in PNe.

Several chemical models of PNe predict abundances of HNC. For example, Redman et al. (2003) estimate the fractional abundance of HNC to be ∼1.4 × 10⁻⁸ at 2550 years, decreasing significantly to 2.7 × 10⁻¹² at 6300 years and further dropping to 2.2 × 10⁻¹² at 10,050 years. While the abundance at 2550 years is on par with the average observed values, the predictions of Redman et al. at 6300 and 10,050 years fall far below any of the fractional abundances determined for the PNe studied here. Kimura et al. (2012) predict HNC abundances as high as 10⁻⁹, in somewhat better agreement, though they did not study evolution with age.

Our data also indicate that there is no strong variation of HNC abundance with nebular age, as illustrated in Figure 4. This figure plots the HNC abundance (cyan circles), assuming clumping, as a function of source age, along with those for HCN and HCO⁺ from Schmidt & Ziurys (in red and blue circles, respectively). Also plotted are abundances in other sources, indicated by stars. Model predictions from Redman et al. are represented as triangles. As shown in the figure, the observed values fall in the range ∼10⁻⁸−10⁻⁷, and do not change significantly from about 1000 to 11,000 years. Although there are uncertainties in the ages of the nebulae, the trend is very clear. The models predict a sharp decrease of 10⁴ by 6000 years for HNC, with a steady abundance of 10⁻¹² thereafter. Indeed, these findings support the idea that clumped material shields the molecules from incoming radiation more effectively than predicted by models, and/or that there is more active chemical production ongoing in PNe than expected, as noted by Schmidt & Ziurys (2016).

Zoom In Zoom Out Reset image size

HNC is a prominent molecule in C-rich AGB envelopes, including IRC+10216, IRAS 15194-5115, IRAS 15082-4808, IRAS 07454-7112, CIT 6, AFGL 3068, and IRC+40540 (Woods et al. 2003). The abundances of HNC in these objects ranged from (0.2–3.3) × 10⁻⁷. In contrast, Smith et al. (2015) found f(HNC/H₂) ≤ (0.5–2.9) × 10⁻⁸ for the envelopes of C-stars II Lup, RAFGL 4211, and AI Vol. HNC has also been identified in a number of proto-planetary nebulae (PPNe). Morris et al. (1987) derived an HNC abundance of 1.8 × 10⁻⁷ for the protoplanetary Calabash Nebula, while in CRL 2688, Zhang et al. (2013) found f(HNC/H₂) ∼ 5.7 × 10⁻⁷—higher than those of the PNe observed here. For CRL 618, the abundance is estimated to be in the range 1.9 × 10⁻⁷–1.9 × 10⁻⁶ (Sopka et al. 1989; Bujarrabal et al. 1994). Comparing these HNC abundances with those measured in PNe, the values on average decrease slightly from the AGB stage through the PN phase.

The [HCN]/[HNC] ratios established here fall within the range 1–8 for all sources. The ratio appears to be highest in K3-17, a young PN, and lowest in M3-55, which is middle-aged. These values are consistent with those found in other PNe. Using the data of Bachiller et al. (1997), NGC 6720, M4-9, and NGC 6781 have ratios ranging from ∼1.4–4.5, while several compact PNe are estimated to have [HCN]/[HNC] ∼ 2.5 (Josselin & Bachiller 2003). The HNC data in the latter case have limited sensitivity. Regardless, it is clear that the ratio is remarkably constant, given the different types of nebulae and their physical characteristics.

There may be a slight correlation of the [HCN]/[HNC] ratio with stellar temperature, which in principle could convert some HNC to HCN. Hb 5 has a ratio of 7 and a high temperature for its central star of ${T}_{* }\geqslant {\rm{150,000}}$ K (Pottasch & Surendiranath 2007). On the other hand, NGC 2440 and M3-28 both have ratios near 4, with respective stellar temperatures of ${T}_{* }\sim {\rm{188,400}}$ K (Heap 1987) and ${T}_{* }\sim {\rm{130,500}}$ K (Preite-Martinez et al. 1989). Temperatures for the other sources are not known. Therefore, it is difficult to determine such a correlation. More data are clearly needed.

Figure 5 shows a plot of [HCN]/[HNC] versus nebular age. There does not appear to be any significant variation in the ratio with age. There are no sharp deviations in the general trend.

Zoom In Zoom Out Reset image size

The [HCN]/[HNC] ratios found in this work are lower than those typically seen in AGB stars. Woods et al. (2003) measured [HCN]/[HNC] ratios of ∼40–300 for the seven carbon stars observed. The value for IRC+10215 is near 300. The ratio apparently drops in the PPN stage. The ratio for CRL2688 is ∼1 (Zhang et al. 2013), while the value in PPN IRAS 16594-4656 is about 3.4 (Woods et al. 2005). Herpin & Cernicharo (2000) found relatively equal column densities for HCN and HNC in the PPN CRL618, implying [HCN]/[HNC] ∼ 1, while Morris et al. (1987) found a ratio of ∼2.2 for the Calabash Nebula.

The rise in HNC with respect to HCN in the PPN phase has been attributed to an increase in ionization. As discussed, HCN is effectively produced by LTE chemistry in the inner envelope of AGB stars, while HNC is not, thereby producing the high [HCN]/[HNC] ratios often observed. In the PPN stage, the increase in ionization drives reactions that lead to the formation of HNC through the effective intermediate, HCNH⁺. The ion HCNH⁺ can form from several different ion-neutral reactions (e.g., Daniel et al. 2012):

$$\begin{eqnarray}&&{\mathrm{HCO}}^{+}+\mathrm{HCN}\to {\mathrm{HCNH}}^{+}+\mathrm{CO}\end{eqnarray} \tag{ 2 }$$

$$\begin{eqnarray}&&{{\rm{C}}}_{2}{{\rm{H}}}_{3}^{+}+\mathrm{HCN}\to {\mathrm{HCNH}}^{+}+{{\rm{C}}}_{2}{{\rm{H}}}_{2}\end{eqnarray} \tag{ 3 }$$

$$\begin{eqnarray}&&{{\rm{H}}}_{3}^{+}+\mathrm{HCN}\to {\mathrm{HCNH}}^{+}+{{\rm{H}}}_{2}.\end{eqnarray} \tag{ 4 }$$

HNC is then produced via dissociative electron recombination, Equation (1), where the branching ratio is estimated to be unity (Talbi 1999). Additionally, increased UV radiation in the PPN and PN phases is effective in creating radical species (Daniel et al. 2012). HNC may also form through the reaction

$$\begin{eqnarray}&&{\rm{C}}+{\mathrm{NH}}_{2}\to \mathrm{HNC}+{\rm{H}}.\end{eqnarray} \tag{ 5 }$$

Over time, these reactions skew the [HCN]/[HNC] ratio toward unity. Observations indicate that the rise of HNC over HCN occurs principally in PPNe. The ratio then appears to "freeze out" into the PN phase. The striking consistency of the [HCN]/[HNC] ratio across the Helix Nebula is evidence of such a freeze-out.

Liszt & Lucas (2001) and Liszt et al. (2006) observed both HCN and HNC (among other molecules) in a set of seven diffuse clouds observed against background extragalactic sources. The HCN abundances ranged from (0.14–0.5) × 10⁻⁸ with respect to H₂, while the HNC abundances fell between (0.03–0.10) × 10⁻⁸. The authors noted both HCN and HNC were much more plentiful in diffuse clouds than predicted by models of quiescent low-density gas-phase chemistry. The reaction rates necessary to form CN-bearing molecules are simply too slow to explain the observed abundances. Based on their measurements, Liszt et al. (2006) estimated an average [HCN]/[HNC] ratio of 5.

The abundances for HNC in diffuse clouds are 1–2 orders of magnitude lower than the abundances measured for PNe. Furthermore, the average [HCN]/[HNC] ratio found for diffuse clouds is remarkably close to that found in PNe: 5 versus 1–8. These results bolster the argument made in Schmidt & Ziurys (2016) and Ziurys et al. (2015, 2016) that the polyatomic molecules observed in diffuse clouds originate in PNe ejecta. The clumps of molecular material, mixed with dust, disperse into the diffuse ISM and later coalesce into clouds.