Preassociation, free-ion, and ion-pair pathways in the electrophilic bromination of substituted cis- and trans-stilbenes in protic solvents

Article abstract of DOI:10.1021/ja9700461

Rates and products of electrophilic bromination of ring-substituted cis- and trans-stilbenes have been investigated in acetic acid, trifluoroethanol, ethanol, methanol, and water-methanol mixtures. The mY(Br) relationships (linear for nucleophilic solvent

Full text of DOI:10.1021/ja9700461

12492
J. Am. Chem. Soc. 1997, 119, 12492-12502
Preassociation, Free-Ion, and Ion-Pair Pathways in the
Electrophilic Bromination of Substituted cis- and
trans-Stilbenes in Protic Solvents
Marie-Franc¸oise Ruasse,*,† Giacomo Lo Moro,‡ Bernard Galland,†
Roberto Bianchini,§ Cinzia Chiappe,*,‡ and Giuseppe Bellucci#
Contribution from Institut de Topologie et de Dynamique des Syste`mes, UniVersite´ Paris
7sDenis Diderot, associe´ au CNRS, URA 34, 1, rue Guy de la Brosse, 75005 Paris, France,
Dipartimento di Chimica Bioorganica, UniVersita di Pisa, Via Bonamo 33, Pisa 56126, Italy,
and Dipartimento di Chimica Organica “U. Schiff”, UniVersita di Firenze, Via G. Capponi 9,
Firenze 50100, Italy
ReceiVed January 6, 1997. ReVised Manuscript ReceiVed October 7, 1997X
Abstract: Rates and products of electrophilic bromination of ring-substituted cis- and trans-stilbenes have been
investigated in acetic acid, trifluoroethanol, ethanol, methanol, and water-methanol mixtures. The mYBr relationships
(linear for nucleophilic solvents only, with m ) 0.8), the deviations of the two nonnucleophilic solvents from the
mYBr plots (∆AcOH and ∆TFE positive, negative, or negligible), the kinetic solvent isotope effects (kMeOH/kMeOD
)
1.1-1.6), the chemoselectivity (predominant dibromide, DB, or solvent-incorporated adducts, MA), and the high
dependence of the stereochemistry on the solvent and the substituents (from stereoconvergency to stereospecificity)
are discussed and interpreted in terms of a mechanistic scheme, analogous to the Jencks scheme for aliphatic
nucleophilic substitutions, in which preassociation, free-ion, and ion-pair pathways compete. In particular, the
stereochemical outcome of these reactions is consistent with a marked change in the nucleophilic partners of the
product-forming ionic intermediate arising from different ionization routes. Return, i.e. change in the rate-limiting
step from ionization to product formation, is shown to be related to substituent-dependent, but not solvent-dependent,
bromine bridging.
Introduction
the similarity between the bromination mechanism and that of
SN1-like solvolysis and nucleophilic substitutions previously
revealed by kinetic solvent and substituent effects.8 Both
reactions involve the formation of an ion pair by ionization of
a neutral substrate, CTC, or alkyl derivative. For both, the
products are obtained by nucleophilic trapping of a cationic
intermediate. Moreover, nucleophilic solvent assistance has
been shown to occur in the ionization steps of the two
reactions.7,9,10 With regard the last steps and, in particular, the
nature of the ionic species from which the products are formed,
the corresponding data for bromination are almost inexistent,
whereas in solvolysis or nucleophilic substitutions, it is widely
The multistep mechanism of electrophilic bromination of
ethylenic compounds of Scheme 1 is well documented by a large
variety of kinetic and product investigations.1 Recent work on
this basic reaction of organic chemistry is mainly focused on
its early steps2 and, in particular, on the formation3 of bromine-
olefin charge-transfer complexes, CTC, and on the structure of
the bromonium ion intermediates.4 Many current studies deal
with the reactivity of crowded double bonds,5 which involves
return enforced by a slow product-forming step because of steric
retardation of the nucleophilic trapping of congested bromonium
ions. In this context, much attention has been paid to the
reversibility of the CTC ionization leading to the ionic
intermediate.5-7 The occurrence of this reversibility emphasizes
(5) (a) Brown R. S.; Slebocka-Tilk, H.; Bennet, A. J.; Bellucci, G.;
Bianchini, R.; Ambrosetti, R. J. Am. Chem. Soc. 1990, 112, 6310. (b)
Nagorski, R. W.; Slebocka-Tilk, H.; Brown, R. S. J. Am. Chem. Soc. 1994,
106, 419. (c) Slebocka-Tilk, H.; Motallebi, S.; Nagorski, R. W.; Turner,
P.; Brown, R. S.; McDonald, R. J. Am. Chem. Soc. 1995, 107, 8769. (d)
Slebocka-Tilk, H.; Gallagher, D.; Brown, R. S. J. Org. Chem. 1996, 61,
3458.
(6) (a) Bellucci, G.; Chiappe, C.; Marioni, F. J. Am. Chem. Soc. 1987,
109, 515. (b) Bellucci, G.; Bianchini, R.; Chiappe, C.; Marioni, F.; Spagna,
R. J. Am. Chem. Soc. 1988, 110, 546. (c) Bellucci, G.; Bianchini, R.;
Chiappe, C.; Brown, R. S.; Slebocka-Tilk, H. J. Am. Chem. Soc. 1991,
113, 8012. (d) Bellucci, G.; Chiappe, C.; Marioni, F.; Marchetti, F. J. Phys.
Org. Chem. 1991, 4, 387. (e) Bellucci, G.; Chiappe, C. J. Chem. Soc., Perkin
Trans. 2 1997, 581.
* Address correspondence to M. F. Ruasse (fax 33 1 44 27 68 14, e-mail
ruasse@paris7.jussieu.fr) or C. Chiappe (fax 39 50 43321).
† Universite´ Paris 7 - Denis Diderot.
‡ Universita di Pisa.
§ Universita di Firenze.
# In memoriam of Professor Giuseppe Bellucci, deceased March 3, 1996.
X Abstract published in AdVance ACS Abstracts, December 1, 1997.
(1) (a) de la Mare, P. B. D.; Bolton, R. Electrophilic Additions to
Unsaturated Systems, 2nd ed.; Elsevier: New York, 1982; pp 136-197.
(b) V’yunok, K. A.; Ginak, A. I. Russ. Chem. ReV. (Engl. Transl.) 1981,
50, 151. (c) Schmid, G. H. In The Chemistry of Double-Bonded Functional
Groups; Patai, S., Ed.; Wiley: New York, 1989; Supplement A, Vol. 3,
Part 1, p 699. (d) Ruasse, M. F. AdV. Phys. Org. Chem. 1993, 28, 207.
(2) Brown, R. S. Acc. Chem. Res. 1997, 30, 131.
(7) (a) Ruasse, M. F.; Motallebi, S.; Galland, B. J. Am. Chem. Soc. 1991,
113, 3440. (b) Ruasse, M. F.; Motallebi, S.; Galland, B.; Lomas, J. S. J.
Org. Chem. 1990, 55, 2298.
(3) (a) Bellucci, G.; Bianchini, R.; Ambrosetti, R. J. Am. Chem. Soc.
1985, 107, 2464. (b) Bellucci, G.; Bianchini, R.; Chiappe, C.; Lenoir, D.;
Attar, A. J. Am. Chem. Soc. 1995, 117, 6243. (c) Bellucci, G.; Chiappe,
C.; Bianchini, R.; Lenoir, D.; Herges, R. J. Am. Chem. Soc. 1995, 117,
12001.
(4) (a) Bennet, A. J.; Brown, R. S.; McClung, R. E. D.; Klobukowski,
M.; Aarts, G. H. M.; Santarsiero, B. D.; Bellucci, G.; Bianchini, R. J. Am.
Chem. Soc. 1991, 113, 8532. (b) Brown, R. S.; Nagorski, R. W.; Bennet,
A. J.; McClung, R. E. D.; Aarts, G. H. M.; Klobukowski, M.; McDonald,
R.; Santarsiero, B. D. J. Am. Chem. Soc. 1994, 116, 2448.
(8) (a) Ruasse, M. F.; Motallebi, S. J. Phys. Org. Chem. 1991, 4, 527.
(b) Ruasse, M. F. Acc. Chem. Res. 1990, 23, 87.
(9) (a) Ruasse, M. F.; Zhang, B. L. J. Org. Chem. 1984, 49, 3207. (b)
Ruasse, M. F.; Lefebvre, E. J. Org. Chem. 1984, 49, 3210.
(10) (a) Bentley, T. W.; Schleyer, P. v. R. AdV. Phys. Org. Chem. 1977,
14, 1. (b) Raber, D. J.; Neal, W. C.; Dukes, M. D.; Harris, J. M.; Mount,
D. L. J. Am. Chem. Soc. 1978, 100, 8137. (c) Harris, J. M.; Mount, D. L.;
Smith, M. R.; Neal, W. C.; Dukes, M. D.; Raber, D. J. J. Am. Chem. Soc.
1978, 100, 8147. (d) Bentley, T. W.; Bowen, C. T.; Morten, D. H.; Schleyer,
P. v. R. J. Am. Chem. Soc. 1981, 103, 5466.
S0002-7863(97)00046-2 CCC: $14.00 © 1997 American Chemical Society

Bromination of Stilbenes in Protic SolVents
J. Am. Chem. Soc., Vol. 119, No. 51, 1997 12493
Scheme 1
Scheme 2
obtained from free ions, 5, ion pairs, 2, or ion-dipole
sandwiches, 4, formed via preassociation or ion-pair pathways.
Species 4 in which bromocation I+ is associated with its counter-
bromide ion and, on its opposite side, with a molecule of a
nucleophilic solvent, is called an ion-dipole sandwich by
analogy with the nomenclature used in nucleophilic substitu-
tions.11 In Scheme 2, the concerted pathway which can occur
in nucleophilic substitution (SN2 route) is not considered since
concerted bromination of carbon-carbon double bonds has
never been observed.1 The reported kinetic and product results
exhibit subtle variations in rate- and product-determining ionic
species which are rationalized in terms of Scheme 2.
Results and Discussion
1. Bromination Kinetics. Bromination rate constants of cis-
stilbenes 6a-11a and of their trans isomers 6b, 7b, 9b, and
11b were measured in a variety of solvents, acetic acid, ethanol,
X
X
X′
H
C
C
C
C
Scheme 3
H
H
H
b
a
X′
6: X = CH3O, X′ = H; 7: X = X′ = CH3; 8: X = CH3, X′ = H
9: X = X′ = H; 10: X = H, X′ = CF3; 11: X = X′ = CF3
trifluoroethanol, methanol, and the aqueous mixtures of the
last: M10, M20, and M30 (10-90, 20-80, and 30-70 H2O-
MeOH, v/v, respectively). The results are shown in Table 1.
The bromination kinetics followed by conventional spectropho-
tometry and/or by stop-flow technique12 were rigorously second
order, first order in olefin, and first order in bromine (eq 1),
regardless of the stilbene or the solvent. In particular in acetic
accepted that free ions, ion pairs, or triple ion complexes can
be the precusors of the reaction products.11
In this paper, we report a kinetic and product investigation
of the bromination of uncrowded ring-substituted stilbenes in
protic solvents which sheds light not only on solvent assistance
and bromine bridging in the ionization step but also on the role
of counterion and solvent association with the ionic intermediate
in the product-forming step. We show that the marked
dependence of kinetics and stereochemistry on the solvent and
the substituents is consistently interpreted in terms of Scheme
2, which is analogous to Scheme 3 well established for aliphatic
nucleophilic substitutions.11 In Scheme 2 for bromination, the
substrate is the CTC, 1; the nucleophile can be the solvent; I+
is the bromocationic intermediate which, depending on the
double-bond substituents, can be open, R-I+, fully bridged, â-I+,
or weakly bridged, γ-I+; the products are the solvent-
incorporated, or mixed adducts, MA, and the dibromides DB.
As in nucleophilic substitutions, the CTC ionization can be
reversible or irreversible and assisted (preassociation route, 1
f 3 f 4) or unassisted (1 f 2), and the products can be
v ) k [Ol] [Br2]
(1)
acid, second-order bromine terms, which have previously been
observed,13 were not found with the small bromine concentra-
tions used. In methanol and its aqueous mixtures, although,
the tribromide ion concentration14 was not controlled by the
addition of external bromide ions (eq 2), eq 1 was always
-
Br2 + Br- h Br3
(2)
rigorously followed, showing that the kinetic term related to
bromide concentration15 (eq 3) was not significant under these
experimental conditions.
kexp (1 + K[Br-]) ) k + KkBr [Br-]
(3)
3
(12) Ambrosetti, R.; Bellucci, G.; Bianchini, R.; Fontana, E.; Ricci, D.
J. Phys. Chem. 1994, 98, 1620.
(11) (a) Ridd, J. H. AdV. Phys. Org. Chem. 1979, 16, 1. (b) Jencks, W.
P. Acc. Chem. Res. 1980, 13, 161. (c) Cox, M. M.; Jencks, W. P. J. Am.
Chem. Soc. 1981, 103, 572. (d) Jencks, W. P. Chem. Soc. ReV. 1981, 10,
345. (e) Richard, J. P.; Jencks, W. P. J. Am. Chem. Soc. 1984, 106, 1373.
(f) Richard, J. P.; Rothenberg, M. E.; Jencks, W. P. J. Am. Chem. Soc.
1984, 106, 1361. (g) Amyes, T. L.; Jencks, W. P. J. Am. Chem. Soc. 1989,
111, 7888. (h) Eldin, S.; Jencks, W. P. J. Am. Chem. Soc. 1995, 117, 4851.
(13) (a) Walker, I. K.; Robertson, P. W. J. Chem. Soc. 1939, 1315. (b)
Yates, K.; McDonald, R. S.; Shapiro, S. A. J. Org. Chem. 1973, 38, 2460.
(14) Ruasse, M. F.; Aubard, J.; Galland, B.; Adenier, A. J. Phys. Chem.
1986, 90, 4382.
(15) (a) Bartlett, P. D.; Tarbell, D. S. J. Am. Chem. Soc. 1936, 58, 466.
(b) Dubois, J. E.; Bienvenue-Goe¨tz, E. Bull. Soc. Chim. Fr. 1968, 2089.
(c) Dubois, J. E.; Huynh, X. Q. Tetrahedron Lett. 1971, 3369.

12494 J. Am. Chem. Soc., Vol. 119, No. 51, 1997
Ruasse et al.
Table 1. Rate Constants,a k (M-1 s-1), for Bromination of X,X′-p-Substituted cis- and trans-Stilbenes in Various Solvents at 25 °C
compd
X
X′
EtOH
YBr -2.4
Nf
0.09
AcOH
-2.1
MeOH
M10b
M20b
0.70
-0.02
M30b
1.42
-0.07
TFEc
2.53
-3.93
KSIEd
e
-1.1
0.01
-0.14
-2.05
0.02
6a
6b
7a
7b
8a
9a
9b
10a
11a
11b
p-OMe
p-OMe
p-Me
p-Me
p-Me
H
H
H
p-Me
p-Me
H
H
H
H
p-CF3
p-CF3
1.9 × 103
3.16 × 10
4.2 × 103
3.3 × 104
2.9 × 104
3.0 × 102
2.8 × 102
1.4 × 102
1.0 × 106
2.0 × 103
16
2.2 × 103
1.0 × 103
1.0 × 104
9.2 × 104
1.33
2.2
3.98
5.01
0.95
3.9 × 103
1.8 × 104
8.6 × 105
1.42
1.10
1.0 × 10-1 12.5
80
4.0 × 102
1.8 × 103
1.2 × 104
3.7 × 102
H
5.5 × 10-2 13.4
p-CF3
p-CF3
p-CF3
3.0 × 10-2
2.8 × 10-3
7 × 10-2
7.9 × 10-1
7.3
4.3 × 10
2.0 × 102
1.52
7.8 × 10-4
7.0 × 10-5
1.0 × 10-2
1.1 × 10-2
6.0 × 10-2
3.0 × 10-1
1.1
2.1 × 10-1 1.55
a Reproducibility better than (5%. b M10, M20 and M30: 10-90, 20-80, and 30-70 H2O-MeOH (v/v) mixtures. c 3% (w/w) aqueous
trifluoroethanol. d Kinetic solvent isotope effects (kMeOH/kMeOD). e Reference 17. f Reference 18a.
Table 2. mBrYBr Relationships for cis-Stilbene Bromination
interpretation for all the stilbenes investigated. If nucleophilic
assistance occurs, deviations, increasingly negative with reactiv-
ity decrease, are expected, as the result of a demand on the
nucleophilic solvent increasing with the decrease in the charge
stabilization of the transition states by the substituents.10 While
the ∆AcOH trend from 6a to 9a agrees with this requirement,
the negligible values of this deviation for 10a and 11a suggest
a change in the bromination mechanism. Analogously, the
positive value of ∆TFE for 8a cannot be readily interpreted in
terms of a variable assistance by the nucleophilic alcohols.
The kinetic solvent isotope effects (KSIE ) kMeOH/kMeOD) for
7a-9a reactions (Table 1) are usual as compared to those
previously observed in bromination7a,19 which are generally close
to 1.35 or smaller when return occurs, i.e., when the product-
forming step is rate limiting.6e However, the KSIEs found for
10a and 11b are markedly larger and point to a mechanism
different from those involved in the bromination of the more
reactive olefins.
Finally, for the most reactive stilbene 6a, for which nucleo-
philic solvent assistance is not expected,9 a markedly upward
curvature of the mY plot is observed, which is also unusual in
bromination.
Therefore, depending on the solvent and the substituents, other
reaction pathways prevail or compete with the preassociation
mechanism involving nucleophilically solvated transition states
of the ionization step of stilbene bromination.
p-OMe p,p′-di-Me p-Me
H
p-CF3 p,p′-di-CF3
a
mBr
g1.4b
0.85c
3.43
-0.45
0.83
3.07
0.85
2.02
0.92
1.00
0.81
-1.10
-0.30
-1.63
log ko
AcOH
d
∆
0.0
-0.73 -1.24 -0.20
d
∆
0.76 -0.10 -0.75
TFE
i
e
1.1
3.6
1.1
3.0
1.1
1.3
mBr
log kio
a Calculated from k in M-1 s-1 at 25 °C, in EtOH, MeOH, and H2O-
MeOH mixtures only, using YBr as solvent parameters; standard errors
e 0.03; correlation coefficients g 0.998. b Estimation from MeOH and
M10 only (see text). c From MeOH, M10, and M20; M30 deviates
markedly (see text). d log kexp - log kcalc with log kcalc calculated from
the corresponding mBrYBr relationships; ∆ values smaller than 0.3 l.u.
are negligible. e Coefficients of the mBrYBr relationships for the unas-
sisted pathways estimated from k in AcOH and TFE only (see text).
Rate constants for free bromine addition to stilbenes 6a-
11a in methanol were also obtained by extrapolation to [Br-]
) 0 of kinetic measurements carried out in the presence of
several concentrations of added bromide ions, using eq 3 (Table
S1). The agreement between the two sets of results is within
the experimental errors and confirms that bromide ions do not
play a significant role when they are released by formation of
solvent-incorporated products (vide supra).
2. Kinetic Solvent Effects and mBr Values. The kinetic
solvent effects for 7a-11a are satisfactorily described by
Winstein-Grunwald equations,10a,16 log k/ko ) mBrYBr, using
YBr parameters,17 when data in acetic acid and trifluoroethanol
are omitted. The coefficients of these relationships are given
in Table 2. An extended Winstein equation,10a log k/ko ) mY
+ lN, in which the effect of the solvent nucleophilicity is also
taken into account, is not applied for statistical reasons. Ethanol,
methanol, and its aqueous mixtures having very similar nucleo-
philicities,18 the lN term is constant and l can, therefore, be
evaluated from the deviations of two barely nucleophilic solvents
only. Therefore, the positions of the rate data in acetic acid
and trifluoroethanol, ∆AcOH and ∆TFE, with respect to the log k
- YBr plots, are also shown in Table 2.
3. Product Analysis. Authentic samples of erythro (or
meso) and threo (or d,l) dibromides (DB) were synthesized by
-
reaction of trans- or cis-olefins 6-11 with Bu4N+Br3 in 1,2-
dichloroethane in the presence of an excess of Bu4N+Br-.
X
X
H
C
Y
Y
H
C
C
C
Br
Br
H
X
H
X
Y = Br: erythro-DB
Y = OMe, OCOMe or OCH2CF3: erythro-MA
threo-DB
threo-MA
All the m values, close to 0.8, are consistent with nucleo-
philically assisted rate-limiting ionizations, as previously found
in solvolysis10 and in brominations of noncongested olefins.7,9
Nevertheless, the values of ∆AcOH and ∆TFE do not support this
Because this reaction is known to give only the products from
anti addition,20 the stereochemistry of these dibromides was
unequivocally established. The erythro- and threo-acetoxy
bromides21 were obtained by acetylation of the corresponding
(16) (a) Grunwald, E.; Winstein, S. J. Am. Chem. Soc. 1948, 70, 846.
(b) Winstein, S.; Fainberg, A. H.; Grunwald, E. J. Am. Chem. Soc. 1957,
79, 4146. (c) Grunwald, E. Chemtech 1984, 14, 698.
(17) (a) Bentley, T. W.; Carter, G. E. J. Am. Chem. Soc. 1982, 104,
574. (b) Bentley, T. W.; Llewellyn, G. Prog. Phys. Chem. 1990, 17, 121.
(18) (a) Schadt, F. L.; Bentley, T. W.; Schleyer, P. v. R. J. Am. Chem.
Soc. 1976, 98, 7667. (b) Bentley; T. W. In Nucleophilicity; Harris, J. M.,
McManus, S. P., Eds; American Chemical Society: Washington, DC, 1987;
pp 255-268.
(19) Garnier, F.; Donnay, R. H.; Dubois, J. E. J. Chem. Soc., Chem.
Commun. 1971, 829.
(20) Fieser, L. F. J. Chem. Ed. 1954, 31, 291.
(21) The terms “erythro” and “threo” are not strictly applicable to these
diastereoisomers. They are used here as a convenient way of describing
the stereochemistry of dibromide and mixed adduct formation.

Bromination of Stilbenes in Protic SolVents
J. Am. Chem. Soc., Vol. 119, No. 51, 1997 12495
bromohydrins. These bromohydrins were prepared, with the
exception of the p,p′-bis(trifluoromethyl) derivatives, by reacting
the trans- and cis-olefins, respectively, with N-bromosuccin-
imide (NBS) in aqueous dimethyl sulfoxide, a reaction known
to give products from anti addition stereospecifically.22 The
stereochemistry of these bromohydrins was further confirmed
by submitting samples to dehydrobromination, which gave the
corresponding epoxides with the same stereochemistry as the
starting olefins. The regiochemistry of the reactions of 6a, 6b,
8a, and 8b was assumed on the basis of the obvious formation
of an open bromocarbenium ion at the benzylic carbon bearing
the p-substituted phenyl ring, while that of 10a and 10b was
achieved by assuming water attack at the benzylic carbon
farthest from the electron-withdrawing CF3-substituted phenyl
group of the intermediate bromonium ion. The latter assignment
was also consistent with the course of HBr ring opening of the
epoxides derived from 10a and 10b, giving the bromohydrins
of opposite regiochemistry.
In the case of 11a and 11b the NBS reactions yielded
essentially the meso and d,l-dibromo adducts, instead of the
expected bromohydrins. It is likely that the strong electron-
withdrawing effect of the two para substituents deactivates the
olefins toward the NBS reagent so much as to make competitive
its decomposition to Br2, which adds to the double bond. The
threo- and erythro-bromohydrins were, therefore, prepared by
HBr ring opening of the corresponding epoxides, obtained by
m-chloroperoxybenzoic acid oxidation of olefins 11a and 11b.
The anti opening of the oxirane ring was proved by recyclization
of the bromohydrins to the starting epoxides.
further direct evidence came, in the case of unsubstituted
stilbenes, from the preparation of both erythro and threo
diastereoisomers by O-methylation of the corresponding bro-
mohydrins of known configurations22 with methyl triflate. The
same method was used to identify the threo-p-methoxy-
substituted adduct, which was formed in amounts too small to
be isolated from the brominations of 6a and 6b in methanol.
The brominations of all olefins 6-11 were carried out in
acetic acid, methanol, and trifluoroethanol at 10-2 M olefin and
5 × 10-3 M Br2, at 25 °C. The olefin to Br2 ratios were always
kept at 2:1, so that during the reactions of olefins 6a-11a the
cis isomers were always in large excess over the eventually
formed trans-olefins 6b-11b, which could, therefore, ac-
cumulate in the reaction medium. The assumption of an
isomerization of the cis-olefins to their trans forms, because of
the presence of amounts of hydrogen bromide equivalent to
those formed during the course of the reaction as a consequence
of solvent incorporation, was preliminarily investigated. No
cis-trans isomerization was ever observed by HPLC analysis.
The stability of erythro- and threo-dibromides and mixed
adducts toward solvolysis and Br2- or HBr-promoted isomer-
ization was checked by exposing the products to solutions of
Br2 at the concentration employed, and to HBr at the maximum
concentrations released in the corresponding reactions. All
products were quantitatively recovered after times comparable
to those required for the olefin brominations. This assured that
the product distributions were actually obtained under kinetic
control. The products were stable to solvolysis, with the
exception of the p-methoxy-substituted dibromo adducts, which
were slowly transformed to the corresponding methoxy bromides
on standing in the reaction medium. This problem was obviated
by extracting the products from the mixtures immediately after
the end of the brominations, which were much faster than
solvolysis. This procedure was followed for all the olefins
examined. Since significant amounts of trans-olefins 9b (in
trifluoroethanol), 10b (in trifluoroethanol), and 11b (in acetic
acid, methanol, and trifluorethanol) were found in the reactions
of the corresponding cis isomers, the brominations of the latter
olefins were also carried out in the presence of trans-1,2-
dichloroethylene. No isomerization to cis-1,2-dichloroethylene,
which would have been produced in the presence of free
radicals,26 was ever observed. This excluded that the transfor-
mations 9a f 9b, 10a f 10b, and 11a f 11b were due to a
competing free-radical bromination process,41 and left the ionic
process as the only one responsible for these isomerizations.
The erythro- and threo-methoxy bromides were isolated as
the solvent-incorporated mixed products (MA) from the anti
addition of bromine to the trans- and cis-olefins, respectively,
performed at [Br2] ≈ 10-3 M, under conditions in which the
methoxy bromide yields were always >90%. Only in the case
of 6a and 6b were the reactions stereoconvergent to the
formation of the erythro adduct, whose regiochemistry was
easily predicted from the carbocationic nature of the p-methoxy-
substituted carbon of the intermediate. This structure was also
consistent with the resistance of the product to solvolysis, which
can take place easily when a bromine is bonded preliminary to
a methoxyphenyl-substituted carbon, as in 1-bromo-2-methoxy-
1,2-bis(p-methoxyphenyl)ethane.23 The stereochemistry was
attributed on the basis of an equilibration of the open bromocar-
benium ion intermediate formed from both 6a and 6b to the
more stable conformation with anti oriented phenyl rings,
followed by backside attack by the solvent. The regiochemistry
of the reactions of 8a, 8b, 10a, and 10b was assumed on the
basis of the regiochemical behavior of the reactions with NBS-
H2O, discussed above.
The product yields obtained from all the olefins in acetic acid
and methanol were determined by HPLC after addition of an
appropriate standard. For the least reactive olefins the product
ratios were determined at several conversions and found to be
independent of the progress of the bromination. This provided
further evidence for the kinetic control of these reactions. The
total yields of products and unreacted olefin (in the case of
incomplete reactions) were always >95%, showing that no
significant amounts of products other than those identified were
formed. For the reactions in trifluoroethanol the erythro- and
threo-trifluoroethoxy bromides were not isolated, but identified
1
The H NMR spectra of all threo-bromohydrins, acetoxy
bromides, and methoxy bromides so obtained showed 3J values
between the R protons slightly higher than those of their erythro
isomers, opposite to the generally reported behavior of erythro-
threo pairs.24 The same trend was found also for the erythro-
and threo-acetoxy bromides and chlorides derived from â-me-
thylstyrenes25 and lent support to the assigned configurations
of all mixed adducts. In particular, for methoxy bromides
1
by the H NMR spectra of the reaction mixtures, from which
1
the distribution of all products was also obtained. These H
(22) Dalton, D. L.; Dutta, P.; Jones, D. C. J. Am. Chem. Soc. 1968, 90,
5498.
(23) (a) Attempts to brominate p,p′-dimethoxystilbenes in methanol
yielded 1,2-dimethoxy-1,2-di(p-methoxyphenyl)ethane arising from the
solvolysis of the first formed methoxybromo adduct. (b) Ruasse, M. F.;
Argile, A. J. Org. Chem. 1983, 48, 202.
(24) Jackman, L. M.; Sternhell, S. Nuclear Magnetic Resonance Spec-
troscopy in Organic Chemistry, 2nd ed.; Pergamon Press: Oxford, 1969; p
291.
NMR spectra showed, besides the signals due to the erythro-
and/or threo-dibromides, an AB quartet between 4.5-5.0 ppm,
attributable to the benzylic protons of the trifluoroethoxy
bromides, and a complex signal centered on δ 4.0 consisting
of four quartets, interpretable as an AB quartet due to the
(26) Walling, C.; Rieger, A. L.; Tanner, D. D. J. Am. Chem. Soc. 1963,
85, 3129.
(25) Rolston, J. H.; Yates, K. J. Am. Chem. Soc. 1969, 91, 1469.

12496 J. Am. Chem. Soc., Vol. 119, No. 51, 1997
Ruasse et al.
Table 3. Chemoselectivity (% MA)a and Stereoselectivity (% anti addition)b of the Reaction of cis- and trans-Stilbenes (10-2 M) with
Bromine (5 × 10-3 M) at 25 °C
CHC6H4X + Br2
XH4C6HC CHC6H4X + XH4C6HC CHC6H4X
XH4C6CH
OS Br
MA
Br Br
DB
MeOH
AcOH
MA
TFE
stereob
stereob
stereob
Xc
%MAa
MA
DB
%MAa
DB
%MAa
MA
DB
p-OMe (6)
a
b
97
96
5
89
-
-
1
7
-
-
22
85
stereoconvergentd
stereoconvergentd
p,p′-di-Me (7)
a
b
90
94
85
95
-
-
17
17
55
100
35
44
stereospecifice
p-Me (8)
a
b
98
98
85
95
-
-
7
12
70
85
30
60
50
50
18
82
20
80
stereospecifice
stereoconvergentd
H (9)
a
b
97
100
97
100
-
-
25
28
85
100
45
60
55
65
25
99
54
80
stereospecifice
p-CF3 (10)
a
b
98
98
98
100
-
-
37
40
95
95
40
60
60
70
33
100
50
85
stereospecifice
p,p′-di-CF3 (11)
a
b
81
75
94
100
80
100
20
4
80
95
95
100
50
45
93
98
93
91
stereospecifice
Stereospecifice
stereospecifice
a %MA ) [(MA)/(MA+DB)]% ) chemoselectivity to ( 2%. b Stereochemistry of MA and DB ) % anti addition to ( 2%; anti addition to cis-
and trans-stilbenes gives threo and erythro adducts, respectively (see Table S2). c X: mono or disubstitution by X: a ) cis-stilbene; b ) trans-
stilbene. d Stereoconvergent: the two stereoisomeric olefins give the same products. e Stereospecific: the two stereoisomers give different products.
Table 4. Isomerizationa of cis-X-Substituted Stilbenes during
the two adducts are obtained in similar yields in trifluoroethanol,
although TFE is markedly less nucleophilic than AcOH.
Their Brominationb
X
MeOH
AcOH
TFE
The bromination of trans-stilbenes 6b-11b leads almost
exclusively to the erythro adducts (Table 3) via a 100% anti
addition, regardless of the substituents and the solvent. In
contrast, the stereochemistry of the reaction of the cis-stilbenes
exhibits a considerable dependence on the substituents and on
the solvent, the adducts, either threo or erythro resulting from
anti or syn addition, respectively. Therefore the reaction of a
couple of cis- and trans-stilbenes is either stereoconvergent (6a-
6b) or stereospecific (11a-11b) or less than 100% stereose-
lective (9a-9b, for example, in AcOH or TFE).
7a
8a
9a
10a
11a
p,p′-di-Me
p-Me
0.00
0.00
0.00
0.02
0.41
0.03
0.05
0.05
0.05
0.51
(0.00)
0.01
0.14
0.65
0.45
H
p-CF3
p,p′-di-CF3
a Isomerization of cis- to trans-olefin after bromination in the
presence of bromine deficiency ([Ol] ) 10-2 M and [Br2] ) 5 × 10-3
M). The reported figures are the ratios of the trans-olefin formed to
the total adducts; reproducibility better than 1%. b Return is inferred
from isomerization. When isomerization is 0.00, return is not significant.
A priori, these results could be rationalized within the usual
framework of a spectrum of intermediates going from fully
bridged bromonium ions, â-I+, affording stereospecific and
stereoselective product formation, to fully open carbocation,
R-I+, leading to stereoconvergency because of ready strain
release by C+-C bond rotation, in the conformation first
obtained from the cis-olefin (Scheme 4). In agreement with
this, the reaction of p-methoxy stilbenes 6a-6b, necessarily8b
via an open bromocarbocation, is stereoconvergent, whereas that
of 11a-11b via bromonium ions is stereospecific. In AcOH
and TFE, there is a progressive change from stereoconvergency
to stereospecificity for the reaction, implying an increase in
bromine bridging when the substituents are more and more
electron withdrawing. However, the role of the solvent is not
readily understood in terms of variable bromine bridging. In
MeOH, the bromination of 7-11 is stereospecific, regardless
of the substituents. In contrast, in AcOH and TFE, the
stereochemistry is solvent dependent but this dependence is not
the same in the two solvents and, moreover, the two adducts
diastereotopic methylene protons of the trifluoroethoxy bro-
mides, each line of which was further split by three H-F
couplings. The relative configurations of the two trifluoroethoxy
bromides were attributed on the basis of the larger J value
always found for the threo diastereoisomers of the analogous
methoxy bromides and acetoxy bromides.
The distributions of dibromides (DB) and mixed adducts
(MA) found for all cis- and trans-olefins 6-11 are reported in
Tables 3 and S2, while the ratios of the trans-olefins formed
during the bromination of the cis isomers to the total addition
products are reported in Table 4. The physical constants and
the 1H NMR data of the olefins and of their bromination
products are given in Tables S3-S7.
4. Stereochemistry and Chemoselectivity. In methanol,
the main products are the solvent-incorporated adducts, MA,
and in acetic acid, the dibromides, DB, in agreement with the
relative nucleophilicities of the two solvents.18 Nevertheless,
3

Bromination of Stilbenes in Protic SolVents
J. Am. Chem. Soc., Vol. 119, No. 51, 1997 12497
Scheme 4
are obtained with different stereoselectivities. There are several
possible interpretations of these results: solvent-dependent
bromine bridging, changes in the reaction pathways, in the nature
of the product-forming intermediates and of their associated
partners (2, 4, or 5), and in their lifetimes, depending on the
solvent and the substituents. These suggestions are discussed
now.
5. Kinetic Substituent Effects and Substituent- But Not
Solvent-Dependent Bromine Bridging. The marked solvent
dependence of the chemo- and stereoselectivities (Table 3)
cannot be related to a change in the magnitude of bromine
bridging with the solvent since this bridging is not solvent
dependent.8b This is supported by the following kinetic and
product data.
(i) If the magnitude of bridging of the bromination intermedi-
ates is the result of a competition among bromine, the substit-
uents, and the solvent in the stabilization of the cationic charge,
a stereoselectivity smaller in the nucleophilic methanol than in
acetic acid or trifluoroethanol is expected. The greater the
nucleophilicity of the solvent, the larger the stabilization of the
carbocationic intermediates by solvation, the smaller the elec-
tronic demand on their bromine atoms and the smaller the
bridging. In contrast, the reaction is fully stereoselective in
methanol but not in the other two solvents.
(ii) Changes in bromine bridging with the substituents lead
to markedly curved Fσ plots for arylolefin bromination8b and,
in particular, for that of trans-stilbenes27 in MeOH. The F+
value decreases smoothly as the substituents are less and less
electron donating, from -4.3 in the usual range for benzylic
carbocations to about -1 when the intermediates are bridged,
i.e. when their positive charge is on the bridging bromine.
Similar curvatures are observed in the other solvents of Table
1 since there are rigorously linear log-log relationships (i)
between the rate data of substituted cis- and trans-stilbenes in
MeOH (log ktrs ) 0.90 log kcis -0.10, correlation coefficient )
0.999) and (ii) between the data for cis-stilbenes in MeOH and
in the other solvents (Figure 1). Therefore, the substituent
dependence of the bridging is very similar in all these solvents.
These stereochemical and kinetic data are the first compelling
evidence for a significant substituent dependence but insignifi-
cant solvent dependence of bromine bridging. This result has
two kinds of consequences. First, if the intermediate is an open
â-bromocarbocation, R-I+, in a given solvent as shown by
stereoconvergency, the intermediate is also R-I+ in the other
solvents. In particular, since the reactions of 6-8 are stereo-
convergent in AcOH and/or TFE, their intermediates are
necessarily the corresponding carbocations R-I+ in these solvents
Figure 1. log-log relationships comparing substituent effects on cis-
stilbene bromination in MeOH (x axis) and in other solvents (y axis).
In nucleophilic solvents, the slopes, FS/FMeOH, are close to unity (0.92,
1.07, 0.98, and 1.03 in EtOH, M10, M20, and M30, respectively;
standard errors e 0.05; correlation coefficients g 0.994). In TFE, FS/
FMeOH is 1.56 ( 0.04 and in AcOH, 1.35 ( 0.03 excluding 10a and
11a for which there is a change in the rate-limiting step. The linearity
of these relationships shows that bromine bridging is not solvent
dependent (see text).
and also in MeOH. Therefore, the stereospecificity of 7 and 8
in MeOH cannot be attributed to bromine bridging. Secondly,
since R-I+ from 7 and 8 react stereospecifically with MeOH
only, the usual conformer rotation (Scheme 4) does not occur
before they react with MeOH. In other words, the lifetimes of
these intermediates are very short in this nucleophilic solvent
as compared to those in the other solvents. A reasonable
interpretation, in agreement with the m values and the deviations
of ∆AcOH and ∆TFE, is that in MeOH bromination is nucleo-
philically assisted, as expressed in a preassociation mechanism,
whereas in AcOH and TFE other pathways are preferred.
6. Preassociation Pathways in Nucleophilic Alcohols.
There are several pieces of kinetic evidence for nucleophilic
assistance to the ionization step by nucleophilic solvents in the
bromination of 7-11 (vide supra). Now the questions are (i)
Is this assistance satisfactorily described by a preassociation
mechanism (eq 4)? (ii) If yes, what are the rate-limiting steps?
1 h 3 h 4 h products
(4)
And (iii) are the products consistent with the collapse of
complexes 4?
There are three main requirements for a preassociation
pathway to occur.11 The nucleophile, which in bromination is
the solvent, must preassociate with the substrate. Methanol is
not a strong nucleophile but bromine-olefin charge-transfer
complexes 1 are highly polarizable, so the two species can
interact strongly to form dipole-dipole complexes 3. A second
condition is that return from 4 to the preassociation complex is
more rapid11,29 than diffusion away of methanol leading to ion
pair 2, k-a < k′-1 (Scheme 2). In bromination, MeOH reacts
(27) (a) Ruasse, M. F.; Dubois, J. E. J. Org. Chem. 1972, 37, 1770. (b)
Ruasse, M. F.; Dubois, J. E. J. Org. Chem. 1973, 38, 473; 1974, 39, 2441.
(c) These curvatures cannot be interpreted in the usual terms of the
Yukawa-Tsuno equation;28 see footnote 14 in ref 27a.
(28) Tsuno, Y.; Fujio, H. Chem. Soc. ReV. 1996, 25, 129.

12498 J. Am. Chem. Soc., Vol. 119, No. 51, 1997
Ruasse et al.
Scheme 5
between the counter-bromide ion and the I+-bromine atom
disfavor capture of I+ by Br-, which should be syn since the
usual anti addition would require methanol expulsion after Br-
rotation around I+ (trans location). On the contrary, methanol
can interact strongly with the empty orbital of the cationic centre,
in particular when I+ is a bromocarbenium ion. In agreement
with this proposal, more dibromide (up to 25%) is obtained when
I+ is a bromonium ion (11) which interacts weakly with MeOH
because of the negligible charge of its carbon atoms.
7. Free-Ion Pathway. In contrast with the other stilbenes,
bromination of 6 is stereoconvergent in methanol and in acetic
acid. This is readily interpreted in terms of the absence of any
nucleophilic assistance and of fully open intermediates. In other
terms, a free ion pathway (eq 5) in which the ionization is rate
limiting (Scheme 2; kd > k-1, kBr > k-1) since I+ is unbridged,
is involved. In the moderately ionizing MeOH where MA is
rapidly with I+ to give the mixed adduct (vide supra) and,
therefore, does not diffuse rapidly from I+, whereas the rate
constant for return, the reaction of the counter-bromide ion with
I+, can be5d as large as 1011 s-1. Finally, the collapse of I+ to
the products must have a significant barrier for 4 to exist as an
intermediate. Preassociation mechanisms in some nucleophilic
substitutions11g,30 have been excluded because the lifetimes of
complexes analogous to 4 are too short. In contrast, the lifetimes
of ionic intermediates for alkene bromination in MeOH, which
are also nucleophilically assisted,7 are short but significant31
(kMeOH ) 109-1011 s-1). Therefore, in nucleophilic solvents,
assisted bromination can occur via a preassociation pathway.
When the intermediates are open carbocations (R-I+) as those
from 7 and 8, return by reaction of the bromide ion with the
bromine atom of I+ is not likely6c because the carbon atom of
I+, and not the bromine, is charged. In contrast, bromonium
ions from 10 and 11 can readily undergo return (Scheme 5).
Evidence for this will be given (vide infra). Therefore, there is
a change in the rate-limiting step from the 4-forming step to
the product-forming step when bromine bridging, i.e., the
electron-withdrawing character of the substituents, increases.
The anti stereoselectivity of cis-stilbene bromination in
MeOH and its independence of bromine bridging is consistent
with product formation by collapse of complexes 4, since
methanol reacts necessarily with I+ on the site opposite to the
entering bromine. Therefore, bromine bridging is not the only
prerequisite for stereospecific bromination; preassociation can
be the source of stereospecificity also. Stereospecific bromi-
nations of trans-â-methylstyrenes via open â-bromocarboca-
tions, in MeOH but not in halogenated solvents,32 can also be
attributed to preassociation.
1 f 2 f 5 f MA
(5)
V
DB
the only product, the dissociation of the ion pair, kd, is faster
than its collapse, kBr, whereas it is the the reverse in AcOH in
which DB is the main adduct. The upward curvature of the
corresponding mY plot is, therefore, surprising and results
probably from differences in the mechanisms of bromination
and of the YBr-defining reaction.17 As there is no return in the
bromination of 6, a number of results29,35 suggests that 2-bro-
moadamantane solvolysis is reversible. Therefore, the kinetic
data for the two reactions are not related to the same microscopic
events.
Free-ion pathways are also involved in the nonnucleophilic
and highly ionizing trifluoroethanol, regardless of the substit-
uents. The trapping of free ions by TFE is not very fast34 and,
therefore, ion-pair collapse36 competes extensively, as shown
by the competitive formation of the two adducts (DB:MA )
1:1). This is consistent with the relative rates of carbocation
trapping by MeOH and TFE34d,e (kMeOH/kTFE ) 106), since in
MeOH no dibromide is obtained (kMeOH . kBr, whereas kTFE
≈
kBr). This result, in agreement with usual findings in bromi-
nation,33 is in contrast with the diffusion-controlled rate of Br-
reaction with nonbrominated benzylic carbocations34 and shows
that the requirements for the collapse of ion pairs 2, either trans
(trans location of Br-) or syn (bromine-bromine repulsions),
are energetically expensive.37
Complex 4 collapses almost exclusively by reaction of I+
with methanol and not with bromide, since dibromide is a very
minor product regardless of the substituents. This is in
agreement with usual findings that alkene bromination in
methanol gives methoxybromo adducts but not dibromide, in
the absence of added bromide ions.31,33 Nevertheless, these
results contradict solvolytic data which show that bromide and
not methanol reacts at the diffusion-controlled rate with benzylic
carbocations,34 so collapse of 4 should afford mainly dibromide.
In bromination intermediates, steric and electrostatic repulsions
Kinetic data in TFE and the corresponding ∆TFE, i.e., the
relative rates of the preassociation and free-ion pathways,
suggest a possible competition between the two routes (eq 6).
4 r r 1 f 2
(6)
In particular, for the irreversible bromination of 8a, the
unassisted route in TFE is favored over the assisted route in a
nucleophilic solvent of similar ionizing power by about 1 kcal
mol-1 (∆TFE ) +0.76), whereas for 9a the two routes have
similar barriers (∆TFE, negligible). The extent of the competition
can be estimated by comparing the previously obtained mY
relationships (Table 2) for the assisted process (1 f 4) to those
for the unassisted pathway (1 f 2). The coefficients of the
mY equations related to this latter path (miBr and log koi in
(29) Cox, B. G.; Maskill, H. J. Chem. Soc., Trans. Perkin 2 1983, 1901.
(30) (a) Richard, J. P.; Amyes, T. L.; Vontor, T. J. Am. Chem. Soc. 1991,
113, 5871. (b) Toteva, M. H.; Richard, J. P. J. Am. Chem. Soc. 1996, 118,
11434.
(31) Nagorski, R. W.; Brown, R. S. J. Am. Chem. Soc. 1992, 114, 7773.
(32) Ruasse, M. F.; Argile, A.; Dubois, J. E. J. Am. Chem. Soc. 1978,
100, 7645.
(33) (a) Dubois, J. E.; Chre´tien, J. R. J. Am. Chem. Soc. 1978, 100, 3506.
(b) Ruasse, M. F.; Coudert, D.; Chre´tien, J. R. J. Org. Chem. 1993, 58,
1917. (c) Ruasse, M. F.; Argile, A. J. Org. Chem. 1983, 48, 202.
(34) (a) Richard, J. P. J. Am. Chem. Soc. 1989, 111, 1455. (b) McClelland,
R. A.; Kanagasabapathy, V. M.; Steenken, S. J. Am. Chem. Soc. 1988, 110,
6913. (c) McClelland, R. A.; Chan, C.; Cozens, F.; Modro, A.; Steenken,
S. Angew. Chem. Int. Ed. Engl. 1991, 30, 1337. (d) Das, P. K. Chem. ReV.
1993, 93, 119. (e) McClelland, R. A. In Organic ReactiVity: Physical and
Biological Aspects; Golding, B. T., Griffin, R. J., Maskill, H., Eds; The
Royal Society of Chemistry: Cambridge, 1995; p 301.
(35) Paradisi, C.; Bunnett, J. F. J. Am. Chem. Soc. 1985, 107, 8223.
(36) Dibromide formation by Br- capture of free ions 5 is excluded
because of the very low concentration of bromide ions resulting from the
dissociation of the ion pair.
(37) It should be noted also that dibromide is obtained with a stereose-
lectivity better than that of MA, showing that the lifetime of the free ion is
longer than that of the ion pair.

Bromination of Stilbenes in Protic SolVents
J. Am. Chem. Soc., Vol. 119, No. 51, 1997 12499
bridged, â-I+, but not that of 9a which must closely resemble
R-I+, since return and bridging are related.
Product data, and in particular the stereochemistry of DB
formation as compared to that of MA from the open intermedi-
ates, are consistent with product formation from complexes 4.
Whereas MA results mainly from an anti addition, DB is
obtained via a process close to syn. For example, in the reaction
of 10a, MA results from a 95% anti addition, whereas the
formation of DB is 60% syn. It is not expected that complexes
4 live long enough for complete conformer rotation to occur.
Therefore, Br--syn addition cannot arise only from strain release
in the initially formed R-I+ conformation. The most reasonable
interpretation is that products are obtained by collapse of 4 in
which Br--anti addition is hindered by the presence of AcOH.
Br--syn addition is, therefore, favored despite its drastic
requirements.
9. Isomerization, KSIEs, and Return. Reversibility of the
ionization steps of the bromination of 10a and 11a is evidenced
by the zero values of their ∆AcOH discussed above. Return is
also supported by a series of experiments (Table 4) carried out
with a bromine deficiency in which cis-trans isomerization of
the starting cis-olefin can be observed. As previously discussed,6c
large isomerization ratios are indicative of significant return.
According to these results, return is important for 11a in every
solvent and for 10a in TFE but not for the other stilbenes, in
agreement with the already mentioned relationship between
bridging and return. The isomerization involved in the bromi-
nation of the two poorly reactive stilbenes can be interpreted
qualitatively in terms of competition between nucleophilic
trapping of the bromonium ions and their cis-trans equilibra-
tion, as shown in Scheme 6. The trapping of bridged ions is
extremely slow because of the absence of charge on their carbon
atoms and, therefore, the barrier for their isomerization is
probably in the same range as that for their reaction with
nucleophiles. The fact that this phenomenon is observed for
10a in the nonnucleophilic TFE but not in the other solvents
suggests similar heights for the two barriers, while for 11a,
isomerization would be uniformly faster than nucleophilic
trapping, even by methanol. It is reasonable to assume that
the isomerization of the bromonium ions goes through open
transition states or intermediates (Scheme 6), the strain of which
is readily released by fast C+-C bond rotation.40 This is
supported by the incomplete stereoselectivity of 11a and 10a
(Table 3), which is unexpected from full bridging. Nucleophilic
trapping of the highly unstable open ions probably occurs.
Alternatively, the isomerization of the two poorly reactive
stilbenes could suggest a change from an ionic to radical
mechanism. Nevertheless, this interpretation is not plausible
because the particular kinetic and stereochemical behavior41 of
radical bromination of stilbenes (zeroth order in stilbene,
stereoconvergency) is not observed under the presently used
reaction conditions, even for 10a and 11a.
Figure 2. mBrYBr relationships for the bromination of 7a and 9a via
assisted (full line from EtOH, MeOH, and H2O-MeOH mixtures; mBr
and log ko in Table 2) and unassisted (dotted line from AcOH, TFE or
M30; miBr and log kio) rate-limiting ionizations.
Table 2) are evaluated from the rate data in the two nonnu-
cleophilic solvents. The miBr values, 1.1, agree with those
previously found for unassisted brominations.7 The comparison
of the two series of mY plots (Figure 2) shows that the rate
differences between the two types of ionization are not large in
the solvents investigated. Nevertheless, it is observed that, for
stilbene 8a for example, the unassisted pathway would be the
only one in nucleophilic solvents more ionizing than TFE. This
result explains fairly well why, in highly aqueous trifluoroet-
hanol (YBr > 3), no evidence was found for a preassociation
pathway in the solvolysis of tert-cumyl derivatives.30
8. Ion-Pair Pathways in Acetic Acid. In AcOH, a Sneen
mechanism38 (eq 7) involving nucleophilic collapse of an ion
pair is the most likely. In stilbene bromination, this pathway
1 h 2 h 4 h products
(7)
is supported by kinetics and product data. The F value in AcOH,
which is larger than that in MeOH and in the same range as
that in TFE (FAcOH ) 0.9FTFE ) 1.35FMeOH), is further evidence
for the absence of nucleophilic assistance in the two barely
nucleophilic solvents, as compared to what occurs in the other
solvents39 (Figure 1). In terms of Scheme 2, the first ionization
steps are similarly unassisted in AcOH and TFE. However,
the deviations of 10a and 11a from the log-log relationship
(AcOH-MeOH in Figure 1) and their almost zero values for
∆AcOH (Table 2) are unexpected. Since these two least reactive
stilbenes fit the mY plots for the nucleophilic alcohols, the rate-
limiting steps are the same in the two kinds of solvent. The
only step common to the preassociation (eq 4) and ion-pair (eq
7) pathways is the product formation from complexes 4. In
other words, the mechanism in AcOH is necessarily that
described by eq 7 involving solvent association with the ion
pair (2 f 4) before product formation. Moreover, this product-
forming step is rate-limiting in the two kinds of solvent for the
reactions of 10a and 11a but not for the other stilbenes. Finally,
the bromination intermediates of 11a and also of 10a are
KSIEs in methanol (Table 1) for the reversible bromination
of 10a and 11a are unexpectedly large as compared to those
usually found in this addition7,19 (1.1-1.3). Moreover, small
KSIEs have been taken previously as evidence for return in the
reaction of highly congested alkenes.7 Insofar as KSIEs are
related to the extent of the solvation of the leaving bromide ion
in the rate-limiting step,7 these results imply highly solvated
Br- in the product-forming step of the bromination of 10a and
11a, but highly desolvated Br- in that of the reaction of
congested alkenes.42 Inspection of the transition states of these
(38) (a) Sneen, R. A. Acc. Chem. Res. 1973, 6, 46. (b) Raber, D. J.;
Harris, J. M.; Schleyer, P. v. R. In Ions and Ion-Pairs in Organic Reactions;
Szwarc M., Ed.; Wiley: New York, 1974; Vol. 2, p 247.
(39) The same F ratio, FAcOH/FMeOH ) 1.35, was previously observed
for styrene bromination.32
(40) The rate of the conformer equilibration can be in the 109 s-1 range,
according to preliminary measurements of lifetimes of stilbene bromination
intermediates (Chiappe, C.; Lo Moro, G.; Ruasse, M. F, unpublished results).
(41) Bellucci, G.; Chiappe, C. J. Phys. Org. Chem., in press.

12500 J. Am. Chem. Soc., Vol. 119, No. 51, 1997
Ruasse et al.
Scheme 6
two reactions shows that this suggestion is reasonable. On the
one hand, the rate-limiting step for the poorly reactive stilbenes
involves Br- diffusion from complexes 4 and, therefore, an
increase in Br- solvation. On the other hand, for crowded
alkenes giving vinylic products,5 the counterion which acts as
a base in the proton elimination is transformed into hydrobromic
acid.
philic trapping, except when they are strongly destabilized
bromonium ions. Moreover, the severe energetic requirements
of syn vs anti addition, which involve strong bromine-bromine
repulsions and Br- trans location, respectively, are fairly well
exhibited by the absence of Br- collapse of the ion-dipole
sandwiches 4, in contrast with the diffusion-controlled Br-
reactions with analogous non-brominated carbocations.34
(iii) Neither the kinetic results (m values, ∆AcOH, ∆TFE, KSIEs)
nor the chemoselectivity dependence on the solvent (major
solvent-incorporated adducts in MeOH, major dibromide in
AcOH but not in TFE), nor the large variety of stereochemical
outcomes (stereoconvergency to stereospecificity) can be ra-
tionalized without taking into account all the pathways of
Scheme 2. Alternatively, changes in bromine bridging with the
solvent and the substituents which were assumed previously
from more limited data sets,45 do not provide a satisfactory
interpretation. Our results point to the need for extensive data
on a very wide range of solvents and substituents in order to
show the consistency of Scheme 2. The same comment
probably applies in aliphatic nucleophilic substitutions for which
Scheme 3 is not necessarily required in every investigation.
Scheme 2 is not only a useful working assumption but also
necessary since pathway crossings are observed. For instance,
it is obvious a priori that preassociation and ion-pair paths are
preferred in MeOH and AcOH, respectively, but the kinetic data
for the two least reactive stilbenes can be interpreted only by a
crossing between these two paths. Furthermore, the fact that
the free ion route is sometimes energetically favored in
nucleophilic solvents is compelling evidence for a competition
between preassociation and free ion paths, which is readily
understood within Scheme 2. Finally, a route corresponding
to a concerted bromination, which has never been supported
by any data,1d,45 can be suggested by analogy with the SN2
pathway of nucleophilic substitution. The positive deviation
from EtOH to the mY plot of the least reactive stilbene could
be attributed to a fully concerted bromine addition, although
there is no direct evidence for it.
Concluding Remarks
From this kinetic and product study of stilbene bromination
over an exceptionally large reactivity range (10 powers of 10),
several original conclusions on the mechanism of this reaction
emerge.
(i) In contrast with the widely accepted postulate, bromine
bridging is not the only stereochemistry-determining factor. The
large solvent dependence of the stereochemistry does not arise
from bridging variations with solvent, since this latter is
substituent- but not solvent-dependent, as shown by the
comparison of kinetic substituent effects in the several solvents
studied. The stereochemical outcome of bromination is con-
trolled not only by the bridged or unbridged structure of the
cationic intermediate but also by its association with its
nucleophilic partners and its lifetime. In this context, an
important result is the substituent-independent stereospecificity
of the reaction of cis-stilbenes and of trans-â-methylstyrenes32
in MeOH, but not in nonnucleophilic solvents, which is the
consequence of their preassociation with the nucleophile, well
before the product-forming step itself.
(ii) Return, which is significant in protic and halogenated
solvents6c when the intermediates are bridged but not when they
are open, is also a result of interest. Until now, much work
has been devoted to return enforced by steric inhibition of the
nucleophile approach to crowded bromination intermediates.2,43
A second origin, namely the magnitude of the charge borne by
the bromine atom of these intermediates, is well established by
the present results. Negligible values of ∆AcOH, large KSIEs,
and large isomerization ratios are converging pieces of evidence
for this return. These conclusions, related to the still-open
question of the equilibrium between bridged and unbridged
structures of the bromination intermediates,44 support a barrier
for this equilibrium markedly higher than that for their nucleo-
(iv) Lifetimes of the ionic intermediates of nucleophilic
substitutions determine the pathway followed under given
reaction conditions.11 The longer these lifetimes, the more
favored the pathways of the top of Scheme 3. Although the
lifetimes of the bromination intermediates31 are not measured
(42) A reviewer suggested an alternative interpretation which assumes
reasonably the occurrence of a primary isotope effect on the rate-determing,
product-forming step involving hydron departure in the I+-MeOH reaction.
This primary effect would superimpose to the KSIE on the ionization step
resulting in the observed increase in the experimental value.
(43) Return in the absence of crowding was also supported by results
on solvolysis of â-bromotriflates in the presence of bromide ions. Zheng,
C. Y.; Slebocka-Tilk, H.; Nagorski, R. W.; Alvarado, L.; Brown, R. S. J.
Org. Chem. 1993, 58, 2122.
(44) (a) Poirier, R. A.; Mezey, P. G.; Yates, K.; Cszimadia, I. G. J. Mol.
Struct. 1983, 94, 137. (b) Galland, B.; Evleth, E.; Ruasse, M. F. J. Chem.
Soc., Chem. Commun. 1990, 898. (c) Hamilton, T. P.; Schaefer, H. F. J.
Am. Chem. Soc. 1990, 112, 8260. (d) Klobukowski, M.; Brown, R. S. J.
Org. Chem. 1994, 59, 7156. (e) Cossi, M.; Persico, M.; Tomasi, J. J. Am.
Chem. Soc. 1994, 116, 5373.
(45) (a) Buckles, R. E.; Bader, J. N.; Thurmaier, R. J. J. Org. Chem.
1962, 27, 4523. (b) Heublein, G. J. Prakt. Chem. 1966, 303, 84.

Bromination of Stilbenes in Protic SolVents
J. Am. Chem. Soc., Vol. 119, No. 51, 1997 12501
500 mL of a 3 × 10-3 M solution of the trans- or cis-olefin,
respectively, in the same solvent. After 4 h at room temperature for
6-10, or 24 h for 11, a saturated solution of NaHCO3 was added and
the products were extracted with dichloromethane. Evaporation of the
dried (MgSO4) extracts and crystallization of the crude solid residues
from hexane, or Kugelrohr short-path distillation in the case of oils,
gave pure methoxy bromides.
The erythro- and threo-1-bromo-2-methoxy-1,2-diphenylethanes
were also obtained by methylation of erythro- and threo-1,2-diphenyl-
2-bromoethanols, respectively, with methyl triflate as follows: The
bromohydrin (1 mmol), methyl triflate (10 mmol), and 2,6-di-tert-butyl-
4-methylpyridine (8 mmol) were added to anhydrous dichloromethane
(10 mL) under argon, the mixture was refluxed for 48 h, and then
filtered. The solution was washed with 5% HCl and saturated aqueous
NaHCO3, dried (MgSO4), and evaporated to give a solid residue from
which the erythro- or threo-methoxy bromide was obtained pure by
preparative TLC using 9:1 hexane-ethyl acetate as the eluent.
The same procedure was used with threo-1-(p-methoxyphenyl)-2-
bromo-2-phenylethanol to obtain the corresponding threo-methoxy
bromide.
in this work, the stereochemical results can be interpreted in
these terms. In particular, the trapping rates of the open
intermediates can be estimated by comparison with the rate of
their conformational rotation40 (Scheme 4). The intermediate
from 6a is long-lived in MeOH and also in AcOH since full
rotation is achieved before its trapping by these solvents;
accordingly, the enforced path is the free ion route. For 8a,
the lifetime of its intermediate is very short in MeOH (no
rotation) but very long in TFE, in agreement with preassociation
and free ion paths, respectively, and also in agreement with the
relative rates of carbocation trapping by these solvents.34 Our
results suggest, therefore, that the conformational barriers can
be used as clocks for measuring carbocation lifetimes, as an
alternative to the more familiar azide clock.11b-f,31,34a More
work is in progress to obtain data on the lifetimes of these
bromination intermediates and on their conformational equilib-
ria.
Experimental Section
Bromination Procedures and Product Analysis. Solutions (5.5
× 10-2 M) of bromine in acetic acid, methanol, or trifluoroethanol
(0.5 mL) were rapidly mixed with 5 mL of 1.1 × 10-2 M solutions of
cis- and trans-stilbenes in the same solvent, and the reaction mixtures
were stored in the dark at 25 °C. At the end of the reactions, or after
stopping the reactions by addition of cyclohexene which rapidly
consumed all the unreacted Br2, the mixtures were diluted with water,
Solvents (ethanol, methanol, acetic acid) were purified before use
as previously described;7 trifluoroethanol from Aldrich was used without
further purification. Bromine (1 mL sealed ampules, Carlo Erba
>99.5%) was used as supplied.
Melting points were determined on a Kofler apparatus and are
uncorrected. 1H NMR spectra were registered in CDCl3 with a Bruker
AC 200 instrument and TMS as the internal reference. HPLC analyses
were carried out with a Waters 600E apparatus equipped with a diode
array detector.
Olefins 6-11. Commercial trans-stilbene (9b) (Schuchard, >99%)
was crystallized from ethanol, mp 124-125 °C. Commercial cis-
stilbene (9a) (Aldrich, >97%) was fractionally distilled with a fraction
with bp 93 °C (5 mmHg) collected resulting in >99% purity by HPLC.
Olefins 6b-11b were obtained in mixtures with the cis isomers 6a-
11a by Wittig reactions of the corresponding para-substituted benzal-
dehydes and para-substituted benzyltriphenylphosphonium chloride.47
The separation of the resulting cis-trans mixtures was performed by
column chromatography over alumina (aluminum oxide S, 100-290
mesh ASTM), with hexane as eluent. The cis isomers were always
eluted first. All olefins were finally checked by HPLC and were found
to be >99% pure.
1
repeatedly extracted with dichloromethane and analyzed by H NMR
and HPLC under the following conditions: Spherisorb S5CN (ps phase
Sep), 25 cm, with hexane-THF (95:5 v/v) as the eluent, at a flow rate
of 1.5 mL/min, for the reactions of 6a and 6b in acetic acid; Spherisorb
S5CN (ps phase Sep), 25 cm, with hexane-THF (99:1 v/v) as the
eluent, at a flow rate of 1.5 mL/min, for the reactions of 7a, 8a, 10a,
11a, 7b, 8b, 10b, and 11b in acetic acid; Spherisorb ODS2 (ps phase
Sep), 25 cm, with methanol-water (75:25 v/v) as the eluent, at a flow
rate of 1.0 mL/min, for the reactions of 9a and 9b in acetic acid;
Spherisorb S5CN (ps phase Sep), 25 cm, with hexane-THF (95:5 v/v)
as the eluent, flow rate of 1.5 mL/min, for the reactions of 6a and 6b
in methanol; Spherisorb S5CN (ps phase Sep), 25 cm, with hexane-
acetonitrile (99:1 v/v) as the eluent, flow rate of 1.5 mL/min, for the
reactions of 7a, 8a, 10a, 11a, 7b, 8b, 10b and 11b in methanol; and
Spherisorb ODS2 (ps phase Sep), 25 cm, with methanol-water (75:
25 v/v) as the eluent, flow rate of 1.0 mL/min, for the reactions of 9a
and 9b in methanol.
For the HPLC quantification of the products and of the unreacted
olefins, erythro-1,2-dibromo-1-phenylpropane was added as an internal
standard.
The product mixtures obtained in trifluoroethanol were analyzed by
1H NMR on the basis of the signals of the benzylic protons of the
dibromides and trifluoroethoxy bromides.
erythro-1-Bromo-2-(p-methylphenyl)-1-phenyl-2-trifluoroethoxy-
ethane: δ 3.62 (qq, 2H, CH2CF3), 4.90 (d, J ) 6.8 Hz, 1H, CHBr or
CHO), 5.02 (d, J ) 6.8 Hz, 1H, CHO or CHBr).
erythro-1-Bromo-1,2-diphenyl-2-trifluoroethoxyethane: δ 4.00 (qq,
2H, CH2CF3), 4.95 (d, J ) 5.4 Hz, 1H, CHBr or CHO), 5.04 (d, J )
5.4 Hz, 1H, CHO or CHBr).
erythro-1-Bromo-2-phenyl-2-(trifluoroethoxy)-2-p-(trifluoromethyl)-
phenyl]ethane: δ 3.70 (qq, 2H, CH2CF3), 4.90 (d, J ) 6.5 Hz, 1H,
CHBr or CHO), 5.02 (d, J ) 6.5 Hz, 1H, CHO or CHBr).
erythro-1-Bromo-1,2-di-[p-(trifluoromethyl)phenyl]-2-(trifluoroethoxy)-
ethane: δ 3.65 (qq, 2H, CH2CF3), 5.00 (d, J ) 7.0 Hz, 1H, CHBr or
CHO), 5.03 (d, J ) 7.0 Hz, 1H, CHO or CHBr).
threo-1-Bromo-1,2-diphenyl-2-(trifluoroethoxy)ethane: δ 4.13 (qq,
2H, CH2CF3), 4.80 (d, J ) 6.6 Hz, 1H, CHBr or CHO), 5.05 (d, J )
6.6 Hz, 1H, CHO or CHBr).
threo-1-Bromo-2-phenyl-2-(trifluoroethoxy)-1-[p-(trifluoromethyl)-
phenyl]ethane: δ 3.80 (qq, 2H, CH2CF3), 4.80 (d, J ) 7.8 Hz, 1H,
CHBr or CHO), 5.03 (d, J ) 7.8 Hz, 1H, CHO or CHBr).
threo-1-Bromo-1,2-bis-[p-(trifluoromethyl)phenyl]-2-(trifluoroethoxy)-
ethane: δ 3.80 (qq, 2H, CH2CF3), 4.85 (d, J ) 7.2 Hz, 1H, CHBr or
CHO), 5.06 (d, J ) 7.2 Hz, 1H, CHO or CHBr).
Dibromides. All cis- or trans-stilbenes were brominated with
Bu4N+Br3- in 1,2-dichloroethane using the reported procedure.6c The
crude products were crystallized from chloroform to give pure erythro
(or meso) or threo (or d,l) dibromides.
Bromohydrins. erythro- and threo-bromohydrins were prepared
from trans- and cis-olefins, respectively, with N-bromosuccinimide in
DMSO, using the procedure reported by Dalton.22 The crude products
were purified by TLC (PSC Fertigplatten Kiesel-gel 60 F254, Merck,
9:1 hexane-ethyl acetate), followed by crystallization from hexane.
erythro- and threo-1,2-bis-[p-(trifluoromethyl)phenyl]-2-bromoet-
hanol were prepared by HBr ring opening of trans- and cis-p-
bis(trifluoromethyl)stilbene oxides respectively, obtained by treatment
of the corresponding olefins with m-chloroperoxybenzoic acid in
dichloromethane for 24 h. The opening reactions were carried out in
HBr-saturated chloroform solution. After 4 h at room temperature the
solutions were washed with water, dried (MgSO4), and evaporated. The
crude products were purified by TLC, as described above.
Acetoxy Bromides. erythro- and threo-acetoxy bromides were
obtained by treating the corresponding bromohydrins with a 10-fold
excess of acetic anhydride in pyridine. After 10 h at room temperature
toluene was added and the mixtures were evaporated at reduced
pressure. The crude residues were crystallized from hexane to give
the pure products.
Methoxy Bromides. erythro- and threo-methoxy bromides were
obtained by addition of a methanolic solution of Br2 (0.4 M, 5 mL) to
(46) De Young, S.; Ehrlich, S. J.; Berliner, E. J. Am. Chem. Soc. 1977,
99, 290. (b) Ehrlich, S. J.; Berliner, E. J. Am. Chem. Soc. 1978, 100, 1525.
(c) De Young, S.; Berliner, E.; J. Org. Chem. 1979, 44, 1088.
(47) Novotny, J.; Collins, C. H.; Starks, F. W. J. Pharm. Sci. 1973, 62,
610.

12502 J. Am. Chem. Soc., Vol. 119, No. 51, 1997
Ruasse et al.
All reactions were carried out at least in quadruplicate. Reactions
stopped at several different conversions gave similar product ratios.
Table S2 reports the average product distributions of dibromides (DB)
and mixed adducts (MA). They were reproducible to (1% for values
<10, and to (2% for higher values.
For each cis-olefin, reactions were also carried out in the presence
of an equimolar amount of trans-1,2-dichloroethylene. No isomeriza-
tion to cis-1,2-dichloroethylene, which would have indicated the
occurrence of free radical processes,26 was ever observed. Furthermore,
HPLC analysis showed that no change in products was produced in
any case by prolonged contact with Br2 or HBr.
and large olefin excess) were employed. The absorbance/time data
obtained at several different wavelengths were fitted to the appropriate
rate equations in order to obtain the pertinent second-order rate
constants. All reactions were carried out at least in triplicate. The k
values are reported in Table 1. The rate constants for the bromination
in methanol were obtained both directly from measurements carried
out in the absence of external Br- and, in several cases, by plots15 of
eq 3 (Table S1). Very satisfactory overall second-order kinetics were
always found and very similar or identical values of k were obtained
by the two methods. The kinetic constants for bromination in all the
other solvents were, therefore, measured in the absence of Br-.
Kinetic Measurements. Rate constants below 1 M-1 s-1 were
obtained by monitoring bromine concentration by conventional spec-
trophotometry under second-order conditions, reagent concentrations
being less than 5 × 10-4 M. For the more reactive stilbenes, a
multichannel stopped-flow apparatus12 was used under pseudo-first-
order conditions with a large olefin excess compared to bromine, the
concentration of which was about 5 × 10-4 M. Bromine solutions (5
× 10-4 to 5 × 10-3 M), prepared shortly before use, were protected
from daylight and adjusted to twice the initial concentration desired in
the kinetic runs. Aliquots of these solutions, thermostated at 25 (
0.05 °C, were mixed with equal volumes of thermostated solutions of
olefins of suitable concentrations. A Cary 2200 spectrophotometer (for
k < 1 M-1 s-1), equipped with a 1 cm cell, or the stopped-flow
apparatus12 (for k > 10 M-1 s-1) was used to monitor the reactions.
Second-order conditions (first-order in both reagents, [Br2] ) [olefin]
Acknowledgment. This work was supported in part by
grants of Consiglio Nazionale delle Ricerche (CNR, Roma) and
Ministerio dell’Universita´ e della Ricerca Scientifica e Tech-
nologica (MURST, Roma). Dr. Iva Blagoeva and Professor John
P. Richard are thanked for their helpful criticism and comments.
Supporting Information Available: Kinetic bromide ion
effects, product distributions, physical constants, and 1H NMR
data of olefins 6-11, of the corresponding dibromides, bromo-
hydrins, acetoxy bromides, and methoxy bromides (13 pages).
See any current masthead page for ordering and Internet access
instructions.
< 5 × 10-4 M) or pseudo-first-order conditions ([Br2] ≈ 5 × 10-4
M
JA9700461