Ind. Eng. Chem. Res. 2004, 43, 4073-4079
4073
Study of Configuration and Coating Thickness of Vanadium on Carbon-Coated Monoliths in the SCR of NO at Low Temperature E. Garcı´a-Bordeje´ ,* L. Calvillo, M. J. La´ zaro, and R. Moliner Instituto de Carboquı´mica (C.S.I.C.), Miguel Luesma Casta´ n 4, 50015-Zaragoza, Spain
Different vanadium loadings, viz., 3, 4, 6, and 8 wt %, have been supported on mesoporous carboncoated monoliths. These catalysts have been tested in the SCR of NO at low temperature. Experiments with different monolith configurations and area velocities have been performed. The conversions of the monoliths placed in series, in parallel, and ground were found to be similar. The latter result indicates that the reaction rate in the monolith is not controlled by diffusion. A one-dimensional model has been applied that predicts reasonably well the results of these experiments. Finally, with this model, we have simulated the effects of varying the carbon coating thickness and cell density on the geometric parameters of the monolith, the pressure drop, the area velocity, and the conversion. As a result, an optimal coating thickness of around 30-40 µm was found. 1. Introduction The major technology for reducing nitrogen oxide emissions from stationary sources is the selective catalytic reduction (SCR) of NOx by ammonia. The industrial operations are carried out on V2O5 + WO3 (MoO3)/TiO2 catalysts at 573-673 K.1,2 However, the high concentrations of SO2 and ash in the flue gas reduce the catalysts performance and longevity. To overcome this problem, the focus has recently been placed on developing catalysts active at low temperature (below 473 K). These catalysts can be implemented in existing plants downstream of the desulfurizer and electrostatic precipitators where most of the SO2 and ash have been removed. Even after the desulfurizer, 100-1000 ppm SO2 remains. SO2 can lead to the formation of ammonium sulfate salts such as NH4HSO4 and (NH4)2S2O7 that can poison the catalyst and upstream equipment. Therefore, the catalyst has to be stable to the remaining SO2. Catalysts supported on carbon have shown very high activities at low temperature and resistance to the remaining SO2.3-13 The only works found in the literature in which the catalyst is in monolithic shape are those reported by Marba´n et al.11-13 The commercial catalyst used in SCR at high temperature is in monolithic form because it has some inherent advantages over conventional fixed-bed catalytic reactors, especially when high gas flow rates have to be treated.14-18 This is the case for exhaust gases from combustion. Some of these advantages are a pressure drop lower by about 2 orders of magnitude, a higher tolerance to dust, a greater resistance to attrition, and a better gas distribution resulting in more uniform access of reactants to the catalyst surface. Carbon is a special catalyst support material showing unique characteristics, such as adjustable porosity and surface chemistry; it is also relatively inert, so that undesired side reactions catalyzed by the support surface hardly occur. To increase the potential application area of carbon, carbon-based monolithic structures * To whom correspondence should be addressed. Tel.: +34 976733977. Fax: +34 976733318. E-mail: jegarcia@ carbon.icb.csic.es.
have been developed.19-26 Two different types of carbon monolithic structures are distinguished: coated and integral monoliths. The integral type is prepared by extrusion of the carbon precursor, e.g., phenolic resin, mixed with various additives, e.g., ceramic powders, cellulose, or polyester fibers. Tennison22 described a method for producing pure carbon monoliths using Novolak resin. The absence of ceramic filler results in significant shrinkage upon carbonization and in more difficult control of the macroscopic structure. Different methods for preparing carbon-coated monoliths have been described in the literature. The most frequently used is the dip-coating method. Different carbon precursors have been used for this process, such as phenolic23 and furanic resins,24 saccharides,25 and poly(furfuryl alcohol).26 In this type of carbon monolith, the monolithic structure, either ceramic or metallic, supplies the mechanical and geometric properties, whereas the carbon layer has to provide the catalytic properties. In this work, mesoporous carbon-coated monoliths have been prepared as supports for different vanadium loadings. These catalysts have been tested in the SCR of NO at low temperature. In the experiments, we have used different monolith configurations, area velocities, and lengths. A one-dimensional model has been applied that predicts the experimental results with a high degree of agreement. Finally, we have simulated the influence of the carbon coating thickness and cell density on some parameters, such as the geometric parameters of the monolith, pressure drop, area velocity, and conversion. From this simulation, an optimum carbon coating thickness can be derived. 2. Experimental Section 2.1. Preparation of Carbon-Coated Monoliths. Cordierite monoliths (400 cpsi, 1 cm diameter, 5 cm length) were coated with a polymer blend by the method described elsewhere.7,21 The polymers used were furan resin (Huttenes-Albertus) and poly(ethylene glycol) with a molecular weight of 6000 (Sigma-Aldrich). After thermal curing, the monoliths were carbonized at 973 K, activated with CO2 at 1173 K for 8 h (30% carbon
10.1021/ie0498854 CCC: $27.50 © 2004 American Chemical Society Published on Web 06/18/2004
4074 Ind. Eng. Chem. Res., Vol. 43, No. 15, 2004
burnoff), and treated for 24 h with 1 N HNO3 at room temperature to create oxygenated surface groups. 2.2. Vanadium Catalyst Preparation. The carbons prepared were loaded with different amounts of vanadium, i.e., 3, 4, 6, and 8 wt %. The impregnation was carried out by equilibrium adsorption of ammonium metavanadate in the stoichiometric amount to get the desired vanadium loading on carbon. To facilitate the dilution of ammonium metavanadate, ca. 10 mg of oxalic acid was added. Under these conditions, the pH of solution remained neutral, and the solution held a yellow color indicative of the presence of VO2+. The monoliths placed in a holder were introduced into a vessel with 100 mL of the impregnating solution. The vessel had a stirrer at the bottom that created a continuous flux of the vanadium solution through the monoliths. This guaranteed that vanadium was deposited inside the channels. This process was allowed to proceed for 18 h. It was visually observed that the solution turned from yellow to colorless in a few hours, indicating that all vanadium in solution had been adsorbed. After this process, the monoliths were rinsed with distilled water in the same setup. The V content of the solutions before and after impregnation was analyzed by ICP. From the difference in V content of these solutions, the V loading was calculated. Finally, the catalyst was calcined in Ar at 673 K. 2.3. Catalytic Tests. The catalysts prepared were tested in the selective catalytic reduction of NO with ammonia at four temperatures, viz., 363, 393, 423, and 453 K. The gas composition was 700 ppm NO, 800 ppm NH3, 3% O2, and the balance Ar. The catalytic tests were performed in a quartz reactor in which two cylindrical monoliths (1 cm in diameter, 5 cm in length) were placed in parallel. The gas was forced to flow through the monolith channels by fixing the monoliths to the inner walls of the reactor using a Teflon stripe. The total amount of carbon coating was ca. 0.3 g, and the total flow rate was 100 mL/min STP, yielding a GHSV of ∼17 000 h-1 (relative to carbon weight). To analyze the gases, a mass spectrometer (Balzels) was used. The mass spectrometer was calibrated with cylinders of gases of known composition. The following main mass-to-charge (m/e) ratios were used to monitor the concentrations of products and reactants: 17 (NH3), 18 (H2O), 28 (N2), 30 (NO), 32 (O2), 44 (N2O), and 46 (NO2). The fragmentation pattern of each gas was treated as follows: Using the m/e 12 signal and taking into account the fact that the contribution of CO2 to m/e 12 is 6%, the amount of CO2 desorbed at steady state was calculated. The N2O concentration was calculated from m/e 44 after the contribution of CO2 had been subtracted. The NO concentration was calculated from m/e 30 after subtraction of the contribution to this mass of N2O (30%). The concentration of N2 was calculated from m/e 28 after subtraction of the contributions of N2O (11%) and CO2 (11%). Finally, the NH3 concentration was obtained from m/e 17. From the concentrations of the gases at steady state, both the conversion and the selectivity to N2 were calculated according to the standard formula. All experiments were repeated, and the maximum difference between the results was 1% of the value, indicating that the catalyst is not deactivated significantly after 6-h experiments. 2.4. Model Description. In the literature, several models have been developed for monolithic catalyst in
Figure 1. Structure and geometric parameters of coated monoliths, following ref 39. dch is the channel diameter, dw is the wall thickness, dc is the coating thickness, and R is the fill radius (the radius of the circle that fits in the corner or the coating). Table 1. Geometric Parameters for a Monolithic Structure with Channels of Square Section parameter
units
expression
cell density geometric surface area void fraction hydraulic dameter catalyst volume fraction
m-2
n ) 1/Dch am ) 4n[(Dch - δw - 2δc) - (4 - π)(R/2)]
m-1 m
2
m ) n[(Dch - δw - 2δc)2 - (4 - π)R2] Dh ) 4m/am xc ) n(Dch - δw)2 - m
the SCR of NO at high temperature.27-36 To the best of our knowledge, there has been no application of monolithic reactor models to the SCR of NO at low temperature. In the model developed in this work, monolithic supports with square-shaped channels are considered. The structure of a monolithic catalyst is depicted in Figure 1 (from ref 37). The geometric parameters of monoliths, such as cell density, geometric surface area, void fraction, hydraulic diameter, and volume fraction of catalyst, can be calculated with the equations presented in Table 1. The model developed in this work is a one-dimensional catalyst model similar to that proposed by Tronconi and Forzatti27 for the SCR of NO at high temperature. One-dimensional means that the monolithic catalyst is modeled using a slab geometry, with mass transfer to and from the catalyst layer and reaction and diffusion occurring in the catalyst layer. When the reaction is carried out at high temperature (>500 K), it is controlled by diffusion, resulting in low effectiveness coefficients.38 On the other hand, when the reaction is carried out at low temperature (
