Computational Evidence Suggests that 1 ... - ACS Publications

Subscriber access provided by Macquarie University

A: Spectroscopy, Molecular Structure, and Quantum Chemistry

Computational Evidence Suggests That 1-Chloroethanol May Be an Intermediate in the Thermal Decomposition of 2-Chloroethanol into Acetaldehyde and HCl Zoi Salta, Agnie Mylona Kosmas, Oscar N. Ventura, and Vincenzo Barone J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.8b11966 • Publication Date (Web): 22 Feb 2019 Downloaded from http://pubs.acs.org on February 26, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Computational Evidence Suggests that 1-Chloroethanol May be an Intermediate in the Thermal Decomposition of 2-Chloroethanol into Acetaldehyde and HCl

Zoi Salta1*, Agnie M. Kosmas2, Oscar N. Ventura3, and Vincenzo Barone1 1SMART

Lab, Scuola Normale Superiore di Pisa, Piazza dei Cavalieri, 7, 56126 Pisa, Italy Chemistry Sector, Department of Chemistry, University of Ioannina, Ioannina, Greece 3Computational Chemistry and Biology Group, CCBG, DETEMA, Facultad de Química, Udelar, Montevideo, Uruguay 2Physical

Abstract The dehalogenation of 2-chloroethanol (2ClEtOH) in the gas phase with and without the participation of catalytic water molecules has been investigated using methods rooted into the density functional theory. The well-known HCl elimination leading to vinyl alcohol (VA) was compared to the alternative elimination route towards oxirane and shown to be kinetically and thermodynamically more favorable. However, the isomerization of VA to acetaldehyde in the gas phase, in the absence of water, was shown to be kinetically and thermodynamically less favorable than the recombination of VA and HCl to form the isomeric 1-chloroethanol (1ClEtOH) species. At the B97X-D/cc-pVTZ level of calculation, this species is more stable than 2ClEtOH by about 6 kcal mol-1 at 298K, and the reaction barrier for VA to 1ClEtOH is 23 kcal mol-1 vs 55 kcal mol-1 for the direct transformation of VA to acetaldehyde. In a successive step, 1ClEtOH can decompose directly to acetaldehyde and HCl with a lower barrier (29 kcal mol-1) than that of VA to the same products (55 kcal mol-1). The calculations were repeated using a single ancillary water molecule (W) in the complexes 2ClEtOH_W and 1ClEtOH_W. The latter adduct is now more stable than 2ClEtOH_W by about 8 kcal mol-1 at 298K, implying that the water molecule increased the already higher stability of 1ClEtOH in the gas phase. However, this catalytic water molecule lowers dramatically the barrier for the interconversion of VA to acetaldehyde (from 55 to 7 kcal mol-1). This barrier is now smaller than the one for the conversion to 1ClEtOH (which also decreases, but not so much, from 23 to 13 kcal mol-1). Thus, it is concluded that while 1ClEtOH may be a plausible intermediate in the gas phase dehalogenation of 2ClEtOH, it is unlikely that it plays a major role in water complexes (or, by inference, aqueous solution). It is also shown that neither in the gas phase nor in the cluster with one water molecule, the oxirane path is more favorable than the VA alcohol path. Additionally, a direct conversion of 2ClEtOH to 1ClEtOH through a transition state which resembles a VA molecule in a complex with a chlorine atom and a hydrogen atom on both sides of this planar species was found. This reaction path has also lower activation energy than the conversion to oxirane, but not as low as the conversion to VA.

*Corresponding author Dr. Zoi Salta SMART Lab, Scuola Normale Superiore di Pisa, Piazza dei Cavalieri, 7, 56126 Pisa, Italy E-mail: [email protected]

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 40

1. Introduction Halogenated hydrocarbons are organic species of considerable interest in environmental chemistry. The role of chlorofluorocarbons (CFCs) in the depletion of the ozone layer1-3 made them a frequent subject of study in atmospheric chemistry. Adsorbable organic halogen (AOX) species produced during chlorine dioxide bleaching of pulp in the Kraft process4,5 as well as the byproducts of water disinfection6-8 contribute to surface water contamination. Many other products used in the industry, like degreasing fluids, and agriculture, like pesticides, further contribute to the presence of halogenated hydrocarbons in the environment. One of the simplest halogenated hydrocarbon families is that of halohydrins (halogenated alcohols) and one of their important reactions is dehalogenation, i.e. HX elimination, which can occur by pyrolysis in gas phase912, or by base catalysis13-15 or oxidation16 in solution. Thermal decomposition of halohydrins has been investigated repeatedly. In particular, single pulse shock wave experiments on the thermal decomposition of fluoroethanol17 as well as chloro- and bromoethanol10 have been performed. Thermal elimination of HCl in pyrolysis occurs through the intermediary vinyl alcohol, which is assumed to be unstable and to evolve later to acetaldehyde10. However, the tautomerization process has a high barrier (about 56 kcal mol-1 experimentally) and it can be kept up to half an hour in a Pyrex flask without decomposition18. At the same time, the reaction occurs easily in the presence of catalytic agents (for instance inorganic19 or carboxylic20 acids, or in concentrated conditions where dimerization can proceed). In clusters, like those present in atmospheric aerosols, water molecules can act as general catalysts to lower the barrier for the conversion of vinyl alcohol to acetaldehyde, as shown by some of the present authors several years ago21, and confirmed more recently by other authors22. In aqueous solution finally, either base catalysis or general base catalysis by water molecules may eliminate HCl from halohydrins by a two-step nucleophilic attack, producing the oxirane molecule, which later evolves towards the diol13,15. It has been shown experimentally that either oxiranes or diols can be the final products depending on the branching of the hydrocarbon lateral chains of the chlorohydrins15. This reaction channel seems to be unfeasible in the case of thermal decomposition or in cluster reactions. The purpose of this work was to study the dehalogenation mechanism of 2-chloroethanol (2ClEtOH) in the gas phase (complete absence of water) and in a complex where only one ancillary water molecule is present. The goal was to assess carefully the interplay between the reaction channels leading to oxirane and to vinyl alcohol in the absence of water and in the presence of only one catalytic water molecule. In particular, we wanted to assess whether acetaldehyde is directly obtained from vinyl alcohol when water molecules are absent. The results obtained in this study suggest that a rarely studied isomer of 2-chloroethanol, 1-chloroethanol (1ClEtOH), may play a role in the interconversion because the Markovnikoff addition of HCl to VA to give 1ClEtOH has a much lower barrier than the direct conversion of VA to acetaldehyde in the absence of catalysts. We found that dehalogenation of 1ClEtOH leads directly to acetaldehyde by abstraction of the oxygen on the hydroxyl atom, contrary to the dehalogenation of 2ClEtOH, where the hydrogen atom involved in the HCl product is attached initially to a carbon atom. The presence of just a single water molecule produces a nine-fold decrease of the barrier for the conversion of VA to acetaldehyde, making this route more favorable than that going through 1ClEtOH, even if the latter is also accelerated by the presence of water.


2

Page 3 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2. Chemical models and methods The general reaction scheme studied is shown in Fig. 1. The opening of oxirane to give ethylene glycol was studied only in the presence of the ancillary water molecule in the chloroethanol complexes since it is essential to form the dihydroxylated EGly species. Labeling of the transition states will be shown in the scheme by using a W to indicate the corresponding transition state when the ancillary water molecule is present. The path from 1ClEtOH to 2ClEtOH through carbene was not investigated in detail because of the very high energy barrier associated with TS3. The reverse path from ethylene glycol plus HCl to 2ClEtOH plus water was not investigated either since, both sterically and electronically, the transition state for such a reaction would be very unfavorable.

INSERT FIGURE 1

Calculations were performed using density functional theory (DFT)23. The method chosen was B97XD24,25, both because of its general accuracy and because it includes long-range and dispersion corrections, which are important for the accurate calculation of loose transition states like those considered in this work. Two basis sets were used. Exploratory work was performed with Pople’s 6-31G(d) basis set26. The structures optimized at this level were further re-optimized using the more extended Dunning’s cc-pVTZ basis set27,28. It is known that, in general, DFT methods are not so sensitive to the choice of the basis set as molecular orbitals methods are. However, there is a certain effect, especially for transition states, and a reasonably large basis set is needed to obtain accurate values. We have shown in other studies (see for instance references 29 and 30) that a basis of the size and composition of the cc-pVTZ is enough to produce reasonable results. Moreover, contrary for instance to post-Hartree-Fock, DFT methods do not exhibit a monotonic behavior with the increase of the basis sets. For example, using basis sets as extended as aug-cc-pV6Z produces often worse results than cc-pVTZ (depending on the problem and the exchange-correlation functional used, see reference 29). We included some more information about this subject in the Additional Information section. Optimizations were performed until all cartesian coordinates were converged to at least 10-4 Å. Analytical second derivatives were calculated for all species, both to obtain the IR spectra as well as for checking the correct number of negative eigenvalues. A Natural Bond Orbital (NBO) analysis31 was performed and NBO charges, as well as charges derived from the electrostatic potential (using the Hu, Lu, and Yang charge-fitting method32) were calculated. The Laplacian of the density was calculated for all complexes and transition states to assess the characteristics of chemical bonds33. Electrostatic potential surfaces were also calculated and displayed in some cases to discuss specific bonding situations. All calculations were performed using the G09 set of computer codes34.


3


Page 4 of 40

3. Results and discussion 3.1. Structures The structure and geometries calculated at the B97X-D/cc-pVTZ level for the isolated species are shown in Fig. 2. In a few cases the comparison with experiment or highly accurate geometrical data is made, to assess the correctness of the method employed here. The same information for the 1:1 complexes with water is displayed in Fig. 3. Important changes in internal coordinates with respect to those in the isolated species are depicted.

INSERT FIGURE 2 INSERT FIGURE 3

Obviously, the most important species is 2ClEtOH, whose geometrical structure has been the subject of two experimental studies. Azrak and Wilson35 analyzed the microwave (MW) spectra of 2ClEtOH and its bromine analog, while Almemmingen et al36 performed more recently an electron diffraction (ED) study with the aim of determining the anti-gauche ratio as a function of temperature. The agreement between both studies is reasonable concerning the C-C, C-O and O-H bond distances, as well as the dihedral angle between the ClCC and CCO planes, but is much worse for the C-Cl distance: 1.7886 ± 0.0038 Å versus 1.801 ± 0.001 Å. Although the discrepancy is not very large, it is significant. Durig et al37 examined these experimental values and concluded that one must assume that the ED data is more accurate, and, since chlorine is the heaviest atom in the molecule, that the C-Cl distance should be the best determined structural parameter in this study. Thus, on the basis of their arguments, we conclude that our theoretical values are in reasonable agreement with experiment. From the theoretical point of view, we did not find in the literature high-level calculations (i.e. CCSD(T) or better) of the geometry of 2ClEtOH. The published calculations used HF, B3LYP or MP2 with at most 6-311++G(d,p) or aug-cc-pVTZ basis sets. We provide here our own B97X-D/cc-pVTZ calculations used throughout this work and, for comparison purposes, MP2/aug-cc-pVTZ and B97X-D/aug-cc-pVQZ calculations performed for another study on microhydration of chloroethanols30. In all cases, the calculations provided the gauche Gg’ structure37 as the most stable. The decomposition analysis of the interactions was performed by Baranac-Stojanović et al38, who determined that the effect is caused mainly by electrostatic attraction; we arrived at the same conclusion in reference 30. The dihedral angle between the ClCC and CCO planes is well reproduced by all the calculations. DFT results for the CCl bond are nearer to the ED data than MP2 results. The MP2 value is nearer to the MW experimental result than to that obtained by ED. Two aspects in which calculations fail are the OH bond length, which is theoretically predicted much shorter than the experimental value, and the hydrogen bond Cl-H, where the computed value is longer than its experimental counterpart. In this specific case, however, MP2 performs better than DFT. The discrepancy with respect to the OH bond length can be ascribed to several effects which, in order of importance would be (a) thermal effects, since experimental values are r0 while the theoretical ones are re, (b) anharmonicity effects, which are most noticeable in bonds involving hydrogen because of its small mass


4

Page 5 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and which have not been considered in this paper, and (c) the fact that DFT methods have more difficulty in describing the small electronic density of the H atom than that of heavier atoms. Infrared studies of 2ClEtOH were performed experimentally several times in different conditions. The spectrum was determined in an Ar-matrix by Perttilä et al39, studied in CCl4 by Chitale et al40 and the far IR spectrum was recorded in the gas phase by Durig et al37. The spectrum was also studied by Skvortsov and Vilesov41 who used He droplets for the measurements of interconversion enthalpy of 2EtOH conformers. The O-H torsional fundamental was reported at 344.0 cm-1 in gas phase37 and at 325 cm-1 in solid Ar39. Both experimentally and in our own calculation (where it falls at 371.8 cm-1, see Fig. 4a) this is the most intense band in the whole spectrum. The C-Cl stretching band is found at 667 cm-1 in Ar while in our calculations it appears at 685 cm-1. Finally, the OH stretching band for the gauche isomer appears at 3614 cm-1 in Ar, 3624.5 cm-1 in He and at 3877 cm-1 in our calculations, again pointing out a significant effect of anharmonicity on this bond. The Raman lines were also recorded experimentally in liquid phase39 and the results of our calculations (see Fig. 4b) agree with those.

INSERT FIGURE 4

Another point to be considered is whether there is a real hydrogen bond or a purely electrostatic attraction between Cl and H atoms. To that end, we present the Laplacian of the electronic density both in the plane of the three atoms Cl-H-O (which cannot be co-linear), Fig. 4c, and in a perpendicular plane to this one, midway the perpendicular line from H to the Cl-O line, Fig. 4d. Both images show clearly that there is no discernible sharing of electronic density, meaning that the interaction between Cl and H is mostly electrostatic and/or hyperconjugative42. Atomic charges obtained from the NBO analysis were 0.46 for H and -0.11 for Cl, while using the Hu, Lu, Yang method of electrostatic potential fitting, they are similar but not identical, 0.38 for H and -0.12 for Cl. The electrostatic attraction is thus obvious, as well as the absence of a hydrogen bond. Nonetheless, observe that this attraction produces a deep consequence. Since the OH and Cl groups are syn to each other, the formation of the oxirane is disfavored because the posterior attack of the oxygen on the Cl-bearing carbon would be less probable. We will discuss this point later on. 2ClEtOH can be present as five different rotational conformers, which have been studied several times3647. We present in Table 1 our own results compared to those published so far and obtained at different experimental and theoretical levels.

INSERT TABLE 1

Our results are in good agreement with the available experimental data36,37,47,48 and previous theoretical calculations41-46,48 using other methods and basis sets. The Gg’ structure is the most stable one, while the Gg structure is the least stable. Tg is the second most stable structure when only energetic effects are included, but


5


Page 6 of 40

the free energy is almost identical to Tt. MP2 predicts Tt as more stable than Tg45,46, while our calculations predict the opposite, in agreement with the CCSD/cc-pVSZ calculations41. The abundances determined with our method of calculation agree well with the available experimental data36 as well as with the MP2 calculation with the more extended basis set45. We believe then that we can use our theoretically derived abundances to obtain the combined IR spectra. The independent IR spectra of the five conformers have been superimposed and shown in Fig. 6. It is clear from this figure that the most affected band is the HOC torsion plus OH stretching combination band appearing in the far infrared below 500 cm-1. In our B97X-D/cc-pVTZ calculations, the band appears at 376 cm-1 for the Gg’ rotamer, while experimentally it was recorded at 344.0 cm-1 by Durig et al37 and 344.2 cm-1 by Soliday et al48. The analogous bands in Tg and Tt were obtained by us at 326 cm-1 and 290 cm-1 respectively, while Durig et al37 reported them experimentally at 281 and 200 cm-1 respectively. The discrepancy between our values and the experimental ones is probably due to anharmonic effects. This hypothesis is supported by the fact that Soliday et al48 at the MP2/aug-cc-pVTZ level and including anharmonicity obtained a value of 340.9 cm-1, much nearer to the experimental one. The theoretical intensities of each frequency in the spectra for each of the rotamers was multiplied by the abundance factor determined from our free energy data (see Table 1) and added to obtain the spectrum shown in Fig. 7 superimposed to the spectrum of the Gg’ conformer. It is immediately obvious that at 298K there would be no difference between both spectra. Almmenningen et al36 reported that the abundance of Tt and Tg conformers increase with the temperature, but the concentration is significant only over 500K. Therefore, we believe that in most experimental situations there will be no need to consider more than the Gg’ isomer. Contrary to 2ClEtOH, there is no experimental information on the structure of 1ClEtOH. The B97X-D/ccpVTZ calculations predict a longer CCl bond and a shorter OH bond, with the ClCOH in a gauche conformation. The distance between Cl and H is significantly larger than in 2ClEtOH. All these elements put together point toward a more favorable disposition for obtaining vinyl alcohol in 1ClEtOH than in 2ClEtOH, while the opposite would be true for obtaining oxirane. The IR spectrum of 1ClEtOH presents similarities with that of 2ClEtOH, but also differences which would allow the experimental differentiation of both species in a reaction mixture. Instead of presenting separately the spectrum of 1ClEtOH, Fig. 5 shows both spectra superimposed. It is simple to see that the relation between the CCl and OH stretching bands do give away the isomer. A ratio vCCl/vOH larger than one indicates the presence of 1ClEtOH, while a ratio smaller than one indicates the presence of 2ClEtOH.

INSERT FIGURE 5 INSERT FIGURE 6 INSERT FIGURE 7


6

Page 7 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Vinyl alcohol (VA) is a molecule that has also been studied several times and which, besides its interest in multiple industrial reaction paths, is also of considerable interest as an interstellar species. The most recent publications on this molecule, to our knowledge, are those of Bunn et al49 who studied the far IR spectra of both anti- and syn-VA using a CCSD(T)/cc-pVQZ chemical model to obtain the structure of both species, an experimental far-infrared spectroscopic characterization of anti-VA, also by Bunn and coworkers50 using CCSD(T)/cc-pVTZ calculations on anti-VA, and a paper from Schaefer’s group51 using methods as demanding as CCSDT(Q)/cc-pV5Z. Experimental microwave spectra of vinyl alcohol were reported by Saito18 for the syn- conformer and by Rodler52 for the anti- conformer. Previous molecular orbital theoretical determinations49-51 agree well with the experimental values, while our own calculations afford C-O distances slightly too short (i.e the C-O bond is slightly too strong). The overall discrepancy, however, is not very significant. Probably just by chance, the experimental value of the CC double bond in anti-VA was reproduced better by our calculations than by the CCSD(T) ones. Meanwhile, the best experimental and theoretical determination of the structure of oxirane can be found in a paper by Demaison et al53. Our B97X-D /cc-pVTZ calculations afford results for the CC and CH bonds very near the experimental ones, but a too short CO bond which, on the contrary, is well predicted at the highest level of theory used by Demaison et al53. This implies that in the present B97X-D /cc-pVTZ calculations the oxygen atom is slightly too strongly bound to the carbon atoms and thus the cycle will be more difficult to open by the water molecule (the same behavior as that observed for VA). The ethylene glycol molecule is another simple species that has attracted considerable attention because of its presence in multiple conformers (ten, taking into account symmetries), in some of which, weak hydrogen bonds are present. Bastiansen54 was perhaps the first to describe correctly the gauche form. Although he assumed standard CC, CO and CH bond distances, he was able to determine without doubt that the species present was the gauche structure (known as g’Ga or g’Gt, referring to the gauche conformation of both the Hs and Os and the anti or syn disposition of the OHs) with an OO distance of 2.97 Å and a dihedral angle of 74 deg between the O1C1C2 and C1C2O2 planes. The microwave spectrum of ethylene glycol has been studied repeatedly for more than 40 years, starting by the microwave experiment of Marsokk and Møllendal55 and up to modern times with the experiments of Christen et al56,57. The barriers for internal rotation around the C-C and C-O bonds were determined to be about 10 kcal mol-1 and 3-4 kcal mol-1 in early experiments58 suggesting that the g’Ga is fairly rigid due to (weak) hydrogen bonding between the OH groups (see, for instance, ref. 59). In all these papers it has been assumed that the bonding between the hydroxyl groups is actually a hydrogen bond, but there has been considerable discussion about whether this bond exists or not59-64. The accepted situation nowadays is that there is not such a bond63,64. Our own calculations show the absence of a bond critical point (BCP) in the same way as in the case of 2ClEtOH (see Fig. 8), supporting this view.

INSERT FIGURE 8

Many papers67-79 have been published that quote the original microwave experiment of Bastiansen54. However, we were unable to find any precise determination of the experimental geometry in the gas phase (a “plausible” r0 structure was derived from a microwave study by Caminati and Corbelli80). Some information is


7


Page 8 of 40

available on the ethylene glycol in the solid phase64,68, both with and without water. Of course, in this case the structure of the lattice influences the geometrical parameters and they cannot be compared to the theoretical ones. The most approximate experimental data (although obtained with some help of ab initio calculations) seem to be those by Kazerouni et al71, included in the figure as experimental results, although in the case of the CC bond the value differs significantly from that of Caminati and Corbelli80. With respect to the IR spectrum, the bound OH stretch is found at 3874 cm-1 (observed harmonic59 3803 cm-1), while the free one is found at 3931 cm-1 (observed harmonic59 3856 cm-1). The red shift we calculated, 57 cm-1, is significantly larger than the one found experimentally (33 cm-1 in vapor58, 33 cm-1 in Ar65 and 26 cm-1 in CO2 65). However, this can easily be explained by the lack of anharmonicity in our calculations, since the difference for the observed harmonic frequencies59 is 53 cm-1. In fact, Howard and Kjaergaard66 performed QCISD/6311++G(2d,2p) calculations adding anharmonic corrections and found a red shift of 39 cm-1 while in a previous paper, using CCSD(T)/aug’-cc-pVQZ calculations (aug-cc-pVQZ only in the hydroxyl groups), they found a harmonic red shift of 53 cm-1. Thus, one can conclude that most of the discrepancy in the red shift is due to anharmonic corrections which would not play an important role in the reactivity in which we are interested. Finally, the geometry of acetaldehyde was determined experimentally by Iijima and Kimura82 and the most recent theoretical determination (at the CCSD(T)/cc-pVQZ level) was performed by Chong83. We calculated neither the structure of carbene nor those of the TS3 and TS5 transition states (see Fig. 1) because of the very small possibility of this reaction path ever occurring. Also included in Fig. 1 are the final complexes between the given product and HCl. Although these complexes will not be present in any appreciable amount at high temperatures under which thermal dehalogenation occurs, we found it interesting to report them here also, in order to evaluate the influence of the HCl species bound to the respective product. Both for these complexes and the transition states, geometries given are those obtained with the B97X-D /cc-pVTZ calculations. We have listed also the difference between the geometrical parameters of each TS with respect to the reactants and the products it connects, in order to estimate whether the TS is more reactant- or product-like. The presence of a water molecule results in several types of hydrogen-bonded complexes. Although in general, several water complexes are possible for each species, we chose to report only the most stable ones. A more complete discussion of the microhydration of both 1ClEtOH and 2ClEtOH, with up to four water molecules, can be found in our previous work30. All complexes involving water in Fig.3, as well as the diol-HCl complex in Fig. 2 (where the water is already incorporated into the structure) exhibit also HCl hydrogen bonded to the water molecule. Except for the case of the oxirane-water-HCl complex, the HCl residue is always interacting with the main species. The most interesting situation arises with respect to the two isomers, syn- and anti-, of VA. The hydrogen bonded water in the latter isomer is bound to the hydroxyl group and interacts with one of the hydrogen atoms of the CH2 group. Attempts to modify the structure so that the free H of the water molecule interacts with the chlorine atom failed. We think that this is due to the greater stability of the planar structure of VA, which must be distorted in order to obtain such a complex. In the syn- isomer instead, the water molecule acts as a hydrogen acceptor from the hydroxyl group and it can bind as a hydrogen donor to the chlorine atom. This structure is very


8

Page 9 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


important because it provides a hydrogen migration chain that would lead directly from syn-VA to acetaldehyde. This is very clear in the structure of the transition state TS4.W in Fig. 3.

3.2 Energetics The general profile of the reactions, starting from ethylene and hypochlorous acid, through 2ClEtOH and 1ClEtOH, and ending in acetaldehyde and HCl in the absence of water, is shown in Fig. 9. Electronic energies with zero-point corrections are displayed relatively to 1ClEtOH. The results obtained with both the small and large basis sets are shown to assess the role of basis set extension. The same information is collected in Fig. 10 for the case of the addition of a single water molecule. For purposes of comparison, both energy profiles (with and without water) are shown, using only the larger basis set. The total energies and thermochemical data are given in the Supplementary Information. Relative energies are shown in Table 2.

INSERT TABLE 2 INSERT FIGURE 9

The data in Fig. 9 show that the effect of basis set extension is non-negligible, but does not change the qualitative aspects of the profiles. Therefore, we can simply discuss the most accurate ones. The first feature of the scheme is the reaction of hypochlorous acid with ethylene to give 2ClEtOH. The reaction is not spontaneous (∆𝑟𝐺𝑜298=58.0 kcal mol-1) and has a large barrier (Ea=49.6 kcal mol-1). Once formed, 2ClEtOH can evolve toward VA, oxirane or 1ClEtOH (by simultaneous H and Cl transfers). The transition states in the three cases are very high (barriers of 57.8, 63.5 and 61.5 kcal mol-1 respectively, using the larger basis set), but there is a slight preference for the evolution toward the anti-VA. Even if it can be argued that the small difference could allow oxirane to be formed to some extent, the reaction is much more endothermic than the conversion to anti-VA (27.4 vs 13.0 kcal mol-1). It could then be concluded that even if oxirane were to be formed, the equilibrium would shift back to antiVA. While this paper was under revision, a just-accepted paper by Soliday et al48 reported an experimental and theoretical study of 2ClEtOH. In this context, the authors also calculated the conversion of 2ClEtOH to oxirane. They used the MP2 method and the aug-cc-pVTZ basis set to obtain a barrier of 68.0 kcal mol-1 (against our 63.5 kcal mol-1 at the DFT level) and an endothermicity of 28.0 kcal mol-1 (against our own 27.4 kcal mol-1) in agreement with the results in this study. On the other side, the direct conversion of 2ClEtOH to 1ClEtOH through a simultaneous H and Cl transfer (see the structure of the transition state TS11) is exergonic but the barrier is slightly larger than that for the conversion to anti-VA. The difference is nonetheless small, 3.7 kcal mol-1. Notice that in the transition state corresponding to this isomerization, TS11, the general appearance is that of anti-VA interacting with a chlorine atom on one face of the planar molecule, and with a hydrogen atom on the other. The rotational barrier for the syn-anti conformers of vinyl alcohol is very small (3.3 kcal mol-1) and easily overcome. Two reaction channels are then open. On one side, it is the well-known isomerization of VA to acetaldehyde, which implies a slightly smaller barrier than for the conversion of 2ClEtOH to vinyl alcohol (55.2 kcal


9


Page 10 of 40

mol-1 vs 57.8 kcal mol-1). On the other side, HCl and VA could react exothermically to give 1ClEtOH, with a much smaller barrier of 23.3 kcal mol-1. The reaction of 1ClEtOH toward acetaldehyde and HCl is now direct, implying the abstraction of the hydrogen atom from the hydroxyl group with a slightly larger barrier (29.0 kcal mol-1) although the reaction is endothermic by 6.4 kcal mol-1. The full reaction then, from 2ClEtOH to acetaldehyde plus HCl would be very slightly endothermic (less than 1 kcal mol-1) and probably better described as thermoneutral. It must be noticed that the reaction of VA with HCl to give 1ClEtOH is exothermic (-17.3 kcal mol-1) and exergonic (-6.9 kcal mol-1). The reverse reaction is unlikely because the reverse barrier associated with TS2 (+39.3 kcal mol1) is substantially larger than the barrier for the decomposition to acetaldehyde and HCl (+29.0 kcal mol-1). The endothermicity of both reaction paths also favors the formation of acetaldehyde and HCl (+6.4 kcal mol-1) instead of VA + HCl (+16.0 kcal mol-1). Moreover, the reaction from 1ClEtOH to acetaldehyde plus HCl is exergonic (-4.5 kcal mol-1). There are two questions in relation to the suggestion of 1ClEtOH being an intermediate in the reaction path. On one side, one should see whether this channel is really plausible. It is clear from the height of the barrier that the process is feasible. A comparison of the structure of TS7 and TS2 transition states with that of the VA.HCl complex (see Fig. 3) shows clearly that the addition of Cl on the carbon atom carrying the hydroxyl group is the favored process, following Markovnikoff’s rule. However, it is clear that if in the reaction media there is enough energy to overcome the barrier associated with TS7, then there is a very slight probability that the charge transfer complex is having more than a fleeting existence. In fact, as can be seen from the free energy of the reaction, the entropy at 298 K is already sufficient to overcome the negative enthalpy and to make the reaction endergonic. It is more probable that the relatively long lifetime of VA allows further collisions with free HCl, collisions that will lead to 1ClEtOH as the favored product. The conclusion of the study of this reaction channel is that it is possible that 1ClEtOH (a much less studied molecule than the 2ClEtOH isomer) is formed in the decomposition of the latter, as an intermediate in the route to acetaldehyde. Some experimental observations in mixtures of VA and HCl seem to point to this explanation because HCl catalyzes the fast decomposition of VA. It does not seem plausible that HCl participates as a catalyst of the VA  acetaldehyde reaction. Instead, the formation of 1ClEtOH and further decomposition is more likely. We suggest that it may be possible to identify it as a transient in the reaction mixture, using infrared spectroscopy for instance. Experimental and theoretical work in this direction is underway. It must also be taken into account the fact that the direct conversion of 2ClEtOH to 1ClEtOH requires a transition state with a similar barrier; thus the possibility to obtain 1ClEtOH becomes more plausible. The comparison of the mechanisms with and without water is reported in Fig. 10. Full lines show the energies (E+ZPE) relative to 1ClEtOH, while dashed lines show the energies (E+ZPE) with respect to 1ClEtOH.W when water is present.

INSERT FIGURE 10


10

Page 11 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The first important difference with the reaction in the absence of water is the possibility of forming ethylene glycol (EGly) from oxirane by incorporating the hydrogen bound water molecule, a channel closed in the absence of water. The transition state leading from 2ClEtOH to oxirane (TS8) remains unchanged in the presence of the water molecule. The barrier decreases by just 3% and stays above the barrier for the transformation of 2ClEtOH to VA. If the oxirane-water complex would be formed, overcoming the TS8.W transition state (a barrier of 61.9 kcal mol-1), then it could surmount the TS9 transition state (with a smaller barrier of 51.4 kcal mol-1) to give EGly, which through the transition state TS10, with an even lower barrier (44.7 kcal mol-1) would give anti-VA and continue through 1ClEtOH to acetaldehyde. The reverse barrier, i.e. that for the reaction HCl + oxirane → 2ClEtOH + H2O, was calculated simultaneously with this work by Soliday et al48 who found a value of 38 kcal mol-1 at the MP2/cc-pVTZ level, in agreement with our own value of 37.2 kcal mol-1. The barrier for the transformation of 2ClEtOH to VA instead decreases more. When water is present, the barrier falls from 57.8 kcal mol-1 to 48.7 kcal mol-1 (16% reduction), about 13 kcal mol-1 lower than the barrier for the transformation to oxirane. The profile is then uniformly lowered with respect to the reaction without water. For instance, the barrier for the transformation from VA to 1ClEtOH decreases from 23.3 to 13.0 kcal mol-1 while the barrier for the transformation of 1ClEtOH to acetaldehyde is reduced from 29.0 to 12.0 kcal mol-1. Those important reductions are however dwarfed by the decrease in the barrier for the direct transformation of VA to acetaldehyde. In fact, the presence of the water molecule causes a decrease of the barrier from 55.2 to 7.1 kcal mol-1. The direct reaction of VA to acetaldehyde is now about twice more favorable than the transformation to 1ClEtOH. Finally, it must be noted that there is also the possibility that 2ClEtOH.W transforms directly into EGly.HCl when the water molecule is present. This barrier, corresponding to the transition state TS12, is however 15% larger than the barrier for the transformation of 2ClEtOH to anti-VA (57.1 vs 48.7 kcal mol-1). The reverse barrier, i.e. that for the reaction HCl + HOCH2CH2OH → ClCH2CH2OH + H2O, was calculated simultaneously to this work by Soliday et al48 who found a value of 52 kcal mol-1 at the MP2/cc-pVTZ level, in agreement with our own value of 53.4 kcal mol-1.

Conclusions We have performed a systematic study of the dehalogenation of 2ClEtOH using the B97X-D/cc-pVTZ chemical model. Geometry optimization of the different species along the reaction paths from 2ClEtOH to several products have been determined both for the isolated molecules and attached to one water molecule, in an effort to understand reaction energetics in water clusters present in atmospheric aerosols. There are three main conclusions in this work. In the first place, the barrier for the dehalogenation towards vinyl alcohol is about 6 kcal mol-1 smaller than the barrier for the reaction to oxirane. This occurs in the absence of water, but also when a single water molecule is present. In the second place, in the absence of water molecules, the path leading to acetaldehyde from a previous reaction of vinyl alcohol to 1ClEtOH by collision with an HCl molecule seems to be more favorable than the direct transformation from vinyl alcohol to acetaldehyde. The reaction barrier for the direct transformation amounts to 55.0 kcal mol-1, while the barrier for the reaction from vinyl alcohol to 1ClEtOH is only 22.1 kcal mol-1. Finally, if a water molecule is present, the barrier for the direct transformation is greatly reduced, from 55.0 kcal mol-1 to 5.8 kcal mol-1 while the barrier for the reaction


11


Page 12 of 40

towards 1ClEtOH is reduced to 11.9 kcal mol-1. The direct reaction becomes then more favorable than the path going through 1ClEtOH, and this species should not be observed in any significant concentration. As a global conclusion, it seems reasonable to think that in the absence of water the route through the direct transformation of VA to acetaldehyde is less probable, favoring the appearance of 1ClEtOH instead. The catalytic effect of one single water molecule is sufficient to lower the probability of this route, favoring the direct conversion of VA to acetaldehyde instead. Even if 2ClEtOH can react directly either to oxirane, 1ClEtOH, or ethylene glycol, both in the absence and presence of an ancillary water molecule, the reaction toward anti-VA exhibits the lowest barrier.

Acknowledgments ONV thanks the permanent support of ANII, CSIC (UDELAR) and Pedeciba to his research. Some of the calculations reported in this paper were performed in ClusterUY, a newly installed platform for high-performance scientific computing at the National Supercomputing Center, Uruguay.


12

Page 13 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References (1) Crutzen, P. J. My Life with O3, NOx, and Other YZOx Compounds (Nobel Lecture). Angew. Chem., Int. Ed. Engl. 1996, 35, 1758-1777. (2) Molina, M. J. Polar Ozone Depletion (Nobel Lecture). Angew. Chem., Int. Ed. Engl. 1996, 35, 1778-1785. (3) Rowland, F. S. Stratospheric Ozone Depletion by Chlorofluorocarbons (Nobel Lecture). Angew. Chem., Int. Ed. Engl. 1996, 35, 1786-1798. (4) Yao, S.; Gao, C.; Wang, C.; Shi, L.; Qin, C.; Wang, S. Kinetics of Adsorbable Organic Halogen Formation During Chlorine Dioxide Bleaching of Eucalyptus Kraft Pulp. J. Biobased Mater. Bioen. 2018, 12, 218-222. (5) Xie, Y-W; Chen, L.-J.; Liu, R.; Tian J.-P. AOX Contamination in Hangzhou Bay, China: Levels, Distribution and Point Sources. Environ. Pollution 2018, 235, 462-469. (6) Beauchamp, N.; Dorea, C.; Bouchard, Ch.; Rodriguez, M. Use of Differential Absorbance to Estimate Concentrations of Chlorinated Disinfection By-product in Drinking Water: Critical Review and Research Needs. Critical Revs. Environ. Sci. Technol. 2018, 48, 210-241. (7) Li, X-F; Mitch, W. A. Drinking Water Disinfection Byproducts (DBPs) and Human Health Effects: Multidisciplinary Challenges and Opportunities. Environ. Sci. Technol. 2018, 52, 1681-1689. (8) Richardson, S. D.; Thruston, Jr., A. D.; Collette E. W.; Schenck-Patterson, K.; Lyklns, Jr., B. W; Majetich, G.; Zhang, Y. Multispectral Identification of Chlorine Dioxide Disinfection Byproducts in Drinking Water. Environ. Sci. Technol. 1994, 28, 592-599. (9) Skingle, D. C.; Stimson, V. R. The Thermal Decomposition of 2-Chloroethanol. Aust. J. Chem. 1976, 29, 609615. (10) Chakravarty, H. K. Thermal Decomposition of Haloethanols and Ignition of JP-10. Ph.D. Dissertation, Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore, India, 2011. (11) Osman, O. I.; Elroby, S. A. K.; Hilai, R. H.; Aziz, S. G. Theoretical Characterization of Gas-phase Thermolysis Products of Ethane-1, 2-diol, 2-Chloroethanol and 2-Fluoroethanol. Mol. Phys. 2013, 111, 643-659. (12) Joo, D. L.; Merer, A. J.; Clouthier, D. J. High-Resolution Fourier Transform Infrared Spectroscopy of Vinyl Alcohol: Rotational Analysis of the nu(13) CH(2) Wagging Fundamental at 817 cm-1. J. Mol. Spectrosc. 1999, 197, 68-75. (13) Ketola, M.; Lyytinen, M.-R.; Hotokka, M.; Pihlaja, K. Kinetics and Mechanisms of Dehydrochlorination Reactions of Simple 1,3-Chlorohydrins in Aqueous Sodium Hydroxide Solutions. Acta Chem. Scandinav. 1978, B32, 743-748. (14) Phillips, D. L.; Zhao, C.; Wang, D. A Theoretical Study of the Mechanism of the Water-catalyzed HCl Elimination Reactions of CHXCl(OH)(X= H, Cl) and HClCO in the Gas Phase and in Aqueous Solution. J. Phys. Chem. A 2005, 109, 9653-9673.


13


Page 14 of 40

(15) Pihlaja, K.; Kiuru, M.-L., Sippola, A. Dehydrochlorination of 2-Chloroethanol, 2-Chloro-1-propanol, 1-Chloro2-propanol, 2-Chloro-2-methyl-1-propanol and 1-Chloro-2-methyl-2-propanol. Arkivoc 2012, 5, 120-133. (16) Gounden, A. N.; Jonnalagadda, S. B. Ozone Facilitated Dechlorination of 2-Chloroethanol and Impact of Organic Solvents and Activated Charcoal. Environ. Monit. Assess. 2013, 185, 8227-8237. (17) Rajakumar, B.; Reddy, K. P. J.; Arunan, E. J. Thermal Decomposition of 2-Fluoroethanol: Single Pulse Shock Tube and Ab Initio Studies. J. Phys. Chem. A 2003,107, 9782-9793. (18) Saito, S. Microwave Spectroscopic Detection of Vinyl Alcohol, CH2=CHOH. Chem. Phys. Lett. 1976, 42, 399402. (19) Karton, A. Inorganic Acid Catalyzed Tautomerization of Vinyl Alcohol to Acetaldehyde. Chem. Phys. Lett. 2014, 592, 330-333. (20) Da Silva, G. Carboxylic Acid Catalyzed Keto-Enol Tautomerizations in the Gas Phase. Angew- Chem. Int. Ed. 2010, 49, 7523-7525. (21) Ventura, O. N.; Lledós, A.; Bonaccorsi, R.; Bertrán, J.; Tomasi, J. Theoretical Study of Reaction Mechanisms for the Ketonization of Vinyl alcohol in Gas Phase and Aqueous Solution, Theor. Chim. Acta 1987, 72, 175-195. (22) González-Rivas, N.; Cedillo, A. Solvent Effects on the Energetic Parameters and Chemical Reactivity in the Keto–enol Tautomeric Equilibrium of Substituted Carbonyl Compounds. Comput. Theoret. Chem. 2012, 994, 4753. (23) Koch, W. A Chemist’s Guide to Density Functional Theory; Wiley-VCH, 2008. (24) Chai, J.-D.; Head-Gordon, M. Systematic Optimization of Long-Range Corrected Hybrid Density Functionals. J. Chem. Phys., 2008, 128, 084106. (25) Chai, J.-D.; Head-Gordon, M. Long-Range Corrected Hybrid Density Functionals with Damped Atom–Atom Dispersion Corrections, Phys. Chem. Chem. Phys. 2008, 10, 6615-20. (26) Rassolov, V. A.; Ratner, M. A.; Pople, J. A.; Redfern, P. C.; Curtiss, L. A. 6-31G* Basis Set for Third-row Atoms. J. Comp. Chem. 2001, 22, 976-984, and references therein. (27) Kendall, R. A.; Dunning Jr., T. H.; Harrison, R. J. Electron Affinities of the First-row Atoms Revisited. Systematic Basis Sets and Wave Functions. J. Chem. Phys. 1992, 96, 6796-6806. (28) Davidson, R. E. Comment on “Comment on Dunning's Correlation-Consistent Basis Sets”. Chem. Phys. Lett. 1996, 260, 514-518, and references therein. (29) Irving, K.; Kieninger, M.; Ventura, O. N. Basis Set Effects in the Description of the Cl-O Bond in ClO and XClO Isomers (X=H,O,Cl) Using DFT and CCSD(T) Methods. J. Chem. (2019, in press) DOI: 10.26434/chemrxiv.6135389.v1.


14

Page 15 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(30) Petsis, G.; Salta, Z.; Kosmas, A. M.; Ventura, O. N. Theoretical Study of the Microhydration of 1-Chloro and 2-Chloro Ethanol as a Clue for Their Relative Propensity Toward Dehalogenation, sent, Nov. 2018, DOI: 10.26434/chemrxiv.7378445 (31) Weinhold, F.; Carpenter, J. E. in The Structure of Small Molecules and Ions; Naaman, R.; Vager, Z., Eds.; Plenum, 1988, pp. 227-236. (32) Hu, H.; Lu, Z.; Yang, W. J. Fitting Molecular Electrostatic Potentials from Quantum Mechanical Calculations. Chem. Theory Comput. 2007, 3, 1004-1013. (33) Bader, R. F. W. Atoms in Molecules - A Quantum Theory; Oxford University Press: Oxford, UK, 1990. (34) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al, Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford CT, 2013. (35) Azrak, R. G.; Wilson, E. B. Microwave Spectra and Intramolecular Hydrogen Bonding in the 2-Haloethanols: Molecular Structure and Quadrupole Coupling Constants for 2-Chloroethanol and 2-Bromoethanol. J. Chem. Phys. 1970, 52, 5299-5316. (36) Almenningen, A.; Fernholt, L.; Kveseth, K. An Electron Diffraction of the Anti-Gauche Ratio as a Function of Temperature for Ethylene Chlorohydrin, 2-Chloroethanol. Acta Chem. Scandinav. A 1977, 31, 297-305. (37) Durig, J. R.; Zhou, L.; Gounev, T. K.; Klaeboe, P.; Guirgis, G. A.; Wang, L.-F. Far Infrared Spectra, Conformational Equilibria, Vibrational Assignments and Ab Initio Calculations of 2-Chloroethanol. J. Mol. Struct. 1996, 385, 7-21. (38) Baranac-Stojanović, M.; Aleksić, J.; Stojanović, M. Energy Decomposition Analysis of Gauche Preference in 2-Haloethanol, 2-Haloethylamine (Halogen = F, Cl), their Protonated Forms and Anti Preference in 1-Chloro-2fluoroethane. RSC Advances 2015, 5, 22980. (39) Perttilä, M.; Murto, J.; Halonen, L. Infrared Spectra and Normal Coordinate Analysis of 2-Chloroethanol. Spectrochim. Acta A 1978, 34, 469-474. (40) Chitale, S. M.; Jose, C. I. Infrared Studies and Thermodynamics of Hydrogen Bonding in 2-Chloroethanol, 3Chloropropanol and 4-Chlorobutanol. J. Chem. Soc. Faraday II 1980, 76, 233-249. (41) Skvorstov, D. S.; Vilesov, A. F. Using He Droplets for Measurement of Interconversion Enthalpy of Conformers in 2-Chloroethanol. J. Chem. Phys. 2009, 130, 151101. (42) Trindle, C.; Crum, P.; Douglass, K. G2(MP2) Characterization of Conformational Preferences in 2-Substituted Ethanols (XCH2CH2OH) and Related Systems. J. Phys. Chem. A 2003, 107, 6236-6242. (43) Buemi, G. Molecular Conformations and Rotation Barriers of 2-Halogenoethanethios and 2Halogenoethanols: an Ab Initio Study. J. Chem. Soc. Farady Trans. 1994, 90, 1211-1215. (44) Souza, F. R.; Freitas, M. P. Conformational Analysis and Intramolecular Interactions in 2-Haloethanols and their Methyl Esthers. Comput. Theoret. Chem. 2011, 964, 155-159.


15


Page 16 of 40

(45) Wang, K.; Zhang, H.; Liu, Y. Ab Initio Study of Hyperconjugative Effect on Electronic Wavefunctions of 2Chloroethanol. Chin. J. Chem. Phys. 2011, 24, 434-438. (46) Tian, S. X.; Kishimoto, N.; Ohno, K. Intramolecular Hydrogen Bonding in 2-Chloroethanol and 2Bromoethanol and Anisotropic Interactions with He*(23S) Metastable Atoms: Two-Dimensional Penning Ionization Electron Spectroscopy Combined with Quantum Chemistry Calculations. J. Phys. Chem. A 2003, 107, 53-62. (47) Buckley, P.; Giguère, P. A.; Schneider, M. Infrared studies on rotational isomerism. III. 2-Chloro- and Zbromo-ethanol. Can. J. Chem. 1969, 47, 901-910. (48) Soliday, R. M., Bunn, H., Sumner, I., Raston, P. L. Far-Infrared Synchrotron Spectroscopy and Quantum Chemical Calculations of the Potentially Important Interstellar Molecule, 2-Chloroethanol. The Journal of Physical Chemistry A. 2019, doi:10.1021/acs.jpca.8b11333 (49) Bunn, H.; Hudson, R. J.; Gentleman, A. S.; Raston, P. I. Far-Infrared Synchrotron Spectroscopy and Torsional Analysis of the Important Interstellar Molecule, Vinyl Alcohol. ACS Earth Space Chem. 2017, 1, 70-79. (50) Bunn, H., Soliday, R. M., Sumner, I., Raston, P. L. Far-infrared Spectroscopic Characterization of Anti-vinyl Alcohol. Astrophys. J. 2017, 67, 1-5. (51) Estep, M. L., Morgan, W. J., Winkles, A. T., Abbott, A. S., Villegas-Escobar, N., Mullinax, J. W., Turner, W. e., Wang, X., Turney, J. M., Schaefer, H. F. Radicals derived from acetaldehyde and vinyl alcohol, 2017, 19, 2727527287. (52) Rodler, M. J. Microwave Spectrum, Dipole Moment, and Structure of Anti-Vinyl Alcohol. Mol. Spectrosc. 1985, 114, 23−30. (53) Demaison, J.; Császár, A. G.; Margulès, L. D.; Rudolph, H. D. Equilibrium Structures of Heterocyclic Molecules with Large Principal Axis Rotations upon Isotopic Substitution. J. Phys. Chem. A 2011, 115, 14078-14091. (54) Bastiansen, O. Intra-Molecular Hydrogen Bonds in Ethylene Glycol, Glycerol, and Ethylene Chlorohydrin. Acta. Chem. Scand. 1949, 3, 415-421. (55) Marstokk, K.-M.; Møllendal, H. On the Microwave Spectrum of Ethylene Glycol. J. Mol. Struct. (Theochem) 1974, 22, 301-303. (56) Christen, D.; Coudert, L. H.; Suenram, R. D.; Lovas, F. J. The Rotational/Concerted Torsional Spectrum of the g′Ga Conformer of Ethylene Glycol. J. Mol. Spectrosc. 1995, 172, 57-77. (57) Christen, D.; Coudert, L. H.; Larsson, J. A.; Cremer, D. The Rotational–Torsional Spectrum of the g′ Gg Conformer of Ethylene Glycol: Elucidation of an Unusual Tunneling Path. J. Mol. Spectrosc. 2001, 205, 185-196. (58) Buckley, P.; Giguère, P. A. Infrared Studies on Rotational Isomerism. I. Ethylene Glycol. Can. J. Chem. 1967, 45, 397-407.


16

Page 17 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(59) Howard, D. L.; Jørgensen, P.; Kjaergaard, H. G. Weak Intramolecular Interactions in Ethylene Glycol Identified by Vapor Phase OH− Stretching Overtone Spectroscopy. J. Am. Chem. Soc. 2005, 127, 17096-17103. (60) Klein, R. A. Hydrogen Bonding in Diols and Binary Diol–Water Systems Investigated using DFT Methods. II. Calculated Infrared OH-Stretch Frequencies, Force Constants, and NMR Chemical Shifts Correlate with Hydrogen Bond Geometry and Electron Density Topology. A Reevaluation of Geometrical Criteria for Hydrogen Bonding. J. Comput. Chem. 2003, 24, 1120-1131. (61) Crupi, V.; Maisano, G.; Majolino, D.; Migliardo, P.; Venuti, V. Dynamic Evidence of Chemical and Physical Traps in H-bonded Confined Liquids. J. Chem. Phys. 1998, 109, 7394-7404. (62) Crupi, V.; Maisano, G.; Majolino, D.; Migliardo, P.; Venuti, V. J. Anharmonic Effects and Vibrational Dynamics in H-Bonded Liquids by Attenuated Total Reflectance FT-IR Spectroscopy. J. Phys. Chem. A 2000, 104, 3933-3939. (63) Klein, R. A. Hydrogen Bonding in Strained Cyclic Vicinal Diols: the Birth of the Hydrogen Bond. Chem. Phys. Lett. 2006, 429, 633-637. (64) Chopra, D.; Guru Row, T. N.; Arunan, E.; Klein, R. A. Crystalline Ethane-1, 2-diol does not have Intramolecular Hydrogen Bonding: Experimental and Theoretical Charge Density Studies. J. Mol. Struct. 2010, 964, 126-133. (65) Renault, B.; Cloutet, H.; Cramail, H.; Tassaing. T.; Besnard, M. On the Perturbation of the Intramolecular HBond in Diols by Supercritical CO2: A Theoretical and Spectroscopic Study. J. Phys. Chem. A 2007, 111, 41814187. (66) Howard, D. L.; Kjaergaard, H. G. Influence of Intramolecular Hydrogen Bond Strength on OH-Stretching Overtones. J. Phys. Chem. A 2006, 110, 10245-10250. (67) Naganathappa, M.; Chaudhari, A. Mono, di, tri and Tetraethylene Glycol: Spectroscopic Characterization using Density Functional Method and Experiment. Vibrat. Spectrosc. 2018, 95, 7-15. (68) Fortes, A. D.; Suard, E. Crystal Structures of Ethylene Glycol and Ethylene Glycol Monohydrate. J. Chem. Phys. 2011, 135, 234501. (69) Crittenden, D. L.; Thompson, K. C.; Jordan, M. J. T. On the Extent of Intramolecular Hydrogen Bonding in Gas-Phase and Hydrated 1, 2-Ethanediol. J. Phys. Chem. A 2005, 109, 2971-2977. (70) Chaudhari, A.; Lee, S.-L. A Computational Study of Microsolvation Effect on Ethylene Glycol by Density Functional Method. J. Chem. Phys. 2004, 120, 7464-7469. (71) Kazerouni, M. R.; Hedberg, L.; Hedberg, K. Conformational Analysis. 21. Ethane-1,2-diol. An ElectronDiffraction Investigation, Augmented by Rotational Constants and ab Initio Calculations, of the Molecular Structure, Conformational Composition, SQM Vibrational Force Field, and Anti-Gauche Energy Difference with Implications for Internal Hydrogen Bonding. J. Am. Chem. Soc. 1997, 119, 8324-8331.


17


Page 18 of 40

(72) Bultinck, P.; Goemmine, A.; Van de Vondel, D. Ab Initio Conformational Analysis of Ethylene Glycol and 1, 3Propanediol. J. Mol. Struct. (Theochem) 1995, 357, 19-32. (73) Van Alsenoy, C.; Van den Enden, L.; Schäfer, L. Ab Initio Studies of Structural Features not Easily Amenable to Experiment: Part 31. Conformational Analysis and Molecular Structures of Ethylene Glycol. J. Mol. Struct. 1984, 108, 121-128. (74) Nagy, P. I.; Dunn III, W. J.; Alagona, G.; Ghio, C. Theoretical Calculations on 1, 2-Ethanediol. 2. Equilibrium of the Gauche Conformers with and without an Intramolecular Hydrogen Bond in Aqueous Solution. J. Am. Chem. Soc. 1992, 114, 4752-4758. (75) Takeuchi, H.; Tasumi, M. Infrared-Induced Conformational Isomerization of Ethylene Glycol in a LowTemperature Argon Matrix. Chem. Phys. 1983, 77, 21-34. (76) Walder, E.; Bauder, A.; Gunthard, Hs. H. Microwave Spectrum and Internal Rotations of Ethylene Glycol. I. Glycol-O-d2. Chem. Phys. 1980, 51, 223-239. (77) Krishnan, K.; Krishnan, R. S. Raman and Infrared Spectra of Ethylene Glycol. Proc. Indian Acad. Sci. 1966, 64, 111-115. (78) Podo, F.; Némethy, G.; Indovina, P. L.; Radics, L.; Viti, V. Conformational Studies of Ethylene Glycol and its two Methyl Ether Derivatives: I. Theoretical Analysis of Intramolecular Interactions. Mol. Phys. 1974, 27, 521539. (79) Krueger, P. J.; Mettee, H. D. Spectroscopic Studies of Alcohols: Part VII. Intramolecular Hydrogen Bonds in Ethylene Glycol and 2-Methoxyethanol. J. Molec. Spectrosc. 1965, 18, 131-140. (80)) Caminati, W.; Corbelli, G. Conformation of Ethylene Glycol from the Rotational Spectra of the Nontunneling 0-monodeuterated Species. J. Mol. Spectrosc. 1981, 99, 572-578. (81) Legon, A. C., Rego, C. A., & Wallwork, A. L. Rotational spectrum of a short-lived dimer of oxirane and hydrogen chloride: Evidence for a bent hydrogen bond. J. Chem. Phys. 1992, 97, 3050–3059. (82) Iijima, T.; Kimura, M. Zero-point Average Structure of Acetaldehyde. Bull. Chem. Soc. Japan 1969, 42, 21592164. (83) Chong, D. P. MP2 or B3LYP: computed bond distances compared with CCSD (T)/cc-pVQZ. Can. J. Chem. 2018, 96, 336-339.


18

Page 19 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47


Table 1. Experimental and theoretical abundances (in %) and relative energies with respect to th Gg' structure (in kcal mol-1) of the rotational conformers of 2ClEtOH

Property Abundancea

Method

Gg’ 81.4% 87.2% 75.5%

B97X-D/cc-pVTZ MP2/aug-cc-pVTZ MP2/6-31++G(d,p) Exp.c b 0.00 ET B97X-D/cc-pVTZ MP2/6-31G(d,p) 0.00 B3LYP/aug-cc-pVTZ 0.00 G2(MP2) 0.00 b,c Exp. 0.00 b,d 0.00 (E+ZPE) B97X-D/cc-pVTZ CCSD/cc-pVDZ 0.00 MP2/aug-cc-pVTZ 0.00 MP2/6-31++G(d,p) 0.00 b,e 0.00 H B97X-D/cc-pVTZ b,f Exp. 0.00 Exp.b,g 0.00 0.00 Gb,e B97X-D/cc-pVTZ a Relative abundance of the rotamers, using Boltzmann’s distribution b Relative to the Gg´ rotamer in Kcal mol-1 c Derived from electron diffraction data d including zero point energy e At 298K f Derived from spectroscopic studies in the far infrared region

2ClEtOH Rotational Tg Tt 6.9% 7.0% 5.9% 3.5% 12.4% 8.3% 7.0-7.6% 1.70 1.93 1.51 1.71 1.35 1.07 1.32 1.72 2.42 1.59 1.69 1.50 1.63 1.59 1.48 1.35 1.07 1.74 1.91 1.19 1.20 1.46 1.46


Gt 3.2% 2.3% 3.3%

Gg 1.6% 1.0% 1.1%

2.55 2.48 1.85 1.87

2.68 2.60 2.50 2.34

2.23 2.42 2.14 1.85 2.42

2.48 2.51 2.62 2.50 2.58

1.92

2.32

Reference This work 45 46 36 This work 43 46 42 36 This work 41 45 46 This work 37 47 This work

19

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Page 20 of 40

Table 2. Relative energies (including ZPE), enthalpies and Gibbs free energies (at 298.15K) of the different stable species and transition states studied in this work (in kcal mol-1) calculated with the BD97X-D method. Species

1ClEtOH 2ClEtOH (Gg’) Ethylene + HOCl a-Vinyl alcohol s-Vinyl alcohol Oxirane Ethylene glycol Acetaldehyde TS1 TS2 TS4 TS6 TS7 TS8 TS9 TS10 TS11 TS12 TSas

Without water (ET+ZPE) 0 6.0 69.9 25.1 23.2 34.2 8.7 7.3 30.5 42.1 77.8 113.8 67.6 70.5

6-31G(d) H(298) 0 6.0 71.2 26.5 24.5 35.2 8.5 8.8 30.6 42.2 79.0 114.5 67.9 70.4

G(298) 0 5.9 59.8 15.8 14.1 25.3 8.9 -2.2 29.9 41.6 68.8 112.1 66.9 70.3

(ET+ZPE) 0 5.8 64.1 17.4 16.0 32.2 9.5 4.9 29.0 39.3 71.2 113.8 63.6 69.4

cc-pVTZ H(298) 0 5.8 65.4 18.8 17.3 33.2 9.3 6.4 29.1 39.4 72.3 114.5 63.9 69.2

G(298) 0 5.7 54.1 8.2 6.9 23.4 9.7 -4.5 28.4 38.9 62.2 112.1 63.0 69.2

65.9

66.2

65.3

69.1

69.3

68.6

28.2

29.3

19.1

20.7

21.8

11.7


With water cc-pVTZ (ET+ZPE) H(298) G(298) 0 0 0 7.6 7.7 7.1 72.1 73.8 53.7 14.4 15.3 11.7 11.8 12.3 10.5 26.6 26.9 25.1 11.3 10.9 11.1 -0.3 0.5 -2.7 12.0 11.5 12.3 24.8 24.3 25.2 18.9 18.2 19.5 98.0 98.7 96.5 56.3 56.1 56.3 69.5 69.1 69.8 78.0 77.5 78.3 56.0 55.4 56.5 57.1 16.1

56.4 16.4

58.1 14.8

20

Page 21 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Captions for the figures Figure 1. Scheme of the species and reaction paths explored in this work. An ancillary water molecule complexed with the different species was considered for the full reaction path involving ethylene glycol. Notice that the path from ethylene glycol to acetaldehyde (which was not investigated in this work) formally proceeds by the loss of one water molecule, but possibly several elementary radical reactions are involved, since a hydrogen transfer from one carbon to the other or from oxygen to carbon is needed to complete the process. TSas is the transition state between the anti and syn conformers of vinyl alcohol. Figure 2. Geometrical structure of the main species studied in this work and the transition states connecting them according to the scheme in Fig. 1. For the stable structures, when they are available, the first entry correspond to the experimental values, followed by other accurate calculations and, at last, our own calculations at DFT and, eventually, MP2 levels, according to the following specifications: (a) 2ClEtOH, microwave spectroscopy35, electron diffraction36, B97X-D/cc-pVTZ calculations in this paper, B97X-D/aug-cc-pVQZ and MP2/aug-cc-pVTZ calculations30; (b)1ClEtOH, B97X-D/cc-pVTZ calculations in this paper, B97X-D/aug-cc-pVQZ and MP2/aug-cc-pVTZ calculations30; (c) vinyl alcohol syn-conformer microwave spectrum18, CCSD(T)/ANO251, CCSD(T)/cc-pVQZ50 and our own B97X-D/cc-pVTZ calculations; (d) vinyl alcohol anti-conformer microwave spectrum52, CCSD(T)/ANO251, CCSD(T)/cc-pVQZ50 and our own B97X-D/cc-pVTZ calculations; (e) vinyl alcohol syn-conformer complexed with the HCl molecule arising from the dehalogenation of 1ClEtOH; the signed italicized numbers show the difference of geometries in the complex with respect to those in the isolated molecules, B97X-D/cc-pVTZ this work; (f) same as (e) but for the anti-conformer; (g) experimental and theoretical data for oxirane from ref. 53, our own B97X-D/cc-pVTZ calculations; (h) structure of the oxirane HCl complex, B97X-D/cc-pVTZ this work (signed italicized values are geometry differences with respect to the isolated monomers). This complex was studied experimentally by Legon et al81 who predicted a structure similar to the one find here but with a significantly larger O-H bond length (i) acetaldehyde, experimental82, CCSD(T)/cc-pVQZ83, our own B97X-D/cc-pVTZ calculations; (j) acetaldehyde-HCl complex, B97X-D/cc-pVTZ, this work; (k) ethylene glycol, semi-experimental data71, CCSD(T)/aug’-cc-pVQZ calculations67, our own B97X-D/cc-pVTZ calculations; (l) ethylene glycol HCl complex, B97X-D/cc-pVTZ, this work. (m)-(s) transition states, B97X-D/cc-pVTZ calculations. In the case of the HCl complexes, the charge transfer between monomers is shown at the NBO and HLY levels. In the case of the transition states, the first italicized entry is the difference with respect to the reactant, the second one is the difference with respect to the product. All distances in Å and angles in degrees. Figure 3. Same as in Figure 2, reactants products and transition states on the path including an ancillary water molecule. Some important geometrical parameters at the B07X-D/cc-pVTZ level are given. To the best of our knowledge, no experimental information is available on these systems. All distances and angles are in Å and degrees. Figure 4. (a) Gas phase infrared spectrum of 2ClEtOH (upper panel); (b) Gas phase Raman spectrum of 2ClEtOH (middle panel); (c) isovalue contour lines of the Laplacian of the electronic density in a plane containing the Cl atom and the OH group (bottom panel, left); (d) same as in (c) but in a perpendicular plane to the former one (bottom panel, right). All calculations at the B97X-D/cc-pVTZ level. Figure 5. Superposition of the infrared spectra of 2ClEtOH (blue) and 1ClEtOH (red). Calculations performed at the B97X-D/aug-cc-pVTZ level. Figure 6. Superposition of the IR spectrum for each of the five stable conformers of 2ClEtOH. See the Additional Information section for the structures optimized at the B97X-D/cc-pVTZ level. Figure 7. Boltzmann’s weighted addition of the spectra for the five stable conformers of 2ClEtOH (see Additional Information section) superposed to the IR spectrum of the Gg’ conformer. Figure 8. Isovalue contours of the Laplacian in different planes containing both hydroxylic groups and the OHO substructure are shown to display the absence of a BCP between oxygen and hydrogen, thus supporting the absence of a hydrogen bond. The surface of electrostatic potential displayed in the last panel shows the positively charged region around the hydrogens and that negatively charged around the oxygens, displaying then the electrostatic attraction between the hydroxyl groups. Figure 9. Energy profiles of the reactants, intermediates, products and transition states (see Fig. 2) participating in the scheme in Fig. 1. All values expressed in kcal mol-1 relative to 1ClEtOH. Profiles shown were obtained using the B97X-D method with the 6-31G(d,p) basis set (dotted lines) and the cc-pVTZ basis set (full lines). A table with the actual barriers for the direct transformation is included in the figure for clarity. Figure 10. Same as in Fig. 9, but now including an ancillary water molecule (energies in kcal mol-1). Full lines are the profiles with respect to the 1ClEtOH complex, dashed lines are the profiles with respect to the 1ClEtOH.W, both taken independently as 0.0 kcal mol-1.The path going through the oxirane-HCl-water complex ends up at the anti-VA.W complex, the same as the direct path from 2ClEtOH.W to the vinyl alcohol water complex, but the lines have not been joined there, in order to keep the graph cleaner. The abbreviations used in the graph


21


Page 22 of 40

are as follows: aVA = anti-vinyl alcohol-HCl complex, Oxi = oxirane-HCl complex, sVA = syn-vinyl alcohol-HCl complex, EGly = ethylene glycol-HCl complex; Acet = acetaldehyde.


22

Page 23 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FIGURE 1


23


Page 24 of 40

FIGURE 2


24

Page 25 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



25



Page 26 of 40

26

Page 27 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



27


Page 28 of 40

FIGURE 3


28

Page 29 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



29



Page 30 of 40

30

Page 31 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FIGURE 4


31



Page 32 of 40

32

Page 33 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FIGURE 5


33


Page 34 of 40

FIGURE 6


34

Page 35 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FIGURE 7


35


Page 36 of 40

FIGURE 8


36

Page 37 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47


FIGURE 9


37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47


Page 38 of 40


38

Page 39 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47


FIGURE 10


39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47


Page 40 of 40


40

TOC Graphic

Computational Evidence Suggests that 1 ... - ACS Publications

Recommend Documents