Selective hydrogenation of furfuryl alcohol to tetrahydrofurfuryl alcohol over Ni/γ-Al2O3 catalysts

Article abstract of DOI:10.1007/s11164-016-2691-8

A series of nickel-based catalysts (with <5?nm Ni particle size) with γ-alumina as a support (x wt% Ni/γ-Al₂O₃, x represents the Ni loading amount) were synthesized by the impregnation method, which was successfully applied for the catalytic hydrogenation of furfuryl alcohol to tetrahydrofurfuryl alcohol. The effects of reaction time, reaction temperature, nickel loading amount, solvent, and hydrogen pressure on conversion of furfural alcohol as well as selectivity for tetrahydrofurfuryl alcohol were investigated systematically. The conversion of furfural alcohol over 15?wt% Ni/γ-Al₂O₃ was up to 99.8?% with a selectivity of 99.5?% toward tetrahydrofurfuryl alcohol, when the reaction was carried out at 353?K with an initial H₂ pressure of 4.0?MPa and reaction time of 2?h. In addition, there was an increase of turnover frequency (TOF) value with the decrease of Ni particle size. The features of the Ni/γ-Al₂O₃ catalysts were investigated by characterization of XRD, TPR, BET, and SEM.

Full text of DOI:10.1007/s11164-016-2691-8

Res Chem Intermed
DOI 10.1007/s11164-016-2691-8
Selective hydrogenation of furfuryl alcohol
to tetrahydrofurfuryl alcohol over Ni/c-Al2O3 catalysts
Shengya Sang1 Yuan Wang1 Wei Zhu1
Guomin Xiao1
•
•
•
Received: 14 April 2016 / Accepted: 5 August 2016
Ó Springer Science+Business Media Dordrecht 2016
Abstract A series of nickel-based catalysts (with \5 nm Ni particle size) with
c-alumina as a support (x wt% Ni/c-Al2O3, x represents the Ni loading amount)
were synthesized by the impregnation method, which was successfully applied for
the catalytic hydrogenation of furfuryl alcohol to tetrahydrofurfuryl alcohol. The
effects of reaction time, reaction temperature, nickel loading amount, solvent, and
hydrogen pressure on conversion of furfural alcohol as well as selectivity for
tetrahydrofurfuryl alcohol were investigated systematically. The conversion of
furfural alcohol over 15 wt% Ni/c-Al2O3 was up to 99.8 % with a selectivity of
99.5 % toward tetrahydrofurfuryl alcohol, when the reaction was carried out at
353 K with an initial H2 pressure of 4.0 MPa and reaction time of 2 h. In addition,
there was an increase of turnover frequency (TOF) value with the decrease of Ni
particle size. The features of the Ni/c-Al2O3 catalysts were investigated by char-
acterization of XRD, TPR, BET, and SEM.
Keywords Ni/c-Al2O3 Á Hydrogenation Á Furfuryl alcohol Á Tetrahydrofurfuryl
alcohol
Introduction
Recently, with the development of the economy and the decrease of petroleum
reserves, researchers have realized that renewable energy is of great importance for
sustainable development. Biomass was regarded as significant substitute goods for
the production of fuels and chemicals [1–4]. Furfural (FFR) is one kind of important
platform compound derived from biomass, which is contained in the hemicellulose
&
Guomin Xiao
xiaogm426@gmail.com
1
School of Chemistry and Chemical Engineering, Southeast University,
Nanjing 211189, Jiangsu, China
123

S. Sang et al.
part of lignocellulosic biomass [5–10]. The FFR contains both C=O and C=C double
bonds and with hydrogenation of FFR, it can produce a series of downstream
compounds such as furfuryl alcohol (FA), tetrahydrofurfural, and tetrahydrofurfuryl
alcohol (THFA) [11]. Yoshinao Nakagawa et al. [11–13] also reported about FFR
concerning the conversion of FFR to tetrahydrofurfuryl alcohol (THFA) with silica-
support nickel catalyst. In some reports, they also report about conversion of FFR to
1,5-pentanediol (1,5-PeD) [14, 15].
Furfuryl alcohol (FA) is a downstream product of furfural [16], which can be
obtained by selective hydrogenation of FFR, and the catalyst has been successfully
industrialized. Tetrahydrofurfuryl alcohol (THFA) is an important downstream
product of FA, which is an important organic solvent, and is widely used in the
industrial and agricultural fields [17, 18]. Because of its benign nature and low
toxicity, it is also widely used as resin in the industry. THFA can be also obtained
from the direct hydrogenation of FFR [19, 20]. Many other different useful
chemicals can be also produced from THFA, such as d-valerolactone [21, 22],
tetrahydrofuran [22–25], and 4-penten-1-ol [21, 26–28]. Koso et al. [29, 30] and
Nakagawa et al. [31–33] have reported that the conversion of THFA to 1,5-
pentanediol (1,5-PeD) with different support catalysts and the selectivity of 1,5-PeD
of up to 70 %. In their reports, two steps were reported about the conversion of
THFA to 1,5-PeD [34, 35]. A multi-step method was proposed by Schniepp’s [34]
team, but this method needs the isolation and purification of intermediates, and the
selectivity of 1,5-PeD was 70 %. A one-step method was proposed by Adkins’s
team, although this method can make the THFA direct to 1,5-PeD; however, the
selectivity of 1,5-PeD was only 20 %, with 1,2-PeD as the main product [35]. In
addition, some other authors also report about conversion of THFA to 1,5-
pentanediol (1,5-PeD) [36–40]. In this paper, our team mainly discussed that THFA
can be obtained by selective hydrogenation of FA in liquid phase (Scheme 1).
Some works have been done in catalytic conversion of FA to THFA. Winterle
et al. [41] and Tike et al. [42] have found that FA can be converted to THFA with
noble-catalysts (Pd, Pt and Rh) and has a high conversion and selectivity. The
selectivity toward THFA was over 97 % below 333 K with 0.68–4.08 MPa reaction
pressure. Considering the high price of noble-metals, Chen et al. [43] have used a
series of support Ni catalysts for catalytic hydrogenation of FA to THFA with a
reaction temperature within 433–453 K, reaction pressure 3–4 MPa, reaction time
of 8 h, and catalyst loading 20 g/L. The conversion of FA was nearly 99.9 % with a
selectivity of 98.3 % toward THFA. However, much work still needs to be done in
exploring low-cost catalysts with good catalytic performance under moderate
reaction conditions.
Nickel-base catalysts were widely used in hydrogenation, hydrodesulfurization,
and steam reforming for its various advantages such as low cost and easy
Scheme 1 Possible reaction path for hydrogenation of FA in liquid phase
123

Selective hydrogenation of furfuryl alcohol to…
preparation and high stability [44, 45]. c-alumina (c-Al2O3) was confirmed to be a
suitable support, owing to its intrinsic properties such as chemical inertness and
thermal stability. Due to Ni metal having a facile hydrogen activation, it has been
widely recognized as a promising candidate for hydrogenation. Sitthisa et al. [46]
reported Ni metal as more favorable for the ring opened products. The Kong et al.
[47] team reported that due to the acid sites adjacent to Ni metal, this leading to Ni
metal played an important role in the hydrogenation. In addition, in the process of
preparation, the high nickel is easily reduced to low nickel in the hydrogenation
reaction. All of this illustrates that Ni plays an important role in contributing to the
selectivity and conversion of FA and THFA.
In this paper, the impregnation method was chosen to prepare c-Al2O3-support
Ni catalysts, because the deposition–precipitation method was reported to affect the
catalytic performance of the metal particles [48]. We found that with the increase of
the Ni loading amount, the particle size also increased, and, when the Ni loading
amount was 15 wt%, the catalysts showed the best catalytic performance. The
catalytic hydrogenation of FA to THFA was improved by high Ni dispersal, which
also showed high turnover frequency (TOF) values based on the numerous surface
metal sites. A series of nickel-base catalysts were developed for hydrogenation of
FA to THFA, mainly focusing on the reaction time, reaction temperature, nickel
loading amount, and solvent and hydrogen pressure on conversion of furfural
alcohol and selectivity for tetrahydrofurfuryl alcohol. In comparison with previous
work [43], it was found that the catalyst we adopted allowed a much lower reaction
temperature (353 K) and shorter reaction time (2 h) with satisfactory conversion
and selectivity, which could not only save energy greatly, but also improve
environmental quality.
Experimental
Chemicals and equipment
FA, alcohol, methanol, Ni(NO3)2Á6H2O, and c-alumina particles were all purchased
from the local Sinopharm Chemical Reagent (purity more than 98 %), which were
directly used without further treatment.
Catalyst preparation
c-alumina particles were pretreated by mall bills and calcined at 673 K for 4 h.
Then, 10 g c-alumina and a certain amount of Ni(NO3)2Á6H2O were added into
200 mL deionized water and stirred for 24 h at room temperature. The mixture was
dried at 393 K for 12 h and calcined at 673 K for 4 h subsequently. Then, a nickel-
based catalyst was obtained after reduction in H2 at 623 K for 4 h.
123

S. Sang et al.
Catalyst characterization
The prepared nickel-based catalyst was characterized by powder X-ray diffraction
(XRD) which measurement performed on a Rigaku D/max-A instrument with a Cu
Ka radiation at 40 kV and 20 mA and a scan speed of 0.02 °C/min.
The structure of the catalyst was examined with a scanning electron microscope
(SEM).
The catalyst with transmission electron microscopy (TEM) analysis was
performed on a JEM-2100 instrument. The samples were prepared by mortar and
a droplet of ethanol to dilute the samples, then ultrasonically treated for 2 h make
the sample disperse completely.
The catalyst for X-ray photoelectron spectroscopy (XPS) analysis was performed
on a ESCALAB250Xi instrument. For in situ XPS analysis, the catalyst was placed
inside the XPS chamber and heated to 550 °C in the presence of 0.15 mbar H2
pressure. The temperatures were kept constant for 2 h in 550 °C under 0.15 mbar
H2 pressure prior to recording. The XPS spectra were recorded at a pass energy of
60 eV.
Fifty milligrams of catalyst was processed in the quart tube reactor by H2
temperature-programmed reduction (H2-TPR TP-5076), then it was pretreated in
helium (40 mL/min) at 400 °C for 1 h. After cooling to 100 °C, the sample was
heated from 50 to 800 °C at a heating rate of 30 °C/min in H2 (40 mL/min).
The catalyst of H2-chemisorption analysis was conducted on the same apparatus
as H2-TPR (TP-5076). The step is the same as the H2-TPR, when cooled to 100 °C,
hydrogen (40 mL/min) pulses were poured into the helium stream to make the
effluent area of consecutive pulses constant.
The BET surface was carried out on the Brunauer-Emmett-Teller (BET 3H-
2000PS1).
Catalytic hydrogenation of FA
Before the reaction, the catalysts were reduced by H2 in the tube furnace. After
reduction, the catalysts were still kept in the tube avoiding oxidation of catalysts’
surfaces in case of exposure to air. The surface oxidation can affect the catalytic
performance. When the reaction began, we took plastic wrap to protect the catalyst
and quickly added the catalysts into the reaction equipment. In this process, we
usually spent a minute or two. During the reaction, we also found that when the
catalyst exposure to air was for 5 min, it did not have any effect on the reaction. All
the liquid-phase catalytic hydrogenation of FA was carried out in a 100 mL stainless
autoclave equipped with an electromagnetically driven stirrer. For each run, 42.5 g
alcohol, 7.5 g FA, and 1 g catalyst were added to the reactor vessel. After displacing
the air, the hydrogen pressure was raised to a certain value. Then the reactor was
heated to the desired temperature, and the stirring speed was fixed at 600–800 rpm
to eliminate diffusion effects [49]. The hydrogenation reaction was carried out in a
reaction time range of 2–10 h to estimate the effect of reaction time. Meanwhile the
reaction was carried out between 353 and 423 K to estimate the effect of reaction
temperature. In addition, other effect factors such as hydrogen pressure, Ni-loading
123

Selective hydrogenation of furfuryl alcohol to…
amount, and solvent were also taken into consideration. When all the operating
conditions reached the settings, pressure drop was recorded every 10 min until there
was no obvious H2 pressure drop in a 30 min period. Finally, the reactor was
quickly cooled to room temperature, and the product samples were centrifuged to
separate the liquid products from the catalyst particles. The liquid phase was
analyzed by gas chromatography (Ouhua GC-9160) equipped with an SE-54
capillary column (30 m 9 0.32 mm 9 0.5 lm), with a flame ionization detector
(FID). The vaporization temperature was 473 K, the detector temperature was
473 K, and the oven temperature was 403 K. The main product was THFA, and the
product yield and selectivity were calculated and defined as follows:
n0FA À nFA
XFA
¼
 100 %;
n0FA
ni
n0FA
yi ¼
Si ¼
 100 %;
yi
 100 %;
XFA
where i represents the product FA in the reaction; n0FA and nFA describe the amounts
of FA before and after the reaction, respectively, in mol; and ni is the amount of
product i, in mol.
Results and discussion
Catalyst characterization
Figure 1a shows the XRD patterns of the calcined nickel-based catalyst before
reduced by hydrogen with a different Ni loading amount. The peaks at 45.5° and
68.5° could be assigned to the diffraction peaks of the (002), (003) planes of
c-Al2O3 (PDF No. 75-921). Three diffraction peaks of NiO at 38.5° (111), 45.5°
(200), and 65.5° (220) were found when the Ni loading amount was up to 15 wt%.
When the Ni loading amount was below 15 wt%, there was only one peak of NiO at
38.5° (111) observed. The reason for this phenomenon is that the surface of support
has little Ni, which cannot be found. Figure 1b shows the XRD patterns of the
reduced nickel-based catalyst with different Ni loading amounts. All the samples
had the characteristic reflections of Ni and c-Al2O3. The sharp and symmetric peaks
clearly indicated that the nickel-based catalysts crystallized well. The peak at 45.5°
and 68.5° could be assigned to the diffraction peaks of the (002), (003) planes of
c-Al2O3 (PDF No. 75-921). Except for the c-Al2O3 feature, the feature of Ni in the
nickel-based catalyst is clearly revealed. As we can see from Fig. 1b, the peak at
38.5° could be assigned to the diffraction peak of the (111) planes of Ni (PDF No.
47-1049). The curve of 15 wt% Ni/c-Al2O3 shows more sharp and symmetric peaks
than other curves. The reason for this phenomenon is that Ni in the surface of the
support has a good crystal shape (Fig. 8). In addition, from the figure, the obvious
123

S. Sang et al.
Fig. 1 a XRD patterns of nickel-based catalyst with different Ni loading amounts: (a) 5 wt% Ni/c-
Al2O3, (b) 10 wt% Ni/c-Al2O3, (c) 15 wt% Ni/c-Al2O3, (d) 20 wt% Ni/c-Al2O3, (e) 25 wt% Ni/c-Al2O3,
and (f) 30 wt% Ni/c-Al2O3. b XRD patterns of reduced nickel-based catalyst with different Ni loading
amounts: (a) 5 wt% Ni/c-Al2O3, (b) 10 wt% Ni/c-Al2O3, (c) 15 wt% Ni/c-Al2O3, (d) 20 wt% Ni/c-
Al2O3, (e) 25 wt% Ni/c-Al2O3, and (f) 30 wt% Ni/c-Al2O3. c XRD patterns of used catalyst and unused
catalyst
difference in Fig. 1a, b is that metal diffraction peak is reduced. In Fig. 1a, there are
three different NiO oxide diffraction peaks, however, in Fig. 1b, only the Ni metal
diffraction peak existed ascribed to Ni2? that reduced Ni0 in the tube furnace. From
Fig. 1b, there is no NiO oxide diffraction peak, it was determined that the Ni2? was
reduced to Ni0 completely at 623 K. In addition, from the data of XRD, our team
also calculated the particle size of Ni (see Table 1) and NiO. When the Ni loading
amount is at the low level, the peak of Ni is not clear indicating the good dispersion
of Ni. With the increase of Ni loading amounts, the peak of Ni began to become
Table 1 Particle size of NiO and Ni in the calcined and reduced states
Ni-loading amount
Ni particle size (nm)
NiO particle size (nm)
5 wt% Ni/c-Al2O3
10 wt% Ni/c-Al2O3
15 wt% Ni/c-Al2O3
20 wt% Ni/c-Al2O3
4.9
4.7
4.6
4.3
7.4
6.5
5.6
5.0
The particle size were calculate by XRD
123

Selective hydrogenation of furfuryl alcohol to…
intensive. When the Ni loading amount is up to 15 wt%, the peak of Ni can be
detected. It also showed that the 15 wt% nickel-based catalyst contributed to
conversion of FA and assured high selectivity toward THFA (Fig. 5). In addition,
Fig. 1c shows the XRD patterns of used catalyst and unused catalyst. We can see
that the diffraction peaks position of catalyst has not obviously changed in the
process of the hydrogenation reaction. The diffraction peaks show intensity
enhancement after the nickel based catalyst was used. The reason may be caused by
the particle size increase after use (see Fig. 4). In Table 1, we can see that the
particle size of NiO and Ni decreased with the increase of nickel loading amounts.
Figure 1 shows the diffraction peak of c-Al2O3 is not changed before and after
reduction. It can be seen that the support of c-Al2O3 is stable. The diffraction peak
of Ni has a large change after the reduction by H2. When the Ni loading amount is
under 15 wt%, the peak of NiO does not appear completely. When the Ni loading
amount is up to 15 wt%, the diffraction peak of NiO obviously completely appears.
Figure 2 shows the XPS pattern of the catalyst. Our team analyzed the
composition of prepared 15 wt% Ni/c-Al2O3. Absorption peaks were identified for
Ni, O and C. Figure 2a shows the general XPS spectra of nickel-based catalyst. It
perfectly shows the absorption peak of Ni, O, and C. From Fig. 2b, binding energies
at 855.9 and 861.3 eV were observed for nickel-based catalyst, corresponding to Ni0
(2p3/2) and Ni2? (2p3/2), respectively. The binding energies at 872.5 and 880.1 eV
correspond to the main lines of Ni0 (2p1/2) and Ni2? (2p1/2) [50].
The H2-TPR results are presented in Fig. 3. It can be seen that all the catalysts
have a relatively intensive H2 consumption peak at about 350–400 and 520–570 °C
with the increase of Ni loading amount in the catalysts. The peak of 350–400 °C
was ascribed to the reduction of bulk NiO oxides in the interaction with Al2O3
[51–53]. Meanwhile, the peak of 520–570 °C was ascribed to complex NiO species,
which have stronger interaction with the c-Al2O3 support [54]. Manikandan’s team
and Kong’s team also report about NiO species having stronger interaction with Ni-
supports [55, 56]. With the increase of Ni-loading amounts, the figure exhibited
broad reduction profiles, which could be ascribed to the reduction of NiO species.
From the figure, we can see that under 350 °C the profile has no H2 consumption,
and it also shows that the reduction temperature should higher than 350 °C. In
addition, our team also studied the impact of reduced temperature on conversion and
Fig. 2 a General XPS spectra of 15 wt% Ni/c-Al2O3. b XPS pattern of Ni2p of 15 wt% Ni/c-Al2O3
123

S. Sang et al.
Fig. 3 H2-TPR of Ni/c-Al2O3
catalysts with different loading
amounts: (a) 10 wt% Ni/c-
Al2O3, (b) 15 wt% Ni/c-Al2O3,
(c) 20 wt% Ni/c-Al2O3,
(d) 25 wt% Ni/c-Al2O3, and
(e) 30 wt% Ni/c-Al2O3
selectivity (see Table 2). Profile (a) shows that the H2 consumption peak at
350–400 °C is not obviously found due to little of the bulk NiO oxides (Fig. 3).
Figure 3 shows that with increase of Ni loading amounts the H2 consumption peak
at 350–400 °C can be obviously found. It indicates that the increase of Ni loading
amounts in the Ni/c-Al2O3 catalysts can lead to the increase of bulk NiO species.
Table 3 shows the fundamental physical properties of the nickel-based catalysts,
including BET surface area and pore volume. The BET surface areas for the Ni/c-
Al2O3 catalysts range from 163.7 to 188.3 m2/g concerning the Ni loading amount.
From the data, we can make a conclusion that the BET surface area is not
considered changed with the change of Ni loading amounts. The Hou et al. [50]
team reported that the increase of Ni content in the Ni/c-Al2O3 catalysts led to the
decrease of the BET surface areas. The decreased BET surface area was mainly
caused by the coverage of Ni on the c-Al2O3 surface. In our report, the increase of
Ni-loading amount led to a slight change of BET area; therefore, we ignore the
impact of area on catalytic activity.
Table 4 shows the particle size of nickel-based catalysts and the reduction degree
calculated by three different methods,such as XRD, TEM, and H2 adsorption. The
XRD method used Scherrer’s equation. The method of H2 adsorption is based on the
dispersion and reduction degree. The reduction degree represents that consumed in
TPR after decomposition of nitrate. We can see from the picture that the particle
size increase from 4.3 to 4.9 nm with the Ni loading amount decreased from 20 to
15 wt%. These particle sizes were smaller than the Montes team’s report, the Louis
team’s report, and the Pilar team’s report [48, 57, 58]. In the left column, the
reduction degree has almost no change.
Table 2 Effect of reduction temperature on conversion and selectivity
Reduction temperature (°C)
200
250
300
350
Conversion (%)
Selectivity (%)
34.7
70.4
50.8
86.5
70
99.8
99.5
99.2
Reaction conditions: 1 g 15 wt% Ni/c-Al2O3 catalyst, 42.5 g alcohol, 7.5 g furfuryl alcohol; reaction
temperature 353 K; reaction time 2 h; and reaction pressure 4 MPa
123

Selective hydrogenation of furfuryl alcohol to…
Table 3 Physical properties of
Catalyst
SBET (m2/g)
VP (mL/g)
Ni/c-Al2O3 catalysts
5 wt% Ni/c-Al2O3
188.3
182.4
174.4
163.7
0.9052
0.7963
0.7433
0.6593
10 wt% Ni/c-Al2O3
15 wt% Ni/c-Al2O3
20 wt% Ni/c-Al2O3
Table 4 Particle size of Ni/c-Al2O3 catalysts
Catalyst
Reduction degree
H2 adsorption
XRD
TEM
5 wt% Ni/c-Al2O3
10 wt% Ni/c-Al2O3
15 wt% Ni/c-Al2O3
20 wt% Ni/c-Al2O3
0.9
1.0
1.0
0.9
2.4
3.5
4.3
5.7
4.9
4.7
4.6
4.3
–
–
6.52
–
The H2 adsorption, XRD, and TEM are three different methods to calculate particle size (nm)
Table 5 TOF value of catalysts
Catalyst
TOF
TOF [11]
5 wt% Ni/c-Al2O3
10 wt% Ni/c-Al2O3
15 wt% Ni/c-Al2O3
20 wt% Ni/c-Al2O3
0.7
1.0
1.1
0.9
–
–
0.63
–
TOF value. Reaction condition: 42.5 g alcohol, 7.5 g furfuryl alcohol; reaction temperature 353 K;
reaction time 1 h; reaction pressure 4 MPa
Table 5 shows the TOF values that are calculated by the H2 adsorption amount
(Table 4). For the data from the table, we calculated when the catalysts were not in
the optimum reaction condition. We adopted the reaction time of 1 h. From the
table, we can see that the TOF is similar to the Ni/c-Al2O3 (Ni loading amount is 5
to 20 wt%) with the Ni particle size 2.4 to 5.7 nm (Table 4). In the right column is
the TOF from Yoshinao’s team [11]. In the reference, the preparation method of
catalyst was different from our team. Our catalyst calcined at 773 K which is similar
to the reference; therefore, our team only compared the 15 wt% Ni/c-Al2O3 with the
reference. We can see that our TOF value is larger than the value obtained by
Yoshinao’s team. Che’s team and Shin’s team reported that, if the activity is
structure-sensitive, the catalytic activity will greatly change when the particle size
of the catalysts is in the typical region [59, 60]. Therefore, we also adopt this
method in the hydrogenation reaction.
Figure 4 shows the TEM image of nickel-based catalyst (Ni loading amount up to
15 wt%) and reused nickel-based catalyst (Ni loading amount up to 15 wt%). From
the picture of (a) and (b), the range of particle size from 4.5 to 6 nm, this result
corresponded fairly well with XRD and H2 adsorption. It also verified the particle
123

S. Sang et al.
Fig. 4 a, b TEM of Nickel-based catalyst (15 %); c, d TEM of reused Nickel-based catalyst (15 %)
sizes of Ni determined by various methods agreeably well. From the picture of
(c) and (d), we can see that it has a large support and that a large number of Ni metal
got together. The reason for this is because the reused catalyst lead to Ni metal
active component agglomeration. From the picture of (c) and (d), the particle size of
the used 15 wt% Ni/c-Al2O3 catalyst is in the range of 8–11 nm, which is larger
than the fresh 15 wt% Ni/c-Al2O3 catalyst. This indicates that when the
agglomeration happened it leads to catalyst activity decreasing. This was the
reason that the catalytic activity decreased with the number of used nickel-based
catalysts.
Catalytic hydrogenation of furfuryl alcohol
Influence of Ni loading amount on the conversion and selectivity
First, we show in Table 6 that when the reaction was not carried out at optimum
reaction conditions, we adopted 1 h as the reaction time. We found that the
123

Selective hydrogenation of furfuryl alcohol to…
Table 6 Effect of Ni loading amount on conversion and selectivity
Ni loading
5 %
10 %
15 %
20 %
25 %
30 %
Conversion (%)
Selectivity (%)
10.7
65.5
65.2
80.7
79.3
92.1
80.1
93.3
80.9
93.9
81.1
94.1
Reaction conditions: 42.5 g alcohol, 7.5 g furfuryl alcohol; reaction temperature 353 K; reaction time
1 h; and reaction pressure 4 MPa
Fig. 5 Effect of Ni loading
amount on conversion and
selectivity. 42.5 g alcohol, 7.5 g
furfuryl alcohol; reaction
temperature 353 K; reaction
time 2 h; and reaction pressure
4 MPa
conversion of FA and selectivity toward THFA increased with the increase of the Ni
loading amount. In addition, Fig. 5 shows the effect of Ni loading amounts on the
conversion of FA and selectivity toward THFA. As can be seen from Fig. 5, the
reaction conversion increases from 25.1 % to about 99.7 % with the increase of Ni
loading amounts. The conversion of FA was up to 99.7 % over 15 wt% Ni/c-Al2O3.
When the Ni loading amount continues to increase, the reaction conversion has
almost no change. In addition, Fig. 1b shows that 15 wt% Ni/c-Al2O3 has good
dispersibility, which is helpful for increasing the reaction conversion and selectivity.
The selectivity toward THFA has a high level no matter what the Ni loading amount
(it varied from 5 to 30 wt%). The amount of by-product in this reaction can be
ignored, which was very subtle (\0.02 %). The increase of the Ni loading amount
could lead to the improvement of FA conversion due to the increase of Ni in the
catalysts, from which it could be shown that the diffraction peak Ni (Fig. 1a) is in
favour of catalytic hydrogenation of FA. The figure shows that when the Ni loading
amount is 5 and 10 wt%, the conversion of FA obviously is lower than 100 %. The
reason is the Ni loading amount in the surface of catalysts is very little and it
directly lead to catalytic performance. When the Ni loading amount in the low
phase, the complex NiO species has little interaction with c-Al2O3 support. It was
also support by the H2-TPR (Fig. 3). So, the optimum loading amount of Ni in Ni/c-
Al2O3 is 15 wt%.
123

S. Sang et al.
Fig. 6 Effect of reaction
temperature on the conversion
and selectivity. 1 g 15 wt% Ni/
c-Al2O3 catalyst, 42.5 g alcohol,
7.5 g furfuryl alcohol; reaction
time 2 h; and reaction pressure
4 MPa
Influence of reaction temperature on conversion and selectivity
Figure 6 shows the influence of the reaction temperature on the conversion of FA
and the selectivity of THFA. As can be seen from Fig. 6, the conversion of FA
increased from the 57.3 to 99.7 % when the temperature increased from 70 to
150 °C, and the conversion reached a peak at 80 °C (about 99.7 %). Then the
conversion has no obvious change with further increase of the temperature. It also
can be seen that the selectivity of THFA has almost no change (about 99.5 %) with
the change of reaction temperature, which shows that the reaction temperature has
little impact on the selectivity of the reaction when the reaction temperature is up to
80 °C. The reaction has almost no by-product when the temperature changes from
80 to 140 °C. When the temperature was higher than 140 °C, dark and viscous
compounds began to appear due to the side reactions (such as polymerization) that
happened during the reaction. Therefore, considering the energy consumption and
the by-product during the reaction, the optimum reaction temperature is selected as
80 °C.
Influence of reaction time on conversion and selectivity
As can be seen in Fig. 7, the reaction time was a great important factor in the
catalytic hydrogenation of FA. Obviously, when the reaction time was prolonged,
the conversion of FA increased gradually from 63.4 to 99.7 %. However, the
selectivity toward THFA had no obvious change when the time ranged from 1 to
10 h. When the reaction was carried out for 2 h, the reaction could give both better
conversion and selectivity toward the wanted THFA (about 99.5 %). Therefore, 2 h
was finally selected as the optimum reaction time.
Influence of hydrogen pressure on conversion and selectivity
Figure 8 shows the effect of the hydrogen pressure on the conversion of FA and
selectivity toward THFA. As can be seen from Fig. 8, the reaction conversion
increases with the increase of hydrogen pressure. When the hydrogen pressure
reached 4 MPa, the conversion was up to 99.7 %, but with the further increase of
123

Selective hydrogenation of furfuryl alcohol to…
Fig. 7 Effect of reaction time
on conversion and selectivity.
1 g 15 wt% Ni/c-Al2O3 catalyst,
42.5 g alcohol, 7.5 g furfuryl
alcohol; reaction temperature
353 K; and reaction pressure 4
Mpa
Fig. 8 Effect of reaction
pressure on conversion and
selectivity. 1 g 15 wt% Ni/c-
Al2O3 catalyst, 42.5 g alcohol,
7.5 g furfuryl alcohol; reaction
temperature 353 K; and reaction
time 2 h
the reaction pressure, there is almost no change in the conversion. The increase of
conversion can be explained by the pushing effect of excess hydrogen on the
reaction balance as the hydrogen pressure increased from 2 to 4 MPa. However, the
selectivity toward THFA was always kept at a high level (about 99.5 %). The results
show that the hydrogen pressure has a large impact on the conversion of the reaction
and little impact on the selectivity. The optimal hydrogen pressure is 4 MPa.
Influence of solvent on conversion and selectivity
As can be seen in Table 7, the solvent has a great effect on the hydrogenation of FA.
It can be seen that when water was used as solvent, the reaction conversion and
selectivity had a low level. The conversion of FA is only 51.5 %, the selectivity
toward THFA is only 87 %. When ethanol was chosen as the reaction solvent, the
reaction conversion and selectivity were up to 99.9 and 99.7 %, respectively. When
methanol was selected as the reaction solvent, the selectivity toward the product had
a high level, however, the conversion of FA was lower than 99.7 %. The better
conversion and selectivity might be caused by the hydrogen-donor effect of ethanol,
which has been reported by Ouyang’s group [61]. We have also investigated the
hydrogen-donor effect of ethanol. In addition, from the price perspective, ethanol is
123

S. Sang et al.
Table 7 Effect of solvent on the conversion of FA and selectivity toward THFA
Solvent
Selectivity (%)
Conversion (%)
Water
87.0
99.9
99.7
51.5
99.7
93.4
Ethanol
Methanol
Reaction conditions: 1 g 15 wt% Ni/c-Al2O3 catalyst, 42.5 g alcohol, 7.5 g furfuryl alcohol; reaction
temperature 353 K; reaction time 2 h; and reaction pressure 4 MPa
Table 8 Effect of support on the conversion of FA and selectivity toward THFA
Support
Selectivity (%)
Conversion (%)
None
54.1
99.9
33.6
99.7
c-Alumina
Reaction conditions: 1 g catalyst, 42.5 g alcohol, 7.5 g furfuryl alcohol; reaction temperature 353 K;
reaction time 2 h; and reaction pressure 4 MPa
cheaper than methanol and pollution-free for the environment. So, the optimum
solvent is ethanol.
Influence of support on the conversion and selectivity
We can see from Table 8 that the support has a great effect on conversion of FA and
selectivity toward THFA. In this part, our team mainly investigated the impact of c-
alumina and no support on the hydrogenation of FA. In Table 8, when we chose c-
alumina as support, the conversion and selectivity was up to 99.9 and 99.7 %,
respectively. It illustrates that the Ni metal has a stronger interaction with c-
alumina. In addition, when we do not choose any support, we found that the
conversion and selectivity were both in the low phase. This illustrates that the
support has a great effect on the reaction.
Catalyst recyclability
To study the recyclability of the catalyst, a batch of 15 wt% Ni/c-Al2O3 was used
repeatedly for the catalytic hydrogenation of FA at 353 K in alcohol. All the used
catalyst was washed in alcohol then dried at 333 K. It can be seen from Table 9 that
the conversion has slightly decreased after the catalyst was reused twice; however,
the selectivity toward THFA decreased obviously indicating the decrease of catalyst
activity, which might be caused by catalyst agglomeration in the catalytic
hydrogenation process of FA (see Fig. 4) [62]. From Fig. 3, we can see that the
particle size of catalyst increased after repeated use. In addition, Fig. 9a shows that
the catalysts have a good crystal shape before reuse. It was found from Fig. 9b that
123

Selective hydrogenation of furfuryl alcohol to…
Table 9 Results of repeated use of Ni/c-Al2O3 catalysts
Entry
Selectivity (%)
Conversion (%)
1
2
3
97.6
85.2
60.1
99.7
99.6
98.4
Reaction conditions: Reaction conditions: 1 g 15 wt% Ni/c-Al2O3 catalyst, 42.5 g alcohol, 7.5 g furfuryl
alcohol; reaction temperature 353 K; reaction time 2 h; and reaction pressure 4 MPa
Fig. 9 a SEM of Nickel-based catalyst (15 wt%); b SEM of reused oNickel-based catalyst (15 wt%)
we only see large support and a little metal. This can illustrate that the activity
component of the catalyst agglomeration phenomenon happened. In addition, our
team found that some by-products exist in the reaction when the catalysts were
reused twice and for a third time. For example, the main by-products are furan and
2-methylfuran [63]. The reason for this is caused by C=C and C–O bonds rupturing
in the hydrogenation reaction. When the catalysts were reused twice, the selectivity
toward 2-methyfuran was up to 10 %, and the selectivity toward furan was only
4.3 %. When the catalysts were reused a third time, the selectivity toward
2-methyfuran was up to 30.4 %, and the selectivity toward furan was 9.5 %. As for
the reaction, this reaction time of 2 h was selected; prolonged reaction time did not
affect the conversion and selectivity in catalyst recyclability.
Conclusion
Ni/c-Al2O3 catalysts were prepared by the impregnation method and applied in
catalytic hydrogenation of furfural alcohol to tetrahydrofurfuryl alcohol. FA could
be converted to THFA over 15 wt% Ni/c-Al2O3 with the yield up to 99.9 % under a
reaction temperature of 353 K and reaction pressure of 4 MPa and reaction time of
2 h. The nickel-based catalyst in this reaction has some advantages such as the
catalyst’s easy preparation, high catalytic activity, and easy separation by simple
123

S. Sang et al.
centrifugation. The reaction temperature dropped from 433 to 353 K, and reaction
time dropped from 8 to 2 h. The experimental results could provide a good research
basis for the catalytic preparation of tetrahydrofurfuryl alcohol from furfural alcohol
under mild reaction conditions.
Acknowledgments This work was financially supported by the National Natural Science Foundation of
China (Nos. 21276050 and 21406034), Fundamental Research Funds for the Central Universities (No.
3207045414), Key Laboratory Open Fund of Jiangsu Province (JSBEM201409), and the Priority
Academic Program Development of Jiangsu Higher Education Institutions.
References
1. G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106, 4044 (2006)
2. P. Gallezot, Chem. Soc. Rev. 41, 1538 (2012)
3. I. Delidovich, K. Leonhard, R. Palkovits, Energy Environ. Sci. 7, 2803 (2014)
4. R.A. Sheldon, Green Chem. 16, 950 (2014)
5. R. Karinen, K. Vilonen, M. Niemela, ChemSusChem 4, 1002 (2011)
6. L. Hu, G. Zhao, W.W. Hao, X. Tang, Y. Sun, L. Lin, S.J. Liu, RSC Adv. 2, 11184 (2012)
7. I. Agirrezabal-Telleria, I. Gandarias, P.L. Arias, Catal. Today 234, 42 (2014)
8. H. Li, Q.Y. Zhang, P.S. Bhadury, S. Yang, Curr. Org. Chem. 18, 547 (2014)
9. S.P. Teong, G.S. Yi, Y.G. Zhang, Green Chem. 16, 2015 (2014)
10. R.F. Perez, M.A. Fraga, Green Chem. 16, 3942 (2014)
11. Y. Nakagawa, H. Nakazawa, H. Watanabe, K. Tomishige, ChemCatChem 4, 1791–1797 (2012)
12. Y. Nakagawa, K. Tomishige, Catal. Commun. 12, 154–156 (2010)
13. Y. Nakagawa, K. Takada, M. Tamura, K. Tomishige, ACS Catal. 4, 2718–2726 (2014)
14. S. Liu, Y. Amada, M. Tamura, Y. Nakagawa, K. Tomishige, Green Chem. 16, 617–626 (2014)
15. S. Liu, Y. Amada, M. Tamura, Y. Nakagawa, K. Tomishige, Catal. Sci. Technol. 4, 2535 (2014)
16. R.S. Rao, R.T.K. Baker, M. Vannice, Catal. Lett. 60, 51 (1999)
17. N. Merat, C. Godawa, A. Gaset, J. Chem. Technol. Biotechnol. 48, 145 (1990)
18. O. Levenspiel, Chemical reaction engineering, vol. 25, 2nd edn. (Wiley, New York, 1972), p. 265
19. A. Corma, S. Iborra, A. Velty, Chem. Rev. 107, 2411 (2007)
20. N. Merat, C. Godawa, A. Gaset, Chem. Technol. Biotechnol. 48, 145 (1990)
21. H.P. Thomas, C.L. Wilson, J. Am. Chem. Soc. 73, 4803 (1951)
22. H.Y. Zheng, Y.L. Zhua, B.T. Teng, Z.Q. Bai, C.H. Zhang, H.W. Xiang, Y.W. Li, Mol J. Catal.
A Chem. 246, 18 (2006)
23. W.H. Bagnall, E.P. Goodings, C.L. Wilson, J. Am. Chem. Soc. 73, 4794 (1951)
24. E.P. Goodings, C.L. Wilson, J. Am. Chem. Soc. 73, 4798 (1951)
25. E.P. Goodings, C.L. Wilson, J. Am. Chem. Soc. 73, 4801 (1951)
26. S. Sato, R. Takahashi, N. Yamamoto, E. Kaneko, H. Inoue, Appl. Catal. A Gen. 334, 84 (2008)
27. F. Sato, H. Okazaki, S. Sato, Appl. Catal. A Gen. 419, 41 (2012)
28. F. Sato, S. Sato, Catal. Commun. 27, 129 (2012)
29. S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K. Kunimori, K. Tomishige, Chem. Commun. 15,
2035 (2009)
30. S. Koso, N. Ueda, Y. Shinmi, K. Okumura, T. Kizuka, K. Tomishige, J. Catal. 267, 89–92 (2009)
31. Y. Nakagawa, M. Tamura, K. Tomishige, ACS Catal. 3, 2655–2668 (2013)
32. Y. Nakagawa, K. Tomishige, Catal. Today 195, 136–143 (2012)
33. Y. Nakagawa, M. Tamura, K. Tomishige, Catal. Surv. Asia 19, 249–256 (2015)
34. L.E. Schniepp, H.H. Geller, J. Am. Chem. Soc. 68, 1646 (1946)
35. H. Adkins, R. Connor, J. Am. Chem. Soc. 53, 1091 (1931)
36. K. Chen, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, ChemCatChem 2, 547–555 (2010)
37. S. Koso, Y. Nakagawa, K. Tomishige, J. Catal. 280, 221–229 (2011)
38. S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, Appl. Catal. B Environ. 111–112,
27–37 (2012)
39. S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige, J. Phys. Chem. C 116, 3079–3090
(2012)
123

Selective hydrogenation of furfuryl alcohol to…
40. B. Aeijelts, G. Mul, M. Makkee, J. Moulijn, J. Catal. 243, 171–182 (2006)
41. S. Winterle, A. Kraynov, J. Klankermayer, W. Leitner, M.A. Liauw, Chem. Ing. Tech. 82, 1211
(2010)
42. M.A. Tike, V.V. Mahajani, Ind. Eng. Chem. Res. 46, 3275 (2007)
43. X.C. Chen, W. Sun, N. Xiao, J.Y. Yan, S.W. Liu, J. Chem. Eng. 126, 5 (2007)
44. M.V. Rajashekharam, J. Sci. Ind. Res. 56, 595 (1997)
45. H.R. Reinhoudt, R. Troost, A.D. van Langeveld, J.A.R. van Veen, S.T. Sie, J.A. Moulijn, J. Catal.
203, 509 (2001)
46. S. Sitthisa, D.E. Resasco, Catal. Lett. 141, 784–791 (2011)
47. X. Kong, Y. Zhu, H. Zheng, X. Li, Y. Zhu, Y.W. Li, ACS Catal. 5, 5914 (2015)
48. M.P. GonzalezMarcos, J.I. GutierrezOrtiz, C.G. deElguea, J.A. Delgado, J.R. GonzalezVelasco,
Appl. Catal. A 162, 269–280 (1997)
49. M.H. Zhou, H.Y. Zhu, L. Niu, G.M. Xiao, R. Xiao, Catal. Lett. 144, 235 (2014)
50. T. Hou, L.X. Yuan, T.Q. Ye, L. Gong, J. Tu, M. Yamamoto, Y. Torimoto, Q.X. Li, Int. J. Hydrogen
Energy 34, 9095 (2009)
51. C. Wu, L. Wang, P.T. Williams, J. Shi, J. Huang, Appl. Catal. B Environ. 108, 6 (2011)
52. Z. Yuan, L. Wang, J. Wang, S. Xia, P. Chen, Z. Hou, X. Zheng, Appl. Catal. B 101, 431–440 (2011)
53. C. Rudolf, B. Dragoi, A. Ungureanu, A. Chirieac, S. Royer, A. Nastro, E. Dumitriu, Catal. Sci.
Technol. 4, 179–189 (2014)
54. Z. Hou, O. Yokota, T. Tanaka, T. Yashima, Appl. Catal. A Gen. 253, 381 (2003)
55. X. Kong, R. Zheng, Y. Zhu, G. Ding, Y. Zhu, Y.W. Li, Green Chem. 17, 2504–2514 (2015)
56. M. Manikandan, A.K. Venugopal, K. Prabu, R.K. Jha, R. Thirumalaiswamy, J. Mol. Catal. A: Chem.
417, 153–162 (2016)
57. M. Montes, C.P. deBosscheyde, B.K. Hodnett, F. Delannay, P. Grange, B. Delmon, Appl. Catal. 12,
309–330 (1984)
58. C. Louis, Z.X. Cheng, M. Che, J. Phys. Chem. 97, 5703–5712 (1993)
59. M. Che, C.O. Bennett, Adv. Catal. 36, 55–172 (1989)
60. E.J. Shin, M.A. Keane, Ind. Eng. Chem. Res. 39, 883–892 (2000)
61. X.P. Ouyang, X.Z. Huang, Y. Zhu, X.Q. Qiu, Energy Fuels 29, 5835 (2015)
62. H.Y. Zhu, M.H. Zhou, Z. Zou, G.M. Xiao, R. Xiao, Korean J. Chem. Eng. 31, 593 (2014)
63. S. Bhogeswararao, D. Srinivas, J. Catal. 327, 65–77 (2015)
123