Graphical Abstract
The partial oxidation of methane (POM) reaction: CH4 + O2
t°C catalyst
NdCaCoO4-derived catalyst is active, selective and stable in POM reaction 100 90 80 70
%
60
X CH4
50
S CO
40 S H2 30 20
10 0 0
20
40
60
80
Time on stream, h
100
120
140
CO + 2H2
Research Highlights
A catalytic activity of NdCaCoO4 in POM reaction was investigated; A decomposition of NdCaCoO4 at POM conditions results in the formation of finely dispersed Nd2O3, CaO and Co particles; The NdCaCoO4-derived catalyst showed conversion of CH4 and selectivity of CO and H2 higher than 90%; No activity and selectivity decrease was observed for 140 h of POM reaction.
Partial oxidation of methane to produce syngas over a neodymium-calcium cobaltate-based catalyst A.G. Dedov1*, A.S. Loktev1, D.A. Komissarenko1, G.N. Mazo2, O.A. Shlyakhtin2, K.V. Parkhomenko3, A.A. Kiennemann3, A.-C. Roger3, A.V. Ishmurzin4 and I.I. Moiseev1 1
Department of General and Inorganic Chemistry, Gubkin Russian State University of Oil and Gas, 65
Leninsky prosp., 119991, Moscow, Russia 2
Department of Chemistry, M.V. Lomonosov Moscow State University, 119991, Moscow, Russia
3
Institut de Chimie et Procédés pour l’Energie, l’Environnement et la Santé, UMR7515 CNRS ECPM,
University of Strasbourg, 25 Rue Becquerel, 67087 Strasbourg Cedex 2, France 4
Joint-Stock company “Gazprom” 16 Nametkina St., 117997, Moscow, Russia
*Corresponding author. Tel.: +7 499 507 82 80; fax: +7 507 82 65 E-mail address:
[email protected] (A.G. Dedov)
Abstract Synthesis gas production by partial oxidation of methane (POM) (CH4/O2 = 1.8-4.5/1) at 850-960 oC over NdCaCoO3,96 was investigated using a fixed bed flow-type reactor. The NdCaCoO3.96-derived catalyst demonstrated conversion of CH4 and O2 up to 90% while CO and H2 selectivities were over 90% at a H2/CO ratio ~ 2. The decomposition of NdCaCoO3.96 at POM conditions results in the formation of finely dispersed Nd2O3, CaO, cobalt oxides and cobalt metal. The as-prepared catalyst exhibited excellent stability of the catalytic properties, and no activity decrease was observed for 140 h of reaction. Synthesis gas formation was accompanied by carbon deposition on the catalyst. However, this process has little or no influence on the catalytic properties of the material.
Keywords: Syngas; Hydrogen; Partial oxidation of methane; Layered perovskite; Nd2-xCaxCoO4±δ
1. Introduction The conversion processes that convert natural gas to hydrogen, synthesis gas, or ethylene have attracted significant attention in recent years [1-5]. The production of hydrogen and synthesis gas has been widely studied due to their potential application as strategic sources of clean energy [6–8]. Hydrogen can be used in fuel cells, which are currently considered to be the lowest pollutant emission energy source. Simultaneously, the most economically available route from natural gas to valuable chemicals is the conversion of methane to synthesis gas [9]. Production of synthesis gas is currently produced on an industrial scale by steam reforming of 1
methane [10]. However, the partial oxidation of methane (POM) has some advantages compared to steam reforming. First, the POM process produces syngas with a molar ratio of CO: H2= 1:2, which is favorable for the Fisher-Tropsch and methanol synthesis. In addition, POM is a more energy efficient process, which reduces the environmental impact. The most well-known catalysts for the POM are transition metals including nickel [11,12]. The partial oxidation of methane at 700-900C over nickel metal on a refractory support was first described in [13]. Aging of the catalyst in a mixture of methane and oxygen at these temperatures results in the fast deactivation of the catalytic layer due to carbon deposition and significant loss of metal during the catalytic process [2,12,14]. Cobalt catalysts typically produce lower methane conversion and synthesis gas selectivity compared to nickel catalysts [11,12]. However, cobalt has higher melting and vaporizing points, and cobalt-based catalysts are less active for syngas methanation process. The addition of CoO especially to NiO/ZrO2 caused a dramatic reduction in carbon deposition [15]. In addition, no carbon formation was detected on the supported CoO catalysts [12]. The alternative approach for catalyzing POM involves the application of mixed perovskite-like oxides. These materials are able to produce finely dispersed metallic particles during the reduction process and to prevent the deactivation of the catalyst by suppressing coke formation [12,16]. Lago et al. reported that the preliminary reduction of GdCoO3 in situ resulted in cobalt metal being finely dispersed over a rare earth sesquioxide support. The as-prepared composite exhibited a methane conversion of 73% and a selectivity of 79 and 81% for CO and H2, respectively, at 730C. The La-Co-O-based catalyst that was obtained using a similar prereduction was active for methane combustion, and only trace amounts of CO and H2 were observed under the reaction conditions. XRD and XPS analyses of the spent La-Co-O catalyst indicated that under the reaction conditions, the metallic cobalt was completely reoxidized regenerating the original LnCoO3 perovskite structure [17]. According to [18], a second metal should be included in the perovskite structure to stabilize the small metal clusters formed on the catalyst surface. Apart from the perovskite-type ABO3 complex oxides, the application of oxides with K2NiF4-like structures in POM catalysis has been very limited. NdSrCu1-xCoxO4-, Sm1.8Ce0.2Cu1-xCoxO4+ [19] and La2-xKxCuO4 [20] are more useful as total combustion catalysts of methane and diesel soot, respectively, due to their chemical stability under the reaction conditions and the high mobility of the lattice oxygen. For (La0.5Sr0.5)2Ni1-xFexO4+ with a K2NiF4-like structure, efficient POM catalysis can be attributed to the cooperative effect of the lattice oxygen in the complex oxide and nanodomains of nickel obtained by the decomposition of precursor double-perovskite type (La0.5Sr0.5)2FeNiO6+ in the POM environment [21]. 2
The possibility of complex and individual oxides being reduced to metals under POM conditions makes it difficult to identify the contributions of various components to the catalytic effect of the whole system. Even for a single catalyst, the reaction between CH4 and O2 may follow different pathways depending on the state of the catalyst and the conditions applied [12]. In addition, identification of the various contributions is necessary to further optimize the catalytic system. In our preliminary communication [22], we reported the design of highly active and rather stable POM catalysts with 100% selectivity to synthesis gas based on the K2NiF4-like cobaltates Ln2–xMxCoO4±δ (Ln = La, Nd; M = Sr, Ca) complex. Comparative analysis of the catalytic properties of Nd2–xCaxCoO4±δ (x = 1; 0.75) demonstrated the superior CO selectivity and better (H2+CO) yield during POM reaction over the NdCaCoO3,96-based catalyst. The current paper examines the details of the catalytic performance and chemical transformations of NdCaCoO3,96 in the POM reaction.
2. Experimental 2.1. Catalyst preparation The preparation of NdCaCoO3,96 was performed using a solid-state synthesis method, as described in [23]. Corresponding amounts of preliminarily calcined Nd2O3, Co3O4 and CaCO3, were mixed in a stoichiometric ratio in the planetary ball mill at 600 rpm in heptane for 30 min and then annealed at 1000-1200C for 24 h in air with intermediate grinding. After pressing the powders into pellets ( = 8 mm), the samples were annealed at 1000-1200 C in air followed by slow cooling to room temperature. After crushing the pellets in an agate mortar, a fraction of 100-250 mesh was selected by sieving.
2.2. Catalyst characterization The as-prepared catalyst was characterized by XRD using a Huber IMAGE FOIL G670 diffractometer (CuK1 radiation). Phase analysis and crystal structure refinement were performed using the WinXPow software (STOE) and PDF-2 ICDD database. Oxygen stoichiometry of the initial single-phase sample was determined by iodometric titration. The powder of the complex oxide was dissolved in diluted HCl with an excess amount of KI, and then, the solutions were titrated using standard thiosulfate solutions. SEM analysis with secondary and backscattered electrons was performed with LEOSUPRA 50 VP scanning electron microscope with a field emission cathode at U = 10-20 kV. The powder of the sample was fixed at the copper substrate using a carbon-based conducting 3
adhesive. EDX analysis at U = 15 kV by means of an energy dispersive INCA x-SIGHT spectrometer attached to a scanning electron microscope. The specific surface area of the samples was measured by the single-point BET method at liquid nitrogen temperature with argon as adsorbate (volumetric system “Sorbtometer”, Catacon, Russia). The chemical composition of the particle surface and oxidation state of elements were analyzed by X-ray photoelectron spectroscopy (XPS) using a ThermoScientific Multilab 2000 employing Al K and Mg K radiations at a residual pressure of 10-9 mbar. The binding energy spectra were calibrated using the C1s line (E = 284.6 eV) internal standard. Temperature programmed reduction (TPR) was studied using a Micromeritics AutoChem 2 Chemisorption Analyzer in an Ar/H2 (10/1) gas mixture at a flow rate of 50 cm3min-1. The temperature of the samples (~ 0.05 g) was increased to 1000C at a heating rate of 10 C min -1. The carbon content in the spent catalyst was determined using an Elementar Vario Micro Cube CHN analyzer. Thermal analysis of this sample was performed in air with a heating rate of 5 C min-1 with a Netzsch STA 409 PC Luxx thermal analyzer equipped with a QMS 403 Aeolos quadrupole mass spectrometer.
2.3 Catalytic activity The partial oxidation of methane was performed using a single-pass plug flow setup (Fig. 1) including a flow-fixed bed reactor (8 mm internal diameter, 650 mm length). The reactor was equipped with an internal pocket for a thermocouple ( = 4 mm). The catalyst was placed in a hot zone of the reactor at a distance of 500 mm from the inlet. The reactor was filled with quartz glass inserts to reduce the free volume upstream and downstream from the catalyst. These quartz inserts were tightly held in the thermocouple pocket. 5 mm quartz wool layers (mixer-filter) were placed at the inlet of the reactor and upstream and downstream from the layer of the catalyst. All of the tests were carried out with 0.1 g of catalyst. The catalyst bed height was only 1 mm to prevent temperature changes in the catalyst bed. All of the tests were carried out at atmospheric pressure using CH4/O2 mixtures in ratios of 1.8–4.5/1 without any dilution. The catalysts were tested in the temperature range of 850–940 ◦C with a GHSV 15–75 Lg-1h-1. A 140 h run of the catalyst was carried out periodically. The catalyst was heated in methane-oxygen flow to the reaction temperature for 2-3 h and held for 1-5 h at the steady state. After analysis the furnace was switched off and catalyst was cooled to room temperature during 3-4 h. A total runtime of 140 h corresponds to the time 4
for the POM to achieve steady state reaction conditions but does not include the heating and cooling time. The initial gas mixture and outlet gas were analyzed using two gas chromatographs. Helium was used as a carrier gas in both chromatographs. The first chromatograph was equipped with a TCD and two stainless steel columns that were 3 m in length and 3 mm in diameter. The first column was filled with Porapak Q (50–80 mesh, Waters Assoc., USA) and used to determine the content of air, methane, CO2, ethane and ethylene. The second column was filled with 8% Na2CO3 supported on alumina and used to determine the content of air, methane, ethane, ethylene and C3+ hydrocarbons. The temperature of the columns during analysis was 80 ◦C. The chromatograms were treated with an internal normalization method with a correction for the molecular weight of the components. The second chromatograph was equipped with a TCD and a stainless steel column that was 2 m in length and 3 mm in diameter and packed with NaX zeolite (Gazprom Neftekhim Salavat, Russia). This column operates at ambient temperature and allows for the determination of the contents of hydrogen, oxygen, nitrogen, carbon monoxide and methane. Due to the small difference in thermal conductivities of He and H2, negative H2 peaks can be occasionally observed. To solve this problem, we used the H2 peak inversion procedure and special software for H2 content calculation based on the results from analysis of model gas mixtures. The methane conversion (X), product selectivity (S) and yield (Y) of products are defined as follows:
X CH4 (%) = [(moles of CH4 converted)/(moles of CH4 in feed)]×100 S CO or CO2 (%) = [(moles of C in products)/(moles of CH4 converted)]× 100 S H2 (%) = [(moles of H2 produced)/(2×moles of CH4 converted)]× 100 S C2+ products (%) =[(n×moles of Cn in products)/(moles of CH4 converted)]×100 Y products (%) = [X CH4 (%)×S products (%)]/100 The ProSim suite was used as a tool for to perform the thermodynamic calculations. The thermodynamic analysis was carried out by minimizing the Gibbs free energy (G). This method determines the composition of the gas mixture at equilibrium by minimization of G at given values of temperature and pressure. For the following analysis, the approach that requires specification of the different compounds that could exist in equilibrium was chosen.
3. Results and Discussion 3.1 Catalytic activity studies 5
The experimental CH4 conversion, CO yield values over the NdCaCoO3,96-based catalyst at various temperatures and the corresponding equilibrium values of CH4 conversion calculated by the ProSim software are shown in Fig. 2a. According to these data, the CH4 conversion increases abruptly at 850-865 °С followed by a steep increase with temperature. At T < 860 °C, the full oxidation of CH4 to CO2 dominates over the POM reaction (Fig. 2b), and the amount of CO2 produced is higher than the corresponding amounts of CO and H2. At T > 860 °C, the yield of synthesis gas increases with temperature approaching the equilibrium values, and the concentration of CO2 decreases. This behavior could be associated with a two stage combustionreforming POM mechanism where CH4 is first oxidized to CO2 and H2O followed by a second stage where CH4 reacts with the as-formed deep oxidation products [12, 24]. The first stage of the process is exothermic and relatively fast, but the second stage is endothermic and less thermodynamically favorable, which hinder these reactions at lower temperatures. The amount of hydrocarbon byproducts (C2+) is significant at T = 850 °C while at T > 880 C, this amount approaches zero (Fig. 2b). The formation of these species could be associated with the recombination of methyl radicals in the gas phase at reduced temperatures, and at higher temperatures, these byproducts are rapidly oxidized. The level of methane conversion over this catalyst at 915C was 89% (Fig. 2a), which corresponds to 78% yield of synthesis gas (H2/CO = 1.9). The maximum synthesis gas selectivity equal to the equilibrium value (100%) was achieved at 925 ºС at 90% CH4 conversion (Fig. 2b). It should be noted also that H2 and CO selectivities at T= 925 ºС approached the values corresponding to the thermodynamic equilibrium, and, hence, since that point the POM reaction became thermodynamically, not kinetically controlled. This effect is rather common for Ni-based POM catalysts where it can be observed at temperatures down to 500C, though usually the equilibrium values of H2 and CO selectivities are observed at 600-750C [12]. A similar effect was observed also for cobalt metal - based catalysts at slightly higher temperatures (T = 800-850 C, [25, 26]). Equilibrium yields in the lack of carbon deposition can be achieved at Rh-based catalysts only at elevated pressures (5-8 bar) and T ~ 1000C [27]. The contact time of the reactants with the catalyst surface substantially affects the POM results. The highest CH4 conversion value was observed at WHSV 17 L×g-1×h-1. A further increase in the gas flow rate causes a significant decrease in the methane conversion (Fig. 3). In addition to CH4 conversion, the variations in the selectivity at 17-37 L×g-1×h-1 were negligible (Fig. 3). However, a further increase in the flow rate resulted in a substantial decrease in the CO and H2 selectivity and a corresponding increase in the selectivity to CO2. Similar results were also observed in [14,28,29] for Ni and Co metal catalysts on various supports. These features may be due to the two stage POM reaction mechanism when shorter contact times for very 6
narrow catalyst beds primarily allow for the faster and more thermodynamically favorable combustion stage of the POM reaction and suppresses the endothermic and slower stage of the syngas formation process. The only reaction parameter affected by the CH4/O2 ratio is the CH4 conversion, which progressively decreases as the methane concentration increases (Fig. 4). In addition, the selectivity of synthesis gas formation and the H2/CO ratio remain nearly stable at CH4/O2 = 2.14.6. This type of behavior indicates that at CH4/O2 > 2, the level of methane conversion is limited by the progressively reduced amount of oxygen in the gas mixture, and the methane oxidation mechanism remains the same at least until CH4/O2 = 4. An outstanding feature of the NdCaCoO3.96–based catalyst is the exceptional stability of its catalytic properties under POM reaction conditions (Fig. 5). High levels of CH4 conversion and CO selectivity remain nearly constant at least for 140 h of use at the optimum reaction conditions. It should be noted that the H2 and CO selectivity values demonstrate much less variation in time compared to CH4 conversion. These small variations may be due to features of solid-state reactions in the catalyst that can be studied by analyzing the prepared catalyst and spent catalyst at various reaction times.
3.2 Composition and structure of catalyst prior to the reaction XRD analysis of NdCaCoO4- demonstrated that the sample was a single phase with a layered perovskite-like structure that is similar to K2NiF4. The lattice parameters (a = 3.7358(3) Å, c = 11.912(2) Å) correlate well with data from previous studies [30,31]. According to the iodometric titration data, the average oxidation state of Co is 2.92, which corresponds to = 0.04 for NdCaCoO4-. Therefore, NdCaCoO3.96 contains significant numbers of defects in the oxygen sublattice. A comparison of the bulk and surface composition data of the catalyst particles obtained by EDX and XPS analysis, respectively, indicate a significant difference (Table 1). The component ratio in the bulk was close to the stoichiometric value, and the surface of the particles was substantially enriched by an alkaline-earth element. Such a difference can be explained by the well-known tendency of calcium oxide to form carbonates in air and by their migration to the reaction interface from the core of the particles. Similar enrichment of the particle surface by strontium was recently observed for Sr-substituted perovskites [32,33]. The XPS spectrum of the sample in the Nd 3d5/2 zone (Fig. 6a) with the center of a single peak at 982 eV is common for the Nd (III) oxidation state [19]. The position of the main peak in the Co 2p area (780.0 eV, Fig. 6b) corresponds to a mixture of Co(II) and Co(III). The binding energy of Co in Co3O4 is 779.6 eV, and the corresponding value for CoO is 780.5 eV [34]. 7
3.3 Temperature programmed reduction of the catalyst The temperature programmed reduction profile of NdCaCoO3.96 is shown in Fig. 7. The results indicate reduction bands are located at 571С and 747C, which correspond to a total H2 consumption of 2.36 mmol/gcat. This value of H2 consumption is lower than the corresponding theoretical value (4.72 mmol/gcat) estimated according to the bulk reduction of NdCaCoO3.96 to cobalt metal, Nd2O3 and CaO. According to [35], the first broad peak at 400-650C with a maximum at 571C may be due to both the reduction of adsorbed oxygen in NdCaCoO3.96 vacancies and the reduction of Co(III) to Co(II). The second peak at 747C corresponds to further partial reduction of the Co (II) oxide to Co metal particles. This model of the NdCaCoO3.96 behavior during TPR correlates with the results of XRD analysis where all of the peaks in the XRD pattern correspond to Co metal, CoO, Nd2O3 and CaO (Fig. 8). According to [12, 24], both Co metal and Co oxides may be the active centers of POM reaction.
3.4 Analysis of spent catalyst The reduction potential of the gas mixture during the POM reaction is smaller than the corresponding value for hydrogen. Therefore, if NdCaCoO3.96 is unstable in the POM environment, it is reasonable to expect the appearance of Co oxides among the products from NdCaCoO3.96 decomposition. Indeed, XRD analysis of the NdCaCoO3.96 catalyst after 12 h of processing in the POM environment (Fig. 9) revealed significant amorphization of the catalyst and partial degradation of the initial K2NiF4-like phase. Several of the most intense peaks of the decomposition products were due to Nd2O3. Further processing of the catalyst in the POM environment results in the appearance of peaks related to Co metal. CHN analysis of the spent NdCaCoO3.96 catalyst indicated that after 140 h of POM processing, various portions of the spent catalyst contain 7-11 mass % of carbon. The main features of the TG-MS analysis curves of the 140 h processed catalyst in air (Fig. 10) correlate well with the XRD and CHN analysis data. The mass intake at 350-400C caused by the oxidation of cobalt metal to oxides occurs prior to the two step mass loss at 400-700C. This temperature range corresponds to the oxidation temperatures of amorphous carbon in air while the mass decrease of the sample corresponds to the amount of carbon obtained by CHN analysis. The observed two step mass loss, which is due to the oxidation of amorphous carbon, was confirmed by MS analysis of the evolved gases. The main gaseous product of the reaction that appeared during this process was CO2 with little or no H2O evolution (MS analysis of CO is
8
complicated by the significant coincidence of the CO2 and CO mass spectra). Burning of carbon at 430-500C caused a reduction in the oxygen amount in the air flowing through the TG cell. Carbon deposition during the POM reaction typically results in fast deactivation of the catalysts [12,36]. However, a stability of the corresponding values as a function of time (Fig. 5) indicates that in the current case, the carbon deposition has no effect on the activity and selectivity of the NdCaCoO3.96-derived catalyst for at least 140 h. A similar absence in the influence of heavy carbon deposition on the activity and selectivity of the Ni-based POM catalyst has been previously observed [37]. This unusual phenomenon can be explained with the results from the SEM analysis. According to Fig. 11, the decomposition of the initial K2NiF4-like perovskite cobaltate during the POM reaction was accompanied by the complete transformation of the initial microstructure of NdCaCoO3.96. Reductive decomposition of the complex oxide results in the delamination of the initial micron-size grains to finer particles that primarily contain Nd2O3, CaO and Co (see Fig. 9). These observations agree with the results of surface area analysis. Thermal processing of fresh catalyst in POM environment results in the increase of its specific surface area since 0.4 m2 g-1 to 1.3 m2 g-1. According to the CHN analysis, the spent catalyst contains also significant amounts of amorphous carbon. It appears reasonable that the smallest particles in Fig. 11b may be due to pyrolytic carbon. These agglomerated particles form a porous, easily penetrated layer on the coarse grains of the metal/oxide composites that could not prevent access of CH4 and O2 to the surface of the catalyst, which should not significantly affect the catalytic properties of the composite.
3.5 Evaluation of catalytically active components Numerous studies of various efficient and highly selective POM catalysts demonstrated that the chemical stability of ABO3- and A2BO4-based complex oxides during high temperature catalytic oxidation of hydrocarbons can vary. Several authors reported the retention of the initial crystallographic structure of the catalyst during the course of the reaction [19,21,38,39], and other studies indicated their complete decomposition [18,24]. A comparison of the catalyst performance conditions demonstrated that in the first case, the oxidation reaction typically occurred at high dilution of (CH4+O2) by inert gases. This type of catalytic process ensures rather a moderate reduction potential of the reaction gas mixture but demands expensive separation of the reaction products from inert gas. In the second group of studies, oxidation was performed in the undiluted mixtures of hydrocarbons with air or oxygen and resulted in the formation of a gas mixture with high 9
reductive potential that promotes the decomposition of complex oxides in most of these cases. Taking into account these features, preliminary reduction of ABO3 precursors in H2/inert gas mixtures was employed. However, further processing of these catalysts was not accompanied by the recovery of the complex oxide. The reaction conditions applied in the current study and the behavior of NdCaCoO3.96 in the POM environment correlates fairly well with previously reported results obtained under similar conditions. Studies of the catalytic POM reaction in the undiluted CH4/O2 mixtures typically applied in industry allows for a more realistic evaluation of the potential industrial application of these catalysts. Filling the free volume of reactor with inert inserts is also useful for the application of undiluted gas mixtures due to efficient inhibition of undesired chain oxidation reactions. High temperature processing of NdCaCoO3.96 at these conditions results in the rapid destruction of the K2NiF4-like complex oxide. Because the NdCaCoO3.96-derived catalyst becomes active in the POM environment within ~ 1 h and demonstrates stable catalytic properties for at least 140 h (Fig. 5), its high activity and selectivity may be due to the products of its reductive decomposition. Continuous topochemical transformations of NdCaCoO3.96 due the POM reaction could promote the continuous formation of active catalytic species that can promote the high performance of the catalyst. However, the affiliation of active centers with particular decomposition products is not obvious. According to [12], Co-containing POM catalysts primarily consist of Co metal and Co oxide particles on various carriers. The effect of cobalt oxidation during the POM is typically considered negative because the oxidation is often accompanied by the chemical reaction of the catalyst with the carrier [40]. It is known also that Co oxide particles may be also partially reduced in the POM environment [12] while rare earth oxides may be efficient promoters of Cobased POM catalysts [12,41]. Therefore, the Nd2O3 particles that appeared during the reductive decomposition of NdCaCoO3.96 may significantly contribute to the activity and stability of the as-obtained catalyst. It should be noted that similar to other complex oxide-derived POM catalysts, the interplay of the particular decomposition products of NdCaCoO3.96 in the catalytic process is rather complex. In addition, more detailed analysis of the contribution of various components to the catalytic properties of the composite catalyst is needed along with careful study and optimization of the NdCaCoO3.96 decomposition process to ensure maximum catalytic performance of the obtained composite catalysts.
4. Conclusions
10
Analysis of the catalytic properties of NdCaCoO3.96 for the partial oxidation of methane indicated its superior performance. Both the methane conversion and CO and H2 selectivity over this catalyst significantly increased at 850 to 865ºС and approached 100% selectivity and 85% methane conversion at 925 C. The catalyst demonstrates excellent stability, and the levels of CH4 conversion and CO and H2 selectivity remain nearly constant for 140 h of use at 915 to 925C. Analysis of the spent catalyst demonstrates the complete decomposition of the micronsized NdCaCoO3.96 precursor particles and the formation of an ultrafine mixture of Nd2O3 and cobalt particles due to the significant reductive properties of the produced gas mixture. The POM under periodic conditions is accompanied by noticeable carbon formation. However, due to the porous structure of the obtained carbon deposits, this process has little or no influence on the activity or selectivity of the catalyst. The outstanding catalytic performance of the as-obtained material for the POM may be due to the formation of a close mixture of finely dispersed Co catalyst and Nd2O3 promoter particles.
Acknowledgments
This work was supported by the Research Programs of the Russian Academy of Sciences, the Russian Foundation for Basic Research (projects 13-03-00381 and 13-03-12406), the Russian Scientific Foundation (project 14-13-01007), Joint-Stock Company “Gazprom” and by the Ministry of Education and Science of Russia within the performance of a basic unit of the state task "Organization of Carrying Out Scientific Research", questionnaire No. 1422 and the framework for the implementation of the project of the State job in the field of scientific activity, task No. 4.306.2014/K. References [1] A. Holmen, Catal. Today 142 (2009) 2-8. [2] Y.H. Hu, E. Ruckenstein, Adv. Catal. 48 (2004) 297-345. [3] V.S. Arutyunov, O.V. Krylov, Russian Chemical Reviews 74 (2005) 1111-1137. [4] A.G. Dedov, G.D. Nipan, A.S. Loktev, A.A. Tyunyaev, V.A. Ketsko, K.V. Parkhomenko, I.I. Moiseev, Appl. Catal. A: Gen. 406 (2011) 1-12. [5] A.G. Dedov, A.S. Loktev, I.I. Moiseev, A. Aboukais, J.-F. Lamonier and I.N. Filimonov, Appl. Catal. A: Gen. 245 (2003) 209- 220. [6] J.H. Jun, T.H. Lim, S.W. Nam, S.A. Hong, K.J. Yoon, Appl. Catal. A: Gen. 312 (2006) 27– 34. 11
[7] F. Pompeo, N.N. Nichio, O.A. Ferreti, D. Resasco, Int. J. Hydrogen Energy 30 (2006) 1399– 1405. [8] Q.G. Yan, T.H. Wu, W.Z. Weng, H. Toghiani, R.K. Toghiani, H.L. Wan, C.U. Pittman Jr., J. Catal. 226 (2004) 247–259. [9] W. Shan, M. Fleys, F. Lapicque, D. Swierczynski, A. Kiennemann, Y. Simon, P.-M. Marquaire, Appl. Catal. A: Gen. 311 (2006) 24-33. [10] A.P.E. York, T. Xiao, M.L.H. Green, Top. Catal. 22 (2003) 345-358. [11] V.A. Tsipouriari, Z. Zhang, X.E. Verykios, J. Catal. 179 (1998) 283-291, 292-299. [12] B. C. Enger, R. Lødeng, A. Holmen, Appl. Catal. A: Gen. 346 (2008) 1-27. [13] M. Prette, C. Eichner, M. Perrin, Trans. Faraday Soc. 43 (1946) 335-346. [14] X. X. Gao, C. J. Huang, N. W. Zhang, J. H. Li, W. Z. Weng, H. L. Wan, Catal. Today 131 (2008) 211-218. [15] V.R. Choudhary, A.M. Rajput, B. Prabhakar, A.S. Mamman, Fuel 77 (1998) 1803-1807. [16] A. García, N. Becerra, L. García, I. Ojeda, E. López, C. M. López, M. R. Goldwasser, Adv. Chem. Eng. Sci. 1 (2011) 169-175. [17] R. Lago, G. Bini, M.A. Pena, J.L.G. Fierro, J. Catal. 167 (1997) 198-209. [18] A. Slagtern, U. Olsbye, Appl. Catal. A: Gen. 110 (1994) 99-108. [19] J. Deng, L. Zhang, H. Dai, H. He, C.T. Au, Appl. Catal. B: Env. 89 (2009) 87-96. [20] L. Zhu, X. Wang, C. Liang, J. Rare Earths 26 (2008) 254-257. [21] X. Yin, L. Hong, Appl. Catal. A: Gen. 371 (2009) 153-160. [22] A.G. Dedov, A.S. Loktev, G.N. Mazo, L.S. Leonova, D.A. Komissarenko, Yu.A. Mamaev, M.S. Kaluzhskikh, O.A. Shlyakhtin, E.P. Kuznetsova, M. N. Kartasheva, I.I. Moiseev, Doklady Physical Chemistry 441 (2) (2011) 233-236. [23] O.A. Shlyakhtin, G.N. Mazo, M.S. Kaluzhskikh, D.A. Komissarenko, A.S. Loktev, A.G. Dedov, Mater. Lett. 75 (2012) 20-22. [24] L.D. Vella, J.A. Villoria, S. Specchia, N. Mota, J.L.G. Fierro, V. Specchia, Catal. Today 171 (2011) 84– 96. [25] H.Y. Wang and E. Ruckenstein, Catal. Lett. 75 (2001) 13-18. [26] A. Slagtern, H.M. Swaan, U. Olsbye, I.M. Dahl, C. Mirodatos, Catal. Today 46 (1998) 107115. [27] J.C. Slaa, R.J. Berger, G.B. Marin, Catal. Lett. 43 (1997) 63-70.. [28] K.Nakagawa, N.Ikenaga,Y.Teng,T.Kobayashi,T.Suzuki, Appl. Catal. A: Gen. 180 (1999) 183-193. [29] S.Yang, J.N. Kondo, K. Hayashi, M. Hirano, K. Domen, H. Hosono, Appl. Catal. A: Gen. 277 (2004) 239-246. 12
[30] L. Borovskikh, G. Mazo, E. Kemnitz, Solid State Sci. 5 (2003) 409-417. [31] L.V. Borovskikh, S.N. Putilin, L.S. Leonova, G.N. Mazo, Inorg. Mater. Appl. Res. (5) (2008) 54-59. [32] J.N. Kuhn, U.S. Ozkan, J. Catal. 253 (2008) 200-211. [33] N. Lakshminarayanan, H. Choi, J. N. Kuhn, U. S. Ozkan. Appl. Catal. B: Env. 103 (2011) 318-325. [34] T.J. Chuang, C.R. Brundle, D.W. Rices, Surf. Sci. 59 (1976) 413-429. [35] W. Shan, M. Luo, P. Ying, W. Shen, C. Lee, Appl. Catal. A: Gen. 246 (2003) 1-9. [36] J.B. Claridge, M.L.H. Green, S.C. Tsang, A.P.E. York, A.T. Ashcroft, P.D. Battle, Catal. Lett. 22 (1993) 299-305. [37] V.R. Choudhary, V.H. Rane, A.M. Rajput, Catal. Lett. 22(1993) 289-297. [38] M. S. S. Khine, L. Chen, S. Zhang, J. Lin, S. P. Jiang, Int. J. Hydrogen Energy 38 (2013) 13300-13308. [39] E. Campagnoli, . A. C. Tavares, . L. Fabbrini, I. Rossetti, . Yu. A. Dubitsky, . A. Zaopo, L. Forni, J. Mater. Sci. 41 (2006) 4713–4719. [40] V.D. Sokolovskii, J.C. Jeannot, N.J. Coville, D. Glasser, D. Hildebrandt, M. Makoa, Stud. Surf. Sci. Catal. 107 (1997) 461-465. [41] V.R. Choudhary, A.M. Rajput, V.H. Rane, Catal. Lett. 16 (1992) 269-272.
13
Table captions Table 1. Bulk composition (EDX) and surface composition (XPS) of the synthesized material. Figure captions Fig. 1. Single-pass plug-flow setup. 1 – Quartz reactor, 2 – Quartz-glass insert, 3 – Quartz wool, 4 – Catalyst, 5 – Internal pocket for thermocouple, 6 – Electric furnace, 7 – Condensate receiver (round-bottom flask), 8 – Thermocouple, 9 – Water cooler, 10 – Flow meters, 11 – Gas mixer, 12 – Ball-in-tube flow meters, 13 – Adsorbent, 14 – Gas flows from META-CHROM, 15 – Furnace controller, 16 – Temperature indicator. Fig. 2. Catalytic activity (a) and selectivity (b) of the NdCaCoO3.96-based catalyst during the partial oxidation of methane at various temperatures (CH4/O2 = 1,8-1,9; GHSV = 17-18 L×g-1×h1
).
Fig. 3. Influence of GHSV on the behavior of NdCaCoO3.96 (t=920-925 oC, CH4/O2 = 2-2,2). Fig. 4. Influence of the CH4/O2 ratio on the behavior of NdCaCoO3.96 (t= 915-925 oC; GHSV = 20 L×g-1×h-1). Fig. 5. Variation of the catalytic properties of the NdCaCoO3.96-based catalyst as a function of time (t= 915-920 oC; CH4/O2 = 2,1-2,7; GHSV = 20-22 L×g-1×h-1). Fig. 6. XPS spectra of the (a) Nd3d5/2 level and (b) Co2p level for NdCaCoO3.96. Fig. 7. TPR profile of NdCaCoO3.96. Fig. 8. XRD patterns of NdCaCoO3.96 before and after TPR. Fig. 9. XRD pattern of the NdCaCoO3.96 catalyst before, after 12 h and after 140 h of processing in the POM environment. Fig. 10. TG-MS curves of the spent catalyst in air after 140 h of processing. Fig. 11. SEM images: a – before catalytic test; b – after 12 h of processing in the POM environment.
14
Table
Table 1. Bulk composition (EDX) and surface composition (XPS) of synthesized material Composition NdCaCoO3.96
EDX XPS
Nd, %
Ca, %
Co, %
36.4 16.4
31.5 62.4
32.1 21.2
Figure
8 1 2
3
To GC
10
4
11
5 6
12
9 13 Water
16 7
Oxygen Methane
15
14
Fig. 1
Figure
100 90
Thermodynamic equilibrium value of X CH4
80
X CH4
70
Y CO
60
Y H2
% 50
Y CO2
40 30 20 10 0 830
840
850
860
870
880
890
900
910
920
930
Temperature, ºC Fig. 2a
Figure
100 90
Thermodynamic equilibrium value of S CO and H2
80
S CO
70
S H2 S CO2
60
%
S C2+
50
40 30 20 10 0 830
840
850
860
870
880
890
900
910
920
930
Temperature, ºC Fig. 2b
Figure
100 90 80 70 60 X CH4
%
50
S CO S H2
40
S CO2 30
S C2+
20 10 0 15
25
35
45
55
Space velocity, L*g-1*h-1
65
75
Fig. 3
Figure
100 90 80 70 60
%
X CH4
50
S CO 40
S H2 S CO2
30
S C2+ 20 10 0 2,1
2,6
3,1
3,6
Ratio CH4/O2
4,1
4,6
Fig. 4
Figure
100 90 80 70 60
%
X CH4 50 S CO 40 S H2 30 20 10 0 0
20
40
60
80
Time on stream, h
100
120
140
Fig. 5
Figure
Intensity (a.u.)
982.0
960
965
970
975
Binding energy (eV)
980
985
990
Fig. 6a
Figure
Intensity (a.u.)
780.0
775
777
779
781
Binding energy (eV)
783
785
787
Fig. 6b
Figure
H2 consumption intensity (a.u.)
747
200
571
300
400
500
600
Temperature, ºC
700
800
900
1000
Fig. 7
Figure
+
NdCaCoO3.96
+
Nd2O3 CaO
*
Co metal
Intensity (a.u.)
+
+
+
+
Before TPR
+ +
+
+
+
+
+
+
*
10
CoO
20
30
*
+
+ +
After TPR
40
º2θ
50
60
70
Fig. 8
Figure
+
+
NdCaCoO3.96 Nd2O3
+
CoO *
Co
+
+
+ +
Before catalytic test
+
+
+
+
Intensity (a.u.)
+
+
+ +
+
*
30
35
+
After catalytic test (140h)
*
25
+
After catalytic test (12h)
*
20
CaO
40
45
2ºθ
50
55
60
65
70
Fig. 9
Figure Click here to download high resolution image
Figure
a
b
Fig. 11