The Role of Metallic Copper in the Selective Hydrodeoxygenation of Glycerol to 1,2-Propanediol over Cu/ZrO2

Article abstract of DOI:10.1002/cctc.201701748

A series of Cu/ZrO₂ catalysts with nominal CuO loadings of 5, 10, 18 and 31 wt.% was synthesized by co-precipitation, characterized and applied in the hydrodeoxygenation of glycerol under mild reaction conditions (200 °C, 25 bar H₂). These catalysts were highly selective for the cleavage of C?O bonds while preserving C?C bonds leading to 95 % selectivity to 1,2-propanediol. The conversion of glycerol was observed to be linearly correlated with the specific copper surface area derived from N₂O frontal chromatography. The reaction was found to occur through the dehydration of glycerol to acetol followed by its hydrogenation to 1,2-propanediol. Metallic copper was identified as the active site for both reactions suggesting the acid ZrO₂ sites to be blocked by water. Reusability studies showed that the catalyst was relatively stable and the conversion decreased by only 18 % after three cycles.

Full text of DOI:10.1002/cctc.201701748

Accepted Article
Title: The role of metallic copper in the selective hydrodeoxygenation of
glycerol to 1,2-propanediol over Cu/ZrO2
Authors: Thomas Gabrysch, Baoxiang Peng, Sorin Bunea, Gerald
Dyker, and Martin Muhler
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.201701748
Link to VoR: http://dx.doi.org/10.1002/cctc.201701748
A Journal of
www.chemcatchem.org

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
The role of metallic copper in the selective hydrodeoxygenation
of glycerol to 1,2-propanediol over Cu/ZrO2
Thomas Gabrysch[a], Baoxiang Peng*[a,b], Sorin Bunea[a], Gerald Dyker[c] and Martin Muhler*[a,b]
A series of Cu/ZrO2 catalysts with nominal CuO loadings of 5, 10, 18
and 31 wt.% was synthesized by co-precipitation, characterized and
applied in the hydrodeoxygenation of glycerol under mild reaction
conditions (200 °C, 25 bar H2). These catalysts were highly selective
for the cleavage of C-O bonds while preserving C-C bonds leading
to 95% selectivity to 1,2-propanediol. The conversion of glycerol was
observed to be linearly correlated with the specific copper surface
area derived from N2O frontal chromatography. The reaction was
found to occur through the dehydration of glycerol to acetol followed
by its hydrogenation to 1,2-propanediol. Metallic copper was
identified as the active site for both reactions suggesting the acid
ZrO2 sites to be blocked by water. Reusability studies showed that
the catalyst was relatively stable and the conversion decreased by
only 18% after three cycles.
Introduction
Due to the gradual depletion of fossil resources, numerous
efforts have been made to develop sustainable processes
converting renewable feedstocks into value-added products.[1,2]
In the biodiesel process, glycerol (1,2,3-propanetriol) is obtained
as by-product with approximately 10 wt.% of the total biorefinery
products.[3,4] The sharp increase of the biodiesel production
1,2-PDO is the selective cleavage of the primary C-O bond while
preserving the C-C bonds as well as suppressing the formation
of overhydrogenolysis products to increase the selectivity to
1,2-PDO. Catalysts applied for the HDO are usually based on
noble metals (Pt, Ru, Rh) or transition metals (Cu, Ni, Co). It is
found that noble metals offer a higher activity for hydrogenolysis
reactions, but the selectivity towards 1,2-PDO is considerably
lower because of the parallel activation of C-C bond hydro-
genolysis. On the contrary, transition metal-based catalysts are
less active, but offer higher selectivity for the HDO of the primary
hydroxyl group.[18]
during the last decades resulted in
a
considerable
overproduction of glycerol,[5] exceeding the global demands by
far and causing a strong decrease of the market price.[6] The
high availability, low cost, non-toxicity and high functionality turn
glycerol into a versatile platform chemical for the production of
value added biomass-derived products.[7-9]
Multiple HDO reaction mechanisms are reported in
literature including a three-step dehydrogenation-dehydration-
hydrogenation,[19,20] a two-step dehydration-hydrogenation,[3,21]
Recently, many studies focused on the catalytic trans-
formation of glycerol, such as esterification,[10-12] etherification,[13]
hydrogenolysis/hydrodeoxygenation,[14-16] dehydration and refor-
ming,[17] of which the hydrodeoxygenation (HDO) to
1,2-propanediol (1,2-PDO) is one of the most attractive valo-
rization approaches. 1,2-PDO is widely used in the production of
unsaturated polyester resins and polyurethanes and has direct
applications as humectant in foods, cosmetics and
pharmaceuticals. The challenging aspect of glycerol HDO to
and
a
direct HDO mechanism over ReOx-modified Ir
catalysts.[22,23] The reaction is strongly affected by the type of
catalyst used and the applied reaction conditions. It is widely
accepted that the dehydration step of glycerol proceeds on the
Brønsted or Lewis acid-base sites of the support, while the
metallic sites are solely required for the consecutive
hydrogenation to 1,2-PDO.[24,25] Moreover, mechanistic studies
have shown that the addition of acidic catalysts like zeolites,[26]
[27]
Nb2O5
and Amberlyst ion exchange resins[28-30] as well as
[a]
T. Gabrysch, Dr. B. Peng, S. Bunea, Prof. Dr. M. Muhler
Laboratory of Industrial Chemistry
Ruhr-University Bochum
alkaline additives[31] such as NaOH and KOH significantly
increase glycerol conversion, indicating a strong influence of the
acid-base properties on the dehydration step.
D-44780 Bochum (Germany)
E-mail: baoxiang.peng@techem.rub.de
muhler@techem.rub.de
ZrO2 is frequently used as support due to its high
mechanical and thermal stability,[32] high specific surface area,[32]
suitable acid-base[33] as well as redox properties.[34] For example,
Cu nanoparticles supported on ZrO2 are applied in methanol
synthesis,[35] in the hydrogenation of ethyl acetate[36] and in the
isomerization of n-butane.[37] The Lewis acid-base properties of
ZrO2 originate from alternating unsaturated Zr4+ and O2- pairs at
the surface, while Brønsted acidity is due to surface OH groups.
The influence of the bifunctional properties of Cu/ZrO2 due to the
presence of metallic Cu0 and acidic Zr-OH groups on the kinetics
of the hydrogenation of ethyl acetate in the gas phase was
investigated in detail by Schittkowski et al.[36]
[b]
[c]
Dr. B. Peng, Prof. Dr. M. Muhler
Max Planck Institute for Chemical Energy Conversion
45470 Mülheim an der Ruhr (Germany)
Prof. Dr. Gerald Dyker
Laboratory of Organic Chemistry II
Ruhr-University Bochum
D-44780 Bochum (Germany)
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
Here, the HDO of glycerol and the conversion of the
intermediates were systematically studied over a series of
Cu/ZrO2 with different CuO loadings to elucidate the reaction
mechanism of the polyol transformation in the aqueous phase
involving C-O and C-C bond cleavage.
Results and Discussion
Catalyst Characterization
Table 1. Physicochemical properties of the co-precipitated CuO/ZrO2
catalysts.
CuO loading
[wt.%]
Surface area
[m² g-1]
Cu surface area
[m² g-1]
Degree of reduction
[%]
0
140
158
157
144
142
0.0
2.9
3.9
4.2
5.7
-
n.d.*
95
5
10
18
31
96
98
* n.d. = not determined
The specific surface areas of the co-precipitated CuO/ZrO2
catalysts determined by N2 physisorption, the Cu surface areas
determined by N2O reactive frontal chromatography (RFC) and
the degrees of reduction derived from the temperature-
programmed reduction (TPR) in H2 are summarized in Table 1.
The specific surface areas are found to be in the range from 140
to 158 m² g-1, which decrease slightly with increasing CuO
loading from 5 wt.% to 31 wt.%. For CuO loadings higher than
5 wt.%, the degree of reduction is close to 100%. The specific
Cu surface areas increase with increasing CuO loading, but do
not show a linear correlation, which is probably due to the
agglomeration of Cu nanoparticles especially at higher CuO
loading. The highest Cu surface area is found to be 5.7 m² g-1 for
31 wt.% CuO/ZrO2. These Cu surface areas are low, suggesting
that the Cu is in the form of nanoparticles embedded in the ZrO2
support.[36]
The X-ray diffraction (XRD) patterns of the calcined
CuO/ZrO2 samples with CuO loadings of 5, 10, 18 and 31 wt.%
as well as the Cu-free ZrO2 support are shown in Figure 1. Pure
ZrO2 crystallized in the tetragonal structure, indicated by the
characteristic reflections at 30.2°, 35.5°, 50.3° and 60.2° 2θ. The
reflections observed for 5 wt.% CuO/ZrO2 also indicate the
presence of tetragonal ZrO2. A further increase of the CuO
loading up to 31 wt.% leads to strong structural changes: the
presence of CuO prevented the crystallization of ZrO2 resulting
in an essentially X-ray amorphous state. For the 31 wt.%
CuO/ZrO2 two additional reflections at 35.6° and 38.9° 2θ are
observed, indicating the formation of a second CuO species on
the ZrO2 support.[36]
Figure 1. XRD patterns of pure ZrO2 and the CuO/ZrO2 samples. All samples
were synthesized by co-precipitation at pH 10.5 and calcined at 490 °C for 3 h
in 20% O2 in N2.
The TPR profiles of the calcined samples are shown in
Figure 2. For pure ZrO2 no reduction peak was observed,
whereas the TPR profiles of the 10 wt.% and 18 wt.% CuO/ZrO2
samples show one symmetric reduction peak at 130 °C due to
the reduction of Cu2+ to metallic Cu0. Two reduction peaks were
observed for 31 wt.% CuO/ZrO2 at temperatures around 127 °C
and 144 °C, indicating a less homogeneous dispersion of the Cu
nanoparticles. Based on the low Cu surface areas in spite of the
high CuO loading and the structural change of ZrO2 from
tetragonal ZrO2 to an amorphous state with increasing CuO
loading, it can be concluded that most Cu nanoparticles are
embedded homogeneously in the ZrO2 support.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
phenol and concluded that at increased pressures, water
molecules preferentially bind to the surface of the catalyst, thus
blocking the active sites for the adsorption of reactants.
The conversion of glycerol depends strongly on the CuO
loading of the catalysts (Figure S1). Up to 10 wt.% CuO loading
conversion increased almost linearly to 25%, yielding 1,2-PDO
as main product with a selectivity of about 95%. Although the
nominal CuO loading was further increased by a factor of 3 to
31 wt.%, the conversion only slightly increased from 25% to 35%,
while the selectivity to 1,2-PDO remained almost constant. The
main by-product was ethylene glycol, probably originating from
the overhydrogenolysis of 1,2-PDO or from the direct cleavage
of the C-C bond through direct hydrogenolysis of glycerol. For
the catalysts with higher CuO loadings the selectivity to ethylene
glycol slightly increased from 2.1% to 4.7%. The highest
conversion of 35% was achieved with the 31 wt.% CuO/ZrO2
catalyst, which is only 4% higher in comparison to 18 wt.%
CuO/ZrO2. 18 wt.% CuO/ZrO2 was identified as an optimum
catalyst with appropriate CuO loading and activity and was
chosen for further mechanistic studies. In contrast, the catalytic
activity was found to be correlated linearly with the specific Cu
surface areas (Figure 3), indicating that metallic Cu is the active
site for the HDO of glycerol.
Table 2. Glycerol HDO over Cu/ZrO2 catalysts with varying CuO loading.[a]
Figure 2. TPR profiles of calcined pure ZrO2 and the CuO/ZrO2 samples.
Selectivity [%]
CuO
loading
[wt.%]
Conversion
[%]
1,2-
PDO
Ethylene
glycol
2-
Influence of the CuO loading on the catalytic properties
Acetol
MeOH[b]
PrOH[b]
The Cu/ZrO2 samples with nominal CuO loadings of 5, 10, 18
and 31 wt.% were applied in the liquid-phase hydro-
deoxygenation of glycerol using aqueous 4 wt.% glycerol
solutions at 200 °C under 25 bar H2 pressure for 8 h (Table 2).
Prior to the catalytic tests, all catalysts were pre-reduced in the
autoclave at 200 °C and 20 bar H2 pressure for 10 h in the
absence of solvent and reaction solution. The identified liquid
products were 1,2-PDO, ethylene glycol, acetol as well as trace
0
<0.3
14.7
25.0
31.3
35.0
-
-
-
-
-
5
95.2
95.2
94.5
93.3
2.1
3.0
3.5
4.7
1.6
1.2
0.8
0.9
-
1.1
1.0
1.0
0.8
10
18
31
0.3
0.3
0.3
amounts
of methanol
and 2-propanol.
Interestingly,
[a] Reaction conditions: 4 wt.% aqueous glycerol solution (20 mL); T = 200 °C;
p(H2) = 25 bar, 400 mg catalyst; 8 h reaction time; 750 rpm stirring speed; all
catalysts pre-reduced at 200 °C for 10 h.
1,3-propanediol was not detected, suggesting a higher reactivity
of the primary than the secondary hydroxyl group. According to
the ab initio calculations (B3LYP-6-31G*) an intermediary
propen-1,2-diol is thermodynamically favored by about
30 kJ mol-1 due to its intramolecular hydrogen bond compared
with the isomeric propen-1,3-diol. Thus, a detailed mechanistic
interpretation has to consider a concerted E2 elimination,[38]
which is typical for the dehydration of primary alcohols, as well
as glycidol as an alternative dehydration product, which can be
hydrogenated to 1,2-PDO, too.[39]
[b] 2-PrOH: 2-propanol; MeOH: methanol
Pure ZrO2 led to less than 0.3% conversion without the
formation of 1,2-PDO and acetol, indicating that Cu-free ZrO2
can neither catalyze the hydrodeoxygenation of glycerol to
1,2-PDO nor the dehydration of glycerol to acetol under the
applied aqueous-phase conditions presumably due to the
coverage of the acidic sites by water. This phenomenon was
also observed by Engelhardt et al.,[40] who studied the influence
of water on molybdenum carbide catalysts applied in the HDO of
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
Figure 4. Glycerol conversion and liquid product selectivity in the HDO of
glycerol as a function of time. Reaction conditions: 20 mL of 4 wt.% aqueous
glycerol solution, 400 mg 5 wt.% CuO/ZrO2, 200 °C, 25 bar H2 pressure and
750 rpm stirring speed.
Figure 3. Correlation of the degrees of glycerol conversion with the specific
Cu surface areas.
Reaction pathway of glycerol HDO to 1,2-PDO
To explore the reaction pathways of glycerol HDO, the
kinetics was investigated with 5 wt.% CuO/ZrO2 catalyst under
identical reaction conditions (Figure 4). Conversion increased
linearly to 15% as the reaction time increased to 8 h, indicating
that the catalytic HDO of glycerol is a zero-order reaction. The
selectivity of 1,2-PDO and ethylene glycol remained almost
constant at 95% and 2.5%, respectively. More interestingly, the
selectivity of acetol gradually decreased from 5% at 2% glycerol
conversion to 1.6% with the progress of reaction, indicating that
acetol originating from the dehydration of glycerol is the main
intermediate and primary product during 1,2-PDO production. To
prove this conjecture, the reaction was performed with the pre-
reduced 18 wt.% CuO/ZrO2 catalyst under inert N2 atmosphere
at 200 °C for 8 h. As expected, acetol was obtained as the main
product reaching 79% selectivity at 18% glycerol conversion.
1,2-PDO and ethylene glycol were found as by-products with
selectivities of 12% and 8%, respectively. In addition, the
intermediate acetol was further converted over Cu/ZrO2 under
identical conditions in a separate experiment (Figure 5a). The
result showed that acetol was almost exclusively converted into
1,2-PDO within only 1 h, attaining 98% yield at 98.5%
conversion. Thus, the HDO of glycerol proceeds via the slow
dehydration of glycerol to acetol as rate-determining step
followed by the fast hydrogenation of acetol to 1,2-PDO. The
amount of 1.4% unconverted acetol, which is consistent with the
low acetol concentration after 8 h during glycerol conversion
(Figure 4), is presumably due to the equilibrium between acetol
and 1,2-PDO at high H2 pressure. Other by-products produced
by consecutive overhydrogenolysis reactions were identified as
methanol (0.1%), ethanol (0.2%), 1-propanol (0.15%) and
2-propanol (0.25%).
Figure 5. Product distributions of the reactions using acetol (a) and
1,2-PDO (b) as the reactant. Reaction conditions: 20 mL of 4 wt.% aqueous
reactant solutions, 400 mg of 18 wt.% CuO catalyst (pre-reduced), 200 °C,
25 bar H2 pressure and 750 rpm stirring speed.
It is generally accepted that over metal-supported catalysts
dehydration occurs on the acidic support (e.g., Al2O3, ZrO2), and
hydrogenation is catalyzed by the metal (e.g., Pt, Cu).[16] In the
present work, the hydrogenation step occurs on metallic Cu;
while the active site for the rate-determining dehydration step is
still unclear. Therefore, pure metallic Cu derived from the
reduction
of
commercial
CuO
(Sigma
Aldrich,
nanopowder < 50 nm) at 225 °C and 20 bar H2 for 10 h was
applied in the glycerol dehydration under inert N2 atmosphere.
The corresponding XRD patterns of the commercial CuO before
and after reduction (Figure S2) indicate the complete reduction
of Cu2+ to metallic Cu0 prior to reaction. The TPR profile of pure
o
CuO shows one reduction peak at 181 C (Figure S3), which is
about 40 oC higher than all the supported catalysts (Figure 2),
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
suggesting that ZrO2 can efficiently increase the reducibility of
CuO. The catalytic results showed that glycerol was converted
into acetol with 100% selectivity during the whole reaction time
yielding 6.5% conversion after 8 h (Figure S4). In comparison,
205 mg CuO nanoparticles were used vs. 400 mg of 18 wt.%
CuO/ZrO2. This observation suggests that pure metallic Cu is an
efficient and active catalyst for the dehydration of glycerol. This
conjecture is consistent with literature reports that pure metallic
Cu catalyzes the dehydration of glycerol to acetol in the gas
phase under inert conditions, attaining a yield of 64% acetol at
250 °C.[24] Cu-catalyzed dehydration was proposed to proceed
via the formation of Cu alkoxide species.[24] In the liquid phase
the dehydration of glycerol to acetol was also reported using
metallic Cu foil under ambient N2 atmosphere, leading to 45%
selectivity of acetol at a very low conversion of less than 1% due
to the extremely low surface area.[41]
Combined with the observation that pure ZrO2 cannot
catalyze the dehydration of glycerol presumably due to the
blockage of acid sites by water in the aqueous phase, it can be
concluded that metallic Cu is not only the active species for the
hydrogenation, but also the active site for the dehydration of
glycerol to acetol, resulting in a linear correlation of the
conversion of glycerol with the specific Cu surface area
(Figure 3). ZrO2 probably acts mainly as support and increases
the dispersion of Cu through embedding the Cu nanoparticles
efficiently in its porous structure. Furthermore, the bifunctional
nature of Cu/ZrO2 in the gas-phase hydrogenation of ethyl
acetate[36] and the dominant role of metallic Cu in the
hydrodeoxygenation of glycerol under the present liquid-phase
conditions point to the important role of water as inhibiting
adsorbate in the aqueous-phase reaction.
The reactivity of 1,2-PDO was also investigated to
determine its stability and possible overhydrogenolysis reaction
products under the present reaction conditions (Figure 5b). The
use of 1,2-PDO as reactant resulted in a tiny conversion of
0.80% after 8 h reaction time yielding trace amounts of acetol
(0.6%), 1-propanol (0.16%) and methanol (0.04%). Since
1,2-PDO is very stable under the chosen reaction conditions, we
can conclude that ethylene glycol is not the overhydrogenolysis
product of 1,2-PDO, suggesting that it originates from the direct
hydrogenolysis of glycerol. As the primary and main product of
1,2-PDO conversion, acetol is mainly formed through the
dehydrogenation of 1,2-PDO catalyzed by metallic Cu, being
limited by the employed high H2 pressure. This observation of
the different catalytic behavior of Cu/ZrO2 in the dehydration of
glycerol to acetol and in the dehydrogenation of 1,2-PDO to
acetol is consistent with literature reports on Raney Cu and
Cu/Al2O3.[24]
Scheme 1. Reaction pathways of the hydrodeoxygenation of glycerol over
Cu/ZrO2 catalysts.
The identified reaction pathways of the glycerol
hydrodeoxygenation and hydrogenolysis reactions are shown in
Scheme 1. The high selectivity to 1,2-PDO is mainly determined
by the stability of the product 1,2-PDO, the selective scission of
C-O over C-C bonds, and the selective dehydration of the
primary hydroxyl group instead of the secondary hydroxyl group.
First, using 1,2-PDO as reactant demonstrated the high stability
of 1,2-PDO, avoiding undesired further degradation products.
Second, the dominant selectivity of 1,2-PDO compared with
ethylene glycol (95% vs. < 5%) suggests the selective cleavage
of C-O over C-C bonds, which can be attributed to the relatively
weak ability of Cu0 to catalyze hydrogenolysis as well as the
stronger C-C bond strength of glycerol compared with that of the
C-O bond (Table S1, i.e., 347 vs. 335 kJ mol-1). Third, the
absence of 1,3-propanediol in the formed products and the
identification of acetol as the primary and dominant intermediate
indicate that the selective dehydration of the terminal hydroxyl
group to acetol is favored over the secondary hydroxyl group
dehydration to 3-hydroxypropanal. This conjecture cannot be
explained by the C-O bond strengths of primary and secondary
OH groups (Table S1, i.e., 333.0 vs. 335.6 kJ mol-1), which are
quite similar, but points to the different reactivities of these two
OH groups. Our previous study showed that the overall reaction
rates of C3 alcohols over Pt/Al2O3 decrease in the sequence
1,3 propanediol ≈ glycerol > 1,2-propanediol ≈ 1-propanol, indi-
cating that the primary OH group is much more reactive for
dehydration than the secondary OH group.[42] Furthermore, the
ability of Cu/ZrO2 to catalyze dehydration is dramatically
decreased due to the blockage of acidic Zr-OH sites by water in
the aqueous phase, leading to the elimination of the terminal OH
group rather than the more difficult secondary OH group
elimination.
Catalyst stability
Owing to the superior performance of the Cu/ZrO2 catalyst
for the HDO of glycerol, the stability of the catalytic performance
was also investigated (Figure 6). During the recycling test, the
mass loss of the catalyst after each cycle was approximately
27 mg (8%) in average. The reusability results showed that the
catalyst lost around 18% of its activity after 3 cycles, indicating
the catalyst was rather stable for the HDO of glycerol.
Additionally, the selectivity was not affected by catalyst recycling.
The reaction solutions after the first and third cycle were
analyzed by AAS to determine if Cu leaching into the solution
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
relatively stable with only a slight decrease of the catalytic
activity after three reaction cycles.
Experimental Section
Catalyst preparation
Precursors of the Cu/ZrO2 catalysts were prepared by co-precipitation
method. Calculated amounts of copper nitrate (Sigma Aldrich, 99.999%)
and zirconium oxynitrate (Sigma Aldrich, 99.0%) were dissolved in
100 mL deionized water (HPLC purity) and precipitated at a constant pH
value of 10.5 using a 25 wt.% sodium hydroxide solution as precipitating
agent. The resulting precipitates were filtered and washed with deionized
water until the filtrate was free of nitrates. After drying at 105 °C overnight,
the precursors were calcined at 490 °C for 3 h in synthetic air
(20% O2/N2), and a series of CuO/ZrO2 catalysts with nominal CuO
loadings of 5, 10, 18 and 31 wt.% was obtained.
Figure 6. Reusability of Cu/ZrO2 in the hydrodeoxygenation of glycerol. The
compensation of the catalyst mass loss is taken into account in the relative
conversion calculation. Reaction conditions: 20 mL of 4 wt.% aqueous
glycerol solution, 400 mg of 18 wt.% CuO/ZrO2 (pre-reduced), 25 bar H2
pressure, 750 rpm stirring speed and 8 h reaction time.
Characterization
Temperature-programmed reduction (TPR) measurements were carried
out in a flow set-up using 5 vol.% H2 in Ar at a flow rate of 84 mL min-1.
For each experiment 100 mg of catalyst in a sieve fraction of 250 µm to
355 µm were loaded in a stainless-steel U-tube reactor. The samples
were heated up to 240 °C with a heating rate of 1 °C min-1. The H2
concentration in the exhaust gas was measured by a thermal conductivity
detector.
had occurred. Only 0.74% (20 ppm) and 0.54% (15 ppm) of the
total Cu was found to be leached into the solution after the first
and third cycle, respectively. This low degree of Cu leaching
points to an efficient stabilization of the metallic Cu nanoparticles
in the ZrO2 support under the applied reaction conditions. The
XRD patterns (Figure S5) of the freshly reduced catalyst and the
catalyst after three cycles reveal structural changes of metallic
Cu from small particle sizes to larger agglomerates. This
agglomeration of metallic Cu could lead to the decrease of the
specific Cu surface area and thus cause the deactivation of the
catalyst. A similar deactivation mechanism was also reported by
Vasiliadou et. al. for HDO of glycerol with Cu/SiO2.[43]
N2O reactive frontal chromatography (N2O-RFC) measurements were
carried out after the TPR measurements with a flow rate of 1% N2O in He
at room temperature. The exhaust gas stream was analysed with an
Advance Optima AO2020 continuous gas analyser (ABB).
X-Ray powder diffraction (XRD) patterns were recorded with an
PANalytical theta-theta powder diffractometer equipped with a Cu Kα
radiation source in a 2θ range of 10° to 80°. Qualitative phase analysis
was carried out using HighScore Plus V.3.0 software and ICDD powder
diffraction database.
Conclusions
A
series of co-precipitated Cu/ZrO2 catalysts was
For the N2 physisorption measurements (BET and BJH) 200 mg of the
calcined catalysts in a sieve fraction of 250 µm to 355 µm were used in a
BELSORP-max device. Prior to the measurements, the catalysts were
heated to 200 °C for 2 h under vacuum to remove residual water. The
measurements were carried out at -196 °C with a relative pressure (p/p0)
in the range of 0 to 1. The specific surface area, average pore volume
and average pore diameters were calculated from the obtained isotherms.
synthesized and applied in the aqueous-phase hydro-
deoxygenation of glycerol (4 wt.% in water) at 200 °C and 25 bar
H2 pressure. The catalysts exhibited high activity and selectivity
of about 95% for the terminal C-O bond cleavage to
1,2-propanediol while preserving the C-C bonds, in this way
avoiding the formation of by-products. The HDO of glycerol was
found to proceed via a two-step dehydration-hydrogenation
reaction pathway forming acetol as dehydration intermediate,
which is further hydrogenated to 1,2-propanediol. As major by-
product ethylene glycol was produced from the direct C-C
hydrogenolysis of glycerol with a selectivity of less than 5%.
Glycerol conversion was linearly correlated with the specific Cu
surface area of the catalysts. In addition, using pure metallic Cu
revealed the dominant role of metallic Cu in the dehydration of
glycerol to acetol. The acidic Zr-OH sites did not contribute to
the catalytic activity presumably due to blockage by abundant
water under the employed aqueous-phase reaction conditions.
Studies of the catalyst reusability showed that the catalyst was
Atomic absorption spectroscopy (AAS) of the liquid reaction solutions
were carried out using a VARIAN AA 300 spectrometer equipped with a
flame ionization system to determine the Cu leaching.
Catalytic tests
The catalytic performance of the Cu/ZrO2 catalysts was tested in a Parr
5050 stainless-steel autoclave reactor (100 mL). Prior to the catalytic test,
the catalyst was pre-reduced in the reactor in the absence of liquid
reaction solution in order to form metallic Cu active sites. For each test
400 mg catalyst was loaded in the reactor. Subsequently, the reactor was
flushed with H2 (99.999%) to remove ambient air and pressurized and
subsequently heated to 200 °C at 20 bar for 10 h to reduce the catalysts.
After cooling and depressurization, 20 g of 4 wt.% aqueous glycerol
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
[1]
[2]
E. M. Shahid, Y. Jamal, Renew. Sustain. Energy Rev. 2008, 12, 2484–
2494.
V. B. Borugadda, V. V. Goud, Renew. Sustain. Energy Rev. 2012, 16,
solution was loaded through a sampling tube (without opening the
reactor) to minimize the exposition of the reduced catalyst to air. After
flushing the reactor with H2 again, the reaction was performed at 200 °C,
25 bar H2 partial pressure at a stirring speed of 750 rpm for 8 h. During
reaction liquid samples of 1.5 mL were taken out after 1, 3, 5 and 8 h.
The samples were filtered using membrane filters and analyzed by gas
chromatography.
4763–4784.
[3]
[4]
[5]
[6]
[7]
Y. Nakagawa, K. Tomishige, Catal. Sci. Technol. 2011, 1, 179–190.
Y. Zheng, X. Chen, Y. Shen, Chem. Rev. 2008, 108, 5253–5277.
D. T. Johnson, K. A. Taconi, Environ. Prog. 2007, 26, 338–348.
Y. Gu, F. Jérôme, Green Chem. 2010, 12, 1127-1138.
M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, P. C. Della, Angew.
Chem., Int. Ed. 2007, 46, 4434–40.
[8]
[9]
S. Bagheri, N. M. Julkapli, W. A. Yehye, Renew. Sustain. Energy Rev.
2015, 41, 113–127
J. Dam, U. Hanefeld, ChemSusChem 2011, 4, 1017–1034.
Prior to the catalytic test under inert nitrogen gas atmosphere, the
catalyst was pre-reduced in the reactor in the absence of the reaction
solution, according to the method described above. After cooling and
depressurization, 20 g of 4 wt.% aqueous glycerol solution was loaded
through a sampling tube (without opening the reactor) to minimize the
exposition of the reduced catalyst to ambient air. After flushing the
reactor several times with N2, the reactor was pressurized with N2 to
12 bar and heated to 200 °C. The reaction was run for 8 h at a stirring
speed of 750 rpm. During reaction 1.5 mL liquid samples were taken out
after 1, 3, 5 and 8 h. The samples were filtered using membrane filters
and analyzed by gas chromatography.
[10] M. Trejda, K. Stawicka, M. Ziolek, Appl. Catal., B 2011, 103, 404–412.
[11] W. D. Bossaert, D. E. De Vos, W. M. Van Rhijn, J. Bullen, P. J. Grobet,
P. A. Jacobs, J. Catal. 1999, 182, 156–164.
[12] J. M. Clacens, Y. Pouilloux, J. Barrault, Appl. Catal., A 2002, 227, 181–
190.
[13] A. M. Ruppert, J. D. Meeldijk, B. M. W. Kuipers, B. H. Erne, B. M.
Weckhuysen, Chem. - Eur. J. 2008, 14, 2016–2024.
[14] Y. L. Wang, J. X. Zhou, X. W. Guo, RSC Adv. 2015, 5, 74611–74628.
[15] A. M. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem., Int. Ed. 2012,
51, 2564-2601
[16] E. S. Vasiliadou A. A. Lemonidou, WIREs Energy Environ. 2015, 4,
486-520.
[17] P. D. Vaidya, A. E. Rodrigues, Chem. Eng. Technol. 2009, 32, 1463–
1469.
[18] H. Zhao, D. S. Tong, L. M. Wu, W. H. Yu, Catal. Rev. 2013, 55, 369–
For catalytic reusability tests, the HDO reaction was performed without
sampling during the reaction and the catalyst was recycled after 8 h from
the reaction solution by centrifugation, washing with deionized water and
freeze-drying overnight. The mass loss of the catalyst after each cycle
was approximately 8%. After each cycle the catalyst was pre-reduced
prior to the reaction in the absence of any solvent or solution and
subsequently reused for the HDO of glycerol.
453.
[19] C. Montassier, D. Giraud, J. Barbier, J. P. Boitiaux, Bull. Soc. Chim. Fr.
1989, 2, 148-155.
[20] C. Montassier, J. C. Menezo, L. C. Hoang, C. Renau, J. Barbier, J. Mol.
Catal. 1991, 70, 99–110.
[21] M. A. Dasari, P. P. Kiatsimkul, W. R. Sutterlin, G. J. Suppes, Appl.
Catal., A 2005, 281, 225–231.
[22] Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige, Appl.
Catal., B 2010, 94, 318-326.
Analysis of liquid-phase products
[23] Y. Amada, Y. Shinmi, S. Koso, T. Kubota, Y. Nakagawa, K. Tomishige,
Appl. Catal., B. 2011, 205, 117-127.
[24] S. Sato, M. Akiyama, R. Takahashi, T. Hara, K. Inui, M. Yokota, Appl.
Liquid-phase reaction products were analysed by an Agilent 7820A gas
chromatography
equipped
with
a
capillary
column
(ZB-
Catal., A 2008, 347, 186–191. [25]
M. Balaraju, V. Rekha, P. S.
WAXplus, 30 m × 0.32 µm × 0.25 µm) and a flame ionization detector
(FID). The GC was calibrated using external standard calibration
resulting in relative standard deviations (RSD%) of 2-4% depending on
the compound. The conversion of glycerol as well as the selectivity and
yield were calculated based on liquid products using the following
equations:
Sai Prasad, R. B. N. Prasad, N. Linghaiah, Catal. Lett. 2008, 126, 119–
124.
[26] Y. Nakagawa, X. H. Ning, Y. Amada, K. Tomishige, Appl. Catal., A
2012, 433, 128-134.
[27] R. Rodrigues, N. Isoda, M. Gonçalves, F. C. A. Figueiredo, D. Mandelli,
W. A. Carvalho, Chem. Eng. J. 2012, 198–199, 457–467.
[28] T. Miyazaw, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 2006,
240, 213–221.
[29] P. Centomo, V. Nese, S. Sterchele, M. Zecca, Top. Catal. 2013, 56,
822-830.
[30] T. Miyazawa, S. Koso, K. Kunimori, K. Tomishige, Appl. Catal., A 2007,
329, 30-35.
[31] A. Marinoiu, C. Cobzaru, E. Carcadea, C. Capris, V. Tanislav, M.
Raceanu, React. Kinet., Mech. Catal. 2013, 110, 63–73.
[32] L. Castro, P. Reyes, C. Correa, J. Sol-Gel Sci. Technol. 2002, 25, 159-
168.
[33] M. D. Rhodes, A. T. Bell, J. Catal. 2005, 233, 198–209.
[34] J. Matos, F. Anjos Junior, L. Cavalcante, V. Santos, S. Leal, L. Santos
Junior, M. Santos, E. Longo, Mater. Chem. Phys. 2009, 117, 455–459.
[35] K. D. Jung, A. T. Bell, J. Catal. 2000, 193, 207–223.
[36] J. Schittkowski, K. Tölle, S. Anke, S. Stürmer, M. Muhler, J. Catal. 2017,
352, 120–129.
where νi is the stoichiometric factor, 1 for glycerol, acetol, 1,2-PDO and
propanols, 2/3 for ethylene glycol and ethanol and 1/3 for methanol.
[37] C. R. Vera, C. L. Pieck, K. Shimizu, J. M. Parera, Appl. Catal., A 2002,
230, 137–151.
[38] F. Sato, S. Sato, Y. Yamada, M. Nakamura, A. Shiga, Catal. Today
2014, 226, 124-133.
[39] R. Cucciniello, C. Pironti, C. Capacchione, A. Proto, M. Di Serio, Catal.
Commun. 2016, 77, 98-102.
Acknowledgements
[40] J. Engelhardt, P. Lyu, P. Nachtigall, F. Schüth, A. M. Garcίa,
ChemCatChem 2017, 9, 1985-1991.
[41] R. B. Mane, A. Yamaguchi, A. Malawadkar, M. Shirai, C. V. Rode, RSC
Adv., 2013, 3, 16499-16508.
The authors are grateful to Dr. Thomas Reinecke at Ruhr-
University Bochum for the XRD measurements and analyses.
[42] B. Peng, C. Zhao, I. Mejía-Centeno, G. A. Fuentes, A. Jentys, J. A.
Lercher, Catal. Today 2012, 183, 3–9.
[43] E. S. Vasiliadou, A. A. Lemonidou, Appl. Catal., A, 2011, 396, 177-185.
Keywords: Hydrodeoxygenation of glycerol • 1,2-propanediol •
Cu/ZrO2 • metallic copper
This article is protected by copyright. All rights reserved.

10.1002/cctc.201701748
ChemCatChem
FULL PAPER
FULL PAPER
Cu - Commonly underrated:
Cu/ZrO2 catalysts exhibit high
selectivity for the
Thomas Gabrysch, Baoxiang Peng*,
Sorin Bunea, Gerald Dyker, Martin
Muhler*
hydrodeoxygenation of glycerol to
1,2-propanediol under mild reaction
conditions. Metallic Cu is identified
as the active site for the sequential
dehydration - hydrogenation reaction
steps.
Page No. – Page No.
The role of metallic copper in the
selective hydrodeoxygenation of
glycerol to 1,2-propanediol over
Cu/ZrO2
This article is protected by copyright. All rights reserved.