Ambient-pressure thermogravimetric characterization of four different coals and their chars

Ambient-pressure thermogravimetric characterization of four different coals and their chars was performed to obtain fundamental information on pyrolysis and coal and char reactivity for these materials. Using a Perkin-Elmer TGS-1 thermobalance, weight loss as a function of temperature was systematically determined for each coal heated in helium at 40 and 160 °C/min under various experimental conditions, and for its derived char heated in air over a temperature range of 20 to 1000 °C. The results indicate that the temperature of maximum rate of devolatilization increases with increasing heating rate for all four coals. However, heating rate does not have a significant effect on the ultimate yield of total volatiles upon heating in helium to 1000 °C; furthermore, coupled with previous data⁹ for identical coal samples, this conclusion extends over a wide range of heating rate from 0.7 to 1.5 × 10⁴ °C/s. Using the temperature of maximum rate of devolatilization as an indication of relative reactivity, the devolatilization reactivity differences among the four coals tested that were suggested by this criterion are not large. For combustion in air, the overall coal/char reactivity sequence as determined by comparison of sample ignition temperature is: N. Dakota lignite coal ≈ Montana lignite coal > North Dakota lignite char > III. No. 6 bituminous coal ≈ Pittsburgh Seam bituminous coal > Montana lignite char > III. No. 6 bituminous char > Pittsburgh Seam bituminous char. The reactivity differences are significantly larger than those for devolatilization. The reactivity results obtained suggest that coal type appears to be the most important determinant of coal and char reactivity in air. The weight loss data were fitted to a distributed-activation-energy model for coal pyrolysis; the kinetic parameters so computed are consistent with the view that coal pyrolysis involves numerous parallel first-order organic decomposition reactions.     

Experimental study on mass transfer from pyrolysing coal particles

The molecular weight distributions of coal tars and coal char extracts were examined in an effort to learn more about the process of mass transfer during coal pyrolysis. Evidence was obtained which suggests that the majority of the tar evolved during rapid pyrolysis of pulverized coal escapes by a process limited by gas film diffusion. However, there is also some evidence that the tar includes a small amount of heavy material which could have been ejected from the particle in a condensed phase. Data were also obtained which suggest that the tar precursors (within the parent coal) are formed over a wide range of temperature and do not seem to be present as such in the raw coal. The rather large effect of pressure on yields of tar from bituminous coal pyrolysis has previously been attributed to the effect of pressure on evaporation rates of tar precursors from the particle surface. This study shows that the molecular weight distributions of both the tar and extractable tar precursors within the particle are consistent with such a mechanism.     

Reactivity of hydrolysed coals in liquefaction

A hvA bituminous, a subbituminous and a lignite coal have been hydrolysed by 20–30% aqueous caustic solution at 100–300 °C and total pressure from ambient to 8.3 MPa (1200 psi). Reactivity of these pretreated coals toward liquefaction has been examined. The conversion to benzene-soluble material (BS) and oil increases, and the preasphaltene and char residue decreases after pretreatment. Improvement in the conversion to the BS fraction is only marginal for the pretreated bituminous coal, but substantial for the low-rank coals. For the subbituminous coal, the liquefaction reactivity (conversion to BS) increases with the severity of hydrolysis pretreatment. Analyses of chemical compositions, ¹H n.m.r. nuclei distributions and hydroxyl concentrations of the acid-insoluble hydrolysis coal extracts indicate that both O and S are enriched in the extracts with half of the oxygen atoms being in hydroxyl forms. The hydroxyl concentrations of the extracts (acid-insoluble) are ≈2 to 3 times higher than their parent coals. Coal activation by this alkali pretreatment is explained by the hydrolytic attacks on ether C–O linkages, and the removal of some constituents rich in oxygen functional groups which are responsible for poor liquefaction behaviour.     

Cross-linking of macromolecules in coal

Cross-linking in bituminous coals has been studied by internal friction methods. A low temperature damping peak occurring at 130 K for a frequency of 1 Hz has been used to deduce the character of cross-linking in coal. The molecular units responsible for the internal friction are found to be methylene chains connecting ring structures of the macromolecular material. The number varies progressively from a Braunkohle with 67wt% carbon to anthracites of > 90 wt% carbon. The number goes through a maximum at ≈80 wt%C, showing that the cross-link density reaches a maximum at that carbon level. It falls of markedly at higher carbon concentrations and remains at a low level > 85 wt%C. These findings correlate well with models of coal structure.     

Study of the reaction of maleic anhydride with Illinois bituminous coal

Maleic anhydride reacted with Illinois No. 6 bituminous coal (mf) at 200 and 250 °C resulted in a 22.5 and 39.3% increase in weight, respectively, which was based on the benzene-insoluble residue. The carboxyl group content of this coal residue, after it had been refluxed with water and demineralized, was 2.01 meq g^?1 (459% increase) at 200 °C, and 2.74 meq g^?1 at 250 °C (626% increase). If one assumes an average molecular weight of 2000 for bituminous coal, these results correspond to approximately 2 mol dienophile reacting with each mol of coal. Carboxyl group content of the insoluble coal residue was determined by the barium acetate method.     

1H n.m.r. spin—lattice relaxation study in bituminous coal

¹H n.m.r. studies of spin-lattice relaxation of a bituminous coal were performed at temperatures between 128 and 308 K. The relaxation function was observed to be two-component over the whole temperature range. The component with the shorter relaxation time was assumed to be the molecular phase in coal and the component with the longer relaxation time to be the macromolecular phase. Mechanisms of spin-lattice relaxation through spin diffusion to paramagnetic centres and through modulation of dipolar interactions by molecular motions are suggested.     

Reactivity of bituminous coal chars during the initial stage of pulverized-coal combustion

The reactivities of pyrolysed and partially burned char particles prepared in an entrained-flow reactor have been investigated. The results indicate that chars collected at the end of the active devolatilization stage are more reactive than those collected before or after this stage. Deactivation of the pyrolysed chars was accompanied by the development of micropores. The chars produced from a liptinite-rich fraction of an HVA bituminous coal showed higher reactivities than those generated from an inertinite-rich fraction. It is suggested that residual volatiles play a more important role in determining char reactivity than the microporosity and the optical anisotropy of the chars. A new expression for TGA reactivity is suggested for use in deriving char combustion kinetics. Relatively constant activation energies of ≈ 125 kJ mol⁻¹ were obtained for chars prepared from a wide range of coal precursors. Calculated char combustion rates at high temperatures extrapolated from such reactivity parameters were in agreement with experimentally determined rates.     

Structural analysis of NaOH-alcohol treated coals

Six vitrinite-concentrated coal samples were treated with NaOH-alcohol at 300 or 350 °C for 1 h. The products were nearly all soluble in pyridine except those from the oldest coal (52.4% of extraction yield). Structural indices of these pyridine extracts were calculated from ultimate analytical values, molecular weight and ¹H-n.m.r. data. The younger coals have tetralin-type nuclei and bituminous coal has 5–6 rings, about 2 of naphthene type. The younger coals were estimated to have more frequent ether linkages.     

Solubilization of Illinois bituminous coal: the critical importance of methylene group cleavage

The extent of solubilization of an Illinois bituminous coal achieved by hydroliquefaction at various temperatures with and without catalyst was determined. The coal was also solubilized by treatment in suspension in THF with potassium in the presence of naphthalene. The original coal was also extracted with THF without prior treatment. After removal of solvent the soluble fractions were examined by ¹H n.m.r. spectroscopy. It is concluded that there are two temperature regimes which affect coal conversion: in the 350–400 °C region, hydroliquefaction proceeds principally by ether cleavage; whereas in the 400–450 °C range, hydroliquefaction results in additional cleavage of methylene groups joining aromatic or heteroaromatic units. Comparison of solubilization by hydroliquefaction with solubilization achieved with potassium in THF seems to indicate the initial degradation of coal by both techniques proceeds by ether cleavage and that further degradation is achieved in both cases by attack at the methylene bridge. The n.m.r. studies on hydroliquefaction products indicate that Illinois bituminous coal is structurally quite homogeneous because the product composition is largely independent of the degree of solubilization. Thus, solubilization results from molecular weight degradation rather than from preferential-degradation to different structural units.     

Heat of wetting of a bituminous coal

Heat of wetting with water has been measured for samples of an Australian low-rank bituminous coal with different initial moisture contents. Although dry coal showed a substantial heat of wetting (≈6 kJ kg^?1), values decreased rapidly with increasing moisture content and it is calculated that for a normal moist coal (65% relative humidity) heat of wetting could cause a temperature rise of only ≈2 °C, which would be unlikely to contribute significantly to self-heating in stored coal.     

Product evolution during rapid pyrolysis of Green River Formation oil shale

The devolatilization of rich (209 cm³ kg⁻¹), rapidly heated (1000 K s⁻¹) Green River Formation, Colorado, oil shale has been studied at temperatures from ≈ 600 to 1100K, under helium (170 kPa) and vacuum. Oil yields decrease with increasing pressure at temperatures above the 800–900 K range. This is attributed to an increased probability of condensed phase cracking reactions involving oil precursors with increased pressure. The fact that escape of the oil from shale particles is vaporization-rate limited is supported by a comparison of the molecular weights of oil collected outside the particle and that left within the particle. The measured nature of particle porosity is consistent with the view that external gas pressure can affect vaporization rates within pores. Elemental and infrared characterization of different molecular weight fractions of oil are presented.     

Quantitative aspects of solid state 13C n.m.r. of coals and related materials

The quantitative aspects of cross-polarization (CP), which is used in conjunction with dipolar decoupling and magic-angle rotation to obtain high resolution ¹³C n.m.r. spectra of coals, have been studied using a bituminous coal (82 wt% C, dmmf basis) and asphaltenes from an extract of the same coal. The condition for obtaining reliable quantitative data, that rotating frame ¹H relaxation times (T_1p′ these govern the extent of CP) are much longer than the time required to polarize the carbons present (≈1 ms), was met for the asphaltenes. In contrast, about half the protons in the coal have T_1p5 of ≈ ? 1 ms, these times being too short to allow CP of all the carbons. Although the aromaticities obtained for this coal were fairly constant (≈0.75) using (CP) contact times > 0.5 ms, the total peak intensity decreased markedly as the contact time was increased and was much less than that for the asphaltenes. These results indicate that not all the carbons in bituminous coals are observed by CP and, as a consequence, aromaticities reported in the literature for some bituminous coals appear to be low.     

Influence of coal preoxidation and the relation between char structure and gasification potential

A study was carried out to ascertain the effects of coal preoxidation and carbonization conditions on the structure and relative gasification potential of a series of bituminous coal chars. Chars were prepared from two freshly mined bituminous coals and preoxidized samples derived from them. Carbonization conditions included a wide range of heating rate (0.2–10000K s^?1), temperature (1073–1273 K) and time (0.25–3600 s). Char properties were characterized in terms of analysis of char morphology, surface area, elemental composition, and gasification reactivity in air. Over the range of conditions used, preoxidation substantially reduced coal fluid behaviour and influenced macroscopic char properties (char morphology). Following slow heating (0.2 K s^?1), preoxidized coals yielded chars having higher total surface areas and higher reactivities toward gasification in air than did similar chars prepared from fresh coal. Following rapid heating (10000 K s^?1) and short residence times (0.25 s), chars prepared from preoxidized and fresh coals exhibited similar microstructural and chemical properties (surface area, $\text{C}\text{H}$ ratios, gasification rates). Carbonization time and temperature were found to be the critical parameters influencing char structure and gasification potential.     

Observation of transient free radicals during coal pyrolysis

Electron spin resonance studies of the early stages of carbonization of a high-volatile A bituminous coal have been used to detect and quantify transient free radicals present in the coal pyrolysate. Free radical (spin) populations are converted from susceptibilities measured at elevated temperatures assuming that the population present after ≈1 h of pyrolysis does not change when the coal pyrolysate is cooled to 20 °C. Co-carbonization of 9–10 dihydroanthracene and anthracene with the coal suggest that disproportionation pathways may be preferred over combination pathways as temperatures of pyrolysis and of liquefaction processes are increased.     

Structure in cokes from coals of different rank

A range of bituminous coals has been carbonized to 1273 K. Polished surfaces of the solid products, carbons or cokes, are examined for optical texture by optical microscopy. Fracture surfaces of the carbons are examined by scanning electron microscopy (SEM). The carbon from the lowest rank coal (NCB Code No. 702) is isotropic and fracture surfaces are featureless. Carbons from coals of ranks 602, 502 are optically isotropic but fracture surfaces are granular (size 0.1–0.2 μm), indicating small growth units of mesophase. In the carbon/coke from a 401 coal, the anisotropic optical texture and grain size are both ≈0.5–10 μm diameter. Coke from a coking coal (301a, 301b) has a layered structure extending in units of at least 20 μm diameter with sub-structures ~ 1.5 μm within the layers, indicating perhaps that the bedding anisotropy of these coals is not totally lost in the fluid phase of carbonization. The carbons from the higher rank coals have the bedding anisotropy of the parent coal. The combined techniques of optical microscopy and SEM (both before and after etching of the fracture surfaces of coke in chromic acid solution) reveal useful detail of structure in carbons/cokes and of the mechanism of carbonization of coking coals.     

Analysis of sub-micron mineral matter in coal via scanning transmission electron microscopy

The fine-scale mineral matter in three US coals has been analysed via scanning transmission electron microscopy (STEM). The samples observed were a North Dakota lignite, a Kentucky bituminous, and a Pennsylvania semi-anthracite. Specific mineral types, differing among the three coals examined, appear to predominate at this fine size scale (particles ? 200 nm in diameter). Fe-rich and Ba-rich minerals in the lignite, a Ti-rich mineral in the bituminous and Ca-rich and Ti-rich minerals in the semianthracite were the predominant species found. The inherent mineral content in the observed organic background also differed from coal to coal. The distributions of mineral species in the size range ? 200 nm reported herein do not reflect the distributions in the larger size ranges obtained by more macroscopic techniques.     

Initial stage (short residence time) coal dissolution

An investigation of the coal dissolution phenomena of bituminous coal during the initial stage is reported. The experimental data obtained from pilot plants and laboratory units are analysed. Coal dissolution reactions can be classified into two distinct stages. In the initial stage of coal dissolution, a first-order reaction scheme is proposed, whereby coal undergoes a fast thermal reaction producing preasphaltene as the predominant product. Conversion for this reaction is calculated, based on the amount of pyridine solubles formed. The rate coefficient for this stage is found to follow an Arrhenius type temperature dependence with an activation energy of 200 ≈ 230 kJ/mol. The second stage reaction, characterized by the formation of asphaltene and coal oil, has been discussed elsewhere⁶.     

Rapid pyrolysis and hydropyrolysis of Canadian coals

A range of Canadian coals were subjected to variable heating-rate conditions in a variety of atmospheres. Heating rate was found to have little effect on total weight loss of the coal, but a dramatic effect on the actual composition of products. High heating rates substantially increased the yield of light hydrocarbons. Operation in ≈100 KPa (1 atm) H₂ at high heating rate resulted in 5% conversion to light hydrocarbon gas and liquid products. Operation in ≈10 MPa (100 atm) H₂ at a heating rate of 600 Ks^?1 gave 10% coal conversion to light liquid products (benzene, xylene, toluene).     

Flash pyrolysis of coals: behaviour of three coals in a 20 kg h−1 fluidized-bed pyrolyser

Flash pyrolysis of Loy Yang brown coal, and Liddell and Millmerran bituminous coals has been studied using a fluidized-bed reactor with a nominal throughput of 20 kg h^?1. The apparatus and its performance are described. The yields of tar and hydrocarbon gases are reported for each coal in relation to pyrolysis temperature, as also are analytical data on the pyrolysis products. The peak tar yields for the dry, ash-free Loy Yang and Millmerran coals were respectively 23% wt/wt (at ≈ 580 °C) and 35% wt/wt (at $\text{\$}\text{?}$600 °C). The tar yield from Liddell coal was 31% wt/wt at ≈ 580 °C. Hydro-carbon gases were produced in notable quantities during flash pyrolysis; e.g. Millmerran coal at 810 °C gave 6% wt/wt (daf) methane, 0.9% wt/wt ethane, 6% wt/wt ethylene, and 2.5% wt/wt propylene. The atomic $\text{H}\text{C}$ ratios and the absolute levels of hydrogen in product tars and chars decreased steadily with increasing pyrolysis temperature.     

Fractionation of Linear Polyethylene with Gel Permeation Chromatography. Part III

Abstract

Many commercial linear polyethylenes have very broad distributions of molecular weight. The high molecular weight fractions often extend beyond the highest molecular weight calibration standard of GPC. For this reason the reliability of information obtainable from GPC has been examined with attention to the average molecular weights. Calibration range is a serious limitation for the accurate determination of the - weight-average and the higher averages of molecular weight. Uncertainty in the baseline at the high molecular weight region, however, does not produce a significant error. With a four-column GPC having 10³ to 10⁷ A nominal capacity, improved resolution is needed in the high molecular weight range. In order to examine the resolution and to improve the calibration, a polyethylene standard of ca. 3–4 million molecular weight is required. With the present limitation of GPC the greatest amount of information can be obtained by examining and intercomparing the cumulative distribution curves. With this representation ca. 95% or more of the cumulative weight range is free from uncertainty in calibration and resolution. A question is raised as to whether melt index is precisely a function of the weight-average molecular weight. This question is pertinent when significantly different molecular weight distributions are involved. GPC offers an opportunity to resolve the question.