Development of cobalt-based catalyst prepared from natural kaolin for production of green diesel by Fischer-Tropsch synthesis Authors: Sreypov Ke, Atthapon Srifa, Wanida Koo-Amornpattana, Suwimol Wongsakulphasatch, Suttichai Assabumrungrat, Choji Fukuhara, Sakhon Ratchahat Published: 2026 ABSTRACT AbstractFischer–Tropsch (FT) synthesis is a catalytic process that converts syngas into primarily linear hydrocarbons. The roles of active metals and supports are critical in determining catalytic performance. In this study, porous silica supports were synthesized from natural kaolin via a facile acid leaching method using different HCl concentrations (2 M, 5 M, and 8 M), denoted as SiO2-2 M, SiO2-5 M, and SiO2-8 M. Cobalt catalysts supported on the prepared silicas were evaluated for FT synthesis and compared with those on conventional supports (Al2O3, ZSM-5, and SiO2). Among the catalysts, Co/SiO2-5 M exhibited superior performance, achieving 83.23% CO conversion and 42.40 wt% selectivity toward diesel-range hydrocarbons (C12–C20) at 220 °C and 20 bar. This enhanced performance is attributed to an optimal balance between cobalt particle size and metal–support interaction, where moderate acid treatment (5 M HCl) results in favorable textural properties and improved reducibility of cobalt species without significant structural degradation. The improved catalytic behavior is associated with enhanced cobalt reducibility and moderated metal–support interaction, as confirmed by H2-TPR analysis. Overall, the results demonstrate a strong correlation between catalyst physicochemical properties, particularly pore structure, cobalt particle size, and reducibility and FT performance. Controlled acid leaching is therefore identified as an effective strategy for optimizing catalyst structure–performance relationships for diesel-range hydrocarbon production.Keywords: Natural kaolin, Silica, Green diesel, Fischer-Tropsch, Cobalt catalystSubject terms: Hydrogen energy, Catalyst synthesis, Heterogeneous catalysis, Chemical engineering, Chemical engineering Introduction Nowadays, diminishing petroleum reserves, increasing crude oil prices, stricter environmental regulations, and the growing global demand to reduce dependence on petroleum resources for producing fuels and chemicals have become major concerns1. The Fischer-Tropsch (FT) synthesis has recently been positioned as one of the promising catalytic processes for the production of clean and sustainable energy from renewable resources. In Fischer–Tropsch synthesis, the product distribution typically follows the Anderson–Schulz–Flory (ASF) mechanism, resulting in a broad spectrum of hydrocarbons with varying carbon chain lengths. These products are generally classified into light hydrocarbons (C1–C4), gasoline-range hydrocarbons (C5–C11), diesel-range hydrocarbons (C12–C20), and heavier waxes (C20+)3,3. Among these fractions, diesel-range hydrocarbons are particularly desirable due to their high energy density, low sulfur content, and compatibility with existing diesel engines, making them highly suitable for clean fuel applications4. Therefore, controlling product selectivity toward C12–C20 hydrocarbons is a key objective in the design of efficient FT catalysts. Basically, the FT synthesis is a catalytic reaction usually performed on iron- or cobalt-based catalysts5,6. However, cobalt catalysts are more appealing from an industrial perspective due to their higher hydrocarbon productivity, better stability, and commercial availability7. The FT reactions involve the strongly exothermic hydrogenation of carbon monoxide (CO) to produce paraffins and olefins, as presented by the following chemical reactions (Eqs. 1–2)8. Aside from alkanes and alkenes, oxygenated compounds and carbon dioxide (CO2) occur as byproducts in the industrial FT process. The mechanism underlying the FT reaction is complex and involves several elementary steps, such as CO dissociation, carbon (C) hydrogenation, CHx coupling reactions, hydrogenation and dehydrogenation reactions that lead to the desorption of hydrocarbon products, as well as the removal of oxygen9. Due to the diverse mixture of hydrocarbon species with different chain lengths and functional groups, the critical factors in the FT reaction are not only the catalytic activity but also the selectivity of the products5. Generally, the FT reaction can be catalyzed by cobalt (Co), iron (Fe), nickel (Ni), and ruthenium (Ru). Cobalt is a unique metal used as a component in a multi-metallic catalyst system in different kinds of chemical reactions. Moreover, the interest in the study of cobalt-based catalysts has increased due to their low cost and readily availability, compared to the traditional noble metal-based catalysts. Cobalt has a wider operating temperature range (200 –350 °C) and typically produces more olefins and oxygenates, especially at the higher end of the temperature range (320 –350 °C)10,11. Supported cobalt catalysts are also preferred for FT synthesis when high conversion, longer lifetimes, and higher selectivity to paraffinic products are needed. To increase activity, cobalt is usually deposited on a high surface area oxide such as alumina or silica supports to obtain high metal dispersion in the chain growth condition. C5+ hydrocarbon selectivity is also related to the diffusion of reactants and products through the pores of the catalyst support12,13. It is generally accepted that the support property has a crucial effect on the Co-based catalysts for enhancing FT performance14. The textural characteristics of the support, including surface area, pore structure, and surface chemistry, also relate to dispersion of cobalt metal particles. For instance, supports with high surface areas tend to disperse cobalt particles efficiently and are likely to produce highly active Fischer-Tropsch catalysts15. Silica possesses several suitable physical and chemical properties to serve as a perfect support. The nature of hydroxyl groups (silanols) and their concentration and distribution on the silica surface play an important role in cobalt particle dispersion. The high surface area of silica favors high metal dispersion at reasonably high metal loadings16. However, Iglesia et al.17 showed that Co site-time yield is independent of Co particle size between 10 and 210 nm using aluminum oxide (Al2O3), silicon dioxide (SiO2), and titanium dioxide (TiO2) as supports. Bezemer et al.18 reported that CO hydrogenation activity is insensitive when the (carbon-nanofiber-supported) Co nanoparticles are larger than 6 nm but demonstrated that smaller Co nanoparticles (< 6 nm) have lower CO hydrogenation activity and higher CH4 selectivity. The metal crystallite size and degree of cobalt reduction increased with an increasing pore diameter of silica support. Furthermore, the C5+ and methane selectivity exhibited a maximum and minimum, respectively, for FT synthesis (20 wt% Co, 200–220 °C, 15 bar, H2/CO = 2) for cobalt catalyst supported on a silica support with pore diameter of 10 nm19. Therefore, this research aims to synthesize a high surface area silica from kaolin clay by a simple method. Although cobalt-based catalysts for Fischer–Tropsch synthesis are well established, the novelty of this work lies in the development of silica supports derived from natural kaolin with tunable physicochemical properties via controlled acid leaching. Unlike conventional silica supports, the kaolin-derived silica allows systematic adjustment of surface area, porosity, and metal–support interaction through the variation of HCl concentration. Silica synthesis requires active kaolin, known as metakaolin, to provide the Si-O and Al-O active sites, which are the sources of silica and alumina20,21. The synthesized silica supports were made of silica derived from natural kaolin, with variations based on different HCl concentrations (SiO2-2 M, SiO2-5 M, and SiO2-8 M). The physicochemical properties of the catalysts were investigated using various techniques, providing deeper insights into their structure and overall characteristics. In addition, the cobalt catalysts supported on commercial supports including alumina (Al2O3), silica (SiO2), and zeolite (ZSM-5) were prepared and tested in FT synthesis for comparison. Experimental Catalyst preparation The Co/SiO2 catalyst was synthesized by impregnation method. Firstly, the silica as a catalyst support was prepared by a template-free method using kaolin as a silica precursor. Kaolin clay was locally provided by Mineral Resources Development Co., Ltd., Ranong Province, Thailand. The as-received kaolin was activated by calcination in air at 850 °C for 6 h (10 °C /min) to obtain metakaolin (MK). Then, 10 g of MK was added into 200 ml of 4 M NaOH aqueous solution in a sealed Teflon vessel and stirred for 30 min at 80 °C to remove impurities. After vacuum filtration, the solid sample was repeatedly washed with deionized water on the filter until the pH of the filtrate reached neutral (pH ≈ 7), ensuring the removal of residual alkali species and then dried at 110 °C overnight, to obtain an alkali-treated metakaolin (AMK). Finally, 10 g of AMK was etched with 200 mL of HCl aqueous solution (2, 5, or 8 M) at 80 °C for 6 h in a sealed Teflon vessel. Then, the etched AMK was filtered, washed with deionized water, and dried at 110 °C overnight to obtain the final silica (SiO2). The concentration of HCl solution was varied at 2, 5, and 8 M to produce different SiO2 samples. The synthesized silicas are abbreviated as SiO2-2 M, SiO2-5 M, and SiO2-8 M, respectively. The cobalt supported on the as-synthesized silica was prepared by impregnation method. Briefly, 9.8 g of cobalt nitrate hexahydrate (Co(NO3)2⋅6H2O, 98% purity, Carlo Erba) was dissolved in 50 ml of deionized water at room temperature to obtain a clear pink solution. The solution was then mixed with 8 g of the as-prepared catalyst support under continuous stirring for 2 h at room temperature to ensure uniform dispersion of the cobalt precursor. The resulting slurry was then evaporated in an evaporating dish at 90 °C for approximately 1 h until it became dry. The obtained powder was calcined in air at ramp rate of 1 °C/min to 400 °C and held for 6 h. Furthermore, the cobalt catalysts supported on different commercial supports were also prepared with the same impregnation method. The commercial supports include γ-Al2O3 (210 m2/g, Sasol Germany), as confirmed by its high surface area and characteristic broad XRD features, along with SiO2 (200 m2/g, fumed silica, Sigma-Aldrich) and ZSM-5 (425 m2/g, SiO2/Al2O3= 80/1, Alfa Aesar). A granular form of catalyst was made before packing in the reactor to avoid a plugging or high pressure drop during the reaction testing. The cobalt catalyst powder was pressed at 40 MPa for 10 min to form a tablet, and the tablet was cracked into granules with a size range of 0.85–1.18 mm. The obtained granular catalyst was stored in a closed bottle for testing. The cobalt loading was fixed at 20 wt%, which is commonly reported as an optimal range for Fischer–Tropsch catalysts22,23. Catalytic performance test by FT synthesis The catalytic performance of the as-prepared catalysts was tested by the FT reaction in a vertical fixed-bed reactor made of stainless steel (i.d.1/2″, 49 cm effective length, Swagelok SS316). The FT reaction system used in this study is schematically shown in Fig. 1. In each run, 5 g of catalyst granules (0.85–1.18 mm) was mixed with silicon carbide (SiC, 0.85–1.18 mm) at ratio of 1:1 (v/v). Before use, SiC was calcined in air at 800 °C for 2 h to remove impurities. To keep the catalyst bed in place, it was loaded between layers of quartz wool inside the reactor. For the activation process, the catalyst was reduced by hydrogen (100 mL/min, 1 atm) at a ramp rate of 2 °C/min from room temperature to 400 °C, held for 3 h, and then cooled down to room temperature. For the FT run, a mixture of CO and H2 was employed as a model syngas (99.999% purity, UHP, moisture ≤ 5 ppm, United Industrial Gas Co., Ltd.) with CO/H2 mol ratio of 1/2. The feed was introduced to the reactor at flow rate of 100 mL/min and pressure of 20 bar. A leak test of the reactor system was performed at 20 bar. To begin the FT reaction, the reactor was then heated to the FT reaction temperature of 220 °C at ramp rate of 2 °C/min. As the reactor reached the reaction temperature, the reaction time for FT synthesis was counted and set for 8 h for each run. The initial composition of syngas at the inlet was taken and analyzed at least 3 times before starting the FT reaction. The gas products were analyzed time-on-stream by gas chromatography (TCD, GC-2014, Shimadzu). Meanwhile, heavier liquid products including liquid hydrocarbons and water byproducts were collected from the condenser tank at the bottom of the reactor after completing a run for the 8 h run. The oil phase was separated from water byproducts. The composition of oil products was analyzed by gas chromatography (FID, GC-2014, Shimadzu). Schematic diagram of the Fischer–Tropsch synthesis experimental setup. Feed gases (H2 and syngas) are controlled by mass flow controllers and introduced into a vertical fixed-bed reactor. The reactor pressure is regulated by pressure gauges and a back-pressure regulator. The effluent passes through a condenser for gas–liquid separation, with gaseous products analyzed online by GC-TCD. A soap film meter was used to measure the outlet gas flow rate Product analysis The CO conversion and hydrocarbon selectivity were measured at the steady state condition and calculated by using Eqs. 3–6. The average values of all 8-hour operations were reported. The selectivity is divided into two categories; the molar selectivity to each product based on carbon number; the mass selectivity of gasoline range (C5−11) and diesel range (C12−20) in the oil phase. Equation (4) was used to calculate molar selectivity to CO2 and C1-C4 species. Fci are the molar flowrates of species Ci. ni is the number of carbon atoms in each species. Equation (5) was used to calculate the C5+ selectivity. Equation (6) was used to analyze the quantity of the diesel range products due to the impracticality of direct quantification using traditional calibrated peak area procedure5. Oil phase product selectivity in this work was based on the corresponding alkane peak area from C8-C20 calibration that performed C8-C20 alkane standard solution (Sigma-Aldrich). This was calculated as the percentage of peak area of diesel range products to the total peak area obtained from GC-FID chromatogram. Catalyst characterization The catalyst samples for analysis wrer reduced to provide the actual active state of catalyst. The elemental composition of raw materials, supports and catalyst material was determined by X-ray fluorescence (XRF, S8 BRUKER TIGER). The phase composition and crystallite size of the catalysts were analyzed using Bruker AXS D2 Phaser X-ray powder diffractometer of CuKα1 excitation source (λ = 0.154060 nm) with filter working at 30 kV and 10 mA. The XRD pattern was collected in 2θ = 10–90° with scanning speed of 1.4°/min and a step size of 0.020265. Field-emission scanning electron microscopy (FE-SEM) was conducted using a JEOL JSM-IT700HR equipped with a field-emission electron gun, enabling high-resolution imaging at an accelerating voltage of 30 kV and a maximum probe current of 300 nA. High-resolution transmission electron microscopy (HR-TEM JEOL, JEM-2100 Plus, Japan) operating at an accelerating voltage of 200 kV was employed on selected catalysts to identify the detailed microstructure of the catalysts. Prior to TEM characterization, a small amount of the sample was finely ground, dispersed in ethanol, and then subsequently deposited onto a copper grid. The textural properties of catalysts including the isotherm type, BET surface area, pore volume, average pore size and pore size distributions were examined by N2 sorption measurement at -196 °C using Micromeritics TriStar II Plus analyzer. Barrett-Joyner-Halenda (BJH) technique in the desorption curve was used to calculate average pore volume and pore diameter, whereas the BET method was used to determine the surface area. For H2-TPR analysis (BELCAT-B, BEL, Japan), the samples were first pretreated in a flowing pure He gas (50 mL/min) for 60 min at 150 °C for the removal of physiosorbed species. After the sample was cooled to 50 °C, a flow of 5% H2/Ar (30 mL/min) was introduced. The temperature was then ramped from 50 °C to 1000 °C with a heating rate of 10 °C/min for the temperature programmed reduction process. The TCD signal was then calibrated, and the H2 consumption in mmol/g was reported. Results and discussion Cobalt catalysts supported on different commercial supports Textural properties The textural properties of the as-prepared catalysts were determined by N2 adsorption desorption, and the results are summarized in Table 1. The cobalt supported on alumina and ZSM-5 catalysts exhibited small pore diameters of 7.4 nm and 3.7 nm, respectively. In contrast, the cobalt supported on silica catalyst displayed a much larger pore diameter of 18.9 nm which resulted in the smallest BET surface area of 129.8 m2/g. This suggests that the silica-supported cobalt catalyst possesses a larger pore structure, allowing the incorporation of cobalt metal particles into the pore structure. This characteristic has a positive impact on the FT catalytic activity, as shown by its performance in terms of selectivity to heavy hydrocarbons and CO conversion. Basically, large particle sizes and large pore diameters can improve cobalt reducibility during the catalyst activation process via hydrogen reduction. The catalyst textural properties, such as specific surface area and pore diameter, strongly impact its activity3. Table 1Textural properties of the as-prepared cobalt catalysts supported on different commercial supports.CatalystSBET (m2/g)Pore volume (cm3/g)Pore diameter (nm)Co/Al2O3148.50.3317.4Co/ZSM-5309.40.4623.7Co/SiO2129.80.54418.9 Textural properties of the as-prepared cobalt catalysts supported on different commercial supports. Crystallinity Figure 2 shows XRD patterns of the reduced cobalt catalysts supported on different commercial supports including Al2O3, ZSM-5 and SiO2. The characteristic diffraction peaks at 2θ = 44.32°, 50.07° and 75.79° could be assigned to metallic cobalt (Co0) (JCPDS 15–0806). Peaks corresponding to Co0 were detected in all reduced catalysts, indicating that the catalysts were reduced from oxide form to metallic form as active phase in FT synthesis. However, the presence of Co3O4 was also observed. The peaks observed at 2θ = 37.29°, 59.60° and 68.66° were in accordance with the different crystal planes of Co3O4 (JCPDS 42-1467). This was also observed in the other catalysts such as Co/Al2O3 and Co/ZSM-5. This suggests that the catalysts were not completely reduced at 400 °C. Diffraction peaks at 2θ = 9.55° and 23.88° are attributed to the ZSM-5 crystal structure (MFI-type, JCPDS no. 44 − 0003)24, which is consistent with previously reported diffraction patterns of ZSM-5. XRD patterns of reduced cobalt catalysts supported on different commercial supports The crystallite size of metallic cobalt (Co0) for all the catalysts was calculated from XRD using the Scherrer equation using full width at half maximum (FWHM) of the most intense diffraction peak at 2θ ≈ 44.3°, corresponding to the (111) plane of metallic cobalt and the results are listed in Table 2. The Co/Al2O3 catalyst shows the largest Co0 crystallite size (12.4 nm), while the Co/ZSM-5 catalyst has the smallest size (10.9 nm). Table 2Crystallite size and cobalt content in the as-prepared cobalt catalysts supported on different commercial supports.CatalystCo content a (wt%)Co0 crystallite sizeb (nm)Co/Al2O327.6812.4Co/ZSM-526.6810.9Co/SiO225.8011.2a Determined by ICP analysis.b Calculated from XRD using the Scherrer equation of 2θ ≈ 44.3°. Crystallite size and cobalt content in the as-prepared cobalt catalysts supported on different commercial supports. a Determined by ICP analysis. b Calculated from XRD using the Scherrer equation of 2θ ≈ 44.3°. Reducibility The reducibility of all the fresh calcined catalysts used in this study was determined by H2-TPR and the results are shown in Fig. 3. It shows a different reduction profile for all three catalysts. Two distinct reduction peaks were observed: a low-temperature peak (150–400 °C) and a high-temperature peak (400–800 °C). The low-temperature peak was assigned to the reduction of cobalt (II, III) oxide, which occurs in two stages (Co3+ Co2+ Co0)6,25,26. The reduction of the Co/SiO2-2 M catalyst occurs in two main stages. Firstly, the Co/SiO2 catalyst displays a TPR peak at a low temperature of 325 °C, followed by a second peak at 390 °C, indicating complete reduction from Co3O4 to Co0 below 400 °C. A small third reduction peak is observed at approximately 415 °C for the Co/SiO2 catalyst. This feature can be attributed to the reduction of cobalt silicate-like species formed via interaction between cobalt oxide and the silica support, likely involving CoO and surface Si–OH groups6. The relatively moderate reduction temperature suggests a moderate metal–support interaction compared to strongly interacting systems such as Co/Al2O3. In contrast, the TPR profile of the Co/Al2O3 catalyst exhibits a very low first reduction peak at 240 °C and a broad second reduction peak ranging from 330 °C to 450 °C. Additionally, a strong third peak is observed at a higher temperature around 700 °C, suggesting that the cobalt supported on alumina is not completely reduced at 400 °C using H2. Co/ZSM-5 also shows a high-temperature peak (~ 692 °C), suggesting relatively strong interaction, although slightly weaker than of Co/Al2O3. Thus, at the reaction temperature (400 °C), Co/SiO2 is more readily reducible than Co/Al2O3 and Co/ZSM-5. H2-TPR profiles of the as-prepared cobalt catalysts on different commercial supports Morphology Figure 4 shows SEM mapping of the morphology and elemental distributions on catalyst granules. The results for Co/Al2O3, Co/SiO2, and Co/ZSM-5 catalysts were presented. On the Co/Al2O3 catalyst granule, cobalt dispersion was non-uniform in certain areas (Fig. 4a). For the Co/SiO2 catalyst, elemental Si was consistently distributed across the substrates (Fig. 4b), with cobalt found inside the mesopores and showing low signals on the support surface. However, for the Co/ZSM-5 catalyst, the elemental maps revealed uneven elemental distribution (Fig. 4c). SEM mapping of the reduced catalysts: (a) Co/Al2O3, (b) Co/SiO2, and (c) Co/ZSM-5 Cobalt catalysts supported on silica prepared from natural kaolin Composition and crystallinity of catalyst The chemical compositions of kaolin and synthesized silicas were analyzed using XRF instrument, and the obtained results are shown in Table 3. It was determined that kaolin contains SiO2 and Al2O3 at approximately 57.9 wt% and 27.5 wt%, respectively. The impurities with a total content of around 3 wt% consists of TiO2, Fe2O3, K2O, CaO, and MgO. The presence of iron oxide (Fe2O3) is responsible for the typical orange color observed in natural kaolin. Elemental analysis of all obtained silicas reveal that SiO2-8 M has the highest SiO2 percentage of 87.90 wt%. The acid treatment could reduce alumina content by leaching of Al3+ ions from an octahedral layer under acidic condition27. Moreover, it was observed that even at a low acid strength, the fractions of Fe2O3, CaO, K2O, and Na2O decreased, while the content of TiO2 slightly increased. When the acid strength was 2 M, the aluminum content in the synthesized silica decreased considerably. This can be attributed to the shielding of aluminum cations by silicon-oxygen networks where the presence of Al-O-Si bonds results in low solubility at low acid strength27. Table 3Elemental composition of kaolin and synthesized silicas with various acid concentrations from XRF analysis.CompositionPercent composition (wt%)KaolinSiO2-2 MSiO2-5 MSiO2-8 MCo/SiO2-2 MCo/SiO2-5 MCo/SiO2-8 MCoO––––23.623.023.9SiO257.9077.386.687.9060.865.265.6Al2O327.5012.83.343.507.884.042.56TiO21.462.103.202.311.791.961.85Fe2O30.810.6840.560.3520.8250.9210.579K2O0.370.4060.250.2780.3240.20.206CaO0.21––––0.122–MgO0.17––––––Na2O0.14–––––– Elemental composition of kaolin and synthesized silicas with various acid concentrations from XRF analysis. Figure 5 shows the XRD patterns of kaolin and metakaolin. The patterns clearly demonstrate that kaolin primarily consists of kaolinite, along with some minor impurities. However, after undergoing calcination, the kaolinite phase was significantly reduced. This indicates that thermal treatment effectively disrupts the crystalline structure of kaolinite. The characteristic diffraction peaks at 2θ of 12.3º, 20.2º, 24.8º, and 36.9º indicated the presence of kaolinite in the kaolin sample28,29. The diffraction peak at 26.6° indicated the presence of quartz, which is a highly crystalline silica (JCPDS 3-0490). It could be observed that the kaolinite peaks had generally lower intensities compared to the quartz peak. This implied that the raw kaolin inherently contained a high concentration of quartz30. On the other hand, the XRD patterns of metakaolin indicated a highly amorphous material. In addition, the peak at 2θ = 12.3º present in the raw kaolin disappeared due to collapse of the kaolinite structure caused by the calcination effect. The kaolinite and mica peaks were apparently lost in the metakaolin due to the presence of quartz. Meanwhile, the quartz, being crystalline silica, remained structurally stable after the calcination process. Past work has shown that crystalline silica is inert to calcination treatment even at 1050 °C31. XRD patterns of as-received kaolin and metakaolin after calcination Figure 6 shows XRD patterns of the as-synthesized silicas. It can be observed that the materials were amorphous, and the only obvious crystalline peak appeared at an angle of 26.6º. This peak was attributed to the presence of residual quartz. The other features were the background peaks due to the amorphous silica30. A study by Li et al2 also stated that synthesized silica mainly consists of silica and kaolinite, while traces of stable quartz impurity are also observed. It can be observed that the quartz remained after synthesis, surviving in the subsequent alkali and acid treatments. Therefore, it is evident that the thermal and subsequent acid treatments destroyed all the crystalline kaolinite phases. XRD patterns of synthesized silicas at various acid concentrations Figure 7 shows XRD patterns of the reduced catalysts. The characteristic diffraction peaks at 2θ = 44.32°, 50.07°, and 75.79° could be assigned to metallic cobalt (Co0) (JCPDS 15–0806). The peaks at 2θ = 36.48°, 42.39°, and 73.41° were assigned to CoO (JCPDS 43-1004) and might remain unreduced after the reduction treatment at 400 °C with hydrogen. The impurity, such as quartz in kaolin, was still observed at the peak of 26.6°. XRD patterns of reduced cobalt catalysts supported on the silica synthesized from kaolin with various acid concentrations The crystallite size of Co0 was estimated by the Scherrer equation using full width at half maximum (FWHM) of a prominent XRD peak. The metallic Co crystalline size of the catalyst prepared with a lower acid concentration of 2 M (Co/SiO2-2 M) was the smallest at 24.7 nm, while the 5 M acid concentration one shows the largest crystalline size of metallic Co at 29.2 nm. Meanwhile, the Co/SiO2-8 M results in a cobalt metallic crystalline size of 27.2 nm. It can be concluded that the SiO2 prepared by 5 M acid concentration enhances the formation of larger crystallite size of cobalt during impregnation. Textural properties and reducibility of catalyst The obtained products from acid leaching are predominantly silica with an amorphous structure, as confirmed by XRF (shown in Table 3) and XRD (shown in Fig. 7). The N2 adsorption-desorption isotherms and pore size distribution of cobalt catalyst supported on silica synthesized from kaolin with different HCl concentrations are shown in Fig. 8. All catalysts exhibit isotherm profiles of a Type I isotherm with an H3-type hysteresis according to the IUPAC classification, indicating a microporous structure of the materials. The hysteresis loops are observed at intermediate to high relative pressures (P/P₀ ≈ 0.4–1.0). The H3-type hysteresis loop is typically associated with slit-shaped pores formed by the aggregation of plate-like particles, which is consistent with the nature of silica derived from kaolin clay. This suggests that the prepared catalysts possess a hierarchical pore structure. N2 adsorption-desorption isotherms (a) and BJH pore size distributions (b) of the reduced cobalt catalysts supported on the synthesized silica at various acid concentrations. The specific surface area and porosity properties of the as-prepared catalysts are summarized in Table 4. The pore size of all catalysts is less than 2 nm, indicating the presence of a microporous structure that correlates with the isotherms as shown in Fig. 8. The cobalt catalysts show a lower specific surface area than the silica supports due to pore blockage by cobalt species during the catalyst impregnation process. Moreover, the specific surface areas of Co/SiO2-2 M, Co/SiO2-5 M, and Co/SiO2-8 M catalysts are 106.8, 122.1, and 119.6 m2/g, respectively, which correlate with the specific surface area of the synthesized silica supports. The acid concentration during the silica synthesis process has a significant effect on the specific surface area and pore volume, which would have a corresponding effect on the catalytic performance. The decrease in specific surface area observed after cobalt loading is attributed to the partial blockage of pores and the coverage of the support surface by cobalt oxide species during the impregnation and calcination processes, which reduces the accessible surface area and pore volume. Similar behavior has been widely reported for supported metal catalysts prepared via impregnation methods27,32. Furthermore, the increase in surface area with increasing HCl concentration is associated with enhanced acid leaching of aluminum species (dealumination) and removal of impurities from the kaolin structure. This process promotes the formation of a more porous silica framework and increases the available surface area. The extraction of Al3⁺ ions from the octahedral layers leads to structural disruption and generation of additional porosity27,32. Table 4Summary of specific surface area and porosity properties of kaolin-derived SiO2 supports with different acid concentrations and corresponding Co/SiO2 catalysts.SampleSBET (m2/g)Pore volume (cm3/g)Pore diameter (nm)SiO2-2 M162.90.17601.2SiO2-5 M224.20.43601.1SiO2-8 M209.70.28001.3Co/SiO2-2 M106.80.16501.5Co/SiO2-5 M122.10.18201.2Co/SiO2-8 M119.60.15301.1 Summary of specific surface area and porosity properties of kaolin-derived SiO2 supports with different acid concentrations and corresponding Co/SiO2 catalysts. H2-TPR profiles of the calcined cobalt catalysts supported on silica synthesized from kaolin with different HCl concentrations are shown in Fig. 9. The Co/SiO2-2 M exhibits the two main reduction peaks. The first peak occurs in the temperature range of 150–400 °C, and the second peak is a broad peak in the temperature range of 400–800 °C. This is attributed to the reduction step of Co3O4 to CoO, and then to Co metal3,4,25. When the acid concentration is increased, both reduction peaks shift to lower reduction temperatures due to the weaker interaction between Co metal and SiO2 support. The second peak exhibits a particularly dramatic change. This suggests that increasing the acid concentration decreases the interaction between metal and support. Co/SiO2-2 M shows reduction peaks at higher temperatures (528 and 630 °C), indicating stronger MSI, which limits reduction at 400 °C. Co/SiO2-5 M exhibits lower reduction temperatures (340, 361, and 480 °C), suggesting weakened MSI and improved reducibility, making it more favorable under reaction conditions. Co/SiO2-8 M shows slightly lower temperatures than 5 M (346, 388, and 464 °C), indicating moderate MSI, but still better reducibility than 2 M. Regarding hydrogen consumption (Table 5), all acid-treated catalysts show similar total H2 uptake (~ 4.2–4.5 µmol g⁻1), indicating comparable cobalt loading. However, the distribution of reduction temperatures differs, confirming that HCl treatment primarily tunes MSI rather than total reducible cobalt content. At 400 °C, Co/SiO2-5 M demonstrates the most favorable reducibility, as a larger fraction of cobalt species can be reduced below or near the reaction temperature. H2-TPR of fresh cobalt-based catalysts supported on the silicas synthesized from kaolin with different acid concentrations Table 5Reduction temperatures and corresponding H2 consumption of Co-based catalysts determined from H2-TPR analysis.SampleReduction Temp (oC)/ H2 consumption(µmol/g)Tempµmol/gTempµmol/gTempµmol/gTotal µmol/gCo-Al2O32820.0274010.7407101.3502.117Co-SiO23290.9073772.6754150.6384.220Co-ZSM-53892.1884451.7166920.5224.426Co-SiO2-2 M3551.1025282.3016301.0994.502Co-SiO2-5 M3401.1603611.7744801.5134.447Co-SiO2-8 M3461.2143881.7554641.2394.208 Reduction temperatures and corresponding H2 consumption of Co-based catalysts determined from H2-TPR analysis. The XRD results indicate incomplete reduction at 400 °C, as evidenced by the presence of residual Co3O4/CoO phases. Therefore, the degree of reduction at 400 °C (DR400) was calculated from H2-TPR data and included in Table 6. The results show a strong effect from both the choice of support and the acid leaching process. Acid leaching of the kaolin-derived silica support also significantly influences reducibility. Mild acid treatment (e.g., 2 M) leads to a stronger metal–support interaction, resulting in lower reducibility. Higher acid concentrations (5 M and 8 M) further improve cobalt dispersion and enhance reducibility. Since metallic Co0 is the active phase for Fischer–Tropsch synthesis, incomplete reduction may limit the number of active sites, particularly for catalysts with strong metal–support interaction (e.g., Co/Al2O3). However, reduction at 400 °C was selected to balance sufficient activation with the prevention of cobalt particle sintering at higher temperatures. The high-temperature peak (~ 700 °C) observed for Co/Al2O3 is attributed to the reduction of strongly interacting cobalt species, which are well known to reduce only at elevated temperatures (> 600 °C). This further confirms the strong metal–support interaction in this catalyst. Table 6Cobalt loading, hydrogen consumption, and degree of reduction of supported catalysts derived from H₂-TPR analysis.SampleCo (wt%)Co (µmol/g)H2 ≤ 400 °C (µmol/g)Total H2 (µmol/g)DR400a (%)Co/Al2O327.684700.7672.11736.2Co/ZSM-526.684532.1884.42649.4Co/SiO225.804383.5824.22084.9Co–SiO2–2 M18.53142.3014.50251.1Co–SiO2–5 M18.13072.9344.44766.0Co–SiO2–8 M18.83192.9694.20870.6a degree of reduction at reduction temperature during reaction (400 °C). Cobalt loading, hydrogen consumption, and degree of reduction of supported catalysts derived from H₂-TPR analysis. a degree of reduction at reduction temperature during reaction (400 °C). Surface morphology and internal structure of catalyst Figure 10 presents SEM images along with their corresponding elemental mapping of the reduced catalysts. The dispersion of elements, including cobalt (Co) and silicon (Si), is mapped as red and green, respectively. It was observed that the particle size and surface morphology of the catalysts remain unchanged, as evident in the SEM images. SEM mappings of reduced catalysts: (a) Co/SiO2-2 M, (b) Co/SiO2-5 M, (c) Co/SiO2-8 M. To provide further insight, both reduced and spent catalysts were characterized using TEM analysis, as shown in Fig. 11. The images show Co particles dispersed on the surface of synthesized silica support. Additionally, increasing the acid concentration resulted in significant Co agglomeration and sintering after reduction at 400 °C under pure hydrogen (H2), even though the silica support is prepared using the hydrothermal method. The reduced Co/SiO2-8 M catalyst exhibits high cobalt crystal agglomeration with silica, forming clusters with a size of approximately 85.2 nm. The cobalt agglomeration observed in the Co/SiO2-8 M catalyst can be attributed to excessive acid leaching during support synthesis. A high HCl concentration leads to extensive dealumination and partial structural degradation of the silica framework, resulting in a reduction in the number of anchoring sites for cobalt species. Consequently, cobalt particles tend to migrate and aggregate during impregnation and reduction, forming larger crystallites. This agglomeration negatively affects catalyst performance by reducing metal dispersion and the number of accessible active sites. Although larger cobalt particles may enhance reducibility, excessive particle growth can limit effective metal utilization and lead to less favorable catalytic performance compared to the Co/SiO2-5 M catalyst. TEM images and corresponding particle size distributions of cobalt catalysts (a) Co/SiO2-2 M, (b) Co/SiO2-5 M, (c) Co/SiO2-8 M. (d–f) spent catalysts after Fischer–Tropsch reaction (d) Co/SiO2-2 M; (e) Co/SiO2-5 M; (f) Co/SiO2-8 M. The histograms show the cobalt particle size distribution with average particle size (Dav) and standard deviation (SD). The particle sizes of the spent catalyst are smaller than those of the reduced catalyst, except for the spent Co/SiO2-5 M catalyst, whose particle size remained largely unchanged. This result indicates that the Co dispersed on SiO2-5 M is stable. During the pressurized syngas (CO/H2) reaction, residual cobalt oxide was not completely reduced in the reduction step may undergo reduction and reconstruction, resulting in smaller particles. The supported Co nanoparticles can distribute on the support during FTS, a phenomenon that can be explained by the presence of isolated Co atoms or smaller Co nanoparticles33. Catalytic performance of the prepared cobalt catalysts The catalytic testing of all catalysts was carried out in FT synthesis at 220 °C for 8 h. The CO conversion and hydrocarbon selectivity are shown in Table 7. For the cobalt on commercial supports, Co/SiO2 exhibits the highest CO conversion of 82.82%, while Co/Al2O3 and Co/ZSM-5 show lower values at 76.40% and 77%, respectively. The SiO2-supported cobalt catalyst also demonstrates the highest C5+ selectivity and product distribution to C12-C20, reaching 93.41% and 29.89%, respectively. Moreover, the SiO2-supported cobalt catalyst produced higher amounts of oil products as well as selectivity to the diesel range. For cobalt on silica derived from kaolin with different concentrations of acid treatment, it was found that the CO conversion of Co/SiO2-5 M was higher than those of Co/SiO2-2 M and Co/SiO2-8 M catalysts. Moreover, there is a slightly higher hydrocarbon selectivity toward liquid products, including gasoline (C5-C11) and diesel (C12-C20), following the Anderson-Schulz-Flory (ASF) distribution, with lower methane selectivity. The chain growth probability (α) of Co/SiO2 catalysts prepared from silica derived from kaolin is found to be in a relatively high range (0.90–0.93) for all catalysts. Among them, Co/SiO2-8 M exhibits the highest α value (α = 0.9269), followed by Co/SiO2-5 M (α = 0.9189) and Co/SiO2-2 M (α = 0.9043). The high α values indicate a strong tendency for chain growth, resulting in an increased formation of long-chain hydrocarbons. This is consistent with the experimental results, where an increase in α corresponds to a decrease in gas products, while the oil product yield increases. In addition, the CH4 selectivity decreases, whereas the C5+ selectivity increases, indicating a shift in product distribution from short-chain to long-chain hydrocarbons. Furthermore, the increasing trend of α with increasing acid concentration suggests that the structure of the support plays an important role in promoting hydrocarbon chain growth during Fischer-Tropsch synthesis. Table 7Catalytic performance of cobalt-based catalysts on different commercial support materials and kaolin derived SiO2 supports.CatalystXCO (%)cProduct selectivity (%)Alkane distribution (wt%)CH4CO2C2H6C5+C5-C11C12-C20aCo/Al2O376.4022.143.720.1573.99nd.nd.aCo/ZSM-577.007.632.800.7587.5775.3224.67aCo/SiO282.824.461.850.2893.4172.5229.89bCo/SiO2-2 M80.773.601.260.3395.1472.3527.64bCo/SiO2-5 M83.232.911.160.2595.9357.5942.40bCo/SiO2-8 M83.182.280.510.1797.0164.6635.33a Co/commercial support (Al2O3, ZSM-5, and SiO2).bCo/SiO2 derived from kaolin with different concentrations of HCl (2 M, 5 M, and 8 M).cThe standard deviation (SD.) of CO conversion for Co/SiO2 = 2.48%, and for Co/SiO2-5 M = 0.84%. Catalytic performance of cobalt-based catalysts on different commercial support materials and kaolin derived SiO2 supports. a Co/commercial support (Al2O3, ZSM-5, and SiO2). bCo/SiO2 derived from kaolin with different concentrations of HCl (2 M, 5 M, and 8 M). cThe standard deviation (SD.) of CO conversion for Co/SiO2 = 2.48%, and for Co/SiO2-5 M = 0.84%. Regarding the selectivity for diesel-range hydrocarbon of cobalt-based catalysts, many factors could influence the catalytic performance in Fischer-Tropsch (FT) synthesis. These factors include pore size diameter, specific surface area, and reducibility. Based on previous results, the Co/SiO2−5 M catalyst exhibits larger cobalt particles and higher reducibility, offering a greater number of active sites on its surface. This promotes higher hydrocarbon selectivity in the diesel range. In comparison to Co/SiO2-2 M and Co/SiO2-8 M catalysts, the Co/SiO2-5 M catalyst demonstrates a better activity for producing hydrocarbons. This result agrees with Bezemer et al., who reported that the selectivity toward C5+ hydrocarbons increased, as the size of cobalt particles increases14. Catalytic and characterization results show that the support porosity strongly influenced the structure, reducibility, and FT catalytic performance of silica-supported cobalt catalysts. Larger pore sizes can cause the Co/SiO2 to form larger Co3O4 crystallites, while the dispersion of catalysts decreases with increasing pore size. Smaller pores lead to smaller Co0 crystallites and lower reducibility. As the pore size of the support increases, the interaction of cobalt oxide species with the support weakens. Catalysts with different pore sizes exhibit distinct CO adsorption properties. In addition, the superior performance of Co/SiO2-5 M can be further explained by surface chemical properties of the support. The presence of silanol (Si–OH) groups on the silica surface provides anchoring sites for cobalt precursor species, promoting uniform dispersion34. The optimal acid treatment (5 M HCl) is expected to provide a suitable density of silanol groups, whereas insufficient or excessive acid treatment may lead to limited surface functionality or structural degradation. Moreover, residual alumina content, as observed from XRF analysis, may contribute to improved cobalt anchoring sites11. A moderate amount of residual Al in Co/SiO2-5 M enhances dispersion and stability, while excessive dealumination at higher acid concentration reduces anchoring sites and promotes cobalt agglomeration. These effects result in an optimal balance of dispersion, particle size, and reducibility, leading to enhanced catalytic performance. Reaction conditions: T = 220 °C, P = 20 bar, H2/CO = 2, total flow rate = 100 mL min⁻1, catalyst weight = 5 g, GHSV = 1200 mL/g⋅h, reaction time = 8 h. All catalysts were catalytic testing in FT synthesis at 220 °C for 8 h, and the results are shown in Table 7. Among them, Co/SiO2 exhibited the highest CO conversion (82.82%), while Co/Al2O3 showed the lowest (76.40%). Additionally, Co/SiO2 had the highest C5+ selectivity and product distribution to C12-C20 (29.89 wt% and 24.67 wt%, respectively). Although the reaction times for these catalysts were similar, their CO conversion values differed, likely due to the formation of cobalt silicates leading to a reduction in catalytic FTS. Particle size of cobalt also affected catalytic activity. Co/SiO2 and Co/ZSM-5 had higher C5-C11 selectivity, but Co/SiO2 outperformed in hydrocarbon chain growth, producing the highest amounts of C12-C20. This suggests that the characteristics of the silica surface can significantly influence catalyst performance, thanks to its high surface area and narrow pore size that aid in metal dispersion and catalytic activities. The scalability and industrial relevance of kaolin-derived silica are significant due to the abundance and low cost of kaolin as a natural resource. The synthesis approach used in this study, based on acid leaching, is simple, scalable, and does not require expensive templating agents, offering a cost advantage over conventional silica supports. From an industrial perspective, kaolin-derived silica shows strong potential for Fischer–Tropsch catalysts due to its low cost, scalability, and tunable textural properties. However, further evaluation under industrial conditions is required to confirm its long-term stability and applicability. In addition, the environmental impact of the acid leaching process should be considered. The use of HCl may generate acidic waste streams containing dissolved metal ions (e.g., Al species), which require proper treatment before disposal. In practice, such effluents can be managed through neutralization and partial recycling processes, and their relatively simple composition facilitates treatment compared to more complex industrial waste streams. Conclusion In this study, the cobalt catalysts supported on silica prepared from natural kaolin were synthesized and tested for diesel production by Fischer-Tropsch synthesis. The silica was prepared by simple acid leaching at various concentrations. The as-prepared silica has a high SiO2 content. Kaolin was chosen as the silica source for the synthesis of the silica support due to its low price and local availability. Moreover, it could be thermally transformed into a more reactive form, metakaolin, through calcination. The synthesized silica derived from kaolin clay was used as a support and was compared with other commercial supports such as SiO2, Al2O3, and ZSM-5 in this work. It was observed that the synthesized silica had a significant effect on the cobalt-based catalyst, exhibiting better catalytic activity due to its higher specific surface area. The Co/SiO2-5 M catalyst provided better catalytic performance in FT synthesis in terms of selectivity to the diesel range, which is attributed to the higher dispersion of cobalt particles on the synthesized silica (SiO2-5 M) support, as confirmed by XRD and H2−TPR analysis. BET results proved that an increased surface area of SiO2−5 M enabled a greater dispersion of cobalt particles on the support surface. A larger pore size of the silica support facilitated the formation of larger cobalt particles. These increased particle sizes of cobalt oxide improved the reducibility of the catalyst by weakening the interaction between cobalt and the support. The CO conversion and C12-C20 hydrocarbon production increased with the initial increase in the acid concentration of the synthesized silica support (HCl = 2 M, 5 M, and 8 M). The physicochemical properties of the catalysts demonstrated a strong influence on FT catalytic performance, with the moderate acid treatment (5 M HCl) providing the optimal balance of surface area, cobalt dispersion, and reducibility for diesel-range hydrocarbon production. Acknowledgements The authors would like to acknowledge the Office of National Higher Education Science Research and Innovation Policy Council – PMU C, PTT Exploration and Production Public Company Limited (PTTEP), CHERDCHAI Corporation Co., Ltd., POLAWAT ENGINE Co., Ltd., and VERASUWAN Co., Ltd., (C10F640103) for funding support. The authors also acknowledge the partial support from Mahidol University (Fundamental Fund: Fiscal year 2024 by National Science Research and Innovation Fund (NSRF), grant number: FF-098/2567). The authors would like to thank Nawapol Udpuay, Chawalit Takoon, Suwilai Chaveanghong, and Bancha Panyacharoen, scientists at the Mahidol University Frontier Research Facility (MU-FRF) for their kind assistance in instrumental operation of X-ray Diffraction and Raman microscope. Author contributions Sreypov Ke: Investigation, Methodology, Visualization, and Writing—Original draft preparation. Atthapon Srifa : Resources. Wanida Koo-Amornpattana : Resources. Suwimol Wongsakulphasatch : Validation and Funding acquisition. Suttichai Assabumrungrat : Validation. Choji Fukuhara : Resources. Sakhon Ratchahat : Conceptualization, Methodology, Validation, Data curation, Supervision, Funding acquisition, and Writing- Reviewing and Editing. Data availability All data related to the finding of this study are accessible upon request from the corresponding author Sakhon Ratchahat. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement All data related to the finding of this study are accessible upon request from the corresponding author Sakhon Ratchahat.