Oxidation of tertiary benzamides by 5,10,15,20-tetraphenyl- porphyrinatoironIII chloride-tert-butylhydroperoxide

Article abstract of DOI:10.1039/b404169k

Tertiary benzamides are oxidized by the 5,10,15,20- tetraphenylporphyrinatoiron(III) chloride-Bu^tOOH system at the α-position of the N-alkyl groups. The major products are N-acylamides, although small amounts of secondary amides, the products of dealkylation, are also formed. Plots of initial rate versus initial substrate concentration for these reactions are curved, suggesting formation of an oxidant-substrate complex. The reaction rates are almost insensitive to the substituent in the benzamide moiety, but there is a kinetic deuterium isotope effect of 5.6 for the reaction of the N,N-(CH₃)₂ and N,N-(CD₃) ₂ compounds. Comparison of the reaction products from N-alkyl-N-methylbenzamides reveals that, for all compounds studied except N-cyclopropyl-N-methylbenzamide, oxidation of the alkyl group is preferred, strongly so (by a factor of ca. 8) for N-allyl-N-methylbenzamide. In contrast to microsomal oxidation, there is no steric hindrance to oxidation of an isopropyl group. Thus, we propose that these reactions proceed via hydrogen atom abstraction to form an α-carbon-centred radical and we attribute the observed diminished reactivity of the N-cyclopropyl group to its known reluctance to form a cyclopropyl radical. Oxidation of N-methyl-N-(2,2,3,3- tetramethylcyclopropyl)methylbenzamide provides preliminary evidence for rearrangement of an intermediate radical. While it remains unclear how these reactions proceed directly to the N-acyl products, we have established that N-hydroxymethyl, N-alkoxymethyl and N-alkylperoxymethyl intermediates are not involved.

Full text of DOI:10.1039/b404169k

View Article Online / Journal Homepage / Table of Contents for this issue
Oxidation of tertiary benzamides by 5,10,15,20-tetraphenyl-
porphyrinatoironIII chloride–tert-butylhydroperoxide
Luis Constantinoa and Jim Iley*b
a CECF, Faculdade de Farmácia de Lisboa, Avenida das Forças Armadas, 1600 Lisboa,
Portugal
b Chemistry Department, The Open University, Walton Hall, Milton Keynes, MK7 6AA, UK.
E-mail: j.n.iley@open.ac.uk
Received 22nd March 2004, Accepted 27th April 2004
First published as an Advance Article on the web 4th June 2004
Tertiary benzamides are oxidized by the 5,10,15,20-tetraphenylporphyrinatoiron() chloride–ButOOH system at
the α-position of the N-alkyl groups. The major products are N-acylamides, although small amounts of secondary
amides, the products of dealkylation, are also formed. Plots of initial rate versus initial substrate concentration for
these reactions are curved, suggesting formation of an oxidant–substrate complex. The reaction rates are almost
insensitive to the substituent in the benzamide moiety, but there is a kinetic deuterium isotope effect of 5.6 for the
reaction of the N,N-(CH3)2 and N,N-(CD3)2 compounds. Comparison of the reaction products from N-alkyl-N-
methylbenzamides reveals that, for all compounds studied except N-cyclopropyl-N-methylbenzamide, oxidation
of the alkyl group is preferred, strongly so (by a factor of ca. 8) for N-allyl-N-methylbenzamide. In contrast to
microsomal oxidation, there is no steric hindrance to oxidation of an isopropyl group. Thus, we propose that these
reactions proceed via hydrogen atom abstraction to form an α-carbon-centred radical and we attribute the observed
diminished reactivity of the N-cyclopropyl group to its known reluctance to form a cyclopropyl radical. Oxidation
of N-methyl-N-(2,2,3,3-tetramethylcyclopropyl)methylbenzamide provides preliminary evidence for rearrangement
of an intermediate radical. While it remains unclear how these reactions proceed directly to the N-acyl products,
we have established that N-hydroxymethyl, N-alkoxymethyl and N-alkylperoxymethyl intermediates are not
involved.
Introduction
Certain N,N-dialkylamides have important biological activity.
For example N,N-diethyl-3-toluamide is the most effective
insect repellent in world-wide use,1 N,N-diethylnicotinamide is
a respiratory stimulant2 and N,N-dimethylformamide is an
industrially important solvent that is toxic on occupational
exposure.3 All these compounds are known to undergo in vivo
oxidative N-dealkylation in the liver. We have studied the
mechanism of this biochemical N-dealkylation reaction using
substituted N,N-dialkylbenzamides as model substrates and
employing a combination of aryl substituent, kinetic deuterium
isotope effects and N-alkyl structural effects on the kinetic
constants together with product studies to probe which of the
possible pathways is followed.4,5,6 These studies were carried
out using microsomes and, since the enzyme responsible, cyto-
chrome P450, is membrane bound, access by the substrate to
the active site often influences the kinetics of the reactions. For
example, log Vmax/KM values for total dealkylation correlate
with the log P value of the substrate.7 Moreover, the large
intrinsic isotope effect for the dealkylation reaction (≥ 6) is
entirely suppressed,5 which suggests that processes like
membrane diffusion to the active site mask the rate at which the
oxidative dealkylation proceeds. For this reason, we decided to
study the oxidation of tertiary dialkylamides 1 by chemical
model systems that avoid the physical constraints of the micro-
somal incubations. The two systems chosen were tetraphenyl-
porphyrinatoiron() chloride 2/tert-butyl hydroperoxide
(TPPFe/ButOOH)
and
tetraphenylporphyrinatoiron()
chloride/iodosobenzene (TPPFe/PhIO), because they are
known to be useful tools in the study of the mechanism of
cytochrome P450 biooxidations.8,9
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
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 4
1894

View Article Online
Table 1 Values of Vmax, Vmax/K and K for the oxidation of N,N-dimethylbenzamides 1a–e by TPPFe/ButOOH in CH2Cl2 at 30 ЊC. [TPPFe] =10Ϫ3
mol dmϪ3, [ButOOH] = 0.2 mol dmϪ3
Model system
Microsomesb
Substrate
IEa/eV
106 Vmax/mol dmϪ3 sϪ1
K/mol dmϪ3
106 Vmax/K/sϪ1
Vmax/KM/hϪ1 (nmol P450)Ϫ1
1a
1b
1c
1d
1e
9.04 (9.55)
8.40 (9.26)
9.16 (9.66)
9.46 (10.0)
(9.55)
1.22
1.59
1.43
0.814
0.223
0.47
0.70
0.59
0.33
0.47
2.61
2.29
2.45
2.43
0.465
1.33
1.15
1.6
1.11
0.97
a Values are the vertical IEs taken from ref.10. Values in parentheses are the corresponding IEs calculated by the AM1 semi-empirical method. b Data
taken from ref. 5.
small quantities of N-methylbenzamides that are formed in
these kinetic reactions, the most appropriate method
Results and discussion
Products of reaction
of quantifying the total extent of reaction was to convert the
N-formyl-N-methyl products to the corresponding N-methyl
compounds. This was accomplished by treating samples of the
reaction mixture with ethanolic sodium hydroxide and then
measuring the total N-methylbenzamide present. Since the
porphyrin and ButOOH undergo unwanted side reactions that
compete with N-alkyl oxidation, it proved necessary to use the
initial rate method to determine reaction velocities. These were
obtained by plotting product concentrations versus time over a
2 h period. At low extents of conversion these plots were found
to be linear, the slopes giving the initial rate of reaction, νi.
However, plots of νi determined using varying initial substrate
concentrations but at fixed concentrations of TPPFe and
ButOOH are curved rather than linear (Fig. 1), implying that
some kind of complex involving the substrate and the active
oxidant is formed. Interpolation of the curves enables values of
Vmax (= k[oxidant], where k is the rate constant for the form-
ation of the products from the catalysis complex) and Vmax/K to
be determined in a manner analogous to those obtained from
the Michaelis–Menten kinetics in enzyme catalysed reactions.
Values of Vmax, Vmax/K and K are given in Table 1, along with
the microsomal Vmax/Km values5 for comparison.
Oxidation of N,N-dimethylbenzamides 1a–d by TPPFe/
ButOOH in CH2Cl2 gives rise to the corresponding N-formyl-N-
methylbenzamides 3 (Scheme 1) as the major product, together
with small quantities of the secondary amide 4, the product of
dealkylation. Yields were unaffected when the reactions were
carried out under a nitrogen atmosphere. Oxidation by the
TPPFe/PhIO system led to the same product distribution but in
lower yields (4.1% vs. 7.4% after 72 h at 30 ЊC; for full reaction
conditions see the Experimental). For this reason, and also
because the oxidations in the presence of iodosobenzene are
more difficult to analyse by HPLC, we decided to proceed using
the TPPFe/ButOOH system.
Scheme 1 Products from the reaction of N,N-dimethylbenzamides
1a–d with TPPFeCl and ButOOH.
Microsomal incubation of N,N-dimethylbenzamide yields
N-hydroxymethyl-N-methylbenzamide that decomposes subse-
quently to N-methylbenzamide.5 The pattern of products found
in the reactions with TPPFe/ButOOH would appear to suggest
the intermediacy of the N-hydroxymethyl-N-methylbenzamide
5. This compound could form the formyl derivative 3 by further
oxidation of the hydroxymethyl group, and the secondary
amide 4 by the loss of formaldehyde. However, even using the
HPLC system that we have previously used to analyse for 5,5 we
have been unable to detect such an intermediate in any of our
reactions. Thus the chemical model system used in the present
reactions appears not to mimic directly the microsomal system
that we have used to investigate the oxidative dealkylation of
N,N-dialkylbenzamides.
Fig. 1 Plot of initial rate, νi versus substrate concentration for 1a.
Effect of substituent in the aryl moiety
Inspection of the data in Table 1 for the aryl substituted com-
pounds 1a–d reveals that there is remarkably little variation in
the magnitude of Vmax, Vmax/K and K with the aryl substituent,
particularly so for the pseudo-first-order rate constant, Vmax/K.
What little variation there is appears not to follow any simple
pattern with the electron withdrawing ability of the substituent,
which contrasts with the experimental10 and calculated (AM1
SFC MO) ionisation potentials for these compounds (Table 1).
Omitting either the porphyrin catalyst or the hydroperoxide
from the incubation mixtures of the N,N-dialkylamides results
in no reaction whatsoever. It seems reasonable, therefore, to
suggest that TPPFe and ButOOH react to form the active
oxidant.
The rates at which the methyl groups were oxidised in the
amide substrates 1a–d were best determined by monitoring the
reactions at timed intervals using HPLC, which is able to
separate and quantify (using appropriate standards) the prod-
ucts and starting materials of the reactions. Given the very
Isotope effect
Interestingly, the data for the N,N-di(trideuteriomethyl) sub-
strate 1e reveal that the rate differences observed for this
compound as compared with 1a are entirely due to the rate at
which the complex breaks down to products. Thus, the
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
1895

View Article Online
Table 2 Relative rates of alkyl versus methyl oxidation, corrected for
the number of oxidisible hydrogen atoms, for compounds 1a,f–l, using
TPPFe/ButOOH at 30 ЊC in CH2Cl2
magnitude of the kinetic deuterium isotope effect (KDIE)
obtained for both Vmax and Vmax/K is 5.6 whereas no isotope
effect is obtained for K, as is to be expected if K is related to the
formation of a substrate/active oxidant complex.
Relative rate, N-R/N-Me
Such a large primary KDIE together with the absence of any
correlation of the rates of reaction with aryl substituent tends
to point to a mechanism that involves direct hydrogen atom
abstraction from the substrate (Scheme 2, path a) rather than
formation of an amide radical cation via electron transfer
(Scheme 2, path b).
Microsomes
a
a
Substrate
TPPFeCl/ButOOH
Vmax
Vmax/KM
1a
1f
1g
1h
1i
1j
1k
1l
1.0
1.11 ( 0.20)
1.40 ( 0.25)
1.25 ( 0.34)
7.89 ( 0.45)
0.65 ( 0.11)
1.61 ( 0.20)
5.63 ( 0.98)
1.0
0.9
0.08
5.1
0
0.76
—
0.44
0.82
0.04
1.75
0
0.13
—
a Data taken from refs. 4 and 7.
group (Vmax), or a preference for oxidation of the methyl group
(Vmax/KM) (Table 2).4 Similar results were obtained for micro-
somal incubation of the other amides 1g–1k,4,11 except for 1i
which showed a distinct preference for the oxidation of the
alkyl group. The microsomal results showed that most probably
the amide approaches the active site of the enzyme at a con-
formation that favours the metabolism of the smallest group,
though in a substrate like 1i, the stability of the allyl radical7
could result in a higher rate constant for the metabolism of the
allyl group.
Relative rate data for the TPPFe/ButOOH model oxidations
of the amides containing other N-alkyl groups are also con-
tained in Table 2. These show that, in general, the larger the N-
alkyl group (1g,h) the greater the preference for the oxidation of
the alkyl group over the methyl group. This would be entirely
consistent with the stabilisation of the carbon-centred radical
afforded by the inductive effect of the larger alkyl group.
Unfortunately, it is also consistent with the formation of an
iminium ion (see Scheme 2), the larger alkyl group being able to
preferentially form the more stable iminium ion due to greater
inductive stabilization. Interestingly, the reaction of the iso-
propyl group in compound 1h is clearly not subject to steric
restraints. This is in contradistinction to the microsomal
reaction where the oxidation of the Pri group was clearly
disfavoured.4
Scheme 2 Pathways for the oxidation of N,N-dimethylbenzamide.
Effect of the alkyl group
We extended our studies to the oxidation of several different
substituted N-alkyl-N-methylbenzamides, 1f–i, to examine the
effect of the alkyl group. For these compounds, by determining
the relative amounts of products formed we were able to
quantify the relative rate of alkyl versus methyl oxidation. For
example, for N-ethyl-N-methylbenzamide, 1f, oxidation using
the chemical model system gave rise to two major products,
N-formyl-N-ethylbenzamide and N-acetyl-N-methylbenzamide,
and two minor products, N-methylbenzamide and N-ethyl-
benzamide (Scheme 3).
As in the microsomal oxidations, a more marked effect was
seen with N-allyl-N-methylbenzamide, 1i. In this substrate the
relative rate of allyl versus methyl oxidation after correction
for the number of potential hydrogen atoms that can be
removed is 7.89 ( 0.45) (Table 2). This reveals a marked
preference for the oxidation of an allyl group, consistent with
the enhanced stability of the allyl carbon-centred radical.
Mechanistic probes
Included in Table 2 are compounds 1j, 1k and 1l that were
synthesised as possible probes for the reaction mechanism.
Compound 1j is N-cyclopropyl-N-methylbenzamide. In a
mechanism where electron abstraction to form an amide radical
cation is the first step, it might be expected that the radical
cation from 1j would rearrange (Scheme 4) thereby giving
products originating from the opening of the cyclopropyl ring.
This strategy has been adopted to intercept radical cations in
the oxidative dealkylation of tertiary amines by cytochrome
P-450.12,13 However, no ring opening products could be detected
in the chemical model oxidation of 1j. Moreover, formation
of a carbon-centred radical at the α-carbon of the cyclopropyl
ring is unlikely given that the difficulty of abstracting a
hydrogen atom in a cyclopropane ring is well known, as illus-
trated by the relative rate of 0.2 for the reactivities of C–H
Scheme 3 Products obtained from the oxidation of 1f by TPPFe/
ButOOH.
After treatment of an aliquot of the reaction mixture with
NaOH (to promote deacylation) the relative extent of N-methyl
versus N-ethyl oxidation is determined by HPLC quantitation
of the two secondary amides. For compound 1f, the relative rate
of ethyl versus methyl oxidation, after correction for the
number of potential hydrogen atoms that can be removed, (two
for Et, three for Me) is 1.11 ( 0.20) (Table 2), which reveals
a slight preference for the oxidation of the ethyl group. This
ratio was found to be invariant with either time or substrate
concentration.
t
ؒ
In contrast, when 1f was subjected to microsomal incubation
it displayed either no preference for the oxidation of the alkyl
bonds in cyclopropane and neopentane towards the Bu OO
radical.14 Thus, for a mechanism that proceeds via hydrogen-
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
1896

View Article Online
Fig. 2 Mass spectra of 1l and 1l-a.
system, a product that had an Mr of two mass units lower was
identified by GC-MS of the reaction mixture (Fig. 2), which we
tentatively assign to the ring opened diene 1l-a (Scheme 5). In
both substrates 1k and 1l, reaction in the cyclopropylmethyl
group was clearly favoured over the methyl group, the reactivity
ratios being 1.61 ( 0.20) for 1k and 5.63 ( 0.98) for 1l, respect-
ively, which is consistent with a carbon-centred radical mechan-
ism. However, it is not clear why the ratio for 1l is so much
larger than that for 1k, which is similar to those for other
N-alkyl-N-methylamides (Table 2). AM1 SCF MO calculations
are unable to reveal a reason for this difference (Table 3). Thus,
the ∆∆Hf between the heats of formation of the amides 1a, 1f,
1k, and 1l and their corresponding radicals show that (i) radical
formation from the Me group is essentially equi-energetic for all
four compounds (ii) radical formation from the alkyl groups is
favoured over radical formation from the Me group, and (iii)
radical formation from the alkyl group in 1l is the most
favoured, but not enough compared to the other amides to
account for the higher relative ratio for 1l in Table 2.
Scheme 4 Mechanistic probes for amide dealkylation.
atom abstraction to form a carbon-centred radical, it is to be
expected that the cyclopropyl group would be more difficult to
oxidise than the methyl group. Our results are consistent with
this; the relative rate of cyclopropyl versus methyl N-dealkyl-
ation, after correction for the number of potential hydrogen
atoms that can be removed, is 0.65 ( 0.11) (Table 2). While
steric arguments could be invoked to explain the lower reactiv-
ity of the cyclopropyl group, the isopropyl group in 1h is more
susceptible to oxidation than a methyl group and twice as likely
to undergo reaction than the cyclopropyl group (Table 2).
Compounds 1k and 1l were studied because a cyclopropyl
ring adjacent to a methylene group can potentially intercept a
radical located in the methylene group, via opening of the
cyclopropyl ring and diversion of the reaction to other products
(Scheme 4). Methylcyclopropanes have been successfully
used to probe a carbon-centred radical mechanism in the
cytochrome P450-catalysed hydroxylation of alkanes and to
measure the rate of oxygen rebound in this oxidation.15 Bowry
et al. calibrated a series of α-alkyl-substituted cyclopropyl-
carbinyl radicals by measuring the rate constants for ring
opening of methylcyclopropyl radicals.16 These authors also
measured rate constants for hydrogen-atom abstraction by the
tert-butoxyl radical from various positions in twelve substituted
cyclopropanes. The reactivity of the methylene groups adjacent
to the cyclopropyl ring is well illustrated by the relative rates of
1.63 to 0.053 for H atom abstraction from methylene versus
methyl group in the ethyl group of ethylcyclopropane (cyclo-
propane C–H bonds have a relative rate lower than 0.02).
If a carbon-centred radical is formed in the model amide
oxidations then we might expect oxidation to occur faster in the
methylene part of the cyclopropylmethyl group than in the
N-methyl group. Besides, if the ring opening of the radical
occurs at a rate faster than the rate of oxygen rebound to the
carbon-centred radical then ring-opening products would also
be expected. When substrate 1k was subjected to the TPPFe/
ButOOH oxidation no products originating from ring opening
were detected. When 1l was subjected to the same oxidation
Scheme 5 Possible pathway for formation of product 1l-a in the
chemical model oxidation of 1l.
Possible intermediates
The data point to the involvement of a carbon-centred radical
intermediate in the oxidation of N,N-dialkylbenzamides by
TPPFe/ButOOH. However, in contrast to the microsomal
reaction, which follows a similar mechanism,5,6 the products
identified in the model reactions were either the N-acyl-N-
alkylbenzamides or the corresponding N-alkylbenzamides
rather than the N-hydroxyalkyl-N-alkylbenzamide (such as 5).
The latter compound is an attractive intermediate in these
reactions because such hydroxylated compounds are the
products of many reactions catalysed by similar oxidation
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
1897

View Article Online
systems.17 Further reaction of the N-hydroxyalkylbenzamides
could be expected to follow one of two routes: oxidation to the
corresponding N-acyl-N-alkylbenzamide or decomposition to
the N-alkylbenzamide.
the small amounts of N-methylbenzamide that are formed
concomitantly.
N-tert-Butoxymethyl-N-methylbenzamide (7, R = But) is a
further possible intermediate in these amide oxidation reactions
that could be formed by reaction of a carbon-centred radical
We investigated several compounds, including compound 5,
that could be possible intermediates in these model amide
oxidations (Table 4). The rather unstable N-hydroxymethyl-
N-methylbenzamide, 5, the product of microsomal oxidation
of N,N-dimethylbenzamide, formed only the corresponding
N-methylbenzamide 4. In the absence of either TPPFe or
ButOOH, compound 5 also decomposed to N-methylbenz-
amide but at rates that were five-fold lower. Thus, the
N-hydroxymethyl compound may be an intermediate for the
production of the minor reaction product, N-methylbenzamide,
but it is not an intermediate in the formation of the N-formyl
compound 3.
t
t
ؒ
with Bu OOH or, less likely, Bu O . Unfortunately, all attempts
to synthesise 7, (R = But) failed. Consequently, we synthesised
the corresponding methyl ether 7 (R = Me) and reaction of this
led exclusively to the formation of N-methylbenzamide but not
N-formyl-N-methylbenzamide. This eliminates the N-alkoxy-
methyl-N-methylamide as an intermediate for the formation of
N-formyl-N-methylbenzamide.
The N-tert-butylperoxy ether 6 is another possible inter-
mediate in the formation of 3; this compound can be envisaged
to be formed by hydrogen atom abstraction from N,N-
dimethylbenzamide and subsequent combination of the
Nature of the oxidant
t
ؒ
carbon-centred radical with a Bu OO radical. Indeed, we were
able to identify (HPLC retention time, mass spectrum and
diode array UV spectrum) traces of this compound in the
TPPFe/ButOOH catalysed oxidation of N,N-dimethylbenz-
amide. Incubation of compound 6 with TPPFe/ButOOH led to
the formation of N-methylbenzamide as the main product; only
small amounts of N-formyl-N-methylbenzamide were formed.
The ratio of N-methylbenzamide/N-formyl-N-methylbenz-
amide was found to be 49 : 1, thus eliminating the N-tert-
butylperoxy ether as an intermediate in the formation of
N-formyl-N-methylbenzamide, but not as an intermediate in
The results obtained with our substrates are consistent with
a hydrogen atom abstraction mechanism that leads to the
formation of a carbon-centred radical. Using TPPFe/ButOOH
this carbon-centred radical could be formed either by direct
hydrogen atom abstraction by the tert-butoxyl radical, as
proposed by Mansuy for alkane hydroxylation,18 or altern-
V
IV
ؒ
atively by an iron-oxo complex of the type Fe ᎐O or Fe –O
᎐
formally equivalent to compound I of peroxidase (an oxene), as
proposed by Mansuy for the same reactions in which iodoso-
benzene is used as oxygen source.19
It is accepted that an oxene is the active species in biomimetic
systems employing porphyrins and iodosobenzene or per-
acids.18 However, there is still some controversy as to whether
Table 3 AM1SCF MO calculated heats of formation of the amides
1a, 1f, 1k and 1l and the corresponding ∆∆Hf values for formation of
the corresponding carbon-centred radicals
high-valent iron-oxo complexes of the type FeV᎐O can be
᎐
formed in model oxidations employing porphyrins and hydro-
peroxides as oxygen source. An oxene is considered to be the
active species in the epoxidation of alkenes in the presence of
TPPFe, hydroperoxides and imidazole.18 Here imidazole acts
as an electron-donating ligand to the iron of the porphyrin
nucleus and catalyses the heterolytic scission of hydroperoxide
O–O bond that is crucial for the formation of the oxene. Traylor
∆∆Hf/kJ molϪ1
Amide
∆Hf/kJ molϪ1
Me-H loss
Alkyl-H loss
1a
1f
Ϫ21.7
(E)a Ϫ47.1
(Z)a Ϫ46.9
(E)a 74.4
(Z)b 103
(E)b 98.2
103
—
82.1 (transoid)c
96.4 (cisoid)c
86.1 (transoid)c
91.8 (cisoid)c
81.8 (transoid)c
95.7 (cisoid)c
86.0 (transoid)c
94.7 (cisoid)c
80.1 (transoid)c
94.9 (cisoid)c
82.8 (transoid)c
93.1 (cisoid)c
has proposed that an oxene of the type FeV᎐O is formed even in
᎐
the absence of imidazole, and that the reaction could be
accelerated by the presence of water, alcohols or imidazole.20,21
Amides are known to act as donor molecules in metallic
complexes and have been shown to coordinate to Fe3ϩ via the
98.1
105
1k
1l
(Z)a 74.1
(E)a Ϫ6.4
(Z)a Ϫ5.5
98.6
104
oxygen atom.22 DMF, like imidazole, can stabilise FeIV᎐O and
᎐
possibly FeV᎐O iron-oxo complexes of porphyrins by co-
᎐
ordinating axially with the iron.23 Spectophotometric titrations
of TPPFe solutions in the presence of imidazole demonstrated
that the formation of a complex between imidazole and
porphyrin led to a type II reverse spectrum.24,25 Results
obtained by us show that the addition of N,N-dimethylbenz-
amide to a solution of TPPFe has a similar effect on the
spectrum of TPPFe (Fig. 3). N,N-Dialkylamides could thus act
in a manner similar to imidazole facilitating the formation of
98.2
a Conformation of the amide. b Stereochemical relationship of
the radical centre to the carbonyl oxygen atom. c Stereochemical
relationship between the groups on the carbon-centred radical and
those on the amide nitrogen atom.
Table 4 Formation of N-formyl-N-methylbenzamide 3 and N-methylbenzamide 4 in the oxidation of N,N-dimethylbenzamide 1a and of the
possible intermediates 5–7. [Substrate] = 0.5 × 10Ϫ3 M.
Relative yield (%)
Substrate
X
R1
R2
N-formylbenzamide 3
N-methylbenzamide 4
1a
5
H
H
H
H
Me
Me
Me
Me
Me
95
nd
2
5
100
98
CH2OH
CH2OOBut
CH2OMe
6
7
nd
100
nd = not detected.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
1898

View Article Online
Alternatively, the hydroperoxide 9 is an intermediate that
could give rise to the acyl amides. A compound of this type was
found to be formed in the dioxygen oxidation of N-methyl-
pyrrolidone to N-methylsuccinimide.27 However, we consider
this unlikely since the N-acyl products are produced in identical
yields when the reactions were carried out under an atmosphere
of nitrogen. Unfortunately, all our attempts to make compound
9 failed.
Fig. 3 UV-visible spectra of a solution of 2 × 10Ϫ5 M TPPFeIII (a) and
a solution of 2 × 10Ϫ5 M TPPFeIII in the presence of 0.02 M N,N-
dimethylbenzamide (b).
The small amounts of N-alkylamides formed directly in the
oxidation of N,N-dialkylamides can be explained by formation
from the carbon-centred radical and subsequent reaction
ؒ
ؒ
ؒ
of this radical either with tert-ButOO , tert-ButO or HO . All
three possible products, viz N-tert-butylperoxymethyl-N-meth-
ylbenzamide, N-tert-butoxymethyl-N-methylbenzamide, and
N-hydroxymethyl-N-methylbenzamide, lead to the formation
of N-methylbenzamide when subjected to the conditions of the
model oxidation. As N-tert-butylperoxymethyl-N-methylbenz-
amide can be identified in trace amounts in the oxidation of
N,N-dimethylbenzamide we tentatively suggest this compound
as an intermediate in the formation of the N-alkylbenzamides
that appear as minor products in the reaction.
the oxene. Furthermore, we found the addition of water or
imidazole increased the initial rate of formation of N-formyl-
N-methylbenzamide from N,N-dimethylbenzamide (not shown),
and when reactions were performed using iodosobenzene as the
oxygen donor a similar product distribution was obtained
suggesting a common oxene oxidant in these systems.
Mechanism for the formation of N-acyl-N-methylbenzamides
The carbon-centred radical must react to form the N-acyl-N-
alkylbenzamides and the N-alkylbenzamides. However, the
exact mechanism for this is still uncertain. Since none of the
postulated intermediates 5–7 are responsible for the formation
of the N-acylamide, we suspect an intermediate of the type of
8 (Scheme 6), similar to that implicated in the formation of
alkylated porphyrins during the oxidation of alkenes,26 to be
involved.
Experimental
Instrumentation
NMR spectra were recorded using JEOL FX90Q or JNM-
EX400FT spectrometers. UV/vis and IR spectra were recorded
using Shimadzu UV 160 and Nicolet 205 FTIR spectrophoto-
meters, respectively. Mass spectra were recorded with a VG
Mass Lab 25–250 spectrometer. GC-MS was performed using a
Hewlett-Packard 5809-A gas chromatograph connected to the
VG Mass Lab 25–250 spectrometer. HPLC was performed
using a Varian 5000 instrument, a Cecil 2112 variable wave-
length detector and a Spectra Physics Chrom Jet integrator.
Diode array detection was performed using a Varian 9060 Poly-
chrom detector. Elemental analyses were obtained from Medac
Ltd, Brunel Science Centre, Cooper’s Hill Lane, Englefield
Green, Egham, TW20 0JZ, UK.
Reagents
Tetraphenylporphyrinatoiron() chloride was obtained from
Aldrich. N,N-Dialkylamides 1a–h and the product N-alkyl-
amides were synthesised by the Shotten–Baumann28 method
from the appropriate amine and acid chloride. The dialkyl-
amides were purified by column chromatography using ethyl
acetate : hexane 5 : 2 as eluent and the N-alkylamides were
recrystallised from ethanol. N-Alkyl-N-methylamides 1i–l were
synthesised from the appropriate monoalkylamine using n-
butyllithium and iodomethane4 and purified by column
chromatography using ethyl acetate : hexane 5 : 2 as eluent.
Iodosobenzene was synthesised by a published method.29
tert-Butylhydroperoxide was obtained from Aldrich as a 70%
aqueous solution, extracted with dichloromethane, dried (4 Å
molecular sieves) and, after evaporation of the solvent, stored
at Ϫ20 ЊC over 4 Å molecular sieves. Dichloromethane used for
the model reactions was obtained from Merck (GLC grade)
and was distilled before use.
Scheme 6 Possible mechanism for the formation of 3 from 1a.
One can conceive of this intermediate being formed by
reaction between a carbon-centred radical and the oxene FeV᎐O
᎐
present in the medium. This intermediate could form the N-
formylamide and regenerate TPPFeIII to restart the cycle. Either
the tert-butoxyl radical or a complex of the type FeIV–OH
could abstract the hydrogen needed to form N-formyl-N-
methylbenzamide and regenerate TPPFeIII.
N-Methoxymethyl-N-methylbenzamide (7, R ؍ Me). N-
Methylbenzamide (0.018 mol, 2.5 g) and paraformaldehyde (2.3
g) were added to chlorotrimethylsilane (30 cm3) and refluxed.
After 18 h chlorotrimethylsilane was evaporated and formation
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
1899

View Article Online
of N-chloromethyl-N-methylbenzamide was confirmed by
NMR spectroscopy (δH/ppm 3.10 (3H,s), 5.28 (2H,s), 7.45
(5H,m)). This (0.50 g, 0.00273 mol) was used without further
purification and added to a solution of sodium methoxide
(prepared by dissolving metallic sodium (0.25 g) in methanol
(50 cm3)). After 12 h, water 50 (cm3) was added, the methanol
evaporated and the residue extracted with dichloromethane
(3 × 20 cm3). The organic phase was washed with water, (2 ×
10 cm3), dried with MgSO4 and purified by column chromato-
graphy using ethyl acetate : hexane 1 : 1 as eluent. The com-
pound was obtained as an oil in 65% yield: νmax/cmϪ1 2934,
1647, 1446, 1436, 1180, 701; δH/ppm (CDCl3) 3.07 (3H, s), 3.20
(3H, s), 4.59 (2H, s), 7.36 (5H, m); δ13C/ppm (CDCl3) 32.7, 55.1,
82.5, 127.3, 128.4, 128.5, 130.0, 125.6, 135.7, 172.5; m/z (%) 179
(9), 164 (39), 148 (13), 105 (100), 77 (42), 51 (13). C10H13NO2
requires: C, 67.0; H, 7.31; N, 7.89%. Found: C, 67.1; H, 7.38; N,
7.54%.
Mass spectral assay for reaction products
Substrates (usually 0.1 M) were subjected to oxidation as
described above. The reactions were followed by HPLC and
after 2 and 4 h one further portion of anhydrous ButOOH was
added. After 6 h an aliquot of the reaction mixture was diluted
with ethanol and analysed in the GC-MS system using a
25 m × 0.5 mm BP-5 column 25 m, (start temperature 80 ЊC
then 10 ЊC/ min up to 200 ЊC).
Molecular orbital calculations
Ionisation potentials for the dimethylbenzamides were
calculated using the semi-empirical AM1 SCF MO program
within the MOPAC 4.0 package.30 All structures were geometry
optimised using the Broyden–Fletcher–Golfarb–Shanno
approach and performed on a VAX cluster.
N-tert-Butylperoxymethyl-N-methylbenzamide, 6. N-Chloro-
methyl-N-methylbenzamide (0.50 g, 0.00273 mol), synthesised
as described above and used without further purification, was
dissolved in dry THF (20 cm3). tert-Butyl hydroperoxide (0.49
g, 0.00546 mol) was added and the reaction left for 12 h. The
solution was evaporated, the residue dissolved in dichloro-
methane, washed with water (3 × 10 cm3), dried with MgSO4,
concentrated and purified by column chromatography using
ethyl acetate as eluent. The compound was obtained as a solid
in 63% yield: mp 42–45 ЊC; νmax/cmϪ1 2979, 2933, 1651, 1390,
1364, 1285, 1259, 925, 700; δH/ppm (CDCl3) 1.15 (9H, s), 3.17
(3H, s), 5.00 (2H, s), 7.33–7.46 (5H m); δ13C/ppm (CDCl3) 23.2,
33.3, 80.7, 84.3, 126.9, 127.7, 128.0, 129.7, 131.0, 135.9, 172.9;
m/z (%) 238 (6), 181 (2), 148, (92), 105 (100), 77 (36), 51 (14).
C13H19NO3 requires: C, 65.8; H, 8.07; N, 5.90%. Found: C, 65.4;
H, 8.12; N, 5.86%.
References
1 P. J. Robbins and M. G. Cherniack, J. Toxicol. Environ. Health, 1986,
18, 503.
2 J. E. F. Reynolds (Ed.), Martindale, the Extra Pharmacopoeia, 29th
Edn. The Pharmaceutical Press, London, 1989.
3 A. Gescher, Chem. Br., 1990, 26, 435.
4 J. Iley and L. Constantino, Biochem. Pharmacol., 1994, 47, 275.
5 L. Constantino, E. Rosa and J. Iley, Biochem. Pharmacol., 1992, 44,
651.
6 J. Iley, R. Tolando and L. Constantino, J. Chem. Soc., Perkin Trans.
2, 2001, 1299.
7 L. Constantino, PhD Thesis, University of Lisbon, 1994.
8 H. Masumoto, K. Takeuchi, S. Ohta and M. Hirobe, Chem. Pharm.
Bull., 1988, 37, 1788.
9 J. R. Lindsay Smith and D. N. Mortimer, J. Chem. Soc., Perkin
Trans. 2, 1986, 1743.
10 J.-F. Gal, S. Geribaldi, G. Pfister-Guillouzo and D. G. Morris,
J. Chem. Soc., Perkin Trans. 2, 1985, 103.
11 L. Constantino and J. Iley, Proceedings of the 8th International
Conference on Cyt P 450, ed. M. C. Lechner, John Libbey Eurotext,
Paris, 1994, p. 677.
12 T. L. Macdonald, K. Zirvi, L. T. Burka, P. Peyman and
F. P Guengerich, J. Am. Chem. Soc., 1982, 104, 2050.
13 R. P. Hanzlik and R. H. Tullman, J. Am. Chem. Soc., 1982, 104,
2048.
14 H. M. Walborsky, Tetrahedron, 1981, 37, 1625.
15 V. W. Bowry and K. U. Ingold, J. Am. Chem. Soc., 1991, 113, 5699.
16 V. W Bowry, J. Lusztyk and K. U. Ingold, J. Am. Chem. Soc., 1991,
113, 5687.
17 J. T. Groves and E. T. Nemo, J. Am. Chem. Soc., 1983, 105, 6243.
18 D. Mansuy, P. Battioni and J. P. Renaud, Chem. Commun., 1984,
1255.
Substrate oxidations
Amide oxidations were performed in CH2Cl2 using a solution
that was 1 × 10Ϫ3 M in TPPFe and 0.2 M in tert-ButOOH.
Usually, to 1 cm3 of CH2Cl2 containing the required con-
centration of substrate, 0.5 cm3 of a CH2Cl2 TPPFe solution
(4 × 10Ϫ3 M) and 0.5 cm3 of a CH2Cl2 ButOOH solution (0.8 M)
were added. Oxidation of amides 1a–e was performed using
initial substrate concentrations between 0.1 and 0.5 M. Oxid-
ation of other amides was performed using a concentration of
0.5 M. Oxidations performed using iodosobenzene as oxidant
were carried out by adding 4 × 10Ϫ4 mol of iodosobenzene to
2 cm3 of a CH2Cl2 solution containing the substrate and
TPPFe. Oxidation of 1a in the presence of imidazole was
19 D. Mansuy, J. F. Bartoli and M. Momenteau, Tetrahedron Lett.,
1982, 23, 2781.
20 T. G. Traylor, S. Tsuchiya, Y.-S. Byun and C. Kim, J. Am. Chem.
Soc., 1993, 115, 2775.
performed using 10Ϫ2 M imidazole, 0.5 M 1a and 1 × 10Ϫ3
M
TPPFe. All reaction mixtures were incubated at 30 ЊC. At timed
intervals aliquots (0.050 cm3) were removed, diluted with
ethanol 1 cm3and injected into the HPLC (for amides 5, 6 and
7) or diluted with a solution of 0.1 × 10Ϫ3 M NaOH in ethanol,
left for 10 min and then neutralised with HCl before injection
into the HPLC.
21 T. G. Traylor, C. Kim, J. L. Richards, F. Xu and C. L. Perrin, J. Am.
Chem. Soc., 1995, 117, 3468.
22 Y. J. Kim and C. R. Park, Inorg. Chem., 2002, 41, 6211.
23 A. Gold, K. Jayaraj, P. Doppelt, R. Weiss, G. Chottard, E. Bill,
X. Ding and A. X. Trautwein, J. Am. Chem. Soc., 1988, 110, 5756.
24 R. Labeque and L. J. Marnett, J. Am. Chem. Soc., 1989, 111, 6621.
25 M. J. Beck, E. Gopinath and T. C. Bruice, J. Am. Chem. Soc., 1993,
115, 21.
HPLC analysis
26 P. R. Ortiz de Montellano, K. L. Kunze, K. S. Keilan and
C. Wheeler, Biochemistry, 1982, 21, 1331.
Detection of the amides was performed using a 5 µm C-18
25 cm × 5 mm Jones chromatography column and an eluent
consisting of 20% acetonitrile in water for 1a and 1e, 25%
acetonitrile in water for 1b, 27.5% acetonitrile in water for 1d,
30% acetonitrile on water for 1c and a gradient using 15%
acetonitrile in 0.05 M pH 2.6 phosphate buffer at t = 0 to 42%
acetonitrile in 0.05 M pH 2.6 phosphate buffer at t = 22 min for
the other amides.
27 R. S. Drago and R. Riley, J. Am. Chem. Soc., 1990, 112, 215.
28 Vogel’s Textbook of Practical Organic Chemistry, 4th edn., ed.
B. S. Furniss, A. J. Hannaford, V. Rogers, P. W. G. Smith and
A. R. Tatchel, Longman, London, 1978, p. 682.
29 Vogel’s Textbook of Practical Organic Chemistry, 4th edn., ed.
B. S. Furniss, A. J. Hannaford, V. Rogers, P. W. G. Smith and
A. R. Tatchel, Longman, London, 1978, p. 541.
30 MOPAC 4.0 Quantum Chemistry Exchange Program, QCPE
program nЊ 455, Indiana University.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 8 9 4 – 1 9 0 0
1900