Dielectric barrier discharge plasma reactor for non-oxidative coupling of methane having a controlled gap distance between dielectric particles and regeneration method of deactivated bed in the same

Abstract

Provided are a dielectric barrier discharge (DBD) plasma reactor including dielectric particles in a packed-bed in a discharge zone, e.g., a DBD plasma reactor for non-oxidative coupling of methane in which an average gap distance between dielectric particles in the packed-bed is adjusted to improve methane conversion and/or product selectivity; a method of regenerating dielectric particles including removing coke, which sis produced by side reactions, from the dielectric particles deactivated by the coke by using a low temperature plasma in an oxidizing atmosphere in the reactor; a method of manufacturing C.sub.2+ hydrocarbons, the method including converting methane into C.sub.2+ hydrocarbons including ethylene and/or ethane by non-oxidative coupling of methane in the reactor; and a method of manufacturing hydrogen, the method including generating hydrogen from methane by non-oxidative coupling of methane in the reactor.

Claims

1. A dielectric barrier discharge (DBD) plasma reactor comprising: a tube having a first end and a second end; a gas inlet coupled to the first end; a gas outlet coupled to the second; dielectric particles that are disposed in the tube to form a plasma bed, the dielectric particles being packed in a discharge zone of the tube, the dielectric particles being packed between layers of supporting material, wherein the dielectric particles having diameter of 53 μm to 100 μm, wherein the DBD plasma reactor is a reactor for non-oxidative coupling of methane designed to cause a non-oxidative coupling reaction of methane, wherein a mean value of a gap distance between the dielectric particles in the bed which are polarized by external electric field is 4 μm to 5 μm.

2. The DBD plasma reactor of claim 1, wherein the gap distance between the dielectric particles in the bed is determined by particle size.

3. The DBD plasma reactor of claim 1, wherein the gap distance between the dielectric particles is selected to obtain desired methane conversion rate and C2 selectivity.

4. The DBD plasma reactor of claim 1, wherein C—H bonds are activated without additional thermal energy and oxidant molecules to produce methyl radicals and directly produce C2-C4 light hydrocarbons.

5. The DBD plasma reactor of claim 1, wherein a particle size of the dielectric particles is selected to inhibit formation of coke on the dielectric particles during the reaction or to control timing of removing coke formed on the dielectric particles during the reaction.

6. The DBD plasma reactor of claim 1, wherein the reactor is designed to remove coke formed on the dielectric particles by side reactions by plasma treatment in an oxidizing atmosphere.

7. The DBD plasma reactor of claim 1, wherein the reactor is designed to remove coke inevitably accompanied by side reactions during the reaction occurring in the DBD plasma reactor by using the same type of plasma as that used in the reaction by supplying an oxygen-containing mixture, instead of a reaction mixture, in a regeneration process of removing the coke.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) FIG. 1 is a schematic conceptual diagram illustrating streamers and reactive intermediates between dielectric particles by an induced microelectric field.

(2) FIG. 2 shows scanning electron microscope (SEM) images of (a) fresh α-Al.sub.2O.sub.3 (S), (b) fresh α-Al.sub.2O.sub.3 (M), (c) fresh α-Al.sub.2O.sub.3 (L), (d) fresh sea sand (S), (e) fresh sea sand (M), (f) fresh sea sand (L), (g) fresh KIT-6 (S), (h) fresh KIT-6 (M), and (i) fresh KIT-6 (L).

(3) FIG. 3 shows graphs illustrating distributions of gap distance between particles obtained from the SEM images of fresh samples.

(4) FIG. 4 is a schematic diagram of a bed plasma reactor system.

(5) FIG. 5 shows Lissajous curves of measured voltage (V) and charge (Q) of (a) α-Al.sub.2O.sub.3 samples, (b) sea sand samples, and (c) KIT-6 samples.

(6) FIG. 6 shows reaction performance of cases without dielectric particles (blank) and with α-Al.sub.2O.sub.3, sea sand, and KIT-6 at TOS 60 min: (a) of FIG. 6 shows CH.sub.4 conversion rates and product selectivities, and (b) of FIG. 6 shows CH.sub.4 conversion rates and yields of products.

(7) FIG. 7 shows voltage-current profiles of the DBD plasma reactor packed with (a) α-Al.sub.2O.sub.3 (S), (b) α-Al.sub.2O.sub.3(M), (c) α-Al.sub.2O.sub.3(L), (d) sea sand (S), (e) sea sand (M), (f) sea sand (L), (g) KIT-6 (S), (h) KIT-6 (M), and (i) KIT-6 (L).

(8) FIG. 8 shows a reaction pathway for describing methane activation, dehydrogenation, coupling, and chain-growth reactions, in which C.sub.4 and C.sub.5+ indicate hydrocarbon molecules having 4 carbon atoms and 5 or more carbon atoms, respectively.

(9) FIG. 9 shows XRD patterns of (a) spent α-Al.sub.2O.sub.3, (b) spent sea sand, and (c) spent KIT-6 samples, in which (S), (M), and (L) respectively indicate sizes of the samples, (F) indicates a fresh sample not used, and α, θ, and . are peaks of α-Al.sub.2O.sub.3, θ-Al.sub.2O.sub.3, and quartz, respectively.

(10) FIG. 10 shows small angle X-ray scattering (SAXS) pattern of (a-b) fresh & spent KIT-6 (S), (c-d) fresh & spent KIT-6 (M), and (e-f) fresh & spent KIT-6 (L).

(11) FIG. 11 shows transmission electron microscope (TEM) images of (a) fresh α-Al.sub.2O.sub.3, (b) fresh sea sand, (c) fresh KIT-6, (d) spent α-Al.sub.2O.sub.3, (e) spent sea sand, and (f) spent KIT-6.

(12) FIG. 12 shows SEM images of (a) spent α-Al.sub.2O.sub.3 (S), (b) spent α-Al.sub.2O.sub.3 (M), (c) spent α-Al.sub.2O.sub.3(L), (d) spent sea sand (S), (e) spent sea sand (M), (f) spent sea sand (L), (g) spent KIT-6 (S), (h) spent KIT-6 (M), and (i) spent KIT-6 (L).

(13) FIG. 13 shows TEM image analysis results of (a) spent α-Al.sub.2O.sub.3, (b) spent sea sand, and (c) spent KIT-6 by EDS.

(14) FIG. 14 shows TG/DTA results of (a) spent α-Al.sub.2O.sub.3, (b) spent sea sand, and (c) spent KIT-6.

(15) FIG. 15 shows FT-IR spectra of (a) α-Al.sub.2O.sub.3, (b) sea sand, and (c) KIT-6. .box-tangle-solidup., •, .square-solid., and * indicates CH.sub.3 stretch vibration mode, C≡C stretching mode or cyano group, C═C stretching mode, and asymmetric C—H bending mode of methylene group in a long aliphatic chain.

(16) FIG. 16 is a schematic diagram of a dielectric barrier discharge plasma reactor.

(17) FIG. 17 shows TGA results of packing materials collected after Example 2, Example 3, and Comparative Example 1.

(18) FIG. 18 shows FR-IR analysis results of packing materials collected after Preparation Example 3, Example 2, Example 3, and Comparative Example 1.

(19) FIG. 19 shows TEM image analysis results of packing materials collected after Example 2, Example 3, and Comparative Example 1.

(20) FIG. 20 shows small angle X-ray scattering (SAXS) analysis results of packing materials collected after Preparation Example 3, Example 2, Example 3, and Comparative Example 1.

MODE OF DISCLOSURE

(21) Hereinafter, the present invention will be described in more detail with reference to the following examples. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present invention.

Preparation Example 1: Preparation of Materials

(22) The α-Al.sub.2O.sub.3 was prepared by thermal treatment of γ-Al.sub.2O.sub.3(Sigma-Aldrich, USA) at 1000° C. for 8 hours.

(23) Sea sand was purchased from Fisher Chemical, USA.

(24) Ordered mesoporous silica (KIT-6) was prepared in the following method.

(25) 6 g of a triblock copolymer P123 (EO.sub.20PO.sub.70EO.sub.20, MW=5800 g/mol, Sigma-Aldrich, USA) was dissolved in 217.64 g of deionized water and 11.16 g of HCl (37%, Sigma-Aldrich, USA) at 35° C. while agitating to prepare a homogenous solution. Then, 6 g of 1-butanol (Sigma-Aldrich, USA) was added to the mixture and stirred for 2 hours. 12.77 g of a silica precursor, tetraethoxy silane (Alfa Aesar, USA) was slowly added to the mixture and stirred for 24 hours. This mixture was added to a polypropylene bottle and hydrothermally treated at 100° C. for 24 hours and washed three times with deionized water and ethanol. Subsequently, the washed sample was placed in an oven and dried at 110° C. for 24 hours. The dried sample was calcined at 550° C. for 5 hours.

(26) All materials were separated to three groups (S, M and L; 0<S<53<M<100<L<150 m) in terms of particle size by using stainless steel sieves.

Experimental Example 1: Analysis of Characteristics of Materials

(27) For each material prepared in Preparation Example 1, 3 groups with different sizes were prepared and the groups were named S, M, and L as shown in Table 1. SEM images are provided in FIG. 2. Gap size distribution between particles are shown in FIG. 3 and Table 1. In this regard, the gap is significantly related to the particle size. In the case of the α-Al.sub.2O.sub.3 particles, mean values of gap between S, M, and L particles were 4.30 m, 4.73 m, and 9.87 μm, respectively.

(28) In the cases of the sea sand and KIT-6 particles, the gap increases with the particle size. As described above, the particles of each group appeared to have log-normal distribution.

(29) TABLE-US-00001 TABLE 1 Standard Mean Minimum deviation Sample (μm) (μm) (μm) α-Al.sub.2O.sub.3 (S) 4.30 0.55 3.22 α-Al.sub.2O.sub.3 (M) 4.73 1.12 2.37 α-Al.sub.2O.sub.3 (L) 9.87 3.52 4.15 sea sand (S) 3.04 0.77 0.94 sea sand (M) 4.10 1.05 1.09 sea sand (L) 10.87 4.28 4.44 KIT-6 (S) 2.07 0.59 0.78 KIT-6 (M) 4.23 2.14 0.97 KIT-6 (L) 8.43 3.16 3.15

Example 1: Packed-Bed DBD Plasma Reactor and Activity Test

(30) Non-oxidative methane coupling reaction was conducted in a lab-made packed-bed DBD plasma reactor system (FIG. 4) at atmospheric pressure and near room temperature.

(31) A volumetric flow rate of a methane mixture (CH.sub.4:N.sub.2=1:1) was 40 standard cubic centimeter per minute (sccm), and a total time for the reaction was 1000 minutes. An alumina tube having an internal diameter of 6 mm and a thickness of 2 mm was used as a dielectric barrier for the plasma bed. A stainless steel rod having a diameter of 3 mm was used as a powered electrode, and a steel wire was used as a ground electrode. A 150 mm-long discharge zone was covered with the ground electrode. A discharge gap between the inner surface of the alumina tube and a high-voltage electrode was 1.5 mm, a volume of the plasma discharge zone was fixed to 3.181 cm.sup.3, and a space velocity (SV) based on the volume was set to 754.5 h.sup.−1. Each dielectric packing material was fully packed in this region. Each dielectric packing material was fully packed in this region. A sinusoidal AC power supply (0-220 V, 60-1000 Hz) was connected to a transformer (0-20 kV, 1000 Hz), and this electrical system continuously applied a high voltage to the plasma bed. The applied voltage and the frequency to the plasma bed were fixed as 15 kV and 1 kHz, respectively. A capacitor with capacitance of 1 μF was connected in series between the plasma bed and the ground. The voltage applied to the plasma bed was measured by employing a high-voltage probe (1000:1, P6015A, Tektronix). The voltage across the 1 μF capacitor was measured by employing a voltage probe (10:1, P6100, Tektronix) connected to each side of the capacitor. A current probe (TCP202, Tektronix) was connected on the ground electrode to evaluate the current profile across the DBD plasma bed. The probes were connected to a digital oscilloscope (TDS 3012C, Tektronix). The accumulated electric charge in the plasma bed was calculated by multiplying the voltage across the capacitor and the capacitance of the capacitor (1 μF). FIG. 5 shows Lissajous curves of measured voltage (V) and charge (Q) of (a) α-Al.sub.2O.sub.3 samples, (b) sea sand samples, and (c) KIT-6 samples. Table 2 shows discharge power calculated by Q-V Lissajous method, calculated energy yields of total C2 and unsaturated C2 hydrocarbon products per discharge power, weight of particles in each bed, breakdown voltage in each bed, and calculated average threshold electric potential difference between particles.

(32) During the reaction, the reactor temperature was measured with an IR temperature detector. The temperature at inlet was nearly room temperature and the temperature of the bed was monitored. The maximum temperature was observed in the central region of the reactor as 100° C. The observed temperatures in the other regions were below 100° C. and most were close to room temperature. An external insulation or an oven was used in this reactor system.

(33) An effluent gas from the plasma bed was analyzed by an on-line gas chromatograph (6500GC Young Lin Instrument Co., Korea) employing a Porapak-N and a Molecular Sieve 13× columns connected with a thermal conductivity detector (TCD) and a GS-GasPro column connected with a flame ionization detector (FID). H.sub.2, N.sub.2, and CH.sub.4 in the effluent were detected by using the TCD. CH.sub.4, C.sub.2H.sub.2, C.sub.2H.sub.4, C.sub.2H.sub.6, C.sub.3H.sub.6, C.sub.3H.sub.8, 1-C.sub.4H.sub.8, and n-C.sub.4H.sub.10 in the effluent gas were detected by using the FID. The energy yield (Yi) of product i and the corresponding discharge power was rigorously calculated in Table 2.

(34) TABLE-US-00002 TABLE 2 Discharge Weight of Breakdown Threshold power Y.sub.total C2 V.sub.unsaturated C2 particles voltage .sup.a ΔV .sup.b Sample (W) (g/kWh) (g/kWh) (g) (kV) (V) α-Al.sub.2O.sub.3 (S) 44.1 11.26 9.21 5.09 5.56 91.4 α-Al.sub.2O.sub.3 (M) 43.5 8.61 5.98 5.06 5.53 90.9 α-Al.sub.2O.sub.3 (L) 42.5 8.11 4.27 3.77 5.60 109 sea sand (S) 43.0 7.28 5.32 5.37 6.52 84.0 sea sand (M) 42.0 6.98 4.56 5.27 6.49 79.0 sea sand (L) 42.3 5.50 2.35 4.87 6.93 106 KIT-6 (S) 39.0 4.65 3.02 1.28 4.91 143 KIT-6 (M) 38.5 6.83 4.32 1.00 4.56 81.5 KIT-6 (L) 36.1 7.66 3.62 0.91 5.95 95.8 .sup.a Measured from Q-V curves in FIG. 5. .sup.b Threshold electric potential difference (ΔV) between particles to initiate plasma discharges, calculated from Equation 1.

(35) 1.1. Activity Tests in DBD Plasma Bed

(36) As shown in FIG. 6, the packed beds showed the higher conversion rate in all the cases than the conversion rate of blank test at the early stage of reaction due to the enhanced intensity of electric field between dielectric particles. In (a) of FIG. 6, selectivities of C2 compounds such as ethylene and acetylene were far higher in the packed-bed tests than those in the blank test. In contrast, the selectivity for ethane in the blank test was shown to be higher than those of packed-bed tests. In terms of yield, the unsaturated C2 compounds in the packed-bed tests were produced more than those in the blank test as shown in (b) of FIG. 6. Molar balances on hydrogen and carbon in each test were calculated in Table 3. Due to the high initial activity, significant amount of coke was generated and a few of carbon balances were measured less than 100%.

(37) TABLE-US-00003 TABLE 3 Sample Blank α-Al.sub.2O.sub.3 sea sand KIT-6 Size — S M L S M L S M L TOS 60 min CB (%) .sup.a 94.22 98.15 82.70 86,68 79.97 75.02 81.10 96.84 75.86 85.32 HB (%) .sup.b 97.01 98.57 93.51 93.65 95.46 95.09 92.41 99.77 92.03 94.47 TOS 300 min CB (%) .sup.a 93.50 94,00 95.44 95.89  99.12 98.09 90.61 98.71 97.49 94.12 HB (%) .sup.b 96.09 93.33 95.11 97.19 100.00 99.79 97.43 99.06 98.77 97.14 ${\hspace{0.17em}}^a\mathrm{Carbon}\mathrm{Balance}\left(\mathrm{CB}\right)\left(\%\right)=\frac{\begin{array}{c}\mathrm{Moles}\mathrm{of}{\mathrm{CH}}_4\mathrm{not}\mathrm{converted}+\\ .Math.\left(x\times \mathrm{Moles}\mathrm{of}{C}_2{H}_y\mathrm{produced}\right)\end{array}}{\mathrm{Moles}\mathrm{of}{\mathrm{CH}}_4}\times 100$${\hspace{0.17em}}^b\mathrm{Hydrogen}\mathrm{Balance}\left(\mathrm{HB}\right)\left(\%\right)=\frac{\begin{array}{c}4\times \mathrm{Moles}\mathrm{of}{\mathrm{CH}}_4\mathrm{not}\mathrm{converted}+\\ 2\times \mathrm{Moles}\mathrm{of}{H}_2\mathrm{produced}+\\ .Math.\left(y\times \mathrm{Moles}\mathrm{of}{C}_2{H}_y\mathrm{produced}\right)\end{array}}{4\times \mathrm{Moles}\mathrm{of}{\mathrm{CH}}_4\mathrm{in}\mathrm{the}\mathrm{feed}}\times 100$

(38) It was interesting to find that the conversion seemed to have its maximum when the middle-sized particles (size M) were used, irrespective of materials. As described above, by using the new concept of micro-electrodes between dielectric particles, the threshold electric potential difference between polarized dielectric particles may be estimated by employing a slightly modified calculation method from the original Paschen's law, which may be used to estimate a breakdown voltage between electrodes.

(39) Through experimental observation, the maximum conversion rate of methane may be obtained by packing M particles. The gap distance between M particles was in the range of 4 m to 5 m as shown in FIG. 2 and Table 2.

(40) To evaluate the value of threshold electric potential difference, the value of γ was estimated by applying the modified Paschen's equation and the gap at the maximum conversion rate. Since γ is affected by numerous factors, it is known that it is very difficult to estimate the exact value of γ. In the embodiment, the Paschen's equation was applied to estimate secondary electron emission coefficient, γ, for all employed materials for the reaction. This application was based on experimental observation, particularly methane conversion rate in the plasma bed and each gap distance between the particles. By comparing the results of the packed-bed tests with the blank tests, the increased CH.sub.4 conversion rate and product selectivity were observed as shown in this study. The level of conversion rate and selectivity seem to be dependent on reaction conditions, types of materials, applied power, and the like.

(41) 1.2. Effects of Micro-Discharge on Performance

(42) FIG. 7 shows voltage-current profiles in each packed-bed test. In all cases, several current pulses were observed. These current pulses indicate generation of micro-discharges in the plasma bed.

(43) Table 4 shows average number and average intensity of micro-discharge current pulses per sample measured with voltage-current profile in the packed-bed DBD reactor.

(44) TABLE-US-00004 TABLE 4 Average number of valid current pulses Average intensity corresponding to of microdischarge microdischarges current pulses Sample per one cycle (—) (mA) Blank 3.3 39.0 α-Al.sub.2O.sub.3 (S) 10.0 47.5 α-Al.sub.2O.sub.3 (M) 13.0 54.0 α-Al.sub.2O.sub.3 (L) 9.3 61.8 sea sand (S) 12.0 48.9 sea sand (M) 13.0 67.5 sea sand (L) 11.0 94.8 KIT-6 (S) 12.3 63.5 KIT-6 (M) 13.3 73.7 KIT-6 (L) 6.0 85.9

(45) Referring to FIG. 7 and Table 4, the increased number and intensity of micro-discharges were observed in the cases of packed-bed tests compared with those in the case of blank test.

(46) Also, as shown in Table 4, when large particles were packed, the intensity of micro-discharges increased. As illustrated in FIG. 7, when M particles were used, the number of valid micro-discharges was greatest. Besides the number of contact points, the number of micro-discharges seemed to have strong relation with electric property of dielectric particles such as capacitance, as previously explained. As a result, the maximum conversion rate was observed in the cases of M particles due to the greatest number of micro-discharges with medium intensity.

(47) Table 2 shows discharge power and weight of packed particles. As the particle size decreases, the weight of particles as well as the discharge power seemed to increase in each material (α-Al.sub.2O.sub.3, sea sand, and KIT-6). The tendency of discharge power in the cases of sea sand samples appeared to be very slightly deviated at L particles, but it may be due to experimental error. It was understood that the increase in bed weight may require more discharge power to polarize the dielectric particles. In addition, Table 2 shows the breakdown voltage in the bed and the threshold electric potential difference between particles. The breakdown voltage was estimated by using Lissajous curves in FIG. 5, and the threshold electric potential difference was computed by using the modified Paschen's equation (Equation 1) to explain the potential difference to initiate streamer or discharge between dielectric particles. The breakdown voltage and the threshold electric potential difference showed their minima in the cases of M particles. This seems highly related to the fact that the conversion showed its maximum in the cases of M particles. The low threshold electric potential difference between the particles seemed to lead to the low breakdown voltage in the entire bed. This reduced breakdown voltage must have facilitated easier formation and a greater number of micro-discharges in the bed as shown in FIG. 7 and Table 4, which resulted in higher converting capability.

(48) 1.3. Reaction Pathway Under DBD Plasma

(49) FIG. 8 shows a reaction pathway under the DBD plasma for describing methane activation, dehydrogenation, coupling, and chain-growth reactions.

(50) Regarding unsaturated C2 compounds, the selectivity was shown to increase as the size of particles decreased. The CH.sub.x species dehydrogenated by plasma seemed to have two different routes to be coupled into unsaturated C2 compounds. In the first route, methane is dehydrogenated to a CH.sub.3 radical and coupled with another CH.sub.3 radical to form C.sub.2H.sub.6 (ethane) or C.sub.2H.sub.5 species. A further dehydrogenation takes place stepwise to produce C.sub.2H.sub.4 (ethylene) and C.sub.2H.sub.2 (acetylene). In the second route, the CH.sub.3 radical further dehydrogenated to CH.sub.2 or CH, which are directly coupled into C.sub.2H.sub.4 and C.sub.2H.sub.2, respectively.

(51) For S particles, the total selectively for hydrocarbons (selectivity for C2-C4) was found to be the highest. It may result from the highest specific surface area (highest surface to volume ratio for the smallest particles in the cases of nonporous materials such as α-Al.sub.2O.sub.3 and sea sand particles). This was also valid for highly porous KIT-6 particles. In general, the specific area increases as the size of particle decreases. The specific surface area is proportional to the amount of surface oxygen ion vacancy sites, which are named V-centers and highly related to the generation of methyl radicals (CH.sub.3). According to Liu et al., when electronically exited states return to the ground states, energy is emitted in the form of electromagnetic radiation. Such radiation accounts for the ultraviolet to visible emissions of the gas discharge, and the V-center is formed due to the irradiation. The methyl radicals may be formed by interaction with the V-center and methane. According to Ozin et al., the V-center photoactivated by UV irradiation plays a role to dehydrogenate alkane through hydrogen abstraction, and a further dehydrogenation may facilitate the formation of unsaturated forms of hydrocarbons from saturated (i.e., formation of ethylene from ethane). The effect of V-center seems very similar to a catalyst. However, the V-center is thought to be an initiator generating radicals rather than a traditional catalyst.

(52) Besides, the decreased fraction of void space due to the small size particles turned out to be a denser environment and a higher pressure. Under the more compressed condition, the number of effective collisions between intermediate radicals seemed to be increased. In consideration of very short lifetimes of radicals, the small gap distance might have helped the increase in the number of effective collisions avoiding termination without chain-growth.

(53) As a result, the selectivity for unsaturated hydrocarbons and the total hydrocarbon selectivity in the case of S particles were found to be the highest among the 3 different sizes (S, M, and L) regardless of the type of material (α-Al.sub.2O.sub.3, sea sand, and KIT-6). This was also valid for either practically nonporous or highly porous material.

(54) As a result of observing the phenomenon in the case of the small size particles, the present inventors have found that the dehydrogenated species such as CH.sub.2 and CH had higher probability to collide with each other and with their types, and be subsequently coupled into unsaturated C2 compounds. In contrast, if the size of particles was large (L) and the space between particles was also large, the dehydrogenation seemed to occur more frequently than the coupling, which resulted in additional carbon deposition. The increased amount of carbon deposition due to the dehydrogenation seemed to be the result of increased capacitance of large particles. As observed in FIG. 7, the intensity of discharges between L particles was increased and it has been known to increase with capacitance of particles and this led to the amount of charge transferred by an individual micro-discharge was increased, although the number of discharges was decreased due to decreased specific surface area in the cases of L particles.

(55) Due to the large size, the specific surface area is relatively small compared with smaller particles. This is directly related to the number of V-centers. Because of the relatively small number of V-centers, the amount of ethylene from ethane at V-centers seemed to be smaller than that in the cases of smaller particles. In addition, the radicals had relatively low possibility to collide effectively for coupling. Instead, quite a few methyl radicals seemed to have followed the dehydrogenation route (the second route in the above-described mechanism).

(56) 1.4. Analysis Results after Plasma Coupling Reaction

(57) In the mechanism of micro-discharge generation described above, the generated C.sub.xH.sub.y radicals may collide with surfaces of the dielectric particles. Due to these collisions, the radicals seemed to be attached to the surfaces and left to form a carbonaceous deposition, which was observed in the spent samples. According to Table 3, the molar balances on carbon in a few samples were quite lower than 100% at the initial stage (TOS 60 min) comparing to the data of TOS 300 min. These results indicate that this type of carbonaceous deposition seemed to be generated dominantly at the initial stage.

(58) FIG. 9 shows results of wide angle X-ray diffraction spectroscopy (WAXRD) for spent samples and original fresh samples (not size-controlled). The alumina samples in (a) of FIG. 9 showed that the most dominant phase in the original fresh sample (F) was α-phase, but a small fraction of 0-phase was also detected in all the samples. In (b) of FIG. 9, the quartz phase was clearly seen in the fresh and the spent samples. The XRD result in (c) of FIG. 9 showed that all the fresh and spent KIT-6 samples were found amorphous. As shown herein, the phase and the crystallinity of each sample (alumina and sea sand samples) did not change after the plasma coupling reaction. Regarding the KIT-6, the SAXS experiment was utilized to verify the structural stability after the reaction. All the fresh samples showed highly ordered mesoporous structure known as Ia3d bicontinuous phase (a, c, and e of FIG. 10). After the plasma coupling reaction, they successively retained their highly ordered structure, and the spectrum of each spent sample (b, d, and f) showed no significant alternations. The spectrum intensity of spent KIT-6 (L) seemed to be a little bit weakened possibly due to carbon deposition.

(59) FIG. 11 shows transmission electron microscope (TEM) images of fresh samples (a to c) and spent samples (d to f). The structural changes were barely observed in the spent samples. In FIG. 12, the particle surface of each spent sample was observed by the SEM images. The changes on the surface of spent samples could be observed and the amount of deposition seemed to be increased as the particle size increased, although the deposited elements such as carbon were hardly identified with these images. FIG. 13 shows the results of TEM imaging analyses with EDS of spent samples. In the cases of spent α-Al.sub.2O.sub.3 (a) and spent sea sand (b), the carbon deposition on the surface did not seem to be significant, whereas in the case of the spent KIT-6 (c), a significant amount of carbon was observed on the spent KIT-6.

(60) To assess the carbon deposition, thermogravimetric analysis (TGA) and differential thermal analysis (DTA) were conducted for the spent samples. As shown in FIG. 14 and Table 5 (TG/DTA results of spent α-Al.sub.2O.sub.3, spent sea sand, and spent KIT-6 samples), the amount of carbon deposition was increased as the size of particles increased. It was understood that the amount of carbon deposition was increased since the increased intensity of micro-discharges for large particles accelerated dehydrogenation.

(61) As shown in the result of TG/DTA, two major peaks were observed at around 400° C. and at a temperature of 500 to 600° C. indicating two different carbon species were deposited. From the XRD results, most of the carbon species were found amorphous, and may be easily removed through traditional oxidation treatments or plasma irradiation techniques. Particularly, the amount of coke in the highly porous KIT-6 was significantly greater than those in the spent α-Al.sub.2O.sub.3 and sea sand samples. This was possibly due to the fact that the carbon species might have grown into the numerous pores of KIT-6 samples as the reaction proceeded.

(62) TABLE-US-00005 TABLE 5 lower higher Weight DTA peak DTA peak loss (%) (° C.) (° C.) α-Al.sub.2O.sub.3 (S) 4.32 387 505 α-Al.sub.2O.sub.3 (M) 4.67 373 506 α-Al.sub.2O.sub.3 (L) 9.67 378 506 sea sand (S) 2.52 376 518 sea sand (M) 3.54 379 529 sea sand (L) 5.61 379 533 KIT-6 (S) 9.95 396 648 KIT-6 (M) 31.3 395 624 KIT-6 (L) 37.3 398 606

(63) FIG. 15 shows the result of FT-IR observation. The similar species of carbon were detected in all the spent samples. At 2850 and 3000 cm.sup.−1, a CH.sub.3 stretch vibration mode was observed. The peak at 2200 cm.sup.−1 may be assigned to the carbon-carbon triple bond stretching mode, or cyano group (—CN) due to chemical reaction between CH.sub.4 and N.sub.2. The peaks at 1640 cm.sup.−1 and 1463 cm.sup.−1 may be ascribed to the carbon-carbon double bond stretching mode, and the asymmetric C—H bending mode of methylene group in a long aliphatic chain, respectively. Considering all these features, the carbon species in the FT-IR spectra could be classified as carbon deposition having long-chain hydrocarbons.

Preparation Example 2

(64) An experimental device of dielectric barrier discharge plasma is shown in FIG. 1. As a plasma reactor, an alumina tube having a length of 700 mm, an external diameter of 10 mm, and an internal diameter of 6 mm was used. This alumina tube served as a dielectric barrier. A stainless steel rod having a diameter of 3 mm was used as a high-voltage electrode, and a steel wire having a thickness of 0.5 mm and wound on the alumina tube in a spring form by 150 mm was used as a ground electrode. An AC power supply (0 to 220 V, 60 to 1000 Hz) was connected to a transformer (0 to 20 kV, 1000 Hz), and an oscilloscope and a high-voltage probe were used to measure voltage. In addition, a 1 μF capacitor was connected to the reactor in series, and an amount of charge was measured by measuring a potential difference between both ends of the capacitor.

(65) The reactants were quantitatively analyzed by using an on-line gas chromatography (6500GC Young Lin Instrument Co., Korea). The thermal conductivity detector (TCD) and the flame ionization detector (FID) were used as detectors, and a Porapak-N, Molecular Sieve 13× column was used for the TCD and a methanizer was used for the FID. H.sub.2, N.sub.2, and CH.sub.4 were detected using the TCD, and CH.sub.4, C.sub.2H.sub.2, C.sub.2H.sub.4, C.sub.2H.sub.6, and C3 and C4 hydrocarbons were detected using the FID.

Preparation Example 3: Porous Silica KIT-6 Used in Paced Layer

(66) Mesoporous silica KIT-6 was prepared according to the following process. A copolymer pluronic p-123 was used as a structure inducer to form a three-dimensional structure of the mesoporous silica KIT-6. This copolymer forms a micelles in an aqueous solution and serves to form a mesoporous silica structure through interactions with silicon ions and self-assembly. 6 g of pluronic p-123 was added to an aqueous solution prepared by mixing 217.64 ml of distilled water and 11.16 g of a 37 wt % HCL solution and rapidly stirred at about 35° C. until the pluronic p-123 was completely dissolved. Subsequently, 6 g of n-butanol was added to the prepared mixed solution and further stirred for 2 hours while maintaining the temperature to form an Ia3d structure that is an intrinsic structure of KIT-6. Thereafter, 12.77 g of tetraethoxysilane (TEOS) was added dropwise to the mixed solution while stirring and then rapidly stirred at about 35° C. for 24 hours. After this process, a solution in which white silica deposits are formed was added to a polypropylene container and transferred to a hydrothermal synthesis device, and hydrothermal synthesis was performed at about 100° C. for one day without stirring. Then, a washing process was performed to remove solvents remaining in the reaction solution. After washing with distilled water for 30 minutes, washing with ethanol was performed three times to remove the pluronic p-123 and remaining impurities. Then, the resultant was dried in an oven at 110° C. for one day. The dried white silica powder was heated to 400° C. at a heating rate of 2° C./min and maintained for 3 hours, and then the powder was heated to 550° C. at a heating rate of 1° C./min and maintained for 5 hours to perform a calcination process. After the calcination was completed, the white silica powder was separated in terms of particle size of 100-150 μm by using sieves.

Example 2: Oxygen-free Methane Coupling Reaction with Packing Material

(67) The dielectric barrier discharge plasma reactor (FIG. 16) prepared in Preparation Example 2 was used. Plasma was generated in an alumina tube between the high-voltage electrode and the ground electrode wound by a length of 150 mm. The alumina tube was packed with the packing material prepared according to Preparation Example 2 except for a volume of 3.181 cm.sup.3 occupied by the high-voltage electrode. The oxygen-free methane coupling reaction was performed at room temperature and atmospheric pressure. NFC for CH.sub.4 and N.sub.2 was connected to the DBD plasma bed to inject CH.sub.4 and N.sub.2 in a ratio of 1:1, and the reaction was performed at a gas hourly space velocity (GHSV) of 2,500 mL.Math.g.sup.−1.Math.h.sup.−1. The experiment was performed while maintaining a voltage of 15 kV and a frequency of 1000 Hz. The oxygen-free methane coupling reaction was performed for 1000 minutes.

(68) Products were analyzed at 20 minutes after the reaction was initiated by an online gas chromatography and analysis was repeated at intervals of 40 minutes.

Example 3: Regeneration Using Low Temperature Plasma

(69) After Example 2 described above, a regeneration reaction of the deactivated packing material was performed by using the DBD plasma. The configuration of the plasma reactor was the same as that shown in FIG. 16, and the experiment was performed by packing the alumina tube with the packing material in the same volume as that of Example 2. In addition, the regeneration reaction was performed at room temperature and atmospheric pressure. An NFC was connected to the DBD plasma bed to inject Air into the DBD plasma bed, and the reaction was performed at a gas hourly space velocity (GHSV) of 1,250 mL.Math.g.sup.−1.Math.h.sup.−1. The experiment was performed while maintaining a voltage of 15 kV and a frequency of 1000 Hz. The regeneration reaction was performed for 720 minutes.

Comparative Example 1: Regeneration by Heat Treatment

(70) After the Example 2 described above, the deactivated packing material was heated to 700° C. in a muffle furnace at a heating rate of 5° C./m while flowing air at a flow rate of 50 sccm and maintained for 5 hours.

Example 4: TGA for Deactivated Packing Material and Regenerated Packing Material

(71) The packing materials collected after Example 2, Example 3, and Comparative Example 1 were subjected to TGA, and results are shown in FIG. 17, and weight losses, total amounts of coke, and removal rates thereof are summarized in Table 6. Comparing the total amounts of coke between Example 3 and Comparative Example 1, it was confirmed that a difference of the removal rate between Example 3 and Comparative Example 1 was only 3.23% p although the effect of Example 3 was not clearly identified. This indicates that most of coke was removed according to Comparative Example 1, and coke was removed according to Example 3 at the same level.

(72) TABLE-US-00006 TABLE 6 Total amount of coke (amount of coke per weight of packing Weight material Removal loss (%) (g-coke/g-SiO.sub.2)) rate.sup.a (%) Note Example 2 37.32 0.60 — Amount of coke generated by reaction Example 3 2.30 0.02 96.67 Amount of coke removed by plasma Comparative 0.06 0.0006 99.90 Amount of coke Example 1 removed by heat treatment .sup.aRemoved amount of coke/total amount of coke according to Example 1

Example 5: Confirmation of Removal of Coke by FT-IR

(73) After Preparation Example 3, Example 2, Example 3, and Comparative Example 1, collected packing materials were subjected to Fourier-transform infrared spectroscopy (FT-IR) analysis, and results are shown in FIG. 18. In the case of Example 2, three types of vibration modes were identified. .box-tangle-solidup. indicates a CH.sub.3 stretch vibration mode, .square-solid. indicates an asymmetrical C═C stretching mode, and • indicates an asymmetrical C—H bending mode of methylene groups in a long aliphatic chain. On the contrary, in Example 3 and Comparative Example 1, these vibration modes were not observed. Although peaks were observed nearby at 1630 cm.sup.−1 in the cases of Preparation Example 3, Example 3, and Comparative Example 1, it is an O—H bending mode which was not adsorbed and not detached therefrom even after being sufficiently dried.

(74) In addition, while Si—O—Si bonding was observed at 806 cm.sup.−1 in the cases of Preparation Example 3, Example 2, and Example 3, it was observed at 816 cm.sup.−1 in the case of Comparative Example 1.

Example 6: Analysis of Position of Coke Using TEM Image

(75) The packing materials collected after Example 2, Example 3, and Comparative Example 1 were subjected to TEM analysis, and results are shown in FIG. 19. A TEM image of the packing material collected after Example 2 was shown in FIG. 19-A, a TEM image of the packing material collected after Example 3 was shown in FIG. 19-B, and a TEM image of the packing material collected after Comparative Example 1 was shown in FIG. 19-C. Although coke was stacked on the surface of the packing material through FIG. 19-A, it may be confirmed that large amounts of coke were removed from the stack structures in the cases of FIGS. 19-B and 19-C.

(76) Regarding FIG. 19, energy-dispersive X-ray (EDX) mapping analysis results are shown in Table 7. Referring to Table 7, a mass percent of carbon was reduced by 19.12% p in the case of FIG. 19-B, and the mass percent of carbon was reduced by 31.01% p in the case of FIG. 19-C. This indicates that coke may be effectively removed from the deactivated packing material by using the low temperature DBD plasma and the muffle furnace.

(77) TABLE-US-00007 TABLE 7 Mass percent (%) C O Si FIG. 19-A (Example 2) 60.60 12.98 26.42 FIG. 19-B (Example 3) 41.48 25.73 32.79 FIG. 19-C 29.59 30.23 40.18 (Comparative Example 1)

Example 7: Structural Stability by XRD Characteristic Analysis

(78) The packing materials collected after Preparation Example 3, Example 2, Example 3, and Comparative Example 1 were subjected to small angle X-ray scattering (SAXS) analysis, and results are shown in FIG. 20. It was confirmed that SAXS graphs of the packing materials collected after Example 2 and Example 3 were identical to the SAXS graph of the fresh packing material. It was also confirmed that the oxygen-free methane coupling reaction using the DBD plasma and the regeneration reaction of the packing material did not affect structural stability of the packing material. On the contrary, it was confirmed that the SAXS graph of the packing material collected after Comparative Example 1 was different from the SAXS graph of the fresh packing material. Therefore, it was confirmed that the regeneration reaction using the muffle furnace affected the structural stability of the packing material.

Example 8: Structural Stability by BET Characteristic Analysis

(79) Table 8 shows BET surface areas, pore volumes, and pore sizes according to Preparation Example 3, Example 3, and Comparative Example 1. Referring to Table 8, it was difficult to maintain the structural stability in the case of the regeneration reaction using the muffle furnace when compared with the case of the regeneration reaction using the plasma.

(80) TABLE-US-00008 TABLE 8 BET Surface Pore Volume Pore Size Area (m.sup.2/g) (cm.sup.3/g) (nm) Preparation 774.92 1.12 6.40 Example 3 Example 3 749.04 0.91 6.05 Comparative 546.68 0.76 6.01 Example 1

Example 9: Reaction Performance of Packing Material for Plasma Regeneration

(81) Fresh KIT-6 and the packing material collected after Example 3 were subjected to reaction performance tests and results at TOS 60 min are shown in Table 9. Reaction conditions are the same as those in Example 2. When reaction performance of Fresh KIT-6, as the packing material before the reaction, was compared with that of Example 3, the methane conversion rate was slightly increased but the selectivity for C2 hydrocarbons was slightly decreased. It is considered that a small amount of unremoved coke remaining on the surface of KIT-6 as a coating layer affected the performance. However, the overall yield of light hydrocarbons was maintained at a similar level to that of Fresh KIT-6, and it is considered that plasma-regenerated KIT-6 may be used in a plasma bed for coupling of methane.

(82) TABLE-US-00009 TABLE 9 Selectivity (%) CH.sub.4 Conversion C3 C4 (%) Acetylene Ethylene Ethane hydrocarbons hydrocarbons Fresh KIT-6 41.80 15.53 6.39 21.64 10.74 11.54 Plasma-treated 48.85 10.15 3.81 16.13 10.57 9.08 KIT-6