C/2016 R2 (PANSTARRS): A Comet Rich in CO and Depleted in HCN

Important modeling parameters, such as expansion velocity, v_exp, and outgassing patterns, can be extracted from spectral line profiles and maps that have sufficiently high resolution (Biver et al. 2002). Comet activity models typically consider two different sources for CO in comets: one emanating from the subsolar point where solar heating is greatest, and one from an isotropic source in the coma. Here, we briefly address how the R2 CO data aligns with the models.

First, we examine the CO line profile, which has a single velocity component and is slightly blueshifted from the comet's ephemeris velocity by δv = −0.12 ± 0.20 km s⁻¹ (see Figure 1). The line is between 2–3 channels wide (corresponding to FWHM of 0.66–0.99 km s⁻¹), and by fitting a Gaussian, we derived a FWHM linewidth of Δv_FWHM = 0.85 ± 0.33 km s⁻¹. Thus, the data are consistent with having a FWHM linewidth of ∼0.8 km  s⁻¹. The small velocity shift we observe of ∼−0.1 km s⁻¹ for the comet with a low phase angle is consistent with at least some of the outgassing taking place on the Earth-facing side. The half-width half-maximum linewidth measured on the blue-ward wing is proportional to the outflow velocity of the gas (see Biver et al. 1999), and thus, we estimate the gas expansion velocity to be v_exp = 0.50 ± 0.15 km s⁻¹. This is comparable to what was measured in other comets at this same heliocentric distance, such as C/1995 O1 (Hale–Bopp) and C/2006 W3 (Christensen) (Gunnarsson et al. 2003; Bockelée-Morvan et al. 2010). de Val-Borro et al. (2018) also report seeing CO emission from this comet approximately three weeks after our first detection but with a velocity redshift of ∼+1 km s⁻¹ and a somewhat broader linewidth of 1.0–1.3 km s⁻¹. We also observed the comet during this time and do not confirm their redshifted velocity component.

The nearly centered and narrow CO spectrum of R2 is consistent with a simple model of cool gas expanding isotropically with v_exp ∼ 0.50 km s⁻¹, which is what we used to calculate production rates. The data are also consistent with a more detailed scenario if one looks more carefully at the spectral line profile. First, we revisit Hale–Bopp and Christensen, which also produced substantial CO at r = 3–4 au. The CO emission in these comets was also slightly offset from the ephemeris velocity by δv ∼ −0.1 km s⁻¹ (Biver 1997; Womack et al. 1997; Bockelée-Morvan et al. 2010). For Hale–Bopp, the CO line profiles were fit by a detailed two-component model comprised of isotropic outgassing of cold CO gas, combined with a blueshifted velocity component associated with a sunward-side active area (Gunnarsson et al. 2003). This model produces two peaked lines in comets beyond 4 or 5 au and single peaked lines for comets within 4 au. Therefore, the CO emission is also consistent with CO arising from a mix of a sunward-side active area and a symmetric source either from the coma or nucleus.

We also examined the maps for clues about the CO outgassing. Figure 2 shows that the emission peaks at the nucleus position provided by the ephemeris. Furthermore, CO emission was readily detected at all positions out to at least 45'' with a decrease in intensity by 20%–40% relative to the line at the center position, consistent with isotropic outflow of CO. There is evidence for a slight sunward enhancement of emission on the sunward side compared to the tailward side. This provides further support for contribution from an active area releasing CO on the sunward side, as is also indicated from the spectral line profile.

It is not clear what could generate isotropic CO emission in the coma. Given that CO₂⁺ was not detected in the optical spectrum of R2 (Cochran & McKay 2018a) and that we did not see CH₃OH or H₂CO down to significant limits (K. Wierzchos 2018, in preparation), it is not likely that CO was produced in high amounts by photodissociation of these species, which are plausible secondary sources for CO in other comets. Also, this comet has an almost nonexistent dust coma, and thus, CO is not likely to come from refractory cometary grains in the coma. Perhaps additional CO is released by sublimating water–ice grains in the coma that were ejected from the nuclear sunward-side facing CO source. Measurements of OH or H₂O emission would be useful constraints to the CO production model in R2. Much higher-resolution spectra, ≤0.1 km s⁻¹, would also be valuable for testing models of CO production.

In principle, more detailed modeling of the spectral line profile and mapping data could significantly constrain models of the release mechanisms for CO and physical conditions in the coma, but this requires higher spectral and spatial resolution data and is beyond the scope of this paper.

We calculated column densities assuming the excitation was dominated by collisional and fluorescence contributions, following the modeling described in Crovisier & Le Bourlot (1983), Bockelee-Morvan & Crovisier (1985), and Biver (1997). We assumed a rotational and excitation temperature of 25 K, which is consistent with the empirical fit to Hale–Bopp CO data described in Biver et al. (2002).

The column density for CO was fairly constant, with an average value of N(CO) = (1.89 ± 0.14) × 10¹⁴ cm⁻². In order to calculate production rates, we assumed a gas expansion velocity of 0.50 km s⁻¹, which is consistent with the CO spectral line profile and values from other comets at this distance, as described in Section 3.1. Using a photodissociation decay model (Haser 1957) and assuming isotropic outgassing of CO, we find an average production rate of Q(CO) = (4.6 ± 0.4) × 10²⁸ mol s⁻¹ between 2017 December 22 and 2018 January 16 (see Table 1). de Val-Borro et al. (2018) report higher CO production rates, despite reporting similar line intensities. We think the different production rates are the result of using different modeling parameters, such as expansion velocity. There is insufficient detail about the modeling in their preliminary announcement to warrant further comments.

As discussed in Section 3.1, CO₂ is not likely to be a significant parent of CO in this comet. Infrared observations of CO₂ would be very useful in quantifying any contributions from CO₂ to the coma and/or CO emission. Also, to date, no searches for OH or H₂O emission have succeeded, which implies that water–ice sublimation is not responsible for most of R2's activity. Thus, the major driver at this distance is probably CO outgassing.

The CO production rate of R2 is very high and approximately half that of C/1995 O1 Hale–Bopp at this same distance from the Sun. Based on the high Q(CO) values, we consider R2 to be "CO-rich." Other CO-rich comets, typically have CO/H₂O > 8% (Dello Russo et al. 2016) and the ratio for R2 may be substantially higher than 8%, as water has yet to be detected. We point out that these observations were obtained when the comet was ∼3 au from the Sun, which is too far for water–ice to sublimate efficiently. Thus, the relative CO/H₂O content in the nucleus may be much higher than we can determine at this heliocentric distance.

The nucleus' radius, R_R2, can be estimated based on the assumption that Q(CO) is proportional to the nucleus surface area and the insolation received, and then compared to comets at the same heliocentric distance for which both Q(CO) and radius are independently known. For example, if we use R_HB = 30 km (Fernández 2002) and Q(CO) = 1.8 × 10²⁹ mol s⁻¹ for Hale–Bopp (at 3 au; Biver et al. 2002), then the average Q(CO) = 4.6 × 10²⁸ mol s⁻¹ at ∼3 au corresponds to a radius of R ∼ 15 km. Similarly, from a comparison with comet Christensen's value of R_Ch < 13 km (Korsun et al. 2016) and Q(CO) = 3.9 × 10²⁸ mol s⁻¹ (Bockelée-Morvan et al. 2010) at 3 au, we derive R < 14 km. CO activity may not, in fact, scale directly with surface area for this comet, but if it does, then we find that the nucleus radius need not exceed R_R2 ∼ 15 km in order to explain the measured CO production rate.

We derived a 3σ upper limit of Q(HCN) < 8.0 × 10²⁴ molecules s⁻¹ from all the HCN data. For comparison, this is ∼100 times lower than observed for Hale–Bopp at the same distance (Biver et al. 2002). Our non-detection of HCN emission is consistent with the absence of the CN band at 3880 Angstroms, as reported by Cochran & McKay (2018a), assuming that CN is caused by photolysis of HCN.

We briefly compare the CO and HCN production rates with other comets, as emission from these two species is commonly detected and their ratio may provide insights to the chemical composition of the nucleus and/or coma (Table 2). The comets in the Table that are identified by name are those reported to be CO-rich. Also listed in the table are average values for larger groups of comets, such as Oort Cloud, Jupiter Family, and finally an "all comets" average value. As the table shows, the relative production rate value derived for R2 is extraordinarily high: Q(CO)/Q(HCN) > 5000 at r ∼ 3 au. The average value for all comets measured is Q(CO)/Q(HCN) ∼25 and this ratio varies by less than a factor of three between Jupiter Family Comets and OCCs. The only group where it noticeably departs from the average value is for CO-rich comets, which are listed individually in the top panel of the table. It is perhaps not surprising that the comets designated as CO-rich also have elevated Q(CO)/Q(HCN) values, but even among these CO-rich comets, the limit derived for R2 is the highest values to date for any comet.

The very high Q(CO) and very low Q(HCN) in R2 is difficult to understand in terms of typical comet compositions. At 3 au, R2's comet nucleus has not received much solar heating and so it will preferentially release CO over HCN, due to its higher volatility. This behavior was seen in the abundance ratio of CO/HCN in Hale–Bopp, which decreased as the comet got closer to the Sun and more HCN was released (see Biver et al. (2002) and Table 2). There are not many measurements of both CO and HCN in comets at ∼3 au, but R2's value is substantially higher than those measured for the CO-rich comets Hale–Bopp, Christensen or C/2010 G2 Hill in the range of 2.5–3.0 au, suggesting that R2's high CO/HCN ratio cannot be explained solely due to volatility differences between the two molecules. Interestingly, the highest CO/HCN values were obtained in comets known to be both distantly active and CO-rich (R2, 29P, Christensen, and Hale–Bopp). This is worth looking into further, but the data are sparse. Even for a comet at 3 au, the HCN upper limit that we derived is extraordinarily low. Another possible clue is that R2's coma is largely gaseous with very little dust (Cochran & McKay 2018a), and this may be related to the significantly decreased amounts of HCN and other volatiles. The chemical composition of R2's coma is noticeably atypical when compared to other comets.

Searching for additional clues to the unusual chemical composition of this comet, we now turn our attention to molecular nitrogen. Measuring cometary N₂ is important for many reasons, including testing models of the condensation and incorporation of ices in the protosolar nebula, and calculating the N₂/NH₃ abundance ratio, which is a key diagnostic of primordial physical and chemical conditions (Fegley & Prinn 1989; Womack et al. 1992). N₂ is also a highly volatile molecule and can contribute to comet activity if significantly incorporated in the nucleus. Furthermore, N₂ is trapped and released in a manner similar to Argon, and thus, detecting ${{\rm{N}}}_{2}^{+}$ emission in any coma suggests that Ar may also be present in high amounts (Owen & Bar-Nun 1995). Despite its importance, it has been difficult to measure the N₂ abundances for all but a few comets (Lutz et al. 1993; Cochran 2002; Korsun et al. 2014; Rubin et al. 2015; Ivanova et al. 2016). Strikingly, ${{\rm{N}}}_{2}^{+}$ optical emission is clearly detected in R2, with a measured abundance ratio of N(N₂)/N(CO) = 0.06 (Cochran & McKay 2018a, 2018b).

The N₂/CO ratio is an important observational constraint for testing the formation environment for cometary ice; it is determined by the temperature of the gases when they were incorporated into the ice as well as any subsequent processing that may have preferentially affected one volatile over the other. In order to place this in context, we briefly review two general scenarios for comet formation. The first is one where comets agglomerated from pristine amorphous water–ice grains originating from the interstellar medium onto which N₂ and CO condensed in the protosolar nebula (Owen & Bar-Nun 1993). This model proposes that the N₂/CO ratio in ices strongly depends on the temperature of the materials at the time the volatiles condensed or were trapped. The ratio derived from R2's optical spectra of ${{\rm{N}}}_{2}^{+}$ and CO⁺ agrees very well with the predicted value of N₂/CO = 0.06 for icy planetesimals forming in the solar nebula at about 50 K (Owen & Bar-Nun 1995; Iro et al. 2003). In the second scenario, comets may have agglomerated from crystalline water–ice grain clathrates that trap N₂ and CO. Due to its relatively small size, N₂ is not readily trapped by clathrates, which leads to a lower predicted ratio of N₂/CO ranging from ∼0.002 to 0.02 (Mousis et al. 2012). Thus, perhaps the measured value of N₂/CO = 0.06 in R2 is not consistent with a clathrate model.

In addition to being relatively rich in N₂, we note that this comet is severely depleted in HCN, as discussed in the previous section. Physicochemical models of nitrogen chemistry in protostellar disks show that photodissociation of N₂ leads to production of HCN (Hily-Blant et al. 2017). It is interesting to consider that comet R2 may have formed in the protosolar nebula disk where there was significant N₂ shielding that led to the high N₂ and decreased HCN abundances.

N₂ cannot undergo rotational transitions due to lack of a permanent dipole moment, and thus emits no radiation at millimeter-wavelengths. However, because the N₂ abundance can put such strong constraints on comet formation models, we derived an N₂ column density and production rate using the N₂/CO abundance ratio calculated from optical spectra and our CO results. We chose the CO data from 2017 December 22, because it is closest in time to the Cochran & McKay (2018b) N₂/CO value on 2017 December 8–10. We derived an N₂ column density of N(N₂) = (1.1 ± 0.2) × 10¹³ cm⁻² and a production rate of Q(N₂) = (2.8 ± 0.4) × 10²⁷ molecules s⁻¹. This production rate corresponds to a mass loss rate of 130 kg s⁻¹ for N₂. Determining the N₂ production rate in this manner can be useful in comparison with those of water and/or NH₃, if these volatiles' abundances are established through direct observation or via their daughter products (Tegler et al. 1992; Crovisier 1989).

We report the first detection of neutral CO emission, and an upper limit of HCN, in comet C/2016 R2 PANSTARRS at r ∼ 3 au. The CO line profile shape is characteristic of CO emission seen in other comets at this distance, with a linewidth of ∼0.8 km s⁻¹ and is slightly blueshifted from the ephemeris velocity. A 64'' × 64'' map (∼96000 km × 96000 km at the comet's projected distance) shows that the CO emission peaks in intensity at the ephemeris position and decreases by 20%–40% at the off-centered positions. The spectra and map are consistent with CO arising from a combination of an isotropic source and an active area on the sunward side.

If comet R2's CO output is proportional to surface area of the nucleus, then we find that the radius need not exceed R_R2 ∼ 15 km in order to explain the measured CO production rate. Thus, R2 may be larger-than-average in size, but need not be a giant comet in order to explain the measured CO production rates.

The very large amount of CO, and the apparent absence, or very low outgassing rate, of HCN, leads to a CO/HCN production rate ratio over 5000. This is remarkably high compared to other comets at r = 3 au, even when including other comets known to be CO-rich. The high CO/HCN ratio cannot be explained solely due to volatility differences between the two, and may represent a compositional difference between R2 and most other comets. One possible explanation is that comet R2 formed in a region of the protosolar nebula with substantial N₂ shielding, which could have led to higher N₂ and decreased HCN abundances.

N₂ production rates were derived from the N₂/CO ratio (Cochran & McKay 2018b) and our CO production rates, and were calculated to be Q(N₂) = (2.8 ± 0.4) × 10²⁷ molecules s⁻¹ at 3 au. N₂ production rates will be valuable for comparison with those of water and/or NH₃, if detected in this comet.

R2's coma composition is clearly very different from other comets observed thus far, both in the high N₂ abundance, and significant decrease in other typically abundant molecules, such as HCN. Further observations of this comet along all spectral ranges are highly encouraged, especially those of NH₂ in the optical, and NH₃, CO₂ and CH₄ in the infrared in order to measure the key diagnostic ratios N₂/NH₃, CO/CO₂ and CO/CH₄, which will provide observational tests for formation models of the environment of this comet.