3-1- Reaction gels:
The most apparent dissimilarity in the preparation of the reaction mixture in the aminothermal method compared to hydrothermal is the alteration in the rheology and pH of the gel. The pH values of the synthesis solution before and after the crystallization process were measured at room temperature. During the hydrothermal crystallization process, the pH value increases from 5.8 to approximately 9 due to the incorporation of precursor species into the framework. [15][16], while in the aminothermal treatment, the crystallization of SAPO occurs in an alkaline medium (~ 9.5) with a slight pH increase. Additionally, in contrast to the hydrothermal technique [17], the order of adding starting materials in the aminothermal method did not significantly affect the final pH of the reaction gel. The viscosity and rheology of the reaction mixture also differed between the two methods. In the hydrothermal method, a smooth gel was obtained due to the aqueous reaction media. However, the low amount of water in the aminothermal process resulted in a thick paste of a reaction mixture that was difficult to homogenize. At the stage when phosphoric acid is added to the gel, it begins to thicken. This may be attributed to the rapid interactions of phosphoric acid with depolymerized Al source facilitated by amine solution and the formation of an amorphous Al-O-P-based gel phase, which is the building unit of the SAPO crystal [18][12]. Therefore, adding H3PO4 to the SA1 solution before silica causes the formation of a thicker gel than SA2(i).
It is noteworthy that sample SA2(iii) with the silica source of TEOS had a smoother gel compared to SA2(i), despite having the same batch composition and gel preparation conditions. The reason for this lies in the presence of ethanol, which is a byproduct of the hydrolysis of TEOS. The polarity and interfacial tension between the solvent and reactants play a crucial role in the gel formation, as well as the nucleation and growth of crystals. Solvents with high polarity and high interfacial tension, such as water, prevent the strong interactions between reactants, thereby slowing down the polymerization/condensation reactions. As a result, the hydrothermal system forms a smoother synthesis gel than the aminothermal system. Ethanol as a polar additive in the amine media of sample SA2(iii) has a smoothing effect on the gel-based mixture.
The study also conducted aminothermal synthesis using various other amines, including TEOA, DGA, DEA, TEA, and DIPA, under the same conditions as sample SA2(i). The results showed that TEOA, as a polar solvent (more polar than morpholine), produced a smooth mixture, while DEA and TEA, with less polarity, formed a completely dry and lumpy gel after mixing with aluminum and phosphorus sources. The large-scale preparation of aminothermal reaction mixture for templates like DEA and TEA poses a serious challenge due to their high volatility and non-uniform dry gel formation. In such cases, the dry gel conversion method may be a more effective synthesis technique than aminothermal [19].
3-2- XRD results:
The XRD patterns obtained from all samples demonstrate the characteristic peaks of the CHA phase (Fig. 2). Sample SH, which underwent a longer crystallization time, exhibits more intense peaks compared to samples SA1 and SA2(i). The crystal size calculated from Scherer's equation for SH, SA1, and SA2(i) is 40, 40 and 37 nm, respectively. sample SH displays two shoulders at 2 thetas of 20.6° and 21.7°, which arise from tridymite impurity [20]. As reported in the literature, dense phases such as AlPO4-tridymite are formed in the first step of SAPO synthesis[21]. Therefore, the factors decreasing the crystallization rate facilitate the formation of tridymite and cristobalite impurity phases: the use of alumina and silica sources that have low reactivity, such as gibbsite and water glass, Al/P < 1, very dilute reaction mixture, and insufficient template [22][23][21][24]. It is worth noting that this impurity phase was not observed in any of aminothermally synthesized samples, which may be attributed to the higher amount of template and the alkaline synthesis solution.
Samples A1 and A2(i) were prepared under similar conditions except for the order of adding ingredients. XRD patterns of these samples confirm that the crystallinity of SAPO is affected by the order of adding materials in the aminothermal technique. The addition of phosphoric acid before silica in sample A1 leads to a thicker and drier gel compared to sample A2(i). A thick, viscous synthesis gel can reduce convection currents and mass transfer, leading to less nucleation and larger crystals [25][26]. Furthermore, due to the limitations of mass transfer in the aminothermal system, it was observed that increasing the duration and intensity of mixing during aging can significantly accelerate the growth of crystals and increase crystallinity (as observed in the XRD pattern of SA(ii)). Sample SA2(ii), synthesized in a shorter crystallization time, has a higher crystallinity than sample SH. This highlights the potential of the aminothermal method in promoting faster crystallization of SAPO structures than the hydrothermal method. The crystal size for samples SA(ii) and SA(iii) is 54 and 44 nm, respectively.
It is interesting to note that XRD patterns of samples SA1 and SA2(ii), which possess higher crystallinity and are subjected to conditions that facilitate faster crystallization, reveal the presence of an impurity phase of SAPO-20 (SOD). Furthermore, the color of these samples after calcination is considerably darker than that of the other samples. The incomplete calcination of these samples can be attributed to the very small size of SAPO-20 pores (i.e., 0.3 nm) in comparison to SAPO-34. The strong diffusion resistances in the structure of SAPO-20 make it difficult for oxygen to reach the morpholine molecules trapped in the cavities, thereby hindering their effective removal. To the best of our knowledge, the synthesis of SAPO-20 with morpholine has not been reported in the literature. The formation of this phase is favored by increasing the alkalinity of the reaction mixture, small amine molecules, high temperature, and longer crystallization time, as reported in [27][28]. Therefore, it can be inferred that the crystallization temperature of 200°C and the presence of large amounts of morpholine in the aminothermal method are the primary reasons for the formation of SAPO-20. This phase exhibits good stability at high crystallization temperatures, whereas for CHA structures such as SAPO-34, crystallization must be halted at the optimal time to prevent the formation of more stable phases. The disappearance of SAPO-20 peaks in SA(iii) confirms this claim.
Figure S1 displays the XRD patterns of the samples directed by other alkyl and alkanol amine templates. As evidenced, using DIPA resulted in the formation of SAPO-34 and SAPO-11 phases, while TEOA produced the structure of SAPO-5. The utilization of DGA, DEA, and TEA templates yielded dense phases. These outcomes are inconsistent with the findings reported in references [8] and [10], which may be due to different preparation conditions and static crystallization.
3-3- SEM and EDX results
As depicted in Fig. 3, the SEM images reveal that all samples consist of cubic crystal particles, which are the characteristic shape of SAPO-34. Samples SH and SA1 exhibit a smooth surface, with an average particle size of 3.3 and 2.2 µm, respectively. Conversely, in samples SA2(ii) and SA2(iii), which were aged under vigorous stirring, crystalline particles with broken surfaces and an average size of 5.8 and 9.3 µm, respectively, were observed. These interfacial and surface defects can result in an increase in Lewis acid sites within the structure [29]. Surprisingly, the particle size of SA2(iii) is greater than that of the other samples, despite having a significantly shorter crystallization time. According to the literature, the rapid hydrolysis of TEOS in the hydrothermal system leads to a higher nucleation rate and, consequently, smaller particle size compared to silica sol [30]. This contradicts the results observed in the aminothermal method. As previously stated in section 3.1, the liberation of ethanol from TEOS alters the solution's behavior and enhances the uniformity of the gel. This, in turn, can improve mass transfer within the system and promote crystal growth.
The results of the EDS analysis, which detail the elemental composition, are presented in Table 2. The (Si + P)/Al ratio was greater than unity for all samples except for SA1, indicating the incorporation of silica in the structure via SM2–SM3 mechanism combination and the formation of silica island [31]. The molar composition of SA1 suggests the existence of extraframework aluminum, which may be due to the distinct order of material addition. The quantity of silica present in the sample synthesized with DIPA (Table S1) is lower than that of samples synthesized with morpholine. This is likely due to the smaller size of morpholine (90.39 A³) in comparison to DIPA (124.94 A³). The small size of the amine results in more amine being trapped within the cages, necessitating a greater amount of silica to balance the protonated amines [10]. Furthermore, P/Al > 1 observed in this sample, which may be attributed to the presence of terminal P = O and P-OH groups within the SAPO structure. This is likely a result of the reduction in dissociation of H3PO4 and subsequent reduction in condensation with aluminum species, which increases the likelihood of forming terminal P-O bonds [32].
3-4- Catalytic performance
The catalytic efficacy of all samples was assessed in the MTO reaction at a temperature of 400°C and a WHSV of 2 h− 1. All samples synthesized with morpholine (both aminothermal and hydrothermal) exhibited a methanol conversion exceeding 98% throughout the catalyst's active life. Figure 4 illustrates the product distribution of the samples when selectivity to light olefins is at its maximum value. The columns display selectivity values to light olefins (C2=, C3=, and C4=) and light alkanes (C2 and C3). In all samples, the selectivity to ethylene was greater than that of propylene, although this ratio was slightly higher for hydrothermal samples than aminothermal samples. According to the literature, the C2=/C3 = ratio is dependent on the intermediates formed within the catalysts, which are influenced by structural factors such as cage size, crystal size, density and strength of acid sites, and operating factors such as reaction temperature and methanol throughput [33]. Propane was the dominant alkane across all SAPO-34 samples. DeLuca et al. suggest that alkylcarbenium stability plays a significant role in alkene hydrogenation. Due to the greater stability of secondary and tertiary alkylcarbenium ions compared to primary carbenium ions, hydrogenation barriers for propylene and butylene are lower than for ethylene [34]. However, structural constraints and diffusion resistances limit the production of butane, resulting in propane being the primary alkane observed in the MTO process on SAPO-34.
The catalytic efficacy of the hydrothermal sample was compared to the aminothermally synthesized samples with two distinct addition orders (Fig. 5). As demonstrated, the performance of the aminothermal samples in the MTO process is comparable to that of the hydrothermal sample. The highest selectivity to light olefins for all three samples exceeds 80%, with the exception that sample SA2(i) displays a consistent trend, while for samples SH and SA1, the selectivity to light olefins is initially low due to alkene hydrogenation and alkane production. However, as the reaction progresses, the selectivity to alkenes increases, and subsequently, with catalyst deactivation (concomitant with the appearance of DME), the selectivity value drops sharply. SA1 exhibits greater stability compared to SH and SA2(i), demonstrating over 6 hours of high activity. The performance of the sample synthesized using DIPA is also depicted in Figure S2. Due to its moderate acidic properties and low silica content, this sample exhibits a consistent trend in selectivity towards olefins. However, the presence of the SAPO-11 phase results in a lower yield of light olefins in comparison to the MOR-templated sample. Owing to the larger pore size of SAPO-11(4.0 × 6.5 Å) in comparison to SAPO-34 (3.8 Å), butylene was produced in greater quantities than propylene and ethylene on this catalyst, which contrasts with the outcomes observed in other samples.
Moreover, a comparison was made between the catalytic behavior of SA2 samples synthesized under different aging conditions and crystallization durations (Fig. 6). The presence of SAPO-20 and coarse particle size in SA2(ii) samples increased the diffusion resistance, leading to accelerated coke formation and catalyst deactivation. Although sample SA2(iii) exhibited good stability, its selectivity to alkanes was initially high (approximately 45%). As the reaction progressed, the trend for alkanes decreased while that for olefins increased, indicating that alkanes are derived from olefins and that strong acid sites play a crucial role in the hydrogenation reaction of olefins. As these sites gradually deactivate and coke forms, the rate of alkane production decreases.
In general, two primary factors have been identified as contributing to the formation of alkanes. The first factor is high diffusion resistance and residence time, which create favorable conditions for secondary reactions [35] [36] [37] [38]. Research indicates that olefins have a greater tendency to interact with internal acidic hydroxyl groups than aromatics and paraffins[39]. The second factor is the high acid density in SAPO-34. Results demonstrate that a high silica incorporation in the structure increases the likelihood of alkane formation [36] [40] [37] [38]. Recently, the role of Lewis acid sites (LAS) in hydrogen transfer has also been established [41] [42]. LAS in SAPOs can exist as threefold-coordinated Al or silicon and/or extraframework Al. It is also believed that the adsorption of methanol on the Brønsted acid site can induce frustrated Lewis pairs and create a synergetic effect in catalyzing the cascade reaction [43][44][45].
As evident from the results, SA2(iii) exhibited a higher amount of silica (acid density) and particle size (diffusion resistance) compared to the other samples, leading to increased alkane formation. However, intense post-treatment involving washing and calcination to remove amines from the structure can result in the formation of Lewis acids, which necessitates further investigation in this regard.