High Bronsted βnuc values in SNAr displacement. An indicator of the SET pathway?

Article abstract of DOI:10.1039/b301031g

Nucleophilic substitutions of 4-chloro-7-nitrobenzofurazan (NBD-Cl) and 3-methyl-1-(4-nitrobenzofurazanyl)-imidazolium ions (NBD-Im⁺) with a series of 4-X-substituted anilines have beeen kinetically investigated in 70-30 (v/v) and 20-80 (v/v) H

Full text of DOI:10.1039/b301031g

View Article Online / Journal Homepage / Table of Contents for this issue
High Brønsted ꢀnuc values in SNAr displacement. An indicator of
the SET pathway?†
François Terrier,*a Malika Mokhtari,a,b Régis Goumont,a Jean-Claude Halléa and
Erwin Buncel*c
a Laboratoire SIRCOB, Institut Lavoisier-Franklin, ESA CNRS 8086, Université de Versailles,
45 Avenue des Etats-Unis, 78035-Versailles cédex, France. E-mail: terrier@chimie.uvsq.fr;
Fax: 00139254452; Tel: 00139254450
b Laboratoire COSNA, Département de Chimie, Université Abou Bekr Belkaid, BP 119,
13000-Tlemcen, Algérie
c Department of Chemistry, Queen’s University, Kingston, ON K7L 3N6, Canada
Received 27th January 2003, Accepted 27th March 2003
First published as an Advance Article on the web 10th April 2003
Nucleophilic substitutions of 4-chloro-7-nitrobenzofurazan (NBD-Cl) and 3-methyl-1-(4-nitrobenzofurazanyl)-
imidazolium ions (NBD-Imϩ) with a series of 4-X-substituted anilines have been kinetically investigated in 70–30
(v/v) and 20–80 (v/v) H2O–Me2SO mixtures. The rate-limiting step in these reactions is nucleophilic addition
with formation of Meisenheimer-type σ-adducts followed by fast expulsion of the leaving group (ClϪ or Im). The
reactions are characterized by a notable sensitivity to basicity of the aniline nucleophiles, with Hammett ρ values
of Ϫ2.68 and Ϫ3.82 in 30% and 80% Me2SO, respectively, for NBD-Cl and even more negative values, Ϫ3.43 and
Ϫ5.27, respectively, for NBD-Imϩ. This is consistent with significant development of positive charge at the nitrogen
atom of the zwitterionic σ-adduct. Unexpectedly, the Brønsted-type plots reveal abnormally high βnuc values, ca. 1.0
and 1.3–1.4, respectively. Satisfactory correlations between the rates of the reactions and the oxidation potentials of
the respective anilines support a SET mechanism for this process, i.e. initial (fast) electron-transfer from the aniline
donor to the nitrobenzofurazan acceptor moiety and subsequent (slow) coupling of the resulting cation and anion
radicals within the solvent cage with formation of the σ-adduct. An alternative possible explanation of the high βnuc
values being related to the strong ϪI effect exerted by the negatively charged 4-nitrobenzofurazanyl structure, which
would induce a greater positive charge at the developing anilinium nitrogen atom in the σ-adduct-like transition state
as compared with the situation in the reference protonation equilibria of anilines, is considered less probable. It is
thus proposed that obtention of abnormal βnuc values may be an indicator of electron-transfer in nucleophilic
aromatic substitution and highlights the transition from the polar (SNAr) to the single electron-transfer (SET)
mechanism.
changes attendant on the formation of the resulting carbon
Introduction
bases.5–8
The significance of structure–reactivity correlations such as
Hammett (ρ) and Brønsted (α,β) coefficients, as probes of reac-
tion mechanisms in nucleophilic substitution or addition reac-
Imbalanced transition states have also been recognized in
nucleophilic reactions.6,9 Thus, a negative βnuc value has been
reported by Jencks for acyl group transfers between p-nitro-
tions, as well as in proton transfer processes, has been elegantly
phenyl phosphates and various quinuclidine bases, which has
discussed by Jencks.1 Thus, α/β values are commonly accepted
been interpreted as reflecting a need for desolvation of the
as measures of the degree of charge transfer, from the base
amine functionality prior to attack at the electrophilic carbon
(nucleophile) to the acid (electrophile) partner, at the transition
center.4 Other solvational imbalances have recently been
state (TS).1,2 In that sense, it could be anticipated that the nor-
mal range of α/β values would be between 0 and 1. However, it
reported in the reactions of p-nitrophenyl acetate with oximate
α-nucleophiles.9,10 At the other extreme, an abnormally high βnuc
was discovered through the work of Bordwell, Jencks, Bernas-
value of 1.6 has been found for the aminolysis of acetylimid-
coni and others,3–5 that certain processes were characterized
azole.11 In this instance, however, the most reasonable explan-
by α/β values out of this normal range. Importantly, such
unexpected α/β values have actually provided new insights into
the understanding of transition state structures. Among other
ation appears to be in terms of a greater sensitivity to polar
substituents of the incoming amino nitrogen in the transition-
state as compared to the reference equilibrium protonation.11
factors, the lack of synchronization in the progress of various
On the other hand, while most SN2 reactions are characterized
events contributing to an elementary reaction has been recog-
by βnuc values in the 0.2–0.5 range,12 values close to or greater
nized as a major source of anomalous α/β values, a situation
than 1 have been observed by Bordwell in some SN2-type reac-
now referred to as a transition-state imbalance.5–9 In this regard,
tions of carbanions and nitranions with sulfonyl- and nitro-
the ionisation of many carbon acids by oxygen or nitrogen
activated halides. These results were interpreted as indicative of
bases, is typical of this behaviour, with the conclusion that the
the occurrence of complete electron transfer.13
proton transfer itself occurs ahead of the resonance/solvation
Regarding SNAr reactions, it has been emphasized by Bord-
well based on the numerous available results of βnuc values fall-
† Electronic supplementary information (ESI) available: first-order
ing in the range 0.5–0.7, that these entail a relatively large trans-
rate constants for the reactions of NBD-Cl (Table S1) and NBD-Imϩ
fer of electronic charge in the TS.13,14 In contrast, two examples
(Table S2) with p-X-substituted anilines and figures showing the influ-
of βnuc values close to 1 have been recently found by Williams
et al. in reactions of amines or phenoxide ions with 4-aryloxy-
1,3,5-triazines.15 Other evidence, however, indicated that these
ence of the oxidation potential on the rates of reaction of NBD-Imϩ
with 1b–e and 1f in 70–30 (Fig. S1) and 20–80 (Fig. S2) (v/v) H2O–
Me2SO. See http://www.rsc.org/suppdata/ob/b3/b301031g/
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1757

View Article Online
reactions proceed through a concerted mechanism rather than
the stepwise SNAr pathway.‡
As part of continuing investigations of SNAr reactions in the
nitrobenzofurazan and nitrobenzofuroxan series,17–20 we report
here a kinetic study of the reactions of para-X-substituted anil-
ines, 1a–f, on the one hand with 4-chloro-7-nitrobenzofurazan,
commonly known as NBD-Cl, and on the other hand with
3-methyl-1-(4-nitrobenzofurazan-7-yl)imidazolium
cations
(NBD-Imϩ) to give the 4-anilino-7-nitrobenzofurazans, 4a–f, at
25 ЊC in aqueous Me2SO solution (eqn. (1)).21 Interestingly,
analysis of the results in terms of Brønsted-type structure–
reactivity relationships reveals that these nucleophilic aromatic
substitutions are associated with some unprecedentedly high
βnuc values. As will be discussed, these results may be regarded
as indicative of the incursion of a single electron transfer-type
SET mechanism in the overall SNAr displacements under
study.§
Fig. 1 Plot of kobsd vs. [aniline; B] for the reaction of NBD-Cl in 70–30
(v/v) H2O–Me2SO; I = 0.2 mol dmϪ3 (Me4NCl); pH = 4.15, ([B]/[BHϩ] =
1); T = 25 ЊC.
(1)
Results
Fig. 2 Plot of kobsd vs. [p-methoxyaniline; B] for the reaction of NBD-
Imϩ in 70–30 (v/v) H2O–Me2SO; I = 0.2 mol dmϪ3 (Me4NCl); pH =
4.66 (᭜), 4.96 (×) and 5.26 (᭹), ([B]/[BHϩ] = 1/2, 1 and 2, respectively);
T = 25 ЊC.
The reactions of NBD-Cl and NBD-Imϩ with the series of
4-X-substituted anilines 1a–f have been kinetically studied at
25 ЊC in 70–30 and 20–80 (v/v) H2O–Me2SO mixtures with the
ionic strength being maintained constant by addition of 0.2 M
Me4NCl (eqn. (1)). The reactions were conducted in buffers
made up from 1a–f, varying the concentrations of the anilinium
and aniline components at constant pH while maintaining
pseudo-first-order conditions with the buffer being in large
excess throughout, i.e. [NBD-Cl] or [NBD-Imϩ] = 3 × 10Ϫ5 mol
dmϪ3, [1a–f]total in the range 2 × 10Ϫ3–0.14 mol dmϪ3. All the
reactions were monitored spectrophotometrically by following
the increase of absorbance at the wavelength corresponding to
the absorption maximum of the resulting 4-anilino-7-nitro-
benzofurazans 4a–f which are all formed quantitatively at com-
pletion of the disappearance of the electrophilic reagents.21
Plots of the first-order rate constants, kobsd, versus the aniline
concentration were linear with negligible intercepts for all reac-
tion systems studied. Examples of these plots are given in Figs 1
and 2 which also show that the points determined at different
pH fall on the same line for a given buffer. This indicates that
kobsd is simply given by eqn. (2), allowing a facile determination
of the second-order rate constants k. These rate constants,
together with the pKa values measured for the aniline reagents
in the two solvent mixtures studied (see Experimental section)
are summarized in Table 1. The complete set of kobsd data
obtained in 1 : 1 buffers are available as supplementary data.†
Discussion
Rate-limiting nucleophilic addition
A major feature of our results is the absence of general base
catalysis in the overall substitution reactions of NBD-Cl and
NBD-Imϩ with the aniline nucleophiles.17,23,24,27 This indicates
that the proton transfer step involved in the reactions is
subsequent to the rate-limiting step. Keeping with the trad-
itional interpretation of nucleophilic aromatic substitution by
amines in general, this behaviour is in accord with the SNAr
mechanism shown in Scheme 1 where rate-limiting form-
ation of the intermediate σ-complexes 2a–f is followed by fast
expulsion of the chloride or imidazole leaving groups.17,24,27
This implies that deprotonation of the adducts 2a–f to
give the conjugate bases 3a–f as well as the subsequent re-
aromatisation of these species are fast processes relative
to the nucleophilic addition step. This situation calls for some
discussion, especially because in the NBD-Imϩ reactions the
basicity of the leaving N-methylimidazole group (pKa = 6.65
and 5.52, in 30 and 80% Me2SO, respectively)28 is in most cases
greater than that of the incoming anilines (pKa in the ranges
3.39–5.80 and 2.37–5.91 in 30 and 80% Me2SO, respectively;
see Table 1).
kobsd = k[1a–f]
(2)
‡ While a βnuc value of 0.91 has been reported for the reaction of
1-chloro-2,4-dinitrobenzene with phenoxides in methanol,16 this value
MeOH
is not very meaningful, since the pKa
the correlation were estimated from the pKa values instead of being
directly measured in methanol.
values used in constructing
H2O
§ It may be noted that nucleophilic substitution in the nitro-
benzofurazan structures should strictly be termed as heteroaromatic
substitution, contrasting with purely aromatic displacements in the
nitrobenzene series. However, ample evidence has been amassed in
the literature, including numerous reviews,17,22–25 that the two types of
systems are subject to very similar principles in terms of reactivity and
mechanism. Therefore, we deem it justified to discuss both our systems
in terms of the SNAr mechanism.
Regarding the proton transfer step corresponding to the con-
version of the anilinium “acids” 2a–f into the conjugate bases
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1758

View Article Online
Table 1 Second-order rate constants (k) for the reactions of NBD-Cl and NBD-Imϩ with p-X-substituted anilines 1a–f in H2O–Me2SO–0.2 mol
dmϪ3 Me4NCl mixtures (eqn. (1))a
102 × k /dm3 molϪ1 sϪ1
σXp
H2O–Me2SO (% v/v)
pKa
L = Cl
L = Imϩ
b
c
p-X-substituted anilines (X)
1a (Cl)
0.23
0
70–30
20–80
70–30
20–80
70–30
20–80
70–30
20–80
70–30
20–80
70–30
20–80
3.39
2.37
4.15
3.31
4.63
4.08
4.96
4.56
5.24
4.87
5.80
5.91
1.35
0.82
3.80
2.93
1b (H)
1.03
1.75
3.87
8.62
9.5
1c (Me)
1d (OMe)
1e (OH)
1f (NH2)
Ϫ0.17
Ϫ0.27
Ϫ0.37
Ϫ0.57
14.3
11.6
30.5
42.7
52.7
161
306
35.4
182
3192
1314
a T = 25 ЊC. b Ref. 26 .c Apparent pKa values of the p-X-substituted anilines referenced to the ternary mixture (see Experimental section); T = 25 ЊC.
Scheme 1
3a–f, it is noteworthy that Crampton and Rabbitt have recently
reported that the deprotonation of the intermediate σ-adducts
of type 5 involved in the addition of various amines to 4-nitro-
benzofuroxan is both a thermodynamically and kinetically
favorable process.29 Based on the finding that the acidifiying
effect exerted by a negatively charged 4-nitrobenzofuroxanyl
structure on the ammonium site directly bonded to the sp3
carbon amounts to 2.5–3 pKa units,29 ¶ the pKa values for
deprotonation of the adducts 2a–f should be in the range 0–3,
as compared with pKa values in the range 2.37–5.91 for the
parent anilinium cations 1H؉,a–f. This is consistent with essen-
tially complete and fast formation of the adducts 3a–f from the
initially formed anilinium complexes 2a–f, notably in the case
of the NBD-Imϩ reactions where the repulsive interaction
between the imidazolium and anilinium moieties in 2a–f
(L = Imϩ) will further contribute to the ease of conversion of
the species into the conjugate bases 3a–f.
That the conversion of the adducts 3a–f into the substitution
products 4a–f is also a fast process compared to the addition
step can then be understood simply in terms of the extremely
poor leaving group ability of an aniline moiety (pKa = 30.6 in
Me2SO)30 as compared with a chloride anion (pKa ∼ Ϫ7) or a
N-methylimidazole molecule (pKa = 6.65 and 5.52 in 30% and
80% Me2SO, respectively).28 The situation is especially relevant
to the NBD-Imϩ reactions since the formation of 4a–f through
spontaneous rearomatisation of the initially formed adducts
2a–f would clearly be more difficult as a result of the greater
basicity of the N-methylimidazole structure relative to most of
the parent anilines 1a–f. However, some contribution of a rapid
and direct decomposition of 2a–f to 4a–f cannot be excluded in
the case of NBD-Cl reactions.
Structure–reactivity correlations
As is apparent from Table 1, increasing the basicity of the anil-
ine nucleophile causes a marked increase in the ease of substi-
tution of NBD-Cl and NBD-Imϩ. This increase is roughly two
orders of magnitude for NBD-Cl and three orders of magni-
tude for NBD-Imϩ. Analysis of these variations in terms of
Hammett plots gives very good correlations (R ≥ 0.99) using σp
substituent constants noting that due to a rather low reactivity
of the related anilines no substituent with an appreciable ϪM
effect is included in this series. In the case of the chloro deriv-
ative, the ρ values, which are equal to Ϫ2.68 and Ϫ3.82 in 30%
and 80% Me2SO, respectively (Table 2), are comparable with
the value of Ϫ3.70 reported by Ryan and Humffray for substi-
tutions of picryl chloride by a similar set of anilines in 75–25
(v/v) EtOH–H2O mixture.31
Rather similar ρ values have also been obtained for these
reactions in methanol (ρ = Ϫ3.50) and in acetonitrile (ρ =
Ϫ3.46).32 The ρ values obtained in the present work are clearly
consistent with significant development of positive charge at
¶ There is evidence that the negatively charged 4-nitrobenzofurazanyl
and 4-nitrobenzofuroxanyl structures exert similar acidifying (ϪI )
effects.17,18b
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1759

View Article Online
Table 2 Hammett (ρ) and Brønsted (βnuc) values for the reaction of
NBD-Cl and NBD-Imϩ with p–substituted anilines in H2O–Me2SO–
0.2 mol dmϪ3 Me4NCl mixtures (eqn. (1))
Moiety
H2O–Me2SO (% v/v)
ρ
βnuc
NBD-Cl
70–30
20–80
70–30
20–80
Ϫ2.68
Ϫ3.82
Ϫ3.43
Ϫ5.27
0.96
0.93
1.43
1.30
NBD-Imϩ
the nitrogen atom of the aniline moiety in the TS for formation
of the zwitterionic intermediate σ-complex 6. Interestingly, the
values determined for the NBD-Imϩ system in a given solvent
mixture are more negative than those for the NBD-Cl system:
ρ = Ϫ3.43 and Ϫ5.27 in 30% and 80% Me2SO, respectively. This
greater sensitivity to substituent changes of the rates of nucleo-
philic addition may be related to the significant repulsive inter-
action between the positively charged imidazolium moiety and
the developing positive charge in the TS for formation of the
corresponding intermediate σ-complexes. Clearly, this repulsive
interaction, represented in structure 7, will be subject to a
greater modulation by the nature of the X substituent in the
aniline moiety. A clear manifestation of this effect is seen in the
data in Table 1 which show that the NBD-Imϩ reactions pro-
ceed faster than the NBD-Cl reactions with the most basic
amines but become slower with the less basic ones.
Fig. 3 Brønsted plots for the reactions of NBD-Cl and NBD-Imϩ
with p-X-substituted anilines 1a–f in 70–30 (v/v) H2O–Me2SO;
I = 0.2 mol dmϪ3 (Me4NCl); T = 25 ЊC. Numbering from Table 1.
Fig. 4 Brønsted plots for the reactions of NBD-Cl and NBD-Imϩ
with p-X-substituted anilines 1a–f in 20–80 (v/v) H2O–Me2SO;
I = 0.2 mol dmϪ3 (Me4NCl); T = 25 ЊC. Numbering from Table 1.
changes of the anilinium reaction site, it is a major feature
of the present system that the negatively charged structure of
4-nitrobenzofuroxan σ-adducts nevertheless exerts an especially
strong ϪI effect, comparable to that of a negatively charged
2,4,6-trinitrocyclohexadienylide structure. As discussed above,
this effect enhances any acidic functionality attached at the sp3
carbon of the σ-adduct by 2.5–3 pKa units.29 Due to this effect,
it is therefore possible that the σ-complexation of NBD-Cl and
NBD-Imϩ by anilines 1a–f is accompanied by the generation
of a greater positive charge at the nitrogen atom of the adducts
2a–f than does the protonation of 1a–f. Should this be true, one
can anticipate that the degree of positive charge develop-
ment on the nitrogen atom at the σ-adduct-like TS will also
be much greater in reactions (1) than in SNAr reactions of
amines with less electrophilic nitroaromatic substrates.14 This
would explain the observed high βnuc values, notably in the
NBD-Imϩ reactions where the development of the positive
charge will be accentuated by the additional ϪI effect exerted
by the N-methylimidazolium moiety.
Further information as to the intimate mechanisms of reac-
tions (1) becomes available through analysis of the results in
terms of Brønsted-type plots, Figs. 3 and 4, because in the pres-
ent systems we have determined the pKa values of the various
anilines 1a–f in the two H2O–Me2SO solvent mixtures studied
(Table 1). In this respect, Figs. 3 and 4 show that the reactivity
of our systems is described by nicely linear correlations which
lead, however, to unexpectedly high βnuc values: for NBD-Cl,
βnuc = 0.96 and 0.93 in 30% and 80% Me2SO, respectively, while
for NBD-Imϩ, βnuc = 1.43 and 1.30 in 30% and 80% Me2SO,
respectively. Reports of βnuc values close to or greater than unity
have so far been very sparse for nucleophilic aromatic substi-
tutions.14,15,17 Our findings pertaining to the NBD-Cl and NBD-
Imϩ–aniline systems are therefore worthy of comment.
Although it accounts reasonably well for the βnuc values
associated with reactions (1), it is more difficult to reconcile the
above reasoning with other experimental findings. As further
confirmed by the similarity of the rate constants for reaction
of picryl chloride and NBD-Cl with MeOϪ in methanol—the
second-order rate constants k are equal to 17.7 and 7.7 dm3
Why high ꢀnuc values?
As recalled in the introduction, βnuc values close to or greater
than 1 have been commonly interpreted either in terms of a
similar or greater sensitivity to substituent changes on the reac-
tion at hand relative to the reference ionisation equilibrium,2a,11
or in the case of SN2 reactions in terms of the advent of a SET
pathway where full electronic transfer occurs prior to the coup-
ling of the electrophilic and nucleophilic partners.12–14 Focusing
first on the argument dealing with the sensitivity to substituent
molϪ1
s
Ϫ1, respectively33,34—the electrophilic character of
4-nitrobenzofurazan and 2,4,6-trinitrobenzene moieties is
comparable. On this ground, substitutions involving these two
types of structures should exhibit a similar sensitivity to substi-
tuent changes and therefore similar βnuc values. In addition,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1760

View Article Online
Scheme 2
Dixon and Bruice have previously reported that the βnuc value
While schemes of this type have previously been proposed for
associated with the reaction of picryl chloride with various
amines is equal to 0.64 in aqueous solution,35 i.e. it falls in the
normal range recognized for substitutions adhering to a normal
polar SNAr mechanism.14,17
In fact, it is only in two nucleophilic aromatic substitutions
proceeding through a concerted mechanism that meaningful
βnuc values out of the range 0.5–0.7 have been previously
reported and explained in terms of an increased sensitivity of
the transition state to substituent changes.16 These reactions
deal, respectively, with the displacement of 4-nitrophenoxide ion
substitution of nitro-substituted benzene derivatives with vari-
ous nucleophiles,17,37–42 including much weaker electron donors
than anilines, the studies did not include determination of βnuc
values, relying chiefly on spectroscopic identification of radical
anions with frequent attendant formation of side products.
Thus, these studies did not permit the obtention of direct
information concerning the progress of the reactions and the
structure of the rate-determining transition-state.
We would like to emphasize two notable features that are
consistent with the formulation of Scheme 2. A first salient
feature which contrasts the present NBD-Cl–aniline systems
with the previously investigated ones is the more facile 1eϪ-
reduction of the nitrobenzofurazan structures relative to simi-
larly electrophilic benzenoid systems.43–47 || EЊ values pertaining
to the reduction of a number of nitrobenzofurazans and related
1-oxides (nitrobenzofuroxans) have been recently measured by
us and found to fall in the range ϩ0.10 to Ϫ0.44 V (relative to
SCE).43 For example, we have EЊ = Ϫ0.44 V for the reduction of
4-nitrobenzofurazan in Me2SO solution.43 Such EЊ values are
less negative than the corresponding EЊ values of Ϫ0.58 to
Ϫ1 V (relative to SCE) which have been reported for radical
anion formation from a large set of dinitro- and trinitrobenzene
derivatives, e.g. EЊ = Ϫ0.68 V for the reduction of tri-
nitrotoluene.45 While attempts to characterize the postulated
radical anion of NBD-Cl through EPR have failed due to the
low lifetime of this type of species in the benzofurazan and
benzofuroxanseries,43 **itcanreasonablybeconcludedfromthe
available EЊ values that the NBD-Cl–aniline system is charac-
terized by a greater affinity to take part in an electron transfer
process compared to picryl chloride–aniline systems. It is this
difference in the affinity for electron capture which can be one
of the driving forces for the proposed mechanistic change and
favor the SET mechanism in our NBD-Cl system. Of interest
also is the recent proposal that the formation of radical anions
from nitroarenes and related substrates would be greatly facili-
tated in systems where the electron transfer between the
electrophilic and nucleophilic partners could take place within
an initially formed π-complex.37,39–42,47,48 Then, the high pro-
pensity of nitrobenzofuroxans to form such π-complexes with a
from 2-(4-nitrophenoxy)-4,6-dimethoxy-1,3,5-triazine (βnuc
=
0.95) and of 3,4-dinitrophenoxide ion from 4-(3Ј,4Ј-di-
nitrophenoxy)-2,6-dimethoxy-1,3,5-triazine by a series of sub-
stituted pyridines (βnuc = 1.08). Interestingly, most other 1,3,5-
triazinyl transfers studied have been found to obey a normal
SNAr mechanism with βnuc values in the range of 0.5–0.7,
including those occurring between two pyridine nucleophiles
with the intermediacy of σ-adducts having two positively
charged moieties bonded at the sp3 carbon.36 According to
Williams and coworkers,15 a concerted nucleophilic aromatic
substitution can only take place in systems where the putative
intermediate σ-adduct is of particularly high energy, a situation
which can be rejected in reactions involving such powerful
stabilizing σ-adduct structures as a 4-nitrobenzofurazanyl or a
picryl moiety. As a result, the explanation of βnuc values close to
or greater than 1 in reactions (1) on the basis of a very high
sensitivity of the transition states to the nature of the X-substi-
tuent of the incoming anilines, is not so convincing if these
reactions really proceed via a normal polar SNAr mechanism.
Can single electron transfer (SET) be a reasonable pathway?
The above inconsistencies lead us to envision that the high βnuc
values associated with our reactions may be the reflection of a
SET pathway, as described in Scheme 2 for the NBD-Cl system
as an example. This scheme depicts the formation of the
σ-complex intermediates 2a–f through initial (fast) electron
transfer from the aniline donor to the nitrobenzofurazan
acceptor moiety and subsequent (slow) coupling of the
resulting cation and anion radicals within the solvent cage
via the transition state 8. The latter will have a full positive
charge on the anilino nitrogen atom thus accounting for the
observed corresponding equivalence of rate and equilibrium
protonation, i.e. βnuc ∼ 1.2b,14 In accord with the lack of
observation of base catalysis in the present system, conversion
of the σ-adducts to the final substitution products will then
occur rapidly, as discussed earlier in this paper.
|| EЊ values of the same order of magnitude have been measured for
1,3,5-trinitrobenzene and 2,4,6-trinitroanisole. We thank Dr Illuminada
Gallardo (Universitat Autonoma de Barcelona) for this personal
communication.
** Cyclic voltammetry experiments have shown that the reduction of
nitrobenzofuroxans is an irreversible process in CH3CN or Me2SO, thus
preventing a determination of the related EЊ values. These are therefore
approximated from the observed reduction peaks.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1761

View Article Online
number of electron-rich arenes, is an additional argument
which can favor Scheme 2 for the reactions of NBD-Cl with the
electron-rich anilines 1a–f.49
Even though the above features are clearly consistent with
Scheme 2, we will not claim that they provide sufficient evidence
for the implication of SET in our reactions. Based on the
reasonable assumption that in this mechanistic route the aniline
moiety in the transition state should closely resemble the radical
cation, one would expect a better description of the reactivity
of our systems in terms of σpϩ rather than σp Hammett substi-
tuent constants, which is not really borne out by experiments. It
remains that the finding of satisfactory log k–EЊ correlations is
unprecedented for SNAr substitutions and hence justifies to a
possible transition from the polar to the SET mechanism in this
field. Other investigations are currently being carried out in
our laboratory to shed further light on this important facet of
nucleophilic aromatic substitutions.
A second and perhaps most important feature in favor of
Scheme 2 is that the rate constants for substitution of NBD-Cl
by the anilines 1a–f are found to correlate satisfactorily with
the oxidation potentials EЊ of these species, as measured by
Jonsson et al. in aqueous solution.50 This is illustrated by the
linear log k vs. EЊ plots of Fig. 5 (R = 0.970) and 6 (R = 0.983)
which describe the variations in reactivity in 30% Me2SO and
80% Me2SO, respectively. In this regard, it is worth noting that
the quality of these correlations is largely related to the surpris-
ing behaviour of 4-chloroaniline which has the same EЊ value
as aniline while being less basic and therefore less reactive than
the unsubstituted amine. As a matter of fact, if the points for
4-chloroaniline are omitted the correlation coefficients of
Figs. 5 and 6 become equal to 0.993 and 0.991, respectively.
Conclusions
In conclusion, the present study of the reactions of substi-
tuted anilines with NBD-chloride and -imidazolium sub-
strates, which are characterized by βnuc values in the region of
1.0 and 1.3–1.4, respectively, highlights the possible role of the
Brønsted βnuc parameter as an indicator of the SET process.
A salient feature of the present system which can induce the
transition from the normal SNAr pathway to the SET pathway,
is the much more facile 1eϪ reduction of the nitrobenzofurazan
structure relative to similarly electrophilic benzenoid systems
such as picryl chloride. A significant result which is consistent
with the SET mechanism is the finding that the rate constants
for the substitutions of NBD-Cl and NBD-Imϩ correlate well
with the oxidation potentials of the aniline nucleophiles. In
addition, it has been shown that the marked sensitivity of rates
to substituent changes in the present system is related to signifi-
cant repulsive interactions in the SNAr transition-states, which
would be conducive to a transition to the SET mechanism as an
energetically more favorable pathway.51
Fig. 5 The influence of the oxidation potential EЊ of anilines on the
rates of reaction of NBD-Cl with 1a–f in 70–30 (v/v) H2O–Me2SO;
I = 0.2 mol dmϪ3 (Me4NCl); T = 25 ЊC. Numbering from Table 1.
Experimental
Materials
Solvents were purified and solutions made up as previously
described.21,52,53
4-Chloro-7-nitrobenzofurazan
(NBD-Cl;
Fluka BioChemika) was used as received. The preparation
of the various substitution products 4a–f was previously
reported.21
3-Methyl-1-(4-nitrobenzofurazanyl)imidazolium
chloride (NBD-ImϩClϪ) was prepared as follows: 2 g (10 mmol)
of NBD-Cl and 0.8 mL (10 mmol) of N-methylimidazole were
dissolved in 50 mL butyl acetate and the mixture was heated to
40 ЊC. After ∼10 minutes, a yellow precipitate began to form.
The crystals deposited after 30 minutes were filtered, washed
thoroughly with pentane and dried under vacuum. NBD-
ImϩClϪ was thus obtained in 70% yield; mp 190 ЊC (with
decomposition).
Fig. 6 The influence of the oxidation potential EЊ of anilines on the
rates of reaction of NBD-Cl with 1a–f in 20–80 (v/v) H2O–Me2SO;
I = 0.2 mol dmϪ3 (Me4NCl); T = 25 ЊC. Numbering from Table 1.
1
NMR data for NBD-ImϩClϪ: H NMR (δ in ppm, Me2SO-
d6, Me4Si): 10.32 (s, 1H, H2Ј), 8.99 (d, 1H, 3JH5H6 = 8.19 Hz, H5),
3
3
8.66 (d, 1H, JH5ЈH4Ј = 1.65 Hz, H5Ј), 8.39 (d, 1H, JH5H6
=
Returning to the imidazolium substrate, NBD-Imϩ, satis-
factorily linear log k–EЊ correlations can also be drawn from
the data of Table 1 (Figs. S1 and S2), making it reasonable to
postulate that Scheme 2 can also operate for substitutions by
the anilines 1a–f in this derivative. However, the coupling step
would then involve the approach of a positively charged aniline
radical towards a ring position bearing a positively charged
leaving group moiety. Again, such an approach would involve a
significant repulsive interaction which would be modulated by
the nature of the X substituent in the aniline moiety, as dis-
cussed above for a normal polar SNAr mechanism. The greater
electronic effect which this calls for will result in a steeper
dependence on the electron-donating/electron-withdrawing
capability of X, i.e. a larger slope in the Brønsted plot or EЊ
correlations, as observed.
3
1.65 Hz, H6), 8.17 (d, 1H, JH5ЈH4Ј = 1.65 Hz, H4Ј), 4.08
(s, 1H, CH3). 13C NMR (δ in ppm, Me2SO-d6, Me4Si): 145.55
(C8, JC8H6 = 9.0 Hz), 143.78 (C9, JC9H5 = 8.5 Hz), 138.38 (C2Ј,
JC2ЈH2Ј = 227.2 Hz, JC2ЈH5Ј = JC2ЈH4Ј = 4.5 Hz), 136.05 (C4, JC4H6
=
8.5 Hz, JC4H5 = 5.1 Hz), 132.66 (C5, JC5H5 = 174.0 Hz), 128.58
(C7, JC7H5 = 9.6 Hz, JC7H6 = 1.7 Hz), 125.23 (C4Ј, JC4ЈH4Ј
206.6 Hz, JC4ЈH2Ј = 11.2 Hz, JC4ЈH5Ј = 3.4 Hz), 123.62 (C6, JC6H6
=
=
174.0 Hz), 121.83 (C5Ј, JC5ЈH5Ј = 208.0 Hz, JC5ЈH2Ј = 13.0 Hz,
JC5ЈH4Ј = 5.1 Hz), 36.64 (CH3, JCH3 = 144.7 Hz). HRMS calcd for
C10H8N5O3: 246.0627 [M Ϫ ClϪ]ϩ, found m/z 246.0631.
Measurements
Kinetic measurements were carried out with either a con-
ventional UV-visible Cary 1E spectrophotometer or an Applied
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1762

View Article Online
19 F. Terrier, M.-J. Pouet, J.-C. Hallé, E. Kizilian and E. Buncel,
Photophysics SX. 18 MV stopped-flow spectrometer being both
equipped with a thermostated cell compartment (25 ЊC 0.2).
All rates were reproducible to within 3%. The acidity con-
stants of the p-X-substituted anilines 1a–f in the 70–30 and
20–80 (v/v) H2O–Me2SO mixtures containing 0.2 mol dmϪ3
Me4NCl have been measured by potentiometry at 25 ЊC, using
an electronic pH meter (Tacussel ISIS 20000) equipped with an
hydrogen electrode and a Tacussel C8 reference calomel elec-
trode. The cell and the procedures, previously described in
detail,52 allow the determination of apparent pKa referenced
to an infinitely diluted buffer solution in the H2O–Me2SO–
Me4NCl (0.2 mol dmϪ3) ternary mixtures.
J. Phys. Org. Chem., 1998, 11, 707 and references cited therein.
20 R. Goumont, M. Sebban, P. Sepulcri, J. Marrot and F. Terrier,
Tetrahedron, 2002, 58, 3249.
21 J.-C. Hallé, M. Mokhtari, P. Soulié and M.-J. Pouet, Can. J. Chem.,
1997, 75, 1240.
22 G. Illuminati and F. Stegel, Adv. Heterocycl. Chem., 1983, 34, 305.
23 J. F. Bunnett, Q. Rev. Chem. Soc., 1958, 12, 1.
24 J. Miller, in Aromatic Nucleophilic Substitution, Elsevier;
Amsterdam, 1968.
25 G. Consiglio, V. Frenna and D. Spinelli, Topics in Heterocyclic
Systems – Synthesis, Reactions and Properties. Base Catalysis
in Aromatic Nucleophilic Substitutions: Current Views. Research
Signpost, Trivandrum (India), 1996, Vol. 1, p. 169.
26 O. Exner, in Correlation Analysis in Chemistry: Recent Advances. ed.
N. B. Chapman and J. Shorter, Plenum Press, New York, 1978.
Chapter 10: “A Critical Compilation of Substituent Constants.”
27 C. F. Bernasconi, MTP Int. Rev. Sci.: Org. Chem., Ser. 1, 1973, 3, 33.
28 J.-C. Hallé, C. Pichon and F. Terrier, J. Biol. Chem., 1984, 259, 4142.
29 M. R. Crampton, J. Delaney and L. C. Rabbitt, J. Chem. Soc.,
Perkin Trans. 2, 1999, 2473.
Acknowledgements
Authors are very indebted to CNRS for financial support
(CNRS/DEF nЊ 12109). We also thank Professor J. Pinson
(University Denis Diderot) for helpful comments.
30 F. G. Bordwell, Acc. Chem. Res., 1988, 22, 456.
31 J. J. Ryan and A. A. Humffray, J. Chem. Soc. B, 1967, 1300.
32 J. Hirst and K. U. Rahman, J. Chem. Soc., Perkin Trans. 2, 1973,
2119.
33 J. Kavalek, M. Pastrnek and V. Sterba, Collect. Czech. Chem.
Commun., 1978, 43, 1401.
34 L. Di Nunno, S. Florio and P. E. Todesco, J. Chem. Soc., Perkin
Trans. 2, 1975, 1469.
References
1 (a) W. P. Jencks, Bull. Soc. Chim. Fr., 1988, 218; (b) W. P. Jencks, in
Nucleophilicity, ed. J. M. Harris and S. P. Mc Manus, Advances in
Chemistry Series 215, American Chemical Society, Washington,
DC, 1987, p. 155.
2 (a) A. Williams, Chem. Soc, Rev., 1994, 93; (b) M. Page and
A. Williams, in Organic and Bio-organic Mechanisms, Longman,
Harlow, Chapter 3; (c) M. Eigen, Angew. Chem., Int. Ed. Engl., 1964,
3, 1; (d ) A. J. Kresge, Chem. Soc. Rev., 1973, 2, 475.
3 (a) F. G. Bordwell and W. J. Boyle, J. Am. Chem. Soc., 1972, 94, 3907;
(b) F. G. Bordwell and J. A. Hautala, J. Org. Chem., 1978, 43, 3116
and references therein.
4 W. P. Jencks, M. T. Haber, D. Herschlag and K. L. Nazaretian,
J. Am. Chem. Soc., 1986, 108, 479.
5 C. F. Bernasconi, Adv. Phys. Org. Chem., 1992, 27, 119 and
references therein.
6 (a) W. P. Jencks, S. R. Brant, J. R. Gandler, G. Friedrich and
C. Nakamura, J. Am. Chem. Soc., 1982, 104, 7045; (b) D. J. Hupe
and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 451.
7 D. J. Hupe and D. Wu, J. Am. Chem. Soc., 1977, 99, 7653.
8 C. F. Bernasconi, Acc. Chem. Res, 1987, 20, 301.
35 J. E. Dixon and T. C. Bruice, J. Am. Chem. Soc., 1972, 94, 2052.
36 (a) N. R. Cullum, A. H. M. Renfrew, D. Rettura, J. A. Taylor,
M. J. Whitemore and A. Williams, J. Am. Chem. Soc., 1995, 117,
9200; (b) A. H. M. Renfrew, J. A. Taylor, M. J. Whitemore and
A. Williams, J. Chem. Soc., Perkin Trans. 2, 1994, 2383.
37 L. Grossi and S. Strazzari, J. Chem. Soc., Perkin Trans. 2, 1999,
2141.
38 C. Paradisi and G. Scorrano, Acc. Chem. Res., 1999, 32, 958.
39 T. Abe and Y. Ikegami, Bull. Chem. Soc. Jpn, 1977, 99, 6677.
40 L. A. Blyumenfeld, L. V. Bryukhovetskayar, G. V. Fomin and
S. M. Shein, Russ. J. Phys. Chem., 1970, 44, 518 and references
therein.
41 (a) R. Bacaloglu, C. A. Bunton and G. Cerichelli, J. Am. Chem. Soc.,
1987, 109, 621; (b) R. Bacaloglu, C. A. Bunton and F. Ortega, J. Am.
Chem. Soc., 1988, 110, 3503; (c) R. Bacaloglu, C. A. Bunton and
F. Ortega, J. Am. Chem. Soc., 1988, 110, 3512.
9 (a) F. Terrier, G. Moutiers, L. Xiao, E. Le Guével and F. Guir, J. Org,
Chem., 1995, 60, 1748; (b) E. Buncel, C. Cannes, A.-P. Chatrousse
and F. Terrier, J. Am. Chem. Soc., 2002, 124, 8766 and references
therein.
10 F. Terrier, P. Mc Cormack, E. Kizilian, J.-C. Hallé, P. Demerseman,
F. Guir and C. Lion, J. Chem. Soc., Perkin Trans. 2, 1991, 153.
11 (a) M. I. Page and W. P. Jencks, J. Am. Chem. Soc., 1972, 94,
3263; (b) M. I. Page and W. P. Jencks, J. Am. Chem. Soc., 1972, 94,
8818.
42 (a) R. Bacaloglu, C. A. Bunton and F. Ortega, J. Am. Chem. Soc.,
1989, 111, 1041; (b) R. Bacaloglu, A. Blasko, C. A. Bunton,
E. Dorwin, F. Ortega and C. Zucco, J. Am. Chem. Soc., 1991, 113,
238.
43 G. Moutiers, J. Pinson, F. Terrier and R. Goumont, Chem. Eur. J.,
2001, 7, 1712.
44 (a) I. M. Sosonkin and G. L. Kalb, Zh. Org. Khim., 1974, 10, 1341;
(b) I. M. Sosonkin, G. N. Strogov, A. Ya. Kaminskii, G. E. Troshin
and F. F. Lakomov, Zh. Org. Khim., 1980, 16, 1711.
45 I. Gallardo, G. Guirado and J. Marquet, Chem. Eur. J., 2001, 7,
1759.
12 F. G. Bordwell, J. C. Branca and T. A. Cripe, Isr. J. Chem., 1985, 26,
357.
13 F. G. Bordwell and A. H. Clemens, J. Org. Chem., 1981, 46, 1037.
14 F. G. Bordwell and D. L. Hugues, J. Am. Chem. Soc., 1986, 108,
5991.
15 (a) A. Hunter, M. Renfrew, D. Rettura, J. A. Taylor, J. M. Whitmore
and A. Williams, J. Am. Chem. Soc., 1995, 117, 5484; (b) J. Shakes,
C. Raymond, D. Rettura and A. Williams, J. Chem. Soc., Perkin
Trans. 2, 1996, 1553.
16 G. D. Leahy, M. Liverio, J. Miller and A. J. Parker, Aust. J. Chem.,
1956, 9, 382.
17 F. Terrier, in Nucleophilic Aromatic Displacement, ed. H. Feuer,
VCH, New-York, 1991, Chapters 1 and 2.
18 (a) F. Terrier, E. Kizilian, J.-C. Hallé and E. Buncel, J. Am.
Chem. Soc., 1992, 114, 1740; (b) F. Terrier, Chem. Rev., 1982, 82, 77;
(c) E. Buncel, R. A. Renfrow and M. J. Strauss, J. Org. Chem., 1987,
52, 488; (d ) E. Buncel, R. A. Renfrow and M. J. Strauss, Can. J.
Chem., 1983, 61, 1690; (e) R. A. Renfrow, M. J. Strauss, S. Cohen
and E. Buncel, Aust. J. Chem., 1983, 36, 1843; ( f ) E. Buncel,
J. M. Dust, K. T. Park, R. A. Renfrow and M. J. Strauss, in
Nucleophilicity, ed. J. M. Harris and S. P. Mc Manus, Advances in
Chemistry Series 215, American Chemical Society, Washington,
DC, 1987, p. 369.
46 (a) I. Gallardo, G. Guirado and J. Marquet, J. Org. Chem., 2002, 67,
2548; (b) I. Gallardo, G. Guirado and J. Marquet, Eur. J. Org.
Chem., 2002, 251; (c) I. Gallardo, G. Guirado and J. Marquet, Eur. J.
Org. Chem., 2002, 261; (d ) I. Gallardo, G. Guirado and J. Marquet,
J. Electroanal. Chem., 2000, 488, 64.
47 L. Grossi, Tetrahedron Lett., 1992, 33, 5645.
48 V. Arca, C. Paradisi and G. Scorrano, J. Org. Chem., 1990, 55, 3617.
49 (a) A. S. Bailey and J. R. Case, Tetrahedron, 1958, 3, 113; (b) C. Boga
and L. Forlani, J. Chem. Soc., Perkin Trans. 2, 2001, 1408;
(c) L. Forlani, J. Phys. Org. Chem., 1999, 12, 417.
50 (a) M. Jonsson, J. Lind, T. E. Ericksen and G. Merényi, J. Am.
Chem. Soc., 1994, 116, 1423; (b) J. Bacon and R. N. Adams, J. Am.
Chem. Soc., 1968, 90, 6596.
51 (a) A. Pross, Acc. Chem. Res., 1985, 18, 212; (b) A. Pross and
S. S. Shaik, Acc. Chem. Res., 1983, 16, 363; (c) L. Eberson and
F. Radner, Acc. Chem. Res., 1987, 20, 53.
52 (a) H. A. Achassi Sorkhabi, J.-C. Hallé and F. Terrier, J. Chem. Res.,
1978, 196 (S), 2545 (M); (b) H. A. Achassi Sorkhabi, J.-C. Hallé and
F. Terrier, J. Chem. Res., 1978, 1371 (M).
53 J.-C. Hallé, R. Gaboriaud and R. Schaal, Bull. Soc. Chim. Fr., 1969,
1851.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 7 5 7 – 1 7 6 3
1763