Energy & Fuels 1989,3, 528-533
528
Characterization of Coal Macerals Using Combined Chemical and NMR Spectroscopic Methods Chol-yoo Choi,*yt John V. Muntean, Arthur R. Thompson,$and Robert E. Botto* Chemistry Division, Argonne National Laboratory, 9700 South Cuss Avenue, Argonne, Illinois 60439 Received January 3, 1989. Revised Manuscript Received May 31, 1989
Resinite, sporinite, vitrinite, and inertinite macerals were separated from two high-volatile bituminous coals obtained from the Argonne Premium Coal Sample Program and one high-volatile bituminous coal from The Pennsylvania State University Coal Sample Bank. A survey of the chemical nature of these macerals was carried out by using chemical techniques in combination with solid 13C NMR spectroscopy. Alkylation with 13C-enrichedmethyl iodide followed by NMR analysis was used to determine the concentrations of acidic OH and CH sites in these materials. The hydroxyl and carboxylic acid contents of these macerals were also determined.
Introduction The combined use of alkylation with 13C-enriched reagents and solid 13CNMR spectroscopy has been shown to be an effective procedure for determining different types of acidic sites in coals.l-' Alkylation using tetrabutylammonium hydroxide as the base occurs predominantly at the acidic oxygen functional groups in coal, such as phenols and carboxylic acids, to produce the corresponding ethers and esters! Certain acidic carbon sites in structures such as fluorene, indene, and benzanthrene are also alkylated under the reaction condition^.^*^ The distinct differences in chemical shifts of methyls on carbon and oxygen allow an estimation of the degree of alkylation occurring at both sites. Furthermore, the 0-methyl region of the lSC spectra can be resolved further into three distinct chemical shift regions facilitating the assignment of methyl carboxylates,unhindered aryl methyl ethers, and hindered aryl methyl ethers. Macerals were separated from three high-volatile bituminous coals by density gradient centrif~gation.~The distributions of various types of hydroxyl and carboxylic acid groups in these macerals as estimated by 13C-enriched methylation and solid 13CNMR spectroscopy are reported. Experimental Section Coal and Maceral Samples. The coals used in this study were obtained from The Pennsylvania State University Coal Sample Bank and the Argonne Premium Coal Sample Program. These were the West Virginia Upper Kittanning seam hvA bituminous coal (PSOC-732), the Utah Blind Canyon seam hvB bituminous coal (APCS-6), and the West Virginia Lewiston-Stockton seam hvA bituminous coal (APCS-7). The Upper Kittanning coal was ground to less than 3 fim by using a planetary ball mill and a fluid energy mill. The Upper Kittanning and Lewiston-Stockton coals were demineralized with hydrochloric and hydrofluoric acids. Maceral groups from the Utah Blind Canyon and the West Virginia Lewiston-Stockton coals were liberated from each other by using a base-catalyzed maceral separation method developed previously.1° The coal particles were then separated into density fractions by using the density gradient centrifugation procedure as described by Dyrkacz and co-worker~.~ The macerals were extracted with benzenemethanol (3:l v/v) at room temperature. Proton NMR of the
* To whom correspondence should be addressed. Present address: Rohm and Haas Company, Spring House, PA 19477.
*
Present address: United States Department of Agriculture, Peoria, IL 61604. 0887-0624/89/2503-0528$01.50/0
extracts showed that most resonances could be attributed to Brij-35, the surfactant used in the density gradient centrifugation procedure. Between 1and 4 wt % of the macerals were removed by the benzene-methanol extraction, with the exception of the Blind Canyon resinite, which was almost completely soluble in benzene-methanol. Thus, this resinite was extracted with methanol alone (which did not dissolve the resinite) in order to exhaustively remove the Brij-35. The elemental and maceral analyses of the density-separated macerals are presented in Table I. The coals and macerals were kept under inert atmosphere whenever possible to minimize air oxidation. Alkylation. The macerals were alkylated by using Liotta's procedure: as described in previous reports.6J1 In a typical experiment, between 40 and 50 mg of the maceral concentrate was alkylated with methyl-13C iodide (98% 13C;Cambridge Isotopes). The samples, which were particularly susceptible to oxidation during the alkylation procedure, were handled under argon atmosphere. The methylated macerals were extensively washed with 50% aqueous methanol to remove the tetrabutylammonium salt. The elemental analyses of the methylated macerals are presented in Table 11. The number of added methyl groups was obtained from the carbon and hydrogen elemental analyses of the original and alkylated macerals by using the procedure described in a previous report.6 Due to the difficulty in determining small quantities of added methyls for the highly aliphatic Utah resinite from elemental data, an alternative method using NMR was adopted. The difference in the carbon aromaticity (f,) between the methylated and original resinite, taking into account the amount of I3C enrichment, was used to estimate the number of added methyls. The reliability of this method depends on the quantity of carbon spins observable by NMR.12 The percentage of observable carbons is high for highly aliphatic macerals such as resinite and (1)Liotta, R.;Brons, G. J. Am. Chem. SOC.1981,103, 1735-1742. (2)Hagaman, E. W.; Woody, M. C. Fuel 1982,61,53-57. (3)Chambers, R. R.,Jr.; Hagaman, E. W.; Woody, M. C.; Smith, K. E.; McKamey, D. R. Fuel 1986,64,1349-1354. (4) Hagaman, E. W.; Chambers, R. R., Jr.; Woody, M. C. Anal. Chem. 1986,58,387-394. (5)Botto, R. E.; Choi, C. Y.; Muntean, J. V.; Stock, L. M. Energy Fuels 1987,I , 270-273. (6)Hagaman, E. W.; Chambers, R. R., Jr.; Woody, M. C. Energy Ruek 1987,1, 352-360. (7) Chambers, R. R., Jr.; Hagaman, E. W.; Woody, M. C. In Polynuclear Aromatic Compounds; Ebert, L. B., Ed.; Advances in Chemistry Series 217;American Chemical Society: Washington, DC, 1988. (8)Liotta, R.Fuel 1979,58,724-728. (9)Dyrkacz, G.R.;Horwitz, E. P. Fuel 1982,61, 3-12. (IO) Choi, C. Y.; Dyrkacz, G . R. Manuscript in preparation. (11)Choi, C. Y.;Dyrkacz, G . R.; Stock, L. M. Energy Fuels 1987,1, 280-286. (12)Botto, R. E.; Wilson, R.; Winans, R. E. Energy Fuels 1987,1, 173-181.
0 1989 American Chemical Society
Energy & Fuels, Vol. 3, No. 4, 1989 529
Characterization of Coal Macerals
maceral whole coal sporinite vitrinite inertinite
Table I. Elemental and Maceral Analysis of Density Separated Macerals elem anal., wt % maceral anal., vol % d, g C H N S ash liptinite vitrinite inertinite West Virginia Upper Kittanning (PSOC-732) 1.172-1.213 1.287-1.317 1.353-1.396
82.4 86.4 82.6 82.4
4.6 6.4 4.9 3.9
1.065-1.084 1.177-1.195 1.263-1.281
74.0 83.9 78.1 74.2
5.6 9.9 7.0 5.7
1.155-1.193 1.265-1.290 1.339-1.378
80.9 81.3 80.2 82.3
1.4 1.2 1.6 1.3
0.7 e e e
0.5 e e e
12 95" 2 0
50 5 97 3
38 0 1 97
5 96b 83e 2
87 4 17 92
8 0 0 6
12 93d 0
73 7 95 5
15 0 5 95
Utah Blind Canyon (APCS-6) whole coal resinite sporinite vitrinite
1.4 0.5 1.2 1.4
0.6 e e e
4.7 e
e e
West Virginia Lewiston-Stockton (APCS-7) whole coal sporinite vitrinite inertinite
5.1 6.6 4.7 3.4
1.5 1.1 1.5 1.1
0.7
e e e
19.8 e e e
0
"Resinite 1%; sporinite 94%. *Resinite 92%; sporinite 4%. cResinite 9%; sporinite 74%. dResinite 1%; sporinite 92%. ONot determined. Table 11. Elemental Analyses of the Methylated Macerals elem anal., wt % maceral C H N West Virginia Upper Kittanning (PSOC-732) whole coal 78.7 4.8 1.4 sporinite 78.5 6.2 1.1 vitrinite 76.8 4.9 1.5 inertinite 79.5 3.9 1.2 Utah Blind Canyon (APCS-6) whole coal 71.4 5.5 resinite 80.0 9.4 sporinite 72.9 6.6 vitrinite 70.5 5.6
1.3 0.5 1.2 1.3
West Virginia Lewiston-Stockton (APCS-7) whole coal 73.8 4.8 1.3 sporinite 72.8 6.2 1.0 vitrinite 74.7 4.6 1.3 inertinite 77.9 3.5 1.0 some sporinites.12 Values obtained from the NMR and elemental data for sporinites were found to be similar to within 5%. NMR Analysis. The solid 13C NMR spectra were recorded a t 2.35 T (25.18 MHz for 13C) on a Bruker Instruments Model CXP-100 spectrometer in the pulse Fourier transform mode with quadrature phase detection. About 50 mg of sample was loosely packed into a 300-pL ceramic sample spinner and spun a t approximately 4 kHz. The operating conditions used in CP experiments included a spectral width of 10 kHz,a 90° proton pulse width a 4.5 ps (BO-kHz proton-decoupling field), an acquisition time of 50 ms, a 2-s pulse repetition rate, and a total accumulation of 4K to 30K transients. Contact time studies carried out on the entire suit of macerals were used to determine the optimum CP conditions. It was shown that using a 2-ms contact time allowed good estimates of the distribution of methoxyl groups to be obtained. In a typical experiment, 400 words of memory were allocated for data acquisition and then increased to 4K (2K real data) by zero-filling. Chemical shifts are reported in ppm with respect to tetramethylsilane (TMS), with tetrakis(trimethy1sily1)silane (TKS)13 used as a secondary reference. The distribution of added methyl groups was determined from relative areas that were calculated from line simulation of experimental spectra. The Bruker Pascal line simulation program LINESIM (version 880101) was used to calculate simulated spectra. The simulation provided a best fit of peak positions, intensities, line widths, and calculated areas from regions of overlapping resonances in the methoxyl and methyl carbon regions of .experimental spectra. Typically, a best fit was obtained by simulating spectra with resonances centered at 51.3,55.1, and 60.9 ppm, (13) Muntean, J. V.; Stock, L. M.; Botto, R. E.J.Magn. Reson. 1988, 76, 540-642.
Table 111. Elemental Data of the Macerals Expressed as Molar Ratios maceral H/C N/C O/C" West Virginia Upper Kittanning (PSOC-732) whole coal 0.67 0.015 0.11 sporinite 0.88 0.012 0.05 vitrinite 0.71 0.017 0.10 inertinite 0.56 0.014 0.11 Utah Blind Canyon (APCS-6) whole coal 0.90 0.016 resinite 1.41 0.005 sporinite 1.07 0.013 vitrinite 0.92 0.016
0.19 0.05 0.13 0.19
West Virginia Lewiston-Stockton (APCS-7) whole coal 0.75 0.016 0.12 sporinite 0.97 0.012 0.10 vitrinite 0.07 0.016 0.13 inertinite 0.49 0.011 0.12
" The oxygen content was determined by difference from the C, H, N microanalyses. having line widths on the order of 80-160 Hz, and having a 30% Lorentzian/70% Gaussian line shape. The best fit was achieved by using the "simplex" fitting routine. The process terminated when the "residual s u m of squares" value showed no improvement over 20 iterations.
Results and Discussion The elemental data for the coals and macerals, which can conveniently be discussed in terms of mole ratios, are shown in Table 111. The H/C values for the macerals follow the order sporinite > vitrinite > inertinite for the West Virginia Upper Kittanning and the West Virginia Lewiston-Stockton macerals and resinite > sporinite > vitrinite for the Utah Blind Canyon macerals. Nitrogen is more concentrated in the vitrinites than in the other macerals. The nitrogen content of the Utah Blind Canyon resinite is substantially lower than the values for other macerals from this coal, and this finding is in accord with previous reports for other Utah resinite~.'~J~ The O/C values follow the order: vitrinite = inertinite > sporinite for all macerals. It should be noted that the oxygen con(14) Winans, R. E.;Hayatsu, R.; Scott, R. G.; McBeth, R. L. In Chemistry and Characterization of Coal Macerals; Winans, R. E., Crelling, J. C., Eds.;ACS Symposium Series 252; Americal Chemical Society: Washington, DC, 1984. (15) Bodily, D. M.; Kopp, V. Prepr. Pap-Am. Chem. SOC.,Diu. Fuel Chem. 1987,32(I),554-547.
530 Energy & Fuels, Vol. 3, No. 4,1989
Choi et al.
A A
A 200
150
100
50
0
-50
Chemical Shift, p p m Figure 1. Solid I3C CP/MAS spectra of West Virginia Upper Kittanning coal (PSOC 732) from top to bottom: liptinite; vitrinite; inertinite; whole coal. tent was calculated by difference from the C, H, N microanalyses. Due to the limited availability of the maceral samples, the sulfur and mineral-matter contents of these samples were not measured. Thus the oxygen contents, which were used to calculate the O/C ratios shown in Table 111, do not take into consideration sulfur or mineral-matter content, even though differences in the amount of sulfur have been reported for different types within a coa1.1618 Solid 13C-CP/MAS spectra of the macerals and whole coals are shown in Figures 1-3. The NMR spectra are scaled to the signal of largest intensity. In general, NMR spectra of the coals and macerals show two broad resonance bands. The maximum signal intensity in the aromatic region occurs at 128 ppm, which is characteristic of unsubstituted aromatic carbon atoms. The maximum signal intensity in the aliphatic region is centered around 30 ppm, which is characteristic of methylene carbons in alkyl chains. A shoulder in the aromatic region around 140 ppm, which is particularly pronounced in the spectra of the sporinites but also is discernible for the other macerals, can be attributed to alkyl-substituted aromatic carbons.lS Less intense shoulders appearing at 150-160 ppm and at 105-120 ppm are compatible with unsaturated carbons that are a and /3 to oxygen substituents in structures such as phenols and benzofurans.20.21 Differences seen in the relative intensities of these shoulders reflect the variations in aromatic substitution among the maceral groups. (16) Raymond, R., Jr. Proc.--lnt. Koohlenwiss. Tag. 1981, 857-862. (17) Dpkacz, G. R.; Bloomquist, C. A. A,; Ruscic, L. Fuel 1984, 63,
- -- - -
1 AR7-1.172. . .-.
'
l " " l " " 1 " " l "
I " " " ' " " ' ' I " ' ' " ' ' l ' ' " i
250
(18) Tseng, B. H.; Buckentin, M.; Hsieh, K. C.; Wert, C. A.; Dyrkacz, G. R. Fuel 1986,65,386-389. (19) Maciel, G. E.; Sullivan, M. J.; Petrakis, L.; Grandy, D. W. Fuel 1982, 61, 411-414. (20) Zilm, K. W.; Pugmire, R. J.; Larter, S. R.; Allan, J.; Grant, D. M. Fuel 1981,60,717-720. (21) Pugmire, R. J.; Zilm, K. W.; Woolfenden, W. R.; Grant, D. M.; Dpkacz, G. R.; Bloomquist, C. A. A.; Horwitz, E. P. Org. Geochem. 1982, 4, 79-82.
250
200
150
100
I
" " I " '
50
0
-50
Chemlcal Shlft, p p m CP/MAS spectra of Utah Blind Canyon coal (APCS-6)from top to bottom: resinite; sporinite; vitrinite,whole
Figure 2. Solid coal.
I \
A
A A r " " i " " 250 200
I
~
150
100
~
50
~
0
~
I
-50
Chemical Shift, p p m Figure 3. Solid 13C CP/MAS spectra of West Virginia Lewiston-Stocktoncoal (APCS-7) from top to bottom: liptinite; vitrinite; inertinite; whole coal. The aliphatic region for the sporinites displays a maximum signal intensity near 30 ppm with a less intense shoulder appearing at approximately 20 ppm. The aliphatic resonances of the vitrinites and inertinites are also centered near 30 and 20 ppm, although the former resonance is much less pronounced in the spectra of the two West Virginia coals. These resonances are compatible with methylene groups in aliphatic chains and pendant methyl groups on aromatic structures, respectively. Notable differences can be seen in the spectra of resinites and sporinites separated from the Utah Blind Canyon
~
~
Energy & Fuels, Vol. 3, No. 4, 1989 531
Characterization of Coal Macerals
Table V. Distribution of Added Methyl Groups As Estimated by Solid NMR Spectra methyl groups/100 C maceral tot.n on oxygen on carbon West Virginia Upper Kittanning (PSOC-732) whole coal 4.8 3.8 1.0 sporinite 6.0 5.0 1.0 vitrinite 5.2 4.4 0.8 inertinite 1.6 1.3 0.3
Table IV. Carbon Aromaticities Estimated by Cross-Polarization NMR maceral f." maceral f." West Virginia Upper Kittanning (PSOC-732) whole coal 0.78 vitrinite 0.79 sporinite 0.59 inertinite 0.87 whole coal resinite
Utah Blind Canyon (APCS-6) 0.60 sporinite 0.16 vitrinite
0.46 0.63
West Virginia Lewiston-Stockton (APCS-7) whole coal 0.74 vitrinite 0.76 sporinite 0.54 inertinite 0.87
whole coal resinite sporinite vitrinite
CP experiments: 2-msmix time.
0.6-
c
E"
U
0.1
0.3
Values obtained from elemental analysis data; estimated error 5% (see ref 5). *From NMR data.
% +I
0.1
vitrinites > sporinites > resinites. Changes in carbon aromaticity with structural properties are revealed when fa values are plotted against the corresponding H/C ratios for macerals and coals, as shown in Figure 4. A regular trend is seen with the correlation similar to that observed previously for maceral aromaticity (22) Mukhopadhyay, P. K.; Gormly, J. R. Org. Geochem. 1984, 6, 439-464. (23) Reference deleted in proof.
The Utah Blind Canyon vitrinite has a fairly low fa value (0.63) for a hvB bituminous coal. Because the sample was extracted with organic solvent prior to the NMR analysis, it is unlikely that any contamination from resinite is responsible for this apparently anomalous result. Rather, the fa if an indication of the unique geological history of the Utah coal, which is not revealed in its rank classification designated by the calorific value. Macerals from the three coals were alkylated with 13Cenriched methyl iodide (98% 13C) by using tetrabutylammonium hydroxide as the basic catalyst in tetrahydrofuran following Liotta's procedure? This widely used alkylation procedure derivatizes acidic hydroxylic and carboxylic acid groups in coal to their corresponding alkyl ethers and alkyl esters. The degree of alkylation estimated from the elemental data and from NMR are reported in Table V as the number of methyl groups added to each 100 carbon atoms of maceral or coal. Alkylation also occurs on acidic carbon centers in derivatives of fluorene, indene, phenalene, and benzanthrene under these reaction condition^.^ Methylation on carbon occurs to the extent of 10-2590 percent of the total in macerals from the West Virginia Upper Kittanning and Lewiston-Stockton coals. Methylation on carbon occurs to a lesser degree in macerals from the Utah Blind Canyon coal; this trend is to be expected for a less mature coal. The observation that less methyl groups were added to the Utah and Lewiston-Stockton coals rather than their individual macerals may be due to the larger particle size of the whole coals relative to the macerals. Hence, reagent accessibilitymay be an important factor in these alkylation reactions. However, the relative 0-vs C-methylation ratio for the whole coal is similar to that of its corresponding vitrinite, which suggests that the majority of alkylatable oxygen and carbon sites must be somewhat randomly dispersed within the coal matrix. CP/MAS spectra of the Utah Blind Canyon macerals alkylated with 13C-enrichedmethyl iodide are shown in Figure 5. The resonances of the 0-methylated products appear within three distinct chemical shift regions that reflect differences in the chemical environment around the methoxy groups. The relative areas of methyl signals (24)
Dyrkacz, G. R.; Bloomquist, C. A. A.; Ruscic, L. Fuel 1984, 63,
1166-1173.
Choi et al.
532 Energy & Fuels, Vol. 3, No. 4, 1989
I‘
a
n
0
0
0.2
o
. 0.4
0.6
o 0.8
t 1
1.2
, 1.4
1.6
H/C Ratio Figure 6. Variation of the fraction of hindered methoxy groups with H/C ratio.
, 250
I
200
, , ,
,
, 150
I
,
,
,
7 -
100
50
0
-50
Chemical S h i f t , ppm Figure 5. Solid 13C C P / W spectra of Utah Blind Canyon coal (APCS-6)methylated with %-enriched methyl iodide from top to bottom: resinite; sporinite; vitrinite; whole coal. Table VI. Relative Quantities of Methyl Ethers and Methyl Esters quantities for carbon chem shift of maceral 60.9 ppm 55.1 ppm 51.3 ppm West Virginia Upper Kittanning (PSOC-732) whole coal 1.1 2.6 0.1 sporinite 2.2 1.8 1.0 vitrinite 1.0 3.4 0 inertinite 0.4 0.9 0 Utah Blind Canyon (APCSB) whole coal resinite sporinite vitrinite
0.9
0.5
0.4
0.4
0.2 0.8 0.8
0.2 0.6
1.0 1.4
0.3
West Virginia Lewiston-Stockton (APCS-7) 0.5 0.7 0.1
whole coal sporinite vitrinite
inertinite
1.6 1.0 0.7
1.3
1.5 1.3
0.3 0.3 0
within the three regions (calculated from line simulation of experimental spectra) are summarized in Table VI for the three coals and their isolated macerals. The most plausible structural elements in the methylated products giving rise to resonances centered at 51.3 ppm are methyl carboxylates, and those resonances centered at 55.1 and 60.9 ppm can be assigned to aryl methyl ethers.25 In general, methyl carboxylates represent a major class of 0-methylated products of liptinites from all three coals, while proportionately fewer methyl carboxylates are found in vitrinites or inertinites. The exception is the high carboxylic acid content of the Utah vitrinites; this high level may reflect its unique geological history. The higher concentrations of carboxylic acids in liptinites relative to vitrinites have been noted by others using infrared spectroscopy26imbut comparative estimates of the carboxylic (25) Stock, L. M.; Willis, R. S.; J. Org. Chem. 1985, 50, 3566-3573. (26) Bent, R.; Brown, J. K. Fuel 1961, 40, 47-56. (27) Allan, J. Ph.D. Dissertation, University of Newcastle upon Tyne, Great Britain, 1975.
acid concentrations have not been reported previously for macerals. The hydroxyl content can be expressed as the sum of the resonance areas for aryl methyl ethers at 55.1 and 60.9 ppm. Given and his c o - w ~ r k e r sacetylated ~ ~ * ~ macerals isolated fro-m a set of British coals and reported that liptinites have lower hydroxyl contents than vitrinites and that inertinites have considerably lower hydroxyl contents than those of the corresponding vitrinites or l i p t i n i t e ~ . ~ ~ ~ ~ Our results show that liptinites and, in particular, the Utah resinite have lower hydroxyl contents than vitrinites from the West Virginia Upper Kittanning and Utah coals. However, hydroxyl contents of liptinites and vitrinites from the Lewiston-Stockton coal are comparable. The two inertinites are found to have the lowest hydroxyl contents. The two resonances centered at 55.1 and 60.9 ppm may be assigned respectively to unhindered aryl methyl ethers and hindered aryl methyl ethers.31 Their relative intensities vary among the maceral types, with the ratio of hindered to unhindered aryl methyl ethers following the order liptinites > vitrinites > inertinites. The presence of a larger proportion of hindered phenolic sites in liptinites is consistent with their having a greater abundance of aliphatic structures and smaller quantities of aromatic rings. A higher occurrence of aromatic ring substitutions should lead to a higher incidence of hindered phenolic sites. Conversely, the low abundance of aliphatic structures in inertinites results in there being less substitution on aromatic rings and hence less hindered phenols in inertinites. The preceding structural implications become more evident when the fraction of hindered methoxy groups (of the total methoxyls added) is plotted against H/C ratios. As shown in Figure 6, a clear trend is found for the entire set of coals and maceral samples despite the diverse and varied chemical nature of the different macerals that are represented. It supports the notion that the macromolecular network of coal evolves in a truely nonrandom fashion, whereby the relative amounta of the various structural subunits that comprise a coal or its individual macerals are related statistically to the frequency of in(28) Dyrkacz, G. R.; Bloomquist, C A. A.; Solomon, P. R. Fuel 1984, 63,536-542. d (29) Given, P. H.; Peover, M. E.; Wyw,W.F.Fuel 1960,39,323-340. (30) Given, P. H.; Peover, M. E.; Wyss,W.F.Fuel 19$6,44,425-435. (31) Hindered structures as defined here refer to aryl methyl ethers in which both ortho hydrogens on the aryl ring have been substituted with
carbon functional groups. Steric interactions resulting from substitution at both ortho positions are known to cause a downfield shift of the methoxyl carbon resonance; see ref 5 and references therein.
Energy & Fuels 1989, 3, 533-535
terunit linkages connecting them, in this case the degree of aromatic substitution. Acknowledgment. We gratefully acknowledge the support of this work by the Office of Basic Energy Sci-
N
533
ences, Division of Chemical Sciences, US.Department of Energy, under Contract No. W-31-109-ENG-38. The ele m e n d microanalyses were performed by Steve Newnam of the Analytical Chemistry Laboratory at Argonne National Laboratory.
. *
Communtcattons Aliphatic Structural Elements of a Pocahontas No. 3 Coal
Sir: During the past few years, we have used ruthenium tetraoxide oxidation reactions to convert coal macromolecules to a wide array of aliphatic carboxylic and benzenecarboxylic acids'" to gather qualitative and quantitative information about their aliphatic and aromatic structural elements. We recently obtained some very novel new information about the aliphatic carbon atom distribution in Pocahontas No. 3 coal with this oxidation method. A premium sample of Pocahontas No. 3 coal was obtained from the Argonne National Laboratory. This lowvolatile material contains 4.77% ash, and its elemental composition is C100H62,9N1.15~o.1,03.26. Each sample was prepared for oxidation by extraction using aqueous hydrochloric acid, aqueous sodium hydroxide, benzenemethanol, and chloroform as described previously.6 The oxidation was carried out by using coal (1g), ruthenium(111) trichloride trihydrate (27 mg), and sodium periodate (21.4 g) in a mixture of carbon tetrachloride (20 mL), acetonitrile (20 mL), and water (30 mL). The black organic phase became yellow-brown after 2 h. The reaction was continued for 24 h at room temperature before the products were collected and analyzed. Duplicate analyses by the method that we have previously used3 indicated that the oxidation reaction produced 25.7 f 0.7 mol of carbon dioxide/100 mol of C in the coal. The abundances of the volatile carboxylic acids were determined in one set of experiments by using isotope dilution procedure^,^^^ and the less volatile acids were methylated prior to analysis in another set of experiments by gas chromatography-mass spectroscopy as described previously3 (Figure 1). In the first series of duplicate experiments, we observed that the ruthenium tetraoxide oxidation of this coal produced 7.28 f 0.04 mol of ethanoic acid/100 mol of C. This yield is far greater than the yields obtained in the oxidation of the other coals: lignite (Rockdale, TX), 1.1; subbituminous coal (Rawhide, WY), 1.2; bituminous coal (Pittsburgh No. 8), 3.4.2 The Pocohontas No. 3 coal also gives about 0.25 mol of propanoic acid/100 mol of C and about 0.05 mol of butanoic acid/100 mol of C. The yields of these three and four carbon atom materials are similar to the yields of these acids obtained with the other coals.2 Thus, the methyl group concentration in Pocahontas No. 3 coal is unusually high. We have previously inferred that eth-
14
-
7
.
20
30
40
Time (min.)
Figure 1. GC/MS chromatography of the oxidation products of Pocahontas No. 3 coal: (1)1,2-benzenedicarboxylic acid; (2) 3-methylbenzene-1,2-dicarboxylicacid; (3) 4-methylbenzene1,2-dicarboxylicacid; (4)dimethylbenzenedicarboxylic acid; (5) dimethylbenzenedicarboxylic acid; (6) dimethylbenzenedicarboxylic acid; (7) 1,2,3-benzenetricarboxylicacid; (8) 1,2,4benzenetricarboxylic acid; (9) methylbenzenetricarboxylic acid; (10) methylbenzenetricarboxylic acid; (11)dimethylbenzenetricarboxylic acid; (12) trimethylbenzenetricarboxylic acid; (13) 1,2,4,5-benzenetetracarboxylicacid; (14) 1,2,3,4-benzenetetracarboxylic acid; (15) 1,2,3,5-benzenetetracarboxylicacid; (16) methylbenzenetetracaboxylic acid + 2,3,2'-biphenyltricarboxylic acid; (17) methylbenzenetetracarboxylic acid; (18) biphenyltetracarboxylic acid; (19) 2,6,2',6'-biphenyltetracarboxylic acid; (20) benzenepentacarboxylic acid.
anoic acid is derived primarily from the oxidation of arylmethanes.2 Additional work was directed toward the identification of the benzenecarboxylic acids that were formed in the reaction. They were converted to their methyl esters for convenient analysis. The observations, Figure 1, revealed that a significant amount of 1,2-benzenedicarboxylicacid was produced with traces of 1,3- and 1,4-benzenedicarboxylic acids. This observation established the predominance of the well-recognized 1,Bfusion pattern in this coal. It was also evident that methyl- and dimethyl-
0887-0624/89/2503-0533$01.50/00 1989 American Chemical Society
