WorksPhysics and Astronomy
Detection of HNCOH+ in the interstellar medium
Abstract
Only five out of the 350 molecules detected in the interstellar medium (ISM) carry the isocyanate functional group (NCO), the essential substructure of peptide bonds in proteins. Among them are isocyanic acid (HNCO) and its most stable N-protonated isomer HNCO. Previous studies have predicted that the O-protonated isomer HNCOH should be present in the ISM; yet this species has remained elusive. We search for the metastable HNCOH towards the Galactic Center molecular cloud G+0.693-0.027, where HNCO has already been detected. We use an ultra-deep broadband spectral survey carried out towards G+0.693-0.027 with the Yebes 40-m, IRAM 30-m and APEX telescopes covering nearly 150 GHz of bandwidth between 31 and 275 GHz. We report the first detection of HNCOH in the ISM with a column density of = (9.9 0.5)10 cm, which corresponds to a molecular abundance relative to H of (7.3)10. With the measured HNCO column density of = (2.6 0.3) cm, this yields a HNCOH/HNCO ratio of 3.8. This result is consistent with recent kinetic calculations that favour the formation of HNCOH over HNCO. The detection of HNCOH in the ISM broadens the inventory of molecules known to contain the isocyanate functional group and reinforces recent works showing that isomer abundances in the ISM do not necessarily correlate with thermodynamic stability.
Introduction
A substantial number of molecules detected in space can be assembled from only the four elements H, C, N, and O. These elements constitute the fundamental building blocks of organic and prebiotic chemistry and, when arranged appropriately, give rise to key functional groups such as carboxyl (-COOH) and amino (-NH), present in biologically relevant molecules such as amino acids. Alternatively, they can link together forming a peptide bond -C(=O)-N(H)-, enabling the formation of large proteins [13]. The detection of simple molecules containing the isocyanate group (-N=C=O) in the interstellar medium (ISM) is particularly significant, as they may offer insight into the earliest chemical steps leading to peptide synthesis.
Within the current inventory of molecules identified in the ISM or in circumstellar envelopes1, only a very small subset carries the isocyanate functional group. To date, this subset includes: isocyanic radical (NCO; [13]); isocyanic acid (HNCO; [34]); N-protonated isocyanic acid (HNCO; [13]); methyl isocyanate (CHNCO; [5]); and ethyl isocyanate (CHNCO; [26]).
HNCO exhibits several metastable isomers in the singlet ground electronic state: HOCN, HCNO, and HONC, which lie at 24.7, 70.7, and 84.1 kcal mol (12430, 35578, and 42321 K) above HNCO, respectively [31]. These species can be formed through a variety of chemical pathways. In the gas phase, the protonation and subsequent hydrogenation of the radicals NCO and CNO yield the protonated [C,H,N,O] ions, including HNCO, HNCOH, HCNOH, HOCN, and HCNO [19, 6] (Marcelino et al. 2010; Quénard et al. 2018). By dissociative recombination (DR) reactions, these ions produce all the [C,H,N,O] isomers, among them HONC, the highest-energy isomer, which has not yet been detected in space [13].
Quantum-chemical calculations reveal that the N-protonated variant of isocyanic acid, HNCO, is the global minimum, followed by the O-protonated variant HNCOH (18 kcal mol, K; [10, 6]). Furthermore, Murakami et al. [17] investigated the thermal rate coefficients and branching ratios for the gas-phase reactions H + HNCO H + HNCOH/HNCO which is expected to be one of the main pathways for forming these species, since protonation by H in the ISM has been shown to be one of the most efficient mechanisms [18]. Their calculations showed that the channel leading to HNCOH is favoured over the formation of HNCO, with branching ratios of and , respectively. These results are consistent with laboratory rotational spectroscopy measurements, which reveal that the abundance of HNCOH in the original Fourier-transform microwave spectroscopy (FTMW) study is twice that of HNCO [10]. Moreover, the proton-affinity-induced isomerisation from the meta-stable HNCOH to the stable HNCO is unlikely to occur due to the substantial high barrier (100 kcal mol, K; [7]) between them. Despite this, only HNCO has been detected in the ISM [13, 26]. Therefore, the detection of HNCOH yields a key step that had been missing in our understanding of the interstellar chemistry of HNCO and its protonated isomers.
In this Letter, we report the first detection of HNCOH in the ISM towards the Galactic Center molecular cloud G+0.693-0.027 (hereafter, G+0.693), a source known for its exceptionally rich chemistry driven by large-scale shocks, likely induced by a cloud–cloud collision [38]. These shocks efficiently sputter material from both the icy mantles and refractory cores of dust grains, injecting large amounts of molecular material into the gas phase. As a result, G+0.693 stands out as one of the most prolific astronomical environments for identifying new interstellar molecules.
Observations
We have used an ultra-deep broadband spectral survey obtained towards the molecular cloud G+0.693 at 7 mm, 3 mm, 2 mm, and 1 mm with the Yebes 40-m, IRAM 30-m, and APEX radio telescopes. The Yebes 40-m observations were conducted using the ultra-broadband Nanocosmos Q-band (7 mm) HEMT receiver to cover the entire Q-band frequency range –50.4 GHz (projects 20A008 and 21A014). The IRAM 30-m observations were carried out using the Eight MIxer Receiver (EMIR) and the Fast Fourier Transform Spectrometer (FFTS) with 200 kHz of resolution across the spectral ranges 71.8–116.7 GHz, 124.77–175.5 GHz, 199.8–238.29 GHz, 252.520–260.30 GHz, and 268.2–275.98 GHz (projects 172-18, 018-19, 133-19 and 123-22), providing complementary coverage. For the APEX observations, the NFLASH230 receiver was connected to two FFTS backends, which provided a simultaneous coverage of two sidebands of 7.9 GHz each separated by 8 GHz. The covered spectral ranges were 217.9–225.9, 234.18–242.18, 243.94–252.12, and 260.18–268.37 GHz (see Rivilla et al. [21], project E-0108.C-0306A-2021). Data were collected in position-switching mode, centered at , , with the off position shifted by and .
The spectra of Yebes, IRAM and APEX observations were finally smoothed to a frequency resolution of 256 kHz (1.5-2.5 km s), 615 kHz (1.0-2.2 km s), and 1 MHz (1.1-1.4 km s), respectively. The half power beam width (HPBW) was –55 for Yebes 40-m, –29 for IRAM 30-m, and –28 for APEX in the observed frequency ranges. Note, however, that the molecular emission towards G+0.693 is extended over the beam of the three telescopes [39] (Li et al. 2020) and hence, the measured spectra are given in units of antenna temperature (). We refer to Rivilla et al. [21, 24] and Sanz-Novo et al. [28] for further information on all observations and data reduction.
Detection of HNCOH
The astronomical line identification of HNCOH has been accomplished using frequency predictions based on the reported laboratory microwave spectrum observed by Lattanzi et al. [10]. As only the two lowest a-type rotational transitions – and – at about 20 and 40 GHz have been studied in the laboratory so far [10], the spectroscopic entry was extrapolated to 300 GHz using SPCAT, and then incorporated into MADGUBA package (version 12.1.1). The identification of the molecular lines was carried out using the Spectral Line Identification and Modelling (SLIM) tool from MADGUBA [14], which assumes local thermodynamic equilibrium (LTE) to compute the synthetic spectra. To derive the physical parameters of the emission, we used the AUTOFIT tool of SLIM that provides a nonlinear least-squares LTE fit to the data using the Levenberg-Marquardt algorithm. The fit provides information about the column density, , excitation temperature, , radial velocity, , and line width, FWHM, of the molecular emission. We note that owing to the typically broad FWHM of the molecular emission observed in G+0.693 (line width of 15-22 km s; e.g. [8, 23]), the quadrupole hyperfine structure from the presence of the N () nucleus is not resolved but collapsed to just one line.
Eight a-type rotational transitions (i.e., with and ; D) of HNCOH are covered by the observational survey, spanning much of the progression with to 13. However, note that the predicted frequencies for the transitions with , 6, 9 and 10 fall within the gaps of the observed frequency coverage of our survey (see Appendix § 6). Moreover, the – transition at 258.9 GHz was excluded from the analysis due to blending with an atmospheric line.
The detected transitions are shown in Figure 1 and their spectroscopic information is listed in Table B.1. The best LTE fit provided a value of FWHM = 21.9 1.4 km s, and a velocity of = 66.9 0.6 km s, consistent with the values found in G+0.693 for other molecular species, specially molecular ions (see e.g. Sanz-Novo et al. 2024). The excitation temperature derived is = 24.4 1.3 K, while the derived column density is = (9.9 0.5) cm. Assuming an H column density of = (1.35 0.20) cm inferred from CO emission at the velocity component of G+0.693 [15], the HNCOH column density translates into a molecular abundance with respect to H of towards G+0.693. As shown in Table B.1, the detected transitions cover a wide range of upper-level energies, , which allowed us to derive an accurate estimate on . The consistency across these features, together with the detection of a nearly complete rotational progression within the spectral coverage of the G+0.693 survey, yields a robust identification of HNCOH.
Figure 1. rotational transitions of HNCOH identified towards G+0.693 and sorted by increasing frequency (see also Table B.1). The LTE fit of HNCOH is plotted with a red line, while the blue line depicts the emission from all the molecules identified to date in our survey, including HNCOH (observed spectra shown as gray histograms). The molecular structure of HNCOH is shown in the bottom right. The quantum numbers of each rotational transition () are indicated in the upper left of each panel. Blue labels show the rest of identified species contributing to the observed spectra.
The astronomical lines observed here, except transition –, represent a first record of the millimetre-wave spectrum, identified in space towards G+0.693 and not measured before in the laboratory. This implies that, although the uncertainties can be as high as 1.2 MHz at 238 GHz, which translates into velocity uncertainties between 1.0–1.5 km s, these would not have an impact on the analysis as they are small compared to the typical line widths of the molecular line emission measured towards G+0.693 (between 15–22 km s). An analogous extrapolation approach was used in the interstellar discovery of HOCS (Sanz-Novo et al. 2024), and then corroborated in the laboratory by Lattanzi et al. [9]. Furthermore, the inclusion of the isolated –, –, and – transitions together with the microwave data set by Lattanzi et al. [10] provides a first estimate of the centrifugal distortion constant of MHz. This value is in very good agreement with the value of MHz obtained from the harmonic force field calculated at the CCSD(T)/cc-pVQZ level of theory [10]. Nevertheless, a comprehensive description of the pure rotational spectrum still requires a thorough targeted laboratory study.
We have also performed a complementary rotational diagram analysis (Goldsmith & Langer 1999) with MADCUBA, using the transitions shown in Figure 1. As expected, given the optically thin emission of the observed transitions, the derived physical parameters of HNCOH are consistent with the ones obtained from the Autofit analysis: cm and K (Figure C.1).
Discussion
To understand the implications of finding interstellar HNCOH in the chemistry of HNCO, we compare the abundance of HNCOH with HNCO. Due to the availability of new, higher-sensitivity data for the studied source as compared to those of Rodríguez-Almeida et al. [26], we have recalculated the LTE fit of HNCO for G+0.693 using the same transitions reported by Rodríguez-Almeida et al. [26], which yield a column density of cm, consistent with previous results within a factor of 1.5 (see Appendix § 9). Thus, we find a HNCOH/HNCO ratio of towards G+0.693. The quantum-chemical calculations employed by Lattanzi et al. [10] show that the HNCO isomer is the most favourable product from a thermodynamic point of view, with a difference in energy of 18 kcal mol (9060 K). Following the minimum energy principle (MEP) formulation [11], the abundance ratios between structural isomers must be governed by thermodynamic equilibrium. Therefore, the abundance of HNCOH based on thermodynamical considerations is predicted to be more than 20 orders of magnitude lower in comparison to HNCO at the of G+0.693 (70–140 K; Zeng et al. [36]). Thus, the derived abundance of HNCOH must be explained in terms of kinetics instead. The theoretical kinetic results derived by Murakami et al. [17] indicate that the gas-phase reactions H + HNCO H + HNCOH/HNCO preferentially proceed via the mechanism leading to the less stable isomer, HNCOH, rather than to the more thermodynamically favourable product HNCO. This result does not emerge as the unique exception for the MEP, since recent detections in the ISM of many other isomeric families have demonstrated that high-energy structural isomers can share similar abundances to their more stable counterparts (e.g. Shingledecker et al. [33]; García de la Concepción et al. [3]; San Andrés et al. [27]; Zeng et al. [37]; Rivilla et al. [25]).
Furthermore, we compared the abundance ratios of HNCOH and HNCO over HNCO [36]. The HNCOH/HNCO ratio is , whereas the HNCO/HNCO is . Taking into account that the proton affinity of HNCO is 759 kJ mol [35], this result is consistent with the abundance ratios found in cold dense clouds for neutral molecules with proton affinities above 700 kJ mol [1].
Moreover, it is worth noting that the of HNCOH is higher than the one found for HNCO ( of 24 K vs. 9 K, respectively), which is consistent with a similar trend observed in for other O-protonated cations in G+0.693 (Sanz-Novo et al. 2024). A possible explanation for this lies in the different dipole moments. Molecules with larger dipole moments have correspondingly larger Einstein coefficients (), which imply higher critical densities. This is expected to yield lower because the emission is likely sub-thermally excited given that the H gas densities in G+0.693 are of a few cm [2]. The dipole moment of HNCO is D, whereas the dipole moment components of HNCOH are D and D [10]. Therefore, the higher value found for HNCOH is consistent with its lower dipole moment compared to HNCO.
Conclusion
We report the first detection of HNCOH in the ISM towards the Galactic Center molecular cloud G+0.693. The derived molecular abundance is with respect to H. This species represents the sixth identified molecule containing the isocyanate functional group, the backbone structure of peptides. Moreover, the detection reveals that the fractional abundance of HNCOH is approximately four times larger than that of HNCO, the most stable isomer of isocyanic acid. This finding demonstrates that, within the ISM, the thermodynamically most stable isomers are not necessarily the most abundant ones.
As the pioneering microwave laboratory study of Lattanzi et al. [10] covered only the two lowest rotational transitions of HNCOH, the present study provides first insight into the millimetre-wave spectrum. The centrifugal distortion parameter, derived by considering three virtually unblended transitions (Table B.1), is in near-perfect agreement with the value derived from high-level force-field calculations. A more detailed characterization of the rotational spectrum, including transitions with , will require dedicated laboratory studies in the future.
Acknowledgements. This work is supported by ERC grant OPENS, GA No. 101125858, funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. M.R.-M., I.J.-S. and A.M. acknowledge funding from the Spanish National Research Council (CSIC) through the Bilateral project SOULMATE (BIL2023-25017). I.J.-S., V.M.R, L.C., M.S.-N, A.M. and D.S.A acknowledge funding from grant PID2022-136814NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities / State Agency of Research MICIU/AEI/10.13039/501100011033 and by “ERDF/EU”. M.S.-N. acknowledges funding from the Marie Sklodowska Curie fellowship MSCA-PF 101272259 funded by the European Union. V.M.R, M.S.-N and D.S.A also acknowledge the grant CNS2023-144464 funded by MICIU/AEI/10.13039/501100011033 and by “European Union NextGenerationEU/PRTR”. D.S.A. thanks the financial support provided by the Comunidad de Madrid through the Grant PIPF-2022/TEC-25475. D.S.A. and V.M.R. acknowledge the funds provided by the Consejo Superior de Investigaciones Científicas (CSIC) and the Centro de Astrobiología (CAB) through the project
20225AT015 (Proyectos intramurales especiales del CSIC). S.Z. and A.M.-H. acknowledges the support by RIKEN Special Postdoctoral Researchers Program. B.T. thanks projects PID2022-137980NB-I00 and PID2023-147545NB-I00 funded by the Spanish Ministry of Science and Innovation/State Agency of Research MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe". At Cologne, this work has been supported by an ERC Advanced Grant (MissIons: 101020583) and by the Deutsche Forschungsgemeinschaft (DFG) via the Collaborative Research Centre 1601 (project ID: 500700252, subprojects A4 and B8). L.C. acknowledges support of a fellowship from the “la Caixa” Foundation (ID 100010434). The fellowship code is LCF/BQ/PR25/12110012.
References
Spectral coverage of the observations
In the following table we show the central frequency of the rotational () level progression in the ladder for HNCOH. For a-type transitions in the R-branch (i.e. ), the line frequencies are approximately distributed according to , where and are the rotational constants. Using the value GHz reported by Lattanzi et al. [10], we computed the expected central frequency for each transition in the ladder (Table A.1). It can clearly be seen that the missing transitions in the analysis correspond to the gaps in the observational data. However, it is worth noting that 3 of the missing transitions lie in the absorption bands of the atmosphere (marked with an asterisk in Table A.1) and therefore, cannot be observed with ground-based radio telescopes.
| Frequency (GHz) | Telescope | |
| 2 - 1 | 39.8 | Yebes 40-m |
| 3 - 2 | 59.7 | * |
| 4 - 3 | 79.6 | IRAM 30-m |
| 5 - 4 | 99.5 | IRAM 30-m |
| 6 - 5 | 119.5 | * |
| 7 - 6 | 139.4 | IRAM 30-m |
| 8 - 7 | 159.3 | IRAM 30-m |
| 9 - 8 | 179.3 | * |
| 10 - 9 | 199.2 | - |
| 11 - 10 | 219.1 | APEX |
| 12 - 11 | 239.0 | APEX |
| 13 - 12 | 258.9 | IRAM 30-m |
Table A.1. Spectral coverage of the ladder for HNCOH.
NOTE. The ‘-’ symbol corresponds to the gaps in the observational data, whereas the ‘*’ symbol corresponds to transitions that lay in the atmospheric absorption bands, which were not covered either in our observations.
Spectroscopic information of HNCOH detection.
In Table B.1, we report the rotational transitions of HNCOH computed from the spectroscopic constants of Lattanzi et al. [10] and considered in the present study. We note that the N nuclear quadrupole hyperfine structure has been omitted, as it cannot be resolved owing to the broad line widths of the molecular emission observed in G+0.693. All transition frequencies but that of the transition are extrapolated based on the data given in Lattanzi et al. [10], for which the uncertainties can be as high as 1.2 MHz at 238 GHz, which translates into velocity uncertainties between 1.0-1.5 km s. These uncertainties are however small compared to the line widths of 15-22 km s measured for the molecular emission toward G+0.693 ([36, 8]).
To identify the unblended transitions, we adopt the criteria used in recent molecular detections towards G+0.693 (Jiménez-Serra et al. 2026). A spectral feature is considered predominantly unblended when its positive residual contribution (reported as residual excess in Table B.1) over the velocity interval km s, is . Following this criterion, we identify 4 unblended spectral features (see Table B.1). However, if we account for the emission of the species already detected towards G+0.693 (reported as "Residual All species" in Table B.1), we find that 5 features blended with known molecules are well reproduced in the observed spectra, with a contribution on the residual of . Note that features blended with only unidentified species (U-lines in Table B.1) remain blended under this criterion, although they are fully consistent with the observed spectra.
Rotational diagram
We have carried out a rotational diagram analysis (Figure C.1) of the emission of HNCOH with MADCUBA excluding transition - at 238.91 GHz due to heavy blending with unknown species. The calculation was done considering the observed velocity-integrated intensity over the line width (Rivilla et al. [22]). As described in San Andrés et al. [27], MADCUBA’s rotational diagram also considers the predicted emission from all of the molecular species previously detected in the source, and accounts for the blending in each transition by subtracting the predicted line contribution of the blending species from the observed spectrum.
Figure C.1. Rotation diagram of HNCOH towards G+0.693 (red dots). The black dashed line corresponds to the best linear fit to the data points. The derived values for the molecular column density, , the excitation temperature, , and the fitting correlation coefficient, , are shown in blue.
The derived physical parameters of HNCOH from the rotational diagram are: cm and K.
Re-analysis of HNCO towards G+0.693
We have revised the previously detected HNCO towards G+0.693. Following the approach carried out by Rodríguez-Almeida et al. [26] for HNCO, we carry out the analysis for the same transitions reported in that work, separating the ortho NOTE. We provide the transition frequencies, quantum numbers, base 10 logarithms of the integrated intensity at 300 K (log ), energies of the upper levels (), noise level (rms), integrated area (), signal to noise ratio (S/N) calculated with the integrated intensity and residual of each transition. The last column indicates the line blending with other molecular species.
| Frequency (GHz) | Transition () | (nm MHz) | (K) | rms (mK) | (mK km s) | S/N | Residual excess | Residual All species | Blending |
| 39.8236960 | -4.8525 | 2.6 | 0.37 | 86 | 16 | 26% | 25% | U-line + aGg′-(CHOH) | |
| 79.6467092 | -3.9578 | 7.8 | 1.9 | 268 | 14 | 54% | 39% | sp-CHNHCHO + U-line | |
| 99.5577465 | -3.6733 | 11.4 | 1.2 | 347 | 32 | 8% | 7% | Unblended | |
| 139.3784555 | -3.2515 | 20.6 | 1.3 | 416 | 42 | 0.8% | 0.8% | Unblended | |
| 159.2879566 | -3.0879 | 26.2 | 2.1 | 399 | 27 | 25% | 11% | CHCOCH | |
| 219.0120224 | -2.7124 | 47.0 | 1.2 | 236 | 32 | 18% | 17% | U-line + NHCHCHOH | |
| 238.9182808 | -2.6150 | 55.3 | 1.3 | 176 | 23 | 47% | 47% | U-line |
Table B.1. List of observed transitions of HNCOH in order of increasing frequency.
( = odd) and para ( = even) states (energy difference of K; Gupta et al. [4]). We used the Madcuba script reported in Zeng et al. [37] to perform the ladder separation analysis. It first separates the ladder automatically, and then, for each ladder, the script corrects the energy of the rotational levels by subtracting the energy of the lower-state energy level. Finally, it recalculates the partition function as a direct summation of all the energy levels within each ladder. The Autofit tool was then applied to obtain the best-fitting LTE model for the transitions of each ladder, and the total column density was derived as a sum of the column density of each ladder.
For the o-HNCO state we obtain cm, K, with fixed value of km s and FWHM = 18 km s to allow the convergence. For the para-state, we assumed the same and obtained cm with fixed km s and FWHM = 18 km s.
These results yield a total column density of cm for HNCO, derived from the summed contributions of the para and ortho states, which translates into a molecular abundance with respect to H of . This is consistent with the column density derived by Rodríguez-Almeida et al. [26], cm with a factor of 1.5. Note that the difference arises from the recalculation of the partition function taking into account the correction of the energy of the rotational levels for the separation of the ladder as well as the new astronomical data. This yields a slightly higher compared to that of Rodríguez-Almeida et al. [26], but consistent within the error with the value derived in that work.
Note that the associated uncertainty for all calculations have been calculated using the Python library RichValues (Megías et al. [16]), and arise from the error propagation of the observational data.
References
- [1]Agúndez, M., Cabezas, C., Marcelino, N., et al. 2022, A&A, 659, L9
- [2]Colzi, L., Martín-Pintado, J., Zeng, S., et al. 2024, A&A, 690, A121
- [3]García de la Concepción, J., Jiménez-Serra, I., Carlos Corchado, J., Rivilla, V. M., & Martín-Pintado, J. 2021, ApJ, 912, L6
- [4]Gupta, H., Gottlieb, C. A., Lattanzi, V., Pearson, J. C., & McCarthy, M. C. 2013, ApJ, 778, L1
- [5]Halfen, D. T., Ilyushin, V. V., & Ziurys, L. M. 2015, ApJ, 812, L5
- [6]Hochlaf, M., Lingueri, R., Kandi, N. C., Alessandrini, S., & Puzzarini, C. 2026, Journal of Physical Chemistry A, 130, 3140
- [7]Ijjaali, F., Alcami, M., Mó, O., & Yáñez, M. 2001, Molecular Physics, 99, 1129
- [8]Jiménez-Serra, I., García de la Concepción, J., Cuppen, H. M., et al. 2020, Nature Astronomy
- [9]Lattanzi, V., Sanz-Novo, M., Rivilla, V. M., et al. 2024, A&A, 689, A260
- [10]Lattanzi, V., Thorwirth, S., Gottlieb, C. A., & McCarthy, M. C. 2012, Journal of Physical Chemistry Letters, 3, 3420
- [11]Latteais, M., Pauzat, F., Ellinger, Y., & Ceccarelli, C. 2009, ApJ, 696, L133
- [13]Marcelino, N., Agúndez, M., Cernicharo, J., Roueff, E., & Tafalla, M. 2018, A&A, 612, L10
- [14]Martín, S., Martín-Pintado, J., Blanco-Sánchez, C., et al. 2019, A&A, 631, A159
- [15]Martín, S., Requena-Torres, M. A., Martín-Pintado, J., & Mauersberger, R. 2008, ApJ, 678, 245
- [16]Megías, A., Jiménez-Serra, I., Dulieu, F., et al. 2025, A&A, 702, A87
- [17]Murakami, T., Nakagawa, S., Ota, H., Okada, K., & Takayanagi, T. 2024, ACS Earth and Space Chemistry, 8, 2294
- [18]Oka, T. 2006, Proceedings of the National Academy of Science, 103, 12235
- [19]Quan, D., Herbst, E., Osamura, Y., & Roueff, E. 2010, ApJ, 725, 2101
- [21]Rivilla, V. M., García De La Concepción, J., Jiménez-Serra, I., et al. 2022a, Frontiers in Astronomy and Space Sciences, 9, 829288
- [22]Rivilla, V. M., Jiménez-Serra, I., Martín-Pintado, J., et al. 2021, Proceedings of the National Academy of Science, 118, e2101314118
- [23]Rivilla, V. M., Jiménez-Serra, I., Martín-Pintado, J., et al. 2022b, Frontiers in Astronomy and Space Sciences, 9, 876870
- [24]Rivilla, V. M., Sanz-Novo, M., Jiménez-Serra, I., et al. 2023, ApJ, 953, L20
- [25]Rivilla, V. M., Sanz-Novo, M., & San Andrés, D. 2026, ACS Earth and Space Chemistry, 10, 1155
- [26]Rodríguez-Almeida, L. F., Rivilla, V. M., Jiménez-Serra, I., et al. 2021, A&A, 654, L1
- [27]San Andrés, D., Rivilla, V. M., Colzi, L., et al. 2024, ApJ, 967, 39
- [28]Sanz-Novo, M., Rivilla, V. M., Jiménez-Serra, I., et al. 2023, ApJ, 954, 3
- [31]Schuurman, M. S., Muir, S. R., Allen, W. D., & Schaefer, H. F. 2004, J. Chem. Phys., 120, 11586
- [33]Shingledecker, C. N., Álvarez-Barcia, S., Korn, V. H., & Kästner, J. 2019, ApJ, 878, 80
- [34]Snyder, L. E. & Buhl, D. 1972, Annals of the New York Academy of Sciences, 194, 17
- [35]Wight, D. & Bostick, F. 1980, in ICASSP ’80. IEEE International Conference on Acoustics, Vol. 5, 156
- [36]Zeng, S., Jiménez-Serra, I., Rivilla, V. M., et al. 2018, MNRAS, 478, 2962
- [37]Zeng, S., Rivilla, V. M., Sanz-Novo, M., et al. 2025, A&A, 703, A73
- [38]Zeng, S., Zhang, Q., Jiménez-Serra, I., et al. 2020, MNRAS, 497, 4896
- [39]Zheng, S., Li, J., Wang, J., et al. 2024, ApJ, 961, 58

