366
Energy & Fuels 1987,1, 366-376
amazing acceleration of these reactions by pressure. Kinetic isotope effect studies confirm the conclusion based on AV* measurements. The value of 2.3 at 400 "C indicatesa large primary deuterium isotope effect showing one hydrogen in transit in the activated complex. Similar effecta were observed for hydride transfer from 2-propanol to bromine.ls Also in quinone oxidations of triphenylm e t h a n e ~ ~and ' dihydroaromatic hydrocarb~ns'~J~ the (16)Swain, C. G.; Wiles, R. A,; Bader, R. F. W. J. Am. Chem. SOC. 1961,&3, 1945. (17)Lewis, E. S.;Perry, J. M.; Grinstein, R. H. J. Am. Chem. SOC. 1970,92,899. (18)Braude, E. A.; Jackman, L. M.; Linstead, R. P. J. Chem. SOC. 1964,3548. (19)Braude, E.A.; Brook, A. G.; Linstead, R. P. J.Chem. SOC.1954, 3569.
strongest evidence supporting a hydride transfer mechanism was large deuterium and tritium kinetic isotope effects. Another objective of this study was to identify a class of model compounds that resemble coal in their pattern of isotope effect in hydrogen transfer reactions with decalin. It is apparent that anthracene, with AV* = -60 cm3/m01and KIE = 2.1 is such a compound. This not only suggests a similarity of the reaction mechanism but also implies that anthracene (and maybe other poIycyclic aromatic hydrocarbon) units are the part of the coal structure that is reactive in decalin at high temperature.
Acknowledgment. J.P. thanks the Research Division of the New Mexico Institute of Mining and Technology for financial support.
Influence of Weathering and Low-Temperature Preoxidation on Oil Shale and Coal Devolatilizationt M. Rashid Khan Morgantown Energy Technology Center, US.Department
of Energy,
Morgantown, West Virginia 26505 Received February 6, 1987. Revised Manuscript Received May 5, 1987
The influences of preoxidation on the composition of oil shale and the yield of shale processing are not well-known. In this study, a western (Colorado) and an eastern (Kentucky) oil shale and their corresponding organic fractions (and a Pittsburgh seam coal, for comparison) were preoxidized under various conditions. Detailed pyrolysis behavior of these materials was subsequently determined by using techniques including thermogravimetric analysis combined with maas spectrometry (TGA/MS) , microdilatometry, and Fourier transform infrared spectroscopy. The feedstock H/C atomic ratios, weight loss during pyrolysis, kerogen fluidity, and hydrocarbon product yield and quality are significantly reduced by preoxidation, while the yield of oxygenated products (H20and COJ increased. These findings led to the hypothesis that preoxidation of oil shale organics introduces oxygen cross-linking bridges and/or reduces the aliphatic hydrogen contents of the shale samples. The hypothesis was tested by Fourier transform infrared spectroscopy, which confirmed that preoxidation of the shale organics introduces oxygenated functional groups such as carbonyl and carboxylic while there is a marked reduction in the aliphatic hydrogen content of the samples. The results suggest that part of the oxygen functional groups undergo decomposition and evolve during subsequent pyrolysis as H20, COP,and heavier liquids at the expense of desirable hydrocarbon fuels and overall weight loss (conversion) during pyrolysis.
Introduction and Background Previous work on weathering and low-temperature preoxidation has been reported for coals (refer to Lowry' and Van Krevelen2 for earlier work in coal and Khan and Jenkins3t4for recent developments in coal), but information for oil shale and kerogen is relatively sparse. Coburn and Ganeson5 studied oxygen uptake by the eastern and western shales and noted significantly more oxygen sorption by an eastern shale compared to that taken by a western shale. Leythaeusers monitored organic carbon and soluble organic matter (in common solvents) contents for Utah shales at various bed depths and reported significant variations in shale composition at various depths. The shale at shallow bed was considered more "weathered" (as was evident by a lower organic carbon and a soluble organic 'Presented in part, by invitation, at the 193rd National Meeting of the American Chemical Society, Denver, CO, April 1987.
matter content) compared to that observed at the deeper regions. Relatively little additional data are reported on the influence of weathering on shale composition and devolatilization mechanisms. Therefore, confusion still exists in this area. For example, Shaffer, Leininger, and Ennis' advocated weathering as a means to increase oil yield from shales. (1)Lowry, H.H.,Ed. Chemistry of Coal Utilization; Wiley: New York, 1963. (2)Van Krevelen, D. W. In Coal: Typology-Chemistry-Physics-Constitution; Elsevier: Amsterdam, New York, 1961. (3)Khan, M. R.;Jenkins, R. G. Fuel 1985,65,189. (4)Khan, M.R.;Jenkins, R. G. Fuel 1985,65,1.Also, see: Given, P. H.; Stockman, W.; Davis, A,; Zoeller, J.; Jenkins, R.; Khan, M. R. Fuel 1984,63,1655. (5)Coburn,T.T.;Ganeaon, P. Liquid Fuels Technol. 1983,I , 173-198. (6)Leythaeuser, D. Geochim. Cosmochim. Acta 1973,37, 113-120. (7)Shaffer, N. R.; Leininger, R. K.; Ennis, M. V. 1984 Eastern Oil Shale Symposium Proceedings; Institute for Mining and Minerals Research Lexington, KY, 1984;pp 401-412.
This article not subject to U.S.Copyright. Published 1987 by the American Chemical Society
Energy &Fuels, Vol. 1, No. 4,1987 367
Oil Shale and Coal Deuolatilization
Table 11. Analyses of Organic Fractions of Colorado and Kentucky Shales feed shale Colorado Kentucky Chemical Analysis (wt % Dry) mma 6.6 20.0 C 74.1 54.7 S 4.1 12.4
Table I. Characterization Data A. Characteristics of Untreated Oil Shales Colorado Modified Fischer Assay: oil yield, gal/ton 32.8 density of oil, g/cm3 0.920 Chemical Analysis (wt % Dry) mineral matter 78.7 moisture" 0.2 organic C 17.0 mineral C 5.3 H 2.2 H/C (organic) 1.55 nitrogen 0.6 total sulfur 0.7 organic sulfur 0.2 0.0 sulfate sulfur pyritic sulfur 0.5
Kentuckv 11.5 0.956
Ultimate Composition* (wt % Dry Mineral Matter Free (dmmf)) C 79.6 74.5 H 10.4 7.5 N 2.7 2.2 S 1.1 4.7 0 (by difference) 6.0 11.3
81.1 1.0
14.1 0.0
1.6 1.36 0.4 3.3 0.9 0.3
a Mineral matter as determined directly from low-temperature ashing. bBased upon elemental composition of the feed shale. Additional descriptions of the shales and their organic fractions are available.lb
2.1
Table 111. Principal Minerals in Shales (as Percent of LTA) As Determined by X-ray Diffraction Colorado Kentucky illite 22.2 57.0 a 6.3 kaolinite 16.5 27.9 quartz 0.8 0.5 plagioclase 15.4 a calcite 29.4 ankerite and/or dolomite 2.2 siderite 7.8 0.6 orthoclase 0 1.7 bassanite pyrite 5.8 5.6
Elemental Analysis (wt % of the High-Temperature Ash) CaO 27.9 0.4 KZO 2.0 4.1 FeZ03 2.0 7.0 7.0 2.1 MgO 42.7 59.2 SiOz A1Z03 12.9 18.9
B. Characterization Data for Pittsburgh No. 8 Coal Ultimate Analysis (daf) % C %H %N %S H/C Btu/lb 83.75 5.46 1.56 2.15 0.79 13976 Proximate Analysis % ash % volatile matter 7.27 37.86
% moisture
0.57
a
C. Comparison of Aromaticity of Various Feedstocks sample aromaticity, fa Pittsburgh No. 8 Coal 0.705 Sunbury Oil Shale 0.45 0.29 Colordao Oil Shale As received.
The first step in the oxidation of coal is generally considered to be peroxide formation, which is facilitated by the oxidation of aliphatic/olefmic/aromaticstructures (R') via free-radical mechanisms:
R'
+
+0 2
-
R-0-0'
(2)
R-0-0' R-H -* R-0-0-H R' (a peroxide or hydroperoxide) (3)
+
Reaction 1can be considered as the initiation of a chain reaction that is further propagated by reactions such as (2) or (3). The peroxide or hydroperoxide formed in step 3 can undergo further decomposition,producing relatively stable oxygenated products (e.g., alcohols, ketones, aldehydes, esters, and ethers) as well as CO, COz, and HzO. The mechanisms and decomposition chemistry of peroxides into other products have been discussed by various authors.&l0 The comparatively more stable oxygen func-
Denotes amounts too small to determine.
tional groups such as hydroxyl, carboxyl, carbonyl, and ether have been identified in oxidized coals both chemicallyll and spectroscopically.12 Weathering of coal has generally been studied in two ways. First, simulated or artificial weathering of coal is effected under laboratory conditions that involve oxidizing at a certain temperature for a precise time period. Second, the coals were compared in the "original" state with the coals of the same deposit that have been exposed to the atmosphere for a prolonged and often undefined time period. In this investigation, a detailed systematic study of weathering/preoxidation of oil shales (Colorado and Kentucky) and the organic fraction of these oil shales was performed. For comparison, a Pittsburgh seam coal was also preoxidized.
Experimental Section Sample Preparation and Characterization. A Colorado shale (freshly mined) from the Pilot Colony mine operated by Exxon and a Sunbury shale from Kentucky (obtained from Kentucky Center for Energy Research Laboratory, KCERL) were used for this study. The Colorado shale is representative of the Mahogany zone of the Green River formation. A coal sample (Pittsburgh No. 8) freshly obtained from the mine mouth, was also investigated for comparison. Large chunks of samples were ground to -8 mesh and stored in N2 until their usage. Acid demineralization of these shales was performed in Nz by using procedures described by Durand and Nicaise13 to separate the organic-rich fractions. The demineralization was achieved by a series of treatments with acids: HCl (9) Chamberlain, E. A. C.; Barrass, G.; Therlaway, J. T. Fuel 1976,55,
n. n
111.
Volborth, A. "Oxidation of Coal"; Report, June 16, 1976, DOE Contract No. EY-'76-5-02-2898;Chem. Abstr. 1977, 47200. (11) Liotta, R.; Brons, G.; Isaacs, J. Fuel 1983, 62, 781. (12) Painter, P. C.; Snyder, R. W.; Starsinic, M.; Coleman, M. M.; Kuehn, D. W.; Davis, A. Appl. Spectrosc. 1981,35, 475. (13) Durand, B.; Nicaise, G. Kerogen; Editions Tech Paris, 1980. (10)
(8) (a) deVries, Von H. A. W.; Bokhoven, C.; Domans, K. N. M. Brennst.-Chem. 1969,50,289. (b) Jones, R. E.; Townend, D. T. A. J . SOC. Chem. I d . , London 1949,68,197. (c) Chakravarty, S. L. J. Mines, Met. Fuels 1960, 8, 1. (d) Studebaker, M. L. Proceedings Third Biennial
Carbon Conference; Pergamon: Oxford, England, 1967; p 289.
368 Energy & Fuels, Vol. 1, No. 4, 1987 for the removal of carbonates and HF for the removal of silicates. Float-sink separations were performed in a solution of specific gravity 1.5 to minimize pyrite content in the organics. The nature of the organic fractions w a ~tested with a microdilatometer to ensure that the demineralization process did not cause dramatic alterations in the thermoplasticity of materials.14JS8 The characterization data for the raw shales (and the coal), the organic fractions, and the principal minerals identified in shales are presented in Tables 1-111 respectively. Table IB provides characterization data for the Pittsburgh No. 8 coal. A comparison of aromaticity of various feeds is given in Table IC. The nature of organic fractions was tested with a microdilatometer to ensure that the demineralization procedure did not destroy the thermophysical properties of the shale organics. More detail on thermophysical behavior of shale or their organic fractions is the subject of a separate r e p ~ r t . l ~ * ' ~ ~ , ~ Samples of eastem (Kentucky Sunbury) and western (Colorado Mahogany) shales were ground and sieved (to -74 pm size fraction) in Nz atmosphere (using a mortar and a pestle) to minimize weathering by using a technique similar to that reported by Khan and Jenkinsa3s4Simulated accelerated weathering of shales and their organic fractions to various severities was achieved by treatment at 150 "C in dry air for up to 6 days (and at 50 "C for 292 days for the Colorado shale). Elemental analyses (H, C) of selected preoxidized shale organics were performed. Detailed pyrolysis behavior of the untreated and weathered materials was subsequently determined in He by using a thermogravimetric analysis (TGA) system (heat treated to -650 "C) interfaced with a mass spectrometer, more details of which are presented elsewhere.14 The TGA data were generally reproducible to within 1% of the initial sample weight utilized. Additional pyrolysis experiments (at 500 "C, residence time 1h) on the untreated and preoxidized eastern and western shales were performed by using a slow heating rate organic devolatilization reactor (SHRODR). Details of this system are reported elsewhere.l& For SHRODR experiments, the feedstocks were preoxidized at 50 "C for 95 days. The particle size used for SHRODR work was -8 mesh. Additional SHRODR work is in progress. The SHRODR liquids were fractionated by using sequential elution solvent chromatography (SESC), more details of which are available.lBb The liquid chromatographic separation was achieved by gravity flow on a silica gel column by sequential elution of the liquids with solvents of increasing polarity according to a procedure described elsewhere.lBbIn this work six fractions were collected with 95-100% recovery of the sample. The silica gel (Baker Analyzed 3405-5,74-250 pm particle size) was washed twice with absolute methanol and dried overnight at 130 "C. Its water level, after it was cooled to room temperature, was adjusted to 4 wt % and stored in a closed container. One hundred (100) milliliters of this material was placed in a glass column and conditioned with hexane. Then, to about 1.5 g of pyrolysis liquid were added 10 mL of silica gel and 5 mL of sand, the mixture was stirred well and placed on top of the column, and the column was eluted with 200 mL of each of (previously dried HPLC grade) the following solvents at a flow rate of 4 mL/min: (a) hexane, (b) hexane/l5% toluene, (c) chloroform, (d) chloroform/4% diethyl ether, (e) diethyl ether/3% ethanol, and (f) methanol. The solute and the solvent were collected in a flask. After collection, the solvent was removed by rotary evaporation at 40 "C under partial vacuum. The residue was transferred to a tared vial, the flask washed with methylene chloride, and the excess wash solvent removed in a stream of nitrogen a t ambient temperature to constant weight. Efficient separation was achieved according to an earlier experiment by Seshadri et al.16bInfrared and NMR spectra verified that good separation of fractions was achieved. (14)Khan, M. R. Proceedings of the First Annual Oil ShalelTar Sand Contractors Meeting; Morgantown Energy Technology Center: Morgantown, WV, 1985;pp 104. Khan, M. R.Prepr.-Am. Chem. Soc., Diu. Pet. Chem. 19 32(1), 49-56. (15)(a)Khan, M. R.Fuel 1987,66,415.(b) Khan, M.R.1985 Eastern Oil Shale Symposium Proceedings; 1985;pp 195-206. (c) Khan, M. R. Fuel Sci. Technol. Znt. 1987,5 , 185-231. (16)(a) Seshadri, K. 5.;Cronauer, D. Fuel 1984,62,1436.(b)Shadle, L. J. et al. 1985 Eastern Oil Shale Symposium Proceedings; 1985;pp 185-194.
Khan Table IV. Influence of Preoxidation on the Hydrogen t o Carbon Ratios of Untreated and Oxidized Samples (Preoxidized at 150 " C ) organic fractions Pittsburgh no. sample Colorado Kentucky No. 8 coal 1 untreated sample 1.57 1.2 0.79 2 preoxidized 2 days 1.25 0.89 0.71 3 preoxidized 3 days 1.17 4 preoxidized 6 days 0.85 0.75 0.65 5 "oxidation index lnn46 37 19
'[(I) - (4)1/(1)X 100.
z
I
A
UNTREATEDSUALE
B
0
100
200
300
400
500
600
OXIDIZED 6 DAYS AT 15OOC
700
800
900
1000
1101
TEMPERATURE (OC)
Figure 1. Effects of air preoxidation of Kentucky Sunbury oil shale on weight loss during TGA pyrolysis, heating rate 50 "C/min. In addition to using the TGA/MS and the high-pressure microdilatometric (HPMD) systems, Fourier transform infrared spectroscopy (FTIR)spectra of untreated and preoxidized samples of the Colorado shale kerogen were obtained. A detailed description of the HPMD procedures is presented elsewhere.lk The FTIR spectra of the samples (mixed with KBr and pressed at 20 000 psi to produce pellets) were obtained by a Nicolet 170SX FTIR spectrometer using a procedure similar to that described by Painter, et a1.I2 The spectra were obtained at the same equipment setting. In addition, experiments were repeated with SiOz as an internal standard to facilitate comparison of various peaks.
Results and Discussion Influence of Weathering on the Elemental Composition of Shales. Elemental analyses of t h e oxidized a n d untreated samples show that the H / C ratios of the samples were significantly reduced during 6 days of oxidation (Table IV). While the H/C ratio of t h e Colorado oil shale organic fraction was reduced by 46% (and by 37% for t h e Kentucky shale compared t o unoxidized samples) during 6 days of air oxidation, t h e reduction in this ratio for Pittsburgh No. 8 coal was -18% when preoxidized under t h e same condition. These results demonstrate that t h e organic fractions of shales are more apt t o undergo oxidation than the Pittsburgh No. 8 cod. T h a t is, t h e highest "oxidation index 1" (defined in Table IV) was noted for t h e Colorado organic fractions, which was significantly greater than that noted for Pittsburgh No. 8 coal.
Thermogravimetric Analyses of Weathered and Untreated Shales. T h e influences of air oxidation of Kentucky Sunbury oil shale on weight loss during T G A pyrolysis (heating rate 50 OC/min) are shown in Figure 1. It can be seen that at 700 OC, the overall weight loss during pyrolysis is significantly reduced by 6 days of preoxidation at 150 "C. In addition, t h e maximum rate of weight loss
Oil Shale and Coal Devolatilization
Energy & Fuels, Vol. I, No. 4, 1987 369
110A UNOXlDlZED
100-
110-
-
100-
90
80 -
-
70
60 50
90
-
80
-
70
-
60
-
OXIDIZED 6 DAYS AT 150%
~
OXIDIZED 2 DAYS AT 150°C
-
UNTREATED
A
20 L
UNTREATED
B OXIDIZED 2 DAYS AT 15OOC
16C
OXIDIZED 6 DAYS AT 150'C
12-
lir 1 4 t
B OXIDIZED 1 DAY LT 15O'C C OXlOlZED 2 DAYS AT 15O'C D OXIDIZED 3 DLYS AT 150°c E OXIDIZED 6 DLYS AT 150'C
/A
8-
0
100
200
300
400
500
600
700
800
900
1000
1100
TEMPERATURE (OC)
Figure 2. Effects of air preoxidation of Kentucky Sunbury shale separated organic components on weight loss during TGA pyrolysis, heating rate 50 "Cfmin. is also significantly reduced during oxidation. Although the weight loss by the preoxidized samples is lower when compared to that of the untreated shale above -450 "C, below this temperature the preoxidized shale actually loses more weight when compared to the unoxidized shale. The effects of preoxidation of Kentucky Sunbury shale organic components on weight loss during TGA pyrolysis is shown in Figure 2. The data shown in this figure are qualitatively similar to that noted for the parent shale (Figure 1). Preoxidation of the oil shale organic fractions leads to a monotonic reduction in weight loss and the maximum rate of weight loss during subsequent pyrolysis. This effect of preoxidation on weight loss during subsequent pyrolysis of shale or organic fraction of shale is qualitatively similar to that noted for Pittsburgh No. 8 coal, although the pyrolysis of coal continues to a much higher temperature (Figure 3). The influence of air preoxidation on weight loss during subsequent pyrolysis of the Colorado oil shale is shown in Figure 4. Preoxidation of this oil shale, like the Kentucky shale, significantly reduces weight loss during subsequent pyrolysis. In addition, as the oxidation time increases, the weight loss monotonically decreases. The effects of air oxidation on weight loss behavior of the organic fractions of the Colorado shale are shown in Figure 5 ( t h e format in this figure is changed from the previous cases to better illustrate the changes occurring during the initial stage of pyrolysis). In this figure, the fractional weight loss is defined by W / Wo. Here, W denotes the weight loss by the oxidized material at any given temperature, while Wostands for the weight loss by the untreated material at 600 "C. As the oxidation time progresses, the fractional weight loss decreases monotonically during subsequent pyrolysis. For comparative purposes, He gas atmosphere was also used during heat treatment at 150 "C. Heat treatment of the oil shale organic fractions in He for 6 days led to a slight reduction in fractional weight loss during subsequent pyrolysis. The small reduction in weight loss is probably attributable to loss of bitumen by evaporation during the
TEMPERATURE I'C)
Figure 3. Influence of air preoxidation of Pittsburgh No. 8 coal on weight loss during pyrolysis, heating rate 50 OCfmin. 1081
80
-
T TEMPERATURE i'Cl
Figure 4. Effects of air preoxidation of Colorado oil shale on weight loss during TGA pyrolysis, heating rate 50 OC/min. heat treatment in He at 150 "C. Obviously, heat treatment in air (i.e., oxidation) influences the overall weight loss
370 Energy & Fuels, Vol. 1, No.4, 1987
Khan
Table V. Comparison of the Influence of Air Oxidation (at 160 "C) on Weight Loss (%) During Subsequent Pyrolysis of Oil Shale, Shale Organics, and Pittsburgh No. 8 Coal (Heated at 50 "C/min to 600 "C) Colorado oil Kentucky oil shale Colorado shale Kentucky Pittsburgh parameter no. (Mahogany) kerogen (Sunbury) kerogen No. 8 coal 1 weight loss by untreated sample, wt % 17 80 14 44 30 2 weight loss by the oxidized (6 days) sample, wt % 14 58 11 32 22 3 22 3 12 8 3 differences in weight lossa 81.1 20.0 8.6 4 mineral matter (mm) dry content,bwt % 78.7 6.6 21.3 93.4 18.9 80.0 91.4 5 organic fraction contentc 6 "oxidation index 2"d 14.1 23.5 15.9 15.0 8.75
-
OAW (1)- (2). bThe mineral matter contents of the shale samples were determined by acid demineralization while those for the kerogens were determined directly from low-temperature ashin More details regarding the demineralization procedures and the characteristics of the shales are presented elsewhere.'" 100 - mm. %aW/(laO - mm) X 100.
g In
9
O"-
0.6
NOT OXiDlZED B HEATED 6 DAYS IN He AT 150°C C OXlDiZED 6 HOURS AT 150T D OXIDIZED 2 DAYS, 150% E OXIDIZED 3 DAYS, 150'C F OXIDIZED 6 DAYS, 150'C A
-'.O
-
Y
3
N e w Albany Shale
MATERIAL AT 600'C W = WT. LOSS BY OXIDIZED MATERIALS AT ANY GiVEN TEMPERATURE
c
$
.
- --
Shale Oxidized, 292 day8 at 50'' C 4 0 mesh New A b m y 1R.w)
D
E
96
-.
200 mesh
F
0.5-
92
0.4-
90
A
5
88 I
I
I
I
l
\
I
I
I
I
I
N e w Albany Shale
200
300
400
500
600
TEMPERATURE ('C)
-
Shale Oxidized, 292 days a1 50'C Ahany (Raw) - - - Newmsih --200 mesh -SO
Figure 5. Effects of air oxidation on fractional weight loss of Colorado oil shale organics during TGA pyrolysis, heating rate 50 OC/min.
during pyrolysis more dramatically than heat treatment in He. The results also suggest that decomposition of oxygen complexes such as carboxylic acid groups that may also be present in the shales does not introduce significant variations in the subsequent TGA weight loss. Quantitatively, preoxidation influences organic fractions more severely than the parent Colorado shale (on organic matter basis). When the parent shale was heated to 600 "C at 50 "C/min, the reduction in weight loss during pyrolysis after 6 days of oxidation was 3%. For the organic fractions of this shale, however, there was about 12% reduction in weight loss during pyrolysis compared to that noted for untreated material. When the differences in the weight loss by the preoxidized and untreated samples are normalized on mineral-matter-free basis (Table V), it is confirmed that the organic fraction of the Colorado shale is more susceptible to oxidation compared to the parent shale. In contrast, the preoxidation effect on weight loss during pyrolysis for the whole Kentucky shale is comparable to that noted for the organic fractions of this shale. The data also demonstrate that the shales or their organic fractions are markedly more affected by 6 days of air oxidation than the Pittsburgh No. 8 coal (i.e., the "oxidation index 2" [defined in Table VI is higher for shales and their organic fractions compared to that for the Pittsburgh No. 8 coal). These trends hold in spite of small variations in TGA weight loss data (maximum of 1% based on initial sample weight). The trends reported in Table V, for the organic fractions and the Pittsburgh No. 8 coal, are qualitatively similar to that observed in Table IV (based on H/C ratios). The reduced weight loss for the preoxidized samples at elevated temperatures is attributable to polynuclear aromatic condensation reactions including formation of diary1 ether linkages (discussed later). Additional TGA Data on Samples Weathered at 50 "C. Additional data for Colorado oil shale that was oxidatively weathered at 50 "C over 292 days confirmed the
-
100 200 300 400 500 600 700 800 900 1000 1100 1
Temperature P C )
Figure 6. Influence of air preoxidation on New Albany shale pyrolysis weight lms and rate of weight loss. A shale particle size of -60 or -200 mesh was used for oxidation.
results described above. The influence of shale particle size on oxidative degradation and on weight loss during subsequent pyrolysis was also studied. The shale particle size fractions investigated were -8, +16, -60, and -200 mesh. The weathered samples were ground to -200 mesh prior to devolatilization in a TGA. Weathering reduced the weight loss during subsequent pyrolysis at 550 "C by 25-36% compared to that for raw shale. The smallest size fraction had the greatest reduction in weight loss, although the observed variations for the three particle size fractions were within the experimental errors of measurements. The weathered shales also demonstrated a reduction in the hydrogen to carbon ratio of 6-15% as compared to the raw shale. The effect of preoxidation on weight loss during pyrolysis of a New Albany shale (of various particle size) was also investigated (Figure 6). The influence of preoxidation on the composition of this shale is summarized in Table VI. These results are qualitatively similar to those noted for the other samples just discussed. Selected Gaseous Product Analysis. The effects of preoxidation on the nature of products generated during pyrolysis were investigated using a mass spectrometer interfaced with the TGA. The effects of preoxidation of Colorado shale organic fractions on COz and H20 evolution during pyrolysis (50 "C/min to -650 "C) are shown in Figure 7. The data show that as the oxidation time increases, the rates of evolution of C02 and H 2 0 at any temperature between 200 and 600 "C monotonically increase. Our MS data were not sufficiently quantitative to correlate the increase in weight loss below 450 "C to the
Oil Shale and Coal Deuolatilization
Energy & Fuels, Vol. 1, No. 4, 1987 371
TEMPERATURE
(OC)
Figure 7. Influence of air preoxidation of Colorado (Mahogany) shale organic fractions on C 0 2 and H20evolution,heating rate 50 OC/min. ”.”-”
Table VI. Influence of Air Preoxidation at 50 O C for 292 Days on New Albany Shale Composition
%C %H %N H/C (atomic)
raw shale 11.11 1.31 0.46 1.41
preoxidized shale -200 mesh -60 mesh 10.17 10.15 1.11 1.09 0.46 0.50 1.31 1.29
increase in C02 and H 2 0 evolution noted for the oxidized materials. Figure 8 demonstrates that with the increase in the extent of preoxidation, the yield of CHI is significantly reduced. These results are similar to that noted in the case of Pittsburgh No. 8 coal (Figure 9). Thermophysical Properties. The influence of preoxidation on the thermophysical nature of the Colorado shale kerogen was investigated by using a high-pressure microdilatometer (HPMD), a detailed description of which * * ~the is available e l ~ e w h e r e . ~ It * ~was J ~ ~~ b s e r v e d l ~that kerogen from a Colorado shale undergoes significant contraction when heated, reflecting development of a highly fluid stage. This fluid behavior (as reflected by large contraction of the sample during heat treatment) was markedly reduced as the kerogen sample was progressively preoxidized (Figure 10). Indeed, upon 6 days of preoxidation, the extent of contraction is reduced to zero, suggesting that the fluid intermediate state was essentially destroyed by oxidative cross-link formation. The loss of fluid intermediate behavior of kerogen by preoxidation can be caused by the formation of ether linkages, which can increase the size of kerogen melt or “thermoplast” (defined in ref 15a,b) during pyrolysis. Ether functional groups may undergo additional condensation reactions forming furan ring structures at increasing temperatures by reactions such as
Furthermore, ether groups joining two saturated carbons are generally weak and can form C-C structures. All these reactions would increase the molecular weight of the pyrolysis liquids generated from the preoxidized materials. Such heavier molecular weight material would have diminished molecular mobility and fluidity. This influence
0.036
-
0.032
-
0.028
-
0.024
-
0.020
-
Q
5= K >
< 1 K
s 0 3 I-
6 D
I’ 0
0.016r
0.000 100
,
200
I *
300
400
I
I
I
500
600
700
TEMPERATURE (OC)
Figure 8. Effects of preoxidation on CHI evolution for the Colorado shale organic fractions, heating rate 50 OC/min. of weathering on fluid behavior during pyrolysis is consistent with that iioted for Pittsburgh and lower Kittanning seam coals by Khan and J e n k i n ~ . ~ ~ ~ Structural Changes. The results of TGA, elemental, and evolved gas analyses suggest that the preoxidation mechanisms of oil shale and their organics are qualitatively similar to that noted for the Pittsburgh No. 8 coal. These data led to a hypothesis that preoxidation of oil shales or their organic components introduces oxygen cross-linking (bridges) in their structures and, thus, “ties-up” functional groups that evolve during subsequent devolatilization primarily as HzO and COP rather than desirable hydrocarbons. To verify this hypothesis, Fourier transform infrared spectroscopy (FTIR) studies of untreated and preoxidized samples were performed. The FTIR spectra for the raw (unoxidized) and preoxidized organic fractions of Colorado oil shale are presented
Khan
372 Energy & Fuels, Vol. 1, No. 4, 1987
TEMPERATURE I'Ci
TEMPERATURE I'Ci
Figure 9. Influence of preoxidation of Pittsburgh No. 8 coal on C 0 2and CHI evolution during TGA pyrolysis, heating rate 50 OC/min. 1 .o
0 75
nI\
c=o
0
f
8
0.50
i
e 0.25
L
\ 4\.
ALIPHATIC ETHERS
I
AROMXC C.H OUT OF PLANE BENDING
PREOXIOATION TIME (days)
Figure 10. Effect of preoxidation on Colorado kerogen volume contraction parameter (Vc, %) as determined by a microdila-
tometer. Contraction is a measure of kerogen softening.
in Figure 11. The raw sample showed absorptions at 2924 and 2853 cm-l due to its aliphatic C-H stretching modes. Equal intensity bands at 1707 and 1636 cm-l can be assigned as C=O stretching of ketone, aldehyde and/or acids, and highly conjugated C=O, respectively. The spectra demonstrated that as the extent of preoxidation progresses from 0 to 3 days and to 6 days, the band intensity ratio of C = O to aliphatic C-H increased, as shown in Figures 11B,C suggesting a decrease of aliphatic C-H group and/or an increase of C = O groups. In addition, the absorption in the 1330-1110-~m-~ region of the preoxidized samples increased, suggesting an increase in C-0 bands. The intensity ratio between the 1714-cm-' band and the 1635-cm-' band increased, suggesting some C=C may be preoxidized. In addition, the relative intensity of the 1456-cm-l band (aliphatic CH2bending and some aromatic ring mode) decreased, especially for the sample preoxidized for 6 days. For the more severely preoxidized (6 days) sample, there appear to be more absorptions due to C-0 groups and unconjugated C=O groups in the ketone, aldehyde, or acid forms (Figure 11B). Obviously, the extent
I
,
4000 3 6 0 0 3 2 0 0 2800 2 4 0 0 2000
I
1600
1200
I
,
800
400
WAVENUMBERS (cm'')
Figure 11. FTIR spectra of untreated and preoxidized Colorado kerogen samples: (A) unoxidized; (B)air oxidized at 150 O C for 3 days; (C) air oxidized at 150 O C for 6 days.
of oxidative weathering is significantly more severe for the 6-days samples compared to that noted for the 3-days oxidized sample. These results are qualitatively similar to those observed for the Sunbury kerogen (Figure 12) or Pittsburgh No. 8 coal (Figure 13). This is the first concrete evidence that preoxidation can cause changes in the organic structures of eastern shale. Coburn and Ganesons attributed the weathering effects of eastern shales to the changes in the inorganic matter only (e.g., oxidation of pyrite to iron sulfate). Furthermore, it was suggested that the influences of weathering on pyrolysis yield could be reversed by removal of soluble sulfates by water washing. In our work, we could not regenerate the weight loss profile by water washing of preoxidized eastern shales. This again confirms that preoxidation of the eastern shale brings about changes in the organic structures of shales that could not be reversed simply by water washing. Figure 13.1shows that the Pittsburgh No. 8 coal samples contained aliphatic C-H groups (2950-2800 cm-l), aromatic
Energy & Fuels, Vol. 1, No. 4, 1987 373
Oil Shale and Coal Devolatilization
OUT OF PLANE BENDING 2 ~
4000
3600
3200
2000
2400
2000
1600
1200
800
400
WAVENUMBERS (cm ')
Figure 12. ITIR spectra of untreated and preoxidized Sunbury
4000
kerogen samples: (A) unoxidized; (B) air oxidized a t 150 "C for 6 days.
rings or C=C bonds (1602,900-700 cm-l), -CH3 groups (1375 cm-l), and kaolinite clay (1035 and 1010 cm-l). Because it is safe to assume the kaolinite clay remained unchanged under the mild air oxidation conditions, the kaolinite absorption was used as an internal standard for the difference spectra between the oxidized and unoxidized coal samples that were obtained. The two difference spectra (Figure 13.2 [6-day oxidized-unoxidized] , and Figure 13.3 [3-day oxidized-unoxidized]) showed that there were negative absorbances in the 2950-2800-~m-~ region (CH stretching), 1610-cm-l region (C=C band), and 1441-cm-' region (CH2 and CH3 bending), and positive absorbance in the 1700-cm-l region, indicating the loss of aliphatic C-H groups and the increase of carbonyl groups after oxidation. The changes appeared to be larger after 6 days of oxidation. The aromatic ring modes in the 900-700-~m-~ region did not show many changes in the difference spectra, indicating that the aromatic rings probably remained intact. These observations are consistent with the earlier study of Painter et al.12 The 3050-cr~~ region, in the difference spectra, would have also confirmed any change in the aromatic C-H groups. Due to the high background above 2800 cm-', this could not be done. Influence of Preoxidation on Pyrolysis Liquid Composition. In order to understand the influence of low-temperature preoxidation on the quality of liquid products, pyrolysis experiments were performed by using a slow heating rate organic devolatilization reactor (SHRODR).Some preliminary results are being presented. Figure 14 provides a comparison of the feedstock properties influenced by low-temperature preoxidation (50 "C, 95 days in air). The results demonstrate that preoxidation under this mild condition significantly deteriorates the quality of the feedstocks by reducing the heating value, volatile matter contents, and H/C (organic) atomic ratio and by increasing the oxygen content (direct oxygen, determined by Hoffman Lab, Golden, CO). A comparison of pyrolysis liquids derived from coal and two oil shale samples is presented in Figure 15. The results confirm that while the H/C ratios of the liquids
3600
I
I
I
I
I
I
I
3200
2800
2400
2000
WOO
OOO
800
400
Wavenumbers (cm')
3=0-A .09
-
3 -.OB
-
A
c
E s
e
-21
f
- -
9 -36 .61
iooo
I
1
I
~(IW
3zoo
ZBOO
I
zu#
I
I
I
I
zoo0
WOO
im
Em
400
Wavenumberr (cm-1)
Figure 13. (1) FTIR spectra of Pittsburgh No. 8 coal: (A) untreated; (B) preoxidized a t 150 "C 3 days; (C) preoxidized at 150 "C 6 days. (2) Difference spectrum of the spectrum of untreated coal from the spectrum of 6 days preoxidized coal. (3) Difference spectrum of the spectrum of untreated coal from the spectrum of 3 days preoxidized coal. 18
1.6
-
1.5
1.4
2
E 0
1.3
E
0
-
14
p1
"
E
1.2 1.1
10
0
I 1
c
.E D
0.0
6
0.8 0
07
Pill 8
Col Shale
Ken Shale
Pill 8
Pill 8
Col Shale
Ken Shale
Pill
8
Col Shale
Ken Shale
Col Shale Ken Shale
Figure 14. Influence of low-temperature air preoxidation (50 "C, 95 days) on composition of feedstocks.
are significantly reduced, the oxygen content (measured directly) and the refractive index (measured by the Hoffman Lab) of the liquids derived from the preoxidized feedstocks are markedly increased. This finding suggests, for the first time, the potential to use "refractive index"
Khan
374 Energy & Fuels, Vol. 1, No. 4, 1987
--
7.88
8
1.6 1 1.59
7
1.58 5 6 3
-
$ 1.57
'
$ 5
1.56 > 1.55
z
G 4
m
f
c
; 3
1.551
1.54
a
1.53
0
2
1.52
1
1.51 1.5
0 Pitt 8
Col Shale
K e n Shale
Pitt 8
Col Shale
K e n Shale
0Raw Preoxidized
Figure 15. Influence of low-temperature air preoxidation (50 "C,95 days) on properties of SHRODR liquids (obtained by pyrolysis a t 500 "C).
Raw Preoxidized
Colorado Shale Oil
Table VII. Variations in Physical Properties for Eastern and Western Shales (As Received Basis)"
co2 N2 surface surface areab areaC Colorado shale 4.7 2.0 preoxidized Colorado shale (95 days at 50 "C) 3.6 12.8 Colorado spent shale (SHRODR, 500 "C) Colorado kerogen 33.2 6.7 Sunbury shale 21.5 8.4 preoxidized Sunbury shale (95 days at 50 "C) 43.9 41.8 Sunbury spent shale (SHRODR, 500 "C) Sunbury kerogen 84.1 8.2 New Albany kerogen 87.2 11.2 "The particle size for the kerogen samples was -74 pm (-200 mesh). The particle size for the remaining samples was -8 mesh. C02 surface area was determined by applying the Dubinin-Polanyi equation. CN2BET surface area.
E0,
.-
Aromatics
30
-
20
-
Hetercyclics
Neutral. Aroma1ks
SunbutyShale Oil
n
10 -
0 -
1
6
Fraction Number
Figure 16. Relative quantities of fractions in sequential elution solvent chromatography of liquids generated from raw and preoxidized (in air a t 50 "C for 95 days) shales.
of liquids as a rapid technique to characterize the oxidative weathering or stability of fossil fuel liquids. The liquid samples generated from the untreated and preoxidized feedstocks were separated by using sequential elution solvent chromatography (SESC). The weight percent of liquid obtained in each fraction and the identity of each fraction is presented in Figure 16. The results demonstrate that the alkane/neutral aromatic fraction is lower and the polyfunctional component is larger for the liquids derived from the preoxidized oil shale. A significant increase in the concentration of fraction 6, a t the expense of fraction 1 can be attributed to the following possible reactions: (a) formation of ketones through oxidation of methylene units in benzylic positions; (b) oxidation in the acid groups followed by esterification with a phenol. The changes in the organic structures of feedstocks and the liquids derived from the untreated and mildly preox-
idized (50 OC for 95 days, Figure 14)samples suggest that the eastern shales are more susceptible to oxidation than the western shale, contrary to the findings presented in Tables IV and V. This apparent contradiction can be resolved by the following arguments. Surface area and porosity measurements of eastern and western shales (Table VII) demonstrate that the eastern shales compared to the western shales are of more open pore structure and contain significantly greater BET N2or C02 (micropore) surface areas. It is known3that low-temperature oxidation of coal is manifested by slow diffusion of oxygen into the cm2/g has pore structures. A diffusion coefficient of been estimated for oxygen diffusion into coals. It is also reported3 that for microporous materials like coal or char, diffusion is activated and is strongly dependent on temperature. Keeping this in mind, I propose that, under low-temperature conditions (as performed at 50 "C in this study, or at 30 "C used by Coburn and Ganeson5), the weathering potential of feedstocks would be a strong function of accessibility of oxygen into their structures. Indeed, our calculations of diffusion coefficients of O2into the eastern and western shales (based on O2chemisorption rate data obtained at a relatively low temperature [50 "C]) demonstrated that the diffusion coefficient was an order of magnitude greater for the eastern Sunbury shale compared with that for the Colorado shale. It is, therefore, expected that the eastern shales would be more accessible to oxygen under low-temperatureconditions compared to the western shales. However, if one overcomes the diffusion limitations by performing the oxidation work at a relative higher temperature (as performed at 150 "C in this study), the western shales with relatively more hydro-
Oil Shale and Coal Deuolatilization
Energy & Fuels, Vol. 1, No. 4, 1987 375
Table VIII. Origin of Oil Shales, Their Grades, and the Data Sources Utilized in Figure 17 lenend no. ref no. t m e of shale grade. gallton Maier/Zimmerley (1924) 1 19 2 Hubbard/Robinson (1950) 20 Diricco/Barrick (1956) 21 4 Johnson et al. (1975) 22 10 Braun/Rothman (1975) 23 11 Campbell et al. (1978) 24 15 Shih/Sohn (1980) 25 18 26 19 Rajeshwar (1981) Wallman et al. (1981) 27 20 21 Rostam-Abadi (1982) 28 22 Rajeshwar/Dubow (1982) 29 30 23 Wang/Noble (1983) 24 Joshi/Lee (1983) 31 31 Joshi/Lee (1983) 25 Elder/Reddy (1983) 32 26 Shen et al. (1984) 33 27 34 Yang/Sohn (1984) 28 35 29 Sohn/Yang (1985) 35 Sohn/Yang (1985) 30 Sohn/Yang (1985) 35 31 32 present study present study 33 present study 34 present study 35
aromatic (e.g., tetralin) structures (and a higher H/C) may be more susceptible to oxidation (defined by TGA weight loss) compared to the eastern Sunbury shale. This finding is consistent with our unpublished data on the oxidation potential of extract and residue separated from the Pittsburgh No. 8 coal by pyridine extraction. The oxidation index (defined similar to that described in Table V) was markedly greater for the extract compared to that for the residue, although the residue had a significantly higher aromaticity compared to that for the coal or the extract. Influence of Preoxidation on Pyrolysis Kinetic Parameters. To determine the influence of preoxidation on pyrolysis kinetic parameters, a distributed activation energy modell' was applied to evaluate the TGA data. The model assumes that pyrolysis occurs by a large number of parallel first-order reactions having a normal distribution of activation energies. The weight loss during pyrolysis process can be represented by an integral relationship
v* - v V*
t
1
1
I
I
I
l
I
l
I
I
A EU WU C
I
= AUSTRALIAN
= EASTERN U.S = WESTERN U S = CHINESE
10.'
10-2 -
(E - Eo)2/2u2] dE
(17)Anthony, D. B.; Howard, J. B.; Hottel, H. C.; Meissner,H. P. Fuel
/
28
10.3
-
10.4
-
10"
-
exp(-E/RT) dt -
where Eo = mean activation energy (kJ/(g mol), u = standard deviation of E about Eo (kJ/(g mol), V* = ultimate (fractional) weight loss ([initial - final weight]/ initial sample weight), and A = preexponential factor, (8-9. This method involves (1)searching over a range of values for Eo, u, V*, and KO,(2) calculating weight loss at various temperatures by using the integral relationship, (3) comparing calculated and experimental weight loss, and (4) adjusting the four parameters until satisfactory agreement is obtained. A nonlinear least-squares approach was used to perform this part of the study. Applying the model we get the following kinetic parameters: (1) activation energy; (2) preexponential factor; (3) activation energy deviation; (4) ultimate weight loss. Detailed procedures for this analysis are available."Ja
1
l0I "6'9
I
\
L o
1'0
1'1
1'2
1'3
1'4
1'5
1'6
1'7
1'8
1'9
2'0
2'1
1950, No. 4744.
213
Figure 17. Arrhenius plot for various oil shale pyrolysis results. The data sources are described in Table VIII.
These parameters facilitate comparison of global kinetics among shales, their organic fractions, and the preoxidized (21) DiRicco, L. Ph.D. Thesis, University of Colorado, 1955.
(22) Johnson, W. F.; Walton, D. K.; Keller, H. H.; Couch, E. J. Q.Colo. Sch. Mines 70(3), 1975, 237. (23) Braun, R. L.; Rothman, A. J. Fuel 1975, 54, 129. (24) Campbell, J. H.; Koskinas, G. H.; Stout, N. D. Fuel 1978,57,372. Also, see: Campbell, J. H.; Koskinas, G. H.; Stout, N. D.; Coburn, T. T. In Situ 1978,2,1. Also, see: Campbell, J. H.; Burnham, A. K. Lawrence Liuermore Lab. [Rep.],UCRL 1978, UCRL-80545. (25) Shih, S.-M.; Sohn, H. Y. 2nd. Eng. Chem. Process Des. Deu. 1980,
1976, 55, 121.
(18)McCown, M. S.; Harrison, D. P. Fuel 1982, 61, 1149. (19) Maier, C. G.; Zimmerley, S. R. Uniu. Utah Bull. 1924, No. 14, 62. (20) Hubbard, A. B.; Robinson, W. E. Rep. Inuest. U.S.Bur. Mines
2'2
1 0 3 1 ~K"
I
American Chemical Society: Washington, DC, 1981;~93.
376 Energy & Fuels, Vol. 1, No. 4, 1987
Khan TEMPERATURE, T, OC
Table IX. Oxidation Effects on the Pyrolysis Model Parameters for Colorado Oil Shale and Kerogen (Distributed Energy Model) unoxidized oxidized 3 davs Colorado Oil Shale A , s-l 3.4 x 109 8.02 x 109 167.6 E , kJ/k mol) 166.6 15.71 u, kJ/k mol) 4.53 V* (at 600 "C) 0.1866 0.1748 k," s-l 1.876 X 3.788 X lo-*
10
600
550
500
450
300
350
400
PREOXlDlZED UNTREATED l3.DAYl
7A
COLORADO SHALE
0
COLORADO KEROGEN
A
4\
10
\ \
x
Colorado Oil Shale Organics A, s-l 3.38 X 10" 3.21 X 10" E , k J / k mol) 189.4 182.7 u, kJ/k mol) 12.65 23.61 V* (at 600 "C) 0.861 0.6374 k," s-l 5.37 x 10-2 1.446 X lo-'
k
i?
10'
z
8w U
z
0
OAt 500 "C.
2
1 0
W
derivatives. Such results are significant to process design, raw material handling, and pretreatment schemes. Application of a model derived for untreated coal may be a simplification for oxidized samples whose structures are obviously more complicated than the untreated shales. Nevertheless, the model may serve as a practical guide to compare further the influence of preoxidation on the shale decomposition kinetics. The results of the analyses of data collected in this study for unoxidized materials are summarized in Figure 17 and compared with the available literature data for various shales. The origin of shales (their grades) and the legends utilized in Figure 17 are described in Table VIII. Figure 17 suggests that the literature data on untreated samples agree reasonably well with the rate constants obtained from the present study, using the model. A comparison of reaction rate constants of untreated and preoxidized oil shale samples is presented in Figure 18. Table IX also summarizes the model parameters for the untreated and preoxidized shale and their organic fractions. The results demonstrate that the preexponential factors, A , and the mean activation energies, E,, are only slightly altered upon preoxidation. However, the rate constant, k , is roughly twice as great for the preoxidized samples (shale or the shale organic fractions) compared to the corresponding untreated samples. The increased rate constant for the pyrolysis of preoxidized sample may reflect the observation that oxidation introduces certain oxygen functional groups into the structure that break at a rapid rate compared to the functional groups present in the original structure. This observation is in line with the example that C-0 bond strength is 83 kcal/mol compared to 103 kcal/mol for the C-H bond. Additional bond 78 and R-OH strengths to consider include R-0-R 91 kcal/mol. Presence of these functional groups in preoxidized shales would facilitate greater decomposition of
J
s 10'
10' 1 1
12
1 3
14
15
16
1 7
'
1 O W K"
Figure 18. Arrhenius plot of reaction rate constants of the untreated and preoxidized shale and kerogen samples.
preoxidized samples compared to the unoxidized materials at a relatively lower temperature.
Diu. Pet. Chem. 1984, 29(3),127.
Conclusions On the basis of the results of this study, it can be concluded that preoxidation of oil shale/shale organic components can reduce significantly the weight loss (and the maximum rate of weight loss) when samples are subsequently devolatilized. The FTIR results demonstrate that oxidation introduces oxygen cross-linking bridges in the organic structures, and thus "ties-up" functional groups that evolve as H 2 0 and C02 rather than desirable hydrocarbons. Such cross-link reactions may also cause the reduced fluid behavior of the organic fractions upon preoxidation (reflected by reduced volume contraction for the preoxidized samples compared to that of the untreated samples during pyrolysis). Low-temperature (50 OC, 95 days) preoxidation of shales and coal deteriorates the quality of feedstocks and produces liquid, upon pyrolysis, or poorer quality. Weathering of shale or ita organic fractions influences pyrolysis kinetic parameters. Finally, the results suggest that preoxidation effects of oil shale and shale organic fractions are qualitatively similar to those found with coal. The results of this study stress the importance of proper storage/handling of oil shale materials to minimize their oxidative degradation, keeping in mind that various sophisticated characterization techniques require finely ground shales. Heat generated during many grinding operations (which is generally performed in air; hand ground under N2 in this study) may further accelerate the oxygen chemisorption and degradation process.
24, 265.
Acknowledgment. The author thanks K. Seshadri and G. Wang for their input in various aspects of this work.
-
-
(28)Ratam-Abadi,M.Ph.D.Thesis, Wayne Stab University,Detroit, MI., 1982. .~ (29) Rajeshwar, K.;Dubow, J. Thermochim. Acta 1984,54, 71. (30)Wang, C. C.;Noble, R. D. Fuel 1983, 62, 529. (31)Joshi, R.; Lee, S.Liquid Fuels Technology 1983, 1 , 17. (32)Elder,J. P.;Ready, V. 1983 Eastern Oil Shule Symposium; 1983; ~~
~
nn 255-262 r r ---
(33)Shen, M.S.;Fan,L. S.;Casleton, K. Prepr.-Am. Chem. SOC., (34)Yang,H.S.;Sohn,H.Y.Fuel 1984,63, 1511. (35)Sohn,H.Y.;Yang.H.S.I d .Eng. Chem. Process Des. Deu. 1986,
