Selective hydrogenation of citral over a carbon-titania composite supported palladium catalyst

Article abstract of DOI:10.1002/cjoc.201190137

A novel carbon-titania composite material, C/TiO₂, has been prepared by growing carbon nanofibers (CNFs) on TiO₂ surface via methane decomposition using Ni-Cu as a catalyst. The C/TiO₂ was used for preparing supported palladium catalyst, Pd/C/TiO₂. The support and Pd/C/TiO₂ catalyst were characterized by BET, SEM, XRD and TG-DTG. Its catalytic performance was evaluated in selective hydrogenation of citral to citronellal, and compared with that of activated carbon supported Pd catalyst. It was found that the Pd/C/TiO₂ catalyst contains 97% of mesopores. And it exhibited 88% of selectivity to citronellal at citral conversion of 90% in citral hydrogenation, which was much higher than that of activated carbon supported Pd catalyst. This result may be attributed to elimination of internal diffusion limitations, which were significant in activated carbon supported Pd catalyst, due to its microporous structure. A novel palladium catalyst supported on carbon-titanium composite, Pd/C/TiO ₂, has been prepared via chemical reduction process. It contains 97% mesopores and exhibited higher selectivity to citronellal in citral hydrogenation compared to formed activated carbon supported palladium catalyst, which was attributed to elimination of internal diffusion limitations.

Full text of DOI:10.1002/cjoc.201190137

FULL PAPER
Selective Hydrogenation of Citral over a Carbon-titania
Composite Supported Palladium Catalyst
Zhu, Jie(朱劼)
Lu, Mohong(鲁墨弘)
Li, Mingshi*(李明时)
Zhu, Jianjun(朱建军)
Shan, Yuhua(单玉华)
School of Petrochemical Engineering, Changzhou University, Changzhou, Jiangsu 213164, China
A novel carbon-titania composite material, C/TiO
TiO surface via methane decomposition using Ni-Cu as a catalyst. The C/TiO
palladium catalyst, Pd/C/TiO . The support and Pd/C/TiO catalyst were characterized by BET, SEM, XRD and
2
, has been prepared by growing carbon nanofibers (CNFs) on
2
2
was used for preparing supported
2
2
TG-DTG. Its catalytic performance was evaluated in selective hydrogenation of citral to citronellal, and compared
with that of activated carbon supported Pd catalyst. It was found that the Pd/C/TiO catalyst contains 97% of
2
mesopores. And it exhibited 88% of selectivity to citronellal at citral conversion of 90% in citral hydrogenation,
which was much higher than that of activated carbon supported Pd catalyst. This result may be attributed to elimi-
nation of internal diffusion limitations, which were significant in activated carbon supported Pd catalyst, due to its
microporous structure.
Keywords carbon-titania composite, carbon nanofibers, hydrogenation, citral, internal diffusion
Introduction
commercial catalyst, i.e., activated carbon supported
palladium.
Citral, a typical α,β-unsaturated aldehydes, is impor-
tant in fine chemical and pharmaceutical industries in
that there are three sites of hydrogenation: a conjugate
double bond (C＝C), a carbonyl group (C＝O) and an
isolated double bond (C＝C). The reduction of citral
would lead to a consecutive reaction network (Figure 1).
Citronellal, an intermediate product of the consecutive
reactions, can be hydrogenated easily to citronellol and
Experimental
Preparation of C/TiO composite
2
The cylindrical TiO
2
(Sinopec Yangzi, China, φ2×6
2
－
1
mm; 18.2 m •g ) was used in this study to prepare
2 2
C/TiO composite material. Specifically, C/TiO was
prepared by impregnation of TiO
of Ni(NO •6H O and Cu(NO
2
with aqueous solution
•3H O (A.R., Sino-
1
-4
finally to 3,7-dimethyloctan-1-ol. Several systematic
studies on citral selective hydrogenation have been re-
ported. These studies covered almost all important as-
pects such as active metals including mono-cata-
3
)
2
2
3
)
2
2
pharm, China), in which the nickel loading was fixed at
18
6
wt% while the molar ratio of Ni to Cu was 8 to 1.
Then, the samples were dried overnight at 393 K prior
to calcination in muffle furnace at 773 K for 2 h. The
samples were subsequently loaded inside a quartz tube
2
,5-11
12-15
lysts
cally, Yilmaz et al. reported the highest selectivity
90%) to citronellal so far, at complete conversion of
and the roles of catalyst supports.
Specifi-
1
2
(
and reduced at 873 K in a gas mixture of N
2
and H
99.999%, Wuxi Tianhong, China) for 3 h. Methane
99.999%, Wuxi Tianhong, China) was passed through
2
citral over Pd/natural zeolite, which was explained by
the active metal properties and the absence of signifi-
cant internal diffusion limitations.
(
(
the tube to decompose at 873 K for 5 h. After being
cooled down to the room temperature under N protec-
tion, the composite support C/TiO was obtained. Fi-
nally, the exposed nickel and copper metal particles
were removed by stirring the composite in concentrated
Titania and their composites have usually been fo-
2
16,17
cused on the study of photocatalytic properties.
While in this work, we prepared carbon-titania compos-
ite as a novel catalyst support, named C/TiO . The pre-
pared C/TiO and its supported palladium catalyst,
Pd/C/TiO , have been characterized using various
methods. Meanwhile, the catalytic performance of the
Pd/C/TiO catalyst has been evaluated in selective hy-
drogenation of citral to citronellal, and compared with a
2
2
2
HNO
for 2 h.
3
2
Preparation of C/TiO
2
supported Pd catalyst
2
0.5 wt% of Pd catalyst was prepared via chemical
reduction process. Firstly, ethylenediaminetetraacetic
*
E-mail: mingshili@yahoo.com; Tel.: 0086-0519-86330360; Fax: 0086-0519-86330360
Received September 14, 2010; revised December 17, 2010; accepted December 24, 2010.
Project supported by the National Natural Science Foundation of China (No. 20976076) and the Natural Science Foundation of Jiangsu Province (No.
BK2009536).
Chin. J. Chem. 2011, 29, 655— 660
© 2011 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
655

Zhu et al.
FULL PAPER
acid (EDTA) solution (A.R., Sinopharm, China) was
reaction samples were withdrawn from the reactor at
different reaction times and analyzed by gas chroma-
tography (VARIAN CP3800) equipped with a FID de-
tector and a capillary column HP-5 using nitrogen as
carrier gas. A formed activated carbon supported cata-
lyst, named Pd/AC, (Pd loading 0.5 wt%, Sinopec
Yangzi, China) was tested at the same conditions for
comparison. In examining the influence of the structure
of the catalyst on citral hydrogenation, we defined the
citral conversion X and the selectivity towards reaction
products S as
added dropwise into an aqueous solution of PdCl
2
(Pd≥
59%, Sinopharm, China) and stirred for 3 h to obtain
Pd-EDTA chelates. Secondly, the support was impreg-
nated with Pd-EDTA chelates, and stirred continuously
overnight. Then, the pH of the solution was adjusted to
9
—10 and the catalyst was reduced at 363 K for 2 h
using hydrazine hydrate. Finally, the reduced catalyst
was collected via filtration followed by subsequent
－
washing to remove Cl . The catalyst prepared was
named Pd/C/TiO
Characterization of supports and catalysts
C/TiO support and Pd/C/TiO catalyst were charac-
2
.
0
n
－n
Citral
Citral
X＝
× 100%
(1)
(2)
2
2
0
n
Citral
terized to test the textural and structural properties.
X-ray powder diffraction data were collected on Rigaku
D/max 2500 diffractometer using CuKα (40 kV, 40 mA)
radiation and a graphite monochromator. The BET spe-
cific surface areas were carried out with Micromeritics
ASAP 2010 apparatus. Scanning electron images were
recorded using a JEOL JSM-6360 LA scanning electron
microscope. Thermogravimetric (TGA) and differential
thermal (DTA) analysis were performed using a TA
SDT Q600 instrument.
ni
S＝
× 100%
i
n
∑
i
where n is the temporal molar amount of the reactant or
0
the product during citral hydrogenation; n is the initial
molar amount of the reactant; i is the reaction product
shown in Figure 1 including citronellal, citronellol,
3,7-dimethyloctanol, geraniol, nerol and isopulegol.
Activity test of Pd/C/TiO
2
The catalytic activity of Pd/C/TiO
ated in selective hydrogenation of citral in liquid phase.
The reaction was conducted in a 100 mL autoclave at
2
has been evalu-
Results and discussion
Characterizations of the catalyst
3
63 K and 3 MPa. By dispersing the catalysts (1 wt%)
BET analysis The novel C/TiO
2
composite can be
host by
methane decomposition. The surface properties of such
prepared C/TiO composites were different from the
corresponding TiO hosts, as shown in Table 1. Specifi-
into the isopropanol (A.R., Sinopharm, China) solution
of citral (10 wt%, A.R., Sinopharm, China), the auto-
clave was sufficiently flushed with nitrogen flow to re-
move dissolved oxygen. The reaction pressure was
maintained by injecting hydrogen. Small amounts of
produced by depositing carbon onto TiO
2
2
2
Figure 1 The consecutive reaction network for citral hydrogenation.
6
56
www.cjc.wiley-vch.de
© 2011 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2011, 29, 655— 660

Selective Hydrogenation of Citral over a Palladium Catalyst
Table 1 Textural properties of the catalysts
Weight increased after methane BET surface area/ Micropore area/
Sample
Average pore diameter/nm
2
－
1
2
－
1
decomposition/wt%
(m •g )
18.3
(m •g )
3.1
TiO
2
support
composite
—
38
—
14.2
6.2
C/TiO
2
60.3
1.8
Pd/C/TiO catalyst
58.1
1.6
9.7
2
Pd/AC (commercial) catalyst
—
810
720
2.1
cally, depositing carbon on TiO
2
hosts can cause a sig-
TG-DTG analysis
Thermogravimetric analysis
nificant increase in BET surface area. The BET surface
has been conducted to determine the percentage of car-
bon deposited into the catalyst as well as its structural
properties. The results are shown in Figure 3.
2
2
area of C/TiO
2
was 60 m /g, much larger than 18 m /g
of its TiO host while its micropore area decreased sig-
2
2
nificantly to 1.8 m /g, accounting for only 3% of its to-
tal BET surface area. In contrast, the micropore area of
original TiO
face area. Therefore, it can be inferred that some mi-
cropores in TiO may be blocked due to the formation
2
accounted for 16.9% of its total BET sur-
2
of carbon deposits, and this may decrease internal diffu-
sion during multiphase catalytic reactions. It was noted
that after Pd loading over C/TiO
of the Pd/C/TiO catalyst were similar to the C/TiO
support, which suggested that the preparation method of
Pd/C/TiO catalyst in this work would not damage the
2
, the surface properties
2
2
2
structure of the catalyst.
SEM analysis As shown by SEM images in Figure
2a, a significant amount of CNFs with diameters of
5
0—80 nm covered the surface of TiO
2
. The thickness
of the deposition layer identified from an SEM image of
C/TiO cross section in Figure 2b is about 1.5 µm. Par-
2
ticularly, some CNFs appeared even in the interior of
the hosts.
Figure 3 TG and DTG curves of Pd/C/TiO
2
(Solid line corre-
sponds TG curve; dashed line corresponds to DTG curve).
2
The total weight loss of Pd/C/TiO catalyst in tem-
perature range from 573 to 973 K was over 35%, as
calculated from TG curve, being near to that of weight
increase in the support (Table 1). Generally, amorphous
carbon and graphite carbon have their own characteris-
tic oxidation temperatures. A higher oxidation tempera-
ture always indicates a purer and less defective material.
The oxidation temperature of amorphous carbon is rela-
tively low, i.e., in a range of 573 to 673 K, while that of
1
9
graphite carbon is ranging from 673 to 973 K. There-
fore, the structure of carbon deposits can be well esti-
mated from the DTG curve. Particularly, it can be in-
ferred from the TG curve in Figure 3 that over 95%
weight loss occurred between 673 and 973 K with the
DTG peak at 853 K while only a little weight loss oc-
curred between 573 and 673 K. Therefore, it is reason-
able to conclude that about 95% carbon deposits pro-
duced by methane decomposition are graphite. It should
1
7
be noted that CNFs have graphite structure. However,
the CNF structure in the interior of C/TiO was not
2
clearly observed by using SEM (Figure 2b) probably
because CNF structure might be covered by a small
percentage of amorphous carbon. Therefore, we sup-
posed that the graphite in the interior of Pd/C/TiO
mainly CNFs.
2
was
Figure 2 SEM images of C/TiO prepared by decomposition of
2
methane over Ni-Cu/TiO : (a) the surface of C/TiO ; (b) the cross
2
2
XRD analysis The structure of Pd/C/TiO
2
has
section of C/TiO2.
Chin. J. Chem. 2011, 29, 655— 660
© 2011 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cjc.wiley-vch.de
657

Zhu et al.
FULL PAPER
been further analyzed using XRD. The XRD spectrum
is shown in Figure 4, in which the characteristic peaks
of graphite (002) and graphite (101) at 2θ (26.38°,
Pd/AC, 35%. The difference in selectivity between
Pd/C/TiO
2
and Pd/AC is very likely due to the differ-
mainly con-
ence of the catalyst structures. Pd/C/TiO
2
4
4.39°) were clearly evident, indicating the dominance
sists of mesopores and the micropore area only accounts
for 3% of its total BET surface area, compared with
Pd/AC (89%) according to Table 1.
of graphite carbon deposit in host material after methane
decomposition. However, the characteristic palladium
crystal peaks at 2θ (40.1°, 46.6°, 68.1°, 82.1°, 86.6°)
were not obvious. It was supposed that the low Pd load-
ing (0.5 wt%) may result in small Pd crystal, which
cannot be measured by XRD.
Figure 5 Citral conversion versus reaction time in citral hydro-
genation at 363 K and 3 MPa over Pd/C/TiO (■) and commercial
2
Pd/AC (▲).
Figure 4 XRD pattern of Pd/C/TiO : ▲ Anatase-TiO ; ★
2
2
graphite (002); ■ graphite (101).
In summary, during methane decomposition, a sig-
nificant amount of carbon was deposited on the surface
of TiO host, among which nearly 95% were CNFs. The
2
external surface of the host was covered by a thick
CNFs layer, while internal pores were filled mostly with
CNFs along with a small percentage of amorphous car-
bon, which cannot be identified using SEM. Comparing
2 2
to TiO host, the BET surface area of C/TiO increased
greatly after carbon depositing. Furthermore, the decline
in micropore area will lower the internal diffusion dur-
ing multiphase catalytic reactions.
2
Evaluation of Pd/C/TiO catalytic activity
The catalytic performance of C/TiO supported Pd
2
2
catalyst (Pd/C/TiO ) in hydrogenation of citral was
evaluated and the results were compared with those of
commercial Pd catalyst Pd/AC (Figures 5, 6). As shown
in Figure 5, the catalytic activity of Pd/C/TiO was
2
much lower than that of Pd/AC. For example, the citral
conversion over Pd/AC was 90% after 4 h, while it took
3
2 h for Pd/C/TiO
2
catalyst to reach the same conver-
might be
sion. The relatively low activity of Pd/C/TiO
2
due to the low specific surface area of the catalyst (Ta-
ble 1). However, the major products over Pd/AC were
citronellal and 3,7-dimethyloctanol, while citronellal
over Pd/C/TiO
ronellal selectivity of Pd/C/TiO
higher than that of commercial Pd/AC. Particularly, at
the citral conversion 90%, the selectivity of Pd/C/TiO
can be as high as 88%, being much higher than that of
2
, as shown in Figure 6a and 6b. The cit-
Figure 6 The product distribution versus citral conversion in
2
was significantly
citral hydrogenation at 363 K and 3 MPa over Pd/C/TiO
2
(a) and
commercial Pd/AC (b): Citronellal; citronellol;
■
▲
▼
2
3
,7-dimethyloctanol; ◄ geraniol＋nerol; ○ isopulegol.
6
58
www.cjc.wiley-vch.de
© 2011 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2011, 29, 655— 660

Selective Hydrogenation of Citral over a Palladium Catalyst
Palladium catalysts are very active in the hydrogena-
easily desorbed to the solution, as the same on external
surface of the catalyst. However, if the reaction took
place inside the micropore, citronellal was confined in it
long enough to be further hydrogenated to 3,7-dimethyl-
octanol (Figure 7c) due to the limitation of the internal
diffusion.
It was noted that during the initial reaction over
Pd/AC with citral conversion less than 10%, the selec-
tivity to citronellal was comparatively high (80%) (Fig-
ure 6b). This was attributed to the citral hydrogenation
on external surface of Pd/AC. With the reaction going
on, citral was deeply hydrogenated to 3,7-dimethyl-
octanol on internal surface of Pd/AC due to its high mi-
cropore area. Therefore, the selectivity to citronellal
decreases rapidly with the increase of citral conversion
2
0
tion of the C＝C bond. Therefore, citral was firstly
hydrogenated to citronellal on external surface of the
catalyst (Figure 7a), and then citronellal is easily de-
sorbed from Pd crystal to the solution. When citral dif-
fused from external surface to different pores of the
catalyst, it hydrogenates to different products. If the
reaction took place in the macro- and mesopore (Figure
7
b) of the catalyst, the intermediate product, citronellal,
(
Figure 6b). However, the influence of internal diffusion
limitation on Pd/C/TiO was not significant since
Pd/C/TiO had much fewer micropores than Pd/AC.
2
2
Therefore, the selectivity to citronellal was almost con-
stant with the increase of citral conversion over
Pd/C/TiO (Figure 6a).
2
Conclusions
A novel carbon-titania composite material C/TiO
2
that can be utilized as a catalyst support has been pre-
pared by decomposition of methane over TiO surface
using Ni-Cu as a catalyst. The carbon in Pd/C/TiO
2
2
catalyst is composed of nearly 95% of CNFs and 5% of
amorphous carbon. And 97% of its textural structures
are mesopores. The catalyst Pd/C/TiO exhibited 88%
2
of selectivity to citronellal at citral conversion of 90% in
citral hydrogenation, which was much higher than that
of activated carbon supported Pd catalyst (35%). This
was attributed to elimination of internal diffusion limi-
tations, which was significant in activated carbon sup-
ported Pd catalyst, due to its microporous structure.
References
1
Vilella, I. M. J.; De Miguel, S. R.; Salinas-Martínez, L. C.;
Linares-Solano, Á.; Scelza, O. A. Appl. Catal. A: Gen. 2005,
2
81, 247.
2
3
Singh, U. K.; Vannice, M. A. J. Catal. 2001, 199, 73.
Álvarez-Rodríguez, J.; Cerro-Alarcón, M.; Guerrero-Ruiz,
A.; Rodríguez-Ramos, I.; Arcoya, A. Appl. Catal. A: Gen.
2
008, 348, 241.
4
5
6
Kiwi-Minsker, L.; Yuranov, I.; Holler, V.; Renken, A.
Chem. Eng. Sci. 1999, 54, 4785.
Li, C. J.; Ma, B. X.; Huo, X. X. New. Carbon. Mater. 1999,
1
4, 19 (in Chinese).
Salmi, T.; Mäki-Arvela, P.; Toukoniitty, E.; Neyestanaki, A.
K.; Tiainen, L. P.; Lindfors, L. E.; Sjöholm, R.; Laine, E.
Appl. Catal. A: Gen. 2000, 196, 93.
7
8
9
Aramendia, M. A.; Borau, V.; Jimenez, C.; Marinas, J. M.;
Porras, A.; Urbano, F. J. J. Catal. 1977, 172, 46.
Trasarti, A. F.; Marchi, A. J.; Apesteguía, C. R. J. Catal.
2004, 224, 484.
Figure 7 The process of citral hydrogenation over Pd/C/TiO
2
:
(a) on external surface of the catalyst; (b) in the macro- and
mesopores; (c) in the micropores.
Singh, U. K.; Sysak, M. N.; Vannice, M. A. J. Catal. 2000,
Chin. J. Chem. 2011, 29, 655— 660
© 2011 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.cjc.wiley-vch.de
659

Zhu et al.
FULL PAPER
1
91, 181.
Kumar, N.; Sjöholm, R.; Ollonqvist, T.; Väyrynen, J.; Salmi,
T.; Murzin, D. Y. Appl. Catal. A: Gen. 2003, 241, 271.
16 Chen, Q. F.; Shi, W. M.; Jiang, D.; Xu, Y.; Wu, D.; Sun, Y.
H. Acta Chim. Sinica 2010, 68, 301 (in Chinese).
10
Arvela, P. M.; Tiainen, L. P.; Neyestanaki, A. K.; Sjöholm,
R.; Rantakylä, T. K.; Laine, E.; Salmi, T.; Murzin, D. Y.
Appl. Catal. A: Gen. 2002, 237, 181.
11
12
13
14
15
Aumo, J.; Wärnå, J.; Salmi, T.; Murzin, D. Y. Chem. Eng.
Sci. 2006, 61, 814.
Yilmaz, S.; Ucar, S.; Artok, L.; Gulec, H. Appl. Catal. A:
Gen. 2005, 287, 261.
Zhang, Y. J.; Li, J. Y.; Li, W. Z.; Xin, Q. Chem. J. Chin.
Univ. 2005, 26, 1345 (in Chinese).
Malathi, R.; Viswanath, R. P. Appl. Catal. A: Gen. 2001,
1
1
1
7
8
9
Wu, J. M.; Yao, J. J.; Yang, H. P.; Fan, Y. N.; Xu, B. L. Acta
Chim. Sinica 2010, 68, 1349 (in Chinese).
Wu, F.; Zhu, J.; Li, M. S.; Zhu, J. J.; Shan, Y. H. Fine. Chem.
2
009, 26, 561 (in Chinese).
Tsoufis, T.; Xidas, P.; Jankovic, L.; Gournis, D.; Saranti, A.;
Bakas, T.; Karakassides, M. A. Diam. Relat. Mater. 2007,
1
6, 155.
2
08, 323.
20
Delbecq, F.; Sautet, P. J. Catal. 1995, 152, 217.
Mäki-Arvela, P.; Tiainen, L. P.; Lindblad, M.; Demirkan, K.;
(E1009141 Sun, H.; Fan, Y.)
6
60
www.cjc.wiley-vch.de
© 2011 SIOC, CAS, Shanghai, & WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chin. J. Chem. 2011, 29, 655— 660