Energy & Fuels 1988,2, 645-651
645
Hydrodenitrogenation of Carbazole on a Mo/A1203 Catalyst. Effects of Sulfiding and Sulfur Compounds Masatoshi Nagai,* Takashi Masunaga,t and Nobuaki Hana-oka* Department of Chemical Engineering, Tokyo University of Agriculture and Technology, Koganei, Tokyo 184, Japan Receiued June 16, 1987. Revised Manuscript Receiued April 22, 1988
The effects of sulfidation and sulfur compounds on the hydrodenitrogenation of carbazole, which contains one of the organonitrogen components of coal liquids and petroleum residue, have been studied. The reaction was carried out in a continuous flow microreactor with reduced and sulfided Mo/A1203 catalysts a t temperatures of 280-360 "C and 10.1 MPa total pressure. Sulfiding of the reduced catalyst enhanced the hydrogenation of the carbazole and the formation of denitrogenated compounds such as bicyclohexyl, a-ethylbicyclo[4.4.0]decane,and hexylcyclohexane at all temperatures. However, the presence of sulfur compounds depressed the C-N bond scission on both reduced and sulfided catalysts, which was the rate-determining step a t 360 0C. On the other hand, at low temperatures the sulfur compounds depressed the hydrogenation, which was not in complete equilibrium, and consequently caused a decrease in the C-N hydrogenolysis. These results showed that the C-N hydrogenolysis took place competitively with the desulfurization on the same sites of the Mo/A1,03 catalyst.
Introduction Hydrodenitrogenation has received increasing attention during the past decade. The high nitrogen contents of synthetic oils derived from shale oil, oil sand, and coal make hydrodenitrogenation imperative if the liquids are to be refined in order to replace products presently obtained from petroleum. Nitrogen is present largely in the form of heterocyclic compounds having five- or six-membered rings.14 The heterocyclic and other nitrogen compounds are often grouped into ir-deficient nitrogen compounds, i.e. strong bases (pyridines, quinolines, acridines), and a-excessive nitrogen compounds, i.e. weak bases or nonbasic compounds (pyrroles, indoles, carbazoles). An extensive research background exists on the hydrodenitrogenation of pyridine,- quinoline,*" and indole,12-14 but there are few reports concerning the hydrodenitrogenation of acridine11J5and c a r b a ~ o l e ,which ~~,~~ are mainly present in these liquids. Hydrodenitrogenation occurs simultaneously with hydrodesulfurization during hydroprocessing. Along with studies on the interactions between hydrodesulfurization and hydrodenitrogenation, increasing attention has been given to mixture studies involving both synthetic (model compounds) and real feedstocks. Sever1 rep~rtsl'-'~reveal common and severe inhibitory effects of nitrogen-containing compounds on the catalytic removal of sulfur, but relatively little attention has been paid to how sulfur compounds may affect hydrodenitrogenation: CS2 enhanced the hydrodenitrogenation reactions of pyridinez0 and quinolinez1p22 but depressed acridine hydrodenitr~genation.'~ In this study, the effects of sulfur compounds and sulfiding on carbazole hydrodenitrogenation on a Mo/A1203 catalyst were determined in order to elucidate the relationship between the presence of sulfur compounds and changes in the hydrodenitrogenation activity of the Mo/ A1203catalyst. The behavior of carbazole was evaluated 'Present address: Kumon Publishing Co., 3-6-9Misakicho, Chiyodaku, Tokyo 160,Japan. f Present address: Okamoto Co. Ltd., Yoshida-cho, Shizuoka 421-03,Japan.
0887-0624/88/2502-0645$01.50/0
as a function of sulfiding of the catalyst and the effect of sulfur compounds and compared with the behavior of acridine, which is one of the ir-deficient three-ring heterocycles. The active sites of a Mo/A1203 catalyst in C-N hydrogenolysis are also discussed. Experimental Section Materials. The catalyst was t h e same as previously used,15 namely, 12.5% MooB o n -pA1203. Catalyst particle sizes of 0.84-1.19 mm granules were employed. Hydrogen (To-ei Chemical Co., 99.9%) was dried by passing it through a Linde 13X molecular sieve trap prior t o use. The dibenzothiophene used in these experiments was also t h e same as material previously synthesized.17 Carbazole (Eastman Organic Chemicals, Ultrapure Grade), all other reagents (Tokyo Kasei Chemicals, E x t r a P u r e Grade), and H2S (To-ei Chemical Co., 99.9%) were used without puri(1) Mckay, J. F.; Weber, J. H.; Latham, D. R. Anal. Chem. 1976,48, 891. (2)Holmes, S. A.;Thompson, L. F. Fuel 1983,62,709. (3)Schmitter, J.; Ignatiadis, I.; Dorbon, M.; Arpino, P.; Guiochon, G.; Toulhoat, H.; Huc, A. Fuel 1984,63,557,565. (4)Novotny, M.; Konishi, M.; Hirose, A.; Gluckman, J.; Wiesler, D. Fuel 1985,64,523. (5)Sonnemans, J.; Mars, P. J. Catal. 1973,31,209. (6)Cerny, M. Collect. Czech. Chem. Commum. 1979,44,85. (7)Sonnemans, J.; Janus, J. M.; Mars, P. J. Phys. Chem. 1976,80, 2107. (8) Mcllvried, H. G. Ind. Eng. Chem. Process Des. Deu. 1971,10,125. (9)Glola, F.;Lee, V. Ind. Eng. Chem. Process Des. Deu. 1986,24918. (10)Miller, J. T.; Hineman, M. F. J. Catal. 1984,85, 117. (11)Shih,S.S.;Reiff, E.; Zawadzki, R.; Katzer, J. R. Prepr. Pap.-Am. Chem. SOC., Diu. Fuel Chem. 1978,23,99. (12)Rollmann, D. D.J. Catal. 1977,46,243. Ollis, D. F. J. Catal. 1983,80,76. (13)Odebunmi, E.0.; (14)Stern, E.W. J. Catal. 1979,57,390. (15)Nagai, M.; Masunaga, T.; Hana-oka, N. J. Catal. 1986,101,284. (16)Nagai, M.; Sawahiraki, K.; Kabe, T. Nippon Kagaku Kakhi 1980, GFI
(17)Nagai, M.; Sato, T.; Aiba, A. J. Catal. 1986,97,52. (18)Massoth, F.E.;Miciukiewicz, J. J. Catal. 1986,101,505. (19)Ramachandran, R.; Massoth, F. E. Chem. Eng. Commum. 1982, 18,239. (20)Satterfield, C. N.; Modell, M.; Wilkens, J. A. Ind. Eng. Chem. Process Des. Deu. 1980,19, 154. (21)Shih, S.; Katzer, J. R.; Kwart, H.; Stiles, A. B. Prepr.-Am. Chem. SOC.,Diu. Pet. Chem. 1977,22, 919. (22) Nelson, S. H.; Levy, R. B. J. Catal. 1979,58,485.
0 1988 A m e r i c a n C h e m i c a l Society
646 Energy & Fuels, Vol. 2, No. 5, 1988
Nagai et al.
n I 14
LiJ10 products
Figure 1. Schematic diagram of high-pressure flow system: (1) hydrogen; (2) hydrogen sulfide; (3) molecular sieves; (4) feed tank; (5) high-pressure liquid pump; (6) reactor; (7) high-pressure gas-liquid separator; (8) rotameter; (9) pressure regulator; (10) stop valve; (11)needle valve; (12) pressure gauge.
fication. Apparatus and Procedure. The reactor and associated flow system used in this study are shown in Figure 1. The reactor was constructed from stainless-steeltubing with an 11-mm inside diameter and 30-cm length. The catalyst was held in place by means of steel b e a r i i and steel bars engraved with spiral ditches The reactor, which was placed in a vertical position, was extemally heated by using an oven connected to a thermal controller. Two grams of catalyst was charged in the reactor for a typical series of experiments. Liquid was fed by high-pressure liquid-chromatography pump through a preheater coil and then to the top of the catalyst bed. The hydrogen and liquid feeds were mixed at the reactor inlet and passed downward through the catalyst bed. The catalyst was first preheated in air for more than 24 h at 450 "C in the reactor. Reduction was carried out in situ by exposing the oxidized catalyst to hydrogen at 400 "C and 10.1 MPa pressure for 4 h. Sulfiding was accomplished with a 10% H2S/H2atmosphere flowing at 30 L/h and 400 "C for 3 h after reduction. After reduction (or sulfidation),the reactor was cooled to the desired temperature in a H2 (or a 10% HzS/H2)stream. Carbazole was dissolved in xylene to form a solution containing 1.3 X lo-, mol/L (-0.22 wt %) carbazole. All experiments described here were performed at 280-360 "C and 10.1 MPa total pressure. Hydrogen was fed at a rate of 30 L/h. The solution was then allowed to flow at 20 mL/h into the reactor, producing 1.86 kPa of initial carbazole partial pressure. This gas-to-liquid ratio was high enough tb insure complete vaporization of the liquid above 280 0C.z3 The reaction reached a steady state in about 3 h. A solution containing 1.3 X lo-, mol/L of carbazole and a concentration of sulfur compounds was then introduced. After 2 h, samples of the liquid products were collected 3 times every 15 min. The next feed, which contained carbazole and a more concentrated sulfur compound than the previous one, was then introduced. Analysis. Liquid-phase samples of the reactor effluent were collected downstream from the gas-liquid separator, and concentrations were related to the starting liquid feed. The quantitative analysis of the nitrogen compounds and the hydrocarbons was carried out by using a Hitachi 163 gas chromatograph equipped with a flame ionization detector using a 2 % Dexsil400 GC/uniport HP. Gaseous products, such as NH3 and any light cracking products, could not be determined because they were dilute. 1,2,3,4-Tetrahydrocarbazole,bicyclohexyl, a-ethylbicyclo[4.4.0]decane,and hexylcyclohexane were isolated by gas chromatography equipped with a thermal conductivity detector using a Silicone OV-17 column. a-Ethylbicyclo[4.4.0]decaneand hex~~~~~
(23) Satoh, K.; PoyntKg, K. Handbook for Chemical Engineers, 4th ed.; The Society of Chemical Engineers, Japan: Tokyo, 1978; pp 25,39.
ylcyclohexane were synthesized separately in the hydrogenolysis of bicyclohexyl at 340 "C and 10.1 MPa hydrogen pressure. The 25-MHz 13C NMR spectrum was obtained on a JEOL JNM-FX 100 spectrometer. Mass spectra were obtained with a Hitachi RMU-7M spectrometer (ion source temperature, 150 "C; ionizing voltage, 70 eV). The infrared spectra of the compounds were obtained by a Nihon Bunko spectrometer. The other hydrogenated compounds and hydrocarbons, e.g., 1,2,3,4,4a,9a-hexaperhydrohydrocarbazole, 1,2,3,6,7,8,8a,9a-octahydrocarbazole, carbazole, and cyclohexylbenzene were analyzed by mass spectroscopy with a Shimazu LK-9000 GC-MS instrument (column packing, 1.5% silicone OV-17; column temperature, 100-220 "C; ionizing voltage, 70 eV), since they were observed in very small amounts in the reaction products. Identification of Reaction Products. The denitrogenated products were bicyclohexyl, a-ethylbicyclo[4.4.0]decane,and hexylcyclohexane. Bicyclohexyl has fragment ions at m / z 166 (P'), 82 (CsHlo),and 83 (C&11), which can be obtained from the pure compound. The IR spectrum of bicyclohexyl are in agreement with those of a known compound. The mass spectrum of hexylcyclohexane shows prominent peak series at m/z 168 (P'), 139 (CloHlg),125 (CgH17 + CH2),111 (CsHls + 2CH2),and 96 (C7H13 + 3CH2). The characteristic pattern of the methylene (-CHz-) units indicates the presence of a normal alkyl side chain in conjugation with hexane. Hexylcyclohexane has the same IR spectrum as the pure compound. Furthermore, the fragment ions of a-ethylbicyclo[4.4.O]decaneare not present at m / z 151 (P' CH3,CllHls) in the spectra but at m / z 137 (P+- C2Hs,C10H17). The 13C NMR (CDCl,) spectrum for this compound is 6 44.325 (CHJ, 40.504 (CH),37.248 (CH),36.723 (CH),35.475 (CHJ, 33.799 (CHZ),33.058 (CH,), 32.843 (CH,), 28.860 (CH,), 26.548 (CH,), 25.261 (CH3),and 25.183 (CHZ). The major hydrogenated product was 1,2,3,4-tetrahydrocarbazole for both catalysts, together with a low percentage of 1,2,3,4,4a,9a-hexahydrocarbazole, 1,2,3,6,7,8,8a,9a-octahydrocarbazole, and perhydrocarbazole in the reaction products. The mass spectrum of tetrahydrocarbazole has peaks at m/z 171 (P+, Cl2H1sN), 143 (CloHgN),130 (CgHsN),and 115 (C9H7). In addition, the melting point of this compound was 114-116 "C (115-115.5 oC).24 1,2,3,4,4a,9a-hexahydrocarbazolehas peaks at m/z 173 (P'), 143 (P+- C2Ha),130 (P+- C3H7),and 117 (P+C4H&.The mass spectrum of 1,2,3,6,7,8,8a,9a-octahydrocarbazole had a peak series at m/z 175 (P+),132 (P+- CzHsN),118 (P+C3H7N),and 106 (P+- CIH7N). Perhydrocarbazole has a major peak at m/z 179 (P'). The conformation of each compound is as reported by Masamune et al.26,26
Results Hydrodenitrogenation of Carbazole. The product distributions in the hydrodenitrogenation of carbazole on the reduced and sulfided catalysts a t 10.1 MPa total pressure are shown in parts a and b of Figure 2, respectively. The major product, bicyclohexyl, was produced from the C-N bond scission of perhydrocarbazole through the successive hydrogenation of carbazole, although hexahydrocarbazole, octahydrocarbazole, and perhydrocarbazole were barely observed as reaction products. CYEthylbicyclo[4.4.0]decane and hexylcyclohexane were also formed in the hydrogenolysis of bicyclohexyl on the sulfided Mo/A1203catalyst a t 340 "C and 10.1 MPa total pressure as described in the Experimental Section. These observations showed that bicyclohexyl was isomerized to give cr-ethylbicyclo[4.4.O]decane and decomposed to hexylcyclohexane above 300 OC. The reaction pathway for carbazole hydrodenitrogenation on the Mo/A1203catalyst is proposed in Figure 3, although there was little or no observation of the hydrogenated compounds except for tetrahydrocarbazole. (24) Adkins, H.; Coonradt, H. L. J. Am. Chem. SOC.1941, 63, 1563. (25) Masamune, T.; Koshi, M. Bull. Chem. SOC.Jpn. 1957, 30, 309. (26) Masamune, T.; Honma, G.; Ohno, M. Nippon Kagaku Zasshi 1956, 77, 1017, 1471.
Energy & Fuels, Vol. 2, No. 5, 1988 647
Hydrodenitrogenation of Carbazole
1001
1
Reaction Temperature, 'C
Figure 4. Percent hydrodenitrogenation of carbazole on the
Reaclion Temperature, 'C
reduced
( 0 )and
sulfided (0) catalyst.
Temperature, ' C
i
b
/
-
.-6 0.5 E
005
$ 1
@
aJ
K
C V
-
0 280
300
320
340
360
Reaction Temperature, 'C
Figure 2. Hydrodenitrogenation of carbazole on the (a) reduced and (b) sulfided Mo/A1203catalysts. Reaction conditions: 10.1 MPa of total pressure; 1.3 X lo-' mol/L carbazole concentration; 500 mL(STP) of H2/mol;20 mL/h of the feed solution. Key: (0) Carbazole;(0)1,2,3,4-tetrahydrocarbazole;(0) other hydrogenated compounds; (A)bicyclohexyl; (A)a-ethylbicyclo[4.4.0]decaneand cyclohexylhexane.
IS
16
17
18
1 9
10001T, K-'
Figure 5. Arrhenius plot of the denitrogenation rate of the
carbazole hydrodenitrogenation on the reduced ( 0 )and sulfided (0) catalysts.
@@*@p(*m 2% oa) H
H
Carbazole
1,2,3,4-Tetrahydrocarbazole
1,2,3,4,4ar9aHexahvdrocarbazole
H
Perhydrocarbazole
+ NH3 d-Ethylbicyclo[4.4.0ldecane
Figure 3. Reaction scheme for the hydrodenitrogenation of
carbazole.
The extent of possible hydrogenation of xylene alone on the reduced Mo/A1203 catalyst was studied a t various temperatures and a total pressure of 10.1 MPa. The extent of hydrogenation to dimethylcyclohexanes was less than 2% a t temperatures below 360 "C in the presence of carbazole under the same conditions as carbazole hydrodenitrogenation. In the absence of carbazole, the conversion of xylene to dimethylcyclohexanes rose to 11% and 18% a t 340 and 360 "C, respectively. Thus, the hydrogenation of xylenes in carbazole hydrodenitrogenation is inhibited, apparently because they are adsorbed on the catalytic sites less strongly than are the nitrogen compounds. Sulfiding of the Reduced Mo/A1203Catalyst. The percent overall hydrodenitrogenation of carbazole on the reduced and sulfided catalyst is shown in Figure 4. Sulfiding of the reduced catalyst greatly increased the percent overall hydrodenitrogenation at low temperature.
As shown in Figure 2, the formation rate of bicyclohexyl was 2.2 and 1.2 times that with the reduced catalyst at 280 and 310 "C, respectively. Sulfidation increased the concentrations of all hydrogenated compounds and bicyclohexyl at all reaction temperatures. It appears that sulfidation promotes the hydrogenation of carbazole as well as the C-N hydrogenolysis of perhydrocarbazole. Arrhenius plots of the formation rate of bicyclohexyl on both catalysts are shown in Figure 5. From the linear portions of these plots, below ca. 310 "C, the apparent activation energy for the formation of bicyclohexyl on the reduced catalyst was 140.2 kJ/mol, whereas the activation energy with the sulfided catalyst dropped to 87.3 kJ/mol (62% of that with the reduced catalyst). Thus, sulfiding led to a decrease in the activation energy for the C-N bond scission, accompanied by an increase in the formation of hydrocarbons and the hydrogenation of carbazole. The activation energy for bicyclohexyl formation was calculated on the basis of a zero-order reaction, since the power law kinetic orders varied from -0.5 to +0.8 in the range of initial concentration of carbazole of 1.5 X to 2.7 X mol/L a t temperatures of 280-310 "CS2' The hydrodenitrogenation of other model compounds such as quinolineZ1and pyridine5r8 was frequently presented in the literature in terms of first-order rate constants, but it then decreased with increasing partial pressure of the nitrogen compound.10p28 (27) To be submitted for publication. (28) Satterfield, C. N.; Yang, S. H. Znd. Eng. Proces Des. Deu. 1984, 23, 11.
Nagai et al.
648 Energy & Fuels, Vol. 2, No. 5, 1988
"
i
C
0
V
0
0
2
6
4
8
10
0
12
$ 0
Time, h
$
$
0 05
0.1
Concenlration of CS2,
I mol I L
Figure 6. Changes of the reaction products during the hydrodenitrogenation of carbazole on the sulfided catalyst at 320 "C and 10.1 MPa total pressure: (a) carbazole introduced in the reactor at 320 "C; (b) 0.13 mol/L CS2added to the xylene solution
containing carbazole; (c) CS2removed. The symbols we the same as those in Figure 2.
1
I
YEEEE 0 0
0.1
0.2
0.3
04
Concentration of Thiophene, mollLxl0'
-
0
-8I
,
1 -
0
0.05
0.1
mol I
Concentration of CS,,
;
L
0 2 c
-
n
n
P 7
0
I
0
01
Figure 8. Effect of (a) CS2 and (b) thiophene on carbazole hydrodenitrogenation on the sulfided catalyst at 320 "C. The symbols are the same as those in Figure 2.
- ,
I
u
02
Concentrat ion of CS2, mol I L
Figure 7. Effect of CS2on carbazole hydrodenitrogenationon the (a) reduced and (b) sulfided catalysts at 360 "C. The symbols are the same as those in Figure 2.
Catalyst Activity. The change in catalyst activity by the presence or absence of CS2 during the carbazole hydrodenitrogenation on the sulfided Mo/A1203 catalyst at 320 OC and 10.1 MPa is illustrated in Figure 6. Addition of 0.13 mol/L CS2to the liquid feed caused a decrease in the formation rates of the hydrocarbons, e.g., bicyclohexyl, a-ethylbicyclo[4,4.O]decane,and hexylcyclohexane. When the CS2 was removed, the formation rates of the hydrogenated compounds rose to their former levels. However, the rates of the formation of bicyclohexyl and the con-
version of carbazole did not entirely return to the same level as before the sulfur compound was introduced, as shown in Figure 6c. The catalyst was deactivated gradually, since it was used for 12 h after introduction of the carbazole solution. The phenomenon should be separated from the effect of H2S on the catalyst surface for hydrodenitrogenation during the presulfiding procedure of the reduced catalyst: after the sulfiding procedure, CS2 is competitively adsorbed on active sites to form H2S, resulting in a decrease in C-N hydrogenolysis. On the other hand, the surface of the reduced catalyst became more active during C-N hydrogenolysis by sulfiding with a small amount of CS2 (H,S) at high temperature. Effect of Sulfur Compounds. The change in the reaction product of carbazole hydrodenitrogenation on the reduced Mo/A1203 catalyst as a function of CS2 concentration at 360 "C and 10.1 MPa total pressure is shown in Figure 7a. The bicyclohexyl formation was increased slightly up to 3.3 X mol/L of CS2 together with decreasing carbazole and the hydrogenated carbazole compounds. Further addition of CS2decreased the formation of bicyclohexyl while increasing the amounts of carbazole and the hydrogenated compounds. From these results, a dual effect was observed on addition of CS2 in carbazole hydrodenitrogenation at 360 "C. Since sulfidation of the reduced catalyst increased the bicyclohexyl formation a t 360 OC as shown in Figure 7a, the addition of a small amount of CS2probably improved the activity for the C-N bond scission by sulfiding the surface of the reduced catalyst. For the sulfided catalyst even a small amount of CS2 decreased the formation of bicyclohexyl at 360 OC as shown in Figure 7b. This result showed that the presence of CS2 depressed the formation of bicyclohexyl from perhydrocarbazole when the surface of the catalyst
Energy & Fuels, Vol. 2, No. 5, 1988 649
Hydrodenitrogenation of Carbazole 0.06 1
I .o
1
z t rl, 0.01
-
a
0 2
0.1
0
Concentration of
CS,. mollL
0
0
10
20
30
Sulfur Compound, kPa
Ethanethiol; (@) thiophene; (@) dimethyl sulfide; cyclohexane. (0) (0) dibenzothiophene; (A)CS2. had been presulfided. Thus, sulfur compounds retarded the C-N hydrogenolysis in carbazole hydrodenitrogenation on both the reduced and sulfided catalysts a t high temperature. A t 320 "C, CS2 was added during carbazole hydrodenitrogenation on the sulfided catalyst in Figure 8a. The presence of CS2 increased the amounts of both the unreacted carbazole and the hydrogenated carbazole compounds and decreased the formation of bicyclohexyl and c~-ethylbicyclo[4.4.0]decane.This result showed that the C-N hydrogenolysis on the fully sulfided catalyst was inhibited by CS2. Furthermore, the effects of various sulfur compounds on carbazole hydrodenitrogenation were examined a t 320 "C. The addition of thiophene showed the same changes as that of CS2, as shown in Figure 8b. Similar observations were seen with the addition of ethanethiol, thiophene, dimethyl sulfide, dibenzothiophene, and CS2. The change in the C-N hydrogenolysis rate as a function of the amount of the additive is shown in Figure 9. CS2was the strongest inhibitor among the added sulfur compounds, since CS2 reacted with hydrogen to release two H2S molecules that inhibited active sites of the catalyst surface on the sulfided catalyst. On the other hand, Figure 10a shows the changes in the product distribution on the reduced catalyst with the addition of CS2 a t 300 "C. The addition of CS2 decreased the formation of bicyclohexyl and all hydrogenated compounds, in contrast to the result obtained at 360 "C. This same tendency was observed with the sulfided catalyst at 300 "C as shown in Figure lob. Therefore, the sulfur compounds retarded the hydrogenation of carbazole to tetrahydrocarbazole on both catalysts a t lower temperatures and probably also retarded C-N hydrogenolysis. Thus, the effect of sulfur compounds depends on the reaction temperature as well as the sulfided state of the catalyst surface. Equilibrium between Carbazole and Tetrahydrocarbazole. The equilibrium constant for carbazole hydrodenitrogenation was reported by Cocchetto and Satterfield.29 Their calculation was based on a mechanism of direct C-N hydrogenolysis of carbazole to biphenyl without preferential hydrogenation. However, since the Mo/A1203 catalyst did not directly catalyze the C-N hydrogenolysis of carbazole as shown in Figure 3, their cal(29) Cocchetto, J. F.; Satterfield, C. N. Ind. Eng. Chem. Process Des. Dew 1976, 15, 272.
b
10
Figure 9. Change in hydrocarbons formation as a function of the amount of various additives at 320 O C : The hydrocarbons and hexylcontained bicyclohexyl, ~~-ethylbicyclo[4.4.O]decane,
:I-
.o 0.5 c
e V
0
0.05
01
Concentration of CS2,
mol/L
Figure 10. Effect of CS2 on carbazole hydrodenitrogenation on the (a) reduced and (b) sulfded catalysts at 300 "C. The symbols are the same as those in Figure 2. Temperature,
4
'C
N
' 4 10-2
/
I
t 1.5
1.6
17
1.8
19
10001T. K-'
Figure 11. Estimation of the equilibrium constants for the hydrogenation of carbazole (CA) to tetrahydrocarbazole (4HCA)
on the reduced ( 0 )and sulfided (0) Mo/A120, catalyst (hydrogen partial pressure = 8.96 MPa). culation could not be used in our study. Hence, the equilibrium constants were estimated for the reaction steps in the hydrogenation of carbazole by using the ratio of the reaction products in Figure 2a,b. The solid line in Figure 11 shows the equilibrium constant for the hydrogenation of carbazole to tetrahydro-, carbazole as a function of reaction temperature. The equilibrium values at 360 "C may be inaccurate in measuring low concentrations. Equilibrium was reached a t approximately 340 "C with the reduced Mo/A1203catalyst and at about 320 "C with the sulfided catalyst. Thus, sulfidation lowered the temperature required to attain hydrogenation equilibrium. Above these temperatures, our results appeared to demonstrate the attainment of hy-
650 Energy & Fuels, Vol. 2, No. 5, 1988
2
$b0 6
t -
d b e
E o d i
I
0 5
I
1 0
Sulfur Compound, w t %
Figure 12. Tetrahydrocarbazole/carbazole ratio vs the addition of (a) CS2 for the reduced catalyst at 360 "C, (b) CS2 for the sulfided catalyst at 360 "C, (c) CS2 for the sulfided catalyst at 320 O C , (d) thiophene for the sulfided catalyst at 320 "C, (e) CS2 for the reduced catalyst at 300 "C, and (0 CS2 for the sulfided catalyst at 300 O C .
drogenation equilibrium between carbazole and tetrahydrocarbazole. The change in the tetrahydrocarbazole/carbazoleratio by the addition of the sulfur compound is presented in Figure 12. The ratios in curves a and b remained constant with both catalysts when CS2was added to the feed at 360 "C. The equilibrium ratio of 0.24 calculated from the data in Figure 11 is close to the value of 0.3 for the sulfided catalyst and that of 0.26 for the reduced catalyst. These results suggested that carbazole remained in full equilibrium with tetrahydrocarbazole a t 360 "C even in the presence of the sulfur compounds. However, at 320 "C, addition of CS2 (curve c) to the sulfided catalyst decreased the ratio from 0.39 to 0.32, while addition of thiophene (curve d) to the sulfided catalyst did not change the ratio from 0.39. These ratios were lower than the equilibrium value (0.46) calculated from Figure 11even without sulfur compound addition. Thus, the hydrogenation equilibrium among tetrahydrocarbazole and carbazole was not always affected by the presence of the sulfur compounds a t 320 "C, since the hydrogenation of carbazole to tetrahydrocarbazole was nearly in equilibrium in the absence of the sulfur compounds. At 300 OC, the reduced and sulfided catalysts yielded initial ratios of 0.24 and 0.41, respectively, compared to the calculated equilibrium value of 0.64 in Figure 11. The ratios (curves e and f) were gradually decreased with increasing amounts of added CS2. From these results, the extent of hydrogenation was far below the equilibrium level for both reduced and sulfided catalysts even in the absence of added CS2.
Discussion Is Sulfur Compound Addition an Inhibiting or Enhancing Effect? Addition of sulfur compounds decreased the C-N hydrogenolysis rate in carbazole hydrodenitrogenation at all temperatures from 300 to 360 "C. However, the presence of the sulfur compounds depressed the hydrogenation of carbazole to tetrahydrocarbazole as well as the C-N hydrogenolysis a t 300 "C, but it did not depress the hydrogenation a t 320 and 360 "C. Thus, the reaction temperature is important as to how sulfur compounds will affect the hydrodenitrogenation of carbazole. The results shown in Figures 11and 12 suggest that the hydrogenation equilibrium between carbazole and tetrahydrocarbazole is approximately reached a t 320-360 "C on both reduced and sulfided catalysts in the absence of added sulfur compounds. However, the hydrogenation is far from equilibrium even in the absence of added CS2 at 300 "C. Thus, if hydrodenitrogenation is considered to be a sequential process, with hydrogenation followed by C-N
Nagai et al. hydrogenolysis, hydrodenitrogenation is limited by hydrogenation a t 300 "C, but not at 320-360 "C under the conditions of the experiments in this study. Therefore, when the intermediate hydrogenation of carbazole was in full equilibrium at 360 "C, the sulfur compounds depressed the formation of bicyclohexyl, since the C-N hydrogenolysis became the rate-determining step in the reaction sequence a t high temperature. However, a t low temperature the additives depressed the hydrogenation which was not in equilibrium and resulted in a reduced C-N hydrogenolysis. The inhibiting effect is in agreement with the effect of sulfur and oxygen compounds for acridine hydrodenitr~genation;'~ the sulfur and oxygen compounds inhibited the C-N hydrogenolysis of perhydroacridine on the sulfided catalyst at higher temperatures, when the preceding hydrogenation of acridine to perhydroacridine attained equilibrium. A t lower temperatures, the additives decreased the hydrogenation and the C-N hydrogenolysis. Such results were also presented by Shih et al.," who reported that the C-N hydrogenolysis of o-phenylaniline, which was a final product of quinoline hydrodenitrogenation, was depressed by the presence of CS2at 342 "C and 3.4 MPa total pressure. Therefore, the presence of sulfur compounds depresses the C-N hydrogenolysis of polynuclear aromatic nitrogen compounds regardless of the type of nitrogen compound, i.e., a r-excessive nitrogen compound or a r-deficient nitrogen compound. Sulfiding Effect and Active Sites. Two types of nitrogen compounds display a different reactivity for hydrogenation, which precedes C-N bond breaking; the T deficient nitrogen compounds (acridine) are more readily hydrogenated to partially hydrogenated compounds than the r-excessive nitrogen compounds, carbazole. In spite of the reactivity difference, sulfur compounds have similar inhibiting effects on the hydrodenitrogenation of both carbazole and acridine. The sulfiding effect of the reduced catalyst, however, on carbazole hydrodenitrogenation was different from that of acridine. Sulfiding of the reduced catalyst enhanced the activity for C-N hydrogenolysis in carbazole hydrodenitrogenation, while the apparent activation energy for the C-N hydrogenolysis dropped considerably from 140.2 to 87.3 kJ/mol. More of the hydrogenated compounds were produced on the sulfided catalyst than on the reduced catalyst. A previous study revealed that a heterocyclic molecule is adsorbed o p Bransted sites and hydrogenated by chemisorbed hydrogen atoms and/or by hydrogen from the SH gr0up.l' Therefore, sulfiding of the reduced catalyst appeared to increase the number of Bransted acid sites (H+ and -SH groups). During acridine hydrodenitrogenation, on the other hand, the sulfiding greatly depressed the activity for the C-N bond scission, although the activation energy with the sulfided catalyst (156 kJ/mol) was slightly higher than that with the reduced catalyst (135 kJ/mol).16 This previous study suggested that sulfiding decreased the number and the strength of denitrogenation sites as well as increased the number of hydrogenation sites on the catalyst. From these results of the present study, sulfidation appeared to enhance the hydrogenation of carbazole and to result in acceleration of the rate of formation of bicyclohexyl. The rate of C-N hydrogenolysis with the reduced catalyst seems to be slightly limited by the hydrogenation rate a t 330 "C, since the tetrahydrocarbazole/carbazole ratio is close to the peak of the line (Figure 11) and is slightly below the equilibrium value. In the case of acri-
Energy & Fuels 1988,2, 651-653 dine, the activity for C-N hydrogenolysis diminished with a decrease in the number of denitrogenation sites, even though the intermediate hydrogenation of acridine was enhanced by sulfiding.
Conclusions (1)Sulfiding of the reduced catalyst enhanced the activity for the C-N hydrogenolysis at all reaction temperatures during carbazole hydrodenitrbgenation. (2) The addition of the sulfur compounds depressed the C-N hydrogenolysis, which became the rate-determining step when the intermediate hydrogenation attained equilibrium a t 360 "C.
651
(3) The presence of a small amount of CS2in the solution promoted the catalytic activity of the reduced catalyst for the C-N hydrogenolysis at 360 "C since CS2was converted to hydrogen sulfide, which sulfided the catalyst surface during the reaction. (4) When the intermediate hydrogenation of carbazole was not in full equilibrium with the reduced and sulfided catalyst at 300 "C, the sulfur compounds depressed the hydrogenation of carbazole, which led to decreased C-N hydrogenolysis. Registry No. Mo,7439-98-7;carbazole, 86-74-8;bicyclohexyl, 92-51-3; c~-ethylbicyclo[4.4.0]decane, 1008-17-9; hexylcyclohexane, 4292-75-5.
Coagglomeration of Athabasca Petroleum Cokes with Sulfur Sorbents as a Means of Reducing Sulfur Emissions during Combustion+ Abdul Majid," V. P. Clancy, and Bryan D. Sparks Division of Chemistry, National Research Council of Canada, Ottawa, Ontario K I A OR9, Canada Received November 9, 1987. Revised Manuscript Received May 16, 1988
The low volatiles and high sulfur content of the coke produced during the upgrading of Athabasca bitumen makes it unsuitable as a fuel. However, the coke can be used as an adsorbent to recover heavy oil or bitumen from process waste streams by using a liquid-phase agglomeration technique. It has been found that bitumen can act as a collector for certain sulfur dioxide capture agents, allowing them to be incorporated into the coke agglomerates during the oil recovery operation. Three sulfur sorbents-lime, hydrated lime, and limestone-and cokes from both Suncor delayed coking and Syncrude fluid coking operations were used in this investigation. This coagglomerated material could be used as an ancillary fuel for generation of process energy for bitumen recovery. During combustion of the coagglomerates, sulfur dioxide capture was found to depend mainly on the calcium to sulfur mole ratio, the combustion temperature, and the type of coke. The efficiencies of the three agents used in this investigation for reducing sulfur dioxide emissions have been compared.
In previous work the authors have used petroleum coke
to recover bitumen or heavy oil from waste water produced in oil sands surface mining or in situ recovery operations by means of the spherical or liquid-phase agglomeration technique.l" In this process the addition of fmely divided hydrophobic collector particles to the agitated waste water results in adsorption of the residual organic liquids by the hydrophobic solids and then the formation of agglomerates of the coke particles due to cohesion between the adsorbed liquid films. The end result of the waste-water treatment process is a cleaned water stream and a particulate solid containing the coke and oil. The latter material has improved combustion properties relative to those of the original low-volatile coke. However, the high sulfur content of this material makes it unsuitable for use as an ancillary fuel for process energy production. The objective of this investigation was to develop an economically attractive method by which the agglomerates produced during waste water treatment could be utilized 'Issued as NRCC No. 28992. Presezlted in part a t the 194th National Meeting of the American Chemical Society, New Orleans, LA, September, 1987. 0887-0624/88/2502-0651$01.50/0
directly as a boiler fuel without serious environmental damage. The combustion of these agglomerates with limestone addition, such as in a fluidized-bed reactor, could be one way to achieve the required reduction in sulfur dioxide emissions.6i6 However, it has been demonstrated'~~ that this approach requires relatively high calcium to sulfur mole ratios, even with ash recycle, to produce acceptable reductions in sulfur dioxide emissions. The development of combined fuel-sorbent pellets or briquettes for use as a sulfur dioxide control method has been reported7i8to give superior sulfur dioxide emission control during combustion. (1)Majid, A.; Ripmeester, J. A. J. Sep. Process Technol. 1983,4, 20-22. (2)Majid, A.;Ripmeester, J. A.; Sparks, B. D. Procedings of the 4th International Symposium on Agglomeration; 1985;pp 927-935. (3)Majid, A.; Sirianni, A. F.; Ripmeester, J. A. Canadian Patent 1 _. 200.778. -, 1985. ----
(4)Kumar, A.; Sparks, B. D.; Majid, A. Sep. Sci. Technol. 1986,21, 31.6-326. 315-326. - - - - - -. (5)Anthony, E. J.;Desai, D. L.; Friedrich, F. D. CANMET Report No. EPR/ERL 81-27;CANMET: Ottawa, Canada, 1981. (6) Lee, D. C.; Georgakis, C. AIChe J. 1981,27,472-481. (7) Conkle, H. H. N.; Dawson, W. J.; Rising, B. W. Proc-Inst. Briquet. Agglom., Biennial Conf. 1983,lath, 33-54. Agglom., ( 8 ) Sidney, H. M. US.Patent 4 173 454,1979. (8)
Published 1988 by the American Chemical Society
