Amidine Nitrosation

Article abstract of DOI:10.1021/jo035884u

The acidic nitrosation chemistry of nine acyclic secondary and tertiary amidines (Ph-N=C(R₁)-NR₂R₃; R₁ = H, CH₃, Ph; R₂, R₃ = H, Ph or (CH ₃)₂ or C(CH₂)₄) and several N-acylamidines was investigated. The principal nitrosation products were amides derived from the amino moiety and compounds derived from the benzenediazonium ion, which was independently trapped for quantitation in several cases. Tertiary amidines also produce nitrosamines in minor, but significant, yields. The benzamidines did not react, and the N-acylamidines hydrolyzed much more rapidly than they nitrosated. The data support the hypothesis that the reaction occurs by nitrosation on the imino nitrogen, followed by the addition of H ₂O to give a tetrahedral intermediate (α-hydroxynitrosamine) for which the main decomposition pathway generates an amide and a diazonium ion. In the case of the pyrrolidine-derived amidines, about 25% of the decomposition results in cleavage of the amine moiety, which nitrosates to give N-nitrosopyrrolidine. Pseudo-first-order rate constants for amidine nitrosation in aqueous acetic acid with excess nitrite at 25 °C ranged from (3 to 106) × 10^-5 s^-1, while the amidine basicity ranged over 5 pK_a units. Rate constants corrected for amidine basicity showed the pyrrolidine derived amidines to be most reactive. The lack of benzamidine nitrosative reactivity is attributed to a very slow rate of H₂O additon to the N-nitrosoamidinium ion and reversible nitrosation.

Full text of DOI:10.1021/jo035884u

Am id in e Nitr osa tion
Richard N. Loeppky* and Hongbin Yu
Department of Chemistry, University of MissourisColumbia, Columbia, Missouri 65211
loeppkyr@missouri.edu
Received December 29, 2003
The acidic nitrosation chemistry of nine acyclic secondary and tertiary amidines (Ph-NdC(R1)-
NR2R3; R1 ) H, CH3, Ph; R2, R3 ) H, Ph or (CH3)2 or C(CH2)4) and several N-acylamidines was
investigated. The principal nitrosation products were amides derived from the amino moiety and
compounds derived from the benzenediazonium ion, which was independently trapped for
quantitation in several cases. Tertiary amidines also produce nitrosamines in minor, but significant,
yields. The benzamidines did not react, and the N-acylamidines hydrolyzed much more rapidly
than they nitrosated. The data support the hypothesis that the reaction occurs by nitrosation on
the imino nitrogen, followed by the addition of H2O to give a tetrahedral intermediate
(R-hydroxynitrosamine) for which the main decomposition pathway generates an amide and a
diazonium ion. In the case of the pyrrolidine-derived amidines, about 25% of the decomposition
results in cleavage of the amine moiety, which nitrosates to give N-nitrosopyrrolidine. Pseudo-
first-order rate constants for amidine nitrosation in aqueous acetic acid with excess nitrite at 25
°C ranged from (3 to 106) × 10-5 s-1, while the amidine basicity ranged over 5 pKa units. Rate
constants corrected for amidine basicity showed the pyrrolidine derived amidines to be most reactive.
The lack of benzamidine nitrosative reactivity is attributed to a very slow rate of H2O additon to
the N-nitrosoamidinium ion and reversible nitrosation.
SCHEME 1
In tr od u ction
The chemistry literature reveals little about the ni-
trosation of the carbon-nitrogen double bond 1 f 2.
Reference to Scheme 1 shows why this might be true. In
aqueous solution, the resulting N-nitrosoiminium ion 2
will rapidly react with water to give the R-hydroxyni-
trosamine 3. This compound, of course, resembles inter-
mediates in the hydrolysis of imines and is expected to
decompose to the corresponding carbonyl compound and
the diazonium ion 4. R-Hydroxynitrosamines 3 are well-
known in the literature relating to the carcinogenesis of
nitrosamines. They are formed from the metabolic oxida-
tion of these compounds.1 They have been prepared, and
their chemistry has been studied.2-7 R-Hydroxynitro-
samines are unstable in aqueous solution and decompose
as shown. Thus, the nitrosation of an imine in aqueous
solution is expected to ultimately give the same products
that would be derived from the nitrosation of the corre-
sponding primary amine. Moreover, with the exception
of those cases where the nitrosoiminium ion 2 can be
trapped either by intramolecular or intermolecular nu-
cleophilic addition,7-13 the transformation is expected to
be of little synthetic value.
From another perspective, however, that of chemical
carcinogenesis, transformations such as those shown in
Scheme 1 could be of significance. Many drugs contain
carbon-nitrogen double bonds. The focus of this paper
is on the nitrosation chemistry of the amidine functional
(1) Yang, C. S.; Smith, T. J . Adv. Exp. Med. Biol. 1996, 387, 385-
394.
(2) Mochizuki, M.; Anjo, T.; Okada, M. Tetrahedron Lett. 1980, 21,
3693-3696.
(3) Mochizuki, M.; Sone, T.; Anjo, T.; Okada, M. Tetrahedron Lett.
1980, 21, 1765-1766.
(8) Rajamaeki, M.; Vigroux, A.; Chahoua, L.; Fishbein J . Org. Chem.
1995, 60, 2324-5.
(9) Vigroux, A.; Kresge, A. J .; Fishbein, J . C. J . Am. Chem. Soc. 1995,
117, 4433-4.
(10) Cai, H.; Fishbein, J . C. Tetrahedron 1997, 53, 10671-10676.
(11) Roller, P. P.; Keefer, L. K. IARC Sci. Publ. 1975, 9, 86-9.
(12) Wiessler, M. Tetrahedron Lett. 1975, 2575-8.
(13) Eiter, K.; Hebenbrock, K. F.; Kabbe, H. J . Liebigs Ann. Chem.
1972, 765, 55-77.
(4) Mesic, M.; Revis, C.; Fishbein, J . C. J . Am. Chem. Soc. 1996,
118, 7412-7413.
(5) Mesic, M.; Peuralahti, J .; Blans, P.; Fishbein, J . C. Chem. Res.
Toxicol. 2000, 13, 983-992.
(6) Kim, H.-J .; Fishbein, J . C. Chem. Res. Toxicol. 2003, 16, 715-
720.
(7) Revis, C.; Rajamaki, M.; Fishbein, J . C. J . Org. Chem. 1995, 60,
7733-8.
10.1021/jo035884u CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/25/2004
J . Org. Chem. 2004, 69, 3015-3024
3015

Loeppky and Yu
SCHEME 2
SCHEME 3
group, which can be found in many drugs, as well as in
other commercial compounds. Nitrosation occurs in the
human body at variable levels.14-17 All humans excrete
the nitrosated amino acid N-nitrosoproline in their
urine,14 a significant portion of which arises from endog-
enous nitrosation and dietary consumption of this
N-nitrosoamino acid.15 There are three principle sources
of endogenous nitrosating agents:18 nitrite, nitrate, and
products of the oxidation of NO, which is formed in
various cells from arginine.19 The human stomach,
because of its acidity, is an ideal place for the formation
of nitrous acid from dietary nitrite.17,20-26 Dietary nitrate,
through its bacterially mediated temporal biphasic re-
duction to nitrite in the saliva, significantly contributes
to endogenous nitrosation, because the major nitrite
pulse does not occur while eating (phase 1), but from
blood-circulating nitrate secreted into the saliva
glands.27-31 Other nitrate-reducing bacteria, such as may
exist in an infected stomach,23-26 or another organ, are
capable of mediating the nitrosation of nitrogen com-
pounds, such as amines through processes which both
do or do not involve direct production of nitrite. Because
of these well-documented processes, it behooves the
chemist to better understand the reactivity of nitrogen
substrates toward nitrosation.
chemistry of the common drug chlordiazepoxide 5 and
related derivatives (Scheme 2).32 The reaction of 5 with
slightly more than one equivalent of sodium nitrite in
glacial acetic acid generates the N-nitrosoamidine 6. This
compound has interesting chemistry and has been uti-
lized in the synthesis of numerous derivatives.32-34 On
the other hand, the nitrosation of 5 in hydrochloric acid
takes another course and results in the formation of the
oximino quinoxaline 7. The N-nitrosoamidine 6 forms
under gastric nitrosation conditions in animals. It pos-
sesses genotoxic activity but is not carcinogenic toward
mice.35-38
Although several N-nitrosoamidines derived from sec-
ondary amidines have been reported in the literature,
little appears to be known about the nitrosation chem-
istry of amidines. One exception involves the nitrosation
N-Nitroso derivatives of 2-arylimidazolines have been
prepared by the reaction of cyclic secondary amidines
with N2O4. Another group of N-nitrosoamidines have
39
been prepared in order to study the migration of the
N-nitroso group from one nitrogen to another by an
intramolecular process, but the nitrosation chemistry of
the parent compounds has not been discussed.40
We have previously shown that three tertiary amidines
reacted with nitrous acid under acidic conditions to give
nitrosamines, N-nitrosoamides, and an array of products
derived from the decomposition of diazonium ions (see
Scheme 3).41 At the time that research was done, our
attention was principally focused on the deleterious
biological action which could result from the formation
of nitrosamines or other N-nitroso compounds. We imag-
(14) Ohshima, H.; Bartsch, H. Cancer Res. 1981, 41, 3658-3662.
(15) Wagner, D. A.; Schultz, D. S.; Deen, W. M.; Young, V. R.;
Tannenbaum, S. R. Cancer Res. 1983, 43, 1921-1925.
(16) Wagner, D. A.; Tannenbaum, S. R. Food Technol. (Chicago)
1985, 39, 89-90.
(17) Spiegelhalder, B.; Preussmann, R. Carcinogenesis (London)
1985, 6, 545-548.
(18) Marletta, M. A. Chem. Res. Toxicol. 1988, 1, 249-257.
(19) Marletta, M. A.; Yoon, P. S.; Iyengar, R.; Leaf, C. D.; Wishnok,
J . S. Biochemistry 1988, 27, 8706-8711.
(20) Sander, J .; Seif, F. Arzneim. Forsch. 1969, 19, 1091-1093.
(21) Lijinsky, W.; Greenblatt, M. Nature (London) 1972, 236, 177-
178.
(22) Mirvish, S. S. Toxicol. Appl. Pharmacol. 1975, 31, 325-351.
(23) Bartsch, H.; Ohshima, H.; Pignatelli, B.; Calmels, S. Cancer
Survey 1989, 8, 335-362.
(32) Walser, A.; Fryer, R. I.; Sternbach, L. H.; Archer, M. C. J .
Heterocycl. Chem. 1974, 11, 619-621.
(33) Walser, A.; Fryer, R. I. J . Org. Chem. 1975, 40, 153-157.
(34) Walser, A.; Benjamin, L. E.; Flynn, T.; Mason, C.; Schwartz,
R.; Fryer, R. I. J . Org. Chem. 1978, 43, 936-944.
(35) Brambilla, G.; Robbiano, L.; Martelli, A.; Cajelli, E.; Allavena,
A.; Mazzei, M. Toxicol. Appl. Pharmacol. 1989, 97, 480-4888.
(36) Robbiano, L.; Carlo, P.; Finollo, R.; Brambilla, G. Toxicol. Appl.
Pharmacol. 1990, 102, 186-190.
(37) Mereto, E.; Brambilla Campart, G.; Ghia, M.; Petrogalli, F.
Cancer Lett. 1990, 53, 61-65.
(38) Giner-Sorolla, A.; Greenbaum, J .; Last-Barney, K.; Anderson,
L. M.; Budinger, J . M. Food Cosmet. Toxicol. 1980, 18, 81-3.
(39) Iley, J .; Norberto, F.; Rosa, E. J . Chem. Soc., Perkin Trans. 2
1989, 1471-5.
(40) Mikhailov, I. E.; Dushenko, G. A.; Zhunke, A.; Miigge, K.;
Minkin, V. I. Russ. J . Org. Chem. 1969, 34(8), 1127-1130.
(41) Loeppky, R. N.; Yu, L. Tetrahedron Lett. 1990, 31, 3263-3266.
(24) Mackerness, C. W.; Leach, S. A.; Thompson, M. H.; Hill, M. J .
Carcinogenesis (London) 1989, 10, 397-399.
(25) Calmels, S.; Bereziat, J . C.; Ohshima, H.; Bartsch, H. IARC
Sci. Publ. 1991, 105, 187-191.
(26) Parke, D. V. Toxicol. Ecotoxicol. News/ Reviews 1997, 4, 132-
137.
(27) Green, L. C.; De Luzuriaga, K. R.; Wagner, D. A.; Rand, W.;
Istfan, N.; Young, V. R.; Tannenbaum, S. R. Proc. Natl. Acad. Sci.
U.S.A. 1981, 78, 7764-7768.
(28) Tannenbaum, S. R.; Sinskey, A. J .; Weisman, M.; Bishop, W.
J . Nat. Cancer Inst. 1974, 53, 79-84.
(29) Spiegelhalder, B.; Eisenbrand, G.; Preussmann, R. Food Cosmet.
Toxicol. 1976, 14, 545-548.
(30) Tannenbaum, S. R.; Archer, M. C.; Wishnok, J . S.; Bishop, W.
W. J . Natl. Cancer Inst. 1978, 60, 251-253.
(31) Eisenbrand, G.; Spiegelhalder, B.; Preussmann, R. Oncology
1980, 37, 227-231.
3016 J . Org. Chem., Vol. 69, No. 9, 2004

Amidine Nitrosation
CHART 1
compounds, which were also used as standards in the
quantitative determination of product yields by GC or
HPLC. Reaction conditions and product yields are sum-
marized in Table 1. The transformations result in scission
of the amidine into two sets of products derived from the
two nitrogen containing moieties as minimally depicted
in Scheme 3. For example, the nitrosation of the N,N′-
diphenylformamidine 8a (run 1) at 0 °C gives formanilide
(97%) and the benzenediazonium ion derived products
4-azophenylphenol (23%) and 2,4-bis(4-azophenyl)phenol
(47%), giving material balances for the two fragments of
97% and 90%, respectively. In general, material balances
were quite good for all runs. The exceptions were for
those runs where a highly water-soluble amide (DMF)
or dimethylnitrosamine was a product. In runs 10, 12,
and 13, we modified our analytical procedure to detect
these compounds as effectively as possible in order to
make valid comparisons. In some cases, our emphasis
was on establishing the nature of the products, e.g., runs
6-9. Even at long reaction times, the more basic amidines
(8b,c,e,f) reacted incompletely (substrate recoveries are
given in Table 1). Regardless of basicity, the benz-
amidines 8g-i did not nitrosate. As a control, hydrolyses
of all amidines were examined under conditions similar
to those for nitrosations, and in no case was hydrolysis
observed on the time scale and conditions of the at-
tempted transformation.
The products of two typical nitrosation reactions, runs
2 and 10, are illustrated in Scheme 4. Substrates 8a , 8d ,
and 8g are secondary amidines and could produce N-
nitrosoamidines as products, but none were observed.
Prior work39 and our own experience43 suggest that
N-nitrosoamidines derived from 8a and 8d will not
survive the acidic reaction conditions due to their rapid
decomposition. In other work,43 we have shown that
N-nitrosoamidines readily nitrosate and decompose, but
our product analyses and material balances indicate that
this transformation is occurring to only a very minor
extent at best because N-nitrosoamides are expected
products (see below). In the two illustrated cases of
Scheme 4, and in all runs where nitrosative reactivity
was observed, products characteristic of the intermediacy
of diazonium ions were observed. Most significantly, the
nitrosation of 8a at 37 °C gives 4-hydroxyazobenzene 19.
This product is formed by the hydrolysis of the benzene-
diazonium ion to phenol which then undergoes azo
coupling with the diazonium ion. As noted above, the
reaction of 8a at 0 °C, which prolongs the lifetime of the
benzenediazonium ion, produces 2,4-bis(phenylazo)phe-
nol in 46% yield. In the case of 8e and related tertiary
amidines, the benzenediazonium ion derived reaction
products phenyl acetate 18, phenol 17, and nitrophenols
22 and 23 are observed. Similar products were observed
in all other amidine (8a -f) nitrosation transformations.
In all cases, the major fragment from the amino moiety
of the amidine was an amide 11 (specifically 16 from 8e),
which for the secondary amidines is a formanilide (here
11 ) 15).
ined the diazonium ion products to arise mainly from the
decomposition of N-nitrosoamides. In reconsideration of
that research, and with new insight gained from our
nitrosation studies on oxazolines and N-alkylimidazolines
(cyclic tertiary amidines),42 we simplified and reformu-
lated our hypotheses to that shown in Scheme 3. We
postulate that amidines react by effective addition of
nitrous acid across the CdN- bond to give a tetrahedral
intermediate, which is also an R-hydroxynitrosamine.
This intermediate is proposed to undergo competitive
decomposition to generate a less stable diazonium ion 10
(path A) or an N-nitrosoamide 12 and a nitrosatable
amine (path B). A goal of the research reported here has
been to test these postulates, to quantify as much as
possible the partitioning between paths A and B, and to
determine how these transformations are affected by
structure, with emphasis being paid to the role of the
carbon substituent R1. Here, we demonstrate that path
A, which gives diazonium ions, is the predominant one
and that as R1 is changed from H to CH3, to Ph, the
amidines become progressively less reactive. Benz-
amidines do not nitrosate at all. We also show that N-
or N′-acyl acyclic amidines hydrolyze much more rapidly
than they nitrosate.
Resu lts a n d Discu ssion
Syn th esis of Am id in es. The nine amidines 8a -i
(Chart 1) used in this work are all known compounds and
were either procured or synthesized by literature proce-
dures as documented in the Experimental Section. Their
physical and spectroscopic properties are given in the
Supporting Information.
Nitr osa tion Rea ction s. In our prior experiments
with amidines, significant consumption of nitrous acid
was observed.41 As a result, only partial reaction of the
substrate occurred, which could be bolstered by the
addition of NO2-. Accordingly, in this work we used a
large excess of nitrous acid in all of the transformations
that we studied. In a typical experiment, the amidine of
interest was dissolved in glacial acetic acid or buffered
acetic acid, and a 5-20 × excess of nitrite was added to
the stirring mixture. Transformations were allowed to
go for variable times depending upon the reactivity of
the amidine and/or the goal of the experiment. The
reactions were quenched by dilution with water followed
by bicarbonate basification. Following extraction into an
organic solvent and concentration, the reaction products
were separated by flash chromatography. In some cases,
the reaction mixtures were analyzed by GCMS or HPLC
after extraction. In addition to the ancillary information
from MS, all products were characterized by chromato-
graphic and spectroscopic comparison with authentic
One of the goals of this research was to determine
whether diazonium ions are primary products of amidine
nitrosation and to see if we could quantitate the diazo-
nium ion formation. Naturally, this required the use of
substrates which would generate a stable aryl diazonium
ion that could be trapped. Accordingly, we performed a
(42) Loeppky, R. N.; Cui, W. Tetrahedron Lett. 1998, 39, 1845-1848.
J . Org. Chem, Vol. 69, No. 9, 2004 3017

Loeppky and Yu
TABLE 1. P r od u ct Yield s fr om th e Nitr osa tion of Am id in esm
a
b
Two equivalents of phenol were added, and 4-nitrosophenol was also a product. Two equivalents of o-cresol were added, and p-nitroso-
o-cresol was also a product. c The analytical method used for product quantitation in each of these runs was identical to ensure valid
comparison of yields of both more and less water-soluble products. Yields corrected for unreacted substrate. e Amide structure that of
d
amidine where O replaces Ph-N (imino). f Either formanilide or acetanilide as indicated by R1. Dimethylnitrosamine for 8b and 8e or
g
N-nitrosopyrrolidine for 8f. Ar ) 4-HOPh except for runs 4 and 5 where Ar ) 4-hydroxy-2-methylphenyl. i % unreacted substrate recovered.
h
j
k
-
2,4-Bis(phenylazo)phenol (46%, based on the reaction of three moles of substrate to yield 1 mole of product) was also isolated. NO2
l
molar equivalents relative to substrate. In a separate run with a duration of 45 h, >92% of 8f was recovered but products were not
quantitated. General abbreviations: na, not applicable; nd, not determined; +, trace; -, not observed.
m
set of experiments employing the N,N′-diphenylforma-
midine 8a where the benzenediazonium ion was trapped
by azo coupling to either phenol or o-cresol. Data from
these experiments are given Table 1 (runs 3-5). The
experiments with o-cresol were most meaningful because
phenol is a decomposition product of the benzenediazo-
nium ion. In a typical trapping experiment, the nitrosa-
tion reaction was allowed to proceed for 15 min at 30 °C,
at which time a 2-fold excess of the trapping phenol was
added and stirring was continued for 1 h. Workup,
product identification, and quantitation were performed
in the same manner as described above with the use of
authentic standards. The major products are shown in
Scheme 5 for o-cresol. In all runs (3-5), the benzenedia-
zonium ion was trapped as the corresponding azo com-
pound in yields ranging from 89 to 93%. Similarly, the
yields of formanilide 15a , which is derived from the other
amidine molecular fragment, were 89-94%. As we dis-
cuss below, these data argue for, but do not require, the
formation of the diazonium ion by path A of Scheme 3.
In the discussion which follows, we reasonably assume
that the nitrosation of the amidine free base at either
nitrogen atom is fast and, in the case of the tertiary
amidines, reversible (see below). Nitrosation at either
nitrogen of the secondary amidines examined will ulti-
mately give the same products, even though nitrosation
of the imino nitrogen for all amidines will give rise to
the more stable cation due to the delocalization of the
charge over the two nitrogens and carbon which make
up the amidine moiety. The additon of water to the
central carbon followed by deprotonation then gives the
R-hydroxynitrosamine 9 (tetrahedral intermediate) de-
picted in Scheme 3. A key point supporting the argument
that the decomposition of this intermediate in the case
of the secondary amidines is occurring almost exclusively
by path A is the near equality of the yields of azo
compound and formanilide in each run (3-5). In our
preliminary work on amidine nitrosation, we isolated
N-nitrosopyrrolidine and N-benzyl-N-nitrosoformamide
from the nitrosation of the tertiary amidine N-benzyl-
formimidoylpyrrolidine.41 Because N-nitrosoamides de-
compose to diazonium ions with facility, it was unclear
as to which path was being followed. N-Aryl-N-nitroso-
amides are much less stable than their N-alkyl ana-
logues. The kinetics of decomposition of N-nitrosoforma-
nilide, N-nitrosoacetanilide, and related N-nitroso amides
have been examined in both protic and aprotic solvents.
3018 J . Org. Chem., Vol. 69, No. 9, 2004

Amidine Nitrosation
SCHEME 4
importantly, the near equivalent yields of the azo com-
pound and formanilide make this unlikely.
Further evidence on the partitioning of paths A and B
can be obtained from the data from the nitrosation of
tertiary amidines 8b, 8e, and 8f, as well as our prior
published data from the nitrosation of 8c, its p-chlo-
rophenyl derivative, and N-benzylformimidoylpyrroli-
dine.41 In these cases, the respective yields of the
nitrosamine 14 and the dialkylamide 11 (16 for 8e) give
minimal markers of the percent of the transformation
which is occurring by each path, A or B. The nitrosamine
arises from path B and the amide from path A. Material
balances for these dialkyl products range from 82 to 88%
(runs 10, 12, and 13). Because of their complete water
miscibility, dimethylnitrosamine and DMF are more
difficult to accurately quantitate in these assays than are
their pyrrolidine analogues, yet the material balances are
good. The greatest yield of nitrosamine (26%) was
observed for 8f, where the dialkylamine moiety is pyr-
rolidine. Thus, assuming that the transformations occur
as shown in Scheme 3, at least 26% of the reaction of 8f
must occur by path B. We previously reported that
N-nitrosopyrrolidine was obtained in 23% yield from the
nitrosation of N-benzylformimidoylpyrrolidine,41 which
agrees well with what we see for 8f. On the other hand,
the dimethylamino amidines 8b and 8e give less nitro-
samine, 8% and 2%, respectively, and correspondingly
larger yields of DMF and N,N-dimethylacetamide (Table
1, runs 10 and 12). The large relative yields of the amides
require that at least 74% and 82% of the nitrosative
transformations of 8b and 8e, respectively, are giving the
diazonium ion by path A. Thus path A, appears to be the
major transformation route for all compounds 8a -f.
SCHEME 5
The participation of path B increases as the amino
portion is changed in the order aniline < dimethylamine
< pyrrolidine. These limited data suggest that path B
becomes more prominent as the amino portion of the
amidine becomes more basic. Decomposition of the tet-
rahedral intermediate 9 by path B almost certainly
involves prior protonation of the amine nitrogen to
convert it into a better leaving group, as is observed in
both acid-catalyzed amidine and amide hydrolysis.45,46
The electron withdrawing PhNNO and OH substituents
attached to the central carbon will significantly reduce
the basicity of adjacent amino moiety of 9, but the
basicity of the amino nitrogen will still be proportional
to that of the related amine. The relative proportion of
the amino nitrogen protonated species, required for
transformation by path B, will then be greatest for the
pyrrolidine-derived amidines that show the largest yields
of nitrosamine.
It is, of course, possible that the observed transforma-
tions occur by pathways other than those depicted in
Scheme 3. The key steps of several possibilities are given
in Scheme 6. Nitrosation, rather than protonation of the
dialkylamino moiety of 9 (path C) could lead to the
nitrosamine directly by decomposition of the tetrahedral
intermediate 26. It is also possible, as shown in path D,
that nitrosation of the tetrahedral intermediate 27 at
either nitrogen atom, following reversible addition of H2O
Half-lives are typically less than 10 min at 25 °C.44 These
facts explain why we were unable to detect N-nitroso-
amides in this study compared to or prior work where
were able to detect N-benzyl-N-nitrosoformamide. In the
case of the nitrosation of 8a , decomposition by path A of
Scheme 3 gives the benzenediazonium ion and formanil-
ide, but the benzenediazonium ion would also form
readily in the medium by the diazotization of aniline
generated by decomposition of the tetrahedral intermedi-
ate 9 by path B. The latter transformation, however,
would produce N-nitrosoformanilide 12 (R1 ) H). In order
for 12 to give formanilide, the N-nitrosoamide would have
to undergo nearly quantitative denitrosation, a known
reaction in the presence of nucleophilic anions such as
chloride, in competition with the well-known decomposi-
tion of these intermediates to yield diazonium ions. The
weaker nucleophilicity of the acetate ion, and most
(43) Loeppky, R. N.; Shi, J . Unpublished results.
(44) Hey, D. H.; Stuart-Webb, J .; Williams, G. H. J . Chem. Soc. 1952,
4657-4665.
(45) Brown, R. S.; Bennet, A. J .; Slebocka-Tilk, H. Acc. Chem. Res.
1992, 25, 481-488.
(46) DeWolfe, R. H.; Keefe, J . R. J . Org. Chem. 1961, 27, 493-496.
J . Org. Chem, Vol. 69, No. 9, 2004 3019

Loeppky and Yu
SCHEME 6
to the protonated amidine, could lead to products. We
may expect path D1 to predominate because of the greater
nucleophilicity of the dialklyamino N, which may give
more 28 than 9. This would lead to large amounts of
nitrosamine being formed, which we do not observe. The
pathway has the virtue of generating the amide as a
coproduct without the necessity of being formed by
denitrosation of an N-nitrosoamide as is foreseen in path
B. In our prior publication,41 we postulated that the
chemistry shown in path E was operative. This process
also avoids N-nitrosoamide formation, but the yields of
the amides 15 are much smaller than those of the
nitrosamines, suggesting that the pathways in Scheme
3 are operative, rather than those of paths D1 and E.
While we have shown that an increase in the basicity
of the amino moiety of the amidine increases the par-
ticipation of path B in the competitive decomposition of
9, we cannot state that these intermediates will always
preferentially decompose by path A. The factors which
control decomposition by path A vs path B are subtle and
sensitive to structure. We have previously shown that
1-methyl-2-phenylimidazoline, a five membered cyclic
amidine, nitrosates and decomposes by path B.42 We
invoked stereoelectronic arguments, which are not ap-
plicable here, to explain this observation, but the basicity
of the nitrogen is also enhanced by the methyl substitu-
tion, favoring decomposition by path B through its
protonation. On the other hand, the acid-catalyzed
decomposition of both 2-benzyl- and 2-phenyl-N-nitrosoim-
idazoline occurs by path A. An examination of the 1,2,4-
triazole-catalyzed decomposition of N-methyl-N′-nitro-N-
nitrosoguanidine shows that the amine catalyst departs
as its anion from the tetrahedral intermediate at a rate
greater than the less basic methyldiazotate (basic me-
dia).47 In the reactions reported here, the preferred
leaving group from the tetrahedral intermediate 9 is
phenyldiazohydroxide and the acidic medium likely
ensures some form of acid catalysis for this process.
Conjugative interaction between the benzene ring and
the forming NdN bond may enhance the rate and thus
the “leaving group ability” of this moiety.
Rea ctivity a n d Kin etics. Reaction rates for the loss
of the starting amidine upon nitrosation, as determined
by HPLC or GC, were measured at 25 °C in glacial acetic
acid employing a 10- or 20-fold excess of sodium nitrite
so as to ensure pseudo-first-order conditions. Plots of ln-
([amidine]) vs time showed good linearity in all cases,
and are the result of two-three independent determina-
tions. The rate constants for the nitrosation of amidines
8a -f are reported in Table 2 in two ways. In the first
case the observed rate constant (kobs), as determined from
linear regression (slope) of the plotted data, is given. In
the second entry, k has been determined by calculation
from kobs by correction for the amidine basicity (free
amidine concentration). The values of amidine pKa’s used
were taken from the literature48,49 and are also given in
Table 2. Although pKa data for some amidines in aqueous
solution are available, for the sake of consistency we used
data where pKa’s were determined in 95% ethanol, which
have been reported for all of the compounds studied
except 8i. As a control we also made cursory determina-
tions of the approximate hydrolysis rates (data not given)
for amidines 8a -i under the acidic conditions used for
(47) Wichems, D. N.; Nag, S.; Mills, J .; Fishbein, J . C. J . Am. Chem.
Soc. 1992, 114, 8846-8851.
(48) Oszczapowicz, J .; Ciszkowski, K. J . Chem. Soc., Perkin Trans.
2 1987, 663-668.
(49) Oszczapowicz, J .; Krawczyk, W.; Lyzwinski, P. J . Chem. Soc.,
Perkin Trans. 2 1990, 311-314.
3020 J . Org. Chem., Vol. 69, No. 9, 2004

Amidine Nitrosation
TABLE 2. Am id in e Nitr osa tion Ra te Con sta n ts
a
compd
R1
R2
Ph
CH3
R3
pKa
kobs × 105 b (s-1
)
( × 105
28
1
k × 10c
5.2
7.4
36
2.2
5425
5425
( × 10
8a
8b
8c
8d
8f
H
H
H
CH3
CH3
CH3
Ph
Ph
Ph
H
CH3
5.38
1061d
13
0.1
0.8
2
0.1
556
556
7.45
8.12
6.97
10.6
10.6
6.15
7.8
(-CH2-)4
13.6
12.0d
6.8
6.8
no reaction
no reaction
0.8
0.8
0.7
0.7
Ph
H
(-CH2-)4
8f
(-CH2-)4
8g
8h
8i
Ph
H
CH3
CH3
(-CH2-)4
no reaction
Literature values, measured in 95% ethanol, see text. [amidine] ) 0.0363 M, [NO2-] ) 0.726 M, 25 °C in acetic acid/H2O, pH ) 3.7.
a
b
c kobs × [H+]/Ka. Because of the high reaction rate of 8a , kobs was determined at [amidine] ) 0.038 M and [NO2-] ) 0.414 M (10.9 equiv
d
of NO2-) and found to be 376 ( 10 × 10-5 s-1. kobs ) 4.3 ( 0.3 × 10-5 s-1 for 8 d was determined at [amidine] ) 0.0363 M and [NO2
]
-
) 0.392 M as well as the table entry. The ratio of these rate constants is 2.82 and kobs for 8a (entry 1) at the higher [NO2-] was estimated
by multiplying by 2.82 × 376 × 10-5 s-1
.
SCHEME 7
the important steps in amidine nitrosation. Step 1
involves N-nitrosation of the amidine free base by
50,51
N2O3
to give the more thermodynamically stable
N-nitrosoamidinium ion 31.52 In the case of trialkyl
tertiary amines, the initial nitrosation step is reversible,53
and it seems reasonable that that some of our substrates
will exhibit this property as well.
While the hydration of N-nitrosoiminium ions (2 f 3,
Scheme 1, where the C substituents are H or C) has been
shown to be very fast,9 depending upon R1, this may or
may not be the case for the analogous hydration of 31
shown as step 2 of Scheme 7, which is akin to the acid-
catalyzed hydration of a carboxylic acid derivative or an
amidine. The effective reversibility of this step, which in
the case of carboxylic acid derivatives can be measured
using 18O exchange, depends on the nature of the leaving
group, or in kinetic terms the relative magnitudes of k3
45
and k-2
.
Step 3 likely proceeds by either prior proton
loss from O and/or transfer to the O-atom of the NO of
the incipient diazohydroxide or, as discussed above, to
the amino nitrogen.
In the case of the tertiary benzamidines (R1 ) Ph), we
propose that their lack of their nitrosative reactivity is
due to a very small value of k2 in comparison to k-1 and
k-2. In other words, the N-nitrosoiminium ion 31 forms
reversibly but does not hydrate at an appreciable rate.
At analysis time, any 31 present will be converted back
to 8. This supposition is well supported by the low
reactivity to benzoic acid derivatives toward nucleophilic
acyl substitution reactions, which was discovered rela-
tively early in the history of physical organic chemis-
try.54,55 There are many examples. Benzamides and
benzamidines undergo acid-catalyzed hydrolysis slowly
in comparison to their aliphatic analogues.56 There is
significant evidence that this low reactivity stems from
the nitrosation. In all cases, hydrolysis did not compete
with nitrosation.
The completely inert character of the benzami-
dines 8g-i toward nitrosation is the most striking result
to emerge from this study. The nitrosation rates of many
nitrogen compounds in acid have been shown to be
inversely proportional to their basicity.22 N-Nitrosation
requires an unshared electron pair on an N-atom. Coor-
dination of this electron pair with a proton prevents
nitrosation until it is removed by some, even weakly basic
species. While N-nitrosation of many amines and some
other nitrogen compounds has been shown to effectively
be encounter controlled,50 the high basicity of these
compounds effectively makes the nitrosation of these
compounds rate limiting. The data of Table 2 show that
this is not the case for the benzamidines. The benz-
amidines 8g-h are less basic than their respective
acetamidine analogues 8d ,e which do nitrosate, so ami-
dine basicity cannot be the sole determinant of nitrosa-
tion rate and reactivity.
(51) Casado, J .; Castro, A.; Mosquera, M.; Rodriguez-Prieto, M.
Monatsh. Chem. 1984, 115, 669-682.
(52) Numerous studies in similar acid media strongly support our
assumption that the nitrosating agent is N2O3. At modest to high
[NO2-] relative to the [substrate], ONOAc is converted to N2O3. The
-
reverse step need not involve NO2 directly but may involve reaction
with H2O or another nucleophile. Nevertheless, the nature of nitrous
acid chemistry is such that the result will be the same, that is, the
regeneration of N2O3 and the amidine.
(53) Gowenlock, B.; Hutcheson, R. J .; Little, J .; Pfab, J . J . Chem.
Soc., Perkin Trans. 2 1979, 1110-1114.
(54) Hammett, L. P. Physical Organic Chemistry: Reaction Rates,
Equilibria, And Mechanisms, 1st ed.; McGraw-Hill: New York, 1940.
(55) Hine, J . S. Physical Organic Chemistry; McGraw-Hill: New
York, 1956.
In Scheme 7, and in the rate equations (1 and 2)
derived from it, we have sketched what we believe are
(50) Williams, D. H. L. Nitrosation; Cambridge University Press:
Cambridge, 1988.
(56) Smith, C. R.; Yates, K. J . Am. Chem. Soc. 1972, 94, 8811-8817.
J . Org. Chem, Vol. 69, No. 9, 2004 3021

Loeppky and Yu
the conjugative stabilization of the carbonyl, protonated
carbonyl, or their imino analogues by the aromatic ring.
The transformation may also be slowed by steric com-
pression in the tetrahedral intermediate.
differently to structural change of R1. From the hydrolysis
data we see that replacement of H by CH3 decreases the
rate of that reaction which is very similar to step 2 of
our transformation.46 On the other hand, CH3 substitu-
tion for H in R-hydroxynitrosamines, analogous to 31,
significantly increases the rate of their decomposition to
diazonium and carbonyl products.57 In addition to the
diazohydroxide, the immediate precursor to the diazo-
nium ion, this transformation generates a carbonyl
compound, which in our case is an amide. The thermo-
dynamic stability of the carbonyl compound is experi-
mentally observed to influence the rate of decomposition
of the R-hydroxynitrosamine from which it is derived.57
The more stable the incipient carbonyl compound is, the
more rapid the decomposition is, suggesting that these
transformations are thermodynamically neutral or have
some endothermic character. The transition state struc-
ture appears to be influenced byproduct structure. Thus,
CH3 for H substitution at the central carbon of the
amidine will retard the rate of H2O attack (k2) but is
expected to enhance the rate of 9 decomposition (k3).
While the individual rate constants may change signifi-
cantly with structure, their product may not.
The pyrrolidine-derived amidines 8c and 8f are unusu-
ally reactive as measured by their corrected rate con-
stants k. The acetamidine 8f is 150 times more reactive
than the formamidine 8c. These reactivity differences are
unexpected.58 While the diversity of substrates examined
here do not permit a complete understanding of the
reactivity differences, we offer the following rationaliza-
tion. As discussed above, these amidines react by path
B, producing N-nitrosopyrrolidine, to the extent of at
least 25%. But path B participation by itself cannot be
the sole reason for the rate enhancement or it would out
compete path A. It is probable that the more rapid
decomposition of the tetrahedral intermediate 9 leads to
the rate enhancement by both paths A and B. This could
result from the release of steric compression and “I
strain” in the conversion of the tetrahedral intermediate
to products of trigonal geometry. Reactions involving the
pyrrolidine ring system have been observed to be acceler-
ated when the nitrogen goes from tetrahedral to trigonal
geometry as in the case of path A where it is converted
from amine N to amide N, or in path B where it is
converted from protonated amine N to free amine N.59
This is due to the ring size enforced preference for atoms
to have a more trigonal geometry in the five-membered
ring.59
Pairwise comparison of the data of Table 2 for the
formamidines (8a -c) with their structurally analogous
acetamidines (8d -f) shows in each case that the forma-
midines are less basic and have higher values of kobs for
nitrosation than the corresponding acetamidines, as may
be expected. On the other hand, when the rate data are
corrected for the basicity of the amidine, so as to give
the rate constant k for the nitrosation of free amidine
(the intrinsic reactivity), the picture is less clear. De
Wolfe and Keefe showed that the hydrolysis of 8a is 1548
times more rapid than its acetamidine analogue 8d in
20/80 dioxane-H2O (0.42 M HCl) at 86 °C.46 Under these
conditions, the amidines are completely protonated and
the rate depends on the rate constant for the attack of
H2O on the amidinium ion. Both the inductive stabiliza-
tion of the cation by the methyl group and its greater
steric bulk retard the reaction of the acetamidine. We
might expect these structural differences to influence
amidine nitrosation rates in a similar way, but this is
not observed. Amidine nitrosation, however, is more
complicated and involves, at least, the three steps shown
in Scheme 7, and the competitive protonation equilibrium
of the amidine, which prevents nitrosation, also plays a
role. Different kinetic and mechanistic scenarios arise
depending upon how the equilibrium and rate constants
change with structure.
k1k2k3[Am][N2O3][H2O]
dP
dt
)
(1)
(2)
k-1(k-2 + k3)[NO2-] + k2k3[H2O]
- 2
k1k2k3[H+]2[Am][NO2
]
dP
dt
)
Ki2(k-1(k-2 + k3)[NO2-] + k2k3[H2O])
For discussion we assume the operation of rate eq 1
for the rate of product formation from the tertiary
amidines, where [Am] is the amidine concentration and
the equilibrium and rate constants are as defined in
Scheme 7. The nitrosative decomposition of the reactive
secondary amidines, even though they may form N-
nitrosoamidines, is presumed to obey a very similar rate
law since their protonation will generate a species 9.
Substitution for the equilibrium concentration of N2O3
([N2O3] ) KN[H+]2[NO2-]2/Ki2), ignoring its homolytic
decomposition, gives eq 2. We have assumed that the
nitrosation is rapid and reversible and that steady-state
concentrations can be defined for both the N-nitroso-
amidinium ion 31 and the tetrahedral intermediate 9.
Without more extensive experimentation and detailed
kinetic data, we can only make some pertinent observa-
tions. Because [NO2-] appears both in the numerator and
We also investigated the nitrosation of several imino-
N-acylamidines and amino-N-acylamidines. These com-
pounds hydrolyze much more rapidly than they nitrosate,
however. Details are given in the Supporting Informa-
tion.
Con clu sion s
-
denominator of eq 2, the reaction order in NO2 may
Both secondary and tertiary amidines are modestly
reactive toward nitrosation. Nitrosation rates are only
change from substrate to substrate and mask intrinsic
reactivity. For example, if k3 . k-2 and k-1[NO2-] . k2-
-
(57) Chahoua, L.; Cai, H.; Fishbein, J . C. J . Am. Chem. Soc. 1999,
121, 5161-6169.
(58) While an error in, or the inappropriate use of, pKa values
determined in 95% ethanol to correct our kobs values for amidine
protonation could be responsible, the pKa values were determined by
the same group using the same methodology.
[H2O] then the rate will be first order in NO2 and
hydration of 31 will be rate limiting (see the Supporting
Information).
The similar magnitudes of k most plausibly result from
cases where k is dominated by k2k3. Literature precedent
suggests that these two rate constants will respond
(59) Kresge, A. J .; Fitzgearld, P. H.; Chiang, Y. J . Am. Chem. Soc.
1974, 96, 4698-4699.
3022 J . Org. Chem., Vol. 69, No. 9, 2004

Amidine Nitrosation
dine,65 8h ; 1-[phenyl(phenylimino)methyl]-pyrrolidine,66 8i;
and N,N-dimethyl-N′-benzoylformamidine.67
roughly correlated with the basicity of the amidine. The
values of kobs varied at most by a factor of 100 where the
amidine Ka values differed by as much as 105, where the
more basic amidines reacted more rapidly than expected
in comparison to a variety of other nitrogen compounds
where N-protonation effectively reduces the nitrosation
rate. Diazonium ions are the primary products of amidine
nitrosation for all substrates examined here and are
formed by path A of Scheme 3. The tertiary amidines also
form nitrosamines by path B where the pyrrolidine
derived amidines show a greater tendency to produce
products by this route. The factors which control decom-
position by path A vs path B are subtle and sensitive to
structure. The benzamidines examined are unreactive
toward nitrosation. This must be due to a greatly reduced
rate of addition of H2O to the N-nitrosoamidinium ion
31. Formamidine and acetamidine nitrosation rates are
comparable, in contrast to their acid-catalyzed hydrolysis
rates. We attribute this phenomenon to the counteraction
of substituents at R1 on the rate of H2O addition (k2)
compared to the rate of tetrahedral intermediate decom-
position (k3). N-Acylamidines hydrolyze more rapidly
than they nitrosate, but some of the hydrolysis products
are easily nitrosated. Although all of the work presented
here was done at high nitrite concentrations, it is clear
that most N-arylamidines react with nitrosating agents
to produce electrophilic diazonium ions, while nitro-
samines are generated in lesser yields. We have also
shown previously that N-nitrosoamides are products of
these transformations. Thus, amidines must be viewed
as nitrosatable compounds capable of generating mu-
tagens and possible carcinogens.
Am id in e Nitr osa tion . All amidines 8a -i were nitrosated
under similar conditions in glacial acetic acid with an excess
of aqueous sodium nitrite. The precise molar equivalents of
sodium nitrite, the reaction temperature, and the reaction time
are given for each run along with the product yields in Table
1. The reaction products are all known, well-characterized
commonly available compounds. 2,4-(Bis)phenylazophenol68
and 2-methyl-4-azophenylphenol 2569 were prepared by the
literature methods as cited. In every case, the compounds were
obtained and their 1H, 13C NMR spectra and mass spectra
examined to ensure conformity with literature values. These
authentic compounds were used as chromatographic stan-
dards. In all cases, the reaction mixtures were submitted to
flash column chromatography following extraction to isolate
and characterize the major components. Depending upon the
goal of the particular run, the products were also submitted
to GC-MS analysis and/or HPLC RP chromatography. Quan-
titation was done through the use of external standards.
Typ ica l Nitr osa tion . N,N-Dimethyl-N′-phenylformami-
dine 8b (0.57 g, 3.8 mmol) was dissolved in 10 mL of glacial
acetic acid. To the solution was added dropwise 7 mL of 5.5 M
aqueous sodium nitrite solution (38.0 mmol) over a 3 min. The
mixture was shielded from light by aluminum foil and stirred
at room temperature for 69 h. The solution was neutralized
to pH 7 with 6 g of potassium carbonate. The aqueous phase
was extracted with methylene chloride (3 × 50 mL). The
combined organic layers were washed with brine and dried
(MgSO4). The filtered solution was diluted to 250 mL in a 250
mL volumetric flask (extract 1). The aqueous phase was
basified with 30 g of potassium carbonate and 2 g of sodium
hydroxide. This mixture was extracted with methylene chlo-
ride (3 × 50 mL). The combined organic layers were dried
(MgSO4), filtered, and diluted to 250 mL in
a 250 mL
volumetric flask (extract 2). To determine the amount of
phenyl acetate and DMF in extract 1, the solution was diluted
five times. To determine the amount of unreacted starting
material in extract 2, the solution was diluted 10 times. The
amounts of dimethylnitrosamine 14b (0.2 mmol, 8%), forma-
nilide 15 (0.08 mmol, 3%), DMF 16 (2 mmol, 74%), phenol 17
(0.2 mmol, 8%), phenyl acetate 18 (0.8 mmol, 30%), 2-nitro-
phenol 23 (0.14 mmol, 5%), and 4-nitrophenol 22 (0.1 mmol,
4%) were determined by GC-FID (40 °C, 5 °C/min to 250 °C).
Retention times (min): 14b, 2.8; 16 (DMF), 3.4; 17, 7.8; 18,
10.2; 23, 12.2; 15 (formanilide), 17.7; 8b, 19.4; and 22, 23.3
min). After vacuum-assisted removal of the solvent, the
resulting oil was subjected to flash column chromatography
on silica gel.
La ck of Ben za m id in e 8g-i Nitr osa tive Rea ctivity. The
benzamidines 8g-i were nitrosated under the conditions
described above for at least 4 h. In each case, the parent
amidine was recovered in near-quantitative yield and no
products could be observed in the extracts.
Dia zon iu m Ion -Tr a p p in g Exp er im en ts. In runs 3-5
(Table 1), N,N′-diphenylformamidine 8a was subjected to
nitrosation under conditions where the benzenediazonium ion
could be trapped as quantitatively as possible by azo coupling
with either added phenol or o-cresol as described. Phenol (2
equiv), which is also a reaction product, was used in run 3,
and o-cresol was utilized in runs 4 and 5. The conditions for
run 4 were as follows. Compound 8a (1.0 g, 5.1 mmol) was
dissolved in 7.5 mL of acetic acid. The solution was stirred in
Exp er im en ta l Section
Ca u t ion : Nitrosamines, N-nitrosoamides, N-nitrosoami-
dines, 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 bring the solution to pH 12-13 and reaction with
Raney-nickel.
Syn th esis of Am id in es 8b-i a n d N-Acyla m id in es 32b
a n d 33a -c. All of the amidines and N-acylamidines used in
this work are known compounds. These substrates, as indi-
vidually cited, were synthesized for use in this work by
literature methods or slight variations thereof. In all cases the
physical and spectral properties were consistent with those
reported in the literature. Additional spectral data for each
compound are given in the Supporting Informaiton. The
compounds are as follows: N,N′-diphenylformamidine, 8a
(commercially available), N,N-dimethyl-N′-phenylformami-
dine,60 8b; 1-[1-(phenylimino)methyl]-pyrrolidine,61 8c; N,N′-
diphenylacetamidine,62 8d ; N,N-dimethyl-N′-phenylacetami-
dine,60 8e; 1-[1-(phenylimino)ethyl]-pyrrolidine63 8f; N,N′-
diphenylbenzamidine,64 8g; N,N-dimethyl-N′-phenylbenzami-
(65) Naulet, N.; Filleux, M. L.; Martin, G. J .; Pornet, J . Org. Magn.
Reson. 1975, 7, 326-330.
(60) Wawer, I. Pol. J . Chem. 1988, 62, 223-229.
(61) Bredereck, H.; Gompper, R.; Klemm, K.; Rempfer, H. Chem.
Ber. 1959, 92, 837-849.
(66) J ensen, K. G.; Pedersen, E. B. Acta Chem. Scand., Ser. B 1979,
B33, 319-321.
(62) Barker, J .; J ones, M.; Kilner, M. Org. Mass Spectrom. 1985,
20, 619-623.
(67) Lin, Y.-i.; Lang, S. A.; Lovell, M. F.; Perkinson, N. A. J . Org.
Chem. 1979, 44, 4160-4164.
(63) Oszczapowicz, J .; Raczynska, E.; Osek, J . Magn. Reson. Chem.
1986, 24, 9-14.
(64) Hontz, A. C.; Wagner, E. C. In Organic Syntheses; Rabjohn,
N., Ed.; J ohn Wiley & Sons: New York, London, 1963; Vol. IV, p 383.
(68) Gore, T. S.; Inamdar, P. K.; Patwardhan, A. V. Indian J . Chem.
1970, 8, 195-197.
(69) Lippmaa, E.; Pehk, T.; Saluvere, T.; Magi, M. Org. Magn. Reson.
1973, 5, 441-444.
J . Org. Chem, Vol. 69, No. 9, 2004 3023

Loeppky and Yu
an oil bath at 30.5 °C. To the solution sodium nitrite (1.8 g,
26.1 mmol) aqueous solution (3.0 mL water) was added all at
once. After the addition of sodium nitrite, the mixture was
stirred for 15 min. To the mixture was added o-cresol (1.1 g,
10.2 mmol). The resulting mixture was stirred for 1 h. To the
mixture was added 20 mL of ice water. The resulting mixture
was extracted with benzene (4 × 20 mL). The combined organic
layers were washed with brine and dried (MgSO4). An aliquot
of the benzene extract was injected in GC-MS. 4-(Phenylazo)-
2-cresol 25, formanilide 15, and 4-nitroso-2-cresol were identi-
fied in the chromatogram. Benzene was evaporated in vacuo.
The resulting oil was purified by column chromatography
(ethyl acetate-hexane, 1:6) to give 0.98 g of 25 (91%), 0.58 g
of 15 (94%), and 0.65 g of 4-nitroso-2-cresol.
Kin etics of th e Nitr osa tion th e Am id in es 8a -f w ith a
La r ge Excess of Sod iu m Nitr ite. Typical kinetics proce-
dure: The amidine was dissolved in 12.0 mL of glacial acetic
acid and 10.0 mL of water. The pH of the solution was adjusted
to 3.7 by 5 N NaOH solution (the concentration of the amidine
was 0.036 M). The amidine solution was stirred in an oil bath
at 25.0 °C. To the solution 5 mL of sodium nitrite aqueous
solution (20 equiv) was added all at once. During the reaction
course, 10-12 aliquots (∼500 µL) were taken from the bulk
solution at specific times (12 min intervals for 8a and 1 h
intervals for the other amidines). Each sample solution was
quenched by base (for 8a, sodium carbonate; for other amidines,
sodium hydroxide). The resulting basic aqueous solution was
extracted with methylene chloride (3 × 15 mL). The combined
organic layers were washed with brine and dried (MgSO4).
Methylene chloride was evaporated in vacuo. The residue was
dissolved in acetonitrile in a 50 mL volumetric flask (for 8a
the methylene chloride extract was diluted to 50 mL with
methylene chloride in a 50 mL volumetric flask). The concen-
tration of the amidine was determined by HPLC (for 8a by
GC-FID) with external standards of the amidine (mobile
phase: acetonitrile & 0.025 M sodium phosphate buffer; flow
rate: 1 mL/min). Four standard solutions of the amidine were
used to make a calibration curve. The concentration of the
amidine in each sample solution was determined by the
calibration curve. The rate constants reported in Table 2 were
obtained from the slopes of linear regressions of ln([amidine])
vs time and are the averages of several runs for each substrate.
Ack n ow led gm en t. The support of this research by
Grant Nos. R37 CA26914 and RO1 CA85538 from the
National Cancer Institute, NIH, is gratefully acknowl-
edged. R.N.L. also expresses his gratitude to Dr. J im
Iley of the Open University, U.K., for his most helpful
discussions regarding reactivity issues.
Su p p or tin g In for m a tion Ava ila ble: Details of N-acyl-
amidine preparation, attempted nitrosation, and hydrolysis;
additional spectral data for the compounds used in this
research; control hydrolysis experiment; abbreviated deriva-
tion of eq 1 and presentation of limiting cases. This material
is available free of charge via the Internet at http://pubs.acs.org.
J O035884U
3024 J . Org. Chem., Vol. 69, No. 9, 2004