Microwave-assisted direct synthesis of butene from high-selectivity methane

Methane was directly converted to butene liquid fuel by microwave-induced non-oxidative catalytic dehydrogenation under 0.1–0.2 MPa. The results show that, under microwave heating in a two-stage fixed-bed reactor, in which nickel powder and NiOx–MoOy/SiO2 are used as the catalyst, the methane–hydrogen mixture is used as the raw material, with no acetylene detected. The methane conversion is more than 73.2%, and the selectivity of methane to butene is 99.0%. Increasing the hydrogen/methane feed volume ratio increases methane conversion and selectivity. Gas chromatography/electron impact ionization/mass spectrometry chromatographic analysis showed that the liquid fuel produced by methane dehydrogenation oligomerization contained 89.44% of butene, and the rest was acetic acid, ethanol, butenol and butyric acid, and the content was 1.0–3.0 wt%.


Introduction
After ethylene and propylene, butene is an important intermediate in chemical production; butene is the raw material of butadiene, and butadiene is the main raw material for the production of synthetic rubber, such as styrene butadiene rubber, butadiene rubber, nitrile rubber and chloroprene rubber. With the development of styrene plastics, the use of styrene and butadiene copolymer can produce widely used resins such as acrylonitrilebutadiene-styrene copolymer, styrene-butadiene-styrene block copolymer, butadiene-styrene copolymer, and methyl methacrylate-butadiene-styrene terpolymer. At present, butene is generally obtained by petroleum cracking; however, with ever-decreasing and depleted oil reserves, potential alternatives will come from the dehydrogenation coupling of methane from natural gas or biogas. Because the main component of natural gas is methane, it is very stable, and the formation of ethylene, acetylene and high-hydrocarbon aromatics requires a reaction temperature of 1273 K or higher [1][2][3][4]. Although the high reaction temperature is favourable for high conversion rates, e.g. methane conversion of up to 50%, higher temperatures will a colourless oily liquid that is easy to volatilize. X-ray diffractometer (XRD) analysis was performed using a TTR-III type XRD manufactured by the Rigaku Corporation, having a Cu-Kα source (wavelength λ = 1.54178 Å) operated at a voltage of 40 kV and a current of 200 mA; the scanning step is 0.02°, the scanning speed is 8°min −1 , and the scanning range 2θ is 5-80°. The gas chromatograph model GC-2060 was produced by the Shandong Huifen Co., Ltd.; a six-way valve injection was used, the carrier gas was hydrogen, the reaction gas product was analysed by the thermal conductivity, thermal conductivity detector (TCD), determination, a column for the stainlesssteel column was used, the fixed phase was GDX-502, 80-100 mesh, the column outer diameter was 3 m × 3 mm (length × inner diameter), the carrier gas I was 0.08 MPa, the carrier gas II was 0.05 MPa, the column temperature was 60°C, the gasification (injection) temperature was 70°C, the thermal conductivity (detector) was 100°C, and the bridge current was 80 mA using the N2000 chromatographic workstation, with the area normalization method used to calculate the component content.
The resulting liquid product was analysed using a Thermo Scientific Company Q Exactive GC gas chromatography time-of-flight mass spectrometer (EI) instrument (Orbitrap). The column was base peak integration 60 m × 0.25 mm, 0.25 µm, the carrier gas was He, the control mode was 1 ml s −1 , and the splitting ratio was 100 : 1. The furnace temperature was initially at 30°C, increased at a rate of 2°C min −1 , and then kept at 210°C for 5 min (m/z: 30-400). The automatic gain control target was 1 × 10 6 with resolution of 120 000, the electron bombardment voltage was 70 eV, the sample injection volume was 1 μl, and the ion source temperature is 250°C; in the electro-spray positive ion mode EI + , the emission current was 50 mA, and interface temperature was 250°C. Scanning electron microscopy (SEM) was analysed using a Gemini SEM 500 Schottky field emission SEM. The LABRAM-HR800 Raman spectrometer of the French JY company was used to perform Raman spectroscopy using the following parameters: an excitation wavelength of the argon ion laser of 514.5 nm, a spot diameter of approximately 10 μm, a sample laser power of less than 1 mW, and a back-scattering configuration.

Experimental setup
The experimental device diagram of methane synthesis of butene is shown in figure 1. The device is composed of three parts: a feed gas flow control device, a microwave reactor and a reaction product condensed gas-liquid detection system. The microwave frequency is 2.45 GHz, and the power is 700 W. The microwave reactor was composed of a cylindrical quartz glass tube having a length of 200 mm and an inner diameter of 10 mm. The first stage of the reactor was filled with pure nickel powder catalyst and connected to the second stage of the reactor filled with NiO x -MoO y /SiO 2 composite catalyst.

Preparation of the nickel-molybdenum/SiO 2 catalyst (impregnation method)
First, 4.460 g (0.0153 mol) of nickel nitrate hexahydrate and 1.656 g (Mo: 0.0094 mol) of ammonium heptamolybdatetetrahydrate were dissolved in 8 ml of deionized water to obtain a metal salt aqueous solution and then mixed and stirred for 30 min at room temperature. The measured 8.212 g of silica (SiO 2 ) was added to the aqueous solution of metal salt, immersed at 80°C for 16 h, dehydrated at 90°C for 8 h, and then dried at 120°C for 8 h; subsequently, it was calcined at 500°C for 8 h in air atmosphere to remove water, oxygen and nitrogen dioxide. The nickel oxide and molybdenum oxide and its silica carrier catalyst NiO-NiMoO 4 (0.6277: 1.0 mol)/SiO 2 were obtained, and then the catalyst was reduced at 700°C in a hydrogen atmosphere for 1 h. With the diameter of the catalyst particles of approximately 200 nm and the hydrogen feed flow rate of 60 ml min −1 , the NiO x -MoO y(x=0-1,y=0-3) /SiO 2 catalyst was obtained, with the nickel and molybdenum content of 9(wt%) each.

Methane microwave catalytic production of the butene test
In the microwave reactor, the methane hydrogen mixed gas (CH 4 :H 2 = 1 : 5 v/v) was passed through a Teflon tube. First, the gas was flowed through the first stage of the reactor (quartz tube) filled with 1.000 g of pure Ni powder catalyst; the reaction intermediate product gas continues to flow through the second stage of the reactor (quartz tube), and then 0.500 g of the NiO x -MoO y /SiO 2 composite catalyst powder was added to the tube, followed by microwave heating at 700 W for 30 min, with the methane-hydrogen mixed gas inlet pressure of 0.1-0.2 MPa, the inlet flow rate of methane of 10 ml mi −1 and that of hydrogen of 50 ml min −1 . When the product gas flows through the second stage of the reactor (quartz tube), the reaction gas is collected by a six-way valve and then analysed online by a GC/TCD. Next, the product mixture is passed into a cold trap (−30°C) containing frozen ethanol solution; after cooling, 1.01 g of colourless liquid product was collected and analysed using GC/EI/MS, and then the non-condensed gas was vented.

Results and discussion
3.1. X-ray diffraction determination of the nickel powder The crystal structure and phase analysis of the Ni powder in the first stage of the reactor was characterized by powder XRD. The XRD patterns of the Ni powder shown in figure 2 were indexed to monoclinic Ni according to the JCPDS database no. 04-0850 [21]. The average crystallite size of Ni sample was calculated by using the Debye-Scherrer formula given in equation (3.1): where d is the crystallite size, k is 0.89 (CuK), λ is the wavelength of the X-rays (λ = 1.54178 Å), θ is the Bragg diffraction angle, and β is the full width at half maximum (FWHM). The average crystallite size d before and after reaction calculated from the diffraction peaks was found to be approximately 8.013 and 9.158 nm, respectively. After the reaction, the average grain size of the nickel powder increased by 1.145 nm. The diffraction peaks of nickel before and after the reaction are at 2θ = 44.48°, 51.78°and 76.46°, which correspond to the (111), (200) and (220) planes of the nickel, respectively. The diffraction peak 2θ (°) before and after the reaction did not change substantially, but the peak intensity after the reaction increased slightly, and the baseline shift of the FWHM (half full width) is small, indicating that the grain agglomeration diameter is increased.

Determination of the nickel powder by scanning electron microscopy
Figure 3a-d are SEM images of the nickel powder in the first stage before and after the reaction, respectively. As shown in the figure, when the nickel particles are of thickness of 2 μm before the reaction, nickel powder particles were in a linear aggregation state; compared with the spent nickel from the used granular aggregates are found to have a linear appearance. The results show that the linear morphology of the nickel powder has no change, but the agglomeration between the nickel particles is obviously enhanced, the linear nickel powder has a polycrystalline structure and nanometre thorns of single-crystal structure were found.    produces poor crystal form. The characteristic peaks appearing at 2θ = 25.5°, 28.9°, 32.7°, 43.9°and 47.5°a re attributed to Ni-Mo/SiO 2 [24]. Compared with the fresh catalyst, the characteristic peaks of the spent catalyst (sample 1) at 2θ = 76.0°are found to be attributed to the reduced Ni, and its peak intensity decreased significantly, showing that the reduced Ni metal in the NiO x -MoO y /SiO 2 catalyst is involved in the reaction, and the active sites are partially covered. The peak intensity of the spent catalyst at 2θ = 26.7°is significantly higher than that of the fresh catalyst; this observation is attributed to the surface of NiMoO 4 producing a poor crystal form, indicating that the reaction on the surface of NiMoO 4 crystal form have an impact. By contrast, the characteristic peaks and peak intensity of NiO and MoO 3 have almost no change, indicating that they have corresponding catalytic stability in ethylene dimerization. Powder XRD patterns of the as-prepared NiMoO 4 /SiO 2 NPs shown in figure 4 were indexed to NiMoO 4 according to the JCPDS database no. 33-0948.
3.4. Determination of the NiO x -MoO y /SiO 2 catalyst by scanning electron microscopy Figure 5a shows the SEM images of the NiO x -MoO y /SiO 2 catalyst in the second stage of the reactor; it can be seen from the figure that, at 200 nm, the fresh catalyst NiO x -MoO y /SiO 2 is elliptical microspheres with macroscopic structure, and the edge is partially transparent, with a typical silica microsphere structure [25], indicating that the formation of the catalyst did not change the macroscopic structure of silicon oxide. Figure 5b shows an SEM image of NiO x -MoO y /SiO 2 after the second stage of the reaction. Before the reaction, the spherical particles were less than 200 nm, the particle size of the catalyst decreased after the reaction, and the appearance showed obvious agglomeration.   [26,27] and is in agreement with the XRD results. After the reaction, the peak intensity of the Mo=O band at 959, 910 and 706 cm −1 decreased significantly, indicating that Mo=O was involved in the ethylene oligomerization reaction. A new band appeared at 461 cm −1 , and the peak intensity is strong, which could be attributed to the reduced surface molybdenum oxide species [28]. Therefore, the Raman analysis results show that the NiO x -MoO y(x=0-1,y=0-3) /SiO 2 catalyst has the characteristic peaks of α-NiMoO 4. After the reaction, the increase of the characteristic peak intensities of 959, 910 and 706 cm −1 of the α-nickel molybdate shows that the large amount of α-nickel molybdate formed leads to the inhibition of the ethylene oligomerization reaction. The intensity of the characteristic reduction peak based on 461 cm −1 is related to the surface reduction property of molybdenum oxide after the reaction decreased significantly. This result shows that the formation of reduced molybdenum is beneficial to the ethylene oligomerization and the formation of butene.

Gas reactants and liquid products
In the first stage of the reactor, under the action of microwave heating and pure nickel powder, methane dehydrogenation coupling to produce ethylene was performed. The hydrogen atmosphere inhibited the further dehydrogenation of ethylene because the reaction gas effluent did not detect acetylene. In the second stage of the reactor, under the action of microwave heating and nickel-molybdenum/SiO 2 catalyst, the intermediate ethylene was coupled to produce butene, which contained 89.44%, and the rest are ethanol, acetic acid, butenol and butyric acid, with contents of 1.0-3.0 wt%. According to GC/EI/MS analysis of the liquid products, in addition to high levels of butene, there are trace amounts of oxygen mixed into the reaction system. Thus, butene is possibly partially oxidized to ethanol, acetic acid, butenol and butyric acid. Table 1 shows the feed composition and the total flow rate of mixed gas. As the feed gas CH 4 /H 2 ratio increases from 1 : 5 v/v to 1 : 4 v/v, that is, the methane flow rate does not change, the hydrogen flow rate reduced, and the methane conversion decreased slowly from 73.2 to 71.0%. When the total gas flow rate is 60 ml min −1 , the volume ratio of methane to hydrogen is 1 : 5, the selectivity of butene is 99.0%, and the conversion rate of methane is 73.2%; this may be caused by sufficient hydrogen partial pressure, resulting in the high reaction conversion rate. However, with the hydrogen flow decreases,    such as the volume ratio of methane to hydrogen reduced from 1 : 4 to 1 : 3, because of the reduced partial pressure of hydrogen, a slight decrease in the conversion of methane from 71.0 to 67.2% occurs, implying that it is necessary to maintain a certain partial pressure of hydrogen. This outcome may occur because the methane conversion reaction rate is not only related to the velocity constant, the methane partial pressure and the percentage of active sites but also to the partial pressure of the hydrogen. We believe that the amount of H 2 also directly determines the amount of regenerated H-Ni-O-H catalyst in situ.
The obtainable yield depends on the catalytic cycle. In fact, hydrogen is not only a reaction raw material but also an inhibitor of coke and acetylene and plays the role of carrier gas, the latter promoting the flow of reactants through the catalyst layer; as a result, the amount of hydrogen should be excessive. The results also show that, when the methane content is low and the hydrogen is in excess, the conversion of methane and the selectivity of butene depend only on the activity of nickel and NiO x -MoO y /SiO 2 composite catalysts and are independent of the hydrogen partial pressure. Figure 7 shows the gas chromatogram of the total reaction gas product before condensation. As seen from the figure, the methane content was 26.8% at the retention time of 0.5 min, and when the retention time was 10.5 min, the butene content was 73.2%; see the GC-MS analysis. Therefore, the methane conversion was greater than 73.2%, and no further larger impurity peaks were observed from the gas-phase reaction effluent.

Determination of liquid products by gas chromatography/electron impact ionization/mass spectrometry
The reaction process collects the condensed liquid; its GC/MS mass spectrum of total ion current diagram is shown in figure 8.      Table 2 shows the chromatographic ion outflow peak time; the content of the peak was greater than 1.0 wt% for retention times of 4.95, 9.2, 17.28 and 28.4 min, and the remaining 1.0% or less had a total of 4.90%. Figure 9a shows the mass-charge ratio; its m/z interval was 30-90, and the relative abundance of large ion molecules is as follows: m/z 55.0, m/z 45, m/z 60.02 and m/z 73. 03. The former is butyric acid debris, with chemical formula of acetic acid (C 2 H 4 O 2 ); the latter is butyrate (removal of methyl, (CH 3 )    50  45  40  35  30  25  20  15  10  5  0  30  35  40  45  50  55  60  65  70  75  80  85  90   100  95  90  85  80  75  70  65  60  55   relative abundance   50  45  40  35  30  25  20  15  10  5  0  30  35  40  45  50  55  60  65  70  75  80  85 Figure 9c shows the mass-to-charge ratio, and its interval is 30 Figure 9e shows the mass-to-charge ratio, and its m/z interval is 30-55; the relative abundance of molecular ion of m/z 43.02 and m/z 45.03 correspond to ethanol fragments.

Methane non-oxidative dehydrogenation coupling into butene mechanism
The non-oxidative dehydrogenation of methane to produce butene is extremely complex. The mechanism involves the nickel powder in the first stage of the reaction and the NiO x -MoO y /SiO 2 composite in the second stage of the reaction. Under the combined action of nickel metal and NiO x -MoO y /SiO 2 catalyst and microwave heating, unlike conventional microwave plasma methane conversion [29][30][31][32][33], the possible mechanism of methane non-oxidative dehydrogenation oligomerization in this reaction is shown in schemes 1-3.

Path (A)
In the first stage of the reactor, the nickel powder absorbs methane by polarization of the C-H bond. Under the microwave action, methane is transformed into carbene (H 2 C:) and then ethylene (CH 2 =CH 2 ). The likely reactions are absorption of methane by Ni metal, according to equation (3.2):