Application of Al2O3/AlNbO4 in the oxidation of aniline to azoxybenzene

Article abstract of DOI:10.1007/s11696-019-00897-5

Al₂O₃/AlNbO₄ powder was fabricated by a facile high-energy milling process. The precursor materials, Al₂O₃ and Nb₂O₅, are readily available and have very attractive properties. Moreover, the catalytic activity of the sample in the liquid phase oxidation of aniline (OA) in the presence of hydrogen peroxide as oxidant was evaluated. The catalyst was found to be highly efficient and selective in the oxidation of aniline to azoxybenzene under mild conditions. When mixed with 28% AlNbO₄ the alumina-based catalyst achieved high conversion and selectivity and very similar to the pure Nb₂O₅.

Full text of DOI:10.1007/s11696-019-00897-5

Chemical Papers
https://doi.org/10.1007/s11696-019-00897-5
ORIGINAL PAPER
Application of Al O /AlNbO in the oxidation of aniline
2
3
4
to azoxybenzene
1
1
1
1
2
Daniel C. Batalha · Sulusmon C. Luz · Jason G. Taylor · Humberto V. Fajardo · Bruno S. Noremberg ·
2
2
2
3
4
Igor J. S. Cherubin · Ricardo M. Silva · Margarete R. F. Gonçalves · Carlos P. Bergmann · Antoninho Valentini ·
Neftalí L. V. Carreño2
Received: 9 October 2018 / Accepted: 7 August 2019
©
Institute of Chemistry, Slovak Academy of Sciences 2019
Abstract
Al O /AlNbO powder was fabricated by a facile high-energy milling process. The precursor materials, Al O and Nb O ,
2
3
4
2
3
2
5
are readily available and have very attractive properties. Moreover, the catalytic activity of the sample in the liquid phase
oxidation of aniline (OA) in the presence of hydrogen peroxide as oxidant was evaluated. The catalyst was found to be highly
eꢀcient and selective in the oxidation of aniline to azoxybenzene under mild conditions. When mixed with 28% AlNbO the
4
alumina-based catalyst achieved high conversion and selectivity and very similar to the pure Nb O .
2
5
Keywords Heterogeneous catalysis · Aniline oxidation · Azoxybenzene · Simple high-energy milling · Niobium
Introduction
(Acharyya et al. 2014; Shukla et al. 2016; Qadir et al. 2015;
Ghosh et al. 2015; Gontier and Tuel 1995; Jagtap and
Ramaswamy 2006). Thus, there is a demand for heteroge-
neous catalysts capable of selectively converting aniline to
azoxybenzene in a tandem oxidation–condensation reaction
(Chang and Liu 2009; Trautwein et al. 2015). Some hetero-
geneous catalytic systems developed for the oxidation of
aniline have been reported, including CuCr O spinel nano-
Azoxybenzenes are valuable compounds widely applied in
the industry as dyes, pigments, pharmaceuticals, polymeri-
zation inhibitors, chemical stabilizers, and food additives.
One possible route to obtaining azoxybenzenes is by a tan-
dem oxidation–condensation reaction. This is a challenging
reaction for heterogeneous catalysis and achieving a highly
selective process is hampered by competitive oxidation and
condensation reactions that lead to the formation of phe-
nylhydroxylamine, nitrosobenzene, nitrobenzene, azoben-
zene, and azoxybenzene as possible products (Scheme 1)
2
4
particles (Acharyya et al. 2014), Cu-nanoparticles supported
on nanocrystalline CeO (Cu–CeO ) (Shukla et al. 2016),
2
2
titanium dioxide (P25) (Qadir et al. 2015) silver nanoparti-
cles supported on nanostructured tungsten oxide (Ag/WO )
3
(
(
Ghosh et al. 2015) cobalt-containing mesoporous silica
Chang and Liu 2009), titanium (IV) oxide (Tumma et al.
*
Neftalí L. V. Carreño
neftali@ufpel.edu.br
2009), titanium silicate molecular sieves (TS-1 and TS-2)
(Selvam and Ramaswamy 1995), titanium-loaded MCM-48
(
Das and Talukdar 2017) and LiNbO (Carreno et al. 2018).
3
1
Department of Chemistry, Institute of Exact and Biological
Sciences, Federal University of Ouro Preto, Ouro Preto,
MG 35400-000, Brazil
However, despite these important results, current catalytic
protocols have yet to achieve maximum yield and selectiv-
ity towards azoxybenzene, highlighting a demand for more
eꢁective methods.
2
3
4
Graduate Program in Materials Science and Engineering,
Technology Development Center, Federal University
of Pelotas, Pelotas, RS 96010-610, Brazil
Although the use of niobium base materials in electro-
chemical applications as anode materials is well known
(Wang et al. 2016; Xie et al. 2016), its versatility motivated
this study to explore potential applications in catalysis, espe-
cially with the advancements in knowledge regarding the
capacity of the material to enhance the catalytic activity of
Graduate Program in Mining, Metallurgical and Materials
Engineering, Department of Materials, Federal University
of Rio Grande do Sul, Porto Alegre, RS 90040-060, Brazil
Department of Analytical Chemistry and Physical Chemistry,
Federal University of Ceara, Campus Pici, Fortaleza,
CE 60455760, Brazil
Vol.:(0
1
123456789)
3

Chemical Papers
Scheme 1 Illustrative chemical
design showing the probable
products of OA
known catalysts (Tanabe 2003). Not only is niobium a read-
ily available commercial material, but niobium oxide and
aluminum oxide are useful precursors for the synthesis of
aluminum niobate using a solid-state reaction, which could
be applied to novel catalytic applications. Thus far, the appli-
cation of Al O /AlNbO as a heterogeneous catalyst for the
for 3 h. Afterward, the material was milled again, however,
using a ball mill (Quimis Q298), and sifted after milling
using a sieve #200. Lastly, the material was sintered again,
in a furnace at 1200 °C aiming to reduce the porosity and
allow the particle size growth.
2
3
4
oxidation of aniline in the presence of H O as oxidant had
2
2
Characterization techniques
not been reported. Given that these metal oxides have known
acidic properties, they were chosen for the investigation of
their potential application in the present reaction. We pre-
sent herein, a simple preparation of powder composed by
aluminum and niobium oxides by the high-energy milling
process, and their application as an active catalyst for the
conversion of aniline to azoxybenzene.
The crystalline structure of Al O /AlNbO powder was
2
3
4
characterized by X-ray diꢁraction (Shimadzu XRD 6000)
using diꢁractometer with CuKa radiation CuKα radiation
(
λ=1.54056 Å) in the scan range of 10°–60°. In addition,
crystallite size was calculated through the most intense
peak, using a simple expression from X-ray diꢁraction peaks
known as the Scherrer equation:
Experimental
K ⋅ ꢀ
L =
,
(
1)
ꢁ
⋅ cosꢂ
Catalyst preparation
where λ is the X-ray wavelength in nanometer (nm), β is the
peak full width at half maximum height and K is a constant
related to crystallite shape, normally taken as 0.9. The value
of β in the 2θ axis of diꢁraction proꢂle must be in radians
(Monshi et al. 2012).
To obtain Al O /AlNbO , a weight ratio of 80:20 between
2
3
4
Al O (ALCOA) and Nb O (CBMM), were placed in
2
3
2
5
a high-energy mill attritor type (Netzsch PE 075-60 Hz),
using 1-mm-diameter zirconia as grinding medium and
ethanol (SYNTH) as a process control agent. The ball to
powder ratio used was 10:1. Diꢁerent grinding times were
employed. After milling, the material was dried in an oven at
The particle size analyses were performed using a laser
diꢁractometer (CILAS 1064). The morphologies of Al O /
2
3
AlNbO powder were characterized by ꢂeld emission scan-
4
7
0 °C for 48 h and then was sintered in a furnace at 1400 °C
ning electron microscopy (FESEM; JEOL, model 7500F),
1
3

Chemical Papers
equipped with energy-dispersive spectroscopy (EDS), which
was used to investigate the elementary composition of the
particles.
of saturated sodium bicarbonate solution, transferred to a
separatory funnel and extracted with CHCl (2 × 10 mL).
3
The organic layers were combined, dried with anhydrous
The number of acid sites of the prepared sample and the
precursor materials was determined by the titration method,
follow the process in (Ventura et al. 2017). In the experi-
ments, 20 mL of a sodium hydroxide aqueous solution
Na SO , ꢂltered and transfer to a volumetric ꢃask (25 mL)
2
4
with CHCl . Quantiꢂcation was performed on Gas Chroma-
3
tograph (Shimadzu, GC-2010 Plus Tracera) equipped with
SH-Rtx-5 capillary column and barrier discharge ioniza-
tion (BID) detector. Aniline conversion and product ratios
were determined by external calibration curves obtained by
known concentrations of authentic samples. The obtained
products were also confirmed by GC–MS (Shimadzu,
GCMS-QP2010).
−
1
of known concentration (0.01 mol L ) was added to the
powder (using a mass of 10 mg). The mixture was stirred
during 24 h at room temperature. Upon separation by cen-
trifugation, the supernatant solution was then titrated with
−
1
hydrochloric acid (0.05 mol L ) aqueous solution using
phenolphthalein as an indicator. Posteriorly, acid sites quan-
tity within the catalysts was determined by the following
equation:
Conversion of aniline was calculated by:
moles of reactant reacted
Conversion (% ) =
× 100,
initial moles of reactant used
(
3)
NAS = (NNaOH −NHCl) × L,
(2)
whereas selectivity of products using the following
expression:
where NAS=number of acid sites, NNaOH=initial quantity
of NaOH added (moles), NHCl=quantity of HCl consumed
(
moles) and L is the Avogadro constant. The relative acid
Selectivity (%)
strength of the samples was determined by temperature-
programmed desorption of pyridine (TPD-Py) experiments.
Prior to TPD, the samples were heat treated under He ꢃow
total moles of product formed
=
× 100.
the sum of total moles of all oxidation products formed
(4)
−
1
(
20 mL min ) at 700 °C for 30 min and cooled under the
Results and discussion
same gas ꢃow to 40 °C. Pyridine adsorption was carried out
using a ꢃow of He containing pyridine vapor (20 mmHg)
through the reactor for 30 min at 40 °C. After baseline stabi-
The inꢃuence of milling time on mean particle size was
monitored (Table 1) and sizes of 10, 50 and 90% of the
particles (D10, D50, and D90, respectively) were obtained.
After 120 min, the main particles sizes were not aꢁected and
remained stable at around 14 μm.
−
1
lization under He ꢃow (20 mL min ) at room temperature,
the Py-TPD run was carried out with a heating rate of 10 °C
−
1
min from room temperature up to 750 °C. The pyridine
desorption was monitored using a thermal conductivity
detector (TCD).
Figure 1a shows the X-ray spectrum of the powder sam-
ple after milling and calcination process. The peaks are
intense and sharp, which can be correlated with its crystal-
linity. The higher peak counts are indexed according to the
JCPDS card no. 41-0347 associated with Al O . The peaks
N2-adsorption/desorption isotherms were obtained at liq-
uid nitrogen temperature (− 196 °C). Before the analysis, the
degassing was carried out at 200 °C for 2 h under vacuum.
The Brunauer–Emmett–Teller (BET) equation was applied
to the results to calculate surface area.
2
3
in smaller amounts are indexed according to the JCPDS card
no. 46-1212 associated with AlNbO . It may conꢂrm the
4
The H temperature programmed reduction (H -TPR)
2
2
ꢂnal product mixture which has Al O and AlNbO with
2
3
4
measurement was carried out in a quartz reactor heating
~
72% and~28%, respectively. The mean crystallite size is
from 45 to 990 °C using an 8% H /N gas mixture ꢃow
2
2
−
1
−1
(
20 mL min ) at a heating rate of 10 °C min . Before
the H -TPR tests, the sample (300 mg) was heat-treating at
2
Table 1 Particle size distribution as a function of high-energy milling
2
00 °C under N ꢃow for 0.5 h. The H consumption was
2 2
time
monitored by a thermal conductivity detector (TCD).
The liquid phase oxidation of aniline was performed
according Ventura et al. (2017) and Carreno et al. (2018).
In a one-necked round bottom ꢃask (10 mL) equipped with
a stir bar, the following were added: 10 mg of catalyst,
Time (min)
Mean diam- D10 (μm)
eter (μm)
D50 (μm)
D90 (μm)
0
19.68
17.95
14.53
14.42
13.59
13.74
14.33
1.36
1.14
1.06
1.01
0.91
0.99
0.96
17.07
14.61
12.24
11.64
10.57
11.35
10.84
43.1
40.6
31.9
32.5
31.5
30.6
34.2
3
6
9
1
1
1
0
0
3
mL of the speciꢂc organic solvent, 0.10 mL of aniline and
0
ꢂnally 0.15–1.0 mL of H O (35% aq. solution). The reac-
2
2
20
50
80
tion was maintained at 25 °C using a hotplate stirrer with
oil bath for times varying from 3 to 48 h. Once the reaction
was completed, it was quenched by the addition of 10 mL
1
3

Chemical Papers
Fig. 1 a XRD pattern; b N
2
adsorption–desorption iso-
therms and c micrograph of the
Al O /Nb O after gridding d
2
3
2
5
micrograph of the catalyst
5
3.21 nm, measured using Eq. 1. Figure 1b shows the iso-
therm from the catalyst, which presents a proꢂle of type II
isotherm with a small type H3 loop (Thommes et al. 2015)
indicating a nonporous and low surface area material (Sing
1
985). The SEM micrographs of the catalyst powder are
shown in diꢁerent stages, after milling (Fig. 1c) and calci-
nation process (Fig. 1d). They conꢂrmed the presence of
two diꢁerent phases observed by the X-ray diꢁraction and
permitted the identiꢂcation of two distinct morphologies,
spherical and rectangular.
The analysis of EDS, Fig. 2, showed that the rectangular
particles are composed of Al, Nb, and O, and the spherical
particles composed of Al and O, indicating the Al O com-
2
3
position for spherical particles and AlNbO composition for
4
rectangular particles.
The number of acid sites found for the prepared Al O /
2
3
AlNbO sample as well as the precursor materials (Nb O
4
2
5
and Al O ) was determined by titration in aqueous solutions
2
3
1
9
+
(
Table 2). The Nb O sample had 114,4 × 10 H sites/g,
2 5
which was the highest value among all, indicating that this
sample has a higher acidity on its surface (De Oliveira et al.
2
014). Al O /AlNbO sample presented an intermediate
2 3 4
value, due to the presence of Al O , which is the less acidic
Fig. 2 FESEM image with EDS spectrum of Al2O3/AlNbO4
2
3
sample.
In Fig. 3a, the temperature-programmed desorption spec-
tra of pyridine on the materials are depicted. The pyridine
desorption curves (TPD) show that the Nb O sample is
540 °C. The Al O sample, however, has two desorption
2
3
peaks at low temperature (120 and 195 °C), followed by two
high-temperature peaks (610 and 700 °C). However, unlike
Nb O sample, it is not observed the predominance of the
2
5
composed mainly of weak acid sites, with a peak centered
at 95 °C. This sample also shows two relatively low-inten-
sity pyridine desorption events at temperatures of 380 and
2
5
low temperature of pyridine desorption peaks for the Al O
2
3
1
3

Chemical Papers
Table 2 Acid sites amount determined by a titration method for the
lower energy for reduction. The Al O sample did not show
2 3
samples and N adsorption–desorption measurements
2
reduction peaks in the studied temperature range.
+
To optimize the catalytic activity of the obtained sample,
the catalytic reaction OA was carried out under diꢁerent
conditions. Initially, the reactions were conducted without
catalyst under the same conditions. No appreciable aniline
conversion was detected by gas-chromatography in this case.
The conversion of aniline and the selectivity of the reac-
tion catalyzed by Al O /AlNbO at 25 °C in acetonitrile as
Sample
Acidic sites (H sites/g)
Surface area
2
(
BET) (m /g)
1
1
1
9
9
9
Nb O5
114.4 × 10
102.4 × 10
111.4 × 10
0.6
1.5
0.5
2
Al O
2
3
Al O /AlNbO
4
2
3
2
3
4
solvent varied with the amount of H O (Fig. 4b) and reac-
sample. The pyridine TPD proꢂle for the Al O /AlNbO
4
2
2
2
3
tion time (Fig. 4a). The catalyst was very active for the oxida-
tion of aniline under mild conditions that employed the mini-
mum amount of hydrogen peroxide at ambient temperature.
Depending on the amount of oxidizing agent used, diꢁerent
conversions and selectivities were obtained. Catalytic tests,
for reactions carried out over 12 h revealed a tendency for an
increase in the aniline conversion in the presence of larger
sample shows a maximum near 300 °C, which diꢁerenti-
ates it from the other samples; followed by another maxi-
mum at 580 °C. This proꢂle shows that the formation of the
AlNbO phase leads to a change in the acidic properties of
4
the material surface. The desorption with the maximum at
5
80 °C can be attributed to pyridine adsorbed on the Al O
2 3
phase present in the sample composition, as pointed out in
Figs. 1a and 2.
H O quantities. The best result reached 82.1% with 1.2 mL
2
2
of oxidant, probably due to the larger number of active oxi-
dant species available for catalysis (Shukla et al. 2016). In
contrast, reaction times of 24 and 48 h resulted in complete
Figure 3b shows the H -TPR proꢂles. The reduction pro-
2
ꢂle of the Nb O sample shows two main H consumption
2
5
2
peaks with the maximum at 700 and 905 °C. The ꢂrst H
2
conversion of aniline and increasing the amount of H O
2
consumption peak can be assigned to a partial reduction of
the fraction of the material corresponding to the surface
layer of the particles. The second peak (905 °C) corresponds
to the partial reduction of the inner layers of the sample
2
was not signiꢂcantly beneꢂcial. Regarding the selectivity
of the reaction, Al O /AlNbO preferentially catalyzed the
2
3
4
formation of azoxybenzene, with some concomitant forma-
tion of nitrobenzene and nitrosobenzene in smaller amounts.
The formation of the minor products increased when hydro-
gen peroxide was present in large excess. It is interesting to
(
Yang et al. 2018; Rodrigus et al. 2012). The change in the
niobium oxide reduction proꢂle is perceptible for the Al O /
2
3
AlNbO sample, which presents a single well-deꢂned and
4
note that, after 12 h of reaction and with 0.2 mL of H O ,
low-intensity peak. Due to the presence of niobium in the
2
2
4
1
8.0% of aniline conversion was achieved and an almost
Al O /AlNbO , the absence of the reduction peaks at 700
2
3
4
:1 mixture of mainly azoxybenzene (49.5%) and nitroso-
and 900 °C points to strong interaction between the oxides.
benzene (36.9%) was observed, in addition to the appear-
ance of phenylhydroxylamine and nitrobenzene in smaller
amounts (8.4% and 5.3%, respectively). The selectivity of
the reaction could be slightly improved by utilizing greater
This result indicates the absence of the Nb O phase, which
2
5
is consistent with the result of XRD, which indicates the
presence of Al O and AlNbO phases (Fig. 1a). On the
2
3
4
other hand, the presence of H consumption peak at 400 °C,
2
amounts of H O , and in this case, the formation of phenyl-
temperature lower than the observed for the sample Nb O ,
2
2
2
5
hydroxylamine was no longer detected. However, there was
indicates the formation of a low fraction of sites that need
a considerable decrease in the formation of nitrosobenzene
Fig. 3 Py-TPD curves of Al O ,
2
3
Nb O , and Al O /AlNbO sam-
2
5
2
3
4
ples. a H -TPR curves of Nb O
5
2
3
2
and Al O /AlNbO samples (b)
2
4
1
3

Chemical Papers
Fig. 4 Aniline conversion (a)
and selectivity (b) for oxidation
products as a function of time
and volume of H O for Al O /
2
2
2
3
AlNbO employed as catalyst.
4
Reaction temperature=25 °C,
mass of catalyst=10 mg and
acetonitrile as solvent
and an increase in azoxybenzene which begins to appear as
the main product (Gontier and Tuel 1994).
contrast, azoxybenzene formation was favored in the more
polar solvents: acetone and DMF (in the range of 80–90%).
Tests were performed at 25, 40 and 50 °C and revealed
that catalytic activity remains quite high, with virtually com-
plete conversion of the aniline at all the tested temperatures
(Fig. 8a). Higher temperatures seem to accelerate the con-
densation between phenylhydroxylamine and nitrosobenzene
and rendering the reaction more selective towards azoxyben-
zene (Fig. 8b) (Selvam and Ramaswamy 1995).
Total conversion of aniline was achieved within 12 h for
all the reactions conditions evaluated (Fig. 5a). The complete
conversion could be attained after 3 h when large excesses
of the oxidizing agent were present and these conditions
also favored the formation of nitrosobenzene, with almost
1
00% selectivity (Fig. 5b) (Acharyya et al. 2014; Waghmode
et al. 2001).
Figure 6 shows the variation of aniline conversion
Each oxide precursor used in the preparation of Al O /
2
3
(
Fig. 6a) and selectivities (Fig. 6b) as a function of diꢁerent
AlNbO was evaluated for catalytic activity under the opti-
4
catalytic loadings at two main hydrogen peroxide volumes
mized conditions employing acetonitrile as solvent at 25 °C
(
0.4 mL and 1.2 mL). At the lowest catalytic loading tested,
for 24 h (Fig. 9a). Both Al O /AlNbO and Nb O promoted
2
3
4
2
5
the conversion was low, around 50%, and favored the for-
mation of azoxybenzene as the major product (Waghmode
et al. 2001).
the complete conversion of aniline whereas Al O was found
2 3
to be the less eꢀcient and only achieved 38.5% conversion of
aniline. Signiꢂcant diꢁerences can also be observed for the
selectivities of the reaction products, which were strongly
inꢃuenced by the nature of the catalyst. It is interesting to
note that Al O was a less selective catalyst towards the
It is well established that oxidation reaction in the liq-
uid phase can be inꢃuenced by the type of solvent used.
For this reason, three more solvents were evaluated: chlo-
roform, dimethylformamide (DMF) and acetone, maintain-
ing 1.2 mL of H O and the catalyst mass of 10 mg. The
2
3
formation of azoxybenzene (34.4%) and performed better
than the other catalysts in furnishing phenylhydroxylamine
(27.9%). The samples presented very low speciꢂc surface
area values. Comparing the speciꢂc surface areas of each
catalyst (Table 2), it could be seen that this property had no
inꢃuence on the catalytic activities presented in this study.
These results may be related to the high capacity of Al O /
2
2
catalytic activity decreased in the following order: acetoni-
trile>DMF~acetone>chloroform (Fig. 7b). Aniline con-
version was only sluggish when chloroform was used as a
solvent (58.1%) and all other solvents performed reasonably
well (Fig. 7a). Interestingly, the reaction was more selective
for the formation of nitrosobenzene in chloroform but in
2
3
AlNbO and Nb O in forming reactive oxygen, superoxo,
4
2
5
1
3

Chemical Papers
Fig. 5 Aniline conversion (a)
and selectivity (b) for oxidation
products as a function of reac-
tion time (3–48 h) and volume
of H O for Al O /AlNbO
2
2
2
3
4
employed as catalyst. Reaction
temperature=25 °C, mass of
catalyst=10 mg and acetonitrile
as solvent
Fig. 6 Aniline conversion (a)
and selectivity (b) for oxida-
tion products as a function
of catalyst mass and volume
of H O for Al O /AlNbO
2
2
2
3
4
employed as catalyst. Reaction
temperature=25 °C, reaction
time=24 h and acetonitrile as
solvent
1
3

Chemical Papers
2
017). In addition, the changes in the acidic properties,
observed by means of TPD using pyridine as a base, can
be correlated with the catalytic performance observed in
Fig. 8a. High-strength acid sites may adsorb reagents irre-
versibly leading to the low conversion of aniline and inhib-
iting the condensation and advanced oxidation reactions,
whereas sites of low acidic strength may not act intensively
in the catalytic process. Therefore, the Nb O sample with a
2
5
small fraction of acidic sites of moderate strength, observed
by the band at 380 °C in the TPD (Fig. 3a), presents bet-
ter catalytic performance for azoxybenzene production than
the Al O sample. On the other hand, the Al O /AlNbO
2
3
2
3
4
sample, which presents high content of sites of moderate
acid strength, showed similar activity but a lower tendency
for azoxybenzene formation than Nb O catalyst. It is also
2
5
worth mentioning that, according to H -TPR results, the
2
greater reducibility of Nb-based samples could be translated
into higher aniline oxidation activity.
The reusability of Al O /AlNbO was evaluated in etha-
2
3
4
nol as a reaction solvent and with 1.2 mL of the oxidant,
H O (Fig. 10). The ꢂrst reusability test revealed good con-
2
2
version for the oxidation of aniline but upon evaluation in
the second reuse, the conversion level decreased by almost
2
0% and a further 15% for the third cycle. Although the
catalytic activity has decreased signiꢂcantly after the ꢂrst
reuse, the selectivity towards azoxybenzene was maintained
for all 3 cycles tested.
Fig. 7 Aniline conversion (a) and selectivity (b) for oxidation prod-
ucts as a function of solvent for Al O /AlNbO employed as cata-
2
3
4
lyst. Reaction temperature=25 °C, reaction time=24 h, H O vol-
2
2
ume=1.2 mL and mass of catalyst=10 mg
Conclusions
peroxo and hydroxyl species upon exposure to hydrogen per-
oxide. The eꢀciency of these catalysts in activating H O
In this study, catalytic tests were carried out with Al O /
2
3
AlNbO (72:28), a low-cost material produced by simple
2
2
4
can be attributed to the higher content of free hydroxyl
groups present on their surfaces. The catalytic action of mill-
ing powder on liquid phase oxidation of aromatic amine is
illustrated in Fig. 9b. Thus, since the peroxo groups replace
the acid groups on the surface (De Oliveira et al. 2014), from
the data presented in Table 2, we can speculate that the low
content of hydroxyl groups on the surface of Al O leads to
high-energy milling of Al O with Nb O , with subsequence
2
3
2
5
annealing treatment. The material presented a high degree
of crystallinity and found to be a nonporous and low sur-
face area material with rectangular and spherical morphol-
ogy. Through the investigation of several parameters and
processes of catalysis, it was concluded that the material is
very attractive for the proposed reaction because the reaction
outcome reached conversions of up to 100% of the aromatic
amine (aniline). It should be highlighted that the process
could be made selective towards the formation of azoxy-
benzene when employing acetone as solvent at ambient con-
ditions or acetonitrile, at a reaction temperature of 50 °C.
When compared with the precursorʼs materials the Al O /
2
3
a decrease in the formation of peroxo species (Ventura et al.
2
017). According to Ziolek et al. (2013, 2015), oxides that
are more eꢀcient in generating peroxo radicals and active
oxygen species are best suited for catalyzing oxidation
reactions. The high eꢀciency of niobium-based catalysts in
generating these species favors fast consumption of aniline.
Al O /AlNbO and Nb O , which were the most eꢁective in
2
3
AlNbO catalyst showed similar selectivity with Nb O and
2
3
4
2
5
4
2
5
the oxidation of the aniline, have more acidic sites and, as
a result, could catalyze the oxidation of the phenylhydroxy-
lamine and subsequently allow the condensation reaction
to occur which yielded azoxybenzene (Das and Talukdar
higher conversion than Al O , indicating the importance of
2 3
the addition of niobium to alumina, which even in an addi-
tion of 28% produces better results than pure Al O .
2
3
1
3

Chemical Papers
Fig. 8 Aniline conversion (a)
and selectivity (b) for oxida-
tion products as a function of
reaction temperature for Al O /
2
3
AlNbO employed as catalyst.
4
Reaction time=24 h, H O vol-
2
2
ume=0.4 and 1.2 mL, mass of
catalyst=10 mg and acetonitrile
as solvent
1
3

Chemical Papers
Fig. 10 Reusability of Al O /AlNbO : aniline conversion (a)
2
3
4
and product selectivities (b). Ethanol=3 mL, Aniline=100 µL,
Time=24 h, H O =1.2 mL, catalyst=10 mg, temperature=25 °C
2
2
Fig. 9 a Aniline conversion and selectivity for oxidation products for
Al O /AlNbO , Nb O and Al O employed as catalysts. Reaction
2
3
4
2
5
2
3
temperature=25 °C, reaction time=24 h, H O volume=0.4 mL, the
2
2
References
mass of catalyst=10 mg and acetonitrile as a solvent. b Scheme of
action of Al O /AlNbO powder fabricated by a facile high-energy
2
3
4
milling process with catalytic activity in the liquid phase oxidation of
Acharyya SS, Ghosh S, Bal R (2014) Catalytic oxidation of aniline to
azoxybenzene over CuCr O spinel nanoparticle catalyst. ACS
aromatic amine
2
4
Sustainable Chem Eng 2:584–589. https://doi.org/10.1021/sc500
0
545
Carreno NLV, Deon VG, Silva RM, Santana LR, Pereira RM, Orlandi
Acknowledgements This study was ꢂnanced in pt arby the Coorde-
MO, Ventura WM, Dias A, Taylor GT, Fajardo HV, Mesko MF
nação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil
(2018) Feasible and clean solid-phase synthesis of LiNbO by
3
(
CAPES)—Finance Code 001, FAPERGS PqG 2017 (2551-0001157-
microwave-induced combustion and its application as cata-
lyst for low-temperature aniline oxidation. ACS Sustainable
Chem Eng 6:1680–1691. https://doi.org/10.1021/acssuschem
eng.7b02921
0
), FINEP/CT-PETRO (2653/09–01.11.0091.00) and FAPEMIG (CEX-
APQ-00369-14), the Brazilian National Council for Scientiꢂc and
Technological Development (CNPq). The author gratefully acknowl-
edges the Brazilian Company of Metallurgy and Mining (CBMM) and
Center for Characterization and Development of Protocols for Nano-
technology (CCDPN-SisNano/UNESP).
Chang CF, Liu ST (2009) Catalytic oxidation of anilines into azoxy-
benzenes on mesoporous silicas containing cobalt oxide. J Mol
Catal A Chem 299:121–126. https://doi.org/10.1016/j.molca
ta.2008.10.032
Das DR, Talukdar AK (2017) Facile synthesis of titanium-loaded
MCM-48 as an eꢀcient heterogeneous catalyst for selective oxi-
dation of aniline to azoxybenzene. Chem Select 2:8983–8989.
https://doi.org/10.1002/slct.201701581
Compliance with ethical standards
Conflict of interest The authors declare no conꢃict of interest.
De Oliveira LC, Costa NT, Pliego JR Jr, Silva AC, De Souza PP,
Patrícia SDO (2014) Amphiphilic niobium oxyhydroxide as a
hybrid catalyst for sulfur removal from fuel in a biphasic sys-
tem. Appl Catal B Environ 147:43–48. https://doi.org/10.1016/j.
apcatb.2013.08.003
Ethical approval The founding sponsors had no role in the design of
the study; in the collection, analyses, or interpretation of data; in the
writing of the manuscript, and in the decision to publish the results.
Ghosh S, Acharyya SS, Sasaki T, Bal R (2015) Room temperature
selective oxidation of aniline to azoxybenzene over a silver
1
3

Chemical Papers
supported tungsten oxide nanostructured catalyst. Green Chem
catalysts for the selective oxidation of aniline. J Catal 331:110–
117. https://doi.org/10.1016/j.jcat.2015.09.004
1
7:1867–1876. https://doi.org/10.1039/C4GC02123A
Gontier S, Tuel A (1994) Oxidation of aniline over TS-1, the tita-
nium substituted silicalite-1. Appl Catal A Gen 118:173–186.
https://doi.org/10.1016/0926-860X(94)80312-9
Tumma H, Nagaraju N, Reddy KV (2009) Titanium (IV) oxide, an
eꢀcient and structure-sensitive heterogeneous catalyst for the
preparation of azoxybenzenes in the presence of hydrogen per-
oxide. Appl Catal A Gen 353:54–60. https://doi.org/10.1016/j.
apcata.2008.10.046
Gontier S, Tuel A (1995) Liquid-phase oxidation of aniline over
various transition-metal-substituted molecular-sieves. J Catal
1
57:124–132. https://doi.org/10.1006/jcat.1995.1273
Ventura WM, Batalha DC, Fajardo HV, Taylor JG, Marins NH,
Noremberg BS, Tanski T, Carreño NL (2017) Low temperature
liquid phase catalytic oxidation of aniline promoted by niobium
pentoxide micro and nanoparticles. Catal Commun 99:135–140.
https://doi.org/10.1016/j.catcom.2017.06.004
Jagtap N, Ramaswamy V (2006) Oxidation of aniline over titania pil-
lared montmorillonite clays. Appl Clay Sci 33:89–98. https://doi.
org/10.1006/jcat.1995.1273
Monshi A, Foroughi MR, Moshi MR (2012) Modiꢂed Scherrer equa-
tion to estimate more accurately nano-crystallite size using XRD.
WJNSE 2:154–160. https://doi.org/10.4236/wjnse.2012.23020
Qadir MI, Scholten JD, Dupont J (2015) Ionic liquid eꢁect: selective
Waghmode SB, Sabne SM, Sivasanker S (2001) Liquid phase oxi-
dation of amines to azoxy compounds over ETS-10 molecular
sieves. Green Chem 3:285–288. https://doi.org/10.1039/B1053
16G
aniline oxidative coupling to azoxybenzene by TiO . Catal Sci
2
Technol 5:1459–1462. https://doi.org/10.1039/C4CY01257G
Rodrigus R, Isoda N, Gonçalves M, Figueiredo FCA, Mandelli D, Car-
valho WA (2012) Eꢁect of niobia and alumina as support for Pt
catalysts in the hydrogenolysis of glycerol. Chem Eng J 199:457–
Wang Q, Wang FC, Cheng XW (2016) Electrochemical performance of
aluminum niobium oxide as anode for lithium-ion batteries. Rare
Met 353:256–261. https://doi.org/10.1007/s12598-016-0690-y
Xie K, Wei W, Yu H, Deng M, Ke S, Zeng X, Li Z, Shen C, Wang J,
Wei B (2016) Use of a novel layered titanoniobate as an anode
material for long cycle life sodium ion batteries. Catal Today
78:65–77. https://doi.org/10.1039/C6RA02530G
4
677. https://doi.org/10.1016/j.cej.2012.06.002
Selvam T, Ramaswamy AV (1995) Selective catalytic oxidation of ani-
line to azoxybenzene over titanium silicate molecular sieves. Catal
Lett 31:103–113. https://doi.org/10.1007/BF00817037
Yang P, Fan S, Chen Z, Bao G, Zuo S, Qi C (2018) Synthesis of Nb O
2
5
Shukla A, Singha RK, Konathala LS, Sasaki T, Bal R (2016) Cata-
lytic oxidation of aromatic amines to azoxy compounds over a
Cu–CeO catalyst using H O as an oxidant. RSC Adv 6:22812–
based solid superacid materials for catalytic combustion of chlo-
rinated VOCs. Appl Cat B Environ 239:114–124. https://doi.
org/10.1016/j.apcatb.2018.07.061
2
2
2
2
2820. https://doi.org/10.1039/C5RA25699B
Ziolek M, Sobczak I, Decyk P, Wolski L (2013) The ability of Nb O
2
5
Sing KS (1985) Reporting physisorption data for gas/solid systems
with special reference to the determination of surface area and
porosity (Recommendations 1984). Pure Appl Chem 57:603–
and Ta O to generate active oxygen in contact with hydrogen
2 5
peroxide. Catal Commun 37:85–91. https://doi.org/10.1016/j.
catcom.2013.03.032
6
19. https://doi.org/10.1351/pac198557040603
Ziolek M, Sobczak I, Decyk P, Sobańska K, Pietrzyk P, Sojka Z
(2015) Search for reactive intermediates in catalytic oxidation
with hydrogen peroxide over amorphous niobium (V) and tanta-
lum (V) oxides. Appl Catal B Environ 164:288–296. https://doi.
org/10.1016/j.apcatb.2014.09.024
Tanabe K (2003) Catalytic application of niobium compounds. Catal
Today 78:65–77. https://doi.org/10.1016/S0920-5861(02)00343
-
7
Thommes M, Kaneko K, Neimark AV, Olivier JP, Rodriguez-Reinoso
F, Rouquerol J, Sing KS (2015) Physisorption of gases, with
special reference to the evaluation of surface area and pore
size distribution (IUPAC Technical Report). Pure Appl Chem
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional aꢀliations.
8
7:1051–1069. https://doi.org/10.1515/pac-2014-1117
Trautwein G, El Bakkali B, Alcañiz-Monge J, Artetxe B, Reinoso S,
Gutiérrez-Zorrilla JM (2015) Dimeric assemblies of lanthanide-
stabilised dilacunary Keggin tungstogermanates: a new class of
1
3