Ti functionalized hierarchical-pore UiO-66(Zr/Ti) catalyst for the transesterification of phenyl acetate and dimethyl carbonate

Article abstract of DOI:10.1039/c9nj04241e

Titanates are frequently used as precursors to prepare transesterification catalysts with Ti^IV species. Unfortunately, it is challenging to control the dispersity of Ti^IV active sites on supports. Herein, a series of Ti^IV species is anchored on abundant linker vacancy sites by introducing point and large scale defects in UiO-66(Zr/Ti) with hierarchical-pore structure. The catalyst functionalized by titanium(iv) oxide bis(2,4-pentanedionate) shows excellent catalytic performance in the transesterification of dimethyl carbonate with phenyl acetate. The catalysts are characterized by XRD, FT-IR, N₂ adsorption-desorption, XPS, SEM and STEM-HAADF techniques. The results demonstrate that the delicate mesopores in the support can not only exhibit a large surface area for the distribution of the active sites, but also provide better mass transfer performance. Meanwhile, the introduction of octahedral Ti^IV ions raises the activity of the catalyst via more coordinatively unsaturated Zr^IV sites. Furthermore, using titanium(iv) oxide bis(2,4-pentanedionate) as a Ti source can effectively prevent the condensation of tetrahedral Ti^IV species anchored on the hierarchical-pore UiO-66(Zr/Ti) support.

Full text of DOI:10.1039/c9nj04241e

NJC
View Article Online
View Journal | View Issue
PAPER
Ti functionalized hierarchical-pore UiO-66(Zr/Ti)
catalyst for the transesterification of phenyl
acetate and dimethyl carbonate†
Cite this: New J. Chem., 2019,
43, 16981
ab
ab
a
a
Bingying Jia,
Gongying Wang
Miaojiang Wu,
*
Hua Zhang,
Yi Zeng
and
a
Titanates are frequently used as precursors to prepare transesterification catalysts with TiIV species.
Unfortunately, it is challenging to control the dispersity of TiIV active sites on supports. Herein, a series
of TiIV species is anchored on abundant linker vacancy sites by introducing point and large scale defects
in UiO-66(Zr/Ti) with hierarchical-pore structure. The catalyst functionalized by titanium(IV) oxide
bis(2,4-pentanedionate) shows excellent catalytic performance in the transesterification of dimethyl
carbonate with phenyl acetate. The catalysts are characterized by XRD, FT-IR, N2 adsorption–desorption,
XPS, SEM and STEM-HAADF techniques. The results demonstrate that the delicate mesopores in the
support can not only exhibit a large surface area for the distribution of the active sites, but also provide
better mass transfer performance. Meanwhile, the introduction of octahedral TiIV ions raises the activity
of the catalyst via more coordinatively unsaturated ZrIV sites. Furthermore, using titanium(IV) oxide
bis(2,4-pentanedionate) as a Ti source can effectively prevent the condensation of tetrahedral TiIV
species anchored on the hierarchical-pore UiO-66(Zr/Ti) support.
Received 15th August 2019,
Accepted 8th October 2019
DOI: 10.1039/c9nj04241e
rsc.li/njc
control the types and distributions of hydroxyl groups on their
surface. To our knowledge, the uniform distribution of TiIV
Introduction
Diphenyl carbonate (DPC) has been widely applied as a green species on catalyst supports is difficult, mainly due to the
carbonyl source for various carbonylation reactions.1–3 As a extremely high reactivity of the titanium dioxide precursors.17
highly reactive reagent, it can be conveniently used for the Organo-titanium, for example, Ti(OR)4, tends to hydrolyze
non-phosgene preparation of polycarbonates (PCs) by a melt quickly and form dense precipitates containing Ti(OR)4ꢀx(OH)x
polymerization process.4 Several methods including oxidative species and corresponding oligomers.18 The emerging dense
carbonylation of phenol (PhOH), and transesterification of precipitates lead to partial blocking of the pores in catalyst
dimethyl carbonate (DMC) and PhOH or phenyl acetate (PA) supports and result in the reduction of Ti atom utilization.
have been applied to synthesize DPC.5–10 Among them, the Consequently, the key issues in the synthesis of heterogeneous
transesterification of DMC and PA is regarded as one of the catalysts containing TiIV species is how to effectively prevent the
most suitable synthesis routes for commercial production of aggregation of TiIV active sites by developing a suitable Ti source
DPC due to the large equilibrium constant and atom-economy. and supports.
Generally, heterogeneous catalysts with TiIV species have been
Numerous studies have been conducted concentrating upon
developed to produce DPC through transesterification.11–14 the construction and modification of Zr-based catalyst supports
To date, the reported heterogeneous catalysts are mainly and their application in transesterification processes.19–22
derived from depositing TiO2 on supports including metal Especially, UiO-66(Zr) has attracted particular interest due to
oxide, SiO2 and carbon.11,15,16 However, these traditional inert its remarkable thermal and chemical stability and superior
carriers are unfavourable to achieve the uniform dispersion of catalytic activity based on the Lewis acid sites derived from
the TiIV species, because it is quite difficult to locate and introducing missing-linker and missing-cluster defects.23–25
In general, the surface defects, internal imperfections, and
the interaction with metal and support are in favor of enhan-
a Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences,
cing the catalytic activity of UiO-66(Zr).26 UiO-66(Zr) consists
Chengdu 610041, China. E-mail: wanggongying1102@126.com
b National Engineering Laboratory for VOCs Pollution Control Material &
of hexanuclear clusters in which each ZrIV ion is connected
to linear 1,4-benzenedicarboxylate (BDC) ligands. Kim et al.27
Technology, University of Chinese Academy of Sciences, Beijing 101408, China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9nj04241e reported that TiIV ions could be incorporated into the node of
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16981--16989 | 16981

View Article Online
Paper
NJC
UiO-66(Zr) by exchanging with partial ZrIV ions. Nguyen et al.28
and Santaclara et al.29 found that TiIV ions could also be
conjectured to occupy the missing-linker sites of UiO-66(Zr).
The modification of the electronic situation at linker and/or
metal node vacancies could significantly affect the reactive
properties of the catalyst.26,30,31 Vermoortele et al.31 demon-
strated that the excellent catalytic activity of UiO-66(Zr) can be
ascribed to the coordinatively unsaturated ZrIV sites, acting as
Lewis acid sites, rather than a modified Lewis acid strength.
Moreover, it was proved that a large number of missing-linker
defects could improve the specific surface area and expose
more Lewis acid sites for UiO-66(Zr).30 However, the pure
microporous structure of UIO-66(Zr) might restrict the accessi-
bility of bulky molecules within it. Therefore, the development
of hierarchical-pore UiO-66(Zr) containing both micropores
and mesopores is of great significance. In the interconnected
network structure of UiO-66(Zr), micropores are beneficial to
maintaining the functionality of the catalyst support, while the
presence of mesopores can bring a positive impact on the
molecular diffusion.32
In this work, the catalysts derived from Ti functionalized
hierarchical-pore UiO-66(Zr/Ti) (H-UiO-66(Zr/Ti)) were prepared.
The starting H-UiO-66(Zr) sample was synthesized by using
an in situ self-assembly template strategy. TiIV ions were then
incorporated in the node and at hydroxyl sites of the H-UiO-66(Zr)
framework by ion-exchange. Furthermore, different TiIV species
were grafted on supports for preparing Ti functionalized H-UiO-
66(Zr/Ti) catalysts. The catalysts were used for the transesteri-
fication of DMC and PA, and the reaction conditions were
investigated. The as-prepared Ti functionalized H-UiO-66(Zr/Ti)
catalysts in this work exhibit high activity and small consump-
tion, resulting from the appropriate distribution and state of
TiIV ions, better mass transfer and a considerable density of
structural defects of the support. Furthermore, H-UiO-66(Zr/Ti)
as a catalyst support shows catalytic activity to a certain extent in
Scheme 1 Schematic representation for the preparation of H-UiO-66(Zr).
tetrabutyl titanate (Ti(OBu)4, 98%) were purchased from Aladdin
chemistry Co. Ltd. TS-1 was purchased from XFNANO Materials
Tech Co. Ltd. All the reagents and solvents were used as received
without further purification. Titanium-Citrate was self-made and
the ORTEP structure is shown in Scheme S1 (ESI†).
Preparation of hierarchical-pore UiO-66 (H-UiO-66)
The synthesis of H-UiO-66 was achieved as per the reported
method with some modifications (Scheme 1).33 Zn(NO3)2ꢁ6H2O
(0.592 g, 2 mmol), HBC (7.320 g, 60 mmol), ZrCl4 (0.480 g,
2 mmol), and TPA (0.664 g, 4.0 mmol) were dissolved in 80 mL
of DMF. The obtained suspension was sonicated for 20 minutes
and then transferred into a 100 mL autoclave in a preheated
oven at 120 1C for 24 hours. After cooling to room temperature,
the white product was centrifugated, washed several times with
a mixture of DMF and methanol, and dried at 100 1C for 12 h.
Then, the H-UiO-66(Zr) precursor was dispersed in HCl solution
(pH = 1.0) and stirred for about 2 h to dissolve the acid-sensitive
metal–organic assembling template (Zn4O(BC)6) and obtain
the hierarchical-pore structure. The product was separated by
centrifugation and washed with DMF and methanol several
times. Finally, the H-UiO-66(Zr) support was dried at 100 1C
overnight.
this transesterification reaction. The aims of this work are Preparation of mixed-metal H-UiO-66(Zr/Ti)
described as follows: (1) the possibility of the preparation of
H-UiO-66(Zr/Ti) was prepared following a modification of a
previously reported procedure (Scheme 2, Step 1).27,34 Generally,
0.75 g of TiCp2Cl2 (3 mmol) was added into 60 mL of
N,N-dimethylformamide (DMF), and the mixture was treated by
ultrasonic wave for 40 minutes. Then, 0.8 g of the prepared H-UiO-
66(Zr) was introduced into it and the suspension was transferred
to a 100 mL autoclave, which was heated for 5 d at 85 1C. After
cooling, the solid was obtained by filtration and washed with a
mixture of DMF and methanol several times. The solid was left to
soak in methanol for 2 days to remove the TiIV species adsorbed
the H-UiO-66(Zr/Ti) sample with abundant point and mesopore
defects was explored by the incorporation of TiIV ions into
H-UiO-66(Zr). (2) Investigating the effect of Ti sources on the
distribution and state of TiIV active sites in order to prevent the
condensation of TiIV species in the Ti functionalized catalysts.
(3) The impact of different TiIV coordination environments on
the activity of the transesterification catalysts was evaluated.
Experimental
Materials
Zirconium tetrachloride (ZrCl4, 98%) and Titanium(IV) oxide
bis(2,4-pentanedionate) (TiO(acac)2, 98%) were purchased from
Shanghai Macklin Biochemical Co. Ltd. Zinc nitrate hexahydrate
(Zn(NO3)2ꢁ6H2O, 99%), 2-methylimidazole (98%), N,N0-dimethyl-
formamide (DMF, 99.5%), and benzoic acid (HBC, 99%) were
purchased from Chengdu Chron Chemical Co. Ltd. Terephthalic
acid (TPA, 99%), titanocene dichloride (TiCp2Cl2, 98%) and
Scheme 2 Schematic representation for the preparation of Ti function-
alized H-UiO-66(Zr/Ti).
16982 | New J. Chem., 2019, 43, 16981--16989 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
on the surface. Finally, the desired product was centrifuged and the reaction was conducted at 170–190 1C for a prescribed time.
dried at 100 1C overnight.
During the reaction, the unreacted DMC and the generation of
methyl acetate, as the light components, were collected in a
receiver flask. After completion of the reaction, the reaction
Preparation of Ti functionalized H-UiO-66(Zr/Ti)
Ti functionalized H-UiO-66(Zr/Ti) with different Ti sources was products were obtained by centrifugation and determined by
synthesized using the reported method; a slight modification GC-MS using a HP-6890/5973 system. Moreover, the products
was made to achieve a better effect (Scheme 2, Step 2).35 were quantitatively analysed using an Agilent Technologies
TiO(acac)2, TiCp2Cl2, Ti(OBu)4 and Ti-Citrate were selected as 7820 A gas chromatograph equipped with a DB-35 capillary
Ti sources, respectively, for TiO(acac)2-H-UiO-66(Zr/Ti) with a column (30 m ꢂ 320 mm ꢂ 0.25 mm) and a flame ionization
Ti : H-UiO-66(Zr/Ti) weight molar ratio of 0.2 : 1 (denoted 20% detector (FID).
TiO(acac)2-H-UiO-66(Zr/Ti)). 0.9 g of TiO(acac)2 was dissolved in
a mixture of 150 mL of methanol and 1 mL of deionized water.
Then, this solution was mixed with 0.8 g of H-UiO-66(Zr/Ti) in a
Results and discussion
Catalyst characterization
250 mL flask. The suspension was heated to reflux at 65 1C in
an oil bath for 12 h. After cooling, the powder was obtained
through centrifugation and immersed in methanol (100 mL) for
12 h to remove unreacted metal ions. The mixture was centri-
fuged and dried to get the 20%TiO(acac)2-H-UiO-66(Zr/Ti)
sample. For comparison, 20%TiO(acac)2-TS-1 was prepared
using a similar method.
Fig. 1(a) presents the XRD patterns of the Ti functionalized
H-UiO-66(Zr/Ti) catalysts. All of the Ti functionalized samples
show XRD patterns similar to that of pure H-UiO-66(Zr),
indicating that the structure of the H-UiO-66(Zr) framework
is retained. However, for TiIV ion exchanged samples, the most
intense peak observed at around 7.31 shifts to a higher value
(Fig. S1, ESI†). The phenomenon is a consequence of the
shrinking of the crystal lattice due to the smaller radius of TiIV
compared to that of ZrIV,36,37 which indicates that some TiIV
species are exchanged rather than grafted on linker vacancy
Characterization
X-ray diffraction (XRD) patterns were recorded using a Rigaku
X-ray diffractometer with monochromatic Cu Ka in a scan
range of 41 to 511. N2 adsorption–desorption isotherms were
determined by using an ASAP 2460 at 77 K. Before adsorption
measurements, the samples were degassed under vacuum at
120 1C for 5 h. The surface area was estimated by the Brunauer–
Emmett–Teller (BET) equation, and micropore and mesopore
size distributions were calculated using the H–K and Barrett–
Joyner–Halanda (BJH) methods, respectively. Fourier transform
infrared (FT-IR) spectroscopy was performed using a Nicolet-
6700 instrument with samples prepared as KBr pellets at
wavenumbers ranging from 4000 to 450 cmꢀ1. X-ray photo-
electron spectroscopy (XPS) spectra were recorded using an
Escalab 250Xi instrument with monochromatic Al Ka radiation.
The morphological properties of the samples were imaged by
using a JEOL JEM-5410LV scanning electron microscope (SEM)
operated at 20 kV. The high-angle annular dark-field scanning
transmission electron microscopy (HAADF-STEM) images and
energy dispersive X-ray spectroscopy (EDX) mappings were
generated using an FEI Tecnai G2 F20 transmission electron
microscope equipped with an energy-dispersive X-ray spectro-
meter system and a high-angle angular dark-field detector was
used at 200 kV. Thermogravimetric analysis (TGA) was performed
using a TGA Q500 V20.10 Build 36 instrument with heating at
10 1C minꢀ1 in a N2 atmosphere.
Activity evaluation of the catalysts
The transesterification reaction of DMC (40 g, 0.44 mol) and PA
(30 g, 0.22 mol) was carried out in a 100 mL three-neck glass
flask. Additionally, the apparatus consisted of a magnetic
stirring bar, a dropping funnel, a thermometer and a fractio-
nating column. A mixture of 30 g of PA and 0.6 g of catalyst was
charged into the flask. When the reaction mixture reached a
temperature of 180 1C, DMC was dropwise added into it, and
Fig. 1 (a) XRD patterns and (b) FT-IR spectra of the Ti functionalized
H-UiO-66(Zr/Ti) catalysts.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16981--16989 | 16983

View Article Online
Paper
NJC
at around 821 cmꢀ1 can be observed and demonstrates the
stretching vibration of –CH in the Cp ring in the 20%TiCp2Cl2-
H-UiO-66(Zr/Ti) sample. Consequently, a partial Cp ring can
still be retained in the reaction products of TiCp2Cl2 and
H-UiO-66(Zr/Ti).39 For the 20%TiO(acac)2-H-UiO-66(Zr/Ti) sample,
a new peak at approximately 1721 cmꢀ1 appeared, corresponding
to the CQO stretching vibration of keto,40 which resulted from
the proton migration from OH to acetylacetonate groups.41
Furthermore, Ti(OMe)4ꢀx(acac)x is formed via the reaction
between TiO(acac)2 and methanol, as reported.42 The spectrum
of 20%Ti(OBu)4-H-UiO-66(Zr/Ti) shows broad bands at
450–800 cmꢀ1, which are the characteristic frequencies of
Ti–O–Ti bonds.43 However, no Ti–O–Ti bond derived from
oligomers containing TiIV can be detected when using TiCp2Cl2
or TiO(acac)2. Additionally, for the 20%Ti-Citrate-H-UiO-66(Zr/Ti)
catalysts, the antisymmetric stretching vibrations in the range of
1663–1580 cmꢀ1 can be attributed to the presence of carboxylate
carbonyls, indicating the formation of free citrate ligand or
monodentate Ti-Citrate.44–47
As shown in Fig. 2(a) and Fig. S2(a) (ESI†), the N2 sorption
isotherms of all samples at low relative pressure correspond to
the existence of micropores, and those at relative pressure
higher than 0.6 indicate the generation of mesopores in the
framework of the catalysts. Moreover, for all samples, the
hystereses are characteristic of solids consisting of nonuniform
size and shape pores, because Zn4O(BC)6, as a mesopore
template, can undergo decomposition or rearrangement in
the in situ self-assembly reaction.33,48 Some structural features
of Ti functionalized H-UiO-66(Zr/Ti) catalysts are listed in
Table 1 and Fig. S3 (ESI†). The specific surface area of
Fig. 2 (a) N2 adsorption–desorption isotherms at 77 K and (b) mesopore
size distributions of Ti functionalized H-UiO-66(Zr/Ti) prepared with
different Ti sources.
sites of the H-UiO-66(Zr) framework. Additionally, the vanishing of H-UiO-66(Zr/Ti) and micropore size slightly increase to 1220
the characteristic diffraction peak of TiO2 crystalline phase demon- and 0.62 from the original 1108 m2 gꢀ1 and 0.60 nm, which
strates the amorphous state of the TiIV species. Moreover, the low suggests that the TiIV ions are incorporated into the secondary
relative crystallinity of 20%Ti(OBu)4-H-UiO(Zr/Ti) can be ascribed building unit (SBU) clusters and few TiIV species are blocking
to the rapid hydrolysis of Ti(OBu)4 and the generation of more the pores. The increase in the surface area and pore size of
Ti(OBu)4ꢀx(OH)x oligomers.
H-UiO-66(Zr/Ti) can be ascribed to the lighter atomic weight
The reactions between different Ti sources and H-UiO-66(Zr/Ti) and lower maximum coordination number of TiIV compared
were explored by FT-IR spectroscopy (Fig. 1(b)). Two strong absorp- to those of ZrIV.35,49 However, the BET surface area of 20%TiO-
tion bands at approximately 1550 and 1410 cmꢀ1 are associated (acac)2-H-UiO-66(Zr/Ti) and 20%TiCp2Cl2-H-UiO-66(Zr/Ti) drops
with asymmetric stretching and symmetric stretching vibrations of slightly, indicating that some TiIV ions can be anchored on the
O–CQO groups in the skeletons of H-UiO-66(Zr/Ti).38 The broad nodes of UiO-66(Zr/Ti). When Ti(OBu)4 is selected as the Ti source,
bands at 3400 cmꢀ1 are ascribed to hydroxyl vibration. The peak the surface area of the sample is significantly reduced. The result
Table 1 Pore features of Ti functionalized H-UiO-66(Zr/Ti) catalysts
Sample
SBET (m2 gꢀ1
)
Smicro (m2 gꢀ1
)
Smeso (m2 gꢀ1
)
Micropore size (nm)
Mesopore size (nm)
TS-1
413
982
575
1108
1220
1031
1092
588
270
917
423
755
865
694
771
326
573
143
65
0.56
0.58
0.64
0.60
0.62
0.61
0.61
0.50
0.59
29.7
18.2
25.4
12.9
13.9
13.9
13.3
17.4
13.7
Conventional UiO-66(Zr)
H-UiO-66(Zr) precursor
H-UiO-66(Zr)
152
353
355
337
321
262
286
H-UiO-66(Zr/Ti)
20%TiCp2Cl2-H-UiO-66(Zr/Ti)
20%TiO(acac)2-H-UiO-66(Zr/Ti)
20%Ti(OBu)4-H-UiO-66(Zr/Ti)
20%Ti-Citrate-H-UiO-66(Zr/Ti)
859
SBET is the Brunauer–Emmett–Teller (BET) specific surface area. Smicro is the t-plot-specific micropore surface area calculated from the N2
adsorption–desorption isotherm. Smeso is the specific mesopore surface area estimated by subtracting Smicro from SBET. The micropore diameter
represent the median pore width which was calculated by Horvath–Kavazoe method.
16984 | New J. Chem., 2019, 43, 16981--16989 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
supports the argument that many Ti(OBu)4ꢀx(OH)x oligomers tetrahedral sites. Therefore, the incorporation of TiIV ions into the
are deposited in the pores of the support. Additionally, the H-UiO-66(Zr) framework is achieved by using TiCp2Cl2 as a Ti
introduction of Ti-Citrate can cause a decrease in the catalyst source under solvent thermal conditions. Moreover, the contents
surface area due to its larger molecular diameter and stability of octahedral coordinated TiIV and tetrahedral coordinated TiIV
in solution. Moreover, in Fig. 2(b) and Fig. S2(b) (ESI†), the are approximately 2.78 at% and 1.05 at% in the sample (Table S2,
mesopore size distributions (calculated by BJH method) of ESI†). When H-UiO-66(Zr/Ti) was treated by refluxing with a
20%TiO(acac)2-H-UiO-66(Zr/Ti) and 20%TiCp2Cl2-H-UiO-66(Zr/Ti) different Ti source, more surface tetrahedral coordinated TiIV
are similar to that of H-UiO-66(Zr/Ti), whereas the mesopore species appeared. Besides, as shown in Table S2 (ESI†), the
volume is slightly smaller than that of unmodified H-UiO- octahedral coordinated Ti/Zr molar ratios almost remain constant
66(Zr/Ti). These results indicate that few TiIV species can be for 20%TiCp2Cl2-H-UiO-66(Zr/Ti) and 20%TiO(acac)2-H-UiO-
observed on the oligomers.
66(Zr/Ti) samples. However, the central TiIV ion of mononuclear
The chemical states and contents of the Ti and O species Ti-Citrate has adopted a six-coordinated arrangement to three
over the surface of samples were studied by using X-ray photo- a-carboxyl groups derived from three citrate ligands, forming
electron spectroscopy (XPS). According to the high resolution a distorted octahedral geometry.47 There is an equilibrium of
scan spectra of Ti atoms illustrated in Fig. 3(a), the Ti 2p3/2 dissociation of the tetradentate and hexacoordinate Ti-Citrate
signal of the samples can be decomposed into two contributions complexes in PH E 7 solution.46 Furthermore, Ti-Citrate possesses
at 458.1 eV and 458.9 eV, which correspond to octahedral excellent hydrolytic stability and less tetradentate hydrolysis
coordinated TiIV and tetrahedral coordinated TiIV, respectively.50 products. Consequently, for the 20%Ti-Citrate-H-UiO-66(Zr/Ti)
The two different contributions of TiIV of the a photoelectron sample, the percentage of octahedral coordinated TiIV species is
signal were determined by the percentage of their areas (Table S1, significantly increased as compared to that of 20%TiO(acac)2-
ESI†). After ion exchange treatment, the TiIV cations of H-UiO- H-UiO-66(Zr/Ti). Additionally, in Fig. 3(b), the samples containing
66(Zr/Ti) are distributed as 72.5% in octahedral sites and 27.5% in TiIV species has asymmetric O 1s signals that can be decomposed
into four contributions locating at 532.65, 531.7, 530.7 and
530.1 eV. The four peaks are assigned to hydroxyl, CQO, Zr–O,
and Ti–O species on the surface of the samples, respectively.34,51,52
The existence of the Ti–O bond implies that the TiIV moieties are
probably introduced into the framework or linker vacancy sites of
H-UiO-66(Zr) via the formation of an oxo-bridge.
As shown in Fig. S4 (ESI†), the hierarchical-pore UiO-66(Zr)
shows irregular spherical agglomerations instead of the layered
structure of conventional UiO-66(Zr) with pure micropores. And
the morphology of Ti functionalized H-UiO-66(Zr/Ti) is dis-
played in Fig. 4. The samples present spherical agglomerations,
which are similar to that of H-UiO-66(Zr). In Fig. 4(a) and (b),
the outer surface of the spheres is smooth, indicating that
almost no Ti oligomer particles can be observed on the surface
of 20%TiCp2Cl2-H-UiO-66(Zr/Ti) and 20%TiO(acac)2-H-UiO-
66(Zr/Ti). However, the surface of 20%Ti(OBu)4-H-UiO-66(Zr/Ti)
(Fig. 4(c)) became rough due to the rapid hydrolysis of Ti(OBu)4
Fig. 3 (a) Ti 2p XPS spectra of the synthesized samples and (b) O 1s XPS
Fig. 4 SEM images of samples.
spectra of the synthesized samples.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16981--16989 | 16985

View Article Online
Paper
NJC
catalysts were investigated by TG analysis over the temperature
range of 30–800 1C under a N2 atmosphere. As displayed in
Fig. S5 (ESI†), all catalysts show a slight loss of mass upon
heating at 100–200 1C due to the removal of solvent. The result
indicates that the structure of all catalysts can be retained at the
reaction temperature (170–190 1C). Additionally, the industria-
lized production of DPC is divided into two steps: the first step
is transesterification to obtain DPC and methyl phenyl carbo-
nate (MPC) by using transesterification catalysts, and the next
step is disproportionation of MPC to get high purity DPC.
Therefore, MPC and DPC were considered as the main products
in this work. The different parameters, such as reaction time
and the performance of reused catalyst, affecting the reaction
were discussed.
Fig. 5 (a), (c), (e), and (g) STEM-HAADF image and (b), (d), (f), (h) corresponding
Ti elemental maps of 20%TiCp2Cl2-H-UiO-66(Zr/Ti), 20%TiO(acac)2-H-UiO-
66(Zr/Ti), 20%Ti(OBu)4-H-UiO-66(Zr/Ti) and 20%Ti-Citrate-H-UiO-66(Zr/Ti).
and the generation of lots of Ti(OBu)4ꢀx(OH)x oligomers.17 Addi-
tionally, some floccules appeared on the surface of 20%Ti-Citrate-
H-UiO-66(Zr/Ti) because of the centrosymmetric oligomers trans-
forming from the linear hydrogen bonding of the uncoordinated
protonated b-carboxylic acid to the deprotonated b-carboxyl
group.46 The results were in good agreement with those from
N2 adsorption–desorption.
The STEM-HAADF and Ti elemental mappings of Ti functio-
nalized H-UiO-66(Zr/Ti) catalysts are shown in Fig. 5. The Ti
elemental mappings of 20%TiCp2Cl2-H-UiO-66(Zr/Ti) and
20%TiO(acac)2-H-UiO-66(Zr/Ti) demonstrate the homogeneous
distribution of TiIV species in the whole catalyst crystals.
Therefore, the uniformly distributed tetradentate TiIV species
can be realized by using TiCp2Cl2 or TiO(acac)2 as a Ti source.
In contrast, 20%Ti(OBu)4-H-UiO-66(Zr/Ti) displays a slightly
higher Ti content inside the particles than the surface amount
of Ti, indicating that more TiIV species block the internal pores
in the particles. In addition, the content of Ti species on the
surface is higher than that in the inner part of the particles,
which suggests that more Ti-Citrate or its oligomers were
anchored on the surface of H-UiO-66(Zr/Ti) due to the larger
molecular diameter.
Effect of Ti source of catalysts on transesterification
The effect of the Ti sources of catalysts on the yield of MPC and
DPC in the transesterification of PA and DMC was investigated.
As shown in Table 2, H-UiO-66(Zr) exhibits higher PA conver-
sion and MPC and DPC yields than UiO-66(Zr), which may be
due to the mesopores of H-UiO-66(Zr) with large-scale defects
and decreased molecular diffusion resistance (as shown in
Fig. S5, ESI† and Table 1).26 Moreover, in the TG curves (as
shown in Fig. S5, ESI†), the weight loss of H-UiO-66(Zr) in the
450–700 1C interval is lower than that of UiO-66(Zr), which
implies more miss-linking sites that can be ascribed to the
increase of structural disorder.53 In addition, the catalyst
activity can be enhanced by introducing TiIV ions, which
are mainly six-coordinated with octahedron geometry, in the
framework of H-UiO-66(Zr). Although six-coordinated TiIV ions
show no Lewis acidity, the appearance of them can lead to the
formation of more structural defects to increase the Lewis acid
sites of catalysts.53 The reason is that the ordering of the
framework structure of H-UiO-66(Zr) is further decreased by
different cation radiuses of TiIV and ZrIV in the octahedral coor-
dination sites of SBU.27 More importantly, adequate hydroxyl
groups can be provided by the defect sites derived from the miss
links and they can be used to coordinate with the TiIV species.35
For 20%TiO(acac)2-H-UiO-66(Zr/Ti) catalyst, the conversion of PA
Properties of the catalysts
A series of Ti functionalized H-UiO-66(Zr/Ti) as heterogeneous reaches a maximum value (81.32%); the yields of DPC and MPC
catalysts was employed for the transesterification of PA and are 54.05% and 19.70%, respectively. Furthermore, compared
DMC to synthesize DPC. The stabilities of the prepared to other catalysts reported in the literature, the as-synthesized
Table 2 Effect of Ti source on the activity of different catalysts
Yield (%)
PA
Transesterification
selectivity (%)
Cat
conversion (%)
DPC
MPC
PhOH
UiO-66(Zr)
H-UiO-66(Zr)
H-UiO-66(Zr/Ti)
20%TiCp2Cl2-H-UiO-66(Zr/Ti)
20%TiO(acac)2-H-UiO-66(Zr/Ti)
20%Ti(OBu)4-H-UiO-66(Zr/Ti)
20%Ti-Citrate-H-UiO-66(Zr/Ti)
TS-1
35.94
42.19
51.24
79.35
81.32
55.87
45.48
42.82
55.93
16.95
24.00
30.14
50.28
54.05
32.04
25.85
22.46
30.58
13.68
12.63
15.93
22.23
19.70
16.08
14.39
15.31
19.55
5.31
5.56
5.17
6.84
7.57
7.75
5.24
5.05
5.80
85.22
86.82
89.91
91.38
90.69
86.13
88.48
88.21
89.63
20%TiO(acac)2-TS-1
Reaction conditions: n(DMC) = 0.44 mol, n(PA) = 0.22 mol, m(Cat) = 0.6 g, reaction temperature 170–190 1C, reaction time 7 h.
16986 | New J. Chem., 2019, 43, 16981--16989 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
catalyst showed better catalytic activity at lower dosage. For
example, Cao et al.8 found that the yields of DPC and MPC are
29.1% and 41.7% for the MoO3 catalyst in a high pressure
autoclave, and the amount of catalyst reached up to 5 wt% with
respect to the mass of PA. Wang et al.54 disclosed that the
conversion of PA was 79.21% when the dosage of TiO2/SiO2
was 4 wt%. We previously reported that the conversion of
PA could achieve 74.65% using mesoporous amorphous TiO2
shell-coated ZIF-8 as a catalyst, and the catalyst concentration
was 3 wt%.55 As expected, the tetra-coordinated titanium, as an
active site, can interact with the CQO group of DMC and MPC to
improve the catalytic activity.11 The result can be attributed to a
suitable hydrolysis rate of TiO(acac)2 and tetra-coordinated
TiIV products. However, no obvious improvement of the PA
conversion and product yield can be observed using Ti(OBu)4
as a Ti source, because Ti(OBu)4 is highly sensitive to water or
alcohols, and tends to generate dense precipitates fast, which
can block the pores prior to reaction with the OH group of
H-UiO-66(Zr/Ti).18 Besides, Ti-Citrate cannot provide enough
tetra-coordinated TiIV active sites to impel the proceeding
of transesterification due to hydrolysis resistance. Additionally,
TS-1 zeolite and 20%TiO(acac)2-TS-1 were applied in this
reaction, because the framework of TS-1 possesses tetra-
coordinated TiIV species (as shown in Fig. S6, ESI†).56 As shown
in Table 2, when using TS-1 or 20%TiO(acac)2-TS-1 as a catalyst,
the PA conversion was lower than 60%, which can be ascribed to
the fewer TiIV species and smaller surface area than those of
Ti functionalized H-UiO-66(Zr/Ti) (see Table 1 and Table S3,
Fig. S7, ESI†).
Fig. 6 (a) Effect of reaction time on the catalytic activity of
20%TiO(acac)2-H-UiO-66(Zr/Ti) and (b) the reusability of 20%TiO(acac)2-
H-UiO-66(Zr/Ti).
Effect of reaction time of catalysts on transesterification
leaching is the main cause of catalyst deactivation. Significantly,
The reaction time has a significant influence on the activity of after 5 runs, the catalytic activity of 20%TiO(acac)2-H-UiO-
20%TiO(acac)2-H-UiO-66(Zr/Ti) (Fig. 6(a)). The yields of MPC 66(Zr/Ti) is close to that of H-UiO-66(Zr/Ti). Furthermore, the
and DPC increase with the extending of the reaction time from conversion of PA and the yield of products gradually stabilized
0 to 7 h. However, when the reaction time comes to 8 h, no with the cycle number of catalysts increasing from 5 to 7. These
significant effect on the conversion of PA and yield of products results can arise from the different stability of octahedral and
is observed. When the reaction time takes longer than 6 h, the tetrahedral coordinated TiIV ions.
MPC yield starts to decline slightly, but the yields of DPC and
The used 20%TiO(acac)2-H-UiO-66(Zr/Ti) catalysts were
phenol (PhOH) keep increasing. These results suggest that analyzed by means of XRD and XPS. As shown in Fig. 7(a),
the generated MPC is disproportionated or reacted with DMC the XRD patterns confirm that the framework structure of the
to DPC.51,57 Moreover, the generation of PhOH by-product 20%TiO(acac)2-H-UiO-66(Zr/Ti) is unchanged, but the relative
is mainly attributed to the rearrangement or hydrogenolysis crystallinity of it is decreased after 7 runs. Besides, there is no
of the transesterification products through using Lewis acid shift of diffraction angle at 2y E 7.31 (Fig. S8, ESI†), indicating
catalysts.
that some octahedral TiIV ions, which are introduced by post-
synthetic exchange, still remain in the metal cluster SBUs.
However, there are only octahedral coordinated TiIV ions left
Reusability of 20%TiO(acac)2-H-UiO-66(Zr/Ti)
Since 20%TiO(acac)2-H-UiO-66(Zr/Ti) shows the highest activity in the catalyst after 7 consecutive cycles (Fig. 7(b)). The results
of the Ti-functionalized H-UiO-66(Zr/Ti) catalysts, the reusability support the hypothesis that the tetrahedral coordinated TiIV ions
of it was investigated by performing the reaction under exhibit a higher catalytic activity than the octahedral coordinated
optimized conditions (n(DMC) = 0.44 mol, n(PA/DMC) = 1 : 2, TiIV ions for the transesterification process. Nevertheless, after 7
m(Cat) = 0.6 g, reaction temperature 170–190 1C, reaction time consecutive runs, the PA conversion remains at 47.25%, which
7 h). The used catalyst was centrifuged, washed with DMC is higher than that of H-UiO-66(Zr). This can be related to more
several times, and dried at 100 1C after each run. Obviously, miss-linking sites in the H-UiO-66(Zr/Ti) support.37 In the catalyst
with the increase of the cycle number of catalysis, the conver- support, the point defects, caused by the introduction of octa-
sion of PA and the yields of DPC and MPC show a gradual hedral coordinated TiIV ions, are well retained after being used in
decline tendency. We conclude that tetrahedral TiIV active site 7 consecutive cycles and help to increase the activity of the catalyst.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16981--16989 | 16987

View Article Online
Paper
NJC
TiO(acac)2 exhibits excellent distribution of tetra-coordinated TiIV
species on H-UiO-66(Zr/Ti), which can be attributed to the suitable
hydrolytic rate and hydrolysates. However, the excess Ti–O–Ti
bonds and TiIV oligomers, which are usually provided by
deposited amorphous TiO2 on catalyst supports, impede the
catalytic efficiency in this reaction due to a lower utilization of
TiIV active sites and a larger diffusion resistance in the pores
of catalyst supports. Consequently, this study provides a
promising route to overcome diffusion limitations, increase
the concentration of Lewis acid sites in catalysts and effectively
inhibit the generation of TiIV oligomers for catalysts containing
tetra-coordinated titanium.
Conflicts of interest
The authors declare that they have no conflicts of interest.
Acknowledgements
We gratefully acknowledge the financial support from the
Innovation Academy for Green Manufacture, Chinese Academy
of Sciences (No. IAGM-2019-A10) and the Department of
Science and Technology of Sichuan Province (No. 18ZHSF0011).
References
1 Z. Du, Y. Xiao, T. Chen and G. Wang, Catal. Commun., 2008,
9, 239–243.
2 P. Wang, S. Liu, F. Zhou, B. Yang, A. S. Alshammari and
Y. Deng, RSC Adv., 2015, 5, 84621–84626.
Fig. 7 (a) XRD patterns of fresh and used catalysts; (b) XPS spectra of fresh
and used catalysts.
3 W. Sun, J. Shao, Z. Xi and L. Zhao, Can. J. Chem. Eng., 2017,
95, 353–358.
4 Z. Wang, X. Yang, S. Liu and G. Wang, J. Mol. Catal. A:
Chem., 2016, 424, 77–84.
Conclusions
The hierarchical-pore UiO-66(Zr/Ti) was prepared by using acid-
sensitive Zn-based metal–organic fragments as templates. And
then the octahedral coordinated TiIV ions were incorporated
into the framework of H-UiO-66(Zr) through ion exchange. The
properties of the catalysts demonstrate that the presence of
defects, including large-scale defects (mesopores) and point
defects (linker vacancies or M–OH), in H-UiO-66(Zr/Ti) can
enhance the catalytic activity for transesterification. Moreover,
the different tetra-coordinated TiIV species were anchored on
the linker vacancy sites of H-UiO-66(Zr/Ti) via a post-grafting
method. The influence of coordination environment of TiIV
ions on the catalytic activity for the transesterification of DMC
with PA was also discussed. The tetra-coordinated TiIV ions are
the main factor responsible for the catalytic activity during the
synthesis of DPC from DMC and PA. With small catalyst
consumption, 20%TiO(acac)2-H-UiO-66(Zr/Ti) exhibits the best
activity and selectivity, indicating that the TiIV species grafted
on H-UiO-66(Zr/Ti) can have a remarkable influence on the
activity of catalysts. Furthermore, after 7 consecutive runs,
the yields of transesterification products can remain at 42.3%
due to the Lewis acid sites of the catalyst derived from the
abundant structure defects. Among the four kinds of Ti sources,
5 S. Luo, T. Chen, D. Tong, Y. Zeng, Y. Lei and G. Wang, Chin.
J. Catal., 2007, 28, 937–939.
6 J. Gong, X. Ma and S. Wang, Appl. Catal., A, 2007, 316, 1–21.
7 S. Luo, Y. Chi, L. Sun, Q. Wang and C. Hu, Catal. Commun.,
2008, 9, 2560–2564.
8 P. Cao, X. Yang, C. Tang, J. Yang, J. Yao, Y. Wang and
G. Wang, Chin. J. Catal., 2009, 30, 853–855.
9 T. Murayama, T. Hayashi, R. Kanega and I. Yamanaka,
J. Phys. Chem. C, 2012, 116, 10607–10616.
10 G. Fan, Z. Wang, B. Zou and M. Wang, Fuel Process. Technol.,
2011, 92, 1052–1055.
11 H. Yang, Z. Xiao, Y. Qu, T. Chen, Y. Chen and G. Wang, Res.
Chem. Intermed., 2018, 44, 799–812.
12 K. Zhou, X. D. Xie and C. T. Chang, Appl. Surf. Sci., 2017,
416, 248–258.
13 X. Zhang, X. Ke, Z. Zheng, H. Liu and H. Zhu, Appl. Catal., B,
2014, 150–151, 330–337.
14 S. Y. Chen, T. Mochizuki, Y. Abe, M. Toba and
Y. Yoshimura, Appl. Catal., B, 2014, 148–149, 344–356.
15 X. Ma, J. Gong, S. Wang, F. He, H. Guo, X. Yang and G. Xu,
J. Mol. Catal. A: Chem., 2005, 237, 1–8.
16988 | New J. Chem., 2019, 43, 16981--16989 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
16 R. Tang, T. Chen, Y. Chen, Y. Zhang and G. Wang, Chin. 35 H. G. T. Nguyen, L. Mao, A. W. Peters, C. O. Audu, Z. J.
J. Catal., 2014, 35, 457–461.
Brown, O. K. Farha, J. T. Hupp and S. B. T. Nguyen, Catal.
Sci. Technol., 2015, 5, 44444451.
17 D. Chen, L. Cao, F. Huang, P. Imperia, Y. B. Cheng and
R. A. Caruso, J. Am. Chem. Soc., 2010, 132, 4438–4444.
18 R. Zhang, A. A. Elzatahry, S. S. Al-Deyab and D. Zhao, Nano
Today, 2012, 7, 344–366.
´
´
36 A. S. Portillo, H. G. Baldovı, M. T. G. Fernandez, S. Navalon,
P. Atienzar, B. Ferrer, M. Alvaro, H. Garcia and Z. Li, J. Phys.
Chem. C, 2017, 121, 7015–7024.
´
´
´
19 S. Wang, C. Li, Z. Xiao, T. Chen and G. Wang, J. Mol. Catal. 37 A. Santiago-Portillo, S. Navalon, M. Alvaro and H. Garcıa,
A: Chem., 2016, 420, 26–33. J. Catal., 2018, 365, 450–463.
20 J. Kansedo and K. T. Lee, Biomass Bioenergy, 2012, 40, 38 X. Min, X. Wu, P. Shao, Z. Ren, L. Ding and X. Luo, Chem.
96–104. Eng. J., 2019, 358, 321–330.
21 C. M. Granadeiro, S. O. Ribeiro, M. Karmaoui, R. Valenca, 39 J. H. Toney and T. J. Marks, J. Am. Chem. Soc., 1985, 107,
J. C. Ribeiro, B. de Castro, L. Cunha-Silva and S. S. Balula,
Chem. Commun., 2015, 51, 13818–13821.
947–953.
40 C. Shen and L. L. Shaw, J. Sol-Gel Sci. Technol., 2010, 53,
22 V. N. Panchenko, M. M. Matrosova, J. Jeon, J. W. Jun,
571–577.
´
´
M. N. Timofeeva and S. H. Jhung, J. Catal., 2014, 316, 41 I. O. Acik, J. Madarasz, M. Krunks, K. Tansuaadu, D. Janke,
¨
251–259.
G. Pokol and L. Niinisto, J. Therm. Anal. Calorim., 2007, 88,
23 A. Dhakshinamoorthy, A. Santiago-Portillo, A. M. Asiri and
H. Garcia, ChemCatChem, 2019, 11, 899–923.
557–563.
42 C. T. Chen and Y. S. Munot, J. Org. Chem., 2005, 70, 8625–8627.
24 M. Vandichel, J. Hajek, F. Vermoortele, M. Waroquier, 43 X. Zeng, L. Huang, C. Wang, J. Wang, J. Li and X. Luo, ACS
D. E. D. Vosb and V. V. Speybroeck, CrystEngComm, 2015,
17, 395–406.
Appl. Mater. Interfaces, 2016, 8, 20274–20282.
44 M. Dakanali, E. T. Kefalas, C. P. Raptopoulou, A. Terzis,
G. Voyiatzis, I. Kyrikou, T. Mavromoustakos and A. Salifoglou,
Inorg. Chem., 2003, 42, 4632–4639.
45 J. M. Collins, R. Uppal, C. D. Incarvito and A. M. Valentine,
Inorg. Chem., 2004, 44, 3431–3440.
46 Y. F. Deng, Z. H. Zhou and H. L. Wan, Inorg. Chem., 2004,
43, 6266–6273.
47 E. T. Kefalas, P. Panagiotidis, C. P. Raptopoulou, A. Terzis,
T. Mavromoustakos and A. Salifoglou, Inorg. Chem., 2005,
44, 2596–2605.
25 L. Liu, Z. Chen, J. Wang, D. Zhang, Y. Zhu, S. Ling,
K. W. Huang, Y. Belmabkhout, K. Adil, Y. Zhang, B. Slater,
M. Eddaoudi and Y. Han, Nat. Chem., 2019, 11, 622–628.
26 Z. Fang, B. Bueken, D. E. D. Vos and R. A. Fischer, Angew.
Chem., Int. Ed., 2015, 54, 7234–7254.
27 M. Kim, J. F. Cahill, H. Fei, K. A. Prather and S. M. Cohen,
J. Am. Chem. Soc., 2012, 134, 18082–18088.
28 H. Song, J. A. You, B. Li, C. Chen, J. Huang and J. Zhang,
Chem. Eng. J., 2017, 327, 406–417.
29 J. G. Santaclara, A. I. Olivos-Suarez, A. Gonzalez-Nelson, 48 G. Leofanti, M. Padovan, G. Tozzola and B. Venturelli, Catal.
D. Osadchii, M. A. Nasalevich, M. A. van der Veen, Today, 1998, 41, 207–219.
F. Kapteijn, A. M. Sheveleva, S. L. Veber, M. V. Fedin, 49 U. Schubert, J. Mater. Chem., 2005, 15, 3701–3715.
´
A. T. Murray, C. H. Hendon, A. Walsh and J. Gascon, Chem. 50 M. A. Arillo, M. L. Lopez, C. Pico, M. L. Veiga, A. Jimenez-
´
Mater., 2017, 29, 8963–8967.
Lopez and E. Rodıguez-Castellon, J. Alloys Compd., 2001,
30 H. Wu, Y. S. Chua, V. Krungleviciute, M. Tyagi, P. Chen,
317–318, 160–163.
T. Yildirim and W. Zhou, J. Am. Chem. Soc., 2013, 135, 51 C. Tao, X. Zou, K. Du, G. Zhou, H. Yan, X. Yuan and
10525–10532. L. Zhang, J. Alloys Compd., 2018, 747, 43–49.
31 F. Vermoortele, B. Bueken, G. L. Bars, B. Van de Voorde, 52 Z. Niu, Q. Guan, Y. Shi, Y. Chen, Q. Chen, Z. Kong, P. Ning,
M. Vandichel, K. Houthoofd, A. Vimont, M. Daturi, S. Tian and R. Miao, New J. Chem., 2018, 42, 19764–19770.
M. Waroquier, V. Van Speybroeck, C. Kirschhock and 53 Y. Han, M. Liu, K. Li, Q. Sun, W. Zhang, C. Song, G. Zhong,
D. E. D. Vos, J. Am. Chem. Soc., 2013, 135, 11465–11468. Z. C. Zhang and X. Guo, Inorg. Chem. Front., 2017, 4, 1870–1880.
32 L. G. Qiu, T. Xu, Z. Q. Li, W. Wang, Y. Wu, X. Jiang, 54 L. Wang, X. Yang, L. Liu and G. Wang, Petrochem. Technol.,
X. Y. Tian and L. D. Zhang, Angew. Chem., Int. Ed., 2008,
47, 9487–9491.
2012, 41, 770–777.
55 B. Jia, P. Cao, H. Zhang and G. Wang, J. Mater. Sci., 2019, 54,
33 H. Huang, J. R. Li, K. Wang, T. Han, M. Tong, L. Li, Y. Xie,
9466–9477.
Q. Yang, D. Liu and C. Zhong, Nat. Commun., 2015, 6, 8847. 56 C. Pang, J. Xiong, G. Li and C. Hu, J. Catal., 2018, 366, 37–49.
34 A. Wang, Y. Zhou, Z. Wang, M. Chen, L. Sun and X. Liu, RSC 57 X. Yin, Y. Zeng, J. Yao, Z. Du, H. Zhang, T. Sun and G. Wang,
Adv., 2016, 6, 3671–3679.
Chem. Eng. Sci., 2019, 199, 478–485.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16981--16989 | 16989