versión impresa ISSN 0366-1644
Bol. Soc. Chil. Quím. v.45 n.2 Concepción jun. 2000
CATALYTIC COMBUSTION OF TOLUENE ON
1: Facultad de Ciencias Químicas, Universidad de Concepción,
Casilla 160-C, Concepción, Chile.
2: Instituto de Catálisis y Petroleoquímica, CSIC, Madrid, Spain
(Received: November 10, 1999 - Accepted: February 23, 2000)
In memoriam of Doctor Guido S. Canessa C.
Catalytic combustion of toluene on silica supported Pd, Pd-Cu and Cu catalysts has been investigated. The solids were prepared by the impregnation method and they were characterized by evaluation of surface area, metal dispersion and surface composition. The reaction was studied under stoichiometric conditions in the presence and absence of thiophene as poison. It was found that the presence of copper reduces the activity due to a copper-palladium interaction, which modifies the nature of the active site. The monometallic Pd and the bimetallic Pd-Cu catalysts showed high resistance to the sulfur poisoning.
KEY WORDS: Palladium, copper, combustion, toluene, catalysis.
Se estudió la combustión catalítica de tolueno sobre catalizadores de Pd, Pd-Cu y Cu soportados sobre sílice preparados por el método de impregnación. Los sólidos se caracterizaron por medidas de superficie específica, dispersión metálica y composición superficial. La actividad catalítica de ellos se midió en la reacción de combustión de tolueno realizada en condiciones estequiométricas de oxígeno en ausencia y presencia de un veneno azufrado, como tiofeno. Se encontró que la presencia de cobre disminuye la actividad del paladio debido a una modificación electrónica paladio-cobre, que modifica la naturaleza del sitio activo. El catalizador monometálico de Pd y los bimetálicos Pd-Cu muestran una alta resistencia al envenenamiento por el compuesto azufrado.
PALABRAS CLAVES: Paladio, cobre, combustión, tolueno, catálisis.
The catalytic combustion is one of the several industrial processes for pollution abatement. Catalysts able to perform combustion are divided in two groups: noble metals for which reactions may start at temperatures as low as room temperature and transition metal oxides which are less efficient, but also more resistant towards high temperatures. With regard to the noble metal supported catalysts, most of the studies have been focussed on platinum and palladium 1,2). The superiority of palladium as a catalyst for the complete oxidation to carbon dioxide and water has been well known for years 3). The industrial emissions are normally carried out in conditions in which the catalysts are exposed at high temperatures and also to the presence of water vapor and sulfur compounds in the feed. It has been reported that water vapor may induce the migration of the metal crystallites, leading to the sinterization of the active phase. This may be minimized using the sol gel method in the preparation of the metal supported catalysts 4,5). It is also widely reported that sulfur compounds may deactivate the active phase by an irreversible adsorption of the sulfur compounds on the metal. Different efforts have been made in order to reduce the metal deactivation, such as, changing the support, the precursors, by addition of some promoters and by adding a second metal having large ability to chemisorb the poison.
In this work it has been chosen the couple Pd-Cu because of the excellent properties of Pd as total combustion catalysts 6,7) and Cu as second metal. It has been studied the effect of Pd/Cu ratio in catalysts containing 0.5wt.% Pd and different Cu loading (0.3 up to 1.2 wt%). The catalysts have been characterized by nitrogen adsorption at 77K, H2 chemisorption measurements at 343 K, TPR, TEM studies and XPS experiments. The activity of the catalysts was measured in the combustion of toluene at different temperatures from 373 to 673 K with a stoichiometric amount of oxygen in the presence and absence of thiophene.
The Pd/SiO2, Pd-Cu/SiO2 and Cu/SiO2 catalysts containing 0.5wt% Pd and variable amount of Cu ranging from 0.3 to 1.2 wt% were prepared by the impregnation of an aqueous solution of copper nitrate on a commercial SiO2 (BASF D11-11, SBET= 154 m2g-1) at 308 K and calcined in air at 673 K for 4 h. When necessary, the Pd was subsequently impregnated with a toluene solution of Pd(acac)2 at the same temperature of impregnation. The reduction was carried out in situ for 1 h at 773 K.
Surface area and porosity were obtained from nitrogen adsorption at 77 K in a Micromeritics Model Gemini 2370 and hydrogen chemisorption measurements at 343 K in a greaseless volumetric system to evaluate the hydrogen uptake and H/Pd ratio. Prior the chemisorption experiment, the samples were reduced in situ at 773 K for 2 h and then outgassed for 4 h at the same temperature. TEM was used for the observation of the supported palladium particles and performed in a JEOL Model JEM-1200 EXII system. The samples were prepared by the extractive replica procedure. TPR experiments were carried out in a TPR/TPD 2900 Micromeritics system provided with a thermal conductivity detector. The reducing gas was a mixture of 5%H2/Ar (40 cm3min-1) and a heating rate of 10 Kmin-1. XPS studies were recorded using an Escalab 200R spectrometer provided with a hemispherical analyzer operated in a constant pass energy mode and unmonochromatized MgKa X-ray radiation (hn=1253.6 eV) operated at 10 mA and 12 kV. The surface Pd/Cu and Cu/Si ratios were estimated from the integrated intensities of Pd 3d5/2, Si 2p3/2 and Cu 2p3/2 after background subtraction and corrected by the atomic sensitivity factors 9). The line of Si 2p at 103.4 eV was used as an internal standard. Palladium and copper peaks were decomposed into several components assuming that the peaks had Gaussian-Lorentzian shapes.
The catalytic activity in the combustion of toluene was evaluated in a conventional flow reactor at atmospheric pressure using 200 mg of catalysts and a space velocity of 3000 h-1. The calcined samples were reduced in situ in flowing H2 (50 cm3 min-1) up to 773 K for 1 h. Then the samples were cooled down to 473 K and the reducing gas was switched to He as carrier. After 30 min of stabilization at this temperature, the carrier gas was switched to the reactant gaseous mixture containing O2:He = 20:80 (molar) which passes through a toluene bath at 273 K. The activity was measured at different temperatures from 423 up to the temperature required for a total conversion using a heating rate of 1 K min-1. Additionally experiments were carried out dopping the feed with 200 ppm of thiophene. The effluents of the reactor were analyzed by an on-line gas chromatograph. 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 HAL 20 was used to detect small traces of products.
RESULTS AND DISCUSSION
Nitrogen adsorption isotherms on the studied catalysts correspond to type IV in the BDDT classification's, with almost no change in the extent of adsorption with the metal loading as shown in Table 1.
|Catalyst||SBET||H/Pd||d, nm||d, nm TEM|
|m2g-1||Chem H2||Pd|| |
In fact, the surface area obtained for the catalysts is close to the support (154 m2/g). TPR experiments were run in the temperature range 195 to 775 K and the profiles of the calcined samples are displayed in Fig. 1.
|Fig. 1 Temperature programmed reduction |
profiles of Pd-Cu/SiO2 catalysts.
In all samples a single peak centered on 293 K which may be attributed to the reduction of PdO is observed. The fact that this reduction process occurs at temperatures close to room temperature is indicative of well-dispersed Pd particles. Additionally, at about 350 K an inverse peak due to the decomposition of the palladium hydride is also observed. With regard to the reduction of copper species, a single peak centered at 520 K was observed for Cu/SiO2 catalyst, which is attributed to the reduction of Cu2+ to Cu°. This result is in agreement with those previously reported by Robertson et. al 11). For the bimetallic Pd-Cu catalysts, the reduction of copper particles takes place at lower temperatures due to that the reduced palladium particles may catalyze the reduction of copper oxides species by H2 spillover. At lower copper loading the Pd-Cu interactions are stronger, making easier the reduction of CuO particles. As Cu content increases, the position of the reduction temperature peak also increases, being close to that of the monometallic Cu catalyst. The hydrogen consumption is close to that required for the stoichiometric reduction of the metal components. The observed shift in the copper reduction may suggest the formation of Pd-Cu alloys in some extent.
Table 1 summarizes H/Pd ratio and the estimated particle size by chemisorption and TEM. Metal particle size obtained from H2 chemisorption at 343 K was evaluated assuming cubic metal particles in which one face remains on the support and the other five are exposed to hydrogen, by the equation d=5/Sr and a stoichiometry of adsorption H/Pds=1. In the equation, S represents the metal surface area and r the metal specific density. It can be seen almost no changes in Pd dispersion and the particle size is close to 2.0 nm. The small differences observed in the particle size estimated from chemisorption and TEM results may be a consequence of alloy formation in a slight proportion, remaining most of the palladium and copper particles as segregated phases. This is clearly shown in the micrographs obtained by TEM.
In fact, copper particles are larger than palladium ones in a factor of 2 to 10, which is expected considering the difficulties to get well-dispersed copper particles.
Figure 2A and 2B show the Pd 3d5/2 and the Cu 2p3/2 core-level spectra of the reduced catalysts. Only small changes in the binding energies for Pd 3d5/2 and Cu 2p3/2 were found. The results are compiled in Table 2
|Fig. 2 XPS spectra for Pd-Cu/SiO2 catalysts. |
A: Pd 3d5/2 core level B: Pd 2p3/2 core level
a) 0.5Pd/SiO2 or 1.0Cu/SiO2,
b) 0.5Pd-0.3Cu/SiO2,c) 0.5Pd-0.4Cu/SiO2,
d) 0.5Pd-0.6Cu/SiO2,e) 0.5Pd-1.2Cu/SiO2.
|Catalyst||Catalyst B. E., eV.||Surface Ratios,|
The B.E values of the Pd 3d5/2 are 335.2 and 337.1 eV for Pd° and Pd2+ species. In the monometallic Pd catalyst the observed value is in agreement with the expected value for Pd° species 12,13). Although no significant changes in the B.E. are observed in the Pd-Cu bimetallic catalysts, the full width at half maximum of Pd 3d peaks increased markedly from 3.0 eV in the monometallic Pd sample to 3.6 eV in the bimetallic 0.5Pd-1.2Cu sample, and this increase was less marked at intermediate Pd/Cu compositions. This finding may suggest a Pd-Cu interaction. With respect to the atomic surface ratios, the slight increases observed in the Pd/Si as copper content increases, suggest that Cu may induce a redispersion of Pd particles or because of a surface coverage of silica by copper. The Cu/Si ratio increases with Cu loading in similar proportion compared with the copper loading, indicating similar copper dispersion, in agreement with TEM results.
The catalytic activity for toluene oxidation as a function of the reaction temperature up to complete combustion is shown in Fig.3.
Fig. 3 Catalytic activity in the combustion of toluene in absence of thiophene for Pd, Pd-Cu and
In this figure, typical sigmoidal curves can be seen in which the reaction starts at about 520 K, then the conversion increases drastically as the temperature increases and a complete conversion is reached. Differences in the conversion level at a given temperature for catalysts having different Cu content are detected. In fact, as copper loading increases the curve shifted towards higher temperatures indicating lower activities. The ignition temperatures (Ti50), defined as the temperature required to obtain 50% of conversion, under stoichiometric conditions and in presence of thiophene are compiled in Table 3.
|Ti50, K||Ti50, K||TOF, 490 K|
|in absence||in presence||in absence|
|of thiophene||of thiophene||of thiophene|
The ignition temperatures are lower for the monometallic Pd catalysts, whereas in the bimetallic Pd-Cu system these values are almost constant. It can be seen that no great differences with the corresponding ignition temperature in absence of thiophene were found, indicating that Pd and Pd-Cu catalysts are resistant to the poisoning. The only exception is the monometallic Cu catalyst that has a small activity and a high sensibility to be poisoned by thiophene. This fact can be explained considering that during the oxidation reaction on the Pd and Pd-Cu catalysts the hydrogenolysis of thiophene can take place and the formed H2S transformed on sulfate, which is not a poison of these catalysts. This effect cannot occur in the monometalic Cu catalysts, because at low temperatures the hydrogenolysis of the thiophene does not occur and consequently, the sulfur compound is irreversible adsorbed on copper. Carbon dioxide and water were the only products of the reaction, no carbon monoxide and coke were detected probably due to the low acidity of the catalysts. In order to compare the activity per site, the turnover frequencies (TOF), expressed as number of toluene molecules converted per Pd surface atom per second, for toluene oxidation at 490 K were calculated. The TOF values are shown in Table III. The bimetallic catalysts have almost the same TOF, which are lower than the value exhibit for the monometallic palladium catalyst. This may be attributed to a change in the nature of the active sites produced by the copper addition and not to a partial coverage of Pd particles, as quantitative XPS data suggest.
Pd, Pd-Cu and Cu/SiO2 catalysts were prepared, characterized and studied in the oxidation of toluene. It was found that the dispersion of both metals is quite different. Palladium metal dispersion is comparable in all the studied catalysts, having palladium crystals size in the range 1.7 to 2.0 nm, whereas copper particles are larger and most of these particles are in the range 6 to 10 nm. TPR and XPS results reveal that it is likely the formation of alloy in a slight extension. The monometallic Pd/SiO2 is very active in the total combustion of toluene. The addition of copper decreases the activity drastically due to that Cu catalysts have poor activity in this reaction. When the combustion of toluene is carried out in the presence of thiophene, the activity remains almost constant, with the exception of the monometallic Cu/SiO2 in which the activity drops drastically due to the sulfur poisoning of the metallic phase.
The authors thank to the D.I. Universidad de Concepción for the financial support. Grant 98-022.015
1. R. Prassad, L. Kennedy, E. Ruckenstein, Catal. Rev. Sci. Eng., 26 (1984) 1. [ Links ]
2. G. Pecchi, P. Reyes, I. Concha and J. L. G. Fierro, J. Catal., 179 (1998) 309. [ Links ]
3. R. B. Anderson, K. C. Stein, J. J. Feenan and L. E. J. Hofer, Ind. Eng. Chem., 53 (1961) 809. [ Links ]
4. T. López, A. Romero and R. Gómez, J. Non-Cryst. Solids, 127 (1991) 105. [ Links ]
5. P. Reyes, M. Morales and G. Pecchi, Bol. Soc. Chil. Quím., 43 (1998) 391. [ Links ]
6. R. J. Farrauto, M.C. Hobson, T. Kennelly and E. M. Waterman, Appl. Catal. A, 81 (1992) 27. [ Links ]
7. T. López, P. Bosch, M. Asomoza, E. García-Figueroa and R. Gómez, J. Catal., 138 (1992) 463. [ Links ]
8. G. Pecchi, M. Morales and P. Reyes, React. Kinet. Catal. Lett., 61 (1997) 237. [ Links ]
9. C. D. Wagner, L. E. Davis, M. V. Zeller, J. A. Taylor, R.H. Raymond and L. H. Gale, Surf. Interf. Anal., 3, 211 (1981) [ Links ]
10. J. G. Mc. Carty, Catal. Today, 26 (1995) 283. [ Links ]
11. S. D. Robertson, B. D. Mc Nicol, J. H. De Bass, S.C. Kloet and J. W. Jenkins, J. Catal., 37 (1975) 424. [ Links ]
12. N. Iwasa, S. Masuda, O. Ogawa and N. Takesawa, Appl. Catal. A. General: 125, 145 (1995) [ Links ]
13. J. Goetz, M. A. Volpe, A. M. Sica, C. E. Gigola and R. Touroude, J. Catal., 153, 86 (1995) [ Links ]