Synthesis of polyoxometalates supported on HZSM-5 for the photocatalytic purification of crude terephthalic acid under mild conditions

Article abstract of DOI:10.1016/j.jphotochem.2019.03.048

Phosphomolybdic (PMo) and vanadium-containing phosphomolybdic (PV₂Mo) acids were synthesized and supported on HZSM-5 zeolite via a wet impregnation technique and used as photocatalysts for the purification of terephthalic acid (TA) via an econo

Full text of DOI:10.1016/j.jphotochem.2019.03.048

Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
Contents lists available at ScienceDirect
Journal of Photochemistry & Photobiology A: Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Synthesis of polyoxometalates supported on HZSM-5 for the photocatalytic
purification of crude terephthalic acid under mild conditions
a
a,b,⁎
c
b
a,⁎
A.M. Hussein , A.H. Mady
, S. Mahmoud , J.-J. Shim , F.Z. Yehia
a
Department of Petrochemicals, Egyptian Petroleum Research Institute, Nasr City, Cairo, 11727, Egypt
School of Chemical Engineering, Yeungnam University, Gyeongsan, Gyeongbuk 38541, Republic of Korea
Department of Chemistry, AlAzhar University, Nasr City, Cairo, Egypt
b
c
A R T I C L E I N F O
A B S T R A C T
2
Phosphomolybdic (PMo) and vanadium-containing phosphomolybdic (PV Mo) acids were synthesized and
Keywords:
Polyoxometalate
Terephthalic acid
Purification
Photocatalysis
Persulfate
supported on HZSM-5 zeolite via a wet impregnation technique and used as photocatalysts for the purification of
terephthalic acid (TA) via an economic process. Nowadays, purified TA can be obtained from crude TA due to
their great commercial importance by separating 4-carboxybenzaldehyde (4-CBA), which represents high en-
vironmental toxicity and negatively affects the applications of TA. For these reasons, conversion of 4-CBA to TA
is highly desirable for both enhancing the efficiency of purified TA production and environmental reasons. The
synthesized catalysts were characterized by XRD, FTIR, Raman spectroscopy, BET, SEM, and TEM. The pur-
ification process was carried out by under different parameters as VUV irradiation, persulfate (PS), time, and
different types of catalysts (PMo or PV
2
Mo)/HZSM-5. This kind of combination showed complete removal of the
4-CBA with a high yield of PTA in 2 h. The high efficiency attributed to the acidity of the polyoxometalates/
HZSM-5, which increase the selectivity to oxidize 4-CBA at room temperature and much shorter reaction time
compared to other common methods.
1. Introduction
176–225 °C and pressures ranging from 1500 to 3000 kPa using Co/Mn
catalysts, acetic acid as solvent, and bromine as promoter [7]. In this
Purified terephthalic acid (PTA) is considered as the prime raw
process, crude terephthalic acid (CTA) is produced as the major product
and 4-carboxybenzaldehyde (4-CBA) as the major impurity [8]. This
impurity deteriorates the quality of the product by terminating the
polymerization process, resulting in polymers with low average mole-
cular weights and with color, hence, the TA should be purified to
contain less than 25 ppm 4-CBA to be considered of polymer grade
[6,9].
As the 4-CBA has the same chemical and physical properties with
that of TA, it is very difficult to separate it from TA using extraction,
adsorption, or membrane methods. It is also difficult to employ the
crystallization process using a solvent with good selectivity to 4-CBA
[9], hence, using heteropolyoxometalates, which have a Keggin struc-
ture, as photocatalysts is considered a good alternative solution. Het-
eropolyacids (HPA) have been used widely in catalysis and homo-
geneous photocatalysis due to their good solubility in polar solvents,
high proton mobility, facile photo excitation using near-visible or UV-
light, and thermal and oxidation stability besides their capability to
material in the production of polyethylene terephthalate (PET), which
is used in numerous industries, such as manufacturing of drink bottles,
film making for balloons, pliant food packaging, and support for pres-
sure sensitive adhesive tape [1,2]. Numerous methods of producing
terephthalic acid (TA) have been reported. Ciamician and Silber re-
corded the first production of terephthalic acid by using sunlight and
oxygen for the oxidation of p-xylene for a year, which resulted in the
generation of a small amount of terephthalic acid with other byproducts
[
3]. Stephens’ method involved the use of dimmed light instead of
sunlight at 100 °C, and the reaction time was 24–60 day with zero
percent of terephthalic acid production and only monoaldehydes pro-
duced [4]. Another method is the nitric acid method, which is con-
sidered a commercial method for p-xylene oxidation to dicarboxylic
acid (TA) using nitric acid as oxidant under temperatures of 160–200 °C
and pressures of 850 to 1350 kPa, but the drawback of this method is
the explosion hazards [5]. The spray reactor method [6] is also used,
but the most widely used industrial process is the AMOCO process,
which is based on the oxidation of p-xylene using air at temperatures of
2 2 2
activate oxidants such as O and H O for selective oxidation [10–12].
Therefore, it is expected that the presence of persulfate (PS) as a cheap
⁎
Corresponding authors at: Department of Petrochemicals, Egyptian Petroleum Research Institute, Nasr City, Cairo, 11727, Egypt.
E-mail addresses: amr_mady2005@yahoo.com (A.H. Mady), dr.famatzakaria@gmail.com (F.Z. Yehia).
https://doi.org/10.1016/j.jphotochem.2019.03.048
Received 13 November 2018; Received in revised form 27 March 2019; Accepted 28 March 2019
Available online 29 March 2019
1010-6030/ © 2019 Elsevier B.V. All rights reserved.

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
source of radical will enhance the photocatalytic activity for oxidizing
4
was stirred for 3 h at 80 °C, and the resulting NH -ZSM-5 was converted
to H type by calcination in air at 540 °C for 4 h [20]. By using the
+
4-CBA. Owing to the good solubility of HPA in the reaction system, the
separation and recovery of HPA will be more difficult. So, the hetero-
geneity of this catalyst with solid support will be very important for
facile catalyst/product separation as well as for improving the appor-
tionment of the active sites for high catalytic activity [12] [13]. HZSM-
impregnation method, PMo of different loadings (5 wt %, 15 wt %, and
25 wt %) and PV Mo (25 wt %) were supported on HZSM-5 by mixing
2
the appropriate amount individually with deionized water via stirring
for 5 h at room temperature. The catalysts were subsequently dried at
110 °C for 10 h.
5
zeolite was applied as support for the Keggin phosphomolybdic (PMo)
and vanadium-containing phosphomolybdic (PV Mo) acids, which en-
2
hanced the photoactivity of the prepared catalysts owing to its ability to
control both electron- and charge-transfer processes. The application of
zeolites stabilizes the charge-transfer state of transient species, such as
2.3. Characterization
The prepared catalysts were characterized via the following analy-
tical techniques. XRD patterns were recorded on an XPERT X-ray dif-
fractometer using CuKα radiation (λ =0.1542 nm). The structural in-
formation of the samples was recorded by using Fourier-transform
infrared (FTIR) spectrometer (Nicolet IS-10) and Raman spectro-
photometer (Brucker RFS 100/S). Field-emission scanning electron
microscopy (FE-SEM; Hitachi S-4800) and transmission electron mi-
croscopy (TEM; Philips, CM 200) were used to determine the mor-
phology and structure of the as-prepared materials. The Brunauer
Emmett and Teller (BET) surface area (S) of the prepared materials was
measured via the adsorption of nitrogen gas at -195.8 °C using a volu-
metric apparatus of the conventional type using a Quantachrome. The
%
−%
OH and O
2
; in addition, its higher acidity plays an important role in
the stability of the Keggin structure [12].
In this work, the use of phosphomolybdic, H
or vanadium-containing phosphomolybdic,
3
PMo12
O
40, (PMo) and/
40, acids
5
H PMo10V O
2
(
2
PV Mo) supported on HZSM-5 in the photocatalytic oxidation process
was reported as a novel method for the removal of 4-CBA by oxidizing it
to terephthalic acid in the presence of PS and low power vacuum UV
(
VUV) light at ambient temperature. Based on the industrial point of
view, this method is more promising than other methods of purifica-
tion, such as thermal treatment, supercritical fluid extraction, post-
oxidation purification, and hydropurification [9,14–16], which are
normally carried out under elevated temperatures and pressures,
compared with our method, which is accomplished under mild condi-
tions.
SBET was estimated from the adsorption curve, and the pore size was
calculated from the BJH desorption branch. The products of the pho-
tocatalytic oxidation of 4-CBA were monitored by an HPLC system
equipped with a pump (waters 515), a sample injector, and a Waters
UV/Visible detector (2489, USA).
2. Experimental
2.1. Materials
2.4. Photo catalytic experiments
All chemicals were of the highest purity available and used as re-
The photocatalytic purification of terephthalic acid (TA) was per-
formed in a glass photoreactor, which was equipped with an internal
vacuum UV lamp (VUV, 185 nm, 6 W, Heraeus) as a light source inside
a synthetic quartz jacket immersed in the TA solution. The reactions
ceived without further purification. Concentrated sulfuric acid (H
Merck), concentrated nitric acid (HNO , Merck), concentrated hydro-
chloric acid (HCl, 36%, Janssen Chimica), diethyl ether, concentrated
phosphoric acid (H PO , 85%, Merck), Na-ZSM-5 (Aldrich), sodium
molybdate dehydrate (Na MoO , Aldrich), vanadium pentoxide (V
Aldrich), molybdenum trioxide (MoO , Merck), and ammonium nitrate
NH NO , Janssen Chimica).
2 4
SO ,
3
3
4
3 2
were carried out in a mixture of CH CN and H O (40:60) as solvent in
−
1
−1
2
4
2
O
5
,
the presence of 10 mg L catalyst and 20 mg L PS as oxidant. First,
the mixture was stirred in dark for 30 min to establish ad-
sorption‒desorption equilibrium. Then, the mixture was irradiated
under the previously described conditions. During the course of the
experiments, samples were periodically withdrawn, filtered, and then
analyzed. The variation in the concentration of 4-CBA and TA as a
function of the reaction time was measured using HPLC. The con-
centration change was calculated using the area percentage method.
3
(
4
3
2
.2. Synthesis of catalytic materials
The Keggin (PMo) was prepared using the ‘etherate method’ [17].
Briefly, 5 mL of H PO (85%) and 50 mL of HCl (36%) were added to a
solution of 45.55 g Na MoO in 100 ml deionized water. The resulting
solution was then transferred into a 1 L dropping funnel containing
5 ml of diethyl ether with shaking. After letting the solution stand still
3
4
2
4
3. Results and discussion
7
for 10–15 minutes, three layers were formed. The bottom layer was
transferred into another funnel with 50 mL of deionized water with
shaking, followed by the addition of 50 mL of HCl and 75 mL of diethyl
ether, with the resulting solution being shaken and left undisturbed
until three layers were again formed. The bottom layer was again
transferred to another dropping funnel, and the same procedure was
followed until the bottom layer was perfectly clear. The clear solution
was then transferred into a beaker containing 12.5 mL of deionized
3.1. Catalyst characterization
2
As shown in the SEM images in Fig. 1a, PV Mo/HZSM-5 has a ty-
pical hexagonal-shaped MFI crystal structure with crystal sizes of about
1–2 μm in length. The TEM images confirmed that the MFI crystalline
structure of the sample consisted of a micro‒mesoporous channel
structure. The XRD patterns of HZSM-5, PMo, PV
Keggin catalysts (PMo/HZSM-5 and PV Mo/HZSM-5) with different
loadings are shown in Fig. 2. The unsupported PMo and PV Mo samples
2
Mo, and supported
2
water, a few drops of HNO
3
were added into it, and the resulting so-
2
lution was evaporated to produce the PMo crystals, which were filtered,
washed, dried, and stored [18].
exhibit characteristic XRD peaks at 9, 10.3, 19.5, and 28.4° 2θ that
matched well with the reported Keggin structure of PMo, highlighting
the crystalline phase obtained from the synthesis method [21,22]. The
The (PV
.01 mol of V
heated up to 120 °C with strong stirring. 0.01 mol of 85% H
2
Mo) was prepared by dissolving 0.10 mol of MoO
in 250 mL deionized water. The resulting solution was
PO was
3
and
0
2
O
5
peaks corresponding to PMo and Pv
2
Mo were not observed on the XRD
Mo catalysts. The ab-
3
4
patterns of the HZSM-5-supported PMo and PV
2
added to the solution, which was then refluxed for 24 h. The produced
powder washed with deionized water, cooled, and dried for further
purification [19].
sence of these peaks prove that the units of POMs are highly dispersed
in the pores of HZSM-5, which was confirmed by the elemental map-
ping images in Fig. 1 that show the good dispersion of POMs. Fur-
thermore, the two characteristic peaks of HZSM-5 that are located at
7–9 and 23‒25° 2θ are well maintained in all the samples highlighting
the limited disturbance of the zeolite framework of HZSM-5 even after
HZSM-5 zeolite with micro‒mesoporous structure was prepared by
the conversion of Na-ZSM-5 into NH
method. 50 mL of 1 M ammonium nitrate containing 5 g of Na type
4
-ZSM-5 using ion exchange
+
174

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
2
Fig. 1. (a) SEM images, (b and c) TEM images, and elemental mapping images of PV Mo/HZSM-5.
Fig. 3. FTIR spectra of PMo supported on HZSM-5 at various loadings, 25%
PV Mo/HZSM-5, and HZSM-5.
2
2
Fig. 2. XRD patterns of different load of PMo/HZSM-5, 25% PV Mo/HZSM-5,
and HZSM-5.
−
1
and 1230 cm−1 were assigned
Si–O–Si, while the bands at 1099 cm
to the Si–O–Si asymmetric stretching [25]. The existence of absorption
bands at 1625 cm
−1
-1
and 3445 cm implies the bending vibrations of
HeOH (physically adsorbed water) and the existence of the hydroxyl
groups of the HZSM-5 structure, respectively [26]. Fig. S1 shows the
Mo, which revealed four imprint
region centered at 1080, 990, 870,
and 790 cm , which are attributed to the central phosphorous species
PeO) and the three Mo–O species, namely, Mo–O
sharing oxygen), Mo–O –Mo (O = terminal oxygen), and Mo–O
= edge sharing oxygen), respectively [27]. The FTIR spectra of the
supported POM catalysts highlight the weak intensities for their
the impregnation of POM under our preparative conditions [23].
The FTIR spectra of POM in the loaded forms (PMo/HZSM-5,
2
PV Mo/HZSM-5, and the pure HZSM-5) are shown in Fig. 3. In the
FTIR spectra of pure PMo and PV
2
spectrum of HZSM-5, the bands were observed at 450, 550, 796, 1230,
−1
bands in the 1100 and 700 cm
−
1
1
625, and 3445 cm , which coincide with those reported in the lit-
−1
−
1
erature [13]. The band concerning 450 cm is relevant to the internal
tetrahedron units of AlO
(
c
–Mo (O
c
= corner
−
1
4 4
and SiO , and the band at 550 cm confirms
t
t
e
–Mo
the preparation of HZSM-5 with five-membered rings [24]. The band
(O
e
−1
centered at 796 cm
was assigned to the symmetric stretching of
175

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
Fig. 5. N
2
gas adsorption–desorption isotherms, the inset figure presents the
pore size distribution of HZSM-5, different loadings of PMo supported on
HZSM-5, and 25% PV Mo/HZSM-5.
2
Table 1
Specific surface area, pore volume, and pore diameter of HZSM-5 and (PMo or
PV Mo)/HZSM-5.
2
Fig. 4. Raman spectra of different loadings of PMo supported on HZSM-5, 25%
Catalyst
Surface area
Pore volume
Pore Diameter
(nm)
PV
2
Mo/HZSM-5, and HZSM-5.
2
−1
3 −1
(m
g
)
(cm
g
)
HZSM-5
377.1
2.46
2.26
1.74
1.64
1.74
3.8
3.8
3.8
3.8
3.8
characteristic bands at the low loading percentage of 5%, but the in-
tensities of these bands were clearly observed to increase with in-
creasing loading percentage. The appearance of these bands as finger-
print for POM strongly indicates that the Keggin structure of POM is
completely retained even after impregnation on the support.
5
% PMo/HZSM-5
368.81
284.31
266.60
304.83
15% PMo/HZSM-5
25% PMo/HZSM-5
2
5% PV
5
2
Mo/HZSM-
2
The Raman spectra after the impregnation of PMo and PV Mo on
the HZSM-5 in Fig. 4 show no significant shift for the sharper and more
2
surface area of 5 wt % PMo/HZSM-5 is 368.81 m /g, whereas the sur-
intense bands of the symmetric and asymmetric stretching vibration
2
face area was reduced to 266.60 m /g in the case of 25 wt % PMo/
HZSM-5. In addition, the pore volume decreased from 2.26 m /g to
. The gradual decrease in the surface area and pore volume
can be attributed to the fact that the POMs are deposited inside the
micro-channels and dispersed well on the surface of the HZSM-5.
−1
modes of Mo-O
t
at 995 and 880 cm , respectively. The resolution
3
between these Raman peaks became broader and relatively weaker
with respect to the PMo spectrum. These findings can be attributed to
the H-bonding interaction between the oxygen atom of the Keggin
anion and the hydroxyl groups on the HZSM-5 surface, as observed
before [28,29].
.64 m3
−1
1
g
The nitrogen adsorption‒desorption isotherms of PMo/HZSM-5,
PV Mo/HZSM-5 and the pure HZSM-5 are shown in Fig. 5 to study the
2
3
3
.2. Purification activity
loading effect of PMo on the HZSM-5 structure. According to the IUPAC
classification, the prepared materials were similar to type I isotherm at
low pressure range, which are typical of microporous materials. How-
ever, an increase in the isotherm between relative partial pressures (P/
P0) of 0.2 and 0.4 during adsorption and desorption indicates the ca-
.2.1. VUV purification
Fig. 6a shows the purification of terephthalic acid using VUV irra-
diation only. Water strongly absorbs light in the VUV range at 185 nm.
As a result of the absorption of the VUV radiation, the photochemical
ionization (eqs. 1 and 2) and homolysis of water take place (eq. 3). The
pillary condensation of N
2
in the mesopores, which is characteristic of
•
dissociation of water generates OH radicals with quantum yields, Φ, of
mesoporous materials [30]. Furthermore, it is remarkable that all the
isotherms are of type H3 hysteresis loop between P/ P0 = 0.45 and
.80, which is typical in solids with slit-shaped pores [31]. The inset
0
.045 and 0.33, respectively, for these reactions, which in turn oxidize
the aldehyde to the corresponding carboxylic acid (4-CBA to TA) (Eq.
4)) [29,33].
0
(
figure shows the pore size distribution curves for HZSM-5 and the
HZSM-5-supported PMO catalysts. All the synthesized materials exhibit
a unique peak located at a pore diameter of 3.8 nm, indicating a mi-
cro‒mesopore structure, which highlights that the preparation proce-
dure does not change the primary structure of HZSM-5 [32].
%
−
VUV (185nm) + H
VUV (185nm) + H
2
2
2
O → OH + H+
+
e
(Φ=0.045)
(1)
(2)
(3)
(4)
aq
%
2
O → H O
%
%
%
VUV (185nm) + H
O → OH + H (Φ=0.33)
Table 1 presents the textural parameters of HZSM-5 and POM/
HZSM-5 materials. These parameters include the surface area, pore
diameter, and pore volume. The results revealed that the pure form of
%
4
-CBA + OH → TA
As a result, the concentration of 4-CBA decreases with the increase
2
−1
%
HZSM-5 has the utmost surface area and pore volume (377.1 m
g
in irradiation time due to the continuous generation of OH radicals,
implying that more oxidation is taking place (the concentration of 4-
CBA dropped from 75,000 to 800 ppm after 7 h irradiation). Increasing
the irradiation time beyond 7 h resulted in no significant enhancement
3
−1
and 2.46 cm g , respectively). Gradually increasing the loading per-
cent of POM reduces the microporosity, which in turn results in the
gradual decrease in the pore volume and surface area. For example, the
176

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
Fig. 6. (a) Effect of VUV irradiation and (b) effect of PS and VUV irradiation on the removal of 4-CBA from CTA.
for the removal of 4-CBA.
highest yield of terephthalic acid (94%) as shown in Fig. 7b. By de-
−1
creasing the oxidant dose amount to 10 mg L , the remaining con-
centration of 4-CBA increased to 13,500 ppm, and this can be attributed
−2
3.2.2. VUV/S
2
O
8
purification
•
‾
The results displayed in Fig. 6b show the effectiveness of the VUV/
to the deficiency in the number of produced sulfate radical (SO
4
),
−
8
2
S
2
O
system on the purification process. The data revealed that the
which is responsible for the oxidation of 4-CBA to TA (Fig. 8a). Further
−1
reaction takes place faster than that of the system without PS, and the
reaction time for the oxidation of 4-CBA to the corresponding ter-
ephthalic acid is reduced. Moreover, the remaining amount of 4-CBA
was stable at 300 ppm after 4 h of VUV irradiation with PS, whereas the
remaining amount of 4-CBA for the system using VUV irradiation only
was stable at a higher concentration of 800 ppm after 7 h. The increase
in reactivity is attributed to the photolysis of PS, which produces sulfate
increase in the PS dose (30 mg L ) yielded no significant change in the
removal of 4-CBA as shown in Fig. 8b. This can be explained by the
presence of the competitive reaction (self-coupling) between the gen-
•‾
erated excess SO
4
radicals, which attack each other based on the fol-
lowing equation (eq. 6):
SO• + SO → S
–
•–
O
2–
8
4
4
2
(6)
•
‾
radical anions, SO
4
, based on the following reaction (Eq. (5)) [31]:
reducing the probability of 4-CBA removal. In addition, the reduction in
the yield of terephthalic acid can be attributed to the excessive pho-
tocatalytic degradation which produced undesired products [34].
2
-
•−
•−
S2O +hυ → 2SO
(SO , E0 = 2.60 V)
(5)
8
4
4
•
in addition to the generation of OH radicals from the photolysis of
water (eqs. 1–3). However, these results are still not suitable to produce
the purified terephthalic acid (PTA) with less than 25 ppm of 4-CBA as
impurities. Therefore, the addition of a POM as catalyst in the form of
bulk (homogeneous catalytic purification) or supported (heterogeneous
catalytic purification) may enhance the selective oxidation of 4-CBA to
TA in the presence of UV irradiation and PS.
3.2.5. Heterogeneous catalytic purification
Homogeneous catalytic purification showed poor conversion and
considerable decrease in the selectivity toward the removal of 4-CBA
due to the large number of molybdenum atoms, which provide many
active sites on the oval-shaped molecule that render the catalyst ef-
fective but non-selective [35]. To resolve these issues, heterogeneous
photocatalytic purification reactions were carried out using PMo/
HZSM-5 with different loadings from 5 to 40% as shown in Fig. 9 and
Table 2. It is obvious that the rate of 4-CBA removal varied considerably
depending on the loading percentage. At 5% loading, the remaining
concentration of 4-CBA was 700 ppm after 3 h. Increasing the loading to
15% resulted in 100 ppm of unreacted 4-CBA. Further increasing the
amount of PMo supported on HZSM-5 zeolite to 25% resulted in 100%
removal of 4-CBA (0 ppm remaining) after 2.5 h. It is worth noting that
at 25% loading show highest yield of TA 98.1% compared with the
starting percent 92.5% of CTA. These results are due to the increase in
the total number of active sites of the catalyst available for the reaction
with the increase in the loading percent of PMo. On the other hand,
increasing the catalyst loading beyond 25% did not have any effect on
the removal of 4-CBA (Fig. 9d), but resulted in comparatively lower TA
yields.
3.2.3. Homogeneous catalytic purification
The results of the homogeneous photocatalytic purification of TA by
using different doses of PMo (5, 10, and 20 mg L ) in the presence of
VUV irradiation and 20 mg L
show that the highest rate of 4-CBA removal (600 ppm remaining) was
−1
−
1
PS are presented in Fig. 7. The results
−
1
obtained by using 10 mg L
PMo with 93.92% yield of terephthalic
−1
acid after 2 h. By decreasing the amount of catalyst to 5 mg L , the
rate of 4-CBA removal decreased, resulting in 2600 ppm of 4-CBA re-
maining after 2 h. This is attributed to the decrease in the active sites,
which resulted in the decrease in the oxidation rate of 4-CBA. By in-
creasing the catalyst dose to 20 mg L , the remaining amount of 4-CBA
increased to 4300 ppm, and the yield of terephthalic acid decreased to
−
1
88.7%. These results suggest that at high catalyst concentrations, light
dispersion can occur. This phenomenon leads to decreased photo-
catalytic activity. In addition, terephthalic acid was oxidized to un-
desirable byproducts due to the large number of molybdenum atoms,
which provide many active sites on the oval shape of the molecule,
enhancing the catalyst efficiency and oxidation power but reducing the
selectivity for 4-CBA.
Vanadium-substituted PMo’s have been found to be efficient cata-
lysts for the selective oxidation of aldehyde group. Their unique feature
as oxidation catalysts is due to their tunable redox properties and cat-
alytic performance by changing the substituents. Therefore, rational
design at the molecular and atomic scale can control the active sites.
2
The result in Fig. 10 shows that the 25% PV Mo/HZSM-5 achieved
3
.2.4. Effect of PS dose on the purification process of TA in presence of VUV
higher photocatalytic activity than that of the vanadium-free PMo/
HZSM-5, where it exhibited 100% conversion of 4-CBA (0 ppm 4-CBA
remaining) and the highest yield of PTA (98.2%) in 2 h. This can be
attributed to the decrease in the energy level of the LUMOs due to the
incorporation of vanadium atoms into the Keggin framework, which in
turn increases the oxidizing power of the vanadium-substituted phos-
phomolybdic acid [36].
−1
light and 10 mg L
PMo
To evaluate the effect of oxidant dose on 4-CBA removal, a range of
−
1
different concentrations (10, 20, and 30 mg L ) of persulfate was used
in the presence of VUV irradiation and 10 mg L
data revealed that the optimum oxidant dose is 20 mg L , which
showed the highest rate of 4-CBA removal (600 ppm remaining) and the
−
1
PMo (Fig. 8). The
−1
177

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
Fig. 7. Effect of different PMo catalyst loadings: (a) 5 mg L 1, (b) 10 mg L−1, and (c) 20 mg L−1, in the presence of 20 mg L-1 PS and VUV light on the removal of 4-
−
CBA.
3
.3. Reaction mechanism
with visible and UV light. Under VUV irradiation, POM can generate
photoexcited species (POM*) by electron transfers from the HOMO to
the LUMO leading to the creation of holes in the HOMO (Eq. (7)). The
holes have strong oxidative properties, which enable the oxidization of
the aromatic aldehyde resulting in the multi-electron reduction of
The photocatalytic conversion process of 4-CBA to PTA on the
PV
2
Mo/HZSM-5 catalyst in the presence of PS and VUV light involves
•‾
•‾
%
the generation of SO , O and OH as reactive species. To determine
the main oxidants responsible for the photocatalytic conversion pro-
cess, individual scavengers were added to the reaction medium. Ethanol
4
2
n−
n−
POM* to POM (Eq. (8)). Accumulated electrons on the POM can be
reoxidized by electron acceptors, such as persulfate and water, which
•
‾
%
for holes (h+),
•‾
%
was added as scavenger for both SO and OH, NaHCO
3
4
produce foolproof oxidants (SO and OH, respectively) and return the
4
%
n−
and tert-butyl alcohol (TBA) for OH. Fig. S2 presents the results after
adding the scavengers to the quenching tests. The addition of TBA af-
fected the rate of the 4-CBA removal slightly, implying the minor role of
POM
to its original oxidation state (POM) (Eq. (9)). The reactive
%
‾
•
SO
4
and OH species enhance the photocatalytic conversion process
through the oxidation of 4-CBA to PTA as represented by Eq. (10).
These steps comprise a full photocatalytic cycle [13,36–38] (Scheme 1).
%
OH on the photocatalytic conversion process. Whereas, the addition of
3
both ethanol and NaHCO resulted in a significant decline in the re-
POM + hv → POM*(h + e)
(7)
(8)
moval of 4-CBA, implying the major role of SO• and h on the pho-
‾
+
4
•‾
+
POM*(h+) + 4-CBA → POMn− + PTA
tocatalytic conversion process of 4-CBA. Therefore, SO4 and h are
considered the dominant contributors in this process.
POMn−+ (H
O or PS) → POM + (SO
%‾
and OH)
%
(9)
2
4
Based on these results, a mechanism for the conversion of 4-CBA to
PTA can be proposed. Recently, POMs have received attention as
photochemical active materials and are considered powerful oxidizing
reagents for a huge number of organic compounds upon irradiation
%‾
%
4-CBA + (SO
4
and OH) → PTA
(10)
−
1
and (b) 30 mg/l, in the presence of 10 mg L−1 PMo and VUV light on the removal of 4-CBA.
Fig. 8. Effect of different PMS doses, (a) 10 mg L
178

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
−
1
Fig. 9. Effect of different loadings, (a) 5 wt % (b) 15 wt % (c) 25 wt % (d) 40 wt %, of PMo supported on H-ZSM5 in the presence of 20 mg L PS and VUV light on
the removal of 4-CBA.
Table 2
Comparison between the efficiency of the unsupported PMo and the supported
−
1
catalysts with varying loadings (reaction conditions: catalyst load =10 mg L
PS load =20 mg L , and 6 W VUV lamp).
,
−
1
Catalyst
Remaining 4-CBA
concentration (ppm)
TA yield
(%)
Reaction time
(h)
PMo
600
700
100
0
0
0
93.9
96.2
98.2
98
97
98.2
2
3
3
2
2.5
2
5
1
2
4
2
% PMo/HZSM-5
5% PMo/ HZSM-5
5% PMo/ HZSM-5
0% PMo/ HZSM-5
5% PV
5
2
Mo/ HZSM-
Scheme 1. The photocatalytic cycle of a polyoxometalate (POM) in the pre-
sence of persulfate and VUV light irradiation. POM* = photoexcited POM,
n−
POM
= reduced POM, and PS = persulfate.
3.4. Catalyst reusability
2
To evaluate the reusability of the catalyst, PV Mo/HZSM-5 was
filtered from the reaction suspension after the first cycle, washed, dried,
and then reused for three more consecutive cycles. The recovered cat-
alysts displayed no significant loss of catalytic activity for the oxidation
of 4-CBA, and the conversions obtained for the four consecutive runs
are 98%, 97%, 96%, and 96%, respectively, as shown in Fig. 11. The
reused catalyst was characterized by FTIR spectroscopy. The FTIR
spectrum of the reused catalyst was found to be similar to that of the
fresh catalyst with no appreciable shift in the FTIR spectroscopic band
positions, which indicates the retention of the Keggin polyanions in the
2
Fig. 10. Removal of 4-CBA using 25% loading PV Mo/HZSM-5 in the presence
−
1
of 20 mg L
PS and VUV light.
179

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
concept for catalytic oxidation of p-xylene to produce high-purity terephthalic acid,
Chem. Eng. Sci. 104 (2013) 93–102.
[
[
[
7] R.J. Sheehan, Terephthalic Acid, Dimethyl Terephthalate and Isophthalic Acid.
Ullmann’s Encyclopedia of Industrial Chemistry vol. 35, WileyVCH Verlag GmbH &
Co, Weinheim, 2005, pp. 639–651.
8] N.A.M. Fadzil, M.H.A. Rahim, G.P. Maniam, A brief review of para-xylene oxidation
to terephthalic acid as a model of primary C–H bond activation, Chinese J. Catal. 35
(2014) 1641–1652.
9] S. Tourani, F. Khorasheh, A. Rashidi, A. Safekordi, Hydro-purification of crude
terephthalic acid using palladium catalyst supported on multi-wall carbon nano-
tubes, J. Ind. Eng. Chem. 28 (2015) 202–210.
[
[
10] D.-L. Long, E. Burkholder, L. Cronin, Polyoxometalate clusters, nanostructures and
materials: from self assembly to designer materials and devices, Chem. Soc. Rev. 36
(
2007) 105–121.
11] G. Marcì, E. García-López, V. Vaiano, G. Sarno, D. Sannino, L. Palmisano, Keggin
heteropolyacids supported on TiO used in gas-solid (photo) catalytic propene hy-
2
dration and in liquid-solid photocatalytic glycerol dehydration, Catal. Today 281
(2017) 60–70.
12] Y. Zhou, G. Chen, Z. Long, J. Wang, Recent advances in polyoxometalate-based
heterogeneous catalytic materials for liquid-phase organic transformations, RSC
Adv. 4 (2014) 42092–42113.
13] C.L. Marchena, L. Lerici, S. Renzini, L. Pierella, L. Pizzio, Synthesis and char-
acterization of a novel tungstosilicic acid immobilized on zeolites catalyst for the
photodegradation of methyl orange, Appl. Catal. B 188 (2016) 23–30.
14] A.L. Stautzenberger, Purification of terephthalic acid by thermal treatment, Google
Patents, US3862218 (1975).
15] A.S. Myerson, Purification of terephthalic acid by supercritical fluid extraction,
Google Patents, US4550198 (1985).
16] R.B. Sheppard, B.A. Tennant, T.E. Woodruff, R. Lin, Process for the oxidative pur-
ification of terephthalic acid, Google Patents, 2006.
17] W. Hsein, Contribution to chemistry of phosphomolybdic acid, phosphotungstic
acid and allied substances, J. Biol. Chem. 43 (1920) 189–220.
18] E. Arendt, K. McEvoy, E.M. Gaigneaux, Structural rearrangement and catalytic
2
Fig. 11. Catalytic reusability of PV Mo/H-ZSM-5 for the removal of 4-CBA in
−
1
the presence of 20 mg L
PS and VUV light.
[
reused catalyst even after four cycles (Fig. S3). The SEM of the reused
catalyst also showed no significance change in its morphology after 4
cycles (Fig. S4).
[
[
[
[
[
[
4. Conclusion
Compared to other existing purification methods, which are con-
ducted under high pressures and temperature, the photocatalytic pur-
ification of CTA under mild conditions provides valuable contribution
in the field of terephthalic acid purification. The successful preparation
of pure HZSM-5 and bulk and supported POMs was affirmed via XRD,
which affirmed the crystalline phase of polyoxomolebdate and its good
dispersion inside the mesopores of HZSM-5. FTIR and Raman spectro-
scopy confirmed that the characteristic Keggin ion of HPA remains in-
tact in the HZSM-5. VUV light irradiation and persulfate anion are
important factors in the purification process, which showed higher ef-
ficiency in the presence of the POM/HZSM-5 catalysts. Homogeneous
purification with PMo showed lower activities compared with the
heterogeneous one. In this study, the vanadium-containing PMo/HZSM-
properties of the Wells–Dawson (NH4)
6 2
P Mo18O62 heteropolycompound in the 2-
butanol reaction, Appl. Catal. A Gen. 357 (2009) 115–124.
[19] Y. Leng, J. Wang, D. Zhu, L. Shen, P. Zhao, M. Zhang, Heteropolyanion-based ionic
hybrid solid: a green bulk-type catalyst for hydroxylation of benzene with hydrogen
peroxide, Chem. Eng. J. 173 (2011) 620–626.
20] A. Popa, V. Sasca, D. Bajuk-Bogdanović, I. Holclajtner-Antunović, Acidic nickel salts
of Keggin type heteropolyacids supported on SBA-15 mesoporous silica, J. Porous
Mater. 23 (2016) 211–223.
21] H.R. Ghalebi, S. Aber, A. Karimi, J. Mol. Catal. A Chem. 415 (2016) 96–103.
22] T. Ressler, O. Timpe, F. Girgsdies, J. Wienold, T. Neisius, In situ investigations of the
bulk structural evolution of vanadium-containing heteropolyoxomolybdate cata-
lysts during thermal activation, J. Catal. 231 (2005) 279–291.
23] B. Liu, L. France, C. Wu, Z. Jiang, V.L. Kuznetsov, H.A. Al-Megren, M. Al-Kinany,
S.A. Aldrees, T. Xiao, P.P. Edwards, Methanol-to-hydrocarbons conversion over
[
[
[
[
5
catalyst was considered the most promising catalyst among the syn-
thesized materials for the purification of terephthalic acid by the oxi-
3
MoO /H-ZSM-5 catalysts prepared via lower temperature calcination: a route to
dation of 4-CBA.
tailor the distribution and evolution of promoter Mo species, and their corre-
sponding catalytic properties, Chem. Sci. 6 (2015) 5152–5163.
[
[
24] A.M. Rabie, M.A. Betiha, S.-E. Park, Direct synthesis of acetic acid by simultaneous
Acknowledgement
2
co-activation of methane and CO over Cu-exchanged ZSM-5 catalysts, Appl. Catal.
B 215 (2017) 50–59.
25] Y. Meng, H.C. Genuino, C.-H. Kuo, H. Huang, S.-Y. Chen, L. Zhang, A. Rossi,
S.L. Suib, One-step hydrothermal synthesis of manganese-containing MFI-type
zeolite, Mn–ZSM-5, characterization, and catalytic oxidation of hydrocarbons, J.
Am. Chem. Soc. 135 (2013) 8594–8605.
This study was supported by the National Research Foundation
(
NRF) of the Republic of Korea under the frameworks of Priority
Research Centers Program (NRF-2014R1A6A1031189) and Basic
Science Research Program (NRF-2015R1D1A1A09060292), funded by
theMinistry of Education of Korea, and International Cooperation
Program (NRF-2015K2A2A7053101).
[
26] S. Zhang, X. Liu, Y. Zhang, T. Lv, J. Zheng, W. Gao, X. Liu, M. Cui, C. Meng, Study
on the synthesis of MFI and FER in the presence of n-butylamine and the property of
n-butylamine in a confined region of zeolites, RSC Adv. 6 (2016) 114808–114817.
[27] V. Balaga, J. Pedada, H.B. Friedrich, S. Singh, Tuning surface composition of Cs
exchanged phosphomolybdic acid catalysts in CH bond activation of toluene to
benzaldehyde at room temperature, J. Mol. Catal. A Chem. 425 (2016) 116–123.
Appendix A. Supplementary data
[
28] A.K. Cuentas, C. Frausto, L.A. Ortiz-Frade, G. Orozco, Vib. Spectrosc. 57 (2011)
4
9–54.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.jphotochem.2019.03.
[29] F. Jin, M. Wei, C. Liu, Y. Ma, The mechanism for the formation of OH radicals in
condensed-phase water under ultraviolet irradiation, J. Chem. Soc. Faraday Trans.
1
9 (2017) 21453–21460.
048.
[
30] M. Thommes, K. Kaneko, A.V. Neimark, J.P. Olivier, F. Rodriguez-Reinoso,
J. Rouquerol, K.S. Sing, Physisorption of gases, with special reference to the eva-
luation of surface area and pore size distribution (IUPAC Technical Report), Pure
References
Appl. Chem. 87 (2015) 1051–1069.
•
−•, and
[
31] X.-Y. Yu, Z.-C. Bao, J.R. Barker, Free radical reactions involving Cl , Cl
2
2-
−
4
•
[
1] S. Tourani, M. Baghalha, F. Khorasheh, A. Behvandi, Equilibrium modeling of xy-
lene adsorption on molecular sieves, Fluid Phase Equilib. 298 (2010) 54–59.
2] S. Tourani, A. Safekordi, M. Rashidzadeh, A. Rashidi, F. Khorasheh, Hydrogenation
of crude terephthalic acid by supported Pd and Pd–Sn catalysts on functionalized
multiwall carbon nanotubes, Chem. Eng. Res. Des. 109 (2016) 41–52.
SO
2 8
in the 248 nm photolysis of aqueous solutions containing S O and Cl, J.
Phys. Chem. A 108 (2004) 295–308.
[
[32] A. Kurhade, A.K. Dalai, Physiochemical characterization and support interaction of
alumina‐supported heteropolyacid catalyst for biodiesel production, Asia-Pac, J.
Chem. Eng. 13 (2018) e2249.
[3] G. Ciamician, P. Silber, Chemische Lichtwirkungen. XXII. Autooxydationen, I,
European Journal of Inorganic Chemistry 45 (1912) 38–43.
[33] K. Zoschke, H. Börnick, E. Worch, Vacuum-UV radiation at 185 nm in water
treatment-a review, Water Res. 52 (2014) 131–145.
[
4] H. Stephens, Oxidations in the benzene series by gaseous oxygen ii. Alkyl benzenes
with two or more carbon atoms in the side chain, J. Am. Chem. Soc. 48 (1926)
[34] A. Shafaei, M. Nikazar, M. Arami, Photocatalytic degradation of terephthalic acid
using titania and zinc oxide photocatalysts: comparative study, Desalination 252
(2010) 8–16.
2
920–2922.
[
5] J. Scheirs, T.E. Long, Modern Polyesters: Chemistry and Technology of Polyesters
and Copolyesters, John Wiley & Sons, 2005.
6] M. Li, F. Niu, X. Zuo, P.D. Metelski, D.H. Busch, B. Subramaniam, A spray reactor
[35] L. Xu, G. Wang, F. Ma, Y. Zhao, N. Lu, Y. Guo, X. Yang, Photocatalytic degradation
of an aqueous sulfamethoxazole over the metallic silver and Keggin unit codoped
titania nanocomposites, Appl. Surf. Sci. 258 (2012) 7039–7046.
[
180

A.M. Hussein, et al.
Journal of Photochemistry & Photobiology A: Chemistry 377 (2019) 173–181
[
36] W. JinYue, F.X. Hong, W. Jian, H.C. We, Effects of vanadium(V)-substitution on the
oxidative properties of α-Keggin-type heteropolyanion clusters—progress in DFT
theoretical studies, Sci China Ser B-Chem 52 (2009) 2096–2105.
[38] J. Zhang, Y. Tang, G. Li, C. Hu, Room temperature direct oxidation of benzene to
phenol using hydrogen peroxide in the presence of vanadium-substituted hetero-
polymolybdates, Appl. Catal. A Gen. 278 (2005) 251–261.
[37] M. Sun, J. Zhang, P. Putaj, V. Caps, Fdr. Lefebvre, J. Pelletier, J.-M. Basset, Catalytic
1 4
oxidation of light alkanes (C –C ) by heteropoly compounds, Chem. Rev. 114
(2013) 981–1019.
181