Boletín de la Sociedad Chilena de Química
versión impresa ISSN 0366-1644
Bol. Soc. Chil. Quím. v.47 n.2 Concepción jun. 2002
Bol. Soc. Chil. Quím., 47, 191-197 (2002)
EFFECT OF CHLORINE PRECURSOR IN SURFACE AND
CATALYTIC PROPERTIES OF Fe/TiO2 CATALYSTS
1 Facultad de Ciencias Químicas, Universidad de Concepción, Casilla 160-C, Concepción,
2 Universidad Autónoma Metropolitana-Iztapalapa, 09340 México, D.F. A.P.55-534.
3 Instituto de Catálisis y Petroleoquímica, CSIC, Cantoblanco, 28049 Madrid, Spain.
(Received: July 24, 2001 - Accepted: April 8, 2002)
Titania-supported iron (1wt%) catalysts were prepared by the sol-gel method using different gelation pH (3 and 9), metal precursors (FeCl2 and FeCl3) and calcination temperatures (873 and 1073K). Characterization data of calcined catalysts revealed that in all samples the dominant iron species is Fe3+ and the crystalline phase of the TiO2 substrate depends on the gelation pH and the metal precursor used. It was found that in the Fe/TiO2 ex-FeCl3 samples an important part of the iron ions became inserted in titania lattice, which makes the anatase phase stable even at high calcination temperatures. The solids were tested in the combustion of methane under stoichiometric conditions at atmospheric pressure. It was observed that the Fe/TiO2 ex-FeCl2 sample displays the highest activity. This behaviour is explained on the basis of the highest fraction of iron oxide deposited on the external surface of the catalyst particles.
KEYWORDS: Iron catalysts, sol-gel, titanium oxide, combustion, methane.
Se prepararon catalizadores de hierro soportados sobre titania utilizando el método sol-gel y contenido de metal de 1% en peso. Se utilizaron diferentes pH de gelación (3 y 9), precursores metálicos (FeCl2 y FeCl3) y temperaturas de calcinación (873 y 1073K). Los resultados de caracterización indican que en todos los casos el hierro está presente fundamentalmente como Fe3+ y la fase cristalina de titanio presente depende del pH de gelación y de la naturaleza del precursor utilizado. Se encontró que cuando se usa FeCl3 como precursor de hierro, hay una mayor inserción de iones hierro en la red de la titania que estabiliza a la fase anatasa aún a altas temperaturas de calcinación. Se midió la actividad catalítica de los sólidos preparados en la reacción de combustión de metano en relaciones estequiométricas metano/oxígeno y a presión atmosférica. Los catalizadores preparados a partir de FeCl2 mostraron la mayor actividad. Este comportamiento se explica considerando que en estos catalizadores la mayor parte del hierro se encuentra en la superficie externa del sólido presentando sólo una pequeña fracción de iones hierro insertos en la red de titanio.
PALABRAS CLAVES: catalizadores de hierro, sol-gel, óxido de titanio, combustión, metano.
The synthesis of metallic oxides by using the sol-gel method is a widely accepted procedure for their preparation because it provides an attractive alternative to obtain catalysts in a single preparation step (1). Additionally, controlled porosity and high resistance to deactivation (2, 3) are among others desirable characteristic of these materials. The method implies basically the solubilization of a metal alkoxide precursor in an organic solvent, which is then hydrolyzed with a controlled amount of water. Acids or bases catalyze the hydrolysis reaction, and the final properties of the solids depend on the gelation pH, calcination temperature and other factors such as: (i), the hydrolysis catalysis (4), (ii), doping by deliberately adding impurities before gelation, (iii), the metal precursor (5). Furthermore, the sol-gel method is widely used because it allows the incorporation of metallic ions into the gels immediately before of their formation remaining them in a highly dispersed state (6, 7).
Important advantages of some properties of the solids obtained by the sol-gel method compared with those prepared by more conventional methods have been reported. High purity of the solids, lower temperatures required for their synthesis, homogeneous distribution of the active phase are, among others, the most noticeable advantages (8-10). The calcination temperature plays an important role, particularly in the case of crystalline solids such as zirconia (11) and titania (12), in which different crystallographic phases are developed within a rather narrow range temperature. Titania exhibits a wide range of applications as photocatalyst (13), support in heterogeneous catalysis (14) and also as electrode in electrochemistry. Stoichiometric titania is a dielectric material, but it turns into a semiconductor when their chemical, electrical and catalytic properties are modified by the presence of metal ions. Thus, titania exposure to a reducing atmosphere or doping it with metal ions (Mn+) in a valence state different than 4+ may result in oxygen vacancies or in the reduction of Ti from +4 to +3 (15, 16). At normal pressures titania can exhibit three different crystalline phases brookite, anatase and rutile, all phases in which the Ti ions are placed inside a distorted octahedron of oxygen ions. The number of sharing edges of these octahedra allow to distinguish the different crystalline phases. Thus, three octahedra edges are shared in brookite, four in anatase and two in rutile (17). As a consequence of this crystallographic ordering the different mass density varies among these crystalline phases. The phase transformations between brookite, anatase and rutile depend on the thermal treatment procedure. As the calcination temperature increases, irreversible transformation from brookite or/and anatase into rutile occurs (18), although anatase can be stabilized by incorporating dopant cations during the preparation (19, 20). It should be mentioned that brookite phase has not been detected at calcination temperatures higher than 873K.
As the properties of titania depend on the crystalline structure (21), the understanding of the catalytic behaviour of Fe/TiO2 system requires not only the identification of the crystalline phases of titania but also their quantification. For iron supported catalysts, both catalytic performance and surface structures have been studied. It has been demonstrated that the iron cations are almost saturated by adventitious oxygen after calcination and removal of these oxygen ions is difficult due to the geometric stability of the octahedrally coordinated iron sites (22-24). Differents metal or metal oxides supported catalysts have been used for the complete combustion of organic compounds (VOCs). It has been observed that Pd an Pt show the best performance (25, 26). However, the high sensitivity of the metal phase to deactivation in the presence of water or sulphur compounds present in the feed streams or by sintering of the metallic phase, attempts have been made to modify noble metal catalysts or to use other less expensive metal oxides. Suported Cu, Mn, Co and Fe oxide catalysts have been studied for the combustion of CO and hydrocarbons (27). Eventhough, these systems are more strongly deactivated by water and sulphur poisoning, the low cost of the catalytic ingredient provides an attractive way for VOCs abatement. Cooper and iron oxides exhibited the highest activity in CO oxidation in the latter series (28-30).
When used as photocatalyst, the Fe/TiO2 system was investigated by looking at the influence of the preparation technique, calcination temperature and metal loading oxide (31, 32) on the catalytic activity. It has been pointed out that the preparation technique seems to be one of the most important factors in determining the catalytic activity (33) specially if it is related with the crystallographic phase of the titania through modifications in the band gap energy. In line with the above, this work was undertaken with the aim to study the effect of some preparation variables on the surface structure and catalytic properties of a Fe/TiO2 sol-gel catalysts. Characterization data of the samples were obtained by nitrogen adsorption isotherms at 77 K, UV-vis spectroscopy, temperature-programmed reduction and photoelectron spectroscopy and the catalytic activity measured in the combustion of methane under stoichiometric condition in a wide temperature range.
Fe/TiO2 catalysts containing 1 wt.% of iron as metal loading were prepared by the sol-gel method. The catalysts were prepared in one step in a reflux system at a constant temperature of 333 K at two different gelation pH values (3 and 9), using HCl and NH3(aq) as hydrolysis catalysts and two different chloride metal precursors: iron (II) chloride tetrahydrate and iron (III) chloride hexahydrate. For each preparation, titanium (IV) butoxide dissolved in absolute ethanol was dropped, stirred and refluxed continuously until gelling. All the fresh samples were dried overnight at 333 K and calcined in air at 873 K and 1073 K for 4 h.
Specific area and porosity were determined from the N2 adsorption-desorption isotherms. These isotherms were recorded with an automatic Micromeritics system ASAP 2001, using nitrogen gas as adsorbate at the liquid nitrogen temperature in the 0.050.995 relative pressure range. Temperature-programmed reduction (TPR) experiments were carried out in a TPR/TPD 2900 Micromeritics system provided with a thermal conductivity detector. A reducing gas mixture of 5% H2/Ar (40 cm3 min-1) at a heating rate of 10 K min-1 was employed. X-ray diffraction (XRD) patterns were recorded in a Rigaku diffractometer using a Ni filter and Si, and Cu Ka1 radiation. Photoelectron spectra (XPS) were recorded using an Escalab 200R electron spectrometer with a Mg Ka X-ray radiation (hn =1253.6 eV) operated at 10 mA and 12 kV. The binding energy of Ti 2p3/2 at 458.5 eV was used as an internal standard. Fe 2p3/2 and Ti 2p3/2 peak areas were calculated by fitting the experimental curves to Gaussian and Lorentzian shapes. The Fe/Ti and O/Ti surface ratios were estimated from the intensity of the peaks after background subtraction and corrected by the atomic sensitivity factors (34).
The catalytic activity in the combustion of methane was evaluated in a conventional flow reactor operated at atmospheric pressure using 100 mg of catalysts, space velocity of 6000 h-1 a flow of a mixture containing 1% CH4, 2%O2 and He as balance at a temperature range 423 K to 973 K. A heating rate of 1K min-1 was used. The effluents of the reactor were analyzed by an on-line gas chromatography. A single column containing molecular sieve (5A) was used and the chromatographic separation was carried out isothermally at 333 K with helium as carrier gas. In some experiments a quadrupole mass spectrometer Hiden HPT 20 was used to detect small traces of products.
RESULTS AND DISCUSSION
Table I compiles the TPR results of the prepared catalysts. It can be seen that the reduction of iron oxides of Fe/TiO2 ex-FeCl2 samples takes place at lower temperatures than in their Fe/TiO2 ex-FeCl3 counterparts. All the samples exhibit only a single peak centered at 650 and 750 K for the catalysts calcined at 873 and 1073 K, respectively. With regard to the H2 consumption during TPR, no significant changes were detected with the gelation pH at a given calcination temperature. However, a significant decay in the H2 consumption was observed as calcination temperature increases. In all the samples the H2 consumption was lower than that required for the stoichiometric reduction. The lowest H2 consumption was detected for all the samples calcined at 1073 K. This feature is explained considering that after calcination at 1073 K, iron ions may be inserted into the titania lattice making the reduction process more difficult.
Adsorption-desorption isotherms at 77 K were performed on all the studied catalysts in order to evaluate surface area and pore size distribution. Figure 1 diplays the adsorption-desorption isotherms for a representative catalyst calcined at 873 and 1073 K. Significant differences in N2 uptake and also in the histerises loops were observed by varying calcination temperature. Similar behaviour has been observed for all the studied samples. Table I compiles the specific area and average pore radius obtained from these data. It can be seen that the catalysts obtained from FeCl3 calcined at 873 K display the highest surface area, between 50 and 70 m2g-1, whereas slightly lower areas, in the range 41-46 m2g-1, are shown by the Fe/TiO2 ex-FeCl2 catalysts calcined under the same conditions. All the samples calcined at 1073 K show very low surface area, in the range 3-5 m2g-1, indicating sinterization of the metal oxide by the high temperature treatment. In parallel with this change, an increase in the pore radius was observed upon increasing calcination temperature, ranging from 12-16 nm for the samples calcined at 873 K and 22-24 nm for those calcined at 1073 K.
Figure 1. Nitrogen adsorption isotherms at 77 K for: (a) FeCl2-3-873; (b) FeCl2-3-1073.
The values of the forbidden gap Eg are also compiled in Table I. These values were calculated using the value of the wavelength at the onset of absorption in the electronic spectra using the equation Eg = hn - a2/A2, where Eg is the energy of the band gap, a the absorption coefficient and A the absorbance. It can be seen that the Fe/TiO2 ex-FeCl2 samples calcined at 873 K show the lower band gap values. This fact is related to the catalytic behavior of these systems as discussed below.
Previous studies dealing with the effect of the gelation pH and calcination temperatures on the crystallinity of TiO2 revealed that at the same calcination temperature the relative proportion of anatase-to-rutile is highly dependent of the the gelation pH. As it is well known, the formation of rutile phase is favoured at higher temperatures whereas brookite and anatase are usually present at lower calcination temperatures. Thus, it has been reported that, at the calcination temperature of 873 K, the extent of anatase phase varies from 5 to 94% when the gelation pH changes from 3 to 9 (18). This behaviour is caused by the higher proportion of hydroxyl groups in the solid gelated in basic medium which contribute to stabilize the anatase phase. It is also documented that an increase in the calcination temperature yields a higher proportion of rutile whatever the gelation pH.
Figure 2 displays the X-ray diffraction patterns of a representative Fe/TiO2 catalyst calcined at 873 and 1073 K. Significant differences in the XRD difractograms may be noted. In fact, those catalysts calcined at 873 K exhibit the main diffraction lines corresponding to anatase (2q =27.46° giving an interplanar distance of 3.245 Å) whereas in that calcined at 1073 K the main lines are those corresponding to rutile (2q = 25.35° giving an interplanar distance of 3.51 Å). No hints of the presence of iron oxide-related phases, for which the most intense peaks appear in zones where no significant interference with TiO2 anatase or rutile peaks is produced (in 2q values, at 33.2-33.5o for Fe2O3JCPDS 84-0311 and 79-1741- or at 35.4o for magnetiteJCPDS 79-0418-), were detected in all the samples.
Figure 2. X-ray diffraction patterns for: (a) FeCl3-9-873; (b) FeCl3-3-873; (c) FeCl3-9-1073 ; (d) FeCl3-3-1073.
As it can be seen in Table I, all the Fe/TiO2 samples calcined at 873 K exhibit a higher proportion of anatase, which may account for the stabilization of this phase. Reasons for this can be: (a), presence and nature of iron ions, (b), gelation pH, and (c), calcination temperature.
Fe/TiO2 catalysts calcined at 873 K prepared by differents iron precursors show differences in the anatase extent, indicating stabilization of the anatase phase by partial insertion of the iron ions during the gelation step. This explanation may be supported by analyzing the (100) plane, which usually appears at d value of 3.510. However, a slight increase in the d values was found for all the catalysts, indicating insertion of iron in the TiO2 lattice. In the Fe/TiO2-ex FeCl3 catalysts the insertion of iron ions into the lattice is more likely because of its similar ionic radius. In fact, taking into account that the ionic radius of Fe2+, Fe3+ and Ti4+ are 8.3, 6.7 and 6.4 nm, respectively, the insertion of Fe3+ ions in the titania lattice during the synthesis seems more likely. Therefore, a higher stabilization of the anatase phase may be produced. On the other hand, an increase in the proportion of anatase with the gelation pH is observed in the Fe/TiO2 samples calcined at 873 K, thus confirming that the hydroxyl groups contribute to the stabilization of this phase. Using FeCl2 as iron precursor, the extent of anatase phase formation varies from 61 to 94% at gelation pH of 3 and 9, respectively, whereas when FeCl3 is used, the extent of anatase formation is higher, and varies between 91 to 100% at gelation pH of 3 and 9. For both precursors, the higher proportion of anatase in the samples gelated at pH 9 is attributed to the higher proportion of hydroxyl group that stabilize the anatase phase during the synthesis (26). This is explained in terms of the larger size of the octahedral channels present in the anatase phase that can easily accommodate surface hydroxyls groups. Similarly, the effect of the calcination temperatures shows the same expected trend to that displayed by pure titania, at 1073 K. Rutile is the only observed phase for the samples gelated at pH = 3, whereas only a slight contribution (lower than 5%) of the anatase phase is present in the samples gelated at pH = 9. These results suggest that, for all the studied catalysts, a large iron dispersion is obtained in all the catalysts because iron ions became incorporated into the titania network.
The Fe 2p3/2 core-level spectrum of a representative catalyst is shown in Figure 3 and the corresponding binding energies of the most intense Fe2p3/2 level are summarized in Table II. In all the studied catalysts the mean peaks related with iron (III) species were detected at 711.2 eV, whereas the component of Fe 2p3/2 core level of the studied samples are wider and slightly shifted to lower BE suggesting the presence of a small proportion of Fe2+ ions. It can be seen that the (Fe/Ti) surface ratio is comparable with the (Fe/Ti) bulk ratio (0.014) only for Fe/TiO2-ex FeCl3 catalysts calcined at 873K. This fact can be interpreted as due to a rather homogeneous distribution of iron incorporated into the TiO2 lattice. For the Fe/TiO2 ex-FeCl2 calcined at 873K samples the surface Fe/Ti ratio is higher than the bulk one, suggesting that in these catalysts the iron enrichment at the surface took place, and therefore a lower proportion of iron ions were incorporated into the lattice with respect to the Fe/TiO2 ex-FeCl3 samples. This fact is also supported by the XRD results for the catalysts prepared using FeCl3 in which a partially stabilization of the anatase phase was detected. For the samples calcined at 1073 K, the high (Fe/Ti) surface ratio is essentially due to a decrease of the surface area.
Figure 3. Fe 2p3/2 core level spectra of FeCl3-3-873 catalyst.
The activity data for methane oxidation under stoichiometric conditions (CH4:O2 = 1:2 molar) were evaluated as a function of the reaction temperature up to 973 K. As expected, the combustion curves displayed a S-shaped profile (29). The conversion groups up slowly (at temperature about 600 K) and then increases drastically as the temperature increases. The activity measurements were carried out under both, reduced and oxidic form, and no differences in the catalytic behavior were observed. Table II compiles the temperatures required to obtain 20 and 50% of CH4 conversion. This latter is defined as the ignition temperatures (Ti50). The highest activity is exhibited by the catalysts calcined at 873 K specially those prepared from FeCl2 precursor. This behavior may be explained considering a high proportion of surface iron oxide species in these solids, as previously discussed. For all the catalysts calcined at 1073 K the catalytic activity is lower than the corresponding at 873K. This feature is attributed to a large decrease in the surface area of these catalysts, due to sintering of particles at high temperatures. This fact reduces significantely the catalytic activity and therefore almost no changes in the activity with other variables except of the calcination temperature should be expected.
The nature of the metal precursor and the gelation pH affect both the structure and catalytic activity of 1wt% Fe/TiO2 sol-gel catalyst. It was found that insertion of iron ions into the lattice of the titania inhibits the appearance of rutile phase at lower calcination temperature and stabilization of anatase form at high calcination temperatures. Iron sol-gel catalysts do not show any diffraction line attributed to iron oxide species indicating the incorporation of the iron ions in the titania lattice or a high dispersion of the iron oxide phase. Catalytic results indicate that isolated iron oxides species display high activity in the combustion of methane mainly in those catalysts calcined at lower temperatures.
The authors thank FONDECYT, Grant 1990484 for the financial support.
1. T.López, E. Sánchez, React. Kinet. Catal. Lett., 48 (1992) 295. [ Links ]
2. G. Pecchi, M. Morales, P. Reyes, React. Kinet. Catal. Lett., 61 (1997) 237. [ Links ]
3. R.D. González, T. López, R. Gómez, Catal. Today, 35 (1997) 293. [ Links ]
4. W. Zou, R.D. González, T. López, R. Gómez, Mater. Letters, 24 (1995) 35. [ Links ]
5. T. López, E. Sánchez, P. Bosvh, Y. Meas, R. Gómez, Mater. Chem. Phys. 32 (1992) 141. [ Links ]
6. Y.I.Yermakov, B.N. Kutznetsov, J. Mol. Catal., 9 (1980) 13. [ Links ]
7. P. Reyes, M. Morales, G. Pecchi, J.L.G.Fierro, Bol. Soc. Chil. Quím., 41 (1996) 221. [ Links ]
8. G. Pecchi, P. Reyes, F. Orellana, T. López, R. Gómez, J.L.G.Fierro, J. Chem. Tech. Biotechnol., 74 (1999) 1. [ Links ]
9. H. Kominami, Y. Takada, H. Yamagiwa, Y. Kera, M. Inoue, T. Inui, J. Mater. Sci. Lett., 15 (1996) 197. [ Links ]
10. C. J. Brinker, G. W. Scherer, Sol-gel science. The physics and chemistry of sol-gel procesiing, Academic Press, New York, 1990. [ Links ]
11. G. Pecchi, P. Reyes, R. Gómez, T. López, J.L.G.Fierro, Appl. Catal. B, 17 (1998) L7. [ Links ]
12. R.J.H. Clark,The Chemistry of Titanium and Vanadium, Elsevier, New York, Chap. 9 (1968). [ Links ]
13. G. Pecchi, P. Reyes, P. Sanhueza, J. Villaseñor, Chemosphere, 43 (2000) 141. [ Links ]
14. C. Cristiani, M. Belloto, P. Forzatti, F. Bregani, J. Mater. Res., 8 (1993) 2019. [ Links ]
15. B. E. Yoldas, J. Mater. Sci., 21 (1986) 1087. [ Links ]
16. H. Al-Ekabi, N. Serpone, J. Phys. Chem., 92 (1988) 5726. [ Links ]
17. L. Pauling, J. Am. Chem. Soc., 51 (1929) 1010. [ Links ]
18. A. Bokhimi, A. Morales, O. Novaro, T. López, E. Sánchez, R, Gómez, J. Mater. Res., 10 (1995) 2788. [ Links ]
19. R. D. Shanon, J. A. Pask, J. Am. Ceram. Soc., 48 (1965) 391. [ Links ]
20. V.T. Zaspalis, W. van Praag, K. Keizer, J.R.H. Ross, A.J. Burggraaf, J. Mater. Sci, 27 (1992) 1023. [ Links ]
21. X. Bokhimi, J.L. Boldú, E. Muñoz, O. Novaro, T. López, J. Hernández-Ventura, R. Gómez, A. García-Ruiz, Chem Mat., 11 (1999) 2716. [ Links ]
22. B. J. Tatarchuk, J. A. Dumesic, J. Catal., 70 (1981) 308. [ Links ]
23. B. J. Tatarchuk, J. A. Dumesic, J. Catal., 70 (1981) 323. [ Links ]
24. B. J. Tatarchuk, J. A. Dumesic, J. Catal., 70 (1981) 335. [ Links ]
25. R. Prassad, L. Kennedy, E. Ruckenstein, Catal. Rev. Sci. Eng. 26 (1984) 1. [ Links ]
26. G. Pecchi, P. Reyes, I. Concha and J.L.G. Fierro, J. Catal., 179 (1998) 309. [ Links ]
27. P. Reyes, A. Figueroa, G. Pecchi, J.L.G.Fierro, Catal. Today, 62 (2000) 209. [ Links ]
28. J. J. Spivey, J.B. Butt, Catal. Today, 11 (1992) 465. [ Links ]
29. P.-O. Larsson, H. Berggren, A. Andersson, O. Augustsson, Catal. Today, 35 (1997) 137. [ Links ]
30. A. S.C. Brown, J. S. J. Hargreaves, B. Rijniersce, Topics in Catal., 11/12 (2000) 181. [ Links ]
31. M.A.Vannice, R.L. Garten, J. Catal., 63 (1980) 255. [ Links ]
32. Y. Kou, Z. Suo, J. Niu, W. Zhang, H. Wang, Catal. Lett., 35 (1995) 271. [ Links ]
33. M.I. Litter, J. A. Navío, Photochem. Photobiol. A, 98 (1996) 171. [ Links ]
34. C.D. Wagner, L.E. Davis, E.V. Zeller, J.A.Taylor, R.H. Raymond, H.L. Gale, Surf. Interf. Anal., 3 (1981) 211. [ Links ]
·To whom correspondence should be addressed. Email: firstname.lastname@example.org