2I/Borisov: An Interstellar Comet

• 2I/Borisov is an interstellar comet defined by a strongly hyperbolic orbit and clear cometary activity including a coma, tail, and outburst events.
• Observations across multiple wavelengths reveal that while its morphology and dust properties resemble Solar System comets, its volatile inventory, especially the high CO/H2O ratio, is distinct.
• Nucleus size estimates vary by method, and its dynamic evolution—including rotational changes and seasonal effects—reflects complex outgassing and activity patterns during its solar approach.

2I/Borisov is an interstellar comet and the second recognized interstellar object after 1I/‘Oumuamua, but it differs from 1I in having displayed an unambiguous coma, tail, and volatile-driven activity. Observations spanning optical, ultraviolet, infrared, millimeter, and high-resolution spectroscopic regimes established a body on a strongly hyperbolic orbit with perihelion near $2.006$–$2.012$ au in December 2019, while also showing that its physical appearance was often close to that of ordinary Solar System comets even when its detailed volatile inventory was not (Guzik et al., 2019, Mugrauer et al., 2020, Opitom et al., 2021).

1. Discovery and dynamical status

2I/Borisov was discovered by Gennady Borisov on 2019 August 30 UT. Early astrometric analysis with 447 positions from 45 stations over 21.1 days yielded a purely gravitational solution with $e = 3.3790 \pm 0.020$, $q = 2.0119 \pm 0.0044$ au, $i = 44.004 \pm 0.041^\circ$, and hyperbolic excess speed $v_\infty \sim 32\ {\rm km\,s^{-1}}$; parabolic solutions required non-gravitational terms so large as to be physically implausible (Guzik et al., 2019). Follow-up astrometry from Jena over 11 epochs in October–November 2019 refined the osculating orbit to $e = 3.3570 \pm 0.0006$, $q = 2.00657 \pm 0.00008$ au, $i = 44.0524 \pm 0.0004^\circ$, with perihelion at $2019\ {\rm Dec}\ 8.555 \pm 0.002$ and closest Earth approach at $2.012$0 au on 2019 Dec 28.2 (Mugrauer et al., 2020).

The interstellar interpretation follows from the orbit itself. The eccentricity is not marginally but strongly hyperbolic, and the Jena analysis states that very close encounters with the giant planets before perihelion can clearly be ruled out, so the observed hyperbolicity cannot plausibly have been generated inside the Solar System (Mugrauer et al., 2020). A separate traceback study used 916 astrometric observations, including pre-discovery detections from December 2018, together with multiple outgassing force models, and found that the incoming asymptote was only weakly sensitive to the assumed non-gravitational prescription. Its preferred barycentric asymptote was $2.012$1, $2.012$2, $2.012$3, with $2.012$4 uncertainties of $2.012$5, $2.012$6, and $2.012$7 respectively (Bailer-Jones et al., 2019).

Searches for a recent stellar origin have remained inconclusive. The closest identified encounter in a 7.4 million-star Gaia DR2 traceback was with the M0V star Ross 573, at a median separation of $2.012$8 pc about 909 kyr ago and relative speed $2.012$9, but that velocity is high for standard planetary ejection and the star shows no evidence for the kind of binarity that would make such an ejection straightforward (Bailer-Jones et al., 2019). A plausible implication is that 2I/Borisov’s parent system may be unrecognized, too distant, or too old to recover with current six-dimensional stellar catalogs.

2. Dust coma, morphology, and photometric behavior

Early imaging immediately showed that 2I/Borisov was morphologically cometary. WHT and Gemini North observations in September 2019 revealed an extended coma and a faint, broad antisolar tail, with $e = 3.3790 \pm 0.020$0 mag and a normalized reflectance slope $e = 3.3790 \pm 0.020$1 per 100 nm, both compatible with ordinary comet dust; Monte Carlo modeling favored a differential size-distribution exponent $e = 3.3790 \pm 0.020$2 and ejection speed $e = 3.3790 \pm 0.020$3 for $e = 3.3790 \pm 0.020$4 grains (Guzik et al., 2019). A later follow-up imaging campaign in $e = 3.3790 \pm 0.020$5 band found stable large-scale morphology over five deep epochs, with average coma diameter $e = 3.3790 \pm 0.020$6 km and tail length $e = 3.3790 \pm 0.020$7 km, assuming anti-solar orientation (Mugrauer et al., 2020).

Multiband pre-perihelion monitoring from October to December 2019 reinforced the impression of an optically normal comet. Simultaneous $e = 3.3790 \pm 0.020$8 imaging gave average colors $e = 3.3790 \pm 0.020$9, $q = 2.0119 \pm 0.0044$0, $q = 2.0119 \pm 0.0044$1, while spectrophotometric imaging yielded $q = 2.0119 \pm 0.0044$2; the coma was described as homogeneous and reddish, with no obvious temporal color change (Prodan et al., 2024). In the same campaign, surface-brightness profiles gave a coma length scale of 8000–13000 km and a dust terminal-velocity law

$q = 2.0119 \pm 0.0044$3

implying terminal speeds of order $q = 2.0119 \pm 0.0044$4 for grains in the 200–250 nm range; under the adopted dust model, the mean pre-perihelion dust production was $q = 2.0119 \pm 0.0044$5, falling to a post-perihelion net mass loss of $q = 2.0119 \pm 0.0044$6 (Prodan et al., 2024).

Higher-resolution HST and ALMA analyses emphasized larger particles. Near perihelion, HST dust-dynamical modeling inferred ejection of particles larger than $q = 2.0119 \pm 0.0044$7m at speeds $q = 2.0119 \pm 0.0044$8, while ALMA continuum data near 0.85–2.15 mm were close to blackbody-like and were interpreted as requiring compact pebbles with radii $q = 2.0119 \pm 0.0044$9 mm; porous grains were found inconsistent with the observed millimeter spectral energy distribution (Kim et al., 2020, Yang et al., 2021). These results do not produce a single unique grain size, but they do show that the inferred dominant scale depends on the diagnostic: optical scattering, HST tail geometry, and millimeter thermal emission probe different parts of the dust population.

The inner coma was not perfectly isotropic. HST imaging between 2019 October and 2020 January showed persistent asymmetry that Kim et al. modeled as afternoon-peaked dust emission caused by thermal lag on a rotating nucleus, implying a preferred pole at RA $i = 44.004 \pm 0.041^\circ$0, Dec $i = 44.004 \pm 0.041^\circ$1, obliquity $i = 44.004 \pm 0.041^\circ$2, and subsolar latitude changing from $i = 44.004 \pm 0.041^\circ$3 near discovery to $i = 44.004 \pm 0.041^\circ$4 by January 2020 (Kim et al., 2020). A later 16-epoch MUSE campaign instead described the dust coma as predominantly smooth, with a persistent westward-to-northwest jet-like feature but no clear seasonal effect in the dust morphology; it also measured dust color in a 5000 km aperture typically between $i = 44.004 \pm 0.041^\circ$5 and $i = 44.004 \pm 0.041^\circ$6/1000 Å, with slight reddening after the March 2020 outburst/splitting event (Deam et al., 7 Jul 2025). Taken together, these studies agree on long-lived asymmetry but differ on how strongly seasonal illumination controlled the visible dust structure.

3. Nucleus size and physical constraints

The nucleus size of 2I/Borisov has been unusually difficult to determine because optical brightness was strongly contaminated by coma dust. Early characterization placed the nucleus at roughly $i = 44.004 \pm 0.041^\circ$7 km in radius from scaling arguments based on inferred water production and on $i = 44.004 \pm 0.041^\circ$8 cm relative to Hale-Bopp, but these estimates were explicitly described as very uncertain (Guzik et al., 2019). Subsequent work produced a broad range of upper limits and constraints.

The most stringent direct profile-based limit came from HST/WFC3 imaging on 2019 October 12. Convolution modeling of the inner coma profile showed that any larger unresolved central point source would overproduce the observed cusp, leading to a robust spherical-equivalent nucleus constraint $i = 44.004 \pm 0.041^\circ$9 km for assumed geometric albedo $v_\infty \sim 32\ {\rm km\,s^{-1}}$0 (Jewitt et al., 2019). The same paper used a momentum-balance argument for non-gravitational acceleration,

$v_\infty \sim 32\ {\rm km\,s^{-1}}$1

together with $v_\infty \sim 32\ {\rm km\,s^{-1}}$2, $v_\infty \sim 32\ {\rm km\,s^{-1}}$3, $v_\infty \sim 32\ {\rm km\,s^{-1}}$4, and $v_\infty \sim 32\ {\rm km\,s^{-1}}$5, to infer $v_\infty \sim 32\ {\rm km\,s^{-1}}$6 km (Jewitt et al., 2019). The overlap $v_\infty \sim 32\ {\rm km\,s^{-1}}$7 km was then used to argue that densities below $v_\infty \sim 32\ {\rm km\,s^{-1}}$8 are incompatible with both constraints, excluding an ultra-low-density fractal aggregate of the type proposed for 1I/‘Oumuamua.

Infrared photometry yielded a markedly larger scale. Gemini South FLAMINGOS-2 $v_\infty \sim 32\ {\rm km\,s^{-1}}$9-band imaging on 2019 November 30 measured $e = 3.3570 \pm 0.0006$0 mag inside a 6.8 arcsec aperture, and—using a reflected-light radius relation, a linear phase slope $e = 3.3570 \pm 0.0006$1, and adopted $e = 3.3570 \pm 0.0006$2-band albedo $e = 3.3570 \pm 0.0006$3—inferred $e = 3.3570 \pm 0.0006$4 km (Lee et al., 2019). That work argued that the near-IR image appeared morphologically much cleaner than optical images and therefore provided a tighter practical nucleus estimate than many earlier optical limits, but it also stated that no formal PSF-plus-coma decomposition, unresolved-flux fraction, or full systematic error budget was provided (Lee et al., 2019). The result is therefore best read as an infrared-based constraint under the assumption that residual $e = 3.3570 \pm 0.0006$5-band coma contamination was modest.

Other methods bracketed the same problem differently. Swift/UVOT water-production modeling constrained a minimum radius of 0.37 km and an active fraction of at least 55% of the surface (Xing et al., 2020). Near-perihelion Indian spectroscopy and imaging converted dust cross-section and sublimation arguments into a much broader interval $e = 3.3570 \pm 0.0006$6 km, explicitly separating an activity-based lower bound from a coma-contaminated photometric upper bound (Krishnakumar et al., 2021). The nucleus size question is therefore not represented by a single settled number in the literature; it is defined instead by method-dependent constraints whose differences reflect residual coma, adopted albedo, and whether the analysis isolates a true nuclear point source.

4. Volatiles, radicals, and coma chemistry

The first compositional result was strong CN with little else. Optical spectra from late September to mid-October 2019 detected CN at $e = 3.3570 \pm 0.0006$7–$e = 3.3570 \pm 0.0006$8 while giving $e = 3.3570 \pm 0.0006$9 limits $q = 2.00657 \pm 0.00008$0, $q = 2.00657 \pm 0.00008$1–$q = 2.00657 \pm 0.00008$2, and $q = 2.00657 \pm 0.00008$3 on 2019 Oct 2, implying $q = 2.00657 \pm 0.00008$4 and placing 2I/Borisov in the carbon-chain depleted class by the A’Hearn et al. criterion (Opitom et al., 2019). Low-resolution spectroscopy in late October and early November then detected C$q = 2.00657 \pm 0.00008$5 explicitly, with $q = 2.00657 \pm 0.00008$6, $q = 2.00657 \pm 0.00008$7, and $q = 2.00657 \pm 0.00008$8, showing that the comet was carbon-chain depleted but not carbon-chain empty (Lin et al., 2019).

By late November, the picture had become more nuanced. MUSE plus TRAPPIST observations detected clear C$q = 2.00657 \pm 0.00008$9, abundant NH$i = 44.0524 \pm 0.0004^\circ$0, and red CN, yielding $i = 44.0524 \pm 0.0004^\circ$1, $i = 44.0524 \pm 0.0004^\circ$2, and $i = 44.0524 \pm 0.0004^\circ$3 near $i = 44.0524 \pm 0.0004^\circ$4 au, so that $i = 44.0524 \pm 0.0004^\circ$5 and $i = 44.0524 \pm 0.0004^\circ$6 (Xing et al., 2020). The authors interpreted this as only barely carbon-chain depleted, in contrast to stronger earlier depletion, and as relatively NH$i = 44.0524 \pm 0.0004^\circ$7-rich compared with most Solar System comets (Xing et al., 2020). Around perihelion, Indian observatories measured $i = 44.0524 \pm 0.0004^\circ$8 pre-perihelion and $i = 44.0524 \pm 0.0004^\circ$9 post-perihelion, classifying the comet as moderately depleted and suggesting chemical heterogeneity in the active source regions (Krishnakumar et al., 2021). A later 126-day MUSE campaign returned to a more strongly depleted classification, with $2019\ {\rm Dec}\ 8.555 \pm 0.002$0 between 0.1 and 0.3 and average $2019\ {\rm Dec}\ 8.555 \pm 0.002$1, concluding that 2I/Borisov is carbon-depleted and relatively NH$2019\ {\rm Dec}\ 8.555 \pm 0.002$2-rich (Deam et al., 7 Jul 2025). The literature therefore records genuine epoch-dependent compositional evolution rather than a single static taxonomic state.

Water production was eventually measured directly enough to supersede early optical non-detections. Swift/UVOT obtained OH-based water production rates of $2019\ {\rm Dec}\ 8.555 \pm 0.002$3 on 2019 Nov 1, $2019\ {\rm Dec}\ 8.555 \pm 0.002$4 on Dec 1, and $2019\ {\rm Dec}\ 8.555 \pm 0.002$5 on Dec 21, with a post-perihelion decline described as faster than that of all previously observed comets (Xing et al., 2020). High-resolution UVES spectroscopy later measured OH lines near 309 nm and derived $2019\ {\rm Dec}\ 8.555 \pm 0.002$6 near Dec 24–26, while noting that the exact value depends strongly on whether a Haser or vectorial treatment is adopted (Opitom et al., 2021). These measurements mean that early reports or hypotheses centered on water non-detection were specific to observing epoch and sensitivity, not a durable absence of water.

The most distinctive volatile property was carbon monoxide. HST/COS ultraviolet spectroscopy detected several CO Fourth Positive bands in every epoch from 2019 December 11 to 2020 January 13, deriving $2019\ {\rm Dec}\ 8.555 \pm 0.002$7 to $2019\ {\rm Dec}\ 8.555 \pm 0.002$8 and concluding that the coma contained at least 173% as much CO as H$2019\ {\rm Dec}\ 8.555 \pm 0.002$9O, more than three times the previous inner-Solar-System record-holder inside 2.5 au (Bodewits et al., 2020). ALMA measured a lower pre-perihelion value, $2.012$00, corresponding to $2.012$01 near Dec 2–3; comparison with later values rising to $2.012$02 post-perihelion was then interpreted as evidence for nucleus heterogeneity and radial mixing of materials formed beyond different snow lines (Yang et al., 2021). UVES spectroscopy provided an independent oxygen-line diagnostic: the forbidden-line ratio

$2.012$03

was $2.012$04 near perihelion, unusually high relative to ordinary comets and consistent with a high CO/H$2.012$05O abundance ratio (Opitom et al., 2021).

Several higher-order compositional diagnostics looked comparatively normal. UVES derived $2.012$06, implying $2.012$07 and $2.012$08, all described as consistent with typical Solar System comet values (Opitom et al., 2021). The same study found $2.012$09, in agreement with the Solar System comet average within uncertainties (Opitom et al., 2021). Independently, X-shooter spectroscopy detected nine Ni I lines in the cold coma at $2.012$10 au, inferred a short-lived nickel-bearing parent with lifetime $2.012$11 at 1 au, and derived a nickel production rate $2.012$12 atoms s$2.012$13, equal to 0.002% of OH and 0.3% of CN (Guzik et al., 2021). These measurements support a recurrent conclusion of the literature: 2I/Borisov was chemically distinctive mainly in its volatile balance, especially CO richness, rather than in every measurable cometary diagnostic.

5. Outburst, splitting, and rotational or seasonal evolution

2I/Borisov did not remain quiescent after perihelion. Ground-based observations reported an outburst between 2020 March 4 and 9, and HST follow-up showed that the event increased the 15,000 km-aperture brightness by $2.012$14 mag, corresponding to an excess cross-section $2.012$15 and, for 0.1 mm grains of density $2.012$16, a dust mass $2.012$17 kg (Jewitt et al., 2020). This was only about $2.012$18 of the mass of a $2.012$19 m nucleus, so the outburst was dynamically noticeable but globally minor (Jewitt et al., 2020).

The same HST program detected a transient “double nucleus” on 2020 March 30: a secondary condensation separated by $2.012$20, or 230 km, from the primary, with cross-section about $2.012$21 and inferred dust mass about $2.012$22 kg (Jewitt et al., 2020). It was absent on and before March 28 and on and after April 03. Jewitt et al. argued that the delayed appearance and rapid disappearance are inconsistent with simple ballistic drift of a visible fragment from the March outburst; instead they favored rotational bursting of one or more meter-sized boulders after ejection from the main nucleus, driven by outgassing torques, with the bright secondary being an unresolved cloud of freshly generated debris (Jewitt et al., 2020). The broader conclusion was that 2I/Borisov survived the event largely unscathed.

Rotational evolution had been anticipated even before the outburst. The HST nucleus study showed that for $2.012$23 km, spin-up timescales from outgassing torques are comparable to or shorter than the time Borisov spent inside the Sun’s water-sublimation zone, implying that the spin angular momentum should change significantly during the solar fly-by (Jewitt et al., 2019). Kim et al.’s thermal-lag model added a seasonal framework, with illumination shifting from southern solstice near discovery toward equinox by January 2020 and with subsequent activity plausibly linked to freshly illuminated northern terrain (Kim et al., 2020). A later MUSE analysis of the March 2020 event found that CN and C$2.012$24 rose after the outburst but with large uncertainties, whereas NH$2.012$25 increased significantly above the 95% prediction interval and the dust reddened slightly; no new gas-morphology feature was detected (Deam et al., 7 Jul 2025). These results jointly place the outburst in a context of a small, active nucleus susceptible to both torque-driven and illumination-driven evolution.

6. Interpretation, controversies, and comparative significance

The scientific importance of 2I/Borisov lies in the conjunction of familiarity and difference. Early optical imaging already emphasized that, aside from its orbit, the comet looked “indistinguishable from the native Solar System comets,” with ordinary coma morphology, ordinary dust colors, and plausible kilometer-to-subkilometer scale (Guzik et al., 2019). Subsequent spectroscopy strengthened that view for several diagnostics: NH$2.012$26/NH$2.012$27 nuclear-spin properties, Ni/Fe behavior, the presence of common daughter radicals, and the general structure of dust and gas activity all have close Solar System analogues (Opitom et al., 2021, Guzik et al., 2021). At the same time, the ultraviolet CO measurements and derived elemental ratios made clear that the comet’s volatile inventory is not simply a Solar System duplicate. In particular, the coma’s CO/H$2.012$28O abundance reached at least 173%, and the volatile $2.012$29 ratio was reported as $2.012$30, nearly six times the average reported for measured Solar System comets (Bodewits et al., 2020).

Several apparent controversies in the literature are best understood as consequences of evolving activity and method dependence rather than of irreconcilable contradiction. Strong early C$2.012$31 depletion gave way to later detections and then to perihelion and post-perihelion values spanning moderate to strong depletion, indicating that the ratio $2.012$32 changed with time (Opitom et al., 2019, Xing et al., 2020, Deam et al., 7 Jul 2025). Early non-detections of OH or claims of water poverty were superseded by later OH and H$2.012$33O detections, so the water budget proved to be observationally accessible but temporally variable (Xing et al., 2020, Opitom et al., 2021). Nucleus-size estimates ranged from sub-0.5 km HST limits to an approximate 1.5 km infrared reflected-light radius because different analyses either modeled the inner coma explicitly or assumed that longer-wavelength morphology sufficiently suppressed coma contamination (Jewitt et al., 2019, Lee et al., 2019). These are differences of technique and assumption, not merely disagreements of arithmetic.

A more speculative line of interpretation proposed that 2I/Borisov could be a “stardust comet,” or post-main-sequence object condensed from AGB ejecta rather than a conventional planetesimal, with specific proposed tests involving persistent water poverty, unusual CNO isotopes, SiC emission near $2.012$34, $2.012$35 or $2.012$36 features, PAH bands at 3.28 and $2.012$37, and possible Na or Li enhancement (Eubanks, 2019). That hypothesis was explicitly predictive and recognized as speculative even in its original formulation. Later compositional work did not yield isotopic ratios, because the relevant bands were too weak, but the accumulating evidence for ordinary comet-like NH$2.012$38 OPR, ordinary Ni/Fe, measured water production, and otherwise familiar coma composition except for high CO/H$2.012$39O shifted the balance of evidence toward an object better described as a volatile-rich comet from another planetary system than as a chemically exotic outlier (Opitom et al., 2021).

The most durable synthesis in the literature is therefore that 2I/Borisov was an active interstellar comet whose morphology, dust properties, and many spectroscopic diagnostics overlap strongly with Solar System comet phenomenology, while its volatile ratios—especially the prominence of CO relative to H$2.012$40O—indicate formation or preservation conditions uncommon among local comets (Bodewits et al., 2020, Deam et al., 7 Jul 2025). In that sense, 2I/Borisov is not chiefly important for being radically unfamiliar. It is important because it demonstrated that extrasolar planetesimals can be recognizably cometary in the ordinary sense, yet still encode a different protoplanetary disk chemistry and thermal history.