The effect of inert salts on the structure of the transition state in the SN2 reaction between thiophenoxide ion and butyl chloride

Article abstract of DOI:10.1021/jo026879d

The effect of inert salts on the structure of the transition state has been determined by measuring the secondary α deuterium and the chlorine leaving group kinetic isotope effects for the S_N2 reaction between n-butyl chloride and thiophenoxide ion in both methanol and DMSO. The smaller secondary α deuterium isotope effects and very slightly larger chlorine isotope effects found in both solvents when the inert salt is present suggests that the S_N2 transition state is tighter and more product-like, with a shorter S - C_α and very a slightly longer C_α - Cl bond when the added salt is present. The salt effect on the reaction in methanol where the reacting nucleophile is the solvent-separated ion-pair complex is much greater than the salt effect on the reaction in DMSO where the reacting nucleophile is the free ion. This greater change in transition-state structure found when the inert salt is present in methanol is consistent with the solvation rule for S_N2 reactions. The greater change in the S - Cα bond is predicted by the bond strength hypothesis. A rationale for the changes found in transition-state structure when the inert salt is present is suggested for both the free-ion and the ion-pair reactions.

Full text of DOI:10.1021/jo026879d

Th e Effect of In er t Sa lts on th e Str u ctu r e of th e Tr a n sition Sta te
in th e SN2 Rea ction betw een Th iop h en oxid e Ion a n d Bu tyl
Ch lor id e
Kenneth Charles Westaway,* Ying Gao, and Yao-ren Fang
Department of Chemistry and Biochemistry, Laurentian University, Sudbury, Ontario P3E 2C6
kwestaway@laurentian.ca
Received December 19, 2002
The effect of inert salts on the structure of the transition state has been determined by measuring
the secondary R deuterium and the chlorine leaving group kinetic isotope effects for the SN2 reaction
between n-butyl chloride and thiophenoxide ion in both methanol and DMSO. The smaller secondary
R deuterium isotope effects and very slightly larger chlorine isotope effects found in both solvents
when the inert salt is present suggests that the SN2 transition state is tighter and more product-
like, with a shorter S-CR and very a slightly longer CR-Cl bond when the added salt is present.
The salt effect on the reaction in methanol where the reacting nucleophile is the solvent-separated
ion-pair complex is much greater than the salt effect on the reaction in DMSO where the reacting
nucleophile is the free ion. This greater change in transition-state structure found when the inert
salt is present in methanol is consistent with the solvation rule for SN2 reactions. The greater
change in the S-CR bond is predicted by the bond strength hypothesis. A rationale for the changes
found in transition-state structure when the inert salt is present is suggested for both the free-ion
and the ion-pair reactions.
TABLE 1. Effect of Ion ic Str en gth on th e Secon d a r y r
Deu ter iu m a n d Nitr ogen Lea vin g Gr ou p Kin etic Isotop e
Effects for th e SN2 Rea ction betw een
Ben zyld im eth ylp h en yla m m on iu m Nitr a te a n d a Sod iu m
Th iop h en oxid e Ion P a ir in DMF a t 0 °C
In tr od u ction
The effect of salts on reaction rates have been inves-
tigated in some detail.1-7 Although the added salts
affected the rate constants for the reaction, they were
not thought to affect the reaction or the structure its
transition state. Recently, however, Pham and West-
away8 reported that the transition state for the SN2
reaction between benzyldimethylphenylammonium ni-
trate and thiophenoxide in DMF at 0 °C, eq 1, was
ionic
strength
(M)
rel transition
state structure
(kH/kD)R
k14/k15
0.90
1.215 ( 0.011a 1.0166 ( 0.0004b
0.64
1.179 ( 0.007 1.0200 ( 0.0007
PhCH2N+(CH3)2Ph NO3- + C6H5S- Na f
PhCH2-SC6H5 + (CH3)2NPh (1)
a
The error in the KIE 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
substrates, respectively.8 Standard deviation on five different
measurements of the KIE.
b
strongly influenced by the addition of inert salts. In
fact, the secondary R deuterium and the nitrogen leav-
ing group kinetic isotope effects (KIEs), Table 1, were
significantly larger and smaller, respectively, when the
ionic strength of the solvent was increased from 0.64 to
0.90 M by adding the inert salt, sodium nitrate, to the
reaction mixture. The marked changes in both KIEs
indicated that the transition state, Table 1, is signifi-
cantly different when the ionic strength is changed in
this reaction. These changes in transition-state structure
were thought to arise because this was a type II SN2
reaction where the incoming nucleophile and the leaving
group have different charges,9 i.e., the thiophenoxide ion
has a negative charge while the leaving group is neutral.
The transition states for type II SN2 reactions have been
found to be very sensitive to a change in solvent.9,10 This
* To whom correspondence should be addressed.
(1) Lowry, T. H.; Richardson, K. S. In Mechanism and Theory in
Organic Chemistry, 3rd ed.; Harper and Row: New York, 1987; pp
345-346.
(2) Dias, P. J .; Gregoriou, G. A. Tetrahedron Lett. 1974, 3827.
(3) J ones, P.; Harrison, R.; Wynne-J ones, L. J . Chem. Soc., Perkin
Trans. 2 1979, 1679.
(4) Alunni, S.; Pero, A.; Reichenbach, G. J . Chem. Soc., Perkin Trans.
2 1998, 1747.
(5) Alunni, S.; Pica, M.; Reichenbach, G. J . Phys. Org. Chem. 2001,
14, 265.
(6) Pregel, M. G.; Buncel, E. J . Am. Chem. Soc., 1993, 115, 10.
(7) Pregel, M. G.; Buncel, E. J . Chem. Soc., Perkin Trans. 2 1991,
307.
(9) Westaway, K. C. Can. J . Chem. 1978, 56, 2691.
(10) Hargreaves, R. T.; Katz, A. M.; Saunders, W. H., J r. J . Am.
Chem. Soc. 1976, 98, 2614.
(8) Pham, T. V.; Westaway, K. C. Can. J . Chem. 1996, 74, 2528.
10.1021/jo026879d CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/12/2003
3084
J . Org. Chem. 2003, 68, 3084-3089

SN2 Reaction between Thiophenoxide Ion and Butyl Chloride
mNa+ + mC6H5S- h [Na+(solvent)n(SC6H5-)]m (4)
presumably occurs because a change in solvent changes
the solvation of the two (a neutral and a negatively
charged) nucleophiles in the reaction differently. This
changes the relative nucleophilicity of the two nucleo-
philes and, therefore, the structure of the transition state
significantly.9 Since adding the inert salt changes the
solvent making it more ionic,8 the structure of the type
II SN2 transition state is altered by the added salt.
This study was undertaken to find out whether a
change in the ionic strength resulting from the addition
of an inert salt would also affect the transition state of a
type I SN2 reaction that has a negatively charged nu-
cleophile and a negatively charged leaving group, eq 2.
Another discovery was that the concentration where
ion pairing begins is different in DMSO and in methanol.
For instance, sodium thiophenoxide exists as a solvent-
separated ion-pair complex at concentrations greater
than 2 × 10-2 M in DMSO and at concentrations greater
than 1 × 10-4 M in methanol, which has a smaller
dielectric constant.13 Since sodium thiophenoxide only
exists as the solvent-separated ion pair complex at
concentrations greater than 2 × 10-2 M in DMSO and
the maximum (initial) sodium thiophenoxide concentra-
tion in the DMSO study was 2.3 × 10-3 M or 10 times
smaller than the concentration where ion pairing begins,
the reacting nucleophile in the DMSO reaction must be
the free thiophenoxide ion. In methanol, on the other
hand, the sodium thiophenoxide concentration was 1.8
× 10-1 M or 3 orders of magnitude greater than the 1 ×
10-4 M concentration where ion pairing begins. As a
result, the reacting nucleophile in methanol is the sodium
thiophenoxide solvent-separated ion-pair complex.
An investigation of the salt effect on a type I SN2
reaction was of interest because these reactions do not
seem to be affected significantly by a change in sol-
vent.9,11,12 This presumably occurs because the change in
solvent changes the solvation of the two reacting nucleo-
philes in the transition state in the same way. As a result,
one would not expect the relative nucleophilicity of the
nucleophile and leaving group or the structure of the
transition state to change significantly with solvent.9 If
the addition of an inert salt can be considered as a change
in solvent as has been suggested,8 one would expect the
added inert salt to have little or no effect on the structure
of a type I SN2 transition state.
The obvious question, however, is whether the inert
salt, sodium nitrate, affects the form of the reacting
nucleophile in DMSO and methanol. This was investi-
gated by adding large amounts (enough to make the
sodium nitrate approximately 1.0 M in DMSO and 0.21
M in methanol14) of sodium nitrate to the 2 × 10-3 and
1.8 × 10-1 M sodium thiophenoxide solutions in DMSO
and in methanol, respectively. In both solvents, the
absorption of the sodium thiophenoxide solvent-separated
ion-pair complex ion was not affected by the addition of
the sodium nitrate. This clearly indicates that the added
salt does not affect the form of the reacting thiophenoxide
ion (the amount of ion pairing). This presumably occurs
because the solvent preferentially solvates the sodium
nitrate rather than the sodium thiophenoxide; i.e., the
solvation energy must be greater for the nitrate ion than
for the thiophenoxide ion in both these solvents. Because
the sodium nitrate does not affect the form of the reacting
thiophenoxide ion significantly, the reacting nucleophile
is the free thiophenoxide ion in DMSO and a solvent-
separated sodium thiophenoxide ion-pair complex in
methanol in both the presence and absence of the inert
salt.
Secon d a r y R Deu ter iu m Kin etic Isotop e Effects.
The secondary R deuterium KIEs for the reactions with
and without the inert salt in DMSO and in methanol are
presented in Table 2. In both solvents, the KIE is smaller
in the reaction with the inert salt, i.e., by 3.8% in DMSO
and by 10.3% in methanol. Since a smaller secondary R
deuterium KIE is found when the SN2 transition state is
tighter, i.e., with a shorter nucleophile-leaving group
distance in the transition state,15,16 the smaller secondary
R deuterium KIEs found in the reactions with the sodium
nitrate indicate that the SN2 transition state is tighter
with shorter S-CR and/or CR-Cl transition-state bonds
when sodium nitrate is present in both DMSO and
methanol.
Resu lts a n d Discu ssion
The effect of adding an inert salt on the transition-
state structure of a type I SN2 reaction was determined
by measuring the secondary R deuterium and the chlo-
rine leaving group KIEs for the SN2 reaction between
n-butyl chloride and sodium thiophenoxide, eq 3, in the
presence and absence of the inert salt, sodium nitrate,
in two solvents, DMSO and methanol. This type I SN2
reaction was chosen for this study because it has been
studied extensively.11-13
Th e F or m of th e Rea ctin g Nu cleop h ile. One dis-
covery resulting from our investigation of the SN2 reac-
tion between n-butyl chloride and thiophenoxide ion was
that the nucleophile, thiophenoxide ion, can react as a
free ion or as a solvent-separated ion-pair complex, eq 4,
in DMSO and in methanol.12,13
(14) These sodium nitrate solutions were almost saturated.
(15) Poirier, R. A.; Wang. Y.; Westaway, K. C. J . Am. Chem. Soc.
1994, 116, 2526.
(16) Barnes, J . A.; Williams, I. H. J . Chem. Soc., Chem. Commun.
1993, 1268.
(11) Graczyk, D. G.; Taylor, J . W.; Turnquist, C. R. J . Am. Chem
Soc. 1978, 100, 7333.
(12) Westaway, K. C.; Lai, Z.-g. Can. J . Chem. 1989, 67, 345.
(13) Westaway K. C.; Lai, Z.-g. Can. J . Chem. 1988, 66, 1263.
J . Org. Chem, Vol. 68, No. 8, 2003 3085

Westaway et al.
TABLE 2. Secon d a r y r Deu ter iu m Kin etic Isotop e
Effects for th e SN2 Rea ction betw een n -Bu tyl Ch lor id e
a n d Sod iu m Th iop h en oxid e in DMSO a n d in Meth a n ol a t
20.000 ( 0.002 °C w ith a n d w ith ou t Ad d ed Sod iu m
Nitr a te
TABLE 3. Ch lor in e (Lea vin g Gr ou p ) KIEs for th e SN2
Rea ction s of Sod iu m Th iop h en oxid e a n d Bu tyl Ch lor id e
in DMSO a n d in Meth a n ol a t 20.000 ( 0.002 °C
ionic
strength
NaNO3 (M)
kH × 102 (L/M‚s)
2.576a ( 0.038b
2.622 ( 0.014
2.321 ( 0.004
kD × 102 (L/M‚s)
(kH/kD)R
δ
average f
Ro/Rf
k35/k37 b
a
DMSO, No Salt Added
DMSO, No Salt Added
2.335a ( 0.033b
1.103 ( 0.023c
1.065 ( 0.007
1.057 ( 0.007
0.954 ( 0.008
0.000
-12.19
-11.59
-11.52
-11.89
-11.94
0.1334
0.1530
0.1667
0.1858
0.2268
0.992655 1.00796
0.993243 1.00740
DMSO + 1.073 M NaNO3
2.462 ( 0.010
0.993312 1.00739
0.992949 1.00788
0.992900 1.00816
MeOH, No Salt Added
average
1.00
1.00776 ( 0.00035c
2.195 ( 0.015
MeOH + 0.2074 M NaNO3
1.983 ( 0.007
DMSO + 1.00 M NaNO3
1.892 ( 0.014
-12.44
-11.68
-11.94
-11.11
-12.04
0.1451
0.1531
0.1792
0.2199
0.2240
0.992409 1.00828
0.003155 1.00750
0.992900 1.00791
0.993715 1.00718
0.992802 1.00825
a
Average of three different rate constants measured on different
days. Standard deviation. c The error in the KIE is 1/kD[(∆kH)2
b
+ (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 substrates, respectively.8
average
0.000
1.00782 ( 0.00048
Methanol, No Salt Added
-12.39
-12.05
-11.78
-11.97
0.1273
0.1459
0.1933
0.2159
0.992212 1.00841
0.992546 1.00814
0.992811 1.00808
0.992624 1.00841
Ch lor in e (Lea vin g Gr ou p ) Kin etic Isotop e Ef-
fects. Although the chlorine KIEs in the reactions with
and without the added salt, Table 3, are identical
considering the experimental errors, the average KIE is
slightly larger when the inert salt is present in both
solvents. Because a larger chlorine (leaving group) KIE
is observed when the amount of CR-Cl bond rupture in
the transition state is greater,17,18 the chlorine KIEs
indicate that the CR-Cl bond is the same or very slightly
longer when sodium nitrate is present in both DMSO and
methanol.
average
0.212
1.00826 ( 0.00018
Methanol + 0.212 M NaNO3
-12.09
-12.25
-12.22
-12.30
-11.72
0.1165
0.1612
0.1898
0.2226
0.2491
0.992507 1.00804
0.992350 1.00843
0.992379 1.00855
0.992301 1.00882
0.992870 1.00832
average
1.00843 ( 0.00029
a
The average Ro from five different samples was 3.13046
relative to a standard methyl chloride sample. The Rf values were
obtained from Rf ) Rr(1 - δ/1000) where Rr is the 35Cl/37Cl ratio
Effect of th e In er t Sa lt (Ion ic Str en gth ) on th e
Str u ctu r e of th e SN2 Tr a n sition Sta te. The smaller
secondary R deuterium KIE found in the presence of the
inert salt in both solvents indicates that the SN2 transi-
tion state is tighter, i.e., has a shorter nucleophile-
leaving group distance when the inert salt is present. The
chlorine leaving group KIEs on the other hand, are
identical or very slightly larger in both solvents when
the inert salt is present. This indicates that the CR-Cl
transition-state bond is the same or very slightly longer
when sodium nitrate is present in both solvents (vide
supra). If the CR-Cl transition-state bond is the same or
very slightly longer and the nucleophile-leaving group
distance is shorter, the SN2 transition state must be
tighter when the inert salt is present in both solvents;
i.e., adding the inert salt must lead to an SN2 transition
state with a much shorter S-CR bond and an equal or
very slightly longer CR-Cl bond whether the reacting
nucleophile is the solvent-separated ion-pair complex in
methanol or the free thiophenoxide ion in DMSO.
It is worth noting that the change in transition-state
structure found when the sodium nitrate is added to the
solvent for both the free-ion reaction in DMSO and the
solvent-separated ion-pair complex reaction in methanol
is predicted by the bond strength hypothesis.19 The bond
strength hypothesis suggests that changing a substituent
in the nucleophile, the substrate, or the leaving group of
an SN2 reaction will cause a significant change in the
b
of the standard methyl chloride sample. The isotope effect is
calculated using the formula k35/k37 ) ln(1 - f)/ln(1 - (Ro/Rf ) f)
where f is the fraction of reaction, Rf is the 35Cl/37Cl ratio in the
chloride ion isolated after the fraction of reaction f, and Ro is the
35Cl/37Cl ratio of the chlorine in the starting material. c Standard
deviation.
weaker reacting bond and little or no change in the
stronger reacting bond in the SN2 transition state.
Although adding the sodium nitrate to the solvent is not
a change in substituent, it leads to the predicted change
in transition-state structure, i.e., the weaker S-CR bond
shortens significantly while there is only a small or zero
change in the stronger CR-Cl bond. Thus, the results
demonstrate that the strength of the reacting bonds plays
a significant role in predicting how the bonds to the
nucleophile and the leaving group will be altered in these
SN2 reactions. Unfortunately, the bond strength hypoth-
esis does not indicate how each reacting bond will change
in a reaction.
Although the changes in transition-state structure on
the addition of the inert salt in both the ion-pair and free-
ion reactions are in accordance with the bond strength
hypothesis and are also in the same direction, i.e., a
tighter, more product-like, transition state is found when
the sodium nitrate is present, they are of a different
magnitude. The changes found in the secondary R deu-
terium and the chlorine KIEs when sodium nitrate is
added to the ion-pair reaction in methanol are ap-
proximately three times larger than those found in the
free-ion reaction in DMSO; i.e., the secondary R deute-
rium KIE decreases by 10.3% in the ion-pair reaction in
(17) Saunders, W. H., J r. Chem. Scr. 1975, 8, 27.
(18) Shiner, V. J ., J r.; Wilgis, F. P. In Isotopes in Organic Chemistry;
Buncel, E., Saunders, W. H., J r., Eds.; Elsevier: New York, 1992; Vol.
8, pp 239-335.
(19) Westaway, K. C. Can. J . Chem. 1993, 71, 2084.
3086 J . Org. Chem., Vol. 68, No. 8, 2003

SN2 Reaction between Thiophenoxide Ion and Butyl Chloride
methanol whereas it only decreases by 3.8% in the free-
ion reaction in DMSO. The change in the chlorine KIE
upon addition of the sodium nitrate is also three times
larger in methanol; i.e., it increases by 0.00017 in
methanol whereas it only increases by 0.00006 in DMSO.
This much larger salt effect on the ion-pair reaction
in methanol was expected. Several studies have shown
that the structure of a type I SN2 transition state is not
affected significantly by a change in solvent. For instance,
Graczyk et al.11 found identical chlorine leaving group
KIEs (the average KIE was 1.00944 ( 0.00012) for the
SN2 reaction between thiophenoxide ion and n-butyl
chloride in several solvents ranging from the dipolar
aprotic solvents DMSO and DMF, to the aprotic solvents
tetraglyme and diglyme, to the protic solvents propanol
and methanol. It is worth noting that the rate constant
for the reaction changed by over 3 orders of magnitude
over this range of solvents. In another study, Westaway
and Lai12 found that the secondary R deuterium KIE was
1.125 ( 0.008 for this type I SN2 reaction when the
solvent was DMSO, DMF, or methanol and the free
thiophenoxide ion was the reacting nucleophile. Since
neither the R carbon-chlorine nor the total nucleophile-
leaving group distance (the magnitude of the secondary
R deuterium KIE) change when the solvent is altered and
the rate constant changes drastically, the structure of the
transition state of this type I SN2 reaction is not altered
by a change in solvent.
Very different results were found for the same reaction,
however, when the solvent was changed in the reactions
where the nucleophile was the solvent-separated ion-pair
complex. Here, the secondary R deuterium KIEs for the
thiophenoxide ion and n-butyl chloride SN2 reaction
changed from 1.034 to 1.084 to 1.110 to 1.141 when the
solvent was changed from DMSO to DMF to methanol
to diglyme. It is worth noting that the rate constants for
the ion-pair complex reactions also changed by 3 orders
of magnitude when the solvent was changed. These
results demonstrate that the ion-pair reaction is very
susceptible to a change in solvent.
The greater sensitivity to the added salt found in the
ion-pair complex reaction can be rationalized by an
extension of the solvation rule for SN2 reactions12 if one
assumes adding the inert salt only changes the solvent
by increasing its dielectric constant. The SN2 transition
states for the free-ion and ion-pair complex reactions,
Figure 1, are very different from the point of view of
solvation. The SN2 transition state for the free-ion
reaction will be mainly, if not exclusively, solvated at the
two negatively charged nucleophiles. As a result, the
relative nucleophilicity of the two nucleophiles will not
be changed significantly by a change in solvent and the
transition-state structure and KIEs will not be altered.9
In the ion-pair complex reactions where the nucleophile
is a tight solvent-separated ion pair,12 Figure 1b, the
solvation will be mainly at the sodium ion, which has a
full positive charge and blocks the normal solvation of
the partially charged sulfur anion, and the developing
chloride ion which has a partial negative charge. As a
result, changing from the protic solvent, methanol, to the
dipolar aprotic solvent, DMSO, changes the solvation
from hydrogen bonding to an ion-dipole interaction at
the negatively charged chloride ion. Since hydrogen
bonding lowers the electron density of an anion, the loss
F IGURE 1. SN2 transition states for the reaction between
n-butyl chloride and (a) the free thiophenoxide ion and (b) the
solvent-separated sodium thiophenoxide ion-pair complex.
of hydrogen bonding to the developing chloride ion will
increase the electron density (the nucleophilicity) of the
chloride ion. The methanol molecule(s) in a tight solvent-
separated ion-pair complex cannot hydrogen bond to the
sulfur anion in the SN2 transition state because the
distance between the sodium and sulfur ions is too small
to form linear hydrogen bonds to methanol. Thus, no
hydrogen bonding is lost at sulfur when the solvent
changes, i.e., the change in solvent from methanol to
DMSO only changes one dipole-ion interaction for
another at the sulfur ion. As a result, the electron density
(the nucleophilicity) of the sulfur anion is not markedly
affected by the change in solvent. Since the solvent
change alters the electron density on the chloride anion
much more than that on the sulfur ion, the relative
nucleophilicity of the chloride and sulfur anions is altered
significantly by a change in solvent (adding the inert
salt). As a result, the ion-pair SN2 transition state
behaves like a type II system,9 and the transition-state
structure and the secondary R-deuterium KIE is altered
markedly by a change in solvent. It is worth noting that
this effect has been reported previously.12 However, the
important observation is that the changes in the KIEs
(transition-state structure) caused by adding the inert
salt to the free-ion reaction in DMSO were significantly
smaller than (approximately one-third) those found when
the salt was added to the ion-pair complex reaction in
methanol.
Th e Role of th e In er t Sa lt. Although the greater
change in the S-CR bond and in transition-state struc-
ture on the addition of the sodium nitrate in the ion-pair
reaction has been rationalized, it is important to try to
understand why adding an inert salt leads to a tighter,
more product-like SN2 transition state in both the ion-
pair and the free-ion reactions. One possibility is that
the sodium nitrate simply changes the solvent by in-
creasing its dielectric constant. In Westaway and Lai’s
earlier study of the n-butyl chloride-sodium thiophe-
noxide SN2 reaction where the nucleophile was the ion-
pair complex,12 a smaller secondary R deuterium KIE and
tighter transition state was found in the solvent with the
J . Org. Chem, Vol. 68, No. 8, 2003 3087

Westaway et al.
This is precisely the effect that is observed in both the
free-ion and ion-pair reactions when the inert salt is
added, i.e., the stronger CR-Cl bond lengthens very
slightly while the weaker S-CR bond shortens signifi-
cantly.
F IGURE 2. Interaction of the cation in the added salt with
the developing chloride ion leaving group in the SN2 reaction
between n-butyl chloride and thiophenoxide ion in DMSO.
largest dielectric constant (vide supra). If one assumes
that the added salt simply increases the dielectric
constant of the solvent, the transition state for the ion-
pair reaction in methanol should be tighter when the salt
is present. This is what is observed in methanol where
the reacting nucleophile is the ion pair, i.e., the very
slightly larger chlorine leaving group KIE and markedly
smaller secondary R deuterium KIE indicate the transi-
tion state is tighter, and more product-like, when the salt
is present. This means that the tighter transition state
is found in the reaction with the solvent with the higher
dielectric constant as Westaway and Lai observed.
While this explanation is acceptable for the ion-pair
reaction in methanol, it is important to note that a
change in the solvent does not explain why the salt affects
the structure of the transition state in the free-ion
reaction. The identical secondary R deuterium and chlo-
rine leaving group effects found in several very different
solvents by Westaway and Lai and by Grazcyk et al. (vide
supra) for this reaction indicate that changing the solvent
does not alter the transition-state structure of the type I
SN2 reaction between n-butyl chloride and the free
thiophenoxide ion significantly. This means the inert salt
must play another role in the free-ion reaction.
One possibility is that the inert salt provides electro-
philic catalysis for the reaction by bonding to the partial
negative charge on the developing chloride ion in the SN2
transition state, Figure 2. It is worth noting that this type
of catalysis to a chloride ion leaving group using hydrogen
bonding to protic solvents in SN reactions has been
suggested by Thornton and Gajewski.20,21 Electophilic
catalysis for the removal of a chloride ion leaving group
by metal ions in E2 elimination reactions of alkyl
chlorides was suggested by Smith et al.22 and by Zavada
and co-workers.23,24 If this type of catalysis occurred, it
would stabilize the developing chloride ion making it a
better leaving group. A tighter, more product-like, transi-
tion state would be expected because Westaway and Ali25
found that changing to a better leaving group in the SN2
reactions between benzyldimethyl-para-substituted phe-
nylammonium ions and thiophenoxide ion in DMF, eq 5,
led to a change in transition-state structure where the
length of the stronger CR-N bond to the leaving group
increased slightly while the weaker S-CR nucleophile-
alpha carbon transition-state bond shortened markedly.
This action of the sodium nitrate would, therefore,
account for the change in transition-state structure found
in the free-ion reaction. It would presumably also have
an effect on the ion-pair reaction as well. Thus, the free-
ion reaction would not be affected by the change in the
solvent caused by adding the inert salt, but only by the
electrophilic catalysis by the added sodium nitrate. The
ion-pair reaction, on the other hand, is presumably
affected by both the change in solvent and by the
electrophilic catalysis of the sodium nitrate.
Con clu sion s
The important discovery is that adding an inert salt
to an SN reaction affects the structure of the SN2
transition state whether the reaction is a type I or a type
II SN2 reaction. The results also show that the transition
state is tighter, and more product-like, with the same or
very slightly longer R carbon-leaving group (CR-Cl) bond
and a much shorter nucleophile-R carbon (S-CR) bond
when the inert salt is present whether the nucleophile
is the free ion or a solvent-separated ion-pair complex.
The much greater salt effect on transition-state structure
found when the nucleophile is a solvent-separated ion-
pair complex rather than a free ion can be rationalized
by the solvation rule for SN2 reactions.12 The major
change in transition-state structure is at the weaker
reacting bond in the SN2 transition states as the bond
strength hypothesis19 suggests. It is suggested that the
change in transition-state structure caused by the added
salt is due to electrophilic catalysis by the cation of the
added salt to the developing anion of the leaving group
in a type I SN2 reaction where the nucleophile is a free
ion and is probably due to a change in the ionizing power
of the solvent a n d the electrophilic catalysis to the
leaving group in a type I SN2 reaction where the nucleo-
phile is a solvent-separated ion-pair complex.
Exp er im en ta l Section
P r ep a r a tion of Rea gen ts. The synthesis of butyl-1,1-d2
chloride and sodium thiophenoxide have been described previ-
ously.13 NMR analysis indicated that the deuterated substrate
was at least 99% deuterated at the R carbon. The sodium
nitrate was dried at 110 °C overnight and then stored in a
vacuum desiccator until it was used. The anhydrous dimethyl
sulfoxide was used as purchased. The reagent-grade methanol
was distilled before use.
Kin etics. In An h yd r ou s Dim eth yl Su lfoxid e. Extra dry
nitrogen was bubbled through the anhydrous DMSO or freshly
distilled methanol for 0.5 h to remove any oxygen. Then, all
the required glassware that had been dried overnight in an
oven at 110 °C and stored in a desiccator, the solvent, the
sodium thiophenoxide, and the sodium nitrate were trans-
(20) Thornton, E. R. Solvolysis Mechanisms; Ronald Press: New
York, 1964.
(21) Gajewski, J . J . J . Am. Chem Soc. 2001, 123, 10877.
(22) Smith, P. J .; Crowe, D. A. J .; Westaway, K. C. Can. J . Chem.
2001, 79, 1145.
(23) Zavada, J .; Pankova, M.; Vitek, A. Collect. Czech. Chem.
Commun. 1981, 46, 3247.
(24) Zavada, J .; Pankova, M.; Vitek, A. Collect. Czech. Chem.
Commun. 1990, 55, 695.
(25) Westaway K. C.; Ali, S. F. Can. J . Chem. 1979, 57, 1089.
3088 J . Org. Chem., Vol. 68, No. 8, 2003

SN2 Reaction between Thiophenoxide Ion and Butyl Chloride
ferred to an I2R glovebag. After the glovebag had been filled,
emptied, and refilled with extra-dry nitrogen, two stock
solutions were prepared. One stock solution was prepared by
dissolving approximately 0.05 g of accurately weighed sodium
thiophenoxide in 150 mL of anhydrous DMSO. The other stock
solution was prepared by adding approximately 0.08 g (ac-
curately weighed in a gastight syringe with the end of the
needle sealed by a rubber septum) of butyl chloride or butyl-
1,1-d2 chloride to 15 mL of anhydrous DMSO. Then, 50 mL of
the sodium thiophenoxide stock solution was pipetted into each
reaction flask and the reaction flask was sealed with a serum
cap fitted on a ground glass adapter. Each reaction flask and
the butyl chloride stock solutions were placed in the constant-
temperature bath at 20.000 ( 0.002 °C for 1 h. In the reactions
where sodium nitrate was present, the appropriate amount
of dry sodium nitrate (vide supra) was added to the sodium
thiophenoxide stock solution. The reaction was started by
injecting 1.00 mL of the butyl chloride stock solution to the
reaction flask. One milliliter samples of the reaction mixture
were removed at various times throughout the reaction and
quenched by injecting the sample into 2.5 mL of cold methanol
containing a known amount (accurately weighed) of the
internal standard, butyl phenyl ether. Each sample was
diluted to 5 mL with methanol, and 10 µL of the 5 mL solution
was injected onto a Zorbax ODS 4.6 mm × 15 cm HPLC
column, and the concentration of the product, butyl phenyl
sulfide, was determined from the [area ratio (internal standard/
butyl phenyl sulfide)] versus the [mole ratio (internal standard/
butyl phenyl sulfide)] calibration curve.26 The HPLC separa-
tion was carried out using 85% methanol-15% water as the
eluant at a flow rate of 1 mL/min. The UV detector was set at
250 nm.
substrate were prepared by dissolving approximately 10.1 g
of accurately weighed sodium thiophenoxide in 400 mL of
purified methanol. The other stock solution was prepared by
adding approximately 2.4 g (accurately weighed in a gastight
syringe with the end of the needle sealed by a rubber septum)
of butyl chloride or butyl-1,1-d2 chloride to 10 mL of purified
methanol. Then, 50 mL of the sodium thiophenoxide stock
solution was pipetted into each reaction flask, and the reaction
flask was sealed with a serum cap fitted on a ground glass
adapter. Each reaction flask and the butyl chloride stock
solutions were placed in the constant-temperature bath at
20.000 ( 0.002 °C for 1 h. In the reactions where sodium
nitrate was present, the appropriate amount of sodium nitrate
was added to the sodium thiophenoxide stock solution. The
reaction was started by injecting 1.00 mL of the butyl chloride
stock solution into 50 mL of the sodium thiophenoxide stock
solution. The individual kinetic points and the analyses were
done using the procedure described above.
Ch lor in e KIEs. These KIEs were measured by starting the
reactions as described above. Then, the partial reactions were
quenched by pouring the reaction mixture into 30 mL of
methanol containing 4 drops of a 1:4 nitric acid solution after
various extents of reaction ranging from 11% to 25% of
completion, Table 3. The complete reaction was started as
described, but it was allowed to react for over 10 half-lives
before it was quenched and the chloride ion recovered. The
procedures used for the analysis of the chloride ion, the
recovery of the silver chloride, and its conversion into methyl
chloride for isotope ratio mass spectrometry are described in
ref 27.
Ack n ow led gm en t. We are indebted to the Natural
Science and Engineering Research Council of Canada
(NSERC) for funding for this research.
In Meth a n ol. The same general procedure was used for
the reactions in methanol. The only changes were that the
stock solution containing the sodium thiophenoxide and the
J O026879D
(26) Dal Nogare, S.; J uvet, R. S., J r. In Gas Liquid Chromatography;
Interscience: New York, 1962; p 256.
(27) Westaway, K. C.; Koerner, T.; Fang, Y.-r.; Rudzinski, J .; Paneth,
P. Anal. Chem. 1998, 70, 3548.
J . Org. Chem, Vol. 68, No. 8, 2003 3089