Thermally stable imidazole/heteropoly acid composite as a heterogeneous catalyst for m-xylene ammoxidation

Article abstract of DOI:10.1007/s11164-020-04342-1

Ammoxidation of m-xylene is evaluated in the presence of a customized heteropoly acid catalyst as an imidazole/molybdovanadophosphoric acid (imidazole/PMoV). Imidazole is employed to maintain its heterogeneous phase during the ammoxidation reaction and to provide the thermal stability of PMoV with the expectation that imidazole can generate strong electronic interactions with terminal molybdenum-oxygen on PMoV. The characterizations of the prepared catalysts are performed using SEM–EDX, XRD, FT-IR, Raman, XPS, and TGA to prove the physical and chemical changes by incorporating imidazole to PMoV, respectively. Also, the thermal stability of the developed catalyst is confirmed by the means of heat treatment test at relatively high temperature. The composite catalyst, imidazole/PMoV, shows an excellent conversion rate of over 98% with high selectivity of isophthalonitrile in m-xylene ammoxidation. Moreover, while the imidazole-free PMoV catalyst is deactivated and washed out during the reaction, the catalyst durability of the imidazole/PMoV is preserved without significant activity loss after 5 reaction cycles at 380 °C.

Full text of DOI:10.1007/s11164-020-04342-1

Research on Chemical Intermediates (2021) 47:287–302
https://doi.org/10.1007/s11164-020-04342-1
Thermally stable imidazole/heteropoly acid composite
as a heterogeneous catalyst for m‑xylene ammoxidation
1
2
3
3
3
Yukwon Jeon · Chanmin Lee · Gicheon Lee · Ohchan Kwon · Jinsol Kim ·
Sang Sun Park3 · Kyeongseok Oh · Yong‑Gun Shul
,4
5
3
Received: 16 October 2020 / Accepted: 27 October 2020 / Published online: 18 January 2021
Springer Nature B.V. 2021, corrected publication 2021
©
Abstract
Ammoxidation of m-xylene is evaluated in the presence of a customized heteropoly
acid catalyst as an imidazole/molybdovanadophosphoric acid (imidazole/PMoV).
Imidazole is employed to maintain its heterogeneous phase during the ammoxida-
tion reaction and to provide the thermal stability of PMoV with the expectation that
imidazole can generate strong electronic interactions with terminal molybdenum-
oxygen on PMoV. The characterizations of the prepared catalysts are performed
using SEM–EDX, XRD, FT-IR, Raman, XPS, and TGA to prove the physical and
chemical changes by incorporating imidazole to PMoV, respectively. Also, the ther-
mal stability of the developed catalyst is conꢀrmed by the means of heat treatment
test at relatively high temperature. The composite catalyst, imidazole/PMoV, shows
an excellent conversion rate of over 98% with high selectivity of isophthalonitrile
in m-xylene ammoxidation. Moreover, while the imidazole-free PMoV catalyst is
deactivated and washed out during the reaction, the catalyst durability of the imi-
dazole/PMoV is preserved without signiꢀcant activity loss after 5 reaction cycles at
3
80 °C.
Keywords Composite catalyst · Thermal stability · m-Xylene · Ammoxidation
Yukwon Jeon and Chanmin Lee both authors contributed equally.
Supplementary Information The online version contains supplementary material available at https://
doi.org/10.1007/s11164-020-04342-1 .
*
*
*
Yukwon Jeon
ykjeon@yonsei.ac.kr
Kyeongseok Oh
kyeongseok.oh@inhatc.ac.kr
Yong-Gun Shul
shulyg@yonsei.ac.kr
Extended author information available on the last page of the article
Vol.:(012
1
3456789)
3

288
Y. Jeon et al.
Introduction
Since multifunctional catalysts are highly needed in various reactions, there have
been wide attempts to provide multi-catalytic centers in catalyst structures [1, 2].
In particular, heteropoly acid (HPA) could be one candidate applicable to acid
and mild oxidation reactions [3–5]. HPAs retain Keggin family structures that are
suitable for oxidation and acid reaction. They are easily modiꢀed on a molecu-
lar level due to their acidic and redox properties. Additionally, the variation of
central hetero-atoms and/or substituted transition metal atoms provides a partial
oxidation state [6, 7]. Therefore, HPA can be purposely customized by replacing
central hetero-atoms (e.g., P, Si, and etc.), addenda atoms (e.g., W, Mo, V, Fe,
and etc.) or counter ions [8, 9].
In an ammoxidation reaction, vanadia-based catalysts have been used to obtain
aromatic nitriles through the ammoxidation reaction of hydrocarbons because of
their high activity originating from vanadium site. An example of an aromatic
nitrile, isophthalonitrile (IPN), is a widely applicable intermediary in producing
pharmaceuticals, paints, resins, dyestuꢁs and pesticides, and can be produced by
ammoxidation of m-xylene [10, 11]. Vanadium-substituted molybdophosphoric
acid, a modiꢀed HPA, was recently challenged in selective oxidation reaction
with the anticipation that this modiꢀed HPA would enhance its oxidizing abil-
ity through partial substitution of vanadium for existing molybdenum [12–14].
However, the high solubility in polar solvents of the free acids of HPA is a cru-
cial problem at oxidation reactions. Also, the rapid depreciation of thermal stabil-
ity became a problematic issue when it was applied to heterogeneous gas phase
reactions, which would result in the reduction of surface area and catalytic activ-
ity. High thermal stability of HPA is vital because reaction temperatures in gas
phases often go higher than their decomposition temperatures. Unfortunately,
it was reported that the vanadium atom in H PMo VO was easily eliminated
4
11
40
from polyanion structures by thermal treatment and that catalyst acid strength
could be diminished along with its thermal stability [15–17]. In order to modify
the property of HPA catalysts, a concept of designing complexes and inorganic
supports with HPA deposition has been widely studied [18–20]. Furthermore, a
replacement of counter ions from hydrogen to other molecules, such as cesium,
ammonium and Zn, has been also attempted these days [21–24]. With its rein-
forced stabilization, vanadium-incorporated ammonium salt of 12-molybdophos-
phoric acid over titania support could be used for 2-cyanopyrazine production
through an ammoxidation reaction of 2-methylpyrazine [10]. Nevertheless, only
a few have been reported regarding organically incorporated HPA to improve its
thermal stability [25–27].
In this paper, we prepared a thermally stabilized HPA by incorporation of
imidazole into vanadium-substituted molybdophosphoric acid (PMoV). Physi-
cal and chemical characterizations were performed by scanning electron micros-
copy (SEM), X-ray diꢁraction (XRD) measurement, FT-IR spectroscopy, electron
spectroscopy for chemical Analysis (ESCA), Raman spectroscopy and thermo-
gravimetric analysis (TGA). We also evaluated the ammoxidation reaction of
1
3

Thermally stable imidazole/heteropoly acid composite as…
289
m-xylene in the presence of the imidazole/PMoV catalyst in regard to m-xylene
conversion, product selectivity, and durability.
Experiment
Catalyst preparation
A single species vanadium-incorporated heteropoly acid (H PMo VO ·xH O,
4
11
40
2
PMoV) was prepared by the impregnation method at a ratio of 1:11:1=P:Mo:V [10,
1
1]. First, molybdenum (VI) oxide (MoO , ACS reagent 99.5%, Sigma-Aldrich) and
3
Vanadium (V) oxide (V O , 99.6%+, Sigma-Aldrich) were dissolved as received in
2
5
de-ionized water. This mixture was heated for 24 h under reꢂux at mild stirring con-
ditions accompanied with the addition of Phosphoric Acid (H PO , 85%, Duksan).
3
4
Molybdophosphoric acid (PMoV) was collected by vacuum heating environment at
1
00 °C and was re-crystallized three times to obtain a ꢀne PMoV powder. Second,
for the imidazole/PMoV composite, PMoV powder was mixed with aqueous imida-
zole (C H N , 99%, Acros organics) by dropping and stirring for 5 h. The total ratio
3
4
2
of PMoV and imidazole was set to 1:4. Final imidazole/PMoV powder was obtained
by centrifuging the suspended imidazole/PMoV solution. After that, the collected
powder was dried at 80 °C. Then, the imidazole/PMoV powder was calcined at
4
00 °C for 6 h. For comparison of the thermal stability, both prepared imidazole-
free PMoV and imidazole/PMoV powder were also calcined at 400 °C for 6 h.
Characterization
Various measurements were performed to characterize imidazole-free PMoV and
imidazole/PMoV, heteropoly acid catalysts. Scanning electron microscopy (SEM,
operated at 15 kV, JEOL) equipped with an energy-dispersive X-ray (EDX) spec-
trometer was used to examine the morphology and element compositions of the
PMoV-based catalysts. The BET (Brunauer–Emmett–Teller) surface area of the
prepared catalysts was determined by N adsorption–desorption at -196 °C (77 K)
2
(
Micromeritics ASAP 2010). The X-ray diꢁraction (XRD) measurement at Multi-
ꢂex (Rigaku) was conducted to analyze the catalyst’s crystal structure with CuKα
radiation (λ=1.5405 nm) at 30 kV and 30 mA. The vibration spectra of the cata-
lyst were investigated by FT-IR spectroscopy using KBr pellet. All FT-IR spectra
−1
−1
were collected at the wavelength of 600–3400 cm with a resolution of 4 cm . The
oxidation states of each atom were analyzed by electron spectroscopy for chemical
analysis (ESCA) with Sigma probe (ThermoVG, UK). The normal X-ray scan was
conducted with monochromatic Al-Kα (1486.6 eV) at 12 kV, 100 W and 400 µm.
For accurate observations of the shifts, a calibration at C 1s (285 eV) was carried
–
8
out. This procedure ensured a pressure below 5×10 mbar during all experiments.
Raman spectroscopy (LabRam Aramis, Horriba) was performed in backscattering
geometry with a JY LabRam HR ꢀtted with a liquid-nitrogen-cooled CCD detec-
tor. The Raman spectra were collected under ambient conditions by the 514.6 nm
1
3

290
Y. Jeon et al.
line of an Ar ion laser. The catalyst’s thermal property was characterized by ther-
mogravimetric analysis (TGA, Perkin-Elmer Pyris Diamond) under air atmosphere
at a temperature range of 30 °C to 800 °C with a heating rate of 10 °C/min. For
the post-characterization, the imidazole/PMoV catalyst after the cycling tests of
m-xylene ammoxidation was also characterized by SEM, XRD, and FT-IR to prove
the stability.
Catalytic reaction test
Ammoxidation of m-xylene (MX, 98%, Sigma-Aldrich) was carried out in a ꢀxed-
bed reactor at atmospheric pressure and in the temperature range of 300–400 °C
with 20 °C intervals. About 1 g of the catalyst was loaded in the reactor with an
inner diameter of 3/8 inches. Then, a 0.3 ml/h of MX (liquid) and the gas mixture
of NH /Air with a molar ratio of MX:NH :Air=1:11:25.4 were fed into the reactor
3
3
at a gas hourly space velocity(GHSV) of approximately 600/h. The inlet part of the
reactor was suꢃciently preheated to over 200 °C and the temperature of the outlet
part was maintained at over 250 °C using heat trace. After reaching a steady state at
each reaction temperature, the gas products were obtained from the outlet and then
analyzed by gas chromatography (Acme 6000 GC, Younglin) with a HP-5 column
using a ꢂame ionization detector (FID). Calibration against standards permitted ana-
lytical accuracy within ±0.5%. The catalytic performances were determined mainly
regarding m-xylene conversion, isophthalonitrile (IPN) and tolunitrile (TN) selectiv-
ity. The reaction was repeated ꢀve times for a cycling test to prove the thermal and
chemical stability of the composite catalyst. For every cycle, the operating condi-
tions in all experiments were the same for the m-xylene ammoxidation carried out
with the composite imidazole/PMoV catalyst. The cycling process had four steps
for one cycle and started at 200 °C followed by the catalyst activation step for 2 h.
Then the temperature was raised to 380 °C where the best performance was obtained
according to the pre-experiment. During 24 h for one cycle, the ꢀnal products were
analyzed by gas chromatography several times to get the average performance.
Lastly, the temperature was lowered to room temperature and the reactor was purged
with N gas. The cycling test was repeated ꢀve times under the same conditions to
2
compare the durability of the catalyst.
Results and discussion
Catalyst characterization
The element compositions and morphologies of the imidazole-free PMoV and the
imidazole/PMoV composite are compared by SEM images and EDX analysis, as
in Fig. 1 and S1. The elemental measurements were performed by the SEM–EDX,
and the calculated compositions are summarized in Table 1. As from the elementary
results, both samples indicate almost equal composition ratios with the designed cat-
alyst structures of the H PMo VO (1:11:1=P:Mo:V) and imidazole(3:4=C:N)/
4
11
40
1
3

Thermally stable imidazole/heteropoly acid composite as…
291
Fig. 1 SEM images and EDX analysis of A imidazole/PMoV composite and B imidazole-free PMoV
Table 1 Elementary and BET surface area analysis of A. imidazole/PMoV composite (where A’ is the
average elementary data from diꢁerent sites in Fig. S1) and B. imidazole-free PMoV
Sample
Element compositions (%)
BET surface area
2
P
V
Mo
O
C
N
m /g
A
A’
B
1.27
1.22
1.53
1.31
1.42
1.66
14.68
15.23
18.18
45.79
48.23
77.02
19.93
18.28
–
15.87
14.99
–
17.3
4.9
PMoV, respectively. The composition of the developed composite was measured
and averaged by analyzing diꢁerent sites (Fig. S1 and Table 1), which conꢀrms the
desired ratio of 1:4 of PMoV and imidazole. By the morphological comparison from
the SEM images (Fig. 1), the imidazole-free PMoV catalyst does not have a speciꢀc
shape, but imidazole/PMoV composite has particle agglomerates in a polyhedron
structure with a uniform size distribution of approximately 3–10 µm. This indicates
that PMoV particles can be shaped in a certain morphology by the addition of het-
erocyclic material (here, imidazole). As the morphology of the catalyst inꢂuences to
control the speciꢀc catalyst surface area, both imidazole-free and imidazole/PMoV
are measured to have an increase in the surface area for more than 3 times from
2
4
.9 to 17.3 m /g, as listed in Table 1. Thus, the higher surface area and mechani-
cally stable shape of the imidazole/PMoV particle allow the particle to become an
active conꢀguration in oxidation reactions that occur under relatively severe reaction
conditions.
XRD measurement was employed to verify the catalyst structural changes by add-
ing heterocycle to heteropoly acid (HPA). Figure 2a shows XRD patterns of imida-
zole-free PMoV as well as the imidazole/PMoV composite. In the case of imidazole-
free PMoV, the XRD curve displayed a typical pattern of pure hydrogen vanadium
molybdenum oxide phosphate hydrate (JCPDS No. 45-0611). On the other hand,
the addition of imidazole into PMoV resulted in a considerable structural change to
form a stable Keggin structure. During the calcination process, XRD curve patterns
of the imidazole/PMoV composite shifted to cubic symmetry (Pn3m), one of the
known stable HPA secondary structures, which is similar to the PWA ammonium
1
3

292
Y. Jeon et al.
Fig. 2 A X-ray diꢁraction patterns, and B Infrared spectra recorded for (a) imidazole/PMoV composite
and (b) imidazole-free PMoV; PMoV (hydrogen vanadium molybdenum oxide phosphate hydrate, green
down-pointing triangle), cubic Keggin structure (blue ꢀlled circle)
salt structure of [(PW O )·[N(CH ) H ] ] during the calcination process [3,
1
2
40
3 x 4−x 3
2
5]. Thus, it could be interpreted that the distortion of crystal structure is caused by
strong interactions between imidazole and PMoV.
To investigate the interaction between imidazole and PMoV, FT-IR measurement
was carried out. Figure 2B shows FT-IR transmittance spectra of both imidazole-free
PMoV and the imidazole/PMoV composite. FT-IR transmittance spectra of imida-
−
1
zole-free PMoV have characteristic vibration bands at 1063, 961, 863 and 784 cm
which can be assigned to the stretching mode of (P–O ), (Mo–O ), (Mo–O –Mo)
a
d
b
and (Mo–O –Mo), respectively (see Fig. 2B). Here, subscripts a, b, c and d represent
c
individual oxygen placed in diꢁerent locations in the tetrahedron structure. These
four bands conꢀrm the existence of a Keggin structure in the vanadium-substituted
molybdophosphoric acid as presented in other literature [25–29]. The addition of
the heterocycle (imidazole) to imidazole-free PMoV led to slight shifts and changes
in the FT-IR spectral bands. However, no shift was observed on the P–O stretch-
−1
ing band at 1064 cm , which may suggest that there are no interactions between
imidazole and the P–O bond in PMoV. Since P–O bonds are surely located at the
center of PMoV structure, imidazole molecules have diꢃculty penetrating PMoV
structure. When compared with stretching modes of Mo–O–Mo intra-bridges placed
in corner-sharing MoO , individual stretching modes of (Mo–O ), (Mo–O –Mo),
6
d
b
−
1
and (Mo–O –Mo) were shifted to 970, 871, and 791 cm . The shifts and switching
c
−
1
eꢁects of transmittance intensities of FT-IR spectra at 970 and 791 cm are due to
the structural distortion of PMoV crystals apparently caused by the incorporation
of imidazole into the Keggin structure. Additionally, FT-IR results provide clues as
to the strong electrostatic interaction between non-bonding nitrogen, i.e., the con-
jugation of π electrons of heterocycles into the inorganic framework of PMoV. A
nitrogen atom with two pairs of non-bonding electrons in an imidazole molecule
may be capable of interacting predominantly with oxygen atoms of the Mo–O–Mo
intra-bridge in PMoV, but interact indirectly with the oxygen atoms of PO ion in
4
a PMoV cage. Therefore, the stretching modes of PMoV shifted to higher frequen-
cies in FT-IR spectra and the Keggin structure of PMoV was marginally distorted
1
3

Thermally stable imidazole/heteropoly acid composite as…
293
after composite formation with imidazole. The results fully support that organic
π electrons may extend their conjugation to an inorganic framework and thus dra-
matically modify the acidity and redox properties of the cluster. Hence, the shift
in the Mo–O band of PMoV might be mainly attributed to the interaction of the
cationic heterocycle species with polyanions. Furthermore, new broad peaks near
−
1
−1
2
930 cm appeared at the frequency range of 2500 to 3100 cm in Fig. 2B, due to
the stretching mode of the N–H aromatic ring of imidazole after a composite forma-
tion with imidazole. This demonstrates that imidazole molecules clearly interacted
with PMoV [26, 30].
The change in the oxidation states and chemical environment from the XPS
data of the V2p , Mo 3d, N 1s–Mo 3p and O 1s regions are shown and sum-
3
/2
3/2
marized in Fig. 3 and Table 2 for the imidazole-free PMoV and the imidazole/
PMoV composite. The XPS peak of N1-Mo3p in Fig. 3A indicates the existence
of imidazole molecules binding on PMoV1 through the N 1s peak even after the
heat treatment. Generally, imidazole has three diꢁerent nitrogen atoms including
+
–
N=species, –NH– species and two N species, and their binding energies were
reported as 398.4± 0.1 eV, 399.4± 0.1 eV and 400.7–402.5 eV, respectively [31].
The N 1 s peak revealed, mainly, the binding energy region for non-bonding nitro-
+
gen, (N ) since it may aꢁect an ionic interaction between the organic π electrons and
the PMoV, while the other two binding energies cannot be clearly deconvoluted due
to the overlapped Mo 3p peak.
Meanwhile, the oxidation states and chemical environments of molybdenum,
vanadium, and oxygen were changed due to shifts in the binding energy of the PMoV
crystals owing to the structure distortion seemingly aꢁected by the incorporation of
Fig. 3 XPS spectra of A Mo 3p3/2 and N 1 s, B Mo 3d, C V 2p , and D O 1 s for (a) imidazole/PMoV
3/2
composite and (b) imidazole-free PMoV, as well as E illustration of the plausible chemical bonding
structure for the imidazole/PMoV composite catalyst for the ammoxidation of m-xylene
1
3

294
Y. Jeon et al.
1
3

Thermally stable imidazole/heteropoly acid composite as…
295
imidazole into the Keggin structure. Remarkably, the Mo states including Mo 3p
and Mo 3d shifted toward the lower binding energy. The 3d orbital of molybdenum
consisted of two peaks including Mo 3d and Mo 3d with binding energies as
3
/2
5/2
shown in Fig. 3B. The observed binding energy values of Mo 3d doublet were inter-
preted as the molybdenum species having a valence state of 6+ in PMoV [32, 33].
Thus, prepared imidazole/PMoV mainly contains molybdenum ions with 6+. The
slight shift of the Mo doublet indicates the conjugation between imidazole and the
inorganic framework of PMoV, which was conꢀrmed by the FT-IR results.
Moreover, the oxidation state of the vanadium atoms can be estimated from the
V2p level in the XPS data. In previous studies, the XPS peak of V2p with a
3
/2
3/2
valence of 5+ appeared in the binding energy range between V2p binding energy
3
/2
for 5+ oxidation state ranges between 517.4 and 518.4 eV, while that with a valence
state of 4+ appeared in the range of 516.1 to 516.8 eV [34–36]. As shown in Fig. 3C,
vanadium ions of PMoV showed mixed valance states of 4+ and 5+ but mainly fea-
tured a valence state of 5+ (79.4%) rather than 4+ (20.6%). The imidazole/PMoV
composite catalyst still has a large portion of vanadium ions with a valence state
of 5+ (51.6%) but also shows a shift toward a region of oxidation state with 4+
(
48.4%) due to the partial reduction of vanadium ions after interaction with nitrogen
atoms containing non-bonding electron pairs.
As for the O1s spectra, PMoV can be deconvoluted into three diꢁerent con-
tributions with the binding energy ranges of 530.4–530.6, 531.4–531.6, and
2
−
−
5
32.6–533.1 eV, which are assignable to oxide (O ), hydroxide (OH ), and struc-
tural water (H O) [37, 38], respectively. As shown in Fig. 3D and Table 2, during
2
the procedure in our experiment, the portions of three contributions changed accord-
ing to the shifts of the O1s peaks. In the case of imidazole/PMoV, all these results
+
suggest that the N in the imidazole changed to an N state, which induces struc-
tural distortion of Mo and V in the Keggin structure along with the protonation of
2
−
O
and OH in the PMoV catalyst. It seems that this conꢀguration of the imidazole/
PMoV composite demonstrates the prevalent role of the redox cycles on the cata-
lytic activity.
Overall, the FT-IR and XPS results provide a strong clue on the changes of the
chemical bonding structure by the ionic interaction between the outer O element
in the HPA and non-bonded N in the imidazole, as illustrated in Fig. 3E. From this
new bonding structure, insolubility in popular hydrophilic solvents and an additional
enhancement of the HPA thermal stability were achieved, which are important char-
acteristics in heterogeneous catalysis.
TGA measurement was performed under air atmosphere to evaluate the thermal
stability of the imidazole/PMoV composite. Figure 4 shows the TGA curves of both
imidazole-free PMoV and the imidazole/PMoV composite. In the case of imidazole-
free PMoV, four step changes in weight loss were recorded. First, weight loss was
observed about 2 wt% at around 170 °C, and this is attributed to the water dissocia-
tion on the surface of PMoV. Consecutive weight loss was about 10 wt% observed
at a temperature range of 200–500 °C, which is related to the further elimination of
water molecules during the oxide crystallization process followed by the decomposi-
tion of PMoV into P O , MoO and V O [17]. Compared with the TGA curve of
2
5
3
2
5
imidazole-free PMoV, the imidazole/PMoV composite curve showed a clear proof
1
3

296
Y. Jeon et al.
Fig. 4 Thermogravimetric
curves under air atmosphere of
(
a) imidazole/PMoV composite
and (b) imidazole-free PMoV
of thermal stability enhancement up to 400 °C as shown in Fig. 4. Since the imida-
zole/PMoV composite does not contain physical or chemical adsorbed water mole-
cules, weight loss of about 2% up to 400 °C is presumed to be due to the elimination
of structural water and heterocycle molecules. Two pairs of non-bonding electrons
from imidazole may participate in the interaction with PMoV and then are tightly
bound with PMoV, which maintains almost the same crystal structure even after
ammoxidation reaction at temperatures of up to 400 °C.
To further investigate the thermal stability of the imidazole/PMoV composite,
XRD and FT-IR studies were performed for the imidazole-free PMoV and the imi-
dazole/PMoV composite catalysts prepared after heat treatment at 400 °C for 6 h
(
see Fig. 5). As from Fig. 5A, the notable point is that the imidazole/PMoV compos-
ite catalyst calcined at high temperature retained the Keggin structure with a stable
cubic secondary arrangement. However, the imidazole-free PMoV catalyst showed
a pattern mainly corresponding to α-MoO , which indicates the decomposition of
3
Keggin ions. Moreover, in vanadium incorporated catalysts, an elimination of vana-
dium atom is also possible at higher temperatures in oxygen atmosphere depending
on the extent of V insertion, and may form unidentiꢀed polymeric and monomeric
Fig. 5 A X-ray diꢁraction patterns, and B Infrared spectra recorded for (a) imidazole-free PMoV and (b)
after 400 °C treatment, as well as (c) imidazole/PMoV composite and (d) after 400 °C treatment; PMoV
(
hydrogen vanadium molybdenum oxide phosphate hydrate, green down-pointing triangle), cubic Keggin
structure (blue ꢀlled circle) and MoO (brown diamond)
3
1
3

Thermally stable imidazole/heteropoly acid composite as…
297
vanadium species [21, 39]. The FT-IR results in Fig. 5B are also in accordance
with this observation. No positional change in the FT-IR transmittance spectra of
imidazole/PMoV composite catalysts was observed even after heat treatment with
−
1
the characteristic vibration bands at 1066, 973, 871 and 790 cm assigned to the
stretching mode of (P–O ), (Mo–O ), (Mo–O –Mo) and (Mo–O –Mo), respectively.
a
d
b
c
On the other hand, the distinct vibration bands of the imidazole-free PMoV cata-
−1
lyst changed to the 988 and 874 cm , indicating a breakdown of the heteropolyacid
structure to individual chemical phases with a stretching mode of the Mo–O–Mo
and Mo=O units for the speciꢀcation of a layered orthorhombic α-MoO phase [40].
3
Therefore, this ꢀnding of the stable phase structure of the imidazole/PMoV compos-
ite catalysts without any transformation at high temperature testiꢀes the suitability
of the catalyst for the air oxidation reactions.
Catalytic reaction and cycling test
As the desirable ꢀnal product is IPN, the reaction pathway of m-xylene ammoxida-
tion was categorized in a parallel reaction, which consists of the (1) step-by step
ammoxidation of m-xylene to IPN via m-tolunitrile and (2) direct formation of IPN
from m-xylene [33]. Therefore, it is important to consider the operational condition
of ammoxidation reaction especially reaction temperatures for either reaction (1) or
reaction (2) to completely occur. To demonstrate the catalytic activity, an ammoxi-
dation reaction with the imidazole/PMoV catalysts was performed in a variety of
temperatures. From the experiment data in Table 1, Imidazole-free PMoV showed
big problems of low surface area and high solubility during the reaction, which is
conꢀrmed and has to be overcome for applications [15–17]. Figure 6A presents
m-xylene conversion, IPN selectivity, and TN selectivity after the reaction that was
performed at temperatures ranging from 300 to 400 °C with intervals of 20 °C.
M-xylene conversion and IPN selectivity were increased gradually with increasing
reaction temperatures up to 380 °C. The conversion of m-xylene was 98% and IPN
selectivity was over 77% at 380 °C under a higher surface area than the imidazole-
free PMoV. From the results, imidazole/PMoV is considered to be a promising
Fig. 6 Performance result of m-xylene ammoxidation for imidazole/PMoV composite catalyst. A correla-
tion between reaction temperature and catalytic performance, as well as B cycling data of (a) conversion
for m-xylene, (b) selectivity to isophthalonitrile, (c) selectivity to m-tolunitrile and (d) selectivity to COx
1
3

298
Y. Jeon et al.
Table 3 Catalyst activity results of m-xylene ammoxidation at 380 °C of m-xylene conversion and selec-
tivities to isophthalonitrile, m-tolunitrile and COx for A. imidazole/PMoV composite and B. imidazole-
free PMoV
Sample
m-xylene conver-
TN selectivity
(%)
IPN selectivity COx selectivity
sion (%)
(%)
(%)
A
B
98.8
71.3
13.5
43.6
83.7
41.6
5.1
9.5
Washed out
composite catalyst for ammoxidation reaction as it maintains high conversion of
m-xylene and good selectivity of IPN even at high temperatures. The results also
showed that the optimum catalytic activity was obtained at a slightly lower tempera-
ture of 380 °C, rather than the temperature that has been previously reported in other
works (averaging from 375 to 420 °C) [32–34].
In addition, stability of the imidazole/PMoV composite was evaluated by cycling
test at a temperature of 380 °C. Figure 6B shows the results of the cycling test with
imidazole/PMoV under the same operation conditions. Although no regeneration
steps were included during the cycle test, it was observed that the same quantities
of m-xylene conversion and IPN selectivity were constantly achieved for 5 cycles
without signiꢀcant changes, while the imidazole-free catalyst was washed out dur-
ing the reaction (Table 3). This may be due to the sustainable polyhedron structure
in the morphology of the imidazole/PMoV composite without any agglomeration
of the particles, as is consistent with the results observed from SEM in Fig. 7A. A
Fig. 7 A SEM images of imidazole/PMoV composite after stability test, B XRD patterns, and C FT-IR
curves of imidazole/PMoV composite before and after the m-xylene ammoxidation cycling test
1
3

Thermally stable imidazole/heteropoly acid composite as…
299
stable chemical conꢀguration was also observed with no structural change in the
cubic symmetry and crystallinity as from the XRD analysis in Fig. 7B. Further-
more, no diꢁerence in the chemical structure compared to the FT-IR result after
ammoxidation reaction of m-xylene was revealed, except a slight decrease in the
intensity which is normally seen at the used catalysts. The FT-IR results in Fig. 7C
showed exactly the same peak positions of the stretching mode for (P–O ), (Mo–O ),
a
d
(
Mo–O –Mo) and (Mo–O –Mo), respectively, for the used imidazole/PMoV com-
b c
posite catalyst. Therefore, it should be noted that signiꢀcant enhancement in the
thermal stability of PMoV was achieved by the addition of imidazole into the PMoV
structure.
Conclusion
A customized heteropoly acid, molybdovanadophosphoric acid (PMoV), was syn-
thesized from binary oxides (molybdenum oxide and vanadium oxide) using an
impregnation method. Imidazole was then employed to reinforce the thermal stabil-
ity of PMoV and form an imidazole/PMoV composite by ion exchange. The char-
acterizations of both the imidazole-free PMoV and the imidazole/PMoV compos-
ite were examined using SEM, XRD, FTIR, XPS, and TGA measurements. The
tetrahedron shape of the imidazole/PMoV composite was veriꢀed and resulted in
well-improved thermal stability, which was conꢀrmed by the existence of strong
interactions between imidazole and PMoV. In addition, imidazole/PMoV showed a
high m-xylene conversion and a considerable improvement in IPN selectivity during
ammoxidation reactions at 380 °C. Imidazole/PMoV maintained its thermal stability
as well as its activity even after ꢀve cycling tests at 380 °C.
Memories with Prof. Michel Che
My memories with Prof. Michel Che have been nothing less than priceless. I will
always remember him as a kind, humorous soul with big glasses. 1994 was my
ꢀ
rst sabbatical year as a professor at Yonsei University. During which, I decided to
become a visiting researcher at the Pierre and Marie Curie University, where Michel
was a member of the faculty. I still remember our intellectually stimulating discus-
sions continuing until late in the night. He always seems to raise essential and criti-
cal points giving much-needed light to challenging subjects. On one occasion, I had
encountered diꢃculty in one of my experiments with students when an inorganic
material synthesis wasn’t yielding reproducible results. It was here where Prof.
Che’s consult had been invaluable, pointing out an overlooked factor. While pas-
sionate about work, he still made brightening jokes, which made him a very likable
colleague. While general stereotype regards people of science to be far from arts, on
the contrary, I recall Michel being a very artistic person. He would give me presents
that were quite fancy in a sense, colorful and eye-catching. I still cherish these little
gifts.
1
3

300
Y. Jeon et al.
I was tremendously saddened to hear that Michel had passed away. Not only
because he was an exceptional researcher with numerous contributions to the sci-
entiꢀc community, but also because he was a joy for those who knew him. I feel
obliged to say that Prof. Michel Che will be deeply missed. May his soul rest in
peace.
Acknowledgements This article is dedicated to the memory of Prof. Michel Che, who sadly passed away
on August 7th, 2019. This work was supported by the National Research Foundation of Korea (NRF)
grant funded by the Korean government (MSIT) (No. 2020R1F1A107320111), and by the Development
of eco-friendly production system technology for total periodic resource cycle funded by the Korea Insti-
tute of Industrial Technology (kitech EO-20-0022).
References
1
2
3
.
.
.
M. Misono, Chem. Comm. 13, 1141 (2001)
T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. 41, 113 (1996)
K. Mohan Reddy, L. Nakka, S.N. Rao Pasupuleti, R. Pasupu, Nagaraju, P. S. Sai Prasad, I.
Suryanarayana, Catal. Lett. 130, 154 (2009)
4
5
6
7
8
9
.
.
.
.
.
.
N. Mizuno, M. Misono, Chem. Rev. 98, 199 (1998)
H.-J. Kim, Y.-G. Shul, H. Han, Appl. Catal. A Gen. 299, 46 (2006)
F. Cavani, Catal. Today 41, 73 (1998)
H.-J. Kim, Y.-G. Shul, H. Han, J. Power Sour. 158, 137 (2006)
K. Yamaguchi, S. Shinachi, N. Mizuno, Chem. Commun. 424 (2004)
I.V. Kozhevnikov, Catal. Rev. Sci. Eng. 37, 311 (1995)
1
0. K. Mohan Reddy, N. Lingaiah, K.N. Rao, N. Rahman, P.S. Sai Prasad, I. Suryanarayana, Appl.
Catal. A Gen. 296, 108 (2005)
1
1
1
1
1. Y. Zhang, A. Martin, H. Berndt, B. Liicke, M. Meisel, J. Mol. Catal. A Chem. 118, 205 (1997)
2. H. Liu, E. Iglesia, J. Phys. Chem. B 107, 10840 (2003)
3. N. Laronze, C. Marchal-Roch, N. Guillou, F.X. Liu, J. Catal. 220, 172 (2003)
4. C. Marchal-Roch, C. Julien, J.F. Moisan, N. Leclerc-Laronze, F.X. Liu, G. Herve, Appl. Catal. A
Gen. 278, 123 (2004)
1
5. C. Rocchioccioli-Deltcheꢁ, A. Aouissi, M. Bettahar, S. Launay, M. Fournier. J. Catal. 164, 16
(
1996)
1
1
1
1
6. K. Inumaru, A. Ono, H. Kubo, M. Misono, J. Chem. Soc. Faraday Trans. 94(12), 1765 (1998)
7. A. Micek-Ilnick, J. Mol. Catal. A Chem. 308, 1 (2009)
8. S. Sadjadi, M.M. Heravi, M. Daraie, Res. Chem. Intermed. 43, 2201 (2017)
9. A.K. Shah, S. Park, H.A. Khan, U.H. Bhatti, P. Kumar, A.W. Bhutto, Y.H. Park, Res. Chem.
Intermed. 44, 2405 (2018)
2
2
0. E. Raꢀee, M. Kahrizi, M. Joshaghani, P. Ghaderi-Sheikhi Abadi, Res. Chem. Intermed. 42, 5573
(
2016)
1. C. Marchal-Roch, N. Laronze, R. Villanneau, N. Guillou, A. Teze, G. Herv, J. Catal. 190, 173
(
2000)
2
2
2
2
2
2
2
2
3
3
3
2. P. Villabrille, G. Romanelli, P. Vázquez, C. Cáceres, Appl. Catal. A Gen. 270, 101 (2004)
3. Z.D. Fang, D. Fang, Q. Chen, Res. Chem. Intermed. 39, 4205 (2013)
4. R. Jahanshahi, B. Akhlaghinia, Res. Chem. Intermed. 44, 2451 (2018)
5. J.D. Kim, I. Honma, Solid State Ionics 176, 547 (2005)
6. H. Ge, Y. Leng, F. Zhang, C. Zhou, J. Wang, Catal. Lett. 124, 250 (2008)
7. H.J. Kim, Y. Jeon, J.-I. Park, Y.-G. Shul, J. Mol. Catal. A Chem. 378, 232 (2013)
8. Z. Fumin, G. Maiping, G. Hanqing, W.J. Chin, J. Chem. Eng. 15(6), 895 (2007)
9. L. Dermeche, R. Thouvenot, S. Hocine, C. Rabia, Inorganica Chim. Acta 362, 3896 (2009)
0. C. Rocchiccioli-Deltcheꢁ, M. Fournier, J. Chem. Soc. Faraday Trans. 87, 3913 (1991)
1. R. Dziembaj, Z. Piwowarska, Synth. Met. 63, 225 (1994)
2. E. Blouet-Crusson, M. Rigole, M. Fournier, A. AboukaõÈs, F. Daubrege, G. Hecquet, M. Guelton,
Appl. Catal. A Gen. 178, 69 (1999)
1
3

Thermally stable imidazole/heteropoly acid composite as…
301
3
3
3
3
3
3
3
3. K.P. Han, J.L. Fang, J. Electron. Spectros. Relat. Phenomena 83, 93 (1997)
4. K.V. Narayana, A. Martin, U. Bentrup, B. Lücke, J. Sans, Appl. Catal. A Gen. 270, 57 (2004)
5. B.M. Weckhuysen, D.E. Keller, Catal. Today 78, 25 (2003)
6. I.-J. Lee, C.-H. Lee, S.-D. Lee, D.J. Suh, J. Ind. Eng. Chem. 11, 918 (2005)
7. M. Nakayama, H. Tagashira, S. Konishi, K. Ogura, Inorg. Chem. 43, 8215 (2004)
8. A. Predoeva, S. Damyanova, E.M. Gaigneaux, L. Petrov, Appl. Catal. A Gen. 319, 14 (2007)
9. N. Lingaiah, K. Mohan Reddy, N. Seshu Babu, K. Narasimha Rao, I. Suryanarayana, P.S. Sai
Prasad, Catal. Commun. 7, 245 (2006)
4
0. A. Klinbumrung, T. Thongtem, S. Thongtem, J. Nanomater. 2012, 930763 (2012)
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional aꢃliations.
Authors and Aꢀliations
1
2
3
3
3
Yukwon Jeon · Chanmin Lee · Gicheon Lee · Ohchan Kwon · Jinsol Kim ·
Sang Sun Park3 · Kyeongseok Oh · Yong‑Gun Shul
,4
5
3
1
Department of Environmental Engineering, Yonsei University, 1 Yonseidae-gil, Wonju,
Gangwon-do 26493, Republic of Korea
2
3
Research Institute of Clean Manufacturing System, Green and Sustainable Materials R&D
Department, Korea Institute of Industrial Technology, 89 Yangdaegiro-gil, lpjang-myeon,
Seobuk-gu, Cheonan-si, Chungcheongnam-do 31056, Republic of Korea
Department of Chemical and Biomolecular Engineering, Yonsei University, 50 Yonsei-ro,
Seodaemun-gu, Seoul 03722, Republic of Korea
1
3

302
Y. Jeon et al.
4
LOTTE Chemical Corporation, #24-1, Jang-dong, Yuseong-gu, Daejeon 305-726,
Republic of Korea
5
Department of Chemical and Environmental Technology, Inha Technical College, 100 Inha-ro,
Michuhol-gu, Incheon 22212, Republic of Korea
1
3