cis-bromination of alkynes without cationic intermediates

Article abstract of DOI:10.1002/anie.200461632

Surprising insight into a classical mechanism is provided by theoretical and experimental investigations on the bromination of alkynes. in nonpolar solvents the bromination of acetylene via a covalent tribromide adduct is strongly favored over the textbook mechanism via a bridged bromirenium ion. The structurally interesting intermediate explains the syn selectivity of the bromination of strained cycloalkynes.

Full text of DOI:10.1002/anie.200461632

Communications
Reaction Mechanisms
cis-Bromination of Alkynes without Cationic
Intermediates**
Rainer Herges,* Andrea Papafilippopoulos,
Kirsten Hess, Cinzia Chiappe, Dieter Lenoir,* and
Heiner Detert*
Dedicated to Professor Herbert Meier
on the occasion of his 65th birthday
The textbook mechanism of the trans-bromination of alkenes
and alkynes via a more or less symmetrically bridged
bromonium ion does not always apply. Even in the parent
system, the reaction of bromine with ethylene, this mecha-
nism is not correct. Highly pure bromine reacts with ethylene
in anhydrous dichloromethane only very slowly at room
[*] Prof. R. Herges, Dr. A. Papafilippopoulos, K. Hess
Institut fꢀr Organische Chemie
Christian-Albrechts-Universitꢁt zu Kiel
Otto-Hahn-Platz 4, 24118 Kiel (Germany)
Fax: (+49)431-880-2440
E-mail: rherges@oc.uni-kiel.de
Prof. D. Lenoir
Institut fꢀr ꢂkologische Chemie
GSF-Forschungszentrum fꢀr Umwelt und Gesundheit
Postfach 1129, 85778 Neuherberg (Germany)
Fax: (+49)89-3187-3371
E-mail: lenoir@gsf.de
Dr. H. Detert
Institut fꢀr Organische Chemie
Johannes Gutenberg-Universitꢁt
Duesbergweg 10–14, 55099 Mainz (Germany)
Fax: (+49)6131-3925396
E-mail: detert@mail.uni-mainz.de
Prof. C. Chiappe
Dipartimento di Chimica Bioorganica e Biofarmacia
Universitꢃ di Pisa
Via Bonanno 33, 56126 Pisa (Italy)
[**] We are grateful to the Fonds der Chemischen Industrie and the
Deutsche Forschungsgemeinschaft for financial support.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
1412
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200461632
Angew. Chem. Int. Ed. 2005, 44, 1412 –1416

Angewandte
Chemie
temperature. Such mixtures are perfectly stable up to several
days. As soon as light or traces of acids initiate the reaction,
formation of 1,2-dibromoethane proceeds very rapidly and
autocatalytically (t1/2 < 2 min).[1–3]
tetramethylthiacycloheptyne-1,1-dioxide) gives almost exclu-
sively syn products, and we present theoretical evidence that
the intermediate is not a contact ion pair of Br3ꢀ and the vinyl
ꢀ
cation B’. Instead, in solvents of low polarity the Br3 anion
According to the generally accepted mechanism, the
stereoselective anti addition of bromine to alkenes and
alkynes proceeds via bridged bromonium ions A or bromiren-
ium ions A’ (Scheme 1). Aberrations from trans selectivity are
forms a covalent bond with the cationic center (Scheme 1, C’).
This tribromide adduct is formed from a p complex and
rearranges directly to the product (dibromide) without
passing through a cationic intermediate.
According to calculations of Bianchini, Lenoir, and
Goldberg et al.[10] the substituents at the alkyne determine
whether the bromirenium ion A’ or the b-bromovinyl cation
B’ will be more stable. In phenylacetylene the vinyl cation is
favored because the positive charge at the benzylic position is
stabilized by delocalization. Hence, the reaction proceeds in a
nonstereospecific way. In p-nitrophenylacetylene the positive
charge at the benzylic position is destabilized. Therefore the
bridged bromirenium ion A’ is more stable, and the reaction
proceeds with anti selectivity. Apart from the anti-selective
and the nonstereoselective additions there are a number of
syn additions of bromine to alkynes. Uemura et al. observed
the exclusive formation of the cis-dibromo adduct after the
reaction of tert-butylphenylacetylene (1) with bromine in
chloroform at ꢀ508C (Scheme 2).[11] Only after long reaction
Scheme 1. Generally assumed intermediates in the addition of bromine
to alkenes and alkynes A (bromonium ion), A’ (bromirenium ion), B
(b-bromocarbenium ion), B’ (b-bromovinyl cation), and the tribromide
adducts C and C’ identified in the present work by theoretical
calculations.
Scheme 2. Alkynes that selectively add bromine in a syn fashion
forming a cis-dibromoalkene.
explained by the intermediacy of the nonbridged cationic
species (B, B’).[4,5] Polar solvents and cation-stabilizing sub-
stituents favor the nonstereospecific reaction by stabilization
of the b-bromocarbenium ion B or bromovinyl cation B’. In
extreme cases this can result in cis/trans ratios of 1:1. The syn
adduct is formed after a rotation of the bromoalkyl group or
by a frontside attack to the trigonal-planar cation. In alkynes
only a marginal geometrical reorientaion of the counterions
times and higher temperatures is the thermodynamically
more stable and sterically less hindered trans-dibromoalkene
formed. Selectively syn additions of bromine are also
observed for the reaction with alkyl propiolate 2 (which is
deactivated compared to alkyl- and aryl-substituted
alkynes).[12,13] When the bromination is conducted in chloro-
form at 708C and with reaction times of 1–3 h only the cis-
dibromo adducts are formed. trans-Dibromopropenoates can
be obtained selectively by reaction with pyridinium tribro-
mide and bromine.
The syn selectivity obviously holds also for the addition of
bromine to strained cycloalkynes. Krebs et al. observed the
formation of only the cis-dibromo adducts in the bromination
of dibenzocyclooctyne 3,[14] and 3,3,6,6-tetramethylthiacyclo-
heptyne-1,1-dioxide (4)[15] in chloroform or dichloromethane.
According to our investigations, cyclooctyne (5) gives 25% of
the cis-dibromide at ꢀ408C in CDCl3 (detected by independ-
ent synthesis and X-ray analysis, see the Supporting Informa-
tion) and only 3% of the trans product. Besides that,
bromocyclooctene is formed by addition of HBr (for exper-
imental details see the Supporting Information). A single
Brꢀ or Br3 is necessary to initiate a syn addition to give the
ꢀ
cis-dibromo adduct. There are, however, a number of halogen
additions that proceed predominantly or even selectively cis
even in nonpolar solvents. In these cases a totally different
mechanism must be operative. The intermediacy of free
cationic intermediates without stabilization by solvation is
unlikely. A number of authors interpret the predominantly
observed syn addition as the collapse of a contact ion pair
composed of the tribromide anion and the corresponding
carbocation B or B’, which is faster than the reorientation of
ꢀ
the Br3 and a subsequent backside attack.[6–9]
We now present experimental work showing that the
addition of bromine to strained alkynes (cyclooctyne and
Angew. Chem. Int. Ed. 2005, 44, 1412 –1416
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1413

Communications
electron transfer (SET) mechanism leading to the cis product
acetylene·2Br2.[19] According to the calculations three differ-
ent 1:2 complexes are in equilibrium; these complexes are
8.62–10.79 kcalmolꢀ1 (B3LYP/6-31G*) and 6.40–8.02 kcal
molꢀ1 (B3LYP/6-31G(2df)) more stable than the separated
reactants (Scheme 3). In the L-shaped complex 6, a bromine
molecule is coordinated perpendicularly to one of the p bonds
and an additional bromine molecule is bound to the first at an
angle of 908 (Scheme 3, Figure 1). According to our calcu-
lations this is the starting point for the formation of the
bromirenium ion and the polar mechanisms.
Starting from the most stable p complex 7 and proceeding
with an activation energy ((7!9)°) of 23.7 kcalmolꢀ1
(B3LYP/6-31G*) or 24.7 kcalmolꢀ1 (B3LYP/6-31G(2df)),
the structurally interesting tribromide adduct 10 is formed.
There is a very flat minimum 9 between the transition state
(7!9)° and 10, which rearranges to 10 with an activation
barrier of only 0.3 kcalmolꢀ1 (B3LYP/6-31G*) or 0.1 kcal
molꢀ1 (B3LYP/6-31G(2df)) and which therefore does not
have any chemically interpretable significance.
is very unlikely. Kinetic measurements show that the reaction
is of first order in cyclooctyne and second order in bromine.
Moreover, neither the reaction rate nor the reaction order is
changed upon addition of radical scavengers.[16] The addition
of tetrammonium tribromide in polar solvents yields only the
trans-dibromide (see the Supporting Information). Wittig and
Dorsch observed the formation of the cis product upon
reaction with bromine and iron powder.[17]
To elucidate the mechanism of the syn addition we
performed extensive density functional theory (DFT) calcu-
lations first on the potential energy surface of the parent
system acetylene+2Br2 and also on the potential energy
surface of cyclooctyne+2Br2 at the B3LYP/6-31G* + ZPE
and B3LYP/6-31G(2df) + ZPE levels of theory.[18] The sta-
tionary points were characterized by harmonic frequency
analyses, and the topology of the potential energy surface was
checked by IRC calculations. To locate the stationary points,
we used numerous grid calculations as well as other methods.
In agreement with experimental observations of Bianchini
et al.[10] p complexes are formed initially. As in the case of
alkenes the reactive complexes have the stoichiometry
The tribromide adduct 10 diverges from the usual valence
schemes of bromine in organic compounds. The central
ꢀ
bromine atom in the Br3 unit is bound to the neighboring
Scheme 3. Comparison of the bromirenium ion with intermediates in nonionic mechanisms for the bromination of acetylene. The energies of the
stationary points in a vacuum (left) and the SP solvent energies in dichloromethane (B3LYP/6-31G*+ZPE in kcalmolꢀ1, right) are relative to the
most stable p complex 7. The energies in vacuo and in dichloromethane at B3LYP/6-31G(2df) are given in parentheses (relative to the p complex
7 at the same level of theory).
1414
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 1412 –1416

Angewandte
Chemie
the reactants and is a well-defined minimum (also at the MP2/
6-31G* level of theory).
The dibromo adduct 11 is formed from 10 by a “shift” of
the Br3ꢀ unit and elimination of Br2. The activation barrier for
this reaction is 20.3 kcalmolꢀ1 (B3LYP/6-31G*) or 21.8 kcal
molꢀ1 (B3LYP/6-31G(2df)). The product complex 11 is more
stable than the isolated molecules (dibromoethene and
bromine) by 6 kcalmolꢀ1. There is also a concerted six-
center pericyclic transition state (8!11)°, which leads to the
product 11; however, with an activation barrier of DH° =
26.3 kcalmolꢀ1 (B3LYP/6-31G*) or 28.0 kcalmolꢀ1 (B3LYP/
6-31G(2df)), this reaction pathway is significantly less favor-
able. The “textbook intermediate”, the bromirenium ion 12, is
considerably more difficult to describe theoretically than the
less polar species. The relative energies given in Scheme 3 and
Table 1 are therefore less reliable. The relative energy
(134.6 kcalmolꢀ1 (B3LYP/6-31G*) or 133.4 kcalmolꢀ1
(B3LYP/6-31G(2df)) of the separated ion pair 12a can be
considered as the upper limit. As the lower limit, one can
consider the twofold hydrogen-bonded ion pair 12b[22] with
21.2 or 22.2 kcalmolꢀ1.
Single-point energy (SP) solvent calculations using Toma-
siꢁs Polarized Continuum Model (PCM)[23–25] reveal that the
activation barrier (7!9)° leading to the tribromide adduct 10
is drastically reduced even by moderately polar media
(Table 1). The same applies to the bromirenium ion 12. The
very large energy difference between 12a and 12b in vacuo
(103.5 kcalmolꢀ1) is, as expected, drastically reduced in polar
solvents (10.8 kcalmolꢀ1, B3LYP/6-31G(2df)). However, even
in water the energy of 12b is still 4.5 kcalmolꢀ1 higher than
the transition state (7!9)° that leads to the tribromide
adduct 10. The latter is at least 40 kcalmolꢀ1 more stable than
the bromirenium ion 12.
Figure 1. Geometries (B3LYP/6-31G*) of selected stationary points on
the potential energy surface of the addition of bromine to acetylene.
Table 1: Relative energies for the transition states and the ion pairs 12a and 12b in various solvents.[a]
carbon atom with a bond length of
only 1.921 ꢀ. (Figure 1). This bond
is significantly shorter than that in
the p complex (3.081 ꢀ) and not
Solvent
(7!9)°
(10!11)°
(8!11)°
12a
12b
cyclohexane
dichloromethane
methanol
DMSO
water
vacuum
19.48 (20.19)
17.93 (17.43)
16.02 (16.71)
16.90 (16.04)
15.05 (15.74)
23.72 (24.66)
4.96 (5.15)
4.82 (4.35)
3.35 (4.34)
1.64 (3.60)
2.72 (3.57)
9.41 (9.21)
25.43 (27.07)
24.22 (25.16)
22.50 (24.81)
20.87 (24.16)
21.89 (24.09)
28.43 (29.65)
83.31 (75.45)
24.08 (25.86)
21.22 (22.16)
18.89 (21.23)
17.31 (20.58)
18.01 (20.17)
28.52 (29.94)
46.66 (32.99)
26.93 (23.98)
37.77 (22.70)
24.41 (22.03)
134.58 (133.43)
ꢀ
much longer than a “normal” C Br
bond (ꢁ 1.89 ꢀ). Also the charges
(NBO analysis)[20] indicate a cova-
ꢀ
[a] Energies relative to the p complex 7; PCM-SP B3LYP/6-31G* in kcalmolꢀ1. The energies at the B3LYP/
6-31G(2df) level of theory are given in parentheses.
lent C Br bond and do not support
a contact ion pair. The two C atoms
with a charge of ꢀ0.04 are almost
neutral. The central Br bearing the
ꢀ
C Br bond has a positive charge of + 0.57, which is almost
compensated by the two terminal Br atoms (ꢀ0.32 and
Considering the relatively high barriers separating the
tribromide adduct 10 from the reactants and the products, it
should be possible to isolate this intermediate under suitable
conditions. There is indeed evidence that a tribromide
intermediate analogous to 10 is formed in the reaction of
bromine with 4. During the reaction of 4 with two equivalents
of bromine at ꢀ708C a yellow-orange precipitate formed.
When the reaction mixture was allowed to warm up to 08C,
we observed the conversion of the complex to the cis-dibromo
adduct. In the 13C NMR spectrum at ꢀ408C there are only
two signals at 129.79 and 135.19 ppm, which indicate olefinic
rather than cationic carbon atoms. The NMR data indicate
ꢀ
ꢀ0.31). The Br Br bonds with 2.575 and 2.583 ꢀ are similar
ꢀ
in length to that in the tribromide anion (2.606 ꢀ), and the C
C bond length of 1.327 ꢀ corresponds that in 11 (1.330 ꢀ).
Interpreted within the simple valence bond model, the
tribromide anion is a linear 22-electron species with a
trigonal-planar sp3d-hybridized central Br atom. The three
lone pairs occupy the equatorial positions. The interaction of
one of the lone pairs with a cation therefore generates a T-
shaped molecule. Despite its unusual structure[21] the tribro-
mide adduct 10 is more than 10 kcalmolꢀ1 more stable than
Angew. Chem. Int. Ed. 2005, 44, 1412 –1416
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1415

Communications
that the observed intermediate is not symmetric, in agreement
[1] H. S. Davis, J. Am. Chem. Soc. 1928, 50, 2769 – 2780.
[2] D. M. Williams, J. Chem. Soc. Abstr. 1932, 2911 – 2915.
[3] S. Kammermeier, R. Herges, unpublished results.
[4] Review on alkenes: M.-F. Ruasse, Adv. Phys. Org. Chem. 1993,
28, 207.
with an intermediate with a tribromide structure analogous to
10; this is in contrast to the symmetric reactant and the
product of the bromination reaction. Unfortunately, however,
the spectroscopic data are not sufficient to assign the structure
unambiguously. All attempts to determine a crystal structure
were unsuccessful.[26]
[5] Review on alkynes: G. Melloni, G. Modena, U. Tonellato, Acc.
Chem. Res. 1981, 14, 227 – 233.
[6] M.-F. Ruasse, G. L. Moro, B. Galland, R. Bianchini, C. Chiappe,
G. Bellucci, J. Am. Chem. Soc. 1997, 119, 12492 – 12502.
[7] V. F. Anikin, V. V. Veduta, A. Merz, Monatsh. Chem. 1999, 130,
681 – 690.
[8] G. Bellucci, C. Chiappe, R. Bianchini, P. Lemmen, D. Lenoir,
Tetrahedron 1997, 53, 785 – 790.
[9] K. Zates, H. W. Leung, J. Org. Chem. 1980, 45, 1401 – 1406.
[10] R. Bianchini, C. Chiappe, G. L. Moro, D. Lenoir, P. Lemmen, N.
Goldberg, Chem. Eur. J. 1999, 5, 1570 – 1580.
Because of the high stability of the tribromide adduct one
could expect that this species is also an intermediate in
bromination reactions of alkenes. Even in very polar solvents
the two conformations of the tribromide adducts 15 and 16
are significantly more stable than the bridged bromonium ion,
both as a separated ion pair 18a as well as the contact ion pair
18b (see Scheme 4). However, the activation barrier for the
[11] S. Uemura, H. Okazaki, M. Okano, J. Chem. Soc. Perkin Trans. 1
1978, 1278 – 1282.
[12] F. Bellina, A. Carpita, M. de Santis, R. Rossi, Tetrahedron Lett.
1994, 35, 6913 – 6916.
[13] R. Rossi, F. Bellina, A. Carpita, R. Gori, Gazz. Chim. Ital. 1995,
125, 381 – 392.
[14] J. Odenthal, Dissertation, Universitꢂt Heidelberg, 1975.
[15] U. Hꢃpfner, Dissertation, Universitꢂt Heidelberg, 1979.
[16] C. Ciappe, unpublished results.
[17] G. Wittig, H.-L. Dorsch, Liebigs Ann. Chem. 1968, 711, 46 – 54.
[18] Gaussian98 (RevisionA.11.4), M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, R. E. Stratmann, J. C. Burant, S.
Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain,
O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A.
Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J.
V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W.
Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle, J. A.
Pople, Gaussian, Inc., Pittsburgh, PA, 2002.
[19] R. Bianchini, C. Chiappe, D. Lenoir, P. Lemmen, R. Herges, J.
Grunenberg, Angew. Chem. 1997, 109, 1340 – 1343; Angew.
Chem. Int. Ed. Engl. 1997, 36, 1284 – 1287.
[20] A. E. Reed, L. A. Curtis, F. Weinhold, Chem. Rev. 1988, 88, 899 –
926.
Scheme 4. Nonionic mechanisms for the bromination of acetylene.
The energies of the stationary points (B3LYP/6-31G*+ZPE in
kcal molꢀ1) are relative to the most stable p complex 13. The SP
solvent energies in dichloromethane (B3LYP/6-31G* in kcalmolꢀ1) are
also relative to the p complex 13.
[21] M. L. Munzarova, R. Hoffmann, J. Am. Chem. Soc. 2001, 123,
4787 – 4795.
[22] There is no minimum for a contact ion pair composed of the
bromirenium cation and a tribromide anion at the levels of
theory that we applied (B3LYP/6-31G*, B3LYP/6-31(2d,f),
MP2/6-31G*). The stationary point of lowest order that was
found for such a species is the C2v-symmetrical structure 12b
(saddle point). Structure 12b is suitable as a model to estimate
the lower level of the relative energy because the charges come
very close to each other and hence minimize the dipole moment.
[23] S. Miertus, J. Tomasi, Chem. Phys. 1982, 65, 239 – 245.
[24] S. Miertus, E. Scrocco, J. Tomasi, Chem. Phys. 1981, 55, 117 – 129.
[25] M. Cossi, M. Persico, J. Tomasi, J. Am. Chem. Soc. 1994, 116,
5373 – 5378.
formation of 15 from the p complex 14, at least in the parent
system acetylene, is too high to compete with the classical
mechanism via the bromonium ion. In substituted systems
like acenaphthylene, which adds bromine preferentially cis,
however, an intermediate analogous to 15 is likely.[7,8]
Received: August 13, 2004
Revised: November 3, 2004
Published online: January 21, 2005
[26] We are grateful to Prof. Dr. R. Boese, Universitꢂt Essen, for his
attempts to determine the X-ray structure of the intermediate in
the bromination of 4.
Keywords: alkynes · bromination · density functional
calculations · electrophilic addition · reaction mechanisms
.
1416
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 1412 –1416