Hydrodesulfurization Activity of MoS2 and Bimetallic Catalysts ...

42 downloads 6634 Views 867KB Size Report
Mar 2, 2007 - reduce the maximum allowable sulfur content in diesel fuel. In most advanced ... automotive exhaust emissions stimulates growing study for better catalysts for ..... The authors appreciate support from the Graduate School,.
1874

Ind. Eng. Chem. Res. 2007, 46, 1874-1882

Hydrodesulfurization Activity of MoS2 and Bimetallic Catalysts Prepared by in Situ Decomposition of Thiosalt W. Trakarnpruk*,† and B. Seentrakoon‡ Department of Chemistry, Faculty of Science, and Program of Petrochemistry and Polymer Science, Faculty of Science, Chulalongkorn UniVersity, Phyathai Road, Bangkok 10330, Thailand

Ammonium thiomolybdate [(NH4)2MoS4, ATM] and tetrabutylammonium thiomolybdate were used as precursors to prepare MoS2 catalyst. The precursor was decomposed in situ under hydrogen pressure and in the presence of decalin solvent during the hydrodesulfurization (HDS) of dibenzothiophene (DBT) and 4,6dimethyldibenzothiophene (4,6-DMDBT). The as-synthesized catalysts and the spent catalysts were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDX), and BET specific surface area measurements. All catalysts demonstrated type IV adsorption-desorption isotherms of nitrogen, and a mesoporous structure with a pore size of 3-4 nm. X-ray diffraction showed the structure of MoS2. The effect of water addition was investigated. The results revealed that the MoS2 prepared from ATM precursor showed a large increase in surface area and change in morphology when water was added, but for the MoS2 prepared from TBATM precursor, water had a negative effect. The catalytic activity is enhanced when the catalyst promoted with Co or Ni is used. The catalyst from ATM efficiently reduces sulfur content in straight run gas oil (SRGO) and light cycle oil (LCO). Introduction Stringent environmental regulations are exerting pressure to reduce the maximum allowable sulfur content in diesel fuel. In most advanced countries, the allowable limit on diesel sulfur will be restricted to 50 ppm. Sulfur content is usually reduced through a hydrodesulfurization (HDS) process using aluminasupported CoMo-based catalysts. The need to minimize the negative environmental effects of automotive exhaust emissions stimulates growing study for better catalysts for hydrodesulfurization (HDS). In the industrial process, the fuel was treated with hydrogen gas at approximately 350 °C over a cobalt- or nickel-doped molybdenum (or tungsten) sulfide catalyst supported on alumina. It is accepted that in the active structures of these catalysts, the promoter atoms are located at the edges of the MoS2 sheets in the form of so-called Co(Ni)-Mo-S structures.1 Ammonium thiomolybdate is a wellknown precursor for MoS2 catalyst via thermal decomposition.2 Some alternative methods for preparing MoS2 include lowtemperature precipitation3 and hydrothermal and solvothermal processes.4 Other precursors include alkylthiomolybdate, which can be prepared from the reaction of ammonium thiomolybdate with alklyammonium hydroxide or halide in organic solvent or in aqueous solution.5 The use of tetraalkylammonium thiometallates generates carbon-containing MoS2 and WS2 catalysts with high surface areas and high catalytic activities.6 The carboncontaining thiosalt precursor is thermally decomposed. The role of the carbon is not completely understood, but recent studies have shown that carbon is included in the arrangement of active sites, in the form of surface carbides.7 MoS2 and WS2 catalysts prepared from tetraheptylammonium thiometallate precursor showed a high selectivity along the direct desulfurization pathway, while the one with a cetyltrimethyl organic chain showed selectivity to hydrogenation.5 An unsupported MoW-Ni trimetallic catalyst (called NEBULA) was shown to have * To whom correspondence should be addressed. Tel.: (662)2187620. E-mail: [email protected]. † Department of Chemistry. ‡ Program of Petrochemistry and Polymer Science.

Figure 1. TGA curve of the decomposition of TBATM precursor.

Figure 2. XRD patterns of MoS2 catalysts formed by the decomposition of ATM and TBATM precursors: (a) ATM, (b) ATM + H2O, (c) TBATM, and (d) TBATM + H2O.

high activity, suitable for production of fuels with ultralow sulfur content (∼10 ppm).8 In this work, MoS2 was generated in situ by the thermal decomposition of ammonium thiomolybdate (ATM) and tetrabutylammonium thiomolybdate (TBATM). The bimetallic catalysts (CoMo and NiMo sulfide catalysts) were also prepared using cobalt or nickel salts and thiomolybdate precursors. They were tested for activity in the hydrodesulfurization (HDS) of

10.1021/ie061176y CCC: $37.00 © 2007 American Chemical Society Published on Web 03/02/2007

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007 1875

Figure 3. XRD patterns of CoMo sulfide catalysts formed by the decomposition of ATM and TBATM precursors and cobalt acetate: (a) Co/ATM, (b) ATM, (c) Co/ATM + H2O, (d) ATM + H2O, (e) Co/TBATM, and (f) TBATM. Table 1. Specific Surface Area, Total Pore Volume, and Elemental Analysis of the in Situ Generated Mo Sulfide and CoMo Sulfide Catalysts elemental analysis

entry

catalyst precursora

specific surf. area (m2/g)

1 2 3 4 5 6 7

ATM ATM + water Co/ATM Co/ATM + water TBATM TBATM + water Co/TBATM

120 544 77 146 270 83 219

a

total pore vol (cm3/g)

S/Mo

C/Mo

0.18 1.16 0.14 0.53 0.23 0.12 0.37

1.8 1.7 1.8 2.1 1.6 2.0 1.8

4.7 5.9 2.4 5.3 3.8 2.9 2.7

Co/Mo

0.12 0.36 0.10

Co ) cobalt acetate.

Table 2. BJH Pore Size Distributions and Mean Pore Size of the Catalysts entry

catalyst precursor

pore size distribution (nm)

mean pore size (nm)

1 2 3 4 5 6 7

ATM ATM + water Co/ATM Co/ATM + water TBATM TBATM + water Co/TBATM

3-5 3-7 3-5 3-5 3-4 2-4 2-4

3.71 4.76 3.71 4.25 3.71 3.28 3.71

dibenzothiophene (DBT) and 4,6-dimethyldibenzothiophene (4,6-DMDBT). Experimental Procedures Chemicals and Apparatus. Ammonium thiomolybdate (ATM), dibenzothiophene (DBT), and tetrabutylammonium bromide were purchased from Fluka, and 4,6-dimethyldibenzothiophene (4,6-DMDBT) was purchased from Aldrich. All other reagents were of reagent grade and were used as received. Straight run gas oil (sulfur 6100 ppm, 9.531 mmol) and light cycle oil (sulfur 310 ppm, 0.484 mmol) were donated from Thai Oil Plc., Thailand. The synthesis of tetrabutylammonium thiomolybdate (TBATM) was performed under a nitrogen atmosphere using standard Schlenk techniques. Solvents were distilled from dark purple solutions of benzophenone ketyl and stored under nitrogen atmosphere. The surface area was measured in a Quantachrome Autosorb-1 nitrogen adsorptometer by nitrogen adsorption at -196 °C using the BET method. The samples were outgassed in a vacuum at 250 °C before adsorption of nitrogen. The pore

size distribution was obtained from the desorption isotherm following the Barrett-Joyner-Hallenda (BJH) method. The X-ray diffraction patterns were obtained on a Rigaku DMAX 2002 Ultima Plus X-ray powder diffractometer with Cu KR radiation operating at 43 kV and 30 mA. The diffractograms were analyzed using standard JCPDS files. Scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDX), and X-ray mapping were performed using a JEOL JSM-5800LV for the morphology and elemental analysis. The NMR spectra were measured on a Bruker ACF 200 MHz spectrometer in deuterium oxide (D2O) at room temperature. The center peak of deuterium oxide were used as the internal reference at 4.6 ppm. Thermogravimetric analysis (TGA) was made using a Netzsch STA 409 thermobalance from TA Instruments with an increment of 20 °C/min heating rate under a N2 flow of 20 cm3/ min. IR spectra were recorded on a Nicolet FT-IR Impact 410 spectrophotometer. Total sulfur content in oil feedstocks before and after HDS reaction was determined using a SISONS X-ray fluorescence spectrometer ARL 8410 (ASTM D4294 method). For model sulfur compounds, quantitative analyses of the reaction products were performed by a DB-1 capillary column gas chromatograph GC-14B, Shimadzu, equipped with a flame ionization detector (FID) or a Varian CP-3800 gas chromatograph equipped with a flame ionization detector and a 30 m (0.25 mm i.d., 0.25 µm film thickness) CP-5 capillary column and by gas chromatography-mass spectrometry (GC-MS) using a Varian Star 3400CX and mass spectrometer Varian Saturn 4D. The GC response factor for starting material and the products were determined using pure compounds. A mass balance of about 95% was achieved in most experiments. All runs were performed in duplicate. Synthesis of (N[CH3(CH2)3]4)MoS4 (TBATM). Ammonium thiomolybdate, (NH4)2MoS4 (0.6 g, 2.3 mmol), was dissolved in 30 mL of water, and a solution of tetrabutylammonium bromide, [CH3(CH2)3]4NBr (1.48 g, 4.6 mmol), in 20 mL of water was added. The precipitating red crystals were filtered and stored under nitrogen atmosphere. The 1H and 13C NMR results for TBATM are in good agreement with the literature.9 Mo Sulfide Catalyst Preparation. The MoS2 catalyst was prepared from in situ decomposition of tetrabutylammonium thiomolybdate (TBATM) or ammonium thiomolybdate (ATM) in a Parr reactor. The catalyst precursor and decalin solvent were placed in the reactor, which was then pressurized with hydrogen to 30 atm. It was heated to 350 °C and stirred at 600 rpm for 1 h. The as-formed catalysts were separated by filtration, washed with isopropyl alcohol to remove hydrocarbon, and dried at 120 °C. In some cases a determined volume of water was also added to study the effect of water addition on the morphology of the catalyst. CoMo and NiMo Sulfide Catalyst Preparation. The above procedure was followed except that an aqueous solution of cobalt or nickel salt (cobalt acetate Co(CH3COO)2, cobalt acetylacetonate Co(C5H7O2)2, cobalt chloride CoCl2, cobalt nitrate Co(NO3)2, cobalt oxide CoO, nickel acetylacetonate Ni(C5H7O2)2, or nickel nitrate Ni(NO3)2) was also added with ATM or TBATM precursor (Co or Ni/Mo molar ratio ) 0.3) to prepare bimetallic catalysts. Catalyst Activity and Selectivity. (A) Model Compound. The HDS reactions were performed in a Parr Model 4842 highpressure batch reactor. It was loaded with substrate (DBT or 4,6-DMDBT, 2 mmol), catalyst precursor ATM, 0.052 g (0.2 mmol), or TBATM, 0.1416 g (0.2 mmol), and 50 mL of decalin solvent. Different molar ratios of water to substrate were also added. The reactor was purged three times with hydrogen gas

1876

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007

Figure 4. Adsorption-desorption isotherms of the catalysts generated from ATM + water and Co/ATM + water.

and then pressurized to 30 atm. Temperature was increased to 350 °C. The stirring speed was kept at 600 rpm. After the reaction, the catalysts were separated from the reaction mixture by filtration, washed with isopropyl alcohol to remove residual hydrocarbons, and dried at room temperature. The contents of the reaction were analyzed by GC (using 2-bromoethylbenzene as internal standard). The GC response factor for starting material and the products were determined using pure compounds. All runs were performed in duplicate. In the case of Co- or Ni-promoted MoS2 catalysts, the required amount of cobalt or nickel salt (Co or Ni/Mo molar ratio ) 0.3) was also added. (B) Oil Feedstock. The same procedure as in the model compound was performed with 50 mL of oil feedstock (straight run gas oil or light cycle oil). Remaining sulfur content was determined by the X-ray fluorescence (XRF) method. Activity was reported as percent conversion of total sulfur in oil. Results and Discussion Synthesis of (N[CH3(CH2)3]4)MoS4 (TBATM). TBATM was synthesized as in eq 1. It was characterized by FT-IR spectroscopy; the results are in good agreement with those reported.9

(NH4)2MoS4 + 2[CH3(CH2)3]4NBr f (N[CH3(CH2)3]4)2MoS4 + 2NH4Br (1) Thermogravimetric analysis (TGA) was used to characterize the thermal decomposition of TBATM (Figure 1). It was shown that the decomposition occurred in two steps: the first step in the region of 165-173 °C with a weight loss of 24.63% involved elimination of tributylamine and butylamine sulfide; the second step in the temperature range of 220-254 °C with

a weight loss of 55.07% corresponded to the elimination of dibutyl sulfide and octane. Mo Sulfide Catalyst Characterization. The MoS2 catalysts prepared from ATM or TBATM precursor decomposition were characterized by XRD, BET, and SEM. (A) X-ray Diffraction. Figure 2 shows the XRD patterns of the MoS2 catalysts formed from ATM and TBATM precursors (with and without added water). The XRD patterns are quite similar; they are in good agreement with those reported for poorly crystalline structure of MoS2 (JCPDS-ICDD 3701492, 2θ (Å) ) 14 (002), 33 (100), 40 (103), and 58 (110).10 Even though the diffraction patterns of catalysts formed from ATM and TBATM precursor with added water have similar XRD patterns, it was observed that the (002) peak of the latter is broader and has low intensity, indicating a marked decrease in the stacking along the c direction when an alkyl group is incorporated into the precursor.11 The XRD patterns of CoMo sulfide catalysts from the decomposition of ATM or TBATM precursor mixed with cobalt acetate, with and without added water, are shown in Figure 3. The XRD patterns exhibited poorly crystalline structure of MoS2, and cobalt sulfide phase was also observed. (B) Surface Area, Pore Size Distribution, and Elemental Analysis. Specific surface areas, total pore volume from BET analysis, and elemental analysis for all prepared catalysts are reported in Table 1. The catalyst from TBATM has higher surface area and pore volume compared to ATM (entry 1 vs 5). This indicates the effect of butyl group in the TBATM precursor. However, when water was added, the surface area and pore volume of the catalyst prepared from ATM precursor increased from 120 m2/g and 0.18 cm3/g to 544 m2/g and 1.16 cm3/g, respectively (entry 1 vs 2). Therefore, it is clear that water addition led to a high surface area catalyst. This finding agrees well with that reported.12 On the contrary, for TBATM,

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007 1877

Figure 5. SEM micrographs of the MoS2 catalysts prepared from ATM or TBATM precursor with and without water addition. Table 3. Conversion of DBT and Selectivitya

Scheme 1. Reaction Network for the HDS of DBT

water/catal molar ratio

DBT conversion (%)

selectivity (HYD/DDS)b

ATM ATM ATM

0 450 600

68.84 86.19 91.87

0.50 0.24 0.25

TBATM TBATM TBATM

0 450 600

88.65 30.12 19.80

0.60 0.35 0.20

catalyst precursor

a Conditions: DBT 0.368 g (2 mmol), DBT/catalyst precursor molar ratio ) 10, decalin solvent 50 mL, 30 atm H2, 350 °C, time 1 h. b HYD/ DDS ) CHB/BP; CHB ) cyclohexylbenzene, BP ) biphenyl.

it was seen that the addition of water decreased the surface area of the catalyst (from 270 to 83 m2/g, entry 5 vs 6). When the ATM or TBATM precursor was promoted with Co, the results showed that both surface area and pore volume were decreased compared with the unpromoted one. The results determined from EDX in Table 1 showed the S/Mo ratios in the range of 1.6-2.1. C/Mo ratios show strong variation (2.4-5.3) depending on the precursor, water addition, and Co promoter. The carbon detected on the surface of catalysts derived from ATM precursor comes from the organic solvent. A higher C/Mo ratio was observed in cases of ATM and Co/ ATM precursor with water addition. The deposited carbon is coke formed by the polymeric arrangement of the aromatic ring. In TBATM-derived catalyst, it comes from two sources: the organic solvent and the carbon-containing precursor.13 It was shown that the Co/Mo ratio in the catalyst prepared from Co/ ATM with added water is close to the ratio used in the precursor

preparation (Co/Mo ) 0.3), suggesting that cobalt is well dispersed on the MoS2 surface. All the adsorption-desorption curves of catalysts correspond to a type IV isotherm with desorption curves characteristic of mesoporous materials. The hysteresis loops shown by the catalysts correspond to cylindrical pores open at both ends. Only one exception is in the case of the catalyst generated from Co/ATM with added water (shown in Figure 4), which presents the adsorption-desorption isotherms of type I. It was reported that the unpromoted MoS2 prepared in situ from ATM precursor presents a type I isotherm.6 In another report, it was found that Co-promoted catalyst resulted in a type I isotherm with a low surface area of 53 m2/g.24 The BJH pore size distributions are shown in Table 2. A slightly narrower pore size distribution was obtained for the catalyst prepared from TBATM precursor.24 (C) Scanning Electron Microscopy (SEM). The MoS2 catalysts in situ generated from different thiomolybdate precur-

1878

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007

Figure 7. Conversion (A) and selectivity (B) plots showing effect of water on HDS of DBT using ATM or TBATM precursor. Table 4. Conversion of 4,6-DMDBT and Selectivitya catalyst precursor

water/catal molar ratio

4,6-DMDBT conversion (%)

selectivity (HYD/DDS)b

ATM ATM ATM ATM

0 750 1200 2000

47.20 93.02 100.00 93.70

0.55 0.36 0.23 0.18

a Conditions: DMDBT 0.212 g (1 mmol), 4,6-DMDBT/catalyst precursor molar ratio ) 10, decalin solvent 50 mL, 30 atm H2, 350 °C, time 2 h. b HYD/DDS ) MCHT/DMBP; MCHT ) methylcyclohexyltoluene, DMBP ) dimethylbiphenyl.

Scheme 2. Reaction Network for the HDS of 4,6-DMDBT

Figure 6. SEM micrographs of Co-promoted MoS2 catalysts.

sors with and without water addition show different morphologies, as shown in Figure 5. The SEM micrographs of the in situ generated MoS2 catalyst prepared by decomposition of ATM (Figure 5a) shows a flat and smooth surface with stacking layers of MoS2. On the other hand, the catalyst formed by adding water (Figure 5b) reveals a highly porous and rough surface. This is in good agreement with that reported.14 The catalyst prepared from decomposition of TBATM (Figure 5c) has a cheeselike morphology with a porous system. This is produced by the internal pressure generated by the vaporization of the organic alkyl groups during the decomposition steps of TBATM precursor under high pressure of hydrogen and high temperature.15 However, when water is added (Figure 5d), cavities appear much smaller, indicating that the water addition retards the formation of an organized porous system. SEM micrographs of the Co-promoted MoS2 catalysts are shown in Figure 6. The catalyst prepared from ATM precursor

(Figure 6a) shows different morphology from that of the unpromoted one (in Figure 5a): stacking layers were lost and agglomerates were observed on the surface. When water was added (Figure 6b), the catalyst exhibited a porous and rough surface. For the Co-promoted catalyst prepared from TBATM (Figure 6c), smaller cavities (compared to that in Figure 5c) were observed and walls between cavities appear thicker than those in the MoS2 catalyst. Catalyst Activity and Selectivity. (A) Mo Sulfide Catalyst. Results of activity and product selectivity of the HDS of DBT are reported in Table 3, and the conversion and selectivity plots are displayed in Figure 7. The main reaction products are biphenyl (BP) and cyclohexylbenzene (CHB), as shown in Scheme 1. The mean standard deviation for catalytic measurements was about 5%. The HDS of DBT yields biphenyl through

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007 1879

Figure 8. Conversion (A) and selectivity (B) plots showing effect of water on HDS of 4,6-DMDBT. Table 5. Conversion of DBT and Selectivity Using MoS2 Catalysts Promoted with Different Metal Saltsa catalyst precursor ATM ATM ATM + water TBATM TBATM TBATM TBATM TBATM TBATM TBATM

promoter

DBT conversion (%)

selectivity (HYD/DDS)b

none Co(CH3COO)2 Co(CH3COO)2 none Co(CH3COO)2 CoO Co(NO3)2 Co(C5H7O2) 2 Ni(NO3)2 Ni(C5H7O2) 2

42.12 73.93 94.35 63.48 85.14 80.71 78.76 73.40 73.69 70.58

0.50 0.38 0.39 0.60 0.32 0.36 0.25 0.38 0.39 0.31

a Conditions: DBT 0.368 g (2 mmol), DBT/catalyst mole ratio ) 20, decalin solvent 50 mL, 30 atm H2, 350 °C, time 1 h. b HYD/DDS ) CHB/ BP; CHB ) cyclohexylbenzene, BP ) biphenyl.

the direct desulfurization pathway (DDS) and cyclohexylbenzene (CHB) through the hydrogenation pathway (HYD). CHB is a secondary product along this pathway obtained by a C-S bond breaking reaction from tetrahydrodibenzothiophene, an intermediate product formed by hydrogenation of one of the aromatic rings of DBT. Since these two pathways are parallel,16 the ratio between HYD and DDS can be approximated in terms of the selectivity by taking the ratio of CHB/BP. It should be mentioned that in this work tetrahydrodibenzothiophene was not detected. In this work, the reaction without catalyst was also conducted; the DBT conversion was very low (6.7%). For the catalyst prepared from ATM precursor, the DBT conversion is 68.84%. Our result is higher than that reported (34% conversion at the reaction condition of 350 °C, 30 atm H2 in 5 h).6 The catalyst derived from TBATM showed higher activity (88.65%) than ATM. This might be due to the higher surface area of TBATM. Our result is in contrast with that reported, in which the in situ decomposition of tetraalkylammonium precursors do not improve the HDS activity for unpromoted catalysts.6 Upon addition of water to the catalytic runs in the case of ATM precursor, the conversion increased significantly. This can

Figure 9. Conversion (A) and selectivity (B) plots showing HDS of DBT using promoted catalyst from ATM precursor.

be explained as that ATM was dissolved in water which was dispersed in fine droplets in decalin under agitation and then decomposed upon heating under hydrogen pressure to produce finely dispersed MoS2 catalyst.17 In addition, water might affect the activity of the catalyst by altering the phase equilibria in the reactor or directly modifying the nature of the active catalytic species.18 In the case of TBATM precursor, water has a negative effect on conversion. This can be explained as that TBATM is not soluble in water, so the dispersion did not occur. As seen from the surface area and pore volume results previously mentioned (in Table 1), both values are decreased markedly when water was added in the reaction. In addition, the porous structure was destroyed as seen from the SEM (in Figure 5). Considering the selectivity, it can be observed that hydrogenolysis (DDS) is dominant over hydrogenation (HYD). This result agrees with previous literature findings.19 It was observed that addition of water decreased the HYD/DDS ratios from 0.50.6 to 0.20-0.35, indicating more hydrogenolysis. For the more refractory sulfur compound, 4,6-DMDBT (for which the HDS was conducted in a longer time (2 h) than in DBT), it can be seen from Table 4 and the conversion and selectivity plots displayed in Figure 8 that the same results as in the case of DBT were obtained. The major product from the HDS of 4,6-DMDBT was dimethylbiphenyl (DMBP) with methylcyclohexyltoluene (MCHT) as a minor product. DMBP is the product from the DDS route. MCHT is a secondary product obtained by C-S bond breaking reaction from tetrahydrodimethyldibenzothiophene (THDMDBT) (which was not detected in this work), an intermediate product formed by hydrogenation of one of aromatic rings of 4,6-DMDBT, shown in Scheme 2. Desulfurization from this molecule is faster than

1880

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007

Figure 10. Conversion (A) and selectivity (B) plots showing HDS of DBT using promoted catalyst from TBATM precursor. Table 6. Specific Surface Area, Total Pore Volume, and Elemental Analysis of the Spent Catalysts after HDS Reaction elemental analysis

Figure 11. XRD patterns of spent catalysts after HDS reaction: (a) Co/ ATM, (b) ATM, (c) Co/ATM + H2O, (d) ATM + H2O, (e) Co/TBATM, and (f) TBATM.

entry

catalyst precursora

specific surf. area (m2/g)

1 2 3 4 5 6 7

ATM ATM + water Co/ATM Co/ATM + water TBATM TBATM + water Co/TBATM

111 332 68 122 105 48 183

a

that from 4,6-DMDBT, because the methyl group in the hydrogenated ring can be rotated away from the sulfur atom and because the C-S bond is weaker after hydrogenation.20 As in the case of DBT, addition of water changes the HYD/DDS ratio (MCHT/DMBP) from 0.55 to 0.18. Moreover, further increasing water addition results in a decrease of conversion. It might be that water could hinder the creation of the active sulfide species or compete with the reactant via competitive chemisorption on surface sites of the catalyst. (B) Co- and Ni-Promoted MoS2 Catalysts. The catalysts promoted with Co or Ni were synthesized in situ by mixing a solution of thiomolybdate with metal compound (Co or Ni/Mo mole ratio ) 0.3) and tested their activities were tested. In this set of experiments, the DBT/catalyst precursor molar ratio was set to 20 (ratio of 10 in the former experiments) and reaction time of 1 h in order to see the difference in conversion change more clearly. The catalytic results are shown in Table 5 and displayed in Figures 9 and 10.

total pore vol (cm3/g)

S/Mo

C/Mo

0.15 0.63 0.14 0.45 0.17 0.12 0.36

1.6 1.6 1.6 2.2 1.3 1.8 2.0

5.4 7.1 4.8 5.6 2.7 2.3 3.5

Co/Mo

0.10 0.36 0.30

Co ) cobalt acetate.

From the results in Table 5, it was revealed that both Coand Ni-promoted catalysts present higher catalytic activities than their MoS2 counterparts. Cobalt shows higher activity than Ni. Similar results reported that Co/MoS2 catalyst usually shows higher activity than Ni/MoS2 catalyst for HDS of DBT.21 The high activity of bimetallic catalysts can result from a highly dispersed heterometallic sulfide phase. The synergetic effect of Co or Ni on the catalytic activity of the MoS2 has been well documented.22 The order of the promoter activities is Co(CH3COO)2 > CoO ∼ Co(NO3)2 > Co(C5H7O2)2 > CoCl2. It was reported that among different cobalt precursors (nitrate, chloride, acetate, and acetylacetonate), the use of cobalt acetate resulted in a number of small cobalt particles dispersed throughout the MCM-41 support.23 The run using ATM precursor mixed with cobalt acetate and added water shows the highest activity. It might be due to a much better dispersion of Co sulfide phase,

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007 1881

Figure 12. SEM micrographs of spent catalysts after HDS reaction.

evidenced from the XRD pattern (intensity of Co sulfide peaks decreased in Figure 3c). Co or Ni promotion not only increases the activity but also modifies the selectivity along the two pathways. Similar to the effect of water addition, Co or Ni promoter enhances the direct desulfurization route to bipohenyl. This result is in good agreement with that reported.24 Characterization of the Spent Catalysts after the HDS Reaction. After running the HDS reaction, the spent catalysts were separated from the reaction mixture by filtration, washed with isopropyl alchool to remove residual hydrocarbons, and dried at room temperature. They were characterized by XRD, SEM, and elemental analysis (the data are shown in Figures 11 and 12). From XRD in Figure 11, the patterns of the unpromoted and promoted catalysts after the reaction present weaker peaks, indicating less crystallinity. In the case of Co-promoted catalysts, intensities of cobalt sulfide phase decreased; this is in good agreement with the previous report.25

From SEM in Figure 12, the surface morphology of all catalysts after the reaction shows agglomerates. The specific surface area, total pore volume, and elemental analysis are collected in Table 6. The BET surface area of the catalysts decreased considerably after HDS reaction. It was evident that carbon deposits were responsible for the decrease in the surface area. Additionally, the decrease of surface area might be attributed to sintering of MoS2 crystallites.26 Hydrodesulfurization of Real Oil Feeds. To utilize the prepared catalyst in real oil feeds, hydrodesulfurization of straight run gas oil (SRGO) or light cycle oil (LCO) was performed using ATM precursor (0.248 g, 0.9 mmol), 50 mL of oil, at 350 °C at 30 atm H2 pressure, 2 h. Water was added in the reaction using a water/catalyst precursor molar ratio of 1200. The results demonstrated that the conversion of total sulfur (determined by XRF method) is 53.28% and 67.74% for SRGO and LCO, respectively.

1882

Ind. Eng. Chem. Res., Vol. 46, No. 7, 2007

Conclusions In this work a simple method was used to prepare unsupported MoS2 catalysts without the use of sulfiding agent by decomposition of ATM or TBATM. It was shown that both thiomolybdate precursors catalyzed the hydrodesulfurization of DBT and 4,6DMDBT, the carbon-containing precursor. TBATM showed higher activity; this might be due to its higher surface area. Addition of water to the reaction caused a morphology change in the catalysts, and showed a positive effect in catalytic activity in the case of ATM precursor but a negative effect in the case of TBATM. It was found that the water/catalyst precursor molar ratio is a factor influencing percent conversion. In addition, cobalt and nickel were shown to promote catalytic activity and change product selectivity. Acknowledgment The authors appreciate support from the Graduate School, Chulalongkorn University, and oil feedstocks from Thai Oil Plc., Thailand. Literature Cited (1) Chianelli, R. R.; Daage, M. Structure-function relations in molybdenum sulfide catalysts: The “Rim-Edge” model. J. Catal. 1994, 149, 414. (2) Kalthod, D. G.; Weller, S. Studies of molybdenum sulfide catalysts: Effects of pretreatment on sintering, stoichiometry, and oxygen chemisorption. J. Catal. 1985, 95, 455. (3) Chianelli, R. R.; Dines, M. B. Low-temperature solution preparation of Group 4B, 5B and 6B transition-metal dichalcogenides. Inorg. Chem. 1978, 17, 2758. (4) (a) Peng, Y. Y.; Men, Z. Y.; Zhong, C.; Lu, J.; Yu, W. C.; Yang, Z. P.; Qian, Y. T. Hydrothermal synthesis of MoS2 and its pressure-related crystallization. J. Solid State Chem. 2001, 159, 170. (b) Li, W. J.; Shi, E. W.; Ko, J. M.; Chen, Z. Z.; Ogino, H.; Fukuda, T. Hydrothermal synthesis of MoS2 nanowires. J. Cryst. Growth 2003, 250, 418. (5) Alonso, G.; Siadati, M. H.; Berhault, G.; Aguilar, A.; Fuentes, S.; Chianelli, R. R. Synthesis of tetraalkylammonium thiometallate precursors and their concurrent in situ activation during hydrodesulfurization of dibenzothiophene. Appl. Catal., A: Gen. 2004, 263, 109. (6) Alvarez, L.; Espino, J.; Omelas, C.; Rico, J. L.; Cortez, M. T.; Berhault, G.; Alonso, G. Comparative study of MoS and Co/MoS catalysts prepared by ex situ/in situ activation of ammonium and tetraalkylammonium thiomolybdates. J. Mol. Catal. A: Chem. 2004, 210, 105. (7) Alonso, G.; Valle, M. D.; Cruz, J.; Petranovskii, V.; Licea-Claverie, A.; Fuentes, S. Preparation of MoS2 catalysts by in situ decomposition of tetraalkylammonium thiomolybdates. Catal. Today 1998, 43, 117. (8) Soled, S. L.; Miseo, S.; Krycak, R.; Vroman, H.; Ho, T. C.; Riley. K. (Exxon Research & Eng. Co.). U.S. Patent 6,299,760, 2001. (9) Alonso, G.; Valle, M. D.; Cruz, J.; Petranovskii, V.; Licea-Claverie, A.; Fuentes, S. Preparation of MoS2 catalysts by in situ decomposition of tetraalkylammonium thiomolybdates. Catal. Today 1998, 43, 117. (10) Song, C.; Yoneyama, Y.; Kondam, M. R. Method for preparing a highly active, unsupported high-surface area MoS2 catalyst. U.S. Patent 6,156,693, 2000. (11) Alonso, G.; Berhault, G.; Aguilar, A.; Collins, V.; Ornelas, C.; Fuentes, S.; Chianelli, R. R. Characterization and HDS activity of

mesoporous MoS2 catalysts prepared by in situ activation of tetraalkylammonium thio-molybdates. J. Catal. 2002, 208, 359. (12) Yoneyama, Y.; Song, C. Promoting effect of H2O addition on C-O bond cleavage and hydrogenation of dinaphthyl ether over MoS2 catalyst in situ generated from ATTM. Energy Fuels 2002, 16, 767. (13) Alonso, G.; Petranovskii, V.; Valle, M. D.; Cruz-Reyes, J.; LiceaClaverie, A.; Fuentes, S. Preparation of WS2 catalysts by in situ decomposition of tetraalkylammonium thiotungstates. Appl. Catal., A: Gen. 2000, 197, 87. (14) Alonso, G.; Aguirre, G.; Rivero, I. A.; Fuentes, S. Synthesis and characterization of tetraalkylammonium thiomolybdates and thiotungstates in aqueous solution. Inorg. Chim. Acta 1998, 274, 108. (15) Alonso, G.; Del Valle, M.; Cruz, J.; Licea-Claverie, A.; Petranovskii, V.; Funtes, S. Preparation of MoS2 and WS2 catalysts by in situ decomposition of ammonium thiosalts. Catal. Lett. 1998, 52, 55. (16) Topsoe, H.; Clausen, B. S.; Massoth, F. E. Hydrotreating Catalysis; Springer: Berlin, Heidelberg, New York, 1996. (17) Song, C.; Saini, A. K. Strong synergistic effect between dispersed Mo sulfide catalyst and H2O for low-severity coal hydroliquefaction. Energy Fuels 1995, 9, 188-189. (18) Siewe, C. N.; Ng, F. T. Hydrodesulfurization of cold lake diesel fraction using dispersed catalysts: Influence of hydroprocessing medium and sources of H2. Energy Fuels 1998, 12, 598. (19) (a) Girgis, M. J.; Gates, B. C. Reactivities, reaction networks, and kinetics in high-pressure catalytic hydroprocessing. Ind. Eng. Chem. Res. 1991, 30, 2021. (b) Pawelec, B.; Navarro, R.; Fierro, J. L. G.; Vasudevan, P. T. Studies of molybdenum sulfide catalyst ex ammonium tetrathiomolybdate: effect of pretreatment on hydrodesulfurization of dibenzothiophene. Appl. Catal., A: Gen. 1998, 168, 205. (20) Prins, R.; Egorova, M.; Rothlisberger, A.; Zhao, Y.; Sivasankar, N.; Kukula, P. Mechanism of hydrodesulfurization and hydrodenitrogenation. Catal. Today 2006, 111, 84. (21) Knudsen, K. G.; Cooper, B. H.; Topsoe, H. Catalyst and process technologies for ultra low sulfur diesel. Appl. Catal., A: Gen. 1999, 189, 205. (22) Furimsky, E. Role of MoS2 and WS2 in hydrodesulfurization. Catal. ReV.sSci. Eng. 1980, 22, 371. (23) (a) Panpranot, J.; Kaewkun, S.; Prasertthdam, P.; Goodwin, J. G. Effect of cobalt precursors on the dispersion of cobalt on MCM-41. Catal. Lett. 2003, 91, 95. (b) Martinez, A.; Lopez, C.; Marques, F.; Diaz, I. FischerTropsch synthesis: the influence of various cobalt compounds applied in the preparation of unsupported cobalt catalysts on their performance. J. Catal. 2003, 220, 486. (24) Nava, H.; Ornelas, C.; Aguilar, A.; Berhault, G.; Fuentes, S.; Alonso, G. Cobalt-molybdenum sulfide catalysts prepared by in situ activation of bimetallic (Co-Mo) alkylthiomolybdates. Catal. Lett. 2003, 86, 257. (25) Alonso, G.; Espino, J.; Berhault, G.; Alvarez, L.; Rico, J. L. Activation of tetraalkylammonium thiotungstates for the preparation of Nipromoted WS2 catalysts. Appl. Catal., A: Gen. 2004, 266, 29. (26) Kouzu, M.; Uchida, K.; Kuriki, Y.; Ikazaki, F. Micro-crystalline molybdenum sulfide prepared by mechanical milling as an unsupported model catalyst for the hydrodesulfurization of diesel. Appl. Catal., A: Gen. 2004, 276, 241.

ReceiVed for reView September 7, 2006 ReVised manuscript receiVed January 23, 2007 Accepted January 27, 2007 IE061176Y