Prereactive Complexes in Chlorination of Benzene, Triazine, and

Article pubs.acs.org/JPCA

Prereactive Complexes in Chlorination of Benzene, Triazine, and Tetrazine: A Quantum Chemical Study Davor Šakić and Valerije Vrček* Faculty of Pharmacy and Biochemistry, University of Zagreb, A. Kovačića 1, HR-10000 Zagreb, Croatia S Supporting Information *

ABSTRACT: In order to perform a complete search for prereactive complexes between arenes and chlorine, the stochastic search method was employed. Stationary points are optimized at B3LYP, M05-2X, and MP2 levels, while improved energetics are calculated using the B2PLYP-D method, which includes corrections important for accurate description of dispersion forces. New intermediates were located and their mechanistic relevance has been discussed. It has been suggested that, at least in the gas-phase, the T-shaped complex precedes the formation of classical benzene/chlorine π-complex. No σ-complex is found on the energy surface, unless explicit counterions are included in calculations. Neither π- nor σ-complexes were located on the reactant side of chlorination of triazine, but only linear and T-shaped complexes were identified as stable minima. These structures represent important prereactive complexes for chlorination of triazine. In the case of tetrazine, which is unlikely to undergo direct chlorination, only two complexes (resting and T-shaped) were located. revisited.15−19 Whereas π- and σ-complexes are commonly postulated as the only possible intermediates, other transient structures were usually ignored or oversighted. Herewith, we propose that σ-complex, in the case of chlorination of benzene, but also π-complex, in the case of chlorination of triazine and tetrazine, are still elusive species. In addition, we have performed an extensive screening for prereactive complexes, which can be included as new sequential intermediates along the corresponding EAS reaction coordinate.

1. INTRODUCTION Despite the huge body of data that has been accumulated, the mechanism of halogenation of benzene continues to be the subject of active research and some controversy. For example, several authors have claimed that the σ-complex, commonly accepted as transient Wheland intermediate in electrophilic aromatic substitutions (EAS), can successfully be located at the energy surface by using quantum-chemical approach in gas phase.1−3 In contrast, others have claimed that no σ-complex is existent until Lewis acid catalyst or high-polarity model solvents are included in calculations.4−6 Similar debates have been ongoing over the existence of benzenium/ethene complex7−9 and the exact nature of chlorine atom/benzene complex,10,11 to name but a few examples that are closely related to prereactive complexes in EAS. Recently, Schleyer et al. have challenged the reaction mechanism paradigm for the halogenation of benzene, demonstrating that alternative pathways, such as the addition− elimination route, may compete successfully with those depicted in textbooks for EAS reactions.12 In their work, the universal appropriateness of considering the classic aromatic bromination mechanism has been questioned. It has also been emphasized that for the complete description of the EAS mechanism, all reactive intermediates are to be considered.13 In this regard, prereactive complexes between a halogen molecule and an aromatic compound are of special importance. Although they play limited roles in the overall energetics of EAS, they can provide mechanistic insight from a structural point of view, i.e., prereactive complexes can lead to mechanistically viable transition state structures.14 This renewed interest in EAS mechanism has initiated detailed investigations of the corresponding potential energy landscapes. Comprehensive searches for stationary points, especially on the reactant side of EAS reactions, have been resumed and © 2012 American Chemical Society

2. COMPUTATIONAL DETAILS The quantum chemical calculations were performed using the Gaussian0320 and Gaussian0921 suites of programs. All structures were fully optimized using density functional theory (DFT) methods employing three nonlocal functionals: B3LYP,22 MPWB1K,23 and M05-2X.24 B3LYP (χ = 20) is the most popular hybrid method, whereas MPWB1K (χ = 44) and M05-2X (χ = 56) are hybrid meta functionals with a successively increased fraction of HF exchange (χ) in the corresponding functional. The standard 6-311+G(d,p) basis set and the augmented correlation consistent double-basis set (AUG-cc-pVDZ) basis set was used for geometry optimizations and frequency calculations. For comparison, all structures were reoptimized at the MP2/AUG-cc-pVDZ level.25 Improved energetics have been calculated using the double hybrid B2PLYP-D/AUG-cc-pVTZ method with empirical dispersions, which combines the exact HF exchange with an MP2-like correlation to a DFT calculation.26 Thermochemical corrections to enthalpies at 298.15 K have been calculated at Received: November 15, 2011 Revised: January 3, 2012 Published: January 4, 2012 1298

dx.doi.org/10.1021/jp210993k | J. Phys. Chem. A 2012, 116, 1298−1306

The Journal of Physical Chemistry A

Article

prereactive complexes obtained by Olah and Wheeler, respectively. In Olah’s study,16 on the reactant side, only two minima were located at the B3LYP/6-31++G(d,p) level. This result we have easily reproduced by using the stochastic search method. In addition, two more prereactive complexes were located not reported in their work (Chart 1). In the case of

the corresponding levels of theory using the rigid rotor/ harmonic oscillator model. Analytical vibrational analysis were performed to characterize each stationary point as a minimum (NImag = 0) or first-order saddle point (NImag = 1). Intrinsic reaction coordinate (IRC) calculations were performed at the corresponding level of theory to identify the minima connected through the transition state. The initial geometries used were that of the corresponding transition states structures, and the paths were followed in both directions from that point. This method verified that a given transition structure indeed connected the presumed energy minimum structures.27 The solvent effects were calculated using the self-consistent reaction field (SCRF) method based on the integral equation formalism model (PCM).28 The solvent relative permittivity of ε = 2.228 (carbontetrachloride, CCl4) was used. A complete search for minima at the potential energy surface was performed by the stochastic search method.29 The method is analogous to the stochastic random search procedure for finding conformers previously developed and extensively used for molecular mechanic surfaces.30 An important feature of these two methods is that intensive enough application is virtually certain to find all of the structures. In applying this method, a starting structure is subjected to a kick, which moves each atom in a random direction over a random distance within a sphere of given radius, which represent the maximal kick size. Optimization of the kicked structure with a quantum mechanical optimizer could take it back to the initial structure or it can be refined to give a different isomer. For a locally modified version of the original kick procedure, see details in our earlier report.31

Chart 1. Prereactive Complexes of Benzene and NO2+ Cation Located by Stochastic Search Procedure at the B3LYP/6-31++G(d,p) Levela

a

New minima found, not reported earlier, are designated by an asterisk.

Wheeler’s study,17 three different interaction complexes were located as minima at the B3LYP/6-31G(d) level. Again, the outcome of this search we have successfully reproduced by the kick procedure. Moreover, four additional minima were located, and all of them were calculated within the energy range reported by the authors (Chart 2). Therefore, all previously identified stationary points were reproduced, and various new structures were found, which confirm that the kick procedure could be superior in comparison to other methods that search for all minima.36 3.2. Structures and Energies of Prereactive Cl2/ Benzene Complexes. The initial kicked structures of chlorine/benzene complexes generated by the kick procedure were fully optimized at B3LYP, MPWB1K, and M05 levels (employing 6-311+G(d,p) basis set) followed by computation of the analytical frequencies at corresponding levels. These calculations provide a relatively quick screen of all possible configurations (see Chart 3). In fact, these optimized structures correspond to various ways by which Cl2 can approach the π-system of benzene. Eleven different B3LYP-optimized structures (Chart 3) were located as stable intermolecular complexes between Cl2 and benzene. A number of stationary points on the product side of reaction (different substitution or addition products, as well as transition state structures) were also found by the kick procedure at the corresponding PES but were not considered in this study (see Supporting Information). As expected, π-complexes 1a and 1b were located as the most stable isomers (at the B3LYP level of theory).37 The π-complex 1a is located as a genuine minimum at all levels of theory employed (see below), whereas the π-complex 1b is characterized by one imaginary frequency (except for B3LYP

3. RESULTS AND DISCUSSIONS 3.1. Stochastic Search for Complexes between Electrophiles and Benzene. In order to perform an efficient search for all possible prereactive interaction complexes between chlorine and benzene (i.e., corresponding minima on the potential energy surface), the stochastic search methodology was used.29 The importance of an automated unbiased isomer search method, such as stochastic search methodology, to help find isomers on the potential energy surface of molecules in the gas phase is emphasized recently.32,33 Complexes between benzene and chlorine were explored using the locally modified version of Saunders’ kick procedure.34,35 Two chlorine atoms are initially placed at a common point in geometrical space and then kicked in random directions, while coordinates of all benzene atoms were kept frozen. A number of starting geometries obtained this way were then optimized automatically to stationary point structures using different DFT methods (see below). The efficiency of the kick procedure was tested against a hand-made set of different NO2+/benzene and NO+/benzene

Chart 2. Prereactive Complexes of Benzene and NO+ Cation Located by Stochastic Search Procedure at the B3LYP/6-31G(d) Levela

a

New minima found, not reported earlier, are designated by an asterisk. 1299



Article

Chart 3. Prereactive Complexes of Benzene and Chlorine Located by Stochastic Search Procedure at the B3LYP/6-311+G(d,p) Level

level). It is in agreement with earlier experimental14,38,39 and theoretical40−43 studies on halogenation of benzene, which find that this overcenter complex (η1) is not a stable structure, while the overbond complex (η2) is a real minimum.44 In this work, we demonstrate that the overcenter structure (π-complex 1b) is a transition state of the 6-fold degenerate equilibrium (1a ⇆ 1b ⇆ 1a ⇆ 1b...), i.e., corresponds to transition state for ambulatory motion of the chlorine about the perimeter of the benzene molecule. The calculated energy barrier (ΔG#, at 298.15 K) for this process is 4 kJ/mol, suggesting a flat PES and the ease of the conversion between the two degenerate π-complexes. If the thermal energy corrections are applied to the electronic energy, an inversion in enthalpies for the complex 1a and the transition state 1b occurs (ΔH# < 0). The energy level of 1a becomes higher than that of 1b because of the higher vibrational energy contribution to the intermediate than to the transition state in the ΔH#. This situation is typical of energy surfaces that are flat in the region of the transition state.45 The relative energy levels become restored to normal (ΔG# > 0) only when the entropy effect is accounted for. The similar activation energy of only 2 kJ/mol has been calculated for the chlorine atom to travel around the benzene ring in Cl/benzene complex.11 It is in contrast to the migration of the chloronium ion (Cl+) around the aromatic ring for which a rather high activation barrier was determined (>70 kJ/mol).46 Interestingly, in the case of nitrosation of benzene no transition state structure for ambulatory motion of NO+ is located,17 due to the flatness of the PES in the vicinity of the π-complexes. The authors suggested a barrierless interconversion of the six identical π-complexes, but this conclusion is valid only for results obtained at the HF/3-21G level of theory. At the B3LYP or M05 level, the corresponding transition state structure is easily located by using the stochastic search procedure (see Supporting Information). In addition to π-complexes 1a, 1b, and 1c,47 structures 1d and 1e were located as the first-order saddle points (NImag = 1). They are characterized with a displaced resting conformation in which the chlorine molecule straddles the C−C bond (1d) or overlaps the C−H bond (1e). Collin et al. ruled out these resting structures for the Cl2/benzene complex on the basis of IR experiments.48 Similar complexes between iodine and benzene were considered earlier and were found to be less stable than axial π-complexes.49,50 Several more complexes (1f−1k) have emerged from the stochastic search at the B3LYP level, which were not described earlier (Chart 3). These prereactive interaction complexes of benzene with chlorine can be classified into two sets. The first one is characterized by T-shaped geometries (1f and 1g),

whereas the second set includes four different in-plane structures (1h−1k). The structures 1h−1k in which the chlorine approaches the benzene in a linear fashion are related to a number of in-plane structures characteristic for complexes between benzene and small molecules. It has been shown previously that benzene forms stable complexes with anions (Cl−, NO3−, and ClO4−) through in-plane hydrogen-bonding interactions.51−54 As well, in-plane hydrogen-bonded complexes with neutral molecules (HF, water, ammonia, or methanol) can be formed if benzene55 or fluorinated benzenes are used.56,57 No similar in-plane complexes of benzene and halogen molecules were described earlier. Furthermore, we have found that in-plane type of complexation between chlorine and triazine is the most viable interaction step preceding the chlorination of heteroaromatics (see below). All complexes (1a−1k) are relatively close in energy (eleven B3LYP-optimized structures span a range of ca. 7 kJ/mol), but only π-complexes (1a, 1b, and 1c) and conformer 1e were calculated more stable than the separated reactants Bz and Cl2 (Table 1) in the gas phase. It is known that DFT does not include dispersion forces that have an important contribution to the interaction energies of weakly bounded complexes. To remedy the situation, B2PLYP-D method26 was used to treat long-range dispersion interactions. If empirical van der Waals corrections are included, all complexes become favored in the gas phase (ΔH < 0 kJ/mol) except structures with linear configurations (1h−1k). Solvation energies for all complexes were calculated by using PCM/UFF model with the dielectric constant of ε = 2.228 to account for bulk effects of carbontetrachloride (CCl4), which has been used as a solvent in earlier experimental studies.58 When both dispersion forces and solvent effects are included (ΔH + ΔGsolv), the calculated complexation energy for the formation of π-complex 1a is −3.0 kJ/mol, which is somewhat lower than the experimental value of −4.6 kJ/mol.59,60 Three different π-complexes (1a−1c) and all unexpected (not described earlier) geometries 1d−1k, which showed up in stochastic search, were re-examined at M05-2X and MP2 levels of theory. It has been shown earlier that the M05-2X functional gives an excellent performance for the calculation of noncovalent interactions found in charge-transfer complexes.61,62 Both 6-311+G(d,p) and AUG-cc-pVDZ basis sets were used (see Supporting Information) in combination with this hybrid meta DFT method. It has been found that B3LYP/6-311+G(d,p) is somewhat better than the M05-2X/AUG-cc-pVDZ method in the description of geometry for π-complex 1a, but the latter method gives more accurate complexation energy for its formation as compared to the experiment.59,60 1300



Article

Table 1. Relative Energies ΔH (in kJ/mol, at 298.15 K) for Prereactive Complexes between Chlorine (Cl2) and Benzene (Bz) Calculated at Different Levels of Theory B3LYPa

Bz + Cl2 1a 1b 1c 1d 1e 1f 1g 1h 1i 1j 1k

B2PLYPD// B3LYPb

M05-2Xc

ΔH (NImag)f

ΔH + ΔGsolvg

ΔH

ΔH + ΔGsolvg

ΔH (NImag)f

0.0 −3.3 (0) −3.3 (0) −3.8 (2) +0.2 (1) −2.3 (1) +0.5 (0) +0.5 (0) +2.5 (0) +2.5 (0) +3.6 (0) +3.5 (0)

0.0 +2.5 +2.4 +7.0 +11.8 +7.8 +7.9 +9.0 +9.5 +9.6 +7.8 +8.7

0.0 −8.8 −8.7 −8.9 −5.6 −6.7 −2.7 −4.6 +0.4 +1.2 +3.7 +3.6

0.0 −3.0 −2.9 +1.9 +6.0 +3.4 +4.7 +3.9 +7.4 +8.3 +7.9 +8.8

0.0 −10.7 (0) −13.0 (1) −11.1 (2) Converged to 1a Converged to 1a −3.2 (0) −0.1 (0) −0.1 (0) −2.8 (1) +0.3 (1) +3.8 (1)

B2PLYPD// M05-2Xd

MP2e

ΔH + ΔGsolvh

ΔH

ΔH + ΔGsolvh

ΔH (NImag)f

ΔH + ΔGsolvi

0.0 −4.8 −7.1 −1.2

0.0 −10.3 −12.8 −10.7

0.0 −4.4 −6.9 −0.8

0.0 −20.2 (0) −22.5 (1) −24.0 (2)

0.0 −16.1 −18.4 −15.9

−1.0 +6.5 +5.4 +2.9 +5.2 +8.3

−2.6 −0.7 −0.9 −3.5 −1.1 +3.4

−0.8 +5.9 +4.6 +2.2 +3.8 +7.9

−10.7 (0) −8.2 Converged to 1f −11.9 (2) −9.7 Converged to 1f −6.8 (2) −3.5 −7.8 (1) −4.6

a

At the B3LYP/6-311+G(d,p) level. bAt the B2PLYP-D/TZVPP//B3LYP/6-311+G(d,p) level. cAt the M05-2X/AUG-cc-pVDZ level. dAt the B2PLYP-D/AUG-cc-pVTZ//M05-2X/AUG-cc-pVDZ level. eAt the MP2/AUG-cc-pVDZ level. fNumber of imaginary frequencies. gPCM/UFF/ B3LYP/6-311+G(d,p). hPCM/UFF/M05-2X/AUG-cc-pVDZ. iPCM/UFF/MP2/AUG-cc-pVDZ method in the model solvent of ε = 2.228 (CCl4).

distance in π-complex 1a (see above). T-shaped complex 1f is only 3.6 kJ/mol (at the B2PLYP-D//M05 level) less stable than the global minimum (1a) on the reactant side of EAS. The transition state structure TS1a−1f (Figure 1) connecting these two intermediates is located and is characterized by a displaced chlorine molecule positioned between two bordering configurations. The calculated barrier for 1f → 1a is only 1.1 kJ/mol, suggesting a flat energy surface and a fast conversion of the T-shaped complex to π-complex. Herewith, we propose that T-shaped complex 1f may precede the formation of the classical π-complex 1a and that it is involved in an extended continuum of structures along the EAS reaction coordinate. It is not likely that the weakly bonded complex 1f can be studied directly experimentally, and therefore, the main source of information for its existence and properties remains in the field of theory. 3.3. Quest for σ-Complex. Cationic σ-complex (or benzenium ion) of arenes with various electrophiles are commonly accepted as transient (Wheland) intermediates in EAS. Olah and co-workers were the first to obtain the direct NMR-spectroscopic evidence for the existence of the NO2+ and Cl+ complexes in magic acid solution at −70 °C by using hexamethylbenzene as an aromatic donor.66 Isolation and X-ray structure of chloroarenium cations as Wheland intermediates was also reported,67 while the identity of chlorohexamethylbenzenium cation is confirmed by comparison of the 1H NMR spectrum with that published by Olah. It is also shown that the spectral observation and isolation of chloroarene cations is only feasible with the fully substituted hexaalkylbenzenes.67 Therefore, it is of interest to see whether the Wheland type of complex can be, at least computationally, identified in the case of chlorination of benzene. Ben-Daniel et al. have used the B3LYP method to search for stable σ-complex involved in the chlorination of benzene.1 However, the structure A they assigned as σ-complex (Chart 4) is not a minimum in the gas phase nor in the solvent model calculation. It collapses, during geometry optimization at several different levels of theory, to the cis-addition product.68 In another study, Smith has calculated the gas phase energy diagram for the bromination of benzene in which the σ-complex intermediate B (Chart 4) is supposedly involved.2

The distance between the centers of mass of the benzene and chlorine molecule in Cl2/benzene π-complex is 4.6 Å,63 whereas the calculated values for 1a are 4.55 and 4.26 Å at B3LYP and M05-2X, respectively. However, the experimental interaction energy for π-complex (ΔH = −4.6 kJ/mol) is reproduced by the M05-2X/AUG-cc-pVDZ method only (ΔH = −4.4 kJ/mol in Table 1). A number of minima found at the B3LYP level have disappeared at the corresponding potential energy surface calculated by the M05-2X/AUG-cc-pVDZ method. Along with the π-complex 1a, only the T-shaped complexes 1f and 1g and the linear complex 1h have been located as real minima (NImag = 0). All other structures have been characterized as the first- or higher-order stationary points, or have converged, during geometry optimization, to the global minimum structure 1a. In the case of the MP2/AUG-cc-pVDZ model, all stationary points were calculated more stable than the separated reactants Bz and Cl2. It is known that the MP2 method overestimates electron correlation effects in complexes that leads often to overbinding.64 Out of eight minima located at the B3LYP level, only 1a and 1f have survived at the energy surface calculated by the MP2 method. However, the MP2 results have to be taken with care, as recently shown by Kolboe and Svelle7 and Moran et al.65 According to these studies, the MP2 failure to describe benzenium/ethene complex7 as a local minimum is not the basis set dependent problem but is method-specific.9 However, the MP2 failure to locate the expected planar minimum of benzene is found to be the basis set dependent problem, which can be solved by using the traditional Dunning basis sets (e.g., AUG-cc-pVDZ in our case), instead of standard Pople’s basis sets.65 It comes out that, in addition to the π-complex 1a, only the T-shaped structure 1f is a stable structure at all theoretical levels employed (different levels of theory are presented in Supporting Information). The similar was found for benzene/ iodine complex, which emerged from molecular dynamic simulations.49 In the T-shaped geometry, the center of mass of chlorine is sandwiched between two hydrogen atoms of benzene (Figure 1). The calculated distance between centers of masses is 4.8 Å, which is slightly longer than the corresponding 1301



Article

Figure 1. Intermediates on the reactant side of benzene chlorination coordinate calculated at the B2PLYPD/AUG-cc-pVTZ//M05/AUG-cc-pVDZ level in the gas phase (dashed line) and in the model solvent (solid line). Implicit solvent effects have been calculated at the PCM/UFF/M05/AUGcc-pVDZ level (ε = 2.228 for CCl4). All distances are in angstroms.

Interestingly, electrophilic aromatic substitution is generally considered to occur via the sequential formation of π- and σ-intermediates despite the fact that in most cases the nature of these stationary points is not clear.13,40,67,70,71 In this work, we provide additional computational evidence that σ-complex is not a stable intermediate in the uncatalyzed chlorination of benzene. No σ-complex is located by the stochastic search method suggesting that this arenium ion does not exist as a minimum on the corresponding PES. Our attempts to optimize geometry of a stable σ-complex failed at all levels applied (see above). All starting geometries, presumed to describe σ-complex, converged back to the more stable π-complex 1a and substitution product (chlorobenzene) or to cis- and trans-addition products (5,6-dichlorocyclohexa-1,3-dienes). It is in agreement with experimental and computational study, which showed that σ-complex is not a stable intermediate in the gas phase or in the solvent of low dielectric constant.4,12,17 Zhang and Lund4 have suggested that, in the case of uncatalyzed chlorination of toluene, only with increasing the solvent polarity (above dielectric of ca. 10) does the generally accepted arenium ion pathway becomes feasible. While not found in the gas phase or in the solvent continuum model, the σ-complex can be located as a minimum in the presence of Lewis acid catalyst.72 We have found that BF4− and AlCl4−anions are suitable counterions, which can effectively stabilize σ-complex C6H5Cl+ (Figure 2). It is in agreement with analogous results from molecular orbital studies of chlorination of toluene5 and benzene6 catalyzed by AlCl3.

Chart 4. Different Reaction Intermediates A, B, and C Assigned Earlier As σ-Complexes in the Aromatic Halogenations

However, this σ-complex B is not a real minimum since the bromide ion was fixed at the van der Waals contact distance (ca. 4 Å) to the ring-bound bromine during geometry optimization. When this geometry constraint is removed, the optimization of cationic σ-complex coordinated to Br− results in the formation of the corresponding π-complex.69 There is one more computational study on electrophilic aromatic halogenation in which an optimized structure has been erroneously assigned as σ-complex.3 In the study of the iodination of methoxybenzene by iodine monochloride, Wang et al. suggested that a planar aromatic structure C (Chart 4) corresponded to the arenium ion intermediate. However, such a structure cannot be classified as an arenium σ-complex according to criteria set by others. The bond angles of the ipsocarbon in C do not reveal its sp3 hybridizationm and altered bond lengths in the ring do not result in a cyclohexadienyl structure. This represents evidence for the lack of the σ-character of this complex. 1302



Article

a weak interaction, which is mainly electrostatic in nature. The calculated Cl−Cl bond distance (2.049 Å) in the molecular complex is slightly elongated relative to that in a free Cl2 molecule (2.039 Å). As expected, the Cl−Cl stretching frequency (556.9 cm−1 in free form) is the normal mode that suffers the greatest change with the interaction in the molecular complex 4a (520.3 cm−1). This red-shift of 36.6 cm−1 is calculated for 17 cm−1 larger than the corresponding shift associated with the formation of classical π-complex 1a between benzene and chlorine (see above). Prereactive complex 4a has not been identified earlier in the chlorination of triazine, but an analogous molecular complex between 1,3,5-triazine and bromine has been described experimentally (with N−Br distance of 2.515 Å).75 As well, pyrazine, which is closely related to triazine, forms a weak linear complex with iodine (with N−I distance of 2.733 Å).76 These experimental results suggest a mechanistic relevance of prereactive complex 4a, which could precede the chlorination reaction of triazine. Two additional linear complexes 4b and 4c have been located, and both were calculated to be less stable than 4a (21.7 and 16.1 kJ/mol at the B2PLYPD//M05-2X level, respectively). The complex 4b, characterized by a close N−H···Cl contact, is structurally related to linear complexes between triazine and chloride (Cl−) anion.77,78 At finally, the T-shaped complexes 4d and 4e were successfully located, supporting the claim that this type of prereactive configurations have to be considered in description of a detailed mechanism for chlorination of heteroaromatics. They are analogous to T-shaped complexes 1f and 1g located as prereactive complexes between benzene and chlorine. Complexes 4d and 4e were calculated 0.7 and 2.1 kJ/mol, respectively, more stable than the separated reactants (Trz and Cl2) in the gas phase, but their formation becomes unfavorable in the model solvent (Table 2). When solvent effects are

Figure 2. M05-2X/AUG-cc-pVDZ optimized minima of chlorobenzenium ion complexed to BF4− and AlCl4− anions (σ-complex intermediates 2 and 3, respectively). All distances are in angstroms.

These computational results indicate that not only is the reaction rate increased with Lewis acid catalyst but also the general mechanism of the aromatic chlorination is changed. In the gas phase or in the low polarity solvent, the reaction occurs without formation of an intermediate arenium cation, whereas in the Lewis acid catalyzed chlorination, the formation of a stable σ-complex intermediate is possible. 3.4. Prereactive Complexes between Cl2 and 1,3,5Triazine. It is known that 1,3,5-triazine and its alkylated derivatives are not readily chlorinated.73,74 Therefore, it is of interest to search for prereactive complexes of these heteroaromatics with Cl2 and to compare them to those found on the reactant side of the parent reaction, i.e., the chlorination of benzene. In the case of 1,3,5-triazine, five different prereactive complexes with chlorine were located as stable minima using stochastic search at different levels of theory (Chart 5). No

Table 2. Relative Energies ΔH (in kJ/mol, at 298.15 K) and Solvation Energies ΔGsolv (in kJ/mol, at 298.15 K) for Prereactive Complexes between Chlorine (Cl2) and Triazine (Trz) Calculated at Different Levels of Theory

Chart 5. Prereactive Complexes of 1,3,5-Triazine and Chlorine Located by Stochastic Search Procedure at the M05-2X/AUG-cc-pVDZ Level

B2PLYPD// M05-2Xb

M05-2Xa

Trz + Cl2 4a 4b 4c 4d 4e

ΔH (NImag)d

ΔH + ΔGsolve

0.0

0.0

−14.8 (0) +4.2 (0) +0.1 (0) −0.1 (0) −2.0 (0)

−11.6 +8.2 +6.7 +5.1 +2.8

MP2c

ΔH + ΔGsolve

ΔH (NImag)d

ΔH + ΔGsolvf

0.0

0.0

0.0

0.0

−16.6 +3.3 −3.9 −0.7 −2.1

−13.4 +8.3 +2.7 +4.5 +2.7

−22.7 (0) −6.8 (2) −7.6 (0) −10.9 (1) −7.7 (0)

−20.3 −4.4 −4.0 −7.4 −4.1

ΔH

a

At the M05-2X/AUG-cc-pVDZ level. bAt the B2PLYP-D/AUG-ccpVTZ//M05-2X/AUG-cc-pVDZ level. cAt the MP2/AUG-cc-pVDZ level. dNumber of imaginary frequencies. ePCM/UFF/M05-2X/AUGcc-pVDZ method in the model solvent of ε = 2.228 (CCl4). fPCM/ UFF/MP2/AUG-cc-pVDZ method in the model solvent of ε = 2.228 (CCl4).

π-complex between heteroaromatic ring of triazine and chlorine molecule was located as minimum. During the optimization procedure, the overcenter (η1) and the overbond geometries (η2) converge to the linear complex 4a. The structure 4a is the most stable complex between triazine and chlorine. It is characterized by a close chlorine/nitrogen contact. The calculated N−Cl distance in 4a is 2.681 Å (at the M05-2X/AUG-cc-pVDZ level), which is notably shorter than the corresponding van der Waals radius (N−Cl distance of ca. 3.4 Å). Structural changes upon formation of this prereactive complex are consistent with

included in calculations at M05-2X and B2PLYPD//M05-2X levels, only complex 4a is more stable than the separated reactants. In contrast, formation of all complexes (4a−4e), calculated at the MP2/AUG-cc-pVDZ level, are exothermic. 1303



Article

new complexes have been located on the reactant side of the corresponding EAS coordinate. It is proposed that these intermediates, not described earlier, can precede the formation of classical π- or σ-complexes. In the case of chlorination of benzene, the T-shaped complex 1f may be present prior to the subsequent formation of the π-complex 1a. At the B2PLYP-D/ M05-2X level of theory, this intermediate is 0.8 kJ/mol more stable than the separated reactants and only 3.6 kJ/mol less stable than the global minimum structure 1a. The transition state structure TS1a−1f connecting these two complexes has been located and thus is involved in an extended continuum of structures along the EAS reaction coordinate. As expected, the energy barriers for benzene chlorination in the gas phase or in the low-polarity solvents (e.g., CCl4 in our case) are too high for preparative purposes.58,87,88 However, such conditions (i.e., in isolation) allow the inherent mechanism (described herewith) to be established. In order to evaluate the intrinsic reactivity of heteroaromatics, the uncatalyzed chlorination reactions in the gas phase and in a nonpolar solvent model have been considered. In the case of the interaction between 1,3,5-triazine and Cl2, neither π- nor σ-complexes have been located as stable minima (the same is observed in the case of tetrazine), which implies that different prereactive complexes are present at the reactant side of the corresponding EAS. Both linear 4a and T-shaped complex 4e have been identified as candidates that may be formed in the early stages of the chlorination of heteroaromatics.

Similar effects have been observed and discussed in the case of benzene/chlorine complexes (see above). 3.5. Complexes between Cl2 and 1,2,4,5-Tetrazine. 1,2,4,5-Tetrazine (or s-tetrazine) is an aromatic compound with a high-nitrogen content, being the least electron-rich class of neutral CN heterocycles.79,80 Because of the electron-deficient ring, it is not expected to undergo electrophilic aromatic substitution, but its chlorinated derivatives can easily react with nucleophiles. It has been shown that tetrazine can form different complexes with neutral molecules and anions. Linear in-plane complexes have been reported for noncovalent interactions of tetrazine with HCl and H2O,81,82 while an axial geometry has been suggested for complexes between tetrazine and anions (Cl−, F−, and Br−).83 Recently, T-shaped and resting-type complexes, which are relevant for our study, have been described for triazine dimers.84 In the case of chlorine, we have located three different complexes (Chart 6) at the M05-2X level: resting (5a), T-shaped (5b), and in-plane (5c). Chart 6. Complexes of 1,2,4,5-Triazine and Chlorine Located by Stochastic Search Procedure at the M05-2X/ AUG-cc-pVDZ Level

■

Table 3. Relative Energies ΔH (in kJ/mol, at 298.15 K) and Solvation Energies ΔGsolv (in kJ/mol, at 298.15 K) for Prereactive Complexes between Chlorine (Cl2) and Tetrazine (Ttz) Calculated at Different Levels of Theory B2PLYPD// M05-2Xb

M05-2Xa ΔH (NImag)d Ttz + Cl2 5a 5b 5c

ΔH + ΔGsolve

ΔH

ΔH + ΔGsolve

S Supporting Information *

Optimized coordinates and calculated energy data for all stationary points studied. This material is available free of charge via the Internet at http://pubs.acs.org.

■

MP2c ΔH (NImag)d

ΔH + ΔGsolvf

0.0

0.0

0.0

0.0

0.0

0.0

−4.7 (0) −0.6 (0) +2.0 (0)

+0.3 +4.5 +3.6

−9.0 −5.2 −2.2

−4.0 −0.1 −0.6

−17.0 (0) −11.9 (0) −6.3 (2)

−14.2 −8.4 −0.1

ASSOCIATED CONTENT

AUTHOR INFORMATION

Corresponding Author

*Tel: +385-1-6394441. Fax: +385-1-4856201. E-mail: valerije@ pharma.hr.

■

ACKNOWLEDGMENTS We thank the Computing Center of the University of Zagreb SRCE for allocating computer time on the Isabella cluster.

a

At the M05-2X/AUG-cc-pVDZ level. bAt the B2PLYP-D/AUG-ccpVTZ//M05-2X/AUG-cc-pVDZ level. cAt the MP2/AUG-cc-pVDZ level. dNumber of imaginary frequencies. ePCM/UFF/M05-2X/AUGcc-pVDZ method in the model solvent of ε = 2.228 (CCl4). fPCM/ UFF/MP2/AUG-cc-pVDZ method in the model solvent of ε = 2.228 (CCl4).

■

REFERENCES

(1) Ben-Daniel, R.; de Visser, S. P.; Shaik, S.; Neumann, R. J. Am. Chem. Soc. 2003, 125, 12116−12117. (2) Smith, W. B. J. Phys. Org. Chem. 2003, 16, 34−39. (3) Wei, Y.; Wang, B.-W.; Hu, S.-W.; Chu, T.-W.; Tang, L.-T.; Liu, X.-Q.; Wang, Y.; Wang, X.-Y. J. Phys. Org. Chem. 2005, 18, 625. (4) Zhang, M.; Lund, C. R. F. J. Phys. Chem. A 2002, 106, 10294− 10301. (5) Spencer, S. R.; Zhang, M.; Lund, C. R. F. J. Phys. Chem. A 2003, 107, 10335−10345. (6) Osamura, Y.; Terada, K.; Kobayashi, Y.; Okazaki, R.; Ishiyama, Y. J. Mol. Struct. 1999, 461−462, 399−416. (7) Kolboe, S.; Svelle, S. J. Phys. Chem. A 2008, 112, 6399−6400. (8) van Mouorik, T. J. Phys. Chem. A 2008, 112, 11017−11020. (9) Schwabe, T.; Grimme, S. J. Phys. Chem. A 2009, 113, 3005−3008. (10) Croft, A. K.; Howard-Jones, H. M. Phys. Chem. Chem. Phys. 2007, 9, 5649−5655. (11) Tsao, M.-L.; Hadad, C. H.; Platz, M. S. J. Am. Chem. Soc. 2003, 125, 8390−8399.

All complexes have been calculated to be more stable than the separated reactants (Table 3), but only 5a and 5b structures have been characterized as genuine minima at all computational levels employed. Although less relevant for the description of EAS in tetrazine, these structures are still of importance in the spectroscopy and photochemistry of weakly bounded complexes of tetrazine.85,86

4. CONCLUSIONS In this work, the importance of prereactive complexes between aromatic compounds (the parent benzene and high-nitrogen heterocycles triazine and tetrazine) and chlorine has been revisited. Using the stochastic search method, a number of 1304



Article

(34) Sakic, D.; Zipse, H.; Vrcek, V. Org. Biomol. Chem. 2011, 9, 4336−4346. (35) Vrcek, V.; Kronja, O.; Saunders, M. J. Chem. Theory Comput. 2007, 3, 1223−1230. (36) Bera, P. P.; Sattelmeyer, K. W.; Saunders, M.; Schaefer, H. F. III; Scheleyer, P. v. R. J. Phys. Chem. A 2006, 110, 4287−4290. (37) Matsuzawa, H.; Osamura, Y. Bull. Chem. Soc. Jpn. 1997, 70, 1531−1537. (38) Vasilyev, A. V.; Lindeman, S. V.; Kochi, J. K. Chem. Commun. 2001, 909−910. (39) Vasilyev, A. V.; Lindeman, S. V.; Kochi, J. K. New J. Chem. 2002, 26, 582−592. (40) Lenoir, D. Angew. Chem., Int. Ed. 2003, 42, 854−857. (41) Ammal, S. S. C.; Ananthavel, S. P.; Venuvanalingam, P.; Hegde, M. S. J. Phys. Chem. A 1998, 102, 532−536. (42) Mebel, A. M.; Lin, H. L.; Lin, S. H. Int. J. Quantum Chem. 1999, 72, 307318. (43) Milano, G.; Guerra, G.; Cavallo, L. Eur. J. Inorg. Chem. 1998, 1513−1517. (44) The descriptor η is the arene hapticity as defined in Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry; Wiley: New York, 1988. (45) Lee, I.; Kim, C. K.; Li, H. G.; Lee, B.-S.; Lee, H. W. Chem. Phys. Lett. 2000, 320, 307−310. (46) Olah, G. A.; Lin, H. C.; Mo, Y. K. J. Am. Chem. Soc. 1972, 94, 3667−3669. (47) Highly symmetrical π-complex (η6) 1c in which chlorine lies in an axial orientation relative to the benzene plane is characterized as a higher-order saddle point. (48) Collin, J.; D’Or, L. J. Chem. Phys. 1955, 23, 397. (49) Grozema, F. C.; Zijlstra, R. W. J.; Swart, M.; van Duijnen, P. T. Int. J. Quantum Chem. 1999, 75, 709−723. (50) Su, J. T.; Zewail, A. H. J. Phys. Chem. A 1998, 102, 4082−4099. (51) Bryantsev, V. S.; Hay, B. P. J. Am. Chem. Soc. 2005, 127, 8282− 8283. (52) Schneider, H.; Vogelhuber, K. M.; Schinle, V.; Weber, J. M. J. Am. Chem. Soc. 2007, 129, 13022−13026. (53) Thompson, C. D.; Poad, B. L. J.; Emmeluth, C.; Bieske, E. Chem. Phys. Lett. 2006, 428, 18−22. (54) Bryantsev, V. S.; Hay, B. P. Org. Lett. 2005, 7, 5031−5034. (55) Li, S.; Cooper, V. R.; Thonhauser, T.; Puzder, A.; Langreth, D. C. J. Phys. Chem. A 2008, 112, 9031−9036. (56) Singh, P. C.; Ray, M.; Patwari, G. N. J. Phys. Chem. A 2007, 111, 2772−2777. (57) Mishra, B. K.; Sathyamurthy, N. J. Phys. Chem. A 2007, 111, 2139−2147. (58) de la Marre, P. B. D. Acc. Chem. Res. 1974, 7, 361−368. (59) Lippert, J. L.; Hanna, M. V.; Trotter, P. J. J. Am. Chem. Soc. 1969, 91, 4035−4044. (60) Kortüm, G.; Walz, H. Z. Elektrochem. 1953, 57, 73. (61) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory Comput. 2006, 2, 364−382. (62) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167. (63) Hassel, O.; Stromme, K. O. Acta Chem. Scand. 1959, 13, 1781− 1786. (64) Grimme, S. J. Chem. Phys. 2003, 118, 9095. (65) Moran, D.; Simmonett, A. C.; Leach, F. E. III; Allen, W. D.; Schleyer, P. v. R.; Schaefer, H. F. III. J. Am. Chem. Soc. 2006, 128, 9342−9343. (66) Olah, G. A.; Lin, H. C.; Mo, Y. K. J. Am. Chem. Soc. 1972, 94, 3667−3669. (67) Rathore, R.; Hecht, J.; Kochi, J. K. J. Am. Chem. Soc. 1998, 120, 13278−13279. (68) As well, the transition state structure, they claimed to connect the π-complex and σ-complex, is, in fact, the transition structure that corresponds to the direct conversion of the π-complex to the cisaddition product. The IRC calculation, not performed in their study, clearly shows that this transition state connects the two minima.

(12) Kong, J.; Galabov, B.; Koleva, G.; Zou, J.-J.; Schaeffer, H. F. III; Schleyer, P. v. R. Angew. Chem., Int. Ed. 2011, 50, 6809−6813. (13) Hubig, S. M.; Kochi, J. K. J. Org. Chem. 2000, 65, 6807−6818. (14) Rosokha, S. V.; Kochi, J. K. J. Org. Chem. 2002, 67, 1727−1737. (15) Fievez, T.; Pinter, B.; Geerlings, P.; Bickelhaupt, F. M.; De Proft, F. J. Org. Chem. 2011, 16, 2958−2968. (16) Esteves, P. M.; Carneiro, J. W. M.; Cardoso, P. S.; Barbosa, A. G. H.; Laali, K. K.; Rasul, G.; Prakash, G. K. S.; Olah, G. A. J. Am. Chem. Soc. 2003, 125, 4836−4849. (17) Skokov, S.; Wheeler, R. A. J. Phys. Chem. A 1999, 103, 4261− 4269. (18) Lu, Y.-X.; Zou, J.-W.; Wang, Y.-H.; Yu, Q.-S. Int. J. Quantum Chem. 2007, 107, 1479−1486. (19) Roohi, H. J. Phys. Org. Chem. 2008, 21, 971−978. (20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision E.01; Gaussian, Inc.: Wallingford, CT, 2003. (21) 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.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision A.02; Gaussian, Inc.: Wallingford, CT, 2009. (22) For a general performanceof B3LYP methods for a large set of organic molecules, see Tirado-Rives, J.; Jorgensen, W. L. J. Chem. Theory Comput. 2008, 4, 297−306. (23) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2004, 10, 6908−6918. (24) Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory Comput. 2006, 2, 364−382. (25) Frisch, M. J.; Head-Gordon, M.; Pople, J. A. Chem. Phys. Lett. 1990, 166, 275−280. (26) Schwabe, T.; Grimme, S. Phys. Chem. Chem. Phys. 2007, 9, 3397−3406. (27) Gonzales, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523− 5527. (28) Scalmani, G.; Frisch, M. J. J. Chem. Phys. 2010, 132, 114110. (29) Saunders, M. J. Comput. Chem. 2004, 25, 621−626. (30) Saunders, M. J. Am. Chem. Soc. 1987, 109, 3150−3157. (31) Budimir, A.; Weitner, T.; Kos, I.; Sakic, D.; Gabricevic, M.; Besic, E.; Birus, M. Croat. Chem. Acta 2011, 84, 133−147. (32) Wheeler, S. E.; Schleyer, P. v. R.; Schaefer, H. F. III. J. Chem. Phys. 2007, 126, 104104. (33) Bachrach, S. M. J. Phys. Chem. A 2008, 112, 3722−3730. 1305



Article

(69) For the same reason, the transition state structure in that work, proposed to connect the π-complex and σ-complex, does not exist at the corresponding PES if the full optimization of its geometry is performed. (70) Gwantley, S. R.; Rosokha, S. V.; Head-Gordon, M.; Kochi, J. K. J. Am. Chem. Soc. 2003, 125, 3273−3283. (71) Heidrich, D. Angew. Chem., Int. Ed. 2002, 41, 3208−3210. (72) Volkov, A. N.; Timoshkin, A. Y.; Suvorov, A. V. Int. J. Quantum Chem. 2005, 104, 256−260. (73) Grundmann, C.; Kreutzberger, A. J. Am. Chem. Soc. 1955, 77, 44−48. (74) Schaefer, F. C.; Ross, J. H. J. Org. Chem. 1964, 29, 1527−1537. (75) Mascal, M.; Richardson, J. L.; Blake, A. J.; Li, W.-S. Tetrahedron Lett. 1996, 37, 3505−3506. (76) Bailey, R. D.; Buchanan, M. L.; Pennington, W. T. Acta Crystallogr. 1992, C48, 2259−2262. (77) Berryman, O. B.; Bryantsev, V. S.; Stay, D. P.; Johnson, D. W.; Hay, B. P. J. Am. Chem. Soc. 2007, 129, 48−58. (78) Mascal, M.; Armstrong, A.; Bartberger, M. D. J. Am. Chem. Soc. 2002, 124, 6274−6276. (79) Katritzky, A. R. Handbook of Heterocyclic Chemistry; Pergamon Press: Oxford, U.K., 1986; p 47. (80) Clavier, G.; Audebert, P. Chem. Rev. 2010, 110, 3299−3314. (81) Haynam, C. A.; Morter, C.; Young, L.; Levy, D. H. J. Phys. Chem. 1987, 91, 2519−2525. (82) Odelius, M.; Kirchner, B.; Hutter, J. J. Phys. Chem. A 2004, 108, 2044−2052. (83) Garau, C.; Quinonero, D.; Frontera, A.; Costa, A.; Ballester, P.; Deya, P. M. Chem. Phys. Lett. 2003, 370, 7−13. (84) Mishra, B. K.; Arey, J. S.; Sathyamurthy, N. J. Phys. Chem. A 2010, 114, 9606−9616. (85) Gong, Y.-H.; Miomandre, F.; Méallet-Renault, R.; Badré, S.; Galmiche, L.; Tang, J.; Audebert, P.; Clavier, G. Eur. J. Org. Chem. 2009, 6121−6128. (86) Audebert, P.; Miomandre, F.; Clavier, G.; Vernieres, M.-C.; Badré, S.; Méallet-Renault, R. Chem.Eur. J. 2005, 11, 5667−5673. (87) Stock, L. M.; Himoe, A. J. Am. Chem. Soc. 1961, 83, 4605−4609. (88) Brown, H. C.; Stock, L. M. J. Am. Chem. Soc. 1957, 79, 5175− 5179.

1306


Prereactive Complexes in Chlorination of Benzene, Triazine, and

Recommend Documents