versión On-line ISSN 0717-9707
J. Chil. Chem. Soc. v.50 n.1 Concepción mar. 2005
J. Chil. Chem. Soc., 50, N 1 (2005)
STUDY OF PEROVSKITES OBTAINED BY THE SOL-GEL METHOD
LILIAN BOUYSSIÉRES1*, RENÉ SCHIFFERLI1 , LORETO URBINA1, PAULO ARAYA2, JOSÉ MARIA PALACIOS3.
1 Facultad de Química y Biología / Universidad de Santiago de Chile, Casilla 40, Correo 33, Santiago, Chile.
Perovskites are very thermally stable mixed oxides, frequently used as oxidation catalysts in spite of their relatively low surface area. In order to increase the area the sol-gel method has been used, with malic acid as complexing agent and pH adjustment of the starting solution, to prepare the LaCoO3 perovskite, used as a reference, and other perovskites in which the La cation has been partially replaced (20% molar) by Ca, Sr or Ba cations. Additionally, the cooling rate of the fresh catalysts, from the calcination temperature to room temperature, has been used as a parameter in the method of preparation.
The catalysts were characterized by BET area, DRX, TG, DTA, and SEM-EDX. Catalytic activity was tested in a fixed bed flow reactor using the oxidation of CO with O2 reaction.
The DRX studies confirmed the presence of a single phase with perovskite structure in the catalysts calcined at 550 C, and the additional studies confirmed that very uniform catalysts with great thermal stability were obtained. Substitution of La by group II cations and the method of preparation make it possible to obtain perovskites with large surface area and excellent catalytic activity for CO oxidation, reaching maximum conversion values, close to 100%, at temperatures of about 200 C. Changes in post-calcination cooling rate did not produce significant changes in the surface area or in the catalytic activity of the catalysts, except when Ba-substituted catalysts were used.
Key words: perovskites, CO oxidation, perovskite preparation, sol-gel method
Perovskites are mixed oxides with an ABO3 pattern, where A typically corresponds to lanthanides, actinides, alkaline or alkaline earth metals, and B generally corresponds to transition metals (1). They are mainly used in partial oxidation reactions of methane to obtain synthesis gas (2-3), oxidation of CO (4-6), and hydrocarbon combustion among others (7-10). Their high thermal stability makes them appropriate for use as catalyst in catalytic converters (1,11-12) and for catalytic combustion in thermal plants, while their low surface area is the main limitation for their use. In order to overcome this problem the use of large surface area supports has been tried with unequal results (11-15), as well as the use of different preparation methods (11,16-18). The sol-gel method has allowed perovskites to be obtained at temperatures slightly above 500 C, i.e. about 300 C less than those used in the previous methods, leading to perovskites with larger surface areas.
Y.Teraoka et al. (19-20) have shown that using malic acid as the complexing acid and adjusting the pH of the starting solution it is possible to obtain the perovskites La0.8Sr0.2MnO3 and La0.8Sr0.2CoO3 at a lower calcination temperature, with the consequent increase in surface area with respect to an analogous preparation with citric acid.
In later studies the same authors found that of the La0.8Sr0.2MnO3 catalysts prepared with different complexing agents and used for the combustion of methane at 400 C, the one that had the largest surface area and greater catalytic activity (ca. 75% conversion) was that obtained from malic acid and calcined at 600 C.
In both experiments calcination time was five hours, and the authors found that as calcination temperature increases, surface area decreases, as has been widely reported. (1,18,21). However, C.K.Lee et al. (22) found that for the perovskite La0.6Ca0.4CoO3 prepared with citric acid, the surface area versus calcination temperature curves always show a maximum whose location depends on the different cooling rates used. For example, a maximum at 600 C is found for slow cooling in the furnace, whereas for faster cooling the maximum is located at 650 C, although their method differed from those used by other authors in that the calcination time was only two hours. The largest surface area, 33 m2/g, was obtained when calcination was done at 650 C for two hours, followed by quenching with water until room temperature.
The object of the present work was to study the variation of the surface area and its possible effect on the catalytic activity for CO oxidation of different La and Co perovskites prepared by the sol-gel method, using malic acid as complexing agent and adjusting the pH of the starting solution. In these perovskites, La is partially replaced in the lattice by 20% (mole ratio) of Ca, Sr or Ba. Calcination temperature and cooling rate to room temperature have been used as variables.EXPERIMENTAL
The study samples were prepared from mixed aqueous solutions of nitrates of the metal cations in their respective molar ratios. In all cases, a molar ratio of malic acid as complexing agent to total metal ions was set at 3:2 (19,20). The pH was adjusted to different values with a 28% w/w ammonia solution under vigorous agitation (19). The resulting solution was evaporated to dryness in a vacuum dryer, followed by calcination in a platinum crucible at 250 C for decomposition of the nitrates (4), and calcination in air at higher temperatures during 5 h for the formation of mixed oxides with perovskite structure. The prepared catalysts, in agreement with their nominal composition, were designated as LaCoO3, La0.8Ca0.2CoO3, La0.8Sr0.2CoO3 and La0.8Ba0.2CoO3. For the first one, calcination temperature was varied from 500 C to 800 C, whereas for the partially replaced catalysts calcination temperature was fixed at 550 C.
In all cases the product calcined at the corresponding temperature was cooled at different speeds by slow cooling in the furnace, immersing the outside of the crucible immediately in water at room temperature, or in liquid nitrogen (-195.8 °C). The pH of the starting solution was varied between 2.4 and 3.5 for the LaCoO3 perovskite (Tables I and II) and was kept constant at 3.5 for the replaced perovskites. The catalyst pellets were made under a pressure of 10 ton, ground and sifted, selecting the sizes between 16 and 20 mesh/cm2.
The BET areas were determined by adsorption of N2 at 77 K in an ASAP 2010 Micromeritics equipment. Catalytic activity for the oxidation of CO with O2 was measured in a fixed-bed flow reactor at atmospheric pressure, fixing the Gas Hourly Space Velocity (GHSV) at 2370 h-1 with 18% oxygen, 14% carbon monoxide and 68% nitrogen. A 1.3-g catalyst sample was used in each case, diluted with 16-20 mesh carborundum in a 1:1 volume ratio to avoid hot spots. This gave a volume of approximately one milliliter for the catalytic bed.
Analysis of incoming and outgoing reactor gases was made by gas chromatography in a Perkin Elmer 8500 chromatograph provided with a thermal conductivity detector and a 100/120 stainless steel Carbosieve SII, 10"x 1/8" column. Determination of CO percentage conversion was made from room temperature to 200 C, temperature at which the perovskites reached their maximum activity. Each measurement was made at least twice, and a third measurement was made in case of discrepancy between the other two.
The structural studies were made by X-ray powder diffraction (XRD) in a Seifert 3000 diffractometer using Cu Ka radiation filtered with Ni, and a graphite monochromator. The corresponding X-ray diffraction analyses were handled with Rietveld methods for the refinement of several structural parameters such as the degree of occupation of each site in the perovskite structure, the hexagonal network parameters a and c, and the mean size of the corresponding crystalline domain.
The thermal stability of the catalysts was determined by thermogravimetric analysis (TG) and differential thermal analysis (DTA), using a Netzsch STA 409 simultaneous thermal analyzer. The sample was heated from room temperature to 1000 C at a heating rate of 10 C/min in an air atmosphere, followed by cooling to room temperature at the same rate.
RESULTS AND DISCUSSION
Table I shows the BET areas of the LaCoO3 sample prepared at different calcination temperatures, with different precursor solution pH.
As expected, there is a clear decrease of the BET areas as calcination temperature increases, in agreement with the results of Teraoka et al. (19.20) and of many other authors (1,21), and in clear disagreement with the results obtained by Lee et al. (22). This behaviour may be explained considering that the formation of the perovskite structure requires a minimum calcination temperature and a minimum calcination time that depend, among other factors, on the preparation method (5,11). Thus K.S. Song et al. (5) found for the LaMnO3 perovskite a maximum in the surface area versus temperature curve when the catalysts were prepared by the coprecipitation method. However, when they used the spray decomposition method the curve followed the classical behaviour, i.e. surface area decreased as calcination temperature increased. Probably surface area increases with calcination temperature when the perovskite phase is not yet formed, due to insufficient calcination temperature and time to form the perovskite phase, and decreases by thermal sintering when heating continues after the perovskite phase has been formed. Additionally, Table I shows that surface area is little affected by the pH of the precursor solution. Values of areas obtained with citric acid as catalyst precursor and therefore calcined at 800 C are seen to compare with those prepared with malic acid. The same is shown in Table II.
This table also shows that for a calcination temperature of 550 C and at pH 2.4 the largest surface area is obtained when the cooling is done in water. It is probable that for higher calcination temperatures and cooling with water greater surface areas may be obtained, as Lee et al. (22) found when cooling a La0.8Ca0.4CoO3 catalyst from 650 C. However, it must be considered that the surface area of perovskites depends, among other factors, on the nature of A-site and B-site ions, their partial replacement, and their percentage substitution in the perovskite (1,5,23), as well as on the method of preparation (5,11,12,16-18,20), factors that in this work are different from those used by Lee et al. Thus, for example, K.S. Song et al. (5) obtained specific areas that go from 2.11 to 14.62 m2/g for the LaMnO3 perovskite prepared by spray decomposition. The specific areas of the same perovskites prepared by coprecipitation vary between 2.6 and 9.41 m2/g. When La is partially replaced by Ag, Sr or Ce, the areas vary between 16.34 and 18.69 m2/g. For our catalysts, the surface area of the LaCoO3 perovskite obtained at pH 3.5, slow cooling and calcined at 550 C is the same as that obtained by B. Delmon et al. (24), and the surface area of the La0.8Sr0.2CoO3 perovskite obtained at pH 3.5, slow cooling and calcined at 550 C is the same as that obtained by Y.Teraoka et al. (19) at the same pH and calcination temperature.
With very fast cooling with liquid nitrogen, the specific area decreases to values even smaller than those obtained with slow cooling.
Table III shows the BET areas of the replaced perovskites prepared at pH 3.5 and calcined at 550 C with different cooling rates. A clear increase in the surface area of the replaced perovskites with respect to the nonreplaced perovskite is seen, and within these there is a clear dependence on the type of substituting cation that is inversely proportional to the ionic radius of the cation.
The samples studied show x-ray diffractograms with intense reflections indicating that the compounds obtained are crystalline. Analysis of the difractograms corresponding to the LaCoO3 sample calcined at 500 C showed the presence of several crystalline phases, including the simple oxides of La and Co corresponding to the respective nitrates.
This indicates that decomposition of the nitrates used as precursor has taken place, but the calcination temperature has not been enough for the formation of the desirable mixed oxides with the perovskite structure. Diffractograms of the compounds calcined at 550 C shows the presence of only one crystalline phase with the structure of the LaCoO3 perovskite for all the pH impregnation values used, in agreement with the work of Teraoka et al. (19) with catalysts of La0.8Sr0.2CoO3. Consequently, all the replaced perovskites studied later were prepared using a calcination temperature of 550 C.
Structural studies by DRX
Profiles of the elements of interest carried out with scanning electronic microscopy with dispersive x-ray energy (not shown) indicate that the catalysts prepared at a calcination temperature of 550 C are homogeneous samples, therefore probably made of only one phase. X-ray diffractograms show that these materials consist of only one crystalline phase. For the unsubstituted catalyst the crystalline phase has a deformed perovskite structure: LaCoO3, SG=167:H, R-3c, a=0.5441 nm, c=1.3128 nm with La in position 6a, Co in 6b and O in 18e. The replaced perovskites show exactly the same crystalline structure, but the position and intensity of the corresponding reflections are slightly different. This allows the refinement of some structural parameters using Rietveld methods. The refined parameters were the degree of occupation of the respective Wyckoff sites, the hexagonal lattice parameters a and c of the perovskite, and the size of the crystalline domains.
Table IV shows the results of the structural study with the parameters obtained for the fresh catalysts and the used ones, (after the tests for activity in the CO oxidation reaction), prepared with different cooling rates.
In the first place it is seen that the real composition of the perovskites, deduced from the determination of the crystalline occupation degree, is not very different from that expected for the nominal composition deduced from the concentrations of the components used in the preparation. This, along with the apparent homogeneity of the samples deduced by SEM-EDX, implies that they are perovskites in which the Wyckoff 6a site occupied by La in the LaCoO3 perovskite has been replaced by Ca, Sr or Ba in the expected 20 mole% concentration.The substituent effect of the cation is translated into a modification of the lattice parameters, a and c, and globally into a change in the volume of the unit cell. The average volumes obtained for the unit cell, shown in Table IV, clearly indicate that the cationic substitution is translated into an increase of this volume as the radius of the substituting ion increases, from Ca to Ba. However, even though small variations are seen when comparing the results corresponding to samples prepared at different cooling rates or between the fresh samples and after their use in the catalytic activity tests, these changes are not sufficiently important nor consistent to be considered. It is therefore concluded that the substituent effect of Ca, Sr or Ba in the perovskite lattice causes an expansion or contraction of the lattice in agreement with the relative size of the ion with respect to La as the element replaced. However, there are no important structural changes derived either from the cooling rate used in the preparation of the catalyst or from the measurement of catalytic activity.
Thermo gravimetric and differential thermal analysis.
The TG and DTA studies in a thermal scan show that the prepared catalysts have great thermal stability without significant loss of mass over the whole range considered, andd neither are there any phase changes or loss of water (the sample was heated in air from room temperature to 1000 C at a rate of 10 C/min and cooled down at the same rate).
Catalytic activity tests.
We do not have an explanation yet for this behaviour, but the curve was verified several times and the catalyst has a lower catalytic activity than the rest of the replaced catalysts. In the same graph it can be seen that the LaCoO3 catalyst has lower catalytic activity than all the replaced catalysts, except for the Ba substituted catalysts at a reaction temperature higher than 170 C.
On the other hand, it is necessary to emphasize the relatively low temperature, 200 C, at which all of them reach 100% conversion (except for the barium catalyst, which reaches 75% conversion), with results similar to those obtained by N.K. Labhsetwar et al. (12) using LaRuO3 catalysts.CONCLUSIONS
From the results obtained we may conclude that the sol-gel method, using malic acid as complexing agent and with regulation of the initial solution pH, makes it possible to obtain homogenous compounds, well crystallized and with a deformed perovskite structure in which the La of the crystalline lattice is partially replaced by elements of group II using relatively low calcination temperatures. It makes it possible to obtain perovskites of larger surface area and great catalytic activity in the oxidation of CO at relatively low temperature. The cationic substitutions promote expansion or compression of the perovskite lattice correlated with the corresponding ionic radius of the substituent ion that is reflected in the surface area and in the catalytic activity, unlike the cooling rate whose structural, textural and, consequently, catalytic effects are less significant.ACKNOWLEDGEMENTS
Financial support by the Dirección de Investigaciones Científicas y Tecnológicas of the Universidad de Santiago de Chile, Project number 059942SD and 050042BM, is gratefully acknowledged. Special thanks to BASF (Chile) for the donation of the silica used in this research.REFERENCES
1. T. Nitadori, T. Ichiki and M. Misono, Bull. Chem. Soc. Japan 61 (1988) 621 [ Links ]
2. J. Marcos, R.H. Buitrago and E.A. Lombardo, J. Catal.105 (1987) 95 [ Links ]
3. M.A.Ulla, R.A.Migone, J. O. Petunchi and E. A. Lombardo, J. Catal.105 (1987) 107 [ Links ]
4. Y. Teraoka, H. Nii, S. Kagawa, K. Jansson and M. Nigren, J. Mat. Chem. 6 (1996) 97 [ Links ]
5. K.-S. Song, H.X.Cui, S.D. Kim, S.K. Kang, Catal. Today 47 (1999) 155-160 [ Links ]
6. P. Ciambelli, S. Cimino, G. Lasorella, L.Lisi, S. De Rossi, M. Faticanti, G. Minelli, P. Porta, Appl. Catal. B: Environmental 37 (2002) 231-241. [ Links ]
7. G. Saracco, F. Geobaldo, G. Baldi, Appl. Catal. B: Environ. 20 (1999) 277. [ Links ]
8. S. Ponce, M. A. Peña, J.L. Fierro, Appl. Catal. B:Environ. 24 (2000) 193 [ Links ]
9. C. Batiot-Dupeyrat, F. Martínez-Ortega, M. Ganne, J.M. Tatibouët, Appl. Catal. A: Ge. 206 (2001) 205. [ Links ]
10. P. Ciambelli, S. Cimino, S. De Rossi, L.Lisi, G. Minelli, P. Porta, G. Russo, Appl. Catal. B: Environ. 29 (2001) 239. [ Links ]
11. N. K. Labhsetwar, A. Watanabe, T. Mitsuhashi, Appl. Catal. B: Environmental 40 (2003) 21-30 [ Links ]12. N. K. Labhsetwar, A. Watanabe, R. B. Biniwale, R. Kumar, T. Mitsuhashi, Appl. Catal. B: Environmental 867 (2001) 1-9. [ Links ]
13. G. Saracco, G. Scibilia, A. Ianibello and G. Baldi, Appl. Catal. B: Enviromental 8 (1996) 229 [ Links ]
14. J. N. Nudel, B. S. Umansky and E. A. Lombardo, Appl. Catal. 31 (1987) 275-289 [ Links ]
15. S.V.Nguyen, V.Szabo, D. Trong On, S. Kaliaguine, Microporous and Mesoporous Materials 54 (2002) 51-61 [ Links ]
16. H. M. Zhang, Y. Teraoka, N. Yamazoe, Chemm. Lett. (1987) 665. [ Links ]
17. Y. Teraoka, H.M. Zhang, N. Yamazoe, Proc. 9th Int. Congr. on Catal., Calgary, 1988, The Chemical Institute of Canada, Ottawa, 1988, p. 1984. [ Links ]18. Y. Shimizu, K. Uemura, N. Miura, N. Yamazoe, Chem Lett., (1988) 1978. [ Links ]
19. Y. Teraoka, H. Kakebayashi, I. Moriguchi, and S. Kagawa, Chemistry Letters (1991) 673-676 [ Links ]
20. Y. Teraoka, H. Kakebayashi, I. Moriguchi, and S. Kagawa, J. of Alloys and compounds, 192 (1993) 70-72 [ Links ]
21. D. Trong On, S. V. Nguyen, S. Kaliaguine, Phys. Chem. Chem. Phys. 2003,5, 2724-2729 [ Links ]
22. C. K. Lee, K. Striebel, Frank R. McLarnon, and J. Cairns, J. Electrochem. Soc., Vol 144, No 11, Nov. 1997 [ Links ]
23. S.-T. Shen, H.-S. Weng, Ind. Eng. Chem. Res. 37 (1998) 2654- 2661 [ Links ]
24.- M. Alifanti, R. Auer, J. Kirchnerova, F. Thyrion, P. Grange, B.Delmon, Appl. Catal. B: Environmental 41 (2003) 71-81. [ Links ]