Kinetics of Partial Methane Oxidation Process

over the Fe-ZSM-5 Catalysts*

B. MICHALKIEWICZ

Institute of Chemical and Environment Engineering, Szczecin University of Technology, PL-70 322 Szczecin
e-mail: beata.michalkiewicz@ps.pl

Received 1 April 2005

Kinetics of a partial methane oxidation over the Fe-ZSM-5 catalysts was investigated. A set
of rate equations describing the reaction system was suggested and verified by comparison with
experimental data. It was found that all the reactions considered were of the zero order with
respect to oxygen and of the first order with respect to the oxidized species (methane, methanol, or
formaldehyde). Values of the reaction rate constant for the temperature range 350—550 ◦C as well
as the reaction apparent activation energy values were determined.
The studies on the process of partial methane oxidation conducted over the Fe-ZSM-5 catalysts
confirmed the assumption that methanol and carbon dioxide are the primary products of the reac-
tion. As a result of the methanol oxidation, formaldehyde is generated, the oxidation of which leads
to receiving slight quantities of CO2 .
The decline of the iron content results in the decrease of the rate of every reaction. In order to
impede the oxidation of methanol to formaldehyde, the H catalysts should be employed. This form is
responsible for the production of methanol; yet it also leads to the direct methane to carbon dioxide
oxidation. Bearing such dependences in mind, it could be concluded that the introduction of the
Fe-ZSM-5 catalyst demonstrating certain content might result in a favoured reaction of synthesis
of methanol, with the process of its further oxidation being hampered. Unfortunately, this makes
the elimination of the reaction of oxidation to CO2 virtually impossible.
It is also evident that the activation energy of methane to methanol or carbon dioxide oxidation
is markedly higher than the activation energy of the unwelcome reactions, which are methanol and
formaldehyde oxidations.

Studies regarding the process of direct methane ox- [10—13]. Until now, the highest yield of formalde-
idation to oxidation products different from those ob- hyde was obtained when molybdenum and vanadium
tained when preparing synthesis gas were conducted compounds supported on silica were applied [5—8].
already at the beginning of the 20th century [1]. How- Unfortunately, selectivity of methane oxidation to
ever, finding of a catalyst allowing obtaining methanol formaldehyde never exceeded 10 %, i.e. highly ac-
or formaldehyde with a satisfactory yield was found tive catalysts demonstrated low selectivity towards
to be quite complicated task. Since, methane combus- products of partial methane oxidation, and highly se-
tion to carbon oxides is thermodynamically favoured, lective catalysts showed extremely low activity. De-
formaldehyde or methanol synthesis by direct methane spite this fact literature offers a few examples of cat-
oxidation, although thermodynamically feasible, is alysts that allow production of methanol by a di-
less probable. The most complicated is the direct rect methane oxidation with oxygen, e.g. Fe-sodalite
methanol production. [14], ZSM-5 [15], Fe-ZSM-5 [16]. Methanol, formalde-
Number of catalysts showing different composition hyde, and carbon dioxide are the main products of
and structure were examined over the last century methane oxidation carried out with the use of this
[2]. It was found that carbon oxides and formalde- catalyst.
hyde are the principal products of methane oxidation The aim of this paper was to investigate kinetics of
with oxygen [3—9]. If nitric oxide N2 O was used to the process of partial methane oxidation in the pres-
oxidize methane, higher portion of formaldehyde was ence of Fe-ZSM-5 catalysts, containing iron and silicon
obtained and also methanol production was possible in their framework only.

*Presented at the 32nd International Conference of the Slovak Society of Chemical Engineering, Tatranské Matliare,
23—27 May 2005.

Chem. Pap. 59 (6a) 403—408 (2005) 403

 B. MICHALKIEWICZ

EXPERIMENTAL at low temperature resulted in higher methanol yield.

At high temperature the lower CH3 OH yield was ob-
Metallosilicates with the ZSM-5 structure were served when the contact time was longer (Fig. 3).
used as catalysts with atomic ratio n(Si) : n(Fe) of The same phenomenon was observed with regard
45, 34, and 22. Both forms, H-ZSM-5 and Na-ZSM- to Fe-HZSM-5. However, in this case carbon dioxide
5, were used for the catalyst preparation. Catalyst always turned out to be the principal reaction prod-
preparation, apparatus, in which the catalysts were uct, regardless of experimental conditions. Extending
prepared, as well as analysis of the reaction mixture the contact time caused a very slow increase of the
was described in [16]. formaldehyde content in the products. On the other
The reaction was carried out in the vertical flow re- hand, a slight decline of the HCHO yield was ob-
actor containing a catalyst. Methane and oxygen were served when the Fe-NaZSM-5 catalysts were employed
supplied from the gas cylinders equipped with micro- at higher temperatures.
metric valves. Both reagents were passed through the Based on a thorough analysis of composition of the
columns with the molecular sieves 5A in order to re- post-reaction mixture obtained under various reaction
move the moisture. The flow rates of gases were mea- conditions, a set of reactions occurring in the reactor
sured using flow meters. The mixing of reagents was was assumed
performed in a mixer. The reagents were passed by
k
valve to the quartz reactor of an inside diameter 11 CH4 + 0.5O2 −−−A−→ CH3 OH (A)
mm. The catalyst bed was placed on the grate from kB
CH4 + 2O2 −−−−→ CO2 + 2H2 O (B )
quartz wool. A resistance furnace heated the reactor. kC
The measurement and control of temperature was en- CH3 OH + 0.5O2 −−−−→ HCHO + H2 O (C )
kD
sured by the system: a thermocouple Fe—CuNi and HCHO + O2 −−−−→ CO2 + H2 O (D )
temperature regulator with a digital meter. A receiver
was used for the collection of the liquid fraction. The Then, the following reaction rate expressions have
gas products were sampled through the membrane. been suggested
Oxygen that comprised 15.5 % of the reaction mix-
ture volume was used as an oxidant.
Contact time of the reaction mixture with the cat- rCH3 OH = kA (pCH4 )N (ACH4 ) (pO2 )N (AO2 ) −
alyst varied between 0.5 s and 2.5 s (1440—7200 h−1 − kC (pCH3 OH )N (CCH3 OH ) (pO2 )N (CO2 ) (1)
space velocity at NTP) at a temperature ranging from
N B N B
350 to 590 ◦C. rCO2 = kB (pCH4 ) ( CH4 ) (pO2 ) ( O2 ) +
Kinetic studies were carried out in an integral-type N (DHCHO ) N D
+ k (p
D HCHO ) (p ) ( O2 )
O2 (2)
reactor. Two series of measurements were conducted.
N (CCH3 OH ) N (CO2 )
During the first one the temperature was kept con- rHCHO = kC (pCH3 OH ) (pO2 ) −
stant while the contact time was changed. In the other N (DHCHO ) N D
one, the process was carried out at different tempera- − kD (pHCHO ) (p ) ( O2 )
O2 (3)
tures, with the same contact time.
The composition of products mixture was deter- rx being the rate of generating the product x, kA , kB ,
mined chromatographically. From the values of the kC , kD pressure reaction rate constants for the corre-
mole fractions of the gases the values of partial pres- sponding reactions (A—D), px partial pressure of the
sure of the compounds were calculated. component x at the reactor outlet, and N (Ix ) order of
reaction (1) related to the component x.
RESULTS AND DISCUSSION As a result of the numerical differentiation of the
px = f (τ ) equation, the instantaneous rate of gen-
In Figs. 1 and 2 variation of selectivity of the erating the product for every contact time that was
methane oxidation to methanol, formaldehyde, and used during the measurements, was obtained. The re-
carbon dioxide with the contact time is shown for the action rate coefficient was determined with the help
lowest and the highest reaction temperature. Prolong- of the quasi-Newton method. It was concluded that
ing the contact time for all catalysts at both temper- satisfactory results were arrived at only when it was
atures resulted in the increase of carbon dioxide con- assumed that all the four reactions were of the zero or-
tent in the post-reaction mixture. The higher the reac- der with respect to oxygen and of the first order with
tion temperature, the more vigorous increase of CO2 respect to the other substrate (methane, methanol, or
yield was noted. For the Fe-NaZSM-5 catalysts, the formaldehyde). It is worth mentioning that Spencer
carbon dioxide yield was lower than that of methanol and Pereira came up with a similar assumption while
when the reaction was carried out at low temperatures investigating the kinetics of methane partial oxida-
and for short contact time. Raising the temperature tion with the molybdena-silica catalyst [3]. Hence, the
and Fe content resulted in the increase of methanol kinetic equations were simplified giving the following
and carbon dioxide yield. Elongating the contact time forms

404 Chem. Pap. 59 (6a) 403—408 (2005)

KINETICS OF PARTIAL METHANE OXIDATION

Fig. 1. The changes in the selectivity of methane oxidation to particular products with the extension of the contact time for
the catalyst’s hydrogen form. a – Fe-HZSM-5(45), b – Fe-HZSM-5(34), c – Fe-HZSM-5(22), 1 – temperature 350 ◦C, 2 –
temperature 590 ◦C, diamond and dotted line – methanol, rectangle and straight line – formaldehyde, triangle and dashed
line – carbon dioxide.

rCH3 OH = kA pCH4 − kC pCH3 OH (4) same atomic ratio n(Si) : n(Fe) resulted in an increase
rCO2 = kB pCH4 + kD pHCHO (5) of the rate of methane oxidation to methanol.
rHCHO = kC pCH3 OH − kD pHCHO (6) The rate of methanol oxidation to formaldehyde,
reaction (C ), was significantly higher when the Fe-
Fig. 4 presents Arrhenius plot of reaction rate con- NaZSM-5 catalysts were used. Iron content in both
stants for the six catalysts. Their rectilinear course forms of catalysts (Na and H) was directly propor-
confirms the accuracy of the previously assumed set tional to the observed rate of methanol oxidation.
of equations. When Fe-HZSM-5(45) was used as catalyst, the un-
Analysis of the value of reaction rate coefficients welcome process of methanol oxidation was performed
obtained at different temperatures in the presence the most slowly.
of various catalysts allowed formulating the follow- The dependence between the iron content, the cat-
ing conclusions. The highest rate of methanol produc- alyst form (Na or H), and the rate of formaldehyde
tion, reaction (A), was observed when the Fe-HZSM- oxidation, reaction (D ), was similar to that of the re-
5(22) catalyst was used. The decline of the iron con- action (C ).
tent in the catalysts results in the decrease of methane Direct methane oxidation to carbon dioxide, reac-
to methanol oxidation rate. Use of H-ZSM-5 instead tion (B ), was the most intense in the presence of Fe-
of the corresponding sodium zeolite form having the HZSM-5(22) catalyst. The decline of the iron content

Chem. Pap. 59 (6a) 403—408 (2005) 405

 B. MICHALKIEWICZ

Fig. 2. The changes in the selectivity of methane oxidation to particular products with the extension of the contact time for
the catalyst’s sodium form. a – Fe-NaZSM-5(45), b – Fe-NaZSM-5(34), c – Fe-NaZSM-5(22), 1 – temperature 350 ◦C, 2 –
temperature 590 ◦C, diamond and dotted line – methanol, rectangle and straight line – formaldehyde, triangle and dashed
line – carbon dioxide.

Fig. 3. The changes in the methanol yield with the extension of the contact time. 1 – H form, 2 – Na form, dotted line – 350 ◦C,
straight line – 590 ◦C, rectangle – Si : Fe = 45, cross – Si : Fe = 34, house – Si : Fe = 22.

caused the decrease of the reaction (B ) rate. It is clear ted to give concentration rate coefficients, kci , in order
that this reaction was much faster in the presence of to determine the activation energy for reactions (A—
H-ZSM-5-based catalysts. D ). Next, the activation energy for reactions (A—D )
The pressure rate coefficients, ki , were recalcula- was calculated on the basis of the inclination angle

406 Chem. Pap. 59 (6a) 403—408 (2005)

KINETICS OF PARTIAL METHANE OXIDATION

Fig. 4. Arrhenius plot of the reaction rate constants kA (plus and short dashed line), kB (square and long dashed line), kC
(house and dotted line), and kD (asterisk and straight line) for the catalysts: a – Fe-NaZSM-5(45), b – Fe-NaZSM-5(34),
c – Fe-NaZSM-5(22), d – Fe-HZSM-5(45), e – Fe-HZSM-5(34), and f – Fe-HZSM-5(22).

of the straight line: ln(kci ) = ln(A) − E/RT . Ta- same iron content revealed insignificant differences be-
ble 1 summarizes the values of apparent activation tween the corresponding values of apparent activation
energy, E, of all reactions in the presence of all cata- energy. The activation of methane (reactions (A) and
lysts investigated. Increase of the iron content in the (B )) took place on the same catalytically active cen-
catalyst leads to the decline of the values of appar- tres.
ent activation energy. The value of activation energy The values of activation energies for the oxida-
depends on the number of active species generated tion of methanol and formaldehyde were significantly
by iron. Different forms of iron have been identified lower than the values of EA and EB . The activation of
in Fe-ZSM-5 [17]: isolated ions either in framework methane and oxygenates (CH3 OH and HCHO) needed
positions or in cationic positions in the zeolite chan- different active species. Markedly lower values of EC
nels, binuclear and oligonuclear iron complexes in ex- and ED were observed when sodium form of the cat-
traframework positions, iron oxide FeOx nanoparticles alyst was used compared to the corresponding Fe-
and large iron oxide particles (Fe2 O3 ). It is difficult HZSM-5 one. Because of acid properties of methanol
to distinguish which one is important for the reac- and formaldehyde these compounds could be more
tions (A—D ). It is evident that the activation energy easily oxidized on Na than on H forms of ZSM-5.
values EA and EB for the same catalyst were simi- The highest methanol and formaldehyde yield was
lar. Both forms (Na and H) of the catalyst having the 3.16 % and 4.52 %, respectively. These values were

Chem. Pap. 59 (6a) 403—408 (2005) 407

 B. MICHALKIEWICZ

Table 1. Apparent Activation Energy of Reactions (A—D) in REFERENCES

the Presence of Investigated Catalysts
1. Newitt, D. M. and Haffner, A. E., Proc. R. Soc. A134,
Apparent activation energy, Ei /(kJ mol−1 ) 591 (1932).
Si : Fe atomic 2. Bone, W. A. and Wheeler, R. V., J. Chem. Soc., Trans.
ratio (A) (B ) (C ) (D)
81, 535 (1902).
45 64776 65042 35873 41834
3. Spencer, D. and Pereira, C. J., AIChE J. 33, 1808
H-ZSM-5 34 60237 60305 34534 35972 (1987).
22 55416 54241 31933 29839 4. Kalucki, K. and Michalkiewicz, B., Pol. J. Chem. Tech-
nol. 2, 17 (2000).
45 65026 73680 31178 30719 5. Parmaliana, A., Frusteri, F., Mezzapica, A., Scurrel, M.
Na-ZSM-5 34 60610 66995 26553 26198 S., and Giordano, N., J. Chem. Soc., Chem. Commun.
22 59150 63247 25823 25395
1993, 751.
6. Weng, T. and Wolf, E. E., Appl. Catal. 96, 383 (1993).
7. Otsuka, K. and Wang, Y., Appl. Catal., A 222, 145
(2001).
achieved over Fe-HZSM-5 at 590 ◦C and after 2.5 s 8. Spencer, N. D., J. Catal. 109, 143 (1988).
contact time. 9. Otsuka, K., Komatsu, T., Jinno, K., Uragami, Y., and
Morikawa, A., in Proceedings of the 9th International
Congress on Catalysis, Ottawa, Canada, 1988.
SYMBOLS 10. Dubkov, K. A., Sobolev, V. I., and Panov, G., Kinet.
Catal. 39, 72 (1998).
A frequency factor 11. Ovanesyan, N. S., Dubkov, K. A., Pyalling, A. A., and
k pressure rate coeffi- Shteinman, A. A., J. Radioanal. Nucl. Chem. 246, 149
cient mol dm−3 s−1 MPa−1 (2000).
kc concentration rate coefficient s−1 12. Teng, Y., Ouyang, F., Dai, L., Karasuda, T., Sakurai,
E activation energy J mol−1 H., Tabata, K., and Suzuki, E., Chem. Lett. 1999, 991.
T temperature K 13. Teng, Y., Sakurai, H., Tabata, K., and Suzuki, E., Appl.
t time s Catal., A 190, 283 (2000).
r reaction rate mol dm−3 s−1 14. Durante, V. A., Walker, D. W., Gussow, S. M., and
Lyons, E. J., Eur. 0393895 (1990).
p component partial pressure MPa
15. Kudo, H. and Ono, T., Appl. Surf. Sci. 121/122, 413
S product selectivity mol % (1997).
R universal gas constant kJ mol−1 K−1 16. Michalkiewicz, B., Appl. Catal. 277, 147 (2004).
17. Perez-Ramirez, J., Mul, G., Kapteijn, F., Moulijn, J.
A., Overweg, A. R., Domenech, A., Ribera, A., and
Arends, I. W. C. E., J. Catal. 207, 113 (2002).

408 Chem. Pap. 59 (6a) 403—408 (2005)