THE CHEMISTRY AND KINETICS OF DONOR SOLVENT COAL LIQUEFACTION

In this work the chemistry and kinetics of donor solvent coal liquefaction at temperatures from 398(DEGREES)C to 440(DEGREES)C was studied. At 440(DEGREES)C two solvents, tetralin and 9, 10 dihydrophenanthrene, were used. The effects of temperature and the solvent upon liquefaction were examined with a Pennsylvania HVB type bituminous coal (Penn. State #330). Also, at 440(DEGREES)C in tetralin, the liquefaction of a MV bituminous Pennsylvania coal (Penn. State #256) was examined. A micro-batch reactor which held about one gram of coal and four grams of solvent was used to conduct the experiments. A fluidized bed sand bath provided rapid heating to the desired reaction temperature. Helium gas was used to pressurize the reactor to 1500 psig. No H(,2) was used. Gas chromatography was used to determine the solvent dehydrogenation and to analyze the gaseous products produced. Elemental analyses, proton NMR and hydroxyl group analyses by acetylation produced information about the structure of the SRC product. A chemical structural model was developed based on known coal structure and liquefaction chemistry. Important chemical structural groups which could be determined from the information about the product and the initial coal were identified and defined as state variables. The behavior of the state variables during liquefaction under the different conditions of this study was examined and compared with expectations from the literature. The coals used in this study were particularly chosen because of the large amount of data in the literature concerning their structure and liquefaction behavior. PSOC 330 coal liquefied primarily by cleavage of ethylene bridges, with ether linkages relatively uninvolved in the process. Significant retrogression of the conversion was observed in tetralin, but this was inhibited in 9, 10 dihydrophenanthrene. In PSOC 256, ether linkages contributed to the liquefaction and no retrogression was observed. A kinetic model was formulated using simple reactions to describe the chemistry, but included the solvent concentration. The model, with fitted parameters, was able to describe the individual reactions and the overall conversion quite well for PSOC 330. For PSOC 256, the slower liquefaction rate was not predicted from the individual reactions, nor was a lack of retrogression.