Corrected Proof

FUPROC-02245; No of Pages 6

ARTICLE IN PRESS F U E L P RO C ES S I NG T EC H NOL O G Y X X (2 0 0 7) XX X–XXX

www.elsevier.com/locate/fuproc

Kinetic modeling of plasma methane conversion in a dielectric barrier discharge Antonius Indarto a,b,⁎, Nowarat Coowanitwong b , Jae-Wook Choi a , Hwaung Lee a , Hyung Keun Song a a

Clean Technology Research Center, Korea Institute of Science and Technology, P.O. Box 131, Cheongryang, Seoul, Republic of Korea School Environmental Resources Development, Asian Institute of Technology, P.O. Box 4, Klong Luang, Pathumthani, Thailand

b

AR TIC LE I N FO

ABS TR ACT

Article history:

Methane conversion by plasma offers a promising route to produce higher value-added

Received 7 February 2007

products. As plasma reaction is a relatively complex process, kinetic modeling is necessary

Received in revised form

to obtain a general pattern of the complex interaction on the basis of chemical reaction and

13 September 2007

products. In this paper, we present a method to obtain the kinetic rate coefficient (k) from

Accepted 17 September 2007

the experimental data. Although plasma reaction was classified as chemically complex interaction, the reactions showed a certain pattern of the mechanism. In pure methane

Keywords:

injection, the decomposition of methane by plasma could initiate coupling reactions and

Plasma

produce C2H6, C3H8, and C4H10. Dehydrogenation of C2H6 into C2H4 and then to C2H2 could be

Dielectric barrier discharge

clearly seen by the higher value of the reaction rate constant of C2Hn + 2 to C2Hn − 2. Using the

Methane

rate constant values (k) obtained by this method, the pathways of the methane conversion

Kinetic reaction

by a dielectric barrier discharge can be drawn. © 2007 Elsevier B.V. All rights reserved.

1.

Introduction

The use of natural gas as a feedstock in the chemical and pharmaceutical industry is an alternative to crude oil whose supplies might run out in the next century. Methane (CH4), the major component of natural gas, is widely distributed at sites around the world and have a potential to be used as the precursor to produce higher value-added products, for example acetylene and ethylene. There are mainly three approaches to converting CH4 into higher hydrocarbons in a high temperature process: the direct oxidative methane conversion (OMC), the Fischer–Tropsch process via syngas, and the non-oxidative methane conversion [1]. Over the last decades, many researches have mainly focused on oxygen-containing processes, either indirect ones as in the Fischer–Tropsch process or direct ones as in the oxidative methane conversion. In the Fischer–Tropsch pro-

cesses, CH4 is converted into hydrocarbons, via synthesis gas, which is subsequently hydrogenated, e.g., to methanol [2]. In the presence of oxygen, the direct CH4 conversion to C2H6, C2H4 (one of the most important raw materials in industrial production cycles), C3Hn, C4Hn is thermodynamically feasible (exothermic), whereas the oxygen-free or dehydrogenative conversion is endothermic (two-step polymerization). OMC has been investigated extensively [3,4] and is still most frequently used due to its higher methane conversion although suffering from low selectivity of C2+ [5]. The oxygen-free methane conversion, however, has only been studied by a few groups [6–9]. The state of the art for this process was recently summarized by Guczi et al. [10]. A major disadvantage of the OMC is its low selectivity towards C2+, because considerable amounts of CH4 are consumed forming CO and CO2. In contrast to OMC, the non-oxidative methane conversion (NOMC) by thermal process gives numerous amounts of solid carbon and hydrogen [11] with less C2+

⁎ Corresponding author. Clean Technology Research Center, Korea Institute of Science and Technology, P.O. Box 131, Cheongryang, Seoul 130-650, Republic of Korea. Tel.: +82 10 2296 3748. E-mail address: [email protected] (A. Indarto) 0378-3820/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.fuproc.2007.09.006

Please cite this article as: A. Indarto, et al., Kinetic modeling of plasma methane conversion in a dielectric barrier discharge, Fuel Process. Technol. (2007), doi:10.1016/j.fuproc.2007.09.006

ARTICLE IN PRESS 2

F U E L P RO C ES S IN G T EC HN O L O G Y X X (2 0 0 7) XXX –XXX

production. Higher hydrocarbon production via methyl coupling reactions occurred only with very short time and high temperature reaction [12]. Non-oxidative plasma methane conversion (NOPMC) offers a high selectivity towards C2+ products [11,13]. The reason for focusing on oxygen-free methane conversion is therefore to maintain the high C2+ selectivity, while simultaneously enhancing the CH4 conversion and the C2+ yield, respectively. The optimization of the NOPMC necessitates a detailed knowledge of the gas-phase reactions that occur and their kinetics. In this present work, we develop a kinetic model for plasma methane conversion to higher hydrocarbons. By finding the kinetic rate constant (k), the global pathway of the reactions could be drawn easily. The model includes the gas-phase transformation of molecules that exist in the reactant and the products. Based on the comparison between experimental and simulation data, it was found that the proposed kinetics well-accounted for the reaction products of C2H2, C2H4, C2H6, C3H6, C3H8, i-C4H10, and n-C4H10. The constructed model was applicable for the plasma reaction using dielectric barrier discharge at atmospheric pressure (c.a. 1 atm) and ambient temperature (c.a. 25 °C). The presented kinetic model is a part of efforts encompassing the oxygen-free plasma chemistry for methane conversion.

2.

Experimental setup and algorithm

The schematic diagram of the experimental setup is shown in Fig. 1. Details of each part of the system are described in the following sections.

2.1.

Plasma reactor and power system

The reactor used in the experiments is a cylindrical pyrex tube (ID of 7.5 mm) with 2 parallel-straight wires (0.2 mm diameter, stainless steel) as the inner metal electrode and silver film coated at the outer side of tube as the outer electrode. The effective volume and length of the reactor were 8.8 ml and 200 mm, respectively. A high frequency alternating current (AC) power supply with a maximum voltage of 20 kV was connected to the electrodes. The supplied power to the reactor was maintained constant at 60 W, measured by a watt meter (Metex model M-3860M).

In order to maintain the similarity of the reactor configuration, e.g. electrodes gap distance, the reactor capacitance was checked by an RCL meter (Fluke model PM6304) before and after experiments. The reactor capacitance was in the range of 9.0–10.0 pF at ambient air condition.

2.2.

Input gas and measurement system

All experiments were carried out by introducing eight different hydrocarbons: CH4, C2H2, C2H4, C2H6, C3H6, C3H8, i-C4H10, n-C4H10. CH4 has purity of 99.97% while other gases have purity of above 99.9%. A Gas Chromatography (YoungLin M600D, column: 30 ft of Hayesep D) has been used to quantitatively analyze the amount of products by flame ionization detection (FID) and thermal conductivity detector (TCD). The FID system was capable to detect compounds in hundreds of ppm level. Considering the volume expansion and the compression of products, the output line was connected to a bubble flow meter to measure the flow difference of the products before and after reaction. The products concentration and reactants conversion were formulated as:

Conversion of Cx Hy ¼

moles of Cx Hy converted  100 moles of initial Cx Hy

Concentration of Cm Hn ¼ Selectivity of Cm Hn ¼

moles of Cm Hn produced  100 total moles products

ð1Þ

ð2aÞ

m  moles of Cm Hn produced  100 : moles of Cx Hy converted ð2bÞ

2.3.

Model algorithm

In the kinetic model, the possibility of molecular transformation to any possible compounds that exist in the products or the reactants was calculated. The reaction rate (dC / dt) of single hydrocarbon was measured experimentally by injecting pure compound to the reactor. Then, the product distribution was calculated and the rate of molecular transformation converted to other molecules could be obtained. From the experimental analysis, eight hydrocarbons existed in the products and those were (1) CH4, (2) C2H2, (3) C2H4, (4) C2H6, (5) C3H6, (6) C3H8, (7) i-C4H10, and (8) n-C4H10. As we assume

Fig. 1 – Schematic diagram of experimental setup. Please cite this article as: A. Indarto, et al., Kinetic modeling of plasma methane conversion in a dielectric barrier discharge, Fuel Process. Technol. (2007), doi:10.1016/j.fuproc.2007.09.006

ARTICLE IN PRESS 3

F U E L P RO C ES S I NG T EC H NOL O G Y X X (2 0 0 7) XX X–XXX

that the reaction will occur in a 1st order reaction, the calculation of each compound is as follows: n n X X dCi ¼ kij Ci þ kji Cj dt j¼1 j¼1

ð3Þ

where n is the maximum value of i and j for k-set which is 8 and kij denotes the reaction rate constant of i molecule that converts into a specific j molecule. The assumption was chosen to simplify the calculation and make the data comparable to each other. The value of kij will be 0 if i = j. 2

3 dC1 2 k11 : : : 6 dt 7 6 7 ⋱ 6v 7 ¼ 4 v 4 dC 5 8 k81 : : : dt

32

C1 : : : 54 v ⋱ C1 : : : k88 k18

v

C8

3

2

k11 : : :

v 5 þ 4v C8

⋱

k18 : : :

32

C1 : : : 54 v ⋱ C8 : : : k88 k81

v

C1

3

v 5

C8

ð4Þ By solving Eq. (4), the calculation will result a set of kinetic rate constant (k) which can be arranged as: 2

0 k12 : : : ::: 6 k21 0 k¼6 4v v ⋱ k81 : : : k87

3 k18 k28 7 7: v 5 0

ð5Þ

The value of k was obtained by an error minimization between model data and experimental data. In this particular case, the minimization calculation was done employing Matlab sub-routine module ‘fminsearch’ after simultaneous integration of Eq. (4) using ‘ode23s’. The uniqueness of the result was measured by calculating the mean-squared error of each set value of kij obtained from eight different input gases which is expressed as:

s¼

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi uPP 2 u k  ki t

ð6Þ

n1 P

where n is the maximum size of k-set which is eight, k is the mean value of k-set.

3.

Results and discussion

In order to obtain the rate constant k in Eq. (5), eight hydrocarbon gases: CH4, C2H2, C2H4, C2H6, C3H6, C3H8, i-C4H10, and n-C4H10 were introduced separately to the plasma reactor at atmospheric

pressure (∼1 atm) and room temperature (∼25 °C). The flow rate and supplied power were maintained similar at 30 ml/min and 60 W, respectively. These values were obtained from the optimization of the previous research [14] in order to avoid large amount of unknown products and carbon deposition which were difficult to measure. Fragmentation reactions of methane into smaller molecules were more favorable than synthesis reactions of higher hydrocarbons at lower input flow rates and higher supplied power [11,15,16]. As higher hydrocarbons are more useful and valuable products, increasing flow rates and lowering supplied power can be the way to avoid above problems. Table 1 shows the conversion rate of input gas and the distribution of products after treated by dielectric barrier discharge at flow rate of 30 ml/min and supplied power of 60 W. Except for the experiments initiated by C2H2 (run 2) and C3H6 (run 5), the total C and H atom balance between the reactants and products of all plasma process approximately closed to 100% which means that the molecular transformation of reactants to the products were almost perfectly identified. Small amounts of carbon deposition on the wall and inner electrode were found but, as the values were relatively small, those could be neglected in the calculation. It was a good achievement, since to obtain correct calculation of the kinetics, the C and H balance between reactants and products should be equal. In the case when C2H2 and C3H6 were used as the reactant, the reaction produced a lot of solid carbon and/or soot. We could not confirm whether the solid product was coming from the polymerization of C2H2 and C3H6 or agglomeration of monomolecular carbon. However, C2H2 and C3H6 are well-known as soot precursor in the pyrolysis or combustion process [17]. In this case, the similar phenomenon has likely occurred in the plasma process. From Table 1, it can be figured out that when methane was the reactant, the products were dominated by ethane (C2H6), with selectivity of 37.5%, propane (C3H8), 25.6%, and butane (normal- and iso-C4H10), 25.4%. This product composition has a similarity with the result of Thanyachotpaiboon et al. [18], who used DBD system and AC power supply similar to ours, with slightly higher selectivity of propane and butane. The higher value of the selectivity was due to the use of a different method in the selectivity calculation. Thanyachotpaiboon et al. was using fractional method instead of carbon balancebased calculation that we used.

Table 1 – The reactant conversion and products distribution of plasma process Reaction

CH4 C 2H 2 C 2H 4 C 2H 6 C 3H 6 C 3H 8 i-C4H10 n-C4H10

Conversion [%] 13.8 12.0 15.7 9.4 9.8 11.7 16.7 8.4

Products concentration [%]

Atom balance

CH4

C2H2

C2H4

C2H6

C3H6

C3H8

i-C4H10

n-C4H10

Cout/Cin

Hout/Hin [1]

86.20 3.03 2.67 0.96 0.70 2.31 2.36 1.27

0.17 81.95 3.31 1.27 1.60 0.47 2.35 1.08

0.45 1.45 84.30 2.10 2.48 2.03 4.02 2.29

2.04 0.00 0.68 90.57 0.46 0.95 1.65 1.28

0.00 0.14 0.38 0.16 80.16 2.54 3.19 1.36

0.93 0.00 0.67 0.89 2.23 88.27 1.39 1.46

0.26 0.00 0.67 0.05 0.99 0.09 83.33 2.33

0.43 0.28 1.84 0.85 0.57 0.00 5.94 91.63

0.95 0.87 0.95 0.97 0.89 0.97 1.04 1.03

1.01 1.13 0.98 0.98 1.01 0.96 1.00 0.98

of Cm Hn produced Note: [1]The calculation of H balance includes the H2 in the output and the selectivity of component follows: selectivity of Cm Hn ¼ n4  moles moles of CH4 converted . All data were obtained at a flow rate of 30 ml/min and a supplied power of 60 W.

Please cite this article as: A. Indarto, et al., Kinetic modeling of plasma methane conversion in a dielectric barrier discharge, Fuel Process. Technol. (2007), doi:10.1016/j.fuproc.2007.09.006

ARTICLE IN PRESS 4

F U E L P RO C ES S IN G T EC HN O L O G Y X X (2 0 0 7) XXX –XXX

Higher ethane selectivity could be obtained by coupling reaction of methyl radical which also suggested by Zhao et al. [19]. Abundant amounts of methyl radical could combine with C2 molecules to form C3 hydrocarbons or with C3 molecules to form C4 hydrocarbons [20]. In the case of other products, e.g. acetylene and ethylene, those compounds could be produced due to stepwise dehydrogenation reaction of ethane. The tendency of acetylene formation from ethylene and ethylene from ethane by dehydrogenation reactions was clearly seen from run-4 of Table 1. When ethane was the reactant, the selectivity for ethylene was 29% which is two times higher compared when acetylene was used as the reactant. A similar phenomenon also occurred in the case of acetylene formation. The selectivity of acetylene production reached 27.7% when ethylene was the reactant and only the selectivity was 3.2% when methane was the reactant. An interesting phenomenon occurred when acetylene was used as the reactant which could be an important part to describe the kinetic pathways. Acetylene was converted mostly into methane, 40%, and at the same time, it also produced ethylene (39%), and n-butane (15%). Those products required hydrogen to react with acetylene. C2 H2 þ 3H2 →2CH4

ð7Þ

C2 H2 þ H2 →C2 H4

ð8Þ

2C2 H2 þ 3H2 →iC4 H10

ð9Þ

2C2 H2 þ 3H2 →nC4 H10 :

ð10Þ

In order to supply H2, fragmentation of C2H2 as the intermediate reaction was needed. C2 H2 →2C þ H2 :

ð11Þ

As the total C atom of gaseous products was only 87% compared to the total C atom of reactant, reaction 11 high possibly occurred in plasma zone and responsible for the formation of molecular solid carbon formation. The proposed reaction 11 was also supported by hydrogen atom ratio of the products to the reactant which closed to 1 that means there was no loss of hydrogen atom in the process. Coupling reactions between two C2 compounds could form butane and the long-chain type (n-C4H10) was more favorable than the iso-type (i-C4H10). It was simply because the adduct reaction of two C2 moieties occurred without any molecular steric hindrance although later on, it converted to iso-form as iso-form is thermodynamically more stable than normal-type. C2 H4 þ C2 Hx þ ð8−xÞ=2 H2 →nC4 H10 :

ð12Þ

In the case of C3H6 consumption process, the reactions occurred by two main parallel ways: C3 H6 þ H2 →C3 H8

ð13Þ

C3 H6 →C2 H4 þ C þ H2 :

ð14Þ

From these two reactions, we would obtain less amount of total C atom number in the gaseous products due to carbon solid formation in Eq. (14). It was confirmed by experimental result that the total C atom in the gaseous products was only 89%

compared to the original reactant. Instead of above conversion reactions, C3H6 could be formed by dehydrogenation of C3H8 (reverse reaction of Eq. (13)) or fragmentation of i-C4H10. This idea was supported by results of pure C3H8 and i-C4H10 injection. When C3H8 and i-C4H10 were the reactant, the selectivity of C3H6 reached 44% and 17%, respectively. The proposed reactions mechanism follows: C3 H8 →C3 H6 þ H2

ð15Þ

iC4 H10 →C3 H6 þ CH4

ð16Þ

iC4 H10 →C3 H6 þ CH2 þ H2 :

ð17Þ

C3H6 produced from Eq. (16) could be suffered from reverse reaction of n-C4H10 formation C3 H6 þ CH4 →nC4 H10 :

ð18Þ

In this case, the probability of Eq. (17) is higher than Eq. (16) as high concentration of H2 was found in the product stream. Moreover, coupling reaction of :H2 diradical could form C2H4, as its selectivity reaches 14%, follows [21]: CH2 þ CH2 →C2 H4 :

ð19Þ

Taken from all above explanation, the reaction pathways of methane conversion in a dielectric barrier discharge at flow rate of 30 ml/min and supplied power of 60 W are shown in Fig. 2.

4.

Kinetic model

In order to quantify the rate of the reaction, a kinetic model was built according to the algorithm (Section 2.3). Some assumptions were used to reduce the complexity of the reactions: (i) all the reactions were first order; (ii) the formation of coke and hydrogen was negligible in the model. All calculations were performed using Matlab by exploiting ‘ode23s’ to solve a set of differential equations in Eq. (4). The reaction rate constants, k (shown in Eq. (5)) were obtained by least-square calculation of the calculated values and the experimental results using ‘fminsearch’ module. The estimated reaction rate constants (k) are listed in Table 2. Higher k value means the reactant transformation to certain product is easier. From Table 2, the value of the k-set supports our previous statements that the decomposition of methane will produce C2H6 (k14), C3H8 (k16), and C4 (k17 and k18). Low value of k13 (CH4 → C2H4) and k12 (CH4 → C2H2) shows that acetylene and ethylene were not coming from the direct methane conversion. Losing two H's and coupling with another triplet :CH2 to form ethylene or losing three H's and coupling with duplet ·CH required many steps. There was also a possibility of attack by

Fig. 2 – The pathway reaction of methane conversion.

Please cite this article as: A. Indarto, et al., Kinetic modeling of plasma methane conversion in a dielectric barrier discharge, Fuel Process. Technol. (2007), doi:10.1016/j.fuproc.2007.09.006

ARTICLE IN PRESS 5

F U E L P RO C ES S I NG T EC H NOL O G Y X X (2 0 0 7) XX X–XXX

Table 2 – reaction rate coefficient (k) kij (1) CH4 (2) C2H2 (3)C2H4 (4) C2H6 (5) C3H6 (6) C3H8 (7) i-C4H10 (8) n-C4H10

(1) CH4 0 1.31 0.08 0.11 0.09 0.06 0.06 0.74

(0.0) (8.1) (8.5) (17.3) (21.8) (13.5) (20.6) (7.2)

(2) C2H2

(3) C2H4

(4) C2H6

(5) C3H6

(6) C3H8

0.06 0 1.79 0.3 0.23 0.97 0.67 1.07

0.18 0.17 0 1.03 0.33 0.16 1.28 2.27

0.5 0.18 0.01 0 0.14 0.49 0.89 0.39

0.02 0.31 0.22 0.07 0 0.48 1.28 1.05

0.3 0.06 0.02 0.67 0.53 0 1.45 0.33

(5.3) (0.0) (8.6) (6.1) (35.1) (11.3) (36.2) (21.8)

(11.6) (12.0) (0.0) (9.2) (7.8) (28.2) (7.7) (6.4)

(6.5) (8.4) (18.9) (0.0) (7.0) (8.1) (21.7) (8.0)

(7.5) (38.6) (6.2) (18.0) (0.0) (10.8) (8.2) (18.3)

(9.8) (20.7) (15.6) (21.7) (10.9) (0.0) (8.3) (23.8)

(7) i-C4H10 0.13 0.14 0.25 0.15 0.54 0.7 0 1.38

(8) n-C4H10

(10.7) (16.1) (9.1) (4.7) (8.5) (7.9) (0.0) (7.7)

0.29 1.47 1.01 0.94 0.53 0.97 2.45 0

(11.1) (9.1) (9.9) (13.7) (7.1) (13.3) (8.9) (0.0)

−1

Note: The value of kij refers to the value of reaction rate coefficient of molecule i to molecule j. All k values are in mol min ; the standard deviation values inside the parentheses are in % unit.

other species before completing the radical coupling reaction. Acetylene and ethylene were formed by stepwise dehydrogenations of ethane. The dehydrogenation reactions of C2H6, occurred in the plasma reaction, could be detected by k43 (C2H6 → C2H4), 1.03 min− 1, then k32 (C2H4 → C2H2), 1.79 min− 1. The trend of the result was similar to that of Jeong et al. [22] which was also suggested by Kozlov et al. [23] that pathways of C2 dehydrogenation reactions will follow: CH4 YC2 H6 Y C2 H4 YC2 H2 : The above stepwise dehydrogenation reaction was also identified in the thermal coupling of methane [24]. Interestingly, n-butane hydrocarbon could be formed from any hydrocarbons species, confirmed by higher value of kx8, N0.5 min− 1, except for methane. This result could answer the high yield production of C4 compounds that usually produced in DBD using AC power supply. This result confirmed the previous research that the addition of higher hydrocarbons, e.g. ethane, on plasma methane conversion would increase the selectivity of butane [18]. Later on, intermolecular transformation of n-butane to i-butane occurred as i-butane is thermodynamically more stable than n-butane, showed by k87 (1.38 min− 1). In a parallel way, homolytic dissociation of n-butane to C2 hydrocarbons (acetylene and ethylene) is confirmed by k82 (1.07 min− 1) and k83 (2.27 min− 1) or into propene by k85 (1.05 min− 1). Usually, decomposition of higher hydrocarbon to lower hydrocarbons results higher reaction rate constant k (around 0.5–2 min− 1) compared to the formation of higher hydrocarbons from the lower ones. It was acceptable since the dissociation reaction, e.g. decomposition of one butane molecule, will produce double C2 molecules or even more if it decomposed to C1 molecules. Instead of coming from the intermolecular transformation of n-butane, the formation of i-butane could be formed from C3 compounds, k58 = 0.54 min− 1 and k68 = 0.7 min− 1. Although i-butane will be consumed in further reaction mostly to n-butane, by k78 = 2.54 min− 1, we have to consider the conversion comparison between n-butane and i-butane. The conversion rate of i-butane is 10% lower than that of n-butane, which means that i-butane is more stable than n-butane. The reverse reaction to produce methane was mostly coming from acetylene, confirmed by k21 (1.31 min− 1) which is higher than for other reactants. From this data, we are also able to determine that the decomposition of higher hydrocarbons will not produce CH4 as the product but instead will

produce ·CH3, :CH2, or ·CH. The existence of C1 radical hydrocarbons could help to increase the selectivity of C2, C3, and C4 compounds by coupling reactions. Unfortunately, as this method is just counting the tendency of the reaction pathway, the presence of intermediate species could not be confirmed scientifically. By taking only the important and significant k values, we can draw a global reaction pattern diagram of methane conversion (presented in Fig. 2). It shows that C2H4 has an important role as a branching agent to diversify the reaction mechanism. This idea is supported by the data because C2H4 was easier to produce than other C2 compounds from the cracking of higher hydrocarbons (C3 and C4). To form C2H6, the reactions should be terminated by recombination reaction with H2. On the other hand, C2H2 formation required further dehydrogenation reaction to remove two hydrogen atoms. Instead of dehydration reaction into C2H2, coupling reaction of C2H4 could form n-C4H10 (Eq.(12)) or react with CH2 radical to form C3H6. In order to check the correctness of the model, a binary mixture of methane and acetylene was injected to the reactor at the total flow rate of 30 ml/min and supplied power of 60 W. As shown in Fig. 3, the model calculation was found to be in good agreement with the experimental data. It also means that the proposed values of reaction rate constant, k, as well as the model can rationalize the plasma mechanism of methane and/or hydrocarbon conversion.

Fig. 3 – The comparison between experimental and model simulation results of binary mixture methane + acetylene at the flow rate of 30 ml/min and supplied power of 60 W.

Please cite this article as: A. Indarto, et al., Kinetic modeling of plasma methane conversion in a dielectric barrier discharge, Fuel Process. Technol. (2007), doi:10.1016/j.fuproc.2007.09.006

ARTICLE IN PRESS 6

F U E L P RO C ES S IN G T EC HN O L O G Y X X (2 0 0 7) XXX –XXX

5.

Conclusions

The kinetics of methane conversion in a dielectric barrier discharge at a supplied power of 60 W was studied. It shows that C2H4 has an important role on the global mechanism, having a function as a branching molecule which is able to diverse the global reaction mechanism. Similar to thermal methane decomposition reaction, dehydrogenation of C2H6 has also occurred and producing C2H4 then continued by formation of C2H2. Coupling reaction of C2H4 with other C2 hydrocarbons would yield butane which usually has a longchain molecule (normal-form). Using this method which is relatively simple, the pathway pattern of CH4 conversion could be obtained. It will be a good advantage to predict or optimize the end products of direct methane conversion by a dielectric barrier discharge. Further model improvement is still necessary to be able to conduct the calculation in various conditions, e.g. different flow rates, which are not covered in this research.

Acknowledgments The authors would like to thank the Korea Institute of Science and Technology (KIST) and the Korea University (KU) for the study supports. The first author would like to express his appreciation to the Università degli studi di Torino for the support during the study period in Turin, Italy.

REFERENCES [1] M.J. Gradassi, N.W. Green, Economics of natural gas conversion processes, Fuel Process. Technol. 42 (1995) 65–83. [2] A.T. Bell, Mechanisms of Fischer–Tropsch Synthesis, 1984, NTiS report number: DE84014608. [3] H. Zanthoff, M. Baerns, Oxidative coupling of methane in the gas phase. Kinetic simulation and experimental verification, Ind. Eng. Chem. Res. 29 (1990) 2–10. [4] M. Baerns, K. van der Wiele, J.R.H. Ross, Catal. Today 4 (1989) 471. [5] J.A. Labinger, Quimica (1993) 48. [6] M. Belgued, H. Amariglio, P. Pareja, A. Amariglio, J. Saint-Just, Catal. Today 13 (1992) 437. [7] T. Koerts, R.A. van Santen, New Frontiers in Catalysis, in: L. Guczi, et al., (Eds.), Proc. 10th Int. Congress on Catalysis, Elsevier, Amsterdam, 1993, p. 1065.

[8] M. Belgued, A. Amariglio, P. Pareja, H. Amariglio, J. Catal. 159 (1996) 441. [9] F. Solymosi, A. Erdöhelyi, J. Cserenyi, Catal. Letters 16 (1992) 399. [10] L. Guczi, R.A. van Santen, K.V. Sarma, Catal. Rev., Sci. Eng. 38 (1996) 249. [11] A. Indarto, J.W. Choi, H. Lee, H.K. Song, Methane conversion using dielectric barrier discharge: comparison with thermal process and catalyst effects, J. Nat. Gas Chem. 15 (2006) 87–92. [12] A. Holmer, O. Olsvik, O.A. Rokstad, Fuel Process. Technol. 42 (1995) 249. [13] Y. Yang, Direct non-oxidative methane conversion by non-thermal plasma: experimental study, Plasma Chem. Plasma Process (2003) 23. [14] S.S. Kim, H. Lee, B.K. Na, H.K. Song, Reaction pathways of methane conversion in dielectric-barrier discharge, Korean J. Chem. Eng. 20 (2003) 869–872. [15] A. Indarto, J.W. Choi, H. Lee, H.K. Song, Kinetic modeling of plasma methane conversion using gliding arc plasma, J. Nat. Gas Chem. 14 (2005) 13–21. [16] A. Indarto, J.W. Choi, H. Lee, H.K. Song, Effect of additive gases on methane conversion using gliding arc discharge, Energy 31 (2006) 2650–2659. [17] M. Frenklach, Reaction mechanism of soot formation in flames, Phys. Chem. Chem. Phys. 4 (2002) 2028–2037. [18] K. Thanyachotpaiboon, S. Chavadej, T.A. Caldwell, L.L. Lobban, R.G. Mallinson, Conversion of methane to higher hydrocarbons in AC nonequilibrium plasmas, AIChE J. 44 (1998) 2252–2257. [19] G.-B. Zhao, S. John, J.-J. Zhang, L. Wang, S. Muknahallipatna, J.C. Hamannb, J.F. Ackerman, M.D. Argyle, O.A. Plumb, Methane conversion in pulsed corona discharge reactors, Chem. Eng. J. 125 (2006) 67–79. [20] M. Kraus, W. Egli, K. Haffner, B. Eliasson, U. Kogelschatzb, A. Wokaun, Investigation of mechanistic aspects of the catalytic CO2 reforming of methane in a dielectric-barrier discharge using optical emission spectroscopy and kinetic modeling, Phys. Chem. Chem. Phys. 4 (2002) 668–675. [21] H.S. Kim, S.J. Oh, A study of the synthesis of C2-hydrocarbons using methane activation, J. Ind. Eng. Chem. 4 (1998) 12–18. [22] H.K. Jeong, S.C. Kim, C. Han, H. Lee, H.K. Song, B.K. Na, Conversion of methane to higher hydrocarbons in pulsed DC barrier discharge at atmospheric pressure, Korean J. Chem. Eng. 18 (2001) 196–201. [23] K.V. Kozlov, K. Michel, H.E. Wagner, Synthesis of organic compounds from mixtures of methane with carbon dioxide in dielectric-barrier discharges at atmospheric pressure, Plasma Polymer. 5 (2000) 129–150. [24] T.V. Choudhary, E. Aksoylu, D.W. Goodman, Nonoxidative activation of methane, Catal. Rev. 45 (2003) 151–203.

Please cite this article as: A. Indarto, et al., Kinetic modeling of plasma methane conversion in a dielectric barrier discharge, Fuel Process. Technol. (2007), doi:10.1016/j.fuproc.2007.09.006