Catalytic activities of ion-exchanged nickel and iron in low temperature hydrogasification of raw and modified birch chars

Catalytic activities of ion-exchanged nickel and iron in low temperature hydrogasification of raw and modified birch chars

Catalytic activities of ion-exchanged nickel and iron in low temperature hydrogasification of raw and modified birch chars Tsutomu Suzuki, Hiroshi ...

678KB Sizes 0 Downloads 0 Views

Catalytic activities of ion-exchanged nickel and iron in low temperature hydrogasification of raw and modified birch chars Tsutomu

Suzuki,

Hiroshi

Minami,

Tetsuo Yamada

Department of Environmental Engineering, Kitami, Hokkaido, 090 Japan (Received 22 November 1993)

Kitami

Institute

and Tsuneyuki of Technology,

Homma

165 Koen-cho,

Raw, HNO, oxidized and carboxymethylated birch woods loaded with nickel or iron by the ion-exchange method were carbonized at 500°C in a flow of nitrogen, and the resulting chars were hydrogasified in a thermobalance to examine their reactivities below 700°C. The amounts of ion-exchanged metals on raw char were too small to give high gasification reactivity. However, oxidized and carboxymethylated woods with increased ion-exchange capacity produced much more reactive chars. Both nickel and iron exhibited larger catalytic activities on carboxymethylated chars than on oxidized chars, because better metal dispersion could be achieved on carboxymethylated wood with its larger cation exchangeability. It was noteworthy that only 1 wt% loading of iron, as well as nickel, on carboxymethylated char was sufficient to attain a gasification of 90 wt% at 700°C. It was also noted that the catalytic effect, up to 6Oo”C, of iron on the gasification of oxidized and carboxymethylated chars was larger than that of nickel. This is ascribed to two factors; greater catalytic activity of metallic iron formed during the gasification than that of nickel metal, and low ash level in the chars. Above 6OO”C, however, serious loss of activity of the iron was observed in the absence of wood ash. This showed the different influence of wood ash on the catalysis of iron in the low and high temperature regions (Keywords: wood char; catalysts; hydrogasification)

The hydrogasification of wood char has been studied to develop a catalytic conversion process of lignocellulosic biomass to fluid energy ’ *; the main aim being to produce highly reactive wood char with the aid of nickel or iron as catalysts. For high gasification reactivity of wood char, the choice of metal precursor, its addition method, the metal loading and the condition of carbonization are all important. Thus far it has been found that nickel and iron nitrates are good catalyst precursors, and their loadings on raw wood by wet impregnation prior to 500°C carbonization lead to wood chars of high gasification reactivity toward hydrogen. The catalytic effect of nickel on birch3g4 and larch6,’ chars was sufficient to attain a conversion of 90 wt% at 600°C when the loading was about 3 wt%. The reactivities of iron loaded birch’ and larch’ chars at about 5 wt% loading were also satisfactorily high at 650°C. However, from the viewpoint of practical synthetic natural gas (SNG) production from wood char, it is more attractive to use a smaller amount of the cheaper iron catalyst. Ion-exchanged catalysts are known to act more effectively than impregnated catalysts because of better dispersion in gasification of coal by steam’.” and hydrogen”. There is a similar report that in biomass gasification, cottonwood char with ion-exchanged cobalt exhibited very high reactivity toward carbon dioxide at 600”C’2. However, usual woods contain few carboxyl groups as ion-exchangeable sites and the carboxyl 0016-2361,!94;12,:1836-06 (‘ 1994 Butterworth-Heinemann

1836

Fuel 1994

Volume

Ltd

73 Number

12

content of the cottonwood was only 8-9 meq lOOg-’ (Ref. 12). It is uncertain whether this level is adequate for low temperature hydrogasification. If the original carboxyl content of wood is too small, certain modifications may be helpful to increase the cation-exchange capacity. The present paper describes the effectiveness of HNO, oxidation and carboxymethylation as modification methods. Catalytic activities of ion-exchanged nickel and iron for the hydrogasification of raw and modified birch chars were examined and compared with those based on impregnation. Different catalytic effects observed between nickel and iron are also discussed in terms of reducibility into metals and the influence of ash level in the char. EXPERIMENTAL Birch (Betula ermanii Cham) powder (OS&O.25 mm in diameter) was used as the wood material. Demineralization of raw wood was done by washing with 0.1 N HCl. HNO, oxidation and carboxymethylation were carried out to increase the ion-exchange capacity. Oxidized wood was obtained by treatment with 6 N HNO, at 40°C for 1 h. Carboxymethylated wood was prepared as described by Nakano et al. (NaOH and ClCH,COOH dissolved in 80% of C,H,OH, 6O”C, 2 h)13. Carboxyl contents in raw, demineralized, oxidized and carboxymethylated woods are given in Table 1, together with their recoveries and

Hydrogasification

of char: T. Suzuki et al. COOH

HN03

oxidation -,KYr”-

CH2OH

CHO

CHO

O-

0 OH -0

JZ?

CHZOCHZCOOH

OH

0

O-

Carboxymethylation OCHZOOH -0

*

0 OCH2COOH

Figure 1

Increase

of carboxyl

group

Table 1 Content of carboxyl mineralization, HNO, oxidation

by HNO,

oxidation

group, recovery and and carboxymethylation

Treatment

Carboxyl group” (meq 100 8-l)

Raw (none, UT) Demineralization (DM) HNO, oxidation (OX) Carboxymethylation (CM)

I 8 40 163

Recoveryb (%) _ 86 67 87

and carboxymethylation

ash after deof birch

Ash’ W) 0.24 0.02 0.01 0.18(0.02”)

“ Determined by treating with 0.2 N calcium acetate to liberate H+ ion and then titrating with 0.05 N NaOH up to pH 8.3 (Ref. 14) *Expressed relative to raw wood as 100% ‘ Incombustible residue at 6OWC “Obtained after HCI washing. This value was used in calculating the char yield on a daf basis

ash levels. Figure I depicts models of the increase of carboxyl groups in cellulose by HNO, oxidation and carboxymethylation. If the reaction is assumed to occur on only cellulose molecules, the number of carboxyl groups introduced by HNO, oxidation and carboxymethylation would be about 0.28 and 0.68, respectively, per glucose unit of C,H,,O,. Ni(NO,), .6H,O and Fe(NO,), .9H,O as catalyst precursors were loaded on raw and demineralized woods by wet impregnation. These salts were also used for metal loading on demineralized, oxidized and carboxymethylated woods by ion-exchange. Wood samples obtained were designated as UT(Imp), DM(Imp), DM(Ion), OX(Ion) and CM(Ion), respectively. In wet impregnation, woods were soaked in an aqueous solution of nickel nitrate or iron nitrate, and they were heated under a pressure of 4 x lo3 Pa at 40°C in a rotary evaporator. Ion-exchange of nickel or iron was made by successive washings of woods packed in a glass column with 0.1 N HCl, deionized water, 0.01 M solution of the nickel or iron salt, and deionized water*. The metal loading was less than 2 wt% as metal in char. *Ion-exchange on OX(ion) and CM(ion) was ascertained by the variation of Fourier transform infrared (FT-i.r.) spectra; the absorption of COOH at about 1765 cm-’ decreased with the increase of COOabsorption near 1600 cm-‘. Such a spectral change could not be recognized on UT(imp) and DM(ion) samples

of cellulose

After vacuum drying at 60°C to remove residual water, all the woods were carbonized in a quartz tube reactor at 500°C for 1 h under a flow of nitrogen. The resulting chars were then characterized in terms of the composition of C, H, N and 0, average crystallite sizes of nickel particles, &, and iron particles, LFer aromaticity of char carbon, and specific surface area, in addition to the yield on a daf basis. Carbon, hydrogen and nitrogen in char were determined by ultimate analysis, and the oxygen was obtained by difference. LNi and L,, for all the Ni and Fe chars, in which metallic nickel and Fe,O, were the main chemical forms, were calculated using X-ray diffraction (XRD) lines with Cu Ka radiation at 20= about 45 and 35”, respectively. The broad carbon crystallite peak was also measured after treating Ni chars with warm 4N HCl for the purpose of obtaining more accurate values of LNi3. Aromaticity of char carbon was defined as the ratio of FT-i.r. spectroscopic absorbance of the aromatic C-H band (Ksod5) to that of the aliphatic C-H band (K 292,,). Specific surface area was measured at - 78°C by the BET method using CO, as an adsorbant. The gasification reactivity was evaluated by temperatureprogrammed reaction using thermogravimetry at ambient pressure under the following conditions: sample weight, 10 mg; hydrogen flow, 100 ml (STP) min-‘; heating rate, 1O’C mini ‘; and maximum temperature, 1000°C. Prior to the hydrogasification test, all the specimens were vacuum dried overnight at 50°C. Weight decreases up to 600 and 700°C were determined and used as the reactivity parameters. Weight loss due to the reduction of Fe,O, to iron metal for Fe chars during the gasification was negligible because it was less than 1 wt% of the char weight. RESULTS

AND

DISCUSSION

Catalytic activities of ion-exchanged nickel and iron on raw wood char Both gasification reactivities and properties of UT(Imp), DM(Imp) and DM(Ion) chars are summarized in Table 2. Metal loading on these UT(Imp) and DM(Imp) chars was adjusted to 0.7-0.8 wt% to match with the amount on DM(Ion) char. Improvement in the initial dispersion of nickel or iron particles on wood char will lead to an

Fuel 1994

Volume

73 Number

12

1837

Hydrogasification Table 2

of char:

Gasification

reactivities

T. Suzuki

et al.

and properties

of UT(imp),

DM(imp)

and DM(ion)

chars loaded

with nickel or iron

Conversions Metal loaded (WWO)

Charsa

600°C

700°C (%I

(%I

UT-None

_

5

8

UT-Ni(lmp)

0.7

16

54

Tb (‘Cl

Char yield (daf, %) 21

540

22

Oxygen conterWd

Ash (wt%)

Surface area” (m2 g-‘1

(%)

Aromaticity” (H

1.2

11

0.9

300

0.9

12

1.2

330

DM-Ni(Imp)

0.8

10

43

550

22

0.1

11

1.1

320

DM-Ni(lon)

0.8

13

47

550

22

0.1

12

1.2

340

UT-Fe(Imp)

0.7

15

36

540

25

0.7

15

1.0

3to

DM-Fe(lmp)

0.7

20

48

550

25

0.1

16

1.0

310

DM-Fe(Ion)

0.7

17

44

540

24

0.1

15

1.1

320

“See the text ‘Temperature where rapid weight decrease begins ‘Except catalyst dNot including the oxygen from Fe,O, in Fe chars

Table 3

Properties

of UT(imp),

Chars”

Metal loaded (wt%)

DM(imp),

OX(ion) and CM(ion)

Char yield (daf, %)

Ashh (wt%)

UT--Ni(lmp)

0.9

23

DM-Ni(lmp)

1.1

22


1.1

OX-Ni(Ion)

1.0

22

CM-Ni(Ion)

1.1

29

UT-Fe(Imp)

1.0

26

DM-Fei(

1.0

25


OX-Fe(Ion)

1.0

24

10.1

CM-Fe(Ion)

1.1

35

Imp)

chars loaded

with about

1 wt% nickel or iron Surface area” tm2 g- ‘1

Oxygen content’ (%I

LNi or Lpeud (nm)

12

4.5

1.3

370

12

5.0

1.3

360

Aromaticity” (-1

13

4.0

1.2

370

0.1

13

4.0

1.4

420

0.9

16

7.0

1.1

330

17

7.5

1.1

320

17

6.0

1.1

310

18

5.5

1.3

340

0.1

“See the text ‘Excluding catalyst ‘Not including the oxygen from Fe,O, in Fe chars ‘Average crystallite size of nickel, LNi, and iron, L,,, particles

increase in the hydrogasification reactivity, particularly in the low temperature region3-8. Although LNi or L,, could not be determined for all the chars because no nickel and iron XRD lines appeared, it is probable that DM(Ion) chars would carry better dispersed metal particles than UT(Imp) and DM(Imp) chars. However, neither DM-Ni(Ion) char nor DM-Fe(Ion) char showed a satisfactory conversion in the low temperature region. This may be due to insufficient ion-exchange on DM(Ion) chars, because of too small a carboxyl content of raw wood. Therefore, it was necessary to increase the cation-exchange capacity of raw wood by suitable modifications. It was also found that DM(Ion) chars did not always give higher conversions than those of UT(Imp) and DM(imp) chars. This indicates that the initial metal dispersion would not be the only factor in governing the gasification reactivity. Since there were no great differences in oxygen content, aromaticity and surface area among the three chars, attention should be given to the difference in ash content. Although ash content would have no great influence on Tb, temperature where the rapid weight decrease begins, for both Ni and Fe chars, demineralization permitted Fe chars to increase the conversion and showed an opposite effect on Ni chars, in accord with a previous report7,8. Since silica, a major wood ash component, markedly lessens the catalytic

1838

Fuel 1994

Volume

73 Number

12

activity of iron, as well as nicke17, ash removal was favourable for iron catalyst. The lowering in the activity of nickel could be caused by the concurrent loss of calcium’.“, the most abundant metal in wood ash. According to Haga and Nishiyamal’, calcium prevents nickel particles from agglomerating during the hydrogasification of coal. Thus, when a comparison of the catalytic activity between nickel and iron is made, the different influence of ash cannot be ignored. Catalytic HN03

activities

ofion-exchanged

oxidized and carboxymethylated

nickel and iron on wood chars

Tcrble 3 compares yields and properties of UT(Imp), DM(Imp), OX(Ion) and CM(Ion) chars loaded with about 1 wt’% metal. Thermogravimetric analysis (TGA) profiles for the four kinds of metal-loaded chars are illustrated in Figure 2, to which those for chars with no treatment are added for reference. The influence of the wood modifications on the uncatalysed gasification was not large. The reactivity of Ni chars was CM(Ion) > OX(Ion) > UT(Imp)> DM(Imp); and that of Fe chars was CM(Ion) > OX(Ion) > DM(Imp) > UT(Imp). Since each order of the gasification reactivity was almost consistent with the order of LNi or LFe, the differences in oxygen content, aromaticity and surface area among the four chars were not so important for the gasification reactivity. However, the higher reactivity of DM-Fe(Imp) char than that

Hydrogasification

(A) Ni-chars

--t ++ 80 - -o+

400

600

800

1C 3

(“C)

Figure 2 TGA profiles for the hydrogasification of (A) Ni chars DM, demineralized; OX. HNO, oxidized; and CM, carboxymethylated

0 A V 0

5.0

1.0

UT-Ni DM-Ni OX-Ni CM-Ni

(Imp) (Imp) (Ion) (Ion)

20

Nickel or iron loaded (%) Figure 3

et al.

--c OX-None * CM-None

Temperature

2-

T. Suzuki

(B) Fe-chars

UT-None DM-None UT(Imp),0.9% DM(Imp), 1.1%

200

of char:

The change of .LNior L,, with the loading

of nickel or iron

of UT-Fe(Imp) char could be partly attributable to increased catalytic activity of iron by demineralization. It is noticeable that Tb was dependent on the type of Fe char: about 430, 480, 530 and 535°C for CM(Ion), OX(Ion), DM(Imp) and UT(imp), respectively. Tb for the Fe chars tended to decrease with increasing amount of iron, and was in the order of CM(Ion)
I

I

0

200

400

I

I

600

800

Temperature and (B) Fe chars with a metal wood samples, respectively

loading

of

1000

(‘C)

1wt%: UT, untreated

(raw);

the gasification temperature will induce the rapid weight decrease of carbon’ 7,1*, the reactivity of Fe chars could be closely associated with the reducibility of Fe,O, to metallic iron during hydrogasification. In other words, the variation of Tb shows that active iron metal is formed under the great influence of the initial dispersion of iron oxide particles on char to trigger the iron-catalysed gasification. On the other hand, all the Ni chars, including DM(Ion) char (Table 2) gave approximately equal Tb of 550°C irrespective of the type of char and loading. This is in accordance with the fact that almost complete reduction to nickel metal takes place during the carbonization for all Ni chars. As for wood carbonization, it is interesting that CM(Ion) wood gave a much higher char yield than the other three woods for either metal loading. This will be due to its higher thermostability. For carboxymethylated wood, the recovery (Table I) would have amounted to 109 wt% on the basis of the increase of carboxyl content, if no weight loss had occurred during the treatment. Most of the weight loss, 22 wt% (109-87 wt%) resulted from the solubilization of thermolabile hemicelluloses. Thus, of the four woods, carboxymethylated wood was most enriched with thermostable constituents like lignin and cc-cellulose. On the contrary, the low recovery of oxidized wood arose from considerable loss of both lignin and o! cellulose (their contents decreased from 26.2 and 41.1 wt% in raw birch to 16.2 and 21.3 wt%), leading to the lowest char yield. Lower yields of Ni chars with lower oxygen content and larger aromaticity and surface area compared to Fe chars agree well with earlier findings which have been discussed elsewheresq7. Changes of LNi or L,, and weight decreases at 600 and

Fuel 1994

Volume

73 Number

12

1839

Hydrogasification

of char: T. Suzuki et al.

(B) at 700°C

(A) at 600°C 100

UL

0.0

I

I

1.0

2.0

Nickel or iron loaded (%) Figure 4

Weight

decreases

during

hydrogasification

Fuel 1994

Volume

73 Number

1

1.0

0.0

UT-Ni(Imp)

A 0 0 v A W

OX-Ni(Ion) CM-Ni(Ion) UT-Fe(Imp) DM-Fe(Imp) OX-Fe(Ion) CM-Fe(Ion) I

2.0

Nickel or iron loaded (%)

at (A) 600 and (B) 700-C for Ni and Fe chars

700°C with metal loading for the four chars are presented in Figures 3 and 4, respectively. It is clear that L,, increased in the order of CM(Ion) < OX(Ion) < UT(Imp) < DM(Imp), and this order also held for LNi with less difference. As expected, ion-exchanged metals on OX and CM chars were better dispersed than impregnated metal on UT and DM chars. Better metal dispersion on CM(Ion) chars than on OX(Ion) chars could be achieved by a larger cation-exchangeability of carboxymethylated wood. Somewhat larger values of LNi and L,, on DM(Imp) chars than on UT(Imp) chars revealed that demineralization was rather disadvantageous to the dispersion of both metals. Increasing metal loading resulted in increases of oxygen content, aromaticity and surface area: their variations were from 11 wt%, 0.9 and 300 m2 g-l of raw char without catalyst (UT-None, Table 2) to 17 wt%, 1.5 and 440m2 g-’ of CM-Ni(Ion, 1.89 wt%) char and 23 wt%, 1.4 and 360 m2 g- ’ of CM-Fe(Ion, 2.04 wt%) char, respectively. Although these properties were in general CM(Ion) > UT(Imp) > DM(Imp)>OX(Ion) at equal loadings for both metals, this order was insignificant for their gasification reactivities shown in Figure 4. In the whole range of each metal loading, the reactivity order determined by weight decrease at 600 and 700°C was the same as in Figure 2. This confirms the close correlation of the gasification reactivity with initial catalyst dispersion. In order to achieve a conversion of 90 wt% at 7OO”C, the loading of nickel should be about 0.9 and 1.3 wt% for CM(Ion) and OX(Ion) chars, respectively. Iron loadings at about 1.0 and 1.5 wt% also were adequate for CM(Ion) and OX(Ion) chars. As was predictable from earlier results3-7, both reactivities of UT-Ni(Imp) and UT-Fe(Imp) chars

1840

I

0

12

were unsatisfactory at 2.0 wt% metal loading. DM(Imp) char, likewise, needed more than 2.0 wt% of iron loading, in spite of its increased catalytic activity. Evidently metal loading on HNO, oxidized and carboxymethylated woods by ion-exchange offered much saving in the amount of nickel or iron for the low temperature hydrogasification over impregnation of catalyst on raw and demineralized woods. It is also noteworthy that iron surpassed nickel in the catalytic effect up to 600°C for all OX(Ion) and CM(Ion) chars; and, particularly, the 600°C conversion of CM(Ion) char with 1 wt% iron approached 90 wt% gasification. When this is considered in relation with the above-mentioned variation of Tb for Fe chars, it seems reasonable to infer that metallic iron formed during the gasification exhibited a larger cataytic activity than nickel metal which already existed before the gasification. In addition, it should be taken into account that iron will raise its catalytic activity and nickel will diminish in the absence of wood ash. The present data gave no definite evidence for oxygen in char being significant in the gasification reactivity. However, there is a possibility that higher oxygen content in Fe chars than in Ni chars would be more favourable to accelerate the catalysis of iron. An important role of oxygen in carbon materials in enhancing the rate of catalysed hydrogasification has been pointed out by several workers6,8,‘9m21. The excellent reactivity of CM-Fe(Ion) char with 1 wt% iron up to 600°C may be a synergy of high catalytic activity of iron itself and a promotive effect by low ash and high oxygen contents in char. Nevertheless, the 700°C conversions of OX-Fe(Ion) and CM-Fe(Ion) chars were almost equal or inferior to those of the corresponding Ni chars. Above 7OO”C, these Fe

Hydrogasification

chars rapidly lost their reactivities with no complete conversion even at lOOO”C, as seen in Figure 2. A similar situation was observed for DM-Fe(Imp) char. In contrast, the gasification reaction of UT-Fe(Imp) char steadily proceeded in the high temperature region. Therefore, the pronounced activity loss of iron above 600°C is characteristic of demineralized char, and shows the different influence of wood ash on the catalysis of iron for the low and high temperature regions.

of char: T. Suzuki

et al.

ACKNOWLEDGEMENTS We would like to express our appreciation to Professor Akira Tomita at Tohoku University, Sendai, Japan, for valuable advice and discussion.

REFERENCES CONCLUSIONS

1

The present study described catalytic activities of ion-exchanged nickel and iron on raw, HNO, oxidized and carboxymethylated birch chars against their hydrogasification, particularly in the low temperature region. The carboxyl content of raw birch was too small to hold sufficient ion-exchanged metals for producing highly reactive char. However, HNO, oxidized and carboxymethylated wood chars showed satisfactory reactivities at much smaller loadings of nickel and iron than that required for raw and demineralized chars with impregnated catalysts. Carboxymethylation was more effective in enhancing the catalytic activity of each metal than HNO, oxidation, because it provided a larger ion-exchange capacity of the wood, leading to more highly dispersed metals on char. These results show that the metal loading required for successful hydrogasification will be smaller on a material with higher carboxyl content. It was also observed that iron exhibited a greater catalytic effect than nickel on the gasifications of HNO, oxidized and carboxymethylated chars up to 600°C. This could be related to the formation of iron metal during gasification and to low ash content. Iron-catalysed gasification is more attractive than nickelcatalysed gasification because of higher reactivity at temperatures below 600°C and higher char yield, in addition to cheaper catalyst cost. Above 600°C however, pronounced loss of activity at iron was observed for low ash chars.

2 3 4 5 6 7 8 9 10 11 12 13 14

15 16 17 18 19 20 21

Suzuki, T., Katabuchi, H., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1984, 30, 684 Suzuki, T., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1985.31, 595 Suzuki, T., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1986, 32, 730 Suzuki, T., Takahashi, A., Kuwahara, E., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1987, 33, 423 Suzuki, T., Kuwahara, E., Takahashi, A., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1988, 34, 537 Suzuki, T., Saitoh, M., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1990, 36, 754 Suzuki, T., Mugishima, M., Yamada, T. and Homma, T. Mokuzai Gakkaishi 1992, 38, 509 Suzuki, T., Katagiri, A., Kobashi, K., Funaki, M. and Yamada, T. Moktcai Gakkaishi in press Hengel, T. D. and Walker, P. L. Jr Fuel 1984, 63, 1214 Nabatame, T., Ohtsuka, Y., Takarada, T. and Tomita. A. J. Fuel Sot. Jpn 1986, 65, 53 Egashira, M., Honda, M. and Kawazumi, S. J. Chem. Sot. Jpn 1982, 323 Degroot, W. F. and Richards. G. N. Fuel 1988. 67, 345, 352 Nakano, T., Honma, S., Ehata, S. and Matsumoto, A. Mokuzai Gakkaishi 1990, 36, 193 ‘Experimental Book of Wood Science, Part II Chemistry’ (Eds M. Usuda, H. Mizumachi, K. Iiyama, N. Maroboshi and A. Yamaguchi), Chugai Sangyo, 1989, pp. 173-174 Haga, T. and Nishiyama, Y. Ind. Eng. Chem. Res. 1987,26,1202 Haga, T. and Nishiyama, Y. Ind. Eng. Chem. Res. 1989,28,724 Yamada, T., Tomita, A., Tamai, Y. and Homma, T. Fuel 1983, 62, 246 Yamada,T.,Suzuki,T. andHomma,T. Chem. Left. 1984,2013 Niiyama, H., Sato, H., Nakamura, R. and Echigoya, E. Chem. Lett. 1981, 605 Tarki, H. T., Kiennemann, A. and Chornet, E. Fuel 1984,63,30 Treptau, M. H. and Miller, D. J. Carbon 1991, 29, 531

Fuel 1994

Volume

73 Number

12

1841