Relationships between the efficiency of cyclohexane oxidation and the electrochemical parameters of the reaction solution

Article abstract of DOI:10.1016/j.molcata.2011.07.003

Vanadyl(IV)-acetylacetonate-catalyzed oxidation of cyclohexane with H ₂O₂, at 40 °C under air atmosphere, has been studied in the presence of small quantities of oxalic acid. The process efficiency is increased by this additive and depends on the nature of the solvent (MeCN ≥ MeOH > Me₂CO ≥ 2-PrOH > EtOH). The relationships between the results (conversion, yield) and the electrochemical characteristics of the reaction solution (relative permittivity, redox-potential) are highlighted and discussed.

Full text of DOI:10.1016/j.molcata.2011.07.003

Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Relationships between the efficiency of cyclohexane oxidation and the
electrochemical parameters of the reaction solution
Alexander Pokutsaa,∗, Orest Fliuntb, Yulia Kubaja, Tomasz Paczesˆniakc, Pawel Blonarzc,
Ruslan Prystanskiya, Jacques Muzartd, Roman Makitraa, Andriy Zaborovskyia, Andrzej Sobkowiakc
a Department of Physical Chemistry of Combustible Fossils, Institute of Physical Organic Chemistry and Chemistry of Coal NAS of Ukraine, Naukova Str., 3A, Lviv 79053, Ukraine
b Lviv National University, Dragomaniv Str., 52, Lviv 79000, Ukraine
c Rzeszow University of Technology, P.O. Box 85, 35-959 Rzeszow, Poland
d CNRS – Université de Reims Champagne-Ardenne, Institut de Chimie Moléculaire de Reims, UMR 6229, UFR des Sciences Exactes et Naturelles, BP 1039, 51687 Reims Cedex 2, France
a r t i c l e i n f o
a b s t r a c t
Vanadyl(IV)-acetylacetonate-catalyzed oxidation of cyclohexane with H2O2, at 40 ◦C under air atmo-
sphere, has been studied in the presence of small quantities of oxalic acid. The process efficiency
is increased by this additive and depends on the nature of the solvent (MeCN ≥ MeOH > Me2CO ≥ 2-
PrOH > EtOH). The relationships between the results (conversion, yield) and the electrochemical
characteristics of the reaction solution (relative permittivity, redox-potential) are highlighted and dis-
cussed.
Article history:
Received 18 March 2011
Received in revised form 30 June 2011
Accepted 5 July 2011
Available online 14 July 2011
Keywords:
© 2011 Elsevier B.V. All rights reserved.
Cyclohexane oxidation
Vanadyl(IV)-acetylacetonate
Oxalic acid
Electrochemistry
1. Introduction
ature data [7]. Nevertheless, in all previous reports, the kinetics
of oxidation was studied exclusively in connection with εr values
The oxidative functionalization of hydrocarbons is important
from both practical and fundamental points of view. Cyclohexane
oxidation is the basic process of the industrial-scale production of
cyclohexanol and cyclohexanone [1,2] – these compounds being
the precursors for Nylon-6 and Nylon-66 manufacturing [3,4]. The
commercial homogeneous Co(II)-based process, which occurs at
160–180 ◦C, 1.3–1.5 MPa of air, leads to low (4–6%) conversions and
moderate (75–80%) selectivities. Therefore, the elaboration of bet-
ter protocols remains a challenge and their achievement can afford
valuable information to carry out oxidative transformation of other
hydrocarbons.
Previously, we disclosed that the efficiency of C6H12 oxidation
into C6H11OH, C6H10O, and C6H11OOH, at 40 ◦C and 0.1 MPa, using
H2O2 as oxidant, VO(acac)2 (1) as catalyst, and MeCN as the sol-
vent is noticeably increased by addition of small amounts of glyoxal
[5] and oxalic acid (H2C2O4) (2) [6]. As appeared, the exchange
of MeCN by EtOH was detrimental to this process [6]. Therefore,
we assumed that this effect can be, among others, referred to the
electrochemical parameters of the reaction solution (e.g., relative
permittivity εr). Such an assumption is in agreement with liter-
inherent to pure solvents. The possible modification of εr and the
redox-potential of the reaction mixture by catalyst and additives
has not been studied. Using low-frequency (10–105 Hz) dielectric
spectroscopy (DS) [8,9] and cyclic voltammetry (CV) [10] tech-
niques, we have tried to estimate the impact of the catalyst and
additives on εr and the redox-potentials of the solutions, and to
determine plausible relationships between these parameters and
the reaction efficiency. These two methods have emerged as the
most powerful noninvasive electrochemical tools to investigate the
mechanism of oxidation processes [11,12]. The solvents for the oxi-
dation, CV, and DS experiments have been chosen according to their
various relative permittivity values as well as to their ability to be
inert under the experimental conditions and to dissolve reagents
and products. MeCN, MeOH, Me2CO, 2-PrOH, and EtOH agree with
these requirements.
2. Experimental
2.1. Chemicals
The commercial aqueous solutions of hydrogen peroxide
(35 wt.%, Fluka) have been concentrated by vacuum distillation at
45 ◦C/10 mm Hg to 70 wt.% (caution: risk of explosion. All glassware
has to be thoroughly rinsed by distilled water to remove traces
∗
Corresponding author. Tel.: +380 32 263 51 74; fax: +380 32 263 51 74.
E-mail address: apokutsa@ukr.net (A. Pokutsa).
1381-1169/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.07.003

16
A. Pokutsa et al. / Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
of any heavy metals). As the commercially available vanadyl(IV)-
acetylacetonate (Aldrich) is of ca. 95% purity only, the crude
VO(acac)2 was dissolved in MeCN to give a saturated solution,
which was then filtrated through a micro-porous paper filter. The
filtrate was evaporated at 40 ◦C under reduced pressure, and the
resulting solid was dried under vacuum at 20 ◦C for 48 h, lead-
ing to dark-green tiny (0.5–1 mm) crystals which were used in
the oxidation experiments. Cyclohexane, acetonitrile, methanol,
ethanol, 2-propanol, acetone, triphenylphosphane, and tetra-n-
butyl ammonium perchlorate (all from Aldrich) were used as
purchased.
vs SCE. The later was contained in a Pyrex tube with a cracked
soft-glass tip, which was placed inside a Luggin capillary [10].
Before each CV measurement, argon was bubbled through the reac-
tion solution for at least 30 min to remove dissolved oxygen. After
the experiment, the deposits, contaminating the active edge of
the working electrode, have been removed by polishing it with a
Gamma Micropolish II Alumina (0.05 ␮m, Buehler) paste, followed
by copiously rinsing with distilled water and then thorough dry-
ing. A microportion of H2O2 (usually 0.3 mmol) has been injected
into the mixture of reactants under vigorous stirring; that moment
was considered as the starting point for current–voltage depended
function measurements.
2.2. Experimental installation and oxidation reactions
2.3. Analysis of the reaction products
The catalytic experiments were carried out in 25-mL round-
bottom glass flasks equipped with reflux condenser and magnetic
stirrer. In a typical experiment, 3.36 g (40 mmol, 1.8 mol L−1) of
cyclohexane were dissolved in 15 mL of MeCN containing 0.0032 g
(0.012 mmol, 5 × 10−4 mol L−1) VO(acac)2 and 0.038 g (0.3 mmol,
0.014 mol L−1) oxalic acid. The mixture was heated at 40 ◦C for
5 min before 1.34 g (40 mmol, 1.8 mol L−1) of H2O2 were added. This
moment corresponds to the beginning of the oxidation reaction.
Usually, the reaction has been carried out for 5 h under atmospheric
pressure.
The reaction mixture was sampled at regular intervals. The
content of products in the samples was analyzed using a GLC
Hewlett–Packard 5890, series II (flame-ionization detector, cap-
illary column 30 m × 0.25 mm, immobile phase HP-Innowax) as
well as a LKhM-80 one (flame-ionization detector, packed column
3 m × 3 mm, immobile phase OV-17 on Chromaton N-AW). Prior
to the analyses, a 2-butylbenzene solution in ethanol was added
as internal standard to the withdrawn sample. The identification
of the oxidation products was performed by comparison with the
retention times of commercial cyclohexanol and cyclohexanone.
Subsequently, both the conversion and the selectivity towards
the obtained products were determined by the common calcula-
tion procedure. The quantification of the oxygenated products was
performed using a multi-points calibration line plotted using ref-
erences of C6H11OH and C6H10O. In order to determine the proper
content of cyclohexanol, cyclohexanone, and cyclohexyl hydroper-
oxide, each probe was analyzed twice by GC – before and after
the addition of triphenylphosphane. In the presence of PPh3, the
C6H11OOH is rapidly and quantitatively transformed into C6H11OH;
then the true content of cyclohexylhydroperoxide can be calculated
as the difference between the C6H10O and C6H11OH concentration
before and after the triphenylphosphane addition [13]. Titration of
probes of the reaction mixture by an aqueous 0.10 mol L−1 solution
of Na2S2O3 in presence of KI allows to determine the amount of
H2O2 which has been consumed.
The values of relative permittivities (εr) of the reaction solution
have been obtained by dielectric spectroscopy (DS) measurements
[8]. The dielectric response of any medium is characterized by
the spectral shape of the real (in-phase) ε1(ω) and imaginary
(quadrature) ε2(ω) components of the dielectric complex constant
ε*(ω) = ε1(ω)–jε2(ω) or by the corresponding complex capacitance
C*(ω) = C1(ω) − jC2(ω), where j is the imaginary unit and ω is the
angular frequency; C1(ω) and C2(ω) are the real and imaginary
components of the complex capacitance, respectively. Complex
capacity and complex εr are connected via the constant factor S/d
by the equation: ε∗r = C ∗ /d(ε0S), where S is the area of contacts of
the plane capacitor, m2; d the thickness of polarization region, m;
ε0 the relative permittivity of vacuum (8.85 × 10−12 F m−1) [8,9]. In
some cases the use of complex capacity instead of complex dielec-
tric constant have some advantages, particularly, when electrode or
double layer polarization effects dominate on the spectra. Dielectric
spectra in the frequency range of 10–105 Hz have been measured
by transducing the complex capacity into sinusoidal voltage on the
basis of operational amplifier [8]. This method eliminates the influ-
ence of stray capacitors of connecting cables. Sinusoidal voltage
100 mV rms was applied to the measuring cell. Phase shifts were
measured using a digital phase-meter. Calibration of the equipment
using etalon elements permits to exclude the constant bias errors.
A 10-mL glass vial equipped with two parallel stainless steel elec-
trodes (1 mm × 10 mm × 20 mm) with fixed 2 mm inter-electrode
distance has been used as the measuring cell. Usually, the total
volume of reaction mixture was 8 mL. All experiments have been
conducted at room temperature and atmospheric pressure. After
each measurement, the electrodes have been copiously rinsed by
distilled water and then thoroughly dried.
3. Results and discussion
Oxidation of C6H12 has been carried out using from 1 to 10 equiv-
alents of H2O2, a catalytic amount of VO(acac)2 (1), and various
quantities of oxalic acid (2). The main results are collected in Table 1
(for all experiments, the process selectivity calculated on the sum
of aimed products was 97–99%). Accordingly to the results shown
in Table 1, the addition of 2 led to a noticeable improvement of the
process. For example, in MeCN, the presence of 2 increases both the
conversion and TON (compare entries 1 and 2), the optimal amount
of 2 being between 0.011 and 0.022 mol L−1 (Fig. 1). Exceeding this
amount drops the alkane conversion.
Cyclic voltammetry (CV) measurements of the redox poten-
tial of the reaction mixtures have been carried out at room
temperature using an EG & G Princeton Applied Research poten-
tiostat/galvanostat model 273A. The experiments were conducted
in a 15-mL glass electrochemical cell containing 10 mL (total vol-
ume) of reagents. In order to provide proper electrolytic conditions,
the voltammograms have been taken in the presence of tetra-n-
butyl ammonium perchlorate (0.1 mol L−1). The working electrode
was a Bioanalytical Systems glassy-carbon (area, 0.09 cm2) inlay,
the auxiliary electrode a platinum wire, and the reference elec-
trode an Ag/AgCl/Cl− wire, in an aqueous tetramethylammonium
chloride solution that was adjusted to give a potential of 0.00 V
Studies on the positive influence of pyrazine-2-carboxylic
[14,15], sulfuric or oxalic acid [16] as a co-catalyst on the V2O5
and NaVO3-catalyzed oxidations of alkanes led us to suspect that
oxalic acid acts also with the reduced form of catalyst (V4+), this
plausibly leading to the coordination of its monoanion to the metal
cation. Diminution of the oxidation efficiency caused by an excess
of 2 can be attributed to the acceleration of the heterolytic cleav-
age of H2O2 at very low pH values [17] (pH equal −2 was reached
in the case of amounts of 2 higher than 0.010 mol L−1). The alkane
conversion increases to 30 % in the presence of 2 when the C6H12
concentration was decreased to 0.18 mol L−1 (entry 4). According
to these data, the most appropriate medium for C6H12 oxidation is

A. Pokutsa et al. / Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
17
Table 1
Oxidation of C6H12 in five solvents in the presence of VO(acac)2 and oxalic acid.a
Entry
c (2), mol L−1
Solvent
Conversion, %
Selectivity (product spread), %b
TON, mol mol−1 (1)
ꢀH2O2c, %
Ecd, mol%
C6H11OH
C6H10
O
C6H11OOH
1
2
–
MeCN
MeCN
MeCN
MeCN
MeOH
MeOH
2-PrOH
2-PrOH
EtOH
1.6
5.3
5.2
30
1.6
3.6
1.1
2.2
0.1
0.7
0.2
2.5
33
48
47
22
65
83
43
52
11
33
57
17
62
7
5
45
43
34
15
<1
28
17
28
29
35
69
33
169
48
96
80
34
21
66
27
43
2
3.3
5.5
6.5
8.8
7.6
5.5
4.1
5.1
5
0.014
0.14
0.014
–
0.014
–
0.014
–
0.014
–
3
10
44
20
17
29
31
61
38
8
173
944
52
120
35
72
3
22
6
85
4e
5
6
7
8
9
10
11
12
EtOH
Me2CO
Me2CO
10
78
84
7
0.3
3.0
0.014
14
a
c(C6H12) = c(H2O2) = 1.8 mol L−1, c (1) = 5 × 10−4 mol L−1, total reaction volume 22 mL, 40 ◦C, 5 h.
b
So far the process selectivity was ca. 99 % the %yield of each products was equal it selectivity and the total yield (in mol L−1) can be estimated as conversion of
C6H12 × c(C6H12)/100.
c
Amount of consumed H2O2 (i.e., conversion of H2O2).
Effectiveness (Ec) of the H2O2 consumption: Ec = c(C6H12converted)/{c(C6H12initial) × ꢀH2O2}.
d
e
c(C6H12) = 0.18 mol L−1, c(H2O2) = 1.8 mol L−1, c (1) = 5 × 10−5 mol L−1, total reaction volume 22 mL, 40 ◦C, 5 h.
acetonitrile whereas the worst one is EtOH. The average difference
in the conversion of cyclohexane and TON between these two sol-
vents was more than 10 times. The solvent-dependent oxidation
efficiency, including the effectiveness of H2O2 consumption (Ec,
Table 1) was in the following order: MeCN ≥ MeOH > Me2CO ≥ 2-
PrOH > EtOH, regardless of the presence or absence of the additive.
The additive, besides its influence on the hydrocarbon conver-
sion and yield, also modify the proportion of the products. For
example, the C6H11OH/C6H10O/C6H11OOH ratio in MeCN varies
from 6.6/12.4/1.0 (entry 1) in the absence of an additive to
1.0/0.2/1.0 when 2 is added (entry 2). In contrast to acetonitrile, 2-
propanol, ethanol, and acetone, the amount of C6H11OOH formed in
MeOH was very low (Table 1, entry 6). The relative yield of C6H11OH
and C6H10O can also be modified by the nature of the solvent and
the presence of 2 (Table 1). Compared to EtOH, the better results
in MeOH are a consequence of the well-known free-radicals termi-
nation by the ethanol molecules [18]. Such feature of the oxidation
in EtOH can refer to its noticeably lower value of εr (24.5) and its
higher donor number (32.0) in comparison to those of MeOH (32.7
and 19.0, respectively) [19]. That trait in conjunction with the abil-
ity of alcohols to form strong intermolecular hydrogen bonds (as
in water which is another powerful quencher of the free-radical
species) leads to the better solvation of free-valent species (see part
devoted of the DS studies below), hence lowers the generation of
exclusively radical-dependent products (C6H11OOH, Table 1).
It is known that the process rate is linked by some manner to the
solvent’s relative permittivity, εr [7,12,19]. Nevertheless, a possibly
existing relationship between εr of the reaction mixture and the
main process characteristics has not been examined yet. Accord-
ing to the data of Table 1, both conversion and yield depend on
solvent, catalyst, and additive. It was logical to assume that cata-
lyst and additive can modify εr of the reaction medium, and thus
affect the whole process effectiveness. In order to elucidate this
assumption, low-frequency dielectric spectra of the reaction mix-
tures used have been measured and four typical ones are presented
at Fig. 2. The values of εr have been calculated from the curved
lines (the real part of spectra) whereas the electric conductivity
of the reaction medium can be obtained from the straight lines
(imaginary part of spectra). According to Fig. 2, the dependences of
C1(ω) vs frequency show a strong dispersion towards the low fre-
quencies region, which is related to electrode polarization. In the
high-frequency region, the dispersion of the real part of capacity
was not observed; this allows to calculate the relative permittivity
constant of the liquid using plane capacitor formula. The imagi-
nary part of capacity C2(ω) decreases with increasing frequency
accordingly to the 1/ω law. Such relationships give an opportunity
to characterize the solutions by the values of non-dispersive con-
ductivities. Given the data of Table 1, the oxidation in methanol
and 2-propanol occurs in yields which are only slightly lower than
Fig. 2. Low-frequency dielectric spectra of 1 (a) and 1 + 2 (b) solutions in MeCN at
20 ◦C. Solid and open symbols correspond to the real (C1) and imaginary (C2) parts
of capacity. Concentration of catalyst and activator was the same as in the oxidation
experiments (see Table 1, footnotes).
Fig. 1. Cyclohexane conversion (1) and process TON (2) depending
on the oxalic acid content. Oxidation conditions: c(C6H12) = 1.8 mol L−1
c(VO(acac)2) = 5 × 10−4 mol L−1, c(H2O2) = 1.8 mol L−1, in MeCN at 40 ◦C, 5 h.
,

18
A. Pokutsa et al. / Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
Despite the observation that relative alteration of the εr value
caused by addition of 2 was quite negligible (1.1–1.2 times), this
addition sets off a conspicuous increase (in 2.5–7 times) of the con-
version and yields in all solvents (Table 1 and Fig. 3). This finding can
be related to the properties of particular cations and anions stud-
ied in solution. The matter is that the solvation of VO(acac)2 and
H2C2O4 is favored by the relative permittivity of the solvent and
the charge on the solute species in accordance to the Born model
[21] (albeit rather in the first approximation due to the data [22]).
On the other hand, in agreement with the HSAB concept [23], har−d
VO2+ cations coordinate better to the hard (and smaller) HC2O4
anion than to the soft (and bulky) acac− anion. The dissociation of
VO(acac)2 is facilitated in polar solvents (MeCN, Me2CO, MeOH).
For solvents possessing a high relative permittivity and a relatively
small DN (MeCN, MeOH), we assume that the metal cation (VO2+
)
interacts preferably with HC2O4 than with acac− [20]. Hence,
the rate of ligands rearrangement should be determined by the
medium polarity as postulated in a number of reports [24–26].
The putative adduct of the first-stage interaction between H2O2
and VO(HC2O4)2 ([VO(HC2O4)2. . .H2O2]) possesses a higher den-
sity of the positive charge than [VO(acac)2. . .H2O2] (see the results
of calculation below). Hence, this primary adduct would be better
solvated by polar solvents in accordance to the general concepts:
if the charge density of an activated complex increases, the corre-
sponding process rate would be enhanced with increasing solvent
polarity because of the lowering of the activation energy of the reac-
tion [18,27]. Given the quantum-chemical calculations (PM6) [28],
a noticeable benefit in energy was developed when the two acac lig-
−
Fig. 3. The total yield of products vs Kirkwood parameter in dependence of compo-
nents presented in the started reaction mixture: squares – in presence of 1, triangles
– in presence of 1 + 2. (a) MeCN, (b) MeOH, (c) Me2CO, (d) 2-PrOH, (e) EtOH. Con-
centration of catalyst and activator was the same as in the oxidation experiments.
in MeCN, whereas the process ceased almost completely in EtOH.
The effect has been asserted to the values of the Kirkwood parame-
ter (εr − 1)/(2εr + 1) of the reaction medium, which is linked to the
oxidation products yields by the linear law shown in Fig. 3. The
addition of 2 caused the slight increase in εr of the solutions of 1
in MeCN (from 42.9 to 47.3) and MeOH (from 41.6 to 45.8). Con-
trary to these solvents, adding of 2 into EtOH or 2-PrOH decreases
the relative permittivity from 32.1 to 26.3 and 34.7 to 27.3, respec-
tively, whereas in acetone this value remains almost constant and
is equal to 34.2 (when 2 is present in the reaction mixture the
difference in relative permittivity between MeOH and EtOH solu-
tions grows from about 30 % for the 1-based process up to 74 %
for the 1 + 2 one). The linearity displayed by these dependencies,
regardless of the absence or presence of 2 (Fig. 3), can be referred
to the predominance of the electrostatic solute/solvent interaction
(non-specific solvation) over the electron-pair donor/electron-pair
acceptor (EPD/EPA) interaction (specific solvation). In general, their
ratios are considered to about 4:1 [20]. The activated complexes
formed in the absence and presence of 2 have rather similar dipole
moments because of the same slope of the lines in Fig. 3 [18]. Some
deviation from linearity, particularly observable when 2 is present
in the reaction mixture (Fig. 3), can be a result of specific (non-
electrostatic) solvation of the reactive species [18]. The observation
that the oxidation is more efficient in MeCN than in MeOH and is
extremely slow in EtOH, can also be due to other characteristics of
the solvents used: (i) their donor numbers (DN) which are equal
to 14.1 for MeCN, 19.0 for MeOH, and 32.0 for EtOH, and (ii) their
dipole moments which are actually proportional to εr and dropped
from ꢁ = 3.53 D (MeCN) to 2.87 D (MeOH) and 1.66 D (EtOH) [20]. It
is found that the efficiency of C6H12 oxidation is only slightly lower
in MeOH as compared to MeCN (Fig. 3) which are characterized by
close numbers of both DN and dipole moment, but it is dramati-
cally diminished in EtOH which possesses the highest DN and the
lowest dipole moment. According to the literature [19,20], solvents
with a relatively small DN (MeCN, MeOH) solvate metal ions rather
weakly particularly those in their higher oxidation state. Hence, in
such solvents, the Mn+1 can be easily reduced to Mn+, making Mn+1
a strong oxidizing agent. The experimental data reported in Fig. 3
support these assumptions and provide a good reason to subject the
yield to εr of the reaction solution. Nevertheless, the dependencies
of C6H12 oxidation may be related, besides the parameters of non-
specific solvation, also to DN. Furthermore, the contribution of the
solvent’s electrophilicity (proportional Lewis acidity), basicity (pro-
portional Lewis basicity), and the intermolecular hydrogen bond
donor/acceptor capability (specific solvation’s parameters) cannot
be completely removed from our consideration.
ands of VO(acac)2 were substituted by two HC2O4 (ꢀH◦ decreased
f
from −239.25 kcal mol−1 for VO(acac)2 to −442.26 kcal mol−1 for
VO(HC2O4)2) [29]. Simultaneously, such a ligands exchange lowers
the charge density on the metal from +1.08 eV to +1.22 eV [29]. The
easier coordination of H2O2 by VO(HC2O4)2 can lead to both quick
generation of hydroxyl radicals and formation of alternative (to
the HO•) high-valent vanadium-oxo species active and selective in
hydrocarbon oxidation [6,13,14]. In the last case H2O2 is consumed
much more sufficiently (Table 1) as compared to the homolytic way
of its decomposition (more details of this assumption are reviewed
in the CV part below). Therefore, the acceleration of the oxidation
process in the presence of 2 would be a result of two factors –
the medium polarity and the modification of the electron−density
of the metallocomplex cation caused by the acac−/HC2O4 ligand
exchange. The exclusive assignment of the effect of the acid to the
transformation of the vanadate anion into oligomeric oxovanadates
as postulated in [16] looks questionable. Indeed, the reduction of
the oligomeric vanadium(V) ion into vanadium(IV) by hydrogen
peroxide as it was assumed in [16] is rather difficult because of the
high oxidizing and low reducing abilities of H2O2. In contrast, a rea-
sonable explanation is plausible if we take into consideration the
particular reducing properties of oxalic acid (see also the CV part
below).
Changes in the charge of these ionic solutes in the reaction mix-
ture occurring in the course of the redox process affect the value
of electric current flows between the electrodes when the applied
potential is varied. Application of the CV technique for the gen-
erated current registration can supply valuable information about
the concentration of the redox species and their modification in the
course of the chemical process. Therefore, some important conclu-
sions concerning the process kinetics and the reactions mechanism
can be made. In Figs. 4–6 and Figs. S1–S3 the voltammograms of 1
and 1 + 2 in MeCN are presented which have been taken before and
after the H2O2 addition. The forward anodic scan of the initial light-
blue solution of VO(acac)2 (Fig. 4, curve 1) displayed the oxidation
peak at 0.98 V, which is coupled with the reduction one appearing
at 0.92 V in the reverse scan. The second reduction peak emerges
at −1.84 V and is irreversible (Fig. 4, curve 1). During the second

A. Pokutsa et al. / Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
19
cycle of this initial positive scan four oxidation peaks are appeared
at −1.45, −0.74, 0.30, and 0.72 V, respectively (Fig. 4, curve 1). The
forward catodic scan of the same solution (Fig. S1, curve 1) resulted
in one irreversible reduction peak at −1.84 V. Four oxidation peaks
appear at the same places (−1.45, −0.74, 0.30, and 0.72 V) in anodic
scanning when the second cycle of this catodic scan was draw-
ing. Simultaneously, a reduction peak at 0.66 V, reversible to the
one appeared at 0.72 V, becomes visible in the second cycle of this
negative scan too. Due to the data reported for similar systems by
Kitamura and Reichel [30–32], the distinct reversible couple at 0.98
and 0.92 V (Fig. 4 and Fig. 1S, curve 1) exhibits without doubt the
VIV ꢀ VV process, while the couple at 0.72 and 0.66 V (Fig. S1, curve
1) can be referred to the VIV ꢀ VIII one. A corroboration that the last
redox reaction can take place is that the ions in their lower oxi-
dation state (VIII) are oxidized (and respectively reduced) at a less
positive potential [31] (a discussion of the last assumption proceeds
below). These facts also point out that the redox reactions studied
involve in fact a one electron exchange. The peaks at −1.84 V and
0.30 V presented in the voltammograms 1 result from both anodic
(Fig. 4) and catodic (Fig. S1) scans and can be attributed to the redox
reaction of the acac ligands, respectively (Fig. S2, inset) [31,33].
The addition of H2O2 into the initial solution of VO(acac)2 causes
a dramatic and rapid increase of the peak situated near −1.0 V
(Fig. 4 and Fig. S1, curves 2). Nevertheless, in 5 min this peak is
reduced (Fig. 4, curves 2 and 3) back to the value found before the
H2O2 addition with almost the same rate as it was growing (Fig. 5,
curve 1). Saturation of the VO(acac)2 solution by O2 resulted in
an extremely high peak appearance which was observed at about
−1.0 V (Fig. 4, inset). Therefore, the discussed peak with a maximum
at −1.0 V (Fig. 4 and Fig. S1, curves 2) can be assigned to the dioxy-
gen released in the course of the fast H2O2 decomposition catalyzed
by VO(acac)2. Furthermore, the last mentioned reaction leads to a
quick transformation of the parental catalyst ions (VIV) into their
oxidized form (VV) that substantiates the noteworthy diminishing
of the oxidation peak at 0.98 V (Fig. 4, curve 2). The surplus wide
oxidation peak appeared in the voltammogram near 0.5 V as well as
the broad reduced peak at about −1.7 V (Fig. 4, curve 2) are inherent
to H2O2 (Fig. S2, curve 1).
Fig. 4. CV diagram of the 1 × 10−3 mol L−1 VO(acac)2 solution in MeCN (1) and the
same solution drawn, respectively, at 2 (2) and 3 (3) min after of 0.3 mmol H2O2
injection (anodic scan). Inset: CV of 1 × 10−3 mol L−1 VO(acac)2 solution preliminary
saturated by O2.
The profile of CV for the 1 + 2 mixture sufficiently differs from
the one depicting solely VO(acac)2 (compare curves 1 and 2 in Fig. 6
and Fig. S3). First of all, the peak was assigned to the oxidation of
VIV (0.98 V), steadily diminished when the H2C2O4 concentration
increases from 0 via 0.2 to 1.0 mmol–L−1 (Fig. 6, inset). Continua-
tion of the second anodic (Fig. 6) and catodic (Fig. S3) scans results
in some broad heel extended from 0.3 to 1.2 V (curves 2) which
eventually hides the peak at 1.0 V. The height of this last peak is
lower almost twice as much compared to the one presented in
curves 1. It means that vanadium ions in oxidation state IV are
not attended in the mixture any more if the amount of oxalic acid
added exceeds one equivalent of VO(acac)2. Secondly, the peak pre-
viously attributed to the oxidation of VIII (0.72 V), in the presence
of oxalic acid noticeably increased (compare curve 1 and 2, Fig. 6).
The last marked changes as well as the fact of almost complete
disappearance of the oxidation peak at −0.98 V in the presence of
oxalic acid (Fig. 6, curve 2) confirm the possibility of the chemi-
cal reduction of initial VIV species towards VIII by 2. Interestingly,
the addition of H2O2 into the solution of 1 + 2 causes a rather neg-
ligible changes (compared to the oxalic acid-free process) in the
peak attributed to released O2 (near −1.0 V) as well as to the one
at −1.7 V assigned to free H2O2. The kinetics of the −1.0 V peak
height changes (Fig. 5) depicts clear relationships between the pres-
ence of oxalic acid and the concentration of O2 in the solution (i.e.,
rate of the H2O2 cleavage). In the absence of 2, the sharp enhanc-
ing of the dioxygen content in the solution followed by its abrupt
decrease was observed almost instantly after the H2O2 injection
(Fig. 5, curve 1). Contrary to the previous case, the presence of about
Fig. 5. The kinetic of the −1.0 V peak height changes on CV of the 1 + 2 solution after
addition of 0.3 mmol H2O2 in dependence of the oxalic acid content: squares – 0 (1),
triangles – 1.0 (2), circles – 5 × 10−3 mol L−1 (3) of 2; c(VO(acac)2) = 1 × 10−3 mol L−1
.
Fig. 6. CV diagram of the 1 × 10−3 mol L−1 solution of VO(acac)2 in MeCN (1),
1 × 10−3 mol L−1 VO(acac)2 + 5 × 10−3 mol L−1 oxalic acid (2), and the last solution
drawn after 5 min of 0.3 mmol H2O2 injection (3), (anodic scan). Inset: CV of the VIV
peak shape vs content of oxalic acid: 1 – 0, 2 – 0.2, 3 – 1.0 × 10−3 mol L−1 concentra-
tion of 2.

20
A. Pokutsa et al. / Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
1.0 mmol L−1 oxalic acid leads to the much more gentle slope of the
curve characterizing the kinetics of O2 (curve 2). Increasing the con-
centration of 2 up to 5 mmol L−1 causes a further flattening of the
curvature of the respective line (compare curve 2 and 3).
The revealed differences in the behavior of the electrochemical
processes in the absence and presence of 2 and H2O2 help to get a
proper reply concerning the mechanism of the selective cyclohex-
ane oxidation enhanced by oxalic acid. Based on the experimental
data we are now able to conclude that: (i) so far the acid-free pro-
cess is characterized by the rapid O2 release, one can assume that it
can be the result of quick homo- and heterolytic decomposition of
the peroxide. The homolysis of H2O2 via the Haber–Weiss mecha-
nism resulted in free-radicals (mainly HO•). The extremely active
(although non-selective) hydroxyl-radicals react with any sur-
rounded molecule, that is substrate, solvent and H2O2. Therefore,
in the case of radical-based oxidation, the percentage of productive
consumption of hydrogen peroxide should be lower compared to
the non-radical (enzyme-like) mechanism. The last mechanism is
presumably favored under 2 assistance [6] since its presence in the
VO(acac)2-based reaction improves the process, i.e., yield, TON, and
Ec (Table 1). Due to the kinetics, electrochemical, and quantum-
chemical calculation [29] data, the activating ability of 2 can be
consisted in the efficient chelation of the transition-metal cations
by 2, this leading to withdraw the charge density from the metal
core of the catalyst. The consequences could be an easier access
of peroxide molecules to the catalyst, and thus an increase of the
rate of homolytic decomposition of H2O2 is accelerated [29]. On the
other hand, the coordination of 2 to the catalyst cation can simulta-
neously facilitate the formation of the high-valent vanadium-oxo
species as it is proposed in [6,34]; (2) the reducing properties of
oxalic acid prompt the reactions like Mn+1 + 2 → Mn+ + CO2 [35], par-
Scheme 1.
ticularly if the redox potential of metal cations (VV/VIV = 1.0 V) is
lower than that of H2O2 (1.77 V) [36]. The electrochemical data by
preferable reduction of VIV over its oxidation in the presence of 2 are
in good agreement with the above-mentioned quantum-chemical
calculations and also match well with previously published results
[6]. In summary, one can conclude that the redox properties of
Scheme 2.

A. Pokutsa et al. / Journal of Molecular Catalysis A: Chemical 347 (2011) 15–21
21
the catalyst can be modulated by the activator used. This fact is
Appendix A. Supplementary data
imprinted in the electrochemical characteristics of the reaction
mixture which are affected by the presence of oxalic acid and have
a distinct impact on the whole catalytic oxidation process.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcata.2011.07.003.
The key putative steps accompanying both the stoichiometri-
cal and electrochemical reactions of reduction and oxidation of
VO(acac)2 in the presence of oxalic acid [35] are presented in
Scheme 1. Reaction 1 by which free •CO2H radicals are formed,
as well as the electrochemicalReaction (8) can be considered as
those which determined the rate of the process. Reaction (2) is
responsible among others for the quick growing of the O2-peak
on the CV diagram (Fig. 4, curve 2) after the addition of H2O2
into the VO(acac)2 solution. The formyl-oxo radicals are the reduc-
ing agent for the oxidized form of metal cation (Reaction (3)) =/(in
reactions (5) and (8) the intermediated species are marked by ).
Vanadyl(IV)oxalate formed in situ from vanadyl(IV)acetylacetonate
and oxalic acid (reaction (4)) can lead to the oxo- (D) and peroxo-
(F) species due to the putative Scheme 2 [6]. Such species are
presumably responsible for the acceleration of the radical and non-
radical steps of the oxidation process in presence of oxalic acid. In
accordance to the Scheme 2, the process would occur via a rad-
ical pathway represented by A → B → C → D → E → A [14,34,37],
and a non-radical one illustrated by A → B → C → F → E → A. The
F → E transformation would involve interactions of cyclohexane at
the second coordination sphere of F. Under our oxidation condi-
tions, cyclohexanol, cyclohexanone and cyclohexyl hydroperoxide
are generated simultaneously, with different rates, from the begin-
ning of the oxidation process. The clear co-catalytic nature of 2 in
the process (besides of it reduction properties) can be displayed by
the next simple calculation. As follows from the Table 1 (entry 2)
the total amount of products formed from C6H12 can be estimated
as: TON × c(1) = 169 × 0.0005 = 0.0845 mol L−1, whereas the initial
amount of 2 was noticeably lower (0.014 mol L−1). The co-catalytic
function of oxalic acid is also depicted in Scheme 2.
References
[1] Y. Ishii, S. Sakaguchi, T. Iwahama, Adv. Synth. Catal. 343 (2001) 393–427.
[2] U.F. Kragten, H.A.C. Baur: EP Pat. 659,726 (1995);
U.F. Kragten, H.A.C. Baur, Chem. Abstr. 123 (1995) P86595r.
[3] A.P. Tomilov, S.K. Smirnov, Adiponitrile and Hexamethylenediamine , Khimiya,
Moskwa, 1974 (in Russian).
[4] M. Eissen, J.O. Metzger, E. Schmidt, U. Schneidewind, Angew. Chem. 114 (2002)
402–425;
M. Eissen, J.O. Metzger, E. Schmidt, U. Schneidewind, Angew. Chem. Int. Ed. 41
(2002) 414–436.
[5] A. Pokutsa, J. Le Bras, J. Muzart, Kinet. Catal. 48 (2007) 26–31.
[6] A. Pokutsa, Yu. Kubaj, A. Zaborovskyi, D. Maksym, J. Muzart, A. Sobkowiak, Appl.
Catal. A: Gen. 390 (2010) 190–194.
[7] R.G. Makitra, R.E. Prystanskii, R.B. Sheparovich, Russ. J. Phys. Chem. 78 (2004)
522–525.
[8] E. Barsoukov, J.R. Macdonald, Impedance Spectroscopy Theory, Experiment,
and Applications , second ed., Wiley, Hoboken, 2005.
[9] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and Applica-
tions , second ed., Wiley, New York, 2001.
[10] D.T. Sawyer, A. Sobkowiak, J.R. Roberts Jr., Electrochemistry for Chemists ,
Wiley, New York, 1995.
[11] G.R. Facer, D.A. Notterman, L.L. Sohn, Appl. Phys. Lett. 78 (2001) 996–999.
[12] C.R. Keese, J. Wegener, S.R. Walker, I. Giaever, PNAS 101 (2004) 1554–1559.
[13] G.B. Shul’pin, Y.N. Kozlov, L.S. Shul’pina, P.V. Petrovskiy, Appl. Organomet.
Chem. 24 (2010) 464–472.
[14] G. Shulpin, Yu. Kozlov, G. Nizova, G. Suss-Fink, S. Stanislas, A. Kitaygorodskiy,
V. Kulikova, J. Chem. Soc. Perkin Trans. 2 (2001) 1351–1371.
[15] R. Khaliullin, A. Bell, M. Head-Gordon, J. Phys. Chem.
17984–17992.
[16] L.S. Shul’pina, M.V. Kirillova, A.J.L. Pombeiro, G.B. Shul’pin, Tetrahedron 65
(2009) 2424–2429.
B 109 (2005)
[17] C. Walling, Acc. Chem. Res. 8 (1975) 125–131.
[18] N. Emanuel, G. Zaikov, Z. Maizus, Oxidation of Organic Compounds. Medium
Effects in Radical Reactions , Pergamon Press, Oxford, 1984.
[19] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry , third ed.,
Wiley-VCH, Weinheim, 2003.
[20] K. Izutsu, Electrochemistry in Nonaqueous Solutions , Wiley-VCH, Weinheim,
2002.
[21] M. Born, Z. Phys. 1 (1920) 45–48.
4. Conclusions
[22] J.E. Gordon, The Organic Chemistry of Electrolyte Solutions , Wiley, New York,
1975.
[23] R.G. Pearson, in: N.B. Chapman, J. Shorter (Eds.), Advances in Linear Free Energy
Relationships, Plenum Press, London/New York, 1972 (Chapter 6).
[24] A.J. Parker, Chem. Rev. 69 (1969) 1–32.
[25] N. Greenwood, A. Earnshaw, Chemistry of the Elements , Pergamon, Oxford,
1984.
[26] A. Cotton, G. Wilkinson, C. Murillo, M. Bochmann, Advanced Inorganic Chem-
istry , sixth ed., Wiley, New York, 1999.
[27] A.A. Frost, R.G. Pearson, Kinetics and Mechanism , Wiley, New York, 1961.
[28] J.J.P. Stewart, J. Mol. Model. 13 (2007) 1173–1213.
[29] A. Pokutsa, Yu. Kubaj, A. Zaborovskyi, R. Prystanskyi, J. Muzart, R. Makitra,
T. Mysakovych, D. Maksym, A. Sobkowiak, Unpublished results, prepared for
submission to Kinet. Catal.
The acceleration effect induced by the addition of small amounts
of oxalic acid on the rate of the cyclohexane oxidation as well as the
relationships between the electrochemical properties of the reac-
tion mixture (relative permittivity and redox potential) from one
side and the oxidation process characteristics from the other one
have been highlighted and explicated. The yield of products and
their ratio depend on the nature of the solvent and the additive.
The practical interest of the above study lies in the ability to fore-
cast the oxidizing system reactivity, thanks to the determination of
the electrochemical properties of the reaction mixtures.
[30] M. Kitamura, K. Yamashita, H. Imai, Bull. Chem. Soc. Jpn. 49 (1976) 97–100.
[31] M.-A. Nawi, T.S. Reichel, Inorg. Chem. 20 (1981) 1974–1978.
[32] M. Kitamura, K. Yamashita, H. Imai, Chem. Lett. (1975) 1071–1974.
[33] M.-A. Nawi, T.S. Reichel, Inorg. Chem. 21 (1982) 2268–2271.
[34] M.V. Kirillova, M.L. Kuznetsov, V.B. Romakh, L.S. Shul’pina, J.J.R. Fraústo da Silva,
A.J.L. Pombeiro, G.B. Shul’pin, J. Catal. 267 (2009) 140–157.
[35] V. Srinivasan, J. Rocˇek, J. Am. Chem. Soc. 96 (1974) 127–133.
[36] R.A. Sheldon, J.K. Kochi, Oxid. Combust. Rev. 5 (1973) 135–242.
[37] G.B. Shul’pin, Mini Rev. Org. Chem. 6 (2009) 95–104.
Acknowledgements
Authors are indebted to Prof. J. Kowalski and Dr. O. Pereviznyk
for helpful discussions as well as to Mrs. D. Maksym for her assis-
tance in the manuscript preparation. This work was supported by
NATO CLG No. 982510.