Coupling Ru nanoparticles and sulfonic acid moieties on single MIL-101 microcrystals for upgrading methyl levulinate into Γ-valerolactone

Article abstract of DOI:10.1016/j.apcata.2018.06.027

Heterogeneous catalytic conversion of biomass-derived methyl levulinate (ML) to γ-valerolactone (GVL) is highly attractive as a sustainable route to mitigating the dependence on nonrenewable fossil resources. However, this route is limited by its low selectivity toward GVL and severe reaction conditions. Here, we present a synergistic catalytic process that offers exclusive GVL selectivity with complete ML conversion under mild conditions (80 °C, 0.5 MPa of H₂ pressure). This is achieved over a dual-functionalized catalyst of Ru nanoparticles (Ru-NPs) and Br?nsted acid SO₃H groups in a Cr-based metal-organic framework (MIL-101). The Ru-NPs served as metal sites for the hydrogenation of ML to methyl-3-hydroxyvalerate (MHV) and the acidic SO₃H moieties promoted the cyclization of MHV to generate GVL via lactonization. In contrast, GVL production was significantly suppressed over the physically mixed congener (2.0 wt% Ru/MIL-101 + MIL-101-SO₃H). This result demonstrates an unrivalled advantage of the dual integration of metal/acid sites in a single MOF crystal, and the relative amount of sulfonic acid groups in the catalyst can exert significant influence on the overall catalytic performances. The activation energies for the hydrogenation and lactonization steps with the developed catalyst were 29.5 and 42.1 kJ/mol, respectively, which are much lower than those observed with the traditional 5.0 wt% Ru/AC catalyst.

Full text of DOI:10.1016/j.apcata.2018.06.027

Applied Catalysis A, General 563 (2018) 54–63
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Coupling Ru nanoparticles and sulfonic acid moieties on single MIL-101
microcrystals for upgrading methyl levulinate into γ-valerolactone
⁎
Zhenzhen Lin, Mengting Luo, Yindi Zhang, Xiaoxue Wu, Yanghe Fu, Fumin Zhang ,
⁎
Weidong Zhu
Key Laboratory of the Ministry of Education for Advanced Catalysis Materials, Institute of Physical Chemistry, Zhejiang Normal University, 321004 Jinhua, People’s
Republic of China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Heterogeneous catalytic conversion of biomass-derived methyl levulinate (ML) to γ-valerolactone (GVL) is
highly attractive as a sustainable route to mitigating the dependence on nonrenewable fossil resources. However,
this route is limited by its low selectivity toward GVL and severe reaction conditions. Here, we present a sy-
nergistic catalytic process that offers exclusive GVL selectivity with complete ML conversion under mild con-
Metal-organic framework
Dual-functionalized catalyst
Biomass conversion
Synergistic catalysis
γ-Valerolactone
Sulfonic acid-functionalized MIL-101
ditions (80 °C, 0.5 MPa of H
Ru-NPs) and Brønsted acid SO
as metal sites for the hydrogenation of ML to methyl-3-hydroxyvalerate (MHV) and the acidic SO
promoted the cyclization of MHV to generate GVL via lactonization. In contrast, GVL production was sig-
nificantly suppressed over the physically mixed congener (2.0 wt% Ru/MIL-101 + MIL-101-SO H). This result
2
pressure). This is achieved over a dual-functionalized catalyst of Ru nanoparticles
H groups in a Cr-based metal-organic framework (MIL-101). The Ru-NPs served
H moieties
(
3
3
3
demonstrates an unrivalled advantage of the dual integration of metal/acid sites in a single MOF crystal, and the
relative amount of sulfonic acid groups in the catalyst can exert significant influence on the overall catalytic
performances. The activation energies for the hydrogenation and lactonization steps with the developed catalyst
were 29.5 and 42.1 kJ/mol, respectively, which are much lower than those observed with the traditional 5.0 wt
%
Ru/AC catalyst.
1
. Introduction
Conventionally, GVL is obtained via catalytic hydrogenation and
subsequent intramolecular dehydration (or dealcoholization) of levu-
linic acid (LA) and its esters, where LA can be obtained in high yields
from the acid-catalyzed alcoholysis of biomass-based sugars, furfural,
and furfuryl alcohol [10–13]. In contrast to LA, alkyl levulinates (AL)
possess relatively lower boiling points and are acid-free and easily re-
coverable. Consequently, the catalytic transformation of ALs such as
methyl leuvlinate (ML) to GVL is an attractive option for biofuel pro-
duction [20]. Supported metal catalysts [21–34], including Ir, Au, Pt,
Pd, Ni, Co, and Ru, have been explored for the conversion of LA or ALs
to GVL. Among the various metal catalysts evaluated, Ru was found to
be the most active catalyst in liquid-phase catalytic reactions [35].
Moreover, by adding an acidic co-catalyst such as ZSM-5 to Ru/graphite
[36], the catalytic activity and selectivity can be greatly enhanced in
the conversion of ML to GVL under mild conditions. Yadav and cow-
orkers reported that the acidity of the catalyst plays an important role
in obtaining high conversion activity [37]. Bond and coworkers per-
formed a detailed kinetics study of the different steps in the formation
Biomass is a readily available renewable carbon source that is
economical to utilize and abundant [1–4]. The use of biomass as an
appropriate feedstock to produce various biochemicals and biofuels for
transportation not only contributes to alleviating global climate change
2
caused by increasing CO emissions, but also offers a sustainable route
for replacing dwindling fossil fuel reserves [5–9]. Among the numerous
compounds that can be obtained through the conversion of biomass, γ-
valerolactone (GVL) has been identified as a sustainable platform mo-
lecule for the production of renewable fuels and value-added chemicals
due to its versatile functions and benign properties [10–13]. For ex-
ample, GVL can be used directly as a fuel additive in gasoline to en-
hance the combustion process and as a green solvent that is especially
useful in biomass valorization [14–16]. Moreover, GVL can be con-
verted into many value-added chemicals (e.g., methyl pentenoate, ε-
caprolactam, and adipic acid) and bio-based liquid fuels, including 2-
methyltetrahydrofuran, valeric esters, and long-chain liquid alkenes
[
17–19].
of GVL, and concluded that by integrating acidic sites and a
⁎
Corresponding authors.
E-mail addresses: zhangfumin@zjnu.edu.cn (F. Zhang), weidongzhu@zjnu.cn (W. Zhu).
https://doi.org/10.1016/j.apcata.2018.06.027
Received 3 May 2018; Received in revised form 18 June 2018; Accepted 20 June 2018
Available online 28 June 2018
0926-860X/ © 2018 Elsevier B.V. All rights reserved.

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
hydrogenation metal, the GVL yield could be significantly improved at
low temperatures [38]. These studies indicate that designing dually
functionalized catalysts coupling hydrogenation functionality and
acidity is favorable for upgrading LA and its esters. The dual func-
tionality enhances the process efficiency (i.e., catalytic activity and
features in one crystal give rise to higher productivity than that of the
corresponding physically mixed catalyst (i.e., 2.0 wt% Ru/MIL-101 +
MIL-101-SO H), confirming that 2.0 wt% Ru/MIL-101-SO H improved
3 3
the overall catalytic performance in a bifunctional mode.
selectivity) and moderates the reaction temperature and H
25].
As a recently developed class of ordered crystalline material, metal-
2
pressure
2. Experimental Section
[
2.1. Chemicals
organic frameworks (MOFs) have potential applications in several fields
because of their large internal surface area, adjustable surface func-
tionality, and outstanding hydrothermal/chemical stability [39–42].
Chromium trioxide (CrO , Sinopharm Chemical Reagent Co., Ltd.,
3
≥99.0 wt%), chromium(III) nitrate nonahydrate (Cr(NO ) ·9H O,
3
3
2
The incorporation of sulfonic acid (eSO
3
H) groups into MOFs is pro-
Sinopharm Chemical Reagent Co., Ltd., ≥99.0 wt%), terephthalic acid
mising as the uniformly distributed and readily accessible Brønsted acid
sites present throughout the MOFs endow the resultant MOFs with
exceptionally high efficiency as solid acid catalysts [43–46]. In a pio-
neering study, Kitagawa and coworkers synthesized free sulfonic acid-
(H BDC, TCI, > 99%), 2-sulfonylterephthalic acid monosodium salt (2-
2
NaSO -H BDC, TCI, > 98%), hydrochloric acid (HCl, Sinopharm
3
2
Chemical Reagent Co., Ltd., 37 wt%), hydrofluoric acid (HF, Sinopharm
Chemical Reagent Co., Ltd., ≥40%), acetone (Sinopharm Chemical
Reagent Co., Ltd., ≥99.5%), polyvinyl pyrrolidone (PVP, Mw = 55000,
Xiya Reagent Co., Ltd., ≥99.0%), ruthenium(III) chloride (RuCl3,
Aladdin Industrial Inc., 43–44%), Ru/AC (Aladdin Industrial Inc.,
5.0 wt%), ethylene glycol (EG, Sinopharm Chemical Reagent Co., Ltd.,
≥99.0%), methanol (Sinopharm Chemical Reagent Co., Ltd., ≥99.5%),
ethanol (Sinopharm Chemical Reagent Co., Ltd., ≥99.7%), ML
(TCI, > 99%), GVL (Aladdin Industrial Inc., 99%), and H2 gas
(Shanghai Pujiang Special Gas Co., Ltd, 99.999%) were used in the
experiments. Deionized water with a resistance exceeding 18.2 MΩ was
obtained from a Millipore Milli-Q ultrapure water purification system.
functionalized chromium-based MIL-101 (MIL-101-SO
lecular formula Cr (H O) O[(O C)-C (SO H)-(CO
(SO -(CO )]·nH2O (n ≈ 38)), in which the SO H groups are
3
H with the mo-
3
2
3
2
6
H
3
3
2
)] [(O C)-
2
2
C
H
6 3
3
2
3
mainly located on the pore surface of MIL-101, using monosodium 2-
sulfoterephthalic acid as a functionalized organic linker [43]. The de-
veloped MIL-101-SO H exhibited high catalytic activity in the hydro-
3
lysis of cellulose to glucose and other sugar derivatives.
3
In a previous work, we reported a supported Ru/UiO-66-SO H
catalyst for the transformation of ML to GVL [34]. However, at least
three important scientific questions were not mentioned and need to be
further clarified. First of all, the Ru/UiO-66-SO
that work is prepared by a wet impregnation and reduction method
34]; thus, the spatial distribution of the resultant Ru nanoparticles (Ru-
3
H catalyst reported in
2.2. Catalyst preparation
[
NPs) would be difficult to control, which means that the Ru-NPs are
inevitably deposited on both the external surfaces and internal cavities
of the UiO-66-SO H support. Therefore, the relative location (or
3
proximity) between the Ru-NPs and sulfonic acid groups is complicated.
It is difficult to clarify the mechanism of synergistic catalysis taking
3
place between Ru and eSO H. Moreover, it is noticed that the Brønsted
3
Cr-based MIL-101 samples with various SO H group contents were
synthesized using the hydrothermal method according to the procedure
reported by Kitagawa and coworkers [43]. In a typical procedure, CrO
2.5 g), 2-NaSO -H BDC, and H BDC (total 25 mmol with different ra-
tios) were dissolved in a mixed solution containing concentrated aqu-
eous HCl (1.82 g) and H O (100 mL), and then transferred to a Teflon-
3
(
3
2
2
2
acid sites are indispensable for accelerating the intramolecular ester-
ification of the formed methyl-3-hydroxyvalerate (MHV) in the con-
version of ML into GVL. Thus, the effect of acid density on the overall
catalytic activities need be further analyzed. Finally, catalytic kinetics
and possible reaction mechanism studies are important aspects from the
viewpoint of catalysis engineering; however, few relevant studies have
been conducted on the catalytic conversion of ML to GVL.
Around the above-mentioned scientific questions, herein we de-
signed and prepared a series of dually functionalized catalysts based on
a MOF to demonstrate the concept of tandem conversion of ML to GVL
lined stainless-steel autoclave. This acidic solution was hydrothermally
reacted at 180 ℃ for 6 d without stirring. To remove the residual Na
ions within the MOFs, the crude microcrystalline powders were treated
by ion exchange in a mixed solution of diluted HCl (0.05 M) in me-
thanol, and were further rinsed in methanol (200 mL) and water
+
(
700 mL) to remove residual HCl. Finally, the solids were dried at
50 °C for 12 h prior to further use. The obtained solids are denoted as
MIL-101-S (x represents the mole fraction of the sulfonated ligand), in
which x was varied from 25 to 100 (mol%). For comparison, pure MIL-
01, with H BDC as a connecting organic linker, was prepared using the
hydrothermal method according to a previously reported procedure
42].
The Ru-NPs were fabricated using a reported polyol reduction
method with slight modifications [47]. The Ru precursor (RuCl ) and
1
x
1
2
(
Scheme 1). The metallic Ru Ru-NPs are well dispersed on the outer
surface of the MIL-101-SO H MOF microcrystals, where the frameworks
are decorated with SO H groups. The pore aperture of the MIL-
01 MOF is large enough to allow the reactants H and ML to contact
the encapsulated Ru-NPs and SO H groups to undergo hydrogenation to
3
[
3
1
2
3
3
PVP stabilizer in a stoichiometric ratio of Ru/PVP = 50:1 were added to
EG (10 mL) at 25 °C. The solution was then heated to 80 °C with mag-
netic stirring and maintained at this temperature for 30 min under an
Ar atmosphere to remove water and oxygen. Subsequently, the solution
was heated to 180 °C at a ramp rate of 10 °C/min and the temperature
was maintained for 120 min under an Ar flow. When the reaction was
complete, excess acetone was poured into the solution, and the re-
sulting black suspension was centrifuged. The precipitated Ru-NPs were
generate MHV, which is further converted into GVL through lactoni-
zation. In such a scenario, the sulfonic acid moieties are most effective
when positioned in the framework of the MOF microcrystals, where
they can promote intramolecular esterification of the intermediate
MHV. In this manner, the dual catalyst functionalities in the MOF
crystals can accelerate the desired cascade in the catalytic hydrogena-
tion-lactonization reaction. Notably, the combined tandem catalyst
Scheme 1. Schematic representation of
tandem conversion of ML to GVL over metal/
acid. dual-functionalized catalyst.
55

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
then separated, collected, and redispersed in ethanol.
To prepare the 2.0 wt% Ru/MIL-101-S
dried MIL-101-S (0.2 g) was dispersed in a Ru nanoparticle solution
0.6 mmol/L, 66 mL) at 35 °C under vigorous stirring for 24 h.
x
(or MIL-101) catalyst, pre-
x
(
Afterwards, the solid was isolated by centrifugation. The final sample
was obtained after drying under vacuum at 120 °C for 4 h.
2.3. Characterization
The structure and phase of the samples were examined by powder X-
ray diffraction (XRD) analyses using
a Philips PW3040/60 dif-
fractometer with Cu K radiation. The textural properties including
α
specific Brunauer–Emmett–Teller (BET) surface area and pore volume
were determined by nitrogen adsorption isotherms at −196 °C using a
Micromeritics ASAP 2020 instrument. Infrared (IR) spectra were col-
lected using a Nicolet NEXUS670 Fourier transform IR (FT-IR) spec-
trophotometer. The surface electronic states were investigated by X-ray
photoelectron spectroscopy (XPS) using a Thermo VG ESCALAB250
−
9
system at a pressure of approximately 10 Torr. The morphology and
energy dispersive X-ray (EDX) spectra were obtained by scanning
Fig. 1. XRD patterns of 2.0 wt% Ru/MIL-101 and 2.0 wt% Ru/MIL-101-S
x = 25, 50, 75, and 100).
x
(
electron microscopy (SEM) on
a
Hitachi S-4800 apparatus.
Transmission electron microscopy (TEM) images were obtained using a
JEOL JEM-1200 instrument at 200 kV. The actual Ru content in the
samples and the amount leached into the solution were determined
using an IRIS Intrepid II XSP inductively coupled plasma-atomic emis-
sion spectrometer (ICP-AES). The amount of sulfonic acid groups in the
characteristic peaks related to the Ru species were indistinguishable for
all the Ru/MIL-101-S samples (Figs. 1 and S2 in SI), which is likely
attributable to the high dispersion of Ru-NPs within the MOFs and/or
the small size of the Ru particles.
x
The specific surface areas and pore volumes of the MIL-101-S
samples as well as those of the corresponding catalysts derived from the
adsorption isotherms are listed in Fig. S3, Table S1, and Fig. 2, re-
spectively, for comparison. The pristine MIL-101 support had a large
x
various MIL-101-S
typical procedure, dried MIL-101-S
x
samples was determined by acid-base titration. In a
(1.0 g) was suspended in an aqu-
x
N
2
eous NaCl saturated solution (40 g). The resulting suspension was
stirred at room temperature for 36 h until equilibrium was reached. The
solution was then filtered, and the filtrate was titrated with 0.1 M
NaOH.
2
specific surface area (SBET: 3933 m /g) and total pore volume (Vtotal
1
:
3
.89 cm /g), which matched well with literature values [42]. A sub-
stantial decrease in SBET and Vtotal was observed when the amount of 2-
NaSO -H BDC linker used in the synthesis of MIL-101-S was increased.
For example, the MIL-101-S100 support containing 100 mol% of the 2-
3
2
x
2.4. Catalytic conversion of ML to GVL
2
3
NaSO
Table S1) and exhibited a type I isotherm (Fig. S2), indicating abun-
dant micropores in the MOF structure [43–46]. The reduced textural
parameters of all the MIL-101-S samples compared to those of the
pristine MIL-101 is possibly due to the pore-filling effect of the -SO
groups tethered to the MOF frameworks. As expected, after the in-
troduction of Ru-NPs to the surfaces of the MIL-101-S supports, the
specific surface areas and total pore volumes of the obtained Ru/MIL-
01 and Ru/MIL-101-S samples decreased (Table 1) due to pore
3 2
-H BDC linker had SBET of 1480 m /g with a Vtotal of 1.07 cm /g
The tandem reaction was performed in a 50-mL high-pressure au-
toclave equipped with a sampling valve and a magnetic stirrer.
Typically, 40 mg of the catalyst (0.5 mol% Ru), 0.206 g of ML
(
x
(
1.6 mmol), and 15 mL of H
clave, which was then sealed, purged, and pressurized with H
.5 MPa and heated to 80 °C. After initiating agitation at 980 rpm,
2
O as a solvent were placed in the auto-
3
H
2
to
0
x
steady pressure was maintained throughout the reaction. Aliquots of
the reaction mixture (0.3 mL) were removed at specific reaction times
and analyzed by gas chromatography (GC, GC-2014, Shimadzu)
equipped with an FID detector and a DB-5 capillary column. The con-
version of ML and selectivity toward GVL were quantified by using bis
1
x
blockage by the highly dispersed Ru-NPs that migrated into the cavities
of the supports.
x
The FT-IR spectra of the series of MIL-101-S samples and those of
(
2-methoxyethyl)ether as an internal standard, and the mass balance
errors were within ± 2% for all the samples based on the repeated tests.
Besides MHV and GVL, equal mole of methanol to GVL was produced as
the only by-product. To examine catalyst reusability, the used catalyst
was recovered by centrifugation, washed with ethanol, dried under
vacuum at 120 °C for 5 h, and then subjected to the next reaction cycle.
3. Results and Discussion
3.1. Catalyst characterization
The structures of the prepared MOFs and the corresponding cata-
lysts were examined by XRD. As shown in Fig. S1 in the Supporting
Information (SI), the XRD patterns of the MIL-101-S samples with
x = 25, 50, 75, and 100 coincided well with that of the pristine MIL-
01 [42–46], and no impurity phases associated with Cr species were
observed. After introducing the Ru-NPs onto the surface of the MOFs,
no significant loss of crystallinity was observed for the Ru/MIL-101-S
x
1
x
Fig. 2. N
2
adsorption isotherms of 2.0 wt% Ru/MIL-101 and 2.0 wt% Ru/MIL-
samples compared to Ru/MIL-101 (Fig. 1), suggesting preservation of
the MIL-101 structure after Ru introduction [48]. Additionally,
1
01-S (x = 50, 75, and 100).
x
56

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
Table 1
a
Comparison of physicochemical properties and catalytic performances of various investigated catalysts .
b
c
Acidityd
mmol/g
Acid densitye
μmol/m
Ru loadingf
wt%
Con.g
%
Sel.h
%
Entry
Catalyst
S
BET
2
V
total
3
2
m /g
cm /g
1
2
3
4
5
6
7
5.0 wt% Ru/AC
855
0.44
1.38
1.21
1.10
1.09
1.06
–
–
–
–
–
5.07
2.05
2.03
2.08
2.06
1.99
–
46.6
38.5
94.9
95.7
96.8
98.7
63.2
11.2
45.7
37.3
44.5
57.8
86.6
70.5
2.0 wt% Ru/MIL-101
2.0 wt% Ru/MIL-101-S25
2.0 wt% Ru/MIL-101-S50
2.0 wt% Ru/MIL-101-S75
2.0 wt% Ru/MIL-101-S100
2.0 wt% Ru/MIL-101 +
MIL-101-S100i
2571
2045
1589
1514
1454
–
0.046
0.81
1.03
1.65
–
0.022
0.51
0.68
1.13
–
a
Reaction conditions: ML (0.206 g), catalyst (40 mg), H
By BET method.
2 2
O (15 mL), 180 min, 0.5 MPa H , 80 °C.
b
c
d
e
f
Reported at P/P
0
= 0.99.
By acid-base titration method.
Calculated by the equation: [Acid density (μmol/m )] = [Acidity (mmol/g)]/[SBET (m /g)] × 1000.
2
2
By ICP-AES analysis.
Conversion of ML.
Selectivity toward GVL.
g
h
i
2.0 wt% Ru/MIL-101 (40 mg), MIL-101-S100 (40 mg).
x
content of Brønsted acid sites in the MIL-101-S samples were measured
as 0.046, 0.81, 1.03, and 1.65 mmol/g, respectively, and the density of
acidic sites monotonically increased as the content of SO H groups in
3
the samples increased (Table 1). Although the coordination of un-
saturated Cr3+ sites within MIL-101 may serve as acidic sites [42], their
strength is exceedingly weaker than that of the eSO
measured acidity can be mainly attributed to the Brønsted acidic
eSO H sites tethered to the MIL-101 frameworks
Figs. 4A and S5A–S7A show the XPS survey data, which demon-
strates that the 2.0 wt% Ru/MIL-101-S contains five elements (Cr, O, C,
3
H sites. Thus, the
3
x
S, and Ru). Interestingly, the S 2p spectrum of the catalysts shows two
different peaks (Figs. 4C, S5C, S6C, and S7C), which are associated with
the SeO (168.7 eV) and S]O bonds (169.9 eV) with a peak separation
of 1.2 eV. These bonds indicate that the S in 2.0 wt% Ru/MIL-101-S
present mainly in the form of SO H sites within the MOF framework
44,45], in accordance with the FT-IR data. Moreover, the sample ex-
x
is
3
[
hibits a Ru 3p band at ca. 461.6 eV, which is the characteristic of zero-
valent Ru species (Figs. 4D, S5D, S6D, and S7D) [25]. The TEM images
demonstrated that the Ru-NPs were highly distributed on the surface of
the MIL-101-S100 support with a uniform size of 2–4 nm, as shown in
Figs. 5 and S8. Further, EDX analyses also confirmed that the Ru-NPs
were evenly dispersed within MIL-101-S100, as shown in Fig. 6.
Fig. 3. FT-IR spectra of 2.0 wt% Ru/MIL-101 and 2.0 wt% Ru/MIL-101-S
x = 25, 50, 75, and 100).
x
(
the corresponding 2.0 wt% Ru/MIL-101-S
x
catalysts are shown in
−1
3.2. Catalytic conversion of ML to GVL
Figs. 3 and S4. The absorption peaks centered at 1180 and 1253 cm
can be attributed to the symmetric and asymmetric stretching modes of
−
1
The use of water as an environmentally friendly green solvent of-
fered high activity for the hydrogenation of ML to GVL over the Ru
catalysts. The specific role of water in promoting hydrogenation of the
C]O group of LA accounts for this excellent performance [26,35].
Therefore, after full characterization, the prepared catalysts were tested
in an aqueous solution for the hydrogenation of ML in an autoclave at a
the O]S]O bonds, the peak at 624 cm
can be ascribed to the CeS
−1
stretching vibration, and the peak at 1080 cm
SeO stretching vibration [43–46]. As expected, an increase in the in-
tensity of these peaks was observed upon increasing the fraction of the
corresponds to the
2
-NaSO
with variation in the amount of sulfonated ligand in the MIL-101-S
samples, suggesting that the sulfonic acid groups were present mainly
3 2 x
-H BDC linker in MIL-101-S . However, no shift was observed
x
2
H pressure of 0.5 MPa and temperature of 80 °C. As shown in Table 1,
the commercially available 5.0 wt% Ru/AC catalyst exhibited only
limited catalytic activity (46.6%) and GVL selectivity (11.2%) under the
employed conditions (Table 1, entry 1). In contrast, in the presence of
as free eSO
3
H sites in MIL-101-S
x
[43]. Notably, the FT-IR peaks cor-
−1
responding to eSO
3
H groups were preserved in the 600–1400 cm
range for all the Ru/MIL-101-S
x
samples (Fig. 3).
2.0 wt% Ru/MIL-101, the selectivity for GVL was greatly enhanced at
The S/Cr atomic ratios of the various MIL-101-S
x
samples are
the expense of a decrease in the ML conversion. The Ru particle sizes of
the 2.0 wt% Ru/MIL-101 and 5.0 wt% Ru/AC were very similar, as
observed by TEM (Figs. S9–S11 in SI). Thus, the lower catalytic activity
of 2.0 wt% Ru/MIL-101 compared to that of 5.0 wt% Ru/AC is possibly
due to the lower Ru loading in MIL-101 than that in the activated
carbon (see Table S1 for detailed characterization of the control cata-
lysts). The significantly higher selectivity toward GVL for the reaction
performed over 2.0 wt% Ru/MIL-101 than that over 5.0 wt% Ru/AC is
summarized in Table S1. The calculated S/Cr ratios were slightly lower
than that of the starting solution, which is possibly attributable to the
2 3 2
difference in the bonding strength of H BDC and 2-NaSO -H BDC with
3
+
the Cr site. In addition, ICP-AES analysis verified that no sodium ions
were present in the synthesized MOF samples, demonstrating that the
+
Na ions from the 2-NaSO
-H
3 2
BDC ligand were fully exchanged with
hydrogen protons from HCl during sample preparation [44–46]. The
57

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
Fig. 4. XPS spectra of 2.0 wt% Ru/MIL-101-S100: survey spectrum (A), high-resolution Cr spectrum (B), high-resolution S spectrum (C), and high-resolution Ru
spectrum (D).
possibly due to the appropriate acidity of MIL-101, derived from the
coordinatively unsaturated Cr sites [42]. These sites are beneficial for
intramolecular esterification of the MHV intermediate. Importantly,
3
SO H was evaluated, which exhibited a considerably low catalytic ac-
tivity (entry 7). This suggests that the close positioning of the metal/
acid active species in the support is important for the enhanced activity
of the tandem catalyst.
3
with a gradual increase in the amount of SO H groups in the MIL-101
frameworks, the conversion of ML and selectivity toward GVL improved
remarkably (entries 3–6 in Table 1). For example, an 86.6% GVL se-
lectivity with nearly complete ML conversion was achieved over 2.0 wt
We hypothesized that: 1) the Brønsted acid sites stemming from the
SO H groups can dramatically accelerate the intramolecular ester-
3
ification of MHV to generate GVL; and 2) the acids sites should be
adjacent to the metal sites for synergetic catalysis. Furthermore, in
addition to activating the lactonization step, the Brønsted acid sites
were also likely to promote ML hydrogenation, confirming that 2.0 wt%
Ru/MIL-101-S100 improved the upgradation of ML to GVL activity in a
bifunctional way. Impressively, the yield of GVL obtained by the hy-
drogenation of ML and the subsequent intramolecular esterification
positively correlated with the Brønsted acidity sites in MIL-101. Thus,
very good catalysis results could be achieved over the developed 2.0 wt
%
Ru/MIL-101-S100, which has the highest acid density (entry 6 in
Table 1) and similar Ru-NP dispersion compared with 2.0 wt% Ru/MIL-
01 and 5.0 wt% Ru/AC (Figs. 5 and S8–S11 in SI). These results de-
1
monstrate that conversion of ML to GVL is a tandem reaction, requiring
both metal hydrogenation and acid sites. The transformation of the
intermediate MHV to generate the final GVL product is probably the
dominant reaction step in this tandem reaction [38]. For comparison,
the activity of a physical mixture of 2.0 wt% Ru/MIL-101 and MIL-101-
Fig. 5. TEM images of the prepared 2.0 wt% Ru/MIL-101-S100 catalyst.
58

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
Fig. 6. SEM image and EDX elemental maps of 2.0 wt% Ru/MIL-101-S100 catalyst.
%
Ru/MIL-101-S100. Recently, Medlin and coworkers reported that or-
accompanied by a drastic decrease in ML concentration to 37.5%. As
the reaction progressed, the concentrations of ML and the formed MHV
diminished gradually, whereas the concentration of the target GVL
product accumulated incrementally. After 360 min of reaction, 100%
yield of GVL was achieved, and no further by-product was detected
when the reaction time was increased to 480 min. This implies that
MHV is an intermediate in the conversion of ML to GVL under the
current reaction conditions, in accordance with previous reports [33].
In contrast, rather low activity and selectivity were observed over
2.0 wt% Ru/MIL-101, where only 37.7% ML was converted and the
selectivity for GVL was 53.6% after 600 min (Fig. 7B). The strong
2 3
ganophosphonic acid modification of Pd/Al O significantly enhanced
the catalytic activity and selectivity in the liquid-phase hydrogenation-
dehydration of vanillin, in which vanillyl alcohol was formed as an
intermediate. They attributed the performance improvement to the
creation of metal/acid bifunctional sites upon organophosphonic acid
modification [50].
To further clarify the influence of catalyst acidity on its activity and
selectivity, the conversion of ML as a function of reaction time over
2.0 wt% Ru/MIL-101-S100 and 2.0 wt% Ru/MIL-101 was compared. As
shown in Fig. 7A, over 2.0 wt% Ru/MIL-101-S100, the maximum con-
centration of MHV reached 38.1% at an initial reaction time of 30 min,
3
acidity of the SO H groups in 2.0 wt% Ru/MIL-101-S100 promoted the
Fig. 7. Conversion of ML to GVL as a function of reaction time over 2.0 wt% Ru/MIL-101-S100 (A) and 2.0 wt% Ru/MIL-101 (B). Reaction conditions: ML (0.206 g),
catalyst (40 mg), H O (15 mL), 0.5 MPa H , 80 °C.
2
2
59

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
2 2
Fig. 8. ML conversion and GVL selectivity at different reaction temperatures (A) and H pressures (B). Reaction conditions: ML (0.206 g), catalyst (40 mg), H O
2
(15 mL), 0.5 MPa H pressure for 120 min (A), and at 80 °C for 120 min (B).
intramolecular esterification of MHV, thereby affording a higher yield
of GVL within a short reaction time [33,34].
Fig. 8A demonstrates that the conversion of ML to GVL was affected
positively by the reaction temperature, and MHV was the major inter-
mediate according to the reported reaction route, as presented in Sec-
tion 1 [33]. The first step of rapid hydrogenation of ML to MHV was not
significantly affected by temperature (from 50 to 100 °C). However, as
expected, at lower temperatures, the subsequent lactonization step for
formation of GVL from MHV was slow. As the temperature increased to
100 °C, the reaction rate increased significantly, with a major en-
hancement in the selectivity for GVL, indicating that the hydrogenation
of ML to generate MHV was much easier than the intramolecular es-
terification of MHV to GVL at a low temperature. The effect of H
pressure on the hydrogenation of ML was also investigated at the op-
timized temperature of 80 °C (Fig. 8B). The details of the influence of H
2
2
pressure on the reactant and product distributions as a function of re-
action time over 2.0 wt% Ru/MIL-101-S100 are displayed in Fig. S12.
Obviously, the concentration of hydrogen in the reaction solution in-
Fig. 9. Recycling test of 2.0 wt% Ru/MIL-101-S100 in the cascade conversion of
ML to GVL. Reaction conditions: ML (0.206 g), catalyst (40 mg), H O (15 mL),
0 °C, 0.5 MPa H pressure, 180 min.
2
8
2
2
creased with an increase in H pressure in the reactor; therefore, the
hydrogenation of ML to MHV remarkably enhanced. Moreover, the
lactonization of MHV to GVL is positive to the concentration of MHV
and can be accelerated by strong Brønsted acid sites in the catalyst [38];
thus, the overall catalytic activity and selectivity of ML to GVL was
kinetic analysis, internal and external transport limitations were
eliminated following the procedure reported previously (see SI) [49].
Subsequently, the dependency of the initial reaction rate on the initial
significantly improved with an increase in H
2
pressure.
concentration of ML at 80 °C and 2.0 MPa H pressure was examined, as
2
The reusability of the 2.0 wt% Ru/MIL-101-S100 catalyst for the
conversion of ML was evaluated under the optimized conditions. The
lost catalyst was replaced with fresh catalyst (less than 5 wt%, due to
filtration) in the subsequent run. Fig. 9 shows that 2.0 wt% Ru/MIL-
shown in Fig. 10. The reaction rate increased linearly within the se-
lected lower initial ML concentration range (0.1–0.3 M), which sug-
gested that the catalytic reaction followed an apparent first-order de-
pendence on ML [27]. However, the reaction rate was almost constant
at relatively higher ML initial concentrations, implying that the ad-
sorption of ML on the catalyst surface achieved saturation, resulting in
an apparent zero-order dependence on ML [27]. Additionally, because
the H2 pressure remained constant at 2.0 MPa over the entire hydro-
101-S100 was reusable and stable for up to 5 cycles without significant
activity loss. The ICP-AES analysis of the reaction liquor revealed that
leaching of the Ru species was negligible. The recovered catalyst was
also characterized by different techniques (including XRD, N adsorp-
2
tion, FT-IR, XPS, and EDX) and the results were compared with those of
the fresh catalyst (Figs. S13–S17). The reused catalyst showed no sig-
nificant changes in crystalline structure, composition, or textural
properties, which confirmed that the 2.0 wt% Ru/MIL-101-S100 catalyst
was stable.
genation process, the influence of H concentration on the conversion of
ML would be negligible. Thus, by further simplifying the kinetic model
developed by Palkovits and coworkers for the catalytic conversion of
ALs to GVL over 5.0 wt% Ru/AC under similar conditions [27], the
following differential equations were used to describe the evolution of
ML, MHV, and GVL concentrations in the current reaction system as a
2
function of time over 2.0 wt% Ru/MIL-101-S100
:
3.3. Kinetics and possible reaction mechanism
dCAL
dt
=
1 AL
k C
The aqueous-phase catalytic conversion of ML to GVL over a solid
catalyst is performed in a three-phase system, where the mass transfer
may impose limitations on the overall reaction rates. Therefore, before
(1)
60

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
Table 2
Values of reaction rate constants and activation energies for the catalytic con-
version of ML to GVL over 2.0 wt% Ru/MIL-101-S100
.
Temp. (°C)
40
60
80
Ea (KJ/mol)
R2
−
2
k
k
1
(10 /min)
4.3
0.53
8.3
1.2
16.0
3.4
29.5
42.1
0.99
0.98
−2
2
(10 /min)
hydrogenation of ML to GVL over Ru/AC (41 kJ/mol for hydrogenation
and 50 kJ/mol for lactonization) [27].
Based on the above results, a possible reaction mechanism for the
formation of GVL from ML over the 2.0 wt% Ru/MIL-101-S100 catalyst
was proposed, which is illustrated in Scheme 2. Briefly, following the
division of the H diatom, hydrogen is adsorbed on the surface of Ru and
forms a bond with Ru, and ML is subsequently adsorbed on the acid
Fig. 10. Initial reaction rate of ML as a function of initial ML concentration.
3
sites through combination of Ru with the SO H site [37]. When the first
Reaction conditions: 2.0 wt% Ru/MIL-101-S100 (40 mg), H
.0 MPa H pressure, 15 min.
2
O (15 mL), 80 °C,
hydrogen atom is added to ML, an intermediate is generated, linking to
the surface of Ru by a bond formed between sulfur and Ru. Addition of
another H atom results in the formation of the adsorbed MHV species.
Finally, the produced MHV forms GVL via lactonization [38], catalyzed
by the sulfonic acid sites.
2
2
dCMHV
dt
=
k
1
C
AL − k
2 MHV
C
(2)
(3)
dCGVL
dt
=
k
2
C
AHV
4. Conclusions
where C
i
represents the concentration of the component and k
i
is the
In summary, we demonstrated that the coupling of two sites (metal/
acid) in a single MOF microcrystal could provide a highly efficient
heterogeneous catalyst for the tandem hydrogenation-lactonization of
biomass-derived ML to produce GVL using water as an environmentally
benign solvent. This significant promotion was derived from acid/metal
nanoparticle synergistic catalysis that required both sites integrating
within one support. Quantitative yield of GVL was obtained from the
reaction over the developed 2.0 wt% Ru/MIL-101-S100 catalyst under
reaction rate coefficient. Using Athena Visual Studio, the numerical
solutions of the differential equations at different temperatures were
calculated. The curve fittings for the experimental data for the con-
version of ML to GVL at 40 and 80 °C are shown in Fig. 11, and the
derived apparent reaction rate constants are listed in Table 2. The rate
constant of the lactonization step is smaller than that of the hydro-
genation step, indicating that lactonization is the rate-determining step
and controls the overall reaction rate. Subsequently, Arrhenius plots
were constructed using the calculated first-order rate constants
2
mild reaction conditions (80 °C, 0.5 MPa of H pressure for 360 min).
The developed catalyst significantly outperformed other catalysts with
low sulfonic acid loadings as well as the previously developed 5.0 wt%
Ru/AC. This can be attributed to the cooperative effect between the Ru
(
Fig. 12). The estimated activation energy for hydrogenation was
9.5 kJ/mol and that for lactonization was 42.1 kJ/mol (Table 2),
which are much lower than the previously reported values for aqueous
2
3
NPs and Brønsted acid eSO H groups uniformly distributed in the
Fig. 11. Comparison of the outcomes of model fitting with the experimental data. Reaction conditions: ML (0.206 g), 2.0 wt% Ru/MIL-101-S100 (40 mg), H
15 mL), 2 MPa H , at (A) 40 °C and (B) 80 °C.
2
O
(
2
61

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
Fig. 12. Arrhenius plots for the hydrogenation step (A) and the lactonization step (B).
Scheme 2. Possible reaction mechanism of hydrogenation-lactonization of ML to GVL over 2.0 wt% Ru/MIL-101-S100
.
confined nanospace of MIL-101, where the composite can catalyze the
hydrogenation of ML (to produce the intermediate MHV) and the suc-
cessive intramolecular dealcoholization (to form GVL). Additionally,
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (21576243), Zhejiang Provincial Natural Science
Foundation of China (LY18B060006 and LY18B030006), and New
Talent Project for Undergraduates in Zhejiang Province
(2017R404067).
3
the developed Ru/MIL-101-SO H catalyst was reusable for at least 5
successive cycles under the specified reaction conditions without sig-
nificant loss of catalytic activity or selectivity. This research highlights
new perspectives for MOF-based materials as dual-functional hetero-
geneous catalysts for efficient transformation of biomass into useful
chemicals and fuels.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
62

Z. Lin et al.
Applied Catalysis A, General 563 (2018) 54–63
online version, at doi:https://doi.org/10.1016/j.apcata.2018.06.027.
[25] Y. Wang, Z. Rong, Y. Wang, T. Wang, Q. Du, Y. Wang, J. Qu, ACS Sustain. Chem.
Eng. 5 (2016) 1538–1548.
[
[
26] J. Tan, J. Cui, T. Deng, X. Cui, G. Ding, Y. Zhu, Y. Li, ChemCatChem 7 (2015)
08–512.
27] L. Negahdar, M.G. Al-Shaal, F.J. Holzhäuser, R. Palkovits, Chem. Eng. Sci. 158
2017) 545–551.
References
5
(
[1] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827–1870.
[2] X. Zhang, K. Wilson, A.F. Lee, Chem. Rev. 116 (2016) 12328–12368.
[3] A.N. Kay Lup, F. Abnisa, W.M.A.W. Daud, M.K. Aroua, Appl. Catal. Gen. 541 (2017)
[
[
28] Z. Yang, Y.B. Huang, Q.X. Guo, Y. Fu, Chem. Commun. 49 (2013) 5328–5330.
29] H. Zhou, J. Song, X. Kang, J. Hu, Y. Yang, H. Fan, Q. Meng, B. Han, RSC Adv. 5
(
2015) 15267–15273.
30] J. Yuan, S.S. Li, L. Yu, Y.M. Liu, Y. Cao, H.Y. He, K.N. Fan, Energy Environ. Sci. 6
2013) 3308–3313.
31] C. Xiao, T.W. Goh, Z. Qi, S. Goes, K. Brashler, C. Perez, W. Huang, ACS Catal. 6
2016) 593–599.
8
7–106.
4] W. Zhang, T. Meng, J. Tang, W. Zhuang, Y. Zhou, J. Wang, ACS Sustain. Chem. Eng.
(2017) 10029–10037.
[
[
[
(
5
[5] K. Dong, J. Zhang, W. Luo, L. Su, Z. Huang, Chem. Eng. J. 334 (2018) 1055–1064.
[6] M.J. Climent, A. Corma, S. Iborra, Green Chem. 16 (2014) 516–547.
[7] K. Yan, Y. Yang, J. Chai, Y. Lu, Appl. Catal. B: Environ. 179 (2015) 292–304.
[8] Y.S. Raupp, C. Yildiz, W. Kleist, M.A.R. Meier, Appl. Catal. Gen. 546 (2017) 1–6.
[9] R.S. Malkar, G.D. Yadav, Appl. Catal. Gen. 560 (2018) 54–65.
(
[
[
[
[
[
32] A.M. Hengne, N.S. Biradar, C.V. Rode, Catal. Lett. 142 (2012) 779–787.
33] Y. Kuwahara, Y. Magatani, H. Yamashita, Catal. Today 258 (2015) 262–269.
34] Z. Lin, X. Cai, Y. Fu, W. Zhu, F. Zhang, RSC Adv. 7 (2017) 44082–44088.
35] C. Michel, P. Gallezot, ACS Catal. 5 (2015) 4130–4132.
[10] F. Liguori, C. Moreno-Marrodan, P. Barbaro, ACS Catal. 5 (2015) 1882–1894.
[11] W.R. Wright, R. Palkovits, ChemSusChem 5 (2012) 1657–1667.
[12] K. Hengst, M. Schubert, H.W.P. Carvalho, C. Lu, W. Kleist, J.D. Grunwaldt, Appl.
Catal. A: Gen. 502 (2015) 18–26.
36] J.M. Nadgeri, N. Hiyoshi, A. Yamaguchi, O. Sato, M. Shirai, Appl. Catal. A: Gen. 470
(2014) 215–220.
[37] J. Molleti, M.S. Tiwari, G.D. Yadav, Chem. Eng. J. 334 (2018) 2488–2499.
[38] A. Abdelrahman, J.Q. Heyden, Bond, ACS Catal. 4 (2014) 1171–1181.
[39] N.L. Rosi, J. Eckert, M. Eddaoudi, D.T. Vodak, J. Kim, M. O’keeffe, O.M. Yaghi,
Science 300 (2003) 1127–1129.
[13] F.D. Pileidis, M.M. Titirici, ChemSusChem 9 (2016) 562–582.
[14] I.T. Horváth, H. Mehdi, V. Fábos, L. Boda, L.T. Mika, Green Chem. 10 (2008)
2
38–242.
[
[
[
40] D. Jagadeesan, Appl. Catal. Gen. 511 (2016) 59–77.
[15] J.S. Luterbacher, J.M. Rand, D.M. Alonso, J. Han, J.T. Youngquist, C.T. Maravelias,
41] Y.B. Huang, J. Liang, X.S. Wang, R. Cao, Chem. Soc. Rev. 46 (2017) 126–157.
42] G. Férey, C. Mellot-Draznieks, C. Serre, F. Millange, J. Dutour, S. Surblé,
I. Margiolaki, Science 309 (2005) 2040–2042.
B.F. Pfleger, J.A. Dumesic, Science 343 (2014) 277–280.
[16] X. Tang, X. Zeng, Z. Li, L. Hu, Y. Sun, S. Liu, T. Lei, L. Lin, Renew. Sustain. Energy
Rev. 40 (2014) 608–620.
[
[
[
43] G. Akiyama, R. Matsuda, H. Sato, M. Takata, S. Kitagawa, Adv. Mater. 23 (2011)
[17] D.M. Alonso, S.G. Wettstein, J.A. Dumesic, Green Chem. 15 (2013) 584–595.
3
294–3297.
44] Y. Zang, J. Shi, F. Zhang, Y. Zhong, W. Zhu, Catal. Sci. Technol. 3 (2013)
044–2049.
45] Y.X. Zhou, Y.Z. Chen, Y. Hu, G. Huang, S.H. Yu, H.L. Jiang, Chem. Eur. J. 20 (2014)
4976–14980.
[18] A.M.R. Galletti, C. Antonetti, V. De Luise, M. Martinelli, Green Chem. 14 (2012)
6
88–694.
19] J.Q. Bond, D.M. Alonso, D. Wang, R.M. West, J.A. Dumesic, Science 327 (2010)
110–1114.
2
[
[
1
1
20] J.P. Lange, R. Price, P.M. Ayoub, J. Louis, L. Petrus, L. Clarke, H. Gosselink, Angew.
[
[
46] Z. Hasan, J.W. Jun, S.H. Jhung, Chem. Eng. J. 278 (2015) 265–271.
47] L. Wang, J. Chen, L. Ge, V. Rudolph, Z. Zhu, J. Phys. Chem. C 117 (2013)
Chem. Int. Ed. 49 (2010) 4581–4585.
[21] X. Du, Y. Liu, J. Wang, Y. Cao, K. Fan, Chin. J. Catal. 34 (2013) 993–1001.
[22] X.L. Du, Q.Y. Bi, Y.M. Liu, Y. Cao, K.N. Fan, ChemSusChem 4 (2011) 1838–1843.
[23] A.M. Ruppert, J. Grams, M. Jędrzejczyk, J. Matrasmichalska, N. Keller, K. Ostojska,
P. Sautet, ChemSusChem 8 (2015) 1538–1547.
4
141–4151.
[48] R. Fang, H. Liu, R. Luque, Y. Li, Green Chem. 17 (2015) 4183–4188.
[49] X. Zhao, Y. Jin, F. Zhang, Y. Zhong, W. Zhu, Chem. Eng. J. 239 (2014) 33–41.
[50] P. Hao, D.K. Schwartz, J.W. Medlin, Appl. Catal. A: Gen. 561 (2018) 1–6.
[24] F. Ye, D. Zhang, T. Xue, Y. Wang, Y. Guan, Green Chem. 16 (2014) 3951–3957.
63