3.1. Characterization of fresh catalysts
3.1.1. Crystallinity, functional group, and textural analysis
Figure 1a shows the wide-angle X-ray diffractograms of both the commercial SAPO-34 and the synthesized FSAPO-34 catalysts. Several distinct peaks are detected at 2θ = 9.6°, 13°, 16.3°, 17.9°, 20.9°, 25.1°, and 31°, which correspond to the (101), (110), (021), (003), (12–1), (220) and (401) crystallographic planes, respectively, that ascribed to a typical diffraction peak of SAPO-34 (CHA-type zeolite). Both samples exhibit a rhombohedral SAPO-34 structure with the most substantial XRD diffraction peak at 2θ = 9.6 (JCPDS file No. 01-087-1527) (Soheili et al. 2021). However, the intensity of the XRD pattern of FSAPO-34 was reduced compared to that of commercial SAPO-34, which indicates a substantial loss of crystallinity during the modification process. This could be due to the growth of silica species during the formation of dendrimeric silica fiber. This phenomenon was consistent with previous studies for other fibrous silica zeolite such as fibrous silica Beta (Hussain et al. 2020a), fibrous silica ZSM-5 (Teh et al. 2016), and fibrous silica MOR (Hussain et al. 2020c). Besides, a broad diffraction peak centered 2θ = 23° is observed for FSAPO-34 due to the amorphous feature of the silica shell (Qureshi and Jaseer 2020). This broad peak was not detected in other fibrous silica zeolite material, possibly due to the different mechanisms for the formation of dendrimeric silica fiber on the SAPO-34 surface, as depicted in Fig. S1.
The FTIR method is frequently utillized to determine the stretching and bending vibrations of various types of bonds present in the structure of catalysts. Figure 1b illustrates the FTIR-KBr spectrum of commercial SAPO-34 and FSAPO-34 in a 400–1400 cm− 1 frequency range. Clearly, all the examined catalysts exhibited five almost identical IR bands spanning from 400 to 1400 cm− 1. The absorption peaks observed at wavenumbers of 1092 cm− 1, 972 cm− 1, 802 cm− 1, and 467 cm− 1 correspond to the vibration of asymmetric Si-O-Si stretching, external Si-OH, symmetric Si-O-Si stretching, and Si-O-Si bending vibration, respectively (Azami et al. 2021; Bahari et al. 2022). While the band at 635 cm− 1 is a result of the structurally sensitive double 6-membered ring vibrations, which are distinctive features of the CHA framework (Liu et al. 2008; Soltanali and Darian 2019). The synthesized FSAPO-34 shows a higher intensity, which is dominated by a strong band at 1092 cm− 1, 802 cm− 1, and 467 cm− 1, confirming the presence of silica fibers surrounding the SAPO-34 seed, consistent with XRD results. An additional peak was noticed at 960 cm–1 only for the FSAPO-34, which was ascribed to bending vibrations of Si-OH. This occurrence could be attributed to the increased abundance of terminal silanol groups formed during the synthesis of silica fibers.
The textural characteristics of the catalysts were investigated via N2 physisorption isotherms, with pore size distributions determined using non-local density functional theory (NLDFT) for clearer analysis. The commercial SAPO-34 exhibited a distinct type I isotherm, affirming the catalyst's microporous structure in accordance with IUPAC classification and aligning with findings in the literature (Chen et al. 2016; De Araujo et al. 2022). Conversely, FSAPO-34 catalysts displayed a type IV isotherm with an H3 hysteresis loop, which verifies a distinctive adsorption profile for mesoporous catalysts with slit-shaped pores (Fig. 1c) (Numpilai et al. 2021).
Both catalysts demonstrate nearly identical N2 adsorption at lower relative pressure, indicating the presence of microporosity in the catalysts. However, FSAPO-34 shows a greater N2 adsorption at higher relative pressures, suggesting the existence of mesopores (Hussain et al. 2020b). In addition, the N2 uptake at relative pressures of 0.3 and 0.9 are assigned to intra- and interparticle pores, respectively (Hamid et al. 2017). A notable increase in N2 uptake at a relative pressure of 0.9 was observed for FSAPO-34 compared to SAPO-34. This finding indicates that the fibrous silica form on the surface of SAPO-34 significantly contributed to the abundance of interparticle pores.
The hysteresis loops observed in the case of FSAPO-34 substantiate the effective formation of mesopores. This observation aligns with the pore distributions derived from the NLDFT method illustrated in Fig. 1d, which shows that a peak within the less than 2 nm range signifies the existence of micropores, whereas a peak within the greater than 2 nm range indicates the presence of mesopores. Although FSAPO-34 displayed a similar pore size distribution to SAPO-34 at a lower range (< 2 nm), the emergence of a broader peak in the 3–6 nm range indicates the formation of larger mesopores which is undoubtedly due to fibrous dendrimeric morphology. This improvement in mesoporous characteristics could potentially improve the mass transfer properties and negate the diffusion limitations associated with the commercial SAPO-34 catalyst. Detailed information regarding the physical characteristics of SAPO-34 and FSAPO-34 are tabulated in Table S1.
3.1.2. Morphological studies
Figure 2 reveals the FESEM and TEM images of SAPO-34 and FSAPO-34. From Fig. 2a, it can be seen that the unmodified SAPO-34 presents the common cubic morphology, characterized by a smooth and compact surface, which is identical to the conventional SAPO-34 (Bakhtiar et al. 2018). After undergoing synthesis using the microemulsion method, the FESEM image of FSAPO-34 reveals the development of spherical silica with well-ordered dendrimeric morphology, characterized by a uniform size of 150–200 nm (Fig. 2b-d). This distinctive spherical morphology covers the surface of the cubic particles of SAPO-34, as shown in Fig. 2c.
Notably, a TEM image of FSAPO-34 (Fig. 2e-f) shows an accumulation growth of fibrous silica on one side of the cubic SAPO-34 surface. Clearly, the EDX mapping of FSAPO-34 shows a high intensity of silica surrounding the outer surface of the core SAPO-34 (Fig. 2h), indicating a good dispersion of fibrous silica. This result may claim the formation of a core and shell structure of FSAPO-34. In addition, a TEM image of a single fibrous silica (Fig. 2g) further confirms the particle's spherical form, enriched with dendrimeric fibers on its external surface. These spherical morphological characteristics closely resemble those of KAUST Catalysis Center-1, as described by Polshettiwar (Polshettiwar et al. 2010). It should be noted that this result differs from other fibrous zeolite materials, where the center of the sphere consists of an aluminosilicate framework of zeolite, while the dendrimer was fully composed of silica (Teh et al. 2016; Abdul Jalil et al. 2019). This observation further clarifies the distinctive broad diffraction peak revealed in XRD results (Fig. 1a) exclusively in the case of FSAPO-34.
3.2. Catalytic performance and lifetime of catalysts
The MTO reaction was carried out in a micro-catalytic reactor over the commercial SAPO-34 and synthesized FSAPO-34. The conversion of methanol, selectivity of products, and light olefins yield were displayed in Fig. 3 (a-c) and computed in Table S2. At lower temperatures (300–350°C), the parent catalyst shows a better conversion compared to FSAPO34. However, the conversion at higher temperatures (> 400°C) shows little disparity between both catalysts, which achieved above 95% methanol conversion. The reduced methanol conversion of FSAPO-34 can be associated with the decrease in weak acid sites, as shown in Fig. S2 (a-b), since these sites are involved in facilitating the conversion of methanol to DME as well as alkylation and methylation reactions prior to the formation of olefin products (Wu et al. 2011; Yang et al. 2012).
It is worth mentioning that as the reaction temperature increases for both catalysts, the selectivity towards light olefins increases, where ethylene is the major product compared to propylene, accompanied by the gradual decrease of the concentration of other hydrocarbons appear among the effluents, which is similar to a pattern observed in previous work involving SAPO-34 (Castellanos-Beltran et al. 2018). The high light olefin yield was obtained over SAPO-34 at 500°C, which is 90.1% with a total light olefin selectivity of 93.1%, while for FSAPO-34, there was a slight 1.4% and 1.1% reduction in light olefin yield and selectivity, respectively, at the same temperature. At the lower reaction temperature (300°C), commercial SAPO-34 demonstrates the capability to produce a higher amount of ethylene and propylene (49.2%) compared to FSAPO-34 (15.2%). At this temperature, FSAPO-34 tends to produce more DME compared to SAPO-34, which highlights the deficiency of strong acid sites, as the olefin formation is only feasible on strong acid sites (Ghavipour et al. 2023). It is well known that the appropriate strength and amount of strong-acid sites on the surface of SAPO-34 is crucial to achieving high selectivity of light olefin, as demonstrated in the previous report (Wang et al. 2013; Hasnain Bakhtiar et al. 2024). The inactivity of silica fiber employed in FSAPO-34 resulted in the dilution of a strong acid site in SAPO-34 and thus led to a lower selectivity of ethylene and propylene.
The catalytic lifetime of the SAPO-34 and FSAPO-34 catalysts for continuous MTO reaction was examined, and the conversion of methanol was plotted as a function of the time on stream, as illustrated in Fig. 3d. The catalyst lifetime was defined as the duration of continuous reaction maintaining the catalyst activity until the methanol conversion lower than 50% (Wu and Hensen 2014). During the catalytic stability test, FSAPO-34 demonstrated an extended catalytic lifetime of 29.5 hours, sustaining a methanol conversion above 50% in contrast to commercial SAPO-34, which has a shorter catalytic lifetime of 19.2 hours. Although the strong acid sites of commercial SAPO-34 led to better light olefin selectivity, it also can expedite the formation of carbon-carbon bonds, leading to the generation of aromatics, which are side reactions that can cause a buildup of carbon deposits within the zeolite's pore structure (Zhang et al. 2018, 2019). On the other hand, the microporous nature of commercial SAPO-34, as shown in N2 physisorption results (Fig. 1c), often introduces diffusion limitations for the aromatics and branched hydrocarbons formed within the pores (Wang et al. 2013; Han et al. 2021a). All of these factors ultimately led to the rapid deactivation of commercial SAPO-34. Thus, introducing mesopores into the synthesized FSAPO-34 structure can enhance the diffusion of products out of the pores, effectively reducing diffusion resistance, which impedes blockage of the pores caused by coke deposition (Sun et al. 2015). Hence, the extended stability duration observed in FSAPO-34 signifies a remarkable resistance to deactivation.
3.3. Analysis of spent catalysts
Catalytic processes using zeolite-type catalysts frequently involve side reactions resulting in the development of carbonaceous material known as coke, which is the primary factor responsible for the deactivation of catalyst in the MTO reaction (Guisnet et al. 2009). The TGA and DTG analysis was used to measure the amount of coke deposits over spent SAPO-34 and FSAPO-34 catalysts at the temperature range of 30 to 900°C, and the results are shown in Fig. 4 (a-b). The initial weight loss at temperatures below 200°C is attributed to the removal of water and decomposition of volatile materials from the catalyst (Dai et al. 2013; Varzaneh et al. 2014). The second weight loss occurring between 200 to 800°C primarily results from the combustion of coke deposits on the catalyst (Varzaneh et al. 2016; Aghamohammadi and Haghighi 2019). These carbonaceous deposits can be categorized into three distinct types based on temperature ranges: 200–400°C, 400–600°C, and above 600°C. The initial range is identified as the soft coke category, characterized by a higher H/C ratio with higher combustion reactivity and a lower activation energy value. The second weight loss classifies as hard coke region with a lower H/C ratio relative to soft coke. The third category of weight loss, which occurs at temperatures exceeding 600°C, is referred to as laid coke. Laid coke is particularly hard to decompose in the de-coking process (Ahmed et al. 2011; Aghamohammadi and Haghighi 2019; Díaz et al. 2021).
It can be seen that both SAPO-34 and FSAPO-34 showed a decrease in weight as the temperature increased (Fig. 4a). However, the synthesized FSAPO-34 showed a weight loss of 24%, significantly lower than the 35% weight loss observed for SAPO-34, which indicates that less amount of carbon deposits formed on FSAPO-34 during MTO catalytic reaction. In addition, according to DTG curves presented in Fig. 4b, the spent FSAPO-34 mostly shows a formation of hard coke within the temperature range of 400 to 600°C while SAPO-34 shows an additional peak at temperatures above 600°C implying the generation of laid coke.
The O2-TPO analysis was carried out to further confirm the amount and characteristics of carbon deposition of spent SAPO-34 and FSAPO-34. Figure 4c shows the O2-TPO profiles over the 100– 850°C oxidation temperature range. Similar to TGA analysis, the coke species were classified by different oxidation temperatures into three types, including CA decomposed at low temperature (350–450°C), CB decomposed at moderate temperature (450–600°C), and CC decomposed at high temperature (600–750°C) (Fakeeha et al. 2013; Díaz et al. 2021). These sections correspond to amorphous, filaments, and graphite carbon, respectively (Hussain et al. 2020c). As expected, comparing the area of the coke-burning peak of both samples, it is evident that the area of the coke-burning peak of the SAPO-34 is substantially greater than that of the FSAPO-34 with 10.2 mmol/gCat and 2.6 mmol/gCat consumption of O2, respectively.
Furthermore, the maximum temperature of the coke-burning over the SAPO-34 shifted to the CC region which resulted in a more substantial peak area in the CC region compared to the FSAPO-34. The CC species is characterized by its thermal stability, making it challenging to decompose at high temperatures. Consequently, there is a possibility that this species could obstruct active sites and pores of the catalyst, leading to a reduction in the surface reaction during the MTO process (Yaripour et al. 2015; Liu et al. 2017). These observations are in agreement with the N2 physisorption (Fig. S3), which shows a smaller reduction in surface area and pore volume for FSAPO-34 compared to SAPO-34 after 30 hours of continuous catalytic reaction, which indicates less carbon deposition within the surface and pores of the spent FSAPO-34.
To differentiate the various carbon species present in SAPO-34 and FSAPO-34 catalysts after 30 hours of reaction, Raman spectroscopy was utilized, and the corresponding results are illustrated in Fig. 4d. The existence of two separate bands at 1350 cm− 1 can be ascribed to the disorder-induced band (D-band) resulting from carbon atoms bonded through sp2 hybridization and the graphite carbon band (G-band) resulting from by C-C stretching at 1590 cm− 1, respectively. In this context, the coexistence of D and G bands indicates that both SAPO-34 and FSAPO-34 exhibited the presence of amorphous and graphite carbon during the MTO reaction. However, compared with the SAPO-34, the G band and D band spectra of the FSAPO-34 exhibit lower intensities, and this intensity of the bands reflects the quantity of carbon deposited during the reaction (Hussain et al. 2019; Owgi et al. 2023; Hatta et al. 2023).
Furthermore, the degree of carbon deposit graphitization is often measured by the ratio of these two bands, known as IG/ID. A higher IG/ID ratio indicates a higher rate of graphitic carbon formation, which is a stable form of carbon deposits compared to very unstable and amorphous carbon. This led to the rapid deactivation of the catalyst and a tougher regeneration process (Rahman et al. 2022). The calculated value of the IG/ID ratio for FSAPO-34 and SAPO-34 was found to be 0.68 and 0.82, respectively, which confirmed that the synthesized FSAPO-34 contained less graphitic carbon than SAPO-34 owing to its distinctive fibrous structure. These observations were in accordance with the fact that FSAPO-34 displayed better stability relative to SAPO-34 during the stability test (Fig. 3d).
TEM analysis was carried out to examine the morphological characteristics of spent FSAPO-34, as depicted in Fig. 4 (e-f). As depicted in Fig. 4e, a portion of the fibrous silica shell has been relocated from the core SAPO-34 after 30 hours of MTO reaction. This relocation could be attributed to the aggregation of fibrous silica since it is not fully intact with the core SAPO-34, as observed in the TEM image of the fresh FSAPO-34 sample (Fig. 2e-f).
However, it is clear that the fibrous silica structure underwent no morphological change and retained its fibrous nature after the stability test, proving the strong integrity of the fibrous silica structure (Fig. 4f). In addition, the filamentous carbon species (whisker carbon) with similar morphology from the literature (Helveg et al. 2011; Simakov et al. 2015) were visible and extensively formed on the spent FSAPO-34, as evidenced by the TGA, O2-TPO, and Raman data. This type of carbon is expected to be produced from the MTO reaction since the light olefins produced from the reaction could further react to form higher olefins, paraffins, aromatics, and naphthenes through hydrogen transfer reaction or aromatization (Stö 1999; Olsbye et al. 2012). These hydrocarbons finally condense to form carbon deposition, which is usually seen as filamentous carbon deposition (Dong et al. 2022). On the basis of the above results, it is concluded that the synthesized FSAPO-34 displays a high resistance toward coke deposition, which leads to high catalytic stability in the MTO reaction due to the formation of fibrous silica on SAPO-34.