The role of N-hydroxyphthalimide in the reaction mechanism of liquid-phase oxidation of p-cymene

Article abstract of DOI:10.1134/S0965544113030092

The reaction of the liquid-phase oxidation of p-cymene to hydroperoxide with a selectivity of up to 95% at a hydrocarbon conversion of 25-30% has been studied at a reaction temperature of 80-120 C in the presence of N-hydroxyphthalimide as the catalyst. The catalytic role of N-hydroxyphthalimide has been substantiated. A kinetic model of the liquid-phase oxidation reaction in the presence of the catalyst N-hydroxyphthalimide has been developed and adequately described.

Full text of DOI:10.1134/S0965544113030092

ISSN 0965ꢀ5441, Petroleum Chemistry, 2013, Vol. 53, No. 3, pp. 171–176. © Pleiades Publishing, Ltd., 2013.
Original Russian Text © V.N. Sapunov, G.N. Koshel’, Yu.B. Rumyantseva, E.A. Kurganova, N.D. Kukushkina, 2013, published in Neftekhimiya, 2013, Vol. 53, No. 3, pp. 193–198.
The Role of
NꢀHydroxyphthalimide in the Reaction Mechanism
of LiquidꢀPhase Oxidation of ꢀCymene
p
a
b
b
b
b
V. N. Sapunov , G. N. Koshel’ , Yu. B. Rumyantseva , E. A. Kurganova , and N. D. Kukushkina
a
Mendeleev University of Chemistry and Technology, Moscow, Russia
b
Yaroslavl State Technical University, Yaroslavl, Russia
eꢀmail: koshelgn@ystu.ru
Received December 10, 2012
Abstract—The reaction of the liquidꢀphase oxidation of
to 95% at a hydrocarbon conversion of 25–30% has been studied at a reaction temperature of 80–120°C in
the presence of ꢀhydroxyphthalimide as the catalyst. The catalytic role of ꢀhydroxyphthalimide has
been substantiated. A kinetic model of the liquidꢀphase oxidation reaction in the presence of the catalyst
ꢀhydroxyphthalimide has been developed and adequately described.
pꢀcymene to hydroperoxide with a selectivity of up
N
N
N
Keywords: liquidꢀphase oxidation,
pꢀcymene, Nꢀhydroxyphthalimide, pꢀcresol manufacture
DOI: 10.1134/S0965544113030092
Practical implementation of the “cymene” process (315–400 kJ/mol). It is supposed that the mechanism
for the production of
by a low selectivity for
particular, the difficulty of its concentration. Developꢀ
ment of methods for increasing the selectivity of oxidaꢀ
tion without catalysis by metal ions would form the real
basis for practical implementation of this process.
p
p
ꢀcresol and acetone is impeded of chain initiation by the «classical» initiator (In2
)
ꢀcymene hydroperoxide and, in
ki
In2
2In*.
(III)
can be duplicated as well according to the following
reaction:
N
ꢀHydroxyphthalimide (NHPI), a nonmetallic cataꢀ
kii
NHPI + O2
PINO* + HO2*.
(IIIa)
lyst for radical oxidation, has recently attracted the
attention of researchers [1–4], since it dramatically
increases the selectivity of processes, in particular, the
It is clear that in the absence of metal ions in the
reaction mixture, NHPI additives can facilitate not
only the selective production of hydroperoxides, but
also a more secure way of their concentration. In this
context, studying the kinetics and mechanism of the
liquidꢀphase oxidation of pꢀcymene is of great theoretꢀ
ical and practical importance.
p
ꢀcymene hydroperoxide preparation process [5]. The
aforementioned studies have shown that the principal
role of ꢀhydroxyphthalimide as a radical oxidation
N
catalyst reduces to duplication of the chain propagation
cycle according to Scheme I:
+
O2, k1
+RH, k2
In this study, we investigated the role of Nꢀhydroxꢀ
yphthalimide in various steps of the radical oxidation of
ꢀcymene.
R* ⎯⎯⎯ ⎯→ ROO* ⎯⎯⎯⎯→ROOH + R* (I)
In this scheme, initial NHPI undergoes a cycle of alterꢀ
nating transformations into the corresponding
p
N
ꢀoxyphthalimide radical (PINO*) and back (by
EXPERIMENTAL
ꢀcymene ( = 134.22 g/mol;
density = 0.8575 g/cm , i.e., 1 kg makes 1.166181 L;
molarity of the neat liquid is 6.463 mol/L) was carried
out with atmospheric oxygen in a continuousꢀflow
Scheme II):
Oxidation of
p
M
k21
(II)
3
NHPI + ROO* ⎯⎯ ⎯→ ROOH + PINO*
k22
PINO* + RH ⎯⎯ ⎯→ HNPI + R*
(IIa)
The new route of reactions (II, IIa) is energetically closed glass reactor with constant vigorous stirring in
more favorable than the traditional chain propagation the presence of the catalyst ꢀhydroxyphthalimide
pathway (I) [6]. Calculations show that the enthalpies of = 163.00 g/mol) taken in an amount of 2.7% of the
N
(M
reactions (I) and (II) are low, <20 and 80 kJ/mol, loaded hydrocarbon mass (0.1656 mol per kg of hydroꢀ
respectively [7]. Furthermore, a study of the thermodyꢀ carbon). The reaction was monitored by following the
namics of the nitroxide radical PINO* [8] showed that oxygen uptake; the oxidation was run in the kinetic
the O–H bond energy of NHPI is 375
± 10 kJ/mol, regime in which the reaction rate did not depend on the
which is comparable, e.g., with the C–H bond energy intensity of stirring. The oxidation product was anaꢀ
171

172
SAPUNOV et al.
Table 1. Temperature dependences of change in the current concentrations of hydroperoxide and
tion time
pꢀcymene with the reacꢀ
Temperature, °C
Time
8
0
90
100
110
120
80
90
ꢀCymene concentration, wt %
100 100 100
99.09 98.55 98.01
100
110
120
(min)
Hydroperoxide concentration, wt %
p
0
5
0
0
0
0
0
100
100
0.196
0.50
0.90
1.30
1.90
2.40
0.89
3.0
5.2
7.7
1.4
1.9
2.7
99.8
97.16
92
10
15
20
25
30
3.7
4.8
7.6
99.49
99.08
98.67
98.05
97.53
96.91
94.63
92.00
88.65
85.28
96.15
92.59
88.9
94.96
90.43
85.56
79.11
72.41
7.1
9.1
12.4
17.8
23.2
27.8
86.92
81.18
75.39
70.42
10.6
14.3
16.5
13.7
19.8
26.1
10.9
14.1
84.98
82.63
lyzed for
tion.
p
ꢀcymene hydroperoxide by iodometric titraꢀ where [ROOH]i is the current hydroperoxide concenꢀ
tration (wt %); is the proportionality factor
Table 2); and [RH]0 and [RH]i are the initial and curꢀ
rent concentrations of ꢀcymene (wt %), respectively.
α
0
(
p
RESULTS AND DISCUSSION
The directly proportional relationship between the
hydroperoxide yield and the hydrocarbon consumption
shows that there has been no noticeable degradation of
ROOH (at least within the accuracy of determination of
the reactant concentrations) during the test period. This
result might be an indication of constancy of oxidation
The oxidation of ꢀcymene in the presence of
p
N
ꢀhydroxyphthalimide was examined in the temperaꢀ
ture range of 80–120°C and until a hydrocarbon conꢀ
version of <30%. As shown in Table 1, the hydroperoxꢀ
ide concentration increases and the concentration of
ꢀcymene correspondingly decreases with the increasꢀ
ing temperature from 80 to 120°С. Under these condiꢀ
tions, ꢀcymene hydroperoxide can be produced with
selectivity (i.e., α ≡ S0) during the experiment. In other
0
p
words, the proportion of the hydrocarbon that was not
consumed for the formation of hydroperoxide (Table 2)
depends on the reaction temperature alone, but it
remains constant during the reaction and corresponds
to the consumption of RH for the formation of chain
p
high selectivity in as large an amount as 27.8% at 120°С
for 30 min.
Indeed, the confidence level (R2, determined by the termination products [Pr0]:
standard program Excel 93_2003) of linear correlation
ki
of experimental values for the concentration of
the forming hydroperoxide (wt %) and the correspondꢀ
2
ROO*
Pr0
(IV)
ing conversion of p
ꢀcymene (wt %) for all the temperaꢀ with maintaining the balance (2)
2
tures was very high (
R = 1), and the relationship
Δ
[RH] = [ROOH] + [Pr ] .
i i 0 i
(2)
obtained had the form:
However, the linear regression method used does not
[
ROOH] = α*([RH]0
⎯
[RH]i)
≡
α*
Δ
[RH] , (1) allow systematic deviations to be detected in quantiꢀ
i
0
0
i
ties to be matched unless these deviations are large in
2
absolute value. For example, the confidence level (
R
)
Table 2. Values for various parameters of the
p
ꢀcymene oxꢀ of a polynomial correlation of the same experimental
values of [ROOH]i and [RH]i was also very high
= 1). This relationship was as follows:
idation reaction, depending on the reaction temperature
Δ
2
(R
Equation (1)
Equations (3), (4)
Run temperaꢀ
ture,
[
ROOH] = (
i
α
–
α Δ[RH]i)
2
Δ
[RH]i.
(3)
°
C
1
α
α
α
0
1
2
An analysis of the coefficients (Table 2) for correlaꢀ
8
0
0.974
0.960
0.952
0.946
0.943
0.983
0.972
0.964
0.957
0.945
– 0.0046
– 0.001
tion equation (3) shows that the decrease in the integral
selectivity (Si, Eq. (4)) is defined by two terms, of which
one is independent of and the other is dependent on the
conversion of the hydrocarbon.
90
1
00
10
– 0.0008
– 0.0005
– 0.0005
1
120
Si
≡
[ROOH]/
Δ
[RH] = (α1
–
α2Δi).
(4)
PETROLEUM CHEMISTRY Vol. 53
No. 3
2013

THE ROLE OF
N
ꢀHYDROXYPHTHALIMIDE IN THE REACTION MECHANISM
173
[
Pr ] /
Σ
Δ
[RH] (%)
i
i
(a)
(b)
7.0
5
[
Pr ] /Δ[RH] (%)
i
Σ
i
6
.0
7
.0
.0
.0
.0
.0
.0
.0
5
4
4
3
6
5
4
3
2
1
5.0
3
2
4.0
3.0
2.0
1.0
2
1
1
0
0
20
40
Time (min)
10
20
[ROOH] (%)
30
i
Fig. 1. Change in the proportion of byproducts ([Pr ] ) relative to the amount of
p
ꢀcymene consumed in the reaction depending
Σ
i
on (a) time and (b) the amount of hydroperoxide formed at different temperatures: ( ) 80, ( ) 90, ( ) 100, ( ) 110, and ( ) 120°C.
1
2
3
4
5
It can be assumed that the term {
α
2 [RH]i} reflects relation to hydroperoxide buildup is illustrated in
Δ
the further consumption of the hydroperoxide via Figs. 1a and 1b, respectively.
reaction (V) at a rate t
r :
Analyzing the data presented in Table 1 and Figs. 1a
and 1b, we can come to the following conclusions:
k3
2
ROOH
Pr1
(rt),
(V)
(1) there is no induction period of the reaction; (2) the
intercepts on the ordinate give the extrapolative ratio of
the chain termination products ([Pr0]0) to the propagaꢀ
The data presented in Table 2 show that the change
in the hydroperoxide formation selectivity dependent
on the hydrocarbon conversion is very small, being only
a few percent. As the temperature increases, the selecꢀ
tivity decreases. However, a certain specific feature of
the process should be noted: the selectivity decrement
due to hydroperoxide degradation during the reaction
depends on the temperature in a manner other than the
usual, expected fashion. Evaluating the relative contriꢀ
bution of this hydrocarbon consumption route by a
tion products at a time
t
→
0, i.e., when the resulting
hydroperoxide is not involved in the process; and (3) the
proportion of the reaction products, increasing during
the process ([Pr ] – ([Pr ] ), decreases with the
Σ
i
0 0
increasing temperature and is much less than [Pr0]0 at
20 up to a 25–30% conversion of the hydrocarbon.
1
°С
Consider the kinetic treatment of the reaction of
interest, describing its “classical” propagation mechaꢀ
nism (Eqs. (6)–(9)). The notation of the rate constants
for different steps of the process has been given above. In
addition, the initiation rate (ri) responsible for the onset
numerical value of
is an order of magnitude greater than the contribution at
20
α
2, we see (Table 2) that at 80°С, it
1
°С.
Taking into account the high selectivity of the reacꢀ of the process is introduced into the scheme. This rate
tion and statistical nondistinguishability of correlation can be considered an infinitesimal, without which the
equations (2) and (3), for the further work we selected set of differential equations cannot be integrated. Typiꢀ
data on the formation of all byproducts ([Pr ]i), since cally, it is assumed for noncatalytic systems that the
Σ
they are the most informative.
chain initiation is determined by reaction (VI), called
the “thermal initiation”, with a rate ri
:
[
Pr ] = [Pr ] + [Pr ] ;
0
(5)
Σ
i
i
1 i
k0
where [Pr0] refers to the byproducts formed via the
chain termination reaction,
i
RH + O2
R* + HOO*
(ri)
(VI)
[
Pr1]i refers to the byproducts due to the decomposiꢀ
tion of hydroperoxide, and
Pr ] is the total amount of byproducts.
The set of differential rate equations is as follows:
d[ROOH]/dt = k [RH][ROO*] – k [ROOH] (6)
2
3
[
Σ
i
The share of total byproducts [Pr ]i in
p
ꢀcymene
Σ
2
d[Pr ]/dt = k ([ROO*]) = r + 2k [ROOH] (7)
(
[Pr ] /
Δ
[RH]i) consumed during the reaction and in
Σ
t
i
3
Σ
i
PETROLEUM CHEMISTRY Vol. 53
No. 3
2013

174
SAPUNOV et al.
The ratio of the byproducts to the hydroperoxide
formation rates depends much strongly on temperature
and is defined by Eq. (12) with a confidence level of
[
Pr ] (%)
Σ
i
2
.0
.8
.6
.4
.2
.0
2
R
= 0.942:
1
1
1
1
1
0
0
0
0
5
(
d[Pr ]/d[ROOH])
= ( r k /k [RH])
i t 2
0
t → 0
(12)
=
exp(13.15 – 4500/T) (wt %).
We compared this result with similar data on the oxiꢀ
dation of ꢀcymene reported by Makgwane et al. [9],
p
4
who used another initiator, a phoshovanadate (VPO)
complex. It turned out that the numerical values of the
product ratio in the initial period of the reaction are
almost identical to each other and are satisfactorily
described Eq. (12). The results show that the initiation
rate ri is not infinitesimal and is comparable with the
.8
.6
.4
.2
0
3
2
rate of initiation by
Nꢀhydroxyphthalimide (rii
≡
ri)
(Scheme III). The relatively small values of the temperꢀ
ature coefficients for the selectivity and product ratio
functions (Eqs. (12) and (13)) indirectly confirm that
the catalyst introduced (NHPI) might initiate the genꢀ
eration of radicals.
1
0
5
10
15
20
25
30
[
ROOH] (%)
i
However, the process characteristics measured at the
propagation step in this study are quite different from
those obtained in [9]. For example, calculations show
that in the case of initiation of the reaction by VPO, the
Fig. 2. Experimental data (symbols) and calculated curves
for the ratio of byproducts ([Pr ]i) and hydroperoxide
Σ
formed at different temperatures: (
1, ᭛) 80, (2, ᭹) 90, (3,
᭡
) 100, ( ) 110, and ( 120°С.
4,
ᮀ
5,
᭜
)
proportion of byproducts [Pr ] /
Δ
[RH]i at 120
30% increases from
25% (Fig. 8 in the paper cited, curve at 120
8) whereas it varies from
to 7% in our case (in Fig. 1b,
curve ). Even if the formation of [Pr ] is attributed to
°
С
and a
5% to
),
Σ
i
p
ꢀcymene conversion of
≈
≈
≈
°
С
–
d[RH]/dt
(
≈
5
=
k [RH][ROO*] + k [ROOH] + r .
2
3
i
5
Σ
the termination products only, it is evident that the
chain length of propagated oxidation in the presence of
NHPI is greater in comparison with one of the classical
initiators (VPO), i.e., metal ions.
Based on the condition that the initiation rate ri is
equal to the termination rate t, i.e., the rate of formaꢀ
tion of chain termination products (Eq. (7)), we obtain
an expression for the concentration of peroxide radiꢀ
cals:
r
In the case of participation of NHPI in the propagaꢀ
tion step (Scheme II), the rate of this step is defined by
[ROO*] = ri + 2k3[ROOH] / kt.
(9) Eq. (6a)
At the very beginning of the process with extrapolaꢀ
–
d[ROOH]/dt
(6a)
tion to a time of when the hydroperoxide has not
t
→
0
=
(k [RH] + k [NHPI])[ROO*] – k [ROOH].
2
21
3
been formed yet, the concentration of the peroxide radꢀ
ical is defined as
The appearance of the new term with a constant k21
in Eq. (6a) can be interpreted as an increase in the rate
constant of the reaction between the radical {ROO*}
(
[ROO*])t → 0
=
r / k .
(9a)
i
t
The selectivity of the reaction
conditions is expressed by Eq. (10):
(
S
)
i t
under these and the hydrocarbon. At a constant rate of the terminaꢀ
tion reaction, the effect of NHPI formally reduces to an
increase in the chain length, i.e., a decrease in the yield
→
0
_
(
Si)
≡ (d[ROOH]/ d[RH])
t → 0
t → 0
of undesired products as compared with that of hydropꢀ
eroxide (Eq. (13)).
k [RH]
k [RH]
(10)
2
2
=
ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ = ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ ꢀꢀ .
k [RH] + k [ROO*]
d[Pr ]/d[ROOH]
Σ
2
t
k [RH] + k r
i
2
t
(
13)
ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ ꢀꢀ ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ ꢀꢀ ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ ꢀꢀ .
(k [RH] + k [NHPI])[ROO*] – k [ROOH]
r + 2k [ROOH]
As can be seen in Figs. 1a and 1b, the initial selectivꢀ
weakly depends on temperature and, after
i
3
=
ity
(S )
i t
→
0
2
21
3
statistical treatment, is expressed by Eq. (11) with a
2
It may be assumed that at high temperatures
>100 ), there is a competition between the chain
propagation routes defined by Schemes I and II. Theoꢀ
retical and experimental studies [10, 11] showed that
confidence level of
R = 0.975:
(
°С
(S )
i t
= exp(4.22 + 135/T) (wt %),
(11)
→
0
where
Т in the reaction temperature in kelvins.
PETROLEUM CHEMISTRY Vol. 53
No. 3
2013

THE ROLE OF
N
ꢀHYDROXYPHTHALIMIDE IN THE REACTION MECHANISM
175
Table 3. Calculated values for the rate constants
ence of ꢀhydroxyphthalimide and their temperature dependence
K and k3 of the liquidꢀphase oxidation reaction of pꢀcymene in the presꢀ
N
Temperature,
°C
K
, L1/2 mol–1/2 s–1
k3, s–1
8
0
0
0.16 (0.035)*
0.42
0.5 (0.5)*
0.63
9
100
110
120
0.50 (0.054)*
0.60
0.75 (0.75)*
0.86
0.65 (0.071)*
1.02 (1.02)*
80–120
exp(7.3–3000/
T
)
R2 = 0.8
exp(6.14–2400/T)
R2 = 0.996
*
Values of the constants as given in [9].
reactions (II) and (IIa) are thermoneutral or slightly this paper determined the value of the constant
endothermic. As the temperature increases, the 120°С
К
at
as 0.093 L1 mol s . Somewhat later, the
replacement of the peroxide radical by the more stable Hattori model of isopropylbenzene oxidation was
ꢀoxyphthalimide radical (Scheme II) becomes faster refined by Bhattacharya [13]. The calculated value of
T =
/2
–1/2 –1
N
К
1/2
–1/2 –1
and, as a consequence, there is a relative decrease in the was 0.083 L mol s . Finally, Makgwane et al. [9],
increment of the byproduct yield. It is apparently this whose study has been discussed above, attempted to
development that is responsible for the unusual pattern apply the model developed by Hattori and Bhattacharya
of the curves in Fig. 1b; i.e., the contribution of reacꢀ to the oxidation reaction of
pꢀcymene in the presence of
tions (II) and (IIa) increases with the increasing temꢀ a vanadium catalyst. It turned out that the model agrees
perature as compared with the reactions of Scheme I.
well with the experimental data and adequately
describes the oxidation process. The calculated value of
Our experimental data presented above do not allow
for the complete kinetic analysis of the process and the
determination of the rate constants for each of the reacꢀ
1/2
–1/2 –1
K
was 0.071 L mol s , which is consistent with the
results obtained by Hattori and Bhattacharya. The
lower value of for ꢀcymene is explained in terms of a
somewhat lower rate of oxidation of ꢀcymene comꢀ
K
p
tions. However, to reveal the role of
imide (NHPI) in the liquidꢀphase oxidation reaction of
ꢀcymene, we performed a comparative analysis of the
description of the system by the set of Eqs. (6)–(8),
8a), and (9) presented above and the description of the
oxidation of isopropylbenzene by the mathematical
Nꢀhydroxyphthalꢀ
p
pared with isopropylbenzene. The rate constant k3
retains its numerical values in both the cases of isoproꢀ
p
pylbenzene and
pꢀcymene oxidation.
(
Based on the above considerations, we treated our
model originally developed by Hattori [12]. Except for experimental data in terms of the mathematical model
minor changes, the kinetics of the process in the work by Makgwane et al. [9]. At the initial step, we used the
cited is described by identical differential equations. For numerical values of
example, Hattori gives the hydroperoxide formation authors for 80, 100, and 120
K
and
k
obtained by the cited
3
°С. However, we managed
equation in the following form:
d[ROOH]/dt
K[RH] [ROOH] – k [ROOH].
to obtain a good description of the data using the least
squares technique, by minimizing the difference
between the experimental and calculated values for the
–
(14)
hydroperoxide and the undesired products with varying
=
3
2
the value of
К
. The confidence level (
R
) for the correꢀ
2
It is easy to see that the constant
К
in Eq. (14) is lation of the experimental and calculated data is R =
identical to the ratio of the constants of the set of differꢀ 0.976 for the entire set of experiments. The values for
ential equations (6)–(8):
the rate constants are given in Table 3.
It turned out that the numerical values of k3
remained almost unchanged, whereas the calculated
K ≅ k2 2k / k ,
(15)
3
t
values of
К were an order of magnitude higher than the
and takes the following form (Eq. (15a)) when the catꢀ
alyst (NHPI) is used:
value reported by Makgwane et al. [9]. A comparison of
the functional dependence of the constant defined by
Eqs. (15) and (15a) makes the cause of this discrepancy
obvious.
K ≅ (k + k [NHPI]/[RH]) 2k / k ,
(15a)
2
21
3
t
According to Hattori [12], Eq. (14) satisfactorily
The calculated data on the dependence of the yield
describes the isopropylbenzene oxidation kinetics in the of byproducts and the buildup of
temperature range of 80 to 120 . Based on the proꢀ oxide in the temperature range of 80 to 120
posed mathematical model of oxidation, the authors of are in good agreement with the experimental data.
p
ꢀcymene hydroperꢀ
°
С
°С
(Fig. 2)
PETROLEUM CHEMISTRY Vol. 53
No. 3
2013

176
SAPUNOV et al.
The results obtained in this study demonstrate the
adequacy of the kinetic model in question and the validꢀ
ity of the proposed reaction mechanism for the oxidaꢀ
tion of pꢀcymene to hydroperoxide in the presence of
ꢀhydroxyphthalimide.
4. Y. Ishii and S. Sakaguchi, Catal. Surveys Jpn.,
1999).
5. E. A. Kurganova, Yu. B. Rumyantseva, G. N. Koshel’,
et al., Khim. Promꢀst. Segodnya, No. 4, 20 (2012).
3, 27
(
N
6. K. Nobuyoshi, Cai Yang, and E. H. James, J. Phys.
Chem. A 107, 4262 (2003).
Thus, the analysis of these data leads to the concluꢀ
7
8
9
. I. Hermans, L. Vereecken, J. A. Pierre, and J. Peeters,
Chem. Commun., 1140 (2004).
. K. Nobuyoshi, Cai Yang, and H. J. Espenson, J. Phys.
Chem. A, 4262 (2003).
. P. R. Makgwane, N. I. Harmse, E. E. Ferg, and
sion that the driving force of the catalytic activity of
ꢀhydroxyphthalimide are reactions (II) and (IIa),
which result in an increase in the chain length of the
radical oxidation of ꢀcymene.
N
p
B. Zeelie, Chem. Eng. J. 162, 341 (2010).
REFERENCES
1
0. Gabriel da Silva and J. W. Bozzelli, J. Phys. Chem. 111
,
5760 (2007).
1
2
3
. O. Fukuda, S. Sakaguchi, and Y. Ishii, Adv. Synth. Catal.
43, 809 (2001).
. Y. Ishii, K. Nakayama, M. Takeno, et al., J. Org. Chem.
, 3934 (1995).
. Y. Yoshino, Y. Hayashi, T. Wham, et al., J. Org. Chem.
, 6810 (1997).
3
11. N. Koshino, Cai Yang, and J. H. Espenson, J. Phys.
Chem. A 107, 4262 (2003).
12. K. Hattori, Y. Tanaka, H. Suzuki, et al., J. Chem. Eng.
60
Jpn.
3, 72 (1970).
62
13. A. Bhattacharya, Chem. Eng. J. 137, P. 308 (2008).
PETROLEUM CHEMISTRY Vol. 53
No. 3
2013