Energy & Fuels 1988,2, 639-644 a loss of active Fe by sintering and agglomeration.
Acknowledgment. Part of this work was carried out under the "Japan/Canada Joint Research ProgramDeveloping Advanced Process for the Efficient Use of
639
Coal", sponsored by the Ministry of Education, Science and Culture, Japan. Registry No. FeO, 1345-25-1;Fe304,1317-61-9;Fe,Os, 130937-1; Fe, 7439-89-6; Fe&, 12011-67-5;COz, 124-38-9.
Performance of MoC1,-LiCl-KCl and NiCl2-LiC1-KCl Catalysts in Coal Hydroliquefaction with a Hydrogen Donor Vehicle Chunshan Song, Kouji Hanaoka, and Masakatsu Nomura* Department of Applied Chemistry, Faculty of Engineering, Osaka University, 2-1 Yamada-oka, Suita, Osaka 565, Japan Received February 29, 1988. Revised Manuscript Received June 2, 1988
The catalytic performance of MoC1,-LiCl-KC1 and NiC12-LiC1-KC1 salts in coal hydroliquefaction with the H donor tetralin was studied. Yilan subbituminous coal impregnated with the above catalysts and with LiC1-KC1 at different loading levels (weight ratios of salts to coal of 0.1-1.0) was hydroliquefied at 400 OC for 1h with an initial Hzpressure of 4.9 MPa. Increased loading of the MoCl, salt increased oil yield (up to 50 wt %), coal conversion, and H2 consumption, decreased the net hydrogen transfer from tetralin, and lowered the heteroatom content of oil. Addition of the NiC12 salt below a weight ratio of 0.2 led to more pronounced increases in oil yield, coal conversion, and H2 consumption and decreases in net hydrogen transfer from tetralin than with the MoCl, system. Higher loadings (>0.2) led to a small decrease in H2 consumption in spite of a further increase in oil yield (up to 50 w t %), implicating contributions of in situ formed H donors. The use of a LiC1-KC1 mixture slightly improved coal conversion while considerably suppressing H2 consumption. The relation between coal conversion and hydrogen consumption suggests that the action of the MoCl, and NiClz catalysts led to enhanced but selective hydrocracking reactions that contribute mainly to oil production. 'H NMR analysis of the oil revealed that hydroaromatic compounds increased and the average ring sizes of aromatics decreased with increased loading of the catalysts.
Introduction Numerous studies have been conducted of the application or improvement of existing catalysts and of the development of new catalysts for coal liquefaction.'-'l The (1)(a) Donath, E. E. Catalysis: Science and Technology; Anderson, J. R., Boudart, M., Eds.; Springer Verlag: West Berlin, 1982;Vol. 3,pp 1-37. (b) Weller, 5. W. Presented at The 4th International Conference on the Chemistry and Uses of Molybdenum, Golden, CO, August 9-13, 1982. (c) Mochida, I. Shokubai Koza; Catalysis Society of Japan, Ed.; Koddansha: Tokyo, 1985;Vol. 9,pp 150-167. (2)(a) Lee, E. S. Coal Conversion Technology;Wen, C. Y., Lee, E. S., Eds.; Addison-Wesley: Reading, MA, 1979; pp 489-538. (b) Polinski, L. M.; Rao, V. U. S.; Stencel, J. M. The Science and Technology of Coal and Coal Utilization; Cooper, B. R., Ellingson, W. A., Eds.; Plenum: New York, 1984;pp 455-482. (3)(a) Tanabe, K.; Hattori, H. J. Jpn Pet. Inst. 1986,29,280-288.(b) Kamiya, Y.J. Fuel SOC.Jpn. 1979,58, 2-9. (4)(a) Kikkawa, S.;Nomura, M. Shokubai 1980,22,79-87. (b)Nomura, M.; Miyake, M. J. High Temp. SOC.(Suita,Jpn.) 1982,8,138-148. ( 5 ) Nomura, M.; Kimura, K.; Kikkawa, S. Fuel 1982,61,1119-1123. (6)Nomura, M.; Ikkaku, Y.; Nakatsuji, Y.; Kikkawa, S. J . Jpn. Pet. Imt. 1983,26,303-308. (7) Nomura, M.; Yoshida, T.; Morida, Z. Ind. Eng. Chem. Prod. Res. Deu. 1984,23,215-219. (8)Zielke, C. W.; Struck, R. T.; Evans, J. M.; Costanza, P.; Gorin, E. Ind. Eng. Chem. l'rocess Des. Dev. 1966,5,58-164. (9)Larsen, J. W.; Earnest, S. Fuel Process. Technol. 1979,2,123-130. (10)Grens, E. A., II; Hershkowitz, F.; Holten, R. R.; Shinn, J. H.; Vermeulen, T. Ind. Eng. Chem. Process. Des. Deu. 1980,19,396-401.
0887-0624/88/2502-0639$01.50/0
known catalytic materials14 include (1)metal oxides, (2) metal sulfides, (3) metal halides (molten salts), (4) metals and molten metals, etc. Two new coal liquefaction catalysts, MoC13-LiC1-KC1 and NiC12-LiC1-KC1 ternary salts, are effective hydrocracking catalysts for hydroliquefaction of a subbituminous coal without a vehicle.12 Differential thermal analysis indicated that these ternary salts exist in the molten state at a coal liquefaction temperature of 400 OC. The catalytic activities for coal conversion and selectivity for oil formation of these MoC13- and NiClZcontaining salts are higher than those of the ZnC1,- and SnCl,-containing ones. ZnC12, SnC12,and their mixtures with alkali-metal chlorides are effective coal liquefaction catalysts."" Examination of these MoC13- and NiClZcontaining salts as catalysts for solvent-free hydroliquefaction of other brown, subbituminous, and bituminous coals also confirmed their high catalytic effectiveness for hydrocracking these coals and their high selectivity for production of oil (hexane ~ o l u b l e ) . ~ ~TemperatureJ~ programmed pyrolysis experiments revealed that these (11)Tanner, K.I.; Bell, A. T. Fuel 1981,60, 52-58. (12)Song, C.; Nomura, M.; Miyake, M. Fuel 1986,65,922-926. (13)Song, C.; Nomura, M. Bull. Chem. SOC.Jpn. 1986,59,3643-3648. (14)Song, C.; Nomura, M. Chem. SOC.Jpn.--Prepr. Symp. Pet. Chem. 1987,17,122-125.
0 1988 American Chemical Society
Song et al.
640 Energy & Fuels, Vol. 2, No. 5, 1988
VMd 46.7
Table I. Remesentathe Analysis (wt % ) of Yilan Coal ultimate (daf) solubility (daf) moisture C H N 0 + sa hexane benzene 6.9 79.4 5.8 1.4 13.4 1.0 2.3b
proximate FC' ash 44.0 2.4
By difference. Hexane insolubles. Benzene-insolubles. dVolatile matter. 100
1
F
Asphal tene
20
1
Fixed carbon.
IC r
A
Q
2
e
t
I
Oi1
0
0 8
0
0.2
0.4
.
I
0.6
THF 6.2c
.
I
0.8
.
I
I
1.0
0
0.2
0.4
0.6
0.8
.
,
1.0
Weight r a t i o o f s a l t s t o c o a l
Figure 1. Influence of loading level of MoCl3-LiC1-KC1 (A), NiCl2-LiC1-KC1 (B), and LiC1-KCl (C) on coal conversion.
MoC13- and NiC1,-containing salts even promote coal cracking at temperatures in the 300-400 OC range under a Nz atmo~phere.'~J~ Furthermore, the addition of the hydrogen donor tetralin to the catalytic reaction system increased oil yields and coal conversion but reduced the total hydrogen consumption.16 The performance of these new catalysts for coal hydroliquefaction with tetralin was investigated, and the results are reported here. Generally, salt (molten salt) catalysts are used in massive amounts (weight ratio of salt to coal 1 l).4-'5 The hydroliquefaction of a coal impregnated with different amounts of MoCl3-LiC1-KC1 and NiC1,-LiCl-KCl as well as LiC1-KC1 salts (weight ratios of salt to coal of 0.1,0.2, 0.5, and 1.0) was carried out with tetralin present. The catalytic performance of MoCl,-LiCl-KCl and NiClz-LiC1-KC1 in coal hydroliquefaction is reported, including the influence of catalyst loading level on coal conversion, product distribution, and routes and extents of hydrogen transfer. A 'H NMR analytical method13 was used to follow the structural changes of the products, reflecting the catalytic action of the salt. The relationship between product yields and total hydrogen consumption in these catalytic reactions is also discussed. Experimental Section Materials and Reagents. A Chinese subbituminous coal (Yilan) was pulverized to C200 mesh and stored under a N2 atmosphere. Analytical data and solubilites (Soxhlet extraction) of the coal are given in Table I. The coal was dried a t 105 "C in vacuo for 1.5 h before use. MoC13was synthesized and purified according to the procedures described previously>2 Reagent grade NiC12,LCl, and KC1 were used. Catalysts examined in this work were MoCl3-LiC1-KC1 (molar composition, 12:51:37; melting temperature, 387 "C) and NiCl2-LiC1-KC1 (145036; 360 "C) ternary sdta and LiCl-KCl(5842; 353 "C). Reagent grade tatralin was distilled under reduced pressure. Reagent grade n-hexane, benzene, and T H F were used as received. Hydroliquefaction Procedures. Hydroliquefaction reactions were carried out in a SUS-316 70-mL rocking autoclave (68 (15)Song, C.;Nomura, M.; Fuel 1987,66,1225-1229.
strokes/min). The coal (5 g) was impregnated with catalyst by immersing the coal in a methanol solution or suspension of catalyst followed by complete removal of the methanol and subsequent drying a t 105 "C in vacuo for 6 h. Impregnated coal and tetralin (5 g) were transferred into the autoclave. The autoclave was then pressurized to 4.9 MPa (50 kg cm-2) with H2 at room temperature and heated at a rate of 7 "C m i d to 400 "C and held at that temperature for 1 h. After the reaction, the autoclave was quenched to room temperature in an air blast. Product Separation. The gaseous products were collected at room temperature and analyzed quantitatively by gas chromatography. The contents of the autoclave were washed out with excess hexane (loo0 mL),and filtered. The fiiter cake was Soxhlet extracted sequentially with hexane, benzene, and THF. The hexane-soluble liquid fraction is oil (HS). The tetralin and naphthalene in this fraction were removed by vacuum distillation (up to 115 "C, 8 mmHg) after the weight ratio of tetralin/ naphthalene had been determined by gas chromatography. The hexane-insoluble but benzene-soluble fraction, benzene-insoluble but THF-soluble fraction and THF-insoluble fraction are denoted as asphaltene (BS), preasphaltene (THFS), and residue, respectively. The amount of H2 consumed was determined by P-V-T and gas chromatographic methods. The amount of hydrogen transferred from tetralin was determined from the tetralin:naphthalene ratio. Yields of products, H2 consumption and net hydrogen transfer (apparent hydrogen transfer) from tetralin were calculated on a daf coal basis. Some experiments (5 runs) were repeated and the deviation in product yields and total hydrogen consumption were within i 2 and i0.2 w t % , respectively. Analysis of Products. The oil and asphaltene products were subjected to elemental and 'H NMR analyses. 'H NMR spectra were measured in CDCl, solution with MelSi as an internal standard on a JNM-PS-100 spectrometer a t 100 MHz. D20 was added to the solution to identify the phenolic OH, whose contribution was subtracted. The total NMR signal was divided into several parts and assignedi3 to protons in different chemical environments, as shown in Table 11. The distribution of hydrogen atoms was calculated from the 'H NMR and elemental analyses and was divided into three groups: HCH*(HF + H,1+ H B ~ & ) ,H ~ (Hu2 H, H,) and H,. The average structural parameters of oils and asphalteneswere calculated by using the Brown-Ladner equati~ns.'~J~
+
+
(16)Brown, J. K.;Ladner, W.R. Fuel 1960,39,87-95.
p-
Catalysis i n Coal Hydroliquefaction I
A
O F r m H2 0 From t e t r a l i n
Energy & Fuels, Vol. 2, No. 5, 1988 641
L 1
0
0.2
0.4
0.6
0.8
1.0
0
0.2 0.4 0.6 0 . 8 1.0 Weight r a t i o o f s a l t s t o coal
0
.
1
.
0.2
1
,
0.4
I
.
I
0.8
0.6
I
f
1.0
Figure 2. Influence of loading level of MoCl3-LiC1-KC1 (A), NiC12-LiC1-KC1 (B),and LiC1-KC1 (C) on hydrogen transfer. Table 11. Assignments of Bands in IH NMR Spectra symbol chem shifta assignts of hydrogen type aromatics H, 9.20-6.00 CH2 CY to two aromatic rings HF 5.00-3.40 Hal 3.40-2.35 CH2 CY to an aromatic ring H2. 2.35-2.00 CH3 CY to an aromatic ring CH2 j3 to an aromatic ring Hgl 2.00-1.60 1.60-1.05 CH3 j3 to an aromatic ring H, CH2 y or further from an aromatic ring H, 1.05-0.20 CH3 y or further from an aromatic ring "In ppm from TMS.
Results and Discussion Effects of Catalysts on Coal Conversion. Figure 1 presents the yields of products as a function of the loading level (weight ratio of salt to coal) of MoCl,-LiCl-KCl (A), NiC12-LiC1-KC1 (B) and LiC1-KC1 (C), respectively. The amounts of Mo and Ni loaded on coal (metal basis) ranged from 1.7 to 17 wt % and from 1.2 to 12 wt %, respectively. Without a catalyst, yields of gas, oil, asphaltene, and preasphaltene were as low as 4.4, 24.3, 18.9, and 10.9 wt % , respectively. When the MoCl,-LiCl-KCl catalyst was used (Figure lA), the oil yield increased sharply with loading levels up to 0.2, rose gradually with loading levels between 0.2 and 0.5, and rose rapidly with loading levels higher than 0.5. Yields of gas, asphaltene, and preasphaltene increased by about 1,6, and 3 wt % ,respectively, a t the loading level of 0.1, relative to a noncatalytic run. For loading levels of 0.1-1.0, yields of gas and asphaltene did not change notably, but preasphaltene yields gradually decreased with increased loading level. The MoCl, catalyst exhibited some proportionality between its loading and catalytic effects on coal conversion. For the NiC12-LiC1KC1 catalyst (Figure lB), the oil yield increased steadily with loading levels up to 1.0, with a sharp increase a t the loading level of 0.2, where a fairly good oil yield (44.6 wt % ) was obtained. A small amount of the NiC12 catalyst shows high catalytic effects, while alkali-metal chlorides (LiC1-KC1) alone enhanced coal conversion only slightly at all loading levels (Figure IC). The catalytic activity of the MoCl,-LiCl-KCl and NiCl2-LiC1-KC1 systems can be attributed mainly to the action of some species of molyb-
denum and nickel. Nevertheless, the LiC1-KC1 salt did exert some influence upon the reaction (see below). Comparison of MoCl, (Figure 1A) and NiC12 (Figure 1B) catalysts reveals that at loading levels below 0.5, the later afforded much higher oil yields and somewhat higher asphaltene and preasphaltene yields than the former. Liquefaction reactions in the presence of both MoCl, and NiC1, catalysts were carried out to check for the synergistic effects. Their equivalent-weight mixture, MoC13-NiC12LiC1-KC1 gave medium oil yields and conversions between those obtained with MoCl, and NiClz catalysts at loading levels below 0.5. These MoCl,-LiCl-KCl and/or NiC12LiC1-KC1 salts tend to give similar oil yields at a loading level of 1.0 (the common loading level of salt catalysts); see Table 111. Good selectivity of the MoCl, and NiC12 catalysts for oil production can be seen clearly from the fact that oil yields (49-50 wt %) from the catalytic runs (runs 10,14,18) are almost twice that of the noncatalytic run (run 1) or the run with LiCl-KC1 (run 6). Another desirable feature of these catalysts is that they did not enhance gas formation to any considerable extent. Effects of Catalyst on Hydrogen Transfer. The amounts of hydrogen consumed from gas-phase H2 and tetralin and their sum are presented in Figure 2 as a function of the loading level of MoCl,-LiCl-KCl (A), NiC12-LiC1-KC1 (B) and LiC1-KC1 (C). In the noncatalytic run, Hzconsumption and net hydrogen transfer from tetralin are 0.6 and 1.1 wt %, respectively. A second noncatalytic run gave almost the same values. An increase in the loading level of MoCl,-LiCl-KCl up to 1.0 (Figure 2A) resulted in a slow increase in H2 consumption from 0.6 to 2.5 wt % and a progressive decrease of net hydrogen transfer from tetralin (from 1.1 to 0.4 wt %). Use of NiClZ-LiC1-KC1 (Figure 2B) up to the loading level of 0.2 greatly increased consumption of H2 (from 0.6 to 2.4 wt % ) and caused a sharp decrease in the net hydrogen transfer from tetralin (from 1.1to 0.2 wt %). Within this range of loading levels, the amounts of Hz consumed are unambiguously higher, and the amounts of hydrogen transferred from tetralin are lower with the NiC1, catalyst than with the MoCl, catalyst. However, higher loading levels (>0.2) of NiC12-LiC1-KC1 (Figure 2B) resulted in
Table 111. Catalytic Hydroliquefaction of Yilan Coal with Tetralin Vehicle
run no. 1
6 10 14 18 a
catalystn none LiCl-KCl MoCl3-LiC1-KCl NiC1,-LiCl-KC1 MoCl3-NiCl2-LiC1-KC1
At the loading level of 1.0.
CO, 3.0 4.2 2.7 3.5 3.1
yields of products, w t % (daf) CrCr oil asph preasph 1.4 24.3 18.9 10.9 28.6 20.5 2.0 10.5 50.1 2.7 24.2 6.6 50.0 26.4 15.4 2.4 49.3 25.5 8.3 2.8
Sum of the yields of producta.
conv,b % 58.5 65.8 86.3 97.7 89.0
H consumed, w t % (daf) Hz tetralin tot. 0.6 1.1 1.7 0.1 1.2 1.3 2.5 0.4 2.9 2.2 0.2 2.4 2.3 0.4 2.7
642 Energy & Fuels, Vol. 2, No. 5, 1988
a small decrease in Hz consumption. This trend is very interesting, since the oil yield increased with an increase in catalyst loading (Figure 1B). The runs with the equivalent mixture of MoC13and NiCl, catalysts generally consumed moderate amounts of hydrogen from Hz and tetralin (see Table 111). In contrast to the catalytic runs, the use of LiC1-KC1 (Figure 2C), even at low loading, considerably suppressed the consumption of gas-phase H2 and total hydrogen consumption and slightly increased net hydrogen transfer from tetralin. It has been observed1' in other catalytic coal liquefaction reactions that tetralin consumption (net hydrogen transfer) in H,/noncatalytic and H,/catalytic systems remained unchanged. Figures 1 and .2 show that MoC13 and NiCl, catalysts not only promoted coal conversion but also changed the routes and extents of hydrogen transfer. Three possibilities for the increase in consumption of H2 and the decrease in net hydrogen transfer from tetralin as compared to the noncatalytic run are (1) the catalysts enhanced regeneration of the spent H-donor vehicle by Hz (indirect hydrogen transfer via tetralin/naphthalene intercon~ersion),'~ (2) the catalysts promoted direct hydrogenation of coal by H,, which increased Hz consumption,12-14and (3) the catalytic liquefaction produced more hydroaromatic compounds that acted, partially in place of tetralin, as in situ H-donor components in the reaction stage, and thus the contribution of tetralin to H donation became lower with higher catalyst loading. Probably the first and second possibilities occurred in the catalytic runs, with the first occurring more readily with the MoC13 catalyst than with the NiC1, catalysts. To what extent they contributed depends upon the activity of the catalysts1g2oand the relative reactivity of the spent H donor and coal toward hydrogenation.21 Moreover, the results in Figure 2A,B suggest that the third possibility also occurred to some extent at higher catalyst loading levels, especially with the NiCl2 catalyst, because the increased oil yields were obtained without any increase in external hydrogen consumption (Figure lB, Figure 2B). Evidence for the enhancement of the formation of hydroaromatics with increasing catalyst loading was found in the 'H NMR analysis of liquefaction products (see below). Recent studies22-26have established that some partially hydrogenated polyaromatics have higher H-donating activity than tetralin. The decrease of H2 consumption with LiC1-KC1 as compared to the noncatalytic run is a curious phenomenon. Previously, alkali-metal chlorides alone in solvent-free coal liquefaction resulted in a small decrease in Hzconsumption, as compared to the reaction without any additive.12 The results in Figure 2C indicate that such an effect becomes much more significant in the presence of an H-donor tetralin vehicle. Relation between Coal Conversion and Hydrogen Consumption. Coal conversion processes can be divided (17) Moritomi, H.; Nagaishi, H.; Naruse, M.; Sanada, Y.; Chiba, T.J. Fuel SOC.Jpn. 1983, 62, 254-262. (18) Makabe, M.; Ohe, S.; Itoh, H.; Ouchi, K. Fuel 1986,65, 296298. (19) Kabe, T.;Nitoh, 0.;Marumoto, M.; Kawakami, A.; Yamamoto, K. Fuel 1987,66, 1321-1325. (20) Kabe, T.;Nitoh, 0.;Funatsu, E.; Yamamoto, K. Fuel 1987, 66, 1326-1329. (21) Ohe, S.;Itoh, H.; Makabe, M.; Ouchi, K. Fuel 1985,64,110&1111. (22) Ruberto, R. G.;Cronauer, D. C.; Jewell, D. M.; Seshadri, K. S. Fuel 1977,56, 25-32. (23) Derbyshire, F. J.; Whitehurst, D. D. Fuel 1981, 60, 655-662. (24) Kamiya, Y.; Yao, T.; Nagae, S. Bull. Chem. SOC.Jpn. 1982, 3873-3877. (25) Kamiya, Y. J . Fuel SOC.Jpn. 1984, 63, 224-238. (26) Stock, L.M. Chemistry of Coal Conuersion; Schlosberg, H., Ed., Plenum: New York, 1985; pp 253-316.
501
Song et al.
Gas
0 Oil
A Asphaltene
i
0 Preasphaltene
40
C
017
0'6
OR
012 0 9
1 .o 2.0 3.0 Total hydrogen consumption (wt%,daf)
Figure 3. Plot of product yields against total hydrogen consumption for liquefaction of Yilan coal. Numbers attached to plots of oil yields correspond to the following runs: (1) none; (2) none, 0 h; (3) NiCl2-LiC1-KC1 (loading level O.l), 0 h; (4-6) LiCl-KCl(O.1,0.2,1.0); (7-10) MoCl3-LiC1-KCl (0.1,0.2,0.5, 1.0); (11-14) NiCl,-LiCl-KCl (0.1, 0.2, 0.5, 1.0); (15-18) MoC13-NiClZ-LiC1-KC1 (0.1, 0.2, 0.5, 1.0).
into (1)cracking coal and coal-derived heavy products, (2) hydrogenation and hence stabilization of the reactive fragments formed during cracking steps, and (3) the simultaneous removal of heteroatoms., From the viewpoint of stabilization of catalytically and thermally derived fragments, H,, H-donor solvent, and transferable hydrogen present in coal and coal-derived products are different but may play essentially the same roles. Figure 3 shows the yields of gas, oil, asphaltene, and preasphaltene as a function of total hydrogen consumption for all the liquefaction runs, including runs with MoC13-LiC1-KC1, NiCl2-LiC1-KC1, LiC1-KC1, and MoC13-NiC12-LiC1-KC1 at various loading levels and two 0 h runs (with the NiClz catalyst and without any salt) as well as the Soxhlet extraction of the raw coal. Oil yields increased with total hydrogen consumption, and asphaltene yields reached maximum values at certain amounts of total hydrogen consumption. For preasphaltenes, a bimodal pattern is observed. The patterns appear to be different from those in the l i t e r a t ~ r e . ~ ~ , ~ ~ Figure 3 shows the existence of some regions where yields of individual products increased sharply with a relatively smaller increase in hydrogen consumption. Since literature data have demonstrated that the production of oil requires considerably higher hydrogen consumption than asphaltenes and p r e a s p h a l t e n e ~ , ' -the ~~~ most ~ ~ ~striking feature of Figure 3 is that oil yields increased dramatically with (27) Ohe, S.;Itoh, H.; Makabe, M.; Ouchi, K. Fuel 1985,64,902-905. (28) Ouchi, K.;Ohe, S.; Makabe, M.; Itoh, H. Fuel 1985,64,1391-1393. (29) Mochida, I.; Otani, K.; Korai, Y. Fuel 1985, 64, 906-910. (30) Hertan, P. A.; Jackson, W. R.; Larkins, F. P. Fuel 1985, 64, 1251-1259.
Catalysis in Coal Hydroliquefaction
Energy & Fuels, Vol. 2, No. 5, 1988 643
Table IV. Elemental Analysis and Structural Parameters elemental anal., w t 70 catalyst product C H N 0 + Sb atomic H/C none oil 84.4 8.5 0.8 6.3 1.20 asph 80.0 6.7 1.3 12.0 1.00 1.20 LiC1-KC1 oil 85.2 8.6 1.0 5.3 1.6 7.8 0.89 84.3 6.3 asph 1.17 MoC1,-LiC1-KC1 oil 0.6 4.2 86.7 8.5 81.3 6.4 1.5 10.8 0.94 =Ph 1.20 86.1 8.7 0.9 4.4 NiCl2-LiC1-KC1 oil 83.3 6.6 1.4 8.6 0.94 asph 86.9 8.6 0.7 3.8 1.18 MoC1,-NiC12-LiC1-KC1 oil 1.3 8.6 0.91 83.6 6.4 asph
run no. 1
6 10 14 18
params"
f.
HJCn. 0.65 0.50 0.67 0.54 0.71 0.50 0.73 0.50 0.70 0.57
u
0.54 0.62 0.54 0.69 0.56 0.65 0.54 0.64 0.56 0.68
0.44 0.41 0.43 0.40 0.44 0.44 0.46 0.49 0.44 0.40
'Based on the Brown-Ladner equations: fa, aromaticity; u, degree of substitution in aromatic systems; H,/C,, ratio of the number of aromatic hydrogen atoms to the number of carbon atoms for a hypothetical unsubstituted aromatic system. bBy difference. run no.
Table V. Hydrogen Distribution of Oil and AsDhaltene hydrogen distribn/ 100 carbon atoms' product Hm HF Ha1 Hu2 HPl HP2 H, HCHz HCH3 Har 8.9 42.8 14.5 33.1 66.9 19.6 oil 19.6 3.2 21.0 9.6 asph 18.0 6.2 14.5 6.1 8.6 34.7 11.4 29.3 52.2 18.0 oil 20.6 3.1 21.7 9.2 8.3 43.5 13.7 33.1 66.4 20.6 asph 22.7 6.1 17.7 8.8 6.9 20.4 6.7 30.7 36.0 22.7 oil 22.4 4.6 25.2 8.2 9.7 34.7 12.0 39.5 54.9 22.4 asph 18.2 6.3 17.5 7.4 8.4 25.6 9.8 32.2 42.8 18.2 oil 21.4 5.8 24.3 9.5 10.3 36.4 12.5 40.4 58.4 21.4 asph 16.3 7.5 19.2 6.9 9.3 25.3 9.9 36.0 42.2 16.3 asph 23.3 7.3 19.3 7.7 8.1 19.0 6.8 34.6 33.4 23.3
catalyst
1
none
6
LiCl/KCl
10
MoClR-LiC1-KC1
14
NiC1,-LiC1-KCl
18
MoC1,-NiC12-LiC1-KC1
HCHl = HF + Hal + HP1; HCH3 = Hu2 + H,
atomic ratiob HCHz/Har
5.10 4.53 4.83 2.94 4.21 4.12 4.62 4.80 2.92
1.69 1.63 1.61 1.35 1.76 1.77 1.89 2.21
1.48
HCHz/HCH3
0.49 0.56 0.50 0.85 0.72 0.75 0.69 0.85 1.04
+ H,. * Hd = HcH~+ HCH3. 0.75
A
MoC13-LiC1-KC1
P
NiCl.-LiCl-KCl
0.70
J
n
m
2 0.65
EI. 8 [
0 None
0
\ V
L
nil
X
0.60
0
LiC1-KCl
A
MoC 1 3 - L i C1 -KC1 NiCl2-LiC1-KC1
0
0.2
0.4
0.6
0.8
1 .o
Weight r a t i o o f s a l t s t o c o a l
Figure 5. Influence of salt loading level on H,/C, I
0
I
I
I
I
I
I
0.2 0.4 0.6 0.8 Weight r a t i o o f s a l t s t o c o a l
I
I
1.0
Figure 4. Influence of salt loading level on heteroatom (N + S + 0) content of oil and asphaltene.
only a slight increase in total hydrogen consumption at 1.8-2.5 w t % ' for the MoCl, and the NiC1, catalysts as well as their equivalent-weight mixture (see Figure 3). The action of the catalysts brought about enhanced, selective hydrocracking reactions, which mainly contributed to increasing oil product. Some insights into the activity of these catalysts were found by characterizing the liquefaction products (see below). Characterization of Products. Table IV gives the typical results of elemental analysis and average structural parameters of oil and asphaltene. Figure 4 shows the change of heteroatom (N + S + 0) content of oil and asphaltene. Increases in the loading level of both MoCl, and NiCl, catalysts decreased the heteroatom content of
value of oil.
the oil, as might be expected. In this respect the MoCl, catalyst is superior to the NiC1, catalyst. The oil from the catalytic runs show a similar H/C atomic ratio and higher H,,/C, values than that from noncatalytic runs. Figure 5 shows the increase of H,,/C, values of oils with increased MoCl, and NiC1, catalyst loading, suggesting a progressive decrease in the average ring size of aromatics. Asphaltenes from runs with either the MoC13or the NiCl, catalyst and LiC1-KC1 have a lower heteroatom content (Figure 4)and show higher aromaticity (f,) than those from a noncatalytic run. Figure 6 shows the changes of fa for asphaltenes. The change in distribution of well-defined hydrogen functional groups of the products reflects the action of the added catalyst^.'^ Table V shows the distribution of hydrogen atoms per 100 carbon atoms of oil and asphaltene. The oil from catalytic runs 10, 14, and 18 contains larger amounts of H,, HF, Hal, and H,, and smaller amounts of H, and H, than that from noncatalytic run 1or run 6 with LiC1-KC1 alone. Asphaltenes from the catalytic runs contain larger amounts of Hal and smaller amounts of H,
0.70k0.70k==
644 Energy & Fuels, Vol. 2, No. 5, 1988
Song et al.
+ HY)represents CY-, 0-,and y-methyl hydrogens, methy-
0.65
0.60
-
J
0 None
7
m
0
m
L
LiC1-KC1
d MoCl 3-LiC1 -KC1 0 NiC12-LiC1-KC1
0.55
0.2 0.4 0.6 0.8 Weight r a t i o o f s a l t s t o coal
0
1 .o
Figure 6. Influence of salt loading level on fa value of asphaltene.
I*
i! 0
2
50t
Conclusions The performance of MoCl,-LiCl-KCl and NiClZ-LiClKC1 in the hydroliquefaction of Yilan subbituminous coal with tetralin was studied in detail. These catalysts not only promote coal conversion, but also change the routes and extents of hydrogen transfer (Figures 1and 2; Table 111). Examination of the relationship between coal conversion and hydrogen consumption suggests that the action of these catalysts leads to enhanced but selective hydrocracking reactions that contribute mainly to oil production (Figures 1-3). Characterization of liquefaction products revealed that the use of these catalysts cause pronounced elemental and structural changes of both oils and asphaltenes (Figures 4-7; Tables IV and V). Comparison of the resuls indicate that the NiC12catalyst is more effective than the MoCl, catalyst at low loading levels (
