Solvolysis of substituted benzyl azoxyarenesulfonates: Characterisation of the transition state and the selectivity of benzylic intermediates in 50% aqueous 2,2,2-trifluoroethanol

Article abstract of DOI:10.1039/b106887n

Thirteen substituted benzyl azoxyarenesulfonates have been prepared and rate constants for solvolysis of twelve have been measured in 50% (v/v) aqueous 2,2,2-trifluoroethanol over a range of temperatures from which activation parameters and rate constants at 25°C have been determined. A six-point Hammett correlation for substituents in the benzylic electrofuge, with azoxytoluene-p-sulfonate as the common nucleofuge, gives ρ(σ⁺) = -3.27. With 4-methylbenzyl as a common electrofuge, four substituents in the azoxybenzenesulfonate nucleofuge give ρ(σ) = 1.07. Comparison with model reactions indicates a synchronous concerted rate-limiting fragmentation for these substrates. With the less reactive 3-chlorobenzyl as electrofuge, ρ for substituents in the nucleofuge is only ca. 0.7 which indicates a lower degree of charge development in the arenesulfonate in the transition state for these reactions. Product analyses have been carried out for compounds of high, intermediate, and low reactivity. These indicate a reactivity-selectivity relationship for capture of the substituted benzylic electrophiles by water and 2,2,2-trifluoroethanol. Substituted benzaldehydes are formed from the less reactive substrates, indicating trapping of the first-formed benzylic carbenium ion by nitrous oxide and subsequent elimination of nitrogen and a proton; but yields are low and decrease as the substituted benzyl cation becomes increasingly stable.

Full text of DOI:10.1039/b106887n

View Online / Journal Homepage / Table of Contents for this issue
Solvolysis of substituted benzyl azoxyarenesulfonates:†
characterisation of the transition state and the selectivity of
benzylic intermediates in 50% aqueous 2,2,2-trifluoroethanol‡
2
Isobel M. Gordona and Howard Maskill*b
a Chemistry Department, University of Stirling, Stirling, UK FK9 4AU
b Chemistry Department, University of Newcastle, Newcastle upon Tyne, UK NE1 7RU
Received (in Cambridge, UK) 30th July 2001, Accepted 18th September 2001
First published as an Advance Article on the web 17th October 2001
Thirteen substituted benzyl azoxyarenesulfonates have been prepared and rate constants for solvolysis of twelve
have been measured in 50% (v/v) aqueous 2,2,2-trifluoroethanol over a range of temperatures from which activation
parameters and rate constants at 25 ЊC have been determined. A six-point Hammett correlation for substituents
in the benzylic electrofuge, with azoxytoluene-p-sulfonate as the common nucleofuge, gives ρ(σϩ) = Ϫ3.27. With
4-methylbenzyl as a common electrofuge, four substituents in the azoxybenzenesulfonate nucleofuge give ρ(σ) = 1.07.
Comparison with model reactions indicates a synchronous concerted rate-limiting fragmentation for these substrates.
With the less reactive 3-chlorobenzyl as electrofuge, ρ for substituents in the nucleofuge is only ca. 0.7 which indicates
a lower degree of charge development in the arenesulfonate in the transition state for these reactions. Product
analyses have been carried out for compounds of high, intermediate, and low reactivity. These indicate a reactivity–
selectivity relationship for capture of the substituted benzylic electrophiles by water and 2,2,2-trifluoroethanol.
Substituted benzaldehydes are formed from the less reactive substrates, indicating trapping of the first-formed
benzylic carbenium ion by nitrous oxide and subsequent elimination of nitrogen and a proton; but yields are low
and decrease as the substituted benzyl cation becomes increasingly stable.
cated by parallel bimolecular reactions,6 and that these
Introduction
reactions represent a mechanistic link between alkyl arene-
Whether the various changes which constitute a chemical reac-
sulfonate solvolysis and alkylamine deamination.7 We estab-
tion are concerted or step-wise is a fundamental question which
lished that they involve an initial rate-determining concerted
continues to be investigated for reactions newly discovered and
fragmentation of the alkyl azoxyarenesulfonate, followed by
ones long known.1 Mechanisms of solvolyses of alkyl halides
capture of the carbenium ion intermediate by the solvent. In
and arenesulfonates were amongst the earliest to be investi-
other words, these are unambiguously step-wise reactions
gated; they provided much of the technical vocabulary of the
involving intermediate carbenium ions generated in an initial
subject still in current use (SN1, SN2, E1, E2, etc.) and illustrate
rate-limiting concerted fragmentation. The benzyl carbenium
well the issues involved.2 SN2 and E2 mechanisms are concerted
ion, generated along with nitrous oxide and the nucleofuge
and SN1 and E1 mechanisms are step-wise proceeding through
within the same solvent cage, from the parent benzyl azoxy-
carbenium ion intermediates. The degree to which a concerted
toluene-p-sulfonate (structure 1a) was shown to be too short-
mechanism is synchronous is an issue which has been appreci-
lived to be captured by dilute nucleophiles.8 But about half is
ated only relatively recently.1 Benzyl arenesulfonates were first
trapped by the proximate nitrous oxide molecule generated in
the initial fragmentation (a form of internal return) in competi-
tion with capture by the solvent, aqueous TFE (Scheme 1,
thought to react simply via step-wise SN1 mechanisms on the
grounds that nucleophilic solutes had no effect upon rates but
led to substitution products, and electron-donating substituents
in the electrofuge lead to appreciably greater reactivity.3 More
recent work indicated that substituted benzyl arenesulfonates
may be SN1 or SN2 according to the nature of the substituents
and experimental conditions.4 The parent benzyl toluene-p-
sulfonate and analogues with electron-withdrawing substituents
react by solvent-induced SN2 mechanisms in 1 : 1 (v/v) aqueous
2,2,2-trifluoroethanol, and coupling between the departure
of the nucleofuge and the bonding of the nucleophile depends
upon the substituent in the electrofuge.5 The mechanistic
characterisation of a solvolytic reaction of a benzylic arenesul-
fonate, for example, as SN1, SN2, or both concurrently, remains
a difficult area.
Some years ago, we reported that secondary alkyl azoxy-
arenesulfonates undergo unimolecular solvolysis uncompli-
Scheme 1
Ar = Ph, ArЈ = 4-tolyl). The product of trapping by nitrous
oxide itself then undergoes further reaction, mainly nucleo-
philic capture.
For substituted benzyl azoxyarenesulfonates 1a–i, we also
reported preliminary kinetic effects of substituents X (in the
† The IUPAC name for benzyl azoxyarenesulfonate is N-benzyl-NЈ-
arylsulfonyloxydiazene N-oxide.
‡ Electronic supplementary information (ESI) available: synthesis and
analytical data of azoxyarenesulfonates. See http://www.rsc.org/
suppdata/p2/b1/b106887n
DOI: 10.1039/b106887n
J. Chem. Soc., Perkin Trans. 2, 2001, 2059–2062
This journal is © The Royal Society of Chemistry 2001
2059

View Online
Fig. 1 Hammett plots for substituents Y in 3-chlorobenzyl azoxy-
(4-Y)benzenesulfonates: ᭜ for σϩ parameters and ᭹ for σ parameters.
eters were determined using the Eyring equation.13 Results for
six azoxytoluene-p-sulfonates (1a–f ), i.e. with a common
nucleofuge, are shown in Table 1. Rate constants at 25 ЊC lead
to an excellent 6-point Hammett correlation and ρ(σXϩ) =
Ϫ3.27 (r > 0.999);9 4-cyanobenzyl azoxytoluene-p-sulfonate
(1m) did not give reproducible rate constants for solvolysis. The
enthalpies of activation decrease regularly as the electron-
supplying ability of the substituent in the benzyl residue
increases, but no trend in the small negative entropies of
activation is evident. Results for 4-methylbenzyl azoxyarene-
sulfonates (1c,g–i), i.e. reactants with the same reactive electro-
fuge, are also included in Table 1 and lead to a simple linear
Hammett correlation, ρ(σY) = 1.07 (r = 0.996). Again, a trend in
∆H‡ is evident; values decrease as the substituents in the
nucleofuge become more electron-withdrawing, but the ∆S‡
values are similarly small and negative. Additionally, results for
four 3-chlorobenzyl azoxyarenesulfonates (1f,j–l), i.e. reactants
with the same unreactive electrofuge, were determined and are
also included in Table 1; no trends are evident in either ∆H‡
or ∆S‡. For these reactants, the Hammett plot using all
four points and the normal σ constants is curved, but a
linear correlation is obtained using σϩ parameters (Fig. 1)
ρ(σYϩ) = 0.71 (r = 0.99).
electrofuge) and Y (in the nucleofuge) which indicated that the
initial concerted fragmentation is essentially synchronous.9 We
now present details of that study and product analysis results
for solvolysis of 1b, 1c, and 1f in 50% (v/v) aqueous 2,2,2-
trifluoroethanol (50TFE, molar ratio H2O : TFE = 4 : 1). We
have also investigated substituent effects upon the kinetics of
four 3-chlorobenzyl azoxyarenesulfonates (1f and 1j–l) to learn
how ρ(σY) depends upon the nature of the electrofuge, i.e. to
investigate the second-order effect of substituents in the electro-
fuge upon the ρ-value for substituents in the nucleofuge on the
rate-limiting initial fragmentation. Finally, we report activation
parameters for solvolyses of all compounds 1a–l in 50TFE
which complement results based upon product analysis and
substituent effects upon rate constants at a single temperature
in this overall mechanistic investigation.
Methods and results
Syntheses of substituted benzyl azoxyarenesulfonates followed
our previous route starting from substituted benzaldehydes
which react with hydroxylammonium chloride and sodium
ethanoate in aqueous methanol to give oximes in high yields.6,8
The oximes were then reduced with sodium cyanoborohydride
under controlled acidic conditions in methanol;10 the hydroxyl-
amine products had to be purified to ensure good yields in
the subsequent nitrosation with sodium nitrite in aqueous
hydrochloric acid containing methylated spirit. CAUTION:
The nitrosated products, actually substituted N-benzyl-NЈ-
hydroxydiazene oxides, were treated as potent carcinogens. Six
substituted benzyl azoxytoluene-p-sulfonates (and 4-methyl-
benzyl azoxy(4-bromo)benzenesulfonate) were prepared by the
Tipson method from the substituted N-benzyl-NЈ-hydroxy-
diazene oxides and tosyl (or 4-bromobenzenesulfonyl) chlor-
ide.11 4-Dimethylaminobenzyl azoxytoluene-p-sulfonate proved
too reactive to allow isolation (effervescence in the reaction
mixture indicated that it had formed but decomposed with the
evolution of nitrous oxide). The Tipson method proved less
successful for the preparation of other azoxyarenesulfonates
and an aqueous method was used.12 The azoxyarenesulfonates
were isolated as crystalline products and purified by trituration
with ether–pentane mixtures for kinetics; the three compounds
used for product analytical studies were further purified by
recrystallisation at low temperatures from ether and/or pentane.
All preparative experimental details are available as electronic
supplementary information.‡
Product analyses
The products from three reactants (1b, c, and f ) which span a
range in reactivity of 5.5 × 103 at 25 ЊC in 50TFE were investi-
gated. For 1c and 1f, absolute yields of alcohols in the com-
pleted reaction mixtures were analysed by direct injection
HPLC using calibration curves constructed from pure authen-
tic samples of the alcohols. Absolute yields of 2,2,2-trifluoro-
ethyl ethers and aldehydes from 1c and 1f were analysed using
molar response factors with respect to the alcohols for the
HPLC UV detector.8 A low yield (<1%) of N-(3-chlorobenzyl)-
hydroxydiazene oxide was obtained (using an estimated value
for the molar response factor) from the reaction of 1f. Thus,
absolute yields for all products were obtained for the reactions
of 1c and 1f (Table 2). 4-Methoxybenzyl azoxytoluene-p-
sulfonate (1b) was too reactive and unstable to obtain pure, so
accurate absolute yields could not be determined. However,
no N-(4-methoxybenzyl)hydroxydiazene oxide was detected
amongst the products, and there was only a trace of the alde-
hyde (<0.01%). Consequently, relative yields of the alcohol and
2,2,2-trifluoroethyl (TFEt) ether were obtained from direct
HPLC analysis using the molar response factor for the alcohol
and ether (Table 2). We observe that the ratio of alcohol : ether
from (1f ) is smaller (3.2) than the ratio of water : TFE in the
solvent (4 : 1), but it increases with increasing reactivity of the
substrate (4.0 for 1c and 8.1 for 1b). The yield of aldehyde from
1f is very low and decreases further as the reactivity of the
substrate increases.
Kinetics
Discussion
Average rate constants for solvolysis in 50TFE at four temper-
atures were determined by our usual computer-controlled UV
method,6 from which values at 25.0 ЊC and activation param-
The Hammett results based upon (i) different substituents in
the benzyl residue and a constant azoxytoluene-p-sulfonate
2060
J. Chem. Soc., Perkin Trans. 2, 2001, 2059–2062

View Online
Table 1 Rate constants at 25 ЊC and activation parameters for solvolysis of X-benzyl azoxy(4-Y)benzenesulfonates (1a–l) in 1 : 1 (v/v) 2,2,2-
trifluorothanol–water
∆S‡/J KϪ1
Substituent X
4-Substituent Y
Compound
105k/sϪ1 (25 ЊC)
∆H‡/kJ molϪ1
molϪ1
3-Cl
3-Cl
3-Cl
3-Cl
4-Cl
H
3-Me
4-Me
4-Me
4-Me
4-Me
4-MeO
MeO
Me
Br
CN
Me
Me
Me
MeO
Me
Br
(1j)
0.013
0.031
0.074
0.14
0.232
0.467
0.764
4.24
5.95
19.1
46.5
174
114
107
109
114
103
100
101
95
92
87
89
86
4
Ϫ11
4
26
Ϫ6
Ϫ12
Ϫ6
Ϫ11
Ϫ17
Ϫ25
Ϫ9
(1f )
(1k)
(1l)
(1e)
(1a)
(1d)
(1g)
(1c)
(1h)
(1i)
CN
Me
(1b)
Ϫ9
Table 2 Products of solvolysis of X-substituted benzyl azoxytoluene-p-sulfonates in 1 : 1 (v/v) trifluoroethanol–watera
Substituent X (Compound)
XC6H4CH2OH (%)
XC6H4CH2OTFEt (%)
XC6H4CHO (%)
XC6H4CH2N(O)NOH (%)
3-Cl (1f )b
74
80
89
23
20
11
0.12
0.01
0.84
4-Me (1c)c
4-MeO (1b)e
<0.01d
<0.01d
<0.01f
a Estimated errors for alcohols and 2,2,2-trifluoroethyl ethers, 3%. b Normalised absolute values; total recovery = 98%. c Normalised absolute
values; total recovery = 93%. d Not detected. e Average relative yields from four reactions each analysed five times. f A trace was detected.
nucleofuge, and (ii) different substituents in the nucleofuge
and a constant 4-methylbenzyl electrofuge, have already been
discussed in a preliminary communication, and indicate a syn-
chronous concerted rate-determining fragmentation.9,14 The
rigorous method of investigating how the ρ-value for substitu-
ents in the nucleofuge is affected by substituents in the electro-
fuge, and vice versa, is to determine the cross interaction
coefficients championed by Lee.15 However, this would have
required the non-trivial synthesis of a wider range of substrates
than was practicable, so we restricted this aspect of our study
to a second ρ(σY)-value determination using 3-chlorobenzyl as
the common electrofuge in place of 4-methylbenzyl. The
simple Hammett plot shown in Fig. 1 for four 3-chlorobenzyl
azoxyarenesulfonates is curved principally because 1j is less
reactive than its simple σY-value predicts. If that single result is
omitted, a fair three point correlation is obtained (ρY = 0.75,
r = 0.99). However, the reduced reactivity of 1j and, to decreas-
ing extents, 1f and 1k (compared with 1c and 1g–i) could be due
to stabilisation of the initial state by resonance interaction
between the 4-substituents in the nucleofuge and the sulfonyl
group whose electron deficiency in 1f and 1j–l is enhanced by
the 3-chloro-substituent in the electrofuge. If so, use of σϩ sub-
stituent parameters in the nucleofuge is justified for 1f and 1j–l
which leads to a good four-point linear correlation, ρ(σYϩ) =
0.71, r = 0.99. Regardless of which correlation is better justified,
however (and we did not have sufficient data to justify use of a
proper Yukawa–Tsuno treatment),16 the ρ-values for substitu-
ents in the nucleofuge of these less reactive 3-chlorobenzyl
azoxyarenesulfonates are much the same (0.71 or 0.75) and
smaller than that obtained for the corresponding reactions of
the more reactive 4-methylbenzyl analogues where ρ(σY) = 1.07
for 1c and 1g–i. This indicates a lower development of negative
charge in the incipient nucleofuge in the transition structures
from 1f and 1j–l than from 1c and 1g–i. This cannot be due to
an earlier transition state in a fully synchronous concerted
fragmentation in the less reactive substrates (a violation of the
Hammond principle),17 but could be due to a decrease in the
synchronicity of the concerted fragmentation. Thus, cleavage
of bond (a) begins before cleavage of bond (b) in 1 with
deactivating substituents in the electrofuge, and the reaction
path in Fig. 1 of our earlier report is curved towards the upper
left for such reactants, i.e. towards route (i).9
The activation parameters for the solvolysis of substituted
benzyl azoxyarenesulfonates shown in Table 1 in 50TFE are
generally similar to those reported earlier for 2-adamantyl
azoxytoluene-p-sulfonate in a range of solvents,6 i.e. large ∆H‡
values and small mostly negative ∆S‡ values. The former are as
expected for reactions in which bond breaking is unassisted by
appreciable bond making. The increase in entropy due to con-
version of the reactant into a loosely bonded transition struc-
ture on the verge of fragmenting into three species is just about
balanced by the loss of entropy associated with the increased
solvation expected as covalent substrate becomes dipolar tran-
sition structure. We also see clearly that the change in reactivity
along the two Hammett series previously reported, for which
ρ(σXϩ) = Ϫ3.27 and ρ(σY) = 1.07, is due entirely to the trends in
∆H‡. For the new Hammett series, σY ca. 0.7 for the 3-chloro-
benzyl azoxyarenesulfonates, the trend in reactivity does not
seem quite so easily attributable to enthalpy or entropy com-
ponents of ∆G‡. The same was observed with 3-chlorobenzyl
arenesulfonates where a satisfactory linear free energy relation-
ship did not allow dissection into interpretable enthalpy and
entropy components.5
Only a very low yield of N-(3-chlorobenzyl)hydroxydiazene
oxide (<1%) is obtained from 1f indicating a very low extent of
nucleophilic attack by solvent at the sulfonyl group in contrast
to what would be observed under basic conditions.8 There is no
evidence of bimolecular reactions from 1c and 1b. As reactivity
increases along the series 1f, 1c, and 1b, fragmentation leads to
increasingly stable benzyl carbenium ions. This series may be
unique in that all three first-formed intermediate electrophiles
are unambiguously benzylic carbocations. In previous investi-
gations into the selectivity of benzylic electrophiles, it has not
always been demonstrated that the products are derived from
carbenium ions in step-wise reactions. Some benzyl arene-
sulfonates, for example, are now known to react by a solvent-
induced SN2 mechanism (in which rate-limiting and product-
determining steps are one and the same)5 rather than by an SN1
mechanism involving benzyl carbenium ions.3 The increasing
stabilities of carbenium ions from the present series of sub-
strates (1f, 1c, and 1b) appear to be reflected in increasing selec-
tivities for reaction with water rather than TFE. The ratio of
alcohol : ether ranges from 3.2 for 1f (i.e. an apparent selectivity
in favour of TFE) to 4.0 for 1c and 8.1 for 1b. For the first two,
J. Chem. Soc., Perkin Trans. 2, 2001, 2059–2062
2061

View Online
the selectivities are lower than for the corresponding benzylic
arenesulfonates (7.3 for 3-chlorobenzyl arenesulfonates and 4.9
for 4-methylbenzyl toluene-p-sulfonates). However, we know
that about half of the initially formed benzyl carbenium
ion from the parent benzyl azoxytoluene-p-sulfonate (1a) is
trapped by the adjacent nitrous oxide molecule to give the
benzyloxydiazonium ion, 2 in Scheme 1.8 This then leads on to
give principally a further solvent-derived substitution product
and, in the absence of a base, only a very low yield of benzalde-
hyde. Thus, we are unable to ascribe the present overall ratio
of alcohol to ether wholly to the selectivity of the benzylic
carbenium ion, and we have insufficient data from the
reactions of 1f, 1c, and 1b to calculate the extents of involve-
ment of the corresponding substituted benzyloxydiazonium
ions (2).
Spherisorb) HPLC was used with manual injections (Rheodyne
valve with 20 µl loop), aqueous methanol flow rate = 1.5 cm3
minϪ1; the UV detector was set at 257 nm.
Acknowledgements
We thank EPSRC for an equipment grant (H. M.) and a
studentship (I. M. G.), and Professor M. Mishima (Kyushu
University) for helpful discussions.
References
1 A. Williams, Concerted Organic and Bio-organic Mechanisms,
CRC Press, Boca Raton, 2000.
2 C. K. Ingold, Structure and Mechanism in Organic Chemistry,
Bell, London, 2nd edn., 1969.
3 G. S. Hammond, J. Peloquin, F. T. Fang and J. K. Kochi, J. Am.
Chem. Soc., 1960, 82, 443; J. K. Kochi and G. S. Hammond, J. Am.
Chem. Soc., 1953, 75, 3445.
4 M. Fujio, T. Susuki, M. Goto, Y. Tsuji, K. Yatsugi, Y. Saeki,
S. H. Kim and Y. Tsuno, Bull. Chem. Soc. Japan, 1994, 67, 2233;
M. Fujio, T. Susuki, K. Yatsugi, Y. Saeki, M. Goto, S. H. Kim,
Y. Tsuji, Z. Rappoport and Y. Tsuno, Bull. Chem. Soc. Japan, 1995,
68, 22619; Y. Tsuji, S. H. Kim, Y. Saeki, K. Yatsugi, M. Fujio and
Y. Tsuno, Tetrahedron Lett., 1995, 36, 1465.
5 H. Maskill, J. Chem. Soc., Perkin Trans. 2, 1986, 1241; I. M. Gordon
and H. Maskill, J. Chem. Soc., Perkin Trans. 2, 1991, 1951.
6 H. Maskill, J. T. Thompson and A. A. Wilson, J. Chem. Soc., Perkin
Trans. 2, 1984, 1693; J. K. Conner, J. Haider, H. Maskill,
M. Pestman and M. N. S. Hill, Can. J. Chem., 1998, 76, 862. If the
conjugate base of the solvent is generated and does lead to a
bimolecular reaction, the site of attack is at the sulfonyl, so the
products are different and do not over-complicate investigation of
the solvolytic process; see H. Maskill, J. Chem. Soc., Chem.
Commun., 1986, 1433.
7 J. K. Conner and H. Maskill, Bull. Soc. Chim. Fr., 1988, 342.
8 H. Maskill and W. P. Jencks, J. Chem. Soc., Chem. Commun., 1984,
944; H. Maskill and W. P. Jencks, J. Am. Chem. Soc., 1987, 109,
2062.
Experimental
TFE (Fluorochem) was refluxed over polyphosphoric acid for
1.5 h then fractionally distilled from molecular sieves (type 3A).
Preparative details and structural assignments are available as
electronic supplementary information.‡
Kinetics
Rates of solvolysis in 1 : 1 (v/v) TFE–H2O were measured by
monitoring the decrease in UV absorbance at a suitable wave-
length in the range 230–260 nm in the thermostatted cell com-
partment of a Pye-Unicam SP-800 spectrophotometer fitted
with a platinum resistance thermometer and interfaced to an
Apple microcomputer.5,6 Initial substrate concentrations were
usually ca. 10Ϫ5 mol dmϪ3. Each reaction was monitored for 4–5
half-lives with the collection of 80–100 absorbance readings.
First-order rate constants were calculated by a non-linear least
squares programme; standard deviations on individual rate
constants were generally <1% and reproducibility was normally
3%. Activation parameters and rate constants at a common
temperature of 25.0 ЊC were calculated by a computer version
of the Eyring equation using average rate constants from tripli-
cate determinations at each of four temperatures covering a
30 degree range.
9 I. M. Gordon and H. Maskill, J. Chem. Soc., Chem. Commun., 1989,
1358.
10 R. F. Borch, M. D. Bernstein and H. D. Durst, J. Am. Chem. Soc.,
1971, 93, 2897.
11 R. S. Tipson, J. Org. Chem., 1944, 9, 235.
12 T. E. Stevens, J. Org. Chem., 1964, 29, 311.
13 H. Maskill, The Physical Basis of Organic Chemistry, Oxford
University Press, 1985.
14 4-Cyanobenzyl azoxytoluene-p-sulfonate was also prepared and
fully characterised, but did not yield a solvolytic rate constant;
presumably, its very slow solvolytic reaction was compromised by
competing non-solvolytic modes of decomposition.
Product analysis5,8
HPLC was carried out using Gilson instrumentation and a UV
detector connected to a Pye-Unicam PU 4810 computing inte-
grator. Glass distilled water and HPLC grade methanol were
filtered before use. The three substituted benzyl azoxytoluene-p-
sulfonates were each solvolysed at least twice for at least 10 half-
lives, and each reaction mixture was analysed directly at least
5 times to give averaged product analyses. Reverse phase (C-18
15 I. Lee, Adv. Phys. Org. Chem., 1992, 27, 57.
16 Y. Tsuno and M. Fujio, Adv. Phys. Org. Chem., 1999, 32, 267.
17 G. S. Hammond, J. Am. Chem. Soc., 1955, 77, 334; D. Fa˘rcasiu,
J. Chem. Educ., 1975, 52, 76; H. Maskill, Structure and Reactivity in
Organic Chemistry, Oxford University Press, 1999.
2062
J. Chem. Soc., Perkin Trans. 2, 2001, 2059–2062