24.02.2013 Views

HYDRODESULFURIZATION OF THIOPHENE OVER BIMETALLIC ...

HYDRODESULFURIZATION OF THIOPHENE OVER BIMETALLIC ...

HYDRODESULFURIZATION OF THIOPHENE OVER BIMETALLIC ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>HYDRODESULFURIZATION</strong> <strong>OF</strong> <strong>THIOPHENE</strong> <strong>OVER</strong><br />

<strong>BIMETALLIC</strong> Ni–Mo SULFIDE CATALYSTS PREPARED<br />

BY DIFFERENT METHODS<br />

Hamid A. Al-Megren*<br />

Petroleum and Petrochemicals Research Institute, King Abdulaziz City for Science and<br />

Technology, P.O. Box 6086 Riyadh 11442, Saudi Arabia<br />

Hamid A. Al-Megren<br />

: ةـصلاخلا<br />

ة���يكيناكيملا (IMPR) بير���شتلا ق���يرط ةطا���سوب ند���عملا يئا���نث Ni–MO/Al2O3 د���يتيربك ن���م تاز ّ◌ّـ���فحم ري���ضحت - ث���حبلا اذ���ھ ي���ف - م���ت د���قل<br />

ةع�شأ فار�حناو ،نيجورتينلا�ب يئا�يزيفلا زاز�تملاا ةطا�سوب تاز�فحملا صاو�خ د�يدحت ﱠمت امك . (OMXC) ةيوضعلا Matrix قرـح ةقيرطو (MECH)<br />

ت�يربكلا عزن ةيلمع يف اھتيلاعف ةساردو ةرضحملا تازفحملل ةترْبَك ةيلمع تيرجأ دقو . نامار فايطمو ينورتكللإا سجملا ةطاسوب قيقدلا ليلحتلاو<br />

سكإ<br />

ة��يلاعف ر�ثكأ (OMXC) ة��قيرطب هري�ضحت م��ت يذ�لا ز��فحملا نأ�ب د��جُو د�قو . م٣٥٠<br />

ةرار��ح ة�جردو د��حاو يو�ج طغ��ض ت�حت نيفو��يثلل (HDS) ة�جردھلاب<br />

. IMPR و MECH قرطب ةرضحملا تازفحملا عم ةنراقم % ٥٧ يلاوحب<br />

ABSTRACT<br />

Bimetallic Ni–Mo sulfide Al2O3 catalysts were prepared by impregnation (IMPR), mechanical (MECH) and<br />

organic matrix combustion (OMXC) methods. The prepared catalysts were characterized by nitrogen physisorption,<br />

X-ray diffraction, electron probe microanalysis, and Raman Spectroscopy. The samples were sulfided and their<br />

activity in hydrodesulfurization (HDS) of thiophenes at a pressure of 1 atm and a temperature of 350ºC was tested. It<br />

was found with Ni–Mo that the catalyst prepared by the OMXC method is more active (by upto 57%) as compared<br />

with the corresponding catalysts prepared by (MECH) and conventional impregnation methods.<br />

Key words: bimetallic HDS catalyst, preparation, characterization<br />

* E-mail: almegren@kacst.edu.sa<br />

Paper Received: 17 November 2007; Accepted 12 November 2008<br />

The Arabian Journal for Science and Engineering, Volume 34, Number 1A January 2009 55


56<br />

Hamid A. Al-Megren<br />

<strong>HYDRODESULFURIZATION</strong> <strong>OF</strong> <strong>THIOPHENE</strong> <strong>OVER</strong> <strong>BIMETALLIC</strong> Ni–Mo SULFIDE<br />

CATALYSTS PREPARED BY DIFFERENT METHODS<br />

1. INTRODUCTION<br />

At present, the hydrodesulfurization (HDS) process is an efficient method for sulfur removal from gasoline and<br />

diesel of high sulfur content [1,2]. In the HDS process, organic sulfur compounds in the liquid fuels are broken<br />

down to H2S using NiMo–CoMo/Al2O3 catalysts [3–6], followed by the removal of H2S, and the sulfur content in<br />

fuels can be reduced to about 10 ppm. Sulfided Ni–Mo catalysts are widely used in hydrotreating reactions [7,8].<br />

Several studies have shown that monometallic cobalt catalyst are active for hydrodesulfurization (HDS) reactions<br />

[9–11]. For example, Duchet et al. investigated carbon supported molybdenum and tungsten catalysts for HDS of<br />

thiophenes [12]. In some cases, however, their HDS activities are low and unimportant due to a low degree of cobalt<br />

dispersion and/or strong interactions with some support materials like alumina [13]. Therefore it is of great interest<br />

to examine the influence of the preparation method on the properties and activity of the catalyst, as well as on the<br />

interaction between promoters and supports [14]. In the present investigations, emphasis was focused on studying<br />

the effect of the preparation method on the hydrodesulfurization activity of Ni–Mo catalysts.<br />

To improve the performance of the HDS catalyst, all steps in the preparation procedure should be considered.<br />

The key parameters are the choice of a precursor of the active species, support, synthesis, and post-treatment of the<br />

synthesized catalyst. The nature of the active phase can be modified by changing the amount of active component<br />

[15], and by the introduction of additives. Numerous additives have been studied, and phosphorus [16, 17] and<br />

fluorine [18–21] have received special attention. In some studies, the sulfides of transition metals were replaced by<br />

nitrides [22] or carbides [23, 24]. Noble metals are also used as the active phase in hydrotreating catalysts for second<br />

stage HDS [25].<br />

Various supports have been used to enhance the catalytic activity in HDS: carbon [26–28], silica [29, 30],<br />

zeolites [31–34], titania and zirconia [35–37], and silica–alumina [38, 39]. Combining new types of catalytic species<br />

with advanced catalyst supports such as ASA (amorphous silica–alumina) can result in an extremely high<br />

desulfurization performance. The application of ASA-supported noble metal-based catalysts for the second-stage<br />

deep desulfurization of gas oil is an example [25]. The Pt and Pt–Pd catalysts are very active in the deep HDS of prehydrotreated<br />

straight run gas oil under industrial conditions. These catalysts are able to reduce the sulfur content to 6<br />

ppm, while simultaneously reducing the aromatics to 75% of their initial amount. At high sulfur levels, the ASA<br />

supported noble metal catalysts are poisoned by sulfur and Ni–W–ASA catalysts become preferable for deep HDS<br />

and dearomatization.<br />

It has long been recognized that the optimal performance of a catalyst depends on its preparation method [40].<br />

The common preparation methods require a combination of different unit operations. Some of these are not entirely<br />

understood. They can be described as: (i) introduction of the metal precursor on the support by impregnation or ionexchange<br />

[41], co-precipitation and deposition precipitation [42]; (ii) drying and calcination, and (iii) reduction. The<br />

size and distribution of metal crystallites in the support, the homogeneity of the multi-component catalyst, the<br />

porosity, surface area, and pore size distribution are all functions of the precursors used in the preparation. Some<br />

preparation parameters such as temperature of treatment, pH of the solution and the aqueous or organic medium also<br />

contribute to the catalyst preparation. Recently, significant progress has been made towards understanding the<br />

relationship between the preparation method and final properties of the catalyst and supports [41]. Generally, it is<br />

believed that the catalyst selectivity depends on the particle size of the dispersed metal and spatial distribution of the<br />

active component in the porous support. The stability of the catalyst depends on the interaction of the metal and the<br />

supports.<br />

The controlled preparation of heterogeneous catalysts has been studied for more than 100 years [42]. However,<br />

perfectly designed catalysts still remain a challenge. Impregnation, sol–gel, ion-exchange, precipitation, and<br />

deposition-precipitation are the commonly used methods [43].<br />

The impregnation method requires the support to be contacted with a certain amount of solution of a metal<br />

precursor, usually a salt, and then it is aged, usually for a short time, dried, and calcined [41]. This method appears to<br />

be simple, and economical (especially when using solutions of costly active components), and able to yield a<br />

reproducible catalyst. However, when drying, the dissolved metal compounds often crystallize in the support, which<br />

leads to larger crystallites and thus leading to low metal dispersion.<br />

The ion–exchange method consists of exchanging either hydroxyl groups or protons from the supports, with<br />

cations and anions in the solutions [41]. In this method, it is important to adjust the pH values to control the<br />

interaction between the support and metal precursors. A knowledge of the isoelectric point is essential if meaningful<br />

dispersions are to be obtained.<br />

January 2009 The Arabian Journal for Science and Engineering, Volume 34, Number 1A


Hamid A. Al-Megren<br />

The sol–gel method is the latest technique for catalyst preparation, in which the metal–organic precursors are<br />

mixed with metal precursor to form a homogeneous solution [43]. The solution is passed through the hydrolysis and<br />

polymerization process by adding water while carefully controlling the pH and reaction temperature. This results in<br />

the formation of colloidal particles and micelles of approximately 10 nm diameter. These particles continue to<br />

increase in size until a metal oxide gel is formed.<br />

However, each of these techniques is either difficult to control or the preparation process is complex and difficult<br />

to use on an industrial scale. Also, none of them ever takes into consideration the adjustment of the interaction<br />

between the active metal and support, which is regarded as one of the properties of supported catalysts.<br />

Also, for hydrotreating catalyst preparation, the content and particle sizes of the active components such as Mo or<br />

W are crucial for the catalyst performance. No methods are available to load the high content of the Mo or W needed<br />

with desirable particle size. Hence, selective deep removal HDS now needs noble metal catalysts, which are much<br />

more costly and easy to deactivate.<br />

Recently, a new catalyst preparation method, i.e. organic matrix combustion (OMXC) has been developed [44,<br />

45]. Some organic compounds are mixed with soluble metal compounds to form a homogeneous solution with<br />

addition of minimum water (or other solvent) where necessary. The particle size of the catalyst is controlled by the<br />

ratio of the organic compounds to the metal in the solution. Interaction between the active component and support is<br />

controlled by the calcinations temperature and atmosphere.<br />

The aim of this work is to develop more robust catalysts with high activity for selective removal of sulfur from<br />

thiophene, to hydrogenate the C–S bond; and with low ability to hydrogenate the mono C=C bonds. The key to<br />

achieving this goal is to control the distribution and particle size of the active component, e.g. MoS2, and optimize<br />

the promotion of Ni and Mo in the catalysts supported over alumina. Active component precursors will be chosen<br />

from the ammonium salts. This catalyst will be prepared and compared by three different preparation methods.<br />

2. EXPERIMENTAL<br />

2.1. Catalyst Preparation<br />

In this work, 15% Ni–Mo bimetallic material supported over alumina was prepared using three different<br />

methods: impregnation (IMPR), mechanical (MECH), and organic matrix combustion (OMXC). All catalysts were<br />

prepared in oxide phase then sulfurized using a stream of hydrogen saturated by thiophene at room temperature.<br />

Nickel nitrate hexahydrate Ni(NO3)2.6H2O(99.0% Fluka) was mechanically mixed with molybdenum trioxide at<br />

a Ni/Mo atomic ratio of 0.43. An amount of 15% from the mixture was added to alumina (150 mesh, 150 m 2 g -1 , from<br />

Aldrich). The mixture was dried in a vacuum oven at 120ºC and calcined at 550ºC for 15 hours. The resulted powder<br />

sample was named as MECH.<br />

Nickel nitrate hexahydrate was dissolved in small amount of water, mixed with ammonium<br />

heptamolybdatetetrahydrate (NH4)6Mo7O24-4H2O (99% Riedel-de Haën) then impregnated with alumina. The<br />

mixture was dried in a vacuum oven and calcined at 550ºC for 15 hours. The resulted powder sample was named as<br />

IMPR.<br />

The same ratio of Ni/Mo supported over alumina was prepared using OMXC method by mixing calculated<br />

amount of urea (Fisher Chemicals) with ammonium heptamolybdate-tetrahydrate and nickel nitrate hexahydrate<br />

(Ca.1g/2ml = (Ni- and Mo-Precursors+urea)/ water ratio. This mixture was stirred to form a homogeneous slurry,<br />

then mixed with Al2O3 support and heated at 50ºC for 2-3 hours to obtain urea-based slurry containing molybdenum<br />

fully loaded over alumina. This paste is ignited at 500ºC in static air for 10 min. The resulted powder sample was<br />

named as OMXC.<br />

2.2. Catalytic Tests<br />

All catalysts were activated by sulfurization using 20 ml of hydrogen through a jacketed saturator containing<br />

thiophene (10.9 kPa). The flow of hydrogen with thiophene passed through 200mg of catalyst placed in a fixed bed<br />

reactor and heated at 5ºC/min to maximum temperature 450ºC, and was then held at this temperature for 3 hours.<br />

The catalytic behavior tests of the prepared supported catalysts (MECH, IMPR, and OMXC) for the<br />

hydrodesulfurization of thiophene were carried out in a laboratory bench scale pilot plant fitted with mass flow<br />

meters and a fixed bed down-pass flow quartz reactor placed in a cylindrical furnace, equipped with a coaxial<br />

thermocouple. The inner diameter for the reactor was 4 mm and its length 300 mm. The catalyst zone in the middle<br />

of the reactor had a length of 30 mm and was filled with 200 mg of catalyst diluted with an equal amount of quartz<br />

particles (50 – 100 mesh) in order to minimize temperature gradients. In the center of the catalyst bed, a<br />

thermocouple was installed in contact with the catalyst particles to measure the reaction temperature. The heating<br />

zones at the inlet and the outlet of the reactor were measured by a thermocouple located inside the furnace and were<br />

controlled by a temperature controller (Cole Parmer Digi-Science).<br />

The Arabian Journal for Science and Engineering, Volume 34, Number 1A January 2009 57


58<br />

Hamid A. Al-Megren<br />

Normally, the flow rate of hydrogen was 20 ml/min at atmospheric pressure. The activity of nickel–molybdenum<br />

supported catalysts (MECH, IMPR, and OMXC), and thiosophene conversion were studied at a temperature of<br />

350°C<br />

Blank reactor runs were conducted as described above in the absence of catalysts. No significant conversions of<br />

thiophenes were observed under the conditions of each experiment. Carbon mass balancing closed within 8%, most<br />

within 5%.<br />

2.2.1. Gas Chromatography<br />

The analysis of the reaction mixture was carried out using, a Hewlett–Packard 5890 series II. The instrument<br />

equipped with 1 × 1/8" O.D. stainless steel packed column (80 – 100 mesh range Porpak S); and flame ionization<br />

detector (FI), and thermal conductivity detector (TCD).<br />

The instrument was calibrated by external standardization method using standard mixture of (CH4, C2H6, C2H4,<br />

C3H8, C3H6, C4H10, C4H8, and C4H8).<br />

2.3. Catalyst Characterization<br />

Electron probe micro-analyzer (EPMA) (JEOL JXA-8900R/RL) was used for identification of constituent<br />

elements in the prepared samples and to study the elements distribution over the support. Analysis was performed<br />

by illuminating the sample surface with a finely focused electron beam and measuring the wavelengths and<br />

intensities of characteristic X-rays from the specimen and the quantities of secondary electrons and backscattered<br />

electrons from it.<br />

The X-ray diffraction (XRD) was carried out in Philips PW1710 diffractometer with CuKα radiation. The data<br />

were collected at room temperature, in θ/θ reflection mode, scanning the specimen between 3º–70º in 2θ, using steps<br />

of 0.05º, with a time per step of 1.25 s.<br />

The structure of the samples was studied using laser Raman spectroscopy from a Jobin Yvon Labram<br />

spectrometer with a 632.8 nm HeNe laser, run in a back-scattered confocal arrangement. The samples were pressed<br />

in a microscope slide; the scanning time was set to 30s. Raman spectra were recorded in air at room temperature<br />

with a resolution of 2 cm –1 and a scanning range of 100 to 2500 cm –1 .<br />

2.3.1. Physisorption<br />

Nitrogen adsorption–desorption isotherm of MECH, IMPR, and OMXC catalyst samples were measured on a<br />

Micromeretics ASAP2010 system using nitrogen sorption at 77K. Prior to the experiments, approximately 0.2 g of<br />

sample was degassed at 523K for 16h. The surface areas were calculated using a multipoint Brunauer–Emmett–<br />

Teller (BET) model. The pore size distribution was obtained using BJH and BET models and the total pore volumes<br />

were estimated at a relative pressure of 0.99 atm, assuming full surface saturation with nitrogen.<br />

For characterization purpose, samples of the fresh and used catalyst were passivated at room temperature in flow<br />

of 1% O2/He before they were exposed to air.<br />

3. RESULTS AND DISCUSSION<br />

The surface area measurements for the three prepared bimetallic catalyst samples (MECH, IMPR, and OMXC) in<br />

oxide phase, before sulfurization, are shown in Table 1. The surface areas for the samples increase in following<br />

order: MECH (125.5 m 2 /g) > OMXC (122.5 m 2 /g) > IMPR (119.3 m 2 /g); the differences are relatively small. The<br />

highest surface area for the sample prepared by mechanical method could be attributed to less dispersion between the<br />

active component and surface of the support. In Table 1, the average pore diameter and the total volume, measured<br />

for the three samples, were the highest for the samples prepared by mechanical and the lowest for the sample<br />

prepared by combustion method. The particle size distribution analysis shows that the samples prepared by IMPR<br />

and OMXC have almost similar size (~60μm), much higher than the particle size for sample prepared by MECH<br />

(~27μm).<br />

Table 1. Physical Properties of NiMoOx Catalyst Prepared by Impregnation, Mechanical, and Organic Matrix<br />

Methods<br />

Preparation Surface<br />

Method Area m 2 g -1<br />

Total Pore Volume<br />

Cm 3 g -1<br />

Average Pore Particle Size<br />

Diameter A º Distribution μm<br />

IMPR 119.27 0.16529 55.436 59.81<br />

OMXC 122.50 0.16420 53.617 63.53<br />

MECH 125.51 0.18621 59.343 26.81<br />

January 2009 The Arabian Journal for Science and Engineering, Volume 34, Number 1A


Count/arb, units<br />

500<br />

400<br />

300<br />

200<br />

100<br />

Hamid A. Al-Megren<br />

0<br />

0 10 20 30 40 50 60 70<br />

2 q<br />

Figure 1. XRD pattern for Ni3Mo7Ox/Al2O3 catalysts prepared by Impregnation, Mechanical and OMXC methods<br />

The Arabian Journal for Science and Engineering, Volume 34, Number 1A January 2009 59<br />

Al<br />

Al<br />

Al<br />

Al<br />

Al<br />

OMXC<br />

Al<br />

The X-ray diffraction patterns for the three samples in oxide phase are shown in Figure 1. Sample MECH shows<br />

an appearance of MoO3 peaks at 2θ values of 14.4, 23.6, and 26.8º. Sample IMPR also shows slight appearance of<br />

MoO3 peak at 2θ values of 23.6º, while there is no appearance of these peaks on the sample prepared by OMXC<br />

method. These appearances of the MoO3 peaks suggest that both methods, mechanical and impregnation, lead to<br />

lower dispersion of the active component compared to OMXC method, which enables a high dispersion of<br />

Ni3Mo7Ox.<br />

Raman spectroscopy is a powerful tool to study the structure of catalytically active phases on a support. Typical<br />

support such as alumina shows weak Raman scatters, with consequence that adsorbed species can be measured at<br />

frequencies as low as 50 cm –1 [10]. Figures 2, 4, and 8 show the Raman spectra of the oxides and sulfided phase for<br />

the catalysts in the 200–2000 cm –1 range. The spectra of Ni3Mo7Ox/Al2O3 catalysts do not exhibit the features of<br />

Ni3Mo7Sx/Al2O3 catalysts. The Raman spectra of NiMoOx/Al2O3 bimetallic supported oxide samples prepared using<br />

the three methods are shown in Figure 2. It shows that the main phase on the surface of samples IMPR and MECH<br />

is MoO3. The Raman band on these samples at 995 cm –1 corresponds to the stretching mode of Mo=O, and the band<br />

at 830 cm –1 corresponds to the anti-symmetric stretching mode of Mo–O–Mo, and the band at 295 cm –1 corresponds<br />

to the bending mode of Mo=O in the polycrystalline MoO3. The strong band at 965 cm –1 on sample MECH is due to<br />

the A1 mode of Mo=O in the MoO4 tetrahedral units. The broad band for samples IMPR and OMXC at 955 cm –1 ,<br />

which can be unfolded into two peaks at 955 and 965 cm –1 , highlights the band characteristic of NiMoO4. The<br />

Raman results suggest that the main oxide present clearly on the surface of sample MECH and IMPR is MoO3. Our<br />

findings agree with those reported by Sergio et al. [46, 47]. In addition, the shoulder of 950 cm –1 and the weak<br />

Raman bands at about 825 cm –1 are associated with the Mo–O–Mo for the polymolybdate species [48–51]. On the<br />

other hand, the weak intensity band at 950 cm –1 is characteristic of Ni–Mo oxide.<br />

Ramman Pand Intincety<br />

4500<br />

4000<br />

3500<br />

3000<br />

2500<br />

2000<br />

1500<br />

1000<br />

500<br />

MECH<br />

OMXC<br />

I MPR<br />

Al<br />

MECH<br />

I MPR<br />

0<br />

0 500 1000 1500 2000<br />

Ramman Shift, cm -1<br />

Figure 2. Laser Raman spectra for Ni 3Mo 7O x/Al 2O 3 catalysts prepared by Impregnation, Mechanical and OMXC methods


60<br />

Hamid A. Al-Megren<br />

Figure 3 shows the X-ray diffraction patterns for the three samples after the sulfidation reaction. It is clear that<br />

the main peaks for the oxide phase, i.e. MoO3 at 2θ values 14.4, 23.6, and 26.8, have disappeared from IMPR and<br />

OMXC samples while they have weakened on the MECH sample. This disappearance was attributed to the higher<br />

formation of sulfide phase on IMPR and OMXC, while some oxides still remain on MECH sample. Our results<br />

agreed with those reported by some investigators [48]. They have shown that during the primary period of<br />

thiophenes hydrodesulfurization on MoO3, a definite amount of sulfur was consumed by sulfidation of the oxide.<br />

This means that oxidic molybdenum catalyst can be sulfided by thiophenes, or by the hydrogen sulfide produced by<br />

DHO of thiophenes.<br />

Count/unit<br />

400<br />

300<br />

200<br />

100<br />

0<br />

10 20 30 40 50 60 70<br />

January 2009 The Arabian Journal for Science and Engineering, Volume 34, Number 1A<br />

2 q<br />

MECH<br />

OMXC<br />

I MPR<br />

Figure 3. XRD pattern for Ni 3Mo 7S x/Al 2O 3 catalysts prepared by Impregnation, Mechanical and OMXC methods<br />

The laser Raman spectra of the three samples (IMPR, MECH, and OMXC) after converting them from oxide to<br />

sulfide phase by sulfurization reaction are shown in Figure 4. The spectrum intensity of oxide phase (Figure 2) was<br />

highly reduced after sulfurization reaction. The vibration frequency at band 965 cm –1 Mo=O in the MoO4 tetrahedral<br />

units is strongly reduced after sulfurization. This band, i.e. 965 cm –1 disappeared completely on OMXC sample but it<br />

was partially reduced on rest of the samples. This result suggests that the molecular symmetry mode in the surface of<br />

MoO3 was changed after sulfurization [46–47]. In addition the spectrum of Figure 4 shows two bands at about 440<br />

and 450 cm –1 which could be attributed to MoS2 structure [13]<br />

1000<br />

Raman band intensity / Arb. units<br />

800<br />

600<br />

400<br />

200<br />

0<br />

MECH<br />

OMXC<br />

IMPR<br />

500 1000 1500 2000<br />

Raman shift, cm -1<br />

Figure 4. Laser Raman spectra for Ni 3Mo 7S x/Al 2O 3 catalysts prepared by Impregnation, Mechanical and OMXC methods


Hamid A. Al-Megren<br />

In the present study, the catalytic activity of sulfided NiMo/Al2O3 catalysts (IMPR, OMXC, MECH) were<br />

examined in the hydrodesulfurization of thiophenes at a temperature of 350°C and under atmospheric pressure. The<br />

conversion of thiophenes is shown in Figure 5. It can be observed that thiophenes conversion is constant over about<br />

8 h time on stream. Thus, high catalytic activity (57% of thiophenes conversion) is reached for the catalyst OMXC.<br />

Catalytic activities of IMPR and MECH catalysts are lower (45% and 21% of thiophene conversion respectively).<br />

The rate of reactions also show in order 73.80, 58.16, and 26.05 µmol (g min) –1 for OMXC, IMPR, and MECH<br />

respectively. The superior activity of OMXC catalyst is attributed to the incorporation of Ni–Mo onto the support<br />

surface. This provides a stronger metal–support interaction allowing a better dispersion of nickel and molybdenum<br />

oxidic and sulfided phases. On the other hand, too strong metal–support interaction makes the reduction and<br />

sulfidation of Ni and Mo species more difficult.<br />

Conversion, %<br />

60<br />

50<br />

40<br />

30<br />

20<br />

10<br />

OMXC<br />

0<br />

0 1 2 3 4 5 6 7 8<br />

Time, hour<br />

The Arabian Journal for Science and Engineering, Volume 34, Number 1A January 2009 61<br />

IMPR<br />

MECH<br />

Figure 5. HDS of thiophene over Ni 3Mo 7S x/Al 2O 3 catalysts at 350ºC, atmospheric pressure and 20ml/min flow of hydrogen<br />

This is in agreement with the results of Klimova et al. [53] who found NiMo catalysts supported on Al-SBA-16,<br />

prepared by grafting with AlCl3, show high activity in HDS of 4,6-DMDBT which attributed a good dispersion of Ni<br />

and Mo active phases.<br />

The catalytic activity for these catalysts was tested on hydrodesulfurization of thiophene at 350 o C and<br />

atmospheric pressure. Figure 5 shows that OMXC catalyst has superior activity (57%) compared to IMPR (45%) and<br />

MECH (21%) catalysts after seven hour of the reaction time. The rates of reaction also show in order 73.80, 58.16,<br />

and 26.05 umol (g min) –1 for OMXC, IMPR, and MECH respectively. This superior activity for OMXC sample<br />

attributed to the well dispersion of Ni–Mo over alumina as well as the higher sulfide formation compare to IMPR<br />

and MECH [46].<br />

= Ratio<br />

C 4 /C 4<br />

2.6<br />

2.5<br />

2.4<br />

2.3<br />

2.2<br />

2.1<br />

IMPR<br />

MECH<br />

OMXC<br />

2.0<br />

0 1 2 3 4 5 6 7 8<br />

Time, hour<br />

Figure 6. C4/C4= mol ratio for HDS of Thiophene over Ni3Mo7Sx/Al2O3 catalysts at 350ºC, atmospheric pressure and 20ml/min<br />

flow of hydrogen.


62<br />

Hamid A. Al-Megren<br />

Figure 6 shows the ratios of butane and butenes (C4/C4 = ), which are produced from HDS of thiophene reaction.<br />

The OMXC catalyst yielded the highest content of olefins, as compared to the IMPR catalyst. Products distribution<br />

data provide some insight into the possible pathway of thiophene desulfurization, which is difficult to extract from<br />

similar catalysts in which products are substantially altered by high rates of hydrogenation and isomerization<br />

reactions.<br />

More important are the differences between the three catalysts (IMPR, MECH, and OMXC) when we consider<br />

the product distributions. GC analysis showed the formation of mainly C4 and C4 = fractions, but it has not showed<br />

the formation of tetrahydrothiophene and butadiene. However many authors have argued for the preferential direct<br />

desulfurization route, by the absence of tetrahydrothiophene [50].<br />

Therefore our result shows that the pathway of reaction is the direct hydrodesulfurization route with successive<br />

hydrogenation, as shown in Scheme 1.<br />

S<br />

HDS<br />

HD<br />

S<br />

HD<br />

HDS<br />

HD<br />

HDS<br />

Scheme 1<br />

With the three catalysts, the ratio of C4 to C4 = is around 2.5.<br />

This suggests butane generation by hydrogenation of C4 compounds. These compounds are produced through the<br />

hydrogenolysis of thiophene and probably from β-H elimination from intermediate hydrogenated thiophenes [55].<br />

Table 2 shows the chemical analysis for the three samples analyzed before and after HDS reaction for more than<br />

seven hours. The sulfur content for the three samples increased after the HDS reaction. This sulfur content increase<br />

could be attributed to formation of sulfur species during the HDS reaction by the oxide phases that remain in the<br />

sample. The sample prepared by OMXC method shows more chemical stability compared to the samples prepared<br />

by other methods. The reaction rate was calculated for the HDS reaction over the three different catalysts. The<br />

catalyst prepared using OMXC method a shows higher reaction rate than IMPR and MECH.<br />

Table 2. Chemical Analysis for NiMoS Catalyst before and after HDS of Thiophene Reaction at 350ºC and<br />

Atmospheric Pressure<br />

Methods Before Reaction % After Reaction % Reaction rate<br />

Al Mo Ni S Al Mo Ni S μmol (g min) -1<br />

IMPR 47.28 24.49 7.41 20.81 41.61 24.36 5.61 28.42 58.16<br />

OMXC 46.63 18.60 9.68 25.10 42.46 19.31 9.50 28.74 73.80<br />

MECH 45.34 18.64 11.35 24.67 36.52 19.74 12.45 31.29 26.05<br />

The structure stability after the HDS reaction for the three samples was analyzed by X-ray and Raman<br />

spectroscopy. The X-ray patterns (Figure 7) for all the samples show higher structure stability after the HDS<br />

reaction. The remaining peaks of MoO3 over the MECH sample after sulfurization reaction were converted to the<br />

sulfide phase.<br />

Raman spectra for the three samples after HDS reaction are shown in Figure 8. It is shown that all bands for<br />

Mo=O and Mo–O–Mo vibrations at 995 cm –1 and 830 cm –1 on the IMPR sample are eliminated during the HDS<br />

reaction. There is also a reduction in band 965 cm –1 for the MECH sample.<br />

January 2009 The Arabian Journal for Science and Engineering, Volume 34, Number 1A<br />

HD


Count/arb. units<br />

500<br />

400<br />

300<br />

200<br />

100<br />

0<br />

0 10 20 30 40 50 60 70<br />

Hamid A. Al-Megren<br />

Figure 7. XRD pattern for Ni 3Mo 7S x/Al 2O 3 catalysts used for HDS reaction at 350ºC, atmospheric pressure and 20ml/min flow of<br />

hydrogen.<br />

Ramman Band Intincity<br />

1000<br />

800<br />

600<br />

400<br />

200<br />

0<br />

2 q<br />

The Arabian Journal for Science and Engineering, Volume 34, Number 1A January 2009 63<br />

MECH<br />

OMXC<br />

I MPR<br />

MECH<br />

OMXC<br />

I MPR<br />

500 1000 1500 2000<br />

Ramman Shift. cm -1<br />

Figure 8. Laser Raman spectra for Ni 3Mo 7S x/Al 2O 3 catalysts used for HDS at 350ºC, atmospheric pressure and 20ml/min flow of<br />

hydrogen.<br />

4. CONCLUSION<br />

The Ni–Mo–Al2O3 catalysts were prepared by impregnation (IMPR), Mechanical (MECH), and organic matrix<br />

combustion (OMXC) methods. The differences in the surface area of the catalysts are relatively small and range<br />

between 119.3, 122.5, and 125.5 m 2 /g for MECH, OMXC and IMPR respectively. XRD examination of IMPR,<br />

MECH, and OMXC catalysts indicated high dispersion of Ni3Mo7Ox in the OMXC catalyst as compared to the<br />

MECH and IMPR catalysts. XRD also indicated the disappearance of MoO3 after sulfidization of the catalysts. This<br />

could be attributed to higher formation of the sulfide phase. The disappearance of MoO3 was also observed in the<br />

Raman spectra.<br />

It was found that the OMXC catalyst has superior activity towards HDS of thiophenes, as compared to IMPR and<br />

MECH catalysts.


64<br />

Hamid A. Al-Megren<br />

In summary, a new method has been developed to prepare HDS catalyst and this catalyst has been compared with<br />

catalysts prepared by the conventional methods for selective thiophene HDS. This method enables a high dispersion<br />

of the molybdenum oxide and sulfide, resulting in higher activity and high yield of olefin in thiophene HDS.<br />

ACKNOWLEDGMENT<br />

The author would like thank Eng. Saeed Alshihri, Eng. Saud Aldrees, and Mr. Rasheed Al-Rasheed for their help<br />

in experimental work in the laboratories of King Abdulaziz City for Science and Technology. The author would<br />

also like to thank the Inorganic Chemistry Laboratory (ICL) at the University of Oxford, England, for their<br />

assistance in this research.<br />

REFERENCES<br />

[1] C. Song, X. Ma, “New Design Approaches to Ultra-Clean Diesel Fuels by Deep Desulfurization and Deep<br />

Dearomatization”, Appl. Catal. B: Environ.,41(2003), p. 207.<br />

[2] V. Babich and J.A. Moulijn, “Science and Technology of Novel Processes for Deep Desulfurization of Oil Refinery<br />

Streams: a Review”, Fuel, 82(2003), p. 607.<br />

[3] H. Topsøe, B.S. Clausen, and F.E. Massoth, Catalysis, Science and Technology, vol. 11. Berlin: Springer-Verlag,<br />

1996.<br />

[4] R. Prins, H.J. de Beer, and G.A. Somorjai, “Structure and Function of the Catalyst and the Promoter in Co–Mo<br />

Hydrodesulfurization Catalysts”, Catal. Rev.-Sci. Eng., 31(1989), p. 1.<br />

[5] P.T. Vasudevan and J.L.G. Fierro, “A Review of Deep Hydrodesulfurization Catalysis”, Catal. Rev.-Sci. Eng.,<br />

38(1996), p. 161.<br />

[6] R. Prins, “Catalytic Hydrodenitrogenation”, Adv. Catal., 46(2001), p. 399.<br />

[7] M. Zdrazil, “Recent Advances in Catalysis over Sulphides”, Catal. Today, 3(1988), p. 269.<br />

[8] R. Prins, V.H.J. de Beer, and G.A. Somorjai, “Structure and Function of the Catalyst and the Promoter in Co–Mo<br />

Hydrodesulfurization Catalysts”, Catal. Rev.-Sci. Eng., 31(1989), p. 1.<br />

[9] J.T. Kloprogge, W.J.J. Welters, et al. “Catalytic Activity of Nickel Sulfide Catalysts Supported on Al-Pillared<br />

Montmorillonite for Thiophene Hydrodesulfurization”, Appl. Catal. A Gen., 97(1993), p. 77.<br />

[10] E. Hayashi. et al. “High Surface Area Smectite Supported Cobalt Oxides as Active Catalysts for Thiophene<br />

Hydrodesulfurization”, Chemistry Letters, 5(1997), pp. 433–434.<br />

[11] E. Hayashi. et al., “Characterization of High Surface Area Smectite Supported Cobalt Oxides Catalysts for<br />

Hydrodesulfurization by Means of TPR, TPS, and ESR”, Appl. Catal. A Gen., 179(1999), p. 203.<br />

[12] J.C. Duchet. et al. “Carbon-Supported Sulfide Catalysts”, J. of Catal., 80(1983), p. 386.<br />

[13] S. L. Gonzalez-Cortes. et al., “Urea–Organic Matrix Method: an Alternative Approach to Prepare Co–MoS2–Al2O3 HDS Catalyst”, Appl. Catal., A: General, 270(2004), p. 209.<br />

[14] S. Damyanova, A. Spojakina, and K. Jiratova, “Effect of Mixed Titania–Alumina Supports on the Phase<br />

Composition of NiMo/TiOz–Al2O3 catalysts”, Appl. Catal. A, 125(1995), p. 257.<br />

[15] H. Qabazard, F. Abuseedo, A. Stanislaus, M. Andari, and M. Absihalabi, Fuel Sci. Technol. Int., 13(1995), p. 1135.<br />

[16] C. Papadopoulou, H. Matralis, A. Lycourghiotis, P. Grange, and B. Delmon, “Florinated Hydrotreatment Catalysts.<br />

Acidity and Carbon Deposition on Fluorine-Nickel-Molybdenum/γ-Alumina Catalysts”, J. Chem. Soc.– Faraday<br />

Trans., 89(1993), p. 3157.<br />

[17] V. Zuzaniuk and R. Prins, “Synthesis and Characterization of Silica-Supported Transition-Metal Phosphides as<br />

HDN Catalysts”, J. Catal., 219(2003), p. 85.<br />

[18] C. Kwak, J.J. Lee, J.S. Bae, K. Choi, and S.H. Moon, “Hydrodesulfurization of DBT, 4-MDBT, and 4,6-DMDBT<br />

on Fluorinated CoMoS/Al2O3 Catalysts”, Appl. Catal. A, 200(2000), p. 233.<br />

[19] H. Kim, J.J. Lee, and S.H. Moon, “Hydrodesulfurization of Dibenzothiophene Compounds using Fluorinated<br />

NiMo/Al2O3 Catalysts”, Appl. Catal. B, 44(2003), p. 287.<br />

[20] W.P. Zhang, M.Y. Sun, and R. Prins, “A High-Resolution MAS NMR Study of the Structure of Fluorinated NiW/γ-<br />

Al 2O 3 Hydrotreating Catalysts”, J. Phys. Chem. B, 107(2003), p. 10977.<br />

January 2009 The Arabian Journal for Science and Engineering, Volume 34, Number 1A


Hamid A. Al-Megren<br />

[21] V. Schwartz, M.Y. Sun, and R. Prins, “An EXAFS Study of the Influence of Fluorine on the Structure of Sulfided<br />

W/Al2O3 and NiW/Al2O3 Catalysts”, J. Phys. Chem. B, 106(2002), p. 2597.<br />

[22] M. Nagai, Y. Goto, H. Ishii, and S. Omi, “XPS and TPSR Study of Nitrided Molybdena–Alumina Catalyst for the<br />

Hydrodesulfurization of Dibenzothiophene”, Appl. Catal. A, 192(2000), p. 189.<br />

[23] P. Da Costa, C. Potvin, J.M. Manoli, J.L. Lemberton, G. Perot, and G. Djega Mariadassou, “New Catalysts for<br />

Deep Hydrotreatment of Diesel Fuel Kinetics of 4,6–Dimethyldibenzothiophene Hydrodesulfurization over<br />

Alumina Supported Molybdenum Carbide”, J. Mol. Catal. A, 184(2002), p. 323.<br />

[24] P. Da Costa, C. Potvin, J.M. Manoli, M. Breysse, and G. Djega-Mariadassou, Catal. Lett., 86(2003), p. 133.<br />

[25] H.R. Reinhoudt, M. van Gorsel, A.D. van Langeveld, J.A.R. van Veen, S.T. Sie, and J.A. Moulijn, Hydrotreatment<br />

and Hydrocracking of Oil Fractions, 127(1999), p. 211.<br />

[26] H. Topsøe, R. Candia, N.Y. Topsøe, and B.S. Clausen, Bull. Soc. Chim. Belg., 93(1984) p. 83.<br />

[27] H. Topsøe, B.S. Clausen, N.Y. Topsøe, and E. Pedersen, Ind. Eng. Chem. Fundam., 25(1986) p. 25.<br />

[28] H. Farag, D.D. Whitehurst, K. Sakanishi, and I. Mochida, “Carbon versus Alumina as a Support for Co–Mo<br />

Catalysts Reactivity Towards HDS of Dibenzothiophenes and Diesel fuel”, Catal. Today, 50(1999), p. 9.<br />

[29] A.A. Spozhakina et. al., “Synthesis of Molybdenum/silica and Cobalt–Molybdenum/Silica Catalysts and their<br />

Catalytic Activity in the Hydrodesulfurization of Thiophenes”, Appl. Catal., 39(1988), p. 333.<br />

[30] R. Cattaneo, T. Shido, R. Prins in: Hydrotreatment and Hydrocracking of Oil Fractions, Amsterdam: Elsevier,<br />

1999, p. 421.<br />

[31] M. Breysse, E. Furimsky, S. Kasztelan, M. Lacroix, and G. Perot, “Hydrogen Activation by Transition Metal<br />

Sulfides”, Catal. Rev.-Sci. Eng., 44(2002), p.651.<br />

[32] F. Bataille et al., “Sulfided Mo and CoMo Supported on Zeolite as Hydrodesulfurization Catalysts: Transformation<br />

of Dibenzothiophene and 4,6– Dimethyldibenzothiophene”, Appl. Catal. A, 220(2001), p. 191.<br />

[33] P.W. de Bont et al., “Molybdenum-Sulfide Particles Inside NaY Zeolite as a Hydrotreating Catalyst Prepared by<br />

Adsorption of Mo(CO) 6”, Appl. Catal. A, 202(2000), p. 99.<br />

[34] G. Plazenet, E. Payen, and J. Lynch, “Cobalt-Molybdenum Interaction in Oxidic Precursors of Zeolite-Supported<br />

HDS Catalysts”, Phys. Chem. Chem. Phys., 4(2002), p. 3924.<br />

[35] J. Vissenberg et al., “The Effect of Support Interaction on the Sulfidability of Al2O3- and TiO2-Supported CoW and<br />

NiW Hydrodesulfurization Catalysts”, J. Catal., 198(2001), p. 151.<br />

[36] S.K. Maity, M.S. Rana, S.K. Bej, J. Ancheyta-Juarez, G.M. Dhar, and T. Rao, Catal. Lett., 72(2001), p. 115.<br />

[37] P. Afanasiev, M. Cattenot, C. Geantet, N. Matsubayashi, K. Sato, and S. Shimada, “(Ni)W/ZrO2 Hydrotreating<br />

Catalysts Prepared in Molten Salts”, Appl. Catal. A, 237(2002), p. 227.<br />

[38] L. Qu, W. Zhang, P.J. Kooyman, and R. Prins, “MAS NMR, TPR, and TEM Studies of the Interaction of NiMo<br />

with Alumina and Silica–Alumina Supports”, J. Catal., 215(2003), p. 7.<br />

[39] L.D. Sharma, M. Kumar, J.K. Gupta, M.S. Rana, V.S. Dangwal, and G.M. Dhar, Indian J. Chem. Technol.,<br />

8(2001) p. 169.<br />

[40] F. Schuth and K. Unger, “Precipitaion and Coprecipitation” in Handbook of Heterogeneous Catalysis, Vol. 1, ed.<br />

G. Ertl, H. Knozinger, and J. Weitkamp. Wiley–VCH, 1997, p. 72.<br />

[41] M. Che, O. Clause, and Ch. Marcilly, “Impregnation and Ion Exchange”, in Handbook of Heterogeneous Catalysis,<br />

Vol. 1, ed. G. Ertl, H. Knozinger, and J. Weitkamp. City: Wıley–VCH, 1997, p. 191.<br />

[42] B. H. Davis, in Handbook of Heterogeneous Catalysis, Vol. 1, ed. By G. Ertl, H. Knozinger, and J. Weitkamp.<br />

City: Wıley–VCH, 1997, p. 13.<br />

[43] E. I. Ko, “Sol-Gel Process”, in Handbook of Heterogeneous Catalysis, Vol. 1, ed. G. Ertl, H. Knozinger, and J.<br />

Weitkamp. Wiley–VCH, 1997, p. 86.<br />

[44] S.L. Gonzalez-Cortes et. al., “Impact of the Urea–Matrix Combustion Method on the HDS Performance of Ni–<br />

MoS2/γ-Al2O3 Catalysts”, J. Mol. Catal. A: Chem, 240(2005), p. 214.<br />

[45] S. L. Gonzalez-Cortes, T-C Xiao, and M.L.H. Green, Proceedings of the 9 th International Symposium, Scientific<br />

Bases for the Preparation of Heterogeneous Catalysts, Louvain-la-Neuve, Belgium, September 10-14, 2006, p. 817.<br />

The Arabian Journal for Science and Engineering, Volume 34, Number 1A January 2009 65


66<br />

Hamid A. Al-Megren<br />

[46] Sergio L. González-Cortés et al., “Influence of Double Promotion on HDS Catalysts Prepared by Urea-Matrix<br />

Combustion Synthesis”, Applied Catalysis A: Gen., 302(2006), p. 264–273.<br />

[47] Sergio L. González-Cortés et. al., “Impact of the Urea–Matrix Combustion Method on the HDS Performance of Ni–<br />

MoS2–Al2O3 Catalysts”, J. of Molecular Catalysis A: Chemical, 240(2005), p. 214–225.<br />

[48] G. Mestl and T.K.K. Srinivasan,“Raman Spectroscopy of Monolayer–Type Catalysts: Supported Molybdenum<br />

Oxides,” Catal. Rev., Sci. Eng., 40(1998), p. 451.<br />

[49] D.S. Kim, K. Segawa, T. Soeya, and I.E. Wachs, “Surface Structures of Supported Molybdenum Oxide Catalysts<br />

Under Ambient Conditions”, J. Catal., 136(1992), p. 539.<br />

[50] H. Hu and I.E. Wachs, “Catalytic Properties of Supported Molybdenum Oxide Catalysts: In situ Raman and<br />

Methanol Oxidation Studies”, J. Phys. Chem., 99(1995), p. 10911.<br />

[51] G. Mestl, “In Situ Raman Spectroscopy–a Valuable Tool to Understand Operating Catalysts”, J. Mol. Catal. A:<br />

Chem., 158(2000), p. 45.<br />

[52] G.L. Schrader and C.E Chen, “In Situ Laser Raman Spectroscopy of the Sulfiding of MO/γ-A1203 Catalysts”, J.<br />

Catal., 80(1983), p. 369.<br />

[53] T. Klimova, L. Lizama, et al., “New NiMo Catalysts Supported on Al Containing SBA-16 for 4,6-DMDBT<br />

Hydrodesulfurization: Effect of the Alumination Method”, Catalysis Today, 98(1-2), p. 141–150.<br />

[54] H. Topose, B.S. Clausen, and F.E. Mssoth, Hydrotreating Catalysis. Berlin: Springer, 1996, p. 116.<br />

[55] R. Prins, in Handbook of Heterogeneous Catalysis, ed. G. Ertl, H. Knozinger, P.J. Weitkam, vol. 4. Wiley-VCH,<br />

1997, pp. 1908.<br />

January 2009 The Arabian Journal for Science and Engineering, Volume 34, Number 1A

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!