Reaction of imidazole with toluene-4-sulfonate salts of substituted phenyl N-methylpyridinium-4-carboxylate esters: Special base catalysis by imidazole

Article abstract of DOI:10.1039/b301936p

The reaction of imidazole in aqueous solution with toluene-4-sulfonate salts of substituted phenyl N-methylpyridinium-4-carboxylate esters obeys the rate law: k_obs - k_background = k₂[Im] + k₃[Im]² where [Im] is the imidazole concentration present as free base. The parameters k₂ and k₃ fit Bronsted type free energy correlations against the pK_a of the leaving phenol with β_Lg values of -0.65 and -0.42 respectively. The imidazolysis is insensitive to catalysis by general bases and yet k₃ for the 3-cyanophenyl ester possesses a deuterium oxide solvent isotope effect of 4.43 consistent with rate limiting proton transfer. A special catalytic function is proposed for decomposition of the tetrahedral addition intermediate (T^±) via k₃ whereby the catalytic imidazole interacts electrophilically with the leaving phenolate ion and removes a proton from the nitrogen in the rate limiting step with subsequent non-rate limiting ArO-C bond fission. This is consistent with the change in effective charge on the leaving oxygen in the transition structure of k₃ which is more positive (-0.42) than that expected (-0.60) for the equilibrium formation of the zwitterion intermediate. The catalytic function at the leaving oxygen is likely to be an electrophilic role of the NH as a hydrogen bond donor. In the k₂ step the deuterium oxide solvent isotope effect of 1.51 for the 3-cyanophenyl ester and the β_Lg of -0.65 are consistent with rate limiting expulsion of the phenolate ion from the T^± intermediate. The absence of general base catalysis of imidazolysis rules out the established mechanism for aminolysis of esters where T^± is stabilised by a standard rate limiting proton transfer. The kinetically equivalent term for k₃ where T^- reacts with the imidazolium ion as an acid catalyst would require this step to be rate limiting and involve proton transfer not consistent with departure of the good aryl oxide leaving group.

Full text of DOI:10.1039/b301936p

View Article Online / Journal Homepage / Table of Contents for this issue
Reaction of imidazole with toluene-4-sulfonate salts of substituted
phenyl N-methylpyridinium-4-carboxylate esters:
special base catalysis by imidazole
Matthew J. Colthurst and Andrew Williams*
University Chemical Laboratory, Canterbury, Kent, UK CT2 7NH. E-mail: aw@ukc.ac.uk
Received 18th February 2003, Accepted 27th March 2003
First published as an Advance Article on the web 16th April 2003
The reaction of imidazole in aqueous solution with toluene-4-sulfonate salts of substituted phenyl
N-methylpyridinium-4-carboxylate esters obeys the rate law:
k
obs Ϫ kbackground = k2[Im] ϩ k3[Im]2
where [Im] is the imidazole concentration present as free base. The parameters k2 and k3 fit Brønsted type free energy
correlations against the pKa of the leaving phenol with βLg values of Ϫ0.65 and Ϫ0.42 respectively. The imidazolysis
is insensitive to catalysis by general bases and yet k3 for the 3-cyanophenyl ester possesses a deuterium oxide solvent
isotope effect of 4.43 consistent with rate limiting proton transfer. A special catalytic function is proposed for
decomposition of the tetrahedral addition intermediate (T ) via k3 whereby the catalytic imidazole interacts
electrophilically with the leaving phenolate ion and removes a proton from the nitrogen in the rate limiting step with
subsequent non-rate limiting ArO–C bond fission. This is consistent with the change in effective charge on the leaving
oxygen in the transition structure of k3 which is more positive (Ϫ0.42) than that expected (Ϫ0.60) for the equilibrium
formation of the zwitterion intermediate. The catalytic function at the leaving oxygen is likely to be an electrophilic
role of the NH as a hydrogen bond donor. In the k2 step the deuterium oxide solvent isotope effect of 1.51 for the
3-cyanophenyl ester and the βLg of Ϫ0.65 are consistent with rate limiting expulsion of the phenolate ion from the
T intermediate. The absence of general base catalysis of imidazolysis rules out the established mechanism for
aminolysis of esters where T is stabilised by a standard rate limiting proton transfer. The kinetically equivalent term
for k3 where TϪ reacts with the imidazolium ion as an acid catalyst would require this step to be rate limiting and
involve proton transfer not consistent with departure of the good aryl oxide leaving group.
non-specific solvent effect as well as making the measurements
Introduction
of k3 less accurate than those of k2.
Rates of hydrolysis of phenyl esters catalysed by imidazole
Termolecular terms are well known in aminolysis reactions at
usually refer to the formation of an acyl-imidazole intermediate
acyl functions bearing relatively poor leaving groups,6–12 where
which hydrolyses to produce acid and imidazole in a subsequent
the catalytic function of the additional base is to trap, by pro-
slower step.1,2 Caplow and Jencks observed that, for the reac-
ton transfer, the zwitterionic tetrahedral adduct from addition
tion of imidazole with 4-nitrophenyl esters of benzoic acids, the
of primary or secondary amines; this circumvents the rapid
plot of the pseudo first order rate constant against imidazole
concentration has a slight upward curvature. This was inter-
return to reactants competing favourably with leaving group
departure (Scheme 1).6 The direct expulsion of the leaving group
preted as an additional term (k3) involving a second imidazole
from T leads to an unstable N-protonated amide as the initial
molecule acting as a general base catalyst (Eqn. (1)).3 The ter-
product; unless the leaving group propensity is exceptional the
molecular term, k3, was observed by other workers4,5 but the
process:
general base catalysed route via TϪ is generally favoured. Only
highly activated esters such as 1-acyloxypyridines do not
require trapping because their leaving groups are good enough
to provide a favourable path direct from T via R2NHϩ–CO–R
to product.
Rate = kbackground[ester] ϩ k2[imidazole][ester] ϩ
k3[imidazole]2 [ester] (1)
In the case of imidazolysis the proton in T does not reside
on the nitrogen adjacent to the nascent carbonyl. Direct expul-
sion of the leaving group from T does not therefore incur the
heavy energy penalty suffered by k in aminolysis because
the product acyl-imidazolium ion has only a weak acidity
could not be studied in detail because the relative magnitude of
k3 and k2 demanded concentrations of imidazole up to 1 M for
the k3 effect to be measured; the high imidazole concentration
introduces the possibility that the curvature could be due to a
Scheme 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
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 9 5 – 2 0 0 0
1995

View Article Online
Table 1 Analytical data for the toluene-4-sulfonate salts of substituted phenyl N-methylpyridinium-4-carboxylate esters
Found
N(%)
Calculated
N(%)
Substituent
Mp/ЊC
Formula
C(%)
H(%)
C(%)
H(%)
4-CH3CO–
Parent
3-Cl–
198–200dec
153–155
3.13
3.50
3.31
6.9
61.46
61.01
56.91
61.2
4.76
4.93
3.96
4.1
C22H21NO6S
C20H19NO5S
C20H18ClNO5S
C21H18N2O5S
3.28
3.63
3.36
6.80
61.81
62.3
57.20
61.45
4.95
4.97
3.60
4.4
201–203
3-Cyano–
200–203dec
compared with that of acyl-ammonium ions (RCO–NHR2ϩ).
The advantage gained by general base trapping of the analo-
gous T from imidazole addition is unexpected on the basis of
the mechanism for aminolysis by primary and secondary
amines.
into 2.5 mL of sample solution in a 1 cm path-length silica
cell in the thermostatted cell-compartment of a UV/Visible
spectrometer. A few vertical strokes of the rod effected mixing
and the reaction was followed by monitoring the change in
absorbance with time at the optimal wavelength, determined by
repetitive scanning of the spectrum of a test solution undergo-
ing reaction. The kinetics of the reactions in D2O solvent were
carried out as above using 1 mL capacity silica cells with 1 cm
path-length.
The pH values of the solutions were measured before and
after each kinetic run using a Radiometer PHM 62 meter fitted
with a Russell CMAWL combined electrode, calibrated with
BDH buffers. Measurements of pD were carried out with the
This study shows that the reaction of imidazole with substi-
tuted phenyl N-methylpyridinium-4-carboxylate esters has a
much more substantial termolecular term relative to the bi-
molecular term, k2, (Scheme 2) than the previously observed
values for benzoate or acetate esters. The possession of a sub-
stantial termolecular effect means that k3 values may be deter-
mined with good accuracy under conditions of relatively low
imidazole concentration where non-specific solvent effects will
be minor. Calculations from previous work show that at 0.1 M
imidazole the percentage of total reaction flux passing through
the termolecular pathway is only 8.06 and 2.86 for the imid-
azolysis of 4-nitrophenyl 4-nitrobenzoate3 and 4-methoxy-
phenyl acetate respectively.5 For this reason, the system of
Scheme 2 is ideal as a vehicle to investigate the origins of the
termolecular term of imidazolysis.
hydrogen electrode using the formula pD = pHmeter
reading
ϩ0.37.14 Data from experiments where the pH changed by more
than 0.1 were discarded. Pseudo first order rate constants were
obtained from the data by fitting the value At to the equation
At = A∞ Ϫ (A∞ Ϫ Ao)exp(Ϫkt).
The formation and decay of the intermediate N-acyl-
imidazole was followed by trapping it with hydroxylamine to
give the hydroxamic acid which was detected as its complex
with Fe() ions using the absorption at 540 nm.15 Solutions of
3.5 M NaOH and 4 M HONH3Cl were prepared and mixed in
the ratio 2 vol. NaOH to 1 vol. HONH3Cl to give a stock
hydroxylamine solution. A solution of FeCl3ؒ6H2O was pre-
pared in 0.7 M HCl to use as the developing agent. A kinetic
run was initiated by mixing 1.25 mL, 51.6 mM ester (50% v/v
CH3CN–H2O) with 1.25 mL, 0.2 M imidazole stock at pH 7.
Aliquots of 0.05 mL were taken at intervals and added to
0.5 mL of the stock hydroxylamine solution. Each sample
was then diluted, after 3 min, with 2 mL FeCl3 solution and the
absorbance read at 540 nm. An experiment to determine
the molar absorption coefficient at 540 nm had water replacing
the imidazole to obtain an absorbance corresponding to a
Scheme 2
known amount of ester; the molar absorption coefficient, ε540
,
Experimental
under the conditions of these experiments is 1002. A further
control experiment under the same conditions measured the
release of the 4-nitrophenolate ion by following its absorbance
at 400 nm.
Materials
The toluene-4-sulfonate salts of substituted phenyl N-methyl-
pyridinium-4-carboxylate esters were either already available or
were prepared according to methods recorded in a previous
investigation.13 New esters were characterised by NMR and
elemental analysis; melting points and analytical data are
recorded in Table 1 and the ester salts with toluene-4-sulfonic
acid were recrystallised from DMF. D2O (99.9% D) and DCl in
D2O (20%, 99.5% D) were obtained from Goss Scientific.
Results
The change in UV/Visible absorbance in the reactions of the
esters in buffers containing imidazole obeyed excellent pseudo
first order kinetics over at least 90% of the total reaction, except
2Ϫ
for the kinetics with the 3-cyanophenyl ester with HPO4 fol-
lowed at 275 nm. The total absorbance changes at the wave-
lengths used for the kinetic measurements were compared with
the expected changes and this indicated complete release of
phenol from the phenyl esters; there is no evidence for substi-
tution in the phenolic aromatic nucleus or in the pyridinium
ring.
The experiments using the hydroxamic acid assay technique
(Fig. 1) show that an activated acyl function is still present after
all the 4-nitrophenolate ion has been expelled from the 4-nitro-
phenyl N-methylpyridinium-4-carboxylate esters in imidazole
buffers. The release of 4-nitrophenolate ion measured at 400 nm
was found to be substantially complete after about 20 s. Fig. 1
illustrates the time dependence of the hydroxamic acid assay on
an identical solution. The decay of the active acyl function is
Kinetics
All solutions were prepared with water that was doubly distilled
from glass and degassed before use. Two solutions of identical
bufferconcentrations(HEPES,4-(2-hydroxyethyl)-1-piperazine-
ethanesulfonic acid at 0.05 M), ionic strength (0.5 M), pH and
solvent composition were prepared, only one of which con-
tained imidazole. The sample solutions were prepared by dilu-
tion of the stock solution to give a series of solutions with
different imidazole concentrations but with the other param-
eters remaining constant.
Reactions were initiated by the addition of an aliquot (0.01
to 0.05 mL) of a stock solution of the ester salt (in 50% v/v
water–acetonitrile) on the tip of a glass or polypropylene rod
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 9 5 – 2 0 0 0
1996

View Article Online
Fig. 3 Values of k3Ј as a function of FB2 for the reaction of
3-cyanophenyl N-methylpyridinium-4-carboxylate in imidazole-
containing buffers. Conditions as in Table 2.
Fig. 1 Assay of active ester in imidazolysis of 4-nitrophenyl ester (᭿);
ꢀ show the absorbance at 400 nm of the release of 4-nitrophenolate
ion.
phenyl ester and it is assumed to hold for the other esters
studied here. Bruice and Benkovic demonstrated this rate law
for the imidazole-catalysed imidazolysis of 4-methylphenyl and
4-methoxyphenyl acetates.5
much slower than the release of the 4-nitrophenolate ion and
this is judged to be consistent with the formation of the N-(NЈ-
methylpyridinium-4-carbonyl)imidazole intermediate followed
by its gradual decay.
The effect of general bases on the reaction of imidazole with
N-methylpyridinium-4-carboxylate esters was tested by the use
of 2,6-lutidine. This base was chosen as it is relatively non-
nucleophilic13 and therefore does not compete with imidazole
for attack at the ester function. Changing the concentration of
2,6-lutidine did not affect the observed rate constant for the
3-cyanophenyl ester (measured at 300 nm) in the presence of
0.1 M free imidazole base (Fig. 4). Assuming that the k3 term
for 2,6-lutidine acting as a general base is the same as that for
imidazole, the calculated rate constant at pH 7.00 for 0.126 M
free 2,6-lutidine (calculated from the pKa = 6.77 and total 2,6-
lutidine of 0.2 M) should be 1.7-fold that without the base. This
calculated change is well outside the error limits of the experi-
ments. A similar insensitivity was obtained in the case of imid-
azolysis of the 4-nitrophenyl ester with 2,6-lutidine as the
potential base catalyst. The rate constants for formation of
acyl-imidazole in the presence of increasing concentrations of
phosphate buffer (ka, see Eqn. (2)) are 0.055, 0.050 and 0.051 sϪ1
for 0.04, 0.1 and 0.2 M respectively in total phosphate at pH
7.00 (at a fraction of free base at 0.381 based on a pKa of 7.21
for phosphate). The argument used for the 2,6-lutidine experi-
ments indicates that 0.2 M in total phosphate should raise the
observed rate constant by 1.4-fold if there were general base
catalysis. The reaction is thus insensitive to the free base,
HPO42Ϫ, within the error limits of the measurements.
Imidazolysis of 3-cyanophenyl N-methylpyridinium-4-carb-
oxylate was followed at 275 nm in the presence of increasing
concentrations of phosphate at pH 7.00 and an intermediate
was detected. The kinetics were analysed according to the
biexponential equation (2) for the consecutive two step process
(A
B
C) where L1, L2 and L3 are functions of ka, kb,
molar absorption coefficients of A, B and C and the initial
concentration of A.
Abst = L1exp(Ϫkat) ϩ L2exp(Ϫkbt) ϩ L3
(2)
The value ka corresponds to formation of the acyl-imidazole
and kb to its decomposition.
The pseudo first order rate constants measured spectro-
photometrically for the reactions of the esters with imidazole fit
a rate law involving a squared term in total imidazole concen-
tration (Eqn. (3)).
kobs = kbackground ϩ k2Ј[total imidazole] ϩ
k3Ј[total imidazole]2 (3)
The parameter kbackground is essentially due to the reaction of
hydroxide ion with the ester. The fit to Eqn. (3) is illustrated in
Fig. 2 for the 4-nitrophenyl ester. The k2Ј term for the 3-cyano-
phenyl ester is demonstrated to be due to reaction of the base
form (free imidazole) as k2Ј is proportional to the fraction of
free base (FB); this result is expected from previous work with
other phenyl esters.1–5 The k3Ј term for the 3-cyanophenyl ester
is proportional to (FB × FB) as illustrated in Fig. 3. Eqn. (1)
therefore governs the reaction of imidazole with the 3-cyano-
Fig. 4 Effect of increasing concentration of 2,6-lutidine on the rate
of reaction of imidazole with 3-cyanophenyl N-methylpyridinium-4-
carboxylate ester. Conditions as in the text.
The kinetic parameters of Eqn. (1) fitting the observed rate
constants of the esters are collected in Table 2 together with the
conditions of pH, buffer, ranges of imidazole concentrations
and observed rate constants.
Fig. 2 Reaction of 4-nitrophenyl N-methylpyridinium-4-carboxylate
in buffers containing increasing concentrations of imidazole.
Conditions as in Table 2.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 9 5 – 2 0 0 0
1997

View Article Online
Table 2 Kinetic parameters and conditions for the reaction of imidazole with substituted phenyl N-methylpyridinium-4-carboxylate esters in
aqueous solutiond, e
Substituent
pKArOH
k3/MϪ2 sϪ1
k2/MϪ1 sϪ1
pH
104 kOH/MϪ1 sϪ1 g
N
[Imidazole]f
kobs × 103/sϪ1
λ/nm
a
H
4-Ac
3-Cl
3-CN(H2O)
3-CN(D2O)
3-NO2
4-NO2
9.95
8.05
9.02
8.61
0.641 0.068
5.13 0.46
2.33 0.24
3.32 0.47
0.749 0.203
8.29 0.68
9.44 0.65
0.0161 0.0066
0.32 0.038
0.115 0.021
0.172 0.035
0.114 0.038
0.409 0.059
1.18 0.060
8.00a
0.309 0.034
1.21 0.17
0.872 0.035
1.45 0.07
—
1.68 0.09
3.24 0.48
27
12
21
43
17
21
15
9–92
16–80
4–90
3–80
40–120
6–90
3.4–10.3
13–70
8.6–38
6.5–44.8
1.7–8.95
20–120
6.0–190
285
325
290
275
275
330
350
7.99a
7.99a
6.47–7.50a, c
7.00 (pD)
8.00a
8.35
7.14
7.00b
2.6–92
a 0.05 M HEPES. b No added buffer. c Fractions of base (data points): 0.23 (6), 0.5 (14), 0.73 (7) and 0.76 (22); k3Ј : 0.159, 0.623, 1.38 and 1.46 MϪ2 sϪ1
respectively. d Solvent composition was between 1 and 2% v/v CH3CN–water. e Temperature 25 ЊC, ionic strength adjusted to 0.5 M with KCl.
f Total imidazole concentration/mM. g kOH for 4-cyano ester is 2.30 0.04 × 104 MϪ1 sϪ1
The parameters k2 and k3 fit Brønsted type equations (4) and
.
Discussion
(5) respectively and these are illustrated in Fig. 5. The second
order rate constants for alkaline hydrolysis of the esters were
obtained from the values of kbackground and are recorded in Table
2; they fit the Brønsted equation (6) which has a βLg similar to
that for alkaline hydrolysis of substituted phenyl esters of other
acids.
The reaction between substituted phenyl N-methylpyridinium-
4-carboxylate esters and imidazole in aqueous buffers involves
nucleophilic substitution at the ester followed by hydrolysis of
the N-(NЈ-methylpyridinium-4-carbonyl)imidazole intermedi-
ate. Fig. 1 shows that a hydroxylamine-active species (the acyl-
imidazole) persists after all the 4-nitrophenolate ion has been
expelled from the ester. The reactivity of 2,6-lutidine with the
13
4-nitrophenyl ester (< 0.08 MϪ1 sϪ1
)
is significantly less than
that of k2 for imidazole attack (1.19 MϪ1 Ϫ1). General base
s
catalysed hydrolysis by imidazole is therefore of minor import-
ance because the hindered 2,6-lutidine, which has a similar
basicity (pKa = 6.77) to that of imidazole (pKa = 7.05), should
have approximately the same reactivity. Kinetics of the reaction
of 3-cyanophenyl N-methylpyridinium-4-carboxylate ester in
imidazole buffers in the presence of phosphate dianion exhibit
an intermediate when followed at 275 nm (Eqn. (2)). The inter-
mediate is assumed reasonably to be the acyl-imidazole and
decays due to its reaction with phosphate dianion.
Substitution reactions between heterocyclic bases (such as
pyridines) and aryl esters and other unsaturated centres can be
stepwise17 involving an addition intermediate which breaks
down to give the product (see for example Eqn. (7)). The
stability of the zwitterionic intermediate
Fig. 5 Brønsted lines for the parameters k2 (ꢀ) and k3 (᭿) for the
reaction of substituted phenyl N-methylpyridinium-4-carboxylate
esters in imidazole containing buffers. Lines are calculated from Eqns.
(4) and (5).
(7)
log k2 =
Ϫ0.65 0.077pK Aa rOH ϩ 4.83 0.66 (r = 0.9729) (4)
is thought to be the cause of the reaction taking a step-
wise pathway18 rather than the concerted route often taken by
substitution reactions of phenolate ions at unsaturated centres
(Eqn. (8)).19 Guthrie and Pike calculated that the energies of the
putative tetrahedral and acylium ion
log k3 =
Ϫ0.42 0.090pK Aa rOH ϩ 4.16 0.77 (r = 0.9208) (5)
log kOH
=
Ϫ0.352 0.053pK Aa rOH ϩ 7.08 0.15 (r = 0.9486) (6)
The proportion of reaction flux taken by the k3 term relative
to that via the k2 term in the reaction of imidazole with the
N-methylpyridinium-4-carboxylate ester of 4-nitrophenol is
much greater than that for the other published systems. Calcu-
lations from the data from Table 2 show that at 0.1 M imid-
azole, the percentage reaction flux taken through the k3 term by
the 4-nitrophenyl N-methylpyridinium-4-carboxylate ester is
44% and at 1 M imidazole it is 89%.
(8)
intermediates are such that the displacement of 4-nitrophenol-
ate ion from the acetate ester by imidazole involves a transition
structure close to that of the tetrahedral intermediate.20 The
present results for the N-methylpyridinium-4-carboxylate esters
can be explained on the basis of the stepwise mechanism
(Scheme 3) where the k3 term arises from the transition state ‡3
for proton transfer to the second imidazole with subsequent
non-rate limiting ArO–C bond fission; the k2 term corresponds
to ‡2. The effective charge map17,21,22 of the system (derived
from the Brønsted exponents, Scheme 3) shows that the leaving
oxygen is more negatively charged in the transition structure ‡2
(Ϫ0.10) than in ‡3 (ϩ0.13).
Deuterium oxide solvent isotope effects of kH2 /D = 1.51 and
kH3 /D = 4.43 for the 3-cyanophenyl ester are well within the
ranges expected for reactions involving no rate limiting
proton transfer (for k2) and proton transfer (for k3).16 However,
both isotope effects are larger than those reported by Caplow
and Jencks for the imidazolysis of 4-nitrophenyl 4-nitro-
benzoate.3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 9 5 – 2 0 0 0
1998

View Article Online
Scheme 3
The D2O solvent isotope effect on k2 (kH/D = 1.51) is consist-
2
ent with ‡2 in Scheme 4; such values of the isotope effect (1.0–
1.8) have been observed for other imidazolysis reactions where
no proton transfer step is rate limiting.5
Scheme 5
the anionic leaving group. This is possible in a transition struc-
ture ‡3 shown in Scheme 4 where the catalytic imidazole can
remove a proton from the imidazolium NH of (1) as well as
interacting (for example by H-bonding or proton transfer) with
the phenolic oxygen. Model-building demonstrates that the
stereochemistry is possible and catalysis should occur when the
planes of the two imidazolyl nuclei are roughly parallel. It is
possible that some advantage is achieved by a stacking inter-
action23 between the two imidazolyl functions.
The proton transfer from 1 to the catalytic imidazole in ‡3
cannot occur with linear geometry. The transfer involves over-
lap of the lone pair HOMO on the catalytic imidazole with the
LUMO of the Nϩ–H bond in T . The diagram, ‡3, in Scheme 4
illustrates a notional MO configuration for the transition struc-
ture (filled orbitals shaded and the empty orbital on the proton
unshaded). Although this geometry appears to be unprece-
dented it is analogous to known hydrogen-bond systems where
the donor–H–acceptor bonding is non-linear.24–26 The transi-
tion structure, ‡3, would involve a relatively symmetrical
arrangement of the hydrogen between the nucleophilic nitrogen
atoms as described in the structure in Scheme 4 and this could
be considered qualitatively as an interaction between the empty
σ orbital of the proton and the filled sp-orbitals comprising the
lone pairs on the imidazolyl nitrogens. The proton transfer,
and ANDN process, is analogous to the front-side nucleophilic
aliphatic substitution mechanisms (SNF) of contemporary
interest.27
Scheme 4
General bases do not catalyse the imidazolysis reaction and
this excludes the mechanism shown in Scheme 5 analogous to
that for aminolysis by primary and secondary amines. Never-
theless, the k3 term involves rate limiting proton transfer as
evidenced by the D2O solvent isotope effect (kH/D) of 4.43.
3
Satterthwait and Jencks6 estimated that the change in effective
charge on the ether oxygen of an aryl ester from the ester react-
ant to an addition intermediate (1) is Ϫ0.6 (∆ε), indicating a
build-up of charge more negative than that observed in the
transition structure of the rate limiting step (∆ε = Ϫ0.42). The
mechanism of Scheme 5 with a rate limiting transition state
after formation of (1) would not be expected to have a smaller
build-up of negative effective charge on the aryl oxygen than
that in the addition intermediate (1). Moreover, the mechanism
for k3 represented by Scheme 5 predicts catalysis by general
bases at a level not observed for the present system.
A mechanism for k3 where the imidazole base owes its
reactivity solely to stacking with the imidazolium group in (1)
would explain the absence of catalysis by the phosphate dianion
which is unable to complex by stacking. Although the preferred
The effective charge on the leaving oxygen in ‡3 (more posi-
tive than expected in the intermediate (1)6) is consistent with a
mechanism where the catalyst interacts electrophilically with
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 9 5 – 2 0 0 0
1999

View Article Online
model includes a stacking process the effective charge results
indicate that a significant component of ‡3 comes from an
electrophilic interaction at the phenolic oxygen atom. The
mechanism involving ‡3 in Scheme 4 explains why 2,6-lutidine is
not a catalyst because, although it can stack with 1 and has its
lone pair available for the proton transfer, there is no electro-
philic component available for interaction at the phenolic oxy-
gen. Proton transfer at the leaving oxygen is not a reasonable
activation process because consideration of the microscopic
reverse mechanism reveals that attack by phenol would involve
general base catalysis, an unlikely requirement under the
prevailing conditions of pH.
A process where the proton transfer from 1 to catalytic imid-
azole is “linear” would not allow any interaction with the leav-
ing aryl oxygen and would therefore not explain why the effect-
ive charge on the oxygen in the transition structure is more
positive than that in the transition structure for the uncatalysed
process (‡2).
k3 (ammonia catalysed ammonolysis) is 1.84. The value of 1.84
corresponds to the formation of the T intermediate as the
proton transfer step should not depend on σ. The lower value
of ρ2 is due to the k step being rate limiting having a transition
structure with substantial carbonyl bond formation bringing
the structure back closer to that of the reactant than T . The
value of 1.84 is similar to that observed for the addition of
nucleophiles to substituted benzoyl groups such as benzalde-
hydes.28 The existence of catalysis in a termolecular term by
tertiary amines precludes the mechanism described here for the
ammonolysis reactions of esters.
On the basis of the arguments for the special base catalytic
function of imidazole in imidazolysis of N-methylpyridinium-
4-carboxylate esters it is reasonable to postulate a similar
mechanism for the k3 term of the acetate5 and benzoate3 esters.
References
Although the mechanism involving rate limiting proton
transfer from T is excluded, rate limiting attack by the imid-
azolium ion on T (kϪ step in Scheme 1) would give the
1 T. C. Bruice and G. L. Schmir, J. Am. Chem. Soc., 1957, 79, 1653.
2 M. L. Bender and B. W. Turnquest, J. Am. Chem. Soc., 1957, 79,
1656.
3 M. Caplow and W. P. Jencks, Biochemistry, 1962, 1, 883.
4 J. F. Kirsch and W. P. Jencks, J. Am. Chem. Soc., 1964, 86, 833.
5 T. C. Bruice and S. J. Benkovic, J. Am. Chem. Soc., 1964, 86, 418.
6 A. C. Satterthwait and W. P. Jencks, J. Am. Chem. Soc., 1974, 96,
7018.
7 J. J. Morris and M. I. Page, J. Chem. Soc., Perkin Trans. 2, 1980, 220.
8 J. J. Morris and M. I. Page, J. Chem. Soc., Perkin Trans. 2, 1980, 685.
9 M. I. Page and W. P. Jencks, J. Am. Chem. Soc., 1972, 94, 8818–8828.
10 T. C. Bruice and M. F. Mayahi, J. Am. Chem. Soc., 1960, 82, 3067.
11 J. F. Kirsch and A. Kline, J. Am. Chem. Soc., 1969, 91, 1841.
12 T. C. Bruice, A. Donzel, R. W. Huffman and A. R. Butler, J. Am.
Chem. Soc., 1967, 89, 2106.
observed rate law (k3). The proton transfer step T
TϪ
would not give rise to a substantial solvent deuterium oxide
effect because it would be an equilibrium step. Thus, the
observed isotope effect of 4.43 would come from kϪ and indi-
cate rate limiting proton transfer to the departing aryl oxygen
leaving group. However a good leaving group such as an aryl
oxide ion should not require assistance from proton transfer as
indicated above.
The decomposition of 1 via ‡3 does not necessarily involve
concerted ArO–C bond fission and proton transfer steps. The
simplest explanation is that the decomposition involves proton
transfer to the catalytic imidazole in the rate limiting step and
the ArO–C fission occurs in a subsequent non-rate limiting
step. The effective charge on the aryl oxygen is more positive
than that in k2 due to an electrophilic interaction rather than to
a bond fission process.
The termolecular term is not seen with 4-nitrophenyl esters
of aliphatic acids nor with those of 4-methoxy- or 4-methyl-
benzoic acids and increasing the reactivity of the acyl function
promotes its observation. The termolecular term provides an
alternative route to direct expulsion of the phenolate anion (the
k2 term). As the acyl group becomes more reactive (electron
withdrawing) the decomposition of the intermediate 1 via k2
becomes less efficient (relative to k3) because the product acyl-
imidazolium ion becomes less stable. The proton transfer step
(k3) provides a more efficient pathway than that represented by
k2 because the initial product is the more stable acyl-imidazole
(rather than acyl-imidazolium ion) product.
13 M. J. Colthurst, M. Nanni and A. Williams, J. Chem. Soc.,
Perkin Trans. 2, 1996, 2285.
14 T. Deacon, A. Steltner and A. Williams, J. Chem. Soc., Perkin Trans.
2, 1975, 1778.
15 S. Hestrin, J. Biol. Chem., 1949, 180, 249.
16 S. L. Johnson, Adv. Phys. Org. Chem., 1967, 5, 237.
17 A. Williams, Concerted Organic and Bio-Organic Mechanisms,
CRC Press, Boca Raton, 2000, Section 4.1.
18 E. Chrystiuk and A. Williams, J. Am. Chem. Soc., 1987, 109, 3040.
19 S. A. Ba-Saif, A. K. Luthra and A. Williams, J. Am. Chem. Soc.,
1987, 109, 6362.
20 J. P. Guthrie and D. C. Pike, Can. J. Chem., 1987, 65, 1951.
21 A. Williams, Adv. Phys. Org. Chem., 1992, 27, 1.
22 S. Thea and A. Williams, Chem. Soc. Rev., 1986, 15, 625.
23 Stacking phenomena are common between heterocyclic bases in
water particularly when there is potential for a π-donor–π-acceptor
interaction. In this article we refer to stacking as the result of an
interaction between two planar molecules rather than multiple
stacking as observed between heterocyclic bases in nucleic acids.
24 S. Scheiner, Acc. Chem. Res., 1994, 27, 402.
25 H. A. Staab and T. Saupe, Angew. Chem., Int. Ed. Engl., 1988, 27,
865.
26 M. A. Zirustein and H. A. Staab, Angew. Chem., Int. Ed. Engl., 1987,
26, 460.
27 E. Uggerud and J. Beache-Andreassen, Chem. Eur. J., 1999, 5,
1917; M. N. Glukhovtsev, A. Pross, H. B. Schlegel, R. D. Bach and
L. Radom, J. Am. Chem. Soc., 1996, 118, 11258; E. Uggerud,
J. Org. Chem., 2001, 66, 7084; R. R. Sauers, J. Org. Chem., 2002, 67,
1221; P. R. Schreiner, P. v. R. Schleyer and R. K. Hill, J. Org. Chem.,
1994, 59, 1849; S. Harder, A. Streitwieser, J. T. Petty and P. v. R.
Schleyer, J. Am. Chem. Soc., 1995, 117, 3253.
Bruice and Mayahi10 showed that the ammonolysis reaction
of substituted phenyl acetates has terms corresponding to gen-
eral base catalysis by ammonia (k3) and direct expulsion (k2).
The Brønsted slopes are Ϫ0.25 and Ϫ0.84 respectively; the
value of Ϫ0.84 measures the change in charge from reactant to
the transition structure of the k step. The βLg of Ϫ0.25 meas-
ures the change in charge to T although this value is very
uncertain due to the paucity of data and the small spread
of pKa values. Kirsch and Kline11 studied the ammonolysis of
4-nitrophenyl esters of substituted benzoic acids and found a
value of ρ2 (rate via k ) (see Scheme 1) of 1.08 whereas that for
28 S. L. Shames and L. D. Byers, J. Am. Chem. Soc., 1981, 103,
6170.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 9 5 – 2 0 0 0
2000