ZSM-5 zeolite chemical composition (mass %, EDS).
The nanoporosity in zeolite ZSM-5 was analyzed as a function of SiO2/Al2O3 molar ratio (MR). The internal pore structure was studied by high-resolution adsorption. Surface areas, microporous volume, characteristic energy of sorption, and pore-size distributions were calculated from N2 sorption isotherms by the BET, Langmuir, t-method of de Boer, αS-plot of Sing, direct comparative plots of Lee, Newnham, Dubinin-Astakhov, differential adsorption curves, and nonlocal density functional theory methods. The results indicated that MR dependence in these zeolites caused structural defects through micropore opening and widening as well as the emergence of further slit-like mesopores.
- ZSM-5 zeolite
- nanopore measurements
The nanoporous, ordered, and three-dimensional structure of zeolites makes them materials of great practical importance in the hierarchy. The broad use of microporous zeolites (pore diameter w < 2 nm) makes them very important in very specific areas, such as acid catalysts and adsorbents, as well as in refining processes and the basic petrochemical industry due to their unique properties both in activity and in selectivity . The great majority of zeolites possess two types of porosity: primary and secondary. Primary porosity with a well-defined size is associated with the crystalline structure of the zeolite and fundamentally depends on its structure type. Imperfections, including defects occurring during the growth of zeolite crystals, as well as defects generated by various treatments, cause secondary porosity, that is, the presence of mesopores (2 < w < 50 nm) and macropores (w > 50 nm) . The secondary porosity differs from the framework porosity, since it does not directly depend on the crystalline structure of the zeolite. Taken together, these comprise the texture of a sorbent. The primary porosity is characterized by а microporous volume (W0) with pore size as seen in the zeolite structure; secondary—by the pore-size distribution (PSD) and the external surface area (AE). These parameters are usually calculated from nitrogen sorption isotherms . Many, but not all, catalysts are porous materials, in which most of the surface area is internal. Sometimes it is convenient to talk about the structure and texture of such materials. The structure is defined both by the distribution in space of atoms or ions in the material part of the catalyst and by the distribution on the surface. The texture is defined by the detailed geometry of the void space in the catalyst particles. Porosity is a concept related to texture and refers to the porous space in the material. However, with zeolites, most of the porosity is determined by the crystal structure. To accurately describe the texture of the porous catalyst, a very large number of parameters will be required. With respect to porous solids, the surface associated with the pores can be called the internal surface. Since the availability of pores can depend on the ratio of the dimensions of the channel and molecules, the extent of the accessible internal surface may depend on the size of the molecules contained in the mixture and may be different for various components of the mixture (molecular sieve effect) .
ZSM-5 and, its purely siliceous analog, silicalite (both have a structural code “MFI” in accordance with the IZA database) are among the most widely studied zeolites. MFI is one of the most versatile and commercially significant zeolites; it is widely used in the petroleum industry to convert methanol into complex hydrocarbons in methanol-to-gasoline processes, as well as in the alkylation of aromatic compounds and their subsequent separation . The microporous network of this zeolite consists of intersecting straight and sinusoidal channels. The straight channels have pore openings defined by a cross-section of 10-member rings of 0.54–0.57 nm and sinusoidal channels by elliptic pores of 0.51–0.54 nm in cross-section. The intersections are cavities of 0.8 nm in diameter  (see Figure 1).
A detailed study on the different types of adsorption sites that constitute the structural skeleton of this zeolite was carried out by Cho et al. . They classified the sorption sites into three types: (1) the SS sites located in straight channels; (2) the SZ sites located in zigzag (sinusoidal) channels; and, finally, (3) the SI sites located at the intersections (Figure 2). One of the most important catalytic properties of ZSM-5 is its shape selectivity. This is a consequence of its primary microporous structure and is the basis for most of its successful applications .
Another important parameter that allows to adjust the zeolite properties is their chemical composition, that is, their SiO2/Al2O3 molar ratio (MR). The amount of Al in the framework is proportional to the number of exchangeable cations, H+ among others, which affects both Lewis and Brønsted acidity. The main interactions of the sorbate molecules in the pores of the zeolite are realized through the oxygen atoms of the lattice and extra-framework cations. Microporosity and secondary porosity in zeolites and similar materials can be determined from the low- and medium-pressure regions of the sorption isotherm using various approaches . The shape-selective activity of MFI can be attributed to the presence of active sites in micropores. It was shown that the shape-selective properties of zeolites may be greatly reduced due to the presence of active sites in the secondary porosity with a wide distribution of the pore diameter and on the external surface of their crystallites, so zeolites with a large outer surface area are less selective than those with fewer imperfections.
The presence of molecules that blocks the pores of the zeolite or a partial destruction of its structure can drastically decrease its activity by reducing the microporous volume accessible for the reactants. The effect that the external surface area of ZSM-5 zeolite crystals used for shape-selective reactions causes, was reported previously . Some authors in reported works have used the αS-plot of Sing as alternative method to evaluate the external surface area and the true intra-crystalline capacity .
The aim of this study was to accurately describe the dependence of all the different types of ZSM-5 porosity on MR and to show which methods are best suited for measuring them in each range. This will allow us to develop an approach to the application of various existing methods of texture characterization for samples of zeolite with mixed porosity.
A set of ZSM-5 zeolites in their sodium form (Na-ZSM-5) with a SiO2/Al2O3 molar ratio (MR) varying from 30 to 120 was synthesized using a template of tetrapropylammonium bromide (TPABr) following the methodology reported by Ghiaci et al. . Through the text and figures, these samples are called Z, followed by the MR value (30, 70, 95, or 120), for example, Z30 means an Na-ZSM-5 sample with an MR equal to 30. For comparison, a set of Na-ZSM-5 samples supplied by TOSOH Co., Japan, with MR 20, 23.3, and 30 were also studied. These TOSOH samples are called ZT, followed by MR value. A reference macroporous solid material required to estimate micropore volumes was obtained from the Tehuacan area in the state of Puebla, Mexico. This reference substrate was identified by X-ray powder diffraction (XRD) as α-SiO2. X-ray powder diffraction of ZSM-5 samples was obtained in the 2θ ranges of 5–50 degrees using diffractometer Bruker D8, using nickel-filtered Cu Kα (λ = 0.154 nm) radiation. Scanning Electron Microscopy images were collected from a JEOL JSM-6610LV electron microscope with tungsten filament and an electron detector operated at 20 kV. N2 adsorption isotherms were measured at the boiling point of liquid N2 (76.4 K at the 2200 m altitude of Puebla City, México) in the interval of relative pressures, p/p0 extending from 10−6 to 1 in an automatic volumetric adsorption system (Quantachrome AutoSorb-1C) in order to determine the textural parameters of ZSM-5 samples in addition to the evaluation of microporosity, which was analyzed through the determination of pore-size distributions calculated by the differential adsorption curves (DAC), Dubinin-Astakhov equation (D-A), and nonlocal density functional theory (NLDFT) approaches.
3. Results and discussion
3.1. X-ray analysis
The XRD patterns of all samples (Figure 3) are typical of ZSM-5 zeolites . In general, all the samples showed reasonably sharp diffraction patterns, indicating good crystallinity. Please note that commercial TOSOH samples and those prepared in the laboratory are nearly identical. The main peaks appear at the following 2θ angles: 8.0°, 8.9°, 9.8°, 14.0°, 14.8°, 20.9°, 23.2°, 23.9°, 24.5°, 29.4°, and 30.0° (Figure 3). Most of these peaks are not resolved; usually, one peak is a superposition of several closely located reflections. For example, the [−101], , and  reflections positioned at 2θ = 7.92°, 7.93°, and 8.01°, respectively, gave rise to a total peak at ~8.0. The most important difference between the standard XRD pattern and the observed for both sets of samples is the relative intensity of the various peaks, but a detailed discussion of the changes in the structure of ZSM-5 due to MR variations and synthesis conditions is beyond the scope of the present work and will be discussed elsewhere. Three peaks that appear at 2θ = 16.0°, 26.4°, and 30.9° in the ZT23.3 sample (marked with asterisks) are most probably associated with an unidentified impurity.
3.2. Scanning electronic microscopy
In Figure 4, it can be seen that the effect of the templates used during the synthesis process affects the morphology of the zeolite crystals obtained. Thus, for example, in Figure 4(a) and (b) corresponding to zeolites ZT-20 and ZT-23.3, it can be seen that the crystals obtained have lath-like shapes. In the case of the ZT-30 and ZT-23.3 zeolites, clusters of spheroidal crystals are observed where the crystals of the zeolites coexist, as seen in Figure 4(c) and (d). Finally, the SEM images of the zeolites ZT-30 and ZT-23.3 do not exhibit a predominant or defined geometry, as seen in Figure 4(e) and (f) .
3.3. High-resolution adsorption
N2 sorption isotherms at 77 K for both sets of samples are shown in Figure 5 as sorbed volume at standard temperature and pressure (STP) in cm3 per gram of zeolite versus p/p0. Figure 5 shows the sorption isotherms using a logarithmic p/p0 scale in the range of 105 ≤ p/p0 ≤ 1. The hysteresis loops shown by ZSM-5 zeolites are of the Type H3 or H4, characteristic of capillary condensation in the slit-like pores attributed to intercrystallite adsorption within aggregates. Table 1 gives the values of some important parameters obtained from the analysis of isotherms. All the N2 isotherms are of Type I according to the IUPAC classification . They indicate: (1) a high sorption at a very low relative pressure caused by the enhanced sorption potential of the ZSM-5 channel system and (2) formation of a monolayer at 0.1 ≤ p/p0 ≤ 0.8.
3.3.1. External surface area
To calculate the volume of the micropores from the sorption data, De Boer t-plots (thickness plots) and Harkins-Jura estimates are given in Table 2. An accurate estimate of these values can be influenced by the choice of the standard isotherm of a nonporous material selected to estimate the statistical thickness of the adsorbed layer (t) and the range of t values considered for the linear fitting  (Figure 6).
|Sample||ASL (m2 g−1)||ASB (m2 g−1)||AE (m2 g−1)||VΣ (cm3 g−1)||BET p/p0 range||CB||dp (nm)||W0/VΣ|
The total micropore volumes in cm3 g−1 for all the samples are given in Table 2. These values were calculated from: (1) αS-plots, (2) t-plots, and (3) the D-A equation (in this case, optimizing the values of the parameters n and E0). The ratio of the micropore-filling capacity to the total sorption uptake, W0/VΣ, a parameter that somehow indicates the degree of crystallinity of the zeolite being analyzed, is also included in Table 2 . For the construction of the αS and direct comparison plots, the adsorption volumes of the α-quartz without thermal processing were used as reference values; α-quartz was chosen as a reference material, since adsorption on these substrata occurs similarly as on a flat surface; access to the underlying microporous structure is impeded by water molecules in the pore openings. The standard nitrogen isotherms  are very similar to the adsorption isotherms of α-quartz along the adsorption branch up to a relative pressure of about 0.8. Since the same material was used as a reference for the αS direct comparison plots, similar microporous volumes were obtained from all of these methods. However, t-plots give slightly different results because the reference isotherm corresponds to de Boer equation.
ASL is the Langmuir-specific surface area; ASB is the BET-specific surface area; AE is the external surface area; VΣ is the volume sorbed at p/p0 = 0.95; p/p0 is the range used for the BET plot; CB is the BET constant; dp is average particle diameter; and W0/VΣ is the degree of crystallinity.
22.214.171.124. High-resolution αS-plots
The filling of macro, meso, and micropores can be proved by analyzing high-resolution αS-plots starting at low relative pressures, that is, 10−5; see Figure 7. There are some significant differences in the form of αS-plots as a function of MR, mainly for Z120. A pronounced distortion of the isotherm shape is observed at a very low p/p0, which can be explained by the enhancement of the sorbent-sorbate interaction in the pores of molecular dimensions, that is, the process of micropore filling . This type of αS-plot is characteristic of microporous adsorbents having a wide range of pore sizes and results in two or more separate stages of micropore filling. Figure 7 shows three linear ranges. Region III, with αS > 1.6, corresponds to sorption in the mesopores and adsorption on the external surface of the zeolite. Extrapolation of the line to the ordinate at p/p0 = 0 allows to estimate the total microporous volume W0. Region II with αS = 0.6–1.6 can be a sorption in the porosity created by partial removal of the constituents of the zeolite matrix with the formation of structural defects. This type of porosity is typically developed by acid leaching. Region I, with αS < 0.25, is due to the stages of final filling of the volume of ultramicroporous elliptical sinusoidal channels (0.55 × 0.51 nm) and nearly circular straight channels (0.54 × 0.56 nm). This behavior is mainly due to the combined filling of channels. However, this region is related to the filling of the ultramicropores corresponding to the narrowing and to the initial stages of channel filling. Zones of this αS-plot for Z120 appear because the substratum has mesopores, supermicropores, and uniform micropores with elliptical and nearly circular free openings. The diameter of these channels corresponds to approximately 1–3 diameters of molecules. The micropore filling regions obtained through high-resolution plots for all samples are presented inTable 2 (Figures 8 and 9).
3.3.3. Pore-size distributions calculated by the DAC, D-A, and NLDFT approaches
126.96.36.199. DAC method
Calculation of pore-size distributions from desorption branches of N2 isotherms using the differential adsorption curves (DAC)  method yields bimodal distributions (Figure 10), with the thickness of the pore size of ca. 0.36 and 0.55 nm for all samples. The plots are unimodal with the pore ca. 0.36–0.40 nm. This approach correctly describes the essential qualitative features of N2 sorption in the microporous zeolites, such as ZSM-5, that is, pores in the range of 0.3–0.6 nm. The results of these estimates are shown in Table 3.
|Sample||αS||DAC||t||D-A||Vmeso||E0 (kJ mol−1)||n|
188.8.131.52. D-A method
The pore-size distributions obtained by the D-A method  are shown in Figure 11. The average pore diameter, seen as a maxima on the curves by this method, varies according to MR. Table 3 lists the optimized W0, n, and E0 values using the D-A equation. The filling of elliptical channels of ZSM-5 with a length of 1.98 nm (width 0.51 × 0.57 nm) and connected through a zigzag path with a length of 0.665 nm (width 0.54) is the main contribution to the volume adsorbed. Figure 11 shows ZSM-5 pore-size distributions obtained by the D-A method, assuming a cylindrical microporous channel; these plots provide average diameters very similar to the width of the porous cavities of the zeolites ZSM-5 (5th column of Table 3). The D-A results shown in Table 2 suggest that MR in ZSM-5 zeolites promotes the opening and widening of their micropores. Nevertheless, the microporous volumes calculated from the D-A equation are somewhat different from the microporous volumes calculated by the αS and t methods. The fact that these D-A microporous volumes are always larger than the volumes calculated by the other methods suggests that the uptake at low relative pressures should be corrected for mesopore adsorption. This correction will result in a lower extrapolated value of the micropore volume from the D-A equation, and better agreement with other (αS and t) methods will be reached. It can be seen there that the D-A treatment, while overestimating somehow the pore sizes, still provides an approximate estimate of the micropore volumes and their corresponding pore sizes. The E0 values obtained by the D-A method decrease as MR increases: ZT20 > ZT23.3 > ZT30; Z30 > Z70 > Z95 > Z120 (Table 2). These values reflect the dependence of this parameter on the SiO2/Al2O3 ratio. Due to the crystalline nature of the zeolites, the force field created by the oxygen atoms in their structure must have symmetrical properties. Therefore, the E0 values are affected by the aluminum content, which has changed some atoms of the zeolite framework in the channels to be displaced, as well as changing the sorption potential. From the comparison of the E0 values obtained for these zeolites, this parameter is influenced not only by the pore sizes. Consequently, the Al content and the geometry of the pores have modified the electric field within the pores of the zeolites and thus have influenced the characteristic sorption energy.
184.108.40.206. Nonlocal density functional theory method
Nonlocal density functional theory (NLDFT) was developed to take into account pore sizes in voids of well-defined geometry . With this approach, the molecules adsorbed in the pores tend to be packaged in accordance with the adhesion forces established with the substrate (i.e. attractive forces between adsorptive and adsorbent molecules) and interactions with the remaining fluid molecules. The molar density of the adsorbed phase varies as a function of pore size. The adsorption isotherm is calculated from a given pore shape (spherical, cylindrical, slit-like, etc.), and the experimental isotherm is given as the sum of a series of individual single-pore isotherms multiplied by their relative abundance over a range of pore sizes. In the present case, the microporous structure of ZSM-5 zeolite can be approximated as a bundle of parallel cylindrical pores and the nature of the adsorbent can be assumed as that of the silica. In this way, the distribution of supermicroporous zeolitic adsorbents can be calculated from high-resolution adsorption isotherms. The results of the analysis of the size of the supermicropores using the NLDFT method are shown in Figure 12 and are listed in Table 4. The pore-size distributions obtained from the N2 isotherms using the NLDFT cylindrical pore model yield bimodal distributions with pore size characteristics of 1.8 and 5.0 nm. It is observed from this figure that the intensity of the distribution at 5.0 nm is poorly developed; however, it is represented in all distributions. A possible explanation is that the structures are not homogeneous and that it contains a significant amount of slit-like pores and pores of other irregular shapes. Based on the NLDFT method, one can get an idea about the actual widths and pore sizes of the supermicropore voids existing in ZSM-5 zeolites for which high-resolution N2 isotherms are available.
αS is Sing’s αS method, DAC is direct comparison plots, t is the t-plot employing the de Boer adsorption equation, D-A represents the Dubinin-Astakhov equation, Vmeso is calculated by subtracting W0αS from VΣ (Table 2), E0 is the characteristic energy of sorption, and n is the order of the sorption energy distribution from the Dubinin-Astakhov equation.
The obtained samples exhibit reasonable diffraction patterns, indicative of good crystallinity. The most important difference between the standard XRD pattern and those observed for both sets of samples is the relative intensity of the various peaks. The ZSM-5 samples synthesized are composed of crystals with different geometry in a range of sizes 5–10 μm. N2 isotherms have been measured, starting at a relative pressure of 10−5 and up to 1. To evaluate the texture properties of ZSM-5 zeolites, BET, Langmuir, Ast, surface areas, and external surface area were used. A significant amount of micropores was found in all ZSM-5 zeolites. Such methods as αS, t, and comparative plots (DAC) were used to estimate micropores in all zeolites. Nanopore size distributions (NSD) obtained from the N2 adsorption at 77 K are in perfect agreement with the type of present pores for all the samples, that is, micropores, mesopores, and macropores. Thus, adsorption is probably the most sensitive tool for evaluating quality and structural properties of the microporous materials such as ZSM-5 zeolites. To characterize these nanomaterials, a combination of comparative methods based on reference isotherms on well-characterized ZSM-5 zeolites is recommended, as well as the results of DAC, D-A, and the NLDFT.
This work was partially supported by DGAPA-UNAM IN107817 Grant, VIEP, and the Academic Body “Investigación en zeolitas,” CA-95 (PROMEP-SEP).
Conflicts of interest
The authors declare no conflict of interest.