Rationalization of the conflicting effects of hydrogen bond donor solvent on nucleophilic aromatic substitution reactions in non-polar aprotic solvent: Reactions of phenyl 2,4,6-trinitrophenyl ether with primary and secondary amines in benzene-methanol mixtures

Article abstract of DOI:10.1016/j.tet.2004.03.079

The kinetics of the reactions of phenyl 2,4,6-trinitrophenyl ether with piperidine and cyclohexylamine respectively were studied at different amine concentrations in benzene. The reaction of cyclohexylamine was not base-catalysed while that of piperidine was catalysed by one molecule of the nucleophilic amine. Addition of small amounts of hydrogen-bond donor solvent, methanol to the benzene medium of the reactions produced different effects - rate diminution followed by rate increase in one and continuous rate diminution in the other. These effects are compared with that of aniline (previously studied) in which a continuous rate increase was observed. The results are rationalized in terms of the effect of amine-solvent interaction on the nucleophilicity of the amines in addition to some other factors operating through cyclic transition states leading to products. It is evident from the rationalization that the idea of 'dimer nucleophile' in nucleophilic aromatic substitution reactions is erroneous.

Full text of DOI:10.1016/j.tet.2004.03.079

Tetrahedron 60 (2004) 4645–4654
Rationalization of the conflicting effects of hydrogen bond donor
solvent on nucleophilic aromatic substitution reactions in
non-polar aprotic solvent: reactions of phenyl 2,4,6-trinitrophenyl
ether with primary and secondary amines in
benzene–methanol mixtures
*
Olayinka Banjoko and Ibitola A. Babatunde
Department of Chemistry, University of Lagos, Akoka-Yaba, Lagos, Nigeria
Received 10 October 2003; revised 22 March 2004; accepted 25 March 2004
Abstract—The kinetics of the reactions of phenyl 2,4,6-trinitrophenyl ether with piperidine and cyclohexylamine respectively were studied
at different amine concentrations in benzene. The reaction of cyclohexylamine was not base-catalysed while that of piperidine was catalysed
by one molecule of the nucleophilic amine. Addition of small amounts of hydrogen-bond donor solvent, methanol to the benzene medium of
the reactions produced different effects—rate diminution followed by rate increase in one and continuous rate diminution in the other. These
effects are compared with that of aniline (previously studied) in which a continuous rate increase was observed. The results are rationalized in
terms of the effect of amine-solvent interaction on the nucleophilicity of the amines in addition to some other factors operating through cyclic
transition states leading to products. It is evident from the rationalization that the idea of ‘dimer nucleophile’ in nucleophilic aromatic
substitution reactions is erroneous.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
scheme and equation have, however, been made depending
on whether the reaction is taking place in protic, dipolar
aprotic3,4 or non-polar aprotic5–8 solvents or whether the
reaction is catalysed by one or two amine molecules, or an
entirely different catalytic entity.
The general mechanism for aromatic bimolecular nucleo-
philic substitution reactions in all solvents when either
primary or secondary amines are the nucleophiles1,2 is given
in Scheme 1.
In a previous paper,9 we subjected to test the claim by
Nuldelman and Palleros10 that the vulnerability of amine–
amine hydrogen bonding (dimer) is the factor responsible
for the diminishing-rate effect observed on the addition of
small amounts of methanol to the reaction of cyclo-
hexylamine with 2,6-dintroanisole in benzene. We found
that the addition of small amounts of methanol to the
reaction of another primary amine, aniline with another
substrate, phenyl 2,4,6-trinitrophenyl ether (1) in the same
solvent showed no such behaviour, the rate of reaction
Application of the steady state hypothesis to Scheme 1 gives
Eq. 1,
k1ðk2 þ k3½BŠÞ
kA ¼
ð1Þ
k21 þ k2 þ k3½BŠ
where kA is the observed second-order rate constant and B is
either a second molecule of the nucleophile or an added base
acting as the catalyst. Specific modifications of the above
Scheme 1.
Keywords: Aprotic solvent; Aromatic substitution reaction; Benzene–methanol mixtures.
*
Corresponding author. Tel.: þ234-5454891x1710; fax: þ234-4932669; e-mail address: yinkabanjoko@yahoo.com
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.03.079

4646
O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
increasing progressively in this case with increasing
amounts of methanol. From this, we inferred that the
diminution in rate observed by Nuldelman and Palleros
could not have been due to the effect of the incursion of
methanol on the supposed ‘dimerization of the amine’. For,
if this were so, the effect should have been more pronounced
with the weaker aromatic amine, aniline. We then suggested
an alternative interpretation that could explain their results.
increased almost linearly with the methanol
content.
(b) For the uncatalysed reaction of the same substrate
with cylohexylamine, the rate diminution con-
tinued to 100% methanol.
(ii) Reaction in which addition of small amounts of
methanol to the benzene medium increases the rate
continuously.
In order to have a more general view of the phenomenon in
dispute, we decided to examine the reactions of some other
amines—primary and secondary, under the same conditions
by adding small amounts of methanol to the benzene
solutions of their reactions. We have studied the reactions of
1 with piperidine and cyclohexylamine in benzene.
This was observed in the base-catalysed reaction of the
substrate with aniline in benzene.9
2.1. Effects of methanol additions
The reaction of 1 with piperidine was observed to be first-
order in the substrate and second-order in the amine. The
effects of addition of small amounts of methanol to the
benzene medium of the reaction at constant amine
concentrations are shown in Table 2.
2. Results and discussion
The reactions were studied spectrophotometrically at 29 8C
in the presence of varying excesses of the respective amine
over the substrate to ensure first-order kinetics. The
observed second-order rate constants, kA were calculated
from the first-order rate constants. The reactions in pure
solvent, as well as in benzene–methanol mixtures, pro-
ceeded straightforwardly to give the expected 2,4,6-tri-
nitrophenylamine and phenol with no side-products.
The observed effects of methanol addition on the reactions
in the present investigation and also of the previous ones in
the literature must be satisfactorily accommodated by a
sound reaction mechanism. The effects could be due to the
interaction of methanol with (i) the non-polar aprotic
solvent or (ii) the substrate or (iii) intermediates on the
reaction pathway or (iv) the nucleophile:
The effects of methanol addition to the benzene medium of
the reactions studied are in Table 1.
(i) Addition of methanol to non-polar aprotic solvent
should increase the dielectric constant of the medium
thereby increasing the rate of SNAr reactions due to the
extra stabilization of the intermediate first formed from
the reaction of the nucleophile with the substrate.
(ii) Complex formation between the substrate and added
methanol could be a probable factor. Observations
have shown, however, that a particular substrate in a
non-polar aprotic medium could display rate
The reactions fall into two categories:
(i) Reactions in which addition of small amounts of
methanol to the benzene medium diminishes the rate.
(a) For the base-catalysed reaction of 1 with piperi-
dine, the rate diminution reached a minimum at
about 40–50% methanol after which the rate
Table 1. Effects of methanol additions to the benzene medium of the reactions studied
Type of reaction Base catalysis
Effect of methanol additions
Initial rate decrease reaching a minimum value then followed by rate
increase
Catalysed
Not catalysed
Continuous rate decrease
Continuous rate increase9
Catalysed

O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
4647
Table 2. Second-order rate constants, kA for the reaction of 1 with piperidine in benzene and benzene–methanol mixtures at 29 8C. [Substrate]¼5£1025
mol dm23
103[Pip]/mol
dm23
% MeOH
(v/v)
10kA/dm3
mol21 s21
103[Pip]/mol
dm23
% MeOH
(v/v)
10kA/dm3
mol21 s21
103[Pip]/mol
dm23
% MeOH
(v/v)
10kA/dm3
mol21 s21
1.0
1.0
1.0
1.0
1.0
1.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
0
1
2
3
5
10
0
1
2
3
2.06
1.12
0.83
0.58
0.36
0.21
3.97
2.35
1.45
1.04
0.68
0.41
0.35
0.33
0.32
0.32
2.0
2.0
2.0
2.0
2.0
3.0
3.0
3.0
3.0
3.0
3.0
4.0
4.0
4.0
4.0
4.0
50
60
75
90
100
0
1
2
3
5
10
0
1
2
3
0.34
0.41
0.71
1.39
2.00
5.90
3.60
2.23
1.55
1.08
6.30
7.60
4.63
2.90
2.08
1.34
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
5.0
5.0
5.0
5.0
5.0
5.0
10
15
20
30
40
50
60
75
90
100
0
1
2
3
5
10
0.80
0.72
0.67
0.64
0.60
0.64
0.71
1.05
1.85
2.98
9.60
5.54
3.70
2.60
1.70
1.01
5
10
15
20
30
40
5
diminution in its reaction with one amine and a
continuous rate increase in its reaction with another
amine both in the presence of small amounts of
methanol. For example, adding methanol to the
benzene medium of the reaction of 1 with piperidine
(present study) produced a rate diminution while the
same addition of methanol produced a continuous rate
increase for the reaction of the same substrate with
aniline in our previous study.9
reported in literature11 displayed rate diminution on
addition of methanol to the non-polar aprotic medium.
The methanol–amine aggregate would slow down the
reaction due to its reduced nucleophilicity. It is therefore
proposed that the amine–methanol aggregate reacts with
the substrate in the first step of the two-step SNAr reaction
leading to reduced k1.
2.2. Causes of rate increase and rate diminution
Also the reaction of 2,4-dinitrofluorobenzene with
piperidine by Bernasconi and Zollinger11 in benzene
displayed a rate increase with added methanol while
the reaction of the same substrate produced the
opposite effect in its reaction with cis and trans-1,2-
diaminocyclohexylamine12 on the addition of metha-
nol. It can therefore be reasonably assumed that a
methanol-substrate interaction is not responsible for
the observed diminution in rate in each case.
There are only a few cases of rate increase in SNAr reactions
on addition of methanol to non-polar aprotic solvent. One is
our previous study of the reaction of 1 with aniline in
benzene9 and the other is the reaction of 2,4-dinitrofluoro-
benzene with piperidine by Bernasconi and Zollinger11 also
in benzene.
The common features in these two reactions are:
(iii) Interaction of added methanol with any zwitterionic
intermediate formed as in Scheme 1 would assist the
simultaneous extraction of a proton and expulsion of
the leaving group through hydrogen bonding in the
transition state thus leading to an increase in the rate of
reaction.
(iv) The formation of aggregates via hydrogen-bonding
between amines and hydrogen-bond donor solvent,
methanol has been widely studied.13 – 15 The
methanol molecule acts as a proton donor to the
amine resulting in the formation of an aggregate as
shown in Scheme 2.
(a) Base-catalysis with high catalytic effectiveness. This is
evident in the base-catalysed reaction of 2,4-dinitro-
fluorobenzene with piperidine11 for which the catalytic
effectiveness, k3/k2¼1230 dm3 mol21 and the base-
catalysed reaction of 1 with aniline7 for which
k3/k2¼1414 dm6 mol22
.
(b) Susceptibility to hydrogen bonding. It is noteworthy
that since a continuous rate increase is only observed
for strongly base-catalysed reactions on the addition of
methanol to their non-polar aprotic medium, it implies
that the factor responsible for this observation comes
into play in the second step of the SNAr reaction, that is,
Scheme 2.
The nitrogen atom, having thus used its lone pair of
electrons partially for hydrogen-bond formation becomes
less nucleophilic compared with the free amine. The
amine–methanol aggregate of reduced nucleophilicity can
either react with the substrate in the first step of the SNAr
reaction or be the catalysing entity in the decomposition of
the zwitterionic intermediate complex in the second step.
The first assumption is obviously more likely, since an
uncatalysed reaction in the present study as well as the one
Figure 1.

4648
O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
the decomposition of the zwitterionic intermediate
which sometimes involves catalysis.
reduced nucleophilicity of the resulting nucleophile
which is now the amine–methanol aggregate thus
causing diminution in rate and
(ii) the second step of the reaction which is the catalytic
step is influenced by amine and methanol molecules
respectively catalysing through hydrogen-bonding,
thus causing an increase in rate.
A close scrutiny of the zwitterionic intermediates involved
in the above two reactions reveals a high probability of
strong hydrogen bonding in the cyclic transition state
involving methanol and either a strong conjugate acid of a
weak base (aniline) in the reaction of aniline with 1 (Fig. 1)
on the one hand or methanol and a strong hydrogen-bond
forming atom (fluorine) as a leaving group in the reaction of
2,4-dinitrofluorobenzene with piperidine (Fig. 2) on the
other. Electrophilic catalysis of fluoride ion departure
through hydrogen bonding has been demonstrated by Pietra
and Fava16 in the reaction of 1-fluoro-2,4-dinitrobenzene
with piperidine in benzene where they observed catalysis by
methanol but no catalysis by added triethylamine.
The observed overall rate would depend on which of the two
effects—(i) and (ii) predominates. For reactions in which
(i) predominates over (ii), Class A, the observed overall
effect is rate diminution and for reactions in which
(ii) predominates over (i), Class B, the observed overall
effect is rate increase on addition of small amounts of
methanol to the aprotic medium.
Thus, the reaction of 1 with piperidine in benzene in the
present study and the reaction of 2,6-dinitroanisole with
cyclohexylamine in benzene studied by Nudelman and
Palleros10 belong to Class A, while the reaction of 1 with
aniline in our previous study9 belongs to Class B.
It is now clear, therefore, that the explanation given above
for Class A reactions is the reason for the rate diminution
observed by Nudelman and Palleros in the reaction of
2,6-dinitroanisole with cyclohexylamine in benzene–
methanol medium and not our previously suggested
explanation of possible reversibility of the reaction.9,17
Figure 2.
From the above observations, it is clear that the addition of
small amounts of methanol to the non-polar aprotic medium
of SNAr reactions would produce two effects—(a) decrease
in rate due to the reduced necleophilicity of the amine in the
formed amine–methanol aggregate and (b) methanol
catalysis through hydrogen-bonding.
2.3. Mechanism of the base-catalysed reaction
SNAr reactions in non-polar aprotic solvents on addition of
small amounts of methanol can generally be assumed to
involve the attack of the amine–methanol aggregate as well
as the free amine on the substrate to produce the zwitterionic
intermediate. Since amine–methanol aggregate formation
via hydrogen bonding is likely to be a very rapid
equilibrium process, three possible routes for conversion
of the zwitterionic intermediate to products are proposed.
For SNAr reactions in which the first step is rate determining
(Scheme 1), k2þk3[B]qk21 in Eq. 1, giving kA¼k1 and
thus, only effect (a) will be observed on addition of
methanol because of the reduced nucleophilicity of the
attacking amine–methanol aggregate which will be
reflected in the reduced value of k1.
In Scheme 3, S stands for the substrate, B for the
nucleophilic base, B· · ·HOMe for the amine–methanol
aggregate, SB for the zwitterionic intermediate, kB3 and
kM3 eOH for the catalytic rate constants for the conversion of
the intermediate into products by the base and methanol,
respectively.
For reactions in which the second step is rate determining,
k21qk2þk3[B], thus Eq. 1 becomes Eq. 2
k1k2
k1k3½BŠ
kA ¼
þ
ð2Þ
k21
k21
Since the amine may exist in free or hydrogen-bonded forms
as given by Eq. 3,
and the above two opposing effects (a) and (b) will be in
operation because
K
B þ MeOH O B· · ·HOMe
the stoichiometric measured concentration, BStoich will be
ð3Þ
(i) the first step leading to the formation of the
zwitterionic intermediate will be influenced by the
Scheme 3.

O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
4649
k1k3B½BŠ
related to the free base [B]Free by Eq. 4.
½B· · ·HOMeŠ þ ½BŠ ¼ ½BŠ
k00 ¼
ð12Þ
2
k21ð1 þ K½MeOHŠ Þ
ð4Þ
Free
Stoich
For the reaction of 1 with piperidine, which is catalysed by
one amine molecule in benzene, as well as in benzene–
methanol mixtures at low methanol concentration, Eq. 9
applies.
From Eqs. 3 and 4, the unmeasurable quantities
[B· · ·HOMe] and [B]Free are derived in terms of the
measurable quantity [B]Stoich
.
Application of the steady-state hypothesis to Scheme 3 in
terms of the stoichiometric base concentration leads to Eq. 5
The plot of kA against [amine] for the reaction gave straight
lines giving credence to Eq. 10. The values of the intercepts
!
ꢀ
ꢁ
k01K½MeOHŠ
k1
k3B½BŠ
þ
k2 þ
k3B½BŠ
þ k3MeOH½MeOHŠ
1 þ K½MeOHŠ
1 þ K½MeOHŠ
1 þ K½MeOHŠ
kA ¼
ð5Þ
k21 þ k021½MeOHŠ þ k2 þ
þ k3MeOH½MeOHŠ
1 þ K½MeOHŠ
where [B] is the total (stoichiometric) base concentration
and K is the association constant for amine–methanol
aggregate formation.
and slopes are listed in Table 3.
The table shows that the intercepts obtained on addition of
methanol are all within the proximity of the origin and are,
within experimental error, equal to zero. In this reaction, the
overall effect of methanol addition is rate diminution. This
is because the conjugate acid of the strong base, piperidine
(pKa¼11.06) is not sufficiently acidic to promote strong
hydrogen-bonding with methanol in the cyclic transition
state involving the zwitterionic intermediate. The resulting
small value of methanol catalytic rate constant, kM3 eOH for
For the base-catalysed reaction, when the second step is rate
determining, Eq. 6 holds and Eq. 5 becomes Eq. 7.
k3B½BŠ
k21 þ k021½MeOHŠ q k2 þ
þ k3MeOH½MeOHŠ
ð6Þ
1 þ K½MeOHŠ
!
ꢀ
ꢁ
k01K½MeOHŠ þ k1
1 þ K½MeOHŠ
k3B½BŠ
k2 þ
þ k3MeOH½MeOHŠ
1 þ K½MeOHŠ
ð7Þ
kA ¼
k21 þ k021½MeOHŠ
On the assumption that hydrogen bonding with methanol
substantially reduces the nucleophilicity of the amine,
k1qk10 , and Eq. 7 then reduces to Eq. 8 which can also be
expressed in the form of Eq. 9.
the decomposition of the zwitterionic intermediate into
products is thus not comparable with the rate diminishing
effect of the reduced nucleophilicity of the amine–methanol
aggregate.
k1
It is important to note that addition of small amounts of
methanol to the reactions of 1 with piperidine (present
study) and with aniline (previous study9) is expected to
produce similar effects of rate diminution in both reactions
due to the reduced nucleophilicity of the amine–methanol
aggregate. However, the former reaction shows the
expected rate decrease while on the contrary the latter
reaction shows a continuous rate increase.9 This is due to the
strong methanol catalysis kM3 eOH which comes into play
(through a cyclic transition state—Figure 1) in the latter.
kA ¼
1 þ K½MeOHŠ
!
k3B½BŠ
ꢀ k2 þ
þ k3MeOH½MeOHŠ
k1k3B½BŠ
ð8Þ
1 þ K½MeOHŠ
k1k2
kA ¼
þ
2
k21ð1 þ K½MeOHŠÞ
k1k3MeOH½MeOHŠ
k21ð1 þ K½MeOHŠ Þ
þ
ð9Þ
k21ð1 þ K½MeOHŠÞ
Table 3. Values of intercepts and slopes of the plots of kA against
[piperidine] at constant methanol concentrations for the reaction of 1 with
piperidine in benzene–methanol at 29 8C
At constant methanol concentration, this equation reduces to
Eq. 10
% Methanol
102k0/dm3 mol21 s21
k00/dm6 mol21 s21
r
kA ¼ k0 þ k00½BŠ
ð10Þ
0
1
2
3
5
10
2.13^0.78
1.12^1.44
0.65^0.65
0.46^0.25
0.30^0.35
0.15^0.15
187.10^2.35
111.434^4.34
71.90^1.66
50.80^0.76
33.40^1.05
19.90^0.40
0.9998
0.9982
0.9990
0.9997
0.9980
0.9990
where k0 and k00 are defined by Eqs. 11 and 12, respectively.
k1k2
k1k3MeOH½MeOHŠ
k0 ¼
þ
ð11Þ
k21ð1 þ K½MeOHŠÞ
k21ð1 þ K½MeOHŠÞ

4650
O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
This is more than enough to compensate for the reduced
nucleophilicity of the amine–methanol aggregate because
the conjugate acid of the weak base, aniline (pKa¼4.61) is
sufficiently acidic to promote strong hydrogen-bonding with
methanol in the cyclic transition state involving the
zwitterionic intermediate, hence the rate increase.
temperature variation ranging from 15 to 35 8C. It was
observed that the rate of the reaction decreased slightly with
increasing temperature (Table 4), thus resulting in a small
negative activation energy of 21.20 kJ mol21
.
Similar negative temperature effects have been observed by
E. F. Caldin et al.18 in the reaction of 2,4-dinitrophenol with
tri-n-octylamine in chlorobenzene and by Banjoko and
Ezeani7 in the reaction of 1 with some substituted anilines in
benzene.
The aniline reaction with 1 in benzene and benzene–
methanol mixtures is catalysed by two amine molecules and
so the corresponding rate Eqs. 8 and 9 involving the free
amine and the amine–methanol aggregate would have [B]2
terms instead of [B].
The observation in each case has been attributed to
hydrogen-bond formation in the transition state as the
strength of hydrogen-bonding is known to decrease with
increasing temperature.
2.4. Effect of temperature
The observations from the study of SNAr reactions in a non
polar aprotic solvent on addition of small amounts of a
hydrogen-bond donor solvent, methanol, to the reaction
medium have brought into focus the difference in the
mechanisms of these reactions in non-polar aprotic and
polar protic solvents. The common feature in some of these
studies is the parabolic curve sometimes obtained in a plot
of the second-order rate constants, kA against % methanol
for some of the base-catalysed reactions (Fig. 3) as observed
by us in the present investigation as well as by Nudelman
and Palleros.10,17
At the right side of the plot, however, a region where the
addition of methanol has considerably increased to make the
medium polar (say 60–75% methanol), increase in
temperature from 15 to 35 8C led to an increase in rate
(Table 4). This thus resulted in a positive activation energy
of 3.20 kJ mol21. This change from negative activation
energy to positive activation energy is clearly indicative of a
change in mechanism of the reaction in the medium. It could
be rightly inferred, therefore, that the reaction represented
by the right hand side of the curve does not involve
hydrogen-bonding in the transition state. Since the medium
is polar in this section, it could now support ionic charges,
hence the Specific Base-General Acid (SB-GA) or proton
transfer mechanism can be assumed to be operating in this
now polar protic medium. Evidence in recent times,
however, is in support of a proton transfer mechanism
over the widely accepted SB-GA mechanism in polar protic
and dipolar aprotic solvents19,20 when the leaving group is
fairly good.
On the left side of these curves, a region in which the
medium is predominantly non-polar aprotic is an effect
symbolic of a mechanism different from that on the right
side which is predominantly polar protic. This is because an
increase in the percentage of methanol results in rate
diminution on the left while it is rate increase on the right
side of the curves. The reaction represented by the left side
of the plot (say 3–40% methanol content) was subjected to
The above observation of a change in mechanism as the
medium changes from a non-polar aprotic to a polar protic is
a strong support for the cyclic transition state mechanism in
non polar aprotic medium. The mechanism thus differs
remarkably in this respect from the hetero/homoconjugate
mechanism,21 which does not differentiate between
mechanisms in the two media.
2.5. Mechanism of the uncatalysed reaction
For a reaction that is not base-catalysed and occurs in the
presence of small amounts of methanol in non-polar aprotic
solvent, Scheme 4 applies.
Figure 3. Plots of second-order rate constants kA against % methanol at
constant piperidine concentrations for the reaction of phenyl 2,4,6-
trinitrophenyl ether with piperidine in benzene–methanol mixtures at
29 8C.
Application of the steady-state hypothesis to Scheme 4,
working in terms of the stoichiometric base concentration,
Table 4. Second-order rate constants for the reaction of 1 with piperidine at constant amine concentration [4£1023 M] in benzene–methanol mixtures at
varying temperatures
3% Methanol
60% Methanol
Temperature (8C)
10kA/dm3 mol21 s21
40% Methanol
15
2.14
20
2.12
25
2.10
35
2.07
Temperature (8C)
10 kA/dm3 mol21 s21
15
0.66
75% Methanol
20
0.68
25
0.70
35
0.73
Temperature (8C)
10kA/dm3 mol21 s21
15
0.57
20
0.59
25
0.61
35
0.63
Temperature (8C)
10 kA/dm3 mol21 s21
15
0.94
20
0.96
25
0.99
35
1.05

O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
4651
Scheme 4.
gives the observed overall second-order rate constant, kA as
ꢀ
ꢁ
k1
k01K½MeOHŠ
k2
þ
1 þ K½MeOHŠ
1 þ K½MeOHŠ
kA ¼
k21 þ k021½MeOHŠ þ k2
Since the reaction is not base-catalysed, the first step is rate
determining and inequality Eq. 13 holds.
k2 q k21 þ k021½MeOHŠ
ð13Þ
ð14Þ
k1
k01K½MeOHŠ
[ kA ¼
þ
1 þ K½MeOHŠ
1 þ K½MeOHŠ
Figure 4. Plot of second-order rate constants, kA against % methanol at
constant cyclohexylamine concentration for the reaction of phenyl 2,4,6-
trinitrophenyl ether with cyclohexylamine in benzene–methanol mixtures
at 29 8C.
The reaction of 1 with cyclohexylamine in benzene is not
base-catalysed and so conforms with Scheme 4 and Eq. 14
derived from it. When no methanol is added to the reaction
medium, Eq. 14 reduces to
It is worth noting that the value of the second-order rate
constant, kA, in 100% methanol (0.92 dm3 mol21 s21) is
much less than that in 100% benzene (12.84 dm3 mol21
s21). This is because the first step of the uncatalysed
reaction involves the amine–methanol aggregate of reduced
nucleophilicity in the former, while in the latter, it involves
the free amine molecule with unreduced nucleophilicity.
kA ¼ k1
ð15Þ
Table 5 shows the constancy of the second-order rate
constant, kA with [cyclohexylamine] when the reaction is
carried out in benzene without methanol, thus giving
credence to Eq. 9. Addition of small amounts of
methanol to the benzene medium of the reaction
showed a sharp decrease in kA [Table 6] from the initial
value of 12.84 dm3 mol21 s21 in pure benzene to
1.88 dm3 mol21 s21 in benzene–10% methanol after
which the decrease became gradual.
2.6. Effect of temperature
As the above observation is an unusual one, we felt it
worthwhile to investigate the effect of temperature on the
reaction. Temperature probe ranging from 15 to 35 8C was
therefore, carried out. The rate of the reaction at low
methanol content (3% methanol) as well as that at high
methanol content (60% methanol) increased appreciably
with increase in temperature (Table 7), thus resulting in
Table 5. Second-order rate constants, kA for the reaction of phenyl 2,4,6-
trinitrophenyl ether with cyclohexylamine (CHA) in benzene at 29 8C
104[CHA]/mol dm23
2.5
12.83
3.0
12.85
3.5
12.83
4.0
12.84
kA/dm3 mol21 s21
positive activation energies of 30.49 and 32.11 kJ mol21
respectively.
,
The experimental data in Table 6 indicate that up to 10%
methanol (2.5 mol dm23), values may be accommodated by
the expression for k0A given by Eq. 8. With k1¼
12.84 dm3 mol21 s21, k1 ca. 1.5 dm3 mol21 s21 and K ca.
7 dm3 mol21. At higher methanol concentrations, however,
values of rate and equilibrium constants will be expected to
be affected by medium effects and so these values will
change.
The facts to be considered in explaining the above
observations are:
(i) diminution in rate on addition of methanol at both low
and high methanol contents,
(ii) increase in rate with increase in temperature also at
both low and high methanol contents.
Unlike in the base catalysed reactions, the diminution in rate
continued at higher methanol content, reaching a minimum
value at 100% methanol (Fig. 4).
Since the reaction is not base catalysed, the first step of the
Table 6. Second-order rate constants for the reaction of 1 with cyclohexylamine (CHA) in benzene–methanol mixtures.at constant cyclohexylamine
concentration (0.02 mol21) at 29 8C [Substrate]¼2.5£1025 mol dm23
% Methanol (v/v)
0
1
3
5
10
20
30
40
60
85
100
kA (dm3 mol21 s21
)
12.84
5.48
3.51
2.84
1.88
1.50
1.32
1.25
1.14
0.99
0.92

4652
O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
Table 7. Effect of Temperature on the reaction of phenyl-2,4,6-trinitrophenyl ether with cyclohexylamine at constant cyclohexylamine concentration
[2.5£1024 M] in benzene–methanol mixtures
3% Methanol
60% Methanol
Temperature (8C)
kA/dm3 21 s21
15
2.04
20
2.51
25
3.02
35
3.76
Temperature (8C)
kA/dm3 mol21 s21
15
0.61
20
0.82
25
1.06
35
1.52
reaction involving the amine–methanol aggregate of
reduced nucleophilicity (as well as that involving the
remaining free amine molecules) is rate-determining. It is
therefore reasonable to assume that any factor that represses
the amine–methanol aggregate formation will lead to an
increase in rate while a factor that promotes it will result
in rate diminution. Hydrogen-bonds are known to be
weakened with an increase in temperature, hence an
increase in rate with an increase in temperature is expected
at both low and high methanol contents as was indeed
observed for the reaction with 3 and 60% methanol content.
On the other hand, an increase in methanol content that
promotes amine–methanol aggregate formation (of reduced
nucleophilicity) should produce rate diminution. The rapid
diminution in rate at low methanol content can thus be
attributed to the rapid involvement of amine molecules
in amine–methanol aggregate formation with the little
available methanol molecules and so the concentration of
these aggregates increases with additional small amounts of
methanol. At reasonably high methanol content, most amine
molecules are in the aggregate forms. Further addition of
methanol will only result in little increase in their
concentration, thus resulting in less rapid diminution in
rate, hence the tailing off of the curve (Fig. 4) which reaches
a minimum at 100% methanol.
observations. It is remarkable that while the first view is
of general applicability, the second one is not. It is difficult,
for example, to comprehend how the dimer mechanism
would explain the first-order dependence of kA on the
concentration of amines in SNAr reactions.
If the dimer, claimed to be the effective nucleophile, reacts
with the substrate in the first step of SNAr reactions, then,
every reaction of the amine should display at least second
order dependence of the rate constant kA on amine
concentration, but this is not always the case. For example,
the reaction of n-butylamine with 2,4,6-trinitrophenyl
ether23 in benzene and that of cyclohexylamine with
2,4-dinitrochlorobenzene,24 also in benzene, are first order
in the amines. The base-catalysed reaction of 2,6-dinitro-
anisole with cyclohexylamine in toluene10which is third-
order in amine was explained on the basis of the dimer
mechanism but the reaction of the same amine in benzene
with 2,4-dinitrochlorobenzene24 and with 1 (present study,
Table 5) are first order in amine. The observation of first-
order dependence of kA on amine concentration in a number
of SNAr reactions23–25 in non-polar aprotic solvents shows
that the ‘dimer nucleophile’ mechanism is not of general
applicability.
The other notable instance in which the dimer mechanism
had been found to be inapplicable is in the base-catalysed
reaction of dinitrofluorobenzene with piperidine in benzene
by Bernasconi and Zollinger11 in which added ethanol
caused an increase in rate. Nudelman and Mentserrat
also carried out the same reaction and made similar
observations12 of rate increase on the addition of ethanol,
instead of the rate decrease they expected on the basis of the
dimer mechanism. To explain this anomaly, the authors
attributed the rate increase to lack of self-association or
dimerization of piperidine in benzene, an assertion that is
inconsistent with their stand on the dimerization of amines
generally. No reason was given for this inconsistency. The
authors, however, asserted that the rate increase with small
additions of a protic solvent was expected on the basis of
hydrogen-bonding assistance to the nucleofuge departure.
The above observation, which could not be explained by the
dimer mechanism is easily explained on the basis of the
cyclic transition state mechanism. This has been extensively
dealt with in the section under which SNAr reactions
exhibiting rate increase on addition of methanol are
classified as Class B while those exhibiting rate decrease
are classified as Class A. In the second step of the reaction in
question, methanol catalysis involving hydrogen-bonding
between methanol and the zwitterionic intermediate in the
cyclic transition state (Fig. 2), more than compensates for
the effect of reduced nucleophilicity of the amine–methanol
aggregate operating in the first step of the reaction, hence
the rate increase.
2.7. Deductions
There is no doubt that the factor responsible for the
diminution in rate on addition of small amounts of methanol
to an SNAr reaction in an aprotic solvent of low relative
permitivity is the reduced nucleophilicity of the resulting
nucleophilic entity. How this comes about, however, is not
without some controversy. While from the explanation
given earlier, we strongly feel that the reduced nucleo-
philicity results from the interaction of the amine with the
added methanol to form an amine–methanol aggregate,
Nudelman and Palleros maintain10,12,17 that it results from
the incursion of the added methanol on the supposed
‘amine–amine dimer’ thus forming the same amine–
methanol aggregate which they termed ‘mixed dimer’. It
is the contention of these authors that amines exist in aprotic
solvents largely as dimers and that it is the dimer that is the
nuclophile in SNAr reactions. We, on the other hand, believe
that in dilute amine solutions as is generally the case in SNAr
reactions, amines exist largely as free molecules and that it
is the free amine molecule that is the nucleophile in SNAr
reactions. Our contention is buttressed by the fact that the
formation constant for aliphatic amine dimers is in the range
0.02–0.1.21,22a,b In concentrated amine solutions, however,
self-aggregation is likely to occur.
The above two differing views will no doubt result in
different interpretations of similar experimental

O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
4653
3. Conclusion
reactions were carried out under conditions of excess of
nucleophile over substrate and, in all cases, excellent first-
order plots were obtained. The second-order rate constant,
kA were obtained by dividing the first-order rate constants
by the amine concentration. All rate determinations were
carried out at least in duplicate and the rate constants are
accurate to within ^2%.
Addition of hydrogen-bond donor (hbd) solvent to SNAr
reactions involving a substrate and an amine in a non-polar
aprotic solvent results in the formation of amine-solvent
aggregates of reduced nucleophilicity. The effect should
normally result in diminution in rate of reaction but could
instead result in an increase in rate if the nature of the
zwitterionic intermediate first formed between the substrate
and the amine is such that could promote strong hydrogen-
bonding between it and the hbd solvent in the cyclic
transition state thus leading to its catalytic decomposition
into products.
Acknowledgements
Grateful acknowledgement is made by O.B. for a grant-in-
aid of research from the Central Research Committee of the
University of Lagos.
The interpretation of this phenomenon in the literature by a
group of co-workers in the field, as being due to the
formation of an ‘amine–amine dimer’ nucleophile is
erroneous. That our contention above, is the correct
position, is further buttressed by the fact that added
hydrogen-bond acceptor (hba) co-solvent, which by impli-
cation (or rationalization), should increase the rate of
reaction did, in fact, increase the rate. This is due to the
increase in the nucleophilicity of the attacking nucleophilic
amine through hydrogen-bonding between the amine and
the co-solvent. The results of the reactions thus form the
basis of our next publication.
References and notes
1. Bunnett, J. F. Quart. Rev. 1958, 12, 1–16.
2. Bernasconi, C. F. M.T.P. International Reviews in Science and
Organic Chemistry, Series One; Butterworths: London, 1973;
Vol. 3. pp 33–61.
3. Bunnett, J. F.; Davies, G. T. J. Am. Chem. Soc. 1960, 82,
665–674.
4. Bernasconi, C. F.; de Rossi, R. H.; Schmid, P. J. Am. Chem.
Soc. 1977, 4090, 4100.
5. Capon, B.; Rees, C. W. Annu. Rep. Prog. Chem. 1964, 60,
279–284.
4. Experimental
4.1. Materials
6. Ayediran, D.; Bamkole, T. O.; Hirst, J.; Onyido, I. J. Chem.
Soc., Perkin Trans. 2 1977, 597–600.
7. Banjoko, O.; Ezeani, C. J. Chem. Soc., Perkin Trans. 2 1982,
1357–1360.
The preparation of phenyl 2,4,6-trinitrophenyl ether (1) and
the purification of benzene and methanol were described
previously.26 Analar piperidine was heated under reflux
with sodium wire for 4 h and then distilled. The process was
repeated twice and the middle fraction distilling at 106 8C
was collected and kept in a desicator, protected from light
(lit.27 bp 105–106 8C). Cyclohexylamine was purified by
the same method bp 132 8C (lit.27 132–133 8C).
8. Nudelman, N. S.; Palleros, D. J. Org. Chem. 1983, 48,
1607–1613.
9. Banjoko, O.; Bayeroju, I. A. J. Chem. Soc., Perkin Trans. 2
1988, 1853–1857.
10. Nudelman, N. S.; Palleros, D. J. Chem. Soc., Perkin Trans. 2
1984, 1277–1280.
11. Bernasconi, B.; Zollinger, H. Helv. Chim. Acta 1966,
103–111.
The product of each reaction studied was prepared by
standard methods previously described.28
12. Nudelman, N. S.; Montsserat, J. M. J. Chem. Soc., Perkin
Trans. 2 1990, 1073–1076.
13. Kollma, P. A.; Allen, L. C. J. Am. Chem. Soc. 1971, 93,
2991–2995.
N-(2,4,6-Trinitrophenyl) piperidine, mp 100 8C, lmax
(C6H6) 390 nm.
14. Bene, J. E. J. Am. Chem. Soc. 1973, 95, 5460–5465.
15. Ratkonics, F.; Salmon, T. Acta Chim. Acad. Hung. 1977, 90,
93–97.
N-(2,4,6-Trinitrophenyl) cyclohexylamine, mp 90–91 8C,
lmax (C6H6) 345 nm.
16. Pietra, F.; Fava, A. Tetrahedron Lett. 1963, 1935–1940.
17. Nudelman, N. S.; Marder, M.; Gurevich, A. J. Chem. Soc.,
Perkin Trans. 2 1993, 229–233.
Kinetic procedure. The rates of formation of the products of
the reactions were determined spectrophotometrically by
the procedure previously described.26 In some cases,
however, the reactions were followed directly in the
thermostated cell of the spectrophotometer. The reactions
were carried out at 29 8C. For reactions in mixed solvents,
the methanol content (v/v) refers to its final volume in the
reaction mixture. Optical densities were recorded at the
absorption maximum wavelength (lmax) of each product. In
all cases the absorption spectrum of the reaction mixture at
‘infinity time’ corresponded within 2% to the ‘mock’
infinity prepared by using the respective N-(2,4,6-trinitro-
phenyl)amine obtained as a product of the reaction. The
18. Caldin, E. F.; Tortschanof, K. J. Chem. Soc., Faraday Trans. 1
1978, 1804–1810.
19. Chamberlin, R. A.; Crampton, M. R. J. Chem. Soc., Perkin
Trans. 2 1995, 1831–1838.
20. Crampton, M. R.; Lord, S. D. J. Chem. Soc., Perkin Trans. 2
1997, 369–376.
21. Hirst, J. J. Phys. Org. Chem. 1994, 7, 68–79.
22. (a) Kern, M.; Servais, B.; Abello, L.; Panneltier, G. Bull. Soc.
Chem. Fr. 1968, 2763–2767. (b) Wolff, H.; Gamer, G. J. Phys.
Chem. 1972, 76, 871–875.
23. Banjoko, O.; Rahman, K. U. J. Chem. Soc., Perkin Trans. 2
1981, 1105–1107.

4654
O. Banjoko, I. A. Babatunde / Tetrahedron 60 (2004) 4645–4654
24. Nudelman, N. S.; Garrido, D. J. Chem. Soc., Perkin Trans. 2
1976, 1256–1260.
27. CRC handbook of chemistry and physics, 53rd ed., 1972–
1973.
25. Pietra, F.; Del Cima, F. Tetrahedron Lett. 1967, 1457–1462.
26. Banjoko, O.; Otiono, P. J. Chem. Soc., Perkin Trans. 2 1981,
399–402.
28. Banjoko, O.; Ezeani, C. J. Chem. Soc., Perkin Trans. 2 1986,
531–536.