© The Authors 2024

1 Introduction

Comets are the most pristine remnants of the formation of the Solar System 4.6 billion years ago. They sample some of the oldest and most primitive material in the Solar System, including ices, and are thus our best window on the volatile composition of the solar proto-planetary disk. Comets may also have played a role in the delivery of water and organic material to the early Earth (see Hartogh et al. 2011, and references therein). The latest simulations of the early Solar System’s evolution (Brasser & Morbidelli 2013; O’Brien et al. 2014) suggest a more complex scenario. On the one hand, ice-rich bodies that formed beyond Jupiter may have been implanted early in the outer asteroid belt and participated in the supply of water to the Earth. On the other hand, current comets are coming from either the Oort Cloud or the scattered disk of the Kuiper belt and may have formed in the same trans-Neptunian region, sampling the same diversity of formation conditions. Understanding the diversity in composition and isotopic ratios of cometary materials is thus essential for assessing such scenarios (Altwegg & Bockelée-Morvan 2003; Bockelée-Morvan et al. 2015).

Comet C/2021 A1 (Leonard) is a long-period dynamically old Oort-Cloud comet (OCC, with an initial semi-major axis of 2000 au and an inclination of 132.7°) that reached perihelion at a heliocentric distance r_h=0.615 au on 3.3 January 2022 UT. It came as close as 0.234 au to the Earth on 12.6 December 2021. It was anticipated to become a bright comet visible to the naked eye, but it under-performed until 14 December when it was in solar conjunction. Then monitoring of the outgassing rate via observations of the OH radical (Crovisier et al. 2021) and visual magnitudes showed strong outbursts of activity, repeating on a ~5 day period from 15 December to 7 January 2022¹. Then, it developed spectacular ion and dust tails while being mostly observable from the southern hemisphere. Later on, the comet faded rapidly, and images taken in April–May 2022 suggest that it was disintegrating during this outbursting phase. The derived pre-disintegration nucleus radius was 0.6 ± 0.2 km (Jewitt et al. 2023). We observed comet C/2021 A1 with the Institut de radioastronomie millimétrique (IRAM) 30-m telescope briefly between 12.0 and 16.4 November, and extensively between 8.4 and 13.4 December 2021 UT, before it reached declinations that are too low for the northern observatories.

Comet C/2022 E3 (ZTF) is also a long-period dynamically old OCC (with an initial semi-major axis of 1400 au and an inclination of 109.2°). It reached perihelion at r_h = 1.112 au on 12.8 January 2023 UT. It was closest to the Earth on 1.8 February 2023 at 0.284 au, and reached naked eye visibility (total magnitude of 4.8). It attracted public attention and was the target of a worldwide campaign because it was discovered 11 months ahead of its peak brightness, expected to happen at the end of January 2023 when the comet was circumpolar for the northern hemisphere and visible to the naked eye all night². The comet was further advertised in NASA news releases at the end of 2021³. We observed comet C/2022 E3 with the Odin space telescope from 19.3 to 20.3 January 2023 (Biver et al. 2023b) and with the IRAM-30m radio telescope between 3.7 and 7.1 February. Due to its high activity and a favourable apparition, this comet was the focus of an international observing campaign, from the radio – OH observed with the Nançay Radio Telescope (NRT) and the Green Bank Telescope – to the infrared (ground-based observations with IRTF and Keck-NIRSPEC and from space with the James Webb Space Telescope on 28 February and 4 March, Milam et al. 2023).

In this paper, we report clear detections of HCN, HNC, CH₃CN, HNCO, NH₂CHO, CH₃OH, H₂CO, HCOOH, H₂S, and CS in both comets, the more marginal detections of CO, HC₃N, CH₃CHO, OCS, (CH₂OH)₂, CH₂OHCHO, and C₂H₅OH, obtained by averaging several lines, as well as significant upper limits on the abundances of SO, SO₂, H₂CS, CH₂CO, PH₃, and other species.

In addition, following the revised photo-destruction rates of several molecules by Hrodmarsson & van Dishoeck (2023), especially of HNCO and HCOOH for which the lifetime has been reduced by nearly one order of magnitude in comparison to previously published or assumed destruction rates (Heays et al. 2017; Huebner & Mukherjee 2015; Biver et al. 2021a), we have reevaluated their abundances in all comets in which they were observed or searched for.

In Sect. 2, we present the observations and spectra of the detected molecules. The information extracted from the observations to analyse the data and compute production rates is provided in Sec. 3. In Sec. 4, we present the retrieved production rates and abundances or upper limits, which are discussed and compared to other comets. The new analysis of HNCO and HCOOH observations are detailed in Sec. 5 followed by the conclusions in Sec. 6.

2 Observations

2.1 Observations of comet C/2021 A1 with the IRAM-30m

Comet C/2021 A1 (Leonard)⁴ was the focus of a worldwide campaign as it was expected to become very bright at its closest approach to Earth (total visual magnitude m₁ ~ 4). It was less active than anticipated until 14 December, when it started to undergo a series of recurrent outbursts bringing it to a maximum brightness of around m₁ = 3 between 15 and 25 December 2021.

It was the target of the observing proposal 100-21 scheduled at the IRAM 30-m telescope between 8 and 13 December 2021. Weather conditions were marginal during the first four days (half of the day too icy, windy, or foggy to observe, with precipitable water vapour (pwv) in the 3 to 7 mm range otherwise). On the first day (observations were taking place during daytime), when observations resumed after an ice storm, the leftover ice or the impact of de-icing on the antenna reduced the beam efficiency by a factor of 3 (at 1 mm wavelength) to 1.5 (at 2 mm). The last two days offered good weather conditions with pwv = 1–3 mm (Table A.1). A few other snapshots (less than 0.5 h of observations) were obtained between 12 and 16 November during observing run 001-21 when comet 67P/Churyumov-Gerasimenko was too high in elevation for tracking (Biver et al. 2023a).

Observations were obtained in wobbler switching mode with the secondary (wobbling) mirror alternating pointing between the ON and OFF positions separated by 180″ every 2 seconds. The wobbler was not working on 14–16 November 2021 and we had to use to the position switching mode (PSW). The reference OFF positions in PSW mode were at 300″ from the source, alternating ON and OFF every 15 seconds. We used the EMIR (Eight MIxer Receiver, Carter et al. 2012) 3 mm, 2 mm, and 1 mm band receivers in 2SB mode connected to the FTS (Fast Fourier Transform Spectrometer) and the VESPA (VErsatile SPectrometer Array) high-resolution spectrometer (Table A.1). The FTS offers an instantaneous bandwidth of 16 GHz in two polarisations with 200 kHz spectral resolution. The VESPA autocorrelator was optimised to provide 4–6 windows of 20–40 kHz spectral resolution on lines of interest in the centre of the Intermediate Frequency windows (6.25 ± 0.25 GHz).

Comet C/2021 A1 was tracked with the IRAM-30m control software (NCS) using the latest JPL Horizons⁵ orbital elements available:#15 in November, #17 on 8 December, and #18 on 9–13 December 2021. Offsets (up to 7″) were added during the observation to take into account the difference between the position computed by the NCS and the very latest (1–3 days old) astrometric measurements. This was a critical point as for such a new comet relatively close to Earth, positions errors can easily reach the beam size (10″ at 240 GHz). The comet position was also checked in real time using coarse (5–9 points) mapping of HCN J=3–2 (Fig. 1). Ephemeris offsets were computed afterwards using the JPL#19 orbit solution (yielding values in agreement with the astrometry that was used). Final pointing offsets for each observation were computed after taking into account reconstruction of pointing errors and finally from intensity maps of HCN, that generally yielded residual offsets of less than 1.2″.

Representative spectra of comet C/2021 A1 (Leonard) are provided in Figs. 2 and 3 and Figs. A.1–A.3. They show individual lines or averages of several lines for the complex organics from the observations centred on the nucleus. Detailed line intensities are provided in Tables online and for some molecules with useful spatial information in Table B.5 of Sec. 4.3. Spectra showing the full spectral coverage with the wide band FTS are available online.

Fig. 1 Coarse map of the HCN(3–2) line integrated intensity in comet C/2021 A1 on 13.4 December 2021. The beam size is 9.3″. Solar phase angle was 158.6°. The direction of the Sun (PA=193.1°) is indicated at the lower right.

2.2 Observations of comet C/2022 E3 (ZTF) with Odin

Comet C/2022 E3 (ZTF) was favourably placed (solar elongation between 60° and 120°) for Odin during one of its yearly astronomy science operations. 15 orbits (~24h) were dedicated to observe the H₂O(1₁₀ − 1₀₁) line at 556.9 GHz in this comet. Odin (Frisk et al. 2003) is a small satellite in a polar orbit (period 95 min) equipped with a 1.1 m sub-millimetre radiometer. Odin houses four sub-mm receivers covering the 486–504 GHz and 541–581 GHz range with 3 backends: one acousto-optical-spectrometer (AOS) with 1 GHz band width and 1.2 MHz spectral resolution and two autocorrelators (AC1 and AC2), covering 120 MHz with 140 kHz resolution in their highest resolution mode. Two of the receivers, 555B2 and 549A1, were supposed to observe the 556.9 GHz line, but the 549A1 receiver did not lock, and we only obtained data from the 555B2 receiver connected to the AOS and AC2. We aimed at alternating between pointing on the comet position and coarse 3 × 3 points maps spaced by 60 or 120″. The beam size was 127″. Odin achieved successful tracking of the comet on most of 14 of the 15 orbits scheduled from 19.3 to 20.3 January 2023 UTC (Table A.2). Part of the data from the 13th and 14th orbits was lost also due to memory overload. Reconstruction of the attitude of Odin did not always converge with good accuracy due to the lack of reference stars in the star-trackers field of view, but most of the pointings were within 15″ (≈12% of the beam width) of the comet position. Figure 4 shows the result from the combination of all the maps.

Fig. 2 Millimetre lines observed with IRAM-30m in comet C/2021 A1 during 12–16 November 2021 (average intensity). The CH3OH line is the sum of the J=1 to 4 J1 − J0 E lines at 165050.175, 165061.130, 165099.240, and 165190.475 MHz observed with the high resolution backend VESPA. The vertical axis is main beam brightness temperature in K and the horizontal axis is Doppler velocity in the rest frame of the comet with respect to the main line.

2.3 Observations of comet C/2022 E3 (ZTF) with the IRAM-30m

Comet C/2022 E3 (ZTF) was the target of the observing proposal 097-22 scheduled at the IRAM 30-m telescope between 3 and 8 February 2023. Weather conditions were very good to good during the first four days (but too windy followed by a technical issue with the antenna temperature control during 2/3 of the first night). The last night (February 7/8) was lost due to bad weather (snow + wind). Pointing and focus stability was not very good at the beginning of the nights. Very low opacity on 5.0 February (0.2–1 mm pwv) enabled a short observation at 177–185 GHz, but the weather conditions degraded and the observing time was too limited to get a useful result on the H₂O line at 183.3 GHz. Table A.2 provides a log of the observations.

Wobbler switching mode with the secondary (wobbling) mirror alternating pointing between the ON and OFF positions separated by 120″ every 2 seconds could only be used on the last night as it had again issues as in November 2021. We had to rely on position switching mode (PSW) with reference OFF positions at 240″, alternating ON and OFF every 15 seconds. Even using symmetric mode (ON OFF OFF ON sequence), alternating the reference position in +RA and -RA, leaves some residual at the position of the strong ozone atmospheric lines, where the spectra are noisier with poorer baselines, especially around 231.13, 237.17, 242.34, 249.80, 249.97, and 267.28 GHz, As for comet C/2021 A1, we used the EMIR receivers and the FTS and VESPA spectrometers (Sec. 2.1).

Comet C/2022 E3 was tracked with the IRAM-30m control software (NCS) using the latest JPL Horizons orbital elements available at the beginning of the observations, JPL#41, adding up offsets (~5″ in declination) from the latest astrometric measurements available and confirmed by coarse mapping of the strongest lines in the setup. Final offset in the reduced data take into account a more recent (JPL#45) orbit solution and reconstructed pointing corrections of the antenna to cope with potential approximations during real time observations (Fig. 5).

Representative spectra of comet C/2022 E3 (ZTF) are provided in Fig. 6 and Figs. A.4–A.6. They show individual lines or averages of several lines for the complex organics from the observations centred on the nucleus. Detailed line intensities are provided online, and for some molecules with useful spatial information in Table B.6 of Section 4.3. Spectra showing the full spectral coverage with the FTS are provided online.

Figs. A.3 and A.6 show the average of 3 to 97 lines, weighted according to the noise in each spectrum. We have selected the strongest lines with expected similar S/N (within a factor around four) which are not blended with other lines. This is only for the purpose of highlighting detection and line shapes.

Fig. 3 Individual lines observed with IRAM-30m in comet C/2021 A1 during 8–13 December 2021 (weighted averages). The vertical axis is main beam brightness temperature in K and the horizontal axis is Doppler velocity in the rest frame of the comet with respect to the line. The position and relative intensity of the hyperfine components of the HCN(3–2) line have been drawn below the line.

Fig. 4 Map of the average spectra of the H2O(110−101) line obtained with Odin between 19.45 and 20.16 January 2023 in comet C/2022 E3 (ZTF), as a function of the pointing offset (in arcsec). A baseline in red is plotted on each spectrum. Solar phase angle was 61.6°, and the direction of the Sun (PA=120.2°) is indicated at the upper right with scales for individual AC2 spectra.

2.4 Observations with the Nançay Radio Telescope

The NRT performances and the observing procedure used for comets were described in Crovisier et al. (2002). The integration time is usually about one hour per day. However, from 12 September to 14 December 2021, it was limited to 30 min due to work on the focal track. The beam size at 18-cm wavelength is 3.5 × 18′.

Fig. 5 Coarse (3×3 points) map of the HCN(3–2) line integrated intensity on 4.67 February 2023 in comet C/2022 E3. Solar phase angle was 45.5°. The direction of the Sun (PA=287°) is indicated at the upper right.

2.4.1 Comet C/2021 A1 (Leonard)

C/2021 A1 (Leonard) was observed at the NRT from 1 October 2021 to 14 February 2022 almost every day or every other day⁶.

A succession of outbursts was observed, beginning on 13 December just at the end of the IRAM observations (Crovisier et al. 2021; Jehin et al. 2021). The OH production rate, which was 2.5 ± 0.1 × 10²⁸ molecules s⁻¹ on average for 8–12 December, rose to 6.3 ± 0.3 × 10²⁸ molecules s⁻¹ on 13.5, up to 22.1 ± 0.2 × 10²⁸ molecules s⁻¹ on 15.5 December.

Representative spectra for selected dates are shown in Fig. 7. The evolution of the retrieved OH production rate is plotted in Fig. 8.

2.4.2 Comet C/2022 E3 (ZTF)

C/2022 E3 (ZTF) was observed at the NRT from 17 October to 23 December 2022⁷ The retrieved OH production rate rose from 3.3 to 8.1 × 10²⁸ molecules s⁻¹. Then the observations were interrupted due to a technical failure on 24 December 2022. They were resumed from 17 to 24 February 2023 in a degraded mode, resulting in an upper limit Q(OH) ≤ 6 × 10²⁸ molecules s⁻¹.

3 Data analysis

3.1 Expansion velocity and outgassing pattern

The lines profiles (Figs. 2–3) of comet C/2021 A1 do not show systematic asymmetry, excepted for small shifts with respect to the rest velocity in the comet frame. In November the mean Doppler shift of the line is slightly negative (δv = −0.03 ± 0.03 km s⁻¹ for HCN(3–2)) and positive for most lines in December (δv(HCN(3–2))= +0.18 ± 0.01 km s⁻¹). The two-Gaussian fit (Biver et al. 2021a) used to estimate the expansion velocity from the velocity at half maximum intensity VHM, yields VHM = −0.75 ± 0.04 km s⁻¹ and VHM = +0.53 ± 0.03 km s⁻¹ in November (from HCN) and VHM = −0.63 ± 0.01 km s⁻¹ and VHM = +0.97 ± 0.01 km s⁻¹ in December (all lines). Those asymmetries are expected for a preferential outgassing at a higher rate and velocity on the sunward hemisphere, since in November the phase angle was ~51° (the sunlit hemisphere is mostly facing us) and in December it was between 116° and 159° (Table A.1) so that we were mostly facing the night side of the comet. We could have simulated a two-component outgassing pattern with a higher production rate and expansion velocity on the day side, but this would not change significantly the retrieved production rates for the optically thin lines from using an isotropic model with a constant expansion velocity equal to the mean of the day and night sides. Thus, we assumed isotropic outgassing with v_exp = 0.60 km s⁻¹ in November and v_exp = 0.73 km s⁻¹ in December. The actual expansion velocities needed to fit the observed profiles are 5–10% lower than the VHM (that is −0.58 and 0.88 km s⁻¹ in December) due to thermal broadening. An example of simulated profile with asymmetric outgassing is shown in Fig. 9.

In the case of comet C/2022 E3 (ZTF), the full width at half maximum (FWHM) of the water line at 556.9 GHz is of the order of 1.60 km s⁻¹, but the line is asymmetric and redshifted due to opacity effects (see for example Lecacheux et al. 2003). In addition the uncertainty on the frequency calibration is of the order of 0.05 km s⁻¹, so it is difficult to derive information on the gas velocity asymmetry from these data, but the average expansion velocity should be of the order of 0.8 km s⁻¹. IRAM data obtained two weeks later provide more accurate line profiles but also show evidence of variation with time. On the basis of the lines with the best signal-to-noise ratio (HCN, CH₃OH, H₂S, CS, H₂CO, and CH₃CN), the weighted average VHMs are −0.83 ± 0.01 and +0.64 ± 0.01 km s⁻¹, with a mean Doppler shift of the lines of −0.12 ± 0.01 km s⁻¹. Since the phase angle was not too large (45°, Table A.2), we deduce that outgassing was larger on the day-side with a larger expansion velocity. From the VHMs and fits to line profiles (Figs. 10, 11) we estimate that the expansion velocity was of the order of 0.76 km s⁻¹ on the day side and 0.52 km s⁻¹ on the night side, with a production rate two times higher on the day side to get the average measured Doppler shift. Nevertheless, assuming isotropic outgassing with the average v_exp = 0.68 km s⁻¹ yields very similar production rates and we assume isotropic outgassing at this velocity to compute and compare the production rates. Model fits with constant velocities in Figs. 9, 10, and 11 are underestimating the signal around zero velocity. Better fits to the line shapes are obtained when simulating radial acceleration of the expansion velocity, using the parameter x_acc = 4 from the formula in Biver et al. (2011, 2022) and v_exp,0 = 0.9 × v_exp. This illustrate a possible way to improve the fit, with a slightly lower (≈3–10%) production rate for most molecules. But there are other possible explanations, like variation of the gas temperature and different azimutal gas distribution. We keep the simpler model with constant velocity and isotropic outgassing for the computation of production rates and abundances of the numerous molecules observed.

3.2 Gas temperature

Several species such as methanol are detected through multiple transitions coming from different rotational upper energy levels E_u. In the case of local thermal equilibrium (LTE) we expect the relative population levels (p_u) to follow the Boltzmann law (p_u ∝ exp(-E_u/kT)). When we plot the logarithm of the populations p_u versus the upper energy levels E_u (the socalled rotational diagram) the mean slope of the linear fit of the data is 1/T_rot, where T_rot is the rotational temperature, equal to the gas kinetic temperature T in LTE. Due to radiative decay and infrared pumping, deviations from LTE can be observed in some series of lines, especially the CH₃OH lines at 242 GHz, for which T_rot < T. In such a case the full non-LTE modelling of the evolution of the population of the rotational level throughout the cometary atmosphere is required to estimate the value of T resulting in the measured T_rot within the radio telescope beam. We provide in Figs. C.1–C.13 the rotational diagrams for species for which we observed several transitions with sufficient signal-to-noise ratio (S/N>3) and in Table 1 the measured T_rot and inferred T.

The average gas kinetic temperature T for comet C/2021 A1 inferred from CH₃OH and CH₃CN rotational temperatures (Table 1) is T = 25 ± 5 K in November and T = 61 ± 3 K in December. In December the temperature inferred from HNCO lines (Figs. C.4) suggests a lower value of T, but we have not taken into account the effect of infrared pumping that could modify the value expected for T_rot(HNCO) for a given T. The spatial distribution of HNCO is also not well constrained: if it comes from a distributed source then the observed rotational temperature will be compatible with a higher value of T. We have adopted T = 30 K in November and T = 60 K in December.

In the case of comet C/2022 E3 (ZTF), due to higher signal-to-noise ratios, we have numerous measurements of rotational temperatures (rotational diagrams in Figs. C.5–C.13). Inferred gas temperatures T are provided in Table 1. From methanol data, the weighted average and dispersion is T = 58 ± 4 K, from CH₃CN we get T = 68 ± 9 K, and T = 53 ± 16 K from the four other species in Table 1. We adopt the average T = 60 K to derive production rates of comet C/2022 E3 for all species. All measurements, given their uncertainties, are within 10 K of this value (Table 1) but due to likely different collision rates for the different molecules, small deviations are not surprising. Also for NH₂CHO, HCOOH, HNCO, and CH₃CHO we have not taken into account infrared pumping that could introduce more differences between T_rot and T. We have also investigated possible day versus night differences of the coma temperature. The positive parts of the 166 GHz methanol lines only suggest marginally higher temperature on the night side than on the day-side (blueshifted parts of the lines, Table 1).

Fig. 6 Individual lines observed with IRAM-30m in comet C/2022 E3 during 3–6 February 2023 (average intensity, excluding offset positions). The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet with respect to the line. A galactic contamination of CO has been blanked out. The position and relative intensity of the hyperfine components of the HCN(3–2), HCN(2–1), and H13CN(3–2) lines have been drawn below the lines.

Fig. 7 Selected averages of the OH lines observed at the NRT in comet C/2021 A1 (Leonard) in November-December 2021 (averages of the 1667 and 1665 MHz lines scaled to 1667 MHz).

Fig. 8 Time evolution of comet C/2021 A1 (Leonard), single days, and averages of selected periods in November-December 2021. The OH production rates from the NRT are plotted in black. The daily water production rates inferred from SOHO/SWAN observations of the wide H Lyman-α coma, which tends to smooth out short term variations, are plotted in red. (Combi et al. 2023a).

Fig. 9 Average spectrum of the HCN(3–2) line observed from 8.5 to 13.4 December in comet C/2021 A1, with simulated profiles. The model in blue assumes a production rate of 2.1 × 1025 molecules s−1 at 0.88 km s−1 on the sunward hemisphere and 1.1 × 1025 molecules s−1 at 0.58 km s−1 on the opposite hemisphere mostly facing the observer. The mean phase angle is ~140°, and the tilt of 40°(or 140°) with respect to the comet-observer line of sight is taken into account in this 3D simulation, as depicted in the upper right. The model in purple uses variable velocities in both hemispheres (see text), with production rates of 2.3 × 1025 molecules s−1 and 0.9 × 1025 molecules s−1, respectively. The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet.

Fig. 10 Average FTS spectrum of the HCN(3–2) line observed from 3.7 to 6.7 February in comet C/2022 E3 with simulated profiles. The model in blue assumes a production rate of 3.0 × 1025 molecules s−1 at 0.76 km s−1 on the sunward hemisphere and 1.5 × 1025 molecules s−1 at 0.52 km s−1 on the other hemisphere, as depicted in the upper right. The model in purple uses variable velocities in both hemispheres (see text), with production rates of 3.0 × 1025 molecules s−1 and 1.0 × 1025 molecules s−1, respectively. The mean phase angle is ~45°. The vertical axis is main beam brightness temperature in K, and the horizontal axis is Doppler velocity in the rest frame of the comet.

Fig. 11 Average spectrum of the CH3OH line at 266.838 GHz observed from 3.7 to 6.7 February 2023 in comet C/2022 E3 at the same time as HCN(3–2). The model in blue assumes a production rate of 5.6 × 1026 molecules s−1 at 0.76 km s−1 on the sunward hemisphere and 2.8 × 1026 molecules s−1 at 0.52 km s−1 on the other hemisphere, as depicted in the upper right, while the red profile comes from symmetric outgassing at 0.68 km s−1 and same total production rate (8.4 × 1026 molecules s−1). The model in purple uses the same variable velocities in both hemispheres as in Fig. 10, with production rates of 6.4 × 1025 molecules s−1and 2.0 × 1025 molecules s−1, in sunward and other hemispheres, respectively. The mean phase angle is ~45°. The vertical axis is main beam brightness temperature in K, and the horizontal axis is Doppler velocity in the rest frame of the comet.

4 Production rates and abundances

Production rates are computed using our excitation and radiative transfer codes, and parameters as in previous papers (Biver et al. 2021a). We assumed isotropic outgassing and a constant velocity and temperature (Sects. 3.1 and 3.2). We provide both daily production rates when the molecules are detected with a sufficiently high signal-to-noise (Tables B.1 and B.2) and averages over the 13–16 November and 8–13 December 2021 periods for comet C/2021 A1 (Leonard) and 3–7 February 2023 for comet C/2022 E3 (ZTF) (Tables B.3 and B.4). For the molecules for which we observed several lines ‘i’, either detected individually with a S/N >5 or not, the final production rate is the weighted average of all production rates Q_i. They are computed on each considered line (even when Q_i is negative) and averaged with weighting according to 1/σ(Q_i)² were σ(Q_i) is the uncertainty in production rate deduced from the line ‘i’.

4.1 Reference water production rate

Water production rates are inferred from the monitoring of OH lines at 18-cm with the NRT (Sec. 2.4) and H₂O observations at 557 GHz with Odin (Sec. 2.2). Water productions rates for these comets were also obtained from SOHO/SWAN (Combi et al. 2023a,b). For the period of observations with the IRAM-30m radio telescope, and to compute relative abundances, we estimate $Q_{{\mathrm{H}}_2\mathrm{O}}=2,3$, and 4×10²⁸ molecules s⁻¹ for the 12, 13–16 November, and 8–13 December 2021, respectively, for comet C/2021 A1 (Leonard). For comet C/2022 E3 (ZTF), extrapolation from Odin production rates in Table 2, following $1/r_h^2$, yields $Q_{{\mathrm{H}}_2\mathrm{O}}=5\times {10}^{28}$ molecules s⁻¹ for the 3–7 February period. These values also follow the longer term trend of brightness evolution. We use these values to determine collisional rates and abundances relative to water.

4.2 Temporal variations in the production rates

Evidence for short-term variability in the outgassing is present both in Odin and IRAM data for comet C/2022 E3 (ZTF), but not for comet C/2021 A1. Visual inspection of the production rates folded on a single period seems to indicate a periodic pattern of about 0.35 days in $Q_{{\mathrm{H}}_2\mathrm{O}}$ and 0.38 days in $Q_{{\mathrm{C}\mathrm{H}}_3\mathrm{O}\mathrm{H}}$ while optical observations have reported a periodic pattern in the CN structures of 0.363 ± 0.004 days (Knight et al. 2023) and 0.354 ± 0.004 days (Manzini et al. 2023). We fitted (Figs. 12 and 13) a sine-wave variation (for simplicity – a more complex profile would require more parameters) to our water and methanol production rates and found similar periods. Table 2 provides the fitted parameters and their uncertainty based on χ² minimisation for a simple sinusoidal pattern. The relative amplitudes are 18% for methanol and 5% for H₂O. H₂O being observed with a beam about ten times larger (the Odin beam radius of 33000 km covers molecules emitted over a full period of 0.38 days), and having optically thick lines, it is expected to display shallower variations. The amplitude of the variations for production rate of methanol are more likely to represent the full extent of variations of outgassing of the nucleus and are similar to the variations (±20%) observed for HCN. This has to be taken into account when deriving precise ratios (e.g. isotopic ratios in Sec. 4.6), but for general abundances averaged over the four days the impact is small, especially as on the 4, 5, and 6 February evenings (Table A.2) the observations covered more than a full period of ~0.36 days.

For other species day-to-day variations of the production rates are provided in Tables B.1 and B.2, and plotted in Figs. 14 and 15, together with some estimates of the Afρ quantity⁸ in order to have a rough idea of the relative evolution of the dust production rate.

4.3 Molecules coming from a distributed source

The coarse mapping done during the observations provides some information on the spatial extent of the emission for some molecules, within 14″ from the nucleus (2400 km at the distance of comet C/2021 A1 in December 2021) to ~20″ (4500 km) for comet C/2022 E3 (ZTF). Tables B.5 and B.6 provide information on the spatial extent of the emission of some molecules (average line intensities as a function of pointing offset) and constraints on the scale length and production from a distributed source. We adjusted a Haser daughter species density profile to the data and provide the result of the χ² minimisation with 1σ uncertainty on the scale length (L_p) and production rate (Q_p). Values for other fixed scale lengths (Q_p and corresponding χ²) are provided. These results have to be taken with caution since: (i) the excitation of the rotational levels is not perfectly known due to variable temperature in the coma, unknown collisional cross-sections, and other poorly estimated processes that can mimic extended emission, (ii) the comet was variable in activity (it underwent a huge outburst on 14 December just after the last observation), (iii) the observations are mostly probing the 1600–5000 km spatial scales (beam size – extent of the mapping), so that information on very different scale lengths is not well constrained.

HCN: it has been shown to be mostly released from nucleus ices (Cordiner et al. 2014), but the IR observations (Dello Russo et al. 2016; Lippi et al. 2021) that probe the inner part of the coma also generally find higher production rates of HCN than the radio ones (Biver et al. 2024). The observations of both comets (data were treated day by day to limit time variability effects) suggest that a fraction of HCN is produced in the coma. The best fit to the 12.5 and 13.4 December observations of comet C/2021 A1 yield a parent scale length of the order of 350 km with Q_p(HCN) larger by 15 to 35% than with L_p = 0. HCN(3–2) data from comet C/2022 E3 (ZTF) yield a scale length of 600 ± 500 km while HCN(2–1) suggest L_p(HCN) < 780 km, implying also Q_p(HCN) larger by ≈20% (Tables B.5 and B.6). In both comets the retrieved parent scale length is smaller than the beam size, so the extended production of HCN needs to be further investigated at higher spatial resolution. The retrieved production rate would only be increased by ≈20%, but other issues show that the excitation of HCN is not fully understood. We also cannot exclude that an excitation process, for example a larger gas temperature or electron collision rate that would decrease the J=3 rotational level population close to the nucleus, is responsible for this apparent distributed production. Also the imperfect modelling of the line shape (underestimation of the J(3–2) F = 2–2 and F = 3–3 hyperfine satellite lines and a signal strength around zero velocity in Figs. 9, 10 being too high), suggest that there are other issues. A much larger opacity could provide a better fit to the line shape, although requiring unrealistic production rate and line intensity.

HNC: the mean of offset observations of HNC(3–2) on the whole period does not yield a significant detection (signal-to-noise is below 2) but L_p = 0 ± 700 km from χ² minimisation is the best fit to comet C/2021 A1 data. In comet C/2022 E3 we find L_p > 2800 km at 1-σ, also poorly constrained with signal below 3-σ at offsets larger than 5″. Interferometric observations of HNC (Cordiner et al. 2017; Roth et al. 2021) suggest a parent scale length of the order of ~2000 km at 1 au (scaled as $r_h^{-2}$), somewhat compatible with our data. Therefore, we provide in the tables Q_p(HNC) for nuclear source L_p = 0 and L_p(HNC) = 1000 km and 2000 km for comets C/2021 A1 and C/2022 E3, respectively.

CS: in comet Leonard we obtained data at offset positions on three different dates: 10.5, 12.5, and 13.6 December. The only two points obtained on the first date do not yield significant constraints (L_p(CS) > 800 km at 1σ) and are not listed in Table B.5. The larger dataset of J=5–4 line observations of 12.48 December (Table B.5) yields a reliable value (reduced ${\chi }_{\nu }^2=0.94$): L_p(CS) = ${500}_{-390}^{+580}\text{\hspace{0.17em}km}$. We tried to combine the J=5–4 and J=3–2 observations of 13.6 December but they yield a large value for L_p(CS) > 2700 km (mostly driven by the CS(5–4) data that yield L_p(CS) > 2400 km at 1σ). This implies a too large value Q_p(CS) = 12.6 × 10²⁵ molecules s⁻¹ for the nominal fitted value of L_p(CS) = 7000 km. Since CS(3–2) and CS(5–4) were not observed simultaneously, the constraint may not be reliable. CS(3–2) data taken alone suggests L_p(CS) ~ 1400 km, with Q_p(CS) = 5.2 × 10²⁵ molecules s⁻¹ although compatible with any value of L_p at 1σ.

In comet C/2022 E3, taken separately, CS(3–2) and CS(5–4) maps yield parent scale lengths of L_p(CS) = ${1400}_{-930}^{+2000}$ and ${700}_{-620}^{+1450}\mathrm{k}\mathrm{m}$, or L_p(CS) = ${1900}_{-570}^{+900}\text{\hspace{0.17em}km}$ combining all data (larger since constrained by a unique value for Q_p(CS), Table B.6).

Around these heliocentric distances (0.8 and 1.2 au), the CS₂ photo-dissociation scale lengths are expected to be ~260 and 560 km. The average of all measurements obtained, yield L_p(CS) >800 and 1300–2200 km, respectively. Most other observations also suggest a parent scale length for CS that is about 3–5 times the CS₂ dissociation scale length (Biver et al. 2022; Roth et al. 2021). Hence, we use L_p(CS)=1000 and 2000 km for C/2021 A1 and C/2022 E3, respectively. This is equivalent to assuming that CS comes from a parent molecule X-CS with a photo-dissociation rate at 1 au β₀(X − CS) ~ 4.5 × 10⁻⁴ s⁻¹, which is comparable to the revised photo-destruction rate of H₂CS (β₀(H₂CS) = 4.9 × 10⁻⁴ s⁻¹, Hrodmarsson & van Dishoeck 2023). However upper limits on the H₂CS abundance are much lower than for CS parent. We note that assuming that CS comes directly from the nucleus reduces the production rate by only ~14% when comparing to the case of production by dissociation of CS₂ with its known lifetime.

H₂CO: combining the two $J_{K_a{K}_c}=3_{1K_c}$ lines at 211.211 and 225.698 GHz observed on 12.6 December provides some constraints on the spatial distribution of H₂CO in the coma of comet C/2021 A1. The retrieved parent scale length is ${1500}_{-970}^{+3700}\text{\hspace{0.17em}km}$ (Table B.5) within the range of values (0.2–1.6× the photo-destruction scale length L_d of H₂CO, which is 2200 km around 12 December), found from previous observations (Biver et al. 2022, 1999; Roth et al. 2021; Cordiner et al. 2014). The +1σ uncertainty is large and the upper limit not very precise (χ² does not increase as steeply as for the lower limit). In addition, if there is some nucleus contribution, we can mostly say that the H₂CO coming from a distributed source must have a parent scale length larger than 1400 km. We use L_p(H₂CO) = 1500 and 2800 km (0.68 × L_d(H₂CO)) to retrieve the (parent) production rate of H₂CO in December and November, respectively.

For comet C/2022 E3 (Table B.6), we made observations at offset positions up to 20″ on 5.00 February and 28″ on 5.95 February (4400–6500 km with a 2400 km beam) that provide more stringent constraints on the spatial distribution of H₂CO. For the assumed Haser daughter distribution profile, we found L_p(H₂CO) = 1000 ± 430 and 1700 ± 1100 km, respectively. The observations of 6.94 February as well as those obtained for other lines (2₁₁ − 1₁₀ line at 150.498 GHz) provide looser constraints that fully encompass those values. Fig. 16 shows the combination of all H₂CO observations at 211 and 226 GHz in comet C/2022 E3 versus expected evolution of line intensity as a function of pointing offset for various parent scale lengths – the retrieved L_p(H₂CO) is 1700 ± 500 km which corresponds to 0.36 ± 0.11 × L_d(H₂CO) at 1.18 au from the sun, with v_exp = 0.68 km s⁻¹.

NH₂CHO and CH₃CHO: we have also studied the spatial distribution of these two species (only NH₂CHO in C/2021 A1): results are not conclusive and may be affected by excitation effects (especially for the high K_a transitions of formamide). Slightly distributed production (with L_p ~ 1500 km) could be favoured for comet C/2022 E3, but a nuclear source is not clearly excluded (within the 2 − σ uncertainty) as for formaldehyde.

In addition, if a significant production of the molecules comes from the sublimation of icy grains in the coma, all species could also show some distributed production. But the least distributed species like HCN, could put an upper limit on the spatial extent of the distributed production. The production of CS and H₂CO remains clearly more distributed and they are thus daughter products.

Fig. 12 Water production rates (in molecules s−1) of comet C/2022 E3 (ZTF) between 19.3 and 20.3 January 2023 derived from observations with Odin (central pointings). A best fit sinusoidal variation is plotted in blue.

Fig. 13 Methanol production rates (in molecules s−1) of comet C/2022 E3 (ZTF) between 3.7 and 7.1 February 2023 derived from observations with IRAM 30m. An average 5% calibration uncertainty has been added to the formal uncertainty due to system temperature noise. A best fit sinusoidal variation is plotted in blue.

Fig. 14 Production rates of comet C/2021 A1 (Leonard) in November and December 2021. Water production rates are 1.1 × QOH from Sec. 2.4. Afρ were measured from images of the comet by N. Biver, not corrected for the phase angle.

Fig. 15 Production rates of comet C/2022 E3 (ZTF) in January and February 2023. Water production rates are from Odin observations (Sec. 2.2, Fig. 12). We note that Afρ(0°) were measured from images of the comet by N. Biver and have been corrected for the phase angle.

Fig. 16 Spatial distribution of formaldehyde in comet C/2022 E3 (ZTF). This plot shows for the average of all observations of the twin 211 GHz and 226 GHz lines observed together in comet C/2022 E3 (ZTF) from 3 to 7 February 2023: (i) in black squares (with error-bars) the observed line integrated intensity, (ii) in solid and dotted lines the expected intensity for parent scale length of 0 or 1000–8000 km. For better visibility a normalisation has been applied: at each offset the line intensity is multiplied by the column density at the first offset divided by the one expected at the considered offset point, for a parent distribution. Production rate of the models are adjusted to get the same integrated intensity at the first point. Deviation from a flat line for Lp = 0 (black continuous line) is due to excitation effects. Vertical axis is normalised integrated intensity in mK km/s and horizontal axis is pointing offset converted into km at the distance of the comet.

4.4 Relative abundances

Abundances relative to water, assuming $Q_{{\mathrm{H}}_2\mathrm{O}}=2-3\times {10}^{28}$ molecules s⁻¹ in November and $Q_{{\mathrm{H}}_2\mathrm{O}}=4\times {10}^{28}$ molecules s⁻¹ in December 2021 for comet C/2021 A1 and $Q_{{\mathrm{H}}_2\mathrm{O}}=5\times {10}^{28}$ molecules s⁻¹ in early February 2023 for comet C/2022 E3 are provided in Table 3. Abundances have also been measured for several molecules in comet C/2021 A1 in the infrared by Faggi et al. (2023), but during the outbursting period of the comet in December-January. While these authors found very similar abundances for HCN, OCS, H₂CO, and compatible with our upper limit on CO, they did not detect CH₃OH. Their inferred CH₃OH/H₂O is very low, down to one order of magnitude lower than our abundance (based on the secure detection of over 42 lines). Either the comet had exhausted its methanol content during the outburst phase at the end of December, or the infrared observations missed part of the methanol emission, following a trend also seen in millimetre investigations (Biver et al. 2011) which show an apparent decrease in methanol abundance as comets get closer to the Sun. The methanol observations of Faggi et al. (2023) were obtained at 0.62 au from the Sun, where the sampled gas coma gets warmer (assumed to be 120 K) and the fraction of methanol in an excited torsional state may also depopulate the ground vibrational levels.

4.5 Upper limits

The surveys cover part of the 2 mm wavelength range and most of the 1 mm (210–272 GHz) as shown in spectra online. Most of the known organic molecules have many transitions in these wavelength ranges. We have not noticed clearly unidentified lines (above the 5 − σ level). We checked the sensitivity of the observations for the following molecules: CH₃SH, CH₃OCH₃, CH₃COCH₃, c-C₂H₄O, c-C₃H₂, CH₃NH₂, C₂H₃CN, CH₃Cl, and propanal. Averaging the lines expected to be the strongest did not yield any significant signal. 3 − σ upper limits on abundances and production rates are provided in Tables 3, B.3, and B.4, and for the species not listed the upper limits on abundances are not expected to be better than in previous comets such as 46P/Wirtanen (Biver et al. 2021a).

4.6 Isotopic ratios

Isotopic ratios of H, C, N, and S and upper limits on their abundances in some molecules are provided in Table 4. Regarding HCN isotopologues, none are clearly detected but upper limits or marginal signal at the 2–3-σ level are compatible with values observed in other comets (Biver et al. 2024). Note that for ¹⁵N/¹⁴N, the Earth value of 272 corresponds already to an enrichment compared to the estimated solar value of 450 (Marty et al. 2011), showing that fractionation was important already for Earth nitrogen but even more for the material that was incorporated into cometary ices 4.5 billion years ago.

The ³²S/³⁴S could be measured in CS in both comets, but in comet C/2022 E3 it seems lower than the Earth value while compatible with terrestrial value for H₂S: this comet seems enriched in C³⁴S, at the 3-σ level. However the PSW observing mode used for comet C/2022 E3 is more prone to produce ripples that affect the signal of marginal lines. We also provide upper limits on abundances of deuterated species. None yields a D/H below the Earth VSMOW (Vienna Standard Mean Ocean Water) value, but since some enrichment has been observed in the interstellar medium and in some molecules ejected by comet 67P/Churyumov-Gerasimenko observed by the ROSINA mass spectrometer on board the Rosetta spacecraft (Drozdovskaya et al. 2021; Müller et al. 2022), we provide those upper limits.

5 The abundance of HNCO and HCOOH in comets

The molecules HNCO and HCOOH were first detected in comets C/1996 B2 (Hyakutake) and C/1995 O1 (Hale-Bopp) in 1996–1997 (Lis et al. 1997; Bockelée-Morvan et al. 2000). They have been detected in several comets since (Biver et al. 2023a) but there were large uncertainties on their destruction rates and lifetimes and a simple excitation model was used to derive their abundances. Hrodmarsson & van Dishoeck (2023) have recently published photo-dissociation and photo-ionisation rates for HNCO and HCOOH that significantly differ from those assumed in the past (Biver et al. 2021a, and references therein). The new values (β₀(HNCO) = 38 × 10⁻⁵ s⁻¹, β₀(HCOOH) = 54 × 10⁻⁵ s⁻¹, in the solar radiation field at 1 au) are 5 to 10 times higher than previously assumed, with a stated 2-σ uncertainty of the order of 20–30% (Hrodmarsson & van Dishoeck 2023). We have decided to revisit all millimetre observations of comets reporting a detection of at least one of these molecules and recomputed production rates and abundances relative to water assuming values of:

• β₀(HNCO) = 3.5 × 10⁻⁴ s⁻¹;

• β₀(HCOOH) = 5.0 × 10⁻⁴ s⁻¹.

The contribution of Solar Lyman-α to their photo-dissociation is minor and their photo-dissociation cross-section extends well into the 130–200 nm range where the solar flux is less affected by solar activity (Huebner et al. 1992). Therefore, we do not expect a large variation with solar activity. An overview of the observations analysed in this work, partly published in previous papers is given in Table A.3.

Table 5 provides the abundances relative to water of HNCO and HCOOH in these comets. In these calculations infrared pumping via the rotational bands is not taken into account, but with those reduced lifetimes, the photo-dissociation of the molecules takes place in a region where collisional excitation dominates (within ~2000 km from the nucleus) and the photo-dissociation process is as efficient as vibrational excitation (g-factors comparable to β). So the new reduced lifetimes (compared to previous calculations) should also reduce the uncertainties from the neglected infrared pumping.

HNCO is relatively well detected in the two comets via its K_a = 0 lines at 2 mm and 1 mm wavelengths, and marginal signals at an offset of 10″ (Tables B.5, B.6) suggest that it may be produced by a distributed source with a scale length L_p(HNCO) larger than 2000 km. Indeed, assuming a parent scale length of 5000 km also provides a better agreement between the observed rotational temperature and the predicted one for T = 60 K (Table 1).

HCOOH: In Tables B.5 and B.6 we have selected the lower energy lines that are expected to be the strongest, that is those with K_a = 0 to 2. For comet C/2021 A1, since at the offset position the weighted average intensity is similar to the one at the central position, we do not find a solution for a reasonably small parent scale length, but data are compatible within 1σ with L_p(HCOOH) = 2000 km, and L_p(HCOOH) = 0 km cannot be fully excluded. Similarly, for comet C/2022 E3, we derive L_p(HCOOH) = 0–5200 km in the ±1σ interval, so that distributed source as well as nuclear source are not excluded. Hence, if HCOOH is produced by the sublimation and dissociation of ammonium formate, then this must take place within a few thousands of kilometers from the nucleus at 0.78–1.18 au from the Sun.

The revised measured abundances in comets (increased by a factor 1.2 to 5) of these two molecules are plotted in Fig. 17 as a function of heliocentric distance at which they were detected. This plots suggest that their abundance relative to water increases for comets observed closer to the Sun. This needs to be confirmed, but can suggest that the molecules could be at least in part produced by the degradation of parents in the coma such as ammonium salts. Their combined abundance relative to water in comets (~0.1–1%) is comparable to that of NH₃(0.2–5%, Biver et al. 2024), the other decomposition product of ammonium salts. Interestingly, a similar trend with heliocentric distance has been observed for the abundance of ammonia (see, e.g. Dello Russo et al. 2016).

6 Conclusions

We have undertaken a comprehensive and sensitive spectroscopic survey at 2 mm and 1 mm of two long period comets, C/2021 A1 (Leonard) and C/2022 E3 (ZTF) when they were relatively close to the Earth (0.2–0.3 au). We collected in addition some useful spatial information from coarse mapping, to probe the spatial distribution of selected molecules. The main results can be summarised as follows:

• Both comets exhibited relatively stable (within ±20%) production rates during the observational periods, although C/2021 A1 underwent a series of disintegration outbursts just after our IRAM-30m observing run;

• We determined the abundance relative to water or a significant upper limit for 26 molecules. Comets C/2021 A1 and C/2022 E3 share similar compositions: they are depleted in hypervolatiles (low CO and H₂S abundances relative to water) and have a relatively low methanol abundance relative to water (0.9 and 1.8% respectively) compared to the mean of other comets;

• Constraints on the presence of distributed sources have been obtained. A slightly distributed source (scale length smaller than 1/3 beam) for HCN fits better the data than direct release from the nucleus. The parent source of formaldehyde was found to have a scale length L_p(H₂CO) = 0.68 and 0.36 ± 0.11 times the photo-dissociation length of formaldehyde (L_d(H₂CO)) in C/2021 A1 and C/2022 E3 respectively. We measured L_p(CS) ≈ 4 × L_d(CS₂), suggesting that the dissociation rate of the parent producing CS is β₀ = 4.5 × 10⁻⁴ s⁻¹ at 1 au from the Sun. This is comparable to the photo-dissociation rate of H₂CS, but the H₂CS upper limit is more than 3 times lower than the CS abundance;

• Both comets are relatively depleted in sulphur-bearing species compared to other comets, with H₂S/H₂O in the low range (0.15–0.18%) and the sum of sulphur-bearing molecule abundances below ~0.5%. SO and SO₂ are not detected with abundances below the lowest measured in a comet and OCS is detected in both comets but with a low abundance, too;

• We obtained only shallow constraints on isotopic ratios: ¹²C/¹³C, ¹⁴N/¹⁵N, and ³²S/³⁴S are compatible with values observed in other comets (90, 150, and 23, respectively), although the C³²S/C³⁴S seems marginally twice lower than the terrestrial value in comet C/2022 E3;

• HNCO and HCOOH are well detected in both comets. Because of recently published lifetimes for these molecules that are significantly shorter than previous estimates, we have revised their inferred abundances in all comets for which detections of these species have been reported. The values found in C/2021 A1 (0.07% and 0.19%) and C/2022 E3 (0.04% and 0.19%) are close to their median abundances relative to water in other comets, 0.05% and 0.25%, respectively. Their total abundance is also about half the median abundance of NH₃ in comets. We cannot exclude that these species are produced in the coma by a distributed source, and their abundance in the comae seems to increase at lower heliocentric distances. This favours the hypothesis that HNCO, HCOOH, and consequently an important fraction of NH₃ could come from the decomposition of ammonium salts.

These observations have provided complementary and new insights regarding the source of several cometary molecules that are unlikely present as such in cometary ices. Further investigations at higher spatial resolution (e.g. interferometric investigations undertaken on those comets under reduction or on future comets), will be very useful to pin down the production processes and the scale lengths of these molecules (CS, H₂CO, HNCO, HCOOH).

Fig. 17 Abundances relative to water of HNCO (green) and HCOOH (blue) in comets as a function of the heliocentric distance. Even though there is dispersion between comets, the trend appears to be that these species are more abundant in the coma close to the Sun.

Data availability

The radio spectra are available at the CDS via anonymous ftp to cdsarc.cds.unistra.fr (130.79.128.5) or via https://cdsarc.cds.unistra.fr/viz-bin/cat/J/A+A/690/A271. The tables of line intensities are available at https://doi.org/10.5281/zenodo.13332037. The spectra of comets are available at https://doi.org/10.5281/zenodo.13341404.

Acknowledgements

This work is based on observations carried out under projects number 001-21, 100-21, and 097-22 with the IRAM 30-m telescope. IRAM is supported by the Institut Nationnal des Sciences de l’Univers (INSU) of the French Centre national de la recherche scientifique (CNRS), the Max-Planck-Gesellschaft (MPG, Germany) and the Spanish IGN (Instituto Geográfico Nacional). We gratefully acknowledge the support from the IRAM staff for its support during the observations. The data were reduced and analysed thanks to the use of the GILDAS, class software (https://www.iram.fr/IRAMFR/GILDAS). This research has been supported by the Programme national de planétologie de l’Institut des sciences de l’univers (INSU). The Nançay radio telescope (NRT) is part of the Observatoire radioastronomique de Nançay, operated by the Paris Observatory, associated with the French Centre national de la recherche scientifique (CNRS) and with the University of Orléans. Part of this research was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration (80NM0018D0004). D.C.L. acknowledges financial support from the National Aeronautics and Space Administration (NASA) Astrophysics Data Analysis Program (ADAP). M.N. Drozdovskaya acknowledges the Holcim Foundation Stipend. B.P. Bonev and N. Dello Russo acknowledge support of NSF grant AST-2009398 and NASA grant 80NSSC17K0705, respectively. N.X. Roth was supported by the NASA Postdoctoral Program at the NASA Goddard Space Flight Center, administered by Universities Space Research Association under contract with NASA. M.A. Cordiner was supported in part by the National Science Foundation (under Grant No. AST-1614471). S.N. Milam and M.A. Cordiner acknowledge the Planetary Science Division Internal Scientist Funding Program through the Fundamental Laboratory Research (FLaRe) work package, as well as the NASA Astrobiology Institute through the Goddard Center for Astrobiology (proposal 13-13NAI7-0032).

Appendix A Additonal material

Fig. A.1 CH3CN series of four lines observed with IRAM-30m in comet C/2021 A1 on 8–13 December 2021 (average intensity). The vertical axis is main beam brightness temperature in K, horizontal axis is Doppler velocity in the rest frame of the comet with respect to the (J, 0) − (J − 1, 0) line, with frequencies in the rest frame of the comet given on the upper axis.

Fig. A.2 CH3OH series of lines around 166, 242, and 252 GHz observed with IRAM-30m in comet C/2021 A1 on 8–13 December 2021 (average intensity). The vertical axis is main beam brightness temperature in K, horizontal axis is Doppler velocity in the rest frame of the comet with respect to the main line, with frequencies in the rest frame of the comet given on the upper axis.

Fig. A.3 Spectra of molecules obtained by averaging several lines observed with IRAM-30m in comet C/2021 A1 between 8 and 13 December 2021. The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet. The number of lines averaged is provided for each molecule, either in the 2 mm band (147–153 and 163–171 GHz) or 1 mm band (209–272 GHz).

Fig. A.4 CH3CN series of lines at 147, 165, 220, and 257 GHz observed with IRAM-30m in comet C/2022 E3 on 3–6 February 2023 (average intensity). The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet with respect to the main line, with frequencies in the rest frame of the comet given on the upper axis.

Fig. A.5 CH3OH series of lines around 165, 242, and 252 GHz observed with IRAM-30m in comet C/2022 E3 on 3–6 February 2023 (average intensity). The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet with respect to the main line, with frequencies in the rest frame of the comet given on the upper axis.

Fig. A.6 Spectra of molecules obtained by averaging several lines observed with IRAM-30m in comet C/2022 E3 between 3 and 6 February 2023. The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet. The number of lines averaged is provided for each molecule, either in the 2 mm band (147–155, 163–171, and 176–184 GHz) or 1 mm band (209–272 GHz).

Appendix B Tables of production rates

Appendix C Rotational diagrams

The logarithm of a quantity proportional to the line intensity (line area divided by column density) is plotted against the energy of the upper level for each observed rotational transition. The upper levels of the transition are given at the top. The inverse of the slope of a fitted line provides the rotational temperature (given in pink) of the group of lines.

Fig. C.1 Rotational diagram of the CH3OH lines around 166 GHz observed between 8.6 and 13.7 December 2021 in comet C/2021 A1 (Leonard).

Fig. C.2 Rotational diagram of all the NH2CHO lines observed between 8.6 and 13.7 December 2021 in comet C/2021 A1 (Leonard). Lines with same J number and Ka = 0 to 2 or 3 to 4 are averaged together. Scales as in Fig. C.1). The values in green are those expected from modelling with T = 60 K.

Fig. C.3 Rotational diagram of the CH3CN lines at 147, 165, 220, 239, and 257 GHz observed between 8.4 and 13.7 December 2021 in comet C/2021 A1 (Leonard). Scales as in Fig. C.1). The values in green are those expected from modelling with T = 60 K.

Fig. C.4 Rotational diagram of the HNCO lines at 1 mm and 2 mm observed between 8.4 and 13.7 December 2021 in comet C/2021 A1 (Leonard). Scales as in Fig. C.1.

Fig. C.5 Rotational diagram of the CH3OH lines around 166 GHz observed between 4.9 and 7.0 February 2023 in comet C/2022 E3 (ZTF). The values in green are those expected from modelling with T = 60 K.

Fig. C.6 Rotational diagram of the CH3OH lines at 242 GHz observed between 4.7 and 6.8 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.

Fig. C.7 Rotational diagram of the CH3OH lines around 252 GHz observed between 3.7 and 6.7 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.

Fig. C.8 Rotational diagram of the CH3CN lines at 147 and 165 GHz observed between 4.9 and 7.0 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.

Fig. C.9 Rotational diagram of the CH3CN lines at 257 GHz observed between 3.7 and 6.7 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.

Fig. C.10 Rotational diagram of all the NH2CHO lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF) at 1 mm and 2 mm. Some lines with the same J and Ka levels (and/or similar energy levels and Einstein A coefficients) have been grouped together (Kc given as 0 in the third quantum number of the line). Scales as in Fig. C.5.

Fig. C.11 Rotational diagram of all the HNCO lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF). The lines with the same J for Ka = 1 levels have been grouped together. Scales as in Fig. C.5.

Fig. C.12 Rotational diagram of all the HCOOH lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF). Scales and grouping of lines as in Fig. C.10.

Fig. C.13 Rotational diagram of all the CH3CHO lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF). Scales and grouping of lines as in Fig. C.10.

References

All Tables

All Figures

	Fig. 1 Coarse map of the HCN(3–2) line integrated intensity in comet C/2021 A1 on 13.4 December 2021. The beam size is 9.3″. Solar phase angle was 158.6°. The direction of the Sun (PA=193.1°) is indicated at the lower right.
In the text

Fig. 2 Millimetre lines observed with IRAM-30m in comet C/2021 A1 during 12–16 November 2021 (average intensity). The CH3OH line is the sum of the J=1 to 4 J1 − J0 E lines at 165050.175, 165061.130, 165099.240, and 165190.475 MHz observed with the high resolution backend VESPA. The vertical axis is main beam brightness temperature in K and the horizontal axis is Doppler velocity in the rest frame of the comet with respect to the main line.

In the text

	Fig. 3 Individual lines observed with IRAM-30m in comet C/2021 A1 during 8–13 December 2021 (weighted averages). The vertical axis is main beam brightness temperature in K and the horizontal axis is Doppler velocity in the rest frame of the comet with respect to the line. The position and relative intensity of the hyperfine components of the HCN(3–2) line have been drawn below the line.
In the text

	Fig. 4 Map of the average spectra of the H2O(110−101) line obtained with Odin between 19.45 and 20.16 January 2023 in comet C/2022 E3 (ZTF), as a function of the pointing offset (in arcsec). A baseline in red is plotted on each spectrum. Solar phase angle was 61.6°, and the direction of the Sun (PA=120.2°) is indicated at the upper right with scales for individual AC2 spectra.
In the text

	Fig. 5 Coarse (3×3 points) map of the HCN(3–2) line integrated intensity on 4.67 February 2023 in comet C/2022 E3. Solar phase angle was 45.5°. The direction of the Sun (PA=287°) is indicated at the upper right.
In the text

Fig. 6 Individual lines observed with IRAM-30m in comet C/2022 E3 during 3–6 February 2023 (average intensity, excluding offset positions). The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet with respect to the line. A galactic contamination of CO has been blanked out. The position and relative intensity of the hyperfine components of the HCN(3–2), HCN(2–1), and H13CN(3–2) lines have been drawn below the lines.

In the text

	Fig. 7 Selected averages of the OH lines observed at the NRT in comet C/2021 A1 (Leonard) in November-December 2021 (averages of the 1667 and 1665 MHz lines scaled to 1667 MHz).
In the text

	Fig. 8 Time evolution of comet C/2021 A1 (Leonard), single days, and averages of selected periods in November-December 2021. The OH production rates from the NRT are plotted in black. The daily water production rates inferred from SOHO/SWAN observations of the wide H Lyman-α coma, which tends to smooth out short term variations, are plotted in red. (Combi et al. 2023a).
In the text

Fig. 9 Average spectrum of the HCN(3–2) line observed from 8.5 to 13.4 December in comet C/2021 A1, with simulated profiles. The model in blue assumes a production rate of 2.1 × 1025 molecules s−1 at 0.88 km s−1 on the sunward hemisphere and 1.1 × 1025 molecules s−1 at 0.58 km s−1 on the opposite hemisphere mostly facing the observer. The mean phase angle is ~140°, and the tilt of 40°(or 140°) with respect to the comet-observer line of sight is taken into account in this 3D simulation, as depicted in the upper right. The model in purple uses variable velocities in both hemispheres (see text), with production rates of 2.3 × 1025 molecules s−1 and 0.9 × 1025 molecules s−1, respectively. The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet.

In the text

Fig. 10 Average FTS spectrum of the HCN(3–2) line observed from 3.7 to 6.7 February in comet C/2022 E3 with simulated profiles. The model in blue assumes a production rate of 3.0 × 1025 molecules s−1 at 0.76 km s−1 on the sunward hemisphere and 1.5 × 1025 molecules s−1 at 0.52 km s−1 on the other hemisphere, as depicted in the upper right. The model in purple uses variable velocities in both hemispheres (see text), with production rates of 3.0 × 1025 molecules s−1 and 1.0 × 1025 molecules s−1, respectively. The mean phase angle is ~45°. The vertical axis is main beam brightness temperature in K, and the horizontal axis is Doppler velocity in the rest frame of the comet.

In the text

Fig. 11 Average spectrum of the CH3OH line at 266.838 GHz observed from 3.7 to 6.7 February 2023 in comet C/2022 E3 at the same time as HCN(3–2). The model in blue assumes a production rate of 5.6 × 1026 molecules s−1 at 0.76 km s−1 on the sunward hemisphere and 2.8 × 1026 molecules s−1 at 0.52 km s−1 on the other hemisphere, as depicted in the upper right, while the red profile comes from symmetric outgassing at 0.68 km s−1 and same total production rate (8.4 × 1026 molecules s−1). The model in purple uses the same variable velocities in both hemispheres as in Fig. 10, with production rates of 6.4 × 1025 molecules s−1and 2.0 × 1025 molecules s−1, in sunward and other hemispheres, respectively. The mean phase angle is ~45°. The vertical axis is main beam brightness temperature in K, and the horizontal axis is Doppler velocity in the rest frame of the comet.

In the text

	Fig. 12 Water production rates (in molecules s−1) of comet C/2022 E3 (ZTF) between 19.3 and 20.3 January 2023 derived from observations with Odin (central pointings). A best fit sinusoidal variation is plotted in blue.
In the text

	Fig. 13 Methanol production rates (in molecules s−1) of comet C/2022 E3 (ZTF) between 3.7 and 7.1 February 2023 derived from observations with IRAM 30m. An average 5% calibration uncertainty has been added to the formal uncertainty due to system temperature noise. A best fit sinusoidal variation is plotted in blue.
In the text

	Fig. 14 Production rates of comet C/2021 A1 (Leonard) in November and December 2021. Water production rates are 1.1 × QOH from Sec. 2.4. Afρ were measured from images of the comet by N. Biver, not corrected for the phase angle.
In the text

	Fig. 15 Production rates of comet C/2022 E3 (ZTF) in January and February 2023. Water production rates are from Odin observations (Sec. 2.2, Fig. 12). We note that Afρ(0°) were measured from images of the comet by N. Biver and have been corrected for the phase angle.
In the text

Fig. 16 Spatial distribution of formaldehyde in comet C/2022 E3 (ZTF). This plot shows for the average of all observations of the twin 211 GHz and 226 GHz lines observed together in comet C/2022 E3 (ZTF) from 3 to 7 February 2023: (i) in black squares (with error-bars) the observed line integrated intensity, (ii) in solid and dotted lines the expected intensity for parent scale length of 0 or 1000–8000 km. For better visibility a normalisation has been applied: at each offset the line intensity is multiplied by the column density at the first offset divided by the one expected at the considered offset point, for a parent distribution. Production rate of the models are adjusted to get the same integrated intensity at the first point. Deviation from a flat line for Lp = 0 (black continuous line) is due to excitation effects. Vertical axis is normalised integrated intensity in mK km/s and horizontal axis is pointing offset converted into km at the distance of the comet.

In the text

	Fig. 17 Abundances relative to water of HNCO (green) and HCOOH (blue) in comets as a function of the heliocentric distance. Even though there is dispersion between comets, the trend appears to be that these species are more abundant in the coma close to the Sun.
In the text

	Fig. A.1 CH3CN series of four lines observed with IRAM-30m in comet C/2021 A1 on 8–13 December 2021 (average intensity). The vertical axis is main beam brightness temperature in K, horizontal axis is Doppler velocity in the rest frame of the comet with respect to the (J, 0) − (J − 1, 0) line, with frequencies in the rest frame of the comet given on the upper axis.
In the text

	Fig. A.2 CH3OH series of lines around 166, 242, and 252 GHz observed with IRAM-30m in comet C/2021 A1 on 8–13 December 2021 (average intensity). The vertical axis is main beam brightness temperature in K, horizontal axis is Doppler velocity in the rest frame of the comet with respect to the main line, with frequencies in the rest frame of the comet given on the upper axis.
In the text

Fig. A.3 Spectra of molecules obtained by averaging several lines observed with IRAM-30m in comet C/2021 A1 between 8 and 13 December 2021. The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet. The number of lines averaged is provided for each molecule, either in the 2 mm band (147–153 and 163–171 GHz) or 1 mm band (209–272 GHz).

In the text

	Fig. A.4 CH3CN series of lines at 147, 165, 220, and 257 GHz observed with IRAM-30m in comet C/2022 E3 on 3–6 February 2023 (average intensity). The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet with respect to the main line, with frequencies in the rest frame of the comet given on the upper axis.
In the text

	Fig. A.5 CH3OH series of lines around 165, 242, and 252 GHz observed with IRAM-30m in comet C/2022 E3 on 3–6 February 2023 (average intensity). The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet with respect to the main line, with frequencies in the rest frame of the comet given on the upper axis.
In the text

Fig. A.6 Spectra of molecules obtained by averaging several lines observed with IRAM-30m in comet C/2022 E3 between 3 and 6 February 2023. The vertical axis is main beam brightness temperature in K. The horizontal axis is the Doppler velocity in the rest frame of the comet. The number of lines averaged is provided for each molecule, either in the 2 mm band (147–155, 163–171, and 176–184 GHz) or 1 mm band (209–272 GHz).

In the text

	Fig. C.1 Rotational diagram of the CH3OH lines around 166 GHz observed between 8.6 and 13.7 December 2021 in comet C/2021 A1 (Leonard).
In the text

	Fig. C.2 Rotational diagram of all the NH2CHO lines observed between 8.6 and 13.7 December 2021 in comet C/2021 A1 (Leonard). Lines with same J number and Ka = 0 to 2 or 3 to 4 are averaged together. Scales as in Fig. C.1). The values in green are those expected from modelling with T = 60 K.
In the text

	Fig. C.3 Rotational diagram of the CH3CN lines at 147, 165, 220, 239, and 257 GHz observed between 8.4 and 13.7 December 2021 in comet C/2021 A1 (Leonard). Scales as in Fig. C.1). The values in green are those expected from modelling with T = 60 K.
In the text

	Fig. C.4 Rotational diagram of the HNCO lines at 1 mm and 2 mm observed between 8.4 and 13.7 December 2021 in comet C/2021 A1 (Leonard). Scales as in Fig. C.1.
In the text

	Fig. C.5 Rotational diagram of the CH3OH lines around 166 GHz observed between 4.9 and 7.0 February 2023 in comet C/2022 E3 (ZTF). The values in green are those expected from modelling with T = 60 K.
In the text

	Fig. C.6 Rotational diagram of the CH3OH lines at 242 GHz observed between 4.7 and 6.8 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.
In the text

	Fig. C.7 Rotational diagram of the CH3OH lines around 252 GHz observed between 3.7 and 6.7 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.
In the text

	Fig. C.8 Rotational diagram of the CH3CN lines at 147 and 165 GHz observed between 4.9 and 7.0 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.
In the text

	Fig. C.9 Rotational diagram of the CH3CN lines at 257 GHz observed between 3.7 and 6.7 February 2023 in comet C/2022 E3 (ZTF). Scales as in Fig. C.5.
In the text

	Fig. C.10 Rotational diagram of all the NH2CHO lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF) at 1 mm and 2 mm. Some lines with the same J and Ka levels (and/or similar energy levels and Einstein A coefficients) have been grouped together (Kc given as 0 in the third quantum number of the line). Scales as in Fig. C.5.
In the text

	Fig. C.11 Rotational diagram of all the HNCO lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF). The lines with the same J for Ka = 1 levels have been grouped together. Scales as in Fig. C.5.
In the text

	Fig. C.12 Rotational diagram of all the HCOOH lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF). Scales and grouping of lines as in Fig. C.10.
In the text

	Fig. C.13 Rotational diagram of all the CH3CHO lines observed between 3.6 and 7.1 February 2023 in comet C/2022 E3 (ZTF). Scales and grouping of lines as in Fig. C.10.
In the text