Hydrotreating of Light Cycle Oil over Supported on Porous Aromatic Framework Catalysts

The hydroprocessing of substituted naphthalenes and light cycle oil (LCO) over bimetallic Ni-W-S and Ni-Mo-S catalysts that were obtained by decomposition of [N(n-Bu)4]2[Ni(MeS4)2] (Me = W, Mo) complexes in situ in the pores of mesoporous aromatic frameworks (PAFs) during the reaction, was studied. The promotion of acid-catalyzed processes by PAF-AlCl3, synthesized by impregnation of a PAF with AlCl3 from its toluene solution, was investigated. It has been found that Ni-W-S catalytic systems were more active in the hydrodearomatization reactions, while Ni-Mo-S catalytic systems were more active in hydrodesulfurization and hydrocracking reactions. The introduction of sulfur into the reaction medium enhanced the activity of the catalysts and the presence of PAF-AlCl3 led to an acceleration of the hydrocracking processes.


Introduction
Light cycle oil (LCO) is a middle distillate that is produced by fluid catalytic cracking of heavy gas oil or vacuum gas oils. It distils between 120 • C and 380 • C and contains high amounts of aromatic hydrocarbons (60-90 wt %) and sulfur (up to 9000 or 25,000 ppm depending on the quality of feedstock). LCO can be used as a blending component for diesel or fuel oil and it can be used as a feedstock for the manufacturing of diesel fuel. However, the ongoing toughening of ecological standards and restrictions for the levels of sulfur and aromatics in fuels makes the use of LCO for these purposes an increasingly unattractive choice.
LCO can be upgraded using hydrocatalytic processes-hydrodearomatization, hydrodesulfurization, hydrocracking. One of the most suitable technologies for the conversion of heavy oils is slurry-phase hydrocracking. This process includes the conversion of large, bulky molecules into smaller ones at the active sites of the catalyst. Therefore, it is crucial to have a system of macro-and meso-scale pores in the catalyst to allow for the mass transport of heavy feedstock to the active sites of catalyst to maximize its activity. Conventional catalysts for various hydroprocesses are mainly based on microstructured inorganic materials, such as zeolites and alumosilicates, in which pore sizes hamper the diffusion of large molecules into the material [1,2]. That is why mesostructured hierarchical materials are considered as a perspective base for the development of new generation catalysts.
In the last decade, various strategies for the synthesis of numerous mesoporous materials, both inorganic, organic and of a hybrid nature, were developed [3][4][5][6][7][8]. Inorganic supports are the most common ones due to the presence of acid sites in their structure that are necessary for the

Characterization of Precursors and Catalysts
Thermal decomposition of [N(n-Bu) 4 ] 2 [Ni(MeS 4 ) 2 ] (Me = W, Mo) within the pores of porous aromatic frameworks PAF (Figure 1) occurred over a wide temperature range in several steps. The probable decomposition pathway, according to Reference [20], with an amendment for the presence of nickel, can be represented as [N(n-Bu) 4 ] 2 [Ni(MeS 4 ) 2 ] = NiMe 2 S 2−n C x N y H z + 2 N(n-Bu) 3 + Bu 2 S 2 + . . . (1) The coefficients n, x, y and z depend on the chemical composition of the salt. As for [N(n-Bu) 4 ] 2 [WS 4 ], the decomposition product could be presented as WS 2 C 1.28 N 0.07 H 1. 14 [20], and for [N(n-Bu) 4 4 ] the composition of the product is MoS 2.03 C 1.13 N 0.04 H 0.8 [21].

] 2 [MoS
The first two peaks of the differential thermal analysis and thermogravimetry (DTA-TG) curves can probably be related to the decomposition of the ammonium cation by the formation of N(n-Bu) 3 . It should be noted that there was a lag between the changing of the mass of the samples and the following decomposition processes. Thus, a mass decrease of approximately 16-19%, which was attributed to the formation of N(n-Bu) 3 , occurred right after the end of the exothermic peaks at 230-235 • C. This could be because of the diffusion limitations during the evaporation of the gaseous products of the decomposition as the [N(n-Bu) 4 ] 2 [Ni(MeS 4 ) 2 ] salts are within the pores of the PAF. The endothermic peaks at 395 and 325 • C were attributed to the structural rearrangements within the products of the initial decomposition and/or the formation of a NiW 2 S 5−n C x N y H z phase. Further decomposition peaks, above 400 • C, were attributed to the decomposition of the support material [22].
High-resolution transmission electron microphotographs (HRTEM) of particles of the PAF-NiWS catalyst, after the reaction of LCO hydroconversion, are seen in Figure 2. Such lamellar structures of particles, that consist of multiple layers, are typical for sulfide catalysts. Analogously to the previous work [14], the length, L, and the number, N, of three-dimensional stacked layers were measured ( Figure 3). The stacking number for the PAF-NiWS catalyst varied from three to nine with a maximum of four to five slabs. The slab length varied from 3 nm to 19 nm with a maximum of 7 nm. of nickel, can be represented as [N(n-Bu)4]2[Ni(MeS4)2] = NiMe2S2−nCxNyHz + 2 N(n-Bu)3 + Bu2S2 + … The coefficients n, x, y and z depend on the chemical composition of the salt. As for [N(n-Bu)4]2[WS4], the decomposition product could be presented as WS2C1.28N0.07H1.14 [20], and for [N(n-Bu)4]2[MoS4] the composition of the product is MoS2.03C1.13N0.04H0.8 [21].   The first two peaks of the differential thermal analysis and thermogravimetry (DTA-TG) curves can probably be related to the decomposition of the ammonium cation by the formation of N(n-Bu)3. It should be noted that there was a lag between the changing of the mass of the samples and the following decomposition processes. Thus, a mass decrease of approximately 16-19%, which was attributed to the formation of N(n-Bu)3, occurred right after the end of the exothermic peaks at 230-235 °C. This could be because of the diffusion limitations during the evaporation of the gaseous products of the decomposition as the [N(n-Bu)4]2[Ni(MeS4)2] salts are within the pores of the PAF. The endothermic peaks at 395 and 325 °C were attributed to the structural rearrangements within the products of the initial decomposition and/or the formation of a NiW2S5−nCxNyHz phase. Further decomposition peaks, above 400 °C, were attributed to the decomposition of the support material [22].
High-resolution transmission electron microphotographs (HRTEM) of particles of the PAF-NiWS catalyst, after the reaction of LCO hydroconversion, are seen in Figure 2. Such lamellar structures of particles, that consist of multiple layers, are typical for sulfide catalysts. Analogously to the previous work [14], the length, L, and the number, N, of three-dimensional stacked layers were measured ( Figure 3). The stacking number for the PAF-NiWS catalyst varied from three to nine with a maximum of four to five slabs. The slab length varied from 3 nm to 19 nm with a maximum of 7 nm. (a) (b)   The first two peaks of the differential thermal analysis and thermogravimetry (DTA-TG) curves can probably be related to the decomposition of the ammonium cation by the formation of N(n-Bu)3. It should be noted that there was a lag between the changing of the mass of the samples and the following decomposition processes. Thus, a mass decrease of approximately 16-19%, which was attributed to the formation of N(n-Bu)3, occurred right after the end of the exothermic peaks at 230-235 °C. This could be because of the diffusion limitations during the evaporation of the gaseous products of the decomposition as the [N(n-Bu)4]2[Ni(MeS4)2] salts are within the pores of the PAF. The endothermic peaks at 395 and 325 °C were attributed to the structural rearrangements within the products of the initial decomposition and/or the formation of a NiW2S5−nCxNyHz phase. Further decomposition peaks, above 400 °C, were attributed to the decomposition of the support material [22].
High-resolution transmission electron microphotographs (HRTEM) of particles of the PAF-NiWS catalyst, after the reaction of LCO hydroconversion, are seen in Figure 2. Such lamellar structures of particles, that consist of multiple layers, are typical for sulfide catalysts. Analogously to the previous work [14], the length, L, and the number, N, of three-dimensional stacked layers were measured ( Figure 3). The stacking number for the PAF-NiWS catalyst varied from three to nine with a maximum of four to five slabs. The slab length varied from 3 nm to 19 nm with a maximum of 7 nm.     The average dimensions and stacking numbers of the sulfide particles for current and previous work are listed below in Table 1, where N is the number of slabs per stack, D-MeS 2 dispersion, and f e /f c is the edge-to-corner ratio of a sulfide slab. Calculations were performed using the previously listed equations [14,15]. The chemical composition of the surface of the catalysts was evaluated by means of X-ray photoelectron spectroscopy (XPS) analysis. The XPS spectra of the samples contained peaks for tungsten, molybdenum, nickel, sulfur, carbon, nitrogen and oxygen. The decomposition of W4f and Mo3d spectra showed the presence of three forms of metal on the surface of the catalyst: MeS 2 , MeS x O y and Me 6+ . The weight ratios and binding energies of the resultant phases are listed in Table 2. The obtained values for the binding energies were compared with those reported in the literature [23,24]. Recently, we described the synthesis of a PAF-AlCl 3 catalyst and its use in the hydrogenation-hydrocracking of naphthalene [19]. It is well known that Lewis acids, particularly aluminum chloride, react with aromatic substances and form complexes of the type PhH σ+ -(AlCl 3 ) σ− through charge-transfer mechanisms [25][26][27]. We showed that this mechanism was realized also in the case of the interaction between AlCl 3 and PAF. Thus, PAF-AlCl 3 , likewise for the PAF-SO 3 H [28][29][30] material, has blue color-appearing when electron-withdrawing groups interact with the polyaromatic skeleton of PAF. The FTIR spectrum of PAF-AlCl 3 (Figure 4), similarly for the spectrum of PS-AlCl 3 material in Reference [26], contains absorption bands with a maximum at 1580 cm −1 , proving the formation of charge-transfer structures. In addition, the solid-state 27 Al NMR spectrum ( Figure 5

Catalytic Experiments with Model Substrates
To investigate the properties of PAF-NiMeS/PAF-AlCl3 catalytic systems for the hydroconversion of substituted naphthalenes, experiments with 1-, 2-methylnaphthalenes, 2ethylnaphthalene and 1,3-, 1,5-dimethylnaphthalenes were performed. These substances were chosen as LCO mainly contains different alkyl-derivatives of naphthalene and tetraline, and we assume that general patterns of activity for PAF-NiMeS/PAF-AlCl3 systems in hydrodearomatization, hydrocracking, isomerization and dealkylation reactions for LCO will be the same, as for model substrates. The aim of this part of the study was to elucidate the peculiarities of the conversion of molecules with different numbers of alkyl groups, their length and position. The simplified mechanism for the conversion of naphthalene and methylnaphtalenes into the reaction products is presented in Figure 6.

Catalytic Experiments with Model Substrates
To investigate the properties of PAF-NiMeS/PAF-AlCl3 catalytic systems for the hydroconversion of substituted naphthalenes, experiments with 1-, 2-methylnaphthalenes, 2ethylnaphthalene and 1,3-, 1,5-dimethylnaphthalenes were performed. These substances were chosen as LCO mainly contains different alkyl-derivatives of naphthalene and tetraline, and we assume that general patterns of activity for PAF-NiMeS/PAF-AlCl3 systems in hydrodearomatization, hydrocracking, isomerization and dealkylation reactions for LCO will be the same, as for model substrates. The aim of this part of the study was to elucidate the peculiarities of the conversion of molecules with different numbers of alkyl groups, their length and position. The simplified mechanism for the conversion of naphthalene and methylnaphtalenes into the reaction products is presented in Figure 6.

Catalytic Experiments with Model Substrates
To investigate the properties of PAF-NiMeS/PAF-AlCl 3 catalytic systems for the hydroconversion of substituted naphthalenes, experiments with 1-, 2-methylnaphthalenes, 2-ethylnaphthalene and 1,3-, 1,5-dimethylnaphthalenes were performed. These substances were chosen as LCO mainly contains different alkyl-derivatives of naphthalene and tetraline, and we assume that general patterns of activity for PAF-NiMeS/PAF-AlCl 3 systems in hydrodearomatization, hydrocracking, isomerization and dealkylation reactions for LCO will be the same, as for model substrates. The aim of this part of the study was to elucidate the peculiarities of the conversion of molecules with different numbers of alkyl groups, their length and position. The simplified mechanism for the conversion of naphthalene and methylnaphtalenes into the reaction products is presented in Figure 6.  All reactions can be formally divided into four types: Hydrogenation reactions, also called hydrodearomatization reactions; isomerization reactions; hydrocracking reactions, which result in the destruction of C-C bonds in the carbon structure of molecule and the formation of new C-H bonds; and dealkylation reactions, which result in the destruction of C-C bonds between carbon atoms of the benzene ring, and the alkyl chain, and removal of the alkyl substituent from the structure of the molecule. Dealkylation reactions always result in a decrease in the number of carbon atoms in the molecule structure, while cracking reactions do not. For each molecule, the probability of entry into a particular type of the reaction depends on its structure, concentration, the activity of the catalytic system in this type of reaction, and other factors. It should be noted that the scheme in Figure  6 represents the main processes that occur during the hydrotreating of methylnaphthalenes, but does not include intramolecular condensation and transalkylation, coke formation and other side reactions.
The results of hydrotreating of 1-and 2-methylnaphthalenes over the PAF-NiWS catalyst are presented in Figures 7 and 8. 2-Methylnaphthalene underwent conversion to methyltetralines and methyldecalines-hydrogenation reaction products. The introduction of both sulfur and PAF-AlCl3 dramatically accelerated the rate of the dealkylation process and, to a lesser extent, the rate of the hydrocracking processes. In the case of experiments with 1-methylnaphthalene, the yields of products of isomerization, dealkylation and hydrocracking reactions were higher than in the case of 2-methylnaphthalene. The introduction of PAF-AlCl3 accelerated the isomerization of 1methylnapthalene into 2-methylnaphthalene. All reactions can be formally divided into four types: Hydrogenation reactions, also called hydrodearomatization reactions; isomerization reactions; hydrocracking reactions, which result in the destruction of C-C bonds in the carbon structure of molecule and the formation of new C-H bonds; and dealkylation reactions, which result in the destruction of C-C bonds between carbon atoms of the benzene ring, and the alkyl chain, and removal of the alkyl substituent from the structure of the molecule. Dealkylation reactions always result in a decrease in the number of carbon atoms in the molecule structure, while cracking reactions do not. For each molecule, the probability of entry into a particular type of the reaction depends on its structure, concentration, the activity of the catalytic system in this type of reaction, and other factors. It should be noted that the scheme in Figure 6 represents the main processes that occur during the hydrotreating of methylnaphthalenes, but does not include intramolecular condensation and transalkylation, coke formation and other side reactions.
The results of hydrotreating of 1-and 2-methylnaphthalenes over the PAF-NiWS catalyst are presented in Figures 7 and 8. 2-Methylnaphthalene underwent conversion to methyltetralines and methyldecalines-hydrogenation reaction products. The introduction of both sulfur and PAF-AlCl 3 dramatically accelerated the rate of the dealkylation process and, to a lesser extent, the rate of the hydrocracking processes. In the case of experiments with 1-methylnaphthalene, the yields of products of isomerization, dealkylation and hydrocracking reactions were higher than in the case of 2-methylnaphthalene. The introduction of PAF-AlCl 3 accelerated the isomerization of 1-methylnapthalene into 2-methylnaphthalene.  The catalytic system, consisting of PAF-NiWS, PAF-AlCl3, and sulfur, was tested for the conversion of 1,3-dimethylnaphthalene, 1,5-dimethylnaphthalene and 2-ethylnaphthalene ( Figure 9). Notably, the yields of dealkylation and hydrocracking products grew by increasing of the number of alkyl groups in the structure of naphthalenes and their length. At the same time, sterically more hindered substrates were more likely to undergo dealkylation and hydrocracking reactions. In this regard, it could be concluded that when PAF-AlCl3 is present in the catalyst system, sterically hindered substrates are more prone to dealkylation and hydrocracking reactions compared to hydrogenation reactions.  The catalytic system, consisting of PAF-NiWS, PAF-AlCl3, and sulfur, was tested for the conversion of 1,3-dimethylnaphthalene, 1,5-dimethylnaphthalene and 2-ethylnaphthalene ( Figure 9). Notably, the yields of dealkylation and hydrocracking products grew by increasing of the number of alkyl groups in the structure of naphthalenes and their length. At the same time, sterically more hindered substrates were more likely to undergo dealkylation and hydrocracking reactions. In this regard, it could be concluded that when PAF-AlCl3 is present in the catalyst system, sterically hindered substrates are more prone to dealkylation and hydrocracking reactions compared to hydrogenation reactions. The catalytic system, consisting of PAF-NiWS, PAF-AlCl 3 , and sulfur, was tested for the conversion of 1,3-dimethylnaphthalene, 1,5-dimethylnaphthalene and 2-ethylnaphthalene ( Figure 9). Notably, the yields of dealkylation and hydrocracking products grew by increasing of the number of alkyl groups in the structure of naphthalenes and their length. At the same time, sterically more hindered substrates were more likely to undergo dealkylation and hydrocracking reactions. In this regard, it could be concluded that when PAF-AlCl 3 is present in the catalyst system, sterically hindered substrates are more prone to dealkylation and hydrocracking reactions compared to hydrogenation reactions.

Catalytic Experiments with Light Cycle Oil (LCO)
The use of the catalytic system PAF-NiMeS/PAF-AlCl3 could be most promising for hydrotreating LCO from the position of both hydrodearomatization, hydrocracking, hydroisomerization and hydrodesulphurization reactions. LCO contains a large number of aromatic compounds, mostly diaromatic (35-55 wt %) and monoaromatic (10-35 wt %) hydrocarbons, and to a lesser extent polyaromatic hydrocarbons (5-10 wt %) with three or more benzene rings in the structure [34]; the predominant diaromatic hydrocarbons are represented basically by different alkylsubstituted naphthalenes. The source of sulfur in LCO is various benzothiophenes and dibenzothiophenes, as well as disulfides in trace amounts. Depending on the fractional distillation temperature range and the feedstock and features of the manufacturing process, the content of the components in LCO could vary. LCO, which was used in the current work, is characterized by a high content of aromatic compounds (approximately 60 wt %). However, the distribution of mono-and diaromatic hydrocarbons was shifted towards the former. The sulfur content in the used feedstock was 0.35 wt %, or 3500 ppm, which is common for LCO.
The results of the catalytic hydrotreating of LCO in the presence of PAF-NiWS and PAF-NiMoS catalysts are shown in the Figures 10-12. The best results for the dearomatization process were obtained using the PAF-NiWS catalyst. Particularly, the application of the catalytic system PAF-NiWS + PAF-AlCl3 + S led to the highest yield of saturated compounds and the lowest content of diand polyaromatic hydrocarbons. Remarkably, the reaction with PAF-NiMoS without PAF-AlCl3 did not lead to a notable decrease in the quantity of biaromatic and polyaromatic hydrocarbons. However, since the fractional composition of LCO shifted towards light hydrocarbons, this result may indicate the dominance of the reactions of dealkylation and cracking of alkyl chains over hydrogenation reactions. When PAF-AlCl3 was used as a promoter with the PAF-NiMoS catalyst, the yields of monoaromatic and saturated hydrocarbons increased, confirming the cracking activity of PAF-AlCl3. Thus, we can conclude that W-based catalysts are more active in hydrogenation/hydrodearomatization processes, while Mo-based catalysts are more active in hydrocracking and dealkylation processes. The obtained results are consistent with reported data for sulfide catalysts [35][36][37], according to which tungsten catalysts are more active in hydrodearomatization reactions compared to molybdenum ones.

Catalytic Experiments with Light Cycle Oil (LCO)
The use of the catalytic system PAF-NiMeS/PAF-AlCl 3 could be most promising for hydrotreating LCO from the position of both hydrodearomatization, hydrocracking, hydroisomerization and hydrodesulphurization reactions. LCO contains a large number of aromatic compounds, mostly diaromatic (35-55 wt %) and monoaromatic (10-35 wt %) hydrocarbons, and to a lesser extent polyaromatic hydrocarbons (5-10 wt %) with three or more benzene rings in the structure [34]; the predominant diaromatic hydrocarbons are represented basically by different alkyl-substituted naphthalenes. The source of sulfur in LCO is various benzothiophenes and dibenzothiophenes, as well as disulfides in trace amounts. Depending on the fractional distillation temperature range and the feedstock and features of the manufacturing process, the content of the components in LCO could vary. LCO, which was used in the current work, is characterized by a high content of aromatic compounds (approximately 60 wt %). However, the distribution of mono-and diaromatic hydrocarbons was shifted towards the former. The sulfur content in the used feedstock was 0.35 wt %, or 3500 ppm, which is common for LCO.
The results of the catalytic hydrotreating of LCO in the presence of PAF-NiWS and PAF-NiMoS catalysts are shown in the Figures 10-12. The best results for the dearomatization process were obtained using the PAF-NiWS catalyst. Particularly, the application of the catalytic system PAF-NiWS + PAF-AlCl 3 + S led to the highest yield of saturated compounds and the lowest content of di-and polyaromatic hydrocarbons. Remarkably, the reaction with PAF-NiMoS without PAF-AlCl 3 did not lead to a notable decrease in the quantity of biaromatic and polyaromatic hydrocarbons. However, since the fractional composition of LCO shifted towards light hydrocarbons, this result may indicate the dominance of the reactions of dealkylation and cracking of alkyl chains over hydrogenation reactions. When PAF-AlCl 3 was used as a promoter with the PAF-NiMoS catalyst, the yields of monoaromatic and saturated hydrocarbons increased, confirming the cracking activity of PAF-AlCl 3 . Thus, we can conclude that W-based catalysts are more active in hydrogenation/hydrodearomatization processes, while Mo-based catalysts are more active in hydrocracking and dealkylation processes. The obtained results are consistent with reported data for sulfide catalysts [35][36][37], according to which tungsten catalysts are more active in hydrodearomatization reactions compared to molybdenum ones.  Fractional composition of LCO after hyrdoconvertion with PAF-NiMeS catalytic systems is shown in Figure 12. The use of the PAF-NiWS catalyst, as well as of PAF-NiMoS catalyst, led to the displacement of the fractional composition of LCO towards lighter products. However, tungsten catalysts were more effective for the conversion of "heavy" fractions, especially of fractions that boil above 300 °C. The addition of sulfur did not lead to a significant change in the distribution of fractions, confirming its role as a promoter in hydrogenation reactions. In contrast, when the cracking-promoting additive PAF-AlCl3 was added to the reaction, more light hydrocarbons were formed. Combining data of the fractional composition of hydroconverted LCO and the distribution of aromatic hydrocarbons in it, it can be concluded that in the presence of tungsten catalysts, the reaction proceeds by hydrogenation of aromatic substances followed by their hydrocracking. At the  Fractional composition of LCO after hyrdoconvertion with PAF-NiMeS catalytic systems is shown in Figure 12. The use of the PAF-NiWS catalyst, as well as of PAF-NiMoS catalyst, led to the displacement of the fractional composition of LCO towards lighter products. However, tungsten catalysts were more effective for the conversion of "heavy" fractions, especially of fractions that boil above 300 °C. The addition of sulfur did not lead to a significant change in the distribution of fractions, confirming its role as a promoter in hydrogenation reactions. In contrast, when the cracking-promoting additive PAF-AlCl3 was added to the reaction, more light hydrocarbons were formed. Combining data of the fractional composition of hydroconverted LCO and the distribution of aromatic hydrocarbons in it, it can be concluded that in the presence of tungsten catalysts, the reaction proceeds by hydrogenation of aromatic substances followed by their hydrocracking. At the Fractional composition of LCO after hyrdoconvertion with PAF-NiMeS catalytic systems is shown in Figure 12. The use of the PAF-NiWS catalyst, as well as of PAF-NiMoS catalyst, led to the displacement of the fractional composition of LCO towards lighter products. However, tungsten catalysts were more effective for the conversion of "heavy" fractions, especially of fractions that boil above 300 • C. The addition of sulfur did not lead to a significant change in the distribution of fractions, confirming its role as a promoter in hydrogenation reactions. In contrast, when the cracking-promoting additive PAF-AlCl 3 was added to the reaction, more light hydrocarbons were formed. Combining data of the fractional composition of hydroconverted LCO and the distribution of aromatic hydrocarbons in it, it can be concluded that in the presence of tungsten catalysts, the reaction proceeds by hydrogenation of aromatic substances followed by their hydrocracking. At the same time, in presence of molybdenum catalysts, the reaction proceeded through hydrocracking and dealkylation mechanisms, and PAF-AlCl 3 promoted the hydrocracking of heavy aromatics.  In addition, we determined the sulfur content in LCO after the reactions to assess the hydrodesulfurization activity of the catalysts (Figure 13). Hydrotreating over both tungsten and molybdenum catalysts allowed us to achieve a high level of desulfurization, reducing the sulfur concentration up to 10-20% from the initial content. The best results were obtained for the molybdenum catalyst PAF-NiMoS, for which the extent of desulfurization reached up to 90%. The use of tungsten catalysts made it possible to decrease the sulfur content to a level of 630-690 ppm, which could be explained by the lower activity of tungsten catalysts in hydrodesulfurization reactions compared to molybdenum catalysts [35,38].  In addition, we determined the sulfur content in LCO after the reactions to assess the hydrodesulfurization activity of the catalysts (Figure 13). Hydrotreating over both tungsten and molybdenum catalysts allowed us to achieve a high level of desulfurization, reducing the sulfur concentration up to 10-20% from the initial content. The best results were obtained for the molybdenum catalyst PAF-NiMoS, for which the extent of desulfurization reached up to 90%. The use of tungsten catalysts made it possible to decrease the sulfur content to a level of 630-690 ppm, which could be explained by the lower activity of tungsten catalysts in hydrodesulfurization reactions compared to molybdenum catalysts [35,38].  In addition, we determined the sulfur content in LCO after the reactions to assess the hydrodesulfurization activity of the catalysts (Figure 13). Hydrotreating over both tungsten and molybdenum catalysts allowed us to achieve a high level of desulfurization, reducing the sulfur concentration up to 10-20% from the initial content. The best results were obtained for the molybdenum catalyst PAF-NiMoS, for which the extent of desulfurization reached up to 90%. The use of tungsten catalysts made it possible to decrease the sulfur content to a level of 630-690 ppm, which could be explained by the lower activity of tungsten catalysts in hydrodesulfurization reactions compared to molybdenum catalysts [35,38]. Figure 13. Sulfur content in LCO before and after the hydroconversion over the PAF-NiMeS catalyst. Figure 13. Sulfur content in LCO before and after the hydroconversion over the PAF-NiMeS catalyst.

Materials
All starting materials were purchased from commercial suppliers and used as received. The purification of solvents was performed according to standard methods. Porous aromatic frameworks were synthesized as described earlier [9]. LCO was received from Gazprom Nephtekhim Salavat, Ltd. Characteristics of LCO are listed in Table 3.

Synthesis of Precursors and Catalysts
The ammonium thiomolybdate and thiotungstate were prepared as described elsewhere [10]. The tetraalkylammonium nickel tetrathyometallate salts and PAF-NiMeS catalysts were obtained using the previously described method [6]. Both the tungsten and molybdenum content in the catalysts was 15 wt %.
Synthesis of the PAF-AlCl 3 catalyst was performed by impregnation of PAF with AlCl 3 from its hot toluene solution. The typical procedure includes dissolving of 2 g of AlCl 3 in 200 mL of toluene at 100-105 • C, cooling of the solution to 80 • C and the addition of 1 g of PAF. Then, the mixture was stirred for 4 h, allowing it to cool to room temperature. After the removal of toluene, under reduced pressure, a dark-blue well-dispersed powder was obtained.

Characterization of Catalysts
DTA-TG measurements were conducted using the Paulik-Erdey MOM Q-1500D derivatograph (MOM, Budapest, Hungary) with the use of alumina crucibles under an air atmosphere within the temperature interval of 30-1000 • C and a heating rate of 10 • C/min.
The solid-state NMR 27 Al spectra were recorded using a Brucker AVANCE-II 400 (Bruker UK Ltd., Coventry, UK) instrument with a magic angle spinning (MAS) at a resonance frequency of 104.23 MHz. The spinning speed was 12 kHz, 1 M aqueous solution of Al(NO 3 ) 3 was used as a standard.
The textural properties of the catalysts were determined from HRTEM microphotographs, obtained on a JEOL JEM-2100/Cs/GIF microscope (JEOL, Tokyo, Japan) with a 0.19 nm lattice fringe resolution and an accelerating voltage of 200 kV. The samples were prepared on a perforated carbon film, mounted on a copper grid, and 10-15 representative micrographs were obtained for each catalyst in high-resolution mode. Typically, the lengths of at least 400 slabs were measured for each catalyst.
The XPS analysis of the catalysts was performed on a VersaProbeII, ULVAC-PHI (ULVAC-PHI, Inc., Kanagawa, Japan) instrument using excitation with Al Kα X-ray radiation at 1486.6 eV. The calibration of the photoelectron peaks was based on the C 1s line with a binding energy of 286.7 eV. The transmission energy of the energy analyzer was 160 eV (a survey spectrum) or 40 eV (individual lines).

Catalytic Testing Procedure
The reactions were carried out in a steel reactor, equipped with a stir bar, at 380 • C under a 5 MPa H 2 pressure. Typically, the stir bar, the PAF-NiMeS catalyst (200 mg), LCO (2 mL) or 30% solution of model substrates in undecane (2 mL), sulfur (200 mg, in several experiments), and PAF-AlCl 3 (100 mg, in several experiments) were introduced into the reactor. Then the reactor was sealed and pressurized with H 2 at the demanded pressure. The mixture was stirred at 380 • C and, after the reaction was finished, the reactor was cooled down and depressurized. The reaction products were analyzed by GC, HPLC and X-ray fluorescence analysis.

X-ray Fluorescence Analysis
The determination of the sulfur content of the reaction products was performed on Specroscan S equipment, using standard methods [40].

High Performance Liquid Chromatography
The content and distribution of aromatic hydrocarbons were determined using "GOST R EN 12916-2008: Petroleum products. Determination of the types of aromatic hydrocarbons in the middle distillates. The method of high-performance liquid chromatography with detection by the refractive index" standard. The analysis was performed using Knauer apparatus (Knauer, Berlin, Germany) with a SmartLine 2300 refractometric detector and a diasphere-amine column at a temperature of 20 • C in accordance with GOST R EN 12916-08 (ASTM 3591-06) standard [41].

Conclusions
The hydroconversion of various naphthalenes and LCO using the catalysts PAF-NiMeS (Me = Mo, W) with the addition of PAF-AlCl 3 and sulfur was studied. It was found that the structure of the substrate molecule strongly influenced the mechanism of its transformation to the reaction products. Obviously, increasing the length of the alkyl chain increased the probability of the cracking reactions, and increasing of the number of the alkyl substituents increased the probability of the dealkylation reactions. However, the position of the substituent in the alkylnaphthalene molecule was an important factor. The probability of dealkylation and isomerization processes was higher when the substituent was in the α-position, whereas β-substituted naphthalenes underwent mainly hydrogenation.
We have also shown that the addition of PAF-AlCl 3 to the catalytic system promoted various acid-catalyzed processes. Moreover, when PAF-AlCl 3 was present in the catalyst system, sterically hindered substrates were more prone to dealkylation and hydrocracking reactions than hydrogenation reactions.
Based on the results of LCO hydrotreating, we can conclude that the PAF-NiWS catalyst was more active in hydrogenation/hydrodearomatization processes, while the PAF-NiMoS catalyst was more active in hydrocracking, dealkylation and desulfurization processes. In the case of the PAF-NiWS catalyst, the reaction followed through the hydrogenation of aromatic substances and further dealkylation, cracking and isomerization of hydrogenated products. In the case of the PAF-NiMoS catalyst, hydrocracking of alkyl chains in the substrate molecules prevailed.
Author Contributions: E.K. and A.M. conceived and designed the experiments; M.V. performed the experiments; Y.K. contributed reagents/materials/analysis tools; L.K. wrote the paper.
Funding: This research was funded by the Russian Science Foundation, grant number 15-19-00099.