Bromide assisted addition of hydrogen bromide to alkynes and allenes

Article abstract of DOI:10.1039/a703569a

The addition of 0.1 M quaternary ammonium bromide to a solution of 20% trifluoroacetic acid in methylene chloride causes a large rate increase in the reaction of non-conjugated alkynes. The initial vinyl bromide product reacts further to provide a mixture of isomeric vinyl bromides and dibromides. At high salt concentrations however, the secondary reactions are prevented and only the initial vinyl bromide is found. In contrast to the corresponding addition to alkenes, the predominant mechanism is proposed to involve a one-step, concerted, nucleophilic attack by halide ion upon an acid-alkyne π-complex which produces the vinyl halide directly. A minor path involving syn addition is also seen. At all salt concentrations, allenes produce significant amounts of rearrangement products suggesting the significant involvement of a cationic mechanism.

Full text of DOI:10.1039/a703569a

View Article Online / Journal Homepage / Table of Contents for this issue
Bromide assisted addition of hydrogen bromide to alkynes and
allenes
Hilton M. Weiss* and Kim M. Touchette
Department of Chemistry, Bard College, Annandale-on-Hudson, New York 12504, USA
The addition of 0.1 M quaternary ammonium bromide to a solution of 20% trifluoroacetic acid in
methylene chloride causes a large rate increase in the reaction of non-conjugated alkynes. The initial
vinyl bromide product reacts further to provide a mixture of isomeric vinyl bromides and dibromides. At
high salt concentrations however, the secondary reactions are prevented and only the initial vinyl bromide
is found. In contrast to the corresponding addition to alkenes, the predominant mechanism is proposed
to involve a one-step, concerted, nucleophilic attack by halide ion upon an acid–alkyne ð-complex
which produces the vinyl halide directly. A minor path involving syn addition is also seen. At all salt
concentrations, allenes produce significant amounts of rearrangement products suggesting the
significant involvement of a cationic mechanism.
Table 1 Half-lives (min) for the appearance of product at 20 ЊC
Introduction
The addition of hydrogen halides to alkenes1 and alkynes2 has
20% TFA–CH2Cl2
been postulated to occur by bimolecular and termolecular
mechanisms depending upon whether or not the halide ion is
No added
TFA (100%) salt
Compound
0.1  BrϪ 1.0  BrϪ
specifically involved in the rate determining step of the reaction.
The termolecular reactions had been thought to involve con-
certed bond formation to both elements of the hydrogen halide
leading directly to anti addition products of the alkene and the
alkyne. In a previous paper,3 we showed that this is only true for
alkenes under the most extreme reaction conditions that we
could design. Under most conditions, the slow step involves a
halide assisted protonation of the alkene leading to a secondary
carbocation tightly sandwiched between the attacking halide
anion and the anion derived from the proton-donating
acid. This carbocation then undergoes the usual carbocationic
reactions leading to unrearranged and rearranged halides,
trifluoroacetates and alkene isomers. Only in the most concen-
trated halide solutions with weakened acidity,3 do alkenes
appear to undergo a concerted addition. The most persuasive
evidence for the concerted mechanism with alkenes was the lack
of rearrangement found in the addition of hydrogen iodide to
3,3-dimethylbut-1-ene.
The absence of rearrangement in the addition of hydrogen
halides to alkynes is less significant since the rearrangement
of vinyl cations is not facile. Solvent capture to form vinyl
esters and ketones is a signature of a vinyl cation4 but these
reactions too may be minimal in a tight ion pair mechanism.
The concerted addition of HCl to alkynes has been reported2
to produce only the Markovnikov product with essentially
100% anti stereoselectivity and we believe that this is the best
chemical evidence for a concerted addition to an alkyne.
Since protonation of an internal alkyne would produce a vinyl
cation with a steric preference for halide capture syn to the
added proton, an anti adduct would be symptomatic of a
Oct-1-yne
Oct-1-yne
Oct-1-yne
Octa-1,2-diene
Oct-1-ene
1600
1400
480
62 000
82 000
9000
2800
5730
360
72
60
15
29
80
5200
1100
1000
3400
7500
100
2-Bromooct-1-ene
89 000
400 000
Results and discussion
Effect of bromide ion† on rates
Non-conjugated alkynes have been found4 to react with neat
trifluoroacetic acid (TFA) at 60 ЊC to form a mixture of cis and
trans trifluoroacetates, ketones and arene trimers. The mechan-
ism is thought to involve an intermediate vinyl carbocation.
The half-life reported4 for the reaction of a terminal alkyne is
43 min at 60 ЊC [and we have found a half-life of 27 h for oct-1-
yne at room temperature (20 ЊC)]. If a solution of 20% TFA in
methylene chloride is used, the half-life for TFA addition
increases to 1040 h producing a similar mixture of products.
When we include 0.1  tetra-n-butylammonium bromide in the
latter reaction mixture, the half-life for the appearance of oct-1-
yne products drops to 6 h and only HBr adducts are found.
With the rate increasing by a factor of 170 and the exclusive
formation of bromide products, it seems safe to assume that the
bromide ion is involved in the rate determining step of all prod-
ucts. Similar results are found for internal alkynes and these
results are shown in Table 1.
In each case, the inclusion of 0.1  bromide increases the rate
by two orders of magnitude indicating significant participation
of the bromide ion. It should also be noted that the higher
concentration of bromide slows the reaction of the alkyne by
an order of magnitude but slows the subsequent reaction of the
vinyl bromide product by over three orders of magnitude. Thus,
the initial products of the reaction are stable to the concen-
trated bromide conditions and can easily be isolated using these
conditions.
R
C
C+
R
H
X–
concerted addition. Although such results have been reported
previously, we wished to analyze these reactions more care-
fully using our previously reported3 reaction conditions.
We therefore undertook this study to ascertain whether an
intermediate carbocation is also involved in these reactions
or whether the alkenes and alkynes react by different
mechanisms.
Because our parallel study with alkenes showed similar evi-
† In this paper, we use the term ‘bromide ion’ to include all ion pairs
and free bromide ions.
J. Chem. Soc., Perkin Trans. 2, 1998
1523

View Article Online
dence for bromide ion participation in the transition state, we
conducted a more detailed kinetic analysis of alkyne disap-
pearance relative to an internal decane standard. These results
parallel those obtained during the product studies but the
half-lives are 30 to 50% shorter.
Identification of products by GC–MS
Over a dozen different reaction products were detected and ana-
lyzed by GC–MS. All peaks above 1% of total were identified
and many peaks below that level could also be identified. Their
identification was achieved by a combination of GC retention
times and mass spectral analyses. In general, vinyl bromides
have large molecular ion peaks whereas allyl bromides show no
peaks containing bromine. Fragmentation of vinyl bromides
appears to pass through an intermediate with radical character
on the β-carbon. This intermediate undergoes a radical frag-
mentation where possible or an internal hydrogen abstraction
followed by an alkene cleavage (Scheme 1). The two routes are
Br+
Br
Br+
Mol Wt
is Odd
H
Fig. 1 NMR spectra of the vinyl region. (a) 1H of (E)-2-bromo-1-
H
H
deuteriooct-1-ene, (b) 2H of (E)-2-bromo-1-deuteriooct-1-ene, (c) 1H of
2-bromooct-1-ene.
Further
Fragmentation
3,3-dimethylbut-1-yne 5.37 ppm (syn to bromine). The addition
of HBr to oct-4-yne affords the (Z)-4-bromooct-4-ene with a
single proton resonance at 5.61 ppm (anti to bromine). The
addition of HI to dec-1-yne gives 2-iododec-1-ene with peaks at
5.99 ppm (anti to iodine) and 5.66 ppm (syn to iodine).
Br+
+Br
Mol Wt
is Even
H
H
Product distributions
Using 1.0  bromide, we found that oct-4-yne produces only
one product, 4-bromooct-4-ene (Scheme 2) and that greater
Scheme 1
rarely possible in the same molecule. For example, 2-bromooct-
2-ene only loses the butyl radical generating a cation with mol-
ecular weights of 133 and 135. The 3-bromooct-2-ene preferen-
tially abstracts a hydrogen atom and loses butene generating
a cation with molecular weights 134 and 136. The corre-
sponding trifluoroacetates appear to follow the same fragmen-
tation paths. Because the (Z)-2-bromooct-2-ene and the (Z)-3-
bromooct-2-ene have very similar retention times, a clear separ-
ation of these isomers was only feasible where similar amounts
of each isomer were present. The area ratios of mass peaks 134
to 133 were calculated from these and used to calculate the
isomer ratios in other mixtures. All pairs of (E)- and (Z)-
isomers give virtually identical mass spectra. The (Z)-vinyl
bromides are the predominant isomers and have shorter reten-
tion times than the (E)-isomers. In some cases, the (Z)-isomers
were isolated and identified by NMR spectroscopy.
Br
Br
Z4B4
E4B4
Oct-4-yne
Br Br
44diB
Reaction with 20%
TFA–CH2Cl2 and
0.1  Bu4Nϩ BrϪ.
Product distribution
from oct-4-yne
Reaction with 20%
TFA–CH2Cl2 and
1.0  Bu4Nϩ BrϪ.
Product distribution
from oct-4-yne
t/h
%Rxn Z4B4 E4B4 44diB
t/h
%Rxn Z4B4
The allyl bromide retention times and mass spectra were
reproduced by the reaction of oct-1-ene with N-bromo-
succinimide. The stereochemistry of the 1-bromooct-2-enes was
assigned solely on the basis of retention time and relative yields.
0.58 45
1.0 58
1.75 72
96.3
95.6
95.9
84.9
3.7
3.6
3.2
3.4
—
0.8
0.9
11.9
3
17
25.4 72
144 100
100
100
100
23.5 100
Scheme 2
Identification of products by NMR spectroscopy
The addition of HBr to oct-1-yne produces 2-bromooct-1-ene
with two olefinic resonances. The hydrogen syn to the bromine
is a doublet of triplets centered at 5.37 ppm. The hydrogen anti
to the bromine is an overlapping doublet of triplets centered at
5.54 ppm. The addition of DBr to oct-1-yne produces (E)-2-
bromo-1-deuteriooct-1-ene with a single proton peak at 5.36
ppm and a deuterium absorbance at 5.52 ppm. These peaks are
shown in Fig. 1.
than 99.9% of this product is the (Z)-isomer (Z4B4). No other
products can be detected after 10 half-lives. When 0.1  brom-
ide is used, 4-bromooct-4-ene remains the initial product but it
contains a few percent of the (E)-isomer (E4B4). The ratio of
(Z)- to (E)-isomer remains constant throughout 90% of the
reaction suggesting that both isomers are formed directly from
the alkyne. No ketone or trifluoroacetate esters could ever be
detected and we have shown that such compounds are stable
and detectable under the conditions of the reaction.
The addition of DBr to the following alkynes gives deuter-
ium decoupled NMR spectra with single proton peaks at the
following frequencies: hex-1-yne 5.38 ppm (syn to bromine);
The reaction of oct-2-yne with 1.0  bromide is quite similar,
producing a mixture of (Z)-2-bromooct-2-ene (43%) (Z2B2)
1524
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
Oct-2-yne
Br
Br
Z2B2
E2B2
Br
Z3B2
Br
Br
E3B2
22diB
Br
Z3B3
Br
Br
Br
33diB
Product distribution from oct-2-yne
reaction with 20% TFA–CH2Cl2 and
0.1  Bu4Nϩ BrϪ
Product distribution from oct-2-yne
reaction with 20% TFA–CH2Cl2 and
1.0  Bu4Nϩ BrϪ
t/h
%Rxn Z2B2 Z3B2 E2B2 E3B2 Z3B3 22diB 33diB
t/h
26
336
720
%Rxn Z2B2 Z3B2
0.58 35
2.8 78
28.3 100
39.6
40.1
46.6
58.0
54.9
53.4
1.0
0.6
—
1.4
1.4
—
—
0.5
—
—
0.6
5.8
—
1.9
17.4
87
100
100
43.4
41.7
42.7
56.3
58.3
57.3
Scheme 3
and (Z)-3-bromooct-2-ene (57%) (Z3B2) as the only products
after all alkyne has reacted (Scheme 3). No (E)-isomers can be
detected. Doubling the reaction time affords no further reaction.
Trace amounts of a dibromooctene can be detected early in the
reaction but it does not increase with time and it was found to
derive from a small amount of bromine which is generated
initially and quickly adds to the alkyne. The GC retention time
and mass spectrum can be reproduced by adding bromine to
oct-2-yne in carbon tetrachloride. When the HBr additions are
performed in the dilute bromide solutions, the same mixture of
(Z)-vinyl bromides is found early in the reaction along with a
small amount of the (E)-isomers but these are subsequently
converted to an increasing amount of 2,2- and 3,3-dibromides
(22diB and 33diB).
The reaction of oct-1-yne is slightly more complicated. As
expected, oct-1-yne reacts with the 1.0  bromide solution to
form 2-bromooct-1-ene quite cleanly, showing only trace
amounts of the 2-bromooct-2-ene. This product can be isolated
easily and subjected to analysis and further reaction. It is also
possible, using deuterated trifluoroacetic acid, to add DBr to
oct-1-yne and to show by NMR spectroscopy that the addition
occurs with 97 3% anti steroselectivity. In this experiment, we
are unable to detect any deuterium exchange into the unreacted
starting material. Unlike the internal alkenes, the 2-bromooct-
1-ene undergoes some further reaction (isomerization and
addition) before the alkyne disappears completely. These sec-
ondary reactions remain slow and account for less than 5% of
the products.
acetates) coupled with the high proportion of anti addition
products suggests that the addition to the alkyne proceeds by a
concerted mechanism (i.e. both covalent bonds to the alkyne
form simultaneously). The small amount of anti proton
detected in the NMR spectrum of 2-bromo-1-deuteriooct-1-ene
may be an artifact arising from protic impurities in the reaction
mixture. Thus, the NMR analysis is less dependable than the
GC results for assessing stereochemical purity.
An important difference, however, is seen in the reaction of
the allene octa-1,2-diene (Scheme 5). Using 1.0  bromide, the
simple addition of HBr to the allene system can account for the
2-bromooct-1-ene (2B1), the (E)- and (Z)-2-bromooct-2-ene
(E2B2 and Z2B2) and the 3-bromooct-1-ene (3B1). However, a
significant amount of oct-2-yne is also detected early in the
reaction. That the oct-2-yne concentration increased to
approximately 7% of the products suggests that at least 7% of
the reaction proceeds through a carbocation. The absence of
3-bromooct-2-ene in the product mixture suggests that oct-2-
yne is not an intermediate on the route to other products
and that the other products derive from a short lived cation
sandwich which can also form the oct-2-yne.
In the more acidic3 0.1  bromide solution, the presumed
carbocation is less associated and produces a wider variety of
products including oct-2-yne and its addition products, 3-
bromooct-2-ene and
a proportionate amount of the 2-
bromooct-2-ene along with some trifluoroacetates. These prod-
ucts account for over 40% of the total products and support the
intermediacy of a carbocation intermediate. The rate enhance-
ment found with the bromide ion coupled with the paucity of
trifluoroacetate products‡ (Tf) suggests that the ion sandwich
mechanism, rather than a free carbocation, is the predominant
path for the allene reaction. That the allene forms a vinyl cation
while the alkynes do not form the same cation is presumably
due to the greater instability of the allene as evidenced in the
predominance of alkynes in the acid catalyzed isomerization5
of allenes and alkynes.
When 0.1  bromide is used, these secondary reaction prod-
ucts are seen in the earliest stages of the reaction and become
the principal products before the alkyne is half reacted. After
approximately 5 h (45% reaction), the product mixture contains
2-bromooct-1-ene (28%) (2B1), and (Z)-2-bromooct-2-ene
(Z2B2), (E)-2-bromooct-2-ene (E2B2) and 2,2-dibromooctane
(22diB) in the ratio 2:1:4 (Scheme 4). When the 2-bromooct-1-
ene is isolated and subjected to the 0.1  bromide reaction con-
ditions, a very similar mixture of dibromide and vinyl bromides
is produced. Eventually, the dibromide becomes the major
product.
‡ The allyl trifluoroacetates may well be converted to bromides under
these reaction conditions. The vinyl trifluoroacetates, however, are
stable to these reaction conditions.
The absence of cation derived products (ketone, trifluoro-
J. Chem. Soc., Perkin Trans. 2, 1998
1525

View Article Online
Oct-1-yne
Z2B2
2B1
Br
Br
Br
Br
E2B2
Z3B2
Br
Br
22diB
Product distribution from oct-1-yne reaction with
20% TFA–CH2Cl2 and 0.1  Bu4Nϩ BrϪ
Product distribution from oct-1-yne reaction with
20% TFA–CH2Cl2 and 1.0  Bu4Nϩ BrϪ
t/h
0.58
1.2
4.9
28.3
75
%Rxn
2B1
Z2B2
Z3B2
E2B2
22diB
t/h
48
336
720
%Rxn
2B1
100
98.0
94.0
Z2B2
E2B2
22diB
6
25
56
94
100
75.3
57.8
28.0
4.7
8.9
14.1
19.9
24.4
15.9
—
—
3.0
2.7
0.6
6.2
9.5
10.7
11.1
7.0
9.6
18.6
38.4
57.5
73.5
52
100
100
—
0.4
2.3
—
0.6
1.0
—
1.0
2.7
3.0
Products from 2-bromooct-1-ene
1.2 56 44 23.6
—
11.9
64.5
Scheme 4
Conclusions
This work was aimed at answering the question of whether, as
suggested, hydrogen halides can add to alkynes by a concerted
mechanism. Our previous work shows that such a mechanism is
extremely rare for alkenes since skeletal rearrangements are
pervasive in these reactions and the reaction rates were found to
be independent of halide ion concentration. Using the same
reaction conditions that we employed in our alkene study, we
have found strong evidence for the concerted mechanism with
alkynes. Persuasive evidence for this is the stereospecific
addition of HBr to the internal alkynes which can be shown to
occur with 95 to 99.9% anti stereospecificity. The NMR analy-
ses of the DBr adducts to the terminal alkynes, hex-1-yne, oct-
1-yne and 3,3-dimethylbut-1-yne are relatively less sensitive and
more vulnerable to protic impurities in the reaction mixture as
well as acid catalyzed isomerization of the vinyl halide products.
We realize that the 97 2% stereoselective anti additions to
oct-2- and -4-yne in the less concentrated bromide solutions
could be interpreted as coming from a very tight ion pair with
some leakage to produce the syn adduct. However, the stereo-
specific results obtained with the more concentrated bromide
solutions generate an unreasonable rate requirement for the ion
pair collapse and a concerted mechanism seems unavoidable.
Although it might be argued that a tight ion pair might always
collapse to the anti adduct, we can see no difference between
this situation and an unsymmetrically bonded concerted mech-
anism. In either case, there is halide bonding to the developing
cationic center and a single downhill path to one product.
The kinetic study of the reactions of oct-1- and -4-yne with a
range of bromide ion concentrations shows behaviour similar
to that seen with the alkenes. However, while the logarithms
of the alkene rate constants are found3 to be proportional
to Ho Ϫ pKa, this relationship is not linear with the alkynes.
When the pseudo first order rate constants for the octynes are
divided by the concentrations of bromide ion, the negative
logarithms of these values are found to show a linear relation-
ship to Ho Ϫ pKa with proportionality constants of 0.8 and 0.9
(Fig. 2).
Fig. 2 Pseudo first order rate constants for oct-1-yne and oct-4-yne
divided by bromide versus the acidity of the solutions (Ho Ϫ pKa)
We also found that the rate constants for HBr addition to
alkynes are not proportional to those found for the alkene add-
itions. Dividing the rate constants for alkyne addition by the
concentrations of bromide used produces values which are
proportional to the alkene rate constants. These results are
shown in Figs. 3 and 4.
This somewhat surprising result supports our thesis that,
while the alkenes react by an AdE2 mechanism independent of
bromide, the alkynes show termolecular kinetics with the add-
itional dependence upon the tetrabutylammonium bromide.
This provides additional support for the concerted reaction of
alkynes and the cationic mechanism for alkenes. That the lines
are straight and intersect the origin suggests that no other
mechanisms are significantly involved at any bromide concen-
tration. The surprising linear dependence of the rates upon the
concentration of added tetrabutylammonium bromide suggests
that either, 1) the effectiveness of the halide ion is independent
1526
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
Octa-1,2-diene
Br
Oct-2-yne
Z2B2
2B1
Br
E2B2
Br
Br
Br
Z3B2
E1B2
3B1
Br
Br
E3B2
Br
Br
Z3B3
Z1B2
Z2Tf2
TfO
TfO
Br
E2Tf2
Br
Br
Br
22diB
33diB
Product distribution from octa-1,2-diene reaction with 20% TFA–CH2Cl2 and 0.2  Bu4Nϩ BrϪ
t/h
%Rxn 2B1
Z2B2
Z3B2
E2B2 E3B2 Z3B3 Z1B2 E1B2 3B1 E2Tf2
Z2Tf2
22diB
33diB Oct-2-yne
3.5
48
99.5
100
0.8
0.6
46.8
42.7
20.0
16.3
22.0
17.6
0.7
0.6
—
1.7
0.9
0.7
1.0
1.8
1.7
—
0.7
0.5
1.3
1.2
1.0
10.1
0.4
5.4
2.7
—
Product distribution from octa-1,2-diene reaction with 20% TFA–CH2Cl2 and 1.0  Bu4Nϩ BrϪ
t/h
%Rxn
2B1
Z2B2
Z3B2
E2B2
Z1B2
E1B2
3B1
Oct-2-yne
1
3.5
28
4.8
7.3
38
15.5
13.0
10.9
30.0
34.8
29.1
—
—
4.2
32.0
29.3
30.0
—
—
3.7
—
—
3.6
16.2
15.7
16.8
6.0
7.2
1.7
Scheme 5
Fig. 3 Pseudo first order rate constants for oct-1-yne (ᮀ) and oct-1-
yne divided by bromide concentrations (᭹) versus pseudo first order
rate constants for oct-1-ene
Fig. 4 Pseudo first order rate constants for oct-4-yne (ᮀ) and oct-4-
yne divided by bromide concentrations (᭹) versus pseudo first order
rate constants for oct-4-ene
J. Chem. Soc., Perkin Trans. 2, 1998
1527

View Article Online
of its degree of association in these solutions or, 2) the salt
maintains the same degree of association throughout the range
of salt concentrations which were studied. The latter possibility
seems more likely with the salt existing as ion pairs in this
solvent system.
ml) were removed and quenched with 15 ml of water and 10 ml
of hexanes. The hexanes layer was washed with another 10 ml
of water and dried over anhydrous potassium carbonate prior
to GC–MS analysis. Chromatographic peaks were identified by
their mass spectrum as well as by retention time of compounds
synthesized and analyzed by NMR spectroscopy. Detector
responses for the vinyl bromides relative to oct-1-yne were
determined by synthesis of a bromide mixture adding a known
amount of oct-1-yne and analyzing the mixture by GC–MS.
Detector responses for all the vinyl bromides were assumed to
be equal. The vinyl trifluoroacetates and halides were shown to
be stable under the reaction conditions. Unless otherwise noted,
all reactions were run at room temperature (20 2 ЊC) and
showed no evidence of exothermicity. Decane was used as an
internal standard in all kinetic studies designed to determine
the overall rate of disappearance of alkyne.
NMR spectra were recorded on a JEOL FX90Q spectro-
meter. Proton spectra utilized a deuterium lock and TMS as
internal reference. Deuterium spectra utilized a lithium lock
and CDCl3 (δ 7.24) as the internal reference. Mass spectra and
chromatographic analyses were performed on a Hewlett-
Packard 5890 Chromatograph with a 12 m HP-1 capillary
column and a 5971A mass selective detector.
This paper addresses the question of the factors which
determine whether the electrophilic attack will generate a carbo-
cation or directly produce the halide product. We believe that
the greater instability of the vinyl cation formed from an alkyne
requires a greater association of the nucleophile to the incipient
cationic center before proton transfer to the alkyne can be
completed. The proven ability of alkynes6 to form π-complexes
with strong acids may also facilitate the concerted mechanism.
The five-fold greater reactivity of the internal alkynes over the
terminal alkyne may derive from their greater π-donor poten-
tial. In contrast, the (less stable) terminal alkenes are found to
be slightly more reactive than the internal alkenes supporting
the idea of a slow formation of the cation.
In our previous paper,3 we asserted that the ionic additions of
all protic reagents to alkenes probably occur via cationic inter-
mediates. In this work, we find that the addition of hydrogen
halides to alkynes easily occurs by a concerted mechanism. We
must point out, however, that there is ample evidence for cat-
ionic intermediates in the addition of other reagents to alkynes.
The most direct evidence is the isomerization of alkynes by
strong acids,5 the rearranged products coming from the reaction
of 3,3-dimethylbut-1-yne with liquid HCl7 and the non-
stereospecific addition of trifluoroacetic acid to hex-3-yne.4 It
should be noted that, in each of these cases, there were strong
acids and weak nucleophiles present in the reaction. Thus, in
contrast to the reaction of alkenes, most termolecular additions
to an unconjugated alkyne are unlikely to proceed through
cationic intermediates.
Finally, we would like to point out the synthetic potential of
this reaction. Using the most concentrated salt solution, the
reactions proceed in essentially 100% yield and formed only the
Markovnikov products with anti stereospecificity. There was no
evidence of the radical reactions seen2a in other HBr additions.
A similar reagent has also been reported8 for the addition of
HBr to alkynes.
General method for preparative HBr (or DBr) additions
Tetra-n-butylammonium bromide (16.12 g, 50 mmol) was put
into a 50 ml volumetric flask and dried overnight in a vacuum
desiccator. The salt was dissolved in approximately 25 ml of
methylene chloride followed by trifluoroacetic acid (10 g, 6.7
ml, 89 mmol). The appropriate alkyne (10 mmol) was added.
The reaction mixture was protected from light and allowed to
react at room temperature for 5 to 7 days (monitored by
chromatography). The entire reaction mixture was transferred
to a separating funnel, washed with 2 × 100 ml water, 1 × 100
ml saturated NaHCO3, 1 × 100 ml water, then dried over
K2CO3 and concentrated to give a colorless oil. The solution
was diluted with 50 ml of pentane and washed with 2 × 50 ml
water and 1 × 20 ml brine then dried over K2CO3, filtered
and concentrated under reduced pressure. At this point, the
products were sufficiently pure for GC and NMR analyses.
Experimental
References
1 (a) R. C. Fahey and C. A. McPherson, J. Am. Chem. Soc., 1971, 93,
2445; (b) R. C. Fahey, M. W. Monahan and C. A. McPherson, J. Am.
Chem. Soc., 1970, 92, 2810; (c) R. C. Fahey and M. W. Monahan,
J. Am. Chem. Soc., 1970, 92, 2816; (d) R. C. Fahey, C. A. McPherson
and R. A. Smith, J. Am. Chem. Soc., 1974, 96, 4534.
2 (a) R. C. Fahey and D.-J. Lee, J. Am. Chem. Soc., 1968, 90, 2124;
(b) R. C. Fahey and D.-J. Lee, J. Am. Chem. Soc., 1967, 89, 2780; (c)
R. C. Fahey, M. T. Payne and D.-J. Lee, J. Org. Chem., 1974, 39, 1124.
3 H. M. Weiss and K. T. Touchette, preceding paper.
4 P. E. Peterson and J. E. Duddey, J. Am. Chem. Soc., 1963, 85, 2865;
1966, 88, 4990.
5 B. J. Barry, W. J. Beale, M. D. Carr, S.-K. Hei and I. Reid, J. Chem.
Soc., Chem. Commun., 1973, 177.
6 D. Mootz and A. Deeg, J. Am. Chem. Soc., 1992, 114, 5887.
7 K. Griesbaum and Z. Rehman, J. Am. Chem. Soc., 1970, 92, 1416.
8 J. Cousseau, Synthesis, 1980, 805.
The oct-1-yne (99.5% purity), oct-2-yne, oct-3-yne and oct-4-
yne used in these experiments were obtained from Farchan
Chemical Co. The 3,3-dimethylbut-1-yne, trifluoroacetic acid
and solvents (HPLC grade) were obtained from Aldrich
Chemical Co. and were used without further purification. The
octa-1,2-diene was synthesized by the method previously9
described. The quaternary ammonium salts were obtained from
Fluka Chemical and from Aldrich Chemical Co. and were kept
in a desiccator prior to use. The tetrabutylammonium bromide
was dried under vacuum at regular intervals.
Reactions were performed in glass-stoppered flasks by add-
ing 1 or§ 2 drops of the alkyne¶ to 5 to|| 100 ml of the 20%
trifluoroacetic acid in methylene chloride solution containing
the quaternary ammonium salt. Aliquots (approximately 0.5
9 D. J. Pasto, S.-K. Chou, A. Waterhouse, R. H. Shults and G. F.
Hennion, J. Org. Chem., 1978, 43, 1385.
§ Results from 1 or 2 drops of alkyne (approximately 10Ϫ4 mol) were
identical within our experimental method.
Paper 7/03569A
Received 23rd May 1997
Accepted 9th February 1998
¶ In all rate studies, a mixture of alkyne and decane standard was used
and analyzed after the standard workup.
|| At least 10Ϫ3 mol of bromide ion were used in all rate studies.
1528
J. Chem. Soc., Perkin Trans. 2, 1998