A chemical link between methylamine and methylene imine and implications for interstellar glycine formation

Methylamine CH3NH2 is considered to be an important precursor of interstellar amino acid because hydrogen abstraction might lead to the aminomethyl radical •CH2NH2 that can react with •HOCO to form glycine, but direct evidence of the formation and spectral identification of •CH2NH2 remains unreported. We performed the reaction H + CH3NH2 in solid p-H2 at 3.2 K and observed IR spectra of •CH2NH2 and CH2NH upon irradiation and when the matrix was maintained in darkness. Previously unidentified IR spectrum of •CH2NH2 clearly indicates that •CH2NH2 can be formed from the reaction H + CH3NH2 in dark interstellar clouds. The observed dual-cycle mechanism containing two consecutive H-abstraction and two H-addition steps chemically connects CH3NH2 and CH2NH in interstellar media and explains their quasi-equilibrium. Experiments on CD3NH2 produced CD2HNH2, in addition to •CD2NH2 and CD2NH, confirming the occurrence of H addition to •CD2NH2.

The article relies on the comparison between measured IR spectra and DFT calculations. In order to identify which species is present, the authors compute the spectrum of various possible molecules. The comparison itself seems to rely on "visual inspection": lines are close in frequency and intensities etc... I found this methodology rather unreliable, as the "agreement" is subjective. I would like to encourage the authors to define more "objective" criteria (percentage of agreement or a match score) that would define more precisely allow to compare the agreement between the experimental spectrum and theoretical spectra.
The way the data are represented makes the comparison difficult to read (see for instance Figure1). For instance the authors claim: « The lines in group A at 3500.5, 3403.6, 3143.3, 3042.6, 1609.9, 1213.6, and 685.5 cm-1 106 107 agree well, in terms of wavenumbers and relative intensities » "Agree well" does not have any significance, the relative intensity and frequency match is hard to quantify by a simple visual inspection of this figure. I therefore recommend the author to use a different representation where the spectra are zoomed in and where the comparison exp/theo is direct (superimpose experimental spectra and calculated sticks).
Why calculated lines at 1300 and 1450 cm-1 are not present in experimental data? overall the work is interesting but it requires a better presentation of the results and more quantitative criteria for comparison.

Reviewer #2 (Remarks to the Author):
The work presented in the manuscript is the follow-up of previous similar investigations on other systems. As in those cases, I have found the present study very interesting and extremely relevant to the cold chemistry of interstellar objects. In particular, the authors have investigated in solid p-H2 the effects caused by exposing methylamine to hydrogen atoms. An interesting isotopic effect has also been noted when using partially deuterated methylamine. The employed experimental technique is state-of-the-art and the results are interpreted in the light of dedicated electronic structure calculations. I have only a few comments that the authors might address.
1)There has been a recent theoretical investigation on the reaction CH3 + NH2 assisted by a cluster of 18-or 33-water molecules to simulate the effect of amorphous ice. I think it would be nice to compare part of the present theoretical results with those reported by Enrique-Romero et al. as the real system we aim to understand is not only cold but features the presence of water molecules. The paper is in press in ApJSS, but can be found also here https://arxiv.org/pdf/2201.10864.pdf 2) I think that the astrophysical implications of this study have been stretched a little bit too far. I refer to the relation between methanimine and methylamine. First of all, methanimine is almost ubiquitous and has been observed in many different kinds of interstellar objects. Its first detection dates back to 1973. On the contrary, methylamine has been detected in few objects. In other words, there are many environments where methanimine is observed and methylamine is not. This is inconsistent with the mother-daughter relation that is suggested here. Non only that: there are numerous hints that methanimine can be considered a signpost for chemistry occurring in the gas phase, while there no gaseous route of methylamine formation, see, for instance, Suzuki et al. (ApJ 825, 79, 2016; https://doi.org/10.3847/0004-637X/825/1/79).
3) I think that there is an important previous study that should be cited here, that is the study on hydrogenation of solid hydrogen cyanide HCN and methanimine CH2NH at low temperature by Theule and coworkers (A&A 534, A64, 2011; https://doi.org/10.1051/0004-6361/201117494). Interestingly, in that study, the authors reported they have been unable to observe methanimine formation in the hydrogenation of HCN because it easily converts into methylamine. Apparently, they were unable to see the same effect caused by H-abstraction that has been seen here. A comment is in order.
Reviewer #3 (Remarks to the Author): Manuscript number: COMMSCHEM-22-0003-T Paper title: A chemical link between methylamine and methylene imine: identification of aminomethyl radical and implications for interstellar glycine formation Authors: Prasad Ramesh Joshi and Yuan-Pern Lee The authors report on identification of aminomethyl radical (CH<sub>2</sub>NH<sub>2</sub>) in solid p-H<sub>2</sub> matrix. The authors combined IR measurements of both normal and partially deuterated CH<sub>2</sub>NH<sub>2</sub> with quantum chemical calculations. The impressive results have a potential to accelerate our understanding of glycine formation in the interstellar molecular clouds. The authors also claim that there is a linear correlation between column densities of CH<sub>3</sub>NH<sub>2</sub> and CH<sub>2</sub>NH in the interstellar molecular clouds. However, I found major issues in the paper, described below. I would like the authors to reconsider and to fix these issues. I would like to review the revised manuscript again.
Major Comments: 1. The experimental conditions may not be suitable for comparing with the interstellar conditions. I have found that the experiments were designed very well. It is a good step to utilize partially deuterated CH<sub>3</sub>NH<sub>2</sub> towards secure identification of the CH<sub>2</sub>NH<sub>2</sub> radical. The experiments were conducted in the solid p-H<sub>2</sub> matrix at 3.2 K. This environment is very different from that in the actual physical conditions in the interstellar dust. It is widely understood in astrophysics that the surface of interstellar dust particles, especially deep inside dense molecular clouds, such as hot cores, are covered with amorphous water ice (ice mantle). Further it is also thought in astrophysics that the dust temperature is determined from the balance of their radiative heating (by absorbing far-IR interstellar radiation penetrating into the cores) and cooling (via the continuum emission). Theoretical consideration with these assumptions concluded that the dust temperature will not be lower than 8 K.
I wonder if the beautiful experimental results by the authors could be applied to actual interstellar dust particles. The temperature, 3.2 K, is much lower than the theoretical lower limit of dust temperature. Further, interaction between solid H<sub>2</sub> and CH<sub>2</sub>NH<sub>2</sub> would be different from those between water ice and CH<sub>2</sub>NH<sub>2</sub>. If the authors claim that it is possible, I would ask the authors to demonstrate the possibility with convincing rationales.
2. Applicability of the quantum chemical calculations to species in solid phase I fully understand that the quantum chemical calculations are powerful tools for understanding energies of reactants, products as well as transition states. Quantum chemical calculations are often used to predict IR spectra in the "gas phase". The experiments are conducted in the solid H<sub>2</sub> matrix, not in the gas phase. Since I was not able to find a text that the quantum chemical calculations were made for the sold phase, I wonder if the quantum chemical calculations shown in the paper considred that the reactants, products and transition state species are in the solid phase. It seems to me that the calculations correspond to the gas phase. If the quantum chemical calculations corresponded to the gas phase, the authors should not compare the calculated IR spectra and the experimental ones. If the authors made the calculations for the solid phase, I would ask the authors to add new text showing that the quantum chemical calculations are for the solid phase.

Dual-cycle mechanism
The authors propose the "dual-cycle mechanism" in Figure 4. However, the right-hand part of the figure has already been pointed out by Garrod (2013; reference 5) and Suzuki et al. (2018;reference 6). Therefore I would suggest the authors to discard Figure 4. If the authors think this figure is valid, I would like the authors to add a new text that similar considerations were made in the past by referring the papers above, and to add another text with regard to new insights contained in this figure.
In radio astronomical observations, astronomers observe atoms and molecules only in the gasphase. Suzuki et al. (2016, the Astrophysical Journal, 825, id.79) reported that CH<sub>2</sub>NH is primarily formed in gas-phase reactions, primarily between CH<sub>3</sub> and NH radicals. Solidphase CH<sub>2</sub>NH is immediately converted into CH<sub>3</sub>NH<sub>2</sub>. Thus CH<sub>3</sub>NH<sub>2</sub> is formed in the solid-phase, then are evaporated from the dust surfaces. This would mean that the apparent linear correlation has no physical meanings. I would also point out that the compiled observational results were taken by many telescopes with largely different spatial resolutions. A celestial object shows a spatial structure. When a data is taken with a large spatial resolution, it means an average of molecular distribution within that resolution. When astronomers use small (fine) spatial resolution, it is usually expected to have higher column density.

Response/revisions to the reviewers' comments
We appreciate very much the valuable comments and suggestions from the reviewers; these comments really helped us to improve the manuscript. Below are the detailed responses to the reviewers' comments. The reviewers' comments are in black, our responses are listed in blue color after each comment, and the revised text are highlighted in yellow.
Reviewer #1 (Remarks to the Author): The article « A chemical link between methylamine and methylene imine: identification of aminomethyl radical and implications for interstellar glycine formation" proposed by Prasad Ramesh Joshi1 and Yuan-Pern Lee deals with the identification of radical species involved in the formation of one of the most important prebiotic molecule (glycine). The authors used solid p-H2 matrix IR spectroscopy to perform H + CH3NH2 reactions and identify the formation of •CH2NH2 and CH2NH through their vibrational spectra. The formation of •CH2NH2 is observed in activation-less environment, which supports its formation in dark interstellar clouds. A multiple step cycle mechanism is given that explains the kinetics and stability of the species. These findings are supporting the hypothesis by Ioppolo et al for the formation of glycine with irradiation involved. The topic is of great interest and I find the methodology relevant. The article is globally well presented and it might be of interest for Commschem. However I have a few comments that the authors should take into account. The article relies on the comparison between measured IR spectra and DFT calculations. In order to identify which species is present, the authors compute the spectrum of various possible molecules. The comparison itself seems to rely on "visual inspection": lines are close in frequency and intensities etc... I found this methodology rather unreliable, as the "agreement" is subjective. I would like to encourage the authors to define more "objective" criteria (percentage of agreement or a match score) that would define more precisely allow to compare the agreement between the experimental spectrum and theoretical spectra. The way the data are represented makes the comparison difficult to read (see for instance Figure1). For instance the authors claim: « The lines in group A at 3500.5, 3403.6, 3143.3, 3042.6, 1609.9, 1213.6, and 685.5 cm-1 106 107 agree well, in terms of wavenumbers and relative intensities » "Agree well" does not have any significance, the relative intensity and frequency match is hard to quantify by a simple visual inspection of this figure. I therefore recommend the author to use a different representation where the spectra are zoomed in and where the comparison exp/theo is direct (superimpose experimental spectra and calculated sticks). Why calculated lines at 1300 and 1450 cm-1 are not present in experimental data? overall the work is interesting but it requires a better presentation of the results and more quantitative criteria for comparison.
Response/Revision: Because theoretical predictions typically have errors about ±20 cm −1 and intensities ±50 % or even larger, it is not easy to have superimposed spectra with zoomed scale. Because of this limitation, usually a pattern recognition (wavenumbers and relative intensities) was employed to aid in making the assignments. We agree with the reviewer that the description was subjective, so in Table 1 we listed the observed and predicted wavenumbers and intensities for comparison. Now we have included information on deviations between experiment and predicted vibrational wavenumbers for all observed wavenumbers in Table 1 and included the average absolute deviations in the text.
As far as predicted lines at 1300 and 1450 cm −1 are concerned, the intensity of both lines is 2 km mol -1 , which is ~1 % of the most intense line, and thus these lines are difficult to observe experimentally.
We added two sentences on page 6, which reads: "The average absolute deviation between experiment and prediction is 18.7 ± 15.2 cm -1 (1.06 ± 0.8 %) for •CH 2 NH 2 . The large deviation for ν 6 (CH 2 wag) is typical for this mode because of the inadequacy in describing the double-well potential experienced by N atom, similar to NH 3 . All lines of •CH 2 NH 2 located in our detection spectral range with predicted IR intensity greater than 6 km mol -1 were observed; predicted lines near 1448 and 1292 cm -1 have intensity ~2 km mol -1 too small to be observed." We also added one column in Table 1 to show the deviations between experiments and predictions.

Reviewer #2 (Remarks to the Author):
The work presented in the manuscript is the follow-up of previous similar investigations on other systems. As in those cases, I have found the present study very interesting and extremely relevant to the cold chemistry of interstellar objects. In particular, the authors have investigated in solid p-H2 the effects caused by exposing methylamine to hydrogen atoms. An interesting isotopic effect has also been noted when using partially deuterated methylamine. The employed experimental technique is stateof-the-art and the results are interpreted in the light of dedicated electronic structure calculations. I have only a few comments that the authors might address. 1) There has been a recent theoretical investigation on the reaction CH3 + NH2 assisted by a cluster of 18-or 33-water molecules to simulate the effect of amorphous ice. I think it would be nice to compare part of the present theoretical results with those reported by Enrique-Romero et al. as the real system we aim to understand is not only cold but features the presence of water molecules. The paper is in press in ApJSS, but can be found also here https://arxiv.org/pdf/2201.10864.pdf Response/Revision: We thank the reviewer for providing this information. Following this idea to understand the influence of surrounding H 2 on •CH 2 NH 2 in solid p-H 2 , we performed quantum-chemical calculations with eighteen H 2 molecules surrounding •CH 2 NH 2 ; the H 2 molecules were placed either in a hexagonal-closed pack (hcp) lattice or randomly (i.e. freely optimized). As expected, the simulated IR stick spectra in both solid-phase cases are similar to the gaseous phase, within uncertainties of calculations. We added the following text on page 5. "To understand the perturbations of H 2 on the IR spectrum of •CH 2 NH 2 , we performed calculations also on •CH 2 NH 2 surrounded by eighteen H 2 molecules, either in a hexagonal-closed pack (hcp) lattice or randomly (free optimization). The resultant vibrational wavenumbers and IR intensities are compared in Supplementary Table 1 and the simulated IR stick spectra of •CH 2 NH 2 are presented in Supplementary Fig. 2 to compare with calculations for gaseous •CH 2 NH 2 and experiments. The perturbation by H 2 is small (with average absolute deviations 8.8 ± 6.0 and 14.8 ± 8.5 cm −1 from the gaseous phase; listed errors represent one standard deviation in fitting) and within calculation errors. This is in line with the fact that observed IR spectra of matrix-isolated species typically showed <1 % matrix shifts so that comparison of observed vibrational wavenumbers with predictions of gaseous species was generally performed." The suggested reference simulated the stability and reactivity of species on ice surface, which is different from our calculations, which simulated the spectrum of the radicals. To include this reference, we added a few sentences on page 11 after Fig. 4 as: "We understand that our experimental conditions do not mimic the ISM conditions closely, so our results cannot be applied directly to the reactions in the ISM. For example, in the case of water ice environments, the interaction between water and the guest species might be stronger so that the stability and reactivity of radicals are different from the gaseous phase, as demonstrated by the simulations of radical-radical reactions on icy surfaces by Enrique-Romero et al. 33 ".
2) I think that the astrophysical implications of this study have been stretched a little bit too far. I refer to the relation between methanimine and methylamine. First of all, methanimine is almost ubiquitous and has been observed in many different kinds of interstellar objects. Its first detection dates back to 1973. On the contrary, methylamine has been detected in few objects. In other words, there are many environments where methanimine is observed and methylamine is not. This is inconsistent with the motherdaughter relation that is suggested here. Non only that: there are numerous hints that methanimine can be considered a signpost for chemistry occurring in the gas phase, while there no gaseous route of methylamine formation, see, for instance, Suzuki et al.

Response/Revision:
We thank the reviewer for point this out. We have removed Figure 5 and the associated text discussing interstellar [CH 2 NH]/[CH 3 NH 2 ] ratio from the manuscript.
3) I think that there is an important previous study that should be cited here, that is the study on hydrogenation of solid hydrogen cyanide HCN and methanimine CH2NH at low temperature by Theule and coworkers (A&A 534, A64, 2011; https://doi.org/10.1051/0004-6361/201117494). Interestingly, in that study, the authors reported they have been unable to observe methanimine formation in the hydrogenation of HCN because it easily converts into methylamine. Apparently, they were unable to see the same effect caused by H-abstraction that has been seen here. A comment is in order.
1. The experimental conditions may not be suitable for comparing with the interstellar conditions. I have found that the experiments were designed very well. It is a good step to utilize partially deuterated CH3NH2 towards secure identification of the CH2NH2 radical. The experiments were conducted in the solid p-H2 matrix at 3.2 K. This environment is very different from that in the actual physical conditions in the interstellar dust. It is widely understood in astrophysics that the surface of interstellar dust particles, especially deep inside dense molecular clouds, such as hot cores, are covered with amorphous water ice (ice mantle). Further it is also thought in astrophysics that the dust temperature is determined from the balance of their radiative heating (by absorbing far-IR interstellar radiation penetrating into the cores) and cooling (via the continuum emission). Theoretical consideration with these assumptions concluded that the dust temperature will not be lower than 8 K. I wonder if the beautiful experimental results by the authors could be applied to actual interstellar dust particles. The temperature, 3.2 K, is much lower than the theoretical lower limit of dust temperature. Further, interaction between solid H2 and CH2NH2 would be different from those between water ice and CH2NH2. If the authors claim that it is possible, I would ask the authors to demonstrate the possibility with convincing rationales.

Response/Revision:
We thank the reviewer for pointing this out. Firstly, we would like to emphasize that the main goal of the present study is to demonstrate the formation of •CH 2 NH 2 radical which was considered to be an intermediate for the reaction CH 3 NH 2 + HOCO leading to the formation of interstellar glycine. We do not claim that our experimental conditions mimic the ISM conditions, rather, we would like to explore key reactions and radical intermediates that are difficult to study under interstellar relevant conditions because of the complicated interactions with species in ice. The temperature used for our experiments is lower than the theoretically assumed dust temperature of 8 K because solid p-H 2 matrix evaporated at 5 K. However, typically, if the reaction occurs at 3.2 K, it is expected to occur also at 8 K. Secondly, we have presented that H abstraction of CH 3 NH 2 is significant for the formation of •CH 2 NH 2 . Earlier laboratory (e.g. Ioppolo, S. et al. Nat. Astron. 5, 197−205 (2021) and theoretical (e.g. Garrod, R. T. Astrophys. J. 765, 1−29 (2013)) models assumed the formation of •CH 2 NH 2 via H abstraction of CH 3 NH 2 by H atoms or •OH. Three-phase model demonstrated by Garrod included Habstraction reactions taking place between complex molecules (precursors) and the most significant grain-surface radicals/atoms such as H, •OH, and •NH 2 . Additionally, the mobility of chemical reactants in the bulk ice is assumed through a swapping mechanism that was supported by laboratory work (e.g. Oberg et al. Astron. Astrophys. 505, 183-194 (2009 (2013)); this mechanism likely brings H atom and CH 3 NH 2 in close proximity to react.
To clarify this, we have revised the paragraph after Fig. 4 (pages 11 and 12) as: "We understand that our experimental conditions do not mimic the ISM conditions closely, so our results cannot be applied directly to the reactions in the ISM. For example, in the case of water ice environments, the interaction between water and the guest species might be stronger so that the stability and reactivity of radicals are different from the gaseous phase, as demonstrated by the simulations of radical-radical reactions on icy surfaces by Enrique-Romero et al. 33 Nevertheless, our results clearly indicate that reaction of H with methylamine CH 3 NH 2 produces •CH 2 NH 2 , an important radical precursor for the formation of glycine, directly supporting the mechanism, reaction (1), proposed by Ioppolo et al. 19 for the formation of glycine under conditions similar to dark interstellar clouds with no need for UV irradiation or cosmic rays. The mobility of chemical reactants in the bulk ice is assumed through a swapping mechanism that was supported by laboratory work 34,35 and theoretical investigations; 5 this mechanism likely brings H atom and CH 3 NH 2 in close proximity to react." 2. Applicability of the quantum chemical calculations to species in solid phase I fully understand that the quantum chemical calculations are powerful tools for understanding energies of reactants, products as well as transition states. Quantum chemical calculations are often used to predict IR spectra in the "gas phase". The experiments are conducted in the solid H 2 matrix, not in the gas phase. Since I was not able to find a text that the quantum chemical calculations were made for the sold phase, I wonder if the quantum chemical calculations shown in the paper considred that the reactants, products and transition state species are in the solid phase. It seems to me that the calculations correspond to the gas phase. If the quantum chemical calculations corresponded to the gas phase, the authors should not compare the calculated IR spectra and the experimental ones. If the authors made the calculations for the solid phase, I would ask the authors to add new text showing that the quantum chemical calculations are for the solid phase.

Response/Revision:
The nice thing about matrix isolation for IR spectroscopy is that, even it is in the solid phase, the perturbation by the matrix host is small because of negligible interactions between guest and host molecules. The vibrational wavenumbers in solid Ar differed by those in the gaseous phase by less than 1 % (Jacox M.E. Chem.