The Nature and Possible Significance of Particulate Structure in Alkali

20 The Nature and Possible Significance of Particulate Structure in Alkali-Treated Brown Coal R. J. CAMIER , S. R. SIEMON, H . A. J. BATTAERD, and B. R. STANMORE Downloaded by UNIV LAVAL on July 12, 2016 | http://pubs.acs.org Publication Date: June 1, 1981 | doi: 10.1021/ba-1981-0192.ch020

1

Department of Chemical Engineering, University of Melbourne, Parkville, Vic toria 3052, Australia

An investigation has been made into the fine structure of Victorian brown coal by means of a particle size analysis of alkali-treated coal at pH 13. Examination of the —40 μm + 1 μm fraction revealed periodic maxima in Stokesian diameter, which corresponded to repetitive mass units, and it was found that microscopic rod shapes 1 μm in diameter and 6-8 μm long were present in large quantities. These aggregated into clusters to give the particle size increments. In the smaller-sized fraction —1 μm, no particles were detected until around the 2-nm mark, revealing a huge particle size gap. The sizes found indicate molecules of mass less than 10,000, i.e., humic acids. The mass of detritus and rods, etc., is surrounded by a matrix of humic acid gel. Three possible explanations for the presence of rodlike particles have been considered. Thefirstsuggests that the rods could be remnants of microbial protoplasm. The second considers the rods to be the heavily lignified remains of plant cell wall material. The third postulates a modification of the theory of coal genesis and assumes that lignin has been digested to humic acid types of molecules, which reploymerize by condensation. As the carbon con tent of the polymers rises, phase separation occurs and cylindrical domains form in a way similar to those found in copolymers of petrochemical origin. Present address: State Electricity Commission of Victoria, Herman Research Laboratory, Howard Street, Richmond, Victoria 3121, Australia. 1

0065-2393/81/0192-0311$05.00/0 © 1981 American Chemical Society

Gorbaty and Ouchi; Coal Structure Advances in Chemistry; American Chemical Society: Washington, DC, 1981.

312

COAL

STRUCTURE

I ^he b e h a v i o r of coal d u r i n g p r o c e s s i n g is d e t e r m i n e d by its physicochemical composition and structure. The examination of the coal molecule has been hampered by the inability to find techniques that measure any meaningful properties of such large complex structures. Most attacks on the problem have been by means of breaking down the structure into smaller, more tractable pieces, examining these, and infer ring the original structure. W i t h bituminous coals the severity of the treatment needed to rupture the molecules raises doubts as to the validity of the method. There is uncertainty even with brown coals that are geologically younger and bear more resemblance to the molecules of classical organic chemistry. This uncertainty is reflected, for instance, in the diversity of models proposed for basic molecular arrangement (1-5). Brown coals have the advantage that they can be broken down by the comparatively gentle treatment of alkali digestion (6) into fragments in the micron and submicron range. This breakdown results in a soluble fraction of humic acids and an insoluble residue, humins. With Victor ian coals maximum digestion occurs at p H 13, to give humic acid yields ranging from 15% to 40% of the dry coal mass (7). This paper reports on a study of digested coal fractions that were subjected to particle size analysis using sedimentation techniques. For the humins fractions a gravitational sedimentation technique was adopted, while the more finely divided humic acids required an ultracentrifuge to generate a sufficiently large force field. The nature of the fragments generated by this technique has resulted in a modified hypothesis of coal genesis.

Downloaded by UNIV LAVAL on July 12, 2016 | http://pubs.acs.org Publication Date: June 1, 1981 | doi: 10.1021/ba-1981-0192.ch020

r

Experimental The coal examined was a sample of about 10 kg of medium light, earthy coal from the Yallourn mine in the Latrobe Valley, Victoria. Its ultimate analysis on a dry, mineral matter- and inorganic-free (DMIF) basis is given in Table I. A stock sample was prepared by wet ball milling followed by further size reduction in a domestic food pulper and was then stored under water in a closed vessel. Table I. Particle Size Fractions Fraction Original coal +43-|xm fraction Rod concentrate (-43, 4- 1.2 yon) Humic acids (-1.2 |xm) Fulvic acids

Yield

Carbon Hydrogen Oxygen Nitrogen

Sulfur

(%)

(%)

(%)

(%)

(%)

(%)

— 20.3 46.4

65.6 62.7 65.8

5.18 4.88 5.94

27.9 31.6 27.2

0.75 0.63 0.78

0.4 0.3 0.3

31.0

59.7

3.90

35.3

0.70

0.3

2.2

23.1

8.48

68.5



20.

CAMIER

E T AL.

Particulate Structure in Brown Coal

313

For each test a 15-mL quantity of slurry was mixed with a 500-mL quantity of 0.1M NaOH solution to maintain pH at 13. After digesting overnight, the slurry was wet screened on a B.S. 350-mesh screen to remove +43-|im oversize par ticles. The underflow then was passed through a micropore filter of nominal pore size of 1.2 |xm. The coal thus was fractionated into three particle size ranges. Some of each of the fractions was acid-washed to remove sodium and reprecipitate the humic acids. A yellow supernatant liquid remained after precipita tion of the humic acids, caused by small quantities of fulvic acid. An elemental analysis then was carried out on the dried solids of each fraction. Other samples of alkaline slurry were subjected to particle size analysis by sedimentation. With the —43 u,m + 1.2 u,m fraction this analysis was done in a 50-mm-diameter settling column of dilute slurry with a tared pan at the base to record continuously the mass of sedimented solid. The data were analyzed by the method of Oden (8), and the particle size distribution (Stokesian diameter), expressed on a mass percent basis, was calculated. The humic acid fraction (—1.2 jxm), which was a dark brown suspension containing 4.7 mg/L of coal, did not settle even after standing for 6 months. This slurry was spun in a Beckman ultracentrifuge with special long tubes to generate high g values. Alkali-resistant polyallomer tubes were used so that the solids collected in the base could be removed in a special guillotine, then dried and weighed. The heights of suspension charged varied from 10 to 80 mm, and rotational speeds up to 40,000 rpm were used. The data were analyzed again by Oden's method, modified according to Brown (9). Results Results of the fractionation experiments are summarized in Table I. A typical output from the sedimentation balance for —43 |xm + 1.2 u,m material is shown i n Figure 1. The occurrence of distinct peaks indicates that groups of closely sized particles are present, the smallest being about 6 |xm in effective (Stokesian) diameter. The frequent occur-

Sfoktsian

Particle

Ofamtttr

(jtm)

Figure 1. Particle size distribution (—43 + 1.2 u.m size fraction)


COAL

STRUCTURE


314

Fuel

Figure 2. Photomicrographs of brown coal in pH 13 solution: (a) pair of typical rods; (b) cross section of cell remains; (c) rod peeling from xylitic fragment (14). rence of similar-sized groups in tests with p H values ranging from 7 to 14, and with similar tests on five other Latrobe Valley coals (10), suggested that some fundamental unit was present. A microscopic examination of the material revealed that cylindrical rods about 0.9 |xm in diameter and



20.

CAMIER

ET

AL.

Particulate Structure in Brown Coal

315

6-8 |xm long were common (Figure 2). When the drag coefficient for such particles was calculated from L a m b s formula for cylinders at low Reynold's number [see Prandtl (11)], the terminal setting velocity was the same as for a spherical particle of about 6-|xm diameter. The larger particle sizes thus could be accretions of these basic units, and several such agglomerations were noted. The rods were arranged side by side, closely packed in bundles. Apparently the alkali peels these rods from the coal mass, and they subsequently agglomerate in solution, similar to tactoid formation (12), according to a crystal growth type of clustering theory (13, 14).) Elemental analyses of the fractions showed that the rod fraction (—43 u,m -I- 1.2 |xm) is rich in both hydrogen and carbon when compared with the original coal, Figure 3, although the effect is obscured partly by the slight oxidation that takes place in alkaline solution (10). The higher carbon and hydrogen contents of the rod-concentrate fraction are attri butable mainly to the liptinite material (pollen, spores, waxes, etc.) that predominates in this size range. The particle size distribution for the humic acid fraction is depicted in Figure 4. No material sedimented out until the most extreme con ditions were applied (40,000 rpm for 24 hr), when some lightening of color at the top of the solution was observed. The sedimented particles had a Stokesian diameter of around 2 nm, which means that a particle size gap of three orders of magnitude exists between these and the next largest particles detected (5 |xm). From the experimentally determined coal particle density of 1.43 g/cm , it was calculated that a solid sphere of diameter 2 nm would have a molecular mass of 4000. If the molecules were rod-shaped, even smaller molecular masses would be predicted. Literature values of the molecular mass of regenerated humic acids range between 800 and 20,000, with the values clustering around 1,000 and 10,000 (15, 16, 17). Since the humic acid fraction constitutes 30% of the dry coal mass, about one-third of the coal is in the form of small macromolecules that are bound to the coal structure with bonds weak enough to be disrupted by dilute alkali. It is of interest to note that the particle size gap supplies a rational basis to the traditional German classification scheme of defining humic acid and humins on the basis of a particle size separation (filtration). 3

Discussion The presence of geometrically uniform rods and the absence of particles over such a wide particle size range have implications for our understanding of coal chemistry and genesis. The following discussion attempts to harmonize these observations.


316

COAL

STRUCTURE

80 Carbon


60

JS m

40

>%

c

The Nature and Possible Significance of Particulate Structure in Alkali

Recommend Documents