VLT observations of interstellar comet 3I/ATLAS (catalog ) II. From quiescence to glow: Dramatic rise of Ni i emission and incipient CN outgassing at large heliocentric distancesFacilities: ESO:VLT/XSHOOTER, VLT/UVES.

Characterizing the volatile composition of interstellar objects (ISOs) passing through the Solar System provides a unique window into the chemical and physical processes operating in distant stellar systems [53, e.g.]. ISOs preserve signatures of both nebular formation conditions and primordial interstellar chemistry that potentially predate Solar system formation. When heated by solar radiation, cometary ISOs release solids and gas due to “cometary activity” [76, e.g.].

The discovery of 1I/2017 U1 (‘Oumuamua, hereafter 1I) on October 19th, 2017 marked the first confirmed detection of an ISO passing through our Solar System. With a hyperbolic orbit (eccentricity $e\!\approx\!1.2$) and an unusual elongated shape (aspect ratio $>$5:1), 1I displayed several peculiar characteristics [62]. Despite intensive observational campaigns, no coma or outgassing was directly detected around 1I, with upper limits placed on the production rates of CN, C₂, and C₃ [96]. However, non-gravitational acceleration was observed in its trajectory, suggesting that some outgassing was occurring below the detection threshold [63]. The lack of observable volatiles, combined with its discovery after perihelion when the object was faint and already receding from the Sun, limited our ability to characterize its composition [34, 76].

The second confirmed ISO, 2I/Borisov (hereafter 2I), was discovered on August 30th, 2019. Unlike 1I, this object was discovered before perihelion ($q\!=\!2.0$ au; December 8th, 2019) and exhibited a significant dust coma [35]. Initial photometry found broad-band optical colors similar to other active comets and the discovery of this active interstellar comet allowed spectroscopic investigations of the sublimated ices. Fitzsimmons et al. [35] reported the first detection of gas emitted via the near-UV emission of CN (B-X) at a heliocentric distance of $r\!=\!2.7$ au, with a production rate of $Q(\mathrm{CN})\!=\!(3.7\!\pm\!0.4)\times 10^{24}$ s^-1. Subsequent observations revealed additional molecular species including C₂ [56], [Oi] [60] and OH [93], NH₂ [8], and exceptionally high abundances of CO [14, 22]. Perhaps most surprisingly, Guzik & Drahus [44] and Opitom et al. [74] detected atomic nickel vapor and in addition Opitom et al. [74] detected iron vapor in 2I, species that had just recently be discovered to be ubiquitous in the comae of Solar System comets [58], and even far from the Sun. Overall, these observations demonstrated that the gas, dust, and nuclear properties of the first active interstellar object were broadly similar to those of Solar System comets, though with some distinctive compositional characteristics.

Cometary activity is primarily driven by the sublimation of major volatiles – H₂O, CO₂, and CO – from the solar-heated nucleus. While numerous other species such as CH₄, C₃H₆, HCN, NH₃, CH₃OH, and CH₄CN are detected in cometary comae, these three dominant volatiles account for the bulk of the gas production that generates observable activity [11]. Water ice is a major component of comets and one that dominates cometary activity and physical evolution, but CO and CO₂ appear to be the major “super volatiles” capable of driving activity at great solar distances as seen in the case of Hale-Bopp [23, 46] and transient brightening of 29P/Schwassman-Wachmann 1 [43]. The relative abundances of different molecular species can reveal information about formation temperature, radiation processing history, and storage conditions in the outer regions of planetary systems and interstellar space.

Spectroscopic observations of comets reveal a diverse array of volatile species with different sublimation temperatures, providing insights into their formation conditions and thermal histories [61, 68, 13, 3, 90, 12]. Highly volatile species such as CO and CO₂ can sublimate at temperatures as low as 25-80 K, driving activity at large heliocentric distances beyond 5 au [1]. At intermediate distances (2-4 au), volatile species including HCN, CH₃OH, and H₂CO (which sublimate at temperatures of 90-100 K) become detectable in cometary comae. Closer to the Sun, at $\lesssim$ 2 au, water ice dominates the activity as temperatures exceed 150 K. Until recently, metallic species such as iron and nickel were thought to be released only under extreme conditions, as they are typically bound in refractory minerals with high sublimation temperatures of $>$700 K [77]. However, the unexpected detection of atomic nickel and iron in the cold comae of Solar System comets at heliocentric distances up to 3.5 au has challenged this understanding, suggesting the existence of previously unrecognized metal-bearing molecules with low sublimation temperatures [58, 16]. This diverse inventory of molecular and atomic species in cometary comae provides crucial chemical fingerprints for understanding their formation environments and subsequent evolution.

On July 1st, 2025, the Asteroid Terrestrial-impact Last Alert System [88, ATLAS;] discovered the third ISO, 3I/ATLAS (catalog )¹¹ 1 https://minorplanetcenter.net/mpec/K25/K25N51.html, at a heliocentric distance of 4.5 au on an incoming, strongly hyperbolic trajectory having an eccentricity of $e\!\approx\!6.14$ [26, 15, 84, 48]. The discovery occurred shortly before the beginning of the Vera C. Rubin Observatory’s Legacy Survey of Space and Time [51, LSST;], proving both timely and unexpected, since previous studies have suggested a low ISO occurrence rate [67, 30, 27, e.g.].

Initial observations revealed a developing coma and tail, confirming the cometary nature of 3I/ATLAS (catalog )  [75, 18, e.g.]. Preliminary spectroscopic studies have shown a red reflectance spectrum [84, 75, 78], possible evidence for OH emission [94] and the presence of water ice [95]. With an upper limit on its diameter of D$\lesssim\!5.6$ km [52], 3I/ATLAS (catalog ) could potentially be larger than the previous ISOs, though its exact size remains uncertain.

What makes 3I/ATLAS (catalog ) particularly valuable for scientific study is the combination of its early discovery and favorable orbital parameters. Unlike previous ISOs, it was detected several months before perihelion, allowing for a comprehensive monitoring campaign to track the evolution of its activity as it approaches perihelion on October 29th, 2025 at $q=1.356$ au²² 2 JPL#25, data arc: 2025-05-22 to 2025-08-12. This extended observational timeline provides an unprecedented opportunity for detailed characterization of an extrasolar minor body throughout its journey in the inner Solar System.

Here we report the first detections of the onset of both atomic nickel (Ni i) and CN gas emission in spectroscopic observations of 3I/ATLAS (catalog ) obtained at the ESO Very Large Telescope (VLT) on Cerro Paranal in Chile. The observations span heliocentric distances from r${}_{h}=4.4$ au to $2.85$ au. For our latest observation on 2025-08-21, we derive production rates of $\log~Q(\mathrm{CN})=23.61\pm 0.05$ molecules s^-1 and $\log~Q$(Ni i) $=22.67\pm 0.07$ atoms s^-1, and examine their evolution as the comet approaches perihelion. The detection of both molecular and atomic species in 3I/ATLAS (catalog ) provides important constraints on its volatile composition and enables direct comparison with both 2I/Borisov and Solar System comets, offering new insights into the universality of cometary material across different planetary systems.

In the following, we provide description of the methods used to obtain the X-shooter products analyzed in this work. We focus here on UVB-arm spectroscopy, while a complete reduction and analysis of all three arms will be presented in Paper I (Puzia et al., in prep.).

We designed the monitoring program for a cadence of $\sim$3-5 days, subject to scheduling constraints from VLTi-UT runs, technical time slots, moon separation, weather conditions, and synchronization with other VLT instruments observing 3I/ATLAS (catalog ). We observed with the 1.6″-wide slit, oriented at the parallactic angle, providing a spectral resolution $R\!\simeq\!3200$. We did not employ nodding, but used small dithers along the 11″ long slit to mitigate detector artifacts.

We reduced the raw frames with the ESO X-shooter pipeline [65, v3.8.3;] within the esoreflex environment [39]. Standard calibration steps included bias subtraction, flat-fielding, arc-lamp wavelength calibration, 2D rectification, order merging, and flux calibration. The pipeline delivered flux-calibrated, rectified 2D slit frames. In several epochs, 3I/ATLAS (catalog ) crossed crowded stellar fields toward the Galactic bulge, complicating extraction. To address this, we monitored the sky traces and spatial profiles of each exposure and developed a custom pipeline for PSF tracing and extraction. The routine adaptively bins the spectrum along the wavelength axis, fits a Moffat profile to the spatial cross-section in each bin, ensuring convergence, and models the wavelength dependence of the PSF center and FWHM. We then use this model to extract 1D spectra within a fixed aperture (in units of FWHM), minimizing contamination from field stars and instrumental artifacts, while fully propagating the variance spectra. Full details of the algorithm are provided in Paper I of this series.

We subtracted the residual background using the slit itself (11″ long), which provided sufficient uncontaminated sky area, although the varying sky background occasionally introduced spatial systematics in the wavelength solution.

From the extracted 1D spectra of each exposure, we discarded frames contaminated by field stars, identified through slit profiles and finding charts. We then averaged the remaining spectra by visit (total integration times 1800-3600 s) to produce a time series of 3I/ATLAS (catalog ) spectra (see Fig. 1). Table 1 summarizes the observations, observing conditions, and heliocentric distance of 3I/ATLAS (catalog ) at the time of observation. Visit X2 was strongly affected by field star contamination in all exposures. To recover this epoch, we applied a narrower extraction aperture (of $1\times$FWHM, instead of the regular $1.5\times$FWHM); however, the continuum slope from this visit remains systematically biased.

We obtained observations of solar analogs in short single exposures and reduced them in the same way. The first two visits lacked dedicated solar analogs and were paired with the analog from Visit X5, which was acquired under similar conditions. All the subsequent visits included one or two solar analog stars observed close in time to the target, among which HD 150469 (catalog ) was always one of them.

The high S/N of the solar analog spectra and the low covariance among fitted parameters render continuum model uncertainties negligible compared to the noise of the 3I/ATLAS (catalog ) spectra. Thus, the propagated errors in the continuum-subtracted spectra are dominated by the statistical uncertainties of the science frames. This continuum removal enables clear identification and measurement of gas and refractory emission features in the coma, which we analyze in the following sections.

We obtained spectra of 3I/ATLAS (catalog ) on 2025 August 11 and 15 UT with VLT/UVES [25], using two blue settings chosen to span the OH near-UV band ($\sim$3100 Å) through the CN violet band ($\sim$3880 Å). On the first night, we employed the 346 nm standard setting with a 1.8″ slit, covering 3030–3880 Å and achieving a resolving power of $R\!\simeq\!35{,}000$ while maximizing throughput. On the second night, we used the 390 nm setting with a 0.6″ slit, covering 3260–4540 Å and reaching $R\!\simeq\!80{,}000$ on the CN band. For each setting, we obtained three consecutive 1600 s exposures of the faint target and averaged them by night.

Table 1 lists the UVES observing dates, set-ups, and observational circumstances. We reduced the raw frames with the ESO UVES pipeline (v6.5.3; Ballester et al. 7), including wavelength calibration, extinction correction, and flux calibration using master response curves. We then applied custom routines for spectrum extraction over the full slit length and for cosmic-ray removal. The spectra were corrected for the Doppler shift due to the comet’s topocentric velocity. Finally, we removed the dust continuum using the Kurucz solar spectrum³³ 3 http://kurucz.harvard.edu/sun.html, resampled to match the resolution of each setting. More details concerning the reduction methodology can be found in Manfroid et al. [59].

In both X-shooter and UVES 3I/ATLAS (catalog ) spectra, we identified both atomic and molecular emission features above the underlying continuum. In X-shooter data, to isolate these emission features, we removed the continuum by subtracting a scaled and reddened solar reference spectrum. This process revealed several emission features, most notably multiple Ni i atomic lines in the 3400-3600 Å region (beginning in X4 on 2025-07-20) and the CN (B-X) band around 3880 Å with detection since X9 2025-08-14.

For both the molecular and atomic species detected in our spectra, we followed established methodologies to convert measured emission line fluxes, from the continuum-subtracted spectra, to gas production rates [58]. To extract line fluxes, for Ni i lines we fit individual Gaussian profiles to each identified atomic line, while for CN, we approximated the asymmetric band shape with two free-fitted Gaussians in X-shooter data. We adopted an outflow velocity of $v=0.85\times r_{h}^{-0.5}$ km s^-1 [20] for all calculations. Due to our rectangular spectroscopic aperture ($1.6\arcsec\times 11\arcsec$), we numerically integrated the spatial distribution model within the slit to derive the corresponding production rates.

The CN(0-0) emission band at 3880 Å was first detected when the comet was at a heliocentric distance (r _h) of 3.07 au in visit X9 on 2025-08-14 and also in U2 the following night on 2025-08-15. The right panel of Figure 1 shows the continuum-subtracted spectrum in the 3860-3930 Å region, highlighting the characteristic asymmetric profile that confirms the identification as CN. In addition, individual lines of CN are detected in higher resolution UVES observations, adding to the confidence of the CN detection.

Following the methodology described in Manfroid et al. [58], we derived CN production rates using a Haser model. We used the fluorescence efficiencies from Schleicher [82] for the B${}^{2}\Sigma^{+}-$X${}^{2}\Sigma^{+}$ (0,0) band near 3880 Å, adjusted for our observing circumstances (heliocentric velocity of $-55.03$ km s^-1 at r${}_{h}\!=\!3.07$ au). The parent scale length of $1.3\times 10^{4}$ km and daughter scale length of $2.1\times 10^{5}$ km scaled $r^{-2}$ at 1 au are adopted from [2]. The CN radical has a photodissociation lifetime that results in these characteristic scale lengths in the cometary coma. The profile distribution was integrated over the area of each spectrograph slits (X-shooter: $1.6\arcsec\times 11\arcsec$, UVES:$0.6\arcsec\times 8\arcsec$) and compared to the observed flux to yield the production rate. For X-shooter X9 taken on (2025-08-14), the measured CN band flux of $(4.45\pm 0.23)\times 10^{-16}\,\mathrm{erg}\,\mathrm{s}^{-1}\,\mathrm{cm}^{-2}$ yielded a CN production rate of $\mathrm{Log}~Q(\mathrm{CN})=23.33\pm 0.10$ molecules s^-1. Table 2 lists the production rates for all visits where CN was detected and upper limits.

We identified 22 emission lines corresponding to Ni i transitions in our 3I/ATLAS (catalog ) spectra, with the strongest ones appearing in visit X4 on 2025-07-20. The middle panel of Figure 1 shows the continuum-subtracted spectrum in the 3350-3630 Å region with identified Ni i lines marked. The strongest lines were detected at 3414.8 Å and 3524.5 Å, consistent with transitions previously observed in 2I/Borisov [44] and Solar System comets [58].

Following the methodology of Manfroid et al. [58], we computed theoretical intensities for all Ni i lines using a multilevel fluorescence model that accounts for the detailed structure of the solar spectrum. For each visit, we compared these theoretical intensities with the observed line intensities to derive column densities. This approach, fully detailed in Manfroid et al. [58], provides a robust determination of atomic nickel abundances without relying on g-factor approximations.

For the strongest line at 3524.5 Å in X11, we measured a flux of 1.82 $\pm$ 0.02 $\times 10^{-15}$ ergs s^-1 cm^-2, using the full fluorescence model. The spatial distribution was modeled following Manfroid et al. [58], who demonstrated that metallic lines show a radiance inversely proportional to the projected distance to the nucleus, corresponding to a parent species with a short lifetime. This yielded a nickel production rate of $\log~Q$(Ni) $=22.67\pm 0.07$ atoms s^-1 for X11. Table 2 summarizes the derived production rates for all visits where Ni i was detected.

To characterize the dust in the coma of 3I/ATLAS (catalog ) in our X-shooter observations, we refer back to the reflectance model described in Sect. II.1. Although we fit a second-order polynomial for the relative reflectance, the resulting slope is nearly linear. Between 3900 and 5550 Å, the dust reflectance exhibits a reddened spectrum with a color slope of $(22\pm 1)$%/1000 Å, which remained constant across all visits within this uncertainty. At shorter wavelengths (3200–3900 Å), the average slope is consistent with this value, but individual visits show higher scatter that we attribute primarily to the higher noise floor in this spectral region.

Our spectroscopic monitoring of 3I/ATLAS (catalog ) provides a unique opportunity to track the evolution of gas emissions as the comet warms toward perihelion [78, see thermal model in]. Figure 2 shows the derived production rates of Ni plotted against heliocentric distance. Ni i emission was first detected in visit X4 (on 2025-07-20) at a heliocentric distance of r${}_{h}=3.88$ au, with a production rate of $\log\,Q($Ni$)=21.26\pm 0.30$ atoms s^-1. As 3I/ATLAS (catalog ) approaches perihelion (2025-10-29), we observe a rapid and steady increase in nickel production, reaching $\log\,Q($Ni$)=22.67\pm 0.07$ atoms s^-1 by visit X11 at r${}_{h}=2.85$ au. Notably, the relative strengths of different Ni i lines remained consistent throughout our observing campaign, so far, suggesting that the excitation conditions and fluorescence mechanism did not change significantly over this period.

The Ni production rate follows a power-law dependence on heliocentric distance of the form Q(Ni) $\propto r_{h}^{-8.43}$, which is far steeper than any purely radiative scaling. The equilibrium surface temperature of an airless body scales as $T\propto r_{h}^{-0.5}$, and between $r_{h}\simeq$ 3.9 and 2.85 au this raises $T$ by $\sim 15-20\%$. Photon-driven rates, e.g. via UV desorption or resonance fluorescence per parent, scale at most with the solar flux, i.e. $\Phi_{\odot}\propto r_{h}^{-2}$. Our 3I/ATLAS (catalog ) measurements yield a scaling far steeper than both $r_{h}^{-0.5}$ (temperature) and $r_{h}^{-2}$ (insolation), implying that the supply of Ni atoms to the coma is growing super-linearly as the object approaches the Sun. This in turn, indicates that the rate-limiting step is not the availability of photons, nor a linear response to temperature increase, but may instead likely be a temperature-activated and/or threshold process that is being switched on.

If the parent of Ni is released by a process with the Arrhenius form $Q\propto\exp(-E_{a}/kT)$ and $T\propto r_{h}^{-0.5}$, then

where $E_{a}$ is the activation energy and $k$ the Boltzmann constant. If over a limited range $Q(r_{h})$ behaves like a power law then with the previous result we obtain

At moderate temperatures $T\!\simeq\!150\!-\!200$ K and a measured slope $n\simeq 8.43$, we estimate an activation energy of $E_{a}\!=\!2kTn\approx 0.22\!-\!0.29$ eV or $21.02-28.02$ kJ mol^-1, which is typical for desorption or diffusion barriers for weakly bound species on ices or organics [40, e.g.], and crucially does not imply metal sublimation [32, e.g.]. This points to Ni-bearing organometallics [58, 50, 44, metal carbonyls, see e.g.] or grain-surface complexes (metal-PAHs) as candidates [54, 80, 16], rather than direct sputtering or simple photon-stimulated desorption [45, 24, e.g.].

If Ni i  is predominantly released from refractory grains via thermal/UV processes in the coma, then once volatile outgassing (CO₂/H₂O) exceeds the cohesion threshold, the lifted dust mass can rise very steeply with decreasing $r_{h}$ [41, 42, e.g.]. Other interesting potential mechanisms are the temperature crossing of the amorphous-to-crystalline H₂O transition at $T\!\simeq\!120\!-\!150$ K, polymerization of ices, and/or devolatilization of surface organics which can trigger both gas releases and fragmentation, effectively increasing the active area and the stock of Ni-bearing carriers [9, 64, 92, 81, 31, e.g]. Furthermore, changing sub-solar latitude or erosion can expose fresh terrains enriched in the Ni carrier, compounding any thermal activation.

The power-law index $n\!\simeq\!8.43$ is steeper than typically observed for volatile species in Solar System comets [2, 55, 69, 70, 21], which mostly show $r_{h}^{-1}$ to $r_{h}^{-6}$ rate dependencies, albeit at shorter heliocentric distances. We note that inbound slopes are typically steeper than post-perihelion slopes. The steeper dependence observed for nickel may indicate a distinct production onset mechanism, possibly related to its hypothesized origin from a short-lived parent molecule such as nickel carbonyl. Future monitoring observations are advised to test for super-linearity correlations with dust tracers (continuum, Na, etc.) or with Q(H₂O), Q(CO₂) and Q(CO), where a strong correlation would support dust-entrainment and/or phase-transition drivers.

3I/ATLAS (catalog )’s CN emission was first measured in our X-shooter spectra on 2025-08-14 (visit X9) at a heliocentric distance of 3.07 au, with a production rate of $\log\,Q(\mathrm{CN})=23.33\pm 0.10$ molecules s^-1. Figure 3 shows the derived production rates of CN plotted against heliocentric distance. The later appearance of CN compared to Ni i, suggests different thermal thresholds for the release of their respective parent molecules, with the putative nickel parent Ni(CO)₄ sublimating at lower temperatures than HCN [38, $\lesssim$ 76-79 K, see e.g.], the likely parent of CN [37]. We recognize that other effects might also be at play, such as differences in detection sensitivity, relative abundances, and stratification of parent molecules at different depths within the nucleus.

With only three X-shooter visits showing detectable CN emission, we cannot robustly constrain the heliocentric distance dependence of the Q(CN) production rate. However, the available data suggests a power-law index of approximately $r_{h}^{-9.38\pm 1.20}$, which is steep within the range observed for Solar System comets ($-10$ to $-1.5$, see A’Hearn et al. 2).

We note that small discrepancies between X-shooter and UVES measurements are expected given the use of solar-analog calibration in the former [85] and the limited spectral resolution of X-shooter in resolving the CN band. Nevertheless, the X-shooter results remain self-consistent across visits, as all reductions relied on the same solar-analog reference star observed under similar conditions to 3I/ATLAS (catalog ) (see Tab. 1).

The detection of atomic nickel and CN in 3I/ATLAS allows us to place this ISO [86] in context with both 2I and the diverse population of Solar System comets. Figure 2 presents our nickel production rate measurements alongside those from other comets, revealing distinct patterns across different comet populations. The Ni production rates we measure for 3I/ATLAS ranging from $\log\,Q($Ni$)=21.26\pm 0.30$ atoms s^-1 at 3.8 au to $\log\,Q($Ni$)=22.67\pm 0.07$ atoms s^-1 at 2.85 au extend the heliocentric distance range for nickel detections in comets, providing new constraints on production mechanisms. For context, scaling our measurement at 2.85 au using a canonical $r_{h}^{-2}$ relationship would yield $Q$(Ni)$|_{1\,\mathrm{au}}\approx 3.80\times 10^{23}$ s^-1, already surpassing values reported for Solar System benchmarks such as notably bright comet C/1996 B2 (Hyakutake) with $Q$(Ni)$\simeq 10^{22}$–$10^{23}$ s^-1 [50, 16].

Figure 2 reveals clear clustering patterns among different comet populations. External and Oort cloud comets (OCCs) show higher Ni production rates ($10^{21.5}$-$10^{23.1}$ atoms s^-1) compared to Jupiter-family comets (JFCs) ($10^{20.8}$-$10^{22.3}$ atoms s^-1) at similar heliocentric distances. These are based on data from Manfroid et al. [58], who demonstrated that Jupiter-family comets display larger variance in Ni i/Fe i  ratios compared to Oort cloud comets. Our analysis of their data yields an F-statistic of 4.3 with a p-value of approximately 7%, indicating that Jupiter-family comets show significantly more heterogeneity in their Ni i/Fe i  ratios compared to Oort cloud comets. This may suggest that JFCs formed under more diverse conditions in the outer solar nebula. However, it should be noted that OCCs also tend to show higher activity levels in general, not restricted to Ni, potentially due to a reduction of active surface area as comets become more processed. In addition, observational biases (e.g., the smaller number of JFCs observed at larger heliocentric distances) can complicate direct comparisons between the two families. With these caveats in mind, 3I/ATLAS’s relatively high nickel production rates at our closest approach (2.85 au) are more consistent with external and Oort cloud comet levels, providing potential clues to their formation environment and/or evolutionary history.

Compared to 2I, which showed $Q(\mathrm{Ni})=(0.9\pm 0.3)\times 10^{22}$ atoms s^-1 at 2.3 au [44], 3I/ATLAS exhibits higher nickel production at comparable distances. The average Ni i/Fe i  ratio in Solar System comets is approximately 15.5 times higher than solar values [58]. This preferential nickel enhancement in cometary volatile reservoirs may extend to ISOs as well. However, since no Fe i lines were detected in our X-shooter and UVES data, additional observations of 3I/ATLAS are needed to determine its Ni i/Fe i  ratio and make a direct comparison.

The sequential detection of Ni i (3.9 au) and CN (3.1 au) in 3I/ATLAS (catalog ) differs from the reported observations of 2I/Borisov, where CN was detected at 2.7 au [35] and metals at 2.3 au [44]. While these detection sequences might suggest compositional differences, they could also reflect observational biases such as sensitivity or different observing circumstances (e.g, pre/post-perihelion positioning). Nevertheless, the detection of both species in ISOs strengthens the broader hypothesis that metal atoms in cometary comae may originate from molecules with low sublimation temperatures, possibly metal carbonyls [58, 16].

Our CN detection in visit X11 on 2025-08-21 with $Q(\mathrm{CN})=4.1\times 10^{23}$ s^-1 at 2.85 au can be compared with 2I’s $Q(\mathrm{CN})=3.7\pm 0.4\times 10^{24}$ s^-1 at 2.7 au [35]. Among Solar System comets, CN production spans a wide range from $10^{23}$ to $10^{27}$ molecules s^-1 depending on comet size and heliocentric distance [2, 33, 19]. Representative examples include C/2013 A1 (Siding Spring) with $Q(\mathrm{CN})=1.4\times 10^{25}$ molecules s^-1 at $r_{h}=1.40$ au [72], and 67P/Churyumov-Gerasimenko with peak production of $Q(\mathrm{CN})=7.5\pm 0.9\times 10^{24}$ molecules s^-1 approximately two weeks after perihelion [71].

The detection of similar metal species and comparable production rates in both ISOs observed to date suggests that there may be commonalities in the processes governing proto-planetary disk chemistry across different stellar systems. With an estimated age of older than the Solar System [48, 86], 3I/ATLAS provides a window into organometallic chemistry from a significantly different and likely lower-metallicity epoch of galactic evolution compared to our 4.6 billion-year-old environment, potentially revealing how these processes operate in planetary systems formed much earlier than our own.

Detection of Ni i emission in a cold coma with absent Fe i imply short-lived, low-temperature Ni i-bearing parent molecules that photodissociate close to the nucleus. Such detections were previously made in 2I near $r_{\rm h}\!\approx\!2.3$ au [44]. Modeling emission profiles requires a parent lifetime of a few $\times 10^{2}$-$10^{3}$ s at 1 au and a Ni i  production rate $\sim 10^{22}$ s^-1 ($\sim$$2\times 10^{-3}$ of OH), i.e. a species volatile enough to be released at $\lesssim 200$ K and photolabile enough to yield atoms near their source [58, 50, e.g.]. The absence of Fe i  under similar conditions demands either i) a Ni-specific parent and/or ii) a release/processing pathway that favors Ni over Fe at low $T$, modulo differences in fluorescence efficiency.