Kinetics of the Catalytic Decomposition of Ammonia over a Bimetallic Copper-cerium Nanoparticle Catalyst

This study examines the kinetics of selective catalytic oxidation of ammonia to nitrogen over a bimetallic copper-cerium nanoparticle catalyst in a tubular fixed-bed reactor. Transmission electron microscopy results showed that the catalyst containing copper and cerium is well dispersed and presents in the form of nano-sized particles. The formation of CuO/CeO2 active sites was confirmed by Fourier Transform Infrared Spectroscopy. Moreover, the catalytic behavior of ammonia oxidation can be expressed with a simple kinetic model. The model-predicted values were then compared to the observed data. *Corresponding author Email: [email protected] INTRODUCTION Ammonia is widely found in industrial processes and is a valuable chemical product for a variety of purposes. It originates from different sources, including ammonium nitrate production industry, livestock feedlots, urea manufacture plants, nitrogen fertilizer application industry, fossil fuel combustion or petroleum refineries, and refrigeration industry. Ammonia can be toxic and its pungent malodorous characteristic under ambient conditions can have the potential harmful effects on the public [1-3]. It is also one of the controlled compounds of the Environmental Protection Administration, Taiwan. Moreover, conventional biological, physical and chemical treatment processes, such as biofilters, stripping, water scrubbing, postcombustion control technologies, microwave-plasma discharge, electrochemical oxidation, and activated carbon fibers, only make a phase transformation and may yield a contaminated sludge and/or an adsorbent, both requiring for further disposal, and the maintenance/operation costs of ammonia are pretty high by physical and/or chemical methods. Therefore, the removal, control, and prevention of ammonia emission from air and waste streams are important. Catalytic oxidation has been known to enhance the possibilities of the advanced oxidation processes technology through the use of catalyst, which can potentially promote oxidation in shorter reaction time under milder operating conditions. The selective catalytic oxidation (SCO) for ammonia-containing stream to molecular nitrogen and water serves as one of the methods to solve ammonia pollution problems [4-5]. There are many types of catalysts, which have been used for oxidation of ammonia in gaseous phase. For example, Amblard et al. [6] proposed an excellent selective conversion of ammonia to nitrogen (> 90%) by γ-Al2O3-supported Ni in the SCO processes. Furthermore, Wang et al. [5], who developed a Ni-based catalyst for fuel gas oxidation from biomass gasification, found the fresh Ni-based catalyst to be more active at lower temperature for the decomposition process of ammonia, and the partial pressure of hydrogen in the flue gas was a key factor to model ammonia oxidation. Liang et al. [7] studied ammonia oxidation in a fixed-bed microreactor in the temperature range of 600-750 °C at Gas Hourly Space Velocity (GHSV) 1800-3600 h. They found that the conversion of ammonia reached 98.7% and 99.8% on nitrided MoNx/α-Al2O3 and NiMoNy/α-Al2O3 catalysts, respectively. Olofsson et al. [8] demonstrated an excellent catalytic conversion of ammonia for nitrogen formation and by γ-Al2O3-supported Pt/CuO in the SCO processes. Among these relevant researches, Schmidt-Szalowski et al. [4] also published a paper presenting a hypothetical model to explain the effect 270 J. Environ. Eng. Manage., 18(4), 269-274 (2008) Fig. 1. The schematic diagram of the tubular fixed-bed reaction (TFBR) system. and the activity and selectivity in ammonia oxidation of the cobalt oxide catalyst’s macrostructure on its properties. As the reaction kinetic model is concerned, Lou and Chen [9] use a catalyst composed of Pt, Ni and Cr alloy of foam type to study the kinetics of catalytic incineration of butanone and toluene. They found that the Mars and van Krevelen model was suitable to describe the catalytic incineration of those volatile organic compounds. Lou and Lee [10] used a Pt/Al2O3 alloy catalyst to study the kinetics of catalytic incineration of trichloromethane. They adopted the powerrate law kinetics and found that the reaction was firstorder in trichloromethane concentration and the activation energy was 67.8 kJ mol. Hung [3] employed a CuO-CeO2 binary oxide catalyst to describe the kinetics of catalyzed NH3 oxidation by using the rate expression of the Eley-Rideal kinetic model. Hung [11] also utilized a nanoscale copper-cerium mixed-metal oxide to investigate the kinetics of the catalytic incineration of NH3. According to their results, the Langmuir-Hinshelwood model was appropriate for describing the catalytic incineration of NH3. However, the integral kinetic studies of catalytic oxidation of NH3 on metal composite catalysts have not been thoroughly investigated. In this study, we investigated the nature of the adsorbed species formed on the catalyst surface using an interpretation of the kinetic data. The integral kinetic model was evaluated in driving the rate expression for ammonia oxidation. Hence, we sought to study the activity of the bimetallic copper-cerium nanoparticle catalyst on oxidation of ammonia with various parameter values and the kinetic behavior of ammonia removal in the effluent stream. Our results might provide valuable information for designing and treating an ammonia-related system. MATERIALS AND METHODS The bimetallic copper-cerium nanoparticle catalyst used in this study was prepared by coprecipitation of Cu(II) nitrate (GR grade, Merck, Darmstadt, Germany), and Ce(III) nitrate (GR grade, Merck, Darmstadt, Germany) at the molar ratio of 6:4. It was calcined at 500 °C in an air stream for 4 h. The resulting powder was made into tablets using acetic acid as a binder. The tablets were later reheated at 300 °C to burn the binder out of the nanoscale coppercerium bimetallic tablets. And then they were crushed and sieved into various particle sizes ranging from 0.15 to 0.25 mm for later use. The experiments were carried out on a tubular fixed-bed flow quartz reactor, as shown in Fig. 1. The flowing gases, namely ammonia, prepared the feeding mixture and the diluent’s gas, helium, at the inlet of the reactor, and each of the gases was independently controlled with a mass flow regulator. High-purity dry air was used as the carrier gas and controlled with a mass flow meter (Sierra instrument Inc., USA) in the range of 8-13 L min. The weight of catalyst was 1 g (empty bed volume approximately at 1.2 cm). An inert material of γ-Al2O3 spheres (a hydrophilic inert material) was used to increase the interfacial area in gas phase for better mass transfer of ammonia from air. Hence, the limitations of external mass transfer and inter-particle diffusion were negligible in this reactor. The reaction tube (length 300 mm and inner diameter 28 mm) was placed inside a split tube furnace, and the tube containing the catalyst was also placed inside anHung: Ammonia Decomposition over a Cu-Ce Catalyst 271 other spilt tube furnace. Two thermocouples of type K (Omega, USA) and a diameter of 0.5 mm were mounted and equally spaced along the catalyst bed. The thermocouples were also connected to a PID controller (FP21, Shimaaen Co Ltd, Japan) to maintain a uniform temperature in the tube within ±0.5%. The feed gas (GHSV, 92 L h g) was controlled at the concentration of 1000 ppm NH3 under a GHSV of 92 L h g and a concentration of O2 of 4%. Samples before and after the reaction were injected automatically with a sampling valve into a gas chromatograph (Shimadzu GC-14A) equipped with a thermal conductivity detector. A stainless-steel column (Porapak Q 80/100 mesh) was used for separation and analysis of the concentrations of N2O isothermally at 100 °C. NH3 was injected automatically before and after the reaction with a sampling valve into a gas chromatograph (Shimadzu GC-14A) equipped with a thermal conductivity detector (Shimadzu, Kyoto, Japan). The signal areas were measured electronically with a data integrator (CR-6A, Shimadzu). The concentrations of NO, NO2 and O2 in the gas samples were measured by a portable flue gas analyzer (IMR3000, IMR, Germany), which was linked to the designed location for continuous monitoring during the combustion. Gas samples with known concentrations of NH3 were used for the calibration. Transmission electron microscopy (TEM) (Philips CM-200 Twin, Netherlands) elucidated the morphology of the catalyst and yielded information on the distribution of copper-cerium species on the catalyst surface. Diffuse reflectance Fourier Transform Infrared Spectroscopy (FTIR) spectra of species adsorbed on the catalyst were measured at room temperature using a Bruker Vector 22 FTIR spectrometer equipped with a diffuse reflectance attachment with 4 cm resolution (Bruker, USA). RESULTS AND DISCUSSION Figure 2 presents the surface morphological changes of nanoscale copper-cerium bimetallic catalyst elucidating using TEM to examine the fresh/aged catalyst surface structure. The mean particle size converged at approximately 70 nm for the fresh coppercerium bimetallic catalyst. However, the diameters of the catalyst decreased during the activity test and the mean particle size was about 100 nm, indicating that the metal diameters of the catalyst decreased when the catalyst surface was aged. Furthermore, Fig. 2a indicates that the nanoscale copper-cerium bimetallic catalyst exhibited more aggregation and crystallin. These crystal phases might explain the high activity of the catalyst. Figure 2b indicates that the disaggregated and dispersed phases were formed when the surface of the catalyst was aged or when poisoning occurred because of plugging, implying that the porosity of the particles had changed. These results also confirmed Fig. 2. TEM photographs of (a) fresh and (b) after activity test a nanoscale copper-cerium bimetallic catalyst. Test conditions: 1,000 ppm NH3 in He, GHSV = 92 L h g.