Characterization of vitrinite concentrates. 2. Magic-angle 13C nuclear magnetic resonance studies

The fraction aromaticity determined by ¹³C n.m.r. with cross-polarization and magic-angle spinning of 19 vitrinite concentrates obtained from the Lower Kittaning seam shows a range of values from ≈ 0.65 for the samples of lowest rank (83 wt% C (dmmf) to about 0.83 for those of highest rank (91 wt%C (dmmf)). It was determined that the wt% aromatic carbon correlates to the wt% fixed carbon and is in good agreement with the results reported by other authors. The combination of the ¹³C n.m.r. results with FTIR measurements allows a number of coal parameters to be estimated. The atomic ratios of aliphatic hydrogen to carbon were demonstrated to vary from 1.8–2.0 to between 2.4–2.6 and previous assumptions that a single value can be used in calculating structural parameters for coal of any rank are not strictly valid. The calculation of the aromatic $\text{H}\text{C}$ ratio indicates that in mean structural units there is approximately one aromatic hydrogen atom for every six carbons in vitrinites of carbon content 83 wt%C (dmmf) and that this ratio changes progressively with rank to a value of about one aromatic hydrogen for every four carbons for vitrinites of carbon content 91 wt%C (dmmf).     

Heavy media separation of SRC-II feedstocks: 1. Effect on coal properties

The objectives of this investigation were to determine the effects of coal preparation on the properties of Run-of-Mine (ROM) and washed Powhatan and Ireland Mine coals and to assess the potential effects on SRC-II liquefaction yields. The effect of washing on the two coals was found to be quite similar. For both coals, the properties were altered more significantly by changes in separation media gravity than by changes in the coal size. The elemental composition of the Powhatan and Ireland washed coals was correlated with carbon content. It was shown that both the hydrogen and oxygen levels increased linearly with the carbon content of the coal samples. However, the $\text{H}\text{C}$ and $\text{O}\text{C}$ ratios were not changed significantly by coal cleaning. Only small variations in the nitrogen and organic sulphur levels were observed while the sulphate sulphur and chlorine levels were not affected by coal cleaning. The major impact of the coal cleaning was to reduce the pyritic sulphur (and hence the total sulphur) content of the coals. Most of the pyritic sulphur was shifted into the middling coal and refuse fractions while the clean coals had much lower contents and the pyritic sulphur level decreased with increasing carbon content. Coal cleaning did not significantly alter the maceral contents of vitrinite, exinite, total reactive macerals (TRM), or the reflectance of vitrinite; all these parameters varied over a very narrow range, probably within the precision of the measurement technique.     

Conversion of bituminous coal in COH2O systems: 2. pH Dependence

Under $\text{CO}\text{H}{}_2\text{O}$ systems at initial pH values s> 12.6, an Illinois No. 6 coal, PSOC-26, was converted to a fully pyridine-soluble product, with benzene and hexane solubilities of 50% and 18%, respectively. The product gases were H₂ and CO₂. However, the expected $\text{H}{}_2\text{CO}{}_2$ ratio of 1.0 based on the water gas shift reaction was not observed, but the deficit in hydrogen was found in the increased hydrogen content of the coal product. 95% coal carbon recovery and good hydrogen balances were obtained, and the coal products were found to be very similar to those from conventional tetralin systems. The results suggest an efficient base-catalysed process, and that $\text{CO}\text{H}{}_2\text{O}$ systems are useful for coal studies.     

Properties of vitrinite concentrates of South African coals

Low-ash vitrinite concentrates of representative South African coals have been prepared and analysed. Various properties when plotted against carbon content show inflections at $\text{1}\text{1}\text{2}$ to 2% lower carbon content than normal British coals. These differences are ascribed to the heating effects of dolerite intrusions.     

Relation between coal aromatic carbon concentration and proximate analysis fixed carbon

Good agreement has been obtained between measured proximate analysis values for fixed carbon (FC) and the predictions of a thermal decomposition model. The model provides a basis for understanding the relation between FC and coal structure and between FC measured under proximate analysis conditions and coke or char measured in other thermal decomposition experiments. The key parameters in the model are the aromatic carbon concentration (C_ar) and the tar yield. C_ar has been determined for 43 coals using quantitative infrared analysis. The aliphatic hydrogen concentration is measured from the absorption near 2900 cm^?1 and the aliphatic carbon concentration is computed assuming a stoichiometry of CH_1.8 C_ar is then computed by difference. The results verify the good correlation between C_ar and FC discussed by van Krevelen. To explain this correlation, use has been made of a coal thermal decomposition model which has been successful in simulating the quantity and composition of volatile components yielded under vacuum pyrolysis conditions. To apply the model to proximate analysis, it was necessary to estimate the tar yields obtained with thick beds and the amounts of O, N, H, and S which remain with the FC. The tar yields for proximate analysis conditions have been estimated to be $\text{1}\text{3}$ to $\text{1}\text{4}$ the yields for thin beds in vacuum. To determine the composition of the FC, measurements were made on a lignite and a bituminous char produced in a thin bed heated by a wire grid for the time (7 min) and temperature (950 °C) used in the proximate analysis, and on the FC residues from a proximate analysis volatile matter determination. Both residues give similar results, showing that approximately 10% of the ‘fixed carbon’ is not carbon. Values of FC computed with the model adjusted for the above conditions are in good agreement with the measured values.     

Flash pyrolysis of coals: comparison of results from 1 g h−1 and 20 kg h−1 reactors

The performances of 1 g h^?1 and 20 kg h^?1 flash pyrolysers are compared for three Australian coals: Loy Yang brown coal (Victoria), Liddell bituminous coal (New South Wales), and Millmerran sub-bituminous coal (Queensland). The two reactors gave comparable yields of tar, char and C₁–C₃ hydrocarbon gases over a range of operating conditions for each particular coal. The yield of total volatile matter from Millmerran coal was similar from both reactors, as were the compositions of chars from Loy Yang coal and tars from the Liddell and Millmerran coals. For Millmerran coal, the yields of tar, C₁–C₃ gases and volatiles from the large reactor below 650 °C, were slightly lower than for the small reactor, possibly owing to a shorter retention time of Millmerran coal particles in the large-scale reactor. At a temperature near 600 °C tar yields were independent of tar concentration in the effluent gas, over a range 0.0025–0.1 kg m^?3 for Liddell coal, 0.005–0.26 kg m^?3 for Millmerran coal and 0.0045–0.09 kg m^?3 for Loy Yang coal. The tar yields from Millmerran and Liddell coals at 600 °C in the large reactor, correlate directly with the atomic $\text{H}\text{C}$ ratio of the parent coal, in the same manner as that found for a wider range of bituminous coals in the small-scale reactor.     

Re-examination of the phenolic hydroxyl contents of coals

In connection with studies of the dependence of liquefaction behaviour on coal characteristics, parameters were needed that might effectively characterize the organic chemical structures. Accordingly, the phenolic hydroxyl contents of 37 coals from three geological provinces of the USA have been determined, by acetylation with ¹⁴C-labelled acetic anhydride. The results, when expressed as fractions of the total organic matter in the coals, showed a good inverse correlation with the carbon contents (dmmf). However, application of a stepwise multiple regression analysis to the data developed a linear equation relating hydroxyl content to the vitrinite reflectance, calorific value and vitrinite content, the fraction of variance explained being 92%. When the hydroxyl contents were expressed in the alternative manner, as fractions of the total oxygen content, no correlation could be seen with carbon content. However, in the wide scatter of points on the graph, the data are seen to fall into three reasonably distinct populations such that at the same level of rank, hydroxyl contents typically decrease for coals from the three provinces in the order Interior > Eastern > Rocky Mountain, implying that coals from these areas differ in structure as a result of differing antecedents. Statistical analyses showed that $\text{OH}\text{O}$ has some significance in determining liquefaction behaviour, but it is not among the coal properties found most significant.     

DTGA combustion of coals in the Exxon coal library

Coal combustion experiments were carried out over the temperature range 25–900 °C using air at atmospheric pressure in a derivative thermogravimetric analysis system. Sixty-six coals high in vitrinite (> 80% mineral-matter-free basis) and low in inorganics (all but 12 samples < 10%) were examined as part of a coal characterization programme. The coals varied in rank from lignites (69% carbon on a dry, mineral-matter-free basis) to low-volatile bituminous (91% carbon). Combustion rates increased progressively with increasing temperature, passed through maxima and then declined. The rate data were fitted to an Arrhenius equation and plots showed four distinct regions of combustion. Apparent activation energies were calculated for each region and varied from ≈4 kJ mol^?1 in the high-temperature, diffusion-controlled region to 290 kJ mol^?1 in the chemical-reaction controlled, low-temperature region. The temperatures at which 50% of the sample had burned away ($\text{sol1}\text{2}\text{-}\text{life}$) were rectilinearly related to oxygen and carbon contents (correlation coefficient squared values of 0.88 and 0.86, respectively).     

Excited triplet states in bituminous coal

The intensities of the broad and narrow lines constituting the e.p.r. spectrum of coal depend on the temperature of measurement. The system of paramagnetic centres related to the broad line does not obey the Curie law and consists of free radicals ($\text{S =}\text{1}\text{2}$) and of centres with singlet ground state (S = 0) and lowest-excited triplet state (S = 1). The singlet-triplet excitation energy, J, is comparable with thermal energy, kT, of lattice vibrations. The system of paramagnetic centres related to the narrow line obeys the Curie law ($\text{S =}\text{1}\text{2}$).     

Fixed bed catalytic reactor modelling: The heat transfer problem

The issue of radial heat transport in fixed beds is discussed and discrepancies between observed and computed hot spots are attributed to neglect of axial dispersion of heat, which neglect also accounts for observed length-dependent radial conductivities.The Nusselt number for heat transport at the wall is inferred to be Nu_w = 5.73($\text{D}{}_{\mathrm{t}}\text{d}{}_{\mathrm{p}}$)^{$\text{1}\text{2}$}$\text{P}\text{r}$(0.11Re_p + 20.64)/Re_p^0.262 which is in accord with recently secured Sherwood numbers for mass transport at a packed wall.The recent success of unconventional reactor models is discussed.     

Raman spectra of supported molybdena catalysts: II. Sulfided,used, and regenerated catalysts

Raman spectroscopy has been used to investigate the structural changes that two supported molybdenum oxide catalysts undergo upon specific chemical treatments. Molecular MoS₂ structures are indicated after sulfidation by a mixture of $\text{H}{}_2\text{H}{}_2\text{S}$. Catalyst samples used in a coal hydrodesulfurization process yield spectra dominated by intense scattering from carbon deposited in the pores of the catalyst. Spectra of used catalyst samples, subjected to controlled air-firing to 600 °C, show that all of the spectral features of the unused catalyst are not recovered after this “regeneration” procedure.     

Mineral matter as a measure of oxidation of a coking coal

The petrography, agglomerating characteristics, low-temperature ash (LTA) mineralogy and chemistry of ten consecutive channel samples from D seam, Natal Ridge, Crowsnest coalfield, British Columbia, were compared. It was found that there is a direct correlation between oxidation, as indicated by absence or limited caking character, and the presence of the mineral bassanite ($\text{CaS0}{}_4\text{·}\text{1}\text{2}\text{H}{}_2\text{O}$) in LTA. An attempt to estimate quantitatively the extent of oxidation by the anhydrite content in LTA samples further heated to 500 °C has resulted in the detection of partial oxidation of a coal sample having an FSI of $\text{6}\text{1}\text{2}$, reduced from a normal value of 8. The recognition and quantification of oxidation using epigenetic gypsum and its derivatives, however, probably only applies to the weakly pyritic, usually freshwater, coals typical of the Rocky Mountain coalfields of British Columbia and the Gondwana coalfields of the southern hemisphere.     

Fluidized-bed gasification of some Western Canadian coals

Three Western Canadian coals were gasified with air and steam in a fluidized bed of 0.73 mm sand and coal, at atmospheric pressure and temperatures of 1023–1175K to produce a low-calorific-value gas. One non-caking and two caking coals were tested. The effects of temperature, coal feed rate, $\text{air}\text{coal}$ ratio, $\text{steam}\text{coal}$ ratio, coal quality, coal particle size and bed depth on gas composition, gas calorific value and operating stability of the gasifier were established. Results are compared with those previously obtained for the same three coals when gasified in essentially the same equipment, but operated as a spouted bed.     

In situ laser raman spectroscopy of the sulfiding of Moγ-Al2O3 catalysts

Raman spectra of sulfided $\text{Mo}\text{γ-}\text{Al}{}_2\text{O}{}_3$ catalysts were obtained using in situ techniques for two sulfiding methods. For samples sulfided by 10% $\text{H}{}_2\text{S}\text{H}{}_2$ at 400 °C, MoS₂ structures were observed. A stepwise sulfiding using 10% $\text{H}{}_2\text{S}\text{H}{}_2$, with spectra recorded at 150, 250, and 350 °C, resulted in observation of molybdenum oxysulfide, reduced molybdate, and surface “MoS₂” phases. Reexposure of these samples to air led to radical modification of the oxysulfide structures as well as transformation of some sulfide phases. A model incorporating terminal and bridging MoS bonding and anion vacancies is proposed. This model is based on the conversion of isolated and aggregated molybdate and MoO₃ species to oxysulfide and reduced molybdenum phases. Conversion of reduced molybdenum phases to sulfides is observed to be slow.     

Reduction of coal in hexamethylphosphoramide-Na-t-BuOH and the molecular weight distribution of pyridine-soluble material

Eight coals were reduced by hexamethylphosphoramide (HMPA)-Na-t-BuOH at room temperature and atmospheric pressure. More than 70 wt % of the coal is soluble in pyridine after this reduction. Coals having ≈88 wt% of carbon content show a maximum solubility in pyridine after reduction. The pyridine-soluble material has a higher $\text{H}\text{C}$ ratio after reduction. l.r. and n.m.r. spectra show that the reaction product is richer in aliphatic and olefinic structures and that the aromatic ring systems are extensively saturated. From g.p.c. separations and molecular weight measurements, it is found that the molecular weight distribution has a maximum at lower and higher molecular weight values. The average molecular weight of the reduced pyridine-soluble material decreases with increase in rank of the coal. Coking properties of these coals can be related to proportions of material in the two modes of distribution of molecular weight.     

The production of active carbon from brown coal in high yields

The acid-leached char produced in high yield by the pyrolysis of the potassium salt of a brown coal (Yallourn seam, Victoria, Australia) is shown to have an internal surface area in excess of $\text{1100}\text{m}{}^2\text{g}$ without the need for activation. Tabletting, or extrusion, of the potassium salt of the coal permits production of a hard, active carbon as shaped granules. Of the cations investigated — sodium, potassium and magnesium — potassium was unique with regard to the formation of such a highly active char.     

Structural study of hexane-soluble products from hydroliquefaction of three Yallourn brown coal lithotypes over molten salt catalysts

The hexane-soluble fractions of hydroliquefied products from three Yallourn brown coal lithotypes have been separated into five fractions by combined silica-alumina packed column chromatography. Analyses of various fractions by g.c.-m.s. permitted the identification of ≈50 components in the saturate fraction and 40 components in the diaromatic fraction, together with 30 components in the monoaromatics. The components identified were quite similar among hexane-soluble portions of all three lithotypes. A marked predominance of even carbon number alkane (C₂₃-C₂₉) was observed in the hydrocarbon fractions from pale lithotype over $\text{ZnCl}{}_2\text{KCl}$, and medium light lithotype over both pure ZnCl₂ and $\text{ZnCl}{}_2\text{KCl}$. However, medium dark lithotype over both melt catalysts produced a saturate fraction with an odd carbon number(C₂₂-C₂₈) preference. Based on spectral methods, Soxhlet extracts obtained from untreated lithotypes (hexane and benzene solubles) were characterized as complex mixtures of higher molecular weight(300–1000) aliphatic hydrocarbons, which were supposed to be a precursor of the saturates produced from the corresponding lithotype in the catalytic hydroliquefaction.     

The characterization of asphaltenes and preasphaltenes obtained under various coal liquefaction conditions

A CP-MAS ¹³C NMR study of asphaltenes and preasphaltenes obtained under various coal liquefaction conditions is reported. The carbon aromaticity, f_a, of the solid extracts from the reaction products has a close relation with the reaction conditions. By plotting f_a against the atomatic $\text{H}\text{C}$ ratio for these solid products on the characterization chart for model polycyclic aromatic compounds, the molecular structures with relation to the liquefaction pathway from coal to oil can be proposed.     

Interpretation of the chemical composition of bituminous coal macerals

Various methods of identifying and classifying macerals are evaluated on the basis of literature references and our own experimental results. Comparison of the diagrams $\text{H}\text{C}$ vs $\text{O}\text{C}$, atomic $\text{C}\text{H}$ ratio vs weight percentage of carbon, and atomic $\text{C}\text{H}$ ratio vs atomic percentage of carbon demonstrates the advantages of the characterization of maceral composition by means of the atomic carbon percentage, which is a function of the content of all five organogenic elements that enter into the composition of the combustible matter of caustobioliths. Illustration by means of maceral analyses from the same seam, of results from pyridine extraction, and of humic acid composition further confirms this view.