Modification of Zeolite Morphology via NH4F Etching for Catalytic Bioalcohol Conversion

Various commercial zeolites, including FER, MOR, ZSM‐5, BEA, and FAU frameworks, were treated with NH4F aqueous solutions to study the effects of fluoride etching on different zeolite frameworks. NH4F‐treated small‐medium pore FER, MOR, and ZSM‐5 samples showed much higher mesoporosities than the untreated ones without alteration of the structural compositions and acidic properties. On the other hand, the 12‐membered ring zeolites BEA and FAU showed severe dissolution of the framework aluminosilicate structure after NH4F etching due to the high accessibility of fluoride species into the framework structures. The effect of NH4F concentration on the fluoride treatment of H‐ZSM‐5 zeolite was specifically studied. From the results, we observed that structural etching with 20 wt % NH4F was optimal for fabricating open‐pore H‐ZSM‐5 zeolite and resulted in a high mesoporosity with comparable relative crystallinity and acidity with respect to the untreated H‐ZSM‐5. The catalytic activities of the open‐pore H‐ZSM‐5 were evaluated with acid‐catalyzed methanol and bioethanol conversions. Remarkably, the hierarchical open‐pore H‐ZSM‐5 zeolite fabricated via fluoride etching exhibited an enhanced catalytic performance in bioethanol conversion with >85 % conversion over 34 h TOS and a higher catalytic stability in methanol conversion than the parent H‐ZSM‐5 (~50 % of bioethanol conversion at 34 h TOS).


Introduction
Zeolites are composed of silica and alumina tetrahedral units that share an oxygen atom between two adjacent units to build 3-dimensional crystalline aluminosilicate structures. [1]There are various zeolite types with different microporous framework architectures, such as 8-membered ring ferrierite (FER), 8membered ring mordenite (MOR), 10-membered ring Mobil Five (MFI) or ZSM-5, 12-membered ring beta (BEA), and 12membered ring faujasite (FAU). [2]They are very useful in heterogeneous catalysis because of their promising properties, including high surface areas, high thermal stabilities, ionexchange capabilities, acidities, and shape selectivities. [3]How-ever, there are limitations on mass transfer through the long diffusion paths of zeolitic microporous systems, leading to high amounts of coke deposition and fast catalyst deactivation. [4]onsequently, hierarchical zeolites have been designed to overcome these problems of conventional zeolites. [5]ierarchical zeolites are characterized by their incorporation of at least two out of the three pore size classes: micropores (with pore diameters less than 2 nm), mesopores (ranging from 2 to 50 nm in diameter), and macropores (exceeding 50 nm in diameter). [6]Given that traditional zeolites are predominantly microporous, the presence of both micropores and either mesopores or macropores within a zeolite sample is indicative of a hierarchical structure.This combination of pore sizes in the zeolite framework defines it as a hierarchical zeolite.The additional mesopores and/or macropores of hierarchical zeolites provide shorter diffusion path lengths, leading to improved molecular transport and resulting in catalyst lifetimes longer than those of traditional microporous zeolites. [7]Many studies have reported enhanced catalytic stability of hierarchical zeolites due to their higher mesoporosities compared to conventional zeolites. [7]For example, the hierarchical nanosized ZSM-5 zeolites fabricated by a post-synthetic treatment with NaHCO 3 treatment exhibited improved catalytic activity and less coke deposition in n-pentane catalytic cracking due to the reduced diffusion resistance and suppression of undesired reactions by the microporous and mesoporous systems of the synthesized hierarchical ZSM-5 with respect to the untreated one. [8]These behaviors confirmed the extreme advantages of hierarchical structures for the fabrication of highly efficient zeolite catalysts.
The syntheses of hierarchical zeolites are mainly divided into two approaches: the bottom-up or templating approach and the top-down or post-synthetic treatment method. [9]The [a] Dr. P. Iadrat, A. Prasertsab, Prof. C. Wattanakit  School of Molecular Science and Engineering (MSE) and School of Energy Science and Engineering (ESE) Vidyasirimedhi Institute of Science and Technology (VISTEC)  Rayong 21210 (Thailand) E-mail: chularat.w@vistec.ac.th bottom-up approach is the more popular method for the syntheses of hierarchical zeolites because the use of organic templates allows more control of the morphology and pore structure. [10]For instance, hierarchical ZSM-5 nanosheets were successfully synthesized by using tetrabutylphosphonium hydroxide (TBPOH) as a dual structure-directing agent (SDA) for building an MFI zeolite framework and creating a nanosheet morphology. [11]Nevertheless, most organic SDAs are expensive, so the syntheses of hierarchical zeolites in industry are not practical. [12]This synthetic procedure produces greenhouse gases such as CO 2 during the calcination step used to remove the organic templates, contributing to the environmentally unfriendly process. [13]Therefore, top-down or post-synthetic treatment methods are needed.3a,14] Although desilication and dealumination of the zeolites provides additional mesopores and/or macropores, the acidic properties can also be changed due to the changed structural composition (Si/Al ratio) after the treatments. [15]oreover, the crystalline framework structures of zeolites can be destroyed by severe alkaline and acid leaching. [16]Promisingly, fluoride etching has been proposed to overcome these limitations of desilication and dealumination. [17]Fluoride etching is a post-synthetic treatment of zeolites with an aqueous solution of ammonium fluoride (NH 4 F) and/or hydrofluoric acid (HF) to create an additional mesoporous network or hierarchical architecture without collapsing the zeolite framework structure and changing the chemical composition. [18]The NH 4 F and/or HF aqueous solution provides various fluoride species under equilibrium conditions, leading to similar rates for dissolution of silica and alumina sites in the zeolite frameworks while retaining the crystallinities and acidities of the parent zeolites. [19]nterestingly, the hierarchical house-of-cards-like ZSM-5 zeolite was successfully fabricated by NH 4 F etching, and it exhibited high catalytic performance in ethanol-to-hydrocarbon conversion due to access of the guest molecules to the active sites of the mesoporous ZSM-5 network. [20]The 3-dimensional openpore MOR structure created by the fluoride treatment was clearly illustrated by electron tomography, which showed large mesoporous connectivity from the outer to inner MOR crystals, indicating successful building of the hierarchical open-pore MOR zeolite by NH 4 F etching. [21] Many studies have explored the synthesis of hierarchical zeolites using fluoride treatments, primarily focusing on the ZSM-5 (with approximately 5 μm crystal size) [18a,19a,20] and MOR (~1.5 μm of crystal size) [21] zeolite frameworks.However, there is a noticeable gap in the literature regarding the limitations and broader impacts of this postsynthetic treatment method.Specifically, the effect of NH4F etching on diverse zeolite frameworks has not been thoroughly investigated, highlighting a significant area for further study and understanding.
Consequently, in this work, various zeolite frameworks, including those of FER, MOR, ZSM-5, BEA, and FAU, were treated with NH 4 F aqueous solutions to study the effects of fluoride etching on different microporous zeolite structures during the creation of additional mesopores.The limitations of this NH 4 F treatment method were systematically investigated.Moreover, the effects of various NH 4 F concentrations were also studied.The catalytic activities of the synthesized open-pore zeolites for methanol and bioethanol conversion were studied.These reactions are very important in producing nonpetrochemicalbased precursors or platform molecules, such as dimethyl ether or diethyl ether, for the production of various value-added chemicals, including biofuels. [22]

Effects of NH 4 F etching on the different zeolite frameworks
Various zeolite frameworks, including 8-membered ring (MR) FER (CP914C, Si/Al = 10), 8-MR MOR (CBV21A, Si/Al = 10), 10-MR ZSM-5 (CBV2314, Si/Al = 11.5),12-MR BEA (CP814E, Si/Al = 12.5), and 12-MR FAU (CBV712, Si/Al = 6) containing similar Si/Al ratios (in the range of 6 to 13), were treated with 20 wt % NH 4 F aqueous solutions to investigate the effects of fluoride etching on the morphologies of the different zeolite microporous systems.The XRD results in Figure 1 show that while the small and medium pore FER, MOR, and ZSM-5 zeolites displayed unchanged XRD patterns after the NH 4 F treatments, the XRD patterns of the fluoride-treated large pore BEA and FAU zeolites were significantly different from those of the untreated samples.The XRD intensity of the NH 4 F-etched BEA zeolite was clearly lower than that of the parent BEA sample (Figure 1B), demonstrating the lower relative crystallinity after the NH 4 F etching (Table 1).In the case of FAU, the characteristic XRD pattern disappeared after fluoride etching (Figure 1B).This result indicated that NH 4 F etching cannot be used to create hierarchical structures in large pore zeolite frameworks, especially FAU zeolite, because it destroys the microporous framework.On the other hand, the crystalline structures of the small and medium pore zeolites, such as FER, MOR, and ZSM-5, were retained after NH 4 F etching.In the morphological investigations, the SEM images illustrated insignificant changes in the crystalline morphologies after NH 4 F etching of all zeolite frameworks (Figure S1).Upon comparing the morphological changes of MOR zeolite following fluoride etching in our study with previous literature findings, [21] a distinct observation emerged.While the SEM images in the reported studies depicted the disintegration of nanorods from the assembled nanorod MOR crystals after NH4F treatment, our current work revealed minimal morphological alterations in MOR crystals post fluoride etching.This contrast suggests that the impact of NH4F etching is not uniform but is significantly influenced by the initial crystal morphology of the parent zeolites.To determine the textural properties, the prepared zeolite samples were characterized by N 2 sorption, and the results are shown in Figure S2 and Table S1.The N 2 isotherms of all samples displayed a combination of type I and type IV characteristics, indicating microporous and mesoporous materials except for the NH 4 F-etched FAU sample (Figure S2). [23]The fluoride-treated FAU (NH 4 F-H-FAU) presented a type V isotherm, indicating mesoporosity without micropores (Figure S2B), as shown by the textural properties in Table S1. [24]These results confirmed that NH 4 F etching destroyed completely the microporous structure of the FAU zeolite framework, consistent with the XRD results (Figure 1B).The NH 4 F-treated zeolites also displayed much higher V meso / V micro ratios and external surface areas compared with the parent zeolites, indicating the creation of additional mesopores and a higher degree of hierarchical structure obtained by fluoride etching (Table S1).On the other hand, micropores were clearly lost after fluoride etching of the large pore FAU framework; this confirmed the destruction of the characteristic microporous structure of the FAU zeolite via NH 4 F treatment, which was consistent with the XRD results (Table S1 and Figure 1B).To examine the acidic properties, the zeolite samples with different framework types were characterized by temperature-programmed desorption of ammonia (NH 3 -TPD), as shown in Figure S3 and Table S2.The NH 3 -TPD profiles were deconvoluted to give two peaks at lower and higher temperatures, which were assigned to weak and strong acid sites, respectively (Figure S3).While the total numbers of acid sites and Si/Al ratios of the NH 4 F-treated FER, MOR, and ZSM-5 zeolites did not differ significantly from those of the parent samples, the BEA and FAU zeolites showed much lower acid densities and higher Si/Al ratios due to the collapsed framework structure after fluoride etching (Table S2).Given the small size of the NH3 molecule, with a kinetic diameter of 2.6 Å, [25] it can readily penetrate the entire porous structure of various zeolites, including dimensions such as 3.5×4.8Å and 4.2×5.4Å for FER, 2.9×5.7 Å and 6.7×7.0Å for MOR, 5.1×5.5 Å and 5.3×5.6 Å for ZSM-5, 5.6×5.6 Å and 6.6×6.7 Å for BEA, and 7.4×7.4Å for FAU. [26]This accessibility is effective even in untreated microporous parent zeolites.Consequently, the minimal change in acidity observed after fluoride treatment (as determined by NH3-TPD) is a noteworthy result, indicating that the acidic properties of the parent zeolites are maintained post-NH4F etching.This finding is particularly evident in the FER, MOR, and ZSM-5 zeolite frameworks, as illustrated in Figure S3 and Table S2.To further investigate the kinetic of fluoride etching over different zeolite frameworks, the NH 4 F etching of ZSM-5, MOR, and FAU zeolites was performed with various treatment times (0-6 h).The relationship between relative crystallinity (%) and etching time (h) is shown in Figure S4.While the ZSM-5 did not show an insignificant change in relative crystallinity (stable at ~98 % over 6 h of NH 4 F treatment time), a dramatic loss of crystallinity was observed in FAU zeolite within 40 min of fluoride etching time.This result clearly demonstrates that the large microporous system of FAU framework (12-membered ring) is extremely accessible by various fluoride species deeply into the framework aluminosilicate sites, leading to a severe and fast dissolution of both silica and alumina which results in the collapse of the FAU framework.On the other hand, MOR retained the framework structure and showed only a slight drop of relative crystallinity (~93 % at 6 h of etching time) due to its 8-and 12-membered ring structure that prevents deep penetration of fluoride species.It confirms that the NH 4 F etching method is applicable to MOR framework for generating opened pores.The MOR zeolite is characterized by its unique structure, which includes onedimensional pore channels along the c-axis -specifically, 8-MR (2.6×5.7 Å) and 12-MR (6.5×7.0Å) -as well as 8-MR side pockets (3.4×4.8Å) along the b-axis. [27]This structure suggests that NH 4 F etching during post-treatment may impact the 8-MR more significantly than the 12-MR.However, the presence of the 12-MR in the MOR topology also plays a role.It influences the dissolution of the framework's aluminosilicate species during fluoride treatment.This effect is evident in the comparative analysis of the relative crystallinity decrease in 8-and 12-MR MOR zeolites and that of the 10-MR ZSM-5 zeolite, as the NH 4 F etching time increases (refer to Figure S4).

Sample
Si/Al ratio [a] Relative crystallinity (%) [b] H-BEA-12.[a] Si/Al ratio was determined by XRF analysis.[b] Relative crystallinity (%) was calculated based on the summation of the representative peak area of two diffraction peaks at 2θ = 7.7°and 22.2°, using untreated BEA as a standard (i.e., relative crystallinity of 100 %). [30]he transportation of fluoride species into the framework aluminosilicate sites with faster rate than the smaller pore zeolites.This explanation is in agreement with the NH 4 F etching rates on various zeolite framework shown in Figure S4.The results indicated a more rapid crystallinity loss for FAU compared with MOR and ZSM-5.Additionally, FAU zeolite underwent treatment with a lower NH 4 F concentration (10 wt %).When comparing this with the normal condition (20 wt % NH 4 F), as illustrated in Figure S5, it is evident that neither concentration preserves the characteristic crystalline XRD peaks of FAU zeolite.This outcome indicates the complete destruction of the FAU framework structure by fluoride treatment, regardless of whether 10 or 20 wt % NH 4 F was used.Such results suggest that fluoride etching is not suitable for creating hierarchical structures in large-pore zeolites, e. g., 12-MR.This ineffectiveness is due to the complete destruction of the microporous framework, even at low NH4F concentrations, a consequence of the high accessibility within the large pores of FAU zeolite.Furthermore, inherent properties of the FAU framework, such as its low framework density or stability, [29] might also contribute to its susceptibility to structural damage by NH 4 F etching.

Effects of NH 4 F etching on the Si/Al ratios of the BEA and FAU zeolites
To confirm that NH 4 F etching is not appropriate for creating additional mesopores in the BEA and FAU zeolite frameworks because it destroys the microporous structures of the large pore BEA and FAU systems, NH 4 F etching was systematically performed with BEA (CP814E;Si/Al = 12.5, CP814C;Si/Al = 19, and CP811C-300;Si/Al = 150) and FAU (CBV300;Si/Al = 2.55, CBV712; Si/Al = 6, and CBV760;Si/Al = 30) zeolites containing different Si/ Al ratios.As expected, the XRD peak intensities for crystalline BEA decreased after fluoride etching of the BEA zeolites with different Si/Al ratios, indicating decreased crystallinity of the fluoride-etched BEA samples with respect to the parent BEA samples (Figure S6A and Table 1).The XRD patterns for all NH 4 F-treated FAU zeolites with various Si/Al ratios demonstrated a broad peak without the characteristic pattern for the crystalline FAU zeolite (Figure S6B).This result indicated that collapse of the BEA and FAU framework structures due to NH 4 F etching was not dependent on the structural composition (Si/Al ratio) but was due to access of fluoride into the 12-membered ring microporous systems and subsequent dissolution of the aluminosilicate species.The differential impact of NH 4 F etching on BEA and FAU may be attributed to their distinct structural characteristics.Specifically, the BEA zeolite, with its slightly smaller microporous system (6.6×6.7 Å and 5.6×5.6 Å), compared to the FAU zeolite (7.4×7.4Å), results in a higher framework density or stability for the BEA structure relative to the FAU. [26,29]We hypothesized the increased stability makes the BEA framework more resistant to NH 4 F etching.As a consequence, the BEA structure undergoes only partial destruction, whereas the FAU structure experiences complete breakdown under fluoride treatment, as shown in Figure S6.The acid densities of the BEA and FAU zeolites were also obviously reduced after NH 4 F etching, confirming the destruction of the BEA and FAU framework structures by the fluoride treatment, which was consistent with the XRD results (Figures S6, S7, and S8 and Tables S3 and S4).In addition, the acidities of zeolites are related to the structural compositions (Si/Al ratio), and a lower Si/Al ratio leads to higher acidity of the zeolite, as shown in Figures S7 and S8 and Tables S3 and S4.

Effects of different NH 4 F concentrations on fluoride etching of the ZSM-5 zeolite
Since ZSM-5 zeolite is a popular medium pore-sized zeolite and presents promising results in the construction of hierarchical structures by fluoride etching, the NH 4 F concentration was varied from 10 to 30 wt % to optimize the pretreatment conditions in which the commercial ZSM-5 (CBV2314, Si/Al = 11.5) was used as the parent ZSM-5 sample.In the XRD patterns as shown in Figure 2A, all of the ZSM-5 samples presented the characteristic crystalline pattern of the MFI zeolite framework, indicating retention of the framework structure after NH 4 F etching.However, the relative crystallinity (%) calculated from the integrated XRD peak areas at 2θ of 22.5-25° [ 31] decreased slightly with increases in the NH 4 F concentration, demonstrating partial destruction of the zeolite structure (Table 2).There were no significant changes in the zeolite framework compositions (Si/Al ratio) after fluoride etching with various NH 4 F concentrations, as indicated by the similar rates for Si and Al dissolution by the fluoride species (Table 2). [20]he SEM images of the NH 4 F-treated ZSM-5 samples illustrated the smaller particle sizes compared to those of the precursors (Figure 2B).18a] To investigate the textural and acidic properties, the prepared ZSM-5 zeolites were characterized by N 2 -sorption and NH 3 -TPD, respectively.The N 2 isotherms and BJH pore size distributions are shown in Figure 3.All ZSM-5 samples exhibited a combination of Type I and Type IV adsorption isotherms, as evidenced by the hysteresis loops indicative of both microporous and mesoporous materials (Figure 3A). [32]Notably, the ZSM-5 samples subjected to treatment with 20 and 30 wt % NH 4 F displayed significantly larger hysteresis loops compared to their counterparts.This observation is attributed to the substantially increased mesoporosity resulting from the fluoride etching process, as clearly demonstrated in Figure 3A. [33]The amounts of N 2 adsorbed increased with increasing NH 4 F concentrations, indicating higher pore volumes after treatment with higher NH 4 F concentrations (Figure 3A).It should be mentioned that the 20 and 30 wt % NH 4 F-etched ZSM-5 zeolites displayed much higher mesopore volumes (in the range of 5-45 nm pore diameters) and external surface areas than the untreated and 10 wt % NH 4 F-treated ZSM-5 samples (Figure 3B and Table 2).This result indicated that a 10 wt % NH 4 F concentration was not sufficient to create the additional mesopores of the ZSM-5 zeolite, but the hierarchical ZSM-5 structure was successfully constructed by fluoride etching with 20 and 30 wt % NH 4 F. In addition, 20NH 4 F-H-ZSM-5 and 30NH 4 F-H-ZSM-5 samples also demonstrated increased specific surface areas (S BET ) and micropore surface areas (S micro ) compared to the parent sample.This enhancement could be attributed to the NH 4 F etching process.Normally, NH 4 F etching creates additional mesopores, thus contributing to a higher external surface area.However, in certain cases, NH 4 F etching extends its influence beyond merely enhancing the external surface area; it also significantly improves the accessible micropore surface area.This is achieved through the pore-opening observed after NH 4 F treatment, which results in the exposure of more surface areas.This dual effect of NH 4 F etching, i. e., splitting of the large zeolite crystals into various small particles and the opening of pores, is evidenced in Table 2 and Figure 2B. [34]This phenomenon suggests that the accessibility to the microporous framework can be improved without progressive structure destruction or loss of crystallinity.For the acidic properties, the NH 3 -TPD profiles and the numbers of acid sites are demonstrated in Figure 4 and Table 3, respectively.Interestingly, ZSM-5 samples etched with 10 and 20 wt % NH 4 F showed negligible differences Table 2. Structural compositions (Si/Al ratios), relative crystallinities (%), and textural properties of ZSM-5 zeolites before and after NH 4 F etching with different NH 4 F concentrations.
[d] S ext : external surface area.
[g] V micro : micropore volume.[a] The numbers of acid sites were determined with NH 3 -TPD analyses.
in NH 3 -TPD profiles and acid densities when compared to untreated ZSM-5 (H-ZSM-5).However, the sample treated with 30 wt % NH 4 F, referred to as 30NH 4 F-H-ZSM-5, exhibited a notable reduction in acid sites.This suggests a partial structural collapse of the ZSM-5 zeolite due to etching with the higher NH 4 F concentration, as evidenced in Figure 4 and Table 3.The results suggest that 20 wt % NH 4 F is the optimal fluoride treatment condition for producing hierarchical ZSM-5 zeolite.This concentration achieves high mesoporosity and crystallinity, while preserving the distinctive properties of the ZSM-5 zeolite.
On the other hand, excessive NH 4 F etching, e. g., 30 wt %, may lead to framework destabilization.It is possible that, at 30 wt %, the increased availability of fluoride ions may interact strongly with both the silicon and the aluminum sites.This interaction could disrupt the crucial spatial arrangement that sustains acid sites, culminating in the removal of aluminum and consequently weakening the structural framework.Additionally, factors like alterations in pore structure and shifts in surface chemistry might also play a role in the diminution of acid sites.
However, further research is necessary to confirm these assertions. 27Al MAS NMR spectroscopy was used to investigate the changes in the Al species after fluoride treatment.The 27 Al MAS NMR spectrum exhibited two bands at 55 and 0 ppm, which were attributed to the tetrahedral framework Al and octahedral extra-framework Al species, respectively (Figure S9). [23]The tetrahedral-to-octahedral Al peak area ratio of the 20NH 4 F-H-ZSM-5 sample was lower than that of the parent H-ZSM-5, indicating the dissolution of framework Al sites and creation of extra-framework Al species during the NH 4 F etching process (Table S5).It is possible that, in the NH 4 F etching process, the aqueous solution of NH 4 F, under equilibrium conditions, produces a range of fluoride species, notably F À and HF 2 À .This chemical environment facilitates a mild and controlled dissolution of the Si and Al species within the framework.18a] NH 4 F etching affects zeolite crystals across their entire surface.The process initially targets the interfaces between intergrown crystals and grain boundaries, as these are the preferred zones for chemical attack due to their defect characteristics.Subsequently, fluoride ions penetrate deeper into the crystals.18a]

Catalytic activities in methanol and bioethanol conversion
Both methanol and ethanol are important feedstocks for biofuel production [6a,35] and can be prepared from biomass via fermentation of sugar cane, wheat, and corn or wastes. [36]To investigate the catalytic activities of the fluoride-treated zeolites systematically, methanol and bioethanol conversions were catalyzed by the ZSM-5 catalysts.Figure 5 shows the methanol conversions (%) as a function of time-on-stream (h) over the synthesized ZSM-5 catalysts before and after fluoride etching with different NH 4 F concentrations.Although all catalysts initially show a 100 % methanol conversion rate in the first hour, as depicted in Figure 5, the catalyst treated with 20 wt % NH 4 F stands out.It uniquely maintains this complete methanol conversion rate (100 %) over an extended period (1-4 h timeon-stream, TOS).This higher catalytic activity correlates with the increased mesoporosity and comparable acidity of the 20 wt % NH 4 F-treated ZSM-5 catalyst, relative to the parent ZSM-5, as detailed in Figure 5 and Tables 2 and 3.
On the other hand, the 30NH 4 F-H-ZSM-5 catalyst gave the lowest methanol conversion (71-98 %) during the initial stage of the reaction (1-5 h time-on-stream) according to the lowest acid density (total number of acid sites ~0.929 mmol g À 1 ) resulted from the loss of acidity after the fluoride treatment with high NH 4 F concentration (30 wt %) (Figure 5 and Table 3).It was found that various hydrocarbon species, including C1-C7 alkane and alkene hydrocarbons and aromatics (BTX: benzene, toluene, and xylene), were produced with a total carbon selectivity of 100 % for the catalytic methanol conversion occurring over the prepared ZSM-5 catalysts at the early reaction stages (1-5 h); dimethyl ether (DME) became a major product (with a carbon selectivity of ~96 %) after catalyst deactivation, demonstrating that DME was not converted into other hydrocarbon products due to the loss of available active sites caused by coke deposition and pore blocking (Figure S10). [37]The catalytic activity results in methanol conversion were also analyzed based on the oxygenate (methanol and DME) conversion following the literature as shown in Figure S11. [38]The results distinctly indicate that after catalyst deactivation, the oxygenate conversion rate is markedly low (less than 5 %).This observation confirms that DME conversion into other hydrocarbons is significantly hindered post-deactivation.The primary reason for this reduction in activity is attributed to the loss of active sites, which is caused by coke deposition and pore blocking within the zeolite catalysts.In addition, given methanol (kinetic diameter of 3.6 Å) is not significantly restricted by mass transfer inside ZSM-5, there is no conclusive evidence to suggest that the ZSM-5 catalysts after fluoride etching may contribute positively to methanol con- version from the catalyst stability perspective (which should be analyzed at differential conversion regime).
Figure 6 compares the catalytic activities of the synthesized H-ZSM-5 and 20NH 4 F-H-ZSM-5 catalysts for bioethanol conversion at 623 K.It can be seen that while the bioethanol conversion over the H-ZSM-5 catalyst dramatically dropped from ~90 to ~50 % between 1 to 34 h time-on-stream (Figure 6A), the 20NH 4 F-H-ZSM-5 catalyst displayed a stable bioethanol conversion with more than ~85 % over 34 h TOS (Figure 6B).7b] It was observed that the product selectivity of diethyl ether (DEE) and ethylene increased and decreased, respectively, with increasing TOS over the untreated ZSM-5 catalyst (H-ZSM-5).The change of this product distribution correlates with the decrease in bioethanol conversion observed with increasing TOS.These changes indicate that ethylene production, which typically occurs via the DEE decomposition pathway into ethylene and ethanol, is significantly reduced following catalyst deactivation.The primary cause of this reduction is likely related to the loss of active sites due to coke deposition.In contrast, the fluoride-treated ZSM-5 catalyst (20NH 4 F-H-ZSM-5) achieved a high ethylene selectivity with more than ~70 % over 34 h TOS possibly due to the facile desorption of desired product and the short diffusion path length resulted from the hierarchical open-pore architecture.These high catalytic activities of the 20NH 4 F-H-ZSM-5 catalyst (> 85 % of bioethanol conversion and > 70 % of ethylene selectivity over 34 h TOS) is comparable to the catalytic activities of the hierarchical ZSM-5 nanosheets catalyst reported in the literature (> 75 % of bioethanol conversion over 15 h TOS and > 60 % of ethylene selectivity at 15 h TOS), [39] which were resulted from the high external surface area and mesoporosity of the hierarchical pore-opened and nanosheet structures.2c] Moreover, the fluoride-treated ZSM-5 zeolite has also been evaluated for its catalytic activity in ethanol conversion processes in the literature. [20]These studies reveal that the fluoride-treated zeolite demonstrates superior catalytic performance compared to its microporous ZSM-5 counterpart, a finding that aligns with our current research.Notably, while large hydrocarbons (C 3 + ) were predominantly produced in the ethanol conversion processes documented in previous studies, [20] our work with NH 4 F-treated ZSM-5 catalysts primarily yielded ethylene as the major product from bioethanol conversion.This suggests that variations in the operational conditions of catalytic activity tests, along with differences in crystal morphology and porosity of the fluoride-treated zeolite catalysts, can lead to diverse product distributions from the same catalytic reaction.From the reported catalytic methanol and bioethanol conversion results, it was revealed that the hierarchical open-pore structure created by the NH 4 F etching presents a better effect on reactions with larger reactants, such as ethanol.Furthermore, the catalytic activity results of methanol and bioethanol conversions were also reported as numeric values in Tables S6 and S7, respectively.
The catalytic activity results suggest that the accessibility of reactant molecules, such as methanol (3.6 Å) and ethanol (4.3 Å), [41] to the active sites of zeolite catalysts is significantly improved by the increased mesoporosity resulting from fluoride treatment.This enhancement is due to the additional mesopores creating shorter diffusion paths and facilitating easier molecular transport.As a result, zeolite catalysts treated with NH 4 F exhibit greater accessibility for these guest molecules to reach the acid sites, compared to their untreated counterparts.This increased accessibility likely contributes to the observed improvements in catalytic performance.

Conclusions
The effects of different zeolite frameworks, including those of FER, MOR, MFI, BEA, and FAU zeolites, on the fabrication of hierarchical structures via NH 4 F etching were systematically investigated.It was found that the fluoride-treated small and medium porosity FER, MOR, and ZSM-5 zeolites exhibited higher mesoporosities with insignificant changes in their structural compositions (Si/Al ratio) and acidic properties compared to the parent zeolites, indicating successful fabrication of hierarchical open-pore zeolites via NH 4 F etching.Our hypothesis suggests that fluoride etching has a detrimental impact on the development of hierarchical structures within large pore zeolite frameworks.This is particularly evident in FAU zeolite, where the microporous framework structure is compromised.The hypothesis posits that this structural collapse is due to the heightened accessibility of fluoride species to the framework's aluminosilicate sites.The effects of different NH 4 F concentrations (10-30 wt % NH 4 F aqueous solutions) on fluoride etching of the ZSM-5 zeolite were investigated to optimize the treatment conditions.The results showed that a 10 wt % NH 4 F concentration was not sufficient to create additional mesopores in the ZSM-5 zeolite structure and that 30 wt % NH 4 F led to significant losses of crystallinity and acidity.The 20 wt % NH 4 Fetched ZSM-5 sample displayed a much higher mesoporosity (V meso /V micro ratio � 1.5) with a comparable acidity (total number of acid sites � 1.285 mmol g À 1 ) compared to the parent sample (V meso /V micro ratio � 1.1 and a total number of acid sites of 1.323 mmol g À 1 ), indicating that 20 wt % NH 4 F was the optimum concentration.The 20NH 4 F-H-ZSM-5 sample exhibited a much higher catalytic performance in bioalcohol conversion with over 85 % bioethanol conversion after 34 h compared to the untreated ZSM-5 (~50 % of bioethanol conversion at 34 h TOS).These results indicated higher accessibility of the guest molecules into the active sites due to the hierarchical openpore structure of the fluoride-treated ZSM-5 sample.Consequently, this work provides important information on the limitations and benefits of post-synthetic NH 4 F treatments for efficient fabrication of hierarchical zeolite catalysts.

Post-synthetic treatments of zeolites via NH 4 F etching
Etching of the zeolites with NH 4 F was performed with the procedure reported in our previous publication. [21]A certain amount of NH 4 F was dissolved in deionized (DI) water with the assistance of ultrasonication.Commercial zeolite powder was added to the prepared NH 4 F aqueous solution (1 g of zeolite per 20 g of NH 4 F solution).After that, the mixed solution was stirred at 323 K for 2 h in an oil bath.The fluoride-treated zeolite solid was washed with a large amount of DI water (200 ml per g of zeolite) with centrifugation at 9000 rpm.Subsequently, the zeolite product was dried at 373 K overnight in an oven and calcined at 823 K for 6 h using a horizontal furnace with an air flow (50 ml min À 1 g-cat À 1 ) to convert the zeolites into their protonated (H) forms.In the series of zeolite frameworks, the samples were treated with 20 wt % NH 4 F, and the untreated and fluoride-treated samples were denoted as Hzeolite and NH 4 F-H-zeolite, respectively.In the case of BEA and FAU zeolites with various Si/Al ratios, the samples before and after fluoride etching were denoted as H-zeolite-X and NH 4 F-Hzeolite-X, respectively, where X is the Si/Al ratio of each sample.The ZSM-5 samples made by etching with 10, 20, and 30 wt % NH 4 F were designated 10NH 4 F-H-ZSM-5, 20NH 4 F-H-ZSM-5, and 30NH 4 F-H-ZSM-5, respectively.

Characterization
The crystalline structure of the zeolites were investigated by powder X-ray diffraction (PXRD) performed with a SmartLab Xray diffractometer (Rigaku, Tokyo) with a CuÀ Kα radiation source (λ = 0.154 nm) operated at 40 kV and 30 mA.Scanning electron microscopy (SEM) images were recorded with a JSM-7000F field emission scanning electron microscope (JEOL, Japan) at 15.0 kV.The structural compositions (Si/Al ratio) were determined with a wavelength-dispersive X-ray fluorescence spectrometer (ZSX Primus IV, Rigaku).The textural properties were analyzed via N 2 adsorption/desorption performed with a N 2 -AutoSorb IQ instrument (Anton Paar).The specific surface areas (S BET ), external surface areas (S ext ), and micropore volumes (V micro ) were calculated with the Brunauer-Emmett-Teller (BET) and t-plot methods.The total pore volumes (V total ) were determined at a P/P 0 of 0.98.The acidic properties were investigated with temperature-programmed desorption of ammonia (NH 3 -TPD) performed on a BELCAT II instrument equipped with a thermal conductivity detector (TCD).The NH 3 -TPD profiles were recorded in the temperature range 323-1073 K.The acid density was calculated based on the amount of desorbed NH 3 molecules using the ChemMaster software. 27Al MAS NMR spectra were collected with a Bruker Avance III HD/Ascend 400 WB.

Catalytic activity testing in methanol conversion
The catalyst powder was pelletized with a hydraulic press and sieved to 20-40 mesh size.Catalytic testing was conducted with a continuous fixed-bed flow reactor packed with 0.2 g of catalyst with glass wool used as the catalyst support.Before testing, the catalysts were pretreated with 40 ml min À 1 N 2 gas at 673 K for at least 2 h.Then, methanol was injected at 2 ml h À 1 with a syringe pump and vaporized at 423 K with 80 ml min À 1 N 2 gas as a carrier.The mixture was introduced into the reactor tube at 673 K.The outlet gases were analyzed with an online gas chromatograph (GC-2030, Australia_MTG System, Shimadzu Corporation) equipped with a flame ionization detector (FID).The conversion (%) and product selectivity (%) were calculated based on the relative response factor (RRF) of the FID detector. [42]The turnover frequency (TOF) was defined as the number of methanol molecules converted per number of acid sites per second. [21,43]

Catalytic activity testing in bioethanol conversion
The catalyst powder was pelletized with a hydraulic press and sieved to 20-40 mesh size.Catalytic testing was conducted in a continuous fixed-bed flow reactor packed with 0.2 g of catalyst with quartz wool used as a catalyst support.Before testing, the catalysts were pretreated with 40 ml min À 1 N 2 gas at 623.15 K for at least 2 h.Then, the bioethanol vapor feed was introduced into the reactor with a N 2 saturation system (293.15K, 15 ml min À 1 N 2 flow, and a weight hourly space velocity (WHSV) of 0.54 h À 1 ) to perform the reaction at 623.15 K.The outlet gases were analyzed with an online gas chromatograph (8860 GC System, Agilent Technologies) equipped with an FID detector and a PoraBOND Q capillary column.The conversion (%) and product selectivity (%) were calculated based on the relative response factor (RRF) of the FID detector according to the following Equations 1-3: [42] Mi ¼ Ai � RRFi � 100 P ðA � RRFÞ (1) where Mi , Ai , and RRFi are the mass percentage, integrated GC area, and relative response factor of an identified hydrocarbon compound i, respectively. [42]

Figure 1 .
Figure 1.XRD patterns of various zeolite samples with different zeolite frameworks before and after NH 4 F etching; (A) FER, MOR, and ZSM-5 and (B) BEA and FAU.

Figure 2 .
Figure 2. (A) XRD patterns and (B) SEM images of ZSM-5 zeolites before and after NH 4 F etching with different NH 4 F concentrations.

Figure 3 .
Figure 3. (A) N 2 isotherms and (B) BJH pore size distributions for ZSM-5 zeolites before and after NH 4 F etching with different NH 4 F concentrations.

Figure 4 .
Figure 4. NH 3 -TPD profiles of ZSM-5 zeolites before and after NH 4 F etching with different NH 4 F concentrations.

Figure 5 .
Figure 5. Methanol conversion rate (%) as a function of time-on-stream (TOS) in the catalytic methanol conversions over ZSM-5 catalysts before and after NH 4 F etching with different NH 4 F concentrations (reaction temperature of 673 K, WHSV ~303 h À 1 , catalyst loading ~0.2 g).

Table 1 .
Structural composition (Si/Al ratio) and relative crystallinity (%) of BEA zeolite with different Si/Al ratios before and after NH 4 F etching.

Table 3 .
Number of acid sites in the ZSM-5 zeolites before and after NH 4 F etching with different NH 4 F concentrations.