Based on Fig. 2, diffractograms (a) and (b) each show diffraction patterns with sharp peaks of high intensity, indicating that ZSM-5 and Ni/ZSM-5 catalysts are crystalline materials. The impregnation of Ni metal into ZSM-5 does not alter the crystalline structure of the catalyst. However, the presence of Ni metal causes a slight decrease in the peak intensity of ZSM-5 in diffractogram (b). This indicates that the crystallinity of the catalyst decreases after metal impregnation. The reduction in crystallinity degree may be attributed to pore blockage of ZSM-5 by the metal and the interaction between the metal and the pore walls of the support [23]. The degree of crystallinity and crystal size of each catalyst are presented in Table 1..

The use of ultrasonic waves in the synthesis of Ni/ZSM-5 catalysts leads to a reduction in crystal size due to the cavitation effect. Cavitation generates localized high pressure and temperature, which induces particle collisions and micro-explosions, thereby preventing agglomeration and breaking growing crystals. In addition, cavitation can accelerate the nucleation process and inhibit further crystal growth, resulting in smaller and more uniformly distributed crystals.

Based on Table 1, the average crystal size of ZSM-5 decreased by 2.81 nm after Ni metal impregnation. The reduction in crystal size may be attributed to the interaction between Ni metal and ZSM-5 during ultrasonic impregnation and subsequent processes (drying, calcination, and reduction), which causes slight distortions in the ZSM-5 lattice structure around the NiO particles. This distortion may also lead to the broadening of the ZSM-5 XRD peaks, contributing to a smaller crystallite size. The smaller particle size of the active phase (Ni) correlates with higher catalytic activity due to better metal dispersion and a larger active surface area.

The diffraction peaks of the ZSM-5 catalyst at 2θ = 7.81°; 8.67°; 22.92°; 23.76°; and 26.50° match the Crystallography Open Database (COD) entry No. 96-154-8625. Meanwhile, for the Ni/ZSM-5 catalyst, slight differences in diffraction patterns are observed at 2θ = 7.85°; 8.73°; 22.99°; 23.84°; and 26.55°. These small variations in 2θ values indicate a slight shift in the diffraction peak positions. Such shifts may be attributed to changes in crystal lattice strain caused by the impregnation of nickel metal ions into the zeolite structure. The incorporation of Ni²⁺ ions into the pores or lattice of ZSM-5 can induce local distortions in the crystal structure, thereby affecting the interplanar spacing (d-spacing) and resulting in peak shifts at 2θ angles. In addition, new peaks for the Ni/ZSM-5 catalyst appear at 2θ = 37.23°; 43.25°; 52.24°; 62.83°; and 75.35°, which correspond to COD entry No. 96-101-0096 for NiO and No. 96-901-1604 for Ni. These peaks confirm that the impregnation of Ni metal into the ZSM-5 support assisted by ultrasonic waves was successfully achieved. However, the reduction of Ni metal did not proceed completely, as evidenced by the presence of oxidized Ni in the form of NiO. The incomplete reduction process may be due to strong interactions between NiO and the ZSM-5 surface, which hinder H₂ dissociation, the relatively large NiO particle size that slows down H₂ diffusion throughout the particles, and suboptimal reduction conditions (low temperature, short duration, and insufficient H₂ flow rate).

The presence of functional groups in a material can be identified using a Fourier Transform Infrared Spectroscopy (FT-IR) spectrometer. The functional groups in a catalyst can be determined by matching molecular vibrations with wavenumbers. The wavenumbers analyzed for the detection of functional groups present in the catalyst material are within the range of 4000 to 400 cm⁻¹. The infrared spectra of ZSM-5 and Ni/ZSM-5 catalysts are shown in Fig. 3.

Based on Fig. 3, the FT-IR spectra of ZSM-5 and Ni/ZSM-5 catalysts show no significant changes in absorption peaks. This indicates that the metal impregnation process does not substantially affect the presence of functional groups on the support. The absorption at a wavenumber of around 432 cm⁻¹ corresponds to the bending vibration of T–O–T (T = Si or Al) [24]. The bands at approximately 790 and 1203 cm⁻¹ are assigned to the symmetric stretching vibration of T–O–T (T = Si or Al) and the asymmetric stretching vibration of T–O–T (T = Si or Al), respectively [25]. The asymmetric stretching vibration of Si–O–Si is observed at a wavenumber of 1060 cm⁻¹. The absorption band at 1060 cm⁻¹ represents the main framework of zeolite [26]. The absorption band around 1631 cm⁻¹ arises from the bending vibration of adsorbed water molecules on the ZSM-5 surface. Due to its hygroscopic nature, zeolite readily binds water at acidic sites or oxygen atoms on its surface [26]. The absorption at 3649 cm⁻¹ is attributed to the stretching vibration of internal silanol hydroxyl groups (Si–OH) located at structural defects or crystal edges. These groups are acidic and often act as catalytic active sites. The intensity of the absorption band at 3649 cm⁻¹ in Ni/ZSM-5 is slightly lower compared to ZSM-5. This decrease may be attributed to the interaction between Ni metal precursors and silanol groups on the ZSM-5 surface during the impregnation process. The interaction of Ni with silanol groups leads to the formation of NiOH⁺ species [27].

Surface Area Analyzer (SAA) is used for the textural analysis of a material. The methods applied for textural analysis are Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH). The BET method is used to determine the specific surface area of the material through nitrogen adsorption measured as a function of relative pressure. The BJH analysis is employed to determine pore size and volume using adsorption and desorption techniques. The surface characteristics of ZSM-5 and Ni/ZSM-5 catalysts are presented in Table 2.

Based on Table 2, the specific surface area and total pore volume of the ZSM-5 catalyst decreased after Ni metal impregnation. This was due to the partial blockage of the ZSM-5 pores by Ni metal particles. The Ni particles entered the pore surfaces, thereby hindering access into the ZSM-5 framework. However, the average pore diameter increased after Ni loading, which can be attributed to the interaction between Ni metal and the ZSM-5 structure during the impregnation process. The MFI-type hexagonal structure of ZSM-5 has non-uniform pores. The smaller pores of ZSM-5 were blocked by Ni particles, resulting in the opening of pores in the mesoporous region of the catalyst [28]. The classification of diameter pore size is divided into three categories: micropores (< 2 nm), mesopores (2–50 nm), and macropores (> 50 nm) [29]. Both ZSM-5 and Ni/ZSM-5 catalysts are categorized as mesoporous materials, with average pore radius of 2.10 nm and 2.53 nm, respectively. The adsorption–desorption isotherm in Fig. 4 illustrates the pore type of the catalyst.

Based on Fig. 4, the ZSM-5 and Ni/ZSM-5 catalysts exhibit a type IV(a) isotherm, which indicates that the catalytic material is mesoporous [30]. This is consistent with the results obtained in Table 2. The red line represents adsorption, while the black line represents desorption. The continuous increase in adsorption and the decrease in desorption eventually converge at a single point, forming a ring known as a hysteresis loop. The hysteresis loop appears at a relative pressure of 0.2–1.0 and belongs to the H2(a) type, similar to an ink bottle. This type indicates pore blocking or percolation of ZSM-5 by metals due to cavitation effects from ultrasonic waves [29]. The irregular pore shape of ZSM-5 causes an uneven distribution of the catalyst pore radius. The pore radius distribution of the catalyst is presented in Fig. 5.

The surface acidity of the catalyst can be evaluated using NH₃-TPD. In this method, the sample is saturated with ammonia (NH₃), a basic molecule, and subsequently heated in a programmed manner. The desorbed ammonia from the catalyst surface is detected as a function of temperature. A higher desorption temperature generally indicates the presence of stronger acid sites. The concentration and strength of acid sites are determined from the integrated peak areas in the NH₃-TPD profile. Acid sites are typically classified into two categories: weak acid sites, which are desorbed at temperatures of 100–350°C, and strong acid sites, which are desorbed at temperatures above 350°C [31]. The NH₃-TPD curve is presented in Fig. 6.

Figure 6 illustrates that the ZSM-5 catalyst exhibits two significant desorption peaks. The first peak appears at 150–200°C, indicating the presence of weak acid sites. The second peak emerges at 350–400°C, corresponding to stronger acid sites. The Ni/ZSM-5 catalyst also displays two desorption peaks. The first peak occurs within a similar temperature range to that of ZSM-5, but with slightly higher intensity. The second peak appears at a higher temperature (400–450°C) compared to ZSM-5 and shows greater intensity. Moreover, Ni/ZSM-5 exhibits a broader desorption area at temperatures above 600°C relative to ZSM-5. The acidity of both ZSM-5 and Ni/ZSM-5 catalysts is summarized in Table 3.

The NH₃-TPD analysis results presented in Table 3 reveal that nickel impregnation on ZSM-5 significantly enhances the total acidity of the catalyst, increasing from 1.398 mmol/g to 2.331 mmol/g, encompassing both weak acid sites (from 0.681 to 1.003 mmol/g) and strong acid sites (from 0.717 to 1.328 mmol/g). Weak acid sites facilitate the removal of oxygen-containing groups through mild deoxygenation reactions without inducing excessive carbon chain cleavage, thereby preserving hydrocarbon yield. In contrast, strong acid sites contribute to cracking and isomerization, regulating the carbon chain length (C₇–C₁₆) to meet bio-jet fuel specifications. Moreover, the incorporation of Ni not only enhances acidity but also provides active metallic sites for hydrogenation reactions, thus rendering Ni/ZSM-5 an effective and selective bifunctional catalyst for the conversion of vegetable oils into bio-jet fuel.

X-ray Photoelectron Spectroscopy (XPS) is an analytical technique employed to determine the chemical composition and oxidation states of elements on the surface of a material. XPS analysis is also utilized to evaluate the effectiveness of nickel impregnation assisted by ultrasonic waves. The spectra obtained from XPS are generally classified into two types: survey spectra (wide scan) and high-resolution spectra (narrow scan). The wide scan and narrow scan spectra of Ni species in the Ni/ZSM-5 catalyst are presented in Fig. 7.

The XPS wide-scan spectra were analyzed within the binding energy range of 0–1200 eV. The detected chemical compositions included Si, Al, O, and Ni. The binding energies of Al 2p, Si 2p, O 1s, and Ni 2p were observed at 66.22, 102.05, 529.14, and 853.01 eV, respectively. These binding energy values are consistent with previous studies by [32–34] in which the peaks of Al, Si, O, and Ni were reported at approximately 69, 103, 529, and 854 eV, respectively. The O 1s signal exhibited the highest intensity in the wide-scan spectra, which can be attributed to the ability of ZSM-5 to adsorb H₂O from the atmosphere or during the calcination process onto the surface and within the pores of the material. This adsorption contributes to the XPS spectra by forming hydroxyl groups, consistent with FT-IR analysis (Si–OH and Al–OH). Additionally, the presence of

Si–O–Si and Si–O–Al bonds in ZSM-5 further contributes to the strong O 1s intensity

[34]. Narrow-scan spectra were employed to analyze the chemical composition of specific elements with higher resolution, particularly focusing on the Ni peaks. Overlapping peaks were deconvoluted to obtain more detailed information on the composition of each peak. Following deconvolution, the J-coupling values were calculated based on quantum numbers to determine the term symbols of Ni 2p in its various oxidation states, resulting in the identification of 2p₃/₂ and 2p₁/₂.

The binding energy of 2p₃/₂ was found to be lower than that of 2p₁/₂, indicating that the 2p₁/₂ electrons are more strongly bound to the nucleus and therefore more difficult to remove. The degeneracy of states with the same energy was then used to determine the number of electrons in each state, with 2 electrons in 2p₁/₂ and 4 electrons in 2p₃/₂. The detection of Ni species confirms the successful impregnation of Ni into the ZSM-5 support. However, Fig. 7 (b) shows the presence of both Ni⁰ and Ni²⁺ species, indicating that Ni was not fully reduced during the calcination–reduction process. This observation is consistent with the XRD results, which revealed peaks corresponding to metallic Ni and nickel oxide species. The binding energies of Ni⁰ at 2p₃/₂ and 2p₁/₂ were observed at 853.01 and 870.63 eV, respectively, while the Ni²⁺ species exhibited binding energies of 859.25 and 875.71 eV for 2p₃/₂ and 2p₁/₂, respectively. The spin–orbit splitting of Ni 2p₃/₂ was found within the typical binding energy range of 850–868 eV, with Ni⁰ detected at ~ 853 eV and Ni²⁺ at ~ 855 eV [35].

The metal content and morphology of a catalyst can be analyzed using SEM-EDX mapping. Scanning Electron Microscopy (SEM) provides detailed information on surface topography, texture, the presence of particles, cracks, and layers within a material. Energy Dispersive X-ray (EDX) mapping data reveal the elemental composition and distribution on the imaged catalyst surface. The results of elemental composition analysis for the ZSM-5 catalyst, Ni/ZSM-5 before use, and Ni/ZSM-5 after use (upper and lower sections) are presented in Tables 3 and 4.

According to Table 3, the Ni metal content detected by the EDX instrument was 5.52%. This result is close to the theoretical calculation, in which the Ni/ZSM-5 catalyst is expected to contain 6% Ni. The data also confirm that the impregnation of Ni metal assisted by ultrasonic waves was successfully achieved. However, the slight deviation from the theoretical value suggests suboptimal deposition of Ni on the support surface. This may be attributed to the less-than-ideal catalyst preparation process, resulting in non-uniform dispersion of Ni throughout the catalyst. In addition, weak interactions between Ni and the ZSM-5 support may also reduce the number of active metal sites on the surface, making the Ni species more prone to leaching from the catalyst pores [20]. Table 4, further shows that the Ni content decreases after the catalyst is reused three times. The Ni content in the upper and lower parts of the Ni/ZSM-5 catalyst decreased by 1.25% and 5.22%, respectively. The greater reduction in the lower part of the catalyst is likely due to its direct exposure to heat during the hydrocracking process. Moreover, a significant increase in carbon content was observed after three catalytic cycles, indicating coke formation during the cracking process.

Figure 8 (a) shows that the distribution of Ni metal prior to the use of the catalyst in the hydrocracking process remains uniform. However, after three cycles of catalytic use,

a reduction in the content of Ni as well as the constituent elements of ZSM-5, such as Si, Al, and O, is observed. This is evidenced by the mapping results in Figures IV.8(b) and IV.8(c), where the distribution of Ni (blue) and the ZSM-5 framework elements appears diminished, accompanied by the formation of black voids. The non-uniform distribution and sintering of Ni particles, caused by agglomeration into larger clusters at specific sites, lead to the blockage of support pores, thereby reducing catalytic activity [36]. The morphology of ZSM-5 and Ni/ZSM-5 catalysts before and after use is presented in Fig. 9.

Figure 9 (a) and (b) display the morphology of the ZSM-5 and Ni/ZSM-5 catalysts prior to use. The interstitial regions of the ZSM-5 catalyst (point a) appear to be occupied by small adhering particles, resulting in a rougher and more pronounced surface (point b). These small particles correspond to Ni species impregnated onto the ZSM-5 support. This observation indicates that the incorporation of Ni metal can alter the morphology of ZSM-5 [20]. No significant differences in the surface appearance between the ZSM-5 and Ni/ZSM-5 catalysts were observed, suggesting that no bulk phase formation occurred after the metal impregnation process [37].

The morphology of the Ni/ZSM-5 catalyst after use in a double-bed configuration is presented in Fig. 9 (c) and (d). Compared to the fresh Ni/ZSM-5 catalyst (point b), the spent catalyst exhibits substantial morphological changes. Both the upper and lower layers of the spent Ni/ZSM-5 catalyst show the presence of agglomerates as a result of being subjected to three successive cracking cycles. The cracking process requires high operating temperatures, which promote the agglomeration of active catalytic particles, leading to metal sintering. The lower Ni/ZSM-5 catalyst appears darker due to coke deposition. Moreover, the structure of the lower spent catalyst shows greater degradation than that of the upper layer, as the lower catalyst bed is directly exposed to the feed during the cracking process.

The catalytic activity of ZSM-5 and Ni/ZSM-5 was evaluated through a hydrotreatment reaction using hydrogen gas at a flow rate of 20 mL/min. The hydrotreatment process was conducted in a one-pot double-decker reactor, with the temperature gradually increased within the range of 200–550°C over a duration of 3 hours. A thermal test was carried out to determine the initial temperature at which the liquid product underwent condensation. This temperature was subsequently employed as a reference parameter for assessing the catalytic activity of both single-layer and dual-layer Ni/ZSM-5 catalysts. The experimental results revealed that, under thermal conditions, the liquid product began to condense at 395°C. The overall conversion of liquid products obtained from thermal cracking, catalytic cracking, as well as from single-layer and dual-layer Ni/ZSM-5 catalysts, is presented in Fig. 10.

Based on Fig. 10, it can be observed that the cracking process conducted without a catalyst (thermal cracking) produced 38.7% liquid products. When ZSM-5 was used as the support, the liquid product conversion increased significantly to 54.71%. The Ni/ZSM-5 catalyst, arranged in both single- and double-bed configurations, was tested in three consecutive runs to evaluate catalyst lifetime. The single-bed catalyst exhibited a generally increasing trend in liquid product conversion, from 56.24% in run 1 to 62.48% in run 3, despite a slight decrease in run 2 (60.31%). Overall, the single-bed Ni/ZSM-5 catalyst demonstrated superior performance compared to ZSM-5 alone. In the double-bed configuration, the Ni/ZSM-5 catalyst achieved liquid product conversions of 59.97% in run 1 and 64.93% in run 2, representing the highest conversions among all conditions. However, a significant decline was observed in run 3, with the conversion dropping to 36.6%. This reduction can be attributed to coke deposition. During the cracking process, particularly at elevated temperatures, solid carbonaceous deposits (coke) form on the catalyst surface. Such coke accumulation progressively covers the active sites, hindering the access of reactant molecules to the catalyst and markedly reducing catalytic activity. The double-bed configuration tends to accelerate coke formation in the lower catalyst layer due to exposure to heavier by-products or differences in temperature distribution within the reactor. While coke deposition remained relatively limited in the first two runs, repeated use resulted in substantial accumulation, leading to severe performance degradation. In addition, deactivation of nickel sites further contributed to the reduced conversion in the third run of the double-bed catalyst.

Nickel serves as an active metal that plays a crucial role in hydrogenation and hydrodeoxygenation reactions essential for bio-jet fuel production. At high temperatures, small Ni nanoparticles may undergo sintering, agglomerating into larger particles. This increase in Ni particle size reduces the metal’s active surface area, thereby decreasing catalytic activity. Overall, both single- and double-bed Ni/ZSM-5 catalysts outperformed ZSM-5 alone. Nickel is a group 10 (VIII B) transition metal with atomic number 28. The electron configuration of nickel atom shows two relatively stable states, namely [Ar] 3d⁸ 4s² and [Ar] 3d⁹ 4s¹. Analysis of atomic energy levels shows that the configuration [Ar] 3d9 4s1 has a lower average energy compared to [Ar] 3d⁸ 4s². The existence of incompletely filled d and s orbitals (partially filled) in the electronic configuration of nickel provides an opportunity for unpaired electrons from other species to form coordinate bonds with the Ni atom.

This enhanced performance can be ascribed to the presence of Ni, which increases the number of acid sites in the catalyst. This statement is supported by NH₃-TPD characterization results, which revealed that Ni/ZSM-5 possesses higher acidity compared to ZSM-5. The higher the metal content and acid site density of the catalyst, the greater its catalytic activity and selectivity [20, 39–41]. A higher acidity value indicates a greater number of acid sites, which play a key role in the hydrocracking reaction. A larger specific surface area may indicate an increased size of the catalyst portion that can absorb hydrogen gas and feed molecules, thus making it a favorable medium for the hydrocracking reaction to proceed [42].

The hydrotreatment process of white butter into bio-jet fuel with carbon chain lengths of C₇–C₁₆ depends on the catalyst employed. An analysis of the performance of ZSM-5 and Ni/ZSM-5 catalysts in terms of selectivity is required to evaluate the ef fect of metal impregnation assisted by ultrasonic waves on the liquid bio-jet fuel products obtained. Catalyst selectivity tests were conducted based on Gas Chromatography–Mass Spectrometry (GC–MS) analysis, specifically the peak area. The compounds identified from the GC–MS analysis were classified into four categories: (i) bio-jet fuel compounds (all compounds with carbon chain lengths of C₇–C₁₆, including alkanes, alkenes, alkynes, cycloalkanes, cycloalkenes, aromatics, and isoalkanes); (ii) major bio-jet fuel compounds (alkanes, cycloalkanes, isoalkanes, and aromatics); (iii) oxygenated compounds (alcohols, esters, aldehydes, carboxylates, and ketones); and (iv) non-bio-jet fuel compounds (compounds outside the C₇–C₁₆ carbon chain range). The selectivity of the liquid products obtained from the hydrotreatment of white butter is presented in Table 6.

Based on Table 6 presents the selectivity data of liquid products from the hydrotreatment of white butter into bio-jet fuel under various reaction conditions, namely thermal (without catalyst), with ZSM-5 catalyst, and with Ni/ZSM-5 catalyst arranged in single and double configurations, as well as up to three consecutive reuses. In general, the use of catalysts enhances selectivity toward bio-jet fuel compounds compared to the thermal process. Under thermal conditions, the selectivity of bio-jet fuel compounds was only 63.22%, accompanied by relatively high contents of oxygenated compounds (20.63%) and non-bio-jet fuel compounds (16.14%). The use of ZSM-5 catalyst increased the bio-jet fuel selectivity to 65.33%; however, the oxygenated compound content also increased to 27.29%, indicating the limited capability of ZSM-5 in removing oxygen-containing groups. In contrast, the Ni/ZSM-5 catalyst demonstrated significantly superior performance, with a notable increase in bio-jet fuel selectivity across all runs, particularly in the double-stacked configuration. The single-stacked catalyst yielded selectivity values in the range of 77.66–81.72%, while the double-stacked configuration further improved selectivity up to 87.86% (double run 2). Moreover, the oxygenated compound content decreased to 5.46% in the first run of the double-stacked catalyst, confirming the effectiveness of Ni active sites in facilitating hydrodeoxygenation reactions. The non-bio-jet fuel compounds also decreased, with the lowest value of 4.15% observed in the second run of the double-stacked catalyst. The Ni/ZSM-5 catalyst in a double-stacked configuration proved to be more efficient and stable over three consecutive reuses, maintaining high performance in the third run (83.53% bio-jet fuel selectivity). These findings suggest that the layered catalyst configuration improves the distribution of active sites and enhances resistance to deactivation. Based on the obtained conversion and selectivity data, the liquid product yield was subsequently determined. The calculated yields are presented in Table 7.

Based on Table 7, catalytic product contained oxygenated compounds which meant that catalytic cracking had high activation energy for decarboxlylation of fatty acids [43]. The yield of bio-jet fuel under purely thermal conditions reached only 24.47% and increased to 35.74% when using the unmodified ZSM-5 catalyst. A significant improvement was observed with the dual-layer Ni/ZSM-5 catalyst, where the bio-jet fuel yield reached 51.71% in run 1 and 57.06% in run 2. These results represent the best performance among all conditions tested. The yield of the main bio-jet fuel fraction also achieved the highest value of 36.1% with Ni/ZSM-5 in run 2. This indicates that the incorporation of Ni and the application of a dual-layer catalyst system enhanced both conversion and selectivity toward the desired product. Furthermore, oxygenated compounds, which are undesirable by-products, showed a substantial decrease, particularly with the dual-layer Ni/ZSM-5 in run 1 and run 3, with yields of only 3.28% and 2.06%, respectively, reflecting high efficiency in the deoxygenation reaction. However, a considerable decline in yield was observed with the dual-layer Ni/ZSM-5 in run 3, where the bio-jet fuel yield dropped to only 30.58%. This suggests the possible occurrence of catalyst deactivation due to repeated use. In hydrotreatment catalysts, the end of the cycle is determined when the catalyst’s activity level declines [44].

Based on Table 8, it is evident that the use of the Ni/ZSM-5 catalyst with ultrasonic assistance in the hydrotreatment of white butter produced a bio-jet fuel yield of 51.71%. This value is significantly higher than that obtained [19], which reached only 24.34% (using the same catalyst but without ultrasonic assistance). The substantial difference in results indicates that ultrasonic activation enhances reactant diffusion, expands the catalyst’s active surface area, and accelerates the reaction rate, thereby directly improving the conversion and yield of the bio-jet fuel product

The analysis of bio-jet fuel compounds revealed that tetradecane (C₁₄H₃₀) was the dominant product. In addition to C₁₄, another predominant carbon number observed in the product distribution was C₁₅. This indicates that the hydrocracking process proceeded effectively, yielding compounds within the carbon number range characteristic of bio-jet fuel (C₇–C₁₆). The fatty acids derived from white butter undergo hydrodeoxygenation and hydrocracking reactions. Hydrodeoxygenation (HDO) to remove oxygen, whereas hydrocracking to obtain a hydrocarbon chain of the required size [50]. In contrast, decarbonylation and decarboxylation reactions can only produce hydrocarbon compounds with carbon chains shortened by one carbon atom relative to the parent fatty acid. The addition of hydrogen gas (H2) to the reactor system aims to minimize coke formation on the catalyst surface, thereby extending the catalytic life. This process produces products in the form of saturated hydrocarbon compounds and occurs under high pressure and relatively high temperature conditions. The hydrocracking reaction mechanism occurs in stages, beginning with the hydrogenation stage followed by the catalytic cracking stage.

Shorter-chain hydrocarbons may also be generated through subsequent cracking reactions. The histogram illustrating the carbon number distribution is presented in Fig. 13.

The bio-jet fuel product obtained was compared with commercial aviation turbine fuel (avtur) from PT Pertamina, in accordance with the study by [52]. The histogram of carbon atom distribution in commercial avtur shows a dominant range at C₁₀ to C₁₂, beginning at C₈. Compared to the bio-jet fuel derived from the hydrotreatment of white butter, commercial avtur tends to contain lighter carbon fractions. This difference is most likely attributed to the suboptimal cracking reaction during the hydrotreatment process, as evidenced by the higher fraction of C₁₄–C₁₅. Nevertheless, the variation is not particularly significant, indicating that the Ni/ZSM-5 catalyst is effective in producing bio-jet fuel.

The chemical composition of commercial aviation fuels, Jet A-1 (EU) and Jet A (US), consists of a mixture of alkanes (n-, cyclo-, i-), alkenes, and aromatics (benzene and naphthalene) [22]. As shown in Fig. 13, the dominant compounds in the bio-jet fuel product are alkanes (paraffins), followed by olefins. The liquid product obtained does not contain aromatic compounds but does contain isoalkanes, which play an important role in lowering the freezing point. However, a significant proportion of alkenes is still present. Alkenes are generally unstable and prone to oxidation. On the other hand, cycloalkanes offer certain advantages, as they belong to one of the principal compound groups of bio-jet fuels. The predominance of alkanes in the bio-jet fuel product can enhance the specific energy [53]. Specific energy is a key parameter in assessing aviation fuel efficiency during flight, as it determines the achievable aircraft range.

The bio-jet fuel product was subsequently subjected to ASTM testing to evaluate its compliance with aviation fuel standards. The tested sample was obtained from a dual-bed catalyst, followed by distillation. Distillation was performed to separate the bio-jet fuel fraction within the boiling point range that matches aviation fuel specifications. The American Society for Testing and Materials (ASTM) analysis revealed that the bio-jet fuel product met ASTM specifications, except for API gravity. The relatively high isoalkane content in the liquid product rendered the fraction lighter, thereby exceeding the standard API gravity limit. The freezing point of the liquid product was relatively low due to the presence of isoalkanes, as detected by GC-MS analysis. Conversely, the high olefin content poses a drawback, as it can promote oxidation leading to gum formation and fuel degradation during storage. The presence of n- and i-alkanes is more desirable, while cycloalkanes should be minimized because they can increase the freezing point of bio-jet fuel. Low viscosity is also required in aviation fuel specifications, as it facilitates faster atomization of the fuel spray [54]. Kinematic viscosity, measured according to ASTM D4052, reflects the resistance of fluid flow under gravitational force through pipelines, pumps, and injectors in aircraft fuel systems, whereas dynamic viscosity is used to evaluate the performance of fuel pumps in aircraft engines. ASTM density testing was conducted to obtain precise data on fuel mass, which is critical for calculating aircraft payload, determining the center of gravity, and planning fuel requirements during flight. A higher density within the specification range contributes to an increased aircraft flight range. The results of the ASTM analysis are summarized in Table 9.

The products derived from the dual-layer catalyst were analyzed using FTIR, as this catalyst exhibited superior performance compared to the single-layer catalyst. The liquid product was compared with the feedstock (white butter) to identify differences in functional groups after undergoing cracking and hydrodeoxygenation reactions. The FTIR spectra of the feedstock and the liquid product are presented in Fig. 15.

Based on the FT-IR spectra shown in Fig. 14, the characteristic peaks of C = O and C–O functional groups were observed to decrease after the production of bio-jet fuel. This indicates that the hydrodeoxygenation reaction during the hydrotreatment process proceeded effectively, as evidenced by the peaks at 1168 and 1743 cm⁻¹ appearing only in white butter. The peak at 1168 cm⁻¹ corresponds to the C–O functional group, while the peak at 1743 cm⁻¹ corresponds to the C = O functional group. However, a peak shift from 1743 cm⁻¹ to 1716 cm⁻¹ was observed, accompanied by a decrease in peak intensity, suggesting that the oxygen bond in C = O weakened due to the influence of nickel on the ZSM-5 material in interaction with the liquid cracking products. Furthermore, the peak at 2925 cm⁻¹, with consistently high intensity, indicates that the liquid product predominantly consists of aliphatic hydrocarbon chains, corresponding to the C₇–C₁₆ range.