The mechanistic origin of regiochemical changes in the nitrosative N-dealkylation of N,N-dialkyl aromatic amines

Article abstract of DOI:10.1039/b418457b

The regioselectivity of the nitrous acid mediated dealkylation of 4-substituted-N-ethyl-N-methylanilines is a function of the acidity of the reaction mixture. At high acidity deethylation predominates, whereas demethylation is the predominant reaction in nitrosamine formation at pH 2 and above. In some cases the regioselectivity of nitrosative dealkylation changes as the run proceeds. Through the use of the corresponding 4-nitroaniline as the primary substrate, CIDNP, kinetics, kinetic deuterium isotope effects and other transformations involving nitrosations with NO₂ or NOBF₄ in aprotic solvents, a new mechanism of tertiary amine nitrosation has been deduced and proposed to explain regioselective deethylation. The mechanism involves the oxidation of the substrate to the amine radical cation by NO ⁺. This is followed by the abstraction of a hydrogen atom from the carbon adjacent to the amine nitrogen by NO₂ to produce an iminium ion which reacts further to produce the corresponding aldehyde and the nitrosamine. Depending upon the acidity, this process competes with three other mechanistic pathways, two of which give the nitrosamine through the iminium ion, and one leads to the formation of C-nitro compounds. The competing pathways to nitrosamine formation involve NOH elimination from a nitrosammonium ion and deprotonation of the radical cation to give an α-amino radical which rapidly oxidized to the iminium ion. Predominant, but not highly regioselective demethylation occurs by these pathways. Nitro compound formation principally arises from the reaction of NO₂ with the radical cation followed by deprotonation, but also occurs by para C-nitrosation followed by oxidation. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b418457b

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
The mechanistic origin of regiochemical changes in the nitrosative
N-dealkylation of N,N-dialkyl aromatic amines†
Emma L. Teuten and Richard N. Loeppky*
Department of Chemistry, University of Missouri, Columbia, MO, 65211, USA.
E-mail: LoeppkyR@Missouri.edu; Tel: 573 882 4885
Received 8th December 2004, Accepted 1st February 2005
First published as an Advance Article on the web 18th February 2005
The regioselectivity of the nitrous acid mediated dealkylation of 4-substituted-N-ethyl-N-methylanilines is a function
of the acidity of the reaction mixture. At high acidity deethylation predominates, whereas demethylation is the
predominant reaction in nitrosamine formation at pH 2 and above. In some cases the regioselectivity of nitrosative
dealkylation changes as the run proceeds. Through the use of the corresponding 4-nitroaniline as the primary
substrate, CIDNP, kinetics, kinetic deuterium isotope effects and other transformations involving nitrosations
with NO2 or NOBF4 in aprotic solvents, a new mechanism of tertiary amine nitrosation has been deduced and
proposed to explain regioselective deethylation. The mechanism involves the oxidation of the substrate to the amine
radical cation by NO+. This is followed by the abstraction of a hydrogen atom from the carbon adjacent to the amine
nitrogen by NO2 to produce an iminium ion which reacts further to produce the corresponding aldehyde and the
nitrosamine. Depending upon the acidity, this process competes with three other mechanistic pathways, two of which
give the nitrosamine through the iminium ion, and one leads to the formation of C-nitro compounds. The competing
pathways to nitrosamine formation involve NOH elimination from a nitrosammonium ion and deprotonation of
the radical cation to give an a-amino radical which rapidly oxidized to the iminium ion. Predominant, but not highly
regioselective demethylation occurs by these pathways. Nitro compound formation principally arises from the reaction
of NO2 with the radical cation followed by deprotonation, but also occurs by para C-nitrosation followed by oxidation.
Introduction
Numerous nitrosamines have been shown to be carcinogenic
at low doses in laboratory animals.1,2 They have been found in
foods, drugs, tobacco and tobacco smoke, cosmetics, personal
care formulations, rubber products, metalworking fluids, and
many other materials to which humans are exposed.3 Often
Scheme 1 Nitrosation of Padimate-O.
they are formed from nitrosating agents of uncertain origin, by
chemical processes which are poorly understood.4 Nitrosamines
The most prominent mechanism of tertiary amine nitrosation,
and other DNA damaging compounds can also be formed by en-
here designated as Mechanism A, is shown in Scheme 2.8,9 It
involves the reversible N-nitrosation of the substrate (3→4)
followed by the elimination of NOH to give an iminium ion 5.
The hydrolysis of the latter and the nitrosation of the resultant
secondary amine 6 lead to the nitrosamine 8. A distinguishing
feature of this process is the formation of N2O, which arises
from NOH as shown. The transformation is perceived to
proceed by the syn cyclic elimination of NOH, a process
which results in the conformational eclipsing of the substituents
on N and those on the carbon from which the H is being
eliminated. Thus the pathway which conformationally leads
dogenous, unsuspected nitrosation transformations in humans.5
In order to better understand the chemistry underlying the
formation of 2-ethylhexyl 4-N-nitrosomethylaminobenzoate 2
from 2-ethylhexyl 4-N,N-dimethylaminobenzoate (Padimate-O)
1 (see Scheme 1), a common sunscreen ingredient, during its pro-
duction and formulation, we initiated an investigation of rates
and mechanisms of N,N-dialkylaromatic amine nitrosation.6
In the course of this work we discovered that regiochemistry
of nitrosative dealkylation of unsymmetrical dialkylaromatic
amines, N-ethyl-N-methyl-4-carboethoxyaniline, for example,
sometimes changed during a run.7 As an illustration, by
following the acidic nitrosation of this amine by HPLC we
found, under certain conditions, that initially N-demethylation
occurred and then after a short time N-deethylation became
dominant. The factors, and the mechanistic changes which give
rise to this unexpected chemistry and the regiochemistry of N,N-
dialkylaromatic amine nitrosative dealkylation are the subject of
this paper. In all but a few cases, a change in the regiochemistry
of a transformation involves a change in mechanism, which
often involves altered rates of reaction. An understanding of the
factors which affect the rates and mechanisms of tertiary amine
nitrosation, particularly as they relate to changes in structure
or reaction conditions, are important in preventing or limiting
human exposure to potentially carcinogenic nitrosamines.
† Electronic supplementary information (ESI) available: Explanations
of CIDNP, and additional experimental details for kinetics experiments,
syntheses of several known compounds, and N2O determinations. See
http://www.rsc.org/suppdata/ob/b4/b418457b/
Scheme 2 Mechanism A: NOH elimination.
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 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 0 9 7
©

View Article Online
to the lesser “crowding” of the transition state is perceived to
control the regiochemistry of the transformation. This concept
is well supported by a number of intramolecular competition
experiments. The transformation proceeds predominantly by
cleavage of the least sterically hindered substituent from N.8
heating the tertiary amine with 4 equivalents of butyl nitrite,
one equivalent of H2O and 0.1 equivalent of NH4Cl at reflux
in an inert atmosphere. This process avoids the production of
significant quantities of nitro compound and exhibits a regios-
electivity for N-dealkylation of 74–100%.12 Except for benzyl
which is selectively cleaved, the smaller alkyl group is selectively
removed just as it is in the nitrous acid nitrosative dealkylations.
However, the mechanism of the transformation under the Ver-
ardo conditions does not appear to be known.12,13 On the other
hand, Hodgson and Nicholson reported in 1941 that the major
nitrosamine arising from the reaction of N-ethyl-N-methyl-
4-nitroaniline 15 with sodium nitrite in concentrated HCl is
N-methyl-N-nitroso-4-nitroaniline 17 (deethylation favored).14
Verardo et al. reported that the nitrosation of 15 under their
conditions resulted in 74% demethylation.12 Verardo et al. used
modern analytical instrumentation, which, of course, Hodgson
and Nicholson did not have at their disposal.
Because of this, we initially sought to verify the early
workers reported regioselectivity as a starting point toward
understanding why we occasionally observed alterations in
the regioselectivity of nitrosative N,N-dialkyl aromatic amine
dealkylation, and how these changes may relate to the regio-
chemical proclivities of mechanistic paths A and B, the latter
of which was unknown at the inception of this research. We
report here that within the limits of our experimentation (i.e.
demethylation vs. deethylation) that Mechanism B proceeds with
preferential cleavage of the less substituted group from N, as does
Mechanism A. Regioselectivity, however is linked to the acidity
of the reaction media. Deethylation is overwhelmingly preferred
at high acid strength. We propose that this results from the
incursion of another radical cation mediated mechanism where
the a-hydrogen atom of the N-alkyl substituent is removed by H-
atom abstraction rather than mild base induced deprotonation
(Mechanism B).
K
N
H+ + NO− ꢀ HNO2
2HNO2 ꢀ N2O3 + H2O
(1)
(2)
2
N2O3 ꢀ NO + NO2
(3)
K
d
HNO2 + H+ ꢀ NO+ + H2O
(4)
Our investigation of the mechanisms of N,N-dialkylaromatic
amine 9 nitrosation revealed that nitrosamines were also com-
petitively arising by at least one pathway which does not generate
N2O as a co-product.10 In this case an aromatic nitro compound
is formed as a reaction by-product. The nitrosamine/nitro
compound ratio could be manipulated by changes in reaction
conditions and we determined that there was a linkage in the
paths leading to these products. The pertinent “nitrous acid
chemistry” is given in eqns. (1)–(4). Extensive investigation
involving 15N-CIDNP NMR, kinetics, deuterium isotope effects,
transformations employing a confined headspace and/or added
NO, NO2, O2, or N2, and reactions of prepared aromatic
amine radical cations with NO, NO2, and NO2−, led us to
propose that this chemistry was occurring through a radical
cation intermediate 11 (Mechanism B, Scheme 3).10,11 The
radical cation forms reversibly from the nitrosammonium ion 10.
Deprotonation to 12 followed by rapid oxidation results in the
generation of the iminium ion 13, which gives the nitrosamine
14 by the same pathway as Mechanism A. Thus mechanisms
A and B differ only in the way in which the iminium ion is
formed. This difference could, of course, result in differences
in the regioselectivity of N-dealkylation. In related work, we
demonstrated that the nitrosation of N-alkyl-N-cyclopropyl
aromatic amines resulted in exclusive cleavage of the cyclopropyl
substituent from nitrogen, a transformation which is indicative
of the intermediacy of amine radical cations.11 Other than
this, however, we did not define the regiochemistry of N-alkyl
cleavage by Mechanism B. As stated above, the regiochemistry of
cleavage was sometimes found to vary as the run proceeded, and
in other experiments we determined that nitrosamine formation
by Mechanisms A and B were always competing to variable
extents depending upon reaction conditions and substrates. In
almost all cases, however, at the end of the run, the nitrosamine
which resulted from cleavage of the least sterically hindered
group from N predominated.
Results and discussion
Preliminary experiments showed that the nitrosation of N-ethyl-
N-methyl-4-nitroaniline 15 (see Scheme 4) with 5 eq. of NaNO2
in 50% HCl, the conditions utilized by Hodgson and Nicholson,
resulted in preferred deethylation to give 17 in preference to 16 as
they had reported.14 On the other hand, nitrosation of the same
amine in glacial or aqueous acetic acid, conditions frequently
used in our work, resulted in preferred demethylation. As a
result we examined the regiochemistry of this transformation
more carefully as a function of acidity. The data of Table 1
clearly show that the preference for deethylation increases with
acidity of the medium. The % deethylation (% deEt) in Table 1
and elsewhere in this presentation is given on a per hydrogen
atom basis and has been corrected for the differing number
Verardo et al. have developed a preparative method for the
production of nitrosamines from dialkyl aromatic amines by
Scheme 3 Mechanism B: deprotonation of radical cation.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
Scheme 4
1 0 9 8

View Article Online
Table 1 The effect of acidity on the regiochemistry of nitrosative
dealkylation of N-ethyl-N-methyl-4-nitroaniline 15
were: 17, 45.7 3.3%, CH3CHO 43.4 3.2%, 16 12.7 1.7%,
CH2O 11.7 1.5%. These data show that the loss of the alkyl
group can be quantitatively accounted for in the corresponding
aldehyde products.
Runa pH/Ho % deEtb,c,d Acid mixture
In our prior work we showed that the contributions of
mechanistic paths A and B could be distinguished by comparing
the nitrosamines yields to the N2O yields.10 Pathway A produces
0.5 mole of N2O for every mole of nitrosamine, while N2O is
not a product of nitrosamine production by Mechanism B. As
a result we next sought to use this technique to determine
how the contribution of Mechanism A changed as the acidity
of the mixture was increased. Using a previously developed
calibrated procedure, N2O yields were determined by measuring
the headspace N2O concentration by GC equipped with an
electron capture detector.10 Nitrosamine yields were determined
by HPLC. The data are given in Table 2. The contribution
of Mechanism A decreases significantly as the acidity of the
reaction mixture is increased, and, as seen already, the % deEt
increases with acidity. These transformations were conducted
in 75% HOAc and HCl was added to increase the acidity.
A comparison of runs 3 and 4 (Table 2) shows that halving
the substrate concentration makes little difference to the %
deEt. It is important to note that run 5 was done with the
deuterated substrate N-(1ꢀ1ꢀ-D2-ethyl)-N-methyl-4-nitroaniline
(15-D2). As expected, deuteration at the ethyl group results in
some mechanistic switching and a decrease in the % deEt under
the same conditions (run 4).
The data presented so far indicate that a different mechanism
of nitrosative dealkylation is responsible for the increased %
deEt with increased acidity of the reaction media. While it may
be tempting to ascribe this to the sole incursion of mechanism
B, literature data on the regiochemistry of N-dealkylation
of chemically or electrochemically generated aromatic amine
radical cations are inconsistent with such a supposition. These
transformations result in the preferential, but not exclusive
removal of the least substituted alkyl group from N by a
process which involves a-deprotonation of the radical cation by
a weak base through a process akin to that shown in Scheme 3,
Mechanism B.
1
2
3
4
5
6
7
8
2
1.5
1.1
0.3
−0.2
−0.5
−0.7
−0.8
28.6
41.2
33.3
63.0
75.6
81.8
84.8
87.8
75% HOAc–2.7% HClO4–0.7 M NaOAc
75% HOAc
75% HOAc–0.2 M NaCl
75% HOAc–5% HClO4/0.8 M NaOAc
75% HOAc–1 M HCl
75% HOAc–5% HClO4/0.7 M NaOAc
1.1 M HCl
75% HOAc–5% HClO4
a [15]i = 11.5 mM, [NaNO2]i = 0.115 M, 36 min reaction time, 23 ◦C.
b deethylation c corrected for the number of CH3 vs. CH2 a-H-atoms.
d Extents (%) of reaction varied and N-ethyl-N-methyl-2,4-dinitroaniline
18 was also a product in runs 4–8.
of a-H on Et vs. Me. The effect produced by changes in acid
strength on the regiochemistry of N-dealkylation is remarkable
and, to the extent of our knowledge, unprecedented in this kind
of chemistry. Since all reactions were stopped at 36 min., the
acidity of the mixture did affect the extent of the reaction,
which increased from around 4% (runs 1–4) at lower acidity
to 70–90% (runs 5,6, and 8). Thus, with the exception of the
run in 1.1 M HCl (all other runs utilized 75% HOAc as a
base “solvent”), increased acidity also increased reaction rate.
These experiments utilized various acid counter ions, OAc−, Cl−,
and ClO4−, not only to manipulate the acidity of the mixture
but to determine whether they had any dramatic effect on the
reaction regiochemistry. While we did not systematically pursue
the question because the primary experimental determinant of
regiochemistry appeared to be acidity, the only effect noted is
seen in a comparison of runs 2 and 3 where the addition of 0.2 M
Cl− to 75% HOAc decreased the pH but decreased the % deEt by
8% instead of increasing it, as we observed for other increases in
acidity. No other significant counter ion effects were observed.
As they are in other nitrosative dealkylation reactions,8,10
aldehydes are produced from the cleaved alkyl groups under
the high acid conditions which give rise to preferential deethy-
lation (3.6 M H2SO4 in 90% acetic acid). Formaldehyde and
acetaldehyde, determined as their 2,4-dinitrophenyl hydrazones
(DNPs), are the respective products of methyl vs. ethyl cleavage.
After accounting for the extraction efficiency of DNPs under the
reaction conditions, the yields of the aldehydes and nitrosamines
The resulting radical is rapidly oxidized to the iminium ion. In
order to determine the regiochemical N-dealkylation preference
for the radical cation derived from 15 under various conditions,
we examined the nature of the products arising from both
chemical and electrochemical oxidation of the parent amine
under various conditions. The data are given Table 3. The
Table 2 Comparison of nitrosamine and N2O yields with acidity for the nitrosation the of 15
Run
Acid in 75% HOAc [15]/mMi
Rxn. time/min % N2O
% 17
% 16
% rxna
% deEt
% Mechanism Ab
1
2
3
none
11.5
11.5
11.5
5.9
36
36
36
21
21
0.7 0.1
2.9 0.2
0.6 0.1 1.2 0.2
7.7 0.2 4.0 0.2 32.4 1.0 74.3
3.6 0.2 42.9
80
50
13
9
7
4
1
1
1
0.1 M HCl
1.0 M HCl
1.0 M HCl
1.0 M HCl
1.2 0.6 14.6 1.4 4.5 0.4 71.1 2.3 83.0
0.6 0.1 10.7 0.8 3.2 0.4 39.5 2.8 83.4
0.5 0.1
4
5c
5.9
2.2 0.2 2.8 0.1 24.2 3.7 54.1
19
a Total conversion of starting material into products. The unlisted product is 18. b Percent of nitrosamine formed by Mechanism A = 200[N2O]/([16]
+ [17])%. c Data for the nitrosation of the deuterated substrate N-(1ꢀ1ꢀ-D2-ethyl)-N-methyl-4-nitroaniline (15-D2).
Table 3 Product distribution from the oxidation of N-ethyl-N-methyl-4-nitroaniline 15
Oxidation methoda
Solvent
[20]/[19]
% deEtb
% rxn
Ce(NH4)2(NO3)6
Ce(NH4)2(NO3)6
Ce(NH4)2(NO3)6
Electrochemical
Electrochemical
Electrochemical
75% HOAc
0.18
0.33
0.32
0.3
0.36
0.54
21.3
33.1
32.4
31.0
35.1
44.8
15.5
6.6
1.6
3.4
0.6
75% HOAc–5% HClO4–0.8 M NaOAc
75% HOAc–5% HClO4
75% HOAc
75% HOAc–5% HClO4
Acetonitrile–NaBF4 (sat.)
0.94
a [15]i = 2.6 mM. b Corrected for the number of H atoms on the respective substituents.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 0 9 9

View Article Online
Scheme 5
oxidations were conducted as described in the Experimental
section and amines were analyzed by HPLC.
their conditions is from 21 to 15 and 22 and then to 16–18.
We also have confirmed the preference for cleavage of the ethyl
group, as noted above. The % deEt for the nitrosation of each
substrate (21, 15 and 22) was also comparable in 75% HOAc–
1 M HCl using a 10 fold excess of NaNO2 (8, 63%; 10, 64%; 15,
67%).
All of the oxidation processes listed in Table 3 are known to
proceed by one electron loss followed by deprotonation of the
a-carbon of the alkyl group.15–17 These hydrogen atoms are rela-
tively acidic, depending upon the aromatic ring substituent.18,19
The resulting carbon free radical 12 suffers a second one electron
oxidation to generate the iminium ion 13, which hydrolyses
in the media to give one of the secondary amines (Scheme 5,
19 or 20). The data show, in concert with literature precedent,
that demethylation is preferred regardless of the acidity of the
reaction media.19 Thus this pathway cannot be leading to the
regiochemical preference for deethylation which we observe
at high acidity. This does not mean, however, that aromatic
amine radical cations are not involved in the reaction leading
to preferred deethylation. In order to test for the possible
involvement of aromatic amine radical cations, and to gain
more information on the pathway to preferred deethylation we
performed a number of experiments.
In our prior work where we showed that aromatic o-nitro
compounds arise in reactions of 4-substituted aromatic amines
by radical cation–NO2 recombination,10 15N-NMR CIDNP
experiments were critical in reaching these conclusions. The
presence of radical cations in the media, as detected by these
nitration reactions, was a strong indicator of their role in
nitrosamine formation by Mechanism B. Accordingly, we
examined the reaction of 21 with acidic Na15NO2 by time-based
15N NMR. Enhanced emission signals (CIDNP) were observed
for 15 and 23 during the first 3 min of the reaction (Fig. 1), which
is indicative of their formation by a process involving radical
recombination of the amine radical cation 11 and NO2. A de-
tailed explanation of the emission using the Kaptein equation,26
which is essentially the same as that published by us and
others,27–31 previously, is given in the Supporting Information.†
Numerous investigations have shown that the nitrite catalyzed
or nitrite mediated nitration of phenols and aromatic amines
does not involve the nitronium ion NO2+, but proceeds by
radical recombination.32,33 Thus, the p-nitroamine 15, and the
o-nitroamine 23 are arising by radical recombination, yet we
did not see any signals for the formation of the dinitroaniline
18, which would arise from the radical cations derived from
either 15 or 23. This was true also when 15 alone was used as a
substrate. This may be a consequence of the very short life time
of the 4-nitrodialkylaniline radical cation due to the electron
15N CIDNP NMR experiments
Hodgson and Nicholson investigated the nitrosation of N-
ethyl-N-methylaniline 21 (see Scheme 6).14 After repeating the
nitrosation of 21 under the conditions described in the original
literature report (50% HCl, 5 equivalents NaNO2), we separated
six products by flash column chromatography and identified
1
them, by GC-MS and H NMR. All of the compounds are
known. As can be seen from the yields given in Scheme 6,
determined by HPLC after a reaction time of 4.5 hours, the
major products are the nitro compounds 15 and 23. Our material
balance is 93%. As shown in Scheme 4, nitrosation of 15 gives
the nitrosamines 16 and 17 and the dinitro compound 18.
Nitrosation of the C-nitroso compound 22 results in its oxidation
to 15 and thence conversion to 16–18. There is extensive
literature precedent for the conversion of p-nitrosoanilines to
their nitro derivatives under nitrosating conditions,20 although
this is probably not the major pathway to 15, 23, or 18. We
have previously demonstrated that aromatic o-nitro compounds
arise in reactions of 4-substituted aromatic amines by radical
cation–NO2 recombination.10 Several recent reports support our
conclusions.21–25 We have shown, as suspected by Hodgson and
Nicholson, that the progression of product formation under
Fig. 1 15N NMR CIDNP spectra for the reaction of 21 with Na15NO2
in 83% HOAc–0.6 M HCl. Line assignments: a, H15NO2; b, 17; c, 22; d,
Ph15NO2 standard; e, 23; f, 15.
Scheme 6 Reinvestigation of Hodgson–Nicholson nitrosation.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 1 0 0

View Article Online
withdrawing destabilizing effect of the nitro group, and its very
rapid recombination with NO2 to give 18. Yet, given what we
know about the mechanisms of nitro compound formation
under these reactions conditions it is highly probable that 18 is
formed from the radical cation of 15 by reaction with NO2.10,21–25
We did not observe 15N CIDNP spectra for the formation
of the 4-nitrosoaniline 22, which is likely formed by a classical
electrophilic aromatic nitrosation mechanism. Surprisingly, no
evidence exists for ortho C-nitrosation of aromatic amines.10 This
may be due to reversibility coupled to rate limiting deprotona-
tion of the Wheland intermediate.34–36 o-Nitrosodialkylanilines
are unknown and we have been unable to make them. It is
possible, but unlikely that they are involved in the chemistry
we are observing.37
is not rate determining (first order in amine). We propose that
NO+ is the active “nitrosating agent”, as we discuss further
below. While acetyl nitrite could be the nitrosating agent, the
careful kinetic work of Casado and his colleagues using amines
of comparable basicity to ours, suggest that this is not likely.39
Acetyl nitrite has only been shown to be important in acetic
acid nitrosations when the amine concentration is very high
compared to the initial [NO2−].
The KDIE experiments were conducted in two ways. Rates
were determined by following the change in UV absorbance
as a function of time, and by using HPLC to follow the time
course of the loss of amine and the appearance of each product.
The rate constants determined for amine loss by either method
are comparable and yield KDIE data which are essentially the
same. Neither reaction in 82% HOAc–3.3 M H2SO4, preferred
deethylation, (kH/kD = 1.3 0.23,), nor reaction in 92% HOAc,
Kinetics and deuterium isotope effects
preferred demethylation, (kH/kD = 1.03
0.23) results in a
significant primary deuterium isotope effect for the loss of
the amine. These are the mechanistically significant KDIE
observations and show that the a-CH bond of the ethyl group
(the only one tested here) is not broken in the rate determining
step. Entries 1c and 2c of Table 4 reveal a significant isotope
effect for deethylation as measured by the formation of the
nitrosamines, but these are “product deuterium isotope effects”
that result from product partitioning after the rate determining
step. a-Deuteration of the ethyl group results in increased
demethylation (the % deEt changes from 88% to 59% with
deuteration), and increased nitro compound formation while
maintaining the same approximate rate of amine loss, showing
a linkage between dealkylation and nitro compound formation
such as we have observed before.10 The KDIE data presented
here contrast with those determined for the nitrosation of 4-
chloro-N,N-dimethylaniline (kH/kD = 4.38, 21 ◦C, 60% HOAc,
pH 3.4) and 4-carboethoxy-N,N-dimethylaniline (kH/kD = 3.69,
21 ◦C, 60% HOAc, pH 3.4) where the transformation was
occurring by a combination of mechanisms A and B.10 These new
data strongly suggest that preferential deethylation is occurring
by a different mechanism.
To further explore the role of possible intermediates in the
preferential deethylation of 15 we turned to kinetics and kinetic
deuterium isotope effect (KDIE) experiments. The nitrosation of
15 in 75% HOAc–3.3 M H2SO4, where the nitrosation proceeds
with a preference for deethylation (83% deEt), is first order in
both the amine and in NaNO2 (rate = kobs[Amine][NaNO2]n,
where n = 1.15 0.14). Kinetics were determined at variable
but known excess [NO2−]. To further investigate the nitrosation
mechanism giving preferential cleavage of the larger alkyl group,
KDIEs were determined for the nitrosation of 15 and 15-D2.
The deuterated compound was prepared by mild reduction of
N-methyl-4-nitroacetanilide with lithium aluminium deuteride.
The relevant kinetic data are given in Table 4. Isotope effects
were determined both under conditions where demethylation
dominated (92% HOAc), Table 4 entries 2a–c, and where deethy-
lation dominated (82% HOAc–3.3 M H2SO4), Table 4; entries
1a–c. We give the rates and isotope effects observed for loss of
15 and the formation of nitrosamines 16 (demethylation) and 17
(deethylation). The other major product in these reactions is the
dinitro compound 18, for which we have not given the rate data.
Evaluation of the data in Table 4, as well as comparison
of it with those we8,10 have previously obtained for similar
substrates reveal several important features. Most tertiary amine
nitrosation reactions investigated so far exhibit kinetics where
the reaction is first order in amine and second order in
[NO2−].9,10,38 The second order [NO2−] term arises because N2O3
is the effective nitrosation agent. The nitrosation of 15 in 75%
HOAc–3.3 M H2SO4 is first order in [NO2−]. We noted above
that the extent of the reaction increased with increasing acidity
(Table 1). Here we see quantitatively by comparing the rate
constants for loss of the amine, entries 1a and 2a of Table 4,
that the nitrosation in the stronger acid is 27 times more
rapid. This is a highly unusual phenomenon. Normally amine
nitrosation rates decrease with decreasing pH below pH 3.4 due
to the protonation of the amine which “protects” it against
nitrosation. Thus increasing the acidity increases the rate and
changes the regioselectivity to preferred deethylation. The fact
that the reaction is first order in NO2− indicates that N2O3 is not
the nitrosating agent, but the rate of nitrosating agent formation
The possible role of NO+
Many nitrosation reactions at high acidity are known to involve
NO+ as the active nitrosating agent.38,40 It is formed from nitrous
acid by protonation of the OH oxygen to generate the nitrous
acidium ion which then dissociates to NO+ and H2O [see eqn.
(1) to (4)]. The reaction kinetics (first order in [NO2−]) for
the nitrosation of 15 are consistent with the hypothesis that
NO+ is the nitrosating agent. Using literature data41 for various
equilibrium constants we calculated the [NO+] as a function of
the acidity of the mixtures reported in Table 1 where the % deEt
is demonstrated to increase with increasing acidity. The relevant
plots are shown in Fig. 2 and clearly demonstrate that the extent
of deethylation increases with the [NO+], which increases with
increasing acidity.
To further explore the possible role of NO+ in the changing
regiochemistry of the reaction we examined the reaction of
Table 4 Rate and KDIE data for the nitrosation of 15
Entry
Compounda
Solvent
kH/10−5 s−1b
kD/10−5 s−1b
kH/kD
1a
1b
1c
2a
2b
2c
15c
16
17
15c
16
17
82% HOAc, 3.3 M H2SO4
82% HOAc, 3.3 M H2SO4
82% HOAc, 3.3 M H2SO4
92% HOAc
92% HOAc
92% HOAc
263 42
202 16
1.30 0.23
0.93 0.20
4.83 0.63
1.03 0.07
0.95 0.22
5.99 1.04
4.36 0.44
21.7 1.3
9.60 0.61
5.59 0.80
0.73 0.05
4.70 0.91
4.47 0.51
9.35 0.33
5.92 1.08
0.19 0.02
a The unlisted product is 18. b Rate of loss of 15, rates of formation of 16 and 17. c [15]i = 5.9 mM, [NaNO2] = 59 mM.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 1 0 1

View Article Online
NO formed to NO2. Its yield was increased significantly by the
addition of a small amount of H2O (compare runs 3 and 4),
and by the use of less NOBF4 compared to amine (runs 3 and 4
compared to runs 1 and 2).
Several pieces of evidence presented above now converge to
strongly support a role for NO+ in the changing regiochemistry
of N,N-alkyl aromatic amine nitrosative dealkylation. Unlike
most amine nitrosation reactions the rate of the transformation
increases at higher acidity where the NO+ concentration is
greatest. The reaction kinetics show that the reaction is 1st order
in “NO2−” as is well established for transformations involving
NO+, but not N2O3, the main nitrosating agent at modest
acidities, which requires second order kinetics in “NO2−”.
Both the nitrosation reactions at high acidity and thence high
relative [NO+] and the reactions with NOBF4 result in highly
regioselective deethylation of 15.
k
1
Am + NO+
AmNO+
(5)
ꢀ
k
−1
k
2
AmNO+ → Am•+ + NO•
(6)
(7)
d[Am]
k1k2KaK4[H+][AmT][NO−2 ]
−
=
dt
(k−1 + k2)KN[H2O]
At this point it is helpful to reconcile our rate data with
a derived rate equation. For the purposes of clarity, we only
consider the process which is giving predominant deethylation,
which occurs at high acidity. We propose the operation of the
simple scheme illustrated by eqns. (5) and (6), which gives rise
to the rate eqn. (7), where Am is the amine, AmNO+ is some
complex of NO+ and the amine, Ka is the dissociation constant
of the protonated amine (AmH+), AmT is the analyzable
concentration of the amine = [Am] + [AmH+], and KN and K4
are defined by eqns. (1) and (4) respectively. We have assumed
that Ka ꢁ [H+] and that [AmNO+] is at steady state. This
rate equation is in complete agreement with our experimental
observations, those being that the reaction is first order in amine
and NO2−. The Equation also predicts that the transformation
will become more rapid as the acidity increases. Depending upon
the relative magnitudes of k−1 and k2, either the nitrosation step
or the decomposition of AmNO+ could be rate limiting. As we
discuss below, we believe it is that latter process.
Fig. 2 Panel A: The calculated variation in [NO+] with acidity is shown.
Panel B: The experimental variation in the nitrosamine ratio (17/16)
(deethylation/demethylation) with acidity is shown. The acid mixture at
each point is given.
We propose that NO+ is acting as a one electron acceptor
and is oxidizing the aromatic amine to a radical cation similar
to the process depicted in Mechanism B. However, we do not
propose that NO+ itself is involved in the product determining
step which is giving rise to the altered regiochemistry where
preferential deethylation is observed. We have come to this
conclusion through several lines of evidence and reasoning. The
oxidation potential of 15 can be estimated from its dimethyl
analog, for which two values have been reported, E0ox = 1.57 V
vs. NHE21 and E0ox = 1.43 V vs. NHE,15 and its oxidation by NO+
(E0red = 1.51 V vs. NHE42) is certainly plausible. The reaction
of NOBF4 with N-cyclopropyl-N-methyl-4-chloroaniline 24
gives exclusive cleavage of the cyclopropyl group to give 4-
chloro-N-methyl-N-nitrosoaniline 25 as the sole product in
27% yield, with 56% recovery of the starting material. This
NOBF4 with 15 in CH3CN under several sets of conditions.
The data are given in Table 5. In runs 1 and 2 the goal was to
determine the extent of deethylation. In all cases the % deEt
is 85–87% as it is when the transformation is conducted at the
highest acid concentrations. We were somewhat surprised to
find that the reaction proceeds with predominant deethylation
because we anticipated that Mechanism A, which involves NOH
elimination, would play a larger role. To test this, we measured
the % N2O formed and compared its yields to those of the
nitrosamines as we had done in the acidic nitrosations. The data
for runs 3 and 4 show that only 3–4% of nitrosamine formation
is occurring by Mechanism A. It is noteworthy that the dinitro
compound 18 is formed in all reactions despite the fact that
we took precautions to exclude O2, which would oxidize any
Table 5 Products from the reaction of 15 with NOBF4
Run
Conditions
% deEta
% Mechanism Ab
%17
%16
%18
%15
%20
% N2Oc
1
2
3
4
Dry CH3CNd
Dry CH3CNd
Dry CH3CNf
Wet CH3CNf,g
85.5
85.1e
85.8
86.7
—
—
4.4 0.9
3.5 1.3
38.5
34.3
9.7
9.8
10.4
2.4
26.7
20.2
47.9
61.0
0.6
20.4
28
—
7.2
—
—
—
—
0.3
0.1
5.2
1.2
5.7
a Corrected for the number of a-H atoms. b Percent of nitrosamine formed by path A = 200[N2O]/([16] + [17])%. c Determined by headspace analysis.
d 0.012 M 15 with 5.3 equivalents of NOBF4, under N2. e The isolated secondary amine was incorporated into the calculation of the % deEt. f 0.0088 M
15 with 2.5 equivalents of NOBF4, under N2. g 1 lL H2O added to simulate conditions of N2O headspace calibration (see Experimental section).
1 1 0 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8

View Article Online
suggests formation of a radical center on the amine nitrogen;
followed by a rapid ring opening and ultimately cleavage
of the cyclopropyl substituent.10,11 We have observed this for
the aqueous nitrosation reactions of other N-cyclopropyl-N-
alkyl aromatic amines, previously.10,11 It is important to note
that the ring opening of the N-cyclopropyl substituent in the
radical cation is so fast as to preclude recombination of the
radical cation with NO2 to generate aromatic nitro compounds.
However, in both the acidic nitrosation reactions and in the
NOBF4 reactions in CH3CN we do observe nitro compound
formation in competition with nitrosative dealkylation. These
observations, as well as the 15N-CIDNP data, strongly imply a
role for an amine radical cation in these transformations and
NO+ is a logical oxidant.
preclude the operation of this pathway at higher acidity or with
NOBF4? The most obvious answer, is that the “effective base”
concentration is very low at high acidity and in the NOBF4
chemistry. A weak base such as a weakly basic anion, or H2O
must be present in sufficient concentration to accept the proton
in the acid–base reaction at the a-position of the N-alkyl group.
These hydrogens are known to be acidic,18,19 but this reaction
must be increasingly precluded as the acidity increases. As a
result of these arguments, we have been required to consider
other reactants and pathways of N-dealkylation. The presence
of nitro products suggests a possible role for NO2 in the N-
dealkylation chemistry.
Nitrogen dioxide
Using Marcus theory (see Supporting Information†) we have
performed calculations which indicate that electron transfer
from the aromatic amine system to NO+ will occur by an
inner sphere rather than by an outer sphere process. A priori,
there would appear to be only three types of molecular sites
in the aromatic amine for transient covalent bond formation
between that moiety and NO+, the nitrogen unshared pair, one
of the ortho ring carbons (r-complex), or p-coordination with
the aromatic ring face. By examining the reaction of NO with
aromatic amine radical cations and related transformations,
we previously produced evidence for the reversible homolytic
dissociation of the nitrosammonium ion.10 However, we always
observed competitive nitrosative dealkylation by NOH elimina-
tion in these processes (competition between mechanistic paths
A and B). Here we observe little reaction by NOH elimination
in either very strong acid or with NOBF4. This suggests that if
the nitrosammonium ion 10 forms by NO+ coordination of the
N unshared pair, and that it dissociates to the radical cation and
NO nearly exclusively, which seems rather implausible. While
we certainly cannot rule out this reaction channel, it seems
likely that the electron exchange in more probably occurring by
through homolytic dissociation of either the p- or r-complexed
intermediates mentioned above to the amine radical cation and
NO.
We examined the reaction of 15 with NO2 in three different
solvents, CH3CN, glacial HOAc, and CH3CN containing a
small amount of H2O. The product data are given in Table 6.
As can be seen, in CH3CN the first formed product is the
dinitroaniline 18 which then reacts further to give N-methyl-2,4-
dinitroaniline 26 by deethylation and N-ethyl-2,4-dinitroaniline
27 through demethylation (entries 1–3). The data show, as we
found with the nitrosation at high acidity, and with NOBF4, that
the regioselectivity is high and that the transformation proceeds
with predominant deethylation. The % deEt does not change
with time (compare entries 2 and 3, and 5 and 6). A high degree
of deethylation is even seen in glacial acetic acid (entry 4). On
the other hand a dramatic difference is seen for the reactions
in CH3CN–2% H2O (entries 5 and 6). Here the predominant
reaction is demethylation! We propose that this shift in reaction
regiochemistry results from the presence of the base, H2O, which
is able to deprotonate the amine radical cation by Mechanism
B or produce the nitrosamine by Mechanism A. The formation
of 18 is indicative of a radical cation intermediate. The radical
cation could form by one of the routes described above for the
acidic nitrosation reactions. There is evidence for the formation
of NO+ from NO2 via N2O4 as shown in eqn. (8). Evidence
for the dissociation of N2O4 as shown comes in the form of
bands observed for NO2, NO+ and NO3 in Raman43 and
−
Another role for NO+ could possibly involve direct hydride
abstraction of the a-H of the N-alkyl group to form the iminium
ion 13. This process would almost certainly give rise to a large
KDIE for amine loss by deethylation, which is not observed, and
the formation of NOH, which does not accompany preferred
deethylation. (a-Deuteration of the ethyl group does result in
a small increase in N2O production (Table 2, entry 5) and a
small KDIE (1.3, Table 4, entry 1c) because of an increased
contribution of Mechanism A). Thus we can exclude hydride
abstraction as the principal pathway giving rise to preferred
deethylation. These arguments support a role for NO+ in the
production of the radical cation, but do not lead us to conclude
that it is directly responsible for the altered regiochemistry. In
our prior work we invoked the formation of a radical cation
from the homolysis of nitrosammonium which was produced
from a NO+ donor (likely N2O3), not NO+ itself. Dealkylation
was then proposed to occur by deprotonation of the alkyl group
in the product determining step (Mechanism B).10 We have
demonstrated that this does not lead to preferential deethylation.
What is different about the current system which seems to
IR44,45 spectra. Nitrogen dioxide bubbled into a solution of
18-crown-6 in CH2Cl2 has been isolated and characterized as
the [NO+.18-crown-6 (NO3)2−] salt.46 Formation of nitrosamines
from the reaction of NO2 with secondary amines has been well
documented.47–50 Nitrosation of tertiary amines by NO2, has also
been mentioned,51 but there is little discussion in the literature of
the mechanism by which this transformation occurs. Because we
observe the formation of nitro compounds in all transformations
where the regioselectivity is highly biased toward deethylation,
and because nitration in our systems involves NO2, we believe
it is logical to consider that NO2 may play an important role
in altering the regiochemistry of the N-dealkylation toward
preferred deethylation.
2NO2 ꢀ N2O4 ꢀ NO+ + NO3
(8)
+
From a physical organic chemical perspective, a transforma-
tion that removes a hydrogen from the a-carbon of the N-alkyl
substituent of a radical cation must be endothermic, and must
involve the development of electron deficiency at that carbon.
Table 6 The reaction of NO2 with 15 in various solvents
Entry
Solvent
Rxn. time/min 18
17
16
26
27
% deEt
1
2
3
4
5
6
CH3CN
CH3CN
CH3CN
HOAc
CH3CN–2% H2O
CH3CN–2% H2O
1
3
20
20
1
95.9
76.5
—
78.5
1.3
15.7
—
—
—
6.8
2.2
4.7
—
—
—
2.7
7.1
—
3.5
70.8
—
—
—
0.6
13
—
—
—
—
89.2
90.0
78.9
31.0
31.0
20
17.5
—
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 1 0 3

View Article Online
In the hydrogen yielding ethyl group, the developing electron
deficiency at its a-carbon is inductively stabilized by the adjacent
CH3, compared to the same process at the N-CH3. Endothermic
H atom abstractions by radicals are well known to proceed
with a high degree of regioselectivity. Consider, for example,
the Br-atom abstraction of a hydrogen from an alkane. The
hydrogen abstraction step is endothermic. The transition state
structure resembles the product electron deficient radical, and a
high degree of regioselectivity is observed where the H-atom is
removed from the carbon most able to stabilize the developing
electron deficiency, normally the most highly substituted one. In
our system H atom abstraction from the a-carbon of the radical
cation results in the formation of a carbocation or more precisely
an iminium ion, a carbocation stabilized by resonance donation
of the unshared pair on N.
Nitrogen dioxide arises from nitrous acid through the for-
mation and homolysis of N2O3 and by the air oxidation of
NO. It is known to abstract hydrogen atoms from a variety
of substrates,52,53 and it is plausible that it does so in this
reaction. The low homolytic O–H bond dissociation energy of
HNO2 (78 kcal mol−1)54,55 suggests the H atom abstraction by
NO2 will be endothermic, and thus occur via a more product-
like transition state, and as we have explained above lead to
preferential deethylation.
Hydrogen atom abstraction from the a-position of amine
radical cations by poor H atom abstractors has been reported
under conditions where the effective base concentration is very
low.56,57 Some studies have shown that this reaction proceeds with
a preference for reaction of the more substituted alkyl group.58
As a result of these considerations, our evidence and relevant
literature data, we propose that the mechanism of aromatic
amine N-dealkylation that gives rise to highly preferential N-
deethylation, and which occurs in strong acid, with NOBF4,
and with NO2 in the absence of significant H2O, proceeds as
depicted in Scheme 7. This represents a new mechanism of
nitrosative N-dealkylation. The key steps involve generation
of the radical cation 28 in a rate determining step, and NO2
mediated a-H atom abstraction from the more substituted N
alkyl group of this species to give the iminium ion 29 and nitrous
acid. The conversion of 29 to 17 by nitrous acid has literature
precedent.59 The incursion of this mechanism is undoubtedly
driven by the absence of a base concentration significant enough
to provide for competitive deprotonation of this acidic radical
cation, as well as the presence of a free radical, NO2, able
to abstract an H-atom from the alkyl group. When 15 was
reacted with nitrous acid in open vessel, the % deEt (62%)
was lower than when the identical reaction was carried out in
a sealed flask (82% deEt) using 75% HOAc–5% HClO4–0.7 M
NaOAc. Sealing the vessel minimizes the NO and NO2 loss to the
atmosphere, and the increase in % deEt under sealed conditions
illustrates the importance of a volatile reagent in the mechanism
giving preferential deethylation. Control experiments in both dry
CH3CN and in strong acid showed that there was no significant
reaction when NO and 15 were mixed, as expected.
While the radical 12 (see Scheme 3) could occur via a direct
a-CH abstraction by NO2, in order for this path to predominate,
12 would have to be oxidized to the iminium ion 13 more
rapidly than it recombines with any radical species, since the
latter processes would give different products, none of which
are observed. We did not measure kinetics or KDIE for the
reaction of 15 with NO2, which is very fast, but the arguments
presented and the similarities with the nitrosations in strongly
acidic media suggest that this pathway is unlikely here. The
NO2 reaction probably takes a slightly different course when
more electron rich amines are employed as substrates because
these compounds have a lower oxidation potential and electron
transfer to NO2 to form a radical cation is more likely than it is
with 15.
Variation of regiochemistry within a run
As noted above, we had observed changes in the regio-
chemistry of nitrosative N-dealkylation of ethyl 4-N-ethyl-N-
methylaminobenzoate during the course of a run. In the course
of our investigation we also observed this phenomenon for the
nitrosation of 15 under some conditions. A more thorough
investigation yielded the data given in Fig. 3. The nitrosation
of 15 in 75% HOAc–5% HClO4–0.8 M NaOAc was examined
in two ways. In the first case a concentrated solution of sodium
nitrite was added to 15 in the acid solution. In the other case the
nitrous acid solution was prepared in the same acid mixture and
15 added to it. In both cases the regiochemistry of dealkylation
changes from predominant demethylation to deethylation with
increasing run time. The data presented in Fig. 3 show that the
change in the mode of addition results in three effects. When the
amine is added to the preformed nitrous acid, 1) higher initial
Fig.
3 Product yields arising from the nitrosation of N-ethyl-
N-methyl-4-nitroanaline 15 in 75% HOAc–5% HClO4–0.8 M NaOAc
are given as function of time. The transformation was conducted in two
ways. The lines with the closed symbol show product yields when a 10 eq.
of sodium nitrite was added to a solution of the substrate amine in acid.
The lines with the open symbols show product yields when the substrate
amine was added to a freshly prepared solution of nitrous acid. The inset
is an expansion of the first 15 minutes of the transformation and clearly
shows that the yield of the initial major product 16 is overtaken by the
production (deethylation) of 17 in each case. The yield cross over point
is reached more rapidly, 2 min, in the freshly made solution of nitrous
acid, than it is (10 min) when nitrite is added to the acidic solution.
Scheme 7 Mechanism C: H atom abstraction from radical cation.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 1 0 4

View Article Online
product yields result; 2) the change over time from demethylation
to deethylation is shorter; and 3) the percent yield of the nitro
compound, 18, is much greater. Since both transformations are
conducted at the same acidity, the changes must result from
changes in NO2 concentrations. When NaNO2 is added to the
acidic amine solution, more time is required for equilibration
and the generation of NO2 both by the decomposition of N2O3
and the oxidation of NO. Evidence in support of this is seen in the
yields of 18, which forms from NO2. As the concentration of NO2
increases, dealkylation through H-atom abstraction becomes
more competitive, and this results in predominant deethylation.
The yields given in Fig 3 are, of course, cumulative. To determine
how the change in regiochemistry changes with time (data not
shown) we made plots of the change in product yield per unit
time as a function of time. To do this we simply subtracted the
cumulative yield of 17 from the prior yield and divided it by the
time interval between points in Fig. 3. This type of data analysis
showed that the extent of deethylation peaks at 8 min for amine
addition to HNO2 and at about 13 min for the other case. Thus,
as the reaction goes on and NO2 is either lost to reaction or the
atmosphere the % deEt decreases after an initial increase.
between deprotonation (Mechanism B) and H-atom abstraction
from the radical cation at the same acidity will lead to a greater
degree of deprotonation for the p-NO2 amine radical cation
because of its greater acidity. From another perspective, higher
acidities are required to suppress the deprotonation process for
the p-NO2 compound compared to its p-Cl analog. This is why
we observe more deethylation for the p-Cl compound at the
same acidity. The comparison of the nitrous acid chemistry
of these two substrates supports our hypothesis regarding the
mechanistic origin of the regiochemical change in the nitrosative
N-dealkylation of aromatic amines.
NOBF4 Reactions
As discussed above, the incursion of the H-atom abstraction
mechanism in aqueous acid, Mechanism C, results principally
from three factors, generation of the amine radical cation, sup-
pression of its a-H deprotonation at high acidity by a decrease
in the available base concentration, and H-atom abstraction of
the a-H by NO2, which competes with recombination of the
radical cation with NO2 to generate the nitro compound. As
noted above in our discussion of the data presented in Table 5,
NOBF4 also reacts in CH3CN with 15 to give mainly 17 by highly
preferred deethylation. Application of our new mechanism to
this process requires the presence of NO2, the source of which is
not immediately obvious. Yet these transformations also result
in the formation of the o-NO2 product 18. The generation of
the radical cation 28 by electron transfer from 15 to NO+ is
supported by their respective oxidation potentials. The NO
produced in this reaction could react with O2 to give NO2, a
rapid process,60 yet we took strong measures to prevent oxygen
contamination. Hydrolysis of NOBF4 by traces of H2O resulting
from either trace contamination or production in the reaction
mixture provides a more likely explanation for the formation of
NO2 under these conditions, and is well supported by our data.
We believe that significant amounts of NO2 form in this media
by homolysis of N2O3, which is generated from the reaction of
NOBF4 with the HNO2 produced in the hydrolysis of NOBF4
[see eqns. (9), (10) and (3)]. Water is produced from NOH
decomposition (Scheme 2). If NOH elimination occurs under
these conditions, then H2O equivalent to the amount of N2O
formed will be produced. For determination of the amount of
N2O evolved in the reaction of 15 with NOBF4, it was necessary
to add 1 lL of water to the reaction since aqueous NaN3 was
required for accurate preparation of a calibration curve. This re-
sulted in a higher yield of 18 (61.0%) than under anhydrous
conditions, but a similar nitrosamine ratio (86% deEt), as shown
in Table 5. Thus, the small amounts of H2O in this reaction
are only changing the percentage of the path that takes
the nitrosamine course, and favor nitro compound formation
through the hydrolytic, then homolytic production of NO2 from
N2O3. Analysis of the N2O yield (0.1%) showed that under these
Because we had examined the N,N-dialkyl-4-chloroaniline
system extensively,10 we also chose to examine the nitrosative
dealkylation of 4-chloro-N-ethyl-N-methylaniline 30 at higher
acidities than employed previously. Comparison of the com-
pounds with the nitro and chloro substituents, 15 and 30, is com-
plicated by their immense difference in nitrosation reaction rates.
Under identical conditions, with a 10-fold excess of NaNO2
in acetic acid, 30 is consumed entirely in less than 5 minutes,
whereas the majority of 15 remains after 2 hours. This is likely
due to either differences in the nitrosation rates and/or the ease
of oxidation of the substrates, since 30 is more basic, more
nucleophilic and has an oxidation potential which is ∼0.7 V
less than that of 15, by comparison to the dimethyl analogues.19
The regiochemistry of the nitrosative dealkylation of 30 is
also dependent on the reaction acidity and initial [NaNO2].
Nitrosation of 30 with a 2 fold excess of NaNO2 in 75% acetic
acid, gave 4-chloro-N-ethyl-N-nitrosoaniline 31 as the major
nitrosamine after 6 min (2.4% 31, 2.1% 32, 57% deEt). When a
10-fold excess of NaNO2 is used, the % deEt increases to 71%
at a reaction time of 2 min. In each case, the other products
were 4-chloro-N-methyl-N-nitrosoaniline 32, and 4-chloro-2-
nitro-N-ethyl-N-methylaniline 33, the major product in all cases.
However, on the addition of 5% HClO4, the % deEt changed with
the run time. Early in the reaction demethylation dominated
(giving 31, 49% deEt), but as the reaction progressed, the
proportion of 32 (deethylation) increased, so that at the end of
the run 32 was the major nitrosamine product (63% deEt). These
observations support the conclusions made for the nitrosation of
15; that at higher acidity an alternative mechanism participates
resulting in preferential cleavage of the larger alkyl group, and
that several different mechanisms, which are subject to subtle
changes in reaction conditions, are operative in nitrosamine
formation. Increased deethylation with greater concentrations
conditions, which favor deethylation, only 3.5
1.3% of the
total nitrosamine formed by Mechanism A (Table 5).
−
of initial NO2 can be explained by the larger amounts of NO2
NOBF4 + H2O → HNO2 + HBF4
(9)
produced by these conditions.
Comparison of the data above for the nitrosation of 30 in
75% HOAc with that for 15 (run 2, Table 1) in the same acid
mixture (10 × NaNO2) shows that the ring substituent influences
the % deEt. For p-Cl deethylation is predominant, whereas the
major nitrosamine is formed by demethylation when the ring
substituent is p-NO2. The large excess of NaNO2 ensures large
concentrations of NO2. For the p-Cl amine, which has the lower
oxidation potential, the role of the radical cation is evident by
the large relative yields of nitro compound (76–83% at reaction
completion). Thus as evidenced by rate and relative product
yields, more radical cation forms from the p-Cl amine than does
from the p-NO2 amine under the same conditions. The a-H
atoms of the amine radical cations also have different pKa’s,
the p-NO2 being more acidic (3 vs. 9 for p-Cl).18,19 Competition
NOBF4 + HNO2 → N2O3 + HBF4
(10)
Summary and conclusions
By using the p-NO2 amine 15, as our principal substrate, but
by also employing the p-Cl analog 30, we have demonstrated
that the regiochemistry of N-dealkylation leading to nitrosamine
formation changes with the acidity of the media and propose
that the N-deethylation observed at high acidity arises from the
NO2 mediated H-atom abstraction from an intermediate radical
cation produced from the effective one electron oxidation of the
amine by NO+. The fact that preferred deethylation of these
substrates does not occur with significant N2O production as
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 1 0 5

View Article Online
required by Mechanism A, that it does not involve observation
of a primary KDIE for the loss of the amine, and the that
the regiochemistry is different from that of electrochemically or
chemically generated radical cations which are undergoing N-
dealkylation by amine radical cation deprotonation (preferred
demethylation) requires the incursion of a new mechanism of
N-dealkylation. Evidence supporting the involvement of NO+
in this process is provided by the kinetics [see eqn. (7)] which
are first order in amine, nitrite, and [H+], by the fact that the
% deEt increases with the calculated [NO+], which increases
with increasing acidity as does the rate, and by the fact that
reactions of the amine with NOBF4 in acetonitrile give a high %
deEt, as occurs in strong acid. The proposal that NO+ is acting
as an effecting one electron oxidant through an inner sphere
process to produce an amine radical cation is supported by
respective reactant oxidation potentials, application of Marcus
theory, extensive chemical precedent, particularly mechanism
studies of nitrite mediated aromatic nitration reactions, the
lack of a primary kinetic deuterium isotope effect for the
loss of the amine, and our observation of 15N-CIDNP for
the formation of the nitro compounds 15 and 23, which arise
from the recombination of NO2 with a radical cation in the
“nitrosation” of 21. Evidence that the dealkylation at high
acidity occurs through NO2 mediated H-atom abstraction from
the more substituted N-alkyl substituent (here ethyl vs. methyl)
is provided by the observations that: 1) the % deEt of 15 is
increased when the reaction is sealed confining volatile reactant;
2) reaction of 15 with NO2 in CH3CN and glacial HOAc results
in preferred deethylation (79–90%) but changes to 31% upon
addition of 2% H2O to CH3CN because the transformation
changes to deprotonation of the radical cation by H2O; 3) the
% deEt of 30 is greater at higher initial [NO2−] where the [NO2]
is greater; 4) the change in the regiochemistry of dealkylation
from demethylation to deethylation occurs more rapidly in the
nitrosation of 15 when NO2− and acid are pre-mixed allowing for
higher concentrations of NO2 to develop prior to the addition of
the amine; 5) physical organic chemical precedent which predict
that the NO2 mediated H-atom abstraction is endothermic and
will result in a product-like transition state structure where the
H-atom abstraction will occur so as to generate the incipient
positive charge at the C-atom most able to electronically stabilize
it; and 6) that significant competitive ring nitration involving
NO2 always occurs when preferential deethylation is observed
showing that NO2 is present.
In our prior work we presented evidence for the formation
of the amine radical cation 11 by reversible homolysis of the
nitrosammonium ion 10.10 Here we are purposefully less specific
about the exact nature of the oxidation process because Mech-
anism A, loss of NOH from the nitrosammonium ion, does not
compete effectively in strong acid. The stronger acid, resulting
in higher concentrations of NO+ and the much less basic (p-
NO2 vs. p-Cl) amine substrate may combine to provide a more
competitive oxidation channel to the radical cation. The radical
cation produced enters into three competitive transformations,
recombination with NO2 to generate a nitro compound, N-alkyl
a-CH deprotonation, and N-alkyl a-CH H-atom abstraction,
which ultimately lead to iminium ions and then nitrosamines.
No prior studies reveal the detailed factors underlying how the
first of these processes competes with the other two. We observe
much more ring nitration as the aromatic ring substituents
become less electron withdrawing. In the case of 21, the
unsubstituted amine, recombination with NO2 to generate 15
and 23 appeared to be the only reaction of the radical cation.
When the ring is substituted with Cl or NO2, then the other
processes compete. The more electron rich unsubstituted radical
cation has a longer life time and radical–radical recombination
becomes more probable leading to more nitro product. We
have previously shown for the p-Cl and p-EtO2C substituted
amines that an increase of the basicity of the medium, e.g.
addition of sodium acetate, results in increased rates of radical
cation a-CH deprotonation (Mechanism B). Here we have
demonstrated that this process is effectively shut down as the
acidity of the medium increases. This leads to the incursion
of the H-atom abstraction process which gives the iminium
ion directly. Very high acidities are required for the relatively
acidic radical cation derived from 15 compared to the less acidic
radical cation with p-Cl derived from 30. Thus we may anticipate
a greater role for the H-atom abstraction mechanism as the
aromatic amines become more electron rich and their derived
radical cations become less acidic and less amenable to facile C–
H deprotonation. On the other hand, pathway A also appears
to become more competitive with these substrates at lower
acidity.10
We believe that the data and interpretation presented here
significantly clarify the mechanism and factors which result
in the regiochemistry and the change thereof with changing
reaction conditions involved in the nitrosative dealkylation of
aromatic dialkyl amines. While no one piece of evidence actually
defines the H-atom abstraction mechanism, the body of work
and literature precedent strongly support the rationality of our
proposal for the occurrence of this new mechanism of nitrosative
amine dealkylation. This mechanism is only anticipated to
occur when radical cations can be formed and structural or
media factors prevail to reduce the rate of the apparently more
favorable a-CH deprotonation pathway.
Experimental
Caution
Nitrosamines, and nitrosation reaction mixtures which produce
them, should be considered carcinogenic and appropriate care
taken in their handling. We performed all operations in well
ventilated hoods. Nitrosamines are effectively destroyed by
treatment with 30% HBr–glacial acetic and we routinely treat
all of our glassware with this solution prior to further cleansing.
Dilute aqueous solutions of nitrosamines can be destroyed by
bringing the solution to pH 12–13 and reaction with aluminium
foil or Raney-nickel.
Synthesis
Synthesis of amines and nitrosamines. In most cases, known
compounds were prepared by literature procedures, or mod-
ifications thereof. Any significant changes to syntheses are
described in the Supporting Information.† The nitrosamines
were prepared as described by Hodgson and Nicholson.14 Slight
modifications of the procedures used by Lamm were used for
the secondary amines.61 The hydrazones were also prepared
by a literature method.62 Other compounds: N-ethyl-N-methyl-
4-nitrosoaniline 22,63 N-ethyl-N-methyl-2-nitroaniline 23,64
N-ethyl-N-methyl-2,4-dinitroaniline 18,65 N-ethyl-N-methyl-4-
chloro-2-nitroaniline 33.66
Synthesis of N-ethyl-N-methyl-4-nitroaniline (15). This is
a known compound,67 however, no efficient methods for its
synthesis have been described previously. In this synthesis, 4-
bromonitroaniline (3.0 g, 0.015 mol) and N-methylethylamine
(2.8 g, 47 mmol) were dissolved in 10 mL pyridine, in a 15 mL
pressure tube. The cap was screwed on tightly and the vessel
heated to 115 ◦C for 48 hours in an oil bath. The solution
was cooled, added to 50 mL water, the precipitate filtered and
recryst◦allized from ethanol yielding 15 (2.35 g, 87%), melting at
85–86 C. dH (250 MHz, CDCl3, Me4Si) 8.11 (d, 2H), 6.60 (d,
2H), 3.49 (q, 2H), 3.06 (s, 3H), 1.21 (t, 3H). dC (250 MHz, CDCl3,
Me4Si) 153.14, 136.55, 126.23, 110.01, 46.89, 37.74, 11.49.
Synthesis of N-(1,1-D2)ethyl-N-methyl-4-nitroaniline (15-D2).
N-Methyl-4-nitroacetanilide was prepared by a standard litera-
ture procedure,68 and reduced with lithium aluminium deuteride
by a known mild reduction method.67 The product was purified
by silica column (10% ethyl acetate in hexanes), giving 15-D2
1 1 0 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8

View Article Online
(0.30 g, 16%); mp 82–84 ◦C. dH (250 MHz, CDCl3, Me4Si) 8.11
(d, 2H), 6.60 (d, 2H), 3.06 (s, 3H), 1.20 (s, 3H), isotopic purity
>95%.
are probably a result of either incomplete derivatization or
extraction.
Measurement of N2O evolved during the nitrosation reaction.
The method and calibration was the same as used previously,10
except the headspace was analyzed by GC-ECD following
sampling through an acid trap containing NaOH and anhydrous
CaSO4.
Synthesis of 4-chloro-N-ethyl-N-methylaniline (30). This
is
a
known compound and was prepared from 4-
bromochlorobenzene and N-methylethylamine by a published
Pd coupling method,69 in 92% yield. The spectral data are
consistent with those in the literature.
Reaction of 15 with NO2. A 100 mL round bottom flask
was charged with 6 mL of 15 (11.2 mM) in of acetic acid.
Methyl 3-nitrobenzoate was added as an internal standard. The
headspace above the reaction was replaced with nitrogen dioxide
at atmospheric pressure. Samples (0.2 mL) were subjected to
standard work up and analysis. This reaction was also carried
out in dry acetonitrile and in 98% acetonitrile. These samples
were analyzed directly by HPLC, after blowing nitrogen through
the solution.
Nitrosation reactions
Nitrosation of 21. The nitrosation was repeated as described
14
by Hodgson and Nicholson, with a modified procedure for
product analysis. The reaction mixture was made basic with
KOH, the products extracted into 3 × 15 mL ethyl acetate,
washed with NaCl, then with water, dried over Na2SO4 and
the solvent removed. The product mixture was separated by
silica column (10% ethyl acetate in hexanes), and the fractions
identified by 1H NMR and GC–MS. To quantify the products,
21 (26 mg, 0.20 mmol) in 100 mL of 0.12 M HCl was reacted
with 20 mL of NaNO2 (0.05 M) which was added rapidly at
Nitrosation of N,N-dialkylanilines with NOBF4. An oven
dried, stoppered, round bottom flask was charged with a 3 mL
solution of 15 (3.7 mM) and methyl 3-nitrobenzoate (5.0 mM)
in dry, degassed acetonitrile. The headspace was replaced with
nitrogen, and 3 mL nitrosonium tetrafluoroborate (0.13 M) in
dry, degassed acetonitrile was added by syringe. After 40 min,
a 0.6 mL was sample was subjected to work up and analysis
by the standard procedure. This procedure was repeated using
N-cyclopropyl-N-methyl-4-chloroaniline 24 (0.12 M), which
had been prepared previously in this laboratory, with 1,3-
dinitrobenzene (0.04 M) as an internal standard. The sole
product, 4-chloro-N-methyl-N-nitrosoaniline 25, was identified
by GC-MS, and by spiking the HPLC run.
To quantify the amount of N2O evolved in the reaction of
15 with NOBF4, the method used for N2O determination under
acidic conditions was modified slightly. The calibration curve
was prepared by addition of various concentrations of aqueous
NaN3 (1 lL) to NOBF4 in acetonitrile. For consistency, the
nitrosation of 15 was repeated with the addition of 1 lL distilled
water.
◦
0 C. Samples (5 mL) were worked up as described above and
the organic extract made to 3 mL with benzonitrile (2.81 mM)
in acetonitrile, as an external standard, and analyzed by HPLC.
General Procedure for the nitrosation of 15. In a typical
experiment, a 0.0138 M solution of 15 was prepared in an
acidic solvent. Methyl 3-nitrobenzoate was added as an internal
standard. Of this solution, 9 mL was transferred to a round
bottom flask with stir bar, and subsequently sodium nitrite
(1 mL, 0.115 M) added rapidly by syringe.
Standard workup and analysis procedure. At a designated
time, a 0.6 mL aliquot of the reaction mixture was quenched
with 5 mL saturated sodium bicarbonate, extracted into 3 ×
5 mL ether, washed with 5 mL water and dried over Na2SO4.
The solvent was removed, and the sample dissolved in 0.5 mL
acetonitrile for HPLC analysis.
Oxidation of 15 with Ce(IV). To 5 mL of 15 (2.2 mM) and
methyl 3-nitrobenzoate (0.5 mM) in 75% acetic acid was added
1 mL of aqueous Ce(NH4)2(NO2)6 (1.7 mM), with stirring. After
consumption of the oxidant, as determined by starch/KI paper
(<20 s), 0.6 ml of the reaction was subjected to standard work
up and analysis. The products were identified by LC-MS. The
reaction was also conducted in 75% acetic acid–5% HClO4.
Nitrosation of 15 with prior equilibration of acid and nitrite.
A 50 mL round bottom flask with a stir bar was charged with
8.9 mL of a mixture of 75% HOAc–5% HClO4–0.8 M NaOAc
and stoppered. Sodium nitrite (1 mL, 0.115 M) was added by
syringe through the septum. After 15 minutes, 15 (2.83 M)
and methyl 3-nitrobenzoate (0.75 M), were added by syringe,
in 0.2 mL of 75% HOAc–5% HClO4–0.8 M NaOAc. Samples
(0.6 mL) were removed throughout the course of the reaction,
worked up and analyzed by the standard procedure.
Electrochemical Oxidation of 15. Electrochemical oxida-
tions were carried out in a split cell using a platinum anode
and cathode, and a standard calomel reference electrode. In a
typical experiment, the cathode and buffer region were filled with
75% acetic acid. The anode region was filled with 12 mL of 15
(11.7 mM) and methyl 3-nitrobenzoate (5.0 mM) in 75% acetic
acid. A constant voltage of 1.91 V was applied to the system for
24 h, after which a 0.6 mL sample of the solution in the anode
region was subjected to standard work up and analysis.
General procedure for nitrosation of 30. A procedure identi-
cal to that used for the nitrosation of 15 was employed, using
1,3-dinitrobenzene as the internal standard. The products were
identified by LC-MS and by spiking the HPLC run.
Quantification of aldehyde byproducts. An adaptation of a
method used previously was utilized.62 The general procedure
was used to nitrosate 15 (3.6 M H2SO4 in 90% acetic acid).
After 5 min, a 1 mL aliquot was added to 8 mL of NaOH
(1 M) followed by, 1 mL sulfamic acid (0.17 M). After
1 min, 2 mL of 2,4-dinitrophenylhydrazine (DNP, 2 mM 18%
phosphoric acid in ethanol) was added and stirred for 10 min.
The organic products were extracted into 3 × 10 mL diethyl
ether, washed with 5 mL distilled water, dried over K2CO3, and
analyzed by HPLC. To determine the extraction efficiency of the
aldehyde trapping procedure, control recovery experiments were
conducted using formaldehyde and acetaldehyde. In a typical
experiment, the aldehyde (2.3 mM) and internal standard in
3.6 M H2SO4–90% HOAc was stirred for 5 min following the
addition of 1 mL NaNO2. Work up and analysis was identical
to that used for the nitrosation. Recoveries although low (CH2O
Kinetic experiments
Determination of KDIE for 15-D2. In a typical experiment,
a 100 mL flask was charged with 9.2 mL of 90% acetic acid
containing 15 (6.9 mM) and internal standard (1.7 mM) at
◦
28 C. While stirring, 1 mL of aqueous NaNO2 (57 mM) was
added rapidly. Samples (0.2 mL) were taken at frequent intervals
and subjected to standard work up and analysis. At least eight
samples were taken during each run, and the concentrations
of the organic components determined from their peak areas,
relative to the standard. The runs were repeated 3 times, and the
rate of loss of 15 and the rate of product formation determined
from plots of ln ([15]0 − [P]) vs. time, where [P] = [product]. The
errors were estimated from the standard deviation of the slope,
determined by linear regression. Similar procedures were used
9.8
0.9%, CH3CHO 35.2
2.9%), are reproducible and
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8
1 1 0 7

View Article Online
for other kinetic runs, including 15-D2, and kH/kD determined.
Rates of substrate consumption were also determined by UV,
and were comparable.
25 T. Gunnlaugsson, H. Q. N. Gunaratne, M. Nieuwenhuyzen and J. P.
Leonard, J. Chem. Soc., Perkin Trans. 1, 2002, 1954–1962.
26 R. J. Kaptein, J. Chem. Soc. D, 1971, 732–733.
27 A. H. Clemens, P. Helsby, J. H. Ridd, F. Al-Omran and J. P. Sandall,
J. Chem. Soc., Perkin Trans. 2, 1985, 1217–1225.
28 A. H. Clemens, J. H. Ridd and J. P. B. Sandall, J. Chem. Soc., Chem.
Commun., 1983, 343–344.
29 A. H. Clemens, J. H. Ridd and J. P. B. Sandall, J. Chem. Soc., Perkin
Trans. 2, 1984, 1667–1672.
30 A. H. Clemens, J. H. Ridd and J. P. B. Sandall, J. Chem. Soc., Perkin
Trans. 2, 1984, 1659–1665.
Determination of the reaction order in nitrite. In a typical
experiment, 15 (13.7 mM) and internal standard (4.2 mM) in
75% HOAc–3.6 M H2SO4 was stirred at 23 ◦C, and reacted with
the following concentrations of NaNO2: 0.69 M, 0.35 M, 0.38 M,
0.45 M, 0.48 M, 0.55 M. A plot of ln (k) against ln [NaNO2]
gave straight lines for 15 and for both nitrosamines. The results
were corroborated by UV.
31 A. H. Clemens, J. H. Ridd and J. P. B. Sandall, J. Chem. Soc., Perkin
Trans. 2, 1985, 1227–1231.
32 J. H. Ridd, Chem. Soc. Rev., 1991, 20, 149–165.
33 L. Eberson, M. P. Hartshorn and F. Radner, Acta Chem. Scand.,
1994, 48, 937–950.
Supporting Information
Explanations of CIDNP, and additional experimental details for
kinetics experiments, syntheses of several known compounds,
and N2O determinations are given.
34 B. C. Challis, R. J. Higgins and A. J. Lawson, J. Chem. Soc. D, 1970,
1223–1224.
35 B. C. Challis, R. J. Higgins and A. J. Lawson, J. Chem. Soc., Perkin
Trans. 2, 1972, 1831–6.
36 B. C. Challis and R. J. Higgins, J. Chem. Soc., Perkin Trans. 2, 1973,
1597–1604.
Acknowledgements
We thank Dr Jim Iley of the Department of Chemistry, The Open
University, UK, for his most thoughtful and careful critique of
this work. We also wish to acknowledge the generous support
of this research by the National Cancer Institute, NIH, under
grants R37CA26914 and R01CA85538.
37 In strong acid, e.g. 75% HOAc containing 3 M HClO4, we have
observed the ratio of the nitro product to the nitrosamines to increase
with time while neither the reaction rate nor the nitrosamine product
ratio changes. The mechanism of this process is uncertain.
38 D. H. L. Williams, Nitrosation, Cambridge University Press, Cam-
bridge, 1988.
39 J. Casado, A. Castro, M. Mosquera and M. Rodriguez-Prieto,
Monatsh. Chem., 1984, 115, 669–682.
40 L. R. Dix and D. L. H. Williams, J. Chem. Res., S., 1982, 190–191.
41 K. H. Becker, R. Kleffmann and P. Kurtenbach, J. Phys. Chem., 1996,
100, 14984–14990.
42 F. D. Lewis and T.-I. Ho, J. Am. Chem. Soc., 1980, 102, 1751.
43 J. D. S. Goulden and D. J. Millen, J. Chem. Soc., 1950, 2620–2627.
44 A. Givan and A. Loewenschuss, J. Chem. Phys., 1989, 91, 5126–
5127.
45 D. J. Millen and D. Watson, J. Chem. Soc., 1957, 1369–1372.
46 M. a. Zolfigol, M. H. Zebarjadian, G. Chehardoli, H. Keypour, S.
Salehzadeh and M. Shamsipur, J. Org. Chem., 2001, 66, 3619–3620.
47 P. Gray and A. D. Yoffe, Chem. Rev., 1955, 55, 1069–1154.
48 B. C. Challis and S. A. Kyrtopoulos, J. Chem. Soc., Perkin Trans. 1,
1978, 1296–302.
References
1 H. Druckrey, R. Preussmann, S. Ivankovic and D. Schmaehl,
Z. Krebsforsch., 1967, 69, 103–201.
2 H. Bartsch and R. Montessano, Carcinogenesis (London), 1984, 5,
1381–1393.
3 R. N. Loeppky and C. J. Michejda, Nitrosamines and related N-nitroso
compounds: chemistry and biochemistry, American Chemical Society,
Washington DC, 1994.
4 R. N. Loeppky, Y. T. Bao, J. Y. Bae, L. Yu and G. Shevlin, in
Nitrosamines and Related N-Nitroso Compounds: Chemistry and
Biochemistry, R. N. Loeppky and C. J. Michejda, Eds., American
Chemical Society, Washington, DC, 1994, p 52–65.
5 H. Bartsch, H. Ohshima, B. Pignatelli and S. Calmels, Cancer Surv.,
1989, 8, 335–362.
6 R. N. Loeppky, R. Hastings, J. Sandbothe, D. Heller, Y. Bao and D.
Nagel, IARC Sci. Publ., 1991, 105, 244–252.
7 R. Hastings and R. N. Loeppky, unpublished laboratory report.
8 P. A. S. Smith and R. N. Loeppky, J. Am. Chem. Soc., 1967, 89,
1147–1157.
9 B. Gowenlock, R. J. Hutcheson, J. Little and J. Pfab, J. Chem. Soc.,
Perkin Trans. 2, 1979, 1110–1114.
10 R. N. Loeppky, S. P. Singh, S. Elomari, R. Hastings and T. E. Theiss,
J. Am. Chem. Soc., 1998, 120, 5193–5202.
11 R. N. Loeppky and S. Elomari, J. Org. Chem., 2000, 65, 96–103.
12 G. Verardo, A. G. Giumanini and P. Strazzolini, Tetrahedron, 1991,
47, 7845–7852.
13 G. Verardo, A. G. Giumanini and P. Strazzolini, Tetrahedron, 1990,
46, 4303–4332.
14 H. H. Hodgson and D. E. Nicholson, J. Chem. Soc., 1941, 470–475.
15 J. R. Lindsay Smith and L. A. V. Mead, J. Chem. Soc., Perkin Trans.
2, 1973, 206–210.
49 B. C. Challis and S. A. Kyrtopoulos, J. Chem. Soc., Perkin Trans. 1,
1979, 299–304.
50 B. C. Challis, D. E. G. Shuker, D. H. Fine, E. U. Goff and G. A.
Hoffman, IARC Sci. Publ., 1982, 41, 11–20.
51 I. Tohru, T. Kawabata and M. Matsui, Nippon Suisan Gakkaishi,
1984, 50, 1425–1429.
52 W. A. Pryor, J. W. Lightsey and D. F. Church, J. Am. Chem. Soc.,
1982, 104, 6685–6692.
53 E. T. Denisov, T. G. Denisova, Y. V. Geletii and J. Balavouane, Kinet.
Catal., 1998, 39, 312–319.
54 V. D. Parker, J. Am. Chem. Soc., 1992, 114, 7458–7462.
55 H. Suzuki and N. Nonoyama, J. Chem. Soc., Perkin Trans. 1, 1997,
2965–2971.
56 S. F. Nelsen and J. T. Ippoliti, J. Am. Chem. Soc., 1986, 108, 4879–
4881.
57 R. W. Alder, M. Bonifacic and K. D. Asmus, J. Chem. Soc., Perkin
Trans. 2, 1986, 277–284.
58 S. F. Nelsen and J. T. Ippoliti, J. Org. Chem., 1986, 51, 3169–3176.
59 L. K. Keefer and P. P. Roller, Science, 1973, 181, 1245–1247.
60 J. C. Treacy and F. Daniels, J. Am. Chem. Soc., 1955, 77, 2033–2036.
61 B. Lamm, Acta. Chem. Scand., 1966, 19, 2316–2322.
62 R. N. Loeppky and P. Goelzer, Chem. Res. Toxicol., 2002, 15, 457–
469.
63 R. L. Bent, J. C. Dessloch, F. C. Duennebier, D. W. Fassett, D. B.
Glass, T. H. James, D. B. Julian, W. R. Ruby and J. M. Snell, J. Am.
Chem. Soc., 1951, 73, 3100–3125.
16 J. R. Lindsay Smith and D. Masheder, J. Chem. Soc., Perkin Trans.
2, 1976, 47–51.
17 C. A. Audeh and J. R. Lindsay Smith, J. Chem. Soc. B, 1970, 1280–
1285.
18 J. P. Dinnocenzo and T. E. Banach, J. Am. Chem. Soc., 1989, 111,
8646–8653.
19 V. D. Parker and M. Tilset, J. Am. Chem. Soc., 1991, 113, 8778–8781.
20 B. G. Gowenlock, J. Pfab and V. M. Young, J. Chem. Soc., Perkin
Trans. 2, 1997, 1793.
21 N. Nonoyama, K. Chiba, K. Hisatome, H. Suzuki and F. Shintani,
Tetrahedron Lett., 1999, 40, 6933–6937.
64 T. W. Campbell, J. Am. Chem. Soc., 1949, 71, 740.
65 E. W. Day, T. Golab and J. R. Koons, Anal. Chem., 1966, 38, 1053–
1057.
22 B. P. Cho, L. R. Blankenship, J. D. Moody, D. R. Doerge, F. A.
Beland and S. Culp, Tetrahedron, 2000, 56, 7379–7388.
23 N. Nonoyama, K. Hisatome, C. Shoda and H. Suzuki, Tetrahedron
Lett., 1999, 40, 6939–6940.
24 N. Nonoyama, H. Oshima, C. Shoda and H. Suzuki, Bull. Chem.
Soc. Jpn., 2001, 74, 2385–2395.
66 G. R. Clemo and J. M. Smith, J. Chem. Soc., 1928, 2415.
67 R. F. Nystrom and C. R. A. Berger, J. Am. Chem. Soc., 1958, 80,
2896–2898.
68 J. A. Moore, D. L. Dalrymple, Experimental Methods in Organic
Chemistry, 2nd edition, W. B. Saunders Co., Philadelphia, 1976.
69 W. Cui and R. N. Loeppky, Tetrahedron, 2000, 57, 2953–2956.
1 1 0 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 0 9 7 – 1 1 0 8