Arenium ions are not obligatory intermediates in electrophilic aromatic substitution

Article abstract of DOI:10.1073/pnas.1405065111

Our computational and experimental investigation of the reaction of anisole with Cl₂ in nonpolar CCl₄ solution challenges two fundamental tenets of the traditional S_EAr (arenium ion) mechanism of aromatic electrophilic substitution. Instead of this direct substitution process, the alternative addition-elimination (AE) pathway is favored energetically. This AE mechanism rationalizes the preferred ortho and para substitution orientation of anisole easily. Moreover, neither the S _EAr nor the AE mechanisms involve the formation of a σ-complex (Wheland-type) intermediate in the rate-controlling stage. Contrary to the conventional interpretations, the substitution (S_EAr) mechanism proceeds concertedly via a single transition state. Experimental NMR investigations of the anisole chlorination reaction course at various temperatures reveal the formation of tetrachloro addition by-products and thus support the computed addition-elimination mechanism of anisole chlorination in nonpolar media. The important autocatalytic effect of the HCl reaction product was confirmed by spectroscopic (UV-visible) investigations and by HCl-augmented computational modeling.

Full text of DOI:10.1073/pnas.1405065111

Arenium ions are not obligatory intermediates in
electrophilic aromatic substitution
Boris Galabova,1, Gergana Kolevaa, Svetlana Simovab, Boriana Hadjievaa, Henry F. Schaefer IIIc,1
and Paul von Ragué Schleyerc,1
,
aDepartment of Chemistry and Pharmacy, University of Sofia, Sofia 1164, Bulgaria; bInstitute of Organic Chemistry with Centre of Phytochemistry, Laboratory
Nuclear Magnetic Resonance, Bulgarian Academy of Sciences, Sofia 1113, Bulgaria; and cCenter for Computational Quantum Chemistry and Department of
Chemistry, University of Georgia, Athens, GA 30602
Edited* by Martin Saunders, Yale University, New Haven, CT, and approved May 28, 2014 (received for review March 18, 2014)
computed at some levels for the related arene nitrosation, but
reaction medium and counter ion effects were not considered.
Gwaltney et al. (28) reported a single concerted transition state
after reoptimizing all saddle points at CCSD(T)/6-31G(d,p) and
modeling bulk solvation by the Onsager approximation, and
Rzepa (www.ch.imperial.ac.uk/rzepa/blog/?p=2423, accessed
March 10, 2013) also found a concerted transition state including
a trifluoroacetate counterion. Instead, one-step reactions via
single transition states take place (Fig. 2). Our experimental
investigations of the chlorination of anisole in CCl4 solution
revealed tetrachloro by-products, which must have arisen by fur-
ther reaction of intermediate dichloro-adducts. Both our UV-vis-
ible (UV-VIS) spectroscopic investigation and our theoretical
modeling of this reaction clearly verified the autocatalytic effect of
the HCl by-product, in harmony with Andrews and Keefer’s (29,
30) early experimental kinetic studies of the chlorination of arenes,
which found that HCl reduces the activation barriers significantly.
We also applied reliable theoretical methods to model a typical
experimental example of the highly investigated SEAr electro-
philic aromatic halogenations, the electrophilic chlorination of
anisole by molecular chlorine in simulated CCl4 solution (15, 16).
Although the elucidation of the classic SEAr mechanism [Fig. 1,
involving the initial formation of a π-complex, followed by a
transition state leading to a σ-complex (arenium) intermediate in
the rate-controlling stage, and, finally, proton loss from the ipso-
position leading to the reaction product] is considered to be
a triumph of physical organic chemistry (1, 31–37), an alternative
addition–elimination pathway leading to substitution products has
Our computational and experimental investigation of the reaction
of anisole with Cl2 in nonpolar CCl4 solution challenges two funda-
mental tenets of the traditional SEAr (arenium ion) mechanism of
aromatic electrophilic substitution. Instead of this direct substitution
process, the alternative addition–elimination (AE) pathway is fa-
vored energetically. This AE mechanism rationalizes the preferred
ortho and para substitution orientation of anisole easily. Moreover,
neither the SEAr nor the AE mechanisms involve the formation of
a σ-complex (Wheland-type) intermediate in the rate-controlling
stage. Contrary to the conventional interpretations, the substitution
(SEAr) mechanism proceeds concertedly via a single transition state.
Experimental NMR investigations of the anisole chlorination reac-
tion course at various temperatures reveal the formation of tetra-
chloro addition by-products and thus support the computed
addition–elimination mechanism of anisole chlorination in nonpolar
media. The important autocatalytic effect of the HCl reaction product
was confirmed by spectroscopic (UV-visible) investigations and by
HCl-augmented computational modeling.
organic chemistry reaction mechanisms alternative computed routes
|
|
|
autocatalysis
nterest in the chemistry of electrophilic aromatic substitution
reactions continues because of their widespread application for
I
the production of a great variety of chemicals and materials (1–
4). Electrophilic substitution, considered to be the most char-
acteristic reaction of aromatic systems, is typically described in
textbooks, monographs, and reviews by the two-stage SEAr
mechanism depicted in Fig. 1 (5–11). Arenium ion (σ-complex)
intermediates are often ascribed to Wheland (9) inaccurately,
since Pfeiffer and Wizinger (10) laid out the principles of such
species for bromination in 1928. Following Brown and Pearsall
(11), they are widely believed to have σ-complex structures.
Arenium ions (σ-complexes) (9–11) are widely accepted to be
obligatory intermediates and are used to rationalize ortho/para
vs. meta position orientation preferences (6–11).
Significance
Electrophilic substitution is universally regarded as the charac-
teristic reaction of aromatic compounds. Arenium ions are widely
accepted as obligatory intermediates in the two-stage (SEAr)
mechanism typically described in textbooks, monographs, and
reviews. We now challenge this mechanistic paradigm. Our
combined computational and experimental investigation of the
exemplary halogenation of anisole with Cl2 in CCl4 finds that
addition–elimination pathways compete with the direct sub-
stitution process and also account for the ortho–para orientation
preferences easily. Moreover, the SEAr processes do not involve
arenium ion pair intermediates, but proceed instead via con-
certed one-stage single transition state routes. We question not
only the generality of the accepted SEAr mechanism, but also the
involvement of arenium ion intermediates, when counter ions
are present.
We now reinforce our challenges (12, 13) of this conventional
“reaction mechanism paradigm” (14) by a combined computa-
tional and experimental study of the facile chlorination of anisole
(methoxybenzene) with Cl2 in CCl4 solution (15, 16). We find that
Fig. 1 is not the favored pathway. Instead, addition reactions of
Cl2 to anisole have the lowest activation energies (Fig. 2). Ready
HCl elimination from the initially formed adducts leads to ortho-
and para-chloroanisole as the predominate products. This addi-
tion–elimination (AE) mechanism (the historical antecedent to
Fig. 1) (17–26) predicts the same positional orientation as the
usually assumed direct substitution (“SEAr”) alternative. Instead
of this classic SEAr mechanism (Fig. 1), we find that direct con-
certed substitution, not involving an arenium ion, σ-complex
(“Wheland”) (9–11) intermediate, competes energetically with the
AE route. Like some earlier computational studies on aromatic
substitution (12, 13, 27, 28) (Rzepa H, www.ch.imperial.ac.uk/
rzepa/blog/?p=2423, accessed March 10, 2013), our study finds no
such intermediates in the direct substitution of anisole by Cl2. A
concerted mechanism without an arenium ion intermediate was
Author contributions: B.G., H.F.S., and P.v.R.S. designed research; G.K., S.S., and B.H.
performed research; B.G., G.K., S.S., B.H., H.F.S., and P.v.R.S. analyzed data; B.G., H.F.S.,
and P.v.R.S. wrote the paper; and S.S. described the NMR results.
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
1To whom correspondence may be addressed. E-mail: galabov@chem.uni-sofia.bg, ccq@
uga.edu, or schleyer@uga.edu.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1405065111/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1405065111
PNAS
|
July 15, 2014
|
vol. 111
|
no. 28
|
10067–10072

Fig. 1. Typical depiction of the arenium ion mechanism for SEAr reactions.
been discussed since the 19th century (19–26, 38, 39). Neverthe-
precedes the formation of an intermediate, which they interpreted
less, it is commonly believed that the classic multistep SEAr to be a σ-complex. Instead, the structure of this complex represents
mechanism involving the formation of a σ-complex intermediate a protonated iodobenzene. Volkov et al.’s MP2/LANL2DZ(d)+
in the rate-controlling stage is the only mechanistic route to ar- study (43) of the chlorination of benzene established that dimers of
omatic substitution products. Our present and previous (12, 13) group 13 metal halides catalyzed the processes more effectively.
results challenge the generality of such traditional interpretations. Optimized geometries of π- and σ-complexes as well as transition
Although the initial stages of the alternative AE route seem un- structures were reported. Theoretical investigations by Ben-Daniel
attractive because aromaticity is lost, many arenes are known et al. (44) and by Filimonov et al. (45) of the chlorination of benzene
experimentally to give addition products in considerable amounts with Cl2 (and other related processes) reported structural details of
(19–26, 38, 39). Thus, de la Mare (21, 25, 38, 39) demonstrated transition states purported to lead to the chlorobenzene product.
the formation of halogen adduct intermediates. Polybenzenoid Our reinvestigations revealed errors in major suppositions of
hydrocarbons (PBHs) react with halogens to give isolable addition both these studies. Our IRC computations show clearly that the
products, which then give substitution products easily by hydro- transition states in question lead to 1,2 Cl2–benzene addition
gen halide elimination (23). Our computational investigations of products (rather than to chlorobenzene). Zhang and Lund (46)
arene bromination with molecular bromine (12) and sulfonation investigated the neat chlorination of toluene by Cl2 experi-
with SO3 (13) provided clear evidence that the mechanisms of the mentally and theoretically at B3LYP/cc-pVTZ(-f) [cc-pVTZ
inherent substitution reactions (i.e., uncatalyzed, gas phase, or (-f), correlation consistent polarized triple-zeta without f-func-
weakly solvated) are concerted and do not involve the conven- tions basis set]. Although we verified their reported geometry of
tional σ-complex (or any other) intermediates. Moreover, the the concerted transition state (figure 6 in ref. 46), our stability
energetics of the bromination processes document the significance check revealed that its wavefunction is unstable. This casts doubt
of competition between AE and direct substitution mechanisms on their conclusions because of the homolysis vs. heterolysis
leading to the same substitution products. Thus, the computed issues. In contrast, all wavefunctions in our paper were checked
barrier in a simulated nonpolar (CCl4) medium is 4 kcal/mol lower
and all are stable. Most prior theoretical studies of SEAr halo-
for Br2 addition to benzene (followed by HBr elimination) than genations did not consider the connections between transition
that for the direct substitution pathway to bromobenzene (12).
Previous theoretical studies of electrophilic aromatic haloge-
nation processes have been based on the classic SEAr mecha-
nism, involving arenium ion intermediates (Fig. 1). Osamura et al.’s
(40) Hartree-Fock computations of the AlCl3-catalyzed electro-
philic aromatic chlorination mechanism found an initial π-com-
plex, a transition state preceding the intermediate σ-complex,
and a second transition state leading to final products. Alumi-
num chloride was important as a Lewis acid catalyst throughout
the process. AlCl3 coordination polarizes Cl2 and thereby assists
its reaction with the arene. Rasokha and Kochi (41) considered
the interaction of Br2 with benzene and toluene in detail in their
survey of theoretical and experimental data on the prereactive
charge-transfer complexes in electrophilic aromatic substitu-
tions. They argued that the structures and properties of the
prereactive complexes provide important mechanistic insights for
the SEAr reactions. Wei et al.’s (42) theoretical study of the
iodination of anisole by iodine monochloride at the B3LYP/6-
311G* and MP2//B3LYP/6-311G* levels (B3LYP, Becke’s three
parameter hybrid functional, using the Lee-Yang-Parr correla-
tion functional; MP2, second order Møller-Plesset perturbation
theory computations) found that the highest energy transition state
states, intermediates, and products explicitly, as we have done.
Experimental findings not always have been in accord with the
prevailing mechanistic assumption for aromatic halogenation: that
arenium ion formation is the rate-limiting step. Thus, Olah et al.
(47), Kochi and coworkers (48), and Fukuzumi and Kochi (49) have
emphasized that substrate and positional selectivity are inconsistent
(e.g., low toluene/benzene reactivity ratios but high toluene ortho–
para vs. meta regiospecificity) for some electrophiles under certain
conditions. This disparity indicates the existence of at least one
other mechanistic pathway. It has been suggested that π-complexes
may control product formation. Olah et al.’s (47) kinetics of the
ferric chloride-catalyzed bromination of benzene and alkyl ben-
zenes provided strong evidence for low substrate selectivity in the
rate-determining step, which precedes the formation of a σ-com-
plex intermediate (Fig. 1). High positional selectivity is governed by
the transition state associated with the second step of the reaction.
However, our earlier study (50) examined the possible partici-
pation of π-complexes in the key mechanistic steps of SEAr bro-
mination reactions in detail but found no link between the energy of
formation of these complexes and the overall reactivity. Although
there is no doubt that π-complexes form easily (via essentially
barrierless processes) in most SEAr reactions after mixing the
Fig. 2. The HCl-catalyzed concerted and addition–elimination pathways of para-chlorination of anisole in nonpolar media.
10068
|
www.pnas.org/cgi/doi/10.1073/pnas.1405065111
Galabov et al.

electrophile and the aromatic substrate, it is unlikely that these Methods
low-energy “bystander” structures influence rates of SEAr reac-
tions significantly. Thus, the lack of accord between substrate and
positional selectivity, established by Olah et al. (47), Kochi and
coworkers (48), and Fukuzumi and Kochi (49) may be due to
other mechanistic differences. De la Mare and Bolton (21) and de
la Mare (51) have stressed the plurality of aromatic substitution
mechanisms, depending on the substrate and the conditions.
Reactive substrates are known to undergo uncatalyzed aromatic
substitution in nonpolar solvents at room temperature. Thus, our
computational investigations modeled Watson’s careful experiments
on the chlorination of anisole in CCl4 at 25 °C (15, 16). His low con-
version (25%) conditions for chlorophenol permitted more accurate
determination of the initial product ratios (and avoided further Cl2
additions to 4-chloroanisole, which ultimately gave 1,3,4,5,6-penta-
chloro-4-methoxycyclohexene). After introduction of gaseous Cl2 into
a CCl4 solution of anisole for 1 h, the products were 4-chloroanisole
(76%), 2-chloroanisole (13.6%), 2,6-dichloro anisole (2.1%), 2,4-
dichloroanisole (3.0%), and 2,4,6-trichloroanisole (0.4%).
Our Gaussian 09 (54) computations used the 6-311+G(2d,2p) basis set (55, 56)
with the B3LYP hybrid functional (57–59) and the Perdew–Burke–Ernzerhof
(PBE) functional (60, 61) augmented with Grimme et al.’s (62) density functional
theory with added Grimme’s D3 dispersion corrections (DFT-D3). Single-point
energies of all optimized structures were obtained with the B2-PLYP [double-
hybrid density functional of Grimme (63)] and applying the D3 dispersion cor-
rections. Intrinsic reaction coordinate (IRC) (64) computations connected the
critical structures along the reaction path: reactants, transition states, inter-
mediates, and products. The integral equation formalism of the polarizable
continuum model (IEF-PCM) method (65) modeled solvent effects. Unrestricted
broken symmetry (UBS B3LYP) computations on saddle-point structures along
the potential energy surfaces gave the same energies as those from the re-
stricted B3LYP method. Therefore, no diradical structures were involved along
the reaction pathways investigated. All saddle-point wavefunctions were stable.
Extensive further testing assessed the methodological integrity of the B3LYP
computations. We used the PBE method, which includes no contribution from
exact exchange and thus is very different from B3LYP. All transition states were
reoptimized with PBE and all intrinsic reaction coordinates were recomputed. As
discussed below, the PBE results are qualitatively similar to those at B3LYP,
strengthening the reliability of both methods for the present studies.
Analogous chlorinations of phenol, 2-methylphenol, and 2-
chlorophenol in CCl4 also have been carried out with high con-
version rates at the reflux temperature (79 °C) (16). Chlorination
of phenol with Cl2 in CCl4 has been reported by other groups
(52, 53).
Our theoretical findings were supported by the results of parallel ex-
perimental investigations of chlorination reactions by NMR monitoring: 1H
(600.13 MHz) and 13C (150.92 MHz) spectra were acquired during the course
of the reactions. Unambiguous assignments of the signals were based on 1D
Table 1. Relative Gibbs free energies (Gibbs+Edisp) of species involved in the autocatalyzed by HCl anisole–Cl2 reaction at B3LYP/6-
311+G(2d,2p), PBE/6-311+G(2d,2p), and RB2PLYP/6-311+G(2d,2p)//RB3LYP/6-311+G(2d,2p) and for the uncatalyzed reaction at
RB2PLYP/6-311+G(2d,2p)
B3LYP*, kcal/mol,
catalyzed, gas phase
B3LYP*, kcal/mol,
catalyzed, CCl4
PBEPBE*, kcal/mol, B2PLYP//B3LYP*, kcal/mol, B2PLYP†, kcal/mol,
Species
Description
catalyzed, CCl4
catalyzed, CCl4
uncatalyzed, CCl4
π-ortho
π-para
TS1
TS2
TS3
2.63
2.76
2.81
2.86
22.24
1.20
1.53
15.97
25.09
14.96
17.72 cis add.
24.58 trans add.
13.15
2.98
3.26
23.71
2.99
3.23
o substitution
m substitution
p substitution
2,3 addition
32.19
39.65
30.84
32.97 cis add.
39.19 cis add.
30.07
40.61
29.24
25.48
35.11
17.15
35.56 trans eli.
13.27
23.31
—
—
32.14
34.58
—
20.99‡
21.00 trans add.
30.51 trans add.
17.78
22.56§
22.08 trans add.
32.31 trans add.
18.93
—
TS4
TS4meta
TS5
TS5meta
TS6
34.75 cis add.
—
{
2,3 addition
3,4-cis addition
3,4-cis addition
2,3-cis–trans
3,4-cis–trans
3,4-cis–trans
2,3-cis HCl eli.
2,3 HCl eli.
3,4-cis HCl eli.
3,4-cis HCl eli.
cis–cis rot.
trans–trans rot.
o product
m product
30.65
—
{
31.73
25.26
33.85
—
14.66
—
—
TS7
TS7meta
TS8
TS8meta
TS9
TS9meta
TS10
TS11
P1
14.49§
26.96
10.44
20.13 cis eli.
6.60
19.83
10.61
8.08
−21.52
−22.80
−23.63
—
11.73§
22.78
14.80§
28.15
11.51
21.55 cis eli.
7.43
21.49
7.73
5.65
−24.16
−25.26
−26.08
—
21.21
—
{
{
{
6.35
14.05 cis eli.
3.36
13.41
8.20
16.85
25.86 cis eli.
12.73
24.73
5.21
7.48
5.36
2.20
−21.45
−23.35
−23.47
7.60
−22.75
−23.46
−23.21
3.15
1.01
4.47
3.52
−1.02
2.56
−0.32
2.06
−0.69
−27.47
−28.37
−28.28
—
P2
P3
P4
P5′
p product
2,3-cis adduct
3,4-cis adduct
3,4-cis adduct
3,4-cis adduct
2,3-trans adduct
2,3-trans adduct
3,4-trans adduct
3,4-trans adduct
3,4-trans adduct
—
5.50
8.26
7.08
2.42
6.12
1.53
5.18
2.31
2.50
5.57
3.81
−0.33
3.69
−0.89
3.09
−0.28
−0.42
3.01
P5′′
P5meta
8.47
6.05
2.10
k
—
P6
P6meta
−3.36
—
k
—
P7′
1.15
5.11
2.00
0.09
−3.54
—
P7′′
P7meta
k
rot., rotation; eli., elimination; add., addition.
*Energies are relative to anisole, Cl2, and HCl, except for the π-complexes, where the energies are relative to anisole and Cl2.
†Only data for stable wave functions obtained for the uncatalyzed chlorination are shown in this column (see text for additional explanations). Energies are
relative to anisole and Cl2.
‡From B3LYP/6-311+G(2d,2p)//B3LYP/6-31+G(d,p) computations.
§Δ(G+Edisp) was determined from IEF-PCM (integral equation formalism of the polarizable continuum model) single-point computations at the gas-phase
optimized structure.
{High-energy transition states, leading to formation of adducts, in which the orientation of the HCl leads to formation of m-substituted product;
kHigh-energy adducts, leading to m-substituted product formation.
Galabov et al.
PNAS
|
July 15, 2014
|
vol. 111
|
no. 28
|
10069

Fig. 3. Computed PES for anisole–Cl2 reactions,
catalyzed by HCl, in simulated CCl4 solution at
B2-PLYP+D3/6-311+G(2d,2p)//B3LYP/6-311+G(2d,2p).
Cis–trans isomerization and elimination transition
states, leading to formation of m-substituted prod-
uct, are not included.
and 2D gradient-enhanced versions of correlation spectroscopy (COSY), total
correlation spectroscopy (TOCSY), nuclear Overhauser and exchange spec-
troscopy (NOESY), heteronuclear single quantum coherence (HSQC) and
heteronuclear multiple bond correlation (HMBC) experiments. Full details of the
NMR experiments are provided in SI Appendix. The catalytic effect of HCl was
confirmed experimentally by monitoring the reaction of anisole with Cl2 in
CCl4, using UV-VIS spectroscopy.
obtained. Note that the barriers for the rate-controlling stages of
the catalyzed AE pathways are more than 10 kcal/mol lower than
the barriers of the respective uncatalyzed processes.
Fig. 3 shows the evaluated potential energy surfaces and SI
Appendix, Fig. S3 illustrates key PES saddle-point structures in
simulated CCl4 media. Fig. 3 reveals two extraordinary features. In
contrast to traditional interpretations, the direct substitution
pathways (both at the ortho- and at the para-positions) proceed
through single concerted transition states (TS1 and TS3) leading
directly to products. No σ-complex arenium ion intermediates are
involved in the reaction path, even as ion pairs. Low-energy
π-complexes are formed at the initial step. The IRC computations
backward from the first transition states end at three-component
Results and Discussion
Our computations of the potential energy surfaces (PES) of HCl-
catalyzed anisole–Cl2 reactions in simulated CCl4 solution at
B2-PLYP-D3/6-311+G(2d,2p)//B3LYP/6-311+G(2d,2p) model the
experimental conditions closely (15, 16). Four competing processes
were examined: direct substitutions at ortho [transition state (TS1)] π-complexes (with free energies of formation about 7 kcal/mol
higher than the initial reactants). These three-component π-com-
plexes decompose easily into the usual two-component Cl2–anisole
complexes that have about 4 kcal/mol lower energy (Table 1). The
forward IRC computations from the concerted transition states
(TS1 and TS3) lead to product complexes, which then transform
into ortho- and para-substitution products. The results show that
para-position attack is favored energetically (TS3 energy, 22.56
kcal/mol) over that of the ortho position (TS1 energy, 23.7
kcal/mol). As expected, the meta-substitution barrier (TS2, 34.6
kcal/mol) is distinctly higher than that of ortho and para. The
structures of these direct substitution transition states (SI Ap-
pendix, Fig. S3) show no significant bond length changes in the
aromatic ring. The CC bond lengths neighboring the position of
attack are elongated only slightly, and the ring preserves its aro-
matic character. Thus, these transition states do not have are-
nium-type cyclohexadienyl geometries.
and para (TS3) positions, as well as 2,3- and 3,4-cis additions of Cl2.
The vicinal 2,3- and 3,4-dichloro adducts undergo cis–trans isom-
erization followed by HCl elimination to give ortho- and para-sub-
stitution products. The potential energy surfaces obtained after full
optimization of all structures and comprehensive IRC computa-
tions at the B3LYP/6-311+G(2d,2p) DFT level are presented in SI
Appendix, Figs. S1 and S2. Comparisons of the energies of critical
structures along the reaction paths in isolation (gas phase) and in
simulated CCl4 solvent (Table 1) show that the bulk solvent effect
(despite low polarity of CCl4) is quite significant for some reaction
stages. The catalytic effect of HCl is illustrated by the theoretically
evaluated transitions state energies for uncatalyzed and catalyzed
processes (Table 1). The comparison was possible only for the AE
routes because no stable wave functions for several TS energies (see
Table 1) of the uncatalyzed direct substitution processes could be
The results reveal that AE routes are favored over direct
substitution. The most pronounced difference (3.6 kcal/mol
lower barrier) is for the processes leading to p-chloroanisole, the
principal reaction product (Fig. 3). The AE route also is favored,
although to a lesser extent (1.6 kcal/mol), for ortho substitution.
Overall, the data obtained suggest that AE pathways dominate
over concerted substitution in anisole chlorination.
The structures of the rate-determining addition transition states
(TS4, TS5) are highly unsymmetrical with one C–Cl bond almost
formed, whereas the other chlorine atom is still quite far from its
eventual bonding site (SI Appendix, Fig. S3). The HCl catalyst
facilitates the addition of the second chlorine atom to the ring
(TS5; Fig. 2 and SI Appendix, Fig. S3; see Table 1), by lowering the
transition state electronic energy considerably, compared to the
uncatalyzed process. The subsequent barriers (TS7–TS9) along
the AE routes shown in Fig. 3 have lower energies. The cis–trans
isomerization process (TS7) is preceded and followed by the facile
Table 2. Ratios (in percent, relative to unreacted anisole) of
reaction products from the chlorination of anisole in CCl4 as
determined by NMR
1
2
3
4
5
Mixture 1*
Mixture 2†
Mixture 3‡
Mixture 4§
100
100
100
100
12
74
115
114
1.0
3.7
15.5
14.0
0.8
2.4
5.8
5.2
0.6
3.0
7.4
7.2
1, anisole; 2, p-chloroanisole; 3, o-chloroanisole; 4 and 5, by-products
formed during the chlorination of anisole in CCl4 (details in SI Appendix).
*0.01 mol/L anisole + Cl2 (25 °C).
†0.05 mol/L anisole + Cl2 (25 °C).
‡0.1 mol/L anisole + Cl2 (25 °C).
§0.1 mol/L anisole + Cl2 (5 °C).
10070
|
www.pnas.org/cgi/doi/10.1073/pnas.1405065111
Galabov et al.

Fig. 4. Orienting methoxy group-carbocation
conjugation effects explaining the HCl elimination
preferences leading to (A) ortho and (B) para (rather
than to meta) substitution products of anisole.
rotation of the methoxy groups (TS10 and TS11; SI Appendix, Fig. 1%, relative to the remaining unreacted anisole, Table 2). Three
S3). The 3,4-addition–elimination pathway yielding para-chlor-
oanisole (P3 via TS7, P7, and TS9) is preferred over the ortho-
diastereomers are possible for the addition product 4 that could
not be distinguished on the basis of the experimental NMR data
chloroanisole (P1) AE sequence (via P6 and TS8). The result is in (details in SI Appendix). The structure shown in SI Appendix, Fig.
qualitative harmony with the experimentally observed greater S4 has the lowest energy among the three possible diastereomers
para- than ortho-chloroanisole product yields (Table 2).
according to B2PLYP+D3/6-311+G(2d,2p) computations (SI Ap-
pendix, Table S3). The initial concentration of anisole as well as the
mixing procedure of the components in CCl4 seems to influence
the product ratios, but 2 always predominated. Although there
were a number of unidentified products in smaller amounts, the
NMR data do not reveal with certainty the presence of a detectable
amount of any dichloro product formed by addition of only one
molecule of chlorine to anisole.
However, our NMR results provide compelling evidence that
addition reactions do take place during the interaction of anisole
with molecular chlorine in CCl4 solution. The tetrachlorinated
products (4 and 5, SI Appendix, Fig. S4) we identified can be ob-
tained only following initial additions of a single Cl2 to anisole.
The very low barriers for the elimination processes [7.43 kcal/mol
for the 3,4-cis HCl elimination (TS9) and 11.51 kcal/mol for the 2,3-
cis HCl elimination (TS8); Fig. 3], however, favor the pathways to
substitution products, thus obviating the chance to detect the in-
termediate Cl2 monoaddition products. In contrast, the tetrachloro
addition products we obtained are quite stable and do not change
during the course of the reaction. The NMR findings provide clear
experimental confirmation that Cl2 addition processes do take
place during anisole chlorination, in remarkable agreement with
the reported theoretical results on the reaction mechanism.
The catalytic effect of the HCl reaction product also was con-
firmed experimentally. We followed the reaction using UV-VIS
The AE mechanism explains the ortho and para orientation effect
of the electron-donating methoxy group simply by the expected
preferential stabilization of the transition states for HCl elimination,
as illustrated in Fig. 4. Thus, the partial positive charges arising
during HCl elimination (TS8 and TS9; SI Appendix, Fig. S3) are
stabilized by resonance conjugation with the methoxy group only in
one of the two alternative TSs. In Fig. 4, the favored processes in-
volve initial loss of the chlorides labeled Cl2. Thus, HCl elimination
from the 2,3-trans dichoro anisole adduct (P6) gives ortho- rather
than meta-chloroanisole (Fig. 4A). Preferential HCl elimination
from the 3,4-trans dichoro anisole adduct (P7) gives para- rather
than meta-chloroanisole (Fig. 4B). Indeed, only traces of meta-
substitution product were found in the Watson experiments (15, 16).
These qualitative interpretations are clearly supported by the
computed energies of the HCl elimination processes (Table 1).
The low-lying relative energy of TS9 for 3,4 elimination is just
7.43 kcal/mol (B2PLYP//B3LYP; Fig. 3), whereas the energies of
the two transition states leading to meta-substituted products are
21.55 kcal/mol (TS8meta) and 21.49 (TS9meta). As shown in Fig. 3,
the energy of TS8 is 11.51 kcal/mol for 2,3 elimination leading
to the ortho-substitution product.
The results of the alternative PBE functional computations
(Table 1) correspond well with the B2PLYP//B3LYP/6-311+
G(2d,2p) data. Whereas the PBE/6-311+G(2d,2p) transition
state energy estimates are typically lower than those at B2PLYP+ spectroscopy at 25 °C in CCl4 solution containing a large excess of
D//B3LYP/6-311+G(2d,2p), the two sets of TS energies corre-
late well (coefficient r = 0.973) and support the mechanistic
conclusions reached for anisole chlorination. Notably, these
two quite different types of DFT functionals yield quite similar
overall potential energy surface features.
anisole with respect to the chlorinating agent (at least 6:1 molar
ratio). The consumption of the Cl2 reactant was monitored by
means of its absorption band at 330 nm. The course of the re-
action in the absence of excess HCl is illustrated in SI Appendix,
Fig. S5A. SI Appendix, Fig. S5B shows the pronounced catalytic
effect when extra HCl was added to the CCl4 solvent before in-
troducing the Cl2 reactant: The consumption of Cl2 in the pres-
ence of extra catalyst is at least five times faster. As discussed
already, these results are in harmony with the early kinetic findings
of Andrews and Keefer for the chlorination of aromatic hydro-
carbons in different solvents (29, 30). These authors established
that the chlorination of pentamethylbenzene in CCl4 is first order
in hydrocarbon and chlorine and mixed first and second order with
respect to HCl (30). The chlorination of toluene with molecular
chlorine in ethylene dichloride was found to be first order
with respect to hydrocarbon, chlorine, and HCl (29). In our
experiments the rate enhancement observed in case B was due
simply to the availability of more catalyst. The energetics of the
reaction, however, are identical for both cases A and B and
reflect a catalyzed process, pure autocatalysis in case A and
The theoretical results were verified by a series of NMR
studies of the reaction of anisole and Cl2 in CCl4 solution, using
Watson’s conditions (15, 16). Moreover, because the reaction at
25 °C was quite fast, we carried out in situ experiments at 5 °C
1
and at −10 °C directly in the NMR tubes by making H NMR
measurements at intervals over extended periods of time. Full
details of the experiments are provided in SI Appendix. As stated
above, various 1D and 2D NMR (1H and 13C) techniques were
used to analyze the mixture of reaction products. The reaction
with Cl2 at 25 °C (mixtures 3 and 4, Table 2) is complete in less
than 5 min. Further product composition does not change with
time, followed up to 1 wk. In SI Appendix, Figs. S6–S15 display
representative NMR spectra of the product mixtures.
More than 25 products from the partial reaction of anisole (1)
with Cl2 were detected in the mixtures by means of the CH3O-
group 1H and 13C peaks present in the HSQC spectra in the 3.4–4.0 combined autocatalytic and added catalyst effects in case B.
ppm proton and 50–60 ppm carbon chemical shift ranges. The
following compounds were identified: substitution products para-
chloroanisole (2) and ortho-chloroanisole (3) and the double Cl2
addition products, 4 and 5 (both of which have four chlorines). The
relative ratios of the reaction products were determined (within
Our new theoretical and experimental investigations of the
pathways and energies of anisole–Cl2 reactions provide dramatic
insights into electrophilic aromatic substitution mechanisms. The
computational results reveal the involvement of two alternative
reaction pathways, an addition–elimination sequence and a direct
Galabov et al.
PNAS
|
July 15, 2014
|
vol. 111
|
no. 28
|
10071

concerted mechanism. The addition–elimination pathways are
more favorable (the barriers are lower by 1.6–3.6 kcal/mol) and
also result in the well-known orientational preference for or-
tho- and para-substituted anisole derivatives. The ortho/para
orientation effect of the methoxy substituent is the natural
consequence of the preferential stabilizing influence of the
OCH3 group. Moreover, no arenium σ-complex (or any other
similar intermediate) is found along the reaction path for direct
substitution. Experimentally, NMR elucidation of the anisole
chlorination products provides clear evidence for the formation
of addition products during the process, in harmony with the
theoretical predictions. Autocatalysis by the HCl reaction
product leads to significant lowering of reaction barriers.
ACKNOWLEDGMENTS. The research in Georgia was supported by the US
National Science Foundation, Grants CHE-1361178 (to H.F.S.) and CHE-
1057466 (to P.v.R.S.). The research in Sofia was supported by European
Union Grant FP7-REGPOT-2011-1 (project BeyondEverest) and Bulgarian
National Science Fund, Grant DRNF02-13/2009.
1. Godula K, Sames D (2006) C-H bond functionalization in complex organic synthesis.
Science 312(5770):67–72.
35. Olah GA (1971) Aromatic substitution. XXVIII. Mechanism of electrophilic aromatic
substitutions. Acc Chem Res 4(7):240–248.
2. Allemann O, Duttwyler S, Romanato P, Baldridge KK, Siegel JS (2011) Proton-catalyzed,
silane-fueled Friedel-Crafts coupling of fluoroarenes. Science 332(6029):574–577.
3. Phipps RJ, Gaunt MJ (2009) A meta-selective copper-catalyzed C–H bond arylation.
Science 323(5921):1593–1597.
4. Dong C, et al. (2005) Tryptophan 7-halogenase (PrnA) structure suggests a mechanism
for regioselective chlorination. Science 309(5744):2216–2219.
36. Ridd JH (1971) Mechanism of aromatic nitration. Acc Chem Res 4(7):248–253.
37. Olah GA, Malhotra R, Narang SC (1989) Nitration. Structure and Mechanisms (VCH,
New York).
38. De la Mare PBD, Johnson MD, Lomas JS, del Olmo VS (1966) The kinetics and mech-
anisms of aromatic halogen substitution. Part XXIV. The stereochemistry of addition
of chlorine to naphthalene. J Chem Soc B 827–833.
5. Taylor R (1990) Electrophilic Aromatic Substitution (Wiley, New York).
6. Brouwer DM, Mackor EL, MacLean C (1970) Carbonium Ions, eds Olah GA, Schleyer
PvR (Wiley-Interscience, New York), Vol 2.
39. De la Mare PBD, et al. (1969) Kinetics and mechanisms of aromatic halogen sub-
stitution. Part XXVIII. Environmental effects on rates and products of chlorination
of some aromatic hydrocarbons. J Chem Soc B 717–724.
7. Koptyug VA (1984) Arenium Ions-Structure and Reactivity (Springer, Heidelberg), pp 1–227.
8. Hubig SM, Kochi JK (2000) Structure and dynamics of reactive intermediates in re-
action mechanisms. σ- and π-complexes in electrophilic aromatic substitutions. J Org
Chem 65(21):6807–6818.
9. Wheland GW (1942) A quantum mechanical investigation of the orientation of sub-
stituents in aromatic molecules. J Am Chem Soc 64(4):900–908.
40. Osamura Y, Terada K, Kobayashi Y, Okazaki R, Ishiyama Y (1999) A molecular orbital
study of the mechanism of chlorination reaction of benzene catalyzed by Lewis acid.
J Mol Struct Theochem 461–462:399–416.
41. Rosokha SV, Kochi JK (2002) The preorganization step in organic reaction mecha-
nisms. Charge-transfer complexes as precursors to electrophilic aromatic sub-
stitutions. J Org Chem 67(6):1727–1737.
10. Pfeiffer P, Wizinger R (1928) Zur theorie der halogensubstitution [On the theory of
halogen substitution]. Justus Liebigs Ann Chem 461(1):132–154. German.
11. Brown HC, Pearsall HW (1952) The action of the catalyst couple aluminum chloride–
hydrogen chloride on toluene at low temperatures; the nature of Friedel–Crafts
complexes. J Am Chem Soc 74(1):191–195.
42. Wei Y, et al. (2005) Theoretical study of the iodination of methoxybenzene by iodine
monochloride. J Phys Org Chem 18(7):625–631.
43. Volkov AN, Timoshkin AY, Suvorov AV (2005) Pathways of electrophilic aromatic
substitution reactions catalyzed by group 13 trihalides: An ab initio study. Int
Quantum Chem 104(2):256–260.
J
12. Kong J, et al. (2011) The inherent competition between addition and substitution
reactions of Br2 with benzene and arenes. Angew Chem Int Ed 50(30):6809–6813.
13. Koleva G, Galabov B, Kong J, Schaefer HF, Schleyer PvR (2011) Electrophilic aromatic
sulfonation with SO3: Concerted or classic SEAr mechanism? J Am Chem Soc 133(47):
19094–19101.
14. Sakic D, Vrcek V (2012) Prereactive complexes in chlorination of benzene, triazine,
and tetrazine: A quantum chemical study. J Phys Chem A 116(4):1298–1306.
15. Watson WD (1974) Chlorination of phenols with chlorine and tert-butyl hypochlorite.
Comparison. J Org Chem 39(8):1160–1164.
44. Ben-Daniel R, de Visser SP, Shaik S, Neumann R (2003) Electrophilic aromatic chlori-
nation and haloperoxidation of chloride catalyzed by polyfluorinated alcohols: A new
manifestation of template catalysis. J Am Chem Soc 125(40):12116–12117.
45. Filimonov VD, Poleshchuk OK, Krasnokutskaya EA, Frenking G (2011) DFT investi-
gation of the thermodynamics and mechanism of electrophilic chlorination and
iodination of arenes. J Mol Model 17(11):2759–2771.
46. Zhang M, Lund CRF (2002) An experimental and computational study of solvent ef-
fects in toluene chlorination. J Phys Chem A 106(43):10294–10301.
47. Olah GA, Kuhn SJ, Flood SH, Hardie BA (1964) Aromatic substitution. XIV. Ferric
chloride catalyzed bromination of benzene and alkylbenzenes with bromine in ni-
tromethane solution. J Am Chem Soc 86(6):1039–1044.
16. Watson WD (1982) Formation of nonaromatic products in the chlorination of simple
substituted aromatic ethers. J Org Chem 47(27):5270–5276.
17. de Queiroz JF, et al. (2006) Electrophilic aromatic nitration: Uunderstanding its
mechanism and substituent effects. J Org Chem 71(16):6192–6203.
18. Baciocchi E, Mandolini E (1987) The model for the transition state of electrophilic
aromatic substitutions: σ-Complex or excited charge-transfer complex? A mise au
point. Tetrahedron 43(17):4035–4041.
19. Holleman AF (1925) Some factors influencing substitution in the benzene ring. Chem
Rev 1(2):187–230.
20. Price CC (1941) Substitution and orientation in the benzene ring. Chem Rev 29(1):37–67.
21. De la Mare PBD, Bolton R (1966) Electrophilic Additions to Unsaturated Systems
(Elsevier, New York), pp 241–251.
22. Jang KS, Shin HY, Chi DY (2008) Electrophilic aromatic addition reaction (AdEAr) to
anthracene. Tetrahedron 64(24):5666–5671.
23. Duan S, et al. (2000) Halogenations of anthracenes and dibenz[a,c]anthracene with
N-bromosuccinimide and N-chlorosuccinimide. J Org Chem 65(10):3005–3009.
24. Mayo FR, Hardy WB (1952) The bromination of naphthalene. J Am Chem Soc 74(4):
911–917.
48. Vasiliev AV, Lindeman SV, Kochi JK (2002) Molecular structures of the metastable
charge-transfer complexes of benzene (and toluene) with bromine as the pre-reactive
intermediates in electrophilic aromatic bromination. New J Chem 26:582–592.
49. Fukuzumi S, Kochi JK (1982) Transition-state barrier for electrophilic reactions. Sol-
vation of charge-transfer ion pairs as the unifying factor in alkene addition and ar-
omatic substitution with bromine. J Am Chem Soc 104(26):7599–7609.
50. Koleva G, Galabov B, Wu J, Schaefer HF, Schleyer PvR (2009) Electrophile affinity:
A reactivity measure for aromatic substitution. J Am Chem Soc 131(41):14722–14727.
51. De la Mare PBD (1976) Electrophilic Halogenations (Cambridge Univ Press, Cambridge, UK).
52. Harvey DR, Norman ROC (1961) The ortho:para-ratio in aromatic substitution. Part II.
Chlorination with t-butyl hypochlorite. J Chem Soc 3604–3610.
53. Kennard WW, Matthews DN (1960) The para:ortho ratio in the monochlorination
of phenol. Aust J Chem 13(2):317–320.
54. Frisch MJ, et al. (2009) GAUSSIAN09, Revision A. 02 (Gaussian, Wallingford, CT).
55. Curtiss LA, et al. (1995) Extension of Gaussian-2 theory to molecules containing third-
row atoms Ga-Kr. J Chem Phys 103(14):6104–6113.
25. De la Mare PBD (1974) Pathways in electrophilic aromatic substitutions. Cyclohexadienes
and related compounds as intermediates in halogenation. Acc Chem Res 7(11):
361–368.
56. Clark T, Chandrasekhar J, Spitznagel GW, Schleyer PvR (1983) Efficient diffuse func-
tion-augmented basis sets for anion calculations. III. The 3-21+G basis set for first-row
elements, Li–F. J Comput Chem 4(3):294–301.
26. Ingold CK (1969) Structure and Mechanism in Organic Chemistry (Cornell Univ Press,
New York), Chap 6, pp 221–306.
57. Becke AD (1993) Density-functional thermochemistry. III. The role of exact exchange.
J Chem Phys 98(7):5648–5653.
27. Skokov S, Wheeler RA (1999) Oxidative aromatic substitutions: Hartree−Fock/density
functional and ab initio molecular orbital studies of benzene and toluene nitrosation.
J Phys Chem A 103(21):4261–4269.
28. Gwaltney SR, Rasokha SV, Head-Gordon M, Kochi JK (2003) Charge-transfer mechanism
for electrophilic aromatic nitration and nitrosation via the convergence of (ab initio)
molecular-orbital and Marcus−Hush theories with experiments. J Am Chem Soc 125(11):
3273–3283.
29. Andrews LJ, Keefer RM (1959) The influence of solvent on the rate of aromatic
chlorination. J Am Chem Soc 81(5):1063–1067.
30. Andrews LJ, Keefer RM (1957) The chlorination of aromatic hydrocarbons in carbon
tetrachloride and in trifluoroacetic acid. J Am Chem Soc 79(19):5169–5174.
31. Carey FA, Sundberg RJ (2007) Advanced Organic Chemistry, Part A, Structure and
Mechanisms (Springer, New York), Chap 10, pp 579–629.
58. Becke AD (1996) Density-functional thermochemistry. IV. A new dynamical correlation
functional and implications for exact-exchange mixing. J Chem Phys 104(3):1040–1046.
59. Lee CT, Yang WT, Parr RG (1988) Development of the Colle-Salvetti correlation-
energy formula into a functional of the electron density. Phys Rev B 37:785–789.
60. Perdew JP, Burke K, Ernzerhof M (1996) Generalized gradient approximation made
simple. Phys Rev Lett 77:3865–3868.
61. Perdew JP, Burke K, Ernzerhof M (1997) Errata: Generalized gradient approximation
made simple. Phys Rev Lett 78:1396.
62. Grimme S, Antony J, Ehrlich S, Krieg H (2010) A consistent and accurate ab initio
parametrization of density functional dispersion correction (DFT-D) for the 94 ele-
ments H-Pu. J Chem Phys 132(15):154104.
63. Grimme S (2006) Semiempirical hybrid density functional with perturbative second-
order correlation. J Chem Phys 124(3):034108.
32. Smith MB, March J (2013) March’s Advanced Organic Chemistry Reactions, Mecha-
nisms and Structure (Wiley, New York).
64. Fukui K (1981) The path of chemical reactions – the IRC approach. Acc Chem Res 14(12):
363–368.
33. Streitwieser A (1961) Molecular Orbital Theory for Organic Chemists (Wiley, New York).
34. Bachrach SM (2014) Computational Organic Chemistry (Wiley, Hoboken, NJ).
65. Tomasi J, Mennucci B, Cammi R (2005) Quantum mechanical continuum solvation
models. Chem Rev 105(8):2999–3094.
10072
|
www.pnas.org/cgi/doi/10.1073/pnas.1405065111
Galabov et al.