Mechanism of the alkaline hydrolysis of O-ethyl S-(2,4,6-trinitrophenyl) thio-and dithiocarbonates

Article abstract of DOI:10.1039/a903143j

The alkaline hydrolysis of O-ethyl S-(2,4,6-trinitrophenyl) thio-and dithiocarbonates, 1 and 2, respectively, were studied kinetically in 5% dioxane in water, at 25.0°C, and ionic strength 0.2 mol dm^-3 (KCl). Two kinetic processes, well separated in time, were detected. The fast process involves the formation of a σ-complex by addition of HO^- to one of the unsubstituted positions of the aromatic ring, followed by fast ionisation of this complex. The slow process leads to the formation of a mixture of 2,4,6-trinitrophenoxide and 2,4,6-trinitrobenzenethiolate ions in a 10:1 ratio, respectively, in the reaction of 2, and to a mixture in a 2:1 ratio in the reaction of 1, independent of KOH concentration. Although the substrates show different kinetic behaviour with KOH concentration, the results can be discussed on the basis of a common reaction mechanism.

Full text of DOI:10.1039/a903143j

View Article Online / Journal Homepage / Table of Contents for this issue
Mechanism of the alkaline hydrolysis of O-ethyl S-(2,4,6-
trinitrophenyl) thio- and dithiocarbonates
Enrique A. Castro,a María Cubillos,a José G. Santos,a Elba I. Buján,b M. Virginia Remedi,b
Mariana A. Fernándezb and Rita H. de Rossib
a Pontificia Universidad Católica de Chile, Facultad de Química, Casilla 306, Santiago 22,
Chile. E-mail: ecastro@puc.cl
b INFIQC, Universidad Nacional de Córdoba. Dpto. de Química Orgánica,
Facultad de Ciencias Químicas, Ciudad Universitaria, 5000 Córdoba, Argentina
Received (in Cambridge, UK) 20th April 1999, Accepted 19th August 1999
The alkaline hydrolysis of O-ethyl S-(2,4,6-trinitrophenyl) thio- and dithiocarbonates, 1 and 2, respectively, were
studied kinetically in 5% dioxane in water, at 25.0 ЊC, and ionic strength 0.2 mol dmϪ3 (KCl). Two kinetic processes,
well separated in time, were detected. The fast process involves the formation of a σ-complex by addition of HOϪ
to one of the unsubstituted positions of the aromatic ring, followed by fast ionisation of this complex. The slow
process leads to the formation of a mixture of 2,4,6-trinitrophenoxide and 2,4,6-trinitrobenzenethiolate ions in
a 10:1 ratio, respectively, in the reaction of 2, and to a mixture in a 2:1 ratio in the reaction of 1, independent
of KOH concentration. Although the substrates show different kinetic behaviour with KOH concentration, the
results can be discussed on the basis of a common reaction mechanism.
Introduction
Although the aminolysis reactions of thio- and dithiocarbonates
have received some attention,1,2 little is known about the mech-
anism of the basic hydrolysis of these compounds.3 We have
recently examined the reactions of O-ethyl S-(2,4,6-trinitro-
phenyl) thiocarbonate, 1, and O-ethyl S-(2,4,6-trinitrophenyl)
dithiocarbonate, 2, with secondary alicyclic amines and pyrid-
ines.2 We have found that, under all conditions, the reactions
Fig. 1 Spectra for the hydrolysis of 1 in basic solution at different
lead mainly to the aminolysis products and only in the reactions
of 1 with nicotinamide2c and 2 with 3-chloropyridine2d and
piperazinium ion,2a was the hydrolysis reaction detected. Like-
wise in the reactions of 1 with anilines and phenoxide ions the
hydrolysis reaction was of minor importance compared to
the aminolysis or phenolysis reactions.4 In all these reactions2,4
the main reaction product is that corresponding to nucleophilic
attack on the carbonyl or thiocarbonyl group with no indi-
cation, within the detection limits, of nucleophilic aromatic
substitution. Since 1 and 2 have an aromatic ring strongly acti-
vated for nucleophilic aromatic substitution,5 we considered it
to be of interest to study the hydrolysis reaction of these com-
pounds in order to see if we could detect the formation of
Meisenheimer intermediates which are often observed in other
reactions of trinitrophenyl derivatives with various substituents
in the 1-position.6–10 Therefore, in the present work we carry out
a kinetic study of the hydrolysis of 1 and 2. In fact, both com-
pounds give substitution on the aromatic ring and in both cases
this is the predominant reaction. Besides, addition of HOϪ to
an unsubstituted position of the aromatic ring takes place and
reaction times. [HOϪ] = 0.1 mol dmϪ3, time interval 1 s, first cycle (A).
the Meisenheimer complex formed, and its conjugate base, can
be detected at short reaction times.
Results
Reactions of O-ethyl S-(2,4,6-trinitrophenyl) thiocarbonate, 1
The spectrum of 1 in 5% dioxane in water showed no absorp-
tion in the visible region. When KOH was added a coloured
species was rapidly formed. The spectrum of the reaction mix-
ture showed an increase in absorbance beyond 450 nm. After 1 s
(in the reaction with KOH 0.2 mol dmϪ3), two isosbestic points
developed, one around 290 nm and another at 450 nm with a
broad absorption band between them (Fig. 1).
Two kinetic processes, well separated in time, were measured
at 400 nm. The results are summarised in Table 1. Analysis of
the products gave 67% of 2,4,6-trinitrophenoxide ion (3) and
33% of 2,4,6-trinitrobenzenethiolate ion (4), eqn. (1). The ratio
J. Chem. Soc., Perkin Trans. 2, 1999, 2603–2607
This journal is © The Royal Society of Chemistry 1999
2603

View Article Online
Table 1 Values of the rate constants for the alkaline hydrolysis of 1 for
the fast (kfast) and the slow (kslow) processesa
Table 2 Values of the rate constants for the alkaline hydrolysis of 2 for
the fast (kfast) and the slow (kslow) processesa
[KOH]/mol
kfast/sϪ1 b
(400 nm)
kslow/sϪ1 c
(400 nm)
[KOH]/mol
kfast/sϪ1
(480 nm)
kslow/sϪ1
(480 nm)
kslow/sϪ1
(300 nm)
dmϪ3
dmϪ3
0.010
0.040
0.050
0.070
0.100
0.125
0.130
0.150
0.160
0.190
0.200
0.037 0.003
0.148 0.001
0.011
0.019
0.040
0.049
0.060
0.079
0.100
0.119
0.120
0.139
0.140
0.159
0.160
0.180
0.190
0.193
0.072 0.003
0.141 0.005
0.281 0.003
11
8.1 0.6
1
5.47 0.03
0.142 0.007
0.338 0.009
0.274 0.001
0.40 0.02
5.64 0.09
5.9 0.04
6.1 0.1
6.1 0.1
6.63 0.04
14.0 0.5
17
0.373 0.004
0.45 0.01
0.494 0.006
0.510 0.006
0.39 0.02
0.49 0.02
0.55 0.01
1
0.47 0.02
0.58 0.05
0.71 0.06
19 0.9
21
28
2
2
0.65 0.02
0.66 0.02
25.2 0.9
6.9 0.3
0.52 0.01
0.67 0.03
0.67 0.04
a Solvent: 5% dioxane in water, temperature 25.0 ЊC, ionic strength 0.2
7.64 0.07
0.524 0.005
0.53 0.01
mol dmϪ3 b Errors represent the deviation of the mean of 5–10
.
determinations. c Errors are the standard deviation of the absorbance vs.
8.34 0.07
8.5 0.2
time plot from a single exponential.
a Solvent 5% dioxane in water, temperature 25.0 ЊC, ionic strength 0.2
mol dmϪ3. Errors are the standard deviation of the absorbance vs. time
plot from a single exponential.
[3]:[4] = 2:1 was independent of HOϪ concentration, within
experimental error.
Reaction of O-ethyl S-(2,4,6-trinitrophenyl) dithiocarbonate, 2
The spectrum of this substrate in 5% dioxane in water showed a
band at 270 nm. When KOH was added, the first spectrum that
could be taken after mixing showed a decrease in absorbance at
270 nm and an increase in absorbance at 500 nm (Fig. 2, line A).
Thereafter, two isosbestic points were observed, at 285 and 430
nm, with an increase in absorbance between them. The band
at 270 nm and the absorbance beyond 430 nm also decreased
(Fig. 2).
At 480 nm, two kinetic processes were measured while at 300
nm there was only one, which corresponds to the slowest
observed at 480 nm. Since the absorbance changes at 480 nm
were too small, better results were obtained from the absorb-
ance changes at 300 nm, which correspond to the maximum
absorption of potassium O-ethyl dithiocarbonate. Both pro-
cesses exhibit a nonlinear dependence on KOH concentration
(Table 2). Products analysis gave 91% of 3 and 9% of 4. The
ratio [3]:[4] = 10:1 was independent of the HOϪ concentration,
within experimental error.
Fig. 2 Spectra for the hydrolysis of 2 in basic solution at different
reaction times. [HOϪ] = 0.1 mol dmϪ3, time interval 1 s, first cycle (A).
dinitrobenzenes.10 For this mechanism two relaxation times
may be expected,11 however only one (kfast in Tables 1 and 2)
is observable since the proton transfer will be a very rapid
equilibrium on the stopped-flow timescale. Assuming that the
process observed corresponds to the formation of 5, in rapid
equilibrium with 6 or 7, the observed rate constant is given
by eqn. (2).
kϪ1
kfast = k1[OHϪ] ϩ
(2)
1 ϩ K2[OHϪ]
By fitting† the data to eqn. (2), the values of k1, kϪ1 and K2 for
the reactions of 1 and 2 could be determined and from the two
former values the equilibrium constant K1 was also calculated;
these values are collected in Table 3. Fig. 3 shows the plot of
kfast vs. [OHϪ] for the reaction of 2. The line was drawn through
eqn. (2) with the data of Table 3, which shows the very good
agreement between the experimental and calculated data. In
Table 3, it can be seen that the K2 value for the reaction of 2 is
about 5 times greater than that of 1, reflecting the higher acidity
Discussion
The fast absorbance increase observed beyond 450 nm for the
reaction of 1 and beyond 430 nm for that of 2 is consistent with
a mechanism involving the formation of a σ-complex by addi-
tion of hydroxide ion to one of the unsubstituted positions of
the aromatic ring (Scheme 1).
Similar adducts have been reported in related systems. Some
examples are the hydrolysis reactions of 2,4,6-trinitrophenyl
aryl sulfides and the corresponding ethers,9a and 1-amino-2,4-
† Sigmaplot, version 1.0, 1995.
2604
J. Chem. Soc., Perkin Trans. 2, 1999, 2603–2607

View Article Online
Table 3 Calculated rate and equilibrium constants for the hydrolysis
reactions of 1 and 2a
1
2
k1/dm3 molϪ1 sϪ1
114.2
4
32.4 0.6
5.8 0.2
5.6 0.2b
8.8 0.8
8.3 0.8
5.4 1.0
0.13 0.06
7.5; 0.8d
k
Ϫ1/sϪ1
4.1 0.6
27.9 4.2b
1.6 0.6
3.9 0.4
4.3 0.3
0.72 0.5
2.6; 1.3c
K1/dm3 molϪ1
K2/dm3 molϪ1
(k3 ϩ kЈ3)/dm3 molϪ1 sϪ1
(k4 ϩ kЈ4)/dm3 molϪ1 sϪ1
(k5 ϩ kЈ5)/dm3 molϪ1 sϪ1
k3; kЈ3/dm3 molϪ1 sϪ1
k4; kЈ4/dm3 molϪ1 sϪ1
k5; kЈ5/dm3 molϪ1 sϪ1
2.9; 1.4c
4.9; 0.5d
0.48; 0.24c
0.12; 0.01d
a Errors shown are standard errors. b Calculated as k1/kϪ1 c Assuming
.
k3/kЈ3 = k4/kЈ4 = k5/kЈ5 = [3]/[4] = 2/1. d Assuming k3/kЈ3 = k4/kЈ4 = k5/
kЈ5 = [3]/[4] = 10/1.
Fig. 3 Dependence of the rate constant for the fast process, kfast, on
the HOϪ concentration for the hydrolysis of 2. The line was calculated
with eqn. (2) and the data of Table 3.
of HOϪ on C-1 of the aromatic ring, and therefore, the form-
ation of intermediates 8, 9 and 10. However, these intermedi-
ates do not accumulate under the reaction conditions and only
an overall rate constant can be determined. Highly charged
Meisenheimer complexes are not unusual in polar solvents. For
instance, trinitrobenzene adds two sulfite ions giving a Meisen-
heimer complex with four negative charges.13
Scheme 1
of the dithio complex 5 (with Z = S). On the other hand, the
value of k1 is larger for the reaction of 1 than that of 2, in
agreement with the findings in aminolysis reactions: the thiol
derivative 1 is more reactive than the dithio derivative 2.2a,b
For the slow hydrolysis of 1 and 2 we suggest the mechanism
depicted in Scheme 2. In this scheme, 1 and 2 have reached
equilibrium with 5. This mechanism is proposed on the basis of:
i) the product analysis, which indicates that the nucleophilic
reaction takes place at the aromatic as well as at the carbonyl
(or thiocarbonyl) carbon; ii) the dependence of the rate con-
stant on HOϪ concentration; and iii) the mechanism just
described for the fast process.
According to Scheme 2, the observed rate constant for the
slow hydrolysis of 1 and 2 (kslow), is given by eqn. (4).
kslow
=
(k3 ϩ kЈ3)[OHϪ] ϩ (k4 ϩ kЈ4)K1[OHϪ]2 ϩ (k5 ϩ kЈ5)K1K2[OHϪ]3
1 ϩ K1[OHϪ] ϩ K1K2[OHϪ]2
(4)
By fitting the kslow vs. [OHϪ] data to eqn. (4) for the reactions
of 1 and 2 the values of (k3 ϩ kЈ3), (k4 ϩ kЈ4) and (k5 ϩ kЈ5)
could be determined; these are shown in Table 3. Fig. 4 shows
the fitting of eqn. (4), using the values in Table 3, to the experi-
mental points for the reaction of compound 1. Assuming that
the [3]:[4] ratios (2:1 and 10:1, for the reactions of 1 and 2,
respectively) reflect the rate constant ratios for the formation of
3 and 4, all the rate microcoefficients in eqn. (4) can be calcu-
lated; they are also summarised in Table 3.
The upper part of Scheme 2 represents the attack of HOϪ at
the 1-aromatic carbon and the lower part of the scheme depicts
its nucleophilic reaction at the carbonyl or thiocarbonyl centre.
A similar reaction to that for the k5 step of Scheme 2 is the
reaction of the 1:1 Meisenheimer complex shown in eqn. (3),
The ratio of the nucleophilic rate constants for attack at the
aromatic/carbonyl carbons (karom/kco) for 1 is lower (2:1) than
the ratio karom/kcs for 2 (10:1). This could be due to the fact that
HOϪ is a hard base that prefers to bind to the relatively hard
CO group compared to the softer CS group.14 The regio-
selectivity observed for these reactions contrasts that observed
with amines where addition to the aromatic ring was never
detected.2 The different behaviour of amines and HOϪ as
nucleophiles may be attributed to the higher steric effect in the
reactions of the amines which are bulkier than HOϪ. Important
steric effects were observed in other nucleophilic aromatic
substitution reactions.15
which has been reported in the hydrolysis of 1-X-2,4,6-
trinitrobenzenes.12
It should be noted that the paths with rate constants k3, kЈ3,
k4, kЈ4, k5 and kЈ5 do not represent elementary steps. All the
steps that lead to product 3 in Scheme 2 must involve the attack
J. Chem. Soc., Perkin Trans. 2, 1999, 2603–2607
2605

View Article Online
Scheme 2
in the case of compound 2 than 1 due to the bulkier C᎐S
᎐
group in the former complex compared to the C᎐O group in the
᎐
latter.
Another possible mechanism for the fast process is that
shown in Scheme 3, where there is an equilibrium between 1 or
2 and 5.
The existence of Meisenheimer 1:2 σ-complexes (similar to
11) is well documented. These have been found in the hydrolysis
of 1,3,5-trinitrobenzene and 1-X-2,4,6-trinitrobenzenes,6b ethyl
dinitrobenzoates6c and 1-amino-2,4,6-trinitrobenzenes,7 and in
the reactions of sulfite ions with alkyl and aryl ethers and
thioethers.9b Aqueous solutions favour the formation of the 1:2
σ-complexes.16
From Scheme 3, eqn. (5) can be derived. Reasonable fits can
also be found with this mechanism.
K1k6[OHϪ]2
1 ϩ K2[OHϪ]
(5)
kfast
=
ϩ kϪ6
For the slow process, assuming equilibrium for the two steps
in Scheme 3, a similar mechanism to that depicted in Scheme 2
and a similar rate law to that in eqn. (4) can be obtained.
The formation of the aromatic substitution product 3 by the
reaction of HOϪ with a 1:2 Meisenheimer complex has been
shown to occur in the hydrolysis reactions of picramides7 and
2,4-dinitroanilines.10
Fig. 4 Dependence of the rate constant for the slow process, kslow, on
the HOϪ concentration for the hydrolysis of 1. The line was calculated
with eqn. (4) and the data of Table 3.
On the other hand, it can be seen (Table 3, k3) that for the
addition of hydroxide anion to the C-1 position of the aro-
matic ring, substrate 2 is three-fold more reactive than 1. This
result may be explained by the fact that in going from the
Meisenheimer complex 8 to 3 there is more steric release
Although satisfactory fits are found for the fast and slow
processes for the mechanisms with and without the dihydroxy
2606
J. Chem. Soc., Perkin Trans. 2, 1999, 2603–2607

View Article Online
mixture of 2,4,6-trinitrophenoxide and 2,4,6-trinitrobenzene-
thiolate ions in a 2:1 ratio, independent of KOH concentration.
This was achieved by comparing the final spectra of the reac-
tions with those of various mixtures of authentic 2,4,6-
trinitrophenoxide and 2,4,6-trinitrobenzenethiolate ions under
the same experimental conditions.
On the other hand, in the reactions of 2, the presence of
potassium O-ethyl dithiocarbonate and a mixture of 2,4,6-
trinitrophenoxide and 2,4,6-trinitrobenzenethiolate ions in a
10:1 ratio as products was inferred by comparison of the final
reaction spectra with those of authentic samples of these
products under identical experimental conditions. The form-
ation of 4 was detected and quantified by HPLC using an
instrument equipped with a Knauer Model 64 pump with an
RP-18 column and Perkin Elmer LC-15 UV detector (254 nm),
under the following conditions: eluant, methanol in the iso-
cratic mode; flow rate, 0.5 ml minϪ1; room temperature. Under
these conditions, the retention times for 2,4,6-trinitrophenol and
2,4,6-trinitrobenzenethiol were 3.2 and 5.7 min, respectively.
Acknowledgements
Financial assistance for this work from Fundación Andes
(Chile) and Fundación Antorchas (Argentina) is gratefully
acknowledged.
Scheme 3
Meisenheimer complex 11, we prefer the mechanisms without it
(Schemes 1 and 2). The reasons for this preference are:‡ i) by
using the values of the microcoefficients found in the fitting
through eqn. (5), the fraction of 11 calculated, after equi-
libration with 5, is very small at low OHϪ concentration. Sub-
stantially larger fractions of 6 or 7 are found for the mechanism
in Scheme 1, using the data in Table 3. ii) The visible spectra
(400–600 nm) of the initially produced adducts in Figs. 1 and 2
show absorption bands that correspond more to a 1:1 adduct
than to a 1:2 complex. iii) The data in Table 1 show a minimum
in a plot of kfast vs. [OHϪ]. This is predicted by eqn. (2) but not
by eqn. (5). iv) The values of K1, for monoadduct formation,
are usually larger than those of K6, for diadduct formation.
Hence, two relaxations should have been found for the fast
process; however, we observed only one relaxation.
References
1 E. A. Castro, N. E. Alvarado, S. A. Peña and J. G. Santos, J. Chem.
Soc., Perkin Trans. 2, 1989, 635; E. A. Castro, F. Ibáñez, M. Salas
and J. G. Santos, J. Org. Chem., 1991, 56, 4819; E. A. Castro,
M. Cubillos and J. G. Santos, J. Org. Chem., 1994, 59, 3572.
2 (a) E. A. Castro, F. Ibáñez, M. Salas, J. G. Santos and P. Sepúlveda,
J. Org. Chem., 1993, 58, 459; (b) E. A. Castro, M. Salas and
J. G. Santos, J. Org. Chem., 1994, 59, 30; (c) E. A. Castro, M. I.
Pizarro and J. G. Santos, J. Org. Chem., 1996, 61, 5982; (d) E. A.
Castro, C. A. Araneda and J. G. Santos, J. Org. Chem., 1997, 62, 126.
3 L. J. Santry and R. A. McClelland, J. Am. Chem. Soc., 1983, 105,
3167; R. J. Millican, M. Angelopoulos, A. Bose, B. Riegel,
D. Robinson and C. K. Wagner, J. Am. Chem. Soc., 1983, 105, 3622;
P. J. S. Harris, S. Afr. J. Chem., 1984, 37, 91.
4 E. A. Castro, L. Leandro, P. Millán and J. G. Santos, J. Org. Chem.,
1999, 64, 1953; E. A. Castro, P. Pavez and J. G. Santos, J. Org.
Chem., 1999, 64, 2310.
Experimental
5 F. Terrier, Nucleophilic aromatic displacement: The influence of the
nitro group, VCH Publishers, Inc., New York, 1991.
6 (a) R. Bacaloglu, C. A. Bunton, G. Cerichelli and F. Ortega, J. Am.
Chem. Soc., 1988, 110, 3495; (b) R. Bacaloglu, C. A. Bunton and
F. Ortega, J. Am. Chem. Soc., 1988, 110, 3503; 1989, 111, 1041;
(c) R. Bacaloglu, A. Blasko, C. A. Bunton and F. Ortega, J. Am.
Chem. Soc., 1990, 112, 9336.
7 E. B. de Vargas and R. H. de Rossi, J. Phys. Org. Chem., 1989, 2, 507.
8 M. M. Nassetta, E. B. de Vargas and R. H. de Rossi, J. Phys. Org.
Chem., 1991, 4, 277.
9 (a) R. Chamberlin, M. R. Crampton and R. L. Knight, J. Chem.
Res. (S), 1993, 444; (b) M. R. Crampton and A. J. Holmes, J. Phys.
Org. Chem., 1998, 11, 787.
10 E. Buján de Vargas, M. V. Remedi and R. H. de Rossi, J. Phys. Org.
Chem., 1995, 8, 113.
11 C. F. Bernasconi, Relaxation Kinetics, Academic Press Inc., New
York, 1976.
12 B. Gibson and M. R. Crampton, J. Chem. Soc., Perkin Trans. 2,
1979, 648; M. R. Crampton, A. B. Davis, C. Greenhalgh and
J. A. Stevens, J. Chem. Soc., Perkin Trans. 2, 1989, 675.
13 C. F. Bernasconi and R. G. Bergstrom, J. Am. Chem. Soc., 1973, 95,
3603.
Materials
Compounds 1 and 2 were available from previous work.2a,b
KOH and KCl, Merck a.r., were used as purchased and dioxane
was purified as described previously.17
Kinetic measurements
Spectral and kinetic measurements were recorded on a
Shimadzu UV-2101PC spectrophotometer or a Hewlett Pack-
ard HP-8453 diode array equipped with a Hi-Tech SFA-20
rapid kinetic accessory. An Applied Photophysics SF 17MV
stopped-flow spectrofluorimeter was used to measure the rate
coefficients of the fast processes. All the reactions were studied
in 5% dioxane in water, 25.0 0.1 ЊC, and ionic strength 0.2 mol
dmϪ3, with KCl used as compensating electrolyte. All kinetic
runs were carried out under pseudo-first-order conditions, the
initial substrate concentration being ca. 2 × 10Ϫ5 mol dmϪ3
.
In a typical stopped-flow experiment, two solutions were
prepared in 5% dioxane in water at twice the concen-
trations required for the final solution: one containing the sub-
strate and the other the nucleophile and KCl. The reaction was
initiated by mixing equal volumes of both solutions. All relax-
ation times represent the average of five to ten determinations.
14 R. G. Pearson, J. Chem. Educ., 1968, 45, 581; R. G. Pearson, J. Org.
Chem., 1989, 54, 1423.
15 R. Chamberlin and M. R. Crampton, J. Chem. Soc., Perkin Trans. 2,
1993, 75.
16 M. R. Crampton, J. Chem. Soc., Perkin Trans. 2, 1967, 1341;
F. Terrier, A. P. Chatrousse and R. Schaal, J. Org. Chem., 1972, 37,
3010; F. Terrier and F. Millot, Bull. Soc. Chim. Fr., 1970, 1743.
17 R. H. de Rossi and E. B. de Vargas, J. Am. Chem. Soc., 1981, 103,
1533.
Product studies
The final products of the reactions of 1 were identified as a
Paper 9/03143J
‡ We thank the referees for their comments regarding these matters.
J. Chem. Soc., Perkin Trans. 2, 1999, 2603–2607
2607