Syngas is a raw material in the production of numerous chemical products, such as synthetic oils and olefins. Catalytic partial oxidation of methane (POM) is the main method used to produce syngas [1, 2, 3, 4, 5, 6, 7]. Designing and developing cheap catalysts with good catalytic performance and high selectivity for POM are important for the production of syngas from CH4.
Supported noble-metal (Pt [8], Pd [9], and Rh [10]), Ni- based [11, 12], and Co-based [5] catalysts are usually used in POM. The support is an important component of the catalyst and significantly affects the catalytic performance. The supports generally used for POM catalysts include Al2O3 [13, 14], SiO2 [15, 16], CeO2 [17], ZrO2 [18], TiO2 [19], and CeZrO2 solid solution [20, 21]. Among these, SiO2 shows good stability because it does not easily react with the other components at high temperature. Moreover, dispersion of the supported active component over SiO2 is improved as a result of its large surface area. For example, Li et al. [22] reported that a Pd/SiO2 catalyst exhibited good activity in POM because of high dispersion of Pd. After reaction for 7 h at 700 °C, the Pd particle size was still 4-5 nm. Xia et al. [23] reported a spherical, single-channel Ni/SiO2 catalyst with good Ni dispersion, synthesized by a sol-gel process. This Ni/SiO2 catalyst showed high activity, selectivity, and long life in POM. Another porous, core-shell Ni@SiO2 catalyst, with a narrow particle size distribution, was synthesized by Li et al. [24]. Activity tests showed that the catalytic performance of this core-shell Ni@SiO2 catalyst is closely related to the size of the Ni core, the porosity of the SiO2 shell, and core-shell interactions.
The catalyst promoter is also crucial in POM because the reduction of the active component is greatly influenced by the interactions between the promoter and the active component. Moreover, such interactions also affect the stability of the catalyst. For example, Li et al. [25] found that Ce doping in Ni/SiO2 effectively inhibited the growth of Ni crystallites and improved Ni dispersion. Moreover, CeO2 easily reacts with other oxides to produce composite oxides, such as CeO2-TiO2, CeO2-ZrO2, CeO2-Al2O3, and so on [26, 27]. The produced composite oxides always have large surface areas and are rich in oxygen vacancies, enabling better promotion by Ce in POM.
In this work, a series of CeO2-SiO2 composite oxides with large surface areas were synthesized using a sol-gel process. The active component, Ni, was loaded on the as-synthesized composite oxides, producing supported Ni catalysts. The effects of the Ce/Si molar ratio in the support, Ni content, calcination temperature, and reaction time on the catalytic performance of Ni/CeO2-SiO2 in POM were investigated.
All the chemicals used were analytically pure and obtained from the Sinopharm Chemical Reagent Co., Ltd. The CeO2-SiO2 composite oxides were synthesized using a sol-gel process. Stoichiometric hexahydrated cerium(ΙΙΙ) nitrate (Ce(NO3)3· 6H2O) was dissolved in ethanol, and the solution pH was maintained at 2-3 using nitric acid. Stoichiometric tetraethyl orthosilicate (TEOS, C8H20O4Si) was then added dropwise to the solution under stirring, producing a transparent gel. The produced gel was aged naturally overnight, dried at 110 °C, ground, and calcined at 700 °C for 4 h in air, giving CeO2-SiO2 composite oxide supports. The obtained supports were impregnated for 12 h with a stoichiometric nickel nitrate (Ni(NO3)2·6H2O) solution, dried, and calcined at different temperatures for 4 h in air. The obtained catalysts were denoted by wNi/xCeO2-(1−x)SiO2-T, where w represents the Ni content (mass fraction, 5%-20%), x refers to the Ce molar fraction (x = 0-1) in the support, and T stands for the calcination temperature (600-900 °C) after impregnation with Ni.
The Brunauer-Emmett-Teller (BET) surface areas of the samples were obtained from N2 adsorption/desorption isotherms determined at liquid-N2 temperature (−196 °C) using an automatic analyzer (V-Sorb 2800P, Beijing Jinaipu Company of Science and Technology). The samples were degassed at 200 °C for 1 h under vacuum before the measurements. X-ray diffraction (XRD) measurements were carried out using a Holland X’Pert Pro MPD X-ray diffractometer (Philips, the Netherlands) equipped with a monochromatized Co Kα radiation source (λ = 0.1790955 nm). The accelerating voltage and the applied current were 40 kV and 120 mA, respectively. Scanning electron microscopy (SEM) was performed using a Nova Nano SEM230 (20 kV) scanning electron microscope (FEI company, USA). Ultraviolet-visible (UV-Vis) diffuse reflectance spectra (DRS) were obtained using a UV-Vis spectrophotometer (UV-2550, Shimadzu). The absorption spectra were referenced to BaSO4. Temperature-programmed reduction by H2 (H2-TPR) was carried out using a chemical adsorption instrument (TP5080, Tianjin Xianquan Industry and Trade Development Co., Ltd.). Samples (0.05 g) were placed in a quartz reactor and treated in situ with a flow of dry N2 (30 mL/min) at 300 °C for 0.5 h. Before a run, the baseline was stabilized in the gas flow. The reactor was heated in a temperature-programmed furnace from room temperature to 800 °C (10 °C/min). A 10% (volume fraction) H2 in N2 mixture was used as the reducing gas. The H2 consumption as a function of the reduction temperature was continuously monitored using a thermal conductivity detector (TCD) cell and recorded. Temperature-programmed desorption of NH3 (NH3-TPD) was carried out using the same chemical adsorption instrument as that used for H2-TPR. Before the tests, the samples (0.1 g) were degassed in situ under a N2 stream at 300 °C for 0.5 h and then cooled to room temperature. The samples were then saturated with an NH3 flow (10 mL/min) at room temperature. The physically adsorbed NH3 was purged under a N2 stream until the TCD baseline was stabilized and then TPD was performed from room temperature to 800 °C at a heating rate of 10 °C/min. Thermogravimetric analysis (TGA) was performed using a TGA Q50 thermogravimetric analyzer (TA Company, USA) to determine the deposited carbon content. After reaction, about 10 mg of each catalyst was degassed in a N2 stream (40 mL/min) and then tested from room temperature to 750 °C at a heating rate of 10 °C/min in an air stream (60 mL/min).
Catalytic performance tests were carried out using a WFSM-3060 reaction apparatus (Tianjin Xianqian co, China) equipped with a continuous-flow fixed-bed quartz-tube reactor of inner diameter 8 mm, operated under atmospheric pressure. Before the reaction, about 0.15 g of each catalyst was placed in the quartz reactor and reduced at 750 °C for 1 h in pure H2 stream (20 mL/min), followed by purging with N2 stream (20 mL/min) for 20 min. The reaction mixture CH4/O2 = 20/10 (30 mL/min) was fed to the reactor at a total gas hourly space velocity of 12000 mL/(h·g). After 30 min the reaction began, the composition of the gaseous products was analyzed using an online GC-2060 chromatograph (Tengzhou Lunan Analysis Instrument Co, LTD, China) equipped with a TCD and a TDX-101 packed column. The CH4 conversion, CO selectivity, H2 selectivity, and H2/CO molar ratio were calculated using the following formula, where Ai,out and Fi,out represent the chromatographic peak area and relative molar correction factor of gaseous component i, respectively. Here,
The BET surface area was calculated using the BET equation and the physical adsorption/desorption isotherms of N2; the surface areas of all the supports and catalysts are shown in Table 1. The data in Table 1 show that the surface area is closely related to the Ce molar fraction. The surface areas of the xCeO2-(1−x)SiO2 composite oxide supports are larger than that of CeO2 alone. As the Ce molar fraction increases (x value from 0 to 1), the surface area decreases (from 470 to 20 m2/g). The Ce molar fraction is the key factor in obtaining CeO2-SiO2 composite supports with large surface area. Increasing the calcination temperature (from 600 to 900 °C) decreases the BET surface area of 0.50CeO2-0.50SiO2 (from 196 to 96 m2/g). Increasing the Ni loading only leads to a slight decrease in the surface area (10%Ni/0.50CeO2-0.50SiO2-700, 100 m2/g; 20%Ni/ 0.50CeO2- 0.50SiO2- 700, 103 m2/g). The Ni is highly dispersed because of the large surface area of the 0.50CeO2-0.50SiO2-700 support (149 m2/g). The loaded Ni could enter some of the pores, causing minor decreases in the surface area [25, 28, 29, 30]. It is notable that even after calcination at 900 °C for 4 h, the surface area of 10%Ni/0.50CeO2- 0.50SiO2- 900 still maintains 59 m2/g, which indicates that the high thermal stability and strong anti-sintering capacity of the catalyst.
The phase structure of the as-prepared supports and catalysts was analyzed using XRD; the XRD patterns are shown in Fig. 1. Figure 1(a) shows that no clear diffraction peak from SiO2 is observed in the single-component SiO2-700 support, indicating that SiO2 is amorphous. Strong diffraction peaks from cubic fluorite CeO2 (Ref. code: 00-034-0394), at around 2θ = 33.3°, 38.7°, 55.8°, and 66.5°, are observed for the xCeO2-(1−x)SiO2- 700 (x = 0.25, 0.50, 0.75, 1) supports, and the peak intensity increases with increasing Ce molar fraction. Figure 1(b) shows that the diffraction peaks of cubic phase NiO (Ref. code: 01-078-0423), at around 2θ = 43.7°, 50.8°, and 74.7°, are detected for 10%Ni/0.50CeO2-0.50SiO2-700, and, as shown in Fig. 1(c), the peak intensity increases with increasing Ni content (from 10% to 20%). The XRD patterns of the 0.50CeO2- 0.50SiO2 and 10%Ni/0.50CeO2-0.50SiO2 catalysts calcined at different temperature are shown in Fig. 1(d) and 1(e), respectively. The intensity of the CeO2 and NiO diffraction peaks increase with increasing calcination temperature.
The Scherrer equation, D = 0.89λ/βcosθ, was used to estimate the average crystallite size of the samples, where β is the width in radians of the XRD peak at half-height for the plane peak, λ is the X-ray wavelength in nanometers (λ = 0.1790955 nm), θ is the angle between the incident and diffracted beams in degrees, and D is the average crystallite size of the powder sample in nanometers. The strongest diffraction peaks of CeO2 (2θ = 33.3°) and NiO (2θ = 50.8°) were used in the calculation. The obtained results are summarized in Table 2. The CeO2 and NiO particles are nanoscale. According to the literature [26, 31], a mixture of CeO2 and SiO2 first forms a Ce9.33(SiO4)6O2 intermediate phase. During calcination, Ce9.33(SiO4)6O2 decomposes to nanoscale CeO2 particles, and the produced CeO2 is highly dispersed on the SiO2 matrix. The average crystallite size of CeO2 in catalysts is slightly larger than that in supports; this may be caused by the second calcination. High-temperature (900 °C) calcination causes an increase in the size of the CeO2 and NiO crystallites, suggesting that aggregation or sintering of these particles occurs, and this could also be a major factor in the decrease in the BET surface area. The average crystallite size of CeO2 in CeO2-700 and 10%Ni/CeO2-700 is 0.5414 and 0.5416 nm, respectively, which are slightly larger than that of CeO2 in the CeO2-SiO2 composite. This could be attributed to lattice constriction of CeO2 when Si4+ enters the CeO2 lattice and forms the CeO2-SiO2 composite oxide, because the Si4+ radius (r = 0.042 nm) is smaller than the Ce4+ radius (r = 0.087 nm) [26].
Scheme 1 shows a proposed mechanism for the formation of the CeO2-SiO2 composite oxide. The formation process could include several steps such as hydrolysis of TEOS, condensation, calcination, and dehydration [32, 33]. First, under acidic conditions, hydrolysis of TEOS takes place, producing orthosilicic acid. The orthosilicic acid is transformed into a silicate polymer by condensation and dehydration. Ce3+ could be adsorbed by the hydroxy (-OH) groups in the silicate polymer. Drying and calcination produce a CeO2-SiO2 composite oxide with good dispersion and a large BET surface area.
The morphology of typical supports and catalysts was analyzed using SEM; the SEM images are shown in Fig. 2. As shown in Fig. 2(a), the SiO2-700 support consists of a large number of aggregated particles. It can be seen from Fig. 2(b) that there are no obvious morphological differences between SiO2-700 and the 0.50CeO2-0.50SiO2-700 composite support. Figure 2(c) shows that after Ni loading (w = 10%), small particles appear on the surface of 0.50CeO2-0.50SiO2-700. These are NiO particles that were reduced to metallic Ni0 by H2 before POM.
DRS was used to identify Ni species on the catalyst surface. Fig. 3 shows the UV-Vis DR spectra of typical catalysts. The UV-Vis DR spectrum of pure CeO2 is also shown. It can be seen from this figure that CeO2 has an obvious absorption peak at around 343 nm, and NiO absorption peaks are observed in the region from 297 to 350 nm. The characteristic absorption peaks of NiO are attributed to charge transfer by octahedral Ni2+ among NiO lattices [25, 34, 35]. A comparison of Fig. 3(a) with Fig. 3(b)-(e) shows that the peaks in the UV-Vis DR spectra of the Ni species in 10%Ni/xCeO2-(1−x)SiO2-700 (x = 0.25, 0.50, 0.75, 1) shift to slightly longer wavelengths, indicating that less energy is needed for charge transfer, as a result of interactions between Ni species and the xCeO2-(1−x)SiO2-700 composite support [25, 30]. These interactions also affect the reduction performance of the catalysts.
The H2-TPR profiles of the catalysts are shown in Fig. 4. The reduction of the CeO2-SiO2 composite support was first tested to determine whether its reduction peaks overlap with the reduction peaks of NiO. Figure 4(a) shows that 0.50CeO2- 0.50SiO2- 700 has a weak H2 consumption peak at around 490 °C. This peak results from the reaction between H2 and surface lattice oxygen on CeO2 or adsorbed oxygen [28, 29, 30, 31, 36]. The reduction peak at 780 °C is ascribed to reduction of bulk CeO2 containing large crystallites [36, 37, 38, 39, 40, 41] because SiO2 cannot be reduced below 800 °C [36]. The NiO reduction peaks appear at 325-580 °C, and they become stronger with increasing Ni content.
The effects of the Ce/Si molar ratio on the reduction performance of the catalysts are shown in Fig. 4(b). Two reduction peaks are observed at 365-670 °C for 10%Ni/SiO2-700 and at 250-470 °C for 10%Ni/CeO2-700. The reduction temperature of NiO supported on the xCeO2-(1−x)SiO2-700 composite support (x = 0.25, 0.50, 0.75) is lower than that for Ni/SiO2-700, suggesting that Ni is more easily reduced over xCeO2- (1−x)SiO2-700. The interactions between CeO2 and NiO could benefit NiO reduction. Moreover, oxygen vacancies in CeO2-SiO2 could increase the lattice oxygen mobility, making the reduction of Ni species easier [20].
Figure 4(c) shows the influence of the calcination temperature on the reduction performance of 10%Ni/0.50CeO2- 0.50SiO2. An increase in the calcination temperature causes the reduction peak of NiO to shift toward higher temperature. This shift indicates that reduction of NiO is difficult; this could be a result of the strong interactions between CeO2 and SiO2 formed during high-temperature calcination [29, 42]. Increases in the size of the CeO2 and NiO crystallites (see Table 2) could also help to make NiO reduction difficult.
The surface acidity of typical supports and catalysts was tested using NH3-TPD; the TPD profiles are shown in Fig. 5. For SiO2-700, two NH3 desorption peaks are observed at around 90 °C and 450 °C, corresponding to weak and strong acid sites, respectively. The TPD profile of 10%Ni/SiO2-700 is consistent with that of SiO2-700. Compared with SiO2-700, the 0.50CeO2- 0.50SiO2-700 has a smaller weak acid desorption peak, and the strong acid desorption peak almost disappears, indicating that the acidity is weakened by the formation of the composite support; this could be caused by the alkalinity of Ce. Ni loading (10%) on 0.50CeO2-0.50SiO2-700 causes no obvious difference in the TPD profile. In POM, carbon species and coke are easily deposited on strong acid centers [2]. The weak acidity of the 10%Ni/0.50CeO2-0.50SiO2-700 catalyst could contribute to its low carbon deposition in POM.
Typical catalysts were analyzed after POM using TGA; the TG profiles are shown in Fig. 6. The mass losses below 130 °C, in the region 130-350 °C, and above 600 °C are produced by desorption of adsorbed water and amorphous carbon, combustion of filamentous carbon, and graphitic carbon combustion, respectively [43, 44]. However, a mass increase is also observed in the region 350-570 °C; this is caused by oxidation of Ni0. The smaller mass losses below 130 °C for 10%Ni/0.50CeO2- 0.50SiO2-700 and 20%Ni/0.50CeO2-0.50SiO2-700 compared with that for 10%Ni/SiO2-700 imply that smaller amounts of water and amorphous carbon are adsorbed on these catalysts, and their more obvious mass increases indicate that there are larger numbers of active Ni0 centers in these two catalysts after reaction. Above 450 °C, no obvious mass loss is observed for 10%Ni/0.50CeO2-0.50SiO2-700, whereas above 440 °C, a mass loss of about 7.1% occurs for 20%Ni/0.50CeO2- 0.50SiO2-700, suggesting that there is a large amount of deposited carbon on 20%Ni/0.50CeO2-0.50SiO2-700. Previous research [45, 46, 47] showed that the average coke formation rate is closely related to the crystallite size and dispersion of Ni. Smaller Ni crystallites and higher dispersion of Ni could significantly decrease coke formation. Moreover, Fig. 6 also shows that carbon deposition increased with increasing Ni content. The increase in Ni content increases the Ni crystallite size, and large Ni crystallites easily become active centers for carbon deposition reactions.
Table 3 shows the CH4 conversion and CO and H2 selectivity for 0.50CeO2-0.50SiO2-700-supported catalyst with different Ni contents. From the data in Table 3, it can be seen that the 0.50CeO2-0.50SiO2-700 support gives low activity. The CH4 conversion is 23.7%, and the CO and H2 selectivity is only 8.6% and 4.5%, respectively, indicating that the major reaction on this support is CH4 combustion. A loading of 5% Ni greatly improves the activity and selectivity: the CH4 conversion is 71.6%, and the CO and H2 selectivity increases to 79.3% and 89.0%, respectively. Loading of 10% Ni gives the highest CH4 conversion (84.3%), CO selectivity (87.4%), and H2 selectivity (93.3%). With further increases in the Ni content (20% Ni), the activity and selectivity decrease again. A very low Ni content (5%) does not provide enough active Ni0 centers for POM. However, a very high Ni content (20%) leads to large Ni crystallites (see Table 2) and higher carbon deposition (in Fig. 6), which impair the catalytic performance.
The effects of Ce/Si molar ratio on the catalytic performance in POM are shown in Table 4. 10%Ni/SiO2-700 and 10%Ni/ CeO2-700 show low catalytic activity, but 10%Ni/ 0.50CeO2- 0.50SiO2-700, which has the same Ni content (10%), shows the highest CH4 conversion and CO and H2 selectivity. The BET surface area, H2-TPR, and NH3-TPD results show that 10%Ni/0.50CeO2-0.50SiO2-700 has a large surface area and weak acidity, and the loaded Ni is easily reduced to Ni0; these are the main reasons for its high catalytic performance.
Table 5 shows the catalytic performance over the catalysts calcined at different temperatures. The highest CH4 conversion and product selectivity were obtained over 10%Ni/ 0.50CeO2- 0.50SiO2-700. A very high calcination temperature results in an obvious decrease in the BET surface area and growth of CeO2 and NiO crystallites. Moreover, after very-high- temperature calcination, reduction of NiO becomes difficult, and the number of Ni0 active centers decreases. The catalytic performance of 10%Ni/0.50CeO2-0.50SiO2-900 was therefore poor.
The changes in CH4 conversion, and CO and H2 selectivity with increasing reaction time over 10%Ni/0.50CeO2-0.50SiO2- 700 and 10%Ni/SiO2-700 are shown in Fig. 7. For 10%Ni/SiO2- 700, when the reaction time is prolonged to 390 min, the CH4 conversion and CO and H2 selectivity decrease sharply, suggesting poor stability. However, for 10%Ni/0.50CeO2- 0.50SiO2- 700, there is almost no change in the CH4 conversion and CO and H2 selectivity during POM for 900 min, implying good stability.
A series of CeO2-SiO2 composite oxide supports with large surface area were synthesized using a sol-gel process. After Ni loading, the produced Ni/CeO2-SiO2 catalysts exhibited high catalytic performance in POM. The catalytic performance of Ni/CeO2-SiO2 was closely related to the Ni content, Ce/Si molar ratio, and calcination temperature. The experimental results showed that the highest CH4 conversion and product selectivity and good stability were obtained over Ni/CeO2-SiO2 calcined at 700 °C, with a Ce/Si molar ratio of 1:1 and a Ni content of 10%. The large BET surface area, weak acidity, ease of NiO reduction, and low carbon deposition are the main reasons for its high catalytic performance.
合成气(CO+H2)被认为是制备合成油和烯烃等众多化工产品的枢纽原料. 而催化甲烷部分氧化(POM)是生产合成气的主要途径, 具有重要的经济和理论意义[1−7]. 目前, 研发低成本、高催化活性和高产物选择性的催化剂仍是甲烷部分氧化制合成气的关键.
POM反应通常使用的是负载型贵金属(Pt[8], Pd[9], Rh[10])、Ni基[11, 12]或Co基[5]催化剂. 载体是催化剂的重要组成部分, 并影响其催化性能. 通常使用的载体有Al2O3[13, 14], SiO2[15, 16], CeO2[17], ZrO2[18], TiO2[19]和CeZrO2固溶体[20, 21]等. 其中SiO2在高温条件下不易和其它组分发生固相反应, 具有较好的稳定性, 同时SiO2的比表面积较大, 能够提高活性组分的分散性能. 如Li等[22]报道了Pd/SiO2催化剂在POM反应中具有优异的活性, 原因是SiO2负载的Pd具有良好的分散性和热稳定性, 经700 °C反应7 h后的Pd粒子的粒径仅为4-5 nm. Xia等[23]采用溶胶-凝胶法制备了易使Ni颗粒分散的单通道、球状Ni/SiO2催化剂, 该催化剂具有较高的POM反应活性和较长的使用寿命. Li等[24]则合成了具有多孔核-壳结构、粒径分布窄的Ni@SiO2催化剂, 其催化性能与Ni核的尺寸、SiO2壳的孔隙率及核-壳间相互作用密切相关. 另一方面, 助剂在POM反应中的作用也极为关键, 因为助剂与活性组分的相互作用可以改变活性组分的还原性能, 并影响催化剂的稳定性. Li等[25]考察了助剂对Ni/SiO2催化剂在CH4和CO2重整反应中的影响. 结果表明Ce的掺杂效果最佳, Ce能够很好地抑制Ni晶粒的高温长大并提高Ni的分散性. 另外, CeO2易与其它氧化物结合形成复合氧化物, 如CeO2-TiO2, CeO2-ZrO2和CeO2-Al2O3等[26, 27]. 这些复合氧化物通常具有大比表面积, 易产生丰富的缺陷氧中心, 在POM反应中常表现出更佳的载体作用.
本论文采用溶胶-凝胶法合成了系列大比表面积的CeO2-SiO2复合氧化物, 然后负载活性组分Ni, 制得Ni/CeO2-SiO2催化剂. 考察了载体中Ce/Si摩尔比、Ni的质量分数、焙烧温度及反应时间等对催化剂在甲烷部分氧化制合成气反应中催化性能的影响.
所使用的试剂均来自上海国药集团, 均为分析纯. 首先, 采用溶胶-凝胶法制备CeO2-SiO2复合氧化物载体. 按计量比称取硝酸亚铈(Ce(NO3)3·6H2O)溶解于适量的乙醇溶液中, 并滴加一定浓度的HNO3, 调节溶液的pH至2-3. 然后以1滴/s的速率滴加正硅酸四乙酯(C8H20O4Si, TEOS), 同时搅拌形成透明凝胶, 待凝胶自然老化过夜后, 于110 °C干燥, 研磨, 700 °C焙烧4 h, 即得到复合氧化物载体CeO2-SiO2. 将载体等体积浸渍于计量的Ni(NO3)2溶液中, 12 h后进行干燥, 于不同温度下焙烧4 h, 自然冷却至室温, 即得到载Ni催化剂, 命名为wNi/xCeO2-(1-x)SiO2-T其中w表示负载Ni的质量分数(5%-20%), x为复合载体中Ce的摩尔分数(x = 0-1), T为负载Ni后进行焙烧的温度(600-900 °C).
比表面积测定(BET, N2物理吸附)在北京金埃谱公司生产的V-Sorb 2800P型比表面积分析仪上进行, 相对压力(p/p0)介于0.05-0.25. 测试前, 样品于150 °C下抽气真空处理2 h, 然后在-196 °C下进行N2等温多点吸附. 依据BET法计算比表面积. X射线衍射(XRD)在荷兰飞利浦公式生产的X'Pert Pro MPD型衍射仪(Co Kα, λ = 0.1790955 nm)上进行, 电压40 kV, 电流120 mA, 扫描范围2θ = 10°-90°, 扫描步长0.02°. 此外, 样品的平均晶粒粒径依据Scherrer公式计算, CeO2的晶胞参数则运用Bragg公式估算. 催化剂的表面形貌在美国FEI公司生产的Nova Nano SEM230型扫描电镜(SEM)仪上观察. 紫外-可见漫反射(DRS)吸收光谱在日本岛津UV-2550型紫外分光光度计上进行室温测试. 波长范围为200-800 nm, 狭缝宽度5 nm, 参照材料为BaSO4. 氢程序升温还原(H2-TPR)测试在天津先权公司产TP-5080型多用吸附仪上进行, 催化剂的用量为50 mg. 测试前, 升温至300 °C并通入N2气吹扫样品表面物理吸附的水分和杂质. 30 min后冷却至室温, 再通入组成为10%H2-90%N2的混合气体(30 mL/min)作还原气. 待基线调零后以10 °C/min的速率开始还原并升温至800 °C. 氨程序升温脱附(NH3-TPD)测试也在TP-5080型多用吸附仪上进行, 样品装载量为100 mg. 表征前的步骤与H2-TPR的一致, 然后以10 mL/min的流量通入NH3进行吸附. 待吸附饱和后关闭NH3, 打开N2吹扫体系约30 min并调零基线, 最后以10 °C/min的速率开始脱附并升温至800 °C. 反应后催化剂的热重(TG)分析在美国TA公司提供的TGA Q50型热重分析仪上进行, 以确定其反应过程中的相对积炭量. 样品在N2气流中吹扫后, 再在空气气流中以10 °C/min的速率开始测试并升温至750 °C.
催化剂的POM反应在天津先权公司生产的WFSM-3060型反应装置中进行, 石英反应管的内径为8 mm, 催化剂的装填量为150 mg. 反应前将催化剂于750 °C下还原1 h, 还原气体为20 mL/min的纯氢. 还原完成后, 用20 mL/min的高纯N2吹扫反应装置20 min. 随后切换成组成为CH4/O2配比为2:1的原料反应气体(30 mL/min), 总空速为12000 mL/(h∙g). 反应10 min后, 经六通阀每30 min取样一次, 联接GC-2060型气相色谱仪分析反应后的尾气组成. 色谱检测条件为: TCD检测器, TDX-01色谱填充柱, 热导池桥电流为70 mA, 载气为高纯N2. 催化剂上CH4转化率、CO和H2选择性及合成气组分的摩尔比按照以下公式计算得出:
X(CH4)= ([CH4]in - [CH4]out)/([CH4]out + [CO]out + [CO2]out) x 100% = (ACO,outFCO,out + ACO2,outFCO2,out)/(ACO,outFCO,out + ACO2,outFCO2,out + ACH4,outFCH4,out) x 100%
S(CO) = [CO]out/([CH4]in-[CH4]out) x 100% =ACO,outFCO,out /(ACO,outFCO,out + ACO2,outFCO2,out) x 100%
S(H2) = [H2]out/2([CH4]in - [CH4]out) x 100% = AH2,outFH2,out /2(ACO,outFCO,out + ACO2,outFCO2,out) x 100%
n(H2)/n(CO) = [H2]out/[CO]out = AH2,outFH2,out/ACO,outFCO,out
式中, Ai和Fi分别表示尾气组分i的色谱峰面积和相对摩尔校正因子, FCH4= 1, FCO = 38.5, FCO2 = 1.9, FH2 = 0.35.
催化剂的比表面积由N2物理吸附结合BET方程计算得到, 表1列出了所有载体和催化剂的比表面积. 可以看出, 复合氧化物载体xCeO2-(1-x)SiO2的比表面积远大于单一CeO2的比表面积, 其比表面积与Ce摩尔分数密切相关, 随着Ce含量的增大, 比表面积逐渐下降. 当Ce摩尔分数x由0增大至1时, 载体的比表面积由470 m2/g逐渐降低至20 m2/g. 为了得到大比表面积的CeO2-SiO2复合载体, 控制Ce/Si比是非常关键的. 固定Ce的摩尔分数x为0.50, 当焙烧温度分别为600, 700, 800和900 °C时, 0.50CeO2-0.50SiO2的比表面积分别为196, 149, 131和96 m2/g, 可见随着焙烧温度的升高, 复合载体的比表面积逐渐下降. 负载10%的Ni后, 催化剂10%Ni/ xCeO2- (1-x)SiO2 (x = 0.25, 0.50, 0.75)的比表面积相对于其载体而相应下降, 这是因为负载的Ni可能占据载体的孔道, 造成催化剂的比表面积下降[25, 28, 29, 30]. 当Ni的含量由10%增大至20%时, 催化剂10%Ni/0.50CeO2- 0.50SiO2- 700 (100 m2/g)和20%Ni/0.50CeO2-0.50SiO2-700 (103 m2/g)的比表面积仅略微变化, 这是因为复合载体0.50CeO2- 0.50SiO2-700具有较大的比表面积(149 m2/g), 能够有效地促进Ni的分散. 催化剂10%Ni/0.50CeO2- 0.50SiO2在900 °C的高温焙烧4 h后, 其比表面积仍然维持在59 m2/g, 表明该催化剂具有较好的热稳定性能.
利用XRD对所制备的载体和催化剂进行了物相分析, 结果示于图1. 图1(a)为不同Ce/Si比载体的XRD谱, 图1(b)为不同Ce/Si比的载体负载10%Ni后催化剂的XRD谱. 由图1(a)可知, 单一载体SiO2-700的XRD谱线中未发现其衍射峰, 表明SiO2呈现无定形态; 其它载体xCeO2-(1-x)SiO2-700 (x = 0.25, 0.50, 0.75, 1)均在2θ为33.3°, 38.7°, 55.8°与66.5°处出现立方萤石型CeO2的衍射峰(Ref. code: 00-034-0394), 并且衍射峰随x的增大而变强和尖锐; 在复合载体中未发现SiO2的衍射峰, 主要原因是SiO2处于无定形态, 衍射峰较弱. 由图1(b)可知, 负载10%Ni后催化剂的XRD谱线中除了CeO2的衍射峰外, 在2θ = 43.7°, 50.8°与74.7°附近还出现了立方相NiO的衍射峰(Ref. code: 01-078-0423). 当负载Ni的含量增大至20%后, NiO的衍射峰逐渐变强而尖锐(图1(c)). 不同温度焙烧后的载体0.50CeO2-0.50SiO2和催化剂10%Ni/0.50CeO2-0.50SiO2的XRD谱图见图1(d)和图1(e). 可以发现, 随焙烧温度升高, 载体和催化剂中CeO2和NiO的衍射峰均逐渐变强.
根据上述XRD谱中CeO2 (2θ ≈ 33.3°)和NiO (2θ ≈ 50.8°)的晶面衍射峰, 利用Scherrer公式和Bragg公式计算分别得到CeO2和NiO的平均晶粒粒径和CeO2的晶胞参数, 结果列于表2. 由计算结果发现, 所有载体和催化剂中的CeO2和NiO均为纳米级晶粒. 已有研究表明, CeO2和SiO2之间可形成一种组成为Ce9.33(SiO4)6O2的中间相态, 该物种在焙烧处理时进一步分解成纳米级的CeO2颗粒和无定形的SiO2, 且CeO2高度分散在SiO2的网格中[26, 31]. 此外, 催化剂中CeO2的晶粒粒径略大于载体中CeO2的晶粒粒径, 可能由负载Ni后经过二次焙烧所引起. 焙烧温度影响载体和催化剂的晶粒粒径, 尤其是经过900 °C高温焙烧后, CeO2和NiO的晶粒粒径增大更为明显, 说明900 °C高温焙烧造成晶粒间发生了一定程度的聚集或烧结, 这也是造成其比表面积降低的主要因素. 另外, 载体CeO2-700和催化剂10%Ni/CeO2-700中CeO2的晶胞参数分别0.5414和0.5416 nm, 而其它载体和催化剂中CeO2的晶胞参数均略小. 这是由于Si4+半径(r = 0.042 nm)比Ce4+半径(r = 0.087 nm)小, 形成的凝胶在焙烧过程中部分Si4+进入CeO2的晶格, 形成CeO2-SiO2复合氧化物, CeO2的晶格发生收缩所致[26].
CeO2-SiO2复合氧化物的形成机理见图式1所示, 该过程包括正硅酸四乙酯的水解、缩合和焙烧脱水等[32, 33]. 首先, 正硅酸四乙酯在酸性环境中水解成原硅酸, 然后原硅酸经缩合失水形成硅酸聚合物, 同时硅酸聚合物表面的-OH与吸附Ce3+表面的-OH产生氢键作用力, 连接在一起. 最后经干燥、焙烧等过程进一步脱水, 产生共价键, 使Ce固定在SiO2相的网格中, 形成分散均匀、具有大比表面积的CeO2-SiO2复合氧化物.
利用扫描电镜对典型样品的表面形貌进行了分析, 结果如图2所示. 可以看出, 载体SiO2-700的表面由大量尺寸不一的颗粒团聚而成, 复合载体0.50CeO2- 0.50SiO2-700的表面形貌没有明显变化. 当载体0.50CeO2-0.50SiO2-700负载10%Ni后, 催化剂的整体形貌虽没有变化, 但其表面有更多细小的颗粒, 这些细小颗粒为沉积的NiO颗粒.
典型催化剂的UV-Vis DRS吸收光谱如图3所示, 同时示出纯CeO2的UV-Vis DRS吸收光谱. 由图可见, 纯CeO2的吸收峰在343 nm左右, 而催化剂中NiO的吸收峰出现在297-350 nm, 此特征吸收谱带归属于NiO晶胞中八面体Ni2+的电荷跃迁[25, 34, 35]. 比较谱线(a)和(b)-(e)发现, 复合载体负载Ni后的10%Ni/xCeO2-(1-x)SiO2-700 (x = 0.25, 0.50, 0.75, 1)催化剂的表面Ni物种的吸收谱带向长波方向移动. 这表明催化剂10Ni/xCeO2-(1-x)SiO2-700中的NiO与复合载体产生了相互作用, 使Ni2+的电荷跃迁所需的能量更低[25, 30]. 这种相互作用还将影响到NiO的还原性能.
图4是催化剂的H2-TPR曲线. 由于复合载体CeO2-SiO2可能产生耗氢峰, 因此先对其进行TPR分析以判断其还原峰与NiO的还原峰是否发生重叠. 由图4(a)可以看出, 载体0.50CeO2-0.50SiO2-700在490 °C左右出现了一个很弱的耗氢峰, 这是CeO2表面晶格氧及吸附氧同H2作用引起的[28, 29, 30, 31, 36]. 由于800 °C以下的SiO2不能被H2还原[36], 因此780 °C附近出现的高温耗氢峰可能是由体相大粒径CeO2的还原所致[36, 37, 38, 39, 40, 41]. 负载活性组分Ni后, NiO的还原峰出现在325-580 °C, 该还原峰随Ni含量的增加而逐渐变大, 但还原峰的位置没有明显改变. 复合载体中x值对催化剂还原性能的影响见图4(b). 单一载体SiO2-700负载Ni后的催化剂仅在365-670 °C出现NiO的还原峰; 而10%Ni/CeO2-700在250-470 °C出现NiO的还原峰, 高于800 °C的峰为大粒径CeO2的还原耗氢峰. 复合载体xCeO2-(1-x)SiO2-700 (x = 0.25, 0.50, 0.75)负载Ni后的催化剂中NiO还原峰的峰温明显低于10%Ni/SiO2-700中NiO还原峰的峰温, 表明Ni负载在xCeO2-(1-x)SiO2复合载体上更容易被还原. 焙烧温度对10%Ni/0.50CeO2-0.50SiO2还原性能的影响见图4(c). 可以看出, NiO还原峰的峰温随焙烧温度的升高而逐渐向高温方向移动, NiO的还原变得更难. 经900 °C焙烧后催化剂的NiO还原峰移动得最明显. NiO还原峰向高温移动的主要原因在于高温焙烧促进了CeO2与SiO2之间强烈的相互作用[29, 42], 导致NiO还原更困难. 同时表2数据也显示, 随焙烧温度的升高, 催化剂中CeO2和NiO的粒径均有所增大, NiO粒径的增大也是造成还原峰向高温移动的原因之一.
利用NH3-TPD分析了典型载体和催化剂表面酸性的强弱, 结果见图5. 图5表明, 单一载体SiO2-700的NH3-TPD曲线中存在两个明显的NH3脱附峰, 分别位于90 °C和450 °C左右, 可分别归属为弱酸和强酸中心吸附氨的脱附峰. 比较发现, 0.50CeO2- 0.50SiO2-700复合载体的弱酸中心脱附峰更小, 且强酸中心的脱附峰消失, 说明形成复合载体后其酸性减弱, 这可能由Ce的碱性所引起的. 载体负载10%Ni后, 催化剂10%Ni/SiO2-700的脱附曲线与其载体的基本保持一致, 同样也存在弱酸中心和强酸中心的脱附峰, 但峰温略微向高温方向偏移; 10%Ni/0.50CeO2-0.50SiO2-700的脱附曲线的峰形与其载体相似, 但是弱酸中心脱附峰明显减小. 在POM反应中, 积炭优先在催化剂表面强酸位置产生和积淀, 且强酸中心的强度和数量越大, 积炭量也相应增加[2]. 对比两个催化剂的NH3-TPD曲线可以看出, 10%Ni/ 0.50CeO2-0.50SiO2-700的酸性比10%Ni/SiO2- 700更弱且几乎没有强酸中心. 因此, 10%Ni/ 0.50CeO2- 0.50SiO2- 700在POM反应过程中可能产生更少的表面积炭.
将反应后的典型催化剂进行了热重测试, 其热重曲线如图6所示. 10%Ni/SiO2-700在低于130 °C, 130- 350 °C和高于600 °C的失重分别来自催化剂表面吸附水的脱除、表面无定形炭的脱除和石墨化炭的燃烧[43, 44], 而350-570 °C的增重则由反应后催化剂中活性中心Ni0被氧化所致. 10%Ni/0.50CeO2-0.50SiO2-700和20%Ni/ 0.50CeO2- 0.50SiO2-700在低于130 °C时的失重较小, 说明其表面吸附的水和无定形炭较少. 另外, 两者测试过程中出现更明显的增重, 说明这两个催化剂反应后仍具有更多的活性中心Ni0. 10%Ni/0.50CeO2- 0.50SiO2- 700在高于450 °C没有明显的失重, 即表面积炭量很小, 然而20%Ni/0.50CeO2-0.50SiO2-700在高于440 °C时出现了大约7.1%的失重, 说明该催化剂表面生成了大量积炭. 据文献[45, 46, 47]报道, 催化剂的平均积炭速率与Ni颗粒粒径和分散性紧密相关, Ni颗粒粒径越小, 分散性越好, 其表面的积炭速率也越低. 此外, 图6还显示, 增大Ni的负载量可产生更多的积炭, 因为随着Ni负载量的增加, Ni晶粒粒径变大, 而大颗粒的Ni更易成为积炭反应的活性中心.
表3列出的是载体0.50CeO2-0.50SiO2-700及其负载不同质量分数的Ni后催化剂的CH4转化率和目的产物CO和H2的选择性. 由该表可知, 载体0.50CeO2- 0.50SiO2-700对CH4转化也具有一定的活性, 其CH4转化率为23.7%, 但CO和H2的选择性仅分别为8.6%和4.5%, 说明在载体表面主要发生的是CH4燃烧反应. 负载5%的活性Ni后催化剂上CH4转化率(71.6%), CO (79.3%)和H2 (89.0%)的选择性明显提高; 继续增大Ni质量分数至10%时, CH4转化率(84.3%), CO (87.4%)和H2 (93.3%)的选择性达到最高值; 但是进一步加大Ni的质量分数至20%时, CH4转化率及CO和H2的选择性开始下降. CH4转化率和目的产物选择性的变化表明, 较低的Ni含量虽然使Ni的分散性能提高, 但由于总体活性中心Ni0的数量较少, 难以提供足够的POM所需的活性中心, 因此甲烷转化率较低; 但过高的Ni含量容易造成Ni晶粒变大(表2)和分散性降低, 生成更多的表面积炭(图6), 同样不利于催化剂整体性能的提高.
不同的载体负载10%Ni后催化剂在POM反应中的催化性能列于表4. 可以看出, 单一载体负载10%Ni后的催化剂10%Ni/SiO2-700和10%Ni/CeO2-700的催化活性均较低; 而复合氧化物载体负载10%Ni后催化剂的CH4转化率和CO, H2选择性大幅度提高, 其中, 10%Ni/ 0.50CeO2- 0.50SiO2-700表现出最高的CH4转化率和CO, H2选择性. 前面的比表面积、H2-TPR和NH3-TPD表征结果表明, 10%Ni/0.50CeO2-0.50SiO2-700具有大比表面积和弱酸性, 同时该催化剂中的NiO较容易被还原成活性中心Ni0, 这是10%Ni/0.50CeO2-0.50SiO2-700催化性能最高的主要原因.
经不同温度焙烧制备的催化剂10%Ni/0.50CeO2- 0.50SiO2对POM反应的催化性能见表5. 由表可知, 焙烧温度对催化剂的催化性能具有显著影响. 其中, 经过700 °C焙烧后的催化剂10%Ni/0.50CeO2-0.50SiO2-700表现出最高的CH4转化率和目的产物选择性. 对比这四个催化剂的比表面积(表1)、晶粒粒径(表2)和还原性能(图4(c))可以看出, 过高的焙烧温度导致催化剂比表面积下降和CeO2, NiO粒子变大, 同时使NiO的还原更加困难, 不能产生充足的催化反应活性中心Ni0, 这是高温焙烧造成催化性能下降的主要原因.
将典型催化剂10%Ni/0.50CeO2-0.50SiO2-700和10%Ni/SiO2-700的反应时间延长至900 min, 考察了CH4转化率和CO, H2选择性随反应时间的变化, 结果见图7. 催化剂10%Ni/SiO2-700在前390 min内催化性能的变化较小, 但随后催化性能迅速下降, 当反应时间为660 min时, 该催化剂CH4转化率及CO和H2的选择性分别降低至58.2%, 67.0%和75.3%, 且H2/CO摩尔比由2.0降至1.54, 说明10%Ni/SiO2-700上的POM反应稳定性较差. 相反, 催化剂10%Ni/0.50CeO2-0.50SiO2-700在900 min的反应时间内, 其CH4转化率基本维持在84%左右, H2/CO摩尔比也一直保持为2.04, 表现出较好的稳定性.
采用溶胶-凝胶法合成了大比表面积的CeO2-SiO2复合氧化物载体. 该系列载体负载Ni后, 对甲烷部分氧化反应具有较高的催化活性. 在Ce/Si摩尔比1:1, Ni的负载质量分数10%, 700 °C焙烧时所得的催化剂Ni/CeO2-SiO2具有最高的CH4转化率和目的产物选择性, 同时该催化剂的稳定性较好. 所得催化剂具有大的比表面积和较弱的酸性, NiO分散性较好且易被还原, 催化剂不容易发生积炭.
2014, Vol. 35 Issue (1):8-20 
PDF (834KB) 2014, Vol. 35 Issue (1):8-20 PDF (834KB)