Energy & Fuels 1991,5,322-327
322
onstrated that the Ca catalyst was initially active but later deactivated rapidly, while the activity of nickel catalyst increased to give a maximum during gasification. Some workers have already noted that the activity of Ca is sensitive to the initial dispersion of Ca on carbon but a rapid loss in activity of Ca during gasification cannot be totally explained by the sintering of calcium parti~1es.I~ A similar problem to this has been pointed out for the catalyst deactivation observed in Ni-catalyzed lower temperature gasification at 797 K ash process performance >797 K conversion, wt % maf feed coal conversion, wt % maf coal H2 consumption, scf/barrel a
0.08 1.8 6.3 26.8 36.3 26.6 0.7a
0.03 2.5 8.4 24.6 30.5 27.0 2.6
56.9
54.5 84.8 10.0
6.3
Ash due to catalyst attrition.
I
I
Analyses
Bitlfmin Analyses
1
I Catalytic Processing
solutility Classes
t
t
Gas
t
Hea y Product
Light Product
'-
Syncrude Blends
1
Distillation
+
t
47lx616K
IBPx47lK
Aliphaltic Aromatic Polar
t
65Qx791K
Aliphaltic
+
,797K
Aromatic Polar
Figure 2. Schematic of analytical workup of samples. of each distillate cut was determined by simulated distillation using the Naphtha method and the resid fractions were run using the Crude method of ASTM 2887. Each distillate cut was analyzed in quadruplicate. The final yields were corrected for overlap, if any. Selected distillate fractions were separated by preparative liquid chromatography into an aliphatic, aromatic, and polar fraction. A 3 ft by 1 in. stainless steel column was packed with dry silica gel (230-400 mesh) and approximately 5 g of sample was applied to the column. The aliphatics were separated by using hexane as the mobile phase, and aromatics were separated by using a 15 vol % benzene in hexane mixture. The column was back-flushed by using T H F to obtain the polar fraction. The separation was monitored by thin layer chromatography using authentic coalderived fractions from a Wilsonville coal liquid as standards. The solvents were removed by rotary evaporation. Recoveries were in the range of 98-99 wt %. The reproducibility of the method is as follows: aliphatics (il.O%), aromatics (i0.7%), and polars (=k1.6 % ).
In addition, selected distillate fractions were separated into the compound classes given above by using an analytical liquid chromatographic method with FID detection to check the purity of the preparative fraction. The method is described in detail el~ewhere.~ Elemental, proximate, ultimate, and petrographic analyses were determined by using standard ASTM procedures. The Soxhlet method utilized to determine the solubility class distribution, based upon pentane, benzene and pyridine solvents, of the feeds and products is described in detail elsewhere! (4) Chawla, B.;Davis, B . H . Fuel Sci. Technol. Znt. 1989, 7(1), 1-14.
(5) Keogh, R.A.; Davis, B. H.J. Coal Quality 1988, 7(1), 27-31.
Carbon Zsotope Analysis Isotopic Analysis. The 13C/12C ratios of the feeds and products were obtained from the following laboratories: (a) University of Alberta (Dr. K. Meuhlenbachs),and (b) Coastal Science Laboratories. Each of the laboratories determined the carbon isotope ratios by quantitatively converting the carbon of the sample to COz and measuring the relative amounts of the different carbon isotopes. The resulting ratios, corrected for oxygen isotopes, are compared to that of a standard material, a Peedee belemnite (PDB).6 All of the carbon isotope ratios are reported relative to the PDB standard in parts per thousand (460) in the following manner:
Results and Discussion Three criteria need to be addressed to obtain accurate results by using the carbon isotope technique. These will be discussed below. The first of the criteria is that the 6I3C ratios of the two feedstocks be substantially different. The 6I3C ratios of a number of coal seams from eastern and western Kentucky and heavy oils were determined by the University of Alberta. The results are given in Table I. As can be seen, there was no significant difference in the 6I3C value of the coals. The average 6% value for all the coals examined was -23.73 f 0.37%. The 613C values of the candidate liquid feedstocks are also given in Table I. The data indicate that there is a significant difference between the 613Cvalues of the coals and some of the heavy oil feedstocks (ca. 70 times the reported experimental error1 of *0.07L). A Western Kentucky No. 9 coal (91864) and the Kentucky tar sand bitumen were selected as feedstocks. The difference in the 613Cvalues of the two feeds is 4.486 which is large when compared to the reported reproducibility of the method.' The second criterion is that the feedstocks must be isotopically homogeneous. The 613C values of the coals given in Table I indicate that there is no relationship between the 6I3C value and rank, as defined by the reflectance. This has been previously observed by Colombo.' The candidate coals also have different petrographic compositions. There appears to be no relationship between the petrographic composition of the coals and their respective 6I3C values. This observation is consistent with what has been observed in the literature. Rigby and cow o r k e r ~ determined ~*~ the 613C values of hand-picked macerals and found no difference in their values. Although the pyridine-soluble fraction of the selected Western Kentucky No. 9 coal (91864)was not available for analysis, it has been reported that no significant difference in the 6I3C values of the parent coal and the pyridine extract was observed.' These data indicate that the coal is homogeneous with respect to the 613C ratios. The bitumen feedstock was separated into fractions by solubility and by distillation to determine whether the bitumen was isotopically homogeneous. The 613C ratio of the distillate and solubility fractions of the bitumen are given in Table V. The data indicate that the bitumen feedstock is homogeneous with respect to the carbon isotopes. The average 613Cvalues for all samples is -28.95 f 0.18% which is essentially equal to the 6I3C ratio of the parent bitumen, -28.81%~ Although only the parent bi( 6 ) Craig, H.Geochim. Cosmochim. Acta 1957,43, 1979-1988. (7) Colombo, V.; Gazzarine, F.; Genhantini, R.; Kneuper, G.; Teichmtiller, M.;Teichmilller, R. 2. Angew. Geol. 1968, 14, 257-265. (8) Rigby, D.; Batta, B. D.; Smith, J. W. Org. Geochem. 1981,3, s 3 6 . (9) Smith, J. W.; Gould, K. W.;Rigby, D. Org. Geochem. 1981, 3,
111-131.
Energy & Fuels, Vol. 5, No. 2, 1991 325 Table V. Carbon Isotope Ratios of the Tar Sand Bitumen Fractions
carbon isotope ratio, per mil (960) University Coastal Science of Alberta Laboratories fraction parent solubility class oils asphaltenes preasphaltenes distillation
IBP X 616 K 616 X 797 K >797 K av of all fractions
-28.81
-29.8
-28.84 -28.93 -28.70
-29.8
-29.09 -29.14 -29.17 -28.95 f 0.18
Table VI. Carbon Isotope Ratios of the Bitumen Only Processing Period carbon isotope ratios, per mil (%) University Coastal Science of Alberta Laboratories bitumen -28.81 -29.8 separator products light product (V-451) -29.44 heavy product (V-401) -28.84 heavy product solubility classes oils -28.66 asphaltenes -28.71 preasphaltenes -29.03 gas CH4 -39.1 cis -32.7 cis a c4+ -31.3
co + coz
syncrude distillate fractions IBP X 477 K 477 X 616 K 616 X 797 K >797 K liquid chromatographic fractions middle distillate (477 X 616 K) aliphatics aromatics polars heavy distillate (616 X 797 K) aliphatics aromatics polars average of all samples (excluding gas)
a
-29.28 -28.65 -28.76 -28.70
-29.7 -29.4 Q
-28.92 f 0.30
-29.5 -29.4 -29.0 -29.40 f 0.25
Insufficient sample for analysis. tumen and the oil fraction were analyzed by the Coastal Science Laboratories, the same conclusion can be made. Therefore, the bitumen is homogeneous with respect to the carbon isotopes. Although the data from the two laboratories result in the same conclusion, there is a consistent difference of approximately 1%between the 613C determination of the samples of Coastal Science Laboratories and the University of Alberta, shown in Table V. This difference has been the subject of much discussion; however, the important observation is not the absolute value, but that the difference in the 613Cvalues is consistent for all the samples determined at a laboratory. Therefore, the difference should not affect the calculation of the contribution of each feedstock to the final product slate. Another criterion which has to be addressed is that the processing should not cause any significant fractionation
Keogh et al.
326 Energy & Fuels, Vol. 5, No. 2, 1991 of the isotope content. The bitumen processing products and the fractions of the products obtained by solubility, distillation, and liquid chromatography were analyzed to determine if processing resulted in any fractionation of the carbon isotopes. The 613C data obtained from the laboratories for these samples are given in Table VI. The average 613C ratio for all the liquid samples analyzed by the University of Alberta is -28.92 f 0.30%; this compares quite well with the 6lF value of the bitumen (-28.81). The average obtained from the Coastal Science data is -29.40 f 0.25 which also compares well to their value for the feed bitumen (-29.8). The major source of variance in the average 613C values was derived from the naphtha fraction (IBP X 477 K). This distillate fraction is isotopically lighter than the heavier distillate fractions. Therefore, it appears that there is a modest amount of fractionation caused by the conversions during processing. As expected, the product gases showed the greatest degree of fractionation. The gaseous components are significantly lighter than the bitumen feedstock as predicted by isotopic fractionation theory.1° The effect of this fractionation on the calculations is discussed below. The 613C values from the aromatic, aliphatic, and polar fractions of the middle (477 X 616 K) and heavy distillates (616 X 797 K) show no significant fractionation when compared to the parent distillate fraction or the parent bitumen. However, there is a small trend observed in the liquid chromatographic fractions; this trend is similar to that reported for petroleum crude oils.l0 For the heavy distillate, the aliphatic fraction is isotopically the lightest (-29.5%) and the aromatic fraction has an intermediate value (-29.4%). This is the same trend observed by Stahl" for crude oils and their extracts." The same trend in the data is observed for the aliphatic and aromatic fractions of the middle distillate (477-616 K). The isotopic data obtained for these samples indicate (a) there is a sufficient difference between the 613C ratios of the coal and bitumen (ca. 4.5760) feedstocks, (b) both feedstocks are isotopically homogeneous, and (c) there is a modest isotopic fraction observed due to the process. The calculation of the amount of coal incorporated into the coprocessing product fractions was accomplished by using three methods (designated method A, method B, and method C). Method A' calculations utilize both the carbon isotope ratios from the bitumen processing period (Table VI) and the coprocessing period (Table VII). In this method, a correction factor is used to account for any isotope fractionation due to processing. The correction factor is derived from the difference between the 6% ratio of the parent bitumen and the 613C ratio of the resulting product fractions of the bitumen processing. These correction factors are added to the 613C ratios of the coprocessing product fractions. The amount of coal carbon incorporated into the coprocessing fractions was calculated by using the following equation: % carbon(coa1) = (613C*(coprocessing)- 613C(coal)) 1x 100% (2) (b13C(bitumen)- 613C(coal)) where 613C*(coprocessing)is the corrected 6 W ratio of the coprocessing product fractions and 613C(bitumen) and 6l3C(coal)are the carbon isotope ratios of the bitumen and coal feedstock, respectively. The carbon(coa1) as the percent of product fraction is calculated by multiplying the carbon content of the fraction by the percent carbon(10) Sackett, W. M. Geochim. Cosmochim. Acta 1978,42, 571-580. (11)Stahl, W. Geochim. Cosmochim. Acta 1978,42,1573-1577.
Table VU. Carbon Isotope Ratios of the Coprocessing Period carbon isotope ratios, per mil (L) University Coastal Science of Alberta Laboratories feedstocks -28.81 -29.8 bitumen -24.33 coal -28.48 coal/bitumen slurry slurry oils -29.13 -28.71 aphaltenes -25.46 preasphaltenes IOM -24.23 products separator products light products (V-451) -28.67 heavy products (V-401) -27.87 -28.8, -28.6 heavy products -27.81 oils -26.77 asphaltenes -26.75 preasphaltenes -25.53 IOM syncrude distillation IBP X 477 K -28.72 -28.24 477 X 616 K 616 X 797 K -27.93 -27.10 >797 K LC fractions (477 X 616 K) -29.4 aliphatics aromatics -28.3 -26.8, -26.9 polars LC fractions (616 X 797 K) aliphatics -29.2 aromatics -28.3 -27.1, -27.2 polars gas -37.1 CH4 -31.7 cz -29.2 c3 -32.5 c4 a coz, co
Insufficient sample for analysis.
(coal) as percent of the carbon of the fraction. Method B uses the same mass balance technique; however, the calculation assumes that little, if any, fractionation occurs during processing. The method calculates the percent coal carbon as the percent of carbon using the following equation 9'0 carbon(coa1) = 613C(bitumen)- 6W(coprocessing) x 100% (3) 613C(bitumen)- 613C(coal) where the terms of the equation are the same as described above with the exception that the 613C ratios of the coprocessing product fractions do not include correction factors. The carbon(coa1) as the percent of the fraction is calculated as described previously. The calculations of method C12utilize the carbon isotope ratios of the individual bitumen-only processing product fractions to account for isotopic fractionation due to processing. The method uses the following equation to calculate the carbon(coa1) as the percent of carbon: X(613C(bitumenprocessing product fraction)) + (1X)613C(coal)= 613C(coprocessingproduct fraction) (4) where X and (1- X)are the fractions of bitumen and coal, (12)Steer, J.; Muehlenbachs, K.; Ohuchi, T.; Careon, D.; Doherty, B.; Ignasiak, B. Paper presented at the 14th Biennial Lignite Symposium on the Technology and Utilization of Low-RankCoals, 1987, D a l h , TX.
Energy & Fuels 1991,5, 327-332 Table VIII. Carbon(coa1) as Percent of Carbon of the CoDrocessina Distillate Fractions IBP X 477 X 616 X calcn method 477 K 616 K 797 K >797 K A 12.5 9.2 18.5 35.7 B 2.0 12.7 19.6 38.2 C 11.3 9.5 18.7 36.6 av of methods 8.6 10.5 18.9 36.8 SD 5.7 1.9 0.6 1.3 Table IX. Carbon(coa1) as Percent of Product Fraction IBP X 477 x 616 X calcn method 477 K 616 K 797 K >797 K 8.2 16.4 28.8 A 10.8 B 1.7 11.3 17.4 38.2 C 9.8 8.4 16.6 29.6
respectively, and the PC(bitumen processing product fraction) and the 613C(coprocessingproduct fraction) are the carbon isotope ratios of the product fractions derived from the bitumen-only and the coprocessing mode of operation. The percent carbon(coa1) as the percent of the fraction is calculated in the same manner as described above. The results of the three methods of calculating the carbon(coal) as the percent of carbon in the coprocessing distillate fractions are given in Table VIII. Comparison of the results calculated from the three methods indicates that the two methods (method A and method C) which attempt to correct the coprocessing 6I3C ratios for the isotopic fractionation caused by the processing result in similar values of coal-carbon incorporation into the distillate fractions. All values calculated by method A and method C agree within 5% and indicate an increasing coal contribution to the distillate fractions as the distillate range increases. In addition, the resulting carbon(coa1) incorporation values calculated with the assumption that no isotopic fractionation occurred during processing (method
327
B)are consistently higher in the calculated carbon(coal) as percent of carbon in the middle, heavy, and resid distillate fractions. From the carbon(coal) as a percent of total carbon and the elemental carbon data, the percent of the carbon(coal) in the product fraction can be calculated. The data derived by using the three methods of calculations are given in Table IX. The same trends are observed for these data as discussed above.
Conclusions The stable carbon isotope technique was utilized to determine the contributions of the coal and the tar sand bitumen to the coprocessing product slate. The experimental design of the operation of the pilot plant indicated that fractionation of the carbon isotopes occurred during processing and must be accounted for in the mass balance calculations. The greatest amount of fractionation occurred in the naphtha fraction and the product gases. The two methods which accounted for fractionation produced the same results, within experimental error, for the coal contribution to the coprocessing product slate. The method which did not take into account the fractionation produced results which consistently indicated larger percentages of the coal contributions to the distillate fractions in which little fractionation was found and smaller coal contributions in the distillate fraction in which the largest fractionation of the carbon isotope was observed. The results of the mass balance calculations which account for the isotopic fraction indicate that the contribution of the coal during coprocessing was greatest in the heavy distillate and resid fraction. Acknowledgment. This work was supported by the Commonwealth of Kentucky and DOE Contract No. DEFC22-88PC8806as part of the Consortium for Fossil Fuel Liquefaction Science (administered by the University of Kentucky).
Ionic Hydrogenation of Organosulfur Compounds Mirjana Eckert-Maksi6* and Davor MargetiE Department of Organic Chemistry and Biochemistry, Rudjer BoBkoviE Institute, Zagreb, Croatia, Yugoslavia Received July 27, 1990. Revised Manuscript Received December 6, 1990
Ionic hydrogenation of the three most abundant types of organosulfur constituents of coal, aromatic sulfides, aromatic disulfides, and benzo[b] thiophene derivatives, in BF3-H20-Et3SiH is studied. Reduction of aromatic sulfides results in partial saturation of the aromatic moiety and cleavage of the corresponding SR group. Aromatic disulfides undergo quantitative sulfur-sulfur bond cleavage, while benzo[ blthiophene derivatives produce 2,3-dihydrobenzo[b] thiophenes in high yields.
Introduction Selective hydrogenation of multiring aromatic sulfur compounb found in the fossil fieh k one of the important steps not only for their upgrading but also in producing useful intermediates for various molecules. Particularly intriguing in this respect is the method of ionic hydrogenation (IH).l This method owes its high selectivity to
the fact that the direction of the hydride ion attack depends exclusively on the structure of the carbonium ion formed through the interaction with the acidic (proton donor) component.2 With the proper choice of a hydro(1) Cheng, J. C.; Maioriello, J.; Lamen, J. W. Energy Fuels 1989, 3, 321-329.
0887-0624/91/2505-0327$02.50/00 1991 American Chemical Society
