A porous inorganic zirconyl pyrophosphate as an efficient catalyst for the catalytic transfer hydrogenation of ethyl levulinate to γ-valerolactone

Article abstract of DOI:10.1002/jccs.201800073

Catalytic transfer hydrogenation (CTH) of ethyl levulinate (EL) to γ-valerolactone (GVL) is an alluring reaction in the field of biomass catalytic conversion, but it normally depends on the consumption of H₂. In this study, we report a porous Zr-containing inorganic pyrophosphate catalyst (ZrOPP), which was used as a catalyst for CTH of EL to GVL in the presence of isopropanol and characterized using FT-IR, py-FTIR, TGA, XRD, BET, XPS, ICP-AES, SEM, TEM, NH₃-TPD, and CO₂-TPD. We achieved a high yield of 94% GVL at 433 K for 11 hr. Furthermore, the ZrOPP has the trait of easy separation and could be reused more than five times without distinct decrease in activity and selectivity. In addition, this catalyst could also be applied to other catalytic hydrogenation reactions, such as those of cyclohexanone, acetophenone, 2-heptanone etc. Its outstanding performance was mainly ascribed to the acid sites from the Zr element and basic sites from phosphate groups interspersing on the surface of the catalyst.

Full text of DOI:10.1002/jccs.201800073

Received: 26 February 2018
DOI: 10.1002/jccs.201800073
Revised: 13 May 2018
Accepted: 19 May 2018
A R T I C L E
A porous inorganic zirconyl pyrophosphate as an efficient
catalyst for the catalytic transfer hydrogenation of ethyl
levulinate to γ-valerolactone
1
1
1
1
2
1
Jianjia Wang | Ruiying Wang | Huimin Zi | Haijun Wang
| Yongmei Xia | Xiang Liu
1
Key Laboratory of Synthetic and Biological
Catalytic transfer hydrogenation (CTH) of ethyl levulinate (EL) to γ-valerolactone
(GVL) is an alluring reaction in the field of biomass catalytic conversion, but it
Colloids, Ministry of Education, School of
Chemical and Material Engineering, Jiangnan
University, Wuxi, Jiangsu, China
normally depends on the consumption of H . In this study, we report a porous Zr-
2
2
State Key Laboratory of Food Science &
containing inorganic pyrophosphate catalyst (ZrOPP), which was used as a catalyst
for CTH of EL to GVL in the presence of isopropanol and characterized using FT-
Technology, Wuxi, Jiangsu, China
Correspondence
IR, py-FTIR, TGA, XRD, BET, XPS, ICP-AES, SEM, TEM, NH -TPD, and CO -
Haijun Wang, Key Laboratory of Synthetic and
Biological Colloids, Ministry of Education, School
of Chemical and Material Engineering, Jiangnan
University, Wuxi 214122, Jiangsu, China.
Email: wanghj329@outlook.com
3
2
TPD. We achieved a high yield of 94% GVL at 433 K for 11 hr. Furthermore, the
ZrOPP has the trait of easy separation and could be reused more than five times
without distinct decrease in activity and selectivity. In addition, this catalyst could
also be applied to other catalytic hydrogenation reactions, such as those of cyclo-
hexanone, acetophenone, 2-heptanone etc. Its outstanding performance was mainly
ascribed to the acid sites from the Zr element and basic sites from phosphate groups
interspersing on the surface of the catalyst.
Funding information
the MOE & SAFEA for the 111 Project, Grant/
Award Number: B13025
KEYWORDS
catalytic transfer hydrogenation, ethyl levulinate, isopropanol, zirconyl
pyrophosphate, γ-valerolactone
1
| INTRODUCTION
improving biomass conversion in ways of accelerating the
initial formation rate of the product and slowing down
product degradation.
[
5]
In the last few years, the growing consumption of fossil
resources and alarming environmental pollution have
received much attention.
most abundant and sustainable resource on the earth, and it
has been discussed to dispose of the energy crisis and
environmental pollution. To date, various value-added
chemicals could be acquired from biomass, such as
Nowadays, the catalytic transformation of ethyl levuli-
nate (EL) into GVL is very interesting in the point of view
of chemistry. The reaction process is generally considered in
such a way: the ketone group on EL is hydrogenated to form
hydroxy levulinic ester, which was ring-closed by intramo-
lecular transesterification to produce GVL and the corre-
[
1–3]
In addition, biomass is the
[
6]
5
-hydroxymethylfurfural (HMF), levulinic acid (LA),
sponding alcohol. Many homogeneous and heterogeneous
catalysts have been used in this process. Yang et al. reported
a smart strategy to fabricate Ru nanoparticle-inserted porous
γ-valerolactone (GVL), 2,5-dimethyfuran, polyol, and so
on.
[
1,4]
Among these chemicals, GVL has been considered
to be a versatile chemical that could be used as a liquid
fuel, perfume, and food additive. Besides, GVL can be
employed not only as a precursor for the production of
gasoline and diesel fuels but also as a green solvent for
carbon, achieving the conversion of LA to GVL with H as
2
[
7]
a hydrogen source. Furthermore, Banerjee et al. put for-
ward a core–shell nanoreactor storing Pd (0) to produce
GVL from EL by decomposing formic acid to offer the
©
2018 The Chemical Society Located in Taipei & Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J Chin Chem Soc. 2018;1–9.
http://www.jccs.wiley-vch.de
1

2
WANG ET AL.
[
8]
hydrogen source. The special carrier structure is helpful
for the loading and dispersion of the metal nanoparticles,
promoting the transfer hydrogenation reaction of EL to
GVL. Although good yields of GVL were obtained, these
(
a)
(b)
606
8
11 582
1
625
hydrogenation methods suffered from high H pressure, the
3421
2
1
180
1
039
use of precious metals, a low catalyst stability, and the cor-
rodibility of formic acid, which limited their large-scale
application. Therefore, it is an ideal task to search for an
efficient way to produce GVL.
4000
3500
3000
2500
2000
1500
1000
500
20
40
60
80
-1
2-Thera (°)
Wavenumber(cm
)
[
9]
220
L
2
1
1
1
1
00
(d)
(
c)
80
60
40
20
L
B
As we all know, the catalytic transfer hydrogenation
L+B
(
CTH) reactions are a conspicuous alternative to achieve the
[10]
100
reduction of carbonyl compounds.
The most critical step
Adsorption
Desorption
8
0
B
in the conversion of EL to GVL is the CTH of EL. Many
catalysts have been prepared for catalytic hydrogenation of
aldehydes and ketones. Mcnerney et al. reported a well-
defined monomeric aluminum complex, which could be
applied as a general catalyst in the CTH reaction. Perfect cat-
alytic activity is attributed to its unique monomeric nature
and the poor coordination of neutral donor isopropanol to
60
4
2
0
0
0
.0
0.2
0.4
0.6
0.8
1.0
1700
1650
1600
1550
1500
1450
1400
Relative Pressure(p/p
0
)
Wavenumber(cm )
-1
(
e)
(f)
[11]
the aluminum center.
Gawande and coworkers applied
the Ag@Ni magnetic core-shell nanocatalyst first to transfer
[
12]
hydrogenation reactions of carbonyl compounds.
Wu
et al. introduced a chiral N,N-dioxide/Y(OTf ) complex to
bring about the reduction of glycosylates. It was found that
3
FIGURE 1 Characterization of the prepared ZrOPP. The FT-IR spectra of
ZrOPP (a), the powder XRD pattern of ZrOPP (b), the N adsorption–
desorption isotherm of ZrOPP (c), pyridine adsorption IR spectra of ZrOPP
2
the adjunction of both molecular sieves and Al(OtBu) was
3
[13]
effective for the reaction.
There are many catalysts used
(d), SEM and TEM images of ZrOPP (e,f )
in the CTH reaction, but for this kind of reaction there still
[
14]
exist limiting factors, such as a high reaction temperature,
a long reaction time,
[15]
complex operation of the catalyst
2 | RESULTS AND DISCUSSION
preparation process, and expensive catalyst materials. There-
fore, it is a vital topic for us to look for an efficient, simple,
cheap, and general catalyst for this kind of reaction.
2
.1 | Characterization of catalysts
The Fourier transform-infrared (FT-IR) technique was used to
characterize the ZrOPP catalyst (Figure 1a). In each spectrum,
There were many reports about connecting phosphate
[16–18]
with zirconium to form catalysts.
But pyrophosphate is
−1
there are a strong broad band at 3,421 cm and a sharp band
at 1,625 cm− corresponding to the surface-adsorbed water and
different from phosphate. Pyrophosphate possesses a sym-
metric structure, it could link with zirconium to become a
relatively regular structure. Zirconium pyrophosphate
1
[22]
−1
hydroxyl groups. A weak band at 1,180 cm is attributed
to the P-O asymmetrical stretching vibration from the
(ZrOPP) has a porous structure comprising of a phosphorus
[23,24]
RPO (OH) groups.
Besides, the symmetrical stretching
oxygen tetrahedron and zirconium atoms. Each phosphorus
forms a tetrahedron with four oxygen atoms and shares oxy-
gen atoms with zirconium to accumulate to form a layered
2
modes of the O P groups are observed because of the strong
3
−1 [25]
absorption at 1,039 cm ,
which was attributed to the
[
19]
stretching vibrations of P-O-Zr and could be taken as evidence
structure.
Compared with zirconyl pyrophosphate, other
[
26]
of the structural formation of zirconium phosphonate. Peaks
Zr-based catalysts have shortcomings in the process of
preparation. For example, the synthesis of Zr-Beta requires a
longer crystallization time and the preparation of Zr(salphen)-
at 811 and 606–562 cm− are assigned to the framework vibra-
1
[27,28]
tion of Zr-O-P and Zr-O bonds, respectively.
[20,21]
The powder XRD pattern of the synthesized catalyst is
shown in Figure 1b. There is no X-ray crystal structure in
the XRD pattern. Furthermore, the XRD pattern revealed
one broad diffraction peak, enucleating that the synthesized
catalyst was amorphous. The textural parameters of the pre-
MCM-41 involves a tedious preparation process.
How-
ever, these catalysts play a role in the conversion of LA and
its ester. In this study, we firstly introduce zirconium pyro-
phosphate (ZrOPP) to achieve a series of CTH reactions,
especially the conversion of EL to GVL. Its excellent perfor-
mance is attributed to the interaction between zirconium and
oxygen providing the Lewis acid sites and basic sites, respec-
tively. At the same time, ZrOPP is convenient to separate and
has high recyclability.
pared ZrOPP were measured using the N
desorption method after the sample was degassed at 120 C
adsorption–
2
ꢀ
for 6 hr. It can be seen that the N
therm of the catalyst exhibits a type IV mode obviously,
adsorption–desorption iso-
2

WANG ET AL.
3
showing pore condensation with palpable adsorption–
desorption hysteresis (Figure 1c). When p/p was less then
0
0
.4, N was mainly monolayer adsorbed on the surface of
2
ZrOPP. At this time, the adsorption amount of p/p and N2
0
was linear. When p/p reached 0.4, the curve entered the
0
jump stage, mainly caused by the capillary condensation of
the N molecules in the mesopores. At the same time, it
2
could be found that the N adsorption capacity increased
2
sharply. The mesoporous structure of ZrOPP was well illus-
trated. In addition, the size and distribution of the pore size
were determined by the abrupt position of the capillary con-
densation. The greater the change rate of the N adsorption
2
and the partial pressure, the more uniform the pore size of
the material was. The results illustrated that the material was
porous. In addition, the pore volume, the average pore diam-
eter, and the Brunauer–Emmett–Teller (BET) surface area of
3
2
the catalyst were 0.24 cm /g, 6 nm, and 159 m /g. To prove
the existence of both Lewis acid sites and Brønsted acid sites
in the ZrOPP, we also further tested the acid properties using
the FT-IR spectra of pyridine adsorption. As shown in
Figure 1d, the obvious bands at 1,446 and 1,608 cm− were
assigned to pyridine adsorbed on Lewis acid sites, which
confirmed the presence of Lewis acid species chiefly stem-
1
[
29]
ming from Zr-O-P and dissociative zirconium ions.
In
−
1
addition, the weak bands at 1,534 and 1,660 cm were
attributed to pyridine adsorbed on Brønsted acid sites, which
testified the existence of Brønsted acid species primarily
originating from –OH on the surface of ZrOPP. The peak at
1
1
,489 cm− belonged to pyridine adsorbed on both Brønsted
[30]
and Lewis acid sites.
These acid sites are favorable for
enhancing the catalytic activity and they play different roles
in the process of catalytic reaction. Scanning electron
microscopy (SEM) and transmission electron microscopy
(
TEM) demonstrated that the catalyst consisted of conglom-
FIGURE 2 Influence of the reaction time and reaction temperature (a).
Reaction conditions: EL 1 mmol; isopropanol 5 mL; ZrOPP 200 mg; effect
of ZrOPP amount (b). Reaction conditions: EL 1 mmol; isopropanol 5 mL;
erated particles to form a laminated structure and had no uni-
form shape (Figure 1e,f ).
ꢀ
reaction temperature 160 C; reaction time 11 hr
2.2 | Influence of different reaction parameters
on the reaction
and the results are shown in Figure 2b. It could be seen that
the yield and selectivity of GVL increased with the incremen-
tal amount of the catalyst. At the beginning, the amount of cat-
alyst was 50 mg, and the selectivity of GVL was 46%. The
main by-product was ethyl 4-isopropoxylpentanote from GC–
MS analysis. 97% conversion of EL and 94% yield of GVL
were achieved when the catalyst amount was 200 mg. After
that, more amounts of catalyst promoted side reactions, which
cut down the yield and selectivity of GVL slightly. In sum-
mary, 200 mg was an appropriate amount of catalyst.
During the reaction process of conversion of EL to GVL. The
influence of the reaction time and reaction temperature was
discussed to achieve optimal reaction conditions in the pres-
ence of 200 mg ZrOPP with isopropanol as a hydrogen source
(Figure 2a). In a relatively short reaction time, the yield of
GVL and conversion of EL were not high. When the reaction
time reached 11 hr, EL was nearly transformed completely
and the yield of GVL is up to 94%. After 11 hr, there was no
obvious augment in terms of the yield and selectivity of GVL.
Meanwhile, with the rise of temperature, the yield and selectiv-
ity of GVL were improved, respectively, and remains almost
ꢀ
ꢀ
invariant after 160 C. Therefore, 11 hr and 160 C were the
optimum time and temperature, respectively, for the reaction.
The impact of the amount of ZrOPP on conversion of EL
2
.3 | Reusability and heterogeneity of the catalyst
The reusability of ZrOPP was also examined in the experi-
ment. After each reaction, the catalyst was recovered by
ꢀ
to GVL was discussed at 160 C with a reaction time of 11 hr

4
WANG ET AL.
centrifugation and washed with ethyl ether for the next run.
As shown in Figure 3a, there was no obvious abatement in
the yield and selectivity of GVL after five cycles. Mean-
while, we also explored the leaching of the catalytic species
and heterogeneity of the catalyst. After 9 hr of reaction, the
catalyst was filtered and the reactor was put back into the oil
bath to keep on the reaction. The result is shown in
Figure 3b. It was obvious that the yield did not continue to
increase after the catalyst filtering, indicating that the active
component in the catalyst was not soluble in the reaction
system and the reaction proceeded with heterogeneous catal-
ysis. In addition, the thermogravimetric analysis (TGA)
showed that the thermogravimetric curve was very steep
ESI). This indicated that the part of loss was because of
water and ZrOPP had a good thermal stability. Therefore,
the catalyst could be stable at the reaction temperature. Fur-
thermore, the catalyst was characterized using FT-IR, XRD,
SEM, and the N adsorption–desorption isotherm (in the
2
ESI) after five cycles. There was no much change in the ele-
ment content and BET surface area (in the ESI). All of these
indicated that ZrOPP was rather stable.
2
.4 | Catalytic activity analysis
ZrOPP has a similar structure to ZrO of the skeleton (see
2
the ESI). However, ZrOPP showed a much better catalytic
activity than ZrO , which was attributed to the higher Lewis
ꢀ
2
within 100 C and then the loss was not obvious (see the
acidity of ZrOPP.
We analyzed the acidity of ZrOPP and ZrO using the
2
NH3-TPD method (Figure 4a). According to the literature,
NH as a smaller molecule compared to pyridine can interact
3
with most of the acidic protons, electron acceptor sites, and
[
31]
hydrogen from neutral or weakly acidic hydroxyls.
As
shown in Figure 4a, ZrOPP exhibited two peaks at 290 and
ꢀ
5
25 C, which were related to the weak Lewis acid sites and
Brønsted acid sites. The Lewis acid sites mainly resulted
from zirconium and the Brønsted acid sites mostly originated
from the hydroxyls in ZrOPP. However, ZrO did not show
2
obvious peaks. Therefore, ZrOPP possessed higher catalytic
acidity than ZrO . At the same time, XPS examinations sug-
2
gested that the binding energies of Zr in ZrOPP were higher
than in ZrO (Figure 4b). The higher effective charge on the
2
[
32,33]
Zr atoms in ZrOPP resulted in a stronger Lewis acidity,
which was beneficial for the activation of the carbonyl
groups in EL improving the reaction activity. In addition,
the CO -TPD method revealed that ZrOPP had much stron-
2
ger basicity than ZrO (Figure 4c). The oxygen atoms attach-
2
ing to phosphate and Zr atoms enhanced the basicity of the
catalyst. The higher basicity was conducive to the dissocia-
tion of the hydroxyl groups of isopropyl alcohol, and
increased the activity of the CTH reaction of EL. Moreover,
ZrOPP had a higher BET surface area, a larger pore diame-
ter, and a larger pore volume (in the ESI), which were propi-
tious to increase the diffusion of the reactants to the activity
center in ZrOPP and accelerated the reaction process.
2
.5 | Mechanism
Through the characterization using NH -TPD and CO -TPD
3
2
techniques, we could discover that there are numerous Lewis
acid sites and basic sites existing in the catalyst, which are
propitious to CTH reactions. The acid sites mainly stemmed
from the zirconium atoms and the basic sites mostly origi-
FIGURE 3 Reusability of the catalyst (a). Reaction conditions: EL
ꢀ
1
mmol; isopropanol 5 mL; ZrOPP 200 mg; reaction temperature 160 C;
nated from oxygen atoms in the pyrophosphate groups.
reaction time 11 hr. Heterogeneity of the catalyst (b). (i) Time-yield plots
for CTH reaction of EL catalyzed by ZrOPP. Reaction conditions: EL
[14]
Based on references,
the possible mechanism of the CTH
reaction has been sorted out in Scheme 1. Firstly, the interac-
tion between acid-basic sites in the catalyst and isopropanol
led to dissociation of isopropanol to the corresponding
1
mmol; isopropanol 5 mL; ZrOPP 200 mg; reaction time 12 hr; reaction
ꢀ
temperature 160 C. (ii) the line shows the GVL yields upon removing the
solid catalyst after 9 hr and continued up to 12 hr

WANG ET AL.
5
SCHEME 1 The plan of possible mechanism for the transformation of EL
to produce GVL catalyzed by ZrOPP via a CTH reaction
alkoxide. Meanwhile, the carbonyl group in EL was acti-
vated by acid sites. After that, proton transfer occurred
between activated carbonyl and alkoxide to form a six-
member intermediate, which further led to produce 4-HPE.
Ultimately, 4-PHE was transformed into GVL through an
intramolecular transesterification under the boost of acid
sites. During the reaction, a part of EL reacted with isopro-
panol to produce isopropyl levulinate first. Then, the same
reaction route occurred and it was similar to EL.
2
.6 | Catalytic performance of different catalysts
Many catalysts were applied in the CTH reaction of EL
using isopropanol as the hydrogen source and solvent. Dif-
ferent catalysts exhibited distinct catalytic activity and the
results are given in Table 1. The reaction did not proceed
without catalyst (entry 1). When ZrO was used as the cata-
2
lyst, the catalytic effect was not satisfying (entry 6). Other-
wise, there were many by-products in the process of
reaction. Comparatively speaking, Zr(OH) and SnO /SBA-
4
2
1
5 were applied to achieve conversion of EL to GVL in iso-
propanol, which exhibited a cracking effect (entry 2–3). In
addition, Ni/MgO was applied to achieve the conversion of
EL with H as the hydrogen source (entry 4). Surprisingly,
2
the reaction proceeded with much higher conversion and
selectivity when Al Zr –300 was used as the catalyst (entry
7
3
5
). But the reaction conditions are too harsh. In our study,
we also synthesized ZrPO to use as a catalyst (entry 7). The
4
catalytic effect was better than that of ZrO . Then, we pre-
2
pared ZrOPP as the catalyst acquiring 94% yield of GVL
and 97% conversion of EL (entry 8). ZrOPP has a high cata-
lytic activity for the CTH reaction of EL and isopropyl levu-
linate. Its excellent performance was attributed to the acid
sites and basic sites interspersing between the layers in the
catalyst, which were conducive to absorb, activate the car-
bonyl group in EL, and intramolecular esterification ulti-
mately. Furthermore, other hydrogen sources and LA were
also used to realize the CTH reaction from EL to GVL with
ZrOPP as the catalyst (entry 9–10). As the catalyst was sen-
sitive to the acidity of LA and was not conducive to the reus-
ability of the catalyst, EL was the best choice as the reaction
3 2
FIGURE 4 NH -TPD spectra (a), XPS spectra of Zr 3d (b), and CO -TPD
spectra (c) of ZrOPP and ZrO
2

6
WANG ET AL.
a
TABLE 1 CTH reactions of EL in isopropanol catalyzed by different catalysts
b
ꢀ
c
Entry
Catalyst
Temp. ( C)
160
Time (hr)
Con.
0
Yield
0
Ref.
1
2
3
4
5
6
7
8
None
11
1
—
Zr(OH)
SnO /SBA-15
Ni/MgO
4
200
94
85
79
96
41
72
97
97
92
45
89
[34]
2
110
8
81
[35]
150
2
70
[36]
Al
ZrO
ZrPO
7
Zr
3
-300
220
4
83
[37]
2
160
11
10
11
11
11
11
11
14
This work
4
160
62
ZrOPP
ZrOPP
ZrOPP
160
94
e
9
160
92
a
1
1
1
0
1
2
160
94
SnOPP
TiOPP
160
20
160
Trace
Trace
a
b
c
d
e
Reaction conditions: a stainless reactor of 22 mL; catalyst, 200 mg; EL, 1 mmol; and isopropanol, 5 mL.
Tem. = temperature.
Conversion (Con.) and GVL yield were determined by GC.
The hydrogen donor was 2-butanol.
The reactant was LA.
substrate. In addition, we also prepared other pyrophosphate
catalysts, such as SnOPP and TiOPP. However, these two
catalysts had little effect on the conversion of EL to GVL
possessing a longer carbon chain afford a relatively long
reaction time and harsh conditions (Entry 3–8). The higher
electron density in the benzene ring results in occupying the
Lewis acid centers and reducing the probability of carbonyl
groups to contact the active centers. In addition, due to the
presence of two benzene rings, the electron cloud density of
carbonyl groups on the two phenyl ketones is greatly
reduced, which leads to the difficulty of the reaction.
(entry 11–12). The amount and strength of acid and basic
sites of ZrPO , SnOPP, and TiOPP were measured using the
4
NH -TPD and CO -TPD (Figure 5). The amount of NH
3
3
2
desorbed from ZrPO was much more than that from SnOPP
4
and TiOPP, which was consistent with their catalytic activ-
ity. Moreover, the pore volume, average diameter, and BET
surface area of ZrPO , SnOPP, and TiOPP were also ana-
4
3
| EXPERIMENTAL
lyzed using the N adsorption–desorption method (in the
2
ESI). The BET surface areas of these three catalysts were
not large, which was unfavorable for bringing into contact
both the catalyst and the reactant sufficiently. However, it
could still illustrate that the synergistic effect of phosphate
and zirconium played a crucial role in the conversion of the
CTH reaction. Moreover, the substitution of pyrophosphate
with a symmetrical structure for phosphate not only
increased the number of acid and basic sites of the catalyst,
but also improved the overall BET surface area, which
enhanced the catalytic effect obviously.
3.1 | Materials and methods
Sodium pyrophosphate (Na P O ꢁ10H O, AR), Zirconium
4
2
7
2
oxychloride (ZrOCl ꢁ8H O, AR), isopropanol (AR), isobuta-
2
2
nol (AR), naphthalene (AR), LA (98%), furfural (AR), benz-
aldehyde (AR), acetophenone (AR), cyclohexanone (AR),
2
- heptanone (AR), 2- octanone (AR), diphenyl ketone
(
[
AR), 4-methyl-2-pentanone (AR), SnCl ꢁ5H O (AR), Ti
4
2
SO ] (AR), phosphoric acid (85%), and ZrO (AR) were
4
2
2
obtained from Sinopharm Chemical Reagent Co. Ltd
Shanghai, China). EL (98%) was gained from Adamas
(
2.7 | Catalytic performance of different reactants
Reagent Co., Ltd. GVL (98%) was purchased from Aladdin
Industrial Inc. (Shanghai, China).The SEM images were
obtained via a HITACHI S-4800 field-emission scanning
electron microscope operated at 15 kV. The TEM was per-
formed on a Jeol JEM model 2,100 microscope operated at
200 kV. FT-IR measurements were recorded on a Nicolet
360 FT-IR instrument (KBr discs) in the wavenumber range
As ZrOPP exhibited outstanding activity toward intermole-
cular hydrogen transfer of EL to GVL, this study was also
extrapolated to explore the reactivity of various carbonyl
compounds (Table 2). All chosen aldehydes and ketones can
be efficiently catalyzed by ZrOPP to produce their corre-
sponding alcohols in high yield. In addition, the reaction
conditions of aldehydes are mild compared with ketones
−
1
of 4,000–500 cm . TGA was performed on a STA409
(Entry 2, 7). The ketones exhibiting a conjugation effect and
instrument under a nitrogen atmosphere at a heating rate of

WANG ET AL.
7
a
TABLE 2 CTH reaction compounds over ZrOPP
b
b
Temp.
( C)
t
Conv.
(%)
Yield
(%)
ꢀ
Entry
Reactant
Product
(hr)
1
100
4
99
99
2
3
150
140
4
99
95
94
90
10
4
5
6
150
100
100
10
10
10
94
95
95
92
90
95
7
8
150
180
11
12
95
90
Trace
Trace
a
Reaction conditions: catalyst, 200 mg; reactant, 1 mmol; and isopropa-
nol, 5 mL.
b
Conversion and yield were determined by gas chromatography (GC).
Following which, it was cooled to room temperature, the
sample was exposed to pyridine vapor under vacuum for
0
.5 hr followed by evacuation of excess pyridine for 0.5 hr.
Then, the cell was heated to 423 K at a rate of 10 K/min and
kept at this temperature for 1.0 hr to remove physisorbed
pyridine. Temperature-programmed desorption of carbon
dioxide (CO -TPD) was performed on a Micromeritics’
2
AutoChem1 II 2920 Chemisorption Analyzer. In the experi-
ment, the catalyst was charged into the quartz reactor, and
the temperature was increased from room temperature to
ꢀ
ꢀ
3
3
00 C at a rate of 10 C/min under a flow of He (50 cm /
ꢀ
min), and then the catalyst was kept at 300 C for 5 hr. After
ꢀ
3
that, the temperature was decreased to 50 C. CO (50 cm /
2
ꢀ
min) was pulsed into the reactor at 50 C under a flow of He
3
(
10 cm /min) until the basic sites were saturated with CO .
2
3
3 2 4
FIGURE 5 NH -TPD spectra (a) and CO -TPD spectra (b) of ZrPO ,
The adsorbed CO was removed by a flow of He (50 cm /
2
SnOPP, and TiOPP
min). When the baseline was stable, the temperature was
ꢀ
ꢀ
increased from 50 to 700 C at a rate of 10 C/min.
Temperature-programmed desorption of ammonia (NH -
3
ꢀ
ꢀ
2
0 C/min from 25 to 600 C. Powder X-ray diffraction
(
XRD) patterns were collected using a Bruker D8 Advance
TPD) was performed on the Micromeritics' AutoChem Auto-
Chem1 II 2920 Chemisorption Analyzer. The catalysts were
charged into the quartz reactor, and the temperature was
ꢀ
powder diffractometer (Cu/Kα) in the 2θ range of 10–90
with a scanning speed of 4 /min at 40 kV and 20 mA. The
ꢀ
ꢀ
increased from room temperature to 300 C at a rate of
X-ray photoelectron spectra (XPS) were recorded using a
Perkin Elmer PHI 5000 ESCT System. The pore volume,
pore size, and specific surface area were measured using N2
adsorption/desorption analysis (Micromeritics ASAP 2020)
at 77 K using BET and Barrett–Joyner–Halenda (BJH)
methods. The contents of Zr and P in ZrOPP were deter-
mined using ICP-AES (VISTA-MPX). The Brønsted and
Lewis acid sites of the sample were determined using the
FT-IR spectra with pyridine as the probe molecule (Py-
FTIR) using a Bruker Tensor 27 spectrometer at 4 cm− res-
olution. Prior to analysis, approximately 25 mg of the sam-
ple was pressed into a 13 mm self-supported wafer and
activated in the IR cell at 673 K for 4.0 hr at 10− Pa.
ꢀ
3
1
0 C/min under a flow of He (50 cm /min), and then the
ꢀ
catalyst was kept at 300 C for 5 hr. After that, the tempera-
ture was decreased to 50 C. NH /He (10/90, 50 cm /min)
was pulsed into the reactor at 50 C under a flow of He
10 cm /min) until the acid sites were saturated with NH .
ꢀ
3
3
ꢀ
3
(
3
3
The adsorbed NH was removed by a flow of He (50 cm /
3
min). When the baseline was stable, the temperature was
ꢀ
ꢀ
increased from 50 to 700 C at a rate of 10 C/min
1
3
.2 | Synthesis of the ZrOPP
In a typical procedure, 1 mmol Na P O ꢁ10H O and 2 mmol
4
2
7
2
3
ZrOCl ꢁ8H O were dissolved in water (25 mL). Then, the
2
2

8
WANG ET AL.
solution of ZrOCl ꢁ8H O was added dropwise to the solu-
4 | CONCLUSIONS
2
2
tion of Na P O ꢁ10H O for about 10 min. After that, the
4
2
7
2
In summary, a porous inorganic Zr-containing catalyst
ZrOPP) was prepared and applied to the CTH reaction with
mixture was continuously stirred for 10 hr. The white precip-
itate was separated by filtration, thoroughly washed with
(
+
−
ꢀ
isopropanol as the hydrogen source. ZrOPP exhibited very
high activity and selectivity for CTH of various carbonyl
compounds, especially for the conversion of EL to GVL,
which was ascribed to the acidic and basic sites interspersing
in the catalyst. In addition, the catalyst can be reused at least
five times without losing activity and selectivity. We believe
that this highly effective and easy to prepare catalyst has
great promise for application in CTH reactions, and the
structures similar to that of ZrOPP can also be used to design
more and better catalysts
H O and ethanol to remove Na and Cl , and dried at 80 C
2
for 12 hr, thus the ZrOPP catalyst was obtained. To deter-
mine the content of Zr and P by ICP, the ZrOPP was dried at
ꢀ
9
0 C for 10 hr. Found: 49.50% Zr, 18.04% P. for Zr P O
2
2
7
(356): 51.12% Zr, 17.42% P.
3.3 | Synthesis of the SnOPP and TiOPP
One mmol Na P O ꢁ10H O and 2 mmol SnCl ꢁ5H O
4
2
7
2
4
2
(
or 2 mmol Ti[SO ] ) were dissolved in water (25 mL).
4 2
Then, the solution of SnCl4 (or Ti(SO ) ) was added drop-
4
2
wise to the solution of Na P O ꢁ10H O for about 10 min.
4
2
7
2
ACKNOWLEDGMENTS
After that, the mixture was continuously stirred for 10 hr.
This work was financially supported by the MOE & SAFEA
for the 111 Project (B13025).
The white precipitate was separated by filtration, thoroughly
+
washed with H O and ethanol to remove Na and
2
−
2−
ꢀ
Cl (or SO4 ), and dried at 80 C for 12 hr, thus SnOPP
and TiOPP catalysts were obtained.
ORCID
Haijun Wang
http://orcid.org/0000-0003-2857-5861
3.4 | Synthesis of the ZrPO4
REFERENCES
First, a certain amount of ZrOCl ꢁ8H O was added in 20 mL
2
2
[
[
1] J. L. Song, L. Q. Wu, B. W. Zhou, Green Chem. 2015, 17, 1623.
2] L. K. Ren, L. F. Zhu, T. Qi, ACS Catal. 2017, 7, 2199.
of distilled water and sonicated to obtain complete dissolu-
tion, afterward H PO (85%) was added dropwise for 10 min
3
4
[3] R. F. Ma, X. P. Wu, T. Tong, ACS Catal. 2016, 6, 2035.
[
[
[
4] M. Al-Naji, A. Yepez, A. M. Balu, J. Mol. Catal. A Chem. 2016, 417, 145.
5] H. Li, Z. Fang, S. Yang, ACS Sustain. Chem. Eng. 2016, 4, 236.
6] D. M. Alonso, S. G. Wettstein, J. A. Dumesic, Green Chem. 2013, 15, 584.
and sonicated to react. When the reaction was completed, a
dispersed white precipitate was obtained. The solid was fil-
tered and washed with distilled water and ethanol several
[7] Y. Yang, C. J. Sun, D. E. Brown, L. Q. Zhang, F. Yang, H. R. Zhao, Green
Chem 2016, 18, 12.
[
ꢀ
times. Subsequently, the catalyst was dried at 80 C for
8] R. Singuru, B. Banerjee, S. K. Kundu, K. Dhanalaxmi, Catal. Sci. Technol.
016, 6, 5102.
9] A. M. Hengne, B. S. Kadu, RSC Adv. 2016, 6, 59753.
ꢀ
1
2 hr and calcined at 500 C for 1 hr to obtain pure ZrPO .
4
2
[
[
[
10] J. R. Ruiz, C. J. Sanchidnan, Curr. Org. Chem. 2007, 11, 1113.
11] B. McNerney, B. Whittlesey, D. B. Cordes, C. Krempner, Chem. Eur. J.
3.5 | Catalytic reactions
2
014, 20, 14959.
In a typical experiment, EL (1 mmol), isopropanol (5 mL),
and the catalyst (200 mg) were charged into a stainless reac-
tor of 22 mL equipped with a magnetic stirrer. After sealing,
the reaction mixture was stirred at a known temperature for
the desired time. Afterward, the products were analyzed
quantitatively by gas chromatography (GC 9700) using
naphthalene as the internal standard, and the identification of
the products was done by GC–MS (SCIONSQ-456-GC).
The yield of GVL was calculated using the following
equation:
[
12] M. B. Gawande, H. Z. Guo, A. K. Rathi, P. S. Branco, Y. Z. Chen,
R. S. Varma, RSC Adv. 2013, 3, 1050.
[13] W. B. Wu, S. J. Zou, L. L. Lin, J. Ji, Y. H. Zhang, B. W. Ma, X. H. Liu,
X. M. Feng, Chem. Commun. 2017, 53, 3232.
[
14] X. Tang, L. Hu, Y. Sun, G. Zhao, W. Hao, L. Lin, RSC Adv. 2013, 3,
0277.
1
[
[
15] M. Chia, J. A. Dumesic, Chem. Commun. 2011, 47, 12233.
16] J. Safaei-Ghomi, H. Shahbazi-Alavi, R. Teymuri, Polycyclic Aromat.
Compd. 2016, 36, 834.
[17] F. K. Li, L. J. France, Z. Cai, Appl. Catal. B Environ. 2017, 214, 67.
18] J. J. Wang, R. Y. Wang, H. M. Zi, J. Chin. Chem. Soc. 2018.
[
[
19] Z. P. Zhu, T. Y. Ma, Y. L. Liu, T. Z. Ren, Z. Y. Yuan, Inorg. Chem. Front.
2014, 1, 360.
[
[
20] J. Wang, S. Jaenicke, G. K. Chuah, RSC Adv. 2014, 4, 13481.
21] R. Y. Wang, J. J. Wang, Mol. Catal. 2017, 441, 168.
Moles of GVL formed
Moles of EL used
GVL yield ðmol%Þ =
× 100%:
[22] T. Z. Ren, Z. Y. Yuan, Chem. Phys. Lett. 2003, 374, 170.
[
23] E. V. Bakhmutova-Albert, N. Bestaoui, V. I. Bakhmutov, A. Clearfield,
A. V. Rodriguez, Inorg. Chem. 2004, 43, 1264.
[
[
24] F. Odobel, B. Bujoli, D. Massiot, Chem. Mater. 2001, 13, 163.
25] C. Jacopin, M. Sawicki, G. Plancque, D. Doizi, F. Taran, E. Ansoborlo,
B. Amekraz, C. Moulin, Inorg. Chem. 2003, 42, 5015.
3
.6 | Reusability of the ZrOPP
In the experiments to test the reusability of the ZrOPP, the
catalyst was recovered by centrifugation, and washed with
ethyl ether (3 × 50 mL). After drying under vacuum at
[
26] Y. Jia, Y. Zhang, R. Wang, J. Yi, X. Feng, Q. Xu, Ind. Eng. Chem. Res.
2012, 51, 12266.
[
27] J. D. Kim, T. Mori, I. Honma, J. Power Sources 2007, 172, 694.
ꢀ
8
0 C for 12 hr, the catalyst was reused for the next run.
[28] J. D. Kim, T. Mori, I. Honma, J. Electrochem. Soc. 2006, 153, A508.

WANG ET AL.
9
[
29] V. V. Ordomsky, J. C. Schouten, T. A. Nijhuis, Chem Sus Chem. 2012, 5,
812.
30] J. José, M. Pedro, O. Pascual, D. J. Jones, R. Jacques, Adv. Mater. 1998,
0, 812.
31] J. Ni, L. Chen, J. Lin, S. Kawi, Nano Energy 2012, 1, 674.
32] J. L. Song, B. W. Zhou, H. C. Zhou, L. Q. Wu, Q. L. Meng, Z. M. Liu,
B. X. Han, Angew. Chem. Int. Ed. 2015, 54, 9399.
SUPPORTING INFORMATION
1
[
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
1
[
[
[
33] B. Tang, W. Dai, X. Sun, G. Wu, N. Guan, M. Hunger, L. Li, Green Chem.
How to cite this article: Wang J, Wang R, Zi H,
Wang H, Xia Y, Liu X. A porous inorganic zirconyl
pyrophosphate as an efficient catalyst for the catalytic
transfer hydrogenation of ethyl levulinate to
γ-valerolactone. J Chin Chem Soc. 2018;1–9. https://
doi.org/10.1002/jccs.201800073
2015, 17, 1744.
[
34] X. Tang, H. Chen, L. Hu, W. Hao, Y. Sun, X. Zeng, L. Lin, S. Liu, Appl.
Catal. B Environ. 2014, 147, 827.
35] S. D. Xu, D. Q. Yu, T. Ye, P. P. Tian, RSC Adv. 2017, 7, 1026.
36] J. Lv, Z. Rong, Y. Wang, J. Xiu, Y. Wang, J. Qu, RSC Adv. 2015, 5,
[
[
72037.
[
37] J. He, H. Li, Y. M. Lu, Y. X. Liu, Z. B. Wu, D. Y. Hu, S. Yang, Appl.
Catal. A Gen. 2016, 510, 11.