Ind. Eng. Chem. Res. 2009, 48, 677–684
677
Kinetic Studies of Carbon Dioxide Reforming of Methane over Ni-Co/Al-Mg-O Bimetallic Catalyst Jianguo Zhang, Hui Wang,* and Ajay K. Dalai Department of Chemical Engineering, UniVersity of Saskatchewan, Saskatoon, SK, Canada S7N 5A9
The kinetics of CO2 reforming with CH4 over a Ni-Co/Al-Mg-O bimetallic catalyst was investigated in a fixed bed reactor at a temperature range of 650-750 °C and the partial pressures of CO2 and CH4 ranging from 30 to 190 kPa. Owing to the simultaneous occurrence of the CO2 reforming reaction and the reverse water-gas shift reaction (RWGS) in the system, the apparent activation energies with respect to reactant consumption and product formation were found different and they are 69.4 and 25.9 kJ/mol for CH4 and CO2 consumption and 85.1 and 61.8 kJ/mol for H2 and CO formation, respectively. It was also found that the reforming rate in terms of CH4 consumption was less sensitive to CO2 partial pressures but had stronger dependence on CH4 partial pressures. At a constant CH4 partial pressure, the increase in CO2 partial pressure did not cause significant change in the reforming rate, whereas at a constant CO2 partial pressure the reforming rate increased with the increase in CH4 partial pressure. The increase in extra CO2 at a constant CH4 pressure led to decreases in hydrogen (H2) formation but increase in carbon monoxide (CO) formation due to the simultaneous occurrence of the reverse water-gas shift reaction. A Langmuir-Hinshelwood (L–H) model was also developed assuming that the dissociation of CH4 and the reaction between the carbon species and the activated carbon dioxide are the rate determining steps over the Ni–Co/Al–Mg–O. It satisfactorily fits the experimental data as well. 1. Introduction The concerns with climate change due to the use of fossil fuels have led to considering production and use of clean fuels and renewable energies. The application of biogas generated from landfills, which is composed of around 50% CH4 and 50% CO2, has drawn attention in both areas of greenhouse gas mitigation and clean fuel production. It has shown exciting promise that landfill gas is converted to synthetic gas, a mixture of H2 and CO, via CO2 reforming of methane (eq 1) of which the energy demand is from solar energy.1 The catalytic reaction of CO2 reforming of CH4 has been extensively studied over a variety of catalysts of both noble metals and non-noble metals.2-4 However, a problem associated with CO2 reforming of CH4 is the severe deactivation of catalysts with time-onstream, which is believed to be the major barrier of wide application of the technologies based on CO2 reforming of CH4. CO2+CH4 f 2CO + 2H2
(1)
It has been found that supported Rh, Ru, Pd, Pt, and Ir catalysts could provide stable operations for CO2 reforming of CH4 with low carbon formation during the reaction.2 However, from an industrial standpoint, it is more practical to develop non-noble metal-based catalysts to avoid the high cost and restricted availability of noble metals. Ni has drawn remarkable attention in this area due to its high activity to the reaction and wide availability.5 However, Ni catalysts suffer more severe catalyst deactivation than noble metal-based catalysts due to sintering, metal oxidation,6 or, especially, the severe carbon formation.2,5,7 Kinetic studies have been performed over different Ni-based catalysts for CO2 reforming of methane. As early as 1960s, Bodrov and Apel’baum8 had reported that the rate of CO2 reforming of CH4 over nickel surface in the temperature range * To whom correspondence should be addressed. E-mail: hui.wang@ usask.ca. Tel.: +306 966 2685. Fax: +306 966 4777.
of 800-900 °C can be described by the same kinetic equation as that for steam (H2O) reforming of CH4. Bodrov’s model is based on the assumption that CH4 does not react with CO2 but with steam; the H2 formed reacts rapidly with CO2 by the reverse water-gas shift reaction and the steam is thus regained. Assuming that CH4 dissociation is the rate determining step, Zhang and Verykios9 developed a Langmuir-Hinshelwood model for CO2 reforming of CH4 over Ni/CaO-Al2O3. Using isotopic studies of CH4/CD4 in CO2 reforming of CH4 over Ni/ SiO2 catalyst, Wang and Au verified the dissociation of CH4 to be the rate-determining step.10 However, Slagtern et al. reported that the surface reaction between carbon species from CH4 cracking and oxygen-adsorbed species derived from CO2 dissociation is the rate-determining step.6 Hu and Ruckenstein11 also concluded that the surface reaction between carbon and oxygen species is the rate-determining step over Ni catalysts through transient response analysis. The above review for literature suggests that the CH4 cracking and the surface reaction between carbon species and oxygen species are generally accepted by different research groups as rate-determining steps over various Ni-based catalysts. Bradfrod and Vannice12 investigated kinetics of CO2 reforming of CH4 over various Ni catalysts and developed a kinetic model based on CH4 dissociative adsorption to form CHx species and CHxO decomposition as rate-determining steps. Also, Tsipouriari and Verykios13 reported another kinetic model by assuming that the CH4 cracking and surface reaction between C and oxycarbonate species are rate-determining steps over Ni/La2O3 catalyst. Even though different rate-determining steps are suggested in mechanism studies, Langmuir-Hinshelwood type of model has been commonly used in the kinetic studies of CO2 reforming of CH4.2,9,12-16 Nevertheless, the L-H model has different formula depending on catalyst systems. Therefore, it is necessary to develop a new kinetic model for our catalyst as there is no general kinetic model available for CO2 reforming of CH4.
10.1021/ie801078p CCC: $40.75 2009 American Chemical Society Published on Web 12/15/2008
678 Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009
Figure 1. Schematic of the reaction system.
In our previous work, we have developed a series of bimetallic catalysts for CO2 reforming of CH4.17 It was found that the Ni-Co/Al-Mg-O bimetallic catalysts have excellent activity and stability for CO2 reforming of CH4. Optimizing Ni-Co content leads to low carbon or carbon-free operation of CO2 reforming of CH4.18 At the conditions of 750 °C, 1 atm, 180000 mL/(gcat · h), and CH4/CO2/N2 ) 1/1/1, 250 h carbon free operation for CO2 reforming of CH4 have been achieved over Ni-Co/Al-Mg-O bimetallic catalysts with lower Ni-Co content.18 In this work, experimental and kinetic studies of CO2 reforming of CH4 have been performed over the Ni-Co/ Al-Mg-O bimetallic catalyst with 4 mol % Ni and 5 mol % Co. The effects of partial pressures of reactants have been investigated at a temperature range of 650-750 °C and total pressure of 310 kPa. Apparent energies for reactant consumption and product formation were estimated. Both power-law model and L-H model were developed and the kinetic parameters for CO2 reforming of CH4 using Ni-Co/Al-Mg-O catalyst were evaluated. 2. Experimental Details The kinetic data of CO2 reforming of CH4 reaction over the Ni-Co/Al-Mg-O bimetallic catalyst was collected using a bench-scale fixed-bed quartz reactor with an inner diameter of 6 mm (Autoclave Engineers). The reactor system is shown in Figure 1. Ni-Co/Al-Mg-O bimetallic catalyst was prepared by coprecipitating a common aqueous solution of nickel nitrate (98%, Lancaster Synthesis Inc.), cobalt nitrate (99%, Aldrich Chemical Co.), magnesium nitrate (EMD Chemicals Inc., Mexico), and aluminum nitrate (EMD Chemical Inc., Germany). The precipitate was washed with deionized water, dried overnight in the air at 120 °C, and then calcined in the air at 900 °C. The prepared catalyst contains 4 mol % Ni, 5 mol % Co, 30 mol % Al, and 61 mol % Mg. The oxygen content is assumed to be stoichiometrically balanced. The catalyst was represented by Ni-Co/Al-Mg-O. Except the runs attempting to select experimental conditions to eliminate mass transfer effects where different space velocities and catalyst particles were used, the experimental design for kinetics measurements and CH4 and CO2 conversions are given in Table 1. For each level of the reaction temperature from 650 to 750 °C, 0.01 g catalyst with an average particle size of 0.165 mm was diluted with quartz sands of the same particle size and placed in the reactor. The Ni-Co/Al-Mg-O bimetallic catalyst was then reduced in the flow of 30 mL/min H2-N2 mixture with 20% H2. The reduction temperature was ramped at 1 °C/ min to 800 °C and held there for 4 h to allow active metals to be formed in the catalyst. Then the temperature was changed
Table 1. Kinetic Experiment Design and Reactant Conversion of the CO2 Reforming of CH4 over Ni-Co/Al-Mg–O Catalyst inlet partial pressure (kPa) run
temperature (°C)
CH4
CO2
N2
1-1 1-2 1-3 1-4 1-5 1-6 1-7 1-8 2-1 2-2 2-3 2-4 2-5 2-6 2-7 2-8 3-1 3-2 3-3 3-4 3-5 3-6 3-7 3-8
750 750 750 750 750 750 750 750 700 700 700 700 700 700 700 700 650 650 650 650 650 650 650 650
62 62 62 62 31 124 186 62 62 62 62 62 31 124 186 62 62 62 62 62 31 124 186 62
31 62 124 186 62 62 62 62 31 62 124 186 62 62 62 62 31 62 124 186 62 62 62 62
217 186 124 62 217 124 62 186 217 186 124 62 217 124 62 186 217 186 124 62 217 124 62 186
conversion (%) CH4 18.2 26.2 32.2 32.8 35.4 17.2 16.8 24.6 15.8 19.4 23.0 23.8 30.3 13.7 11.6 19.0 8.2 10.9 13.2 14.1 19.8 8.6 6.8 11.3
CO2 47.2 41.2 27.2 22.3 29.7 35.1 48.8 39.8 33.7 36.6 20.9 16.6 25.6 28.2 28.6 37.1 20.6 29.7 14.2 11.5 18.3 19.2 18.7 29.5
to the reaction temperature. A gas stream of CH4/CO2/N2 in total flow rate of 300 mL/min and total pressure of 310 kPa was fed to the reactor. The effluent gas stream was cooled down to room temperature to remove produced vapor and the composition of the remaining stream was analyzed using an online GC (Agilent 6890N) equipped with a thermal conductivity detector (TCD). The first kinetic data was collected when the catalyst had been stabilized in a reactant stream for about 24 h. Then the partial pressures of CH4 and CO2 varied as given in Table 1, and the corresponding data were collected. Last, the partial pressures were set back to an earlier setting and one more set of data was obtained to confirm the catalyst stability. N2 was used in the feed gas stream as a balance inert gas to maintain the overall pressure constant when CH4 or CO2 partial pressure was varied. It was also used as an internal standard for GC analysis. The reactor pressure was controlled by a back pressure regulator. And the flow rate of gases was set by mass flow controllers (Brooks Instruments). 3. Results and Discussion 3.1. Catalyst Stability. The catalyst used for kinetic study contained 4 mol % Ni, 5 mol % Co, 30 mol % Al, and 61 mol % Mg (on the metal basis), the same composition as that denoted
Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009 679
Figure 2. Effect of flow velocity on reforming rate over Ni-Co/Al-Mg-O catalyst at 750 °C, 310 kPa, and W/F ) 0.002 (gcat · s)/mL.
as 0.04Ni0.05Co in our previous study, where no detectable carbon deposition was observed on the catalyst after a 250 h test under 750 °C and atmospheric pressure for CO2 reforming of CH4. Thermo-gravimetric/differential thermal analyzer (TG/ DTA) was used to determine the amount of carbon deposited on the spent catalyst. The method can measure as low as 1 × 10-6 g for 0.03 g catalyst loading but it did not observe carbon deposition. In this study, one catalyst loading was used for kinetic measurements at a temperature level (650, 700, or 750 °C) where the partial pressures of CH4 and CO2 varied. With the same catalyst at the same temperature, two identical runs with the same partial pressures of CH4 and CO2, one at the beginning and the other at the end, were conducted to monitor the catalyst stability. The experiments arrangement and the conversions of CH4 and CO2 are shown in Table 1. Comparison of the conversion data between the identical runs, 1-2 and 1-8, 2-2 and 2-8, and 3-2 and 3-8, respectively, indicates no significant catalyst deactivation. Therefore, it is believed that the catalyst was stable during the kinetic measurements. The activity of this catalyst was extremely high and stable. Even though the space velocity in the kinetics study was maintained as high as 1 800 000 mL/(gcat · s), the conversion of CH4 changes from 6.8 to 35.3%, and that of CO2 from 11.5 to 48.8% at temperatures ranging from 650 to 750 °C and CH4 and CO2 partial pressures ranging from 31 to 186 kPa. However, these numbers are significantly less than equilibrium conversions given by Bradford and Vannice.2 We assume they are proper for apparent kinetics study. 3.2. Effects of Mass Transfer and Heat Transfer. The effects of internal and external mass transfer were evaluated experimentally with a feed mixture composed of CH4/CO2/N2 ) 1/1/3 at 750 °C and 310 kPa. In a fixed bed reactor, the limitation of external mass transfer is typically tested through the variation of feed velocity.13,16 At a fixed residence time obtained through maintaining a constant W/F ratio, higher flow velocity leads to higher mass transfer rate. When the reforming rate does not change with the change of flow rates, limitation of external mass transfer is negligible. In this work, the reforming rate of CO2 reforming of CH4 is expressed in terms of CH4 consumption rate. Figure 2 shows the reforming rate as a function of feed flow velocity at a constant W/F of 0.002 gcat · s/ mL. It is clear that the reforming rate increased with the increase in feed flow. However, the reforming rate barely changed when the feed flow rate was increased from 200 to 400 mL/min. It indicats that the external mass transfer effect can be neglected when the feed flow velocity is higher than 200 mL/min. In this work, the kinetic studies were conducted at a feed flow velocity of 300 mL/min. In heterogeneous catalysis, the variation of catalyst particle size is usually used to check the limitation of internal mass transfer.13,16 Internal mass transfer is dependent on particle
Figure 3. Effect of catalyst particle size on reforming rate over Ni-Co/ Al-Mg-O catalyst at 750 °C, 310 kPa, and W/F ) 0.002 gcat · s/mL.
radius but kinetics is independent of radius.19 A clear change in reaction rate with the change in particle size under certain reaction conditions indicates the internal mass transfer controlling regime. As shown in Figure 3, the effect of internal mass transfer was tested using four different catalyst particle sizes to carry out CO2 reforming of CH4 at 750 °C, 310 kPa, and W/F of 0.002 gcat · s/mL. It can be seen that the reforming rate was increased when the catalyst particle size was reduced from 0.300 to 0.200 mm. However, only a slight change in reforming rate was observed when the particle size was further reduced from 0.200 to 0.085 mm. The results shown in Figure 3 suggest that the internal mass transfer can be neglected when catalyst particle size is less than 0.200 mm. Therefore, a catalyst particle size of 0.165 mm was selected for the kinetics study in this work. The effects of heat transfer limitations were also ruled out using Anderson criterion20 which is shown by |(-∆Hr)|〈R〉dp2 TsR < 0.75 4λpTs Ea
(2)
where ∆Hr is reaction enthalpy (275.6 kJ/mol at 750 °C), 〈R〉 is average reaction rate per unit catalyst particle volume (0.6 mol/(m3 · s)), dp is the average catalyst particle diameter (0.000165 m), λp is the thermal conductivity of catalyst particle (7.45 W/m/K at 750 °C, which was estimated using the thermal conductivity of Al2O3 and MgO considering that they are the main compositions of the catalyst), Ts is the surface temperature of catalyst particle (1023 K), R is the gas constant (8.314 Pa · m3/ (mol · K)), and Ea is the apparent activation energy (69.4 kJ/ mol). When substituting the numbers into eq 2, the value of the left-hand side is 1.5 × 10-7, which is far less than the value of the right-hand side, 9.2 × 10-2. It was clearly estimated that the catalyst particle was at isothermal condition under the reaction conditions using Anderson criterion. 3.3. Effect of Partial Pressure on Reforming Rate. The influence of partial pressures of CH4 and CO2 on the rate of CO2 reforming of CH4 reaction rate was carried out over the Ni-Co/Al-Mg-O bimetallic catalyst at 310 kPa and W/F 0.002 gcat · s/mL, and 650-750 °C. The effect of CO2 partial pressure at different temperatures was investigated by applying a constant CH4 partial pressure of 62 kPa while the CO2 partial pressure was varied between 31 and 186 kPa as shown in Figure 4. It can be seen that the reforming rate was increased with the increase in CO2 partial pressure in the range 31-124 kPa. However, the increase of reforming rate became slower at higher CO2 partial pressures in the range 124-186 kPa. The relatively constant reaction rate in the range 124-186 kPa can be ascribed to the restriction of thermodynamic equilibrium at limited availability of CH4 (62 kPa). Also, it can be seen that at a lower temperature, the effect of CO2 partial pressure was less significant in comparison with those at higher temperatures. The effects of CH4 partial pressures on the reaction rate of CH4 were studied at a constant CO2 partial pressure of 62 kPa.
680 Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009
Figure 4. Effect of CO2 partial pressure on the reforming rates of CO2 reforming of CH4 over Ni-Co/Al-Mg-O catalyst at 310 kPa and W/F ) 0.002 gcat · s/mL.
Figure 6. Effect of reactant partial pressures on the formation rates of CO over Ni-Co/Al-Mg-O catalyst at 310 kPa and 0.002 gcat · s/mL: (A) effect of CO2 partial pressure; (B) effect of CH4 partial pressure.
Figure 5. Effect of CH4 partial pressure on the reforming rates of CO2 reforming of CH4 over Ni-Co/Al-Mg-O catalyst at 310 kPa and W/F ) 0.002 gcat · s/mL.
Results at different temperatures are shown in Figure 5. It is clear that the reforming rate was affected by the partial pressure of CH4 in the whole experimental range 31-186 kPa. The acceleration of reforming rate slowly decreased with the increase in CH4 partial pressure at a constant CO2 partial pressure. Comparison of data between Figure 4 and Figure 5 indicates that the reforming rate was more sensitive to CH4 partial pressure than to CO2 partial pressure. It probably can be attributed to that the adsorption of CH4 to the surface of catalyst was stronger than that of CO2 at higher partial pressures. 3.4. Effect of Reactants’ Partial Pressure on Formation Rate of Products. Figure 6 shows the effect of reactant partial pressures on the formation rate of CO at 750, 700, and 650 °C, respectively. From Figure 6A, it can be seen that the formation rate of CO increased with the increase in CO2 partial pressure while it did not change significantly with the increase in CH4 partial pressure (Figure 6B). This observation can be ascribed to the influence of reverse water-gas shift reaction (RWGS). (3)
Figure 7. Effect of reactant partial pressures on the formation rates of H2 over Ni-Co/Al-Mg-O catalyst at 310 kPa and W/F ) 0.002 gcat · s/mL: (A) effect of CO2 partial pressure; (B) effect of CH4 partial pressure.
At constant CH4 partial pressure, the reaction rate of CO2 reforming of CH4 was restricted due to limited availability of CH4. However, the formation of CO still can be enhanced significantly through the RWGS reaction between H2 and excess CO2 when CO2 partial pressure was increased continuously as shown in Figure 6A. At constant CO2 partial pressure, the formation rate of CO only increased slightly with the increase in CH4 partial pressure (Figure 6B). The effects of reactant partial pressures on the formation rate of H2 at 750, 700, and 650 °C are shown in Figure 7. It is clear that at constant CH4 partial pressure the H2 formation rate increased in the low CO2 partial-pressure range 31-62 kPa (Figure 7A). The increase in H2 formation was due to the increase in reaction rate of CO2 reforming of CH4 when more CO2 was available. The formation rate of H2 was relatively stable
in the CO2 partial-pressure range 62-124 kPa (Figure 7A). In this CO2 partial-pressure range, the reaction rate of CO2 reforming of CH4 should be increasing slightly but limited by the availability of CH4. At the same time, the RWGS reaction could be also increasing with the availability of excess CO2. When the increase of CO2 reforming of CH4 is equivalent to the increase of RWGS, the formation of H2 will remain constant. However, the formation rate of CO will continue to increase. This can be confirmed by the increasing formation rate of CO due to the increased CO2 partial pressure in the range of 62-124 kPa (Figure 6A). With further increase in CO2 partial pressure from 124 to 186 kPa, the formation rate of H2 dropped (Figure 7A). It was because of the fact that the increase in H2 formation rate from CO2 reforming of CH4 reaction was less than the consumption rate of H2 from RWGS reaction when CH4 was
CO2+H2 f CO + H2O
Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009 681
Figure 9. Effect of temperatures on the reaction rates of CO2 reforming of CH4 at 310 kPa, CH4/CO2/N2 ) 1/1/3, and W/F ) 0.002 gcat · s/mL. Table 2. Apparent Activation Energy (Ea, kJ/mol) over Ni-Based Catalysts catalyst
Figure 8. Reaction rate of the RWGS reaction as a function of partial pressure: (A) function of CO2 partial pressure; (B) function of CH4 partial pressure.
limited but CO2 was in excess. At constant CO2 partial pressure, the formation rate of H2 first slightly increased and then was constant with the increase of CH4 partial pressure (Figure 7B). It indicated that both CO2 reforming of CH4 reaction and RWGS were restricted when limited CO2 was available. 3.5. Reverse Water-Gas Shift Reaction (RWGS) in the Reforming System. As discussed in the previous section, the effects of RWGS were clearly indicated by the difference between CO and H2 formation rate. Referring to eq 1 and eq 3, it is clear that the production of every one CO molecule through RWGS reaction results in the consumption of one H2 molecule, that is, every one RWGS reaction produces one CO but consumes one H2. The same formation rates of CO and H2 should have been expected if CO2 reforming with CH4 were the only reaction (eq 1). However, the occurrence of RWGS reaction (eq 3), actually, increased the CO formation rate and reduced the H2 formation rate. The RWGS reaction rate can be calculated using the 0.5 times difference between CO and H2 formation rates if assuming RWGS is the only side reaction in the CO2 reforming with CH4 reaction system. It has been reported that carbon formation on the Ni-Co/Al-Mg-O bimetallic catalysts is very low17,18 so that it is negligible in comparison with the formation of CO and H2. Therefore, it is reasonable to assume the RWGS is the only side reaction when the product formation is considered. In this work, the reaction rate of RWGS was estimated on the basis of the previous assumption in order to have a quantitative understanding about its effects on the CO2 reforming with CH4 reaction system. Figure 8A shows the RWGS reaction rate as a function of CO2 partial pressure. At constant CH4 partial pressure, the RWGS reaction rate increased almost linearly with the increase in CO2 partial pressure. At higher temperatures, the reaction rate increased faster as more H2 was available at higher conversion of CH4 at higher temperature. However, as shown in Figure 8B, at constant CO2 partial pressure the reaction rate of RWGS was suppressed as the CH4 concentration increased. It indicated that CH4 reforming reaction was competing for CO2 with the RWGS reaction. The suppression of RWGS reaction at availability of additional CH4 can be expected since the equilibrium
Ni-Co/Al-Mg-O (650-750 °C) Ni/Al2O3 (500-700 °C) Ni/TiO2 (400-550 °C) Ni/CaO–Al2O3 (620-690 °C)
Ea (CH4) Ea (CO2) Ea (H2) Ea (CO) reference 69.4
25.9
85.1
50.9
56.1
108.9
87.9
146.5
106.8
98.8
147.4
61.8
this work
80.5
15
87.9
12
103
21
constant K of CO2 reforming with CH4 is much higher than that of the RWGS reaction at the reaction temperatures studied. 3.6. Effect of Temperature and Apparent Activation Energy. The effect of temperatures on the reaction rate of CO2 reforming of CH4 over Ni-Co/Al-Mg-O bimetallic catalyst was investigated in the temperature range 650-750 °C at 310 kPa, CH4/CO2/N2 ) 1/1/3, and W/F ) 0.002 gcat · s/mL. Results in the form of Arrhenius plots are shown in Figure 9. The apparent activation energies were then estimated for CH4, CO2 consumption, H2, and CO formation. Comparison of apparent activation energies of Ni-Co/Al-Mg-O bimetallic catalyst with reported Ni-based catalysts are listed in Table 2.12,15,21 From Table 2, it is clear that the apparent energy is very different over different catalysts. The disagreement in apparent activation energy is probably due to the difference of catalyst systems applied in the relevant studies. Tsipouriari and Verykios13 reported that the catalyst supports of Ni may influence significantly the activation energy by altering the rate-controlling step in the reaction sequence. The proximity of apparent activation energy values for CH4, CO2, and CO suggests that the ratecontrolling step for CO formation is the same as that for CH4 and CO2 consumption.21 In a review by Bradford and Vannice,2 it was reported that the activation energies of CO2 reforming of CH4 are typically between 33 and 100 kJ/mol while high values up to 160 kJ/mol are obtained for some catalysts. However, for the case of Ni-Co/Al-Mg-O bimetallic catalyst in this work, the apparent activation energy of CO2 consumption was much lower than that of CH4 consumption (Table 2). Also, it was lower than the reported activation energies for CO2 consumption (Table 2). Possibly this difference was due to the presence of strong Lewis base of MgO22 in the support of Ni-Co/Al-Mg-O bimetallic catalyst, which can facilitate the activation of CO2. Also, the existence of RWGS reaction might be an important factor influencing the activation of CO2. Concerning the activation energies for CO and H2, the apparent energy of H2 formation is always higher than that for CO formation (Table 2). It has been ascribed to be the result of the RWGS influence on the reaction mechanism.12
682 Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009 Table 3. Parameters of Power-Law Equation catalyst
m
n
Ni/SiO2 Ni/Al2O3 Ni/MgO
-0.3 to 0.8 0.6 to 1.0 1.0
0 to 0.6 0 to 0.3 0
reference 23, 24 25 26
Table 4. Estimated Parameters of the Power-Law Model of Ni-Co Bimetallic Catalyst k0 (mol/gcat · s · kPa(m+n))
E (kJ/mol)
m
n
0.089
63.3
0.483
0.291
model might not be meaningful. Therefore, other models may be preferred to the power-law form because it is more likely to be amenable to extrapolation. In the following section, a Langmuir-Hinshelwood (L-H) mechanism model is developed. 3.7.2. Consideration of Reaction Mechanism: LangmuirHinshelwood Mechanism. Mechanismic steps plausible to describe the reaction mechanism of CO2 reforming of CH4 over the Ni-Co/Al-Mg-O bimetallic catalyst was proposed in this work based on experimental observation and literature review. (i) The reversible adsorption and dissociation of CH4 can be assumed as the initial steps for the reforming reaction.26 Because of the metallic properties of active sites on Ni-Co/Al-Mg-O bimetallic catalysts,17 the methane was supposed to adsorb and dissociate on metal ensembles (eq 6 and 7), leading to the formation of H2 and C species. K1
CH4 + M 798 M - CH4
equilibrium
k2
M - CH4 98 M - C + 2H2
Figure 10. Comparison plot for the reforming rates of CO2 reforming of CH4 by the power-law model (eq 4).
3.7. Kinetic Models. 3.7.1. Power Model. A general power law model can be expressed by m Pn r ) kPCH 4 CO2
(4)
The values of the power law rate coefficient k and the indices m and n are different for different catalysts. The values of the parameters of Ni-based catalysts are summarized in Table 3.23-26 The power-law model (eq 4) was used to describe the reforming rate of CO2 reforming of CH4 over Ni-Co/ Al-Mg-O bimetallic catalyst. Also, the reaction rate constant, k, was further expressed in an Arrhenius equation (eq 5). The values of the parameters and the R squared value for the powerlaw model are estimated and listed in Table 4. k ) k0e-E⁄RT
(5)
A comparison of the experimental rates of CH4 consumption over the Ni-Co/Al-Mg-O bimetallic catalyst with those predict by the power-law model is shown in Figure 10. It can be seen that the power-law model is able to predict the reaction rate satisfactorily with R2 ) 0.944. Simplicity is the major advantage of the power-law model. However, the power-law
slow
(7)
where M represents the metal ensemble of the reduced Ni-Co/ Al-Mg-O catalyst. It was supposed that the CH4 adsorption was at equilibrium while the dissociation of CH4 was a slow step. The carbon species was present as M-C or carbide-like species which contains no C-H bond.27 (ii) In CO2 reforming of CH4, the activation of CO2 can be assumed to be another important elementary step. CO2 is a linear triatomic molecule and has two sets of π molecular orbitals which are orthogonal. The carbon is a Lewis acid center and the oxygen is a weak Lewis base. The formation of an active site-CO2 complex via direct coordination is one of the most powerful and universal ways to activate the inert CO2 molecule to undergo chemical reactions.28 The support of Ni-Co/ Al-Mg-O bimetallic catalyst is mainly composed of MgO which is a strong Lewis base.22 Therefore, the activation of CO2 was supposed to occur on the Lewis base center of the support (S) and was at equilibrium. K3
CO2 + S 798 S - CO2
equilibrium
(8)
(iii) The CO2 reforming of the CH4 reaction was assumed to terminate through the reaction between carbon species (M-C) derived from CH4 decomposition and activated CO2 (S-CO2) and it is also considered to be a slow step.13 k4
S-CO2 + M-C 98 2CO + M + S
slow
model
catalyst
reference
-rCH4)kPCH4 1+aPH2O / PH2 +bPCO -rCH4)aPCH4PCO22 / (a+bPCO22+cPCH4)2 -rCH4)kˆ1PCH4PCO2 kˆ-1kˆ / kˆ7 PCOPH2(4-x)⁄2+(1+kˆ1 / kˆ7 PCH4)PCO2 -rCH4 ) (kPCH4PCO2)/((a + K1PCH4 + K2PCO)(1 + K3PCO2)) -rCH4 ) (k3K1K2PCH4PCO2PCOPH2)/((PCOPH22 + K1PCH4PCO + K2PCO2PH22)2) -rCH4 ) (k3PCH4PCO2)/((1 + K1PCH4)(1 + K2PCO2)) -rCH4 ) (K1k2K2k4PCH4PCO2)/((K1k2K3PCH4PCO2 + K1k2PCH4 + K3k4PCO2)) -rCH4 ) (k1LPCH4)/((k1LPCH4PCO)/(k7LKaPCO2) + (KbPCO2PH21/2)/(Pco) + (k1LPCH4)/(k7L) + 1)
Ni foil Ni/Al2O3, Ni/CaO–Al2O3 Ni/TiO2 Ni/C Ni/SiO2 Ni/MgO Ni/La/Al2O3 Ni/SiO2
8 9 12 14 29
Ni/Al2O3 Ni/CeO2-Al2O3 Ni/La2O3 Ni-K/CeO2-Al2O3
30 13 16
/
(9)
(iv) RWGS is the most common side reaction in CO2 reforming of CH4. It is important to take its occurrence into
Table 5. Reported L-H Models for CO2 Reforming of CH4 over Ni-Based Catalysts
/
(6)
Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009 683 Table 6. Parameter Estimates of the L-H Model parameter
estimate
unit -8
(mol/(g · s))2•(kPa)-2 kJ/mol (mol/(g · s))•(kPa)-1 kJ/mol (mol/(g · s))•(kPa)-1 kJ/mol
1.35 × 10 25.9 9.25 × 10-8 -40.6 2.46 × 10-7 -38.3
a0 Ea b0 Eb c0 Ec
account. Besides step 9, activated CO2 can also react with the surface hydrogen species which originated from the decomposition of CH4 and was assumed at equilibrium with the gaseous H2.12 The occurrence of RWGS results in the consumption of H2 and the formation of CO, which can be confirmed by the observation that CO/H2 is higher than unity.17 H2 + M T 2M-H
equilibrium
S-CO2 + M-H f CO + M-OH + S M-OH + M-H T H2O + 2M
(10) slow
fast
(11) (12)
3.7.3. L-H Type of Model. Literature survey shows that the L-H type of kinetic model is widely used to describe CO2 reforming of CH4 system.2,10 A summary of rate expressions derived for Ni-based catalysts is tabulated in Table 5.8,9,12-14,16,29,30 From the earlier proposed mechanism, a L-H model (eq 13) for CO2 reforming of CH4 over the Ni-Co/Al-Mg-O bimetallic catalyst was developed based on the following assumptions: (1) step 9 is rate-limiting step; (2) the concentration of total metallic active sites for reforming reaction is constant, CMT; (3) the coverage of metallic active sites for reforming reaction by species of H and OH is negligible; (4) the coverage of active sites for CO2 activation is negligible in comparison with its total concentration CST, which is also supposed to be constant. -rCH4 )
aPCH4PCO2 bPCH4 + cPCO2 + dPCH4PCO2
(13)
Figure 11. Comparison plot for the reforming rates of CO2 reforming of CH4 by L-H model (eq 18).
reaction rates were calculated using the simplified L-H model (eq 18). The comparison plot for the reforming rate of CO2 reforming of CH4 is shown in Figure 11. In comparison with the power-law model (Figure 9), the L-H model showed a better fit with the experimental data with a R2 of 0.96. 4. Conclusion The kinetics of the reaction of CO2 reforming of CH4 was investigated over Ni-Co/Al-Mg-O bimetallic catalyst in the temperature range 650-750 °C through variation of CO2 and CH4 partial pressures. The reaction mechanism was proposed on the basis of literature and experimental observations. The dissociation of CH4 and the reaction between C species and activated CO2 were suggested as rate-determining steps. The L-H kinetic model derived from the proposed mechanism was found to fit the kinetic measurement well. It was also found that the power-law model is able to satisfactorily fit the kinetic experimental results as well.
a ) CMTCSTK1k2K3k4
(14)
b ) K1k2
(15)
c ) CSTK3k4
(16)
Acknowledgment
d ) CSTK1K3k4
(17)
The authors gratefully acknowledge the financial support from the Natural Science and Engineering Research Council of Canada (NSERC) and the Canada Foundation for Innovation (CFI).
Preliminary estimation for the parameters of the L-H model (eq 13) was performed using least-squares analysis and the Marquardt-Levenberg algorithm. Estimation of the parameters was based on the minimization of the sum of residual squares of the observed and the predicated reforming rates. It showed that d (∼10-13) , b (∼10-8) and d (∼10-13), c (∼10-8). Therefore, the L-H model (eq 13) was simplified to -rCH4 )
aPCH4PCO2 bPCH4 + cPCO2
(18)
After the values of all of the parameters were estimated at 750, 700, and 650 °C, the parameters of the simplified L-H model (eq 18) were further expressed in the form of Arrhenius equations (eq 19-21). a ) a0e-Ea⁄RT
(19)
b ) b0e-Eb⁄RT
(20)
c ) c0e-Ec⁄RT
(21)
where a0, b0, and c0 and Ea, Eb, and Ec are constants. The estimated values of these constants are shown in Table 6. The
Nomenclature a, a0, b, b0, c, c0 ) constants CMT ) concentration of total metallic active sites, mol/m3 CST )concentration of total active sites for CO2 activation, mol/ m3 dp, D ) average catalyst particle diameter, m Ea, Eb, Ec ) apparent activation energy, kJ/mol ∆Hr ) reaction enthalpy, kJ k, k0, k1, k2, k3, k4, k1L, k7L, kˆ1, kˆ-1, kˆ7, kˆ-7 ) specific rate of reaction Ka, Kb, K1, K2, K3, K4, K ) chemical equilibrium constant m, n ) reaction order P ) pressure, kPa -r, r ) reaction rate per unit catalyst weight, mol/(g · s) 〈R〉 ) average reaction rate per unit catalyst particle volume, mol/ (m3 · s) R ) ideal gas constant R2 ) square of correlation coefficient Ts ) surface temperature of catalyst particle, K
684 Ind. Eng. Chem. Res., Vol. 48, No. 2, 2009 T ) temperature, K λp ) thermal conductivity of catalyst particle, W/m/K
Literature Cited (1) Worner, A.; Tamme, R. CO2 reforming of methane in a solar driven volumetric receiver-reactor. Catal. Today 1998, 46, 165. (2) Bradford, M. C. J.; Vannice, M. A. CO2 reforming of CH4. Catal. ReV. Sci. Eng. 1999, 41, 1. (3) Hu, Y.; Ruckenstein, E. Binary MgO-based solid solution catalysts for methane conversion t syngas. Catal. ReV. 2002, 44, 423. (4) Hu, Y.; Ruckenstein, E. Catalytic conversion of methane to synthesis gas by partial oxidation and CO2 reforming. AdV. Catal. 2004, 48, 297. (5) Rostrup-Nielsen, J. R. Industrial relevance of coking. Catal. Today 1997, 37, 225. (6) Slagtern, A.; Olsbye, U.; Blom, R.; Dahl, I. M. Stud. Surf. Sci. Catal. 1997, 107, 497. (7) Ashcroft, A. T.; Cheetham, A. K.; Green, M. L. H.; Vernon, P. D. F. Partial oxidation of methane to synthesis gas using carbon dioxide. Nature 1991, 352, 225. (8) Bodrov, I. M.; Apel’baum, L. O. Reaction kinetics of methane and carbon dioxide on a nickel surface. Kinet. Catal. 1967, 8, 326. (9) Zhang, Z.; Verkios, X. E. Mechanistic aspects of carbon dioxide reforming of methane to synthesis gas over Ni catalysts. Catal. Today 1994, 21, 589. (10) Wang, H. Y.; Au, C. T. CH4/CD4 isotope effects in the carbon dioxide reforming of methane to syngas over SiO2 supported nickel catalysts. Catal. Lett. 1996, 38, 77. (11) Hu, Y. H.; Ruckenstein, E. Transient response analysis via a broadened pulse combined with a step change or an isotopic pulse: application to CO2 reforming of methane over NiO/SiO2. J. Phys. Chem. B 1997, 101, 7563. (12) Bradford, M. C. J.; Vannice, M. A. Catalytic reforming of methane with carbon dioxide over nickel catalysts II. Reaction kinetics. Appl. Catal., A 1996, 142, 97. (13) Tsipouriari, V. A.; Verykios, X. E. Kinetic study of the catalytic reforming of methane with carbon dioxide to synthesis gas over Ni/La2O3 catalyst. Catal. Today 2001, 64, 83. (14) Olsbye, U.; Wurzel, T.; Mleczko, L. Kinetic and reaction engineering studies of dry reforming of methane over a Ni/La/Al2O3 catalyst. Ind. Eng. Chem. Res. 1997, 36, 5180. (15) Wang, S.; Lu, G. Q. A comprehensive study on carbon dioxide reforming of methane over Ni/γ-Al2O3. Ind. Eng. Chem. Res. 1999, 38, 2615.
(16) Nandini, A.; Pant, K. K.; Dhingra, S. C. Kinetic study of the catalytic carbon dioxide reforming of methane to synthesis gas over NiK/CeO2-Al2O3. Appl. Catal., A 2006, 308, 119. (17) Zhang, J.; Wang, H.; Dalai, A.K. The development of Ni-Co bimetallic catalyst for CO2 reforming of CH4. J. Catal. 2007, 249, 300. (18) Zhang, J.; Wang, H.; Dalai, A. K. Effects of metal content on activity and stability of Ni-Co bimetallic catalysts for CO2 reforming of CH4. Appl. Catal., A 2008, 339, 121. (19) Forni, L. Mass transfer and heat transfer in catalytic reactions. Catal. Today 1999, 52, 147. (20) Anderson, J. B. A criterion for isothermal behavior of a catalyst pellet. Chem. Eng. Sci. 1963, 18, 147. (21) Lemonidou, A. A.; Vasalos, I. Carbon dioxide reforming of methane over 5 wt.% Ni/CaO-Al2O3 catalyst. Appl. Catal., A 2002, 228, 227. (22) Li, C.; Fu, S.; Zhang, H.; Xin, Q. An infrared spectroscopic study on the lewis base properties of metal oxides by using a novel probe molecule: boric acid trimethylester. J. Chem. Soc. Chem. Commun. 1994, 17, 18. (23) Takanabe, K.; Nagaoka, K.; Nariai, K.; Aika, K.-I. Titania-supported cobalt and nickel bimetallic catalysts for carbon dioxide reforming of methane. J. Catal. 2005, 232, 268. (24) Osaki, T.; Horiuchi, T.; Suzuki, K.; Mori, T. Catalyst performance of MoS2 and WS2 for the CO2-reforming of CH4 -suppression of carbon deposition, Appl. Catal. 1997, 155, 229. (25) Rostrup-Nielsen, J. R.; Hansen, J. H. CO2-reforming of methane over transition metal. J. Catal. 1993, 144, 38. (26) Hickman, D. A.; Schmidt, L. D. Production of syngas by direct catalytic oxidation of methane. Science 1993, 259, 343. (27) Chang, J.; Park, S.; Yoo, J.; Park, J.-M. Catalytic behaviour of supported KNiCa catalyst and mechanistic consideration for carbon dioxide reforming of methane. J. Catal. 2000, 195, 1. (28) Gibson, D.H. η2-coordination of CO2 was found both experimentally and theoretically. Chem. ReV. 2007, 1996, 2063. (29) Kroll, V. C. H.; Tjatjopoulos, G. J.; Mirodatos, C. Kinetics of methane reforming over Ni/SiO2 catalysts based on a stepwise mechanistic model. Stud. Surf. Sci. Catal. 1998, 119, 753. (30) Wang, S.; Lu, G. Q. Reaction kinetics and deactivation of Ni-based catalysts in CO2 reforming of methane. React. Eng. Pollut. PreV. 2000, 75.
ReceiVed for reView July 14, 2008 ReVised manuscript receiVed October 28, 2008 Accepted November 4, 2008 IE801078P
