Biomass gasification to hydrogen and syngas at low temperature: Novel catalytic system using fluidized-bed reactor

Article abstract of DOI:10.1006/jcat.2002.3575

A novel catalytic process for hydrogen and synthesis gas (syngas) production from biomass at very low temperature using an excellent catalyst (Rh/CeO₂/SiO₂) in a continuous-feeding fluidized-bed reactor using cellulose as a model compound was developed. The catalyst simultaneously promoted the reforming and combustion reactions of pyrolyzed products of cellulose, e.g., char and tar, and the water-gas shift reaction. In the presence of a mixture of steam and air, the cellulose completely converted to gas products at 773 K. The study represents a new and energy efficiency approach of synthesis of hydrogen and syngas from biomass.

Full text of DOI:10.1006/jcat.2002.3575

Journal of Catalysis 208, 255–259 (2002)
doi:10.1006/jcat.2002.3575
PRIORITY COMMUNICATION
Biomass Gasification to Hydrogen and Syngas at Low Temperature:
Novel Catalytic System Using Fluidized-Bed Reactor
∗
∗
∗
∗,1
Mohammad Asadullah, Shin-ichi Ito, Kimio Kunimori, Muneyoshi Yamada,† and Keiichi Tomishige
∗
Institute of Materials Science, University of Tsukuba, 1-1-1, Tennodai, Tsukuba, Ibaraki 305-8573, Japan; and †Department of Applied Chemistry,
Graduate School of Engineering, Tohoku University, Aoba 07, Aramaki, Aoba-ku, Sendai 980-8579, Japan
Received December 21, 2001; revised January 29, 2002; accepted February 13, 2002
antly available everywhere in the world. Although biomass
can be converted to hydrogen and syngas, the process is
problematic because of the formation of tar (a complex
mixture of higher hydrocarbons) and char even in the pro-
cess operated at very high temperature (>1173 K). When
the process is carried out at lower temperature (<1123 K)
to get higher energy efficiency, more tar and char are pro-
duced. Forexample, aNicatalystforthehydrocarbonsteam
reforming was used in the secondary reactor for bio-oil re-
forming to hydrogen; however, the catalyst suddenly deac-
tivated. Thus the process remains problematic (21).
Biomass gasification to produce hydrogen and/or syn-
gas (H2 + CO) has been considered to be a promising
method for future energy systems to meet environmental
requirements (1, 11, 21). However, the gasification needs
complete conversion of the biomass to gas as well as high
selectivity of the useful gas. Especially, a novel catalyst with
high performance and a suitable reactor are necessary for
a highly efficient low-temperature process. We have devel-
oped, and describe here, such a process for cellulose gasifi-
cation using the Rh/CeO2/SiO2 catalyst in a continuous-
feeding fluidized-bed reactor at temperatures as low as
773 K.
We have developed a novel catalytic process for hydrogen and
syngas production from biomass at very low temperature using
an excellent catalyst (Rh/CeO2/SiO2) in a continuous-feeding
fluidized-bed reactor using cellulose as a model compound. The
catalyst simultaneously promoted the reforming and combustion
reactions of pyrolyzed products of cellulose, such as tar, char, and
so forth, and the water–gas shift reaction. In the presence of a mix-
ture of air and steam, the cellulose completely converted to gas
products at 773 K. This work can represent a new and energy
efficient approach to the synthesis of hydrogen and syngas from
c
biomass. ꢀ 2002 Elsevier Science (USA)
Key Words: biomass gasification; catalytic gasification; hydro-
gen; syngas; fluidized-bed reactor.
INTRODUCTION
Hydrogen is nowadays the most promising energy source
that can be used in fuel cells and internal combustion en-
gines. And it offers the large potential benefit of reduc-
ing emissions of pollutants and greenhouse gases (1–4),
which are known to be derived from fossil fuel burning
(5). The state-of-the-art of hydrogen production is the
high-temperature steam reforming or the partial oxidation
of fossil resources such as methane, light hydrocarbons,
naphtha, and heavy oils (6–10). To meet the growing de-
mand of hydrogen and to prevent any energy-related pro-
blems, alternative hydrogen sources, which should be re-
newable and sustainable, efficient and cost-effective, and
convenient and safe, must be developed and the produc-
tion method also must be energy efficient (11, 12).
METHODS
A continuous-feeding fluidized-bed reactor system was
used in this work. The experimental system is almost the
same as a batch-feeding system previously described (22).
However, the reactor dimensions and feeding system were
modified for a continuous-feeding gasification system. Here
the gasification reactor is a quartz tube 66 cm high, with
a 1.8-cm i.d. The reactor consisted of a fluidized-bed sec-
tion at the middle of the reactor. The cellulose feeder con-
sisted of a glass vessel with a small pore, about 0.5 mm
in diameter, at the bottom allowing continuous feeding
from vibration of the vessel with a vibrator. The vibration
rate controlled the feeding rate. Cellulose particles (Merck;
A number of different ways of producing hydrogen
from alternative sources have recently been investigated
(13–20). However, hydrogen production from biomass
holds the greatest promise (20, 21), since biomass is abund-
1 To whom correspondence should be addressed. Fax: +81-298-53-5030.
E-mail: tomi@tulip.sannet.ne.jp.
255
0021-9517/02 $35.00
c
ꢀ 2002 Elsevier Science (USA)
All rights reserved.

256
ASADULLAH ET AL.
particle size, 100–160 µm) were transported to the catalyst
bed by the flow of N2 gas through an inner tube of 5 mm i.d.
Air and steam were introduced from the bottom of the re-
actor. Steam was supplied by using a microfeeder. The bed
temperatures at different points were measured by ther-
mocouples. The sample of the product gas was collected
from the sampling port by microsyringe and analyzed by
gas chromatography (GC). The concentration of CO, CO2,
and CH4 was determined by FID-GC and the concentra-
tion of hydrogen was determined by TCD-GC. The amount
of char was determined by the amount of gas (mainly CO2)
formed after stopping the feed of cellulose under the air
flowing at the reaction temperature.
CeO2/SiO2 was prepared by the incipient wetness
methodusingtheaqueoussolutionofCe(NH4)2(NO3)6 and
SiO2 (Aerosil, 380 m2/g). After drying at 393 K for 12 h, the
catalyst was calcined at 773 K for 3 h under an air atmo-
sphere. The Rh was loaded on CeO2/SiO2 by impregnation
of the support with acetone solution of Rh(C5H7O2)3. Af-
ter evaporating the acetone solvent, the catalyst was dried
at 393 K for 12 h. The final catalyst was pressed, crushed,
and sieved to 150- to 250-µm particle size. The loading
amount of Rh was 1.2 × 10−4 mol (g of catalyst)−1. The
loading amount of CeO2 is denoted in parenthesis using
the weight percent. In each run, 3 g of catalyst was used
and pretreated by a hydrogen flow at 773 K for 0.5 h. The
fresh (after H2 treatment) and used catalysts were char-
acterized by Brunauer–Emmett–Teller (BET) analysis and
transmission electron microscopy (TEM). The TEM ob-
servation of the fresh and used catalysts was carried out
by means of a JEM-2010F microscope (JEOL) operated at
200 kV. Samples were dispersed in tetrachloromethane by
a supersonic wave and put on Cu grids for the TEM ob-
servation under air atmosphere. The methane combustion
reaction was carried out on various catalysts within the tem-
perature range of 523–1123 K using CH4/air = 2/98 under
atmospheric pressure. The composition of the commercial
steam reforming catalyst (TOYO CCI, G-91) was 14 wt%
Ni, 65–70 wt% Al2O3, 10–14 wt% CaO, and 1.4–1.8 wt%
K2O. The composition of the dolomite was 21.0 wt% MgO,
30.0 wt% CaO, 0.7 wt% SiO2, 0.1 wt% Fe2O3, and 0.5 wt%
Al2O3. Before reaction the dolomite was calcined at 773 K
for 3 h followed by hydrogen treatment at 773 K for 0.5 h.
FIG. 1. Dependence of time on stream on C conversion and prod-
uct distribution of cellulose gasification on (a) Rh/CeO2/SiO2 (35) and
(b) G-91 at 773 K. Cellulose feeding rate, 85 mg min−1 (C, 3148 µmol
min−1; H, 5245 µmol min−1; and O, 2622 µmol min−1); air flow, 51 cm3
min−1 (O2, 417 µmol min−1); and N2 flow, 51 cm3 min−1
.
ble (Fig. 1a). In contrast, the C-conv decreased with time
on stream on G-91 (Fig. 1b). The rest of the carbon cor-
responded to the tar and char, which were deposited on
the catalyst surface. When the cellulose feeding stopped
after 25 min, the deposited carbon slowly converted to
mainly CO2. The figures clearly show that the deposited
carbon on the G-91 catalyst is much higher than that of
Rh/CeO2/SiO2 (35) catalyst. In this paper, the total amount
of the CO2 is assigned to the amount of char, as listed in
Table 1. The char usually accumulates on the catalyst sur-
face with lower activity and slowly takes part in the com-
bustion reaction. Consequently, in the continuous-feeding
system the amount of char gradually increases on the lower
active catalyst surface and deactivates the catalyst, as ob-
served on G-91. In order to obtain a highly efficient cata-
lyst, various kinds of Rh/CeO2/M-type (M = SiO2, Al2O3,
and ZrO2) catalysts with various loadings of CeO2 were
prepared and tested in the gasification of cellulose in a
RESULTS AND DISCUSSION
Catalyst Development
The results of the activity test of cellulose gasification continuous-feeding fluidized-bed reactor using air as a gasi-
over Rh/CeO2/SiO2 (35) and commercial steam reforming fying agent at 823 K. Among the catalysts Rh/CeO2/SiO2
catalyst (G-91) at 773 K are shown in Fig. 1. On Rh/CeO2/ (35) exhibited the best performance with respect to the for-
SiO2 (35), thecarbonconversion{C-conv =(formationrate mation of syngas and/or hydrogen. As in the batch-feeding
of CO + CO2 + CH4)/(C feeding rate in cellulose) × 100} reaction (22), the Rh/CeO2 catalyst exhibited considerably
to gas and formation of H2, CO, CH4, and CO2 was sta- high C-conv; however, the BET surface area drastically

NOVEL CATALYTIC SYSTEM FOR BIOMASS GASIFICATION
257
TABLE 1
Performance of Various Catalysts in the Gasification of Cellulosea
Formation rate (µmol min−1
)
C-convb
(%)
Charc
(%)
Tard
(%)
Catalyst
T (K)
CO
H2
CH4
CO2
H2/CO
Rh/CeO2/SiO2 (35)
773
823
873
923
973
845
1250
1617
1910
2279
1077
1286
1666
1995
2357
676
653
470
335
211
1178
1050
966
865
615
1.3
1.1
1.1
1.1
1.1
86
94
97
99
99
6
4
3
1
1
8
2
0
0
0
G-91
823
873
973
798
1289
2053
1538
1858
2242
418
393
158
1261
1114
762
1.9
1.5
1.1
79
87
94
18
10
3
3
3
3
Dolomite
823
973
1073
1173
414
1149
1383
1656
112
892
1072
1442
72
294
410
515
747
336
833
750
0.3
0.8
0.8
0.9
39
57
83
93
34
14
4
25
29
13
5
2
None
823
1023
1073
1173
240
1536
1714
1943
76
456
505
592
15
357
462
499
562
457
417
455
0.3
0.3
0.3
0.3
26
75
82
92
7
4
3
2
67
21
15
6
Nonee
823
228
62
11
39
0.3
9
15
76
a Conditions: cellulose, 85 mg min−1 (C, 3148 µmol min−1; H, 5245 µmol min−1; O, 2623 µmol min−1); air, 50 cm3 min−1
N2, 50 cm3 min−1; catalyst weight, 3 g; particle size of catalyst, 150–250 µm.
;
b C conversion to gas = {(formation rate of CO + CO2 + CH4)/C feeding rate} × 100.
c Percentage char = (CO + CO2 formation amount after stopping cellulose feeding/total C feeding) × 100.
d Percentage tar = 100 − (% C conversion + % char).
e Cellulose pyrolysis (N2 flow, 50 cm3 min−1 through distributor and 50 cm3 min−1 with cellulose).
decreased after the reaction because of CeO2 aggregation and H2 formation is the function of temperature and the
(23). In the batch-feeding gasification of cellulose, Pt, Ru, catalyst activity. The increase in temperature favors the
Pd, and Ni on CeO2 and Rh on SiO2, Al2O3, TiO2, MgO, C-conv and syngas formation either in the catalytic or in
and ZrO2 were also tested and the performance was lower the noncatalytic process. Thus, these are increased in all
than that of Rh/CeO2. In the continuous-feeding system, the systems with high temperature. On the other hand, the
the Rh/CeO2 catalyst suddenly deactivated due to a de- performance of the active catalyst also becomes higher with
crease in surface area from 60 to 13 m2 g−1. The loading increasing temperature. Consequently the C-conv and syn-
of CeO2 on the high-surface-area SiO2 inhibited the ag- gas formation were improved drastically on Rh/CeO2/SiO2
gregation of CeO2 and maintained the catalytic activity (35) catalyst. About 94% of the carbon in the cellulose was
of Rh/CeO2/SiO2 (35). Among the various loadings, the converted to gas, with a considerably high yield of CO and
35 mass% of CeO2 on SiO2 totally is the most suitable H2 at temperatures as low as 823 K on Rh/CeO2/SiO2 (35)
support for Rh catalyst in terms of the C-conv, gas yield, catalyst; however, this value was attained on G-91 catalyst
and fast char conversion (Fig. 1a). In addition, no decrease at973K. AlmostthecompleteC-convwasachievedat923K
in the BET surface area of this catalyst was observed. As on Rh/CeO2/SiO2 (35). Remarkably, methane was formed
shown in Table 1, at a particular temperature, such as 823 K, on the highly active Rh/CeO2/SiO2 (35) catalyst from the
Rh/CeO2/SiO2 (35) shows much higher C-conv and syngas CO hydrogenation. On the other hand, the lower C-conv
formation. Furthermore, the tar and char are much lower was achieved on the dolomite and in noncatalyst system
than that of other systems. These results clearly approach even at 1173 K. Especially, in the noncatalytic system, a very
thenoveltyofRh/CeO2/SiO2 (35)catalystforcellulosegasi- small amount of H2 was formed. The reaction conditions of
fication.
Table 1 were adjusted to the low-temperature syngas pro-
duction. In the hydrogen production system, the steam was
introduced in order to proceed to the steam reforming of tar
Effect of the Temperature
The results of the effect of temperature on the gasifica- and char and the water–gas shift reactions (H2O + CO →
tion of cellulose on Rh/CeO2/SiO2 (35), G-91, and dolomite CO2 + H2). The presence of steam in the reaction system
catalysts and on the noncatalytic gasification and pyroly- facilitates the tar and char conversion to gas. Thus, in the
sis are listed in Table 1. The C-conv as well as the CO next experiments, we added various amounts of steam.

258
ASADULLAH ET AL.
achieved at 773 K, whereas, interestingly enough, the 100%
C-conv with the higher hydrogen formation was found
when the steam with H2O/C = 0.35 was introduced. Fur-
thermore, the formation of hydrogen and CO2 expectedly
increased with an increase in the H2O/C ratio. The limit of
the temperature was 773 K for the complete conversion of
cellulose to gas products. No successful report was found
for the cellulosic biomass gasification at such a low tem-
perature. This result indicates that the steam directly takes
part in the gasification of the tar and char on the highly ac-
tive catalyst even at low temperatures, and thus complete
C-conv was achieved at 773 K. The biomass-derived tar
can be converted to gas on the Ni-based catalysts in the
secondary-bed reactor at above 1073 K (24); however, in
the primary-bed reactor the Ni-based catalysts suddenly
deactivated as a result of carbon deposition (25, 26), and a
similar phenomenon was observed for G-91.
The life of the Rh/CeO2/SiO2 (35) catalyst was tested at
823 K with a constant flow of the air and steam for 7 h. It
was observed that in the experimental conditions used here,
the C-conv and the product gas composition remained con-
stant. Figure 3 shows the TEM images of Rh/CeO2/SiO2
(35) after H2 reduction (Fig. 3a) and the reaction for 7 h
(Fig. 3b). The BET surface area measurement and the TEM
observation (Fig. 3b) of the used catalyst were carried out
and almost no structural change was observed. This sug-
gests that the catalyst activity can remain constant over a
prolonged reaction time.
FIG. 2. The influence of the steam-to-gas-formation rate and C con-
version in the gasification of cellulose at 773 K over Rh/CeO2/SiO2 (35).
Cellulose feeding rate, 85 mg min−1 (C, 3148 µmol min−1; H, 5245 µmol
min−1; and O, 2622 µmol min−1); air flow, 100 cm3 min−1 (O2, 818 µmol
min−1); N2 flow, 50 cm3 min−1 (2046 µmol min−1); and H2O, 555–
11,110 µmol min−1
.
Effect of the Steam
The C-conv as well as the selectivity of H2 was dra-
matically improved by the steam addition (Fig. 2) in the
gasification of cellulose on Rh/CeO2/SiO2 (35) catalyst. In
the absence of steam, 86% C-conv with less hydrogen was
FIG. 3. TEM images of Rh/CeO2/SiO2 (35) after the reduction and the reaction. (a) Hydrogen reduction at 773 K for 0.5 h. (b) Cellulose gasification
at 823 K for 7 h. Reaction conditions: cellulose feeding, 85 mg/min (C, 3148 µmol/min; H, 5246 µmol/min; O, 2624 µmol/min); air, 100 cm3/min (O2,
818 µmol/min); N2, 50 cm3/min; and H2O, 4444 µmol/min.

NOVEL CATALYTIC SYSTEM FOR BIOMASS GASIFICATION
259
Effect of the Catalyst Fluidization
think our catalytic process will have resistance to it. In the
fluidized-bed reactor, the catalyst circulates in the oxidizing
and reducing atmosphere. Under this condition, the surface
of the active site can be kept clear by this in situ treatment.
Therefore our catalytic process holds promise for the gasi-
fication of biomass with low levels of impurities, such as
wood.
It is thought that catalyst fluidization plays a major role in
regenerating an active catalyst surface. In the feeding line,
where the oxygen and external steam are absent, the cellu-
lose is first pyrolyzed to tar, char, H2O, and a small fraction
of gas. Then all the products diffused into the catalyst bed,
where the tar and char converted to gas during fluidization
of catalyst. We think that the catalyst is oxidized and it can
contribute to the combustion of the tar and char in the lower
part of the bed since oxygen is available. In the upper part
ACKNOWLEDGMENTS
This research was supported by the Future Program of Japan Society
of the bed it can contribute to the reforming of the tar and for the Promotion of Sciences under the Project “Synthesis of Ecological
High Quality of Transportation Fuels” (JSPS-RFTF98P01001).
char since oxygen is absent. Moreover, the steam directly
takes part in the conversion of tar and char to gas as well as
in the water–gas shift reaction on the catalyst surface. And
REFERENCES
thus 100% C-conv with a high yield of hydrogen even at the
low temperature (773 K) becomes possible. Since the cata-
1. Ogden, J. M., Annu. Rev. Energy Environ. 24, 227 (1999).
2. Steinberg, M., Int. J. Hydrogen Energy 24, 771 (1999).
lyst bed is in the fluidized condition, the catalyst particles
can interact with the oxygen frequently at the lower part of
the reactor. This can contribute to the removal of char with
low reactivity by combustion (Fig. 1). Methane combustion
activity has been tested and it was found that the Rh/CeO2/
SiO2 (35) catalyst had much higher activity than that of Rh/
CeO2, Rh/SiO2, and G-91 (Fig. 4). Furthermore, the tem-
perature in the fluidized bed is usually homogeneous. In
our fluidized reactor, the temperature difference between
the lower and upper parts and also between the outside
and inside of the reactor was found to be less significant
3. Pehr, K., Sauermann, P., Traeger, O., and Bracha, M., Int. J. Hydrogen
Energy 26, 777 (2001).
4. Du¨lger, Z., and Ozc¸elik, K. R., Int. J. Hydrogen Energy 25, 895 (2000).
5. Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., vanderLinden, P. J.,
and Xiaosu, D., Eds., “Intergovernmental Panel on Climate Change,
Climate Change 2001: The Scientific Basis.” Cambridge University
Press, Cambridge, UK, 2001.
6. Pena, M. A., Gomez, J. P., and Fierro, J. L. G., Appl. Catal. A 144, 7
(1996).
7. Armor, J. N., Appl. Catal. A 176, 159 (1999).
8. Saito, M., Shokubai (Catalysis) 35, 485 (1993).
9. Kikuchi, E., Uemiya, S., and Matsudo, T., Stud. Surf. Sci. Catal. 61,
509 (1991).
¨
(only 15 K). And a fluidized-bed reactor is more effective 10. Steinberg, M., and Cheng, C. H., Int. J. Hydrogen Energy 14, 797
(1989).
at removing the less reactive carbonaceous species (27, 28).
Finally, the combination of the high performance of the
Rh/CeO2/SiO2 (35) with the fluidized-bed reactor will pro-
11. Chum, H. L., and Overend, R. P., Fuel Process. Technol. 71, 187 (2001).
12. Scarpellini, S., and Romeo, L. M., Energy Convers. Manage. 40, 1661
(1999).
vide the novel system for hydrogen and syngas production
from biomass at low temperature with a high energy effi-
ciency. In this communication, pure cellulose was used as
the model compound of biomass. When we apply the cata-
lyst to real biomass, it is expected that impurities such as
sulfur and halogen will poison the catalyst. However, we
13. Momirlan, M., and Vezirog˘lu, T., Renew. Sustain. Energy Rev. 3, 219
(1999).
14. Benemann, J. R., Nature Biotechnol. 14, 1101 (1996).
15. Mizuno, O., Dinsdale, R., Hawkes, F. R., Hawkes, D. L., and Noike,
T., Bioresour. Technol. 73, 59 (2000).
16. Minowa, T., and Inoue, S., Renew. Energy 16, 1114 (1999).
17. Arauzo, J., Radlein, D., Piskorz, J., and Scott, D. S., Ind. Eng. Chem.
Res. 36, 67 (1997).
18. Olivares, A., Aznar, M. P., Caballero, M. A., Gil, J., France´s, E., and
Corella, J., Ind. Eng. Chem. Res. 36, 5220 (1997).
19. Caballero, M. A., Corella, J., Aznar, M. P., and Gil, J., Ind. Eng. Chem.
Res. 39, 1143 (2000).
20. Reed, T. B., Ed., “Biomass Gasification Principles and Technology,”
Energy Technol. Rev. Noyes Data Corporation, Park Ridge, NJ, 1981.
21. Garc´ıa, L., Frence, R., Czernik, S., and Chornet, E., Appl. Catal. A
201, 225 (2000).
22. Asadullah, M., Tomishige, K., and Fujimoto, K., Catal. Commun. 2,
63 (2001).
23. Asadullah, M., Fujimoto, K., and Tomishige, K., Ind. Eng. Chem. Res.
40, 5894 (2001).
24. Corella, J., Or´ıo, A., and Toledo, J.-M., Energy Fuel 13, 702 (1999).
25. Baker, E. G., Mudge, L. K., and Brown, M. D., Ind. Eng. Chem. Res.
26, 1335 (1987).
26. Garc´ıa, L., Salvador, M. L., Arauzo, J., and Bilbao, R., Energy Fuel
13, 851 (1999).
27. Tomishige, K., Matsuo, Y., Sekine, Y., and Fujimoto, K., Catal.
Commun. 2, 11 (2001).
28. Tomishige, K., Matsuo, Y., Yoshinaga, Y., Sekine, Y., Asadullah, M.,
and Fujimoto, K., Appl. Catal. A 223, 225 (2002).
FIG. 4. Catalytic activity of methane combustion. Conditions: reac-
tion temperature, 523–1123 K; CH4/Air = 2/98; total pressure, 0.1 MPa;
W/F = 0.4 g h−1 mol−1; hydrogen pretreatment at 773 K for 0.5 h.