The results of the performed characterizations, followed by the observed performances of the catalytic testing, are reported in this section. These data are discussed in the framework of identification and possible correlation of the characteristics of the catalysts’ samples with their catalytic performances.
3.1. Characterization Results
All samples were analyzed using XRD to investigate the effect of introducing Cu and Zn by different methods on the crystalline structure of the zeolite (ZSM-5). In particular, the material phase characteristics of the hybrid synthesized single-pot (FSP) catalyst were studied. Additionally, all samples’ crystalline structures and phases were compared to the parent H–ZSM-5 (commercial).
Figure 1 shows the XRD patterns of all samples, where for all hybrid catalyst samples the feature reflections corresponding to MFI (ZSM-5) can be found at 2θ = 21–25°, according to the data bank JCPDS No. 49-0657. All hybrid samples exhibit the corresponding reflections of CuO (JCPDS No. 41-01254) and ZnO (JCPDS No. 89-1397). The peak assigned for CuO at 2θ of 35.5° is present in all samples, except H–ZSM-5. However, it is sharper in CZZ(33), implying the presence of larger CuO crystallites in this sample, as seen in
Figure 1. Meanwhile, CZZ(66) shows higher ZSM-5 intensity than CZZ(33) due to its twice higher ZSM-5 content (66%). As the metal content increased in CZZ(33), the corresponding Cu and Zn reflections became sharper and narrower in comparison to their counterparts (CZZ(66)). As can be seen for CZZ(33), CZZ(66), and FSP, the 35.43° reflection corresponds to (002) crystal planes inside CuO.
Figure S2 shows the XRD patterns corresponding to acidic ZSM-5 and the unsuccessful attempt to synthesize ZSM-5 without fluorine in the presence of Cu and Zn nitrates. As a result of the experiment, acidic ZSM-5 was obtained. However, the introduction of Cu and Zn metal nitrates under the same hydrothermal conditions did not result in the expected ZSM-5 feature reflections (
Figure S2). Upon introducing fluorine under the selected conditions, ZSM-5 was able to be crystallized in FSP-P (
Figure 1). The FSP-P sample presents an additional composition of zinc metasilicate (ZnSiO
3) that might be attributed to the inter-fusion between ZnO and SiO
2 during the heat treatment at higher temperatures [
32]. In order to investigate the phase transformations and determine the calcination temperature of the as-synthesized FSP sample, a TGA analysis was performed, which covered the temperature range of room temperature to 1100 °C, as can be seen in
Figure S3. Moreover, the XRD patterns of FSP at different calcination temperatures were plotted in
Figure S4. As seen, increasing the calcination temperature decreased the crystallinity of the planes found in ZSM-5. There are reflections associated with ZnSiO
3 in FSP (
Figure 1), possibly as a result of the formation of a solid solution between Zn and Si during the calcination step at above 600 °C, as shown in
Figure S4 where the matched reflection appeared at 650 °C. The catalytic materials showing the highest zeolite crystallinity (i.e., SI, SP) are expected to be more active for methanol dehydration. This was confirmed by analyzing their catalytic performance indicators, primarily through their DME selectivity.
The chemical states of metal species (e.g., Cu) over the catalysts’ samples were studied using XPS characterization. Depending on the applied synthesis method, i.e., precipitation, impregnation, and single-pot synthesis, different levels of dispersion of the metal species [
28] and quantities of active sites over the catalyst surface can be expected. An XPS measurement was performed to reveal the electronic and surface chemical states of the involved metal species. This is visualized through a comparative XPS analysis of the catalysts synthesized via sequential, co-precipitation/impregnation, electrospinning, and single-pot syntheses.
Figure 2a illustrates the XPS spectra of Cu 2p plotted from 925 eV to 965 eV for all studied samples. The Cu 2p spectra mainly show four sub-peaks, including doublets of Cu 2p
1/2 and Cu 2p
3/2 with the spin-orbit splitting of about 19.8 eV and two shakeup satellite peaks [
33]. In order to better understand the chemical shifts in all samples, they were compared with the XPS spectra of pure CuO and ZnO, labeled as CuRef and ZnRef, respectively. It has been reported elsewhere that the shakeup satellite peaks appear at 5–10 eV below the principle line in Cu 2p spectra [
34], whereas such peaks cannot be found in Zn 2p spectra. These peaks particularly represent the presence of Cu(II). Having considered this and as seen in
Figure 2a, the two shakeup satellite peaks located at about 943 eV and 963 eV show the presence of the paramagnetic chemical state of Cu
2+ in all samples (
Figure 2a) [
35]. The doublets of Cu 2p
3/2 and Cu 2p
1/2 at 933.73 eV and 953.27 eV can be clearly seen in
Figure 2a for the CuRef sample. The feature peaks at 933.32 eV and 953.22 eV are observed for the SP catalyst with a chemical shift of 0.05 eV to the lower BEs. The same doublet peaks can be observed for the SI catalyst at 934.34 eV and 954.24 eV, while a shift towards higher BEs can be seen for this catalyst compared to SP. In addition, the shakeup satellite peaks of the SI catalyst are more intense than those of the SP catalyst, which can be correlated to the existence of Cu metallic in SP. For the CP catalyst, the feature Cu 2p peaks are at 933.7 eV and 953.45 eV, similarly attributed to Cu 2p
3/2 and Cu 2p
1/2, respectively. As can be seen from
Figure 2a, the catalysts prepared by the sequential impregnation method, namely SI catalyst, exhibit a higher shift compared to the corresponding spectra (Cu 2p and Zn 2p) in the catalysts prepared by sequential/co-precipitation approach, namely SP and CP catalysts. For electrospun fiber catalysts, including CZZ(33) and CZZ(66), all characteristic peaks of Cu 2p
3/2 and Cu 2p
1/2 can be observed at 933.95 eV and 953.7 eV. Moreover, shakeup satellite peaks are present in the corresponding spectra. A slight shift to higher BE can be seen for CZZ(66), which might be due to the lower amount of metal active sites present in this catalyst compared to the CZZ(33) catalyst. As can be seen for FSP-P XPS in
Figure 2a, it exhibits a triangular line shape, while the other samples display nearly square line shapes. One of the satellite peaks in FSP-P disappeared along with the main peak shifting to the higher BEs of about 936.52 eV. This XPS pattern and shift to higher BEs can be attributed to CuF
2 [
34,
36,
37]. This results from the solid-state effect of hybridization between 3d and ligand orbitals in the FSP-P catalyst [
36,
37,
38], in which NH
4F was used as a mineralizer so that Cu reacted with F in the synthesis medium of the single-pot.
Figure S5 shows the XPS of C 1s for all samples. As can be seen in the C 1s spectra, a different FWHM can be observed for all samples, meaning the different ratio of carbon is available for each sample. The BE shifts can be attributed to the charge transfer that occurred in all samples [
38]. In the Zn 2p XPS spectra,
Figure 2b, the characteristic doublet peaks attributed to Zn 2p
3/2 and Zn 2p
1/2 can be observed in all catalyst samples. The gap between the prominent doublet peaks in Zn 2p XPS is about 23 eV [
39]. For the SP catalyst, the peaks at 1022.1 eV and 1045.3 eV can be ascribed for Zn 2p
3/2 and Zn 2p
1/2, respectively. The synthesis and preparation methods can highly affect the surface chemistry states of the metal contents, especially the Zn species [
40]. There is a shift to the higher BE for SI compared to SP. As reported elsewhere, the Zn species located inside the cation exchange sites in the zeolite structures possess higher BE because the electronegativity of lattice oxygen in zeolite is higher than the O
2− found in ZnO [
40], stabilized on protonic acid sites, a blue shift can occur in such catalysts [
41], as seen in
Figure 2b. All deconvoluted feature peaks along with the corresponding BE shifts for Cu 2p and Zn 2p are shown in
Tables S1 and S2, respectively.
Figure 3 shows the results of the XPS analysis of Si 2p and Al 2p for all catalysts. The XPS data corresponding to Al 2p clearly show the presence of higher Al oxidation states in all samples, further confirmed by the presence of an Al–O reference peak at 75 eV, indicating the presence of Al–O bonds. Al 2p spectra of SP show two main peaks corresponding to aluminum compounds at BEs of 75.2 eV and 78.2 eV, which can be attributed to Al
2O
3 and anhydride Al
2O
3, respectively [
42,
43,
44]. A shift to the higher BEs appeared in the SI catalyst’s spectra, which can be attributed to the higher amount of zeolite present in this catalyst (66.7%wt). Since the same amount of ZSM-5 present in the SI catalyst was utilized in CZZ(66), it showed a similar pattern shifting to the higher BEs compared to CZZ(33). It confirms the higher Al oxidation states in CZZ(66). As compared to other samples, FSP exhibited a significantly higher chemical shift. The chemical shift to the higher BE indicates the possibility of the reaction of some Al with F
− to make AlF
3 with the corresponding peak of 78 eV [
45]. Nevertheless, it is still difficult to conclude that Al is only bound to F
−, as also zeolite was obtained for this sample, according to the XRD pattern (
Figure 1), meaning that Al participated in the zeolite framework as well.
Figure 3b represents the XPS spectra for Si 2p for all samples. Two main peaks expected by the presence of Si 2p are located at 101.4 eV and 102.4 eV corresponding to Si–O–Al and Si–O–Si compounds, respectively. The chemical shift to the higher BEs can be observed for all samples. SP exhibits the peaks mentioned above at 102.6 eV and 103.5 eV [
46]. XPS analysis was also performed for the spent catalysts and compared with the fresh ones, as shown in
Figure S6.
Table 4 shows the phase composition and average crystallite sizes of ZSM-5 and CuO for powder and fiber samples. The data were derived using XRD Rietveld refinement and ICP analysis. The phase compositions derived from XRD confirm the presence of the ZSM-5, CuO, and ZnO compounds in all hybrid powder and fiber samples, which are comparable with the intended compositions, in wt.% The Cu/Zn ratio was calculated based on ICP and XPS measurements, which are about 1 for all samples. As seen, SI and SP possess the smallest CuO crystallites with 7 nm and 8 nm, respectively, compared to the other samples.
Figure 4 presents the Nitrogen adsorption–desorption isotherm performed at 77 K for all samples. As seen there, ZSM-5-P and ZSM-5-F show isotherm Type-I with hysteresis meaning that microporosity exists in both samples. A surface area of 294 m
2·g
−1 was obtained for the ZSM-5-P sample, while after the ES process, a higher surface area of 326 m
2·g
−1 was achieved for ZSM-5-F, confirming that fibers exhibit a higher surface area.
Table 4 provides a comparative overview of such data for all catalysts’ samples. Similar isotherm types were obtained for CZZ(33) and CZZ(66) catalysts. However, introducing Cu and Zn nitrates to ZSM-5 in CZZ(33) and CZZ(66) fiber catalysts reduced the surface area significantly in these samples to 26 and 132 m
2·g
−1, respectively. The trends of CP and CZZ(33) catalysts are similar. However, the CP catalyst has demonstrated a slightly higher surface area of 41 m
2·g
−1, but both show a low portion of microporosity. This can be attributed to the higher amount of CuO and ZnO (each one 33%); through their precipitation, the access to the zeolite structure is limited, for instance, in the CZZ(33) catalyst. In contrast, CZZ(66), with more zeolite and lower metal contents, shows almost five times higher surface area than the CZZ(33) catalyst. All two SP and SI catalysts exhibit relatively similar ranges of surface areas of 183 m
2·g
−1 and 192 m
2·g
−1, respectively. Therefore, exhibiting a higher surface area by pure zeolites than the sequential/co-precipitation and impregnated catalytic samples can be attributed to the free protonic sites and channels found in pure zeolite [
47]. Accordingly, by introducing metal contents in the zeolite structure, these sites and pores get blocked, causing the reduction in the surface area [
47]. The catalyst synthesized by single-pot (FSP) revealed a surface area of 91.8 m
2·g
−1, which is slightly lower than what was obtained for SP and SI catalysts but higher than the CP catalyst. Despite the lower surface area for the FSP-P catalyst, it showed a higher mesoporous volume than the other samples.
Reviewing the recorded values of the external- and micropore surface areas for each sample and their average pore size provides valuable information regarding the desired structural characteristics of such hybrid catalysts. As seen through the data reported in
Table 5, introducing zeolite has caused a reduction in the pore size to below 2 nm. They may trap some of the water molecules in the micropores and accordingly hinder the accessibility of the active sites for the reactants. Electrospinning the samples increases the relative ratio of the external surface area to the micropores’ internal surface area; therefore, a higher conversion could be expected. The CP catalyst consists of CuO and ZnO and can be considered as a reference catalyst, also in this regard representing a low but externally accessible surface structure facilitating a high level of CO
2 hydrogenation. The CZZ(33), consisting of a significant part of the CuO–ZnO material structure, also shows a relatively low specific surface area. The precipitated metal oxides could have even blocked access to the zeolite porous structure, indicated by the recorded low micropore surface area in this hybrid catalyst. It is noteworthy to mention that the surface area of fibers can potentially be increased by optimizing the solution parameters such as precursor concentration, polymer selection, and operating parameters such as needle-to-collector distance, flow rate, and the applied voltage. Moreover, surface acidity of all catalysts was investigated and compared with the one measured for H–ZSM-5-P using temperature programmed desorption of ammonia (NH
3–TPD), as shown in
Figure S7.
SEM and EDS analyses were performed to investigate the microstructure and fine distribution of metal content in the catalysts synthesized by different methods.
Figure 5 illustrates SEM images of the electrospun fibrous samples, including ZSM-5-F, CZZ(33), FSP-F, and CZZ(66), before and after removing the PVP during the calcination step.
Figure 5a shows the as-spun PVP/ZSM-5-F morphology before the calcination step, illustrating a nonwoven network of PVP/ZSM-5 fibers with a diameter of 1.5 μm to 2 μm. As seen for this sample, in some regions, the ZSM-5 particles protruded from the surface of the fiber, which can be attributed to the aggregation of zeolite particles during the electrospinning process. Furthermore, some areas exhibit PVP-rich fibers with a diameter of about 100 nm having a few zeolite particles inside. As shown in
Figure 5b, the diameter of fibers was not significantly changed for ZSM-5-F after the calcination process at 550 °C. However, as seen in
Figure 5b, the fibers were distorted after removing PVP during the calcination process and a few ZSM-5-Fs were aggregated along their length. The microstructure of as-spun PVP/CZZ(33) fibers can be seen in
Figure 5c. A rough fiber mat was obtained for PVP/CZZ(33) with a range of fiber diameters from 500 nm to 2 μm. There is no significant change in the diameter of fibers after the calcination step in CZZ(33) (
Figure 5d). However, some spherical particles of approximately 2 μm appeared in this sample after the calcination step, which can be attributed to their copper and zinc metal contents. To investigate the spherical particles found through the CZZ(33) fibers, EDS analysis was performed (
Figure S8). As seen in
Figure S8, the spherical particles can be attributed to Cu.
Figure 5e presents the SEM image of as-spun PVP/FSP fibers before calcination with a diameter range from 1 μm to 2.5 μm. As seen in
Figure 5e, a smooth fiber structure was obtained for the PVP/FSP sample. However, after the calcination step and removing the PVP, the morphology of the fibers for the FSP-F sample was changed to a discontinuous network, and in a similar way to ZSM-5 fibers they were distorted after removing PVP. As seen in
Figure 5f, the diameter of fibers for FSP-F was reduced from 500 nm to 2 μm. As-spun PVP/CZZ(66) is shown in
Figure 5g, in which continuous fibers with a diameter of 500 nm can be observed, while some regions exhibit an aggregation of ZSM-5 particles in the base of fibers. The reduction in diameter can be attributed to the metal concentrations that are lower in CZZ(66) (with 16 wt.% CuO and 16 wt.% ZnO) than in CZZ(33) (with 33 wt.% CuO and 33 wt.% ZnO) caused by obtaining more continuous fibers of ZSM-5/CuO–ZnO in CZZ(66) (
Figure 5h). Some agglomerated ZSM-5 powder can be found in some areas in CZZ(66) after the calcination step.
The rest of the samples with particle morphology were also investigated by the SEM technique.
Figure 6 shows SEM images of the non-fibrous samples, including SP, SI, CP, and FSP-P in different magnifications of 20 k×, 10 k×, and 5 k×. For SP (
Figure 6a–c), the combination of CuO, ZnO, and ZSM-5 crystallites can be observed. CuO and ZnO have been distributed around ZSM-5 crystallites, yet, in some areas, metal oxides have become agglomerated. As a result, a precise estimation of metal oxide particle size is difficult. The particle size of ZSM-5 is in the range of 500 nm to 1 μm. For the SI catalyst, fine distribution of CuO and ZnO particles across the zeolite structure could be observed, as shown in
Figure 6d–f. As the CuO precursor was sequentially impregnated after impregnating the ZnO precursor on the zeolite particles in SI, it is expected to observe the CuO particles on the surface of this catalyst. Similar to the SP catalyst, the aggregation of particles can also be seen for the SI catalyst, showing the tendency of metal oxide particles to aggregate.
Figure 6g–i is the SEM images of the CP catalyst. More extensive aggregation of metal oxides in some areas occurred while the CP catalyst was synthesized through the co-precipitation method simultaneously using two metal precursors. FSP-P catalyst morphology differs markedly from others, as seen in
Figure 6j–l. As reported elsewhere [
48], CuO/ZnO can emerge in petal-like morphologies. So, it can be concluded that the petal-like structure found in
Figure 6j–l can be ascribed to CuO/ZnO. Particularly in the FSP-P catalyst, some fluorine zeolite crystals appeared in trapezoid/plate shapes and were not entirely covered by metal oxides.
The elemental EDS mapping and image analysis of the cut surface using FIB in different scale bars, namely 1 μm to 5 μm with several ranges of magnifications from 1500× to 18,000×, were conducted for the SP, SI, CP, and FSP catalysts.
EDS mapping analyses of Al and Si in
Figure 7e, and
Figure 7f show zeolite as a component of the sequential precipitation SP catalyst. Well-dispersed Cu and Zn species across the ZSM-5 structure can also be observed here. However, two different areas with different contents of Cu and Zn can be identified, namely Cu-rich and Zn-rich areas, as seen in
Figure 7c,d.
The Cu and Zn species can be seen over the surface of zeolite in the SI catalyst, as is expected from the implemented sequential impregnation method in this case (
Figure 8c,d). The H–ZSM-5 crystals within the size range of 500 nm to 2 μm are covered by Cu and Zn with no visually observed agglomeration, meaning that the impregnation was successfully carried out in this catalyst. As seen in
Figure 8e, the H–ZSM-5 crystals can be distinguished by their representative Si elemental content.
As seen in
Figure 9, the CP catalyst exhibits a large area of precipitated Cu and Zn. Although in co/sequential precipitation methods, the precipitates (metal particles) are supposed to be incorporated into the pores and channels of the zeolite structure, some metal particles might accumulate over the surface due to their bigger particle size compared to the pore size of H–ZSM-5 (
Figure 7) so that the metal particles cannot diffuse into the pores easily. Despite this, the metal particles can be observed in the areas between the zeolite particles (intercrystals).
FIB-SEM and EDS elemental mapping of FSP-P is shown in
Figure 10. It can be observed that fluorine ZSM-5 crystals were covered by Cu and Zn oxides asymmetrically. Cu and Zn were not immigrated or diffused into the fluorine ZSM-5 crystals meaning that these crystals are free of Cu and Zn. This might be due to the synthesis limitation that the above-mentioned metal particles did not participate in the crystallization of zeolite so the crystallization solely occurred by the main relative Al and Si-based precursors. The incorporation of metal active sites in closer proximity to zeolite crystals is anticipated to enhance catalytic activity. Although fluorine ZSM-5 is formed and crystallizes in a single-pot, the limitation of controlling the formation and crystallization of metal particles hinders obtaining a closest distance between them.
FSP-P’s morphology was further investigated through SEM/EDS analyses.
Figure 11 displays the corresponding SEM images along with elemental EDS mappings. The fluorine ZSM-5 crystals are covered with a petal-like aggregation of CuO/ZnO, as shown in
Figure 11. However, some fluorine ZSM-5 crystals were not completely surrounded by CuO and ZnO and emerged from the aggregation of the metal oxides. In some areas, zeolite plates were stacked on top of one another in different directions (
Figure 11a). The EDS mapping (
Figure 11) reveals that the analyzed zeolite area contains only Si and Al, without Cu or Zn. This confirms a pure zeolite region, as there are no other elements extending out of the representative area. As seen, fluorine ZSM-5 crystals can be found in different sizes. For instance, in the direction of the
c-axis, the size of zeolite crystals is between 10 μm and 20 μm, while the size varies from 1 μm to 5 μm in the direction of the
b-axis (
Figure 11a,b).
Figure S9 shows the cross-section SEM images before and after cutting the FSP-P sample by Gallium ion beam in FIB-SEM. The catalytic performance of FSP might be enhanced by decreasing the distance accomplished by introducing smaller crystals, which can be established by tuning the synthesis parameters such as time, temperature, and concentration of metal particles. This requires a separate comprehensive study to be devoted primarily to analyzing this parameter.
Figure S10 shows the schematic of synthesis procedures to fabricate catalysts in this work.