Hydrogen production on Ni–Pd–Ce/g-Al2O3 catalyst by partial

oxidation and steam reforming of hydrocarbons for potential

application in fuel cells

Y.H. Wang, J.C. Zhang*

The Key Laboratory of Science and Technology of Controllable Chemical Reactions, Ministry of Education,

Beijing University of Chemical Technology, Beijing 100029, China

Received 21 November 2004; received in revised form 23 March 2005; accepted 23 March 2005

Available online 18 April 2005

Abstract

A series of catalysts of nickel–palladium–cerium supported upon alumina have been investigated in order to obtain a suitable catalyst that

could be used in the process of producing hydrogen for the potential application in fuel cell by partial oxidation and steam reforming (POSR)

of mixtures of hydrocarbons. Investigated results showed that little addition of cerium into Ni–Pd catalyst could definitely improve its

stability of hydrogen production from mixtures of hydrocarbons by POSR method. The experimental results also showed that the optimum

compositions of Ni–Pd–Ce catalyst were the molar ratio of Ni to Pd as 1:0.09 and containing of Ce 0.5 wt%, shortened as Ni–Pd–Ce-0.5

catalyst. XRD results for the typical catalysts showed that it mainly displayed the g-Al2O3 and Ni peaks. SEM and TG results for the fresh

and used Ni–Pd–Ce-0.5 catalysts, lasted for 540 h, did not show much difference on their surface patterns and TG curves, respectively.

This indicated this catalyst would be a practical catalyst to produce hydrogen from liquid fuel by POSR method for potential application in

fuel cells.

q 2005 Elsevier Ltd. All rights reserved.

Keywords: Hydrogen production; Partial oxidation; Steam reforming; Hydrocarbons

1. Introduction

Fuel cells are a viable alternative for clean energy

generation. Over the past few years, fuel cell and

automotive companies have announced new technologies

or prototype vehicles adopting fuel cells in an effort to

reduce atmosphere pollution [1]. A variety of fuel cells for

different applications is under development, e.g. solid

polymer fuel cells (SPFC), also known as proton-exchange

membrane fuel cells (PEMFCs) operating about 353 K,

alkaline fuel cells (AFC) operating about 373 K, phosphoric

0016-2361/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.fuel.2005.03.023

* Corresponding author. Present address: Institute on Membrane

Technology, CNR-ITM, C/o University of Calabria, Via P. Bucci 17/C,

87030 Rende, CS, Italy. Tel.: C39 0984 49 2050.

E-mail addresses: zhangjc@mail.buct.edu.cn (J.C. Zhang),

zhangitaly@sina.com (J.C. Zhang), zhangjc2005@yahoo.com.cn

(J.C. Zhang).

acid fuel cells (PAFC) about for 473 K, molten carbonate

fuel cells (MCFC) operating around 923 K, solid oxide fuel

cells (SOFC) for high temperature operation, 1073–1373 K.

PEMFCs possess a series of advantageous features that

make them leading candidates for mobile vehicle power

applications or for small stationary power units [2–4], low

operating temperature, sustained operation at high current

density, low weight, long stack life, fast start-ups as well as

suitability to discontinuous operation, etc. However, the

ideal fuel for PEMFCs is pure hydrogen, with less than

50 ppm carbon monoxide, as dictated by the poisoning limit

of the Pt fuel cell catalyst. Size, weight, cost and other

technical limitations make it difficult to store hydrogen in

necessary quantity and density, so the hydrogen gas will

likely be generated on site and on demand. Therefore, the

paramount issue facing fuel cells, which provides power for

the mobile vehicles, right now is how to get the hydrogen to

the vehicles. Investigations have been carried out with

several approaches to produce hydrogen, e.g. steam

reforming [5–7], electrochemical [8,9], photochemical

Fuel 84 (2005) 1926–1932

www.fuelfirst.com

Y.H. Wang, J.C. Zhang / Fuel 84 (2005) 1926–1932 1927

[10], biological [11,12] and thermochemical methods [13].

Steam reforming is one of the least expensive hydrogen

production methods. Hence, many research works are

concentrated on steam reforming of methane [14,15],

methanol [16–20] and ethanol [21]. Compared with the

above fuels, however, hydrocarbons, e.g. naphtha, kerosene

and gasoline, have the following advantages of higher heat

value, large amounts of storage hydrogen and steady supply

as well as convenient transportation [22,23]. Moreover, the

method of producing hydrogen from hydrocarbons through

partial oxidation and steam reforming enjoys the merit of

low energy requirement [24,25], due to the opposite

contribution of the exothermic hydrocarbon oxidation and

endothermic steam reforming. The process, however, still

needs a suitable and active catalyst, which is not well

developed at present time [26,27]. Wang and Wu [26]

primarily reported that the catalyst of Ni–Pd supported upon

Al2O3 had good activity of producing hydrogen from single

n-octane using POSR method. Recently, Zhang et al. [28]

study further indicated this catalyst not only had good

activity but also good stability under similar reaction

conditions reported in [26]. Pino et al. [27] reported that

the catalyst of platinum supported on ceria had good activity

and selectivity in the partial oxidation of methane. Stability

test, lasted for 100 h, indicated that it was stable under

experimental conditions. Yang et al. [29] reported the

important mechanism of ceria and palladium promotion on

Ni supported upon Al2O3 catalyst. They reported that the

addition of CeO2 to Ni/g-Al2O3 catalyst not only decreased

carbon deposition on it but also increased the ability of

elimination of carbon from the catalyst. The addition of

noble metal Pd could further improve its resistance to

carbon deposition upon the catalyst. Those studies provided

the basis for developing a novel catalyst to be used in the

process of hydrogen production from mixture hydrocarbons

by POSR method.

The objective of present investigation is to develop an

effective catalytic system for the selective production of

hydrogen for its potential application in fuel cells by

partial oxidation and steam reforming of hydrocarbons. A

series of catalysts of nickel–palladium–cerium supported

upon alumina have been prepared in order to obtain a

suitable catalyst that could be used in the process of

producing hydrogen by partial oxidation and steam

reforming of mixtures of hydrocarbons. It could provide

the basis for the further investigation of production

hydrogen for fuel cells application through partial

oxidation and steam reforming of naphtha, kerosene and

gasoline, etc. The prepared catalyst of Ni–Pd–Ce-0.5 has

been undergone 540 h stability experiment by using the

hydrocarbon mixtures. Finally, the fresh and used

catalysts were characterized by using SEM, BET, XRD,

TG and ICP methods. This paper reported our primary

investigated results from the preparation of the catalysts

of Ni–Pd–Ce-0.5 to be used in the hydrogen production

by POSR.

2. Experimental methods

2.1. Experimental setup

Investigations have been performed in a fixed-bed

reactor at atmospheric pressure. The reactor was constructed

from a stainless steel pipe of 10 mm inner diameter and

400 mm length. At the bottom a perforated gas distributor

was equipped. The reactor was placed inside an electric

furnace equipped with an electric heater driven by a

proportional-integral-derivative (PID) electronic tempera-

ture controller. The prepared catalyst, 3 ml, was sandwiched

between quartz wool. The temperature of the catalytic bed

was monitored with a thermocouple sliding inside a

stainless steel pipe that was well inserted in the catalyst

bed. Because of the high exothermic hydrocarbon oxidation

and endothermic steam reforming, the temperature of

catalyst bed was varied at the beginning of the reaction.

All the data were collected when the temperature equili-

brium was established. The scheme of the experimental

system could be referred [28].

2.2. Catalyst preparation

The catalysts used in the experiment were obtained by

the impregnation of g-Al2O3 support with the solution

containing 31.1(wt)% Ni (NO3)2 and 2.72(wt)% PdCl2,

respectively. To the solutions of containing different

concentrations of cerium, they were prepared by dissolving

certain amounts of Ce(NO3)3$6H2O into distilled water

listed in Table 2. Afterwards, it was followed by drying in

an oven at 383 K overnight under vacuum condition and

calcined in a muffle furnace at 1073 K for about 6 h. It

should be especially mentioned that the Ni–Pd–Ce–O

catalysts were mainly prepared through two impregnation

steps. The Ni–Ce–O/g-Al2O3 was first prepared through co-

impregnation, followed by drying in an oven at 383 K and

calcining in a muffle furnace at 773 K for 3 h through which

series of catalysts of Ni–Ce–O/g-Al2O3 were obtained.

Then, a series of catalysts of Ni–Pd–Ce–O/g-Al2O3 were

obtained by using the prepared materials of Ni–Ce–O/g-

Al2O3 as support by impregnated method. The reduced

catalyst was obtained through reducing its precursor in the

fixed-bed reactor at 723 K with mixtures of hydrogen

balanced with nitrogen that were fed to the reactor via the

mass controller. All the prepared catalysts were kept in a

drying utensil and used as catalysts for the POSR reaction.

2.3. Experimental conditions

Mixtures of n-pentane, n-hexane, n-octane, isooctane and

little amounts of benzene, which were used as experimental

reactants instead of naphtha or gasoline for the simplicity of

the experimental procedure, were tested as feedstock

through partial oxidation and steam reforming for the

production of hydrogen. The hydrocarbons were mixed

Table 1

Operational parameters of POSR for the mixtures of hydrocarbons

Temperature

(K)

Composition of hydrocarbons (mol%) Feed rates (ml minK1)

n-C5H12 n-C6H14 n-C8H18 i-C8H18 C6H6 Oil Water Air

973 16.0 30.0 25.0 25.0 4.0 0.05 0.13 83.0

Y.H. Wang, J.C. Zhang / Fuel 84 (2005) 1926–19321928

together before the experiment and its compositions were

listed in Table 1. Distilled water was fed through a metering

pump and got vaporized in the preheating section prior to

injection into the reactor. The mixtures of hydrocarbons

were likewise sent through a metering pump into the reactor,

which got vaporized in the preheating zone before being fed

into the reactor. The typical activity tests were carried out at

973 K with liquid space velocities of mixtures of hydro-

carbons (LSVMH) 1 hK1. LSVMH was defined as volume

per hour of the liquid mixtures of hydrocarbons at 298 K

and 0.1 MPa per volume of the catalytic bed. The amounts

of water and air introduced into reaction system were

determined to their molar ratios of molH2O/molC and

molO2/molHydrocarbons, respectively. The molar ratios of

molH2O/molC and molO2/molHydrocarbons were deter-

mined to be equal to 3.0 and 2.0, respectively, according to

the investigated results [26,28]. Thus, the tests reported in

this paper were also carried out at 973 K with gas hourly

space velocities (GHSV) 14200 hK1. Likewise LSVMH

definition, GHSV was defined as volume per hour of the

gaseous feed (including the nitrogen introduced into

the system by air) at 298 K and 0.1 MPa per volume of

the catalytic bed. Amounts of H2, CO, CO2, CH4 and N2

were analyzed with thermal conductivity detector (TCD)

with a column consisting of molecular sieves 5A and 13X

with argon as carrier gas at 393 K. Amounts of C2, C3 and

other organic compounds were determined using an OV-1

capillary column connected to a hydrogen flame ionization

detector (FID) at 303 K, which was reported elsewhere

[26,28]. The calculation for the conversion ratio of

hydrocarbons and dry gas compositions were, therefore,

calculated according to the following equations for the

simplified reason [27]

Ci ZDFi

Fi;in

ZFi;in KFi;out

Fi;in

100 (1)

Total conversion ð%Þ ZX

i

Ci (2)

H2ð%Þ ZFH2

FH2CFCH4

CFC2H4CFC2H6

CFC3H8

% (3)

COð%Þ ZFCO

FCO CFCO2CFCH4

CFC2H4CFC2H6

CFC3H8

%

(4)

CO2ð%ÞZFCO2

FCO CFCO2CFCH4

CFC2H4CFC2H6

CFC3H8

%

(5)

CH4ð%ÞZFCH4

FCO CFCO2CFCH4

CFC2H4CFC2H6

CFC3H8

%

(6)

Where Fi is the molar flow rate of species of i, Ci is the

conversion ratio of species of i, i indicated n-pentane,

n-hexane, n-octane, i-octane, benzene, respectively.

2.4. Characterization of typical catalysts

X-ray diffraction (XRD) characterization was analyzed

by using the apparatus of Rigaku D/MAX-RB. The analyzed

conditions were CuKa radial, 40 kV tube voltage and

100 mA tube current with a scan rate of 2qZ28/min.

Scanning electron microscope (SEM) characterization was

analyzed by using S-3200N HITACH apparatus. The

surface area of preparation catalysts was investigated by

using Micromeritics ASAP-2000 apparatus. The TG data

were obtained using a PCT-1A instrument. The pure

nitrogen was introduced into the system at the beginning

when the temperatures were ranging between room

temperature and 393 K as well as the system was maintained

2 h at the temperature of 393 K. Then, the temperature of

the system was increased to 1173 K at the rate of

283 K minK1, and the mixtures of nitrogen and hydrogen

was simultaneous instead of the pure nitrogen. Finally, the

catalytic components of Ni, Pd and Ce for the prepared

catalysts were determined by means of induced coupled

plasma (ICP) analysis, Perkin–Elmer ICP-500 apparatus.

3. Results and discussion

3.1. Investigation of Ni–Pd–Ce/g-Al2O3 activities

The compositions and BET surface area for the prepared

catalysts of Ni–Pd/g-Al2O3 and series of Ni–Pd–Ce/

g-Al2O3 were listed in Table 2, which showed that the

catalyst surface area was not varied much with the little

addition of cerium into the Ni–Pd/g-Al2O3 catalysts. It has

been reported that the Ni–Pd/g-Al2O3 catalyst has compara-

tively higher activity of production hydrogen from single

hydrocarbon and the mainly reason would be that the Pd is

more stable than Pt during the high vapor pressure at the

working condition [28]. It is also indicated that little

Table 2

Characterization of fresh Ni–Pd/g-Al2O3 and Ni–Pd–Ce/g-Al2O3 catalysts

Catalysts Molar ratio

(Ni:Pd)

Content of Ce

(wt%)

BET surface

area (m2 gK1)

Ni–Pd 1:0.09 – 148.0

Ni–Pd–Ce-0.1 1:0.09 0.1 147.0

Ni–Pd–Ce-0.3 1:0.09 0.3 148.0

Ni–Pd–Ce-0.5 1:0.09 0.5 145.0

Ni–Pd–Ce-1.0 1:0.09 1.0 144.0

Ni–Pd–Ce-1.5 1:0.09 1.5 143.0

Y.H. Wang, J.C. Zhang / Fuel 84 (2005) 1926–1932 1929

addition of Ce into the catalyst of containing noble metals

could improve its reaction property of hydrogen production

[27]. In order to evaluate the prepared Ni–Pd/g-Al2O3 and

Ni–Pd–Ce/g-Al2O3 catalyst activities, the experiments were

carried out at the listed reaction conditions, as shown in

Table 1. The catalyst activity of Ni–Pd/g-Al2O3 and series

of Ni–Pd–Ce/g-Al2O3 in the partial oxidation and steam

reforming process was shown in Fig. 1 as a function of the

reaction time at 973 K. To the Ni–Pd/g-Al2O3 catalyst, its

conversion of hydrocarbon mixtures was maintained around

93%. However, compared with the Ni–Pd/g-Al2O3 catalyst,

the activities for the Ni–Pd–Ce/g-Al2O3 catalyst were

definitely improved by addition of little amounts of cerium

into Ni–Pd/g-Al2O3 catalyst. Moreover, Ni–Pd–Ce/g-Al2O3

catalyst activities were slightly increased with increasing

content of cerium on Ni–Pd/g-Al2O3 catalyst. But, the

catalyst activities were almost maintained constant, around

99%, when the content of cerium on the catalyst was over

0.5 wt%, which indicated that the suitable compositions for

the Ni–Pd–Ce/g-Al2O3 catalyst were 1:0.09 molar ratio

of Ni to Pd and Ce 0.5 wt%, respectively, shortened as

Ni–Pd–Ce-0.5 catalyst. The following detailed researches

for the Ni–Pd–Ce catalyst was carried out on the catalyst of

Ni–Pd–Ce-0.5. It should be mentioned here that the catalyst

activities for the Ni–Pd/g-Al2O3 and Ni–Pd–Ce/g-Al2O3

catalysts were somewhat fluctuated at the start-up, as shown

0 20 40 60 80 100 12080

85

90

95

100

Tota

l con

vers

ion

(%)

Reaction time (hr)

Ni-Pd Ni-Pd-Ce(0.1) Ni-Pd-Ce(0.3) Ni-Pd-Ce(0.5) Ni-Pd-Ce(1.0) Ni-Pd-Ce(1.5)

Fig. 1. Effect of Ce loadings on Ni–Pd–Ce catalyst activity at 973 K.

in Fig. 1, which would be resulted by feeding water and

mixtures of hydrocarbons through pumps thereby slightly

altering the feeding ratio during the experimental process.

3.2. Ni–Pd–Ce-0.5 catalyst stability tests

Those mentioned above results led us to a preliminary

long life test for the partial oxidation and steam reforming of

mixtures of hydrocarbons on the promising Ni–Pd/g-Al2O3

and Ni–Pd–Ce-0.5 catalysts. The reaction conditions were

the same as stated in the above section. The effect of time

on stream, up to 540 h, was presented in Fig. 2 for the

Ni–Pd/g-Al2O3 and Ni–Pd–Ce-0.5 catalysts, respectively.

As shown in Fig. 2, the activity of conversion mixtures of

hydrocarbons for the Ni–Pd/g-Al2O3 catalyst was slightly

decreased with increasing reaction time when it has

been undergone about 300 h. However, the activity for the

Ni–Pd–Ce-0.5 catalyst remained almost constant when it

has been lasted for 540 h. Those results performed on

Ni–Pd/g-Al2O3 and Ni–Pd–Ce-0.5 catalysts during the

stability tests indicated that cerium plays an important role

during the partial oxidation and steam reforming of mixtures

of hydrocarbons, especially, when the mixtures of hydro-

carbons contain aromatic compounds, e.g. benzene. The

result that the Ni–Pd–Ce-0.5 catalyst had a good stability

revealed that addition of little amounts of cerium into Ni–

Pd/g-Al2O3 catalyst not only improved Ni–Pd/g-Al2O3

catalyst activity but also its stability. Some main reasons

that cerium had improved Ni–Pd/g-Al2O3 catalyst property

would be discussed in the following section. The gas

compositions (not concluding the nitrogen) for the Ni–Pd–

Ce-0.5 catalyst was shown in Fig. 3. The molar compo-

sitions for H2, CO, CO2 and CH4 were about ranging

between 65–68, 14–18, 12–17 and 2–7%, respectively. It

was known that the CO would be converted to CO2 through

the steam shift reaction before the mixtures of gases were

0 100 200 300 400 500 60080

85

90

95

100

Tota

l con

vers

ion

(%)

Reaction time (hr)

Ni-Pd Ni-Pd-Ce(0.5)

Fig. 2. Effect of time on stream performance of Ni–Pd and Ni–Pd–Ce

catalyst activity at 973 K.

0 6000

20

40

60

80

100

Com

posi

tion

(%)

Reaction time (hr)

CH4 CO2 CO H2 CO+H2

200100 300 400 500

Fig. 3. Compositions of H2, CO, CO2, CH4 and H2CCO on the Ni–Pd–Ce-

0.5 catalyst at 973 K.

20 30 40 50 60 70

2 Theta (deg)

(1)

(2) a

b

c

d

e

f

(1) Fresh catalyst (2) Used Catalyst

Fig. 4. XRD characterization results of fresh and used Ni–Pd–Ce-0.5

catalysts.

Y.H. Wang, J.C. Zhang / Fuel 84 (2005) 1926–19321930

potential used to fuel cells. Theoretically, it was considered

that the moles of CO were equal to the moles of H2

according to the steam shift reaction. Then, the latent

hydrogen composition would be summed the compositions

of H2 and CO, which was about ranging between 80 and

85%, as shown in Fig. 3. It indicated that ternary catalyst of

Ni–Pd–Ce-0.5 actually could be a promising catalyst for the

production of hydrogen from the mixtures of hydrocarbons,

especially containing aromatic compounds in the feed,

which provides the basis for further investigation of

hydrogen production from gasoline by partial oxidation

and steam reforming method.

3.3. Characterization of Ni–Pd–Ce-0.5 catalyst

From the above results, it was found that the catalyst of

Ni–Pd–Ce-0.5 had good stability of hydrogen production by

using partial oxidation and steam reforming of mixtures of

hydrocarbons, especially containing aromatic compounds,

compared to the catalyst of Ni–Pd/g-Al2O3. In order to

clarify the reasons why the catalyst of Ni–Pd–Ce-0.5 had

good stable activity and the Ni–Pd/g-Al2O3 catalyst

partially lost its activity during the reaction, the fresh and

used on stream last for 540 h catalysts were characterized by

using XRD, SEM, TG and ICP apparatus, respectively.

Some characterization results of Ni–Pd/g-Al2O3 catalyst

could be found in elsewhere [28]. The characterization

results for the Ni–Pd–Ce-0.5 catalyst were mainly discussed

in this paper. Fig. 4 was the XRD characterization results for

the fresh and used catalysts of Ni–Pd–Ce-0.5. From the

XRD results, it could be seen that the characterization peaks

of Ni and g-Al2O3 were clearly appeared for the fresh and

used Ni–Pd–Ce-0.5 catalysts. Although the characterization

peaks of Pd were not obvious, it still could be seen from the

XRD results, and the characterization peaks of Ce could not

be much detected for the fresh and used Ni–Pd–Ce-0.5

catalysts, which would be probably that the contents of Pd

and Ce were comparatively lower as well as they could be

uniformly dispersed on the support. The largest difference

between the fresh and used Ni–Pd–Ce-0.5 catalysts was that

it appeared an unobvious characterization peak of CeO2 for

the used Ni–Pd–Ce-0.5 catalyst. It has been known that

aromatic compounds would easily deposit on the active sites

of catalyst surface during the hydrodesulfurization, which

would be further reacted and form analogous carbon

compounds on the catalyst active sites and then make the

catalyst gradually lose its activity during the reaction. The

unobvious CeO2 characterization peak appeared on the used

Ni–Pd–Ce-0.5 catalyst seemly revealed that the active sites

on the catalyst surface would catalytically adsorb more

oxygen because of addition little amounts of cerium into

Ni–Pd/g-Al2O3 catalyst during the POSR reaction, which

would be much useful to remove those analogous carbon

compounds through oxidation reaction. Then, the cerium

itself would be easily oxidized to oxides. This was probably

the main reason why the Ni–Pd–Ce-0.5 catalyst had good

stability than Ni–Pd/g-Al2O3 catalyst. The further study

should be carried out to explain the cerium role on the

Ni–Pd/g-Al2O3 catalyst during the POSR process. Anyway,

it showed that there was no much difference between the

fresh and reacted catalyst XRD results, which partially

indicated that the Ni–Pd–Ce-0.5 catalyst should have good

stability during the POSR reaction process. Figs. 5 and 6

were SEM results for the fresh and used Ni–Pd/g-Al2O3 and

Ni–Pd–Ce-0.5 catalysts, respectively, lasted for 540 h on the

stream during the POSR process. There was obvious

difference about the surface patterns between the fresh

and used Ni–Pd/g-Al2O3 catalysts, which was primarily

considered by the formation of little amounts of analogous

carbon compounds on the catalyst surface during the POSR

reaction process. However, it did not show much difference

on their surface patterns for the fresh and used Ni–Pd–Ce-

0.5 catalysts, as shown in Fig. 6. TG analytical results

were shown in Fig. 7 for the fresh and used Ni–Pd/g-Al2O3

Fig. 5. SEM characterization of fresh and used Ni–Pd catalysts. (a) Fresh Ni–Pd; (b) used Ni–Pd.

Fig. 6. SEM characterization of fresh and used Ni–Pd–Ce catalysts (a) Fresh Ni–Pd–Ce; (b) used Ni–Pd–Ce.

Y.H. Wang, J.C. Zhang / Fuel 84 (2005) 1926–1932 1931

and Ni–Pd–Ce-0.5 catalysts, respectively. To Ni–Pd/

g-Al2O3 catalyst, as shown in Fig. 7, it showed that

maybe some compounds were produced on the Ni–Pd/

g-Al2O3 catalyst because the used catalyst curve was

obviously decreased since about 250 8C compared with

fresh catalyst. To Ni–Pd–Ce-0.5 catalyst, TG results did not

show any great difference between the fresh and used

catalyst. The BET and ICP results for the Ni–Pd/g-Al2O3

and Ni–Pd–Ce-0.5 catalysts were listed in Table 3. It clearly

400 600 800 1000 120090

92

94

96

98

100

used catalyst (Ni-Pd-Ce)

fresh catalyst (Ni-Pd-Ce)used catalyst (Ni-Pd)

fresh catalyst (Ni-Pd)

Wei

ght (

wt.%

)

Temperature (K)

Fig. 7. TG data for fresh and used Ni–Pd and Ni–Pd–Ce catalysts on stream

lasted for 540 h.

indicated that the main reason to make the Ni–Pd/g-Al2O3

catalyst partially lose its activity could be resulted by the

variation of BET surface area resulted by the analogous

carbon compounds formed on the catalyst surface. The

largest difference between the used Ni–Pd/g-Al2O3 and Ni–

Pd–Ce-0.5 catalysts was the amounts of carbon deposition

on the catalysts have been varied, as shown in Table 3. The

amounts of carbon deposition on the Ni–Pd–Ce-0.5 catalyst

were decreased about 38% compared with Ni–Pd/g-Al2O3

catalyst. It indicated that the main function of Ce on the

catalyst was to prevent the carbon deposition on the surface

of Ni–Pd–Ce catalyst. Thus, the Ni–Pd–Ce-0.5 catalyst had

good stability compared with Ni–Pd/g-Al2O3 catalyst. The

further investigation for the Ni–Pd–Ce-0.5 catalyst to be

used to produce hydrogen by using more complicated

Table 3

Characterization of fresh and used Ni–Pd/g-Al2O3 and Ni–Pd–Ce/g-Al2O3

catalysts

Catalyst

samples

BET surface

area

(m2 gK1)

Carbon

depositions

(mg gK1)

Molar

ratio

(Ni:Pd)

Content of

Ce (wt%)

Ni–Pda 148.0 – 1:0.09 –

Ni–Pdb 139.0 2.9 1:0.09 –

Ni–Pd–Ce-0.5a 145.0 – 1:0.09 0.5

Ni–Pd–Ce-0.5b 143.0 1.8 1:0.09 0.4

a Fresh catalysts.b Used catalysts.

Y.H. Wang, J.C. Zhang / Fuel 84 (2005) 1926–19321932

mixtures of hydrocarbons to simulate the gasoline compo-

sitions would be reported in another paper.

4. Conclusion

A series of Ni–Pd–Ce catalysts were prepared in this

investigation. The investigated results showed that the

activities of Ni–Pd–Ce catalysts were slightly increased

with the increasing content of Ce on the catalyst of Ni–

Pd/ g-Al2O3. Further study indicated that the suitable

compositions for the Ni–Pd–Ce catalyst could be 1:0.09

molar ratio of Ni to Pd and Ce 0.5 wt%, respectively,

shortened as Ni–Pd–Ce-0.5 catalyst. The activity and

stability experiments for Ni–Pd/g-Al2O3 and Ni–Pd–Ce-

0.5 catalysts were carried out at 973 K and 1 hK1 of

liquid space velocities of mixtures of hydrocarbons, total

gas hourly space velocities, 14,200 hK1, which showed

that its activity of conversion of hydrocarbon mixtures,

especially containing aromatic compounds, e.g. benzene,

into hydrogen was higher than that of Ni–Pd/g-Al2O3

catalyst. The further study indicated Ni–Pd–Ce-0.5

catalyst had good stability. The characterization results

for the typical catalyst samples indicated that the main

function of Ce on the catalyst probably was to prevent the

carbon deposition on the surface of Ni–Pd–Ce catalyst.

Thus, the Ni–Pd–Ce-0.5 catalyst had good stability

compared with Ni–Pd catalyst. This implied this catalyst

was favorable for hydrogen production from the hydro-

carbon mixtures and it would be a promising catalyst to

produce hydrogen for its potential application in the fuel

cells by POSR method.

References

[1] Ghenciu AF. Curr Opin Solid State Mater Sci 2002;6:389.

[2] Lokurlu A, Grube T, Hohlein B, Stolten D. Int J Hydrogen Energy

2003;28:703.

[3] Tang LC, Yang TG. J Chem Process 1995;1:18.

[4] Keigh BP. J Power Sources 1994;51:129.

[5] Taylor JD, Herdman CM, Wu BC, Wally K, Rice SF. Int J Hydrogen

Energy 2003;28:1171.

[6] Suzuki T, Iwanami HI, Yoshinari T. Int J Hydrogen Energy 2000;25:

119.

[7] Wang X, Gorte RJ. Appl Catal A 2002;224:209.

[8] Aparicio LM. J Catal 1997;165:262.

[9] Kikuchi E, Uemiya S, Matsuda T. Stud Surf Sci Catal 1991;61:509.

[10] Koenigstein C. Photochem J Photobiol 1995;90:141.

[11] Yamada A, Takano H, Burgess GJ, Matsunaga T. J Mar Biotechnol

1996;4:23.

[12] Willner I, Steinberger B. Int J Hydrogen Energy 1998;13:593.

[13] Lee C, Yoon HK, Moon SH, Yoon KJ. Korean J Chem Eng 1998;15:590.

[14] Menderes L, Gund DJ, El-Bousiffi MA. Int J Hydrogen Energy 2003;

28:945.

[15] Dybkijar I. Fuel Process Technol 1995;42:85.

[16] Jiang CJ. Appl Catal A 1993;93:245.

[17] Jiang CJ. Appl Catal A 1993;97:145.

[18] Lindstrom B, Pettersson LJ. Int J Hydrogen Energy 2001;26:923.

[19] Johan A, Magali B, Ignacio MC, Jose LG. Appl Catal A 2003;253:

201.

[20] Nobuhiro I, Tomoyuki M, Wataru N, Masahiko A, Nobutsune T. Appl

Catal A 2003;248:153.

[21] Cavallaro S, Chiodo V, Freni S, Mondello N, Frusteri F. Appl Catal A

2003;249:119.

[22] Ahmed S, Krumpelt M. Int J Hydrogen Energy 2001;26:291.

[23] Manuel P, Jorge S, John K. Appl Catal A 2003;250:161.

[24] Ronny N, Michal LE. J Catal 1997;106:206.

[25] Newson E, Truong TB. Int J Hydrogen Energy 2003;28:1379.

[26] Wang YH, Wu DY. Int J Hydrogen Energy 2001;26:795.

[27] Pino LD, Antonio V, Manuela C, Vincenzo R, Manajanatha SH. Appl

Catal A 2003;243:135.

[28] Zhang JC, Wang YH, Ma RY, Wu DY. Appl Catal A 2003;243:251.

[29] Yang YL, Xu HY, Li WZ. Acta Phys-Chem Sin 2002;18:321.