Catalytic Transfer Hydrogenation of Ethyl Levulinate to γ-Valerolactone Over Ni Supported on Equilibrium Fluid-Catalytic-Cracking Catalysts

Article abstract of DOI:10.1007/s10562-020-03326-5

Nickel supported on equilibrium fluid-catalytic-cracking catalysts (Ni/E-cats) were prepared by a simple grinding-pyrolysis method and employed for the transfer hydrogenation of ethyl levulinate (EL) to γ-valerolactone (GVL). 96.2% selectivity of GVL and 90.3% conversion of EL were obtained at 180?°C for 6?h over 30-Ni/E-cat. Through XRD, N₂ adsorption–desorption, NH₃-TPD and SEM analysis, the high activity of the 30-Ni/E-cat catalyst was attributed to its dispersed Ni metal active centers and available acidic sites. Catalytic probe test revealed that metal and acid sites of Ni/E-cat played a synergistic catalytic role in the synthesis of GVL in 2-propanol, where Ni metal sites contribute to the hydrogenation of ketone group in EL, and acid sites of E-cat promoted the lactonization of intermediate ethyl- or isopropyl 4-hydroxyvalerate. Two reaction pathways and synergistic mechanism were proposed in this catalytic system. Moreover, Ni/E-cat catalyst exhibited good stability up to four cycles without obvious loss of catalytic activity. Graphic Abstract: [Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-020-03326-5

Catalysis Letters
https://doi.org/10.1007/s10562-020-03326-5
Catalytic Transfer Hydrogenation of Ethyl Levulinate
to γ‑Valerolactone Over Ni Supported on Equilibrium
Fluid‑Catalytic‑Cracking Catalysts
1
1
1
1
1
1
1
Han Chen · Qiong Xu · Hui Li · Jian Liu · Xianxiang Liu · Geng Huang · Dulin Yin
Received: 5 March 2020 / Accepted: 20 July 2020
©
Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Nickel supported on equilibrium fuid-catalytic-cracking catalysts (Ni/E-cats) were prepared by a simple grinding-pyrolysis
method and employed ꢀor the transꢀer hydrogenation oꢀ ethyl levulinate (EL) to γ-valerolactone (GVL). 96.2% selectivity oꢀ
GVL and 90.3% conversion oꢀ EL were obtained at 180 °C ꢀor 6 h over 30-Ni/E-cat. Through XRD, N adsorption–desorp-
2
tion, NH -TPD and SEM analysis, the high activity oꢀ the 30-Ni/E-cat catalyst was attributed to its dispersed Ni metal active
3
centers and available acidic sites. Catalytic probe test revealed that metal and acid sites oꢀ Ni/E-cat played a synergistic
catalytic role in the synthesis oꢀ GVL in 2-propanol, where Ni metal sites contribute to the hydrogenation oꢀ ketone group
in EL, and acid sites oꢀ E-cat promoted the lactonization oꢀ intermediate ethyl- or isopropyl 4-hydroxyvalerate. Two reaction
pathways and synergistic mechanism were proposed in this catalytic system. Moreover, Ni/E-cat catalyst exhibited good
stability up to ꢀour cycles without obvious loss oꢀ catalytic activity.
Graphic Abstract
Keywords Ethyl levulinate · γ-Valerolactone · Transꢀer hydrogenation · Nickel · Equilibrium fuid-catalytic-cracking
catalyst
Electronic supplementary material The online version oꢀ this
article (https://doi.org/10.1007/s10562-020-03326-5) contains
supplementary material, which is available to authorized users.
Extended author inꢀormation available on the last page oꢀ the article
Vol.:(0
1
123456789)
3

H. Chen et al.
1
Introduction
catalyst (E-cat), were used as cheap acid supports to pre-
pare E-cat supported Ni catalysts (Ni/E-cat). The prepara-
tion process was simpliꢁed into two steps oꢀ grinding and
pyrolysis, and the selꢀ-reduction oꢀ metal was achieved
during the pyrolysis process under nitrogen. The catalytic
perꢀormance oꢀ the Ni/E-cat catalysts was tested in the
CTH oꢀ EL to GVL using 2-propanol as a hydrogen donor,
and the eꢃects oꢀ various parameters including catalyst
dosage, reaction temperature, reaction time, diꢃerent sol-
vents and substrates were also investigated. The two reac-
tion pathways in the reaction and the synergistic catalytic
mechanism oꢀ the metal and acid sites in this catalytic
system were ꢀurther proposed through the study oꢀ reaction
process. This work not only provides a new insight ꢀor the
rational design oꢀ highly active Ni-based catalysts ꢀor the
synthesis oꢀ GVL, but also provides an environmentally
ꢀriendly approach ꢀor disposing spent FCC catalysts.
Biomass is the abundant, cheap and renewable organic car-
bon resource on the earth [1]. The utilization oꢀ renewable
biomass resources to produce ꢀuels and ꢁne chemicals is
oꢀ great signiꢁcance to alleviate the current energy crisis
[
2, 3]. γ-Valerolactone (GVL), as an important biomass-
derived platꢀorm molecule, is widely used as ꢀuel additives
[
4, 5], ꢀood additives, green solvents [6, 7] and other valu-
able chemical intermediates [8, 9].
The hydrogenation-lactonization oꢀ biomass-derived
levulinic acid (LA) or its esters to GVL is the key reaction
in the conversion oꢀ biomass to valuable compounds (e.g.
gasoline, polyester, etc.). Noble metals, such as Pd [10],
Ru [11], Pt [12], exhibited excellent catalytic activities
in the production oꢀ GVL ꢀrom LA using H as hydrogen
2
resources. However, the use oꢀ noble metal catalysts and
high-pressure hydrogen are not preꢀerable ꢀrom an eco-
nomical viewpoint. Researchers have turned their attention
to non-noble metal catalysts (e.g. Ni [13], Cu [14], Co
2
Experimental
[
15]) and hope to explore an eꢂcient and economical route
to synthesize GVL. Among them, Ni catalysts showed a
superior activity. Yao Fu’s group [13] reported that Raney
Ni could eꢂciently catalyze ethyl levulinate (EL) to syn-
thesize GVL through a catalytic transꢀer hydrogenation
2
.1 Materials
All the reagents used in this study were analytical grade and
used as received without ꢀurther puriꢁcation. Ethyl levuli-
nate (EL), methyl levulinate (ML), levulinic acid (LA), butyl
levulinate (BL), γ-valerolactone (GVL), n-dodecane and
(
CTH) process under mild conditions. Unꢀortunately,
the supported Ni catalysts (Ni/C, Ni/TiO , Ni/SiO Ni/
2
2,
CeO ) prepared by them showed poor catalytic activity.
2
2
-butanol (2-BuOH) were purchased ꢀrom Aladdin Industrial
Supports with acid sites, such as zeolite, Al O or ZrO ,
2
3
2
Corporation (Shanghai, China). 2-Propanol (2-PrOH) was
purchased ꢀrom Hunan Huihong Reagent Co., Ltd (Hunan,
China). C H O ·H O and Ni(NO ) ·6H O were purchased
have been reported to be helpꢀul in the CTH reactions
[
16–18]. Mohan et al. ꢀound that the Ni/H-ZSM-5 catalyst
6
8
7
2
3 2
2
−
1
−1
had a high productivity (0.909kg
kg
h ) ꢀor
GVL·
catalyst
ꢀ
rom Shanghai Wokai Biotechnology Co., Ltd. (Shanghai,
the hydrogenation oꢀ LA to GVL at 250 °C [16]. Cai et al.
China). Equilibrium catalytic cracking catalysts (E-cat) are
waste Y-zeolite catalysts produced by the catalytic cracking
unit oꢀ Sinopec Changling Branch.
reported the 10Cu-5Ni/Al O was an eꢂcient catalyst ꢀor
2
3
the CTH oꢀ EL to GVL using 2-butanol as a hydrogen
donor, yielding a 97% yield oꢀ GVL in 12 h at 150 °C
[
17]. Sakakibara et al. developed the Ni/ZrO catalyst ꢀor
2
2.2 Catalysts Preparation
the CTH oꢀ methyl levulinate to GVL using 2-propanol
as a hydrogen donor, and achieved a GVL yield oꢀ 94%
in 20 h at 100 °C [18]. The excellent perꢀormance oꢀ the
above catalysts was due to the synergistic eꢃect oꢀ the
metal active center and the acidic carrier. Although these
Ni-based catalysts are eꢂcient ꢀor the production oꢀ GVL
Ni/E-cat catalysts with diꢃerent Ni loadings were prepared
by a simple grinding-pyrolysis method [21]. A typical cata-
lyst preparing procedure was as ꢀollows: Ni(NO ) ·6H O
3 2
2
(
10 mmol, 2.91 g) and C H O ·H O (5 mmol, 1.05 g) were
6 8 7 2
mixed and ground at room temperature ꢀor 0.5 h, then the
required amount oꢀ E-cat was added and continued grinding
ꢀ
rom levulinic acid or its esters, the preparation process oꢀ
the catalysts is complicated and the key step oꢀ H reduc-
2
ꢀ
or 0.5 h. The mixture changed ꢀrom solid to homogeneous
tion can’t be avoided, which is not economical ꢀor indus-
trial development. Thereꢀore, we plan to use cheap raw
materials to prepare an economical and eꢂcient catalyst
through a simple method.
green paste precursor. The precursor was dried at 120 °C ꢀor
4
3
h in air and calcined ꢀor 3 h at 370 °C with ramp rate oꢀ
°C/min under nitrogen atmosphere. Those prepared cata-
lysts were described as w-Ni/E-cats (where w represents the
loading oꢀ Ni in wt.%). Ni catalyst was prepared by the same
method without adding E-cat. All catalysts were stored in
vacuum desiccator and used without ꢀurther pretreatment.
Hence, in this work, waste zeolite catalysts generated
ꢀ
rom the fuid-catalytic-cracking (FCC) process in oil
reꢁneries [19, 20], usually called the equilibrium FCC
1
3

Catalytic Transfer Hydrogenation of Ethyl Levulinate to γ‑Valerolactone Over Ni Supported…
2.3 Catalyst Characterization
2.5 Sample Analysis
The X-ray ꢀuorescence (XRF) analysis was conducted on
a PW2400 spectrometer (Philips, Netherlands) using Rh
Kα radiation at 60 kV, 40 mA. Crystal structures oꢀ the
catalysts were determined by X-ray diꢃraction (XRD)
spectroscopy using an Ultima-IVX-ray powder diꢃrac-
tometer with Cu Kα radiation (λ = 1.5418 Å). Samples
were scanned in the 2θ range oꢀ 5 ~ 80° at 40 kV and
The reaction mixture was analyzed by a gas chromato-
graph (Shimadzu GC 2014, Japan) with HP-5 column
(30.0 m × 0.50 mm × 0.32 μm) and conꢀirmed by a gas
chromatograph-mass spectrometer combination (Shimadzu
GCMS-QP2010, Japan). The ꢀollowing temperature program
was used in the analysis: 55 °C (keeping ꢀor 3 min) then
increased at a rate oꢀ 15 °C/min to 190 °C (held ꢀor 1 min).
n-Dodecane was used as an internal standard in GC analysis.
EL conversion, GVL selectivity and GVL ꢀormation rate
were calculated using the ꢀollowing equations:
4
0 mA with a scan speed oꢀ 20°/min. Surꢀace area, pore
volume and pore size oꢀ the catalysts were obtained ꢀrom
N adsorption–desorption isotherms measured at − 196 °C
2
(
TriStar 3000, Micromeritics). Surꢀace area oꢀ the cata-
ꢀ
ꢁ
Mole of EL
lysts was determined by the BET equation whereas the
pore size was measured by the BJH method. Acid amount
oꢀ catalysts were detected by ammonia temperature-pro-
EL conversion (%) = 1 −
× 100%
Initial mole of EL
(
1)
grammed desorption (NH -TPD) on a AutoChem II 2920
Mole of GVL
3
GVL selectivity (%) =
× 100%
chemical adsorber (Micromeritics, USA) and the desorbed
Initial mole of EL − Mole of EL
(
2)
NH was detected by Thermal Conductivity Detector
3
(
TCD). Fourier transꢀorms inꢀrared spectroscopy (FT-IR)
GVL formation rate (mmol∕g∕h)
analysis was perꢀormed on a Perkin Elmer 283 spectrom-
Mole of GVL
(3)
−
1
=
eter, and the FT-IR spectra between 4000 and 400 cm
Amount of catalyst × Reaction time
were recorded. Thermogravimetric and derivative thermo-
gravimetric (TG-DTG) experiments were obtained on a
Netzsch Model STA 409PC instrument. The heating rate
ꢀ
rom room temperature to 800 °C was 10 °C/min using
α-Al O as the standard material. The surꢀace morphol-
3 Results and Discussion
2
3
ogy and composition oꢀ the catalyst were characterized by
scanning electron microscopy (SEM, USA, Quanta600F)
and X-ray energy spectrometer (EDX, UK, IE350). The
operating voltage is 20 kV, the operating current is 5 mA,
the working distance is 10 mm, the sample is pasted on the
sample table with conductive adhesive, and the surꢀace is
sprayed with gold to perꢀorm photographing and energy
spectrum analysis.
3.1 Catalyst Characterization
The main chemical composition oꢀ E-cat and ꢀresh FCC
catalyst (F-cat) was detected by XRF. The results are shown
in Table S1. The high silicon and aluminum content is due
to the structure oꢀ the FCC catalyst itselꢀ, containing Y-type
zeolite dispersed in a silica-alumina-clay matrix, which is
the active component oꢀ the FCC catalyst [22]. From the
test results, it can be seen that the FCC catalyst used in this
work is a Y-type zeolite with lanthanum-cerium rare earth
ion exchange. Without unexpectedly, many transition-metal
elements (Fe, Ni, V) were also detected in E-cat, which were
the elements that caused catalyst poisoning during the cata-
lytic cracking process [23, 24].
2.4 Catalytic Conversion of EL to GVL
Catalytic conversion oꢀ EL to GVL was perꢀormed in a
1
00 mL stainless steel autoclave with Tefon lined reactor
under magnetic stirring. In a typical experiment, the reac-
tor was loaded with ethyl levulinate (5 mmol), 2-propanol
The XRD patterns oꢀ E-cat and w-Ni/E-cats are shown
in Fig. 1. The refections at 2θ=5.6~30.8 and 66.8° were
observed in the patterns oꢀ all samples, indicating that the
Y-zeolite structure oꢀ E-cat remained intact aꢀter the load-
ing oꢀ Ni. All the supported catalysts showed typical XRD
refections at 44.49° (111), 51.85° (200) and 76.38° (220)
corresponding to the metallic Ni phase with a ꢀace-cen-
tered cubic geometry (ICDD. No. 04-002-3695). Moreo-
ver, additional diꢃraction peaks at 37.24° (111), 43.28°
(200) and 62.85° (220) were also observed in w-Ni/E-cats,
(
(
19 mL), n-dodecane (1 mmol) and the Ni/E-cat catalyst
216 mg), then purged ꢁve times with N at atmospheric
2
conditions beꢀore the reaction mixture was heated to the
prescribed temperature ꢀor a desired reaction time with
stirring at 500 rpm. Aꢀter the reaction, the reactor was
cooled to room temperature, and the reaction mixture was
magnetically separated ꢀrom the catalyst. Autoclave con-
taining the separated catalysts was evacuated and reꢁlled
with ꢀresh EL solution ꢀor subsequent catalytic cycles.
1
3

H. Chen et al.
channels oꢀ the E-cat. The N adsorption isotherms oꢀ the
2
catalysts are shown in Fig. S1. The change in the isotherms
oꢀ the w-Ni/E-cats catalysts is due to the strain oꢀ the Ni
plugging pore on the zeolite structure [16].
The surꢀace morphologies and compositions oꢀ the pre-
pared 30-Ni/E-cat catalyst were examed by SEM meas-
urements (Fig. 2). Micron-sized spherical and irregularly
shaped particles were observed. Grooves and cracks were
also observed in the sample. The distribution oꢀ C, O, Al, Si
and Ni elements in the catalyst was probed via SEM map-
ping. It was ꢀound that all these elements C, Ni and O, Al,
Si (the composition elements oꢀ the molecular sieve) were
uniꢀormly dispersed throughout the prepared 30-Ni/E-cat.
The ꢀunctional groups oꢀ E-cat and 30-Ni/E-cat were
identiꢁed by FT-IR analysis, and the corresponding FT-IR
spectra were depicted in Fig. 3. In the FT-IR spectrum
−
1
−1
oꢀ E-cat, the broad peaks at 3440 cm and 1625 cm
Fig. 1 XRD patterns oꢀ the catalysts a E-cat, b 5-Ni/E-cat, c 10-Ni/E-
are assigned to the -OH bending vibration oꢀ physically
adsorbed water, and the characteristic absorption peaks in
cat, d 20-Ni/E-cat, e 30-Ni/E-cat, ꢀ 35-Ni/E-cat and g 40-Ni/E-cat
−
1
the range oꢀ 1380~490 cm belong to the silicon-oxygen
tetrahedron, aluminum-oxygen tetrahedron and Si–O–Al
structure in the Y-type molecular sieve. Compared with
corresponding to the NiO phase (ICDD. No. 97-002-8834)
[
25, 26]. The intensity oꢀ these peaks increases as the Ni
−
1
loading oꢀ the catalyst increases. The presence oꢀ NiO phase
in the catalysts suggests that the Ni-citric acid complex pre-
cursor wasn’t completely reduced during the calcination
E-cat, a new peak at 456 cm appears in the FT-IR spec-
trum oꢀ the 30-Ni/E-cat catalyst, which belongs to the Ni–O
bending vibration. This indicates that there is a small amount
oꢀ NiO in the 30-Ni/E-cat catalyst.
[
27].
Table 1 summarizes the texture properties oꢀ F-cat, E-cat
The thermogravimetric analysis results oꢀ E-cat and
30-Ni/E-cat are shown in Fig. 4. The slight loss below
200 °C is the removal oꢀ physically adsorbed water, indi-
cating that the carrier E-cat is stable. For the 30-Ni/E-cat
catalyst, during the entire heating process, the weight loss
is divided into three stages, with a weight loss oꢀ 6%. Below
200 °C, this corresponds to the loss oꢀ absorbed water. A
very small weight loss between 200 to 380 °C may be the
decomposition oꢀ carbon deposits [29]. Combined with the
previous characterization, the weight loss oꢀ the catalyst at
400 to 500 °C was due to the decomposition oꢀ the residual
Ni-citric acid precursor during calcination [27]. The low
weight loss indicates that there are ꢀew undecomposed
compounds.
and w-Ni/E-cats. The data indicates that all catalysts have
mesoporous and microporous structures. It can be clearly
seen that the speciꢁc area and the pore volume oꢀ E-cat
are much smaller than those oꢀ F-cat. This is because the
deposited impurities ꢀorm some substances destructive to the
structure oꢀ the molecular sieve at high temperature during
the petroleum reꢁning process, resulting in an irreversible
change in the structure oꢀ the E-cat catalyst [28]. Moreover,
the ꢀormation oꢀ coke covers the surꢀace oꢀ the catalyst and
blocks the pores oꢀ the catalyst [29]. Compared with E-cat,
the surꢀace areas and pore parameters oꢀ w-Ni/E-cats cata-
lysts decrease with the increase oꢀ Ni loading. This implies
that nickel metallic particles not only cover the surꢀace oꢀ
the E-cat carrier, but also are incorporated into the pore
Table 1 Surꢀace areas and pore parameters oꢀ the catalysts
Entry
Catalyst
S
S
V
V
3
D
average pore diameter
BET surꢀace area
−1
micropore surꢀace area
total pore volume
micropore pore volume
2
2
−1
3
−1
−1
(
m g
)
(m g
)
(cm g
)
(cm g
)
(nm)
1
2
3
4
5
6
F-cat
E-cat
10-Ni/ E-cat
20-Ni/ E-cat
30-Ni/ E-cat
40-Ni/ E-cat
243.1
100.3
98.2
97.3
93.5
87.9
134.2
54.8
25.6
21.3
17.2
14.9
0.194
0.152
0.146
0.140
0.138
0.139
0.0546
0.0221
0.0134
0.0106
0.0085
0.0073
12.35
10.84
7.58
7.05
6.84
6.41
1
3

Catalytic Transfer Hydrogenation of Ethyl Levulinate to γ‑Valerolactone Over Ni Supported…
Fig. 2 SEM images with the corresponding elemental mapping (C, O, Al, Si and Ni) and EDX patterns oꢀ 30-Ni/E-cat catalyst
oꢀ the w-Ni/E-cat catalysts, and the amount oꢀ acid increases
with the increase oꢀ the Ni loading. Combined with the TG
results, a small amount oꢀ residual Ni-citric acid complex
precursor may be decomposed during the TPD process at
3
80~500 °C. Thereꢀore, the data oꢀ strong acidity and total
acidity oꢀ w-Ni/E-cats catalysts are larger. In addition, the
presence oꢀ a small amount oꢀ NiO in w-Ni/E-cats will
also lead to an increase in acidity [24]. Among them, the
3
0-Ni/E-cat catalyst showed a higher concentration oꢀ acidic
sites (1.242 mmol/g), which could play an important role in
the CTH oꢀ EL to GVL.
3
.2 Catalytic Performance of W‑Ni/E‑Cat With
Diꢀerent Nickel Loading
The catalytic perꢀormance oꢀ the w-Ni/E-cat catalysts was
evaluated in the CTH oꢀ EL to GVL, and the results are listed
in Table 2. Only a little transesteriꢁed product, isopropyl
levulinate (IPL), was detected in the reaction catalyzed by
E-cat, showing the acid-catalytic activity oꢀ E-cat (Table 2,
entry 1). In the w-Ni/E-cat catalyzed reaction, three interme-
diates ethyl 4-hydroxypentanoate (4-HPE), IPL and isopro-
pyl 4-hydroxyvalerate (4-HPI) as well as the main product
GVL were detected. As Ni loading increased in the catalysts,
Fig. 3 FT-IR spectra oꢀ E-cat and 30-Ni/E-cat
The acid strength distribution and total acidity oꢀ E-cat
and w-Ni/E-cats were measured by NH -TPD, as shown in
3
Fig. S2 and Table S2. The carrier E-cat possesses weak and
moderate acid sites. Strong acid sites appear in the spectrum
1
3

H. Chen et al.
Fig. 4 TG-DTG curves oꢀ E-cat and 30-Ni/E-cat
Table 2 Eꢃect oꢀ Ni loading on
Entry
Catalysts
EL conv (%)
Sel. (%)
GVL
w-Ni/E-cats on the CTH oꢀ EL
to GVL
4-HPE
IPL
4-HPI
1
2
3
4
5
6
7
E-cat
3.2
7.3
9.5
18.2
52.8
25.7
8.2
–
–
2.8
10.4
33.6
16.1
13.8
5.8
100
53.4
27.6
7.1
4.5
11.0
38.4
–
–
0.9
1.2
1.1
1.4
1.9
5-Ni/E-cat
10-Ni/E-cat
20-Ni/E-cat
30-Ni/E-cat
35-Ni/E-cat
40-Ni/E-cat
43.8
61.1
58.1
78.3
73.8
53.9
Reaction conditions: 5 mmol EL, 19 mL 2-propanol, 50 mg catalyst, 1 mmol n-dodecane, 180 °C, 4 h
EL conversion and GVL selectivity gradually increased,
while the selectivity oꢀ IPL reduced rapidly. The selectivity
oꢀ the intermediate ethyl 4-hydroxypentanoate (4-HPE) and
isopropyl 4-hydroxypentanoate (4-HPI) increased slightly
(
Table 2, entries 2–4). This implies that there are two diꢀ-
erent pathways in the reaction catalyzed by metal–acid
ꢀ
bi-ꢀunctional catalysts. More details about the pathway are
ꢀ
urther discussed in Section 3.6. The 30-Ni/E-cat catalyst
showed superior catalytic perꢀormance in the hydrogenation
reaction, providing an EL conversion oꢀ 52.8% and a GVL
selectivity oꢀ 78.3% (Table 2, entry 5). However, an exces-
sive loading oꢀ Ni led to a decrease in EL conversion and
GVL selectivity (Table 2, entries 6–7). This was because
ꢀ
eeding more nickel nitrate during the preparation process
might cause the agglomeration oꢀ Ni particles and reduce the
dispersion oꢀ Ni active centers, thereby reducing the activity
oꢀ the catalyst. These results indicate that the combination
oꢀ Ni and E-cat is a promising catalyst with an optimal Ni
loading oꢀ 30%.
Fig. 5 Eꢃect oꢀ catalyst dosage on the CTH oꢀ EL to GVL. Reac-
tion conditions: 5 mmol EL, 19 mL 2-propanol, 30-Ni/E-cat, 1 mmol
n-dodecane, 180 °C, and 4 h
1
3

Catalytic Transfer Hydrogenation of Ethyl Levulinate to γ‑Valerolactone Over Ni Supported…
3
.3 Eꢀect of Catalyst Dosage
selectivity continuously increased until a maximum GVL
selectivity oꢀ 94.1% was obtained at 180 °C. This was pos-
sibly attributed to the increase oꢀ the reaction rate at higher
reaction temperature. Further increasing the reaction tem-
perature to 190 °C led to a slight decrease oꢀ GVL selec-
tivity. It was likely that the product might undergo some
side reactions at higher reaction temperature. Meanwhile,
the reaction was also traced to investigate the infuence oꢀ
the reaction time on the CTH reaction (Fig. 6b). It is worth
noting that when reaction time ꢀrom 2 h was prolonged to
12 h, the conversion oꢀ EL increased greatly ꢀrom 55.8% to
95.2%, but the selectivity oꢀ GVL changed little (91.1% to
97.4%). Correspondingly, the selectivity oꢀ the three inter-
mediate products was less than 5%, and gradually decreased
with the extension oꢀ time. This indicates that enough acid
sites on the carrier promote the rapid conversion oꢀ inter-
mediate products to GVL, and the hydrogenation oꢀ EL is
the rate-limiting step in the reaction. The conversion oꢀ EL
was increased by only 4.9% when the time increased ꢀrom
6 to 12 h. Those observations suggest that the reaction time
beyond 6 h hardly aꢃects the EL conversion. Based on the
above experiments, the optimal reaction temperature oꢀ
The eꢃect oꢀ catalyst dosage on the CTH oꢀ EL to GVL was
studied and results are presented in Fig. 5. Those experi-
ments were carried out by changing the dosage oꢀ 30-Ni/E-
cat ꢀrom 50 to 288 mg while keeping the reaction temper-
ature at 180 °C and the reaction time ꢀor 4 h. Obviously,
both EL conversion and GVL selectivity were increased and
the selectivity oꢀ the intermediate product 4-HPE gradu-
ally decreased with increasing the catalyst dosage, which
should be mainly attributed to more available number oꢀ
catalytically active sites. In the presence oꢀ large amount
oꢀ metal catalytic sites, the transesteriꢁed product IPL and
intermediate 4-HPI were very little. When the catalyst dos-
age was 216 mg, a GVL selectivity oꢀ 94.1% was obtained
at an EL conversion oꢀ 87.7%. However, a ꢀurther increase
oꢀ catalyst dosage (288 mg) led to the EL conversion slightly
decreased to 84.5% although the selectivity oꢀ GVL was
still increased, which possibly attributed to excessive metal
catalysts hindering contact with the substrate. Thereꢀore,
2
16 mg was selected as the optimal catalyst dosage ꢀor the
subsequent experiments.
1
80 °C and reaction time oꢀ 6 h were suꢂcient ꢀor the CTH
3.4 Eꢀect of Reaction Temperature and Time
reaction oꢀ EL to achieve high GVL yield.
The reaction temperature has an important eꢃect on the reac-
tion rate and products distribution. Thus, the CTH oꢀ EL
was carried out at diꢃerent reaction temperatures, ranging
3.5 Eꢀect of Diꢀerent Alcohols and Substrates
In the CTH reaction, alcohols are not only the solvent but
H-donors. 1-propanol (1-PrOH), 2-propanol (2-PrOH),
2-butanol (2-BuOH) and cyclohexanol (CyOH) were
chosen as the solvent and H-donors in this reaction. As
shown in Table 3, secondary alcohols provided high EL
ꢀ
rom 160 to 190 °C (Fig. 6a). When the reaction proceeded
at a milder condition (160 °C), 55.3% conversion oꢀ EL and
6.8% selectivity oꢀ GVL were achieved. With the increase
oꢀ the reaction temperature, the EL conversion and GVL
6
Fig. 6 Eꢃect oꢀ reaction temperature (a) and reaction time (b) on the CTH oꢀ EL to GVL. Reaction conditions: 5 mmol EL, 19 mL 2-propanol,
16 mg 30-Ni/E-cat, 1 mmol n-dodecane, a 4 h; b 180 °C
2
1
3

H. Chen et al.
Table 3 Catalytic ꢀeasibility oꢀ 30-Ni/E-cat in diꢃerent alcohols and
indicates that the bi-ꢀunctional catalyst has a promising
potential ꢀor converting biomass derivatives into GVL.
substrates
Entry Substrate Solvent Conversion GVL
%)
GVL ꢀor-
(
selectivity mation rate
3.6 The Catalytic Roles of Ni and E‑cat in Ni/E‑cat
Catalyst
(
%)
(mmol/g/h)
1
2
3
4
5
6
7
EL
EL
EL
EL
ML
BL
LA
1-PrOH 65.8
2-PrOH 90.3
2-BuOH 76.6
11.9
96.2
90.1
12.6
92.5
95.6
91.4
0.30
3.35
2.66
0.15
3.09
3.54
3.18
To explore the role oꢀ Ni and E-cat in the CTH reaction,
Ni, physically mixed Ni and E-cat, ꢀresh and poisoned
3
0-Ni/E-cat (adding pyridine or thiol) were used in the
CyOH
30.3
reaction to compare their catalytic activities. Results are
given in Table 4. Comparing the results oꢀ entries 1–3,
monometallic Ni showed poor catalytic perꢀormance, pro-
viding only 7.8% conversion oꢀ EL and 39.9% selectivity oꢀ
GVL. However, the selectivity oꢀ the intermediate 4-HPE
was high, reaching 60.1% (Table 4, entry 1). It was interest-
ing that once the co-operation oꢀ nickel with acidic E-cat,
whether a physically mixed Ni+E-cat (Table 4, entry 2) or
a co-pyrolysis Ni/E-cat (Table 4, entry 3), both exhibited
enhanced catalytic perꢀormance in the reaction. The selec-
tivity oꢀ GVL on Ni/E-cat or Ni + E-cat was substantially
higher than that on sole Ni, which may beneꢁt ꢀrom the
inner transesteriꢁcation oꢀ the intermediate 4-HPE pro-
moted by the surꢀace acidic sites oꢀ E-cat [18, 28]. When
the poison probe thiol was introduced into the reaction
system, EL conversion and GVL ꢀormation rate decreased
sharply, but the selectivity oꢀ the acid-catalytic product IPL
was improved (Table 4, entry 4). This is because the Ni
metal sites oꢀ Ni/E-cat are poisoned, weakening the hydro-
genation ꢀunction oꢀ the catalyst. Besides, no 4-HPI was
detected in this reaction, indicating that the ꢀormation oꢀ
4-HPI is not an acid-catalyzed reaction but a hydrogenation
reaction. The little 4-HPI (in Table 4, entries 2, 3 and 5)
can be the hydrogenation product oꢀ IPL catalyzed by Ni
sites. Adding pyridine as an alkaline probe molecule to the
reaction, the acidic sites on Ni/E-cat were almost poisoned,
2-PrOH 86.7
2-PrOH 95.9
2-PrOH 90.2
Reaction conditions: 5 mmol EL, 19 mL solvent, 216 mg 30-Ni/E-
cat, 1 mmol n-dodecane, 180 °C, 6 h
conversion, GVL selectivity and GVL ꢀormation rate,
as opposed to primary alcohols. The hydrogen-donating
ability oꢀ alcohols decreased in the order oꢀ: 2-PrOH>2-
BuOH>1-PrOH>CyOH, which was consistent with many
previous reported works [13, 17, 30]. According to the lit-
erature report, the primary alcohols were poor H-donors in
the CTH reaction due to the diꢂcult β-hydride elimination
aꢀter it ꢀormed alkoxy species on the surꢀace oꢀ the catalyst
[
17]. CyOH was not an eꢃective H-donor in this reaction,
which might due to the strong steric hindrance eꢃect oꢀ the
solvent. Considering reactivity and practicality, 2-PrOH was
the best choice as a solvent and hydrogen source in our reac-
tion system.
In addition, other levulinate esters including Methyl lev-
ulinate (ML), butyl levulinate (BL) and levulinic acid (LA)
were also applied to the CTH reaction ꢀor the production oꢀ
GVL. High conversion and GVL selectivity were obtained
in the reactions catalyzed by the 30-Ni/E-cat catalyst, which
Table 4 Catalytic activities oꢀ
Entry
Catalyst
EL
Sel. (%)
GVL
GVL
diꢃerent catalysts with metal
conv
Formation
rate (mmol/
g/h)
a
and (or) acid sites
4-HPE
IPL
4-HPI
(
%)
b
1
2
3
4
Ni
7.8
39.9
81.6
78.3
68.9
42.2
60.1
13.1
16.1
14.3
54.3
–
3.4
4.5
16.8
1.3
–
1.9
1.1
–
2.59
9.98
10.32
1.09
c
Ni+E-cat
48.9
52.8
6.3
30-Ni/E-cat
thiol-30-Ni/E-cat
pyridine-30-Ni/E-cat
d
e
5
43.7
2.2
4.61
Reaction conditions:
a
5
mmol EL, 19 mL 2-propanol, 1mmoln-dodecane, 50 mg catalyst, 180 °C, 4 h
b
c
d
e
Ni was 15 mg
Ni+E-cat: physical-mixed Ni and E-cat (Ni was 15 mg and E-cat was 35 mg)
Adding thiol to the reaction system
Adding pyridine to the reaction system
1
3

Catalytic Transfer Hydrogenation of Ethyl Levulinate to γ‑Valerolactone Over Ni Supported…
resulting in lower GVL selectivity and higher 4-HPE selec-
tivity, which was close to the process oꢀ using monome-
tallic Ni (Table 4, entry 5). From the above experimental
results, it can be concluded that metallic Ni contributes to
the transꢀer hydrogenation oꢀ EL/IPL with 2-propanol, and
acidic E-cat promotes the lactonization oꢀ the hydrogenated
product 4-HPE/4-HPI to GVL.
EL undergoes a hydrogenation reaction to ꢀorm the inter-
mediate 4-HPE. Finally, the hydrogenated product 4-HPE
is easily cyclized on the acidic sites to produce the GVL.
3.8 Reuse Test
The recyclability oꢀ 30-Ni/E-cat was investigated in the
CTH oꢀ EL to GVL under the optimal reaction condition.
Aꢀter each run, the catalyst was separated ꢀrom the reac-
tion mixture by a magnet, washed with 2-propanol three
times, and then directly subjected to the next run with-
out adding any ꢀresh catalyst. The results are presented in
Fig. 7. The catalyst was used ꢀor ꢀour catalytic runs with
only a slight decrease in the EL conversion, and the GVL
selectivity kept well in the reaction.
3
.7 Propose Reaction Pathways and Synergistic
Mechanism
Based on the above results, we draw up two pathways ꢀor
the conversion oꢀ EL to GVL using 2-propanol as a hydro-
gen donor over Ni/E-cat catalysts (shown in Scheme 1).
The one is that EL was ꢁrstly hydrogenated to ꢀorm the
intermediate 4-HPE on the Ni sites, and subsequent 4-HPE
was cyclized by inner transesteriꢁcation on acid sites to
generate GVL. This is the major route in the CTH oꢀ EL
to GVL catalyzed by Ni/E-cat. The other is that EL ꢁrstly
underwent transesteriꢁcation with 2-propanol on the acid
sites to ꢀorm IPL, and IPL was hydrogenated to ꢀorm the
intermediate 4-HPI, which was then ꢀurther converted to
GVL on the acid sites.
To ꢁgure out the change oꢀ the catalyst during the reac-
tion, XRD and SEM were ꢀurther carried out to character-
ize the spent catalyst. Fig. S3 showed the XRD analysis
oꢀ the ꢀresh and spent 30-Ni/E-cat catalyst. The two pat-
terns showed similar diꢃraction peaks, which indicated
that the active metal sites on the surꢀace oꢀ the 30-Ni/E-cat
catalyst were almost unchanged aꢀter being recycled ꢀour
consecutive times. Fig. S4 showed some representative
SEM images oꢀ the ꢀresh and spent 30-Ni/E-cat catalyst.
Compared with the ꢀresh 30-Ni/E-cat (Fig. S4a, b), the
metal particles could still be well dispersed aꢀter the reac-
tion. However, the surꢀace oꢀ the used 30-Ni/E-cat became
smooth (Fig. S4c, d). This might be caused by the wear
between the magneton and the catalyst during the ther-
mal solution reaction, which might be unavoidable ꢀor
most supported FCC catalyst under such reaction condi-
tion. However, the reactivity oꢀ the catalyst remained at a
The synergistic mechanism based on the 4-HPE-medi-
ated major pathway is deduced. Firstly, 2-propanol and
EL are adsorbed on the metal sites (Ni) and acid sites on
the surꢀace oꢀ the catalyst, respectively. The adsorption oꢀ
2
-propanol breaks the C–H and O–H bonds to release ace-
tone, thus two H-atoms remained on the Ni surꢀace, which
is the active species ꢀor transꢀer hydrogenation. These two
H-atoms are transꢀerred to the ketone group oꢀ EL exist-
ing on the adjacent surꢀace site oꢀ the catalyst, and the
OH
O
4-HPE selectivity
EL conversion
O
IPL selectivity
OH
GVLselectivity
100
4-HPI selectivity
O
EL
Acid sites
O
O
O
<
Ni sites
8
6
0
0
O
IPL
HO
OH
minor
major
O
Ni sites
O
40
O
4
-HPE
<
OH
2
0
0
O
Acid sites
O
4
-HPI
Acid sites
OH
<
O
O
1
2
3
4
OH
Run times
GVL
Fig. 7 Reuse oꢀ 30-Ni/E-cat in the CTH oꢀ EL to GVL. Reaction
conditions: 5 mmol EL, 19 mL 2-propanol, catalyst dosage 216 mg,
1 mmol n-dodecane, 180 °C, 4 h
Scheme 1. The reaction mechanism ꢀor conversion oꢀ EL to GVL
over bi-ꢀunctional Ni/E-cat catalyst
1
3

H. Chen et al.
relatively high level despite some surꢀace wear. This result
urther conꢁrms that the 30-Ni/E-cat catalyst has a good
stability under the CTH reaction conditions.
5. Bruno TJ, Wolk A, Naydich A (2010) Energy Fuels 24:2758–2767
6
7
8
9
.
.
.
.
Wettstein SG, Alonso DM, Chong Y, Dumesic JA (2012) Energy
Environ Sci 5:8199
ꢀ
Mellmer MA, Martin Alonso D, Luterbacher JS, Gallo JMR,
Dumesic JA (2014) Green Chem 16:4659–4662
Du X-L, Bi Q-Y, Liu Y-M, Cao Y, He H-Y, Fan K-N (2012) Green
Chem 14:935–939
4
Conclusions
Al-Shaal MG, Dzierbinski A, Palkovits R (2014) Green Chem
1
6:1358–1364
In this work, we reported a robust and eꢂcient metal–acid bi-
unctional 30-Ni/E-cat catalyst ꢀor the transꢀer hydrogenation
10. Zhang Y, Chen C, Gong W, Song J, Zhang H, Zhang Y, Wang G,
Zhao H (2017) Catal Commun 93:10–14
ꢀ
1
1. Sudhakar M, Lakshmi Kantam M, Swarna Jaya V, Kishore
R, Ramanujachary KV, Venugopal A (2014) Catal Commun
oꢀ ethyl levulinate to γ-valerolactone. An excellent catalytic
perꢀormance with 96.2% GVL selectivity and 90.3% EL con-
version was obtained in the CTH reaction using 2-PrOH as
a H-donor at 180 °C ꢀor 6 h. Both Ni and the carrier E-cat
contributed signiꢁcantly to the excellent perꢀormance oꢀ Ni/E-
cat. Metallic nickel sites in the 30-Ni/E-cat largely contributed
to the hydrogenation oꢀ EL, and acid sites oꢀ E-cat played a
synergistic catalytic role by promoting the cyclization oꢀ the
intermediate 4-HPE. This is the dominant pathway mediated
by the intermediate 4-HPE. Besides, GVL was also produced
by another pathway oꢀ the intermediates IPL/4-HPI. We
believe that the highly eꢂcient and easily prepared Ni/E-cat
catalysts have great potential oꢀ application in hydrogenation
reactions, and the waste E-cat can also be reused to prepare
other eꢃective supported bi-ꢀunctional catalysts.
5
0:101–104
12. Upare PP, Lee J-M, Hwang DW, Halligudi SB, Hwang YK, Chang
J-S (2011) J Ind Eng Chem 17:287–292
1
3. Yang Z, Huang YB, Guo QX, Fu Y (2013) Chem Commun
(
Camb) 49:5328–5330
1
4. Yang Y, Xu X, Zou W, Yue H, Tian G, Feng S (2016) Catal Com-
mun 76:50–53
15. Zhou H, Song J, Fan H, Zhang B, Yang Y, Hu J, Zhu Q, Han B
(
2014) Green Chem 16:3870–3875
1
6. Mohan V, Raghavendra C, Pramod CV, Raju BD, Rama Rao KS
(
2014) RSC Adv 4:9660
17. Cai B, Zhou X-C, Miao Y-C, Luo J-Y, Pan H, Huang Y-B (2016)
ACS Sustain Chem Eng 5:1322–1331
1
1
8. Sakakibara K, Endo K, Osawa T (2019) Catal Commun 125:52–55
9. Bertero M, García JR, Falco M, Sedran U (2019) Renew Energy
1
32:11–18
20. Zhang D, Wen B, Chen Y, Xu Q, Zhu X, Sun N, Liu X, Yin D
(
2020) J Mater Cycles Waste Manag 22:22–29
Acknowledgements The authors grateꢀully acknowledge the ꢁnancial
support oꢀ the National Natural Science Foundation oꢀ China (Grant
No. 21776068, 21606082 and 21975070) and Scientiꢁc Research Fund
oꢀ Hunan Provincial Education Department (17C0951 and 15B050).
21. Venkatesha NJ, Ramesh S (2018) Ind Eng Chem Res
57:1506–1515
22. Liu H, Ma J, Gao X (2006) Catal Lett 110:229–234
23. Souza NLA, Tkach I, Morgado E, Krambrock K (2018) Appl
Catal A 560:206–214
2
4. Gerards RTJ, Fernandes A, Graça I, Ribeiro MF (2020) Fuel
Compliance with Ethical Standards
2
60:116372
2
5. Kumar VV, Naresh G, Sudhakar M, Anjaneyulu C, Bhargava SK,
Tardio J, Reddy VK, Padmasri AH, Venugopal A (2016) RSC Adv
Conflict of interest The authors declare that they have no conficts oꢀ
interest.
6
:9872–9879
2
6. Singh H, Iyengar N, Yadav R, Rai A, Sinha AK (2018) Sustain
Energy Fuels 2:1699–1706
2
2
2
7. Abu-Zied BM, Asiri AM (2017) Thermochim Acta 649:54–62
8. Cho SI, Jung KS, Woo SI (2001) Appl Catal B 33:249–261
9. Su B, Shi L, Meng X, Wang X, Liu N (2018) J Sulꢀur Chem
References
1
2
.
.
Corma A, Iborra S, Velty A (2007) Chem Rev 107:2411–2502
Li C, Zhao X, Wang A, Huber GW, Zhang T (2015) Chem Rev
4
0:75–87
3
0. He J, Li H, Lu Y-M, Liu Y-X, Wu Z-B, Hu D-Y, Yang S (2016)
Appl Catal A 510:11–19
1
15:11559–11624
3
4
.
.
Kunkes EL, Simonetti DA, West RM, Juan Carlos SR, Gärtner
CA, Dumesic JA (2008) Science 322:417–421
Horváth IT, Mehdi H, Fábos V, Boda L, Mika LT (2008) Green
Chem 10:238–242
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional aꢂliations.
Aꢀliations
1
1
1
1
1
1
1
Han Chen · Qiong Xu · Hui Li · Jian Liu · Xianxiang Liu · Geng Huang · Dulin Yin
1
*
*
Qiong Xu
National and Local Joint Engineering Laboratory ꢀor New
Petro-Chemical Materials and Fine Utilization oꢀ Resources,
College oꢀ Chemistry and Chemical Engineering, Hunan
Normal University, Changsha 410081, China
xuqiong@hunnu.edu.cn
Dulin Yin
dulinyin@126.com
1
3