Using secondary α deuterium kinetic isotope effects to determine the symmetry of S(N)2 transition states

Article abstract of DOI:10.1021/ja962088f

The secondary α deuterium and heavy atom kinetic isotope effects found for two different S(N)2 reactions suggest that the magnitude of secondary α deuterium kinetic isotope effects can be determined by the length of only the shorter (stronger) reacting bond in an unsymmetrical S(N)2 transition state rather than by the usual nucleophile-leaving group distance. Although this means the interpretation of these isotope effects is more complex than has been recognized, the results suggest that they can be used to determine whether an S(N)2 transition state is symmetrical or unsymmetrical.

Full text of DOI:10.1021/ja962088f

3670
J. Am. Chem. Soc. 1997, 119, 3670-3676
Using Secondary R Deuterium Kinetic Isotope Effects To
Determine the Symmetry of SN2 Transition States
Kenneth Charles Westaway,* Thuy Van Pham, and Yao-ren Fang
Contribution from the Department of Chemistry and Biochemistry, Laurentian UniVersity,
Sudbury, Ontario, Canada P3E 2C6
ReceiVed June 21, 1996. ReVised Manuscript ReceiVed January 10, 1997X
Abstract: The secondary R deuterium and heavy atom kinetic isotope effects found for two different SN2 reactions
suggest that the magnitude of secondary R deuterium kinetic isotope effects can be determined by the length of only
the shorter (stronger) reacting bond in an unsymmetrical SN2 transition state rather than by the usual nucleophile-
leaving group distance. Although this means the interpretation of these isotope effects is more complex than has
been recognized, the results suggest that they can be used to determine whether an SN2 transition state is symmetrical
or unsymmetrical.
Introduction
Secondary R deuterium and heavy atom kinetic isotope effects
(KIEs) have been widely used to determine how changing a
substituent on the nucleophile, the substrate, or the leaving group
alters the structure of an SN2 transition state (eq 1).1-5
Recent work has shown that the magnitude of a secondary R
deuterium KIE is determined by the changes that occur in both
the CR-H(D) stretching and out-of-plane bending vibrations
when the reactant is converted into the transition state.6,7
However, a study of the secondary R deuterium KIEs for a series
of SN2 reactions of methyl or ethyl chlorides and fluorides with
several different nucleophiles has shown that the magnitude of
these KIEs for a series of SN2 reactions with the same leaVing
group is determined by the changes that occur in the out-of-
plane bending vibrations when the reactant is converted into
the transition state.7 This means the magnitude of the KIEs
for a series of SN2 reactions with the same leaVing group is
determined by the nucleophile-leaving group distance (tight-
ness) of the transition state7,8 (Figure 1). Although this
relationship has been widely used to relate the magnitude of
secondary R deuterium KIEs to transition state structure,1-5 it
seemed possible that the magnitude of these KIEs might not be
determined by the nucleophile-leaving group distance in a very
unsymmetrical SN2 transition state. For instance, if the SN2
transition state were unsymmetrical (Figure 2), the magnitude
of the secondary R deuterium KIE (the changes in the CR-H(D)
out-of-plane bending vibrations when the reactant is converted
into the transition state) might only depend on the length of the
Figure 1. Relationship between the looseness (the nucleophile-leaving
group distance) of the SN2 transition state and the magnitude of the
secondary R deuterium KIE as determined by the CR-(H)D out-of-
plane bending vibrations.
Figure 2. CR-(H)D out-of-plane bending vibrations for an unsym-
metrical SN2 transition state.
shorter reacting bond rather than on the nucleophile-leaving
group distance.
The secondary R deuterium and heavy atom KIEs found for
two different SN2 reactions (Vide infra) suggest (i) that the
magnitude of some secondary R deuterium KIEs is not
determined by the nucleophile-leaving group distance but only
by the length of the shorter reacting bond in the SN2 transition
state and (ii) that there are two general types of SN2 transition
states. One is a central (nearly symmetrical) transition state
with reasonably short nucleophile-R carbon and R carbon-
leaving group bonds where the magnitude of the KIE is
determined by the nucleophile-leaving group distance. The
second type of SN2 transition state is unsymmetrical with the
* Author to whom correspondence should be addressed.
X Abstract published in AdVance ACS Abstracts, April 1, 1997.
(1) Westaway, K. C.; Ali, S. F. Can. J. Chem. 1979, 57, 1354.
(2) Westaway, K. C.; Waszczylo, Z. Can. J. Chem. 1982, 60, 2500.
(3) Yamataka, H.; Ando, T. J. Am. Chem. Soc. 1979, 101, 266.
(4) Harris, J. M.; Shafer, S. G.; Moffat, J. R.; Becker, A. R. J. Am. Chem.
Soc. 1979, 101, 3295.
(5) Lee, I. Chem. Soc. ReV. 1995, 223.
(6) Wolfe, S.; Kim, C.-K. J. Am. Chem. Soc. 1991, 113, 8056.
(7) Poirier, R. A.; Wang, Y.; Westaway, K. C. J. Am. Chem. Soc. 1994,
116, 2526.
(8) Barnes, J. A.; Williams, I. H. J. Chem. Soc., Chem. Commun. 1993,
1286.
S0002-7863(96)02088-4 CCC: $14.00 © 1997 American Chemical Society

Secondary R Deuterium Kinetic Isotope Effects
J. Am. Chem. Soc., Vol. 119, No. 16, 1997 3671
Table 1. Nitrogen (Leaving Group) KIEs for the SN2 Reactions
between Several Para-Substituted Sodium Thiophenoxides and
Benzyldimethylphenylammonium Nitrate in DMF at 0 °C
Table 2. Secondary R Deuterium KIEs for the SN2 Reactions
between Several Para-Substituted Sodium Thiophenoxides and
Benzyldimethylphenylammonium Nitrate in DMF at 0 °C
para substituent on the
thiophenoxide ion
para substituent on the
thiophenoxide ion
k14/k15
1.01580
1.01536
1.01658
1.01711
(kH/kD)R
CH3O
CH3O
1.238 ( 0.017a
1.219 ( 0.019
1.210 ( 0.012
1.217 ( 0.014
av 1.0162 ( 0.0007a
av 1.221 ( 0.012b
H
1.01599
1.01717
1.01683
1.01623
1.01680
H
1.212 ( 0.015
1.228 ( 0.010
1.210 ( 0.022
av 1.215 ( 0.011
Cl
1.204 ( 0.019
1.229 ( 0.018
1.211 ( 0.014
av 1.0166 ( 0.0004
Cl
1.01625
1.01724
1.01684
1.01707
1.01584
av 1.215 ( 0.013
a The error in each KIE ) 1/kD[(∆kH)2 + (kH/kD)2(∆kD)2]1/2, where
∆kH and ∆kD are the standard deviations for the rate constants for the
undeuterated and deuterated substrates, respectively. In each KIE
experiment, three rate constants for the undeuterated and three for the
deuterated substrates were measured simultaneously to reduce the effect
of any temperature changes during the reaction and or differences
between batches of solvent. b The error is the standard deviation of
the mean.
av 1.0166 ( 0.0005
a The error is the standard deviation of the mean.
strongest reacting bond shorter than the weaker reacting bond.
For these transition states, the magnitude of the secondary R
deuterium KIE is determined by only the shorter (stronger)
reacting bond.
assumes that the nitrogen leaving group KIEs can be interpreted
in the usual fashion, i.e., that the magnitude of the KIE increases
with the percent CR-N bond rupture in the transition state,10
then all three reactions have identical amounts of CR-N bond
rupture in the transition state. This interpretation is reasonable
because the magnitude of a nitrogen leaving group KIE only
depends on the change in the vibrational energy of the CR-N
bond (the amount of CR-N bond rupture11 ) on going from the
reactant to the transition state. If this is the case, then
interpreting the secondary R deuterium KIEs is not straightfor-
ward. In fact, identical secondary R deuterium KIEs could be
found for two different reactions if the transition states were
unsymmetrical and the bond to one of the reacting nucleophiles
in the transition state was very long. In this instance, the
magnitude of the KIE would only be determined by the shorter
reacting bond in the SN2 transition state (Figure 2) because the
other nucleophile in the transition state would be too far away
to affect the CR-H(D) out-of-plane bending vibrations (frequen-
cies). Therefore, when a substituent in the nucleophile is altered,
the change in the magnitude of the KIE would be determined
by only the changes that occur in the shorter reacting bond and
not by the nucleophile-leaving group distance in the SN2
transition state.
Results and Discussion
The first set of KIEs which suggest that the magnitude of a
secondary R deuterium KIE is not always determined by the
nucleophile-leaving group distance in the SN2 transition state
was obtained for the SN2 reactions9 between several para-
substituted thiophenoxide ions and benzyldimethylphenylam-
monium ion in DMF (eq 2).
The nitrogen (leaving group) and the secondary R deuterium
KIEs for these reactions were measured at 0 °C in DMF at an
ionic strength of 0.904. The inert salt, sodium nitrate, was used
to keep the ionic strength constant throughout the reaction so
accurate rate constants (KIEs) could be determined. The
nitrogen KIEs for these reactions (Table 1), are identical. The
secondary R deuterium KIEs found for these reactions (Table
2) are also all the same within experimental error.
The conventional approach to interpreting the identical
nitrogen and secondary R deuterium KIEs would be that the
transition state structure does not change when the substituent
on the nucleophile is altered. This suggestion seems highly
unlikely, however, because (i) no one has observed this behavior
in any study and (ii) it is unreasonable to conclude that a change
in nucleophile, which changes the rate constant by a factor of
6.4, would not alter the energy and, therefore, the structure of
the transition state.
A detailed examination of the nitrogen and secondary R
deuterium KIEs that have been measured for the SN2 reaction
between benzyldimethylphenylammonium ion and thiophenox-
(10) Saunders, W. H. Chem. Scripta 1975, 8, 27.
(11) Some of the vibrational energy lost when the CR-N bond breaks
in the transition state will be partially replaced by the increased vibrational
energy associated with the strengthening of the N-C(phenyl) bond, i.e.,
by conjugation between the dimethylamino group and the benzene ring in
the transition state.1 Although the KIE’s ability to detect a change in the
extent of CR-N bond rupture is reduced slightly by this factor, reactions
with identical KIEs must have identical amounts of CR-N bond rupture in
the transition state. The conjugation effect is thought to be small because
(i) the wavenumber of the absorption of the N-C(phenyl) bond only
increases by approximately 300 cm-1, i.e., from approximately 1000 cm-1
in the reactant to 1300 cm-1 in the product, and (ii) large nitrogen KIEs of
1.0202 have been found in these reactions.1 The largest nitrogen KIE that
has been reported for an SN2 reaction of a quaternary ammonium salt is
1.0226.12
A second, and more likely, possibility is that the change in
nucleophile alters the structure of the transition state but that
the change in transition state structure does not cause a change
in the nitrogen or the secondary R deuterium KIEs. If one
(9) All the reactions were second order, first order in both the substrate
and the nucleophile up to at least 67% of completion. The second order
kinetic plots were linear with correlation coefficients of g0.999. This, the
Hammett F values obtained by changing the substituent on the nucleophile
and the nitrogen (leaving group) KIEs,1 indicates that these reactions are
all SN2 processes.
(12) Smith, P. J.; Pollock, C. A.; Bourns, A. N. Can. J. Chem. 1975, 53,
1319.

3672 J. Am. Chem. Soc., Vol. 119, No. 16, 1997
Westaway et al.
Table 3. Secondary R Deuterium and Primary Nitrogen KIEs for
the SN2 Reaction between Sodium Thiophenoxide and
Benzyldimethylphenylammonium Nitrate at Different Ionic
Strengths in DMF at 0 °C
Table 4. Rate Constants and Hammett F Values for the Reactions
between Para-Substituted Sodium Thiophenoxides and
Benzyldimethylphenylammonium Nitrate at Different Ionic
Strengths in DMF at 0 °C
ionic strength
(kH/kD)R
1.215 ( 0.011a
1.179 ( 0.007
k14/k15
1.0166 ( 0.0004b
1.0200 ( 0.0007
rate constant × 103 (L/(M s))
0.904
0.64
para substituent on
the thiophenoxide ion
ionic strength
ionic strength
µ ) 0.904
µ ) 0.64
a The error in the KIE ) 1/kD[(∆kH)2 + (kH/kD)2(∆kD)2]1/2, where
∆kH and ∆kD are the standard deviations for the rate constants for the
undeuterated and deuterated substrates, respectively. b The error is the
standard deviation of the mean of five different measurements.
CH3O
H
Cl
F
30.1 ( 0.8a
10.9 ( 0.3
4.67 ( 0.07
-1.62 ( 0.01
1.000
46.2 ( 3.3a
12.8 ( 1.0
6.13 ( 0.65
-1.76 ( 0.19
0.994
correlation coefficient
a The error is the standard deviation of the mean of at least three
different measurements.
ide ion (eq 2) at two different ionic strengths shows that the
transition states for the high ionic strength reactions where the
constant KIEs have been observed are, indeed, unsymmetrical.
The identical nitrogen KIEs of 1.0165 found at the high ionic
strength (Table 1) are not large, i.e., they are only approximately
one-third of the theoretical maximum nitrogen KIE of 1.044.13,14
Also, the nitrogen KIE found for the reaction with sodium
thiophenoxide at the high ionic strength (k14/k15 ) 1.0166 (
0.0004) is significantly smaller than the nitrogen KIE of 1.0200
( 0.0007 found1 for the same reaction at an ionic strength of
0.64 (Table 3). Thus, CR-N bond rupture is not well advanced
in the transition states of the high ionic strength reactions. This
suggests that the transition states are reactant-like with reason-
ably short CR-N bonds.
The identical secondary R deuterium KIEs of 1.22 found for
the reactions at high ionic strength, on the other hand, are very
large (approximately 11%/R-D).15 In fact, they are the largest
that have been found for an SN2 reaction of a quaternary
ammonium ion. This indicates that the transition state is very
loose.7 Since the CR-N bond is short and the transition states
are very loose, the S-CR bond must be very long in these
transition states. Thus, the KIE data suggests that the high ionic
strength reactions with constant nitrogen and secondary R
deuterium KIEs have very reactant-like transition states with
short CR-N bonds and very long S-CR bonds.
the secondary R hydrogen-deuterium KIEs are larger in the high
ionic strength reactions, the transition states for these reactions
must have very long S-CR bonds and be unsymmetrical.
The Hammett F value also suggests that the S-CR transition
state bond is very long in the reactions with (kH/kD)R ) 1.22. A
Hammett F ) -1.62 ( 0.01 was observed in the reaction where
(kH/kD)R ) 1.22 whereas a larger Hammett F value of -1.76 (
0.19 was found in the reaction with (kH/kD)R ) 1.179 (Table
4). Since a larger F value is observed when the change in charge
on the sulfur atom on going from the reactants to the transition
state is larger, i.e., when there is more nucleophile-R carbon
bond formation in the transition state, the reaction with the larger
(kH/kD)R must have a longer S-CR transition state bond.22 Thus,
the conclusion based on the F values supports that based on
relative magnitudes of the secondary R deuterium KIEs, i.e.,
that the transition states in the reactions with the identical KIEs
have longer S-CR bonds. Finally, since the reaction at low
ionic strength has a symmetrical transition state,1 the transition
states in the reactions with the constant KIEs must be reactant-
like with very long S-CR and short CR-N bonds.
If one assumes (i) that the transition states for the reactions
with the constant nitrogen and secondary R deuterium KIEs are
reactant-like with very long S-CR and short CR-N bonds, (ii)
that the structure of the transition state is altered by the change
in substituent on the nucleophile, and (iii) that changing the
substituent in the nucleophile does not alter the amount of CR-N
bond rupture in the transition state (the nitrogen KIE) signifi-
cantly (Vide supra), then a change in substituent must change
the length of the S-CR transition state bond. However, since
the transition state is unsymmetrical and very reactant-like, the
changes in the length of the S-CR bond occur too far from the
R carbon to affect the CR-(H)D out-of-plane bending vibrations
in the transition state (the magnitude of the secondary R
deuterium KIE, Figure 2) and identical secondary R deuterium
KIEs26 are found for all three of the SN2 reactions between
benzyldimethylphenylammonium ion and para-substituted
thiophenoxide ions. In this instance, the magnitude of the
secondary R deuterium KIEs in these reactions is only deter-
mined by what happens to the shorter CR-N bond when the
substituent is changed. Since the CR-N bond does not change
Two other pieces of information suggest the S-CR bonds
are long in the high ionic strength transition states. The (kH/
kD)R of 1.215 ( 0.011 found for the thiophenoxide ion reaction
in this study is significantly larger than the (kH/kD)R ) 1.179 (
0.007 found1 for the same reaction at an ionic strength of 0.64
(Table 3). Since the CR-N bond is shorter (Vide supra) and
(13) Maccoll, A. Annu. Rep. A: Chem. Soc. (London) 1974, 71, 77.
(14) Buddenbaum, W. E.; Shiner, V. J., Jr. In Isotope Effects on Enzyme-
Catalyzed Reactions; Cleland, W. W., O’Leary, M. H., Northrop, D. B.,
Eds.; University Park Press: London, 1977; p 18.
(15) The maximum secondary R deuterium KIE for an ammonia leaving
16,17
group is 1.17/R-D
.
This means the minimum KIE expected for a
carbocation SN reaction would be approximately 1.12/R-D
workers suggested that the maximum KIE expected for an SN2 reaction is
.
18 Shiner and co-
19
1.04/R-D
.
The large secondary R deuterium KIEs ) 1.22 (1.105/R-D) found
for these SN2 reactions have been attributed to the relief of steric crowding
around the CR-H(D) bonds in the substrate as the bond to the very sterically
crowded N,N-dimethylaniline leaving group lengthens on going to the
transition state1,20,21 and to the fact that the thiophenoxide ion is too far
away from the R carbon to affect the CR-H(D) out-of-plane bending
vibrations in the transition state.
(16) Hartshorn, S. R.; Shiner, V. J., Jr. J. Am. Chem. Soc. 1972, 94,
9002.
(17) Shiner, V. J., Jr.; Neumann, T. E. Z. Naturforsch. 1989, 44a, 337.
(18) Westaway, K. C. In Secondary Hydrogen-Deuterium Kinetic Isotope
Effects and Transition State Structure in SN2 Processes In Isotopes in
Organic Chemistry; Buncel, E., Lee, C. C., Eds.; Elsevier: New York, 1987;
Vol. 7, p 312.
(22) Hoz et al.23 used theoretical calculations on homolytic cleavage
reactions to question using Hammett F values to determine transition state
structure. However, linear free energy relationships such as F values are
currently used to indicate differences in transition state structure24,25 and
there is no experimental evidence to show which position is correct. The
authors believe that F values from closely related systems can be used to
indicate transition state structure because the changes in the structure of
the SN2 transition state suggested by F values have been confirmed by KIE
data.1,2
(23) Hoz, S.; Basch, S; Goldberg, M. J. J. Am. Chem. Soc. 1992, 114,
4364.
(24) Stefanidis, D.; Jencks, W. P. J. Am. Chem. Soc. 1993, 115, 6045.
(25) Lee, I.; Lee, W. H.; Lee, H. W. J. Phys. Org. Chem. 1990, 3, 545.
(19) Humski, K.; Sendijarevic, V.; Shiner, V. J., Jr. J. Am. Chem. Soc.
1974, 96, 6187.
(20) Westaway, K. C., Tetrahedron Lett. 1975, 4229.
(21) Westaway, K. C. In Secondary Hydrogen-Deuterium Kinetic Isotope
Effects and Transition State Structure in SN2 Processes In Isotopes in
Organic Chemistry; Buncel E., Lee, C. C., Eds.; Elsevier: New York, 1987;
Vol. 7, pp 311-316.

Secondary R Deuterium Kinetic Isotope Effects
J. Am. Chem. Soc., Vol. 119, No. 16, 1997 3673
Table 5. 35Cl/37Cl Leaving Group, the Incoming Group 11C/14C,
the Secondary R Deuterium and the 12C/14C R Carbon KIEs for the
SN2 Reactions between Para-Substituted Benzyl Chlorides and
Cyanide Ion in 20% Aqueous DMSO at 30 °C
when the substituent in the nucleophile is altered, neither the
nitrogen nor the secondary R deuterium KIEs change when the
substituent on the nucleophile is altered.
These conclusions are interesting because they are consistent
with the predictions of the “bond strength hypothesis”27 which
suggests that “there will be a significant change in the weaker
reacting bond but little or no change in the stronger reacting
bond in an SN2 transition state when a substitutent in the
nucleophile, the substrate, or the leaving group is altered in an
SN2 reaction”. Since the C-S bond is weaker than the C-N
bond in these SN2 reactions,27 the “bond strength hypothesis”
would predict that adding an electron-withdrawing substituent
to the nucleophile should not affect the CR-N bond significantly
but should lead to a significant change in the weaker S-CR
bond. These are the exact changes suggested on the basis of
the KIEs.
para
substituent
k35/k37 a
k11/k14
(kH/kD)R
k12/k14 b
CH3
H
Cl
1.0079 ( 0.0004 1.0104 ( 0.0001 1.008 ( 0.003c 1.090
1.0072 ( 0.0003 1.0105 ( 0.002 1.011 ( 0.001 1.102
1.0060 ( 0.0002 1.0070 ( 0.001 1.002 ( 0.003 1.106
a Measured in 20% aqueous dioxane at 30.00 °C.29 b Measured in
20% aqueous dioxane at 40 °C. No error limits were given for these
KIEs.30 c Measured at 30.000 ( 0.002 °C. The error in the KIE )
1/kD[(∆kH)2 + (kH/kD)2(∆kD)2]1/2, where ∆kH and ∆kD are the standard
deviations for the rate constants for the undeuterated and deuterated
substrates, respectively. In each KIE experiment, at least three rate
constants for the undeuterated and three for the deuterated substrates
were measured simultaneously to reduce the effect of any temperature
changes during the reaction and or differences between batches of
solvent.
Finally, another interesting conclusion is that the stronger
CR-N bond is shorter than the weaker S-CR bond in these
reactant-like, unsymmetrical, SN2 transition states.
The same phenomenon, i.e., that the secondary R deuterium
KIE is only determined by the length of the shortest reacting
bond rather than by the nucleophile-leaving group distance in
the SN2 transition state, has been found in a completely different
reaction system. Recently, Matsson, Westaway, and co-
workers28 used 11C/14C incoming group KIEs, Hill and Fry’s
chlorine leaving group KIEs,29 and Fry’s 12C/14C R carbon
KIEs30 to model the transition states for a series of SN2
reactions31 between para-substituted benzyl chlorides and
cyanide ion (eq 3). Both the chlorine29 and the incoming group
11C/14C 28 KIEs (Table 5) decrease when a more electron-
withdrawing substituent is added to the benzene ring of the
substrate. A more detailed analysis of these KIEs28 showed
that the change in the CR-Cl bond (the change in the KIE with
respect to the maximum chlorine KIE) is approximately 6 times
greater than the change in the NtC-CR bond (the change in
the KIE with respect to the maximum 11C/14C incoming group
KIE) when the substitutent on the benzene ring is altered. This
means that adding a more electron- withdrawing substituent to
the benzene ring of the substrate shortens the CR-Cl bond
significantly but has little or no effect on the CR-CtN bond
(Figure 3).
Figure 3. Relative structures for the SN2 transition states for the
reactions of para-substituted benzyl chlorides with cyanide ion.
The secondary R deuterium KIEs for these reactions were
determined in an effort to learn in more detail how the transition
state changes when the para substituent on the benzene ring is
altered. All of these KIEs (Table 5) are small and normal. This
was surprising because (i) the secondary R deuterium KIEs
found for the SN2 reactions of many para-substituted benzyl
substrates, particularly those with electron-donating substituents,
are large and normal32 and (ii) because the KIEs do not decrease
significantly when a more electron-withdrawing substituent is
added to the substrate. In fact, the maximum change in these
KIEs is less than 1%, and the KIE for the unsubstituted benzyl
chloride reaction is greater than that for the p-methylbenzyl
chloride reaction. Normally, the small change observed in the
secondary R deuterium KIE would lead one to conclude that
the nucleophile-leaving group distance in the transition state
does not change significantly when the substituent on the
benzene ring is altered. However, the chlorine KIEs indicate
that the CR-Cl bond shortens significantly and that the NtC-
CR bond shortens slightly when a more electron-withdrawing
substituent is added to the benzene ring.28 Thus, the nucleo-
phile-leaving group distance in these SN2 transition states
decreases when a more electron-withdrawing substituent is
added to the benzene ring of the substrate. If the magnitude of
the secondary R deuterium KIEs were determined by the
nucleophile-leaving group distance in the transition state, the
(26) The magnitude of the secondary R deuterium KIE is determined by
the changes that occur in the both the CR-H(D) stretching and out-of-
plane bending vibrations when the reactant is converted into the transition
state. However, the stretching contribution to the KIE is constant for each
leaVing group regardless of transition state structure.7 Therefore, the
magnitude of the KIE for reactions with the same leaVing group can be
interpreted in terms of the changes in the CR-H(D) out-of-plane bending
vibrations.
(27) Westaway, K. C. Can. J. Chem. 1993, 71, 2084.
(28) Matsson, O.; Persson, J.; Axelsson, B. S.; Långstro¨m, B.; Fang,
Y.-R.; Westaway, K. C. J. Am. Chem. Soc. 1996, 118, 6350.
(29) Hill, J. W.; Fry, A. J. Am. Chem. Soc. 1962, 84, 2763.
(30) Fry, A. In Isotope Effects in Organic Reactions; Collins, C. J.,
Bowman, N. S., Eds.; ACS Publication 167; Van Nostrand Reinhold: New
York, 1970; p 380.
(31) A reviewer suggested that these reactions might shift from an SN2
to a mixed SN2-SN1 mechanism or from a mixed SN2-SN1 mechanism to
an SN1 mechanism when electron-donating groups were on the benzene
ring of the substrate. The second-order kinetic plots for all the substrates
were linear with correlation coefficients of g0.999 up to at least 55% of
completion. Also, a significant shift in mechanism should cause a significant
decrease in the incoming group 11C/14C KIE and a significant increase in
the secondary R deuterium KIE. The kinetics and the almost identical
incoming group 11C/14C and secondary R deuterium KIEs found for these
reactions indicate that all of these reactions are SN2 processes.
(32) Westaway, K. C. In Secondary Hydrogen-Deuterium Kinetic Isotope
Effects and Transition State Structure in SN2 Processes In Isotopes in
Organic Chemistry; Buncel, E., Lee, C. C., Eds.; Elsevier: New York, 1987;
Vol. 7, pp 275-392.

3674 J. Am. Chem. Soc., Vol. 119, No. 16, 1997
Westaway et al.
KIEs should decrease significantly when a more electron-
withdrawing para substituent is added to the substrate.
The only concern with the above interpretation is the large
magnitude of the 11C/14C incoming group KIEs. BEBOVIB-
IV calculations28 suggest that the incoming group KIE should
be inverse (near 0.87) for a product like-transition state with a
short NtC-CR and a long CR-Cl bond. However, only a
product-like transition state with a short NtC-CR and long
CR-Cl bond is consistent with the 35Cl/37Cl, the secondary R
deuterium, the 12C/14C R carbon, and the trend in the 11C/14C
incoming group carbon KIEs found in these reactions. The
explanation for the large magnitude of the 11C/14C incoming
group carbon KIEs found in this system is not obvious.
It is interesting to note that the changes that occur in transition
state structure are consistent with the “bond strength hypoth-
esis”27 even though the change in substituent in these reactions
is in the substrate rather than in the nucleophile as it was in the
first reaction. As expected, the weaker CR-Cl bond changes
significantly and there is little or no change in the stronger
NtC-CR bond when the substituent in the substrate is altered.
It is also worth noting that, again, the weaker CR-Cl reacting
bond is long and the stronger NtC-CR reacting bond is short
in these unsymmetrical transition states. Thus, the behavior is
identical to that observed in the first reaction even though the
change in substituent has been made at different positions.
The secondary R deuterium KIEs reported in this paper
suggest that the magnitude of the KIE for an SN2 reaction can
be determined either (i) by the nucleophile-leaving group
distance in a symmetrical transition state or (ii) by the length
of the shorter (stronger) reacting bond of an unsymmetrical
transition state. A comparison of the secondary R deuterium
KIEs for several SN2 reactions of benzyl substrates1,2,32,35-42
(Table 6) shows that there are two different substituent effects
on the magnitude of secondary R deuterium KIEs. In some
SN2 reactions, these KIEs decrease markedly when a more
electron-withdrawing substitutent is present while in other SN2
reactions they are virtually independent of substituent. On the
basis of the results presented above, it is proposed that the
reactions where the secondary R deuterium KIE varies with
substituent have reasonably tight, symmetrical transition states
while those where the KIE is independent of substituent have
unsymmetrical SN2 transition states where the strongest reacting
bond is short and determines the magnitude of the KIE.
In the SN2 reactions with symmetrical transition states, the
nucleophile-R carbon and R carbon-leaving group transition
state bonds have comparable bond orders, the CR-(H)D out-
of-plane bending vibrations are affected by both the nucleophile
and the leaving group, and the magnitude of the secondary R
deuterium KIE is determined by the nucleophile-leaving group
distance in the SN2 transition state. In these reactions (Table
6), the secondary R deuterium KIEs decrease markedly (by
between 2.8 and 12%) when a more electron-withdrawing
substituent is added to the benzene ring in the substrate2,35-38
or the leaving group.1
The only way to rationalize the almost constant secondary R
deuterium KIEs found for these reactions is to assume that these
benzyl chloride-cyanide ion SN2 transition states are also
unsymmetrical and that only the shorter reacting bond deter-
mines the magnitude of the KIE. A comparison of the chlorine
and the secondary R deuterium KIEs in Table 5 shows that the
magnitudes of the chlorine and the secondary R deuterium KIEs
are clearly not related. The secondary R deuterium and the
incoming group 11C/14C KIEs, on the other hand, change in the
same way with substituent. In fact, although the changes that
are found in the secondary R and the incoming group carbon
KIEs are very small, they parallel each other exactly, i.e., the
largest KIEs are observed when the para substituent on the
benzene ring is hydrogen and the smallest KIEs are found when
the para substituent is chlorine. Obviously, the factors that
affect the secondary R deuterium KIEs also affect the incoming
carbon KIEs. This suggests that these transition states have a
short NtC-CR bond and that the CR-Cl bonds in these
transition states are long, i.e., that the transition states are
unsymmetrical and product-like (Figure 3). As a result, the
changes that occur in the CR-Cl bond when the para substituent
on the benzene ring is altered do not affect the CR-(H)D out-
of-plane bending vibrations significantly and the magnitude of
the secondary R deuterium KIE is only determined by the length
of the shorter NtC-CR transition state bond.
Pearson and Fry30 measured the 12C/14C R carbon KIEs (Table
5) for the SN2 reactions between cyanide ion and three para-
substituted benzyl chlorides at 40 °C in 20% aqueous dioxane,
the solvent that was used to measure the chlorine KIEs for these
reactions. If it is assumed that these KIEs measured in 20%
aqueous dioxane would be similar to those found in the 20%
aqueous DMSO used in this study,28 the 12C/14C KIEs can be
used to model the transition states for these reactions in even
more detail. The R carbon 12C/14C KIE increases as a more
electron-withdrawing group is added to the benzene ring of the
substrate. Since the maximum R carbon KIE is observed when
the SN2 transition state is symmetrical, i.e., when the R carbon
is bonded with equal strength to the nucleophile and the leaving
group in the transition state,33,34 the transition state for the
p-chlorobenzyl chloride reaction is the most symmetrical.
Although the R carbon KIE found for the p-chlorobenzyl
chloride reaction is near the maximum expected for these
KIEs,13,14 these KIEs only indicate that the p-chlorobenzyl
chloride transition state is the most symmetrical and that the
benzyl chloride and p-methylbenzyl chloride transition states
are more unsymmetrical. The chlorine KIEs indicate that the
CR-Cl transition state bond length decreases significantly when
a more electron-withdrawing substituent is present in the
substrate.28 The incoming group carbon KIEs, on the other
hand, suggest there is only a very small decrease in the length
of the NtC-CR bond when a more electron-withdrawing
substituent is on the substrate. The only way a more sym-
metrical transition state can be achieved by adding a more
electron-withdrawing substituent is if these transition states are
product-like, i.e., with short NtC-CR and long CR-Cl bonds.
Then, shortening the CR-Cl bond by adding a more electron-
withdrawing substitutent makes the CR-Cl bond more equal in
strength to the short NtC- - -CR transition state bond (a more
symmetrical transition state) and a larger 12C/14C R carbon KIE
is observed (Figure 3).
In SN2 reactions with an unsymmetric transition state, the
CR-(H)D out-of-plane bending vibrations that determine the
(35) Ashan, M.; Robertson, R. E.; Blandamer, M. J.; Scott, J. M. W.
Can. J. Chem. 1980, 58, 2142.
(36) Shiner, V. J.; Rapp, M. W.; Pinnick, H. R., Jr. J. Am. Chem. Soc.
1970, 92, 232.
(37) Koshy, K. M.; Robertson, R. E. J. Am. Chem. Soc. 1974, 96, 914.
(38) Vitullo, V. P.; Grabowski, J.; Sridharan, S. J. Am. Chem. Soc. 1980,
102, 6463.
(39) Lee, I.; Koh, H. J.; Lee, B.-S.; Sohn, D. S.; Lee, B. C. J. Chem.
Soc., Perkin Trans. 2 1991, 1741.
(40) Ando, T.; Tanabe, H.; Yamataka, H. J. Am. Chem. Soc. 1984, 106,
2084.
(41) Matsson, O.; Persson, J.; Fang, Y.-R.; Westaway, K. C. Unpublished
results.
(33) Sims, L. B.; Fry, A.; Netherton, L. T.; Wilson, J. C.; Reppond, K.
D.; Crook, S. W. J. Am. Chem. Soc. 1972, 94, 1364.
(34) Gray, C. H.; Coward, J. K.; Schowen, B. K.; Schowen, R. L. J.
Am. Chem. Soc. 1979, 101, 4351.
(42) Westaway, K. C.; Koerner, T. Unpublished results.

Secondary R Deuterium Kinetic Isotope Effects
J. Am. Chem. Soc., Vol. 119, No. 16, 1997 3675
Table 6. Some Secondary R Deuterium KIEs for SN2 Reactions with Symmetrical Transition States Where the Magnitude of the KIE Varies
with the Nucleophile-Leaving Group Distance in the Transition State and for Unsymmetrical Transition States Where the Magnitude of the
KIE Is Determined by the Length of Only the Shortest and Strongest Reacting Bond
(kH/kD)R para substituent (Z)
substrate/nucleophile
Z-BzCla/C6H5S-
CH3O
CH3
H
Cl
Br
NO2
∆KIE (%)
ref
SN2 Reactions with Symmetrical Transition States
1.126
1.096
1.056
1.096
1.059
1.124
1.092
1.063
1.024
0.984
1.046
1.061
1.039
8.7
6.3
5.1
12.0
3.1
3.1
2
Z-BzCl/H2O
1.032
1.008
35
35
36
37
38
38
38
Z-BzCl/H2O in 10% aq CH3CN
Z-BzOBs/H2O in 90% aq EtOH
Z-BzCl/H2O
1.004
2-
Z-BzBr + S2O3
1.032
0.996
-
Z-BzBr + N3
2.8
4.4
Z-BzBr + OH-
1.028
1.207
Z/C6H5S–
BzNMe2
1.179
1.151
5.6
1
SN2 Reactions with Unsymmetrical Transition States
BzNMe2C6H5/Z–C6H4S–
1.221
1.215
1.011
1.215
1.002
0.6
this study
this study
Z-BzCl/CN-
1.008
0.99
BzOSO2
BzOSO2
MeOSO2
MeOSO2
EtOSO2
EtOSO2
Z/CH3O
NH2
NH2
NH2
NH2
NH2
NH2
1.096
1.098
0.990
0.971
0.981
0.953
1.102
1.095
0.993
0.974
0.984
0.962
0.6
0.3
0.3
0.3
0.3
0.9
39
39
39
39
39
39
Z/m-O2N
Z/CH3O
Z/m-O2N
Z/CH3O
Z/m-O2N
Secondary r Tritium KIEsb
BzOSO2
m-Br
Cl/Z
NMe2
1.061
1.055
1.042
1.048
1.033
1.9
0.7
40
40
CH2OSO2
Cl/Z
NMe2
1.033
1.026
a
Bz =
CH2
b Secondary R tritium KIEs are much larger and more sensitive to a change in transition state structure than secondary R deuterium KIEs.
magnitude of the secondary R deuterium KIEs are only affected
by the nucleophile in the shortest reacting bond. In all the
reactions with an unsymmetrical transition state found to date,
the stronger reacting bond is short and the weaker reacting bond
is long. Since the “bond strength hypothesis” predicts that little
or no change will occur in the stronger (the shortest) reacting
bond in an SN2 transition state when a substituent on the
nucleophile, the leaving group, or the substrate is altered,27 the
secondary R deuterium KIE will be insensitive to a change in
substituent. In fact, the change in the secondary R deuterium
KIE with substituent (Table 6) is always less than 1% (the
average change is 0.5%) even when the substituent is changed
from strongly electron-donating to strongly electron-withdraw-
ing. Finally, it is worth noting that this behavior has also been
observed for SN2 reactions where the change in substituent is
in the nucleophile39,40 (Vide supra), the leaving group,39,41 or
the substrate42 (Vide supra).
the two types of transition states. While this makes interpreting
secondary R deuterium KIEs more difficult, it appears that the
change in a secondary R deuterium KIE with substituent will
be a good indicator for determining whether an SN2 transition
state is symmetric or unsymmetric.
Conclusions
The KIE data suggest that the magnitude of a secondary R
deuterium KIE is not always determined by the nucleophile-
leaving group distance in an SN2 transition state. In SN2
reactions with unsymmetric transition states, the magnitude of
the secondary R deuterium KIE is determined by the length of
only the shorter (stronger) reacting bond. Finally, it appears
that the change in the secondary R deuterium KIE when a
substituent in the nucleophile, the substrate, or the leaving group
is changed in an SN2 reaction can indicate whether the transition
state is symmetric or unsymmetric.
The data in Table 6 clearly show that there is two different
substituent effects on the secondary R deuterium KIEs of SN2
reactions, and the KIE data in this manuscript suggest that this
is because there are both symmetric and unsymmetric SN2
transition states and that the relationship between the secondary
R deuterium KIE and transition state structure is different for
Experimental Section
Nitrogen KIEs for the SN2 Reactions between Para-Substituted
Sodium Thiophenoxides and Benzyldimethylphenylammonium Ion
in DMF at 0 °C. The procedure used to synthesize the benzyldi-

3676 J. Am. Chem. Soc., Vol. 119, No. 16, 1997
Westaway et al.
methylphenylammonium nitrate and measure the nitrogen KIE has been
described by Westaway and Ali.1 The only difference was that these
KIEs were measured at an ionic strength of 0.90 rather than an ionic
strength of 0.64. The ionic strength was adjusted by adding sodium
nitrate to the solvent.
substrates. Approximately 0.018 g (accurately weighed) of TEACN
was dissolved in 2.50 mL of 20% (v/v) aqueous DMSO (Caledon,
distilled in glass) containing 2.625 × 10-4 mol/L of the internal
standard, p-nitrotoluene (Aldrich), under an extra-dry nitrogen (Praxair)
atmosphere in an IR2 glovebag. After this solution had temperature
equilibrated for at least one hour at 30.000 ( 0.002 °C, the kinetic run
was begun by injecting approximately 0.039 g (0.030 mL) (accurately
weighed) of the appropriate para-substituted benzyl chloride into the
2.50-mL solution. Between 10 and 14 10-µL samples were withdrawn
at various times and injected into an HPLC consisting of a Waters 600
multisolvent delivery system, a 3.9 mm × 15 cm, 3 µm Nova-Pak C18
column, a Waters programmable multiwavelength UV detector, and
an HP 3396A integrator. The mobile phase was 55% aqueous
acetonitrile for the benzyl chloride reaction and 60% aqueous aceto-
nitrile for the reactions of the p-methyl- and the p-chlorobenzyl
chlorides. The flow rate was 1.2 mL/min, and the ultraviolet detector
was set at 260 nm. The concentration of the product, the para-
substituted benzyl cyanide, was calculated from a calibration curve of
the area ratio of the p-substituted benzyl cyanide/internal standard peaks
and the known concentration of p-nitrotoluene.44 Several experiments
showed that the internal standard did not affect the products or the
rate of the reaction. The second-order rate constants were calculated
in the usual way.45
The Secondary r Deuterium KIEs for the SN2 Reactions between
Para-Substituted Sodium Thiophenoxides and Benzyldimethylphen-
ylammonium Ion in DMF at 0 °C. The preparation of the ben-
zyldimethylphenylammonium nitrate and the benzyl-1,1-d2-dimeth-
ylphenylammonium nitrate have been described.1 The procedure used
to measure the second-order rate constants is also described.1 The least-
squares kinetic plots had correlation coefficients of at least 0.999 and
the kinetics were followed to at least 55% of completion. The
secondary R deuterium KIEs were obtained by dividing the average
(from three separate kinetic runs done using the same stock solutions
and at the same time) rate constant for the reaction of the undeuterated
substrate with that for the reaction of the deuterated substrate. The
KIEs measured with different batches of solvent on different days were
identical.
Secondary r Deuterium KIEs for the SN2 Reactions between
Cyanide Ion and Para-Substituted Benzyl Chlorides in 20%
Aqueous DMSO at 30 °C. The undeuterated para-substituted benzyl
chlorides were purchased from Aldrich and purified by treatment with
sodium bisulfite,43 dried, and distilled prior to use. The deuterated para-
substituted benzyl chlorides were synthesized according to the procedure
described by Westaway and Waszczylo.2 The tetraethylammonium
cyanide (TEACN) (Aldrich) was used without further purification. Once
opened, the TEACN was stored in a vacuum desiccator.
Acknowledgment. The authors gratefully acknowledge the
financial support provided by the Natural Sciences and Engi-
neering Research Council of Canada (NSERC).
JA962088F
The secondary R deuterium KIEs were calculated from the second-
order rate constants for the reactions of the undeuterated and deuterated
(44) Snyder, L. R.; Kirkland, J. J. In Introduction to Modern Liquid
Chromatography, 2nd ed.; Wiley-Interscience: New York, 1979; pp 549-
(43) Shriner, R. L.; Fuson, R. C.; Curtin, D. Y. In The Systematic
Identification of Organic Compounds, 4th ed.; J. Wiley and Sons: New
York, 1959; p 149.
555.
(45) Laidler, K. J. In Chemical Kinetics; McGraw-Hill Inc.: New York,
1965; p 8.