Hydrothermally Stable Ruthenium–Zirconium–Tungsten Catalyst for Cellulose Hydrogenolysis to Polyols

Article abstract of DOI:10.1002/cctc.201701113

In this work, we describe a catalytic material based on a zirconium–tungsten oxide with ruthenium for the hydrogenolysis of microcrystalline cellulose under hydrothermal conditions. With these catalysts, polyols can be produced with high yields. High and stable polyol yields were also achieved in recycling tests. A catalyst with 4.5 wt % ruthenium in total achieved a carbon efficiency of almost 100 %. The prepared Zr-W oxide is mesoporous and largely stable under hydrothermal conditions (493 K and 65 bar hydrogen). Decomposition into the components ZrO₂ and WO₃ could be observed at temperatures of 1050 K in air.

Full text of DOI:10.1002/cctc.201701113

Accepted Article
Title: Hydrothermally stable ruthenium-zirconium-tungsten catalyst for
cellulose hydrogenolysis to polyols
Authors: Martin Lucas, Katarina Fabičovicová, and Peter Claus
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201701113
Link to VoR: http://dx.doi.org/10.1002/cctc.201701113
A Journal of
www.chemcatchem.org

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
Hydrothermally stable ruthenium-zirconium-tungsten catalyst for
cellulose hydrogenolysis to polyols
Martin Lucas*[a], Katarina Fabičovicová[a] and Peter Claus[a]
Abstract: In this work, we describe a catalytic material based on a
zirconium-tungsten oxide with ruthenium for the hydrogenolysis of
microcrystalline cellulose under hydrothermal conditions. With these
catalysts, polyols can be produced with high yields. High and stable
polyol yields were also achieved in recycling tests. A catalyst with 4.5
wt% ruthenium in total achieved a carbon efficiency of almost 100%.
The prepared Zr-W oxide is mesoporous and largely stable under
reaction e.g. tungsten compound). It should be mentioned at this
point that the resulting polyols could promote the solvolysis of the
cellulose in the presence of mineral or solid acids.[42,43] In addition
to the activity and the selectivity achieved, the long-term stability
under harsh hydrothermal conditions is the decisive criterion for
the evaluation of the catalysts. For zeolites and supported metal
catalysts several studies were published about hydrothermal
stability in hot liquid water.[37-39] The main reason for deactivation
of zeolite based catalyst in hot water is the dealumination and
hydrolysis of siloxane bonds.[37] The stability of the zeolites can be
increased by modification of the H-form (e.g., lanthanium or
cerium). Unfortunately, the acidity decreases.[40]
hydrothermal conditions (493
K and 65 bar hydrogen). A
decomposition into the components ZrO2 and WO3 could be observed
at temperatures of 1050 K in air.
Further materials with excellent hydrothermal stability are carbon
materials, zirconia, titandioxid and mixed oxides.[37] Furthermore,
sintering of metal particles is a problem that these systems have
in common
In a previous work, we found that because of the high reaction
temperature, leaching of tungsten can occur.[12,13] The ruthenium
– tungsten catalyst with activated carbon as support material[12]
showed a loss of activity after six reaction runs due to the release
of tungsten.
This is consistent with results from the literature.[7,14,15] Thereafter,
tungsten bronzes are responsible for C-C cleavage. These
HxWO3 compounds act homogeneously catalytically and are
formed from the tungsten components of the catalyst. To
compensate the resulting loss of activity, authors give Tungstic
acid to the reaction mixture.[14,15]
A review, published by Tao Zhang, provides a comprehensive
overview of transformation of cellulose to polyols with a short
section on catalyst recycling.[33] A further study describes, among
other things, problems with the catalyst stability in the conversion
of biomass.[36]
In search of tungsten compounds that are stable under
hydrothermal conditions, references to zirconium tungsten oxides
were found. In literature, under hydrothermal conditions produced
tungsten and zirconia containing material can be found.[26-28] The
hydrothermal preparation after Hui[26] produced well crystalline
ZrW2O8 [16,17] grains for the photocatalytic water splitting with low
BET surface of 3.58 m2/g. In this work, preparation conditions
were also described leading to amorphous material. The surface
acidic properties of tungstated zirconia catalysts and ZrW2O8 are
used in the paraffin isomerization, cracking, alkylation, hydration
of cyclohexene and in the esterification of fatty acids.[18-22] The use
Introduction
Using of biomass for the production of chemicals is one of the
solutions to replace fossil raw materials. It solves part of the
problem of global warming. As it is the most frequent occurring
biological raw material, cellulose has been the focus of research
the past years.
The depolymerization of cellulose to soluble oligosaccharides up
to glucose is usually the first step in order for the utilization of
cellulose as a raw material for chemicals. Herein the application
of mineral acids shows good results.[1] These have the
disadvantage that mineral acids are contained in the product
mixtures in a homogeneous phase and complicated further
processes steps. The use of solid acid catalysts has therefore
been proposed as alternative approach.[2]
Many different solid acids like zeolites, sulfated zirconia, silica
(SBA15) and others were tested.[3-5,33,34,35]
All these studies have in common that the de-polymerization of
cellulose in water starts only at higher temperatures. Under these
conditions the final de-polymerization product – glucose – tends
to isomerization and dehydration Fructose by isomerization of
glucose formed, leads via retro aldol reaction to 1,2-propandiol as
main product.[30,31] The dehydration products 5-HMF and other
furans leads to unwanted polymerization products.[7,8] A solution
is the fast hydrogenation of the glucose molecule to sorbitol at
nickel or ruthenium catalysts[3] or the production of short chain
polyols like ethylene glycol via retro aldol reaction and hydrolysis.
The retro aldol reaction is catalysed by tungsten species.[9-11] An
effective catalyst to produce ethylene glycol or 1,2-propandiol
from cellulose contains a hydrogenation function (e.g. nickel or
ruthenium) and a function for the C-C cleavage (retro aldol
[23]
of tungstated zirconia in the aqueous hydrolysis of cellobiose
and cellulose [24] has been reported. Good yields of lactic acid and
2,5-hexanedione were observed with tungstated zirconia.[25]
.
[a]
Martin Lucas, Dr. Katarina Fabičovicová and Prof.Dr. P. Claus
Technische Chemie II, TU-Darmstadt
Ernst-Berl-Institut für Technische und Makromolekulare Chemie
Alarich-Weiss Straße 8
Good yields of ethylene glycol from cellulose were obtained with
a WO3-ZrO2 catalyst in combination with Ru/C.[32]
E-mail: lucas@tc2.tu-darmstadt.de
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
Results and Discussion
poor reflexes (Fig. 1d) and in case of a higher tungsten content,
reflexes are clearly visible (Fig. 1e). The pattern is in this case
very similar to the polymorph of zirconium tungstate hydrate
described in literature.[29] But also here: No x-ray reflexes were
found that could be attributed to ZrW2O8, WO3 or ZrO2. For the
catalysts with the different ruthenium contents, the catalytic
activity (Table 4) was determined in tests under standard
conditions. These standard conditions have proved to be the
optimum in experiments with ruthenium and tungsten-containing
catalysts and facilitate comparability with the results of our
previous investigations (Table 4, last line).[13,14] The ruthenium
content has a significant influence on conversion and product
spectrum. As ruthenium content increases from 2.0 wt% to
4.5 wt%, cellulose conversion drops from 90% to 72%.
With the chosen preparation method – using only small amounts
of hydro chloric acid – a support material (Zr-W-oxide) was
obtained that is similar to a polymorph of zirconium tungstate
hydrate described in literature.[29]
Table 1. Results of N2-physisorption measurement.
catalyst[a]
BET
surface
[m2/g]
Pore
volume[a]
[cm3/g]
Micropore
area[b]
[m2/g]
Pore
diam.[c]
[Å]
Zr-W-oxide
75.5
80.4
79.1
78.0
0.26
0.25
0.25
0.24
7.0
6.7
6.5
5.2
113
131
132
129
2Ru/Zr-W-oxide
3Ru/Zr-W-oxide
4.5Ru/Zr-W-oxide
3Ru/Zr-W-oxide
after Run#10
3Ru/Zr-W-oxide hw[d]
92.8
88.1
0.27
0.26
5.1
9.5
130
130
Figure 1. XRD Patterns for Zr-W-oxide (a), 3Ru/Zr-W-oxide (b) and 3Ru/Zr-W-
oxide after 10 recycling runs (c) with a measurement artefact around 50 degrees,
3Ru/Zr-W-oxide (-25W) (d), 3Ru/Zr-W-oxide (+25W) (e).
3Ru/ZrW2O8 AE[e]
0.3
0.0005
0.06
0
(44)
15
3Ru/Zr-W-oxide(-25W)
3Ru/Zr-W-oxide(+25W)
37.8
49.9
1.4
4.2
0.09
17
No x-ray reflexes were found that could be attributed to ZrW2O8,
WO3 or ZrO2 (Fig. 1a). The mesoporous material in this work, the
so-called Zr-W-oxide has a comparatively high BET-surface area
of 75 m2/g, with a pore volume of 0.26 g/cm3 and a small fraction
of micro pores (Table 1).
[a] Total pore volume < 270 Å. [b] t-Plot Method. [c] Medium pore diameter
(BJH method). [d] catalyst pre-treated four hours without cellulose in hot
water under standard reaction conditions. [e] catalyst was prepared based
on a commercial support material (ZrW2O8; AlfaAesar).
The ratio between tungsten and zirconium is according to REM-
EDX measurement 2:1 (Table 2).
This is strongly correlated with the measured acidity of the
catalysts (Table 3).
In the temperature programmed reduction experiments (Fig. 2) it
was shown that, the material has a well-defined reduction peak
with a maximum at 1050 K. The addition of ruthenium does not
affect the XRD results (Fig. 1b). The BET surface area and the
average pore radius are slightly increased (Table 1). This may be
related to the reduction in the hydrogen stream at 523 K. The
reduction peak of the tungsten is shifted slightly towards low
temperatures of 1034 K. In addition, a reduction peak for
ruthenium at 391 K can be observed.
In order to determine the optimum of zirconium to tungsten ratio,
the amount of Na2WO4 used was varied during preparation. Two
different materials were obtained.
Zr-W-oxide (-25W) with 25 % less tungsten addition and Zr-W-
oxide (+25W) with 25 % more tungsten addition during the
preparation. For these materials, a similar pattern can be seen in
XRD results. In case of a lower tungsten content, XRD shows very
It can also be formulated that the acidity of the support diminishes
with increasing ruthenium surface coverage. Ruthenium
dispersion is nearly constant for catalysts with different ruthenium
content. A dispersion value of 2.9% is low, but also explains the
low formation of sorbitol and the higher formation of ketones with
the catalysts. A higher ruthenium dispersion, but also a lower
acidity and thus a lower conversion, can be observed in case of
the catalysts with a varied Zr / W content. The catalysts with the
nominal W/Zr ratio > 2 show a higher EG yield (35% compared to
30%). This can possible be explained by soluble tungsten species.
Overall, the differences in product yields and conversion are not
very large due to the variation of the W / Zr ratio. A comparison of
long-term stability should provide further insights into W/Zr
system.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
Information about the catalyst stability can be obtained by
recycling experiments. The procedure for the recycling
experiments are described in the experimental section. The
recycling experiments started with a half gram of the catalyst
3Ru/Zr-W-oxide. After seven four-hour cycles, a cycle of eight
hours duration followed. Thereby the cellulose was fully converted
and the catalyst separated. Thus, 75% of the initial catalyst mass
could be recovered. On one hand, the loss of 25% of the catalyst
can be explained by the elimination of catalyst-components under
the hydrothermal conditions; on the other hand, a loss of catalyst
due to the recycling itself is possible. As shown in Table 2, the
tungsten zirconium atomic ratio (at the surface; REM-EDX)
changes from 1.9 to 1.4. However, in the centrifuged and
decanted reaction solutions, small amounts of catalyst together
with cellulose precipitated over time.
Table 2. Results of REM-EDX and XRF measurements.
Ru
[at.%]
Zr
[at.%]
W
W/Zra[a]
[at.%] REM-EDX
W/Zr[a]
XRF
catalyst[a]
Zr-W-oxide[b]
3Ru/Zr-W-oxide
-
33.3
66.7
2.01
1.76
7.4
31.9
60.8
1.91
1.76
3Ru/Zr-W-oxide
after Run#10
5.4
38.6
n.m.
53.8
n.m.
1.40
n.m.
1.47
1.63
3Ru/Zr-W-oxide hw[c]
n.m.
[a] Atomic ratio. [b] The W/Zr atomic ratio derived from the amounts used
in the preparation is 1.66. [c] Catalyst pre-treated four hours without
cellulose in hot water under standard reaction conditions.
The decrease in the cellulose conversion is accompanied by a
significant increase in the yield of polyols from 37 to 45%. In this
context, the decrease in ketone yields, which are the precursors
of the polyols, is also notified. The yield of ketones is very high
(>5%) compared to Ru/W/AC catalyst systems (0.4%) – hinting to
the fact that the hydrogenation activity of the catalysts is lower. It
was to be expected that the hydrogenation rate with the ruthenium
content increases. The jump in carbon efficiency coefficients
(CEL) from 92 to near 100% was unusual. The reason might be
that only small amounts of cellulose were converted to 5-HMF,
which is prone to polymerise and build humins under the reaction
conditions.[8] We obtained a yield of free HMF of approximately
< 0.07% in all experiments with Ru/Zr-W-oxide catalysts, only
about a quarter of what we found with Ru/W/AC catalysts in
previous experiments.[12] It is also noticeable that only small
amounts of sorbitol are formed - a further indication that the
hydrogenation activity of the catalysts is comparatively low. In
comparison, the catalyst with the support ZrW2O8 (AlfaAesar)
shows very low yields to polyols.
Figure 2. TPR profiles of Zr-W-oxide (___) and 3Ru/Zr-W-oxide (- - -)
What can be said about the catalyst activity during recycling? As
shown in Figure 3 and Table 5, the CEL was nearly constant at
90% after two runs. The increase during the first experiments
results from the transfer and accumulation of unreacted cellulose
by the recycling process. The overall selectivity to polyols was
nearly constant at 65%. The EG selectivity decreased from 44 to
28%, while selectivity to 1,2-propandiole, 1,2-butandiole and
sorbitol increased.
In the subsequent eighth cycle, fresh and accumulated cellulose
was completely converted after a longer time of reaction. This
results in an apparent CEL of 108%. After cycle eight the catalyst
recovered after being washed and dried and was used in the
reaction again. For this, the amount of cellulose and water were
adjusted to the lower catalyst mass (0.2688g). The results can
also be found in Figure 3 and Table 5 under Run#9 and #10. The
carbon efficiency achieved is comparable to the previous
experiments. Concerning ethylene glycol, an increase in the
selectivity from 27% to 34% was achieved. This value is close to
the selectivity in the first cycles. Part of the activity changes during
recycling and can be attributed to catalyst losses.
Table 3. Results of CO-chemisorption and NH3-TPD measurement in
[b]
comparision to cellulose conversion (X) and ethylene glycol yield (YEG
) .
[a]
catalyst
DRu
[%]
NH3-TPD
[mmol/g]
XCell
[%]
YPolyols
[%]
Zr-W-oxide
-
0.141
0.106
0.100
0.077
0.085
-
-
2Ru/Zr-W-oxide
3Ru/Zr-W-oxide
4.5Ru/Zr-W-oxide
2.9
2.9
2.8
2.8
90.5
81.9
72.2
-
37.2
43.5
45.1
-
3Ru/Zr-W-oxide
after Run#10
3Ru/Zr-W-oxide (-25W)
5.6
4.5
0.033
0.037
65.1
70.8
43.3
48.6
3Ru/Zr-W-oxide (+25W)
[a] Ruthenium dispersion. [b] extracted from table 3.
The resulting product solution was brown and the product stuck
to all parts of the batch. In comparison the yield of free 5-HMF
was very high at 0.6%.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
Table 4. Cellulose conversion (X), carbon efficiency coefficients (CEL) and product yields of the cellulose hydrogenolysis[a]
.
Yield [%][b]
catalyst
X [%]
CEL [%]
pH
EG
PG
BD
Sor
Σ Polyols
Ketones
Gaseous Others
40.6
26.6
22.4
27.7
19.4
19.0
20.4
5.1
2.80
2.88
2.87
2.94
2.92
2.88
2.12
-
2Ru/Zr-W-oxide
90.5
81.9
72.2
100
92.4
91.8
100
25.5
30.0
31.9
29.9
30.0
34.9
0.9
4.8
4.2
4.5
11.2
3.7
4.0
0.1
4.2
4.0
4.4
4.5
5.7
3.8
4.5
0.2
3.5
0.4
0.9
0.8
4.6
1.6
1.3
0.8
18.5
37.2
43.5
45.1
58.8
43.3
48.6
5.1
9.8
6.8
5.7
0.8
5.0
4.7
5.0
0.4
1.0
0.8
1.3
1.6
0.9
1.0
3.0
1.5
3Ru/Zr-W-oxide
4.5Ru/Zr-W-oxide
3Ru/Zr-W-oxide 8h[c]
3Ru/Zr-W-oxide(-25%W)
3Ru/Zr-W-oxide(+25%W)
3Ru/ZrW2O8 (AlfaAesar)[d]
2Ru/W/AC[e]
90.8
100
65.1
70.8
84.9
90.0
100
34.2
81.8
34.2
66.6
[a] under standard conditions as described. [b] EG = ethylene glycol, PG = propylene glycol, BD = 1,2-butanediol, Sor = Sorbitol, Σ Polyols = total C2…C6
diols and sugar alcohols, Ketones = 1-hydroxy-2-propanone and 1-hydroxy-2-butanone, Gaseous = CH4, CO and CO2, Others = summary of all other
peaks in GC (ketones, acids, furanes, esters, alcohols – identified by GC/MS), pH = pH of the solution after the reaction. [c] after 8 h reaction time –
standard conditions 3 h. [d] catalyst was prepared based on a commercial support material (ZrW2O8; AlfaAesar). [e] Preparation published in a previous
work. [12]
Table 5. Carbon efficiency coefficients (CEL) and product selectivity in the recycling experiments [a]
.
Selectivity [%][b]
Run#t
CEL [%]
pH
EG
PG
BD
7.5
Sor
HDL
Σ Polyols
Ketones
Gaseous Others
1
66.6
79.0
89.4
90.3
89.6
88.0
88.4
108.3
74.0
93.8
44.4
37.6
37.1
33.8
31.2
29.8
28.7
27.1
34.3
34.7
6.1
0.9
2.5
4.0
6.2
7.1
6.3
6.2
6.4
5.3
4.8
2.3
3.6
3.8
4.0
4.7
4.7
4.9
5.8
3.1
3.3
62.9
65.7
67.3
65.1
66.8
64.5
63.7
65.3
66.3
65.8
5.7
3.0
2.3
1.6
1.4
2.5
3.0
0.3
1.6
0.2
1.7
1.6
1.7
2.0
2.0
1.7
1.6
1.9
2.2
1.8
20.1
22.5
21.2
23.3
21.9
24.0
24.6
24.9
22.7
22.9
2.97
3.02
3.14
3.15
3.14
3.19
3.20
3.06
3.05
2.94
2
11.3
11.7
11.2
13.1
12.7
13.0
14.9
12.7
13.0
8.0
7.3
6.5
7.1
6.4
6.3
7.2
6.6
6.6
3
4
5
6
7
8[c]
9[d]
10[e]
[a] under standard conditions as described. [b] EG = ethylene glycol, PG = propylene glycol, BD = 1,2-butanediol, Sor = Sorbitol, Σ Polyols = total
C2…C6 diols and sugar alcohols, Ketones = 1-hydroxy-2-propanone and 1-hydroxy-2-butanone, Gaseous = CH4, CO and CO2, Others = summary
of all other peaks in GC (ketones, acids, furanes, esters, alcohols – identified by GC/MS), pH = pH of the solution after the reaction. [c] 8 h reaction
time for full conversion of cellulose. [d] With reduced catalyst mass (0.2688g) from recycling#8 and accordingly reduced cellulose and water amount
after 4 h reaction time. [e] Same as [d] and 4 h reaction time.
The significant change in the product distribution from the first to
the second cycle - fewer retro aldol reaction products (ethylene
glycol), more isomerization (1,2-propandiole as a product of the
retro aldol reaction of fructose) and hydrogenation products (e.g.
sorbitol)- can’t be explained simply by catalyst lost. The decrease
of retro aldol reaction products (e.g. ethylene glycol) and the
increase in selectivity of sorbitol can be explained by tungsten
losses.[12]
catalyst after recycling tests was almost the same (2.9%). The
acidity decreased significantly (15%) during the recycling tests,
which seems to be associated with the loss of tungsten (Table 3
and 2). Compared to data from the literature, however, this loss is
not very high. Lercher et al. published for different HBEA zeolites
the loss of acid sites in hot liquid water.[41] For example, the
number of acidic sites of HBEA150 decreased from 0.150 mmol/g
to 0.115 mmol/g after 48 hours in hot liquid water at 433 K, a loss
of almost 25%. Compared to this, the loss in the investigated Zr-
W systems at 493 K (50h) in a real reaction system is low. This
The catalyst recovered after Run # 10 was examined for changes.
Interestingly, sintering of the ruthenium particles did not play an
important role. The ruthenium dispersion of the fresh and of the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
also applies to the pore system. HBEA150 lost about 50% of the
micropores in hot water, while the catalyst 3Ru/Zr-W-oxide loses
little pore volume.
preparation were washed out. According to experience made in
the recycling test, the solution was left to sediment. This reduced
the catalyst losses from 4 wt% to 2 wt%. Via REM-EDX
measurements a change in the W/Zr atomic ratio from 1.9 to 1.4
could be observed as described above. This statement applies,
due to the REM-EDS method only for the surface of the sample.
The atomic ratio in the bulk (Table 2; XRF) does not decrease so
drastically from 1.76 to 1.47. The described four-hour hot water
treatment changes the W/Zr atomic ratio from 1.76 to 1.63. This
result is consistent with the measured mass loss and the change
in the pore system of the catalyst.
Fig. 4 shows the XPS of the 3Ru/Zr-W-oxide in fresh state and
after the recycling tests. The changes in the tungsten signals are
not very large. The ratio of the summarized signal intensity
between WO3 and WO2 changes from 44 to 56 (fresh catalyst)
toward 57 to 43 (after recycling). As noted in the REM-EDX and
XRF investigations, the atomic ratio W/Zr changes significantly.
From 1.4 for the fresh catalyst to 1 for the recycled catalyst.
The XRD measurements showed no significant changes in the
location and the width of the reflexes (Figure 1). A decomposition
into WO3 and ZrO2 was observed only after a treatment in air at
1050 K for 12 h.
Figure 3. Results of the recycling experiments under standard conditions as
described. EG = ethylene glycol, PG = propylene glycol, BD = 1,2-butanediol,
Total polyols = total 1,2-C2…C6 diols, Other GC = summary of all other peaks
in gaschromatography (ketones, acids, furanes, esters, alcohols – identified by
GC/MS. a) after 8 h reaction time. b) Recycling test with reduced cellulose
mass – adapted to the catalyst mass found after experiment #8. Results of the
recycling experiments under standard conditions as described.
Conclusions
A ruthenium catalyst based on zirconium and tungsten, which is
of interest for the reaction under hydrothermal conditions, has
been successfully developed and tested. In addition to good
yields of polyols, the carbon efficiency coefficients reached
nearly 100% for the catalyst with 4.5 wt% Ruthenium - under
these harsh conditions a particular highlight.
The N2-physisorption measurement shows a significant larger
surface area for the recycled catalyst (93 m2/g instead of 79 m2/g).
A comparison test, in which fresh catalyst was treated under
standard conditions once without cellulose for 4 hours, showed
also an increase in the surface area in the same order of
magnitude (Table 1). This is an indication that residues of the
This is probably achieved by a slight tendency to form 5-HMF
and its subsequent polymerization to form humines. A decisive
advantage of the investigated system is its stability under the
chosen conditions in any case - 50 hours at 493 K in hot
compressed water under hydrogen pressure with a constant
selectivity of 65 % to polyols. The hydrogenation efficiencies are
weaker compared to other similar catalysts based on ruthenium.
This results in lower ethylene glycol yields and higher yields of
1,2-propandiol. This disadvantage could be compensated by
optimization of the catalyst preparation. In particular, the
calcination temperature, which has a decisive influence on the
activity and selectivity of Zr/W systems, should be optimized in
further studies.[32]
Experimental Section
Catalyst preparation
The carrier for the catalysts was prepared in accordance with.[26] For this
purpose, 100 ml of ZrOCl2 aqueous solution (0.15 mol/L; AlfaAesar) and
Figure 4. Experimental and simulated XPS of 3Ru/Zr-W-oxide (a) fresh and
after recycling run#10 (b)
200 ml of
a Na2WO4 aqueous solution (0.125 mol/L; AlfaAesar;
0.09375 mol/L in case of Zr-W-oxide(-25W); 0.156 mol/L in case of Zr-W-
oxide(+25W)) were added dropwise to 50 ml of deionized water in a beaker
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
at 333 K with constant stirring. After half an hour, 50 ml of 1.5 mol/L HCl
aqueous solution was added dropwise to the mixture. The slurry was
stirred for four hours at 333 K. The partially evaporated white slurry (100
mL) was transferred in a Teflon-lined Parr autoclave (300 mL), purged
three times with Argon and heated under 15 bar Argon for 20 h at 473 K
to carry out the hydrothermal conversion. The now blue suspension was
removed and aged for two days at room temperature. The colour changed
to light blue. After the aging, the product was filtered and washed with
deionized water until it was free of chloride ions and dried for 10 h at 376
K. Finally, the light green powder was crushed and sieved to particle less
than 200 µm and heated under airflow at 873 K for 6 hours in a quartz tube.
6.3 g light yellow powder were obtained.
cellulose conversion and then slows. After three hours (four hours in case
of the recycling experiments) the autoclave was cooled down to room
temperature and the gas phase was analysed via gas-phase FTIR
analyser (Thermo Antaris IGS). The aqueous phase together with the rest
of cellulose was filtered and the products analysed via GC and HPLC as
described in [13]. Only the twenty main products were individually quantified.
For the other products a mean quantification factor based on 1,2-
butandiole was used. The conversion rate (X) of the cellulose was
determined by weighing the dry filter cake.
In this work the yield of products (Yi) are calculated as the ratio of moles
of carbon in the different products and the moles of carbon in cellulose.
The carbon efficiency coefficient (CEL) is the ratio between the sum of
carbon found in the analysed products and the carbon in the converted
cellulose.
To each of three fractions (2g) of the powder different concentration of
ruthenium (III) nitrosyl nitrate solution (AlfaAesar) was added via incipient
wetness method. Catalysts containing 2 wt%, 3 wt% and 4.5 wt%
ruthenium were obtained. After drying at 376 K for 2 hours, the catalyst
was heated to 623 K under hydrogen flow for 2 h.
The recycling experiments started with the above-described procedure.
Since at the end of the reaction cellulose was not completely reacted, it
was separated from the product solution together with the catalyst. For that
purpose the slurry was centrifuged, the product solution was decanted and
weighed for the mass balance. The product yields are based on the mass
of the removed solution. The wet unconverted cellulose and the catalyst
were re-transferred in the batch together with fresh cellulose and water. It
is clear that the amount of cellulose with the uncomplete conversion in the
following recycle experiments rises. This procedure prevent re-oxidation of
the catalyst. The selectivity shown in the recycling experiments is the ratio
between the individual product yield and the sum over all analysed product
yields.
Materials with the variation in the Zr/W ratio were loaded in the same
manner with 3 wt% ruthenium.
Catalyst characterization
N2-Physisorption
measurements
were
performed
by
using
Quantachromes Quadrasorb MP. The Powder XRD measurements were
recorded using StoeCie (Ge[111]-Monochromator, CuKa1-Radiation,
l = 1,54060 Å, Detektor: Mythen1K), REM-EDX measurements (Jeol JSM
6400 with EDAX Apollo) and X-ray flourescence measurements with
Olympus GoldXpert XRF analyser. Temperature-programmed reduction
measurements was carried out on a TPD/R/O 1100 from Thermo Fisher
Scientific. Ruthenium dispersion was determined with CO-chemisorption -
also in the TPD/R/O 1100 equipment. Typically, 100 mg catalyst were pre-
treated in hydrogen at 623 K for one hour and after cooling down to room
temperature CO was pulsed in the hydrogen flow over the catalyst. The
CO concentration was monitored with a TCD. The ruthenium dispersion
DRu is the ratio of ruthenium amount and the amount of chemisorbed CO
in percent. NH3-TPD experiments was carried out in a self-assembled
quartz reactor. The sample (approximately 100 mg) was calcined at 623 K
or at 873 K under a flow of 10 cm3 /min of nitrogen. After cooling down to
373 K the sample was saturated under flow with 2% of ammonia in
nitrogen, subsequently 2 hours flushed with pure nitrogen and heated up
to 873 K with 10 K/min under 50 ml/min nitrogen flow. The desorption of
ammonia was monitored with a FTIR-detector (Thermo Antaris IGS with
2 m gas cell). XPS were taken on a SSX 100 ESCA Spectrometer with
monochromated Al Kα radiation source (aperture slot 0.25 * 1.0mm). The
high resolution spectra were collected with 50 eV and 0.054 eV resolution.
Acknowledgements
The help of Dr. Kathrin Hofmann with the XRD measurements,
Karl Kopp with XPS, Nalan Kalyon and Anne-Marie Zieschang
with the REM-EDX measurements is greatly acknowledged.
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Keywords: Cellulose • Hydrogenolysis • Polyols • Tungsten-
zirconia catalyst • Hydrothermally stable
[1]
[2]
J. A. Geboers, S. Van de Vyver, R. Ooms, B. Op de Beeck, P. A.
Jacobs, B. F. Sels, Catal. Sci. Technol. 2011, 1, 714-726.
X. Liu, X. Wang, S. Yao, Y. Jiang, J. Guan and X. Mu, RSC Adv. 2014,
4, 49501.
Catalytic experiments
[3]
[4]
[5]
A. Fukuoka, P. L. Dhepe, Angew. Chem., Int. Ed. 2016, 45, 5161-5163.
A. Onda, T. Ochi, K. Yanagisawa, Green Chem. 2008, 10, 1033-1037.
S. Suganuma, K. Nakajima, M. Kitano, D. Yamaguchi, H. Kato, S.
Hayashi, M. Hara, J. Am. Chem. Soc. 2008, 130, 12787-12793.
J. Chheda, Y. Roman-Leshkov, J. Dumesic, Green Chem. 2007, 9,
342–350.
A. Wang, T. Zhang, Acc. Chem. Res. 2013, 46, 1377-1386.
P. Sushil, C. Lund, Energy Fuels 2011, 25, 4745-4755.
N. Ji, T. Zhang, M. Zheng, A. Wang, H. Wang, X. Wang J. G. Chen,
Angew. Chem. 2008, 120, 8638-8641.
Typically, 0.5 g of the catalyst, 5 g microcrystalline cellulose (Merck; item
1.02331.500) and 100 g deionized water are filled in a 300 ml stainless
steel autoclave (Parr instruments), purged with argon, 25 bar hydrogen
was added to the solution and the mixture was heated under stirring (1000
rpm) up to 493 K. As the solution reached the desired temperature, the
pressure was increased from 42 bar to 65 bar with hydrogen. Notice that,
the measured gas phase temperature (approx. 425 K) is lower than the
liquid phase temperature. As a result, the partial pressure of the hydrogen
is roughly 59 bar. The pressure in the reactor is kept constant at 65 bar via
a backpressure regulator. The spent hydrogen was fed from an external
tank. The pressure drop in this tank was recorded during the reaction.
Therefore, it is possible to determine the amount of hydrogen consumed
in the reaction. The rate of hydrogen consumption is constant to about 50%
[6]
[7]
[8]
[9]
[10] M.-Y. Zheng, A.-Q. Wang, N. Ji, J.-F. Pang, X.-D. Wang, T. Zhang,
ChemSusChem 2010, 3, 63-66.
[11] Y. Liu, C. Luo and H. Liu, Angew. Chem. 2012, 124, 3303-3307.
[12] K. Fabicovicova, M. Lucas, P. Claus, Green Chem. 2015, 17, 3075-
3083.
[13] K. Fabicovicova, O. Malter, M. Lucas and P. Claus, Green Chem. 2014,
16, 3580–3588.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
[14] Z. Tai, J. Zhang, A. Wang, M. Zheng, T. Zhang, Chem. Commun. 2012,
48, 7052-7054.
[29] J. Colin, D. Camper, S. Gates, M. Simon, K. Witker, C. Lind, J. Solid
State Chem. 2007, 180, 3504-3509.
[15] Z. Tai, J. Zhang, A. Wang, J. Pang, M. Zheng, T. Zhang,
ChemSusChem 2013, 6, 652-658.
[16] J. Graham, A.D. Wadsley, j.H. Weymouth, L.S. Williams, J. Am. Cer.
Soc. 1959, 42, 570.
[17] Mary, T. A.; J. S. O. Evans; T. Vogt; A. W. Sleight, Science 1996, 272,
90-92.
[30] T. Wang, M. Nolte, B. Shanks, Green Chem. 2014, 16, 548.
[31] B.Beeck, M. Dusselier, J. Geboers, J. Holsbeek, E. Morre, S. Oswald,
L. Giebeler, B. Sels, Energy Environ. Sci. 2015, 8, 230-240.
[32] J. Chai, S. Zhu, Y. Cen, J. Guo, J. Wang, W. Fan, RSC Adv. 2017, 7,
8567–8574.
[33] M. Zheng, J. Pang, R. Sun, A. Wang, T. Zhang, ACS Catal. 2017, 7,
[18] F. Di Gregorio, V. Keller, Journal of Catalysis 2004, 225, 45-55.
[19] J.C. Vartuli, J.G. Santiesteban, P. Traverso, N. Cardona-Martinez, C.D.
Chang, S.A. Stevenson, Journal of Catalysis 1999, 187, 131-138.
[20] D.E. Lopez, K. Suwannakarn, D.A. Bruce, J.G. Goodwin Jr., Journal of
Catalysis 2007, 247, 43–50.
[21] P. Yuan, Y. Liu, F. Bai, L. Xu, Z, Cheng, W. Yuan, Catalysis
Communications 2011, 12, 753-756.
[22] P. Wongmaneenil, B. Jongsomijt, P. Praserthdam, Catalysis
Communications 2009, 10, 1079-1084.
1939-1954.
[34] Y.-B. Huang, Y. Fu, Green Chem. 2013, 15, 1095-1111.
[35] P. Rout, A. Nannaware, O. Prakash, R. Rajasekharan, ChemBioEng
Rev 2014, 1, 96-116.
[36] C. Hammond, Green Chem. 2017, 19, 2711-2728.
[37] H. Xiong, H. N. Pham, A. K. Datye, Green Chem. 2014, 16, 4627-4643.
[38] R. Ravenelle, F. Schüßler, A. D’Amico, N. Danilina, J. van Bokhoven, J.
Lercher, C. Jones, C. Sievers, J. Phys. Chem. 2010, 114, 19582-
19595.
[23] R.Kourieh, S. Bennici, M. Marzo, A. Gervasini, A. Auroux, Catalysis
Communications 2012, 19, 119-126.
[39] L. Zhang, K. Chen, B. Chen, J. White, D. Resasco, J. Am. Chem. Soc.
2015, 137, 11810-11819.
[24] F. Chambon, F. Ratboul, C. Pinel, A. Cabiac, E. Guillon, N. Essyem,
Applied Catalysis B: Environmental 2011, 105, 171-181.
[25] F. Chambon, F. Ratboul, C. Pinel, A. Cabiac, E. Guillon, N. Essyem,
Applied Catalysis A: General 2015, 504, 664-671.
[40] I. Bakare, O. Muraza, T. Kurniwan, Z. Yamani, E. Shafei, A. Punetha,
K. Choi, T. Yokoi, Micropor. Mesopor. Maer. 2016, 233, 93-101.
[41] A. Vjunow, M. Derewinski, J. Fulton, D. Camaioni, J. Lercher, J. Am.
Chem. Soc. 2015, 137, 10374-10382.
[26] J. Li, L. Hui, Y. Jian, S. Wenfeng, J.Wuhan, Uni. Sci. Ed. 2010, 25, 919-
[42] M. Grilc, B. Likozar, J. Levec, ChemCatChem 2016, 8, 180-191.
923.
[43] M. Grilc, B. Likozar, J. Levec, Catal. Today 2015, 256, 302-314.
[27] M. Dadachov, R. Lambrecht, J.Mater.Chem. 1997, 7, 1867-1870.
[28] E.Dedova, A.Gubanov, E. Filatov, T. Kardash, S. Kulkov, AIP
Conference Proceedings 2016, 1783, 020036.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701113
ChemCatChem
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
FULL PAPER
Martin Lucas*, Katarina Fabičovicova
and Peter Claus
((Insert TOC Graphic here; max. width: 11.5 cm; max. height: 2.5 cm))
Page No. – Page No.
Hydrothermally stable ruthenium-
zirconium-tungsten catalyst for
cellulose hydrogenolysis to polyols
Solid acid and effective: based on ruthenium, zirconium and tungsten, a
hydrothermal stable catalyst for cellulose conversion has been developed. In
recycling experiments, a consistently high yield of polyols was achieved. The
carbon efficiency achievable with the Ru/Zr-W-oxide is nearly 100%.
This article is protected by copyright. All rights reserved.