Renewable Butene Production through Dehydration Reactions over Nano-HZSM-5 / γ -Al 2 O 3 Hybrid Catalysts

: The development of new, improved zeolitic materials is of prime importance to progress heterogeneous catalysis and adsorption technologies. The zeolite HZSM-5 and metal oxide γ -Al 2 O 3 are key materials for processing bio-alcohols, but both have some limitations, i.e., HZSM-5 has a high activity but low catalytic stability, and vice versa for γ -Al 2 O 3 . To combine their advantages and suppress their disadvantages, this study reports the synthesis, characterization, and catalytic results of a hybrid nano-HZSM-5 / γ -Al 2 O 3 catalyst for the dehydration of n-butanol to butenes. The hybrid catalyst is prepared by the in-situ hydrothermal synthesis of nano-HZSM-5 onto γ -Al 2 O 3 . This catalyst combines mesoporosity, related to the γ -Al 2 O 3 support, and microporosity due to the nano-HZSM-5 crystals dispersed on the γ -Al 2 O 3 . HZSM-5 and γ -Al 2 O 3 being in one hybrid catalyst leads to a di ﬀ erent acid strength distribution and outperforms both single materials as it shows increased activity (compared to γ -Al 2 O 3 ) and a high selectivity to oleﬁns, even at low conversion and a higher stability (compared to HZSM-5). The hybrid catalyst also outperforms a physical mixture of nano-HZSM-5 and γ -Al 2 O 3 , indicating a truly synergistic e ﬀ ect in the hybrid catalyst.


Introduction
Due to increasing environmental awareness, there has been a shift toward bio-based products and the design of sustainable, renewable technologies [1][2][3]. In light of a transition to a more sustainable chemical industry, there has been a lot of interest in the production and use of bio-alcohols [4,5]. Bio-alcohols can be used as additives to fuels [6,7], as solvents, or even as feedstock for chemical processing [8,9]. Alcohols can undergo dehydration to olefins, which are of the most extensively used building blocks in the chemical industry [10,11].
Zeolites and γ-Al 2 O 3 could play a key role in the transition from a fossil-based industry into a renewable industry as they are of the most important materials used in traditional oil processing and heterogeneous catalysis [12][13][14][15]. Zeolites, and, more specifically, HZSM-5, are essential materials in mature technologies such as fluid catalytic cracking (FCC) [16,17] and alkylation reactions [18,19], but also in more recent technologies such as methanol-to-olefins (MTO) [20,21]. HZSM-5 is known for its shape selectivity, strong intrinsic acidity [19], and thermal stability, but is also prone to coking [22,23].
Catalysts 2020, 10 Alumina materials, on the other hand, are of the most important catalyst supports, for example, γ-Al 2 O 3 is used in reforming reactions [24], hydrotreating, and hydroconversion [25][26][27]. Alumina materials are widely used due to their low cost, thermal stability, fairly high surface area, and pore volume [25,28,29]. To convert alcohols, e.g., ethanol or butanol, to olefins, two catalysts, (i) HZSM-5 and (ii) γ-Al 2 O 3 , are the most investigated materials [30][31][32][33][34][35], and are already used at an industrial scale for ethanol dehydration [32,36]. Although HZSM-5 and γ-Al 2 O 3 are extensively used, they have their limits. For the dehydration of bio-alcohols, HZSM-5 typically has a very high activity and selectivity to the corresponding olefins; however, the stability is poor due to coking [37]. Furthermore, beyond full conversion, secondary reactions are also catalyzed [38]. For γ-Al 2 O 3 , it is the opposite: It has a low activity, requiring higher temperatures; moreover, a high selectivity toward olefins also requires sufficiently high temperatures [32]. On the other hand, the catalyst has a high stability under dehydration reaction conditions. Their performance is somewhat the opposite and can be related to their very different properties, summarized in Table 1. Table 1. Characteristic properties of HZSM-5 and γ-Al 2 O 3 [39,40]. To avoid these shortcomings, there have been studies on modifying these catalysts. Decreasing the HZSM-5 crystal size to nanocrystals has shown an increase in the catalyst life time [41] and is often associated with an increase in activity [42]. This increase in stability and activity is often related to the decrease in diffusion pathlength inside the micropores and/or lowered deactivation due to coking. Other strategies include inducing mesoporosity [43] and modification of the acid sites by post-synthesis treatments [44], e.g., phosphorus-modifications decrease the strength of the acid sites and thereby attenuate coke formation [45]. Instead of creating mesopores inside the zeolite particle, it is also possible to synthesize the zeolite on a mesoporous material (such as MCM-41) [46]: This can enhance the performance by suppressing zeolite crystal aggregation, inducing dilution of the zeolitic Brønsted acid centers, and creating mesoporosity, which leads to a lower acid density and can attenuate diffusional restrictions [47,48].

Material
For γ-Al 2 O 3 , the modifications typically aim to increase the activity and allow lower reaction temperatures [32]. By modifying the γ-Al 2 O 3 with TiO 2 , the acidity increased, resulting in a higher activity. Modification of the γ-Al 2 O 3 surface with Zn-Mn-Co increases the acid strength and crystallinity, which results in an increased activity [49,50]. He et al. found that the combination of γ-Al 2 O 3 and HZSM-5 into one catalyst leads to an increased selectivity toward light olefins for the pyrolysis of waste tires [51]. Due to their interesting findings, and the fact that HZSM-5 and γ-Al 2 O 3 are two widely used catalytically active materials for the dehydration of alcohols, the objective of this work is to develop a hybrid catalyst that is a combination of HZSM-5 and γ-Al 2 O 3 . The combination is aimed to be in such a manner that the advantages of both catalysts are combined, namely (i) a high activity and selectivity to olefins related to HZSM-5 nanocrystals, i.e., no high temperatures are needed for bio-alcohol dehydration, and (ii) a strong resistance to deactivation related to the presence of γ-Al 2 O 3 . To achieve these synergistic effects, the nano-HZSM-5 is synthesized upon γ-Al 2 O 3 , which acts as a catalytically active support and leads to mesopores with mildly active sites in the vicinity of the active sites of HZSM-5. To this end, a hybrid catalyst, composed of in-situ-synthesized nano-HZSM-5 crystals dispersed on a γ-Al 2 O 3 active support, was manufactured. To the best of our knowledge, this is the first report where the synergistic effects of combining both catalysts in one hybrid material is exploited for the dehydration of bio-alcohols to olefins. The dehydration of n-butanol at 513 K is used as a test reaction to compare the hybrid catalyst to HZSM-5, γ-Al 2 O 3 , and a physical mixture of both. The catalytic performance (activity/selectivity/stability) is then related to the properties of the studied materials.

Catalyst Characterization
The XRD patterns of the commercially available HZSM-5 (c-HZSM-5), the synthesized nano-HZSM-5 (n-HZSM-5), the physical mixture of nano-HZSM-5 and γ-Al2O3 (PM-50/50), the hybrid catalyst (Hybrid-50/50), and commercially available γ-Al2O3 are presented in Figure 1. The diffraction patterns between 7.5 and 9° and between 22.5° and 26° are attributed to the ZSM-5 structure. The intensity in these ranges of 2θ is the highest for c-HZSM-5 and n-HZSM-5, which is as expected as these are pure zeolite phases, while PM-50/50 and Hybrid-50/50 are mixtures of zeolite phases and γ-Al2O3. Albeit a lower intensity, the diffraction pattern at these angles is identical, which indicates that the Hybrid-50/50 synthesis is successful and the ZSM-5 crystals can be synthesized alongside γ-Al2O3. The peaks related to the HZSM-5 phase are for all catalysts sharp, and have a high intensity, which is typical for highly crystalline HZSM-5 [52]. In the pattern of PM-50/50 and Hybrid-50/50, broader diffraction peaks due to the γ-Al2O3 phases are also present, indicating that both catalysts consist of both HZSM-5 and γ-Al2O3. The mean HZSM-5 crystal size determined by the Scherrer equation, Equation (1), is 160 nm for c-HZSM-5, 118 nm for n-HZSM-5, and 114 nm for Hybrid-50/50. As the HZSM-5 particles in PM-50/50 are the n-HZSM-5 ones, their mean crystal size should also be 118 nm.
N2-sorption isotherms of the various materials are displayed in Figure 2. From the isotherms, it is clear that c-HZSM-5 and n-HZSM-5 are highly microporous and show a type Ia isotherm according to the Brunauer-Deming-Deming-Teller classification. This type of isotherm corresponds to microporous solids with narrow micropores (width < 1 nm) [8]. Where n-HZSM-5 is almost completely microporous, there is some mesoporosity in c-HZSM-5 noticeable. The isotherm of c-HZSM-5 has a hysteresis of type H4, which can be ascribed to the capillary condensation of N2 in interparticles or in crystal agglomerates [53]. The γ-Al2O3, PM-50/50, and Hybrid-50/50 clearly display strong mesoporosity, although there is also microporosity noticeable in the latter two due to the presence of HZSM-5 crystals. The isotherm of γ-Al2O3 is of type IV, while for PM-50/50 and Hybrid-50/50, the isotherm starts off as a type I but evolves into a type IV. The diffraction patterns between 7.5 and 9 • and between 22.5 • and 26 • are attributed to the ZSM-5 structure. The intensity in these ranges of 2θ is the highest for c-HZSM-5 and n-HZSM-5, which is as expected as these are pure zeolite phases, while PM-50/50 and Hybrid-50/50 are mixtures of zeolite phases and γ-Al 2 O 3 . Albeit a lower intensity, the diffraction pattern at these angles is identical, which indicates that the Hybrid-50/50 synthesis is successful and the ZSM-5 crystals can be synthesized alongside γ-Al 2 O 3 . The peaks related to the HZSM-5 phase are for all catalysts sharp, and have a high intensity, which is typical for highly crystalline HZSM-5 [52]. In the pattern of PM-50/50 and Hybrid-50/50, broader diffraction peaks due to the γ-Al 2 O 3 phases are also present, indicating that both catalysts consist of both HZSM-5 and γ-Al 2 O 3 . The mean HZSM-5 crystal size determined by the Scherrer equation, Equation (1), is 160 nm for c-HZSM-5, 118 nm for n-HZSM-5, and 114 nm for Hybrid-50/50. As the HZSM-5 particles in PM-50/50 are the n-HZSM-5 ones, their mean crystal size should also be 118 nm. N 2 -sorption isotherms of the various materials are displayed in Figure 2. From the isotherms, it is clear that c-HZSM-5 and n-HZSM-5 are highly microporous and show a type Ia isotherm according to the Brunauer-Deming-Deming-Teller classification. This type of isotherm corresponds to microporous solids with narrow micropores (width < 1 nm) [8]. Where n-HZSM-5 is almost completely microporous, there is some mesoporosity in c-HZSM-5 noticeable. The isotherm of c-HZSM-5 has a hysteresis of type H4, which can be ascribed to the capillary condensation of N 2 in interparticles or in crystal agglomerates [53]. The γ-Al 2 O 3 , PM-50/50, and Hybrid-50/50 clearly display strong mesoporosity, although there is also microporosity noticeable in the latter two due to the presence of HZSM-5 crystals. The isotherm of γ-Al 2 O 3 is of type IV, while for PM-50/50 and Hybrid-50/50, the isotherm starts off as a type I but evolves into a type IV. From these isotherms, it is possible to gain insights into how the synthesis of the catalysts influences their textural properties: Specific surface area, specific pore volume, and pore size distribution analysis are performed (see Figure 3 and Table 2). The specific surface area of c-HZSM-5 is the highest, followed by n-HZSM-5. PM-50/50 and Hybrid-50/50 have intermediate surface areas, whereas γ-Al2O3 has the lowest surface area. The obtained pore sizes of c-HZSM-5 and n-HZSM-5 (Error! Reference source not found., a) show good agreement with the pore diameter of approximately 0.5 nm for the MFI framework. Furthermore, there are some small mesopores measured (dp about 2-2.5 nm). Analyzing the pore structures (Error! Reference source not found., b) shows that the mesopore volume is largest for γ-Al2O3. The mesoporosity measured for PM-50/50 is attributed to the γ-Al2O3, which is present in the mixture, and the pore dimensions of the mesopores are unaltered for PM-50/50, although the mesopore volume has dropped as expected. For the Hybrid-50/50, a much stronger decrease in mesopore volume is noticeable, although the pore diameter of the mesopores is unaltered, meaning that, most likely, there are no HZSM-5 crystals inside the mesopores of the γ-Al2O3 support, but they are instead formed on the external surface.  From these isotherms, it is possible to gain insights into how the synthesis of the catalysts influences their textural properties: Specific surface area, specific pore volume, and pore size distribution analysis are performed (see Figure 3 and Table 2). The specific surface area of c-HZSM-5 is the highest, followed  Figure 3b) shows that the mesopore volume is largest for γ-Al 2 O 3. The mesoporosity measured for PM-50/50 is attributed to the γ-Al 2 O 3 , which is present in the mixture, and the pore dimensions of the mesopores are unaltered for PM-50/50, although the mesopore volume has dropped as expected. For the Hybrid-50/50, a much stronger decrease in mesopore volume is noticeable, although the pore diameter of the mesopores is unaltered, meaning that, most likely, there are no HZSM-5 crystals inside the mesopores of the γ-Al 2 O 3 support, but they are instead formed on the external surface. From these isotherms, it is possible to gain insights into how the synthesis of the catalysts influences their textural properties: Specific surface area, specific pore volume, and pore size distribution analysis are performed (see Figure 3 and Table 2). The specific surface area of c-HZSM-5 is the highest, followed by n-HZSM-5. PM-50/50 and Hybrid-50/50 have intermediate surface areas, whereas γ-Al2O3 has the lowest surface area. The obtained pore sizes of c-HZSM-5 and n-HZSM-5 (Error! Reference source not found., a) show good agreement with the pore diameter of approximately 0.5 nm for the MFI framework. Furthermore, there are some small mesopores measured (dp about 2-2.5 nm). Analyzing the pore structures (Error! Reference source not found., b) shows that the mesopore volume is largest for γ-Al2O3. The mesoporosity measured for PM-50/50 is attributed to the γ-Al2O3, which is present in the mixture, and the pore dimensions of the mesopores are unaltered for PM-50/50, although the mesopore volume has dropped as expected. For the Hybrid-50/50, a much stronger decrease in mesopore volume is noticeable, although the pore diameter of the mesopores is unaltered, meaning that, most likely, there are no HZSM-5 crystals inside the mesopores of the γ-Al2O3 support, but they are instead formed on the external surface.     The results extracted from the NH 3 -TPD are summarized in Table 3. The spectra at a ramp rate of 10 K min −1 are shown in Figures S2 and S3. From these results, it is clear that the Hybrid-50/50 catalyst, its density of strong acid sites, and its total amount of acid sites are substantially lower than those for the other catalysts. This can be partially explained by the synthesis method: During the synthesis, the γ-Al 2 O 3 is added to the zeolite precursor mixture, which is at a pH of about 12. The hydroxyl groups (originating from tetrapropylammonium hydroxide (TPAOH)) in this mixture can interact with the acid sites on the γ-Al 2 O 3 and even partially dissolve the alumina [54], thereby decreasing the number of acid sites. Furthermore, the HZSM-5 crystals being on the γ-Al 2 O 3 surface can block some of the previously available acid sites. Not only has the amount of acid sites changed, but so has the distribution of the acid strength, where n-HZSM-5 and γ-Al 2 O 3 primarily have "strong" acid sites, and the Hybrid-50/50 has more "medium" acidic active sites than "strongly" acidic. Based on the NH 3 -desorption energy, the NH 3 is adsorbed the strongest on n-HZSM-5 and the weakest on γ-Al 2 O 3 . Both PM-50/50 and Hybrid-50/50 have intermediate NH 3 -desorption energies, although for Hybrid-50/50, which consists of nano-HZSM-5 crystals and γ-Al 2 O 3 , the E d is almost the same as for c-HZSM-5. Through SEM, the morphological details of the catalysts are studied, Figure 4, and further magnified images are shown in Figure S4. The c-HZSM-5 has irregularly shaped, aggregated crystals, confirming the possibility of intercrystalline mesoporosity, which was deducted from N 2 -sorption. The crystal size ranges from about 80 to 600 nm. The n-HZSM-5 has far more regularly hexagonal or coffin-shaped nanocrystals (see Figure S4) and a much smaller range of crystal sizes (about 80-180 nm). For PM-50/50, both separate n-HZSM-5 and γ-Al 2 O 3 phases are noticeable. From the SEM images, no interaction between both phases is observable. For the Hybrid-50/50, on the other hand, it is clearly noticeable that the n-HZSM-5 crystals are dispersed on the surface of the γ-Al 2 O 3 , rather than having two separate phases. Therefore, it seems likely that the in-situ synthesis of HZSM-5 in the presence of γ-Al 2 O 3 leads to an interaction between the HZSM-5 nanocrystals and the γ-Al 2 O 3 support. The γ-Al 2 O 3 particles are regularly shaped spheres with a diameter between 100 and 150 µm.

Activity Analysis
To compare the activity of the catalysts, the n-butanol dehydration was performed at a single temperature of 513 K, as γ-Al 2 O 3 and HZSM-5 display some activity for the dehydration of n-butanol at

Activity Analysis
To compare the activity of the catalysts, the n-butanol dehydration was performed at a single temperature of 513 K, as γ-Al2O3 and HZSM-5 display some activity for the dehydration of n-butanol at this temperature [38], see Figure 5. For the dehydration of n-butanol, the order of activity is c-HZSM-5 > n-HZSM-5 > Hybrid-50/50 > PM-50/50 >> γ-Al2O3.

Activity Analysis
To compare the activity of the catalysts, the n-butanol dehydration was performed at a single temperature of 513 K, as γ-Al2O3 and HZSM-5 display some activity for the dehydration of n-butanol at this temperature [38], see Figure 5. For the dehydration of n-butanol, the order of activity is c-HZSM-5 > n-HZSM-5 > Hybrid-50/50 > PM-50/50 >> γ-Al2O3. n-HZSM-5 show the highest activity, most likely related to their re sizes, which increases the interaction with the reactant [55].
-5 compared to n-HZSM-5 can be attributed to the high ratio sites. The activity, expressed as site time yield (STY, defined in m and strong acid sites) is linearly related to this ratio for every e Figure 6), showing that the catalyst is acting non-linear nction of the ratio of the amount of strong acid sites over medium ack), n-HZSM-5 (, red), PM-50/50 (, purple), Hybrid-50/50 (, rature = 513 K, inlet pressure of n-butanol = 29 kPa, total pressure ctivity generally increases linearly with increasing ratio of the medium acid sites. The only catalyst strongly deviating from that this is due to a synergistic interaction between γ-Al2O3 and brid catalyst. Furthermore, the nano-HZSM-5 crystals being creases the accessibility of the zeolitic or Brønsted acid sites, tive site. Furthermore, Wang et al. found that if there are Lewis f Brønsted acid sites in HZSM-5, the dehydration of ethanol is s of Hybrid-50/50 are more active than what would be expected, . , green), and γ-Al 2 O 3 (•, blue). Temperature = 513 K, inlet pressure of n-butanol = 29 kPa, total pressure = 5 bar. Error bars indicate the 95% confidence interval.
As expected, the c-HZSM-5 and n-HZSM-5 show the highest activity, most likely related to their strong acidity and relatively small pore sizes, which increases the interaction with the reactant [55]. The slightly higher activity of c-HZSM-5 compared to n-HZSM-5 can be attributed to the high ratio of strong acid sites over medium acid sites. The activity, expressed as site time yield (STY, defined in Equation (5), comprising both medium and strong acid sites) is linearly related to this ratio for every Catalysts 2020, 10, 879 7 of 17 catalyst except for Hybrid-50/50 (see Figure 6), showing that the catalyst is acting non-linear compared to its constituents. of n-butanol = 29 kPa, total pressure = 5 bar. Error bars indicate the 95% confidence interval.
As expected, the c-HZSM-5 and n-HZSM-5 show the highest activity, most likely related to their strong acidity and relatively small pore sizes, which increases the interaction with the reactant [55]. The slightly higher activity of c-HZSM-5 compared to n-HZSM-5 can be attributed to the high ratio of strong acid sites over medium acid sites. The activity, expressed as site time yield (STY, defined in Equation (5), comprising both medium and strong acid sites) is linearly related to this ratio for every catalyst except for Hybrid-50/50 (see Figure 6), showing that the catalyst is acting non-linear compared to its constituents. As can be seen in Figure 6, the activity generally increases linearly with increasing ratio of the densities of the strong acid sites over medium acid sites. The only catalyst strongly deviating from this trend is Hybrid-50/50. We expect that this is due to a synergistic interaction between γ-Al2O3 and the nano-HZSM-5 crystals in the hybrid catalyst. Furthermore, the nano-HZSM-5 crystals being dispersed on the alumina surface increases the accessibility of the zeolitic or Brønsted acid sites, leading to a possibly more efficient active site. Furthermore, Wang et al. found that if there are Lewis acid sites introduced in the vicinity of Brønsted acid sites in HZSM-5, the dehydration of ethanol is facilitated [56]. Although the acid sites of Hybrid-50/50 are more active than what would be expected, they are still less active than HZSM-5.
The Hybrid-50/50 has a higher activity than PM-50/50, which could be related to the interaction between the nano-HZSM-5 crystals and the γ-Al2O3 support, which leads to a more efficient conversion of the alcohol. The high activity of Hybrid-50/50 and PM-50/50 compared to γ-Al2O3 is related to the presence of HZSM-5 nanocrystals. Another aspect that influences the catalyst activity is the reaction pathways that are followed. In the work of Gunst et al. it was shown that a stable dibutyl ether surface species can occupy over 95% of the available acid sites in HZSM-5 during the dehydration of n-butanol [31]. The high surface coverage by dibutyl ether decreases the activity as the alternative reaction pathways (e.g., direct dehydration from n-butanol to 1-butene) are suppressed. In our previous work, we showed that dibutyl ether is the main product over γ-Al2O3 below conversions of 0.6 mol mol −1 [38]. This can implicate that if the γ-Al2O3 significantly contributes during the dehydration reaction and mainly forms dibutyl ether, the activity of the nano-HZSM-5 is lowered because of the poisoning by dibutyl ether. To gain more insights into the behavior of Hybrid-50/50 and PM-50/50 and validate our assumption, a selectivity analysis is performed. ty, most likely related to their raction with the reactant [55]. be attributed to the high ratio ite time yield (STY, defined in related to this ratio for every catalyst is acting non-linear ong acid sites over medium , purple), Hybrid-50/50 (, anol = 29 kPa, total pressure y with increasing ratio of the alyst strongly deviating from eraction between γ-Al2O3 and nano-HZSM-5 crystals being eolitic or Brønsted acid sites, . found that if there are Lewis the dehydration of ethanol is than what would be expected, d be related to the interaction h leads to a more efficient 50/50 compared to γ-Al2O3 is nfluences the catalyst activity l. it was shown that a stable d sites in HZSM-5 during the ther decreases the activity as n-butanol to 1-butene) are e main product over γ-Al2O3 l2O3 significantly contributes ctivity of the nano-HZSM-5 is s into the behavior of Hybridis performed.
As can be seen in Figure 6, the activity generally increases linearly with increasing ratio of the densities of the strong acid sites over medium acid sites. The only catalyst strongly deviating from this trend is Hybrid-50/50. We expect that this is due to a synergistic interaction between γ-Al 2 O 3 and the nano-HZSM-5 crystals in the hybrid catalyst. Furthermore, the nano-HZSM-5 crystals being dispersed on the alumina surface increases the accessibility of the zeolitic or Brønsted acid sites, leading to a possibly more efficient active site. Furthermore, Wang et al. found that if there are Lewis acid sites introduced in the vicinity of Brønsted acid sites in HZSM-5, the dehydration of ethanol is facilitated [56]. Although the acid sites of Hybrid-50/50 are more active than what would be expected, they are still less active than HZSM-5.
The Hybrid-50/50 has a higher activity than PM-50/50, which could be related to the interaction between the nano-HZSM-5 crystals and the γ-Al 2 O 3 support, which leads to a more efficient conversion of the alcohol. The high activity of Hybrid-50/50 and PM-50/50 compared to γ-Al 2 O 3 is related to the presence of HZSM-5 nanocrystals. Another aspect that influences the catalyst activity is the reaction pathways that are followed. In the work of Gunst et al. it was shown that a stable dibutyl ether surface species can occupy over 95% of the available acid sites in HZSM-5 during the dehydration of n-butanol [31]. The high surface coverage by dibutyl ether decreases the activity as the alternative reaction pathways (e.g., direct dehydration from n-butanol to 1-butene) are suppressed. In our previous work, we showed that dibutyl ether is the main product over γ-Al 2 O 3 below conversions of 0.6 mol mol −1 [38]. This can implicate that if the γ-Al 2 O 3 significantly contributes during the dehydration reaction and mainly forms dibutyl ether, the activity of the nano-HZSM-5 is lowered because of the poisoning by dibutyl ether. To gain more insights into the behavior of Hybrid-50/50 and PM-50/50 and validate our assumption, a selectivity analysis is performed.

Selectivity Analysis
The selectivity toward the dehydration products of the catalysts is compared in Figure 7 (and Figure S6 with error bars). Of all studied catalysts, only γ-Al 2 O 3 does not actively catalyze the isomerization of the butenes. In earlier work, it is reported that both the butenes and dibutyl ether (DBE) are primary products over HZSM-5, whilst for γ-Al 2 O 3 , only 1-butene and dibutyl ether are directly formed [31,38].

Selectivity Analysis
The selectivity toward the dehydration products of the catalysts is compared in Figure 7 (and Figure S6 with error bars). Of all studied catalysts, only γ-Al2O3 does not actively catalyze the isomerization of the butenes. In earlier work, it is reported that both the butenes and dibutyl ether (DBE) are primary products over HZSM-5, whilst for γ-Al2O3, only 1-butene and dibutyl ether are directly formed [31,38]. For most catalysts, the selectivity toward 1-butene is relatively stable at approximately 0.2 mol mol −1 . Hybrid-50/50 starts at a slightly higher 1-butene selectivity, but with increasing conversion, it approaches a similar value as the other catalysts. The higher initial 1-butene selectivity is most likely related to (i) the presence of γ-Al2O3, which catalyzes the formation of 1-butene but does not catalyze the isomerization reactions of 1-butene, and (ii) because, at low conversion, the surface of the HZSM-5 crystals is saturated with adsorbed alcohol and ether [30], thereby suppressing butene isomerization reactions. At higher conversion, the partial pressure of n-butanol decreases, allowing isomerization of the butenes, leading to a lower 1-butene selectivity. For n-HZSM-5 and c-HZSM-5, the selectivity toward 1-butene is identical and corresponds well with the literature [30,31,38].
The selectivity toward cis-and trans-2-butene is very similar for all studied materials except for γ-Al2O3. Not only the trends in the selectivity profile but also the absolute values of the selectivity agree well for every material containing HZSM-5 crystals. The selectivity toward trans-2-butene is slightly higher for n-HZSM-5 than for c-HZSM-5 and, correspondingly, the selectivity toward dibutyl ether is slightly higher for c-HZSM-5. This is attributed to the lack of mesopores in n-HZSM-5, and the formation of the bulkier dibutyl ether preferentially occurs at the largest voids inside the zeolite. As c-HZSM-5 has some mesoporosity, it has larger voids where dibutyl ether is preferentially formed.  (   7 of 18 show the highest activity, most likely related to their hich increases the interaction with the reactant [55]. ared to n-HZSM-5 can be attributed to the high ratio activity, expressed as site time yield (STY, defined in ng acid sites) is linearly related to this ratio for every 6), showing that the catalyst is acting non-linear e ratio of the amount of strong acid sites over medium SM-5 (, red), PM-50/50 (, purple), Hybrid-50/50 (, 3 K, inlet pressure of n-butanol = 29 kPa, total pressure nerally increases linearly with increasing ratio of the acid sites. The only catalyst strongly deviating from s due to a synergistic interaction between γ-Al2O3 and lyst. Furthermore, the nano-HZSM-5 crystals being e accessibility of the zeolitic or Brønsted acid sites, urthermore, Wang et al. found that if there are Lewis acid sites in HZSM-5, the dehydration of ethanol is d-50/50 are more active than what would be expected, n PM-50/50, which could be related to the interaction γ-Al2O3 support, which leads to a more efficient Hybrid-50/50 and PM-50/50 compared to γ-Al2O3 is s. Another aspect that influences the catalyst activity the work of Gunst et al. it was shown that a stable 95% of the available acid sites in HZSM-5 during the e coverage by dibutyl ether decreases the activity as ect dehydration from n-butanol to 1-butene) are that dibutyl ether is the main product over γ-Al2O3 implicate that if the γ-Al2O3 significantly contributes rms dibutyl ether, the activity of the nano-HZSM-5 is er. To gain more insights into the behavior of Hybridon, a selectivity analysis is performed. For most catalysts, the selectivity toward 1-butene is relatively stable at approximately 0.2 mol mol −1 . Hybrid-50/50 starts at a slightly higher 1-butene selectivity, but with increasing conversion, it approaches a similar value as the other catalysts. The higher initial 1-butene selectivity is most likely related to (i) the presence of γ-Al 2 O 3 , which catalyzes the formation of 1-butene but does not catalyze the isomerization reactions of 1-butene, and (ii) because, at low conversion, the surface of the HZSM-5 crystals is saturated with adsorbed alcohol and ether [30], thereby suppressing butene isomerization reactions. At higher conversion, the partial pressure of n-butanol decreases, allowing isomerization of the butenes, leading to a lower 1-butene selectivity. For n-HZSM-5 and c-HZSM-5, the selectivity toward 1-butene is identical and corresponds well with the literature [30,31,38].
The selectivity toward cis-and trans-2-butene is very similar for all studied materials except for γ-Al 2 O 3 . Not only the trends in the selectivity profile but also the absolute values of the selectivity agree well for every material containing HZSM-5 crystals. The selectivity toward trans-2-butene is slightly higher for n-HZSM-5 than for c-HZSM-5 and, correspondingly, the selectivity toward dibutyl ether is slightly higher for c-HZSM-5. This is attributed to the lack of mesopores in n-HZSM-5, and the formation of the bulkier dibutyl ether preferentially occurs at the largest voids inside the zeolite. As c-HZSM-5 has some mesoporosity, it has larger voids where dibutyl ether is preferentially formed.
γ-Al 2 O 3 has the highest and PM-50/50 has the second highest selectivity toward dibutyl ether, which is most likely related to the high amount of γ-Al 2 O 3 present in PM-50/50. Therefore, both γ-Al 2 O 3 and HZSM-5 are important contributors to the activity and selectivity over PM-50/50. On the contrary, instead of having an intermediate selectivity toward dibutyl ether, Hybrid-50/50 has the lowest selectivity to it of all studied catalysts at low conversion. With increasing conversion, the selectivity toward DBE over Hybrid-50/50 approaches the values obtained over HZSM-5. The low selectivity toward dibutyl ether could explain why the activity of Hybrid-50/50 is higher than the activity of PM-50/50. We propose that the low selectivity of dibutyl ether over Hybrid-50/50 is related to the synthesis procedure, where the γ-Al 2 O 3 is intensely mixed in the zeolite precursor solution at high pH. As mentioned in Section 2.1, we assume that during the synthesis, the basic hydroxyl groups originating from TPAOH interact with the surface of γ-Al 2 O 3 and deactivate some of the strong Lewis acid sites. It is possible that the decrease in strong acid sites during synthesis is reflected in a lowered selectivity toward DBE and a shift in dominant reaction pathways to allow more butene formation, which consequently leads to a higher activity than expected.

Catalyst Stability
Next to activity and selectivity, stability is an important parameter to assess the catalyst performance. To this end, a deactivation study was performed, and the results are shown in Figure 8. The reaction was performed at 513 K and at site times where the conversion was similar for every catalyst.
γ-Al2O3 has the highest and PM-50/50 has the second highest selectivity toward dibutyl ether, which is most likely related to the high amount of γ-Al2O3 present in PM-50/50. Therefore, both γ-Al2O3 and HZSM-5 are important contributors to the activity and selectivity over PM-50/50. On the contrary, instead of having an intermediate selectivity toward dibutyl ether, Hybrid-50/50 has the lowest selectivity to it of all studied catalysts at low conversion. With increasing conversion, the selectivity toward DBE over Hybrid-50/50 approaches the values obtained over HZSM-5. The low selectivity toward dibutyl ether could explain why the activity of Hybrid-50/50 is higher than the activity of PM-50/50. We propose that the low selectivity of dibutyl ether over Hybrid-50/50 is related to the synthesis procedure, where the γ-Al2O3 is intensely mixed in the zeolite precursor solution at high pH. As mentioned in Section 2.1, we assume that during the synthesis, the basic hydroxyl groups originating from TPAOH interact with the surface of γ-Al2O3 and deactivate some of the strong Lewis acid sites. It is possible that the decrease in strong acid sites during synthesis is reflected in a lowered selectivity toward DBE and a shift in dominant reaction pathways to allow more butene formation, which consequently leads to a higher activity than expected.

Catalyst Stability
Next to activity and selectivity, stability is an important parameter to assess the catalyst performance. To this end, a deactivation study was performed, and the results are shown in Figure  8. The reaction was performed at 513 K and at site times where the conversion was similar for every catalyst. Although all catalysts are relatively stable, the strongest deactivation over time is noticeable for c-HZSM-5, for which the conversion decreases from 0.58 to 0.45 mol mol −1 in 48 h. The selfsynthesized n-HZSM-5 is more resistant to deactivation: The conversion decreases from 0.64 to 0.54 mol mol −1 in 48 h. Where the deactivation of c-HZSM-5 is the most severe in the first 15 h and then slows down, the n-HZSM-5 has an overall more gradual deactivation. The increased resistance to deactivation of n-HZSM-5 is most likely related to the, on average, smaller crystal size and the narrower crystal size distribution, resulting in a shorter microporous diffusion pathlength compared to c-HZSM-5. As the zeolite nanocrystals easily aggregate, there can be localized higher densities of strong acid sites prone to coke formation, and, therefore, the n-HZSM-5 can still deactivate significantly over time. On the other side of the spectrum, there is γ-Al2O3, for which no measurable deactivation is observed over 48 h. This is as expected as γ-Al2O3 is known for its high stability under alcohol dehydration conditions [36,57].
PM-50/50 also has a slightly stronger deactivation during the first hours, but then becomes more stable. The initial deactivation is attributed to the presence of nano-HZSM-5, but over time, the show the highest activity, most likely related to their hich increases the interaction with the reactant [55]. ared to n-HZSM-5 can be attributed to the high ratio activity, expressed as site time yield (STY, defined in ng acid sites) is linearly related to this ratio for every 6), showing that the catalyst is acting non-linear e ratio of the amount of strong acid sites over medium SM-5 (, red), PM-50/50 (, purple), Hybrid-50/50 (, 3 K, inlet pressure of n-butanol = 29 kPa, total pressure nerally increases linearly with increasing ratio of the acid sites. The only catalyst strongly deviating from s due to a synergistic interaction between γ-Al2O3 and lyst. Furthermore, the nano-HZSM-5 crystals being e accessibility of the zeolitic or Brønsted acid sites, urthermore, Wang et al. found that if there are Lewis acid sites in HZSM-5, the dehydration of ethanol is d-50/50 are more active than what would be expected, n PM-50/50, which could be related to the interaction γ-Al2O3 support, which leads to a more efficient Hybrid-50/50 and PM-50/50 compared to γ-Al2O3 is s. Another aspect that influences the catalyst activity the work of Gunst et al. it was shown that a stable 95% of the available acid sites in HZSM-5 during the e coverage by dibutyl ether decreases the activity as ect dehydration from n-butanol to 1-butene) are that dibutyl ether is the main product over γ-Al2O3 implicate that if the γ-Al2O3 significantly contributes rms dibutyl ether, the activity of the nano-HZSM-5 is er. To gain more insights into the behavior of Hybridon, a selectivity analysis is performed. Although all catalysts are relatively stable, the strongest deactivation over time is noticeable for c-HZSM-5, for which the conversion decreases from 0.58 to 0.45 mol mol −1 in 48 h. The self-synthesized n-HZSM-5 is more resistant to deactivation: The conversion decreases from 0.64 to 0.54 mol mol −1 in 48 h. Where the deactivation of c-HZSM-5 is the most severe in the first 15 h and then slows down, the n-HZSM-5 has an overall more gradual deactivation. The increased resistance to deactivation of n-HZSM-5 is most likely related to the, on average, smaller crystal size and the narrower crystal size distribution, resulting in a shorter microporous diffusion pathlength compared to c-HZSM-5. As the zeolite nanocrystals easily aggregate, there can be localized higher densities of strong acid sites prone to coke formation, and, therefore, the n-HZSM-5 can still deactivate significantly over time. On the other side of the spectrum, there is γ-Al 2 O 3 , for which no measurable deactivation is observed over 48 h. This is as expected as γ-Al 2 O 3 is known for its high stability under alcohol dehydration conditions [36,57].
PM-50/50 also has a slightly stronger deactivation during the first hours, but then becomes more stable. The initial deactivation is attributed to the presence of nano-HZSM-5, but over time, the catalyst keeps most of its activity due to the presence of γ-Al 2 O 3 . Hybrid-50/50 seems, similar to γ-Al 2 O 3 , to be very stable and shows no measurable deactivation during 48 h on stream. This is quite peculiar because, based on the relatively high activity and the observed selectivity profile, a deactivation behavior somewhere between γ-Al 2 O 3 and HZSM-5 would be expected. While the deactivation behavior is very similar to that of γ-Al 2 O 3 , the selectivity toward the products is very similar to that of HZSM-5 during the deactivation study (see Figure 9). This is quite remarkable as the selectivity versus time on the stream profile implies that the nano-HZSM-5 crystals of Hybrid-50/50 must be dominantly contributing during the complete run. However, there is no measurable deactivation, while there is definitely deactivation noticeable for the pure n-HZSM-5. We assume that this strong increase in resistance to deactivation for Hybrid-50/50 is related to the decrease in aggregation of the HZSM-5 nanocrystals due to the dispersion and interaction with the γ-Al 2 O 3 surface. A similar observation is reported by Castaño et al. where the coke formation during the methanol-to-hydrocarbon conversion was significantly reduced by decreasing the aggregation of HZSM-5 crystals inside catalyst extrudates consisting of HZSM-5, a pseudo-boehmite binder, and α-Al 2 O 3 filler [58]. Next to the decrease in crystal aggregation, the mesopores of γ-Al 2 O 3 in the vicinity of the nano-HZSM-5 crystals allow more efficient transport of the products (i.e., product removal) and could avoid coke formation.
The intermediate activity of Hybrid-50/50, combined with the high selectivity toward olefins and the high stability during the conversion of bio-butanol, indicates a beneficial effect that is possibly related to the close proximity of active sites of HZSM-5 and those of γ-Al 2 O 3 , because of the high dispersion of the HZSM-5 nanocrystals on the γ-Al 2 O 3 .
catalyst keeps most of its activity due to the presence of γ-Al2O3. Hybrid-50/50 seems, similar to γ-Al2O3, to be very stable and shows no measurable deactivation during 48 h on stream. This is quite peculiar because, based on the relatively high activity and the observed selectivity profile, a deactivation behavior somewhere between γ-Al2O3 and HZSM-5 would be expected. While the deactivation behavior is very similar to that of γ-Al2O3, the selectivity toward the products is very similar to that of HZSM-5 during the deactivation study (see Figure 9). This is quite remarkable as the selectivity versus time on the stream profile implies that the nano-HZSM-5 crystals of Hybrid-50/50 must be dominantly contributing during the complete run. However, there is no measurable deactivation, while there is definitely deactivation noticeable for the pure n-HZSM-5. We assume that this strong increase in resistance to deactivation for Hybrid-50/50 is related to the decrease in aggregation of the HZSM-5 nanocrystals due to the dispersion and interaction with the γ-Al2O3 surface. A similar observation is reported by Castaño et al. where the coke formation during the methanol-to-hydrocarbon conversion was significantly reduced by decreasing the aggregation of HZSM-5 crystals inside catalyst extrudates consisting of HZSM-5, a pseudo-boehmite binder, and α-Al2O3 filler [58]. Next to the decrease in crystal aggregation, the mesopores of γ-Al2O3 in the vicinity of the nano-HZSM-5 crystals allow more efficient transport of the products (i.e., product removal) and could avoid coke formation.
The intermediate activity of Hybrid-50/50, combined with the high selectivity toward olefins and the high stability during the conversion of bio-butanol, indicates a beneficial effect that is possibly related to the close proximity of active sites of HZSM-5 and those of γ-Al2O3, because of the high dispersion of the HZSM-5 nanocrystals on the γ-Al2O3.
The synthesis of the nano-HZSM-5 material was based upon the method reported by Song et al. [59], but was slightly adjusted. First, aluminum isopropoxide (≥98%, Merck) was dissolved in tetrapropylammonium hydroxide (TPAOH, 1 M in water, Merck) under continuous stirring to obtain a clear solution, and this was followed by addition of H 2 O, then NaOH (≥98%, Merck), and finally tetraethyl orthosilicate (TEOS, ≥99%, ChemLab) to obtain a synthesis solution, which was hydrolyzed at room temperature for 24 h. The obtained mixture was transferred into a Teflon-lined stainless-steel autoclave, where the solution underwent hydrothermal reaction at 165 • C for 96 h. The molar composition of the synthesis mixture was 1 Al/25 Si/9 TPAOH/495 H 2 O/0.16 NaOH/100 Ethanol. Following the hydrothermal reaction, the solid product was recovered, filtered, washed with deionized water until the pH of the washing water was neutral, dried at 120 • C overnight, and calcined in air at 823 K for 8 h to remove the TPAOH and structure directing agent, and obtain nano-ZSM-5.
The nano-ZSM-5 was then ion-exchanged in a 1 M solution of NH 4 NO 3 (Merck) dissolved in water. The ion exchange took place under continuous stirring at 323 K for 2 h and was repeated three times to obtain nano-NH 4 ZSM-5. The nano-NH 4 ZSM-5 was calcined at 823 K for 4 h, with a temperature ramp of 1 K min −1 , which resulted in nano-HZSM-5 (n-HZSM-5).
For the nano-HZSM-5/γ-Al 2 O 3 (Hybrid-50/50) catalyst, the synthesis route was identical to the synthesis of n-HZSM-5, but γ-Al 2 O 3 was added under continuous stirring 30 min prior to the hydrothermal reaction. The addition of γ-Al 2 O 3 prior to the crystallization of HZSM-5 has also been reported in the work of He et al. [51], but their nano-HZSM-5 synthesis differs from the method reported in this work. The addition of γ-Al 2 O 3 resulted in a synthesis mixture aimed at 50 mass% nano-HZSM-5 and 50 mass% γ-Al 2 O 3 . The introduction of γ-Al 2 O 3 prior to the crystallization allows the in-situ formation of nano-HZSM-5 crystals on the γ-Al 2 O 3 surface, see Figure S1 for a schematic representation of the synthesis route of Hybrid-50/50. The subsequent crystallization and ion-exchange procedures were identical as for the n-HZSM-5.

Catalyst Characterization
Catalyst particles within the 100-150 µm range were prepared by pressing the catalyst powders, followed by sieving. X-ray diffraction (XRD, Diffractometer Kristalloflex D5000 with Cu Kα radiation) was performed to identify the materials after synthesis. All powder diffraction patterns were collected for 2θ between 5 • and 70 • , with a step of 0.02 • and 30 s counting time for each angle. The mean crystal sizes were estimated by application of the Scherrer equation: with d the mean crystal size (nm), K a dimensionless shape factor with a value of 1 [59], λ the wavelength of the X-rays, β the full-width at half-maximum (corrected for the instrumental error), and θ Bragg's angle. The surface area and pore volume of the catalysts were determined by N 2 -sorption at 77 K (Micromeritics Tristar). Before the measurements, the samples were degassed at 573 K for 4 h under a continuous N 2 flow. The BET-method was used to obtain the surface area S BET , the t-plot method for micropore area and micropore volume (V micro ), and the single-point adsorption for the total pore volume (V t ). The external surface area of the self-synthesized catalysts was determined by performing the N 2 -sorption prior to calcining the materials. To increase the accuracy of the S BET of the materials containing micropores (HZSM-5, Hybrid-50/50, and the PM-50/50), Rouquerol's criteria were applied [60]. The pore size distribution (PSD) was determined by analyzing the adsorption branch of the isotherm, by applying the Barrett, Joyner, and Halenda (BJH) method for the mesoporous materials, and by a non-local density functional theory (NLDFT) method, assuming cylindrical pore geometries for ZSM-5 [39,61]. The measurements of the c-HZSM-5 and n-HZSM-5 were initiated at low relative pressure. NH 3 -TPD (Micromeritics Autochem) was used for the determination of the number of acid sites, assuming that each medium and strongly adsorbed NH 3 leads to one catalytically active acid site [62]. Then, 0.1 g catalyst was dried at 573 K under a continuous helium flow during 1 h to remove any surface-bound gases or moisture. Subsequently, the sample was saturated with NH 3 by flowing a 4 mol% NH 3 /He flow for 1 h over the catalyst, and any unbound NH 3 molecules were flushed from the sample by helium at a temperature of 423 K. Desorption of NH 3 was achieved by increasing the temperature to 873 K at a rate of 10 K min −1 and was monitored using a thermal conductivity detector (TCD) and a mass spectrometer (MS), which were calibrated prior to every run. The obtained spectra were deconvoluted into three peaks, attributed to: Weak, medium, and strong acid sites. Integration of the peaks associated with the medium and strongly bound ammonia was assumed to result in the number of catalytically active acid sites (C a ) for each catalyst. To assess the strength of the bonding between the catalyst surface and the NH 3 , the NH 3 -desorption energy was determined through measuring the NH 3 -TPD profiles at different heating rates and then solving the Kissinger equation [63].
where T M is the maximum desorption temperature of the highest temperature peak (typically around 673 K), β is the heating ramp (K min −1 ), E d the NH 3 -desorption energy (J mol −1 ), R the universal gas constant (J mol −1 K −1 ), and A d the pre-exponential factor for the desorption. The calculated E d does not take into account the amount of acid sites, but is solely dependent on the change in the peak position with the change in the heating ramp, similar to how activation energy is calculated. Scanning electron microscopy (SEM, FEI Quanta 200F) was used to study the shape and size of the catalysts. Where XRD gives an average crystal size, SEM can give insights into the range of the crystal sizes. By SEM, it is also possible to qualitatively study the dispersion of the ZSM-5 crystals on the γ-Al 2 O 3 surface for the Hybrid-50/50 catalyst and compare it to the PM-50/50.
Energy-dispersive X-ray spectroscopy (EDX) was used to determine the Si/Al ratio of the c-HZSM-5 and n-HZSM-5, where for c-HZSM-5, this was 28.1, close to the reported 25 by the manufacturer, and, for n-HZSM-5, the determined Si/Al ratio was 25.3, very close to the aimed ratio of 25. For Hybrid-50/50, we could not accurately determine the Si/Al ratio of the zeolite nanocrystals, due to the presence of γ -Al 2 O 3 near the crystals.

Catalytic Testing
The catalytic experiments were performed on an identical setup as previous work [30,31,38]: The reactions were performed in a set of tubular reactors with a length of 0.85 m and an inner diameter of 0.0022 m. The catalyst was loaded in the reactor and diluted with inert α-alumina until 10 mass% of catalyst to avoid temperature deviations across the catalyst bed. The catalyst weight range was limited to between 0.005 and 0.2 g. Liquid n-butanol (Merck, >99.5%) was fed through a Coriolis mass flow controller. The inlet partial pressure of n-butanol in the reactors was regulated to 29 kPa by adding N 2 (Air Liquide) as a carrier gas. Prior to reaction, the reactors and catalyst bed were heated to the reaction temperature under a N 2 flow for 1 h, to ensure the complete reactor tubes were at reaction temperature and remained stable. All downstream lining was heated to 443 K to avoid condensation. To achieve a working pressure of 5 bar, a back pressure regulator was used. Online analysis of the reactor effluent was performed with a GC-FID (with a 100 m PONA column), with CH 4 (Air Liquide) as an internal standard. The only detected species were n-butanol (the feed), dibutyl ether, butenes, and the internal standard. No other (oxygenated) hydrocarbons were detected; therefore, based on reaction stoichiometry, the amount of water being formed was calculated. The resulting material balance of the results shown in this work closed within 5%. To ensure intrinsic kinetics were measured, literature correlations were utilized [64]. The different acid site densities of each catalyst were compensated for by comparing the catalysts based on site time: with W (kg) the mass of the catalyst, C a (mol kg −1 ) the acid site density, and F 0 BuOH (mol s −1 ) the molar flow rate of n-butanol at the inlet.
The activity of the catalysts is based on the conversion of n-butanol (X BuOH ) and is defined as: where F BuOH (mol s −1 ) is the molar flow rate of n-butanol at the outlet of the reactor as determined by GC analysis. The activity can also be expressed as site time yield (STY, in s −1 ): The STY is similar to a turnover frequency but is averaged over the complete reactor. The carbon selectivity (S i ) toward the product is expressed as: with F i the outlet flow rate of product i, and c i the number of carbon atoms per molecule of product i.

Conclusions
HZSM-5 and γ-Al 2 O 3 are widely used materials in heterogeneous catalysis and they could be key materials to process bio-derived compounds. To attenuate their disadvantages and combine their advantages, a hybrid catalyst composed of nano-HZSM-5 crystals dispersed on a γ-Al 2 O 3 surface was in-situ hydrothermally synthesized. The hybrid catalyst's performance for the dehydration of n-butanol was compared with commercially available HZSM-5 and γ-Al 2 O 3 , and with synthesized nano-HZSM-5 and a physical mixture of nano-HZSM-5 and γ-Al 2 O 3 .
Through characterization of the materials, it was found that for the hybrid catalyst, the nanozeolite crystals were dispersed on the alumina surface, while in the physical mixture, there were domains of HZSM-5 and domains of γ-Al 2 O 3 . The synthesis procedure led to a different acid strength distribution compared to nano-HZSM-5 and pristine γ-Al 2 O 3 , where there was a shift toward more mildly acidic sites. The commercially available HZSM-5 had irregular, aggregated crystals with a size ranging from 80 to 600 nm, whilst the synthesized nano-HZSM-5 had a hexagonal shape, with a small crystal size distribution of 80−180 nm. The activity of the commercially available HZSM-5 was the highest, related to its high amount of strong acid sites. The activity of the hybrid catalyst was higher than that of the physical mixture, and both were intermediate between HZSM-5 and γ-Al 2 O 3 .
The selectivity toward dibutyl ether at low conversion was the lowest for the hybrid catalyst, which was related to the synthesis procedure and, consequently, altered the strongest acid sites on γ-Al 2 O 3 , leading to a decrease in ether formation. The selectivity was significantly different from the physical mixture, indicating an interaction between γ-Al 2 O 3 and the nano-HZSM-5 crystals. The lower selectivity to dibutyl ether over the hybrid catalyst could explain the higher activity compared to the physical mixture.
The order of catalytic stability under n-butanol dehydration conditions was: γ-Al 2 O 3 = the hybrid catalyst >> the physical mixture > nano-HZSM-5 > commercial HZSM-5. Whilst the hybrid catalyst was as stable as γ-Al 2 O 3 , it showed much higher selectivity toward butenes. In fact, the selectivity over the hybrid catalyst during the deactivation experiment was very similar to that of the nano-HZSM-5.
The combination of a relatively high activity, a high selectivity toward olefins, and very high stability indicate that the hybrid catalyst's performance benefits from the nano-HZSM-5 crystals being in the vicinity of γ-Al 2 O 3 and being well-dispersed on it. The hybrid catalyst outperformed every tested material when aiming for long-term olefin production under mild conditions. Future work could focus on the finetuning of the acid strength and degree of micropores/mesopores through changing the ratio of nano-HZSM-5/γ-Al 2 O 3 .
The selectivity toward dibutyl ether at low conversion was the lowest for the hybrid catalyst, which was related to the synthesis procedure and, consequently, altered the strongest acid sites on γ-Al2O3, leading to a decrease in ether formation. The selectivity was significantly different from the physical mixture, indicating an interaction between γ-Al2O3 and the nano-HZSM-5 crystals. The lower selectivity to dibutyl ether over the hybrid catalyst could explain the higher activity compared to the physical mixture.
The order of catalytic stability under n-butanol dehydration conditions was: γ-Al2O3 = the hybrid catalyst >> the physical mixture > nano-HZSM-5 > commercial HZSM-5. Whilst the hybrid catalyst was as stable as γ-Al2O3, it showed much higher selectivity toward butenes. In fact, the selectivity over the hybrid catalyst during the deactivation experiment was very similar to that of the nano-HZSM-5.
The combination of a relatively high activity, a high selectivity toward olefins, and very high stability indicate that the hybrid catalyst's performance benefits from the nano-HZSM-5 crystals being in the vicinity of γ-Al2O3 and being well-dispersed on it. The hybrid catalyst outperformed every tested material when aiming for long-term olefin production under mild conditions. Future work could focus on the finetuning of the acid strength and degree of micropores/mesopores through changing the ratio of nano-HZSM-5/γ-Al2O3.