Hydrogenolysis of biomass-derived sorbitol over la-promoted ni/zro2 catalysts

Article abstract of DOI:10.1039/c9ra10394e

Ni/La₂O₃/ZrO₂ catalysts were prepared by a step-by-step impregnation method through regulation of the contents of the active component and alkali. The introduction of an alkaline promoter not only enhanced the alkalinity of the catalyst but also improved the dispersion of Ni on the catalyst owing to the strong interaction between Ni²⁺ and alkali promoter. The synergistic effect between Ni and La₂O₃ was beneficial to selective hydrogenolysis of sorbitol. Under the optimal reaction conditions, sorbitol conversion reached nearly 100% and target products (ethylene glycol, 1,2-propanediol, and glycerol) selectivity reached 74.8%. Metal-alkali coordination mechanism and possible pathways for target products formation were proposed.

Full text of DOI:10.1039/c9ra10394e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Hydrogenolysis of biomass-derived sorbitol over
La-promoted Ni/ZrO2 catalysts†
Cite this: RSC Adv., 2020, 10, 3993
abc
abc
Chiliu Cai,
Qi Zhang,
‡
abc
Haiyong Wang,
Xinghua Zhang,
‡
abc
Haosheng Xin,abc Changhui Zhu,abc
abc
abc
*
*
Chenguang Wang,
Qiying Liu
and Longlong Maabc
Ni/La2O3/ZrO2 catalysts were prepared by a step-by-step impregnation method through regulation of the
contents of the active component and alkali. The introduction of an alkaline promoter not only enhanced
the alkalinity of the catalyst but also improved the dispersion of Ni on the catalyst owing to the strong
interaction between Ni2+ and alkali promoter. The synergistic effect between Ni and La2O3 was
beneficial to selective hydrogenolysis of sorbitol. Under the optimal reaction conditions, sorbitol
conversion reached nearly 100% and target products (ethylene glycol, 1,2-propanediol, and glycerol)
selectivity reached 74.8%. Metal–alkali coordination mechanism and possible pathways for target
products formation were proposed.
Received 11th December 2019
Accepted 8th January 2020
DOI: 10.1039/c9ra10394e
rsc.li/rsc-advances
bonds.14,15 According to previous reports,11–17 owing to their high
activity for C–C bond cleavage, Ni- and Ru-based catalysts have
1. Introduction
been oen applied in experiments. Unfortunately, they also
catalyse excessive C–C bond scission to produce by-products. In
contrast, Cu-based catalysts exhibit good activities for C–O, C]
O, and C]C bond hydrogenation and C–OH dehydrogenation
but poor reactivity activity for C–C bond cleavage.18 Huang
et al.19 observed that, by using Cu–SiO2 catalysts, one could
obtain ꢀ60% yield of glycols and glycerol (GLY) in the presence
of Ca(OH)2. Ni–NaY and Pt–NaY catalysts were used in the
presence of Ca(OH)2. It was found that the main products on
Ni–NaY and Pt–NaY catalysts were 1,2-propylene glycol (1,2-PG)
and GLY, respectively.20
Composite catalysts combining with metallic sites and basic
sites have been used in this reaction.21 A Cu/g-Al2O3 catalyst can
promote C–C and C–O bonds cleavage, leading to the conver-
sion of glucose to glycols (with a selectivity of 67%).22 Ordered
mesoporous 10.9% Ni/6.3% CeO2–Al2O3 showed better catalytic
performance in the hydrogenolysis of sorbitol, obtaining
a 56.9% yield of ethylene glycol (EG) and 1,2-PG.23 A Cu–Ni/ZrO2
catalyst was efficient for xylitol conversion to EG and 1,2-PG
with a 65% total selectivity in water without a base additive.24
Recently, researchers25 found that Ru–(Mn–Al)Ox solid base
catalysts exhibited high performance (with 58% selectivity of
glycols) in the hydrogenolysis of xylitol under base-free condi-
tions. This catalyst, however, was not stable during the reaction
because the Mn species were converted to other compounds.
3% Ru/4.5% MnO/C also showed good activity with 70% selec-
tivity of glycols in a water–alcohol mixture solvent, avoiding
competitive decarbonylation at the Ru sites. However, the
catalyst exhibited poor hydrothermal stability in the hydro-
genolysis of xylitol.26 Zhang27 reported the use of 10% Ni–0.5%
On account of excessive exploitation of fossil fuel sources and
serious contaminative environmental problems, worldwide
efforts have been devoted to exploiting new energy sources,
especially clean and renewable energy sources.1–3 As a clean and
renewable resource, biomass energy is an ideal alternative
resource with its advantages of high reserves, renewability, and
non-greenhouse-gas emission.4–6 Sugar alcohols (sorbitol and
xylitol) are the important platform compounds produced from
biomass and can be converted into high-value-added chemicals
through hydrogenolysis, oxidation, or selective reduction.7–10
It is widely believed that hydrogenolysis of polyols to glycols
involves three key steps:11,12 First, some hydroxyl groups on the
molecule of polyols are dehydrogenated at the metal active sites.
Subsequently, C–C bond cleavage of intermediates occurs by an
alkaline catalyst via retro-aldol condensation. Finally, the
carbonyl intermediates are hydrogenated to target glycols on
the metal surfaces.13 For sorbitol conversion into small-
molecule alcohols, researchers have mainly focused on the
synergistic regulation of metal active sites and alkaline sites,
which are useful for selective cleavage of C–C and C–O
aGuangzhou Institute of Energy Conversion, Chinese Academy of Sciences, 510640
Guangzhou, China. E-mail: liuqy@ms.giec.ac.cn; wangcg@ms.giec.ac.cn; Fax: +86-
20-87057789; Tel: +86-20-87048614; +86-20-37029721
bKey Laboratory of Renewable Energy, Chinese Academy of Sciences, 510640
Guangzhou, China
cGuangdong Key Laboratory of New and Renewable Energy Research and Development,
510640 Guangzhou, China
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c9ra10394e
‡ Dr Chiliu Cai and Dr Haiyong Wang contributed equally to this work.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 3993–4001 | 3993

View Article Online
RSC Advances
Paper
Ir/La2O3 catalyst for the conversion of cellulose to EG and 1,2-PG sample was placed into a U-shaped quartz tube, reduced at
in a higher yield of 63.7%.
500 ꢁC for 4 h under H2 ow and purged at the same temperature
Although the catalytic system reported in the previous liter- for 40 min under He ow. Aer cooling to 50 ꢁC, it was adsorbed
ature can efficiently realize sorbitol hydrogenolysis, the with CO for 80 min. The sample was then purged by a He ow to
commonly used alkaline additives are slightly soluble CaO or remove the physical adsorbed CO for 1 h. The sample was heated
Ca(OH)2, which can easily corrode equipment and cause poor from 50 ꢁC to 850 ꢁC at a heating rate of 10 ꢁC minꢂ1 and the CO
product separation.28,29 Therefore, the development of new desorption was monitored with a TCD detector.
solid-alkali-supported metal catalysts can effectively solve this
CO2 temperature-programmed desorption (CO2-TPD)
problem. In this paper, using the inexpensive alkaline additive measurement was conducted on CHEMBET 3000 equipment
La2O3 and inexpensive active Ni, we prepared Ni/La2O3/ZrO2 (Quantachrome, USA). In a typical run, 50 mg sample was put
catalysts with different Ni/La ratios by using the initial wetness into a U-shaped quartz tube and purged with He at 400 ꢁC for 40
ꢁ
impregnation method. The physicochemical properties of the minutes. Aer cooling to 50 C, it was adsorbed with CO2 for
catalysts were characterized by X-ray diffraction, scanning 1 h. Finally, He was again introduced to remove the physical
electron microscopy, H2 and CO2 temperature-programmed adsorbed CO2 for 1 h. The sample was heated from 50 ꢁC to
reduction techniques, inductively coupled plasma atomic 900 ꢁC at 10 ꢁC minꢂ1 heating rate and the CO2 desorption was
emission spectrometry, X-ray photoelectron spectroscopy and monitored with a TCD detector.
Transmission Electron Microscopy. The catalytic performance
H2 temperature-programmed reduction (H2-TPR) was per-
was investigated in sorbitol hydrogenolysis via different reac- formed on an Auto Chem II 2920 chemical adsorption instru-
tion parameters. The synergistic effect between the metal and ment. For a typical test, 20 mg sample was purged with He ow
the base of the catalysts was discussed and the possible path- at 250 ꢁC for 1 h, then the temperature was decreased to 50 ꢁC.
ways for target products formation were also proposed during The sample was heated from 50 ꢁC to 900 ꢁC at a heating rate of
sorbitol hydrogenolysis.
10ꢁC minꢂ1 under a 5% H2/He ow.
The Ni and La leaching aer reaction was analyzed by the
inductively coupled plasma atomic emission spectrometry (ICP-
2. Experimental section
2.1. Catalyst preparation
AES) measurement using
instrument.
a PerkinElmer OPTIMA 8000
The microstructure of catalysts was analyzed by Trans-
mission Electron Microscopy (TEM) on a JEOL JEM-2100EX
electron microscopy and the electron beam accelerating
voltage was 200 kV.
X-ray photoelectron spectroscopy (XPS) was performed by
using a Thermo Fisher Scientic Escalab 250 Xi photoelectron
spectrometer and values were corrected by C 1s (284.8 eV). The
spectra was excited using Mg Ka (1253.6 eV) radiation (operated
at 200 W) of a twin anode in the constant analyzer energy mode
with a pass energy of 30 eV.
The Ni/La2O3/ZrO2 catalysts were prepared by initial wetness
impregnation. ZrO2 was added to a La(NO3)3$6H2O solution
with different La contents, and the mixture was stirred for 4 h at
room temperature. The excess H2O was evaporated at 80 ꢁC for
12 h. Subsequently, the obtained powder was dried overnight at
100 ꢁC. Finally, the powder was calcined at 550 ꢁC for 4 h at
a heating rate of 5 C min to obtain the La2O3/ZrO2.
The obtained La2O3/ZrO2 was added to a Ni(NO3)2$6H2O
solution with a certain Ni amount, followed by rigorous stirring
at room temperature at 4 h. Then the sample was dried and
calcined following the same procedure mentioned above.
Finally, the Ni/La2O3/ZrO2 catalysts were reduced in the H2
atmosphere at 500 ꢁC for 4 h before reaction.
ꢂ1
ꢁ
2.3. Sorbitol hydrogenolysis and products analysis
Sorbitol hydrogenolysis was performed in a stainless steel
autoclave, into which 20 mL of 100 mg mLꢂ1 sorbitol aqueous
solution and 0.4 g of freshly reduced catalyst were introduced.
2.2. Catalyst characterization
The textural properties of the catalysts were determined using Aer eliminating the air residue by H2 ushing, the reactor was
an automated surface area and pore size analyzer (Quantach- pressurized by H2 and was heated to a certain temperature for
rome Co., USA) at ꢂ196 ꢁC.
a certain period. During the hydrogenolysis process, the H2
X-ray powder diffraction patterns were obtained on a Rigaku pressure was kept at a constant 4 MPa by supplying H2. Aer
D/max-rC X-ray diffractometer operated at 40 kV and 40 mA reaction, the reaction system was cooled to collect the products.
using Cu Ka radiation (l ¼ 0.154 nm) and 2q was scanned from
The stability experiment procedure was described as follows:
aer each reaction cycle, the reaction liquid was centrifuged,
10ꢁ to 80ꢁ with a step size of 0.0167ꢁ.
The surface morphology of the catalyst was observed using the upper layer liquid was removed for product analysis, and the
a Hitachi S-4800 scanning electron microscope (SEM). lower catalysts continued to be used in the next cycle reaction
The thermogravimetry (TG) experiments were operated froꢂm1 without any treatment. This cycle was repeated ve times. For
ꢁ
ꢁ
room temperature to 550 C at a heating rate of 20 C min
under an air atmosphere.
catalyst regeneration, the used catalysts were calcined in an air
atmosphere at 500 C for 1 h and then were reduced in an H2
ꢁ
CO temperature-programmed desorption (CO-TPD) were atmosphere at 500 ꢁC for 4 h.
measured using a CHEMBET 3000 chemisorption/desorption
The aqueous products were analysed with a Waters 2695
apparatus (Quantachrome Co., USA). In a typical test, a 50 mg HPLC containing a refraction index and UV detector (RID-2414)
3994 | RSC Adv., 2020, 10, 3993–4001
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
and SH1011 separation column. Sulfuric acid solution
(0.005 mol Lꢂ1) was used as the eluent at a 0.5 mL minꢂ1 ow
rate. The gas products were analyzed with a Ke Chuang Gas
Chromatography 9800FP with a Porapak Q column.
The sorbitol conversion and selectivity S(i) of product i were
dened by the following formulas:
X1 ꢂ X2
conversion ¼
ꢃ 100%;
X1
niM0
6ðX1 ꢂ X2Þ
S
ðiÞ
¼
ꢃ 100%;
where X1 expressed the initial sorbitol weight and X2 expressed
the sorbitol weight aer reaction, ni represented the carbon
Fig. 1 XRD patterns of La-promoted Ni/ZrO2 catalysts by different Ni
moles of product i (GLY, EG, and 1,2-PG) and M0 was molar and La loadings.
weight of sorbitol.
>10%, the intensity of ZrO2 and NiO decreased remarkably. The
3. Results and discussion
3.1. Characterizations of catalysts
reduction of the diffraction peaks was possibly attributed to the
decrease of the relative content of ZrO2 on the catalysts. No
La2O3 diffraction peaks were observed even the loading was as
high as 20%, indicating that La2O3 existed as an amorphous
form and highly dispersed on ZrO2 surface. The weakness of the
NiO diffraction peak should be due to the good dispersion of
NiO on the catalyst surface, which was promoted by La2O3.
Compared to Ni/ZrO2, the NiO particle sizes in the Ni/La2O3/
ZrO2 catalysts were signicantly reduced, again evidencing Ni
stabilization by La2O3 addition. Fig. 1 and 2 also showed that
the catalysts had a good crystalline phase before reaction,
indicating that the support and Ni particles were unchanged
and demonstrating high hydrothermal stability.
Table 1 listed the texture properties of different catalysts. The
Brunauer–Emmett–Teller (BET) area and pore volume of ZrO2
were 40.9 m2 gꢂ1 and 0.20 mL gꢂ1, respectively, whereas for Ni/
ZrO2, with the introduction of Ni, the specic surface area and
pore volume of 10% Ni/ZrO2 catalyst decreased. However, the
average pore size of Ni/ZrO2 increased aer Ni introduction.
This might be because Ni entered into the inner channels of
ZrO2 aer impregnation, causing a decrease of specic surface
area. The addition of La2O3 did not obviously cause the specic
surface area and pore volume to decrease, indicating that it
could highly disperse on the ZrO2 surface.
To investigate the effect of La2O3 on the reduction charac-
teristics of catalysts, we investigated the H2 reduction behavior
through the H2-TPR technique (Fig. 3). There was a single peak
at ꢀ500 ꢁC for 10% Ni/ZrO2, which corresponded to the
reduction of NiO particles. Aer adding La2O3, the reduction
peaks of NiO shied to lower temperatures, showing that La2O3
weakened the interaction between Ni species and the ZrO2
support. Upon further increase of the La2O3 loadings, the
reduction peak of NiO was further reduced to ꢀ400 ꢁC.
The XRD patterns of different catalysts were presented in
Fig. 1. The diffraction patterns of NiO found in 10% Ni/ZrO2 had
diffraction peaks at 37.3ꢁ and 43.3ꢁ, corresponding to the (111)
and (200) crystal planes of NiO, respectively. With La2O3 intro-
duction, the intensity of NiO decreased, which indicated that
La2O3 improved the dispersion of NiO by the enhanced inter-
action between Ni2+ and La2O3.30 With increasing La2O3 in the
catalysts, no obvious changes in NiO diffraction patterns could
be observed. However, with further increasing La2O3 loading to
Table 1 Textural properties and H2-chemisorption of La-promoted Ni/ZrO2 catalysts
Catalyst
SBETa (m2 gꢂ1
)
Pore volumeb (mL gꢂ1
)
Pore diameterc (nm)
DNid (%)
dNiOe (nm)
ZrO2
10% Ni/ZrO2
40.9
36.8
35.2
36.4
33.5
30.9
37.9
35.7
31.8
0.20
0.17
0.18
0.17
0.17
0.16
0.19
0.17
0.12
16.7
19.8
18.2
20.1
18.8
19.0
19.6
20.3
22.4
—
5.3
10.7
9.6
10.8
10.7
12.6
11.0
7.9
—
25
8
11
7
8
5
9
15
10% Ni/2% La2O3/ZrO2
10% Ni/5% La2O3/ZrO2
10% Ni/10% La2O3/ZrO2
10% Ni/20% La2O3/ZrO2
5% Ni/5% La2O3/ZrO2
7.5% Ni/5% La2O3/ZrO2
12.5% Ni/5% La2O3/ZrO2
a
Calculated by using the BET method. b Determined by the cumulative volume between P/P0 ¼ 0.03–0.99. c Calculated from the desorption branch
by using the Barrett–Joyner–Halenda (BJH) method. d Ni dispersion values were obtained by CO-TPD. e Diameters of NiO particles were calculated
by using the Scherrer formula ((111) crystal plane).
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 3993–4001 | 3995

View Article Online
RSC Advances
Paper
number of basic sites was reduced. However, for the Ni/La2O3/
ZrO2 catalysts, the number of weak basic sites remarkably
increased, suggesting that the La2O3 additive strengthened the
weak basicity.
3.2. Catalytic performance
Table 2 listed the results of sorbitol hydrogenolysis over Ni-
based catalysts with different contents of La2O3 under the
ꢁ
conditions of 220 C and 4 MPa H2 pressure. Throughout the
process, sorbitol was mostly converted to the target products.
By-products such as erythritol, xylitol, etc. were also observed,
and the data for those were not shown here. When a 10% Ni/
ZrO2 catalyst was used, a 66.6% sorbitol conversion with
a 23.7% total selectivity containing GLY, EG, and 1,2-PG was
achieved. With the addition of La2O3, sorbitol conversion and
total glycol selectivity were obviously increased. With increasing
La2O3 content from 2 wt% to 5 wt%, the sorbitol conversion
sharply rose from 63.5% to 96.8%. Meanwhile, the total selec-
tivity for the target products also rose from 37.3% to 74.8%.
However, when we continued to increase the La2O3 content
to 20 wt%, the sorbitol conversion and total selectivity dropped
to 87.5% and 62.2%, respectively. The above results indicated
that the alkaline La2O3 promoted catalytic performance for
obtaining the target products, which was conducive to C–C
scission in the retro-aldol condensation and also helpful for
improving the desired product selectivity. Sorbitol and inter-
mediates might also be adsorbed on the alkaline sites to
improve the reaction.31 However, the excessive La2O3 promoted
abundant reverse C–C scission and formed other by-products,
which led to a decrease in the selectivity of the desired
product. Therefore, it could be concluded that 5% La2O3
content was appropriate for optimizing the reaction (entries 2–
5). In addition, different contents of Ni were investigated by
using 5% La2O3 as a base (entries 6–9). Increasing the Ni
content from 5% to 12.5% led to increased sorbitol conversion
but the total selectivity of target products remained at ꢀ66%. It
was obvious that the basic La2O3 plays a more profound role in
glycol formation from sorbitol hydrogenolysis than Ni. Given
the best performance in sorbitol hydrogenolysis to the target
products, the optimized 10% Ni/5% La2O3/ZrO2 catalyst was
chosen to investigate the inuence of temperature, reaction
time, and H2 pressure.
Fig. 2 SEM image of 10% Ni/5% La2O3/ZrO2 catalyst before reaction.
Fig. 3 H2-TPR profiles of catalysts with different La promoters.
The basicities of the Ni-based catalysts with different La2O3
loadings were displayed in Fig. 4. The desorption peak of CO2 at
ꢁ
ꢁ
ꢁ
150 C for ZrO2 and those obtained at 750 C and 850 C for
La2O3 indicated that ZrO2 and La2O3 presented weak basicity
and strong basicity, respectively. Because Ni was supported on
ZrO2 without La2O3, the weak basicity was kept but the total
As shown in Fig. 5A, the sorbitol conversion increased from
26% toꢁ100% as the reaction temperature increased from 200 ꢁC
to 240 C, which indicated that temperature had an important
effect on sorbitol conversion. The selectivity of GLY, EG, and
1,2-PG increased as the reaction temperature increased from
200 ꢁC to 220 ꢁC.ꢁHowever, when the reaction temperature was
higher than 220 C, the selectivity of GLY and target products
total selectivity decreased while the selectivity of EG and 1,2-PG
still increased, this result demonstrated that GLY acted as the
key intermediate for further conversion to EG and 1,2-PG
possibly through dehydration and hydrogenation.
Fig. 5Bshowed the effect of H2 pressure. The conversion of
sorbitol slightly reduced with the increase of H2 pressure, while
the selectivities of the target products were almost unchanged.
Fig. 4 CO2-TPD profiles of catalysts with different La loadings.
3996 | RSC Adv., 2020, 10, 3993–4001
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 2 Hydrogenolysis of sorbitol to glycols over different Ni/La2O3/ZrO2 catalystsa
Product selectivity (C mol%)
Entry
Catalyst
Conversion (%)
GLY
EG
1,2-PG
Total selectivity (%)
1
2
3
4
5
6
7
8
9
10% Ni/ZrO2
66.6
63.5
96.8
94.2
87.5
55.8
75.4
76.6
71.3
13.1
18.3
27.7
24.5
21.2
30.2
30.3
32.1
31.6
3.9
6.7
13.8
26.8
28.4
26.8
21.7
21.8
20.4
20.8
23.7
37.3
74.8
70.8
62.2
65.3
66.4
66.1
66.3
10% Ni/2% La2O3/ZrO2
10% Ni/5% La2O3/ZrO2
10% Ni/10% La2O3/ZrO2
10% Ni/20% La2O3/ZrO2
5% Ni/5% La2O3/ZrO2
7.5% Ni/5% La2O3/ZrO2
10% Ni/5% La2O3/ZrO2
12.5% Ni/5% La2O3/ZrO2
5.2
20.3
17.9
14.2
13.3
14.3
13.6
13.9
a
Reaction conditions: 10% sorbitol contained in an aqueous solution of 40 mL; catalyst amount ¼ 1 g; temperature ¼ 220 ꢁC; initial H2 pressure ¼
4 MPa. Entries 1–5 are for 4 h; entries 6–9 are for 2 h. GLY ¼ glycerol; EG ¼ ethylene glycol; 1,2-PG ¼ 1,2-propylene glycol.
This indicated that the target products obtained by sorbitol 31% to 10% as reaction time increased from 0.5 h to 8 h. At 4 h,
hydrogenolysis were insensitive to the H2 pressure.
the selectivities of EG and 1,2-PG reached the maximum values.
The effect of reaction time on sorbitol conversion and At the same time, xylitol and erythritol were also detected in the
product selectivity was presented in Fig. 5C. The sorbitol liquid phase, obtaining the selectivity of 5.6% and 1.8%,
conversion increased with the reaction time, reaching almost respectively. Trace ethanol was also detected in the liquid
100% at 4 h. Meanwhile, the selectivity of GLY decreased from phase. There were also a certain amount of gas products such as
Fig. 5 Influence of reaction temperature (A), initial H2 pressure (B), and reaction time (C) on sorbitol hydrogenolysis to glycols over 10% Ni/5%
La2O3/ZrO2. Reaction conditions: 10% sorbitol contained in an aqueous solution of 40 mL; catalyst amount ¼ 1 g; temperature ¼ 220 ꢁC; initial
H2 pressure ¼ 4 MPa; time ¼ 2 h. “TS” means total selectivity of GLY, EG, and 1,2-PG.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 3993–4001 | 3997

View Article Online
RSC Advances
Paper
Fig. 6 Stability and regeneration test of 10% Ni/5% La2O3/ZrO2 in
sorbitol hydrogenolysis to glycols. Reaction conditions: 10% sorbitol
contained in an aqueous solution of 40 mL; catalyst amount ¼ 1 g;
temperature ¼ 220 ꢁC; initial H2 pressure ¼ 4 MPa; time ¼ 4 h.
Fig. 8 XRD patterns of the fresh and the used 10% Ni/5% La2O3/ZrO2
after the fifth used.
CH4, CO, and CO2. Unfortunately, the CH4 selectivity increased
to 27% at 8 h. With increasing of reaction time, the yields of
small molecule monohydric alcohols and gas-phase products
increased. In a word, the optimized conditions were determined
as: initial H2 pressure of 4 Mpa, reaction temperature of 220 ꢁC
and reaction time of 2 h.
The stability of the catalyst was of great importance for
industrial applications. A stability test was conducted using
a 10% Ni/5% La2O3/ZrO2 catalyst and the optimized reaction
conditions (Fig. 6). Aer reaction, the catalyst was separated via
ltration and directly used for the next run without any treat-
ment. As shown in Fig. 6, the sorbitol conversion decreased
rapidly from 98% to 52% aer ve cycles, while the target
products total selectivity slightly decreased to 60%. Aer
regeneration of the used catalyst by calcination and reduction,
the catalytic performance could be regenerated partially based
on the sorbitol conversion and glycols selectivity. The leaching
of Ni2+ was 1.5 wt%, 0.8 wt%, 0.3 wt%, 0.1 wt%, 0.03 wt%,
respectively, aer each run. For La3+, the leaching was 1.9 wt%,
1.2 wt%, 0.8 wt%, 0.6 wt%, 0.3 wt%, respectively. It was inferred
that metal elution might be one factor of catalyst deactivation.26
Fig.
9 TG-DTG profiles of 10% Ni/5% La2O3/ZrO2 catalyst after
sorbitol hydrogenolysis and drying at 100 ꢁC.
According to Fig. 7, the La XPS spectra of the fresh and used
catalysts indicated that their chemical valence was almost the
same, showing the stable La species under the current
Fig. 7 XPS spectra of La 3d (A) and Ni 2p (B) in the fresh and the fifth used 10% Ni/5% La2O3/ZrO2 catalysts.
3998 | RSC Adv., 2020, 10, 3993–4001
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Fig. 10 TEM images of fresh and used catalysts: (a and b) fresh and (c and d) after the fifth used 10% Ni/5% La2O3/ZrO2.
hydrothermal conditions. For the Ni species, partial oxidation caused by carbon deposition. The main reason for deactivation
took place aer the h cycle. It also could be seen from XRD of the catalyst and decrease of target product selectivity was due
pattern (Fig. 8) that the catalyst aer the h cycle was mainly to the metal leaching.
contained La(OH)3 and La2(CO3)3 phase.32,33
Based on the distributions of products in our case and the
TG analysis27 (Fig. 9) showed that weight loss at 350 C was previous reports,11–14 we proposed reaction pathway for sorbitol
ꢁ
2ꢂ
attributed to the removal of surface hydroxyl group and CO3
.
hydrogenolysis to 1,2-PG (Scheme 1) and possible pathways for
TEM analysis was also carried out for the same catalyst aer target products formation from sorbitol hydrogenolysis over
reaction (Fig. 10), there were no obvious change of the catalysts, 10% Ni/5% La2O3/ZrO2 catalyst in Scheme 2. In Scheme 1,
so it could be concluded that catalyst deactivation was not sorbitol was dehydrogenated at C2-position, and then
Scheme 1 Reaction pathway for sorbitol hydrogenolysis to 1,2-PG over the Ni based catalyst.34
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 3993–4001 | 3999

View Article Online
RSC Advances
Paper
Scheme 2 Possible pathways for sorbitol hydrogenolysis to small molecular glycols.
glyceraldehyde was formed under the action of base. C]O of
4. Conclusions
glyceraldehyde was easily hydrogenated to GLY, and nally 1,2-
PG was formed under the action of H2, Ni and base.
Ni/La2O3/ZrO2 catalysts were prepared via initial wetness
impregnation and used as the metal–base bi-functional cata-
lysts for the hydrogenolysis of sorbitol into target small
molecular polyol products. The addition of La2O3 improved the
basicities of the Ni-based catalyst and promoted dispersion of
Ni particles by strong interactions of Ni and La2O3. By opti-
mizing the sorbitol hydrogenolysis process, we could obtain the
highest target products selectivity of 74.8% over 10% Ni/5%
In the current case, metal Ni and basic La2O3 synergistically
promoted the rate-determined sorbitol dehydrogenation to
form the corresponding C6 aldehyde sugar. The C6 aldehyde
sugar could be further isomerized to the relative ketone sugar
with the aid of the base La2O3. Aerward, the aldehyde and
ketone sugars were hydrogenolyzed to the C3–C4 aldehyde and
a ketone containing polyols intermediates via the retro-aldol
reaction, where the C3 counterpart was hydrogenated to GLY
and the C4 intermediate was further converted into the C2 gly-
colaldehyde via a similar retro-aldol reaction followed by
hydrogenation to EG. However, GLY could be further converted
into C1 and C2 intermediates, which were nally transferred to
CH4 and EG. In addition, decarbonylation and aqueous phase
reformation could not be ruled out because trace amounts of
CO and CO2 were detected besides CH4. During this process, the
synergy between metallic Ni and basic La2O3 in the catalyst
played an essential role in obtaining the target C2–C3 polyols
with high selectivity.
ꢁ
La2O3/ZrO2 catalyst, under the conditions of 220 C, 4 MPa H2
pressure and 4 h of reaction time. The possible metal–alkali
coordination mechanism and possible pathways for target
products formation were proposed. Ni metal promoted dehy-
drogenation and hydrogenation reaction, basic promoter was
benet to C–C bond cleavage, and aerwards polyols were
produced by a series of small carbonyl molecules hydrogenation
reactions.
Conflicts of interest
There are no conicts to declare.
4000 | RSC Adv., 2020, 10, 3993–4001
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
16 R. Mane, S. Patil, M. Shirai, S. Rayalu and C. Rode, Appl.
Catal., B, 2017, 204, 134–146.
Acknowledgements
17 X. Chen, X. Wang, S. Yao and X. Mu, Catal. Commun., 2013,
39, 86–89.
This work is nancially supported by the National Key R&D
Program of China (2018YFB1501402), the Natural Science
Foundation of Guangdong Province (2017A030308010), the
National Natural Science Foundation of China (51576199 and
51536009), the DNL Cooperation Fund, CAS (DNL180302), the
“Transformational Technologies for Clean Energy and
Demonstration”, Strategic Priority Research Program of the
Chinese Academy of Sciences (No. XDA21060102), and the Local
Innovative and Research Teams Project of Guangdong Pearl
River Talents Program (2017BT01N092).
´
18 A. F. Trasarti, N. M. Bertero, C. R. Apesteguıa and
A. J. Marchi, Appl. Catal., A, 2014, 475, 282–291.
19 Z. Huang, J. Chen, Y. Jia, H. Liu, C. Xia and H. Liu, Appl.
Catal., B, 2014, 147, 377–386.
20 M. Banu, S. Sivasanker, T. M. Sankaranarayanan and
P. Venuvanalingam, Catal. Commun., 2011, 12, 673–677.
´
˜
21 L. S. Ribeiro, N. Rey-Raap, J. L. Figueiredo, J. J. Melo Orfao
and M. F. R. Pereira, Cellulose, 2019, 26, 7337–7353.
22 C. Liu, Z. Zhang, X. Zhai, X. Wang, J. Gui, C. Zhang, Y. Zhu
and Y. Li, New J. Chem., 2019, 43, 3733–3742.
References
23 Z. Zhou, J. Zhang, J. Qin, D. Li and W. Wu, Russ. J. Phys.
Chem. A, 2018, 92, 456–465.
24 S. Li, Y. Zan, Y. Sun, Z. Tan, G. Miao, L. Z. Kong and Y. Sun, J.
Energy Chem., 2019, 28, 101–106.
1 G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106,
4044–4098.
2 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem.,
Int. Ed. Engl., 2010, 46, 7164–7183.
`
25 M. Riviere, N. Perret, D. Delcroix, A. Cabiac, C. Pinel and
3 S. Ye and J. Cheng, Cheminform, 2003, 83, 1–11.
M. Besson, Catalysts, 2018, 8, 331.
´
4 J. C. Serrano-Ruiz, R. Luque and A. Sepulveda-Escribano,
`
26 M. Riviere, N. Perret, D. Delcroix, A. Cabiac, C. Pinel and
Chem. Soc. Rev., 2011, 40, 5266–5281.
5 X. Tong, M. A. Yang and L. I. Yongdan, Appl. Catal., A, 2010,
385, 1–13.
6 J. J. Bozell and G. R. Petersen, Green Chem., 2010, 12, 539–
554.
M. Besson, ACS Sustainable Chem. Eng., 2018, 6, 4076–4085.
27 R. Sun, T. Wang, M. Zheng, W. Deng, J. Pang, A. Wang,
X. Wang and Z. Tao, ACS Catal., 2015, 5, 874–883.
28 X. Wang, A. K. Beine, P. J. C. Hausoul and R. Palkovits,
ChemCatChem, 2019, 11, 4123–4129.
7 K. Wang, M. C. Hawley and T. D. Furney, Ind. Eng. Chem. Res.,
1995, 34, 3766–3770.
8 E. L. Kunkes, D. A. Simonetti, R. M. West, J. C. Serrano-Ruiz,
29 O. V. Manaenkov, O. V. Kislitsa, V. G. Matveeva,
E. M. Sulman, M. G. Sulman and L. M. Bronstein, Front.
Chem., 2019, 7, 834.
¨
C. A. Gartner and J. A. Dumesic, Science, 2008, 322, 417.
30 X. F. Cao, Q. Zhang, D. Jiang, Q. Y. Liu, L. L. Ma, T. J. Wang
and D. B. Li, Chin. J. Chem. Phys., 2015, 28, 338–344.
31 X. Chen, X. Wang, S. Yao and X. Mu, Catal. Commun., 2013,
39, 86–89.
32 Q. Mu and Y. Wang, J. Alloys Compd., 2011, 509, 396–401.
33 C. Sun, G. Xiao, H. Li and L. Chen, J. Am. Ceram. Soc., 2007,
90, 2576–2581.
9 A. M. Ruppert, K. Weinberg and R. Palkovits, Angew. Chem.,
Int. Ed., 2012, 51, 2564–2601.
10 H. Imamura, T. Tanaka, Y. Sakata and S. Tsuchiya, J. Alloys
Compd., 1999, 293–295, 919–922.
11 H. Liu, Z. Huang, H. Kang, X. Li, C. Xia, J. Chen and H. Liu,
Appl. Catal., B, 2018, 220, 251–263.
12 Y. Jia and H. Liu, Catal. Sci. Technol., 2016, 6, 7042–7052.
13 Z. Long, J. Zhou, C. Hong, M. Zhang, Z. Sui and X. Zhou,
Korean J. Chem. Eng., 2010, 27, 1412–1418.
34 C. Montassier, D. Giraud and J. Barbier, in Studies in Surface
Science and Catalysis, ed. M. Guisnet, J. Barrault, C.
´
Bouchoule, D. Duprez, C. Montassier and G. Perot,
14 J. Sun and H. Liu, Green Chem., 2011, 13, 135–142.
15 J. J. Zhang, F. Lu, W. Q. Yu, J. Z. Chen, S. Chen, J. Gao and
J. Xu, Catal. Today, 2014, 234, 107–112.
Elsevier, 1988, vol. 41, pp. 165–170.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 3993–4001 | 4001