Energy & Fuels 1989,3,641-646
641
Successive Sequential Extractive'Disintegration of Coal in Coal-Derived Solvents under Ambient Pressure D. K. Sharma* and S. Mishra Centre for Energy Studies, Indian Institute of Technology, Delhi, Hauz Khas, New Delhi-110016, India Received August 17, 1988. Revised Manuscript Received June 5, 1989
Assam coal was successively extracted in anthracene oil, quinoline, and liquid paraffin. About 68% of the coal was rendered extractable through successive extraction in anthracene oil followed by quinoline and then finally in liquid paraffin. Preextraction of the coal in anthracene oil is beneficial for subsequent extraction in quinoline. The extractability of anthracene oil extracted coal in quinoline increased with an increase in extraction time in anthracene oil. Thus anthracene oil extraction of coal proceeds through chemical depolymerization (disintegration) of the coal. Preextraction of coal in liquid paraffin showed an adverse effect on successive extraction in other solvents. The combined total extractability of coal in all the solvent systems increased with an increase in the extraction time in successive extractions. Assam, Talcher, and Raniganj coals and Neyveli lignite gave more than 50% extraction yields through the successive extractive disintegration using an anthracene oilquinoline-liquid paraffin solvent system. Possible explanations for successive extractive disintegration of coal in anthracene oil-quinoline-liquid paraffin solvents are given.
Introduction Lately, interest in solvent extraction of coal for upgrading low-grade coals has increased.'+ Most of the research work on solvent extraction of coal was performed a t elevated temperatures and high pressures.' Unfortunately, coal has poor extractability in organic solvents under atmospheric pressure conditions. Pretreatments such as reduction,7*8reductive alkylation?l0 phenolation (dep~lymerization),"-'~alkylation,'"" reductive acylation,18 a ~ y l a t i o n , 'alkaline ~ degradation,20 and graft copolymerization21enhance the extractability of coal in organic solvents. In addition, these reactions have their own problems.= Direct extraction of coal in an organic solvent (without pretreatments) would be more interesting for process development. Relatively shorter extraction times (2-6 h), were used so that a semicontinuous or continuous process may be feasible a t some stages. Attempts were made to chemically depolymerize (disintegrate) the coal in a solvent through extractive disintegration and to extract the disintegrated coal in another solvent. In the present work, solvent extraction studies on Assam (Baragolai) coal are reported. Three coal-derived solvents, quinoline (Qn), anthracene oil (AO), and liquid paraffin (LP), were selected for these studies. Extractions were performed for 2-24 h. The sequence used for successive extractions was altered. Conditions were optimized for solvent extraction of coal through successive extraction. Other coals such as Talcher, Raniganj, and Godavari coals and Neyveli lignite were also used for extractive disintegration studies. Experimental Section Assam, Talcher, Raniganj, and Godavari coals were gound to -60 to +120 BSS (British Sieve Standard) mesh size and Neyveli lignite waa ground to -120 BSS mesh size and dried in a vacuum oven at 105 OC for 24 h. After drying the coal was stored in a vacuum desiccator. Chemicals. Anthracene oil (AO) was obtained from carbonization of the coal and had a 270-300 "C boiling range. Liquid paraffin (LP) was a petroleum-derived solvent available commercially in the market and had a 330-360 "C boiling range. The
* To whom correspondenceshould be addressed. 0887-0624 18912503-0641$01.50/O
elemental analysis of LP and A0 was performed on a Carlo Erba elemental analyzer, Model 1106. Quinoline, a commercially available laboratory-grade reagent, was dried by refluxing over NaOH and then distilled. I3C NMR spectral analysis of A 0 and LP WBS performed in CDC1, solvent by using TMS as an internal standard on a 100-MHz JEOL JNM FX 100 FT-NMR spectrometer. The elemental analysis of coal samples was performed on a Heraeus CHN-RAPID elemental analyzer. Sequential Extractive Disintegration of Assam Coal. Assam coal (20 g) was placed in a three-necked flask fitted with a reflux condenser and containing solvent (340 mL). A coalsolvent ratio of 1:17 was used. The mixture was refluxed and used the first solvent for extraction. The residual coal obtained after filtration and washing was subjected to further extraction in the (1)Chem. Eng. News 1986, 64(April 28),22. (2)Minkowa, V.; Petrov, V.; Angelova, G. Khim. Ind. 1986,88, 67. (3)Gatais, J. G. Prepr. Pap.-Am. Chem. Soc., Diu.Fuel Chem. 1986, 31, 181. (4)Chepa, K.; Tagaya, H.; Kabayashi, T.; Shibuya, Y. Ind.Eng. Chem. Res. 1987,26, 1329. (5)Schlosberg, R.H.; Neaval, R. C.; Maa, P. S.; Gorbaty, M. L. Fuel 1980,59,45. (6) Liotta, R. Fuel 1981,60, 453. (7) Halleux, A.; Greef, H. Fuel 1963,42, 185. (8)Lazarov, L.; Angelova, G. Fuel 1968,47, 333. (9)Sternberg, H.W.; Delle Donne, C. L.; Pantages, P.; Moroni, E. C.; Markely, R. E. Fuel 1971,51,220. (10)Stock, L. M.; Willis, R. S. J . Org. Chem. 1985,50,3566. (11)Heredy, L. A.;Neuworth, M. B. Fuel 1962,41, 221. (12)Ouchi, K.; Imuta, K.; Yamashita, Y. Fuel 1973,52,156. (13)Larsen, J. W.;Kuemmerle, E. W. Fuel 1976,55,162. (14)Sharma, D. K.Fuel 1988,67, 186. (15)Kroger, C.; de Vries, H. Liebigs Ann. Chem. 1962,15, 652. (16)Kroger, C.; Rabe, H. S.; Rabe, B. Erdoel Kohle, Erdgas, Petrochem. 1963,16, 21. (17)Kroger, C. Forschungsber. Landes Nordrhein Westfalen 1965, No. 1488. (18)Sharma, D. K.;Mma, Z. B.; Sarkar,M. K. Fuel Process. Technol. 1982,6, 301. (19)Hodek, W.; Kolling, G. Fuel 1973,52, 220. (20)Mirza, Z.B.;Sarkar, M. K.; Shanna, D. K. Fuel Process. Technol. 1984,9, 149. (21)Polygulf Associates. Combustion 1977,48,34; Chem. Abs. 1977 87, 120341e. (22)Sharma, D. K. Solvent Extraction of Coal through Chemical Reactions at Atmospheric Pressure. In Advances in Coal Chemistry; Vasilakos, N. P., Ed.; Theophrastus Publications: Athens, Greece, in press.
1989 American Chemical Society
Sharma and Mishra
642 Energy & Fuels, Vol. 3, No. 5, 1989
Table I. Analyses of the Coals and the Lignite Used Proximate Analvsis I% on a Drv Basis) moisture ash volatile matter
coal fields Assam Talcher Raniganj Godavari Neyveli lignite
2.8 2.1 1.5 2.2 8.7
8.3 16.9 13.3 27.0 6.1
42.0 38.8 36.4 32.0 46.6
fixed carbon 46.9 42.2 48.8 38.8 38.6
Ultimate Analysis (% on a Daf Basis) coal fields Assam Talcher Raniganj Godavari Neyveli lignite
carbon
hydrogen
sulfur
nitrogen
oxygen
77.0 74.8 73.0 84.5 62.0
5.7 5.3 5.4 5.0 5.3
3.8 1.0 0.9 1.0 1.4
1.5 1.6 2.2 2.0 0.9
12.0 17.3 18.5 7.5 30.4
next solvent following the same procedure as before under reflux conditions. The residual coal obtained after filtration and washing was dried and weighed to record the loss in weight of coal. Filtration of A 0 and L P Extraction Products. The A 0 extraction product was filtered through Whatman Grade 1 filter paper with a double-walled funnel. The steam was passed through this funnel to keep the reaction product hot for easy filtration. For longer extractions in A0 (i.e., for more than 6 h), filtration was performed under vacuum (as otherwise filtrations were slow). When all or most of the A 0 had been filtered, the residue was washed with toluene until a colorless filtrate was obtained. This residual coal product was dried under vacuum at 150 "C for 12 h. In the case of LP extraction, the same procedure was adopted for the filtration of LP. When all or most of the LP had been filtered, the residual coal was washed with hexane to remove LP from the coal. The residual coal was dried in an oven at 105 "C for 12 h. Extraction in Quinoline. Assam coal was subjected to quinoline (Qn) extraction for 2-48 h in a Soxhlet extractor. The Qn-extraded coal reaidue was washed with aqueous HCl to remove Qn and then washed with water to remove HCl and dried under vacuum at 150 "C for 12 h. Calculation of Extractability. The extractabilityof coal was calculated on the basis of percent loss in weight of coal after extraction. The total extractability was calculated from the difference between the weight of the original coal and that of the residual coal remaining after extraction using all three solvents in succession. The overall loss in weight of coal represented the extractability, which was then calculated on a dmmf basis. The experiments were repeated two to three times and the extraction results were found to be within *5%. Swelling Measurement. Swelling measurements were carried out by the procedures reported by Liotta et a1.23and Larsen et al.u The coal (500 mg) was placed in a 15-mL graduated glass tube and centrifuged (dry) and its height measured. A solvent such as LP, Qn, or A 0 (5 mL) was added, and the tube was shaken well. After 24 h, the tube was centrifuged and the height of the coal column was measured again. The swelling ratio is the ratio of coal column heights: column height after swelling with solvent Qv = column height of dried coal without solvent Recovery of A 0 Extract. A 0 extract was recovered after &tilling off the solvent under reduced premure. The concentrated extract obtained was washed with methanol to remove traces of A 0 and the A 0 extract obtained was dried under vacuum at 150 "C for 12 h. (23) Liotta, R.; Brown, G.; Isaaca, J. Fuel 1983, 62, 781. (24) Larsen, J. W.; Lee, D. Fuel 1983, 62, 1351. (25) Oele, A. P.; Water", H. I.; Goedkoop, M. L.; van Krevelen, D. W. Fuel 1951,30, 169. (26) Dryden, I. G. C. Fuel 1950,29, 197. (27) Davis, G. 0.; Derbyshire, F. J.; Price, R. J. Inst. Fuel 1977, 50, 121. (28) Rush, A.; Potyka, W. Koks,Smola, Gar 1978, 23, 73.
atomic ratio H/C OK 0.88 0.87 0.89 0.71 1.02
0.12 0.17 0.19 0.07 0.37
Table 11. Elemental Analyses of Liquid Paraffin, Anthracene Oil, and Assam Coal after Different Extraction Steps atomic ratio sample % C % H % N N/C H/C liquid paraffin 86.90 13.10 1.81 anthracene oil 87.95 6.10 1.65 0.02 0.83 Assam coal 77.0 5.70 1.50 0.017 0.88 AO-extracted residue of Assam 78.5 4.80 2.20 0.024 0.73 coal AO-Qn-extracted residue 78.0 5.10 1.90 0.021 0.78 AO-Qn-LP-extracted residue 79.9 4.10 1.90 0.020 0.61 IR spectra of coal samples were recorded on a Perkin-Elmer infrared spectrophotometer using coal (1 mg) with KBr (49 mg) in the form of pellets.
Results Table I shows analyses of Assam coal and other Indian coals. Table I1 shows analyses of A 0 and LP. Assam coal was extracted by using different organic solvents separately. The extractability of Assam coal was observed to be a maximum in A 0 (36%) followed by LP (28%). The other solvents for extraction of Assam coal under atmospheric pressure conditions were ethylenediamine (EDA), diethanolamine, ethanolamine, and Qn. Among the solvents used, the extractibility was at a maximum in EDA (25%) followed by diethanolamine (23%), monoethanolamine (20%), and Qn (18%). Figure 1 shows the extractability of Assam coal in different combinations of solvents through successive extractions. Assam coal was extracted with Qn for 24 h. The residue can be further extracted in AO. The residue of Qn and subsequent A 0 extraction was further extractable in LP. Earlier, i t was reportedz9that successive extraction in a number of solvents does not exceed the total extractability. An observation contrary to the report was made in extractive disintegration under atmospheric pressure. The extraction sequence of coal using the above three solvents for 24 h (each) was changed to A w n - L P (Figure 1). Approximately 68% of the coal was rendered extractable through this AO-Qn-LP extraction sequence. The minimum extractability (48%) was obtained by using the Qn-LP-A0 extraction sequence. Other extraction sequences showed almost the same yields. These studies showed that it was possible to extract about 68% of the (29) Kiebler, M. W. The Chemistry of Coal Utilization;Lowry, H. W., Ed.; Wiley: New York, 1945; Vol. 1 p 715.
Energy & Fuels, Vol. 3, No. 5, 1989 643
Extractive Disintegration of Coal
-
80
70-
i a
b
36
36 34
C
d
e
Figure 1. Successive sequential extraction of Assam coal using LP, AO, and Qn for 24 h in each solvent: (open box) total extraction (YOon dmmf basis); (slanted lines) A 0 extraction (70on loss in weight basis); (crossed lines) Qn extraction (YOon loss in weight basis); (vertical lines) LP extraction (% on loss in weight basis). Key: (a) LP-AO-Qn solvent sequence; (b) LP-Qn-A0 solvent sequence; (c) AWn-LP solvent sequence; (d) AO-LP-Qn solvent sequence; (e) Qn-AO-LP solvent sequence; (f) Qn-LP-A0 solvent sequence.
coal within 72 h through successive extraction a t atmospheric pressure and without the use of hydrogen under pressure. Effect of Time on Successive Sequential Extractive Disintegration. Assam coal was successively extracted by using different combinations of solvents a t a constant time of extraction (2-24 h). It was observed that the extraction of coal increased with an increase in the extraction time of the original coal and the residual coals of various successions of the extraction (Figure 2). A continuous increase in the extraction yield was observed up to 6 h, and the further extraction yield was comparatively less with an increase in extraction time (up to 24 h). Therefore, it appeared that extractions were mostly complete and exhaustive after 24 h in each solvent. However, in the QnLP-A0 solvent system, extraction yields increased up to the first 6-h interval and then there was a gradual decrease in extraction yields up to 24 h. This decrease in extraction yields may be attributed to the possibility of an increase in cross-linking in the coal structure as a result of prolonged Qn and LP prior extractions. The solvent systems where successive extractions of coal started with A 0 (i.e. A w n - L P and AO-LP-Qn) showed the maximum extraction and, the differences in extraction yields between 12 and 24 h were 14% and 9%, respectively. This could be attributed to the extractive disintegration effect caused by A 0 extraction of the original coal. Successive (Sequential) Extractive Disintegration of Coal for 6 h. Shorter extraction times are desirable in batch and semicontinuous extractions for process development. A considerable amount of coal was found to be extracted within the first 2 h of extraction in each of the solvents tried in successive extractions (Figure 3). Figure 3 shows that extraction was about 48% of the coal within 6 h through successive solvent extraction. The sequences that showed maximum and minimum extraction yields after 72 h were different from those after 6 h (Figures 1and 3). This may be due to the fact that extractions were not complete after 6 h.
70
I-
-1
2r
IO
0 0
I
L
I 8
I
12
I 16
I 20
I 2L
TIME ( h 1
Figure 2. Total successive extraction of Assam coal using LP, AO, and Qn solvent in different sequences for 1-24 h in each
AO-LP-Qn; (0) LP-Qn-AO; ( 0 ) solvent: (X) LP-AO-Qn; (0) Qn-AO-LP; (A)AO-Qn-LP; (A)Qn-LP-AO.
Coal Disintegrating Ability of Solvents. A 0 showed the maximum extraction of Assam coal among the three solvents. Table I11 shows the Qn extractability of residual coals obtained after extraction in AO, LP, or Qn solvents, respectively. The Qn extractability of LP- and Qn-extracted residues was not appreciable, whereas the AO-extracted coal residue had enhanced Qn extractability. Thus, these studies showed that the coal was being chemically
Sharma and Mishra
644 Energy & Fuels, Vol. 3, No. 5, 1989
5045
-
40 35 -
a
b
d
C
e
f
Figure 3. Successive sequential extraction of Assam coal using LP, AO, and Qn for 2 h in each solvent. See Figure 1 for the key.
Table 111. Swelling Ratio and Extractability of the Original Assam Coal and Extracted Assam Coal Residues in Different Solvents 2h
coal sample
solvent LP LP LP LP LP
swelling ratio
OAC" AO-extracted coal residue AO-Qn-extracted coal residue Qn-extracted coal residue Qn-AO-extracted coal residue A0 OAC LP-extracted coal residue A0 LP-Qn-extracted coal residue A0 Qn-extracted coal residue A0 Qn-LP-extracted coal residue A0 Qn OAC LP-extracted coal residue Qn LP-AO-extracted coal residue Qn AO-extracted coal residue Qn AO-LP-extracted coal residue Qn " OAC = original Assam coal. Loss in weight basis.
depolymerized (disintegrated) during A 0 extraction. van Krevelen et al.25had reported that chemical extractive disintegration of coal takes place above 350 "C. However, with A 0 as solvent, chemical extractive disintegration of coal takes place a t 270 "C (and below 350 "C). Assam coal was extracted successively in three steps of 24 h each by using Qn, AO, or LP as a solvent alone. The extraction of the coal was observed in the succeeding two steps after the first step by using Qn or LP solvents alone, respectively. A considerable amount of the coal was extracted in A 0 (i.e., 36%, 22%,and 17% in the first, second, and third steps of the successive extraction, respectively). This c o n f i i the chemical depolymerization of coal during (successive) A 0 extraction. Thus, overall, 64% of the coal was extracted after three successive steps of A 0 extractions, which was less than the best extraction sequence (AO-Qn-LP) (68% 1. Elemental and IR Spectral Analyses of Extracted Coal Residues. The atomic H/C ratio in A 0 was found to be 0.82 (Table 11). The aromaticity of A 0 was 96% as revealed by 13C NMR spectral studies. Thus, this oil will enhance the extractability of the coal by shaking the aromatic ring clusters of coal. This was supported by elemental analysis of the residual coal obtained after A 0 extraction. It was observed that extractive disintegration
1.12 1.10 1.12 1.12 1.12 1.38 1.33 1.71 2.00 1.34 2.64 2.28 2.68 2.67 1.25
24 h %
extractionb 19 15 13 18 17 22 15 11 15 14 18 13 12 17 12
swelling ratio 1.12 1.12 1.12 1.08 1.10 1.38 1.33 1.74 2.24 1.30 2.64 2.20 2.72 2.80 1.32
%
extraction* 28 19 15 17 15 36 20 23 34 17 19 9 18 30 22
of coal resulted in reduction of the H contents of coal. Dehydrogenation of coal as a result of A 0 extraction can be seen from the H/C ratio of AO-extracted coal residue (0.73) which was lower than that of the original coal (0.88) (Table 11). The nitrogen content in the residual coal obtained after A 0 extraction increased in comparison to that of the original coal (Table 11). A 0 has been reported30 to contain a number of nitrogenous compounds. The N/C ratio was at a maximum in the AO-extracted coal residue (Table 11). This confinned the chemical incorporation of N-containing compounds to coal. The basic nitrogenous compounds were stable up to 150 "C under vacuum. The IR spectrum of AO-extracted coal residue also showed some additional absorptions a t 1720, 1760, and 1780 cm-' (imides, cyclic a,p-unsaturated five-membered ring) and 1410 and 1310 cm-I (phenols). These absorptions were absent in the IR spectrum of the original Assam coal. Absorption around 1700 cm-' could also be due to oxidation of coal during extraction or workup. Extractive Disintegration of Other Coals. Other Indian coals such as Talcher, Raniganj, and Godavari coals (30)Chemistry of Coal Utilization;Elliott, M. A., Ed.; John Wiley: New York, 1981; Second Supplementary Volume, p 2006.
Extractive Disintegration of Coal
Energy & Fuels, Vol. 3, No. 5, 1989 645
Table IV. Successive (Sequential) Extraction of Different Coals in AO-Qn-LP Extraction System (Extraction for 2 and 24 h, Respectively in Each Solvent) % extractionn total extraction A0 Qn LP yield: % 2h 24 h 2h 24 h 6h 72 h coal sample 2h 24 h Assam coal 22 36 20 30 13 15 50 68 Talcher coal 16 20 18 18 10 12 46 51 Raniganj coal 14 19 12 16 10 10 37 46 Godavari coal 0 4 10 12 7 6 22 29 Neyveli lignite 23 35 24 26 15 12 52 63 OOn loss in weight basis. *On a dmmf basis.
and Neyveli lignite were studied for successive extractive disintegration by using the A w n - L P sequence for 2 and 24 h in each solvent respectively (Table IV). The increase in extraction time had little effect on the extraction yields of Talcher coal. However, Raniganj coal showed an increase in extraction yields when the extraction time was increased from 6 to 72 h (Table IV). Godavari coal showed no extraction in A 0 after 2 h of extraction. Overall, only 22% of Godavari coal (on a dmmf basis) was extracted after 6 h of successive extraction. The extraction of Godavari coal was only 29% even after 72 h of successive extraction in AO-Qn-LP (Table IV). These studies showed that the extractability of Godavari coal in AO, Qn, and LP solvents was poor, and this shows that it is a low-grade coal. Perhaps mineral matter in the Godavari coal forms organometallic complexes with coal organic matter that render it poorly extractable even through extractive disintegration. In contrast to Godavari coal, Neyveli lignite showed good extraction results through successive extractive disintegration in AO-Qn-LP solvents (Table IV). These studies showed that Neyveli lignite and Assam and Raniganj coals produced good extraction yields through AO-Qn-LP successive extractions. This showed the general applicability of the successive extractive disintegration of coals.
Discussion Possible Explanation for Successive (Sequential) Extractive Disintegration of Coal. Coal is composed of many structures. On treatment with solvent, the various coal compositions interact with solvent depending on the polarity and functionality of the solvent and the composition of the interacting structure of the coal. Qn, being a polar solvent, can extract structural units similar to its structure (e.g., heterocyclic compounds). N-containing solvents have also been reported35to be good solvents for coals, as these contain a lone pair of electrons. The extractability also depends on the temperature of extraction of a solvent and the boiling point of the solvent. Effect of Solvents on Successive Sequential Extraction. LP has a fixed quota for extraction irrespective of its position in the sequence of extraction (Table 111). LP, being a high-boiling solvent, extracts original or residual coal through some hydrogen interchanges in coal. The share of LP extraction is a maximum for the thermally degraded structure of the coal (through free-radical for(31) Chemistry of Coal Utilization; Elliott, M. A., Ed.; John Wiley: New York, 1981; Second Supplementary Volume, p 487. (32) Wender, I.; Heredy, L. A.: Neuworth, M. B.; Dryden, I. G. C. Chemical Reactions and Constitution of Coal. In Chemistry of Coal Utilization; Elliott, M. A., Ed.; John Wiley: New York, 1981; Second Supplementary Volume. (33) Wood, R. E.; Wiser, W. H. Ind. Eng. Chem. Process Des. Deu. 1976,15, 144. (34) van Krevelen, D. W. Presented at the 6th International Conference on Coal Science, Munster, FRG, 1965. (35) The Chemistry and Technology of Coal; Speight, J. G., Ed.; Marcel Dekker: New York, 1983, Vol. 12, p 205.
mation) and the structure which was stabilized by H atom interchange between 330 and 360 0C.4*31It was surmised that coal is being degraded by pyrolysis at the boiling range of LP (i.e. 330-360 "C). Pyrolysis of Assam coal a t 360 "C in a nitrogen atmosphere resulted in a 3% loss of weight of the coal. Therefore, the loss in weight of coal through LP extraction is not entirely due to the pyrolysis of coal; instead, LP was able to extract an appreciable amount of coal through extractive disintegration of coal. These studies showed that the matter volatile at 360 "C was less than the matter extractable in LP a t the same temperature. There may be some overlap, but these are very different fractions. Evolution of some gases was observed during LP extraction of coal, but a detailed discussion of these results will follow in the next paper on this subject. Preextraction with LP had an adverse effect on extraction in the succeeding solvent (Table 111). This could be due to the fact that LP has a high boiling range (i.e., 330-360 "C). Coal is thermally depolymerized in this temperature range and this results in disrupting of the coal structure, including its pore structure (as revealed by scanning electron microscopic studies). On cooling, the residual coal from the LP extraction repolymerizes and reconstitutes physically and chemically through some condensation reactions, forming cross-linkages. On the other hand, when coal is extracted with AO, it heats to only 270 "C, and a t this temperature, coal interacts with AO. Some of aromatic units from A 0 to coal takes place. An interchange of H atoms of coal with solvents a t elevated temperatures has been reported earlier by several res e a r c h e r ~ .Moreover, ~ ~ ~ ~ some reactions between coal and A 0 also take place. As a result, coal gets depolymerized and its extractability in Qn is enhanced. Qn was found to have a neutral effect on the successive sequential extraction of coal (Table II). It was found that it is A 0 which has a major effect in the successive sequential extraction of coal. Three factors can contribute to the diminishing of the H contents of coal after A 0 extraction: (1) extraction of the hydrogen-rich portion of coal in AO; (2) aromatization of the hydroaromatic structure of coal catalyzed by alumina etc. present in coal mineral matter; (3) the addition or interchange of the aromatic portion of AO. In fact, all three factors seem to be responsible. Hydrogen-rich compounds from coals were extracted by A 0 as revealed by elemental analysis and IR spectral analysis of the A 0 extract. The predominance of absorption a t 710 cm-l in the IR spectrum of AO-extracted residual coal showed the presence of aromatic rings having four adjacent H atoms in aromatic rings. Such aromatic rings are present in AO, which might have been added on to coal during the extractive disintegration.
Conclusions The following conclusions can be drawn from present studies.
646
Energy & Fuels 1989,3, 646-647
1. Anthracene oil extraction is a beneficial pretreatment for rendering coal further extractable in other solvents such as quinoline. 2. Preextraction using liquid paraffin has an adverse effect on the extractability of residual coal in other succeeding solvents. 3. It is possible to extract more than 50% of the coal by using an AO-Qn-LP solvent system within 6 h (by extraction with each solvent for 2 h). This may help in developing a process for obtaining solvent-refined (clean) fuel from coal under atmospheric-pressureconditions. The successive extractive disintegration may also be of interest in the development of a process for the liquefaction of coal through solvent extraction under ambient pressure conditions. 4. Coal remains extractable in other solvents even after a 24-h extraction in solvents such as anthracene oil, quinoline, or liquid paraffin. The extractive disintegration/depolymerization of coal in one solvent renders the residual coal (further) extractable in succeeding solvents. 5. Anthracene oil has been found to be the solvent of choice for the successive extraction of coal through the chemical depolymerization/disintegrationof coal.
6. Extraction of coal with anthracene oil results in diminishing of the hydrogen contents of the coal with introduction of nitrogenous units in the residual coal. 7. Extraction of coal in liquid paraffm remains the same irrespective of its position in the sequence of successive extraction. The extraction of coal in liquid paraffin is not a pyrolysis of coal at 360 "C. The extraction of coal in other solvents such as quinoline or anthracene oil varies according to the position in the sequence of extraction. 8. Successive extraction (24 h in each solvent) using the A w n - L P sequence renders 68% of the Assam coal, 51% of the Talcher coal, 46% of the Raniganj coal, 29% of the Godavari coal, and 63% of the Neyveli lignite extractable. This shows the general applicability of successive sequential extractive disintegration of coal in coal technology. Acknowledgment. We are thankful to Prof. J. W. Larsen and the reviewers for their useful comments and suggestions that have helped in the modification of this paper. We are also thankful to the Council of Scientific and Industrial Research, New Delhi, India, for financial assistance to carry out this work. Registry No. Qn, 91-22-5.
Communications Letter to the Editor on Analysis of Gas-Solid Reactions by Use of a Temperature-Programmed Reaction Technique
volumetric model (v)
Sir: In a recent paper in this journal Miura and Silveston' applied the temperature-programmed reaction method to the gasification of flash pyrolysis chars in air to distinguish
grain model (g)
between three rival gasification models: the volumetric model, the grain model, and one proposed by Bhatia and Perlmutter.2 As Miura and Silveston discuss, the temperature-programmed reaction (TPR) is a technique that increases temperature a t a constant rate through an experiment and measures the rate of reaction as a function of temperature. These authors found that they could not distinguish among the three models using a single TPR experiment. Each of the three competing models described the rate vs temperature data equally well (see Figure 2 in ref 1). However, these investigators found that if three TPR experiments with different heating rates were used, only the Bhatia and Perlmutter model fitted the experimental results satisfactory (see Figure 5 in ref 1). They viewed this result with surprise because the TPR method is supposed to reduce the number of experiments needed for model fitting. In this letter, we attempt to explain briefly why the single TPR run was unsuccessful and why it is necessary to employ a t least two TPR measurements. Integral forms of the three models are as follows:
Bhatia and Perlmutter model (b)
(1) Miura, K.; Silveston, P. L. Analysis of GasSolid Reactions by Use of a Temperature-Programmed Reaction Technique. Energy Fuels 1989, 3, 243. (2) Bhatia, S. K.; Perlmutter, D. D. A Random Pore Model for Fluid-Solid Reactions: I. Isothermal, Kinetic Control. AIChE J. 1980, 26, 379.
1 - X = exp(-.r,)
1 - x = (1 - 7,/3)3 1 - X = exp[-Tb(l
+ J..b/4)]
where 7i
kiRP aEi
= -exp(-Ei/RT)
i = v, g, b
and where X,a, k , E, R, and Tare respectively conversion, heating rate, frequency factor, activation energy, gas constant, and temperature. If a single TPR measurement is described satisfactorily by any of three models, the form of any one model equation should be transformable to that of another model. This is possible, indeed. For example, the Bhatia and Perlmutter model can be transformed into the volumetric one by replacing 1 + +rb/4 with the term ae@IRT, where a and P are constants, as may be shown by substitution in the expression above. A result of the transformation is that a and P depend numerically on the model parameters E,, &, k,, tzb. The two models become equivalent if a and @ take on values so that the latter term, aeBIRT,is nearly equal to the former, 1 + #Tb/4, through the range of possible reaction temperatures. This equivalence is shown in Figure 1. The bold curves indicate changes in In (1 + J/rb/4) with reciprocal temperature. Kinetic parameters determined by Miura and Silveston' for char B are used for these curves where solid portions correspond to the actual reaction temperatures used in the
0S8~-0624/89/2503-0646$01.50/0 0 1989 American Chemical Society
