Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on Ni/C and basic oxide-promoted Ni/C catalysts

Article abstract of DOI:10.1016/j.cattod.2013.12.040

The selective hydrogenolysis of xylitol to ethylene glycol and propylene glycol was examined on Ni/C catalysts in the presence of solid bases, e.g. Ca(OH)₂ and CeO₂, physically mixed with or co-supported with Ni on C. Compared with Ru/C, the Ni/C catalysts were more selective to the two target glycols under identical conditions, apparently as a result of their lower hydrogenation activity and consequently favored the CC cleavage of xylose intermediate by the base catalyst over its competitive hydrogenation on the Ni particles. Noticeably, the presence of the solid bases rendered the Ni particles resistant to leaching and sintering, and thus stable in the xylitol hydrogenolysis. Supporting the solid bases, especially CeO₂ and CaO, with the Ni particles on C led not only to a reduction in the amount of solid bases required, but also efficient formation of the two glycols with negligible lactic acid. For instance, on Ni-CaO/C (at a CaO/Ni molar ratio of 0.66), the combined selectivity to ethylene glycol and propylene glycol, together with glycerol, reached 69.5% at nearly 100% xylitol conversion at 473 K, 4.0 MPa H₂. These features of the basic oxide-promoted Ni catalysts show their promising advantages over the noble Ru catalysts, upon optimization of their compositions and structures, and the reaction parameters, for the efficient hydrogenolysis of xylitol and other lignocellulose-derived polyols to produce the two target glycols.

Full text of DOI:10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol
and propylene glycol on Ni/C and basic oxide-promoted Ni/C catalysts
Jiying Sun, Haichao Liu∗
Beijing National Laboratory for Molecular Sciences, State Key Laboratory for Structural Chemistry of Stable and Unstable Species, College of Chemistry and
Molecular Engineering, Peking University, Beijing 100871, China
a r t i c l e i n f o
a b s t r a c t
Article history:
The selective hydrogenolysis of xylitol to ethylene glycol and propylene glycol was examined on Ni/C
catalysts in the presence of solid bases, e.g. Ca(OH)2 and CeO2, physically mixed with or co-supported
with Ni on C. Compared with Ru/C, the Ni/C catalysts were more selective to the two target glycols under
identical conditions, apparently as a result of their lower hydrogenation activity and consequently favored
the C C cleavage of xylose intermediate by the base catalyst over its competitive hydrogenation on the
Ni particles. Noticeably, the presence of the solid bases rendered the Ni particles resistant to leaching
and sintering, and thus stable in the xylitol hydrogenolysis. Supporting the solid bases, especially CeO2
and CaO, with the Ni particles on C led not only to a reduction in the amount of solid bases required,
but also efficient formation of the two glycols with negligible lactic acid. For instance, on Ni-CaO/C (at a
CaO/Ni molar ratio of 0.66), the combined selectivity to ethylene glycol and propylene glycol, together
with glycerol, reached 69.5% at nearly 100% xylitol conversion at 473 K, 4.0 MPa H2. These features of
the basic oxide-promoted Ni catalysts show their promising advantages over the noble Ru catalysts,
upon optimization of their compositions and structures, and the reaction parameters, for the efficient
hydrogenolysis of xylitol and other lignocellulose-derived polyols to produce the two target glycols.
© 2014 Elsevier B.V. All rights reserved.
Received 25 November 2013
Received in revised form
20 December 2013
Accepted 27 December 2013
Available online xxx
Keywords:
Selective hydrogenolysis
Ni catalyst
Solid base
polyols
Propylene glycol
Ethylene glycol
1. Introduction
xylitol conversion in the presence of Ca(OH)2 under relatively mild
conditions of 473 K and 4.0 MPa H2. In term of the Ni catalysts,
Naturally occurring biomass provides the only practical source
of renewable liquid fuels and organic chemicals [1–4]. In this
context, due to their rich chemistry and large availability from
non-edible hemicellulose and cellulose, xylitol and sorbitol are
considered the key primary building blocks in the synthesis of
various value-added chemicals [5–8]. One such example is their
catalytic hydrogenolysis to ethylene glycol and propylene glycol,
which provides a promising route for the sustainable production of
the two important commodity glycols that are manufactured today
in industry from petroleum-based ethylene and propylene via their
epoxide intermediates [9–17].
Recent efforts have demonstrated the efficacy of supported Ru
and Ni catalysts in the presence of basic promoters (e.g. Ca(OH)2)
for the hydrogenolysis of xylitol and sorbitol to ethylene glycol and
propylene glycol [9–15]. Zhou et al. [9] reported an 85.7% sorbitol
conversion and 51.3% selectivity to the two glycols on a carbon
nanofiber-supported Ru catalyst at 493 K and 8.0 MPa H2. We previ-
ously [10] studied the xylitol hydrogenolysis on Ru/C and achieved
a combined selectivity of ∼61% to the two glycols at nearly 100%
Yuan and co-workers [11] achieved sorbitol conversion of above
90% and selectivity to the two glycols of 55–60% on Ce-promoted
Ni/Al2O3 catalysts with Ca(OH)2 at 513 K and 7.0 MPa. Banu et al.
[12] reported that Ni–NaY catalyzes the sorbitol hydrogenolysis in
the presence of Ca(OH)2 to the two glycols with a total selectivity of
76% at 75% sorbitol conversion at 493 K and 6.0 MPa H2. Sotak et al.
[14] found that a Ni2P/C catalyst in the presence of Ba(OH)2 is effi-
cient for the xylitol hydrogenolysis, providing a 71.4% selectivity
to the two glycols at 99% conversion under relatively mild condi-
tions (473 K and 4.0 MPa H2). Chen et al. [15] used a bifunctional
co-precipitated Ni-MgO catalyst with a Ni/Mg molar ratio of 3/7
in sorbitol hydrogenolysis, and at 473 K and 6.0 MPa H2, obtained
80.8% selectivity to ethylene glycol, propylene glycol and glycerol
at 67.8% conversion.
These previous studies show that the Ni catalysts exhibit com-
parable selectivities to the two target glycols to those on the Ru
catalysts. However their potential as efficient polyol hydrogenol-
ysis catalysts is limited due to their inferior activities as they
require harsher reaction conditions (i.e., higher reaction temper-
atures or H2 pressures) and show lower stabilities that need to be
further improved. Meanwhile, these studies apparently encounter
low selectivities to the two glycols particularly from the viewpoint
of industrial practice, which is maybe related, at least partly, to the
∗
Corresponding author.
E-mail address: hcliu@pku.edu.cn (H. Liu).
0920-5861/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.12.040
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
2
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
Table 1
Activities and selectivities in xylitol hydrogenolysis on Ni/C with different loadings in the presence of Ca(OH)2, and average diameters of Ni particles on Ni/C. a
Loading
Activity (h−1
)
Selectivity (on a carbon basis, %)
Diameter (nm)
Ethylene glycol
Propylene glycol
Glycerol
Lactic acid
Arabitol
Threitol
dXRD
dTEM
c
2.4%
4.6%
7.3%
11.4
17.2
24.4
31.3
21.7
32.0
30.8
31.1
24.7
33.7
32.5
31.8
n.d.b
n.d.
1.0
16.4
17.3
17.6
15.0
n.d.
3.2
3.9
5.4
n.d.
n.d.
1.5
2.2
–
11.2 3.4
9.1 1.9
8.5 1.5
8.8 1.5
7.9
6.8
6.6
10.5%
2.6
a
Reaction conditions: 473 K, 4.0 MPa H2, 40 g 10 wt% xylitol aqueous solution, 0.06–1 g Ni/C, 0.26 g Ca(OH)2, 1 h, ∼20% conversion.
b
c
Not detected.
The diffraction peak is too weak for estimating Ni particle size.
use of relatively large amount of the basic promoters, facilitating
grids. The average sizes of Ni particles were calculated by averaging
of more than 300 particles randomly distributed in the TEM images.
Ni loadings and the amount of Ni leaching into the reaction solution
were examined by ICP.
the competitive formation of lactic acid in the form of lactate and
consumption of the bases [10,16]. These issues are strongly rele-
vant to the polyol hydrogenolysis mechanism [10,18]. We recently
proposed that the xylitol hydrogenolysis to ethylene glycol and pro-
pylene glycol proceeds by its kinetically relevant dehydrogenation
of xylitol to xylose intermediate on the metal surfaces, and sub-
sequent base-catalyzed retro-aldol condensation of xylose to form
glycolaldehyde and glyceraldehyde, the intermediates for the two
glycols[10]. Theselectivityto thetwoglycolsis thereforeultimately
controlled by the relative rates between the hydrogenation of the
aldehyde intermediates and their competitive reactions with the
bases, reflecting the bifunctional nature of the xylitol hydrogenol-
ysis.
In this work, we studied the Ni/C catalyst for the xylitol
hydrogenolysis in the presence of a different solid bases such as
Ca(OH)2 and CeO2, aimed at tuning the relative rates between the
aforementioned competitive reactions on the metal surfaces and
basic sites. We found that the basic oxides (e.g. CaO and CeO2)-
promoted Ni/C catalysts exhibit high selectivities to ethylene glycol
and propylene glycol while the amount of the solid bases required
is significantly reduced and lactic acid is essentially not formed.
Xylitol hydrogenolysis reactions were carried out in a stainless
steel autoclave (100 ml) at a stirring speed of 800 rpm. Typically,
40 g of 10 wt% xylitol (99%, Alfa Aesar) aqueous solution, proper
amount of Ni catalysts (varied depending on xylitol conversion)
and also for the Ni/C case, solid base were introduced to the auto-
clave. Afterwards, the reactor was purged with H2 (>99.99%, Beijing
Huayuan) three times, and pressurized with H2 to 4.0 MPa and
heated to 473 K, which was kept constant during the reaction.
The reactant and liquid products, after silylation with hexame-
thyldisilazane (HDMS) and trimethylchlorosilane (TMSCl) (both
≥98.0%, Sinopharm Chemical) in pyridine (AR, Shantou Xilong
Chemical), were analyzed by gas chromatography (Agilent 7890A)
using a capillary column HP-1ms (30 m × 0.25 mm × 0.25 ␮m) and
a flame ionization detector. The detected liquid products included
ethylene glycol, propylene glycol, glycerol, lactic acid, threitol,
arabitol, and dehydroxy-pentitols (mainly 1,2,5-pentanetriol and
1,2,4,5-pentanetetraol), and dehydrated product hydroxyl furan.
Gas products, i.e. CH4 and CO2, were also detected in trace amounts,
and thus not discussed in this work. Xylitol conversion and product
selectivity are reported on a carbon basis, and xylitol reaction activ-
ity is reported as molar xylitol conversion rate per mole of metal
loaded per hour (h−1).
2. Experimental methods
The Ni/C and basic oxide-modified Ni/C (denoted as Ni-Oxide/C)
catalysts were prepared using the incipient wetness impregnation
method. After impregnation of C (AR, Beijing Dali Fine Chemical,
dried at 373 K in air for 12 h) with aqueous solutions of Ni(NO3)2
and, for Ni-Oxide/C catalysts, another metal nitrate (Ce(NO3)3,
La(NO3)3, Mg(NO3)2, Ca(NO3)2, Ba(NO3)2, NaNO3, Al(NO3)3, or
ZrO(NO3)2), the slurry was dried at room temperature until large
amount of water was evaporated and then the resulting sample was
dried at 383 K for at least 12 h. The catalyst precursor was reduced
in a flow of 20% H2/N2 at 673 K for 3 h.
ZrO2 was prepared hydrothermally as reported by Li et al. [19].
CeO2 was synthesized hydrothermally in a similar way. Briefly,
a Teflon inner vessel containing an aqueous solution of 0.4 M
(NH4)2Ce(NO3)6 (AR, Sinopharm Chemicals) and ∼4.0 M urea in a
stainless steel jacket was heated at 413 K for 24 h. The resulting
precipitates were filtered and washed thoroughly with deionized
water until the filtrate was neutral, followed by drying at 383 K
for 12 h. The as-prepared ZrO2, CeO2 and purchased Al2O3 (Condea
Chemie Gmbh) were calcinated at 673 K under a flowing air before
use as additives.
3. Results and discussion
Table
1 shows the activity and selectivity of the xylitol
hydrogenolysis at 473 K and 4.0 MPa H2 on four Ni/C catalysts
with different Ni loadings in the range 2.4–10.5 wt% in the pres-
ence of Ca(OH)2. The xylitol conversion was kept around 20% in
the kinetic-controlled regime. These catalysts were characterized
by XRD and TEM, as shown in Figs. S1 and 2 (in Supporting data),
respectively, and their Ni particle sizes were accordingly estimated.
As listed in Table 1, the average sizes estimated from the XRD pat-
terns (6–8 nm) were comparable to the data derived from the TEM
images (8–11 nm), revealing the similar sizes of the Ni particles on
the Ni/C catalysts especially with Ni loadings of 4.6 wt%, 7.3 wt% and
10.5 wt%. As the Ni loadings increased from 2.4 wt% to 10.5 wt%, the
activity (normalized per Ni atom) increased linearly from 11.4 h−1
to 31.3 h−1. Such increase in the activities with the Ni loadings are
due clearly not to the change in the Ni particle sizes, but most likely
to the increase in the surface density of the Ni particles on C, i.e. the
higher Ni particle density favors the dehydrogenation of xylitol, the
previously proposed rate-determining step involved in the polyol
hydrogenolysis reactions [10,16]. This Ni density effect on the activ-
ity, although the underlying reason is not clear, might be relevant
The catalysts were characterized by X-ray diffraction (XRD),
transmission electron microscopy (TEM), and inductive coupled
plasma emission spectroscopy (ICP). The XRD patterns were
obtained on a Rigaku D/MAX-2400 diffractometer using Cu Kɑ1
˚
radiation (ꢀ = 1.5406 A) operated at 40 kV and 100 mA. The TEM
to the effect on the adsorption configurations of xylitol and its C
bond activation on the Ni particles.
The Ni loadings also strongly influence the product selectivi-
ties in the xylitol hydrogenolysis. On 2.4 wt% Ni/C, the selectivities
H
images were taken on a Philips Tecnai F30 FEGTEM operated at
300 kV. Samples were prepared by uniformly dispersing the cata-
lysts in ethanol and then placing them onto carbon-coated copper
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
3
Fig. 2. Activities and selectivities to ethylene glycol, propylene glycol and lactic acid
for the five reaction cycles of xylitol hydrogenolysis on Ni/C. Reaction conditions:
473 K, 4.0 MPa H2, 1 h, 40 g 10 wt% xylitol aqueous solution, 0.2 g 7.3 wt% Ni/C, 0.26 g
Ca(OH)2.
Fig. 1. Dependence of product selectivities on xylitol conversions on Ni/C. Reaction
conditions: 473 K, 4.0 MPa H2, 1–3 h, 40 g 10 wt% xylitol aqueous solution, 0.05–0.8 g
7.3 wt% Ni/C, 0.26–0.6 g Ca(OH)2.
to the two target glycols, ethylene glycol and propylene glycol,
were 21.7% and 24.7%, respectively, which largely increased to
32.0% and 33.7% with increasing the Ni loading from 2.4 wt% to
4.6 wt%, and then remained almost unchanged at higher Ni load-
ings of 7.3 and 10.5% at similar xyiltol conversions (∼20%). Glycerol
was detected always at low selectivities, which increased to 2.6%
at the Ni loading of 10.5 wt%, while the selectivities to lactic acid
slightly varied between 15.0% and 17.6% in the whole range of the Ni
loadings. Accordingly, the combined selectivities to ethylene gly-
col, propylene glycol, and glycerol increased from 46.4% to 65.7%
with increasing the Ni loadings from 2.4% to 10.5 wt%. Such change
is consistent with the enhanced activities of the Ni/C catalysts at
higher Ni loadings and the consequent effect on the xylitol reac-
tion pathways. Ethylene glycol, propylene glycol and glycerol are
derived, according to the proposed xylitol hydrogenolysis mecha-
nism [10], from glycolaldehyde and glyceraldehyde intermediates
that are formed involving the kinetically relevant dehydrogena-
tion of xylitol to xylose on the metal surfaces, and its subsequent
retro-aldol condensation with bases. The higher activities of the
Ni/C catalysts at higher Ni loadings favor the hydrogenation of
the glycolaldehyde and glyceraldehyde intermediates and their
derivatives (e.g. pyruvaldehyde), which would otherwise undergo
condensation reactions to form unidentified products or react to
form glycolic acid and lactic acid under the basic conditions [10].
The enhanced activities also led to the increase in the selectivities
to arabitol and threitol monotonically from nearly zero to 7.6% with
increasing the Ni loading from 2.4 wt% to 10.5 wt%, as a result of the
improved hydrogenation and decarbonylation of the xylose inter-
mediate relative to its C C cleavage via retro-aldol condensation.
It is noted that the two target glycols, ethylene glycol and pro-
pylene glycol, are stable in the xylitol hydrogenolysis, as shown
by the representative results on 7.3 wt% Ni/C in Fig. 1. At 473 K and
4.0 MPa H2, the selectivities to ethylene glycol and propylene glycol
increased slightly from 29.7% and 34.8% to 31.9% and 35.4%, respec-
tively, as the conversions increased from 5% to 100%, corresponding
to a 67.3% yield for the two target glycols on Ni/C, which was higher
than the yield of 60.0% on Ru/C reported recently under the identical
reaction conditions [10]. Glycerol was always detected at very low
selectivities (∼1%, not shown in Fig. 1). The lactic acid selectivities
kept essentially constant, being around 18.3% even at 100% conver-
sion, which was lower by about 10% than the value (∼28%) on Ru/C.
The higher yields for the target glycols and lower yields for lactic
acid on Ni/C show its advantage over Ru/C for the selective xylitol
hydrogenolysis, together with its good stability and reusability, as
discussed below.
It is known that Ni particles are apt to leach and agglomer-
ate leading to their deactivation. The stability and recyclability
of Ni/C was thus examined at ca. 20% xylitol conversion in the
kinetic-control regime. In the recycling experiments, the catalyst
was washed thoroughly with deionized water and acetone, and
then filtered for the next cycle. As shown in Fig. 2, no significant
decline in the activities and selectivities was observed on 7.3 wt%
Ni/C after five successive cycles. This is consistent with the char-
acterization results for this catalyst. The Ni contents in aqueous
reaction solutions after each cycle were measured by ICP and no Ni
leaching was detected. TEM images in Fig. 3 show that the mean
diameters of the Ni particles and their size distributions remained
essentially unchanged (8.5 1.5 nm vs. 8.3 1.3 nm) after recycling
the 7.3 wt% Ni/C catalyst over 5 times. These results demonstrate
that the Ni/C catalysts are stable and reusable under the reaction
conditions employed in this work.
It is worth mentioning that the observed stability of the Ni/C
catalysts must be related to the presence of Ca(OH)2 and its basic-
ity. Under neutral conditions, the Ni particles on 7.3 wt% Ni/C
agglomerated rapidly from 6.8 nm into 30–40 nm in diameter in
the reaction solutions, as characterized by XRD (Fig. 4), and leached
significantly during the xylitol conversion. Obviously, Ca(OH)2 pro-
tected the Ni particles from agglomeration and leaching. Such a
protecting effect was also observed in the presence of other bases,
Mg(OH)2 and CaCO3, as shown in Fig. 4, maintaining the sizes of
the Ni particles (being around 6–7 nm) almost unchanged after the
xylitol hydrogenolysis, compared to the size (6.8 nm) before the
reaction. When CeO2 was used, the diffraction peak at 2Â = 44.8◦
became very weak, indicative of the dispersion of the Ni particles
on the C support surface although the underlying reason needs to
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
4
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
Fig. 3. TEM micrographs and Ni particle size of 7.3 wt% Ni/C before and after five reaction cycles (scale bar = 50 nm). (a) before reaction, 8.5 1.5 nm; (b) after five reaction
cycles, 8.3 1.3 nm.
4.1 h−1 and 4.2 h−1, respectively, and much lower than the value
(24.4 h−1) in the presence of Ca(OH)2 although the sizes of the Ni
particles were similar or for the case of CeO2 even smaller dur-
ing the reactions. Such inferior activities, compared to that in the
presence of Ca(OH)2, are due to the low pH values in the aque-
ous reaction solutions, analogous to the phenomenon observed on
Ru/C.
As for the product selectivity, Ni/C under neutral conditions
was selective to ethylene glycol (18.3%), propylene glycol (12.5%)
and glycerol (11.0%) with a combined selectivity of 41.8%, which
was much higher than that on Ru/C with only trace amounts of
the C2 and C3 products under identical conditions. Instead, xyli-
tol dominantly converted to other C5 polyols, e.g. arabitol, with a
be clarified. However, the Ni particles aggregated dramatically in
the presence of acidic Al2O3 or amphoteric ZrO2 with weak acid-
basicity, resembling the result under the neutral conditions. These
XRD results confirm the protecting role of bases in stabilizing the
Ni particles under the reaction conditions.
Upon the promotion of Mg(OH)2, CaCO3 and CeO2 as well as
Al2O3 and ZrO2, the xylitol hydrogenolysis was also examined at
473 K and 4.0 MPa H2 on 7.3 wt% Ni/C, for comparison with Ca(OH)2.
As shown in Table 2, the activity was 4.8 h−1 on Ni/C under neutral
condition, i.e. without addition of any promoter, which was similar
to that in the presence of acidic Al2O3 and ZrO2, being 4.0 h−1 and
3.5 h−1, respectively. The activity was not enhanced even in the
presence of basic Mg(OH)2, CaCO3 and CeO2, which was 3.5 h−1
,
Table 2
Effects of different basic and acidic additives on activities and selectivities in xylitol hydrogenolysis on Ni/C.a
Additive
Activity (h−1
)
Selectivity (on a carbon basis, %)
Ethylene glycol
Propylene glycol
Glycerol
Lactic acid
Arabitol
Threitol
Deoxy-
pentitolsb
–
ZrO2
Al2O3
CeO2
CaCO3
Mg(OH)2
4.8
3.5
4.0
4.2
4.1
3.5
18.3
17.0
16.9
26.3
26.6
27.3
12.5
13.4
12.2
30.9
31.1
32.7
11.0
13.9
13.9
4.0
6.0
6.0
n.d.
n.d.
n.d.
3.6
3.0
15.9
12.0
15.7
5.6
6.3
4.0
2.7
2.0
2.0
2.8
2.3
2.0
5.8
6.5
6.7
6.5
5.4
4.7
5.4
a
Reaction conditions: 473 K, 4.0 MPa H2, 3 h, 40 g 10 wt% xylitol aqueous solution, 0.2 g 7.3 wt% Ni/C, 0.1 g additive, ∼15% xylitol conversion.
b
Including 1,2,5-pentanetriol and 1,2,4,5-pentanetetrol.
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
5
selectivities to arabitol and threitol decreased significantly from
18.6% to 6.0-8.6% in the presence of CeO2, CaCO3 and Mg(OH)2.
The combined selectivities to ethylene glycol, propylene glycol and
glycerol are comparable to the data (64.3%) in the presence of
Ca(OH)2, but use of CeO2, CaCO3 and Mg(OH)2, due to their weaker
basicity, led to a dramatic decline in the selectivities to lactic acid by
3–5 folds (3.0–5.4% vs. 17.6%), showing the feasibility for control-
ling the product distributions by tuning the basicity in the xylitol
hydrogenolysis. Based on these results, we next further reduced the
use of the solid bases by directly modifying the Ni/C catalysts with
basic oxides and tune the product selectivities.
Table 3 displayed the activities and selectivities of the xylitol
hydrogenolysis at ca. 15% conversion on a series of 2.4 wt% Ni/C
catalysts modified with equimolar amounts of oxides (designated
as Ni-Oxide/C) at 473 K and 4.0 MPa H2. Ni-CaO/C was selective
to ethylene glycol, propylene glycol and glycerol with a combined
selectivity of 72.8%, which was higher than the aforementioned best
result (around 65%) on Ni/C in the presence of Ca(OH)2 at similar
conversions (Table 1). Moreover, less than 1% selectivity of lactic
acid was produced, in contrast to about 17% of lactic acid on Ni/C
when Ca(OH)2 was separately added into the reaction solutions,
although the activity of Ni-CaO/C was lower (7.0 h−1 vs. 11.4 h−1), as
expected by its lower amount of CaO used. Since CaO is slightly sol-
uble in water, two control experiments were performed to examine
whether CaO on Ni-CaO/C worked homogeneously or heteroge-
neously. The equal amount of CaO (∼0.02 g) was used by physically
mixing with 2.4 wt% Ni/C, as done in the experiments described in
Table 1. The activity decreased to 1.6 h−1, and more significantly,
the selectivity to ethylene glycol, propylene glycol and glycerol
decreased to 48% with a much higher selectivity to lactic acid, simi-
lar to the result in Table 1. In contrast, when extra 0.02 g of CaO was
added together with Ni-CaO/C into the reaction solution, the activ-
ity and product selectivity did not altered significantly, except that
the selectivity to glycerol declined from 11.0% to 8% concurrently
with the increase in the selectivity to lactic acid to about 3%. These
control experiments confirm that CaO on Ni-CaO/C acts as a solid
base in the xylitol hydrogenolysis. For comparison, Ru-CaO/C was
also examined, showing that xylitol was transformed mainly into
arabitol and threitol with a combined selectivity of 54.5%, and only
trace amounts the target glycols were detected, as a result of the
prevailing hydrogenation of the xylose intermediate on the very
active Ru surface over its C C bond cleavage via the retro-aldol
condensation on CaO base, reflecting the competition between the
two steps [10].
Fig. 4. XRD patterns of Ni/C (7.3 wt% Ni loading) catalysts before and after xylitol
hydrogenolysis in the absence and presence of Ca(OH)2, Mg(OH)2, CaCO3, CeO2, ZrO2
and Al2O3 at 473 K and 4.0 MPa H2.
selectivity of 80.4% on Ru/C [10], which was much higher than the
result (15.9%) on Ni/C, reflecting the lower activity of Ni/C than
Ru/C and the consequent effect on the reaction pathways. Accord-
ing to the proposed reaction mechanism, the lower hydrogenation
activity of Ni/C favored the retro-aldol condensation of the xylose
intermediate over its competitive hydrogenation reaction, lead-
ing to higher extent of its C C bond cleavage to ultimately form
the C2 and C3 products. Deoxypentitols, mainly pentane-1,2,5-triol
and pentane-1,2,4,5-tetrol, were also detected on Ni/C, which were
likely formed, by referring to the similar reaction of glycerol conver-
sion to propylene glycol [20], involving dehydration of the xylose
intermediate and subsequent hydrogenation, in competition with
the retro-aldol condensation of xylose. Addition of both ZrO2 and
Al2O3 exerted negligible effect on the product distributions on Ni/C,
as shown in Table 2. However, when CeO2 was added, even with
known weak basicity, the total selectivity to ethylene glycol (26.3%),
propylene glycol (30.9%) and glycerol (4.0%) increased sharply from
41.8% to 61.2%, which further increased to 63.7% and 66.0% in the
presence of CaCO3 and Mg(OH)2, respectively. Concurrently, the
Similar promoting effects on the xylitol hydrogenolysis were
also observed for the other basic oxides, including MgO, BaO, CeO2
and La2O3 (Table 3). Characterization of these Ni-Oxide/C catalysts
by XRD before and after the reactions, as shown in Fig. 5 reveals that
Table 3
Activities and selectivities in xylitol hydrogenolysis on oxide-modified Ni/C and CaO-modified Ru/C.a
Catalyst
Activity (h−1
)
Selectivity (on a carbon basis, %)
Ethylene glycol
Propylene glycol
Glycerol
Arabitol
Threitol
Deoxy-pentitolsb
Ni-CaO/C
Ru-CaO/C
Ni-BaO/C
Ni-CeO2/C
Ni-La2O3/C
Ni-MgO/C
Ni-Al2O3/C
Ni-ZrO2/C
Ni/C
7.0
103
8.2
9.7
6.6
28.3
trace
26.7
25.6
23.2
22.4
n.d.
33.5
trace
30.7
31.6
29.8
27.3
n.d.
11.0
trace
9.6
3.8
4.2
6.7
45.2
7.4
5.0
5.6
n.d.
trace
12.6
n.d.
trace
9.3
trace
3.6
6.3
n.d.
6.9
8.7
8.0
n.d.
4.3
2.6c
1.8c
1.9c
1.0c
3.7
n.d.
n.d.
trace
n.d.
n.d.
n.d.
n.d.
n.d. (40.0)d
n.d.(34.0)d
n.d.
n.d.
n.d.
n.d.
n.d.
a
Reaction conditions: 473 K, 4.0 MPa H2, 40 g 10 wt% xylitol aqueous solution, 0.4 g 2.4 wt% Ni/C modified with equal mol of different oxides, 3 h; 0.1 g 4 wt% Ru/C modified
with equal mol of CaO, 1 h. ∼15% xylitol conversion.
b
Including 1,2,5-pentanetriol and 1,2,4,5-pentanetetrol.
At <4% xylitol conversion.
Data in the parenthesis represents the selectivity to hydroxyl furan.
c
d
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
6
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
Fig. 5. XRD patterns of Ni/C catalysts modified with different oxides of Al2O3, ZrO2, CaO, MgO, La2O3, CeO2, BaO before (a) and after (b) xylitol hydrogenolysis at 473 K and
4.0 MPa H2.
the Ni particles modified with the basic oxides did not agglomerate
essentially, reflecting the protecting role of the bases on the Ni par-
ticles dispersed on C, as found for Ni/C physically mixed with the
bases (Fig. 4). Including Ni-CaO/C, their activities were much higher
than the un-promoted Ni/C (1.0 h−1), and they decreased in the
order of Ni-CeO2/C > Ni-BaO/C > Ni-CaO/C > Ni-La2O3/C > Ni-MgO/C,
being 9.7 h−1, 8.2 h−1, 7.0 h−1, 6.6 h−1, and 2.6 h−1, respectively.
This order, except CeO2, agrees approximately with the decrease
in their relative basicity, demonstrating the basicity effects on
the Ni activities for the kinetically-relevant xylitol dehydrogena-
tion. It is noted that the formation of lactic acid was negligible
(with < 1% selectivity) on these basic oxide-modified Ni/C catalysts.
The selectivity to the sum of C2 and C3 products (i.e. ethylene glycol,
propylene glycol and glycerol) was 67.0% on Ni-BaO/C, similar to the
value (i.e. 72.8%) on Ni-CaO/C, higher than the value on Ni-CeO2/C,
Ni-La2O3/C and Ni-MgO/C, being 61.0%, 57.2% and 53.4% respec-
tively. Clearly, all these basic oxides are efficient for catalyzing the
retro-aldol condensation reaction and the C C bond cleavage to
form the C2 and C3 products, notwithstanding the difference in
their selectivities. For comparison, acidic oxide such as Al2O3 and
amphoteric oxide such as ZrO2 were also supported together with
Ni on C to examine their effects. However, ZrO2 and Al2O3 failed
to stabilize the Ni particles on C, as evidenced from the narrow
and sharp Ni diffraction peaks after the reaction (Fig. 5), similar
to the phenomenon when they were added physically with Ni/C
(Fig. 4). ZrO2 and Al2O3 promoted the Ni activities (1.9 h−1 for Ni-
ZrO2/C and 1.8 h−1 for Ni-Al2O3/C vs. 1.0 h−1 for Ni/C), but they
led to no detectable formation of the C2 and C3 products. A dehy-
drated product, hydroxyl furan, was mainly detected on Ni-ZrO2/C
and Ni-Al2O3/C, apparently as a result of the sequential dehydra-
tion and hydrogenation reactions of the xylose intermediate on the
Ni surface.
To further understand the promoting effects of the basic oxides,
their contents on the Ni-Oxide/C catalysts were varied. Table 4
shows the activities and selectivities of the xylitol hydrogenol-
ysis at 473 K and 4.0 MPa H2 on the representative CeO2 and
CaO-modified 7.3 wt% Ni/C catalysts (noted as Ni-xCeO2/C and Ni-
xCaO/C, in which x stands for the molar ratio of CeO2 and CaO to
Ni). As the CeO2/Ni molar ratio increased from 0.1 to 2.0, the activ-
ities increased by about two folds from 9.2 h−1 to 18.3 h−1. In this
range of CeO2/Ni ratios, the Ni particles remained similar in sizes.
Fig. 6 shows the STEM images of the representative Ni-xCeO2/C
Table 4
Effects of CeO2 and CaO contents on activities and selectivities in xylitol hydrogenolysis on Ni-CeO2/C and Ni-CaO/C.a
Catalystb
Activity (h−1
)
Selectivity (on a carbon basis, %)
Ethylene glycol
Propylene glycol
Glycerol
Arabitol
Threitol
Deoxy-pentitolsc
Ni-0.1CeO2/C
Ni-0.22CeO2/C
Ni-0.45CeO2/C
Ni-0.66CeO2/C
Ni-1CeO2/C
Ni-1.5CeO2/C
Ni-2CeO2/C
Ni-0.33CaO/C
Ni-0.66CaO/C
Ni-1CaO/C
9.2
11.2
22.5
23.1
24.4
25.1
25.2
25.4
26.0
22.4
27.6
27.7
28.4
12.3
18.0
26.1
26.3
26.5
25.4
26.1
18.3
25.8
25.2
26.0
20.0
15.4
7.6
7.5
8.0
9.0
8.2
16.3
15.2
16.4
15.1
15.7
8.6
5.6
5.3
5.8
6.9
7.6
15.1
8.6
8.3
2.7
4.0
3.4
4.0
3.3
3.6
3.4
4.2
2.1
1.3
1.3
0.9
2.9
5.7
7.6
6.8
5.5
4.5
3.8
6.8
6.0
5.3
5.0
11.5
12.9
14.4
20.5
18.3
6.4
6.4
6.7
d
Ni-0.66CaO/C
–
a
Reaction conditions: 473 K, 4.0 MPa H2, 2 h, 7.3 wt% Ni loading, 0.1–0.4 g catalyst, 40 g 10 wt% xylitol aqueous solution, 20–30% xylitol conversion.
Numbers represents molar ratios of CeO2 or CaO to Ni at constant Ni contents.
Including 1,2,5-pentanetriol and 1,2,4,5-pentanetetrol.
At ∼100% conversion.
b
c
d
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
7
Fig. 6. STEM micrographs and histograms of Ni particle size distribution of 7.3 wt% Ni-xCeO2/C (x = 0.22, 0.45, 1) (scale bar = 20 nm). (a) Ni-0.22CeO2/C, 8.1 1.2 nm; (b)
Ni-0.45CeO2/C,9.0 1.5 nm; (c) Ni-1CeO2/C, 7.8 1.3 nm.
samples with CeO2/Ni ratios of 0.22, 0.45, and 1, which possessed Ni
particle sizes of 8.1 1.2 nm, 9.0 1.5 nm and 7.8 1.3 nm, respec-
tively. Meanwhile, elemental analysis of the isolated Ni particles
for these three samples by energy dispersive X-ray spectroscopy
(EDS) shows their Ce/Ni molar ratios increased linearly from 0.05
to 0.25. implying close contact between CeO2 and the Ni particles.
By referring to the Ni-xCaO/C catalysts to be discussed below, the
higher activities at the higher CeO2/Ni ratios is not solely related
to the increase in the basicity, but may be due to other reasons,
such as electronic effect of CeO2 on Ni, as proposed by Xu et al.
[21] and Yuan et al. [11]. The selectivities to the sum of the C2 and
C3 products increased from 54.8% to 58.1%, and the selectivities
to deoxypentitols (mainly pentane-1,2,5-triol and pentane-1,2,4,5-
tetrol) increased from 2.9% to 7.6%, concurrently with the decrease
in the selectivities to C5 and C4 products from 19.7% to 9.6%, as
the CeO2/Ni ratios increased from 0.1 to 0.45. In this ratio range,
specifically, the propylene glycol selectivities increased from 12.3%
to 26.1% with the concurrent decrease of the glycerol selectivi-
ties from 20% to 7.6%. By further increasing the CeO2/Ni ratios
from 0.45 to 2.0, the selectivities to the C2 and C3 products kept
nearly constant (∼59%); the selectivities to the C4 and C5 products
increased from 9.6% to 11.8% while the selectivities to deoxypen-
titols declined from 7.6% to 3.8%. Moreover, lactic acid was not
detectable, irrespective of the CeO2/Ni ratios.
As for the Ni-xCaO/C catalysts, their activities clustered around
a value of 6.5 h−1, independent of the CaO/Ni ratios in the range
0.33 to 1.0. However, similar to the effect of CeO2 on selectivity,
with increasing the Ca/Ni ratios from 0.33 to 1.0, the selectivities to
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040

GModel
CATTOD-8824; No. of Pages8
ARTICLE IN PRESS
8
J. Sun, H. Liu / Catalysis Today xxx (2014) xxx–xxx
the sum of C2 and C3 products increased from 57.0% to 69.3%, and
the selectivities to the C4 and C5 products decreased from 17.2%
to 9.6% with an slight decline in the selectivities to deoxypenti-
tols from 6.8% to 5.3%. This selectivity trend reflects the favorable
retro-aldol condensation at higher CaO contents and consequently
higher basicity. On these catalysts, lactic acid formed at negligi-
ble amounts (<1%). Noticeably, the high selectivity to the C2 and
C3 products can be obtained even at high xylitol conversions. For
example, on Ni-0.66CaO/C, the combined selectivity to ethylene
glycol, propylene glycol and glycerol reached 69.5% at nearly 100%
xylitol conversion with less than 1% selectivity to lactic acid. This
selectivity can be potentially further improved upon optimization
of the reaction parameters and the catalyst compositions, show-
ing the potential advantages of such modified Ni catalysts for the
selective hydrogenolysis of xylitol and other polyols.
National Basic Research Project of China (Grants 2011CB201400
and 2011CB808700).
Appendix A. Supplementary data
Supplementary material related to this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.cattod.2013.
12.040.
References
[1] (a) G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 4044–4098;
(b) A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.
[2] S. Van de Vyver, J. Geboers, P.A. Jacobs, B.F. Sels, ChemCatChem 3 (2011)
82–94.
[3] D.M. Alonso, S.G. Wettstein, J.A. Dumesic, Chem. Soc. Rev. 41 (2012)
8075–8098.
[4] (a) A. Wang, T. Zhang, Acc. Chem. Res. 46 (2013) 1377–1386;
(b) J.P. Ma, W.Q. Yu, M. Wang, X.Q. Jia, F. Lu, J. Xu, Chin. J. Chem. 34 (2013)
492–507.
[5] Y. Liu, C. Luo, H.C. Liu, Angew. Chem. Int. Ed. 51 (2012) 3249–3253.
[6] A.M. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem. Int. Ed. 51 (2012)
2564–2601.
[7] J. Ten Dam, U. Hanefeld, ChemSusChem 4 (2011) 1017–1034.
[8] R.V. Chaudihari, A. Torres, X. Jin, B. Subramaniam, Ind. Eng. Chem. Res. 52 (2013)
15226–15243.
[9] J.H. Zhou, M.G. Zhang, L. Zhao, P. Li, X.G. Zhou, W.K. Yuan, Catal. Today 147S
(2009) S225–S229.
[10] J.Y. Sun, H.C. Liu, Green Chem. 13 (2011) 135–142.
[11] L.M. Ye, X.P. Duan, H.Q. Lin, Y.Z. Yuan, Catal Today 183 (2012) 65–71.
[12] M. Banu, S. Sivasanker, T.M. Sankaranarayanan, P. Venuvanalingam, Catal. Com-
mun. 12 (2011) 673–677.
[13] M. Banu, P. Venuvanalingam, R. Shanmugam, B. Viswanathan, S. Sivasanker,
Top. Catal. 55 (2012) 11–13.
[14] T. Soták, T. Schmidt, M. Hronec, Appl. Catal. A: Gen. 459 (2013) 26–33.
[15] X.G. Chen, X.C. Wang, S.X. Yao, X.D. Mu, Catal. Commun. 39 (2013) 86–89.
[16] Z.W. Huang, J. Chen, Y.Q. Jia, H.L. Liu, C.G. Xia, H.C. Liu, Appl. Catal. B: Environ.
147 (2014) 377–386.
[17] S.P. Chopade, D.J. Miller, J.E. Jackson, T.A. Werpy, J.G. Frye, A.H. Zacher, US
6291725 (2001).
4. Conclusions
Ni/C catalysts in the presence of solid bases, e.g. Ca(OH)2 and
CeO2, physically mixed with or co-supported with Ni on C, can
efficiently catalyze the xylitol hydrogenolysis into ethylene glycol
and propylene glycol. The presence of the solid bases prevents the
agglomeration and leaching of Ni particles on C, and renders them
stable under the reaction conditions. Supporting the solid bases,
especially CeO2 and CaO, with Ni on C reduces the amount of solid
bases required and leads to negligible formation of lactic acid. The
observed efficiency of the Ni catalysts and the reduction in the
use of the solid bases are apparently correlated to the appropri-
ate hydrogenation activity of the Ni catalysts and consequently the
favorable C C cleavage of xylose intermediate with the bases, over
its competitive hydrogenation on Ni surfaces, to form glycolalde-
hyde and glyceraldehyde, the proposed intermediates for the two
target glycols.
[18] K.L. Deutsch, D.G. Lahr, B.H. Shanks, Green Chem. 14 (2012) 1635–1642.
[19] W.Z. Li, H. Huang, H.J. Li, W. Zhang, H.C. Liu, Langmuir 24 (2008)
8358–8366.
Acknowledgements
[20] S. Wang, Y.C. Zhang, H.C. Liu, Chem. Asian J. 5 (2010) 1100–1111.
[21] W. Yu, J. Zhao, H. Ma, H. Miao, Q. Song, J. Xu, Appl. Catal. A: Gen. 383 (2010)
73–78.
This work was supported by the National Natural Science Foun-
dation of China (Grants 21173008, 51121091 and 21373019) and
Please cite this article in press as: J. Sun, H. Liu, Selective hydrogenolysis of biomass-derived xylitol to ethylene glycol and propylene glycol on
Ni/C and basic oxide-promoted Ni/C catalysts, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2013.12.040