Kinetics of the oxidation of aromatic C-nitroso compounds by nitrogen dioxide

Article abstract of DOI:10.1039/a700258k

The oxidation of nitrosobenzene by nitrogen dioxide in carbon tetrachloride has been re-examined, the rate of reaction exhibiting first-order dependence upon the concentrations of both nitrosobenzene and nitrogen dioxide, with a reactant stoichiometry of 1:1. The rate of reaction is found to be highly sensitive to the presence of trace quantities of water and intensive drying procedures for the solvent have been developed such that the reaction rate could be reduced to a constant level. For nitrosobenzene and a range of substituted nitrosobenzenes the kinetics of the oxidation reaction in carbon tetrachloride have been studied over the temperature range 291-313 K using the stopped flow technique and pseudo-first-order conditions with the nitrogen dioxide in excess. Arrhenius parameters have been obtained for 15 nitrosoarenes and the results obtained are discussed with reference to the Hammett σ constants and the electron population at the nitroso nitrogen as determined by CNDO calculations. It is concluded that the reaction occurs by a radical addition mechanism to the nitroso nitrogen atom giving an unstable aminoxyl intermediate which decomposes rapidly giving the corresponding nitrobenzene and nitric oxide.

Full text of DOI:10.1039/a700258k

View Article Online / Journal Homepage / Table of Contents for this issue
Kinetics of the oxidation of aromatic C-nitroso compounds by
nitrogen dioxide
Brian G. Gowenlock,*,† Josef Pfab* and Victor M. Young
Department of Chemistry, Heriot-Watt University, Riccarton, Edinburgh, UK EH14 4AS
The oxidation of nitrosobenzene by nitrogen dioxide in carbon tetrachloride has been re-examined, the
rate of reaction exhibiting first-order dependence upon the concentrations of both nitrosobenzene and
nitrogen dioxide, with a reactant stoichiometry of 1 :1. The rate of reaction is found to be highly sensitive
to the presence of trace quantities of water and intensive drying procedures for the solvent have been
developed such that the reaction rate could be reduced to a constant level. For nitrosobenzene and a range
of substituted nitrosobenzenes the kinetics of the oxidation reaction in carbon tetrachloride have been
studied over the temperature range 291–313 K using the stopped flow technique and pseudo-first-order
conditions with the nitrogen dioxide in excess. Arrhenius parameters have been obtained for 15
nitrosoarenes and the results obtained are discussed with reference to the H ammett ó constants and the
electron population at the nitroso nitrogen as determined by CN D O calculations. It is concluded that the
reaction occurs by a radical addition mechanism to the nitroso nitrogen atom giving an unstable aminoxyl
intermediate which decomposes rapidly giving the corresponding nitrobenzene and nitric oxide.
The oxidation in solution of C-nitroso compounds to C-nitro
compounds [reaction (1)] is well established for a wide variety
Experimental
C-Nitrosocompounds
R᎐N᎐O ϩ [O] → R᎐NO
(1)
These were all made by standard procedures usually from either
the corresponding nitro compound via reduction to the N-
substituted hydroxylamine followed by oxidation using either
ferric chloride or aqueous dichromate (Method 1), or by con-
trolled oxidation of the corresponding amino compound using
Caro’s acid (Method 2). The products were purified by
recrystallisation and in some cases this was followed by sublim-
ation. All compounds were stored at Ϫ20 ЊC. The solid com-
pounds were trans-dimers with the exception of nitrosobenzene
itself which was the cis-dimer. The compounds prepared are
listed in Table 1.
᎐
2
of oxidising agents although kinetic studies of such oxidation
reactions are confined to the oxidants peroxyacetic acid,1 per-
oxychloroacetic acid,1 Caro’s acid,1 dinitrogen tetroxide,2 per-
oxydisulfate anion 3 and nitric acid 4 with a relatively limited
range of substituted nitrosobenzenes, i.e. p-H, Cl, Br, CH3,
OCH3,1 2,5-dimethyl,2 OH 3 and p-Cl, CH3 and m-Cl, NO2 and
CH3.4 In the first of these studies it was reported that the oxida-
tion by peroxyacetic acid in aqueous ethanol was second order
overall, being first order in each reactant. The reaction was
accelerated by electron-releasing para substituents, retarded by
electron-withdrawing para substituents and accelerated also by
increasing the water content of the solvent. The authors1 sug-
gested that the nitroso nitrogen performed a nucleophilic attack
on the outer oxygen atom of the peroxyacid, the transition state
being made up of the peroxyacid, the nitroso compound, and a
solvent molecule. A similar nucleophilic attack was proposed
for the nitroso nitrogen in the reaction of the p-nitroso-
phenoxide ion with the peroxydisulfate dianion.3 Bonner and
Hancock 2 suggested the possibility of a 1,3-cycloaddition
mechanism when nitrogen dioxide, the reactive species in the
equilibrium with dinitrogen tetroxide, reacts with the 2,5-
dimethylnitrosobenzene. Ogata and Tezuka 4 suggested that the
mechanism of reaction (1) involved addition of the radical
species, nitrogen dioxide, to the nitroso group and that this
Nitrogen dioxide–dinitrogen tetroxide
Cylinder nitrogen dioxide (Matheson) was purified by mixing
with a stream of dry oxygen and passing slowly through a
2 × 20 cm column of phosphorus pentoxide before cooling and
collection. The resultant liquid was then slowly distilled under
reduced pressure and a small middle fraction was collected and
stored at 0 ЊC in a glass vessel fitted with a rotaflo tap. Altern-
atively, excess dry oxygen was mixed with nitric oxide which had
been purified by slow distillation at Ϫ150 ЊC. The nitrogen
dioxide product was deoxygenated and stored as before. All
samples were completely free from dinitrogen trioxide as the
solid material produced on cooling to low temperatures was
free from blue colour.
Dry white fuming nitric acid was prepared by the addition of
urea to commercial fuming nitric acid followed by the addition
of at least three times its volume of concentrated sulfuric acid
and the subsequent slow distillation of a fraction of the nitric
acid which was collected and stored at Ϫ20 ЊC in a glass vessel
fitted with a rotaflo tap.
ؒ
involved addition to the oxygen, producing the radical PhN O-
NO2, rather than addition to the nitrogen and production of
ؒ
PhN(ONO)O .
There is an early report 5 that a small quantity of nitro-
benzene was formed when dry chloroform solutions of nitrogen
dioxide and nitrosobenzene were allowed to stand at 22 ЊC for
39 h.
Gas phase studies of reaction (1) have been confined to the
case of C-nitrosoalkanes. Thus for the case of R = tert-butyl it
has been shown 6 that the reaction involved transfer of an oxy-
gen atom from the nitrogen dioxide to the nitroso compound,
and that the reaction obeyed mixed second-order kinetics.7
Commercial ‘white spot’ nitrogen was dried by passing it
through an activated bed of 5 Å molecular sieve.
Solvents
The carbon tetrachloride used as solvent for kinetic experi-
ments was commercially available and was purified by three
methods hereafter named G1, 2 and 3. G1 was purified by distil-
lation only. G2 was the product from G1 obtained by repeated
shaking with aliquots of concentrated sulfuric acid, followed by
† Present address: Department of Chemistry, University of Exeter,
Exeter, UK EX4 4QD.
J. Chem. Soc., Perkin Trans. 2, 1997
1793

View Article Online
Table 1 Preparation of substituted nitrosobenzenes
Substituent
Method
Ref.
Mp (lit. value)/ЊC
λmax(CCl4)/nm
εmax/dm3 molϪ1 cmϪ1
H
1
1
1
b
2
2
2
2
2
2
2
2
2
2
2
e
8
10
9
67 (689)
762
a
50
2,5-di-CH3
4-CH3
4-OCH3
4-Cl
4-Br
4-NO2
4-CN
4-COCH3
4-CO2CH3
3-CH3
3-OCH3
3-Cl
3-COCH3
3-CN
2-CH3
98 (101.511
)
48 (4812
)
760
742
756
760
788
782
780c
776
764
758
764
762
760d
760
53
56
46
49
43
38
42
45
50
50
28
45
23
39
13
15
15
16
16
16
16
16
16
16
16
16
—
32 (32–3414
)
90 (9015
94 (949)
)
119 (118–11917
)
135–136dec (136–13718
110 (11119
)
)
129.5 (128–129.520
)
53 (5312
49 (4821
72 (7212
81.5
)
)
)
107–108 (11122
)
a
Not measured. b Prepared from reaction of 4-nitrosophenol with methanol.13 c Solvent contains added 5% CH2Cl2. d Solvent contains 10% added
CH2Cl2. e Commercial product.
Merck Kieselgel GF254 precoated plates. Columns for prepar-
ative liquid chromatography were slurry packed with Merck
Kieselgel 60, 0.040–0.063 mm (230–400 mesh ASTM). HPLC
was carried out using a Haskell pressure multiplying pump and
the eluted fraction was monitored with a Cecil CE 212 variable
wavelength UV detector connected to a Servoscribe 1s chart
recorder. The HPLC columns were 5 mm bore, internally
polished, stainless steel tubes with a septum injector bolted to
the top and a fine sintered frit at the bottom. A 25 cm column
was filled with 5 µm Spherisorb silica gel by slurry packing from
a reservoir under a pressure of 3000 psi. Disturbance of the
surface of the column was avoided by use of a fine wire disc on
top of the packing. Samples were added through the septum,
directly onto the top of the column. The column was packed
using methanol and this was gradually changed to the required
solvent for analyses, namely isooctane–diethyl ether in a 9:1
ratio.
Following identification of the compounds present in a
sample, the HPLC was calibrated by injection of these com-
pounds in known concentrations at a fixed wavelength. The
amount of compound eluted was determined from the recorded
peak area and a previously determined response factor.
Fig. 1 Vessels used for the spectrophotometric method
passing the separated solvent down a column of commercial
alumina under dry nitrogen to remove any suspended acid
droplets and any remaining water. It was then stored and used
under dry nitrogen. G3 was prepared from G2 by adding a
further purification stage. The solvent was passed, under dry
nitrogen, down a column of highly activated alumina (previ-
ously heated to 250 ЊC under vacuum and handled under dry
nitrogen only), then degassed by repeated freeze–thaw cycles
and distilled into a storage vessel on the vacuum line. For all
subsequent usage it was vacuum distilled into the reaction
vessels and used for reaction under an atmosphere of dry nitro-
gen. For any case where dichloromethane was used as solvent
it was purified as for G3 carbon tetrachloride, stored under
vacuum with exclusion of light and used within two days of the
purification procedure.
UV–VIS absorption spectra were measured using a Pye-
Unicam SP 1800 spectrophotometer. NMR spectra were
obtained from a Perkin-Elmer R12, 60 MHz spectrometer and
IR spectra were recorded on a Perkin-Elmer 157G grating spec-
trometer. Kinetic studies were carried out using two techniques:
either by optical density measurements with the thermostatted
SP 1800 spectrophotometer or by stopped flow spectrometry
using a Nortech SF-1A apparatus.
Experimental method for the spectrophotometric technique
The dried solvents and reactants were introduced into and
handled in an all-glass reaction vessel as shown in Fig. 1. The
complete reaction vessel was evacuated and the interconnecting
taps were closed. The two reservoirs had known capacities and
any volume of solvent which filled the graduated stems could be
accurately read. The approximate weight of nitrosoarene was
weighed into one of the reservoir socket stems and then
degassed on a vacuum line. Solvent was then distilled in from a
storage bulb on the vacuum line and condensed into the socket
stem. The reservoir tap was then opened and all the reactant
was washed into the reservoir. More solvent was distilled in
until the level had reached an appropriate point on the gradu-
ated stem. The reservoir tap was then closed and the other side-
arm connected to the vacuum line and evacuated. Solvent was
distilled into the reservoir until just below the graduated stem.
The reservoir tap was closed and the socket stem evacuated. An
approximate amount of nitrogen dioxide was distilled in from a
small container, as shown in Fig. 1, and condensed into the
socket stem. The top of the nitrogen dioxide reservoir was
cooled using a cotton wool pad dipped in liquid nitrogen and
the reservoir tap was opened allowing the condensed nitrogen
dioxide to flow in rapidly. The volume of the solvent in the
reservoir plus the dissolved nitrogen dioxide now filled the res-
ervoir to a point on the graduated scale and its volume was read
off after thermostatting to 20 ЊC. The concentrations of both
Chromatographic techniques
Thin layer chromatographic plates used were commercial
1794
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
solutions of reactant in carbon tetrachloride were then filled
with dry nitrogen to atmospheric pressure from the vacuum line
and transferred to the stopped flow apparatus. At this point the
glass tubes of the stopped flow apparatus were filled with
carbon tetrachloride.
The oscilloscope scales were then adjusted with reactant
injections so as to obtain a full scale trace on the screen of
nitrosoarene disappearance with time. All traces and light levels
were then stored on the storage oscilloscope and recorded using
a Polaroid camera (black and white film) for future measure-
ment. A reaction run was thus performed at the designated
temperature, the trace recorded, and the flow cell flushed out
with either reactant. The temperature was then adjusted and
allowed to stabilise at the next desired temperature. Following
recalibration of the intensity levels a reaction run was per-
formed as before and then further repeated at additional tem-
peratures. The reaction storage vessels were attached to a sup-
ply of dry nitrogen and as each aliquot of sample or reactant
was drawn off into the thermostatted coils it was replaced by a
corresponding volume in the storage vessels.
From the photographic record the absorbance (A) of the
nitrosoarene was measured at various reaction times thereby
enabling the pseudo-first-order rate constant to be calculated.
The nitrogen dioxide concentration was obtained from the
known values23 for the equilibrium constant and enthalpy of
dissociation of the dinitrogen tetroxide–nitrogen dioxide sys-
tem and the concentration of dinitrogen tetroxide calculated
from its known 24 extinction coefficient of 200 dm3 molϪ1 cmϪ1 at
λmax 342 nm and its measured absorbance. The nitrogen dioxide
concentration remains constant throughout each experimental
run due to the large ‘reservoir’ of dinitrogen tetroxide and the
very fast re-establishment of equilibrium. From the nitrogen
dioxide concentration and the pseudo-first-order rate constant
the second-order rate constant k2 was obtained.
Fig. 2 Reactant storage vessel for the stopped flow apparatus
reactants were now measured spectrophotometrically by meas-
urement of their UV–VIS absorbances at the appropriate wave-
lengths, the extinction coefficients being known from previous
determinations. The apparatus was inverted to fill the two
upper UV cells, the spectrometer being fitted with a shaped
cover to exclude any outside light. To obtain the concentration
of nitrogen dioxide it was necessary to consider the equilibrium
with dinitrogen tetroxide; from knowledge of the volumes of
both reactant solutions it was possible therefore to calculate the
equilibrium concentrations of both NO2 and N2O4 in the mixed
solutions as well as the initial concentration of the nitrosoarene
in the reaction chamber. The mixing of the thermostatically
controlled solutions (20 ЊC) was carried out by opening the taps
connecting the sidearms to the reaction chamber, timing the
start of the reaction, shaking the apparatus to ensure thorough
mixing of the reactants, followed by rapid insertion into the
spectrophotometer where the disappearance of the nitroso-
arene was monitored using the absorption in the range 700–800
nm where neither the nitrogen dioxide–dinitrogen tetroxide nor
the nitroarene product absorb light. This technique enables the
rate constant for the bimolecular reaction between nitrogen
dioxide and the nitrosoarene to be determined, but the stopped
flow technique is both more accurate and more convenient to
use and the description of this technique follows.
Results
The stoichiometry of reaction (2) was established by Bonner
Experimental method for the stopped flow technique
RNO ϩ NO2 → RNO2 ϩ NO
(2)
The reactant storage vessels used in our experiments were
designed to avoid moisture and air contact and are shown in
Fig. 2. A storage vessel was connected to the vacuum line and
evacuated. It was then weighed and a known weight of
nitrosoarene was inserted into the stem of the connection to
the vacuum line and the connection stem was evacuated. Dried
solvent was then distilled in from a storage bulb and the upper
tap opened allowing the nitrosoarene to be washed completely
into the storage vessel. Solvent was then distilled in until the
vessel was approximately three-quarters full. The taps were then
closed and the vessel allowed to warm to room temperature
after which it was weighed. The volume of solvent contained in
the reservoir could then be calculated from its initial weight and
the weight of added nitrosoarene. Alternatively the nitroso-
arene concentration could be determined from the absorbance
of the nitrosoarene at its λmax in the 750 nm region using the
stopped flow spectrophotometer. The agreement between these
two measurements showed that any loss by sublimation of the
nitrosoarene in its introduction procedure was negligible.
The nitrogen dioxide solution was made up in a similar fash-
ion by first evacuating a reactant storage vessel, weighing it,
then three-quarters filling it with dried solvent and reweighing
it. Nitrogen dioxide was then distilled into a small storage vessel
as described for the spectrophotometric method and accurately
weighed. It was then carefully and quantitatively transferred to
the carbon tetrachloride in the reaction storage vessel using
liquid nitrogen cooling of the storage vessel neck to ensure
complete transfer. This vessel and the small nitrogen dioxide
storage vessel were then reweighed to ensure that transfer was
complete. Both reaction storage vessels with their respective
and Hancock 2 for the case of 2,5-dimethylnitrosobenzene: they
showed that the ratio of RNO :NO2 :RNO2 was 1:1:1. We have
confirmed this for nitrosobenzene using quantitative HPLC
analysis and assume that it is operative for all the substituted
nitrosobenzenes which we have studied. Support for this
assumption is afforded by our observation that in all cases TLC
analysis of the products of reaction revealed only one product.
In the Experimental section we have drawn attention to the
different levels of purification of solvent that we have
employed. These considerations are exemplified by the rate con-
stant data collected in Table 2 which relate to the influence of
solvent dryness and influence of air upon the values obtained
for the nitrosobenzene–nitrogen dioxide reaction. Because of
its importance as the only case in the literature where a rate
constant value had been obtained for the oxidation by nitrogen
dioxide of a substituted nitrosobenzene, a rate constant meas-
urement was made for 2,5-dimethylnitrosobenzene using G3
solvent under nitrogen and UV spectrophotometry at 292 nm
as in ref. 2. The reaction was so slow as to be almost unobserv-
able and the maximum possible value for the rate constant at
20 ЊC was 0.028 dm3 molϪ1
s
Ϫ1. Discussion of the difference
between this value and that reported by Bonner and Hancock 2
(1.86 dm3 molϪ1 sϪ1) is deferred to the Discussion section.
A preliminary study of the reaction of nitrosobenzene with a
six-fold excess of dry fuming nitric acid in G3 solvent was made
using spectrophotometric monitoring of the nitrosobenzene
concentration. An initial small drop in the nitrosobenzene con-
centration was followed by no change in absorbance over a
period of 16 hours. A slow increase in absorbance followed over
J. Chem. Soc., Perkin Trans. 2, 1997
1795

View Article Online
Table 2 Dependence of rate constant values at 20 ЊC upon experi-
mental conditions
Experimental conditions
k/dm3 molϪ1 sϪ1
Solvent
Gas phase
Other
Apparatus
G1
G1
G1
G1
G1
G2
G3
G3
G3
G3
G3
Air
Air
Air
Air
Air
N2
N2
N2
N2
N2
a
b
b
b
b
a
a
a
a
b
b
0.79
0.97
1.02
0.81
0.74
0.42
0.18
0.27
0.27
0.13
0.21
c
d
Fig. 3 Variation at 298 K of log10 (k2/dm3 molϪ1 sϪ1) with Hammett σ
constant of substituent X in XC6H4NO, p-X (᭿), m-X(᭜)
e
f
g
N2
our values are similar to those obtained for the m-chloroper-
benzoic acid oxidation of nitrosobenzene in a variety of sol-
vents (44–60 kJ molϪ1) and Diels–Alder addition reactions of
2,3-dimethylbutadiene to para-substituted nitrosobenzenes
(49–60 kJ molϪ1), but higher than those for the addition of alkyl
radicals to nitrosocompounds (4–8 kJ molϪ1).
a
Spectrometer.b Stoppedflow.c PhNO :NO2 = 2:1.d PhNO :NO2 = 1:2.
e Highly purified commercially available NO2. f Highly purified NO2
prepared from NO–O2. g Highly purified NO2 containing traces of
white fuming HNO3.
a two hour period followed by a decrease accompanied by evo-
lution of nitrogen dioxide. In order to test the assumption of an
initial protonation of nitrosobenzene followed by a long induc-
tion period during which decomposition of the nitric acid takes
place, the experiment was repeated with addition of crystalline
sodium nitrite to the mixture. A yellow coloration appeared
followed by slow evolution of nitrogen dioxide accompanied by
a more rapid diminution of the nitrosobenzene concentration.
Some of these studies of the rate constants of reactions of
substituted nitrosobenzenes have also claimed correlations with
Hammett substituent constants.4,12,27,28,30 The major feature of
any such correlation is the expectation that there will be a
smooth correlation between the rate constant and the electron-
donating and -accepting characteristics of the substituents. In
our studies all of the para-substituents (with the exception of
COCH3) show rate constants lower than for nitrosobenzene
including both negative and positive values of σ. A linear
relationship (ρ = Ϫ1.14) is found for the meta-substituents,
there being no evidence for a linear plot for the para-
substituents or for the two sets combined (see Fig. 3). For some
other reactions of substituted nitrosobenzenes the values of ρ
are negative e.g. Ϫ0.5 (nitric acid oxidation 4), Ϫ1.58 (peroxy-
acetic acid oxidation 1), Ϫ1.74 (formaldehyde addition to give
N-phenylhydroxamic acids32), although ρ = ϩ1.22 for the con-
densation reaction of aniline with substituted nitrosobenzenes
in acid solutions.12 For the para-substituents in this study, there
is an unusual feature in that all of them, excepting acetyl, have
rate constants lower than for nitrosobenzene itself. This con-
trasts with all other studies of rates of reaction of para-
substituted nitrosobenzenes. There is however a possible paral-
lel in the 13C NMR spectroscopy of para-substituted nitroso-
benzenes33 where it is found that the π-electron accepting
property of the nitroso group is such that it can turn other π-
electron acceptor groups in the para-position into π-electron
donors. It is clear from previous studies6 that in the overall
reaction (2) an oxygen atom from the nitrogen dioxide is trans-
ferred to the nitrogen of the nitroso compound as distinct from
the nitrogen dioxide molecule displacing the nitroso group. We
suggest that this transfer can be achieved in three ways compat-
ible with the second-order kinetics that we have established. In
the first of these nitrogen dioxide, itself a free radical, adds to
the nitrogen atom of the nitroso compound to give an inter-
mediate aminoxyl radical as in reaction (3), the decomposition
Discussion
The effect of water upon the reaction is exemplified by the pro-
gressive lowering of the rate constant as more stringent experi-
mental conditions for the removal and exclusion of moisture
are introduced. Once the effect had been recognised for
nitrosobenzene itself it was necessary to carry out a study of
the nitrogen dioxide–2,5-dimethylnitrosobenzene reaction as
this was the only reaction of nitrogen dioxide with a nitroso-
arene which had previously been subjected to a kinetic investi-
gation. The very large difference in the values of the rate con-
stant obtained is due to the differences in the experimental
techniques employed. In the method adopted by Bonner and
Hancock,2 the reactants, each dissolved in ‘reasonably dry’
carbon tetrachloride and thermostatted in flasks, were pipetted
together into a reaction cell. Our technique emphasises rigorous
drying, handling the reactants under dry nitrogen and strict
exclusion of the atmosphere. The apparatus used for the
stopped flow technique permits these conditions to be better
achieved than does the use of the spectrophotometric cell
method and consequently our discussion of results obtained
relates entirely to the stopped flow kinetic data. We consider
that we have reduced any water effect to negligible proportions.
The reaction with 2,5-dimethylnitrosobenzene was not investi-
gated further by the stopped flow method because it is too slow
to yield accurate results by this method, as back diffusion of the
reactants into the reaction cell would be considerable over the
necessary reaction time.
O•
There are several studies of the kinetics of bimolecular reac-
tions of substituted nitrosobenzenes1–4,12,25–31 and all of these,
studied over a range of temperatures, have large negative entro-
pies of activation. Thus the oxidation reactions1,25 have ∆S ‡
(3)
R–N–ONO
R–NO2 + NO
RNO + NO2
of the aminoxyl radical being fast. In the second of these the
addition of the nitrogen dioxide gives a transition state as in
reaction (4). The third mechanism is provided by Bonner and
values in the range Ϫ109 to Ϫ138 J molϪ1
K
Ϫ1, the condens-
ation reactions12,26 have ∆S ‡ values in the range Ϫ227 to Ϫ264
J molϪ1
K
Ϫ1, the Diels–Alder reactions29,30 have ∆S ‡ values in
the range Ϫ123 to Ϫ145 J molϪ1 KϪ1 and the radical addition
reactions31 have ∆S ‡ values in the range Ϫ93 to Ϫ129 J
O
RNO + NO2
R–N
O
N
O
(4)
molϪ1
K
Ϫ1. Our values for ∆S ‡ are similar to those for the
oxidation, Diels–Alder and radical addition reactions. Extend-
Hancock 2 who suggested that the large negative entropy of
activation for reaction (2) was consistent with nitrogen dioxide
ing these considerations to the ∆H ‡ values it can be seen that
1796
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
Table 3 Rate constant values and Arrhenius parameters for the reaction of XC6H4NO with NO2 at 298 K using the stopped flow technique and G3
solvent under N2
X
H
k298/dm3 molϪ1 sϪ1
Ea/kJ molϪ1
A/dm3 molϪ1 sϪ1
∆H ‡/kJ molϪ1
∆S ‡/J molϪ1 KϪ1
Ϫ102.9
Ϫ100.8
Ϫ106.7
Ϫ82.4
Ϫ115.6
Ϫ90.4
Ϫ109.2
Ϫ47.7
Ϫ118.4
Ϫ127.2
Ϫ151.9
Ϫ123.8
Ϫ166.9
Ϫ99.2
0.263
0.214
0.089
0.105
0.144
0.098
0.247
0.085
0.264
0.345
0.211
0.110
0.062
0.119
0.142
0.098
48.1
49.4
49.8
56.8
45.6
54.4
46.4
67.4
43.5
40.2
34.3
43.9
32.6
51.0
41.4
46.1
7.0 × 107
9.3 × 107
4.4 × 107
8.4 × 108
1.5 × 107
3.3 × 108
3.4 × 107
4.5 × 1010
1.1 × 107
3.8 × 106
2.0 × 105
5.8 × 106
3.1 × 104
1.1 × 108
2.5 × 106
1.3 × 108
45.6
46.9
47.3
54.0
43.1
51.9
43.9
64.9
41.0
37.7
31.8
41.4
30.1
48.5
38.9
43.6
p-CH3
p-OCH3
p-Cl
p-Br
p-CN a
p-CO2CH3
b
c
p-NO2
d
p-COCH3
m-CH3
m-OCH3
m-Cl
m-CN e
m-COCH3 f
m-COCH3
Ϫ131.0
Ϫ98.0
g
o-CH3
a
Solvent contains 8.0% CH2Cl2. b Solvent contains 4.0% CH2Cl2. c Solvent contains 6.3% CH2Cl2. d Solvent contains 4.4% CH2Cl2. e Solvent contains
7.1% CH2Cl2. f Solvent contains 5.5% CH2Cl2. g We are indebted to Dr B. King for these measurements.
order COCH3 > CN > OCH3 > CH3 > H). If our explanation
of radical addition to nitroso N is correct then it seems likely
that a similar effect would hold for addition reactions of other
radicals to para-substituted nitrosobenzenes.
It has been shown that when nitrogen dioxide reacts with
excess nitrosobenzene in methylene chloride36 the reaction
products are nitrobenzene and benzenediazonium nitrate, the
latter product having been formed by reaction of the nitric
oxide produced in the oxidation process reacting further with
the excess nitrosobenzene as established by Bamberger.5 Our
reaction conditions (excess nitrogen dioxide) avoid this second-
ary reaction of the nitric oxide produced.
The influence of nitric acid and water upon the reaction
systems requires consideration. The observation that addition
of purified anhydrous nitric acid to nitrosobenzene in carbon
tetrachloride led to an initial drop in the absorbance of the
nitrosobenzene is probably due to the formation of a hydrogen
bonded association complex [equilibrium (6)] followed by the
reaction sequence (7)–(10).
Fig. 4 Variation at 298 K of log10 (k2/dm3 molϪ1 Ϫ1) with electron
s
population at nitroso N in p-XC6H4NO as calculated by the CNDO
method
acting as a 1,3 dipole leading to a cyclic transition state as in
reaction (5) followed by oxygen transfer and formation of nitric
oxide.
PhNO ϩ HNO3
PhNOؒHNO3
PhNO2 ϩ HNO2
H2O ϩ 2NO2
(6)
(7)
O
O
+O
–O
slow
+
R
N
O
(5)
RNO2 + NO
N•
R–N:
N•
PhNO ϩ HNO3
HNO2 ϩ HNO3
PhNO ϩ NO2
NO ϩ HNO3
O
(8)
We have noted however that large negative entropies of activ-
ation are experienced for other bimolecular reactions of
nitrosobenzenes without cyclic transition states and that this
includes radical addition reactions. It is difficult to envisage the
oxidation of the nitrosoarene taking place without the attach-
ment of at least one of the oxygen atoms of the nitrogen
dioxide onto the nitroso nitrogen in the rate-determining step.
The long established ability of C-nitroso compounds to act as
radical trapping agents lends further support for radical attack
by nitrogen dioxide at nitroso nitrogen and we suggest that the
first of these possible mechanisms is the most likely. Support is
provided by CNDO calculations of the electron populations at
the nitrogen atom of para-substituted nitrosobenzenes34 which
demonstrate that as this electron population increases so does
the rate constant for reaction (2), (see Fig. 4) and that the
relative order COCH3 > H > CH3 > CN > OCH3 > NO2 is
consistent for the two variables. The figure is drawn with a
linear relationship between log k2 and the electron population
at the nitrogen: a curvilinear relation may equally be appropri-
ate. Whatever the interrelationship may be, we have no know-
ledge of any other substituent parameter that will produce the
relative order given above. The possibility of a correlation with
PhNO2 ϩ NO
HNO2 ϩ NO2
(9)
(10)
After an induction period of some 16 hours the nitroso-
benzene concentration began to decrease due to reaction (7)
and the resultant nitrogen dioxide produced in equilibrium (8)
reacts by reaction (9). Equilibrium (10) is pushed to the right by
the excess of nitric acid and by the reaction of the nitrogen
dioxide in reaction (9). Similarly equilibrium (8) is also pushed
to the right and the production of water also accelerates reac-
tion (9). Addition of sodium nitrite removes the induction
period by production of nitrous acid followed by equilibrium
(8). On the other hand, addition of trace amounts of nitric acid
to a reaction mixture of nitrosobenzene–nitrogen dioxide has
no observable effect upon the rate of removal of the nitroso-
benzene.
The acceleration of the reaction rate by trace quantities of
water in the solvent is possibly due to the formation of nitrous
acid [equilibrium (8)] followed by production of nitric oxide by
equilibrium (10) and reaction of these species with the nitroso
compound adding to the rate of its removal. The acceleration
the substituent constant 35 for radical reactions, σ , has been
explored but found to be inoperative. (The σ values fall in the
ؒ
ؒ
J. Chem. Soc., Perkin Trans. 2, 1997
1797

View Article Online
13 F. E. Williams, USP 3 417 147, 1968; Chem. Abstr. 1969, 70, 47 089w.
14 A. Baeyer and E. Knorr, Ber. Dtsch. Chem. Gesell., 1902, 35, 3034.
15 C. K. Ingold, J. Chem. Soc., 1924, 125, 93.
16 A. Schors, A. Kraaijeveld and E. Havinga, Recl. Trav. Chim. Pays-
Bas., 1955, 74, 1243.
17 E. Bamberger and R. Hübner, Ber. Dtsch. Chem. Gesell., 1903, 36,
3803.
18 J. N. Ashley and S. S. Berg, J. Chem. Soc., 1957, 3089.
19 J. H. Hall and F. W. Dolan, J. Org. Chem., 1978, 43, 4608.
20 F. J. Alway and A. B. Walker, Ber. Dtsch. Chem. Gesell., 1903, 36,
2312.
21 O. Baudisch and R. Furst, Ber. Dtsch. Chem. Gesell., 1915, 48, 1665.
22 A. Darchen and C. Moinet, Bull. Soc. Chim. Fr., 1976, 812.
23 T. F. Redmond and B. B. Wayland, J. Phys. Chem., 1968, 72, 1626.
24 C. C. Addison and J. C. Sheldon, J. Chem. Soc., 1958, 3142.
25 K. M. Ibne-Rasa and J. O. Edwards, Int. J. Chem. Kinet., 1975, 7, 575.
26 R. A. Yunes, A. J. Terenzani, O. D. Andrich and C. A. Scarabino,
J. Chem. Soc., Perkin Trans. 2, 1973, 696.
suggests that in the presence of water a more reactive oxidising
agent than nitrogen dioxide is formed. The possible agents may
include nitrous acid, nitrogen monoxide and nitrogen trioxide.
Whatever the reaction sequences may be, it is clear that only
rates of reaction of nitrosobenzenes with nitrogen dioxide
obtained under scrupulously dry conditions can be taken as
reliable rates for reaction (2).
Acknowledgements
We thank the Ministry of Defence for a research grant and Drs
L. Phillips and A. R. Osborn for helpful discussions. We thank
Dr J. A. Hunter for carrying out CNDO calculations of
p-substituted nitrosobenzenes.
27 M. D. Corbett and B. R. Corbett, Biochem. Pharmacol., 1986, 35,
3613.
28 S. Kazanis and R. A. McClelland, J. Am. Chem. Soc., 1992, 114,
3052.
29 J. Hamer, M. Ahmad and R. E. Holliday, J. Org. Chem., 1963, 28,
3034.
30 M. Ahmad and J. Hamer, J. Org. Chem., 1966, 31, 2829.
31 P. Schmid and K. U. Ingold, J. Am. Chem. Soc., 1978, 100, 2493.
32 S. Ursic, Helv. Chim. Acta, 1993, 76, 131.
33 B. M. Al-Tahou and B. G. Gowenlock, Recl. Trav. Chim. Pays-Bas,
1986, 105, 253.
34 J. A. Hunter, personal communication.
35 J. M. Dust and D. R. Arnold, J. Am. Chem. Soc., 1983, 105, 6531.
36 B. Milligan, J. Org. Chem., 1983, 48, 1495.
References
1 K. M. Ibne-Rasa, C. G. Lauro and J. O. Edwards, J. Am. Chem.
Soc., 1963, 85, 1165; K. M. Ibne-Rasa, J. O. Edwards, M. T. Kost
and A. R. Gallopo, Chem. Ind., 1974, 964.
2 T. G. Bonner and R. A. Hancock, J. Chem. Soc. B, 1970, 519.
3 Y. Ogata and T. Akada, Tetrahedron, 1972, 28, 15.
4 Y. Ogata and H. Tezuka, J. Am. Chem. Soc., 1967, 89, 5428.
5 E. Bamberger, Ber. Dtsch. Chem. Gesell., 1918, 51, 634.
6 P. J. Carmichael, B. G. Gowenlock and C. A. F. Johnson, J. Chem.
Soc., Perkin Trans. 2, 1972, 1379.
7 C. A. F. Johnson and C. M. Mechie, J. Chem. Res. ( S), 1983, 212.
8 G. H. Coleman, C. M. McCloskey and F. A. Stewart, Org. Synth.,
1945, 25, 80.
9 W. J. Mijs, S. E. Hoekstra, R. M. Ulmann and E. Havinga, Recl.
Trav. Chim. Pays-Bas., 1958, 77, 746.
Paper 7/00258K
Received 10th January 1997
Accepted 13th May 1997
10 R. E. Lutz and M. R. Lytton, J. Org. Chem., 1938, 2, 73.
11 E. Bamberger and A. Rising, Liebigs Ann. Chem., 1901, 316, 289.
12 Y. Ogata and K. Takagi, J. Am. Chem. Soc., 1958, 80, 3591.
1798
J. Chem. Soc., Perkin Trans. 2, 1997