Energy & Fuels 1988,2, 118-121
11s
Unfortunately the O(1s)peaks corresponding to carbonyl, ether, and the double-bonded oxygen in carboxylate, as well as hydroxyl oxygen, occur a t about the same binding energy. The presence of carbonyl, ether, or hydroxyls would contribute to the O(1s) difference peak at 531.5 eV. We would anticipate asymmetry to higher binding energies due to the presence of smaller amounts of the singlebonded carboxylate oxygen. IR photothermal beam deflection spectroscopy has been used to study the oxidation of high-temperature chars by 02.41 Evidence of ether-like species appeared as a band centered near 1300 cm-'. The presence of a band attributable to carbonyl species was absent. C(1s) deconvolution schemes based on the simple additivity relationship have been successfully used in the interpretation of oxidized (41) Morterra, C.; Low, M. J. D. Langmuir 1986,1,320 and references therein.
carbon s u r f a c e ~ . ~ ~ - ~ ~In J ' -the ~ ' study of carbon fiber surfaces Takahagi and Ishitmi's use a 2.4-eV shifted peak for carbonyl groups. Proctor and Sherwoodlguse a 3.0-eV shifted peak. We have measured the XPS spectrum of anthraquinone and find a distinct carbonyl peak at 2.6 eV from the main C(1s) line. Following O2and COPoxidation up to moderate coverages, we find little intensity in the C(1s) difference curves in the region corresponding to carbonyls. These C(1s) difference curves correspond to surfaces that subsequently give CO a t high temperature. The high-temperature CO formation observd in TPD is associated with the decomposition of ether species. Carbonyl oxygen thus represent a small minority of the oxygen surface species present after dissociative O2 adsorption, gasification of carbon by C 0 2 ,and combustion of glassy carbon by 02. wstv NO.c , 7440-44-0; 02,7~2-44-7; c o 2 , 124-389;HNO~, 7697-37-2.
Secondary Ion Mass Spectrometry and X-ray Photoelectron Spectroscopy of Derivatized Coal Surfacest R. R. Martin,* N. S. McIntyre,t J. A. MacPhee,* and K. T. Aye Department of Chemistry and Surface Science Western, The University of Western Ontario, London, Ontario, Canada N6A 5B7, and Energy Research Laboratories, CANMET, Energy, Mines & Resources, Ottawa, Ontario, Canada K l A OG1 Received July 7, 1987. Revised Manuscript Received August 19, 1987
Secondary ion mass spectrometry (SIMS) and X-ray photoelectron spectroscopy (XPS) have been used to study the low-temperature oxidation of coal. I s 0 has been used as an isotopic tracer for the oxygen distribution on the coal surface. Several chemical derivatizations have been observed on the oxidized coal surface, and the reactivity of specific regions has been monitored.
Introduction The low-temperature oxidation of coal leads both to loas of coking ability' and autoignition2 of coal stock piles. In this study we report the use of secondary ion mass spectrometry (SIMS) and X-ray photoelectron spectroscopy (XPS) in conjunction with I s 0 2 oxidation to examine the oxygen distribution on a coal surface in relation to specific other elements. The use of both XPS and SIMS in coal studies has been described el~ewhere.~ Experimental Section T h e coal used in this study was supplied by the Canadian Center for Mineral and Energy Technology (CANMET). The sample used was a piece of pyrite-rich fusinite, 26.9% ash, having an elemental composition C = 52.2%, H = 1.6%, N = 0.4%, S = 19.9%) and 0 = 25.9% by difference, which had received no special treatment (such as storage under N2). Fresh surfaces were prepared for SIMS analysis in the following way: the coal was cut with a diamond saw and polished on a silk wheel with diamond *To whom correspondence should be addressed at the Department of Chemistry, The University of Western Ontario. Presented at the Symposium on the Surface Chemistry of Coals, 193rd National Meeting of the American Chemical Society, Denver, CO, April 5-10, 1987. *Surface Science Western, The University of Western Ontario. 8 Energy, Mines & Resources.
0887-0624I88 12502-0118%01.50,I O I
,
grit used as abrasive and water used as lubricant. The resulting coal surfaces were exposed to 0.1atm of le02 for 10 days at room temperature and then subjected to SIMS analysis. The SIMS measurements were made on a Cameca IMS 3f ion microscope using a Cs' primary beam (100i!a t 7.5 keV net energy) rastered over a 400 X 400 pm area. This leads to an ablation rate of approximately 1.25A/s. Ion images were collected from a central area of 170 pm in diameter; these images were obtained at a maas resolution as high as one part in 3500, although for certain ions the secondary ion yield was insufficient to permit imaging. In most cases where SIMS images were taken mass spectral scans were also obtained. In all cases the time required to obtain the images is recorded. Unfortunately there is no simple relation between concentration and image intensity. Fresh surfaces for X P S were obtained in a two-step process. Small pieces of coal were removed from the interior of a large block with a diamond saw, these small pieces were cleaved with a razor blade, and the resulting surface was immediately analyzed. XPS studies were also carried out on regions adjacent to those studied by SIMS on a Surface Science Laboratory Model SSX-100 spectrometer with monochromatized Al Ka radiation with an oval spot size with an approximately 150-pm major axis. The l80was used as an isotopic tracer to distinguish between oxygen added to the coal as a result of this procedure and that already present? (1) Larsen, J. W.; Lee, D.; Schmidt, T.; Grint, A. Fuel 1986,65,595. ( 2 ) Jones, R. E.; Townsend, D. T. A. Nature (London) 1946,155,424. ( 3 ) McIntyre, N. S.; Martin, R. R.; Chauvin, W. J.; Winder, C. G.; Brown, J. R.; MacPhee, J. A. Fuel 1985, 64, 1705.
0 1988 American Chemical Society
Energy & Fuels, Vol. 2, No. 2, 1988 119
Derivatized Coal Surfaces
3 I
cn
I
51
.+
C
s I 1000 A
8 B C
Figure 1. SIMS images, 170-pm diameter, from fresh coal surface exposed to 0.1 atm of lSO2 at room temperature for 10 days.
Binding Energy (eV) XPS Broadscan
Binding Energy (eV) XPS Fe 2p Region
0
698
I
I
The pyrite-rich coal was selected because it provides remarkably well-defined secondary ion images and pyrite is reported to play an important role in the self-heating of coal.5
Results and Discussion The assignment of ionic species to the secondary ion images obtained during high mass resolution imaging was achieved by using two criteria. First the mass spectrum a t a given nominal mass was analyzed to achieve a fit between the mass separation observed and that expected from the species assigned to the spectrum. At mass 28 two peaks were observed separated by 0.019 f 0.003 Da; only %Siand CO are likely to produce this pattern. In addition successive ion images may be used to eliminate some assignments. For example, silicon molecular ions would not be expected in regions where silicon could not be detected. Finally, in one case at mass 30-high mass resolution was not used to distinguish between 3oSi-and Cl80- since the yield of Cl80- was too low for effective imaging. The ion images displayed parts A-E of Figure 1 are assigned to ions C2-, C4-, CO-, S-, and CS-,respectively. The images for C2- and C4- define the organic region of the coal surface while that of S- outlines the pyrite region. The image for CO- is confined entirely to the organic region and originates from a set of sources having a very small area. These images represent oxidized areas on the coal surface, and although the oxygen functional groups cannot be identified with certainty, they may be aldehydes or phenols. The CS- image is of particular interest since it is confined to the organic region of the coal. It has been proposed elsewhere that pyrite oxidizes readily to form a series of hydrated iron sulfates and sulfuric acid.6 Since these (4) Martin, R. R.;McIntyre, N. S.; MacPhee, J. A. ProceedingsInternational Conference on Coal Science, 1985; Pergamnon: New York, 1985; p 796. (5) Herman, R.G.;Simmons, G. W.; Cole, D. A.; Kunicz, V.; Klier, K. Fuel 1984, 63, 673. (6) Huggins, F.E.;Huffman, G. P.; Lin, M. C. Int. J. Coal Geol. 1983, 3, 157.
b
XPS S 2p Region
286 D
Binding Energy (eV)
27
XPS C I s Region
Figure 2. XPS spectra of freshly cleaved coal surfaces.
reactions are exothermic7the surrounding organic material would be exposed to sulfuric acid and simultaneously heated. Under these conditions sulfonation reactions would be expected. In fact sulfonation of coal has been used as a technique for preparing inexpensive ion-exchange m a t e ~ i a l . ~ .Accordingly ~ we suggest that oxidation of (7) Robie, R. A.; Hemingway, B. S.;Fisher, J. R. US.Geol. Sum. Bull. 1978, No. 1452,456.
(8) Holmes, E. L. U.S.Patent 2393249, 1946. (9) Caio, F. A.; Sobrinbo, P. A. Rev. Bras. Technol. 1979, 10, 129. (10) Binder, H. Z. Naturforsch., B: Anorg. Chem., Org. Chem. 1973, 28B, 256.
120 Energy & Fuels, Vol. 2, No. 2, 1988
Martin et al.
Figure 4. Oxygen-related SIMS images compared to siliconrelated SIMS images. See Figure 1 for conditions.
Figure 3. Oxygen-related SIMS images. See Figure 1for conditions.
pyrite is accompanied by sulfonation of the surrounding coal. The XPS spectra shown in Figure 2 are from a typical area on a freshly cleaved surface. There is no evidence of the line broadening that would indicate differential charging of the coal surface. The total spectrum (see Figure 2A) shows that sulfur species comprise about 11% of the surface, while the oxygen concentration a t 13% is typical of other fusinites analyzed. It is difficult to correlate the S(2p) and C(1s) signals since the sulfur binding energy in C-S-0 is different from that in C-S-C while the carbon 1s binding energy is littie changed since the nearest-neighbor effect dominates. In addition, the carbon binding energy will increase in species like C-S-O3 where there is sufficient oxygen to have a pronounced effect on the C 1s binding energy. Only one form of iron is recognized in the detailed Fe(2p) spectrum in Figure 2B. The sharp Fe(2p3/,) peak a t 707 eV is typical of iron pyrite, FeS2.11 The iron signal from sulfates is lost in the background; however, a sulfur peak corresponding to an iron sulfate is identified in Figure 2C. This result is not unexpected since the intensity of the sulfur 2p peak is about an order of magnitude greater than that of Fe 2p. The detailed S(2p) spectrum shown in Figure 2C can be mathematically resolved into contributions from four major chemical species (each species produces a spin-orbit split doublet). The S(2p3/2) peak (no. 1)with the lowest binding energy (BE) (charge-corrected BE = 162.4 f 0.2 eV with BE for C(1s) = 284.9 eV as reference) is clearly assignable to sulfur bonded as inorganic FeS;O while peak 2 at BE = 163.4 f 0.2 eV is ascribed to sulfur bonded as an organic sulfide." Elemental sulfur, another possible species, is eliminated since its binding energy has been measured on our instrument to be about 0.5 eV higher than that in FeS2. The two sulfur species a t higher binding energies are assigned as oxidized sulfur compounds. The S ( ~ P , /peak ~ ) at BE = 168.3 eV (no. 3) is assigned to sulfur bonded as an organic sulfonate (C-S-0,) (2C) while that (11) Clark, D.T.;Wilson, R. Fuel 1983,62, 1034.
a t BE = 169.1 eV (no. 4) is assigned to inorganic FeS04. On the basis of the total oxygen measured on the surface (13%) and the quantity of sulfur-oxygen bonds, it is estimated that 8% of the oxygen is bonded directly to carbon. Figure 2D shows the C(1s) spectrum for this surface. The major peak (no. 1)is due to hydrocarbon,peaks 2 and 3 are representative of C-S and C-S-Ox, respectively. Thus, both part C and part D of Figure 2 present evidence for the C-S bonding in a coal containing pyrite. Peak 4 represents carboxyl groups. Figure 3 displays a series of oxygen-related images. l60-, which is coincident with OH- and H 2 0 (Figure 3A-C) is shown by the l60-image to be widely distributed in the coal; however, there is oxygen enrichment in the organic region and in silicate regions identified in Figure 4. A recurrent feature associated with silicate minerals is identified by arrows in Figures 3 and 4. The persistence of oxygen-related species in this area (particularly l80-and 180H-,parts D and E, Figure 3) provides evidence that the silicate serves as a catalyst during oxidation. The presence of lag-and the molecular ion 180H- is evidence that the 1802 has reacted with the coal surface under very mild conditions. The F- image (Figure 3F) is diffuse within the organic region, suggesting the presence of carbon monofluoride: while other sources of F-are clearly associated with the silicate material. The latter source represents fluoride replacement of OH- in the mineral matter. The image obtained for H20 indicates that traces of water are present in the sample. This could arise from oxidation of FeOOH by the Cs+ primary beam. Figure 4 serves to outline the silicate-enriched regions defined by the image at mass 28- (Figure 4B), which might be caused by a mixture of 28Si-and A1H- since the mass resolution used cannot distinguish between these species. However, the image can be assigned to silicon by comparison with the image obtained a t mass 30- (30Si-with traces of Cl80-, Figure 4D). A representative silicate mineral region is identified with an arrow in Figure 4A. The images assigned to 02-and 03-,parts A and C of Figure 4, also show some enhancement in the silicate region. The 02-probably originates from adsorbed oxygen while the image for 03-may represent ozone on the coal surface. Three signals are present at about 48 Da. These represent C, 03-, and SO-. The latter signal was too low for imaging. It has been suggested that the low-temperature oxidation of coal proceeds by a free-radical chain reaction.12 Min-
Energy & Fuels 1988,2, 121-124 erals would be expected to act as catalysts in such react i o n ~ .Figure ~ 3 shows the presence of l80in molecular ions after mild oxidation while Figure 4 shows reactive oxygen species on the coal surface associated with mineral matter. These results would be expected if a free-radical mechanism is involved in low-temperature oxidation.
Conclusions Detailed SIMS images can be obtained for ions differing in mass by as little as 1:3500. The uptake of l80and molecular ions containing l80can be successfully studied with SIMS after exposure of coal surfaces to an l8OZatmosphere at room temperature. XPS provides additional insights when used on coal surfaces nearly identical with those studied by SIMS. The SIMS results, while not in themselves definitive, provide circumstantial evidence for a free-radical oxidation
121
mechanism in which clay minerals act as catalysts. Indeed the generation of 180H- and the presence of l80on the surface are indicative of low-temperature oxidation while the detection of Os-, 02-, l80and , 180H- in regions rich in clay is suggestive of catalytic activity. These suggestions have been made elsewhere in the l i t e r a t ~ r e . ~ , ~ J ~ - l ~ In a similar fashion the SIMS images for the molecular ion CS-, in conjunction with XPS results showing the existence of C-S' and C-S-Ox, seem to imply that a t least some of the sulfur present in the organic regions adjacent to pyrite arise from sulfonation of the coal by the products of air oxidation of the pyrite. The same result would be expected from in situ sulfonation of coal. Such a mechanism would be entirely consistent with the course of pyrite oxidation6s7and the relative ease of sulfonation of specific
coal^.^*^ ~~~
(12)Cole, D. A.;Herman, R.G.; Simmons, G. W.; Klier, K. Fuel 1985, 64,303.
(13)Liotta, R.; Brons, G.; Isaccs, J. Fuel 1983,62, 781. (14)Holstein, W. L.; Boudart, M. Fuel 1983,62, 162. (15)Liu, K. H.; Johannes, A. H.; Hamrin, C. E. Fuel 1984,63, 18.
Heats of Immersion of Pocahontas No. 3 Coal in n-Alkanes and Oxidized Coal in Watert James B. Hollenhead, J. 0. Glanville, and J. P. Wightman* Department of Chemistry, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 Received August 9, 1987. Revised Manuscript Received September 25, 1987 The heats of immersion of Pocahontas No. 3 coal powder have been determined in a series of n-alkanes (C8-Clg) with a Calvet MS-70 microcalorimeter. The results indicate a similar wetting behavior for the alkanes longer than dodecane; however, the higher heats of immersion of coal in c8+2 alkanes indicate the availability of more surface area to these shorter alkanes. The heats of immersion of Pocahontas No. 3 in water have also been studied as a function of oxidation time at 320 "C and as a function of extraction by pyridine and methanol prior to oxidation. A dramatic increase in the heat of immersion in water was observed as oxidation time increases. The heat of immersion of the oxidized coal in water surprisingly showed no dependence on prior extraction with either methanol or pyridine. However, the time of immersion of oxidized coal in water was reduced by either extraction procedure. X-ray photoelectron spectroscopy (XPS) was employed as a complementary technique to heat of immersion measurements. Oxygen/carbon ratios determined from the XPS spectra obtained for the oxidized coals paralleled the heat of immersion results.
Introduction The interactions that occur between liquids and coal surfaces may be studied by a variety of experimental methods including both thermodynamic and spectroscopic onea. This study incorporates microcalorimetry and X-ray photoelectron spectroscopy to ascertain the effects of nalkane chain length and the degree of oxidation on the heats of immersion of coal. An excellent but dated review of immersional calorimetry has been given by Zettlemoyer,' and a number of researchers have used this technique to examine various
carbon surfaces.2-'0 The magnitude of heat released upon the immersion of a bare solid in a liquid yields information regarding the nature of the interaction. Heat of immersion determinations are sensitive and easily distinguish between the interaction of a hydrophobic surface with a. nonpolar wetting liquid and the interaction of a hydrophilic surface with a polar wetting liquid. This sensitivity has prompted a study of the oxidation of a Virginia coal by measuring the heat of immersion in water? where the increasing heat of immersion in water parallels an increase in surface oxygen content. Several papers'lJ2 have demonstrated the
'Presented at the Symposium on the Surface Chemistry of Coals, 193rd National Meeting of the American Chemical Society, Denver, CO, April 5-10, 1987. (1)Zettlemoyer, A. C.In Chemistry and Physics of Interfaces;Rosa, S., Ed.; American Chemical Society: Washington, DC, 1965. (2)Robert, L.Bull. SOC.Chim. Fr. 1967,7 , 2309-2316. (3)Clint, H. J.; Clunie, J. S.; Goodman, J. R.; Tate, J. R. Nature (London) 1969,223,51~52. (4)Larsen, J. W.; Kuemmerle, E. W. Fuel 1978,57, 55.
(5)Glanville, J. 0.; Wightman, J. P. Fuel 1980,59, 557-562. (6) Widyani, E.;Wightman, J. P. Colloids Surf. 1982,4, 209-212. (7)Nordon, P.; Bainbridge, N. W. Fuel 1983,62, 619-621. (8)Phillips, K. M.;Glanville, J. 0.;Wightman, J. P. Colloids Surf. 1986,21, 1-8. (9)Glanville, J. 0.; Newcomb, K. L.; Wightman, J. P. Fuel 1986,65, 485-488. (10)Barton, S.;Boulton, G. L.; Dacey, J. R.; Evans, M. J. B.; Harrison, B. J. Colloid Interface Sci. 1973,44(1),50-56.
0887-0624/88/2502-Ol21$01.50/00 1988 American Chemical Society
