Detection and kinetic characterization of SNV intermediates. Reactions of thiomethoxybenzylidene Meldrum's acid with thiolate ions, alkoxide ions, OH-, and water in aqueous DMSO

Article abstract of DOI:10.1139/v99-009

The reaction of thiomethoxybenzylidene Meldrum's acid (5-SMe) with thiolate and alkoxide ion nucleophiles is shown to proceed by the two-step addition-elimination S_NV mechanism in which the tetrahedral intermediate accumulates to detectable levels. For the reactions with thiolate ions, rate constants for nucleophilic addition (k₁^RX), its reverse (k_-1^RX), and for conversion of the intermediate to products (k₂^RX) were determined. For the reactions with alkoxide ions, only k₁^RX and k_-1^RX could be obtained; the intermediate in these reactions did not yield the expected substitution products, and hence no k₂^RX values could be determined. The reaction with OH^- and water are believed to follow the same mechanism, but the respective intermediates remain at steady-state levels, and only k₁^OH and k₁H₂^O for nucleophilic attack on 5-SMe were measurable. New insights regarding structure-reactivity behavior in S_NV reactions are gained from comparisons of rate and equilibrium constants in the reactions of 5-SMe with the corresponding parameters in the reactions of methoxybenzylidene Meldrum's acid (5-OMe) and benzylidene Meldrum's acid (5-H). In particular, the relative importance of steric and π-donor effects of the MeS vs. MeO group in 5-SMe and 5-OMe, respectively, and their role in affecting the intrinsic rate constants for nucleophilic addition, has been clarified by these comparisons. Our results also add support to a previous suggestion that soft-soft type interactions tend to increase intrinsic rate constants for thiolate ion addition to vinylic substrates, especially 5-SMe with the soft MeS group.

Full text of DOI:10.1139/v99-009

584
Detection and kinetic characterization of
SNV intermediates. Reactions of
thiomethoxybenzylidene Meldrum’s acid with
thiolate ions, alkoxide ions, OH–, and water in
aqueous DMSO
Claude F. Bernasconi, Rodney J. Ketner, Xin Chen, and Zvi Rappoport
Abstract: The reaction of thiomethoxybenzylidene Meldrum’s acid (5-SMe) with thiolate and alkoxide ion nucleophiles
is shown to proceed by the two-step addition–elimination SNV mechanism in which the tetrahedral intermediate
accumulates to detectable levels. For the reactions with thiolate ions, rate constants for nucleophilic addition (kRX), its
1
reverse (k−R1X), and for conversion of the intermediate to products (k2RX) were determined. For the reactions with
alkoxide ions, only kRX and k could be obtained; the intermediate in these reactions did not yield the expected
RX
1
−1
substitution products, and hence no k2RX values could be determined. The reaction with OH– and water are believed to
follow the same mechanism, but the respective intermediates remain at steady-state levels, and only kOH and kH O for
2
1
1
nucleophilic attack on 5-SMe were measurable. New insights regarding structure–reactivity behavior in SNV reactions
are gained from comparisons of rate and equilibrium constants in the reactions of 5-SMe with the corresponding
parameters in the reactions of methoxybenzylidene Meldrum’s acid (5-OMe) and benzylidene Meldrum’s acid (5-H). In
particular, the relative importance of steric and π-donor effects of the MeS vs. MeO group in 5-SMe and 5-OMe,
respectively, and their role in affecting the intrinsic rate constants for nucleophilic addition, has been clarified by these
comparisons. Our results also add support to a previous suggestion that soft–soft type interactions tend to increase
intrinsic rate constants for thiolate ion addition to vinylic substrates, especially 5-SMe with the soft MeS group.
Key words: nucleophilic vinylic substitution, intrinsic rate constants, transition state imbalances, steric/π-donor/anomeric
effects. 594
Résumé : On a démontré que la réaction du thiométhoxybenzylidène de l’acide de Meldrum (5-SMe) avec des
nucléophiles comme les ions thiolate et alcoolate se produit par un mécanisme d’addition–élimination en deux étapes,
SNV, au cours duquel l’accumulation de l’intermédiaire tétraédrique est telle qu’il est possible de le détecter. Pour les
réactions avec les ions thiolates, on a déterminé les constantes de vitesse de l’addition nucléophile (kRX), de la réaction
1
inverse (k−R1X) ainsi que celle de la conversion de l’intermédiaire en produits (k2RX). Pour les réactions avec les ions
alcoolates, on n’a pu déterminer que les valeurs de kRX et de k−R1X; dans ce cas, l’intermédiaire ne conduit pas aux
1
produits de substitution attendus et on n’a donc pas pu déterminer les valeurs de k2RX. On suppose que les réactions
avec l’eau et avec l’ion OH– se produisent par le même mécanisme; toutefois, les intermédiaires respectifs se
maintiennent aux niveaux de l’état stationnaire et, pour les réactions nucléophiles avec le 5-SMe on ne peut mesurer
que les valeurs de kOH et kH O . On a développé une meilleure compréhension du comportement structure–réactivité
2
1
1
dans les réactions SNV en procédant à des comparaisons des constantes de vitesse de réaction et d’équilibre dans les
réactions du 5-SMe avec les paramètres correspondants dans les réactions du méthoxybenzylidène de l’acide de
Meldrum (5-OMe) et du benzylidène de l’acide de Meldrum (5-H). Ces comparaisons ont permis, en particulier, de
clarifier l’importance des effets stériques et donneurs π des groupes MeS par rapport à MeO respectivement dans les
composés 5-SMe et 5-OMe et de leur rôle sur les valeurs intrinsèques des constantes de vitesse de réaction de
l’addition nucléophile. Nos résultats confirment une suggestion faite antérieurement à l’effet que les interactions de
type mou-mou tendent à augmenter les constantes intrinsèques des vitesses de réactions pour les réactions d’addition de
l’ion thiolate sur des substrats vinyliques, particulièrement le 5-SMe comportant le groupe mou MeS.
Received November 5, 1998.
This paper is dedicated to Jerry Kresge in recognition of his many achievements in chemistry.
C.F. Bernasconi1 and R.J. Ketner. Department of Chemistry and Biochemistry of the University of California, Santa Cruz,
CA 95064, U.S.A.
X. Chen and Z. Rappoport. Department of Organic Chemistry, the Hebrew University of Jerusalem, Jerusalem 91904, Israel.
1Author to whom correspondence may be addressed. Telephone: (831) 459-2035. Fax: (831) 459-2935.
e-mail: bernasconi@chemistry.ucsc.edu
Can. J. Chem. 77: 584–594 (1999)
© 1999 NRC Canada

Bernasconi et al.
585
Mots clés : substitution vinylique nucléophile, constantes intrinsèques de vitesses de réaction, déséquilibre dans état de
transition, effets stériques/donneur-π/anomère.
[Traduit par la Rédaction]
Nucleophilic substitution at vinylic carBboenrn(aSsNcVon)i iestaanl.important reaction that may proceed by a variety of different mecha-
nisms (1–7). With substrates (1) moderately or strongly activated by electron withdrawing substituents (Y,Y′), the preferred
mechanism involves two steps as shown in eq. [1] where Nu– is an anionic nucleophile, and LG is the anionic leaving group.
Recent research in our laboratory has focused on reactions of strong nucleophiles with highly activated substrates that have a
sluggish leaving group where the intermediate accumulates to detectable levels. Early examples include the reactions of β-
methoxy-α-nitrostilbene, 4-OMe, and β-thioalkoxy-α-nitrostilbene, 4-SR, with thiolate ions (8, 9); for other examples see
refs. 10–13. Recently, we have reported that in the reactions of thiolate and alkoxide ions with methoxybenzylidene
Meldrum’s acid, 5-OMe, the respective intermediates are also directly observable (14).
Our main motivation in searching for systems that allow direct observation of the intermediate is that all rate constants in
eq. [1] (k1, k–1, and k2) can be determined. A systematic study of how these rate constants depend on the Nu–, LG, and Y,Y′ is
expected to help unravel the complex interplay of the various factors that affect reactivity in SNV reactions. Apart from the
basicities of nucleophile and leaving group, these factors include π-donor effects of the leaving group as well as the
nucleophile (once attached to the substrate), anomeric, steric, polarizability, inductive/field, and resonance effects of the acti-
vating groups, and possibly others.
In the present paper we report on our study of the reactions of thiomethoxybenzylidene Meldrum’s acid, 5-SMe, with
thiolate and alkoxide ions as well as OH– and water in 50% Me2SO – 50% water (v/v). Just as with 5-OMe, the respective in-
termediates in the reactions of 5-SMe with thiolate and alkoxide ions (but not with OH–) were directly observable. However,
the reactivity patterns with 5-SMe show some important differences compared to those with 5-OMe. Of particular interest
are the insights gained not only from comparisons between 5-SMe and 5-OMe but also from those between 5-SMe and the
unsubstituted benzylidene Meldrum’s acid, 5-H (15), and especially from comparison of intrinsic rate constants2 for the vari-
ous compounds.
Results
Reaction of 5-SMe with OH– and water
The rate of hydrolysis of 5-SMe conforms to eq. [2], as for the hydrolysis of 5-OMe (14). The organic product is
the same as in the hydrolysis of 5-OMe, i.e., 5-OH,3 which under most conditions is present in the form of its anion
2 The intrinsic rate constant of a reaction with a forward rate constant k1 and a reverse rate constant k–1 is defined as ko = k1 = k–1 when the
equilibrium constant K1 = 1 (∆G° = 0).
3 The identification of 5-OH has been described elsewhere (14).
© 1999 NRC Canada

586
Can. J. Chem. Vol. 77, 1999
Fig. 1. Reaction of 5-SMe with HOCH2CH2S– at pH 10.58 in
50% DMSO – 50% water at 20°C. Pseudo-first-order rate
constants of the second (slow) process according to eq. [6].
Fig. 2. Reaction of 5-SMe with CF3CH2O– at pH 14.00 in 50%
DMSO – 50% water at 20°C, [5-SMe]o = 8.0 × 10–5 M.
Spectrum a represents 5-SMe taken in a neutral solution in the
absence of CF3CH2O–; all other spectra are at pH 14.00 in the
presence of 0.15 M CF3CH2O–. Spectrum b taken approximately
5 s after mixing, spectrum c taken 5 s after b, and all other
spectra taken at additional 5 s intervals for 120 s.
6– (pKa(5-OH) = 1.06) (14). The main difference between
the hydrolysis of 5-SMe and 5-OMe is that the reaction of
5-SMe is several orders of magnitude slower.
[2]
kobsd = kH O + kOH[OH− ]
thiolate ions, the spectra of the corresponding intermediates
were easily obtained (14, 15) because the equilibrium con-
stants for thiolate ion addition are much higher, so that much
lower [RS–] could be used.)
2
Reaction of 5-SMe with thiolate ions
When 5-SMe is mixed with a thiolate buffer two kinetic
processes are observed. The first is on the stopped-flow time
scale and leads to a loss of absorption at 335 nm, which is
the λmax of 5-SMe. This is consistent with addition of RS– to
5-SMe according to eq. [3]. The spectrum of the intermedi-
ate, 5-(SMe,SR)–, which is expected to be similar to that of
intermediates derived from the addition of thiolate or
alkoxide ions to 5-OMe (14) and 5-H (15) or alkoxide in ad-
dition to 5-SMe (λmax ≈ 260 nm, see below), could not be re-
[3]
The kinetic data conform to eq. [4]. Some of the plots of
kobsd vs. [RS–] have measurable intercepts that yield an ap-
RS
proximate value for k−R1S; k1RS and k values are reported in
−1
Table 1.
[4]
kobsd = k1RS[RS− ] + k
RS
−1
corded. This is because at the relatively high thiolate ion
concentrations (>10–2 M) required to push the equilibrium of
eq. [3] to the right there is strong interference by the
absorbance of RS–. (In the reactions of 5-OMe or 5-H with
The second process is much slower, with kobsd showing a
nonlinear dependence on [RS–] (Fig. 1). This is consistent
with conversion of 5-(SMe,SR)– to 5-SR according to eq. [5].
[5]
© 1999 NRC Canada

Table 1. Summary of rate and equilibrium constants for the reactions of 5-SMe, 5-OMe, and 5-H with various nucleophiles in 50% DMSO – 50% water (v/v) at 20°C, µ =
0.5 M.
RX–
pKaRXH
kRX (M–1 s–1)
k
(s–1)
−1
K1RX (M–1)
k2RX (s–1)
k2RX/k
RX
RX
1
−1
5-SMea
n-BuS–
11.40
10.56
8.83
15.2
14.0
(1.68 ± 0.02) × 103
(9.22 ± 0.12) × 102
(7.17 ± 0.15) × 102
(1.04 ± 0.04) × 101c
1.41 ± 0.12c
0.74 ± 0.10b
2.78 ± 0.25b
22.3 ± 0.15b
(2.26 ± 0.29) × 103
(3.32 ± 0.27) × 102
(3.21 ± 0.17) × 101
0.404 ± 0.01
0.115 ± 0.002
(2.69 ± 0.01) × 10–2
0.55
4.14 × 10–2
1.21 × 10–3
HOCH2CH2S–
MeO2CCH2S–
HCϵCCH2O–
CF3CH2O–
OH–
(4.93 ± 0.91) × 10–2b
(2.86 ± 0.35) × 101
≤(2.30 ± 0.10) × 10–2f
<0.47
17.33
–1.44
(6.43 ± 0.06) × 10–1
(2.80 ± 0.40) × 10–6e
H2O
5-OMed
0.395
n-BuS–
11.40
10.56
10.40
8.83
15.2
14.0
6.70 × 104
4.40 × 104
4.43 × 104
2.40 × 104
4.61 × 103
1.09 × 103
5.41 × 102
2.98 × 10–2e
1.70 × 105
2.57 × 104
2.22 × 104
1.71 × 103
4.35 × 106
6.81 × 104
1.11 × 10–4
2.16 × 10–4
1.98 × 10–5
2.81 × 10–4
1.76 × 10–4
HOCH2CH2S–
MeO2CCH2CH2S–
MeO2CCH2S–
HCϵCCH2O–
CF3CH2O–
OH–
1.71
2.00
14.0
1.06 × 10–3
1.60 × 10–2
<3.81 × 10–4
17.33
–1.44
H2O
5-H
4.25 × 10–5
2.68 × 10–4
3.35 × 10–3
4.71 × 10–4
3.25 × 10–3
1.57 × 10–7
n-BuS–
11.40
10.56
8.83
15.2
14.0
2.48 × 107
1.44 × 107
8.82 × 106
3.93 × 104
2.06 × 104
1.80 × 103
5.89 × 1011
5.38 × 1010
2.63 × 109
8.34 × 107
6.43 × 106
1.15 × 1010
HOCH2CH2S–
MeO2CCH2S–
HCϵCCH2O–
CF3CH2O–
OH–
17.33
a This work.
b
values based on eq. [4] are 0.48, 2.96, 26.4, and 0.041 s–1 for n-BuS–, HOCH2CH2S–, MeO2CCH2S–, and CF3CH2O–, respectively; the
RX
−1
RX
Calculated as k
= k1RX/K1RX with K1RX from eq. [6]; the k
−1
former values are considered more accurate.
c Average value based on determinations at 260 and 335 nm.
d Reference 14.
e In units of s–1.
f The value 2.30 × 10–2 refers to ku in eq. [9]; depending on which explanation holds for the reaction of 5-(SMe,OCH2CF3) , k2RX is either ku or <ku, see text.

588
Can. J. Chem. Vol. 77, 1999
Fig. 3. Reaction of 5-SMe with CF3CH2O– at pH 14.59
In this case, the nucleophilic addition step acts as a fast
monitored at 260 nm or 335 nm in 50% DMSO – 50% water at
20°C. Pseudo-first-order rate constants of the second (slow)
process according to eq. [9].
preequilibrium, and kobsd is given by eq. [6].
K1RSk2RS[RS− ]
[6]
kobsd =
1 + K1RS[RS− ]
K1RS and k2RS values determined by nonlinear squares fit to
eq. [6] are summarized in Table 1.
The observed spectral changes call for comment. At low
[RS–] (K1RS[RS–] << 1) the change in absorbance at 355 nm
(λmax of 5-SMe) is small. This is because the reaction repre-
sents conversion of 5-SMe into 5-SR with 5-(SMe,SR)– act-
ing as a steady-state intermediate, and the spectra of 5-SR
are similar to that of 5-SMe. At intermediate [RS–] (K1[RS–]
≥ 1), the changes in absorbance are large because the reac-
tion refers mainly to conversion of 5-(SMe,SR)– to 5-SR,
and 5-(SMe,SR)– absorbs very little at 335 nm. As [RS–] is
increased further, the observed changes in absorbance de-
crease again. This is because the product is a mixture of 5-
SR and the symmetrical adduct 5-(SR,SR)– that is rapidly
formed by addition of RS– to 5-SR. This adduct has a simi-
lar absorption spectrum as 5-(SMe,SR)–, which is the “reac-
tant” under these conditions.
Reaction of 5-SMe with alkoxide ions
The reaction of 5-SMe with a CF3CH2O– buffer at high
[CF3CH2O–] is characterized by three kinetic processes. The
fastest one leads to a rapid loss of 5-SMe; it is accompanied
by the formation of a new species at 260 nm. The spectrum
of the latter is characteristic of a tetrahedral intermediate,4
consistent with the formation of 5-(SMe,OR)– (R = CF3CH2),
eq. [7]. The spectral changes are shown in Fig. 2.
one of two possible interpretations. (i) The intermediate, 5-
(SMe,OCH2CF3)–, undergoes a reaction other than loss of
MeS–, leading to an unidentified product that eventually
hydrolyzes to 6–. (ii) 5-(SMe,OCH2CF3)– is converted to 5-
OCH2CF3 but then rapidly reacts and forms an unidentified
[7]
product that is eventually converted to 6–.
With respect to the kinetic determinations, the rates for
the first process were measured at pH 14.59 by monitoring
the reaction at 335 and 260 nm. The data are consistent with
eq. [8].
The second process leads to a product with a λmax
=
306 nm (Fig. 2). It is unclear what this product is. It cannot
be the substitution product 5-OCH2CF3 because its rate of
hydrolysis is expected to be much faster than the rate of the
transformation of 5-(SMe,OR)– into the new species at
306 nm. (The hydrolysis of 5-OCH2CF3 can be expected to
be at least as fast as hydrolysis of 5-OMe. At pH 14.59 and
14.01, the conditions under which the second process was
measured kinetically, kobsd, for the hydrolysis of 5-OMe is
27 and 7.2 s–1, respectively, while the kobsd values for the
second process range from 1.06 × 10–3 to 1.80 × 10–2 s–1,
depending on [CF3CH2O–].) The spectrum of the new spe-
cies is indeed inconsistent with 5-OCH2CF3 whose spec-
trum should be similar to that of 5-OMe (λmax = 278 nm)
(14). But the spectrum is also inconsistent with the hydroly-
sis product 6– (λmax = 284 nm) (14); 6– is eventually formed,
but on a much slower time scale which corresponds to the
third kinetic process mentioned earlier. Our results require
[8]
kobsd = k1RO[RO–] + kOH[OH–]
There is a contribution by the hydrolysis reaction (kOH[OH–]
term) that becomes significant at low [RO–]; k1RO was deter-
mined from the slope of plots of kobsd – kOH[OH–] vs. [RO–]
(not shown). ([RO–] was corrected for the homoassociation
constant (Kassoc = 1.8 M–1) (10) of CF3CH2O– with
CF3CH2OH.) The reported k1RO value in Table 1 is the aver-
age of k1RO = 1.53 ± 0.04 M–1 s–1 at 335 nm and 1.28 ± 0.02
RO
at 260 nm; k is too small to be obtained from these exper-
−1
iments.
The rates of the second process were measured at
pH 14.59 (λ = 260 nm) and pH 14.01 (λ = 300 nm). A plot
4 It is virtually identical to the spectra of 5-(OMe,SR)– generated by the reaction of 5-OMe with RS– (14).
© 1999 NRC Canada

Bernasconi et al.
589
of kobsd vs. [RO–] is shown in Fig. 3. It is consistent with
hydrolysis reaction, i.e., kH O = k1H O and kOH = k1OH, respec-
2
2
eq. [9],
tively (eq. [2]).
K1RO[RO− ]
Structure–reactivity relationships
[9]
kobsd
=
ku
1 + K1RO[RO− ]
A. Equilibrium constants for nucleophilic addition
with K1RO being the equilibrium constant for eq. [7], and ku
referring to the unknown reaction that depletes the interme-
diate. Least-squares analysis of the data according to eq. [9]
yields K1RO = (2.86 ± 0.35) × 101 M–1 and ku = (2.30 ± 0.10)
× 10–2 s–1.
The kinetic behavior of the reaction of 5-SR with
HCϵCCH2O– is similar to that for the reaction with
CF3CH2O–. However, the rate measurements for the second
process gave irreproducible results which are not reported,
and hence only k1RO could be determined.
It is instructive to compare the equilibrium constants for
nucleophilic addition to 5-SMe with those for addition to 5-
H and 5-OMe, which are included in Table 1. Irrespective of
the nucleophile, the relative magnitude of these equilibrium
constants follows the order K1RX (5-SMe) << K1RX(5-OMe)
<< K1RX(5-H). For example, for HOCH2CH2S– as the
nucleophile, the ratios K1RS(5-SMe):K1RS(5-OMe):K1RS(5-H)
are 6.2 × 10–9:4.8 × 10–7:1; for the other thiolate ions, these
ratios are quite similar. For CF3CH2O– as the nucleophile,
the ratios K1RO(5-SMe):K1RO(5-OMe): K1RO(5-H) are 4.5 ×
10–6:1.7 × 10–2:1.
These trends result from an interplay of differences in
steric, inductive/field, π-donor, and anomeric effects on the
three reaction systems. These factors affect the K1RX values
in the following way. (i) The increasing size in the order H
<< OMe << SMe leads to increasing steric crowding in the
adduct and hence lowers K1RX for both 5-OMe and 5-SMe
relative to 5-H, and more so for 5-SMe.5 (ii) The electron-
withdrawing inductive/field effect should enhance both
K1RX(5-OMe) and K1RX(5-SMe) relative to K1RX (5-H), and
somewhat more so for 5-OMe.6 (iii) The π-donor effect of
the MeO and MeS groups that leads to resonance stabiliza-
tion of 5-XMe (see 5-XMe ±) reduces K1RX for 5-OMe and
5-SMe relative to 5-H, with the effect on K1RX(5-OMe) be-
ing the strongest.7 (iv) The anomeric effect is an adduct sta-
bilizing factor that mainly affects the dialkoxy intermediates
in the reactions of 5-OMe with RO– 8 and enhances K1RO(5-
OMe) relative to K1RO(5-SMe) and K1RO(5-H).
Discussion
Detection of the intermediate
As is the case for 5-OMe, the respective intermediates
5-(SMe,SR)– in the reactions of 5-SMe with n-BuS–,
HOCH2CH2S–, and MeO2CCH2S–, and the intermediates 5-
(SMe,OR)– in the reactions with HCϵCCH2O– and
CF3CH2O– all accumulate to detectable levels under the ap-
propriate conditions. As has been discussed in detail else-
where (8–13), the requirements for the intermediate to be
detectable are that the equilibrium of the first step is favor-
able (k1RX[RX–] > k or K1RX[RX–] > 1) and that the rate of
formation of the intermediate is faster than its rate of con-
RX
−1
version to products (k1RX[RX] > k2RX). These conditions are
evidently met in the reactions of 5-SMe with RS– and RO–.
The only intermediate that has remained elusive under all
conditions is 5-(SMe,OH)– in the hydrolysis reaction. The
reasons for this are the same as why the corresponding inter-
mediates such as 5-(OMe,OH)– (14) and 4-(OMe,OH)– (16)
have not been detectable: their conversion to products via
two pathways not available to intermediates lacking the OH
group is faster than their formation and turns them into
steady-state intermediates. One of these additional pathways
is intramolecular acid catalysis of leaving-group departure
by the OH group (10, 11, 16). The other involves rapid
deprotonation of the OH group, generating the dianionic
form of the intermediate which expels the leaving group
much more rapidly than the monoanionic form (10, 11, 16).
Note that the rapid conversion of 5-(SMe,OH)– to products
implies that attack by water and OH– is rate limiting in the
For the thiolate ion reactions, the steric and π-donor ef-
fects appear to be the dominant factors, dramatically reduc-
ing K1RS(5-OMe) and K1RS(5-SMe) relative to K1RS(5-H),
greatly offsetting the inductive/field effect. The fact that
K1RS(5-SMe) is smaller than K1RS(5-OMe) indicates that the
greater steric effect of the MeS group combined with the
somewhat weaker inductive/field effect are dominant in this
comparison and more than offset the weaker π-donor effect
of the MeS group. Overall, when comparing the K1RS values
of the three substrates, the steric effect emerges as the most
important factor.
For the alkoxide ion reactions the K1RO values for 5-SMe
and 5-OMe are not as strongly depressed relative to K1RO for
5 Taft’s steric substituent constants, Es, are –0.55 for MeO and –1.07 for MeS, respectively (17), while Charton’s (18) νef values are 0.36 for
MeO and 0.64 for MeS, respectively.
6 σF = 0.30 and 0.20 for MeO and MeS, respectively (19).
7 σR = –0.43 and –0.15 for MeO and MeS, respectively (19).
8 In the present context, the anomeric effect (20, 21) refers to the stabilization exerted by geminal oxygen atoms (22–25), e.g., in dialkoxy
adducts such as 5-(OMe,OR)–.
© 1999 NRC Canada

590
Can. J. Chem. Vol. 77, 1999
Table 2. Brønsted coefficients and intrinsic rate constants for the
reactions of 5-SMe, 5-OMe, and 5-H with thiolate and alkoxide
ions.
Fig. 4. Reaction of 5-SMe with n-BuS–, HOCH2CH2S–, and
MeO2CCH2S–. Plots of log kRS (᭹) and log k (᭺) vs. log
RS
1
−1
K1RS. The point where the two lines intersect corresponds to log
koRS, see text.
Parameter
5-SMea
RS– nucleophiles
5-OMeb
5-Hc
d
βnuc
0.13 ± 0.06
–0.57 ± 0.04
0.70 ± 0.09
0.19 ± 0.06
–0.81 ± 0.06
2.53 ± 0.14
0.44 ± 0.07
0.17
–0.59
0.76
0.17
–0.72
0.89
e
βlg
f
βeq
g
β nnuc
0.22
0.19
n h
β
–0.78
3.66
–0.81
5.17
lg
log koRSi
j
βpush
0.75
RO– nucleophiles
0.72k
0.51
–0.97
1.48
0.23
–0.81
1.03
d
βnuc
e
βlg
f
βeq
g
β nnuc
ca. 0.5l
–0.6
ca. –0.9l
0.34
0.22
n h
β
–0.66
1.49
–0.79
2.86
lg
log koROi
aThis work.
bReference 14.
cReference 15.
compared to oxygen nucleophiles manifests itself even more
dβnuc = d log k1RX/dpKaRXH
.
dramatically in the reactions of 5-H where steric hindrance
e
βlg = d log k−R1X/dpKaRXH
.
is less important: K1RS = 5.89 × 1011 M for n-BuS– vs. K1RO
=
f
βeq = d log K1RX/dpKaRXH
gβnnuc = d log k1RX/d log K1RX
.
6.43 × 106 M–1 for CF3CH2O–. On the other hand, in the re-
action of 5-OMe with alkoxide ions the anomeric effect re-
duces the advantage of the sulfur bases: K1RS = 1.70 × 105 M
(n-BuS–) vs. K1RO = 6.81 × 104 M (CF3CH2O–).
.
hβnlg = d log k−R1X/d log K1RX
i
.
koRX = k1RX = k
for K1RX = 1.
RX
−1
j
βpush= d log k2RS/dpKaRSH
.
The higher carbon basicity of sulfur compared to oxygen
bases has been commonly attributed to stronger polarizabil-
ity (26–28) or “softness” (29, 30) of the sulfur compounds,
i.e., in the reaction of RS– with polarizable electrophiles
such as 5-H, 5-OMe, 5-SMe, and other alkenes there is a fa-
vorable soft–soft interaction, while in the reactions of RO–
with the same electrophiles there is a less favorable hard–
soft interaction. An additional factor, particularly important
in aqueous media, is the weaker solvation of thiolate com-
pared to that of alkoxide ions (31–33).
kNo standard deviation given (two points only).
lEstimated, see text.
B. Rate constants for nucleophilic addition by RS– and
RO–
5-H as for the thiolate ion reactions. Here, the smaller size
of the alkoxide ion nucleophiles reduces the steric crowding
in the respective intermediates. On the other hand, K1RO (5-
OMe) appears to be much more enhanced relative to K1RO
(5-SMe) than is the case with thiolate ion nucleophiles. This
demonstrates the importance of the anomeric effect, which
stabilizes the dialkoxy but not the alkoxy–thioalkoxy com-
plexes.
The rate constants for nucleophilic attack on 5-SMe, 5-
OMe, and 5-H follow the same qualitative pattern as the
equilibrium constants, i.e., k1RX(5-SMe) << k1RX(5-OMe) <<
k1RX(5-H). Specifically, the ratios k1RS(5-SMe):k1RS(5-OMe):
k1RS(5-H) are 6.4 × 10–5:3.1 × 10–3:1 for HOCH2CH2S–,
which is representative for all thiolate ion nucleophiles. For
the CF3CH2O– reaction the ratios k1RO(5-SMe):k1RO(5-OMe):
k1RO(5-H) are 6.8 × 10–5:5.3 × 10–2:1. For the thiolate ion re-
actions the rate constant ratios are substantially smaller than
the equilibrium constant ratios, consistent with the notion
that the electronic and steric factors discussed above are
only partially expressed at the transition state. On the other
hand, for the alkoxide ion reactions the rate constant ratios
are quite comparable to the equilibrium constant ratios, sug-
With respect to their absolute magnitude, the K1RS values
are significantly larger than the K1RO for addition of alkoxide
ion of comparable pKa. For example, K1RS for n-BuS– addi-
tion to 5-SMe (2.26 × 103 M–1) is 79-fold larger than K1RO
=
28.6 M–1 for CF3CH2O– addition to the same substrate, even
though the pKaRSH of n-BuSH is only 11.40, while that of
CF3CH2O– is 14.0. The enhanced carbon basicity of sulfur
© 1999 NRC Canada

Bernasconi et al.
591
gesting that in moving from the reactants to the transition
state, some or all the factors are nearly as strongly expressed
as in moving from the reactants to the adducts. These results
indicate that there must be significant differences in the in-
trinsic rate constants2 of the various reactions.
conclusion because predictions as to whether development
of steric effects is generally ahead of bond formation or lags
behind it have been difficult to make (37). Specifically, in
our earlier study (14) where only data on 5-OMe and 5-H
were available, no definite conclusion regarding steric ef-
fects on the intrinsic rate constants could be drawn.
The intrinsic rate constants are reported in Table 2. For
the thiolate ion reactions they were obtained by suitable ex-
(iii) Following generally observed behavior (37–39), the
partial desolvation of the nucleophiles that occurs as they
enter the transition state should be more advanced than bond
formation. This has the effect of reducing the intrinsic rate
constant. Because the solvation of highly basic alkoxide ions
is stronger than that of thiolate ions (31–33), koRO decreases
more than koRS, which explains, at least in part, why koRS >>
koRO for all substrates.
RX
trapolation of plots of log k1RX or log k
vs. log K1RX
−1
(Fig. 4); they also yield the normalized Brønsted coefficients
βnnuc and βlng. For the alkoxide ion reactions, log koRX was esti-
mated as log k1RO – 0.5 log K1RO with k1RO and K1RO for
CF3CH2O–. (This is equivalent to applying the simplest ver-
sion of the Marcus equation (34), ∆G‡ = ∆Go‡ + 0.5∆G° +
(∆G°)2/16∆Go‡ and neglecting the third term, which should
be very small in our case. The factor 0.5 in front of ∆G° im-
plies βnnuc ≈ 0.5, which is a reasonable value in view of the
trend in βnnuc for RO– addition to 5-H and 5-OMe discussed
below.)
(iv) An additional PNS effect contributing to the high koRS
values in the thiolate ion reactions may come from the soft–
soft interactions if these interactions developed ahead of
bond formation. As has been discussed in more detail else-
where (40), this is a reasonable possibility.
There are large differences in the koRX values for the vari-
ous reactions. For a given type of nucleophile koRX(5-SMe)
<< koRX(5-OMe) << koRX(5-H), while for a given substrate
the intrinsic rate constant for the thiolate ion reactions is
much higher than for the alkoxide ion reactions. To under-
stand why such large differences in the koRX values exist, we
need to remind ourselves that intrinsic rate constants are
purely kinetic quantities that have been corrected for differ-
ences in the equilibrium constants that arise from different
degrees of steric, inductive/field, π-donor, anomeric, solva-
tion, and polarizability effects. This means that if at the tran-
sition state these factors were being developed or lost in
proportion to the degree of bond formation (“balanced” tran-
sition state), the koRX values should be the same for all the re-
actions. The fact that they are different means that the
transition states are imbalanced in several respects.
The rules of the principle of nonperfect synchronization
(PNS) allow us to describe these effects as follows. (The
PNS (35–37) states that if the development of a product sta-
bilizing factor lags behind bond changes or charge transfer
at the transition state, ko is reduced. The same is true if the
loss of a reactant stabilizing factor runs ahead of bond
changes or charge transfer. For product stabilizing factors
that develop early or reactant stabilizing factors that are lost
late, ko is enhanced. For product or reactant destabilizing
factors, the opposite relationships hold.)
(v) The question whether the anomeric effect in the reac-
tions of 5-OMe with alkoxide ions develops early (increase
in koRO(5-OMe)) or late (decrease in koRO(5-OMe)) is an in-
teresting one. From the fact that for the alkoxide ion reac-
tions log koRO(5-OMe) – log koRO(5-SMe) is large (≈2.39),
while for the thiolate ion reactions log koRS(5-OMe) – log
koRS(5-SMe) is small (1.13), one might conclude that the
anomeric effect enhances log koRO(5-OMe), implying that its
development is ahead of bond formation of the transition
state. However, an alternative interpretation of these differ-
ences is that it is log koRS(5-SMe) for the thiolate ion reac-
tions which is unusually large, making the difference log
koRS(5-OMe) – log koRS(5-SMe) look small. An enhanced log
koRS(5-SMe) value could be the result of the MeS group
making 5-SMe softer than 5-OMe, which would increase
the soft–soft interactions and its koRS(5-SMe) increasing PNS
effect. (This may be similar to findings Bunnett has de-
scribed 40 years ago (41).) Comparisons between 5-OMe
and 5-H render this alternative interpretation more plausible:
the difference log koRO(5-OMe) – log koRO(5-H) = –1.37 is,
within experimental error, indistinguishable from log koRS(5-
OMe) – log koRS(5-H) = –1.51, i.e., there is no enhancement
of log koRO(5-OMe) that could be attributed to early develop-
ment of the anomeric effect.
C. Rate constants for the reactions with OH– and water
The reactivity ratios for the OH– reaction are k1OH(5-
SMe):k1OH(5-OMe):k1OH(5-H) = 3.6 × 10–4:0.3:1, which
compares with k1RO(5-SMe):k1RO(5-OMe):k1RO(5-H) = 6.8 ×
10–5:5.3 × 10–2:1 for the CF3CH2O– reactions. The smaller
reductions in k1OH for the reactions with 5-SMe and 5-OMe
relative to 5-H must be the result of the smaller size of OH–
compared to CF3CH2O–, which reduces the steric effect in
the reactions with 5-SMe and 5-OMe. The fact that the ab-
solute values of k1OH are all lower than for the corresponding
alkoxide ion reactions can be attributed to the even stronger
solvation of OH– than that of the alkoxide ions (38, 39, 42),
which reduces the intrinsic rate constant for OH– addition
(PNS effect) even more than for alkoxide ion addition.
For the water reaction only data for 5-SMe and 5-OMe
(i) The π-donor resonance stabilization of 5-SMe and 5-
OMe is expected to follow the generally observed pattern of
resonance effects (35–37), and hence its loss should be more
advanced than bond formation. This results in a reduction of
the intrinsic rate constants and must be part of the reason
why koRX for both 5-OMe and 5-SMe are lower than for 5-H.
(ii) The observation that the koRX(5-SMe) values are more
strongly reduced than the koRX(5-OMe) values, despite the
smaller π-donor effect of the MeS group compared to that of
the MeO group, implies that an additional PNS effect is op-
erative. We propose that this additional effect is the early de-
velopment of the steric factor. (In the context of the PNS
(35–37), steric crowding is a product destabilizing factor
which depresses ko if it is more advanced than bond forma-
tion.) This factor is stronger for the reactions with 5-SMe
than with 5-OMe and should therefore depress koRX(5-SMe)
more than koRX(5-OMe). This is a particularly significant
are available; they yield k1H O(5-SMe)/k1H O(5-OMe) = 9.4 ×
10–5, which compares with k1OH(5-SMe)/k1OH(5-OMe) = 1.2
2
2
© 1999 NRC Canada

592
Can. J. Chem. Vol. 77, 1999
× 10–3 for the OH– reaction. The greater selectivity in the
water reaction may reflect a more intermediate-like transi-
tion state because the reaction is thermodynamically less fa-
vorable. As a result, k1H O(5-SMe) is more strongly reduced
2
by the steric effect, and k1H O(5-OMe) is more strongly en-
2
hanced by the anomeric effect, consistent with the
Hammond postulate (43–45).
RS
k2RS/k
D. Brønsted coefficients
ratio is close to unity (0.55), reflecting the fact that
−1
pKaRSH ≈ pKaMeSH
Table 2 reports a variety of Brønsted coefficients such as
.
βnuc, β , βeq, βnnuc, and βlng. For the thiolate ion reactions, βnuc
RS
–
lg
(iv) The
value for the reaction of HOCH2CH2S with
k2
and βnnuc follow the well-known pattern of being very small
5-OMe does not fit the correlation of log k2RS with the pKaRSH
defined by n-BuS– and MeO2CCH2CH2S–; it shows a strong
positive deviation (14). This contrasts with k2RS for the reac-
tion of HOCH2CH2S– with 5-SMe which shows no devia-
tion. The exalted k2RS value for 5-OMe has been attributed to
intramolecular hydrogen bonding assistance of methoxide
ion departure by the OH group as shown in 8 (14). The ab-
sence of such assistance in the reaction of 5-SMe is consis-
tent with the weaker susceptibility to acid catalysis of
thiolate ion departure compared to alkoxide ion departure
(12, 13).
(10, 11, 38, 40, 46–48) and |β | or |βnlg| being quite large, im-
lg
plying little bond formation at the transition state. (This is
the traditional view (43, 44, 49), although this view has been
challenged (50–52) as well as defended (45).) It has been ar-
gued previously (15, 40) that the small βnuc (βnnuc) values are
not necessarily the result of a Hammond–Leffler effect (43,
44) arising from very large equilibrium constants for thiolate
ion additions. Our results with 5-SMe for which the K1RS val-
ues are much smaller than for 5-OMe and particularly 5-H
support this view.
(v) The k2RS values for MeS– departure from 5-(SMe,SR)–
are much higher than for MeO– departure from 5-
(OMe,SR)–. For example, k2RS(5-SMe)/k2RS(5-OMe) = 3.6 ×
103 in the reaction with n-BuS–. (The k2RS(5-SMe)/k2RS(5-
OMe) ratio in the reaction with HOCH2CH2S– is approxi-
mately 5 × 103. This ratio is only approximate because for
k2RS(5-OMe) the reaction with MeO2CCH2CH2S– whose
pKaRSH is very close to that of HOCH2CH2S– is used instead
of the reaction with HOCH2CH2S–; the reaction of 5-OMe
with HOCH2CH2S– is subject to intramolecular hydrogen
bonding catalysis and has an exalted k2RS value (see
(iv) above).) This is mainly the result of two factors. One is
the superior leaving group ability of MeS– compared to
MeO–, which is related to the lower proton basicity of the
sulfur base (pKaMeSH ≈ 11.19 vs. pKaMeOH ≈ 17.2 (10)) but
probably attenuated by its stronger carbon basicity. The
other is the greater steric crowding in 5-(SMe,SR)– com-
pared to 5-(SMe,OR)–, which should enhance the k2RS(5-
SMe)/k2RS(5-OMe) ratio. The steric effect on leaving group
departure is also seen in the fact that k2RS for MeS– departure
from the more crowded 5-(SMe,SR)– in the reaction of n-
BuS– with 5-SMe is much larger than k2RS for MeS– depar-
ture from the less crowded 5-(SMe,OR)– in the reaction of
CF3CH2O– with 5-SMe. Without the steric effect, the ex-
pected stronger electronic push by the CF3CH2O group com-
pared to that of the n-BuS group should render k2RO larger
than k2RS. The same comments apply to the comparison of
k2RS with k2RO in the reactions of 5-OMe with n-BuS– and
CF3CH2O–.
The finding that βeq for the reaction of thiolate ions with
5-SMe is smaller than 1 is again consistent with the results
of 5-OMe (14) and 5-H (15) as well as with other examples
(40, 46).
In contrast to the thiolate ion reactions βnuc (βnnuc), alkoxide
ion addition appears to follow the Hammond–Leffler (43,
44) trend towards larger values for thermodynamically less
favored reactions. This is confirmed by the results for 5-
SMe. The unusually large βeq value for alkoxide ion addition
to 5-OMe has been discussed elsewhere and attributed to the
anomeric effect (14).
E. Rate constants for leaving group expulsion
Our data set for the k2RX process is not as complete as for
RX
the nucleophilic addition step. However, by including k
,
−1
which also represents leaving group expulsion, into the dis-
cussion, several insights emerge.
(i) The k−R1X values indicate that there is moderate to strong
inverse correlation between leaving group departure rate
constants and the proton basicity of the leaving group. This
is reflected in the relatively large β values which range
lg
from –0.57 to –0.97 (Table 2).
(ii) The k2RS values for the reactions of thiolate ions with
5-SMe and 5-OMe increase significantly with increasing
pKaRSH, indicating a substantial electronic push by the RS
group left behind. This is reflected in the βpush
=
d log k2RX͞dpKaRXH values of 0.44 (5-SMe) and 0.75 (5-OMe),
respectively (Table 2). This push is the result of the develop-
ing resonance effect in the product (5-SR)±.
(iii) Because a decrease in pKaRSH enhances k−R1S but lowers
RS
k2RS, the k2RS/k ratios in the reactions with thiolate
Experimental section
−1
ions increase strongly with decreasing pKaRSH(d(log
k2RS/k−R1S)/dpKaRSH = 1.01 for 5-SMe and 1.34 for 5-OMe).
Materials
Thiomethoxybenzylidene Meldrum’s acid, 5-SMe, was
synthesized as described by Huang and Chen (54); mp 160–
162°C (lit. (54) 164°C). 1H NMR (CDCl3, 400 MHz) δ: 1.77
(d(log k2RS/k−R1S)/dpKaRSH = βpush – β , with β referring to the
lg
lg
k
step.) For the reaction of n-BuS– with 5-SMe, the
RS
−1
9 Estimated as 0.3 units lower than pKan−BuSH, based on the fact that in pure water the pKa difference between MeSH and n-BuSH is 0.3 units
(53).
© 1999 NRC Canada

Bernasconi et al.
593
(s, 6H), 1.90 (s, 3H), 7.43–7.48 (m, 5H). All reagents were
purified as described earlier (14).
with increasing pKa, moderate to large βpush values), leaving
group atom (higher rate with S than with O), and steric ef-
fects (increased rate with bulkier leaving and (or) remaining
groups).
Methodology
(vii) In the reaction of 5-OMe with HOCH2CH2S–, the
k2RS value is abnormally high because of intramolecular hy-
drogen bonding assistance of MeO– departure by the OH
group. In contrast, the k2RS value for the reaction of 5-SMe
with the same nucleophile is not enhanced, indicating that
intermolecular assistance of MeS– departure is not impor-
tant.
Preparation of solutions, pH measurements, in situ gener-
ation of intermediates, recording of spectra, and kinetic mea-
surements were performed as described before (14).
Conclusions
(i) The conditions necessary for the direct observation of
the SNV intermediate, i.e., K1RX[RX–] > 1 and k1RX[RX–] >
k2RX, are easily met for the reactions of 5-SMe with all
nucleophiles except for OH–. In this latter reaction the acidic
nature of the OH group in 5-(SMe,OH)– leads to additional
pathways that accelerate the conversion of the intermediate
to products to the point of turning 5-(SMe,OH)– into an un-
detectable steady state intermediate.
Acknowledgments
This research was supported by grant CHE-9307659 from
the National Science Foundation (C.F.B.) and a grant from
the U.S.–Israel Binational Science Foundation (Z.R.).
(ii) The ratios of the equilibrium constants for
HOCH2CH2S– addition to 5-SMe, 5-OMe, and 5-H are 6.2
× 10–9:4.8 × 10–7:1; they are representative for all thiolate
ion reactions of this study. The strong reduction in K1RS(5-
SMe) and K1RS(5-OMe) compared to K1RS(5-H) is mainly the
result of the steric and π-donor effects. The fact that K1RS(5-
SMe) < K1RS(5-OMe) indicates that the steric effect is, over-
all, the dominant factor. The corresponding ratios for
CF3CH2O– addition are 4.5 × 10–6:1.1 × 10–2:1. They indi-
cate less severe crowding in the intermediates derived from
5-SMe and 5-OMe. The smaller K1RO(5-SMe)/K1RO(5-OMe)
ratio for the alkoxide ion reactions compared to the K1RS(5-
SMe)/K1RS(5-OMe) ratio for the thiolate ion reactions shows
the importance of the anomeric effect in 5-(OMe,OR)–.
(iii) Because of the greater polarizability and weaker sol-
vation of thiolate compared to alkoxide ions, the equilibrium
constants for thiolate addition to all substrates are much
higher than for addition of alkoxide ions of the same proton
basicity.
(iv) The rate constants for nucleophilic addition to 5-SMe,
5-OMe, and 5-H are affected by the same factors as the
equilibrium constants, but the relative importance of these
factors is different because of multiple transition state imbal-
ances that affect the intrinsic rate constants. For a given type
of nucleophile the koRX values follow the order 5-SMe << 5-
OMe << 5-H, which is mainly the result of early loss of the
π-donor stabilization of 5-SMe and 5-OMe and early devel-
opment of the steric effect at the transition state. For a given
substrate, koRS >> koRO; this is a consequence of early
desolvation of the nucleophile, probably combined with
early development of the soft–soft interactions in the reac-
tions with thiolate ions. These soft–soft interactions appear
to be particularly beneficial in the reaction of 5-SMe due to
the softness of the MeS group.
References
1. Z. Rappoport. Adv. Phys. Org. Chem. 7, 1 (1969).
2. G. Modena. Acc. Chem. Res. 4, 73 (1971).
3. S.I. Miller. Tetrahedron, 33, 1211 (1977).
4. Z. Rappoport. Acc. Chem. Res. 14, 7 (1981).
5. Z. Rappoport. Recl. Trav. Chim. Pays-Bas, 104, 309 (1985).
6. B.A. Shainyan. Usp. Khim. 55, 942 (1986).
7. Z. Rappoport. Acc. Chem. Res. 25, 474 (1992).
8. C.F. Bernasconi, J. Fassberg, R.B. Killion, Jr., and Z. Rappoport.
J. Am. Chem. Soc. 111, 6862 (1989).
9. C.F. Bernasconi, J. Fassberg, R.B. Killion, Jr., and Z. Rappoport.
J. Am. Chem. Soc. 112, 3169 (1990).
10. C.F. Bernasconi, D.F. Schuck, R.J. Ketner, M. Weiss, and Z.
Rappoport. J. Am. Chem. Soc. 116, 11 764 (1994).
11. C.F. Bernasconi, D.F. Schuck, R.J. Ketner, I. Eventova, and Z.
Rappoport. J. Am. Chem. Soc. 117, 2719 (1995).
12. C.F. Bernasconi, A.E. Leyes, Z. Rappoport, and I. Eventova. J.
Am. Chem. Soc. 115, 7513 (1993).
13. C.F. Bernasconi, A.E. Leyes, I. Eventova, and Z. Rappoport. J.
Am. Chem. Soc. 117, 1703 (1995).
14. C.F. Bernasconi, R.J. Ketner, X. Chen, and Z. Rappoport. J.
Am. Chem. Soc. 120, 7461 (1998).
15. C.F. Bernasconi and R.J. Ketner. J. Org. Chem. 60, 6266 (1998).
16. C.F. Bernasconi, J. Fassberg, R.B. Killion, D.F. Schuck, and Z.
Rappoport. J. Am. Chem. Soc. 113, 4937 (1991).
17. S.H. Unger and C. Hansch. Prog. Phys. Org. Chem. 12, 91
(1976).
18. M. Charton. Stud. Org. Chem. 42, 629 (1991).
19. C. Hansch, A. Leo, and R.W. Taft. Chem. Rev. 91, 165 (1991).
20. A.G. Kirby. The anomeric effect and related stereoelectronic
effects of oxygen. Springer Verlag, Berlin, 1983.
21. P.v.R. Schleyer, E.D. Jemmis, and G.W. Spitznagel. J. Am.
Chem. Soc. 107, 6393 (1985).
22. J. Hine and A.W. Klueppl. J. Am. Chem. Soc. 96, 2924 (1974).
23. K.B. Wiberg and R.R. Squires. J. Chem. Thermodyn. 11, 773
(1979).
24. M.P. Harcourt and R.A. More O’Ferrall. J. Chem. Soc. Chem.
Commun. 823 (1987).
(v) The fact that βnuc is very low for thiolate ion addition
to 5-SMe, despite the relatively small equilibrium constants,
confirms previous conclusions that in thiolate additions to
electrophiles the generally observed low βnuc values are not
the result of a Hammond effect.
(vi) The leaving group departure rate constants depend on
the following factors: the pKa of the leaving group (in-
creased rate with decreasing pKa, moderate to large negative
25. M.P. Harcourt and R.A. More O’Ferrall. Bull. Soc. Chim. Fr.
407 (1988).
26. E.G. Sander and W.P. Jencks. J. Am. Chem. Soc. 90, 6154
(1968).
27. J. Hine and R.D. Weimar, Jr. J. Am. Chem. Soc. 87, 3387
(1965).
β
values), the pKa of the remaining group (increased rate
lg
© 1999 NRC Canada

594
Can. J. Chem. Vol. 77, 1999
43. G.S. Hammond. J. Am. Chem. Soc. 77, 554 (1955).
44. J.E. Leffler and E. Grunwald. Rates and equilibria of organic
reactions. Wiley, New York. 1963. p. 156.
45. W.P. Jencks. Chem. Rev. 85, 511 (1985).
46. C.F. Bernasconi and D.F. Schuck. J. Org. Chem. 57, 2365
(1992).
28. J. Hine. Structural effects on equilibria in organic chemistry.
Wiley, New York. 1975. p. 225.
29. R.G. Pearson and J. Songstad. J. Am. Chem. Soc. 89, 1827
(1967).
30. R.G. Pearson. Surv. Prog. Chem. 5, 1 (1969).
31. A.J. Parker. Chem. Rev. 69, 1 (1969).
32. F.G. Bordwell and D.L. Hughes. J. Org. Chem. 47, 3224
(1982).
47. H.F. Gilbert and W.P. Jencks. J. Am. Chem. Soc. 99, 7931
(1977).
33. R.W. Taft. Prog. Phys. Org. Chem. 14, 247 (1983).
34. R.A. Marcus. J. Chem. Phys. 43, 679 (1965).
35. C.F. Bernasconi. Acc. Chem. Res. 20, 301 (1987).
36. C.F. Bernasconi. Acc. Chem. Res. 25, 9 (1992).
37. C.F. Bernasconi. Adv. Phys. Org. Chem. 27, 119 (1992).
38. D.J. Hupe and W.P. Jencks. J. Am. Chem. Soc. 99, 451 (1977).
39. W.P. Jencks, S.R. Brant, J.R. Gandler, G. Fendrich, and C.
Nakamura. J. Am. Chem. Soc. 104, 7054 (1982).
40. C.F. Bernasconi and R.B. Killion, Jr. J. Am. Chem. Soc. 110,
7506 (1988).
48. C.D. Ritchie and J.P. Gandler. J. Am. Chem. Soc. 101, 7318
(1979).
49. A.J. Kresge. Acc. Chem. Res. 8, 354 (1975).
50. A. Pross. J. Org. Chem. 49, 1811 (1984).
51. F.G. Bordwell and D.L. Hughes. J. Am. Chem. Soc. 107, 4737
(1985).
52. A. Pross and S.S. Shaik. New J. Chem. 13, 427 (1989).
53. W.P. Jencks and J. Regenstein. In Ionization constants of acids
and bases, CRC handbook of biochemistry and molecular biol-
ogy. 3rd ed. Edited by G.D. Fasman. CRC Press, Cleveland.
1977. p. 305.
41. J.F. Bunnett. J. Am. Chem. Soc. 79, 5969 (1957).
42. A.J. Kresge. Chem. Soc. Rev. 2, 475 (1973).
54. X. Huang and B.-C. Chen. Synthesis, 480 (1987).
© 1999 NRC Canada