A novel Ni/AC catalyst prepared by MOCVD method for hydrogenation of ethyl levulinate to γ-valerolactone

Article abstract of DOI:10.1016/j.mcat.2020.111155

GVL (γ-valerolactone) is identified as an important biomass platform molecule due to its wide application. In this work, a series of novel supported Ni catalysts with different supports and Ni loading were synthesized via metal-organic chemical vapor deposition (MOCVD) method for the hydrogenation of EL (ethyl levulinate) to GVL. Fourier transform infrared spectroscopy (FT-IR), X-ray powder diffraction (XRD), nitrogen adsorption/desorption, inductively coupled plasma optical emission spectroscopy (ICP-OES) and transmission electron microscopy (TEM) were used to characterize the as-synthesized catalysts. The results showed that the 2 wt.% Ni/AC(MOCVD) presented superior catalytic activity when compared with the catalyst prepared by impregnation method. This behavior is explained in terms of the smaller Ni nanoparticles (4.28 nm) and higher dispersion on 2 wt.% Ni/AC(MOCVD). Among the catalysts, the 2 wt.% Ni/AC catalyst exhibited the best catalytic performance with 99.7 % EL conversion and 79.8 % GVL yield under 1 MPa initial H₂ pressure (measured at room temperature) at 250 °C for 2 h. In addition, the reaction conditions were optimized and the stability of the catalyst were also investigated. The insights gained from this study in the design of high dispersed Ni particles with smaller particle size via MOCVD method will facilitate the metal-catalyzed hydrogenation of EL to GVL.

Full text of DOI:10.1016/j.mcat.2020.111155

Molecular Catalysis 495 (2020) 111155
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
A novel Ni/AC catalyst prepared by MOCVD method for hydrogenation of
ethyl levulinate to γ-valerolactone
a
a
a
a
a
a
a
Na Ji , Zhenyu Liu , Xinyong Diao , Jinrong Bao , Zhihao Yu , Chunfeng Song , Qingling Liu ,
a
a,b,
Degang Ma , Xuebin Lu
*
a
School of Environmental Science and Engineering, Tianjin Key Laboratory of Biomass/Wastes Utilization, Tianjin University, 300350, Tianjin, China
Department of Chemistry & Environmental Science, School of Science, Tibet University, Lhasa, 850000, China
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
GVL (γ-valerolactone) is identified as an important biomass platform molecule due to its wide application. In this
work, a series of novel supported Ni catalysts with different supports and Ni loading were synthesized via metal-
organic chemical vapor deposition (MOCVD) method for the hydrogenation of EL (ethyl levulinate) to GVL.
Fourier transform infrared spectroscopy (FT-IR), X-ray powder diffraction (XRD), nitrogen adsorption/deso-
rption, inductively coupled plasma optical emission spectroscopy (ICP-OES) and transmission electron micro-
scopy (TEM) were used to characterize the as-synthesized catalysts. The results showed that the 2 wt.% Ni/
AC(MOCVD) presented superior catalytic activity when compared with the catalyst prepared by impregnation
method. This behavior is explained in terms of the smaller Ni nanoparticles (4.28 nm) and higher dispersion on 2
wt.% Ni/AC(MOCVD). Among the catalysts, the 2 wt.% Ni/AC catalyst exhibited the best catalytic performance
Metal-organic chemical vapor deposition
Ni/AC catalyst
Hydrogenation
Ethyl levulinate
γ-valerolactone
2
with 99.7 % EL conversion and 79.8 % GVL yield under 1 MPa initial H pressure (measured at room tem-
perature) at 250 °C for 2 h. In addition, the reaction conditions were optimized and the stability of the catalyst
were also investigated. The insights gained from this study in the design of high dispersed Ni particles with
smaller particle size via MOCVD method will facilitate the metal-catalyzed hydrogenation of EL to GVL.
1. Introduction
through ring-opening hydrogenation reaction [7]. Moreover, GVL could
also be converted to epsilon-caprolactam, which is the raw material to
With the rapid depletion of fossil fuels and its subsequent environ-
produce nylon polymers [14]. Therefore, GVL has been known as one of
the most important biomass platform molecules [10].
mental deterioration, searching for clean and renewable energy re-
sources has become particularly important. The research focused on
biomass-derived platform molecules which have attracted considerable
attention in recent years because of its unique physical and chemical
properties. Biomass-derived platform molecules are considered as po-
tential alternative to replace fossil-based resources [1,2], which can
meet the increasing demand of energy for the production of fuels and
chemicals. Biomass-derived platform molecules can be converted into a
variety of valuable chemicals [3–6], such as 5-hydroxymethylfurfural
At present, many reports have focused on the preparation of GVL,
especially on the hydrogenation of LA and its esters to GVL, which is
one of the key steps in biomass conversion reactions [15,16]. The
synthesis of GVL from LA or its esters can be divided as homogeneous
reaction and heterogeneous reaction. Generally speaking, the catalysts
used in homogeneous systems are mainly noble metal coordination
compounds, such as Ru(acac)
showed excellent catalytic performances under mild conditions. For
example, [Ir(COE) Cl] achieved a 99 % yield of GVL under 5.0 MPa H
at 100 °C for 15 h, and Pd(DTBPE)Cl obtained a almost 100 % yield of
GVL under 0.5 MPa H at 80 °C for 5 h. However, it is obvious that the
3 2 2
, [Ir(COE) Cl] and so on [17,18], which
(
(
5-HMF), ethyl levulinate (EL), levulinic acid (LA), γ-valerolactone
GVL), 2-methyl tetrahydrofuran (2-MTHF), etc. Among them, GVL is
2
2
2
2
identified as an important building block due to its wide application
7–11]. As a non-toxic and coconut scented compound, GVL could be
used as an additive of food and perfume, and proved to be an excellent
solvent [12,13]. Besides, in consideration of the carbonyl and cyclic
lactone groups on GVL, it could be directly used as raw material to
prepare high calorific value liquid hydrocarbon fuels and fuel additives
2
[
recovery of the catalysts in the homogeneous system can be extremely
difficult, which significantly limited its large-scale commercial appli-
cation. Moreover, the complex synthesis methods for homogeneous
catalysts could also constrain their application.
Hence, heterogeneous catalyst would be a more suitable choice for
⁎
Corresponding author at: School of Environmental Science and Engineering, Tianjin University, Tianjin, 300350, China.
E-mail address: xbltju@tju.edu.cn (X. Lu).
https://doi.org/10.1016/j.mcat.2020.111155
Received 11 March 2020; Received in revised form 9 July 2020; Accepted 29 July 2020
2468-8231/ © 2020 Elsevier B.V. All rights reserved.

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
hydrogenation of LA and its esters to GVL. In heterogeneous reaction
system, noble metal catalysts such as Ru, Ir, Pd and Au, etc. [19–22],
could reach a considerable GVL yield under mild conditions. Du et al.
Aladdin (99 %). All the chemicals were directly used without pre-
treatment after purchase.
[
19] reported that the Ir/AC catalyst obtained a 95 % GVL production
2.2. Catalyst preparation
under 5.5 MPa H at 150 °C for 2 h. Nonetheless, a common sense is that
2
the noble metal catalysts are too expensive for the industrial synthesis
of GVL. Compared to noble metal catalysts, transition metal catalysts
are more economical and practical, especially for nickel-based catalysts,
MOCVD method: The deposition of Ni particles on the support was
conducted in a schlenk bottle. Specifically, a certain amount of Ni
(acac) precursor was mechanically mixed with the support, then the
2
such as Ni/Al
23] obtained 69.9 % GVL yield and 74.9 % LA conversion with Ni/
MgO catalyst under optimal reaction conditions of 150 °C, 1.0 MPa H
2
O
3
, Ni-Cu/Al
2
O
3
, Ni/MgAlO
2
, Ni/MgO [23–25]. Lv et al.
mixture was transferred into the schlenk bottle. After vacuuming, the
schlenk bottle was put into the drying oven at 200 °C for 12 h, which
resulted in a complete deposition of the precursor onto the support.
Finally, the thermal decomposition and reduction progress were carried
[
2
for 2 h. In recent years, Raney Ni, Ni@NCMs, also had been proved to
be effective for the hydrogenation of LA and its esters to GVL [26, 27].
However, for the traditional preparation process such as co-precipita-
tion and impregnation, the active site cannot be dispersed uniformly on
the catalyst. Moreover, the utilization of nickel was relative low, and
the solvents might block the inner surface and pores of the supports
which leading to a low loading. Fortunately, MOCVD is considered as a
potential alternative to obtain the catalyst with the Ni nanoparticles
well-dispersed [28]. To our knowledge, the nickel-based catalysts pre-
pared by MOCVD method for the hydrogenation of LA and its esters to
GVL was reported minimally. Thus, considerable interest has been paid
on MOCVD synthesis of supported Ni catalyst. The key issue for the
MOCVD located on the selection of suitable Ni-based metal-organic
precursors, which should be adequately volatile and sufficiently stable
at room temperature as well as non-toxic. Highly and homogeneously
−
1
2 2
out at 460 °C for 4 h in the flow of H /N (40 mL min , volume flow
ratio was 1:1), with a heating rate of 5 °C/min. The obtained catalysts
were recorded as x wt.% Ni/support, where x was the theoretical
loading amount of Ni.
Impregnation method (IM): The catalyst prepared by classical im-
pregnation method was using an aqueous solution of Ni(NO ) ·6H O
3
2
2
and activated carbon. Firstly, the activated carbon was dried at 120 °C
for 12 h. Secondly, AC was adequately dispersed in an aqueous solution
of Ni(NO
3 2 2
) ·6H O at room temperature for 24 h. Then the mixture was
dried at 120 ℃ for 12 h. Subsequently, the dried sample was then
−1
2 2
subjected to H /N (40 mL min , volume flow ratio was 1:1) reduction
at 460 °C for 4 h. The obtained catalysts were recorded as 2 wt.% Ni/AC
(IM).
nickel-supported catalysts has been synthesized by using Ni(COD)
the MOCVD precursor [28]. However, the toxicity and instability lim-
ited its wide application. We found that Ni(acac) shoud be a good
choice for the replacement of Ni(COD) , owing to its appropriate sub-
limation and thermal decomposition temperature. Additional, it is re-
lative cheaper, less toxic and more stable than Ni(COD)2. Based on the
above, supported Ni catalysts prepared by MOCVD method with Ni
2
as
2.3. Catalyst characterization
2
Thermogravimetric (TG) and differential thermal analysis (DTA)
experiments were carried out using a QMA200 M instrument. An ap-
propriate amount of samples were taken in the ceramic frame and the
temperature changed from room temperature to 800 °C in the atmo-
2
−
1
sphere of N with the heating rate of 10 °C min . The weight changes
2
(
acac)
2
as the metal-organic precursor were applied to the hydrogena-
of samples that underwent temperature change were recorded.
tion of EL to GVL.
Fourier transform infrared (FT-IR) spectra were recorded at 298 K
on powder samples using the KBr wafer technique on a Nicolet Impact
2
In this work, Ni(acac) (Nickel(II) acetylacetonate) was used as the
−
1
metal-organic precursor to synthesize a series of supported Ni catalysts
via MOCVD method, and the as-synthesized catalysts were manifested
to be efficient for the hydrogenation of EL to produce GVL under mild
410 with a resolution of 4 cm
.
X-ray diffraction (XRD) was measured on an ESCALAB250
(ThermoScientific) diffractometer with monochromatic Cu-Kα as the
radiation source (λ = 1.54056 Å) and operated at 40 kV and 200 mA.
conditions. N
further taken to characterize the catalysts. The results showed that the
wt.% Ni/AC(MOCVD) presented superior catalytic activity when
2
-adsorption/desorption, XRD, FT-IR, ICP, and TEM were
−
1
The scan speed was 3° min
with a scanning angle of 5−80°.
2
Nitrogen adsorption/desorption measurements were performed on
Quantachrome Autosorb iQ2 automated gas sorption system. The spe-
cific surface area of the sample was calculated by the Brunauer-Emmett-
Teller (BET) method and the total pore size distributions were de-
termined by Barrett-Joyner-Halenda (BJH) method from the isotherms.
Before measurements, the samples were deaerated at 100 °C for 1 h and
then outgassed at 300 °C for 3 h.
compared with the catalyst prepared by impregnation method. This
behavior is explained in terms of the smaller Ni nanoparticles (4.28 nm)
and higher dispersion on 2 wt.% Ni/AC(MOCVD). Among the catalysts,
the 2 wt.% Ni/AC catalyst exhibited the best catalytic performance with
99.7 % EL conversion and 79.8 % GVL yield under 1 MPa initial H
2
pressure (measured at room temperature) at 250 °C for 2 h. In addition,
the reaction conditions and pathways were also investigated. This paper
successfully synthesized the Ni-based catalysts with high dispersion and
relatively small particle size via MOCVD method, which presented ex-
cellent catalytic activity and will promote the further commercial ap-
plication of Ni-based catalysts.
To determine the composition of the catalyst, Inductively Coupled
Plasma Atomic Emission Spectrometry (ICP-OES) was conducted on a
thermal 7000 series analyzer. Initially, the samples were pretreated
with aqua regia dissolving solid samples. Subsequently, the microwave
digestion took place in a microwave digestion apparatus for 30 min to
completely dissolve the metal. After cooling and filtration, the sample
was then submitted to metal analysis.
2. Experimental
2
Transition metal catalysts were analyzed using a Tecnai G F30
2.1. Materials
(FEI) transmission electron microscope. The samples were sonicated for
5
min in ethanol to make them highly suspended, then dripped in a
The organic metal precursor nickel acetylacetonate (Ni(acac)
2
), Ni
carbon film supported on a copper grid.
(
3 2 2
NO ) ·6H O, Dodecane (internal standard) were purchased from
Shanghai Dibai Biotechnology Co., Ltd., Tianjin Kemiou Chemical
Reagent Co. and Aladdin. Catalyst support included activated carbon
2.4. Ethyl levulinate hydrogenation
(
AC), SiO
2
, Al
2
O
3
and H-ZSM5 were purchased from NORIT, Tianjin
The catalytic conversion of EL was performed in a 50 mL batch re-
actor (Parr 4597). For each reaction, 0.2883 g (2 mmol) of EL and
0.4258 g (2.5 mmol) of dodecane (internal standard) were dissolved in
20 mL of isopropanol, together with 0.1 g of catalyst. Thenceforth, the
Kemiou Chemical Reagent Co., Ltd., Aladdin, Aladdin and Shen Shi
Mining Co., Ltd., respectively. EL was purchased from Beijing
Bellingway Technology Co., Ltd. (99 %), and GVL was purchased from
2

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
2
reaction was carried out at a certain temperature, H pressure and re-
action time with the mechanical stirring speed of 1000 rpm. After the
reaction, the liquid products were analyzed by GC–MS (Agilent 6890A-
5975C) and GC (SP-7890). The conversion of EL and the yield of GVL
were calculated according to the following equations:
c(residual ethyl levulinate)
c(initial ethyl levulinate)
⎛
⎜
⎞
⎟
Conversion=
1-
×100%
⎝
⎠
(1)
(2)
c(product)
c(initial ethyl levulinate)
Product yield=
×100%
3. Results and discussion
3.1. Catalyst characterization results
Ni(acac)
cause of its adequate volatility nature and relatively wide temperature
gap” between evaporation and decomposition. In addition, it is com-
mercially available and sufficiently stable at room temperature, com-
pared to other precursors such as Ni (COD) , Ni(hfac) and so on.
Thermogravimetric analysis was applied to verify the volatile
characteristics of Ni(acac) (Fig. 1). The precursor showed excellent
2
was used as the precursor for nickel-based catalyst be-
“
Fig. 2. FT-IR spectra of Ni-based catalysts.
band at 1430 cm-1 and 1378 cm , which could be attributed to the
stretching vibrations and flexural vibrations of CeH bond. Sig-
nificantly, this typical spectrum disappeared in 10 wt.% Ni/AC due to
-1
2
2
2
stability up to 210 °C and illustrated a gradual weight loss before 150 °C
thermal decomposition of Ni(acac)
was similar to AC, indicating the purity of the as-synthesized catalyst
was high. When mixing Ni(acac) with AC, the large π bond resulted in
strongly bonded to the functional group of the AC. It was clear
2
. The spectrum of 10 wt.% Ni/AC
due to the removal of water. Moreover, the melting point (226−238 °C)
−
1
and boiling point (220−235 °C 14.67 kPa ) of Ni(acac)
close, which indicated that Ni(acac) was easy to sublimate. Thus, the
sublimation temperature could be set as 200 °C under vacuum in this
study. DTA of the TG profiles of Ni(acac) showed 76.6 % weight loss
2
are very
2
2+
2
Ni
that NiO species was successfully converted to metallic nickel.
XRD analysis was carried out to investigate the crystal structure of
the catalysts (as shown in Fig. 3). The broad peak at 2θ = 14.6° was the
characteristic peak of AC support. The three sharp diffraction peaks
observed at 2θ = 45.5°, 51.9°, and 76.5°are corresponding to the (111),
2
around 220−460 °C, due to the loss of organic ligands. The result was
in a good agreement with the calculated weight percent loss to produce
pure Ni (22.8 % calculated), which confirmed the precursor could be
totally decomposed. Therefore, in this work, the thermal decomposition
temperature was set at 460 °C.
(
200), and (220) lattice planes of Ni° (ICDS 98−064-6090), respec-
tively [21,24,29]. No diffraction peak of NiO was found in the XRD
patterns, further confirmed that the Ni(acac) precursors were com-
In order to investigate the possible formation mechanism of small Ni
particles, FT-IR spectra of catalysts were performed and the results were
2
pletely reduced to Ni° particles. This result was consistent with the TG/
DTA results. Furthermore, the intensity of the diffraction peaks of Ni°
increased with the increasing of Ni loadings from 2 wt.% to 10 wt.%.
To determine the textural parameters and actual loading amount of
shown in Fig. 2. The spectrum of 10 wt.% Ni(acac)
2
/AC (meant that the
mass fraction of Ni atom was 10 %) exhibited the typical broad-band at
−1
-1
1641 cm
and 1530 cm , which could be assigned to telescopic vi-
bration of C]O and the stretching vibration of C]C. It was mainly
attributed to the fact that the C]O double bond of nickel metal and
acetylacetone formed a large π bond resulting in C]C and C]O
stretching vibration peak shifted to the low-frequency direction. Be-
the synthesized Ni/AC catalysts, N
analysis were carried out separately. As shown in Table 1, with the
increase of Ni content from 0 to 10 wt.%, the BET surface area (SBET
2
adsorption/desorption and ICP
)
2
sides, the spectrum of 10 wt.% Ni(acac) /AC exhibited a typical broad-
Fig. 1. TG/DTA curve of Ni(acac)
2
precursor.
Fig. 3. XRD patterns of Ni/AC catalysts with different Ni loadings.
3

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
Table 1
Table 2
a
Textural properties of Ni/AC catalysts with different Ni loadings.
Hydrogenation results of EL over different catalysts .
Catalyst
S
(
BET
Pore sizea Pore volume
Ni contentb
(wt.%)
Entry
Catalyst
Solvent
Conv. (EL)/%
Yield/%
m2 g−1
)
(nm)
(cm
3
g
−1
)
GVL
A
B
AC
980
965
887
795
958
2.78
2.35
2.34
2.32
2.36
0.56
0.56
0.52
0.47
0.56
–
2
5
1
2
wt.% Ni/AC
wt.% Ni/AC
0 wt.% Ni/AC
wt.% Ni/
2.10
5.21
10.49
1.88
1
2
3
4
5
6
7
8
9
Without
AC
10 wt.% Ni/AC
10 wt.% Ni/Al
10 wt.% Ni/SiO
10 wt.% Ni/HZSM-5
10 wt.% Ni/AC
10 wt.% Ni/AC
10 wt.% Ni/AC
2 wt.% Ni/AC
Isopropanol
Isopropanol
Isopropanol
Isopropanol
Isopropanol
Isopropanol
Ethanol
2-butanol
Methanol
Isopropanol
Isopropanol
Isopropanol
10.0
14.0
97.6
79.4
26.3
14.8
39.9
78.3
97.4
99.7
90.2
97.4
7.8
1.3
2.9
27.3
4.3
8.2
0.2
–
39.2
82.4
1.3
9.3
26.4
0.8
0.8
1.8
5.1
0.6
0.2
0.0
0.0
0.0
2.2
2.4
1.9
10.2
67.0
47.3
16.9
14.4
33.8
38.3
12.3
79.8
62.5
67.7
2 3
O
AC(IM)
2
a
Average pore diameter was calculated from the desorption branches by the
BJH model.
b
Determined by ICP analysis.
10
11
12
b
2 wt.% Ni/AC(IM)
5 wt.% Ni/AC
and the pore volume (V) reduced slowly (SBET: from 980 to 795 m g−1,
Pore volume: from 0.56 to 0.47 cm g ). Simultaneously, the pore size
also showed a minimal drop due to the obstruction of the channel by
2
3
−1
a
Reaction conditions: EL (2 mmol), dodecane (2.5 mmol, internal standard
2
7
for GC), solvent (20 mL), catalyst (0.1 g) and initial H
action time (2 h), reaction temperature (250 °C).
2
pressure (1 MPa), re-
the Ni particles . For the catalyst prepared by MOCVD method, the ICP
results demonstrated that the actual loading of Ni was slightly higher
than the theoretical loadings. The difference between actual metal
loadings and the theoretical loadings could contribute to the loss of
bound water and impurities in AC. Compared with the catalyst pre-
pared by MOCVD method, the loss of Ni was inevitable. As a result, we
chose the MOCVD method.
TEM was conducted to investigate the morphologies and metal
dispersions of the Ni/AC catalysts. As depicted in Fig. 4, the average
diameter of 2 wt.% Ni/AC (MOCVD) was around 4.28 nm, while the
average diameter of 2 wt.% Ni/AC (IM) was around 7.58 nm. Addi-
tional, as determined from XRD analysis, the crystal size of 2 wt.% Ni/
AC (MOCVD) calculated by using Scherrier equation through the Ni
to our result [30]. Therefore, the Ni nonoparticles obtained from
MOCVD method exhibited narrower size distribution with high dis-
persion when compared to the impregnation method.
3
.2. Activity and stability of synthesized catalysts
As shown in Table 2, the catalytic performance of the as-prepared
catalysts was examined under different conditions, including different
supports, solvents and Ni loadings. The main product was GVL, ac-
companied by the formation of levulinic esters (A) and valerate esters
B) due to side reaction (Scheme 1). For the 2 wt.% Ni/AC catalyst
prepared by MOCVD method (as shown in entry 10), the EL was almost
completely converted with the conversion of 99.7 % and GVL yield of
(
(
111) diffraction peak was around 4.52 nm, which was consistent with
the TEM results. It was also reported that the mean size of Ni/AC
prepared by impregnation method was about 8.0 nm, which was closed
7
9.8 %. Apparently, the Ni active sites simultaneously promoted the
Fig. 4. TEM images of (a, b, c) 2 wt.% Ni/AC, (d, e, f) 2 wt.% Ni/AC (IM).
4

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
Scheme 1. Main products from hydrogenation
of EL to GVL.
fracture of the CeO bond to form γ-hydroxyl acid or esters. At last, GVL
was formed from the γ-hydroxyl acid or its esters through the process of
dealcoholization and cyclization.
higher GVL yield, the temperature of 250 °C was selected for the next
research.
And then, the effect of H
2
pressure was shown in Fig. 5 (b). It can be
To optimize the best reaction conditions for the hydrogenation of EL
catalyzed by the Ni-based catalyst, the effect of different supports was
investigated. For entry 1 and 2, it demonstrated that the EL conversion
only reached 10.0 % without catalyst and slightly increased to 14.0 %
seen that the H pressure had a slight influence on the transformation of
EL into GVL. This phenomenon may be ascribed to the reason that the
isopropanol could also be the hydrogen donor on the CTH of EL to GVL.
2
When the H
2
pressure was 1 MPa, the conversion and yield achieved
pressure led
on the AC support. Interestingly, Ni/SiO
yield, owing to the low acidity of the supports. When using the mod-
erately strong acidity oxides, such as Al , the result showed that the
conversion and yield could reach to 79.4 % and 47.3 %, respectively.
The diameter of Ni(acac) was around 12 Å, which is bigger than the
micropore of the HZSM-5 (d = 5∼6 Å) [31,32]. Therefore, during the
MOCVD progress, the Ni(acac) can hardly deposit into the micropore
2
exhibited low conversion and
97.6 % and 79.8 %, respectively. Further enhancing the H
2
to the decline of GVL yield, implying more undesirable by-products
generated, such as 1, 4-pentanediol (1, 4-PDO), 2-MTHF and so on,
which was in accordance with previous reports [37,38].
2 3
O
2
For the sake of higher GVL yield, we investigated the effect of re-
action time. As shown in Fig. 5(c), the result exhibited that the reaction
rate was rapid within the first 2 h and the GVL yield reached 79.8 %
with almost complete conversion of EL. Although the conversion of EL
to GVL could be promoted with further prolonging the time, the se-
lectivity of the GVL displayed a downward trend at the meantime.
Above all, the optimized condition for efficient catalytic hydrogenation
of EL to GVL was: 250 ℃, 1 Mpa initial H2 and 2 h.
The stability is significant for the commercial industrial production,
so we studied the reusability of 2 wt.% Ni/AC under optimum condi-
tions (Fig. 6). After each cycle, the catalyst was collected by filtration,
washed with fresh isopropanol, and then dried for the next run. After
the second run, both the conversion of EL and the yield of GVL de-
creased obviously. After four cycle, the conversion of EL and yield of
GVL was 85.3 % and 68.2 %, respectively.
2
of the HZSM-5 which probably leading the surface aggregation of Ni
particles with lower activity. As shown in Table 2 (entry 3, 4, 5 and 6),
it was noteworthy that the activity of Ni/AC was higher than any other
catalysts, which was probably correlated with the large specific surface
area and pore diameter of activated carbon. Furthermore, acid sites on
the surface of the activated carbon also promoted the reaction [33]
Therefore, Ni/AC was selected for further studies.
Moreover, the effect of the solvent on the hydrogenation activity of
the catalyst was investigated. According to entries 3, 7, 8, and 9 in
Table 2, solvents significantly affected the hydrogenation of EL to GVL
over Ni/AC catalyst under optimum conditions. Generally, CTH (cata-
lytic transfer hydrogenation) reaction with second alcohol provided
much higher EL conversion and GVL yield than primary alcohol. The
reducing potential of various alcohols was IPA < 2-butanol <
ethanol < methanol [34]. The yield of GVL was in line with this order.
The highest reduction of Methanol resulted in relatively poor H-donor
abilities. When using methanol as a solvent, the conversion was 97.4 %
whereas little GVL was obtained. As depicted in Table 2 entry 9, it was
clearly that most of EL converted to methyl levulinate (intermediate).
Primary alcohol provided little GVL because of difficult β-H remove-
ment in the CTH reaction. Among all the alcohols, isopropanol showed
better performance, and was chose for further researches.
The catalytic performance of different Ni loadings catalysts was also
listed in Table 2. When a small amount of Ni deposited on the AC, the
conversion and yield both rose obviously. However, the selectivity of EL
further decreased with the increase of Ni loadings. Previous studies
revealed that the hydrogenation and lactonization were two inter-
mediate steps for the transformation of EL to GVL [[23]]]]. The ex-
cessive amount of nickel affected the process of lactonization, which
tended to produce by-product A [[34]]]]. For the 2 wt.% Ni/AC (IM)
shown in entry 11, the catalytic performance was inferior to the Ni/AC
with EL conversion of 90.2 % and GVL yield of 62.5 %. Compared with
the results, it was obviously that the Ni/AC prepared by MOCVD
method had better catalytic performance, which was due to the smaller
particle size and the higher dispersion of Ni/AC [[26]]]].
Thereafter, the effect of reaction temperature on the catalytic ac-
tivity of 2 wt.% Ni/AC was studied. As shown in Fig. 5 (a), the con-
version of EL increased stably and the yield of GVL rose rapidly with
temperature increased from 200 to 250 °C. The highest yield of GVL
could reach 79.8 % within 2 h at 250 °C. In the meantime, A1 (isopropyl
levulinate) decreased. Moreover, B1 (isopropyl valerate) increased.
This phenomenon reflected that the CeO bond fractured easier at high
temperature [35,36]. In order to shorten the reaction time and obtain
XRD diffraction peaks of the fresh and spent Ni/AC catalyst were
shown in the Fig. 7. Apparently, for the spent 2 wt.% Ni/AC, the
characteristic peak of Ni0 at 44.0° and 51.2° became weaker. Moreover,
additional peak at 43.2° which belonged to (200) planes of NiO ap-
peared for the spent 2 wt.% Ni/AC. These phenomenon suggested that
the Ni species possibly undergo oxidation or leaching loss during the
reaction, which was considered as the disadvantage of almost all
monoatomic supported catalysts [[39]] The Brunauer–Emmett–Teller
(BET) surface area, pore volume and pore size for the spent 2 wt.% Ni/
2
AC were displayed in Table 3. The surface area declined from 965 m /g
2
to 886 m /g after the reaction. Simultaneously, the pore size of the
catalyst3 slightly declined and the pore volume decreased from
−1
3 −1
0.56 cm g
to 0.41 cm g . According to the previous literature, it
should be caused by the coke deposition on the surface and in the pores
of the catalyst. ICP-OES was performed to determine the Ni content of
the spent catalyst. The result exhibited that the Ni content decreased
from 2.1 wt.% to 1.86 wt.% after recycling, which further confirm the
leaching loss of Ni species during the reaction. It is reported that ag-
gregation and sintering of Ni particles could also result in the deacti-
vation of the catalyst. Thus, TEM was carried out to observe the mor-
phology change of Ni on the surface of the catalyst. As depicted in
Fig. 8, the Ni particles still distributed uniformly on the surface of AC,
but the particle size has increased from 4.28 nm to 5.07 nm for the
spent catalyst. In summary, we concluded that it is the oxidation, ag-
gregation and leaching loss of the Ni species, together with the coke
deposition during the reaction which leading the deactivation of the Ni/
AC catalyst.
3.3. Proposed mechanism
Based on the previous study [40–43], the possible mechanism of Ni/
5

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
Fig. 5. The effect of the reaction conditions in the conversion of EL in the presence of 2 wt.% Ni/AC catalyst. EL (2 mmol), dodecane (2.5 mmol, internal standard for
GC), catalyst (0.1 g); A1 (isopropyl levulinate); B1 (isopropyl valerate); (a) Reaction temperature (1 MPa H , 2 h); (b) Reaction pressure (250 °C, 2 h); (c) Reaction
time (250 °C, 1 MPa H ).
2
2
6

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
2
Fig. 6. The recycling experiments of 2 wt.% Ni/AC (250 °C, 1 MPa H , 2 h).
Fig. 8. TEM images of (a, b, c) fresh 2 wt.% Ni/AC and (d, e, f) spent 2 wt.% Ni/
AC.
the GVL. Because the rate of transformation from EL to 4-HPE was
rapid, equilibrium of reaction was up to the generation of 4-HPE. Then
GVL was further hydrogenated to form 1, 4-PDO, 2-MTHF and valerate
esters. In addition, EL could produce levulinic esters due to the sub-
stitution reaction at low temperature or pressure.
4. Conclusions
Overall, this study clearly demonstrated that the high dispersed Ni/
AC catalyst prepared by MOCVD method with Ni(acac) as the pre-
2
cursor was successfully applied in efficient catalytic hydrogenation of
EL to GVL. Results revealed that the Ni particles obtained from MOCVD
method exhibited narrower size distribution with high dispersion,
which presented superior catalytic activity in hydrogenation of EL to
GVL when compared to the impregnation method. Therefore, the design
of high dispersed Ni particles with smaller particle size via MOCVD
method will facilitate the metal-catalyzed hydrogenation of EL to GVL.
In addition, the 2 wt.% Ni/AC catalyst was proved to have the best
catalytic performance for the hydrogenation of EL to GVL. The results
showed that the conversion of EL achieved 99.7 % and the yield of GVL
reached 79.8 % via IPA as solvent under the conditions of 250 °C and
Fig. 7. XRD patterns of fresh and spent 2 wt.% Ni/AC catalyst.
AC catalyzed hydrogenation of EL to GVL was proposed above. As de-
picted in Scheme 2, GVL could be generated in two pathways. The first
route showed that γ-hydroxy pentanoic was produced from EL through
the hydrogenation of the ketone group. γ-hydroxy pentanoic, as the
intermediate, was unstable under the reaction condition, therefore, it
could be converted into GVL via the intramolecular transesterification.
The other route involved the ketone group of EL hydrogenated to form
ethyl 4-hydroxyvalerate (4-HPE), which generated GVL upon deal-
coholization and cyclization. As we know, the reaction was driven by
the thermodynamic properties of the intermediates and the products.
Besides this, it was apparently that the reaction was irreversible. Ac-
cording to the GC-MS result, no 4-HPE was detected, which demon-
strated that the 4-HPE were unstable and can quickly be transferred to
1
MPa H2 for 2 h. This novel Ni/AC catalysts possessed excellent cata-
lytic performance was proved to be a promising catalyst in the hydro-
genation of EL to GVL in biomass utilization. Preparation of the cata-
lysts using other non-precious metals and the study of their catalytic
performances are in progress.
CRediT authorship contribution statement
Na Ji: Conceptualization, Methodology, Writing - review & editing.
Zhenyu Liu: Writing - original draft. Xinyong Diao: Writing - review &
editing. Jinrong Bao: Visualization, Investigation. Zhihao Yu:
Software. Chunfeng Song: Writing - review & editing. Qingling Liu:
Table 3
Textural properties of fresh and spent 2 wt.% Ni/AC catalyst.
Catalyst
S
BET (m2 g−1
)
Pore sizea (nm)
Pore volume (cm3 g−1
)
Ni contentb (wt.%)
2
2
wt.% Ni/AC (fresh)
wt.% Ni/AC (spent)
965
886
2.35
2.33
0.56
0.41
2.10
1.86
a
b
Average pore diameter was calculated from the desorption branches by the BJH model.
Determined by ICP analysis.
7

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
Scheme 2. Proposed reaction pathways of EL to GVL.
Writing - review & editing. Degang Ma: . Xuebin Lu: Writing - review
editing.
[12] L. Qi, Y.F. Mui, S.W. Lo, M.Y. Lui, G.R. Akien, I.T. Horváth, Catalytic conversion of
fructose, glucose, and sucrose to 5-(Hydroxymethyl)furfural and levulinic and
formic acids in γ-Valerolactone As a green solvent, ACS Catal. 4 (2014) 1470–1477.
&
[
13] D. Fegyverneki, L. Orha, G. Láng, I.T. Horváth, Gamma-valerolactone-based sol-
vents, Tetrahedron 66 (2010) 1078–1081.
Declaration of Competing Interest
[
14] S. Raoufmoghaddam, M.T. Rood, F.K. Buijze, E. Drent, E. Bouwman, Catalytic
conversion of gamma-valerolactone to epsilon-caprolactam: towards nylon from
renewable feedstock, ChemSusChem. 7 (2014) 1984–1990.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
x y
[15] M. Sun, J. Xia, H. Wang, X. Liu, Q. Xia, Y. Wang, An efficient Ni Zr O catalyst for
hydrogenation of bio-derived methyl levulinate to γ-valerolactone in water under
low hydrogen pressure, Appl. Catal. B 227 (2018) 488–498.
[
[
[
16] G. Novodárszki, H.E. Solt, G. Lendvay, R.M. Mihályi, A. Vikár, F. Lónyi, J. Hancsók,
J. Valyon, Hydroconversion mechanism of biomass-derived γ-valerolactone, Catal.
Today 336 (2019) 50–62.
17] C. Delhomme, L.-A. Schaper, M. Zhang-Preße, G. Raudaschl-Sieber, D. Weuster-
Botz, F.E. Kühn, Catalytic hydrogenation of levulinic acid in aqueous phase, J.
Organomet. Chem. 724 (2013) 297–299.
18] W. Li, J.-H. Xie, H. Lin, Q.-L. Zhou, Highly efficient hydrogenation of biomass-
derived levulinic acid to γ-valerolactone catalyzed by iridium pincer complexes,
Green Chem. 14 (2012).
Acknowledgments
This work was supported by the National Key R&D Program of
China (2018YFB1501500), National Natural Science Foundation of
China (Nos. 21975181; 21690083; 51876180), and Tibet University
2018, 2019 Central Financial Support Special Funds for Local Colleges
and Universities ([2018] No. 54; [2019] No. 19).
[19] X. Du, Y. Liu, J. Wang, Y. Cao, K. Fan, Catalytic conversion of biomass-derived
levulinic acid into γ-valerolactone using iridium nanoparticles supported on carbon
nanotubes, Chin. J. Catal. 34 (2013) 993–1001.
References
[
[
[
[
[
[
[
[
[
[
[
20] T. Pan, J. Deng, Q. Xu, Y. Xu, Q.-X. Guo, Y. Fu, Catalytic conversion of biomass-
derived levulinic acid to valerate esters as oxygenated fuels using supported ru-
thenium catalysts, Green Chem. 15 (2013).
21] K. Yan, T. Lafleur, C. Jarvis, G. Wu, Clean and selective production of γ-valer-
olactone from biomass-derived levulinic acid catalyzed by recyclable Pd nano-
particle catalyst, J. Clean. Prod. 72 (2014) 230–232.
22] X.L. Du, Q.Y. Bi, Y.M. Liu, Y. Cao, K.N. Fan, Conversion of biomass-derived levu-
linate and formate esters into gamma-valerolactone over supported gold catalysts,
ChemSusChem. 4 (2011) 1838–1843.
[
1] V. Menon, M. Rao, Trends in bioconversion of lignocellulose: biofuels, platform
chemicals & biorefinery concept, Prog. Energy Combust. Sci. 38 (2010) 522–550.
2] A. Eldering, P.O. Wennberg, D. Crisp, D.S. Schimel, M.R. Gunson, A. Chatterjee,
J. Liu, F.M. Schwandner, Y. Sun, C.W. O’Dell, C. Frankenberg, T. Taylor, B. Fisher,
G.B. Osterman, D. Wunch, J. Hakkarainen, J. Tamminen, B. Weir, The orbiting
carbon observatory-2 early science investigations of regional carbon dioxide fluxes,
Science 358 (2017).
[
23] J. Lv, Z. Rong, Y. Wang, J. Xiu, Y. Wang, J. Qu, Highly efficient conversion of
biomass-derived levulinic acid into γ-valerolactone over Ni/MgO catalyst, RSC Adv.
[
3] C. Li, X. Zhao, A. Wang, G.W. Huber, T. Zhang, Catalytic transformation of lignin for
the production of chemicals and fuels, Chem. Rev. 115 (2015) 11559–11624.
4] J.L.X. Hong, T. Maneerung, S.N. Koh, S. Kawi, C.-H. Wang, Conversion of coal fly
ash into zeolite materials: synthesis and characterizations, process design, and its
cost-benefit analysis, Ind. Eng. Chem. Res. 56 (2017) 11565–11574.
5
(2015) 72037–72045.
[
24] K. Jiang, D. Sheng, Z. Zhang, J. Fu, Z. Hou, X. Lu, Hydrogenation of levulinic acid to
γ-valerolactone in dioxane over mixed MgO–Al
2 3
O supported Ni catalyst, Catal.
Today 274 (2016) 55–59.
[
[
[
5] M. Zheng, J. Pang, A. Wang, T. Zhang, One-pot catalytic conversion of cellulose to
ethylene glycol and other chemicals: from fundamental discovery to potential
commercialization, Chin. J. Catal. 35 (2014) 602–613.
25] I. Obregón, E. Corro, U. Izquierdo, J. Requies, P.L. Arias, Levulinic acid hydro-
genolysis on Al -based Ni-Cu bimetallic catalysts, Chin. J. Catal. 35 (2014)
56–662.
2 3
O
6
6] L. Landenna, A. Villa, R. Zanella, C. Evangelisti, L. Prati, Gold-iridium catalysts for
26] Z. Yang, Y.B. Huang, Q.X. Guo, Y. Fu, RANEY(R) Ni catalyzed transfer hydro-
genation of levulinate esters to gamma-valerolactone at room temperature, Chem.
Commun. (Camb.) 49 (2013) 5328–5330.
27] D. Liu, L. Zhang, W. Han, M. Tang, L. Zhou, Y. Zhang, X. Li, Z. Qin, H. Yang, One-
step fabrication of Ni-embedded hierarchically-porous carbon microspheres for le-
vulinic acid hydrogenation, Chem. Eng. J. 369 (2019) 386–393.
the hydrogenation of biomass derived products, Chin. J. Catal. 37 (2016)
1
771–1775.
7] P.M. Mortensen, J.D. Grunwaldt, P.A. Jensen, K.G. Knudsen, A.D. Jensen, A review
of catalytic upgrading of bio-oil to engine fuels, Appl. Catal. A Gen. 407 (2011)
1
–19.
[
[
8] Z. Zhang, Synthesis of gamma-valerolactone from carbohydrates and its
Applications, ChemSusChem. 9 (2016) 156–171.
9] D.M. Alonso, S.G. Wettstein, J.A. Dumesic, Gamma-valerolactone, a sustainable
platform molecule derived from lignocellulosic biomass, Green Chem. 15 (2013).
28] L. Brissonneau, A. Reynes, C. Vahlas, MOCVD processed Ni films from Nickelocene.
Part III: gas phase study and deposition mechanisms, Chem. Vap. Depos. 5 (1999)
2
81–290.
29] W. Haojie, C. Chun, W. Haimin, W. Guozhong, Z. Huijun, An efficient and reusable
bimetallic Ni Fe nps@c catalyst for selective hydrogenation of biomass-derived
[
10] K. Yan, Y. Yang, J. Chai, Y. Lu, Catalytic reactions of gamma-valerolactone: a
platform to fuels and value-added chemicals, Appl. Catal. B 179 (2015) 292–304.
11] Y. Roman-Leshkov, C.J. Barrett, Z.Y. Liu, J.A. Dumesic, Production of di-
methylfuran for liquid fuels from biomass-derived carbohydrates, Nature 447
3
levulinic acid to γ-valerolactone, Chin. J. Catal. 39 (2018) 1599–1607.
30] C. Li, G. Xu, Y. Zhai, X. Liu, Y. Ma, Y. Zhang, Hydrogenation of biomass-derived
ethyl levulinate into γ-valerolactone by activated carbon supported bimetallic Ni
[
(2007) 982–985.
8

N. Ji, et al.
Molecular Catalysis 495 (2020) 111155
and Fe catalysts, Fuel 203 (2017) 23–31.
31] Y.Y. Titova, L.B. Belykh, F.K. Shmidt, Ziegler-type nickel-based hydrogenation
Al–Zr mixed oxides, Appl. Catal. A Gen. 510 (2016) 11–19.
[
[
[
[
[
[38] Z. Li, M. Zuo, Y. Jiang, X. Tang, X. Zeng, Y. Sun, T. Lei, L. Lin, Stable and efficient
CuCr catalyst for the solvent-free hydrogenation of biomass derived ethyl levulinate
to γ-valerolactone as potential biofuel candidate, Fuel 175 (2016) 232–239.
[39] Z. Rong, Z. Sun, L. Wang, J. Lv, Y. Wang, Y. Wang, Efficient conversion of levulinic
acid into γ-Valerolactone over Raney Ni catalyst prepared from melt-quenching
alloy, Catal. Lett. 144 (2014) 1766–1771.
[40] P.P. Upare, J.M. Lee, D.W. Hwang, S.B. Halligudi, Y.K. Hwang, J.-S. Chang,
Selective hydrogenation of levulinic acid to γ-valerolactone over carbon-supported
noble metal catalysts, J. Ind. Eng. Chem. 17 (2011) 287–292.
[41] Y. Kuwahara, Y. Magatani, H. Yamashita, Ru nanoparticles confined in Zr-con-
taining spherical mesoporous silica containers for hydrogenation of levulinic acid
and its esters into γ-valerolactone at ambient conditions, Catal. Today 258 (2015)
262–269.
catalysts: the effect of the water content of the nickel precursor on the size and
nature of the resulting particles, Kinet. Catal. 57 (3) (2016) 388–393.
32] M. Popova, P. Djinovic, A. Ristic, H. Lazarova, G. Drazic, A. Pintar, A.M. Balu,
N. Novak Tusar, Vapor-phase hydrogenation of levulinic acid to gamma-valer-
olactone over Bi-Functional Ni/HZSM-5 catalyst, Front. Chem. 6 (2018) 285.
33] F. Li, L.J. France, Z. Cai, Y. Li, S. Liu, H. Lou, J. Long, X. Li, Catalytic transfer
hydrogenation of butyl levulinate to γ-valerolactone over zirconium phosphates
with adjustable Lewis and Brønsted acid sites, Appl. Catal. B 214 (2017) 67–77.
34] X. Tang, H. Chen, L. Hu, W. Hao, Y. Sun, X. Zeng, L. Lin, S. Liu, Conversion of
biomass to γ-valerolactone by catalytic transfer hydrogenation of ethyl levulinate
over metal hydroxides, Appl. Catal. B 147 (2014) 827–834.
35] A.H. Valekar, K.-H. Cho, S.K. Chitale, D.-Y. Hong, G.-Y. Cha, U.H. Lee, D.W. Hwang,
C. Serre, J.-S. Chang, Y.K. Hwang, Catalytic transfer hydrogenation of ethyl levu-
linate to γ-valerolactone over zirconium-based metal–organic frameworks, Green
Chem. 18 (2016) 4542–4552.
[42] M. Ma, H. Liu, J. Cao, P. Hou, J. Huang, X. Xu, H. Yue, G. Tian, S. Feng, A highly
efficient Cu/AlOOH catalyst obtained by in situ reduction: catalytic transfer hy-
drogenation of ML into γ-GVL, Mol. Catal. 467 (2019) 52–60.
[
36] H. Li, Z. Fang, S. Yang, Direct conversion of sugars and ethyl levulinate into γ-
Valerolactone with superparamagnetic acid–base bifunctional ZrFeOx nanocata-
lysts, ACS Sustain. Chem. Eng. 4 (2015) 236–246.
[43] R. Wang, J. Wang, H. Zi, Y. Xia, H. Wang, X. Liu, Catalytic transfer hydrogenation of
ethyl levulinate to γ-valerolactone over zirconium (IV) Schiff base complexes on
mesoporous silica with isopropanol as hydrogen source, Mol. Catal. 441 (2017)
168–178.
[37] J. He, H. Li, Y.-M. Lu, Y.-X. Liu, Z.-B. Wu, D.-Y. Hu, S. Yang, Cascade catalytic
transfer hydrogenation–cyclization of ethyl levulinate to γ -valerolactone with
9