Halide assisted addition of hydrogen halides to alkenes

Article abstract of DOI:10.1039/a703570e

The addition of 0.1 M quaternary ammonium halide to a solution of 20% trifluoroacetic acid in methylene chloride causes a large rate increase in the reaction of simple alkenes leading to a mixture of alkyl halides and trifluoroacetates. The mechanism is proposed to involve a halide assisted protonation of the alkene which produces a carbocation intermediate sandwiched between the attacking halide ion and the trifluoroacetate ion. At higher concentrations of halide ion, the proton donating ability of the solution decreases, slowing the reaction and increasing the efficiency of cation capture by the halide ion. This leads to a greater proportion of unrearranged halide product. At the highest concentration of halide ion, cation rearrangement is virtually eliminated.

Full text of DOI:10.1039/a703570e

View Article Online / Journal Homepage / Table of Contents for this issue
Halide assisted addition of hydrogen halides to alkenes
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 halide to a solution of 20% trifluoroacetic acid in methylene
chloride causes a large rate increase in the reaction of simple alkenes leading to a mixture of alkyl halides
and trifluoroacetates. The mechanism is proposed to involve a halide assisted protonation of the alkene
which produces a carbocation intermediate sandwiched between the attacking halide ion and the
trifluoroacetate ion. At higher concentrations of halide ion, the proton donating ability of the solution
decreases, slowing the reaction and increasing the efficiency of cation capture by the halide ion. This leads
to a greater proportion of unrearranged halide product. At the highest concentration of halide ion, cation
rearrangement is virtually eliminated.
dependent upon the halide ion concentration) often competes
Introduction
with the solvent assisted AdE2 mechanism and the alkyl halide
products usually show increased anti stereoselectivity as the
concentration of halide ion is increased. This mechanism is
generally postulated to involve a nucleophilic attack by the
halide ion upon a π-complex or encounter complex formed
between the alkene and the hydrogen halide molecule ‘bypass-
ing a discrete carbocation’.5 Both of these mechanistic possi-
bilities have been called ‘concerted AdE3’ and we shall continue
that somewhat misleading tradition.
Since the stereochemical integrity and the lack of rearrange-
ment inherent in such a mechanism could be synthetically
useful, we decided to study this type of reaction further. To
minimize the complications of a concomitant AdE2 reaction
occurring, we used methylene chloride as solvent since it is less
capable than acetic acid at solvating the potential carbocation.
We also chose to utilize trifluoroacetic acid as the sole proton
source to control and to limit the acidity of the reaction mixture
while using different halide ions. This allowed us to add the
elements of HBr without the presence of molecular HBr which
is known to cause unwanted radical addition and isomeriz-
ation.6 By avoiding the hydrogen halides as proton donors,
we also avoided the production of syn-generated ion pairs.
Trifluoroacetic acid is also advantageous because of its low
nucleophilicity7 and its availability in deuterated form.
The electrophilic addition of hydrogen halides to alkenes is a
versatile synthetic reaction which has promulgated a wide range
of mechanistic possibilities.1 Perhaps the best known mechan-
ism is the AdE2 path in which slow proton transfer to the alkene
generates a carbocation and halide anion. In polar solvents,
alkenes capable of forming stable, usually conjugated, carbo-
cations undergo this AdE2 mechanism and show little or no
stereoselectivity in their addition products.
In less polar solvents, these conjugated alkenes are less
reactive. When proton transfer to the alkene does occur, the
carbocation, being less well solvated, maintains a strong attrac-
tion for the halide anion and this ion pair is likely to collapse
rapidly to the addition product. In such cases, a syn addition
product often predominates (Scheme 1).2
DBr 0 °C
CH CH CH3
CHBr CHD CH3
CH2Cl2
cis
81% syn addition
88% syn addition
trans
Scheme 1
With simple alkenes which form less stable (secondary) carbo-
cations, stereorandom addition is rare. In fairly nucleophilic
solvents such as acetic acid, proton transfer is slow and is
assisted by the nucleophilic solvation of the developing carbo-
cationic center during the proton transfer. In acetic acid, sol-
vent incorporation is common and the alkyl acetate products
usually show anti stereochemistry3 indicative of ‘backside’ sol-
vation followed by rapid collapse to the alkyl acetate (Scheme
2). In these reactions, the hydrogen halide adducts also show a
Results
Although monosubstituted alkenes are fairly reactive toward
neat trifluoroacetic acid7 (t1/2 ≈ 1 h), we found that the half-life
of a monosubstituted alkene in 20% TFA–methylene chloride
at room temperature is approximately 95 h. The inclusion of a
0.1  tetra-n-butylammonium halide ion in this reaction mix-
ture produces alkyl chloride, bromide or iodide in reactions
with half-lives of 6.6, 0.5 and 0.2 h respectively.
Although kinetic analyses of these reactions are difficult, it
seems clear that the alkene, acid and salts are involved in the
rate determining step. Furthermore, with the tremendous rate
acceleration caused by the added halide ion, it is safe to assume
that virtually all of the products arise from the halide assisted
reaction.
D
H
H
H
D
H
D
H
H
Cl
Cl
OAc
0.21 M DCl–DOAc 25 °C
14%
4%
<4%
80%
30%
0.15 M DCl–DOAc 25 °C
with 0.43 M Bu4NCl
>67%
Effect of bromide concentration on rates
We studied these reactions on a small scale using approximately
10Ϫ4 mol of alkene in a large excess of acid and bromide ion.
Under these conditions, the reaction is first order in alkene for
60% of the reaction. For the product studies, samples were
removed for chromatographic analysis and corrected for the
Scheme 2
preference for anti stereochemistry presumably arising from an
analogous ‘backside’ participation by the halide ion during the
proton transfer step.4 This AdE3 mechanism (additionally
J. Chem. Soc., Perkin Trans. 2, 1998
1517

View Article Online
Table 1 Half-lives, relative product yields and relative acidities for the
reaction of oct-1-ene with 20% TFA in CH2Cl2 with added tetrabutyl-
ammonium bromide
Table 2 Product composition (%) from oct-1-ene at 34 4% reaction
in 20% TFA–CH2Cl2
[BrϪ]/ 2-RBr 3-RBr 4-RBr
2-RTFAc 3-RTFAc 4-RTFAc
RBr/
RTFAc
log [In]/
[HInϩ]
[BrϪ] used/
t1/2/min
0.01
0.1
1.0
0.1
75.8
80.5
99.3
45
6.0
6.5
0.1
0
1.0
0
(from
trans-oct-
2-ene)
49
18.2
9.9
0.6
0
1.7
0
0
0.4
0
0.00
0.01
0.02
0.04
0.06
0.08
0.10
5730
69
44
32
26
24
29
—
0.08
4.5
55
7.6
8.7
12.9
10.0
8.2
0.1
51
(from
trans-oct-
3-ene)
Ϫ1.61
0.20
0.40
0.60
0.80
45
71
185
1300
7500
396
29
9.3
Ϫ1.93
Ϫ2.01
Ϫ1.52
Ϫ0.78
0.40
15.2
18.2
32.6
135
quaternary ammonium salts, small amounts of trifluoroacetate
ion were also added as a buffer to some reactions. Trifluoro-
acetate is found to have a small retarding effect on the hydrogen
halide addition reactions and is assumed to cause the deviation
from first order kinetics found late in the reactions.
1.00
0.10 (ClϪ)
3.5 (ClϪ)
0.10 (BrϪ)
8.2 (BrϪ)
20.3 (IϪ)
0.10 (IϪ)
12
0.10 (Bu4N O2CCF3)
0.10 (Bu4N Br), 0.012
(Bu4N O2CCF3)
0.10 (Bu4N Br), 0.071
(Bu4N O2CCF3), 0.89
(Bu4N BF4)
4800
50
—
9.2
5.5
Effect of bromide concentration on the acidity of the solution
The addition of tetrabutylammonium bromide to the 20% tri-
fluoroacetic acid solution dilutes the acid somewhat but exerts
its major influence on the activity coefficients of the solutes,
especially the ionic solutes. To assess the effect of the added salt
upon the acidity of these solutions, we determined its ability
to protonate a neutral Hammett base. We spectrophoto-
metrically6,8 determined that 4-chloro-2-nitroaniline is 50%
protonated in 20% TFA–methylene chloride. The addition of
small amounts of the salt increases the protonated form of
the Hammett base to 97% but additional salt eventually lowers
the degree of protonation to 37% at 1.0  salt. Although we
have not determined the pKa of 4-chloro-2-nitroaniline in our
solvent system, we can evaluate the relative acidity of these
solutions (Table 1) by determining the ratio of unprotonated to
protonated indicator. Graphing these results (Fig. 1) suggests
that the reaction rates parallel the acidity of the solutions.
15 800
Effect of bromide concentration on product composition
At the lower concentrations of bromide ion, a 10-fold increase
in bromide ion leads to little change in the product com-
positions (Table 1). Higher bromide concentrations cause a
significant increase in the alkyl bromides and the highest con-
centrations also generate a greater proportion of unrearranged
products (Table 2). At all concentrations of bromide ion, the
alkyl bromides are the primary products reinforcing the notion
that the bromide ion is involved in the transition state. The
proportion of bromide products also parallels the acidity of the
reaction (Fig. 1) solutions except when the bromide ion concen-
tration is extremely low.
Fig. 1 Logarithmic graph of t1/2, product ratios and relative acidity
of reaction solution as functions of the bromide concentration in the
oct-1-ene reaction solutions
differential detector responses of reactants and products. In
later kinetic studies, alkene disappearance was referenced to
an internal standard. The resulting rate data parallel the
rates found during the product studies but are 50 to 90%
larger. All kinetic data are treated as pseudo first order reac-
tions converting percent reaction to half-lives for convenience
of comparison.
Table 1 shows the kinetic results of varying the concentration
of bromide ion† and these data are also seen graphically in Fig.
1. It is quite clear that small amounts of bromide cause a great
increase in the rate of reaction, however, increasing the bromide
concentration further has a marked inhibitory effect. This rate
dependence upon bromide concentration suggests large salt
effects in this non-polar medium. The inclusion of an inert
anion (BF4Ϫ) in addition to 0.1  bromide causes the same
inhibition of the reaction seen with higher levels of bromide
ion. To counter the effect of strongly acidic impurities in some
Effect of halide nucleophilicity
Using 0.10  halide ion, we have found half-lives for the reac-
tion of chloride, bromide and iodide to be 396, 28 and 12 min,
respectively. The ratio of alkyl halide products to alkyl tri-
fluoroacetate increases with increasing halide nucleophilicity
(Table 1). Unlike the chloride or bromide ion, the iodide ion is
susceptible to air oxidation and the iodide reactions are best run
under a nitrogen atmosphere. A number of oxyanions show
behavior similar to the bromide ion and their reactions will be
discussed in a later paper.
Effect of alkene structure on rates
The rates of the addition reactions were essentially identical for
all monoalkyl and 1,2-dialkyl alkenes. In contrast, the rate of
reaction of a 2,2-dialkyl alkene was significantly larger (see
Table 3).
Stereochemistry
Because trifluoroacetic acid is the only proton donor in the
† In this paper, we will not differentiate between associated and
unassociated halide ions.
1518
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
Table 3 Half-lives of various alkenes in 0.1  bromide ion, 20%
trifluoroacetic acid in CH2Cl2
Table 4 Half-lives and product ratios from the reaction of halide ion,
20% trifluoroacetic acid in CH2Cl2 and 3,3-dimethylbut-1-enea
Alkene
t1/2/min
Halide
Conc./
T/ЊC
t1/2/min
% Rearrangement
Oct-1-ene
Oct-2-ene
Oct-3-ene
3,3-Dimethylbut-1-ene
2-Methylhept-1-ene
2-Bromooct-1-ene
29
30
25
18
<0.1
69
None
BrϪ
BrϪ
BrϪ
IϪ
0
20
20
Ϫ15
20
20
Ϫ15
20
5100
18
192
9000
8
137
7000
100 (as ester)
0.1
0.1
1.0
0.1
0.1
1.0
77
77
31
66
67
7
IϪ
IϪ
a Uncorrected for detector response.
Fig. 2 NMR spectrum of the methine proton on 2-deuteriobromo-
cyclohexane
Fig. 3 Logarithm of pseudo first order rate constants against Ho Ϫ
pKa of reaction solution
reaction, it is reasonably easy and inexpensive to add DBr to
alkenes. The addition to cyclohexene produces 2-deuterio-
bromocyclohexane with the deuterium-decoupled NMR spec-
trum shown in Fig. 2.
Cl
Cl –
HCl
The proton on the brominated carbon is a triplet of doublets
indicative of two anti protons (J = 8.9 Hz) and one syn proton
(J = 4.0 Hz). Reported values3d are 8.7 and 3.5 Hz. The DBr
addition therefore occurs primarily with anti stereochemistry.
Deuterium NMR of this product produces a spectrum contain-
ing two peaks assigned to the equatorial (δ 2.38) and axial (δ
1.92) deuterium atoms. The relative areas of these peaks are
5.6:1 indicating that at least 85% of the reaction occurs in an
anti manner. The 5 day reaction period in concentrated bromide
may allow for some racemization.
Cl –
H
H
Cl
Scheme 3
Noting the rate accelerating effects of the added chloride,
Fahey stated that ‘Since chloride salts produce no reaction in
the absence of HCl, both the C᎐H and C᎐Cl bonds must be
formed in the transition state and an AdE3 anti addition mech-
anism must be involved.’5 We believe that this conclusion is
exaggerated.
Rearrangement
In order to probe the sensitivity of this reaction toward steric
hindrance and rearrangement, the reaction was run using an
equimolar mixture of oct-1-ene and 3,3-dimethylbut-1-ene.
Analysis of the products after 20 2% reaction gave the
relative rates of reaction and the extent of rearrangement of
the dimethylbutene. The results are shown in Table 4. It should
be noted that the percentage of rearranged bromide increases
with time as the rearranged (tertiary) trifluoroacetate is slowly
converted to the halide. The percentage of total rearranged
product, however, is independent of time.
We found similar behavior under the conditions of our reac-
tion. As did Fahey,3d we find a positive deuterium isotope effect
(kH/kD = 2.2) indicating proton transfer in the transition state.
The lack of reactivity in the absence of halide ion, coupled with
the high proportion of halide product, suggests that these
anions are involved in the transition states leading to virtually
all of the products. The accelerating effect of the halide ion,
however, merely indicates that it is involved in the slow step but
not necessarily in covalent bond formation. Indeed, the variety
of products formed strongly suggests the intermediacy of a
carbocation resulting from a proton transfer from the trifluoro-
acetic acid molecule.
Somewhat more accurate kinetic data for the disappearance
of oct-1-ene were obtained by referencing it to an internal
decane standard. Plotting the negative logarithm of these
pseudo first order rate constants against the negative acidities
(Ho Ϫ pKa) of the solutions gives a reasonably straight line with
a slope of 1.1 (Fig. 3). The corresponding data for oct-4-ene
produce a line with a slope of 1.2. The independence of
this relationship to bromide concentration suggests that the
Discussion
Fahey has reported that, under some conditions, alkenes may
add the elements of HCl3c or HBr3d by a termolecular reaction.
The rate of the (anti) addition of HCl to cyclohexene depends
on the concentrations of the alkene, HCl and added chloride
ion. He postulated that this product is formed by the nucleo-
philic attack of the chloride ion upon an alkene᎐HCl complex
(Scheme 3).
J. Chem. Soc., Perkin Trans. 2, 1998
1519

View Article Online
bromide ion only affects the solution’s ability to protonate the
alkene. This kinetic dependence upon Ho and alkene supports
the AdE2 mechanism leading to the formation of a cationic
intermediate. Other anions affect the rate similarly.
75% is formed through an initial isomerization to the mono-
deuteriooct-2-ene which adds DBr. Only 3% of the 3-bromo-
octane formation occurs through two octene isomerizations.
The small amount of 4-bromooctane found in these reactions
contains similar amounts of singly, doubly and triply deuter-
ated octyl cations suggesting similar competitive pathways.
Approximately 80% of the 2-octyl trifluoroacetate appears to
come from the simple addition of CF3COOD with the remain-
ing ester going through an alkene isomerization. The latter
path also accounts for the majority of the 3-octyl trifluoro-
acetate. The predominance of this pathway has been described
elsewhere.7 Assuming trifluoroacetate to be the base which
deprotonates the sandwiched cation, it deprotonates the cation
nine times as often as it collapses to ester. At this 75% comple-
tion point in the reaction, the oct-2-ene is found to be 98%
monodeuterated while the oct-3-ene is 30% monodeuterated,
68% dideuterated and 2% trideuterated.
An alternative way of grouping these results is to look at
the percentage of products which derives from initial anion
capture, proton loss or cationic rearrangement of the 2-octyl
cation. This analysis indicates that the initial cation sandwich
captures an anion 65% of the time (95% of this time the anion
is bromide). It loses a proton 32% of the time and only 3% of
the 2-octyl cation undergoes a hydride shift. Although less
accurately calculated because of the smaller amounts of these
products, the partitioning of the 3-octyl cation appears very
similar.
We therefore believe that the role of the halide anion in this
transition state is to ‘solvate’ the developing carbocationic
center antiperiplanar to the protonation site. The insensitivity
of the product ratios to the concentration of added bromide ion
militates against a free carbocation. Thus, the intermediate is
a short-lived carbocation sandwiched between an incoming
halide ion and the newly formed conjugate base, trifluoro-
acetate. This form of ‘ion sandwich’ has been described previ-
ously9 in nucleophilic substitutions and as an ‘encumbered ion’
in additions of trifluoroacetic acid to alkenes.7 In the addition
of DBr to cyclohexene, rapid collapse of this unsymmetrical
cation sandwich can produce the anti addition product but we
believe that the stereoselectivity of this reaction derives, in part,
from the inflexibility of the cyclic system.
The similar reactivities of oct-1-, -2- and -4-ene support the
slow formation of a secondary carbocation in all cases. The 2-
methylhept-1-ene rapidly forms a more stable tertiary carbo-
cation (Scheme 4). While rapid collapse of the 2-octyl cation
–OOCCF3
CF3COOH
Br –
Br –
When the concentration of bromide ion in these reactions
approaches 1 , we find a sharp decrease in the rate, coupled
with a marked increase in the proportion of unrearranged
Markovnikov bromide product. Both‡ of these effects parallel
the decreasing acidity§ of the solution (Fig. 1). We believe that
the high salt concentration makes protonation of the Hammett
indicator more difficult and has a parallel effect on the proton-
ation of the alkenes. The effect of this is to require a more
developed participation of the halide ion in the transition state,
a tighter ion sandwich and a greater yield of the Markovnikov
addition product.
Br
OOCCF3
etc.
Scheme 4
sandwich can produce the octyl bromide, many other routes are
also available to it. Bonding to the trifluoroacetate produces the
ester while proton loss, primarily to the more basic trifluoro-
acetate ion, produces cis- or trans-oct-2-ene. Rearrangement to
the 3-octyl cation produces the corresponding 3-substituted
derivatives as well as oct-2-enes and oct-3-enes. That virtually
all of these products derive from a tight ion sandwich can be
seen from the insensitivity of the product composition to the
concentration of bromide ion in the bulk of the solution.
The relative importance of these different pathways was
evaluated further by using deuterated trifluoroacetic acid and
analyzing the deuterium content of each compound as it passed
through the GC–MS. (We chose to examine the reaction of
oct-1-ene with 0.1  bromide looking at the products formed
when the reaction was 75% complete so that we could identify
the alkene products which eventually add DBr.) In the mass
spectrometer, the 2-bromooctane fragments by the loss of
bromine and the resulting 2-octyl cation is detected as singly
and doubly deuterated ions. The former results from the simple
addition of DBr to the oct-1-ene whereas the doubly deuterated
ion comes from an initial isomerization to monodeuteriooct-2-
ene which adds DBr. After correcting for natural deuterium
abundance, 95% of the 2-bromooctane is calculated to result
from the direct addition process while 5% goes through the
oct-2-ene.
Fahey found10 that 3,3-dimethylbut-1-ene undergoes signifi-
cant rearrangement under the conditions of his cyclohexene
study and he suggested that a cation is being formed in this
particular case because steric hindrance prevents chloride
attack on the acid–alkene complex. In our study, oct-1-ene,
which would be expected to show no such hindrance, also
appears to react via a carbocation. Furthermore, our study
of 3,3-dimethylbut-1-ene shows a rate acceleration by 0.1 
bromide ion similar to that seen with the oct-1-ene. This indi-
cates that nucleophilic involvement is not prevented by the tert-
butyl group of the dimethylbutene. However, as in Fahey’s
study, the major product in this reaction is the rearranged
bromide indicating a cationic intermediate. The rearrangement
of the alkene is not diminished by lowering the temperature,
indicating similar activation parameters for the pathways lead-
ing to both rearranged and unrearranged product. The more
nucleophilic iodide ion reacts faster than the bromide ion but
produces the corresponding ‘ion sandwich’ which suffers the
same fate as in the bromide reaction. The great accelerating
effects of the halide ions guarantee that these anions are stabil-
izing the transition state for proton transfer while the extensive
rearrangement guarantees the intermediacy of a carbocationic
intermediate.
Similarly, the 3-bromooctane gives singly, doubly and triply
deuterated octyl cations in the mass spectrometer. The singly
deuterated product derives from rearrangement of the initial 2-
octyl cation followed by bromide capture. The doubly deuter-
ated octyl fragment requires isomerization of the oct-1-ene to
monodeuteriooct-2-ene which adds DBr. The triply deuterated
octyl bromide comes from two deuteration/de-protonation
steps forming doubly deuterated oct-2- and -3-enes which sub-
sequently add DBr. Our results indicate that 22% of the 3-
bromooctane is formed by the cationic rearrangement while
As with the octenes, an increased concentration of bromide
ion slows the reaction and produces a greater proportion of
‡ The lowest concentrations of bromide lead to a lower yield of alkyl
bromide and a significant deviation from the parallel behavior.
§ We are defining our relative acidities as Ho Ϫ pKa = log [In]/[HIn] and
measuring the absorbance of the Hammett indicator, 2-chloro-4-
nitroaniline (see ref. 8). We have not determined the pKa of this indi-
cator in our solvent system.
1520
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
unrearranged bromide. At the highest bromide concentration,
the disappearance of the dimethylbutene is slower than the dis-
appearance of the dimethylbutene in the absence of added salts
and the unrearranged alkyl bromide represents approximately
70% of the product. Using 1.0  iodide ion, greater than 90% of
the addition product is unrearranged.¶ No esters are seen in
either of these cases. The energy barrier for the methyl shift
in the 2,3,3-trimethyl-2-butyl cation has been experimentally
determined11 to be 3.5 0.1 kcal molϪ1. This translates into a
isomerizations were found in any of these reactions although
no efforts were made to exclude oxygen or peroxides. This result
is rarely found6 in HBr addition reactions and is taken as evi-
dence that no HBr is formed during the reaction. It also under-
scores the value of this reagent for the addition of HBr and
DBr to simple alkenes. The ability to add HI to 3,3-dimethyl-
but-1-ene with minimal rearrangement is extremely rare and
suggests that this reagent may find use in similar situations.
rate constant of 1010 Ϫ1 for a methyl shift at room temperature.
s
Experimental
Our methyl shift to a secondary carbocation would be expected
to be faster. A 90% yield of unrearranged product would make
the iodide capture an order of magnitude faster than the methyl
shift. Using 1.5  iodide ion, we could detect no rearranged
product. These considerations put the rate constant for ion pair
collapse to product in the range of 1012 sϪ1 indicating that the
iodide ion must be present when the carbocation is formed.
Rate constants of this magnitude are considered12 to be border-
line between a preassociation mechanism and a concerted
mechanism.
The oct-1-ene (99.5% purity) used in this experiment was
obtained from Fluka Chemical Co. The other alkenes, tri-
fluoroacetic acid and solvents (HPLC grade) were obtained
from Aldrich Chemical Co. and were used without further
purification. The quaternary ammonium salts were obtained
from Fluka Chemical and/or 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 volumetric
flasks by adding 1 or|| 2 drops of the alkene** to 5 to†† 100 ml
of the 20% trifluoroacetic acid in methylene chloride solution
containing the quaternary ammonium salt. Aliquots (approxi-
mately 0.5 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 purchased or synthesized and analyzed by
NMR spectroscopy. Weight averaged detector responses for the
octyl bromides relative to oct-1-ene were determined by syn-
thesis of a bromide mixture, adding a known amount of oct-1-
ene and analyzing the mixture by GC–MS. Detector responses
for these secondary alkyl bromides were assumed to be equal.
Relative detector responses for the secondary octyl trifluoro-
acetates were determined in a similar fashion. Relative detector
responses for the diverse octyl halides were assumed to be pro-
portional to the responses for the corresponding cyclohexyl
halides which were analyzed in the usual manner. The second-
ary alkyl trifluoroacetates and halides were shown to be stable
under the reaction conditions. Unless otherwise noted, all reac-
tions were run at room temperature (20 2 ЊC) and showed no
evidence of exothermicity.
The additions of HI were accompanied by some iodide sub-
stitution into the methylene chloride solvent. When this inter-
fered with purification of the product, 1,1,1-trichloroethane
became the preferred reaction solvent. Flash chromatography
was often useful in product purification.
NMR spectra were recorded on a JEOL FX90Q spec-
trometer. 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.
While most workers in this area have described the electro-
philic addition of hydrogen halides to alkenes as a competition
between bimolecular and termolecular mechanisms, it appears
that the halide ions are not covalently involved in the transition
states leading to ion pair intermediates. We have considered the
possibility of a competition between an AdE2 cation sandwich
mechanism and an AdE3 unsymmetrical concerted mechanism
but the kinetic independence of bromide ion makes the latter
unlikely. The efficient trapping of a 3,3-dimethyl-2-butyl cation
is remarkable and supports the case for concertedness only in
the most extreme situations.
We believe that our results are best explained by the slow
protonation of the alkene assisted by variable halide partici-
pation. In a non-polar solvent, low concentrations of anions
will assist the protonation of neutral bases (e.g. nitroanilines)
and, similarly, assist the proton transfer to an alkene. The
resulting ion pair undergoes the reactions expected from
cations. Higher concentrations of salts make protonation of
both species more difficult. The weaker proton donating solu-
tion inhibits the protonation thus requiring greater halide ion
participation, more efficient trapping of the intermediate cation
and, perhaps, eventually cause a concerted reaction. This will
lead to increasing proportions of unrearranged alkyl halide
product. This mechanism, with variable nucleophilic partici-
pation, is reminiscent of the suggestions13 for a nucleophilic
substitution mechanism intermediate between the SN1 and SN2
limiting cases.
Since we have skewed the reaction conditions much further
toward non-ionic reaction conditions than any other similar
reaction that we can find, we expect that truly concerted ter-
molecular additions of hydrogen halides to alkenes are extremely
rare. Furthermore, since the iodide ion is an optimal nucleo-
phile in strongly acidic solutions, it seems unlikely that there are
many cases of acid catalyzed additions to alkenes which do not
involve a carbocationic intermediate. We believe that the partial
stereoselectivity found in additions of hydrogen halides to
alkenes derives from the ion sandwich mechanism rather than
from a true concerted addition. The inability of most reaction
conditions to prevent the rearrangement of the dimethylbutyl
cation supports this view.
Tetra-n-butylammonium trifluoroacetate
A 25 ml sample of tetra-n-butylammonium hydroxide (Fluka
Chemical) was neutralized with trifluoroacetic acid, then
cooled and extracted with two 50 ml portions of methylene
chloride which were rotary evaporated to a thick syrup. Appli-
cation of a high vacuum for 24 h produced a thick mass which
was broken up and dried by vacuum for another 24 h at 65 ЊC.
The resulting white powder melted at 81–82 ЊC (Found: C,
It should be noted that oct-1-ene, cyclohexene and 3,3-
dimethylbut-1-ene all appear to react by the same termolecular
cation sandwich mechanism at the lower bromide concen-
trations.
No anti-Markovnikov products and no radical induced
|| Results from 1 or 2 drops of alkene (approximately 10Ϫ4 mol) are iden-
tical within our experimental uncertainty.
¶ Under these conditions, the addition of HI to 3-methylbut-1-ene also
produces greater than 90% unrearranged product. This product ratio
remains unchanged in the reaction for more than ten weeks. The 2,3-
dimethyl-2-iodobutane appears to be somewhat less stable.
** In all kinetic studies, a preanalyzed mixture of alkene and decane was
used.
†† At least 10Ϫ3 mol of bromide ion was used.
J. Chem. Soc., Perkin Trans. 2, 1998
1521

View Article Online
2 (a) M. J. S. Dewar and R. C. Fahey, J. Am. Chem. Soc., 1963, 85,
2245; 2248; (b) M. J. S. Dewar and R. C. Fahey, J. Am. Chem. Soc.,
1963, 85, 3645.
60.26; H, 10.27; N, 3.80. C16H36NO2F3 requires C, 60.81; H,
10.21; N, 3.94%).
3 (a) G. S. Hammond and T. D. Nevitt, J. Am. Chem. Soc., 1954, 76,
4121; (b) G. S. Hammond and C. H. Collins, J. Am. Chem. Soc.,
1960, 82, 4323; (c) R. C. Fahey and R. A. Smith, J. Am. Chem. Soc.,
1964, 86, 5035; (d) R. C. Fahey, C. A. McPherson and R. A. Smith,
J. Am. Chem. Soc., 1974, 96, 4534; (e) D. J. Pasto and J. F. Gadberry,
J. Am. Chem. Soc., 1978, 100, 1469.
4 (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.
5 R. C. Fahey, in Topics in Stereochemistry, ed. E. L. Eliel and N. L.
Allinger, Wiley-Interscience, New York, 1968, vol. 3, pp. 237–342.
6 D. J. Pasto, G. R. Meyer and B. Lepeska, J. Am. Chem. Soc., 1974,
96, 1858.
7 P. E. Peterson and G. Allen, J. Org. Chem., 1962, 27, 1505.
8 K. Yates and H. Wai, J. Am. Chem. Soc., 1964, 86, 5408.
9 F. G. Bordwell, P. F. Wiley and T. G. Mecca, J. Am. Chem. Soc.,
1975, 97, 132.
2-Deuteriobromocyclohexane
A solution of tetra-n-butylammonium bromide (5.51 g, 17.1
mmol, 1.88 equiv.) in 20 ml of methylene chloride under a
nitrogen atmosphere was treated with 4.0 ml (51.5 mmol, 5.6
equiv.) of deuteriotrifluoroacetic acid. After stirring the reac-
tion mixture for 10 min, 0.75 g (9.1 mmol) of cyclohexene was
added. The reaction mixture was protected from light with
aluminum foil and allowed to react at room temperature for
5 days. The entire reaction mixture was transferred to a separ-
atory funnel, washed with 2 × 50 ml water, 50 ml saturated
NaHCO3, 50 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 20 ml brine then
dried over K2CO3, filtered and concentrated under reduced
pressure. Distillation gave 0.93 g (63% yield) of a colorless
1
10 R. C. Fahey and C. A. McPherson, J. Am. Chem. Soc., 1969, 91,
3865.
11 M. Saunders and M. R. Kates, J. Am. Chem. Soc., 1978, 100, 7082.
12 W. P. Jencks, Chem. Soc. Rev., 1982, 10, 345.
liquid. H NMR, IR, GC retention times and mass spectrum
were consistent with authentic material (Aldrich Chemical).
13 T. W. Bentley and P. v. R. Schleyer, J. Am. Chem. Soc., 1981, 103,
5466.
References
1 T. H. Lowry and K. S. Richardson, Mechanism and Theory in
Organic Chemistry, Harper and Row, New York, 1987, pp. 576–577
and references therein; F. A. Carey and R. J. Sundberg, Advanced
Organic Chemistry, Plenum Press, New York, 3rd edn., 1990, vol. B,
p. 170.
Paper 7/03570E
Received 23rd May 1997
Accepted 9th February 1998
1522
J. Chem. Soc., Perkin Trans. 2, 1998