www.elsevier.com/locate/apcata

Applied Catalysis A: General 292 (2005) 272–280

Performance of NiO–MgO solid solution-supported Pt catalysts in

oxidative steam reforming of methane

Mohammad Nurunnabi a, Baitao Li a, Kimio Kunimori a, Kimihito Suzuki b,Ken-ichiro Fujimoto b, Keiichi Tomishige a,*

a Institute of Materials Science, University of Tsukuba, 1-1-1, Tennodai, Tsukuba, Ibaraki 305-8573, Japanb Advanced Technology Research Laboratories, Nippon Steel Corporation, 20-1, Shintomi, Futtsu, Chiba 293-8511, Japan

Received 14 March 2005; received in revised form 6 June 2005; accepted 21 June 2005

Available online 16 August 2005

Abstract

The effect of Pt addition to NiO–MgO solid solution catalysts on the performance in oxidative steam reforming of methane was

investigated. In the oxidative reforming of methane, Pt/NiO–MgO gave much higher methane conversion than NiO–MgO and Pt/MgO,

especially under short contact time (low W/F : W = catalyst weight, F = total flow rate). Although the effect of Pt was not clear in temperature

program reduction profiles, the additive effect of Pt is remarkable in oxidative reforming of methane in term of catalyst activation. In the case

of Pt/NiO–MgO catalysts, even when the H2 reduction pretreatment was not done, the catalyst can be activated at a temperature higher than

773 K with a reactant gas including methane, steam, or oxygen. This behavior is related to the methane combustion activity. The order of the

activity was as follows: Pt/NiO–MgO � Pt/MgO > NiO–MgO. High combustion activity is related to methane activation ability, and this can

make the catalyst activation and reduction easier.

# 2005 Elsevier B.V. All rights reserved.

Keywords: Oxidative steam reforming of methane; Pt; Ni; Self-activation; NiO–MgO solid solution

1. Introduction

Steam reforming of natural gas, whose main component

is methane, on a nickel-based catalyst is one of the important

economic processes for the production of hydrogen and

synthesis gas [1]. The reaction can be described below:

CH4 þ H2O!CO þ 3H2; DH ¼ 206 kJ=mol (1)

Although the steam reforming of methane can give hydro-

gen-rich synthesis gas, the composition of the synthesis gas

can be altered by the water–gas shift reaction or reverse

water gas shift reaction, as shown below:

CO þ H2O!CO2 þ H2; DH ¼ �36 kJ=mol (2)

CO2 þ H2 !CO þ H2O; DH ¼ þ41 kJ=mol (3)

* Corresponding author. Tel.: +81 29 853 5030; fax: +81 29 853 5030.

E-mail address: tomi@tulip.sannet.ne.jp (K. Tomishige).

0926-860X/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.apcata.2005.06.022

The produced hydrogen and synthesis gas can be utilized

for the synthesis of ammonia, methanol, hydrocarbons,

dimethyl ether, acetic acid, and for oxo-syntheses. As

described above (Eq. (1)), steam reforming of methane to

synthesis gas requires heat, since the reaction is highly

endothermic. Therefore, in the conventional steam reform-

ing system, severe external heating has been carried out.

However, the efficiency of the external heating with burners

is not so high. In order to enhance the efficiency, some

researchers have added oxygen to the reactant gas to utilize

the exothermic reactions such as methane combustion

(Eq. (4)) and partial oxidation of methane (Eq. (5)).

CH4 þ 2O2 !CO2 þ 2H2O; DH ¼ �803 kJ=mol (4)

CH4 þ 1=2O2 !CO þ 2H2; DH ¼ �36 kJ=mol (5)

This method corresponds to an internal heating system.

The most famous process of this method is ATR

(autothermal reforming of methane) developed by Haldor

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280 273

Topsøe [2]. This reaction system consists of non-catalytic

partial oxidation + combustion and catalytic methane

reforming with H2O and CO2. In this system, the introduced

oxygen was consumed in the non-catalytic part, and does not

contact with the steam reforming catalyst for the inhibition

of the hot spot formation on the catalyst surface. There are a

lot of reports on catalyst development for the combinations

of catalytic combustion and reforming, which correspond to

oxidative steam reforming of methane [3–13]. When oxygen

is introduced to the steam reforming Ni catalyst with

methane and steam, only methane combustion proceeds near

the catalyst bed inlet, since the rate of methane combustion

can be much higher than the rate of methane reforming. This

makes the bed temperature higher and the hot spots can be

formed. This is a general problem in the combination of

catalytic combustion with reforming for the oxidative steam

reforming process. Recently, our group has reported that the

addition of Pt to Ni catalyst drastically flattens the catalyst

bed temperature gradient in oxidative reforming of methane

[4,14]. This effect can be explained by the inhibition of the

oxidation of Ni, even in the region of O2 presence, by alloy

formation of Pt and Ni.

On the other hand, it has been reported that NiO–MgO

solid solution catalysts exhibited high performance in steam

[15] and dry reforming of methane [16–20]. Especially, the

catalysts had high resistance to coke deposition, which is

one of the most serious problems in the reforming process.

Since NiO–MgO solid solution catalysts have low reduci-

bility [15], it is thought that they are not suitable to oxidative

steam reforming of methane. According to our previous

reports, when methane combustion and reforming can

proceed simultaneously in the same region of the catalyst

bed, the hot spot formation can be effectively inhibited [21].

For this purpose, maintenance of the metallic state of Ni is

thought to be very important.

In this research, we investigated the effect of Pt addition

to NiO–MgO solid solution catalysts in oxidative steam

reforming of methane. Especially, the additive effect of Pt is

to enhance the catalyst reducibility and to promote self-

activation of catalyst, which can be due to high combustion

activity caused by synergy between Pt and NiO–MgO solid

solution.

2. Experimental

2.1. Catalyst preparation

NiO–MgO solid solutions were prepared by the solid

reaction method from NiO (Wako Pure Chemical Industries

Ltd, Japan) and MgO (UBE Material Industries Ltd, Japan).

The mixture of NiO with MgO was calcined in air at 1423 K

for 12 h. The formation of NiO–MgO solid solution was

identified by X-ray diffraction (XRD), as described later.

The compositions of solid solution prepared in this study

were NixMg1�xO where x = 0.05, 0.10, 0.15 and 0.20. After

the calcination, the BET surface areas of NixMg1�xO were

determined to be 2.6–4.1 m2/g. As a reference, MgO was

calcined at the same temperature. The loading of Pt on

NixMg1�xO and MgO was performed by the impregnation

of NixMg1�xO and MgO with Pt(NO2)2(NH3)2 aqueous

solution. After removal of the solvent by heating, the

catalysts were dried overnight at 383 K and then calcined in

air at 773 K for 3 h. The loading amount of Pt was 0–

1.0 mass%. Before use, all of these catalysts were pressed

(500 kg/cm2) into disks and then crushed to 60–100 mesh

particles. These catalysts are denoted as 0.1% Pt/NixMg1�xO,

where the value before Pt represents the loading amount of Pt

as mass%, and ‘x’ means the Ni content as a molar ratio in

NiO–MgO solid solution.

2.2. Catalytic reaction

Oxidative steam reforming of methane was carried out in

a fixed-bed flow reaction system. A quartz tube (inner

diameter: 4.4 mm) was used as the reactor. A thin quartz

tube (inner diameter: 1.5 mm) was used as the thermo-

couple-well. The reaction temperature was monitored by

thermocouples located at the catalyst bed inlet and outlet.

This temperature was controlled on the basis of the inlet

thermocouple. Catalyst reduction was carried out by

hydrogen flow under atmospheric pressure at an appropriate

temperature. The pretreatment conditions are shown in each

result. The partial pressure ratio of reactants was usually

CH4/H2O/O2 = 2/1.5/1, where H2O was supplied by using

the microfeeder. The total pressure was 0.1 MPa; 0.03 g

catalyst was used for each experiment. The length of the

catalyst bed was about 4 mm. Contact time W/F (W = cat-

alyst weight/g and F = total flow rate of the introduced

gases/mol/h) was in the range of 0.13–1.2 g h/mol. The

effluent gas was analyzed with an FID gas chromatograph

(GC) (column packing: Gaskuropack 54) equipped with a

methanator for CH4, CO, CO2; a TCD-GC (column packing:

Molecular Sieve 13X) was used for H2 analysis. An ice bath

was set up between the reactor exit and the sampling port in

order to remove water from the effluent gas used for GC

analysis. Methane conversion and CO selectivity in

oxidative steam reforming of methane were calculated as

described below. The amount of deposited coke during the

reaction can be neglected in all the cases here.

Methane conversion ð%Þ ¼ ðCCO þ CCO2Þ

ðCCH4þ CCO þ CCO2

Þ � 100

CO selectivity ð%Þ ¼ CCO

ðCCO þ CCO2Þ � 100

C, concentration of each component in the effluent gas.

2.3. Catalyst characterization

The surface area of the catalysts was determined by the

BET method by using a Gemini (Micromeritics) instrument.

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280274

Fig. 2. XRD patterns of the samples after the calcination at 1423 K for 12 h:

(a) NiO; (b) MgO; (c) Ni0.05Mg0.95O; (d) Ni0.1Mg0.9O; (e) Ni0.15Mg0.85O;

(f) Ni0.2Mg0.8O.

X-ray diffraction (XRD) was measured by using a Philips

analytical X-ray diffractometer with Cu Ka radiation at

40 kV and 20 mA.

The reducibility of the catalysts was characterized by a

temperature-programmed reduction (TPR) technique.

Before the TPR measurement, the samples (about 0.05 g)

were heated at 773 K for 0.5 h (heating rate 10 K/min) under

He gas flow of 30 ml/min in order to remove any adsorbed

species such as CO2. After the sample was cooled down to

room temperature under He flowing, the reactor was heated

again from room temperature to 1123 K for 0.5 h at a heating

rate of 10 K/min in a 5% H2/Ar mixture gas flow of 30 ml/

min. The temperature was measured by using a thermo-

couple located in the catalyst bed. The effluent gas was

analyzed using TCD-GC.

Each chemisorption experiment was carried out in a high-

vacuum system by a volumetric method. Before adsorption

of H2, the catalysts were treated in H2 at 1073 K for 0.5 h in a

fixed-bed reactor. After this pretreatment, the sample was

transferred to the cell for adsorption measurements under air

atmosphere. Before each measurement, H2 pretreatment at

773 K was carried out for 0.5 h in the cell. After evacuation

at 773 K, the sample was cooled down to room temperature.

The amount of H2 adsorption was determined at room

temperature. The adsorption gas pressure was about

26.3 kPa. The sample weight was about 0.25 g. The dead

volume of the apparatus was 63.5 cm3.

3. Results and discussion

Fig. 1 shows the calcination temperature dependence of

XRD patterns of NiO + MgO physical mixture, whose molar

ratio of Ni to Mg is 1:4, and this corresponds to the raw

materials for Ni0.2Mg0.8O. At a calcination temperature

Fig. 1. Calcination temperature dependence of Ni0.2Mg0.8O: (a)

NiO + MgO physical mixture (without calcination); (b) 773 K for 3 h;

(c) 1073 K for 3 h; (d) 1273 K for 3 h; (e) 1423 K for 12 h. (&) MgO,

(D) NiO, (*) Ni–Mg–O.

below 1073 K, the peaks were almost the same as those of

the physical mixture. This means that the solid reaction

between NiO and MgO did not proceed at 1073 K. The solid

reaction started above 1273 K; after calcination at 1423 K

for 12 h, the two peaks assigned to NiO and to MgO changed

into one peak. The peak was located between the peaks

assigned to NiO and to MgO. This indicates that an NiO–

MgO solid solution was formed. Fig. 2 shows the XRD

patterns of the catalysts with various Ni contents. XRD

patterns of NixMg1�xO (x = 0.05–0.20) were located

between those of MgO and NiO; the diffraction angle

increased with increasing Ni content. The diffraction angles

and lattice constants (d) of the solid solution are listed in

Table 1. From the diffraction angles, the lattice constant can

be calculated. When the Ni content in solid solution

increased, the lattice constant was decreased. From the

comparison of the lattice constants from XRD and the

estimation based on Vegard’s rule [22], where the lattice

constant of a solid solution is represented by using that of

each compound, the lattice constants (d) of solid solutions

agreed well within the experimental errors. These results are

Table 1

XRD analysis of NiO–MgO solid solutions

Catalysts Diffraction angle Lattice constant (d) (nm)

2u (8)a XRDb Vegard’s rulec

NiO 43.26 0.20896 –

MgO 42.88 0.21072 –

Ni0.05Mg0.95O 42.90 0.21063 0.21064

Ni0.1Mg0.9O 42.93 0.21049 0.21055

Ni0.15Mg0.85O 42.96 0.21035 0.21046

Ni0.2Mg0.8O 43.00 0.21016 0.21037a Diffraction peak positions on the basis of Fig. 2.b Lattice constant obtained from XRD result.c Lattice constant calculated from Vegard’s rule [22]: dNixMgð1�xÞO ¼

xdNiO þ ð1 � xÞdMgO.

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280 275

Fig. 3. TPR profiles of the samples: (a) Ni0.05Mg0.95O; (b) Ni0.1Mg0.9O; (c)

Ni0.15Mg0.85O; (d) Ni0.2Mg0.8O; (e) 0.1% Pt/Ni0.05Mg0.95O; (f) 0.1% Pt/

Ni0.1Mg0.9O; (g) 0.1% Pt/Ni0.15Mg0.85O; (h) 0.1% Pt/Ni0.2Mg0.8O.

consistent with the previous report. This indicates the solid

solution formation [15].

Fig. 3 shows the TPR profiles of the NixMg1�xO and

0.1% Pt/NixMg1�xO (x = 0.05–0.20) catalysts. In the case of

NixMg1�xO, hydrogen consumption started at about 623 K.

Total amounts of hydrogen consumption are listed in

Table 2. The amount of hydrogen consumption increased

with increasing Ni content almost proportionally. On the

other hand, the TPR profiles of 0.1% Pt/NixMg1�xO

catalysts were rather similar to that of NixMg1�xO. In the

usual cases, the addition of Pt significantly promoted the

reduction of Ni [23]. However, the effect of Pt addition on

the reducibility of NiO–MgO solid solution was small in

terms of TPR profiles. This is probably because the additive

amount of Pt was as small as 0.1 mass%. As described later,

the effect of Pt addition on the catalyst performance in

oxidative steam reforming of methane is very large.

On the basis of TPR results, it is possible to calculate the

composition ratio of reduced Pt to reduced Ni on Pt/NiO–

Table 2

Properties of NiO–MgO and Pt/NiO–MgO catalysts

Catalysts BET

(m2/g)

Ni content

(10�4 mol/g)

Pt content

(10�6 mol/g)

H2 cons

(10�6 m

Ni0.05Mg0.95O 4.1 11.9 0 34

Ni0.1Mg0.90O 3.6 22.9 0 63

Ni0.15Mg0.85O 3.1 33.0 0 100

Ni0.2Mg0.8O 2.6 42.4 0 146

0.1% Pt/Ni0.05Mg0.95O 4.1 11.9 5.1 42

0.1%Pt/Ni0.1Mg0.9O 3.6 22.9 5.1 78

0.1% Pt/Ni0.15Mg0.85O 3.2 33.0 5.1 111

0.1%Pt/Ni0.2Mg0.8O 2.8 42.4 5.1 149a Total amount of hydrogen consumption in TPR profile (Fig. 3).b Molar ratio of Pt to reduced Ni, assuming the following reactions occur

Pt2+ + H2 ! Pt0 + 2H+.c Reduction degree of Ni: the ratio of reduced Ni to total Ni.d Total amount of hydrogen adsorption at 298 K.e Dispersion of metal: (total H2 adsorption � 2 � 100)/(H2 consumption); assf Calculated by 971/(dispersion � 10) [27].

MgO, as listed in Table 2. The ratio of Pt to reduced Ni is in

the range between 0.04 and 0.14. In addition, from the result

of H2 adsorption, we can estimate the dispersion and particle

size. In the case of solid solution without the Pt addition, the

dispersion decreased with increasing Ni content. This

tendency is also observed on Pt/NiO–MgO catalysts;

however, the addition of Pt enhanced the dispersion of

metal particles. From the estimation of metal particle size, it

is expected that XRD peaks can be observed. However, the

peaks due to Ni metal or Pt–Ni alloy were not observed,

although the XRD patterns of the samples after H2 reduction

are not shown here. This is because the amount of reduced

Ni is very small (about 3%). In addition, according to our

previous report [24], Pt–Ni alloy is formed after the

reduction over Pt/NiO–MgO catalysts. The formation of

bimetallic particles on Pt/NiO–MgO was recognized by

transmission electron microscopy combined with energy

dispersive X-ray spectroscopy. By this method, it was found

that each metal particle contained both Pt and Ni [24].

Therefore, we think that the behavior of alloy formation can

be true for this case.

Fig. 4 shows the effect of the contact time (W/F) on the

methane conversion and H2/CO ratio over NiO–MgO

catalysts with various Ni contents after reduction pretreat-

ment at 1073 K. At all W/F conditions, methane conversion

was higher on the catalyst with larger Ni content. The dotted

line represents methane conversion due to methane

combustion under CH4/H2O/O2 = 2/1.5/1. At low W/F,

such as 0.13 g h/mol, methane conversion was as low as this

level, and the formation rates of hydrogen and H2/CO were

smaller. This is related to the oxidation of Ni at low W/F

conditions. In addition, the equilibrium level of methane

conversion was almost 100% and the H2/CO ratio in the

produced gas was approximately 3.

Fig. 5 shows W/F dependences of the methane

conversion and H2/CO ratio in oxidative steam reforming

of methane over 0.1% Pt/NiO–MgO and 0.1% Pt/MgO

catalysts. Results show that methane conversion over 0.1%

a

ol/g)

Pt/Ni0b Dredc

(%)

H2 add

(10�6 mol/g)

Ddispe

(%)

Particle sizef

(nm)

0 2.8 1.6 9.4 10.3

0 2.8 2.8 8.9 10.9

0 3.0 4.0 8.0 12.2

0 3.4 5.7 7.8 12.4

0.14 3.1 4.5 21.3 4.6

0.07 3.2 6.3 16.1 6.0

0.05 3.2 7.4 13.3 7.3

0.04 3.4 9.0 12.0 8.1

during the reduction at 1123 K in TPR profiles: Ni2+ + H2 ! Ni0 + 2H+,

uming that all Pt atoms were reduced.

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280276

Fig. 4. Contact time dependence of methane conversion and H2/CO ratio in oxidative steam reforming of methane over NiO–MgO catalysts with various Ni

contents: (D) Ni0.2Mg0.8O; (&) Ni0.15Mg0.85O; (*) Ni0.1Mg0.9O; (^) Ni0.05Mg0.95O. Reaction conditions: catalyst weight: 0.03 g, CH4/H2O/O2 = 2/1.5/1,

reaction temperature: 1073 K, total pressure: 0.1 MPa, H2 reduction pretreatment: 1073 K for 0.5 h. The dotted line represents methane conversion due to

methane combustion.

Pt/MgO was far from the equilibrium level even at high W/F

(1.2 g h/mol). This indicates that 0.1% Pt/MgO had low

activity in oxidative steam reforming of methane. The

comparison at W/F = 1.2 g h/mol shows that the methane

conversion over 0.1% Pt/MgO was almost the same as that

over Ni0.05Mg0.95O (Fig. 4). However, the comparison at low

W/F shows that the conversion over 0.1% Pt/MgO was much

higher than that over Ni0.05Mg0.95O. In the oxidative

reforming of methane, methane conversion is dependent on

the amount of reduced catalyst and on its performance.

Under the presence of oxygen in the gas phase, metal species

can be oxidized and can lose the reforming activity. High

conversion over 0.1% Pt/MgO is related to high catalyst

reducibility. Pt catalysts usually exhibited much higher

reducibility than Ni catalysts [25]. In the case of

Ni0.05Mg0.95O, Ni species on the catalyst can be oxidized

gradually with decreasing W/F. On the other hand, it has

been reported that Pt catalysts exhibited high performance in

oxidative steam reforming of methane [3]. In these results,

the performance of 0.1% Pt/MgO was rather low; this can be

explained by the low loading of Pt (0.1 mass%). In contrast,

all the 0.1% Pt/NiO–MgO catalysts gave very high methane

Fig. 5. Contact time dependence of methane conversion and H2/CO ratio in oxidat

catalysts: (~) 0.1% Pt/Ni0.2Mg0.8O; (&) 0.1% Pt/Ni0.15Mg0.85O; (*) 0.1% P

conditions: catalyst weight: 0.03 g, CH4/H2O/O2 = 2/1.5/1, reaction temperature: 1

The dotted line represents methane conversion due to methane combustion.

conversion values, which reached the equilibrium level even

at W/F = 0.43 g h/mol. The promoting effect of Pt addition

to NiO–MgO solid solution with all the Ni contents was

clearly observed. It has been reported that the addition of Pt

to NiO–MgO solid solution catalysts promoted the

performance in dry reforming of methane [24]. The methane

conversion of 0.1% Pt/NiO–MgO was much higher than that

over NiO–MgO and 0.1% Pt/MgO. According to the TPR

profiles, the additive effect of 0.1 mass% Pt was not

observed clearly. On the other hand, the effect of Pt was

observed very clearly in oxidative steam reforming of

methane.

Fig. 6 represents the effect of Pt loading amount over Pt/

Ni0.2Mg0.8O catalysts in oxidative steam reforming of

methane. As shown in Fig. 5, methane conversion reached

almost the equilibrium level at higher W/F conditions.

Therefore, the results in Fig. 6 are based on the data at W/

F = 0.13 g h/mol. On the catalyst without Pt addition, which

corresponds to Ni0.2Mg0.8O, low methane conversion was

obtained; the value was almost the same level as the

contribution of methane combustion. On the other hand, by

the addition of a small amount of Pt such as 0.03 mass%,

ive steam reforming of methane over 0.1% Pt/NiO–MgO and 0.1% Pt/MgO

t/Ni0.1Mg0.9O; (^) 0.1% Pt/Ni0.05Mg0.95O; (�) 0.1% Pt/MgO. Reaction

073 K, total pressure: 0.1 MPa, H2 reduction pretreatment: 1073 K for 0.5 h.

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280 277

Fig. 7. Reduction temperature dependence of CH4 conversion, CO selec-

tivity and H2/CO ratio in oxidative steam reforming of methane over the

catalysts at reaction temperature of 1073 K: (a) 0.1% Pt/ Ni0.2Mg0.8O; (b)

0.1% Pt/MgO; (c) Ni0.2Mg0.8O. (~) CH4 conversion, (*) CO selectivity,

(&) H2/CO. Reaction conditions: catalyst weight: 0.03 g, CH4/H2O/O2 = 2/

1.5/1, W/F = 0.43 g h/mol, total pressure: 0.1 MPa. The dotted line repre-

sents methane conversion due to methane combustion.

Fig. 6. Effect of Pt loading over Pt/Ni0.2Mg0.8O catalysts in oxidative steam

reforming of methane: (~) CH4 conversion; (*) CO selectivity; (&) H2/

CO ratio. Reaction conditions: catalyst weight: 0.03 g, CH4/H2O/O2 = 2/

1.5/1, W/F = 0.13 g h/mol, reaction temperature: 1073 K, total pressure:

0.1 MPa, H2 reduction pretreatment: 1073 K for 0.5 h. The dotted line

represents methane conversion due to methane combustion.

methane conversion increased to about 67%. Methane

conversion over Pt/Ni0.2Mg0.8O increased gradually with

increasing Pt content, although the dependence of the

performance on Pt loading is rather small. The presence of Pt

is very important for the promotion, and the effect of Pt

addition was almost saturated at 0.1 mass%.

Fig. 7 shows the reduction temperature dependence in

oxidative steam reforming of methane over 0.1% Pt/

Ni0.2Mg0.8O, 0.1% Pt/MgO and Ni0.2Mg0.8O catalysts at

1073 K. Even when the reduction temperature was as low as

773 K, 0.1% Pt/Ni0.2Mg0.8O exhibited almost 100%

methane conversion. The hydrogen reduction at 773 K

was enough to activate 0.1% Pt/Ni0.2Mg0.8O. Similar

behavior was observed over 0.1% Pt/MgO, although CH4

conversion over 0.1% Pt/MgO was much lower than that

over 0.1% Pt/Ni0.2Mg0.8O. This also indicates that H2

reduction at 773 K is enough to activate 0.1% Pt/MgO. In

contrast, significant reduction temperature dependence over

Ni0.2Mg0.8O was clearly observed. Ni0.2Mg0.8O required a

reduction temperature higher than 973 K. This result

indicates that reduced Ni0.2Mg0.8O solid solution has high

activity in oxidative reforming of methane. However,

Ni0.2Mg0.8O, which is not reduced enough, was active for

only methane combustion. The reduction treatment for

catalyst activation can strongly influence the catalyst

performance in oxidative steam reforming of methane.

Fig. 8 shows the dependence of Pt addition on the catalyst

activation in oxidative steam reforming of methane. In a

practical process, the catalyst pretreatment like H2 reduction

is impossible. Therefore, the catalyst should be reduced and

activated by methane, oxygen and steam. In the results

shown above, the activity test was carried out over the

catalyst after H2 reduction. In this experiment, the reactant

gas (CH4/H2O/O2 = 2/1.5/1) was introduced onto the

catalyst without H2 pretreatment. At first, the catalyst was

heated at 773 K under Ar flowing, and then the reactant

gases (CH4 + H2O + O2) for oxidative steam reforming of

methane were introduced at 773 K. The performance was

observed at the reaction temperature for 30 min. After

30 min, the temperature was increased to 823 K. Further-

more, the catalyst was heated stepwise up to 1073 K in 50 K

steps. At each temperature, the performance was observed

for 30 min. In the case of 0.1% Pt/NixMg1�xO (x = 0.05,

0.20) as shown in Fig. 8(a) and (c), methane conversion was

25% and the main carbon-containing product was CO2 at

773 K. This result indicates that only the methane

combustion reaction proceeded. This can be supported by

the absence of any H2 formation. At a reaction temperature

higher than 773 K, methane conversion was beyond 25%,

and it increased with increasing the reaction temperature. At

the same time, CO selectivity also increased and the H2/CO

ratio decreased. Such behaviors indicate that the reforming

reaction proceeded and that the synthesis gas was formed

under these reaction conditions. Methane conversion over

0.1% Pt/Ni0.2Mg0.8O was higher than that over 0.1% Pt/

Ni0.05Mg0.95O. On the other hand, in the case of 0.1% Pt/

MgO (Fig. 8(e)), in the range of 773–973 K, methane

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280278

Fig. 8. Results of CH4 conversion, CO selectivity and H2/CO ratio in oxidative steam reforming of methane during the reaction temperature increasing over the

catalysts without H2 pretreatment: (a) 0.1% Pt/Ni0.05Mg0.95O; (b) Ni0.05Mg0.95O; (c) 0.1% Pt/Ni0.2Mg0.8O; (d) Ni0.2Mg0.8O; (e) 0.1% Pt/MgO. (~) CH4

conversion, (*) CO selectivity, (&) H2/CO, (�) equilibrium level of CH4 conversion at each reaction temperature. Reaction conditions: catalyst weight: 0.03 g,

CH4/H2O/O2 = 2/1.5/1, W/F = 0.43 g h/mol, total pressure: 0.1 MPa. The dotted line represents methane conversion due to methane combustion.

conversion was much lower than 25%, and CO selectivity was

also low. This indicates that 0.1% Pt/MgO has low activity of

methane combustion and reforming. However, at 1023 K,

methane conversion reached about 50%, although it was

lower than that over 0.1% Pt/NixMg1�xO catalysts. In

addition, in the case of NixMg1�xO (Fig. 8(b) and (d)), in

all the temperature ranges, methane conversion did not reach

25%, and CO selectivity and H2 formation were also very low.

From the comparison in methane conversion, Ni0.2Mg0.8O

catalyst had higher combustion activity than Ni0.05Mg0.95O,

although the combustion activity of these catalysts was much

lower than Pt/MgO and Pt/NiO–MgO under the conditions of

oxidative steam reforming of methane.

Fig. 9 shows the behaviors during the process of decreasing

temperature in oxidative steam reforming of methane over

Ni0.2Mg0.8O and 0.1% Pt/Ni0.2Mg0.8O catalysts. In this

experiment, the reactant gas (CH4/H2O/O2 = 2/1.5/1) was

introduced to the catalyst bed after H2 reduction at 1073 K for

0.5 h. In the case of 0.1% Pt/Ni0.2Mg0.8O in Fig. 9(a), methane

conversion decreased with decreasing the reaction tempera-

ture. Methane conversion at each temperature in the cooling

profile (Fig. 9(a)) was only a little higher than that in the

heating profile (Fig. 8(c)). This indicates that the catalyst in

the cooling profiles can be activated a little more than that in

the heating procedure under the reactant gas flowing

(Fig. 8(c)). In contrast, in the case of Ni0.2Mg0.8O

(Fig. 9(b)), methane conversion was very high at 1073 K;

however, when the reaction temperature decreased, methane

conversion sharply decreased below combustion level. This

means that Ni cannot be maintained in the reduced state at a

temperature lower than 1073 K even when Ni0.2Mg0.8O was

reduced sufficiently with H2 pretreatment. The difference

between heating and cooling profiles in Figs. 8(d) and 9(b) is

so small in all the reaction temperatures, except at 1073 K.

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280 279

Fig. 9. Results of CH4 conversion, CO selectivity and H2/CO ratio in

oxidative steam reforming of methane during the reaction temperature

decreasing over the catalysts with H2 pretreatment: (a) 0.1% Pt/

Ni0.2Mg0.8O; (b) Ni0.2Mg0.8O. (~) CH4 conversion, (*) CO selectivity,

(&) H2/CO. Reaction conditions: catalyst weight: 0.03 g, CH4/H2O/O2 = 2/

1.5/1, W/F = 0.43 g h/mol, total pressure: 0.1 MPa, H2 reduction pretreat-

ment: 1073 K for 0.5 h. The dotted line represents methane conversion due

to methane combustion.

Fig. 10 shows the methane combustion activity of 0.1%

Pt/Ni0.2Mg0.8O, 0.1% Pt/MgO and Ni0.2Mg0.8O catalysts on

CH4/H2O/O2/Ar = 1/1.5/2/3 without H2 pretreatment. In this

experiment, stoichiometric methane and O2 were supplied

under the presence of H2O. It is found that the order of

Fig. 10. Comparison of methane combustion activity under CH4/H2O/O2/

Ar = 1/1.5/2/3: (D) 0.1% Pt/Ni0.2Mg0.8O; (*) 0.1% Pt/MgO; (&)

Ni0.2Mg0.8O. Reaction conditions: catalyst weight: 0.03 g, W/F = 0.43 g h/

mol, total pressure: 0.1MPa, catalysts used without H2 pretreatment.

methane combustion activity was as follows: 0.1% Pt/

Ni0.2Mg0.8O � 0.1% Pt/MgO > Ni0.2Mg0.8O.

From the comparisons of performance of these three

catalysts, we can conclude that the additive effect of Pt on

NiO–MgO solid solution is to promote the activity of

methane combustion and methane reforming. According to

the previous report on the structure of MgO supported on Pt

by means of the extended X-ray absorption fine structure

(EXAFS), Pt ions can be incorporated to the lattice of MgO

with the substitution of Mg ion site in an oxidized state [26].

Therefore, it is expected that Pt ions can be incorporated into

the lattice of NiO–MgO solid solution in this case. In order

to discuss the activity of methane combustion, one must

consider the reactivity of the oxide ions (O2�). In the case of

0.1% Pt/MgO, there are oxide ions between Pt and Mg ions,

which is denoted as Pt–O–Mg. In the case of Ni0.2Mg0.8O,

the Ni–O–Mg structure is present. From the result of the

combustion activity test, it is thought that these structures

can give low combustion activity. In contrast, Pt–O–Ni

structure is present on 0.1% Pt/Ni0.2Mg0.8O as well as Pt–O–

Mg and Ni–O–Mg structures. It is expected that this can

contribute to high methane combustion activity, although

further investigations are necessary for the identification of

the active sites for methane combustion over 0.1% Pt/

Ni0.2Mg0.8O.

4. Conclusions

(1) In oxidative steam reforming of methane, NixMg1�xO

solid solution catalysts showed high methane conver-

sion at high W/F condition such as 1.2 g h/mol;

however, they gave low methane conversion and syngas

formation at low W/F condition. This is due to the

deactivation by oxidation.

(2) T

he catalytic activity of NiO–MgO solid solution in

oxidative steam reforming of methane was improved

significantly by the addition of small amounts of Pt.

0.1% Pt/NiO–MgO catalysts exhibited much higher

performance than NiO–MgO and Pt/MgO, especially at

low W/F condition. This indicates that the synergy

between Pt and NiO–MgO enhances the performance.

In addition, the reduction temperature of H2 pretreat-

ment decreased with the addition of Pt in solid solution

because of the high reducibility of Pt.

(3) H

igh temperature hydrogen reduction was necessary for

the activation of NiO–MgO catalysts. In contrast, the

reactant gases for oxidative steam reforming of methane

above 773 K can activate Pt/NiO–MgO catalysts. This

behavior is related to the high methane combustion and

the high catalyst reducibility.

(4) F

rom the characterization results, we can conclude that

the addition of Pt did not enhance the reduction degree

of Ni so significantly; however, it increased the metal

dispersion remarkably. This suggests the formation of

M. Nurunnabi et al. / Applied Catalysis A: General 292 (2005) 272–280280

Pt–Ni alloy, which can contribute to the promoting

effect in oxidative steam reforming of methane.

References

[1] J.R. Rostrup-Nielsen, in: J.R. Anderson, M. Boudart (Eds.), Catalysis

Science and Technology, vol. 5, Springer-Verlag, New York, 1984, p.

1.

[2] J.R. Rostrup-Nielsen, Sekiyu Gakkaishi 40 (1997) 366.

[3] B. Li, K. Maruyama, M. Nurunnabi, K. Kunimori, K. Tomishige,

Appl. Catal. A: Gen. 275 (2004) 157.

[4] K. Tomishige, S. Kanazawa, S. Ito, K. Kunimori, Appl. Catal. A: Gen.

244 (2003) 71.

[5] K. Tomishige, Y. Matsuo, Y. Yoshinaga, Y. Sekine, M. Asadullah, K.

Fujimoto, Appl. Catal. A: Gen. 223 (2002) 225.

[6] K. Takehira, Catal. Surv. Jpn. 6 (2002) 19.

[7] S. Liu, G. Xiong, H. Dong, W. Yang, Appl. Catal. A: Gen. 202 (2000)

141.

[8] A.M. O’Connor, J.R.H. Ross, Catal. Today 46 (1998) 203.

[9] P.D.F. Vernon, M.L.H. Green, A.K. Cheetham, A.T. Ashcroft, Catal.

Today 13 (1992) 417.

[10] V.R. Choudhary, A.M. Rajput, B. Prabhakar, Catal. Lett. 32 (1995)

391.

[11] T. Inui, K. Saigo, Y. Fujii, K. Fujioka, Catal. Today 26 (1995) 295.

[12] V.R. Choudhary, A.M. Rajput, Ind. Eng. Chem. Res. 35 (1996)

3934.

[13] K. Tomishige, M. Nurunnabi, K. Maruyama, K. Kunimori, Fuel

Process. Technol. 85 (2004) 1103.

[14] K. Tomishige, S. Kanazawa, M. Sato, K. Ikushima, K. Kunimori,

Catal. Lett. 84 (2002) 69.

[15] O. Yamazaki, K. Tomishige, K. Fujimoto, Appl. Catal. A: Gen. 136

(1996) 49.

[16] K. Tomishige, Y.G. Chen, K. Fujimoto, J. Catal. 181 (1999) 91.

[17] Y.G. Chen, K. Tomishige, K. Yokoyama, K. Fujimoto, J. Catal. 184

(1999) 479.

[18] Y.H. Hu, E. Ruchenstein, Catal. Rev. Sci. Eng. 44 (2002) 423.

[19] K. Tomishige, Y. Himeno, Y. Matsuo, Y. Yoshinaga, K. Fujimoto, Ind.

Eng. Chem. Res. 39 (2000) 1891.

[20] K. Tomishige, Catal. Today 89 (2004) 405.

[21] K. Tomishige, S. Kanazawa, K. Suzuki, M. Asadullah, M. Sato, K.

Ikushima, K. Kunimori, Appl. Catal. A: Gen. 233 (2002) 35.

[22] L. Vegard, Z. Phys. 5 (1921) 17.

[23] J. Rynkowski, D. Rajski, I. Szyszka, J.R. Grzechowiak, Catal. Today

90 (2004) 159.

[24] Y.G. Chen, K. Tomishige, K. Yokoyama, K. Fujimoto, Appl. Catal. A:

Gen. 165 (1997) 335.

[25] T.B. Reed, Free Energy Formation of Binary Compounds, MIT Press,

Cambridge, MA, 1971, p. 67.

[26] K. Tomishige, Y. Nagasawa, U. Lee, Y. Iwasawa, Bull. Chem. Soc.

Jpn. 70 (1997) 1607.

[27] C.W. Bartholomew, R.B. Pannell, J.L. Butler, J. Catal. 65 (1980) 335.