Alcohols and aldehydes represent two key classes of interstellar complex organic molecules (COMs). This work seeks to better understand their possible chemical connections, with a focus on such molecules in the sources of the star-forming region Sgr B2 (N).
Methods
The gas-phase reaction between ethanol (CH3CH2OH) and the halogens fluorine and chlorine was investigated via DFT calculations, with the goal of determining whether astrochemically viable chemical pathways leading to acetaldehyde (CH3CHO) exist. The studied reactions were then included in an astrochemical model of Sgr B2 (N) to determine their significance under real interstellar conditions.
Results
Our DFT calculations revealed that both chlorine and fluorine can react barrierlessly with ethanol to abstract a hydrogen atom. We further found that, following this initial step, the resulting ethanol radicals can undergo further reactions with atomic hydrogen, with some routes leading to acetaldehyde. Incorporation of these novel reactions in astrochemical models of hot cores suggest that they are indeed efficient under those conditions, and can lead to modest increases in the abundance of CH3CHO during model times where gas-phase ethanol is abundant. Of the ethanol radicals included in our chemical network, we found CH3CHOH to have the highest abundances in our simulations comparable to that of ethanol at some model times.
Discussion
Overall, this work reveals a novel gas-phase “top-down” link from alcohols to aldehydes that compliments the better studied “bottom-up” routes involving grain-surface H-addition reactions yielding alcohols from aldehydes. Moreover, results from our astrochemical models suggest that the ethanol radical CH3CHOH may be detectable in the interstellar medium.
alcoholsaldehydesastrochemical modelingastrochemistryDFThalogensSgr B2European Science Foundation10.13039/501100000782Fundación General CSIC10.13039/501100006003The author(s) declared that financial support was received for this work and/or its publication. The National Radio Astronomy Observatory and Green Bank Observatory are facilities of the U.S. National Science Foundation operated under cooperative agreement by Associated Universities, Inc. CS, DW, and ES gratefully acknowledge support through the Virginia Military Institute’s Summer Undergraduate Research Institute (SURI) program. GM acknowledges the support of the grant RYC2022-035442-I funded by MCIU/AEI/10.130 39/501100011033 and ESF+. GM also received support from project 20245AT016 (Proyectos Intramurales CSIC). We acknowledge the computational resources provided by the DRAGO computer cluster managed by SGAI-CSIC, and the Galician Supercomputing Center (CESGA). The supercomputer FinisTerrae III and its permanent data storage system have been funded by the Spanish Ministry of Science and Innovation, the Galician Government and the European Regional Development Fund (ERDF).section-at-acceptanceAstrochemistryIntroduction
Hot cores, especially those found towards the Galactic Center in the Sagittarius (hereafter Sgr) B2 (N) complex, are known to be some of the most chemically rich interstellar sources (Belloche et al., 2013; Bonfand et al., 2019; Neill et al., 2014; Loomis et al., 2013; Zaleski et al., 2013). The reason for this complexity is manyfold, though significant factors include the fact that the higher temperatures obtained during core warmup liberate existing complex organic molecules (COMs) already present in dust-grain ice mantles out to the grain surfaces and into the gas-phase (Yang et al., 2022; McClure et al., 2023). This enables a comparatively brief but active period of surface chemistry due to the enhanced mobility of surface species heavier than hydrogen (Garrod et al., 2008; Ioppolo et al., 2020; Qasim et al., 2020).
Underscoring this chemical complexity is the fact that molecules bearing most of the functional groups in organic chemistry have now been detected (McGuire, 2018), with many more being detected annually. Of these, two of the most important and common are alcohols and aldehydes, having R-OH and R-CHO groups, respectively, where R could be any other part of the molecule. One of the essential tasks in astrochemistry is investigating possible chemical connections between detected molecules. Alcohols and aldehydes hold a crucial cornerstone in the evolution of chemical complexity in the ISM, bearing two of the most chemically ubiquitous functional groups. To this end, a number of studies have focused on the chemical connections between alcohols and aldehydes, for example, the theoretical studies of Woon (2002), Rimola et al. (2014), Das et al. (2008), Song and Kästner (2017) and Mondal et al. (2021), as well as the experimental work by Qasim et al. (2019) and Hiraoka et al. (1998). Most of these studies envisioned the formation of the alcohol from the aldehyde via successive reactions with atomic hydrogen on grain surfaces, which represents a “bottom-up” chemical link between the two. This hypothesis is supported by the easy hydrogenation of formaldehyde H2CO to methanol CH3OH (Watanabe and Kouchi, 2002).
While the “bottom-up” route is well supported for methanol and formaldehyde, the picture is more complicated for acetaldehyde and ethanol. Recent computational works suggest that grain-surface formation routes of acetaldehyde are likely to be unfavorable (Perrero et al., 2023). This, combined with the presence of acetaldehyde in cold prestellar cores (Scibelli and Shirley, 2020), points to the main mechanisms of acetaldehyde formation potentially being in the gas-phase (Vazart et al., 2020). Further complicating the bottom-up scenario, results from recent quantum chemical calculations by Molpeceres et al. (2025) suggest that acetaldehyde, at least, is resistant to further hydrogen addition, complicating the bottom-up scenario.
Another possible chemical link between alcohols and aldehydes is a “top-down” route where alcohols are converted to aldehydes through successive hydrogen abstractions. To further investigate the role of gas-phase formation, it is this chemical route that this work seeks to address through a combination of ab initio quantum chemical calculations and astrochemical modeling of Sgr B2(N). However, rather than assuming a series of reactions with, initially, the alcohol with H, we instead assume the initiating co-reactant are the halogen atoms Cl and F in the gas-phase. This choice was motivated in part by the observation by Balucani et al. (2015) of the increased reactivity of these species compared with H. Both HCl and HF have been detected towards Sgr B2 (Zmuidzinas et al., 1995; Neufeld et al., 1997), with observed fractional abundances that exist in the ranges of ∼10−10–10−9 relative to molecular hydrogen. The contributions of chlorine and fluorine chemistry in the context of astrochemical models have previously been studied in various interstellar environments in Acharyya and Herbst (2017), wherein they include reactions of methanol with atomic F and Cl, which can result in either the CH3O or CH2OH radical. In the case of the latter, the CH2OH can react further with a chlorine atom to produce formaldehyde in the “top-down” fashion, but reactions including ethanol were not investigated (Acharyya and Herbst, 2017).
The rest of this paper is structured as follows: in Section 2 we describe our approaches for both the quantum calculations and the astrochemical modeling, in Section 3 we present the results of this work, and finally, we summarize our findings in Section 4.
MethodsQuantum chemical calculations
We determined the outcome of the H-abstraction reactions with fluorine, shown in Equation 1, and chlorine, shown in Equation 2:CH3CH2OH+FCH3CH2OH+Clusing quantum chemical calculations. The first step is to determine the stationary points in each of the potential energy surface (PES) (Section 2.1.1) for the different reaction channels, namely:CH3CH2OH+F−>CH2CH2OH+HFCH3CH2OH+F−>CH3CH2O+HFCH3CH2OH+F−>CH3CHOH+HFCH3CH2OH+Cl−>CH2CH2OH+HClCH3CH2OH+Cl−>CH3CH2O+HClCH3CH2OH+Cl−>CH3CHOH+HCl
Later, we derive the corresponding bimolecular rate constants of each of the reaction channels (Section 2.1.2), that we later introduce in the kinetic models of Sgr B2 (N).
Characterization of stationary points
All geometries were optimized using density functional theory (DFT) with the double-hybrid functional rev-DSD-PBEP86(D4) (Kozuch and Martin, 2011; Santra et al., 2019; Caldeweyher et al., 2019), in combination with the cc-pCVTZ basis set (Hill et al., 2010), which includes functions designed to account for core–valence correlation. Zero-point vibrational energy (ZPVE) corrections were computed within the harmonic approximation by numerical differentiation of the gradient. To obtain more accurate energetics, we refined the energy of each stationary point at the CCSD(T)/cc-pCVTZ level (Raghavachari et al., 1989; Bartlett et al., 1990; Hill et al., 2010) using an unrestricted Hartree-Fock (UHF) reference wavefunction (Neese, 2000). All calculations, including geometry optimizations, ZPVE corrections, and single-point energy refinements, were performed correlating all electrons, i.e., without applying the frozen-core approximation. All our electronic structure calculations employ the Orca package (v.6.0.0) (Neese et al., 2020; Neese, 2022).
To investigate the reactivity of each distinct hydrogen atom in ethanol, we performed exploratory PES scans along well-defined reaction coordinates, followed by full optimization of the resulting stationary points. All reactions were modeled (Section 2.1.2) within a general mechanistic framework involving two van der Waals complexes: one preceding the hydrogen abstraction (pre-reactive complex, PRC) and one following it (post-reactive complex). This scheme was applied to all possible hydrogen abstraction pathways, regardless of whether the corresponding transition states (TS) were submerged. The only exception was the set of reactions CH3CH2OH+F/Cl→CH3CHOH+HF/HCl, where PES scans were inconclusive, suggesting a barrierless abstraction mechanism with no PRC on the electronic energy surface at the rev-DSD-PBEP86(D4)/cc-pCVTZ level. To assess whether these reactions truly proceed without an activation barrier, as proposed in previous studies (Taketani et al., 2005), we carried out nudged elastic band (NEB) calculations (Henkelman et al., 2000) to confirm a strictly downhill energy profile. The NEB paths were discretized into 24 images including endpoints and computed using the Knarr module of the Ash library (Asgeirsson et al., 2021; Bjornsson, 2022).
Kinetic calculations
We compute phenomenological rate constants for all bimolecular channels considered in this study using an ab initio transition state theory master equation (AITSTME) framework. Barrierless processes, such as capture into a pre-reactive complex (PRC) or into a product-side van der Waals (vdW) complex, as discussed above, and the escape from product vdW complexes are modeled using phase space theory (Pechukas and Light, 1965; Chesnavich, 1986). The rigid scans for these barrierless association and dissociation channels are performed at the spin component scaled MP2 (SCS-MP2) level (Grimme et al., 2012) with the aug-cc-pVTZ basis set (Kendall et al., 1992), starting from the corresponding stationary point on the potential energy surface (on the pre-optimized geometries using the double hybrid functional). The scans extend from 4 Å to 15 Å to capture the long range asymptotic behavior of the interactions. This last set of calculations were carried out within the frozen core approximation. The resulting scan energies are fitted to a potential of the form V(r)=−C/rn, with n=6 for entrance channels and n=5 for exit channels. The choice of using a fit with n=5 stems from the change in the nature of the interaction once HF or HCl are formed. While the entrance channel corresponds to an atom–molecule interaction well described by an r−6 dependence, the exit channels are dominated by dipole–dipole interactions, better represented by an r−5 behavior. In fact, the large dipole moment of HF (or HCl) can make some exit channels even better reproduced by an n=4 dependence. Nevertheless, we retain n=5 for consistency, since the kinetics of the reaction of interest are mainly governed by the capture event, for which the r−6 fit provides a good description in all cases (Figure 1). In cases where a barrier exists between the PRC and the product vdW complex, the interwell transition is modeled using RRKM theory, including tunneling corrections based on asymmetric Eckart transmission coefficients. To avoid numerical instabilities arising from vibrational entropy contributions, all vibrational frequencies of weakly bound complexes below 100 cm−1 are raised to that threshold. The elementary rate constants and the resulting phenomenological rate constants for the full process are calculated using the Mess code (Georgievskii et al., 2013), across a temperature range of 30–500 K and at a residual pressure of 1×10−7 atm, low enough to prevent collisional relaxation of the PRC and product vdW complexes. Finally, in a first approximation we consider that all reactions take place in the lowest conformer of ethanol, that is, we do not consider a multistate (or multiconformer) treatment of the reaction (Zheng and Truhlar, 2012).
Capture rigid scans leading to PRCs or VDW complexes, see text. In the legend we show the value of the capture coefficient C6. The scans are not ZPE corrected.
Six graphs display energy change (\( \Delta E \)) against scan distance for fluorine (F) and chlorine (Cl) at three different entrances. In the top row, graphs for fluorine show decreasing \( \Delta E \) values with varying \( C_6 \) coefficients: 41.45, 149.09, and 45.22 a.u. In the bottom row, graphs for chlorine also show decreasing \( \Delta E \) with \( C_6 \) coefficients: 215.81, 390.83, and 119.25 a.u. Each graph demonstrates similar trends of energy reduction over distance.
The phenomenological rate constants are later fitted to a typical 3-parameter Arrhenius-Kooij formula of the type:k=αT300Kβexp−γTwhere α is the pre-exponential factor for the rate coefficient, β gives the temperature dependence, and γ is the energy barrier. The resulting fits to this equation for reactions Equations 3–8 are described below.
Avenues for the improvement of the quantum chemical calculations
The model chemistry employed in our electronic structure calculations is sufficiently accurate for the purposes of this study. Nevertheless, several refinements could be implemented to achieve a fully quantitative description of the reactions, particularly at higher temperatures where finite-temperature effects may influence the kinetics. Possible improvements include going beyond the harmonic approximation to account for anharmonic vibrational frequencies, explicitly considering ethanol conformational effects, or introducing methodological refinements to the model chemistry, such as enlarging the basis set. The kinetic treatment could also be improved by moving beyond a one-dimensional zero-curvature tunneling correction to more accurately capture corner-cutting effects (Nandi et al., 2022). Finally, the most immediate avenue for improvement would be a more accurate description of the capture event in the barrierless channels, since capture theory is known to overestimate the capture rate constant (Marchione et al., 2022; Tsikritea et al., 2022).
Astrochemical modeling
In order to estimate the efficiency of the studied reactions under real astrophysical conditions, astrochemical models were run. For this, the nautilus v1.1 code was used (Ruaud et al., 2016). The physical conditions used in the model, shown in Table 1, were chosen to replicate those of Sgr B2 (N). Briefly, a three-phase (gas, grain surface, and grain bulk) three stage (collapse, warm up, and constant physical conditions) model was employed to simulate the collapse of a prestellar cloud and the formation of a hot core (Garrod, 2013). The initial physical conditions at the beginning of collapse include an AV of 0.3, density of 103cm−3. The collapse stage has a duration of 8×105 yr, at the end of which the model reaches a density of 108cm−1, as shown in Figure 2. During the collapse, the gas temperature was held at a constant value of 10 K and the grain temperature, which starts at 15 K, becomes coupled to the gas temperature of 10 K by the end of this first stage, which can be seen in Figure 3. A cosmic ray ionization rate of ζ=1.3×10−16s−1 was used for all stages to account for the enhanced cosmic ray flux in the Galactic Center, which is likely at least one order of magnitude greater than the standard value of ∼10−17s−1 (Goto, 2013).
Physical parameters used in all astrochemical models. Note: the parameter α is the chemical desorption efficiency (Garrod et al., 2007).
Parameter
Value
ζ
1.3×10−16s−1
UV flux
1 Draine
α
1.0×10−2
Grain radius
1.0×10−5 cm
Diffusion barrier width
1.0×10−8 cm
Surface site density
1.5×1015 sites/cm2
Ebsurface/ED
0.4
Density evolution profile.
Graph showing particle density over time on a logarithmic scale. Density (particles per cubic centimeter) is on the vertical axis from 10^3 to 10^9, and time (years) is on the horizontal axis from 10^4 to 10^7. The graph shows a gradual increase, then a rapid rise near 10^6 years.
Gas and grain temperature evolution profiles.
Graph showing temperature in Kelvin versus time in years on a logarithmic scale. A blue solid line represents gas and a red dashed line represents grain. Both lines are constant until they sharply rise near 10^6 years, stabilizing at 400 Kelvin.
Following the collapse stage, a warm up occurs, during which the density and AV remain constant while, as shown in Figure 3, the gas and grain temperatures rise to ∼400 K over a period of ∼106 yr. Finally, after the warm up stage, the model continues with static physical conditions until a total model time of 107 yr is obtained.
Initial elemental abundances were taken from Laas and Caselli (2019), and correspond to known cosmic standard elemental abundances. The full table of values is given in Laas and Caselli (2019), but we note in particular an initial chlorine abundance of 2.88×10−7 relative to hydrogen (Esteban et al., 2004) and fluorine abundance of 3.63×10−8 relative to hydrogen (Asplund et al., 2009). These values differ slightly from those used in Acharyya and Herbst (2017), which used initial abundances relative to hydrogen of 1×10−7 and 1.8×10−8 of chlorine and fluorine, respectively. To determine if the differences in initial abundances will affect the overall chemistry, additional models were run with starting with initial abundances of 10−7 and 1.8×10−8 for chlorine and fluorine, respectively.
The base chemical network used here is that of Byrne et al. (2024), developed for work by the GOTHAM (GBT Observations of TMC-1 Hunting Aromatic Molecules) project. To this was added reactions (3)–(8), in addition to a number of destruction paths for the ethanol radicals with H, OH, and NH2. The destruction reactions with H added to our network are given in Equations 10–15, namely,CH3CHOH+H−>CH3CHO+H2CH3CH2O+H−>CH3CHO+H2CH3CH2OH+H−>CH3CHO+H2CH3CHOH+H−>CH4+H2OCH3CH2O+H−>CH4+H2COCH3CH2OH+H−>CH4+H2CO.Equations 16–21 show those with OH,CH3CHOH+OH−>CH3CHO+H2OCH3CH2O+OH−>CH3CHO+H2OCH3CH2OH+OH−>CH3CHO+H2OCH3CHOH+OH−>CH3OH+H2COCH3CH2O+OH–>CH3OH+H2COCH3CH2OH+OH−>CH3OH+H2CO.Finally, Equations 22–27 give the destruction reactions with NH2, which are as follows:CH3CHOH+NH2−>CH3CHO+NH3CH3CH2O+NH2−>CH3CHO+NH3CH3CH2OH+NH2−>CH3CHO+NH3CH3CHOH+NH2−>CH3NH2+H2COCH3CH2O+NH2−>CH3NH2+H2COCH3CH2OH+NH2−>CH3NH2+H2CO.
All of the above destruction reactions were assumed to occur barrierlessly at the collisional rate of 10−10cm3s−1, split between the two product channels, corresponding to H abstraction (in the case of reactions leading to acetaldehyde) vs. attack of the alpha carbon. Our choice of rate coefficient is a typical approximate value for barrierless bimolecular gas-phase reactions without significant long-range forces, though follow up studies of these reactions are needed to confirm this choice, a task beyond the scope of this work. If these destruction reactions prove to be less efficient, the abundance of the intermediate radicals would be higher, which would have implications related to their detectability as described below. If the reverse is true, then the chemical connections between alcohols and aldehydes is further strengthened. Either outcome would be of astrochemical interest.
Results and DiscussionQuantum chemical resultsCH<sub>3</sub>CH<sub>2</sub>OH + F
The PES profiles for H abstraction by fluorine at the three inequivalent positions are shown in Figure 4. A quick inspection of the graph reveals that all hydrogen abstraction reactions with fluorine (Reactions 3, 4, and 5) are barrierless with respect to the reaction asymptote CH3CH2OH+F. It is important to note, however, that each reaction is barrierless for a different reason. In the case of Reaction 3, we located a TS on the electronic energy surface (that is, without ZPVE correction) at less than 1.0 kcal mol−1. This shallow barrier disappears upon application of ZPVE. We therefore consider this reaction to be effectively barrierless, proceeding directly toward the van der Waals product complex (VDW1 in Figure 4). Nevertheless, a small barrier might still be present if more accurate electronic structure methods were used to determine the stationary point energies. For Reaction 4, the barrierless character originates from the submersion of the TS below the reactant asymptote. This is due to the stabilization of the PRC, located at approximately −3.6 kcal mol−1, which consequently lowers the energy of the TS as well. Finally, Reaction 5 is the only case that is entirely barrierless, displaying a strictly downhill energy profile on the electronic PES. This behavior is confirmed by the NEB calculations described in Section 2.1.1.
Potential energy profiles for the three abstraction channels in the CH3CH2OH+F reaction. All energies are ZPVE corrected.
Reaction energy diagram showing the potential energy changes of a reaction involving ethanol and fluorine. Energy levels in kilocalories per mole are marked for intermediate and final states, labeled from 0.0 at the starting materials to -43.0 for the final products. Transition states and intermediates such as PRC2 and TS2 are labeled with their respective energy changes. Molecular structures are depicted at each stage.
The PES profiles shown in Figure 4 provide a useful framework to interpret the kinetics of these reactions. The corresponding rate constants are presented in Figure 5 (top panel). Both Reactions 3 and 5 exhibit clearly barrierless behavior, with only a minor dependence on temperature. We attribute this slight temperature dependence to the small enhancement of the final step, VDW → Product, at higher temperatures. In contrast, Reaction 4 shows anti-Arrhenius behavior, which is typically observed in barrierless reactions where the redissociation from the PRC back to reactants becomes increasingly significant at higher temperatures.
Reaction rate constants for the H-abstraction reactions of CH3CH2OH with F (Top panel) and Chlorine (bottom panel).
Two graphs show the rate constant \( k \) (\( \times 10^{-10} \, \text{cm}^3 \, \text{s}^{-1} \)) versus temperature \( T \) (K). The top graph compares reactions \( \text{CH}_2\text{CH}_2\text{OH} + \text{HF} \) (blue), \( \text{CH}_3\text{CH}_2\text{O} + \text{HF} \) (green), and \( \text{CH}_3\text{CHOH} + \text{HF} \) (black). The bottom graph compares the same species with HCl. Insets detail low values, highlighting differences in rates across temperature ranges.CH<sub>3</sub>CH<sub>2</sub>OH + Cl
The CH3CH2OH+Cl reaction is slower than its fluorine analogue. This fact follows directly from the PES cuts in Figure 6. Among the three abstraction pathways, only Reaction 8 is truly barrierless, as confirmed by an NEB calculation, exactly as for the corresponding F reaction. The remaining pathways, Reactions 6 and 7, display well defined TSs above the reactant asymptote, which strongly diminish their rates. The absolute PES profile agrees closely with earlier work on this system (Taketani et al., 2005). It reproduces both the small exothermicity of Reaction 6 and the endothermicity of Reaction 7, the latter making that channel relevant only at high temperature. PRC1 (Reaction 6) and PRC2 (Reaction 7) show inverted stability with respect to their products CH2CH2OH and CH3CH2O. PRC2 is stabilised by H bonding between the hydroxyl group of CH3CH2OH and the Cl atom, making it the most stable PRC in this study and even more stable than PRC2 of Reaction 4. Such deep stabilisation raises the overall barrier for Reaction 7, leading to very slow kinetics. By contrast, Reaction 6 has a much lower barrier; although slower than the barrierless channel (Reaction 8), its rate constants exceed those of Reaction 7 by several orders of magnitude at low temperature. The stabilization of PRC2, in principle, leads to competition between back-dissociation to the reactants and evolution through a roaming mechanism toward VDW3, ultimately forming CH3CHOH+HCl. Our simplified kinetic scheme does not take roaming into account, which may result in an underestimation of the overall rate constant. However, this approximation partially compensates for the possible presence of sub-kcal mol−1 activation barriers for reaction 8, which would otherwise artificially enhance the calculated rate at low temperatures.
Potential energy profiles for the three abstraction channels in the CH3CH2OH+Cl reaction. All energies are ZPVE corrected.
Reaction energy profile diagram showing the potential energy changes (ΔU in kcal per mol) for the reaction of ethanol (CH₃CH₂OH) with chlorine. The diagram includes transition states (TS1 and TS2) and intermediate complexes (PRC1, PRC2, VDW1, VDW2, and VDW3) at specific energy levels. Molecular structures are depicted at each stage, illustrating the changes in the reaction pathway from reactants to products, such as the formation of CH₃CH₂O and HCl.
Looking at the rate constants for the CH3CH2OH+Cl reaction, shown in Figure 5 (bottom panel), we observe that, in contrast to the case with fluorine, the reactivity is entirely dominated by the channel corresponding to Reaction 8. This is unsurprising given its barrierless nature. However, although irrelevant for the overall kinetics, we also find that the gap in rate constants between Reactions 6 and 7 decreases with increasing temperature, and the two channels converge at high temperature. This unexpected trend can be rationalised by a combination of two effects. First, the greater thermal energy at high temperature makes the endothermic CH3CH2O+HCl channel (Reaction 7) accessible, while it is effectively closed at low temperature. Second, the higher imaginary frequency of the TS in Reaction 7 (2013i) compared to that of Reaction 6 (1042i) enhances the tunneling efficiency, as this frequency can be taken as an estimate of the crossover temperature for tunneling (Gillan, 1987), which is significantly higher for Reaction 7. Nonetheless, as noted above, this behaviour does not affect the global kinetics, since both Reactions 6 and 7 remain minor channels compared to the dominant Reaction 8.
Comparison between reactions with fluorine and chlorine
The comparison between the two reactions studied in this work reveals fundamental differences between H abstraction by fluorine and by chlorine, which are worth highlighting to guide further chemical interpretation. As expected from basic principles of inorganic chemistry, fluorine is overall more reactive than chlorine. However, in this case we can quantify the effect. Considering the sum of all reaction channels at low temperatures 8.85×10−10cm−3s−1, we find that the total rate constant for the fluorine system is approximately a factor 2–3 (2.8) larger than that for chlorine. The anti-Arrhenius behavior of Reaction 4, together with a slightly faster capture efficiency in Reaction 8 compared to Reaction 5 prevent the system from reaching the theoretical factor of three that would be expected based solely on the number of barrierless channels. It is also worth noting that while all reactions involving H abstraction by fluorine are strongly exothermic, this is not the case for chlorine, which shows much lower exothermicities and even endothermic channels at low temperatures, as in the case of Reaction 7. This difference suggests caution when extrapolating the behavior of chlorine to other systems or when proposing new H-abstraction reactions involving chlorine.
Astrochemical modeling
In Table 2 we show the values of the Arrhenius-Kooij parameters used to introduce the halogen-mediated H-abstraction reactions considered in Section 3.1. Calculated abundances for acetaldehyde (CH3CHO) for our three-stage model of Sgr B2(N) are shown in Figure 7. Shown in yellow are results from our “control” model, which uses the unmodified network of Byrne et al. (2024) without any of the new reactions mentioned in this work.1 The blue curve in the figure gives the model results using our newly expanded chemical network. A comparison of the two shows that they are largely the same, save for a noticeable increase in acetaldehyde abundance around a model time of 106 yr.
Fitted parameters of Equation 9 for the reactions considered in our quantum chemical calculations. The fit is performed for the rate constants derived between 30 and 500 K.
Reaction
Label
α (cm−3s−1)
β
γ (K)
CH3CH2OH+F→CH2CH2OH+HF
3
4.0(−10)
1.7(−1)
9.8(−3)
CH3CH2OH+F→CH3CH2O+HF
4
6.1(−11)
−1.4(0)
4.5(1)
CH3CH2OH+F→CH3CHOH+HF
5
4.1(−10)
1.7(−1)
1.3(−3)
CH3CH2OH+Cl→CH2CH2OH+HCl
6
4.6(−13)
2.6(0)
−1.9(2)
CH3CH2OH+Cl→CH3CH2O+HCl
7
7.4(−12)
−4.4(−1)
5.2(2)
CH3CH2OH+Cl→CH3CHOH+HCl
8
4.6(−10)
1.7(−1)
5.3(−2)
Abundances of CH3CHO in our three-stage warmup model. Shown in black is the control model without the new reactions. Models including the new reactions are given in blue (high metal abundances) and red (low metal abundances).
Line graph showing the ratio \(n(x)/n(H_2)\) over time in years. The black line represents control, the blue line high metal, and the red line low metal. The curves show varying peaks and trends, with the low metal condition reaching the highest peak before decreasing. Numerical values are on logarithmic scales, with \(n(x)/n(H_2)\) on the left and \(N(x)\) in centimeters squared on the right.
To examine the effect of variations in initial Cl and F abundances, we ran a model using the lower initial abundances for chlorine and fluorine of, respectively, 10−7 and 1.8×10−8 relative to hydrogen. These values have been used in previous investigations of halogen astrochemistry by, e.g., Neufeld et al. (2005) and Acharyya and Herbst (2017). As can be seen in Figure 7, the reduced halogen abundances have only a minor influence on the time-dependent abundances of CH3CHO, with the peak abundance remaining unchanged. The increase in acetaldehyde abundance provides a clue as to the underlying chemistry. The likely answer is implied in Figure 8. Around 106 yr in the model, the gas (and dust) temperature is roughly 175 K, as shown in Figure 3. It is at this point that ethanol, which has hitherto been trapped mostly on grains, is efficiently liberated into the gas. This occurrence becomes the occasion for our new reactions to show their utility, when the desorbed ethanol begins reacting with F and Cl to produce the three ethanol radicals. Of the three, CH3CHOH is the most abundant, reaching approximately the abundance of its ethanol precursor from a few times 105 to 106 yr.
Abundances of CH3CH2OH, CH3CHOH, CH3CH2O, CH2CH2OH, and CH3CHO in our new model. Results shown here are for the high metal abundance model.
Graph showing the temporal evolution of methanol, acetaldehyde, ethanol, ethylene oxide, and acrylic acid abundances relative to hydrogen over three million years. Lines in black, blue, red, green, and pink represent these compounds, respectively. The left y-axis shows abundance ratios, while the right y-axis shows column density in square centimeters.
In previous work done investigating the role of fluorine and chlorine in the chemistry of various interstellar environments (including hot cores), Acharyya and Herbst (2017) included a system of F and Cl reactions in their chemical network, including those where atomic F and Cl react with methanol. From Table 2 and Figure 5, we find that the temperature-dependent rate constants for reactions 3−8 reach into the range of ∼10−13–5×10−10cm3s−1. Comparing these to the chlorine and fluorine reactions with methanol, as summarized in Acharyya and Herbst (2017), the ethanol pathways show a slight increase in reactivity. The most favourable pathway for the methanol-halogen reactions is found in the CH3OH+F channel resulting in CH2OH, with a rate constant of k=1.66×10−10cm3s−1 (Jodkowski et al., 1999).
Despite reactions 6 and 7 not having rate constants reaching as high of values and exhibiting clear barriers, the remaining four reactions all have peak values that are at least a factor of 2 larger than their methanol counterparts Acharyya and Herbst (2017). Furthermore, reactions 3 and 8 both share similar peak values, combined with their overall barrierless nature, this suggests that both chlorine and fluorine can efficiently contribute to the presence of the CH3CHOH radical.
Our calculated abundance presents the interesting possibility for the detection of CH3CHOH in interstellar environments, however, to the best of our knowledge, no laboratory spectrum for this species exists, though a recent study by Williams et al. (2021) did report calculations of relative energetics. Measured laboratory spectra generally provide a more reliable basis for astronomical searches than theoretically calculated spectra. We estimate that the strongest predicted transitions may appear in the radio to centimeter range (typical of cold dark clouds) or in the millimeter to submillimeter range (typical of hot molecular cores). Given its structural similarity to both CH3CH2OH and CH2CHOH - two well-characterized interstellar molecules—it is plausible that the CH3CHOH radical is also present in astronomical environments. Completing the connection, the ethanol radicals, once formed, react with predominantly H to yield the spike in acetaldehyde abundance observed in Figure 7. Thus, our model results show that, in times and places where gas-phase ethanol is present, our new destruction reactions with halogens represent efficient pathways linking alcohols and aldehydes.
Conclusion
In this work, we have investigated a new chemical link between alcohols and aldehydes through use of quantum chemistry paired with astrochemical models. Specifically, rather than focusing on a “bottom-up” formation route involving H-addition to grain-surface aldehydes, we here investigated a “top-down” route involving an initial reaction of an alcohol, in this case ethanol, with the halogens chlorine and in the gas phase. We find that, in particular, the formation of CH3CHOH by this route is efficient, owing to barrierless reaction pathways, and subsequent gas-phase H-abstraction could efficiently yield acetaldehyde in cases where gas-phase ethanol is abundant. An examination of the calculated abundances of our models including these new reactions predicts peak abundances of CH3CHOH comparable with that of the parent species ethanol, which was initially detected in Sgr B2 by Zuckerman et al. (1975). This finding suggests that CH3CHOH may likewise represent a potential target for future studies in this or similar sources, though to our knowledge there is no spectroscopic data on this species.
There exist many potential future directions for this work. Firstly, further study of the subsequent reactions of the ethanol radicals is warranted. Moreover, experimental and theoretical studies of this and similar systems could verify whether or not this reaction route with halogens is accessible for a broad range of alcohols.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
CS: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review and editing. GM: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review and editing. AF: Conceptualization, Methodology, Writing – original draft, Writing – review and editing. DW: Investigation, Software, Visualization, Writing – original draft, Writing – review and editing. ES: Investigation, Writing – original draft, Writing – review and editing. AR: Conceptualization, Investigation, Supervision, Writing – original draft, Writing – review and editing.
Acknowledgements
CS thanks A. Byrne for compiling the model inputs used as a starting point for those used in this work.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Edited by:Ryan C. Fortenberry, University of Mississippi, United States
Reviewed by:Kotomi Taniguchi, National Astronomical Observatory of Japan (NAOJ), Japan
Yong Zhang, Sun Yat-Sen University, China
Kinsuk Acharyya, Physical Research Laboratory, India
Bernardo Ballotta, Trinity College Dublin, Ireland
https://zenodo.org/records/13257329
ReferencesAcharyyaK.HerbstE. (2017). Gas-grain fluorine and chlorine chemistry in the interstellar medium. Astrophysical J.850, 105. 10.3847/1538-4357/aa937eAsgeirssonV.BirgissonB. O.BjornssonR.BeckerU.NeeseF.RiplingerC. (2021). Nudged elastic band method for molecular reactions using energy-weighted springs combined with eigenvector following. J. Chem. Theory Comput.17, 4929–4945. 10.1021/acs.jctc.1c0046234275279AsplundM.GrevesseN.SauvalA. J.ScottP. (2009). The chemical composition of the sun. Annu. Rev. Astronomy Astrophysics47, 481–522. 10.1146/annurev.astro.46.060407.145222BalucaniN.CeccarelliC.TaquetV. (2015). Formation of complex organic molecules in cold objects: the role of gas-phase reactions. Mon. Notices R. Astronomical Soc.449, L16–L20. 10.1093/mnrasl/slv009BartlettR. J.WattsJ.KucharskiS.NogaJ. (1990). Non-iterative fifth-order triple and quadruple excitation energy corrections in correlated methods. Chem. Phys. Lett.165, 513–522. 10.1016/0009-2614(90)87031-LBellocheA.MüllerH. S. P.MentenK. M.SchilkeP.ComitoC. (2013). Complex organic molecules in the interstellar medium: IRAM 30 m line survey of sagittarius B2(N) and (M). Astronomy Astrophysics559, A47. 10.1051/0004-6361/201321096[Dataset]BjornssonR. (2022). ASH – a multiscale modelling program. Available online at: https://ash.readthedocs.io/en/latest/ (Accessed June 21, 2025).BonfandM.BellocheA.GarrodR. T.MentenK. M.WillisE.StéphanG. (2019). The complex chemistry of hot cores in Sgr B2(N): influence of cosmic-ray ionization and thermal history. Astronomy and Astrophysics628, A27. 10.1051/0004-6361/201935523ByrneA. N.XueC.VoorhisT. V.McGuireB. A. (2024). Sensitivity analysis of aromatic chemistry to gas-phase kinetics in a dark molecular cloud model. Phys. Chem. Chem. Phys.26, 26734–26747. 10.1039/D4CP03229B39431338CaldeweyherE.EhlertS.HansenA.NeugebauerH.SpicherS.BannwarthC. (2019). A generally applicable atomic-charge dependent London dispersion correction. J. Chem. Phys.150, 154122. 10.1063/1.509022231005066ChesnavichW. J. (1986). Multiple transition states in unimolecular reactions. J. Chem. Phys.84, 2615–2619. 10.1063/1.450331DasA.ChakrabartiS. K.AcharyyaK.ChakrabartiS. (2008). Time evolution of simple molecules during proto-star collapse. New Astron.13, 457–467. 10.1016/j.newast.2008.01.003EstebanC.PeimbertM.García-RojasJ.RuizM. T.PeimbertA.RodríguezM. (2004). A reappraisal of the chemical composition of the Orion nebula based on very large telescope echelle spectrophotometry. Mon. Notices R. Astronomical Soc.355, 229–247. 10.1111/j.1365-2966.2004.08313.xGarrodR. T. (2013). A three-phase chemical model of hot cores: the formation of glycine. Astrophysical J.765, 60. 10.1088/0004-637X/765/1/60GarrodR. T.WakelamV.HerbstE. (2007). Non-thermal desorption from interstellar dust grains via exothermic surface reactions. Astronomy Astrophysics467, 1103–1115. 10.1051/0004-6361:20066704GarrodR. T.WeaverS. L. W.HerbstE. (2008). Complex chemistry in star-forming regions: an expanded gas-grain warm-up chemical model. Astrophysical J.682, 283–302. 10.1086/588035GeorgievskiiY.MillerJ. A.BurkeM. P.KlippensteinS. J. (2013). Reformulation and solution of the master equation for multiple-well chemical reactions. J. Phys. Chem. A117, 12146–12154. 10.1021/jp406070424053787GillanM. J. (1987). Quantum-classical crossover of the transition rate in the damped double well. J. Phys. C Solid State Phys.20, 3621–3641. 10.1088/0022-3719/20/24/005GotoM. (2013). The cosmic ray ionization rate in the central parsec of the galaxy. Proc. Int. Astronomical Union9, 429–433. 10.1017/S1743921314001070GrimmeS.GoerigkL.FinkR. F. (2012). Spin-component-scaled electron correlation methods. WIREs Comput. Mol. Sci.2, 886–906. 10.1002/wcms.1110HenkelmanG.UberuagaB. P.JónssonH. (2000). Climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys.113, 9901–9904. 10.1063/1.1329672HillJ. G.MazumderS.PetersonK. A. (2010). Correlation consistent basis sets for molecular core-valence effects with explicitly correlated wave functions: the atoms B–Ne and Al–Ar. J. Chem. Phys.132, 054108. 10.1063/1.330848320136306HiraokaK.YamashitaA.MiyagoshiT.OohashiN.KiharaY.YamamotoK. (1998). Reactions of hydrogen atoms with solid, thin films of acetone and 2-Propanol at 13 K. Astrophysical J.508, 423–430. 10.1086/306385IoppoloS.FedoseevG.ChuangK.-J.CuppenH. M.ClementsA. R.JinM. (2020). A non-energetic mechanism for glycine formation in the interstellar medium. Nat. Astron.5, 1–9. 10.1038/s41550-020-01249-0JodkowskiJ. T.RayezM.-T.RayezJ.-C.BércesT.DóbéS. (1999). Theoretical study of the kinetics of the hydrogen abstraction from methanol. 3. Reaction of methanol with hydrogen atom, methyl, and hydroxyl radicals. J. Phys. Chem. A103, 3750–3765. 10.1021/jp984367qKendallR. A.DunningT. H.HarrisonR. J. (1992). Electron affinities of the first-row atoms revisited. Systematic basis sets and wave functions. J. Chem. Phys.96, 6796–6806. 10.1063/1.462569KozuchS.MartinJ. M. (2011). DSD-PBEP86: in search of the best double-hybrid DFT with spin-component scaled MP2 and dispersion corrections. Phys. Chem. Chem. Phys.13, 20104–20107. 10.1039/c1cp22592h21993810LaasJ. C.CaselliP. (2019). Modeling sulfur depletion in interstellar clouds. Astronomy Astrophysics624, A108. 10.1051/0004-6361/201834446LoomisR. A.ZaleskiD. P.SteberA. L.NeillJ. L.MuckleM. T.HarrisB. J. (2013). The detection of interstellar ethanimine (CH3CHNH) from observations taken during the GBT PRIMOS survey. Astrophysical J.765, L9. 10.1088/2041-8205/765/1/L9MarchioneD.ManciniL.LiangP.VanuzzoG.PiraniF.SkouterisD. (2022). Unsaturated dinitriles formation routes in extraterrestrial environments: a combined experimental and theoretical investigation of the reaction between cyano radicals and cyanoethene (C2H3CN). J. Phys. Chem. A126, 3569–3582. 10.1021/acs.jpca.2c0180235640168McClureM. K.RochaW. R. M.PontoppidanK. M.CrouzetN.ChuL. E. U.DartoisE. (2023). An ice age JWST inventory of dense molecular cloud ices. Nat. Astron.7, 431–443. 10.1038/s41550-022-01875-wMcGuireB. A. (2018). 2018 census of interstellar, circumstellar, extragalactic, protoplanetary disk, and exoplanetary molecules. Astrophysical J. Suppl. Ser.239, 17. 10.3847/1538-4365/aae5d2MolpeceresG.NguyenT.ObaY.WatanabeN. (2025). Hydrogenation of acetaldehyde on interstellar ice analogs results in limited destruction. Astronomy and Astrophysics694, A299. 10.1051/0004-6361/202451990MondalS. K.GoraiP.SilM.GhoshR.EtimE. E.ChakrabartiS. K. (2021). Is there any linkage between interstellar aldehyde and alcohol?Astrophysical J.922, 194. 10.3847/1538-4357/ac1f31NandiA.MolpeceresG.GuptaP. K.MajorD. T.KästnerJ.MartinJ. M. L. (2022). “Quantum tunneling in computational catalysis and kinetics: is it really important?,” in Comprehensive computational chemistry (Elsevier). 10.1016/B978-0-12-821978-2.00020-9NeeseF. (2000). Approximate second-order SCF convergence for spin unrestricted wavefunctions. Chem. Phys. Lett.325, 93–98. 10.1016/S0009-2614(00)00662-XNeeseF. (2022). Software update: the ORCA program system—Version 5.0. WIREs Comput. Mol. Sci.12, e1606. 10.1002/wcms.1606NeeseF.WennmohsF.BeckerU.RiplingerC. (2020). The ORCA quantum chemistry program package. J. Chem. Phys.152, 224108. 10.1063/5.000460832534543NeillJ. L.BerginE. A.LisD. C.SchilkeP.CrockettN. R.FavreC. (2014). Herschel observations of extraordinary sources: analysis of the full Herschel/HIFI molecular line survey of sagittarius B2(N). Astrophysical J.789, 8. 10.1088/0004-637X/789/1/8NeufeldD. A.ZmuidzinasJ.SchilkeP.PhillipsT. G. (1997). Discovery of interstellar hydrogen fluoride 1. Astrophysical J. Lett.488, L141–L144. 10.1086/310942NeufeldD. A.WolfireM. G.SchilkeP. (2005). The chemistry of fluorine-bearing molecules in diffuse and dense interstellar gas clouds. Astrophysical J.628, 260–274. 10.1086/430663PechukasP.LightJ. C. (1965). On detailed balancing and statistical theories of chemical kinetics. J. Chem. Phys.42, 3281–3291. 10.1063/1.1696411PerreroJ.UgliengoP.CeccarelliC.RimolaA. (2023). Quantum mechanical modelling of the grain-surface formation of acetaldehyde on H2O:CO dirty ice surfaces. Mon. Notices R. Astronomical Soc.525, 2654–2667. 10.1093/mnras/stad2459QasimD.FedoseevG.ChuangK. J.TaquetV.LambertsT.HeJ. (2019). Formation of interstellar propanal and 1-propanol ice: a pathway involving solid-state CO hydrogenation. Astronomy Astrophysics627, A1. 10.1051/0004-6361/201935217QasimD.FedoseevG.ChuangK.-J.HeJ.IoppoloS.van DishoeckE. F. (2020). An experimental study of the surface formation of methane in interstellar molecular clouds. Nat. Astron.4, 781–785. 10.1038/s41550-020-1054-yRaghavachariK.TrucksG. W.PopleJ. A.Head-GordonM. (1989). A fifth-order perturbation comparison of electron correlation theories. Chem. Phys. Lett.157, 479–483. 10.1016/S0009-2614(89)87395-6RimolaA.TaquetV.UgliengoP.BalucaniN.CeccarelliC. (2014). Combined quantum chemical and modeling study of CO hydrogenation on water ice. Astronomy Astrophysics572, A70. 10.1051/0004-6361/201424046RuaudM.WakelamV.HersantF. (2016). Gas and grain chemical composition in cold cores as predicted by the nautilus three-phase model. Mon. Notices R. Astronomical Soc.459, 3756–3767. 10.1093/mnras/stw887SantraG.SylvetskyN.MartinJ. M. L. (2019). Minimally empirical double-hybrid functionals trained against the GMTKN55 database: revDSD-PBEP86-D4, revDOD-PBE-D4, and DOD-SCAN-D4. J. Phys. Chem. A123, 5129–5143. 10.1021/acs.jpca.9b0315731136709ScibelliS.ShirleyY. (2020). Prevalence of complex organic molecules in starless and prestellar cores within the taurus molecular cloud. Astrophysical J.891, 73. 10.3847/1538-4357/ab7375SongL.KästnerJ. (2017). Tunneling rate constants for H2CO+H on amorphous solid water surfaces. Astrophysical J.850, 118. 10.3847/1538-4357/aa943eTaketaniF.TakahashiK.MatsumiY.WallingtonT. J. (2005). Kinetics of the reactions of Cl*(CH2 P1∕2) and Cl(2 P3∕2) atoms with CH3 OH, C2 H5 OH, n -C3 H7 OH, and i -C3 H7 OH at 295 K. J. Phys. Chem. A109, 3935–3940. 10.1021/jp050055t16833712TsikriteaA.DiproseJ. A.SoftleyT. P.HeazlewoodB. R. (2022). Capture theory models: an overview of their development, experimental verification, and applications to ion–molecule reactions. J. Chem. Phys.157, 060901. 10.1063/5.009855235963719VazartF.CeccarelliC.BalucaniN.BianchiE.SkouterisD. (2020). Gas-phase formation of acetaldehyde: review and new theoretical computations. Mon. Notices R. Astronomical Soc.499, 5547–5561. 10.1093/mnras/staa3060WatanabeN.KouchiA. (2002). Efficient formation of formaldehyde and methanol by the addition of hydrogen atoms to CO in H[TINF]2[/TINF]O-CO ice at 10 K. Astrophysical J.571, L173–L176. 10.1086/341412WilliamsA. E.HammerN. I.TschumperG. S. (2021). Relative energetics of CH3CH2O, CH3CHOH, and CH2CH2OH radical products from ethanol dehydrogenation. J. Chem. Phys.155, 114306. 10.1063/5.006280934551536WoonD. E. (2002). Modeling gas-grain chemistry with quantum chemical cluster calculations. I. Heterogeneous hydrogenation of CO and H2CO on icy grain mantles. Astrophysical J.569, 541–548. 10.1086/339279YangY.-L.GreenJ. D.PontoppidanK. M.BergnerJ. B.CleevesL. I.IIN. J. E. (2022). CORINOS. I. JWST/MIRI spectroscopy and imaging of a class 0 protostar IRAS 15398–3359. Astrophysical J. Lett.941, L13. 10.3847/2041-8213/aca289ZaleskiD. P.SeifertN. A.SteberA. L.MuckleM. T.LoomisR. A.CorbyJ. F. (2013). Detection of E-Cyanomethanimine toward sagittarius B2(N) in the green bank telescope PRIMOS survey. Astrophysical J.765, L10. 10.1088/2041-8205/765/1/L10ZhengJ.TruhlarD. G. (2012). Multi-path variational transition state theory for chemical reaction rates of complex polyatomic species: ethanol + OH reactions. Faraday Discuss.157, 59–140. 10.1039/c2fd20012k23230764ZmuidzinasJ.BlakeG. A.CarlstromJ.KeeneJ.MillerD. (1995). HC[CLC]l[/CLC] absorption toward sagittarius B2. Astrophysical J.447. 10.1086/309570ZuckermanB.TurnerB. E.JohnsonD. R.LovasF. J.FourikisN.PalmerP. (1975). Detection of interstellar trans-ethyl alcohol. Astrophysical J. Lett.196, L99–L102. 10.1086/181753