Characterisation and activity of mixed metal oxide catalysts for the gas-phase selective oxidation of toluene

Article abstract of DOI:10.1016/j.cattod.2019.06.001

Mixed metal bi-component oxide catalysts, including Fe/Mo, U/Mo, U/W, Fe/U, U/V and U/Sb have been prepared, characterised and evaluated for gas phase selective toluene oxidation. Selective toluene oxidation activity to form benzaldehyde was exhibited by Fe/Mo, U/Mo and U/W mixed oxide catalysts. The Fe/Mo catalyst produced the highest benzaldehyde yield. Catalysts that formed benzaldehyde also produced a range of by-products, these were other partial oxidation and coupling products, and preliminary studies of benzaldehyde oxidation suggests they were formed from secondary reactions of benzaldehyde. The Fe/U, Sb/U and U/V catalysts produced only total oxidation to carbon oxides. Catalysts were characterised by X-ray diffraction, laser Raman spectroscopy and temperature programmed reduction. Single molybdate phases were identified for the Fe/Mo and U/Mo catalysts, and a mixture of uranium molybdate and WO₃ was identified for the U/W catalyst. Results suggest that the formation of a molybdate phase is important for the selective oxidation of toluene. In contrast, the U/Fe catalyst was a mixture of U₃O₈ and V₂O₅, whilst the Fe/U catalyst was comprised of highly dispersed iron oxide on UO₃. The presence of U₃O₈ was responsible for toluene total oxidation. The U/Sb catalyst did not exhibit selective toluene oxidation, but previous studies have demonstrated that the catalyst exhibits high activity for selective propene oxidation. Similar behaviour has been observed for the other catalysts in this study, and it is apparent that catalysts that were selective for toluene oxidation were not selective for propene/propane oxidation and vice versa.

Full text of DOI:10.1016/j.cattod.2019.06.001

Accepted Manuscript
Title: Characterisation and activity of mixed metal oxide
catalysts for the gas-phase selective oxidation of toluene
Authors: Maria-Luisa Palacios, Stan Golunski, Graham J
Hutchings, Stuart H. Taylor
PII:
S0920-5861(18)31552-9
DOI:
Reference:
https://doi.org/10.1016/j.cattod.2019.06.001
CATTOD 12248
To appear in:
Catalysis Today
Received date:
Revised date:
Accepted date:
28 September 2018
17 May 2019
1 June 2019
Please cite this article as: Palacios M-Luisa, Golunski S, Hutchings GJ,
Taylor SH, Characterisation and activity of mixed metal oxide catalysts
for the gas-phase selective oxidation of toluene, Catalysis Today (2019),
https://doi.org/10.1016/j.cattod.2019.06.001
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Characterisation and activity of mixed metal oxide catalysts for the gas-
phase selective oxidation of toluene
Maria-Luisa Palacios, Stan Golunski, Graham J Hutchings* and Stuart H. Taylor*
Cardiff Catalysis Institute, School of Chemistry, Cardiff University, Main Building, Park
Place, Cardiff, CF10 3AT, UK.
*Corresponding authors: taylorsh@cardiff.ac.uk; +44 (0)29 2087 4062
hutch@cardiff.ac.uk; +44 (0)29 2087 4059
Graphical Abstract
Benzoate species
OH
OH
O
O
O
O
O
O
O
Metal oxide
Metal oxide
Metal oxide
Toluene selective oxidation
CHO
O
O
Metal oxide
COOH
O
O
O
O
1

Highlights




Uranium mixed oxide catalysts are active for toluene oxidation
Molybdenum uranium oxides give the highest benzaldehyde yield
Fe/U, Sb/U and U/V catalysts give only total oxidation to carbon oxides
U3O8 is responsible for toluene total oxidation
Abstract
Mixed metal bi-component oxide catalysts, including Fe/Mo, U/Mo, U/W, Fe/U, U/V and U/Sb
have been prepared, characterised and evaluated for gas phase selective toluene oxidation.
Selective toluene oxidation activity to form benzaldehyde was exhibited by Fe/Mo, U/Mo and
U/W mixed oxide catalysts. The Fe/Mo catalyst produced the highest benzaldehyde yield.
Catalysts that formed benzaldehyde also produced a range of by-products, these were other
partial oxidation and coupling products, and preliminary studies of benzaldehyde oxidation
suggests they were formed from secondary reactions of benzaldehyde. The Fe/U, Sb/U and U/V
catalysts produced only total oxidation to carbon oxides. Catalysts were characterised by X-ray
diffraction, laser Raman spectroscopy and temperature programmed reduction. Single
molybdate phases were identified for the Fe/Mo and U/Mo catalysts, and a mixture of uranium
molybdate and WO3 was identified for the U/W catalyst. Results suggest that the formation of
a molybdate phase is important for the selective oxidation of toluene. In contrast, the U/Fe
catalyst was a mixture of U3O8 and V2O5, whilst the Fe/U catalyst was comprised of highly
dispersed iron oxide on UO3. The presence of U3O8 was responsible for toluene total oxidation.
The U/Sb catalyst did not exhibit selective toluene oxidation, but previous studies have
demonstrated that the catalyst exhibits high activity for selective propene oxidation. Similar
behaviour has been observed for the other catalysts in this study, and it is apparent that catalysts
that were selective for toluene oxidation were not selective for propene/propane oxidation and
vice versa.
Keywords: catalytic selective oxidation, mixed oxides, toluene, benzaldehyde, uranium,
molybdenum, tungsten
2

1. Introduction
Selective oxidation to partially oxidised products is an important research aim; and provides a
facile route for the functionalisation of hydrocarbons. Within the field of heterogeneous
catalytic oxidation much research effort has focussed on hydrocarbon partial oxidation to
commercially more valuable products. Achieving high selectivity in hydrocarbon oxidation is
especially challenging, as thermodynamically oxidation favours the formation of carbon
dioxide and water. The oxidation of aromatic compounds is potentially a facile process for the
production of oxygenated derivatives. Toluene can be converted to benzaldehyde, benzoic acid
and benzyl benzoate, as well as a range of other products, and these are commercially
significant as versatile intermediates in the manufacture of pharmaceuticals, dyes, solvents,
perfumes, plasticizers, dyestuffs, preservatives and flame retardants. Currently benzaldehyde is
produced commercially by the chlorination of toluene followed by saponification [1], which is
a more complex process than direct partial oxidation. Relatively recently efficient toluene
selective oxidation by oxygen in the liquid phase has been achieved in a batch process using
titania supported gold-palladium nanoparticles [2]. However, interest still remains for selective
toluene oxidation in a gas phase flow process.
The gas-phase selective oxidation of toluene has been studied using heterogeneous catalysts
based on various single and mixed metal oxides, with the majority of studies using molybdenum
and vanadium oxide systems [3],[4],[5],[6]. The aim of this study is to synthesise, characterise
and explore more fully the use of mixed metal oxide systems, especially those containing
uranium, as it has a well-established history as a selective oxidation catalyst component. These
include Fe/Mo, U/Mo, Fe/U, Sb/U and V/U oxide catalysts, as mixed metal oxides with these
elements have been prepared previously. The catalysts selected for this study have been chosen
3

for several reasons; in particular uranium oxide is a central component of some selective
oxidation catalysts, and it is a component that is not often currently studied. The actinide
elements exhibit an electronic structure with 86 electrons, surrounded by valence electrons in
the 7s, 6d and 5f shells. Uranium oxides are generally presented as ionic solids. However, an
alternative interpretation involves covalent bonding with the formation of new molecular
orbitals from effective overlapping of appropriate atomic orbitals of similar energy. These two
descriptions are limiting cases, and actinide compounds are often best described as partially
ionic and partially covalent. The electronic structures of these elements and their compounds
are determined by a strong competition between the 5f, 6d and 7s electrons. In uranium these
shells are so close in energy that hybridisation of the 5f and 6d electrons can occur. Hence the
unique electronic properties of uranium give the element the potential for many interesting
catalytic properties [7].
Uranium has six valence electrons, and it is therefore similar in nature to the transition elements
chromium, molybdenum and tungsten. It can also show structural similarities with MoO3 and
WO3.In particular, molybdenum and tungsten are oxides forming the basis of catalysts for many
selective oxidation reactions, for example Fe2(MoO4)3 is used as a commercial catalyst for the
selective oxidation of methanol to formaldehyde [8]. Uranium oxides also exist in a variety of
different forms, with some of the most common being UO2, U4O9, U3O7, U3O8 and UO3. These
oxides exhibit a range of oxidation states, and in some cases mixed oxidation states.
Furthermore, uranium oxides also have extensive redox properties with the ability to show a
wide range of nonstoichiometry and semiconducting behaviour. Against this background
uranium oxide is a flexible catalyst component, as it have versatile properties, many of which
are important for catalytic selective oxidation reactions.
Pioneering work by Grasselli and co-workers developed mixed uranium-antimony oxide
catalysts for the commercial selective oxidation and ammoxidation of propylene to acrolein and
4

acrylonitrile [9], [10], [11]. The structure of USb3O10 is critical for selective oxidation, and the
phase is structurally complex. Uranium is present in a single environment, whilst antimony is
in two and oxygen in four different environments. The structure comprises of layers, which
contain heavy atoms and oxygen alternating with layers of oxygen only. Five layers of planes
containing heavy metal atoms are required to completely describe the unit cell of USb3O10.
There are two types of lattice oxygen in the structure, one giving high selectivity for partial
oxidation, and the other, which was less active and selective, producing carbon oxides [10].
Optimum performance was only observed with the USb3O10 phase, and catalyst preparation is
crucial to achieve the correct phase as a number of other structures are possible. The USbO5
structure is very similar to that of USb3O10 and it is proposed as a precursor to USb3O10, which
again decomposes at high temperature to USbO5 [10, [12]. The elegant work of Grasselli and
co-workers has inspired many studies, and more recent work has investigated alternative
methods, such as hydrothermal synthesis for the preparation of uranium-antimony mixed
selective oxidation catalysts [13]. We have also used the work of Grasselli and co-workers as
inspiration for investigating the current range of uranium-containing two-component mixed
metal oxide catalysts for selective oxidation, and this approach forms the basis of a large part
of the present study.
2. Experimental
Catalyst preparation
The Fe/U (atomic ratio Fe/U = 0.5/3) was prepared using a co-precipitation method.
Fe(NO3)3.9H2O (1 g, Aldrich 99.99+%) was dissolved in deionised water (10 ml) and added to
a solution of UO2(NO3).2H2O (5.85 g, Strem 99+%), in deionised water (16 ml). The two
solutions were mixed and the excess water evaporated overnight at 150 C. The solid obtained
was calcined (450 C, 3 h) in static air to produce the catalyst.
5

The U/Sb catalyst was synthesised using methods described in the literature, reacting
UO2(NO3).2H2O (8.8 g, Strem 99+%) with Sb2O3 (8.7 g, Aldrich 99.99%) dissolved in 70%
nitric acid and refluxed for 3 h [8,10]. After cooling, the solution was adjusted to pH 8 by adding
28% aqueous NH4OH solution. The solid product was filtered, washed with deionised water
and dried (120 C, 24 h) to produce the precursor of the catalyst. The precursor was then
calcined (425 C, 16 h) and the temperature increased to 925 C for a further 16 h. The catalysts
produced by this procedure had an excess of antinomy, observed by XRD, in the form of the
oxide. The contaminating Sb2O3 was removed by refluxing the catalyst with aqueous HCl at
100 C to produce the final catalyst.
The U/Mo catalysts were synthesised using a method described in the literature [3]. A solution
was prepared by dissolving U(CH3COO)2.2H2O (5.4 g, BDH AnalaR grade) in deionised water
(19.6 ml). The solution was added to a solution of (NH4)6Mo7O24.4H2O, (0.4 g, Aldrich)
dissolved in hot deionised water (20 ml). The excess water was evaporated and the resulting
solid calcined in flowing air (550 oC, 4 h).
The Fe/Mo catalyst was prepared by a precipitation method [14]. Fe(NO3)3.9H2O (2.0 g,
Aldrich 99.99+%) was dissolved in deionised water (15 ml) and added to a solution of
(NH4)6Mo7O24.4H2O (1.3 g, Aldrich) dissolved in deionised water (10.5 ml). The resulting
precipitate was dried and calcined in static air (400 C, 6 h and then 750 C, 24 h).
The U/W catalyst was prepared following a procedure previously described in the literature
[15]. Preparation was via heating a physical mixture of U3O8 (10 g, Strem) and WO3 (8.26 g,
Aldrich) at 700 C for 24 h and then at 800 C for 72 h in static air. The U/V catalyst was also
prepared by heating a physical mixture of V2 O5 (1.82 g, Aldrich 99.99+%), V2O3 (0.3 g, Aldrich
99.99+%) and U3O8 (6.7 g, Strem) at 550 C for 24 h in flowing nitrogen.
6

Catalyst characterisation
Powder X-ray diffraction (XRD) studies were performed using an Enraf Nonius FR590
instrument with a Cu source operated at an X-ray power of 1.2 kW (30 mA, 40 kV). A Ge(111)
single crystal monochromator was used to select Cu KX-rays for analysis. The powdered
samples were compressed into an aluminium sample holder, which was rotated during data
collection to compensate for any ordering of crystallite orientation, which may occur during
sample preparation. The diffraction pattern was measured by means of a position sensitive
detector (Inel PSD120), covering 2 values over the range 4.4-124.6. Raw data were corrected
against a silicon standard, and phase identification was performed by matching the experimental
pattern against standard entries in the ICDD database.
Catalyst surface areas were measured by the BET method, using a Micromeritics ASAP 2010
apparatus. Prior to determination catalysts were degassed at 75 C for 6 h. Measurements were
carried out at -196 °C using multiple point high purity nitrogen adsorption.
Laser Raman spectroscopy (LRS) analysis was performed using a Renishaw system 1000 laser
Raman microscope. An argon ion laser (514.5 nm) was used as excitation source and was
typically operated at a power of 20 mW. All the samples were used in powdered form and were
placed on a microscope slide and the laser focused on the sample to produce a spot size of ca.
3 m in diameter. Spectra were collected using a back-scattering geometry with an 180 angle
between the illuminating and the collected light. In certain studies the laser power was varied
by using a series of filters. The laser was focused onto the sample by means of an Olympus
BH2-UMA optical microscope.
The apparatus used for temperature programmed reduction (TPR) was a TPR/TPO/TPD
Micromeritics Autochem 2900 equipped with a TCD detector. The reducing gas used in all
experiments was 10%H2 in Ar, with a flow rate of 50 ml min-1. The temperature range explored
was from room temperature to 1000 C with a ramp rate of 10 C min-1. The sample mass was
7

varied (25-100 mg) depending on the sample and details of specific conditions are given with
the TPR profiles.
Catalyst activity
Catalytic activity was determined in a fixed bed plug flow microreactor. Helium and oxygen
flow rates were controlled with electronic thermal mass flow controllers. The gases flowed into
a vaporiser thermostatically controlled at 120 C, and were mixed with toluene vapour, which
was injected by a syringe pump. The reactor tube (i.d 13 mm) was held vertically and was
constructed from Pyrex. The catalyst was held in the middle of the reactor tube, supported upon
a porous sintered quartz frit. The catalyst bed was heated by a ceramic tube furnace and the
temperature controlled by a thermocouple placed in intimate contact with the catalyst bed. The
outlet of the reactor was heated to 170 C with a heating tape in order to minimise condensation
of reactants and products. At the end of an experiment the catalyst was cooled down in a flow
of He with all lines heated. The total gas flow rate was 27.75 ml min-1 with the ratio of
toluene/oxygen/helium = 1/16/83. The catalyst was loaded to a constant volume of 0.45 cm3
for all experiments; resulting in a gas hourly space velocity (GHSV) of 3,700 h-1. Product
analysis was performed on-line using a Varian 3400 GC with thermal conductivity and flame
ionisation detectors. Separation was achieved using Poropak Q (2 m x 1/8” o.d.) and Molecular
sieve 13X (2 m x 1/8” o.d.), with the columns connected in a series/by-pass configuration.
Using a suitable temperature programmed regime it was possible to separate and quantify the
products and reactants with a single analysis. Toluene conversion was calculated from the
difference between the concentrations in the reactor exit gas at a given temperature as a
percentage of the toluene concentration in the reactant feed. Selectivities were defined as the
percentage product formed in relation to the total products measured in the gas phase, after
correction for the number of carbon atoms. Carbon balances were in the range of95-105 %.
8

3. Results
Catalyst characterisation
The atomic ratio of the catalyst as synthesised and the BET surface areas are presented in Table
1. The surface areas of the catalysts were diverse, and very low surface areas (<1 m2 g-1) were
determined for the U/Mo, Fe/Mo and U/V catalysts. The surface area of the U/W catalyst was
also very low, however, the surface areas of U/Sb and U/Fe were significantly greater, at 25
and 18 m2 g-1 respectively.
The phases identified from powder X-ray diffraction are summarised in Table 1. The presence
of single mixed oxide phases, with no other contaminating crystalline phases, were confirmed
for the Fe/Mo, U/Mo, and U/Sb catalysts. In the first two systems molybdate phases were
identified, these were iron molybdate (Fe2(MoO4)3) and uranyl molybdate (UO2MoO4)
respectively. In the case of the U/Sb catalyst the USb3O10 phase was observed and there was
no evidence for excess Sb2O3 which was identified during preparation. On the contrary, the
remaining catalysts were a mixture of single oxides or mixtures of single oxides and mixed
oxides. The former has been confirmed for the U/V system as the phases identified were U3O8
and V2O5, and for the U/Fe system were only the UO3 phase was observed. Whilst for the U/W
catalyst a mixture of bulk WO3 and UO2WO4 was identified.
The catalysts were investigated using laser Raman spectroscopy, and by virtue of analysis by
back scattered light LRS is a more surface sensitive technique compared to the bulk analysis of
powder XRD. The combination of both techniques provides an excellent set of tools for
characterisation. In all cases the spectra were compared with the spectra of their respective
single oxide components and appropriate reference compounds.
The Raman spectrum of the U/Mo oxide catalyst exhibited bands in the 955-165 cm-1 region
(Figure 1). The most intense Raman bands were observed at 747 and 761 cm-1, whilst significant
bands were also observed at 955 and 886 cm-1. A further series of peaks were centred around
9

330 cm-1. Raman spectra of (NH4)6Mo7O24.4H2O and MoO3 have also been recorded for
referencing purposes.
The Raman spectra of the Fe/Mo mixed oxide catalyst is presented in Figure 2. The most intense
Raman band was at 783 cm-1, with further major bands at 821, 937, 967 and 991 cm-1. Iron(III)
oxide was also studied by Raman spectroscopy as a reference material. The U/Sb oxide catalyst
is shown in Figure 3. The most intense Raman band was observed at 491 cm-1. The remaining
Raman bands were relatively broad, and with the exception of the peak at 205 cm-1, they were
also of relatively low intensity. The spectrum showed an increasing background indicative of
sample fluorescence from the excitation source. Antimony oxides, Sb2O3 and Sb2O4, were also
studied to aid spectral assignment.
The Raman spectrum of the Fe/U catalyst showed relatively few active bands and these were
also relatively broad (Figure 4). The most intense band in the U/Fe spectrum was at 762 cm-1,
with much less intense broad features at 821 477 and 322 cm-1.
The Raman spectrum of the U/V mixed oxide catalyst is shown in Figure 5. Raman studies
demonstrated sample fluorescence, despite the fluorescence the Raman active bands are
extremely well-defined with the dominant band at 997 cm-1. Strong Raman bands were also
observed at 703, 528 and 286 cm-1. For comparison V2O5 was used as a reference material.
The Raman spectrum of the U/W oxide catalyst showed a large number of Raman bands (Figure
6). Intense bands were observed at 806, 717 and 272 cm-1 and generally these were broader
than the other lower intensity peaks in the spectrum.
Temperature programmed reduction studies were performed on the Mo/U, Fe/Mo, Fe/U and
U/W catalysts. The TPR profile for the U/Mo mixed oxide catalyst showed three peaks at 614,
653 and 780 C (Figure 7). In the case of the Fe/Mo mixed oxide catalyst the TPR profile was
better resolved when a larger sample mass was used. In fact only the profile obtained with 70
mg of catalyst was resolved into four peaks at 670, 767, 831 and 951 C (Figure 8). The TPR
10

profile for the U/W mixed oxide catalyst is shown in Figure 9. Three peaks were observed at
738, 843 and 891 C. For the Fe/U mixed oxide the TPR profile revealed important information,
regarding the structure of the catalyst (Figure 10). The profile exhibited two reduction maxima
at 489 and 576 C.
Catalytic activity
Activity studies were carried out in the temperature range from 150 C to 530 C, although data
summarised in Table 8, are generally from 300 C, as no significant activity was observed
below this temperature. Carbon balances for detected gas phase products were generally 100
5%, although for some experiments when higher yields of benzaldehyde were observed this
value decreased significantly. Blank experiments in an empty reactor showed only trace activity
at 500 °C, suggesting that reactions of toluene in the gas phase and those initiated by the reactor
walls were negligible over the temperature range investigated.
The mixed oxide catalysts Mo/U, Fe/Mo and U/W were all active for toluene selective oxidation
to benzaldehyde in varying degrees. The other gaseous products from these catalysts were
carbon oxides, with trace quantities of aromatic oxidative coupling products, which were
analysed off-line by NMR spectroscopy and GC-MS. The best catalytic results for toluene
selective oxidation were obtained using the Fe/Mo oxide, which produced a conversion of 43%
and selectivity to benzaldehyde of 86% at 400 C. The U/Mo catalyst also demonstrated
appreciable conversion and selectivity to benzaldehyde, and at 500 C a 90% conversion of
toluene and 13% selectivity towards benzaldehyde was observed. Selective toluene oxidation
was also evident over the U/W catalyst, which produced benzaldehyde at 475 C with 5%
selectivity; the toluene conversion under these conditions was 74%. The yields of benzaldehyde
produced by the Fe/Mo, U/Mo and U/W oxide catalysts, albeit relatively low, are comparable
11

with the per pass yields from other studies [3]. The per pass yield of benzaldehyde obtained
from the Fe/Mo catalyst at 400 °C was particularly noteworthy.
In addition to gaseous products, the Fe/Mo, U/Mo and U/W mixed oxide catalysts also formed
solid products, and these were characterised by NMR spectroscopy. There was a significant
quantity of solid products produced from the Fe/Mo catalyst and benzaldehyde, benzoyl
peroxide
and benzoic acid were identified. At 450 °C and above the Fe/Mo catalyst
showed deactivation, and this was due to deposition of solid reaction products throughout the
catalyst bed and the reactor, evidenced by a decrease in the carbon balance measured from the
gaseous products detected. A similar array of solid products was also detected for the U/Mo
and U/W catalysts, although they were not as active for benzaldehyde production, and formed
less deposited solid products, and did not demonstrate the same deactivation as the more active
Fe/Mo catalyst.
The U/Sb, U/V and Fe/U catalysts were active for toluene oxidation, however, they were not
selective for partial oxidation as none produced benzaldehyde. The only gaseous products
detected from these catalysts were those of deep oxidation. The Fe/U and U/V catalysts were
the most active in this study, both exhibited appreciable conversion at 300 C. The catalysts
based on U/Sb and U/V only produced CO2, whilst the U/Sb system was considerably less
active than U/V and only showed 28% toluene conversion at 530 C. The U/V catalyst also
produced CO in addition to CO2, although CO2 selectivity was always >96%. It was also
evident that the U/Sb, U/V and Fe/U mixed oxide catalysts did not produce any of the aromatic
coupling products or solid products which were produced from the catalysts which were active
for toluene partial oxidation to benzaldehyde.
12

4. Discussion
Catalyst structure
The structure of the U/Mo oxide catalyst can be determined from comparison of the Raman
vibrational data with a series of reference molybdenum and uranium compounds, these include
(NH4)6Mo7O24.4H2O, MoO3, uranyl nitrate and various oxides (Table 2). There are two
common coordination states for molybdates, these are tetrahedral (Td) in the MoO42- unit and
octahedral (Oh) in the Mo7O246- unit [16].
2-
6-
7MoO4 + 8H+  Mo7O24 + 4H2O
2-
The position of the Mo=O symmetric stretching vibrational band of MoO4 is at 955 cm-1, and
therefore confirms the presence of the tetrahedral structure [17]. The band at 886 cm-1 is
2-
probably due to the asymmetric stretching vibration of the Mo=O of the MoO4 tetrahedral
molybdate and corresponds well with tetrahedrally coordinated crystalline K2MoO4 []. Some
uncertainty on the assignation of this band still remains due to the fact that the band at 886 cm-
1 could also be due to the symmetric stretching vibration of the uranyl ion [14]. Bands at 799,
793, and 747 cm-1 are due to asymmetric stretching vibrations in Mo-O-Mo bonds. The band at
761 cm-1 is proposed to be the symmetric stretching O-U-O-U vibration, although the frequency
is shifted from that expected for the common oxides of uranium, confirming that discrete
uranium oxide phases were not present [12]. The Mo-O-Mo bending vibration bands are at 368
cm-1 and 224 cm-1, whilst the bending vibration of O-U-O is present at 347 cm-1.
The Raman assignment of the Fe/Mo mixed oxide catalyst is shown in Table 3. The main
Raman band of the Fe/Mo catalyst was at 783 cm-1 and corresponds to the asymmetric
stretching vibration Mo-O-Mo of the MoO42- tetrahedral molybdate. The weak Raman features
due to the symmetric stretching vibrations of Mo=O of the MoO42- tetrahedral molybdate were
13

observed at 967 and 937 cm-1. The band at 821 cm-1 was attributed to asymmetric stretching
2-
vibration Mo=O of the MoO4 . The band at 991 cm-1 is best assigned to an Mo=O vibration of
an MoO3 phase, however, no evidence for a MoO3 phase was apparent from the XRD data. It
is also possible that the band could be due to the Mo=O vibration of the molybdate phase which
is shifted by the incorporation of iron. The absence of MoO3 signatory bands at 917 and 665
cm-1 support the suggestion that MoO3 is not present in addition to the molybdate phase.
The detailed assignment of the Raman active bands in the U/Sb mixed oxide catalyst is more
difficult as there is little previous Raman data on the mixed oxide, which has a large and
complex unit cell [18]. The structure of the mixed oxide has a variety of different coordinations
for both uranium and antimony. General assignments have been made and these are shown in
Table 4. Bands in the region ca. 304 cm-1 are in the bending region and are attributed to U-O-
U and Sb-O-Sb. The other two groups of bands at 345-560 cm-1 and 776-905 cm-1 are assigned
to symmetric and asymmetric stretching vibrations of U-O and Sb-O. The most significant
observation from the Raman data is that the active bands in the mixed oxide spectrum do not
correspond to those in the spectra of the antimony and uranium single oxides. In the case of the
Fe/Mo, U/Mo and U/Sb catalysts the formation of mixed oxide phases were confirmed, and this
was in good agreement with the results obtained by XRD.
Raman data are summarised for the Fe/U catalyst in Table 5. The main band in the spectra at
762 cm-1 was due to U-O stretching vibration, which along with the weaker band at 822 cm-1 is
characteristic of the simple Raman spectra of UO3 (Table 3). Both bands are slightly shifted to
lower wave number when compared to UO3, and this is probably due to a chemical interaction
between iron and uranium components. There is no strong evidence for the presence of a
discrete iron oxide phase, although the low frequency Raman feature at 322 cm-1 is centred in
the region of the most intense features in the spectra of Fe2O3, and bands from UO3 are not
generally observed in this region [12]. The presence of iron oxide is a tentative suggestion based
14

on the Raman studies, and the absence of any X-ray diffracting domains indicated that the iron
oxide is either amorphous or it has a small crystallite size.
Table 6 shows the comparison between the Raman spectra of the U/V oxide catalysts and V2O5
reference material. Both Raman spectra were similar and the positions of the bands were within
1-2 wavenumbers of each other. Bulk U3O8 was also identified by XRD, however, this phase
was not identified from the Raman analysis. U3O8 has an intense dark green/black colour and
Raman spectroscopy is notoriously difficult due to black body effects. The most intense Raman
bands from U3O8 are at 738 and 811 cm-1 and are weak compared to those from V2O5. The
similarity of the Raman data between the U/V oxide catalysts and V2O5 indicates that there is
little interaction between the separate uranium and vanadium oxide phases.
The U/W oxide catalysts showed features corresponding to a WO3 phase, these were clearly
visible at 806, 717 and 272 cm-1 (Table 7), and are consistent with XRD data. Powder XRD
also indicated that the catalyst contained a crystalline uranyl tungstate phase (UO2WO4), and
the Raman bands at 953 and 994 cm-1 are characteristic of the vibrational stretching modes of
W=O in the WO42- unit. Furthermore, the intense bands in the range 751-773 cm-1 correspond
to asymmetric stretching vibrations of W-O-W from the WO42- unit.
Temperature programmed reduction studies were also useful in confirming the structure of the
catalysts and revealed additional information that was not evident from XRD and Raman
studies. For the TPR of the U/Mo catalyst the peak at 614 C was due to the reduction of U6+
in the uranyl ion to U4+, the equivalent U6+ to U4+peak was in the reduction profile of UO3 to
UO2 at 690 C [19]. It appears that the reduction of uranium in the U/Mo catalyst is more facile
than in the single oxide, due to modification of the U6+ ion. The reduction features at 653 and
780 C were most likely due to reduction of Mo6+ to Mo4+ and Mo4+ to Mo0 respectively. The
reduction of molybdenum from Mo6+ in MoO3 is generally observed at higher temperatures
than those observed for the uranyl molybdate phase from this study [14, [20]. It is therefore
15

clear that the formation of the mixed oxide influences the redox behaviour of the uranium and
molybdenum ions.
The formation of the Fe/Mo mixed oxide also facilitated the reduction process compared to the
single component oxides. The mechanism of reduction for the Fe/Mo mixed oxide has been
investigated in detail previously [21],[22]. The four peaks in the TPR corresponded well with
data in previous studies, and the absence of any peaks in the regions 320-350 C and 515-590
C indicated the absence of any single iron oxide species, such as FeO, Fe2O3 and Fe3O4. The
lowest temperature peak in the TPR profile was due to the reduction of Fe2(MoO4)3 (ferric
molybdate), to FeMoO4, (ferrous molybdate) and leads to the formation of MoO3 [21]. The
features at 767 and 831 C corresponded to the reduction of Mo6+ to Mo4+ in the two different
environments of MoO3 and MoO42- [21]. The combination of iron and molybdenum oxide in
the mixed oxide phase retarded the reduction of Fe3+ whilst facilitating the reduction of Mo6+
when the redox behaviour was compared to the component oxides. These data reinforce the
conclusions from XRD and Raman studies, indicating the formation of a pure molybdate phase.
Characterisation of the Fe/U oxide catalyst by XRD and laser Raman spectroscopy concluded
that uranium oxide (UO3) was the only identifiable phase, with no obvious evidence for the
formation of a mixed oxide or iron oxide phase. TPR studies of iron oxide, Fe2O3, showed a
major reduction feature at 586 C, with a much smaller reduction feature at 350 C, whilst
Fe3O4 showed a major reduction feature at 655 C for the oxide alone. Furthermore, studies of
the reduction of UO3 produced a broad reduction feature at 690 C, in a similar temperature
range expected for the iron oxides. It is difficult to unequivocally assign the reduction peaks of
the Fe/U catalyst, however, the peak at 489 C is assigned to the reduction of UO3. The
reduction temperature is significantly lower than bulk UO3 and it is clear that the presence of
iron modifies the redox properties of UO3. Further evidence to support this assignment is
provided by the relative areas of the reduction profiles, uranium oxide is in a large excess
16

compared with the iron component and hence the reduction feature at 489 C is attributed to
the uranium component. We have assigned the reduction peak at 576 C to the reduction of
FeOX, this is close in temperature to the expected reduction for Fe2O3 at 586 C. The smaller
and lower temperature reduction peak expected at ca. 350 C for Fe2O3 is not clearly visible,
although it may be expected that the large reduction feature at 489 C may interfere. Close
analysis of the leading edge of the feature at 489 C indicates that a further reduction feature is
present at ca. 375 C and leads into the main reduction peak. The assignment to FeOX is
consistent with the suggestion from the laser Raman data. The definite assignment of these
reduction features cannot be certain, and it is possible that the order may be reversed. The co-
precipitation procedure used to prepare the Fe/U catalysts resulted in an iron oxide phase with
a highly dispersed small crystallite size. This would explain why the iron oxide phase was
amorphous to X-rays and the increased indeterminate scattering capacity of small crystallites
may also affect the Raman spectroscopy. However, in the case of the uranium and iron
components the mixture of FeOX and UO3 facilitates the reduction of the oxides, and this is
prevalent for UO3 in particular.
Catalyst activity and by-product formation
The U/Mo, Fe/Mo and U/W oxide catalysts showed selective toluene oxidation activity and
exhibited appreciable yields of benzaldehyde. It is interesting to note that all the selective
catalysts contained either a molybdate or tungstate phase. Evidence from XRD, laser Raman
and TPR studies indicated that the U/Mo and Fe/Mo catalysts consisted solely of the
UO2(MoO4)2 and Fe2(MoO4)3 phases. The U/W oxide catalyst was predominantly the UO2WO4
phase with some WO3 identified by XRD. The best catalyst for selective toluene oxidation was
Fe/Mo. In the case of the Fe/Mo catalyst the presence of Fe in the molybdate phase was clearly
beneficial. The Fe3+ ion can be considered as a Lewis acid site and it can aid the adsorption of
17

the aromatic ring onto the catalyst surface [23]. The subsequent abstraction of a hydrogen atom
from the toluene molecule by a basic site, could either occur by O2- present as Mo=O or Mo-
O-Mo bridging species. It is interesting to note that a similar catalyst has also been used
previously for the selective oxidation of toluene [17]. However, direct comparison of catalyst
performance cannot be made because previous studies by Kuang et al. have tested catalysts
under very different conditions. For instance, catalytic tests were performed in the absence of
molecular oxygen and in a pulse microreactor. The best yield of benzaldehyde obtained was
23.2 mol g-1 for a Fe/Mo atomic ratio of 1:1. The results in this study, and the results of Kuang
et al., demonstrate that iron molybdates have an efficacy for selective toluene oxidation to
benzaldehyde.
Although these studies were primarily concerned with the selective oxidation of toluene to
benzaldehyde, it is also important to consider the influence of reaction by-products. The Fe/Mo
catalysts exhibited a decrease in conversion as the temperature was increased and this was
indicative of deactivation by blocking of the catalyst surface. In the reactor studies it was
concluded that when the reaction temperature was reduced quickly from a catalytically active
temperature the catalyst bed became solidified. This was most likely due to heavier molecular
weight species remaining on the catalyst surface. GC-MS analysis of the material extracted
from the solidified catalyst bed indicated that products due to coupling reactions were formed,
and included phthalic anhydride and methyl diphenylmethane. These were in addition to the
more common benzaldehyde, benzoyl peroxide and benzoic acid products, and formation of
these types of products have been reported previously [2,3,16]. Due to the deposition of the
solid reaction products throughout the catalyst bed and the reactor it was not possible to quantify
them. However, it was evident that the quantities of the solid reaction products deposited were
much lower for U/Mo and U/W catalysts compared to Fe/Mo. Catalyst deactivation was not
observed with the more unselective U/Mo and U/W oxidation catalysts, however, the yields of
18

benzaldehyde were lower, and were produced at higher temperatures, therefore, potential
precursors for by-product formation were only present at low concentrations and the higher
temperature could aid the desorption of poisoning species from the catalyst surface.
In order to study the stability of benzaldehyde under reaction conditions and to probe the
reaction pathway during toluene oxidation, benzaldehyde oxidation was attempted using the
range of catalysts under equivalent conditions. Substantial amounts of solid products were
always formed during reaction, and products were identified by NMR spectroscopy. Only one
product was identified and this was benzoyl peroxide. According to these results it is likely
that during toluene oxidation the primary benzaldehyde formed was subsequently oxidised
rapidly to form benzoyl peroxide. Further studies on the partial oxidation of benzaldehyde were
attempted in order to study the mechanism of partial oxidation, but unfortunately reproducible
results could not be obtained, despite many efforts. This was due to the production of large
quantities of solid material, which continually blocked the reactor and deactivated the catalyst
under our reaction conditions. Previously, the study of benzaldehyde oxidation has not been
widely investigated and this may be due to similar reasons of experimental difficulty. However,
Martin et al. [3] observed that when using a vanadyl pyrophosphate catalyst, benzaldehyde was
strongly adsorbed on the catalyst surface, leading to the formation of cyclic anhydrides and
degradation products of the aromatic ring. Various studies have reported the formation of
benzoic acid from toluene oxidation [24], [25], and although we did not detect it as a gaseous
product it was detected as a solid product in the reactor. According to the results discussed
above, and the literature studies, a possible scheme for toluene oxidation and reaction products
can be proposed (Scheme 1).
It was clear from the results that coupling and more easily condensable products to form solids,
such as benzoyl peroxide were only formed during toluene oxidation when benzaldehyde was
observed, indicating that they were either formed by secondary reactions of a benzaldehyde
19

intermediate or from a common surface intermediate. It appears most likely that a benzoate
species could be formed as a reaction intermediate. This intermediate could lead either to the
formation of benzoic acid, or to the formation of benzoyl peroxide through the reaction of two
benzoate groups. The latter reaction was predominant in our benzaldehyde oxidation studies,
and it is reasonable to assume that a similar mechanism also operates for benzaldehyde formed
by toluene oxidation.
The oxide catalysts U/Sb, Fe/U and U/V did not exhibit any toluene selective oxidation activity.
The Fe/U and U/V catalysts were active at the lowest temperatures and showed high selectivity
towards carbon dioxide. The higher activity and absence of selective oxidation products can be
related to the structural phases of these catalysts. Powder X-ray diffraction studies of the U/Fe
oxide catalysts identified only UO3, whilst this phase was also established by laser Raman
spectroscopy, which suggested the presence of iron oxide, and this was confirmed by TPR. It
has been shown that uranium oxides are, in general, very active for deep oxidation of a wide
range of organic compounds to COX [26]. Furthermore, the characterisation studies are
consistent with a dispersed iron oxide phase and it is recognised that catalysts of this type are
also active for total oxidation reactions [27]. The combination of these components would
account for the observation of toluene total oxidation.
Similarly characterisation studies also showed that the U/V oxide catalyst was comprised from
a mixture of the single oxides U3O8 and V2O5. Investigation of toluene oxidation using the same
conditions employed to test the mixed oxide catalysts demonstrated that U3O8 produced 88%
conversion to CO2 at 300C. Previously the total oxidation activity of U3O8 has been well
documented, and using a considerably greater GHSV (70,000 h-1) toluene was readily oxidised
to COX at temperatures above 350 C [28]. It was ther efore evident that in a similar manner to
the Fe/U catalyst the mix of U3O8 and V2O5 single oxides were not suitable for selective toluene
20

oxidation. Again no solid reaction products similar to those detailed above were produced, and
this indicated that toluene was oxidised directly to CO and CO2.
Relationship between structure and activity
It is clear that in this study the catalysts can be grouped into two classes, those, which are
selective for toluene oxidation, such as U/Mo, Fe/Mo and U/W, and those, which only gave
combustion products, such as U/Sb. U/V and Fe/U. The reactivity of the range of catalysts has
also been studied for selective oxidation of saturated and unsaturated C3 hydrocarbons [29]. In
these previous studies Fe/Mo and U/Mo mixed oxide catalysts showing iron and uranium
molybdate phases were relatively inactive for propane and propene oxidation. Outstanding
results for the selective oxidation of propane and propene to formaldehyde were obtained using
Fe/U catalysts prepared with ratios of 0.5/3 and 1/3. At 450 C selectivity towards formaldehyde
of 44% was achieved at 42% propane conversion. The mixed U/Sb catalyst, with an USb3O10
structure, was inactive for propane oxidation, but demonstrated the expected high selective
oxidation activity to form acrolein from propene [9, 10, 11]. The results of propane/propene
oxidation show the reverse trend to toluene oxidation, that is catalysts; that are selective for C3
oxidation are unselective for toluene and vice versa. It can therefore be concluded that catalyst
features required for selective oxidation of aromatic compounds are very different from those
needed for selective oxidation of saturated and linear unsaturated hydrocarbons.
5. Conclusions
Selective gas phase toluene oxidation to benzaldehyde was achieved using Fe/Mo, U/Mo and
U/W mixed oxide catalysts; Fe/Mo was the most active and selective. The activity of Fe/U,
Sb/U and V/U oxide catalysts produced only deep oxidation to carbon oxides. A range of by-
products was observed, these were other partial oxidation and coupling products. By-product
21

formation was particularly apparent for catalysts with high benzaldehyde yields. Preliminary
studies of benzaldehyde oxidation suggest that they were derived from benzaldehyde,
particularly for the production of benzyl peroxide.
Catalysts have been characterised by X-ray diffraction, Laser Raman spectroscopy and
temperature programmed reduction. Pure molybdate phases were identified for the F e/Mo and
U/Mo catalysts, and there was no evidence for the formation of single oxide phases. Data
suggests that the formation of the molybdate ion is important for the selective oxidation of
toluene to benzaldehyde. In contrast, the U/Fe catalyst was a mixture of U3O8 and V2O5, whilst
the Fe/U catalyst was comprised of highly dispersed iron oxide on UO3. The presence of the
single oxides was responsible for toluene total oxidation. The U/Sb catalyst did not exhibit
selective toluene oxidation, but previous studies have demonstrated that the catalyst exhibits
high activity for selective propene oxidation. Similar behaviour has been observed for the other
catalysts in this study, and it is apparent that catalysts that were selective for toluene oxidation
were not selective for propene oxidation and vice versa.
Acknowledgements
The authors would like to thank Darren Lee at BNFL (Springfields, UK) for his assistance with
this work and BNFL for financial support.
References
[1] L. Kesavan, R. Tiruvalam, M.H. Ab Rahim, M.I. bin Saiman, D.I. Enache, R.L. Jenkins,
N. Dimitratos, J.A. Lopez-Sanchez, S.H. Taylor, D.W. Knight, C.J. Kiely, G.J.
Hutchings, Science, 331 (2011), pp 195-199.
22

[2] W. Partenheimer, Catal. Today, 23 (1995), pp 69-158.
[3] J. Zhu, S.L.T Anderson, J. Chem. Soc. Faraday Trans., 85 (1989), 3629-3644.
[4] M. Ai, Heterogeneous Catalysis and Fine Chemicals II”, Elsevier Science Publishers
B.V., Amsterdam, (1991).
[5] H.K. Matralis, C.H. Papadopoulou, C. Kordulis, A.A. Elguezabal, V.C. Coberan, Appl.
Catal. A, 126 (1995), pp 365-380.
[6] A. Martin, U. Bentrup, A. Brückner, C. Lück, Catal. Lett., 59 (1999), pp 61-65.
[7] S.H. Taylor, Heterogeneous catalysis by uranium oxides, Metal Oxide Catalysts: Volume
2, (Eds. S.D. Jackson, J.S.J. Hargreaves), Wiley VCH, (2008), Chapter 13, pp 539-560.
[8] C.J. Machiels, A.W. Sleight, Proceedings of the 4th International Conference on Chemistry
and Uses of Molybdenum, Climax. Molybdenum Co. Ann Arbor, Mich., (1982), p 411.
[9] R.K. Grasselli, J.L. Callahan, J. Catal., 14 (1969), pp 93-103.
[10] R.K. Grasselli, D.D. Suresh, K. Knox, J. Catal., 18 (1970), pp 356-358.
[11] R.K. Grasselli, D.D. Suresh, J. Catal., 25 (1972), pp 273-291.
[12] S.E. Golunski, T.G. Nevell, D.J. Hucknall, J. Catal., 88 (1984), pp 448-456.
[13] T.E. Albrecht-Schmitt, R.E. Sykora, J.E. King, A.J. Illies, J. Solid State Chem., 177
(2004), pp 1717-1722.
[14] K. Otsuka, Y. Wang, I. Yamanaka, A. Morikawa, J. Chem. Soc. Faraday Trans., 89 (1993),
pp 4225-4230.
[15] M. Sundberg, B.D. Marinde, J. Solid State Chem., 121 (1996), pp 167-173.
[16] D.S. Kim, K. Segawa, T. Soeya, I.E. Wachs, J. Catal., 136 (1992), pp 539-553.
[17] N-L. Palacios, S.H. Taylor, Appl. Spectros., 54 (2000), pp 1372-1378.
[18] K. Aykan, A.W. Sleight, J. Am. Ceram. Soc., 53 (1970), pp 427-431.
[19] M-L. Palacios, Selective Oxidation Catalysts Based on Uranium Oxides, PhD Thesis,
Cardiff University, (2000).
[20] P. Arnoldy, C. de Jonge, J.A. Mouljin, J. Phys. Chem., 89 (1985), pp 4517-4526.
23

[21] S. Lars, T. Anderson, J. Catal., 98 (1986), pp 138-149.
[22] W. Kuang, Y. Fan, K. Chen, Y. Chen, J. Chem. Res., 0 (1997), 366-367.
[23] A. Borgna, J. Sepúlveda, S.I. Magni, C.R. Apesteguia, Appl. Catal. A, 276 (2004), pp
207-215.
[24] J. Miki, Y. Osada, T. Konoshi, Y. Tachibana, T. Shikada, Appl. Catal. A, 137 (1996), pp
93-104.
[25] A.J. van Hengstum, J. Pranger, S.M. van Hengstum-Nijhuis, J.G. van Ommen, P.J.
Gellings, J. Catal., 101 (1986), pp 323-330.
[26] G.J. Hutchings, C.S. Heneghan, I.D. Hudson, S.H. Taylor, ACS Symposium Series 638,
Heterogeneous Hydrocarbon Oxidation, (Ed. B.K. Warren, S.T. Oyama), ACS Books
Washington D.C., (1996), pp 58-75.
[27] S. Betteridge, C.R.A. Catlow, D.H. Gay, R.W. Grimes, J.S.J. Hargreaves, G.J.
Hutchings, R.W. Joyner, Q.A. Pankhurst, S.H. Taylor, Topics Catal., 1 (1995), pp 103-
110,
[28] G.J. Hutchings, C.S. Heneghan, I.D. Hudson, S.H. Taylor, Catal. Today, 59 (2000), pp
249-259.
[29] S.H. Taylor, G.J. Hutchings, M-L. Palacios, D.F. Lee, Catal. Today, 81 (2003), pp 171-
178.
24

Table 1. Summary of catalysts BET Surface areas and phases identified from powder X-ray
diffraction studies.
Phases identified
SABET/m2 g-1
Catalyst
At. ratio
U/Mo
1/2
0.7
0.9
UO2(MoO4)2
Fe2(MoO4)3
2/3
1/3
Fe/Mo
U/Sb
25
USb3O10
Fe/U
0.5/3
1/3
18
116
UO3
UO3
U/V
1/3
1/3
0.8
1.5
U3O8, V2O5
U/W
UO2WO4, WO3
25

Table 2. Laser Raman spectroscopy summary and assignment of U/Mo oxide catalyst and comparison with molybdenum and uranium
reference compounds.
a
b
U/Mo oxide (NH4)6Mo7O24.
H2Oa
MoO3
UO2(AcO)2
UO2
Band assignment
-U3O8
-UO3
207 (S)
UO22+ bending
250, 224 (M)
290 (M), 283
(M), 245 (W),
217 (W)
 bending vibration Mo-O- Mo of the
Mo7O246-octahedral molybdate
347 (W)
343 (M)
351 (M)
U-O-IU sigma bending vibration
368 (W),
319 (W)
369-312 (W)
912-861 (M)
938 (VS)
 bending vibration Mo=O
of the Mo7O246-octahedral molybdate
U-O stretching vibration
450 (S)
412 (S)
483 (S)
740 (M)
761 (VS)
767 (S)
O-U-O-U symmetric stretching vibration
U-O stretching vibration
811 (S)
838 (M)
799 (M), 793
(M), 747 (VS)
886 (S)
665 (M)
817 (VS)
994 (S)
 asymmetric Mo-O-Mo stretching
vibration
881 (VS)
UO22+ symmetric stretching
 symmetric stretching vibration Mo-O-
Mo
955 (M)
 symmetric stretching vibration Mo=O
All frequencies in cm-1. (VS) very strong, (S) strong, (M) medium, (W) weak
a [Error! Bookmark not defined.,Error! Bookmark not defined.] used to assist band assignment.
b [Error! Bookmark not defined.,Error! Bookmark not defined.,Error! Bookmark not defined.] used to assist band assignment.
26

Table 3. Laser Raman spectroscopy summary and assignment of Fe/Mo oxide catalyst and comparison with iron
oxide reference compound.
Band assignment
Fe/Mo oxide
Fe2O3
289 (S), 220 (W)
bending vibration Fe-O-Fe
 bending vibration Mo=O of the MoO42- tetrahedral molybdate
372, 362, 349,
300, 258 (M)
 bending vibration Mo-O- Mo of the Mo7O246-octahedral molybdate
asymmetric stretching vibration Fe-O-Fe
405 (M)
603 (W)
symmetric stretching vibration Fe-O-Fe
783 (VS)
821 (M)
 asymmetric Mo-O-Mo stretching vibration
 asymmetric stretching vibration Mo=O of the MoO42- tetrahedral molybd
967 (M),
937 (W)
 symmetric stretching vibration Mo=O
of the MoO42- tetrahedral molybdate
991 (W)
 symmetric stretching vibration Mo=O (MoO3 crystalline phase)
All frequencies in cm-1. (VS)= very strong, (S)=strong, (M)=medium, (W)=weak.
27

Table 4. Laser Raman spectroscopy summary and assignment of U/Sb oxide catalyst and
comparison with antimony oxide reference compounds.
Band assignment
U/Sb oxide
Sb2O3
Sb2O4
205 (M), 304
(W)
259 (VS), 194
(S),
266 (M), 203
(VS), 147 (W)
 bending vibration Sb-O-Sb
(U-O-U)
124 (W)
345 (W), 401
(M), 464 (W),
491 (S), 560
(W)
 symmetric + asymmetric stretching vibration
U-OI,IV, SbI-OI, SbIIOI, OIII, OIV
455 (VS), 378
(S), 361 (W)
651 (M), 618
(VW), 467 (M),
421 (Sh), 405 (S)
828 (VW),
 asymmetric stretching vibration Sb-O-Sb
 symmetric stretching vibration Sb-O-Sb
718 (M)
755 (VW),
720(VW)
776 (M), 824
(W), 905 (M)
603 (W)
 symmetric + asymmetric stretching vibration
U-OIII, SbI-OII, SbIIOII
All frequencies in cm-1. (VS)= very strong, (S)=strong, (M)=medium, (W)=weak.
28

Table 5. Laser Raman spectroscopy summary and assignment of Fe/U oxide catalyst.
Band assignment
Fe/U oxide
asymmetric stretching vibration Fe-O-Fe
332 (M)
477 (W)
762 (S)
821 (W)
asymmetric stretching vibration Fe-O-Fe
U-O stretching vibration
U-O stretching vibration
U-O stretching vibration
All frequencies in cm-1. (VS)= very strong, (S)=strong, (M)=medium, (W)=weak.
29