Catalytic conversion of cellulose to ethylene glycol over a low-cost binary catalyst of Raney Ni and tungstic acid

Article abstract of DOI:10.1002/cssc.201200842

Following our previous report on the selective transformation of cellulose to ethylene glycol (EG) over a binary catalyst composed of tungstic acid and Ru/C, we herein report a new lowcost but more effective binary catalyst by using Raney nickel in place of Ru/C (Raney Ni+H₂WO₄). In addition to tungstic acid, other W compounds were also investigated in combination with Raney Ni. The results showed that the EG yield depended on the W compound: H₄SiW₁₂O₄₀H₃PW ₁₂O₄₀WO₃H₂WO₄, but all the investigated W compounds were selective towards EG. Moreover, both WO ₃ and H₂WO₄ were dissolved partially under the reaction conditions and transformed into H_xWO₃ which is the genuinely active species for the C-C bond breakage of cellulose. This result further confirmed that the reaction that involves the selective breakage of the C-C bonds of cellulose with W species is homogenous. Among various binary catalysts, the combination of Raney Ni and H₂WO₄ gave the highest yield of EG (65 %), which could be attributed to the high activity of Raney Ni for hydrogenation and its inertness for the further degradation of EG. Moreover, Raney Ni+H₂WO₄ showed good reusability; it could be reused at least 17 times without any decay in the EG yield, which shows its great potential for industrial applications.

Full text of DOI:10.1002/cssc.201200842

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CHEMSUSCHEM
FULL PAPERS
DOI: 10.1002/cssc.201200842
Catalytic Conversion of Cellulose to Ethylene Glycol over
a Low-Cost Binary Catalyst of Raney Ni and Tungstic Acid
Zhijun Tai,[a, b] Junying Zhang,[a, b] Aiqin Wang,[a] Jifeng Pang,[a] Mingyuan Zheng,[a] and
Tao Zhang*[a]
Following our previous report on the selective transformation
of cellulose to ethylene glycol (EG) over a binary catalyst com-
posed of tungstic acid and Ru/C, we herein report a new low-
cost but more effective binary catalyst by using Raney nickel in
place of Ru/C (Raney Ni+H2WO4). In addition to tungstic acid,
other W compounds were also investigated in combination
with Raney Ni. The results showed that the EG yield depended
on the W compound: H4SiW12O40 <H3PW12O40 <WO3 <H2WO4,
but all the investigated W compounds were selective towards
EG. Moreover, both WO3 and H2WO4 were dissolved partially
under the reaction conditions and transformed into HxWO3,
which is the genuinely active species for the CÀC bond break-
age of cellulose. This result further confirmed that the reaction
that involves the selective breakage of the CÀC bonds of cellu-
lose with W species is homogenous. Among various binary cat-
alysts, the combination of Raney Ni and H2WO4 gave the high-
est yield of EG (65%), which could be attributed to the high
activity of Raney Ni for hydrogenation and its inertness for the
further degradation of EG. Moreover, Raney Ni+H2WO4
showed good reusability; it could be reused at least 17 times
without any decay in the EG yield, which shows its great po-
tential for industrial applications.
Introduction
The limited reserves of fossil resources and the ever-increasing
emissions associated with their consumption have stimulated
great interest in the search for alternative energy sources,
which include wind, nuclear, solar, and biomass energies.
Biomass, as the only renewable carbon source on earth, is con-
sidered a promising feedstock for the sustainable production
of liquid fuels and chemicals.[1] In regard to the valorization of
biomass, the conversion of cellulose has received a great deal
of attention thanks to its nonfood nature and huge availabili-
of glucose to sorbitol, which circumvents the problem of meta-
stable glucose and allows the reaction to proceed at a relatively
high temperature (above 453 K). To meet the requirements of
both hydrolysis and hydrogenation, various bifunctional cata-
lysts that bear both acid and metal sites on one catalyst have
been designed and evaluated, which include Pt/Al2O3,[2a]
Ni2P/C,[4] Ru/CNT (CNT=C nanotubes),[5] Ni/CNF (CNF=C nano-
fiber),[6] and Ir-Ni/MC (MC=mesoporous carbon).[7] In some
cases, binary catalysts composed of a dilute mineral acid and
a solid metal catalyst were also employed to make the two dif-
ferent functional sites more tunable.[2c] In each case, a good
balance between the two functions is very important to ach-
ieve a high yield of polyols (which include sorbitol and manni-
tol). A key point is that high-temperature water is also able to
afford H+ and, therefore, functions as a reversible acid to pro-
mote the hydrolysis of cellulose.[2b] In this case, that is, at reac-
tion temperatures above approximately 500 K, a high yield of
hexitols could be obtained with only a hydrogenation catalyst
such as Ru/C without the requirement of external acids.
ty.[2] Cellulose is
a biopolymer of d-glucose linked by
b-1,4-glycosidic bonds with a polymerization degree up to
10000.[3] The abundant intra- and intermolecular H bonds in
cellulose protect the b-1,4-glycosidic bonds from attack by for-
eign molecules, which includes water. Such an intrinsically re-
calcitrant structure presents a grand challenge for the efficient
transformation of cellulose under mild conditions.
The one-pot conversion of cellulose into polyols, which was
first reported by Fukuoka and Dhepe in 2006, opened a new
route to the valorization of cellulose.[2a] This process couples
the hydrolysis of cellulose to glucose and the hydrogenation
In 2008, our group developed a new route: the one-pot
transformation of cellulose to ethylene glycol (EG).[8] Compared
with hexitols, EG has a much larger market (ca. 20 MTyearÀ1
)
[a] Z. Tai, J. Zhang, Prof. A. Wang, Dr. J. Pang, Dr. M. Zheng, Prof. T. Zhang
State Key Laboratory of Catalysis
and is widely used as a monomer for the manufacture of poly-
ethylene teraphthalate (PET).[9] Currently, EG is predominantly
manufactured from ethylene by oxidation followed by hydra-
tion, and is, therefore, highly dependent on petroleum
resources. The direct transformation of cellulose to EG opens
a new avenue for the production of bulk chemicals from re-
newable biomass, which will greatly reduce the dependence
on petroleum resources. This new reaction critically relies on
the employment of W-based catalysts. Initially, we found that
Dalian Institute of Chemical Physics
Chinese Academy of Sciences
Zhongshan Road 457, Dalian 116023 (P.R. China)
Fax: (+86)411-84691570
E-mail: taozhang@dicp.ac.cn
[b] Z. Tai, J. Zhang
Graduate University of Chinese Academy of Sciences
Beijing 100049 (P.R. China)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201200842.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 652 – 658 652

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Ni-promoted tungsten carbide (Ni-W2C/AC; AC=active C)
acted as a highly effective catalyst to afford EG yields from 61–
75% depending on the nature of the carbon support.[10] A fur-
ther study revealed that the formation of W2C is not mandato-
ry; bimetal formulations composed of W and group VIII metals
(W–Pt, W–Ir, W–Ru, etc.) also gave rise to an EG yield of up to
75%.[11] These results pose a new question: what is the genu-
inely catalytically active species in this reaction? Very recently,
we used tungstic acid in combination with Ru/C to catalyze
this reaction and found a very interesting phase-transfer be-
havior for tungstic acid during the reaction.[12] Tungstic acid is
insoluble in water at room temperature (RT), but dissolves in
high-temperature water and transforms to tungsten bronze
(HxWO3) in the presence of H2. We propose that this dissolved
HxWO3 acts as the catalytically active species to promote the
CÀC bond breaking reaction of cellulose by homogeneous cat-
alysis to form glycolaldehyde, which is then hydrogenated
over Ru/C to give the target product EG.[12]
composed of Raney Ni and tungstic acid afforded EG yields of
up to 65%, an increase of 10% compared to binary Ru/C and
tungstic acid systems. Moreover, this binary catalyst also shows
excellent stability and reusability. Thus, our present study pro-
vides a low-cost but highly effective catalyst formulation,
which will have great potential for industrial applications.
Results and Discussion
Catalytic performance of the Raney Ni+H2WO4 binary
system
Firstly, we investigated a series of binary catalysts composed of
tungstic acid and Raney Ni in various relative amounts.
As shown in Table 1, the EG yield was strongly dependent on
the relative amounts of Raney Ni and tungstic acid; it first in-
creased with the increasing amount of H2WO4 from approxi-
mately 15–50 mg and then leveled off (ca. 63%) with a higher
amount of H2WO4 if the amount of Raney Ni was fixed at ap-
proximately 80 mg (entries 1–5). According to our previous
study on the phase-transfer behavior of H2WO4,[12] the catalyti-
cally active species in this reaction is the dissolved HxWO3,
which dictates the selective formation of EG from cellulose.
Thus, the concentration of the dissolved HxWO3 in hot water
increases initially with an increase of the amount of H2WO4
and then reaches its saturation when the amount of H2WO4
reaches 50 mg per 50 g of water. An increased amount of
H2WO4 beyond 50 mg did not result in a further increase of
the EG yield because of the dependence of its solubility on
temperature. However, on fixing the amount of H2WO4 at ap-
proximately 50 mg and changing the amount of Raney Ni from
approximately 10–300 mg, one can see a volcano shape in the
One of the remarkable advantages of the binary catalyst
system composed of tungstic acid and Ru/C is its excellent
reusability thanks to the phase-transfer behavior of tungstic
acid and superior stability of Ru/C. Nevertheless, the high cost
of Ru will limit its industrial application. Moreover, as Ru is able
to promote the methanation of EG, the EG yield over the
binary catalyst H2WO4 +Ru was always lower than that over
Ni-W2C/AC reported earlier.[8,10,11] In this work, with an aim to
search for low-cost but highly effective catalyst combinations,
we systematically studied a new binary catalyst composed of
Raney Ni and tungstic acid. Both components are low cost and
commercially available, and Raney Ni was recently reported to
catalyze the hydrogenolysis of glycerol to EG[13] and the up-
grade of bio-oil.[14] In this work, we found that binary catalysts
Table 1. Yields of the main polyols from the conversion of cellulose over different catalysts.[a]
[d]
Entry
Catalyst
Yield[b] [%]
GC[c]
[%]
LC[c]
[%]
Conversion
[wt%]
EG
1,2-PG
Gly
1,2-BDO
Ery
Man
Sor
1
2
3
4
5
6
7
8
Raney Ni (0.0877 g)+H2WO4 (0.0156 g)
Raney Ni (0.0848 g)+H2WO4 (0.0310 g)
Raney Ni (0.0845 g)+H2WO4 (0.0410 g)
Raney Ni (0.0815 g)+H2WO4 (0.0513 g)
Raney Ni (0.0857 g)+H2WO4 (0.1049 g)
Raney Ni (0.0104 g)+H2WO4 (0.0511 g)
Raney Ni (0.0232 g)+H2WO4 (0.0502 g)
Raney Ni (0.0461 g)+H2WO4 (0.0508 g)
Raney Ni (0.0658 g)+H2WO4 (0.0505 g)
Raney Ni (0.3031 g)+H2WO4 (0.0506 g)
Raney Ni (1.0496 g)+H2WO4 (0.0558 g)
3% Pt/AC (0.1515 g)+H2WO4 (0.0514 g)
4% Pd/AC (0.1511 g)+H2WO4 (0.0514 g)
4% Ir/AC (0.1506 g)+H2WO4 (0.0509 g)
1.2% Ru/AC (0.1503 g)+H2WO4 (0.0500 g)
Raney Ni (0.030 g)+H2WO4 (0.032 g)
21.5
55.0
61.0
65.4
62.8
37.6
45.5
55.1
65.0
58.4
14.7
39.6
30.1
34.9
54.4
39.0
2.9
2.8
3.4
3.3
2.6
2.9
4.4
4.2
3.4
3.0
3.1
2.7
1.1
3.0
6.0
9.5
6.2
0
0
0
0
0
0
0
0
1.3
2.8
2.0
0
1.3
1.6
7.8
2.5
4.3
5.5
4.7
1.3
9.1
1.4
7.5
4.3
3.3
3.9
10.7
7.1
6.6
4.7
n.d.
2.2
4.7
4.5
4.1
2.7
1.3
2.0
1.8
3.6
4.7
2.6
0.9
1.6
1.4
2.2
0
3.6
3.6
3.0
3.3
2.6
1.4
1.5
1.8
2.6
4.1
4.3
0.9
1.0
0.8
1.6
4.2
11.2
8.9
8.1
7.5
5.8
1.3
1.2
2.2
5.6
12.3
10.8
1.5
2.8
2.3
4.5
4.3
1.1
1.1
1.0
1.0
1.5
2.1
1.9
1.0
1.2
0.8
2.6
1.3
2.4
2.1
4.4
n.d.
61.0
87.5
94.1
94.9
92.3
86.5
70
100
100
100
100
100
100
100
100
n.d.
n.d.
99
n.d.[e]
88.5
92.6
89.4
52.7
79.8
68.4
72.7
81.3
n.d.
9
10
11
12
13
14
15
16[f]
88
87
100
98.7
[a] Typical reaction conditions: cellulose (0.5 g), a designated amount of catalyst, and water (50 g) were put in an 100 mL autoclave and stirred at 518 K,
6 MPa H2, and 1000 rpm for 30 min. [b] The yields of polyols were calculated by using the equation: yield [%]=(weight of polyol in the products)/(weight
of cellulose put into the reactor)ꢁ100%. EG, 1,2-PG, Gly, 1,2-BDO, Ery, Man, and Sor are abbreviations for ethylene glycol, 1,2-propylene glycol, glycerol,
1,2-butanediol, erythritol, mannitol, and sorbitol, respectively. [c] GC and LC indicate the total carbon in the gas and liquid products, respectively. [d] Cellu-
lose conversion [wt%] was calculated by the weight difference of cellulose before and after the reaction with an uncertainty of Æ3%. [e] Not determined.
[f] Reaction conditions: Miscanthus (0.25 g), a designated amount of catalyst, and water (25 g) were put into a 75 mL autoclave and stirred at 518 K, 5 MPa
H2, and 1000 rpm for 120 min.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 652 – 658 653

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
EG yield, which reached the highest value (ca. 65%) at 60–
80 mg of Raney Ni and then went down slightly to 58.4% at
approximately 300 mg of Raney Ni (entries 6–10). A further in-
crease of the amount of Raney Ni to approximately 1000 mg
led to a sharp decrease of the EG yield to 14.7% (entry 11),
possibly a result of the basicity associated with Raney Ni,
which negatively affects the conversion of cellulose. The basici-
ty of Raney Ni was proven by CO2 temperature-programmed
desorption (TPD, Figure S1 in the Supporting Information). The
low liquid carbon yield (LC in Table 1) also confirmed the poor
conversion of cellulose at too high loading of Raney Ni.
Evidently, a high yield of EG requires a good match between
the two catalyst components. A similar trend was observed in
the case of Ru/C+H2WO4,[12] which reflects the importance of
the balance of the two functions in such cascade reactions.
Compared to the binary catalyst Ru/C+H2WO4,[12] the binary
catalyst Raney Ni+H2WO4 possesses two remarkable advan-
tages: the low cost of the catalyst and the high yield of EG.
Based on the required amount of Ni or Ru to process 1 kg of
cellulose as well as the price of the two metals, it can be
roughly estimated that the catalyst cost of Raney Ni is only ap-
proximately 1/8 of that of Ru/C. In addition, Raney Ni is more
easily separated than Ru/C, for example, by magnetic separa-
tion. This is very important for the commercialization of this
process. Moreover, Raney Ni+H2WO4 always led to a higher EG
yield than Ru/C+H2WO4. As shown in Table 1, with optimized
formulations, the EG yield over the Raney Ni catalyst was
65.4 wt%, which is an increase of 11 wt% over that from Ru/C
(entry 15). We also found that the carbon balance was also
much better in the case of Raney Ni+H2WO4. As shown in
Table 1, the total organic carbon in the liquid phase was as
high as 95%, whereas the total carbon in the gas phase was
only 1%, which suggests high selectivity to liquid polyols over
Raney Ni. To further shed light on the superiority of Raney Ni
in the production of EG from cellulose, we investigated the
degradation of EG over the two catalysts. In this experiment,
a 0.5 wt% EG aqueous solution was used to simulate the EG
concentration in cellulose conversion. As shown in Figure 1,
more than 30% EG was decomposed to form CH4 over
1.2 wt% Ru/C, whereas less than 5% EG conversion was ob-
served over Raney Ni and other noble metal catalysts. Clearly,
the nearly inert nature of Raney Ni for the degradation of EG
should be a reason for the higher EG yield achieved with
Raney Ni+H2WO4. Not only superior to Ru/C, Raney Ni was
also superior to other precious metal catalysts such as Pt/C,
Pd/C, and Ir/C, which is due to its excellent hydrogenation ac-
tivity. As shown in Table 1, in combination with H2WO4, these
noble metal catalysts gave rise to EG yields no higher than
40%, which is more than 20% lower than Raney Ni+H2WO4.
The combination of Raney Ni and H2WO4 was also applied to
the conversion of raw biomass, Miscanthus, as we reported
previously.[2h] As shown in entry 16, the EG yield based on the
weight of Miscanthus was 39.0% under optimized conditions.
Considering that the cellulose content of Miscanthus is approx-
imately 50%, this EG yield is high. This result shows the poten-
tial of the binary catalyst Raney Ni+H2WO4 for the conversion
of raw lignocellulosic materials into the bulk chemical EG.
Figure 1. Degradation of EG over different metal catalysts. The reaction con-
ditions and the catalyst amounts are identical to those in Table 1, entries 9,
12–15. An aqueous solution of EG (50 mL, 0.5 wt%) was used in the
reaction.
Combination of Raney Ni with various W species
With the determination of the optimal relative amounts of
Raney Ni and H2WO4, we subsequently investigated the catalyt-
ic performances of Raney Ni in combination with various W
species, which included heteropolyacids (H4SiW12O40 and
H3PW12O40), WO3, and H2WO4. Figure 2 illustrates the polyol
yields as a function of the different W compounds. It can be
seen that EG is always the predominant product regardless of
which W compound was employed in combination with Raney
Ni. This result provides strong evidence that W species play
a key role in the selective formation of EG from cellulose, as
proposed previously.[12] It was also found that the EG yield in-
creased in the order: H4SiW12O40 <H3PW12O40 <WO3 <H2WO4.
In particular, H4SiW12O40 showed a much lower EG yield
Figure 2. Catalytic performances of different binary catalysts composed of W
compounds with Raney Ni. The reactions were performed with cellulose
(0.5 g), water (50 mL), Raney Ni (40–50 mg), and a W compound (50 mg) at
518 K and 6 MPa H2 for 0.5 h.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 652 – 658 654

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
(32.6%) but a higher sorbitol yield than the other three W
compounds. According to our previous work[12] and the work
of Liu et al.,[15] W species catalyze the cleavage of the CÀC
bonds of cellulose by a retro-aldol reaction. The lower EG yield
and low cellulose conversion achieved with H4SiW12O40 may
suggest that H4SiW12O40 is intrinsically less active for cellulose
CÀC bond cleavage. In contrast, H2WO4 was the most active W
compound in the catalytic transformation of cellulose.
In a previous work,[12] we reported the temperature-con-
trolled phase-transfer behavior of H2WO4, which has been
taken as an indicator of the homogenous catalysis of H2WO4 in
the reaction of cellulose. In this work, we extended our study
to other W compounds, which are either soluble or insoluble
in water. Table 2 shows photographs of various W compounds
Figure 3. (a) Photograph of H2WO4 after hydrothermal treatment at 518 K
and 6 MPa H2 for 30 min. The dark blue solid dispersed on the bottom indi-
cates that the dissolved HxWO3 precipitated upon cooling. (b) The color
change of the transparent blue solution with time over 3 h, which indicates
that the dissolved HxWO3 was gradually oxidized to H2WO4 and precipitated
from the solution.
restored to WO3. To confirm the phase changes of
WO3 during the reaction, we performed XRD exami-
nations on both fresh and used WO3 in comparison
with H2WO4 (Figure 4). The XRD examinations had to
Table 2. Properties of various W compounds before and after treatment.[a]
W Compound
Color of solution
W concentration[c] EG yield
be conducted on a freshly recovered sample to be
able to collect important information about the
phase change with time. For WO3, the freshly recov-
ered sample after the reaction (6 min) showed a dif-
ferent XRD pattern to WO3, with the appearance of
the characteristic XRD pattern of HxWO3 (Figure 4a).
This indicates that WO3 is unstable under the reaction
conditions and subject to phase change from WO3 to
HxWO3. However, HxWO3 was not stable in air and
quickly transformed back to WO3 after its recovery
from the reaction medium (after 12 min). This conclu-
sion is quite different from that drawn by Liu et al.,
who found that the surfaces of all W species (W, W2C,
and WO3) were dominated by WO3 after the reac-
tion.[15] Nevertheless, their characterizations, either by
using Raman or X-ray photoelectron spectroscopy
(XPS), were performed ex situ after the reaction with
dried samples. Consequently the phase changes ob-
served here were missed in their work. Very similar to
WO3, H2WO4 also transformed to HxWO3 after the
reaction (Figure 4b).
before treatment after treatment[b] in liquid [ppm]
[%]
WO3
54
193
711
738
736
8.8[d]
H2WO4
32.4[d]
53.8[e]
49.1[e]
32.6[e]
(NH4)6H2W12O40
H3PW12O40
H4SiW12O40
[a] Treatment conditions: 518 K, 6 MPa H2 for 30 min in 100 mL autoclave. [b] The
blue solutions were obtained by treating W compounds in water (50 mL) at 518 K,
6 MPa H2 for 30 min. [c] The W concentration was determined by using inductively
coupled plasma (ICP). [d] EG was produced by the reaction of cellulose (0.08 g) with
blue solution (30 mL). [e] EG was produced under typical reaction conditions: cellulose
(0.5 g), a designated amount of catalyst, and water (50 g) were put in a 100 mL auto-
clave and stirred at 518 K, 6 MPa H2, and 1000 rpm for 30 min.
By understanding the temperature- and atmos-
phere-dependent phase-transfer behavior of both
WO3 and H2WO4, we subsequently conducted the
cellulose conversion with the transparent blue solu-
tion in combination with fresh Raney Ni to provide
before and after treatment in hot water (2458C for 0.5 h under
H2). Quite different from heteropoly acids and ammonium
metatungstate (AMT), which are soluble in water, neither
H2WO4 nor WO3 dissolve in water at RT. However, upon treat-
ment with hot water, a clear blue solution was observed to-
gether with precipitates in the autoclave. The transparent blue
solution was allowed to settle under ambient conditions for
some time, and a small amount of fine yellow precipitate was
observed as the solution became colorless (Figures 3 and S2).
This phenomenon is an indicator of temperature- and atmos-
phere-dependent phase-transfer behavior (Scheme 1). In other
words, WO3 behaved similarly to H2WO4; it transformed into
HxWO3 in hot water under H2 and, upon cooling to RT, it was
direct evidence for the homogeneous catalysis of the W spe-
cies in this reaction. The reaction conditions and the results
are shown in Table 2. It can be seen that, except for WO3, all
Scheme 1. Illustration of temperature- and atmosphere-dependent phase-
transfer behavior of H2WO4 and WO3 in the conversion of cellulose into EG.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 652 – 658 655