decreased until there was a levelling off below pH ∼3. Below
about pH 1 there is no evidence of any reaction at all.
similar to that encountered earlier using SNCys as the source of
the nitrosating species, ie a fairly rapid build-up followed by
a slower decay. Fig. 6 suggests that we are looking at an equi-
librium process here, generating the highest product yield at the
Discussion
highest [HNO ] in very large excess over uric acid. The pH
2
It is clear from the results with NMA that nitrosation here
dependence experiments gave complicated results, but it was
clear that at the higher pH, reaction is much slower as expected
2ϩ
occurs by initial formation of nitric oxide, via the Cu -
catalysed route (eqns. (1)–(3)), since reactions are completely
inhibited in the presence of EDTA. Subsequent oxidation to
nitrogen dioxide and formation of dinitrogen trioxide N O ,
when [HNO ] is much reduced (pK 3.1). At the lower pH, more
2
a
product is generated (and more rapidly), but the rate of decom-
position is also increased. The result of the balance here is that
no product is seen at pH values <∼1. The crude initial rate
measurements showed as expected that the rate increased from
2
3
allows the possibility that either or both of these nitrogen
oxides can bring about N-nitrosation of NMA; both are well-
19
known nitrosating species. Under these conditions we favour
pH ∼5, as the concentration of HNO
2
increased, and appeared
the pathway via N O , since nitrite ion is generated as the
to level off above pH ∼3. This suggests that when all the nitrous
2
3
ϩ
inorganic product and not a mixture of nitrite and nitrate ions
which would be expected from the hydrolysis of NO /N O .
acid has been released, reaction occurs either between H
2
NO
NO and an ionised form of uric acid, or more likely, given the
low acidity and relatively high [NaNO ], that reaction occurs
between N and a neutral form of uric acid. Clearly more
2
/
ϩ
2
2
4
Reaction rates are drastically reduced in the absence of oxygen.
Thus we can eliminate, under these reaction conditions, the
2
O
3
2
ϩ
slower pathway involving direct NO transfer from RSNO to
experimental work is necessary to get a full mechanistic picture.
A reasonable rationalisation for this unexpected reaction
with uric acid, given our results using SNCys and also with
the amine.
14,15
Recent publications
solutions of nitric oxide in water readily nitrosate both amines
typically morpholine) and thiols (typically cysteine). Further,
have shown that oxygenated
HNO
, is that we are looking here at the nitrosation of an
2
22,23
(
amide (albeit a cyclic one). We and others
have found that
the rate law is identical with that obtained earlier for the
auto-oxidation of NO, and the reaction rate is independent of
the nature and concentration of the thiol or amine present.
In our case the rate limiting step is the release of NO from
there are in general significant differences between the nitro-
sation of amines and of amides. For the latter, there appears to
be no nucleophilic catalysis, but there is base catalysis and a
substantial kinetic solvent isotope effect, and for secondary
amides the reaction is significantly reversible. We proposed a
mechanism which involves rapid and reversible N-nitrosation
followed by rate-limiting proton transfer to the solvent. The
change of rate-limiting step (when compared with amine nitros-
ation) is as a result of the powerful electron-attracting effect
of the carbonyl group. Sulfonamides are believed to react in a
similar fashion as do amines containing powerful electron
attracting substituents, e.g. 2,4-dinitroaniline. The mechanism
is borne out by measurements on the reverse reaction, i.e., the
2ϩ
RSNO. This is clear from the data on the Cu catalysis of the
reactions of SNAP with NMA (Figs. 2 and 3) where the
2ϩ
dependence upon [Cu ] is evident.
Reactions with phenol gave the expected 4-nitrosophenol
product but in much lower yield than for the corresponding
reactions with NMA. This is readily explained in terms of the
expected competition between N O3 nitrosation and its
2
hydrolysis to nitrite (eqn. (5)).
Ϫ
ϩ
22
N O ϩ H O
2HNO2
2NO2 ϩ 2H
(5)
denitrosation of nitrosamides. With this mechanism it is not
2
3
2
possible to make any deductions about the nature of the nitro-
sating species, since the nitrosation step is reversible (and the
next stage is rate-limiting), reaction will always be first-order
It is well-known that NO in aerated water generates a quan-
titative yield of nitrite ion. For the NMA reactions under our
experimental conditions the competition very much favoured
the nitrosation reaction, perhaps as expected since the reaction
of N O with NMA is known to occur at or close to the dif-
ϩ
ϩ
in [HNO ] regardless of whether the reagent is H NO /NO ,
2
2
2
N O , or ClNO etc. It has been argued that initial nitrosation
2
3
takes place at the carbonyl oxygen atom of amides, which is
2
3
20
23
fusion controlled limit. There appears to be no corresponding
literature value for the phenol, but it is known that phenol is
orders of magnitude less reactive than NMA towards other
followed by a O- to N- rearrangement. Such rearrangements
are not uncommon in nitrosation reactions.
We propose that uric acid undergoes N-nitrosation at one of
the amide nitrogen atoms, by N O , derived either from RSNO
21
nitrosating species, such as the nitrosyl halides. Competition
with hydrolysis is then expected to be less effective for phenol.
As expected, increasing [phenol] does increase the percentage
of 4-nitrosophenol (see Table 1).
2
3
species via the intermediate formation of NO, or from relatively
high nitrite ion concentrations in the pH region around 4.
The position of nitrosation is purely speculative at this stage.
The N-nitrosamide then undergoes acid-catalysed denitro-
sation, to such an extent that at pH < ∼1, no N-nitrosamide
can be detected. For this nitrosouric acid derivative, there may
also be some homolytic decomposition, since some nitric oxide
was detected, but was not quantified. This fits in with the fact
that N-nitrosamides generally are more usually synthesised, not
from the more conventional nitrous acid in acidic media (as are
N-nitrosamines), but by using N O , N O or NOCl in, e.g.,
The situation with uric acid is rather unusual. As far as we
are aware there is no reference to the formation of a nitroso
derivative of uric acid in the literature. We were somewhat
surprised therefore to see a fairly rapid build-up of an absorb-
ance at 390 nm from the reaction of S-nitrosothiols with uric
acid at pH 7.4. This product is quite likely to be a nitroso
derivative, but this was quite unstable under the reaction con-
ditions and so no product could be isolated. The maximum
yield of this unknown product increased with [uric acid] when
SNCys was in a large excess. Again there was no discernible
reaction when EDTA was present demonstrating that as for the
2
3
2
4
carbon tetrachloride under non-acidic conditions, alkyl nitrites
in organic solvents or nitrosonium salts in acetonitrile. It was
possible to measure the formation of N-methyl-N-nitrosourea
at low acidities (pH ∼2), and to examine the reverse reaction
2ϩ
reactions of NMA and phenol, Cu is essential, even at the low
level impurity from the distilled water and buffer components,
to generate NO from SNCys as the first step. In order to probe
this further we carried out some preliminary experiments on the
22
at much higher acidities.
Experimental
attempted nitrosation of uric acid using the more usual HNO /
2
ϩ
Ϫ3
H method. In moderately acidic solution (0.4 mol dm ) there
was no evidence of any reaction whatsoever and we could only
see the fairly rapid build-up of the 390 nm absorbance when the
acidity was reduced to pH ∼ 4. Here the behaviour was very
The S-nitrosothiols were all synthesised by conventional
nitrous acid nitrosation of the corresponding thiol. The pH of
the solution was adjusted and the solution used without further
treatment. All other materials were commercial samples of the
J. Chem. Soc., Perkin Trans. 2, 2002, 1834–1838
1837
Noble, Darren R.
