Water and hydroxide ion pathways in the σ-complexation of superelectrophilic 2-aryl-4,6-dinitrobenzotriazole 1-oxides in aqueous solution. A kinetic and thermodynamic study

Article abstract of DOI:10.1039/b201731h

As part of our continuing studies of the highly electron-deficient nature of nitrobenzofuroxans, nitrobenzofurazans and related heterocycles, we report here a kinetic and thermodynamic study of σ-complexation for a series of 2-aryl-4,6 -dinitrobenzotriazole 1-oxides (3a-e) over a large pH range in aqueous solution. The reaction series represents a modulation in electrophilic properties of the benzotriazole moiety in formation of the corresponding hydroxy σ-adducts (4a-e). Analysis of the data has allowed dissection of observed rates into forward (k₁^H2O, k₂^OH-) and reverse (k_-1^H+, k_-2) rate constants as well as the obtention of pK_a values for H₂O addition to the benzotriazole moiety. Our results reveal that 3a-e are superelectrophilic compounds with respect to 1,3,5-trinitrobenzene (TNB) as a standard electron-deficient aromatic, but less superelectrophilic compared to 4,6-dinitrobenzofuroxan (DNBF). Some data pertaining to buffer catalysis of the formation and decomposition of the adducts together with solvent deuterium isotope effects for these pathways are also reported. From these results, it is concluded that adduct formation occurs via general base catalyzed water attack: the general bases include notably H₂O, HCO₃^-, CO₃^2- as well as OH^-. This contrasts with the situation for the σ-complexation of DNBF where HCO₃^- and CO₃^2- were found to act as nucleophilic catalysts whereas OH^- functioned as a general base catalyst. This contrasting behaviour provides further evidence that the dinitro-activated carbocyclic ring of the benzotriazoles 3a-e ranks somewhat lower in electrophilic/superelectrophilic properties compared to that in DNBF. Altogether, the results provide a basis for understanding the relationship between the superelectrophilic reactivities, as evidenced by the contrasting kinetic and thermodynamic properties of the systems at hand, and the varied abilities Of these substrates to react in pericyclic Diels-Alder reactions.

Full text of DOI:10.1039/b201731h

View Article Online / Journal Homepage / Table of Contents for this issue
Water and hydroxide ion pathways in the ꢀ-complexation of
superelectrophilic 2-aryl-4,6-dinitrobenzotriazole 1-oxides in
aqueous solution. A kinetic and thermodynamic study
2
Taoufik Boubaker,a,b Alain Pierre Chatrousse,b François Terrier,b Bahoueddine Tangour,a
Julian M. Dustc and Erwin Bunceld
a Unité de Recherche de Physico-Chimie Moléculaire, Faculté des Sciences de Monastir,
Avenue de l’Environnement 5019 Monastir, Tunisie
b Laboratoire SIRCOB, Institut Lavoisier-Franklin, Université de Versailles,
45, Avenue des Etat-Unis, 78035-Versailles cedex, France
c Department of Chemistry, Sir Wilfred Grenfell College,
Memorial University of Newfoundland, Corner Brook, NF A2H 689, Canada
d Department of Chemistry, Queen’s University, Kingston, ON K7L 3N6, Canada
Received (in Cambridge, UK) 15th February 2002, Accepted 21st June 2002
First published as an Advance Article on the web 24th July 2002
As part of our continuing studies of the highly electron-deficient nature of nitrobenzofuroxans, nitrobenzofurazans
and related heterocycles, we report here a kinetic and thermodynamic study of σ-complexation for a series of 2-aryl-
4,6 -dinitrobenzotriazole 1-oxides (3a–e) over a large pH range in aqueous solution. The reaction series represents
a modulation in electrophilic properties of the benzotriazole moiety in formation of the corresponding hydroxy
Ϫ
σ-adducts (4a–e). Analysis of the data has allowed dissection of observed rates into forward (kH1 O, k 2OH ) and reverse
2
ϩ
(kϪH1, kϪ2) rate constants as well as the obtention of pKa values for H2O addition to the benzotriazole moiety. Our
results reveal that 3a–e are superelectrophilic compounds with respect to 1,3,5-trinitrobenzene (TNB) as a standard
electron-deficient aromatic, but less superelectrophilic compared to 4,6-dinitrobenzofuroxan (DNBF). Some data
pertaining to buffer catalysis of the formation and decomposition of the adducts together with solvent deuterium
isotope effects for these pathways are also reported. From these results, it is concluded that adduct formation
occurs via general base catalyzed water attack: the general bases include notably H2O, HCO3 , CO32Ϫ as well as OHϪ.
Ϫ
This contrasts with the situation for the σ-complexation of DNBF where HCO3Ϫ and CO32Ϫ were found to act as
nucleophilic catalysts whereas OHϪ functioned as a general base catalyst. This contrasting behaviour provides further
evidence that the dinitro-activated carbocyclic ring of the benzotriazoles 3a–e ranks somewhat lower in electrophilic/
superelectrophilic properties compared to that in DNBF. Altogether, the results provide a basis for understanding the
relationship between the superelectrophilic reactivities, as evidenced by the contrasting kinetic and thermodynamic
properties of the systems at hand, and the varied abilities of these substrates to react in pericyclic Diels–Alder
reactions.
strong electrophiles as benzenediazonium cations, including
Introduction
the p-nitrobenzenediazonium cation.10
The last decade has witnessed considerable interest in studies
of nitrobenzofuroxans, a class of electron-deficient heteroaro-
matic compounds that show increased reactivity with nucleo-
philes in the formation of σ-bonded anionic (Meisenheimer)
(1)
complexes.1–8 The high susceptibility of 4,6-dinitrobenzo-
furoxan (DNBF, 4,6-dinitrobenzo[c][1,2,5]oxadiazole 1-oxide)
to undergo σ-complexation in the absence of added base in
aqueous solution is illustrative of this behaviour.1 The pKa for
the formation of the hydroxy adduct 1 according to eqn. (1) is
equal to 3.75 at 25 ЊC, as compared with a pKa value of 13.37
(2)
for formation of the analogous adduct 2 of 1,3,5-trinitro-
benzene (TNB),9 the conventional reference aromatic electro-
phile in σ-complex chemistry.1–8 Use of dilute alkali hydroxide
solutions is in fact necessary to achieve the formation of 2 in
aqueous solution (eqn. (2)). More importantly, DNBF has
been found to react quantitatively at room temperature with
weak nucleophiles such as anilines or hydroxy or methoxy-
substituted benzenes to give stable σ-bonded adducts which are
formally the products of SEAr substitution of the benzene
rings.2–4 Similar substitutions readily occur with π-excessive
heteroaromatics like pyrrole, thiophene, furan or indole deriv-
atives. From kinetic studies of these reactions, it has been
shown that DNBF is in fact more electrophilic than such
In view of the numerous synthetic, biological and analytical
applications which have derived from the discovery of the
superelectrophilic character of nitrobenzofuroxans,6–8,10–14
efforts have been made to design other families of hetero-
aromatic compounds that will show similar enhanced reactivity
with nucleophiles in σ-complex formation and related nucleo-
philic aromatic substitution processes. Thus, promising results
have been reported regarding the 2-aryl-4,6-dinitrobenzo-
1
triazole 1-oxides 3a–c.15,16 On the basis of H and 13C NMR
studies of their reactions with such nucleophiles as ethanethiol,
DOI: 10.1039/b201731h
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633
This journal is © The Royal Society of Chemistry 2002
1627

View Article Online
phenoxide ions and arylamines, compounds 3a–c have been
qualitatively ranked much closer to DNBF than to TNB on a
scale of increasing electrophilicity in Meisenheimer complex
formation.16 However, no detailed kinetic and thermodynamic
studies of the reactivity of these compounds has been made. In
particular, it was of interest to establish whether 3a–c are sus-
ceptible to covalent hydration through water attack since the
importance of this pathway in the formation of the DNBF
adduct 1 (eqn. (1)) has been used as a major criterion to
highlight the superelectrophilic character of this substrate.1,8
In this paper, we report the results of a thermodynamic and
kinetic study of the reactions of water and hydroxide ion with
3a–c as well as with the 2-phenyl and the 2-(4Ј-bromophenyl)
analogues 3e and 3d to give the hydroxide adducts 4a–e accord-
ing to Scheme 1 in aqueous solution. Our results reveal that
pH rate profiles for the interconversion of 3a–e
Using a stopped-flow spectrophotometer, the interconversions
of 3a–e and 4a–e were kinetically investigated under first-order
conditions with a substrate or adduct concentration of 2–3 ×
10Ϫ5 mol dmϪ3. In agreement with the direct equilibrium
approach described by Scheme 1 only one relaxation time
corresponding to the formation (pH > pKa) or decomposition
(pH < pKa) of the adducts 4a–e was observed in all cases.
The logarithmic values of the observed first order rate constant
kobsd for the combined formation and decomposition of 4a–e at
25 ЊC are plotted in Figs. 1 and 2 as a function of pH. In the
Fig. 1 pH dependence of kobsd (sϪ1) for the formation and decom-
position of the adducts 4a, 4c and 4e in water; T = 25 ЊC, I = 0.2 mol
dmϪ3
.
Scheme 1
3a–e are definitely somewhat less reactive than the super-
electrophile DNBF but considerably more reactive than the
conventional electrophile TNB. Some interesting buffer
catalysis and deuterium isotope effect results will be also dis-
cussed. A few data regarding the hydroxide pathway leading
to the adducts 4a, 4c and 4e have been reported previously.17
The relationship between the kinetic and thermodynamic
σ-complexation results of the benzofuroxan and benzotriazole
series to the observed pericyclic reactivities of these substrates
is also discussed.
Results
Fig. 2 pH dependence of kobsd (sϪ1) for the formation and decom-
position of the adducts 4b and 4d in water; T = 25 ЊC, I = 0.2 mol dmϪ3
All rate and equilibrium measurements pertaining to Scheme 1
were made at 25 ЊC and constant ionic strength of 0.2 mol dmϪ3
maintained with KCl. Dilute hydrochloric acid solutions, vari-
ous buffer solutions and dilute potassium hydroxide solutions
were used to cover a pH range of 0.8–13.0. To be noted is that
all pH values have been measured relative to the standard state
in pure water. Accordingly the relation [Hϩ] = 10ϪpH/γ holds
with γ being the mean activity coefficient in 0.2 mol dmϪ3 KCl
(γ = 0.75 at 25 ЊC ).1a
.
experiments where buffer catalysis was observed (vide infra), the
kobsd values used to draw the pH-rate profiles were those
extrapolated to zero buffer concentration.
The observed rate constants at each pH are, of course, the
sum of the individual first-order rate constants for formation
(kf) and decomposition (kd) of the adducts.
pKa values of 3a–e
kobsd = kf ϩ kd
(6)
In contrast with data for the covalent hydration of 3a, 3c and 3e
(vide infra Table 1),17 the pKa values for the σ-complexation of
3b and 3d according to eqn. (3) were unknown. Using appro-
priate buffer solutions, these pKa’s were therefore determined
from the observed absorbance variations at λmax of the resulting
adducts 4b and 4d (λmax = 475 nm) obtained at equilibrium as a
function of pH. These actually describe clear acid–base type
equilibrations, as evidenced by the observation of good straight
lines with unit slopes fitting eqn. (5). From these plots (not
shown), we readily obtained: pKa3b = 7.15 0.05, pKa3d = 9.87
0.05 at I = 0.2 mol dmϪ3.
As shown previously in related studies of the covalent
hydration of DNBF and various heterocyclic cations,1a,18 values
of kf and kd can be readily obtained from kobsd using eqns. (7)
and (8):
(7)
(8)
(5)
1628
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633

View Article Online
Table 1 Kinetic and thermodynamic parameters for the formation and decomposition of hydroxy σ-adducts in aqueous solutiona
ϩ
Ϫ
Adduct
pKa
kH O/sϪ1
kHϪ1/dm3 molϪ1 sϪ1
kOH /dm3 molϪ1 sϪ1
k
Ϫ2/sϪ1
K2/dm3 molϪ1
2
1
2
1b
3.75
13.43
6.70
3.45 × 10Ϫ2
—
146
—
4215
(10900)
7050
(18000)
1.33 × 104
(3.57 × 104)
2.63 × 104
3.30 × 104
OHϪ
3.35 × 104
37.5
392
(435)
1000
(1010)
680
(715)
450
2.5 × 10Ϫ6
9.8
1.34 × 1010
3.82
2c
4ad
1.13 × 10Ϫ3
(6.3 × 10Ϫ4
6.73 × 10Ϫ4
1.96 × 10Ϫ5
2 × 107
)
(1.28 × 10Ϫ5
1.42 × 10Ϫ4
)
4b
7.15
9.00
7.04 × 106
1.92 × 105
4cd
1.80 × 10Ϫ5
3.54 × 10Ϫ3
4de
4ed
9.87
10.73
4.78 × 10Ϫ6
8.3 × 10Ϫ7
1.7 × 10Ϫ2
0.17
2.65 × 104
1860
317
a T = 25 ЊC; I = 0.2 mol dmϪ3 KCl. b Ref. 1a. c Ref. 9. d pKa, k
H2O–DMSO mixture because of a too low solubility of the parent molecule 3d in pure aqueous solution. Values in brackets refer to measurements
and kϪ2 values taken from ref. 17. e Rate and equilibrium data refer to a 90 : 10 (v/v)
2
in D2O.
Each of the corresponding pH profiles, which are shown in
Buffer catalysis and isotope effects
Figs. 3 and 4 is nicely consistent with eqns. (9) and (10):
No appreciable catalysis of the interconversion of 3a–e and
4a–e was found to occur in buffers of pKa < 7, i.e. the formic
acid, benzoic acid, acetic acid, succinic acid and dihydrogen
phosphate buffers, at least under the experimental conditions
used. These are exemplified for the acetic acid–acetate buffers in
Fig. 5 which refers to the N-picrylbenzotriazole system (3a). In
(9)
(10)
Fig. 5 Plots showing lack of dependence of kobsd in acetic buffers at
various pH in the N-picrylbenzotriazole system 3a.
contrast, notable catalysis was observed for the formation of
Fig. 3 pH dependence of the rate constants kd (sϪ1) for decomposition
of 4a–e in water; T = 25 ЊC, I = 0.2 mol dmϪ3
.
the adducts 4a, 4b and 4c in carbonate buffers. In these
2Ϫ
instances, only catalysis by the dianionic species CO3
occurred, as evidenced by the finding of excellent and parallel
linear plots fitting eqn. (11) on plotting kobsd vs. the CO32Ϫ con-
centration at the different pH values studied (Fig. 6).
2Ϫ
kobsd = kOH[OHϪ] ϩ kCO [CO3
]
(11)
2Ϫ
3
2
2Ϫ
From the slopes of these plots, we readily obtained kCO
=
3
2Ϫ
16 molϪ1 dm3 sϪ1 for 4a, kCO = 16 molϪ1 dm3 sϪ1 for 4b,
3
2Ϫ
kCO = 3.1 molϪ1 dm3 sϪ1 for 4c. Within experimental error, the
3
intercepts agreed well with the values of the kO2 H [OHϪ] term, as
Ϫ
calculated at each pH using the kO2 H values (Table 1) deter-
mined from the straight lines of slope ϩ1 in Figs 1 and 2.
Ϫ
Similar catalysis by HCO3 was observed for the 3a and 3b
systems in bicarbonate buffers (pH = 8.38) where the formation
of the adducts 4a and 4b is essentially complete (pH ≥ pKa ϩ
Ϫ
Ϫ
1.2); kHCO = 0.78 molϪ1 dm3 sϪ1 for 4a; kHCO = 0.12 molϪ1
dm3 sϪ1 for 4b. In the other systems, 3c–e, where the decom-
position pathway is largely favored in these buffers (pK3ac–e ≥ 9),
no appreciable catalysis by HCO3Ϫ could be observed.
Only the formation and decomposition of the most stable
adduct 4a has been studied in detail in D2O, using dilute NaOD
or DCl solutions as well as various buffer solutions. The solvent
isotope effects pertaining to the different pathways involving 4a
3
3
Fig. 4 pH dependence of the rate constants kf (sϪ1) for formation of
4a–e in water; T = 25 ЊC, I = 0.2 mol dmϪ3
.
Ϫ
ϩ
in which the rate constants kH1 O, k2OH , kϪH1 and kϪ2 refer to the
various reactions depicted in eqns. (3) and (4). Least-squares
fitting of kf and kd to eqns. (9) and (10) gave the parameters
which are collected in Table 1.
2
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633
1629

View Article Online
Table 2 Deuterium isotope effects for formation and decomposition
of the hydroxy σ-adducts 1 and 4a–ca
markedly from 3a, 3b to the less activated N-phenyl substrate 3e
(pKa = 10.73). The electrophilic character of 3e ranks above that
of TNB by only 2.70 pK units, being somewhat below that of
other aromatic electrophiles like 1,3,6,8-tetranitronaphthalene
(pKa = 9.96)20a 1,2,3,5-tetranitrobenzene (pKa = 9.62)20b or
4-nitrobenzofuroxan (pKa = 10.32).17
The effect of substitution of the N-phenyl ring on the sus-
ceptibility of 3a–e to covalent hydration calls for some
additional comments. The nearly similar pKa values of 3a and
3b suggest that the activating effect exerted by the picryl and
2,4-dinitrophenyl rings is largely derived from inductive/field
effects rather than conjugation effects. This idea agrees with the
results of a X-ray analysis which has shown that the two ring
systems of 3a, and hence of the related adduct 4a, are in almost
orthogonal orientation due to the strong steric compression of
the adjacent ortho-nitro and N-oxide functions.15a In contrast, it
is not unreasonable to assume that some conjugation between
the 2-nitrogen and the para-substituent of the benzene ring
occurs in 3c–e. This would lead in the case of 3e to some
electron-donation of the phenyl ring to the heterocyclic frame-
work, accounting for the lowest acidity of this compound.
ϩ
Ϫ
Ϫ
Dϩ
Ϫ1
2
2
Adduct
kH1 O/kD1 O
kHϪ1 /k
kO2 H /kO2 D
kϪH2 /kϪD2
1b
4a
4b
4c
1.67
1.79
—
0.38
0.386
0.39
0.905
0.90
0.99
0.95
1.69
1.53
—
—
0.372
—
a T = 25 ЊC; I = 0.2 mol dmϪ3 b Ref. 1a.
.
Reactivity of 3a–e
The high susceptibility of DNBF to covalent hydration through
water attack has been the major feature as a result of which this
molecule has been accorded superelectrophilic properties.1,8,21,22
Such a water reaction does not contribute at all to the form-
ation of the TNB–hydroxide adduct 2.9 Interestingly, neither
could a water pathway be detected in the formation of the
adducts of 1,3,6,8-tetranitronaphthalene and 1,2,3,5-tetranitro-
benzene 5 and 6 whose stabilities are of the same order as those
of the adducts 4c–e.19,20 As a matter of fact, Table 1 reveals that
a similar situation prevails for these three complexes. In the case
Fig. 6 Effect of carbonate concentration and pH on kobsd for
formation of the adduct 4a in water; T = 25 ЊC, I = 0.2 mol dmϪ3
.
Ϫ
Ϫ
are given in Table 2. Also given in this Table are the kO2 H /k2OD
ϩ
ϩ
and kϪH1 /kϪD1 values measured, respectively, for the formation and
decomposition of the adducts 4b and 4c
of 4c, for example, the first-order rate constant kH1 O for adduct
2
formation is about 175 times lower than the first-order rate
constant kϪ2 for its spontaneous decomposition. This precludes
Discussion
Our study of substrates 3a–e, representing an increasingly
electrophilic family of electron-deficient heteroaromatic
compounds, has revealed a number of interesting features,
especially when compared to the known superelectrophile
DNBF and the standard electrophile TNB. Salient features
which derive from the results in Tables 1 and 2 and the buffer
catalysis effects described in part in Figs. 5 and 6 can be high-
lighted under the following headings.
even a minor contribution of the water pathway to the form-
Ϫ
ation of 4c under the most favorable conditions where the k2OH
[OHϪ] pathway contributes to minimal extent. In agreement
with this conclusion, the pKa value for formation of 4c appears
at the intercept between the plateau (kϪ2) and the straight line
Ϫ
of slope ϩ1 (kO2 H [OHϪ]) of the related pH-rate profile in Fig. 1.
Relative stabilities of hydroxy adducts
Table 1 reveals that the pKa values for the conversion of 3a–e
into the adducts 4a–e fall in the range 6.70–10.73. Interestingly
the two most activated N-picryl and N-(2,4-dinitrophenyl)
compounds, 3a and 3b, have pKa values (6.70 and 7.15 respect-
ively) which are more than six units lower than the pKa value of
TNB. While this corresponds to a considerable increase in
electrophilic character from that of the common reference
electrophile in σ-complex chemistry, it is rather far from being
comparable to that of DNBF: the hydroxy adducts 4a and 4b
are about 103 less stable than the DNBF analogue. Substituting
the oxygen atom of the annelated ring of DNBF for a less
electronegative N-phenyl group thus reduces the adduct
stability. This situation is reminiscent of the one observed upon
substitution of the oxygen atom for a sulfur or selenium atom
in this heterocycle and it can be explained in terms of the
aromatic character of the parent molecules which increases in
the order O < N–Ph, and correspondingly, of the electron-
withdrawing effect of the annelated ring which decreases in the
order O > N–Ph.19
A similar situation is observed for the formation of the two
less stable adducts 4d and 4e. In contrast, Figs. 1 and 2 show
that a different situation holds for the 2,4-dinitrobenzotriazole
adduct 4b for which the kH O and kϪ2 values are of the same
2
order of magnitude and particularly for the picryl adduct 4a
where the kH1 O value is 40 times greater than kϪ2. In this
2
instance, the pKa value for formation of 4a is seen at the inter-
cept of the plateau for kH1 O and the straight line of slope Ϫ1
2
Hϩ
Ϫ1
(k [Hϩ]) of the pH–rate profile in Fig. 1. It remains that the
kH O value for this electrophilic benzotriazole is 40 times lower
2
1
than for DNBF. These results confirm that the 4,6-dinitro-
benzotriazole series is appreciably below the 4,6-dinitrobenzo-
furoxan in terms of electrophilic character. Concomitantly, the
pH range where the water pathway contributes predominantly
to the formation of 4a is very much reduced compared to the
situation observed in the DNBF system.
On the other hand, it is noteworthy that nitro-substitution
of the N-bonded phenyl ring has a major effect on the ease
of σ-complexation of the 4,6-dinitrosubstituted benzotriazole
moiety as evidenced by the finding that the pKa values increase
Using again the data for DNBF and TNB as the references,
Ϫ
Table 1 shows that the kO2 H rate constants for the hydroxide ion
attack to the five benzotriazoles 3a–e are intermediate between
the two extremes, with rather moderate variations on going
1630
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633

View Article Online
Scheme 2
from 3a to 3e. Because of the especially large steric compression
reactions will therefore go through a transition state such as
which exists between the two picryl and benzotriazole rings, the
activating effect exerted by the picryl moiety on the susceptibil-
ity of the carbocyclic ring of 3a to covalent hydration is reduced
7. Several authors have discussed the occurrence of such
hydroxide catalyzed water attack in reactions of carbonyl
compounds23–27 and evidence for a similar catalytic behaviour
of OHϪ has also been reported in various reaction systems,
including SNAr and related reactions.28–32 Interestingly, our
relative to that of the N-(2,4-dinitrophenyl) compound.15a This
Ϫ
accounts for the observed 2.5-fold decrease in kO2 H on going
from 3a to 3b and the absence of a linear Brönsted correlation
incorporating all the compounds studied. On the other hand, it
is the variations in the kϪ2 value, measuring the tendency of the
adducts 4a–e to undergo spontaneous decomposition, which
govern the changes observed in the K2 values. In this regard, it is
noteworthy than the kϪ2 value for the most stable adduct 4a is
one order of magnitude greater than that for the DNBF adduct
1. Then, kϪ2 increases regularly on going from 4a to 4e, i.e., as
the electron-withdrawing character of the N-bonded phenyl
observed solvent isotope effect of kH1 O/k1D O = 1.79 for water
attack on 3a is also similar in magnitude to that measured for
water attack on DNBF, being too large to support a transition
state which would not involve rate-limiting proton transfer.
This in turn favors structure 8 which represents general-base
catalysed water attack with a second water molecule acting as a
base catalyst. Consistent with this structure, the isotope effect
2
2
ϩ
Dϩ
Ϫ1
kϪH1/k is in the range of values reported for a number of
examples of authentic general acid catalysis, e.g., the hydrolysis
moiety decreases. A similar situation follows when changes are
of orthoesters and some acetals and ketals.33
ϩ
considered in the rate constant kϪH1 for the Hϩ catalyzed decom-
position of 4a–e.
Buffer catalysis and isotope effects
Due to the reduced contribution of the water pathway to the σ-
complexation process of 3a–e, the set of base catalysts involved
in the formation of the adducts is restricted to OHϪ, CO32Ϫ and
Ϫ
HCO3 for the N-picryl and N-(2,4-dinitrophenyl) systems, to
2Ϫ
OHϪ and CO3 for the N-(4-nitrophenyl) system and to OHϪ
for the unsubstituted and N-(4-bromophenyl) systems. A most
interesting question is, however, whether these various catalysts
act as general base or nucleophilic catalysts.
2Ϫ
Returning to the nucleophilic catalytic behaviour of CO3
In the case of DNBF, we previously suggested that OHϪ
ions identified in the DNBF system, we note in Table 3 that
Ϫ
Ϫ
2Ϫ
behaves as a general base catalyst on the basis that (1) the k2OH
rate constant falls nicely on the Brönsted line defined by other
base catalysts such as borate, p-cyanophenoxide and water; (2)
the observed kinetic solvent isotope effect for the reverse direc-
tion, i.e. kϪH2/kϪD2 was equal to 1.69, consistent with a proton
transfer being part of a rate-limiting transition state.1a On the
other hand, the points for CO32Ϫ and HCO3Ϫ showed such large
positive deviations from the aforementioned Brönsted line that
the catalytic contribution of these two general bases was inter-
preted in terms of nucleophilic catalysis (Scheme 2).1a This idea
was in accord with the fact that DNBF is known to displace
CO2 from bicarbonate solutions.10,12
there is a strong but regular increase in the kO2 H /kCO ratio on
3
going from DNBF to 3a to 3b to 3c. This decrease in the
catalytic efficiency suggests that we are dealing with a loss in the
ability of CO3 to act as a nucleophilic catalyst rather than
2Ϫ
with a systematic change in the transition state structure associ-
ated with the mechanism of nucleophilic catalysis. In other
words, it seems that the contribution of the nucleophilic path-
way of Scheme 2 may still be operating to some extent in the
case of 3a, the strongest electrophile in the benzotriazole series,
but that the general base mechanism becomes preferred in the
Ϫ
case of 3b and 3c. A similar conclusion is reached for HCO3
Ϫ
Ϫ
from a comparison of the kO2 H /kHCO ratios in Table 3.
3
Considering the hydroxide reactions of the present work, no
reference to a meaningful Brönsted line can be envisioned. In
The above results on buffer catalysis in the dinitrobenzotri-
azole series may be compared with corresponding data
obtained previously for the archetypal superelectrophile, 4,6-
dinitrobenzofuroxan (DNBF). In the latter case it was found
contrast, it is noteworthy that the measured solvent isotope
Ϫ
effects on the kO2 H and kϪ2 pathways for the 3a–3c systems are
closely similar to those measured in the DNBF systems (Table
2). This suggests that we adhere to our previous reasoning and
therefore to the conclusion that hydroxide ion acts as a general
base catalyst for the reactions of water with 3a–e. These
Ϫ
2Ϫ
that HCO3 and CO3
act as nucleophilic catalysts in
σ-complexation whereas OHϪ functioned as a general base
catalyst.1a This different behaviour provides further evidence
that the dinitro-activated carbocyclic ring of the benzotriazoles
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633
1631

View Article Online
Table 3 Rate constants for catalysis of the formation of the adducts 1 and 4a–c by carbonate and bicarbonate speciesa
Ϫ
kOH
/
Ϫ
3
2Ϫ
Ϫ
Ϫ
2Ϫ
3
3
3
Adduct
kCO /dm3 molϪ1 sϪ1
kHCO /dm3 molϪ1 sϪ1
kOH /kCO
kHCO
1b
4a
4b
4c
2370
16
16
58
11.56
24.5
62.5
472
502
8330
—
0.78
0.12
—
3.1
220
a T = 25 ЊC; I = 0.2 mol dmϪ3 b Ref. 1a.
.
3a–e ranks somewhat lower in electrophilic/superelectrophilic
properties.
Council of Canada (NSERC) (E.B.), and the Principal
Research Fund of Sir Wilfred Grenfell College (J.M.D.).
Superelectrophilic vs. pericyclic reactivity in benzofuroxans and
benzotriazoles
References
1 (a) F. Terrier, F. Millot and W. P. Norris, J. Am. Chem. Soc., 1976,
98, 5883; (b) F. Terrier, Chem. Rev., 1982, 82, 77; (c) F. Terrier,
A. P. Chatrousse, Y. Soudais and M. Hlaibi, J. Org. Chem., 1984, 49,
4176.
2 (a) M. J. Strauss, R. A. Renfrow and E. Buncel, J. Am. Chem. Soc.,
1983, 105, 1555; (b) E. Buncel, R. A. Renfrow and M. J. Strauss,
J. Org. Chem., 1987, 52, 488; (c) R. A. Manderville and E. Buncel,
J. Chem. Soc., Perkin Trans. 2, 1993, 1887.
3 (a) R. J. Spear, W. P. Norris and R. W. Read, Tetrahedron Lett., 1983,
24, 1555; (b) W. P. Norris, R. J. Spear and R. W. Read, Aust. J.
Chem., 1983, 36, 297.
4 (a) M. R. Crampton, L. C. Rabbitt and F. Terrier, Can. J. Chem.,
1999, 77, 639; (b) M. R. Crampton and L. C. Rabbitt, J. Chem. Soc.,
Perkin Trans. 2, 1999, 1669; (c) M. R. Crampton, J. Delaney
and L. C. Rabbitt, J. Chem. Soc., Perkin Trans. 2, 1999, 2473;
(d ) M. R. Crampton and L. C. Rabbitt, J. Chem. Soc., Perkin Trans.
2, 2000, 2169.
5 J. Kind and H. Niclas, Synth. Commun., 1993, 23, 1569.
6 M. J. Evgen’yev, S. Y. Garmonov, I. J. Evgen’yeva and L. S.
Gazizullina, J. Anal. Chem., 1998, 53, 571.
7 (a) S. V. Kurbatov, Z. N. Budarina, G. S. Vaslyaeva, N. J. Borisenko,
A. P. Knyazev, V. I. Minkin, Yu. A. Zhdanov and L. P. Olekhnovich,
Izv. Acad. Nauk, Ser. Khim., 1997, 1509; (b) L. P. Olekhnovich,
Z. N. Budarina, A. V. Lesin, S. V. Kurbatov, G. S. Borodkin and
V. I. Minkin, Mendeleev Commun., 1994, 162.
Recent works have highlighted the potential of benzofuroxans
and benzotriazoles families to react not only as electrophilic
heteroaromatics in σ-complexation and related SNAr reactions
but also in pericyclic manner as dienophiles and heterodienes.
In this regard, a noteworthy finding is that whereas 4,6-dinitro
and 4-nitrobenzofuroxans exhibited both types of reactivities,
in contrast, in the 4,6-dinitrobenzotriazole series, only the most
activated derivative, i.e. the picryl one, was found to react in
Diels–Alder processes. In as much as the 2-(2,4,6-trinitro-
phenyl)-4,6-dinitrobenzotriazole 1-oxide structure is the only
one which resembles DNBF in superelectrophilic behaviour, its
capability to undergo Diels–Alder reactions points to a close
relationship between the degree of electron-deficiency of the
carbocyclic ring and the potential pericyclic reactivity of this
ring. This finding suggests therefore that there is a direct
relationship between superelectrophilic and pericyclic reactiv-
ities, a discovery which should be kept in mind in the design
of new types of structures of inherent low aromaticity but
highly prone towards pericyclic reactivity. This important con-
clusion concerning synthetic applications is here anchored on
firmer ground based on the full kinetic and thermodynamic
properties of both nitrobenzofuroxans and the triazole family
of compounds.
8 (a) F. Terrier, in Nucleophilic Aromatic Displacement, ed. H. Feuer,
VCH, New York, 1991, chapters 1 and 2; (b) E. Buncel, J. M. Dust
and F. Terrier, Chem. Rev., 1995, 95, 2261.
9 C. F. Bernasconi, J. Am. Chem. Soc., 1970, 92, 4682.
10 F. Terrier, E. Kizilian, J. C. Halle and E. Buncel, J. Am. Chem. Soc.,
1992, 124, 1740.
11 F. Terrier, M. J. Pouet, J. C. Halle, E. Kizilian and E. Buncel, J. Phys.
Org. Chem., 1998, 11, 707.
12 (a) A. Gasco and A. J. Boulton, Adv. Heterocycl. Chem, 1981, 29,
251; (b) P. B. Ghosh, B. Ternai and M. W. Whitehouse, Med. Res.
Rev., 1981, 1, 159.
13 (a) A. Chattopadhyay, Chem. Phys. Lipids, 1990, 53, 1; (b) J. W.
Nichols, Biochemistry, 1988, 27, 3925.
14 (a) S. Féry-Forgues, J. P. Fayet and A. Lopez, J. Photochem.
Photobiol. A, 1993, 70, 229; (b) S. Féry-Forgues, C. Vidal and
D. Lavabre, J. Chem. Soc., Perkin Trans. 2, 1996, 73.
15 (a) R. A. Renfrow, M. J. Strauss, S. Cohen and E. Buncel, Aust. J.
Chem., 1983, 36, 1843; (b) J. M. Dust and E. Buncel, Can. J. Chem.,
1991, 69, 978.
Experimental
Materials
The benzotriazoles 3a–e were compounds available from previ-
ous studies.15,16 HCl and KOH solutions were prepared from
Titrisol. Buffer solutions were made up from the best available
commercial grades of reagents. Buffers used were formate
(pH 3–4), benzoate (pH 3.6–4.3), acetate (pH 4–5.2), succinate
(pH 4–5.7), phosphate (6–7.5), bicarbonate (pH 8.38), borate
(pH 8.5–9.5), and carbonate (pH 9.5–10.5); solutions were
prepared and their pH measured as described before.1a
Rate and pKa measurements
16 (a) J. M. Dust and E. Buncel, Can. J. Chem., 1994, 72, 218;
(b) J. M. Dust and R. Manderville, Can. J. Chem., 1998, 76, 662.
17 D. Vichard, T. Boubaker, F. Terrier, M. J. Pouet, J. M. Dust and
E. Buncel, Can. J. Chem., 2001, 79, 1617.
18 (a) J. W. Bunting, Adv. Heterocycl. Chem., 1979, 25, 1; (b) J. W.
Bunting and D. Stefanidis, J. Org. Chem., 1986, 51, 2060, 2068.
19 (a) C. Deicha and F. Terrier, J. Chem. Res. (S), 1981, 312; (b) D. Dal
Monte, E. Sandri, L. Di Nunno, S. Florio and P. E. Todesco,
J. Chem. Soc. B, 1971, 2209.
20 (a) J. H. Fendler, E. J. Fendler and L. M. Casillo, J. Org. Chem.,
1971, 36, 1749; (b) M. R. Crampton and M. El Ghariani, J. Chem.
Soc. B, 1970, 391.
21 W. P. Norris and J. Osmundsen, J. Org. Chem., 1965, 30, 2407.
22 R. J. Gaughran, J. P. Picard and J. V. R. Kaufman, J. Am. Chem.
Soc., 1954, 76, 2233.
Stopped-flow determinations were performed on an Applied
Photophysics spectrophotometer, the cell compartment of
which was maintained at 25Њ 0.2 ЊC. Other kinetic and pKa
measurements were made using a Varian Cary 1E spectro-
photometer. All kinetic runs were carried out in triplicate under
pseudo first-order conditions with a substrate concentration of
2 to 3 × 10Ϫ5 mol dmϪ3. In the case of 2e, the measurements
were made in 90% H2O–10% Me2SO (v/v) to overcome solubil-
ity problems but this low Me2SO content does not have any
major influence on the results obtained, including in buffer
solutions.34 All rate constants are accurate to 3%.
Acknowledgements
23 (a) W. P. Jencks, J. Am. Chem. Soc., 1972, 94, 4731; (b) W. P. Jencks
and J Carriuolo, J. Am. Chem. Soc., 1961, 83, 1743.
24 C. D. Ritchie, D. J. Wright, D. Shing-Huang and A. A. Kamego,
J. Am. Chem. Soc., 1975, 97, 1163, 1170.
We are grateful for the financial support of this research by
CNRS (F.T.), the Natural Sciences and Engineering Research
1632
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633

View Article Online
25 R. A. McClelland and M. Coe, J. Am. Chem. Soc., 1983, 105, 2718.
26 C. D. Hall and C. W. Goulding, J. Chem. Soc., Perkin Trans. 2, 1995,
1471.
27 M. H. Fernandez and R. H. de Rossi, J. Org. Chem., 1999, 64, 6000.
28 J. W. Bunting and W. G. Meathrel, Can. J. Chem., 1974, 52, 303.
29 (a) R. H. de Rossi and A. Veglia, Int. J. Chem. Kinet., 1985, 17, 859;
(b) R. H. de Rossi and A. Veglia, J. Org. Chem., 1983, 48, 1879;
(c) R. H. de Rossi and E. B. de Vargas, J. Am. Chem. Soc., 1981, 103,
1540.
30 R. P. Kelly, R. A. More O’Ferrall and M. O’Brien, J. Chem. Soc.,
Perkin Trans. 2, 1982, 211.
31 I. M. Kovach, A. J. Bennet, J. A. Bibbs and Q. Zhao, J. Am. Chem.
Soc., 1993, 115, 5138.
32 T. W. Bentley, P. J. Morris and J. A. Taylor, J. Chem. Soc.,
Perkin Trans. 2, 2000, 2171.
33 T. H. Fife, Acc. Chem. Res., 1972, 5, 264.
34 J. C. Hallé, R. Gaboriaud and R. Schaal, Bull. Soc. Chim. Fr., 1970,
2047.
J. Chem. Soc., Perkin Trans. 2, 2002, 1627–1633
1633