Spiro-λ4-sulfanes with a N-SIV-O axial bond system. A kinetic study on the mechanism of hydrolysis

Article abstract of DOI:10.1039/a701002h

Kinetics of the hydrolysis of diaryl(acylamino)(acyloxy)spiro-λ⁴-sulfanes (2a-e, 3-5) leading to sulfoxides have been studied under pseudo-first-order conditions in dioxane-water mixtures or aqueous buffer solutions. A sulfonium-carboxylate-typ

Full text of DOI:10.1039/a701002h

View Article Online / Journal Homepage / Table of Contents for this issue
Spiro-ë4-sulfanes with a N᎐SIV᎐O axial bond system. A kinetic study
on the mechanism of hydrolysis
Elemér Vass, Ferenc Ruff,* István Kapovits, Dénes Szabó and Árpád Kucsman
Department of Organic Chemistry, Eötvös University, PO Box 32, H-1518 Budapest 112,
Hungary
K inetics of the hydrolysis of diaryl(acylamino)(acyloxy)spiro-ë4-sulfanes (2a–e, 3–5) leading to sulfoxides
have been studied under pseudo-first-order conditions in dioxane–water mixtures or aqueous buffer
solutions. A sulfonium-carboxylate-type dipolar structure of the starting spiro-ë4-sulfanes is supported
by IR spectroscopic data. Solvent polarity and ionic strength have no significant influence on the rate of
hydrolysis of compounds 2a and 3. Electron-withdrawing para-substituents promote the reaction in
both neutral (ñ 1.43) and acidic media (ñcat 0.90). H ydrolysis is moderately accelerated by strong acids
owing to the protonation of the negatively polarized acyloxy group of the substrate. In 50 :50 (v/v)
dioxane–H 2O(D 2O) the primary deuterium isotope effect is kH O/kD 3.68. Spiro-ë4-sulfanes 2a–e and 3
2O
2
with a five-membered N -containing ring are much more reactive than the six-membered analogues 4 and 5.
A mechanism involving the rate-determining nucleophilic attack of water on the positively polarized
sulfur atom is proposed which is accompanied by a simultaneous O᎐H and S᎐N bond cleavage. Spiro-ë4-
sulfanes 4 and 5 with six-membered N -containing spiro-rings, which are slightly reactive towards water,
undergo a fast parallel reaction with OH Ϫ ions even in neutral solutions. The different reactivities of
(acylamino)(acyloxy)spiro-ë4-sulfanes and their diacyloxy analogues are discussed and interpreted.
reactivity (ρ Ϫ0.52); (iii) the reaction is more or less strongly
Introduction
catalysed by acids, depending on the substituent (for the cat-
alytic constants ρcat Ϫ1.55); (iv) in neutral media the reaction
shows a secondary deuterium solvent isotope effect (kH O/kD
In a previous paper 1 we reported on kinetic investigations of
the hydrolysis of diaryl(diacyloxy)spiro-λ4-sulfanes (e.g. 1a–e,
2O
2
1.66) which is superimposed on the isotope effect of a proton-
ation pre-equilibrium in acidic media, and therefore an inverse
isotope effect is observed (the ratio of catalytic constants was
0.56); (v) diaryl(diacyloxy)spiro-λ4-sulfanes containing two
five-membered spiro-rings B are significantly more stable
towards hydrolysis than the analogues with six-membered ring
(the so-called five-membered ring effect has been observed). On
the basis of these findings a multi-step mechanism has been
proposed, according to which the spiro-λ4-sulfane in solution
first undergoes an equilibrium process like valence isomeriz-
ation to give a reactive monocyclic sulfonium-carboxylate
zwitterion (in acidic media protonation leads to a carboxy-
substituted monocyclic sulfonium cation); the sulfonium
species, attacked in the subsequent step by a water molecule, is
converted finally into sulfoxide by fast proton-transfer steps. An
equilibrium involving ring-opening and recyclization for spiro-
λ4-sulfanes has been proved earlier by isotope labelling2 as well
as by NMR measurements.3 Both the substituent effect and the
ring-size effect observed have been correlated with the stability
of the spiro-rings B, i.e. with the readiness of the formation of a
reactive sulfonium-carboxylate intermediate.
O
O
O
S
O
S
B
B
R
R
B
A
O
N
Me
O
O
1a–e
c
2a–e
c
a
b
d
e
a
b
d
e
R = H NO2 Cl OMe NMe2
R = H NO2 Cl Me OMe
O
O
B
O
O
B
S
A
N
S
A
N
Me
O
Me
O
3
4
O
B
The synthesis and X-ray structure determination of
diaryl(acylamino)(acyloxy)spiro-λ4-sulfanes 2a and 3–5 with
O
S
A
N
O
O
O
-
δ-
O
O
••
O
Me
O
••
S
δ+
+
+
••
S
S
5
••
N
-
N
N
also known as diaryldiacyloxyspirosulfuranes) with two equiv-
alent SIV᎐O(acyl) hypervalent bonds in the axial position. For
this reaction, which leads to bis(carboxyaryl)-sulfoxides with
the opening of both spiro-rings B, we found that (i) the
hydrolysis is considerably assisted by strongly ionizing reac-
tion media, (ii) electron-releasing substituents on the aromatic
ring increase, whereas electron-withdrawing ones decrease the
Me
Me
Me
O
O
O
2a-A
(minor contribution)
2a-B
(major contribution)
2a-C
axial N(acyl)᎐SIV᎐O(acyl) bond system has been reported in an
earlier paper.4 Whereas the diaryl(diacyloxy)spiro-λ4-sulfane 1a
J. Chem. Soc., Perkin Trans. 2, 1997
2061

View Article Online
exhibits an almost perfect trigonal-bipyramidal (TBP) arrange-
ment and medium length axial hypervalent S᎐O bonds (1.84 Å)
about the central sulfur atom,5 the diaryl(acylamino)(acyloxy)
analogues show a similar TBP geometry but with a practically
covalent S᎐N bond (1.71–1.73 Å) in spiro-ring A and with an
unusually long and polarized S᎐O hypervalent bond (2.13–2.25
Å) in spiro-ring B. For 2a the ‘disproportion’ of axial hyper-
valent bonds may be described by a major contribution of the
limiting structure 2a-B. The ‘condensed’ formula 2a-C reflects
the view that the unusually long and weak hypervalent S᎐O
bond in 2a may also be regarded as an effective S ؒ ؒ ؒ O close
contact (cf. refs. 6 and 7). Owing to this interaction the TBP
geometry with the linearity of the axial N᎐S ؒ ؒ ؒ O part is con-
served in the actual conformation, and the charges on the
sulfonium–sulfur and acyloxy–oxygen atoms are somewhat
reduced.
The νC᎐O frequencies obtained for spiro-λ4-sulfanes 1a, 2a
and 3–5 in different solvents are summarized in Table 1. The
᎐
(diacyloxy)spiro-λ4-sulfane 1a with an axial O᎐SIV᎐O array
exhibits a single νC᎐O(acyloxy) band both in acetonitrile and in
dimethyl sulfoxide ᎐(DMSO). The observed value is equally far
from the two extremes reported by Livant and Martin,8 indi-
cating that the two equivalent axial Sδϩ᎐OδϪ (hypervalent)
bonds in 1a are half-polarized as is also shown by S᎐O bond
length values (Table 1). The corresponding νC᎐O(acyloxy) data
for (acylamino)(acyloxy)spiro-λ4-sulfanes 2a and 3–5 with an
axial N᎐SIV᎐O array, however, point to a dramatic shift
towards the zwitterionic form, which is also assisted by polar
solvents (acetonitrile < dimethyl sulfoxide < 50:50 dioxane–
D2O). Extremely low νC᎐O(acyloxy) values were observed in the
most polar dioxane–D᎐2O solvent for compounds 4 and 5,
which also show the longest S᎐O interatomic distances (cf.
Table 1).
In this paper, we report on the mechanism of hydrolysis of
diaryl(acylamino)(acyloxy)spiro-λ4-sulfanes 2–5 leading to the
corresponding sulfoxides 6–9, using data obtained from
detailed kinetic studies. The synthesis of compounds 2b–e is
also described.
The νC᎐O(acylamino) frequencies of spiro-λ4-sulfanes 2–5 are
᎐
comparable to (or somewhat lower than) the amide I frequen-
cies of the corresponding lactams, pointing almost fully
covalent S᎐N bonds, which is also reflected in short S᎐N bond
lengths (cf. Table 1).
The splitting of acyloxy and/or acylamino C᎐O bands in
᎐
CO2H
some cases may be explained by an interaction with solvent
CO2H
molecules. For spectra recorded in dioxane–D2O the lower fre-
R
S
S
quency values may be related to hydrogen-bonded C᎐O groups.
᎐
Data suggest that compounds 4 and 5, in which the N atom is
incorporated in a six-membered ring A, have more pronounced
ionic character than 2a and 3 with a five-membered ring. On the
other hand, ionic character is also increased if the O-containing
five-membered ring B is enlarged to a six-membered one
(2a < 3; 4 < 5).
From both IR and X-ray data we may conclude that the
structure of the (acylamino)(acyloxy)spiro-λ4-sulfanes 2–5 with
highly polarized S᎐O bonds can be well described by a high
contribution of a sulfonium-carboxylate-type limiting structure
(e.g. 2a-B) and the polarity of the structure depends on the
solvent, too. However, the role of steric factors too cannot be
neglected, as is pointed out in refs. 6 and 7.
O
O
CONHMe
CONHMe
6a–e
c
7
a
b
d
e
R = H NO2 Cl Me OMe
CO2H
CO2H
S
S
O
O
MeNHCO
MeNHCO
8
9
As will be shown below, structural differences observed for
spiro-λ4-sulfanes are also reflected in their reactivity. Because the
hydrolysis of (acylamino)(acyloxy)spiro-λ4-sulfanes with five-
membered ring A (2a–e and 3) and that of their analogues with
six-membered rings (4 and 5) does not follow the same kinet-
ics, they will be discussed separately.
The differences in reactivity of acylamino-acyloxy com-
pounds 2–5 relative to the analogous diacyloxy derivatives of
type 1 will be attributed primarily to the ‘disproportion’ of
bond strengths in the highly polarized axial N᎐SIV᎐O array,
which is also reflected in the infrared spectra of compounds
2–5.
Hydrolysis of spiro-ë4-sulfanes 2a–e and 3
Kineticequations. Thehydrolysisreactionsof compounds2a–e
and 3 were followed by UV-spectrophotometric methods in
dioxane–water containing 10–60% v/v of water at 25 ЊC (for
other conditions see Experimental). As water was used in great
excess compared to the spiro-λ4-sulfanes, the hydrolysis fol-
lowed pseudo-first-order kinetics according to rate eqn. (1).
Results and discussion
IR spectra of ë4-sulfanes 2–5
In an earlier paper, Livant and Martin 8 reported that the car-
bonyl frequencies of acyloxy-λ4-sulfanes of type 10 are very
sensitive to the nature of the axial ligand X and thus indicative
of the charge distribution along the highly polarizable O᎐SIV᎐X
axial bond system in λ4-sulfanes. Depending on whether the
structure 10a or 10b makes a higher contribution, the carbonyl
rate = k[spiro-λ4-sulfane]
(1)
The pseudo-first-order rate constants, k2a for 2a and k3 for 3,
are listed in Table 2, indicating that 3 is more reactive toward
hydrolysis in neutral media than 2a. Rate constant k2a showed a
linear dependence on the molar concentration of water, eqn. (2)
O
O
O
-
O
•
O
•
O
+
•
+
•
•
•
S
•
S
S
•
k2a = Ϫ3.47 × 10Ϫ7 ϩ 4.35 × 10Ϫ7 [H2O]
(2)
Ph
Ph
Ph
-
•
•
X
X
X
being valid for reaction mixtures containing up to 45% v/v of
water. The intercept of this equation can be considered as zero
within experimental error, while its slope gives the second-order
rate constant (in dm3 molϪ1 sϪ1) of the hydrolysis.
10
10b
10a
frequencies will range from ~1830 cmϪ1 (a value even higher
than usual for γ-lactones because of the electron-withdrawing
effect of the sulfonium centre) to ~1600 cmϪ1 or even less (a
value characteristic of a carboxylate group).
Medium effects. Spiro-λ4-sulfanes 2a and 3 showed good lin-
ear relationships between log k and the empirical parameters of
the solvent polarity ET 9,1 and Y10 [see Table 2 and eqns. (3)–(6)].
2062
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
Table 1 Selected X-ray structural data 4,5 and IR carbonyl frequencies (vC᎐O) of spiro-λ4-sulfanes 1a, 2a and 3–5
᎐
Axial bond lengths/Å
vC᎐O/cmϪ1
᎐
Compound
S᎐O
1.84
2.132
S᎐N
—
Solvent
Acyloxy
Acylamino
1a
2a
Acetonitrile
Dimethyl sulfoxide
Acetonitrile
1729
1722
1666
—
—
1706
1.734
Dimethyl sulfoxide
50:50 (v/v) Dioxane–D2O
Acetonitrile
Dimethyl sulfoxide
50:50 (v/v) Dioxane–D2O
Acetonitrile
Dimethyl sulfoxide
50:50 (v/v) Dioxane–D2O
Acetonitrile
1663, 1650
1631
1640
1634
1629, 1623
1644
1641
1623, 1611
1627
1702
1702
1705
1701
1698
1671
1666
1679, 1661
1674
3
4
5
2.130
2.247
2.598a
1.734
1.710
1.694a
Dimethyl sulfoxide
50:50 (v/v) Dioxane–D2O
1625
1580
1670
1692, 1676
a
Data obtained for the 1:1 molecular complex of 5 with methanol (see ref. 4).
Table 2 Pseudo-first-order rate constants k2a and k3 for the hydrolysis
of spiro-λ4-sulfanes 2a and 3, respectively, in different dioxane–water
mixtures at 25 ЊC, and parameters of the solvent polarity ET1 and Y10
H2O (% v/v)
k2a/10Ϫ5 sϪ1
k3/10Ϫ5 sϪ1
ET/kcal molϪ1 b
Y
10
20
20a
30
40
45
50
50a
55
60
0.283
0.516
0.156
0.742
1.003
1.126
1.320c
0.359
1.442
1.607
2.055
3.147
—
3.807
4.914
—
6.574
—
—
55.4
57.6
—
59.3
60.6
61.2
62.1
—
Ϫ2.030
Ϫ0.833
—
0.013
0.715
—
1.361
—
—
62.9
63.5
9.006
1.945
a
Determined in dioxane–D2O mixtures. b 1 cal = 4.184 J. c In the pres-
ence of 0.1 mol dmϪ3 LiClO4 k2a 1.175 × 10Ϫ5 sϪ1 was found.
log k2a = Ϫ10.629 ϩ 0.0924 ET (r 0.9939)
log k2a = Ϫ5.143 ϩ 0.1901 Y (r 0.9984)
log k3 = Ϫ8.978 ϩ 0.0773 ET (r 0.9958)
log k3 = Ϫ4.388 ϩ 0.1555 Y (r 0.9919)
(3)
(4)
(5)
(6)
The slopes of the above functions are only about a quarter of
the similar values obtained earlier for diaryl(diacyloxy)spiro-
λ4-sulfane 1a,1 which indicates that the hydrolysis of (acyl-
amino)(acyloxy)spiro-λ4-sulfanes 2a and 3 is less sensitive to
the change in the ionizing power of the solvent. In the case of
2a the rate enhancement caused by the increase in water con-
centration may be attributed to a greater nucleophile concentra-
tion rather than to an effect of solvent polarity which becomes
significant only at higher (>50% v/v) water concentrations.
Similarly, the rate of hydrolysis of 2a is not enhanced by
increasing the ionic strength (Table 2). Although the k3 versus
[H2O] relationship cannot be regarded as linear, eqns. (5) and (6)
are indicative of a weak solvent effect in the case of spiro-λ4-
sulfane 3 as well.
Fig. 1 Dependence of the hydrolysis rate constants k of spiro-λ4-
sulfanes 2a (ᮀ), 3 (᭺), 4 (᭿) and 5 (᭹) on the concentration of HClO4
[in 50:50 (v/v) dioxane–water at 25 ЊC for 2a and 3; in water at 50 ЊC for
4–5]. Solid lines indicate values calculated from eqn. (11).
Acid catalysis. The acid-catalysed hydrolysis of spiro-λ4-
sulfanes 2a–e and 3 was carried out at 25 ЊC in 50:50 (v/v)
dioxane–water containing strong acids, e.g. perchloric acid
(HClO4) or toluene-p-sulfonic acid (TsOH) in different concen-
trations. As shown in Fig. 1, reactions are promoted by acid
catalysis and the relative reactivities of compounds 2a and 3
depend on the concentration of acid (at a higher acid concen-
tration 2a proved to be more reactive than 3).
Substituent effects. In contrast with diaryl(diacyloxy)spiro-λ4-
sulfanes (1a–e),1 the hydrolysis of diaryl(acylamino)(acyloxy)-
spiro-λ4-sulfanes (2a–e) is promoted by electron-withdrawing
and retarded by electron-releasing substituents. Data in Table 3
fit the Hammett equation [eqn. (7)] very well (see σ values in
log k2 = ρσ ϩ constant
(7)
In the (0.2–1) × 10Ϫ2 mol dmϪ3 domain the measured
pseudo-first-order rate constants (k) for 2a–e are linear func-
tions of the acid concentration [eqn. (8)]. The k0 rate constants
ref. 11) giving a reaction constant ρ 1.43 (r 0.9982) which has
the opposite sign and is considerably greater in absolute value
than the reaction constant found for diacyloxy analogues with
ρ Ϫ0.52.
k = k0 ϩ kcat[acid]
(8)
J. Chem. Soc., Perkin Trans. 2, 1997
2063

View Article Online
Table 3 Rate constants k2 for the hydrolysis of compounds 2a–e in dioxane–water mixtures at 25 ЊC, and parameters k0 and kcat calculated from
eqn. (8)
Compound
H2O (% v/v)
k2/10Ϫ5 sϪ1
Acid a
k0/10Ϫ5 sϪ1
kcat/10Ϫ2 dm3 molϪ1 sϪ1
r
2a
2a
2a
2a
2a
2b
2c
2d
2e
50
50
20
50b
20b
50
50
50
50
1.320
HClO4
TsOH
HClO4
TsOH
1.234
1.205
0.968
0.385
0.413
27.10
4.000
0.877
0.406
1.059
1.048
0.791
0.596
0.749
4.510
1.526
0.703
0.461
0.9997
0.9998
0.9995
0.9987
0.9991
0.9997
0.9984
0.9997
0.9959
TsOH
20.05
2.911
0.827
0.663
HClO4
HClO4
HClO4
HClO4
a
Concentration range: (0.2–1.0) × 10Ϫ2 mol dmϪ3 b Measured in dioxane–D2O.
.
obtained from the linear equations can be regarded as practic-
ally identical (within experimental error) with the k2 constants
measured in neutral dioxane–water (Table 3). The susceptibil-
ity for acid catalysis of (acylamino)(acyloxy)spiro-λ4-sulfanes
2a–e is, however, far less pronounced than that of (diacyl-
oxy)spiro-λ4-sulfanes 1a–e. The catalytic constant measured
for 2a in 80:20 (v/v) dioxane–water (Table 3) is about 3000
times lower than that for its diacyloxy analogue 1a.1
sulfanes (i.e. carboxy-substituted sulfonium salts; see 11 in
Scheme 2) are significantly higher than those of spiro-λ4-
O
HO
H2O
Oδ–
C
O
B
B
+H3O+
–H3O+
Sδ+
S+
A
A
Similarly to noncatalysed hydrolysis, the reactions carried
out in acidic media were assisted by electron-withdrawing sub-
stituents but the substituent effect was weaker in this case. By
substituting the catalytic constants kcat (Table 3) and the
appropriate σ values11 into eqn. (7) a reaction constant ρcat 0.90
(r 0.9926) was obtained for the acid-catalysed reaction. The
decrease in ρ values (ρ 1.43 for noncatalysed hydrolysis) may
be explained by the facts that in the latter case protonation is
aided by electron-releasing, while nucleophilic displacement by
electron-withdrawing substituents.
N
N
Me
Me
O
O
2a-C
11
H2O slow
H2O slow
H2O
O
H
H
-
OH
O
S+
HO2C OH
S+
By raising the concentration of acid to 0.1 mol dmϪ3, the k
versus [HClO4] plot for 2a curves slightly downwards (Fig. 1).
Compound 3 gave a similar function in the (0.1–1) × 10Ϫ2 mol
dmϪ3 HClO4 concentration range (Fig. 1). By assuming that the
highly polarized acyloxy group in spiro-λ4-sulfane (S) is proton-
ated in strong acids and the resulting carboxy-substituted sulf-
onium cation (SϩH) reacts faster than the spiro-λ4-sulfane, a
reaction pathway with two parallel steps can be written for the
hydrolysis, as shown in Scheme 1.
A
A
N
N
Me
Me
O
O
12
13
HO2C OHδ–
–H3O+
S δ++
A
Nδ–
Me
O
Ka
S+H
S + H+
14
fast
kS+H
kS
H2O
H2O
sulfoxide
CO2H
Scheme 1
S
Starting from Scheme 1, eqns. (9) and (10) can be derived,
O
CONHMe
kexp[S]st = kS H[SϩH] ϩ kS[S]
(9)
؉
6a
[Hϩ]
[Hϩ] ϩ Ka
Ka
Scheme 2
(10)
؉H
kexp[S]st = kS
[S]st ϩ kS
[S]st
[Hϩ] ϩ Ka
sulfanes with a polar, sulfonium-carboxylate-like structure
؉H
(2a-C) (kS > kS), and the protonated form of 2a is more
reactive than that of 3.
where [S]st = [S] ϩ [SϩH] denotes the stoichiometric concentra-
tion of spiro-λ4-sulfane. The experimental rate constants are
Acid dissociation constants Ka calculated for protonated
spiro-λ4-sulfanes from eqns. (9)–(11) are rather high. The
strength of the sulfonium carboxylic acid is increased not only
by the electron-withdrawing effect of the sulfonium centre but
also by the through-space polar effect of the sulfonium group in
the ortho position, especially in the case of protonated 2a,
in which a sterically favourable five-membered spiro-ring is
performed.
؉
described well by eqn. (11); the calculated kS H, kS and Ka
constants are given in Table 4.
kS H[Hϩ] ϩ kSKa
؉
(11)
kexp
=
[Hϩ] ϩ Ka
The rate constants kS related to the hydrolysis of the nonpro-
tonated spiro-λ4-sulfanes 2a and 3 are in good agreement with
the rate constants k2a and k3 obtained for these compounds in
Deuterium solvent isotope effect. The isotope effect in the
noncatalysed hydrolysis of spiro-λ4-sulfane 2a was studied in
50:50 and 80:20 (v/v) dioxane–H2O(D2O) mixtures, and
kH O/kD values of 3.68 and 3.31 were obtained, respectively
؉H
neutral 50:50 (v/v) dioxane–water (cf. Table 2). The kS data
indicate that the hydrolysis rates of protonated spiro-λ4-
2O
2
2064
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
Table 4 kS H , kS and Ka constants for the hydrolysis of spiro-λ4-sulfanes 2a and 3–5 in acidic media, calculated from eqn. (11)
؉
kS H /10Ϫ4 sϪ1
kS/10Ϫ5 sϪ1
Ka/mol dmϪ3
Solvent a
T /ЊC
؉
Compound
2a
3
4
49.7
9.94
0.521
1.82
1.22
6.51
0.43
2.62
45.6
1.5
2.6
3.1
50:50 (v/v) Dioxane–water
50:50 (v/v) Dioxane–water
Water
25
25
50
50
5
Water
a
The [HClO4] range was 0–0.1 mol dmϪ3 for 2a, 0–0.01 mol dmϪ3 for 3 and 0.02–0.1 mol dmϪ3 for 4–5.
with a slope close to unity (Fig. 2); e.g. at pH 7 k4 0.105 sϪ1 and
k5 0.0102 sϪ1
.
The above findings suggest that the role of OHϪ ions as
nucleophiles cannot be neglected in describing the hydrolysis of
4 and 5. In the absence of a buffer the deviation from pseudo-
first-order kinetics at the early stage of the reaction may be
ascribed to the slight pH shift caused by the acid dissociation of
the carboxy-substituted sulfoxide products (8 and 9, respect-
ively) decreasing the concentration of OHϪ ions.
As is expected, the kinetics become pseudo-first-order in
solutions containing a strong acid such as HClO4 because the
concentration of OHϪ ions as well as the acidity resulting from
the products is not significant in this case. The character of the
k versus [acid] plots obtained for compounds 4 and 5 in aque-
ous HClO4 solutions at 50 ЊC is the same as in the case of 3
(see Fig. 1). Rate and equilibrium constants calculated from
eqn. (11) are listed in Table 4. The rate constants for zero acid
concentration, which were estimated from the linear portion of
the ln[(A Ϫ A∞)/(A0 Ϫ A∞)] versus t plots obtained in pure water
at 50 ЊC (1.9 × 10Ϫ4 sϪ1 for 4 and 5.13 × 10Ϫ5 sϪ1 for 5), were not
included in the calculation because they were significantly
higher than expected, likely due to the participation of OHϪ
ions in the reaction. Comparing the rate constants we found
that compound 4 is more reactive in neutral solutions (Fig. 2),
whereas compound 5 shows a higher reactivity in acidic media
(Fig. 1 and Table 4). We may suppose that water molecules
take part as nucleophiles in the hydrolysis of compounds 4 and
5 in acidic media while in neutral solutions OHϪ ions take
part.
Fig. 2 pH dependence of log k for spiro-λ4-sulfanes 4 (᭿) and 5 (᭹) in
aqueous buffer solutions at 50 ЊC
Mechanism
From X-ray, IR and kinetic data one may draw some conclu-
sions for the hydrolysis of diaryl(acylamino)(acyloxy)spiro-λ4-
sulfanes 2–5; reaction pathways proposed for the noncatalytic
and acid-catalysed cases are shown in Scheme 2 for compound
2a.
(cf. Table 2). Data indicate that the deuterium solvent isotope
effect does not depend essentially on the composition of the
solvent mixture. For the catalytic constants of 2a isotope effects
of kH O/kD 1.76 and 1.65, respectively, were found in the
2O
(i) Spiro-λ4-sulfanes with a major contribution of a limiting
structure such as 2a-B exhibit long (2.13–2.25 Å)4 and highly
polarized S᎐O bonds which may be classified as weak hyper-
valent bonds.† Because ring B with an acyloxy part is closed in
these compounds only by a weak S - - - O bond (see 2a-C), the
opening of this ring, which pertains to the nucleophilic attack
of water on sulfur, should occur very easily and so cannot be
rate-determining. This is why the water content of the solvent
influences the rate of hydrolysis only to the extent of its molar
concentration, and the ionic strength of the medium has
practically no effect on the reaction rate.
2
same solvents containing TsOH (cf. Table 3).
Hydrolysis of spiro-ë4-sulfanes 4 and 5
The hydrolysis of compounds 4 and 5 did not follow pseudo-
first-order kinetics in 50:50 (v/v) dioxane–water mixture at
25 ЊC throughout the whole range of measurements; the
ln[(A Ϫ A∞)/(A0 Ϫ A∞)] versus t plots were not linear in the init-
ial stage of the reaction. From the second, almost linear parts
of the plots the approximate values of k4 9.89 × 10Ϫ7 sϪ1 and k5
8.86 × 10Ϫ7 sϪ1 were calculated for the hydrolysis of 4 and 5,
respectively. Data indicate that compounds 4 and 5, in which
the acylamino part is incorporated in a six-membered ring A,
are considerably less reactive than the 2a–e and 3 analogues
with a five-membered ring. We assumed that the curvature of
the plots was caused by a shift in pH of the medium therefore
the reactions were carried out in aqueous buffer solutions, too
(for details see Experimental). Under these conditions the reac-
tions became first order at pH 6–7.5. The rate increased with
pH, and the log k versus pH plots [eqns. (12)–(13)] were linear
(ii) From the easy opening of ring B it follows that nucleo-
philic attack of water on the positively charged sulfonium
centre should represent the rate-determining step for the non-
catalysed hydrolysis (2a-C → 12 → 14). The positive reac-
tion constant (ρ 1.43) observed for compounds 2a–e provides
strong evidence for this view. If the ring-opening were rate-
† Axial S᎐O bonds occurring in (diaryl)spiro-λ4-sulfane-type com-
pounds may be classified as follows (cf. ref. 6): S᎐O covalent single bond
(1.66–1.71 Å), usual S - - - O hypervalent bond (1.79–1.83 Å) in sym-
metric structures, weak S - - - O hypervalent bond (1.95–2.25 Å) in
unsymmetric structures and S ؒ ؒ ؒ O close contact (2.35–2.75 Å) signifi-
cantly shorter than the sum of the van der Waals radii (3.25 Å).
log k4 = Ϫ7.784 ϩ 0.972 pH (r 0.9997)
log k5 = Ϫ8.863 ϩ 0.981 pH (r 0.9996)
(12)
(13)
J. Chem. Soc., Perkin Trans. 2, 1997
2065

View Article Online
determining, a negative reaction constant would be obtained, as
was the case for (diacyloxy)spiro-λ4-sulfanes 1a–e with two
equivalent and strong axial O - - - SIV - - - O hypervalent bonds
(ρ Ϫ0.52).1
sulfane and its protonated form that can be observed (see data
for 2a and 3 in Table 4).
Comparing kcat data for compound 2a (Table 3) and those
for 1a (Table 2 in ref. 1) one may conclude that (acylamino)-
(acyloxy)spiro-λ4-sulfanes are less sensitive to acid catalysis
than the diacyloxy analogues, the difference in reactivity may
be attributed to the fact that ring-opening is not a part of the
rate-determining step for acylamino derivatives.
(viii) The decrease of the positive reaction constant value for
the acid-catalysed hydrolysis of compounds 2a–e may be
explained by eqn. (14) showing that ρcat can be divided into two
(iii) The significant role of steric factors for compounds
2–5 is reflected by the different order of dipolar char-
acter (2a ≈ 3 < 4 < 5; see Table 1) and reactivity towards non-
catalysed hydrolysis in neutral, unbuffered solutions (4 ≈ 5 Ӷ
2a < 3). Spiro-λ4-sulfanes with five-membered ring A (2a and
3) are much more reactive than the six-membered analogues
(4 and 5). Nevertheless, the ring-size effect is in agreement
with the suggested rate-determining step because other nucleo-
philic displacements occurring at the sulfonium centre are
known to proceed several orders of magnitude more quickly
in five-membered cyclic derivatives than in six-membered
ones.12,13
(iv) The size of the O-containing ring B also has some rate-
controlling role in the noncatalysed hydrolysis of (acyl-
amino)(acyloxy)spiro-λ4-sulfanes. The greater reactivity of
compound 3 compared to that of 2a may be explained by the
fact that the six-membered ring B in 3 is closed by a weaker
hypervalent S - - - O bond than the five-membered ring in 2a.
(As shown in ref. 6 S ؒ ؒ ؒ O close contact is also less effective in a
six-membered ring than in a five-membered analogue.) Thus the
sulfonium centre in 3 is less shielded from nucleophilic attack
than in 2a. It is worth mentioning that (diacyloxy)spiro-λ4-
sulfanes1 having six-membered spiro-rings B proved also to be
more reactive than those with five-membered rings.
(v) Under similar conditions of hydrolysis (acylamino)-
(acyloxy)spiro-λ4-sulfanes are less reactive than their diacyloxy
analogues (cf. Table 2 and Table 1 in ref. 1). The decrease in
reactivity may be ascribed to the acylamino part in rings A
which is a poorer leaving group than the acyloxy part in rings B
therefore the formation of a monocyclic λ4-sulfane with a
polarized S - - - N hypervalent bond (see 2a-C → 12 → 14
in Scheme 2) is less favourable than the analogous process
involving S - - - O hypervalent bond (see 1 → 7 → 8 in
Scheme 1 of ref. 1).
ρcat = ρe ϩ ρN u
(14)
terms. The negative ρe elementary reaction constant reflects the
fact that the protonation equilibrium is promoted by electron-
releasing substituents. The positive ρN u term is related to the
nucleophilic attack at the sulfonium centre, which is assisted by
electron-withdrawing groups. Because ρe does not overcompen-
sate ρN u, the overall reaction constant ρcat remains positive. It is
worth mentioning that ρcat (Ϫ1.55) is negative for the acid-
catalysed hydrolysis of (diacyloxy)spiro-λ4-sulfanes 1a–e, indi-
cating that the negative ρe term is predominant in this case.
(ix) In the acid-catalysed hydrolysis of 2a the deuterium solv-
ent isotope effect is only about half the value obtained for the
noncatalysed reaction because the primary isotope effect due to
the deprotonation of the water nucleophile (which occurs in
the rate-determining step, see e.g. 11 → 13 → 14) and the
inverse isotope effect of the protonation pre-equilibrium (e.g.
2a-C → 11) are superimposed.
(x) The significant acceleration of the hydrolysis observed for
the less reactive spiro-λ4-sulfanes 4 and 5 in nearly neutral buff-
ered solutions [e.g. at pH 7 k4 0.105 sϪ1 cf. eqn. (12)] as com-
pared with the reaction in distilled water (e.g. k4 1.9 × 10Ϫ4 sϪ1
)
suggests that OHϪ ions also take part in the reaction as nucleo-
philes. It is striking, however, that the order of reactivity
(k4 > k5) observed in buffered solutions differs from that
detected in acidic media (k4 < k5). At this moment we do not
have enough information to explain this phenomenon. It seems
likely, however, that reactions with OHϪ nucleophiles do follow
a reaction pathway different to those with water. In the former
case e.g. an associative mechanism similar to that published by
Martin and Balthazor 14 might be taken into account.
(vi) The deuterium solvent isotope effect measured for the
hydrolysis of 2a is compatible with a primary effect, suggesting
that the nucleophilic attack of the water molecule in the rate-
determining step is accompanied by O᎐H bond cleavage, as
shown in the transition structure 12 (Scheme 2). Other water
molecules or the neighbouring carboxylate group may con-
tribute as proton acceptors. It seems very likely that ring A
with an acylamino part needs a stronger nucleophile than
water to undergo splitting, and such a nucleophile can be
formed from the neutral water molecule by proton abstraction.
This view is supported by the fact that the more reactive
(diacyloxy)spiro-λ4-sulfanes (e.g. 1a) with acyloxy-containing
ring B exhibit significantly weaker deuterium solvent isotope
effect; kH O/kD 3.31 for 2a and 1.661 for 1a in 80:20 (v/v)
(xi) As a result of the nucleophilic attack of H2O or OHϪ
reactants on the central sulfur atom the S᎐N covalent bond in
ring A is markedly weakened (see e.g. the transition states 12
and 13 in Scheme 2). Because the basicity of OHϪ and Ar-CO-
N(Me)Ϫ species do not differ significantly, we may assume that
the hypervalent S - - - N and S - - - O bonds are about equally
as strong in the 14-type intermediates from which the sulfoxide
products (e.g. 6) are formed in fast proton-transfer steps.
(xii) We found in other experiments15 that the hydrolysis of
amidosulfonium salts (e.g. 11 without ortho-carboxy substitu-
ents) is not affected by acid catalysis. This suggests that the
amide moiety in spiro-λ4-sulfanes (ring A) is not protonated in
the rate-determining step of the hydrolysis but only in the sub-
sequent fast proton-transfer steps.
2O
2
dioxane–water.
(vii) The values of catalytic constants kcat (Table 3) show that
the hydrolysis of (acylamino)(acyloxy)spiro-λ4-sulfanes is mod-
erately promoted by acid catalysis. Protonation converts the
starting compounds into a more reactive carboxy-substituted
sulfonium ion (e.g. 2a-C → 11). The enhanced reactivity of
the sulfonium centre in 11-type intermediates as compared to
that in 2a-C-type structures may have two origins. (a) The
ortho-carboxy group formed by protonation has a stronger
electron-withdrawing effect than the negatively charged carb-
oxylate group, which increases the positive polarization and so
the reactivity of the sulfonium centre; (b) the carboxy group
setting up S ؒ ؒ ؒ O close contact (11) shields the sulfonium
centre against the attack of water nucleophile less successfully
than the carboxylate group forming a hypervalent S - - - O
bond which may also increase the rate of hydrolysis. The
stronger the S - - - O hypervalent bond in spiro-λ4-sulfanes,
the greater difference between the reactivity of the spiro-λ4-
Experimental
Materials
The preparation of spiro-λ4-sulfanes 2a, and 3–5 has been pub-
lished earlier.4 The synthesis of 2b–e is described below. Melting
points were determined on a Boëtius micro melting point
1
apparatus. H NMR measurements for 2b–e were made on a
Bruker WP-80SY instrument operating at 80 MHz. Toluene-p-
sulfonic acid (TsOH) was recrystallized from concentrated HCl
solution. Spectroscopic grade anhydrous dioxane used in the
kinetics and IR spectroscopic studies was obtained from ana-
lytical grade dioxane (Reanal, Budapest) by an appropriate
2066
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
purification method.16 Before use, traces of peroxides were
removed by distillation from LiAlH4. For the hydrolysis of 4–5
in aqueous solutions, 4-(2-hydroxyethyl)-1-piperazineethane-
sulfonic acid (HEPES, Serva Feinbiochemica, analytical grade,
0.01 mol dmϪ3, pH 6.83–7.27 at 50 ЊC), 4-morpholineprop-
anesulfonic acid (MOPS, Serva Feinbiochemica, analytical
grade, 0.01 mol dmϪ3, pH 6.43–6.62 at 50 ЊC) and 1,4-piper-
azinebis(ethanesulfonic acid) (PIPES, Merck, 99%, 0.01 mol
dmϪ3, pH 6.08–6.30 at 50 ЊC) were used as buffer substances.
The desired pH values were fixed by titration with NaOH, using
Table 5 Initial concentrations (C0), wavelengths of measurements (λ)
and absorbance values for the starting spiro-λ4-sulfanes 2a–e, 3–5 (A0)
and the corresponding sulfoxide products (A∞)
Spiro-λ4-sulfane
C0/10Ϫ4 mol dmϪ3
λ/nm
A0
A∞
2a a
2ba
2ca
2da
2ea
3a
17.5
6.32
9.38
16.7
7.92
1.49
1.19
1.04
313
345
313
316
323
333
333
345
0.08
0.24
0.12
0.01
0.07
0.92
1.02
0.86
0.81
1.01
0.80
0.39
0.99
0.30
0.04
0.04
4b
a
Beckman Φ32 pH meter with automatic temperature
5b
compensation.
Spiro-ë4-sulfanes 2b–e. To a saturated solution of the corre-
sponding sulfoxide 6b–e (0.1 g) in dry AcOH (1–4 cm3), Ac2O
(0.1 cm3) was added and the mixture was heated at 100 ЊC for
1 h. The solvent was evaporated, the residue was triturated with
diethyl ether, filtered off and dried in vacuo over KOH to give
2-methyl-5Ј-nitrospiro[3H-2,1-benzoxathiole-1,1Ј-3H-1,2-benz-
isothiazole]-3(2H),3Ј-dione [2b (79 mg, 79%): mp 210–222 ЊC;
a
b
In 50:50 (v/v) dioxane–water at 25 ЊC. In aqueous buffer (PIPES)
solution of pH 6.2 at 50 ЊC.
cm3) was added dropwise under continuous stirring in 30 min.
The mixture was stirred for an additional 15 min at the same
temperature, then filtered, and the filtrate was poured onto a
mixture of concentrated HCl solution (400 cm3) and ice. The
precipitate formed was filtered off, washed with water and dried
in vacuo to give 5-nitrothiosalicylic acid (104 g, 67%). Being
highly sensitive to oxidation, the crude product was directly
used in the synthesis of 15b.
νmax(KBr)/cmϪ1 1710vs, 1650vs (C᎐O); δ (80 MHz; [2H ]DMSO)
᎐
H
6
7.4–8.75 (m Ar-H), 3.34 (NCH3)], 5Ј-chloro-2-methylspiro[3H-
2,1-benzoxathiazole-1,1Ј-3H-1,2-benzisothiazole]-3(2H ),3Ј-
dione [2c (84 mg, 84%): mp 254–260 ЊC; νmax(KBr)/cmϪ1 1720vs,
1700vs, 1640vs (C᎐O); δ (80 MHz; CDCl ) 7.0–8.6 (m Ar-H),
᎐
H
3
3.41 (NCH3)], 2,5Ј-dimethylspiro[3H-2,1-benzoxathiole-1,1Ј-
3H-1,2-benzisothiazole]-3(2H),3Ј-dione [2d (92 mg, 92%): mp
264–267 ЊC; νmax(KBr)/cmϪ1 1692vs, 1660vs (C᎐O); δ (80 MHz;
IR spectra
Fourier transform IR spectra of spiro-λ4-sulfanes 1–5 [1–5 mg
cmϪ3 in 50:50 (v/v) dioxane–D2O, acetonitrile and dimethyl
sulfoxide (DMSO)] were obtained with a Bruker IFS-55 spec-
trometer using CaF2 cells (0.05–0.25 mm pathlengths). In the
case of spectra recorded in acetonitrile and DMSO, the contri-
bution of the H᎐O᎐H deformation vibration (due to traces of
water) to the 1600–1700 cmϪ1 spectral region was removed by
subtracting water spectra obtained in the same solvents. The
1800–1500 cmϪ1 region of the spectra was analysed by Fourier
self-deconvolution 20 and decomposed into individual bands by
a Levenberg–Marquardt nonlinear curve-fitting procedure21 by
using weighted sums of Lorentzian and Gaussian functions.
The carbonyl frequencies are listed in Table 1.
᎐
H
CDCl3) 7.0–8.6 (m Ar-H), 3.38 (NCH3), 2.47 (5Ј-CH3)] and
5Ј-methoxy-2-methylspiro[3H-2,1-benzoxathiole-1,1Ј-1,2-benz-
isothiazole]-3(2H),3Ј-dione [2e (80 mg, 80%): mp 225–227 ЊC;
νmax(KBr)/cmϪ1 1710vs, 1650vs (C᎐O); δ (80 MHz; CDCl ) 7.0–
᎐
H
3
8.6 (m Ar-H), 3.91 (5Ј-OCH3), 3.38 (NCH3)], respectively.
5-Substituted 2-[2Ј-(N-methylcarbamoyl)phenylsulfinyl]benz-
oic acids [R = NO2 (6b), Cl (6c), Me (6d), MeO (6e)]. The corre-
sponding sulfide 15b–e (1.9 mmol) was dissolved in pyridine (5.7
cm3) and then water (1 cm3) and phenyltrimethylammonium
tribromide (0.75 g, 2 mmol) were added to the solution. The
mixture was stirred at 20 ЊC for 1 h and finally poured into 1 
H2SO4 (57 cm3). After standing at 0 ЊC for 1 h the precipitate
formed was filtered off, washed with water and dried to give 6b
[yield 76%; mp 218–220 ЊC (AcOH–H2O); νmax(KBr)/cmϪ1
Kinetics
Kinetic measurements were performed with a Specord 40M
(Zeiss, Jena) and a Varian Cary3E UV–VIS spectrophotometer
in dioxane–water (10–60% v/v water) at 25 ЊC (spiro-λ4-sulfanes
2–3) and in aqueous buffer or HClO4 solutions at 50 ЊC (spiro-
λ4-sulfanes 4–5), by using quartz cells of 1 cm (initial concen-
trations, wavelengths and absorbance values for the starting
spiro-λ4-sulfanes and the sulfoxide products are listed in Table
5). The reactions performed in organic–aqueous solutions were
started by mixing quickly aliquots of solutions of spiro-λ4-
sulfane in dioxane with adequate quantities of water, D2O,
aqueous LiClO4, TsOH or HClO4 solutions, or with solutions of
TsOH in D2O, as required. The reactions performed in aqueous
media were started by adding small quantities (ca. 1% v/v) of
concentrated stock solutions of spiro-λ4-sulfane in DMSO to
the aqueous buffer or HClO4 solution, directly into the stirred
measurement cells. The initial concentration of the spiro-λ4-
sulfanes was (1.0–17.5) × 10Ϫ4 mol dmϪ3. The effect of the ionic
strength was studied in the presence of LiClO4 (0.1 mol dmϪ3),
whereas acid catalysis was investigated by using toluene-p-
sulfonic acid or HClO4 (1.0 × 10Ϫ3–0.1 mol dmϪ3). Relative
errors of the rate constants were less than ±3%.
3395s (NH), 3100–2200br (OH), 1680s, 1648vs (C᎐O), 970vs
᎐
(S᎐O)], 6c [yield 87%; mp 134–138 ЊC (AcOH–H O); νmax(KBr)/
᎐
2
cmϪ1 3450s (NH), 3100–2100br (OH), 1695vs, 1642vs (C᎐O),
᎐
1015vs (S᎐O)], 6d [yield 66%; mp 213–214 ЊC (AcOH–H O);
᎐
2
νmax(KBr)/cmϪ1 3260s (NH), 3150–2300br (OH), 1680vs,
1650vs (C᎐O), 1030vs (S᎐O)] and 6e [yield 84%; mp 125–127 ЊC
᎐
᎐
(AcOH–H2O); νmax(KBr)/cmϪ1 3400s (NH), 3100–2100br (OH),
1690s, 1655vs (C᎐O), 1000vs (S᎐O)], respectively.
᎐
᎐
5-Substituted 2-[2Ј-(N-methylcarbamoyl)phenylthio]benzoic
acids [R = NO2 (15b), Cl (15c), Me (15d), MeO (15e)]. Sulfides
15b–e were prepared from the corresponding 5-substituted
thiosalicylic acid and 2-iodo-N-methylbenzamide as described
in ref. 4 for 2-[2Ј-(N-methylcarbamoyl)phenylthio]benzoic acid
(15a). Selected data: 15b [yield 48%; mp 209–212 ЊC (EtOH);
νmax(KBr)/cmϪ1 3370s (NH), 3200–2300br (OH), 1713s, 1625vs
(C᎐O)], 15c [yield 87%; mp 88–89 ЊC (AcOH–H O); νmax(KBr)/
᎐
2
cmϪ1 3330s (NH), 3450–2300br (OH), 1695s, 1620vs (C᎐O)], 15d
᎐
[yield 46%; mp 163–169 ЊC (CH2Cl2); νmax(KBr)/cmϪ1 3330s
(NH), 3150–2300br (OH), 1700s, 1612s (C᎐O)], 15c [yield 60%;
᎐
mp 173–175 ЊC (EtOH); νmax(KBr)/cmϪ1 3340s (NH), 3200–
2300br (OH), 1722s, 1700s, 1613vs (C᎐O)]. 5-Methoxy-
᎐
thiosalicylic acid and 5-methylthiosalicylic acid were prepared
from 2-amino-5-methoxybenzoic acid 17 and 2-amino-5-
methylbenzoic acid (Aldrich, 97%), respectively, by analogy to
5-chlorothiosalicylic acid.18
5-Nitrothiosalicylic acid. A mixture of finely powdered 2,2Ј-
dithiobis(5-nitrobenzoic acid)19 (155 g, 0.39 mol) and -glucose
(108 g, 0.56 mol) in 96% EtOH (500 cm3) was heated to 78–
80 ЊC, and a solution of NaOH (85 g, 2.12 mol) in water (160
Acknowledgements
This work was financed by the Hungarian Scientific Research
Foundation (OTKA, Nos. T017187 and T017233) and the
Eötvös University Graduate School (Budapest). One of the
authors (E. V.) is grateful for the grant received from the ‘Foun-
dation for Hungarian Science’ supported by the Hungarian
Credit Bank.
J. Chem. Soc., Perkin Trans. 2, 1997
2067

View Article Online
10 A. H. Fainberg and S. Winstein, J. Am. Chem. Soc., 1956, 78, 2770.
References
11 C. Hansch, A. Leo and R. V. Taft, Chem. Rev., 1991, 91, 165.
12 S. Tamagaki, M. Mizuno, H. Yoshida, M. Hirota and S. Oae, Bull.
Chem. Soc. Jpn., 1971, 44, 2456.
1 E. Vass, F. Ruff, I. Kapovits and D. Szabó, J. Chem. Soc., Perkin
Trans. 2, 1993, 855.
2 I. Kapovits, J. Rábai, F. Ruff and Á. Kucsman, Tetrahedron, 1979,
35, 1875.
13 R. Curci, F. Di Furia, A. Levi and G. Scorrano, J. Chem. Soc.,
Perkin Trans. 2, 1975, 408.
3 K. Ohkata, M. Ohsugi, T. Kuwati, K. Yamamoto and K. Akiba,
Tetrahedron Lett., 1990, 31, 1605.
4 D. Szabó, I. Kapovits, Á. Kucsman, P. Huszthy, Gy. Argay, M.
Czugler, V. Fülöp, A. Kálmán, T. Koritsánszky and L. Párkányi,
J. Mol. Struct., 1993, 300, 23.
5 (a) A. Kálmán, K. Sasvári and I. Kapovits, Acta Crystallogr., Sect.
B, 1973, 29, 355; (b) I. Kapovits, J. Rábai, D. Szabó, K. Czakó,
Á. Kucsman, Gy. Argay, V. Fülöp, A. Kálmán, T. Koritsánszky and
L. Pákányi, J. Chem. Soc., Perkin Trans. 2, 1993, 847.
6 Á. Kucsman and I. Kapovits, Nonbonded Sulfur-Oxygen Intercation
in Organic Sulfur Compounds, in Studies in Organic Chemistry 19.
Organic Sulfur Chemistry: Theoretical and Experimental Advances,
ed. F. Bernardi, I. G. Csizmadia and A. Mangini, Elsevier,
Amsterdam, 1985, pp. 191–245.
14 J. C. Martin and T. M. Balthazor, J. Am. Chem. Soc., 1977, 99, 153.
15 T. Ádám, F. Ruff, D. Szabó, I. Kapovits and Á. Kucsman, submitted
for publication in J. Chem. Soc., Perkin Trans. 2.
16 D. D. Perrin, L. F. Armarego and D. R. Perrin, Purification of
Laboratory Chemicals, 2nd edn., Pergamon Press, Oxford, 1980,
p. 235.
17 P. Friedländer, Chem. Ber., 1916, 49, 963.
18 L. Katz, L. S. Karger, W. Schroeder and M. S. Cohen, J. Org. Chem.,
1953, 18, 1380.
19 R. Ponci, A. Baruffini and F. Gialdi, Farmaco (Pavia) Ed. Sci., 1963,
18, 732; Chem. Abstr., 1964, 60, 2919h.
20 H. H. Mantsch, D. J. Moffatt and H. L. Casal, J. Mol. Struct., 1988,
173, 285.
21 D. W. Marquardt, J. Soc. Ind. Appl. Math., 1963, 11, 431.
7 J. G. Ángyán, R. A. Poirier, Á. Kucsman and I. G. Csizmadia,
J. Am. Chem. Soc., 1987, 109, 2237.
8 P. D. Livant and J. C. Martin, J. Am. Chem. Soc., 1977, 99, 5761.
9 K. Dimroth, C. Reichardt, T. Siepmann and F. Bohlmann, Liebigs
Ann. Chem., 1963, 661, 1; C. Reichardt, Liebigs Ann. Chem., 1971,
756, 64.
Paper 7/01002H
Received 11th February 1997
Accepted 16th June 1997
2068
J. Chem. Soc., Perkin Trans. 2, 1997