Monometallic Pd and Pt and Bimetallic Pd-Pt/Al2O3-TiO2 for the HDS of DBT: Effect of the Pd and Pt Incorporation Method

The effect of the preparation method of monometallic Pd and Pt and bimetallic Pd-Pt/Al 2 O 3 -TiO 2 catalysts on the hydrodesulfurization (HDS) of dibenzothiophene (DBT) was investigated in this study. The synthesis was accomplished using three methods: (A) impregnation, (B) metal organic chemical vapor deposition (MOCVD), and (C) impregnation-MOCVD. The bimetallic PdPt catalyst prepared by the impregnation-MOCVD method was most active for the HDS of DBT compared to those prepared by the single impregnation or MOCVD method due to the synergetic effect between both noble metals. The greater selectivity toward biphenyl indicated that this bimetallic Pd-Pt catalyst preferentially removes sulfur via the direct desulfurizationmechanism. However, the bimetallic Pd-Pt catalyst prepared using the single MOCVDmethod did not produce any cyclohexylbenzene, which is most likely associated with the hydrogenation/dehydrogenation sites.


Introduction
The hydrodesulfurization (HDS) of straight run gasoil (SRGO) is receiving considerable attention due to stringent environmental requirements, such as a lower sulfur content and the elimination of aromatic compounds in diesel fuel [1][2][3][4].The different sulfur-containing compounds in SRGO include dibenzothiophene (DBT), 4methyl-dibenzothiophene, and 4,6-dimethyl dibenzothiophene (46DMDBT) [5,6].The mechanisms for eliminating sulfur compounds follow two routes (i.e., direct desulfurization (DDS) and the hydrogenation (HYD)).However, the reactivity between DBT and 46DMDBT presents differences in terms of the preferential reaction pathways.In general, DBT is desulfurized via the DDS route while 46DMDBT is desulfurized via the HYD route [7,8].In this case, strong hydrogenating catalytic sites are preferable for the successful removal of refractory sulfur compounds, such as 46DMDT.
Conventional supported Ni-Mo sulfide hydrotreating catalysts become active in the HYD route only at high pressure due to the thermodynamic limitation [9].However, precious metals, such as Pd and Pt, have exhibited a higher hydrogenation ability at lower temperatures compared to metallic sulfides in hydrotreating catalysts.Unfortunately, monometallic Pd or Pt are easily poisoned by hydrogen sulfide.Therefore, their use as hydrogenation catalysts is only attractive when the feedstock is free of sulfur or when the sulfur-metal reaction is inhibited [10,11].Pd/Al 2 O 3 catalysts are better than Pt/Al 2 O 3 catalysts for the hydrodesulfurization of either DBT or 46DMDBT, which is due to the higher hydrogenation ability of palladium.Indeed, better hydrodesulfurization results have been observed with the use of bimetallic Pt-Pd/Al 2 O 3 catalysts [12,13].The sulfur resistance of the noble metal catalysts has been increased using bimetallic Pd-Pt supported on different mixed oxides, such as zeolite, SiO 2 -Al 2 O 3 , and amorphous SiO 2 [14][15][16][17][18][19].Recently, Galindo and de los Reyes [20] reported that monometallic Pd and Pt supported on Al 2 O 3 -TiO 2 exhibited higher HDS activity than those supported on Al 2 O 3 .In addition, Guofu et al. [21] reported that the integration of TiO 2 into Al 2 O 3 facilitates the dispersion of supported Pt-Pd bimetallic particles, which leads to higher hydrodesulfurization activity and better stability.
For the synthesis of HDS catalysts, conventional wet impregnation has been the most frequently employed method for metal incorporation.However, this process requires several steps in the preparation and has several limitations to good metallic dispersion [22].In this case, the metallic dispersion is strongly influenced by the metal precursors, temperature pretreatment, gas atmosphere, and other factors [23].By contrast, the metal organic chemical vapor deposition (MOCVD) method has been extended to the preparation of noble metal supported catalysts in a single step.In this case, fine metallic particles can be highly dispersed on the support [24][25][26].However, to our knowledge, there are no studies concerning the effect of bimetallic Pd-Pt catalysts prepared by single MOCVD or combined impregnation-MOCVD method in the HDS of DBT.Therefore, in this study, we report the effect of the integration method of a metallic active phase on the catalytic performance of monometallic Pd and Pt catalysts and bimetallic Pd-Pt catalysts on an Al 2 O 3 -TiO 2 support.The incorporation of metals was performed using different methods including conventional wet impregnation (IMP), single MOCVD, and a combination of the wet impregnation and MOCVD methods (IMP-MOCVD).The product selectivity of dibenzothiophene hydrodesulfurization was modified by the physicochemical properties of the bimetallic Pd-Pt/Al 2 O 3 -TiO 2 catalysts, which were influenced by the metal incorporation method.In this report, we try to explain the HDS catalytic activity as a function of the reducibility or hydrogenating properties and the particle size of the bimetallic catalysts.

Experimental Procedure
2.1.Support and Catalysts Preparation.The Al 2 O 3 -TiO 2 support was prepared by the sol-gel method following a previously published procedure [27] using 15 wt.%TiO 2 in the mixed oxide followed by annealing at 500 ∘ C. Supported monometallic Pd and Pt catalysts were prepared by either the IMP method or the MOCVD method.In the first metal incorporation method, palladium (II) nitrate dihydrate Pd[(NO 3 ) 2 ⋅2H 2 O] and tetraammineplatinum (II) nitrate [Pt(NO 3 ) 2 (NH 3 ) 4 ] were used as the metallic precursors.A proper solution concentration was used to achieve 1 wt.%Pd and 0.5 wt.% Pt in the final catalysts.The metal incorporation was performed using incipient wetness impregnation by spraying the Pd or Pt solutions.The single impregnated samples were dried at 120 ∘ C for 24 h followed by annealing at 500 ∘ C in air for 4 h.These samples were labeled Pd-IMP and Pt-IMP.The second metal incorporation method (i.e., single MOCVD) employed (CH 3 -COCHCO-CH 3 ) 2 Pd and (CH 3 -COCHCO-CH 3 ) 2 Pt as the metal organic complexes.The catalysts were prepared in a horizontal hot-wall type CVD apparatus using the proper quantities of the support and metal precursor to obtain 1.5 wt.% Pd and 0.5 wt.% Pt the monometallic Pd and Pt catalysts.Each metallic precursor (i.e., Pd or Pt) was separately evaporated at 180 ∘ C. The deposition temperature ( dep ) was maintained at 400 ∘ C, and the total pressure ( tot ) was maintained at 1 Torr.These samples denoted Pd-MOCVD and Pt-MOCVD.The bimetallic Pd-Pt catalysts supported on the Al 2 O 3 -TiO 2 carrier were prepared by (A) single wet impregnation (IMP), (B) single MOCVD, and (C) a combined Pd-IMP/Pt-MOCVD method, and the resulting catalysts were labeled Pd-Pt-IMP, Pd-Pt-MOCVD, and Pd-IMP/Pt-MOCVD, respectively.A nominal loading of 1 wt.%Pd and 0.5 wt.% Pt was used for all of the catalysts.The bimetallic Pd-Pt-IMP catalyst was prepared by wet impregnation of the Al 2 O 3 -TiO 2 using the same palladium (II) and platinum (II) solutions.The wet catalyst was dried and annealed as mentioned above for the monometallic catalysts.The bimetallic Pd-Pt-MOCVD catalyst was prepared using proper amounts of both Pd(acac) 2 and Pt(acac) 2 precursors, which were evaporated together at 180 ∘ C followed by drying and annealing according to the abovementioned conditions.The preparation of bimetallic catalyst Pd-IMP/Pt-MOCVD was carried out in two steps.First, the Pd was incorporated using the IMP method with an aqueous solution of palladium (II) nitrate, and then the Pt was integrated using the MOCVD method to yield Pd-IMP/Pt-MOCVD.

Support and Catalysts Characterization.
Characterizations of the alumina-titania support, the monometallic IMP and MOCVD catalysts, and the bimetallic IMP, MOCVD, and IMP/MOCVD catalysts were performed.Additional characterization of Pd-MOCVD, Pt-MOCVD, and Pd-Pt-MOCVD was performed for the freshly prepared samples.

Textural Properties of the Al
O 3 -TiO 2 Support.Textural properties of the support were determined by N 2 physisorption at 77 K with a Micromeritics ASAP-2000 apparatus.Surface areas and pore size distributions were calculated using the BET and desorption BJH methods, respectively.

XRD and Raman Spectroscopy.
Powder X-ray diffraction (XRD) of the alumina-titania support was performed on a Siemens D8 focus diffractometer with Cu K radiation ( = 0.15406 nm).The Raman spectra of the aluminatitania support were obtained at room temperature using a Horiba Jobin Yvon Inc. T64000 spectrometer equipped with a confocal microscope (Olympus BX41) with a 514.5 nm line generated by an Ar + laser (5 mW).

Temperature Programmed Reduction (TPR).
Temperature programmed reduction (H 2 -TPR) of catalysts previously annealed at 500 ∘ C was conducted with Altamira AMI200 TPR equipment.The experiments were performed in a quartz reactor coupled to a thermal conductivity detector (TCD), and the reduction stream included an H 2 (10%vol)/argon certified mixture.50 mg of each sample was reduced from 273 to 1123 K by passing the gas mixture at a flow rate of 30 mL/min (10 ∘ K/min was the heating rate).

High Resolution Transmission Electron Microscopy (HRTEM).
High resolution transmission electron microscopy (HRTEM) was performed with an FEI Tecnai 20 instrument at a point resolution of 1.7 > and an acceleration voltage of 200 keV.The samples of the monometallic and bimetallic Pd and Pt catalysts (fresh, sulfided, and spent) were suspended in acetone.One droplet of the suspension was applied to a 400-mesh carbon-coated copper grid and left to dry in air.

Catalytic Evaluation of the HDS of DBT.
Prior to the catalytic evaluation, all of the catalysts were heated at 400 ∘ C in an N 2 atmosphere and then sulfided in situ by flowing a 10% H 2 S/H 2 gas mixture at 400 ∘ C for 1 h.Next, catalyst activation was tested for the HDS of DBT in a stirred batch reactor (Parr 4562 M).The reaction mixture was prepared by dissolving ∼0.3 g of Aldrich DBT in 100 cm 3 of Aldrich n-hexadecane (98 wt-% and 99 wt-% purity, resp.)followed by the addition of 200 mg of sieved catalyst (80-100 Tyler mesh, 0.165 mm average particle diameter).The operating conditions, which were carefully chosen to avoid external and/or internal diffusion limitations, were  = 5.59 ± 0.03 MPa, T = 350 ± 2 ∘ C, and 1000 rpm (∼105 rad s −1 ) mixing speed.Samples were periodically (1, 2, 3, and 4 h) analyzed in a gas chromatograph Perkin-Elmer AutoSystem XL (flame ionization detector and column Econo-CAP5 from Alltech).HDS kinetic constants were estimated assuming pseudo-first-order kinetics referred to the DBT concentration.

Results and Discussion
3.1.Al 2 O 3 -TiO 2 Support.The BET specific surface area and average pore volume of the mixed oxide Al 2 O 3 -TiO 2 support were 275 m 2 g −1 and 0.76 cm 3 g −1 , respectively.In addition, the mean pore diameter was 110 Å.The X-ray diffraction analysis (not shown) of the Al 2 O 3 -TiO 2 support indicated the presence of well-crystallized -Al 2 O 3 and signals corresponding to crystalline TiO 2 in the anatase phase.The appearance of the signal corresponding to the anatase phase suggested that it is well crystallized and segregated with TiO 2 domains [21].The laser Raman spectra of the support exhibited (not shown) signals corresponding to the anatase phase that were shifted to 395, 517, and 637 cm −1 [28,29].

Temperature Programmed Reduction (TPR) of the Monometallic and Bimetallic Catalysts.
Figure 1 shows the TPR of the monometallic oxide (Pd and Pt catalysts) incorporated on the Al 2 O 3 -TiO 2 support by the IMP method.The monometallic Pd-IMP catalyst was reduced from room temperature to 850 ∘ C. In addition, the negative profiles correspond to palladium hydride (Pd-H x ) formation at 102 ∘ C, and some changes of these types of sites were observed at 381 and 616 ∘ C (Figure 1).For the reduction profile, the PdO over Al 2 O 3 is immediately reduced at 30 ∘ C but when PdO is deposited on TiO 2 , it is expected to be reduced below 0 ∘ C (not shown).The absence of reduction peaks at room temperature suggests that PdO could be deposited onto the TiO 2 rich matrix.However, unimportant variations in the temperature of the Pd-H x formation were observed when Ti 4+ is present in the Al 2 O 3 matrix, as reported by Wang et al. [30].However, the temperature of this event was higher than that reported for the Pd-H x formation on Al 2 O 3 (80 ∘ C) and Al 2 O 3 -TiO 2 (75 ∘ C).This increase in the temperature could be associated with the strong interaction between PdO and the Al 2 O 3 -TiO 2 support.In addition, the high intensity of the main peak (i.e., Pd-H x ) suggests low dispersion of the Pd particles in this catalyst [23,31].
The TPR of the monometallic Pt-IMP catalyst exhibits a first reduction peak at 220 ∘ C, which corresponds to the reduction of the Pt 2+ species to Pt 0 due to interaction with Al 2 O 3 .The PtO reduction on Al 2 O 3 has been reported at 220-250 ∘ C [32,33].In addition, the second reduction peak was observed at 460 ∘ C due to the partial reduction of TiO 2 by hydrogen with the metal-support interaction (MSI), which has been reported to occur in the range of 360-600 ∘ C [34].This H 2 consumption peak for Pt/TiO 2 is shifted to a higher temperature, which implies a strong metal-support interaction (SMSI) between Pt and the TiO 2 rich matrix.This result is associated with the reduction of TiO 2 promoted by the presence of Pt, which suggested that Pt is preferentially deposited onto the TiO 2 rich matrix and only a small amount of Pt is deposited on the Al 2 O 3 surface.
The TPR profiles of the bimetallic Pd-Pt-IMP, Pd-Pt-MOCVD, and Pd-IMP/Pt-MOCVD catalysts are shown in Figure 2. The TPR profiles contain only a few peaks including a positive peak and a negative one for all of the samples.The first peak appears at 25 ∘ C due to PdO reduction of Pd +2 to Pd 0 , and the second peak was observed at 67-97 ∘ C due to the Pd-H x formation on the Pd crystallites [31].Based on the analysis of the TPR profile of the bimetallic Pd-Pt catalysts, we can assume that they are not a simple sum of the monometallic samples.Apparently, the Pd and Pt particles coexist in close proximity, which can be deduced from the shift in the peak corresponding to the Pd-H x formation.The decrease in the formation of Pd-H x is influenced not only by the dispersion of Pd particles but also by the Pt content in the Pd-Pt alloy [31].In our case, the negative peak of the Pd-H x formation for the three bimetallic catalysts appears at different temperature.For the IMP catalysts, the negative peak appears at 67 ∘ C, and for the MOCVD catalysts, the negative peak appears at 75 ∘ C with a shoulder at 67 ∘ C. For the IMP/MOCVD catalyst, the negative peak appears at 99 ∘ C, which suggests that part of the deposited Pd particles in the bimetallic Pd-Pt catalysts is present either as isolated metal particles with low dispersion or alloyed with Pt.For both bimetallic IMP and MOCVD catalysts, the peak corresponding to the Pd-H x formation is in agreement with that reported by Baldovino-Medrano et al. [35], where the interaction between Pd and Pt modifies the temperature of the Pd-H x formation to 63 ∘ C. Navarro et al. [16] detected a broad peak at 379 ∘ C, which was due to the reduction of the Pd particle along with the coimpregnated Pt particles resulting from an interaction between them.However, in our case, these interactions are present only in the IMP catalysts.
Only the bimetallic Pd-Pt-IMP catalyst exhibited reduced PtO on Al 2 O 3 at 216 ∘ C. In the remaining catalysts prepared by the MOCVD method, this reduction was not observed, which suggests that when the Pt is incorporated with the MOCVD method using an organometallic precursor, the Pt could be present as elemental platinum (Pt 0 ), as was reported by Stakheev and Kustov [36].In addition, for the bimetallic IMP/MOCVD catalyst, the presence of Pt 0 slightly modifies the nature of the PdO crystalline particles previously deposited by the IMP method leading to species with a different nature, which results in the slight shift in the peak corresponding to Pd-H x formation (99 ∘ C).However, the temperature of its formation is similar to that of the monometallic Pd-IMP catalyst (102 ∘ C, see Figure 1), which indicates the presence of poorly dispersed Pd particles that can influence the activity and selectivity of the HDS of DBT.
For all of the bimetallic catalysts, the Al 2 O 3 -TiO 2 support did not exhibit an H 2 consumption signal at high temperatures (400-600 ∘ C), which indicated that the partial reduction of Ti 4+ to Ti 3+ did not occur after Pt incorporation [34].This result is not surprising for the bimetallic MOCVD and IMP/MOCVD catalysts because the MOCVD method affords Pt 0 species, which prevent the strong metal-support interaction (SMSI) between PtO and the TiO 2 matrix.By contrast, for the bimetallic IMP catalysts, the PtO particles are most likely formed only on the Al 2 O 3 matrix.

HRTEM of the Freshly Prepared Monometallic Pd and
Pt Alumina-Titania Catalysts.Figures 3(a)-3(d) show the HRTEM images of the Pd and Pt catalysts prepared using the different incorporation methods.Micrograph images revealed that the support contains monometallic nanoparticles of Pd and Pt.The nanoparticles are preferentially formed near or over the nanocrystallites of the well-dispersed TiO 2 in the network.For TiO 2 dispersion of this type of material, we have demonstrated by TEM that the presence of very small TiO 2 particles in the support is homogeneously dispersed over and inside the alumina matrix, which facilitates the formation of more active metallic phases [27].A predominant unimodal size distribution varying from 5 to 10 nm was observed for the monometallic Pd and Pt particles incorporated using the impregnation (IMP) method.However, these particles tend to agglomerate, which resulted in the low dispersion of the metallic particles as indicated by the TPR study.By contrast, when the metal incorporation was performed with the MOCVD method, the monometallic catalysts exhibited good dispersion and homogeneity of the small metallic particles on the Al 2 O 3 -TiO 2 support.The main difference is the presence of a bimodal size distribution with particles varying in size from 5 to 8 nm with an average particle size of 2-5 nm.In both cases, these monometallic particles are preferentially formed near the crystallites of TiO 2 present in the network, as indicated by the TPR results.

HRTEM of the Freshly Prepared Bimetallic
Pd-Pt Catalysts.More information regarding the particle size of both the Pd and Pt particles over the Al 2 O 3 -TiO 2 support could be determined using HRTEM characterization of the bimetallic catalysts.Figures 4(a)-4(c) show the HRTEM micrographs of the bimetallic Pd-Pt-IMP, Pd-Pt-MOCVD, and Pd-IMP/Pt-MOCVD catalysts.From numerous HRTEM micrographs, histograms of the particle size distribution were obtained for the three catalysts.It is important to note that, due to the low metal loading, the particles are difficult to observe in most areas, and these particles are inhomogeneously distributed.Therefore, Figures 4(a)-4(c) are not typical of the overall distribution but instead represent an area with a particular nanoparticles density.Figures 4(a)-4(c) show the influence of the incorporation method on the crystallite size of the bimetallic Pd-Pt catalysts.A predominant unimodal size distribution varying from 5 to 10 nm was observed for the codeposited Pd-Pt particles incorporated by either the single impregnation or MOCVD method.However, there are differences in the size of the bimetallic particles depending on the method of preparation, and although the sample of Pd-Pt obtained by impregnation shows small crystallites with an average size of 5-8 nm, there are some regions where crystallites tend to form agglomerates.  the growth of smaller Pt 0 crystallites on previous and larger impregnated PdO particle, which could account for the increase in the activity for HDS.

HDS of DBT over the Monometallic Catalysts.
Figure 5 shows the DBT hydrodesulfurization rate constants of the monometallic catalysts.The most active catalyst was Pt/ Al 2 O 3 -TiO 2 , especially when prepared using the MOCVD method.In both cases, the low metallic content (0.5% wt.) allows for a high dispersion of Pt on the TiO 2 matrix, and the interactions between Pt and TiO 2 may be responsible for the high catalytic activity that was observed, which is in agreement with the report published by Galindo and de los Reyes [20] and Baldovino-Medrano et al. [35].The opposite effect was observed for the monometallic Pd catalysts where better catalytic behavior was obtained over the Pd-IMP catalyst.In this particular case, the low HDS activity of both the Pd-IMP and Pd-MOCVD catalysts may be associated with the low dispersion of the Pd particles.

Selectivity of the Monometallic Catalysts.
The product reaction species observed in the HDS of DBT over the monometallic Pd and Pt catalysts included biphenyl (BP) as the direct desulfurization product as well as tetrahydrodibenzothiophene (TH-DBT) and cyclohexylbenzene (CHB) as the intermediate and final products of the hydrogenation route, respectively.Figures 6(a)-6(b) show the selectivity profiles for the HDS of DBT over the monometallic Pd catalysts prepared using the IMP and MOCVD method.The HDS of DBT with the Pd catalyst exhibited significant differences in the mechanism pathway depending on the Pd incorporation method.When the Pd was incorporated with the IMP method, a tendency to form THDBT and CHB products was observed (46% and 17%, resp.).In this case, the low selectivity towards the BP product (36%) at the end of reaction suggests that the hydrogenation route (HYD) proceeds more rapidly than the C-S bond breaking (DDS route), which indicates the presence of Pd hydrogenation sites over the monometallic Pd-IMP.According to Baldovino-Medrano et al. [35], the high selectivity towards products in the HYD route over Pd is associated with a stronger interaction of the aromatics with the Pd active phase and with a lower rate of atomic H transfer from the metal to these rings.
By contrast, the amount of products formed (i.e., THDBT and BP products (50%)) in the HDS of DBT using the monometallic Pd-MOCVD catalyst (Figure 6(b)) was the same as the amount produced in the first hour of the reaction.However, as the reaction progressed, the selectivity of the products evolved resulting in a 6% selectivity for THDBT at the end of the evaluation.In addition, this Pd-MOCVD catalyst exhibited less catalytic activity than the analogous one obtained using the IMP method.The most important feature is the dramatic increase in the presence of BP (94%) with a concomitant disappearance of the THDBT and CHB products.Therefore, once formed, THDBT is dehydrogenated due to the reversibility of the reaction leading to the reformation of DBT, which is eventually desulfurized by the direct mechanism.The hydrogenation/dehydrogenation is faster than the desulfurization route in the HDS of dibenzothiophene over the Pd--Al 2 O 3 catalyst at temperatures of 290-330 ∘ C [35].Because our tests were performed at 350 ∘ C, the dehydrogenation route may be taking place, which may explain the absence of CHB due to the reversibility of the reaction.
The HDS of DBT with the monometallic Pt catalysts prepared by either the IMP or MOCVD method exhibited a similar catalytic selectivity (Figure 7).The Pt-IMP catalyst exhibited a higher selectivity for the formation of BP (93%) compared to CHB (3%), which confirms a higher C-S bond hydrogenolysis activity (DDS route).The Pt-MOCVD catalyst exhibited a slightly different behavior to that observed for the Pt-IMP catalyst.In this case, the presence of THDBT (approximately 14%) after 1 hour of reaction is slightly higher compared to the analogous catalyst obtained using the IMP method.In addition, the selectivity to BP was approximately 86%.However, at 4 hours of reaction, the selectivity to BP reached 93%, and the presence of the intermediates of the HYD route precursors decreased to less than 7%, which is very similar to that observed for the Pt-IMP catalyst.Therefore, the high selectivity towards BP indicated that the DDS route is more favorable than the HYD route, which is consistent with the previously reported results [1, 23].

HDS of DBT over the Bimetallic Catalysts. Figure 8(a)
shows the kinetic constants for the HDS of DBT after four hours of reaction for both bimetallic catalysts obtained using either the single IMP or MOCVD method, which indicates that the HDS activity of both bimetallic catalysts is nearly the same.Although the bimetallic Pd-Pt-MOCVD catalyst is slightly more active than the Pd-Pt-IMP catalyst, both catalysts exhibit a lower activity than the monometallic Pt catalysts (Figure 5).This low HDS activity may be due to an inhibitor effect of Pd on Pt or to the HDS reaction being strongly influenced by the presence of Pd.

Selectivity of the Pd-Pt Bimetallic
Catalyst for the HDS of DBT.The product reaction species observed in the HDS of DBT over the bimetallic Pd-Pt catalysts included biphenyl (BP) as the direct desulfurization product as well as tetrahydrodibenzothiophene (TH-DBT) and cyclohexylbenzene (CHB) as the intermediate and final product of the hydrogenation route, respectively.These products are similar to those obtained from the monometallic Pt catalysts.The Pd-Pt-IMP catalyst did not exhibit significant differences during the evaluation period (Figure 8(b)).After one hour of reaction, this catalyst exhibited selectivity toward the formation of BP, which reached 93%.However, the formation of prehydrogenated derivatives was less than 7%.The presence of BP remained constant irrespectively of the reaction time.Based on these results, the selectivity of the products was influenced by the presence of the Pt species on the support surface.As observed for the monometallic Pt catalyst (Figure 7), desulfurized products were obtained, which has been previously reported [1,15,23].
The bimetallic Pd-Pt-MOCVD catalyst exhibited a slightly different trend compared to that observed for the catalyst prepared using the single IMP method.Surprisingly, the Pd-Pt-MOCVD catalyst only produced BP and THDBT without the formation of CHB (Figure 8(b)), which indicates that, for this particular catalyst, the subsequent hydrogenation step of THDBT is rate determining.This result is similar to that observed for the monometallic Pd-MOCVD catalyst.Here, the selectivity may be associated with the presence of isolated Pd species.As proposed by Houalla et al. [7], the reaction pathways for the HDS of DBT are the direct desulfurization to BP and hydrogenation to yield THDBT or HHDBT, which are further desulfurized to CHB.In our case, the formation of HHDBT was not detected with the Pd-Pt-MOCVD catalyst.In addition, for this catalyst, only the formation of BP was observed, which may be result from direct desulfurization as well as dehydrogenation of THDBT due to the ability of Pd to dehydrogenate these types of rings [35].The Pd/(Pd + Pt) molar ratio can modify the competition between the hydrogenation-dehydrogenation C-S-C bond scission reactions occurring during the HDS of DBT, which indicates that the optimal molar ratio is (0.8) for Pd-Pt deposited on Al 2 O 3 .In addition, bimetallic Pd-Pt/Al 2 O 3 prepared from nitrate or organometallic precursors possesses a higher DDS selectivity than those prepared from chloride precursors [35].In our case, the bimetallic Pd-Pt supported on Al 2 O 3 -TiO 2 exhibited DDS selectivity because the Pd/(Pd + Pt) molar ratio (0.75) was close to the optimal value and they were incorporated from organometallic precursors.

HDS Activity of DBT with the Bimetallic IMP/MOCVD
Catalyst.Because the monometallic Pd-IMP and Pt-MOCVD catalysts exhibited individually high HDS activity, we prepared a bimetallic Pd-Pt catalyst by combining the IMP and MOCVD methods.Figure 9(a) shows the kinetic constants for the HDS of DBT for the bimetallic catalysts compared to their respective monometallic catalysts (i.e., Pd-IMP and Pt-MOCVD).After four hours of reaction, the HDS activity of the Pd-IMP/Pt-MOCVD catalyst was twice as high as that of the monometallic Pt-MOCVD catalyst, which indicated that a synergic effect was achieved due to the interaction of the Pd particles over TiO 2 matrix when both Pd and Pt particles were incorporated using a combination of the IMP and MOCVD method, respectively.In addition, the HDS activity of the better bimetallic catalyst was four times higher than that obtained from the catalysts prepared with either the single IMP or MOCVD method (Figure 8(a)), which indicated that the catalytic activity was strongly improved when Pt 0 is dispersed or deposited on the PdO particles that were previously incorporated onto the Al 2 O 3 -TiO 2 support rather than being incorporated in one step.This result indicates that the HDS activity of the bimetallic catalyst was influenced by the metal incorporation methods.
3.6.1.Selectivity of Pd-Pt-IMP/MOCVD for the HDS of DBT.The results in Figure 9(b) show that the selectivity results of this catalyst closely resemble the trend observed for the bimetallic Pd-Pt-IMP catalysts.At a shorter reaction time (1 hour), significant amounts of intermediate prehydrogenated products were detected.However, the selectivity towards BP reached 87%, but after 2 hours of reaction the selectivity increased to 91% for BP and ∼3 and ∼5% for THDBT and CHB, respectively.Then, these values remained constant.This behavior suggests that the HDS selectivity was influenced by the presence of Pt dispersed on either the PdO or TiO 2 rich matrix surface.

Conclusions
In this study, we synthesized Al 2 O 3 -TiO 2 supported bimetallic Pd-Pt catalysts prepared using a single IMP, single MOCVD, and combined IMP/MOCVD methods.The most active bimetallic catalyst for the HDS of DBT was the catalyst prepared using the combined IMP/MOCVD method.This HDS activity of this catalyst was four times higher than that obtained with catalysts prepared using the single IMP or MOCVD methods.Based on the observed selectivity, all of the catalysts exhibited a preference for the removal of sulfur via direct desulfurization.However, the single MOCVD method does not considerably improve the activity of HDS of DBT but influences the selectivity where the reversibility of the reaction appears to favor the direct desulfurization pathway, which is the preferred mechanism in the DBT reaction.The TPR and TEM results indicated that the incorporation of active noble metals by combining techniques influences the particle size and dispersion of Pt over the Pd particles and may be responsible for improving the catalytic activity for the HDS of DBT.

Figure 6 :Figure 7 :
Figure 6: Profiles of HDS selectivity in the HDS of DBT reaction over monometallic Pd catalysts prepared by (a) IMP and (b) MOCVD method.