Oligomerization of Butene Mixture over NiO/Mesoporous Aluminosilicate Catalyst

: This study is aimed at preparing C 8 –C 16 alkene through oligomerization of a butene mixture using nickel oxide supported on mesoporous aluminosilicate. Mesoporous aluminosilicate with an ordered structure was successfully synthesized from HZSM-5 zeolite by combining a top-down and a bottom-up method. MMZ ZSM-5 catalyst showed much higher butene conversion and C 8 –C 16 yield in the butene oligomerization reaction than those with HZSM-5. This is attributed to the pore geometry of MMZ ZSM-5 , which is more beneﬁcial for internal diffusion of reactants, reaction intermediates, and products. The ordered channel-like mesopores were maintained after the nickel-loading on MMZ ZSM-5 . The yield for C 8 –C 16 hydrocarbons over NiO/MMZ ZSM-5 was higher than that of MMZ ZSM-5 catalyst, which seemed to be due to higher acid strength from a higher ratio of Lewis acid to Brønsted acid. The present study reveals that a mesoporous NiO/MMZ ZSM-5 catalyst with a large amount of Lewis acid sites is one of the potential catalysts for efﬁcient generation of aviation fuel through the butene oligomerization.


Introduction
The greenhouse gas (GHG) emitted when an aircraft is in operation has negative impacts on the environment, including global warming [1]. To resolve such problems, the International Civil Aviation Organization announced a policy intended to reduce GHG emission by aircraft using biomass-derived fuels for aircraft [2]. Consequently, research about synthesis of aviation fuel from biomass is drawing attention. The Alcohol to Jet (ATJ) process for making alcohol from biomass as feedstock and for making aviation fuel from alcohol has recently attracted attention [3]. This process involves the production of butene through the dehydration of butanol, the synthesis of higher olefins by oligomerization of the newly produced butene, and hydrogenation of the higher olefins [4][5][6]. Studies on the synthesis of butenes through butanol dehydration in the first step are already under way [7][8][9][10][11]. We have previously reported that silica-based, solid acid catalysts with mesopores and highly dispersed metal oxide catalysts are effective in the production of butene by butanol dehydration [12,13].
Because the demand for aviation fuel is increasing steadily, maximizing the production of aviation fuel is of great interest to oil refining companies [14]. Because butene is an excess in the refineries (FCC), the use of oligomerization to convert butene into aviation fuel is a promising method for the generation of aviation fuel.
We have been studying the application of mesoporous aluminosilicate with zeolite framework for catalytic reaction. In particular, we have published research results in which a framework was constructed using commercial zeolite. From this, mesoporous aluminosilicate with a well-ordered structure was produced by combining a top-down method and a bottom-up method [13,27].
The purpose of this study was to prepare a particular fraction (C8-C16; corresponding to aviation fuel) through oligomerization of butene using nickel oxide supported on mesoporous aluminosilicate with a well-ordered structure. We used a butene mixture having the same composition as the product obtained from dehydration of butanol as raw material for the oligomerization. High-pressure oligomerization of the butene mixture was investigated in a fixed-bed continuous-flow reactor.

Textural Properties of Catalysts
The N2 adsorption-desorption isotherm of MMZZSM-5 ( Figure 1) showed a type IV(a) physisorption isotherm in the IUPAC classification [28,29]. Furthermore, the N2 adsorptiondesorption isotherm of MMZZSM-5 exhibited the H4 type of hysteresis loop, which is typical of mesoporous zeolites and micro-mesoporous materials [29,30]. In the case of NiO/MMZZSM-5 catalyst, the characteristics of N2 physisorption isotherm were not much different from the N2 adsorptiondesorption isotherm of MMZZSM-5. This indicates that the geometry of the ordered mesopores of MMZZSM-5 was not changed by NiO loading.
The pore size distribution curves of the three catalysts were compared in Figure 2. As expected from the nitrogen adsorption isotherm, it was difficult to observe the mesopores of HZSM-5, while a mesopore peak of about 3.0 nm could be observed on MMZZSM-5. Moreover, the pore size somewhat decreased after the nickel loading of 9 wt %. However, the overall characteristic pore-size distribution curve of the material did not change after NiO loading.
The BET surface areas of MMZZSM-5 and HZSM-5 were 745 and 448 m 2 /g, respectively (Table 1). Furthermore, the total pore volumes of MMZZSM-5 and HZSM-5 were 0.47 and 0.19 cm 3 /g, respectively. Therefore, it was clear that the BET surface area and pore size of MMZZSM-5 catalyst were much larger than those of HZSM-5 catalyst. Meanwhile, the BET surface area and pore volume of MMZZSM-5 decreased slightly after the NiO loading, which was due to the fact that the nickel oxide was attached to the pore surface and partially blocked a pore mouth.  Low-angle XRD patterns of the MMZZSM-5 and NiO/MMZZSM-5 catalysts are presented in Figure 3. MMZZSM-5 has three diffraction peaks corresponding to 100, 110, and 200 Bragg reflections, respectively. In particular, the peak at 2θ of 2.3° is the characteristic peak of the 2-D hexagonal structure, and the two weak peaks at 3.9° and 4.5° show the inherent characteristics of the mesoporous material [27]. The NiO/MMZZSM-5 sample represents the characteristic peaks of MMZZSM-5 with three intense peaks at 2θ = 2.3°, 3.9°, and 4.5°, respectively. After nickel loading, the XRD patterns of the material showed a slight reduction in peak intensity but maintained the same patterns as those of MMZZSM-5. This suggests that the ordered mesoporous structures of the MMZZSM-5 are well retained after the nickel loading. The XRD patterns of MMZZSM-5 and NiO/MMZZSM-5 material in the high-angle region ( Figure 4) give a broad line centered at 2θ ≈ 23°, which indicates that the framework of MMZZSM-5 was not atomically ordered and the starting source, HZSM-5, was not present in the MMZZSM-5 material.
A TEM image is presented in Figure 5 to determine whether MMZZSM-5 and NiO/MMZZSM-5 have hexagonal structures. MMZZSM-5 has a well-ordered structure with mesopores that are hexagonal 1-D channels [31]. The NiO/MMZZSM-5 sample maintained the structural features of MMZZSM-5 after the nickel loading.  Low-angle XRD patterns of the MMZ ZSM-5 and NiO/MMZ ZSM-5 catalysts are presented in Figure 3. MMZ ZSM-5 has three diffraction peaks corresponding to 100, 110, and 200 Bragg reflections, respectively. In particular, the peak at 2θ of 2.3 • is the characteristic peak of the 2-D hexagonal structure, and the two weak peaks at 3.9 • and 4.5 • show the inherent characteristics of the mesoporous material [27]. The NiO/MMZ ZSM-5 sample represents the characteristic peaks of MMZ ZSM-5 with three intense peaks at 2θ = 2.3 • , 3.9 • , and 4.5 • , respectively. After nickel loading, the XRD patterns of the material showed a slight reduction in peak intensity but maintained the same patterns as those of MMZ ZSM-5 . This suggests that the ordered mesoporous structures of the MMZ ZSM-5 are well retained after the nickel loading. The XRD patterns of MMZ ZSM-5 and NiO/MMZ ZSM-5 material in the high-angle region ( Figure 4) give a broad line centered at 2θ ≈ 23 • , which indicates that the framework of MMZ ZSM-5 was not atomically ordered and the starting source, HZSM-5, was not present in the MMZ ZSM-5 material.
A TEM image is presented in Figure 5 to determine whether MMZ ZSM-5 and NiO/MMZ ZSM-5 have hexagonal structures. MMZ ZSM-5 has a well-ordered structure with mesopores that are hexagonal 1-D channels [31]. The NiO/MMZ ZSM-5 sample maintained the structural features of MMZ ZSM-5 after the nickel loading.

Acid Properties of Catalysts
The acidic characteristics of HZSM-5, MMZ ZSM-5 , and NiO/MMZ ZSM-5 were analyzed using NH 3 -TPD ( Figure 6). HZSM-5 exhibits two peaks at the desorption temperature of ammonia near 180 and 380 • C, indicating that weak strength acid sites and medium strength acid sites are widely distributed. The NH 3 -TPD curve of MMZ ZSM-5 shows that the peak intensity is greatly reduced compared to HZSM-5. This indicates that MMZ ZSM-5 has both the weak acid sites near 170 • C and medium strength acid near 340 • C, but the total amount of acid sites over MMZ ZSM-5 is greatly reduced compared to HZSM-5. The reason could be that aluminum is relatively less exposed to the amorphous framework that constructs the pore walls of MMZ ZSM-5 . On the other hand, in the case of the MMZ ZSM-5 catalyst, the peak temperature of NH 3 -TPD shifted to a lower temperature than that of HZSM-5, which means that the acid strength was also weakened.
For the NiO/MMZ ZSM-5 catalyst, the desorption peak at a low temperature shifted to high temperature (210 • C), indicating that the acid strength of NiO/MMZ ZSM-5 is higher than that of MMZ ZSM-5 catalyst. It is noticeable that an additional high-temperature (540 • C) peak appeared on the NH 3 -TPD profile of the NiO/MMZ ZSM-5 catalyst, which means that nickel oxide generates a new type of acid site. This high-temperature peak is known to be attributable to Lewis acid sites from nickel oxide [32].

Acid Properties of Catalysts
The acidic characteristics of HZSM-5, MMZZSM-5, and NiO/MMZZSM-5 were analyzed using NH3-TPD ( Figure 6). HZSM-5 exhibits two peaks at the desorption temperature of ammonia near 180 and 380 °C, indicating that weak strength acid sites and medium strength acid sites are widely distributed. The NH3-TPD curve of MMZZSM-5 shows that the peak intensity is greatly reduced compared to HZSM-5. This indicates that MMZZSM-5 has both the weak acid sites near 170 °C and medium strength acid near 340 °C, but the total amount of acid sites over MMZZSM-5 is greatly reduced compared to HZSM-5. The reason could be that aluminum is relatively less exposed to the amorphous framework that constructs the pore walls of MMZZSM-5. On the other hand, in the case of the MMZZSM-5 catalyst, the peak temperature of NH3-TPD shifted to a lower temperature than that of HZSM-5, which means that the acid strength was also weakened.
For the NiO/MMZZSM-5 catalyst, the desorption peak at a low temperature shifted to high temperature (210 °C), indicating that the acid strength of NiO/MMZZSM-5 is higher than that of MMZZSM-5 catalyst. It is noticeable that an additional high-temperature (540 °C) peak appeared on the NH3-TPD profile of the NiO/MMZZSM-5 catalyst, which means that nickel oxide generates a new type of acid site. This high-temperature peak is known to be attributable to Lewis acid sites from nickel oxide [32].  Figure 7 shows the results of the pyridine-FTIR for the NiO/MMZZSM-5 sample. As the desorption temperature increased to 300 °C, the band intensity at 1596 cm −1 diminished drastically and disappeared at temperatures higher than 150 °C, indicating a very weak H acid site [33,34]. Four wellresolved bands can be observed at (1450, 1491, 1542, and 1610) cm −1 at higher than 150 °C. Two of these bands are attributed to Lewis sites (1450 and 1610 cm −1 ), while the band at 1542 cm −1 corresponds to the Brönsted sites [33,34]. Additional band near 1491 cm −1 is associated with a Lewis site and Brönsted site simultaneously.
The pyridine-FTIR results of HZSM-5, MMZZSM-5, and NiO/MMZZSM-5 catalysts are shown in Figure 8. HZSM-5 is rich in Brönsted sites (1542 cm −1 ) and also has Lewis sites (1450 cm −1 , 1610 cm −1 ). In MMZZSM-5 also, both Brönsted and Lewis acidity can be observed. It should also be noted that for the NiO/MMZZSM-5 catalyst, the band corresponding to the Lewis acid site (1450 cm −1 ) was much larger  Figure 7 shows the results of the pyridine-FTIR for the NiO/MMZ ZSM-5 sample. As the desorption temperature increased to 300 • C, the band intensity at 1596 cm −1 diminished drastically and disappeared at temperatures higher than 150 • C, indicating a very weak H acid site [33,34]. Four well-resolved bands can be observed at (1450, 1491, 1542, and 1610) cm −1 at higher than 150 • C. Two of these bands are attributed to Lewis sites (1450 and 1610 cm −1 ), while the band at 1542 cm −1 corresponds to the Brönsted sites [33,34]. Additional band near 1491 cm −1 is associated with a Lewis site and Brönsted site simultaneously.
The pyridine-FTIR results of HZSM-5, MMZ ZSM-5 , and NiO/MMZ ZSM-5 catalysts are shown in Figure 8. HZSM-5 is rich in Brönsted sites (1542 cm −1 ) and also has Lewis sites (1450 cm −1 , 1610 cm −1 ). In MMZ ZSM-5 also, both Brönsted and Lewis acidity can be observed. It should also be noted that for the NiO/MMZ ZSM-5 catalyst, the band corresponding to the Lewis acid site (1450 cm −1 ) was much larger than that associated with the Brönsted site (1542 cm −1 ). That is, the NiO loading shows that the Lewis site is relatively increased. The pyridine-FTIR and NH 3 -TPD results suggest that a new acid site is generated by loading of NiO over MMZ  , which is attributable to the Lewis site. than that associated with the Brönsted site (1542 cm −1 ). That is, the NiO loading shows that the Lewis site is relatively increased. The pyridine-FTIR and NH3-TPD results suggest that a new acid site is generated by loading of NiO over MMZZSM-5, which is attributable to the Lewis site.

Oligomerization of Mixed-Butene
The HZSM-5 catalyst showed good stability for butene oligomerization (Figure 9). In previous studies on trimerization of iso-butene, it was proposed that the catalytic performance of solid acid catalysts and zeolites could be altered by their pore geometry [21,22]. The most common finding is that medium-pore molecular sieves, such as zeolite MFI and zeolite beta, were shown to be suitable for oligomerization of iso-butene or n-butene into diesel and gasoline fuels due to their shape- than that associated with the Brönsted site (1542 cm −1 ). That is, the NiO loading shows that the Lewis site is relatively increased. The pyridine-FTIR and NH3-TPD results suggest that a new acid site is generated by loading of NiO over MMZZSM-5, which is attributable to the Lewis site.

Oligomerization of Mixed-Butene
The HZSM-5 catalyst showed good stability for butene oligomerization (Figure 9). In previous studies on trimerization of iso-butene, it was proposed that the catalytic performance of solid acid catalysts and zeolites could be altered by their pore geometry [21,22]. The most common finding is that medium-pore molecular sieves, such as zeolite MFI and zeolite beta, were shown to be suitable for oligomerization of iso-butene or n-butene into diesel and gasoline fuels due to their shape-

Oligomerization of Mixed-Butene
The HZSM-5 catalyst showed good stability for butene oligomerization (Figure 9). In previous studies on trimerization of iso-butene, it was proposed that the catalytic performance of solid acid catalysts and zeolites could be altered by their pore geometry [21,22]. The most common finding is that medium-pore molecular sieves, such as zeolite MFI and zeolite beta, were shown to be suitable for oligomerization of iso-butene or n-butene into diesel and gasoline fuels due to their shape-selective properties [20,35]. Besides that, these zeolite catalysts were proved to have high coke-resistance due to limited polyaromatic generation in their pore structure. Based on similar analogies, the stability of HZSM-5 catalyst during the conversion of the butene mixture into C 8 -C 16 hydrocarbon could be due to the coke-resistance resulting from the pore geometry. selective properties [20,35]. Besides that, these zeolite catalysts were proved to have high cokeresistance due to limited polyaromatic generation in their pore structure. Based on similar analogies, the stability of HZSM-5 catalyst during the conversion of the butene mixture into C8-C16 hydrocarbon could be due to the coke-resistance resulting from the pore geometry. After the reaction had progressed for six hours, the MMZZSM-5 catalyst showed much higher butene conversion and C8-C16 yield than those with HZSM-5 (Figures 9 and 10). From the NH3-TPD and Pyridine-FTIR results, MMZZSM-5 was found to have fewer and weaker acidities than HZSM-5 catalyst. Thus, it is presumed that the reason why the MMZZSM-5 catalyst is beneficial for butene oligomerization, despite the fact that the acidity of MMZZSM-5 is inferior to HZSM-5, is its pore geometry. MMZZSM-5 has an average pore size of 2.7 nm, which is much larger than the pore size (0.5-0.6 nm) of HZSM-5. Consequently, it can be suggested that the mesoporous structure of MMZZSM-5 plays an important role in overcoming the internal pore diffusion of the molecules participating in the reaction.
Moreover, the yield for C8-C16 hydrocarbons over NiO/MMZZSM-5 was higher than that over the MMZZSM-5 (Figures 9 and 10). The pore size distributions of the two catalysts were not significantly different. On the other hand, according to the NH3-TPD and pyridine-FTIR analysis, it can be confirmed that when NiO is loaded on MMZZSM-5, the Lewis acid sites become more abundant and the acid strength becomes stronger. Therefore, the higher butene conversion over NiO/MMZZSM-5 catalysts seemed to be due to higher acid strength resulting from a higher Lewis site-to-Brønsted site ratio. Consequently, it may be summarized that mesoporous structures, as well as the high ratio of Lewis to Brønsted acid, were beneficial for production of C8-C16 hydrocarbons through butene oligomerization.
The carbon number distribution in the aviation fuel range (C8-C16 hydrocarbons) was not significantly influenced by the catalysts (Figure 11). In our previous study, it was found that dimers were the main reaction products at a low butane conversion over the zeolite catalysts [26]. In this study, because the butene conversion was high over three catalysts, the carbon number distribution in the range of C8-C16 hydrocarbon became broader, which is ascribed to the contribution of consecutive cracking and subsequent oligomerization pathways. After the reaction had progressed for six hours, the MMZ ZSM-5 catalyst showed much higher butene conversion and C 8 -C 16 yield than those with HZSM-5 (Figures 9 and 10). From the NH 3 -TPD and Pyridine-FTIR results, MMZ ZSM-5 was found to have fewer and weaker acidities than HZSM-5 catalyst. Thus, it is presumed that the reason why the MMZ ZSM-5 catalyst is beneficial for butene oligomerization, despite the fact that the acidity of MMZ ZSM-5 is inferior to HZSM-5, is its pore geometry. MMZ ZSM-5 has an average pore size of 2.7 nm, which is much larger than the pore size (0.5-0.6 nm) of HZSM-5. Consequently, it can be suggested that the mesoporous structure of MMZ ZSM-5 plays an important role in overcoming the internal pore diffusion of the molecules participating in the reaction.

Time on stream (h)
Moreover, the yield for C 8 -C 16 hydrocarbons over NiO/MMZ ZSM-5 was higher than that over the MMZ ZSM-5 (Figures 9 and 10). The pore size distributions of the two catalysts were not significantly different. On the other hand, according to the NH 3 -TPD and pyridine-FTIR analysis, it can be confirmed that when NiO is loaded on MMZ ZSM-5 , the Lewis acid sites become more abundant and the acid strength becomes stronger. Therefore, the higher butene conversion over NiO/MMZ ZSM-5 catalysts seemed to be due to higher acid strength resulting from a higher Lewis site-to-Brønsted site ratio. Consequently, it may be summarized that mesoporous structures, as well as the high ratio of Lewis to Brønsted acid, were beneficial for production of C 8 -C 16 hydrocarbons through butene oligomerization.
The carbon number distribution in the aviation fuel range (C 8 -C 16 hydrocarbons) was not significantly influenced by the catalysts (Figure 11). In our previous study, it was found that dimers were the main reaction products at a low butane conversion over the zeolite catalysts [26]. In this study, because the butene conversion was high over three catalysts, the carbon number distribution in the range of C 8 -C 16 hydrocarbon became broader, which is ascribed to the contribution of consecutive cracking and subsequent oligomerization pathways.  Catalytic stability of the NiO/MMZZSM-5 catalyst was evaluated under 1.5 MPa at 350 °C for 30 h on stream (Figure 12). The conversion of butene was more than 80% and did not decrease significantly for 18 h on stream. However, the NiO/MMZZSM-5 catalyst exhibited a decrease in butene conversion after 18 h of time-on-stream (TOS) while having no significant drop in selectivity to C8-C16 fraction with TOS. As a result, the yield of C8-C16 hydrocarbon remained above 63% for 18 h TOS, then gradually decreased to 48.0% after 30 h TOS.
To quantify the coke deposits formed on the NiO/MMZZSM-5 catalyst after reaction for 30 h, the weight contents of these deposits were determined by TGA. According to Figure 13, two ranges of weight losses can be recognized: a low-temperature region at <100 °C and a high-temperature region at >400 °C, which could be attributed to volatile organic compounds and coke deposits, respectively. Consequently, it was found that about 4 wt % of coke was formed on the NiO/MMZZSM-5 catalyst after TOS of 30 h.  Catalytic stability of the NiO/MMZZSM-5 catalyst was evaluated under 1.5 MPa at 350 °C for 30 h on stream (Figure 12). The conversion of butene was more than 80% and did not decrease significantly for 18 h on stream. However, the NiO/MMZZSM-5 catalyst exhibited a decrease in butene conversion after 18 h of time-on-stream (TOS) while having no significant drop in selectivity to C8-C16 fraction with TOS. As a result, the yield of C8-C16 hydrocarbon remained above 63% for 18 h TOS, then gradually decreased to 48.0% after 30 h TOS.
To quantify the coke deposits formed on the NiO/MMZZSM-5 catalyst after reaction for 30 h, the weight contents of these deposits were determined by TGA. According to Figure 13, two ranges of weight losses can be recognized: a low-temperature region at <100 °C and a high-temperature region at >400 °C, which could be attributed to volatile organic compounds and coke deposits, respectively. Consequently, it was found that about 4 wt % of coke was formed on the NiO/MMZZSM-5 catalyst after TOS of 30 h. Catalytic stability of the NiO/MMZ ZSM-5 catalyst was evaluated under 1.5 MPa at 350 • C for 30 h on stream (Figure 12). The conversion of butene was more than 80% and did not decrease significantly for 18 h on stream. However, the NiO/MMZ ZSM-5 catalyst exhibited a decrease in butene conversion after 18 h of time-on-stream (TOS) while having no significant drop in selectivity to C 8 -C 16 fraction with TOS. As a result, the yield of C 8 -C 16 hydrocarbon remained above 63% for 18 h TOS, then gradually decreased to 48.0% after 30 h TOS.
To quantify the coke deposits formed on the NiO/MMZ ZSM-5 catalyst after reaction for 30 h, the weight contents of these deposits were determined by TGA. According to Figure 13, two ranges of weight losses can be recognized: a low-temperature region at <100 • C and a high-temperature region at >400 • C, which could be attributed to volatile organic compounds and coke deposits, respectively. Consequently, it was found that about 4 wt % of coke was formed on the NiO/MMZ ZSM-5 catalyst after TOS of 30 h.

Preparation of Catalysts
The procedure of the catalyst preparation consists of the top-down and bottom-up approaches [13,27]. A cationic surfactant and HZSM-5 zeolite were used as the template and the framework sources, respectively. At first, HZSM-5 is rent into nano-unit in alkali solution (top-down approach). NaOH (0.375 mol) was dissolved in 76.5 mL of deionized water to prepare NaOH aqueous solution (4.9 M, 80.0 mL), and then 33.75 g of HZSM-5 (Zeolyst, SiO2/Al2O3 mole ratio = 50) was mixed into the NaOH aqueous solution and stirred for 1 h. Subsequently, this zeolitic nano-unit is assembled into ordered mesoporous material using the template (bottom-up approach). Hexadecyltrimethylammonium bromide (0.189 mol) was dissolved into 1050 mL of deionized water, and the HZSM-5 solution was added drop-by-drop. This mixture was stirred for 24 h, and the pH was controlled to 10 using CH3COOH solution (50%). After pH adjustment and mixing were repeated three times, a white precipitate was obtained. The resulting products were dried at 110 °C overnight and calcined at 550 °C for 4 h in an air atmosphere to obtain a mesoporous aluminosilicate material. Subsequently, it was ion-exchanged using NH4Cl aqueous solution (1 M) to convert to the NH4 + form and again converted to H + form by calcination in an air atmosphere at 550 °C for 3 h. The newly prepared material was named MMZZSM-5, meaning mesoporous material from ZSM-5.

Preparation of Catalysts
The procedure of the catalyst preparation consists of the top-down and bottom-up approaches [13,27]. A cationic surfactant and HZSM-5 zeolite were used as the template and the framework sources, respectively. At first, HZSM-5 is rent into nano-unit in alkali solution (top-down approach). NaOH (0.375 mol) was dissolved in 76.5 mL of deionized water to prepare NaOH aqueous solution (4.9 M, 80.0 mL), and then 33.75 g of HZSM-5 (Zeolyst, SiO2/Al2O3 mole ratio = 50) was mixed into the NaOH aqueous solution and stirred for 1 h. Subsequently, this zeolitic nano-unit is assembled into ordered mesoporous material using the template (bottom-up approach). Hexadecyltrimethylammonium bromide (0.189 mol) was dissolved into 1050 mL of deionized water, and the HZSM-5 solution was added drop-by-drop. This mixture was stirred for 24 h, and the pH was controlled to 10 using CH3COOH solution (50%). After pH adjustment and mixing were repeated three times, a white precipitate was obtained. The resulting products were dried at 110 °C overnight and calcined at 550 °C for 4 h in an air atmosphere to obtain a mesoporous aluminosilicate material. Subsequently, it was ion-exchanged using NH4Cl aqueous solution (1 M) to convert to the NH4 + form and again converted to H + form by calcination in an air atmosphere at 550 °C for 3 h. The newly prepared material was named MMZZSM-5, meaning mesoporous material from ZSM-5.

Preparation of Catalysts
The procedure of the catalyst preparation consists of the top-down and bottom-up approaches [13,27]. A cationic surfactant and HZSM-5 zeolite were used as the template and the framework sources, respectively. At first, HZSM-5 is rent into nano-unit in alkali solution (top-down approach). NaOH (0.375 mol) was dissolved in 76.5 mL of deionized water to prepare NaOH aqueous solution (4.9 M, 80.0 mL), and then 33.75 g of HZSM-5 (Zeolyst, SiO 2 /Al 2 O 3 mole ratio = 50) was mixed into the NaOH aqueous solution and stirred for 1 h. Subsequently, this zeolitic nano-unit is assembled into ordered mesoporous material using the template (bottom-up approach). Hexadecyltrimethylammonium bromide (0.189 mol) was dissolved into 1050 mL of deionized water, and the HZSM-5 solution was added drop-by-drop. This mixture was stirred for 24 h, and the pH was controlled to 10 using CH 3 COOH solution (50%). After pH adjustment and mixing were repeated three times, a white precipitate was obtained. The resulting products were dried at 110 • C overnight and calcined at 550 • C for 4 h in an air atmosphere to obtain a mesoporous aluminosilicate material. Subsequently, it was ion-exchanged using NH 4 Cl aqueous solution (1 M) to convert to the NH 4 + form and again converted to H + form by calcination in an air atmosphere at 550 • C for 3 h. The newly prepared material was named MMZ ZSM-5 , meaning mesoporous material from ZSM-5.
Nickel nitrate hexahydrate (98.0%) was impregnated on MMZ ZSM-5 catalyst by the dry impregnation method. The loading amount of nickel metal was 3 wt %. After calcination in an air atmosphere at 550 • C for 3 h, the NiO/MMZ ZSM-5 catalyst was obtained.

Characterization of Catalysts
The nitrogen physisorption isotherms of the catalysts were obtained using a BELSORP-mini II device (BEL Japan Inc., Toyonaka, Japan) at −196 • C. XRD patterns of the catalysts were collected at 40 kV and 40 mA using a D/MAX-2200V Diffractometer (Rigaku Co., The Woodlands, TX, USA), and data were reduced using the JADE program. TEM images were collected using a JEM-3010 (JEOL, Tokyo, Japan) operated at 300 kV.
The NH 3 -TPD was performed using BEL-CAT-B device (BEL Japan Inc., Toyonaka, Japan). The sample was outgassed at 550 • C in a He gas flow (50 mL/min) for an hour. The ammonia (5% in He) was adsorbed on the sample, and then the physically adsorbed ammonia was desorbed in a He gas flow (50 mL/min) for 2 h. Subsequently, the amount of ammonia desorbed was detected by a thermal conductivity detector while increasing the temperature of sample holder from 100 to 550 • C at a heating rate of 10 • C/min. FTIR spectra of adsorbed-pyridine were attained with a Spectrum GX device (Perkin Elmer Co., Waltham, MA, USA) equipped with an HgCdTe detector and a custom-made cell. FTIR spectra of pyridine adsorbed on the catalyst disk were measured while the temperature of the cell was increased to 300 • C [12,13].
In order to determine the amount of the coke over the spent catalyst, a TGA analysis was carried out using a SDT Q600 device (TA Instruments, New Castle, DE, USA). N 2 was injected through a mass flow controller (model 5850E from Brooks, Hatfield, PA, USA) operated using a flow pressure controller (ATOVAC, GMC 1200, Yongin, Korea). The sample was loaded onto a Pt holder, and the volume changes and energy flow were recorded while running N 2 (60 mL/min) and increasing the temperature up to 600 • C at a heating rate of 10 • C/min.

Oligomerization of Butene Mixture
The oligomerization reaction of the butene mixture was conducted in a fixed-bed continuous-flow reactor [26]. The reactor set-up and the experimental procedures for butene oligomerization were described in detail in the literature [26]. In our previous research on dehydration reaction of butanol, a mixture of butene (butene-1:butene-2 = 1.0:1.3) could be generated through dehydration of 2-butanol over a solid acid catalyst [13]. Accordingly, a mixture composed of 43.5% of butene-1 (99%, Rigas Korea Co., Daejeon, Korea) and 56.5% of butene-2 (11.8% of cis-2-butene and 87.5% of trans-2-butene, Rigas Korea Co., Daejeon, Korea) was used as a model reactant. The particle size of the catalyst was 25 µm or less. The reaction pressure and temperature were 15 bar and 350 • C, respectively, unless otherwise specified. In this study, most of the reaction products were alkene compounds. Needless to say, in order to use this product as a fuel, it must be converted to a paraffin compound via a hydrogenation reaction. The selectivity to C 8 -C 16 hydrocarbons was calculated based on the alkenes in the range of C 8 -C 16 . The formulae for the conversion of butene, the selectivity to C 8 -C 16 hydrocarbons, and the yield of C 8 -C 16 hydrocarbons are as follows: Conversion of butene (wt %) = (consumed butene/fed butene) * 100 (1) Selectivity of C 8 -C 16 (wt %) = (C 8 -C 16 product/consumed butene) * 100 Yield of C 8 -C 16 (wt %) = (C 8 -C 16 product/fed butene) * 100

Conclusions
Mesoporous aluminosilicate with a well-ordered structure was successfully prepared from HZSM-5 by combining a top-down method and a bottom-up method. The ordered channel-like mesopores were maintained after the nickel-loading on MMZ ZSM-5 catalyst. The NH 3 -TPD and pyridine-FTIR results revealed that a new kind of acid site (Lewis acid sites) was generated by NiO loading on MMZ ZSM-5 . The MMZ ZSM-5 catalyst showed much higher butene conversion and C 8 -C 16 yield in the butene oligomerization reaction than those with HZSM-5. This is attributed to the pore geometry of MMZ ZSM-5 , which is more beneficial for internal diffusion of reactants, reaction intermediates, and products. Moreover, the yield for C 8 -C 16 hydrocarbons over NiO/MMZ ZSM-5 was higher than that over the MMZ  , which seemed to be due to higher acid strength resulting from a higher ratio of Lewis sites to Brønsted sites. The present study reveals that mesoporous NiO/MMZ ZSM-5 catalyst with high concentration of Lewis acid sites is a potential catalyst for production of C 8 -C 16 hydrocarbons through the oligomerization of a butene mixture.