Ind. Eng. Chem. Res. 1993,32, 753-755
753
Deep Desulfurization of Light Oil. 3. Effects of Solvents on Hydrodesulfurization of Dibenzothiophene Atsushi Ishihara a n d Toshiaki Kabe' Department of Chemical Engineering, Tokyo University of Agriculture and Technology, Nakamachi, Koganei, Tokyo 184, Japan
Solvent effects on hydrodesulfurization (HDS) of dibenzothiophene (DBT) catalyzed by Co-Mol A1203 were investigated under deep hydrodesulfurization conditions. It was found that type of solvent affects HDS of DBT into biphenyl (BP). By contrast, the formation of cyclohexylbenzene (CHB) was scarcely affected by solvents such as xylene, decalin, tetralin, and n-hexadecane. From this result, it was suggested that desulfurization and hydrogenation proceed on different catalytic sites and that cyclohexylbenzene would be formed by hydrodesulfurization of hexahydrodibenzothiophene as well as hydrogenation of biphenyl. Introduction
A number of attempts have been made to elucidate the mechanism of HDS of sulfur-containing compounds, especially thiophenes, by a kinetic approach (Gates et al., 1979). Most studies deal with a reaction of thiophene (Vrinat, 1983), benzothiophene (Kilanowski and Gates, 19801,and dibenzothiophene(Broderick and Gates, 1981; Vrinat and de Mourgues, 1982;Singhal et al., 1981;Houalla et al., 1978)with hydrogen catalyzed by Co-Mo/AlzOa and have proposed many kinetic equations which are often different depending on experimental methods or reaction conditions. Although the representative kinetic studies have been performed on the HDS of DBT in the presence of Co-Mo/AlnO3, the retarding effects of components in feedstock, e.g., solvents, on the catalytic activity and the product selectivity remain unknown. We have already reported the deep desulfurization of light oil (Ishihara et al., 1992; Kabe et al., 1992a)and the kinetic study of alkyl-substituted DBTs (Kabe et al., 199213). In the course of our study, we investigated the effects of solvents in HDS under deep desulfurization conditions (below 0.05 wt % sulfur). In this paper, we report the retarding effects of solvents such as xylene, decalin,tetralin, and n-hexadecaneon the catalyticactivity and the product selectivity in deep HDS of dibenzothiophene catalyzed by Co-Mo/AlzOs. Experimental Section Materials. Solvents, xylene (mixture of o-, m-, and p-xylene), decalin (mixture of cis- and tram-decalin), tetralin (Kishida Chemicals), n-hexadecane, and dibenzothiophene (TokyoKasei Chemicals),were of commercial GR grade and were further purified by passing through a column (i.d. 20 mm; length 30 cm) containing activated alumina (0.063-0.200 mm). Dibenzothiophene (Tokyo Kasei Chemicals) was recrystallized from ethanol. Hydrogen (99.99%) was obtained from Tohei Chemicals. Hydrogen sulfide in hydrogen (H2S; 3 % ) was obtained from TakachioChemicals. The commercial Co-Mo/AlzOs (COO,3.8 wt %; Moos, 12.5 wt 5%) was supplied as a 1/32 extrudate which was crushed and screened to provide 0.841.19-mm granules used in this work. Apparatus and Procedure. The reactor was a 10mm-i.d. stainless-steel tube packed with 0.2 g of catalyst particles diluted with quartz sand to 3 cm3. The single charge was used throughout the entire series of experiments. After the catalyst bed was heated for more than 24 h at 450 "C in air, it was presulfided with a mixture of 3 % H2S in H2 flowing at 30 L/h at atmospheric pressure
and 400 "C for 3 h. After these pretreatments, the reactor was cooled in a H2S/H2stream to the expected temperature and was pressurized by hydrogen. Then, the solution containing dibenzothiophene was supplied to the feed pump (Kyowa Seimitsu KHD-16). A typical reaction was carried out under the following conditions: temperature, 180-310 "C; total pressure, 25-100 atm; flow rate of hydrogen (Hz),18L/h; flow rate of solution, 14g/h (weight hourly space velocity (WHSV) 70 h-l); concentration of dibenzothiophene, 0.1-3.0 wt % . Conversion of DBT reached a steady state within about 3 h. Then, samples of products were collected from a gas-liquid separator 4 times every 15 min. In the sampling period, hydrogen sulfide was trapped in an aqueous solution of Pb(CH3C00)~3HzO.PbS formed was filtered, dried at 120 "C for 2 h, and weighed. Sulfur removed by the conversion of DBT was quantitativelyrecovered by the transformation of hydrogen sulfide into PbS in that period. Subsequently, the reaction temperature was changed, and after 1 h sampling was carried out in a similar manner. No sign of deactivation of the catalyst was observed during the run for 16 h. Analysis. Reaction products were analyzed by gas chromatographywith a flame ionization detector (Hitachi 163) using a capillary column, G-column 100 (i.d. 1.0 mm; film thickness 1.0 pm; length 40 m) programmed from 150 to 220 "C (heating rate 8 "C/min; injection temperature 270 "C; carrier gas N2). Commercially available G-column 100 was supplied by the Chemicals Inspection & Testing Institute. Results and Discussion In Figure 1, HDS reactions of DBT, 4-methyldibenzothiophene,and 4,6-dimethyldibenzothiophenein decalin (Ishihara et al., 1992; Kabe et al., 1992a) were compared with those in light oil (Kabe et al., 1992b). In light oil, about 50 "C higher temperature is required in order to obtain the same conversions as those of DBTs in decalin. In HDS of 4-MDBT in decalin, about 50 "C higher temperature is also required in order to obtain the same conversionsas those of DBT in decalin at the temperature range where the differential reactor can be applied. At 220 "C, the rates of HDS of DBT and 4-MDBT in decalin were 7.1 x 10-4 and 6.9 X 103 mol/(g of catalystah), respectively (Kabe et al., 199213). Thus, the rate of HDS of thiophenes in light oil decreases to about 1/10 of that in decalin. This clearly shows the extent of retarding effects of components in light oil. In order to estimate such retarding effectsof components in light oil on the catalytic activity of HDS and the product
0SSS-5~~519312632-0753$04.00/0 0 1993 American Chemical Society
754 Ind. Eng. Chem. Res., Vol. 32, No. 4,1993
Reaction Temperature (%I Temperature (OC)
Figure 1. Retarding effecta of components in light oil on conversions of dibenzothiophenes. Initial concentration of DBT, 4-MDBT, or 4,6-DMDBT in decalin: 0.1 wt % , Initial concentrations of DBT, 4-MDBT, and 4,6-DMDBT in light oil: 0.13, 0.17, and 0.08 wt %, respectively. DBTsindecalin: 0,DBT; A,4-MDBT; 0,4,6-DMDBT. DBTs in light oil: 0 , DBT; A, 4-MDBT; M, 4,6-DMDBT.
Figure 4. Effects of solvents on conversion of dibenzothiophene into cyclohexylbenzene. 50 kg/cm2;WHSV 70 h-1; catalyst 0.2 g; Hz 18 L/h; initial concentration of DBT, 0.1 wt % Solvent: 0,xylene; A, decalin; 0 , tetralin; 0; n-hexadecane.
.
Table I. Partial Pressure of Hydrogen, Dibenzothiophene, and Solvent. ~~
hydrogen dibenzothiophene solvent (kg/cm2) (kdcm2) xylene 42.55 0.043 decalin 44.05 0.045 tetralin 43.82 0.044 n-hexadecane 46.18 0.047
solvent tamp* (kdcm9 ("C) 7.40 240 5.70 289 6.13 3MC 3.78 358
Total pressure 50 kg/cm2; WHSV 70 h-l; catalyst 0.2 g; Hz 18 Lih; initial concentration of DBT, 0.1 wt % Temperature required to evaporate solvent completely. Antoine equation was available. Calculated from data for vapor pressure of tetralin in the literature (Simnick et al., 1977).
.*
Reaction Temperature IT) Figure 2. Effects of solvents on conversion of dibenzothiophene. 50 kg/cm2;WHSV 70 h-l; catalyst 0.2 g; Hz 18L/h; initial concentration of DBT, 0.1 wt %. Solvent: 0, xylene; A, decalin; 0 , tetralin; 0, n-hexadecane.
Reoction Temperature
(OC)
Figure 3. Effects of solvents on conversion of dibenzothiophene into Biphenyl. 50 kg/cm2;WHSV 70 h-l; catalyst 0.2 g; H2 18 L/h; initialconcentrationofDBT,O.lwt % . Solvent: 0,xylene; A,decalin; 0 , tetralin; 0, n-hexadecane.
selectivity, effects of solvents on hydrodesulfurization (HDS) of dibenzothiophene (DBT) catalyzed by Co-Mo/ A1203 were investigated a t 200-310 OC, 50 atm, and 0.1 wt 3' 6 DBT. Solvents used were xylene, decalin, tetralin, and n-hexadecane. Products were biphenyl (BP), cyclohexylbenzene (CHB), hydrogen sulfide, and a trace amount of hexahydrodibenzothiophene. Figure 2 showsthe effect of temperature on the conversionof DBT. The conversion of DBT decreased in the order xylene > decalin.> tetralin > n-hexadecane over every temperature. Figure 3 and 4 show the effects of temperature on the conversionsof DBT into B P and CHB, respectively. Shapesof curves in Figure 3 are very similar to those in Figure 2. This may indicate that solvents competitively adsorbed on the active sites for desulfurization to prevent the adsorption of DBT and
the formation of BP. Another possibility would be the effects of solvents evaporation on HDS of BDT which have been reported by Kocis and Ho (Kocisand Ho, 1986). Table I shows the partial pressures of DBT, Hz, and a solvent. In Table I, it was assumed that all components were gaseous. In order to make a solvent gaseous, the vapor pressure of the solvent should exceed the partial pressure of the solvent. Table I also shows enough temperature to evaporate a solvent completely. In this calculation, the Antoine equation was used. The conversions with the use of saturated hydrocarbons decalin and n-hexadecane were lower than those with aromaticsxylene and tetralin, respectively. This result may be difficult to explain by only competitive adsorption of solvents and DBT since an aromatic compound would adsorb more competitively than a saturated one. In this case, it may be also related to the effect of liquid evaporation. Due to capillary condensation and the conditions used, the reaction should take place predominantly in the liquid phase. Xylene is more volatile than DBT; as a result, the concentration of DBT in the liquid phase increases, which in turn speeds up the reaction. The reverse is true in the case with n-hexadecane. With regard to the effecta of hydrogen solubility on HDS, it can be ignored. Although solubility of hydrogen into saturated hydrocarbon n-hexadecane (ratio of dissolved hydrogen in n-hexadecane, 0.1007 at 269 "C and 50 atm) (Lin et al. 1980) is higher than that into aromatic tetralin (0.0373 at the same condition) (Simnick et al., 19771, the conversion with the use of n-hexadecane was lower than that with tetralin. In those reports, the solubility of hydrogen changed eignificantly depending on pressure. When we changed the partial pressure of hydrogen in the range 22-66 kg/cm2, however, the conversion of DBT was hardly affected. On the other hand, solvent effects on the formation of CHB in Figure 4 were scarcely observed in comparison with that of BP. This means that the formation of CHB is not affected by solvents. The fact that different solvents
Ind. Eng. Chem. Res., Vol. 32, No. 4, 1993 755 Scheme I. Reaction Pathways in Hydrodesulfurization of Dibeqzothiophene. Proposed by Houalla et al. (1978)
Gates, B. C.; Katzer, J. R.; Schuit, G. C. A. Chemistry of Catalytic Processes; McGraw-Hill: New York, 1979. Houalla, M.; Nag, N. K.; Sapre, A. V.; Broderick, D. H.; Gates, B. C. Hydrodesulfurization of Dibenzothiophene Catalyzed by Sulfided CoO-MoOsgamma-Al~03:The ReactionNetwork. AIChE J. 1978, 24, 1015-1021. Ishihara, A.; Tajima, H.; Kabe, T. Deep Desulfurization of AlkylSubstituted Dibenzothiophenes in Light Oil. Chem. Lett. 1992, 669-770. Kabe, T.; Ishihara, A.; Tajima, H. Hydrodesulfurization of SulfurContaining Polyaromatic Compounds in Light Oil. Znd. Eng. Chem. Res. 1992a, 31,1577-1580. Kabe, T.; Ishihara, A.; Zhang, Q. Deep Desulfurization of Light Oil. 2. Hydrodeedfurhation of Dibenzothiophene, 4-Methyldibenzothiophene and 4,6-Dimethyldibenzothiophene.Appl. Catal. 1992b, in press.
influence the desulfurization rate to BP but do not affect the formation of CHB supporta the thesis made in the literature (Houallaetal., 1978;Broderick and Gates, 1981) that desulfurization and hydrogenation proceed on different catalytic sites. The reaction scheme of HDS of DBT is shown in Scheme I, which has been proposed by Houalla et al. (1978). Although the BP to CHB pathway cannot be completely ruled out, the BP to CHB pathway would be kinetically insignificant under our experimental conditions. If CHB is mainly formed from hydrogenation of BP and hydrogenation of BP is not affected by solvents, the differencein the HDS rate of DBT into BP willgenerate the significant difference in the rate of CHB formation. Although the effect of solubility of DBT in solvents was not discussed in this paper, this may affect the catalytic activity of HDS. Retarding effects of componentsin light oil on the reactivity of DBTs contain the s u m of various factors such as competitive adsorption of components, liquid evaporation, solubility of and DBTs. In order to understand these effects, further investigation will be necessary.
Kilanowski, D. R.; Gates, B. C. Kinetics of Hydrodesulfurization of Benzothiophene Catalyzed by Sulfided Co-Mo/AlzOs. J. Catal. 1980,62,70-78. Kocis, G. R.; Ho, T. C. Effects of Liquid Evaporation on the Performance of Trickle-Bed Reactors. Chem.Eng. Res. Des. 1986, 64,288-291. Lin, H.-M.; Sebastian, H. M.; Chao, K.-C. Gas-Liquid Equilibrium in Hydrogen + n-Hexadecane and Methane + n-Hexadecane at Elevated Temperature and Pressures. J. Chem. Eng. Data 1980, 25,252-254. Simnick, J. J.; Lawson, C. C.; Lin, H. M.; Chao, K. C. Vapor-Liquid Equilibrium of Hydrogen/Tetralin System at Elevated Temperature and Pressures. AIChE J. 1977, 23, 469-476.
Singhal, G. P.; Espino, R. L.; Sobel, J. E.; Huff, G. A. Hydrodesulfurization of Sulfur Heterocyclic Compounds. Kinetics of Dibenzothiophene. J. Catal. 1981, 67, 457-468. Vrinat, M. L. The Kinetic of Hydrodesulfurization Process-A Review. Appl. Catal. 1983,6,137-158. Vrinat, M. L.; de Mourgues, L. Kinetic Study on Sulfided Co-Mol &os, Mo/AlzOs and Co/Al203 Catalysts: Hydrodesulfurization of Dibenzothiophene. J. Chim. Phys. 1982, 79-1,4652.
Literature Cited Broderick, D. H.; Gates, B. C. Hydrogenolysis and Hydrogenation of Dibenzothiophene Catalyzed by Sulfided CoO-MoOsgammaAl203: The Reaction Kinetics. AZChE J. 1981,27,663-673.
Received for review August 17, 1992 Revised manuscript received December 18, 1992 Accepted January 11, 1993