- Email: [email protected]
Mgo Catalyst
147
A. Holmen et al. (Editors), Natural Gas Conuersion 1991 Elsevier Science PublishersB.V., Amsterdam
KINETIC STUDIES OF THE OXIDATIVE C O U P L I N G OF METHANE OVER A CE/LI/MGO C A T A L Y S T
S. BARTSCH? Ha-G.PIRKL2, W. BAUMANN:
H. HOFMANN2
Degussa AG, Forschung Organische Chemie Rodenbacher Chaussee 4, 6450 Hanau Instltut fur Technische Chemle I, Universitgt Erlangen-NUrnberg Egerlandstr. 3, 8520 Erlangen Institut flir Technlsche Chemie 11, Universltlt Erlangen-NUrnberg Egerlandstr. 3, 8520 Erlangen
SUMMARY The main reaction pathways of the oxidative coupling of CH4 to CzH6, C2H4, C O and cO2 were determined using a Ce/Ll/MgO catalyst at 750 OC. In non-catalytic runs it was found that homogeneous oxldatlon of CH4 as well as of C O is neglegible, but the conversion of C& to C2H4 and further oxidation of C2H4 into C O occure as homogeneous gas phase reactions. In the presence of the catalyst CH4 1s oxidized selectively to C2H6 as well as unselectively to C02, which is partly produced also from CzH6. C2H6 is converted to C2H4 by oxldatlve dehydrogenation as well as by dehydrogenation. C O is oxidized to C02. The rates of all these reactions were described by power law equations.
INTRODUCTION In the last five years a l o t of work has been done to develop active and selective catalysts f o r the heterogeneously catalyzed oxidative coupling of methane to Cz hydrocarbons. Typical carbon containing products that were found when investigatlng the reaction of Fig.1: General reaction scheme including all carbon containing methane and oxygen uslng approspecies priate catalysts in the temperature range of 650 - 800 OC are C2H6, CzH4, CO, C02. As it is known from the literature (ref.1) homogeneous gas phase reactions may influence the product dlstrlbution. At present it is not clear what reaction pathways are responsible for the production of C2 hydrocarbons as well as carbonoxides. The reason 1s the complex interaction between homogeneous and hetero-
148
geneously catalyzed reactions, some of which are parallel or consecutive steps (see Figure 1). This paper presents the results of our studies on the reaction scheme of the oxidative coupling of methane using a Ce/LiMgO catalyst at 750 OC. The rates of the main reaction steps were mathematically described by power law rate equations.
METHODS Experimental The Ce/Li/MgO system was used in the investigations presented in this paper because it proved to be an effective catalyst for the title react ion yielding more than 20% of C2 hydrocarbons at 750 OC (ref.2 ) . The catalyst preparation was reported previously (ref.3). The experimental set-up consisting of a feed section, a tubular reactor made of catalytically inert ceramic material (a-A1203) and a gas chromatograph was already described in detail (ref. 3). In the course of the examination of the methane c p sampling coupling reaction a new gas rnovutle capillary tube 1 sampling device has been 7 fied capillary tube .bed developed as it is shown in Figure 2. A capillary tube with an outer diameter of 2.0 mm and an inner dia- 2 meter of 1.5 mm was placed 5 - 3 in the fixed-bed of the N b reactor tube. All over the 5 6 length of the catalytic 7 section (z = 0...lo0 mm) it 8 had an axial slit of 0.5 mm 9 10 width. Inside a second capillary tube (outer diameter 1.5 mm, inner diameter 1.0 nun) having a radial hole (0.5 mm diameter) could be Fig.2: gas sampling moved up and down in order to get the gas mixture out of the reaction zone at any axial position. All component parts were made of a-A1203 to avoid undesired catalytic influences of the construction material. The flow rates through this system of capillary tubes were adjusted in such a way that the residence time of the unconverted reactants as well as the pressure drop could be neglec-
I:
1
2
149
ted. With this sampling technique the concentration profiles of the reactants as a function of the reactor length were accurately obtained. For the examination of the reaction scheme, such profiles were determined for different feed gases (CH4/02/Nz, C2H6/Oz/N2, C2H4/02/N2) under catalytic and non-catalytic conditions to distinguish between homogeneous and heterogeneously catalyzed reaction steps. The operating conditions in these runs were atmospheric pressure, T = 7 5 0 OC, F = 6 Nml/s, W = O . 1 g of catalyst per 10 mm of reactor length, particle size 0.8 - 1.0 nun. The feed gas consisted of 10 mol% reaction mixture and 90mol% N2. The ratios of hydrocarbons to oxygen were varied in the range of 2 : l to 10:f. Modeling The integral reactor used for this investigation was described by a one-dimensional pseudohomogeneous plug-flow model
-=-. dz
Tvll' I'
( R = 8.314 J/(K.mol), u = gas velocity, uij = stolchiometric coefficient of component 1 with respect to reaction j, rj = rate of reac-
tion j) that implies no volumetric change of the reaction mixture as well as isothermalandisobarlcconditionswith respect to the reactor length. These preconditions were fulfilled under the operating conditions stated above. The deviations of pressure and molar volume along the catalytic fixed-bed were less than 1% with respect to the values at the inlet of the reaction zone (z = 0 ) . Temperature was constant to within +/- 3 K throughout the whole catalytic section. Insertion of j rate equations of the type
into the reactor model leads to a set of 1 nonlinear differential equations that are simultaneously integrated by a 4 t h order RungeKutta method to calculate the values of pi as a function of z. In this type of equation kj denotes the rate constant of reaction j and ni, the order of reaction j with respect of the reactant i. In the case of heterogeneously catalyzed reactions kj is equivalent to the term k j , r * ~ s( ~ =s 127-3 kgcat/m3reactron volume * In order to f i t the experimental data the values of kj and nrj were systematically varied by a computer program including a least square minimization procedure. The significance of each parameter is strongly affected by the
150
number of unknown parameters which consequently has to be kept as low as possible. Therefore the over all reaction network was devided into four subunits that were considered to be separable without any interaction. In a first step the oxidation of ethylene was examined (ref.4). Next, the homogeneous gas phase oxidation of ethane was studied. Then the catalytic oxidation of ethane and finally the methane oxidation was investigated. This approach leads to a formal mathematical description that is appropriate tocalculate thecontributionof the main reactions in the total reaction system. But the rate equations are not mechanistically founded. RESULTS Ethylene oxidation In the non-catalytic runs the predominant reaction product was CO, which is therefore considered as a primary product, while only trace amounts of C02 were formed. In the catalytic runs the concentration of CO went through a maximum with respect to the reactor length and a remarkable increase of the C02 concentration was found at the same time when CO decreased. The sum of CO plus C02 was only marginally increased compared with the non-catalytic results. Because of these observations it is assumed that C2H4 reacts in the gas phase to form CO, which is oxidized to C02 by a consecutive heterogeneously catalyzed surface reaction.
Pi 1103 .Pal 8
7
-
calculated
4
3 2 1
0
0
1
2
3
-
4
5
6
2
7 Icml
Fig.3: ethylene oxidation (T = 75OoC, p = 1.2 bar, F = 6Nml/s, W = 0.1 g/cm, Ce/Li/MgO catalyst)
151
-(a)nz
Ethylene oxidation was described by the following two equations: CzH4 + 2 0 2
co
+ 0.50,
& r2
2 CO + 2 H z O
rt =
coz
r2
kl
- (m) Pc,n. "t
= kP,..P..
(1)
( pco )"" RT
* (*)"4
RT
( 2)
The best parameter values are: = 31,O
k,
n~~.~/(rnol'*~
k2.S = 0,29 m4*5/(rno10-s * s * kg,,J
nt = 1.0
"2
= 1.4
n3 = 1.0
1l4
= 0.5
In Figure 3 a typical profile of partial pressures with respect to the reactor length is shown. The symbols represent the experimental data, the lines are calculated. Homogeneous ethane oxidation The conversion of ethane in the gas phase without using a catalyst was always less than 10 %. The only product was ethylene, no hydrogen could be detected. The consecutive reaction of ethylene to carbonmonoxide ( 1 ) was not observed because of the very low concentration of ethylene. A strongly linear decrease of the partial pressure of ethane with respect to the reactor length was observed (see Figure 41, indicating that the order of the reaction was zero. The rate of the homogeneous oxidation of ethane CzH6 + 0 . 5 0 2 + CzH4 + HzO
(3)
was determined to r = k = 0.52 mol/(m3.s).
'
6 4
-
z [cml Fig.4: Homogeneous ethane oxidation ( T = 75OoC, p = 1.2 bar, F = 6 Nml/s, W = 0.1 g/cm, Ce/Li/MgO catalyst)
152
Catalytic ethane oxidation The main products found in this experimental series were CzH4, COz, H2 and HzO. On the basis of the following considerations and calculations a set of three reactions was postulated in order to gain an appropriate description: Hydrogen is produced from ethane by a heterogeneously catalyzed step, because it was not found in the non-catalytic runs. The calculated concentration of ethylene corresponding to the amount of hydrogen plus ethylene formationby reaction ( 3 ) is lower than observed. Consequently some ethylene is formed vla oxidative dehydrogenation. C02 production is higher than expected by calculating the effect of the interaction of the reactions (1). (21, (3). Therefore a third equation for the total oxidation of ethane was taken into account .
4 P 0
P
lC2Hg) 1 C2Hl,)
p cco,, A p (021
14
x
12
4-
0
p IHZO) p (HZ1
- calculated
The resulting system of reactions describing the catalytic ethane oxidation is C2H6+O.5O2 C2H6
+
C2H4+ H20
rj =
kj,,.
9.
(*)"I
- (&)n2
(4)
153
wlththeparametervalues:
kt,. = 0.031 m4*'/(kgOat
S
n 2 = 0.5
n1 = I
mo14")
12.8"
0.024 m4.S/(ke,,t.s.molo.S)
n a = 1.5
kj,.=
0.006 ms~1/(kgoat~s~mo10~7)
n4= 1
n s = 0.7
In Figure 5 the experimental data as well as the calculated concentrationprofiles includingthereactions (1). ( 2 ) and ( 3 ) are shown. Methane oxidation Since the conversions of methane and oxygen in the homogeneous gas phase can be neglected (ref. 3 1 , the methane oxidation was investlgated only under catalytic conditions. It was found that two reactions are necessary to complete the total system. Methaneis converted selectively to ethane as well as unselectively to carbondioxide.
CHI
+
C02 + 2 H 2 0
2 O2
12
= k2,.*p.
-(s)n
.(%)ns
(8)
The parameter values were determined to
kl
= 0.0058 m3*6/(mo10.2*s.kg,&
k2,s =
0.012 m1*6-molo.4/(s*kgoat)
0
1
2
3
4
-
5 6
7
8
nl = 2
n2 = 1.2
n3 = 0
"4
= 0.6
9 10 t[cml
Flg.6: Methane coupling (T= 75OoC, p = 1.2 bar, F = 6 Nml/s, W = 0.1 g/cm, Ce/Li/MgO catalyst 1
154
A typical result of the methane coupling reaction that involves all
react ions
( 1)
-
( 8 ) is presented in Figure 6.
Reaction scheme The total reaction system using a Ce/Li/MgO catalyst at eight reactions, two of which tions ( Cprl = Pa, [TI = K, Cr,l 2CH4 + 0 . 5 0 2 CH4 + 2 0 2
of the oxidative coupling of methane 750 OC is quantitatively described by occure as homogeneous gas phase reac= mol/(m3-sl ) .
4
C2H6+ H 2 0
r = 0.0058
__+
C 0 2 +2H2O
r = 0.012 r = 0.031
.
. Q
. (*)2R T . (%)" RT
. .Q . (RT P" )"*6 . (+)"" 9. , (*)'
*
I T
Fig. 7: Catalytic (right hand) and non-catalytic (left hand) reactlon scheme (Ce/Li/MgO catalyst, T = 750 C ) Figure 7 summarizes the main reaction pathways that were found in the catalytic (right hand scheme) and non-catalytic (left hand scheme) experiments. Theproduction of methanol and formaldehyde was observed in very trace amounts. Other side reactions were the formation of methane from ethane and ethylene as well as direct oxidation of ethylene to carbondioxide. But the influences of these effects were negligible. REFERENCES 1 D.J.C. Yates, N.E. Zlotln, J . Catal 111 (1988)317 2 J. Schleblsch, S.Bartsch, H. Hofmann, to be published 3 S. Bartsch, J. Falkowski, H. Hofmann, Catalysis Today, 4 (1989) 421 4 S. Bartsch, H, Hofmann, Catalysis Today, 6 (1990) 527
A. Holmen et al. (Editors), Natural Gas Conuersion 1991 Elsevier Science PublishersB.V., Amsterdam
KINETIC STUDIES OF THE OXIDATIVE C O U P L I N G OF METHANE OVER A CE/LI/MGO C A T A L Y S T
S. BARTSCH? Ha-G.PIRKL2, W. BAUMANN:
H. HOFMANN2
Degussa AG, Forschung Organische Chemie Rodenbacher Chaussee 4, 6450 Hanau Instltut fur Technische Chemle I, Universitgt Erlangen-NUrnberg Egerlandstr. 3, 8520 Erlangen Institut flir Technlsche Chemie 11, Universltlt Erlangen-NUrnberg Egerlandstr. 3, 8520 Erlangen
SUMMARY The main reaction pathways of the oxidative coupling of CH4 to CzH6, C2H4, C O and cO2 were determined using a Ce/Ll/MgO catalyst at 750 OC. In non-catalytic runs it was found that homogeneous oxldatlon of CH4 as well as of C O is neglegible, but the conversion of C& to C2H4 and further oxidation of C2H4 into C O occure as homogeneous gas phase reactions. In the presence of the catalyst CH4 1s oxidized selectively to C2H6 as well as unselectively to C02, which is partly produced also from CzH6. C2H6 is converted to C2H4 by oxldatlve dehydrogenation as well as by dehydrogenation. C O is oxidized to C02. The rates of all these reactions were described by power law equations.
INTRODUCTION In the last five years a l o t of work has been done to develop active and selective catalysts f o r the heterogeneously catalyzed oxidative coupling of methane to Cz hydrocarbons. Typical carbon containing products that were found when investigatlng the reaction of Fig.1: General reaction scheme including all carbon containing methane and oxygen uslng approspecies priate catalysts in the temperature range of 650 - 800 OC are C2H6, CzH4, CO, C02. As it is known from the literature (ref.1) homogeneous gas phase reactions may influence the product dlstrlbution. At present it is not clear what reaction pathways are responsible for the production of C2 hydrocarbons as well as carbonoxides. The reason 1s the complex interaction between homogeneous and hetero-
148
geneously catalyzed reactions, some of which are parallel or consecutive steps (see Figure 1). This paper presents the results of our studies on the reaction scheme of the oxidative coupling of methane using a Ce/LiMgO catalyst at 750 OC. The rates of the main reaction steps were mathematically described by power law rate equations.
METHODS Experimental The Ce/Li/MgO system was used in the investigations presented in this paper because it proved to be an effective catalyst for the title react ion yielding more than 20% of C2 hydrocarbons at 750 OC (ref.2 ) . The catalyst preparation was reported previously (ref.3). The experimental set-up consisting of a feed section, a tubular reactor made of catalytically inert ceramic material (a-A1203) and a gas chromatograph was already described in detail (ref. 3). In the course of the examination of the methane c p sampling coupling reaction a new gas rnovutle capillary tube 1 sampling device has been 7 fied capillary tube .bed developed as it is shown in Figure 2. A capillary tube with an outer diameter of 2.0 mm and an inner dia- 2 meter of 1.5 mm was placed 5 - 3 in the fixed-bed of the N b reactor tube. All over the 5 6 length of the catalytic 7 section (z = 0...lo0 mm) it 8 had an axial slit of 0.5 mm 9 10 width. Inside a second capillary tube (outer diameter 1.5 mm, inner diameter 1.0 nun) having a radial hole (0.5 mm diameter) could be Fig.2: gas sampling moved up and down in order to get the gas mixture out of the reaction zone at any axial position. All component parts were made of a-A1203 to avoid undesired catalytic influences of the construction material. The flow rates through this system of capillary tubes were adjusted in such a way that the residence time of the unconverted reactants as well as the pressure drop could be neglec-
I:
1
2
149
ted. With this sampling technique the concentration profiles of the reactants as a function of the reactor length were accurately obtained. For the examination of the reaction scheme, such profiles were determined for different feed gases (CH4/02/Nz, C2H6/Oz/N2, C2H4/02/N2) under catalytic and non-catalytic conditions to distinguish between homogeneous and heterogeneously catalyzed reaction steps. The operating conditions in these runs were atmospheric pressure, T = 7 5 0 OC, F = 6 Nml/s, W = O . 1 g of catalyst per 10 mm of reactor length, particle size 0.8 - 1.0 nun. The feed gas consisted of 10 mol% reaction mixture and 90mol% N2. The ratios of hydrocarbons to oxygen were varied in the range of 2 : l to 10:f. Modeling The integral reactor used for this investigation was described by a one-dimensional pseudohomogeneous plug-flow model
-=-. dz
Tvll' I'
( R = 8.314 J/(K.mol), u = gas velocity, uij = stolchiometric coefficient of component 1 with respect to reaction j, rj = rate of reac-
tion j) that implies no volumetric change of the reaction mixture as well as isothermalandisobarlcconditionswith respect to the reactor length. These preconditions were fulfilled under the operating conditions stated above. The deviations of pressure and molar volume along the catalytic fixed-bed were less than 1% with respect to the values at the inlet of the reaction zone (z = 0 ) . Temperature was constant to within +/- 3 K throughout the whole catalytic section. Insertion of j rate equations of the type
into the reactor model leads to a set of 1 nonlinear differential equations that are simultaneously integrated by a 4 t h order RungeKutta method to calculate the values of pi as a function of z. In this type of equation kj denotes the rate constant of reaction j and ni, the order of reaction j with respect of the reactant i. In the case of heterogeneously catalyzed reactions kj is equivalent to the term k j , r * ~ s( ~ =s 127-3 kgcat/m3reactron volume * In order to f i t the experimental data the values of kj and nrj were systematically varied by a computer program including a least square minimization procedure. The significance of each parameter is strongly affected by the
150
number of unknown parameters which consequently has to be kept as low as possible. Therefore the over all reaction network was devided into four subunits that were considered to be separable without any interaction. In a first step the oxidation of ethylene was examined (ref.4). Next, the homogeneous gas phase oxidation of ethane was studied. Then the catalytic oxidation of ethane and finally the methane oxidation was investigated. This approach leads to a formal mathematical description that is appropriate tocalculate thecontributionof the main reactions in the total reaction system. But the rate equations are not mechanistically founded. RESULTS Ethylene oxidation In the non-catalytic runs the predominant reaction product was CO, which is therefore considered as a primary product, while only trace amounts of C02 were formed. In the catalytic runs the concentration of CO went through a maximum with respect to the reactor length and a remarkable increase of the C02 concentration was found at the same time when CO decreased. The sum of CO plus C02 was only marginally increased compared with the non-catalytic results. Because of these observations it is assumed that C2H4 reacts in the gas phase to form CO, which is oxidized to C02 by a consecutive heterogeneously catalyzed surface reaction.
Pi 1103 .Pal 8
7
-
calculated
4
3 2 1
0
0
1
2
3
-
4
5
6
2
7 Icml
Fig.3: ethylene oxidation (T = 75OoC, p = 1.2 bar, F = 6Nml/s, W = 0.1 g/cm, Ce/Li/MgO catalyst)
151
-(a)nz
Ethylene oxidation was described by the following two equations: CzH4 + 2 0 2
co
+ 0.50,
& r2
2 CO + 2 H z O
rt =
coz
r2
kl
- (m) Pc,n. "t
= kP,..P..
(1)
( pco )"" RT
* (*)"4
RT
( 2)
The best parameter values are: = 31,O
k,
n~~.~/(rnol'*~
k2.S = 0,29 m4*5/(rno10-s * s * kg,,J
nt = 1.0
"2
= 1.4
n3 = 1.0
1l4
= 0.5
In Figure 3 a typical profile of partial pressures with respect to the reactor length is shown. The symbols represent the experimental data, the lines are calculated. Homogeneous ethane oxidation The conversion of ethane in the gas phase without using a catalyst was always less than 10 %. The only product was ethylene, no hydrogen could be detected. The consecutive reaction of ethylene to carbonmonoxide ( 1 ) was not observed because of the very low concentration of ethylene. A strongly linear decrease of the partial pressure of ethane with respect to the reactor length was observed (see Figure 41, indicating that the order of the reaction was zero. The rate of the homogeneous oxidation of ethane CzH6 + 0 . 5 0 2 + CzH4 + HzO
(3)
was determined to r = k = 0.52 mol/(m3.s).
'
6 4
-
z [cml Fig.4: Homogeneous ethane oxidation ( T = 75OoC, p = 1.2 bar, F = 6 Nml/s, W = 0.1 g/cm, Ce/Li/MgO catalyst)
152
Catalytic ethane oxidation The main products found in this experimental series were CzH4, COz, H2 and HzO. On the basis of the following considerations and calculations a set of three reactions was postulated in order to gain an appropriate description: Hydrogen is produced from ethane by a heterogeneously catalyzed step, because it was not found in the non-catalytic runs. The calculated concentration of ethylene corresponding to the amount of hydrogen plus ethylene formationby reaction ( 3 ) is lower than observed. Consequently some ethylene is formed vla oxidative dehydrogenation. C02 production is higher than expected by calculating the effect of the interaction of the reactions (1). (21, (3). Therefore a third equation for the total oxidation of ethane was taken into account .
4 P 0
P
lC2Hg) 1 C2Hl,)
p cco,, A p (021
14
x
12
4-
0
p IHZO) p (HZ1
- calculated
The resulting system of reactions describing the catalytic ethane oxidation is C2H6+O.5O2 C2H6
+
C2H4+ H20
rj =
kj,,.
9.
(*)"I
- (&)n2
(4)
153
wlththeparametervalues:
kt,. = 0.031 m4*'/(kgOat
S
n 2 = 0.5
n1 = I
mo14")
12.8"
0.024 m4.S/(ke,,t.s.molo.S)
n a = 1.5
kj,.=
0.006 ms~1/(kgoat~s~mo10~7)
n4= 1
n s = 0.7
In Figure 5 the experimental data as well as the calculated concentrationprofiles includingthereactions (1). ( 2 ) and ( 3 ) are shown. Methane oxidation Since the conversions of methane and oxygen in the homogeneous gas phase can be neglected (ref. 3 1 , the methane oxidation was investlgated only under catalytic conditions. It was found that two reactions are necessary to complete the total system. Methaneis converted selectively to ethane as well as unselectively to carbondioxide.
CHI
+
C02 + 2 H 2 0
2 O2
12
= k2,.*p.
-(s)n
.(%)ns
(8)
The parameter values were determined to
kl
= 0.0058 m3*6/(mo10.2*s.kg,&
k2,s =
0.012 m1*6-molo.4/(s*kgoat)
0
1
2
3
4
-
5 6
7
8
nl = 2
n2 = 1.2
n3 = 0
"4
= 0.6
9 10 t[cml
Flg.6: Methane coupling (T= 75OoC, p = 1.2 bar, F = 6 Nml/s, W = 0.1 g/cm, Ce/Li/MgO catalyst 1
154
A typical result of the methane coupling reaction that involves all
react ions
( 1)
-
( 8 ) is presented in Figure 6.
Reaction scheme The total reaction system using a Ce/Li/MgO catalyst at eight reactions, two of which tions ( Cprl = Pa, [TI = K, Cr,l 2CH4 + 0 . 5 0 2 CH4 + 2 0 2
of the oxidative coupling of methane 750 OC is quantitatively described by occure as homogeneous gas phase reac= mol/(m3-sl ) .
4
C2H6+ H 2 0
r = 0.0058
__+
C 0 2 +2H2O
r = 0.012 r = 0.031
.
. Q
. (*)2R T . (%)" RT
. .Q . (RT P" )"*6 . (+)"" 9. , (*)'
*
I T
Fig. 7: Catalytic (right hand) and non-catalytic (left hand) reactlon scheme (Ce/Li/MgO catalyst, T = 750 C ) Figure 7 summarizes the main reaction pathways that were found in the catalytic (right hand scheme) and non-catalytic (left hand scheme) experiments. Theproduction of methanol and formaldehyde was observed in very trace amounts. Other side reactions were the formation of methane from ethane and ethylene as well as direct oxidation of ethylene to carbondioxide. But the influences of these effects were negligible. REFERENCES 1 D.J.C. Yates, N.E. Zlotln, J . Catal 111 (1988)317 2 J. Schleblsch, S.Bartsch, H. Hofmann, to be published 3 S. Bartsch, J. Falkowski, H. Hofmann, Catalysis Today, 4 (1989) 421 4 S. Bartsch, H, Hofmann, Catalysis Today, 6 (1990) 527