- Email: [email protected]
Detailed mechanism of the oxidative coupling of methane
Studies in Surface Science and Catalysis, volume 147 X. Bao and Y. Xu (Editors) 9 Elsevier B.V. All rights reserved.
571
Detailed mechanism of the oxidative coupling of methane Y. Simon*, F. Baronnet, G.M. C6me and P.M. Marquaire D6partement de Chimie Physique des R6actions, UMR 7630 - CNRS, ENSIC-1NPL, 1 rue Grandville - BP 451, 54001 NANCY Cedex (France). ABSTRACT To determine the relative importance of gas-phase and surface reactions in the oxidative coupling of methane (OCM), experimental investigations were performed. Our experimental results were compared to simulated values derived from a kinetic model taking into account heterogeneous and gasphase reactions. We propose an original approach derived from Benson's techniques to estimate the kinetic parameters of surface reactions. 1. I N T R O D U C T I O N Whereas it is well established that the OCM is a homogeneous-heterogeneous reaction, several points are still open on the reaction mechanism, especially the relative importance of gas-phase and surface reactions. Experiments in a continuous flow reactor over La203 as a catalyst were performed in a large temperature range (1023 K to 1173 K) and on a variable quantity of catalyst. The results were compared to simulated values. 2. EXPERIMENTAL The experimental setup is a catalytic jet-stirred reactor and has been previously described [1-4]. The reactor is made of quartz and has a gas-phase of constant volume V = 113 cm3.The catalyst, La203, made from lanthanum carbonate, has a B.E.T. specific area equal to 1.2 m2.g -~. The particles have a diameter of around 60 lain and are compacted into pellets (diameter: 12.6 mm, thickness: 1 mm, mass: 0.45 g). Catalyst pellets are laid on a flat surface inside the reactor. The outlet gas stream was analysed by on-line gas
* Corresponding Author: Email: [email protected] Phone: (33) 383.17.51.22 Fax: (33)383.37.81.20
572
chromatography. The reaction was carried out in the following conditions" inlet composition mixture" CH4/O2 - 4, temperatures from 1023 to 1173 K and number of catalyst pellets from 0 to 8. 3. KINETIC MODELLING Experimental results were compared to simulated values. These simulations were performed by means of Chemkin Surface Program on the basis of a kinetic model which takes into account elementary heterogeneous and gasphase reactions. The mechanism used for simulation includes a heterogeneous part and a homogeneous one. The homogeneous part is described by a set of over 450 elementary reactions. It takes into account all the elementary reactions between molecules and free radicals including less than three carbon atoms. This gas- phase mechanism is well known and has been confirmed by a large amount of experimental data for different hydrocarbon reactions, particularly for the homogeneous oxidation of methane [5]. The heterogeneous mechanism takes into account the results from the mechanistic studies reported in the literature [6-12]. The experimental studies of oxygen chemisorption have shown that this reaction leads to the dissociation of diatomic O2 molecule to form two active atomic oxygen centres. 02 at- 0
<
r
O2(S) + 0
> 20(s)
The rate determining step is assumed to be the 02 dissociation reaction. Heterogeneous mechanism is written assuming that reactive surface oxygen species interact with CH4 and all major reaction products. We presumed that oxygen surface reactions are E l e y - Rideal reactions. The first step is the activation of methane. CH4 + O(s) "-) CH3. + OH(s)
Estimation of the kinetic parameters of surface reactions. In this work, we have chosen a method of estimation based on an original approach derived from BENSON's techniques [13]; it can be an alternative to the ab-initio methods which are more complex to handle. The activation energy of elementary surface reactions is assumed to be the same as that of the equivalent gas phase reaction. The preexponential factors are estimated by using partition functions of reactants and transition state. For example, in a reaction such as:
573
A + B(s) --) ABe(s) --) products the preexponential factor is given by the equation 9
A = k__T Nq,~B h qAqB
k 9Planck constant h 9Bolzman constant Y "Temperature (K) N "Avogadro number
where qA, qB, q#AB are the partition functions of A, B and the activated complex AB ~. Partition functions of gas-phase molecules can be calculated. On the other hand, the determination of the partition function of absorbed species [B(s), ABe(s)] requires some assumptions. Therefore, we suppose that the difference between the partition functions of these two species is only due to the vibrational part qv. Hence, a vibrational analysis of adsorbed species was carried out to estimate the vibrational partition function qv. The frequency values used are those given by Benson [13]. Finally, this mechanism is a set of 20 elementary reactions. 4. R E S U L T S A N D D I S C U S S I O N
The formation of different reaction products (CO, CO2, C2H4, C2H6) and CH4 conversion were simulated at different experimental conditions (temperature, number of catalyst pellets) using the Chemkin Surface Program. In our model the specific area and the site density are not adjustable parameters. The value of the specific area was determined by using B.E.T. method and that of site density was taken in the literature [ 14]. Only the frequency factors have been optimized to adjust calculated and experimental results. The correlation between experimental and simulated values is satisfactory (see Fig. 1 and 2). As a result of the kinetic sensitivity analysis, the set of 20 heterogeneous reactions could be reduced to 6 reactions (Table 1).The production rate analysis of the main species has allowed to write the reaction scheme of OCM (Fig. 3). Table 1 OCM reduced surface mechanism (~) symbolizes surface site and (s) adsorbed species) Reactions A(mol, cm, s) E (cal. rnolq) 0 2 q-
20 --) 20(s)
CH4 + O(s) --) CH3. + OH(s) 2 OH(s) --) H20 + O(s) + 0
CO + O(s) -) CO2 + C2H4 + O(s) "-)' C2H3. + OH(s) C2H6 + O(s) "-) C2H5. + OH(s)
6.1 1016 2.8 1008 3.0 1023 2.7 1009 2.1 10l~ 2.5 10l~
-26700 8400 1O0 3000 5900 5800
574
1.5
0,5"
[] C2H4 9 CO 9
CO
CO2
0,4"
[] C2H4 9 CO 9 CO2
.. CO2
1 0,3" @ I,i
~,
CO2
CO 0,2
0.5 0,1[] II u
0
[]
[]
Z~~..~..~r~.~
C2H4
C2H4
|
|
|
2
4
6
Fig. 1. 1023 K and 1 catalyst pellet
0 8
0
2
4
1: (s)
;
Fig. 2.1173 K and 3 catalyst pellets
Experimental/caculated production of CO, C 0 2 and C2H4 (molar fraction vs gas space time)
According to this scheme, the initiation is exclusively the surface reaction: CH4 + O(s) ") CH3. + OH(s) This heterogeneous step is the main source of the free radicals involved in the chains of the gas-phase reaction. This mechanism shows that the catalytic reactions are also important for CO2 formation from CO and for C2H4 formation from CzH6.We can note that the mechanism of OCM includes two distinct reaction pathways. The first pathway leads to the formation of oxygenated compounds Ox (CO, CO2, HCHO...) and the second one leads to the formation of hydrocarbons C2+ (C2H6, C2H4, C2H2...). Therefore, when the conversion is low, C2+ and Ox selectivities are controlled by the rates of formation of C2H6 and HCHO respectively, through the reactions:
CH3. +
02
"-) HCHO + OH.
2 CH3. "-) C2H6
r = k(CH3.)(O2) r - k (CH3.)2
The C2+ selectivity increases when 02 concentration decreases and when the initiation rate increases. On the other hand, when the conversion is high, a third reaction pathway decomposes C2H4 into oxygenated species and, as a consequence, C2+ selectivity decreases.
575
100% ~.......:....~.....':+ .. * ,>
+CH4
.................~
.....................................
~
~r ~
2-o
'CH3" ~ ~~
~
+CH3.
I C2H6 ~,,=~,,I
I HCH~ 100%
+R.
40% :~i~:=:~, +R.i:~=~::~!!i=~ 60% ~:;~+O* :
/--
C2H5.
CHO.
~ ..dl 29% ~+ 0 ~2 67% + M ~ ~
i
l
Z
/
+HO 6 6 % .IZIII+O*
+0 2
"
63%
+M 35%
\ )
+02
'p,
<
90%
16% & +M
?
CO2
C2H2
......................................................................................................... H., OH., HO2. Fig. 3. Mechanism of the oxidative coupling of methane over La203 (1023 K, CH4 conversion" 10%). Black arrows are gas-phase reactions and grey arrows are surface reactions.
The presence of La203 increases the C2+ selectivity when the gas space time is low by introducing a new surface initiation reaction. Then, for higher gas space time, the secondary reaction of C2H3. oxidation decreases C2+ selectivity. This analysis explains that many authors have found that C2+ selectivity decreases drastically with increasing conversion of methane [1517].
576
5. CONCLUSION The reaction scheme proposed in this work allows to simulate OCM reaction over La203 in a catalytic jet-stirred reactor over a large range of operating conditions. Preexponential factors of heterogeneous reactions were estimated by using an original method. Finally, the reduced surface mechanism contains only 6 elementary reactions and shows that the OCM reaction is a gas-phase chain reaction coupled with surface reactions. This study suggests that the surface mechanism of OCM over La203 essentially accounts for the initiation of the reaction, the production of CO2 and the decomposition of C2H6. The OCM mechanism includes 2 reaction pathways. The first one leads to the formation of oxygenated species and the second one leads to the formation of hydrocarbons. This mechanism also explains why the hydrocarbon selectivity seems to reach a limit with increasing conversion. REFERENCES [1] P.M. Marquaire, P. Barb6, Y.D. Li, G.M. C6me and F. Baronnet, Stud. Surf. Sci. Catal., 81 (1994) 149. [2] P. Barb6, Y.D. Li, P.M. Marquaire, G.M. C6me and F. Baronnet, Catal. Today, 21 (1994) 409. [3] G.M. C6me, Y.D. Li, P. Barb6, N. Gueritey, P.M. Marquaire and F. Baronnet, Catal. Today, 30 (1996) 215. [4] P.M. Marquaire, N. Gueritey, G.M. C6me and F. Baronnet, Stud. Surf. Sci. Catal. 119 (1997) 383. [5] P. Barb6, F. Battin-Leclerc and G.M. C6me, J. Chim. Phys. 92 (1995) 1666. [6] J.G. McCarty, Methane Conversion by Oxidative Processes, E.E. Wolf Ed., Van Nostrand Reinhold, New York (1992). [7] G.A. Martin and C. Mirodatos, Fuel Processing Technology 42 (1995) 179. [8] S. Lacombe, Z. Durjanova, L. Mleczko and C. Mirodatos, Chem. Eng. Technol., 42 (1995) 216. [9] P.M. Couwenberg, Qi Chen and G.B. Marin, Ind. Eng. Chem. Res., 35 (1996) 3999. [10]D. Wolf, M. Slinko, M. Baems and E. Kurkina, Appl. Catal. A: General, 166 (1998) 47. [11]K. Coulter and D.W. Goodman, Catal. Letters, 20 (1993) 169. [12] C. Shi, M. Xu, M.P. Rosynek and J.H. Lunsford, J. Phys. Chem., 97 (1993) 216. [ 13] S.W. Benson, Thermochemical Kinetics, 2nd Edition, Wiley Interscience (1976). [ 14] S.-J. Huang, A.B. Walters, M.A. Vannice, Appl. Cat. B, 17 (1998) 183. [15]Z. Kalenik and E.E. Wolf, Methane Conversion by Oxidative Processes, E.E. Wolf Ed., Van Nostrand Reinhold, New York (1992). [16] T. Le Van, C. Louis, M. Kermarec, M. Che, J.M. Tatibou~t, Catal. Today, 13 (1992) 321. [17] T. Wakatsuki, M. Yamamura, H. Okado, K. Chaki, S. Okada, K. Inabu, S. Susuki and T. Yoshinari, Natural Gas Conversion IV, Studies in Surface Science and Catalysis, Vol. 107, Elsevier (1997).
571
Detailed mechanism of the oxidative coupling of methane Y. Simon*, F. Baronnet, G.M. C6me and P.M. Marquaire D6partement de Chimie Physique des R6actions, UMR 7630 - CNRS, ENSIC-1NPL, 1 rue Grandville - BP 451, 54001 NANCY Cedex (France). ABSTRACT To determine the relative importance of gas-phase and surface reactions in the oxidative coupling of methane (OCM), experimental investigations were performed. Our experimental results were compared to simulated values derived from a kinetic model taking into account heterogeneous and gasphase reactions. We propose an original approach derived from Benson's techniques to estimate the kinetic parameters of surface reactions. 1. I N T R O D U C T I O N Whereas it is well established that the OCM is a homogeneous-heterogeneous reaction, several points are still open on the reaction mechanism, especially the relative importance of gas-phase and surface reactions. Experiments in a continuous flow reactor over La203 as a catalyst were performed in a large temperature range (1023 K to 1173 K) and on a variable quantity of catalyst. The results were compared to simulated values. 2. EXPERIMENTAL The experimental setup is a catalytic jet-stirred reactor and has been previously described [1-4]. The reactor is made of quartz and has a gas-phase of constant volume V = 113 cm3.The catalyst, La203, made from lanthanum carbonate, has a B.E.T. specific area equal to 1.2 m2.g -~. The particles have a diameter of around 60 lain and are compacted into pellets (diameter: 12.6 mm, thickness: 1 mm, mass: 0.45 g). Catalyst pellets are laid on a flat surface inside the reactor. The outlet gas stream was analysed by on-line gas
* Corresponding Author: Email: [email protected] Phone: (33) 383.17.51.22 Fax: (33)383.37.81.20
572
chromatography. The reaction was carried out in the following conditions" inlet composition mixture" CH4/O2 - 4, temperatures from 1023 to 1173 K and number of catalyst pellets from 0 to 8. 3. KINETIC MODELLING Experimental results were compared to simulated values. These simulations were performed by means of Chemkin Surface Program on the basis of a kinetic model which takes into account elementary heterogeneous and gasphase reactions. The mechanism used for simulation includes a heterogeneous part and a homogeneous one. The homogeneous part is described by a set of over 450 elementary reactions. It takes into account all the elementary reactions between molecules and free radicals including less than three carbon atoms. This gas- phase mechanism is well known and has been confirmed by a large amount of experimental data for different hydrocarbon reactions, particularly for the homogeneous oxidation of methane [5]. The heterogeneous mechanism takes into account the results from the mechanistic studies reported in the literature [6-12]. The experimental studies of oxygen chemisorption have shown that this reaction leads to the dissociation of diatomic O2 molecule to form two active atomic oxygen centres. 02 at- 0
<
r
O2(S) + 0
> 20(s)
The rate determining step is assumed to be the 02 dissociation reaction. Heterogeneous mechanism is written assuming that reactive surface oxygen species interact with CH4 and all major reaction products. We presumed that oxygen surface reactions are E l e y - Rideal reactions. The first step is the activation of methane. CH4 + O(s) "-) CH3. + OH(s)
Estimation of the kinetic parameters of surface reactions. In this work, we have chosen a method of estimation based on an original approach derived from BENSON's techniques [13]; it can be an alternative to the ab-initio methods which are more complex to handle. The activation energy of elementary surface reactions is assumed to be the same as that of the equivalent gas phase reaction. The preexponential factors are estimated by using partition functions of reactants and transition state. For example, in a reaction such as:
573
A + B(s) --) ABe(s) --) products the preexponential factor is given by the equation 9
A = k__T Nq,~B h qAqB
k 9Planck constant h 9Bolzman constant Y "Temperature (K) N "Avogadro number
where qA, qB, q#AB are the partition functions of A, B and the activated complex AB ~. Partition functions of gas-phase molecules can be calculated. On the other hand, the determination of the partition function of absorbed species [B(s), ABe(s)] requires some assumptions. Therefore, we suppose that the difference between the partition functions of these two species is only due to the vibrational part qv. Hence, a vibrational analysis of adsorbed species was carried out to estimate the vibrational partition function qv. The frequency values used are those given by Benson [13]. Finally, this mechanism is a set of 20 elementary reactions. 4. R E S U L T S A N D D I S C U S S I O N
The formation of different reaction products (CO, CO2, C2H4, C2H6) and CH4 conversion were simulated at different experimental conditions (temperature, number of catalyst pellets) using the Chemkin Surface Program. In our model the specific area and the site density are not adjustable parameters. The value of the specific area was determined by using B.E.T. method and that of site density was taken in the literature [ 14]. Only the frequency factors have been optimized to adjust calculated and experimental results. The correlation between experimental and simulated values is satisfactory (see Fig. 1 and 2). As a result of the kinetic sensitivity analysis, the set of 20 heterogeneous reactions could be reduced to 6 reactions (Table 1).The production rate analysis of the main species has allowed to write the reaction scheme of OCM (Fig. 3). Table 1 OCM reduced surface mechanism (~) symbolizes surface site and (s) adsorbed species) Reactions A(mol, cm, s) E (cal. rnolq) 0 2 q-
20 --) 20(s)
CH4 + O(s) --) CH3. + OH(s) 2 OH(s) --) H20 + O(s) + 0
CO + O(s) -) CO2 + C2H4 + O(s) "-)' C2H3. + OH(s) C2H6 + O(s) "-) C2H5. + OH(s)
6.1 1016 2.8 1008 3.0 1023 2.7 1009 2.1 10l~ 2.5 10l~
-26700 8400 1O0 3000 5900 5800
574
1.5
0,5"
[] C2H4 9 CO 9
CO
CO2
0,4"
[] C2H4 9 CO 9 CO2
.. CO2
1 0,3" @ I,i
~,
CO2
CO 0,2
0.5 0,1[] II u
0
[]
[]
Z~~..~..~r~.~
C2H4
C2H4
|
|
|
2
4
6
Fig. 1. 1023 K and 1 catalyst pellet
0 8
0
2
4
1: (s)
;
Fig. 2.1173 K and 3 catalyst pellets
Experimental/caculated production of CO, C 0 2 and C2H4 (molar fraction vs gas space time)
According to this scheme, the initiation is exclusively the surface reaction: CH4 + O(s) ") CH3. + OH(s) This heterogeneous step is the main source of the free radicals involved in the chains of the gas-phase reaction. This mechanism shows that the catalytic reactions are also important for CO2 formation from CO and for C2H4 formation from CzH6.We can note that the mechanism of OCM includes two distinct reaction pathways. The first pathway leads to the formation of oxygenated compounds Ox (CO, CO2, HCHO...) and the second one leads to the formation of hydrocarbons C2+ (C2H6, C2H4, C2H2...). Therefore, when the conversion is low, C2+ and Ox selectivities are controlled by the rates of formation of C2H6 and HCHO respectively, through the reactions:
CH3. +
02
"-) HCHO + OH.
2 CH3. "-) C2H6
r = k(CH3.)(O2) r - k (CH3.)2
The C2+ selectivity increases when 02 concentration decreases and when the initiation rate increases. On the other hand, when the conversion is high, a third reaction pathway decomposes C2H4 into oxygenated species and, as a consequence, C2+ selectivity decreases.
575
100% ~.......:....~.....':+ .. * ,>
+CH4
.................~
.....................................
~
~r ~
2-o
'CH3" ~ ~~
~
+CH3.
I C2H6 ~,,=~,,I
I HCH~ 100%
+R.
40% :~i~:=:~, +R.i:~=~::~!!i=~ 60% ~:;~+O* :
/--
C2H5.
CHO.
~ ..dl 29% ~+ 0 ~2 67% + M ~ ~
i
l
Z
/
+HO 6 6 % .IZIII+O*
+0 2
"
63%
+M 35%
\ )
+02
'p,
<
90%
16% & +M
?
CO2
C2H2
......................................................................................................... H., OH., HO2. Fig. 3. Mechanism of the oxidative coupling of methane over La203 (1023 K, CH4 conversion" 10%). Black arrows are gas-phase reactions and grey arrows are surface reactions.
The presence of La203 increases the C2+ selectivity when the gas space time is low by introducing a new surface initiation reaction. Then, for higher gas space time, the secondary reaction of C2H3. oxidation decreases C2+ selectivity. This analysis explains that many authors have found that C2+ selectivity decreases drastically with increasing conversion of methane [1517].
576
5. CONCLUSION The reaction scheme proposed in this work allows to simulate OCM reaction over La203 in a catalytic jet-stirred reactor over a large range of operating conditions. Preexponential factors of heterogeneous reactions were estimated by using an original method. Finally, the reduced surface mechanism contains only 6 elementary reactions and shows that the OCM reaction is a gas-phase chain reaction coupled with surface reactions. This study suggests that the surface mechanism of OCM over La203 essentially accounts for the initiation of the reaction, the production of CO2 and the decomposition of C2H6. The OCM mechanism includes 2 reaction pathways. The first one leads to the formation of oxygenated species and the second one leads to the formation of hydrocarbons. This mechanism also explains why the hydrocarbon selectivity seems to reach a limit with increasing conversion. REFERENCES [1] P.M. Marquaire, P. Barb6, Y.D. Li, G.M. C6me and F. Baronnet, Stud. Surf. Sci. Catal., 81 (1994) 149. [2] P. Barb6, Y.D. Li, P.M. Marquaire, G.M. C6me and F. Baronnet, Catal. Today, 21 (1994) 409. [3] G.M. C6me, Y.D. Li, P. Barb6, N. Gueritey, P.M. Marquaire and F. Baronnet, Catal. Today, 30 (1996) 215. [4] P.M. Marquaire, N. Gueritey, G.M. C6me and F. Baronnet, Stud. Surf. Sci. Catal. 119 (1997) 383. [5] P. Barb6, F. Battin-Leclerc and G.M. C6me, J. Chim. Phys. 92 (1995) 1666. [6] J.G. McCarty, Methane Conversion by Oxidative Processes, E.E. Wolf Ed., Van Nostrand Reinhold, New York (1992). [7] G.A. Martin and C. Mirodatos, Fuel Processing Technology 42 (1995) 179. [8] S. Lacombe, Z. Durjanova, L. Mleczko and C. Mirodatos, Chem. Eng. Technol., 42 (1995) 216. [9] P.M. Couwenberg, Qi Chen and G.B. Marin, Ind. Eng. Chem. Res., 35 (1996) 3999. [10]D. Wolf, M. Slinko, M. Baems and E. Kurkina, Appl. Catal. A: General, 166 (1998) 47. [11]K. Coulter and D.W. Goodman, Catal. Letters, 20 (1993) 169. [12] C. Shi, M. Xu, M.P. Rosynek and J.H. Lunsford, J. Phys. Chem., 97 (1993) 216. [ 13] S.W. Benson, Thermochemical Kinetics, 2nd Edition, Wiley Interscience (1976). [ 14] S.-J. Huang, A.B. Walters, M.A. Vannice, Appl. Cat. B, 17 (1998) 183. [15]Z. Kalenik and E.E. Wolf, Methane Conversion by Oxidative Processes, E.E. Wolf Ed., Van Nostrand Reinhold, New York (1992). [16] T. Le Van, C. Louis, M. Kermarec, M. Che, J.M. Tatibou~t, Catal. Today, 13 (1992) 321. [17] T. Wakatsuki, M. Yamamura, H. Okado, K. Chaki, S. Okada, K. Inabu, S. Susuki and T. Yoshinari, Natural Gas Conversion IV, Studies in Surface Science and Catalysis, Vol. 107, Elsevier (1997).