Format And Type Fonts CHEMICAL ENGINEERING TRANSACTIONS VOL. 39, 2014 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Peng Yen Liew, Jun Yow Yong Copyright © 2014, AIDIC Servizi S.r.l., ISBN 978-88-95608-30-3; ISSN 2283-9216 DOI:10.3303/CET1439202 Please cite this article as: Palma V., Barba D., Ciambelli P., 2014, Vanadium-Ceria structured catalysts for the selective partial oxidation of H2S from biogas, Chemical Engineering Transactions, 39, 1207-1212 DOI:10.3303/CET1439202 1207 Vanadium-Ceria Structured Catalysts for the Selective Partial Oxidation of H2S from Biogas Vincenzo Palma*, Daniela Barba, Paolo Ciambelli University of Salerno, Department of Industrial Engineering, via Giovanni Paolo II, 132 - Fisciano (SA) - Italy vpalma@unisa.it The reaction of H2S selective partial oxidation to elemental sulfur in the range of temperature of 150-200 °C was studied on cordierite honeycomb-structured catalysts. The preparation procedure of catalytic cordierite monoliths has been studied, from the washcoating with CeO2 to the deposition of V2O5 by wet impregnation. Two different preparation washcoating procedures deposition were studied: in one case the washcoat had been already added with the salts precursors of the active species (joint impregnation method), in other case the active species were added only after the deposition step of the washcoat on the monolith, by impregnation in a solution of the salt precursor (distinct impregnation method). The catalysts prepared with the two different methods were characterized and the stability was investigated in catalytic activity tests. The catalysts prepared with the “joint impregnation” method have shown a poor catalytic activity and tendency to the deactivation. Very different results were found for the catalysts prepared with the “distinct impregnation” method, for which the effect of the vanadium content (2-19 V2O5 wt %) was also studied at 200 °C. Good catalytic performances were obtained for both samples that have shown a high H2S conversion (~90 %), low SO2 selectivity (2 %) and a high stability. 1. Introduction Several processes have been described and developed for the elimination of H2S from products or off- gases. For the small scale plant, a very interesting solution could be represented by the direct and selective H2S oxidation to sulphur as reported in the following reaction H2S + ½ O2 = 1/x Sx + H2O and performed at low temperature in the presence of a very active and selective catalyst (Yasyerly and Dogu, 2004). In a previous work, we examined the performances of vanadium-oxide based catalysts supported on the metal oxides in the range of temperature of 50 - 250 °C (Palma et al., 2012). Among the investigated samples, V2O5/CeO2 catalyst showed the most promising catalytic performances at temperature lower than 200 °C in terms of high H2S and O2 conversion with a very low SO2 selectivity at a feed ratio H2S/O2= 2. In order to verify the possibility to further reduce the SO2 selectivity, additional experimental tests were performed by investigating the effect of the inlet H2S concentration, the gas hourly space velocity and the H2S/O2 molar feed ratio (Palma et al., 2013a). Additional catalytic activity tests were also carried out to study the effect of the vanadium content in the range 2.55-20 wt % V2O5 at different temperatures the results of which of this preliminary screening showed that, in the temperature range of 150 - 250 °C, the vanadium load mainly affected the value of SO2 selectivity, with not so relevant effects on the catalytic activity. The most selective catalyst was the 20 wt % V2O5/CeO2 showing a sulfur selectivity of about 99 % at 150 °C (Palma et al., 2013b). The aim of this work is to transfer this formulation on a cordierite (2MgO·2Al2O3·5SiO2) honeycomb structured carrier by using a CeO2 based-washcoat as support for the vanadium oxide deposition (2-19 V2O5 wt %). Monolithic catalysts can be an attractive replacement of conventional carriers in heterogeneous catalysts (Heck et al., 2001). The development of the cordierite monolithic structure has been stimulated by the 1208 requirements of a widely variety of gas-phase reactions such as catalysis in catalytic combustion of VOCs, automotive application, selective reduction of NOx. The main reasons are due to its high mechanical strength, low thermal expansion coefficient, and low-pressure drop (Nijhuis et al., 2001). After the selection of the cordierite as structured carrier, the attention will be focused on the washcoat impregnation with the active phase (V2O5). To this scope, two different techniques will be used in order to verify the possibility to obtain a one-step catalytic washcoat deposition procedure. The different structured samples will be characterized in terms of mechanical stability test by ultrasonic vibration to investigate the adhesion of ceria washcoat on cordierite and compared in terms of catalytic activity tests. 2. Experimental 2.1 Materials Cordierite honeycomb monoliths (9 channels) with 30 mm in length, 6 mm in width, 6 mm in height and 226 cells/in 2 (cpsi) were used as the substrate on which a commercial Ceria-Zirconia washcoat (Ecocat) was deposited. For the active phase deposition (V2O5) we started by aqueous solution of ammonium metavanadate (Aldrich). 2.2 Catalysts preparation Before the washcoat deposition on the cordierite monolith, the carrier was predried and evacuated at 550 °C for 2 h. The monoliths were impregnated with the washcoat by dip coating technique and the excess suspension inside the channels of the cordierite substrate was blown off. More precisely the excess suspension was evacuated from the channels by a vacuum pump and after the monoliths were dried at 120 °C for 30 min. Since the monolith cannot be coated sufficiently by a single impregnation, multi-impregnation were required. In order to get more homogeneous washcoatings, it was preferable to use diluted suspensions and perform more than one immersion. The target of the washcoating procedure was to deposit about 20 wt % of washcoat on the monoliths, in order to completely cover the monoliths channels with a thin layer of washcoat; for the following impregnations the procedure was the same as the first one, until a 20 wt % increase was obtained. Finally the washcoated monoliths were calcined in air at 550 °C for 2 h. Two different washcoating procedures deposition were studied. In a case, the solution of the salt precursor of the active phase is added to the washcoat slurry and the monolith is dipped into a suitable slurry (joint impregnation method). After the removing of the excess suspension, the catalysed monolith was dried and also calcined at 400 °C for 3 h. In the other case the monolith, after washcoat deposition and stabilization by calcination at 550 °C for 2 h, is dipped into the solution of the salt precursor of the phase active (distinct impregnation method). The impregnation/drying steps into the solution of the salt precursor of the phase active were performed for many times until it was reached the desired content of the active phase. Finally the monoliths were calcined at 400 °C for 3 h. Catalysts with a vanadium content variable between 2 wt % and 19 wt % were prepared with this last preparation method. The catalysts prepared was also characterized by different techniques: X-ray diffraction measures, Raman spectroscopy; furthermore the adhesion of the washcoat was evaluated using an ultrasonic technique. One of goals of this study was to find out a good preparation route to obtain a mechanically stable washcoated cordierite monolith with catalytic activity at least as good as the powder catalyst. The catalytic tests were carried out in a fixed bed flow reactor, made of a pyrex glass tube 0.21 m long and a 0.014 m of internal diameter, inserted in an electrical furnace equipped with a PID electronic temperature controller. Preliminary tests were carried out at atmospheric pressure and GHSV of 21,000 h -1 (~170 ms), in the temperature range 150 - 250 °C, with 500 ppm of H2S, 250 ppm of O2 and N2 to balance. The exhaust stream (H2S, O2, H2O) was analyzed by a quadrupole filter mass spectrometer while the concentration of SO2, which may be present in the stream leaving the reactor, is monitored instead by an analyzer FT-IR Multigas in continuous. The H2S conversion and the SO2 selectivity were calculated by using the following equations, by considering negligible the gas phase volume change: xH2S, % = ((H2S IN -H2S OUT )/ H2S IN )·100 ySO2, % = SO2/(H2S IN -H2S OUT )·100 (1) (2) 1209 3. Results and Discussion The Raman spectra of the samples with a V2O5 content of 20 wt % prepared with the different impregnation techniques have shown in the following figures (Figure 1a-b). For the sample prepared with the “joint impregnation” method, it is possible to identify the characteristic peak of CeO2 at 458 cm -1 and other peaks (260, 377, 785, 792, 857 cm -1 ) ascribable to the presence of the cerium vanadate - CeVO4 (Gu et al., 2006) Figure 1a. The Raman spectra of the catalyst prepared by the “distinct impregnation” method have shown in the Figure 1b. Figure.1: Raman spectra of the fresh catalysts prepared with the joint impregnation (a), and with the distinct impregnation (b) From the Raman spectra shown in Figure 1b, it is possible to attribute all the peaks to V2O5 crystalline species, that exhibited Raman characteristic peaks at 144, 203, 283, 302, 404, 489, 523, 697, 994 cm -1 (Daniell et al., 2002). The washcoat adhesion test, performed by exposure to ultrasounds showed only some very low, or almost negligible washcoat weight loss during the first cycles of adhesion test, indicating a good adherence to the cordierite support. The optimal washcoat thickness was determined starting from three different thickness (50, 150, 226 μm) at temperature of 200 °C and under conditions far from the thermodynamic equilibrium to better evaluate a difference of the H2S conversion values (Figure 2). Figure.2: Influence of the washcoat thickness on the catalytic activity (Q=800 Ncm 3 /min, T=200 °C) A decrease of the H2S conversion value was observed for a washcoat thickness greater than 150 µm, likely due to the internal diffusion phenomena inside the washcoat. 1210 For this reason, the catalysts will be prepared by using a washcoat thickness between 50 - 150 µm. Table.1 Effect of the different washcoating procedure on the catalytic activity Joint impregnation Distinct impregnation XH2S, % 77 90 YSO2, % 1 2 The effect of the different washcoating procedure (joint and distinct impregnation) on the catalytic performances was studied on catalysts having a vanadium loading of 2 wt % at 200 °C. The results, in terms of H2S conversion and SO2 selectivity are reported in Table 1. From the results obtained it is possible to observe quite similar values of SO2 selectivity and an higher H2S conversion only for the catalyst prepared with the distinct impregnation of the active phase respect to washcoat. Based on these results, other catalysts will be prepared with the distinct impregnation method of the washcoat with the active phase. The effect of the V2O5 loading variable between 2 wt % and 19 wt % was investigated on the catalytic performances and in particular relatively to the SO2 minimization at temperature of 150-200 °C. The catalytic test performed on the catalyst with the lowest vanadium content (2 wt %), reported in the Figure 3 at 200 °C, shows the concentration of the reactants (H2S, O2) and the products (SO2, H2O) as function of time. The feed stream is sent in by-pass to the reactor for 30 min to obtain stable concentration values of H2S and O2; the strong decrease of the reactants concentration and the increase of the products concentration, denotes the sending of the feed mixture to the reactor. During the test (70 min), it was obtained a good catalytic activity and a high stability. By the concentration profiles of H2S and oxygen from the mass fragment, it is possible to note the achievement of stationary values just after 30 min of the test with a final conversion of H2S equal to 90 %. It was found a low SO2 formation, the concentration of which from the outset has reached a steady state value of 10 ppm. The formation of H2O, stable during the test, is indicated by the presence of mass fragment 18. Figure.3: Catalytic activity test on monolith with 2 wt % V2O5 (T=200°C) 1211 Figure.4: Catalytic activity test on monolith with 19 wt % V2O5 (T=200°C) Similar results were also obtained in the case of the catalytic test performed with the catalyst having a vanadium loading of 20 wt %, where the H2S final conversion was about 92 % with a low value of SO2 selectivity (2 %), corresponding to about 9 ppm of SO2 (Figure 4). Even in this case, during the test (~ 2 h), the sample did not show evident deactivation phenomena. 4. Conclusions The reaction of H2S selective partial oxidation to elemental sulfur in the temperature range of 150-200 °C was studied on specifically prepared honeycomb-structured catalysts obtained starting by honeycomb cordierite monoliths and a commercial Ceria based washcoat, at which the V active species was added. The ultrasound tests showed that the washcoat anchors and interlocks well on the substrate and the high calcination temperature promotes the stability of the washcoat (Boix et al., 2003). Raman spectra obtained on the catalysts prepared with the two different methods, showed the formation of different species, likely responsible of the sensibly altered catalytic activity. In the catalytic tests, the so prepared catalyst showed a moderate H2S conversion, corresponding to about 77 % at 200 °C, and at the same time, a low durability, with a strong tendency to the deactivation. Very different results were found for the catalysts prepared with the “distinct impregnation” method for which the effect of the vanadium content (2 wt % - 19 wt %) was also investigated in the range 150-200 °C. The results obtained at 200 °C for the two Vanadium load were very similar, in terms of H2S conversion (~90 %) and SO2 selectivity (2 %).The absence of a sensible effect of the different V2O5 on the catalytic performances, can be likely due to the very good catalytic activity of the two samples, that in the selected operating conditions (150 – 200 °C) reached a H2S conversion close to the thermodynamic equilibrium. The worst catalytic performances obtained for the samples prepared with the “joint impregnation” method, may be attributed to the presence of CeVO4 instead of the VOx, that are reported in the literature as the most active species for the desired reaction. References Boix A.V, Zamaro J.M., Lombardo E.A., Miro E.E., 2003, Appl. Catal. B, 46, 121-132. Daniell W., Ponchel A., 2002, Characterization and catalytic behavior of VOx–CeO2 catalysts for theoxidative dehydrogenation of propane, Catalysis, 20, 1–4. Gu X., Jiazen Ge J., Zhang H., Aurox A., Shen J., 2006, Structural, redox and acid - base properties of V2O5/CeO2 catalyst, Thermochimica Acta, 451, 84-93. Heck R.M., Gulati S., Farrauto R.J., 2001, Chem. Eng. J., 82, 149. Nijhuis T.A., Beers A.E.W., Vergunst T., Hoek I., Kapteijin F., Moulijn J.A., 2001, Catal. Rev., 43, 345. Palma V., Barba D., Ciambelli P., 2012, Chemical Engineering Transaction, 29, 631-636, DOI: 10.3303/CET1229106 1212 Palma V., Barba D., Ciambelli P., 2013a, Intern. Journal of Hydrogen Energy, 38, Issue 1, 328-335, DOI 10.1016/j.ijhydene.2012.10.010. Palma V., Barba D., Ciambelli P., 2013b, Chemical Engineering Transactions, 32, 631-636, DOI: 10.3303/CET1332106. Yasyerly S., Dogu G., 2004, Chemical Engineering Science, 59, 4001.