Phosphoric acid (H₃PO₄, 85wt%; Aladdin), pseudo-boehmite (Al₂O₃, 70wt%; Macklin), diisopropylamine (DIPA, 99wt%; Aladdin), di-n-propylamine (DPA, 99wt%; Aladdin), acid silica sol (SiO₂, 30wt%; Dezhou Jinghuo technology Glass Co., Ltd.), dodecyltrimethylammonium bromide (DTAB, 99wt%; Aladdin), nickel nitrate hexahydrate [Ni (NO₃)₂·6H₂O, 98wt%; Aladdin], ammonium metatungstate hydrate ((NH₄)₆ H₂W₁₂O₄₀·xH₂O, 99.5wt%; Macklin), hydrogen (H₂ ≥ 99.999%; Beijing Haipu Gas Co., Ltd.), and deionized water were the materials used.

SAPO-11 molecular sieves modified with Ni in situ were synthesized by the two-step crystallization method. In a typical example, first, 60.75 g of deionized water and 12.97 g of phosphoric acid were mixed to form a solution, and then, 11.0 g of pseudo-boehmite and a certain amount of Ni (NO₃)₂·6H₂O were added to the solution and stirred for 2 h. Second, 3.873 g of di-n-propylamine (DPA) and 3.873 g of diisopropylamine (DIPA) were added and continuously stirred for 2 h. Third, 6.76 g of acid silica sol was added dropwise to the system and stirred vigorously for 2 h. Finally, 1.17 g of dodecyltrimethylammonium bromide (DTAB) was added and stirred for 1 h to form an initial gel with a molar composition of 1.0 Al₂O₃: 0.75 P₂O₅: 0.45 SiO₂: x Ni(NO₃)₂ (x/Al₂O₃ = 1, 2, 3, and 4%): 0.5 DPA: 0.5 DIPA: 0.05 DTAB: 45 H₂O. The initial gel was precrystallized at 90°C for 12 h and then crystallized at 190°C for 24 h. The solid products collected by filtration were washed to neutrality with deionized water, dried at 110°C overnight, and calcined at 600°C for 6 h to obtain the Ni-modified SAPO-11 molecular sieves. The Ni-modified SAPO-11 samples were denoted as xNi-SAPO-11 (x = 1, 2, 3, and 4).

For comparison, a conventional SAPO-11 molecular sieve was obtained through a similar procedure used for synthesizing xNi-SAPO-11 above, but without adding Ni(NO₃)₂·6H₂O.

All the SAPO-11 samples were pressed, crushed, and sieved to obtain particles of a 20–40 mesh in size. The NiW-supported catalysts were prepared by the incipient wetness impregnation method with an aqueous solution of Ni(NO₃)₂·6H₂O and (NH₄)₆H₂W₁₂O₄₀·xH₂O and then dried at 110°C for 6 h and calcined at 500°C for 4 h after evaporation at room temperature overnight. For each catalyst, the loading concentration of NiO was 5%, and the loading concentration of WO₃ was 15%. The obtained catalysts were denoted as NiW/xNi-SAPO-11 (x = 1, 2, 3, and 4) and NiW/SAPO-11.

X-ray diffraction (XRD) characterization of the SAPO-11 samples was performed on a Bruker AXS D8 Advance X-ray diffractometer with Cu Kα radiation working under 40 kV and 40 mA and 2θ ranging from 5 to 90° at a scanning speed of 5°/min. Scanning electron microscopy (SEM) images of the SAPO-11 samples were obtained on an FEI Quanta 200 F field-emission environmental scanning electron microscope. N₂ adsorption–desorption characterization was carried out on a Micromeritics ASAP 2020 analyzer. The specific surface area, micropore volume, and mesopore volume were calculated by the Brunauer-Emmett-Teller (BET) method, the de Boer t-plot method, and the Barrett-Joyner-Halenda (BJH) method, respectively. Temperature-programmed desorption of ammonia (NH₃-TPD) characterization was carried out on a Micromeritics auto-chem 2920 instrument. The samples were heated at 600°C for 30 min in an Ar flow, and then, the Ar flow was switched to an ammonia flow and kept for 30 min when the samples were cooled to 70°C. The samples were purged with Ar for 2 h to remove the physically adsorbed ammonia, and the TPD signal was recorded using a thermal conductivity detector (TCD) from 70°C to 600°C with a heating rate of 10°C/min. Pyridine adsorbed infrared (Py-IR) spectra were recorded on a Nicolet 5700 spectrometer, and the samples with pyridine were evacuated at 200°C and 350°C, respectively. X-ray photoelectron spectroscopy (XPS) characterization of the sulfided catalysts was performed on a Thermo spectrometer using Al Kα radiation as the excitation light source. All spectra used the Al 2p peak at a binding energy of 74.6 eV to calibrate the binding energy scale. The XPS spectra of the sulfided catalysts were decomposed by fitting with XPS PEAK, and the deconvolution was realized by Gaussian-Lorentzian band shapes. High-resolution transmission electron microscopy (HRTEM) images of the sulfided catalysts were obtained on a JEM 2100 LaB₆ transmission electron microscope. The average length and average stacking number of WS₂ slabs were calculated by the following equation (Yu et al., 2012): Average   slab   length   L ¯ = ∑ i = 1 n n i l i ∑ i = 1 n n i , (1) Average   Stacking   Number   N ¯ = ∑ i = 1 n n i N i ∑ i = 1 n n i , (2) where l _i is the length of the WS₂ slabs, n _i is the number of the slabs with length l _i , and N _i is the number of layers of the WS₂ slabs. Based on the assumption that WS₂ slabs are standard hexagons, the dispersion degree of the WS₂ active phase was calculated by the following equation (Tao et al., 2014): f W = W e d g e W t o t a l = ∑ i = 1 t 6 ( n i − 1 ) ∑ i = 1 t ( 3 n i 2 − 3 n i + 1 ) , (3) where W _edge denotes the W atoms located on the edge of the WS₂ slabs, W _total denotes the total W atoms, n _i is the number of W atoms along one side of a WS₂ slab determined from its length [L = 3.2 (2n _i −1) Å], and t is the total number of slabs determined by no less than 500 WS₂ slabs obtained from the HRTEM images of different parts of the different catalysts.

The hydroisomerization of n-hexadecane (n-C₁₆) was carried out in a fixed-bed hydrogenation microreactor. In a typical example, 3 g of the catalyst was packed in the middle of the reaction tube, and both ends were filled with silica sand. Before the reaction, the catalyst was presulfurized with cyclohexane solution containing 2 vol% CS₂ for 5 h at 320°C, a liquid hourly space velocity (LHSV) of 7 h⁻¹, and H₂/oil of 100 (v/v). After the presulfurization treatment was completed, the catalytic performance of the catalyst was investigated under the conditions of a reaction pressure of 2 MPa, an LHSV of 1.5 h⁻¹, H₂/oil of 600 (v/v), and a reaction temperature of 320–400°C. The reactant n-hexadecane was fed into the reactor using a syringe pump when the temperature was decreased to the reaction temperature. The products collected were analyzed on a Shimadzu GC-2014 gas chromatograph equipped with a capillary HP-PONA column, and the qualitative analysis of the products was realized by GC-MS. The conversion of n-C₁₆ of each catalyst at 340°C was converted into the TOF (turnover frequency, which is used to evaluate the catalytic activity of the catalyst) of each active site, which was calculated by the following equation based on the number of all available active sites (Wen et al., 2019): T O F = V f e e d ⋅ x n W ⋅ f W , (4) where V _feed is the feed flow rate of the reactant n-C₁₆ in mol/h, x is the conversion of n-C₁₆ at 340°C, n _w is the amount of W atoms of the catalyst in mol, and f _W is the dispersion degree of W species.

Figure 1 presents the X-ray diffraction (XRD) patterns of the Ni-modified SAPO-11 molecular sieves. It can be seen that all the Ni-modified SAPO-11 samples presented characteristic diffraction peaks at 2θ = 8.2°, 9.6°, 13.0°, 15.8°, 20.2°, 21.1°, and 22.0–23.3°, which are attributed to the typical AEL structure, and no characteristic diffraction peaks of other crystalline phases were observed (Jin et al., 2016). The results indicate that the crystal structural units of the Ni-modified SAPO-11 molecular sieves were completely preserved. In addition, no characteristic diffraction peaks of Ni species were observed in the XRD patterns of all the Ni-modified SAPO-11 samples, indicating that Ni species were highly dispersed in SAPO-11. This result can be explained by the fact that Ni species substituted some of the Al species and entered the framework structure of SAPO-11 in the process of in situ modification.

The SEM images of the Ni-modified SAPO-11 samples are presented in Figure 2. All the samples showed pseudo-spherical particles with a particle size of about 3–7 μm, indicating that the introduction of Ni did not change the particle size distribution of SAPO-11 in the synthesis process. Moreover, no amorphous phase was observed in SEM images of all samples, indicating that the introduction of Ni into SAPO-11 by in situ modification did not change the morphology of SAPO-11. Through further careful observation, it can be found that the particles of all samples were assembled by many crystallites, which provides conditions for the formation of more intercrystallite mesoporous structures.

In order to obtain the textural properties such as specific surface area, pore size, and pore volume of the Ni-modified SAPO-11 molecular sieves, N₂ adsorption–desorption tests of all the samples were carried out. The N₂ adsorption–desorption isotherms and pore size distribution curves of all the samples are presented in Figure 3, while the data of texture properties are listed in Table 1. As shown in Figure 3A, all the samples showed type IV isotherms and typical H4-type hysteresis loops (Verma et al., 2015; Tao et al., 2017). The N₂ adsorption amount of all the samples increased sharply at a lower relative pressure (10^–5 < P/P₀ < 10^–2), which was attributed to the adsorption of N₂ in micropores (Liu et al., 2015). All the samples showed obvious hysteresis loops in the range of relative pressures of 0.4–0.9, indicating that there were a large number of mesoporous structures in all the samples. In addition, the hysteresis loops of Ni-modified SAPO-11 samples were larger than those of the unmodified SAPO-11, indicating that Ni-modified SAPO-11 samples possessed more mesoporous structures. Figure 3B presented the pore size distribution curves of all the samples, and it can be seen that the pore size of all samples was mainly distributed between 6 and 10 nm. Table 1 lists the textural property data of the Ni-modified SAPO-11 samples. The BET specific surface area (S_BET) of Ni-modified SAPO-11 samples increased in the order of SAPO-11 (166 m²/g) < 1Ni-SAPO-11 (169 m²/g) < 2Ni-SAPO-11 (171 m²/g) < 3Ni-SAPO-11 (174 m²/g) < 4Ni-SAPO-11 (176 m²/g) and the mesopore volume (V_meso) of Ni-modified SAPO-11 samples also increased in the order of SAPO-11 (0.35 m³/g) < 1Ni-SAPO-11 (0.36 m³/g) < 2Ni-SAPO-11 (0.37 m³/g) < 3Ni-SAPO-11 (0.39 m³/g) < 4Ni-SAPO-11 (0.41 m³/g), which indicate that Ni in situ modification is beneficial to increase the S_BET and V_meso of SAPO-11. Figure 4 and Table 2 present the N₂ adsorption–desorption isotherms, pore size distribution, and textural data of the different catalysts, respectively. It is not difficult to see that the pore size distribution of the catalysts supported with NiW is basically unchanged. However, the BET specific surface area, external specific surface area, micropore volume, and mesopore volume of the catalysts all decreased, and the change order of these data of the different catalysts still followed the change order of these data of the Ni-modified SAPO-11 molecular sieves.

Figure 5 presents the temperature-programmed desorption of ammonia (NH₃-TPD) profiles of Ni-modified SAPO-11 samples. Each curve can be fitted and deconvolved into three peaks with a Gaussian function. The three decomposition peaks near 180°C, 250°C, and 330°C correspond to the desorption of NH₃ adsorbed on weak acid sites, medium acid sites, and strong acid sites, respectively (Liu et al., 2009; Liu et al., 2020). It can be seen that with the increase of Ni content in the process of Ni modification, the intensity of low-temperature desorption peaks of the Ni modified SAPO-11 samples gradually decreased, indicating that the weak acid amount of Ni-modified SAPO-11 samples decreased with the increase of Ni content. The intensity of medium-temperature desorption peaks and high-temperature desorption peaks of Ni-modified SAPO-11 samples increased in the order of SAPO-11 < 1Ni-SAPO-11 < 2Ni-SAPO-11 < 4Ni-SAPO-11 < 3Ni-SAPO-11, indicating that the medium and strong acid amounts also increased in the same order. This result can be explained by the fact that the average electronegativity of Ni (1.91) is higher than that of Al (1.71), and the covalency of Ni is higher than that of Al (Zhou et al., 2017). Therefore, the Brønsted protons and Lewis caves of SAPO-11 formed by substituting part of Al species with Ni species have stronger acid density. However, when the Ni content reached 4%, not only Ni species entered the framework of SAPO-11 but also excessive nickel species located on the surface of SAPO-11 and covered a part of acid sites. Therefore, the medium and strong acid amounts of 4Ni-SAPO-11 are less than that of 3Ni-SAPO-11. In addition, the different catalysts were also characterized by NH₃-TPD. Compared with Ni-modified SAPO-11 molecular sieves, the peak areas of NH₃-TPD profiles of the catalysts all decreased, indicating that the acid amount of the catalysts decreased after being supported with NiW. However, the variation in peak area of the catalysts was still consistent with that of the Ni-modified SAPO-11 samples.

To qualitatively analyze the acidity of the Ni-modified SAPO-11 samples, the samples were characterized by Py-IR. The Py-IR spectra of the Ni-modified SAPO-11 samples are presented in Figure 6. The weak acid amount was calculated from the Py-IR spectra formed by pyridine molecules desorbed at 200°C (Figure 6A), while the medium and strong acid amounts were calculated from the Py-IR spectra formed by pyridine molecules desorbed at 350°C (Figure 6B). The corresponding acid amount data are listed in Table 3. As shown in Figure 6, all the samples showed three peaks in the range of 1,400–1,600 cm⁻¹. The two peaks at 1,455 cm⁻¹ and 1,545 cm⁻¹ correspond to the desorption of pyridine molecules on Brønsted (B) acid sites and Lewis (L) acid sites, respectively, while the peak at 1,490 cm⁻¹ is attributed to the synergistic effect of B acid sites and L acid sites (Guo et al., 2013; Du et al., 2019). Compared with SAPO-11, the weak L acid amount of Ni-modified SAPO-11 samples increased and the weak B acid amount decreased. This result can be explained by the fact that NiO entering the framework of SAPO-11 can form some new weak L acid sites in the process of Ni in situ modification. The weak B acid sites of SAPO-11 are generated by Si-OH, Al-OH, and P-OH (Yang et al., 2017). Ni species substitute some Al species to form Ni-OH species, which cannot form weak B acid sites like Al-OH species. The amount of medium and strong B acid sites of the Ni-modified SAPO-11 samples increased in the order of SAPO-11 (30.73 μmol/g) < 1Ni-SAPO-11 (32.42 μmol/g) < 4Ni-SAPO-11 (32.62 μmol/g) < 2Ni-SAPO-11 (35.25 μmol/g) < 3Ni-SAPO-11 (37.21 μmol/g). The result can be explained by the fact that the medium and strong B acid sites of SAPO-11 are generated by the Si-OH-Al species (Blasco et al., 2006). In the process of Ni in situ modification, Ni species substitute some Al species to form Si-OH-Ni species, and their acid density is greater than that of Si-OH-Al species. The reason for the medium and strong B acid amounts of 4Ni-SAPO-11 being lower than that of 3Ni-SAPO-11 is that excessive nickel species could not completely enter the framework of 4Ni-SAPO-11 and were distributed on the surface and pores of SAPO-11 and covered some acid sites. The results of the acid amount reflected by Py-IR tests well verified the results of the acid amount reflected by NH₃-TPD.

To visualize the morphology and stacking of the active phase on the catalysts, HRTEM characterization of the sulfided catalysts was carried out. The representative HRTEM images of the different sulfided catalysts are presented in Figure 7. The black line-like layers in the images are the typical WS₂ slabs, and the NiS_x phase is too small to be observed in the images. The statistical analysis was carried out on no less than 500 slabs from different parts of different catalysts. The average length and average stacking number of WS₂ slabs on each catalyst were calculated from Eqs 1, 2, respectively, and the results are listed in Table 4. The statistical results showed that the average length of WS₂ slabs on different catalysts increased in the order of NiW/3Ni-SAPO-11 (2.98 nm) < NiW/4Ni-SAPO-11 (3.23 nm) < NiW/2Ni-SAPO-11 (3.46 nm) < NiW/1Ni-SAPO-11 (4.03 nm) < NiW/SAPO-11 (4.72 nm). The length of WS₂ slabs on NiW/SAPO-11 was in the range of 2–8 nm and mainly concentrated in 3–7 nm. The lengths of WS₂ slabs on NiW/1Ni-SAPO-11, NiW/2Ni-SAPO-11, NiW/3Ni-SAPO-11, and NiW/4Ni-SAPO-11 were mainly concentrated in 2–7 nm, 2–6 nm, 1–4 nm, and 2–5 nm, respectively. The results showed that Ni in situ modification decreased the slab length of the active phase. It is believed that the length of active phase slabs is related to the dispersion of the active phase. The longer the length of active phase slabs, the lower the dispersion degree of the active phase (Zhou et al., 2018). The result indicates that the dispersion degree of the so-called NiWS active phase decreased in the order of NiW/3Ni-SAPO-11 > NiW/4Ni-SAPO-11 > NiW/2Ni-SAPO-11 > NiW/1Ni-SAPO-11 > NiW/SAPO-11. The stacking number of the active phase can reflect the interaction between the active phase and support to a certain extent. Generally speaking, the higher the stacking number of the active phase, the weaker the interaction between the active phase and support (Cui et al., 2012). The average stacking number of WS₂ slabs on the different catalysts increased in the order of NiW/SAPO-11 (1.33) < NiW/1Ni-SAPO-11 (1.85) < NiW/2Ni-SAPO-11 (2.68) < NiW/4Ni-SAPO-11 (3.06) < NiW/3Ni-SAPO-11 (3.23), indicating that the interaction between the active phase and support decreased in the order of NiW/SAPO-11 > NiW/1Ni-SAPO-11 > NiW/2Ni-SAPO-11 > NiW/4Ni-SAPO-11 > NiW/3Ni-SAPO-11. In order to further verify the change of dispersion degree of the active phase on the different catalysts, the dispersion degree of tungsten species (f _W ) on the different catalysts was calculated by Equation 3. The results showed that the dispersion degree of tungsten species on different catalysts decreased in the order of NiW/3Ni-SAPO-11 > NiW/2Ni-SAPO-11 > NiW/4Ni-SAPO-11 ≈ NiW/1Ni-SAPO-11 > NiW/SAPO-11, which was completely consistent with the results reflected by the length and stacking number of WS₂ slabs. The result can be explained by the fact that Al in W-O-Al species was substituted by Ni to form W-O-Ni species on the Ni-modified SAPO-11-supported NiW catalysts. However, the W-O-Ni bond is weaker than the W-O-Al bond, which weakened the interaction between the active phase and support and made the dispersion of the active phase poor. However, the average slab length of the active phase of NiW/4Ni-SAPO-11 was longer than that of NiW/3Ni-SAPO-11, and the stacking number of the active phase of NiW/4Ni-SAPO-11 was lower than that of NiW/3Ni-SAPO-11. This may be due to the fact that excessive nickel species did not completely enter the framework of the SAPO-11 molecular sieve and distributed on the surface and pores of SAPO-11, which hindered the dispersion of the so-called NiWS active phase.

The different catalysts were characterized by XPS to analyze the covalent state of nickel species and tungsten species on the catalysts. Figures 8, 9 present the W4f XPS spectra and Ni2p XPS spectra of the sulfided catalysts, respectively. The sulfidation degree of tungsten species and nickel species was calculated by the deconvolution peak area. The results of the binding energy (BE) and sulfidation degree of tungsten species and nickel species of the sulfided catalysts are listed in Tables 5, 6, respectively. The W 4f XPS spectra of different catalysts were decomposed into four peaks, which were attributed to the two overlapping peaks of W⁴⁺ and W⁶⁺. The decomposition peaks in Figure 8 with binding energies of 36.30 ± 0.40 eV and 38.10 ± 0.40 eV were attributed to the W 4f_7/2 levels and W 4f_5/2 levels of W⁶⁺ (WO₃), respectively, while the decomposition peaks with binding energies of 32.40 ± 0.40 eV and 34.80 ± 0.40 eV were attributed to the W 4f_7/2 levels and W 4f_5/2 levels of W⁴⁺ (WS₂), respectively. The binding energies of W 4f_7/2 levels and W 4f_5/2 levels of the investigated catalysts decreased in the order of NiW/SAPO-11 > NiW/1Ni-SAPO-11 > NiW/2Ni-SAPO-11 > NiW/3Ni-SAPO-11 > NiW/4Ni-SAPO-11, which indicates that the interaction between the active phase and support of the investigated catalysts also decreased in the same order, which was consistent with the results reflected by HRTEM images. The sulfidation degree of the tungsten species of different catalysts was defined as W⁴⁺/(W⁴⁺ + W⁶⁺) (Cui et al., 2013), which increased in the order of NiW/SAPO-11 (56.29%) < NiW/4Ni-SAPO-11 (57.23%) < NiW/1Ni-SAPO-11 (57.69%) < NiW/2Ni-SAPO-11 (58.48%) < NiW/3Ni-SAPO-11 (60.93). This result can be explained by the following facts: Ni species substituted some Al species, resulting in some of the W-O-Al bonds being substituted with W-O-Ni bonds in the process of Ni in situ modification, and the strength of W-O-Ni bonds was lower than that of W-O-Al bonds, thus reducing the interaction between the active phase and support and improving the sulfidation degree of tungsten species. However, the excessive addition of Ni in the process of in situ modification of SAPO-11 led to poor dispersion of active metals on the catalyst, which led to the lower sulfidation degree of tungsten species of NiW/4Ni-SAPO-11 than NiW/3Ni-SAPO-11. The Ni 2p XPS spectra of different catalysts could also be decomposed into four peaks, which were attributed to the two overlapping peaks of nickel oxide and nickel sulfide. The decomposition peaks in Figure 9 with binding energies of 858.80 ± 0.40 eV and 877.40 ± 0.40 eV were attributed to the Ni 2p_3/2 levels and Ni 2p_1/2 levels of nickel oxide (NiO), respectively, while the decomposition peaks with binding energies of 853.20 ± 0.40 eV and 870.70 ± 0.40 eV were attributed to the Ni 2p_3/2 levels and Ni 2p_1/2 levels of nickel sulfide (NiS_x), respectively. The decreasing order of the binding energy of the levels of Ni is the same as that of the levels of W described above. This result further confirms that Ni in situ modification weakened the interaction between the active phase and support. The sulfidation degree of the nickel species was defined as NiS_x/(NiO + NiS_x) (Cui et al., 2013), which increased in the order of NiW/SAPO-11 (60.17%) < NiW/1Ni-SAPO-11 (60.69%) < NiW/2Ni-SAPO-11 (61.58%) < NiW/4Ni-SAPO-11 (62.53%) < NiW/3Ni-SAPO-11 (64.93). The reason for this result is the same as the above explanation about tungsten species.

The evaluation of the hydroisomerization performance of all the catalysts was carried out in a fixed-bed hydrogenation microreactor. In each reaction, the catalyst was presulfurized in advance, the reaction pressure was fixed at 2.0 MPa, the liquid hourly space velocity (LHSV) was fixed at 1.5 h⁻¹, and H₂/oil was 600 (v/v). The catalytic performance of the catalysts was evaluated in the range of 320–400°C. Figure 10A presents the conversion of n-C₁₆ on the different catalysts at different reaction temperatures. It can be seen that the n-C₁₆ conversion of all the catalysts increased with the increase of temperature. Among all the investigated catalysts, NiW/3Ni-SAPO-11 showed the highest conversion of n-C₁₆ in the range of investigated reaction temperatures, which indicates that NiW/3Ni-SAPO-11 had the highest catalytic activity among all the investigated catalysts, and the catalytic activities of the catalysts increased in the order of NiW/SAPO-11 < NiW/1Ni-SAPO-11 < NiW/2Ni-SAPO-11 < NiW/4Ni-SAPO-11 < NiW/3Ni-SAPO-11. The result can be explained by the fact that the amount of acid in the support and the properties of the active phase jointly determine the catalytic activity of the catalyst. It is generally believed that the acid amount of the catalyst, especially the medium and strong B acid amounts, determines the isomerization rate of olefin intermediates in the isomerization process. Moreover, the higher the dispersion degree of the active phase, the better the hydrogenation performance of the catalyst, and with the appropriate increase of the number of stacking layers of the active phase, the hydrogenation performance of the active phase also increases. The appropriate interaction between the active phase and support of NiW/3Ni-SAPO-11 and the highest dispersion of the active phase made NiW/3Ni-SAPO-11 show the highest catalytic activity. The selectivity of isomeric hexadecane (i-C₁₆) decreased with the increase of reaction temperature, which indicates that the increase of reaction temperature intensified the cracking reaction. The selectivity of i-C₁₆ decreased greatly when the reaction temperature reached 380°C. Among all the investigated catalysts, the i-C₁₆ selectivity of NiW/3Ni-SAPO-11 was the highest in the range of investigated reaction temperatures, and the i-C₁₆ selectivity of different catalysts at different reaction temperatures basically increased in the order of NiW/SAPO-11 < NiW/1Ni-SAPO-11 < NiW/2Ni-SAPO-11 < NiW/4Ni-SAPO-11 < NiW/3Ni-SAPO-11 (Figure 10B). This result can be explained by the fact that 3Ni-SAPO-11 possessed the most medium and strong B acid sites, which are considered as the active centers of the isomerization of the alkane skeleton (Lyu et al., 2017). Meanwhile, the increase of mesoporous structures alleviated the diffusion resistance of n-hexadecane and isomeric hexadecane in the channels of SAPO-11, thus reducing the occurrence of cracking reactions. 3Ni-SAPO-11 showed the highest selectivity of i-C₁₆ due to its highest medium and strong B acid amount and mesopore volume. Among all the investigated catalysts, the i-C₁₆ yield of NiW/3Ni-SAPO-11 was the highest in the range of investigated temperatures, and the highest yield was 71.18% (Figure 10D).

Table 7 lists the isomer distributions and TOF values of the different catalysts at 340°C. It is not difficult to find that the isomer products of all the catalysts are mainly monobranched isomers and double-branched isomers, and the selectivity of monobranched isomers is much higher than that of double-branched isomers, and the selectivity of monomethyl branched isomers is much higher than that of monoethyl branched isomers. Among the isomer products, monobranched isomers mainly include 2-methylpentadecane (2-MC₁₅), 3-methylpentadecane (3-MC₁₅), 4-methylpentadecane (4-MC₁₅), 5-methylpentadecane (5-MC₁₅), 6-methylpentadecane (6-MC₁₅), 3-ethyltetradecane (3-EC₁₄), 5-ethyltetradecane (5-EC₁₄), and 6-ethyltetradecane (6-EC₁₄), while the double-branched isomers mainly include 2,4-dimethyltetradecane (2,4-DMC₁₄), 2,5-dimethyltetradecane (2,5-DMC₁₄), 2,6-dimethyltetradecane (2,6-DMC₁₄), 3,6-dimethyltetradecane (3,6-DMC₁₄), 3,7-dimethyltetradecane (3,7-DMC₁₄), and 5,8-diethyldodecane (5,8-DEC₁₂). Among all the isomer products, the selectivity of 2-MC₁₅ and 3-MC₁₅ is obviously higher than that of other isomer products, which is the same as that reported by Zhang et al. (Zhang et al., 2018). TOF values of the different catalysts refer to the standardized turnover frequency at each available active phase center, which can effectively reflect the catalytic activity of the catalysts. TOF of the different catalysts increased in the order of NiW/SAPO-11 (27.70 h⁻¹) > NiW/4Ni-SAPO-11 (27.71 h⁻¹) > NiW/1Ni-SAPO-11 (28.44 h⁻¹) > NiW/2Ni-SAPO-11 (29.80 h⁻¹) > NiW/3Ni-SAPO-11 (30.81 h⁻¹), which well verified the results reflected by the above catalytic activity.