Energy Conversion and Management 52 (2011) 1364–1370

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier .com/locate /enconman

Dibenzothiophene hydrodesulfurization over Ru promoted alumina basedcatalysts using in situ generated hydrogen

Yaseen Muhammad a,b, Yingzhou Lu b, Chong Shen a,b, Chunxi Li a,b,⇑a State Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, Beijing 100029, PR Chinab College of Chemical Engineering, Beijing University of Chemical Technology, Beijing 100029, PR China

a r t i c l e i n f o

Article history:Received 26 June 2010Accepted 22 September 2010Available online 27 October 2010

Keywords:DibenzothiopheneHydrodesulfurizationAlumina based catalystIn situ hydrogenOrganic additive

0196-8904/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.enconman.2010.09.034

⇑ Corresponding author: Beijing University of ChemChina. Tel./fax: +86 1064410308.

E-mail address: Licx@mail.buct.edu.cn (C.X. Li).

a b s t r a c t

Catalytic hydrodesulfurization (HDS) of dibenzothiophene (DBT) was carried out in a temperature rangeof 320–400 �C using in situ generated hydrogen coupled with the effect of selected organic additives forthe first time. Four kinds of alumina based catalysts i.e. Co–Mo/Al2O3, Ni–Mo/Al2O3, Ru–Co–Mo/Al2O3 andRu–Ni–Mo/Al2O3 were used for the desulfurization process, which were prepared following incipientimpregnation method with fixed metal loadings (wt.%) of Co, Ni, Mo and Ru. The surface area, averagepore diameter and pore volume distribution of the fresh and used catalysts were measured by N2 adsorp-tion using BET method. Catalytic activity was investigated in a batch autoclave reactor in the completeabsence of external hydrogen gas. Addition and mutual reaction of specific quantities of water and eth-anol provided the necessary in situ hydrogen for the desulfurization reaction. Organic additives likediethylene glycol (DEG), phenol, naphthalene, anthracene, o-xylene, tetralin, decalin and pyridine didimpinge the HDS activity of the catalysts in different ways. Liquid samples from reaction products werequantitatively analyzed by HPLC technique while qualitative analyses were made using GC–MS. Both ofthese techniques showed that Ni-based catalysts were more active than Co-based ones at all conditions.Moreover, incorporation of Ru to both Co and Ni-based catalysts greatly promoted desulfurization activ-ity of these catalysts. DBT conversion of up to 84% was achieved with Ru–Ni–Mo/Al2O3 catalyst at 380 �Ctemperature for 11 h. Catalyst systems followed the HDS activity order as: Ru–Ni–Mo/Al2O3 > Ni–Mo/Al2O3 > Ru–Co–Mo/Al2O3 > Co–Mo/Al2O3 at all conditions. Cost effectiveness, mild operating conditionsand reasonably high catalytic activity using in situ generated hydrogen mechanism proved our processto be useful for HDS of DBT.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Sulfur (noted as S-hereinafter) components present in fuel oilscan lead to a SOx pollution to the environment [1]. In order to min-imize such pollution from the exhaust of fuel oils, increasinglystringent regulations are being imposed to reduce the S-contentto a very low level, such as 10–20 ppm [1,2]. At present, the con-ventional technology for the removal of such S-components fromfuel oils is the catalytic HDS process in which the active compo-nents of the catalysts mostly include Ni or Co with Mo supportedon Al2O3 [2–5]. Besides, addition of some noble metals, e.g. Ru,Rh, Ru–Rh, Pt, Pd and Pt–Pd, has also been reported to improvethe activity of the catalysts [6]. Apart from this, ruthenium (Ru)has been used widely as a promoter to the Co–Mo/Al2O3 andNi–Mo/Al2O3 catalyst system [7,8].

ll rights reserved.

ical Technology, Beijing, PR

The conventional HDS process is often carried out under hightemperature and pressure using external hydrogen gas as desulfu-rizing reagent [9–11]. In this process, the high temperature andpressure conditions are closely associated with the low reactivityof S-compounds with hydrogen molecules. Suppose the hydrogenatoms required for the HDS process can be provided in situ bythe reforming reaction of alcohol, relatively mild reaction temper-ature and pressure can be expected. In effect, the in situ hydroge-nation has been studied extensively, however little work has beenreported for the HDS of fuels using in situ hydrogen mechanism.Liu and Flora [12] for the first time reported the HDS of DBT usingin situ generated hydrogen by water gas shift reaction (WGSR)along with external gaseous hydrogen in the presence of dispersedMo catalysts. In the present work, we studied the in situ HDS pro-cess of DBT in the complete absence of external hydrogen gas usingfour different types of catalysts, i.e. Co–Mo/Al2O3, Ni–Mo/Al2O3,Ru–Co–Mo/Al2O3 and Ru–Ni–Mo/Al2O3. Moreover the effect ofsome selected organic additives on activity of the catalystswas also analyzed. The reaction products were qualitativelyanalyzed using GC–MS technique. Fresh and used catalysts were

Y. Muhammad et al. / Energy Conversion and Management 52 (2011) 1364–1370 1365

characterized in terms of BET adsorption properties. The presentstudy is instructive for developing a novel and efficient catalyticHDS process of fuel oils without using external hydrogen gas.

2. Experimental

2.1. Chemical reagents used

Chemical reagents used in this study were of analytical grade, andwere used without further purification. DBT was purchased fromACROS Organics, New Jersey, USA. Al2O3, phenol, naphthalene,tetrahydronaphthalene (THN) i.e. tetralin and anthracene werepurchased from Tian-jin Guang-fu fine Chemical Institute, China.Octane was purchased from Tian-jin Kermel Chemical ReagentCompany, China. Ethanol, o-xylene, Co (NO3)2�6H2O, pyridine, DEGand n-heptane were purchased from Tian-Jin Shi-Fu-chenChemical Reagent Factory, China. Decahydronaphthalene (DHN)i.e. decalin was purchased from Sinopharm Chemical Reagent Co.Ltd., China. Ammonium heptamolybdate (NH4)6Mo7O24�4H2O andPb(CH3COO)2 were purchased from Beijing Chemical Works, China.

2.2. Catalyst preparation

Alumina with high surface area (279.05 m2/g) was used as cat-alytic support in this study. Catalysts were prepared by successiveincipient wetness impregnation method. Calculated amount of alu-mina support was first impregnated with aqueous solution ofammonium heptamolybdate (NH4)6Mo7O24�4H2O, correspondingto 8 wt.% of Mo, and then with aqueous solution of Co(NO3)2�6H2Oor Ni(NO3)2�6H2O, corresponding to 1 wt.% of Co or 4 wt.% of Ni,respectively. After each impregnation, the support with metal saltsolution was stirred for 20 h at room temperature and then dried inan oven at 120 �C for 15 h. The dried catalysts were calcined inmuffle furnace at a temperature of 500 �C for 5 h to convert metalsinto their respective oxides. After calcinations, all the catalystswere presulfided in a tubular reactor in an environment of 5 wt.%dimethyl sulfide i.e. (CH3)2S (DMS) in n-Heptane (as solvent) at500 �C for 5 h using N2 as career gas. On completion of the presulf-idation step, catalysts were kept under inert environment of N2 soas to avoid any effect of air on its composition. Similar processwas used for the preparation of Ru promoted Co–Mo/Al2O3 orNi–Mo/Al2O3 catalyst with 1 wt.% Ru loadings in all the samples.

2.3. Support and catalysts characterization

The Al2O3 support and all catalysts (fresh and used) were char-acterized by virtue of specific surface area (SBET) and pore size dis-tribution using N2-physisorption method with Micromeritcs-ASAP2020, USA, surface area and pore size analyzer. For the surface areaanalysis, fully dried and sealed samples were first degassed at300 �C for 3 h, and then exposed to a temperature of 77.2 K. Porevolume analysis for all the samples was carried out at a tempera-ture of 77 K after being degassed at 150 �C for 3 h bringing downthe total operational pressure to less than 5 lmHg.

2.4. Catalytic activity measurement and product analysis

A typical experiment for determining catalytic activity of thecatalyst was carried out in a 200 ml magnetically stirred batchautoclave reactor. This reactor is connected with an automatic con-trolling unit for regulating temperature and rotation of the stirrerinside the reactor. The temperature of the reaction was determinedwith a jacketed thermocouple placed inside the reactor. In allexperiments, a 130 rpm stirrer rotation speed was used while tem-perature was varied from 320 to 400 �C. The model fuel used was

900 ppm DBT solution in n-octane for all the experiments. Typi-cally a single run used 100 ml of liquid contents consisting of90 ml model solution, stoichiometric amount of water–ethanolmixture, 10 mmol of a specific organic additive and 0.5 g catalyst.After sealing, the autoclave was allowed to heat up gradually fromroom temperature to reaction temperature, with a heating rate of6.4 �C/min. The reaction was then continued for different durationsof time ranging from 1 to 11 h. Reaction was carried out in thecomplete absence of external hydrogen gas supply, and the inter-nal pressure of the reactor during experiments changed automati-cally with changing temperature. Gaseous products from autoclavewere passed through aqueous Pb(CH3COO)2 solution, and H2S com-ponent was confirmed by the resulting grayish precipitate of PbS.Liquid samples were taken from the reactor after regular intervalsof time, sealed and kept in an inert environment below 0 �C. Thequantitative analyses for these samples were carried out usingHPLC technique while qualitative analyses were carried out usinga GS–MS chromatograph (Trace GC Ultra-DSQ II, Thermo-Scientific,USA), having HP-5MS steel column 30 m in length and 0.25 mm ininner diameter. A GC analysis was started with a starting temper-ature of 80 �C, kept on this temperature for 1 min, and then in-creased up to 250 �C at a rate of 10 �C/min with a sample volumeof 0.2 lL. Temperature was kept at 250 �C for 20 min till the endof the run.

3. Results and discussion

3.1. Textural properties of the support and catalysts

The textural properties of alumina support and catalysts (freshand used) were measured using N2-physisorption method at atemperature of 77.2 K. The data for the BET surface area, pore vol-ume, average pore diameter and nominal metal contents of thefresh samples are presented in Table 1. From Table 1, it is evidentthat, the incorporation of Co, Ni, Mo or Ru into the support greatlydecreased its surface area, i.e. from 279.05 m2/g to about 160 m2/g.This decrease in surface area was attributed to the deposition ofthese metals on surface of the support after impregnation. Thisdata is in good agreement with literature cited elsewhere for cata-lyst systems using alumina as support being impregnated with Mo,Co or Ni [13,14].

There is a great decrease in surface areas for the used catalystsas can be seen by comparing the values in Table 1 and 2. The de-crease in surface area can be attributed to the deposition of sulfurcompounds formed during HDS reaction. Decrease in the surfaceareas of the used catalyst also strongly supports the adsorptivenature (hydrogenation reaction) of HDS reaction over the catalystsurface. The absorptive nature and slowness of HDS is also wellsupported by a greater decrease in surface areas of the used cata-lysts with longer reaction time i.e. 11 h as compared to 5 h. Themaximum conversion at 11 h reaction time and 380 �C tempera-ture is well supported by the maximum decrease in catalyst’s sur-face area at these conditions.

3.2. Catalytic activity

Catalytic activity of the presulfided catalysts was evaluated in abatch autoclave reactor in a temperature range of 320–400 �C atdifferent reaction intervals in the complete absence of externalhydrogen gas. Operating conditions used during catalyst prepara-tion and HDS reaction are listed in Table 3.

3.2.1. Effect of reaction temperature on the activity of catalystThe effect of reaction temperature on the activity of presulfided

catalysts was studied in a temperature range of 320–400 �C. The

Table 1Textural properties of the support and sulfided catalysts (fresh).

Catalyst/Support BET surface area (m2/g) Pore volume (cm3) Average pore diameter (Å) Metal loading (wt.%) over Al2O3 Usage (g)

Co Mo Ni Ru

Al2O3 279.05 0.39 56.60 – – – – -Co–Mo/Al2O3 158.06 0.30 74.79 1 8 – – 0.5Ni–Mo/Al2O3 166.42 0.29 70.76 – 8 4 – 0.5Ru–Co–Mo/Al2O3 160.29 0.28 65.42 1 8 – 1 0.5Ru–Ni–Mo/Al2O3 146.08 0.27 72.84 – 8 4 1 0.5

Table 2Textural properties of the used catalysts.

Catalyst Additive Time (h) Temperature (�C) BET surface area (m2/g) Pore volume (cm3) Average pore diameter (Å)

Co–o/Al2O3 Decalin 5 380 100.3 0.26 105.05Ni–Mo/Al2O3 Decalin 5 380 87.4 0.28 129.03Ru–Co–Mo/Al2O3 Decalin 5 380 99.7 0.30 119.86Ru–Ni–Mo/Al2O3 Decalin 5 380 100.1 0.28 110.59Ru–Ni–Mo/Al2O3 Decalin 11 380 86.8 0.27 126.19

Table 3Catalyst preparation and HDS reaction conditions.

1. Catalyst preparation conditionsStirring time between each impregnation 20 hCatalyst drying temperature/time 120 �C/15 hCalcination temperature/time 500 �C/5 hSulfidation media 5 wt.% DMS in n-heptaneSulfidation temperature/time 400 �C/5 hCareer gas Pure N2 gas

2. HDS reaction conditionsReaction temperature 320–400 �CReaction time 1–11 hStirring speed 130 rpmHeating rate 6.4 �C/minDBT model solution’s concentration 900 ppm in n-OctaneCatalyst weight 0.5 g/runAdditive concentration 10 mmolWater:ethanol ratio (3:1) mol

0

15

30

45

60

75

90

300 320 340 360 380 400Temperature (oC)

DB

T c

onve

rsio

n (%

)

Fig. 1. Influence of reaction temperature on % DBT conversion over (�) Co–Mo/Al2O3, (N) Ru–Co–Mo/Al2O3, (j) Ni–Mo/Al2O3 and (d) Ru–Ni–Mo/Al2O3 catalysts at11 h reaction time.

1366 Y. Muhammad et al. / Energy Conversion and Management 52 (2011) 1364–1370

results obtained are illustrated in Fig. 1. Activity of the catalystswas calculated in terms of percentage DBT conversion.

From Fig. 1, it is evident that temperature has a positive effecton activity of the catalyst, which is consistent with the literatureresults for the similar catalyst systems [3], [15–17]. In those stud-

ies a temperature of 350–360 �C provided the maximum DBT con-version utilizing external H2 gas as hydrogenating media, while inour case the temperature reached up to 380 �C for maximum DBTconversion. This can be explained on the basis of water:ethanolreforming reaction occurring parallel to the HDS in our process.The reforming reaction has been widely studied [18–23] usingNi, Co and some noble metal like Ru, Rh and La based catalystson the basis of reaction A.

C2H5OHþ 3H2O! 6H2 þ 2CO2 ðReactionAÞ

Those studies showed that the reforming process is highly tem-perature dependant and is a fast reaction at high temperature. Inthose studies, reforming reaction was studied in a temperaturerange of 500–1000 �C. On the other hand, HDS is a slow processoccurring in a temperature range of 320–420 �C [2,5,6]. Being acombination of reforming and hydrogenation reaction, the in situHDS reaction must be kept in a temperature range that can bestsuit both the reactions, i.e. 350–400 �C. In Fig. 1 the maximumDBT conversion at 380 �C can be explained by the fact that at thistemperature the reforming and hydrogenation reactions are occur-ring at their best conditions. In contrast, an increase from this tem-perature will inhibit the HDS process although the reformingreaction is favored which is shown by the decline in DBT conver-sion at 400 �C.

From Fig. 1, it can also be seen that at all temperatures, catalyticactivity followed the order Ru–Ni–Mo/Al2O3 > Ni–Mo/Al2O3 > Ru–Co–Mo/Al2O3 > Co–Mo/Al2O3. The high activity of the Ni-based cat-alysts compared to the Co-based ones, is well supported by the factthat Ni is a better reforming metal than Co at such a low temper-ature [24,25] thus giving higher DBT conversion. Incorporation ofRu, on the other hand definitely increases the active sites as com-pared to the mere Ni or Co based catalyst, thus giving higher DBTconversion.

The DBT conversion of 73% by Ru–Ni–Mo/Al2O3 having a lowmetal loading such as 1 wt.% Ru utilizing in situ generated H2

gas, is comparable to the reported results for the similar catalystsystem, using higher metal loadings i.e.16 wt.% Ru coupled with0.25 wt.% Rh and ample amount of external H2 gas supply ashydrogenating agent [6].

3.3. Effect of reaction time on DBT conversion

The effect of reaction time on HDS is depicted in Fig. 2. Fig. 2shows that DBT conversion increases with reaction time and

0

15

30

45

60

75

90

0 2 4 6 8 10 12Time (h)

DB

T c

onve

rsio

n (%

)

Fig. 2. Influence of reaction time on % DBT conversion over catalysts (�) Co–Mo/Al2O3, (N) Ru–Co–Mo/Al2O3, (j) Ni–Mo/Al2O3 and (d) Ru–Ni–Mo/Al2O3 at 380 �Creaction temperature.

Y. Muhammad et al. / Energy Conversion and Management 52 (2011) 1364–1370 1367

reaches to the maximum value after 11 h for all the catalysts. Amaximum DBT conversion of 84% was recorded for Ru–Ni–Mo/Al2O3 catalyst at 11 h reaction time, which is comparable withthe results reported elsewhere utilizing external H2 gas as hydro-genating reagent [26].

Fig. 2 also shows that Ni-based catalysts are more active thanCo-based ones which is consistent to the reported literature [27].The mild increase in DBT conversion with time can be explainedby accumulative production of H2 and slow HDS process. The activ-ity of the catalysts followed the order of Ru–Ni–Mo/Al2O3 > Ni–Mo/Al2O3 > Ru–Co–Mo/Al2O3 > Co–Mo/Al2O3, showing that Ru is an ac-tive promoter to the conventional Co/Ni-based catalysts.

3.4. Effect of organic additives on activity of the catalyst

Experiments were performed in order to examine the influenceof selected organic additives on HDS activity of the catalysts. In to-tal, eight organic additives, namely anthracene, decalin, DEG,naphthalene, phenol, pyridine, tetralin and o-xylene, were tested.For comparison, a contrast experiment was also performed for eachcatalyst. The data for this set of experiments is presented in Fig. 3.

Depending on their chemical nature, each additive affected thecatalytic activity in a different way and up to different extent. From

0

10

20

30

40

50

60

70

80

Anthracene Blank Decalin DEG NaphthalineAdditive

DB

T c

onve

rsio

n (%

)

Fig. 3. Influence of organic additive on % DBT conversio

Fig. 3, it is obvious that among the additives used, only three i.e.anthracene, decalin and o-xylene promoted the catalytic activity,while the rest showed inhibitory effect. These results are consis-tent with the reported literature [28,29] where tetralin and pyri-dine were found to be detrimental while o-xylene and decalinare promoters of HDS. We used three saturated organic additivesi.e. tetralin, DEG and decalin, in which decalin increased the activ-ity to great extent, while Tetraline and DEG showed an opposite ef-fect. This inhibitory effect for tetraline and DEG can be explainedon the grounds that these saturated compounds adsorb competi-tively on active sites as compared to DBT [29], thus blocking the ac-tive sites and inhibiting both the reforming and HDS reaction.Another important factor to be considered here is the boilingpoint/volatility of these additives. Boiling point of DEG, tetralinand decalin is 245 �C, 206 �C and 187 �C respectively, clearly indi-cating that DEG and tetralin with higher boiling points will remainover the catalyst surface for a longer time, covering active sites andhence giving a lower DBT conversion. Decalin being completely invapor state allows a better contact of DBT with catalyst’s activesites, resulting in a higher DBT conversion. The same explanationis true for o-xylene (B.P 144.4 �C). It is assumed that aromaticstructures like anthracene, are quite stable at normal conditions,but if its aromaticity is disturbed by an atom like in situ producedhydrogen, they become unstable (intermediate state), and thusprefer to regain their aromaticity with speed and ease, hencereleasing a more active form of hydrogen, which in turn assiststhe in situ produced hydrogen in HDS process. Naphthalene whichhas been reported to be detrimental [30], also showed inhibitoryeffect. Decalin has been reported to be detrimental [28], but inour work it was found to be a highly promotive additive. Thiscan be explained by the fact that decalin and decalene are inter-changeable at high temperature giving rise to additional in situH2 generation and hence assisting the desulfurization reaction.This phenomenon is also confirmed by GC analysis (Figs. 4 and5). For all types of additives used, catalytic activity followed the or-der Ru–Ni–Mo/Al2O3 > Ni–Mo/Al2O3 > Ru–Co–Mo/Al2O3 > Co–Mo/Al2O3.

3.5. Product distribution and GC analysis

Liquid products were periodically collected from batch reactorand analyzed by a GC–MS chromatograph. GC results are depictedin Figs. 4 and 5 while the peak areas of the products are givenin Table 4. In Figs. 4 and 5, chromatogram ‘‘A” represents the

Phenol Pyridine Tetralin Xylene

Co-Mo/Al2O3

Ni-Mo/Al2O3

Ru-Co-Mo/Al2O3

Ru-Ni-Mo/Al2O3

n at a temperature of 380 �C and 5 h reaction time.

6 7 8 9 10 11 12 13 14 15 1605101520253035404550556065707580859095100 5.54

5.85

13.918.38 15.72

6.34 15.576.577.63

12.748.63 9.02 10.18 10.7012.65

A

Rel

ativ

e ab

unda

nce

Time (min)

S

Fig. 4. GC Chromatograph of products with Co–Mo/Al2O3 (conditions listed in Table 4. S. no. A).

5 6 7 8 9 10 11 12 13 14 15 1605101520253035404550556065707580859095100 5.79

5.49

8.31

6.50 7.56 8.55 15.8110.64 13.778.96 9.74 15.2213.7112.6711.33

Rel

ativ

e ab

unda

nce

H

Time (min)

Fig. 5. GC Chromatograph of products with Ru–Ni–Mo/Al2O3 (conditions listed in Table 4. S. no. H).

1368 Y. Muhammad et al. / Energy Conversion and Management 52 (2011) 1364–1370

minimum% DBT conversion using Co–Mo/Al2O3 catalyst, whilechromatogram ‘‘H” represents the maximum% DBT conversionusing Ru–Ni–MO/Al2O3 catalyst in our experimental data.

In Figs. 4 and 5 the peak at retention time 5.4–5.7 min was as-signed to tertralin while the peak at retention time of 5.8–6.0 wasassigned to naphthalene and decaline, indicating that these are thedehydrogenated products of decalin, which itself was assigned thepeak at 7.5–7.8 min. The peak at 8.2–8.5 min was assigned to BP,which was the major product in all the samples while the peak

at 13.0–14.6 min retention time was assigned to DBT. Presence ofBP peak in all the chromatograms indicates that the mechanismfor DBT conversion is same, and products differ only quantitatively.As BP was formed in major quantity, it was supposed that directdesulfurization (DDS) dominated the hydrogenation process,which is in agreement with the literature cited elsewhere [31–33]. In GC chromatograms (Figs. 4 and 5), cyclohexylbenzene(CHB) was completely absent, which has been reported as a prod-uct along with BP for HDS of DBT using external hydrogen [13]. The

Table 4Products and their relative abundance for different catalysts used at 380 �C and different reaction time.

S. no. Catalyst Temperature (�C) Time (h) Additive DBT peak area (%) BP peak area (%)

A Co–Mo/Al2O3 380 5 Decalin 87.39 12.61B Ru–Co–Mo/Al2O3 380 5 Decalin 77.79 22.21C Ni–Mo/Al2O3 380 5 Decalin 64.70 35.30D Ru–Ni–Mo/Al2O3 380 5 Decalin 32.20 67.80E Co–Mo/Al2O3 380 11 Decalin 66.89 33.11F Ru–Co–Mo/Al2O3 380 11 Decalin 59.46 40.54G Ni–Mo/Al2O3 380 11 Decalin 22.53 77.47H Ru–Ni–Mo/Al2O3 380 11 Decalin 22.59 77.41

Y. Muhammad et al. / Energy Conversion and Management 52 (2011) 1364–1370 1369

absence of CHB in product streams can be explained on twoassumptions. First, it is supposed, that the amount of in situ pro-duced hydrogen was not ample enough to carry out the hydroge-nation reaction of DBT or BP into CHB. Secondly, the catalystsurface may have lost all of its active sites due to DDS pathway,thus not facilitating the adsorption of hydrogen over its surface,and hence not helping DBT or BP to get hydrogenated into CHB.In these chromatograms, if peaks for the solvent and its derivatives(5.4–5.7 min, 5.8–6.0, 7.5–7.8 min) are considered constant/ne-glected, the peak area of BP and DBT directly determines the activ-ity of the catalysts. As can be seen in Table 4A–D, area of BP peakincreases, while that of DBT decreases gradually, showing thatconcentration of BP increases and DBT decreases. In Table 4A–D,area (concentration) of BP peak followed the order of Co–Mo/Al2O3 < Ru–Co–Mo/Al2O3 < Ni–Mo/Al2O3 < Ru–Ni–Mo/Al2O3. Thereverse order was observed for DBT peak area. The same patternwas found from the analysis of products for reaction time of 11 has shown in Table 4E–H. GC results also confirmed the obtainedexperimental data for catalytic activity.

4. Conclusions

A new approach towards catalytic HDS of DBT based on in situgenerated hydrogen by water:ethanol reforming reaction in atemperature range of 320–380 �C was found to be quite usefuland productive. Incorporation of even a very low metal loadingof Ru (1 wt.%) to the conventional Ni/Co–Mo alumina supportedcatalysts greatly promoted the catalytic activity towards DBTdesulfurization. On one hand, the increased catalytic activity wasrecorded by the addition of Ru, at the same time the use ofin situ produced hydrogen also attributed to the novelty of the pro-cess. Effect of some of the selected organic additives coupled withthe in situ hydrogen usage was noteworthy as they did enhancethe HDS activity of the catalysts, while some of them showedinhibitory effect. In situ H2 mechanism, low metal loadings andmild operating conditions are the plus points of our process, mak-ing it prospective in the field of HDS research. The process can beused as an alternative approach to HDS of DBT on industrial scale.Furthermore, the present work opens a new area of research in thisfield and thus can be extended further based on in situ hydrogenmechanism using different catalyst systems.

Acknowledgment

The authors acknowledge financial support from the Funda-mental Research Foundation of Sinopec (X505015).

References

[1] Mochizuki T, Itou H, Toba M, Miki Y, Yoshimura Y. Effects of acidic propertieson the catalytic performance of CoMo sulfide catalysts in selectivehydrodesulfurization of gasoline fractions. Energy Fuels 2008;22(3):1456–62.

[2] Rodríguez-Castellón E, Jiménez-López A, Eliche-Quesada D. Nickel and cobaltpromoted tungsten and molybdenum sulfide mesoporous catalysts forhydrodesulfurization. Fuel 2008;87(7):1195–206.

[3] Chen J, Ring Z. HDS reactivities of dibenzothiophenic compounds in a LC-finerLGO and H2S/NH3 inhibition effect. Fuel 2004;83(3):305–13.

[4] Depauw G, Froment G. Molecular analysis of the sulphur components in a lightcycle oil of a catalytic cracking unit by gas chromatography with massspectrometric and atomic emission detection. J Chromatogr A 1997;761(1–2):231–47.

[5] Song C. An overview of new approaches to deep desulfurization for ultra-cleangasoline, diesel fuel and jet fuel. Catal Today 2003;86(1–4):211–63.

[6] Ishihara A, Dumeignil F, Lee J, Mitsuhashi K, Qian E, Kabe T.Hydrodesulfurization of sulfur-containing polyaromatic compounds in lightgas oil using noble metal catalysts. Appl Catal A Gen 2005;289(2):163–73.

[7] Mochida I, Sakanishi K, Ma X, Nagao S, Isoda T. Deep hydrodesulfurization ofdiesel fuel: design of reaction process and catalysts. Catal Today 1996;29(1–4):185–9.

[8] Wojciechowska M, Pietrowski M, Czajka B. New supported ruthenium catalystfor hydrodesulfurization reaction. Catal Today 2001;65(2–4):349–53.

[9] Kouzu M, Kuriki Y, Hamdy F, Sakanishi K, Sugimoto Y, Saito I. Catalyticpotential of carbon-supported NiMo-sulfide for ultra-deephydrodesulfurization of diesel fuel. Appl Catal A Gen 2004;265(1):61–7.

[10] Rana M, Capitaine E, Leyva C, Ancheyta J. Effect of catalyst preparation andsupport composition on hydrodesulfurization of dibenzothiophene and Mayacrude oil. Fuel 2007;86(9):1254–62.

[11] Kouzu M, Uchida K, Kuriki Y, Ikazaki F. Micro-crystalline molybdenumsulfide prepared by mechanical milling as an unsupported model catalystfor the hydrodesulfurization of diesel fuel. Appl Catal A Gen 2004;276(1–2):241–9.

[12] Liu C, Flora T. HDS of DBT Using in situ Generated Hydrogen in the Presence ofDispersed Mo Catalysts III. Effects of Catalyst Precursors, H�2S, CO and H�2O.Chin J Catal 1999;20(6):591–6.

[13] Pawelec B, Fierro J, Montesinos A, Zepeda T. Influence of the acidity ofnanostructured CoMo/P/Ti-HMS catalysts on the HDS of 4,6-DMDBT reactionpathways. Appl Catal B Environ 2008;80(1–2):1–14.

[14] Solís D, Agudo A, Ramírez J, Klimova T. Hydrodesulfurization of hindereddibenzothiophenes on bifunctional NiMo catalysts supported on zeolite–alumina composites. Catal Today 2006;116(4):469–77.

[15] Al-Zeghayer Y, Sunderland P, Al-Masry W, Al-Mubaddel F, Ibrahim A,Bhartiya B. Activity of CoMo/-Al2O3 as a catalyst in hydrodesulfurization:effects of Co/Mo ratio and drying condition. Appl Catal A Gen 2005;282(1–2):163–71.

[16] Solis D, Klimova T, Ramírez J, Cortez T. NiMo/Al2O3–MgO(x) catalysts: theeffect of the prolonged exposure to ambient air on the textural and catalyticproperties. Catal Today 2004;98(1–2):99–108.

[17] Navarro R, Pawelec B, Fierro J, Vasudevan P, Cambra J, Arias P. Deephydrodesulfurization of DBT and diesel fuel on supported Pt and Ir catalysts.Appl Catal A Gen 1996;137(2):269–86.

[18] Deng X, Sun J, Yu S, Xi J, Zhu W, Qiu X. Steam reforming of ethanol forhydrogen production over NiO/ZnO/ZrO2 catalysts. Int J Hydrogen Energy2008;33(3):1008–113.

[19] Zhang L, Li W, Liu J, Guo C, Wang Y, Zhang J. Ethanol steam reforming reactionsover Al2O3�SiO2-supported Ni–La catalysts. Fuel 2009;88(3):511–8.

[20] Sun J, Qiu X, Wu F, Zhu W, Wang W, Hao S. Hydrogen from steam reforming ofethanol in low and middle temperature range for fuel cell application. Int JHydrogen Energy 2004;29(10):1075–81.

[21] Lima da Silva A, Malfatti C, Müller I. Thermodynamic analysis of ethanol steamreforming using Gibbs energy minimization method: a detailed study of theconditions of carbon deposition. Int J Hydrogen Energy 2009;34(10):4321–30.

[22] Koh A, Chen L, Leong W, Ang T, Johnson B, Khimyak T, et al. Ethanol steamreforming over supported ruthenium and ruthenium–platinum catalysts:comparison of organometallic clusters and inorganic salts as catalystprecursors. Int J Hydrogen Energy 2009;34(14):5691–703.

[23] Llorca J, Homs N, Sales J, de la Piscina P. Efficient production of hydrogen oversupported cobalt catalysts from ethanol steam reforming. J Catal2002;209(2):306–17.

[24] Frusteri F, Freni S, Spadaro L, Chiodo V, Bonura G, Donato S, et al. H2 productionfor MC fuel cell by steam reforming of ethanol over MgO supported Pd, Rh, Niand Co catalysts. Catal Commun 2004;5(10):611–5.

1370 Y. Muhammad et al. / Energy Conversion and Management 52 (2011) 1364–1370

[25] Freni S, Cavallaro S, Mondello N, Spadaro L, Frusteri F. Production of hydrogenfor MC fuel cell by steam reforming of ethanol over MgO upported Ni and Cocatalysts. Catal Commun 2003;4(6):259–68.

[26] Navarro R, Castaño P, Álvarez-Galván M, Pawelec B. Hydrodesulfurization ofdibenzothiophene and a SRGO on sulfide Ni (Co) Mo/Al2O3 catalysts. Effect ofRu and Pd promotion. Catal Today 2009;143(1–2):108–14.

[27] Lecrenay E, Sakanishi K, Mochida I, Suzuka T. Hydrodesulfurization activity ofCoMo and NiMo catalysts supported on some acidic binary oxides. Appl Catal AGen 1998;175(1–2):237–43.

[28] Gül Ö, Atanur O, Artok L, Erbatur O. The effect of additives onhydrodesulfurization of dibenzothiophene over bulk molybdenum sulfide:increased catalytic activity in the presence of phenol. Fuel Process Technol2008;89(4):419–23.

[29] Ishihara A, Kabe T. Deep desulfurization of light oil. 3. Effects of solvents onhydrodesulfurization of dibenzothiophene. Ind Eng Chem Res 1993;32(4):753–5.

[30] Isoda T, Nagao S, Ma X, Korai Y, Mochida I. Hydrodesulfurization of refractorysulfur species. 2. Selective hydrodesulfurization of 4,6-dimethyldibenzo-thiophene in the dominant presence of naphthalene over ternary sulfidescatalyst. Energy Fuels 1996;10(2):487–92.

[31] Farag H, Sakanishi K, Sakae T, Kishida M. Autocatalysis-like behavior ofhydrogen sulfide on hydrodesulfurization of polyaromatic thiophenes over asynthesized molybdenum sulfide catalyst. Appl Catal A Gen 2006;314(1):114–22.

[32] Egorova M, Prins R. The role of Ni and Co promoters in the simultaneous HDSof dibenzothiophene and HDN of amines over Mo/Al2O3 catalysts. J Catal(Print) 2006;241(1):162–72.

[33] Farag H, Whitehurst D, Sakanishi K, Mochida I. Carbon versus alumina as asupport for Co–Mo catalysts reactivity towards HDS of dibenzothiophenes anddiesel fuel. Catal Today 1999;50(1):9–17.