Effect of 2,6-Bis-(1-hydroxy-1,1-diphenyl-methyl) Pyridine as Organic Additive in Sulfide NiMoP/γ-Al2O3 Catalyst for Hydrodesulfurization of Straight-Run Gas Oil

The effect of 2,6-bis-(1-hydroxy-1,1-diphenyl-methyl) pyridine (BDPHP) in the preparation of NiMoP/γ-Al2O3 catalysts have been investigated in the hydrodesulfurization (HDS) of straight-run gas oil. The γ-Al2O3 support was modified by surface impregnation of a solution of BDPHP to afford BDPHP/Ni molar ratios (0.5 and 1.0) in the final composition. The highest activity for NiMoP materials was found when the molar ratio of BDPHP/Ni was of 0.5. X-ray diffraction (XRD) results revealed that NiMoP (0.5) showed better dispersion of MoO3 than the NiMoP (1.0). Fourier transform infrared spectroscopy (FT-IR) results indicated that the organic additive interacts with the γ-Al2O3 surface and therefore discards the presence of Mo or Ni complexes. Raman spectroscopy suggested a high Raman ratio for the NiMoP (0.5) sample. The increment of the Mo=O species is related to a major availability of Mo species in the formation of MoS2. The temperature programmed reduction (TPR) results showed that the NiMoP (0.5) displayed moderate metal–support interaction. Likewise, X-ray photoelectron spectroscopy (XPS) exhibited higher sulfurization degree for NiMoP (0.5) compared with NiMoP (1.0). The increment of the MoO3 dispersion, the moderate metal–support interaction, the increase of sulfurization degree and the increment of Mo=O species provoked by the BDPHP incorporation resulted in a higher gas oil HDS activity.


Introduction
The pollution is a problem in the society today. In this sense, environmental legislations are more rigorous with contaminants in diesel fuel, and the sulfur content in diesel is about 10 ppm [1]. The production of ultra-low sulfur diesel is of paramount importance in the hydrotreatment transmittance mode on a Perkin Elmer Spectrum One (Waltham, MA, USA) with a universal attenuated total reflectance (ATR) sampling accessory in the 1800-1300 cm −1 range. Spectra were measured at room temperature at a 4 cm −1 resolution and an average of 256 scans per sample was used. Raman spectra were obtained at room temperature on a T64000 triple monochromator (Jobin-Yvon-Horiba, Edison, NJ, USA) using the 514.5 nm line of an Ar+ laser (Lexel Laser, Fremont, CA, USA). All the spectra were obtained at a power of 10 mW at the laser head, in the range 10-1600 cm −1 , using an Olympus microscope (Shinjuku, Tokio, Japan) with a 100× objective and 10 accumulations of 60 s each. The spectrum resolution was 1 cm −1 .
Thermal analyses (from room temperature to 1000 • C) of freshly dried samples (ca. 15 mg) were carried out with a Netzch Thermische Analize, STA 409 EP apparatus (Selb, Germany) under a static air atmosphere, operating at 10 • C/min heating ramp. Temperature programmed reduction (TPR) of the catalysts were performed using Altamira Instruments AMI-90 equipped with a thermal conductivity detector (TCD). Approximately 0.1 g of catalyst sample was placed in a quartz sample cell (U-shaped) and then pretreated in situ at 393 K for 1 h under Ar flow. The reduction of catalysts was performed from room temperature to 1273 K, under a stream of 10% H 2 , employing a gas flow rate of 0.84 cm 3 /s and heating rate of 10 • C/min. XPS spectra of the sulfided catalyst samples were measured at room temperature using a VG Escalab 200R spectrometer (Richardson, TX, USA) equipped with a hemispherical electron analyzer and a MG Kα (hν = 1253.6 eV) X-ray source. The details of the XPS measurements by this spectrometer are reported elsewhere [21].

Catalytic Activity
The solid catalysts (1.20 mL; 80-100 mesh) were sulfided in virgin straight run gas oil (2.05 wt % S) spiked with dimethyl disulfide (DMDS) to afford 2.5 wt % of sulfur in the feedstock and then fed to the reactor at a flow rate according to an liquid hourly space velocity (LHSV) of 2.5 h −1 . The pressure was set to 56 kg/cm 2 and the hydrogen rate was adjusted according to a H 2 /oil ratio of 500 m 3 (STP)/m 3 . The temperature of the catalyst bed was increased to 290 • C and held for 15 h. These were tested in the HDS reaction of SRGO and H 2 flow rates were adjusted to LHSV 1.5 h −1 , 370 • C, H 2 /oil ratio equal to 500 m 3 (STP)/m 3 and 56 kg/cm 2 . The sulfur concentration in liquids was obtained with a Tanaka RX-360SH model sulfur analyzer according to the ASTM D4294 method. For each test, 10 mL of sample was fed to the equipment. The reported instrument sensitivity is ±0.001% of the measured value.
The apparent rate constant rates (k hds (Sw%) −0.5 h −1 ) were estimated at steady state conditions (after 12 h), and they were calculated according to the following equation [22,23]: where: k hds : Pseudo 1.5 order HDS kinetic constant (Sw% −0.5 h −1 ); S p : Sulfur in product (wt %); S f : Sulfur in feedstock (wt %); LHSV: Liquid Hourly Space Velocity (h −1 ). Table 1 shows the textural properties for NiMoP (0.0), NiMoP (0.5) and NiMoP (1.0). The reference catalyst showed lower specific area, pore volume and pore size in comparison with the catalysts with the organic additive. The NiMoP (1.0) presented similar pore size and total volume to NiMoP (0.5). However, the specific area for NiMoP (0.5) slightly decreased (156 vs. 164 m 2 /g) in contrast with NiMoP (1.0). In addition, the catalysts with organic additive have more specific area than the NiMoP (0.0) suggesting that the BDPHP slightly modifies the textural properties.

Fourier Transformed Infrared Spectroscopy (FT-IR)
The FT-IR spectroscopy allowed us to identify the vibrations of BDPHP on the catalyst NiMoP/ γ-Al 2 O 3 before the HDS reaction. Figures 2 and 3 show the spectra for 2,6-Bis-(1-hydroxy-1,1-diphenylmethyl) pyridine and NiMoP/γ-Al 2 O 3 at different BDPHP/Ni molar ratio (0.5 and 1.0). The peaks in the 1800-1100 cm −1 frequency region are attributed to the carbon and nitrogen vibrations [28][29][30]. In this region, the spectra of BDPHP exhibited some vibrational frequencies corresponding to the ring vibration (1664 and 1589 cm −1 ), C=C (1570 cm −1 ), C=N (1490 cm −1 ), C-C (1444 and 1353 cm −1 ), C-NH 2 (1334 cm −1 ), C-N (1261 cm −1 ) and C-H (1197 and 1163 cm −1 ) [28][29][30].  The NiMoP (0.5) and NiMoP (1.0) exhibited vibrational frequencies corresponding to C-C vibration at 1462, 1447 and 1384 cm −1 [28][29][30]. Moreover, the NiMoP (0.5) and NiMoP (1.0) exhibited a peak at 1632 cm −1 corresponding to Al-O-C vibration [29][30][31]. Furthermore, the CoMoP (0.5) and CoMoP (1.0) presented C=N vibration at 1494 cm −1 [28][29][30]. In contrast with BDPHP spectra (Figure 2), the catalysts presented the C=N vibration close to the wavenumber of BDPHP. This result discards the formation of Ni or Mo interaction with the free pair of electrons of C=N from the organic additive. In addition, this result is related to the low ability complex for some organic additives with a larger size and some metals [32]. However, the OH groups of the organic additive contribute to the formation of new Al-O-C [31]. In this line, the BDPHP covered the γ-Al2O3 prevented the strong metal-support interaction [31,33].  The NiMoP (0.5) and NiMoP (1.0) exhibited vibrational frequencies corresponding to C-C vibration at 1462, 1447 and 1384 cm −1 [28][29][30]. Moreover, the NiMoP (0.5) and NiMoP (1.0) exhibited a peak at 1632 cm −1 corresponding to Al-O-C vibration [29][30][31]. Furthermore, the CoMoP (0.5) and CoMoP (1.0) presented C=N vibration at 1494 cm −1 [28][29][30]. In contrast with BDPHP spectra (Figure 2), the catalysts presented the C=N vibration close to the wavenumber of BDPHP. This result discards the formation of Ni or Mo interaction with the free pair of electrons of C=N from the organic additive. In addition, this result is related to the low ability complex for some organic additives with a larger size and some metals [32]. However, the OH groups of the organic additive contribute to the formation of new Al-O-C [31]. In this line, the BDPHP covered the γ-Al2O3 prevented the strong metal-support interaction [31,33]. The NiMoP (0.5) and NiMoP (1.0) exhibited vibrational frequencies corresponding to C-C vibration at 1462, 1447 and 1384 cm −1 [28][29][30]. Moreover, the NiMoP (0.5) and NiMoP (1.0) exhibited a peak at 1632 cm −1 corresponding to Al-O-C vibration [29][30][31]. Furthermore, the CoMoP (0.5) and CoMoP (1.0) presented C=N vibration at 1494 cm −1 [28][29][30]. In contrast with BDPHP spectra (Figure 2), the catalysts presented the C=N vibration close to the wavenumber of BDPHP. This result discards the formation of Ni or Mo interaction with the free pair of electrons of C=N from the organic additive. In addition, this result is related to the low ability complex for some organic additives with a larger size and some metals [32]. However, the OH groups of the organic additive contribute to the formation of new Al-O-C [31]. In this line, the BDPHP covered the γ-Al 2 O 3 prevented the strong metal-support interaction [31,33].

Raman Spectroscopy
Raman spectroscopy was used to determine the Mo species in NiMoP/γ-Al2O3 catalysts at different molar ratios of BDPHP/Ni. Figure 4 shows the Raman spectra for NiMoP (0.0), NiMoP (0.5) and NiMoP (1.0), which is supported by γ-Al2O3. The NiMoP (0.0) spectrum exhibiting bands at 176, 255, 376, 638, 714, 853, 975, and 993 cm −1 . The bands at 993, 975, 376, 225, 376 and 176 cm −1 are assigned to Mo7O24 6− [34], while the intense 853 cm −1 signal is ascribed to large MoO3 aggregates [35]. The peaks at 638 and 714 cm −1 are associated with the γ-Al2O3 [36]. The NiMoP with BDPHP/Ni (0.5, 1.0) spectra showed bands at 220, 360, 565, 895, 950 and 1015 cm −1 . The band at 895 cm −1 corresponds to MoO vibration of MoO4 2− while the bands at 220, 360, 565 and 950 cm −1 are associated with Mo7O24 6− species [34]. Additionally, the peak at 1015 cm −1 is associated with P-O vibrations [28,34]. Regarding the band located at 950 cm −1 , it increases gradually conversely with the decrease of the BDPHP/Ni ratio, most probably due to the BDPHP, which promotes the polymerization of the polymeric Mo7O24 6− species. This, in fact, suggests changes in the Mo oxide dispersion with the addition of BDPHP. The Mo species at 220 and 360 cm −1 is associated with the Mo-O-Mo [37]. In addition, the Mo species at 895 and 950 cm −1 is ascribed to the Mo=O vibrations [37]. To obtain information about the Mo oxide dispersion, the Raman spectra were deconvoluted with the Mo=O, Mo-O and Mo-O-Mo vibrations. Figure 5 shows the deconvolution for NiMoP (0.5). Figure Figure 6 showed that the Raman ratio increases until a maximum at BDPHP/Ni = 0.5, and then decreases at BDPHP/ Ni = 1.0. This result suggests the re-dispersion of Mo oxide provoked by BDPHP with different BDPHP/Ni molar ratio. A similar behavior was seen by Diaz de Leon [38] in an analogous system with additives. Regarding this, Diaz de Leon suggested that the increase of the Raman band intensities ratio promoted better dispersion of the metal species and is responsible for the highest HDS activity. The Mo species at 220 and 360 cm −1 is associated with the Mo-O-Mo [37]. In addition, the Mo species at 895 and 950 cm −1 is ascribed to the Mo=O vibrations [37]. To obtain information about the Mo oxide dispersion, the Raman spectra were deconvoluted with the Mo=O, Mo-O and Mo-O-Mo vibrations. Figure 5 shows the deconvolution for NiMoP (0.5). Figure Figure 6 showed that the Raman ratio increases until a maximum at BDPHP/Ni = 0.5, and then decreases at BDPHP/Ni = 1.0. This result suggests the re-dispersion of Mo oxide provoked by BDPHP with different BDPHP/Ni molar ratio. A similar behavior was seen by Diaz de Leon [38] in an analogous system with additives. Regarding this, Diaz de Leon suggested that the increase of the Raman band intensities ratio promoted better dispersion of the metal species and is responsible for the highest HDS activity.

Thermal Analysis (TGA/DTA)
To study the thermal stability of the catalysts with BDPHP, TGA analysis of NiMoP (0.5) and NiMoP (1.0) samples were carried out and the results are shown in Figures 7 and 8. It can be seen that the TGA curves show a continuous weight loss of 87.3 and 85.8 wt % from 30 to 700 °C for CoMoP (0.5) and CoMoP (1.0), respectively. The weight loss for all catalysts around 12.0 wt % lower than 200 °C is associated with the removal of physisorbed H2O [39], and the further weight loss around 75.0 wt % is attributed to the loss of crystallization H2O and the stepwise loss of the organic additive [40].
The heat flow profiles for NiMoP (0.5) and NiMoP (1.0) samples ( Figure 8) exhibited a broad endothermic peak of around 100-200 °C related with dehydration reactions [39]. The exothermic peaks centered in the 430-460 °C range on the heat flow profiles are associated with the decomposition of BDPHP (decarboxylation reactions) for NiMoP (0.5) and NiMoP (1.0) samples [40]. The exothermic peak maximum decreases in the order of NiMoP (0.5) > NiMoP (1.0). It is supposed that a higher decomposition temperature corresponds to major interaction between the support and the organic additive. In contrast, it is suggested that the NiMoP (0.5) catalyst presented higher interaction.

Thermal Analysis (TGA/DTA)
To study the thermal stability of the catalysts with BDPHP, TGA analysis of NiMoP (0.5) and NiMoP (1.0) samples were carried out and the results are shown in Figures 7 and 8. It can be seen that the TGA curves show a continuous weight loss of 87.3 and 85.8 wt % from 30 to 700 °C for CoMoP (0.5) and CoMoP (1.0), respectively. The weight loss for all catalysts around 12.0 wt % lower than 200 °C is associated with the removal of physisorbed H2O [39], and the further weight loss around 75.0 wt % is attributed to the loss of crystallization H2O and the stepwise loss of the organic additive [40].
The heat flow profiles for NiMoP (0.5) and NiMoP (1.0) samples ( Figure 8) exhibited a broad endothermic peak of around 100-200 °C related with dehydration reactions [39]. The exothermic peaks centered in the 430-460 °C range on the heat flow profiles are associated with the decomposition of BDPHP (decarboxylation reactions) for NiMoP (0.5) and NiMoP (1.0) samples [40]. The exothermic peak maximum decreases in the order of NiMoP (0.5) > NiMoP (1.0). It is supposed that a higher decomposition temperature corresponds to major interaction between the support and the organic additive. In contrast, it is suggested that the NiMoP (0.5) catalyst presented higher interaction.   [39], and the further weight loss around 75.0 wt % is attributed to the loss of crystallization H 2 O and the stepwise loss of the organic additive [40].
The heat flow profiles for NiMoP (0.5) and NiMoP (1.0) samples ( Figure 8) exhibited a broad endothermic peak of around 100-200 • C related with dehydration reactions [39]. The exothermic peaks centered in the 430-460 • C range on the heat flow profiles are associated with the decomposition of BDPHP (decarboxylation reactions) for NiMoP (0.5) and NiMoP (1.0) samples [40]. The exothermic peak maximum decreases in the order of NiMoP (0.5) > NiMoP (1.0). It is supposed that a higher decomposition temperature corresponds to major interaction between the support and the organic additive. In contrast, it is suggested that the NiMoP (0.5) catalyst presented higher interaction.

Temperature Programmed Reduction (TPR)
The influence of the organic additive on the nickel-molybdenum-support interaction was investigated by temperature-programmed reduction of NiMoP/γ-Al 2 O 3 . The TPR profiles of NiMoP (0.0), NiMoP (0.5) and NiMoP (1.0) are shown in Figure 9. From Figure 9, the NiMoP (0.0) sample displayed two reduction peaks at 540 • C and 725 • C, which is associated with the nickel-molybdenum and the molybdenum reduction, respectively. NiMoP (0.5) presents two reduction peaks at 400 • C and 555 • C, corresponding to the decomposition/reduction of nickel-molybdenum and molybdenum precursor in weak and strong interactions with the γ-Al 2 O 3 support, respectively [41][42][43][44]. Additionally, a shoulder can be observed at 222 • C, which corresponds to the reduction of nickel in a weak interaction with γ-Al 2 O 3 [43]. The reduction profiles of NiMoP (1.0) exhibit two peaks: one at 410 • C ascribed to the reduction of nickel-molybdenum species, and the second one at about 586 • C due to the reduction of molybdenum species [41][42][43][44]. This catalyst also displays a peak at 271 • C corresponding to the nickel reduction [43].
The increase in the Ni and Mo reduction temperature due to the presence of higher concentration of BDPHP on the NiMoP (1.0) catalyst suggests the interaction between BDPHP and Mo-Ni species in NiMoP/γ-Al 2 O 3 . This effect was observed previously for an analogous system synthesized in the presence of EDTA [8]. Moreover, the lower temperature reduction of Ni and Mo species for NiMoP (0.5) can be rationalized in terms of better interaction of metal-support in comparison with the NiMoP (1.0) and NiMoP (0.0), and, likewise, the easier reduction of the NiMoP (0.5), compared with the NiMoP (0.0) sample, which is attributed to the presence of the organic additive at lower concentration favoring the nucleation step. In this line, the easier reduction improves the sulfidation degree of Mo and Ni species [45].
Molecules 2017, 22, 1332 9 of 17 displayed two reduction peaks at 540 °C and 725 °C, which is associated with the nickel-molybdenum and the molybdenum reduction, respectively. NiMoP (0.5) presents two reduction peaks at 400 °C and 555 °C, corresponding to the decomposition/reduction of nickel-molybdenum and molybdenum precursor in weak and strong interactions with the γ-Al2O3 support, respectively [41][42][43][44]. Additionally, a shoulder can be observed at 222 °C, which corresponds to the reduction of nickel in a weak interaction with γ-Al2O3 [43]. The reduction profiles of NiMoP (1.0) exhibit two peaks: one at 410 °C ascribed to the reduction of nickel-molybdenum species, and the second one at about 586 °C due to the reduction of molybdenum species [41][42][43][44]. This catalyst also displays a peak at 271 °C corresponding to the nickel reduction [43]. The increase in the Ni and Mo reduction temperature due to the presence of higher concentration of BDPHP on the NiMoP (1.0) catalyst suggests the interaction between BDPHP and Mo-Ni species in NiMoP/γ-Al2O3. This effect was observed previously for an analogous system synthesized in the presence of EDTA [8]. Moreover, the lower temperature reduction of Ni and Mo species for NiMoP (0.5) can be rationalized in terms of better interaction of metal-support in comparison with the NiMoP (1.0) and NiMoP (0.0), and, likewise, the easier reduction of the NiMoP (0.5), compared with the NiMoP (0.0) sample, which is attributed to the presence of the organic additive at lower concentration favoring the nucleation step. In this line, the easier reduction improves the sulfidation degree of Mo and Ni species [45].

X-ray Photoelectron Spectroscopy of Sulfided NiMoP/γ-Al2O3 Catalyst (XPS)
To monitor the evolution of the sulfide surface species, XPS was studied in the Mo 3d and Ni 2p levels. As an example, the XPS spectra of Mo 3d emission line and fit decomposition of NiMoP (0.5) and NiMoP (1.0) samples are presented in Figure 10. The Mo 3d core-level spectra of all catalysts were found to be rather complex, suggesting the presence of at least two species. They are fitted satisfactorily with two sets of doublets, each one containing the Mo 3d5/2 and Mo 3d3/2 components coming from the spin-orbit splitting. The observation of two doublets indicates that there are two different Mo-species. The Mo 3d5/2 peak for the first set, at 228.92 eV, is assigned to MoS2 (Mo 4+ ) species, and the other at 232.88 eV is attributed to Mo 6+ species [46][47][48][49] (Table 2). The NiMoP reference catalyst without organic additive exhibited 70% of MoS2 species. For the NiMoP (0.5) catalyst, 86% of Mo species was converted to MoS2 species while the NiMoP (1.0) showed 87%. The XPS results suggest that the organic additive increases the sulfidation degree in the NiMoP/γ-Al2O3 system.

X-ray Photoelectron Spectroscopy of Sulfided NiMoP/γ-Al 2 O 3 Catalyst (XPS)
To monitor the evolution of the sulfide surface species, XPS was studied in the Mo 3d and Ni 2p levels. As an example, the XPS spectra of Mo 3d emission line and fit decomposition of NiMoP (0.5) and NiMoP (1.0) samples are presented in Figure 10. The Mo 3d core-level spectra of all catalysts were found to be rather complex, suggesting the presence of at least two species. They are fitted satisfactorily with two sets of doublets, each one containing the Mo 3d 5/2 and Mo 3d 3/2 components coming from the spin-orbit splitting. The observation of two doublets indicates that there are two different Mo-species. The Mo 3d 5/2 peak for the first set, at 228.92 eV, is assigned to MoS 2 (Mo 4+ ) species, and the other at 232.88 eV is attributed to Mo 6+ species [46][47][48][49] (Table 2). The NiMoP reference catalyst without organic additive exhibited 70% of MoS 2 species. For the NiMoP (0.5) catalyst, 86% of Mo species was converted to MoS 2 species while the NiMoP (1.0) showed 87%. The XPS results suggest that the organic additive increases the sulfidation degree in the NiMoP/γ-Al 2 O 3 system.  The XPS spectra of the Ni 2p emission line for all fresh sulfided catalysts are presented in Figure 11. As can be seen from this figure, spectra show primary satellite peaks due to shake-up electrons. Since the line-shape of the Ni 2p3/2 peak, and especially its satellite, was unsymmetrical and broad, we applied an empirical method to roughly fit the curve using Gaussian/Lorentzian distributions. An example of the decomposition of Ni 2p3/2 profile made for the NiMoP sample is shown in the inlet of Figure 11. Each Ni 2p3/2 profile was resolved in three components at 863.8, 857.0 and 860.5 eV, corresponding to highly dispersed NiS, NiMoS and NiOx phases, respectively [46][47][48][49]. Apparently, the electronegativity of the organic additive has little effect on the binding energies separating satellite and main peaks of the Ni 2p3/2 spectra. The BE at 853.1 eV is close to the value reported in the literature for bulk NiS on unsupported MoS2 (852.9 eV) [38], whereas the BE at 854.4 eV is close to the value reported for NiMoS species formed after sulfidation of the NiMo/γ-Al2O3 catalysts at 673 K [46][47][48][49].
Considering the binding energy position and the relative area related to the three Ni species, the percentages of each species were determined. As seen in Table 2, the reference catalyst displayed 49% of NiMoS species [31]. For the catalysts with organic additives, the NiMoP (0.5) sample exhibited 45% of the NiMoS species, while the NiMoP (1.0) with a higher concentration of BDPHP showed increased percentages of this species to 46%. The Ni and Mo species surface exposure and the catalyst sulfidation degree could be deduced from Table 3 showing (Ni + Mo)/Al and S/(Ni + Mo) atomic ratios, respectively. The samples prepared with BDPHP/Ni = 0.5 exhibit a lower Mo and Ni species surface exposure than the BDPHP/Ni = 1.0, as inferred from the comparison of the (Ni + W)/Al atomic ratio (Table 3). However, the catalysts with BDPHP showed more metal species over the surface in comparison with the reference catalyst. Considering the TPR results (vide supra), this might indicate that catalyst sulfidation at 290 °C led to the migration of the metal species from the support surface to its inner porous structure. Finally, there is a significant variation in the sulfidation degree of the Mo species, as deduced from the comparison of the S/(Mo + Ni) atomic ratios ( Table 3). The NiMoP (0.5) exhibited 30% more sulfidation degree than NiMoP (1.0) and NiMoP (0.0).  The XPS spectra of the Ni 2p emission line for all fresh sulfided catalysts are presented in Figure 11. As can be seen from this figure, spectra show primary satellite peaks due to shake-up electrons. Since the line-shape of the Ni 2p 3/2 peak, and especially its satellite, was unsymmetrical and broad, we applied an empirical method to roughly fit the curve using Gaussian/Lorentzian distributions. An example of the decomposition of Ni 2p 3/2 profile made for the NiMoP sample is shown in the inlet of Figure 11. Each Ni 2p 3/2 profile was resolved in three components at 863.8, 857.0 and 860.5 eV, corresponding to highly dispersed NiS, NiMoS and NiOx phases, respectively [46][47][48][49]. Apparently, the electronegativity of the organic additive has little effect on the binding energies separating satellite and main peaks of the Ni 2p 3/2 spectra. The BE at 853.1 eV is close to the value reported in the literature for bulk NiS on unsupported MoS 2 (852.9 eV) [38], whereas the BE at 854.4 eV is close to the value reported for NiMoS species formed after sulfidation of the NiMo/γ-Al 2 O 3 catalysts at 673 K [46][47][48][49].
Considering the binding energy position and the relative area related to the three Ni species, the percentages of each species were determined. As seen in Table 2, the reference catalyst displayed 49% of NiMoS species [31]. For the catalysts with organic additives, the NiMoP (0.5) sample exhibited 45% of the NiMoS species, while the NiMoP (1.0) with a higher concentration of BDPHP showed increased percentages of this species to 46%. The Ni and Mo species surface exposure and the catalyst sulfidation degree could be deduced from Table 3 showing (Ni + Mo)/Al and S/(Ni + Mo) atomic ratios, respectively. The samples prepared with BDPHP/Ni = 0.5 exhibit a lower Mo and Ni species surface exposure than the BDPHP/Ni = 1.0, as inferred from the comparison of the (Ni + W)/Al atomic ratio (Table 3). However, the catalysts with BDPHP showed more metal species over the surface in comparison with the reference catalyst. Considering the TPR results (vide supra), this might indicate that catalyst sulfidation at 290 • C led to the migration of the metal species from the support surface to its inner porous structure. Finally, there is a significant variation in the sulfidation degree of the Mo species, as deduced from the comparison of the S/(Mo + Ni) atomic ratios ( Table 3). The NiMoP (0.5) exhibited 30% more sulfidation degree than NiMoP (1.0) and NiMoP (0.0).  Table 3. Mo and Ni surface atomic ratios and sulfidation degree of NiMoP/γ-Al2O3 sulfide catalysts.

Catalytic Activity Measurements
The activity of the sulfided catalysts was evaluated in the HDS of straight-run gas oil and the reaction was carried out in a fixed bed reactor (T = 340 • C and 56 kg/cm 2 of pressure). A conventional NiMoP/γ-Al 2 O 3 catalyst with similar metal content (12.0, 3.0 and 1.6 wt % for Mo, Ni and P, respectively) was used as reference. The straight-run gas oil conversion and the reaction rate constants are listed in Table 4. The catalyst prepared with BDPHP at lower concentration (BDPHP/Ni = 0.5) exhibits a higher straight-run gas oil conversion than the NiMoP (0.0) and NiMoP (1.0). Thus, to clarify the effect of organic additive, the reaction rate constants (at steady state) should be compared with their Raman ratios where the HDS activity and the Raman ratio exhibits a relationship ( Table 4). The most active catalyst NiMoP (0.5) exhibits the highest Raman ratio value. When BDPHP/Ni ratio increased at 1.0, the activity decreased. At the same time, the lowering of the Raman ratio occurs. Interestingly, the reaction rate constant of NiMoP (0.5) increased 40% compared with the NiMoP (0.0) and NiMoP (1.0) samples. It is clear that the variations in the catalytic activity are related to the presence of the Mo=O species. However, BDPHP/Ni ratio conditions should be controlled, since the best activity was achieved for the sample prepared at BDPHP/Ni = 0.5.

Discussion
The NiMoP (0.5) catalyst was more active in the HDS of straight-run gas oil than NiMoP (1.0) and NiMoP (0.0). It was demonstrated that the final characteristic of the dried and sulfided NiMoP/γ-Al 2 O 3 samples depends strongly on the BDPHP/Ni molar ratio; with 0.5 being the optimal molar ratio. At this ratio, there is a chemical equilibrium between the BDPHP located in the interface region and -OH groups of the alumina support. The role of the interface region on the adsorption of an organic additive such as BDPHP is extensively discussed in the review by Bourikas et al. [53,54]; therefore, it will be not discussed in this work.
The N 2 -phys study showed that the NiMoP (1.0) presented more surface area than the NiMoP (0.5) and NiMoP (0.0). On the other hand, XRD showed peaks corresponding to Mo species. In addition, NiMoP (1.0) exhibited two additional peaks at 39 • C and 58 • C corresponding to Mo species, indicating that the NiMoP (1.0) presented more Mo species with higher size in comparison with NiMoP (0.5), which has more dispersion than NiMoP (1.0).
The FT-IR results displayed that the BDPHP did not interact with Mo or Ni species. However, we do not discard an interaction between BDPHP and superficial groups from γ-Al 2 O 3 . Furthermore, the increment of the coke signal was seen for NiMoP (1.0). In this sense, the coke affects the MoS 2 species for HDS reaction. On the other hand, the Raman spectroscopy suggests an increment of the amount of superficial Mo=O for NiMoP (0.5) than NiMoP (1.0) and NiMoP (0.0). This result suggests that the NiMoP (0.5) displays higher oxide dispersion than its NiMoP (0.0) counterpart. This finding is in agreement with the XRD and XPS results. Likewise, this increment of MoS 2 dispersion seen by XPS is related to the isolating effect proposed by Li [19]. Accordingly, during the sulfurization process (exothermic reaction [55]), the carbonaceous deposits may isolate the Mo species from the γ-Al 2 O 3 and favored the formation of MoS 2 on the surface.
The TPR profiles showed that the temperature reduction of Ni-Mo species for NiMoP (0.5) shifted to low temperature in contrast with NiMoP (1.0) and NiMoP (0.0). These results suggest a better interaction of metal-support and easier reduction for NiMoP (0.5). This fact could suggest that the addition of BDPHP has an effect on the Ni and Mo sulfidation. In this line, the XPS results displayed an increment of 30% in the sulfidation degree for NiMoP (0.5), in comparison with NiMoP (1.0) and NiMoP (0.0). Consequently, an increment of the Mo sulfidation improves the active phase formation (MoS 2 ). Moreover, the content of coke was higher for NiMoP (1.0) in comparison with NiMoP (0.5). This result is in accordance with FT-IR measurement for NiMoP after HDS reaction.
From the TPR and XPS characterization, a relation was found between the S/(Mo + Ni) and Ni-Mo temperature reduction for NiMoP (0.0), NiMoP (0.5) and NiMoP0 (1.0) catalysts ( Figure 13) in addition to the reaction rate constant and Raman ratio ( Figure 14). The sulfidation degree increases with a decrease in the temperature reduction of Ni and Mo species. This effect is related to a decrease of the metal-support interaction [45]. On the other hand, the catalytic activity increases with the MoOx species as Mo=O. These results are in agreement with the fact that Mo=O species are more available for sulfidation process than the Mo-O and Mo-O-Mo species [38]. The latter species are close to the support; therefore, these are difficult to sulfide. Additionally, the highest activity of the NiMoP (0.5) catalyst with respect to NiMoP (1.0) and NiMoP (0.0) samples are linked to its moderate metal-support interaction, higher sulfidation degree, lower content of coke, and higher amount of Mo=O species. (1.0) and NiMoP (0.0). Consequently, an increment of the Mo sulfidation improves the active phase formation (MoS2). Moreover, the content of coke was higher for NiMoP (1.0) in comparison with NiMoP (0.5). This result is in accordance with FT-IR measurement for NiMoP after HDS reaction. From the TPR and XPS characterization, a relation was found between the S/(Mo + Ni) and Ni-Mo temperature reduction for NiMoP (0.0), NiMoP (0.5) and NiMoP0 (1.0) catalysts ( Figure 13) in addition to the reaction rate constant and Raman ratio ( Figure 14). The sulfidation degree increases with a decrease in the temperature reduction of Ni and Mo species. This effect is related to a decrease of the metal-support interaction [45]. On the other hand, the catalytic activity increases with the MoOx species as Mo=O. These results are in agreement with the fact that Mo=O species are more available for sulfidation process than the Mo-O and Mo-O-Mo species [38]. The latter species are close to the support; therefore, these are difficult to sulfide. Additionally, the highest activity of the NiMoP (0.5) catalyst with respect to NiMoP (1.0) and NiMoP (0.0) samples are linked to its moderate metalsupport interaction, higher sulfidation degree, lower content of coke, and higher amount of Mo=O species.   (1.0) and NiMoP (0.0). Consequently, an increment of the Mo sulfidation improves the active phase formation (MoS2). Moreover, the content of coke was higher for NiMoP (1.0) in comparison with NiMoP (0.5). This result is in accordance with FT-IR measurement for NiMoP after HDS reaction. From the TPR and XPS characterization, a relation was found between the S/(Mo + Ni) and Ni-Mo temperature reduction for NiMoP (0.0), NiMoP (0.5) and NiMoP0 (1.0) catalysts ( Figure 13) in addition to the reaction rate constant and Raman ratio ( Figure 14). The sulfidation degree increases with a decrease in the temperature reduction of Ni and Mo species. This effect is related to a decrease of the metal-support interaction [45]. On the other hand, the catalytic activity increases with the MoOx species as Mo=O. These results are in agreement with the fact that Mo=O species are more available for sulfidation process than the Mo-O and Mo-O-Mo species [38]. The latter species are close to the support; therefore, these are difficult to sulfide. Additionally, the highest activity of the NiMoP (0.5) catalyst with respect to NiMoP (1.0) and NiMoP (0.0) samples are linked to its moderate metalsupport interaction, higher sulfidation degree, lower content of coke, and higher amount of Mo=O species.

Conclusions
In summary, a significant increase in the HDS activity of NiMoP/γ-Al 2 O 3 catalysts was found when the support was impregnated with BDPHP organic additive at BDPHP/Ni = 0.5. The addition of BDPHP covered the Al 2 O 3 surface in order to isolate the metals and the support. The increase of HDS activity of the catalysts with organic additive was associated with the formation of Mo=O species, enhancement of the MoO 3 dispersion, moderate metal-support interaction and an increment of Ni and Mo sulfurization degree.