A kinetic study of S-nitrosothiol decomposition

Article abstract of DOI:10.1002/1521-3765(20020118)8:2<380::AID-CHEM380>3.0.CO;2-P

Under anaerobic conditions S-nitrosothiols 1a-e undergo thermal decomposition by homolytic cleavage of the S-N bond; the reaction leads to nitric oxide and sulfanyl radicals formed in a reversible manner. The rate constants, k₁, have been determined at different temperatures from kinetic measurements performed in refluxing alkane solvents. The tertiary nitrosothiols 1c (k_1(69°C) = 13 × 10^-3 min^-1) and 1d (k_1(69°C) 91 × 10^-3 min^-1) decomposed faster than the primary nitrosothiols 1a (k_1(69°C) = 3.0 × 10^-3 min^-1) and 1b (k_1(69°C) = 6.5 × 10^-3 min^-1). The activation energies (E^#= 20.5 - 22.8 Kcal mol^-1) have been calculated from the Arrhenius equation. Under aerobic conditions the decay of S-nitrosothiols 1a-e takes place by an autocatalytic chain-decomposition process catalyzed by N₂O₃. The latter is formed by reaction of dioxygen with endogenous and/or exogenous nitric oxide. The autocatalytic decomposition is strongly inhibited by removing the endogenous nitric oxide or by the presence of antioxidants, such as pcresol, β-styrene, and BHT The rate of the chain reaction is independent of the RSNO concentration and decreases with increasing bulkiness of the alkyl group; this shows that steric effects are crucial in the propagation step.

Full text of DOI:10.1002/1521-3765(20020118)8:2<380::AID-CHEM380>3.0.CO;2-P

FULL PAPER
A Kinetic Study of S-Nitrosothiol Decomposition
Loris Grossi* and Pier Carlo Montevecchi*[a]
Abstract: Under anaerobic conditions
S-nitrosothiols 1a e undergo thermal
decomposition by homolytic cleavage
(k1(698C) ˆ 3.0  10À3 minÀ1)
and
1b
latter is formed by reaction of dioxygen
with endogenous and/or exogenous ni-
tric oxide. The autocatalytic decomposi-
tion is strongly inhibited by removing
the endogenous nitric oxide or by the
presence of antioxidants, such as p-
cresol, b-styrene, and BHT. The rate of
the chain reaction is independent of the
RSNO concentration and decreases with
increasing bulkiness of the alkyl group;
this shows that steric effects are crucial
in the propagation step.
(k1(698C) ˆ 6.5  10À3 minÀ1). The activa-
tion energies (E# ˆ 20.5 22.8 KcalmolÀ1)
have been calculated from the Arrhe-
nius equation. Under aerobic conditions
the decay of S-nitrosothiols 1a e takes
place by an autocatalytic chain-decom-
position process catalyzed by N2O3. The
À
of the S N bond; the reaction leads to
nitric oxide and sulfanyl radicals formed
in a reversible manner. The rate constants,
k1, have been determined at different
temperatures from kinetic measure-
ments performed in refluxing alkane
solvents. The tertiary nitrosothiols
1c (k1(698C) ˆ 13  10À3 minÀ1) and 1d
(k1(698C) ˆ 91  10À3 minÀ1) decomposed
faster than the primary nitrosothiols 1a
Keywords: chain decomposition
nitrogen oxides radical ions
radicals ¥ S-nitrosothiols
¥
¥
¥
14]
has been suggested as well.[12 In contrast to the homolytic
mechanism, it has been reported that the rate of disappear-
ance of RSNOs depends on their concentration[15] and on the
presence of air.[16] Moreover, Williams et al. reported that, in
aqueous solutions, RSNOs can undergo a copper(ii)-catalyzed
reaction as the main decomposition route.[2, 17]
Introduction
Even though the existence of S-nitrosothiols (RSNOs) has
been known for a long time,[1, 2] not very much was reported
on their chemistry until the early nineties, mainly because
their use for synthetic purposes was thought to be hindered by
their instability. In the past decade the interest in these
compounds has remarkably increased due to the discovery of
their involvement in several biological processes.[3, 4] In
particular, S-nitroso derivatives of essential biomolecules,
like cysteine and glutathione, seem to act as in vivo biological
carriers of NO,[5 the role of which as a neurotransmitter is
now well documented.[9]
Several studies have been conducted with the aim of
clarifying the mechanism of the thermal decomposition of S-
nitrosothiols, which is crucial for the understanding of how the
NO transportation in the human body takes place; but many
aspects still remain unknown.
RSNOs can exist in two different isomeric forms, syn and
anti, due to possible partial double bond character between
the sulfur and nitrogen atoms.[18] Primary and secondary
RSNOs exist preferentially as syn conformers that are red
colored, whilst tertiary RSNOs have a preference for, for
steric reasons, the anti conformation that is green colored;
these latter have been reported to be more persistent than the
red ones, even though a reasonable explanation for this
experimental evidence is not given.[18, 19]
Herein, we report results[20] obtained from kinetic studies
on the decomposition of a number of simple S-nitrosothiols,
synthesized from the parent thiols and peroxynitrite, as we
recently reported.[21] Our aim was to throw light onto the
8]
The decomposition of RSNOs has been generally assumed
À
to take place by a unimolecular mechanism through the
mechanism of the S N bond cleavage and to point out the
effect of the concentration, the presence of dioxygen, nitric
oxide, and antioxidants.
cleavage of the weakS N bond,[10, 11] but a heterolytic process
À
[a] Prof. L. Grossi, Prof. P. C. Montevecchi
Dipartimento di Chimica Organica ™A. Mangini∫
Results and Discussion
¡
Universita di Bologna
Viale Risorgimento 4, 40136 Bologna (Italy)
Fax : (‡390)51-209-3654
For this study we have considered the two primary RSNOs 1a
and 1b, which are red colored, the two tertiary RSNOs 1c and
1d, which are green colored, and the red aryl RSNO 1e. These
E-mail: grossi@ms.fci.unibo.it
montevec@ms.fci.unibo.it
380
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0380 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 2

380 387
compounds were synthesized by treating at 5 108C a solution
of the corresponding thiol in acetonitrile with hydrochloric
acid and a molar equivalent of an aqueous solution (0.5 mm,
pH ˆ 13.5) of peroxynitrite.[21] Under these conditions S-
triphenylmethylnitrosothiol (1d) was separated as a green
solid in 60% yield, whereas S-nitrosothiols 1a c,e were
extracted from the aqueous solution with the appropriate
organic solvent, and the concentration was measured by
spectrophotometry. Kinetic experiments were performed on
solutions of 1a e (12 mm) either under aerobic and anaero-
bic conditions. The decay of RSNOs was monitored by
spectrophotometry at l ˆ 550 nm for 1a[22, 23] and 1b,[22] l ˆ
603 nm for 1c[23] and 1d, and l ˆ 570 nm for 1e.[24]
In a preliminary paper we have already discussed the
[20]
À
possible reversibility of the S N bond scission.
In the
present workwe found definitive evidence based on the
following findings: i) a solution of 1b, carefully deaerated at
room temperature by bubbling argon and then by saturation
with gaseous NO, was found to be fairly stable at 708C over
20 h, whereas complete decomposition occurred within 3 h in
the absence of added nitric oxide; this finding clearly
indicated that, in agreement with Equation (1), the rate of
decomposition of RSNOs decreases with increasing nitric
oxide concentration; ii) a solution of (triphenylmethyl)ni-
trosothiol (1d) decomposed in an argon atmosphere and
then it was saturated with nitric oxide. A green color, due
to the re-formation of 1d, immediately developed. This
finding was explained in these terms: decomposition of
1d leads to nitric oxide, removed by the stream of argon,
Decomposition of S-nitrosothiols under anaerobic conditions:
In preliminary experiments, solutions of 1a c in n-octane
were carefully deaerated by bubbling argon at room temper-
ature and then they were kept in a thermostatic bath. The
solutions were vigorously stirred and kept under a weak
stream of argon until the end of the experiment. Samples were
taken at fixed time intervals up to 200 min. Several experi-
ments were conducted at different temperatures in the range
40 708C. In all cases the decomposition of 1a c was found
to follow a first-order kinetic law.
.
and (triphenylmethyl)sulfanyl radicals (Ph3CS ), which are
known to exist in equilibrium with the corresponding
disulfide 2d.[26] Subsequent addition of gaseous nitric oxide
promoted the re-formation of 1d by reversible coupling of
.
NO with Ph3CS .[27]
From these overall results we argued that the first-order
À
kinetic rate constants, k1, of the unimolecular S N bond
scission could not be obtained under the above experimental
conditions. So, we thought it prudent to run kinetic experi-
ments in refluxing solvents, since under these conditions the
concentration of nitric oxide can be considered to be almost
zero, and then Equation (1) becomes v ˆ k1[RSNO]. Accord-
ingly, solutions of 1a d in the appropriate solvent (cyclo-
pentane: b.p. ˆ 508C; hexane: b.p. ˆ 698C; cyclohexane:
b.p. ˆ 80.58C; and/or methylcyclohexane: b.p. ˆ 1018C) were
at first deaerated at room temperature by bubbling with argon
and then heated to a moderate reflux in the darkand in an
argon atmosphere. Each kinetic experiment was repeated
three times. We obtained first-order kinetic rate constants
(Æ2% error) that we assumed to be the ™true∫ k1 values (see
Table 1).
Apparently, these findings are those expected from an
À
unimolecular homolytic scission of the S N bond, which leads
to nitric oxide and sulfanyl radicals. In agreement with this,
GC-MS analyses of the reaction mixtures, which resulted after
complete decomposition had been monitored by the disap-
pearance of the color, showed the formation of the disul-
fide 2a c as the unique reaction product (Scheme 1).
k1
k–1
k2
RSSR
2
NO
RS•
+
RSNO
1
a: R = PhCH2; b: R = n-hexyl; c: R = tert-butyl; d: R = Ph3C; e: R = Ph
Scheme 1. Formation of the disulfide 2.
From the k1 values at different temperatures we could
#
À
estimate the activation energy for the S N bond scission, E ,
through the Arrhenius equation (ln k ˆ ln A À E#/RT) (Ta-
ble 1 and Figure 1). Our E# values appear to be lower (ca. 3
4 kcalmolÀ1) than the dissociation bond energies (DBEs)
reported in the literature for related nitrosothiols.[28, 29] It is
notable that the reported DBEs do not seem to be affected by
the nature of the alkyl group. In sharp contrast to this report,
as well as to the claim that the bulky, green nitrosothiols are
However, we found that the kinetic constant values
depended a great deal (up to one order of magnitude) on
the experimental conditions. First of all, reactions conducted
under the normal laboratory lighting were faster as compared
with those conducted in the dark, and this showed a significant
contribution of the photochemical-induced decomposition.[25]
Furthermore, the decomposition rate was found to increase
considerably by increasing the speed of stirring as well as the
flow of argon. We accounted for this surprising behavior by
Table 1. Kinetic constants (k1 Æ 2%), activation energies (E# Æ
0.4 KcalmolÀ1), and frequency factors (ln A Æ 0.6) for the decomposition
of RSNOs 1a d under an argon atmosphere in refluxing solvents (A:
cyclopentane, b.p. ˆ 323 K; B: hexane, b.p. ˆ 342 K; C: cyclohexane, b.p. ˆ
353.5 K; D: methylcyclohexane, b.p. ˆ 374 K).
À
assuming that the unimolecular S N bond scission is a
reversible process (Scheme 1). In this assumption, and in the
.
steady-state approximation for RS , the rate of decomposition
of RSNO depends on the NO concentration in Equation (1).
RSNO
k1 Â 103 minÀ1
E#
ln A
A
B
C
D
[KcalmolÀ1
]
v ˆ k1[RSNO]{1 À kÀ1[NO]/(kÀ1[NO] ‡ k2[RS])}
(1)
1a
1b
1c
1d
0.90
6.5
18.0
8.8
37.0
105
53
202
22.3
22.8
21.9
20.5
27.7
27.7
27.7
27.7
3.0
13.0
91
We reasoned that an increase of the rate of stirring and/or
the flow of argon caused an increase in the decomposition rate
by favoring the escape of NO from the reaction vessel.
15.4
Chem. Eur. J. 2002, 8, No. 2
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0381 $ 17.50+.50/0
381

P. C. Montevecchi and L. Grossi
FULL PAPER
Figure 1. ln k versus 1/T for the thermal decomposition of 1a (curve b), 1b
(curve a), 1c (curve c), and 1d (curve d) carried out under anaerobic
conditions and in refluxing solvents (cyclopentane: b.p. ˆ 508C; hexane:
b.p. ˆ 698C; cyclohexane: b.p. ˆ 80.58C; methylcyclohexane: b.p. ˆ
1018C).
Figure 2. Kinetic curves obtained from the decomposition under aerobic
conditions of solutions of S-nitrosothiols 1a e in n-pentane at 08C:
a) 12 mm 1c; b) 12 mm 1d; c) 12 mm 1b; d) 16 mm 1a; e) 12 mm 1a;
f) 6 mm 1a; g) 12 mm 1e.
more stable than the nonbulky, red ones,[18, 19] our findings
clearly showed that the thermal stability of RSNOs decreases
with increasing bulkiness of the alkyl group.
concentration. The higher stability exhibited by the bulky S-
nitrosothiols 1c,d, which contrasts with the lower thermal
stability exhibited under anaerobic conditions (see above),
should be ascribed to steric effects in the chain-propagation
step. Also, it is worth noting the peculiar trend exhibited by
the kinetic curve obtained from 1d (Figure 2, curve b). After
an initial period of time, in which the forecast autocatalytic
behavior was observed, the increase in the chain-reaction rate
tended to slow down. This trend led us to suppose that a
product arising from 1d inhibited the chain-decomposition
reaction (see later for discussion).
In order to verify if the RSNO concentration can influence
the rate of decomposition, as previously reported,[15] kinetic
experiments were also conducted with 16 and 6 mm solutions
of 1a in n-pentane. But, no variation in the kinetic behavior
was observed when compared with the 12 mm solution
(Figure 2, curves d f).
Attempts to perform kinetic experiments in refluxing
solvents with S-benzenenitrosothiol 1e failed because of its
thermal instability. Actually, it was possible to run experi-
ments only at 158C with solutions of 1e prepared under an
argon atmosphere. Under these conditions a first-order
decomposition rate was observed (kexp ˆ 35  10À3 minÀ1).
TLC and GC-MS analyses of the reaction mixture showed
the disulfide 2e to be the unique reaction product. On the
assumption that kexp ˆ k1 and ln A ˆ 27.7 (see Table 1), we
could estimate E# ˆ 19.1 kcalmolÀ1 (the DBE value reported
in the literature is 19.4 kcalmolÀ1).[28] The lower thermal
stability of 1e, as compared with that of 1a d, must be
attributed to the higher stability[29] of arenesulfanyl radicals
.
.
(ArS ) as compared with that alkanesulfanyl radicals (RS ).
To determine the effect of the temperature on the decom-
position rate, kinetic measurements were also conducted at
18, 33, and 508C with solutions of 1a in n-octane. The kinetic
curves (Figure 3) show that an increase of the temperature
from 0 to 508C caused the expected increase of the decom-
position rate (Figure 3, curves c,d), whereas a remarkable
decrease was found on passing from 08C to 18 or 338C
(Figure 3, curves a c). This unexpected behavior of the
decomposition rate led us to argue that two contrasting
effects were arising by increasing the temperature (see later
for discussion).
Decomposition of S-nitrosothiols under aerobic conditions:
Kinetic experiments under aerobic conditions were conducted
at 08C in air-saturated solutions in n-pentane. Under these
conditions a fast decomposition of nitrosothiols 1a e oc-
curred; but, none of the reactions followed a simple first-order
kinetic law. The rate of decomposition was found to be higher
for the red, primary nitrosothiols 1a,b (Figure 2, curves c,e) as
compared with that for the green, tertiary nitrosothiols 1c,d
(Figure 2, curves a,b). Apparently, these results are consistent
with previous claims that the half-life time of RSNOs depends
on the presence of dioxygen[16] and the bulkiness of the alkyl
group.[18, 19] The kinetic curves (Figure 2) show that the rate of
disappearance of RSNOs increases over time. Such kinetic
behavior is characteristic of autocatalytic chain reactions, in
which a reaction product behaves as the chain carrier; the
observed increase over time of the decomposition rate must
be related to the progressive increase of the chain-reaction
rate due to the parallel increase of the chain-carrier
GC-MS and 1H NMR analyses of the reaction mixture
obtained from 1a provided evidence for the formation of
À
benzaldehyde (4) and the thiosulfonate derivative RS SO2R
(R ˆ PhCH2), in addition to dibenzyl disulfide (2a)
(Scheme 1). These products were formed in a 25:5:70 ratio
and in 80% overall yield.
Similarly, GC-MS analyses of the reaction mixtures ob-
tained from 1b,c,e showed the formation of the corresponding
À
disulfide 2b,c,e and thiosulfonate derivative, RS SO2R
382
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0382 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 2

S-Nitrosothiol Decomposition
380 387
From the above kinetic results we reasoned that dioxygen
can induce an autocatalytic chain decomposition of RSNOs
À
that largely predominates over the unimolecular S N bond
scission.
However, we obtained sound evidence that the presence of
dioxygen alone is not sufficient to induce the autocatalytic
decomposition. In fact, when a kinetic experiment was carried
out on a solution of 1a under continuous bubbling of
dioxygen, a remarkable decrease in the rate of decomposition
was observed (Figure 4, curve a). We hypothesized that in this
case the endogenous nitric oxide, produced by homolytic
À
cleavage of the S N bond, was removed from the reaction
medium by the stream of dioxygen. So, we assumed that the
autocatalytic chain decomposition is induced by the simulta-
neous presence of both dioxygen and nitric oxide.
Figure 3. Kinetic curves obtained from the decomposition under aerobic
conditions of 12 mm solutions of S-nitrosothiol 1a at 08C (curve c), 188C
(curve a), 338C (curve b), and 508C (curve d).
[R ˆ n-hexyl (7%), tert-butyl (7%), and phenyl (13%)]
(Scheme 2).
From the reaction mixture obtained from 1d we could
separate small amounts of a yellow oil (A), which rapidly
decomposed to triphenylmethanol (7) on standing in air or by
absorption on silica gel. TLC analysis of the reaction mixture
showed the presence of triphenylmethanol (7) and benzo-
phenone (6) as the main products (Scheme 2), besides trace
amounts of two unidentified compounds and minor amounts
of the yellow product A. Compounds 6 and 7, but not A, could
be separated by subsequent silica gel column chromatography
in 23 and 50% yield, respectively, together with trace amounts
of a mixture of the two unidentified products (Scheme 2). The
oily product A possibly was the nitrite 8, which could be
formed by reaction of 7 with nitric oxide under aerobic
conditions.[30]
Figure 4. Kinetic curves obtained from the decomposition of S-ni-
trosothiols 1a,c,d at 08C: a) 12 mm solution of 1a in n-octane under
bubbling of oxygen; b) 12 mm solution of 1d in n-octane added with nitric
oxide; c) 12 mm solution of 1c in n-octane added with nitric oxide;
d) 12 mm solution of 1a in n-octane added with nitric oxide.
Our assumptions on the key role of nitric oxide were proved
correct by the finding that a noticeable increase in the
decomposition rate was observed when air-saturated solutions
of 1a,c,d were added with exogenous nitric oxide. Under these
conditions a pseudo-first-order kinetic decay was followed by
1a,c (Figure 4, curve c,d). Once more, peculiar behavior was
shown by the nitrosothiol 1d: a pseudo-first-order decay was
initially observed, afterwards the decomposition rate mark-
edly slowed down (Figure 4, curve b).
.
RS
O2
chain initiation
N2O3
NO
NO
RSNO
1
N2O3
.
.
.
O
+
-
.
N O
RS N
1
RSNO
2
3
RS
2
ONO
3
N2O3
.
O
- NO
- H+
S hydrolysis
+
Ph
PhCH2SNO
Ph
H
H
4
We also found that a fast chain decomposition occurred at
room temperature when nitric dioxide was bubbled into a
carefully deaerated solution of 1a in n-hexane. Complete
disappearance of the red color was detected within 1 min.
Since nitrous anhydride (N2O3) was expected to be rapidly
formed under aerobic conditions from the reaction of nitric
oxide and dioxygen,[5, 31] and under anaerobic conditions from
nitric oxide/nitric dioxide coupling, we could infer that N2O3
was most likely to be the real chain carrier involved in the
chain-decomposition reaction of S-nitrosothiols 1a e.
We have recently reported[32] that N2O3 is capable of
oxidizing suitable organic substrates, such as styrene and
.
.
Ph3C
NO
S S
Ph3C
O2
Ph3C S
2d
S
CPh3
+
- NO2
.
.
Ph3CO
Ph3COONO
Ph3COO
5
Ph
.
Ph
Ph
.
- Ph
O Ph
O
Ph
.
6
Ph3CO
N2O3
?
Ph3COH
Ph3CONO
.
+ H
8
7
Scheme 2. Radical reactions of 1 and 2d.
Chem. Eur. J. 2002, 8, No. 2 ¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0383 $ 17.50+.50/0
383

P. C. Montevecchi and L. Grossi
FULL PAPER
p-cresol, through an ET (electron transfer) process; the
resulting radical ion pair decomposes to nitric oxide and nitro-
substituted radicals, from which reaction products 9, 10, and
11, respectively, are eventually formed (Scheme 3). An
analogous ET process between N2O3 and S-nitrosothiols 1
might lead to a radical ion pair, which could decompose to
some intermediate, possibly the radical 3, with the release of
nitric oxide (Scheme 2). The latter could be oxidized by
dioxygen to N2O3 (chain initiation). In turn, the radical
mentioned above, the disulfide 2d is in equilibrium with the
.
corresponding thiyl radical, Ph3CS , as well as perthiyl
.
.
radicals, Ph3CSS , and trityl radicals, Ph3C .[26] The latter are
known to be trapped by dioxygen to give the peroxyl radical,
.
Ph3COO , and then the hydroperoxide, Ph3COOH.[26] Most
likely, under our experimental conditions the peroxyl radical
is rapidly scavenged by NO (k ˆ 109 molÀ1 secÀ1).[34] The
resulting peroxynitrite, Ph3COONO, is expected to fragment
.
to give the trityloxyl radical, Ph3CO , which can lead to trityl
À
intermediate 3 could undergo homolytic S N bond scission
with formation of sulfanyl radicals (precursors of the disul-
fides 2) and N2O3 (chain propagation).
alcohol (7), by a hydrogen-abstraction reaction, and benzo-
phenone (6) by an initial O-neophylic rearrangement fol-
lowed by the b-elimination of a phenyl radical.[35] Subsequent
reaction of alcohol 7 with N2O3 would lead to the nitrite 8
(product A; vide supra) (Scheme 2). The peculiar kinetic
behavior exhibited by 1d (Figure 2, curve b and Figure 4,
curve b) can be explained by assuming that the reaction of
nitric oxide with dioxygen to give the chain carrier N2O3 can
Ph
NO2
9
- H+
- e
.
be partially prevented by NO/Ph3CS coupling to give back
.
N2O3
-
- NO Ph
1d; moreover, the formation of the nitrite 8 would lead to
N2O3 consumption (chain-termination reaction). So, the rate
of the formation of N2O3 (chain carrier) should progressively
decrease by progressively increasing the disulfide 2d and the
alcohol 7 concentration.
Ph
Ph
N2O3
.
.
+
NO2
+ O2
Ph
+ NO
Ph
NO2
To obtain further evidence for the crucial role of N2O3 as
the chain carrier, we performed the decomposition of 1a in
air-saturated solutions in the presence of styrene and p-cresol.
As mentioned before, these substrates react efficiently with
N2O3 through an initial ET process,[32] so they were expected
to behave as potential inhibitors of the chain decomposition
of 1a.
NO2
ONO O
O2N
O
10
OH
OH
OH
OH
NO2
H
NO2
N2O3
- NO
- e
- H+
.
.
+
.
-
N2O3
Me
Me
Me
Me
Actually, we found that the presence of equimolar amounts
of styrene strongly inhibited the autocatalytic reaction (Fig-
ure 5, curve a). The 1H NMR quantitative analysis of the
reaction mixture showed the formation of b-nitrostyrene (9)
and 1-(1-phenyl-2-nitro)ethyl nitrate (10) (Scheme 3) in a 1:1
ratio and in 44% overall yield, based on starting styrene, as
well as dibenzyl disulfide (2a) and minor amounts of
11
Scheme 3. Oxidation by N2O3 and the formation of 9, 10, and 11.
Based on this mechanism, the pseudo-first-order kinetic law
observed when the decomposition of 1a,c occurred in the
presence of an excess of exogenous nitric oxide (Figure 4,
curves c,d) can be explained by assuming that a nearly steady
concentration of N2O3 was achieved. Furthermore, the
unexpected dependence of the decomposition rate on the
temperature (Figure 3) that suggests that two contrasting
effects operate can be accounted for by assuming that both the
chain-propagation reaction and the formation of the chain
carrier, N2O3, are disfavored by increasing the temperature,
since the higher the temperature the lower the concentration
of gaseous NO, O2, and N2O3. But, in the meantime, both
reactions are favored as a consequence of the enhanced
kinetic rate constants.
As for the reaction products obtained from 1a c,e,
disulfides 2a c,e were derived by dimerization of the corre-
sponding sulfanyl radicals, whereas the source of the thiosul-
À
fonate derivatives, RS SO2R, is still unclear. Benzaldehyde
(4) might arise from the radical cation [PhCH2SNO]
.
‡
through a competitive deprotonation process followed by
the release of nitric oxide and hydrolysis of the resulting
unstable thioaldehyde (Scheme 2).[33]
Triphenylmethanol (7) and benzophenone (6), detected as
the main products from the decomposition of 1d, were
probably derived from the disulfide 2d initially formed. As
Figure 5. Kinetic curves obtained from the decomposition under aerobic
conditions of 12 mm solutions of S-nitrosothiol 1a at 08C: a) in the
presence of 12 mm styrene; b) in the presence of 20 mm p-cresol; c) in
the presence of 2 mm p-cresol; d) in the presence of 2 mm BHT; e) in the
absence of antioxidants.
384
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0384 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 2

S-Nitrosothiol Decomposition
380 387
À
benzaldehyde (4), the thiosulfonate derivative, RS SO2R
(R ˆ PhCH2), and two unidentified compounds.
The formation of the hydroxy derivative 14 can be
accounted for by an initial ET process between BHT and
.
‡
Analogously, when the decomposition of a 12 mm 1a
solution was conducted in the presence of both 20.0 and
2.0 mm p-cresol, the autocatalytic reaction rate decreased
with the increase in the p-cresol concentration (Figure 5,
curves b,c). The 1H NMR quantitative analysis of the mixture
from the former reaction showed that the p-cresol was almost
quantitatively converted (>90%) into 2-nitro-p-cresol (11),
as expected from the ET reaction between p-cresol and
N2O3 that leads to the intermediate radical cation BHT . The
latter can give the peroxyl radical 12 through deprotonation
and trapping of the resulting cyclohexadienyl radical by
dioxygen. Coupling between radical 12 and nitric oxide can
give the peroxynitrite 13, from which 14 arises by loss of nitric
dioxide and a subsequent hydrogen-abstraction reaction. The
postulated nitrite 15 (compound B) might derive from the
following nitrosation of 14.[30]
[32]
N2O3 (Scheme 3). In addition, the disulfide 2a, benzalde-
À
hyde (4), and the thiosulfonate derivative, RS SO2R (R ˆ
PhCH2), were detected in a 75:20:5 ratio.
Conclusion
We also found that 2,6-di-tert-butyl-4-methylphenol (BHT),
a well-known antioxidant,[36] is capable of inhibiting the
autocatalytic decomposition of 1a, as shown by a kinetic
experiment carried out in the presence of BHT (2.0 mm)
(Figure 5, curve d).
Under anaerobic conditions S-nitrosothiols 1a e have been
proved to undergo thermal decomposition by homolytic
À
cleavage of the S N bond that leads to nitric oxide and
sulfanyl radicals in a reversible reaction. As a consequence,
the decomposition rate is strongly decreased by the presence
of endogenous and/or exogenous nitric oxide. Rate constants,
When a solution of 1a (12 mm) in n-pentane decomposed in
1
the presence of equimolar amounts of BHT, the H NMR
À
analysis of the reaction mixture showed the formation of 2,6-
di-tert-butyl-4-hydroxy-4-methyl-2,5-cyclohexadienone
(14)[37] and an unidentified compound (B) in 45 and 18%
yield, respectively, as well as the disulfide 2a, unreacted BHT
(30%), and minor amounts of the aldehyde 4 and the
k1, of the unimolecular S N bond scission have been
determined at different temperatures from kinetic measure-
ments performed in refluxing solvents. In sharp contrast to
previous reports, tertiary nitrosothiols 1c (k1(698C) ˆ 13 Â
10À3 minÀ1) and 1d (k1(698C) ˆ 91  10À3 minÀ1) decompose
faster than primary nitrosothiols 1a (k1(698C) ˆ 3.0 Â
10À3 minÀ1) and 1b (k1(698C) ˆ 6.5  10À3 minÀ1).
À
thiosulfonate derivative, RS SO2R. The subsequent silica
gel column chromatography separated, in addition to 2a and
À
the thiosulfonate, RS SO2R, the hydroxy derivative 14 in
Activation energies (E# ˆ 20.5 22.8 KcalmolÀ1), calculated
through the Arrhenius equation, are significantly lower than
52% yield. No traces of the product B were found. In order to
obtain evidence about its identity, a solution of BHT in n-
octane was reacted for three hours with nitric oxide at 08C in
the presence of air. 1H NMR analysis of the reaction mixture
showed complete disappearance of BHT and the presence of
the compound B as the exclusive reaction product. The latter
was found to give the product 14 on standing in air or by
absorption on silica gel. We suggest that the product B might
be the nitrite 15, from which the alcohol 14 can form by
subsequent hydrolysis (Scheme 4).
À
the S N dissociation bond energies (DBEs) reported in the
literature (ca. 25 KcalmolÀ1).
Conversely, under aerobic conditions, the decay of S-
nitrosothiols 1a e takes place by an autocatalytic chain-
decomposition process catalyzed by N2O3. The latter is
formed by reaction of dioxygen with the endogenous nitric
oxide. The rate of the chain reaction is independent of the
RSNO concentration and decreases with increasing bulkiness
of the alkyl group; this shows that steric effects are crucial in
the propagation step.
The autocatalytic decomposition is strongly inhibited by
removing the endogenous nitric oxide or by the presence of
antioxidants, such as p-cresol, b-styrene, and BHT, which are
preferentially oxidized by N2O3. On the other hand, the
addition of exogenous nitric oxide causes a noticeable
increase in the chain-decomposition rate.
.
Me
Me
Me OO
.
O2
tBu
N2O3
- e-, - H+
tBu
tBu
tBu
tBu
tBu
O
12
OH
BHT
O
NO
Me OONO
.
Me OH
Me
O
- NO2
.
+ H
Experimental Section
tBu
tBu
tBu
tBu
tBu
tBu
O
14
O
O
General: NMR spectra were recorded with a Varian Gemini200 instrument
using Me4Si as an internal standard. GC-MS analyses were performed with
a Carlo Erba QMD1000 instrument. Mass spectra were recorded with a
VG7070E instrument using electron impact ionization. Kinetic measure-
ments were performed with Pharmacia Biotech Ultrospec1000 UV/Vis
spectrophotometer.
13
N2O3
Me ONO
hydrolysis
Materials: S-Nitrosothiols 1a e were synthesized from the corresponding
thiols, commercially available, and peroxynitrite.[21] An aqueous solution of
peroxynitrite (0.4m, pH ˆ 13.5) (2.5 mL, 1 mmol) was added at 5 108C
under stirring to a solution (10 mL) in acetonitrile of the appropriate thiol
(1 mmol) and HCl (12m, 0.2 mL); the resulting mixture was stirred for a
further 10 min. In the case of S-nitrosothiols 1a c,e the reaction mixture
tBu
tBu
O
15
Scheme 4. Reaction of 2,6-di-tert-butyl-4-methylphenol (BHT) with N2O3.
Chem. Eur. J. 2002, 8, No. 2 ¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0385 $ 17.50+.50/0
385

P. C. Montevecchi and L. Grossi
FULL PAPER
was extracted with the appropriate organic solvent (30 mL), and the
organic layer was washed twice with water (20 mL), twice with sodium
carbonate (10%, 20 mL) to eliminate the unreacted thiol, and then dried
with Na2SO4. Yields of 1a c,e were determined spectrophotometrically
(1a: lmax (e) ˆ 550 nm (26 molÀ1 dm3 cmÀ1);[22, 23] 1b: lmax (e) ˆ 550 nm
(21 molÀ1 dm3 cmÀ1);[22] 1c: lmax (e) ˆ 603 nm (16 molÀ1 dm3 cmÀ1);[23] 1e:
b-nitrostyrene (9) (22%, based on starting styrene), 2-nitro-1-phenylethyl
nitrate (10)[32] (22%, based on starting styrene), and an unknown product
(7%) [1H NMR (200 MHz, CDCl3, TMS): d ˆ 4.45 (dd, JAB ˆ 12.7 Hz,
JAX ˆ 3.5 Hz, 1H; A part of an ABX system), 4.55 (dd, JAB ˆ 12.7 Hz, JBX
ˆ
9.0 Hz, 1H; B part of an ABX system), 5.40 (dd, J1 ˆ 9.0 Hz, J2 ˆ 3.5 Hz,
1H)].
lmax (e) ˆ 570 nm (42 molÀ1 dm3 cmÀ1
)
[24]), and then the solutions were
Reaction products from the decomposition of 1a under aerobic conditions
1
diluted to reach the desired concentration. In the case of 1d, the green solid
separated from the reaction mixture was filtered, washed with water, and
dried under vacuum [185 mg, 60%; lmax (e) ˆ 603 nm (17 molÀ1 dm3 cmÀ1)].
Peroxynitrite was synthesized following the previously reported
procedure.[21] The concentration (usually in the range 0.45 0.50m)
in the presence of p-cresol: H NMR quantitative analysis (acetophenone
as the internal standard) of the reaction mixture obtained from the
decomposition of 1a (40 mL of a 12 mm solution in n-pentane, 0.48 mmol)
carried out in the presence of air and p-cresol (85 mg, 0.80 mmol) showed
the formation of 4-methyl-2-nitrophenol (11) (100% yield, 90% conver-
À
sion) together with products 2a, 4, and RS SO2R (R ˆ PhCH2) in a 75:20:5
was
determined
spectrophotometrically
[lmax
(e) ˆ 302 nm
(1670 molÀ1 dm3 cmÀ1)].[38] Solutions kept at À188C showed little decom-
ratio (overall yield not determined), and an unknown product showing a
singlet at d ˆ 4.0.
position over several weeks.
Reaction products from the decomposition of 1a under aerobic conditions
in the presence of 2,6-di-tert-butyl-4-methylphenol (BHT): A solution of
1a in n-octane (12 mm, 40 mL) containing BHT (0.48 mmol, 105 mg) was
allowed to react at 08C in a open tube until the disappearance of the red
color (ca. two days), and then the solvent was removed. 1H NMR analysis
of the residue, by using acetophenone as the internal standard, showed the
presence of unreacted BHT (30%), the disulfide 2a (60%), benzaldehyde
Kinetic measurements under anaerobic conditions: A solution of 1a
d
(12 mm) in the appropriate solvent (cyclopentane: b.p. ˆ 508C; hexane:
b.p. ˆ 698C; cyclohexane: b.p. ˆ 80.58C; and methylcyclohexane: b.p. ˆ
1018C) was deaerated by bubbling with argon at room temperature for
15 min, and then introduced into a thermostatic bath kept at a temperature
58C higher than the solvent boiling point in order to maintain a moderate
reflux. The apparatus was kept in the dark under an argon atmosphere.
Samples were taken at fixed periods of time up to 200 min. In all cases a
first-order kinetic decay was followed. The kinetic constant values are
reported in Table 1.
À
(4) (yield not determined), RS SO2R (R ˆ PhCH2) (4%), an unknown
product showing a singlet at d ˆ 4.0, the 4-hydroxycyclohexanedione 14[37]
(45%, based on reacted BHT), and the product B (see below) (18%, based
on reacted BHT). Subsequent silica gel column chromatography gave, by
gradual elution with petroleum ether (b.p. 40 708C)/diethyl ether: i) the
Kinetic measurements under aerobic conditions: Kinetic measurements
were generally performed in an open flask, without stirring, with a solution
of 1a e in n-pentane (12 mm) for experiments carried out at 08C, or
solutions of 1a in n-octane (12 mm) for experiments carried out at 18, 33,
and 508C. Also, kinetic measurements at 08C were performed with
solutions of 1a (6 mm and 16 mm) in n-pentane. Experiments performed in
the presence of added nitric oxide were carried out by bubbling for 10 s
nitric oxide into the appropriate solution of S-nitrosothiol 1a,c,d in n-
octane. Kinetic curves are reported in Figures 2 4.
À
disulfide 2a (65 mg, 55%); ii) RS SO2R (R ˆ PhCH2) (5 mg, 4%); iii) the
cyclohexanedione 14 (40 mg, 0.17 mmol, 52%).
Reaction of BHT with nitric oxide: A solution of BHT in n-octane (40 mL,
90 mg, 0.4 mol) was saturated with gaseous nitric oxide and allowed to react
at 08C in a sealed tube. After 2 h, TLC analysis showed the complete
disappearance of the starting BHT. The residue obtained after removal of
the solvent was composed of the compound B, as evidenced by TLC and
1H NMR analysis [1H NMR (200 MHz, CDCl3, TMS): d ˆ 1.21 (s, 18H),
1.82 (s, 3H), 6.73 (s, 2H)]. Compound B, which tentatively was assigned the
structure of nitrite 15, underwent quantitative decomposition to 14 upon
standing in air or by absorption on silica gel.
Reaction products from decomposition of 1a c,e under anaerobic
conditions: Solutions of 1a c,e were left to react in refluxing n-hexane
until disappearance of the color was detected. GC-MS analysis of the
reaction mixtures detected the exclusive presence of the corresponding
disulfide 2a c,e (yield not determined).
Reaction products from decomposition of 1a e under aerobic conditions:
GC-MS analysis of the reaction mixtures obtained from the decomposition
of solutions of 1b,c,e (12 mm) showed the exclusive presence of the
Acknowledgements
À
À
disulfide 2b,c,e and the thiosulfonate derivative, RS SO2R (2/RS SO2R
relative ratio: 93:7 (R ˆ n-hexyl), 93:7 (R ˆ tert-butyl), and 87:13 (R ˆ
phenyl), overall yield not determined).
This workwas financially supported by the Ministry of the University and
Scientific and Technological Research (MURST), Rome (Funds 60% and
40%), and by the University of Bologna (Funds for selected research topics
A.A. 1999 2001).
GC-MS analysis of the reaction mixture obtained from 1a (40 mL of a
12 mm solution in n-pentane, 0.48 mmol) showed the formation of
À
compounds 2a and 4, and the thiosulfonate RS SO2R (R ˆ PhCH2). The
n-pentane solvent was carefully removed under vacuum. 1H NMR
quantitative analysis of the resulting residue, performed by using aceto-
phenone as the internal standard, showed the above products to be present
in a 70:25:5 ratio and in 80% overall yield together with small amounts of
an unidentified product showing a singlet at d ˆ 4.0.
From the reaction mixture obtained from 1d (40 mL of a 12 mm solution in
n-pentane, 0.48 mmol) small amounts of a yellow oil separated from the
solution; the n-pentane solvent was removed, and the residue was purified
by chromatography on a silica gel column; gradual elution with petroleum
ether (b.p. 40 708C)/diethyl ether gave: i) a mixture of two unidentified
products (6 mg); ii) benzophenone (6) (20 mg, 23%); and iii) triphenyl-
methanol (7) (62 mg, 50%). In a repeated experiment we separated the
above yellow oil [MS (70 eV): m/z (%): 244 (55), 167 (70), 165 (100)] which
rapidly decomposed on standing in air to give 7.
[1] S. Oae, K. Shinhama, Org. Prep. Proced. Int. 1983, 15, 165.
[2] D. H. L. Williams, Chem. Soc. Rev. 1983, 14, 171.
[3] J. S. Stamler, Curr. Top. Microbiol. Immunol. 1995, 196, 19.
[4] D. H. L. Williams, Acc. Chem. Res. 1999, 32, 869.
[5] L. J. Ignarro in Nitric Oxide: Biology and Pathobiology (Ed.: L. J.
Ignarro), Academic Press, San Diego, 2000, Chapter 1.
[6] R. J. Singh, N. Hogg, J. Joseph, B. Kalyanaraman, J. Biol. Chem. 1996,
271, 18596.
[7] W. R. Mathews, S. W. Kerr, J. Pharmacol. Exp. Ther. 1993, 267, 1529.
[8] B. Gaston, J. Reilly, J. M. Drazen, J. Fackler, P. Ramdev, D. Arnelle,
M. E. Mullins, D. J. Sugarbaker, C. Chee, D. J. Singel, D. Loscalzo, J. S.
Stamler, Proc. Natl. Acad. Sci. USA 1993, 90, 10957.
[9] S. Moncada, R. M. Palmer, E. A. Higgs, Pharmacol. Rev. 1991, 43, 109.
[10] P. M. Rao, J. A. Copeck, A. R. Knight, Can. J. Chem. 1967, 45, 1369.
[11] P. D. Josephy, D. Rehorek, E. G. Janzen, Tetrahedron Lett. 1984, 25,
1685.
[12] L. Field, R. V. Dilts, R. Ravichandran, P. G. Lenhert, G. E. Carnahan,
J. Chem. Soc. Chem. Commun. 1978, 249.
[13] S. A. Lipton, Y. B. Choi, Z. H. Pan, S. Z. Lei, H. S. Vincent Chen, N. J.
Sucher, J. Loscalzo, D. J. Singel, J. S. Stamler, Nature 1993, 364, 626.
[14] J. McAninly, D. H. L. Williams, S. C. Askew, A. R. Butler, C. Russell,
J. Chem. Soc. Chem. Commun. 1993, 1758.
Reaction products from the decomposition of 1a under aerobic conditions
in the presence of styrene: The decomposition of 1a (40 mL of a 12 mm
solution in n-octane, 0.48 mmol) carried out in the presence of air and
styrene (50 mg, 0.48 mmol) was complete after approximately two days.
The solvent was evaporated. 1H NMR analysis of the residue, by using
acetophenone as the internal standard, showed the formation of the
À
disulfide 2a (60%), benzaldehyde 4 (yield not determined), RS SO2R
(R ˆ PhCH2) (4%), an unknown product showing a singlet at d ˆ 4.0,
386
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0386 $ 17.50+.50/0
Chem. Eur. J. 2002, 8, No. 2

S-Nitrosothiol Decomposition
380 387
[15] S. Oae, D. Fukushima, Y. H. Kim, J. Chem. Soc. Chem. Commun. 1977,
407.
value was overestimated since experiments were conducted in day-
light.
[16] L. J. Ignarro, H. Lippton, J. C. Edwards, W. H. Baricos, A. L. Hymann,
P. J. Kadowitz, C. A. Gruetter, J. Pharmacol. Exp. Ther. 1981, 218, 739.
[17] a) D. L. H. Williams, Chem. Commun. 1996, 1085; b) S. C. Askew,
D. J. Barnett, J. McAninly, D. L. H. Williams, J. Chem. Soc. Perkin
Trans. 2 1995, 741; c) A. P. Dicks, H. R. Swift, D. L. H. Williams, A. R.
Butler, H. H. Al-Sa×doni, B. G. Cox, J. Chem. Soc. Perkin Trans. 2
1996, 481; d) A. P. Dicks, P. H. Beloso, D. L. H. Williams, J. Chem. Soc.
Perkin Trans. 2 1997, 1429; e) A. J. Holmes, D. L. H. Williams, Chem.
Commun. 1998, 1711; f) D. R. Nobble, H. R. Swift, D. L. H. Williams,
Chem. Commun. 1999, 2317.
[26] F. F. Blicke, J. Am. Chem. Soc. 1923, 45, 1965.
[27] Formation of 1d by direct reaction of nitric oxide with the disulfide 2d
was ruled out since we found that nitric oxide is generally incapable of
reaction with disulfides.
[28] J.-M. L¸, J. M. Wittbrodt, K. Wang, Z. Wen, H. B. Schlegel, P. G.
Wang, J.-P. Cheng, J. Am. Chem. Soc. 2001, 123, 2903.
[29] S. Benson, Chem. Rev. 1978, 78, 23.
[30] Aliphatic alcohols have been reported to react with nitric oxide under
aerobic conditions to give the corresponding nitrites (L. Grossi, S.
Strazzari, J. Org. Chem. 1999, 64, 8076).
[18] M. D. Bartberger, K. N. Houk, S. C. Powell, J. D. Mannion, K. Y. Lo,
J. S. Stamler, E. J. Toone, J. Am. Chem. Soc. 2000, 122, 5889.
[19] S. Oae, K. Shinhama, K. Fujimori, Y. H. Kim, Bull. Chem. Soc. Jpn.
1980, 53, 775.
[20] L. Grossi, P. C. Montevecchi, S. Strazzari, J. Am. Chem. Soc. 2001, 123,
4853.
[21] L. Grossi, P. C. Montevecchi, S. Strazzari, Eur. J. Org. Chem. 2001, 131.
[22] J. Barret, L. J. Fitzgibbons, J. Glauser, R. H. Still, P. N. W. Young,
Nature 1966, 211, 848.
[23] S. Oae, Y. H. Kim, D. Fukushima, K. Shinhama, J. Chem. Soc. Perkin
Trans. 1 1978, 913.
[31] J. S. Stamler, D. I. Simon, J. A. Osborne, M. E. Mullins, D. Jaraki, T.
Michel, D. J. Singel, J. Loscalzo, Proc. Natl. Acad. Sci. USA 1992, 89,
444.
[32] L. Grossi, P. C. Montevecchi, S. Strazzari, Eur. J. Org. Chem. 2001, 741.
À
[33] The formation of thiosulfonate derivatives, RS SO2R, and benzalde-
hyde (4) from 1a c has already been reported (see ref. [19]).
[34] S. Padmaja, R. E. Huie, Biochem. Biophys. Res. Commun. 1993, 195,
539.
[35] W. H. Urry, M. S. Karasch, J. Am. Chem. Soc. 1944, 66, 1438.
[36] E. G. Janzen, A. L. Wilcox, V. Manoharan, J. Org. Chem. 1993, 58,
3597.
[24] A. Cornelis, N. Depaye, A. Gerstmans, P. Laszlo, Tetrahedron Lett.
1983, 30, 3103.
[37] B. M. Benjamin, V. F. Raaen, E. W. Hagaman, L. L. Brown, J. Org.
Chem. 1978, 43, 2986.
[38] M. N. Hughes, H. G. Nicklin, J. Chem. Soc. (A) 1968, 450.
[25] In a previous paper (see ref. [19]) we reported k ˆ 2.8  10À4 minÀ1 for
the thermal decomposition of 1a c at 08C. It is most likely that this
Received: May 7, 2001 [F3239]
Chem. Eur. J. 2002, 8, No. 2
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
0947-6539/02/0802-0387 $ 17.50+.50/0
387