Study on Mesoporous Al-SBA-15 with Enhanced Acidity and Hydrothermal Stability for Heavy Oil Hydrocracking Conversion

To increase heavy oil conversion by hydrocracking, the larger pore opening mesoporous materials with enhanced acidity and hydrothermal stability are required. A series of mesoporous Al-SBA-15 was studied in acid-free and different HCl concentration solutions with various initial SiO2/Al2O3 molar ratio. The final product samples were characterized with N2 adsorption, SAXS/XRD, NH3-TPD, TEM, Si NMR, and Al NMR. The results concluded that the acid-free and weak acidic medium favor the formation of the Al-SBA-15 with better textual properties, higher acidity, and higher wall-thickness. A considerable amount of Al was present in the tetrahedral form.


Introduction
Hydrocracking process is one of the most important processes in modern refineries, due to its high versatility in processing different feedstocks, good product quality, and high liquid yields, especially at times when crude oil properties are getting deteriorated yearly and more environmentally friendly products are demanded. Hydrocracking catalysts play a crucial role in the process, because the catalysts determinate product selectivity, hydrogen consumption, run length, and economics.
Hydrocracking catalysts are bi-functional: cracking function contributed by the acidic sites such as amorphous oxides (silica-alumina) and/or zeolites, and hydrogenation function by the metals such as Co-Mo, Ni-Mo, or Ni-W. Before 1960, the amorphous silica alumina catalysts were widely used. After that, via intense researches and studies, the zeolites (mainly zeolite Y and beta) were found to have much higher acidity, higher thermal/hydrothermal stability, and showed significant improvements in activity, ammonia tolerance, and higher gasoline selectivity. Until now, zeolite Y or zeolite beta is still widely used cracking component for the commercialized catalysts.
Zeolite Y and beta with a 12-membered ring have maximum pore opening 7-8Å which is smaller than the sizes of the most molecules in heavy oils. To improve the large molecule diffusion inside the zeolites, great efforts were made to develop the mesoporous materials [1][2][3][4] that can be applied in heavy oil hydroprocessing. Although the mesoporous materials such as widely studied SBA-15 and M41S family have high surface area, large pore volume, and uniform hexagonal array of cylindrical mesopores [5], their lower acidity and worse thermal and hydrothermal stability compared with zeolites become big challenges in hydroprocessing application [6].
Among all the mesoporous materials studied (such as MCM-41, MCM-48, MCM-50 and SBA-15 etc.), SBA-15 has the thickest wall, and thus has higher thermal and hydrothermal stability than other mesoporous materials [7]. However, the stability is not higher enough to be used in the hydrocracking catalysts. Meanwhile, its acidity is similar to that of amorphous silica-alumina, and too low to hydroprocess most of heavy oils. In the early stage, only pure silicon SBA-15 was synthesized and studied in the strong acid medium. The highly acidic synthesis medium hindered the direct incorporation of metal ions, such as trivalent aluminum, into the silica framework. In order to increase the acidity of the materials, many modifications were made, such as aluminum being introduced into pure silica SBA-15, or post treatment of SBA-15 [8][9][10][11][12][13]. W. Hua and his co-workers reported an acidity enhancement method by SO 4 2-/ZrO2 modification. Zirconium nitrate was impregnated onto SBA-15, after drying and precalcination, the precalcinated samples were treated with 0.5M of H 2 SO 4 . Although the acidity was enhanced, the surface areas and pore volumes were greatly reduced from 683m 2 /g to <500m 2 /g and 0.98ml/g to <0.70ml/g respectively. C. Han et al. treated SBA-15 with diluted solution of aluminosilicate sol-gel, which caused a huge loss in pore volumes and surface areas (reduce half). Campos and his co-workers impregnated SBA-15 with TPABr and NH 4 F to transform the amorphous wall of the mesoporous SAB-15. In acid media, the SBA-15 synthesized contained only traces of aluminum, despite the fairly high aluminum content in the original gel. Meantime, surface areas were reduced from 699 m2/g to <380m2/g. Other methods included post synthesis alumination of mesoporous silica SBA-15 using ammonium hexafluoroaluminate, coating of alumina on the wall of SBA-15 by ammonia/water vapor induced internal hydrolysis, impregnating SBA-15 with aluminum nitrate, and then by vapor hydrolysis, and multiple grafting of Al isopropoxide in organic solvent on SBA-15. With all these methods, the aluminum species were coated on the wall of the meso SBA-15. They could only enhance the acidity to a limited extent, but led to huge loss in pore volumes and surface areas.
To simultaneously improve the acidity and stability, the best and efficient approach is to convert the amorphous silica-alumina wall to crystallized silica-alumina (zeolite type) during the synthesis, at least incorporate as large amounts of Al atoms as possible in a tetrahedral environment. Many attempts were made to achieve the goal. Wang and his co-workers reported a highly ordered SBA-15 with thick pore wall and high hydrothermal stability, which was synthesized by adding PVA powder [14]. Studies showed that direct synthesis under acid-free or very mild acid conditions significantly enhanced the acidity and hydrothermal stability of the mesoporous materials, and thus more ordered wall structure may be formed [15][16][17]. Because different researches used various chemicals and synthesis conditions, it is hard to compare the effect of the initial acid concentration and silica-to-alumina ratio on the main properties of the synthesis Al-SBA-15. Therefore, in this communication, under the same starting silicon and aluminum sources, the effect of initial acid concentration and silica/alumina ratio was investigated.

Synthesis of Al-SBA-15
The mesoporous materials were prepared in different HCl concentrations (0-5.0M) by using different concentrations of SiO 2 /Al 2 O 3 molar ratio (5, 20, and 50). Nonionic triblock copolymer, Pluronic P123 (Aldrich), was used as the structure-directing agent (SDA). Tetraethyl orthosilicate (TEOS, Fluka) and aluminum sulfate 18-hydrate (Al 2 (SO 4 ) 3 ·18H 2 O, Fluka) were used as silica and aluminum sources, respectively. In a typical synthesis, 8g of P123 was added to 60ml of water, and stirred until a clear solution were formed. Different concentrations of HCl (0, 0.1, 0.5, 2, and 5M) were added, and kept stirring for 2 hours. Then 18 g of TEOS and different concentrations of Al 2 (SO 4 ) 3 ·18H 2 O were added respectively, and stirred vigorously at 40°C for 24h. Subsequently, the mixture was transferred to an autoclave and heated at 100°C for 2 days under static condition. The solid precipitate was recovered by filtration, washed several times with distilled water until no Clwas detected. The solid product was dried at 60°C overnight, and calcinated at 600°C for 6 hours in air (ramp 2°C /min). The samples prepared from starting gel SiO 2 /Al 2 O 3 molar ratio 20 were designed as Meso-0, Meso-0.1, Meso-0.5, Meso-2, meso-5, the number indicated the HCl concentration added.

Catalyst Characterization
The sample texture property (surface areas, pore volumes, pore sizes, and pore size distributions) was characterized by using a physisorption analyser of Micromeritics, ASAP 2420 instrument. Before adsorption, the samples were calcined at 823K for 4 hours. 30-40 mg of powder samples were degassed in a sample preparation station under 473K and 1.33×10 -3 Pa for 15 hours, then switched to the analysis station for adsorption and desorption under liquid nitrogen at 77K with an equilibrium time of 2 minutes. The surface area was calculated with the multipoint BET equation with linear region in the P/P o range of 0.05 to 0.35. Pore volume was calculated from the maximum adsorption amount of nitrogen at P/P o = 0.99. The pore size distribution was determined based on the BJH method and the desorption branch of the isotherm.
The ammonia temperature programmed desorption (NH 3 -TPD) test was conducted with Micromeritics AutoChem 2910 instrument. Prior the TPD measurements, the sample was pre-treated with helium carrier gas at 500°C for 60 min, and cooled down to 50°C. After that, the sample was exposed to 10% NH 3 in He for 30 min followed by purged with helium gas at 100°C for 60 min. to remove the physisorbed ammonia. The TPD run was then conducted from 50 to 700°C at a heating rate of 10°C/min.
Small Angle X-ray Scatting (SAXS) and XRD characterization was performed with the RIGAKU ULTIMA-IV multiple-purposed X-ray diffraction system. SAXS and XRD of the calcinated samples were recorded in the 2θ range from 0.5 o to 6 o and from 0.5 o to 70 o respectively using CuKα radiation (λ=1.5406) and proportional counter as detector. 29 Si and 27 Al MAS NMR were performed on Bruker 400 MHz with spinning rate 4000 Hz. The samples were taken in a zirconium rotor. For 29 Si, CP MAS experiment was applied.
For 27 Al, one pulse experiment was applied. The chemical shift scales were referenced against 29 Si NMR resonance of pure powder silicon at 0 ppm for 29 Si NMR, and externally against 27 Al NMR resonance of aluminum sulfate at 0 ppm for 27 Al NMR.
The pore sizes and wall thickness of the samples were determined by transmission electron microscopy (TEM) on a JEOL 2010 STEM, operated at 200 keV. A small amount of powder (0.5-1.0 g) was mixed with methanol and agitated for 1 minute, and then sonicated for 20 minutes. Several drops of the suspension were placed onto a carbon-coated 3.5mm diameter copper mesh grid before TEM characterization.
The silica and alumina content in the samples was analyzed by using VARIAN VISTA PRO ICP-OES (Inductively Coupled Plasma-Optical Emission Spectroscopy) based on ASTM D1976.

Results and Discussion
The isotherms of all product samples from nitrogen adsorption analysis showed the typical type-IV curves. The hysteresis loops indicate the presence of the mesopores. For the samples with initial SiO 2 /Al 2 O 3 molar ratio 20, the effect of acid concentration on relative surface area, pore volume, average pore size, and pore size distribution is illustrated in Figure 1 to Figure 4 respectively. The surface area reaches the maximum at 0.5M HCl and then quickly dropped. The pore volume and average pore sizes exhibit a different trend. They dramatically decreased with increased acid concentration from acid free to 0.1M HCl, then changes become slow or level off while HCl acid concentration changed from 0.5 to 5.0M. With the HCl acid concentration increased, the pore size distribution shifted to smaller pore sizes, and distribution become much wider.    The effect of the SiO 2 /Al 2 O 3 ratio in the initial gels on the textual properties is summarized in Table 1. Under the same HCl acid concentrations, the relative surface areas, pore volumes, and average pore sizes were quite similar. This means that the initial SiO 2 /Al 2 O 3 ratio did not lead to significant textual property change. The higher surface areas of the catalysts can provide more active sites. The larger pore volume and pore sizes favor the large molecules diffusion and mass transfer, and thus favor the hydroconversion of the heavier petroleum feedstocks. Therefore, from the textual property point of view, the optimized acid concentration of the mesoporous materials synthesized is very low acid environment (acid free or <0.1M).
The acidities and their distribution of Meso-0 to Meso-5 are sumamrized in Table 2 and illustated in Figure 5. To compare the acidity distribution among the samples, the weak, medium, and strong acidities were assigned to the peak areas of NH 3 -TPD curves below 623 K, at 623-773 K, and above 773 K, respectively. Meso-0.1 exhibited the highest acidity, about 7.7 times higher than that of Meso-0 and about 35 times higher than that of Meso-0.5 to Meso-5. Total acidity and acid distribution became the same when the HCl concentration was >0.5M. With the HCl concentrations increased from acid-free to 5.0M, the week acidic peak temperature was shifted to lower temperature, while the strong acid peaks were similar. From acidity distribution, it can be concluded that the majority acidity is weak and strong acidic sites. The week acidic sites were mainly contributed by alumina and amorphous silica-alumina, while the strong acidic sites were from tetra-hedral aluminum species. XRD diffraction showed typical amorphous profiles for all samples, indicating not ordered structure in the long range (the XRD spectra are not shown in the communication). The 27 Al NMR provided the evidence of the tetra-hedral aluminum species existence. From the effect of the initial SiO 2 /Al 2 O 3 molar ratio in starting gel on the SiO 2 /Al 2 O 3 molar ratio of the final products (Table 3), it can be seen that the final product SiO 2 /Al 2 O 3 ratio seemed indepent on the intial gel SiO 2 /Al 2 O 3 ratio.
The final product SiO 2 /Al 2 O 3 molar ratio was higher than 55. The higher the HCl concentration was, the higher SiO 2 /Al 2 O 3 molar ratio the final products had. This means that Al is more difficult to be assmbled to become the wall of the mesoporous material than Si. The acid concentration governed the Al amount that can be combined with silica precusors to become the miceles, and finally assembled around the structure direct agents to form mesoporous materials. At higher HCl concentration, the aluminum is more easily existed in the form of Al 3+ , and presented in the solution not in the miceles. Interestingly, the lowest final product SiO 2 /Al 2 O 3 molar ratio was achieved when the HCl concentration was at 0.1M not acid-free evironment.  Table 3 lists the SiO 2 /Al 2 O 3 molar ratio of the final products under different SiO 2 /Al 2 O 3 molar ratio in starting gel. Although the initial SiO 2 /Al 2 O 3 molar ratio in starting gel increased 10 times (from 5 to 50), the SiO 2 /Al 2 O 3 molar ratio of the final products changed less than 30%. This means that much less Al was assembled in the final mesoporous sample than Si. As expected, when the HCl concentration increase from 0.1M to 5.0M, SiO 2 /Al 2 O 3 molar ratio of the final product abruptly increased. The reprehensive TEM images of the samples are shown in Figure 6. In the image, the dark and bright strips are the walls and pores of the amorphous materials. About10-15 such images were selected. The wall thickness and pore sizes were measured. The average wall thickens and average pore sizes are illustrated in Figure 7. The wall thickness is dramatically decreased with the HCl concentration, and then level off when HCl concentration was higher than 0.5M. The trend was in good agreement with the BET result. The result was also in agreement with a study on SBA-15 silica synthesized by direct addition of PVA powder [15]. The mesoporous materials with highly ordered structure and thick wall generally showed the excellent thermal and hydrothermal stability. Therefore, from the wall thickness point of view (e.t. thermal and hydrothermal stability), Meso-0 and Meso-0.1 is better than all other samples.
To further prove the formation of the mesopores and the wall thickness, the SAXS characterization was performed and the results are shown in Figure 8. As seen in Figure 8, all Al-SBA-15 samples exhibit at least two well-resolved peaks indexed to the (100) and (200) reflections which are associated with p6mm hexagonal symmetry. This indicated the samples were of well-ordered meso-structures. With the HCl concentration increased from 0M to 5.0M, the reflection peak (100) was slightly shifted to higher angle, suggesting the unit-cell size of the final sample was decreased. The decreased unit-cell size was one of the reasons that the wall thickness of the samples became thinner when HCl concentration increased.    27 Al MAS NMR spectra of all samples are exhibited in Figure 9. In addition to the band corresponding to tetrahedral Al (the chemical shift: around 40-55 ppm), there was always a band near 0 ppm for all samples, which is attributed to octahedral Al. Appearance of a band in the 25-30 ppm region indicated the presence of penta-coordinated Al [20]. The presentence of the tetrahedral Al in all samples explained the existence of the strong acid sites.
The results from NH 3 -TPD, final product SiO 2 /Al 2 O 3 molar ratio, and 27 Al NMR indicate that the highest acidity of Meso-0.1 is closely associated with its high Al content and most of Al species are in the tetrahedral state. Both Lewis and Bronsted acid sites of the samples are related to the aluminum content in the final product samples and their states (tetrahedral or octahedral coordinated). In the absence of acid and very weak acid environment, the pH value of the P123 solution was about 7, and dropped to about 2-3.5 after dissolution of aluminum sulfate respectively, for initial gel SiO 2 /Al 2 O 3 between 5 and 50. The self-generated acidity from the hydrolysis of aluminum sulfate with the reasonable hydrolysis time was probably effective for the hydrolysis of TEOS. The pH is higher than the isoelectric point of silica (pH=2), and thus positively charged aluminum species is easily interacted with the negatively charged silica species resulting from TEOS hydrolysis. Therefore the aluminum was easily introduced. The initially formed aluminosilcate species possibly are attached to the micelles or even the aluminum cationic species to the micelles facilitating the interaction between micelles and anionic silicate species. This could contribute to the cooperative self-assembly of the organic-inorganic aggregates to form an alumimum rich silicate layer attached to the micelles. At higher acidic medium (HCl concentration >0.5M), Al species exist only in the cationic form [18]. Due to lower than the isoelectric point of silica, the silica species are also positively charged. Therefore the interaction between Al and Si species in the system was restricted, which makes their condensation with the silica species to form Si-O-Al linkages unlikely. As a result, less Al can be incorporated and assembled with silica micelles, and thus a very high SiO 2 /Al 2 O 3 molar ratio and lower acidity of the final products were achieved. Other researchers also found that, despite the increase in pH value to 1.5, only a limited amount of Al from the initial gel was incorporated into the final solid products [19][20].
Current study and previous literature proved that weak acidic and acid-free medium favor the synthesis of the Al-SBA-15 with higher acidity and hydrothermal stability that can be applied to hydrocracking and FCC catalysts. Unfortunately, the high Al content in the final product under such condition is hard to be achieved. To synthesize Al-SBA-15 with high content of Al (preferred in tetrahedral form), many methods have been investigated. Li, Jiang and his co-workers used "The hydrolysis-controlled approach" in which the hydrolysis of TEOS was accelerated by fluoride [21][22]. They also tried tetramethylorthosilicate as silicon source instead of TEOS. Lin and his workers used ''acid-free approach'' (the same as the method in this communication) in which the acidity of the synthesis solution was generated in-situ by the hydrolysis of aluminium sulfate [23]. The disadvantage of all these methods are always still low incorporation of aluminium from the synthesis gel in the final solid and the presence of extra-framework Al species. Therefore, to synthesize mesoporous materials with zeolite structured wall, more studies on proper synthesis conditions, appropriate chemicals, and the order to add them need be carried out to not only ensure the hydrolysis of TEOS, but also the enough interaction between Al and Si species in the synthesis system.

Conclusion
When using TEOS and aluminum sulfate as silica and aluminum source respectively, P123 as structure directing agent, the acid free and weak acid environment favored the enhancement of acidity, the formation of larger mesoporous (higher pore volume, pore size, and surface areas), and formation of thicker wall. A considerable amount of Al was in tetrahedral Al state, which help the further conversion to the zeolite wall. The effect of the original gel Si/Al ratio on textural properties and acidity is not significant. The Si/Al molar ratios are all above 50 when the initial Si/Al ratios changed from 5 to 50.