Removal for Flue Gas Cleanup - ACS Publications - American

Center for Engineering and Environmental Technology, Research Triangle Institute,. P.O. Box 12194, Research Triangle Park, North Carolina 27709-2194, ...
15 downloads 0 Views 321KB Size
Download PDF
Energy & Fuels 1997, 11, 299-306

299

Low-Temperature NOx Removal for Flue Gas Cleanup Ben W.-L. Jang,*,† James J. Spivey,† Mayfair C. Kung,‡ and Harold H. Kung‡ Center for Engineering and Environmental Technology, Research Triangle Institute, P.O. Box 12194, Research Triangle Park, North Carolina 27709-2194, and Department of Chemical Engineering, Northwestern University, Evanston, Illinois 60208 Received September 5, 1996. Revised Manuscript Received December 19, 1996X

NOx reduction with hydrocarbons is studied at temperatures around 150 °C to develop an easy retrofit NOx removal process to be located downstream of the particulate control device for flue gas cleanup. Acetone is the most active hydrocarbon reductant for NO reduction in the presence of O2 and H2O. At 150 °C, 35% NO reduction was obtained over 5% Cu-2% Ag/PCB. The activity, however, is greatly decreased by the presence of SO2. NO oxidation to NO2 is studied for further improving the NO reduction activity because NO2 is known to be more reactive than NO. Active carbon fibers and PCB active carbon are both active for NO oxidation at low temperature, but the activity decreased with temperature. Ten percent Co3O4/Al2O3 and 5% Au/Co3O4 are not active for NO oxidation in the absence of SO2 and at 150 °C, but are very active in the presence of SO2 and H2O. Five percent Au/Co3O4 is especially stable for NO oxidation to NO2 at temperatures from 120 to 150 °C for more than 15 h.

Introduction NOx (NO + NO2) emission from flue gas represents a health hazard and an environmental concern. To protect us and our environment, the Clean Air Act Amendment of 1990 has driven the widespread use of selective catalytic reduction (SCR) with NH3 for electric utilities to control the emission of NOx. Although SCR using NH3 is effective for NOx reduction, the process suffers from many disadvantages.1 Current SCR processes must be located at places within the flue gas treatment process where temperatures are near 400 °C. Retrofit of SCR into these locations in the plant is costly because space and access in many power plants are extremely limited. Therefore, there is a need to develop NOx control technologies that can be located downstream of the particulate control device, near the stack, where temperatures are around 150 °C.2 Recently, it was found that NOx can be selectively reduced by hydrocarbons in excess oxygen.3-8

NO + hydrocarbons + O2 f N2 + H2O + COx The progress in catalytic removal of NOx under oxidizing conditions has been reviewed.9-11 As emphasized in * Author to whom correspondence should be addressed (e-mail [email protected]). † Research Triangle Institute. ‡ Northwestern University. X Abstract published in Advance ACS Abstracts, February 1, 1997. (1) Li, Y.; Battavio, P. J.; Armor, J. N. J. Catal. 1993, 142, 561571. (2) Singoredjo, L.; Slagt, M.; Wees, J. V.; Kapteijn, F.; Moulijn, J. A. Catal. Today 1990, 7, 157-165. (3) Petunchi, J. O.; Gustave, S.; Hall, W. K. Appl. Catal. B 1993, 2, 303-321. (4) Gopalakrishnan, R.; Stafford, P. R.; Davidson, J. E.; Hecker, W. C.; Bartholomew, K. H. Appl. Catal. B 1993, 2, 165-182. (5) Miyadera, T.; Yoshida, K. Chem. Lett. 1993, 1483-1486. (6) Hamada, H.; Kintaichi, Y.; Sasaki, M.; Ito, T.; Tabata, M. Appl. Catal. 1991, 75, L1-L8. (7) Hamada, H.; Kintaichi, Y.; Yoshinari, T.; Tabata, M.; Sasaki, M.; Ito, T. Catal. Today 1993, 17, 111-120. (8) Montreuil, C. N.; Shelef, M. Appl. Catal. B 1992, 1, L1-L8.

S0887-0624(96)00138-7 CCC: $14.00

those publications, NOx removal levels obtained by direct decomposition into N2 and O2 over a catalyst are far from practical. However, selective catalytic reduction of NOx in an oxidizing atmosphere, using hydrocarbons or oxygenates, is a promising approach to meet the stringent worldwide emission standards. Various supported Cu catalysts have shown high selective NO reduction activity with hydrocarbons. These include Cu-ZSM 5, Cu-ZrO2, Cu-Ga2O3, etc.12 CuZSM-5 is the most studied catalyst and has excellent activity for NOx removal in the presence of O2 with various hydrocarbons. However, low activity at low temperature and narrow reaction window prevent the use of Cu-ZSM-5 for NOx reduction in the presence of O2 for low-temperature flue gas cleanup. Other supported Cu catalysts have even lower NOx reduction activity (compared to Cu-ZSM-5) and are subject to poison by SO2.13,14 Supported Pt catalysts have the highest NOx reduction activity at low temperatures; for example, Pt/Al2O3 showed 44% NOx conversion at 200 °C with 0.2 g‚s/cm3 using propene.15 However, the products formed during NOx reduction contain 75% N2O, which is a more harmful greenhouse gas than CO2 and which may contribute to ozone depletion in the upper atmosphere. Although effective NOx reduction has been demonstrated over different catalysts with various hydrocarbon reductants, the temperature required for efficient NOx reduction is much higher than 150 °C, the temperature of the flue gas after the particulate control (9) Burch, R.,Ed. Catal. Today 1995, 26 (2). (10) Heck, R. M.; Farrauto, R. J. Catalytic Air Pollution ControlCommercial Technology; Van Nostrand Reinhold: New York, 1995; pp 192-200. (11) Iwamoto, M., Ed. Catal. Today 1994, 22 (1). (12) Bethke, K. A.; Kung, M. C.; Yang, B.; Shah, M.; Alt, D.; Li, C.; Kung, H. H. Catal. Today 1995, 26, 169-183. (13) Obuchi, A.; Ohi, A.; Nakamura, M.; Ogata, A.; Mizuno, K.; Ohuchi, H. Appl. Catal. B 1993, 2, 71-80. (14) Iwamoto, M.; Yahiro, H.; Tanda, K. Stud. Surf. Sci. Catal. 1988, 37, 219. (15) Hamada, H. Catal. Today 1994, 22, 21-40.

© 1997 American Chemical Society

300 Energy & Fuels, Vol. 11, No. 2, 1997

device. Further development of low-temperature NOx reduction catalysts is needed. Because NOx containing >90% NO and NO2 is more reactive than NO, oxidizing NO to NO2 was investigated as a way of making NOx more reactive with hydrocarbons. Although NO oxidation to NO2 is thermodynamically favorable at flue gas conditions, at temperatures below about 400 °C the rate of reaction is very slow.

NO + 1/2O2 f NO2

∆G°298 ) -8.43 kcal/mol

Generally, there are two types of NO oxidation catalysts reported in the literature: activated carbon16 and inorganic metal oxides.17-23 Carbon catalysts are active for NO oxidation to NO2 at low temperatures, but the activity is low at temperatures above 100 °C. For example, Mochida et al. reported that high surface area carbon yarns are excellent NO oxidation catalysts.16 In the absence of water vapor, high conversion of NO can be obtained at temperatures as low as room temperature. However, water vapor suppresses the NO oxidation activity at low temperatures. The activity increases with increasing temperature, reaching a maximum at about 80 °C, beyond which the activity declines. The decline in activity above 80 °C is due to the oxidation of the carbon via NO2 + C f NO + CO. On the other hand, most metal oxides are active for NO oxidation at temperatures above 250 °C, but the conversion is normally limited by thermodynamics. At temperatures ∼150 °C, only Karisson and Rosenberg reported that NiO-Al2O3, Bi2O3/MoO3-Al2O3, and Fe2O3/ MnO/ZnO are fairly active in the presence of SO2 and H2O.17 However, like many other transition metal oxides,19 some deactivation occurred with time on stream, probably due to the formation of inorganic nitrates.17-20 Therefore, a catalyst is needed for significant conversions of NO to NO2 at low temperatures. This study investigates the low-temperature selective reduction of NO with various hydrocarbons under simulated flue gas conditions over carbon and carbonsupported catalysts because carbon-supported catalysts are reported to be active for low-temperature NOx selective reduction with NH3.24-26 Our NO oxidation investigation focuses on the effect of SO2, H2O, pretreatment, and temperature over carbons and supported Au and cobalt oxide catalysts. (16) Mochida, I.; Kisamori, S.; Hironaka, M.; Kawano, S.; Matsumura, Y.; Yoshikawa, M. Energy Fuels 1994, 8 (6), 1341-1344. (17) Karisson, H. T.; Rosenberg, H. S. Ind. Eng. Chem. Process Des. Dev. 1984, 23, 808-814. (18) Miyadera, T.; Kawai, M.; Hirasawa, S.; Miyajima, K.; Oyama, M.; Kido, N.; Yamasaki, M.; Hattori, H.; Kobayashi, H. Presented at the 32nd Spring Team Annual Meeting of the Japan Chemical Society, Tokyo, Japan, 1975; paper 2005. (19) Kruse, C. W.; Lizzio, A. A.; Debarr, J. A.; Bhagwat, S. B. Prepr. Pap.sAm. Chem. Soc., Div. Fuel Chem. 1995, 40 (4), 838-842. (20) Lawson, A. J. Catal. 1972, 24, 297-305. (21) Arai, H.; Tominaga, H.; Tsuchiya, J. In Proceeding of the 6th International Congress on Catalysis; Royal Chemical Society: London, England, 1976; pp 997-1006. (22) Xue, E.; Seshan K.; Mercera, P. D. L.; Van Ommen, J. G.; Ross, J. R. H. Environmental Catalysts; Armor, J., Ed.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993; pp 250267. (23) Xue, E.; Seshan K.; Van Ommen, J. G.; Ross, J. R. H. Appl. Catal. B 1993, 2, 183-197. (24) Gangwal, S. K.; McMichael, W. J.; Howe, G. B.; Spivey, J. J.; Silveston, P. L. ICHEME Symp. Ser. 1993, No. 131, 181-193. (25) Seki, M.; Sakurai, Y.; Yoshida, K. Prepr. Pap., Natl. Meet.sAm. Chem. Soc., Div. Environ. Chem. 1975, 15 (2), 28-31. (26) Kuehl, H.; Richter, E.; Knoblauch, K.; Juentgn, H. Carbon ’86; Bergbau-Forschung: Essen, Germany, 1986; pp 351-353.

Jang et al.

Experimental Section Four active carbon samples including type GI of Barnebey & Sutcliffe (B&S) Corp. (GI), GmbH from Carbo-Tech (GmbH), PCB from Calgon (PCB), and Rheinbraun AG (AG) and one activated carbon fiber (ACF) (480 m2/g) provided by Professor Mochida (Kyushu University) were used in this experiment. All of the granules were broken into 16/30 mesh particles followed by drying at 120 °C overnight. The carbon-supported catalysts were prepared by pore volume impregnation with copper nitrate, silver nitrate, iron nitrate, and cobalt nitrate from Johnson Matthey. After the impregnation, the carbonsupported catalysts were first dried under vacuum at room temperature for 2 h, then at 60 °C for 2 h, and at 80 °C overnight. Ten weight percent Co3O4 supported on Al2O3 was prepared by impregnation of γ-Al2O3 with cobalt nitrate solution followed by drying and calcination. γ-Al2O3 pellets were also purchased from Johnson Matthey; 5% Au/Co3O4 catalyst was prepared by adding 60 mL of 5.45 × 10-3 M hydrogen tetrachloroaurate(III) hydrate (Aldrich Chemical) and 0.29 M Co(NO3)2 (Aldrich Chemical) drop by drop into 175 mL of 1 M Na2CO3 solution as in the method of Haruta et al.27 The precipitate was suction filtered and washed many times; it was then dried in a 100 °C drying oven and calcined for 5 h at 450 °C. All catalysts were treated with 150 cm3/min of helium at 400 °C for 1 h before reaction. Three pretreatments including heat treatment, H2 reduction, and HNO3 oxidation were applied on PCB carbon for NO oxidation studies. Heat treatment and H2 reduction were carried out at 800 °C in 150 cm3/min of N2 for 4 h, respectively. HNO3 oxidation was tested with concentrated nitric acid solution at 60 °C for 1 h; 10 cm3 of concentrated HNO3 was used per gram of PCB carbon. The oxidized PCB carbon was then washed with distilled water and dried at 120 °C. The schematic of the experimental system used in this study (Figure 1) was also shown in our previous paper.28 Briefly, a Pyrex reactor tube of 1/4 in. on both ends and 3 in. × 3/4 in. on the body was used. Stainless steel (SS) tubing (1/4 in.) was used to connect the Pyrex reactor on both ends to the rest of the system. Two 1/16 in. thermocouples were positioned in the reactor, one at the inlet and one at the outlet. The catalyst bed was positioned in the center between glass/wool supports. An extra SS coil before the inlet reactor was provided to increase the length of the preheat zone in the oven. NO reduction was tested with 400 cm3/min of 0.1% NO, 4% O2, 8% H2O, and hydrocarbons in He over 7 g of catalysts. NO oxidation was carried out with 400 cm3/min flow of 0.1% or 400 ppm NO, 4% O2, and balance He over a various amount of catalysts. The reaction gases were analyzed using an online HewlettPackard Model 5890 gas chromatograph (GC) equipped with a thermal conductivity detector (TCD) and two columns in series. The first column, a 9 ft × 1/8 in. o.d. SS column packed with 60/80 mesh Hayesep R, was used to separate hydrocarbons, CO2, and N2O from O2, N2, and CO. The second column, a 6 ft × 1/8 in. o.d. SS column packed with 45/60 mesh molecular sieve 5A, was used to separate O2, N2, and CO. A chemiluminescent NOx analyzer (Thermo Electric Model 10S) was used to monitor the inlet and outlet NOx concentration continuously. The reaction results are described in terms of NO conversion (oxidation or reduction) and N2 selectivity:

% NO conversion ) (NOin - NOout) × 100/NOin % N2 selectivity ) N2 × 100/(N2 + N2O) (27) Haruta, M.; Yamada, N.; Kobayashi, T.; Iijima, S. J. Catal. 1989, 115, 301-309. (28) Jang, B. W.-L.; Spivey, J. J.; Kung, M. C.; Yang, B.; Kung, H. H. ACS Symp. Ser. 1995, No. 587, 83-95.

Low-Temperature NOx Removal

Energy & Fuels, Vol. 11, No. 2, 1997 301

Figure 1. Schematic of reaction system. Table 1. Surface Areas of the Four Carbons and the NO Reduction Activity with Acetone over Supported 5% Cu Catalyst

Rheinbraun AG Carbo-Tech GmbH Calgon PCB Barnebey Sutcliff GI

surface area (m2/g)

NO reduction (%) 5 h on stream

90 440 927 943

99 >99 >99 >99

a All catalysts were subjected to He treatment at 400 °C for 1 h before reaction. The conversion and selectivity were taken after 5 h on stream. b The concentration of ethyl ether was limited by its solubility in water.

example, it is well-known that the Cu species is highly dispersed on ZSM-5 and the size of CuO clusters is limited by the size of pores in ZSM-5 and Cu loading.29 It is also demonstrated by Kung et al.30 that highly dispersed Cu species on ZrO2 and Ga2O3 are not active in propene combustion, resulting in high activity for selective reduction of NO to N2 in the presence of O2. Normally, carbons with high surface areas would have smaller pores and higher dispersion of Cu; however, we did not measure the dispersion of Cu on carbon supports and the pore size distribution. (2) Reductants. Because the NO reduction activity over Cu supported on PCB carbon was the best among the carbon materials tested, further investigation was focused on PCB-supported Cu-Ag catalysts. The choice of Ag is to minimize the deep oxidation of hydrocarbons to CO2. Table 2 shows the NO reduction activities of 5% Cu-2% Ag/PCB with various hydrocarbons as reductants in the presence of O2 and H2O, along with (29) Kung, M. C.; Bethke, K. A.; Kung, H. H. Prepr.sAm. Chem. Soc., Div. Pet. Chem. 1994, 39 (1), 154-158. (30) Kung, M. C.; Bethke, K. A.; Alt, D.; Yang, B.; Kung, H. H. ACS Symp. Seri. 1995, No. 587, 96-109.

302 Energy & Fuels, Vol. 11, No. 2, 1997

Figure 2. NO reduction conversion over 5% Cu-2% Ag/PCB as function of time with different O2 concentrations. Reaction conditions: 0.1% NO, 0.13% acetone, 8% H2O at 150 °C with 3000 h-1 space velocity.

the results in the absence of any hydrocarbon. The efficiency order of reductant was acetone > 2-propanol > isobutyl alcohol > ethyl ether > propene, ethanol > methanol. The major NO reduction product was N2 with >99% selectivity. With 3000 h-1 space velocity, the NO reduction with acetone was 35% after 5 h on stream. Interestingly, the NO reduction without any reductant was higher than the conversion using propene, ethanol, or methanol as reductant. It is speculated that propene, ethanol, or methanol decreases the direct reduction of NO with carbon which is catalyzed by Cu in the presence of O2 by competitive adsorption. In the case of acetone, 2-propanol, isobutyl alcohol, and ethyl ether, the selective reduction of NO with hydrocarbons in the presence of O2 is the major NO reduction route instead of NO direct reduction with carbon.31 Figure 2 shows that oxygen is required to promote the reduction of NO by acetone. The NO reduction conversion was low in the beginning and decreased to ∼10% after 20 min on stream without O2. With 1% O2, NO reduction conversion was high and slowly decreased to about 25% after 5 h. With O2 concentration of 4%, the reduction activity followed the same trend except with higher NO conversion, roughly 32% after 5 h. This shows that the reaction of NO directly with acetone is not significant in comparison with the NO selective reduction in the presence of O2. Under the reaction conditions, the fast decrease in apparent NO reduction activity during the first hour is not due to the NO adsorption on carbons because the reduction conversion is low in the absence of O2. It is also unlikely that the decrease in NO reduction activity is caused by NO oxidation to NO2 with O2 followed by the adsorption of NO2 because NO2 decomposes quickly to NO on the surface of carbons at 150 °C.32 (3) Effect of H2O and SO2. Most literature shows an inhibiting effect of H2O on NO reduction over various catalysts, e.g., Cu-ZrO232 and Co-ZSM-5.1 In contrast, results on 5% Cu-2% Ag/PCB showed that H2O had no effect on NO reduction.28 Probably the highly hydrophobic property of carbon material surface (compared (31) Yamashita, H.; Tomita, A.; Yamada, H.; Kyotani, T.; Radovic, L. R. Energy Fuels 1993, 7, 85-89. (32) Kung, H. H.; Kung, M. C.; Yang, B.; Spivey, J. J.; Jang, B. W.L. Technical Report for March-May/1994 to the Illinois Clean Coal Institute under ICCI Contract 93-1-3.1A-3P, Urbana, IL, 1994.

Jang et al.

Figure 3. Effect of SO2 on NO reduction with acetone over a Cu supported on PCB carbon catalyst at 150 °C. Reaction conditions: 0.1% NO, 0.13% acetone, 4% O2, and 8% H2O; 3000 h-1.

Figure 4. NO reduction conversion vs time over 5% Cu-2% Ag/PCB and carbon only. Reaction conditions: 0.1% NO, 4% O2, 0.13% acetone, and 8% H2O, 3000 h-1, 150 °C.

to metal oxides) minimizes the inhibition of the reduction reaction by H2O. SO2, on the other hand, has a large negative effect on NO selective reduction with acetone. As shown in Figure 3, the catalyst deactivated very fast with 0.1% of SO2. It is suspected that the presence of SO2 poisons the metals by the formation of metal sulfates which deactivate the NO reduction activity. (4) Effect of Metals. Carbon alone is a catalyst for the selective reduction of NO with ammonia at these low temperatures, though little has been reported on NO reduction by hydrocarbons on carbon catalysts.33 Figure 4 shows the NO reduction conversion versus time for the PCB carbon alone and for the 5% Cu-2% Ag/ PCB catalyst. The enhancing effect of Cu + Ag on the activity of active carbon for NO conversion can be clearly seen with acetone as a reductant. The NO reduction conversion decreased rapidly on PCB active carbon alone (without Cu + Ag) with acetone in the presence of O2 and H2O. NO reduction conversion was negligible after 1 h of reaction. On the other hand, the NO reduction conversion over 5% Cu-2% Ag/PCB slowly decreased to 35% after 5 h. As Figure 2 shows, acetone, in the absence of oxygen, is not an effective reductant for NO reduction, which suggests that the function of (33) Spivey, J. J. Annual Report on Progress in Chemistry, Part C; Royal Society of Chemistry: Cambridge, U.K., 1994; pp 155-176.

Low-Temperature NOx Removal

Energy & Fuels, Vol. 11, No. 2, 1997 303

Table 3. NO Oxidation to NO2 over ACF and PCB Activated Carbon (AC)a oxidation (%) at carbon

30 °C

75 °C

120 °C

ACF PCB AC

90 82

51 42

negligible negligible

a Activities were obtained after 40 h on stream. Reaction conditions: 400 ppm NO, 4% O2, and balance He; 400 cm3/min over 4.0 g of catalysts.

Cu + Ag is to promote the oxidation of acetone to some intermediates which then react with NO. All experiments, with one exception, over 5% Cu-2% Ag/PCB in the presence of O2 showed nearly 100% N2 selectivity; that is, no N2O was formed. The only exception was the NO reduction with acetone in the absence of O2, which showed >80% N2O selectivity, suggesting that the NO reduction pathway in the presence of O2 is different from the pathway in the absence of O2. The formation of N2O is probably from the direct reaction between NO and acetone or from the disproportionation of NO.34 The high selectivity of the 5% Cu-2% Ag/PCB catalyst to N2 may be related to low catalytic activity for the activation of molecular oxygen.7 Other carbon-supported metal catalysts, such as Pt and Pd, were also tested for NO selective reduction with acetone.32 However, the NO reduction activity over supported Pt and Pd catalysts was much lower than the activity with Cu + Ag. Besides lower NO reduction activity with Pt and Pd, lower selectivity to N2 was also noticed. NO Oxidation. (1) Carbon Catalysts. (a) ACF vs PCB Active Carbon. Table 3 shows that both ACF and PCB active carbon have high NO oxidation activities at 30 °C. At this temperature with 400 ppm NO, 4% O2, and balance He, NO conversions are 90% and 82% over ACF and active carbon, respectively. The apparent activity, however, decreases with the temperature. At 75 °C, the NO oxidation conversion decreased to 51% and 42% on ACF and active carbon, respectively. Negligible apparent oxidation was observed at 120 °C. This is consistent with results reported by Mochida et al.16 It is believed that NO oxidation to NO2 and NO2 reduction to NO by carbon are in competition with each other. At low temperatures, NO oxidation is the dominant one and results in high NO conversion. On the other hand, NO2 reduction with carbons is faster than NO oxidation at high temperature; that is, no apparent NO oxidation is observed at 120 °C. Figure 5 shows the NO and NOx (NO + NO2) concentrations versus time during NO oxidation reaction over PCB active carbon at 30 °C. The adsorption/ oxidation pattern is interesting. In the beginning, the adsorption of NOx is severe. The concentration of NO increased with time, and no NO2 (NOx - NO) was detected for the first 7 h. After 7 h, NO2 adsorption sites were saturated and, as a result, NO2 breakthrough was observed. This suggests that NO2 is adsorbed more strongly than NO on carbon materials at low temperatures as reported in refs 35-37. (b) Effect of Pretreatment. Three different pretreatments were tested to increase the NO oxidation (34) Li, Y.; Armor, J. N. Appl. Catal. 1991, 76, L1-L8. (35) Yamashita, H.; Yamada, H.; Tomira, A. Appl. Catal. 1991, 78, L1-L6.

Figure 5. NO and NOx concentrations as function of time during NO oxidation over PCB activated carbon at 30 °C. Reaction conditions: 400 ppm NO, 4% O2, and balance He; 400 cm3/min over 4.0 g of catalysts. Table 4. NO Oxidation to NO2 over PCB Activated Carbon (AC)a oxidation (%) at PCB AC

120 °C

75 °C

fresh heat treatment H2 reduction HNO3 oxidation with 0.1% SO2 with 0.1% SO2 + 10% H2O

negligible negligible negligible negligible negligible negligible

42 35 31 28 negligible not available

a Activities at 120 °C were obtained after 24 h on stream, and activities at 75 °C were obtained after 24 h at 120 °C and 40 h at 75 °C. Reaction conditions: 400 ppm NO, 4% O2, and balance He; 400 cm3/min over 4.0 g of catalysts.

activity of PCB active carbon: heat treatment at 800 °C in N2, reduction with H2 at 800 °C, and oxidation using aqueous HNO3.38 The results are shown in Table 4. None of the pretreatments increased NO oxidation at 120 °C. As reported by Mochida et al.,16 the heat treatment increased the oxidation activity of ACF for NO to NO2 in the presence of H2O. Mochida et al.16 speculated that the increased oxidation activity of ACF is caused by the formation of unsaturated species on the carbon surface by the liberation of CO and CO2 during heat treatment in N2. The removal of the oxygen functional group, which increases the hydrophobic nature of the carbon, also contributes to the increase in NO oxidation activity of carbons in the presence of H2O. Comparing of Tables 3 and 4 for the PCB carbon at 75 °C shows that the heat treatment actually slightly decreased the oxidation activity from 42% to 35% at this temperature. It has been reported that H2 reduction at high temperature can reduce the oxygenated functional groups on carbons.39 The removal of oxygenated functional groups by H2 treatment can result in higher hydrophobicity of carbon surface and may also increase the activity of active carbon for NO oxidation to NO2. To test this hypothesis, PCB carbon pretreated with H2 (36) Cooper, B. J.; Thoss, J. E. SAE Tech. Pap. Ser. 1989, No. 890404, 612-624. (37) Rubel, A. M.; Stencel, J. M. Prepr. Pap.sAm. Chem. Soc., Div. Fuel Chem. 1996, 41 (1), 184-188. (38) Grzybek, T.; Papp, H. Appl. Catal. B 1992, 271-283. (39) Avgul, N. N.; Kisele, A. V.; Kovalyova, N. V.; Khrapova, E. V. In Proceeding of the 2nd International Congress on Surface Activity; Schulman, J. H., Ed.; Academic Press: New York, 1957; Vol. II, p 218.

304 Energy & Fuels, Vol. 11, No. 2, 1997

Jang et al.

Table 5. NO Oxidation to NO2 over PCB Activated Carbon and PCB-Supported Catalystsa oxidation (%) at catalyst

120 °C

75 °C

fresh PCB 1% iron oxide/PCB 5% cobalt oxide/PCB

negligible negligible negligible

42 26 24

a Activities at 120 °C were obtained after 24 h on stream, and activities at 75 °C were obtained after 24 h at 120 °C and 40 h at 75 °C. Reaction conditions: 400 ppm NO, 4% O2, and balance He; 400 cm3/min over 4.0 g of catalysts.

was tested for NO oxidation. The NO oxidation activity, also shown in Table 4, was negligible at 120 °C, but at 75 °C was close to the activity after the heat treatment and was lower than the catalyst without pretreatment. Both heat treatment and H2 reduction treatment were intended to remove the oxygenated functional groups on carbon and to improve the oxidation activity. However, the results showed a negative effect on NO oxidation to NO2. It is likely that the hydrophobic nature of the carbon surfaces can only show advantageous NO oxidation in the presence of H2O but not in the absence of H2O. However, the surface property of PCB carbon was not analyzed to confirm the removal of oxygen functional groups. HNO3 treatment,38 on the other hand, is a well-known procedure to increase the oxygenated functional groups on the surface of carbon materials. Our hypothesis is that if the removal of oxygen functional groups of carbon surfaces does decrease the NO oxidation activity, carbons increasing the oxygen functional groups on carbons should improve it. The effect of HNO3 treatment on NO oxidation activity of PCB active carbon was then tested and summarized in Table 4. The activity was negligible at 120 °C and was only 28% at 75 °C. Thus, all pretreatments resulted in some decrease in the activity of PCB activated carbons for NO oxidation, suggesting that neither the oxygenated functional groups nor the hydrophobicity plays the major role in the oxidation of NO to NO2. (c) Effects of SO2 and H2O. The effect of SO2 on NO oxidation activity over heat-treated PCB active carbons was studied by adding 0.1% SO2 to the inlet gas. The results are also shown in Table 4. The NO oxidation activity was negligible in the presence of SO2 over the catalyst at 120 and 75 °C, compared to 35% NO oxidation conversion in the absence of SO2. SO2 deactivates the carbon catalysts, which is similar to the results reported in the literature.17,21,40 The competitive adsorption of NO and SO2 on carbon surfaces is possibly the cause. (d) Active Carbon-Supported Catalysts. Five percent cobalt oxide and 1% iron oxide supported on PCB active carbons were tested for NO oxidation because cobalt oxide and iron oxide are both active catalysts for NO oxidation at low temperature.17 The results, summarized in Table 5, are compared with the NO oxidation activity of unsupported active carbon. Again, both active carbon-supported catalysts showed negligible activity at 120 °C. It is believed that the reduction of NO2 to NO by carbon is too fast at 120 °C to result in any apparent NO oxidation conversion even (40) Takayasu, M.; An-Nen, Y.; Morita, Y. Rikogatu Weseda Daigaku 1976, 72, 17-21.

Figure 6. NO oxidation to NO2 over 10% Co3O4/Al2O3 as function of time with different SO2 concentrations. Reaction conditions: 0.1% NO, 4% O2, and 10% H2O at 150 °C with 2000 h-1 space velocity.

with high oxidation activity of cobalt oxide and iron oxide. This speculation could have been supported by the formation of COx, but the concentration of COx in the exit gas was not measured. The activities at 75 °C were obtained, followed by the tests at 120 °C. The NO oxidation activities were 26% and 24% over 1% iron oxide/PCB and 5% cobalt oxide/PCB, respectively. The activities of active carbon-supported catalysts were actually lower than the unsupported PCB carbon catalyst at 75 °C. (2) Metal Oxide Catalysts. (a) 10% Co3O4/Al2O3. Figure 6 shows that the 10 wt % Co3O4/Al2O3 was active for NO oxidation at 150 °C. Its activity first increased with time on stream to a maximum of 60% after 6 h and then slowly decreased with longer time on stream to a relatively stable 25%. The effect of SO2 on NO oxidation activity over the 10% Co3O4/Al2O3 catalyst was studied with three different concentrations of SO2 (0%, 0.1%, and 0.2%) in the reactant stream. The result shows that no NO oxidation activity was observed in the absence of SO2, but high activity was obtained with 0.1% SO2. This suggests that the presence of SO2 is necessary for the oxidation of NO to NO2 over 10% Co3O4/Al2O3 catalyst as opposed to the inhibiting effect reported in the literature and our results on PCB carbon (Table 4).17,21,40 The reaction profiles, an initial increase followed by subsequent decline in activity with time on stream, were similar in the presence of 0.1% and 0.2% SO2 in the feed. However, the maximum was higher and reached more quickly for the 0.2% than for the 0.1% SO2 experiment. To understand this activation and deactivation phenomenon, the catalyst was pretreated for different lengths of time in a gas flow of 0.1% SO2, 4% O2, and 10% H2O, with helium as the diluent. Afterward, the NO oxidation activity was tested by adding 0.1% NO into the feed stream. The pretreatment and NO oxidation were both carried out at 150 °C. The results, summarized in Figure 7, showed that the initial NO oxidation activity was a function of the length of pretreatment. The initial NO activities were 0%, 42%, 65%, and 42% for 0, 1, 3, and 16 h of pretreatment, respectively. The NO oxidation activity increased with time on stream for the first 3 h of reaction for the sample with no SO2 pretreatment (Figure 6), while it remained invariant or decreased for all of the samples pretreated with SO2 (Figure 7). This indicates that the chemical

Low-Temperature NOx Removal

Figure 7. Effect of SO2 pretreatment on the activity of NO oxidation to NO2 as function of time over 10% Co3O4/Al2O3. Reaction conditions: 0.1% NO, 4% O2, 0.1% SO2, and 10% H2O with 2000 h-1 space velocity at 150 °C, 400 cm3/min, 7.0 g of catalysts.

Figure 8. NO oxidation to NO2 in the absence of SO2 over 5% Au/Co3O4 as a function of time. Reaction conditions: 400 ppm NO, 4% O2, and 10% H2O, 400 cm3/min, 3.0 g of catalysts. Temperature was varied from 250 to 200 °C, then lowered to 150 °C, and raised back to 200 °C.

reaction of SO2 with cobalt oxide in the presence of O2 is beneficial to the NO oxidation activity and that there is an optimum concentration of such resulting species. It is further supported by the results reported by Pope et al.41 that Co3O4 is capable of incorporating sulfur into its structure during oxidation reactions. However, it is also possible that the reaction pathway is NO oxidation by SO3 to form NO2 and SO2. Although the equilibrium constant for this reaction is 8 × 10-5 at 150 °C, the rapid and favorable subsequent reoxidation of SO2 to SO3 by oxygen in the feed would render the pathway feasible. (b) 5% Au/Co3O4. Catalysts containing Au well dispersed in metal oxides are known to be very active for CO oxidation.27 Because there may be similarities between the mechanism of low-temperature CO oxidation and NO oxidation, a 5 wt % Au/Co3O4 was tested and showed high activity for NO oxidation. Figure 8 shows the activity at a space velocity of about 16 000 h-1 and a feed of 400 ppm of NO and 4% O2 in He with the temperature varied in the following temperature sequence: 250 °C for 450 min, 200 °C for 700 min, 150 °C for 150 min, and then 200 °C for 100 min. Ninetytwo percent NO oxidation conversion was obtained after 7.5 h at 250 °C. The activity decreased when the (41) Pope, D.; Walker, D. S.; Moss, R. L. Atmos. Environ. 1976, 10, 951-956.

Energy & Fuels, Vol. 11, No. 2, 1997 305

Figure 9. NO oxidation to NO2 as function of time over 5% Au/Co3O4 in the presence of SO2 and H2O. Reaction conditions: 400 ppm NO, 4% O2, 0.1% SO2, and 10% H2O, 400 cm3/ min, 3.0 g of catalysts.

temperature was lowered to 200 °C, and it stabilized at 45% oxidation conversion after 12 h on stream. However, the activity dropped to zero within 40 min when the temperature was lowered further to 150 °C. The loss in activity at low temperature was not the result of poisoning, and the activity could be restored by raising the temperature to 200 °C. The effect of SO2 and H2O on NO oxidation was tested with this 5 wt % Au/Co3O4 that had been pretreated with O2 in helium at 430 °C for 1 h. The results, presented in Figure 9, showed that in the presence of SO2 and H2O an NO oxidation conversion of ∼35% was obtained after 4 h at 200 °C. This was lower than the 45% oxidation conversion obtained in the absence of SO2 and H2O. On the other hand, in the presence of SO2 and H2O, 48% NO oxidation conversion to NO2 was observed after 5.5 h on stream at 150 °C. This contrasted dramatically with 0% conversion without SO2 and H2O at this temperature. Further reducing the temperature to 120 °C increased the activity slightly to 50%. The activity was stable for the duration of the test (11 h). These results strongly suggested that Au/Co3O4 is a promising catalyst for NO oxidation at low temperatures (120-150 °C) under flue gas conditions. When SO2 was removed from the feed, the oxidation activity decreased from 50% to 6% within 50 min. Thus, the low-temperature NO oxidation activity of this Au/Co3O4 catalyst, similar to that of the Co3O4/Al2O3, was enhanced by the presence of SO2. However, this catalyst is superior to Co3O4/Al2O3 because of its higher activity and much better stability. Summary Acetone was the most active reductant for NO selective reduction over 5% Cu-2% Ag/PCB among the hydrocarbons tested. Acetone, oxygen, and carbonsupported Cu + Ag catalysts are all required to maintain high NO reduction conversion and high N2 selectivity at these low temperatures. Although 35% NO reduction conversion is relatively high at 150 °C in the presence of O2 and H2O, further improvement is needed for the process to be competitive with other processes. The NO reduction activity of Cu-Ag/PCB with acetone was also greatly decreased by the presence of 0.1% SO2, which may be present even downstream of some flue gas desulfurization systems.

306 Energy & Fuels, Vol. 11, No. 2, 1997

Both ACF and PCB active carbon are active for NO oxidation to NO2 at low temperature. At 30 °C and 0.01 g‚min/cm3 space velocity, the NO oxidation conversions to NO2 were 90% and 82% on ACF and active carbon, respectively. However, the activity decreased with the temperature, becoming negligible at 120 °C on both ACF and active carbon. Different treatments (heat treatment at 800 °C, reduction in H2, and oxidation with HNO3) and additional metals did not improve the NO oxidation activities of carbons at 120 °C. With the addition of 0.1% SO2, the NO oxidation activity of activated carbon was poisoned and the oxidation conversion was negligible at 75 °C. In the case of 10% Co3O4/Al2O3, SO2 has a promoting effect for NO oxidation to NO2 at 150 °C. Although the activity can be as high as 60% in the presence of 0.2% SO2, it decreases to ∼30% after 20 h on stream. The initial NO oxidation activity can be greatly enhanced by the pretreatment in a stream of 0.1% SO2, 4% O2, and 10% H2O. The results suggest possible active species might be formed from the reaction of SO2 with cobalt oxide in the presence of O2. Five percent Au/ Co3O4 showed 92% oxidation conversion of NO to NO2

Jang et al.

at 250 °C but was inactive at 150 °C in the absence of SO2. However, when 0.1% SO2 was added, a 48% NO conversion to NO2 was observed at 120 °C with 40 000 h-1. The activity was stable for at least 15 h. Au/Co3O4 has a potential to be commercialized for NOx cleanup in the presence of SO2 at temperatures between 100 and 200 °C. Acknowledgment. We acknowledge Mr. Robert Nelson for carrying out the experiments and Mr. David Alt for the preparation of Au/Co3O4 catalysts. This work was prepared with support in part by grants made possible by the Illinois Department of Energy and Natural Resources through its Coal Development Board and Illinois Clean Coal Institute and by the U.S. Department of Energy (Contract 94-1/2.1A-2P). However, any opinions, findings, conclusions, or recommendations expressed herein are those of the authors and do not necessarily reflect the views of the IDENR, ICCI, and DOE. EF960138W