It has been found that naturally hydrophobic carbonaceous materials such as coal and graphite can be selectively coagulated and separated from hydrophilic impurities without the use of oily agglomerants, flocculants or electrolytes. The coagulation occurs at ζ-potentials significantly higher than those predicted by the classical DLVO theory, suggesting that it is driven by a hydrophobic interaction energy. Thus, the process is referred to as the selective hydrophobic coagulation (SHe) process. The fundamental development of this process is the focus of this study. In this study, the energy barriers for the coagulation of two different coal samples and a graphite sample have been calculated using the extended DLVO theory, which incorporates the hydrophobic interaction energy in addition to the dispersion and the electrostatic energies. Stability diagrams have been developed from the data, which show that the maximum ζ-potential at which a given coal can coagulate decreases as surface hydrophobicity decreases. For the coagulation of minerals present in coal, the classical DLVO theory has been used for the energy barrier calculations. The results of these calculations provide an excellent correlation with the results from a series of SHC tests conducted with run-of-mine coal. The strength of the coal aggregates have also been investigated by measuring the coagula size distributions under different hydrodynamic conditions. The coagula size distributions were measured using an in-situ particle size analyzer. These results have been used along with models for coagulation rate and breakage rate to determine strength characteristics of the aggregates and to verify the primary parameters controlling the aggregate size. The study found that the coal and graphite aggregates incurred a substantial reduction in size when a small amount of mechanical agitation was applied. Based on this outcome, quiescent continuous processes have been successfully designed and developed to separate the coagulated hydrophobic particles from the dispersed hydrophilic particles. / Ph. D.

M.S.

M.S.

M.S.

On the basis of the results obtained in this investigation, the following conclusions were drawn: 1. Midlothian coal was 31.3 percent liquefied by hydrogenation at atmospheric pressure and 185°C after 72 hours of reaction in a tetralin vehicle and using stannous sulfide as a catalyst. 2. The liquid product of the hydrogenation of Midlothian coal consisted of approximately 50 percent light oil boiling below 200°C., 20 percent middle oil, boiling from 205°C. to 235°C., and 30 percent heavy oil or tar residue. 3. The light oil fraction of the hydrogenated product of Midlothian coal consisted of 5 to 10 percent each of olefins and saturated hydrocarbons, with the remainder being aromatic compounds. 4. The rate of liquefaction of Midlothian coal by hydrogenation researched a nearly constant value of 0.41 grams of liquid product produced per 100 grams of coal per hour after 24 hours of reaction. 5. During the first 24 hours period of hydrogenation, the volatile matter content of the carbonaceous residue increased by 5.0 percent after which the volatile matter decreased by 6.2 percent during the following 48 hour period. 6. During the hydrogenation reaction of 72 hours the fixed carbon content of the coal was reduced by 18 percent. 7. During the latter 48 hours period of reaction, 65 percent of the liquefaction was from the volatile matter constituents and 35 percent of the liquidation was from the fixed carbon components. / M.S.

The understanding of liquefaction behaviors, related to a coal's properties and a recycle solvent's composition, is essential for the development of an efficient direct liquefaction process. In this dissertation, a study of the liquefaction behaviors of an Eastern us bituminous and four Western US subbituminous coals is presented. The experimental approach has been to examine their behaviors under various reaction conditions with in-house microautoclave reactor and Kerr McGee pilot plant liquefaction runs. In-house runs involved surveys of coal types and process solvent compositions with variations in reaction times, temperatures and atmospheres. Runs performed at Kerr McGee examined the use of tetrahydroquinoline (TBQ) as a process solvent with a Wyoming coal. Liquefaction activities were assessed through determinations of coal conversion to both solvent-soluble products and distillate yields. Per the in-house liquefaction studies, a novel microautoclave reactor design and product recovery methods were developed, evaluated and employed. The reaction chemistries of !n !!S!! metal species and basic nitrogen heterocycles were investigated specifically. Changes in trace element concentrations were ascertained by Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES) and Size Exclusion Chromatography/ICP-AES (SEC/ICP-AES). Pates of basic nitrogen components in distillate and solvent-soluble residuum products were examined by nitrogen mass balance determinations, Gas Chromatography/Pourier Transform Infrared Spectrometry (GC/PTIR) and Gas Chromatography/Mass Spectrometry (GC/MS). Conversions to soluble products demonstrated the expected dependencies of liquefaction on coal rank, elemental composition and petrography. The western subbituminous coals showed extreme sensitivity to drying and solvent-soaking pretreatments. Metal content analyses revealed that metals exist as complexed species in the liquefaction process. Higher conversions to toluene-soluble materials were obtained with THQ in contrast to other H-donor solvents. Adduction of THQ was significant in the non-distillate product stream, however. The direct coupling of Reversed Phase HPLC separations with PTIR (RP-HPLC/PTIR) detection through on-line, post-column extraction was developed. Though intended for application to coal-liquefied product (CLP) analysis, this system was evaluated rigorously for both chromatographic and spectral performance. Throughout this investigation, the overall utility of these hyphenated methods for CLP analysis was explored. These methods demonstrated exceptional performance in providing a wealth of qualitative and quantitative information in a rapid manner. / Ph. D. / incomplete_metadata

Call number: LD2668 .T4 1978 H77 / Master of Science

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1980. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND SCIENCE. / Bibliography: leaves 294-308. / by Howard D. Franklin. / Ph.D.

Methods for the preparation of active carbons from many kinds of carbonaceous material have been described in the literature. Many processes of activation for many different raw materials are used to obtain active carbons for specific purposes. In general, all of these processes involve a low temperature carbonization of the raw material followed by a slow, controlled oxidation of the carbonized product. A high temperature of carbonization (above 600°C.) results in a product which is not active and cannot be activated. Any selection of a process or raw material must be based on a knowledge of the ultimate use of the product as well as on economic considerations. Certain physical properties are desirable for certain uses in addition to the general property of being adsorbent to foreign molecules. A gas adsorbent carbon should be dense with a rather small pore size; while liquid adsorbent carbons should be less dense, not triable, easily filterable from solutions, and have a larger pore size than the gas adsorbent type of carbon. In view of the low yield obtained in any process of activation, a cheap and plentiful raw material would be advantageous. Coal is such a raw material and active carbons have been prepared and used to a limited extent from coals. It has been reported that an initial chlorination of a geologically young coal before its carbonization results in a high yield of a good active carbon. The existence of large deposits of such a coal in the Piedmont section of Virginia and North Carolina which has not been exploited to any great extent, because it is not suitable as a fuel, seems to warrant a further investigation of this chlorination process with an idea of its possible economic use in the preparation of an active carbon. The purpose of this investigation in the preparation of a liquid adsorbent active carbon from a high volatile Midlothian coal by a process of chlorination followed by carbonization and steam activation. / M.S.

In recent years supercritical fluid chromatography has gained attention as an alternative technique to high performance liquid chromatography for the analyses of nonvolatile or thermally labile compounds, whose analysis with gas chromatography is impossible. The work presented here demonstrates a system that allows supercritical fluid extraction of the sample with subsequent direct introduction of a fraction of. this extract onto the column for analysis with supercritical fluid chromatography. Such a procedure has several inherent advantages to traditional sampling, where extraction or dissolution of the sample is done in a liquid. A valving scheme is described to accomplish this task and is evaluated for use with several different samples, including fuels and food products, with direct comparisons made between traditional sampling and direct on-line extraction. In most cases, the chromatograms were very similar, however later eluting components were sometimes lower in concentration with this method when compared to traditional sampling techniques. The apparatus was demonstrated in the analysis of high boiling coal derived material. Analysis of this material is accomplished by preliminary class separation with subsequent supercritical fluid extraction and analysis by packed column supercritical fluid chromatography. Detection included variable wavelength UV and FTIR spectrometry. The coal derived products studied were taken from a bench scale coal liquefaction reactor, in which the same catalyst was used for twenty-five consecutive days. The changes that occur as the catalyst decays were determined chromatographically for a portion of the high boiling products. The changes were noted best in the aromatic fraction analysis, where a trend towards molecules with higher numbers of condensed rings was observed as the catalyst decayed. / Master of Science