Kinetic and equilibrium studies of σ-adduct formation and nucleophilic substitution in the reactions of 2-phenoxy-3,5-dinitropyridine and 2-ethoxy-3,5-dinitropyridine with aliphatic amines in dipolar aprotic solvents

Article abstract of DOI:10.1039/b211639c

The reactions of aliphatic amines with 2-phenoxy-3,5-dinitropyridine, 4, and 2-ethoxy-3,5-dinitropyridine, 5, in DMSO result in the rapid reversible formation of anionic σ-adducts at the 6-position. Kinetic studies show that proton transfer from the initially formed zwitterions to base may be rate-limiting. Slower reactions result, except in the case of 5 and piperidine, in displacement of the 2-substituent via intermediates which have lower thermodynamic stabilities than their 6-isomers. Base catalysis of the substitution process is attributed in the case of 4 to rate-limiting proton transfer from zwitterionic intermediates, but in 5 to acid catalysis of ethoxide departure (SB-GA mechanism). X-Ray crystallography of 5 shows a planar non-strained structure although the structure of 2-piperidino-3,5-dinitropyridine, 10c, shows distortion resulting from steric interactions of the 2- and 3-substituents. Kinetic and equilibrium results are compared with those for related reactions of the more sterically strained 2,4,6-trinitrobenzene derivatives. Results for the reactions of 4 and 5 with pyrrolidine in three dipolar aprotic solvents are compared. Values of equilibrium constants for σ-adduct formation decrease in the order DMSO > DMF ? Acetonitrile, while values of rate constants for proton transfer are in the reverse order.

Full text of DOI:10.1039/b211639c

View Article Online / Journal Homepage / Table of Contents for this issue
Kinetic and equilibrium studies of ꢀ-adduct formation and
nucleophilic substitution in the reactions of 2-phenoxy-3,5-
dinitropyridine and 2-ethoxy-3,5-dinitropyridine with aliphatic
amines in dipolar aprotic solvents†
Michael R. Crampton,*a Thomas A. Emokpae,b Judith A. K. Howard,a Chukwuemeka Isanborb
and Raju Mondala
a Chemistry Department, Durham University, Durham, UK DH1 3LE
b Department of Chemistry, University of Lagos, Nigeria
Received 29th November 2002, Accepted 28th January 2003
First published as an Advance Article on the web 20th February 2003
The reactions of aliphatic amines with 2-phenoxy-3,5-dinitropyridine, 4, and 2-ethoxy-3,5-dinitropyridine, 5, in
DMSO result in the rapid reversible formation of anionic σ-adducts at the 6-position. Kinetic studies show that
proton transfer from the initially formed zwitterions to base may be rate-limiting. Slower reactions result, except in
the case of 5 and piperidine, in displacement of the 2-substitutent via intermediates which have lower thermodynamic
stabilities than their 6-isomers.
Base catalysis of the substitution process is attributed in the case of 4 to rate-limiting proton transfer from
zwitterionic intermediates, but in 5 to acid catalysis of ethoxide departure (SB-GA mechanism). X-Ray crystal-
lography of 5 shows a planar non-strained structure although the structure of 2-piperidino-3,5-dinitropyridine, 10c,
shows distortion resulting from steric interactions of the 2- and 3-substituents. Kinetic and equilibrium results are
compared with those for related reactions of the more sterically strained 2,4,6-trinitrobenzene derivatives.
Results for the reactions of 4 and 5 with pyrrolidine in three dipolar aprotic solvents are compared. Values of
equilibrium constants for σ-adduct formation decrease in the order DMSO > DMF ӷ Acetonitrile, while values
of rate constants for proton transfer are in the reverse order.
The reactions of alkyl 2,4,6-trinitrophenyl ethers 1 with nucleo-
philes to give σ-adducts are well documented.1–3 The usual
pattern is the initial formation of adducts at the 3-position
under kinetic control followed by isomerisation to give the
thermodynamically more stable adducts at the 1-position. That
the relative stabilities of 1-adducts may be ca. 104 times that of
the corresponding 3-adducts has been attributed to the relief
of steric strain in 1 accompanying reaction at the 1-position,
together with the stabilising effects of di-alkoxy substitution at
an sp3-hybridised carbon atom. That attack at the 3-position is
faster is thought to be due to the lower F-strain associated with
attack at an unsubstituted position relative to attack at a sub-
stituted position, and to the possibility of partially conserving
the resonance interaction present in 1 between the alkoxy group
and the nitro-groups in the transition state for attack at the
3-position but not at the 1-position.1–4
steric effects, and their relief, in directing the course of σ-adduct
formation.
Here we report on the reactions of 2-phenoxy-3,5-dinitro-
pyridine 4 and of 2-ethoxy-3,5-dinitropyridine 5 with aliphatic
amines in dipolar aprotic solvents. Most results refer to di-
methyl sulfoxide (DMSO) but comparison is made with results
in acetonitrile and in dimethylformamide. As with methoxide
initial attack is at the 6-position, however, the use of amine
nucleophiles allows the slower attack at the 2-position, leading
to irreversible substitution, to be investigated. There have been
no previous reports of the reactions of the alkyl ethers 2 or 5
with amines but kinetic studies have reported base catalysis
in the amino-dephenoxylations of 4 in acetonitrile,8 benzene,9
and methanol.10 There is current interest in the mechanism
of such catalysis. Studies with trinitro-activated substrates
have established that base catalysis may indicate rate-limiting
proton-transfer from a zwitterionic intermediate to yield an
anionic intermediate. Alternatively it may indicate general acid
catalysis of leaving group departure in the SB-GA mechanism
of substitution.3,11–14 Our results provide evidence of both these
pathways in reactions of the dinitropyridyl ethers.
There have been few studies involving the reactions of
2-alkoxy-3,5-dinitropyridines, but intriguingly it has been
reported that reaction of the 2-methoxy compound, 2, with
methoxide yields the σ-adduct, 3, as the kinetically and
thermodynamically preferred product.5–7 Consequently the
adduct at the 2-position has not been observed. This switch
in reactivity pattern compared to 1 confirms the importance of
Results and discussion
The UV–visible spectra of the phenoxy ether, 4, in DMSO con-
taining n-butylamine, pyrrolidine or piperidine showed that the
initial reactions produced species absorbing strongly at 490–495
nm. This absorption rapidly faded to give spectra which were
† Electronic supplementary information (ESI) available: Tables 10–15.
See http://www.rsc.org/suppdata/ob/b2/b211639c/
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
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
1004

View Article Online
Table 1 Spectroscopic data for reactants, products and intermediates in DMSO
δ
UV absorbance
H4
H6
2-substituent
λmax/nm
ε/104 dm3 molϪ1 cmϪ1
4
5
9.14
9.34
9.27
9.14
9.11
8.94
7.2–7.6(5H)
290
290
345
1.05
1.1
1.3
4.63(OCH2), 1.38(CH3)
9.3(NH), 3.65, 1.60
1.33(CH2), 0.88(CH3)
3.58(4H), 1.70(4H)
3.54(4H), 1.62(6H)
4.1(OCH2), 1.2(CH3)
4.1(OCH2), 1.2(CH3)
10a
10b
10c
7b, R=Et
7c, R=Et
9.08
9.11
8.45
8.52
8.83
8.82
5.69
5.50
363
368
498
495
1.8
1.9
2.4
2.4
Scheme 1
identical to those of the expected substitution products, 10a–c,
in the same medium. The reactions of the ethoxy ether, 5, with
n-butylamine and with pyrrolidine were similar to those of 4,
although the initial products, λmax 495–498 nm, were much
longer lived. However, in the reaction of 5 with piperidine the
initial product, λmax 495 nm, was stable for several hours and the
slow fading process did not yield 10c.
equilibration of reactants with 7, while the slower process yields
the substitution product, 10. In the presence of high amine
concentrations the products, 10, were themselves found, as
mentioned later, to be in rapid equilibrium with their 6-amino
adducts. It is worth noting that adducts 9, with lower thermo-
dynamic stabilities than their isomers 7, will not be observable
in the substitution process even though they may be reaction
intermediates.
Rate constants were measured under first order conditions.
For reactions with buffers (amine plus amine salt) the buffer
components were in large excess of the substrate concentration,
which was usually 5 × 10Ϫ5 mol dmϪ3. For adduct formation in
the absence of added salt, sufficient excess of amine was used
to ensure that >95% conversion was achieved, thus isolating the
forward reaction. Making the assumption that the zwitterionic
form 6 may be treated as a steady state intermediate leads12,14,15
to eqn. (1) for the general rate expression for equilibration
between reactants and 7. When kAm [Am] ӷ kϪ6, corresponding
to rapid proton transfer, eqn. (1) reduces to eqn. (2). The
Spectroscopic data for reactants and products are in Table 1.
In the case of the reactions of 5 with piperidine and pyrrolidine
it was possible to record 1H NMR spectra for the initially
produced species. These show resonances at ca. δ 5.6 and
8.5 due to ring hydrogens consistent with nucleophilic attack at
the unsubstituted 6-position to yield adducts of type 7. The
large shift to lower frequency of H6 is consistent with the
change to sp3 hybridisation associated with reaction at this
position.1–3 The observed spectra are not consistent with the
observation of adducts 9 formed by reaction at the 2-position,
but might be compatible with adducts formed by attack at
the 4-position. However, this possibility is rendered unlikely
both by analogy with methoxide attack on 2 which is known,5,6
from the use of the 6-deuterio labelled derivative, to give 3,
and by the knowledge1–3 that a para-aza function is far
less stabilising in σ-adduct forming reactions than a para-
nitro group. Hence our results are interpreted in terms of
Scheme 1.
(1)
Kinetic study of these reactions showed the presence of two
processes well separated in time. The more rapid process,
observable by stopped-flow spectrophotometry, is attributed to
(2)
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1005

View Article Online
Table 2 Kinetic results for reaction of 4 with n-butylamine in DMSO at 25 ЊC
[n-Butylamine]/mol dmϪ3
[n-Butylammonium perchlorate]/mol dmϪ3
kfast a/sϪ1
kslow b/sϪ1
kcalc c/sϪ1
0.006
0.008
0.010
0.002
0.005
0.010
0.020
0.003
0.005
0.010
0.020
0.040
0
0
0
91
110
130
—
—
—
—
—
—
—
—
—
—
—
—
—
0.001
0.001
0.001
0.001
0.010
0.010
0.010
0.010
0.010
0.017
0.038
0.049
0.036
0.023
0.038
0.073
0.123
0.137
0.016
0.033
0.044
0.038
0.024
0.039
0.074
0.121
0.140
—
—
a Measured as a colour forming reaction at 495 nm. b Measurement as a colour forming reaction at 345 nm or 405 nm gave identical results.
c Calculated from eqn. (7) with k2 8 dm3 molϪ1 sϪ1 and Kc,6 8 dm3 molϪ1
.
other limiting form, when kϪ6 ӷ kAm [Am], corresponds to rate
limiting proton transfer and is given by eqn. (3), where K6 = k6/
Results in DMSO
The kinetic data in Table 2 indicate that in the reaction of 4 with
n-butylamine nucleophilic attack is rate limiting in both the
formation of adduct 7a and in nucleophilic substitution. In the
absence of added salt values of kfast increase linearly with n-
butylamine concentration giving, from eqn. (2), a value for k6
k
Ϫ6. The stoichiometric equilibrium constant for formation of
the 6-adducts, 7, is defined by eqn. (4).
(3)
(4)
of 1.5 × 104 dm3 molϪ1 Ϫ1. Values of kslow for the substitution
s
process accord with eqn. (7) with k2 8 dm3 molϪ1 sϪ1 and Kc,6
8
dm3 molϪ1. The same values were obtained for these parameters
at both 10Ϫ3 and 10Ϫ2 mol dmϪ3 salt concentration, indicating
the absence of major salt effects.
The rate expression for the irreversible substitution reaction
is eqn. (5), which may be written as eqn. (6) where K2
In contrast the results in Table 3 for the formation of 7c from
4 and piperidine show that proton transfer is rate-limiting so
that the data are fitted by eqn. (3) with K6kAm 6.6 × 105 dm6
ϩ
=
k2/kϪ2. The limiting forms are eqn. (7) when kB[Am] ӷ kϪ2, and
eqn. (8) when kϪ2 ӷ kB[Am]. Here kB is the rate constant
representing the base catalysed conversion of the zwitterion 8
to product. The significance of kB depends on the mechanism
molϪ2 sϪ1 and kAmH 2.5 × 104 dm3 molϪ1 Ϫ1. Combination of
s
these values gives Kc,6 26 dm3 molϪ1. Proton transfer is partially
rate limiting in the substitution reaction and the data require
the use of eqn. (6) with K2kB 1200 dm6 molϪ1 sϪ1, kB/kϪ2 12 dm3
molϪ1 and Kc,6 26 dm3 molϪ1. Combination of the former two
of base catalysis. When proton transfer from 8 to give 9 is
᎐
rate limiting then k ᎐ kЈAm, while in the case of rate limiting
᎐
B
values yields k2 100 dm3 molϪ1 sϪ1
.
acid-catalysed loss of alkoxide (SB-GA mechanism) kB = kЈAm
ϩ
k4/kЈAmH . It should be noted that values of rate constants for
A situation intermediate between n-butylamine and piper-
idine is encountered in reaction of 4 with pyrrolidine. Here
proton transfer is rate limiting in the formation of 7b, with
proton transfer will vary depending on the ring position
ϩ
at which attack has occurred. Hence kAm ≠ kЈAm and kAmH
≠
K6kAm 1.4 × 107 dm6 molϪ1 sϪ1 and kAmH 4 × 10 dm molϪ1 sϪ1
,
5
3
ϩ
kЈAmH
.
ϩ
and nucleophilic attack is rate limiting in the substitution
reaction, with k2 146 dm3 molϪ1 Ϫ1. Data here are given in
s
Table 10 as supplementary information.†
Results for reaction of 5, the ethyl ether, with n-butylamine
are in Table 4. The rapid reaction forming 7a, is at the limit
of measurement by the stopped-flow method. Data are
consistent with rate limiting nucleophilic attack, eqn. (2), and
yield approximate values for k6 7000 dm3 molϪ1 sϪ1 and
(5)
(6)
(7)
(8)
kϪ6 kAmH /kAm 8000 sϪ1. Combination of these values gives
ϩ
Kc,6 0.9 dm3 molϪ1. The substitution reaction differs from the
corresponding reaction of 4 in that base catalysis is observed
and the data are fitted by eqn. (6). The values required
for optimum fit are given at the foot of Table 4. They vary
slightly with salt concentration, but yield a value of k2 of
5
1 dm3 molϪ1 sϪ1
.
The results (Table 11, supplementary information†) for
the reaction of 5 with pyrrolidine show that in the formation
of 7b proton transfer is rate limiting. The value of K6kAm is
5 × 106 dm6 molϪ1 sϪ1 and equilibrium constant Kc,6 has the
value 2.4 dm3 molϪ1. Combination of these values gives kAmH
ϩ
2 × 106 dm3 molϪ1 sϪ1. Proton transfer is also rate limiting in the
substitution reaction with K2kB 3.3 dm6 molϪ2 sϪ1
.
The formation of 7c from 5 and piperidine follows the
pattern of the reaction with pyrrolidine. The results (Table 12,
supplementary in formation†) yield K6kAm 2 × 105 dm3 molϪ1 sϪ1
and Kc,6 1.6 dm3molϪ1. Combination of these values gives
ϩ
kAmH 1.2 × 105 dm3 molϪ1 Ϫ1. However, the fading reaction is
s
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1006

View Article Online
Table 3 Kinetic results for the reaction of 4 with piperidine in DMSO at 25 ЊC
[piperidine]/mol dmϪ3
[piperidine perchlorate]/mol dmϪ3
kfast a/sϪ1
kcalc b/sϪ1
kslow c/ 10Ϫ2 sϪ1
kcalc d/ 10Ϫ2 sϪ1
0.008
0.010
0.020
0.030
0.004
0.005
0.008
0.010
0.020
0.030
0.040
0
0
0
0
0.01
0.01
0.01
0.01
0.01
0.01
0.01
46
65
260
600
—
—
—
330
520
(840)
—
42
66
260
600
—
—
—
320
510
(850)
—
—
—
—
—
1.6
2.6
6.1
9.4
20
24
24
—
—
—
—
1.7
2.6
6.0
8.5
19
24
25
a Colour forming process at 495 nm. b Calculated from eqn. (3) with K6kAm 6.6 × 105 dm6 molϪ2 sϪ1 and kAmH 2.5 × 104 dm3 molϪ1 sϪ1
.
c Colour
ϩ
forming process at 368 nm. d Calculated from eqn. (6) with K2kB 1200 dm6 molϪ2 sϪ1, kB/kϪ2 12 dm3 molϪ1 and Kc,6 26 dm3 molϪ1
.
Table 4 Kinetic results for reaction of 5 with n-butylamine in DMSO at 25 ЊC
b
[n-Butylamine]/ mol dmϪ3
[n-Butylammonium perchlorate]/mol dmϪ3
kfast a/sϪ1
kcalc
kslow c/10Ϫ2 sϪ1
kcalc d/10Ϫ2 sϪ1
0.020
0.050
0.10
0.001
0.002
0.005
0.010
0.020
0.040
0.10
0.001
0.002
0.005
0.010
0.020
0.040
—
—
—
150
300
500
—
—
—
(820)
600
600
700
—
—
—
140
350
700
—
—
—
(820)
550
500
800
—
—
—
—
—
—
—
—
—
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.01
0.01
0.01
0.01
0.01
0.01
0.13
0.42
1.6
3.4
6.6
8.8
7.1
0.098
0.31
1.3
3.0
6.8
14
0.13
0.41
1.6
3.9
7.3
9.2
6.1
0.095
0.31
1.3
3.1
7.0
14
—
—
—
—
—
—
a Colour forming reaction at 495 nm. b Calculated from eqn. (2) with k6 7000 dm3 molϪ1 sϪ1 and kϪ6kAmH /kAm 8000 sϪ1 c Colour forming process at
.
ϩ
345 nm. d Results at 0.001 mol dmϪ3 salt calculated from eqn. (6) with K2kB 1600 dm6 molϪ2 sϪ1, kB/kϪ2 280 dm3 molϪ1 and Kc,6 0.8 dm3 molϪ1; results
at 0.01 mol dmϪ3 salt with K2kB 1200 dm6 molϪ2 sϪ1, kB/kϪ2 280 dm3 molϪ1 and Kc,6 1.0 dm3 molϪ1
.
extremely slow, and does not, from the NMR evidence, yield the
expected substitution product.
Before discussing the kinetic data it is useful to consider
X-ray crystal structures, which have been determined for the
reactant 5 and the substitution product with piperidine, 10c.
The structure of 5, shown in Fig. 1a, indicates a nearly planar
molecule with the ethoxy group syn to the ring nitrogen. The
nitro-group at the ortho position shows only a small degree
of twisting, 13Њ, from the ring plane and there is evidence
from the shortened 05–C1 (1.319Å) and C2–C3 (1.375Å) bond
lengths for conjugation between ethoxy and nitro groups as
shown in 11. There is similarly evidence for a relatively
unstrained molecule in the corresponding methyl ether.16
Fig. 1 X-Ray crystallographic structures of a) 5 and b) 10c.‡
was a faster process than at the substituted 2- position. The
kinetic results obtained in DMSO are summarised in Table 5.
In the case of n-butylamine the ratio of rate constants k6/k2 is
2000 for 4 and 8000 for 5. An important factor here is likely to
be the stabilisation of reactants by cross-conjugation between
In contrast the structure of 10c, shown in Fig. 1b, indicates
considerable deformation of the ring, with deviations from
planarity of ca 20Њ. The ortho-nitro group is twisted by 30Њ
indicating considerable steric interaction with the adjacent
piperidyl group.
‡ CCDC reference numbers 198285–198286. See http://www.rsc.org/
suppdata/ob/b2/b211639c/ for crystallographic data in .cif or other
electronic format.
In all cases attack by amine at the unsubstituted 6-position
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1007

View Article Online
Table 5 Summary of rate and equilibrium data for reaction of 4 and 5
with amines in DMSO at 25 ЊC
The results in Table 5 show that the ratio of Kc,6 values for 4
and 5 varies between 9 and 16 for the three amines studied. That
the values for 5 are lower than for 4 is attributed to the greater
ground state stabilisation in the ethyl ether, referred to earlier,
which is progressively lost on passage to the transition state
and to the anionic adduct. Comparison of data for the three
amines shows that values of Kc,6 and also of k2 fall in the order
pyrrolidine > piperidine > n-butylamine which is the usual
order in reactions involving nucleophilic attack.12–15
Amine
n-Butylamine
4
5
k6/dm3 molϪ1 sϪ1
Kc,6/dm3 molϪ1
1.5 × 104
8
8
—
—
7 × 103
0.9
5
k2/dm3 molϪ1 sϪ1
K2kB/dm6 molϪ2 sϪ1
kB/kϪ2 dm3 molϪ1
1400
280
Base catalysis
In the reactions forming the σ-adducts 7, R = Ph and 7, R = Et,
from 4 and 5 respectively, nucleophilic attack is rate limiting
when the amine is n-butylamine but proton transfer is rate
limiting with pyrrolidine and piperidine. A similar situation
applies12,13 in reactions of the corresponding trinitro-com-
pounds 1 R = Ph, and 1, R = Et. In these reactions, involving
Pyrrolidine
Kc,6/dm3 molϪ1
40
2.4
K6kAm/dm6 molϪ2 sϪ1
1.4 × 107
5 × 106
3
kAmH /dm molϪ1 sϪ1
4 × 105
146
—
2 × 106
—
3.3
ϩ
k2/dm3 molϪ1 sϪ1
K2kB/dm6 molϪ2 sϪ1
strongly activated substrates, the proton-transfer process 6
7,
is expected to be thermodynamically favoured as discussed
previously.12–15,21 Hence rate constants for proton-transfer will
be largely determined by steric factors. Since proton-transfer
involves approach of an amine molecule to the zwitterion, 6,
steric hindrance is likely to be higher for pyrrolidine and
piperidine than for n-butylamine. This explains the difference
in rate-determining step in the reactions involving secondary
and primary amines. The argument has been used that intra-
molecular hydrogen-bonding between an amino proton and an
ortho-nitro group plays an important role in differentiating the
behaviour of primary and secondary amines.3,22 However, we
have argued previously that in DMSO the acidic protons in 6
will be hydrogen-bonded to the solvent, a good hydrogen-bond
acceptor, rather than to an adjacent nitro-group.12–15
Piperidine
Kc,6/dm3 molϪ1
26
6.6 × 105
2.5 × 104
100
1200
12
1.6
K6kAm/dm6 molϪ2 sϪ1
2 × 105
1.2 × 105
—
3
kAmH /dm molϪ1 sϪ1
ϩ
k2/dm3 molϪ1 sϪ1
K2kB/dm6 molϪ2 sϪ1
kB/kϪ2 dm3 molϪ1
—
—
the ethoxy- and nitro-group4,17 as shown in 11, or phenoxy- and
nitro-group as shown in 12. Such stabilisation will be lost on
attack at the 2-position but may be partially conserved when
attack occurs at the 6-position. The higher reactivity ratio for
5 than for 4 is consistent with somewhat greater stabilisation
expected17 in 11 than in 12.
ϩ
It is interesting that values of kAmH are ca. one hundred
times greater for reactions involving 4 and 5 than for the
corresponding reactions involving 1, R = Ph and 1, R = Et,
respectively. For example in reactions with pyrrolidine values
are 4 4 × 105 dm3 molϪ1 sϪ1 1, R = Ph 3 × 103 dm3 molϪ1 sϪ1 and
5 2 × 106 dm3 molϪ1 sϪ1, 1, R = Et 1 × 104 dm3 molϪ1 sϪ1. These
factors which are likely also to reflect kAm values are attributable
to the lower steric hindrance to the approach of the amine to 6
relative to the approach to the corresponding zwitterions from
ϩ
1. All values of kAmH are about ten times lower for piperidine
than for pyrrolidine reflecting the higher steric requirements of
this amine.12–15
The reactions of amines at unsubstituted ring positions of
4 and 5, to give adducts 7, show similar susceptibility to base
catalysis. Nevertheless, the substitution reactions of 5, to give
10, are far more likely to be base catalysed than the corre-
sponding reactions of 4. Thus the reaction of 5 with pyrrolidine
is wholly base catalysed while in the reaction with 4 nucleophilic
attack is rate-limiting. Also substitutions of 5 are considerably
slower than those of 4. This difference in behaviour is attri-
butable to a change in the nature of the rate-limiting step.
Phenoxide is known to be a considerably better leaving group
than alkoxides;3 for example phenoxide departure is 106 times
faster23 than methoxide departure from 13. Also it is known13
that in the substitution reactions of 1, R = Ph, the trinitro-ana-
logue of 4, with amines base catalysis is due to rate-limiting
proton transfer from the zwitterionic intermediates. Hence
we similarly attribute the base catalysis observed in the reaction
of 4 with piperidine. In the reactions with n-butylamine and
pyrrolidine nucleophilic attack, the k2 step, is rate-limiting,
probably as a consequence of the relatively rapid deprotonation
of the zwitterionic intermediates. Although a value of K2kB
A further factor favouring attack at the 6-position is the
lower F-strain associated with nucleophilic attack at an un-
substituted ring position relative to attack at an atom carrying
an electronegative substituent.18,19 Also, the absence of steric
strain around the 2-position in the reactant molecules 4 and
5 means that there is little energy to be gained from its relief.
All these factors result in the 6-adducts being preferred, both
kinetically and thermodynamically, to the 2-adducts. This
contrasts with the behaviour of 1-ethoxy-2,4,6-trinitrobenzene,
1, R = Et, where there are considerable steric interactions
around the 1-position.20 Here the ratio of k3 : k1 of rate con-
stants for attack by n-butylamine at the unsubstituted 3-
position and the 1-position is only 13, while the relief of
steric strain resulting from adduct formation at the 1-position
results12 in a value for the ratio of equilibrium constants Kc,1
:
Kc,3 of 3 × 104.
Values of Kc,6 for formation of adducts 7a,b,c are between 15
and 30 times lower than values for Kc,3 corresponding to
reaction of the three amines at the unsubstituted 3-position
of 1-ethoxy-2,4,6-trinitrobenzene.12,14 This indicates that here
the ortho-aza function is rather less activating than on ortho-
nitro group, although the difference is not large. Similar values
have been reported7 for equilibrium constants for the reactions
of methoxide in methanol at unsubstituted ring positions in 2
and in 1, R = Me.
᎐
(᎐ K kЈAm) is not directly measurable in the reaction of 4 with
᎐
2
pyrrolidine, the value is likely to be higher than that for the
corresponding reaction with piperidine, and we may estimate a
value of ca 104 dm6 molϪ1 sϪ1. This value is considerably higher
than the value of 3.3 dm6 molϪ2 sϪ1 observed (Table 5) for reac-
tion of 5 with pyrrolidine, indicating a different rate-limiting
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1008

View Article Online
Table 6 Kinetic results for reaction of 4 with pyrrolidine in DMF at 25 ЊC
[Pyrrolidine]/mol dmϪ3
[Pyrrolidinium perchlorate]/mol dmϪ3
kfast a/sϪ1
kslow b/sϪ1
kcalc c/sϪ1
0.005
0.010
0.002
0.005
0.010
0.020
0.030
0.040
0.050
0.050
—
—
450
840
—
—
—
—
—
—
—
—
—
—
—
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.005
0.18
0.49
0.86
1.3
1.3
1.2
1.1
2.5
0.19
0.48
0.86
1.3
1.3
1.2
1.1
2.8
—
a Colour forming at 490 nm. b Measured at 360 nm. c Calculated from eqn. (7) with k2 98 5 dm3 molϪ1 sϪ1 and Kc,6 1.4 0.2 dm3 molϪ1
.
step. In the case of 5 general base catalysis is attributed to acid-
catalysis of ethoxide departure from 9, the k4 step in Scheme 1.
Hence the values of K2kB in Table 5 correspond to Kc,2k4,
where Kc,2 is the equilibrium constant for formation of the
anionic intermediates, 9. The lower value of Kc,2k4 obtained for
pyrrolidine compared with n-butylamine, 3.3 vs 1400 dm6 molϪ1
s
Ϫ1, probably reflects lower values for both Kc,2 and k4 resulting
from the greater steric requirements of the secondary amine.
Scheme 2
Interestingly reaction of 5 with piperidine does not yield 10c
the product of nucleophilic substitution. It is known both from
studies of the reaction of 1, R = alkyl3,12,14,24 and also 1-alkoxy-
2,4-dinitronaphthalenes25,26 that when reaction occurs by the
SB-GA mechanism substitutions involving piperidine are
several orders of magnitude slower than the corresponding
reactions of pyrrolidine. This has been attributed to the diffi-
culty in obtaining the required anti-periplanar arrangement of
entering and leaving groups in the transition state for acid cata-
lysed leaving group expulsion.25,26 The X-ray structure, Fig. 1b,
of 2-piperidino-3,5-dinitropyridine shows that unfavourable
steric interactions between the 2- and 3-substituents remain
in the substitution product also contributing to the slowness of
its formation. The consequence is that 9c, decomposes by an
alternative pathway. We have not investigated this, but it is likely
to involve cleavage of a ring carbon-nitrogen bond. There is
precedent for this possibility from studies27 of the hydrolysis of
2-chloro-3,5-dinitropyridine which show that reaction occurs
by an SN(ANRORC) ring-opening mechanism.
rate determining step they do not affect the kinetics of the
substitution reactions of 4 or 5
Reactions in N,N-dimethylformamide (DMF)
It is known that σ-adduct formation may occur in DMF30
although there have been few quantitative studies. To allow
comparison of solvent effects we examined the reactions of
4 and 5 with one amine, pyrrolidine, in DMF. The results
are qualitatively similar to those obtained in DMSO, and are
similarly interpreted. Thus with both 4 and 5 there was a fast
reaction resulting in increased absorbance at 490–495 nm
attributed to the formation of adducts 7b; this was followed
by a slower reaction giving increased absorbance at 360 nm
corresponding to substitution to give 10b.
Data for reaction of 4 are in Table 6. Values of kfast are only
just within the limits of measurement but indicate that nucleo-
philic attack is rate limiting so that eqn. (2) applies with k6
ca. 9 × 104 dm3 molϪ1 Ϫ1. Values for kslow accord nicely with
s
eqn. (7) with k2 98 5 dm3 molϪ1 sϪ1 and Kc,6 1.4 0.2 dm3
molϪ1 This indicates that nucleophilic attack is rate limiting in
the displacement of phenoxide.
Reactions of 2-amino-3,5-dinitropyridines with amines
In the presence of amine the reaction products 10 were found to
be in rapid equilibrium with amine-adducts. In the case of 10a,
reaction with n-butylamine resulted in a shift in wavelength
from 345 nm to 470 nm. Absorbance values are given in Table
13 (supplementary information†). By analogy with previous
work 28,29 on 1-butylamino-2,4,6-trisubstituted benzenes, the
reaction of 10a, is expected to yield the anion 14 formed by
proton loss at low amine concentrations, and the σ-adduct 15
at higher amine concentrations, as shown in Scheme 2. The
results in Table 13 show that at the concentrations used there is
a squared dependence on the amine concentration indicating
that formation of the σ-adduct is dominant and yielding a
Results for reaction of 5 (Table 14, supplementary infor-
mation†) show that formation of the adduct 7b is too rapid
for measurement. Absorbance values at completion of this
reaction yield a value for Kc,6 of 0.15 0.02 dm3 molϪ1. Values
of kslow give a precise fit with eqn. (8) with Kc,6 0.15 0.02 dm3
molϪ1 and with K2kB 1.15
0.05 dm6 molϪ2 Ϫ1. The inter-
s
pretation, as when the solvent is DMSO, is that the SB-GA
mechanism operates here so that K2kB corresponds to Kc,2k4.
Reactions in acetonitrile
value for K6,But of 0.015 dm3 molϪ1
.
For comparison purposes data were obtained for reactions of 4
and 5 with pyrrolidine in acetonitrile. It is known14 that the
formation of anionic σ-adducts is much less thermodynamic-
ally favourable in acetonitrile than in DMSO. Hence reaction of
Spectroscopic changes were also observed for 10b and 10c
in the presence of high amine concentrations and are attributed
to σ-adduct formation. Since these processes occur after the
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1009

View Article Online
Table 7 Kinetic and equilibrium results for the reaction of 5 with pyrrolidine in acetonitrile at 25 ЊC
[Pyrrolidine]/
mol dmϪ3
[Pyrrolidine perchlorate]/
mol dmϪ3
Absa (495 nm)
Kc,6 b/10Ϫ4 dm3 molϪ1
Koc,6/10Ϫ4 dm3 molϪ1
kslow d/10Ϫ3 sϪ1
kcalc e/10Ϫ3 sϪ1
0.05
0.075
0.1
0.15
0.2
0.3
0.3
0.001
0.001
0.001
0.001
0.001
0.001
0
—
—
0.02
—
0.10
0.30
1.2
—
—
17
—
23
37
—
0.44
1.0
1.9
4.4
7.8
15
0.50
1.1
2.0
4.4
7.4
14
5c
—
4c
5c
—
—
a At completion of rapid reaction. b Calculated as Kc,6
=
.
c It is known that in acetonitrile pyrrolidinium ions are stabilised by
homoconjugation with free pyrrolidine, Kh = 23 dm3 molϪ1 (see ref. 14). Ko is the value of Kc,6 extrapolated to zero amine concentration, where
c,6
pyrrolidinium ions will be unassociated. Kc,6 = Koc,6 (1 ϩ Kh [pyrrolidine]). d Measured at 360 nm. e Calculated from eqn. (8) with K2kB 0.20 dm6 molϪ2
s
Ϫ1 and Kc,6 = 4 × 10Ϫ4 (1 ϩ 23 [pyrrolidine]).
Table 8 Comparison of data for reactions of 4 and 5 with pyrrolidine in DMSO, DMF and acetonitrile at 25 ЊC
DMSO
DMF
Acetonitrile
Substrate 4
Kc,6/dm3 molϪ1
Reaction at 6-position
k2/dm3 molϪ1 sϪ1
kЈAm/kϪ2 dm3 molϪ1
40
1.4
<10Ϫ2
Not observed
85
Rate limiting proton transfer
146
≥1000
Rate limiting nucleophilic attack
98
≥5000
1300
Substrate 5
Kc,6/dm3 molϪ1
2.4
3.3
0.15
1.15
5 × 10Ϫ4
0.20
Kc,2.k4/dm6 molϪ2 sϪ1
4 with low concentrations, ≤5 × 10Ϫ3 mol dmϪ3, of pyrrolidine
resulted in formation of the substitution product 10b without
the formation of transient species, such as 7, at observable
concentrations. Thedata, Table15supplementaryinformation,†
are fitted by eqn. (6) with Kc,6 < 10Ϫ2 dm3 molϪ1, K2kB 1.1 ×
105 dm6 molϪ2 sϪ1 and kB/kϪ2 1300 dm3 molϪ1. Combination
of these values allows the calculation of a value for k2 of 85 dm3
constants for their transfer. This will be expected to reduce
ϩ ϩ
values for kAm, kAmH , kЈAm, kЈAmH and k4 in DMSO relative to
acetonitrile. Although our measurements do not lead to the
direct evaluation of rate constants for proton transfer in DMF,
the results suggest that values are higher than in DMSO. Thus
in the formation of 7b from 4 and pyrrolidine proton transfer,
kAm, is rate limiting in DMSO while nucleophilic attack, k6, is
rate limiting in DMF. Also in the corresponding reaction of
5 proton transfer is again rate limiting in DMSO, while the
reaction is too fast to measure in DMF.
molϪ1 sϪ1
.
In the case of 5, the substitution reaction was considerably
slower and at the higher concentrations of pyrrolidine used a
transient absorption band was observed at 495 nm attributed to
In each solvent the results for the displacement of phenoxide
from 4 indicate that the rate limiting step is either nucleophilic
attack, k2, or proton transfer, kЈAm, from the zwitterion 8 to
base. The values of k2 in Table 8 show small variations from
DMSO to acetonitrile compatible with an early transition state
involving little charge development. In contrast the substitution
reaction of 5 involves rate limiting general-acid-catalysed
expulsion of ethoxide from 9b. Values of the parameter Kc,2k4
again show relatively small variation with solvent. However this
is attributed to large decreases in the value of the equilibrium
constant Kc,2 on going from DMSO to acetonitrile which are
compensated for by large increases in value of k4. In the case of
DMF, if it may be assumed that the solvent effect on Kc,2 is
similar to that on Kc,6 so that Kc,2(DMF) : Kc,2(DMSO) = 2.4 :
0.15, then the ratio of k4(DMF) : k4(DMSO) may be calculated
as ca. 6. This again indicates that proton transfers are likely to
be somewhat faster in DMF than in DMSO although the effect
is not large. Overall the results suggest that values of rate and
equilibrium constants for σ-adduct formation and for nucleo-
philic substitution reactions in DMF are likely to be closer to
those in DMSO than those in acetonitrile.
7, R=Et. Absorbance values, in Table 7, lead to a value for Ko
c,6
of (4 1) × 10Ϫ4 dm3 molϪ1. Values of kslow are accommodated
by eqn. (8) with K2kB 0.2 0.02 dm6 molϪ2 sϪ1 and Koc,6 4 × 10Ϫ4
dm3 molϪ1. This value of K2kB is nearly 106 times lower than the
value obtained for reaction of 4 indicative of a change in the
nature of the rate determining step. With 4 the rate limiting
proton transfer is from the zwitterionic intermediate 8 to base
so that K2kB corresponds to K2kЈAm. With the ethoxy derivative,
5, the SB-GA mechanism applies so that K2kB corresponds to
Koc,2k4.
Comparison of solvent effects
Data for reactions of 4 and 5 with pyrrolidine are compared in
Table 8. Values of Kc,6 fall in the order DMSO > DMF ӷ
acetonitrile. The major factor is likely to be the greater ability
of DMSO than of acetonitrile14,31,32 to solvate the charged spe-
cies which are the products of reaction at the 6-position. Values
in DMF are closer to those in DMSO than in acetonitrile.
Previous work14 has shown that rate constants for proton
transfers between nitrogen atoms may be considerably lower, by
a factor of 104, in DMSO than in acetonitrile. This has been
attributed to the excellent hydrogen-bond acceptor properties
of DMSO33,34 so that acidic ammonium protons, in –NHR1R2ϩ
groups, will be H-bonded to the solvent thus reducing rate
Experimental
Compound 4 was prepared by reaction of 2-chloro-3,5-di-
nitropyridine with one equivalent of base in the presence of 1.3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1010

View Article Online
Table 9 Summary of data collection, structure solution and refinement details
10c
5
Empirical formula
M
Crystal size/mm
Crystal system
a/Å
b/Å
c/Å
α/Њ
C10 H12 N4 O4
252.24
C7 H7 N3 O5
213.16
0.38 × 0.34 × 0.08 mm3
Triclinic
0.52 × 0.24 × 0.08
Monoclinic
6.4904(4)
16.7850(11)
8.2216(5)
90.0
5.9847(12)
7.3833(15)
12.375(3)
94.29(3)
β/Њ
95.28(3)
91.80(3)
102.454(1)
90.0
874.60(9)
P21/n
γ/Њ
V/ Å3
542.5(2)
¯
Space group
P1
Z
2
264
1.544
0.12
4
440
1.619
0.14
F(000)
Dcalc/g cmϪ3
µ/mmϪ1
T /K
ω-scan; max 2θ/Њ
120
55
120
58
Unique reflections
Data/restraints/parameters
R
wR2
2484
2484 / 0 / 211
0.033
0.094
2316
2316 / 0 / 164
0.038
0.096
equiv of phenol in aqueous ethanol, mp 158 ЊC (lit.,35 159 ЊC).
Found, % C 50.1; H 2.7; N 16.1, Calcd, % C 50.5; H 2.7; N 15.8.
5 was prepared by reaction of 2-chloro-3,5-dinitropyridine with
one equivalent of sodium ethoxide in ethanol, mp 64 ЊC (lit.,36
69 ЊC). Found, % C 39.2; H 3.3; N 19.5. Calcd, % C 39.4; H 3.3;
N 19.7. Reaction products 10a,b,c were prepared as before9 by
reaction of the 2-chloro compound with a 4 fold excess of
the appropriate amine in ethanol. Recrystallisation was from
ethanol. Spectroscopic data are in Table 1.
7 A. P. Chatrousse, F. Terrier and R. Schaal, J. Chem. Res. (S), 1977,
228.
8 T. A. Emokpae, P. U. Uwakwe and J. Hirst, J. Chem. Soc., Perkin
Trans. 2, 1993, 125.
9 T. A. Emokpae, P. U. Uwakwe and J. Hirst, J. Chem. Soc., Perkin
Trans. 2, 1991, 509.
10 T. A. Emokpae, J. Hirst and P. U. Uwakwe, J. Chem. Soc., Perkin
Trans. 2, 1990, 2191.
11 J. A. Orvik and J. F. Bunnett, J. Am. Chem. Soc., 1970, 92, 2417.
12 M. R. Crampton and P. J. Routledge, J. Chem. Soc., Perkin Trans. 2,
1984, 573.
1H NMR spectra were measured with Varian Mercury 200
MHz or Varian Unity 300 MHz instruments. UV/visible spectra
and kinetic measurements were made at 25 ЊC with an Applied
Photophysics SX-17 MV stopped-flow instrument, or with
Perkin-Elmer Lambda 2, or Shimadzu UV PC spectrometers.
Rate constants were measured under first order conditions with
substrate concentrations of 5–10 × 10Ϫ5 mol dmϪ3, and were
evaluated by standard methods. Values are precise to 5%.
The X-ray diffraction experiments for 10c and 5 were carried
out on a SMART 3-circle diffractometer with 1K and 6K CCD
area detectors respectively (graphite monochromated Mo–Kα
radiation; hemispheres of reciprocal space were covered by
a combination of 4 sets of ω scans, each set at different φ and
2θ angles). Crystals were cooled using a Cyrostream (Oxford
Cyrosystems) open-flow N2 gas cyrostat. For 10c, absorption
correction was performed by ψ scans. The structures were
solved by direct methods and refined by full-matrix least
squares against F 2 of all data, using SHELXTL software. Data
are in Table 9.
13 R. Chamberlin and M. R. Crampton, J. Chem. Soc., Perkin Trans. 2,
1995, 1831.
14 M. R. Crampton and S. D. Lord, J. Chem. Soc., Perkin Trans. 2,
1997, 369.
15 M. R. Crampton and B. Gibson, J. Chem. Soc., Perkin Trans. 2,
1981, 533.
16 G. Punte, B. E. Rivero, S. Cerdeira and N. S. Nudelman, Can. J.
Chem., 1990, 68, 298.
17 M. R. Crampton, L. M. Pearce and L. C. Rabbitt, J. Chem. Soc.,
Perkin Trans. 2, 2002, 257.
18 G. Baldini, G. Doddi, G. Illuminati and F. Stegel, J. Org. Chem.,
1976, 41, 2153.
19 R. Chamberlin, M. R. Crampton and R. L. Knight, J. Chem. Res.
(S), 1993, 444.
20 C. M. Gramaccioli, R. Destro and M. Simonetta, Acta Crystallogr.,
Sect. B, 1968, 24, 129.
21 C. F. Bernasconi, M. C. Muller and P. Schmid, J. Org. Chem., 1979,
44, 3189.
22 C. F. Bernasconi, MTP Int. Rev. Sci.: Org. Chem. Ser. One,
Butterworths, London, 1973, vol. 3, 33.
23 C. F. Bernasconi and M. C. Muller, J. Am. Chem. Soc., 1978, 100,
5530.
24 R. Chamberlin, M. R. Crampton and I. Robotham, J. Chem. Res.
(S), 1994, 408; (M) 1994, 2232.
25 J. F. Bunnett, S. Sekiguchi and L. A. Smith, J. Am. Chem. Soc., 1981,
103, 4865.
26 H. Fujinuma, M. Hosokawa, T. Suzuki, M. Sato and S. Sekiguchi,
Bull. Chem. Soc. Jpn., 1989, 62, 1969.
Acknowledgements
We thank ICSC-World Laboratory, Lausanne for a scholarship
to enable C. I. to visit Durham, U.K.
27 H. A. Al-Lohedan and A. J. Kirby, J. Chem. Soc., Perkin Trans. 2,
1995, 1283.
28 R. Chamberlin and M. R. Crampton, J. Chem. Res. (S), 1993, 106.
29 Y. Hasegawa, J. Chem. Soc., Perkin Trans. 2, 1998, 1561.
30 M. R. Crampton and V. Gold, J. Chem. Soc., 1964, 4293.
31 J. F. Coetzee and C. D. Ritchie, Solute Solvent Interactions, Marcel
Dekker, New York, 1969.
References
1 F. Terrier, Chem. Rev., 1982, 82, 77.
2 E. Buncel, M. R. Crampton, M. J. Strauss and F. Terrier, Electron
Deficient Aromatic- and Heteroaromatic-base Interactions, Elsevier,
Amsterdam, 1984.
3 F. Terrier, Nucleophilic Aromatic Displacement, VCH, New York,
1991.
4 C. F. Bernasconi, J. Am. Chem. Soc., 1970, 92, 4682.
5 C. Abbolito, C. Iavarone, G. Illuminati, F. Stegel and A. Vazzoler,
J. Am. Chem. Soc., 1969, 91, 6746.
6 M. E. C. Biffin, J. Miller, A. G. Moritz and D. B. Paul, Aust.
J. Chem., 1970, 23, 957.
32 J. F. Coetzee, Prog. Phys. Org. Chem., 1967, 4, 45.
33 M. J. Kamlet and R. W. Taft, J. Am. Chem. Soc., 1976, 98, 377.
34 L. Joris, J. Mitsky and R. W. Taft, J. Am. Chem. Soc., 1972, 94,
3438.
35 Z. Talik and E. Plazek, Bull. Acad. Pol. Sci. Ser. Sci. Chim., 1960, 8,
219.
36 J. Barycki and E. Plazek, Roczniki Chem., 1963, 37, 1443.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 0 0 4 – 1 0 1 1
1011