Energy Efficient Production of Hydrogen and Syngas from Biomass

The Rh/CeO2/M (M = SiO2, Al2O3, and ZrO2) type catalysts with various compositions ... However, the use of cellulose or cellulosic biomass for energy ...
19 downloads 0 Views 215KB Size
Download PDF
Environ. Sci. Technol. 2002, 36, 4476-4481

Energy Efficient Production of Hydrogen and Syngas from Biomass: Development of Low-Temperature Catalytic Process for Cellulose Gasification MOHAMMAD ASADULLAH,† SHIN-ICHI ITO,† KIMIO KUNIMORI,† MUNEYOSHI YAMADA,‡ AND K E I I C H I T O M I S H I G E * ,† Institute of Materials Science, University of Tsukuba, 1-1-1, Tennodai, Tsukuba-shi, Ibaraki 305-8573, Japan, and Department of Applied Chemistry, Graduate School of Engineering, Tohoku University, Aoba, Aramaki, Aoba-ku, Sendai 980-8579, Japan

The Rh/CeO2/M (M ) SiO2, Al2O3, and ZrO2) type catalysts with various compositions have been prepared and investigated in the gasification of cellulose, a model compound of biomass, in a fluidized bed reactor at 500700 °C. The conventional nickel and dolomite catalysts have also been investigated. Among the catalysts, Rh/CeO2/ SiO2 with 35% CeO2 has been found to be the best catalyst with respect to the carbon conversion to gas and product distribution. The steam addition contributed to the complete conversion of cellulose to gas even at 600 °C. Lower steam supply gave the syngas and higher steam supply gave the hydrogen as the major product. Hydrogen and syngas from cellulose or cellulosic biomass gasification are environmentally super clean gaseous fuels for power generation. Moreover, the syngas derived liquid fuels such as methanol, dimethyl ether, and synthetic diesels are also super clean transportation fuels. However, the use of cellulose or cellulosic biomass for energy source through the gasification is challenging because of the formation of tar and char during the gasification process. It is interesting that no tar or char was finally formed in the effluent gas at as low as 500-600 °C using Rh/CeO2/SiO2(35) catalyst in this process.

Introduction The increasing concentration of the CO2 in the atmosphere, which is mainly coming from the fossil fuel burning, causes faster global warming than any time in the past 1000 years (1). Moreover, the NOx, SOx, and particulate matters are very harmful; however, these are increasing in the atmosphere day by day by burning of fossil fuels for energy. Thus, to create a safe environment for human beings, our future energy sources must be clean for environment, renewable and sustainable, efficient and cost-effective, and convenient and safe (2-4). Syngas (CO + H2) for clean liquid fuels such as Fischer-Tropsch liquids, methanol, and dimethyl ether and pure hydrogen for fuel cells and internal combustion * Corresponding author phone and fax: e-mail: [email protected]. † University of Tsukuba. ‡ Tohoku University. 4476

9

+81-298-53-5030;

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 20, 2002

engines are considered to be the most promising energy sources. This is because theses energy sources offer large potential benefits in terms of reduced emissions of pollutants and greenhouse gases (5, 6). Actually, the syngas production is the key step to produce such super clean liquid fuels. The reforming or partial oxidation of hydrocarbons (7-9) is the most common process for the hydrogen and syngas production. However, the use of biomass to produce syngas is the most promising option to share renewable energy sources and to decrease the fossil fuel dependence. A great deal of research has been done in the field of biomass gasification within the past two decades. Most of these studies have utilized fluidized or fixed-bed reactor using air, steam, and/or mixture of steam and oxygen (10-13). From these studies, fluidized bed gasification has been found to be advantageous (14). However, these processes still provide significant amounts of tar (a complex mixture of higher hydrocarbons) in the product gas even it operated at 800-1000 °C. Thus, the product gas cleaning is a challenging point for most of its application. Catalytic cleaning and upgrading of hot dry gas is nowadays the best solution. Calcined dolomites and steam reforming nickelbased catalysts are now being considered as the most common catalysts. These catalysts are usually used in the secondary reactor for hot dry gas clean up (15, 16). Nickel based catalyst is very effective for tar conversion in the secondary reactor at around 700-800 °C, resulting about 98% tar removal from product gas (17). The use of catalyst in the primary reactor simplifies the overall process; however, a limited number of works focused on the direct use of catalyst in the primary bed. Although the nickel-based catalysts have been found to be effective in primary reactor to reduce tar content in the product gas at above 750 °C (18, 19), the catalysts deactivate suddenly because of the carbon deposition on the surface (20). Our interest in this work is to discover an efficient catalyst for complete gasification of cellulose and/or real biomass to gas in primary reactor at around 500-600 °C. However, it is challenging because of severe tar formation at this low temperature, causing serious deactivation of conventional catalysts. Investigation of the efficient catalyst may be the option to overcome such a problem of low-temperature biomass gasification. In the previous report, we have investigated the active catalyst for the gasification of cellulose as a model compound of biomass in a laboratory scale batchfeeding fluidized-bed reactor (21). The study offered us the Rh/CeO2 as a novel catalyst for complete gasification of cellulose to gas at 550 °C. However, the batch-feeding process is not a proper way to evaluate the catalyst performance and is less important in the practical use. Moreover, the Brunauer-Emmett-Teller (BET) surface area of Rh/CeO2 drastically decreased during the reaction (22). Thus, we have further developed the catalyst and the reaction system for continuous-feeding fluidized-bed reactor. This article describes the gasification of model compound, cellulose, to hydrogen and syngas using various types of Rh/CeO2/M catalysts and compares with the results of Rh/CeO2, Rh/SiO2, commercial steam reforming catalyst G-91, and dolomite catalyzed and noncatalyzed reactions.

Experimental Section Procedure. The gasification of cellulose powder (Merk) has been carried out in a continuous feeding fluidized-bed reactor. The operational system is almost the same as before (21); however, the reactor dimension and feeding system have been modified for continuous-feeding gasification. Here 10.1021/es020575r CCC: $22.00

 2002 American Chemical Society Published on Web 09/10/2002

FIGURE 1. Dependence of time on stream on C-conversion and product distribution in cellulose gasification on (a) Rh/CeO2/SiO2(35), (b) Rh/SiO2, (c) G-91, (d) Rh/CeO2, (e) Rh/CeO2/Al2O3(30), and (f) Rh/CeO2/ZrO2(50) at 500 °C. /: C-conversion, 9: CO, [: H2, 4: CO2, and O: CH4. the gasification reactor is a quartz tube of 66 cm high and 1.5 cm i.d. The feeder is a conical glass vessel with a small opening at the bottom, which can be controlled by a screw plug. The cellulose particles can flow through the opening by vibrating the vessel with an electric vibrator. Controlling the opening size and vibration rate controls the cellulose feeding rate. In this system the feeding rate was controlled around 85 mg/min (C, 3148 µmol/min: H, 5245 µmol/min and O, 2622 µmol/min). The feeding rate was corrected by the difference of feeder weight before and after the reaction. Cellulose particles were transported to the catalyst bed by the flow of 51 cm3/min N2 gas at 25 °C and 1 atm pressure (2087 µmol/min) through an inner tube of 7 mm i.d. Air of 51 cm3/min at 25 °C and 1 atm (O2, 417 µmol/min) pressure and steam were introduced from the bottom and reached to the catalyst bed through a quartz distributor. The water was supplied in the hot zone in the reactor from the micro feeder. About 3 g of catalyst was used in the fluidized bed. At the initial run, the catalyst was pretreated with the flow of H2 of 40 cm3/min at 500 °C for 30 min. The temperature gradient between the out and inside of the reactor was measured by thermocouples. The concentrations of CO, CO2, and CH4 were determined by FID-GC equipped with a methanator by using a stainless steel column packed with Gaskuropack 54 and the concentration of hydrogen was determined by TCDGC using stainless steel column packed with Molecular sieve 13X. Carbon conversion (C-conv) was calculated from the following equation: (formation rate of carbon in CO + CO2 + CH4)/(total C feeding as cellulose) × 100. C-conv and the formation rate are average during 25 min reaction. The char was determined by the following equation: (carbon in

the formation of CO2 after feeding was stopped)/(total C feeding as cellulose) × 100. The tar was calculated by the following equation: 100 - C-conv% - char%. Catalyst. CeO2/M type supports in various compositions were prepared by the incipient wetness method using the aqueous solution of Ce(NH4)2(NO3)6 and M (SiO2, Aerosil 380, 200 and 50 m2/g), Al2O3 (Aerosil, 100 m2/g), and ZrO2 (Daiichi-Kigenso Co. Ltd., 100 m2/g)). The pure CeO2 (70 m2/g) was also obtained from Daiichi-Kigenso Co. Ltd. After drying at 110 °C for 12 h, the CeO2/M supports were calcined at 500 °C for 3 h under air atmosphere. The method of Rh loading is the impregnation of the support with acetone solution of Rh(C5H7O2)3. The size of the catalyst particles is in the range of 150-250 µm. The Brunauer-Emmett-Teller (BET) surface areas of fresh (after H2 treatment) and used catalysts were determined by a Gemini (Micrometrics). The detail of the reaction conditions is described in the results.

Results and Discussion Efficient and economic production of hydrogen and syngas from biomass gasification needs complete conversion of biomass to gas and high selectivity of useful gas at low temperature. The novel catalyst with high performance and a suitable reactor could meet the requirements of efficient gasification of biomass. We have recently developed Rh/ CeO2 as an efficient catalyst for cellulose gasification in a batch feeding fluidized-bed reactor (21). However, the catalyst has a serious problem in the catalyst life (22), and this is very common in the catalytic gasification of biomass (23). The surface area of the catalyst drastically decreased because of VOL. 36, NO. 20, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4477

TABLE 1. Performance of Various Catalysts in the Gasification of Cellulose formation rate (µmol/min) catalystc none

dolomite

Rh/CeO2 G-91 Rh/SiO2 Rh/CeO2/SiO2(10) Rh/CeO2/SiO2(20) Rh/CeO2/SiO2(35) Rh/CeO2/SiO2(50) Rh/CeO2/SiO2(80) Rh/CeO2/SiO2(30)a Rh/CeO2/SiO2(10)b Rh/CeO2/ZrO2(10) Rh/CeO2/ZrO2(50) Rh/CeO2/Al2O3(20) Rh/CeO2/Al2O3(30) a

SiO2, 200 m2/g.

b

T (°C)

CO

H2

CH4

CO2

H2/CO

C-conv (%)

500 550 750 800 900 550 700 800 900 500 500 500 500 500 500 500 500 500 500 500 500 500 500

152 240 1536 1714 1943 414 1149 1383 1656 1158 477 970 546 516 845 927 975 1189 1321 842 886 399 448

24 76 456 505 592 112 892 1072 1442 1764 964 838 777 742 1077 1200 1370 1684 1295 816 1212 613 836

5 15 357 462 499 72 294 410 515 35 284 128 377 648 676 750 625 141 170 506 548 574 585

569 562 457 417 455 747 336 833 750 898 1202 632 1255 1253 1178 999 912 1049 710 949 897 1177 1364

0.2 0.3 0.3 0.3 0.3 0.3 0.8 0.8 0.9 1.5 2.0 0.9 1.4 1.4 1.3 1.3 1.4 1.3 1.0 1.0 1.4 1.2 1.8

23 26 65 82 92 39 57 83 93 67 62 55 69 77 86 85 79 76 70 73 74 68 76

9

tar (%)

9 7 4 3 2 34 14 4 2 11 18 9 11 7 6 6 5 15 13 16 14 17 14

68 67 31 15 6 25 29 13 5 22 20 36 20 16 8 9 16 9 17 11 12 15 10

fresh

used

59

13

312 285 250 208 183 82 180 62 87 74 66 63

310 277 247 206 176 77 177 58 86 76 61 56

SiO2, 50 m2/g. c Values in parentheses are the mass% of CeO2 in the support.

sintering of the CeO2 during reaction. Thus, we have modified the support as CeO2/M (M ) SiO2, Al2O3, and ZrO2) in order to enlarge the surface area and to protect the sintering of CeO2. A number of various kinds of Rh/CeO2/M catalysts were prepared and tested in the cellulose gasification. Figure 1 represents the dependence of formation rate of product gases and C-conv with respect to the time on stream. Although the reaction was carried out at 500 °C, the C-conv and formation rate over Rh/CeO2/SiO2(35) was very stable during 25-min reaction time (Figure 1a). The deposited carbon on the catalyst surface as coke/char was burned when cellulose feeding was stopped after 25 min. The CO2 formation by burning of coke/char was measured, and it was decreased with time on stream. The decreasing of CO2 formation over Rh/CeO2/SiO2(35) was much higher than that over other catalysts. This indicates the high combustion activity of Rh/ CeO2/SiO2(35). In contrast, the carbon conversion (C-conv) on Rh/SiO2 suddenly decreased within several minutes (Figure 1b). This may be due to the deactivation of Rh/SiO2 because of coke/char deposition on the surface. Since the methanation reaction usually proceeds on the clean metal surface, the sudden fall of the methane formation also indicates that the catalyst surface was covered by the coke/ char. The C-conv and product gas formation on G-91 gradually decreased with time on stream except the CO2 formation (Figure 1c). The G-91 and Rh/SiO2 showed much lower activity to char combustion. And thus, the CO2 formation due to combustion of deposited coke/char over Rh/SiO2 and G-91 continued for much longer time than that over Rh/CeO2/SiO2(35). As Table 1, the C-conv was lower on Rh/CeO2 at 500 °C. The increase of CO2 formation rate can be related to the increase of coke/char deposition with time on stream (Figure 1d). On the Rh/CeO2/Al2O3 and Rh/CeO2/ZrO2, C-conv decreased with time on stream (Figure 1e,f). The rate of char combustion on Rh/CeO2/Al2O3 and Rh/CeO2/ZrO2 was much lower than that of Rh/CeO2/SiO2(35). In terms of the stability of formation rate and high combustion activity, Rh/CeO2/ SiO2(35) exhibited the highest performance. Furthermore, it gave significant amounts of hydrogen and CO. The BET surface area of Rh/CeO2/SiO2(35) remained almost constant 4478

surface area (m2/g)

char (%)

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 20, 2002

after several hours of reaction (Table 1). This may be due to the interaction of CeO2 and SiO2, which prohibited the sintering of CeO2. As a result the Rh particles remained dispersed on the CeO2 surface and worked efficiently (24). The use of SiO2 with lower surface area such as 200 m2/g and 50 m2/g in Rh/CeO2/SiO2 decreases the performance. This may be due to the lower dispersion of Rh on comparatively lower surface of CeO2/SiO2. At 500 °C the level of C-conv is 23%, and the hydrogen formation is almost the negligible (24 µmol) in the noncatalyst system. In this case the major products were the char and tar. The increasing of temperature increases the C-conv in the 90% level at 900 °C. At this temperature, although the CO formation was increased remarkably, the H2 formation was hardly increased. Although the other catalysts such as dolomite, G-91, Rh/ CeO2, Rh/SiO2, Rh/ZrO2, and Rh/Al2O3 rather enhanced the C-conv than noncatalyst, the level of C-conv is much lower than that of Rh/CeO2/SiO2(35). In addition, the char and tar formation in these cases are much higher than Rh/CeO2/ SiO2. Since the air was used as a gasifying agent and N2 was used for cellulose transportation, the product gas contains almost 50% N2. This diluted gas may be used for ammonia production. For the production of pure syngas for methanol, dimethyl ether, and synthetic fuels syntheses, the pure oxygen and steam may be used. In fact we have already used pure oxygen for real biomass gasification and obtained very promising results. The temperature dependence of gas formation rate and C-conv on various catalysts is shown in Figure 2. The C-conv on Rh/SiO2 (b) was 55% at 500 °C and 91% at 700 °C. However, these values slightly increased on G-91 (c) (62% at 500 °C and 93% at 700 °C, respectively). Rh/CeO2 (d), Rh/CeO2/ Al2O3 (e), and Rh/CeO2/ZrO2 (f) exhibited higher C-conv than Rh/SiO2 and G-91. This indicates that CeO2 is very effective to the enhancement of C-conv. Interestingly enough, the C-conversion was 67% at 500 °C, which reaches to about 100% at 650 °C on Rh/CeO2/SiO2(35) (a). Higher temperature gave higher yield of CO and H2 where the formation of CO2 and CH4 decreased. Table 2 shows the temperature, where the C-conv reaches 90% over various catalysts, which can be expected from the results of temperature dependence (Figure

FIGURE 2. Effect of temperature on the C-conversion and product distribution in cellulose gasification on (a) Rh/CeO2/SiO2(35), (b) Rh/SiO2, (c) G-91, (d) Rh/CeO2, (e) Rh/CeO2/Al2O3, and (f) Rh/CeO2/ZrO2. /: C-conversion, 9: CO, [: H2, 4: CO2, and O: CH4.

TABLE 2. Reaction Temperature for 90% C-conv by Various Catalysts Expected from the Results of Figure 2 Figure 2

catalyst

temp (°C)

a b c d e f

Rh/CeO2/SiO2(35) Rh/SiO2 G-91 Rh/CeO2 Rh/CeO2/Al2O3(30) Rh/CeO2/ZrO2(50)

522 687 632 572 602 627

2). The temperature of Rh/CeO2/SiO2 (35) for 90% C-conv was much lower than that of other catalysts. Figure 3 shows the performance of Rh/CeO2/SiO2 with various mass% of CeO2 on SiO2 (380 m2/g). The C-conv increased to 86% at 500 °C with increasing CeO2 content until 35 mass%. The higher loading of CeO2 than 35% resulted in lower C-conv. However, the char formation gradually decreased with increasing the CeO2 content. This suggests that the CeO2 contributes to the char conversion. The tar formation decreased until 35 mass% of CeO2 and increased when 80 mass% of CeO2 was used. This suggests that the tar conversion needs high surface of CeO2, where Rh can be dispersed properly. As described below, the tar is converted by the reforming with steam. Since SiO2 only ref 21 and also Rh/SiO2 exhibited very poor performance with respect to the C-conv as well as hydrogen and CO formation, the lower loading of CeO2 such as 10 and 20 mass% could not overcome the negative factor of SiO2 on Rh/CeO2/SiO2 in cellulose gasification. Furthermore, the higher loading of CeO2 causes

FIGURE 3. Effect of the amount of CeO2 loading in the Rh/CeO2/SiO2 catalyst with respect to the C-conversion as well as tar and char formation. /: C-conversion, 2: tar-formation and 9: char-formation. Reaction time was 25 min. the crystal formation of CeO2 on SiO2, which decreases the dispersion of CeO2 and Rh. As a result the performance decreases. Figure 4 shows the effect of steam in the cellulose gasification. In the absence of steam, 97% C-conv was achieved at 600 °C; however, 100% C-conv was achieved when at least 1111 µmol/min steam was introduced. The introduction of steam in the reaction system dramatically changed the product distribution. When there was no steam supply, the H amount in the cellulose and in the product H2 and CH4 VOL. 36, NO. 20, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4479

FIGURE 4. Performance of steam in the Rh/CeO2/SiO2 (35) catalyzed gasification of cellulose at 600 °C. /: C-conversion, 9: CO, [: H2, 4: CO2, and O: CH4.

FIGURE 5. Performance of air flow rate in the Rh/CeO2/SiO2(35) catalyzed gasification of cellulose at 600 °C: (a) steam, 832 µmol/ min and (b) steam 5555 µmol/min. /: C-conversion, 9: CO, [: H2, 4: CO2, and O: CH4. was almost the same. This indicates that there was no H2O formed. When 1111, 2222, and 4444 µmol of external steam was introduced 470, 1226, and 3111 µmol of steam were formed in the product gas. The differences of these values indicate that the water gas shift reaction (CO + H2O ) CO2 + H2) proceeded in the presence of steam. As a result the formation of H2 and CO2 constantly increased and conversely the formation of CO decreased with increasing the flow rate of steam. Furthermore, the formation of methane gradually decreased with increasing flow rate of steam because of steam reforming of methane. This also indicates that the composition of product gas is adjustable to that we need. For the syngas production, we carried out experiments with various rates of air flow, when the steam flow rate was lower (833 µmol/min) (Figure 5a). On the other hand, for the hydrogen production the steam flow rate was higher (5555 µmol/min) (Figure 5b). The C-conv is a function of air and steam. This was considerably low (86%) in the case of the 4480

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 20, 2002

FIGURE 6. Relation between hydrogen yield and C yield in useful products in the gasification of cellulose over Rh/CeO2/SiO2(35) under various reaction conditions. O: Figure 4, 4: Figure 5a and 0: Figure 5b. lower amount of steam (833 µmol/min) in the absence of air. In contrast, C-conv drastically increased by the addition of air. This is because the reactivity of the char with air is much higher than that with steam. By the air addition, the CO2 formation rate was increased significantly and this is related to the decrease of the CO formation rate. Methane formation decreased slightly with increasing air flow. In the presence of a large amount of steam, the C-conv was high without the addition of air (Figure 5b). This indicates that the char can be gasified under high partial pressure of steam. The introduction of small air flow with higher steam greatly varied the product distribution. Hydrogen formation suddenly increased to the maximum value when 1043 µmol of air was introduced. However, it decreased gradually with a further increase of the air flow. In the presence of a large excess of oxygen, the tar preferably takes part in the combustion reaction. Consequently, the formation of hydrogen and CO decreased with increasing the air flow. By the same reason the CO2 formation increased and methane formation decreased with increasing the air flow. Here the hydrogen, CO, and methane are thought to be useful products. This is because the syngas can be converted to liquid fuels, and methane can be used as a gas fuel. Therefore, we compare the yield of them under various reaction conditions. Figure 6 shows the relation between H2 + CH4 and CO + CH4 yields from cellulose, steam, and air over Rh/CeO2/SiO2(35). Under some reaction conditions, the fact that the hydrogen yield approaches unity implies that the pyrolysis of the cellulose was completed and that the almost negligible or small final yield of tars and char contain very little hydrogen (because they are aromatic in nature). From this the effective conversion to hydrogen can be realized even at low temperature. At the same time, the efficiency with respect to carbon is also high. In our gasification system Rh/CeO2/SiO2(35) catalyst has been thought to be multifunctional (Figure 7). The fluidized catalyst bed can be divided into three sections such as pyrolysis zone, combustion zone, and reforming zone. Cellulose particles are first fed to the oxygen- and steam-free pyrolysis zone, where the thermal cracking of cellulose to tar, char, steam, and a small fraction of gas proceeded. The pyrolyzed products then contacted with catalyst particles at the lower part of the fluidized bed, where oxygen is present. Since the catalyst remains in the oxidized state in this region, a part of the tar and char are burnt to CO2 and H2O. Then the catalyst particles moved upward by the fluidization and are reduced by the hydrogen and CO. The reduced catalyst then contributed to the steam reforming of tar and char to

achieved in this reported work when catalyst to biomass feeding rate was 0.83 h. However, in our investigation, 100% C-conv with no tar or char resulted even when the catalyst to biomass feeding rate was 0.58 h. Since the use of syngas and hydrogen needs to be completely free of tar, our results have paramount importance in the utilization of product gases.

Acknowledgments This research was supported by the Future Program of Japan Society for the Promotion of Sciences under the Project “Synthesis of Ecological High Quality of Transportation Fuels” (JSPS-RFTF98P01001).

Literature Cited

FIGURE 7. Function of Rh/CeO2/SiO2(35) catalyst and fluidized bed reactor in the catalytic gasification of cellulose with air and steam. produce CO and hydrogen. A number of secondary reactions are also proceeded in the reforming zone. The CO2 and H2 can also be formed by the water-gas shift reaction (CO + H2O ) CO2 + H2). Methane can be formed by methanation reaction (CO + 3H2 ) CH4 + H2O). On the highly active Rh/CeO2/SiO2(35) catalyst the tar is thought to be totally converted; however, some char may be remained on the catalyst surface in the reforming zone. Since the catalyst bed existed in the fluidized condition, the catalyst could interact further with oxygen at the lower part of the reactor and contributed to the further combustion of the rest of the char remained on the catalyst surface. It has been reported that the fluidized-bed reactor is very effective to remove the low reactive carbonaceous species in methane reforming with CO2 and O2 (25, 26). High performance of Rh/CeO2/SiO2(35) catalyst is thought to be due to the smooth redox properties in the combination of Rh and CeO2. In terms of the stability, Rh/CeO2/SiO2(35) also has high performance. The SiO2 prohibits sintering of CeO2, and, thus, Rh metal particles remained highly dispersed on CeO2. In addition, the fluidized bed reactor can promote the heat transfer from exothermic zone to endothermic zone and make the reactor temperature homogeneous. In our system, the lower part of the reactor is an exothermic zone, and the upper part of the reactor is an endothermic zone. However, the temperature difference between lower and upper parts of the reactor was not significant (only 15 °C). Finally, the Rh/CeO2/SiO2 catalyst with 35 wt % of CeO2 exhibited the excellent performance for cellulose gasification at 500-600 °C. Almost complete C-conv was achieved when at least 1111 µmol/min of steam was introduced with 51 cm3/min of air for the gasification of 85 mg/min at 600 °C. The catalyst deactivation was not observed even after 8 h reaction time. However, the Ni-based catalyst was deactivated within a matter of some minutes when it was used in the primary bed even at 750 °C (20, 23). About 90% C-conv was

(1) Intergovernmental Panel on Climate Change, Climate Change 2001; Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linden, P. J., Xiaosu, D., Eds.; The Scientific Basis Cambridge University Press: Cambridge, UK, 2001; pp 944, 2001. (2) Chum, H. L.; Overend, R. P. Fuel Proc. Technol. 2001, 71, 187. (3) Scarpellini, S.; Romeo, L. M. Energy Conv., Management 1999, 40, 1661. (4) Lynch, M. C. Appl. Energy 1999, 64, 31. (5) Ogden, J. M. Annual Rev. Energy Environ. 1999, 24, 227. (6) Steinberg, M. Int. J. Hydrogen Energy 1999, 24, 771. (7) Pena, M. A.; Gomez, J. P.; Fierro, J. L. G. Appl. Catal. A: General 1996, 144, 7. (8) Armor, J. N. Appl. Catal. A: General 1999, 176, 159. (9) Steinberg, M.; Cheng, C. H. Int. J. Hydrogen Energy 1989, 14, 797. (10) Gil, J.; Corella, J.; Aznar, M. P.; Caballero, M. A. Biomass Bioenergy 1999, 17, 389. (11) van der Drift, A.; van Dorn, J.; Vermeulen, J. W. Biomass Bioenergy 2001, 20, 45. (12) Blasi, C. D.; Signorelli, G.; Portoricco, G. Ind. Eng. Chem. Res. 1999, 38, 2571. (13) Rapagna, S.; Jand, N.; Kiennemann, A.; Foscolo, P. U. Biomass Bioenergy 2000, 19, 187. (14) Warnecke, R. Biomass Bioenergy 2000, 18, 489. (15) Caballero, M. A.; Corella, J.; Aznar, M. P.; Gil, J. Ind. Eng. Chem. Res. 2000, 39, 1143. (16) Garcı´a, L.; Salvador, M. L.; Arauzo, J.; Bilbao, R. Fuel Processing Tech. 2001, 69, 157. (17) Caballero, M. A.; Corella, J.; Aznar, M. P.; Gil, J. Ind. Eng. Chem. Res. 2000, 39, 1143. (18) Rapagna, S.; Jand, N.; Foscolo, P. U. Int. J. Hydrogen Energy 1998, 23, 551. (19) Arauzo, J.; Radlein, D.; Piskorz, J.; Scott, D. S. Ind. Eng. Chem. Res. 1997, 36, 67. (20) Baker, E. G.; Mudge, L. K.; Brown, M. D. Ind. Eng. Chem. Res. 1987, 26, 1335. (21) Asadullah, M.; Tomishige, K.; Fujimoto, K. Catal. Commun. 2001, 2, 63. (22) Asadullah, M.; Fujimoto, K.; Tomishige, K. Ind. Eng. Chem. Res. 2001, 40, 5894. (23) Baker, E. G.; Mudge, L. K.; Wilcox, W. A. Catalysis of gas-phase reactions in steam gasification of biomass. In Fundamentals of Thermochemical Biomass Conversion; Overend, R. P., et al., Eds.; Elsevier Applied Science: London, 1985; p 863. (24) Bensalem, A.; Bozon-Verduraz, F.; Delamar, M.; Bugli, G. Appl. Catal. A: General 1995, 121, 81. (25) Tomishige, K.; Matsuo, Y.; Sekine, Y.; Fujimoto, K. Catal. Commun. 2001, 2, 11. (26) Tomishige, K.; Matsuo, Y.; Yoshinaga, Y.; Sekine, Y.; Asadullah, M.; Fujimoto, K. Appl. Catal. A: General 2002, 223, 225.

Received for review February 5, 2002. Revised manuscript received June 25, 2002. Accepted July 25, 2002. ES020575R

VOL. 36, NO. 20, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4481