Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres: Efficient Heterogeneous Catalysts for Acetalization of Benzaldehyde with Ethylene Glycol

Article abstract of DOI:10.1007/s10562-019-03029-6

Abstract: In this study, magnetically-recoverable core–shell catalysts with different amount of H₃PW₁₂O₄₀ loading [Fe₃O₄@C@mTa₂O₅-NH₂-PW₁₂ (w%)] were prepared by the application of phosphotungstic acid supported on amino group functionalized magnetic core–shell mesoporous tantalum pentoxide microspheres. The prepared samples were characterized by FT-IR, N₂-adsorption–desorption isotherms, TEM, SEM, Pyridine-IR analysis, XRD and magnetism. Fe₃O₄@C@mTa₂O₅-NH₂-PW₁₂ samples present both Br?nsted and Lewis acidity, large BET surface area and high magnetization. The catalytic activity was evaluated by the acetalization of different aldehydes with diols, and the results show that Fe₃O₄@C@mTa₂O₅-NH₂-PW₁₂ (14.47%) catalyst exhibits the highest catalytic activity for acetalization of aldehydes with glycols with 94.5% conversion of benzaldehyde and 99% selectivity to benzaldehyde glycol acetal at 80?°C. The catalytic activity of the catalyst for acetalization is related to its total acidity and Br?nsted–Lewis acid synergy. The catalyst Fe₃O₄@C@mTa₂O₅-NH₂-PW₁₂ can be easily recovered and reused for at least 5 times without obvious decrease of catalytic activity.

Full text of DOI:10.1007/s10562-019-03029-6

Catalysis Letters
https://doi.org/10.1007/s10562-019-03029-6
Phosphotungstic Acid Supported on Magnetic Mesoporous
Tantalum Pentoxide Microspheres: Efcient Heterogeneous Catalysts
ꢀor Acetalization oꢀ Benzaldehyde with Ethylene Glycol
1
1
1
1
1
1
Caiting Feng · Qinlong Li · Panfeng Wu · Bin Liu · Huaiming Hu · Ganglin Xue
Received: 22 July 2019 / Accepted: 2 November 2019
©
Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
In this study, magnetically-recoverable core–shell catalysts with diferent amount oꢀ H PW O loading [Fe O @C@
3
12 40
3
4
mTa O -NH -PW (w%)] were prepared by the application oꢀ phosphotungstic acid supported on amino group ꢀunctional-
2
5
2
12
ized magnetic core–shell mesoporous tantalum pentoxide microspheres. The prepared samples were characterized by FT-IR,
N -adsorption–desorption isotherms, TEM, SEM, Pyridine-IR analysis, XRD and magnetism. Fe O @C@mTa O -NH -
2
3
4
2
5
2
PW samples present both Brönsted and Lewis acidity, large BET surꢀace area and high magnetization. The catalytic activity
12
was evaluated by the acetalization oꢀ diferent aldehydes with diols, and the results show that Fe O @C@mTa O -NH -PW
3
4
2
5
2
12
(
14.47%) catalyst exhibits the highest catalytic activity ꢀor acetalization oꢀ aldehydes with glycols with 94.5% conversion oꢀ
benzaldehyde and 99% selectivity to benzaldehyde glycol acetal at 80 °C. The catalytic activity oꢀ the catalyst ꢀor acetali-
zation is related to its total acidity and Brönsted–Lewis acid synergy. The catalyst Fe O @C@mTa O -NH -PW can be
3
4
2
5
2
12
easily recovered and reused ꢀor at least 5 times without obvious decrease oꢀ catalytic activity.
Electronic supplementary material The online version oꢀ this
article (https://doi.org/10.1007/s10562-019-03029-6) contains
supplementary material, which is available to authorized users.
*
Ganglin Xue
xglin707@163.com
1
College oꢀ Chemistry & Materials Science, Northwest
University, 1 Xueꢀu Ave., Guodu, Chang’an District,
Xi’an 710127, Shaanxi, People’s Republic oꢀ China
Vol.:(0
1
123456789)
3

C. Feng et al.
Graphic Abstract
Keywords Heteropoly acid · Ta O · Recyclable catalysts · Brönsted/Lewis acid · Acetalization
2
5
1
Introduction
resulting in the diꢂculty oꢀ catalyst recovery. Thereꢀore,
immobilizing HPAs onto diferent supports with high sur-
ꢀace area, such as silica [13], titania [14], zeolites [15],
polymers [16], and active carbons [17], is an efective
approach ꢀor successꢀul applications in heterogeneous
catalysis. Meanwhile, the type oꢀ support and the extent
oꢀ loading can greatly afect the acidity and catalytic activ-
ity oꢀ the supported HPAs. For example, combining HPAs
with acidic supports, like Lewis metal, the ionic liquid and
MOF, the composite materials would possess both Brön-
sted and Lewis acid centers, and the total acidity and Brön-
sted/Lewis acid site ratio could also be adjusted, thereꢀore
ꢀurther afecting the catalytic activity [18–21].
The acetalization oꢀ aldehydes with diols is a widely used
method in organic chemistry [1]. Acetals are commonly
used as solvents, intermediates or ꢁnal products due to their
unique properties required in practical industrial productions
[
2, 3]. Moreover, acetals are excellent protection groups ꢀor
carbonyl groups during multi-step organic synthesis [4]. The
currently used catalysts ꢀor the acetalization include Lewis
or Brönsted acids, such as hydrochloric acid, sulꢀuric acid,
acidic polymers [5], metal catalysts [6], MOFs [7, 8], and
solid acids [9]. However, most oꢀ the those catalysts sufer
ꢀ
rom some disadvantages to a certain extent, such as diꢂ-
culty oꢀ catalyst recovery, low activity, low selectivity, harsh
conditions, unavoidable equipment corrosion and environ-
mental problems [10]. Thereꢀore, it is imperative to develop
reusable, eꢂcient and environmental-benign catalysts ꢀor
the production oꢀ acetals.
It is well known that the process oꢀ the acetalization
can be efectively promoted by the synergistic efect oꢀ
Brönsted (B) and Lewis (L) acids in a catalyst system
[22–24]. For example, Ni ion-exchanged tungstophos-
phoric acid-based ionic liquid catalyst (Ni[MIMPSH]
PW12O40, MIMPS = methylimidazolium propyl sulꢀobe-
taine) was revealed to be very efective in the acetaliza-
tion oꢀ aldehydes with glycols, afording 94.6% oꢀ acetal
Heteropoly acids (HPAs), especially 12-tungstophos-
phoric acid (HPW), which has the super strong Brönsted
acidity, have been used as solid-acid catalysts ꢀor esteriꢁ-
cation, hydration, alkylation and hydrolysis [11, 12]. How-
ever, most oꢀ HPAs have high solubility in polar medium,
2
−
yield [25]. SnO2-based solid acids catalyst (SO4 /SnO2)
1
3

Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres:…
showed high reactivity and selectivity in the acetalization
oꢀ glycerol with acetone, ꢀurꢀural and its derivatives [26].
As acidic metal oxides, Ta O was applied to the acid
2 Result and Discussion
2
5
2.1 Synthesis and Characterization of Fe O @C@
3
4
catalysis due to its Lewis acidity. For example, hetero-
geneous catalyst Ta O + Pt/SiO was reported to be an
mTa O ‑NH ‑PW Catalyst
2 5 2 12
2
5
2
excellent acid catalyst on the conversion oꢀ 5-hydroxy-
methylꢀurꢀural to 3-hydroxymethylcyclopentanone [27].
Xu et al. prepared mesoporous heteropolyacids-tantalum
pentoxide hybrid catalyst (H PW O /Ta O ), which
The Fe O magnetic microspheres (ca. 300 nm) were
3 4
synthesized via the solvothermal method. Aꢀterwards,
Fe O @C microspheres were prepared through hydro-
3
4
3
12 40
2
5
thermal reaction oꢀ glucose on Fe O microspheres. The
3
4
showed an excellent catalytic perꢀormance in esteriꢁca-
tion and transesteriꢁcation under mild conditions [28].
Inspired by these ꢀacts, it is rational to prepare Brönsted/
Lewis acid biꢀunctional composite catalysts by pairing
Ta O with HPW such that the catalyst simultaneously
Fe O @C microspheres coated with a thin mesoporous
3 4
tantalum pentoxide layer, Fe O @C@mTa O , could be
3
4
2
5
obtained through a surꢀactant-templating sol–gel approach
by the treatment oꢀ Fe O @C with tantalum ethoxide
3
4
2
5
using cetyltrimethylammonium bromide as template agent
oꢀ mesopores in ethanol–water solution. In order to obtain
Fe O @C@mTa O -NH , the as-prepared Fe O @C@
contains Brönsted acid site ꢀrom HPW and Lewis acid
site ꢀrom Ta O in the composite structure. The prepared
2
5
3
4
2
5
2
3
4
catalysts might be very efective ꢀor the acetalization oꢀ
aldehydes with glycols. Although the HPAs-based com-
posites have higher acid catalytic activity, the separation
process is usually regarded as time-consuming and energy-
intensive, and stably immobilizing HPAs onto supports is
still challenging.
mTa O microspheres were covalently modiꢀied with
2 5
amino groups via reacting with aminopropyltriethoxysi-
lane. As previous reports, amino groups ꢀunctionalized
silica may combined with HPAs to construct a stable
catalyst in two ways:one is the electrostatic interaction
oꢀ protonated amino groups with Keggin anions and the
interaction oꢀ the protonated hydroxyl groups oꢀ the sil-
ica surꢀace with Keggin anions; another is the exchange
reaction oꢀ the protonated hydroxyl groups oꢀ the silica
surꢀace with external oxygen atoms oꢀ the Keggin anions
[36, 37]. Similarly, amino groups and hydroxyl groups on
shell oꢀ tantalum pentoxide may also provide suitable sites
interacting with HPAs ions to construct a stable catalyst
Fe O @C@mTa O -NH -PW . The diferent amounts
In recent years, magnetic nanoparticles (MNPs) have
been used as a kind oꢀ catalyst support because the mag-
netic separation is ꢀast, simple and environmentally
ꢀ
riendly [29]. The surꢀace oꢀ MNPs was usually coated
with SiO , ZrO , polymer or other materials to ꢀorm mag-
2
2
netic core–shell nanoparticles, and ꢀurther used as a sup-
port ꢀor HPAs. Various HPAs-based MNPs have been pre-
pared and applied to acid catalysis [30–35]. The catalytic
activity and stability oꢀ these HPAs-based magnetic com-
posite materials depend mainly on the type oꢀ supports,
the extent oꢀ loading, and the interaction modes between
supports and HPAs.
3
4
2
5
2
12
oꢀ HPW were immobilized on Fe O @C@mTa O -NH
3
4
2
5
2
magnetic microspheres in order to check which one has
the highest catalytic activity. The determined H PW O
3
12 40
loading is ca. 8.51, 14.47, 21.20, 23.41% according
to the ICP result. The ꢀollowing results indicated that
Fe O @C@mTa O -NH -PW (14.47%) has the highest
In this work, ꢀor the ꢁrst time, we successꢀully ꢀabricated
the amino group ꢀunctionalized mesoporous tantalum pen-
toxide magnetic core–shell microspheres, which were ꢀurther
used as the support ꢀor immobilizing H PW O . Magnetic
3
4
2
5
2
12
catalytic eꢂciency. The schematic preparation process oꢀ
3
12 40
H PW O graꢀted Fe O @C@mTa O -NH magnetic
3
12 40
3
4
2
5
2
core–shell mesoporous tantalum pentoxide microspheres
mesoporous microspheres is shown in Scheme 1. To obtain
more inꢀormation about the structural characteristics, the
as-synthesized catalyst was studied through FT-IR, SEM,
have high pore volume and large surꢀace area, H PW O
3
12 40
was ꢁrmly ꢁxed on the surꢀaces oꢀ mesoporous tantalum
pentoxide through more amino groups and hydroxyl groups
exposed on shell oꢀ tantalum pentoxide. A series oꢀ mag-
netically-recoverable core–shell catalysts with diꢀꢀerent
amount oꢀ HPW loading [Fe O @C@mTa O -NH -PW
TEM, XRD, N -adsorption–desorption isotherms, magnet-
2
ism, EDX analysis, potentiometric titration and pyridine-
IR analysis.
3
4
2
5
2
12
The FT-IR spectra oꢀ Fe O , Fe O @C@mTa O ,
3
4
3
4
2
5
(
w%)] were synthesized. By changing the amount oꢀ HPW
Fe O @C@mTa O -NH , Fe O @C@mTa O -NH -
3 4 2 5 2 3 4 2 5 2
immobilized, the site ratio oꢀ Brönsted and Lewis acidity
could be adjusted, and the relationship between the acid-
ity oꢀ the catalysts and catalytic activity in the acetalization
oꢀ benzaldehyde with ethylene glycol was investigated. A
possible mechanism ꢀor the Brönsted/Lewis acid-catalyzed
acetalization oꢀ aldehyde with diols was proposed. In addi-
tion, the reusability oꢀ the catalyst was also studied.
PW (14.47%) and H PW O are presented in Fig. 1.
12 3 12 40
−
1
For Fe O , the peak at 584 cm assigns to Fe–O stretch-
3
4
ing vibration (Fig. 1a). In addition, the broad and intense
−
1
−1
bands around 3400 cm and 1640 cm are assignable
to O–H stretching vibration and H–O–H bending vibra-
tion oꢀ water at the surꢀace oꢀ Fe O , respectively. Aꢀter
3
4
1
3

C. Feng et al.
Scheme 1 Schematic illustra-
tion ꢀor the preparation oꢀ
Fe O @C@mTa O -NH -PW
12
3
4
2
5
2
catalyst
Fig. 2 SEM images oꢀ a Fe O , b Fe O @C, c Fe O @C@mTa O ,
3
4
3
4
3
4
2
5
d Fe O @C@mTa O -NH -PW (14.47%)
3
4
2
5
2
12
Fig. 1 The FT-IR spectral pattern oꢀ various materials: (a) Fe O , (b)
3
4
4
Fe O @C@mTa O , (c) Fe O @C@mTa O -NH , (d) Fe O @C@
3
4
2
5
3
4
2
5
2
3
mTa O -NH -PW (14.47%), and (e) H PW O
12 40
2
5
2
12
3
vibration and H–O–H symmetrical bending vibration.
All oꢀ these veriꢁed the amino-ꢀunctionalization oꢀ the
Ta O shell (Fig. 1c). Concurrently, ꢀor the HPAs immo-
coating with mesoporous Ta O shell, there appears a
2
5
2
5
−
1
−1
wide broad peak at 633 cm (Fig. 1b), which was attrib-
bilized samples, the ꢀour bands appeared at 1083 cm
−
1
−1
uted to the superimposition oꢀ the characteristic absorp-
ν (P–O), 984 cm ν (W=O), 892 cm ν (W–O –W)
as
as
as
b
−
1
−1
tion band oꢀ Ta–O vibration appeared at 660 cm and
and 804 cm ν (W–O –W) (Fig. 1d) could be attributed
as c
−
1
Fe–O stretching vibration appeared at 584 cm . When
the surꢀace was ꢀurther treated with 3-aminopropyltri-
ethoxysilane, another two new bands appear at 1517 and
to characteristic peaks oꢀ H PW O (Fig. 1d). The peak
3 12 40
intensity oꢀ diferent amount oꢀ HPW immobilized cata-
lysts increases with the HPW amount (Fig. S1). The FT-IR
spectra indicated that the surꢀace oꢀ the Fe O magnetic
−
1
−1
1
225 cm , the peak at 1225 cm is originated ꢀrom the
3
4
−
1
C–N stretching vibration, a new weak peak at 1517 cm
microspheres have been successꢀully coated and ꢀunction-
−
1
+
and the peak at 1630 cm are ascribed to –NH bending
alized, and ꢀurther immobilized with H PW O .
3
3
12 40
−
1
−1
vibration. Besides, the peaks at 3430 cm and 1630 cm
The morphology oꢀ the as-synthesized catalysts was char-
acterized by SEM and TEM. Figure 2 shows SEM images
oꢀ Fe O , Fe O @C, Fe O @C@mTa O and Fe O @C@
are strengthened due to N–H stretching vibration and N–H
bending vibration, which coincide with O–H stretching
3
4
3
4
3
4
2
5
3
4
1
3

Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres:…
Fig. 4 SEM images oꢀ Fe O @C@mTa O -NH -PW using difer-
3
4
2
5
2
12
ent ethanol/water ratio(v/v): a 10:1, b 5:1, c 1:1, and d1:5
Fe O @C@mTa O -NH -PW microspheres. The ratio
3
4
2
5
2
12
oꢀ ethanol/water (v/v) afects the hydrolysis rate oꢀ tan-
talum ethanol and ꢀurther afects the morphology oꢀ the
Ta O shell. As the water content in the ethanol/water mix-
Fig. 3 TEM images oꢀ a Fe O , b Fe O @C, c Fe O @C@mTa O ,
3
4
3
4
3
4
2
5
d Fe O @C@mTa O -NH -PW (14.47%)
3
4
2
5
2
12
2
5
ture increases, the hydrolysis rate oꢀ tantalum ethanol will
become ꢀaster. It has been conꢁrmed by the TEM images
oꢀ Fe O @C@mTa O -NH -PW obtained using diferent
mTa O -NH -PW (14.47%) microspheres, respectively,
2
5
2
12
indicating that the products are all well dispersed with near-
spherical morphology. Figure 3 shows typical TEM images
oꢀ Fe O , Fe O @C, Fe O @C@mTa O and Fe O @C@
3
4
2
5
2
12
ratio oꢀ ethanol/water (v/v) (Fig. 4). When the ethanol/water
ratio (v/v) is 10:1, almost no Ta O shell was observed, the
3
4
3
4
3
4
2
5
3
4
2
5
mTa O -NH -PW (14.47%) microspheres, proving that the
morphology is similar to Fe O @C (Figs. 2b and 4a). When
2
5
2
12
3 4
coating process is layer-by-layer and uniꢀorm, Fe O @C@
the ethanol/water ratio (v/v) is 5:1, the Ta O shell composed
2 5
3
4
mTa O -NH -PW microspheres possess a sandwich-
oꢀ dense Ta O nanoparticles was ꢀormed (Fig. 4b). How-
2 5
2
5
2
12
like core–shell structure. As shown in Figs. 2a and 3a, the
ever, with the increase oꢀ the water content in the ethanol/
water mixture, the ratio oꢀ ethanol/water (v/v) achieved 1:1,
the Ta O shell is looser and there are small particles on the
obtained Fe O microspheres are uniꢀorm with a mean size oꢀ
3
4
~
300 nm. Aꢀter coated with a thin carbon layer (ca.10 nm),
2
5
the obtained Fe O @C microspheres show smoother surꢀace
surꢀace oꢀ the magnetic microspheres, which may be easy
to ꢀall of and aggregate, probably resulting in a decrease in
catalytic activity when the catalyst is recycled (Fig. 4c, d).
Thereꢀore, it is concluded that adjusting the ratio oꢀ etha-
nol/water and optimizing the appropriate ratios oꢀ ethanol/
water (v/v) are beneꢁcial to construct stable Fe O @C@
3
4
than Fe O microspheres and become a little larger (Figs. 2b
3
4
and 3b). Aꢀter the surꢀactant-assembly sol–gel process, a
uniꢀorm mesoporous Ta O shell (ca.30 nm) was coated on
2
5
Fe O @C microspheres, the obtained microspheres show
3
4
rougher surꢀace (Figs. 2c and 3c). Aꢀter immobilized HPW
3
4
onto Fe O @C@mTa O , the morphology was well kept
mTa O -NH -PW microspheres, and the optimum ratios
3
4
2
5
2
5
2
12
(
Figs. 2d and 3d). The energy dispersive X-ray spectros-
oꢀ ethanol/water (v/v) is 5:1.
copy (EDX) oꢀ the Fe O @C@mTa O -NH -PW (14.47%)
Wide-angle XRD patterns oꢀ Fe O , Fe O @C@mTa O ,
3
4
2
5
2
12
3
4
3
4
2
5
microspheres exhibit the characteristic peaks oꢀ Fe, O, C,
Fe O @C@mTa O -NH -PW (14.47%) and H PW O
3 4 2 5 2 12 3 12 40
Ta, N, P and W (Fig. S2), suggesting that the Fe O cores
are shown in Fig. 5A. The difraction peaks oꢀ the obtained
Fe O microspheres agree with the standard Fe O (JCPDS
3
4
are successꢀully coated by mesoporous Ta O shell and
2
5
3
4
3
4
ꢀ
urther immobilized H PW O . The average diameters
card No. 19-0629). For the wide-angle XRD patterns oꢀ
3
12 40
oꢀ Fe O , Fe O @C, Fe O @C@mTa O and Fe O @C@
Fe O @C@mTa O and Fe O @C@mTa O -NH -PW
3
4
3
4
3
4
2
5
3
4
3
4
2
5
3
4
2
5
2
12
mTa O -NH -PW (14.47%) microspheres represented
(14.47%) (Fig. 5A—b, c), the characteristic difraction peaks
2
5
2
12
in Fig. S3 are ca. 304.45, 314.08, 376.26 and 382.07 nm,
respectively (close to the size measured by TEM), which
provide additional evidence ꢀor the successꢀul coating.
It is worth noting that the ratio oꢀ ethanol/water (v/v) is
an important ꢀactor ꢀor the preparation oꢀ the core–shell
oꢀ Fe O are weakened than the original Fe O because oꢀ the
3
4
3
4
carbon and mesoporous Ta O shell; Meanwhile, two typical
2
5
broad band peaks are observed in the range oꢀ 2θ=15°–40°
and 45°–65°, which originated ꢀrom amorphous Ta O [28].
2
5
The pattern oꢀ Fe O @C@mTa O -NH -PW (14.47%) is
3
4
2
5
2
12
1
3

C. Feng et al.
Fig. 5 Typical Wide-angle (A) and Low-angle (B) PXRD patterns oꢀ (a) Fe O , (b) Fe O @C@mTa O , (c) Fe O @C@mTa O -NH -PW
12
3
4
3
4
2
5
3
4
2
5
2
(
14.47%), (d) H PW O
3 12 40
similar to that oꢀ Fe O @C@mTa O and no difraction
cancellation between X-rays scattered ꢀrom the walls and the
pore regions, and thereꢀore reduces the intensities oꢀ XRD
peaks [39].
3
4
2
5
peaks corresponding to H PW O are observed in as-pre-
3
12 40
pared catalysts because oꢀ high dispersion oꢀ the HPW on
Fe O @C@mTa O (Fig. 5A—c).
The surꢀace area and porous structure oꢀ Fe O @C,
3
4
2
5
3
4
Figure 5B shows the Low-angle XRD patterns oꢀ
Fe O @C@mTa O and Fe O @C@mTa O -NH -PW
3 4 2 5 3 4 2 5 2 12
Fe O @C@mTa O and Fe O @C@mTa O -NH -PW
(14.47%) were examined by N adsorption measurement.
3
4
2
5
3
4
2
5
2
12
2
(
(
14.47%). In these two curves, a broader reꢃection peak
Figure 6 shows the N adsorption–desorption isotherms. It
2
100) indicates the ꢀormation oꢀ mesoporous structure [28].
was ꢀound that the pore size oꢀ Fe O @C is uneven, and the
3
4
Compared to Fe O @C@mTa O , the peak oꢀ Fe O @C@
average pore diameter is mainly at 1.4 nm, the BET surꢀace
3
4
2
5
3
4
2
−1
mTa O -NH -PW (14.47%) becomes broader and the
area is calculated to be 10.1 m g . Compared to Fe O @C,
3 4
2
5
2
12
difraction intensity decreases simultaneously. The reason
or this phenomenon is that HPW was ꢁlled in the pores
and scattered X-rays [38]. This results in increased phase
the other two samples show type-IV curves, indicating their
mesoporous structure. Aꢀter coating with mesoporous
Ta O shell, the average pore size and BET surꢀace area
ꢀ
2
5
Fig. 6 Nitrogen adsorption–desorption isotherms (a) and pore size distribution curves (b) oꢀ Fe O @C, Fe O @C@mTa O and Fe O @C@
3
4
3
4
2
5
3
4
mTa O -NH -PW (14.47%)
2
5
2
12
1
3

Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres:…
−
1
titration with n-butylamine in acetonitrile (0.05 mol L ).
Based on the total acidity displayed in Table S1, it can be
seen that the acid-site density oꢀ diferent amount oꢀ HPAs
immobilized catalysts increases gradually with increasing
the contents oꢀ HPW owing to the ꢀact that HPW has strong
Brönsted acid capacity.
The diferent types oꢀ acid sites oꢀ Fe O @C@mTa O
3
4
2
5
and various Fe O @C@mTa O -NH -PW samples were
3
4
2
5
2
12
determined by Pyridine-FTIR spectra (Fig. 8). The peaks at
−
1
−1
1
450 cm and 1611 cm correspond to the Lewis acid (L)
sites, which is attributed to pyridine coordinatively bonded
5
+
−1
to the unsaturated surꢀace Ta . The peak at 1540 cm cor-
relates with pyridinium ions ꢀormed by protonation oꢀ the
Brönsted acid (B) sites. The characteristic peak oꢀ the co-
−1
existence oꢀ B and L sites can be observed at 1490 cm [40].
−
1
The peak at 1540 cm was not observed on Fe O @C@
3
4
Fig. 7 Magnetization curves oꢀ the samples measured at 300 K:
Fe O , Fe O @C, Fe O @C@mTa O , Fe O @C@
mTa O , but observed on Fe O @C@mTa O -NH -PW .
2
5
3
4
2
5
2
12
a
b
2
c
d
3
4
3
12
4
3
4
2
5
3
4
Thereꢀore, the immobilization oꢀ HPW on Fe O @C@
mTa O -NH -PW (14.47%)
3
4
2
5
mTa O introduced strong B sites at the catalyst surꢀace,
2
5
these B sites are attributed to the protons oꢀ HPW. The
−
1
oꢀ Fe O @C@mTa O increase rapidly to 3.9 nm and
peak at 1490 cm conꢁrms that Fe O @C@mTa O -NH -
3 4 2 5 2
3
4
2
5
2
−1
9
2.6 m g , respectively. The average pore size and BET
PW possesses both B and L sites. On the basis oꢀ the
12
surꢀace area oꢀ Fe O @C@mTa O -NH -PW (14.47%)
peak intensities, it can be inꢀerred that both oꢀ B sites and
the co-existence oꢀ B and L sites increase in the ꢀollowing
order: Fe O @C@mTa O -NH -PW (8.51%)<Fe O @C@
3
4
2
5
2
12
2
−1
decrease to 3.2 nm and 78.5 m g , respectively. The most
likely reason is that HPW was dispersed and deposited on
the surꢀace oꢀ Fe O @C@mTa O , decreasing the pore
3
4
2
5
2
12
3
4
mTa O -NH -PW (14.47%)<Fe O @C@mTa O -NH -
3
4
2
5
2
5
2
12
3
4
2
5
2
diameter and thus diminishing the surꢀace area.
PW (21.20%)<Fe O @C@mTa O -NH -PW (23.41%),
12 3 4 2 5 2 12
Figure 7 shows the magnetization curves oꢀ the as-
prepared Fe O , Fe O @C, Fe O @C@mTa O and
which indicates that the acid intensity oꢀ as-prepared cata-
lyst is associated with the amount oꢀ H PW O immobi-
3
4
3
4
3
4
2
5
3
12 40
Fe O @C@mTa O -NH -PW (14.47%) microspheres.
lized. Thus, it can be inꢀerred that changing the amount oꢀ
3
4
2
5
2
12
The magnetization saturation values oꢀ Fe O , Fe O @C,
H PW O immobilized can control the Brönsted/Lewis
3
4
3
4
3
12 40
Fe O @C@mTa O and Fe O @C@mTa O -NH -PW
acid site ratio ꢀor Fe O @C@mTa O -NH -PW .
3
4
2
5
3
4
2
5
2
12
3
4
2
5
2
12
−
1
(
14.47%) are 80.8, 49.3, 40.8 and 29.7 emu g , respec-
tively. The saturation magnetization values oꢀ Fe O @C,
3
4
Fe O @C@mTa O and Fe O @C@mTa O -NH-
3
4
2
5
3
4
2
5
H PW O (14.47%) signiꢁcantly decrease than that oꢀ
3
12 40
Fe O core due to the existence oꢀ non-magnetite C, Ta O
3
4
2
5
shell and HPW. However, compared to some reported mag-
−
1
netic nanoparticle catalysts, the value oꢀ 29.7 emu g is still
higher [30, 32]. From the partial enlargement oꢀ the mag-
netization curves oꢀ the as-prepared samples, it can be ꢀound
that there are hysteresis loops in the curves (Fig. S4), but
the coercive ꢀorce oꢀ as-prepared samples are much lower
than that oꢀ the magnetic Fe O bulk material (115–150 O ).
3
4
e
These results indicate that the as-prepared samples have
good magnetic properties. The Fe O @C@mTa O -NH -
3
4
2
5
2
PW (14.47%) microspheres can be easily enriched with
1
2
the aid oꢀ a permanent magnet (Fig. 7, inset).
For the acid-catalyzed reaction, the acid types, concen-
tration and strength oꢀ acid sites are the key ꢀactors which
can determine the catalytic perꢀormance oꢀ the catalyst.
Fig. 8 Py-IR spectra oꢀ
a Fe O @C@mTa O , b Fe O @C@
3 4 2 5 3 4
mTa O -NH -PW
(8.51%),
c
Fe O @C@mTa O -NH -PW
2
5
2
12
3
4
2
5
2
12
The total acidity oꢀ Fe O @C@mTa O and Fe O @C@
3
4
2
5
3
4
(14.47%), d Fe O @C@mTa O -NH -PW (21.20%), e Fe O @C@
3
4
2
5
2
12
3
4
mTa O -NH -PW was measured through potentiometric
mTa O -NH -PW (23.41%)
2 5 2 12
2
5
2
12
1
3

C. Feng et al.
2
.2 Catalysis of Fe O @C@mTa O ‑NH ‑PW
the acetalization, magnetic mesoporous tantalum pentox-
ide microspheres with diferent amounts oꢀ HPW immo-
bilizing, Fe O @C@mTa O -NH -PW (w%) (w = 8.51,
3
4
2
5
2
12
The catalytic activity was evaluated by the acetalization oꢀ
benzaldehyde (BzH) with ethylene glycol (EG). A series
oꢀ controlled experiments were carried out to investigate
the catalytic active components oꢀ catalysts, as shown in
Table 1. It is clear that only 42–51% oꢀ benzaldehyde is
converted to benzaldehyde ethylene glycol acetal (BEGA)
without catalyst or using Fe O as catalyst (entries 1 and
3
4
2
5
2
12
14.47, 21.20 and 23.41), were prepared. Among them, the
Fe O @C@mTa O -NH -PW (14.47%) catalyst exhib-
3
4
2
5
2
12
ited the highest activity and the best catalytic perꢀormance
with 94.5% conversion and 99% selectivity. However,
the total acidity and the B/L site ratio oꢀ the Fe O @C@
3
4
mTa O -NH -PW catalysts ꢀollowed the asending order:
3
4
2
5
2
12
2
). Fe O @C@mTa O , which presents predominantly
Fe O @C@mTa O -NH -PW (8.51%) < Fe O @C@
3
4
2
5
3 4 2 5 2 12 3 4
Lewis acidity, aforded 68% conversion oꢀ benzaldehyde
mTa O -NH -PW (14.47%)<Fe O @C@mTa O -NH -
2 5 2 12 3 4 2 5 2
(
entry 3). These indicate that Lewis acidity ꢀacilitates the
PW (21.20%)<Fe O @C@mTa O -NH -PW (23.41%)
12 3 4 2 5 2 12
acetalization oꢀ BzH with EG. When H PW O was taken
(Table S1, Fig. 8). Thus, an appropriate ratio oꢀ Brönsted
and Lewis acidity is preꢀerred ꢀor acetalization. This is well
in agreement with the reported literature [25]. In other word,
the acetalization oꢀ BzH with EG invokes a Brönsted-Lewis
acid synergy efect.
3
12 40
as homogeneous catalyst, which possesses Brönsted acid-
ity, we can obtain 81.5% conversion and 99% selectivity
(
entry 4).This suggests that Brönsted acidity is essential
ꢀ
or acetalization oꢀ BzH with EG. The catalytic activity oꢀ
physical mixture oꢀ H PW O and Fe O @C@mTa O is
A reasonable reaction pathway ꢀor acetalization oꢀ alde-
hydes with diols over the Fe O @C@mTa O -NH -PW
3
12 40
3
4
2
5
close to that oꢀ H PW O (entry 9). Notably, Fe O @C@
3
12 40
3
4
3
4
2
5
2
12
mTa O -NH -PW samples show superior catalytic perꢀor-
composite catalyst is presented in Scheme 2. The catalytic
2
5
2
12
mances, exhibiting 85.3–94.5% conversion in comparison
process involving co-participation oꢀ Brönsted and Lewis
5+
with that oꢀ H PW O catalyst (entries 5–8). This can be
acid is summerized as ꢀollows: the Lewis metal center (Ta )
is prone to combine with the oxygen atom (which possessed
strong electronegativity) in the carbonyl groups to induce
interactions between the carbon atom in the carbonyl group
and the oxygen in the alcohol hydroxyl group. The coupling
3
12 40
explained ꢀrom the ꢀollowing two aspects: one is that Brön-
sted and Lewis acid sites are co-existing in Fe O @C@
3
4
mTa O -NH -PW , the Brönsted–Lewis acid synergy is a
2
5
2
12
key ꢀactor in improving catalytic eꢂciency; another is that
the composite catalysts have a larger surꢀace area, enough
pore distribution, which allows the substrate molecules to
contact with more active sites and promotes the reaction.
In general, Fe O @C@mTa O -NH -PW catalysts show
+
oꢀ Brönsted acidic H with the oxygen atom in the carbonyl
group tends to reduce the negative charge, which ꢀacilitates
5
+
the release oꢀ Ta . The intermediate ꢀormed by coupling
between carbocation and oxygen in the alcohol hydroxyl
3
4
2
5
2
12
+
better catalytic activity than HPW.
group release proton (H ) to obtain the ꢁnal product.
It is noteworthy that the B/L site ratio oꢀ the catalyst
also play an important role on catalytic activity [41]. The
incorporation oꢀ Brönsted and Lewis acidity is ꢀavorable ꢀor
the acetalization oꢀ BzH with EG, and the conversion oꢀ
benzaldehyde was ꢀound to reach a maximum at a moder-
ate B/L site ratio [25]. In order to clearly demonstrate the
relationship between B/L site ratio and catalytic activity in
To veriꢀy the scope and limitation oꢀ the catalyst, tak-
ing Fe O @C@mTa O -NH -PW (14.47%) as cata-
3
4
2
5
2
12
lyst, the acetalization oꢀ diferent aldehydes with diols
were investigated. As displayed in Table 2, Fe O @C@
3
4
mTa O -NH -PW (14.47%) catalyst exhibited a high
2
5
2
12
catalytic activity in most acetalization reactions con-
ducted, indicating general applicability oꢀ the catalyst.
Table 1 Acetalization oꢀ
Entry
Catalyst
Conversion (%)
Selectivity (%)
benzaldehyde with ethylene
glycol catalyzed by diferent
samples
1
2
3
4
5
6
7
8
9
–
42.6
51.4
68.4
81.5
85.3
94.5
92.6
86.7
83.7
98
98
98
99
99
99
99
98
98
Fe O
3
4
Fe O @C@mTa O
5
3
4
2
[
a]
H PW O
3
12 40
Fe O @C@mTa O -NH -PW (8.51%)
3
4
2
5
2
12
Fe O @C@mTa O -NH -PW (14.47%)
3
4
2
5
2
12
Fe O @C@mTa O -NH -PW (21.20%)
3
4
2
5
2
12
Fe O @C@mTa O -NH -PW (23.41%)
3
4
2
5
2
12
[
b]
Fe O @C@mTa O -NH +H PW O
12 40
3
4
2
5
2
3
Reaction condition: benzaldehyde 0.1 mol, glycol 0.15 mol, cyclohexane 12.0 mL, catalyst 0.42 g, 80 °C
or 2 h. [a] the mass oꢀ catalyst 0.06 g, [b] 0.36 g Fe O @C@mTa O mixed with 0.06 g H PW O
ꢀ
3
4
2
5
3
12 40
1
3

Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres:…
Scheme 2 Reaction mechanism
ꢀ
or acetalization oꢀ aldehydes
with diols over the Fe O @C@
3
4
mTa O -NH -PW catalyst
2
5
2
12
(
14.47%)
Table 2 Acetalization oꢀ
Entry
Aldehyde
Diol
Conversion (%)
Selectivity (%)
various aldehydes with diols,
catalyzed by Fe O @C@
3
4
1
2
3
4
5
6
7
8
Benzaldehyde
Ethylene glycol
Propylene glycol
Butylene glycol
Pentylene glycol
Ethylene glycol
Propylene glycol
Butylene glycol
Pentylene glycol
94.5
91.2
89.5
82.6
94.2
90.7
84.6
82.5
99
98
98
98
99
99
98
98
mTa O -NH -PW (14.47%)
2
5
2
12
p-Tolualdehyde
Reaction condition: aldehydes 0.1 mol, glycol 0.15 mol, cyclohexane 12.0 mL, catalyst 4 wt%, 80 °C ꢀor
2
h
Supposing the acetalization occurs mainly in the pores
oꢀ the Fe O @C@mTa O -NH -PW (14.47%) catalyst,
acetalization reaction, probably due to the occurrence oꢀ
the reaction in pores.
3
4
2
5
2
12
the conversion oꢀ the aldehyde decreases along with the
length oꢀ carbon chain in the diols increases, probably
because oꢀ steric hindrance ꢀor the reaction. Thus, diols
with short carbon chains can easily enter into the pores oꢀ
Fe O @C@mTa O -NH -PW (14.47%) and catalytically
We also compared catalytic perꢀormance oꢀ Fe O @C@
3
4
mTa O -NH -PW (14.47%) with some reported heteropo-
2
5
2
12
lyacid composite catalysts in the acetalization oꢀ benzalde-
hyde with ethylene glycol. As shown in Table 3, the cata-
lytic activity oꢀ Fe O @C@mTa O -NH -PW (14.47%) is
3
4
2
5
2
12
3
4
2
5
2
12
react with aldehyde, and the products can also easily diꢀ-
superior to Cs-VImPS-PW and H [SiW O ]/SiO (entries
4 12 40 2
ꢀ
use through the pore canal oꢀ Fe O @C@mTa O -NH -
4 and 5). Although the catalytic activity oꢀ Ni[MIMPSH]
3
4
2
5
2
PW (14.47%) to leave out the catalyst. As the size oꢀ
PW O or HPW@MIL-100(Fe) is a little higher than
1
2
12 40
either reactant or product increases, the steric hindrance
would be essential ꢀor the acetalization, leading to the
lower yields ꢀrom Entries 1 to 4 and ꢀrom Entries 5 to 8
in Table 2 [7]. In contrast, the non-porous Fe O @C@
that oꢀ Fe O @C@mTa O -NH -PW (14.47%) (entries
3
4
2
5
2
12
2 and 3), the reaction conditions oꢀ Ni[MIMPSH]PW O
1
2
40
is environmental-unꢀriendly as the reaction temperature is
much higher, and their separation eꢂciency is lower than
Fe O @C@mTa O -NH -PW (14.47%), which can be
3
4
Ta O -NH -PW (14.82%) catalyst exhibited a low cata-
2
5
2
12
3
4
2
5
2
12
−
1
lytic activity in the acetalization reactions oꢀ benzalde-
hyde with diferent diols (Table S2). Moreover, the conver-
sion oꢀ the benzaldehyde to diferent diols is very close.
These results conꢁrm that the mesoporous structure oꢀ
Fe O @C@mTa O -NH -PW (14.47%) can promote the
quickly magnetically separated. Moreover, the TOF (min )
oꢀ Fe O @C@mTa O -NH -PW (14.47%) catalyst is 12.1,
3
4
2
5
2
12
indicating that the as-prepared catalyst has high catalytic
activity. Thereꢀore, in terms oꢀ perꢀormance and reaction
conditions, although the catalytic activity oꢀ Fe O @C@
3
4
2
5
2
12
3
4
1
3

C. Feng et al.
Table 3 Comparison oꢀ diferent heteropolyacid catalysts in acetalization oꢀ benzaldehyde with ethylene glycol
Entry
Catalyst
Benzaldehyde
mmol)
ethylene
-glycol
Cyclohexane
(mL)
Temp.
(°C)
Time (h)
Conversion (%)
94.5
Reꢀ.
(
(
mmol)
1
Fe O @C@
100
150
12
80
2
This work
3
4
mTa O -NH -PW
12
2
5
2
(
14.47 wt%)
Ni[MIMPSH]PW O
40
2
3
4
5
100
70
1
150
126
1.5
15
12
10
0
110
80
60
2
2
6
2
95.8
98.1
73
[25]
[42]
[10]
[43]
1
2
HPW@MIL-100(Fe)
H [SiW O ]/SiO
4
12 40
2
Cs-VImPS-PW
10
5
80
94.1
mTa O -NH -PW (14.47%) is not the highest, Fe O @C@
ethylene glycol, acetal was detected as the ꢁnal product
(Fig. 9). In the acetalization, water is inevitably ꢀormed.
Cyclohexane was used as the water-carrying agent to
improve the acetal yield by seperating water ꢀrom the reac-
tion mixture at 80 °C. As shown in Fig. 9, the amount oꢀ
catalyst and cyclohexane, EG to BzH molar ratio, and reac-
tion time can signiꢁcantly change the yield oꢀ acetal. The
highest yield (94.5%) was obtained under the ꢀollowing
2
5
2
12
3
4
mTa O -NH -PW (14.47%) could still serve as a more
2
5
2
12
promising catalyst candidate ꢀor the acetal ꢀormation.
2.3 Eꢀects of Reaction Parameters
The synthesized Fe O @C@mTa O -NH -PW (14.47%)
3
4
2
5
2
12
was applied to acetalization reaction oꢀ benzaldehyde and
Fig. 9 The efect oꢀ diferent a molar ratio oꢀ EG/BzH, b reaction
time, c catalyst amount, and d cyclohexane amount on acetal yield.
While changing each experimental variable, the other parameters
were kept constant: molar ratio oꢀ EG/BzH=1.5:1, catalyst 4 wt%,
cyclohexane 12.0 mL, 80 °C ꢀor 2 h
1
3

Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres:…
reaction conditions: a EG/BzH molar ratio oꢀ 1.5:1, the
reaction time oꢀ 2 h, 4 wt% catalyst loading, and 12 mL
cyclohexane. Thereaꢀter, with ꢀurther increasing the EG/
BzH ratio and the reaction time, the increase in yield is neg-
ligible. Further increase in the cyclohexane amount results
in signiꢁcant decrease oꢀ benzaldehyde conversion, which
is ascribed to the dilution efect by excessive cyclohexane.
Further increasing the catalyst amount beyond 4 wt% can-
not promote the reaction and lead to a slight decrease oꢀ
the conversion, the most likely reason is the mass transꢀer
hindrance caused by the low dispersity oꢀ the excess catalyst
in the reactant mixture.
The efect oꢀ reaction temperature on the conversion and
selectivity oꢀ acetalization oꢀ benzaldehyde with ethylene
glycol was also evaluated. As the reaction temperature
increases ꢀrom 70 to 110 °C, the conversion oꢀ benzaldehyde
Fig. 10 Reusability oꢀ Fe O @C@mTa O -NH -PW (14.47%) in
3
4
2
5
2
12
increased ꢀrom 75.6 to 98.9%. High temperature ꢀacilitates
to the ꢀurther acetalization oꢀ benzaldehyde. Compared to
the acetalization oꢀ benzaldehyde with ethylene glycol. Reaction con-
dition: benzaldehyde 0.1 mol, ethylene glycol 0.15 mol, cyclohexane
1
2.0 mL, catalyst 0.42 g, 80 °C ꢀor 2 h
7
0 °C, the conversion oꢀ benzaldehyde at 80 °C increased
signiꢁcantly. Further increasing the reaction temperature
ꢀ
rom 80 to 110 °C, the increase in conversion is slight.
Thereꢀore, considering the perꢀormance and cost, the best
reaction temperature is 80 °C (Fig. S5).
carried out. Aꢀter three cycles, the conversion oꢀ benzalde-
hyde decreased ꢀrom 94.3 to 81.2% due to leaching oꢀ HPW
active species. FT-IR spectra showed that the absorption oꢀ
HPW declined aꢀter three cycles (Fig. S7). These results
clearly show that the Fe O @C@mTa O -NH -PW cata-
The stability and reusability oꢀ Fe O @C@mTa O -NH -
3
4
2
5
2
PW was evaluated through ꢁve catalytic recycles. The
1
2
Fe O @C@mTa O -NH -PW catalyst could be easily
3
4
2
5
2
12
3
4
2
5
2
12
recycled by a permanent magnet aꢀter the reaction. Then,
the recovered catalysts were washed with acetone ꢀor three
times and dried under vacuum and reused ꢀor the next run. It
was ꢀound that the catalytic eꢂciency ꢀor the acetalization oꢀ
benzaldehyde with ethylene glycol oꢀ the recovered catalysts
is almost not changed (Fig. 10). Aꢀter ꢁve cycles, the recov-
ered catalyst was characterized by FT-IR spectra (Fig. S6).
The FT-IR spectrum oꢀ the ꢀresh catalyst is quite similar to
that oꢀ the reused catalyst, indicating that Keggin polyanions
are still retained in the recovered catalyst. To conꢁrm that
the catalytic action oꢀ Fe O @C@mTa O -NH -PW is
lyst is stable and the electrostatic interaction oꢀ protonated
amino groups with Keggin anions plays an important role in
constructing the stable catalyst.
3 Conclusions
In this study, we have successꢀully synthesized phospho-
tungstic acid supported on amino-ꢀunctionalized magnetic
core–shell mesoporous tantalum pentoxide microspheres
with a ꢀacile procedure. The magnetically recoverable cata-
lyst possesses large BET surꢀace area, high stability, strong
Brönsted and Lewis acidity, and present high eꢀꢀicient
catalytic activity ꢀor acetalization oꢀ diferent aldehydes
and diols. The results show that the Brönsted/Lewis acid
site ratio oꢀ the catalyst play an important role on catalytic
activity, and Fe O @C@mTa O -NH -PW (14.47%) with
3
4
2
5
2
12
truly heterogeneous, ICP was used ꢀor determining the con-
tent oꢀ W in the catalytic mixture aꢀter removing magneti-
cally the heterogeneous catalyst. The ICP results revealed
that there is almost no W in the mixture, which means that
H PW O is not leached during the reaction. Meanwhile,
3
12 40
changes oꢀ acidity oꢀ the recovered catalyst are negligible
(
entry 6 in Table S1).
3
4
2
5
2
12
For comparison, Fe O @C@mTa O immobilized with
a moderate Brönsted/Lewis acid site ratio exhibits the high-
est catalytic activity with 94.5% conversion oꢀ benzaldehyde
and 99% selectivity to benzaldehyde glycol acetal. Addi-
tionally, the as-prepared catalyst could be easily recycled
by magnetic separation and reused at least 5 cycles with-
out signiꢁcant loss oꢀ its catalytic activity. To our knowl-
edge, this is the ꢁrst report that H PW O anchored to
3
4
2
5
H PW O were synthesized, denoted as Fe O @C@
3
12 40
3
4
mTa O -PW (15.12%), the preparation process was
2
5
12
shown in Support Inꢀormation. Under the same reaction
conditions, the Fe O @C@mTa O -PW (15.12%) cata-
3
4
2
5
12
lyst exhibited 94.3% conversion ꢀor the acetalization oꢀ
benzaldehyde with ethylene glycol, which is close to that oꢀ
Fe O @C@mTa O -NH -PW (14.47%). The reusability
3
12 40
amino-ꢀunctionalized Fe O @C@mTa O microspheres and
3
4
2
5
2
12
3
4
2
5
oꢀ the Fe O @C@mTa O -PW (15.12%) catalyst was also
used ꢀor acetalization oꢀ aldehydes with diols. As-prepared
3
4
2
5
12
1
3

C. Feng et al.
multiꢀunctionalized hybrid catalysts are considered as prom-
ising candidates ꢀor other acid catalyzed reactions.
4.2.3 Synthesis oꢀ Fe O @C@mTa O Magnetic
3 4 2 5
Microspheres
The Fe O @C@mTa O magnetic microspheres were syn-
3
4
2
5
thesized through a sol–gel process with some modiꢁcation
4
Experimental
[
46, 47]. First, Fe O @C microspheres (0.15 g) were dis-
3 4
persed in the mixture (40 mL) containing oꢀ CTAB (0.64 g),
deioned water and ethanol (v/v=1/3) by ultrasonication ꢀor
4.1 Materials and Chemicals
1
5 min. Then, anhydrous ethanol (20 mL) containing oꢀ
H PW O was purchased ꢀrom Aladdin Chemistry Co.,
3
12 40
tantalum ethoxide (120 μL) was added into the mixed solu-
tion over a period oꢀ approximately 20 min under magnetic
stirring. Aꢀter stirred ꢀor 6 h, the products were magneti-
cally separated and washed with ethanol and water. Next,
the microspheres were redispersed in 80 mL oꢀ acetone and
reꢃuxed at 80 °C ꢀor 48 h to extract the template CTAB.
Aꢀter removal oꢀ CTAB templates, the mesoporous tantalum
pentoxide magnetic microspheres were obtained (denoted
as Fe O @C@mTa O ). Finally, the catalyst was collected,
Ltd.;Ferric chloride (FeCl ·6H O), sodium acetate anhy-
3
2
drous (NaAc), polyethylene glycol, glucose, cetyltrimeth-
ylammonium bromide (CTAB) were supplied by Tianjin
Kemiou Chemical Reagent Co., Ltd.; Tantalum ethoxide
(
99%) was purchased ꢀrom Sigma Aldrich; (3-aminopropyl)
triethoxysilane (APTES, 98%), cyclohexane, benzaldehyde
99%), 3-methylbenzaldehyde (99%), cyclohexanone(99%),
(
ethylene glycol (99%), 1,3-propanediol (99%), 1,4-butan-
ediol (99%), 1,5-pentadiol (98%) were obtained ꢀrom
Adamas. All chemicals were directly used without any ꢀur-
ther puriꢁcation.
3
4
2
5
washed with ethanol and dispersed in ethanol (40 mL).
For comparison, non-porous tantalum pentoxide coated
Fe O @C was also prepared according to the method similar
3
4
to Fe O @C@mTa O , except that no template was added,
3
4
2
5
the sample was denoted as Fe O @C@Ta O .
3
4
2
5
4.2 Catalyst Preparation
4
.2.4 Synthesis oꢀ Fe O @C@mTa O ‑NH Magnetic
3 4 2 5 2
4.2.1 Preparation oꢀ Fe3O4
Microspheres
Fe O magnetic microspheres (ca.300 nm) were synthesized
3
4
Amino-ꢀunctionalized Fe O @C@mTa O microspheres,
3
4
2
5
through the solvothermal reaction [44]. FeCl ·6H O (2.0 g),
3
2
designated as Fe O @C@mTa O -NH , were synthesized
3
4
2
5
2
polyethylene glycol (1.5 g) were added into ethylene glycol
as ꢀollows. First, 350 μL oꢀ APTES and 50 μL oꢀ NH ·H O
3
2
(
60 mL) under vigorous magnetic stirring. Aꢀter completely
were added to Fe O @C@mTa O ethanol solution. Then,
3
4
2
5
dissolved, NaAc (5.4 g) was added to the above solution.
Subsequently, the mixture was stirred ꢀor 1 h and sealed
in a teꢃonlined stainless-steel autoclave (100 mL capacity).
Through hydrothermal treatment at 200 °Cꢀor 8 h, the black
products were obtained. Finally, the precipitate was washed
with ethanol and deionized water and dried under vacuum
at 50 °C ꢀor 8 h.
the mixture was reꢃuxed at 80 °C ꢀor 12 h. The obtained
Fe O @C@mTa O -NH magnetic microspheres were
3
4
2
5
2
washed with ethanol and air-dried.
4
.2.5 Synthesis oꢀ Fe O @C@mTa O ‑NH ‑PW Magnetic
3 4 2 5 2 12
Microspheres
For the immobilization oꢀ H PW O on Fe O @C@
3
12 40
3
4
4
.2.2 Synthesis oꢀ Fe O @C Magnetic Microspheres
mTa O -NH , 60 mL oꢀ methanol solution containing a
3
4
2 5 2
certain amount oꢀ H PW O (0.2–1.0 g) was added to a
3
12 40
The Fe O @C magnetic microspheres were prepared
ꢀreshly activated Fe O @C@mTa O -NH (0.3 g) and the
3
4
3 4 2 5 2
through hydrothermal reaction oꢀ glucose on Fe O magnetic
mixture was reꢃuxed ꢀor 6 h. Subsequently, the products
were gathered aꢀter being washed with deionized water and
dried at 50 °C overnight. The actual amount oꢀ HPW in the
catalyst was determined by ICP/AES. The obtained cata-
lysts were denoted as Fe O @C@mTa O -NH -PW (w%)
3
4
microspheres [45]. Typically, Fe O (0.1 g) microspheres
3
4
were ultrasonicated ꢀor 15 min in 0.1 M HNO aqueous solu-
3
tion (50 mL), then washed with deionized water. Thereaꢀter,
the treated Fe O microspheres were redispersed in 0.5 M
3
4
3
4
2
5
2
12
aqueous glucose solution (80 mL) by ultrasonication ꢀor
(w=8.51, 14.47, 21.20 and 23.41), respectively.
1
5 min. Finally, the mixture was transꢀerred to autoclaves
and heated at 180 °C ꢀor 4 h. The black products were col-
lected and washed with deionized water and ethanol. Aꢀter
drying under vacuum at 80 °C ꢀor 4 h, the ꢁnal sample was
obtained (denoted as Fe O @C).
4.3 Catalyst Characterization
FT-IR spectra were recorded on a TENSOR 27 spectrom-
eter. The XRD data were recorded by a Bruker D8 Advance
3
4
1
3

Phosphotungstic Acid Supported on Magnetic Mesoporous Tantalum Pentoxide Microspheres:…
difractometer with Cu Kα radiation, operating at 40 kV and
0 mA. Morphologies were collected by scanning electron
Acknowledgements This work was supported by the National Natural
Science Foundation oꢀ China (21673176).
4
microscopy (SEM) images using ZEISS SIGMA 500 scan-
ning microscope with an area energy dispersive X-ray spec-
trometer (EDS). Transmission electron microscope (TEM)
images were collected on a GZF2.0 transmission electron
References
1
2
3
4
.
.
.
.
Wuts PGM, Greene TW (2007) Greene’s protective groups in
organic synthesis. Wiley, New York
microscope at an accelerating voltage oꢀ 200 kV. N absorp-
2
tion–desorption experiment was carried out at 77 K on a
Micromeritics TR2 Star3020 analyzer. The surꢀace areas
were calculated using the BET equation, and the pore size
distributions were obtained using the BJH adsorption branch
oꢀ the isotherms. Magnetic properties were determined on
a Quantum Design MPMS SQUID magnetometer at room
temperature. The content oꢀ W was determined by ICP/
AES on a VarioEL III elemental analyzer (German Ellman
Elemental Analysis System Company).
Bauer K, Garbe D, Surburg H (2001) Common Fragrances and
ꢃavor materials: preparation and uses. Wiley, New York
Reddy PS, Sudarsanam P, Mallesham B, Raju G, Reddy BM
(
2011) J Ind Eng Chem 17:377–381
Climent MJ, Corma A, Velty A, Susarte M (2000) J Catal
96:345–351
5. Tan MX, Gu LQ, Li NN, Ying JY, Zhang YG (2013) Green Chem
1
1
5:1127–1132
6
7
8
9
.
.
.
.
Miao ZC, Xu LL, Song HL, Zhao HH, Chou LJ (2013) Catal Sci
Technol 3:1942–1954
Zhang FM, Shi J, Jin Y, Fu YH, Zhong YJ, Zhu WD (2015) Chem
Eng J 259:183–190
Dhakshinamoorthy A, Alvaro M, Garcia H (2010) Adv Synth
Catal 352:3022–3030
4.4 Determination of the Acidity Content
Sinhamahapatra A, Sutradhar N, Ghosh M, Bajaj HC, Panda AB
(
2011) Appl Catal A 402:87–93
The total acid content oꢀ as-synthesized catalysts was meas-
ured by potentiometric titration. In brieꢀ, the catalyst powder
1
1
0. Zhao S, Jia YQ, Song YF (2014) Catal Sci Technol 4:2618–2625
1. da Silva KA, Kozhevnikov IV, Gusevskaya EV (2003) J Mol Catal
A 192:129–134
(
50 mg) was added in acetonitrile (50 mL). Aꢀter stirring ꢀor
2
h, the obtained suspension was titrated with a solution oꢀ
12. de Meireles ALP, Rocha KAD, Kozhevnikov IV, Gusevskaya EV
(2011) Appl Catal A 409:82–86
−
1
n-butylamine in acetonitrile (0.05 mol L ) [31].
1
1
3. Wilke TJ, Barteau MA (2019) J Catal 371:357–367
4. An XQ, Tang QW, Lan HC, Liu HJ, Qu JH (2019) Appl Catal B
2
44:407–413
4.5 Characterization of Brönsted and Lewis Acidity
1
5. Seraꢁm H, Fonseca IM, Ramos AM, Vital J, Castanheiro JE
2011) Chem Eng J 178:291–296
(
Brönsted and Lewis acidity oꢀ the prepared materials was
determined by pyridine-FT-IR spectra. The as-prepared
materials were exposed to pyridine vapour at 60 °C ꢀor 2 h
in vacuum aꢀter drying at 100 °C ꢀor 2 h, the physisorbed
16. Kim YS, Wang F, Hickner M, Zawodzinski TA, McGrath JE
(
2003) J Membr Sci 212:263–282
1
1
7. Obali Z, Dogu T (2008) Chem Eng J 138:548–555
8. Su F, An S, Song DY, Zhang XH, Lu B, Guo YH (2014) J Mater
Chem A 2:14127–14138
pyridine was blown of with a N ꢃow. Finally, FT-IR spectra
19. Zhao PP, Zhang YY, Wang Y, Cui HY, Song F, Sun XY, Zhang
LP (2018) Green Chem 20:1551–1559
2
oꢀ adsorbed pyridine was recorded.
2
2
2
0. Zhang XY, Li Y, Xue LF, Wang ST, Wang XH, Jiang ZJ (2018)
ACS Sustain Chem Eng 6:165–176
1. Bromberg L, Hatton TA (2011) Acs Appl Mater Interꢀaces
4.6 Acetalization Reaction
3
:4756–4764
2. Manjunathan P, Marakatti VS, Chandra P, Kulal AB, Umbarkar
SB, Ravishankar R, Shanbhag GV (2018) Catal Today 309:61–76
Acetalization reactions were perꢀormed in a three-necked
ꢃask equipped with a dean-stark apparatus. In a typical
experiment, a mixture oꢀ 0.1 mol aldehyde, 0.15 mol eth-
ylene glycol/1,3-propanediol/1,4-butanediol/1,5-pentadiol,
23. Li H, Yang TT, Riisager A, Saravanamurugan S, Yang S (2017)
ChemCatChem 9:1097–1104
2
4. Yan W, Chen KK, Zhang XF, Kuang YY, Tang XJ, Han XX
(
2015) J Ind Eng Chem 29:185–193
1
2 mL cyclohexane and 4 wt% oꢀ catalyst Fe O @C@
25. Han XX, Ouyang K, Xiong CH, Tang XJ, Chen Q, Wang KW, Liu
LL, Hung CT, Liu SB (2017) Appl Catal A 543:115–124
3
4
mTa O -NH -PW (0.42 g) were stirred at 80 °C ꢀor 2 h.
2
5
2
12
2
2
2
6. Mallesham B, Sudarsanam P, Reddy BM (2014) Catal Sci Technol
The reaction mixture was sampled periodically and analyzed
with a HP SP-6890 gas chromatograph and identiꢁed via a
gas chromatography-mass spectroscopy (GC–MS).
4
:803–813
7. Ohyama J, Kanao R, Ohira Y, Satsuma A (2016) Green Chem
18:676–680
8. Xu LL, Wang YH, Yang X, Yu XD, Guo YH, Clark JH (2008)
Green Chem 10:746–755
4.7 Recycling of the Catalyst
29. Kucheryavy P, He JB, John VT, Maharjan P, Spinu L, Goloverda
GZ, Kolesnichenko VL (2013) Langmuir 29:710–716
3
0. Wang SG, Zhang ZH, Liu B, Li JL (2013) Catal Sci Technol
Aꢀter reaction, the catalyst Fe O @C@mTa O -NH -PW
12
3
4
2
5
2
3
:2104–2112
was recovered by magnetic separation, thoroughly washed
3
1. Zolꢀagharinia S, Kolvari E, Koukabi N (2017) Catal Lett
147:1551–1566
with acetone and dried at 60 °C.
1
3

C. Feng et al.
3
3
3
3
3
2. Amiri A, Saadati-Moshtaghin HR, Zonoz FM, Targhoo A (2017)
J Chromatogr A 1483:64–70
42. Zhang FM, Jin Y, Shi J, Zhong YJ, Zhu WD, El-Shall MS (2015)
Chem Eng J 269:236–244
3. Kong AG, Wang P, Zhang HG, Yang F, Huang SP, Shan YK
43. Zhang WH, Liu SS, Liu P, Xu J, Xue B, Wei XY, Li YX (2016)
Rsc Adv 6:41404–41409
(
2012) Appl Catal A 417:183–189
4. Tayebee R, Amini MM, Rostamian H, Aliakbari A (2014) Dalton
Trans 43:1550–1563
44. Deng H, Li XL, Peng Q, Wang X, Chen JP, Li YD (2005) Angew
Chem 117:2842–2845
5. Dong XB, Zhang X, Wu PF, Zhang YJ, Liu B, Hu HM, Xue GL
45. Li Y, Wu JS, Qi DW, Xu XQ, Deng CH, Yang PY, Zhang XM
(2008) Chem Commun 5:564–566
(
2016) ChemCatChem 8:3680–3687
6. Duan XX, Liu Y, Zhao Q, Wang XH, Li SW (2013) Rsc Adv
46. Qi DW, Lu J, Deng CH, Zhang XM (2009) J Chromatogr A
1216:5533–5539
3
:13748–13755
3
3
7. Raꢁee E, Eavani S (2011) Green Chem 13:2116–2122
8. Li L, Wu QY, Guo YH, Hu CW (2005) Microporous Mesoporous
Mater 87:1–9
47. Chen YY, Song GS, Dong ZL, Yi X, Chao Y, Liang C, Yang K,
Cheng L, Liu Z (2017) Small 13:1602869
3
4
4
9. Kumar D, Landry CC (2007) Microporous Mesoporous Mater
Publisher’s Note Springer Nature remains neutral with regard to
9
8:309–316
jurisdictional claims in published maps and institutional aꢂliations.
0. Pezzotta C, Fleury G, Soetens M, Van der Perre S, Denayer JFM,
Riant O, Gaigneaux EM (2018) J Catal 359:198–211
1. Han XX, Kuang YY, Ouyang K, Kan RJ, Tang XJ, Hung CT, Liu
LL, Wu PH, Liu SB (2017) J Taiwan Inst Chem E 70:23–31
1
3