New method for preparing highly effective catalyst for hydrodesulfurisation (HDS)

Tiancun Xiao, Jianwen Da, and Sergio L Gonzalez Cortes Malcolm.L.H. Green Wolfson Catalysis Centre, Inorganic Chemistry Laboratory, Oxford University, South Parks Road, Oxford, UK, OX1 3QR, email: xiao.tiancun@chem.ox.ac.uk Hamid Al-Megren, Soliman Al-Khowaiter Petroleum and Petrochemicals Research Institute, King Abdulaziz City for Science and Technology, P. O. Box 6086, Riyadh 11442, Saudi Arabia.

Abstract
A new process, e.g., organic matrix combustion-decomposition (OMXC) has been developed to prepare differently supported molybdenum and cobalt oxide catalyst for hydrodesulfurisation, and the catalysts have been tested under both lab conditions using thiophene HDS as the model reaction, and industrial conditions for gasoline hydrodesulfurisation. The catalysts at different stages were characterized using X-ray diffraction (XRD), Laser Raman, and transmission electron microscope (TEM). The results show that carburization pretreatment of the supported bimetallic oxide give lower HDS performance of resulted catalysts, generally speaking, the alumina supported catalyst has higher HDS activity then the SiO2 and TiO2 supported ones. Among the catalyst tested, gama-alumina supported molybdenum sulfide catalysts prepared suing OMXC method has the highest activity for HDS, and lower hydrogenation, which may result from the higher dispersion of MoS2 and the stack layers of the MoS2. The MoS2 cluster in the catalyst has more uniform distribution. The dispersion and cluster size of MoS2 can be tailored through the fuel/metal ratio, combustion temperature and atmosphere. More than 99% of sulfur can be removed

in the HDS of naphtha at 300oC and 2.0MPa in one pass. The catalyst prepared using OMXC method has a unique surface structure and vibration. The amount of the active component and promoters can be easily controlled in the process, which may be breakthrough in hydrotreating catalyst preparation.

Introduction
The increasing strict environmental regulation requires deep removal of the heteroatoms (S, N, metals) from the crude oil and feedstock, which is regarded as one of the key operations for the production of clean transportation fuels. In a recent American EU and EPA regulatory change,1 a 10-fold reduction in fuel sulfur emission is required, which leads to an urgent demand for the improved hydrodesulfurization (HDS) catalysts. Transition metal sulphide catalysts, which are relatively cheap and easy to prepare, have been used in the hydrotreatment for a long time. However, the currently available catalysts can not deeply remove the heteroatoms particularly sulphur and nitrogen from the crude oil and process raw fractions. Hence it is becoming more and more important and new challenge to develop catalysts with high HDS ability. For this purpose, a lot of efforts have been made to further understand the effect of preparation factors on the catalyst performance and new process or formulations are been developed in the supported sulfide catalyst preparation 2-7. The main aim is to find a catalyst with high performance to remove the heteroatoms from oil, particularly sulfur, which is the source of acid rain. For a comprehensive understanding the catalyst structure and performance of the commercially important catalysts, and the future development, the reviews can be referred to a series of literatures 8-10. Among these efforts, the most promising way is to improve and modify the late group VI metal sulfide system, due to its remarkable activity.3, 4, 6, 11-21 It is proposed that the MoS2 basal plane is largely inert, while the edges of the MoS2 layer expose Mo atoms for reaction.21 Therefore, the catalytic activity is strongly dependent on the specific dispersion of the MoS2 crystal (i.e., the ratio of edge sites to basal plane area) and on the local environment of the active edge sites. Hence the key factor in the catalyst preparation is to increase the dispersion of molybdenum sulfide, and thus expose more edge of MoS2 layer to the reaction. The traditional synthesis of the supported later group VI metal sulfide catalyst promoted by Co or Ni is to co-impregnate or sequentially impregnate the transition metal (such as Mo or W) and promoter salts onto the supports (for example, Al2O3,
2

SiO2, AC, molecular sieves), followed by calcinations (producing oxides), and sulfurisation under the flow of H2S in H2 prior to use.11-16 The HDS activity could be modified by using different transition metal promoter precursors, and also by changing the sequence of impregnation of molybdenum and promoter ions on alumina.3 New materials 22-26 and techniques such as sonochemical preparation11, metathesis reactions
18

and decomposition of metal sulfur compounds 19 or new promoters like noble metal

(Ru, Pt, Rh)23, 24 have also been explored to sulfur removal efficiency through the dispersion and performance of the sulfide catalysts. Although the catalysts prepared in these ways show super performance for HDS, the preparation process is either time consuming and difficult to scale-up, or the cost of the catalyst is too high to be acceptable. Hence the preparation of Mo-based supported catalyst with higher

concentration of active metals on the exterior surface and high dispersion still represents a synthetic challenge in the field of advanced catalysis. 8,19 In our work, on the basis of solid state chemistry study, a new but simple and practical process has been developed to prepare supported transition metal sulfide catalyst promoted by the other transition metals. The catalyst samples have been characterized at different stages, and tested for HDS of thiophene and gasoline under different conditions. The catalysts show unique structure and superior performance in HDS.

Experimental
Catalyst Preparation: Catalysts were prepared using OMXC, which is widely used in sintering and cement solid solution material preparation.26 The difference of our work from the reference is that we used urea as the media, and a specific amount of water is added in the precursor of the catalyst solution. Also the calcinations

temperature and atmosphere is controlled to give the high metal dispersion and exposure over the catalyst surface. For example, in preparing Al2O3 supported Co and Mo oxide catalysts, 1.0 g of dried Al2O3 (-Al2O3, BET surface area: 150m2/g) is mixed with 0.16g ammonium molybdate, 0.1g Co(NO3)6.6H2O and 0.4g urea, the mixture
3

system was then added with a small amount of water and heated to 50oC while stirring until all the cobalt nitrate and ammonium molybdate were loaded on the support. This was then transferred to a 500oC Muffle oven and combusted for 10 minutes. The oxide form of the alumina supported CoMo catalyst was thus obtained. Table 1. The composition of Mo-Based catalysts prepared using combustion method* Catalyst MoO3 wt% Mo wt% Co3O4 Co wt% symbol wt% (3-1) 12.87 8.58 0.00 0.00 (3-2) 12.87 8.58 0.00 0.00 (3-3) 12.59 8.39 2.08 1.64 (3-4) 12.59 8.39 2.08 1.64 (3-3-1) 14.15 9.43 2.17 1.70 (3-3-2) 17.21 11.47 2.09 1.65 (3-3-3) 10.91 7.27 2.24 1.76 (3-3-4) 14.23 9.48 1.63 1.28 (3-3-im) 12.59 8.58 2.08 1.64 *each specific catalyst specification and activation will be described in details in the following parts. Catalyst Characterisation The analysis of the catalysts for carbon and sulfur was carried out on the elemental analyzer Vario EL. The samples were digested through oxidative combustion and the gases analyzed with a thermal conductivity detector. The characterization of crystalline phases of the oxidic precursors and the catalysts before and after reaction was carried out using X-ray diffraction (XRD) at room temperature in a Philips PW1710 diffractometer equipped with an X-ray tube (CuK radiation = 1.5406 ; 40 kV, 30 mA). The specimens were prepared by grinding small quantities of each sample an agate mortar and pestle and then loading into a flat sample holder. The data were collected at room temperature, in / reflection mode, scanning the specimens between 3-70 2, using steps of 0.05, with time per step of 1.25 s. The Raman spectra were recorded in a Yvon Jobin Labram spectrometer with a 632 nm HeNe laser, run in a back-scattered co focal arrangement. The samples were pressed in a microscope slide; the scanning time was set to 30 s. Raman spectra were
4

recorded in air at room temperature with a resolution of 2 cm-1 and scanning range of 100 to 2500 cm-1. Several locations on each catalyst surface were probed to ensure homogeneity of the sample and reproducibility of the data. High resolution transmission electron microscopy (HRTEM) was carried out using a JEOL 4000FX electron microscope with accelerating voltage of 400 kV. The sulfurised bimetallic catalysts were ground to a fine powder and dispersed in AR-grade chloroform. Then, they were placed in an ultrasonic bath for ca. 15 min, before a drop of the suspension was put on a lacey carbon-coated copper grid (Agar, 20 mesh). The samples were examined using a JEOL 2010 electron microscope fitted with a LINK Pentafet energy dispersive X-ray microanalysis (EDX) system to verify semi-quantitative compositional and homogeneity. At least eight representative images were taken for each sample and the length of around 150 slabs was measured in order to determine the distribution of slab lengths and the stacking degree. The average or mean slab length (L) and stacking degree (N) were calculated according to the first moment of the distribution:

niMi / ni
i =1 i =1

Where Mi is the slab length or number of layers (slabs) of a MoS2 unit and ni is the number of slabs or stacks in a determined range of length or stacking number. BET specific surfaces were measured in a Sorptomatic 1990 (ThermoQuest) based on the static volumetric principles, using the multiple-point method and nitrogen as adsorption gas. The sample was degassed at 170 C in vacuum (10-3 torr) for ~16 hours, before exposing to the analysis conditions.
Catalytic performance on thiophene HDS reaction.

The thiophene HDS tests under lab conditions were carried out in a fixed-bed tubular quartz reactor (34 cm long, 4 mm i.d.) with an axial thermowell containing a thermocouple centred in the catalyst bed. The required amount of catalyst precursor to yield 200 mg of oxidic precursor (<250 m) was loaded between quartz wool plugs. The activation or sulfurisation process was made using a hydrogen flow (ultra high
5

purity) of 20 cm3.min-1, controlled by a Brooks mass flow meter (model 5850TR), that pass through a thiophene (99+ %, Aldrich) saturator at room temperature (10.9 kPa). The temperature is increased linearly from room temperature to 450 C at 5 C.min-1 and held at 450 C for 10 h. Then, the sample was treated with pure H2 flow at 450 C for 1 h and cooled down to 350 C. The catalytic tests were carried out under atmospheric pressure (0.1 MPa) and the reactants consisted of flowing H2 at a rate of 20 cm3.min-1 saturated with thiophene vapor at 0 C (2.96 kPa). The effect of catalyst activation ways on the catalyst performance was also studied in this work. The activation ways includes carburisation and sufurisation. The carburisation process was carried out by treating the Al2O3 supported CoMo oxide with flowing 20%CH4/H2. The products were analyzed using an on-line HP 5890 Series II chromatograph equipped with a Hayesep R packed column (80-100 mesh, 3m x 1/8 x 2mm SS) and a flame ionization detector (FID).After completion of the catalytic reaction, the catalyst sample was cooled to room temperature under flowing argon and then passivated before exposure to the atmosphere.

Results and Discussion 1. Catalyst evaluation results

Comparison of different catalyst for thiophene HDS at 350 (thiophene 0, H2 20 ml/min) after sulfurization

Table 2 Catalytic performance of the supported Mo and CoMo catalysts for thiophene HDS at 350oC and 1 bar Catalyst Preparation Thiophene HDS rate, Method Conv, % mmol/s.g Commercial catalyst A Impregnation 66.3 0.27 CoMo/Al2O3 Commercial Catalyst B, Impregnation 46.4 0.19 CoMo/Al2O3 CoMo/Al2O3(3-3-im) impregnation 15.8 0.07 Mo/Al2O3(3.1) OMXC 31.1 0.13 Mo/TiO2(3.2) OMXC 22.1 0.09 CoMo/Al2O3(3.3) OMXC 79.1 0.33 CoMo/TiO2(3.4) OMXC 17.8 0.07 The activity test results show that thiophene conversion over alumina supported
6

CoMoS catalysts prepared using OMXC method is up to 79.1%, and the main product is C4, the catalyst activity is much higher than the two commercial catalysts under the test conditions. While over the alumina supported CoMoS catalysts prepared using co-impregnation method, thiophene conversion is only 15.8%. The results also show that alumina is a better support than titania, may be due to the smaller surface area of the TiO2. It is also shown that the addition of Co in the supported molybdenum catalyst can improve the catalyst performance significantly, which is in agreement with the reference results (3-5).
Comparison of different catalyst for thiophene HDS at 380

Raising the reaction temperature to 380oC increased the thiophene conversion over the all the catalysts, and the catalytic performance trend is similar to these at 350oC, suggesting that the catalyst mechanism is not changed. However, the alumina

supported CoMo catalyst prepared using OMXC method still has the highest thiophene conversion and HDS reaction rate. The one pass conversion of thiophene is up to 80%, which is the best among the reported catalysts till now. Table 3 Catalytic performance of the different catalysts for thiophene HDS at 380oC and 1 bar. Catalyst Preparation Thiophene HDS rate, Method Conv, % mmol/s.g Commercial Catalyst A Impregnation CoMo/Al2O3* Commercial catalyst B Impregnation 52.96 0.22 CoMo/Al2O3 CoMo/Al2O3 (3-3-im) impregnation 24.3 0.10 Mo/Al2O3 (3.1) OMXC 34.1 0.14 Mo/TiO2 (3.2) OMXC 28.5 0.12 CoMo/Al2O3 (3.3) OMXC 79.9 0.33 CoMo/TiO2 (3.4) OMXC 25.4 0.12
Comparison of different catalysts for thiophene HDS at 400oC

When the reaction temperature is increased to 400oC, the conversion of thiophene over CoMoS/Al2O3 (3.3) prepared using OMXC method is up to 80.3%, rising by 0.4% compared with the reaction at 380oC. However, thiophene conversion over the other catalysts increases more than this, particularly the MoS/Al2O3 (3.1). These results
7

suggests that rising the reaction temperature may change the less active sites of the sulfide catalysts into active states. The Co promoted molybdenum sulfide catalyst have been active at low temperature, and the increase of reaction temperature may have less effect on the active sites, but may increase the reaction rate, hence to improve the catalyst performance. However, the CoMoS/Al2O3 (3.3) still has the highest activity and reaction rate among all the catalyst. The activity of commercial catalyst B increased more significantly by increasing the reaction temperature. Table 4 Thiophene HDS over the catalyst prepared using different methods at 400oC and 1 bar. Catalyst Preparation Thiophene HDS rate, Method Conv, % mmol/s.g Commercial catalyst A Impregnation 68.9 0.27 CoMoS/Al2O3* Commercial catalyst B Impregnation 62.5 0.26 CoMoS/Al2O3 CoMoS/Al2O3(3-3-im) impregnation 36.09 0.15 MoS/Al2O3(3.1) OMXC 41.1 0.17 MoS/TiO2(3.2) OMXC 29.2 0.12 CoMoS/Al2O3(3.3) OMXC 80.3 0.33 CoMoS/TiO2(3.4) OMXC 17.8 0.07 Although the increase of the reaction temperature can increase the activity of the catalyst for thiophene HDS, the industrial operation often requires lower temperature to depress the catalyst deactivation and increase the selectivity. Hence obviously the alumina supported CoMoS catalyst prepared using OMXC method has the advantages in industrial application.
Comparison of the CoMo/Al2O3 catalyst with different active metal content prepared by same combustion method for thiophene HDS at 350

The composition of Mo and Co has a significant effect on the catalyst performance for thiophene HDS. The CoMoS/Al2O3 catalyst containing 11.47%Mo and 2.24%Co prepared using OMXC method has the highest thiophene conversion and reaction rate. While decreasing Mo content and increasing Co content is not good for the improvement of the catalyst performance, this is because Mo plays as the active component, and the Co is a promoter for the reaction.
8

Table 5 Effect of composition of the catalysts on thiophene HDS Catalyst 3-3-1 3-3-2 3-3-3 3-3-4 Mo loading wt%
9.43 11.47 7.27 9.48

Co loading wt%
2.17 2.09 2.24 1.63

Preparation Method OMXC OMXC OMXC OMXC

Pretreatment Sulfurization Sulfurization Sulfurization Sulfurization

Thiophene Conv, % 78.4 80.1 60.1 62.9

HDS rate, mmol/s.g 0.33 0.34 0.25 0.26

Comparison of CoMo/Al2O3 catalyst for thiophene HDS (thiophene 0, H2 20 ml/min) after different pretreatment methods

It has been reported that carbide catalysts are very active in HDS and HDN. 20 To compare the activity of the CoMo/Al2O3 catalyst prepared using impregnation and combustion method, the catalyst in its oxide form was carburised and sulfurised

Table 6 Effect of states of the active component on the performance of the catalysts Catalyst Preparation Method 3-3 3-3 3-3 3-3 3-3-im 3-3-im 3-3-im 3-3-im Combustion Combustion Combustion Combustion Impregnation Impregnation Impregnation Impregnation React. Temp, 350 380 350 380 350 380 350 380 Sulfurization Sulfurization Carburization Carburization Sulfurization Sulfurization Carburization Carburization Pretreatment Thiophene Conv, % 79.1 79.3 28.7 34.0 15.8 24.3 12.0 20.4 HDS rate, mmol/s.g 0.33 0.33 0.12 0.14 0.07 0.10 0.05 0.09

according to reference procedures. The catalyst performance was evaluated and the results are listed in table 6. The catalyst prepared using the combustion method has higher activity in both sulfide and carbide form. However, the sulfide catalysts have much higher performance for thiophene conversion than carbide. This suggests that the sulfide is more active than carbide for HDS. The catalyst prepared using combustion method has better performance for HDS than that prepared using impregnation method. The comparison of the carbide catalyst and sulfide catalysts prepared using combustion
9

and impregnation methods for HDS are shown in Fig. 1. Apparently, the sulfide catalyst has higher thiophene conversion, and reaction temperature has less effect on the HDS performance on the sulfide than on the carbide catalysts.
90 80 Thiophene conv., % 70 60 50 40 30 20 10 0
350 380 350 380

Sulfide catalyst
C

I: Impregnation C: Combustion

carbide catalyst

Reaction temperature (oC) Fig. 1. Performances of CoMo/Al2O3 prepared by different methods and pretreated by different conditions tested at different temperature.
The results of the catalyst test under industrial conditions

The above catalyst tests were carried out under atmosphere pressure conditions and using the simple molecule, e.g., thiophene HDS as the model reaction. To study the performance of the catalyst under real condition and compare with some other available industrial catalysts, the catalysts were tested for naphtha hydrodesulfuristion. The reaction conditions and the HDS results are shown in Table 7. It is shown that the commercial catalysts A and B can only decrease sulfur content in the naphtha from 239 ppm to 10.3 and 6.85 ppm in one pass. When using Oxford catalyst with the same composition under the same reaction conditions, sulfur can be decreased to 4.7 ppm, and when increasing the fuel to metal ratio to 3 and 4 in the OMXC preparation, sulfur can be further removed from naphtha, and the catalytic performance is significantly improved. These results suggest that increase the fuel to metal ratio in the preparation stage can improve the catalyst performance, which is believed to be due to the increase of the dispersion of molybdenum and cobalt oxide and leads to higher exposure on the catalyst surface. The Hydrogen temperature programmed reduction results show that the reduction peak
10

temperature of CoMoOx/Al2O3 appears at lower values, and the hydrogen consumption is higher for the catalyst prepared using OMXC with urea to metal ratio of 4. This also suggests that the urea to metal ratio influence the interaction between the molybdenum and cobalt oxide and the support.
Table 7. Comparison of the performances of different HDS catalysts Catalysts Composition Preparation Sulfur content removal method in treated oil efficiency Commercial CoMoS/Al2O3 Imprgnation . 10.3 ppm 95% catalyst A Commercial CoMoS/Al2O3 Co-impregnati 6.85 ppm 96.5% Catalyst B CoMoS/Al2O3 OMXC 4.7 ppm 98.1% (fuel/metal=2) Oxford Catalyst CoMoS/Al2O3 OMXC 3.2 ppm 98.7% (fuel/metal=3) CoMoS/Al2O3 OMXC 1.8 ppm 99.5% (fuel/metal=4) Reaction conditions: GHSV: 8 h-1, H2/oil volume ratio: 200:1, T: 300oC, 2.0MPa. The feedstock is naphtha containing 239ppm of sulfur.

Catalyst characterization and discussion

The XRD patterns of alumina supported CoMoO catalysts with different Co and Mo content in the oxide precursor prepared using OMXC method are given in Fig. 2. When MoO3 content changes from 10-17wt%, the XRD patterns of the catalyst samples are almost the same. Very weak peaks at 11.8, 26.3 were seen, which is due to the diffraction of highly dispersed molybdenum oxide, the peak are very weak and broad, and no peak due to cobalt oxide was seen, suggesting that the Co and Mo oxides are highly dispersed in the alumina support. The peaks appearing at 35.6, 45.9 and 66o are due to the peak of the diffraction of alumina support, and the sharp diffraction peaks and 38, 44 and 64o are due to the sample holder of aluminum badge. These results suggest that the OMXC method can leads to high dispersion of the active component, e.g. MoO3 and promoter such as cobalt oxide on the alumina support.

11

350 300

Intensity, a.u.

250 200 150 100 50 0 10

3-3-1 3-3-2 3-3-3 3-3-4

20

30

40

50

60

70

2
Figure 2. The XRD patterns of CoMoOx/Al2O3 catalysts prepared by OMXC method.

200

3-3, combustion 3-3-im, impregnation Commercial catalyst B

150

Intensity, a.u.

100

50

10

20

30

40 2

50

60

70

Fig. 3 XRD patterns of the CoMoOx/Al2O3 oxide catalyst prepared by different methods. The XRD patterns of the MoCoOx/Al2O3 catalysts prepared using different methods are compared in Fig. 3. The small peaks corresponding to MoO3 are seen in all the samples, however, in the 3-3 combustion sample, the MoO3 is up to 22.4wt%, and Co3O4 2.5wt%, while in the impregnation method, the MoO3 is only 12.6wt%, and commercial catalyst B contains 15wt% MoO3, which also has relatively strong
12

diffraction of MoO3. These results suggest that the OMXC method gives higher MoO3 dispersion over alumina support. The laser Raman spectra (Fig. 4) of the CoMoOx/Al2O3 prepared using different methods show that no Raman peaks of MoO3 (800~900cm-1) in commercial catalyst B (Fig. 4). The peak at 1562.5 cm-1 may be the surface vibration of a kind additive. This indicates that the surface of the commercial catalyst B is covered by a kind of additives in its oxide form. In the CoMoO/Al2O3 prepared using combustion method, strong but broad peaks are seen in the region of 600~900 cm-1, which are assigned to the vibration of surface MoO3. The low frequency and broad peaks may be resulted from the distorted coordination environment of the MoO3 in the catalyst, which is probably the main reason for the high performance of the catalysts. In the CoMoOx/Al2O3 prepared using combustion method, a strong and sharp peak at 937.7 cm-1, which is assigned to the vibration of the Mo-O-Mo in the crystalline MoO3. This suggests that the conventional impregnation method can not disperse the active component MoO3 over the support surface very well.
3-3-1, combustion 3-3-im, im pregnation Commercial Catalyst B Commercial catalyst A

2500 2000 1500

937.7

1602.2

Intensity, a.u.

1000 500 0 -500 -1000 500

812.4 875 410.6 364.6 789.5 846.7

1562.5

1000

1500
-1

Raman Shift, cm

Fig 4. The Laser Raman spectroscopy of the CoMoO/Al2O3 catalysts prepared using different methods. The changes of laser Raman spectra with the composition of the CoMoO/Al2O3 catalyst are shown in Fig. 5. It is seen that the Raman vibration mode and frequency are not changed with the surface composition, but the surface composition has influence on
13

the intensity of the Raman peaks, suggesting that the combustion method can help to disperse most of the active component over the catalyst support.
792.1

1000 900 800 700

3-3-1 3-3-2 3-3-3 3-3-4

Intensity, a.u

600 500 400 300 200 100 0 500 1000 1500

Raman Shift, cm

-1

Fig. 5. The Laser Raman spectroscopy of CoMoO/Al2O3 catalysts prepared using OMXC methods. The TEM graphs o f the sufurised MoCoOx/Al2O3 prepared using different methods and urea/metal ratios are shown in Fig. 6. The multi-layers are the characteristics of MoS2 materials, which can be easily distinguished from the amorphous alumina support. It is seen that the MoCoSx/Al2O3 catalyst prepared using impregnation method has a thick stack of MoS2, which covers alumina support. And the length of the MoS2 is also very long, suggesting a high crystalline phase and lower dispersion of MoS2 over the support (Fig. 6-a). In the catalyst prepared using OMXC method with urea/(Co+Mo) =2, the MoS2 appears more disordered fewer stacks of MoS2, also very small and short MoS2 with fewer layers are dispersed over the alumina support (Fig. 6-b). This suggests that the OMXC method gives higher dispersion of MoS2, which are active sites for the HDS reaction. When urea/(Co+Mo) ratio is increased to 4 in the OMXC preparation, the sulfurised sample has the highest activity for naphtha HDS reaction. The TEM observation also shows that the catalyst has the highest MoS2 dispersion, and the high dispersed MoS2 covers the alumina support, which may be the responsible for the high activity.

14

5 nm

Fig. 6. TEM images of different HDS catalysts after sulfurisation prepared using a) impregnation, b) OMXC urea/(Mo+Co)=2, c) OMXC, with urea/(Mo+Co)=4

a
10 nm

5 nm

c
15

Conclusion
A new method has been developed to prepare supported CoMo bimetallic catalyst for HDS reaction. The method is simple and easy control, and the amount of Mo and Co can be adjustable using this method. The molybdenum oxide is highly dispersed over the alumina support, and less non-active aluminum molybdate was formed in the catalyst. In the sulfide form, it is seen that the catalyst has more MoS2 exposed to the surface, comparing with the commercial catalysts. The prepared MoCoOx/Al2O3 catalysts using OMXC method has much higher activity and stability in thiophene HDS than the commercial catalysts under the reaction conditions, and carburization of the catalyst does not give higher activity for thiophene HDS than the sulfurisation. Alumina is a better support for the HDS catalyst than the others such as TiO2 and SiO2. Under the industrial pilot plant test conditions, the catalyst prepared using OMXC method shows high sulfur removal efficiency and can decrease sulfur in the naphtha to about 2ppm. These results are in agreement with our lab scale test results. Also, the catalyst performance can be adjusted by changing the urea to metal ratio at the starting organic matrix. The catalyst prepared using impregnation method has a long and big particle size of MoS2 distributed over the alumina support. However, the MoS2 in the catalyst prepared using OMXC method show high dispersion with smaller size. This leads to more active MoS2 active sites over the catalyst, which may be responsible for the high performance.

16

References
1. 2. Chem. Eng. News 2000, 78 (1), 927.

Okamoto, Yasuaki; Ochiai, Keiji; Kawano, Masatoshi; Kobayashi, Keiko; Kubota, Takeshi, Applied Catalysis, A: General (2002), 226(1-2), 115-127

3.

Ishihara, Atsushi; Lee, Jeayoung; Dumeignil, Franck; Higashi, Ryuichirou; Wang, Anjie; Qian, Eika W.; Kabe, Toshiaki., Journal of Catalysis (2003), 217(1), 59-68.

4.

Okamoto, Yasuaki; Ishihara, Shin-Ya; Kawano, Masatoshi; Satoh, Masaru; Kubota, Takeshi., Journal of Catalysis (2003), 217(1), 12-22.

5.

Sakashita, Y.; Araki, Y.; Shimada, H., Applied Catalysis, A: General (2001), 215(1-2), 101-110.

6.

Kouzu, Masato; Kuriki, Yasunori; Sakanishi, Kinya; Farag, Hamdy; Saito, Ikuo; Sugimoto, Yoshikazu, Abstracts of Papers, 226th ACS National Meeting, New York, NY, United States, September 7-11, 2003 (2003), FUEL-003.

7.

Kabe, T.; Aoyama, Y.; Wang, D.; Ishihara, A.; Qian, W.; Hosoya, M.; Zhang, Q., Applied Catalysis, A: General (2001), 209(1,2), 237-247.

8. 9.

Chianelli, R. R., Catal. Rev.-Sci. Eng. 1984, 26, 361 (). Prins, R., De Beer, V. H. J., and Somorjai, G. A., Catal. Rev.Sci. Eng. 1989, 31, 1.

10. Wiegand, B. C., and Friend, C. M., Chem. Rev. 1992, 92, 491. 11. (a) Furimsky, E. Appl. Catal. A 1998, 171, 177. (b) Grange, P. Catal. ReV.-Sci.
Eng. 1980, 21, 135.

12. Vakros, J.; Kordulis, C., Appl. Catal., A 2001, 217, 287-293. 13. (a) Harris, S.; Chianelli, R. R. J. Catal. 1986, 98, 17. (b) Halbert, T. R.; Ho, T. C.; Stiefel, E. I.; Chianelli, R. R. J. Catal. 1991, 130, 116. 14. Dhas, N. Arul; Ekhtiarzadeh, Arash; Suslick, Kenneth S. J. Am. Chem. Soc., 2001, 123(34), 8310. 15. Sarbak, Z., Appl. Catal., A 2001, 216(1-2), 9. 16. Vasudevan, P. T.; Fierro, J. L. G. Catal. ReV.-Sci. Eng. 1996, 38, 161.
17

17. (a) Dungey, K. E.; Curtis, M. D. J. Am. Chem. Soc. 1997, 119, 842., (b) Brito, J. L.; Barbosa, A. L. J. Catal. 1997, 171, 467. 18. Bonneau, P. R.; Jarvis, R. F.; Kaner, R. B. Nature 1991, 349, 510. 19. (a) Zdrazil, M. Catal. Today 1988, 3, 269. (b) Vicic, D. A.; Jones, W. D. J. Am.
Chem. Soc. 1999, 121, 7606. (c) Afanasiev, P.; Xia, G. F.; Gilles, B.; Jouguet, B.;

Lacroix, M.; Chem. Mater. 1999, 11, 3216. 20. Norskov, J. K.; Clausen, B. S.; Topse, H. Catal. Lett. 1992, 13, 1. 21. (a) Bouwens, S. M. A. M.; Koningsberger, D. C.; de Beer, V. H. J.; Louwers, S. P. A.; Prins, R. Catal. Lett. 1990, 5, 273. (b) Ozkan, U. S.; Zhang, L.; Shuangyao, N.; Moctezuma, E. J. Catal. 1994, 148, 181. 22. Kushmerick, J. G.; Weiss, P. S. J. Phys. Chem. B 1998, 102, 10094. 23. Pinto, P.; Calhorda, M. J.; Straub, T.; Miikkulainen, V.; Raty, J.; Suvanto, M.; Pakkanen, T. A., J. Mol. Catal. A: Chem. 2001, 170, 209. 24. Cattenot, M.; Geantet, C.; Glasson, C.; Breysse, M., Appl. Catal., A 2001, 213(2), 217. 25. Aruna, S. T.; Muthuraman, M.; Patil, K. C. Mater. Res. Bull. 2000, 35(2), 289. 26. Aegerter, Paul A.; Quigley, Wes W. C.; Simpson, Garth J.; Ziegler, David D.; Logan, John W.; McCrea, Keith R.; Glazier, Samantha; Bussell, Mark E. J. Catal. 1996, 164(1), 109.

18