Promoter effect of Pd in CuCr2O4 catalysts on the hydrogenolysis of glycerol to 1,2-propanediol

Article abstract of DOI:10.1039/c2gc00009a

CuCr₂O₄ catalysts containing various amounts of Pd (Pd-CuCr) were prepared by a co-precipitation method and examined for use in the conversion of glycerol to 1,2-propanediol (1,2-PDO). Pd was observed to be highly dispersed in a CuCr₂O₄ spinel structure and conferred unique reduction characteristics of a CuCr₂O₄ catalyst. After reduction, the amounts of surface exposed Cu⁰ species and occluded hydrogen species were much larger in the case of added Pd, the Pd-CuCr catalyst, compared to a pure copper chromite catalyst. The Pd-CuCr catalyst utilized hydrogen very efficiently, resulting in an enhancement in the catalytic activity for the conversion of glycerol to 1,2-PDO, even at a relatively low hydrogen pressure. The Pd_0.5-CuCr catalyst (containing 0.5 wt% of Pd) showed a total yield of 93.9%, with a selectivity approaching 100% for 1,2-PDO at a hydrogen pressure of 4 MPa. In a kinetic study, the effects of H₂ pressure and the concentration of glycerol on the initial rate were examined. Based on the results, the palladium promoter (Pd-CuCr) enhanced the rate constant by about 1.7 times compared with copper chromite. The results provide a basis for the production of 1,2-PDO from glycerol using a process which is both environmentally benign and has improved economics.

Full text of DOI:10.1039/c2gc00009a

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2012, 14, 2638
www.rsc.org/greenchem
PAPER
Promoter effect of Pd in CuCr O catalysts on the hydrogenolysis of glycerol
2
4
to 1,2-propanediol†
Nam Dong Kim,‡ Jae Ryul Park,‡ Dae Sung Park, Byoung Kyu Kwak and Jongheop Yi*
Received 4th January 2012, Accepted 20th July 2012
DOI: 10.1039/c2gc00009a
CuCr O catalysts containing various amounts of Pd (Pd–CuCr) were prepared by a co-precipitation
2
4
method and examined for use in the conversion of glycerol to 1,2-propanediol (1,2-PDO). Pd was
observed to be highly dispersed in a CuCr O spinel structure and conferred unique reduction
2
4
0
characteristics of a CuCr O catalyst. After reduction, the amounts of surface exposed Cu species and
2
4
occluded hydrogen species were much larger in the case of added Pd, the Pd–CuCr catalyst, compared to
a pure copper chromite catalyst. The Pd–CuCr catalyst utilized hydrogen very efficiently, resulting in an
enhancement in the catalytic activity for the conversion of glycerol to 1,2-PDO, even at a relatively low
hydrogen pressure. The Pd0.5–CuCr catalyst (containing 0.5 wt% of Pd) showed a total yield of 93.9%,
with a selectivity approaching 100% for 1,2-PDO at a hydrogen pressure of 4 MPa. In a kinetic study, the
effects of H pressure and the concentration of glycerol on the initial rate were examined. Based on the
2
results, the palladium promoter (Pd–CuCr) enhanced the rate constant by about 1.7 times compared with
copper chromite. The results provide a basis for the production of 1,2-PDO from glycerol using a process
which is both environmentally benign and has improved economics.
1
. Introduction
1,2-PDO from glycerol by a hydrogenolysis reaction. Among the
various catalysts evaluated, Cu-based catalysts, such as
6–8
9
10
Glycerol is a renewable chemical which is usually produced by
steam splitting or the saponification of animal fats or vegetable
oils. It is also produced as a major by-product from the bio-
diesel production process. As the production of biodiesel has
increased, the amount of produced glycerol has also increased.
Therefore, the catalytic conversion of glycerol into value-added
products provides an alternative use for glycerol and for solving
the problem of the surplus supply of this compound. 1,2-Propa-
nediol (1,2-PDO) is a value-added chemical that is widely used
in manufacturing polyester fibers, antifreeze, pharmaceuticals
and other important products. It is currently produced from pro-
pylene via the selective oxidation of propylene to propylene
oxide and subsequent hydrolysis. The process is limited by the
supply of propylene derived from crude oil. Therefore, the devel-
opment of an alternative method for producing 1,2-PDO from
glycerol that is environmentally benign and with improved
process economics would be highly desirable.
Cu–ZnO,
Cu–SiO , or Cu–Al O , generally have a high
2 2 3
selectivity for 1,2-PDO, due to their intrinsic ability to selec-
tively cleave C–O bonds in preference to the C–C bonds in gly-
cerol. Although Cu-based catalysts show high catalytic activities
for the hydrogenolysis of glycerol to 1,2-PDO, the use of high
hydrogen pressures (8 MPa) limits their further application in
1
2
6
,10,11
industry.
.2 MPa, the conversion to glycerol was greatly reduced
(
In a study using a lower hydrogen pressure of
3
4
8
22.5%). A catalytic process for converting glycerol to 1,2-PDO
in the absence of hydrogen has also recently been reported.
Although the use of hydrogen was thereby reduced, some
organic solvents were used as a source of hydrogen. Catalytic
performance still needs to be improved, if a viable industrial
4
5
,6
11,12
application is to be achieved.
Therefore, a highly active cata-
lyst that can be used at lower hydrogen pressures needs to be
developed for industrial use.
The copper chromite (CuCr O ) catalyst is known to have
2
4
A number of attempts have been reported concerning the
development of more efficient catalysts for the production of
high catalytic activity for the hydrogenolysis of glycerol to 1,2-
PDO and is considered to be a promising candidate for commer-
2,13
cialization.
In our previous study, we reported that it had
some peculiar reduction characteristics, such as the occurrence
of structural changes during the reduction process from a tetra-
gonal spinel structure to a cubic spinel structure and the presence
of occluded hydrogen species. However, the catalyst exhibited
an excellent utilization of hydrogen during the reaction and
such characteristics are mainly responsible for the high catalytic
activities. Although it showed high performance with a selectivity
World Class University (WCU) program of Chemical Convergence for
Energy & Environment (C2E2), School of Chemical and Biological
Engineering, Institute of Chemical Process, Seoul National University
14
(
SNU), Seoul, 151-744, Republic of Korea. E-mail: jyi@snu.ac.kr;
Tel: +82-2-880-7438
†
1
‡
Electronic supplementary information (ESI) available. See DOI:
0.1039/c2gc00009a
Equal contribution.
2638 | Green Chem., 2012, 14, 2638–2646
This journal is © The Royal Society of Chemistry 2012

View Online
of 83.9% and an 80.3% conversion, the hydrogen pressure
required was still high (8 MPa), a situation that would involve an
Temperature programmed desorption (TPD) analyses were
carried out (Micromeritics AutoChemII instrument). Measure-
1
4
additional cost in an industrial process.
In this research, Pd-deposited CuCr O catalysts were pre-
ments of acidic properties were performed by NH -TPD. A 0.1 g
3
sample was reduced with a mixed stream of 10% H in a N2
2
4
2
pared for use in the hydrogenolysis of glycerol to 1,2-PDO
under low hydrogen pressure. A series of catalysts containing
highly dispersed Pd were prepared and their catalytic perform-
ance compared. Pd is widely used in many fields where hydro-
gen is involved, such as in catalysts and in hydrogen storage
balance at 320 °C for 2 h and then cooled to room temperature.
A mixed stream of 10% NH in He was introduced at room
3
temperature for the adsorption of NH and subsequently purged
3
with He at 100 °C for 1 h to remove the physisorbed NH . The
3
TPD measurements were conducted in flowing He from room
15–20
−1
devices.
Due to the peculiar properties associated with the
temperature to 600 °C with a heating rate of 10 °C min .
utilization of hydrogen by Pd, the reducibility of the CuCr O4
Occluded hydrogen species were measured by H -TPD.
2
2
catalyst was significantly increased and more active sites were
produced on the surface of the catalyst. Moreover, it was found
that the amount and chemical properties of the occluded hydro-
gen which can actively participate in hydrogenolysis reactions
were increased and improved. Consequently, a CuCr O catalyst
A given amount of the sample was also reduced by the same
method as for NH -TPD and then cooled to room temperature.
3
The temperature of the reduced catalysts was cooled to room
temperature and then purged with a N stream at the same tem-
2
perature under reduced pressure to remove the physisorbed H2.
2
4
that contained Pd showed much enhanced catalytic activities for
the hydrogenolysis of glycerol to 1,2-PDO under mild reaction
conditions. A kinetic study was also conducted to collect more
The TPD measurements were conducted in flowing N2 from
room temperature to 600 °C at a heating rate of 10 °C min .
−1
precise data on the catalytic process. The effect of H pressure
2
2
.3 Catalytic activity measurement
and the concentration of glycerol on the hydrogenolysis of gly-
cerol were examined and an appropriate power law equation was
developed considering two factors, namely, the pressure of H2
and the concentration of glycerol. From this equation, we were
able to confirm that Pd improved the catalytic activity of copper
chromite.
The hydrogenolysis of glycerol to 1,2-PDO was carried out in a
hastelloy-C autoclave equipped with a magnetic stirrer and
heater. The autoclave was filled with glycerol (50 g, SIGMA-
Aldrich, Tech., 90%) and catalyst (1 g) added. The autoclave
was pressurized with hydrogen and heated at a ramping rate of
5
°C min− . The reaction was conducted at 220 °C, for 12 h at
1
2
. Experimental
various H
employed in this study were reduced by treatment with a mixed
2
pressures. Prior to the reaction, all of the catalysts
2
.1 Catalyst preparation
−1
−1
stream of H (2 ml min ) and N (20 ml min ) at 320 °C for
2
2
2
h. After the reduction process, all catalysts underwent a passi-
CuCr O catalysts containing various amounts of Pd were pre-
2
4
vation treatment with the N2 flow containing a very small
amount of O (<1% O in N ) for 1 h. Reduced and passivated
catalysts were introduced to the reaction test.
The reaction products were analyzed by gas chromatography,
using a DS 6200 (DONAM INSTRUMENT INC.) equipped
with a flame ionization detector. An SGE BP20 (WAX) capillary
column (25 m × 0.53 mm × 1 μm) was used for separation.
pared by a co-precipitation method. A calculated amount
of metal precursor (Cu(NO ) ·2.5H O (Riedel-de Haën),
2
2
2
3
2
2
Cr(NO ) ·9H O (Sigma-Aldrich), and PdCl (Sigma-Aldrich))
3
2
2
2
was dissolved in distilled water with vigorous stirring. Sodium
hydroxide solution (3 M NaOH) was then added until the pH of
the solution reached 12. The resulting precipitates were filtered
and dried overnight. The dried precipitates were ground and cal-
cined at 550 °C in air to obtain the final calcined CuCr O cata-
2
4
lysts containing various amounts of Pd. The prepared catalysts
2
.4 Kinetic study
were designated as follows (Pd –CuCr, x denotes the weight
x
ratio of Pd in the CuCr O catalyst).
The kinetic study of the hydrogenolysis of glycerol was con-
ducted in a batch reactor. The stirring speed of 350 rpm at a
temperature of 220 °C was used in all experiments. In all cases,
2
4
2
.2 Catalyst characterization
6
0 ml of an aqueous solution and 1 g of reduced catalysts were
used. In order to determine the effect of H pressure, the exper-
iments were carried out at 3, 5 and 7 MPa H . The experiments
The X-ray diffraction (XRD) patterns of the powdered catalysts
were recorded using an X-ray diffractometer (D-MAX2500-PC,
Rigaku Corp.) equipped with CuKα radiation. To examine
the reducibility of the synthesized catalysts, temperature-
programmed reduction (TPR) measurements were performed
using a conventional flow system with a moisture trap connected
to a thermal conductivity detector (TCD). The flow rate of the
2
2
were also conducted using various concentrations of glycerol,
0.5, 0.75 and 1 M, to identify the effect of the glycerol concen-
tration on the reaction rate.
−1
3. Results and discussion
mixed gas (10% H in N ) was fixed at 20 ml min and the
2
2
−
1
heating rate was 5 K min . X-ray photoelectron spectroscopy
XPS) spectra were recorded on a KRATOS AXIS electron
3
.1 Characterization of the catalysts
(
spectrometer equipped with MgKα radiation for exciting photo-
electrons. All binding energies (BEs) are referred to the adventi-
tious C 1s line at 284.6 eV.
The crystalline structures of the prepared Pd –CuCr catalysts
were investigated by XRD measurements. XRD patterns of cal-
x
cined Pd –CuCr catalysts are shown in Fig. 1. The Pd –CuCr
x
0
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2638–2646 | 2639

View Online
x
Fig. 1 XRD patterns for calcined Pd –CuCr catalysts (● calcined
x
Fig. 3 TPR profiles of calcined Pd –CuCr catalysts.
CuCr , ■ Cr , ▲ PdCuO ).
2
O
4
2
O
3
2
ca. 29.6° corresponding to the (220) plane and 53.4° to the
422) plane are no longer present. This is due to the fact that the
(
0
metallic Cu phase and the cubic spinel phase are now both
present after the reduction process compared to the initial sample
23
that has a tetragonal spinel phase structure. Diffraction patterns
for the reduced Pd –CuCr catalysts did not show specific charac-
x
teristic peaks for metallic Pd, which also show high dispersion
of Pd. The remaining characteristic peaks for Cr O after the
2
3
reduction process indicate that it is hardly reduced in the
14
reduction conditions used in this study.
It has been reported that the CuCr O spinel structure
2
3,24
4
2
has unusual reduction characteristics.
Briefly, during the
reduction process of the CuCr O spinel structure, some of the
2
4
Cu species, in the tetrahedral position of the spinel structure, are
partially reduced and are released on the surface of the catalyst
0
in the form of Cu , which generally has a high catalytic activity
Fig. 2 XRD patterns for reduced Pd
CuCr , ■ Cr ).
x
–CuCr catalysts (● reduced
for hydrogenolysis of glycerol to 1,2-PDO. Along with this
process, atomic hydrogen species are occluded into the vacant
tetrahedral interstices of the tetragonal spinel structure. The
overall structure of the catalyst is simultaneously transformed
from a tetragonal spinel structure to a cubic spinel structure,
which was previously confirmed by XRD analysis. Such
reduction characteristics of CuCr O are mainly responsible for
2
O
4
2 3
O
catalyst, which contains no Pd, shows sharp characteristic peaks
at 18.6°, 29.6°, 31.1° and 37.7°, which correspond to the (111),
(
220), (022), and (113) planes of the tetragonal spinel structure
2
4
2
1
14
of CuCr O , respectively. When a small amount of Pd was
its high catalytic activity in hydrogenolysis reactions. There-
fore, it is assumed that, if Pd plays an important role during the
reduction process, it can possibly promote the catalytic activity
of a CuCr O catalyst.
2
4
added to the CuCr O catalyst, it was not possible to distinguish
2
4
it by XRD analysis, indicating that the Pd is highly dispersed in
the CuCr O structure. However, when 3 wt% Pd is introduced
2
4
2 4
to CuCr O catalysts (Pd –CuCr), characteristic peaks other than
TPR measurements were carried out to investigate the reduci-
2
4
3
CuCr O4 appear, which indicates the emergence of a Cr O3
bility of the Pd –CuCr catalysts (Fig. 3). The TPR pattern of the
2
2
x
structure (PCPDF 34-0412). Pd can form an alloy structure
CuPdO , PCPDF 44-0185) with Cu at milder conditions than
Pd –CuCr catalyst shows a large sharp peak at ca. 150 °C,
0
(
which originates from the partial reduction of the tetragonal
spinel structure. As Pd is added to the CuCr O catalyst, a small
2
2
2
Cr. Therefore, as Pd is introduced to the CuCr O structure,
2
4
2 4
the excess Cr forms a Cr O structure instead of forming a
peak appears below 100 °C, corresponding to the reduction of
Pd. The intensity of this peak increases with the amount of Pd in
2
3
CuCr O spinel structure. The proportion of Cr O increases as
2
4
2 3
the amount of Pd increases up to 10 wt% (Pd –CuCr).
the catalysts. Moreover, after the addition of Pd, the CuCr O4
1
0
2
Before the reaction test, all of the catalysts were reduced in a
catalyst starts to be reduced at lower temperatures compared to
stream of H . The structures of the reduced catalysts were also
the pure CuCr O4 catalyst, which occurs almost immediately
2
2
characterized by XRD (Fig. 2). After the reduction process, the
after the reduction of Pd. This suggests that the reducibility of
Pd –CuCr catalyst showed a significantly enhanced characteristic
the CuCr O4 catalyst is enhanced by the addition of small
0
2
peak for the (111) plane at ca. 18.6°. In addition, the peaks at
amounts of Pd.
2640 | Green Chem., 2012, 14, 2638–2646
This journal is © The Royal Society of Chemistry 2012

View Online
0
Fig. 4 X-ray photoelectron spectra (Cu 2p region) of (a) Pd –CuCr and (b) Pd0.5–CuCr, before and after the reduction process.
Table 1 Deconvolution results of the Cu 2p region for Pd
Pd0.5–CuCr catalysts
0
–CuCr and
a
Pd
0
–CuCr
Pd0.5–CuCr
Calcined
Reduced
Calcined
Reduced
2
0
+
Cu
Cu
0.73
0.27
0.65
0.35
0.77
0.23
0.46
0.54
a
2+
0
The peak area percentage (%) of Cu and Cu in the XPS Cu 2p
region.
It has been reported that coordinately unsaturated copper
0
+
species (Cu or Cu ) on the catalyst surface are the active sites
for the hydrogenolysis of glycerol. XPS spectra for the Cu 2p
region in both Pd –CuCr and Pd0.5–CuCr are shown in Fig. 4.
0
3 x
Fig. 5 NH -TPD profiles of reduced Pd –CuCr catalysts.
To compare the effect of the reduction process on the Cu state in
both catalysts, XPS spectra were collected, before and after the
reduction process. Deconvolution results show that there are two
components in XPS Cu 2p spectra. Their binding energies were
centered at 934.0 ± 0.2 eVand 932.4 ± 0.2 eV, which correspond
(Fig. S3†). The findings indicate that the total amount of de-
sorbed NH gradually decreases as the amount of Pd in the cata-
3
lysts increases (Table 2). Moreover, the temperature of the peak
maximum at ca. 250 °C decreases with the added Pd. This
implies that, along with the addition of Pd, the total amount and
2
+
0
to oxidized Cu species (Cu ) and metallic Cu species (Cu ),
respectively. Before the reduction process, relative fractions of
each copper state are similar in both catalysts, indicating that the
crystalline structure of both catalysts is mainly a CuCr O spinel
strength of acid sites of Pd
the amount of the weakest acid sites, which shows the lowest
desorption temperature of NH (ca. 145 °C), was significantly
increased in the Pd0.5–CuCr and Pd –CuCr catalysts. It has been
–CuCr catalysts decrease. However,
x
2
4
structure (Table 1). After the reduction process, large amounts of
metallic Cu species are present in both catalysts, along with
structural changes from a tetragonal spinel structure to a cubic
spinel structure. However, the proportion of reduced metallic Cu
3
1
reported that catalysts with a large amount of Brønsted acid sites,
which have milder acidic characteristics than Lewis acids, are
more active and selective for the dehydration reaction of gly-
species is much larger in Pd0.5–CuCr than Pd –CuCr. This
0
27,28
means that the degree of reduction in Pd0.5–CuCr was much
cerol.
Consequently, undesirable dehydration reactions and
larger than that of Pd –CuCr, due to the promotional effect
of Pd.
problems associated with coke deposition are suppressed in the
0
29
case of Brønsted acid sites compared to Lewis acid sites. Thus,
although the total amount and the strength of the acid sites
decrease after a small amount of Pd is added, a significant
increase in Brønsted acid characteristics would be beneficial for
the selective dehydration of glycerol.
The hydrogenolysis of glycerol to 1,2-PDO consists of two
individual reactions, one is the dehydration of glycerol to acetol
and the other is the hydrogenation of acetol to produce propylene
glycol. In the first reaction, the dehydration of glycerol to acetol,
the acidic properties of the catalysts have a major effect on cata-
It has been reported that occluded hydrogen species in the
reduced state of CuCr O play an important role in hydrogen-
2 4
25,26
lytic activity.
Therefore, the strength and nature of the acid
sites on the synthesized catalysts were evaluated by NH -TPD
analysis (Fig. 5) and in situ FT-IR measurement with NH3
ation or hydrogenolysis reactions. To investigate the amount
and the characteristics of occluded hydrogen species in the
3
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2638–2646 | 2641

View Online
Table 2 NH
3
-TPD results for the reduced Pd
x
–CuCr catalysts
3
NH )
desorbed region (mmol g−
1
desorbed (mmol g−
1
)
ca. 145 °C
ca. 200 °C
ca. 250 °C
ca. 320 °C
Catalyst
Total NH
3
Pd
0
–CuCr
0.8718
0.7329
0.6771
0.5946
0.5452
0.0683
0.0565
0.0988
0.0795
0.0558
0.2233
0.1291
0.1295
0.1033
0.1210
0.5467
0.5106
0.4229
0.3819
0.3333
0.0334
0.0367
0.0259
0.0298
0.0354
Pd0.3–CuCr
Pd0.5–CuCr
Pd
Pd
1
–CuCr
–CuCr
3
Fig. 7 Schematic illustration for the effects of Pd on the reduced
CuCr catalyst in hydrogenolysis of glycerol to 1,2-PDO.
2 4
O
Fig. 6
H
2
-TPD profiles of reduced Pd
x
–CuCr catalysts.
Pd in the CuCr O catalyst can be explained by its hydrogen
spillover behaviour. When hydrogen is adsorbed onto the
2
4
Table 3 The amount of desorbed H
experiment
2
measured by the H
2
-TPD
17
metal surface of the catalyst, it spontaneously dissociates into
(mmol g−
1
)
1,17
Catalyst
Pd –CuCr
Pd0.3–CuCr
Pd0.5–CuCr
Amount of desorbed H
2
hydrogen adatoms.
Some of the hydrogen adatoms remain
attached on the metal, while others diffuse to the adjacent
support. It is known that spillover hydrogen adatoms exhibit
higher diffusivity and activity, compared to molecular hydrogen.
It appears that Pd can provide hydrogen to the catalytic active
site via spillover. In addition, it is known that Pd easily forms an
interstitial hydride (β-PdH0.67) which makes it possible to store
0
28.29
34.97
37.76
32.41
30.19
20.41
Pd
Pd
1
–CuCr
–CuCr
3
Pd10–CuCr
20
hydrogen internally. This property would also take advantage
of the hydrogen spillover. Due to such hydrogen spillover
synthesized catalysts, a H -TPD analysis was performed (Fig. 6).
2
The evolution of hydrogen was measured by means of a TCD
detector, and the direction of the characteristic signal was down-
ward under the baseline which is contrary to the results of TPR
experiments. All of the catalysts show a large amount of hydro-
gen evolution during the temperature scanning (Table 3), which
verifies that occluded hydrogen species are present in the
phenomena, the activation energy for the reduction of CuCr O4
2
catalyst is lowered and the overall reducibility of the catalysts is
increased. As the degree of reduction is increased by the pro-
0
motion effect of Pd, the fraction of surface Cu species is also
increased, which means that more active sites are formed. More-
over, Bronsted acid properties significantly increased which
allowed a more active and selective dehydration reaction of
glycerol. It is also expected that Pd would refresh the active site
during the reaction, as long as it keeps providing hydrogen via
spillover. Moreover, it is likely that Pd functions as a route for
the spillover of hydrogen toward the reduced CuCr O , provid-
reduced CuCr O4 catalyst. However, the amount of occluded
2
hydrogen is much larger in Pd –CuCr, where x > 0, catalysts. In
x
addition, the desorption of occluded hydrogen started at a lower
temperature in these catalysts. This indicates that a large amount
of hydrogen can thereby participate in the hydrogenolysis reac-
2
4
tion due to the addition of a small amount of Pd to the CuCr O4
ing occluded hydrogen species which can actively participate in
2
catalyst.
the catalytic reaction. Since Pd plays a role as a ‘center for
Considering all the analysis shown above, it can be concluded
that the addition of a small amount of Pd facilitates the activity
of the CuCr O catalyst with respect to reduction. The effect of
hydrogen utilization’ in the CuCr O4 as described in Fig. 7,
2
the catalytic activity of Pd –CuCr would be highly enhanced
x
2
4
compared to the pure CuCr O .
2 4
2642 | Green Chem., 2012, 14, 2638–2646
This journal is © The Royal Society of Chemistry 2012

View Online
Fig. 9 Results of time on stream in the reaction test over Pd0.5–CuCr
catalysts at 220 °C, 4 MPa of H for 18 h.
2
selectivity at least over ca. 88% even in the Pd –CuCr catalyst.
10
After the reaction test, trace amounts of side products were
detected in the case of Pd0.5–CuCr, where relatively large
amounts of 1-propanol and ethylene glycol were detected in the
case of the CuCr O catalyst. The reason for such high selectiv-
2
4
ity could be found in the nature of less deactivated Pd –CuCr
x
catalysts. As previously discussed, the amount of both partially
0
reduced Cu species and occluded hydrogen could be a major
parameter for determining the activity of the reduced CuCr O4
2
catalyst in the hydrogenolysis of glycerol to 1,2-PDO. In
Pd –CuCr catalysts, Pd can act as a route for providing active
x
hydrogen atoms to access reduced CuCr O . Thus, additional
2
4
hydrogen would, in principle, be available to react with glycerol.
Consequently, the conversion of glycerol to 1,2-PDO could be
accelerated, whereas possible side reactions producing undesir-
able by-products would be suppressed.
Fig. 8 Catalytic performance of Pd
of glycerol to 1,2-PDO, performed at 220 °C, (a) 6 MPa and (b) 4 MPa
in H for 12 h.
x
–CuCr catalysts in hydrogenolysis
2
To further examine the effect of Pd, the H pressure of the
2
hydrogenolysis reaction was reduced to 4 MPa (Fig. 8b). All of
the catalysts showed a slightly reduced yield of 1,2-PDO com-
pared with the results shown in Fig. 5a, which is due mainly to
3
.2 Catalytic reaction
The catalytic performance of the prepared catalysts for the
hydrogenolysis of glycerol to 1,2-PDO at 220 °C in 6 MPa of
H is shown in Fig. 8a. In our previous results, when a CuCr O
2 4
the reduced conversion. The catalytic activity of the Pd –CuCr
catalyst decreased significantly, showing ca. 62.9% conversion,
77.3% selectivity and 48.6% total yield, respectively. On the
0
2
catalyst was used in this reaction at the same temperature but at a
other hand, it is interesting to note that the Pd –CuCr catalysts
x
higher H pressure (8 MPa), a high catalytic activity was found,
maintain their high selectivity for 1,2-PDO production, even
2
showing a conversion of ca. 80.3%, a selectivity of 83.9% and a
under conditions of much reduced H pressure (ca. 85.5% in
2
14
total yield of 67.4%. Although its catalytic activity decreased
slightly as the hydrogen pressure was decreased to 6 MPa, it
exhibited a relatively high catalytic activity showing ca. 78.3%
of conversion, 75.9% of selectivity and giving a total yield of
Pd –CuCr and ca. 93.6% in Pd –CuCr). This result further
10
1
demonstrates the conclusion that Pd plays an important role in
utilizing hydrogen during the hydrogenolysis reaction.
It is also important to inhibit the further conversion of 1,2-
PDO to other chemicals, such as 1-propanol and 2-propanol,
for enhancing the selectivity for 1,2-PDO during the glycerol
hydrogenolysis reaction. During the stream analysis using the
Pd0.5–CuCr catalyst with 4 MPa of hydrogen, the total yield of
1,2-PDO increased rapidly up to 12 h and was then maintained
at this level without any evidence of further degradation. Conse-
quently, the Pd0.5–CuCr catalyst results in a total yield of 93.9%
and a selectivity of nearly 100% after 18 h of reaction time
(Fig. 9). This means that the Pd0.5–CuCr catalyst shows stable
5
9.4%, respectively. As Pd was introduced to the CuCr O cata-
2 4
lysts, catalytic activity increased significantly. In the case of
Pd0.5–CuCr, 83.4% of conversion, 92.9% of selectivity and
7
1
7.4% of yield were attained. The addition of Pd in excess of
wt% to the CuCr O catalyst was detrimental to overall cataly-
2
4
tic performance. This is because the relative fraction of the major
active component, CuCr O spinel structure, decreased, whereas
2
4
the Cr O structure, which is inactive in this reaction, is formed
2
3
with increasing Pd loading. It should be noted that the enhanced
catalytic activity can be largely attributed to an increased
performance even at reduced H pressure and that this catalyst
2
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2638–2646 | 2643

View Online
2
Fig. 10 Effect of H pressure on initial rate of hydrogenolysis of
glycerol (temperature: 220 °C, catalyst: 1 g and 1 M of glycerol).
Fig. 11 Effect of the concentration of glycerol on the initial rate of
hydrogenolysis of glycerol (temperature: 220 °C, catalyst: 1 g and
5
2
MPa of H ).
has some potential for use in the industrial production of
1
,2-PDO from glycerol.
of initial rate (r ) with H pressure was found. The concentration
0 2
of hydrogen in water is directly proportional to the pressure of
hydrogen and a large amount of dissolved hydrogen in water
was involved in the reaction at high hydrogen pressures.
3
.3 Kinetic study
The effect of glycerol concentration was studied using 0.5,
.75 and 1 M glycerol solutions, a temperature of 220 °C and a
Numerous literature references dealing with the kinetics of gly-
cerol hydrogenolysis have appeared. However, most of them
assume the LHHW (Langmuir–Hinshelwood–Hougen–Watson)
model and the kinetic parameters were found to be mainly gov-
erned by the adsorption/desorption properties of glycerol and
0
fixed H pressure (5 MPa). Fig. 11 shows that the initial reaction
2
rate also increased with increasing concentration of glycerol. A
linear relation was found between concentration of glycerol and
initial rate. A line with a slope of 2.28 was obtained, indicating a
.28-order dependence of hydrogenolysis of glycerol on the con-
centration of glycerol.
30–32
1
,2-PDO.
Thus, the effect of hydrogen pressure on kinetic
2
parameters was very difficult to quantify, because the conversion
of glycerol is highly dependent on the initial glycerol concen-
tration. In this study, we attempted to quantitatively analyze the
kinetic parameters, in order to compare the utilization of hydro-
Generally, the adsorption equilibrium constant at a certain
condition is needed to represent the precise rate equation. Instead
of the rate constant, the apparent rate constant was measured in
many previous studies due to the lack of information concerning
gen between Pd –CuCr and Pd0.5–CuCr which showed the
0
optimal catalytic activity. Thus, the effects of glycerol concen-
35–37
adsorption.
The adsorption equilibrium constant during the
tration, H pressure and type of catalyst on the rate of 1,2-PDO
2
hydrogenolysis of glycerol has not been studied and it was
difficult to calculate the adsorption equilibrium constant exper-
formation were estimated using the initial rate method.
The hydrogenolysis of glycerol is a combination of two
sequential reactions, namely dehydration and hydrogenation. It is
known from a previous study that the dehydration of glycerol to
imentally due to high temperature and H pressure of the system.
2
The relation between H pressure and the concentration of dis-
2
3
3
solved hydrogen in water was reported in a previous study and
the concentration of hydrogen increased linearly with respect to
acetol is the rate determining step in the reaction. Acetol was
rarely detected by GC analysis in our experiments and it was
assumed that 1,2-PDO is formed as soon as glycerol is converted
into acetol. In this reaction, the rate of formation of acetol was
assumed to be identical to that for 1,2-PDO. The initial rate of
formation of 1,2-PDO was calculated from the slope of the line
between the starting point and a certain point under a 5% yield
38
H2 pressure. Therefore, the initial rate equation can be
described as
2:28
1:09
r ¼ k ꢀ CGlycerol
ꢀ CHydrogen
where r is initial rate of formation of 1,2-PDO, k is an apparent
rate constant, CGlycerol is initial concentration of glycerol and
CHydrogen is concentration of hydrogen in water.
3
4
to minimize error arising from rate approximation.
The effect of hydrogen pressure on the hydrogenolysis of
glycerol was studied by varying the pressure (3, 5 and 7 MPa),
while maintaining a constant concentration of glycerol in D.I.
water (1 M), the amount of catalyst (copper chromite, 1 g) and
the reaction temperature (220 °C). The yield of 1,2-PDO for
various hydrogen pressures with time is shown in Fig. S1.† The
rate of glycerol hydrogenolysis increased with increasing H2
pressure. Fig. 10 shows a plot of the logarithm of initial
rate against the concentration of glycerol. A linear dependence
The hydrogenolysis of glycerol reactions was carried out
under copper chromite and Pd0.5–CuCr to calculate the initial
rate of the reaction respectively. In Fig. 12, the yield of 1,2-PDO
in the reaction with Pd0.5–CuCr reached over 15% in a few
hours. Pd0.5–CuCr showed higher catalytic activity compared to
copper chromite as determined in a long term test for 12 h.
It might be that two catalysts were well-fitted to the described
initial rate equation, because Pd showed feeble catalytic activity
2644 | Green Chem., 2012, 14, 2638–2646
This journal is © The Royal Society of Chemistry 2012

View Online
constant of Pd0.5–CuCr was 1.7 times larger than that of the pure
material. Considering the ease of the synthetic procedure used
and the high catalytic performance, the Pd –CuCr catalyst can be
x
a potential candidate to be used in an economical and environ-
mentally friendly process for producing 1,2-PDO from glycerol.
Acknowledgements
This research was supported by the WCU (World Class Univer-
sity) program through the National Research Foundation
of Korea funded by the Ministry of Education, Science and
Technology (R31-10013). This work was also partially sup-
ported by the grant from the Industrial Source Technology
Development Programs (10033352-2010-12) of the Ministry of
Knowledge Economy (MKE) of Korea.
Fig. 12 Reaction result of hydrogenolysis of glycerol under copper
chromite and Pd0.5–CuCr (temperature: 220 °C, catalyst: 1 g, 1 M of
glycerol and 5 MPa of H ).
2
References
1
2
3
4
J. Zhou, L. Guo, X. Guo, J. Mao and S. Zhang, Green Chem., 2010, 12,
835–1843.
M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi and C. Della Pina,
Angew. Chem., Int. Ed., 2007, 46, 4434–4440.
A. Behr, J. Eilting, K. Irawadi, J. Leschinski and F. Lindner, Green
Chem., 2008, 10, 13–30.
M. Balaraju, V. Rekha, P. S. S. Prasad, B. L. A. P. Devi, R. B. N. Prasad
and N. Lingaiah, Appl. Catal., A, 2009, 354, 82–87.
1
Table 4 The initial rate and apparent rate constant of copper chromite
and Pd0.5–CuCr
Copper chromite
Pd0.5–CuCr
5.7 × 10−4
Rate (mol l min−1
−
1
3.4 × 10−4
)
−
1
−2.37
−1
−3
−3
5 J. Haveren, E. L. Scott and J. Sanders, Biofuel. Bioprod. Bior., 2008, 2,
1–57.
K (l mol
min
)
5.56 × 10
9.21 × 10
4
Solubility of hydrogen is 0.077 mol l−1 in water at 220 °C and 5 MPa H
6 J. Chaminand, L. a. Djakovitch, P. Gallezot, P. Marion, C. Pinel and
C. Rosier, Green Chem., 2004, 6, 359–361.
2
.
7
M. Balaraju, V. Rekha, P. Sai Prasad, R. Prasad and N. Lingaiah, Catal.
Lett., 2008, 126, 119–124.
8
S. Wang and H. C. Liu, Catal. Lett., 2007, 117, 62–67.
39
for the hydrogenolysis of glycerol in a previous study. The
apparent rate constant for the two catalysts calculated by initial
rate and the power law equation is shown in Table 4. As a result,
the apparent rate constant of Pd0.5–CuCr was 1.7 times larger
than that of copper chromite.
9 Z. Huang, F. Cui, H. Kang, J. Chen, X. Zhang and C. Xia, Chem. Mater.,
008, 20, 5090–5099.
2
1
0 R. Mane, A. Hengne, A. Ghalwadkar, S. Vijayanand, P. Mohite,
H. Potdar and C. Rode, Catal. Lett., 2010, 135, 141–147.
11 A. Y. Yin, X. Y. Guo, W. L. Dai and K. N. Fan, Green Chem., 2009, 11,
514–1516.
1
1
1
1
1
1
2 M. G. Musolino, L. A. Scarpino, F. Mauriello and R. Pietropaolo, Green
Chem., 2009, 11, 1511–1513.
3 M. A. Dasari, P. P. Kiatsimkul, W. R. Sutterlin and G. J. Suppes, Appl.
Catal., A, 2005, 281, 225–231.
4. Conclusions
4 N. D. Kim, S. Oh, J. B. Joo, K. S. Jung and J. Yi, Top. Catal., 2010, 53,
517–522.
5 C. I. Contescu, K. van Benthem, S. Li, C. S. Bonifacio, S. J. Pennycook,
P. Jena and N. C. Gallego, Carbon, 2011, 49, 4050–4058.
6 J. H. Jang, J. Kim, Y. H. Lee, C. Pak and Y. U. Kwon, Electrochim. Acta,
In this study, CuCr O catalysts promoted by small amounts of
2
4
Pd were proposed as a potential candidate for producing 1,2-
PDO from glycerol under conditions of a relatively low H2
pressure. It was found that Pd plays an important role in utilizing
hydrogen during the reduction process and catalytic reaction of
the CuCr O catalyst. Pd facilitated the reduction process of the
2
009, 55, 485–490.
17 C. Lin, T. Xu, J. Yu, Q. Ge and Z. Xiao, J. Phys. Chem. C, 2009, 113,
513–8517.
8
2
4
18 M. H. Rei, G. T. Yeh, Y. H. Tsai, Y. L. Kao and L. D. Shiau, J. Membr.
CuCr O catalyst producing more active sites and a large amount
2
4
Sci., 2011, 369, 299–307.
of occluded hydrogen species. Furthermore, Pd helps to provide
mild acidic properties and the catalytic active site remains intact
during the reaction preventing possible deactivation processes,
resulting in a considerably high selectivity toward 1,2-PDO (ca.
19 H. Wu, D. Wexler and H. Liu, Int. J. Hydrogen Energy, 2011, 36, 9032–
036.
9
2
2
2
2
0 R. Zacharia, K. Y. Kim, A. K. M. Fazle Kibria and K. S. Nahm, Chem.
Phys. Lett., 2005, 412, 369–375.
1 A. M. Kawamoto, L. C. Pardini and L. C. Rezende, Aerosp. Sci. Technol.,
2004, 8, 591–598.
2 M. Y. Wang, X. J. Liu, J. Meng and Z. J. Wu, J. Mol. Struct.
9
3%). Consequently, significantly enhanced catalytic perform-
ance can be obtained with a very small amount of Pd (0.5 wt%),
(THEOCHEM), 2007, 804, 47–55.
at a very low H pressure (4 MPa). Kinetic studies of the pro-
2
3 O. V. Makarova, T. M. Yureva, G. N. Kustova, A. V. Ziborov,
L. M. Plyasova, T. P. Minyukova, L. P. Davydova and V. I. Zaikovskii,
Kinet. Catal., 1993, 34, 608–612.
motion effect of Pd were also conducted. The effect of H2
pressure and the concentration of glycerol were confirmed
respectively under limited reaction conditions and the proper
power law equation for the initial rate was deduced. Based on
reaction results of Pd0.5–CuCr and pure copper chromite, the rate
2
2
4 O. V. Makarova, T. M. Yureva, L. M. Plyasova, T. A. Kriger and
V. I. Zaikovskii, Kinet. Catal., 1994, 35, 371–375.
5 R. B. Mane, A. A. Ghalwadkar, A. M. Hengne, Y. R. Suryawanshi and
C. V. Rode, Catal. Today, 2011, 164, 447–450.
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2638–2646 | 2645

View Online
2
2
2
2
6 M. Balaraju, V. Rekha, P. S. Sai Prasad, B. L. A. Prabhavathi Devi,
R. B. N. Prasad and N. Lingaiah, Appl. Catal., A-Gen, 2009, 354, 82–87.
7 S-H. Chai, H.-P. Wang, Y. Liang and B.-Q. Xu, Green Chem., 2007, 9,
33 J. T. Dam and U. Hanefeldl, ChemSusChem., 2011, 4, 1017–1034.
34 J. Casado, M. A. Lopez-Quintela and F. M. Lorenzo-Banal, J. Chem.
Educ., 1986, 63, 450–452.
35 P. Kluson, P. Kukula, E. Kyslingerova and L. Cerveny, React. Kinet.
Catal. Lett., 1996, 59, 9–13.
36 H. Rojas, G. Borda, M. Brijaldo, P. Reyes and J. Valencia, React. Kinet.
Mech. Catal., 2012, 105, 341–284.
37 M. Han, H. Zhang, Y. Du, P. Yang and Z. Deng, React. Kinet. Mech
Catal.., 2011, 102, 393–404.
38 H. A. Pray, C. E. Schweichert and B. H. Minnich, Ind. Eng. Chem.,
1952, 44, 1146–1151.
1130–1136.
8 R. Weingarten, G. A. Tompsett, W. C. Conner Jr and G. W. Huber,
J. Catal., 2011, 279, 174–182.
9 P. Lauriol-Garbey, J. M. M. Millet, S. Loridant, V. Belliere-Baca and
P. Rey, J. Catal., 2011, 281, 362–370.
3
0 D. G. Lahr and B. H. Shanks, J. Catal., 2005, 232, 386–394.
1 A. Torres, D. Roy, B. Subramaniam and R. V. Chaudhari, Ind. Eng.
Chem. Res., 2010, 49, 10826–10835.
3
3
2 D. G. Lahr and B. H. Shanks, Ind. Eng. Chem. Res., 2003, 42, 5467–
39 C.-H. Zhou, J. N. Beltramini, Y.-X. Fana and G. Q. Lu, Chem. Soc. Rev.,
2008, 37, 527–549.
5
472.
2646 | Green Chem., 2012, 14, 2638–2646
This journal is © The Royal Society of Chemistry 2012