The reactions of 4-nitrobenzofuroxan with amines in DMSO; kinetic and equilibrium studies

Article abstract of DOI:10.1039/a905774i

Kinetic and equilibrium studies are reported for the reactions of six aliphatic amines with 4-nitrobenzofuroxan, 3, in dimethyl sulfoxide. Rapid reaction at the 5-position to yield σ-adducts is followed by slower formation of the thermodynamically more stable adducts at the 7-position. ¹H NMR spectra are reported. The results are compared with those for reaction of the amines with 1,3,5-trinitrobenzene, 2. These reactions involve the initial formation of zwitterionic intermediates which transfer an amino proton to a second molecule of amine. It is found that the proton-transfer step is less likely to be rate determining in the reactions of 3 than of 2; this may be due to the lower steric requirements of 3.

Full text of DOI:10.1039/a905774i

View Article Online / Journal Homepage / Table of Contents for this issue
The reactions of 4-nitrobenzofuroxan with amines in DMSO;
kinetic and equilibrium studies
Michael R. Crampton,* Jeannette Delaney and Lynsey C. Rabbitt
Chemistry Department, Durham University, Durham, UK DH1 3LE
Received (in Cambridge, UK) 19th July 1999, Accepted 20th September 1999
Kinetic and equilibrium studies are reported for the reactions of six aliphatic amines with 4-nitrobenzofuroxan, 3,
in dimethyl sulfoxide. Rapid reaction at the 5-position to yield σ-adducts is followed by slower formation of the
thermodynamically more stable adducts at the 7-position. 1H NMR spectra are reported. The results are compared
with those for reaction of the amines with 1,3,5-trinitrobenzene, 2. These reactions involve the initial formation of
zwitterionic intermediates which transfer an amino proton to a second molecule of amine. It is found that the proton-
transfer step is less likely to be rate determining in the reactions of 3 than of 2; this may be due to the lower steric
requirements of 3.
Kinetic studies of the reactions of aromatic substrates with
amine nucleophiles have provided important evidence for the
SNAr mechanism of substitution1–3 (Scheme 1). It has been
lower than the diffusion-controlled limit.10–14 There is evidence
that steric congestion at the reaction centre may decrease the
value of kAm, so that lower values are obtained with secondary
than with primary amines, and also when the size of the X
group is increased.15
In the present study we report kinetic and equilibrium results
for reaction of six aliphatic amines with 4-nitrobenzofuroxan,
3, in DMSO. One of the aims of this work was to examine
Scheme 1
shown that base catalysis in such reactions may result from rate-
limiting proton transfer4–6 from the zwitterionic intermediate,
1, or, alternatively, may involve rapid interconversion of 1 with
its deprotonated form followed by general acid catalysis of leav-
ing group departure1,7–9 (the SB-GA mechanism). In dimethyl
sulfoxide (DMSO) as solvent the latter possibility becomes
more likely with decreasing nucleofugality of the leaving group.
Useful information about the values of rate constants for
proton transfer between nitrogen atoms in these systems has
come from studies involving substrates such as 1,3,5-
trinitrobenzene, 2, where stable σ-adducts are formed. These
studies have shown that in DMSO the proton transfer step in
Scheme 2 may be rate-limiting, with values of kAm considerably
σ-adduct forming reactions with amine nucleophiles in an acti-
vated system where steric effects are likely to be less pronounced
than in a trinitro-activated benzene ring.
3 is also of interest since it has been shown that its action as
an in vitro inhibitor of nucleic acid and protein biosynthesis in
animal cells probably involves σ-adduct formation.16–18 There
have been previous studies of its reactions with oxygen nucleo-
philes. NMR studies have shown that reaction with methoxide
ions results in rapid attack at the 5-position followed by slower
isomerisation to the thermodynamically more stable 7-methoxy
σ-adduct.19,20 Kinetic and equilibrium data have been reported
for these reactions in methanol,21 and also for similar reactions
of 4-nitrobenzofurazan derivatives.22 NMR studies23 of the
reaction of 3 with aryloxides have shown that reaction occurs
via the oxygen centre of these potentially ambident nucleo-
philes; in contrast with methoxide addition the 7-aryloxy
adducts are preferred both kinetically and thermodynamically
to the 5-adducts. There have been several reports20,24 showing
that, following initial σ-adduct formation, reaction of 3 with
nucleophiles may result in slow irreversible reactions yielding
4-nitrobenzofurazan derivatives.
Results and discussion
Our results for the reactions of 3 with six aliphatic amines in
DMSO provide evidence for the σ-adduct forming reactions
1
shown in Scheme 3. The H NMR spectra, recorded immedi-
Scheme 2 1,3,5-Trinitrobenzene, 2, X = H.
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480
This journal is © The Royal Society of Chemistry 1999
2473

View Article Online
Table 1 1H NMR dataa for 3 and adducts, 7, in [2H6]-DMSO
1H NMR Shifts
H5
H6
H7
J56
7.3
10.2
10.4
10.4
10.4
10.2
10.4
J67
Otherb
3
8.61
6.94
6.99
7.04
6.93
6.99
7.09
7.53
4.92
5.03
5.00
4.94
5.00
5.14
8.13
4.47
4.73
4.61
4.62
4.71
4.73
8.9
4.4
4.6
4.4
4.4
4.2
4.4
7a
7b
7c
7d
7e
7f
2.5(2H), 2.2(2H), 1.3(6H)
2.6(2H), 2.4(2H), 1.6(4H)
3.48(4H), 2.60(2H), 2.35(2H)
2.35(1H), 2.15(1H), 1.40(2H), 0.82(3H)
4.95(NH), 3.61(1H), 3.41(1H)
3.59(2H), 2.05(NMe)
a J values in Hz. b Non-equivalence is observed for hydrogen atoms on the carbon atom(s) α to nitrogen in the adducts.
formation. These are reasonably attributed to the rapid form-
ation of adducts 5 followed by slower formation of the thermo-
dynamically more stable adducts, 7. The UV absorption
spectrum of 3 shows a band at 414 nm, ε 8 × 103 dm3 molϪ1
cmϪ1; on adduct formation the maximum shifts to 350 nm, ε
1.5 × 104 dm3 molϪ1 cmϪ1 with a shoulder at 400 nm. Kinetic
measurements were made at 350 nm under first order condi-
tions. Thus for reactions in buffers (amine plus amine salt) the
buffer components were in large excess of the substrate concen-
tration (2–4 × 10Ϫ5 mol dmϪ3). For reactions with amines in the
absence of added amine salts a sufficient excess of amine was
used to ensure >95% conversion to adduct at equilibrium.
Under these conditions eqn. (1) applies and was used to calcu-
late rate constants.
A∞
ln ͫ ͬ= kobs
t
(1)
A∞ Ϫ A
It is assumed that the zwitterionic forms 4 and 6 may be
treated as steady state intermediates, so that the general rate
expression for reaction at the 5-position to give 5 is eqn. (2).
Scheme 3
ately after mixing, of 3 in the presence of a four-fold excess of
amine in [2H6]-DMSO indicate that the thermodynamically
stable adducts result from reaction at the 7-position to give 7.
The data summarised in Table 1 show that in the adducts, 7,
spin-coupled bands are observed for H7 at ca. δ 4.6, for H6 at
ca. 5.0 and for H5 at ca. 7.0. Interestingly the spectra show that
in the adducts rotation is slow about the nitrogen to α-carbon
bond in the added amine. For example in 7e, the adduct from
benzylamine, the methylene hydrogens give an AB quartet with
δ 3.41 and 3.61 and J 13.4 Hz. In a related manner in the piper-
idine adduct, 7a, the two methylene groups adjacent to nitrogen
give distinct multiplets at δ 2.2 and δ 2.5. The NMR spectra
show that the initial formation of σ-adducts is followed by
slow irreversible decomposition reactions which were not
investigated.
ϩ
k5kAm[Am]2 ϩ kϪ5kAmH [AmH ]
ϩ
kfast
=
(2)
k
Ϫ5 ϩ kAm[Am]
The overall equilibrium constant for formation of the 5-adduct
is defined in eqn. (3). Values of kfast were, in most cases, high
[5][AmHϩ]
[3][Am]2
k5ؒkAm
Kc,5
=
=
(3)
ϩ
kϪ5ؒkAmH
and towards the limit of detection by the stopped-flow method.
In addition, particularly in the presence of amine salts, ampli-
tudes were low. Hence only limited measurements of kfast were
possible.
The general rate equation for the equilibration of 3 and 7,
in the presence of 5, is eqn. (4). If, in the formation of 7, the
proton transfer equilibrium between 6 and 7 is rapid (corre-
sponding to the condition kAm[Am] ӷ kϪ7) then eqn. (4) reduces
to eqn. (5). The overall equilibrium constant for formation of 7
Comparison with data for the isomeric methoxide adducts19
shows that the spectra obtained are not compatible with form-
ation of 5, where H6 is expected to be considerably less shielded
than is observed.
Kinetic measurements by stopped flow spectrophotometry
generally showed two rate processes associated with σ-adduct
2474
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480

View Article Online
Table 2 Kinetic data for reaction of 3 with piperidine in DMSO at 25 ЊC
[Piperidine]/
[Piperidine HϩClϪ]/
Item
mol dmϪ3
mol dmϪ3
kfast a/sϪ1
kcalc b/sϪ1
kslow/sϪ1
kcalc c/sϪ1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0.002
0.004
0.006
0.008
0.010
0.001
0.002
0.004
0.007
0.010
0.001
0.001
0.001
0.002
0.001
0.001
0.001
0.002
0.002
0.003
0.004
0.006
0.008
0.010
0
0
0
0
62
200
380
620
800
—
—
—
—
—
—
—
—
—
—
—
—
—
57
190
380
600
850
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
0
—
—
0.10
0.10
0.10
0.10
0.10
0.06d
0.04d
0.02d
0.02d
0.06
0.04
0.02
0.02
0.005
0.005
0.005
0.005
0.005
0.005
9.8
9.1
12.1
19.4
26.6
5.7
4.3
3.1
5.5
8.0
6.7
5.2
6.6
6.0
8.6
11.1
14.4
17.1
18.9
9.6
8.7
12.1
19.2
26.7
5.8
5.2
4.0
5.9
—
—
—
—
6.1
8.6
10.9
14.5
17.1
18.9
—
—
—
—
—
—
—
—
—
—
—
—
a Where no value is given, rate constants were too high for measurement. b Calculated from eqn. (8) with K5kAm 1.7 × 107 dm6 molϪ2 sϪ1 and kAm/kϪ5
100 dm3 molϪ1 c Calculated from eqn. (5) or (9), with values given in the text. d Salt concentration maintained at 0.10 mol dmϪ3 using tetra-n-
butylammonium chloride.
.
k7kAm[Am]2
Reactions of piperidine
kslow
=
ϩ
Kc,5[Am]2
[AmHϩ]
Data are in Table 2. In the absence of added salt it was possible
to measure rate constants for the rapid process corresponding
to the formation of 5a. Since no amine salt is present the reverse
reaction makes a negligible contribution to the rate expression.
Hence eqn. (2) commutes to eqn. (8) where K5 = k5/kϪ5. The
(kϪ7 ϩ kAm[Am])
ͩ
1 ϩ
ͪ
ϩ
ϩ
kϪ7kAmH [AmH ]
(4)
(5)
k
Ϫ7 ϩ kAm[Am]
ϩ
ϩ
k7[Am]
kϪ7kAmH [AmH ]
K5kAm[Am]2
kslow
=
ϩ
Kc,5[Am]2
[AmHϩ]
kAm
[Am]
(8)
kfast
=
kAm[Am]
ͩ
1 ϩ
ͪ
1 ϩ
kϪ5
is given in eqn. (6). It should be noted that, in general, the
results, items 1–5, show that kfast shows a dependence between
first and second order on the amine concentration. Values cal-
culated with K5kAm 1.7 × 107 dm6 molϪ2 sϪ1 and kAm/kϪ5 100
dm3 molϪ1 give a good fit with the experimental data. These
values lead to a value for k5 of 1.7 × 105 dm3 molϪ1 sϪ1. Since a
value of 27 dm3 molϪ1 is calculated later for the value of Kc,5 at
low salt concentration it is possible to calculate using eqn. (3) a
[7][AmHϩ]
[3][Am]2
k7ؒkAm
Kc,7
=
=
(6)
ϩ
kϪ7ؒkAmH
values of kAm in formation of 5 and 7 will have different values;
ϩ
similarly values of kAmH
.
In order to obtain the maximum kinetic information it was
necessary to make rate measurements at different concen-
trations of amine salt. Measurements were made at low, ≤0.01
mol dmϪ3, salt concentration and also at higher, 0.10 mol dmϪ3,
concentrations using perchlorate and/or chloride salts. It was
found that values of the equilibrium constant Kc,5 and Kc,7 were
slightly higher at the higher salt concentration. This is attribut-
able to a general salt effect expected for the formation of ionic
products from neutral reactants. In agreement with previous
work,7,12,15 a specific effect of chloride ions was also observed.
Thus it is known that stabilisation of substituted ammonium
5
3
value for kAmH of 6 × 10 dm molϪ1 sϪ1.
ϩ
It was found that values of kslow, measured in the presence of
piperidine hydrochloride, are correlated by eqn. (5). This indi-
cates that the proton transfer equilibrium 6a 7a is rapid
with kAm/kϪ7 > 1000 dm3 molϪ1. The results also show that
when [amine salt] ≥ 0.01 mol dmϪ3 the initial formation of 5a
was sufficiently inhibited that eqn. (5) reduces to eqn. (9).
ϩ
ϩ
kϪ7ؒkAmH [AmH ]
(9)
kslow = k7[Am] ϩ
kAm
[Am]
Ϫ
ions occurs by association with chloride ions. The values of KCl
defined in eqn. (7) are ca. 10 dm3 molϪ1. The expected effect of
Items 6–10 in Table 2 give an excellent fit with this equation
with values of k7 2600 200 dm3 molϪ1 sϪ1 and kϪ7ؒkAmH /kAm
ϩ
0.07 0.01 sϪ1. Combination of these values, using eqn. (6),
gives a value for Kc,7 of (3.7 0.5) × 104 dm3 molϪ1. Items 11–
14 obtained with varying concentrations of piperidine hydro-
chloride but with the ionic strength maintained also gave an
acceptable fit with the above parameters. Items 15–18 were used
Ϫ
KCl
ϩ
RRЈNH2ϩ ϩ ClϪ
RRЈNH2 ؒ ؒ ؒ ClϪ
(7)
a 0.1 mol dmϪ3 chloride concentration7,12,15 is to increase values
of Kc,5 and Kc,7 by a factor of ca. 2.0. The main effect on rate
constants is expected to be a ca. two-fold lowering in values of
ϩ
ϩ
to calculate the variation of kAmH kϪ7/kAm with piperidine
kAmH
.
hydrochloride concentration. Values calculated for this par-
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480
2475

View Article Online
Table 3 Kinetic data for reaction of 3 with pyrrolidine in DMSO at
25 ЊC
Table 4 Kinetic data for the reaction of 3 with morpholine in DMSO
at 25 ЊC
[Pyrrolidine
[Morpholine
hydrochloride]/
mol dmϪ3
[Pyrrolidine]/
HϩClO4Ϫ]/
[Morpholine]/
Item
mol dmϪ3
mol dmϪ3
kslow/sϪ1
kcalc a/sϪ1
Item
mol dmϪ3
kslow/sϪ1
kcalc a/sϪ1
1
2
3
4
5
6
7
8
0.001
0.002
0.003
0.005
0.01
0.02
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
9.4
12.4
17.1
25.6
38
38
34
9.4
12.8
17.1
25.4
38
38
32
1
2
3
4
5
6
7
8
9
10
11
12
13
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
0.006
0.010
0.014
0.018
0.020
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.001
0.001
0.001
0.001
0.001
12.7
11.3
10.2
10.0
9.8
10.0
10.1
10.5
1.47
2.42
3.44
4.40
4.96
13.1
11.1
10.2
9.8
9.7
9.9
10.2
10.5
1.5
2.5
3.5
0.04
29
26
[Pyrrolidine HϩClϪ]/mol dmϪ3
9
10
11
12
13
14
15
16
17
18
19
20
21
0.001
0.002
0.003
0.005
0.01
0.02
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.10
0.10
0.10
0.10
0.10
8.5
12.1
16.9
25
33
32
27
21
16
17
8.5
11.9
16.0
24
33
32
26
21
16
19
4.5
5.0
a Calculated from eqn. (9) with values as follows: items 1–8, k7 360 60
dm3 molϪ1
s
Ϫ1, kAmH ؒkϪ7/kAm 0.66 0.06 sϪ1; items 9–13, k7 250 30
ϩ
dm3 molϪ1 sϪ1, kAmH
k
Ϫ7/kAm 0 sϪ1
.
ϩ
0.04
0.002
0.003
0.005
0.01
negligible so that a linear dependence of kslow on amine concen-
tration is observed (Table 4) with a value for k7 of 250 30 dm3
molϪ1 sϪ1.
25
46
54
26
46
54
0.0125
a Calculated using eqn. (5) with the following values: items 1–8, k7
Reactions with n-butylamine
(5500 500) dm3 molϪ1
s
Ϫ1, kϪ7kAmH /kAm (0.39 0.05) sϪ1 and Kc,5
ϩ
(47 5) dm3 molϪ1; items 9–16, k7 (5200 500) dm3 molϪ1 sϪ1
,
Both processes, involving formation of the adducts 5d and 7d,
were measurable. Analysis of the data in Table 5 indicates that
the proton transfer equilibria between zwitterionic intermedi-
ϩ
k
Ϫ7kAmH /kAm (0.33 0.05) sϪ1 and Kc,5 (57 5) dm3 molϪ1; items 17–
21, k7 (4900 500) dm3 molϪ1
s
Ϫ1, kϪ7kAmH /kAm (0.12 0.03) sϪ1 and
ϩ
Kc,5 (100 50) dm3 molϪ1
.
ates and anionic adducts are rapid, and limits may be set at kAm
/
k
Ϫ5 > 2000 dm3 molϪ1 and kAm/kϪ7 > 1000 dm3 molϪ1. Hence
eqn. (2) reduces to eqn. (10). Values of kfast calculated with k5
ameter from eqn. (9), on the assumption that kAm is constant,
increased from 0.07 to 0.14 sϪ1 as the concentration of amine
salt was decreased from 0.1 to 0.02 mol dmϪ3. This allowed the
ϩ
ϩ
kϪ5ؒkAmH [AmH ]
kfast = k5[Am] ϩ
(10)
estimation of a value for kAmH kϪ7/kAm of 0.16 sϪ1 at a salt
kAm
[Am]
ϩ
concentration of 0.005 mol dmϪ3. Use of this value in eqn. (5)
together with values of k7 2900 300 dm3 molϪ1 sϪ1, and Kc,5
27 5 dm3 molϪ1 gave an excellent fit for items 19–24. Combin-
(1.7 0.2) × 104 dm3 molϪ1 sϪ1 and kϪ5ؒkAmH /kAm (155 15)
ϩ
s
Ϫ1 gave a good fit with this equation and lead to a value for Kc,5
of (110 20) dm3 molϪ1.
ϩ
ation of these values of k7 and kAmH kϪ7/kAm using eqn. (6)
leads to a value at low salt concentration of (1.8 0.4) × 104 for
Kc,7.
The values of kslow measured at 0.01 mol dmϪ3 salt concen-
tration correspond to eqn. (5) with k7 (140 20) dm3 molϪ1 sϪ1,
kϪ7ؒkAmH /kAm (0.025 0.005) sϪ1 and Kc,5 (90 10) dm3 molϪ1.
ϩ
Combination of these values leads to Kc,7 (5.6 1) × 103 dm3
molϪ1. In the presence of amine hydrochloride, 0.1 mol dmϪ3,
the values of best fit to eqn. (5) are k7 (115 10) dm3 molϪ1 sϪ1,
Reactions with pyrrolidine
Two reactions were observed, the faster of which was
inconveniently rapid for measurement. Data for the slower
reaction, corresponding to equilibration with adduct 7b, are in
Table 3. Values of kslow measured in the presence of 0.01 mol
dmϪ3 pyrrolidine perchlorate or hydrochloride go through a
maximum with increasing pyrrolidine concentration as pre-
dicted by eqn. (5). Combination of values using eqn. (6) gives
values of Kc,7 (1.4 0.2) × 104 and (1.6 0.3) × 104 dm3 molϪ1
respectively. In the presence of 0.1 mol dmϪ3 pyrrolidine hydro-
chloride the value of Kc,7 is increased to (4 1) × 104 dm3
molϪ1.
kϪ7ؒkAmH /kAm (0.0088 0.0010) sϪ1 and Kc,5 (205 20) dm3
ϩ
molϪ1. These allow the calculation of Kc,7 (1.3 0.2) × 104 dm3
molϪ1. As expected values of Kc,5 and Kc,7 are increased by a
factor of ca. 2 in the presence of 0.1 molar amine hydro-
chloride.
Reactions with benzylamine
Data for the slow reaction are in Table 6. They indicate that the
proton transfer equilibrium 6e 7e is rapid with kAm
k
/
Ϫ7 > 500 dm3 molϪ1. Values of rate constants give a good fit
with eqn. (5) with the parameters appending the Table.
Reactions with morpholine
Morpholine is considerably less basic than piperidine or pyr-
rolidine and the presence of 0.001 mol dmϪ3 amine salt was
sufficient to inhibit formation of the adduct 5c. A limit for
Kc,5 < 1 dm3 molϪ1 may be calculated. Values of kslow in Table 4
measured in the presence of 0.1 mol dmϪ3 morpholine hydro-
chloride give an excellent fit with eqn. (9), and lead to a value of
Kc,7 of (550 100) dm3 molϪ1. We may predict a value at 0.01
mol dmϪ3 amine salt for Kc,7 of 280 dm3 molϪ1. In the presence
of 0.001 mol dmϪ3 amine salt the final term in eqn. (9) becomes
Reactions with N-methylbenzylamine
In contrast with the results obtained for benzylamine the data
in Table 7 indicate that with N-methylbenzylamine proton
transfer is partially rate limiting in the formation of both 5f and
7f. It was possible to follow the formation of the 5-adduct in the
absence of amine salt. Items 1–3 in the Table show an approxi-
mately squared dependence on the amine concentration con-
sistent with eqn. (11) which is the form of eqn. (8) when
2476
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480

View Article Online
Table 5 Kinetic data for reaction of 3 with n-butylamine in DMSO at 25 ЊC
[n-Butylamine
[n-Butylamine]/
hydrochloride]/
Item
mol dmϪ3
mol dmϪ3
kfast/sϪ1
kcalc a/sϪ1
kslow/sϪ1
kcalc b/sϪ1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
0.004
0.006
0.008
0.010
0.002
0.004
0.006
0.008
0.010
0.020
0.040
0.050
0.002
0.003
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.001
0.001
0.001
0.001
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.010
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
118
136
164
195
—
114
136
165
198
—
—
—
—
—
—
—
—
—
0.39
0.59
0.75
0.77
0.78
0.64
0.38
0.33
0.67
0.61
0.66
0.83
0.98
1.09
1.13
1.17
1.22
0.41
0.57
0.70
0.77
0.79
0.64
0.38
0.31
0.67
0.63
0.67
0.79
0.92
1.04
1.14
1.21
1.26
—
—
340
280
300
460
740
810
—
—
—
—
—
340
310
300
390
660
810
—
—
—
—
—
—
—
—
—
—
—
—
—
a Calculated using eqn. (10) with the following values: items 1–4, k5 1.8 × 104 dm3 molϪ1
s
Ϫ1, kϪ5ؒkAmH /kAm 166 sϪ1; items 7–12, k5 1.6 × 104 dm3
ϩ
molϪ1 sϪ1, kϪ5ؒkAmH /kAm 145 sϪ1
.
b Calculated using eqn. (5) with the values: items 5–12, k7 146 dm3 molϪ1 sϪ1; kϪ7kAmH /kAm 0.025 sϪ1 and Kc,5 91 dm3
ϩ
ϩ
molϪ1; items 13–21, k7 115 dm3 molϪ1 sϪ1, kϪ7ؒkAmH /kAm 0.0088 sϪ1, and Kc,5 205 dm3 molϪ1
.
ϩ
Table 6 Kinetic data for reaction of 3 with benzylamine in DMSO at
25 ЊC
Comparisons
Results are summarised in Table 8, where they are compared
with corresponding values for reactions of 1,3,5-trinitro-
benzene, 2.
[Benzylamine
[Benzylamine]/
Hϩ, ClO4Ϫ]/
Item
mol dmϪ3
mol dmϪ3
kslow/sϪ1
kcalc a/sϪ1
The data show, in agreement with NMR results, that the 7-
adducts from 3 are considerably more stable thermodynamic-
ally than the isomeric 5-adducts. Thus values of Kc,7 are ca. 100
times larger than values of Kc,5. The actual value of the ratio
Kc,7/Kc,5 depends slightly on the nature of the amine, and
increases from 50 for n-butylamine and 73 for benzylamine to
300 for pyrrolidine and 600 for piperidine. These increases may
indicate some unfavourable steric interactions with the ortho-
nitro group in the adducts 5a and 5b formed by addition of the
bulkier secondary amines.
Nevertheless the 5-adducts are formed more rapidly than the
7-adducts and for some amines the kinetics of the faster reac-
tion were too fast for measurement by the stopped-flow
method. For the reactions of piperidine and n-butylamine k5/k7
has a value of ca. 100. Although a precise value of k5 is not
obtainable for reaction of N-methylbenzylamine the value of
k5kAm/kϪ5 is considerably higher than that of k7kAm/kϪ7, again
implying a higher value for k5 than for k7. It is also significant
1
2
3
4
5
6
7
8
9
10
11
12
13
0.002
0.003
0.005
0.01
0.02
0.03
0.04
0.06
0.01
0.02
0.03
0.04
0.06
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.10
0.10
0.10
0.10
0.10
0.82
0.55
0.45
0.50
0.77
0.92
0.99
0.98
1.12
1.04
1.24
1.46
1.81
0.72
0.55
0.45
0.51
0.75
0.92
0.99
0.99
1.22
1.10
1.25
1.45
1.83
a Calculated from eqn. (5) with the following values: items 1–8, k7
(40 5) dm3 molϪ1
s
Ϫ1, kϪ7ؒkAmH /kAm (0.13 0.02) sϪ1 and Kc,5
ϩ
(4.1 1) dm3 molϪ1. Items 9–13, k7 (33 5) dm3 molϪ1
s
Ϫ1, kϪ7ؒkAmH
/
ϩ
kAm (0.09 0.02) sϪ1 and Kc,5 (5 1) dm3 molϪ1
.
that, where measurable, the value of kϪ5kAmH /kAm is ca. 104
times larger than the value of kϪ7kAmH /kAm for the correspond-
ing amine. The value of kAm/kAmH , which reflects the acidity of
the adduct relative to that of the parent ammonium ion, will
not be expected to vary greatly with the position of attack.
Hence the implication is that kϪ5 ӷ kϪ7.
The kinetic preference for 5-attack, but the thermodynamic
preference for 7-attack, may be explicable in terms of the extent
of charge delocalisation possible in the adducts. Thus com-
parison of the resonance forms available, Scheme 4, indicates
the possibility of greater charge delocalisation in the 7-adducts.
This will result in greater stability of the 7-adducts coupled with
the higher kinetic barrier to their formation. Thus Bernasconi
has argued that increase in charge delocalisation generally
results in a higher intrinsic barrier to reaction.26
pKa values, measured in DMSO, for the conjugate acids of
the amines are given in Table 8. For the three cyclic secondary
amines, piperidine, pyrrolidine and morpholine, there is a
reasonably good correlation of values of Kc,5 and Kc,7 with bas-
ϩ
kfast = K5kAm[Am]2
(11)
ϩ
ϩ
k
Ϫ5 ӷ kAm[Am]. A value for K5kAm of (1.6 0.4) × 105 dm6
molϪ2 sϪ1 is obtained.
ϩ
K7kAm[Am]2 ϩ kAmH [AmH ]
ϩ
kslow
=
(12)
1 ϩ kAm[Am]/kϪ7
The presence of 0.01 mol dmϪ3 amine salt is sufficient to
inhibit formation of 5f and the data are fitted by eqn. (12)
which is a form of eqn. (4) when Kc,5 Ӷ 1. Items 4–10 in Table 7
yield values for K7kAm 8600 2000 dm6 molϪ2 sϪ1, kAm/kϪ7
25 10 dm3 molϪ1 and kAmH 170 10 dm3 molϪ1 sϪ1. Combin-
ϩ
ation of these values gives k7 340 100 dm3 molϪ1 sϪ1 and Kc,7
50 10 dm3 molϪ1. The only changes in these parameters in the
presence of 0.1 mol dmϪ3 salt, items 11–16, are a reduction in
the value of kAmH to 75 dm3 molϪ1 sϪ1 leading to a value for
ϩ
Kc,7 of 110 20 dm3 molϪ1.
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480
2477

View Article Online
Table 7 Kinetic data for reaction of 3 with N-methylbenzylamine in DMSO at 25 ЊC
[N-Me benzylamine]/
[N-Me benzylamine
HϩClO4Ϫ]/mol dmϪ3
kfast a/[Am]2
Item
mol dmϪ3
kfast/sϪ1
dm6 molϪ2 sϪ1
kslow/sϪ1
kcalc b/sϪ1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0.0125
0.0175
0.025
0.003
0.005
0.01
0.02
0.03
0.04
0.05
0.001
0.002
0.003
0.005
0.010
0.0125
0
0
0
32
41
90
—
—
—
—
—
—
—
—
—
—
—
—
—
2 × 105
—
—
—
—
—
—
1.3 × 105
1.5 × 105
—
—
—
—
—
—
—
—
—
—
—
—
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.10
0.10
0.10
0.10
0.10
0.10
1.64
1.71
2.00
3.2
5.3
7.8
10.3
7.4
7.3
6.9
6.8
6.6
6.6
1.65
1.70
2.05
3.4
5.4
7.7
10.3
7.3
7.2
7.0
6.8
6.7
6.7
—
3
a Corresponds to K5kAm, from eqn. (11). b Calculated from eqn. (12) with K7kAm 8600 dm6 molϪ2 sϪ1, kAm/kϪ7 25 dm3 molϪ1, and with kAmH 170 dm
ϩ
molϪ1 sϪ1 at 0.01 mol dmϪ3 salt and 75 dm3 molϪ1 sϪ1 at 0.1 mol dmϪ3 salt.
Table 8 Summary of results in DMSO at 25 ЊC
Piperidine
10.85
Pyrrolidine
11.06
Morpholine
9.15
n-Butylamine
Benzylamine
10.16
N-Methylbenzylamineb
a
pKa
11.12
9.30c
Reaction with 3d
k5/dm3 molϪ1 sϪ1
1.7 × 105
6.3 × 103
—
—
—
—
1.7 × 104
—
—
—
—
ϩ/sϪ1
kϪ5ؒkAmH
155
kAm
Kc,5/dm3 molϪ1
27
52
—
5300
<1
—
300
110
4.1
—
40
<1
—
340
3
kAmH /dm molϪ1 sϪ1
6 × 105
2900
ϩ
k7/dm3 molϪ1 sϪ1
140
ϩ/sϪ1
kϪ7ؒkAmH
0.16
0.36
1.1
0.025
0.13
6.8
kAm
Kc,7/dm3 molϪ1
1.8 × 104
—
1.5 × 104
—
280
—
5.6 × 103
300
—
50
170
3
kAmH /dm molϪ1 sϪ1
ϩ
Reaction with 2e
k1/dm3 molϪ1 sϪ1
>2 × 105
7.5 × 105
>100
>2.5
4.5 × 104
1.3 × 104
—
—
ϩ/sϪ1
kϪ1ؒkAmH
>100
210
45
120
kAm
Kc,1/dm3 molϪ1
2.1 × 103
280
3.5 × 103
3 × 103
34
250
1 × 103
6 × 104
105
—
—
3
kAmH /dm molϪ1 sϪ1
1.5 × 104
ϩ
a Values for corresponding ammonium ions, from ref. 25. b For N-methylbenzylamine the value of k5kAm/kϪ5 is 1.5 × 105 dm6 molϪ2 sϪ1, and the value
of k7kAm/kϪ7 is 8.6 × 103 dm6 molϪ2 sϪ1
.
c M. R. Crampton and L. C. Rabbitt, unpublished observation. d Values at 0.01 mol dmϪ3 salt concentration.
Ϫ3
e Refer to Scheme 2. Kc,1 = k1ؒkAm/kϪ1ؒkAmH . Data from refs. 11–13, with 0.1 mol dm salt concentration. A statistical correction has not been
ϩ
applied.
Similarly comparison of the data for the two primary amines,
n-butylamine and benzylamine, indicates that the higher
basicity of n-butylamine is reflected in higher values of Kc,5, Kc,7
and k7. Nevertheless when comparing the data for the second-
ary amines with those for the primary amines, the relatively
high nucleophilicities of the secondary amines are noteworthy.
For example the value of k7 for reaction with pyrrol-
idine is larger by a factor of ca. 40 than that for reaction with
n-butylamine, an amine with a similar pKa value. Likewise the
value for k7 for N-methylbenzylamine is almost tenfold higher
than that for benzylamine, despite the higher proton basicity
of the latter. It should, however, be remembered that while
nucleophilicity is measured here by rate constants for attack at
a carbon atom, the pKa values refer to equilibrium protonation.
It has been argued25 that primary ammonium ions are particu-
Scheme 4
larly well stabilised in DMSO by hydrogen bonding of the NHϩ
icity. Thus values for morpholine, the least basic amine, are
considerably lower than those for piperidine. Also nucleo-
philicity, as measured by values of k7, follows the same pattern.
protons to the solvent. Such stabilisation may be less effective in
the transition states for nucleophilic attack, where charge
development is only partial. The variation with the nature of
2478
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480

View Article Online
ϩ
the amine of values of k7 and Kc,7 indicates that steric hindrance
either to nucleophilic attack at the 7-position, or in the adducts
7, is not an important factor. The one exception may be in the
adduct 7f formed from N-methylbenzylamine where the value
of Kc,7 is unexpectedly low.
the parameter kϪ7kAmH /kAm for reaction of 3 are considerably
ϩ
lower than the corresponding values of kϪ1kAmH /kAm for reac-
tion of 2. This indicates that values of kϪ7 are likely to be con-
siderably lower than values of kϪ1. Thus the difference in the
nature of the rate determining step in the reactions of 3 and 2
with amines is likely to be a consequence both of higher rate
constants for proton transfer with 3 and also a reduction in the
value of kϪ7 relative to kϪ1.
Comparison with 1,3,5-trinitrobenzene
1,3,5-Trinitrobenzene, 2, is the bench-mark for comparisons of
reactivity in σ-adduct forming reactions.3 The results in Table 8
show that values of Kc,7 for reaction at the 7-position of 3 are
between three and eight times larger, for corresponding amines,
than the values of Kc,1 for reaction at an unsubstituted position
of 1,3,5-trinitrobenzene. (The values of ratios would be three
times larger if a statistical correction were to be applied to
account for the equivalent positions in 2.) This confirms the
excellent ability, reported previously21,22,27 in reactions with
alkoxides, of the 4-nitrobenzofuroxan system in delocalising
negative charge. Despite the higher thermodynamic stabilities
of adducts from 3, the values of k7 for nucleophilic attack are
considerably lower than values of k1 for reaction with 2. This
may be a consequence of the possibility, referred to earlier, of
the extensive charge delocalisation in the benzofuroxan system
which leads to higher intrinsic barriers to reaction.
The interesting exception is the reaction of 3 with N-methyl-
benzylamine. Here proton transfer is partially rate-limiting in
formation of both the 5-adduct and the 7-adduct. Since these
proton transfers must occur from the respective zwitterions 4f
and 6f to N-methylbenzylamine, large steric effects are likely.
ϩ
Hence values of kAm and kAmH will be reduced.
Experimental
4-Nitrobenzofuroxan was prepared, as previously described,28
by nitration of benzofuroxan: mp 143 ЊC (lit.28 mp 143 ЊC).
Amines and solvents were the purest available commercial
products. Amine salts were prepared as solutions in DMSO by
accurate neutralisation of amines, with the appropriate acid.
1H NMR spectra were recorded in [2H6]-DMSO with a
Varian Mercury 200 MHz spectrometer. UV–vis spectra and
kinetic measurements were made at 25 ЊC with a Perkin-Elmer
Lambda 2 spectrophotometer, a Shimadzu UV-2101 PC spec-
trometer or an Applied Photophysics SX-17 MV stopped-flow
spectrometer. Reported rate constants are the means of several
determinations and are precise 5%.
A further difference in the reactions of 3 and 2 with amines is
in the nature of the rate determining step. With 3 the proton
transfer equilibria between zwitterions and anions 6 7, are
generally rapid so that nucleophilic attack, the k7 step, is rate-
limiting in the forward direction. The only exception is in the
reaction with the sterically hindered amine N-methylbenzyl-
amine where proton transfer becomes partially rate determin-
ing. By contrast in the reaction of 1,3,5-trinitrobenzene with
amines the proton transfer step is generally rate limiting with
secondary amines and may be partially rate-limiting with pri-
mary amines.11 It should be noted that the zwitterionic inter-
mediates for both 2 and 3 are expected to be considerably more
acidic than the corresponding ammonium ions. In the case of 2
Acknowledgements
We thank Zeneca Fine Chemicals, Huddersfield for financial
support.
References
a value for kAm/kAmH of ca. 500 has been estimated,11 and the
ϩ
1 J. A. Orvik and J. F. Bunnett, J. Am. Chem. Soc., 1970, 92,
2417.
2 C. F. Bernasconi, MTP Int. Rev. Sci.: Org. Chem., Ser. One, 1977, 3,
33.
3 F. Terrier, Nucleophilic Aromatic Displacement, VCH, New York,
1991.
4 R. Chamberlin and M. R. Crampton, J. Chem. Soc., Perkin Trans. 2,
1995, 1831.
5 R. Chamberlin and M. R. Crampton, J. Chem. Soc., Perkin Trans. 2,
1994, 425.
6 S. Sekiguchi, M. Hosokawa, T. Suzuki and M. Sato, J. Chem. Soc.,
Perkin Trans. 2, 1993, 1111.
failure to observe zwitterionic intermediates in reactions of 3
ϩ
indicates that kAm/kAmH ӷ 1. Hence the proton transfers from
zwitterions to amines, kAm, are thermodynamically favourable
processes. Nevertheless in the reactions of 2 values of kAm are
considerably below the diffusion limit. For example kAm has the
value 5 × 104 dm3 molϪ1 sϪ1 for the reaction involving piper-
idine.11 It has been argued that steric hindrance at the reaction
centre is responsible, and that this is more severe with second-
ary than with primary amines.11,15 It should be noted that this is
steric hindrance to proton transfer between the zwitterion and
amine as distinct from steric hindrance to nucleophilic attack.
The latter is thought to be unimportant when reaction occurs at
a ring carbon atom carrying hydrogen.
7 M. R. Crampton and P. J. Routledge, J. Chem. Soc., Perkin Trans. 2,
1984, 573.
8 J. F. Bunnett, S. Sekiguchi and L. A. Smith, J. Am. Chem. Soc., 1981,
103, 4865.
The results for adduct formation from 3 show that the proton
transfer equilibria are generally rapid and are not rate-limiting.
This suggests that rate constants for proton transfer may be
faster in the 4-nitrobenzofuroxan system than in the 1,3,5-
trinitrobenzene system. This would be reasonable since 3 is a
considerably less sterically demanding reactant than is 2, where
reaction must occur ortho to two nitro-groups. There is evidence
that this conclusion is valid in formation of the 5-adducts, 5.
9 Y. Hasegawa, J. Chem. Soc., Perkin Trans. 2, 1985, 87.
10 C. F. Bernasconi, M. C. Muller and P. Schmid, J. Org. Chem., 1979,
44, 3189.
11 M. R. Crampton and B. Gibson, J. Chem. Soc., Perkin Trans. 2,
1981, 533.
12 M. R. Crampton and C. Greenhalgh, J. Chem. Soc., Perkin Trans. 2,
1983, 1175.
13 R. Chamberlin, M. R. Crampton and I. A. Robotham, J. Phys. Org.
Chem., 1996, 9, 152.
14 E. Buncel and W. Eggimann, J. Am. Chem. Soc., 1977, 99, 5958.
15 M. R. Crampton, P. J. Routledge and P. Golding, J. Chem. Soc.,
Perkin Trans. 2, 1984, 329.
ϩ
Thus in the reaction with piperidine kAmH has a value of
6 × 105 dm3 molϪ1 sϪ1 compared with a corresponding value
of 280 dm3 molϪ1 sϪ1 in the 1,3,5-trinitrobenzene system.
Although these values refer to the reverse proton transfers,
from piperidinium ions to the anionic adducts, they will largely
reflect the steric situation at the reaction centre. The factor of
2000 is consistent with less steric hindrance in the reaction
involving 3. Nevertheless for reaction at the 7-position of 3 we
must add the caveat that the nature of the rate determining step
depends on the value of the ratio kAm/kϪ7 rather than on the
value of kAm alone. The results in Table 8 show that values of
16 P. B. Ghosh and M. W. Whitehouse, J. Med. Chem., 1968, 11,
305.
17 M. W. Whitehouse and P. B. Ghosh, Biochem. Pharmacol., 1968, 17,
158.
18 P. B. Ghosh, B. Ternai and M. W. Whitehouse, J. Med. Chem., 1972,
15, 255.
19 F. Terrier, F. Millot, A.-P. Chatrousse, M.-J. Pouet and M.-P.
Simonnin, Org. Magn. Reson., 1976, 8, 56.
20 E. Buncel, N. Chuaqui-Offermanns, B. K. Hunter and A. R. Norris,
Can. J. Chem., 1977, 55, 2852.
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480
2479

View Article Online
21 F. Terrier, A.-P. Chatrousse and F. Millot, J. Org. Chem., 1980, 45,
2666.
22 L. Di Nunno, S. Florio and P. E. Todesco, J. Chem. Soc., Perkin
Trans. 2, 1975, 1469.
26 C. F. Bernasconi, Pure Appl. Chem., 1982, 2335; Acc. Chem. Res.,
1987, 20, 301; Adv. Phys. Org. Chem., 1992, 27, 119.
27 D. Dalmonte, E. Sandri, L. Di Nunno, S. Florio and P. E. Todesco,
J. Chem. Soc. (B), 1971, 2209.
23 R. A. Manderville and E. Buncel, J. Chem. Soc., Perkin Trans. 2,
1993, 1887.
28 P. Drost, Justus Liebigs Ann. Chem., 1899, 307, 49.
24 D. W. S. Latham, O. Meth-Cohn and H. Suschitzky, J. Chem. Soc.,
Perkin Trans. 1, 1976, 2216.
25 M. R. Crampton and I. A. Robotham, J. Chem. Res. (S), 1997, 22.
Paper 9/05774I
2480
J. Chem. Soc., Perkin Trans. 2, 1999, 2473–2480