Mixed Co-Mn oxide-catalysed selective aerobic oxidation of vanillyl alcohol to vanillin in base-free conditions

Article abstract of DOI:10.1002/cplu.201300247

Manganese-doped cobalt mixed oxide (MnCo-MO) catalyst was prepared by a solvothermal method. The as-prepared catalyst was characterised by X-ray photoelectron spectroscopy, H₂ temperature-programmed reduction, O₂ temperature-programmed oxidation and XRD. This catalyst gave 62 % conversion with 83 % selectivity to vanillin in 2 hours for the liquid-phase air oxidation of vanillyl alcohol without using base. Three different types of metal oxides were observed in the prepared catalyst, which could be identified as Co₃O₄, Mn₃O₄ and CoMn ₂O₄. Among these, the tetragonal phase of CoMn ₂O₄ was found to be more active and selective for vanillyl alcohol oxidation than Co₃O₄ and Mn₃O ₄. High-resolution TEM characterisation revealed the morphology of MnCo-MO nanorods with a particle size of 10 nm. Successful recycling of the catalyst was also established in this oxidation reaction. It takes two: Mn in the MnCo mixed oxide (MO) forms a metal hydroperoxo complex simultaneously giving the phenolate ion of vanillyl alcohol under base-free conditions, which undergoes smooth air oxidation to give vanillin with >83 % selectivity (see scheme). The high catalytic activity of the MnCo-MO results from a synergic effect of the MO phases, which is absent in single Mn or Co oxides. Copyright

Full text of DOI:10.1002/cplu.201300247

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300247
Mixed Co–Mn Oxide-Catalysed Selective Aerobic Oxidation
of Vanillyl Alcohol to Vanillin in Base-Free Conditions
Ajay Jha,[a] Kashinath R. Patil,[b] and Chandrashekhar V. Rode*[a]
Manganese-doped cobalt mixed oxide (MnCo-MO) catalyst was
prepared by a solvothermal method. The as-prepared catalyst
was characterised by X-ray photoelectron spectroscopy, H2
temperature-programmed reduction, O2 temperature-pro-
grammed oxidation and XRD. This catalyst gave 62% conver-
sion with 83% selectivity to vanillin in 2 hours for the liquid-
phase air oxidation of vanillyl alcohol without using base.
Three different types of metal oxides were observed in the pre-
pared catalyst, which could be identified as Co3O4, Mn3O4 and
CoMn2O4. Among these, the tetragonal phase of CoMn2O4 was
found to be more active and selective for vanillyl alcohol oxi-
dation than Co3O4 and Mn3O4. High-resolution TEM characteri-
sation revealed the morphology of MnCo-MO nanorods with
a particle size of 10 nm. Successful recycling of the catalyst
was also established in this oxidation reaction.
Introduction
Mixed metal oxides play a very important role in academic as
well as industrial research because of their acid–base and
redox properties, and constitute the largest family of catalysts
in heterogeneous catalysis.[1] Mixed metal oxides with a struc-
ture of spinel can be obtained by the substitution of metal cat-
ions from the host spinel oxide with other similar metal cations
(dopants). This substitution may restructure chemical bonding
at the surface of the host oxides, which modifies the electronic
properties as well as the chemistry of the host metal oxides.[2]
For example, doping of manganese into Co3O4 spinel leads to
an increase in catalytic activity compared with the material
that contains only a single metal, either cobalt or manganese,
as an active phase.[3,4] The higher activity of the mixed metal
oxides may be a cooperative effect towards an increment in
the mobility of the oxygen as well as stabilising the more
active species and favouring the redox cycles, which also
permit the reactivation of the catalyst.[5] Zhang et al. used
MnOx-doped Co3O4 for the oxidation of CO in an H2 stream.
They proposed that the incorporation of MnOx into Co3O4 in-
creased the amounts of reactive oxygen species and adsorbed
CO species over catalyst surfaces and enhanced the regenera-
tion ability of the reduced catalyst by O2.[6] Similarly, bimetallic
copper–manganese oxide was reported for the liquid-phase
aerobic oxidation of p-cresol to p-hydroxybenzaldehyde. It
showed an excellent performance with 99% conversion and
96% selectivity to p-hydroxybenzaldehyde.[7] Another bimetal-
lic combination of copper with cobalt (CoCu/C) has been used
for the oxidation of o-cresol to salicylaldehyde.[8] The catalytic
properties of Mn-doped Co3O4 catalyst are attributed to the
ability of manganese to form oxides of different stoichiome-
tries or mixed-valence compounds and their high oxygen stor-
age capacity.[9] These metals (Co and Mn) are commonly used
for oxidation reactions and also are more economical than
noble-metal-based catalysts. Despite the large number of stud-
ies on single-component manganese[10–12] and cobalt
oxides[13–16] and very few on mixed cobalt and manganese
oxides for oxidation of volatile organic compounds (e.g., ben-
zene and toluene), there is no single report addressing the cat-
alytic properties based on the structural attributes of combina-
tions of these two oxides for the aerobic oxidation of vanillyl
alcohol, which is an industrially important phenolic starting
compound.
The oxidation product of vanillyl alcohol is vanillin, which
has a wide range applications in food and perfumes because
of its flavour and also finds use in medicinal applications or as
a platform chemical for pharmaceuticals production.[17–19]
A
great deal of work has been done on the synthesis of vanillin
from guaiacol through the 1) nitrose and 2) glyoxalic acid
methods.[17,20] However, these processes are known to give
lower product yields (57%) associated with some serious draw-
backs, such as formation of undesirable side products (nitrile
and p-aminodimethylaniline) in the case of the nitrose method
and use of toxic oxidants (CuO, PbO2 and MnO2)[21–23] in the
glyoxalic acid method. Recently, Hu et al. reported the synthe-
sis of vanillin from oxidation of 4-methylguaiacol by using [Co-
(salen)(py)][PF6]2 (salen=bis(salicylidene)ethylenediamine, py=
pyridine) as the catalyst, to afford complete conversion with
86% selectivity to vanillin in 18 h with mediation of the reac-
tion by NaOH.[24] Thereafter, a combination of cobaltous chlo-
[a] A. Jha, Dr. C. V. Rode
Chemical Engineering and Process Development
National Chemical Laboratory
Pune 411008 (India)
Fax: (+91)20-2590-2621
E-mail: c.v.rode@ncl.res.in
[b] Dr. K. R. Patil
Centre for Materials Characterizations
NCL, Pune 411008 (India)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201300247.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1384

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
ride and N-hydroxyphthalimide-catalysed oxidation in the pres-
ence of NaOH was also reported for vanillin synthesis with
complete conversion of 4-methylguaiacol and 90% product
yield (vanillin) in 7 hours.[25] However, catalyst preparation in
both the processes was complicated, requiring a long reaction
time, and suffered a major problem in its recovery and recycla-
bility. Hence, it is of great practical importance to develop
a simplified preparation protocol for low-cost, highly active
and reusable heterogeneous catalysts with the capability to
form a strong redox couple and give controlled oxidation
products.
Recently, we have been focussing on the development of ef-
ficient heterogeneous catalysts for liquid-phase oxidations of
phenol derivatives.[26,27] This also includes use of Co3O4 nano-
particles for vanillyl alcohol oxidation in the presence of
sodium hydroxide to give 80% conversion with 98% selectivity
to vanillin in 6 hours.[27] In this oxidation, use of an excess
amount of sodium hydroxide is mandatory for two reasons:
firstly, it reacts with the phenolic ꢀOH and prevents oxidation
on the benzene ring; secondly, it prevents the formation of di-
meric by-products and hence is beneficial for selective oxida-
tion.[28] Nevertheless, the use of a large excess of sodium hy-
droxide has two major drawbacks: 1) catalyst deactivation
owing to the formation of an inactive hydroxyl-bridged cobalt
complex in the presence of water, and 2) formation of huge
amounts of waste inorganic salt during product neutralisation.
Hence, there is a genuine need to develop an oxidation route
catalysed without using base. This would require a multifunc-
tional solid catalyst designed in such a way that different
active sites are present in the same catalyst for performing
more than one function. Hence, we developed a manganese-
doped cobalt mixed oxide (MnCo-MO) catalyst, synthesised by
a solvothermal method, which showed an excellent activity for
liquid-phase air oxidation of vanillyl alcohol to vanillin without
using base (Scheme 1).
Figure 1. XRD patterns of A) MnCo-MO and B) Co3O4 (a), mixed oxide (b) and
Mn3O4 (c).
and manganese samples on calcination showed diffraction
phases that were related to Co3O4 and Mn3O4, respectively. The
cobalt sample exhibited peaks at 2q=19 (111), 31.4 (220), 36.9
(311), 44.9 (400), 59.5 (511) and 65.48 (440), ascribed to the
cubic phase of Co3O4 according to JCPDS No. 74-1657. The
manganese sample exhibited a broad peak at 2q=378 corre-
sponding to the (112) plane of Mn3O4. The other peaks of
lower intensity were also identical and matched those of the
tetragonal hausmannite phase (JCPDS card No. 08-0017). In the
case of MnCo-MO catalyst, the diffraction peaks of Co3O4
became weaker and broader after doping of manganese. The
intensities of broad and small peaks in the XRD patterns of
mixed oxides indicated that manganese caused some deforma-
tion in the crystal structure of Co3O4 that resulted in a decrease
of particle size[29] (the crystallite size of Co3O4 was 14 nm,
whereas it was 10 nm for MnCo-MO catalyst as calculated by
using the Scherrer equation). Doping of manganese into the
Co3O4 resulted in a slight shift in position from 2q=36.98 to-
wards a lower diffraction angle of 2q=36.78 (311), which
Scheme 1. Aerobic oxidation of vanillyl alcohol.
Results and Discussion
The XRD pattern of MnCo-MO catalyst in Figure 1A shows dif-
fraction peaks at 2q=33.4 (103), 39.1 (004) and 54.68 (312) cor-
responding to the tetragonal phase of CoMn2O4. The peaks at
2q=24.7 (101) and 41.58 (103) were attributed to the tetrago-
nal phase of Mn3O4 and those at 2q=19, 31.4, 36.9, 44.9, 59.5
and 65.48 were ascribed to the cubic phase of Co3O4. Thus, the
catalyst contained three different spinel oxides, namely
CoMn2O4, Mn3O4 and Co3O4. Figure 1B shows XRD profiles of
both single and mixed metal oxides. Single-component cobalt
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1385

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
might result from the insertion of manganese oxide in the oc-
tahedral framework of Co3O4.[30,31]
transmission electron microscopy (HRTEM) results. CoMn2O4 is
an incomplete normal spinel with a tetragonal structure in
which Co2+ mainly occupies the tetrahedral sites and Mn3+
mainly occupies the octahedral sites. Another peak at 780.9 eV
was also present owing to Co2+, but from spinel Co3O4. A sat-
ellite peak at 782.8 eV was indicative of Co2+ in both the
spinel Co3O4 and CoMn2O4. In the case of undoped Co3O4, the
Co2p3/2 peak could be deconvoluted into two major peaks,
one at 780.2 eV (Co3+) and the other at 782 eV (Co2+); the
very low intensity peak at 784 eV resulted from the 3d!4s
shake-up satellite (Figure 2b), which was a characteristic of
Co3O4.[33] We observed that doping of manganese into Co3O4
led to a decrease in the binding energy of cobalt species,
which might be a result of formation of a Co–Mn compound
(CoMn2O4), and it also reversed the Co2+/Co3+ ratio in a mixed
oxide relative to that in undoped cobalt oxide (Table 1).
X-ray photoelectron spectroscopy (XPS) was used to monitor
the oxidation states and the relative percentages of different
constituent elements in the MnCo-MO catalyst and the results
are shown in Table 1. A broad signal of Co2p3/2 could be fitted
satisfactorily to three principal peaks and one satellite peak
Table 1. Surface composition calculated from XPS.
Catalyst
Atomic ratio (Co/Mn) CoMn2O4
Co3O4
At.%[a]
Co2+ [%] Co2+ [%] Co3+ [%] Mn Co
MnCo-MO 2.6
40
–
27
33
33
67
27 73
Co3O4
–
–
–
[a] At.% = Atomic percent of Mn and Co in mixed oxide (MnCo-MO).
The MnCo-MO catalyst also showed a broad signal of
Mn2p3/2, which could be fitted satisfactorily to three major
peaks and a satellite peak after deconvolution, as shown in
Figure 2c. The peaks at 640.4, 641.4 and 642.4 eV correspond
to Mn2+, Mn3+ and Mn4+, respectively,[9,34,35] whereas the small
after deconvolution, as shown in Figure 2a. The peak at
779.2 eV was assigned to the Co3+ of Co2O3 whereas the peak
at 780 eV was assigned to the Co2+ of CoMn2O4.[32] Formation
of CoMn2O4 was also confirmed by XRD and high-resolution
Figure 2. XPS of a) Co2p3/2 of MnCo-MO, b) Co2p3/2 of Co3O4, c) Mn2p3/2 and d) O1s of MnCo-MO catalyst.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1386

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
peak at 643.6 eV corresponds to the satellite of Mn4+. It is rea-
sonable, therefore, to conclude that Mn element exists as
Mn2+ and Mn3+ in both Mn3O4 and CoMn2O4 spinel oxides.
The presence of these spinel oxides in MnCo-MO catalyst was
also confirmed by XRD peaks at 2q=24.7 (101) and 33.48
(103), respectively. However, the presence of Mn4+ indicates
the possibility of MnO2, which we did not observe in XRD be-
cause of peak broadening.
To evaluate the reducibility and re-oxidisability as a function
of the composition, both temperature-programmed reduction
(TPR) and temperature-programmed oxidation (TPO) studies
were performed for single and mixed metal oxides from 100 to
8008C and the results are displayed in Figures 4 and 5, respec-
tively. TPR of the manganese sample exhibited two peaks at
The manner in which oxygen is bonded to metals in metal
oxides can be determined through XPS of oxygen. In this
series, the O1s spectrum was used to identify the types of
metal oxides present in MnCo-MO catalyst. In the O1s spec-
trum of MnCo-MO, three peaks were fitted satisfactorily after
deconvolution, as shown in Figure 2d. The peaks at lower
binding energies of 529.2 and 529.7 eV could be assigned to
the lattice oxygen and must therefore correspond to CoꢀO
and MnꢀO bonds, respectively.[25,36] However, the peak at the
higher binding energy of 531.2 eV was assigned to the surface-
adsorbed oxygen (O2ꢀ or Oꢀ) or hydroxyl group or oxygen va-
cancies.[37]
Figure 3 shows HRTEM images of MnCo-MO catalyst, syn-
thesised by the solvothermal method and subsequent calcina-
tion at 4508C. As can be seen in Figure 3b, the calcined
Figure 4. H2-TPR profiles of Mn3O4, Co3O4 and MnCo-MO catalysts.
304 and 4608C ascribed to the two-step reduction of
MnO2.[4,38] The H2-TPR profile of Co3O4 was also characterised
by two reduction peaks located at 345 and 5838C (Figure 4),
which resulted from a progressive reduction of Co3O4 as fol-
lows [Eq. (1)]:[39]
Co3O4 ! CoO ! Co0
ð1Þ
The first reduction peak was very sharp and symmetrical
with the temperature maximum centred at about 3458C. This
peak was associated with the reduction of trivalent cobalt
oxide (Co3O4) to a divalent cobalt oxide (CoO) [Eq. (1)]. The
second reduction peak was broad and unsymmetrical with
a temperature maximum at about 5838C, attributable to the
subsequent reduction of divalent cobalt oxide (CoO) to metal-
lic cobalt (Co0).[40,41]
Figure 3. HRTEM images of MnCo-MO catalysts: a) as-synthesised catalyst;
b) calcined sample. c) Lattice fringe patterns and d) SAED pattern.
sample displayed a roughly nanorod-like structure, whereas
the as-synthesised sample showed a fully rod-like morphology
with 79 nm length and 11 nm diameter (Figure 3a). The aver-
age dimensions of calcined MnCo-MO nanorods were approxi-
mately 20–30 nm in length with a diameter of 4–7 nm. The lat-
tice fringe patterns of the MnCo-MO catalyst are shown in Fig-
ure 3c. The dominant exposed planes of MnCo-MO nanorods
were (103) with a lattice spacing of 0.26 nm, which was related
to the spinel CoMn2O4. The other exposed planes were (111)
and (004) with a lattice spacing of 0.46 and 0.22 nm, related to
the spinel Co3O4 and CoMn2O4, respectively. Figure 3d shows
the selected-area electron diffraction (SAED) pattern of the
MnCo-MO catalyst, which was used to determine the lattice
constants; the results were in excellent agreement with those
obtained from XRD analysis.
Four overlapping peaks at 247, 315, 341 and 3928C were ob-
served in the case of the manganese-doped cobalt oxide
sample in a temperature range from 200 to 4008C. It was no-
ticed that the peak at 2478C was observed only in the MnCo-
MO sample, not for Co3O4 and Mn3O4 samples. It is reasonable
to deduce that the peak at 2478C resulted from the reduction
of surface oxygen species generated by the presence of
oxygen vacancies in MnCo-MO oxides. A similar observation
was also reported by Shi et al. for MnxCo3ꢀxO4 oxide used for
formaldehyde oxidation.[42] The hydrogen consumption in
a lower temperature range of 200–4008C could be a result of
the stepwise reduction of Co3+ to Co2+ and also of Mn4+ to
Mn3+. However, the peak in a higher temperature range of
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1387

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
400–7008C could result from the reduction of Co2+ to Co0 and
Mn3+ to Mn2+ in a MnCo-MO sample.[43] The presence of the
reduction peak at a higher temperature of approximately
6378C in the TPR profiles of MnCo-MO catalysts (Figure 4) is
additional evidence of the formation of mixed Co–Mn oxides.[4]
After reduction of single (Co3O4 and Mn3O4) and mixed
metal oxides (MnCo-MO) to the corresponding metallic forms
during TPR experiments, TPO experiments were also per-
formed and the results are displayed in Figure 5. The addition
Figure 6. Catalyst screening for oxidation of vanillyl alcohol. Reaction condi-
tions: vanillyl alcohol (3.24 mmol), catalyst (0.1 g), 1408C, acetonitrile
(70 mL), air pressure (21 bar), time 2 h.
over-oxidation to vanillic acid. However, the MnCo-MO catalyst
showed superior recyclability (discussed later) and selectivity
to vanillin of 83% relative to catalysts containing only the
single oxide of either cobalt or manganese. The higher activity
of the MnCo-MO catalyst might result from spinel CoMn2O4
along with Mn species of varying oxidation states. A separate
oxidation experiment performed by using a physical mixture
of Mn3O4 and Co3O4 showed only 45% conversion of vanillyl al-
cohol with 74% selectivity to vanillin, which was close to that
obtained for spinel Co3O4. The importance of Co/Mn composi-
tion in the mixed metal oxide was also studied by preparing
and testing two mixed oxide samples with varying molar ratios
of Mn and Co (MnCo 1:1 and 2:1). Both of these catalysts
showed either inferior activity and/or selectivity relative to
MnCo-MO catalyst with 1:2 molar ratio of Mn and Co. MnCo
1:1 catalyst showed 52% conversion and 83% selectivity to va-
nillin, whereas MnCo 2:1 showed 64% conversion of vanillyl al-
cohol with 74% selectivity to vanillin (Table S1 in the Support-
ing Information). The lower selectivity to vanillin in the case of
MnCo 2:1 catalyst was owing to the over-oxidation of vanillin
to vanillic acid. MnCo-MO having a Mn/Co ratio of 1:2, pre-
pared by a co-precipitation method,[44] on testing for vanillyl al-
cohol oxidation showed inferior activity (22% conversion) to
MnCo-MO prepared by a solvothermal method (Table S1). In
spite of the fact that MnCo-MO prepared by co-precipitation
had a pure CoMn2O4 phase, it had a very low activity because
of its lower surface area (17 m2 gꢀ1) and higher crystallite size
(62 nm) than MnCo-MO prepared by the solvothermal method,
which had a higher surface area of 105 m2 gꢀ1 and smaller crys-
tallite size of 10 nm (Figure S1).
Figure 5. O2-TPO profiles of Mn3O4, Co3O4 and MnCo-MO catalysts.
of manganese to cobalt oxide led to remarkable changes in
the TPO profiles of single metal oxides with respect to the
amount of oxygen consumed, number of oxidation peaks and
the oxidation temperature. If the reduced MnCo-MO catalyst
was exposed to oxygen, manganese was oxidised to a great
extent and the resulting manganese oxide covered the cobalt
oxide peaks. So we could not distinguish the separate oxida-
tion peaks for Mn and Co oxides. As can be seen from
Figure 5, the area of the oxygen consumption peak for MnCo-
MO catalyst was higher than that of the oxygen consumption
peak for single metal oxides. This observation indicates that
the addition of manganese to cobalt oxide enhanced the con-
sumption of oxygen in mixed metal oxides.
Catalytic activity
The activity of single metal oxides (Co3O4 and Mn3O4) was com-
pared with that of mixed metal oxide (MnCo-MO) catalysts for
the liquid-phase air oxidation of vanillyl alcohol and the results
are shown in Figure 6. Among these catalysts, Co3O4 showed
significantly lower substrate conversion (48%) with higher se-
lectivity to vanillin (88%). The efficiency of Mn3O4 in terms of
vanillin yield was marginally higher (54%) than that obtained
for MnCo-MO catalyst (51%), but the catalyst recycling experi-
ment with Mn3O4 showed a substantially diminished conver-
sion of 29% and also lower selectivity to vanillin (74%) owing
to the higher oxygen storage capacity and faster oxygen ab-
sorption rates of manganese oxide (MnOx), which led to the
Further, the nature of the solvent was also found to have
a profound influence on vanillyl alcohol oxidation. The sequen-
tial oxidation of aldehyde to acid and other by-products may
be inhibited through careful selection of the solvent. In a non-
polar solvent such as toluene, poor conversion of vanillyl alco-
hol was obtained (35%) owing to the lower solubility of vanill-
yl alcohol in toluene. Among the polar solvents, organic polar
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1388

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Figure 8. Effect of reaction time on vanillyl alcohol oxidation. Reaction con-
ditions: vanillyl alcohol (3.24 mmol), MnCo-MO catalyst (0.1 g), 1408C, aceto-
nitrile (70 mL), air pressure (21 bar).
Figure 7. Effect of solvent on the oxidation of vanillyl alcohol. Reaction con-
ditions: vanillyl alcohol (3.24 mmol), MnCo-MO catalyst (0.1 g), 1408C, sol-
vent (70 mL), air pressure (21 bar), time 2 h. IPA=isopropyl alcohol.
The effect of temperature on vanillyl alcohol oxidation was
examined over the MnCo-MO catalyst in a temperature range
of 100–1608C and the results are shown in Figure 9. The con-
version of vanillyl alcohol increased from 41 to 76% with an in-
crease in reaction temperature from 100 to 1608C. The selec-
tivity to vanillin increased from 57 to 83% initially with an in-
solvents showed a better performance than inorganic solvent
(H2O), which might be because of the higher solubility of
oxygen in organic solvents. As can be seen from Figure 7, the
conversion and selectivity were remarkably different in polar
inorganic and polar organic solvents. In water, conversion of
vanillyl alcohol was 52% and selectivity to vanillin was only
44% with majority formation of tar. Vanillyl alcohol is sparingly
soluble in water; hence significant inhibition of the adsorption
of the substrate onto the catalyst surface restricts the rate of
reaction. However, in polar organic solvent, the solubility of
the substrate was higher allowing easy adsorption of reactant
molecules on the catalyst surface and the reaction was readily
facilitated. Oxidation of vanillyl alcohol in acetonitrile gave the
highest conversion of 62% and selectivity of 83% to vanillin
compared with oxidation in isopropyl alcohol (60% conversion
of vanillyl alcohol with 62% selectivity to vanillin). The higher
dielectric constant of acetonitrile (37.5) than that of isopropyl
alcohol (17.9) facilitates better solubility of substrate and oxi-
dant in the solvent, hence affecting the activity and selectivity
pattern.
The effect of reaction time on vanillyl alcohol conversion
and selectivity to vanillin was studied with MnCo-MO catalyst
under optimised reaction conditions and the results are shown
in Figure 8. The conversion of vanillyl alcohol increased from
43 to 63% and the selectivity to vanillin increased from 79 to
83% with an increase in reaction time from 1 to 2 hours. A fur-
ther increase in reaction time from 2 to 4 hours enhanced the
conversion of vanillyl alcohol from 63 to 80%; however, it led
mainly to the formation of tar. As this could not be detected
by HPLC, the material balance of the overall reaction was
<90%. The visible tarry product was separated out from the
reaction crude, dried, weighed and accounted for determining
the vanillin selectivity as well. The selectivity of vanillin was
found to decrease from 83 to 72% with increase in reaction
time from 2 to 4 hours.
Figure 9. Effect of temperature on vanillyl alcohol oxidation. Reaction condi-
tions: vanillyl alcohol (3.24 mmol), MnCo-MO catalyst (0.1 g), 100–1608C,
acetonitrile (70 mL), air pressure (21 bar), time 2 h.
crease in reaction temperature from 100 to 1408C, but a further
increase in temperature to 1608C caused a decrease in vanillin
selectivity from 83 to 74% owing to the formation of over-oxi-
dation products.
Catalyst concentration has a significant effect on the liquid-
phase oxidation of vanillyl alcohol. Experiments were conduct-
ed by varying the catalyst concentration in the range of
0.0007–0.0017 gcmꢀ3 and the results are displayed in
Figure 10. The conversion of vanillyl alcohol increased with in-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1389

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Figure 10. Effect of catalyst concentration on vanillyl alcohol oxidation. Re-
action conditions: vanillyl alcohol (3.24 mmol), MnCo-MO catalyst (0.0007–
0.0017 gcmꢀ3), 1408C, acetonitrile (70 mL), air pressure (21 bar), time 2 h.
Figure 11. Effect of pressure on vanillyl alcohol oxidation. Reaction condi-
tions: vanillyl alcohol (3.24 mmol), MnCo-MO catalyst (0.1 g), 1408C, acetoni-
trile (70 mL), air pressure (7–28 bar), time 2 h.
crease in catalyst loading from 0.0007 to
0.0014 gcmꢀ3, whereas the selectivity to vanillin re-
mained almost constant at about 83%. The increase
in conversion with increase in catalyst loading was
a result of the increase in the number of active sites
available for participating in the reaction. Further in-
crease in catalyst loading from 0.0014 to 0.0017 had
a marginal effect on vanillyl alcohol conversion, thus
implying that mass-transfer resistance could be sig-
nificant under these conditions.
The effect of partial pressure of oxygen on conver-
sion and selectivity was studied for vanillyl alcohol
oxidation by varying the air pressure in the range of
7 to 28 bar at 1408C and the results are shown in
Figure 11. Both the conversion of vanillyl alcohol and
selectivity to vanillin increased gradually with an in-
crease in air pressure from 7 to 21 bar, beyond which
the conversion increased but the selectivity to vanil-
lin started decreasing from 83 to 78%, thereby imply-
Figure 12. MnCo-MO catalyst recycling studies. Reaction conditions: vanillyl alcohol
(3.24 mmol), MnCo-MO catalyst (0.1 g), 1408C, acetonitrile (70 mL), air pressure (21 bar),
time 2 h.
ing that a higher dissolved concentration of oxygen
increased the formation of other by-products.
The stability of the MnCo-MO catalyst for vanillyl
alcohol oxidation was investigated by recycling stud-
ies in the following way. After the first oxidation run with fresh
MnCo-MO catalyst, the reaction crude was allowed to settle
and the clear supernatant product mixture was removed from
the reactor. The catalyst remaining in the reactor was washed
with methanol twice and then dried at 1008C. The subsequent
run was continued by adding a fresh charge to this catalyst.
Figure 12 shows that the activity decreased if only the dried
reused catalyst was used (recycles 1 and 2), but the original ac-
tivity could be regained (63%) if the reused catalyst (recycle 3)
was calcined again at 4508C for 1 hour. A small loss in the ac-
tivity was observed, which resulted from handling of the cata-
lyst.
was also performed as follows. The MnCo-MO catalyst was sep-
arated from the reaction mixture by simple filtration after a par-
tial conversion of 43% in 1 hour, and the filtrate was then em-
ployed in further reaction under similar conditions (21 bar air
pressure at 1408C), which did not give any further conversion.
The absence of any metal species in the filtrate was also con-
firmed by inductively coupled plasma spectrometry. These re-
sults clearly support the fact that the oxidation of vanillyl alco-
hol catalysed by MnCo-MO is truly heterogeneous in nature.
A plausible mechanistic pathway for oxidation of vanillyl al-
cohol catalysed by mixed oxide (MnCo-MO) to give vanillin is
shown in Scheme 2. In the ground state, oxygen remains in
the triplet form containing two unpaired electrons with parallel
spins. Direct reaction of triplet oxygen with singlet organic
To eliminate the possibility of a homogeneously catalysed
reaction owing to dissolution of metal oxide, a leaching test
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1390

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Scheme 2. Plausible mechanistic pathway for the formation of vanillin.
molecules to give a singlet product is a spin-forbidden process
and has a very slow reaction rate. This can be overcome by
complexation of triplet oxygen with a paramagnetic transition-
metal ion.[45] As manganese with d4 configuration has a more
paramagnetic nature (2.8 BM) than cobalt with d7 configuration
(1.7 BM) in a tetragonal complex because of the Jahn–Teller
effect,[46] it has a strong tendency to react with an oxygen mol-
ecule to give a metal superoxo complex. This then further con-
verts into the metal peroxo complex, which readily forms
a metal hydroperoxo complex by abstracting a proton from va-
nillyl alcohol to give the corresponding phenolate ions. The
role of Mn is thus significant in eliminating the use of a free
base (NaOH). The phenolate ions get oxidised to the phenoxy
radical by means of reaction with CoIII species present in the
reaction medium.[47] The phenoxy radical then undergoes fur-
ther reaction to form the quinomethide, which reacts with the
cobalt superoxo complex to give a peroxycobalt(III) complex
followed by its subsequent decomposition to the correspond-
ing aldehyde with concomitant release of a hydroxycobalt(III)
complex. The formed aldehyde with a free radical abstracts the
proton from the reaction medium and is converted to vanillin.
To examine the involvement of free radicals in the reaction,
we performed the oxidation reaction in the presence of
a small amount of radical scavenger, benzoquinone, under the
same reaction conditions. The conversion of vanillyl alcohol
dropped significantly from 62 to 25% without affecting the se-
lectivity, thus supporting the role of free radicals in the reac-
tion.
work led to the formation of tetragonal CoMn2O4 and Mn3O4
spinel phases. The existence of CoMn2O4 was confirmed by
XRD, HRTEM and XPS. TPR also gave additional evidence of the
formation of mixed Co–Mn oxides by the appearance of a re-
duction peak at a higher temperature (ꢁ6378C). TPO of
a mixed oxide exhibited higher oxygen adsorption capacity
than that of a single metal oxide, which was the reason for the
higher activity of the mixed oxide catalysts. The prepared ma-
terial showed a rod-like morphology, as confirmed by HRTEM.
Our catalyst could be successfully recycled three times and no
metal leaching was observed.
Experimental Section
Catalyst preparation
Manganese-doped cobalt oxide catalyst was prepared by a solvo-
thermal method keeping the molar ratio of Co/Mn=2. In a typical
synthesis, manganese acetate tetrahydrate (2.45 g) and cobalt ace-
tate tetrahydrate (4.98 g) were dissolved in ethylene glycol (60 mL),
the solution was heated gradually from room temperature to
1608C, and then 0.3m aqueous Na2CO3 solution (200 mL) was
added dropwise. The resulting slurry was further aged for 1 h
under a nitrogen atmosphere. The product was isolated by filtra-
tion, washed thoroughly with distilled water until neutralisation,
dried at 1008C for 12 h, then calcined at 4508C for 4 h. The ob-
tained catalyst was labelled MnCo-MO. Single-component Co3O4
and Mn3O4 catalysts were also prepared by the same method. In
addition, MnCo-MO was prepared by a co-precipitation method as
described earlier.[44]
Conclusion
Characterisation
A mixed oxide (MnCo-MO) catalyst has been prepared by a sol-
vothermal method and its catalytic activity was evaluated in
the liquid-phase aerobic oxidation of vanillyl alcohol in the ab-
sence of base. Doping of manganese into the Co3O4 frame-
X-ray diffractograms of the catalysts were recorded in the 2q range
of 10–808 (scan rate 5.38minꢀ1) on a PANalytical PXRD Model X-
Pert PRO-1712 instrument, with Ni-filtered CuKa radiation (l=
0.154 nm) as source (current intensity, 30 mA; voltage, 40 kV) and
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1391

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
´
´
[10] J. Trawczynski, B. Bielak, W. Mista, Appl. Catal. B 2005, 55, 277–285.
[11] M. Baldi, E. Finocchio, F. Milella, G. Busca, Appl. Catal. B 1998, 16, 43–
51.
an Xcelerator detector. Temperature-programmed reduction (TPR)
experiments were performed on a Micromeritics Chemisorb 2720
instrument: the catalyst (0.02 g) was placed in a quartz tube and
treated with argon gas (25 mLminꢀ1) at 2008C for 1h. A gas mix-
ture of 5% hydrogen in argon was then passed through the quartz
reactor at 508C for 1 h. The temperature was raised from room
temperature to 9008C at a heating rate of 10 Kminꢀ1 and held at
9008C for 10 min. The reduced sample was then subjected to tem-
perature-programmed oxidation (TPO). X-ray photoelectron spec-
troscopy (XPS) analysis was performed on a VG Scientific ESCA-
3000 spectrometer by using non-monochromatised MgKa radiation
(1253.6 eV). To correct for possible deviations caused by electric
charge of the samples, the C1s line at 284.6 eV was taken as an in-
ternal standard.
[12] A. Gil, L. M. Gandꢁa, S. A. Korili, Appl. Catal. A 2004, 274, 229–235.
[13] V. S. Kshirsagar, S. Vijayanand, H. S. Potdar, P. A. Joy, K. R. Patil, C. V.
Rode, Chem. Lett. 2008, 37, 310–311.
[14] C. V. Rode, M. V. Sonar, J. M. Nadgeri, R. V. Chaudhari, Org. Process Res.
Dev. 2004, 8, 873–878.
[15] A. C. Garade, N. S. Biradar, S. M. Joshi, V. S. Kshirsagar, R. K. Jha, C. V.
Rode, Appl. Clay Sci. 2011, 53, 157–163.
[16] X. Xie, Y. Li, Z. Liu, M. Haruta, W. Shen, Nature Lett. 2009, 458, 746–749.
[17] S. R. Rao, G. A. Ravishankar, J. Sci. Food Agric. 2000, 80, 289–304.
[18] H. Priefert, J. Rabenborst, A. Steinbꢂchel, Appl. Microbiol. Biotechnol.
2001, 56, 296–314.
[19] A. Sheldrak, Nature 1972, 238, 352.
[20] A. D. Bulam, S. V. Nekrasov, B. V. Passet, V. G. Foshkin, J. Appl. Chem.
USSR. 1988, 61, 859.
[21] Y. Chang, M. Zhang, L. Xia, J. Zhang, G. Xing, Materials 2012, 5, 2850–
2871.
Catalytic reaction
All the catalytic oxidation reactions were performed in a 300 cm3
capacity high-pressure Hastelloy reactor supplied by Parr Instru-
ments Co. USA. The reactor was connected to an air reservoir held
at a pressure higher than that of the reactor. A Thermo Scientific
model AS3000 liquid chromatograph equipped with an ultraviolet
detector was used for the analysis. HPLC analysis was performed
on a 25 cm RP-18 column. The product and reactant were detected
with a UV detector at lmax =254 nm. Aqueous methanol (50%) was
used as mobile phase at a column temperature of 358C and a flow
[22] N. J. Hill, J. M. Hoover, S. S. Stahl, J. Chem. Educ. 2013, 90, 102–105.
[23] J. KamLet, E. Conn, SU Pat. 2640083, 1953.
[24] J. Hu, Y. Hu, J. Mao, J. Yao, Z. Chena, H. Li, Green Chem. 2012, 14, 2894–
2898.
[25] Y. Zhang, X. Li, X. Cao, J. Zhao, Res. Chem. Intermed, 2013 DOI:
10.1007s11164–013–1039-x.
[26] M. Biswal, V. V. Dhas, V. R. Mate, A. Banerjee, P. Pachfule, K. L. Agrawal,
S. B. Ogale, C. V. Rode, J. Phys. Chem. C 2011, 115, 15440–15448.
[27] A. Jha, C. V. Rode, New J. Chem. 2013, 37, 2669–2674.
[28] F. Wang, G. Yang, W. Zhang, W. Wu, J. Xu, Chem. Commun. 2003, 1172–
1173.
[29] J. Li, X. Liang, S. Xu, J. Hao, Appl. Catal. B 2009, 90, 307–312.
[30] F. Morales, D. Grandjean, A. Mens, F. M. F. de Groot, B. T. Weckhuysen, J.
Phys. Chem. B 2006, 110, 8626–8639.
rate of 0.7 mLminꢀ1
.
In a typical experiment, vanillyl alcohol (0.5 g), catalyst (0.1 g) and
acetonitrile (70 mL) were charged into a 300 mL Parr autoclave.
The reaction mixture was heated to 1408C. After the desired tem-
perature was attained, the reactor was pressurised with 21 bar air
pressure. Then the reaction was started by agitating at 1000 rpm.
The reaction was continued for up to 2 h, then the reactor was
cooled to room temperature and the unabsorbed air was vented
out. The content of the reactor was discharged and the final
volume was noted.
[31] Q. Liang, K. D. Chen, W. H. Hou, Q. J. Yan, Appl. Catal. A 1998, 166, 191–
199.
[32] C. D. Wagner, J. F. Moulder, L. E. Davis, W. M. Riggs, Handbook of X-ray
Photoelectron Spectroscopy, 1st ed (Ed.: G. E. Muilenberg), Perkin-Elmer
Corporation, Physical Electronics Division, Eden Prairie, 1979.
[33] B. Tan, K. J. Klabunde, P. M. A. Sherwood, J. Am. Chem. Soc. 1991, 113,
855–861.
[34] S. Ponce, M. A. Pena, J. L. G. Fierro, Appl. Catal. B 2000, 24, 193–205.
[35] D. Delimaris, T. Ioanides, Appl. Catal. B 2008, 84, 303–312.
[36] C. Liu, Q. Liu, L. Bai, A. Dong, G. Liu, S. Wen, J. Mol. Catal. A 2013, 370,
1–6.
[37] J. Li, S. Xiong, X. Lia, Y. Qian, J. Mater. Chem. 2012, 22, 23254–23259.
[38] M. Ferrandon, J. Carnç, S. Jꢃrꢄs, E. Bjçrnbom, Appl. Catal. A 1999, 180,
141–151.
[39] J. Luo, M. Meng, X. Li, J. Catal. 2008, 254, 310–324.
[40] P. Arnoldy, J. A. Moulijn, J. Catal. 1985, 93, 38–54.
[41] W. J. Xue, Y. F. Wang, P. Li, Z. Liu, Z. P. Hao, C. Y. Ma, Catal. Commun.
2011, 12, 1265–1268.
Acknowledgements
A.J. gratefully acknowledges the Council of Scientific and Industri-
al Research (CSIR, New Delhi) for the award of a senior research
fellowship.
Keywords: aerobic oxidation
catalysis · manganese · spinel phases
·
cobalt
·
heterogeneous
[42] C. Shi, Y. Wang, A. Zhu, B. Chen, C. Au, Catal. Commun. 2012, 28, 18–
22.
[43] K. Jirꢅtovꢅ, J. Mikulovꢅ, J. Klempa, T. Grygar, Z. Bastl, F. Kovanda, Appl.
Catal. A 2009, 361, 106–116.
[44] S. A. Hosseini, D. Salari, A. Niaei, F. Deganello, G. Pantaleo, P. Hojati, J.
Environ. Sci. Health Part A 2011, 46, 291–297.
[1] M. B. Gawande, R. K. Pandey, R. V. Jayaram, Catal. Sci. Technol. 2012, 2,
1113–1125.
[2] E. W. McFarland, H. Metiu, Chem. Rev. 2013, 113, 4391–4427.
[45] R. A. Sheldon, R. A. van Santen, Catalytic Oxidation: Principles and Appli-
cations, World Scientific, Singapore, 1995, pp. 12–13.
[46] J. E. Huheey, E. A. Keiter, R. L. Keiter, O. K. Medhi, Inorganic Chemistry:
Principles of Structure and Reactivity, Pearson Education, Upper Saddie
River, NJ, 2008, pp. 489.
˝
[3] Z. Tian, P. H. T. Ngamou, V. Vannier, K. Kohse-Hoinghaus, N. Bahlawane,
Appl. Catal. B 2012, 117 – 118, 125–134.
´
[4] S. Todorova, H. Kolev, J. P. Holgado, G. Kadinov, Ch. Bonev, R. Perenꢁ-
guez, A. Caballero, Appl. Catal. B 2010, 94, 46–54.
[5] D. A. Aguilera, A. Perez, R. Molina, S. Moreno, Appl. Catal. B 2011, 104,
144–150.
[6] Q. Zhang, X. Liu, W. Fan, Y. Wang, Appl. Catal. B 2011, 102, 207–214.
[7] F. Wang, G. Yang, W. Zhang, W. Wu, J. Xu, Adv. Synth. Catal. 2004, 346,
633–638.
[47] B. Barton, C. G. Logie, B. M. Schoonees, B. Zeelie, Org. Process Res. Dev.
2005, 9, 62–69.
Received: July 4, 2013
[8] F. Wang, J. Xu, S. Liao, Chem. Commun. 2002, 626–627.
[9] S. Todorova, A. Naydenov, H. Kolev, J. P. Holgado, G. Ivanov, G. Kadinov,
A. Caballero, Appl. Catal. A 2012, 413–414, 43–51.
Revised: September 3, 2013
Published online on October 14, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 1384 – 1392 1392