Comprehensive computational automated search of barrierless reactions leading to the formation of benzene and other C6-membered rings

Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 1 of 10 CHEMICAL PHYSICS Comprehensive computational automated search of barrierless reactions leading to the formation of benzene and other C6- membered rings Marta Castiñeira Reis1*, Emilio Martínez Núñez2, Antonio Fernández Ramos1,2* We present the systematic exploration of various potential energy surfaces for systems with C6H6–x (x=0, 1, 2, or3) empirical formula using an automatic search approach. The primary objective of this study is to identify reaction pathways that lead to the creation of benzene, obenzyne, and other rings. These pathways initiate with a barrierless recombination reaction and involve subsequent isomerization reactions with submerged transition states until the final product is reached. The reported reaction profiles are consistent with the existing conditions in the interstellar medium if the hot complex formed can cool down through radiative relaxation. Recent studies on the deactivation of polyaromatic hydrocarbons (PAHs) support the possibility of these reactions taking place. The C6- membered rings are considered precursors of PAHs, and our focus is on identifying pathways originating from the barrierless recombination of reactive molecules known to exist in the interstellar medium, with potential implications in other environments. INTRODUCTION Small hydrocarbon molecules are widely present in combustion chemistry, the atmosphere, and the interstellar medium (ISM). They are thought to play a crucial role in the creation of more complex hydrocarbon molecules, such as benzene (C6H6) (1), obenzyne (2), higherorder polyaromatic hydrocarbons (PAHs) such as indene (c- C9H8) (3, 4), or derivatives such as 2- cyanoindene (2- C9H7CN) (5). It is suggested that around 10 to 25% of interstellar carbon exists in the form of PAHs, indicating their substantial presence in space (6, 7). The most accepted hypothesis is that PAHs are formed in the ISM from benzene or the phenyl radical through reactions involving acetylene or other carbonaceous scaffolds (8–10). For instance, the phenyl radical may be key in the formation of PAHs (10–12) and cyanonaphthalenes (9), and the 1,2,3- tridihydrobenzene and the 1- hexene- 3,5- diynyl- 2 radical (l- H2H6H) have been obtained in crossbeam experiments of C2 with vinylacetylene (13). In addition, 1,2,3- tridihydrobenzene (or 1,2,3- cyclohexatriene) due to the ring strain may be key in the synthesis of complex organic molecules (14). One of the pioneering computational works that explore the formation of benzene was carried out by Miller and Klippenstein (15). These authors found that benzene could form by recombination of two propargyl (CH2CCH) radicals after several isomerization reactions at temperatures between 300 and 2200 K and different pressures. The recent discovery of the propargyl radical in the TMC- 1 cold dark cloud (16) has strengthened the hypothesis that this radical may be a key species in the formation of benzene. Other species that have a direct relation to dehydrogenated forms of benzene, such as obenzyne, are vinylacetylene(but- 1- en- 3- yne) (17) and other C4H4 isomers (18), which can be formed by reaction with singlet or triplet spin states of the C2 molecule (13). Because of the low densities of reactants and ultralow temperatures in the ISM, the recombination reaction between two species must be barrierless, forming a hot intermediate that may dissociate back to reactants or rearrange to produce two different fragments. This reaction of the type A+B → C+D may occur in a onestep process where the products are more stable than the reactants or in a multistep process with submerged (below reactants) transition states (19). In principle, the reaction of two propargyl radicals to form benzene seems highly improbable because a hot complex is formed that will dissociate back to reactants or two different products. However, there is also the possibility that the hot complex formed by the collision of A and B loses part of its energy excess leading to only one product. Traditionally, it has been believed that because of the low density of the reacting species, the complex cannot dissipate its energy excess and dissociates back to reactants. However, recent research by Stockett etal. (20) suggests that radiative relaxation is more effective than previously assumed. This process allows the hot complex to cool down efficiently, ultimately forming a cold product. These findings have notable implications, as they explain the resilience of small PAHs, such as cyanonaphthalene and naphthalene (21), from dissociating back to reactants. Furthermore, this evidence opens up the possibility of similar reactions occurring for other cyclic species such as benzene and its derivatives. An additional possibility for this type of reaction to occur is on a grain surface because the latter can dissipate the energy excess of the complex. The aim of this study is to use automated computational methods to explore the potential energy surface of benzene, obenzyne, and other C6- membered rings. The objective is to identify a diverse variety of barrierless bimolecular reactions of the form A+B → C, which exhibit submerged transition states and could potentially lead to one of the aforementioned species. Our focus will be on reactants that have been observed in the ISM or that have been theorized to exist, even if they have not yet been detected. In this work, 1Centro Singular de investigación en Química Biolóxica e Materiais Moleculares (CiQUS), Campus vida, 15782, Universidade de Santiago de Compostela, C/Jenaro de la Fuente s/n, Santiago de Compostela, Spain. 2departamento de Química Física, Facultade de Química, Campus vida, 15782, Universidade de Santiago de Compostela, Avda. das Ciencias s/n, Santiago de Compostela, Spain. *Corresponding author. email: marta. castineira. reis@ usc. es (M.C.R.); qf. ramos@ usc. es (A.F.R.) Copyright © 2024 the Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. no claim to original U.S. Government Works. distributed under a Creative Commons Attribution license 4.0 (CC BY). Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 2 of 10 we identify all these potential paths, but the kinetic mechanism responsible for stabilizing the final product falls outside the scope of this research. For the generation of the potential energy surfaces, we rely on software for the automatic search of stationary points. The benefit of using an automated approach lies in its immunity to human bias, thereby allowing for a more uninhibited exploration of the potential energy surface. In this context, several programs have been designed to automatically find multistep reactions (22–30) and generate intricate reaction networks. Here, we have used AutoMeKin (22, 28, 31), which is a computational code designed to study unimolecular decompositions. The program has demonstrated its proficiency in analyzing systems of comparable size to benzene, as evidenced by its application in studying the decomposition of indole (32). The automated algorithm identifies transition state structures (saddle points) for both isomerization and decomposition reactions. The equilibrium structures of these reactions are obtained by following the minimum energy path (MEP) from the transition state in the backward direction (toward reactants) and in the forward direction (toward products). This allows building a reaction network about the target species that includes the interconversion reactions between different isomers and the products of their decomposition. The search for barrierless decompositions is more involved and requires an additional step, as these reactions do not present a transition state. We use the nudged elastic band algorithm (33) to investigate fragmentation pathways originating from different isomeric forms of the target molecule. This approach makes it possible to identify previously unobserved transition states and processes without energy barriers. By convenience, we are approaching the problem from a different perspective, i.e., the fragments resulting from barrierless reactions will act as the initial reactants that combine to form the desired product. Notice that the reported potential energy surfaces may also serve as a future database for the study of the interconversion reactions between isomers and their decomposition into different products. All the information about the different reaction networks is available at Zenodo (34). RESULTS AND DISCUSSION We follow a general scheme in which each of the C6H6–x (x=0, 1, 2, or 3) rings are analyzed separately, beginning with benzene and ending with the C6H3 radicals. For the latter, some sixmembered rings and some linear carbon chains have similar energies, so it is at this stage when the rings start to form (13). The successive additions of hydrogen atoms are barrierless processes that systematically increase the stability of the ring, starting with the addition to the C6H3 to form benzyne, continuing with the formation of the phenyl radical, and ending with the formation of benzene. The hydrogen additions to the ring provide a common ground for a systematic study of these carbonaceous rings. In the following sections, first, we describe the formation of benzene; second, the formation of the phenyl radical and of benzyne; and last, we deal with the formation of several isomers with C6H3 empirical formula. Benzene Shocktube experiments carried out by Wu and Kern (35) found that two propargyl radicals (PR29) produce benzene as one of the major products. Miller and Melius (36) argued that benzene and the phenyl radical plus a hydrogen atom are the first rings formed after the recombination of the two propargyl radicals. A subsequent series of gas phase (37) and combustion studies (15, 38, 39) confirmed the formation of benzene, as well as that of fulvene and the phenyl radical. In addition, the work of Wilson etal. (40) that simulated the conditions of the atmosphere of Titan pointed out that the reaction of two propargyl radicals is the main route to benzene formation. AutoMeKin found 191 minima, 278 transition states, and 35 products of benzene decomposition at the M08HX/6- 31G(d,p) level. The reaction network generated from these structures predicts that the recombination of two propargyl radicals is a viable path to benzene formation. Figure1 shows two of these paths, but more details are given in the Supplementary Materials. One of them leads to the 1,2,4,5- hexatetraene (MIN15) and then to fulvene (MIN2). The latter is a key species in the recombination of the two propargyl radicals, although it is possible to find a path that avoids fulvene and that passes through the 1,2- hexadiene- 5- yne (MIN26). Labels of stationary points of the initial benzene network start with MIN or TS for equilibrium or transition state structures, respectively. To distinguish the stationary points that originate from different reaction networks, the prefixes A, B, C, etc., are added to the labels with a given empirical formula using the following convention: A (C6H5), B (C6H4), C (C6H3), D (C6H2), E (C4H4), F (C5H3), G (C5H2), H (C5H), and I (C4H4). Benzene can also be formed from the barrierless reaction of the phenyl radical with atomic hydrogen. Vinylacetylene and acetylene can also react to form benzene, but the initial reaction has a barrier of 31.88 kcal mol−1, so this path was discarded. When vinylidene (CCH2) substitutes acetylene the reaction is barrierless. Unfortunately, this carbene remains elusive to detection in the ISM, although the existence of this reactive intermediate has been hypothesized many years ago (41). We also included C5H3 species because the recombination reaction of this radical with the methyl group can be obtained from the photodissociation of benzene (42); in particular, the excitation of benzene at 193 and 248 nm produces 4% of C5H3+CH3. All reactions between these two fragments involve the formation of the hexa- 1,2,3,4- tetraene (MIN22) that evolves by an H- shift to a carbene and undergoes another H- shift to form the hexa- 1,2,3,5- tetraene (MIN11). This intermediate evolves to a carbene that reacts to form a bicyclic species (MIN64) and the subsequent C6- membered ring species (MIN21), which can evolve to MIN44 to form the phenyl radical plus atomic hydrogen or to MIN41 to yield obenzyne plus H2 or benzene. However, the high barrier heights involved (near the dissociation energy of the two fragments) and the large number of elementary steps indicate that this path will marginally contribute to the formation of C6- membered rings, a result that is in agreement with experimental findings. The reactions of different C4H4 isomers (vinyl ethylene, butatriene, and methylenecyclopropene) with acetylene all have initial barriers. The reaction between the vinyl and butatrienyl radicals is barrierless, but none of the two species has been detected in the ISM. A similar situation occurs with the reaction between the propargyl and cyclopropenyl (c- C3H3) radicals because the latter has not been encountered in the ISM. The phenyl radical The number of minima, transition states, and products optimized at the density functional theory (DFT) method are 181, 268, and 356, Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 3 of 10 respectively. The phenyl radical (c- C6H5) can be formed from the reaction of the hydrogen atom with obenzyne (see Fig.2), but the reaction occurs in two steps. The initial barrierless reaction leads to an intermediate (A- MIN13), with an energy of 32.89 kcal mol−1 below reactants, that easily evolves to the final product. However, obenzyne is not the only species with C6H4 empirical formula that can lead to the phenyl radical. A full list of the currently undetected reactants compatible with the ISM conditions is given in the Supplementary Materials. Alternative paths to the obenzyne route involve additions to a triple bond, such as the addition of the methylidyne radical, methylene (43), methyl radical (44), ethynyl radical (45), and propadienylidene (l- H2C3) (46) to C5H4, C5H3, C5H2, C4H4, and C3H3 species, respectively. The reaction of the methylidyne radical (CH) with 1,4- pentadiyne [HCCCH2CCH—an isomer of allenyl acetylene, whose existence in the ISM could not be established (47)], allows forming the phenyl radical in just three steps: (i) addition, (ii) cyclization, and (iii) hydrogen shift. The reaction path involves two lowbarrier transition states because, in this case, there are no hydrogen shift reactions associated with the acyclic species. AutoMeKin predicts that two C5H3 isomers, the 1,4- pentadiynyl- 3 and the ethynyl cyclopropenylidene radicals, can react with methylene by an addition to the triple bond. The first C5H3 isomer forms the A- MIN14 intermediate species, which undergoes two consecutive hydrogen shift reactions and a further cyclization to produce the final product. The second C5H3 isomer leads to the phenyl radical in three steps: addition (to form A- MIN53), hydrogen shift, and cyclization. More information about the most stable C5H3 isomers is given in the Supplementary Materials. The reaction of the ethynyl radical with vinylacetylene offers more possibilities because both molecules present triple bonds. Figure2 shows one of the paths in which it is possible to obtain the phenyl radical in just three steps, i.e., addition to the triple bond, cyclization, and hydrogen transfer. The reaction of the propargyl radical with propadienylidene forms an allenyl propargyl radical (A- MIN11), which, by hydrogen shift, leads to the intermediate A- MIN43 that forms a cycle (A- MIN15) and undergoes a new hydrogen shift to yield (A- MIN1). Another possibility is the addition of the methyl radical to pentatetraenylidene that produces the phenyl radical in five steps, of which the last three are common to the reaction starting with the two propargyl radicals. o- Benzyne Many pathways may lead to obenzyne, and the reaction network is formed by 278 minima, 552 transition states, and 201 products. As CCC C C C C C C C CC H H H H H H· H H H H H H PR25/A-MIN1 111.2 CCC C C CH H H HH MIN1 0.00 H MIN41 90.21 TS6 90.25 TS23 106.28 TS11 94.51 CCC C C C H H H H MIN21 76.34 H HCCCC CC H H H HH H MIN4 62.19 TS17 100.11 TS147 134.02 CCCC CC H H HH H MIN140 129.72 H CCCCCC H HHH HH PR28 168.28 TS242 145.63 CCC C C C H H H HH MIN64 103.98 HTS223 139.22 CCCCH H H H C C MIN112 124.19 H H TS206 143.6 CCCC CC H H H H MIN11 70.61 HH TS200 143.77 CCCC CC H H H MIN149 133.95 HH H TS211 145.84 CCC C C C H H H MIN22 79.89 H H H CCC C C H H H PR35 149.86 ·CH3 C C C C C C H H H H MIN33 86.99 H H TS199 136.22 TS241 144.98 TS219 139.07 CCCC CC H H H H MIN171 84.40 HH C C C C C C H H H H MIN63 100.2 H H CCC C C C H H H H H H PR32 178.76 H H CCC HH H MIN26 80.02 C C C H CCC H H H PR29 146.91 H C C C C H H MIN31 85.24 H TS132 130.3 H C C H TS246 144.03 PR7-vdW 143.21 H C C CC H H H H C C H PR7 146.04 H C C C H H MIN15 75.06 H H C H C C CC C C C C H H H H MIN2 31.56 H H TS178 140.83 TS40 111.52 CC C C C C H H H MIN79 108.85 H H H TS42 111.16 CC C C C C H H H MIN8 72.87 H H H TS15 101.15 2 CCC C C C H H H H PR1 88.64 TS77 122.2 H2 TS44 110.0 CCC C C C H H H H MIN44 92.70 H H Fig. 1. Multistep barrierless reactions encountered by AutoMeKin leading to benzene. in gray are the reactive species that have not been yet detected in the iSM. All F12- CCSd(t) relative energies include the density functional theory (dFt) zeropoint energies. Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 4 of 10 shown in Fig. 3, the most straightforward path is through 1,2,3- tridehydrobenzene because it is a barrierless onestep reactions. However, this species has not yet been detected in the ISM. Another possibility is the reaction of the propargyl radical with l- C3H (B- PR108) or with c- C3H (B- PR150). In both cases, there is a first association step with a large stabilization and a cyclization to a common species that suffers an initial hydrogen transfer to form fulvenylidene (B- MIN9) and a final (nearly barrierless) ring expansion to produce obenzyne. The mechanism proposed by Kaiser etal. (13) that starts with adding C2 (singlet) to the terminal carbon of the double bond of vinylacetylene was also found. The recombination reaction leads to B- MIN157 [si3—nomenclature from (13)], which, by an internal C─C rotation, leads to B- MIN55, and this species undergoes cyclization to form fulvenylidene, which easily evolves to obenzyne (see Fig.3). Figure 3 shows that vinyldiacetylene can be formed from the recombination reaction of the methyl radical and pentadiynylidene (l- C5H) (48) with an energy of 126.01 kcal mol−1 above obenzyne. AutoMeKin indicates that l- H2C6H can be formed from hexapentaenylidene (l- H2C6) (49) plus the hydrogen atom or from butatrienylidene (l- H2C4) (50) plus the ethynyl radical (not displayed in Fig.3) The formation of obenzyne from 1- hexene- 3,5- diynyl- 2 (l- H2C6H) plus atomic hydrogen leads to several possibilities that are explained in more detailed in the Supplementary Materials. In addition, we encountered that two compounds with C5H3 empirical formula can recombine with methylidene and lead to obenzyne after a few steps. The first is the 2,4- pentadiynyl- 1 radical, already mentioned above, and the other is the 1,4- pentadiynyl- 3 radical (HCCCHCCH). The later can be formed from the recombination of the ethynyl radical plus cyclopropenylidene (c- C3H2) (51), followed by a ring opening and H- shift. The 1,4- pentadiynyl- 3 radical C C C C CC H H H H H CC·H C C H CC H C C C C CC H H H H H H H· A-MIN1 0.00 MIN1 A-TS4 61.29 A-MIN13 47.61 A-PR14 80.5 C C C C CC H H H H H C C C C CC H H H H H· A-TS23 80.7 A-MIN15 47.61 C C C C CC H H H H H A-TS109 101.69 H H HH H CCC CC C A-MIN43 70.01 A-TS125 111.92 A-TS220 140.13 A-PR233 134.28 CCCCCC H H H A-MIN5 44.65 H CCC CCC H H H H H H CCCCC H A-PR218 140.93 H A-TS41 89.02 A-MIN62 74.75 C C C C CC HH H H H A-TS119 106.59 HH H H H CC CC A-MIN106 100.19 C C A-PR129 122.88 A-TS12 68.03 CCC C C C H H H H A-MIN24 47.61 H A-TS112 105.98 CCC H H H H A-MIN6 47.1 C C H C A-MIN65 69.93 C C C C CC H H H A-TS35 86.08 H H A-TS221 135.58 CC C C C C H H H A-MIN171 122.31 H H CH CC CCC HH A-PR330 168.27 HH CCC C C C H H H A-MIN11 47.61 H H H H ·CH3 A-TS38 81.0 A-TS100 106.71 A-MIN14 47.61 A-PR150 177.83 H C H CCC H C CH H CH2 C C H C C C H H A-TS209 125.58 A-MIN68 85.78 C C CC CC HH H H H A-TS204 134.1 A-MIN53 77.42 C C C CCC H H H H H A-PR82 220.94 CH2 C C H C CC H H Fig. 2. Same as Fig.1 but for the formation of the phenyl radical. Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 5 of 10 may also lead to mbenzyne (B- MIN2) because it is the previous step to the obenzyne formation. C6H3 structures The reaction path connections between reactive ISM molecules and the most stable C6H3 structures are given in Fig.4. Kaiser etal. (13) considered the C6H3 structures “as the crucial link between energetically favored aromatic and resonantly stabilized radicals for hydrogenrich and poor C6Hx molecules, respectively.” Certainly, this is the case for the 6- bicyclo[3.1.0]hexa- 1,3,5- trienyl radical (C- MIN1), 1,2,3- tridehydrobenzene (C- MIN3), and 1,2,4- tridehydrobenzene (C- MIN4) that are very close energy to l- H2C6H (C- MIN2). Therefore, molecular structures with C6H3 empirical formula are the ones with the least number of hydrogen atoms of the type C6Hx (where x=1, …, 6) for which the closed ring is very close in energy to the open carbon chain (see Fig. 4). The ring C C C C CC H H H H H A-MIN1/PR25 C C C C CC H H H H B-MIN1/A-PR14 0.00 C C C C CC H H H H B-MIN106 86.53 B-TS44 80.69 C C C C C C H B-MIN74 72.54 B-TS92 88.88 H H H C C C C C C H B-MIN71 72.08 H H H B-TS36 82.47 B-TS1 31.25 CC C C CC H H B-MIN9/B-MIN9* 30.84 H H CC C C CC H H B-MIN133 88.78 H B-TS116 98.92 H B-TS75 89.74 B-TS258 117.64 C C C H H B-MIN36 64.61 H H C CC CC C C C C H B-PR108 144.33 H H H CC C C CC HB-MIN19 58.72 H H H CCC H H H CC C H B-PR150 142.49 C C C C CC H H B-MIN66 69.8 B-TS33 78.42 H H B-TS57 87.37 B-TS97 87.81 C CCCCC H B-MIN3 14.12 H HCCCCC B-PR141 126.01 H B-TS14 72.03 C C C C CC H H B-MIN2 15.19 HH B-TS21 72.92 C CC C C C H H H B-MIN34 63.06 H C C C C C C H H H B-MIN178 101.4 H B-TS24 80.3 C C C C CC HH B-MIN13 51.79 H H B-TS84 92.83 C C C C CC HH B-MIN143 95.71 H H B-TS455 137.81 C C C C C C B-MIN146 93.59 HH HH C CCC C H H B-PR134 167.72 H CH C C C C CC H H B-PR100 107.52 H CH H· H· ·CH3 CCCCC B-PR113 166.26 H H H C C C C CC H H B-MIN98 78.09 H H H B-TS41 83.3 C CCCCC H HH HB-MIN53 66.6 B-TS304* 122.44 C C CH H B-MIN55* 73.27 C CC H H B-TS201* 116.38 C C C C CC H B-MIN157* 113.92 H H H C CC C H B-PR213* 154.85 HH H CC C C C C CC H B-MIN68* 74.88 H H H B-TS401* 137.47 B-TS228* 97.9 H· CCCCCC B-PR109 110.05 H H H B-TS154 101.56 Fig. 3. Same as Fig.1 but for the formation of obenzyne. Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 6 of 10 formation is unfavorable from an entropic point of view at high temperatures but competes with the other mechanisms at ultralow temperatures because, in this case, the entropic contribution is negligible. The linear l- H2C6H radical can lead to obenzene by a further H- atom addition (vide supra). Crossed molecular beam experiments carried out by Guo etal. (52) found that this molecule can be formed from C3 (53) and allene (H2CCCH2). We found that the 1- hexene- 3,5- diynyl- 2 radical (Fig.4) can also be formed by the recombination of (i) hexapentaenylidene plus atomic hydrogen (C- PR26), (ii) pentadiynylidene plus methylene (C- PR116), (iii) ethynylcyclopropynylidyne (c- C3C2H) (54) plus methylene (C- PR191), (iv) butatrienylidene plus the ethynyl radical (C- PR29), (v) propadienylidene plus propynylidyne (C3H) (55) (C- PR95), (vi) reactions that involve C3- membered rings (C- PR25, C- PR32, and C- PR44), and (vi) other reactants that go through the C- MIN7 radical. In particular, this radical is a key intermediate between the open and closed forms of C6H3 structures. It can be reached by recombining the reactants mentioned above and by other three paths. The most direct path is the recombination of acetylene and the butadiynyl radical. Another possibility for the formation of cyclic structures from l- H2C6H is by direct cyclization through C- MIN71 and concerted H- shift and C─C bond formation leading to the C- MIN1 bicyclic structure. The addition of atomic hydrogen to 1,2,3- tridehydrobenzene and 1,2,4- tridehydrobenzene may also be a source of formation of obenzyne but not the bicyclo[3.1.0]hexa- 1,3,5- trienyl radical. Adding atomic or molecular hydrogen to the latter presents a barrier height. A new path starts with the recombination of the propargyl radical and c- C3, but c- C3 has not been detected in the ISM (not plotted in Fig.4). Other possibilities involve passing through other stable cyclic structures (C- MIN3 or C- MIN4), but once they are formed, their evolution to C- MIN1 is highly improbable due to the high barriers involved. The 1,2,3- tridehydrobenzene radical, apart from the paths that connect with the C- MIN7 intermediate, can be formed from C3 C C C C CC H H C-MIN3/B-PR100 0.74 H C-TS58 66.52 CCC C CC H C-MIN7 19.09 H H CCCCCCH H H C-MIN2/B-PR109 3.27 C-TS20 59.98 C-PR3 114.45 CCCC H H CCH C-TS88 83.86 C CCCC C H H C-MIN22 43.77 H C CCC CC H H H C-PR95 136.57 C-TS148 94.78 C C CC H C C H H C-MIN25 48.88 HC C CCC CH2 C-PR116 163.03 C-TS53 68.54 C-MIN14 58.44 CC C C C H C H H CCCCCC H H H C-PR26 98.69 CCH H C-PR51 80.72 CCCC H CCC C CC C-MIN91 72.82 H HH C-TS63 74.54 C-TS258 120.26 C CCC C C C-MIN35 53.79 H HH C-TS171 100.74 CCC CH C C C-MIN11 33.73 H H CCC CC C H H H C-PR25 134.28 C-TS45 72.91 H H C CC CCC H C-MIN46 57.5 CCH H CCCCH C-PR29 70.62 C C C C CC H H C-MIN4/B-PR101 3.72 H C-TS74 79.75 C-TS11 45.06 C CC CH C C C-MIN96 79.07 H H C C HC CH CH C C-MIN33 58.12 H C-PR44 120.13 CH CH C C C C H H C C CH CCC H C-MIN5 11.65 C C CH CCC H C-PR20 100.8 H C-TS67 74.84 H C CC CCC H C-MIN58 64.54 H C-TS208 109.84 C C C CCC H C-MIN50 57.37 H H CCHC H H CCC C-PR154 111.67 C-TS210 111.03 C C C C C C C-MIN24 50.74 H H H C-PR32 122.41 CC C HH CCC H C-TS146 99.94 C-TS79 85.89 CC C CC C C-MIN1 0.00 H H HC C C C CC H H H H B-MIN1/A-PR14 (106.78) H· H· C-TS43 73.88 C-TS6 41.48 C-MIN71 65.8 C-TS5 40.83 C C C C CC H H H C-TS73 71.58 HC CC C C CH2 C-PR91 180.05 Fig. 4. Same as Fig.1 but for the formation of C6H3 structures. Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 7 of 10 plus the propargyl radical or from propynylidyne (HC3) plus cyclopropenylidene (c- C3H2). Both paths form the cyclic radical in three steps and have in common the intermediate C- MIN58 in the last step. The 1,2,4- tridehydrobenzene can be formed from the recombination of cyclopropenylidene and the cyclopropanediylidenyl radical (c- C3H) (56, 57) after a ring opening of one of the rings, followed by a ring opening of the carbene and a further cyclization. It can also be formed from ethynylallene (HCCCHCCC) (58) plus the hydrogen atom in only two steps. We present an extensive analysis of both onestep and multistep reaction pathways for the A+B → C type of reactions, derived from a reaction network generated using automated methods. The reaction profiles initiate with the barrierless recombination of two reactive molecules, which, in the case of multistep reactions, pass through submerged transition states before culminating in the formation of aC6- membered ring hydrocarbon. These findings are summarized in Fig.5. With this prescription, we confirmed that benzene can be formed by associating two propargyl radicals. In addition, benzene can be reached by the reaction of H2C5H with the methyl radical or by the reaction of vinylidene (undetected) with vinylacetylene. The most favorable path for benzene formation is the barrierless reaction between the phenyl radical and atomic hydrogen. The analysis of the reaction networks for dehydrogenated forms of benzene such as the phenyl radical, obenzyne, or different cyclic C6H3 structures reveals that the formation of these rings is very involved and comprises a plethora of reactive species. Two encircled species in Fig.5, which are the phenyl and the C6H3 radicals, have not yet been encountered in the ISM, but their existence has been hypothesized. We found that several reactions start with different combinations of radical reactants such as propargyl, cylopropynil, methyl, or acetyl radicals. Other paths initiate with the combination of linear cumulenic carbon scaffolds (usually carbenes) plus a radical or a neutral molecule, depending on the multiplicity of the final product. Some paths consist of simple hydrogen atom additions to a ring; this provides a common ground for the formation of different C6 species. A molecule that is very abundant in the TMC- 1 cloud is vinylacetylene, which, in our scheme, plays a crucial role in the formation of obenzyne by the addition of C2 [in agreement with the calculations of Kaiser etal. (13)] and in the formation of the phenyl radical by the addition of the ethynyl radical (C2H). Another molecule that is ubiquitous in C6 rings formation is the propargyl radical, which appears in the formation of benzene, the phenyl radical, obenzyne, and 1,2,3- tridehydrobenzene by reaction with another propargyl molecule, propadienylidene, C3H (cyclic and linear forms), and C3, respectively. This completes the scheme depicted in Fig. 5. Last, Fig. 5. Summary scheme of different ISM- detected (in black) and undetected (in gray) species encountered by AutoMeKin as possible reactants that lead to benzene and other C6- membered rings. dashed arrows indicate paths that involve several elementary steps. Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 8 of 10 some pathways involve the reaction of up- to- date undetected species with C6H3 and C5H3 empirical formulas plus a small radical or carbene. The recent discovery that some PAHs undergo rapid radiative relaxation adds significance to the recombination paths identified in this study. This finding may serve as a foundation for developing kinetic models that incorporate these types of reactions. The abundance of viable pathways identified by AutoMeKin demonstrates the effectiveness of automated methods in constructing reaction networks and uncovering new reaction pathways that may have relevance in the ISM and other environments. MATERIALS AND METHODS The search for the stationary points (minima and transition states) was carried out with AutoMeKin (28). This program runs highenergy trajectories combined with an algorithm able to identify bondbreaking/forming events. Snapshots along the trajectories provide guess transition state structures that are further optimized at a low electronic structure level. For the lowlevel calculations, we used the PM7 semiempirical method (59) implemented in MOPAC (60). For each successfully obtained transition state, we follow the MEP, so the minima (reactants and products) associated with that particular transition state are also identified. In the event that the current transition state search algorithm fails to locate the appropriate saddle points, the nudged elastic band method (33) is used to validate the absence of saddle points for the identified barrierless pathways. This tool was implemented in AutoMeKin in 2022 (revision 1123). The pipeline starts with a search for bonds whose bond order is less than 1.5. The search is performed for every single minimum in the reaction network. It then runs a constrained Langevin dynamics simulation with external forces to break the selected bonds within each minimum energy structure. An elementary step is considered barrierless if it meets the following criteria: (i) It results in two fragments, (ii) the energy of the resulting fragments is below a specified threshold, and (iii) no alternative pathway with a barrier, connecting the same reactants and products, has been found. For the reactions with a barrier, the lowlevel transition state structures are further optimized at a high level, and new MEP calculations are performed to generate a more accurate reaction network. All the stationary points were optimized at the M08HX functional (61) with the 6- 31+G(d,p) basis set except the C2 plus vinylacetylene channel that used the B3LYP functional (62) and the 6- 311G(d,p) basis set. All the lowlevel and highlevel optimization calculations were performed with MOPAC and Gaussian 16 (63), respectively. For each DFT optimized geometry, we have confirmed that the wavefunction was stable. Highlevel Hessian calculations were performed to check the nature of the stationary points. Accurate electronic energies were obtained by singlepoint calculations carried out at the F12- CCSD(T)/ ccpVTZ level over the DFT optimizations. The F12- CCSD(T) calculations were performed with MOLPRO (64). All reported relative energies include the DFT- calculated zeropoint energy contribution to the F12- CCSD(T) electronic energy. A scaling factor of 0.972 was applied to the zeropoint energies (65). We have chosen M08HX because it has provided a good representation of the very challenging barrierless association of CF2 and the dissociation of C2F4 (66). In addition, it has been used by some of us in the study of the association reactions of OH with methanol (67) and methylamine (68), followed by the hydrogen abstraction reactions. In all cases, the combination F12- CCSD(T)//M08HX produced thermal rate constants in good agreement with experiment. An interesting alternative to these electronic structure calculations could be the junChS and junChS- F12 composite methods of Puzzarini and coworkers (69). The location of reaction paths inside the reaction network generated by AutoMeKin was performed with amk_tools (32, 70). The reaction networks for the different potential energy surfaces were built incorporating some restrictions. During the barrierless fragmentation of the molecule, AutoMeKin only examines the cleavage of a single bond with a bond order smaller than 1.5 (MOPAC calculates the bond order). Barrierless fragmentations that involve the cleavage of multiple bonds, as may be the case of a carbon atom or a C2 molecule dissociating from a threemembered ring, are not considered. Moreover, we included a spin restriction during the dissociation into fragments. In particular, the fragments with an even number of electrons have singlet spin states, so intersystem crossing to different electronic states (for instance, triplet states) is forbidden. Supplementary Materials This PDF file includes: Supplementary text Figs.S1 to S4 tablesS1 to S4 References REFERENCES AND NOTES 1. J. Cernicharo, A. M. heras, A. G. G. M. tielens, J. R. Pardo, F. herpin, M. Guélin, l. B. F. M. Waters, infrared space observatory’s discovery of C4h2, C6h2, and benzene in CRl 618. Astrophys. J. 546, l123–l126 (2001). 2. J. Cernicharo, M. Agúndez, R. i. Kaiser, C. Cabezas, B. tercero, n. Marcelino, J. R. Pardo, P. de vicente, discovery of benzyne, o- C6h4, in tMC- 1 with the QUiJOte line survey. Astron. Astrophys. 652, l9 (2021). 3. J. Cernicharo, M. Agúndez, C. Cabezas, B. tercero, n. Marcelino, J. R. Pardo, P. de vicente, Pure hydrocarbon cycles in tMC- 1: discovery of ethynyl cyclopropenylidene, cyclopentadiene, and indene. Astron. Astrophys. 649, l15 (2021). 4. A. M. Burkhardt, K. l. K. lee, P. B. Changala, C. n. Shingledecker, i. R. Cooke, R. A. loomis, h. Wei, S. B. Charnley, e. herbst, M. C. McCarthy, B. A. McGuire, discovery of the pure polycyclic aromatic hydrocarbon indene (c- C9h8) with GOthAM observations of tMC- 1. Astrophys. J. Lett. 913, l18 (2021). 5. M. l. Sita, P. B. Changala, C. Xue, A. M. Burkhardt, C. n. Shingledecker, K. l. K. lee, R. A. loomis, e. Momjian, M. A. Siebert, d. Gupta, e. herbst, A. J. Remijan, M. C. McCarthy, i. R. Cooke, B. A. McGuire, discovery of interstellar 2- cyanoindene (2- C9h7Cn) in GOthAM observations of tMC- 1. Astrophys. J. Lett. 938, l12 (2022). 6. A. G. G. M. tielens, interstellar polycyclic aromatic hydrocarbon molecules. Annu. Rev. Astron. Astrophys. 46, 289–337 (2008). 7. C. Joblin, G. Mulas, interstellar polycylic aromatic hydrocarbons: From space to the laboratory. Eur. Astron. Soc. Publ. Ser. 35, 133–152 (2009). 8. Z. d. levey, B. A. laws, S. P. Sundar, K. nauta, S. h. Kable, G. da Silva, J. F. Stanton, t. W. Schmidt, PAh growth in flames and space: Formation of the phenalenyl radical. J. Phys. Chem. A 126, 101–108 (2022). 9. B. A. McGuire, R. A. loomis, A. M. Burkhardt, K. l. K. lee, C. n. Shingledecker, S. B. Charnley, i. R. Cooke, M. A. Cordiner, e. herbst, S. Kalenskii, M. A. Siebert, e. R. Willis, C. Xue, A. J. Remijan, M. C. McCarthy, detection of two interstellar polycyclic aromatic hydrocarbons via spectral matched filtering. Science 371, 1265–1269 (2021). 10. e. Reizer, B. viskolcz, B. Fiser, Formation and growth mechanisms of polycyclic aromatic hydrocarbons: A minireview. Chemosphere 291, 132793 (2022). 11. X. Gu, R. i. Kaiser, Reaction dynamics of phenyl radicals in extreme environments: A crossed molecular beam study. Acc. Chem. Res. 42, 290–302 (2009). 12. R. i. Kaiser, d. S. n. Parker, F. Zhang, A. landera, v. v. Kislov, A. M. Mebel, Pah formation under single collision conditions: Reaction of phenyl radical and 1,3- butadiene to form 1,4- dihydronaphthalene. J. Phys. Chem. A 116, 4248–4258 (2012). 13. R. i. Kaiser, M. Goswami, P. Maksyutenko, F. Zhang, Y. S. Kim, A. landera, A. M. Mebel, A crossed molecular beams and abinitio study on the formation of C6h3 radicals. An Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024 Castiñeira Reis et al., Sci. Adv. 10, eadq4077 (2024) 11 September 2024 SCienCe AdvAnCeS | ReSeARCh ARtiCle 9 of 10 interface between resonantly stabilized and aromatic radicals. J. Phys. Chem. A 115, 10251–10258 (2011). 14. A. v. Kelleghan, A. S. Bulger, d. C. Witkowski, n. K. Garg, Strainpromoted reactions of 1,2,3- cyclohexatriene and its derivatives. Nature 618, 748–754 (2023). 15. J. A. Miller, S. J. Klippenstein, the recombination of propargyl radicals and other reactions on a C6h6 potential. J. Phys. Chem. A 107, 7783–7799 (2003). 16. M. Agúndez, C. Cabezas, B. tercero, n. Marcelino, J. d. Gallego, P. de vicente, J. Cernicharo, discovery of the propargyl radical (Ch2CCh) in tMC- 1: One of the most abundant radicals ever found and a key species for cyclization to benzene in cold dark clouds. Astron. Astrophys. 647, l10 (2021). 17. J. Cernicharo, M. Agúndez, C. Cabezas, n. Marcelino, B. tercero, J. R. Pardo, J. d. Gallego, F. tercero, J. A. lópez- Pérez, P. de vicente, discovery of Ch2ChCCh and detection of hCCn, hC4n, Ch3Ch2Cn, and, tentatively, Ch3Ch2CCh in tMC- 1. Astron. Astrophys. 647, l2 (2021). 18. A. M. Mebel, v. v. Kislov, R. i. Kaiser, Ab initio/Rice- Ramsperger- Kassel- Marcus study of the singlet C4h4 potential energy surface and of the reactions of C2 ( X1 Σ + g ) with C4h4 ( X 1 A + 1 g ) and C (1d) with C3h4 (allene and methylacetylene). J. Chem. Phys. 125, 133113 (2006). 19. C. Puzzarini, Gasphase chemistry in the interstellar medium: the role of laboratory astrochemistry. Front. Astron. Space Sci. 8, 811342 (2022). 20. M. h. Stockett, J. n. Bull, h. Cederquist, S. indrajith, M. Ji, J. e. navarro navarrete, h. t. Schmidt, h. Zettergren, B. Zhu, efficient stabilization of cyanonaphthalene by fast radiative cooling and implications for the resilience of small pahs in interstellar clouds. Nat. Commun. 14, 395 (2023). 21. M. Saito, h. Kubota, K. Yamasa, K. Suzuki, t. Majima, h. tsuchida, direct measurement of recurrent fluorescence emission from naphthalene ions. Phys. Rev. A 102, 012820 (2020). 22. e. Martíneznúñez, An automated method to find transition states using chemical dynamics simulations. J. Comput. Chem. 36, 222–234 (2015). 23. C. W. Gao, J. W. Allen, W. h. Green, R. h. West, Reaction mechanism generator: Automatic construction of chemical kinetic mechanisms. Comput. Phys. Commun. 203, 212–225 (2016). 24. P. l. Bhoorasigh, B. l. Slakman, F. S. Khanshan, J. Y. Cain, R. h. West, Automated transition state theory calculations for highthroughput kinetics. J. Phys. Chem. A 121, 6896–6904 (2017). 25. C. Cavallotti, M. Pelucchi, Y. Georgievskii, S. J. Klippenstein, eStoktP: electronic structure to temperature- and pressuredependent rate constants - A code for automatically predicting the thermal kinetics of reactions. J. Chem. Theory Comput. 15, 1122–1145 (2019). 26. i. ismail, h. B. v. A. Stuttaford- Fowler, C. Ochan Ashok, C. Robertson, S. habershon, Automatic proposal of multistep reaction mechanisms using a graphdriven search. J.Phys. Chem. A 123, 3407–3417 (2019). 27. R. van de vijver, J. Zádor, Kinbot: Automated stationary point search on potential energy surfaces. Comput. Phys. Commun. 248, 106947 (2020). 28. e. Martíneznúñez, G. l. Barnes, d. R. Glowacki, S. Kopec, d. Peláez, A. Rodríguez, R. Rodríguez- Fernández, R. J. Shannon, J. J. P. Stewart, P. G. tahoces, S. A. vazquez, AutoMeKin2021: An opensource program for automated reaction discovery. J. Comput. Chem. 42, 2036–2048 (2021). 29. C. Robertson, R. hyland, A. J. d. lacey, S. havens, S. habershon, identifying barrierless mechanisms for benzene formation in the interstellar medium using permutationally invariant reaction discovery. J. Chem. Theory Comput. 17, 2307–2322 (2021). 30. J. Zádor, C. Martí, R. van de vijver, S. l. Johansen, Y. Yang, h. A. Michelsen, h. n. najm, Automated reaction kinetics of gasphase organic species over multiwell potential energy surfaces. J. Phys. Chem. A 127, 565–588 (2023). 31. e. Martíneznúñez, An automated transition state search using classical trajectories initialized at multiple minima. Phys. Chem. Chem. Phys. 17, 14912–14921 (2015). 32. d. Garay- Ruiz, M. Álvarez Moreno, C. Bo, e. Martíneznúñez, new tools for taming complex reaction networks: the unimolecular decomposition of indole revisited. ACS Phys. Chem. Au 2, 225–236 (2022). 33. G. henkelman, h. Jónsson, improved tangent estimate in the nudged elastic band method for finding minimum energy paths and saddle points. J. Chem. Phys. 113, 9978–9985 (2000). 34. M. Castineira Reis, e. Martinez-nunez, A. Fernandez-Ramos, Comprehensive computational automated search of barrierless reactions leading to the formation of benzene and other C6-membered rings (2024). https://zenodo.org/records/11175641. 35. C. h. Wu, R. d. Kern, Shocktube study of allene pyrolysis. J. Phys. Chem. 91, 6291–6296 (1987). 36. J. A. Miller, C. F. Melius, Kinetic and thermodynamic issues in the formation of aromatic compounds in flames of aliphatic fuels. Combust. Flame 91, 21–39 (1992). 37. l. Zhao, W. lu, M. Ahmed, M. v. Zagidullin, v. n. Azyazov, A. n. Morozov, A. M. Mebel, R. i. Kaiser, Gasphase synthesis of benzene via the propargyl radical selfreaction. Sci. Adv. 7, eabf0360 (2021). 38. W. tang, R. S. tranter, K. Brezinsky, isomeric product distributions from the selfreaction of propargyl radicals. J. Phys. Chem. A 109, 6056–6065 (2005). 39. M. e. law, P. R. Westmoreland, t. A. Cool, J. Wang, n. hansen, C. A. taatjes, t. Kasper, Benzene precursors and formation routes in a stoichiometric cyclohexane flame. Proc. Combust. Inst. 31, 565–573 (2007). 40. e. h. Wilson, S. K. Atreya, A. Coustenis, Mechanisms for the formation of benzene in the atmosphere of titan. J. Geophys. Res. Planets 108, 5014 (2003). 41. S. Green, e. herbst, Metastable isomers: A new class of interstellar molecules. Astrophys. J. 229, 121 (1979). 42. A. Yokoyama, X. Zhao, e. J. hintsa, R. e. Continetti, Y. t. lee, Molecular beam studies of thephotodissociation of benzene at 193 and 248 nm. J. Chem. Phys. 92, 4222–4233 (1990). 43. J. M. hollis, P. R. Jewell, F. J. lovas, Confirmation of interstellar methylene. Astrophys. J. 438, 259 (1995). 44. h. Feuchtgruber, F. P. helmich, e. F. van dishoeck, C. M. Wright, detection of interstellar Ch3. Astrophys. J. 535, l111–l114 (2000). 45. K. d. tucker, M. l. Kutner, P. thaddeus, the ethynyl radical C2h–A new interstellar molecule. Astrophys. J. 193, l115–l119 (1974). 46. J. Cernicharo, C. A. Gottlieb, M. Guélin, t. C. Killian, G. Paubert, P. thaddeus, J. M. vrtilek, Astronomical detection of h2CCC. Astrophys. J. 368, l39 (1991). 47. J. Cernicharo, C. Cabezas, M. Agúndez, B. tercero, n. Marcelino, J. R. Pardo, F. tercero, J. d. Gallego, J. A. lópez- Pérez, P. de vicente, discovery of allenyl acetylene, h2CCChCCh, in tMC- 1. A study of the isomers of C5h4. Astron. Astrophys. 647, l3 (2021). 48. J. Cernicharo, M. Guelin, C. M. Walmsley, detection of the hyperfine structure of the C5h radical. Astron. Astrophys. 172, l5 (1987). 49. W. d. langer, t. velusamy, t. B. h. Kuiper, R. Peng, M. C. McCarthy, M. J. travers, A. Kovács, C. A. Gottlieb, P. thaddeus, First astronomical detection of the cumulene carbon chain molecule h2C6 in tMC- 1. Astrophys. J. 480, l63 (1997). 50. J. Cernicharo, C. A. Gottlieb, M. Guélin, t. C. Killian, G. Paubert, P. thaddeus, J. M. vrtilek, Astronomical detection of h2CCCC. Astrophys. J. 368, l43 (1991). 51. P. thaddeus, J. M. vrtilek, C. A. Gottlieb, P. thaddeus, J. M. vrtilek, C. A. Gottlieb, laboratory and astronomical identification of cyclopropenylidene, C3H2. Astrophys. J. 299, l63 (1985). 52. Y. Guo, X. Gu, F. Zhang, A. M. Mebel, R. i. Kaiser, A crossed molecular beam study on the formation of hexenediynyl radicals (h2CCCCCCh; C6h3 (X2A′)) via reactions of tricarbon molecules, C3 ( X1 Σ + g ), with allene (h2CCCh2; X1A1) and methylacetylene (Ch3CCh; X1A1). Phys. Chem. Chem. Phys. 9, 1972 (2007). 53. K. W. hinkle, J. J. Keady, P. F. Bernath, detection of C3 in the circumstellar shell of iRC+10216. Science 241, 1319–1322 (1988). 54. C. Cabezas, M. Agúndez, R. Fuentetaja, Y. endo, n. Marcelino, B. tercero, J. R. Pardo, P. de vicente, J. Cernicharo, discovery of the cyclic C5h radical in tMC- 1. Astron. Astrophys. 663, l2 (2022). 55. P. thaddeus, C. A. Gottlieb, A. hjalmarson, l. e. B. Johansson, W. M. irvine, P. Friberg, R. A. linke, Astronomical identification of the C3h radical. Astrophys. J., Lett. Ed. 294, l49–l53 (1985). 56. S. Yamamoto, S. Saito, M. Ohishi, h. Suzuki, S.- i. ishikawa, n. Kaifu, A. Murakami, laboratory and astronomical detection of the cyclic C3h radical. Astrophys. J. 322, l55 (1987). 57. J. G. Mangum, A. Wootten, Observations of the cyclic C3h radical in the interstellar medium. Astron. Astrophys. 239, 319 (1990). 58. R. Fuentetaja, M. Agúndez, C. Cabezas, B. tercero, n. Marcelino, J. R. Pardo, P. de vicente, J. Cernicharo, discovery of two new interstellar molecules with QUiJOte: hCCChCCC and hCCCCS. Astron. Astrophys. 667, l4 (2022). 59. J. J. P. Stewart, Optimization of parameters for semiempirical methods vi: More modifications to the nddO approximations and reoptimization of parameters. J. Mol. Model. 19, 1–32 (2013). 60. J. J. P. Stewart, “MOPAC2016,” 2016; http://openmopac.net/MOPAC2016.html; 10th May 2024. 61. Y. Zhao, d. G. truhlar, exploring the limit of accuracy of the global hybrid meta density functional for maingroup thermochemistry, kinetics, and noncovalent interactions. J.Chem. Theory Comput. 4, 1849–1868 (2008). 62. A. d. Becke, densityfunctional thermochemistry. iii. the role of exact exchange. J. Chem. Phys. 98, 5648–5652 (1993). 63. M. J. Frisch, G. W. trucks, h. B. Schlegel, G. e. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, v. Barone, G. A. Petersson, h. nakatsuji, X. li, M. Caricato, A. v. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, h. P. hratchian, J. v. Ortiz, A. F. izmaylov, J. l. Sonnenberg, d. Williams- Young, F. ding, F. lipparini, F. egidi, J. Goings, B. Peng, A. Petrone, t. henderson, d. Ranasinghe, v. G. Zakrzewski, J. Gao, n. Rega, G. Zheng, W. liang, M. hada, M. ehara, K. toyota, R. Fukuda, J. hasegawa, M. ishida, t. nakajima, Y. honda, O. Kitao, h. nakai, t. vreven, K. throssell, J. A. Montgomery, Jr., J. e. Peralta, F. Ogliaro, M. J. Bearpark, J. J. heyd, e. n. Brothers, K. n. Kudin, v. n. Staroverov, t. A. Keith, R. Kobayashi, J. normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. iyengar, J. tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. l. Martin, K. Morokuma, O. Farkas, J. B. Foresman, d. J. Fox, Gaussian 16 Revision B.01 (Gaussian inc., 2016). Downloaded from https://www.science.org at Universidade de Santiago de Compostela on December 19, 2024