The Composition of Comet C/2012 K1 (PanSTARRS) and the Distribution of Primary Volatile Abundances Among Comets

Comets are among the most primitive remnants from the formation of the solar system. They were some of the first bodies to accrete in the solar nebula, forming in the outer (>5 AU) giant planet region. The chemical composition of their nuclei should reflect the chemical makeup of the midplane of the protoplanetary disk where (and when) they formed. Gravitational interactions with the giant planets during the final phases of planet formation ejected many comets into either the Oort cloud (Gladman 2005) or the Kuiper disk (scattered disk population, see Morbidelli & Brown 2004). These two regions make up the major dynamical reservoirs of the solar system for comets that become available for remote sensing using high-resolution spectroscopy.

Since their emplacement in the Oort cloud or the Kuiper disk, the interior compositions of cometary nuclei have remained (at least to a large degree) unchanged. Most processes considered to alter the properties of the nucleus during its (∼4.5 billion years) residence in the Oort cloud (or the Kuiper disk) are expected to affect a thin (a few meters deep) layer near the surface (see Stern 2003 for a detailed discussion of these processes for Oort cloud comets). This layer is lost during a typical passage through the inner solar system. Due to the scattering processes that placed comets in their present-day reservoirs, the Oort cloud and Kuiper disk contain comets that may represent widely varying (or, at the other extreme, largely overlapping) formation regions in the solar nebula. Determining the native volatile composition (i.e., as contained as ices in the nucleus) can provide insights into these formation regions and also the formation pathways (Mumma & Charnley 2011).

As comets enter the inner solar system, increasing radiation from the Sun causes native ices to sublimate and release primary volatiles into the coma (a freely expanding atmosphere, or exosphere), a dust tail, and an ion tail. Near-infrared spectroscopy of fluorescent emission can be used to characterize the primary volatile composition of the coma, and by inference the nucleus. Early results led to characterization of (at least) three taxonomic classes: "organics-depleted," "organics-normal," or "organics-enriched" (Mumma & Charnley 2011), based on measured abundance ratios (also termed "mixing ratios") of their primary volatiles relative to H₂O (the most abundant ice in comets). However, the compositions of some comets do not fit into any of these proposed taxonomic classes, challenging and requiring expansion of this classification system (Bonev et al. 2008b; Gibb et al. 2012; Radeva et al. 2013).

To that end, we add the volatile composition of comet C/2012 K1 (PanSTARRS) to the body of work, with the hope of further establishing the taxonomic classification of primary volatiles among comets. In Section 2, we discuss our observations and data analysis. In Section 3, we present our results. In Section 4, we provide a detailed discussion of our results in the context of other comets sampled at infrared wavelengths.

Comet C/2012 K1 (PanSTARRS) (hereafter K1) was a dynamically new Oort cloud comet on its first journey into the inner solar system (Nakano 2013). K1 reached perihelion (1.05 AU) on 2014 August 27 and was closest to Earth (0.95 AU) on 2014 October 31. On 2014 May 22 and 24, we observed K1 with the high-resolution (λ/Δλ ≈ 25,000), near-infrared, long-slit echelle spectrograph NIRSPEC at the 10 m W. M. Keck Observatory (McLean et al. 1998) to characterize its volatile composition.

The observing log is shown in Table 1. Seeing improved from ∼05 to ∼02 over the course of the night on May 22, and was stable at ∼05 on May 24. The column burden of atmospheric water vapor (expressed in precipitable millimeters) retrieved in fitting synthetic telluric absorption models to flux standard star continua was 2.9 on May 22, and 2.2 on May 24.

We targeted nine primary volatiles (CO, H₂O, C₂H₂, C₂H₆, CH₄, H₂CO, CH₃OH, HCN, and NH₃) and two product species (OH* and NH₂). Observations were performed with a 3 pixel (043) wide slit, using a standard ABBA nod pattern, with a 12'' beam separation along the 24'' long slit. Combining spectra of the nodded beams as A–B–B+A cancelled emissions from thermal background, instrumental biases, and "sky" emission (lines and continuum) to second order. The data were dark subtracted, flat fielded, cleaned of cosmic ray hits and high dark current pixels, and corrected for anamorphic optics. A detailed description of the flux calibration (using BS 5447) and reduction procedure can be found in Bonev (2005, Appendix B), Radeva et al. (2010), Villanueva et al. (2011a), and references therein.

Atmospheric spectra were synthesized using the Line-By-Line Radiative Transfer Model optimized for Mauna Kea's atmospheric conditions (Clough et al. 2005; Villanueva et al. 2011b). These models were used to determine column burdens for absorbing species in the atmosphere and to assign wavelength scales to the extracted spectra. The atmospheric models were binned to the resolution of the comet spectrum and normalized to the comet's continuum level. The atmospheric models were then subtracted from each row of the cometary spectra; co-addition of multiple rows resulted in the comet emission spectra shown in Figures 2–3.

Production rates (Q, molecules s⁻¹) were determined using the Q-curve methodology (e.g., DiSanti et al. 2001; Bonev 2005; Gibb et al. 2012), which averages the emission intensity on either side of and equidistant from the nucleus. Averages were stepped in 0.6 arcsec intervals along the slit, resulting in a "symmetric" Q-curve. A spherically symmetric outflow velocity (${v}_{\mathrm{gas}}=0.8\,{{R}_{h}}^{-0.5}$ km s⁻¹) was assumed (Bonev 2005). The symmetric Q-values increase with nucleocentric distance due primarily to atmospheric seeing, until reaching a terminal value, referred to as the "global" production rate (Q_global).

Growth factors (defined as GF = Q_global/Q_NC, where Q_NC is the nucleocentric production rate—see Section 3.2) were determined for both the gas and the dust when the signal-to-noise ratio (S/N) was sufficiently high (i.e., only for water and ethane). These two species had similar spatial profiles (see Figure 1) and therefore, since their photo-dissociation lifetimes are comparable, provided similar growth factors (see Table 2).

Zoom In Zoom Out Reset image size

We determined production rates for six primary volatiles (H₂O, HCN, CH₄, CH₃OH, C₂H₆, and CO) and determined upper limits for three others (C₂H₂, NH₃, and H₂CO) in K1. Synthetic models of fluorescent emission for each targeted species were compared to observed line intensities, after correcting each line for the monochromatic atmospheric transmittance at its Doppler-shifted wavelength (according to the geocentric velocity of the comet at the time of the observations). The g-factors used in synthetic fluorescent emission models in this study were generated with quantum mechanical models for each molecule. These models include CH₄ (Gibb et al. 2003), C₂H₆ ν₇ (Villanueva et al. 2011b), H₂O (Villanueva et al. 2012b), CH₃OH (Villanueva et al. 2012a; DiSanti et al. 2013), HCN (Villanueva et al. 2011a; Lippi et al. 2013), H₂CO (DiSanti et al. 2006), OH* (Bonev et al. 2006), C₂H₂ (Villanueva et al. 2011a), CO (Paganini et al. 2013b), and NH₃ (Villanueva et al. 2013). A Levenberg–Marquardt nonlinear minimization technique (Villanueva et al. 2008) was used to fit fluorescent emission from all species simultaneously in each echelle order, allowing for high-precision results, even in crowded spectral regions that contained many spectral lines within a single instrumental resolution element.

3.1. Determination of Rotational Temperature

Rotational temperatures were determined using correlation and excitation analyses as described in Bonev (2005, pp. 53–65), Bonev et al. (2008a), DiSanti et al. (2006), and Villanueva et al. (2008). In general, well-constrained rotational temperatures can be determined for individual species with intrinsically bright lines and for which a broad range of excitation energies is sampled.

For our observations, the most robust rotational temperature (T_rot = 42 ± 7 K) was found for H₂O in order 26 of our KL2 setting on May 22. We were also able to retrieve a rotational temperature for HCN (${{43}^{+11}}_{-10}$ K), in agreement with that for water. (In general, rotational temperatures agree for different primary species measured at infrared wavelengths (see for example Gibb et al. 2012 and references therein; also see Section 3.2.1 of DiSanti et al. 2016), supporting this approach.) The rotational temperature derived for H₂O was therefore applied to species for which the rotational temperature could not be well constrained.

The H₂O rotational temperature was poorly constrained on May 24, owing to poor S/N (less on-source integration time, see Table 1) in orders with temperature-sensitive water lines. Therefore, the May 22 H₂O rotational temperature was adopted in determining production rates and abundances on May 24. Additionally the water production rate in the KL1 setting on May 24 was significantly higher than that found for the KL2 setting (see Table 2), as well as the water production rate on May 22. However, production rates for all other trace volatiles in the KL1 setting on May 24 agree within uncertainty with those on May 22. McKay et al. (2016) reported a water production rate of 4.35(0.44) × 10²⁸ s⁻¹ on May 24, consistent with our Q(H₂O) from May 22. Such variations are not unknown in comet observations and may be due to short-term variability. Spectra and best-fit fluorescence models are shown in Figures 2 and 3. Best-fit rotational temperatures, growth factors, production rates, and mixing ratios for each date are given in Table 2.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 2.  Volatile Composition of Comet C/2012 K1 (PanSTARRS)

NIRSPEC Setting	Molecule	${T}_{\mathrm{rot}}$ a (K)	Growth Factorb	Qc (1026 mol s−1)	Mixing Ratiod (%)
	2014 May 22, Rh = 1.857 AU, Δ = 1.557 AU
KL2	H2O	42 ± 7	1.6 ± 0.2d	439 ± 31	100
	H2CO	(42)	(1.6)	<0.61 (3σ)	<0.14 (3σ)
	CH4	(42)	(1.6)	2.02 ± 0.37	0.46 ± 0.09
	HCN	(42)	(1.6)	0.66 ± 0.07	0.15 ± 0.02
		${{43}^{+11}}_{-10}$	(1.6)	0.67 ± 0.07	0.15 ± 0.02
	C2H2	(42)	(1.6)	<0.49 (3σ)	<0.11 (3σ)
	NH3	(42)	(1.6)	<7.6 (3σ)	<1.8 (3σ)
KL1	H2O	(42)	1.7 ± 0.1d 2.3 ± 0.5e	446 ± 27	100
	C2H6	(42)	1.7 ± 0.1e 1.0 ± 0.3d	4.37 ± 0.18	0.98 ± 0.07
Order 22	CH3OH	(42)	(1.7)	10.5 ± 1.6	2.36 ± 0.39
Order 23				13.6 ± 1.3	3.06 ± 0.35
				12.4 ± 1.2f	2.74 ± 0.26
	2014 May 24, Rh = 1.834 AU, Δ = 1.574 AU
KL2	H2O	(42)	1.6 ± 0.2d	353 ± 68	100
	HCN	(42)	(1.6)	0.39 ± 0.10	0.11 ± 0.04
KL1	H2O	(42)	(1.6)	595 ± 29	100
	C2H6	(42)	1.6 ± 0.1d	4.63 ± 0.32	0.80 ± 0.06
	CH3OH	(42)	(1.6)	15.0 ± 2.0	2.58 ± 0.36
	CH4	(42)	(1.6)	<5.4 (3σ)	<0.91 (3σ)
MWA	H2O	(42)	(2.0)	361 ± 107	100
	CO	(42)	(2.0)	14 ± 2	3.9 ± 1.2

Notes. For settings with both dust and gas growth factors, bold font indicates the growth factor used for determining production rates.

^aRotational temperature. Values in parentheses are assumed. ^bGrowth factor. Values in parentheses are assumed. ^cGlobal production rate. Errors in production rate include line-by-line deviation between modeled and observed intensities and photon noise (see Dello Russo et al. 2004; Bonev 2005; Bonev et al. 2007). ^dContinuum (dust) growth factor. ^eGas growth factor. ^fWeighted average CH₃OH production rate from KL1 Order 22 and KL1 Order 23.

Download table as:  ASCIITypeset image

3.2. Production Rates and Mixing Ratios

The matter of classifying comets according to their primary volatile composition has proven to be a complex undertaking. Extensive work at optical wavelengths has revealed that comets can be classified as "typical" or "carbon-chain depleted" based on their product species (e.g., A'Hearn et al. 1995; Cochran et al. 2012 and references therein), however their precursors are often uncertain and can be other volatiles or even dust grains. Additional work has been done using radio observations, where no clear taxonomic classes have been found (Crovisier et al. 2009; Mumma & Charnley 2011, and references therein). A similar endeavor began in the infrared with comets 1P/Halley (Mumma et al. 1986), C/1996 B2 (Hyakutake) (Mumma et al. 1996; Dello Russo et al. 2002b; Magee-Sauer et al. 2002b; DiSanti et al. 2003), and C/1995 O1 (Hale-Bopp) (Magee-Sauer et al. 1999; Dello Russo et al. 2000, 2001; DiSanti et al. 2001). The primary volatile compositions of these comets suggest that they are chemically similar objects (Mumma et al. 2003). Subsequent observations of comets D/1999 S4 (LINEAR) prior to its complete disruption (Mumma et al. 2001a) and of the split comet 73P/Schwassmann-Wachmann 3-B (Villanueva et al. 2006; Dello Russo et al. 2007) showed two comets that were highly depleted in virtually all trace primary volatiles relative to water. At the other extreme, comets C/2001 A2 (LINEAR) (Magee-Sauer et al. 2008) and later C/2007 W1 Boattini (Villanueva et al. 2011a) were enriched in the sampled trace primary volatiles. These results formed the basis for the aforementioned (Section 1) three-tiered taxonomy based on primary volatile abundance ratios (organics-enriched, organics-normal, organics-depleted; e.g., see Mumma & Charnley 2011 and references therein).

However, recent work has suggested that the three-fold classification scheme is incomplete and more complex (see Dello Russo et al. 2016 for a recent review of comet taxonomies based on near-infrared spectroscopy). For example, the primary volatile composition of comets 8P/Tuttle, C/2007 N3 (Lulin), and 2P/Encke (Bonev et al. 2008b; Gibb et al. 2012; Radeva et al. 2013) show no systematic enrichment, depletion, or similarity to the mean. Among these three comets, CH₃OH may be seen as a "smoking gun" in that it is "overabundant" compared to other primary volatiles. Specifically, these three comets all had high CH₃OH abundances while being depleted in certain other molecules, for example C₂H₂, and "normal" in others, such as C₂H₆. This suggests that the chemical diversity among comets is more complex than the simple organics-enriched, organics-normal, and organics-depleted framework. This should not be surprising, and may even be expected given that the taxonomy based on product species now suggests as many as seven distinct groupings (Schleicher & Bair 2014; Cochran et al. 2015). However, both dust and gas can contribute product species, complicating the comparison with the emerging taxonomy based on primary species alone.

The next natural question is whether the distribution of primary volatiles among comets is more nearly continuous versus distinct. Figure 4 shows abundances relative to water for HCN, C₂H₆, CH₃OH, and CH₄, respectively, in comets. For each molecule, most comets have abundances close to their median value, with some showing enrichment in certain molecules and depletion in others. Overall, the abundances of well-sampled primary volatiles, such as HCN, C₂H₆, and CH₃OH, suggest the emergence of a continuous distribution among comets. The graphic for HCN (Figure 4(A)), one of the most well measured volatiles in comets, shows an example of a well filled-in continuous distribution of volatile abundances between 0.6% and 0.05%. The results for K1 reinforce this view. Prior to this study, there was a lack of comets with C₂H₆ abundances between 0.87% and 1.70% (between nearly "average" and "enriched"; Figure 4(B)). While the C₂H₆ abundance in K1 on May 24 was within the range observed to date (due to the higher water production rate in that setting, see Section 3.1 and Table 2), on May 22 it was about ∼1% and fell within the previously unsampled range. This also suggests that the apparent gap for CH₃OH abundances between 0.20% and 1.0% (between "depleted" and closer to "average"; Figure 4(C)) may be expected to be "filled in" with additional comet observations.

Zoom In Zoom Out Reset image size

CH₄ has been sampled in fewer (∼20) comets, and it appears that a gap remains between C/1999 T1 and the remaining comet population (Figure 4(D)); however, due to the large uncertainty in the T1 CH₄ abundance, the significance of the gap is unclear. CH₄ is difficult to detect, particularly in Jupiter-family comets (JFCs), due to lack of sensitivity with available instruments and (especially) the requirement that observable comets have a sufficiently large geocentric Doppler shift to displace cometary emission lines from the (opaque) cores of corresponding telluric absorptions. Clearly, more work is needed to characterize CH₄ in comets, and JFCs in particular.

Examination of Figure 4 also shows that the level of enrichment or depletion in a given comet does not necessarily correlate across all molecules sampled. One comet may be enriched in CH₃OH and consistent with normal in HCN (K1—see red arrows in Figure 4) while another may be depleted in CH₃OH but not in HCN (e.g., 73P/SW 3-B), challenging attempts to assign definitive taxonomic classes. In support of this conclusion is the lack of strong correlation between the abundance of CH₃OH and the other species represented in Figure 4. Dello Russo et al. (2016) found correlation coefficients of 0.37, 0.66, and 0.51 between CH₃OH and HCN, C₂H₆, and CH₄, respectively.

There are several unanswered questions that need to be addressed before the distribution of volatile abundances in comets can be understood. First, what is the range of abundances for trace volatiles in comets? Are the currently proposed "taxonomic end-members" (C/2001 A2 on the "enriched" end, and D/1999 S4 on the "depleted" end) truly representative of compositional extremes? On the low abundance end, we are limited by technology and the sensitivity of state-of-the-art techniques. On the upper end, we are limited by the relatively small number of comets measured to date with adequate S/N. Of the ∼10¹¹ cometary nuclei that reside in the Oort cloud (Emel'Yanenko et al. 2007), we have measured primary volatile abundances for only about 30 comets in the infrared. For some molecules, most specifically C₂H₂ and OCS (carbonyl sulfide), that number is much lower, due principally to lack of sensitivity (in the case of C₂H₂) and/or spectral coverage (in the case of OCS) in our "standard" NIRSPEC settings.

However, we expect both areas to be addressed with the availability of a powerful new cross-dispersed spectrograph (iSHELL) at the NASA Infrared Telescope Facility (IRTF; Rayner et al. 2012). Specifically, regarding C₂H₂, iSHELL/IRTF will enable very long on-source integrations, including allowance for daytime observing. Regarding OCS, many lines will be sensed simultaneously with CO and H₂O in a standard iSHELL M-band setting. Conversely, targeting OCS with NIRSPEC requires an additional setting to MWA, one that includes CO lines having higher rotational quantum numbers. For rotational temperatures typically found in comets, these higher-J CO lines are relatively weaker than the low-J lines included in MWA (e.g., see Figure 4 in Gibb et al. 2012). As the answers to these questions become more clear, we may also ask whether the distribution of primary volatile abundances in comets is a primordial effect preserved from cometary formation in the solar nebula, or if we are instead sampling heterogeneous nuclei, such as the binary comet 67P/Churyumov–Gerasimenko (Rickman et al. 2015).

Once we understand the distribution of volatile abundances in comets, we can attempt to interpret the abundance of a given volatile in a comet in terms of the extant conditions during its formation. In principle, this can be accomplished by comparing volatile abundances measured in comets with those of ices as predicted by models of protoplanetary disk midplanes (Drozdovskaya et al. 2016). However, as with adding to the inventory of comet primary volatile abundances, much work remains to be done in improving protoplanetary disk models before firm conclusions can be drawn. Clearly, more studies of the primary volatile compositions of comets are needed to answer these complex questions.

The data presented in this study were obtained using the W. M. Keck Observatory, which is operated as a scientific partnership among the California Institute of Technology, the University of California, and the National Aeronautics and Space Administration. The Observatory was made possible by the generous financial support of the W. M. Keck Foundation. We recognize and acknowledge the very significant cultural role and reverence that the summit of Maunakea has always had within the indigenous Hawaiian community. Keck telescope time was granted by NOAO (Prop. ID: 2014A-0379; PI: B. Bonev). We thank both NOAO and Keck support staff for their significant assistance. This study was generously funded by the NASA Missouri Space Grant Consortium and NSF Planetary Astronomy Grant AST-1211362, AST-1616306, and AST-1615441. NASA supported this work through its Planetary Astronomy Program (proposal 11-PAST11-0045 and grant #NNX12AG24G), Astrobiology Program (awarded by the NASA Astrobiology Institute to the Goddard Center for Astrobiology under proposal 13-13NAI7-0032), Emerging Worlds Program (grant #NNN12AA01C), Planetary Atmospheres Program (grant #NNX12AG60G), and Solar System Observations Program (grant #NNH15ZDA001N-SSO). This work was supported by NASA Headquarters under the NASA Earth and Space Science Fellowship Program—Grant "NNX16AP49H." We gratefully acknowledge support from the International Space Science Institute (team 361).