Isotope effect and kinetic studies of the reaction of tertiary alcohols with triphenylphosphine-carbon tetrachloride: Ion pair or concerted?

Article abstract of DOI:10.1016/S0040-4020(00)00277-5

Kinetics of the reactions of 1,2,3-triphenyl-2-propanol (1), 1,2- diphenyl-2-propanol (2) and 3,3,3-trideuterio-1,2-diphenyl-2-propanol (3) with triphenylphosphine-carbon tetrachloride in the temperature range of 25- 78°C in several solvents are investigated. In a non-polar solvent (CCl₄), the reaction of (2) proceeds via intermolecular anti E2 elimination and/or intermolecular SN2 nucleophilic substitution (28% substitution, ratio of 2- alkene/l-alkene=l.06, E/Z>49). In a polar solvent (CH₃CN) reaction proceeds via El and/or S(N)1 (24% substitution, 2-alkene/1-alkene=1.9, E/Z≥6. At equilibrium, the ratio of 2-alkene/1-alkene is equal to 99 with E/Z≥4.21. The primary kinetic isotope effect (k(H)/k(D)) for the elimination pathway in the non-polar solvent is equal to 4.90 and 3.90 at temperatures of 25 and 60δC, respectively. A small secondary β-isotope effect of 1.10 was observed for substitution reaction at both temperatures. Direction of substitution (S(N)2 vs. E2) depends on temperature and polarity of the solvent. The energetics (δS(+), δG(+), δH(+)), the rate orders, and optimization of molecular geometry of intermediates by semiempirical methods (AM1 and CNDO) all agree with intermolecular E2 and S(N)2 mechanisms. New rules for stereoselectivity and Hofmann-Saytzeff eliminations are considered. 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00277-5

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 3611±3617
Isotope Effect and Kinetic Studies of the Reaction of Tertiary
Alcohols with Triphenylphosphine±Carbon Tetrachloride:
Ion Pair or Concerted?
*
Abduol Hossein Dabbagh and Khalil Faghihi
College of Chemistry, Isfahan University of Technology, Isfahan 84156, I.R. Iran
Received 16 December 1999; revised 17 March 2000; accepted 30 March 2000
AbstractÐKinetics of the reactions of 1,2,3-triphenyl-2-propanol (1), 1,2-diphenyl-2-propanol (2) and 3,3,3-trideuterio-1,2-diphenyl-2-
propanol (3) with triphenylphosphine±carbon tetrachloride in the temperature range of 25±788C in several solvents are investigated. In a
non-polar solvent (CCl4), the reaction of (2) proceeds via intermolecular anti E2 elimination and/or intermolecular SN2 nucleophilic
substitution (28% substitution, ratio of 2-alkene/1-alkeneˆ1.06, E/Z$49). In a polar solvent (CH3CN) reaction proceeds via E1 and/or
SN1 (24% substitution, 2-alkene/1-alkeneˆ1.9, E/Z$6. At equilibrium, the ratio of 2-alkene/1-alkene is equal to 99 with E/Z$4.21. The
primary kinetic isotope effect (kH/kD) for the elimination pathway in the non-polar solvent is equal to 4.90 and 3.90 at temperatures of 25 and
608C, respectively. A small secondary b-isotope effect of 1.10 was observed for substitution reaction at both temperatures. Direction of
³
³
³
substitution (SN2 vs. E2) depends on temperature and polarity of the solvent. The energetics (DS , DG , DH ), the rate orders, and
optimization of molecular geometry of intermediates by semiempirical methods (AM1 and CNDO) all agree with intermolecular E2 and
SN2 mechanisms. New rules for stereoselectivity and Hofmann±Saytzeff eliminations are considered. q 2000 Published by Elsevier Science
Ltd.
Introduction
phosphine±carbon tetrachloride constitutes a mild and ef®-
cient method for the conversion of alcohols into the
corresponding alkyl halides.3±6 The mechanism for the
substitution reaction has been tested by several investiga-
tors.4±6 They have demonstrated that in a polar solvent (ace-
tonitrile), the reaction proceeds with racemization but in a
non-polar solvent (carbon tetrachloride), inversion of
con®guration is observed. Optically active (R)-(1)-2-octa-
nol in a polar-solvent (CH3CN) produces 2-chlorooctane
with 78% racemization but in a non-polar solvent (CCl4)
an inversion of con®guration greater than 90% is observed.5
Recent interest in the stereoselective synthesis of ole®ns
prompted a simultaneous investigation of the elimination
reaction and the mechanism of the reactions of two model
compounds 1,2,3-triphenyl-2-propanol (1) and 1,2-di-
phenyl-2-propanol (2) with heterogeneous aluminum
oxide and thorium oxide,1 semi-heterogeneous polystyryl
diphenylphosphine in carbon tetrachloride,2 and in this
work with homogenous triphenylphosphine in carbon tetra-
chloride.
Table 1. Comparison of product distribution of dehydration reaction of 2
(reaction time 9 h, temperature 788C (unless otherwise stated))
The dehydration of alcohols over metal oxides depends not
only upon the steric interaction of the intermediate and/or
transition state on the surface but is also strongly dependent
on the preparation conditions of the catalyst and the reaction
conditions (Table 1).1
Reagent
Solvent Con % % (6) % (7) % (8) 6/7 617/8
a
P±PPh2
P±PPh2
CCl4
CH3CN
22.5b
29b
100 Trace Trace V Lc
±
a
62
25
19
17
49
81
0
38
64
56
V L 1.630
2.30 0.560
1.2
Al2O3±Hd
Al2O3±Bd
2-Hexanol 100e
2-Hexanol 100e
11
16
26
21
19
The 2% cross-linked polystyryl diphenylphosphine in tetra-
chloromethane causes dehydration similar to that by alumi-
num oxide (i.e. anti-E2 elimination) (Table 1).2
0.630
Al2O3±Pured 2-Hexanol 100e
57 0.650 0.750
30
1
78f
100
2.30
4.3
2.30
99
d
ThO2
PhSO3Hg
2-Hexanol
Benzene
Reaction of primary and secondary alcohols with triphenyl-
a See Ref. 2.
b 36% Alkyl halide.
c Very large.
d See Ref. 1.
e 2308C.
Keywords: stereoselective synthesis; semiempirical calculations; reaction
mechanism; kinetic isotope effect.
* Corresponding author. Tel.: 198-31-891-3257; fax: 198-31-891-2350;
e-mail: dabbagh@cc.iut.ac.ir
f 3508C.
g Equilibrium conditions, 708C, 2 h, (Refs. 7,8).
0040±4020/00/$ - see front matter q 2000 Published by Elsevier Science Ltd.
PII: S0040-4020(00)00277-5

3612
A. H. Dabbagh, K. Faghihi / Tetrahedron 56 (2000) 3611±3617
Dabbagh et al. reported that for the reaction of erythro-3-
deuterio-2-butanol with triphenylphosphine±carbon tetra-
chloride the major pathway is substitution with elimination
(100% anti E2, favoring the Saytzeff rule, and deuterium
isotope effect equal to 2) as a minor pathway. Reaction with
these reaction mechanisms and their stereoselectivity, we
have undertaken a semiempirical calculation (AM1 and
CNDO) of the molecular geometry of the intermediate of
the reactions of alcohol 2 with triphenylphosphine. It is the
purpose of the present report to provide the evidence which
serves to establish the fact that the mechanism of the reac-
tion of tertiary alcohols with triphenylphosphine±carbon
tetrachloride in a non-polar solvent (carbon tetrachloride)
is concerted intermolecular anti E2 and intermolecular SN2.
In a polar solvent (acetonitrile), it proceeds via an ion pair
mechanism (e.g. E1 and SN1). One additional aim of the
report is to show that these reactions are indeed stereoselec-
tive.
1,1,1,3,3,3-hexadeuterio-2-methyl-2-propanol
85±90% isobutylene with kH/kD equal to 2.6
produced
The question raised was whether reaction of tertiary alco-
hols with triphenylphosphine in carbon tetrachloride could
be used in a stereoselective synthesis of hindered ole®ns. If
so, what is then the mechanism of elimination for these
reactions? What is the effect of temperature and polarity
of solvent on the stereoselectivity and regioselectivity?
How comparable are the dehydrations of 2 in homogenous
triphenylphosphine in carbon tetrachloride with that of a
heterogeneous system (aluminum oxide and thorium
oxide), or with that of semi-heterogeneous cross-linked
diphenylphosphine polystyrene in carbon tetrachloride?
We believe we have found just the right system which can
shed more light on the mechanism for the dehydration reac-
tions of tertiary alcohols.
Initially the data for the reaction of tertiary alcohols with
triphenylphosphine±carbon tetrachloride in a non-polar
solvent (carbon tetrachloride) resembled that of acid-cata-
lyzed dehydration of alcohols (e.g. E1 and SN1 mechanism)
(Table 2, Scheme 1). However, a closer examination of the
data and comparison with other acid or base-catalyzed dehy-
dration of alcohols encouraged us to re-examine this system.
The reaction of 2 with other acid or base reagents were less
stereoselective and/or produced undesired side products (UP)
in high yield (Table 2). The reaction of 2 with p-toluenesulfo-
nic acid initially produced the kinetically controlled adducts
(6/8ˆ1.13) but quickly isomerized (within 2±3 h) to reach an
equilibrium composition (6/8ˆ30±40) indicating E1 mechan-
ism, Tables 1 and 2. The reaction of 2 with phosphorus
pentachloride showed unusual Saytzeff selectivity in which
the formation of E-alkene 6 competed almost equally with
UP and a minor amount of substitution. Thionyl chloride
produced trace amounts of substitution and mostly UP. An
To answer these questions, we ®rst studied the effects of
varying temperature and the polarity of solvent on reactiv-
ity, stereoselectivity (i.e. the ratio of syn/anti elimination),
and regioselectivity (i.e. the ratio of Saytzeff/Hofmann
elimination). Second, we calculated the kinetics of elimina-
tion and substitution reactions [i.e. the energy of activation
(Ea), enthalpy (DH ) and the entropy of activation (DS )].
Then, we studied the elimination reaction of 3,3,3-tri-
deutero-1,2-diphenyl-2-propanol (3) and obtained a primary
kinetic isotope effect. In order to gain more insight into
³
³
³
Table 2. Conversion, substitution, and elimination adducts for the reaction of 2 with hydrochloric acid, sulfuric acid, phosphorus pentachloride, thionyl
chloride, and triphenyl-phosphine-tetrachloromethane
Reagentsa (temp.)
Time (h)
% Conv.b
% UPc
Relative percent adducts
Ratio of (6/8)
10
6
7
8
A(788C)d
A(788C)
8
7
66
100
65
43
77
100
±
100
100
100
±
0
0
0
0
0
43.5
±
100
±
±
±
28
31.5
48
57
±
6.5
±
Trace
±
±
±
±
±
37
37
30
24.5
56
50
5
Trace
Trace
Trace
Trace
Trace
Trace
2
0
2
5
19
21
9
35
31.5
22
18.5
44
Trace
46
0
46
1.10
1.20
1.40
1.30
1.30
±
1.13
±
1.13
8.5
A(258C)e
547
19
48
0.75
24
1.25
,1
±
HCl(788C)f
H2SO4
g
h
PCl5
POCl3 (458C)i
j
SOCl2
0
B (708C)ik
B (708C)k
C(708C)kl
D(1398C)km
E(1398C)kn
52
85
81
77
89
10
1
2
±
±
±
39.5
32
44.5
±
61
±
±
2
a AˆPh3P±CCl4.
1
b % conversion (by H NMR).
c % UPˆUndesired products.
d Reaction mixture kept at 2158C for three months, then at 788C.
e Reaction mixture kept at 2158C for three months, then at 258C.
f 36% HCl, CDCl3±CCl4.
g Conc., THF, re¯ux.
h THF, 30±458C.
i Ref. 7.
j Re¯ux.
k BˆTsOH-benzene.
l CˆEquilibrium composition in TsOH-benzene.
m DˆEquilibrium composition in t-BuOK-xylene.
n EˆKNH2±NH3.

A. H. Dabbagh, K. Faghihi / Tetrahedron 56 (2000) 3611±3617
3613
Scheme 1.
ion pair mechanism is consistent with these experimental
observations.
several solvents with various dielectric constants in order
to compare the extent of bond-making and bond-breaking in
polar and non-polar solvents (Tables 4 and 5).
Results
Reaction of 1,2,3-triphenyl-2-propanol (1)
Table 4. Conversion of 2 to alkyl halides and alkenes in acetonitrile
Reaction of alcohol 1 with one or two molar ratios of triphe-
nylphosphine in carbon tetrachloride was studied at 788C
over 200 h. The results are summarized in Table 3 (Scheme
1).
Time (h)
Temp. (8C)
% Conv.
10/61718
6/7
617/8
b
5
45
78
45
78
45
78
35
86.5
51
98
66.5
100
1.60
0.113
4.60
Largea
14.5
1.60
b
13
22
Largea
5.10
b
0.020
0.770
Largea
6.25
2.20
Polarity of solvents
c
b
The decomposition reactions of 2 were investigated in
a Trace of Z-alkene.
b No trace of 1-alkene.
c No alkyl halide.
Table 3. Conversion of 1 to alkenes at 788C with triphenylphosphine±
carbon tetrachloride (relative percentages were calculated by 1H NMR;
alkyl halides were not formed)
Table 5. Comparison of dielectric constant of the solvents on the conver-
sion of 2 to alkyl halides and alkenes in 9±10 h
Time (h)
% Conv. (mol)
Ratio of 4/5 (mol)
1a
2
1a
2
Solvent
e [D] Temp. (8C) %Conv. 10/61718
6/7
617/8
Largea 1.20
Largea 1.10
6.30
b
CCl4
2.2
21
36
49
78
45
78
70
61.5
48
90
37.5
0.550
0.430
0.117
2
,2
3
6
13
30
35
7
16
44
45.5
58
83
3.25
2.70
2.57
4.0
4.30
4.30
(CH3)2COb
CH3CN
(CH3)2SO
22
47
71
144
192
5.70
5.70
70
5.70
5.70
c
Largea 1.75
d
a Trace of Z-alkene.
b Time 23 h and 30 min.
c No 1-alkene.
a Moles of triphenylphosphine.
b Trace amounts.
d No alkyl halide.

3614
A. H. Dabbagh, K. Faghihi / Tetrahedron 56 (2000) 3611±3617
Figure 1. Typical plots of (a) second order rate for the disappearance of ROH 2. (b) First order rate of appearance of E-alkene 6. (c) First order rate of
appearance of RCl 10. (d) Arrhenius plot for the appearance of E-alkene 6.
Table 6. Energetics for the rate of disapearance of alcohol 2 and for the rate of formation of alkenes and alkyl halides
³
DS (cal/mol)
³
DG (kcal/mol)
³
DH (kcal/mol)
Compd (#)
Ea (kcal/mol)
ra
ROH (2)b
16.70
14.90
19.81
16.10
0.9960
0.990
0.990
0.990
216.62
221.390
28.59
19.25
20.6840
21.78
16.11
14.3060
19.21
RCl (10)c
E-alkene (6)c
1-alkene(8)c
217.180
20.5740
15.4510
a Correlation coef®cient.
b Second order rate equations are used to calculate the rate constants.
c First order rate equations are used to calculate the rate constants.
Kinetics
energy, and the enthalpy were calculated using the ®rst
order rate constants as in the following expressions,
³
³
log kˆ2Ea/RT1log A, A ˆ ꢀkkT=h†eDS =R
,
DG ˆDH 2
³
The rates of disappearance of alcohol 2 and the rate of
formation of E-alkene 6, 1-alkene 8, and alkyl halide 10
were measured in pure carbon tetrachloride at various
³
³
TDS , DH ˆEa2RT, respectively (Fig. 1). The principal
results are summarized in Table 6. The rate of decomposi-
tion of alcohol 2 was also ®t by the second-order rate equa-
tion. The rate of formation of alkenes 6 and alkyl halide 10
showed good linearity with ®rst order rate equation (Fig. 1).
1
temperatures by H Nuclear Magnetic Resonance Spectro-
scopy (NMR). The rate of intermediate decomposition was
fast and non-measurable, Scheme 1. The rate constants for
the reactions were calculated using the ®rst (Ln a0/
a02xˆkt) and second order rate equations [(1/a)ˆkt1(1/
a0)]. The energy of activation, the entropies, Gibbs free
Isotope effects in the elimination and substitution
reactions of 1,2-diphenyl-2-propanol in carbon
tetrachloride
Table 7. The corrected kinetic isotope effects for elimination and substitu-
tion reactions of 1,2-diphenyl-2-propanol (2) and 3,3,3-trideuterio-1,2-
diphenyl-2-propanol (3) (average of two values (areas from 1H NMR);
for isotope effect corrections see equation 1 in Ref. 1)
The kinetics were run simultaneously on the deuterated (3)
and nondeuterated 1,2-diphenyl-2-propanol (2) so that even
if they did not turn out to be absolute, the relative kinetic
isotope effect would still be valid. The kinetic data are
summarized in Table 7.
Temp. (8C)
Time (h)
(kH/kD)elim
(kH/kD)subst
25
216
312
712
7
23
27
4.90
3.60
3.60
4.10
3.80
3.90
1.20
1.20
1.10
1.10
1.10
1.10
Calculation of kinetic isotope effect
60
The corrected kinetic isotope effects for the elimination and
substitution steps were calculated using equation 1 in Ref. 1.

A. H. Dabbagh, K. Faghihi / Tetrahedron 56 (2000) 3611±3617
3615
Discussion
product distribution in acetonitrile reached equilibrium
composition in 22 h (Tables 5 and 6). Indeed, in the more
polar solvent (acetonitrile), carbon±oxygen cleavage was
suf®ciently extensive to be responsible for the loss of stereo-
selectivity. This indicates the formation of a carbocation
intermediate and that the reaction proceeds via SN1 and
E1 mechanism. In fact, in acetonitrile, carbon±chlorine
cleavage was suf®ciently extensive to be responsible for
the conversion of alkyl halide 10 to alkenes! We have
begun an investigation on the decomposition of 10 and
hope to report the results in due course.
Initially, we attempted to obtain high yields of ole®ns from
some hindered tertiary alcohols 1 and 2, Scheme 1. We
were, however, surprised to ®nd that under our experimental
conditions, we were able to obtain only 5% ole®ns (E/
Zˆ5.70) from 1 in 20 h and 54% ole®n (E/Zˆ5.70) in
312 h, at 788C. At higher temperatures (.1008C, most
part of solvent was evaporated), 100% ole®ns (E/Zˆ0.60)
were produced in 2 h without producing side products (at
room temperature, no ole®n was formed in 451 h). A two-
fold increase in the concentration of triphenylphosphine
increased the rate by about three times, producing 83%
ole®n (E/Zˆ4.30) in 192 h (Table 2). We, therefore, decided
to examine the less hindered alcohol 2. It was known that
under acidic conditions, kinetically controlled a-benzylstyr-
ene (8) was produced in approximately equal amounts (1:1)
with trans-a-methyl stilbene (6) in 0.40 h. The former
alkene was rapidly converted to an equilibrium mixture of
ole®ns (6 and 7) in 2 h (Table 1).7,8 Much to our surprise, we
found that under our experimental conditions (788C)
compound 2 converted 100% to a mixture of 32% 1,2-
diphenyl-2-chloropropane (10), 36% trans-a-methyl stil-
bene (6), and 32% a-benzylstyrene (8). A trace amount of
cis-a-methyl stilbene (7) was formed. The product ratio,
however, remained unchanged in 50 h (2-methyl-2-pro-
panol produced 90% elimination!).
In the light of what has already been said, stereoselectivity
seems to need rede®nition! A reaction should be considered
stereoselective if the product distribution is far from that of
the thermodynamic equilibrium composition. Most acid-
catalyzed reactions are not stereoselective because they
form products near thermodynamic equilibrium composi-
tion. In other words, a high ratio of E/Z or Z/E does not
necessarily mean high stereoselectivity unless this ratio is
far from that of the equilibrium composition.
Another interesting observation is that the reaction of
2-butanol with triphenylphosphine-tetrachloromethane
favored the Saytzeff elimination, but in the case of
compound 2, Hofmann elimination competed equally with
Saytzeff elimination. This indicates that Hofmann±Saytzeff
elimination does not depend on the reagent nor on the stabi-
lity of products but it depends, instead, on steric interactions
in the transition state and/or the intermediates (see also
Refs. 1,2).
In order to gain absolute certainty of the order of the decom-
position reactions, the activation parameters of the decom-
positions, the kinetic isotope effect for the rate of
decomposition of alcohol and the rate of formation of
products, the reactions of 2 and 3 in various solvents and
at different temperatures were investigated. These results
are of utmost importance when evaluating any proposed
decomposition mechanism. Any proposed mechanism
must be consistent with stereochemistry of the reaction,
the kinetics, the energetics, and the isotope effect.
The mechanism is consistent with the energetics of the reac-
tion. The relatively low DH is characteristic of concerted
³
reactions, in which bond-making accompanies bond-break-
ing. In reactions in which the formation of ion pair cleavage
of C±O bond is the rate-determining step, the DH is mark-
³
edly higher with little new bond-making.
The rate of formation of products which is equal to the rate
of decomposition of the intermediate is ®rst order but the
reaction proceeds via intermolecular SN2 substitution or
intermolecular anti E2 elimination. Of course, a true SN2
or E2 reaction cannot be ®rst order unless both substrate and
nucleophile are one molecule.
The mechanism of elimination is consistent with stereo-
chemistry of the reaction. The product distribution, although
initially similar to the acid-catalyzed reaction, is far differ-
ent in that, with triphenylphosphine±carbon tetrachloride
and a non-polar solvent (CCl4), it remained unchanged
over 500 h whereas in the acid-catalyzed reaction the equi-
librium was reached in less than 2 h (Tables 1±6).
³
The entropy of activation (DS ), on the other hand,
decreases because of the loss of translational and rotational
degree of freedom. The two reacting intermediates must
attain a speci®c geometry to permit the bonding intermedi-
ates that occur as the transition state is approached. The
entropy of activation for the 1-alkene formation step is
much higher than that for the E-alkene or alkyl halide
step. This substantiates that chloride ion and methyl group
are molecularly constrained as close neighbors when
compared to E-alkene or alkyl halide formation.
We examined the system using a more polar solvent, aceto-
nitrile, since it has been reported that halogenation occurs
very rapidly in this particular solvent.9] In addition, Slagle et
al. have demonstrated that the increased polarity of aceto-
nitrile over carbon tetrachloride and chloroform is effective
in disturbing the three-dimensional lattice such that the
kinetic order changes from unimolecular to bimolecular.
This allows for suf®cient separation of the ion pair, promot-
ing anion interchange. There is the possibility that the
increased solvent polarity might affect intermediate decom-
position such that carbon±oxygen cleavage might be
considerably enhanced over carbon±chloride bond forma-
tion or elimination of a proton. This was tested by allowing
(R)-(1)-2-octanol to react with triphenylphosphine-tetra-
chloromethane in acetonitrile and it was found that the 2-
chlorooctane obtained was 78% racemic.10 In this work, the
The magnitude of the kinetic isotope effect is a good repre-
sentation of the position of the transition state along the
reaction coordinate. The maximum isotope effect is
expected from a linear symmetrical transition state with
hydrogen positioned centrally between the two-acceptor
centers. The kinetic isotope effect of 3.6±4.9 observed for

3616
A. H. Dabbagh, K. Faghihi / Tetrahedron 56 (2000) 3611±3617
(2-butanol) favored the Saytzeff rule whereas elimination
of the hindered alcohol 2 favored the Hofmann rule.
Experimental
General
1H NMR spectra were measured on a Varian EM390
(90 MHz). IR spectra were obtained using a Shimadzu
ZU-435. Melting points were taken using the Gallenkemp
melting point apparatus and are uncorrected. Yields were
Scheme 2.
1
calculated from H NMR and/or isolated products are not
elimination pathway indicates that the rate-determining step
is the loss of proton. The transition-states 11 and 12 accord
with this observation. Consequently, transition-state 13 may
be rejected (Scheme 2).
optimized. Geometry optimization was performed by semi-
empirical calculations11 (AM1 and CNDO) with the aid of
the HyperChem computer program. All starting materials or
solvents were obtained from Merck or Fluka Chemicals
companies and were puri®ed with proper puri®cation tech-
niques. All reaction temperatures were that of the oil bath
^0.28C. 1,2,3-Triphenyl-2-propanol (1); 1,2-diphenyl-2-
propanol (2) and 3,3,3-trideuterio-1,2-diphenyl-2-propanol
(3) were prepared as reported earlier.1,2,7,8
The very small kinetic b-isotope effect of 1.15^0.05
observed for the substitution pathway (Table 7) indicates
that this process has nearly balanced bond-making and
bond-breaking in what is usually known as an SN2 reaction.6
We would not consider the process to be kinetically
controlled because of the lack of the formation of Z-alkene
(8). How can a reaction be kinetically controlled when
nearly equal amounts of the thermodynamically controlled
isomer (E-alkene) and the least stable kinetically controlled
isomer (1-alkene) are produced but the isomer with inter-
mediate stability (Z-alkene) is formed in trace amounts?
Typical procedure for the following the disappearance of
alcohols and the appearance of alkyl halides and/or
alkenes by H NMR
1
In a 10 ml ¯ask, 0.5 g (2.35 mmol) of 2, 0.6 g (2.29 mmol)
triphenylphosphine, and 6.3 ml (10.01 g, 65.04 mmol) of
carbon tetrachloride were combined in a pressure vessel
(equipped with Rota¯o stopcock), ®lled with 20 g of
3 mm glass beads topped with calcium chloride and kept
in an oil bath at 788C under dry nitrogen. The reaction
progress was monitored by 1H NMR, in which we removed
and tested a small aliquot from each ¯ask at suitable inter-
vals (tetramethylsilane TMS was added as an internal stan-
dard). For alcohol 1, the following chemical shifts (ppm)
were used to calculate the relative percentage of adducts:
two ±CH2± groups dˆ3.15 (AB-quartet, 2H, Jˆ15 Hz),
one ±OH group dˆ1.80 (s, 1H), for E-alkene 4 a ±CH2±
group dˆ4.10 (s. 2H), and for Z-alkene 5 a ±CH2± group
dˆ3.75 (s, 2H). No peak for alkyl halide 9 was observed.
For alcohol 2 the following chemical shifts were used to
calculate the relative percentage of adducts: One ±CH3
group dˆ1.50 (s, 3H), one-OH group dˆ1.55 (s, 1H), or
±CH2± groups dˆ2.90 (AB-quartet, 2H, Jˆ15 Hz); for E-
alkene 6 the CH3 group dˆ2.25 (s, 3H); for 1-alkene 8 a
±CH2± group dˆ3.80 (s, 2H), two ±CH± groups dˆ5.00
and 5.45 (s, 1H); for alkyl halide 10 the CH3 group dˆ1.85
(s, 3H), or dˆ3.30 ±CH2± (s, 2H) groups. No peak for Z-
alkene 7 was observed.
Examination of molecular geometry of the intermediates
(CNDO and AM1 semiempirical calculations)11 clearly
substantiates the back-side attack by the chloride ion from
one intermediate to the vicinity of the methyl group (±CH3),
methylene group (±CH2), and a-carbon (SN2) of another
intermediate. This model was tested for the elimination of
2 with aluminum oxide.1 The most intriguing fact is that the
semiempirical models of the intermediate and/or transition-
state for aluminum oxide also exactly matched the experi-
mental data as well. Formation of cluster(s) of ion pairs for
such hindered molecules would have formidable energetics
as also suggested by Ramos and Rosen.12 Therefore, we
believe that the semiempirical models provide a good repre-
sentation of actual intermediate and/or transition-state.
Attempts are under way to investigate these systems with
more advanced semiempirical methods.
Conclusion
Kinetics (rate constant, energy of activation, enthalpy of
activation, entropy of activation, free energy of activation),
kinetic isotope effect, the stereochemistry, and semiempiri-
cal calculation of the optimized molecular geometry of the
intermediates of the reaction of triphenylphosphine±carbon
tetrachloride in a non-polar solvent (CCl4) all suggest inter-
molecular anti E2 elimination and intermolecular SN2
substitution. They also suggest that the reaction may not
be kinetically controlled. The reaction is indeed stereoselec-
tive in non-polar solvents (high ratio of E/Z) but stereo-
chemistry is lost in polar solvents in which the E/Z ratio is
similar to that of equilibrium composition at higher
temperatures. Elimination of the less hindered alcohol
Typical competitive kinetic isotope effect as determined
by H NMR
1
To a high precision 1H NMR tube, 0.106 g (0.50 mmol) 1,2-
diphenyl-2-propanol (2), 0.262 g (1.00 mmol) of triphenyl-
phosphine, 1 ml (10.32 mmol) carbon tetrachloride and
tetramethylsilane were added. To another high precision
1H NMR tube, 0.107 g (0.50 mmol) 3,3,3-trideuterio-2-
propanol (3), 0.262 g (1.00 mmol) of triphenylphosphine,
1 ml (10.32 mmol) carbon tetrachloride and tetramethyl-
silane were added. After an initial run, the reactions were

A. H. Dabbagh, K. Faghihi / Tetrahedron 56 (2000) 3611±3617
3617
observed for 3 days at 608C (a similar reaction was
completed at 788C in 10 h).
Acknowledgements
This work was supported by Isfahan University of Technol-
ogy Research Council Grant No. IUT-500-1181.
Reaction of 2 with triphenylphosphine±carbon
tetrachloride in acetonitrile
The procedure was similar to carbon tetrachloride using
2.0 g (9.40 mmol) of 2, 2.40 g (9.16 mmol) triphenyl-
phosphine, and 3.50 ml (36.13 mmol) of carbon tetrachlor-
ide in 52 ml (989 mmol) of acetonitrile at 788C (reaction
was repeated at 458C) under dry nitrogen (see Table 4).
References
1. Dabbagh, H. A.; Mohammed Salehi, J. J. Org. Chem. 1998, 63,
7619±7627.
2. Dabbagh, H. A.; Malekpoor, S. S.; Fagihi, K. H. Iranian Polym.
J. 1998, 7, 149±156.
Reaction of 2 with triphenylphosphine±carbon
tetrachloride in dimethyl sulfoxide (DMSO)
3. Downie, I. M.; Holmes, J. B.; Lee, J. B. Chem. Ind. (London)
1966, 65, 900±903.
The procedure was similar to carbon tetrachloride using
2.0 g (9.40 mmol) of 2, 2.40 g (9.16 mmol) triphenyl-
phosphine, and 2.0 ml (20.65 mmol) of carbon tetrachloride
in 13.5 ml (19.00 mmol) of dimethyl sulfoxide at 788C (see
Table 5).
4. Hooz, J.; Giliani, S. S. H. J. Chem. 1968, 46, 86±87.
5. Jones, L. A.; Sumner Jr, C.; Franzus, B.; Huang, T. T. S.;
Snyder, E. I. J. Org. Chem. 1978, 43, 2821±2827.
6. Dabbagh, H. A.; Franzus, B.; Huang, T. T. S.; Davis, B. H.
Tetrahedron 1991, 47, 949±960.
7. Ho, I.; Smith, J. G. Tetrahedron 1970, 26, 4277±4286.
8. Ho, I.; Smith, J. G. Tetrahedron Lett. 1970, 41, 3535±3538.
9. Weiss, R. G.; Snyder, E. I. J. Org. Chem. 1970, 35, 1627.
10. Slagle, J. D.; Huang, T. T. S.; Franzus, B. J. Org. Chem. 1981,
46, 3526.
Reaction of 2 with triphenylphosphine±carbon
tetrachloride in acetone
The procedure was similar to carbon tetrachloride using
2.0 g (9.40 mmol) of 2, 2.40 g (9.16 mmol) triphenyl-
phosphine, and 3.50 ml (36.13 mmol) of carbon tetrachlor-
ide in 20 ml (27.20 mmol) of acetone at 458C under dry
nitrogen (Table 5).
11. Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. P.
J. Am. Chem. Soc. 1985, 107, 3902.
12. Ramos, S.; Rosen, W. J. Org. Chem. 1981, 46, 3530±3533.