Reactivity of ions and ion pairs in the nucleophilic substitution reaction on methyl p-nitrobenzenesulfonate

Article abstract of DOI:10.1039/a802954g

Nucleophilic substitution on methyl p-nitrobenzenesulfonate has been studied with a series of chloride salts with different structures and solvations: Bu₄NCl, PPNCl or KCl complexed by 18-crown-6 or Kryptofix 2,2,2 as well as with a bromide salt (PPNBr) for comparison purposes. The rate constants and the activation parameters are in accordance with an S_N2 mechanism. The treatment of the data, following the Acree eqaution, shows that the process takes place by way of two reaction paths: the first, due to the chloride ion, has the same rate for all the salts, while the second slower path, involving the ion pair, has a rate related to the dissociation constant of the salts.

Full text of DOI:10.1039/a802954g

View Article Online / Journal Homepage / Table of Contents for this issue
Reactivity of ions and ion pairs in the nucleophilic substitution
reaction on methyl p-nitrobenzenesulfonate
Sergio Alunni,* Agostino Pero and Gustavo Reichenbach*
Dipartimento di Chimica, Università di Perugia, 06100 Perugia, Italy
Nucleophilic substitution on methyl p-nitrobenzenesulfonate has been studied with a series of chloride
salts with different structures and solvations: Bu4NCl, PPNCl or KCl complexed by 18-crown-6 or
Kryptofix 2,2,2 as well as with a bromide salt (PPNBr) for comparison purposes. The rate constants and
the activation parameters are in accordance with an SN2 mechanism. The treatment of the data, following
the Acree equation, shows that the process takes place by way of two reaction paths: the first, due to the
chloride ion, has the same rate for all the salts, while the second slower path, involving the ion pair, has a
rate related to the dissociation constant of the salts.
The importance of ion pairs in chemistry is well recognized.1
Depending on the reacting species, free ions or ion pairs, several
properties of a process are affected, such as reactivity, selectiv-
ity and stereoselectivity. To check the relative reactivity of ions
and ion pairs, we studied the nucleophilic substitution reaction
on a common substrate,2 methyl p-nitrobenzenesulfonate (1) in
CH2Cl2 with a series of salts with the same anion, but with
cations differing in structure and solvation, in order to assess
the kinetic behaviour of a series of ion pairs. Studies to assess
the relative reactivities of free ions and ion pairs in SN2 reac-
tions have been reported in the literature,3 but, to our know-
ledge, no systematic analysis of the reactivity of ion pairs with
different structures is available.
The salts we used were tetrabutylammonium chloride, bis-
(triphenylphosphoranylidene)ammonium chloride (PPNCl),
KCl–18-crown-6 (KCl–crown) or KCl–Kryptofix 2,2,2 (KCl–
Krypto). In order to compare the reactivity of a different anion,
PPNBr was also used. The kinetic data were treated with the
classical Acree equation.4
Table 1 Observed first order rate constants for Bu4NCl in CH2Cl2 at
22 ЊC, unless otherwise stated
[Bu4NCl]/10Ϫ3 
α
f
kobs/10Ϫ3 sϪ1
0.83
2.90
3.97
5.80
6.54
7.53
10.2
11.6
21.5
0.178
0.116
0.106
0.095
0.092
0.088
0.082
0.080
0.072
0.070
0.135
0.106
0.095
0.083
0.077
0.075
0.075
0.075
0.169
0.139
0.123
0.111
0.110
0.106
0.106
0.108
0.111
0.703
0.596
0.565
0.524
0.511
0.495
0.459
0.443
0.362
0.269
0.620
0.541
0.495
0.423
0.364
0.316
0.284
0.257
0.588
0.514
0.454
0.378
0.364
0.316
0.265
0.233
0.206
0.19
0.54
0.66
0.97
1.097
1.10
1.51
1.75
3.03
5.50
0.16
0.29
0.38
0.71
0.95
1.49
1.59
2.05
0.61
0.95
1.13
1.72
2.14
2.78
4.04
4.59
5.10
41.8
2.11a
4.67a
7.00a
12.7a
19.8a
28.1a
35.0a
42.2a
2.11b
4.22b
7.00b
12.7b
14.00b
19.80b
28.10b
35.00b
42.20b
Results and discussion
Pseudo-first-order rate constants (kobs/sϪ1) were measured for
the nucleophilic substitution reactions on methyl p-nitro-
benzenesulfonate induced by various chloride salts, in which the
nucleophile can be both the free ion (ClϪ) and the ion pair
(AϩClϪ), at 22 ЊC in CH2Cl2. The relative data are reported in
Tables 1–4.
a At 12 ЊC. b At 1 ЊC.
The persistence of isosbestic points at 266 nm for two to three
half lives suggests that no intermediates of spectrally significant
concentration accumulated under these conditions. The process
is shown in Scheme 1.
All the reactions were followed in dichloromethane, a very
good solvent for organic and organometallic compounds
because of its very low permittivity (ε = 9.08),5 very low donor
number (DN = 2) and good acceptor number (AN = 20.4).6
While alkali halides and Me4NCl are insoluble in this solvent,
salts with large aliphatic or aromatic cations are soluble, but a
large fraction will exist as ion pairs.7,8
O
S
O2N
OCH3
O
Of the four salts used as a source of chloride, Bu4NCl was
chosen because it is a simple aliphatic salt whose interionic
distance (ca. 4.9 Å) is smaller than the Stokes radius (ca. 5.6 Å)9
and is supposed to form tight ion pairs with a partial compen-
etration of the anion between the aliphatic chains.10 The dis-
sociation constants are 1.6 ( 0.2) × 10Ϫ5, 1.7 ( 0.9) × 10Ϫ5 and
O
O
O –
O2N
S
O –
O2N
S
O
O
+ CH3Cl
+ CH3Cl + A +
Scheme
1
Nucleophilic substitution reaction paths for methyl
p-nitrobenzenesulfonate induced by the free chloride ion (ClϪ) or the
* E-Mail: ORMETAL@UNIPG.IT
ion pair (AϩClϪ)
J. Chem. Soc., Perkin Trans. 2, 1998
1747

View Article Online
Table 2 Observed first order rate constants for PPNCl in CH2Cl2 at
22 ЊC
form solvent-separated ion pairs. Some authors claim that the
salt is ‘highly dissociated’;13 but they did not report the dissoci-
ation constant; we measured it and found Kdiss = 34.1 × 10Ϫ5 ,
which corresponds to a degree of dissociation of about 0.40 at a
concentration of 0.01 . Thus this salt, too, gives both ions and
ion pairs.
[PPNCl]/10Ϫ3 
α
f
kobs/10Ϫ3 sϪ1
0.30
0.30
0.59
0.65
0.89
1.28
3.18
3.18
6.37
6.47
7.03
8.80
9.42
11.67
14.90
17.80
0.767
0.767
0.695
0.686
0.654
0.617
0.533
0.533
0.478
0.477
0.471
0.455
0.451
0.436
0.421
0.409
0.672
0.672
0.601
0.592
0.558
0.518
0.420
0.420
0.349
0.348
0.340
0.319
0.313
0.294
0.274
0.260
0.16
0.16
0.38
0.38
0.60
0.70
1.56
1.52
2.72
2.92
2.90
4.10
3.87
4.77
6.07
7.59
The adduct KCl–18-crown-6 was used on the assumption
that the anion, in the presence of the crown ether, is ‘naked’ and
unencumbered by strong solvation forces, and is therefore very
reactive.14 Even though the potassium is surrounded by the
receptor, it is still accessible from above and below the crown
ether plane.4d,15 It is then tightly connected to the chloride
anion with a very short interionic distance (even shorter than in
crystalline KBr);16 the dissociation constant is rather low
(1.87 × 10Ϫ5 ). Consequently, in this case too, the salt is mainly
in the form of tight ion pairs. With the hope of obtaining a
naked anion, the Kryptofix 2,2,2 was used, because it surrounds
the potassium ion with two cycles (like a football),17 which
should leave the anion completely free. In fact, the distance
between Kϩ and ClϪ is rather high (ca. 6.5 Å)1f,4d,18,19 but they
still form ion pairs with a dissociation constant of 26.1 ×
10Ϫ5 .
A problem arising with such salts in low permittivity solvents
is that the dissociation constants (Kdiss) are measured at low
concentration (10Ϫ5–10Ϫ4 ), while the kinetics are carried out
at higher concentrations (10Ϫ3–10Ϫ2 ). To obtain an accurate
value of the ion and ion pair activities, the degree of dissoci-
ation of the salt (alpha) and the activity coefficients (f ) of
the free ion were calculated at each concentration by the
method of successive approximations using the extended Debye
Huckel expression.20 This treatment assumes that the activity
coefficient of the ion pairs is equal to one, owing to its neutral
nature.
Table 3 Observed first order rate constants for KCl–crowna in CH2Cl2
at 22 ЊC
[KCl]/10Ϫ3 
α
f
kobs/10Ϫ3 sϪ1
1.46
1.62
3.07
4.04
4.09
4.47
7.20
7.30
8.07
12.00
13.40
13.80
15.40
16.00
20.20
22.00
22.67
28.00
30.40
30.80
0.163
0.157
0.129
0.120
0.120
0.117
0.105
0.105
0.103
0.096
0.095
0.095
0.094
0.094
0.094
0.094
0.095
0.097
0.099
0.099
0.637
0.627
0.562
0.531
0.530
0.520
0.460
0.458
0.445
0.389
0.373
0.368
0.351
0.345
0.308
0.294
0.289
0.253
0.238
0.236
0.21
0.21
0.36
0.49
0.49
0.56
0.76
0.83
0.92
1.25
1.38
1.51
1.59
1.58
1.84
1.95
1.98
2.50
2.53
2.86
Since we observed a decrease in the kobs/[salt] versus [salt]
plot, we assumed that the reaction takes place through two dif-
ferent paths: one due to the free ion and the other due to the ion
pair (Scheme 2). So, the kinetic data were treated following the
(A+Cl –)
+ S
kip
A+ + Cl –
+ S
ki
P
a The ratio crown:KCl was 1.7/1, see Experimental section.
P
Table 4 Observed first order rate constants for KCl–Kryptoa in
CH2Cl2 at 22 ЊC
Scheme 2 Representation of reaction paths, one due to free ion at rate
ki, the other due to the ion pair at rate kip
[KCl]/10Ϫ3 
α
f
kobs/10Ϫ3 sϪ1
Acree equation [eqn. (1)].4 We chose to use the equation in the
0.36
0.37
0.72
1.11
3.72
5.21
5.55
7.47
11.1
14.9
22.4
0.713
0.711
0.643
0.603
0.526
0.515
0.513
0.509
0.511
0.518
0.533
0.641
0.639
0.560
0.506
0.347
0.303
0.295
0.257
0.210
0.178
0.138
0.21
0.20
0.33
0.61
1.67
2.15
2.45
2.80
4.39
4.81
6.69
kobs = ki[ion] ϩ kip[ion pair]
form shown in eqn. (2), where c is the total salt concentration,
kobs/cαf = ki ϩ kip(1 Ϫ α)/αf (2)
(1)
α is the dissociation degree and f is the activity coefficient,
evaluated as previously discussed.
The data for the various salts are reported in Fig. 1. The fit by
linear regression analysis is very good and the four straight lines
have the same intercept (within experimental error) and differ-
ent slopes.
The calculated rate constants of the free ions (ki) and of the
ion pairs (kip) are reported in Table 5, together with the dissoci-
ation constants (Kdiss) of the related salts.
From the data of Table 5 and from Fig. 1 several consider-
ations can be made. First of all it can be seen that applying
Acree’s model gave consistent results: the plots of the data gave
four straight lines for the four chlorides, with a common inter-
cept and positive slopes. It follows that both the free ions and
a The ratio Krypto:KCl was 1.7/1, see Experimental section.
2.5 ( 0.5) × 10Ϫ5  at 22, 12 and 1 ЊC respectively. The corre-
sponding thermodynamic parameters (∆H0 = 14.9 3.9 kJ
molϪ1 and ∆S0 = 142.7 14.0 J molϪ1 KϪ1) are similar to those
determined by Song and Trogler.11 The low enthalpic factor and
the high negative entropy are in accord with a substantial inter-
vention of CH2Cl2 in solvating the separated ions. In PPNCl,
the cation has two structures, one linear and the other bent,
with the positive charge being delocalized on nitrogen, on
phosphorus and on the aromatic rings.12 In CH2Cl2 the inter-
ionic distance is great (11.3 Å),9 so that the salt is supposed to
1748
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
Table 5 Second order rate constantsa and dissociation constants in
CH2Cl2 at 22 ЊC, unless otherwise stated
masked by the importance of the bond between anion and
cation of the ion pair. It is reasonable to think that the counter-
ion is far from the reaction center and not interacting with it.
The lower reactivity of the ion pair as compared to the free ion
is reported in the literature,3,4 but studies with a large variety of
counterion are not available.
The activation parameters determined for Bu4NCl are in
agreement with an SN2 mechanism, where association between
the reactants takes place.21 In fact, the ∆H‡ for the free ion is
not large (54.4 2.5 kJ molϪ1), and smaller than that for the ion
pair (79.5 2.9 kJ molϪ1), where the dissociation between ion
and counterion must be taken into account. The ∆S‡ for the free
ion is negative (Ϫ58.6 8.4 J molϪ1) as in most SN2 reactions,
due to an increase of order in the activated complex. The ∆S‡
(7.9 10 J molϪ1) for the ion pair (even though more difficult to
explain) can be justified because the formation of the activated
complex is coupled with a partial bond breaking of the reacting
salt.
The salt PPNBr was also studied in order to compare the
reactivities of two salts with different anions (BrϪ vs. ClϪ). It
can be seen from Table 5 that the reactivity of free bromide is
lower than that of free chloride (0.46 vs. 1). This is consistent
with the known lower nucleophilicity of BrϪ vs. ClϪ in solvents
in which these ions are not strongly solvated.1b,22 It can also be
noticed that the rate constant of the PPNBr ion pair is similar
to the one of PPNCl (see Table 5).
Salt
ki/Ϫ1 sϪ1
kip/Ϫ1 sϪ1
Kdiss/10Ϫ5 
Bu4NCl
1.07 (0.07)
0.83 (0.09)
0.98 (0.12)
1.04 (0.07)
0.46 (0.1)
0.402 (0.094)
0.177 (0.03)
0.12 (0.003)
0.077 (0.004)
0.52 (0.03)
0.51 (0.02)
0.42 (0.03)
0.043 (0.003)
0.009 (0.001)
1.58
1.87
26.1
34.1
25.8
1.7
KCl–crown
KCl–Krypto
PPNCl
PPNBr
Bu4NClb
Bu4NClc
2.5
a Standard deviations in parentheses. b At 12 ЊC. c At 1 ЊC.
Table 6 Observed first order rate constants for PPNBr in CH2Cl2 at
22 ЊC
[PPNBr]/10Ϫ3 
α
f
kobs/10Ϫ3 sϪ1
0.48
0.58
1.10
1.22
2.75
4.45
6.60
8.35
10.00
0.663
0.641
0.568
0.556
0.472
0.428
0.394
0.375
0.361
0.642
0.623
0.561
0.551
0.471
0.426
0.390
0.370
0.355
0.14
0.19
0.36
0.49
0.91
1.41
1.99
2.63
3.56
Experimental
Spectral grade dichloromethane (Carlo Erba) was refluxed over
P2O5 and distilled in the dark under a dry nitrogen atmosphere
immediately before use. Bu4NCl (Fluka) was crystallised from
diethyl ether–ethanol and dried under vacuum before use.
PPNCl (Strem) and KCl were crystallised from boiling water
and dried under vacuum at 80 ЊC and 120 ЊC, respectively. The
Kryptofix 2,2,2 (Fluka) was dried under vacuum before use.
The 18-crown-6 (Fluka) was crystallised from n-hexane
and dried under vacuum. The complexed KCl was prepared by
Pedersen’s method23 using methanol to solubilize the KCl and
crown ether or Kryptofix 2,2,2. The conductometric measure-
ments were performed as described in ref. 11 and 24. While
some dissociation constants can be found in the literature, we
measured all of those used in the experiments in order to have
homogeneous data.
All kinetic measurements were made on a Beckman DU 7500
diode array spectrophotometer. The path length of the glass cell
was 1 cm. The temperature was monitored with a thermo-
couple. The reaction was started by adding with a microsyringe
the appropriate amount of substrate solution to the thermo-
Fig. 1 Using the Acree eqn. [eqn. (2)] to determine ki and kip. Plot of
(kobs/cαf ) vs. (1 Ϫ α)/αf , where α is the dissociation degree and f is
the activity coefficient. ki is given by the intercept, kip by the slope.
᭹, PPNCl; ᭺, KCl–Krypto; ᭢, Bu4NCl; ᭝, KCl–crown.
the ion pairs are reactive species and compete (Fig. 1) to carry
out the nucleophilic substitution reaction on the sulfonate. The
mechanism of the reaction is assumed to be the classical SN2
process, as found in several other studies on similar substrates.2
The ki values, calculated from the intercept of the straight lines
in Fig. 1, are constant (within experimental error) for all four
chlorides, and the average value for the second-order rate con-
stant for the free chloride is ki = 0.98 0.1 Ϫ1 sϪ1. The second-
order rate constants (kip) of the ion pairs (calculated from the
slopes) are different for all the salts, and always smaller than the
rate constant of the free ion, but for PPNCl and KCl–Krypto
it is just half the rate of the free ion. The lower reactivity of
the ion pairs (kip) as compared to that of the free ion can be
explained by the fact that the reacting chloride is bound to the
counterion, and then the nucleophilicity of the chloride within
the ion pair is lower than that of the free ion because the
negative charge is partially neutralized, and because this bond
must be broken during the reaction. It can be noticed that there
is a substantial correlation between the dissociation constant
(Kdiss) and the rate constant (kip) (see Table 5). The effect of
steric hindrance does not appear to be significant, or else it is
statted salt solution in the cuvette. Rate constants, kobs/sϪ1
,
were determined by following the increase in absorbance at 280
nm (or at 290 for PPNCl) due to the formation of the product
p-nitrobenzenesulfonate ion (2), using the classical equation
ln[(A∞ Ϫ Ao)/(A∞ Ϫ At)] = kobst. Linearity was observed up to
90% of the reaction. Duplicate runs showed that the reproduci-
bility was within 5%. The sulfonate and salt concentrations
were 3 × 10Ϫ5  and (0.3–30) × 10Ϫ3  respectively; to obtain a
pseudo-first-order condition, there was always an excess of salt.
The reactions were followed to completion. The same final
absorbance was obtained even with different excesses of nucleo-
phile, showing that the reaction is not an equilibrium but
goes to completion. For each salt the quality of the isosbestic
points was checked (at 266 nm). Blank experiments showed
that the substrate and the salts were stable in the conditions of
the kinetic experiments. The p-nitrobenzenesulfonate product
was also shown to be stable by the constancy of the A∞
value. In some cases, the same rate constants were also obtained
by following the decrease in the absorbance of (1) at 253 nm.
Very good agreement was obtained by both methods (within
3%).
J. Chem. Soc., Perkin Trans. 2, 1998
1749

View Article Online
5 CRC Handbook of Chemistry and Physics, CRC Press Inc., West
Palm Beach, 67th edn., 1986.
Acknowledgements
Thanks are due to CNR(Roma) and the Ministero della
Università e della Ricerca Scientifica e Tecnologica (MURST)
for financial support.
6 (a) V. Gutman and U. Mayer, Adv. Inorg. Chem. Radiochem., 1975,
17, 189; (b) V. Gutman, The Donor-Acceptor Approach to Molecular
Interaction, Plenum, New York, 1978.
7 I. Svorstøl, Ph.D. Thesis, University of Bergen, 1989.
8 P. I. Svorstøl, H. Høiland and J. Songstad, Acta Chem. Scand., 1984,
B38, 885 and references therein.
9 I. Svorstøl and J. Songstad, Acta Chem. Scand., 1985, B39, 639.
10 (a) E. Hirsch and R. M. Fuoss, J. Am. Chem. Soc., 1960, 82, 1018;
(b) G. Boche, Angew. Chem., Int. Ed. Engl., 1992, 31, 731; (c) A. P.
Abbot and D. J. Schriffin, J. Chem. Soc., Faraday Trans., 1990, 112,
6714.
11 L. Song and W. C. Trogler, J. Am. Chem. Soc., 1992, 114, 3355.
12 (a) M. Tilset, A. A. Zlota, K. Folting and K. G. Caulton, J. Am.
Chem. Soc., 1993, 115, 4113; (b) M. Darensbourg, H. Barros and
C. Borman, J. Am. Chem. Soc., 1977, 99, 1647.
13 (a) G. Lavigne and H. D. Kaesz, J. Am. Chem. Soc., 1984, 106, 4647;
(b) J. Lillis, A. Rokicki, T. Chin and P. C. Ford, Inorg. Chem., 1992,
32, 5040.
14 C. L. Liotta and H. P. Harris, J. Am. Chem. Soc., 1974, 96, 2250.
15 J. M. Lehn, Pure Appl. Chem., 1980, 52, 2303.
16 M. A. Bush and M. R. Truter, J. Chem. Soc. (B), 1971, 1440.
17 B. Metz, D. Moras and K. Weiss, J. Chem. Soc., Chem. Commun.,
1971, 444.
18 D. Moras, B. Metz and K. Weiss, Acta Crystallogr., 1973, B29, 383.
19 S. Boileau, P. Hemery and J. C. Justice, J. Solution Chem., 1975, 4,
873.
References
1 See for example: (a) M. Szwarc, Ions and Ions Pairs in Organic
Reactions, Wiley-Interscience, New York, 1974; (b) C. Reichardt,
Solvents and Solvent Effects in Organic Chemistry, VCH, Weinheim,
1990; (c) Z. G. Lai and K. C. Westaway, Can. J. Chem., 1989, 67, 21;
(d) E. Baciocchi, R. Ruzziconi and G. V. Sebastiani, J. Org. Chem.,
1979, 44, 3718; (e) M. Svoboda, J. Hapala and J. Zavada,
Tetrahedron Lett., 1972, 4, 265; ( f ) J. Smid, A. J. Varma and S. C.
Shah, J. Am. Chem. Soc., 1979, 101, 5764; (g) J. Coddington and
S. Wherland, Inorg. Chem., 1996, 35, 4023; (h) R. Romeo, G. Arena,
L. M. Scolaro and M. R. Plutino, Inorg. Chim. Acta, 1995, 240, 81;
(i) S. Harder, A. Streitwieser, J. T. Petty and P. von R. Schleyer,
J. Am. Chem. Soc., 1995, 117, 3253.
2 See for example: (a) I. Lee, H. J. Koh, B. S. Lee, D. S. Sohn and B. C.
Lee, J. Chem. Soc., Perkin Trans. 2, 1991, 174; (b) E. S. Lewis and
D. D. Hu, J. Am. Chem. Soc., 1984, 106, 3292; (c) J. W. Bunting,
J. M. Mason and C. K. M. Heo, J. Chem. Soc., Perkin Trans. 2, 1994,
2291; (d) H. A. Al-Lohedan, J. Phys. Chem., 1990, 94, 8226; (e)
K. R. Fountain, D. P. Fountain, B. Michaels, D. Brenton Myers,
J. K. Salmon, D. A. Van Galen and P. Yu, Can. J. Chem., 1991, 69,
798; ( f ) K. R. Fountain, L. K. Hutchinson, D. C. Mulhearn and
Y. B. Xu, J. Org. Chem., 1993, 58, 7883; (g) P. E. Dietze, R. Hariri
and J. Khattak, J. Org. Chem., 1989, 54, 3317.
3 (a) P. Beronius, Acta Chem. Scand., 1969, 23, 1175; (b) P. Beronius,
U. Isacsson and A. M. Nilsson, Acta Chem. Scand., 1970, 24, 189;
(c) P. Beronius, A. M. Nilsson and G. Wikander, Acta Chem. Scand.,
1970, 24, 2826; (d) P. Beronius and L. Pataki, J. Am. Chem. Soc.,
1970, 92, 4518.
4 (a) S. F. Acree, Am. Chem. J., 1912, 48, 352; (b) P. Cayzergues,
C. Geourgoulis and G. Mathieu, J. Chim. Phys., 1987, 84, 63;
(c) G. Mathieu, Ph.D. Thesis, University Pierre et Marie Curie,
Paris, 1985; (d) M. G. Pregel, E. J. Dunn and E. Buncel, Can. J.
Chem., 1990, 68, 1846; M. G. Pregel and E. Buncel, J. Chem. Soc.,
Perkin Trans. 2, 1991, 307; M. G. Pregel and E. Buncel, J. Am.
Chem. Soc., 1993, 115, 10; (e) K. J. Msayb and C. I. F. Watt, Chem.
Soc. Rev., 1992, 237; ( f ) R. Poli, B. E. Owens and R. G. Link, Inorg.
Chem., 1992, 31, 662.
20 S. Glasstone, An Introduction to Electrochemistry, D. Van
Nonstrand, New Jersey, 1959.
21 L. L. Schalenger and F. A. Long, Adv. Phys. Org. Chem., 1963, 1, 1.
22 (a) S. Winstein, L. G. Savedoff, S. Smith, I. D. R. Stevens and J. Gall,
Tetrahedron Lett., 1960, 9, 24; (b) A. Earle and C. R. Jablonski,
J. Chem. Soc., Dalton Trans., 1986, 2137.
23 C. J. Pedersen, J. Am. Chem. Soc., 1970, 92, 391.
24 G. Bellachioma, G. Cardaci, A. Macchioni and G. Reichenbach,
Gazz. Chim. Ital., 1991, 121, 101.
Paper 8/02954G
Received 20th April 1998
Accepted 3rd June 1998
1750
J. Chem. Soc., Perkin Trans. 2, 1998