Isotope effects in nucleophilic substitution reactions. IV. The effect of changing a substituent at the α carbon on the structure os SN2 transition states

Article abstract of DOI:10.1139/v82-360

Kinetic studies, secondary α-deuterium kinetic isotope effects, primary chlorine kinetic isotope effects (1), Hammett ρ values determined by changing the substituent in the nucleophile, and activation parameters have been used to determine the detailed (relative) structures of the transition states for the S₂ reactions between para-substituted benzyl chlorides and thiophenoxide ion.A rationale for the U-shaped Hammett ρ plots observed when para-substituted benzyl compounds react with negatively charged nucleophiles is also presented.

Full text of DOI:10.1139/v82-360

1067
An unusually large secondary α-deuterium kinetic
isotope effect in hydride ion SN2 reactions
Terry Koerner, Kenneth Charles Westaway, Raymond A. Poirier, and Y. Wang
Abstract: Theoretical calculations at the HF/6–31+G* level suggest the secondary α-deuterium kinetic isotope effects
for hydride ion SN2 reactions are much larger than expected for the structure of the transition state. The secondary
α-deuterium kinetic isotope effects for the SN2 reactions between sodium borohydride (hydride ion) and para-methyl-
and para-chlorobenzyl chlorides are much larger than expected as the theoretical calculations suggest. It appears the
secondary α-deuterium isotope effects are larger than expected for the structure of the SN2 transition state because the
hydride ion is too small to affect the Cα—H(D) out-of-plane bending vibrations in the transition state.
Key words: secondary α-deuterium kinetic isotope effects, SN2, nucleophilic substitution, transition state, substituent
effects.
Résumé : Des calculs théoriques au niveau HF/6–31+G* suggèrent que les effets isotopiques cinétiques secondaires des
atomes deutérium en α dans les réactions SN2 des ions hydrures sont beaucoup plus importants que ceux auxquels on
pourrait s’attendre à partir de la structure de l’état de transition. Les effets isotopiques cinétiques secondaires des
atomes deutérium en α dans les réactions SN2 entre le borohydrure de sodium (ion hydrure) et les chlorures de para-
méthyl- et para-chlorobenzyles sont beaucoup plus élevés que ceux auxquels on pourrait s’attendre sur la base des
calculs théoriques. Il semble que les effets isotopiques cinétiques secondaires des atomes deutérium en α sont beaucoup
plus élevés que ceux auxquels on pourrait s’attendre sur la base de l’état de transition SN2 en raison du fait que l’ion
hydrure est beaucoup trop petit pour affecter les vibrations de déformation hors-plan Cα—H(D) dans l’état de transi-
tion.
Mots clés : effets isotopiques cinétiques secondaires des atomes deutérium en α, SN2, substitution nucléophile, état de
transition, effets de substituant.
Koerner et al. 1072
Introduction
In a theoretical study of the secondary α-deuterium kinetic isotope effects (KIEs) for SN2 reactions, eq. [1], by Poirier et al.,
(1) the KIE was expressed as the
[1]
product of a translational, a rotational, a vibrational, and a tunneling term (eq. [2]).
[2]
(kH/kD)α = (kH/kD)trans (kH/kD)rot (kH/kD)vib (kH/kD)tunnelling
Received November 10, 1999. Published on the NRC Research Press website on July 31, 2000.
T. Koerner and K.C. Westaway.1 Department of Chemistry and Biochemistry, Laurentian University, Sudbury, ON P3E 2C6,
Canada.
R.A. Poirier1 and Y. Wang. Department of Chemistry, Memorial University of Newfoundland, St. John’s, NF A1B 3X7, Canada.
1Authors to whom correspondence may be addressed. KCW: Fax: (705) 675–4844. e-mail: Kwestawa@Nickel.Laurentian.Ca; RAP:
Fax: (709) 737–3702. e-mail: Rpoirier@morgan.ucs.mun.ca
Can. J. Chem. 78: 1067–1072 (2000)
© 2000 NRC Canada

1068
Can. J. Chem. Vol. 78, 2000
Table 1. A comparison of the bending and stretching contributions to the secondary α-deuterium KIE for hydride ion and other
nucleophiles with a similar nucleophile-leaving group distance in the transition state.
Methyl chloride
OH–
Methyl fluoride
F–
H–
% Difference
H–
% Difference
(kH/kD)α
0.907
4.37
0.714
1.112
0.983
4.32
0.707
1.288
8.4
1.1
1.0
0.846
3.69
0.677
0.978
1.013
3.65
0.653
1.389
20.0
1.1
3.5
Nu···LG distance (Å)
(kH/kD)stretching
(kH/kD)bending
15.9
42.0
The vibrational term was factored into three terms (eq. [3]),
[3] (kH/kD)vib
Fig. 1. The relationship between RT.S, RT.S.corr, and RT.S.actual for
the Cα—H(D) bending vibrations in the SN2 transition state.
=
(kH/kD)stretching (kH/kD)bending (kH/kD)other
where (kH/kD)stretching is due to the changes that occur in the
three Cα—H(D)stretching vibrations, (kH/kD)bending is the
contribution due to the changes that occur in the seven
Cα—H(D) bending (primarily the out-of-plane bending)
vibrations, and (kH/kD)other is due to the changes that occur in
the other vibrations, when the reactant is converted into the
SN2 transition state (1). The (kH/kD)stretching and (kH/kD)bending
contributions to the KIE were calculated to determine how
these vibrations affected the magnitude of the secondary
α-deuterium KIE and to determine how the (kH/kD)stretching
and (kH/kD)bending contributions to the KIE were related to the
structure of the SN2 transition state. The initial study
(1) showed that the large and inverse stretching component
of the KIEs was constant, but different, for each leaving
group. For example, the stretching component of the second-
ary α-deuterium KIE for the SN2 reactions between several
nucleophiles and methyl chloride or methyl fluoride, were
(kH/kD)stretching = 0.701 ± 0.014 and 0.677 ± 0.013, respec-
tively. The bending contributions to the KIE on the other
hand, varied markedly with the nucleophile and the
nucleophile-leaving group (Nu···LG) distance (RT.S.) in the
transition state2. For instance, the (kH/kD)bending varied from
1.098 to 1.221 in the methyl chloride reactions and from
0.978 to 1.215 in the methyl fluoride reactions.
pected given the Nu···LG distance in the transition state
(Fig. 3).
The results (Table 1) show that the KIE for the methyl
chloride reaction with hydride ion is 8.4% larger (less in-
verse) than the hydroxide ion reaction that has an almost
identical (4.37 rather than 4.32 Å) Nu···LG distance. In the
methyl fluoride SN2 reactions, the KIE for the hydride ion
reaction is 20% larger than the KIE for the fluoride ion
(identity reaction) even though both reactions have almost
identical (3.65 vs. 3.69 Å) Nu···LG distances in the transi-
tion state. The stretching and bending contributions to the
KIEs for the methyl chloride and methyl fluoride reactions
with hydride ion were calculated in an effort to understand
why the hydride ion reaction KIEs were so large. The
stretching contribution to the methyl chloride and methyl
fluoride reactions with hydride ion are virtually identical to
those found for the reactions with other nucleophiles, i.e.,
the stretching vibration contribution to the KIEs for the H–
and OH– reactions with methyl chloride and for the H– and
F– reactions with methyl fluoride only differ by 1.0 and
Since the (kH/kD)stretching contribution to the KIE is con-
stant for a particular leaving group, the magnitude of both
the total KIE and the bending contribution to the KIE were
related to the Nu···LG distance in the SN2 transition state
(1) (Fig. 2).
Results and discussion
When the calculations described above were extended to
the secondary α-deuterium KIEs for both the methyl chlo-
ride and methyl fluoride SN2 reactions with hydride ion, the
results showed that the hydride ion KIEs were significantly
larger than expected, i.e., the KIE was much larger than ex-
2 Glad and Jensen (2) suggested that the Nu···LG (internuclear) distance (RT.S.) is not the correct parameter to use but that the Nu···LG dis-
tance minus the radius of both the nucleophile and the leaving group atoms (RT.S.corr) should be used for determining the energy of the bend-
ing vibrations (the magnitude of the secondary α-deuterium KIE). We tried correlating the calculated KIE with RT.S.corr before publishing
our first paper (1) but found the relationship between the (kH/kD)α and RT.S.corr was worse than the correlation between (kH/kD)α and RT.S. A
possible explanation for this is that the α-hydrogen only approaches the nucleophile and the leaving group in the Cα—H(D) bending vibra-
tion approximately 0.8 Å above the internuclear axis joining the nucleophile, the α-carbon and the leaving group in the SN2 transition state
(Fig. 1). The curvature of the surfaces of the spherical nucleophile and leaving group nucleophilic atoms means RT.S.actual (the space avail-
able for the Cα—H(D) out-of-plane bending vibrations in the transition state) is greater than RT.S.corr and the higher correlation with RT.S.
suggests RT.S.actual is closer to RT.S. than to RT.S.corr
.
© 2000 NRC Canada

Koerner et al.
1069
Fig. 2. The total KIE (curve a) and the bending vibration contribution to the KIE (curve b) vs. RT.S. for the SN2 reactions of methyl
chloride and methyl fluoride with various nucleophiles. The open circles are for the methyl chloride reactions and the solid circles are
for the methyl fluoride reactions.
3.5%, respectively (Table 1). The major difference between
the KIEs with the almost identical Nu···LG distances in the
transition state, is due to the bending contribution to the
KIE, i.e., the bending contribution to the H– KIE is 15.9%
larger than the bending contribution to the OH– KIE when
the substrate is methyl chloride. The effect is even more
© 2000 NRC Canada

1070
Can. J. Chem. Vol. 78, 2000
Fig. 3. The secondary α-deuterium KIE as a function of the
Nu···LG distance in the transition states for the SN2 reactions of
methyl chloride and methyl fluoride with various nucleophiles.
Fig. 4. A reaction coordinate diagram for an SN2 reaction with
low energy out-of-plane bending vibrations in the transition state.
Fig. 5. The transition state for the hydride ion SN2 reactions
showing that the hydride ion is too small to affect the Cα—H(D)
out-of-plane bending vibrations.
Although the above rationale for the larger secondary α-
deuterium KIEs in the hydride ion reactions seems reason-
able, experimental confirmation of this phenomenon is
needed. The secondary α-deuterium KIE for the SN2 reaction
between sodium borohydride (hydride ion) and para-
methylbenzyl chloride (eq. [4]), was determined to learn
whether the KIE for the hydride ion reaction was, indeed,
larger than expected.
Three labelled and three unlabelled reactions were done
simultaneously. The three rate constants (3) for the unla-
belled (kH) and labelled (kD) substrates were averaged and
divided to obtain the secondary α-deuterium KIE. Each KIE
was determined twice and all the second order (first order in
BH4– and in substrate) kinetic plots were linear to at least
90% of completion and had correlation coefficients ≥0.998.
The average secondary α-deuterium KIE at 30.000 ±
0.002°C (Table 2), is 1.091 ± 0.002 or approximately 1.044/
α-D. The (kH/kD)α per α-deuterium for a carbenium ion SN re-
action of an alkyl chloride where the first step is fully rate-
determining, is 1.11 (4–6). If dissociation of the intimate ion
pair or formation of the free carbocation is rate-determining,
the (kH/kD)α would be approximately 1.15 per α-deuterium
(5). The much smaller secondary α-deuterium KIE of
dramatic for the methyl fluoride reactions. Here, the bending
contribution to the H– KIE is 42.0% larger than the bending
contribution to the F– KIE.
The larger bending vibration contribution to the H– KIEs
obviously occurs because the zero-point energy difference in
the Cα—H(D) out-of-plane bending vibrations is lower than
expected in the SN2 transition state (Fig. 4). The unusually
large (kH/kD)bending contribution to the KIE means the
Cα—H(D) bending vibrations in the transition state are low
energy. One possible explanation for this is that the hydride
ion is too small to affect the Cα—H(D) vibrations in the
transition state (Fig. 5), i.e., the hydride ion is too small to
sterically hinder the Cα—H(D) out-of-plane bending vibra-
tions. This means the bending vibrations are low energy, the
ZPE difference in the transition state is smaller than ex-
pected, and the KIE is larger than expected. It is interesting
to note that the effect of the small hydride ion is greater in
the methyl fluoride reactions which have a shorter Nu···LG
distance in the transition state. Removing the effect of the
hydride in a tighter transition state would be expected to
have a greater effect on the Cα—H(D) out-of-plane bending
vibrations and, therefore, on the KIE.
[4]
© 2000 NRC Canada

Koerner et al.
1071
Table 3. The average rate constants and secondary α-deuterium
KIEs for the sodium borohydride SN2 reactions with p-
chlorobenzyl chloride in DMSO at 30.000 ± 0.002°C.
Table 2. The average rate constants and secondary α-deuterium
KIEs for the sodium borohydride SN2 reactions with p-
methylbenzyl chloride in DMSO at 30.000 ± 0.002°C.
kH × 103
1.735 ± 0.015a
1.698 ± 0.003
kD × 103
1.595 ± 0.004a 1.088 ± 0.010b 1.043
(kH/kD)α
(kH/kD)/α-D
kH × 103
kD × 103
(kH/kD)α
(kH/kD)/α-D
1.806 ± 0.012a 1.655 ± 0.009a 1.091 ± 0.009b 1.045
1.640 ± 0.009
1.553 ± 0.017
Average
1.093 ± 0.012
1.045
1.502 ± 0.002
Average
1.092 ± 0.006
1.092 ± 0.001c 1.045
1.045
1.091 ± 0.002c 1.044
aStandard deviation for three separate kinetic runs.
aStandard deviation for three separate kinetic runs.
bThe error is 1/kD[(∆kH)2 + (kH/kD)2 × (∆kD)2]1/2, where ∆kH and ∆kD are
the standard deviations for the average rate constants for the reactions of
the undeuterated and deuterated nucleophiles, respectively.
bThe error is 1/kD[(∆kH)2 + (kH/kD)2 × (∆kD)2]1/2, where ∆kH and ∆kD are
the standard deviations for the average rate constants for the reactions of
the undeuterated and deuterated nucleophiles, respectively.
cStandard deviation.
cStandard deviation.
1.044/α-D found in this study, confirms that the reaction is an
SN2 process.
ride, would be unusually large as the theoretical calculations
suggest.
The secondary α-deuterium KIE = 1.091 or 1.044/α-D
found for this reaction is near the maximum secondary α-
deuterium KIE = 1.06/α-D found for an SN2 reaction of an
alkyl chloride (7–11). Thus, the hydride ion KIE is large.
However, interpreting the KIE is difficult. Two explanations
are possible. One possibility is that the large secondary α-
deuterium KIE of 1.091 means that the transition state is
loose, and symmetric, with long nucleophile—α- carbon and
α-carbon—leaving group bonds. A large KIE would be ob-
served in this case because the Cα—H(D) out-of-plane bend-
ing vibrations will be low energy in the transition state.
The second possibility is that the hydride ion – p-
methylbenzyl chloride reaction has an unsymmetric transi-
tion state and an unusually large secondary α-deuterium
KIE. Westaway et al. (12) suggested that some benzyl sub-
strates react via an unsymmetric SN2 transition state. In all
the reactions where unsymmetric SN2 transition states have
been characterized in enough detail so that one knows the
relative lengths of the nucleophile—α-carbon and the α- car-
bon—leaving group bonds in the transition state, the stron-
ger reacting bond has always been short and the weaker
reacting bond has been long. Since the H—C bond is much
stronger than the Cα—Cl bond (the stretching frequencies
are 2960 and 650 cm–1, respectively), (13) an unsymmetric
transition state should have a short H—Cα and a long
Cα—Cl bond. If this is the case, one would expect a very
small secondary α-deuterium KIE because the out-of-plane
bending vibrations in the transition state will be determined
by only the out-of-plane bending vibrations to the short re-
acting bond (12). In fact, the secondary α-deuterium KIEs in
reactions with unsymmetric SN2 transition states are all
small because the out-of-plane bending vibrations in the re-
actant or the product and in the transition state are similar
(12). In fact, this behaviour has been found experimentally.
For instance, in the p-substituted benzyl chloride – cyanide
ion reactions and in the cyanide ion – m-chlorobenzyl-p-sub-
stituted benzyl benzenesulphonate reactions which are be-
lieved to have unsymmetric SN2 transition states, the
secondary α-deuterium KIEs are small, i.e., they range from
1.008 to 1.011 and from 1.009 to 1.028, respectively (12,
14). Therefore, if the hydride ion – p-methylbenzyl chloride
transition state is unsymmetric with a short H—Cα bond,
one would expect a small secondary α-deuterium KIE to be
found. The observed KIE of 1.091, which is near the maxi-
mum possible secondary α-deuterium KIE for an alkyl chlo-
The problem then, is to determine whether the KIE is
large because the SN2 transition state is loose but symmetric
or unsymmetric with a short H—Cα bond in the transition
state. Westaway et al. (12) proposed that one could deter-
mine the symmetry of the SN2 transition state if one varied a
substituent on the benzene ring of the substrate. If the transi-
tion state is symmetric, the structure of the transition state
and therefore, the KIE, would change with a change in
substituent. If, on the other hand, the transition state was
unsymmetric with a short H—Cα bond, changing the
substituent on the benzene ring of the substrate would not al-
ter the stronger H—Cα transition state bond significantly
(the Bond Strength Hypothesis (15)) and therefore, the KIE,
which is determined by the short H—Cα bond, will not
change with substituent, i.e., the change in the long Cα—Cl
transition state bond will occur too far away to affect the
Cα—H(D) out-of-plane bending vibrations in the transition
state (the KIE).
The secondary α-deuterium KIE for the SN2 reaction be-
tween p-chlorobenzyl chloride and sodium borohydride was
determined to learn if the SN2 transition state was symmetric
or unsymmetric. The secondary α-deuterium KIE of 1.092 ±
0.001 (Table 3), for the p-chlorobenzyl chloride hydride ion
reaction at 30.000 ± 0.002°C is identical to the KIE = 1.091
± 0.002 for the reaction with p-methylbenzyl chloride.
Therefore, these SN2 transition states must be unsymmetric
with short H—Cα and long Cα—Cl transition state bonds.
Since the H—Cα bond is short in the SN2 transition states
of hydride ion reactions, one would expect the secondary α-
deuterium KIE to be small (vide supra). However, the sec-
ondary α-deuterium KIE is large. The experimental KIEs,
therefore, are larger than expected, presumably because the
hydride ion in the transition state is too small to affect the
Cα—H(D) out-of-plane bending vibrations (Fig. 5). This
means the Cα—H(D) out-of-plane bending vibrations will be
very low energy in the transition state of any hydride ion
SN2 reaction and therefore, the KIE will be much larger than
one would expect given the H—Cα or Nu···LG distance in
the SN2 transition state.
Experimental
The p-methylbenzyl- and p-chlorobenzyl chlorides (Aldrich)
were distilled before use. The preparation of the p-methyl-
and p-chloro-1,1-d2-benzyl chlorides has been described
© 2000 NRC Canada

1072
Can. J. Chem. Vol. 78, 2000
5. V.J. Shiner, Jr., M.W. Rapp, E.A. Halevi, and M. Wolfsberg. J.
Am. Chem. Soc. 90, 7171 (1968).
6. K.C. Westaway. In Isotopes in organic chemistry. Edited by E.
Buncel and C.C. Lee. Elsevier Science Publishing Comp. Inc.,
New York. 1987. p. 312.
previously (9). The description of the kinetic method used to
measure the rate constants is described in ref. (3). The de-
tails of the theoretical calculations are given in ref. (1).
7. K.M. Koshy and R.E. Robertson. J. Am. Chem. Soc. 96, 914
(1974).
Acknowledgement
8. H. Strecker and H. Elias. Chem. Ber. 102, 1270 (1969).
9. K.C. Westaway and Z. Waszczylo. Can. J. Chem. 60, 2500
(1982).
10. K.C. Westaway and Z. Lai. Can. J. Chem. 66, 1263 (1988).
11. K.C. Westaway. In Isotopes in organic chemistry. Edited by E.
Buncel and C.C. Lee. Elsevier Science Publishing Comp. Inc.,
New York. 1987. pp. 313–319.
The authors wish to thank the Natural Sciences and Engi-
neering Research Council of Canada (NSERC) for the fund-
ing that allowed this work to be completed.
References
1. R.A. Poirier, Y. Wang, and K.C. Westaway. J. Am. Chem. Soc.
116, 2526 (1994).
12. K.C. Westaway, T.V. Pham, and Y. Fang. J. Am. Chem. Soc.
119, 3670 (1997).
2. S.S. Glad and F. Jensen. J. Am. Chem. Soc. 119, 227 (1997).
3. K.C. Westaway, T. Koerner, Y.-R. Fang, J. Rudzinski, and P.
Paneth. Anal. Chem. 70, 3548 (1998).
13. J. Dyer. In Application of absorption spectroscopy of organic
compounds. Prentice–Hall, Englewood Cliffs, N.J. 1965.
pp. 25–36.
4. V.J. Shiner, Jr. In Isotope effects in chemical reactions.. A.C.S.
Publication 167. Edited by C.J. Collins and N.S. Bowman. Van
Nostrand Rheinhold, New York. 1970. pp. 90–159.
14. K.C. Westaway, Y.-R. Fang, J. Persson, and O. Matsson. J.
Am. Chem. Soc. 120, 3440 (1998).
15. K.C. Westaway. Can. J. Chem. 71, 2084 (1993).
© 2000 NRC Canada