Molecular structures of the metastable charge-transfer complexes of benzene (and toluene) with bromine as the pre-reactive intermediates in electrophilic aromatic bromination

Article abstract of DOI:10.1039/b110169m

Successful crystallization and X-ray crystallographic analyses of the highly metastable (1:1) complexes of bromine with benzene and toluene establish the unique (localized) structure B that di3ers in notable ways from the long-accepted (delocalized) structure A. Furthermore, we demonstrate the (highly structured) charge-transfer complexes [C₆H₆,Br₂] and [CH₃C₆H₅,Br₂] to be the pre-reactive intermediates that are converted (via an overall Br⁺ transfer) to the Wheland intermediates in electrophilic aromatic bromination. The role of the dative ion pairs [C₆H₆^·+ Br₂^·-] and [CH₃C₆H₅^·+ Br₂^·-] in the rate-limiting activation processes is underscored.

Full text of DOI:10.1039/b110169m

View Article Online / Journal Homepage / Table of Contents for this issue
Molecular structures of the metastable charge-transfer complexes of
benzene (and toluene) with bromine as the pre-reactive intermediates
in electrophilic aromatic brominationy
Alexandr V. Vasilyev, Sergey V. Lindeman and Jay K. Kochi
Department of Chemistry, University of Houston, Houston, TX 77204-5003, USA
Received (in New Haven, CT, USA) 25th October 2001, Accepted 30th November 2001
First published as an Advance Article on the web 15th April 2002
Successful crystallization and X-ray crystallographic analyses of the highly metastable (1 : 1) complexes of
bromine with benzene and toluene establish the unique (localized) structure B that differs in notable ways from
the long-accepted (delocalized) structure A. Furthermore, we demonstrate the (highly structured) charge-
transfer complexes [C6H6 ,Br2] and [CH3C6H5 ,Br2] to be the pre-reactive intermediates that are converted (via
an overall Brþ transferþ) to the Wheland intermediates in electrophilic aromatic bromination. The role of the
ꢀþ
dative ion pairs [C6H6 Br2 À] and [CH3C6H5 Br2 À] in the rate-limiting activation processes is underscored.
ꢀ
ꢀ
ꢀ
More than 52 years ago, Benesi and Hildebrand published
their seminal studies describing the unique spectral (UV-vis)
changes that accompany the spontaneous complexation of
various aromatic hydrocarbons (ArH) with iodine in nonpolar
solvents (CCl4 , C6H14 , etc).1 Keefer and Andrews (and others)
in extending such spectroscopic studies also found the mag-
nitudes of the (thermodynamic) equilibrium constants KCT for
the formation of these intermolecular (1 : 1) complexes
donor/acceptor pair. In this system, the intermolecular (1 : 1)
complex is transient since its diagnostic (charge-transfer)
absorption band with energy hnCT slowly disappears as bromo-
benzene and hydrogen bromide are coproduced. However,
these simultaneous chemical events may not be directly cou-
pled, since Colter and Dack8 correctly pointed out that the
reversible formation of the EDA complex (KEDA) may be an
unrelated side process independent of electrophilic bromina-
tion (kBr):
KCT
!
ArH þ X2
½ArH; X2Š
ð1Þ
KCT
kBr
C6H6 þ Br2 À! C6H5Br þ HBr
½C6H6; Br2Š
ð2Þ
!
to be uniformly limited, typically with KCT < 3 MÀ1 for the
halogens X2 ¼ I2 , Br2 , and Cl2 or the interhalogens XY ¼ IBr,
ClF, etc.2
Mechanistically, such a parallel process in which the EDA
complex is an innocent bystander cannot be kinetically dis-
tinguished from the sequential process [eqn. (3)], in which it
lies squarely on the pathway to electrophilic aromatic bromi-
nation:
Immediately following the Benesi–Hildebrand report, Mul-
liken published another landmark paper in 1950,3 in which he
assigned these new spectral bands to the unusual electronic
(charge-transfer) transition from theþground-state complex
[D,A] to the dative excited state [Dꢀ ,AꢀÀ], where D is the
generic representation of electron donors (such as aromatic
hydrocarbons, etc.) and A identifies the electron acceptors
(such as X2 , XY, etc.) in eqn. (1).
Despite the subsequent explosion in the number and types of
papers dealing with the various facets of electron donor/
acceptor, or EDA, complexes,4–6 reports of their reactivity as
intermediates in (irreversible) chemical reactions are sparse. In
the latter context, there are two reviews7,8—both now more
than 25 years old—that unfortunately failed to kindle wide-
spread interest in the kinetic (as opposed to static) aspects of
these interesting EDA complexes. To make the point, we now
focus simply on the benzene/bromine dyad as a prototypical
KCT
kBr
!
C6H6 þ Br2
½C6H6; Br2Š À! C6H5Br þ HBr
ð3Þ
Various spectroscopic (IR, NQR, NMR, etc.) techniques
have been applied to deduce the structures of [C6H6 ,Br2] and
related complexes,9–11 but to date the classic X-ray crystal-
lographic determination by Hassel and Strꢀmme in 195812
stands alone as the principal structural standard (for the weak
binding of bromine to benzene) by which all others
are invariably compared.13 Their structure A reveals the
non-covalently bound dibromine acceptor to lie in an axial
orientation relative to the benzene plane. The Br–Br bond
(2.28 A), which has essentially the same length as that found in
elemental bromine, lies across an inversion center on (or near)
the 6-fold symmetry axis of benzene at an intermolecular
+
+
bromine–benzene separation of D ¼ 3.36 A that is significantly
y Electronic supplementary information (ESI) available: X-ray crystal-
lographic data for: benzene/bromine charge-transfer complex (Tables
S1–S5), toluene/bromine charge-transfer complex (Tables S6–S10),
bromohexamethylbenzenium tribromide (s-complex) as the tris-
(dibromine) solvate (Tables S11–S15), bromohexamethylbenzenium
hexafluoroantimonate (s-complex) (Tables S16–S20); actual transmit-
tance spectrum of the filter used for the isolation of UV light for the
specific irradiation of the charge-transfer band of arene/bromine
complexes (Fig. S1); and the detailed structure of the donor/acceptor
chains in the crystal structure of the toluene/dibromine charge-
transfer complex (Fig. S2). See http://www.rsc.org/suppdata/nj/b1/
b110169m/
582
New J. Chem., 2002, 26, 582–592
DOI: 10.1039/b110169m
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2002

View Article Online
Fig. 1 (A) Spectral (UV-vis) changes attendant upon the incremental addition of benzene aliquots to a dilute solution of 5 mM bromine in carbon
tetrachloride at bromine : benzene ratios of 1 : 2 to 1 : 40 (bottom to top). For comparison, the spectra of the solutions in CCl4: 5 mM Br2 alone
(- - - -) and 0.1 M C6H6 alone (Á Á Á Á Á Á). (B) Similar to A [except for the insertion of a 5 mM Br2 in CCl4 filter (blank solution) in the reference beam
of the spectrometer] to isolate the progressive growth of the charge-transfer band (lCT ¼ 285 nm).
+
closer than the van der Waals contact distance of 3.55 A.14 As
such, structure A represents the electronic interaction of a
completely delocalized benzene donor with the bromine
acceptor—much in the way predicted by Mulliken theory.3,15
However, our careful perusal of Hassel and Strꢀmme’s
experimental details raised some serious questions as to the
definitiveness of structure A.16 Accordingly, in this paper we
re-examine the X-ray crystallography of the benzene/dibro-
mine complex and extend our consideration to the cor-
responding toluene/dibromine complex for completeness.
Furthermore, the availability of the bromine complexes in
crystalline form allows us to directly effect the electrophilic
bromination of benzene according to eqn. (3), since under
these solid-state conditions only nearest neighbors react, and
diffusional (second-order) processes are largely precluded.17
Toluene. The spectral changes attendant upon the incre-
mental additions of pure toluene to a 5 mM solution of bro-
mine in carbon tetrachloride are shown in Fig. 2. The red-shift
of the charge-transfer absorption band of the toluene/bromine
complex to lmax ¼ 295 nm follows from the Mulliken corre-
lation of its increased donor strength (Eꢁ ¼ 2.25 V) relative to
ox
that of benzene (Eꢁ ¼ 2.62 V).19
ox
Crystallization of the bromine complexes of arenes donors
Benzene. Owing to the low value of the formation constant
KEDA , the benzene/bromine complex was necessarily prepared
in situ by the low-temperature crystallization of the pure
components in a sealed glass capillary.20 For example, the
equimolar mixture of benzene and bromine remained liquid
at À30 ꢁC, but crystal nucleation was readily initiated by
Results
Spectral (UV-vis) changes accompanying the bromine
complexation to arene donors
Benzene. When pure benzene was added incrementally in
small amounts to a dilute (5 mM) solution of bromine in
carbon tetrachloride, the red-brown color changed almost
imperceptibly. However, inspection of the UV-vis spectrum
readily revealed the progressive growth of a new absorption
band at lmax ¼ 285 nm [see Fig. 1(A)]. Benesi–Hildebrand
treatment of the absorbance data yielded the formation con-
stant KEDA ¼ 1.0 MÀ1, in agreement with the earlier determi-
nation.2 In the [C6H6 ,Br2] complex, the ‘‘local’’ band of the
bromine moiety was unchanged relative to the absorption of
free bromine, as shown by the series of invariant spectra at
l > 350 nm in Fig. 1(A). The latter is underscored in Fig. 1(B),
which was obtained by repeating the foregoing experiments
and merely inserting a filter (consisting of the same 5 mM
solution of Br2 in carbon tetrachloride) in the reference beam
of the spectrometer. Such spectral features of the [C6H6 ,Br2]
complex are wholly consistent with Mulliken’s formulation of
weak complexes in which the new UV-vis absorption relates to
the electronic transition (hnCT) corresponding to:18
Fig. 2 UV-vis spectral changes upon the addition of toluene in in-
cremental amounts to a dilute solution of 5 mM bromine in carbon
tetrachloride at bromine : toluene ratios of 1 : 5 to 1 : 20 (bottom to
top), showing the growth of the charge-transfer band (lCT ¼ 295 nm).
For comparison, the spectra of the solutions in CCl4: 5 mM Br2 alone
(- - - -) and 0.1 M toluene alone (Á Á Á Á Á Á).
ꢀ
ꢀ
hnCT
½C6H6; Br2Š
½C6H þ; Br ÀŠ
ð4Þ
!
6
2
New J. Chem., 2002, 26, 582–592
583

View Article Online
carefully brushing liquid nitrogen over the capillary with a
cotton applicator. By a series of local (manual) warmings all
but one small crystal was alternately dissolved/melted, and the
remaining single crystal was allowed to grow along the capil-
lary axis at À40 ꢁC. The brown color of the crystal was almost
indistinguishable from the color of the residual liquid (com-
pare Fig. 1), but its slow growth could be continuously mon-
itored under a microscope using crossed polarizers. Most
interestingly, the crystal exhibited a phase change as the tem-
perature was gradually decreasedÀt1o À70 ꢁC, but only a very
slow cooling rate of ꢂ1 ꢁC min resulted in the apparent
single-crystal-to-single-crystal phase transformation of the
[C6H6 ,Br2] complex.
Toluene. An equimolar mixture of toluene and bromine was
visually indistinguishable from the brown benzene complex.
Most notably, a series of carefully controlled studies showed
that the toluene complex (visually) bleached within 2–3 h in the
temperature range of À40 to À50 ꢁC. In order to successfully
grow a single crystal of the toluene/bromine complex, various
molar mixtures were examined at lower temperatures. When a
2 : 1 molar ratio of toluene and bromine was employed, the
resulting brown liquid began to crystallize at À70 ꢁC to pro-
duce bright orange crystals. After some manual local warm-
ings, all but one crystal was suppressed in the capillary. The
single crystal of the 1 : 1 complex consisted of a bright orange
prism positioned along the capillary axis, and the surrounding
liquid (presumably consisting of the excess of toluene) was pale
yellow and glassy (clear and isotropic under polarized light) at
À150 ꢁC.
Fig. 3 Molecular diagram showing the localized (over-atom/bond)
coordination of Br2 to benzene. Thermal ellipsoids of non-hydrogen
atoms are shown at the 50% probability level.
coordinate to benzene symmetrically. Instead, bromine is
positioned over the rim (not the center) of the benzene ring as
+
in structure B—being shifted by d ¼ 1.44 A from the main (C6)
symmetry axis. In structure B, the dibromine molecule is
essentially oriented perpendicular to the benzene plane, and
tilted by only a ¼ 5.1 deg off the C6 axis.
The molecular structure of the [C6H6 ,Br2] complex in Fig. 3
shows an asymmetric coordination of bromine to benzene as
given by the shortest pair of BrÁ Á ÁC distances of d1 ¼ 3.18 A
X-Ray crystallography of the bromine complexes of benzene
and toluene
+
+
and d2 ¼ 3.36 A, both of which are substantially shorter than
+
the sum of the van der Waals radii of 3.55 A. Otherwise, the
intermolecular complex shows little deviation of the Br–Br
X-Ray crystallographic analyses of the 1 : 1 bromine complexes
of benzene and toluene were uniformly carried out at À150 ꢁC
to obviate the dynamic disorder observed at higher tempera-
tures. As a result, our structural conclusions about the bro-
mine binding in these complexes differ in substantial ways
from those obtained by Hassel and Strꢀmme at higher tem-
peratures (À40 to À50 ꢁC).12a
+
bond of l ¼ 2.30 A, which is only slightly longer than that in
+
free bromine (l ¼ 2.28 A). The precision of the bond-length
+
determination in our experiments (sCC ¼ 0.006 A) is insuffi-
cient to allow the detection of small polarization effects in the
benzene donor since such changes in (multiple) C–C bonds are
+
typically less than 0.005 A.21
Bromine binding to benzene. In striking contrast to the axial
(delocalized) structure A, we found that bromine does not
Bromine binding to toluene. As in the localized structure B,
bromine is also positioned over the rim (not above the center)
of the toluene ring in the form of non-equivalent dyads, the
structural parameters of which are listed in Table 1. The clo-
sest approach of bromine occurs at the normal distances
+
D ¼ 3.01–3.17 A, which are on the average somewhat shorter
than that in the benzene complex. In all cases, there is an
asymmetric coordination of bromine, as given by the pair of
shortest BrÁ Á ÁC distances d1 and d2 in Table 1. More precisely,
the coordination of bromine to the aromatic ring can be
Table 1 Principal geometric parameters of the dibromine complexes of benzene and toluene
+
+
+
+
+
+
Interacting molecules
Da/A
ab/deg
dc/A
d1d/A
d2e/A
Z f
lg/A
lavh/A
Benzene complex
Br1A–Br1Á Á Á(C1...C3A)
Toluene complexes
3.154(8)
5.1(5)
1.44(1)
3.18(1)
3.36(1)
1.52
2.301(2)
1.39(2)
Br2–Br1Á Á Á(C1A...C6A)
Br1–Br2Á Á Á(C1B...C6B)
Br3A–Br3Á Á Á(C1A...C6A)
Br4A–Br4Á Á Á(C1B...C6B)
3.009(3)
3.172(3)
3.099(3)
3.133(3)
5.4(2)
20.7(2)
4.6(2)
1.397(4)
1.472(4)
0.936(4)
1.414(4)
3.053(4) ortho
3.229(4) ortho
3.146(4) para
3.196(4) para
3.150(4) meta
3.292(4) ipso
3.259(4) meta
3.241(4) meta
1.70
1.82
1.70
1.86
2.307(1)
2.307(1)
2.291(1)
2.304(1)
1.389(6)
1.385(6)
1.389(6)
1.385(6)
7.9(2)
a
b
Distance of bromine to the mean aromatic plane. Angle between the vector of the Br–Br bond and the normal to the aromatic
c
d
e
plane. Deviation of the coordinated Br from the main axis of benzene. The shortest BrÁ Á ÁC distance. Second shortest BrÁ Á ÁC
f
distance. Hapticity of the coordination. The Br–Br bond length. The average C–C bond length in the aromatic ring.
g
h
584
New J. Chem., 2002, 26, 582–592

+
evaluated as the hapticity (Z) for coordination,22 so that Z ¼ 1
when d1 ¼ D (‘‘over-atom’’ coordination) and Z ¼ 2 when
d1 ¼ d2 (‘‘over-bond’’ coordination). For intermediate cases,
the hapticity can be estimated as a function of the relative
Although the intermolecular CÁ Á ÁBr separation of D ¼ 3.18 A
View Article Online
+
is 0.37 A closer than the equilibrium van der Waals distance,14
the contraction is perceptibly less than those previously
reported in a series of complexes with slightly polarizable and
weakly nucleophilic donors.23 [For example, the XÁ Á ÁBr dis-
(separation) values: (d1 À D2)1/2 and (d2 À D2)1/2 by using the
2
2
geometric relationship:
tance contraction (relative to the corresponding equilibrium
+
van der Waals separations) is 0.55 A in the acetone/Br2
+
1=2
Z ¼ 1 þ 2ðd1 À D2Þ1=2=½ðd1 À D2Þ
+
2
2
complex (OÁ Á ÁBr 2.82 A),24 0.56 A in the acetonitrile/Br2
1=2
2
+
+
þ ðd2 À D2Þ
Š
ð5Þ
complex (NÁ Á ÁBr 2.84 A),25 0.57 A in the [Te2Cl10]2À
/
/
+
+
Br2 complex (ClÁ Á ÁBr 3.03 A),26 and 0.60 A in the [Se2Br10]2À
In the toluene complex, the hapticities evaluated in this way
vary from 1.70 to 1.86, and thus lie closer to the ‘‘over-bond’’
coordination model. Importantly, the ‘‘over-bond’’ coordi-
nated bromine is shifted toward the ortho- and para- carbons
of toluene [see Fig. 4 and S2 (ESI)].
+
+
Br2 complex (BrÁ Á ÁBr 3.10 A)26,27]. Moreover, the average
CÁ Á ÁBr separation of 3.156 A in the toluene/Br2 complex is
somewhat shorter than that in the benzene complex, as
expected from the better donor strength of toluene.28
The weak C(arene)Á Á ÁBr charge-transfer interaction is
reflected in an almost unperturbed geometry of the coordi-
nated dibromine. [The Br–Br bond lengths are actually very
General structural features of weak arene/Br2 complexes. The
charge-transfer complex [C6H6 ,Br2] is presently the weakest
EDA complex of dibromine studied in the solid state.
sensitive to coordination/polarization effects and readily
elongate from 2.284 A in the non-coordinated molecule (bond
+
+
order n ¼ 1) to 2.53 A in the symmetric [Br3]À anion29 (bond
+
order n ¼ 1/2).] As such, the Br–Br bond lengths of 2.301(2) A
+
in the benzene complex and an average of 2.302(1) A in the
toluene complex do not exhibit much elongation during
complex formation. For comparison, the Br–Br bond lengths
+
vary within a narrow range (2.28 to 2.33 A) in the weakly
coordinated acetone, acetonitrile, dioxane and methanol
complexes.24,25,30,31
In the absence of significant polarization, dibromine can be
coordinated equally well from either end (owing to the
acceptor s*-orbital which is localized on both bromine cen-
ters) and this explains why dibromine has often been found in
crystals to be symmetrically coordinated to a pair of donor
molecules (in a bridging manner), especially in complexes with
weak donors.12b,c [However, it is important to note that in
solution, 2 : 1 complexes of dibromine with benzene (and
toluene) are only found at very high Br2 concentrations.] In the
benzene and toluene complexes, dibromine is also positioned
symmetrically between the coordinated benzene rings forming
infinite (weak) Á Á ÁArÁ Á ÁBr–BrÁ Á ÁArÁ Á ÁBr–BrÁ Á ÁArÁ Á Á chains
through the crystal, and there are no specific interactions other
than van der Waals contacts between the chains. Although the
chains are highly symmetrical in the benzene/dibromine crys-
tals—with 2-fold axes (through the diagonals of the benzene
rings and through the centers of the dibromine molecules)
across the chains—the chains in the toluene/dibromine crystals
are less so. Two of the three dibromines (Br3–Br3A and Br4–
Br4A) occupy inversion centers and are thus symmetrically
coordinated, but the third dibromine (Br1–Br2) does not show
crystallographic symmetry. Indeed, the latter exhibits some
signs of larger polarization as a result of a less symmetric
coordination (Table 1), and it has the shortest contact, CÁ Á ÁBr
+
3.053(4) A, as well as the longest Br–Br bond length, 2.307(1)
A, in the series. Interestingly, a similar asymmetric coordina-
+
tion of dibromine is found in the complex with methanol,30 in
+
which the OÁ Á ÁBr distance is shorter (2.705 vs. 2.723 A) and the
+
Br–Br bond length is longer (2.324 vs. 2.303 A) than those in
the closely related (but symmetric) dioxane complex.31 This
structural effect predicts that polarization in isolated donor/
acceptor dyads (such as those extant in dilute solutions) will be
somewhat stronger than that observed in (crystalline) polymeric
chains.
Solid-state (thermal) transformation of arene/Br2 complexes
via electrophilic bromination
Benzene/bromine. Crystals of the EDA complex are sur-
prisingly reactive, especially if one considers that equimolar
solutions of benzene and bromine dissolved in carbon tetra-
chloride remained unchanged at room temperature for pro-
longed periods if protected from adventitious light. The
Fig. 4 Localized bonding of bromine to the ortho- (top) and para-
(bottom) centers of toluene in the charge-transfer complex.
New J. Chem., 2002, 26, 582–592
585

View Article Online
Hexamethylbenzene/bromine. The complex prepared in a
Table 2 Solid-state (thermal) transformation of the benzene/bromine
complex via electrophilic bromination at different temperaturesa
sealed tube from equimolar amounts of hexamethylbenzene
and bromine in dichloromethane solution was allowed to stand
undisturbed in a cold bath at À40 ꢁC. After more than a week,
the mixture deposited a dark red salt with the composition:
Bromobenzene yield(%)
T/ꢁC
After 3 h
After 6 h
.
C6(CH3)6BrþBr3À 34 X-Ray crystallographic analysis indicated
the formation of a cationic bromoarenium s-adduct:
À78
À60
À40
À20
<0.03
0.05
0.1
0.08
0.1
0.2
1.5
1
ð8Þ
a
In the dark, without solvent, using 2 mmol each of benzene and
bromine.
The unit cell consisted of a honeycomb of anionic polybromine
networks with cages populated by the cationic s-complex.
Since these cages have a local plane of symmetry the s-com-
plex structure was sufficiently disordered to afford poor pre-
cision. However, the molecular diagram of the well-ordered
structure of the same cationic s-complex obtained as the
hexafluoroantimonate salt is illustrated in Fig. 5.35
crystalline 1 : 1 complex consisting of [C6H6 ,Br2] melted at
À14 ꢁC. Nonetheless, even at À78 ꢁC, the brown crystals
slowly evolved hydrogen bromide, and essentially quantitative
yields of bromobenzene were found upon workup:
½C6H6; Br2ŠÀ!C6H5Br þ HBr
ð6Þ
Charge-transfer photoreactions of arene/bromine complexes
Although the solid-state conversion was deliberately kept low
( <0.5%) to minimize disruption of the crystal structure, we
consider the electrophilic substitution in eqn. (6) to represent
a crystalline (first-order) process. The higher conversion
achieved with increasing temperature (Table 2) probably also
represented crystalline transformations of the [C6H6 ,Br2]
complex, although there is some ambiguity owing to the phase
change observed between À60 and À70 ꢁC (vide supra) that
may have allowed some (but limited) diffusional separation of
benzene from bromine for second-order reactivity. Be that as it
may, careful scrutiny revealed the solid-state transformation of
[C6H6 ,Br2] to be singularly uncomplicated by side products.32
The spectral characteristics of the UV-vis absorption of the
arene/Br2 complexes [as described in eqn. (4)] suggested the
possibility of their photoactivation by the deliberate irradia-
tion of the charge-transfer band.36 For the benzene complex,
the charge-transfer band (hnCT ¼ 285 nm) occurs in a well-
defined (UV) window between l ¼ 275 and 350 nm (see Fig. 1),
which was well suited for the filter combination we prepared to
only allow transmission of light with 280 < l < 350 nm—
hereinafter referred to as lexc ¼ 320 nm (see Experimental).
Benzene. The specific irradiation (lexc ¼ 320 nm) of the
charge-transfer absorption band of a crystalline sample of
[C6H6 ,Br2] complex at À78 ꢁC for 6 h led to a 0.10% con-
version to bromobenzene that was uncontaminated by other
by-products. However, the dark control carried out in a side-
by-side experiment led to 0.08% bromobenzene. Moreover,
when an equimolar (liquid) mixture of neat benzene and bro-
mine was similarly irradiated at 0 ꢁC (6 h), it resulted in a 5%
conversion to bromobenzene; at 25 ꢁC (6 h) conversion was
12%. However, both of these were close to the bromobenzene
conversion rates of 4.5% at 0 ꢁC and 11% at 25 ꢁC in the dark
control for the same period of time (vide supra).
Toluene/bromine. Crystals were derived from an equimolar
mixture of pure donor and acceptor. The bright-orange crys-
tals of [PhCH3 ,Br2] slowly evolved hydrogen bromide on
standing at À78 ꢁC in the dark. Workup of the reaction mix-
ture after 6 h yielded a roughly 1 : 2 mixture of ortho- and para-
bromotoluene:
ð7Þ
Toluene. An equimolar mixture of neat toluene and di-
bromine cooled at À78 ꢁC as red-brown crystals was irradiated
with lexc ¼ 320 nm for 6 h. Workup of the partially converted
reaction mixture resulted in a mixture ortho- and para-
bromotoluenes in 5% and 14%, respectively. However, the
dark control resulted in ortho- and para-bromotoluenes in 5%
and 13% yields, respectively (Table 3). When an equimolar
mixture of toluene and bromine was cooled to only À65 ꢁC, it
remained as a clear brown liquid. Irradiation at lexc for 6 h led
to a mixture of ortho- and para-bromotoluenes in 9% and 25%
yields, respectively, together with traces (0.1%) of benzyl
bromide.33 When compared to the thermal control (see
Table 3), the slightly enhanced yields of ortho- and para-
bromotoluenes were 1.6% and 3.0%. Although such conver-
sions were low, they could be carried out reproducibly
(within 1%). Considering the experimental difficulty of car-
rying out such low-temperature photoirradiations, we consider
these experiments to be indicative of the inefficient charge-
transfer photoactivation of the [ArH,Br2] complexes for elec-
trophilic bromination of both benzene and toluene:36
but no benzyl bromide could be detected.33 The conversion
and yields of bromotoluenes obtained at low temperatures are
listed in Table 3. It is noteworthy that the molar ratio of the
ortho and para isomers of bromotoluene obtained from the
solid-state transformation of the charge-transfer complex was
the same as that obtained in carbon tetrachloride solution.
Table 3 Thermal transformation of the neat toluene/bromine com-
plex to bromotoluenes at low temperaturesa
Bromotoluene yieldb(%)
T/ꢁC
ortho
para
À78
À70
À65
À60
À50
5.0
6.2
7.4
12
13
15
22
32
44
16
hnCT
a
In the dark, without solvent, using 2 mmol each of toluene and
bromine. After 6 h; benzyl bromide <0.05% in all cases.
½ArH; Br2Š À! ArBr þ HBr
ð9Þ
b
the quantum yields of which were estimated to be <10À2
.
586
New J. Chem., 2002, 26, 582–592

View Article Online
Fig. 5 Molecular structure of the cationic bromohexamethylbenzenium s-adduct, showing the ion pairing to (A) tribromide anion and (B) to
hexafluoroantimonate anion.
a specific carbon center of benzene and not as in the deloca-
Discussion
lized structure
A originally proposed by Hassel and
The successful crystallization and X-ray crystallographic ana-
lyses of the metastable bromine complexes of benzene and
toluene bear directly on the mechanism of electrophilic aro-
matic bromination in several important ways.
First, the molecular structure in Fig. 3 shows the pre-
organized bromine complex of benzene to have the discrete
localized structure B in which the binding of bromine occurs at
Strꢀmme.12a,37 Such a highly localized structure is strongly
reminiscent of the transition state for electrophilic bromina-
tion. Yet it is formed in a rapid pre-equilibrium step (with
essentially no energy barrier). The dibromine moiety remains
largely intact (with only a slight elongation of the Br–Br bond)
in the pre-reactive benzene complex (structure B). More-
over, the rather close bromine proximity to the benzene
New J. Chem., 2002, 26, 582–592
587

+
View Article Online
chromophore at an intermolecular distance of D ¼ 3.15 A
derives from charge-transfer forces that are sufficient to bind
+
the donor/acceptor pair at a separation ꢂ0.4 A closer than
that allowed by van der Waals contacts.18 Such a significant
charge-transfer interaction is even more clearly shown in the
bromine complexation to toluene. Thus, the molecular struc-
ture in Fig. 4 readily shows bromine to gravitate specifically to
the electron-rich ortho and para carbons. It is singularly
notable that the dibromine is poised over only those carbon
centers in the pre-reactive toluene complex that are expected to
lead to the transition state for the preferential ortho- and para-
brominations. In the benzene complex, a pair of dibromines
coordinates each benzene ring from opposite sides in the meta
positions. Otherwise there is no obvious steric reason to favor
such a coordination, but the meta positions are known to be
relatively more electron-rich in arenes with acceptor sub-
stituents. Despite the quasi-chain structures of the crystalline
[C6H6 , Br2] and [CH3C6H5 , Br2] complexes, there is no doubt
that their charge-transfer character derives from discrete
intermolecular (1 : 1) interactions of Br2 with benzene (and
toluene).
Fig. 6 Linear correlation of the rate (log kBr) of electrophilic aro-
matic bromination with the charge-transfer transition energy (hnCT) of
the bromine complexes with various arene donors (as identified).18
Second, the availability of the crystalline charge-transfer
complex forms the topochemical basis17 for the direct (pair-
wise) interaction of the arene donor and the bromine acceptor
in the absence of diffusion. As such, the bromination results in
Tables 2 and 3 (showing exceptionally high solid-state reac-
tivity relative to that in solution) prove that electrophilic
aromatic bromination of benzene and toluene proceeds via
the corresponding charge-transfer complex as described in
eqn. (3).38 Thus, the complex is not merely an innocent
bystander in the bromination process [as suggested in eqn. (2)].
T hird, the subsequent steps leading to the electrophilic
bromination process are also fairly clear but more difficult to
prove unambiguously. Thus, the observation of the bromo-
arenium s-adduct ion pair from hexamethylbenzene and
dibromine [eqn. (8)], together with the molecular structures in
Fig. 3 and 4, suggests that the bromine attachment coincides
with the collapse of the charge-transfer complex:
donor and the acceptor moieties.42 It is thus particularly
noteworthy that such an electronic transition (hnCT) has been
found to correlate linearly with the activation energy (log kBr
)
for electrophilic bromination of a wide series of aromatic
donors:18
log kBr ¼ ahnCT þ constant
ð12Þ
Indeed, the direct relationship [expressed by eqn. (12) and
illustrated in Fig. 6] indicates that those electronic factors
leading to the charge-transfer excited state of the [ArH,Br2]
complex and to the transition state [ArHÁ Á ÁBr2]z for electro-
philic aromatic substitution are very closely related.43 The
direct relationship between them is difficult to establish
experimentally since the photo-excitation represents a non-
adiabatic (vertical) process whereas the thermal activation is
adiabatic and accompanied by solvation changes.44 None-
theless, the direct photoexcitation of [ArH,Br2] complexes
according to eqn. (9) points to the dative ion-radical pair as the
reactive intermediate:
ð10Þ
ð13Þ
However, its rapid deactivation by back electron transfer
(kBET) expectedly leads to an inefficient photoprocess, owing
to the highly exergonic driving force for relaxation back to
the charge-transfer complex relative to the mesolytic dissocia-
Such an attachment to the fully-substituted hexamethyl-
benzene donor is reversible [eqn. (8)].39 However, when the
point of attachment occurs at an unsubstituted carbon center
(as in benzene or toluene) the subsequent rapid loss of the a-
proton renders the interchange effectively irreversible.40 Thus,
the composite of the molecular structures in Fig. 3 and 5
represents a close-to-ideal transformation adhering to the
principle of least motion.40d
tion of Br2 À, which is relatively slow.45,46 We believe that the
ꢀ
predominant thermal process for electrophilic aromatic bromi-
nations also follows an analogous (adiabatic) pathway,47 but
the final mechanistic proof must await more definitive (time-
resolved spectroscopic) studies36,48 on the temporal behavior
Since the transfer of Brþ in eqn. (10) is most likely to con-
stitute the rate-limiting step, let us consider what the electronic
character of the prereactive arene/bromine complex reveals
about the activation process for electrophilic bromination.
According to Mulliken,3,15 the characteristic new absorption
bands in Fig. 1 and 2 derive from the ground-state polarization
of the weak [ArH,Br2] complex that leads to charge-transfer
upon the absorption of light:41
of the charge-transfer ion pair [C6H6 þ,Br2 À].
ꢀ
ꢀ
Conclusions
The metastable (1 : 1) bromine complexes of benzene (structure
B) and of toluene are established as the critical pre-reactive
intermediates in electrophilic bromination according to
mechanistic eqn. (3). Its subsequent (rate-controlling) trans-
formation to the bromoarenium s-adduct (i.e., Wheland
intermediate) in eqn. (10) evokes the considerable, if not
complete, charge-transfer character established by the corre-
hnCT
ꢀ
ꢀ
½ArH; Br2Š À! ArH þ þ Br2
Such an electronic (nonadiabatic) transition to the dative state
ð11Þ
À
[ArHꢀþ,Br2 À] in eqn. (11) corresponds to the (electron)
ꢀ
lation in Fig. 6. As such, the dative ion pair [ArHꢀþ,Br2 À] is
ꢀ
depopulation of the arene HOMO at the expense of the bro-
mine LUMO, and the resultant destabilization of both the
the best (valence-bond) representation of the rate-limiting
transition state.43a
588
New J. Chem., 2002, 26, 582–592

View Article Online
sensitivity of the toluene/bromine mixture under the spectral
conditions (to interference from free-radical chain reactions).
Experimental
Materials
In situ Crystallization of the bromine complexes of benzene
and toluene
Benzene (EM Science, Merck) and toluene (EM Science,
Merck) were purified by repeated shaking with successive
portions of cold concentrated H2SO4 until the acidic layer was
colorless. The aromatic layer was washed with water, aqueous
NaHCO3 , followed by several washings with water, and dried
over CaCl2 . The arene was then refluxed (ꢂ9 h) and distilled
from sodium under an argon atmosphere and stored in
Schlenk flasks under argon. Hexamethylbenzene (Aldrich) was
purified by recrystallization from absolute ethanol. Bromine
(EM Science, Merck) was initially washed by shaking with
several portions of H2SO4 and it was then refluxed (ꢂ4 h) over
solid KBr and distilled. Predistilled bromine was refluxed
(ꢂ9 h) over P2O5 and distilled under an argon atmosphere and
stored in flasks equipped with Schlenk adapters under an
argon atmosphere. All-glass syringes with Teflon needles or
Teflon cannulas (without any steel elements) were used for all
operations with bromine. Dichloromethane (EM Science,
Merck) and carbon tetrachloride (Aldrich) were repeatedly
stirred with H2SO4 , until the acidic layer was colorless. After
separation, the organic layer was washed with water, aqueous
NaHCO3 , water, and dried over CaCl2 . The solvent was
refluxed (ꢂ9 h) and distilled from P2O5 under an argon atmos-
phere and it was again refluxed (ꢂ9 h) and distilled from CaH2
under an argon atmosphere. Dichloromethane and carbon
tetrachloride were stored in Schlenk flasks equipped with
Teflon valves fitted with Viton O-rings under an argon atmo-
sphere. Authentic samples of bromobenzene, ortho- and para-
bromotoluene, and benzyl bromide for comparison with the
products of photo- and thermo- reactions were from Aldrich.
Equimolar amounts of benzene and dibromine were mixed
(with the aid of a glass microsyringe attached to a Teflon
needle) at þ5 ꢁC under an argon atmosphere and kept at 0 ꢁC.
Small amounts of the mixture were transferred into glass
capillaries (d ¼ 0.4 mm) and the contents of the capillaries
frozen. The sealed capillary was attached (with wax) to a
hollow copper pin, leaving a ꢂ7 mm tip exposed. The pin was
mounted onto the diffractometer equipped with an LT-2 low
temperature device. The capillary was placed at an angle of
w ¼ 54ꢁ under the vertically oriented cooling nozzle, so that the
exposed part of the capillary and the pin tip were both posi-
tioned well within the laminar flow of nitrogen. The brown
color of the crystal was almost indistinguishable from the color
of the residual liquid, and its formation and growth were
continuously monitored under a polarizing microscope. The
initial crystal showed very poor diffraction (similar to the
earlier description by Hassel and Strꢀmme12). However, as
the temperature was gradually decreased through À70 ꢁC, the
crystal exhibited a phase transition, but only a slow cooling
rate (ꢂ1 ꢁC minÀ1) induced a single-crystal-to-single-crystal
phase transformation. The resulting bright orange crystal
(although cracked and surrounded by smaller satellites) was in
a trigonal space group (as opposed to the monoclinic mod-
ification studied by Hassel and Strꢀmme12 at À40 to À50 ꢁC),
but it showed a bright high-angle diffraction pattern of regular
quality at À150 ꢁC.
The crystallization of the toluene complex was in many
details similar to that for the benzene complex. Crystals grown
from an equimolar mixture at higher than À70 ꢁC were brown
and exhibited extremely poor diffraction—much like the
higher-temperature crystalline modification of the benzene/
dibromine complex. Below À70 ꢁC, the color of the crystals
changed to bright orange and the diffraction intensity
increased dramatically (in a manner similar to the transfor-
mation observed for the benzene analog). To grow a single
crystal of the toluene/dibromine complex, we employed a 2 : 1
molar ratio of toluene and dibromine. The resulting brown
liquid began to crystallize below À70 ꢁC (i.e., below the
transformation point of the 1 : 1 mixture) to produce bright
orange crystals. After some manual local warming, all but one
crystal was suppressed in the capillary. The single crystal
consisted of a bright orange prism positioned along the
capillary axis.
General
The X-ray crystallographic analyses were carried out with
a
Siemens–Bruker SMART diffractometer (l MoKa ¼
+
0.71073 A) equipped with a 1K CCD detector and an LT-2
low-temperature device. Gas chromatography was performed
on a Hewlett–Packard 5890A gas chromatograph equipped
with a HP 3392 integrator. Gas chromatography-mass spec-
trometry analyses were carried out on a Hewlett–Packard 5890
gas chromatograph interfaced to a HP 5970 mass spectrometer
1
(EI, 70 eV). H NMR spectra were recorded with a General
Electric QE-300 NMR spectrometer. UV-vis absorption
spectra were recorded on a Hewlett–Packard 8453 diode-array
spectrometer.
The mixtures obtained from the thermal and photo trans-
formations were dissolved in chloroform and the products
(bromobenzene, ortho- and para-bromotoluenes and benzyl
bromide) were identified by GC-MS analysis by comparison of
their retention parameters and mass-spectral checking patterns
with authentic samples, and with the aid of NMR 1H spec-
troscopy. Yield of the products was quantified by gas chro-
matography using the internal standard method.49
X-Ray crystal structure analysis of the arene/dibromine
complexes
The diffraction data were collected at À150 ꢁC. The data were
corrected for absorption and other effects using the SADABS
program.50 The structures were solved using direct methods51
and refined on F2 by a least-squares procedure.52
CCDC reference numbers 162148 and 162149. See http://
www.rsc.org/suppdata/nj/b1/b110169m/ for crystallographic
data in CIFor other electronic format.
Measurement of the charge-transfer spectra of [ArH,Br2]
complexes
In a 1 cm quartz cuvette under an argon atmosphere, the pure
arene (benzene or toluene) was incrementally added to a
solution of 0.005 M bromine in carbon tetrachloride so that
the bromine : arene ratio was increasing from 1 : 1 to 1 : 40.
The growth of the charge-transfer band was observed at
lmax ¼ 285 nm for the benzene/bromine complex and at 295 nm
for the toluene/bromine complex (see Fig. 1 and 2). In the case
of toluene, each spectral measurement was carried out with
fresh portions of the bromine solution, owing to the extreme
Benzene/dibrominecomplex. FormulaC6H6Br2 (M ¼ 237.93);
+
trigonal, space group P32212; a ¼ b ¼ 8.721(2), c ¼ 8.701(2) A,
+
U ¼ 573.1(2) A3, Z ¼ 3; Dcalc ¼ 2.068 g cmÀ3, m(MoKa) ¼ 105.1
cmÀ1; 7294 reflections were collected over a reciprocal hemi-
sphere (ymax ¼ 29ꢁ) of which 605 reflections (Rint ¼ 0.048) were
symmetrically non-equivalent. Bromine atoms were refined
anisotropically, whereas the carbon atoms and the hydrogens
New J. Chem., 2002, 26, 582–592
589

View Article Online
Charge-transfer photoirradiation of the benzene/bromine
(calculated from a riding geometric model) were refined iso-
tropically. The final discrepancy factors were R1 ¼ 0.063 and
wR2 ¼ 0.178 for 549 reflections with I 5 2s(I). The absolute
structure was determined with a Flack parameter of w ¼ À0.1(3).
charge-transfer complex as a fluid mixture of neat compounds.
An equimolar mixture of bromine (0.1 ml, 2 mmol) and
benzene (0.17 ml, 2 mmol) was prepared in a 1 mm quartz
cuvette fitted with a Schlenk adapter under an argon atmo-
sphere at room temperature. The cuvette was placed in a
Dewar equipped with quartz windows and it was irradiated
with UV light from a medium-pressure mercury lamp at
either 0 ꢁC (ice–water bath) or at room temperature (see
Results) under an argon atmosphere for 6 h. UV light was
focused through an aqueous IR filter and the CT-band iso-
lation filter (see above). As the thermal control, the same
mixture was placed in glass tube wrapped with aluminum foil
and the tube was kept in the same Dewar (to ensure the same
time for the photoreaction). After reaction, the mixtures were
analyzed as described above.
Toluene/dibromine complex. Formula C7H8Br2 (M ¼ 251.95);
triclinic, space group P-1; a ¼ 5.516(1), b ¼ 11.715(2), c ¼
+
13.551(3) A, a ¼ 79.76(1), b ¼ 80.89(1), g ¼ 85.56(1)ꢁ, U ¼
+
849.8(3) A3, Z ¼ 4; Dcalc ¼ 1.969 g cmÀ3, m(MoKa) ¼ 94.5 cmÀ1
;
19282 reflections were collected over the reciprocal sphere
(ymax ¼ 29ꢁ) of which 4356 reflections (Rint ¼ 0.077) were sym-
metrically non-equivalent. All non-hydrogen atoms were refined
anisotropically; thehydrogenswere positioned usinga riding and
rotating geometric model and refined isotropically. The final
discrepancy factors were R1 ¼ 0.039 and wR2 ¼ 0.080 for 3051
reflections with I 5 2s(I).
Charge-transfer photoirradiation of toluene/bromine charge-
transfer complex as a fluid mixture of neat compounds.
Bromine (0.1 ml, 2 mmol) was added to toluene (0.2 ml,
2 mmol) cooled to À78 ꢁC in a 1 mm quartz cuvette fitted with
a Schlenk adapter under an argon atmosphere. The mixture
was slightly warmed for homogenization and the cuvette was
placed in the Dewar with quartz windows (dry ice–methanol
bath with temperature À65 ꢁC). The liquid mixture was irra-
diated with UV light from a medium pressure mercury lamp
(see above) at À65 ꢁC under an argon atmosphere for 6 h. The
equivalent thermal control was placed in a glass tube wrapped
with aluminum foil and the tube was kept in the same Dewar
for the same period. After reaction, the mixtures were analyzed
as described above.
Bromohexamethylbenzenium tribromide (s-complex) as the
tris(dibromine) solvate. Formula C12H18Br10 (M ¼ 961.36);
monoclinic (regular twin), space group P21/c; a ¼ 8.7047(4),
+
b ¼ 17.9315(8), c ¼ 15.4610(7) A, b ¼ 90.078(2)ꢁ, U ¼ 2413.3(2)
+
A3, Z ¼ 4; Dcalc ¼ 2.646 g cmÀ3, m(MoKa) ¼ 166.0 cmÀ1; 19 084
reflections were collected over the reciprocal sphere
(ymax ¼ 35ꢁ) of which 9809 reflections (Rint ¼ 0.055) were non-
equivalent. All non-hydrogen atoms were refined aniso-
tropically; the hydrogens were positioned using a riding and
rotating geometric model and refined isotropically. The final
discrepancy factors were R1 ¼ 0.071 and wR2 ¼ 0.1568 for 6223
reflections with I 5 2s(I).
Bromohexamethylbenzenium hexafluoroantimonate (s-
complex). Formula C12H18BrF6Sb (M ¼ 477.92); monoclinic,
space group P21/n; a ¼ 7.0691(3), b ¼ 11.0782(5), c ¼
2.094 g cmÀ3, m(MoKa) ¼ 45.1 cmÀ1; 18705 reflections were
collected over the reciprocal sphere (ymax ¼ 35ꢁ) of which 6610
reflections (Rint ¼ 0.025) were non-equivalent. All non-hydro-
gen atoms were refined anisotropically; the hydrogens were
positioned using a riding and rotating geometric model and
refined isotropically. The final discrepancy factors were
R1 ¼ 0.025 and wR2 ¼ 0.055 for 5641 reflections with I 5 2s(I).
Procedure for the charge-transfer irradiation of bromine
complexes with benzene and toluene in the solid state. The solid
complex, as an equimolar mixture of bromine (0.1 ml, 2 mmol)
and arene [benzene (0.17 ml, 2 mmol) or toluene (0.2 ml,
2 mmol)], was irradiated at the CT band for 6 h in a 1 mm
quartz cuvette under an argon atmosphere with the apparatus
described above, in a Dewar with quartz windows at À78 ꢁC
(dry ice–methanol bath). The dark thermal control was carried
out in a glass tube wrapped with aluminum foil, which was
placed in the same Dewar. After reaction, the mixtures were
dissolved in chloroform and the products analyzed by stan-
dard GC-MS methods. The Quantum Yields of photoreaction
products were measured with the aid of a medium-pressure
(500 W) mercury lamp. The intensity of the lamp was deter-
mined at l ¼ 313 nm with a freshly prepared potassium ferri-
oxalate actinometer solution,53 under the same conditions as
used for the photoreactions of bromine/arene complexes (fil-
ters, apparatus).
+
+
19.5085(9) A, b ¼ 97.180(1)ꢁ, U ¼ 1515.8(1) A3, Z ¼ 4; Dcalc
¼
Thermal transformation of the bromine complexes of benzene
and toluene
Equimolar mixtures of bromine (0.1 ml, 2 mmol) and arene
[benzene (0.17 ml, 2 mmol) or toluene (0.2 ml, 2 mmol)] were
prepared in glass tubes and were kept in the dark at different
temperatures during 3 or 6 h (see Tables 2 and 3) in a dry ice–
methanol bath. The reaction mixtures were analyzed as
described above.
Acknowledgement
Charge-transfer photoexcitation of bromine complexes of
benzene and toluene
We thank a helpful reviewer for insightful comments/correc-
tions of the manuscript, and the National Scientific Founda-
tion and R. A. Welch Foundation for financial support.
Filter for the isolation of UV light for the specific irradiation
of the charge-transfer band of arene/bromine complexes. A
medium-pressure mercury lamp (500 W) was used for the
photoirradiation. For the isolation of UV light from the
medium-pressure mercury lamp in the region of the charge-
transfer band of arene/bromine complexes, we used the com-
bination of a colored glass filter UG-5 (Oriel Instruments) and
filter solutions consisting of: 1 M solution of CoSO4 in 5%
aqueous H2SO4 ; 2 M solution of NiSO4 and 0.05 M solution
of CuSO4 in 5% aqueous H2SO4 ; and 0.05 M solution of Br2
in CCl4 in quartz cuvettes. This filter combination had a
transmittance from 280 nm to 350 nm with a maximum at 320
nm (for the transmittance characteristics, see the ESI).
References and notes
1
2
H. A. Benesi and J. J. Hildebrand, J. Am. Chem. Soc., 1949, 71,
2703.
(a) R. M. Keefer and L. J. Andrews, J. Am. Chem. Soc., 1950, 72,
4677; (b) For a review see: L. J. Andrews and R. M. Keefer,
Molecular Complexes in Organic Chemistry, Holden-Day, San
Francisco, 1964.
3
4
R. S. Mulliken, J. Am. Chem. Soc., 1950, 72, 600.
(a) R. Foster, Organic Charge-Transfer Complexes, Academic,
New York, 1969; (b) G. Briegleb, Elektronen Donor-Acceptor
Komplexe, Springer Verlag, Berlin, 1973.
590
New J. Chem., 2002, 26, 582–592

View Article Online
21 For a discussion of this point, see: (a) S. M. Hubig, S. V.
Lindeman and J. K. Kochi, Coord. Chem. Rev., 2000, 200–202, 831;
5
6
(a) Molecular Interactions. From van der Waals to Strongly Bound
Complexes, ed. S. Scheiner, Wiley, New York, 1997; (b) P. Hobza
and R. Zahradnik, Intermolecular Complexes, Elsevier, New
York, 1988; (c) H. Ratajczak, W. J. Orville-Thomas, Molecular
Interactions, Wiley, New York, 1980–82, vol. 1–3. EDA com-
plexes are observed variously in the gas phase, rare-gas matrix,
solution, or the crystalline solid state.
See also: (a) J. Collin and L. D’Or, J. Chem. Phys., 1955, 23, 397;
(b) H. B. Friedrich and W. B. Person, J. Chem. Phys., 1966, 44,
2161; (c) G. DeBoer, J. W. Burnet, A. Fujimoto and M. A.
Young, J. Phys. Chem., 1996, 100, 14882; (d) P. Y. Cheng,
D. Zhong and A. H. Zewail, J. Chem. Phys., 1996, 105, 6216; (e)
J. T. Su and A. H. Zewail, J. Phys. Chem. A, 1998, 102, 4082.
E. M. Kosower, Prog. Phys. Org. Chem., 1965, 3, 81.
A. K. Colter and M. R. J. Dack, in Molecular Complexes, ed.
R. Foster, Crane, Russak & Co., New York, 1973, vol. 1, ch. 6,
p. 301 and 1974, vol. 2, ch. 1, p. 1.
(a) F. C. Grozema, R. W. J. Zijlstra, M. Swart and P. T. Van
Duijhen, Int. J. Quantum Chem., 1999, 75, 709; (b) A. M. Mebel,
H. L. Lin and S. H. Lin, Int. J. Quantum Chem., 1999, 72, 307.
See also: (b) P. Le Mague
`
Organometallics, 2001, 20, 115; (c) P. Le Mague
res, S. V. Lindeman and J. K. Kochi,
res, S. V. Linde-
`
man and J. K. Kochi, J. Chem. Soc., Perkin T rans. 2, 2001, 1180.
22 (a) The descriptor Z is the arene hapticity as defined by F. A.
Cotton and G. Wilkinson, Advanced Inorganic Chemistry, Wiley,
New York, 5th edn., 1988, p. 38; (b) In structural comparisons
discussed hereinafter, for convenience we make no distinction
among the various halogens (Cl2 , Br2 or I2).
23 According to Mulliken,3,15 the dative contribution from
[Br2À,C6H6þ] plays a role in the noncovalent interaction. As such,
for charge-transfer complexes the intermolecular contraction is
largely due to the Coulombic and dispersion forces. See also:
H. O. Hooper, J. Chem. Phys., 1964, 41, 599.
24 O. Hassel and K. O. Strꢀmme, Acta Chem. Scand., 1959, 13, 275.
25 K.-M. Marstokk and K. O. Strꢀmme, Acta Crystallogr., Sect. B,
1968, 24, 713.
7
8
9
26 S. Hauge and K. Maroy, Acta Chem. Scand., 1996, 50, 1095.
27 V. Janickis, Acta Chem. Scand., 1999, 53, 188.
10 (a) H. Bai and B. S. Ault, J. Phys. Chem., 1990, 94, 199; (b) E. E.
Ferguson, J. Chem. Phys., 1956, 25, 577; (c) L. Fredin and
B. Nelander, J. Am. Chem. Soc., 1974, 96, 1672; (d) S. A. Cooke,
C. M. Evans, J. H. Holloway and A. C. Legon, J. Chem. Soc.,
Faraday T rans., 1998, 94, 2295; (e) L. Fredin and B. Nelander,
Mol. Phys., 1974, 27, 885; ( f ) W. B. Person, C. F. Cook and H. B.
Friedrich, J. Chem. Phys., 1967, 46, 2521.
11 (a) S. S. C. Ammal, S. P. Ananthavel, P. Venuvanalingam and
M. S. Hegde, J. Phys. Chem. A, 1998, 102, 532; (b) A. Matsuzawa
and Y. Osamura, Bull. Chem. Soc. Jpn., 1997, 70, 1531; However,
see: (c) G. Milano, G. Guerra and L. Cavallo, Eur. J. Inorg.
Chem., 1998, 1513; (d) E. Kochanski and J. Prissette, Nouv. J.
Chem., 1980, 4, 509; (e) I. Jano, T heor. Chim. Acta, 1985, 66, 341;
( f ) E. G. Cook and J. C. Shug, J. Chem. Phys., 1970, 53, 723.
12 (a) O. Hassel and K. O. Strꢀmme, Acta Chem. Scand., 1958, 12,
1146; See also: (b) O. Hassel, Mol. Phys., 1958, 1, 241; (c) O.
Hassel and C. Rꢀmming, Quart. Rev., 1962, 16, 1; (d) O. Hassel
and K. O. Strꢀmme, Acta Chem. Scand., 1959, 13, 1781.
28 J. O. Howell, J. M. Goncalves, C. Amatore, L. Klasinc, R. M.
Wightman and J. K. Kochi, J. Am. Chem. Soc., 1984, 106, 3968.
29 Cambridge Crystallographic Database, Release Fall 2000.
30 P. Groth and O. Hassel, Acta Chem. Scand., 1964, 18, 402.
31 O. Hassel and J. Hvoslef, Acta Chem. Scand., 1954, 8, 873.
32 Irradiation of the bromine absorption at l > 380 nm leads to
homolysis of dibromine, and products of bromine-atom addition
to benzene (hexabromocyclohexane, isomeric tribromobenzene,
etc.) are produced (unpublished results).
33 (a) The side-chain bromination of toluene to benzyl bromide was
the competing (radical-chain) side reaction that could be con-
trolled by the addition of 2,3-dimethylbutane, as the bromine-
atom trap; (b) It is also possible that (some) benzyl bromide was
derived from the toluene cation-radical via rapid deprotonation to
benzyl radical, etc.; (c) For the unusual substitution patterns ob-
served in vapor-phase brominations, see: E. C. Kooyman, Pure
Appl. Chem., 1963, 7, 193.
34 (a) T. Dhanasekaran, unpublished results. The crystal also con-
tained 3 dibromines as solvates, which are not shown for clarity of
the structure in Fig. 5(A); (b) It is important to note the me-
chanistic equivalency of second-order and third-order kinetics
(involving the presence of one and two Br2 acceptors, respectively,
in the rate-limiting transition states) for electrophilic bromina-
tions, see: S. Fukuzumi and J. K. Kochi, Int. J. Chem. Kinet.,
1983, 15, 249.
13 It is important to pay tribute (some 40 years later) to the un-
precedented (experimental) achievement of Hassel and Strꢀmme
in obtaining single crystals of such weak (low-melting) unstable
complexes for X-ray crystallographic analysis.
14 A. Bondi, J. Phys. Chem., 1964, 68, 441.
15 (a) R. S. Mulliken, J. Am. Chem. Soc., 1952, 74, 811; (b) R. S.
Mulliken and W. B. Person, Molecular Complexes. A L ecture and
Reprint Volume, Wiley, New York, 1969.
35 In Fig. 5(A), the tribromide anion is located from the partially
À
16 (a) In this temperature range (from À40 to À50 ꢁC), significant
thermal motion caused a weak diffraction pattern and high-angle
reflections were absent. [In our repetition of this experiment we also
observed only very weak diffraction up to y < 15ꢁ owing to large
thermal motion in the crystal.] As a result of such insufficient dif-
fraction data, Hassel and Strꢀmme12a were unable to definitively
assign the space group to either C2/m (centrosymmetric) or C2 and
Cm (noncentrosymmetric). They arbitrarily chose C2/m and the
centrosymmetric structure A, although they recognized that some
other (unsymmetrical) structures such as B (vide infra, with 6-fold
disorder) were also possible. Nevertheless, in their subsequent
papers12b–d Hassel and Strꢀmme put forward structure A as the
true one (which has been widely accepted by others) despite severe
discrepancies in the infrared (spectroscopic) data.10 Recent theo-
retical calculations have also cast doubt on the validity of the
symmetric structure A.9,11; (b) Hassel and Strꢀmme12 chose a
monoclinic space group for the [C6H6 ,Br2] complex at À40 to
À50 ꢁC (vide supra), whereas the trigonal modification pre-
dominates below À70 ꢁC (vide infra). Crystallographically, except
for disorder and symmetry patterns, these two modifications differ
only by a mere shift of donor/acceptor chains relative to each other.
17 For the topochemical control of solid-state reactions, see: (a) M. D.
Cohen, G. M. J. Schmidt and F. I. Sonntag, J. Chem. Soc., 1964,
2000;(b) G. M. J. Schmidt, Pure Appl. Chem., 1971, 27, 647; (c) T. Y.
Fu, Z. Liu, J. R. SchefferandJ. Trotter, J. Am. Chem. Soc., 1993, 115,
12 202; (d) V. Enkelmann, G. Wegner, K. Novak and K. B. Wa-
gener, J. Am. Chem. Soc., 1993, 115, 10 390; (e) H. E. Zimmerman
and M. J. Zuraw, J. Am. Chem. Soc., 1989, 111, 7974; ( f ) J. H. Kim,
S. V. Lindeman and J. K. Kochi, J. Am. Chem. Soc., 2001, 123, 4951.
18 S. Fukuzumi and J. K. Kochi, J. Am. Chem. Soc., 1982, 104, 7599.
19 S. Fukuzumi and J. K. Kochi, J. Org. Chem., 1981, 46, 4116. See
also ref. 41.
disordered crystal structure of C6(CH3)6BrþBr3 in ref. 34. For
the preparation of bromohexamethylbenzenium hexafluoro-
antimonate, see ref. 39a.
36 For a related study of electrophilic aromatic substitutions by
charge-transfer photoactivation of the charge-transfer complexes,
see: (a) J. K. Kochi, Adv. Phys. Org. Chem., 1994, 29, 185; (b) S. M.
Hubig and J. K. Kochi, J. Am. Chem. Soc., 2000, 122, 8279; (c)
E. K. Kim, T. M. Bockman and J. K. Kochi, J. Am. Chem. Soc.,
1993, 115, 7051.
37 (a) If the symmetrical (delocalized) structure A actually exists, it
must correspond to a single potential energy minimum, and
therefore be increasingly populated at lower temperatures, con-
trary to experiment (vide infra). Therefore, the structure A ob-
served only at higher temperatures must correspond to the
superposition of the (6-fold) disordered structure B;37c (b) The
fact that there are two crystalline modifications (at À50
and À150 ꢁC corresponding to monoclinic and trigonal space
groups)16b does not necessarily point to two separate structures (A
and B) since it is well-known that phase transitions often ac-
company low-temperature ordering of (disordered) molecular
structures. See, for example: H. Cailleau, J.-L. Baudour,
J. Meinnel, A. Dworkin, F. Moussa and C. M. E. Zeyen, Faraday
Discuss. Chem. Soc., 1980, 69, 7; (c) To identify the structural
ambiguity, Hassel and Strꢀmme themselves cautiously (and wi-
sely) state:12a ‘‘We think that the structure [A] derived from our
presented material is certainly essentially correct, but hope to be
able to supplement the investigation with material obtained at
lower temperatures. Such investigations might contribute toward
clearing up the question (mentioned above) regarding the poten-
tial energy curve of the bromine atom.’’ Now (40 years later) we
herein deliver this low-temperature study that reveals the asym-
metric shape of the potential energy curve corresponding to
structure B that was previously discarded by Hassel and Strꢀmme;
(d) See also discussion in ref. 10f; (e) Structure B is also favored by
high-level theoretical calculations.9,11
20 This procedure followed the pioneering work of Hassel and
coworkers.12 For a preliminary communication, see A. V. Vasi-
lyev, S. V. Lindeman and J. K. Kochi, Chem. Commun., 2001, 909.
New J. Chem., 2002, 26, 582–592
591

þ
À
View Article Online
or ck(DA*) and the transition energy from C(DA) to Ci(D A ) is
38 (a) Such a high reactivity of crystalline complexes relative to that
in solution occurs despite the somewhat diminished charge-
transfer polarization arising from the quasi-chain structure (vide
infra); (b) This low-temperature observation is tantamount to the
classic criterion established by Kiselev and Miller for the direct
involvement of the charge-transfer complex in a chemical trans-
formation. See: V. D. Kiselev and J. G. Miller, J. Am. Chem. Soc.,
1975, 97, 4036.
given as in eqn. (iii). Thus, any significant deviation from eqn. (iii)
owing to the interaction between c(DA), ci(DþAÀ), cj(D*A), and
ck(DA*) in eqn. (i) and (ii) is unlikely for weak complexes. For
the nature of CN in molecular complexes, see: K. Morokuma, Acc.
Chem. Res., 1977, 10, 294].
42 For a recent review of the enhanced chemical reactivity of donor
cation-radicals and acceptor anion-radicals, see: R. Rathore and
J. K. Kochi, Adv. Phys. Org. Chem., 2000, 35, 193.
43 (a) S. Fukuzumi and J. K. Kochi, J. Am. Chem. Soc., 1981, 103,
7240; (b) Thermal electron transfer from electron-rich aromatic
donors to Br2 has been observed by L. Eberson, M. P. Hartshorn,
F. Radner and O. Persson, Chem. Commun., 1996, 215; See also
J. K. Kochi, Tetrahedron L ett., 1975, 41; (c) See also footnote 34b.
44 S. Fukuzumi and J. K. Kochi, Bull. Chem. Soc. Jpn., 1983, 56,
969.
45 Compare (a) W. A. Chupka, J. Berkowitz and D. Gutman, J.
Chem. Phys., 1971, 55, 2724; (b) S. D. Malone and J. F. Endicot,
J. Phys. Chem., 1972, 76, 2223; (c) The formation of the cationic
s-complex in eqn. (13) is likely to proceed via prior mesolytic
cleavage of Br2ꢀÀ followed by the very fast collapse of the caged
(arene cation radical/bromine atom) pair,46 that is:
39 (a) S. M. Hubig and J. K. Kochi, J. Org. Chem., 2000, 65, 6807; (b)
Additives such as Lewis acids (that can sequester bromide) will
promote the transfer of the bromine cation.
40 See: (a) P. B. D. de la Mare, Electrophilic Halogenation, Cam-
bridge University Press, London, 1976; (b) R. O. C. Norman and
R. Taylor, Electrophilic Substitution in Benzenoid Compounds,
Elsevier, New York, 1965; (c) L. M. Stock, Aromatic Substitution
Reactions, Prentice-Hall, Englewood Cliff, NJ, USA, 1968; (d)
T. H. Lowry and K. S. Richardson, Mechanism and T heory in
Organic Chemistry, Harper and Row Publishers, New York, 3rd
edn., 1987.
41 According to Mulliken,3,15 hnCT corresponds to the electronic
excitation from the ground state CN of the complex to the excited
singlet state CE as described in eqn. (i) and (ii), respectively, where
c(DA) and ci(DþAÀ) represent the wave functions for the
X
X
CN ¼ acðDAÞ þ
biciðDþAÀÞ þ
cjcjðD*AÞ
i
j
X
þ
dkckðDA*Þ
ðiÞ
k
This is followed by the rapid (irreversible) deprotonation of the s-
adduct by the accompanying bromide. The direct collapse of the
ion pair to a hypervalent adduct is also possible.
and
X
X
CE ¼ a*cðDAÞ þ
bi *ciðDþAÀÞ þ
cj*cjðD*AÞ
46 The amelioration of such retarding factors, which result in in-
efficient photoprocesses for [C6H6 , Br2], has been shown to lead
then to a variety of other (efficient) charge-transfer photoprocesses
with other types of donor/acceptor pairs.42
47 For experimental support of such a (thermal) electron-transfer
process, see: L. Eberson, M. P. Hartshorn, F. Radner and
O. Persson, J. Chem. Soc., Perkin T rans. 2, 1998, 59.
48 For an example of a rapid rate of such a homolytic (radical-
radical) combination, see: E. K. Kim, T. M. Bockman and J. K.
Kochi, J. Am. Chem. Soc.ꢀ, 1993, 115, 3091 in the case of the
coupling of (ArHꢀþ þ NO2 ) pertinent to the photoactivation of
electrophilic aromatic nitration.
49 J. P. Novak, Quantitative Analysis by Gas Chromatography,
M. Dekker, New York, 2nd edn., 1988.
50 G. M. Sheldrick, SADABS Bruker–Siemens Area Detector Ab-
i
j
X
þ
dk *ckðDA*Þ
k
ðiiÞ
no-bond structure and ith zero-order electron transfer singlet
state, respectively. cj(D*A) and ck(DA*) are the wave functions
of the zero-order singlet states corresponding to the jth and kth
local excitations within the electron donor and acceptor, respec-
tively. For weak EDA complexes of the type between bromine and
arenes, in which the overlap integrals between the donor and ac-
ceptor orbitals are small, the transition energy can be expressed
(to first order) as
hnCT ¼ IDl À EAm þ o
ðiiiÞ
sorption and Other Corrections, v. 2.03, University of Gottingen,
¨
Germany, 2000..
In eqn. (iii), hniCT corresponds to the transition energy from
C(DA) to Ci(DþAÀ), ID is the lth ionization potential of the
51 G. M. Sheldrick, SHELXS 86–Program for Crystal Structure
Solutions, University of Gottingen, Germany, 1986.
¨
l
m
donor, EA is the mth electron affinity of the acceptor, and o is
the interaction energy between the donor and acceptor moieties in
the ci(DþAÀ) state. [For weak complexes of the type described
here, the nondiagonal terms in the secular equation derived from
eqn. (i) and (ii) are neglected, CE is given as ci(DþAÀ), cj(D*A),
52 G. M. Sheldrick, SHELXL 93–Program for Crystal Structure
Refinement, University of Gottingen, Germany, 1993.
¨
53 J. G. Calvert and J. N. Pitts, Photochemistry, Wiley, New York,
1966, pp. 783–786.
592
New J. Chem., 2002, 26, 582–592