4864 Ind. Eng. Chem. Res.

2004, 43, 4864-4870

Kinetic Studies on Catalytic Decomposition of Methane to Hydrogen

and Carbon over Ni/TiO2 Catalyst
Sharif Hussein Sharif Zein, Abdul Rahman Mohamed,* and P. Sesha Talpa Sai
School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia,
14300 Nibong Tebal, S.P.S, Pulau Pinang, Malaysia

Experiments were conducted in a fixed-bed catalytic reactor to study the decomposition of

methane to hydrogen and carbon over 13 wt % Ni/TiO2-based catalyst. The variables include
the volumetric flow rate of reactant (relative partial pressures of methane and argon of 0.5 each)
(100 < v0 < 450 mL/min) and reaction temperature (823 < T < 1173 K). The experimental data
of conversion versus time were subjected to the integral method of analysis, and the rate law of
the decomposition reaction was found to be first-order. An activation energy of 60 kJ/mol was
obtained. A possible reaction sequence for methane decomposition was validated against experi-
mental data. An analysis of the rate equations indicated that the adsorption of methane on the
surface is the rate-controlling step. This observation was same as that obtained from an analysis
of data by both integral and differential methods. Using the design equation for a packed plug-
flow reactor, predictions were made to obtain the conversion profiles at various operating condi-
tions. Good agreement was obtained upon comparison of the experimental and simulated data.

Introduction In this paper, kinetic studies were carried out using

13 wt % Ni/TiO2-based catalyst for the catalytic decom-
Hydrogen is predicted to become an efficient source position of methane for the production of hydrogen. The
of clean fuel and has been described as a long-term rate equation was obtained and the activation energy
replacement for natural gas. With increasing interest was estimated. A model, which uses the obtained rate
in fuel cells and progress in hydrogen storage technolo- equation for the decomposition of methane to hydrogen,
gies, the use of hydrogen as a source of clean fuel is was simulated to predict the methane conversion pro-
likely to increase in the future.1 There is substantial files.
interest in developing novel methods for the production
of hydrogen, and one such process is the catalytic
decomposition of methane.2-8 Experimental Section
Many experimental studies have been conducted to Methane decomposition was studied using 13 wt %
gain a better understanding of the kinetics of methane Ni catalyst doped onto different supports, viz., TiO2,
decomposition. The majority of the research has been Al2O3, MgO, and SiO2. The catalysts were prepared
focused on determination of the reaction mechanism, using the impregnation method. Desired amounts of the
identification of the primary decomposition products, transition metal oxides were dissolved in deionized
and estimation of the activation energy for methane water. The dopant concentrations are actually relative
decomposition. The activation energy for methane de- to the molar quantity of the support. The resulting paste
composition using a tubular reactor without catalyst was then dried in an oven and subsequently calcined
was found 370 kJ/mol over the temperature range in a ceramic crucible at 900 C. The catalysts were then
1770-2270 K.9 Steinberg10 reported an activation en- sieved to a size of 400-500 m.
ergy of 131 kJ/mol for methane decomposition over the The experiments were carried out at atmospheric
temperature range 973-1173 K, which is considerably pressure in a stainless steel fixed-bed reactor system.
lower than the value for methane decomposition without A schematic diagram of the reactor system is shown in
catalyst. However, he performed the reaction at pres- Figure 1. The reactor was fabricated from a stainless
sures of 28-56 atm. Dahl et al.11 reported an activation steel tube (o.d. ) 12.7 mm, i.d. ) 10.92 mm, and length
energy of 208kJ/mol for methane decomposition using ) 600 mm). A thermocouple of type K in an inconel tube,
a fluid-wall graphite aerosol flow reactor over the 3 mm in diameter and 600 mm long, was used to
temperature range 1533 < T < 2144 K. measure the temperature of the catalyst bed in the
The activation energy of the decomposition of meth- reactor. The catalyst layer was situated in the center
ane with a mixture of hydrogen over nickel catalysts of the reactor. The free space before and after the
was found to be 90 kJ/mol over the temperature range catalyst layer was filled with quartz particles (RDH) to
723-863 K12 and 97 kJ/mol over the temperature range minimize the reactor dead volume. The furnace used
803-863 K.13 Muradov14 studied the decomposition of was a single zone (model Carbolite VST 11) with a
methane over carbon catalysts, reporting activation temperature controller and was supplied by Carbolite,
energies of 236 kJ/mol (valid for a temperature range Hope Valley, U.K. A pressure gauge (Ashcroft, Stratford,
of 1023-1223 K) and 201 kJ/mol (valid for a tempera- CT) located just above the reactor was used to read the
ture range of 873-893 K) for a carbon black catalyst inlet pressure.
and an activated carbon catalyst, respectively. Methane (supplied by Malaysian Oxygen Sdn. Bhd.)
with a purity of 99.999% and argon (supplied by Sitt
* To whom correspondence should be addressed. E-mail: Tatt Industrial Gases Sdn Bhd.) with a purity of
[email protected] 99.999% were mixed in ratio of 1:1 before entering the
10.1021/ie034208f CCC: $27.50 2004 American Chemical Society
Published on Web 07/08/2004
 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004 4865

Figure 1. Schematic diagram of the reactor system.

reactor, i.e., a mole fraction of 0.5 for methane was used column. The gas chromatograph injector temperature
for all of the experiments. Argon was used as a diluent was set at 313 K. The initial and the final temperatures
as nitrogen might react with the hydrogen at high of the oven were set to 313 and 473 K, respectively. A
temperatures. The flow of methane was regulated using heating rate of 5 K/min was used. The detector tem-
a mass flow controller (MKS), and the argon flow was perature was kept at 473 K. Pure argon gas (99.999%)
regulated with Brooks mass flow controllers (model was used as a carrier gas. The total analysis time was
5850E). The outlet gas flow was monitored by a gas flow 25 min for each injection. Standard gas was injected into
meter (Alexander Wright DM3 B). the gas chromatograph, and the area of each of the
For the kinetics study, a total of 64 experiments were components in the standard gas was determined. The
conducted with fresh catalyst in each run. One gram of standard gas mixture was supplied by BOC Gases,
catalyst was packed in the reactor. The furnace was Guilford, U.K. Chromatograms of hydrocarbons such as
switched on, and pure argon was allowed into the methane, ethylene, ethane, and propylene were obtained
reactor to create an inert atmosphere in the reactor. To using a flame ionization detector (FID), whereas hydro-
activate the fresh catalyst, it was subjected to methane gen, oxygen, carbon monoxid,e and carbon dioxide were
for 5 min at 998 K and 2700 h-1 gas hourly space detected using a thermal conductivity detector (TCD).
velocity (GHSV). When the furnace had attained the A carbon balance of 100 ( 2% was obtained for every
required temperature, methane of known concentration run over the catalyst.
was admitted into the reactor. The concentrations of the Pore volume and surface area measurements of the
product gases were analyzed, and the experiment was different samples were determined via nitrogen adsorp-
stopped when the concentration of the product gases tion/desorption isotherms at liquid nitrogen tempera-
attained a steady value. The experiment was repeated ture (77 K) using an automated gas sorption system
for different volumetric flow rates (100-450 mL/min) (Autosorb I, QuantaChrome Corporation, Boynton Beach,
and temperatures (823-1173 K). FL). All samples were degassed at a temperature of 573
The product gases were analyzed using an on-line gas K for 3 h prior to the measurements. Computer pro-
chromatograph (GC) (Hewlett-Packard Series 6890,
grams (Micropore, version 2.46) allowed for rapid nu-
Hewlett-Packard, Wilmington, DE). The GC was con-
merical results for the surface area and pore texture
trolled on-line using HP ChemStation Rev. A. 06.01.-
from the adsorption-desorption isotherm. Spent cata-
[403] software. Porapaq N and molecular sieve 5 (1/
lysts were analyzed using a transmission electron
8-in. diameter, 6-ft length) stainless steel columns,
microscope (Philips TEM CM12). In preparation for
situated in series with the Porapaq N column located
TEM experiments, a few samples of the spent catalyst
in front were used. The Porapaq N column was used to
were dispersed in distilled water, and then a drop of
separate carbon dioxide, ethane, ethylene, and propy-
the dispersion was deposited on a coated copper grid.
lene and the molecular sieve 5 column for hydrogen,
oxygen, carbon monoxide, nitrogen, and methane. Be-
cause higher hydrocarbons and carbon dioxide can ruin Results and Discussion
the molecular sieve 5 column, two valves operated at
temperature 333 K were used to control the outlet gas A blank reaction (i.e., decomposition of methane
from the Porapaq N column to the detectors to avoid without catalyst) study was carried out in the stainless
passing it through the molecular sieve 5 column. steel reactor filled with quartz sand with a feed consist-
Valve 1 functioned as a sampling mechanism, and valve ing of methane and argon in a ratio of 1:1 at a total
2 was used to control the flow of the product through flow rate of 80 mL/min and in the temperature range
the molecular sieve 5 column. When the valve is off, of 673-1173 K (the range used in the present study).
gas is allowed to flow through the molecular sieve 5 The conversion of methane was less than 1% in all of
4866 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004

Table 1. Physical Properties of the Fresh Ni Supported Table 3. Repeatability and Analytical Reproducibility of
on Different Supports the Catalytic Data Obtained in the Present Study at 998
K for 60 Min on Stream Using 13 wt % Ni on TiO2
surface area total pore volume
catalyst (m2/g) (Vp) (cm3/g) trial conversion (%)
Ni on SiO2 11.45 0.003 1 61.34
Ni on Al2O3 5.28 0.010 2 60.08
Ni on TiO2 8.05 0.005 3 62.11
Ni on MgO 7.53 0.007
and only hydrogen was detected because the catalyst
Table 2. Details of Conversions Obtained in the Present
Study with Different Supports at 998 K and 2700 h-1 was reduced further. After 5 min on stream, hydrogen
GHSV and methane were the only components detected in the
product stream for the all types of catalysts. The ratio
conversion (%)
of methane conversion and hydrogen formation was
catalyst 2 min 5 min 60 min 120 min found to be 1:2. In fact, hydrogen was the only gas
13 wt % Ni on SiO2 97.41 74.12 48.67 PBa detected after 5 min on stream. The methane decom-
13 wt % Ni on Al2O3 99.54 72.98 PB PB position obtained within the first 5 min of reaction
13 wt % Ni on TiO2 99.18 64.57 61.34 62.44 decreased in the order SiO2 > Al2O3 > TiO2 > MgO
13 wt % Ni on MgO 99.23 52.23 15.84 8.05 support. After 60 min on stream, the order of activity
a PB ) pressure buildup. changed to TiO2 > SiO2 > MgO support, and a pressure
buildup (PB) was observed on the Al2O3 support. At 120
these experiments, indicating that the decomposition of min on stream, the TiO2 system maintained its activity,
methane without catalyst in the temperature range of and the MgO decreased from 16 to 8%, whereas a
673-1173 K was negligible. Thus, the material stainless pressure buildup occurred in the SiO2-supported cata-
steel of type 316 used in this study is a suitable material lyst system. Thus, it can be concluded that TiO2 is the
for the construction of a reactor for the decomposition best support. This is in line with the observations of
of methane. Takenaka et al.,17,18 who concluded that TiO2 is the best
Selection of Support Material. Table 1 shows the support for either repeated cycles of methane decom-
physical properties of 13 wt % Ni/TiO2 doped onto position or high catalytic activity of Ni/TiO2 for repeated
different supports, viz., SiO2, Al2O3, TiO2, and MgO. The reactions. The results of the present study also confirm
surface areas of Ni on SiO2, Al2O3, TiO2, and MgO were that deactivation is not occurring up to 120 min on
11.45, 5.28, 8.05, and 7.53 m2/g, respectively. The total stream over TiO2-supported Ni catalyst. Because fresh
pore volumes show the same trend as the surface areas catalyst was used each time, the deactivation can be
with values of 0.003, 0.010, 0.005, and 0.007 cm3/g for neglected in the present study. Carbon samples obtained
Ni on SiO2, Al2O3, TiO2, and MgO, respectively. This from Ni catalysts loaded onto various supports were
can be attributed to the change of the structure or further studied using TEM. The results obtained showed
electronic state of the metal species due to the interac- that the introduction of the support markedly influenced
tion with the supports.15 the carbon morphology. The TEM micrographs of carbon
To select a suitable support material, experiments synthesized on Ni supported on SiO2, Al2O3, TiO2, and
were conducted for methane dissociation into hydrogen MgO support are shown in parts a-d, respectively, of
and carbon with 13 wt % NiO catalyst supported on Figure 2. The TEM image of the carbon formed on Ni
different supports, viz., TiO2, Al2O3, MgO and SiO2, at on SiO2 and Al2O3 supports seems to be undeveloped
998 K and 2700 h-1 gas hourly space velocity (GHSV). carbon nanotubes, whereas the carbon formed on Ni on
The results obtained are listed in Table 2. After 2 min MgO looks like small tubes connected to each other. On
on stream, methane conversion of 100, 100, 100, and the other hand, the TEM image of the carbon formed
97% and COx yield of 60, 62, 65, and 62% were using Ni on the TiO2 support (Figure 2c) was the best
obtained using 13 wt % Ni/SiO2, 13 wt % Ni /Al2O3, 13 among the supported catalysts. Thus, the catalytic
wt % Ni/TiO2, and 13 wt % Ni/MgO, respectively. The activity depended strongly the kind of support.
rest of the yield was hydrogen. No higher hydrocarbons To test the repeatability and analytical reproducibility
were detected. The COx production probably occurs via of the raw catalytic data, activity tests for the newly
the reaction of chemisorbed carbon and hydrogen, prepared 13 wt % Ni on TiO2 catalyst were conducted
produced by the decomposition of methane with the three times at 998 K for 60 min on stream using new
lattice of metal oxides.4 Methane reactions were con- catalyst each time, and the results are reported in Table
ducted over catalysts without any added oxygen. There- 3. The ratio of methane conversion to hydrogen forma-
fore, any methane oxidation in these experiments must tion was found to be 1:2. In fact, hydrogen was the only
be derived from oxygen originating from the catalyst. gas detected. The results indicate that the reproduc-
The surface oxygen present on the catalyst was respon- ibility of the data is satisfactory.
sible for the formation of COx as no oxygen was present Kinetics of the Reaction. Among the catalysts
in the feed gas. The reaction takes place between gas- studied in this work, 13 wt % Ni on TiO2 support was
phase or weakly adsorbed methane and strongly ad- found to be the most effective catalyst for the production
sorbed oxygen or lattice oxygen. Surface OH groups and/ of hydrogen and carbon through the decomposition of
or oxygen from the support and/or promoter and/or the methane. Kinetic studies were carried out using this
active metal, in this case NiO, might have participated catalyst. The experimental program for obtaining the
in the conversion of methane to COx. Previously, it was kinetics of the reaction consisted of runs at different
concluded that the surface OH groups were involved in mole ratios and temperatures with a constant weight
the conversion of methane to CO.16 On the basis of Table of the 13 wt % Ni/TiO2, and the results are presented
2, it was estimated that 2-4 min is sufficient for in Table 4. The conversion of methane versus time data
reducing the fresh catalyst. As the time increased, the were subjected to the integral method for the derivation
number of methane pulses over the catalyst increased, of a suitable rate law.
 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004 4867

Figure 2. Transmission electron microscope images of the carbon produced on Ni supported on the different supports (a) SiO2, (b) Al2O3,
(c) TiO2, and (d) MgO at 998 K and GHSV of 2700 h-1.

Table 4. Details of Methane Conversions (%) Obtained in

the Present Study at Various Operating Conditionsa

temperature volumetric flow rate (mL/min)

(K) 100 150 200 250 350 450
823 15.82 11.42 7.04 6.52 - -
873 27.41 20.65 16.55 14.03 7.03 -
923 41.56 33.81 27.43 24.87 16.21 9.87
973 53.99 45.10 37.21 31.58 24.99 17.88
1023 59.07 49.66 39.48 33.24 26.04 21.97
1073 70.54 59.63 49.02 35.10 24.43 21.07
1123 82.49 69.08 54.32 48.77 37.67 24.33
1173 92.03 76.88 65.87 53.46 47.01 34.68
a Catalyst weight ) 1 g.
Figure 3. Plot of conversion as a function of weight time at
For a first-order reaction in a packed-bed reactor, the different temperatures.
relation between conversion and weight time is as
follows19

k ) (1 + ) ln ( 1
1 - Xout )
- AXout (1)

For the present reaction, the above equation reduces to

k ) 1.5 ln
( 1
1 - XCH4 )
- 0.5XCH4 (2)

From the plots of the function on the right-hand side of Figure 4. Plot of values of logarithmic specific reaction rate
eq 2 versus (w/v0), shown typically in Figure 3 for a constant against the reciprocal of the reaction temperature.
1-g sample of catalyst, it follows that the data fall on a
reasonably straight line through the origin. Hence, the catalyst range from 356 to 402 kJ /mol,9,11 and those
decomposition reaction is of first order. with catalyst range from 90 to 236 kJ/mol.12-14 Thus,
From the values of the specific reaction rate constant, the activation energy obtained in this study using 13
k, calculated from the slopes at different reaction wt % Ni/TiO2-based catalyst is considerably the lower
temperatures, the activation energy obtained in the than others.9-14
present study is 60 kJ/mol following the Arrhenius Reaction Mechanism. Experimental observations
equation. The plot is shown in Figure 4. Reported of the present study indicate that the reaction is
activation energies for methane decomposition without elementary. However, the reaction mechanism and the
4868 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004

decomposes on the front surface of certain active sites

of the Ni/TiO2-based catalyst, and the carbon thus
formed diffuses through the metal and precipitates at
the rear surface. The driving force that pushes the
carbon diffusion was suggested to originate from the
concentration gradient of dissolved carbon between the
two interfaces, i.e., the metal-gas interface to the
metal-nanocarbon interface. Therefore, for high values
of diffusivity, the concentration of carbon dissolved
in nickel is almost uniform and small (therefore, CNi
0).
When the adsorption of methane is the rate-limiting
step, the expression relating the rate of reaction and
the concentrations of various species is

CTK1PCH4
Figure 5. Transmission electron microscope image of carbon
formation on the catalyst particle.
-rCH4 ) (10)
1 + K71/2PH21/2
identification of rate-controlling steps were attempted
in the present study following the mechanism proposed
by Snoeck et al.20 They proposed the following mecha- Initially, the concentration of hydrogen is zero, so eq
nism for methane decomposition 10 becomes

k+1
CH4 + S {\
k
} CH4S (3) -rCH4,0 ) CTK1PCH4 (11)
-1

k+2
CH4S +S {\
k
} CH3S + HS (4)
-2 If the methane decomposition is an adsorption-limited
reaction, then the initial rate will be linear with the
k+3
CH3S + S {\ } CH2S + HS (5) initial partial pressure of methane. This is what was
k -3 observed experimentally in the present study. However,
the individual rate laws and individual initial rate plots
k+4
CH2S + S {\ } CHS + HS (6) with the other steps in the mechanism as rate-limiting
k -4 step are also derived below and checked to confirm
whether the methane decomposition reaction is really
k+5
CHS + S {\ } CS + HS (7) adsorption limited.
k -5
When the removal of the first hydrogen atom from
k+6 molecularly adsorbed methane is the rate-limiting step,
CS {\
k
} CNi + S (8) it follows that
-6

k+7
2HS {\ } H2 + 2S (9) CTK1k+2PCH4
k -7 -rCH4S ) (12)
(1 + K1PCH4 + K71/2PH21/2)2
Whereas eq 3 represents the adsorption of methane
on the surface of the catalyst, eqs 4-7 represent the
surface reaction to form adsorbed methyl radicals and Initially, the concentration of hydrogen is zero, and
adsorbed hydrogen atoms on the catalyst surface. Equa- hence eq 12 becomes
tions 8 and 9 represent the deposition of carbon on the
catalyst surface and the desorption of hydrogen from
the catalyst surface, respectively. CT2K1k+2PCH4
For the mechanism shown in eqs 3-9, the rate- -rCH4S,0 ) (13)
limiting step is identified as follows: It is assumed that (1 + K1PCH4)2
one of the equations is the rate-limiting step, and then
the rate law is formulated in terms of the partial
pressures of the methane and hydrogen. From the If the removal of the first hydrogen atom from molecu-
equation obtained, the variation of the initial reaction larly adsorbed methane limits the methane decomposi-
rate with the initial partial pressure of methane is tion, the rate will increase at low initial partial pres-
checked. If the theoretical rate varies with pressure in sures of methane and then decrease at high initial
the same way as observed experimentally, it will be partial pressures of methane. This conclusion is not
assumed that the mechanism and the rate-limiting step
consistent with the experimental observations of the
are correct.
present study.
A mechanistic interpretation of the growth of carbon
on the catalyst has been proposed in the literature.21,22 When the removal of second, third, and fourth hy-
Figure 5 supports the proposed mechanism for the drogen atoms from molecularly adsorbed methane is
growth of carbon nanotubes on the catalyst. Methane the rate-limiting step, the expressions relating the rate
 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004 4869

of reaction and concentrations of various species are as Initially, the concentration of hydrogen is zero, and
follows eq 19 becomes

-rCH3S )
( C2TK1K2k+3PCH4
K71/2PH21/2 )/(1 + K1PCH4 +
-rHS ) 0 (20)

Because the reaction rate cannot be zero, desorption of

)
2 hydrogen from the surface cannot be the rate-limiting
K1K2PCH4 step.
+ K71/2PH21/2 (14)
K71/2
PH2 1/2 An analysis of the above rate equations indicates that
the rate equation with the adsorption of methane on

( )/(
the surface as the rate-controlling step exhibited a linear
CT2K1K2K3k+4PCH4 relationship with the concentration of methane. This
-rCH2S ) 1 + K1PCH4 + observation is same as that obtained from the analysis
K7PH2 of data by the integral method. Hence, the assumed

)
2
mechanism with the adsorption of methane rate-
K1K2PCH4 K1K2K3PCH4 controlling is consistent with the experimental observa-
1/2 1/2
+ + K7 PH2 (15) tions. This can be expected, as the surface reaction is
K71/2PH21/2 K7PH2
unlikely to control the rate because the reaction occurs

( )/(
at relatively high temperatures.
CT2K1K2K3K4k+5PCH4 Simulation. Conversion profiles along the length of
-rCHS ) 1 + K1PCH4 + the reactor were obtained following the design equation
K73/2PH23/2 for packed-bed reactor. Isothermal and plug-flow condi-
tions (the ratio of bed diameter to catalyst particle
K1K2PCH4 K1K2K3PCH4 K1K2K3K4PCH4 diameter is around 25) were assumed in the reactor.
+ + + Because the flow rates were sufficiently high, heat- and
K71/2PH2 1/2 K7PH2 K73/2PH23/2 mass-transfer limitations were assumed to be negligible,

)
2
and the reaction was assumed to be kinetically con-
1/2 1/2
trolled. Experiments were conducted with different flow
K7 PH2 (16) rates keeping all of the other variables constant, and
the conversion (H2 yield) was found to be almost same
in all of these experiments (there was a slight increase
Initially, the concentration of hydrogen is zero, and
at lower flow rates).
the numerators and denominators of eqs 14-16 ap-
proach infinity. Therefore, the limit of these expressions The packed-bed reactor is assumed to have no radial
approaches zero. Because the reaction rate cannot be gradients in concentration, temperature or reaction rate.
zero, the removal of the second, third, or fourth hydro- When the pressure drop through the reactor and
gen atom from molecularly adsorbed methane cannot catalyst decay are neglected, the relation between
be the rate-limiting step. conversion and weight time is

( )
When the deposition of carbon on the catalyst is rate-
limiting step, one obtains kW 1
) 1.5 ln - 0.5 XCH4 (21)
1 - XCH4

( )/(

CTK1K2K3K4K5k+6PCH4
-rCS ) 1 + K1PCH4 + The above equation can be used to determine the
K72PH22
amount of catalyst needed to obtain a specified conver-
K1K2PCH4 K1K2K3PCH4 K1K2K3K4PCH4 sion at a given temperature and volumetric flow rate.
+ + + Similarly, it can also be used to obtain the conversion
1/2
K7 PH2 1/2 K7PH2 K73/2PH23/2

)
for a given amount of catalyst at a particular temper-
K1K2K3K4K5PCH4 ature and volumetric flow rate. The above equation is
+ K71/2PH21/2 (17) valid for 823 < T < 1173 K and 100 < v0 < 450 mL/
K72PH22 min, as these were ranges in which the experiments
were performed in the present study.
Initially, the concentration of hydrogen is zero, and the The model predictions for the conversion of methane
numerator and denominator of eq 17 approach infinity. under conditions of variable volumetric flow rates (100
Therefore, the limit of this expression gives < v0 < 450 mL/min) and variable temperatures (823 <
T < 1173 K) agree well with the experimental observa-
tions. The predictions of the model for methane conver-
-rCS ) CTk+6 (18)
sion are compared to the experimental values at all
temperatures and volumetric flow rates. Figure 6 shows
Because, the reaction rate cannot be CTk+6, the deposi- a satisfactory comparison between the experimental and
tion of carbon on the catalyst cannot be rate-limiting predicted (using eq 21) conversions. The standard
step. deviation (SD) of (0.13 was obtained using the equation
When desorption of hydrogen from the surface is the
rate-limiting step, one finds

x
N (Exp - Cal)
2
CT k-7PH2 Exp
-rHS ) - (19) SD )
i)1
) (0.13 (22)
(1 + K1PCH4)2 N-1
4870 Ind. Eng. Chem. Res., Vol. 43, No. 16, 2004

Literature Cited
(1) Armor, J. N. The Multiple Roles for Catalysis in the
Production of H2. Appl. Catal. A: Gen. 1999, 76, 159.
(2) Muradov, N. N. CO2-Free Production of Hydrogen by
Catalytic Pyrolysis of Hydrocarbon Fuel. Energy Fuels 1998, 12,
41.
(3) Steinberg, M. M. Fossil Fuel Decarbonization Technology
for Mitigating Global Warming. Int. J. Hydrogen Energy 1999, 24,
771.
(4) Avdeeva, L. B.; Kochubey, D. I.; Shaikhutdinov, S. H. K.
Cobalt Catalysts of Methane Decomposition: Accumulation of the
Filamentous Carbon. Appl. Catal. A: Gen. 1999, 177, 43.
(5) Ermakova, M. A.; Ermakov, D. Yu.; Kuvshinov, G. G.
Catalysts for Direct Cracking of Methane to Produce Hydrogen
and Filamentous Carbon. Part I. Nickel Catalysts. Appl. Catal.
Figure 6. Comparison of theoretically and experimentally de-
A: Gen. 2000, 201, 61.
termined conversions of methane.
(6) Aiello, R.; Fiscus, J. E.; Zur Loye, H.-C.; Amiridis, M. D.
Hydrogen Production via Direct Cracking of Methane over Ni/
SiO2: Catalyst Deactivation and Regeneration. Appl. Catal. A:
As consequence, the experimental values are well fitted Gen. 2000, 192, 227.
by the model at low temperature. (7) Monnerat, B.; Kiwi-Minsker, L.; Renken, A. Hydrogen
Production by Catalytic Cracking of Methane over Nickel Gauze
Conclusions under Periodic Reactor Operation. Chem. Eng. Sci. 2001, 56, 633.
For the decomposition of methane to hydrogen and (8) Choudhary, T. V.; Sivadinarayana, C.; Goodman, D. W.
carbon using 13 wt % Ni/TiO2 catalyst, it was found in Production of COx-Free Hydrogen for Fuel Cells via Step-Wise
Hydrocarbon Reforming and Catalytic Dehydrogenation of Am-
the present study that the decomposition reaction is of monia. Chem. Eng. J. 2003, 93, 69.
first order with an activation energy of 60 kJ/mol. This (9) Holmen, A.; Rockstad, O. A.; Solbaklen, A. High-Temper-
activation energy is the lowest activation energy re- ature Pyrolysis of Hydrocarbons. 1. Methane to Acetylene. Ind.
ported in the literature for this reaction. It was also Eng. Chem. Process Des. Dev. 1976, 15, 439.
found that the adsorption of methane on the surface is (10) Steinberg, M. Production of Hydrogen and Methanol from
the rate-controlling step. The model developed indicates Natural Gas with Reduced CO2 Emission. Int. J. Hydrogen Energy
1998, 23, 419.
the volumetric flow rate, catalyst weight, and temper- (11) Dahl, J. K.; Barocas, V. H.; Clough, D. E.; Weimer, A. L.
ature to be decisive parameters in scale-up consider- Intrinsic Kinetics for Rapid Decomposition of Methane in an
ations and in the extrapolation of laboratory-scale ex- Aerosol Flow Reactor. Int. J. Hydrogen Energy 2002, 27, 377.
perimental results. The model predictions for the con- (12) Alstrup, I.; Tavares, M. T. Kinetics of Carbon Formation
version agree well with the experimental observations. from CH4 + H2 on Silica-Supported Nickel and Ni-Cu Catalysts.
J. Catal. 1993, 139, 513.
Acknowledgment (13) Kuvshinov, G. G.; Mogilnykh, Yu. I.; Kuvshinov, D. G.
Kinetics of Carbon Formation from CH4-H2 Mixtures over a
The authors acknowledge for the financial support Nickel Containing Catalyst. Catal. Today 1998, 42, 357.
provided by Ministry of Education, Malaysia and Uni- (14) Muradov, N. Proceedings of the 2000 Hydrogen Program
versiti Sains Malaysia under Fundamental Research Review; Report NREL/CP-570-28890; U.S. Department of Energy,
Grant Scheme (Project: A/C No: 6070014). U.S. Government Printing Office: Washington, DC, 2000.
(15) Takenaka, S.; Ogihara, H.; Yamanka, I.; Otsuka, K.
Nomenclature Decomposition of Methane over Supported-Ni Catalysts: Effects
CA ) concentration of substance A (mol/mL) of the Supports on the Catalytic Lifetime. Appl. Catal. A: Gen.
CA0 ) initial concentration of substance A (mol/mL) 2001, 217, 101.
(16) Soick, M.; Buyevskaya, O.; Hohenberger, M.; Wolf, D.
CCH4 ) concentration of methane (mol/mL) Partial Oxidation of Methane to Synthesis Gas over Pt/MgO
CNi ) concentration of carbon dissolved in nickel Kinetics of Surface Processes. Catal. Today 1996, 32, 163.
CT ) total concentration of sites (17) Takenaka, S.; Kito, E.; Tomikubo, Y.; Otsuka, K. Structural
CV ) concentration of vacant sites Change of Ni Species During the Methane Decomposition and the
k+1, k-1, k+2, k-2, k+3, k-3, k+4, k-4, k+5, k-5, k+6, k-6, k+7, Subsequent Gasification of Deposited Carbon with CO2 over
k-7 ) rate coefficients of the forward (+) and the reverse Supported Ni Catalysts. J. Catal. 2003, 219, 176
(-) reactions (18) Takenaka, S.; Tomikubo, Y.; Kato, E.; Otsuka, K. Sequen-
k ) specific reaction rate constant [mL/(gcat s)] tial Production of H2 and CO over Supported Ni Catalysts. Fuel
K1, K2, K3, K4, K5, K6, K7 ) equilibrium coefficients 2004, 83, 47
(19) Levenspiel, O. Chemical Reaction Engineering; Wiley: New
N ) number of experiments York, 1999.
PCH4 ) partial pressure of methane (20) Snoeck, J. W.; Froment, G. F.; Fowles, M. Kinetic Study
-rA ) rate of species A decomposition [mol(gcat s)] of the Carbon Filament Formation by Methane Cracking on a
-rCH4 ) rate of methane decomposition [mol/(gcat s)] Nickel Catalysts. J. Catal. 1997, 169, 250.
S ) surface active site (21) Zhang, T. T; Amiridis, M. D. Hydrogen Production via the
T ) Temperature (K) Direct Cracking of Methane over Silica-Supported Nickel Cata-
v0 ) entering volumetric flow rate lysts. Appl. Catal. A: Gen. 1998, 167, 161.
Vp ) total pore volume (cm3/g) (22) Ermakova, M. A.; Ermakov, D. Yu.; Chuvilin, A. L.;
W ) Weight of catalyst Kuvshinov, G. G. Decomposition of Methane over Iron Catalysts
at the Range of Moderate Temperatures: The Influence of
XA ) conversion of species A Structure of the Catalytic Systems and the Reaction Conditions
XCH4 ) conversion of methane on the Yield of Carbon and Morphology of Carbon Filaments. J.
Xout ) conversion at the outlet Catal. 2001, 201, 183.
Greek Symbols Received for review October 25, 2003
Revised manuscript received March 25, 2004
 ) fractional change in volume per mole of A reacted
Accepted May 4, 2004
resulting from the change in total number of moles
) weight time (g s/mL) IE034208F