Carbon Chain Depletion of 2I/Borisov

The modern composition and structure of solar system comets retain information about conditions of the early protosolar disk from which they formed (Duncan et al. 1987; Bar-Nun & Kleinfeld 1989; Mumma et al. 1993; A'Hearn et al. 2012). Telescopic surveys of comet molecular abundances (A'Hearn et al. 1995; Fink 2009; Cochran et al. 2012) have revealed there to be multiple taxonomic classes of solar system comets whose differences are likely tied to differences in disk chemistry. Dynamical work also suggests multiple formation regions inside the disk (Biver & Bockelée-Morvan 2015; Dones et al. 2015; Bockelée-Morvan & Biver 2017; Meech et al. 2017) related to the modern dynamical classes of comets (Jupiter-family, Halley-type, long-period, etc.). Connections between dynamical class and composition ("normal," carbon-chain depleted, carbon-chain and NH₂ depleted, etc.) have also been identified (A'Hearn et al. 1995; Fink 2009; Cochran et al. 2012) using spectroscopy of cometary comae. Spectroscopic observations of an active cometary coma, whether the object is from our solar system or interstellar, thus investigates the link between disk properties and produced planetesimals, and provides constraints on models of planetesimal and planet formation.

The discovery of the second-ever interstellar object comet 2I/Borisov on 2019 August 30 (MPEC 2019-R106)⁴ presents a unique opportunity to contrast planetesimal formation in our solar system with formation processes in other stellar systems—in a way not possible with the first interstellar object found, 1I/'Oumuamua. 'Oumuamua is red with a pronounced elongation (Knight et al. 2017; Meech et al. 2017; Fitzsimmons et al. 2018; Trilling et al. 2018; Ye et al. 2018) with an assumed low, but undetermined, visible (<0.15) and infrared (<0.2) albedo (Fitzsimmons et al. 2018; Trilling et al. 2018) in a non-principal axis ("tumbling") rotational state (Belton et al. 2018; Drahus et al. 2018; Fraser et al. 2018). There is strong evidence for nongravitational acceleration (Micheli et al. 2018). No gas emissions were detected from 'Oumuamua: chemical analyses were limited to observing its surface reflectivity (Fitzsimmons et al. 2018) and are consistent with an outer solar system body with appreciable organic content. All of these properties contrast with 2I/Borisov; the object's substantial outgassing provides more data about its formation, thermal history, and chemistry.

We summarize the currently known properties of 2I/Borisov. Its eccentricity and inclination are 331 and 441, respectively (JPL Orbit Solution #11, queried for 2019 October 1.0). Due to its low solar elongation angle it is visible only in the early morning, presenting a brief observing window. Guzik et al. (2019) obtained g'- and r'-band images with Gemini North, finding 2I/Borisov's color to be indistinguishable from solar system comets. de León et al. (2019) reported a red slope between 0.55 and 0.90 μm similar to D-type asteroids and cometary nuclei. Fitzsimmons et al. (2019) detected CN emission from 2I/Borisov (Q ∼ 4 × 10²⁴ molecules s⁻¹) at a level appearing typical for comets at similar heliocentric distances. Fitzsimmons et al. (2019) also constrained the C₂ production (Q < 4 × 10²⁴ molecules s⁻¹) and estimated water production by proxy (Q ∼ 1.7 × 10²⁷ molecules ⁻¹). McKay et al. (2019) derived from the 6300 Å [O i] line a water production rate of ∼6.3 ∗ 10²⁶ mols s⁻¹, which could be high or typical compared to solar system comets depending on size estimated (Jewitt & Luu 2019).

Carbon "typical" comets have production rate ratios of C₂/CN of 0.66–3.0 (A'Hearn et al. 1995; Cochran et al. 2012), while comets with less C₂ are called "depleted." Fitzsimmons et al. (2019) cannot discriminate 2I/Borisov between "typical" and "depleted" for 2I/Borisov, though the detection of CN does effectively rule out an uncommon composition like that of Comet Yanaka (1988r; Fink 1992) or 96P/Machholz (Schleicher 2008). Opitom et al. (2019) found no evidence of C₂ in their spectroscopic observations, suggesting moderate-to-strong (C₂/CN < 0.3) carbon-chain depletion. Spectroscopic observations can place chemical constraints on 2I/Borisov, allowing insight into another solar system's disk chemistry and formation history by examination of the interstellar comet's C₂ and CN abundance in ways that 1I/'Oumuamua could not.

We observed interstellar comet 2I/Borisov on 2019 September 20 with the Boller & Chivens spectrograph at the 2.3 m Bok telescope (Kitt Peak National Observatory, Arizona), on 2019 October 1 and 9 with the Blue Channel spectrograph at the 6.5 m MMT telescope (Mount Hopkins, Arizona; Hastie & Williams 2010), and on 2019 October 10 with MODS on the Large Binocular Telescope (LBT; Pogge et al. 2006) to search for and measure the abundance of molecular species as well as to measure the reflective properties of dust in the object's coma. We also imaged Borisov in the Hale–Bopp CN (Farnham et al. 2000) and Sloan r' filters on 2019 October 26 with the 90 Prime imager (Williams et al. 2004) on the Bok telescope. Observational details, including heliocentric and geocentric distances, spectral resolution, and exposure times among other information, are listed in Table 1. For context, our Bok observations took place approximately 7 hr after Fitzsimmons et al. (2019).

These spectra and images were reduced using standard techniques, including bias subtraction, flat-fielding, cosmic-ray rejection, local sky subtraction, and extraction of one-dimensional spectra. The spectra had wavelength solutions derived using comparison with helium–neon–argon lamps, had a standard extinction correction applied, and were flux-calibrated by comparison with nearby flux standard stars observed just beforehand. For the spectroscopic observations extracted apertures were chosen to maximize the comet signal for each instrument and slit. For the data from 2019 September 20, this was an 1833 (spatial) by 15 (slit width) box, 2019 October 1 had 15575 by 10, 2019 October 9 had 260 by 15, and 2019 October 10 had 18'' by 24. The imaging data were extracted using circular aperture photometry with a radius of 250 and largely followed the reduction procedure outlined in Farnham et al. (2000) using the dust reflectance slope measured from our 2019 October 10 spectrum. The standard star used (HD 81809) is noted in that work to be less than ideal for blue/UV observations, so we have overestimated errors when possible. We note that for the 2019 September 20 and October 1 data sets, insufficient data were collected to allow a median combination of spectra, making the two data sets "look" even noisier than their later counterparts even though cosmic rays were cleaned from them.

In order to search for molecular emission features, it is necessary to produce a flux-calibrated solar spectrum fit to subtract from the flux-calibrated data. This process creates a dust reflectance spectrum for the comet, which can be used to look for a spectral slope. The solar spectrum was taken from Chance & Kurucz (2010) and binned to the resolution of each data set. The original flux-calibrated spectra, best-fit reflectance, and dust-subtracted comet spectra are shown in Figure 1. The noisiness of the Bok data prevents a high-quality solar reflectance subtraction, while the later data become very linear after subtraction indicating a relatively good fit of the solar spectrum to the continuum points.

Zoom In Zoom Out Reset image size

We consider the spectral slope between the blue, green, and red continuum regions defined for the Hale–Bopp filter set (Farnham et al. 2000) to compare reflectance slopes with other workers. These filters are centered at 4450, 5260, and 7128 Å, respectively. If we normalize our highest-quality data from 2019 October 10 at the green continuum point, then we find the blue-to-green slope to be S' = 22+/−4%/micron, and the green-to-red slope to be S' = 11+/−3%/micron. Fitzsimmons et al. (2019) suggested that the dust might be "redder" at shorter wavelengths, and our best data support that conclusion.

After the reflected sunlight from dust was subtracted, we searched for emission from molecules in the coma of 2I/Borisov (Table 2). For each possible feature we subtracted a linear estimate of the background as determined by a fit to data on either side of the feature and then integrated using a trapezoidal method all flux above it. That retrieved flux is compared against an upper limit found through the method described in Cochran et al. (2012), whereby a fraction of the local spectral noise is multiplied over the width of the bandpass that the feature was integrated over to determine blindly if the feature is "real." All retrieved fluxes and upper limits are in Table 2.

The CN B–X (0–0) band near 3883 Å is a relatively narrow two-peaked feature that we integrated from 3861 to 3884 Å based on the extent of the feature as calculated in Schleicher (2010). We detect CN strongly on 2019 October 1, 9, and 10, but the detection on 2019 September 20 is marginal, as background subtraction was challenging so close to the edge of the detector. We thus list it in Table 2 as an upper limit, but if it were a detection, the value should be similar to the value listed.

The C₂ Swan band $({\rm{\Delta }}{\boldsymbol{\nu }}=0)$, with a bandhead at 5167.0 Å, is a much broader feature stretching over 200 Å in wavelength. The overall shape of the band has a sharp peak near the primary 0–0 bandhead and a secondary peak near 5129 Å at the 1–1 bandhead. We integrate from 5067 Å to the 0–0 bandhead at 5167 Å and to match the procedures of Fitzsimmons et al. (2019) and Opitom et al. (2019). There is no C₂ seen above the noise in any of the data. Detailed sections of the 2019 October 1, 9, and 10 spectra of 2I/Borisov show the CN detections and C₂ nondetections on those dates (Figure 2). For the 2019 October 10 data, we also measured upper limits over other ranges. If we integrate from the 0–0 bandhead to the 1–1 bandhead (5129–5167 Å), we get 39% of the 100 Å upper limit, and if we integrate over the full 200 Å, then we get 218% of the 100 Å upper limit. This is consistent with there being no variation in spectral properties or noise throughout the bandpass.

Zoom In Zoom Out Reset image size

To convert the integrated fluxes to production rates, we used a standard Haser model (Haser 1957) with scale lengths and lifetimes from A'Hearn et al. (1995) and outflow velocity from Cochran & Schleicher (1993). We note that the A'Hearn et al. scale lengths and lifetimes assume a different outflow velocity, though this does not change inferred production rate ratios. The g-factor (formally the fluorescence efficiency) for C₂ was taken from A'Hearn et al. (1995) and the g-factor for CN was taken from Schleicher (2010) for the appropriate heliocentric distance and velocity on each date. The lifetimes increased proportional to heliocentric distance squared and the g-factors were scaled down by the same factor. We have verified the output of our Haser model implementation by comparison of its output with a recent paper on the topic, Hyland et al. (2019). The input fluxes and output production rates for all nights are presented in Table 2. The production rate ratio upper limit, Q(C₂)/Q(CN), is measured as <0.1, which indicates that 2I/Borisov is clearly in the "depleted" group of the A'Hearn et al. (1995) taxonomy if there is any C₂ at all. A comparison of our inferred Q(C₂)/Q(CN) ratio with those of the A'Hearn et al. (1995) sample as submitted to the Planetary Data System in 2003 (Osip et al. 2003) as well as that of Opitom et al. (2019) is presented in Figure 3.

Zoom In Zoom Out Reset image size

Fluxes are in ergs cm⁻² s⁻¹, and production rates are in molecules s⁻¹. Production rate ratios are rounded to the nearest tenth and are based on the 100 Å upper limit.

4.1. 2I/Borisov as a Depleted Comet

4.2. Comparison with Other Measurements

We report on spectroscopic and imaging observations of 2I/Borisov between 2019 September 20 and 2019 October 26 using the Bok, MMT, and LBT telescopes to characterize and measure the gas production rates and dust reflectance properties to compare with solar system comets. We find the dust reflectance to be red (S' = 22 (11)%/micron below (above) the Hale–Bopp green continuum filter) and likely shallower at larger wavelengths. We identify CN emission on all 2019 October dates and convert the integrated fluxes to production rates using a standard Haser model with a rectangular (for spectra) or circular (for imaging) aperture. From 2019 October 1 to 26, the production rate ranged from Q(CN) = (1.1–1.9) ∗ 10²⁴ mols s⁻¹ after our initial upper limit on 2019 September 20 (Q(CN) < 5.0 ∗ 10²⁴ mol s⁻¹). These values and dust reflectance properties are all consistent with contemporaneous observations by Fitzsimmons et al. (2019) and Opitom et al. (2019). We do not detect emission from C₂ on any date and set upper limits in the standard method of Cochran et al. (2012). For our most sensitive observations from the LBT on 2019 October 10, we find Q(C₂) < 1.6 ∗ 10²³ using a 100 Å wide integration area. The ratio of C₂ to CN production rates was found on that date to be Q(C₂)/Q(CN) < 0.1, which is extremely depleted if any C₂ is even being produced at the current time. This ratio is used as a diagnostic measure of composition (e.g., "typical" or "depleted"; see A'Hearn et al. 1995; Fink 2009; Cochran et al. 2012), and only three solar system comets have been detected to be more depleted than our upper limit. However, the comparability of our upper limit to detections in surveys with different goals is not ideal. It is unclear whether or not 2I/Borisov has little C₂, if any, or the C₂ it might have is simply not being released yet. More work is needed before we can truly apply a solar system classification onto 2I/Borisov and know that the statement has a physical meaning. More work is very much needed to further constrain this ratio and identify if 2I/Borisov has any long-chain hydrocarbons at all. The recent discovery of the interstellar comet 2I/Borisov presents an unparalleled opportunity to compare its composition to solar system comets and test the applicability of planetesimal formation models to other stellar systems.

We would like to congratulate Gennadiy Borisov on this great discovery and thank the IAU for letting the object retain his name as is typical for comets. We thank Kendall Sullivan among others for encouragement to pursue this project. We thank an anonymous reviewer for comments that greatly improved the Letter as well as Alan Fitzsimmons for very helpful correspondence on the topic. J.E.A. and N.S. receive support from NSF grant AST-1515559. K.V. acknowledges funding from NSF (grant AST-1824869). Some observations in this work were collected at Kitt Peak National Observatory; we are honored to be permitted to conduct astronomical research on Iolkam Du'ag (Kitt Peak), a mountain with particular significance to the Tohono O'odham Nation. Observations reported here were obtained at the MMT Observatory, a joint facility of the University of Arizona and the Smithsonian Institution. We appreciate the expertise of the Kitt Peak National Observatory, Steward Observatory, MMT Observatory, and LBT Observatory staff as well as Hannes Gröller who helped ensure the success of these observations.

This Letter made use of the modsIDL spectral data reduction reduction pipeline developed in part with funds provided by NSF grant AST-1108693 and a generous gift from OSU Astronomy alumnus David G. Price through the Price Fellowship in Astronomical Instrumentation.

modsCCDRed was developed for the MODS1 and MODS2 instruments at the Large Binocular Telescope Observatory, which were built with major support provided by grants from the U.S. National Science Foundation's Division of Astronomical Sciences Advanced Technologies and Instrumentation (AST-9987045), the NSF/NOAO TSIP Program, and matching funds provided by the Ohio State University Office of Research and the Ohio Board of Regents. Additional support for modsCCDRed was provided by NSF grant AST-1108693.

Facilities: MMT(BCS) - MMT at Fred Lawrence Whipple Observatory, Bok (B&C - , 90Prime) - , LBT (MODS). -