Dehydrogenation of Propane to Propylene Using Promoter-Free Hierarchical Pt/Silicalite-1 Nanosheets

Abstract

Propane dehydrogenation (PDH) is the extensive pathway to produce propylene, which is as a very important chemical building block for the chemical industry. Various catalysts have been developed to increase the propylene yield over recent decades; however, an active site of monometallic Pt nanoparticles prevents them from achieving this, due to the interferences of side-reactions. In this context, we describe the use of promoter-free hierarchical Pt/silicalite-1 nanosheets in the PDH application. The Pt dispersion on weakly acidic supports can be improved due to an increase in the metal-support interaction of ultra-small metal nanoparticles and silanol defect sites of hierarchical structures. This behavior leads to highly selective propylene production, with more than 95% of propylene selectivity, due to the complete suppression of the side catalytic cracking. Moreover, the oligomerization as a side reaction is prevented in the presence of hierarchical structures due to the shortening of the diffusion path length.

1. Introduction

Propylene is one of the most important petrochemical building blocks for the production of a diverse range of products, from solvents to plastics and other valuable intermediates, for example, polypropylene, acrylonitrile, acrolein, and acrylic acid. In the present technology, the majority of propylene production is obtained as a byproduct from steam cracking and fluid catalytic cracking of hydrocarbons [1,2]. Because of an increasing demand of propylene and the depletion of petroleum resources, on-purposed propylene production technologies from alternative substrates, such as methanol-to-olefin (MTO), [3,4] metathesis of ethylene and butylene [5] and propane dehydrogenation (PDH) [6] have been intensively developed. Among these technologies, propane dehydrogenation is the most economical and shortest route for propylene production and is commercially available, developed by several licensing processes including the Catofin process from CB&I Lummus, reference [7] and the Oleflex process from Universal Oil Products (or Honeywell UOP) [8]. Because this reaction is a highly endothermic process and equilibrium-limited reaction, the relatively high temperature (above 550 °C) is usually required for promoting the catalytic activity. Thus, this behavior facilitates several side-reactions, in particular, thermal/catalytic cracking and coke formation, resulting in the low propylene selectivity and the rapid deactivation of catalysts [9]. It is, therefore, a great challenge to develop this process, especially in terms of increasing the efficiency of catalysts exhibiting high selectivity and stability in such severe operating conditions for the industrial production of propylene.

A Cr-based alumina supported catalyst is considered as one of the most important catalysts for this reaction due to their promising catalytic activity [6,8] however, the Cr is rather hazardous. An alternative way is to replace by the Pt-based catalysts [10,11,12] and other non-noble metals, including Cu, Fe, Zn and VO_x [13,14,15,16,17,18]. Although the Pt-based catalysts exhibit an excellent activity, they still also suffer from disadvantages, such as a short catalyst lifetime due to the coke formation and the sintering of Pt clusters [19].

Several strategies have been applied to improve the catalytic performance of Pt-based catalysts, including the addition of a second metal (e.g., Sn, Zn, Cu, Ce) as promoters [17,20,21,22,23], the development of the synthesis method to obtain small Pt nanoclusters such as the atomic layer deposition (ALD) technique [24] and the development of metal-support interactions to change the electron density of Pt, which eventually leads to reducing the propylene adsorption, and then suppressing the further catalytic cracking [22,25,26,27,28].

A zeolite is one of the most interesting materials to be widely used as a support owing to its outstanding properties, such as a high surface area and a high thermal/hydrothermal stability [29,30]. However, it often suffers from the metal aggregation on outer surfaces after the thermal treatment, resulting in a very low catalytic selectivity to propylene [31,32]. To prevent the metal aggregation, the introduction of a mesoporous system into the microporous zeolite network can greatly improve the dispersion of metal species due to a large increasing number of defect sites [33]. With the benefits of mesoporous or hierarchical structures [34,35], an improved catalytic efficiency has been observed in some reactions, in particular the processes involving bulky molecules, such as hydroisomerization, aromatization and hydrodesulfurization, as compared to a conventional zeolite [36,37,38,39,40].

Indeed, the acidity of supports also affects the propylene selectivity because of the interferences of the catalytic cracking and alkene oligomerization. Therefore, an alternative way is to apply the concept of using the neutralized surface acidity with basic molecules [41] or the siliceous zeolite with a weak acid. However, the lack of alumina in the zeolite framework results in the weak metal-support interaction, eventually leading to the aggregation of metal species [42,43]. Thus, the development of alternative PDH catalysts, which exhibit the interplay between minimizing the acidity and increasing the surface interaction of support and metal species, is still a great challenge.

In the present study, we report the example of a highly selective propane dehydrogenation over monometallic nanoparticles supported on weakly acidic siliceous MFI-type zeolite having the hierarchical nanosheet structure. This catalyst has been chosen since it not only benefits to the shape-selectivity of MFI structure that is preferable to propylene, but also suppresses the side catalytic cracking on acidic surfaces and prevents the aggregation of metal species owing to the presence of silanol defect sites, resulting in increasing the metal-support interaction. Compared to the conventional Pt supported zeolite, another side reaction such as an oligomerization is also suppressed in the presence of hierarchical structures due to shorten diffusion path lengths. In addition, its catalytic performance will be discussed in terms of the roles of metal-support interaction and acidity. Finally, the degree of propylene selectivity can be easily adjusted by changing the loading content of metal and the weight hourly space velocity (WHSV).

2. Results and Discussion

2.1. Structures and Properties of Catalysts

XRD patterns of Pt supported conventional and hierarchical siliceous zeolites represented the MFI structure including that the impregnation of Pt did not destroy the structural property of zeolites. In addition, there were no detectable PtO₂ characteristics in XRD patterns of all supports indicating the well dispersion of Pt or the relatively low amount of Pt loading (Figure 1A). SEM images of hierarchical siliceous MFI or silicalite-1 nanosheets, conventional silicalite-1 and γ-alumina exhibited uniform particles with a narrow size distribution (see Figure 1B,D and Figure S1) of 250.0 ± 20.0 nm, 2.5 ± 0.3 µm, and 25.0 ± 5.0 nm, respectively. The diameters of Pt particles supported on silicalite-1 nanosheets, conventional silicalite-1 and γ-alumina were approximately 2.0 ± 0.5, 20.0 ± 10.0 and 0.6 ± 0.2 nm, respectively (Figure S2) as shown in TEM images (Figure 1E–G and Figure S2A). The Pt content was confirmed after a successful impregnation of Pt on the supports by using XRF analysis as shown in

Table 1. Textural properties of Pt supported catalysts, including hierarchical silicalite-1.

Sample	SBET a	Smicro b	Sext c	Vtotal d	Vmicro e	Vmeso f	%Ptg
Si-MFI-NS	447	292	154	0.85	0.12	0.74	n.d.
1% Pt-Si-MFI-NS	399	296	104	0.56	0.12	0.45	1.07
Si-MFI-CON	427	334	93	0.25	0.13	0.12	n.d.
1% Pt-Si-MFI-CON	374	322	52	0.22	0.13	0.09	0.90
1% Pt-Al2O3	161	11	150	0.96	0.005	0.95	1.19

^a S_BET: BET specific surface area; ^b S_micro: micropore surface area; ^c S_ext: external surface area; ^d V_total: total pore volume; ^e V_micro: micropore volume; ^f V_micro = V_total − V_ext; All surface areas and pore volumes are in the units of m².g⁻¹ and cm³.g⁻¹, respectively: ^g measured by XRF.

. The Pt loading content on various supports was in the same range of 1 wt.%.

The textural properties of all catalysts were measured using N₂ sorption measurements as presented in Figure 1H and Table 1. The isotherm of γ-alumina exhibited an apparent hysteresis loop at P/P₀ range of 0.6 to 0.9 corresponding to the type-IV isotherm, suggesting the presence of mesoporous structures [44]. In contrast, the isotherm of the conventional silicalite-1 exhibited only the adsorption at the relatively low pressure (P/P₀ < 0.2) due to a microporous structure, corresponding to the type-I isotherm [44]. On the other hand, the silicalite-1 nanosheet exhibited the adsorption at low pressure together with the hysteresis loop at P/P₀ in the range of 0.5 to 0.9, indicating the combination of type-I and IV isotherms [44]. The mesopore sizes of γ-alumina and silicalite-1 nanosheets obtained by the BJH model were approximately 10 and 40 nm, respectively, and no mesoporous structure was observed in the case of the conventional silicalite-1. The BET surface area of γ-alumina was relatively low (161 m²·g⁻¹) compared to silicalite-1 (~400 m²·g⁻¹), while the total pore volume was relatively high (up to 0.96 cm³·g⁻¹) due to a high fraction of mesopores (S_ext 150 m²·g⁻¹). Considering the siliceous zeolite, the silicalite-1 nanosheets showed a relatively higher total pore volume (0.56 cm³·g⁻¹) than that of the conventional one (0.25 cm³·g⁻¹), while the micropore volume was in the same range (~0.12–0.13 cm³·g⁻¹), indicating the presence of a large amount of meso- and macroporosity. In addition, the silicalite-1 with Pt loading slightly changed the textural properties of support. The surface areas (S_BET from 447 to 399 m²·g⁻¹) of Pt supported on the silicalite-1 nanosheets (1% Pt-Si-MFI-NS) and Pt supported on the conventional silicalite-1 (1% Pt-Si-MFI-CON) were decreased by a factor of 11% and 12%, respectively (Table 1).

The alumina support is known to play an important role as metal supported catalysts and provides a significant Pt-support interaction, as compared to a conventional silica support [45]. To verify the reduction behavior of metal-supported catalysts, the influences of different supports on reduced Pt species were characterized by H₂-TPR. The Pt supported on γ-alumina exhibited two reduction peaks at a relatively higher temperature (at 385 °C and 460 °C) compared to siliceous zeolites (Figure 2A) and similar to what has been described in previous reports [46,47], indicating the strong interaction between Pt nanoparticles and γ-alumina support. The Pt supported on conventional silicalite-1 exhibited only one peak at 320 °C, suggesting that Pt species on this support were easily reduced to metallic Pt compared γ-alumina support, confirming a weak interaction between a Pt metal and a silica support. Interestingly, in the case of Pt supported on silicalite-1 nanosheets, another reduction peak at higher temperature of 500 °C was observed. It is therefore reasonable to imply that there are some Pt species having a strong interaction with support due to the hierarchical structure, resulting in the higher resistance towards the reduction of Pt species. All H₂-TPR results confirmed that the different nature of supports can alter the behaviors of reduced Pt species.

One reason for the strong interaction between Pt and silicalite-1 nanosheets is that the hierarchical siliceous zeolite can provide a large number of defect sites, and therefore improve the interaction with metals. To confirm this hypothesis, the NH₃-TPD profiles and ²⁹Si MAS NMR spectra of silicalite-1 nanosheets were also observed, as shown in Figure S3 and Figure 2B. Compared to the conventional silicalite-1, the large number of weak acid sites of Pt supported silicalite-1 nanosheets remarkably increased (1.17 and 0.03 mmol·g⁻¹, for silicalite-1 nanosheets and the conventional silicalite-1, respectively), implying an increase in the surface silanol density due to the hierarchical structures. To further verify the presence of surface silanols of MFI zeolite, the ²⁹Si MAS NMR revealed the environment of Si framework representing as Qⁿ-groups in which n is 1 to 4 according to the number of tetrahedral Si unit as the next-nearest neighbors [48]. Interestingly, the hierarchical silicalite-1 exhibited the higher ratio of Q³ to Q⁴ (2.90) and the presence of Q² species, which was the characteristic of silanol surfaces due to a Si atom linked with two O–H groups. In strong contrast to this, the conventional silicalite-1 showed the significantly lower ratio of Q³ to Q⁴ (1.14) in the absence of Q² characteristic (see Figure 2B). This relates to the fact that the hierarchical nanosheets can enhance the surface silanol density located on external surfaces.

The electronic properties of Pt species were characterized by XPS analysis. The Pt4f XPS spectra of all catalysts exhibited doublet peaks corresponding to a spin-orbit splitting of Pt4f_7/2 and Pt4f_5/2 states (Figure S4) [49]. Unfortunately, the binding energy (B.E.) of Pt 4f_5/2 was very close to that of Al 2p in case of γ-alumina support [50]. The decomposition of such Pt 4f_7/2 and 4f_5/2 XPS peaks were used to analyze the predominant species in catalysts (Figure S4 and Table S1). Accordingly, the B.E.s at 71.1–74.4, 72.8–76.1 and 74.3–77.6 eV were assigned to metallic Pt (Pt⁰), Pt^IIO and Pt^IVO₂, respectively. The Pt 4f_7/2 B.E.s of the Pt supported on hierarchical silicalite-1 nanosheets obviously shifted to a lower binding energy compared with the conventional silicalite-1, indicating that defect sites in hierarchical structures can change the electron density of Pt to be an electron richer due to the strong interaction of metal and support, and thus, this material exhibited the improved metal-support interaction.

To clarify these behaviors, the theoretical study by a computational chemistry approach was used to investigate the electronic properties of catalysts. The interaction of Pt cluster supported on different silicon species (Q⁴, Q³, and Q²) was investigated as shown in Figure S5 and Table S3. The binding energy of Pt₄ cluster supported on the Q³ species (Pt₄/Q³-MFI) was approximately −7.0 kcal·mol⁻¹, whereas the binding energy of Pt₄ cluster supported on the Q⁴ species (Pt₄/Q⁴-MFI) was significantly lower (approximately −4.0 kcal·mol⁻¹). Moreover, the Natural Population Analysis (NPA) charge (Table S4) revealed that the electron transfer from a support to Pt₄ cluster in the Pt₄/Q³-MFI model was more predominant compared to the Pt₄/Q⁴-MFI model. The partial atomic charges of Pt₄ cluster were reduced from 0.000 e (bare Pt₄ cluster) to −0.290 e and −0.252 e for Pt₄/Q³-MFI and Pt₄/Q⁴-MFI models, respectively. The reason of this higher negatively charge over the Pt₄/Q³-MFI compared to the Pt₄/Q⁴-MFI related to the improved charge-transfer from defects or silanol surfaces of zeolites to the Pt₄ cluster. These observations confirmed the improved interaction of metal-support in the presence of defects or silanol surfaces of the Pt supported on silicalite-1 nanosheet. In addition, the lower adsorption energy of propylene on the Pt₄/Q³-MFI model (−29.5 kcal·mol⁻¹) compared to that of the Pt₄/Q⁴-MFI model (−36.1 kcal·mol⁻¹) indicated to a weaken adsorption strength of propylene from active site. It is therefore reasonable to assume that the propylene product is easily desorbed over defects or silanol surfaces, resulting in the suppression of further side reactions. This eventually leads to an increase in overall propylene selectivity over Pt supported silicalite-1 nanosheets.

2.2. Dehydrogenation of Propane over Pt Nanoparticles Supported on Different Supports

The above-mentioned acidity and metal-support interaction makes it clear that the hierarchical silicalite-1 not only improves the metal-support interaction compared to the conventional silicalite-1, but also reduces the acidity of supports with respect to γ-alumina. Therefore, this material is expected to benefit to the catalytic activity and selectivity of propane dehydrogenation.

To illustrate the benefits of Pt supported hierarchical silicalite-1 nanosheets (Pt/MFI-NS) on the propane dehydrogenation (PDH), it clearly demonstrates that the modified catalyst exhibits a relatively high activity with the Turnover Frequency (or TOF) value of 1112.2 h⁻¹ and an improved catalytic stability even after time on stream (TOS) of 12 h (Figure 3A). Interestingly, the high selectivity of propylene above 95% was observed, indicating that the side reactions were diminished on this catalyst. In strong contrast to this, the conventional silicalite-1 (Pt/Si-MFI-con) showed a much lower activity of the propane conversion with the TOF of 157.6 h⁻¹ with the propylene selectivity of 50%. Although the high activity (TOF = 278.4 h⁻¹) was observed over the fresh alumina-based catalyst (Pt/Al₂O₃), the deactivation rate was also very fast with the 20-fold increase in the deactivation rate constant compared to those of over the Pt supported hierarchical silicalite-1 nanosheets (Pt/Si-MFI-NS) (Figure S6). As expected, the selectivity of propylene also decreased from 80 to 55% after 12 h of TOS.

The distribution of products after 10 h of TOS is shown in Figure 3B. Unfortunately, the Pt/Al₂O₃ noticeably exhibited a high selectivity to cracked products (methane and ethane) (~40%). The relatively high selectivity of cracked products was also observed over the conventional zeolite (Pt/Si-MFI-CON) (~50%). These results indicated that the nature of supports plays an important role in the selective production of light cracked hydrocarbons. The reason for this increase in cracked product selectivity relates to the fact that the Al₂O₃ support composes of the high amount of acid sites as shown in NH₃-TPD results (see Figure S3), resulting in the promotion of cracking reaction. Compared to the hierarchical silicalite-1 nanosheets, the cracked product selectivity over the conventional silicalite-1 was also remarkably increased. This is most likely due to the fact that the formation of lighter hydrocarbons via C–C bond cleavage is inevitably excluded using Pt-based catalysts [49]. It is well known that the side reactions, such as catalytic cracking of hydrocarbons over Pt-based catalysts are structure-sensitive reactions and they can be accelerated over large Pt clusters [10]. This behavior eventually leads to the formation of a light C–C bond cracking over the conventional silicalite-1, containing large Pt clusters.

2.3. Effects of WHSV and Pt Loading on the Propylene Selectivity

The degree of propylene selectivity can be simply tuned by changing the loading content of metal and the weight hourly space velocity (WHSV). Figure 4A,B demonstrates the conversion, TOF and selectivity of propylene over the hierarchical silicalite-1 nanosheets obtained at different TOS and WHSV. The conversion of propane was in the range of 30–35% and it agreed well with the theoretical equilibrium value at 550 °C is ~37%, [8,15] (Figure 4A). Obviously, at the low WHSV, the selectivity of propylene was decreased to 89% and the selectivity of cracked products was rather high (~11%, Figure 4B). Such an observation suggested that at low WHSV, the side catalytic cracking was preferentially occurred due to a long time of reaction mixtures being in contact with the catalyst. In contrast, it was also found that the highly selective production of propylene (>95%) was observed when the WHSV was greater than 2.8 h⁻¹, indicating that the short contact time suppressed side reactions. In addition, the TOF increased as a function of WHSV; however, at too high WHSV the TOF becomes smaller (Figure 4B).

The conversion of propane and propylene selectivity as a function of Pt loading over the Pt supported hierarchical silicalite-1 nanosheets at WHSV of 2.8 h⁻¹ is presented in Figure 4C. As expected, the silicalite-1 nanosheets without Pt loading, there was no activity and the selectivity of propylene was in a very low level of 45%, demonstrating that the support itself could not serve as the dehydrogenated active site. It was also found that a small amount of methane and ethane derived from catalytic cracking was observed. In strong contrast to this, the Pt supported catalyst exhibited the selectivity of propylene about 95.2%, although the low Pt loading of 0.5 wt.% was used. The different loading of Pt showed a significant influence on the TOF of propane conversion. When the loading of Pt was relatively low (0.5 wt.%), the low TOF (734.3 h⁻¹) of propane conversion was observed. With a further increase in Pt loading to 1 wt.%, the TOF increased and reached the highest value of 1112.2 h⁻¹, while the high propylene selectivity was still observed (95%). At the relatively high Pt loading level of 1.5 wt.%, the TOF becomes smaller (741.2 h⁻¹). This may be explained by the fact that the excessive loading of Pt leads to the aggregation of Pt particles, and consequently the lower catalytic activity. Such an observation suggested that the Pt supported on silicalite-1 nanosheets containing 1 wt.% of Pt exhibited the superior performance in terms of propane activity and propylene selectivity.

In summary, the obtained results indicate that the superior catalytic performance of Pt supported on hierarchical silicalite-1 nanosheets for propane dehydrogenation (PDH) can be achieved to selectively convert propane to propylene without the catalyst deactivation even after the time on stream (TOS) of 12 h. TEM, H₂-TPR, ²⁹Si NMR, XPS measurements and DFT calculations confirmed the presence of small active Pt particles supported on silanol or defect surfaces of the hierarchical silicalite-1 nanosheet catalyst. This relates to the fact that the silanol surfaces or defect sites of hierarchical zeolite nanosheets can be greatly enhanced compared to the conventional silicalite-1, eventually resulting in enhancing the interaction of metal and support, thus preventing the aggregation of Pt clusters (Scheme 1).

To obtain the high yield of propylene, the effects of metal loading content and weight hourly space velocity (WHSV) were systematically studied to fine-tune the optimal experimental conditions. It was found that the silicalite-1 nanosheets containing 1 wt.% of Pt with the WHSV of 5.64 h⁻¹ exhibited the highest performance with the TOF of 1394.2 h⁻¹ and propylene selectivity of 96%. In strong contrast to this, the Pt supported conventional silicalite-1 and commercial Pt/Al₂O₃ catalysts exhibited a very low activity, propylene selectivity and short catalyst lifetime. In particular, they preferentially facilitate the side-reactions, such as catalytic cracking due to the presence of large Pt clusters and high acidity of supports for the Pt supported conventional silicalite-1 and commercial Pt/Al₂O₃, respectively.

Scheme 1 illustrates the proposed concept of the improved metal-support interaction in Pt supported on hierarchical silicalite-1 to promote the selective propane dehydrogenation to propylene. This is consistent with the view that the modified hierarchical structure enhances the PDH performances by changing the structure of Pt. The high dispersion degree of small Pt particles is responsible for the enhancement of the propane dehydrogenation performances. The smallest size of Pt particles can be obtained in the case of hierarchical silicalite-1 nanosheets due to the high external surface area together with the increasing of silanol or defect surfaces which can enhance the interaction with Pt [51]. In addition, the Pt metal nanoparticles supported on zeolite nanosheets have remained the well-dispersion with the similar size of 2.1 ± 0.7 nm. To confirm this hypothesis, the computational approach was performed to understand the surface interaction and electron transfer in different silanol network systems. It was found that silanol surfaces can greatly enhance the binding energy with Pt clusters, implying that Pt is strongly bound to the hierarchical nanosheet surfaces compared to the conventional one. Therefore, it is reasonable to suppose that the improved interaction of Pt clusters and the silanol groups eventually leads to preventing the aggregation of Pt over hierarchical silicalite-1 nanosheets. In addition, the electronic properties of Pt over different supports were also changed. The negative charge of Pt over the Pt₄/Q³-MFI model is relatively larger than that of the Pt₄/Q⁴-MFI model, suggesting that the silanol groups can promote the electron transfer from the supports to Pt cluster. This is consistent with the view that the Q³ species are predominant in the hierarchical silicalite-1 nanosheets, then facilitating the strong interaction between metal and support. The change of electronic properties of Pt on different supports also affects the desorption of propylene as a product. As expected, the silanol surfaces can directly weaken the adsorption of propylene on Pt sites. This leads to an increase in propylene selectivity due to the suppression of further side-reaction. In addition, the hierarchical structure provided a good benefit for promoting the propylene production, resulting in the suppression of the further oligomerization as a side reaction due to the shortening of the diffusion path length of catalysts. These observations may have the benefit of an increase in metal dispersion and it is complementary to the development of the well-dispersed metallic nanoparticles on solid supports in PDH reaction [52].

3. Materials and Methods

3.1. Synthesis of Silicalite-1 Zeolites

Hierarchical silicalite-1 nanosheets (Si-MFI-NS) and conventional silicalite-1 (Si-MFI-CON) were synthesized by using tetra-(n-butyl)phosphonium hydroxide (TBPOH) and tetrapropylammonium hydroxide (TPAOH) as structure directing agents (SDAs), respectively. The hierarchical silicalite-1 nanosheets were crystallized from a gel precursor in the absence of aluminum source by following the modified literature procedures [53]. The molar composition of the precursor was 60SiO₂ 18TBPOH: 0.75NaOH: 240EtOH: 600H₂O. The resultant gel was stirred at room temperature for 12 h and then heated at 130 °C for 2 days. The conventional silicalite-1 was synthesized from a gel precursor with the composition of 10SiO₂: 1TPAOH: 1.03NaOH: 240EtOH: 400H₂O and crystallized at 180 °C for 3 days. The synthesized products were collected and washed with distilled water until the pH of filtrate less than 9. Finally, the products were calcined in air at 550 °C for 8 h to remove the SDAs.

3.2. Preparation of Pt Supported Silicalite-1 Zeolites

The Pt was loaded on silicalite-1 zeolites by a wet impregnation method. Typically, a 1.2 g of a calcined zeolite catalyst was added to a 10 mL of an aqueous solution of tetraammineplatinum nitrate (Pt (NH₃)₄(NO₃)₂) with various Pt loading contents (0.5, 1 and 1.5 wt.%). The slurry was stirred overnight at room temperature. Afterward, the water was carefully removed by the freeze-drying method.

3.3. Characterizations

X-ray diffraction (XRD) patterns of all catalysts were recorded on Bruker (Massachusetts, United States) D8 ADVANCE instrument with CuKα radiation (40 kV, 40 mA). Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images were obtained from JEOL-JSM-7600F (at 2 kV) (Tokyo, Japan) and JEOL-JEM-2100 (at 200 kV) (Tokyo, Japan) microscopes, respectively. The average particle size of Pt and zeolite was estimated from at least 200 particles. N₂ sorption techniques were used to examine the textural properties of catalysts. The measurements were carried out on Belmax (MicrotracBEL Corp., Tokyo, Japan) apparatus, and all the prepared samples were degassed at 300 °C for 24 h before the measurement. The specific surface area (S_BET) was determined by using the BET theory. The micropore volume (V_micro), the micropore area (S_micro) and the external surface area (S_ext) were calculated from the t-plot method. The total pore volume (V_total) was estimated at P/P₀ = 0.98. The composition and Pt loading content of prepared catalysts were measured by X-ray fluorescence (XRF) spectrometry performed on a Bruker (Massachusetts, United States) AXS S4 Pioneer instrument. The temperature-programmed reduction of H₂ (H₂-TPR) and H₂ pulse chemisorption were performed on the Belcat II (Tokyo, Japan) chemisorption analyzer equipped with thermal conductivity detectors (TCD). In H₂-TPR experiments, the sample was pretreated in Ar (50 mL·min⁻¹) at 300 °C for 1 h and then the temperature was ramped from 50 °C to 700 °C at 5 °C·min⁻¹ under the flow of 2% H₂/Ar (50 mL·min⁻¹). For the H₂ pulse chemisorption, all samples were heated to 300 °C under the Ar flow (50 mL·min⁻¹). Subsequently, the consecutive pretreatment by O₂ and H₂ was introduced. Finally, samples were cooled down to 50 °C and the H₂ pulse measurement was operated under the flow of 2% (v/v) H₂ in Ar with the flow rate of 50 mL·min⁻¹. XPS spectra were obtained using JEOL JPS-9010 (Tokyo, Japan). Non-monochromatic Mg Ka X-rays (1486.6 eV) were used as primary excitation. The C 1s was selected as the reference, which value was accepted equal to 284.7 eV.

3.4. Catalytic Testing

The dehydrogenation of propane to propylene was carried out in a continuous-flow fixed-bed reactor. Prior to the reaction testing, the similar particles size of a catalyst was controlled in the range of 425–850 µm. A 0.2 g of catalyst was placed into the reactor and activated at 550 °C under the flow of H₂ (2.4 mL·min⁻¹) for 2 h. Then, the propane in N₂ with the molar volume ratio of 1:1 was fed with various flow rates to study the effect of weight hourly space velocity (WHSV). The WHSV was varied from 1.4 to 11.3 h⁻¹ by changing the flow rate of the reactant mixture at constant reactor volume. The turnover frequency (TOF) was calculated according to the following Equation:

$$TOF(h^{-1})=\frac{moleofconvertedreactantA}{(moleoftotalmetal)x(D_m)x(unittime)}$$

where the number of active sites was calculated based on the total number of active metal sites derived from the H₂ chemisorption measurement [54]. D_m is the dispersivity of metal. All reactions were carried out at 550 °C, and atmospheric pressure. The analysis of the reaction mixtures was performed by an online gas chromatograph (GC) (Agilent 7820A) equipped with a flame ionization detector (FID) and a GS-GASPRO capillary column (60 m × 0.32 mm) at an interval time of 1 h. To validate the performance of catalysts in the kinetic regime, the experimental criteria due to heat and mass transfer limitation are also discussed as shown in the Supplementary Materials.

4. Conclusions

In conclusion, the results demonstrate the first example, describing the improved metal-support interaction of hierarchical nanosheet structures. A promoter-free Pt on hierarchical silicalite-1 nanosheets exhibited the highly selective production of propylene from propane dehydrogenation above 95% at the propane conversion level of 30%. The observations from TEM, H₂-TPR, ²⁹Si NMR, XPS measurements and DFT calculations confirmed that the hierarchical silicalite-1 nanosheet catalyst promoting the presence of small active Pt particles supported on silanol or defect surfaces of the hierarchical silicalite-1 nanosheet catalyst. Therefore, a novel design of supported catalysts with hierarchical structures can promote active species for the highly selective propane dehydrogenation, which has a great potential to be utilized in the important petrochemical catalytic reactions.