Chemical Equilibrium Controlled Etching of MFI-Type Zeolite and Its

Article pubs.acs.org/cm

Chemical Equilibrium Controlled Etching of MFI-Type Zeolite and Its Influence on Zeolite Structure, Acidity, and Catalytic Activity Z. Qin, L. Lakiss, J.-P. Gilson, K. Thomas, J.-M. Goupil, C. Fernandez, and V. Valtchev* Laboratoire Catalyse & Spectrochimie, ENSICAEN - Université de Caen − CNRS, 6 boulevard du Maréchal Juin, 14050 Caen, France S Supporting Information *

ABSTRACT: Chemical etching with fluoride ions is a new approach for secondary porosity engineering of aluminosilicate zeolite frameworks. We show that diluted HF solutions extract preferentially aluminum from zeolite frameworks. The Brønsted acidity of ZSM-5 treated in such a way decreases, while its structure is unaffected after an HF treatment. With higher HF concentrations, the number of undissociated HF molecules and the concentration of HF2− ions, extracting indiscriminately Al and Si, increase. The addition of NH4F shifts the chemical equilibria to produce more HF2−, avoiding the use of highly concentrated HF solutions; it also suppresses HF dissociation. The etching selectivity of such solutions is concentration-independent and extracts indiscriminately both framework Si and Al. Zeolite dissolution in NH4F-HF solutions starts preferentially at small intergrowth domains and goes deeply in the crystals without a substantial increase of the external surface area. Macropores are produced without altering zeolite acidity. Hierarchical materials obtained by these two approaches are characterized extensively by complementary methods and the catalytic impact illustrated in the m-xylene conversion. KEYWORDS: hierarchical zeolite, fluoride etching, xylene isomerization

1. INTRODUCTION The introduction of secondary porosity consisting of larger pores, i.e., mesopores and/or macropores, to overcome the inherent diffusion limitations of microporous zeolite-type materials has been extensively studied. While steaming and acid leaching tend to remove framework aluminum (dealumination),1 alkaline leaching removes silicon relatively selectively (desilication).2 The later approach, although more complex than a simple desilication,3 attracted considerable attention during the past decade, due to its broad applicability to a variety of high silica zeolites.4,5 However, both dealumination and desilication, as the name indicates, are cationselective approaches. Both of them have limitations as they, at the very least, change the Si/Al ratio of the initial zeolites. While dealumination applies to parent samples of low Si/Al ratio (2.5−5), desilication is more applicable to zeolites with Si/Al ratios from 30 to 50.6 Besides, both approaches change the composition and distribution of silicon and aluminum in the zeolite and thus its acidic properties.3 Therefore, the development of a general approach to design hierarchical zeolites retaining their initial chemical composition is highly desirable. Until now such materials could be prepared with sacrificial porogens, eliminated after the synthesis by high temperature combustion.7,8 Although efficient, this approach could hardly be applied on a large scale. Hence, the development of a chemical process producing hierarchical zeolites without substantial changes in their chemical composition is required. A possible solution is to use a chemical neither Si nor Al selective, thus etching the zeolite without altering its framework composition. © 2013 American Chemical Society

It is well-known that HF reacts with both silicon and aluminum. Recently, we used HF-etching to create very narrow distributions of macropores in ZSM-5 crystals.9 In this case the secondary pore formation was predetermined by high energy ionic bombardment, creating amorphous cylindrical zones crossing the entire zeolite crystal. The kinetics of etching of the amorphous zones is substantially faster than the crystalline ones. Thus a hierarchical material with a composition close to its parent zeolite was obtained. Another previous study reports that diluted HF acid preferentially extracts aluminum and changes substantially the zeolite framework composition.10 Aqueous or nonaqueous HF solutions were also used to treat zeolite crystals prior to structure-related studies.11−15 However, the use of acidic fluoride solutions to create secondary porosity in zeolite crystals has not yet been reported. According to Ghosh and Kydd, the selectivity of aluminum etching decreases in concentrated HF solutions.16 They reported that a treatment with dilute HF solutions (≤0.25 mol L−1) raised the Si/Al ratio of mordenite without affecting its crystallinity. However, the use of a more concentrated HF solution (1.5 mol L−1) did not change dramatically the Si/Al. A closer look at the chemistry of fluoride solutions also showed that there are opportunities to shift equilibrium to equal extraction of zeolite framework cations.17,18 The present study will explore these opportunities further and develop a method Received: March 4, 2013 Revised: June 17, 2013 Published: June 21, 2013 2759

dx.doi.org/10.1021/cm400719z | Chem. Mater. 2013, 25, 2759−2766

Chemistry of Materials

Article

to indiscriminately etch zeolite crystals with fluoride anions and its use in the preparation of hierarchical zeolites without changing their framework composition.

quantitative conditions on the ammonium form of the samples to quantify the majority of the tetrahedral species. 2.3. Catalytic Experiments. Isomerization of m-xylene (0.025 bar of m-xylene) was carried out at 623 K in a dynamic Pyrex reactor under atmospheric pressure. The catalyst bed contains 50 mg of zeolite mixed with 200 mg SiC, both in the 200−400 μm size range. The W/F is varied between 7 and 87 g·h·mol−1. The products are analyzed with a Bruker 430 online gas chromatograph (Varian Capillary Column: 25 m × 0.25 mm, df = 0.25 μm, CP-ChiraSil Dex-CB) fitted with a flame ionization detector.

2. EXPERIMENTAL SECTION A micrometer-sized commercial ZSM-5 sample (Süd-Chemie, NH4form) was used as a parent to prepare hierarchical materials. Fluoride media treatments were carried out with HF and NH4F-HF mixed aqueous solutions. HF treatments were performed as follows: i) 0.5 g of the parent ZSM-5 zeolite was dispersed in 15 mL of 0.5 mol L−1 HF solution and reacted at 338 K for 15 min under stirring; and ii) 0.5 g of the parent ZSM-5 zeolite was dispersed in 15 mL of 1.0 mol L−1 HF solution and reacted at 298 K for 6 min under stirring. Mixed NH4FHF treatment was performed as follows: 2.5 g of NH4F was dissolved in 15 mL of 0.2, 0.5, or 1.0 mol L−1 HF aqueous solution first, and then 0.5 g of the parent ZSM-5 zeolite was dispersed in the mixed solutions and react at 298 K for 6 min under stirring. All the samples were thoroughly washed using distilled water after the fluoride medium treatment. The samples treated by 0.5 and 1.0 mol L−1 HF solutions were denoted HF-0.5 and HF-1.0, and those treated by 0.2, 0.5, and 1.0 mol L−1 NH4F-HF solutions were denoted NH4F-HF-0.2, NH4F-HF-0.5, and NH4F-HF-1.0, respectively. The H-forms of the parent and acid modified zeolites were obtained by calcination in static air at 823 K for 2 h with a heating rate of 4 K min−1. A treatment of the parent ZSM-5 with a flow of NH4F-HF solutions was also carried out, to provide deeper insight in this chemical etching process. Vacuum filtration with a FB65432 mechanical vacuum pump (Fisher Scientific) was used to keep the etching solutions flowing. For details, see subsection 3.3.2. 2.2. Characterization Techniques. Powder X-ray diffraction (XRD) patterns were obtained with a PANalytical X’Pert Pro diffractometer using Cu Kα radiation (λ = 1.5418 Å, 45 kV, 40 mA). Before analysis, the powders (ca. 20 mg) of the as-synthesized and modified samples were loaded on a silicon wafer. The samples were studied in the 5−50° 2θ range with a scanning step of 0.0167° s−1. Electron micrographs were taken on a MIRA-LMH (TESCAN) scanning electron microscope (SEM) equipped with a field emission gun. Elemental analysis was performed by inductively coupled plasma−atomic emission spectroscopy (ICP-AES) using an OPTIMA 4300 DV (Perkin−Elmer) instrument. Nitrogen adsorptions were performed with a Micromeritics ASAP 2020 automated gas adsorption analyzer. Prior to analysis, the samples were outgassed at 373 K for 1 h and 573 K for 10 h. Specific surface areas were determined from the BET equation. The total volume was taken from nitrogen adsorbed volume at P/P0 = 0.99. The t-plot method was used to distinguish the micropores from the mesopores in the samples. Infrared spectra were recorded with a Nicolet Magna 550-FT-IR spectrometer at 2 cm−1 optical resolution. Prior to the measurements, the catalysts were pressed into self-supporting discs (diameter: 1.6 cm, 18 mg) and pretreated in the IR cell attached to a vacuum line at 823 K (2 K/min) for 5 h down to 10−6 Torr. The adsorption of pyridine and collidine was performed at 373 and 473 K, respectively. After establishing a pressure of 1 Torr at equilibrium, the cell was evacuated at 523 K in order to remove physisorbed species. All spectra were normalized to 20 mg wafers. The amount of pyridine adsorbed on Brønsted and Lewis sites was determined using the integrated area of the bands observed at 1545 cm−1 and 1454 cm−1, respectively. For collidine, the bands between 1632 and 1648 cm−1 were considered. The extinction coefficients used in this study were the following: ε(B)1545 = 1.02 and ε(L)1454 = 0.89 cm/mol for pyridine and ε(Me3Py)1632−1648 = 10.1 cm/mol for collidine.19 27 Al MAS NMR spectra were recorded on a Bruker Avance-400 (magnetic field of 9.4 T) spectrometer. Chemical shifts were referenced to a 0.1 M Al(NO3)3 aqueous solution. All NMR experiments were performed using the sample spinning speed of 12 kHz, a pulse recycling delay of 1 s, and a pulse length of 2.5 μs with 20 kHz B1 rf field. The 27Al MAS spectra were recorded under

3. RESULTS AND DISCUSSION 3.1. Physicochemical Properties of ZSM-5 Samples. SEM of the parent material shows that the sample is made of uniform well-defined crystals with the typical ZSM-5 morphology (Figure 1). Their size is about 3 × 1 × 6 μm

Figure 1. SEM micrographs of the parent ZSM-5 crystals.

along the a, b, and c directions, respectively. Typical 90° twin intergrowths are often observed on the (010) face. The Si/Al ratio of this sample is 19.0 (ICP analysis), and it exhibits high X-ray crystallinity without traces of other crystalline phase (Figure 2). 3.2. Etching of the ZSM-5 Zeolite with HF Solutions. 3.2.1. Structural and Compositional Properties of ZSM-5 Etched by HF Solutions. The first experiments on HF etching of ZSM-5 crystals were performed with 0.5 and 1.0 mol L−1 concentrated solutions, following Ghosh and Kydd.

Figure 2. XRD patterns of the parent and HF-treated ZSM-5 samples. 2760



Article

textural characteristics close to their parent (Table 1). While their micropore volume is preserved, their external surface area increases somewhat, a result of a minor increase of the volume of nonzeolitic pores (Vsec in Table 1. Vsec: secondary meso- and macropores formed during chemical treatment). SEM does not show substantial differences between the HF-0.5 and HF-1.0 samples. Intact agglomerated and deeply etched crystals are observed (Figure 4). The presence of severely etched samples could contribute to the increased nonzeolitic pore volume (Table 1). Since all samples are treated under the same conditions, the difference in their etching should be related with their growth mechanism. More precisely, the defects or domains with different chemical composition are more vulnerable to acid attack, and these parts of the crystal are first dissolved. To investigate the influence of temperature and time, the HF etching is carried at temperatures between 273 and 338 K during 2−120 min. Careful SEM inspection of the etched products does not indicate different dissolution patterns, showing that temperature and time do not influence significantly the reaction. 3.2.2. Equilibrium in HF Solutions and Their Etching Performance. To better control the HF etching process, it is important to understand why a simple increase of HF concentration modifies the etching selectivity.16,20 It is known that different species, namely, H+, F−, HF2−, and undissociated HF molecules, are present in HF solutions.17,21 It is also known that the SiO 2 etching rate depends linearly on the concentration of HF and HF2−.22 The activity of these two species is much higher than that of H+ in HF solutions.17 Hence, both the Ghosh and Kydd16 and our studies can be rationalized by considering that increasing the HF concentration not only increases acidity but also the concentration of the species responsible for silicon removal. It also explains why etching with a 1.0 mol L−1 HF solution leads to a zeolite with a Si/Al ratio lower than with a 0.5 mol L−1 treatment. Obviously a fine control of the equilibrium in fluoride solutions is imperative to promote a nonselective extraction of framework elements in a zeolite. 3.3. Etching of ZSM-5 with NH4F-HF Solutions. 3.3.1. NH4F-HF Treatment: Theoretical Considerations. We use a combination of ammonium fluoride − hydrofluoric acid solutions to perform nonselective framework dissolution of ZSM-5 crystals. The approach is based on the analysis of chemical equilibria in the HF−NH4F−H2O system. Diluted HF

The zeolites retain their crystallinity after such a treatment (Figure 2). However, the ICP analysis shows that the Si/Al ratio of the HF-0.5 zeolite is 37.3, almost twice than its parent (Table 1). When treated with a more concentrated solution Table 1. Framework Composition and Porosity of Parent and HF-Treated ZSM-5 Samples samples ZSM-5 parent HF-0.5 HF-1.0

Si/Al ratioa

SBETb (m2 g−1)

Vmicc (cm3 g−1)

Sextc (m2 g−1)

Vsecd (cm3 g−1)

19.0

377

0.18

9

0.02

37.3 32.1

367 381

0.17 0.17

20 29

0.05 0.06

ICP. bBET surface area. ct-plot. dVsec − secondary meso- and macropores formed during chemical treatment (Vsec = Vtotal − Vmic). a

(1.0 mol L−1 HF), the Si/Al ratio decreases to 32.1 (Table 1), confirming that more concentrated solutions are less selective with respect to aluminum. However, this Si/Al ratio still differs substantially from the parent. Like their parent, the isotherms of HF treated ZSM-5 display almost horizontal adsorption and desorption branches coupled with small hysteresis loops in the 0.5−1.0 partial pressure (P/P0) range (Figure 3). The

Figure 3. Nitrogen adsorption/desorption isotherms of the parent and HF-treated ZSM-5.

continuous ascent of the isotherms near saturation pressure is a sign of nitrogen adsorption in large mesopores or macropores. Accordingly, the HF treated samples possess

Figure 4. SEM micrographs of HF-0.5 (A) and HF-1.0 (B). 2761



Article

Figure 5. SEM images of NH4F-HF-1.0 (A), NH4F-HF-0.5 (B), and NH4F-HF-0.2 (C).

operating conditions (Figures 4B and 5A). The particularity of the dissolution process is that the central parts of the crystals where the twin crystal is situated are deeply dissolved. Obviously the intergrowth areas are more vulnerable to chemical attack. Further, the place where the twin crystals are situated is severely attacked by the acid solution. Some of the crystals are sharply cut suggesting the presence of defect zones which dissolve preferentially. Although the crystals are severely etched, the crystallinity of the sample treated with NH4F-HF is retained (Figure 6). The Si/Al ratio of severely etched zeolite crystals is 21.8 (Table 2), very close to that of its parent. Thus, ZSM-5 crystals can be etched with almost no change of their framework composition. This opens the door to more rationally designed hierarchical zeolites.

behaves as a weak acid, in sharp contrast with other hydrogen halides.17,23 The ionization of HF in water can be represented by two steps: HF ⇌ H+ + F−

(1)

HF + F− ⇌ HF2−

(2)

The equilibrium constants at 298 K are 6.85 × 10−4 mol L−1 for 1 and 3.963 L mol−1 for 2.24 NH4F is a strong electrolyte; therefore, its addition to an HF aqueous solution will shift equilibrium 1 to the left. In addition, the large concentration of F− will favor the generation of bifluoride ions, HF2− in 2. It is known that both HF2− and HF are active for silicon removal, and the rate of silicon dissolution with HF2− is at least four to five times as fast as that with HF.22,25 In HF solutions, the concentration of HF2− is very low due to equilibrium 1. An increase in HF concentration raises the number of nondissociated HF molecules and the concentration of HF2−. This shift of equilibrium explains why more concentrated HF solutions are less aluminum selective (Table 1). In the case of NH4F-HF solutions, the presence of NH4F not only suppresses dissociation of HF and decreases the concentration of H+ but also favors the formation of a large number of HF2−. As a result, the silicon extraction is increased and no preferential extraction of Al takes place, in contrast to diluted HF solutions. 3.3.2. NH4F-HF Treatment: Experimental Results. A series of treatments with NH4 F-buffered HF solutions were performed, and the resulting samples carefully studied. As can be seen in Figure 5A, a substantial dissolution of zeolite crystals occurs on the sample treated with a mixture of NH4F and 1.0 mol L−1 HF (NH4F-HF-1.0). The dissolution of zeolite crystals is much more pronounced than with HF alone under the same

Figure 6. XRD patterns of the parent and NH4F-HF treated ZSM-5. 2762



Article

on the respective roles of HF and NH4F during the etching process. While HF is the source of the active species (i.e., undissociated HF molecules and HF2−), NH4F increases the supply of HF and HF2− by providing a large amount of F− anions. This gives also new insights in the dissolution behavior of porous aluminosilicate in fluoride media, a topic so far not thoroughly addressed in the literature. It still remains unclear, however, how NH4F-HF affects the zeolite morphology during the etching process. A more detailed investigation of this type of etching was initiated to clarify the role of experimental conditions on secondary porosity and zeolite crystal morphology. The formation of twin crystals is a well-known feature of MFI-type zeolites.26−28 Typically, the intergrowths nucleate from small areas on (010) faces of growing crystals,11 as can be seen in Figure 1. Since interfaces between twin crystals are potential breaking point rich in defects,29 it is reasonable to assume that the dissolution of crystals takes place preferentially at these locations. It is not clear, however, what level of surface etching NH4F-HF will generate on zeolite crystal faces. SEM pictures of series of samples etched with different concentration of HF (0.05, 0.1, and 0.5 mol L−1) at a constant NH4F (2.5 g per 15 g solution) amount (Table 3) produce

Table 2. Framework Composition and Porosity of ZSM-5 Treated by NH4F-HF Solutions samples ZSM-5 parent NH4F-HF1.0 NH4F-HF0.5 NH4F-HF0.2

Si/Al ratioa

SBETb (m2 g−1)

Vmicc (cm3 g−1)

Sextc (m2 g−1)

Vsecd (cm3 g−1)

19.0

377

0.18

9

0.02

21.8

407

0.18

6

0.03

19.3

427

0.18

16

0.04

19.2

390

0.19

13

0.03

ICP. bBET surface area. ct-plot. dVsec − secondary meso- and macropores formed during chemical treatment (Vsec = Vtotal − Vmic) a

Further etchings were performed with NH4F-HF solutions at lower HF concentrations. As can be seen in Table 2, the Si/Al ratio of sample NH4F-HF-0.5 is identical (Si/Al = 19.3) to its parent ZSM-5 while that of HF-0.5 is 37.3 (Table 1). A further decrease of HF concentration (NH4F-HF-0.2) yields a material with a Si/Al equal to the parent zeolite (Table 2). These results highlight that a relatively large window exists where the chemical etching of ZSM-5 is not dependent on HF concentration. Besides preserving the Si/Al ratio of the parent zeolite, this mode of etching allows an excellent preservation of micropore volume. As can be seen by N2 physisorption, the uptake at low pressure is similar for all samples (Figure 7 and Table 2). The

Table 3. Experimental Conditions for Etching of ZSM-5 with NH4F-HF Solutions samples

mass of parent ZSM-5 (g)

HF concn in mixed solna (mol L−1)

etching temp (K)

etching time (s)

D1 D2 D3 D4

0.5 0.5 0.5 0.5

0.05 0.1 0.5 0.5

293 293 293 293

30 30 10 30

a

The NH4F-HF mixed solutions contain different amounts of HF and a constant amount of NH4F (2.5 g per 15 g solution).

hierarchical zeolites (Figure 8) displaying a clear trend. Figure 8A shows that small intergrown domains on the surface of zeolite crystals are promptly removed (Sample D1, Figure 8A) leaving the crystal faces intact. Increasing HF concentration leads to further crystal dissolution. As a result, some large secondary pores appear on the zeolite crystal surface, located at the negative forms remaining after the dissolution of small twin crystallites (Sample D2, Figure 8B). The crystal domains left after the removal of small intergrown parts are more vulnerable to fluoride attack. For sample D3, although the intergrown area is severely dissolved at the beginning of the etching process, only a few macropores are found in the portion of crystal face that does not contain twin crystals (Figure 8C). By extending the etching from 10 to 30 s, other parts of zeolite crystals dissolve, introducing macropores and/or large mesopores into areas that remained intact at shorter treatment time (Sample D4, Figure 8D). Another specificity of such an etching process is its fast reaction rate. Under the above conditions, an extensive dissolution of zeolite crystals is observed within 10 s (Figure 8C). This enables the selective extraction of the most vulnerable zones of the zeolite crystal, i.e. NH4F-HF dissolves preferentially the defect zones in ZSM-5. Once the etching process starts on the crystals surface the reaction goes further rapidly and reaches the inside of the crystals. Thus the surface etching that usually results in high external surface area is limited, and large meso- and macropores penetrating deeply

Figure 7. Nitrogen adsorption/desorption isotherms of the parent and ZSM-5 treated by NH4F-HF.

parent and treated samples show slight differences at high relative pressure where a small second uptake, due to the formation of very large pores, is observed. The formation of large pores in zeolite crystals is clearly visible on SEM images (Figure 5). On the other hand, no smaller mesopores are detected (Table 2). Another important difference between HF and NH4F-HF etching is the degree of zeolite dissolution. SEM of a zeolite etched with NH4F-HF shows a more homogeneous dissolution of the crystals than with HF (Figures 4 and 5). Further, for similar periods of time, the crystals etched with NH4F-HF are more dissolved than their counterparts treated with HF solutions of similar concentration. The above results unambiguously prove that the combination of HF with NH4F extracts equally Si and Al in zeolites. This is in sharp contrast to the other chemical approaches currently used for postsynthesis modifications. Our results also shed light 2763



Article

Figure 8. SEM images of ZSM-5 crystals treated by flowing NH4F-HF solutions under vacuum at 293 K. The NH4F-HF solutions contain different amounts of HF (0.05, 0.1, and 0.5 mol L−1) and a constant amount of NH4F (2.5 g per 15 g solution): (A) treated by a 0.05 mol L−1 solution for 30 s, (B) treated by a 0.1 mol L−1 solution for 30 s, (C) treated by a 0.5 mol L−1 solution for 10 s, (D) treated by a 0.5 mol L−1 solution for 30 s.

0.57 nm) and 2,4,6-collidine (kinetic diameter: 0.74 nm) are used to probe acid sites located in the micropores and on the external surface of zeolites crystals, respectively. The difference spectra, after pyridine and collidine adsorption, are reported in the Supporting Information (Figure S2 and Figure S3). The accessibility index (ACI) for these molecules is reported in Table 4.

inside the crystals are formed. Such a dissolution mechanism has the advantage to remove defect zones and produces crystals with high levels of perfection. Further refinements of this fluoride process are expected to provide zeolite catalysts with predetermined desirable features, i.e. true zeolite crystal engineering. 3.4. Acidic and Catalytic Properties of Hierarchical Zeolites. 3.4.1. IR Study of Acidic Properties. Fluorination of zeolites was already used to modulate their acidity and hence their catalytic activity.16,30−35 A specific calcination procedure was used to optimize the interaction of fluoride ions with the zeolite framework.33,36 Highly electronegative fluoride ions incorporated in the zeolite lattice polarize the structure and thus substantially increase surface acidity.31 The fluoride treatment in the present study is exclusively performed under hydrothermal conditions. After fluoride etching, the samples are washed thoroughly using distilled water and dried overnight at room temperature before their conversion to an H-form by calcination. 19F NMR analysis (not shown) does not reveal the presence of fluoride species in the sample and they have no effect on its acidity. The IR spectra of the samples after activation are presented in Figure S1. All samples exhibit two distinct absorption bands at 3740 and 3610 cm−1 assigned to silanols and bridged hydroxyls (Brønsted acid sites), respectively. The third broad band that can be observed in the parent sample around 3500 cm−1 is generally assigned to silanol nests interacting through hydrogen bonding. These species are obviously more vulnerable to the etching process since they were not observed after the fluoride media treatment. The acidity is monitored by IR spectroscopy study of probe molecules of different size. Pyridine (kinetic diameter:

Table 4. Brønsted (B) and Lewis (L) Acid Sites by IR of Adsorbed Pyridine (Py) and 2,4,6-Collidine (μmol/g, 523 K) sample

B-Py

L-Py

collidine

ACI Pya

ACI Collb

MFI parent HF-0.5 HF-1.0 NH4F-HF-0.2 NH4F-HF-0.5 NH4F-HF-1.0

643 331 406 717 614 584

30 26 63 44 52 87

12.9 0.8 8.9 2.7 4.0 7.0

0.86 0.87 0.98 0.98 0.86 0.98

0.02 0.002 0.02 0.004 0.005 0.01

a

Accessibility index of pyridine molecules = BPy +LPy/total amount of aluminum. bAccessibility index of collidine molecules = total amount of collidine molecules adsorbed/total amount of aluminum.

IR shows that the parent zeolite and its hierarchical derivatives prepared either by HF or NH4F-HF etching contain mainly Brønsted acid sites (Table 4). This result is consistent with the 27Al MAS NMR spectra in Figure 9 where a single peak at ca. 55 ppm highlights the sole presence of tetrahedral aluminum on all ZSM-5. After HF etching both samples show a substantial loss of Brønsted acidity with respect to their parent (Table 4), which is 2764



Article

decreases the activity of the resulting zeolites (Figure 10), in agreement with the observed loss of Brønsted acid sites (Table 4). The catalytic performance of hierarchical materials obtained by NH4F-HF treatment is substantially improved (Figure 10). Both NH4F-HF-treated samples show a higher conversion their parent. For the NH4F-HF-treated ZSM-5 zeolites, the surface area, microporosity, aluminum content, the number, and distribution of acid sites are similar to their parent. Thus, the most plausible explanation for the increased activity of the NH4F-HF treated zeolites is an increased accessibility to the zeolite micropores. With a NH4F-HF treatment, most zeolite crystals are more fragmented (Figures 5 and 8). Small parts of the crystals are cut by a selective dissolution of defect zones. In addition, large pores penetrating deep in the crystals are created (Figures 5 and 8). These large mesopores and/or macropores lead to an easier access to the micropores, similar to a reduction in crystal size.38 It is worth nothing also that the particularity of the NH4F-HF etching process is that the surface of the zeolite crystals is not severely etched providing small mesopores that substantially increase the external surface area. As already discussed in previous sections the very reactive FHF− anion dissolves namely defect zones in the crystals and removes the structure defects. This way the etching process leads to almost defect-free zeolite catalysts, where the remaining crystalline domains are to a great extent uniform as in chemical composition and structure features. In addition, the increased catalytic activity of NH4F-HF treated zeolites is most probably due to a secondary effect of fluoride media etching. As evidenced in Figure 1 and Figure 8, the zeolite crystals of this study consist of intergrown building blocks. The interfaces of these individual segments act as diffusion barriers due to a local mismatch in the alignment of the micropore network,39 making some portions of the zeolites inaccessible to reactant molecules and affecting their adsorption capacity and catalytic activity.40−43 The small intergrowth domains on the zeolite surface are disintegrated during fluoride treatment. As a result, the diffusion of molecules in and out is easier, which certainly influence the catalytic conversion of m-xylene.

Figure 9. 27Al MAS NMR spectra of the parent and fluorine treated ZSM-5.

in agreement with the aluminum analysis (Table 1). The amount of Lewis acid sites present in the material seems to be related to HF concentration and is almost twice higher than that of initial material. The IR study showed higher pyridine accessibility to hierarchical zeolite crystals. On the other hand, the collidine accessibility to HF-0.5 and HF-1.0 samples is substantially different. It is very low for HF0.5 indicating a substantial removal of external acid sites, whereas HF-1.0 shows accessibility similar to the parent zeolite. These results confirm the direct relation between HF concentration and its selectivity with respect to framework zeolite cations. In sharp contrast, the acidity of the NH4F-HF etched samples does not change substantially compared to nontreated ZSM-5 material. This further proves the advantage of the HF-NH4F combination on the Si and Al framework extraction (Table 2). In addition, the accessibility toward pyridine is improved even at low concentration of the etching solution. 3.4.2. m-Xylene Isomerization on the Parent and Fluorine Etched Zeolites. The conversion of a xylene isomer is a convenient way to monitor changes induced by treatments on ZSM-5.37 The m-xylene conversion on various ZSM-5 zeolites prepared in this study is shown in Figure 10. All catalysts are very stable, after at least 4 h on stream. Little (0.2%) disproportionation to toluene and trimethylbenzenes takes place. Catalytic data, collected after 0.5 h on stream, are selected to compare the different samples with space times between 7−87 g·h·mol−1. Etching the ZSM-5 with HF alone

4. CONCLUSION We present a comprehensive study on the effect of fluoride treatments on the secondary pore formation in ZSM-5 crystals. The etching performance of HF solutions is concentration dependent. Diluted (

Chemical Equilibrium Controlled Etching of MFI-Type Zeolite and Its

Recommend Documents