XRD patterns of calcined samples of zeolite ZSM-5 synthesized by varying time of HT treatment and the reference one [15] are shown in Fig. 1, ranges of 2Θ for the most intensive peaks (7–10 and 22.5–25 °) are increased and placed separately on the right. Location of peaks of synthesized zeolites coincide with those of the reference MFI structure. The peaks referred to synthesized zeolites are broadened and some of them are overlapped resulting from both decreasing coherent scattering region and instrumental broadening (Table 1).
We evaluated the relative crystallinity by the area of peaks in the range from 22.5 to 25.0 (Table 2). It should be noted, that for correct estimation of relative crystallinity XRD patterns were obtained under the same recording conditions and the same mass of samples (0.2 g). The crystallinity grows while prolonging HT treatment up to 51 h and then falls during the following 118 h (Table 2).
The average yield of zeolites synthesized during more than 5.5 h was 87±2 %, while the yield of the sample subjected to HT treatment for 3 h was only 39 %. We also succeeded to separate a negligible amount of the sample prepared for 2 h of HT treatment with yield less than 1 %, being sufficient only for TEM analysis.
TEM images of zeolite materials prepared during different time of HT treatment from 2 to 168 h are presented in Fig. 2. We have estimated the average crystal size of zeolite crystals by measuring size of 100 particles at a number of TEM images (Fig. 3). Monodisperse MFI crystals with the average size of 50 nm uniformly distributed inside amorphous phase are observed in TEM images of the sample subjected to HT treatment for 2 h. The data obtained indicates that nucleation and initial crystal growth of zeolite from our mixture with molar ratio 1.0 SiO2 : 0.01 Al2O3 : 0.2 ТPAОН: 0.05 Na2O: 50 H2O: 4.0 EtOH most likely occurs in the gel matrix through the reorganization of solid amorphous material due to dissolution – condensation processes [16]. After 3 h of HT treatment the average crystal size increased up to 230 nm, the crystal surface became rougher. It should be noted, that amorphous phase was not observed in TEM images in case of the sample synthesized for 3 h, yield of solid product being is only 39 % and XRD crystallinity is 82 %. It means that the amorphous gel either lost stability and decomposed into nanoparticles, which were not separated by centrifugation, or totally dissolved under the consumption of silicate species during crystallization process. The following evolution of ZSM-5 crystals during HT treatment associated both with growing the crystal size and significant modification of the crystal morphology. Zeolite crystals turned into aggregate-like ellipsoid crystals with the average size of 300 nm after 15 h of HT synthesis. The average sizes of domains forming aggregate-like crystals are in the range of 30–50 nm. Aggregate-like crystals appeared to be generated by the attachment of the solid precursor domains оn the zeolite surface followed by their transformation into MFI phase via dissolution – condensation steps resulting in the formation of inseverable parallel planes across each crystal. A mechanism where the precursor 5-nm building units evolve to silicalite-1 crystals through aggregative growth was earlier shown here [17]. Further evolution of aggregate-like ZSM-5 crystals under HT treatment represent both increasing the average crystals size and gradual smoothing the surface of zeolite crystals because of Ostwald ripening. Moreover, the transition into thermodynamically stable “rounded-boat” morphology was observed resulting from the crystal reconstruction driven by the minimization of surface free energy. Significantly, zeolite yield after 15 h of HT treatment remains approximately constant indicating evolution of aggregate-like crystals proceed precisely through the reconstruction of those. Thereby, ZSM-5 crystals with aggregate-like morphology turned out to be metastable and we have succeeded to reproducibly synthesize them.
Before discussing textural characteristics of synthesized ZSM-5 samples peculiarities of application of BET equation to evaluate surface area of zeolites should be considered. Bae and coauthors shown that choice of appropriate pressure range based on consistency criteria allows obtaining surface areas from N2 adsorption-desorption data being in good agreement with those obtained directly from crystal structures [18]. Here we have used two consistency criteria suggested by Rouquerol et al [19] and recommended by IUPAC [20] to obtain the BET surface areas from the linear region of nitrogen isotherms: (a) the quantity C should be positive; (b) application of the BET equation should be restricted to the range where the term V(1–p/p0) continuously increases with p/p0, where V is the volume of nitrogen adsorbed per gram of material. Therefore, p/p0 range for approximation of experimental date by BET equation was chosen from 0.001 to 0.01.
Textural characteristics of ZSM-5 materials calculated from nitrogen adsorption-desorption data are represented in Table 2. All zeolite samples exhibit high BET surface areas (470–590 m2∙g‑1) and micropore volumes (0.14–0.19 cm3∙g-1). The effect of time of HT treatment on textural characteristics of ZSM-5 samples is shown in Fig. 4. The reduction of the BET surface area and micropore volume after 72 h of HT treatment is in accordance with falling their XRD crystallinity. External surface area (the surface area of pores larger than micropores) was simultaneously increased, indicating mesopore formation because of partial destruction of zeolite lattice. Falling the total pore of the zeolite sample during HT treatment took place in two steps. The first step after 3 h of HT treatment appeared to result from decreasing the contribution of mesopore volume between packed crystals, and the second one after 72 h − because of smoothing the surface of zeolite crystals.