Pathways of Nitrosobenzene Reduction by Thiols in Alcoholic Media

Article abstract of DOI:10.1021/jo981889t

The biologically important reaction of nitrosobenzenes with thiols has been investigated in 2-propanol solution at room temperature, experimental conditions which allow for the detection and characterization of key intermediates. Final stable products of such complex reactions include azoxybenzenes and anilines, formed in relative proportions and at a rate which depend on the reagents initial molar ratio. A detailed description of the reaction of 4-chloronitrosobenzene with benzenethiol in 2-propanol was achieved by means of ¹H NMR in situ analysis. The reaction is initiated by rapid and quantitative coupling of the two reagents into a covalent adduct, an N-ydroxysulfenamide (N(OH)S), which decays to N-(4-chlorophenyl)hydroxylamine. This second intermediate is then converted via competing paths to 4,4′-dichloroazoxybenzene and to N-(4-chlorophenyl)benzenesulfenamide (4-ClC₆H₄NHSPh), which in turn decays to 4-chloroaniline. Interestingly, sulfinamides (ArNHS(O)R), major products of the reaction in aqueous media, do not form in 2-propanol.

Full text of DOI:10.1021/jo981889t

3
422
J . Org. Chem. 1999, 64, 3422-3428
P a th w a ys of Nitr osoben zen e Red u ction by Th iols in Alcoh olic
Med ia
Stefano Montanari, Cristina Paradisi,* and Gianfranco Scorrano
Centro Studio Meccanismi di Reazioni Organiche del CNR, Dipartimento di Chimica Organica,
Via Marzolo 1, 35131 Padova, Italy
Received September 16, 1998
The biologically important reaction of nitrosobenzenes with thiols has been investigated in
2
-propanol solution at room temperature, experimental conditions which allow for the detection
and characterization of key intermediates. Final stable products of such complex reactions include
azoxybenzenes and anilines, formed in relative proportions and at a rate which depend on the
reagents initial molar ratio. A detailed description of the reaction of 4-chloronitrosobenzene with
1
benzenethiol in 2-propanol was achieved by means of H NMR in situ analysis. The reaction is
initiated by rapid and quantitative coupling of the two reagents into a covalent adduct, an
N-hydroxysulfenamide (N(OH)S), which decays to N-(4-chlorophenyl)hydroxylamine. This second
intermediate is then converted via competing paths to 4,4′-dichloroazoxybenzene and to N-(4-
chlorophenyl)benzenesulfenamide (4-ClC H NHSPh), which in turn decays to 4-chloroaniline.
6 4
Interestingly, sulfinamides (ArNHS(O)R), major products of the reaction in aqueous media, do not
form in 2-propanol.
In tr od u ction
Nitrosoarenes are key intermediates in oxidation/
reactivity is involved in “damaging” processes leading to
methemoglobinemia, carcinogenesis, or mutagenesis.
Research in this field has focused mostly on the reactions
of N-oxygenated arylamines in red cells, particularly on
the reaction of nitrosobenzenes with glutathione. The
results of mechanistic investigations, which employed
reduction pathways of aza-substitued aromatic com-
pounds. In alcoholic solutions, nitrosoarenes form as
5
1
2
intermediates in the alkoxide - or thiolate -induced re-
duction of nitroarenes to the corresponding azoxyarenes
and anilines. In the course of our work with S-anions2
our interest was captured by the reactions of nitrosoben-
zenes with neutral thiols. While toxicologists have long
been interested in such processes for their important role
in biological systems, relatively little information is found
in the organic chemistry literature about these reac-
tions.3 Nitrosobenzenes, which form in biological sys-
tems either by metabolic N-oxidation of arylamines
8
-12
mostly reduced glutathione
but also 2,3-dihydroxy
propanethiol13 as model sulfhydryl reagents (RSH), in-
dicated that in buffered aqueous media the reaction of
nitrosobenzenes (ArNO) leads to the corresponding N-
arylhydroxylamines (ArNHOH), N-arylsulfinamides
(ArNHS(O)R), N-arylsulfenamides (ArNHSR), and anilines
(
ArNH ) in amounts which depend on the reactants
2
,4
structure and concentration and on the reaction condi-
tions. Notably, sulfenamides, ArNHSR, were not ob-
served in some of these studies because of fast hydrolysis
(
nitrosoarenes and N-hydroxyarylamines, which are in
metabolic equilibrium, are often indicated as N-oxygen-
1
3
to ArNH
2
.
A common feature of the different mecha-
5
a
ated arylamines ) or by reduction of aromatic nitro
compounds, react readily with SH groups in proteins in
vivo.6 Indeed biomonitoring of hemoglobin-bound ni-
trosoarenes (binding to cystein residue) can be used to
control exposure to and toxication of potentially hazard-
nistic proposals emerged from these studies is that the
first product of the reaction between a nitrosobenzene
,5a
6
and a thiol is a covalent adduct, the stability of which
was found to increase with decreasing solvent polarity
and temperature.13 The structure of a semimercaptal, the
N-arylhydroxysulfenamide ArN(OH)SR, was proposed for
this species, which was never isolated and was examined
7
ous arylamines and nitroarenes in persons at risk. This
*
Corresponding author. e-mail: paradisi@mail.chor.unipd.it;
phone: ++39 0498275661; fax: ++39 0498275239.
1) Arca, V.; Paradisi, C.; Scorrano, G. J . Org. Chem. 1990, 55, 3617
and references therein.
2) (a) Montanari, S.; Paradisi, C.; Scorrano, G. J . Org. Chem. 1993,
1
3
in solution by means of UV, FAB-MS, and C NMR
(
13,11
analyses.
Increasing pH was found to increase the
1
0
rate of formation of the adduct, suggesting involvement
of RS formed from RSH in a predissociation equilibrium
(
-
5
1
8, 5628. (b) Montanari, S.; Paradisi, C.; Scorrano, G. J . Org. Chem.
991, 56, 4274. (c) Dario, M. T.; Montanari, S.; Paradisi, C.; Scorrano,
10,4
step.
It was proposed that this adduct decays along
G. Tetrahedron Lett. 1994, 35, 301.
3) Eyer, P.; Gallemann, D. Reactions of Nitrosoarenes with SH
three competing routes: (1) rearrangement to the corre-
sponding sulfinamide (ArNHS(O)R), (2) reaction with
(
Groups. In The Chemistry of Funzional Groups: The Chemistry of
Amino, Nitroso, Nitro and Related Groups. Suppl. F2; Patai, S.,
Rappoport, Z., Eds.; J ohn Wiley & Sons: New York, 1996; pp 999-
1
039.
(8) Kazanis, S.; McClelland, R. A. J . Am. Chem. Soc. 1992, 114, 3052.
(9) D o¨ lle, B.; T o¨ pner, W.; Neumann, H.-G. Xenobiotica 1980, 10, 527.
(10) Eyer, P. Chem.-Biol. Interact. 1979, 24, 227.
(11) Saito, K.; Kato, R. Biochem. Biophys. Res. Commun. 1984, 124,
1.
(4) Zuman, P.; Shah, B. Chem. Rev. 1994, 94, 1621.
5) (a) Eyer, P. Xenobiotica 1988, 18, 1327. (b) Gallemann, D.; Eyer,
(
P. Biol. Chem. Hoppe-Seyler 1993, 374, 51.
6) Eyer, P. In Biological Oxidation of Nitrogen in Organic Molecules;
(
Gorrod, J . W., Damani, L. A., Eds.; Verlag Chemie E. Horwood:
(12) Mulder, G. J .; Unruh, L. E.; Evans, F. E.; Ketterer, B.; Kadluba,
F. F. Chem.-Biol. Interact. 1982, 39, 111.
(13) Klehr, H.; Eyer, P.; Sch a¨ fer, W. Biol. Chem. Hoppe-Seyler 1985,
366, 755.
Chichester, England, 1985; pp 386-389.
(7) (a) Neumann, H.-G. Arch. Toxicol. 1984, 56, 1. (b) Albrecht, W.;
Neumann, H.-G. Arch. Toxicol. 1985, 57, 1.
1
0.1021/jo981889t CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/24/1999

Nitrosobenzene Reduction by Thiols in Alcoholic Media
J . Org. Chem., Vol. 64, No. 10, 1999 3423
thiol to give ArNHOHsthe ratio of N-arylhydroxylamine
cially advantageous, relative to both aqueous and apolar
(benzene)16 media, since clean solutions containing only
the initial intermediate were readily obtained, thus
simplifying the study of its reactivity.
to sulfinamide was found to increase with increasing
6
glutathione initial concentration, and (3) reaction with
thiol to give the sulfenamide ArNHSR. As remarked
4
recently, little is known about the mechanisms involved.
The ArN(OH)SR f ArNHS(O)R “rearrangement” (route
Resu lts
4
1
) requires, according to one proposal, one molecule of
The 2-propanol reaction of nitrosobenzenes with ben-
zenethiol at room temperature gives as final stable
products, besides diphenyl disulfide, the corresponding
aniline and azoxybenzene derivatives, in relative propor-
tions and at a rate which depend on the reagents’ initial
molar ratio. An example of a typical product distribution
is given in eq 4 (isolated yields are reported). As the
concentration of thiol was reduced, the yield of the azoxy
product increased while that of 4-chloroaniline decreased.
The long reaction time indicated in eq 4 is due to slow
reactions of various intermediate species, the reactivity
of 4-chloronitrosobenzene toward benzenethiol being in
fact remarkably high (vide infra).
ionized thiol and occurs either in a single- or in a two-
step sequence. An alternative proposal involves rate-
4
limiting heterolytic N-O cleavage, either uncatalyzed or
+
catalyzed by acids, to a nitrenium ion intermediate ArN -
8
SR, followed by reaction with water. The mechanism by
which sulfenamides are formed has also remained con-
troversial.9
,6
Evidence for the presence of radicals in these systems
has been collected by ESR analysis.1
4-17
In aqueous
media nitrosobenzene reacts with thiols to give the
phenylhydronitroxyl radical, PhN(O•)H, according to eq
1
4,17
1
.
This species was observed in buffered (pH 7.4)
aqueous equimolar solutions of nitrosobenzene and glu-
tathione. No ESR signal was, however, detected in the
presence of larger concentrations of glutathione.14 It was
indeed shown that glutathione rapidly reduces the phen-
ylhydronitroxyl radical to PhNHOH, eq 2.15 Unstable
thiyl radicals (RS•) were also detected in these systems
by spin-trapping techniques.1
5,17
The phenylhydronitroxyl radical was also observed in
a 2-propanol solution of nitrosobenzene and thiophenol.
In benzene solution, on the other hand, the thionitroxyl
1
6
Similar results were also obtained for the reaction of
nitrosobenzene, a somewhat less reactive substrate. It
was reported earlier that reactivity increases with in-
radical (PhN(O•)SPh) was observed, instead of the PhN-
creasing electronegativity of para and meta substitu-
(
O•)H radical, its formation being attributed to the
ents.1 Sulfenamides ArNHSPh (Ar ) Ph, 4-ClC
8,4
16
6 4
H )
sequence of eq 3.
were detected by GC-MS and HPLC analysis. Both
techniques, however, proved unsuitable for monitoring
the reaction course because of the complexity of the
mixtures and of the lability of some of the species
involved. Thus, both N-aryl-hydroxylamines and N-aryl-
sulfinamides decomposed during GC analysis, and HPLC
conditions of adequate resolution and reproducibility
could not be found in the present study. We, therefore,
turned to NMR in situ analysis. Since 4-chloronitrosoben-
Different conclusions were reached concerning the
involvement of such radicals in the processes described
above. Mason and Eyer noted that “clearly there are also
many other reactions of nitrosobenzene with thiols by non
free-radical processes to yield semimercaptals and other
metabolites”.15 In contrast, only radical mechanisms
1
zene (1) has a very characteristic H NMR spectrum
(AA′XX′) and chemical modifications of its nitroso func-
(involving mostly the thionitroxyl radical) were proposed
tion produce significant changes in the signal due to the
ring protons, this compound served as an ideal model to
monitor reactions with both aliphatic thiols and ben-
zenethiol (to avoid interference by the protons of the PhS
group, perdeuterated benzenethiol was prepared and
used in these experiments). Our research focused on the
reaction with benzenethiol for its convenient rate and for
the possibility to access most potential intermediates as
stable compounds to be used as references for the
attribution of NMR signals. The following compounds,
available commercially or through reported synthetic
procedures, constituted our reference basis set: 4-chlo-
roaniline (2), 4,4′-dichloroazoxybenzene (3), N-(4-chlo-
rophenyl)hydroxylamine (4), N-(4-chlorophenyl)benzene-
sulfenamide (5), N-(4-chlorophenyl)benzenesulfinamide
to account for all the different products formed in the
benzene reaction of benzenethiol with nitrosobenzenes.16
We report in this paper the results of a mechanistic
investigation of the reaction of nitrosobenzenes with
1
thiols in 2-propanol, based mainly on in situ H NMR
analysis. The advantages of this analytical approach with
respect to those previously employed (HPLC, UV) are
obvious if one considers that some of the crucial reaction
intermediates are not stable (vide infra) under the
conditions used for chromatographic analysis (HPLC,
4
GC). The choice of 2-propanol as solvent proved espe-
(
14) Takahashi, N.; Fischer, V.; Schreiber, J .; Mason, R. P. Free Rad.
Res. Commun. 1988, 4, 351.
15) Maples, K. R.; Eyer, P.; Mason, R. P. Mol. Pharmacol. 1990,
7, 311.
16) Alberti, A.; Benaglia, M.; Carloni, P.; Greci, L.; Stipa, P.; Rizzoli,
C.; Sgarabotto, P. Gazz. Chim. Ital. 1995, 125, 555.
17) Hiramoto, K.; Ojima, N.; Kikugawa, K. Free Rad. Res. 1997,
7, 409.
(
3
(
6), N-(4-chlorophenyl)-N-(phenylthio)benzenesulfena-
(
(
(18) Diepold, C.; Eyer, P.; Kampffmeyer, H.; Reinhardt, K. Adv. Exp.
Med. Biol. 1982, 136B, 1173.
2

3
424 J . Org. Chem., Vol. 64, No. 10, 1999
Montanari et al.
Ta ble 1. 1H NMR (400 MHz) Da ta for 4-ClC6H4NXY
Com p ou n d s in 2-P r op a n ol-d 8
1
F igu r e 1. Partial H NMR (400 MHz) spectra in oxygen-free
-PrOH-d at room temperature of (a) 1 (7 mM); (b) solution
a) analyzed 2 min after addition of 1 equiv of benzenethiol-d
20 µL of a 0.39 M 2-PrOH-d solution); (c) solution of 1 (7
2
8
(
(
5
8
mM) analyzed 2 min after addition of 0.5 equiv of benzenethiol-
d .
5
a
Center of gravity of the two “doublets” of the AA′XX′ system,
b
relative to TMS. Separation between adjacent peaks of the two
“doublets” of the AA′XX′ system.
mide (7), 4-chloronitrobenzene (8), and 4,4′-dichloro-
azobenzene (9). Relevant NMR spectral data obtained in
2
8
-propanol-d are reported in Table 1. It should be noted
that for some of these compounds the appearance of the
spectrum is highly dependent on the solvent. In the case
of 5, for example, the slightly broad “singlet” observed
in 2-propanol-d
AA′XX′ signal in CDCl
8
(7.04 ppm) resolves into a characteristic
(two “doublets” centered at 6.97
3
and 7.17 ppm). It was therefore essential to compare
spectra recorded in the same solvent. Chemical shifts and
coupling constants values reported in Table 1, as well as
in the text, are approximations derived directly from the
spectra as explained in the table footnotes.
All evidence collected in this study indicates that the
reaction between a nitrosobenzene and benzenethiol is
best described as a two-phase process: (1) formation of
a first intermediate (fast) and (2) decay of this common
intermediate (slow) through a complex sequence of reac-
tions to give eventually anilines and azoxybenzenes as
major stable products. For convenience the two phases
are discussed separately.
F or m a tion of th e In ter m ed ia te. Figure 1 shows
some representative H NMR results relative to the
formation of the intermediate. The spectrum of 4-chlo-
ronitrosobenzene (1), shown in trace a changes, after
1
1
F igu r e 2. Partial H NMR (400 MHz) spectra in oxygen-free
-PrOH-d at room temperature of (a) 1 (7 mM) and of the
2
8
same solution recorded, after addition of 5 equiv of 2-pro-
panethiol, at the indicated times: (b) 50 s; (c) 5 min; (d) 10
min; (e) 25 min.
5
addition of 1 equiv of PhSH-d , into b in which the
original signal has been replaced by a new one (AA′XX′,
two “doublets” at 7.08 and 7.23 ppm; J ) 8.9 Hz). In an
analogous experiment in which only 0.5 equiv of PhSH-
that of i, the product of the benzenethiol reaction.
However, these aliphatic thiols react more slowly than
benzenethiol, the complete conversion of 1 in the presence
of excess thiol requiring ca. 15 min with MeSH and ca.
30 min with 2-PrSH (Figure 2). With t-BuSH the signals
of 1 remained unchanged for 24 h. A very low reactivity
was reported for this thiol also in aqueous media. The
reaction with aliphatic thiols was not further investi-
gated.
5
d was used, trace c was obtained, comprising the signals
of 1 and of the product in approximately a 1 to 1 ratio.
These and other similar experiments show that, upon
addition of benzenethiol, the nitroso compound is rapidly
converted into a new species, named i, the stoichiometry
requiring 1 equiv of each reagent. A few experiments
were also carried out with aliphatic thiols (MeSH,
6
2
-PrSH, and t-BuSH). In each of the reactions with MeSH
Concerning the identity of the intermediate i formed
in the benzenethiol reaction, it appears that the NMR
signal due to this species does not match any of those of
and 2-PrSH, one initial species formed that was charac-
terized by H NMR signal patterns closely resembling
1

Nitrosobenzene Reduction by Thiols in Alcoholic Media
J . Org. Chem., Vol. 64, No. 10, 1999 3425
1
F igu r e 3. Partial H NMR (400 MHz) spectra recorded at the
indicated times for the room-temperature reaction of 1 (7 mM)
1
F igu r e 4. Partial H NMR (400 MHz) spectra recorded at the
with 5 equiv of benzenethiol-d
5 8
in oxygen-free 2-PrOH-d .
indicated times for the room-temperature reaction of 1 (7 mM)
with 1 equiv of benzenethiol-d in oxygen-free 2-PrOH-d .
5 8
Trace d was recorded at 200 MHz.
Trace d was recorded at 200 MHz.
reference compounds 2-9 reported in Table 1. Both the
stoichiometry of the reaction and the NMR chemical
shifts are consistent with the structure of the covalent
adduct N-hydroxysulfenamide.13,11 This conclusion is also
supported by the results of a mass spectrometric inves-
tigation,19 briefly discussed in the next section.
Attempts to isolate i were unsuccessful. It was hoped
that fast removal of the solvent under reduced pressure
should yield either 1 or i depending on whether adduct
6
formation were reversible or not. Instead the isolated
residue contained mostly the final reaction products (eq
4
). An alternative approach aimed at converting i into
some stable derivative by functionalizing the hydroxyl
group (i.e., to a silyl ether) was unsuccessful since i does
not appear to form in appreciable concentrations in
aprotic solvents (CDCl , C D , (CD ) CO, (CD ) SO).
3 6 6 3 2 3 2
Notably, in benzene solution, the first detectable product
1
F igu r e 5. Partial H NMR (400 MHz) spectra recorded at the
indicated times for the room-temperature reaction of 1 (7 mM)
with 0.5 equiv of benzenethiol-d
Trace d was recorded at 200 MHz.
5 8
in oxygen-free 2-PrOH-d .
was the benzenesulfenamide 5.
Deca y of th e In ter m ed ia te. In situ NMR analysis
of many reaction mixtures of different initial composition
allowed us to draw some conclusions concerning the fate
of intermediate i. The results of three representative
experiments in which the concentration of 1 was kept
(
AA′XX′ “doublets” at 6.64 and 6.95 ppm). Also evident
in trace d are the characteristic signals (two AA′XX′
doublets”) of 4,4′-dichloroazoxybenzene (3) as well as
“
some minor signals which could not be assigned.
When equimolar amounts of 1 and benzenethiol were
used in preparing intermediate i, a similar behavior was
observed (Figure 4), except for two main differences.
First, the azoxy compound 3 appears at earlier stages
6 5
constant (7 mM) while the [C D SH]/[1] ratio was fixed
at 5, 1, and 0.5, are summarized in Figures 3, 4, and 5,
respectively. Out of the many NMR recordings taken over
the whole reaction time, four representative traces are
shown for each experiment to illustrate the reaction
course. The data of Figure 3, which refer to reaction of
intermediate i run in the presence of excess benzenethiol,
are the easiest to analyze. It is seen (trace b) that as i
decays, a new system of signals develops which in part
overlaps that of i. This new signal matches exactly that
of the ring protons of N-(4-chlorophenyl)hydroxylamine
(
(
it is detectable after 1 h, trace b) and in larger amounts
compare traces d of Figures 3 and 4). Second, the decay
of the sulfenamide (5) signal appears to be quite slower,
so that a significant amount of this intermediate is still
present after 5.5 days (again compare traces d of Figures
3
and 4).
When only 0.5 equiv of thiophenol was used (Figure
(
4), thus suggesting that 4 is the second reaction inter-
mediate. At longer reaction times a new signal appears
7.04 ppm, singlet), as seen in trace c recorded after 7.5
5), i.e., when the reaction of i was initiated in the
presence of an equimolar amount of 1 (trace a), produc-
tion of the azoxy compound 3 was overwhelming. The
usual conversion of i into the hydroxylamino product 4
is evident in trace b of Figure 5 recorded 20 min after
mixing. Note already at this short reaction time the
presence of some 3, which after 1 day (trace d) constitutes
the major component of the reaction mixture. A small
amount of 1 and a trace amount of an unidentified
compound (“singlet” at 7.09 ppm) are also present at this
(
h, which is attributed to N-(4-chlorophenyl)benzenesul-
fenamide (5). As both i and 4 slowly decay, 5 builds up
(traces not shown) and in turn decays to leave a product
mixture composed mainly (trace d) of 4-chloroaniline (2)
(19) Montanari, S.; O’Carroll, F.; Paradisi, C.; Scorrano, G.; Traldi,
P. Rapid Commun. Mass Spectrom. 1995, 9, 1081, 1473.

3
426 J . Org. Chem., Vol. 64, No. 10, 1999
Montanari et al.
Sch em e 1
The process is initiated by fast condensation of the NO
and SH functionalities to form a covalent adduct, the
N-(4-chlorophenyl)hydroxybenzenesulfenamide i, the H
1
NMR spectrum of which is shown in trace b of Figure 1.
In an earlier investigation19 of a solution of 1 and
benzenethiol in glycerol by means of FAB mass spec-
+
trometry, we observed signals due to species [MH] ,
corresponding in mass to the sum of the two reagents
+
plus one proton, and [MH - H
2
O] . In contrast, the FAB
mass spectrum of an authentic sample of sulfinamide 6,
recorded under the same experimental conditions, con-
tained no signal due to water loss. These observations,
consistent with earlier reports1
3,11
clearly indicate that i
is an isomer of 6. Deeper insight into the FAB-MS process
was achieved by analysis of the time dependence of these
signals coupled to structural analysis by the MIKE
technique. It was concluded that the adduct formed upon
mixing the two reagents in glycerol reacts according to
two independent paths (i) FAB-induced protonation
+
(
presumably on oxygen) to an unstable [MH] followed
6 4
by water elimination to give the ionic species 4-ClC H -
+
8
Nd S-Ph; (ii) isomerization to 6 which, under FAB
+
stage. Interestingly, in contrast to the experiments
described previously, the signal of the sulfenamide 5 (7.04
ppm) was not detected at any time in the course of this
experiment. A control NMR experiment indicated that 4
and 1 undergo condensation to 3 under the typical
reaction conditions used in this study. The spectrum
shown as trace d did not undergo further changes over a
period of 5 days.
conditions, undergoes protonation to give a stable [MH]
species.19
As for the fate of the initial addition product i, however,
our observations and conclusions differ considerably from
those made in related studies carried out in aqueous
solvents and in benzene. Our NMR data suggest that,
although the final stable product composition and the
time required to reach it depend on the reactants molar
ratio, reaction of i is always initiated by its conversion
to the corresponding N-arylhydroxylamine 4. N-Arylhy-
droxylamines were also observed in reactions in aqueous
media. Their formation was attributed to reaction of the
primary adduct N-arylhydroxysulfenamide, ArN(OH)SR,
with free thiol, but no suggestion was provided about the
mechanism involved. In our system, however, any reac-
tion requiring the involvement of free thiol is not suf-
ficient to account for experimental observations such as
the following: the reaction mixture composition after 1
h of experiments run with benzenethiol in equimolar
(Figure 4, trace b) and in 5-fold excess (Figure 3, trace
b) concentrations with respect to the nitrosobenzene are
very similar.
A few experiments were performed to investigate on
the origin of the sulfenamide 5. According to one pro-
9
posal, the dithioketal 7 (4-ClC
6
H
4
N(SPh)
2
) should form
as an intermediate species in the conversion of i into 5.
We searched for compound 7 (“doublets” at 7.32 and 7.48
ppm) in our experiments, but found no detectable amounts
of it under any of the conditions used. Our observations
instead suggest that 5 derives from 4. A hypothetical
route involving reaction of 4 with benzenethiol could be
ruled out since 4 remained unaffected upon prolonged
6
treatment with excess benzenethiol in 2-PrOH-d
8
. Sup-
port for the proposal that 4 is involved in the production
of intermediate 5 came from a crossover experiment in
which 1 equiv of N-(4-bromophenyl)hydroxylamine was
added to a solution of i freshly prepared by mixing 1 (7
mM) with 1.2 equiv of benzenethiol. GC-MS analysis of
the resulting solution revealed the presence of 4,4′-
bromochloroazoxybenzene and 4,4′-dibromoazoxybenzene
besides the usual reaction products. It also indicated the
presence of 5 already after 3 h as opposed to the longer
reaction time required in the absence of added external
hydroxylamino reagent.
Conceivable radical pathways to 4 were then consid-
20
ered, based on known processes adapted as in eqs 5 and
15
6
.
Finally, an important result concerns sulfinamide 6,
which was never observed (“doublets” at 7.38 and 7.42
ppm) at any stage in any of the various experiments
performed with [C
6
D
5
SH]/[1] ratios ranging from 0.15 to
Both involve reaction of the arylhydronitroxide radical,
5
. Interestingly, the sulfinamide is a major product of
a species which has been detected in 2-propanol solutions
of nitrosobenzenes and PhSH. Both can be ruled out,
reaction 6 for the reasons expressed in the preceding
paragraph (it does not account for the production of 4 in
experiments run in the absence of excess free ben-
zenethiol), and reaction 5 because it produces equimolar
6
,8
16
reaction in aqueous media and, to a certain extent, also
in benzene.16
Discu ssion
The results described above establish the sequence of
steps which account for the 2-propanol reaction of ni-
trosobenzenes with benzenethiol. A proposal consistent
with our observations is presented in Scheme 1.
(
20) Aurich, H. G. In Supplement F: The Chemistry of Amino,
Nitroso and Nitro Compounds and Their Derivatives; Patai, S., Ed.; J .
Wiley & Sons: New York, 1982; Part 1, Ch. 14.

Nitrosobenzene Reduction by Thiols in Alcoholic Media
J . Org. Chem., Vol. 64, No. 10, 1999 3427
amounts of 4 and 1, which should, as shown in a control
experiment, undergo condensation to 3.
Sch em e 2
A third radical process can be envisioned based on
unimolecular reaction of i via homolytic S-N fission, a
process which occurs readily in sulfenamides, in some
cases simply upon exposure to diffuse sunlight.21 If one
assumes similar lability for i then the sequence of eqs 7
and 8 could account for the formation 4.
As for the nature of the H-donor species in eq 8, in
experiments run with free benzenethiol, this reagent
would obviously be involved in such a step. On the other
hand, since the i f 4 conversion (Scheme 1, path b) was
observed under any of the conditions used, including the
case in which reaction of i was initiated in the absence
6 5
of free benzenethiol and nitrosobenzene reagents ([C D -
SH]/[1] ) 1), we believe that the solvent itself must be
mechanism proved unlikely also for reactions in aqueous
media since only ArNHSR and no ArNHSR′ formed when
ArN(OH)SR was treated with R′SH.13
involved in this step. It could intervene directly in eq 8
3 2
to produce the strongly reducing (CH ) (HO)C• radical.
Alternatively, an undissociated molecule of i could act
as the H-donating species in eq 8, leading to 4 and to
the 4-ClC H N(O•)SPh radical, which in turn could be
6 4
reduced to i by a molecule of 2-propanol. Notably, such
radical was not detected in 2-propanol solutions of 1 and
benzenethiol.16
Depending on the experimental conditions used, the
next NMR signals to appear are due to benzenesulfena-
Alternative routes to 5 had therefore to be considered.
One, consistent with all the experimental data collected
so far, is that of step d of Scheme 1, in which reduction
of i to 5 is accompanied by oxidation of 4 to the nitroso
precursor 1: the latter can either be trapped by excess
benzenethiol and recycle as i or compete for 4 and enter
path c. Route d, therefore, accounts for the formation of
the azoxy compound in reactions of the intermediate i
initiated in the absence of free nitroso compound, i.e.,
when an original [C D SH]/[1] ratio g 1 is used. Azoxy-
6 5
mide 5 and/or to the azoxy compound 3. When a [C D -
SH]/[1] ratio g 1 is used, the former is observed as a
reaction intermediate which accumulates to significant
concentrations before decaying to aniline 4 (see traces c
and d of Figure 4). The build-up and decay of 5 is also
quite evident in additional traces, not shown, pertinent
to the experiment of Figure 3. In contrast, the azoxy
compound 3 prevails when the nitroso reagent is used
in excess with respect to benzenethiolsnotably, the
benzenesulfenamide 5 was never detected under these
conditions.
6
5
benzenes can also be produced via dimerization of arylni-
troxide radicals, ArN(O•)H.20
Step d is also consistent with the results of a crossover
experiment in which addition of N-(4-bromophenyl)-
hydroxylamine to a solution of i stimulated the produc-
tion of 5 and led to, besides the usual products, also some
4,4′-bromochloroazoxybenzene and 4,4′-dibromoazoxy-
benzene, as summarized in Scheme 2.
For the absence of the sulfinamide 6, the 2-propanol
reaction differs remarkably from that in aqueous media
Three different mechanisms have been proposed to
account for the production of sulfenamides ArNHSR in
the reaction of nitrosobenzenes (ArNO) with thiols (RSH)
in aqueous media: (1) reaction of the original adduct
6
,8
for which sulfinamides are major products of reaction.
Notably PhNHS(O)Ph was also isolated (17% yield) from
1
6
the reaction of nitrosobenzene with PhSH in benzene,
ArN(OH)SR with a molecule of thiol, a molecule of
and 6 was detected by FAB-MS analysis of a solution of
6
19
[
RSOH] being the side product; (2) two-step reaction,
4-chloronitrosobenzene and benzenethiol in glycerol.
each step requiring one molecule of thiol, of the original
adduct ArN(OH)SR via the intermediacy of dithioketal
ArN(SR) ; (3) reaction of ArNHOH with thiol. All three
2
Certainly the role of the solvent on the stability and
reactivity of the various intermediates and, therefore, on
the outcome of these reactions is of major importance.
This was clearly demonstrated in the present study also
for the reaction first step: in sharp contrast with the
behavior in 2-propanol, no indication was obtained by in
situ NMR analysis for the presence of i in benzene
solution, the first detectable product being in this case
5.
9
12
require free thiol as reagent and appear therefore inad-
equate to accommodate our finding that 5 forms also in
the absence of free thiol. In addition, evidence against
the operation of mechanism 3 in our system was obtained
in an independent experiment which proved the stability
of 4 upon prolonged treatment with excess benzenethiol,
6
a finding matching some earlier observations. Lack of
Finally, it should be pointed out that no evidence was
collected, in any of our many different NMR runs, for the
occurrence of radical species. Thus, although radicals
have been detected by EPR spectroscopy in different
detection of the dithioketal 7 in all of our experiments,
on the other hand, speaks against mechanism 2. Such a
(21) (a) Horspool, W. H. In The Chemistry of Sulphenic Acids and
14-16
16
solvents
including 2-propanol, and their involve-
Their Derivatives; Patai. S., Ed.; J . Wiley & Sons: Chichester, 1990;
Ch. 11, p 517. (b) Bayfield, R. F.; Cole, E. R. Phosphorous Sulphur
ment in some of the elementary steps comprised in the
proposed complex reaction scheme (Scheme 1), such as
1
976, 1, 19.

3
428 J . Org. Chem., Vol. 64, No. 10, 1999
Montanari et al.
step b, is very likely, their concentration and/or lifetime
is evidently not sufficient to produce any appreciable line-
broadening phenomena in the NMR signals.
was used for GC-MS analysis, with a 15-m fused silica column
of poly(dimethylsiloxane)-bonded phase.
Ma ter ia ls. Reagent grade 2-propanol was fractionally
distilled from Mg turnings. 2-Propanol-d
deuterated solvents (chloroform, benzene-d
acetone-d , and DMSO-d ) were the products of Cambridge
8
and the other
Con clu sion s
6
, methanol-d ,
4
6
6
In 2-propanol at room temperature nitrosobenzenes are
trapped by benzenethiol to form adducts of N-arylhy-
droxysulfenamide structure. This reaction is fast and
quantitative, and its product is relatively stable, thus
allowing for convenient monitoring of subsequent com-
plex transformations by in situ NMR analysis. In contrast
to the reaction in aqueous media, which involves several
competing steps for the decay of this primary intermedi-
ate, our data suggest one single path leading to the
corresponding N-arylhydroxylamine. A new mechanism
is suggested for this conversion, based on homolytic N-S
fission and reduction of the resulting aminyl radical. The
N-arylhydroxylamine reacts according to either of two
courses, leading to azoxybenzene (step c) and to N-
arylbenzenesulfenamide (step d), respectively, the par-
titioning depending on the initial reagents molar ratio.
When PhSH is used in excess, path d prevails since the
nitroso precursor generated as a side product in this step
is trapped by PhSH and recycled. When the nitroso
reagent is used in excess, path c prevails due to the
efficient condensation between the NO and NHOH func-
tionalities. Another striking difference concerns the
absence in the 2-propanol reaction of benzenesulfina-
mides (ArNHS(O)Ph), major products of the reaction in
aqueous media. The present investigation highlights the
crucial role played by the solvent on the reactivity in
these complex systems. It also shows that, in contrast
with what suggested for reaction in aqueous media,
reaction of the adduct does not require free benzenethiol.
Isotope Laboratories. They were used as received and stored
over molecular sieves. Benzenethiol was freshly distilled before
use and stored under argon. 4-Chloronitrosobenzene (1) was
22
prepared and purified as reported earlier. N-(4-chlorophenyl)-
23
23
hydroxylamine (4), N-(4-bromophenyl)hydroxylamine, N-(4-
24
chlorophenyl)benzenesulfenamide (5), N-(4-chlorophenyl)-
benzenesulfinamide (6), N-(4-chlorophenyl)-N-(phenylthio)-
benzenesulfenamide (7), 4,4′-dichloroazossibenzene (3), and
4,4′-dichloroazobenzene (9)28 were prepared and purified ac-
cording to published procedures. Perdeuterated thiophenol,
2
5
2
6
27
5 5
PhSH-d , was prepared from bromobenzene-d (g99.5%-D,
2
9
Fluka) by a literature method and purified by vacuum-
distillation (Kugelrohr B u¨ chi GKR-51).
P r od u ct Isola tion . A deoxygenated solution of 0.3 g (2.1
mmol) of 4-chloronitrosobenzene and 1.25 mL (10.6 mmol) of
benzenethiol in 25 mL of 2-propanol was stirred at 25 °C for
3
days. After addition of 25 mL of water, the mixture was
extracted with diethyl ether (3 × 50 mL). Flash chromatog-
raphy (petroleum ether/toluene 1:1) of the residue obtained
after evaporation of the solvent yielded 4-chloroaniline (0.19
g, 70% yield) and 4,4′-azoxybenzene (0.07 g, 25% yield).
NMR Exp er im en ts. Most spectra were conducted at 400
MHz (a few at 200 MHz), with a probe temperature of 25 °C
and a sweep width of 5000 Hz. Complete sets of reference
spectra in 2-propanol-d recorded both at 200 and at 400 MHz
8
were acquired for peak attribution. The general procedure was
as follows. A 1 mL aliquot of a 7 mM solution of 4-chloroni-
8
trosobenzene in 2-propanol-d was transferred into a 5 mm
diameter NMR tube fitted with a screw-cap equipped with a
septum. After sonication and gentle bubbling of Ar through
the solution, the NMR spectrum was recorded to provide the
“
time zero” data point. By means of a graduated microsyringe,
a small volume of a degassed concentrated (0.39 M) C SH
solution in 2-propanol-d was quickly added to obtain the
Exp er im en ta l Section
1H NMR experiments were performed on Bruker AC200 and
AC400 spectrometers. A Hewlett-Packard 5890 GC5970 MSD
6
D
5
8
desired final concentration, the tube inverted a few times to
ensure thorough mixing, and the time count started. Spectra
were recorded at desired times either by means of an auto-
mated procedure or manually. Automated aquisition was
especially useful to monitor the initial stages of the process.
The same procedure was used with the aliphatic thiols, except
with MeSH, the concentration of which was not determined.
In this case the reagent was bubbled directly through the
nitrosobenzene solution for a short time.
(
22) Paradisi, C.; Prato, M.; Quintily, U.; Scorrano, G. J . Org. Chem.
981, 46, 5156.
23) Vogel’s Textbook of Practical Organic Chemistry, 4th ed.;
Longman Inc.: New York, 1978; p 722.
24) Davis, F. A.; Friedmann, A. J .; Kluger, E. W.; Skibo, E. B.; Fretz,
E. R.; Milicia, A. P.; LeMasters, W. C.; Bentley, M. D.; Lacadie, J . A.;
Douglass, I. B. J . Org. Chem. 1977, 42, 967.
1
(
(
(
25) (a) Gilman, H.; Morris, H. L. J . Am. Chem. Soc. 1926, 48, 2399.
b) Klamann, C. S.; Zelenka, M. Chem. Ber. 1959, 1910. (c) Rajogola-
(
palan, P.; Andrani, B. G.; Talaty, C. N. Organic Syntheses; Wiley: New
York, 1975; Coll. Vol. 5, p 504.
Ack n ow led gm en t. Financial support by the British
Council and MURST (British-Italian Collaboration in
Research and Higher Education) is gratefully acknowl-
edged. We are grateful to Professors A. Alberti and L.
Greci for helpful discussions.
(
26) Benati, L.; Montevecchi, P. C.; Spagnolo, P. J . Chem. Soc.,
Perkin Trans 1 1985, 2261 and refs therein.
27) Galbraith, H. W.; Degering, F. E.; Hitch, E. F. J . Am. Chem.
Soc. 1951, 73, 1323.
(
(
28) Willgerodt, C. Ber. 1881, 14, 2632.
(29) (a) Wardell, J . L. In The Chemistry of the Thiol Group; Patai,
S., Ed.; J . Wiley & Sons: London, 1974; Part 1, Ch. 4. (b) Gilman, H.;
Fullhart, L. J . Am. Chem. Soc. 1949, 71, 1478.
J O981889T