Communication pubs.acs.org/JACS
Opening the Cages of Faujasite-Type Zeolite Zhengxing Qin,†,§ Katie A. Cychosz,‡ Georgian Melinte,⊥ Hussein El Siblani,† Jean-Pierre Gilson,† Matthias Thommes,‡ Christian Fernandez,† Svetlana Mintova,† Ovidiu Ersen,⊥ and Valentin Valtchev*,† †
Laboratoire Catalyse et Spectrochimie, Normandie Univ, ENSICAEN, UNICAEN, CNRS, 14000 Caen, France, 6 Bd Maréchal Juin, 14000 Caen, France ‡ Quantachrome Instruments, 1900 Corporate Drive, Boynton Beach, Florida 33426, United States ⊥ Institut de Physique et de Chimie de Strasbourg, Université de Strasbourg 23, rue du Loess BP 43, F-67034 Strasbourg, France S Supporting Information *
sodalite cages in FAU). The FAU zeolite is at the core of many important catalytic (e.g., FCC and hydrocracking to convert heavy oil fractions to transportation fuels) and separation (e.g., selective adsorption of aromatic isomers in petrochemistry) processes. There is currently a sustained effort to discover new zeolites with larger pores, able to process bulkier molecules.11 Some zeolitic materials with extra-large pores (up to 2 nm) were synthesized; their framework density (10.5 T/1000 Å3 for ITV-type zeolite) is lower than that of FAU-type (13.3 T/1000 Å3).10 However, none of these materials reached the stage of industrial use due to cost and thermal stability issues. The FAUtype zeolites thus remain the largest pore zeolite currently used. Postsynthesis treatment (steaming, chemical etching) was used, from their industrial beginning in catalysis (FCC), to modify the framework compositions12 and/or introduce a secondary porosity in zeolite crystals.13 Recently, GarciaMartinez et al. used caustic leaching together with a surfactant, to recrystallize part of the zeolite and generate a fairly homogeneous distribution of mesopores in zeolite Y crystals; this material is currently used in some oil refinieries.14 These treatments are based on a preferential extraction of Si (caustic media) or Al (acid media or steaming) framework cations. With such caustic or acid extractions, the concentration of the etching solution decreases steadily and leads to a steep gradient of etchant from the outside to the inside of zeolite crystals. These extractions are difficult to control and quite heterogeneous, i.e. large portions of some crystals can be modified while others remain almost intact. Recently, we developed a new method based on etching with NH4F.15 It takes advantage of the double hydrolysis of NH4F, generating steadily and in situ small amounts of HF2−.15 Its novelty is that the latter species extracts framework Si and Al cations at equal rates. The method is also easier to control (mild conditions) and tunable as the etching species, present in small concentration, is homogeneously distributed inside the crystals. Dissolution kinetics can be controlled by both the concentration and the temperature of the etching solution. To better control the etching rate and extent, we elected to run the reaction at 277 K. We highlight the very significant result that this method, applied to the important FAU zeolite, increases its accessible
ABSTRACT: Zeolites are widely used in industrial processes, mostly as catalysts or adsorbents. Increasing their micropore volume could further improve their already exceptional catalytic and separation performances. We report a tunable extraction of zeolite framework cations (Si, Al) on a faujasite-type zeolite, the archetype of molecular sieves with cages and the most widely used as a catalyst and sorbent; this results in ca. 10% higher micropore volume with limited impact on its thermal stability. This increased micropore volume results from the opening of some of the small (sodalite) cages, otherwise inaccessible to most molecules. As more active sites become accessible, the catalytic performances for these modified zeolites are substantially improved. The method, based on etching with NH4F, is also applicable to other cage-containing microporous molecular sieves, where some of the most industrially relevant zeolites are found.
A
mong the different types of porous materials (micro-, meso-, and macro-porous),1 zeolites have so far the largest impact on our lives.2,3 These crystalline microporous materials are extensively used in petroleum refining, petrochemistry, pollution abatement, household, and more recently in renewable energy production, electronic, optical and medical applications.4−7 Their impact is also expected to widen further. One way to do so would be to provide materials with more micropores (≤2 nm) accessible to a wider variety of molecules. In particular, zeolite molecular sieves with pores/cages in the 1−2 nm range are highly desired. Zeolite structures are formed by corner sharing tetrahedra (TO4) where T is usually Si or Al. Connecting them leads to larger units (3-, 4-, 5-, and 6-member rings) further organized to develop a network of channels and/or cages.8,9 The zeolite with the largest pore currently used in industry is faujasite (FAU-type), with pore openings of 0.74 nm leading to accessible large cages (supercages, size of 1.12 nm) and smaller cages (sodalite cages [SOD], size of 0.63 nm), Figure S1.10 The latter are unfortunately not accessible, even to small molecules, due to their narrow 6-member ring windows (diameter of 0.25 nm).10 The presence of these sodalite cages, where active sites are located, implies that these materials are not yet used to their full potential (40% of their Al is located in supercages vs 60% in © 2017 American Chemical Society
Received: September 27, 2017 Published: November 21, 2017 17273
DOI: 10.1021/jacs.7b10316 J. Am. Chem. Soc. 2017, 139, 17273−17276
Communication
Journal of the American Chemical Society micropore volume. Other zeolites with small cages (LTA, CHA, LTL, RHO, ...) contributing substantially to their microporosity could also benefit from such a treatment. Such confined environments can indeed be unavailable or poorly accessible for adsorption or catalysis due to the very small diameter of their openings (4- and 6-member rings). A commercial Na,NH4-Y zeolite (Y-54 from UOP, hereinafter denoted P-Y) is etched with an aqueous 25 wt % NH4F solution (see Supporting Information). A sample was kept for 1 min in 25 wt % NH4F solution at 277 K, washed with distilled water and further analyzed by TEM-EDS (Figure S2). The TEM-EDS mapping reveals a uniform fluorine distribution, highlighting that NH4F penetrates very fast all the zeolite porosity. The parent zeolite was also etched under similar conditions for 5, 10, and 20 min; the corresponding samples are designated Y-F5, Y-F10, and Y-F20, respectively. Their XRD patterns do not show any loss of crystallinity or presence of other crystalline phases (Figure S3). A careful inspection of their XRD patterns reveals changes in the intensity of certain peaks, suggesting some structural changes. However, all treated zeolites are indexed in the Fd3̅m symmetry, typical of the FAUtype zeolite, and all unit cell parameters are identical (Table S1). A marginal increase in Si/Al ratio (2.7 and 2.9) is observed after 10 and 20 min of treatment. Figure 1 shows nitrogen isotherms (77 K) on the parent and its Y-F5 and Y-F10 derivatives (Figure 1a), as well as argon isotherms (87 K) on the parent and its Y-F10 and Y-F20 derivatives (Figure 1b). Both nitrogen and argon isotherms are of the type Ia for the parent and 5 and 10 min treated samples.1 The micropore volume increases from 0.33 cm3/g (P-Y) to 0.36 cm3/g (Y-F10), ca. 10% increment, in agreement with a tplot analysis of the nitrogen isotherms (Table S2). A negligible mesoporosity is observed on Y-F10 but its isotherm is still very similar to its parent (Figure. 1c). As supercages are the only space in FAU accommodating N2 (0.36 nm) and Ar (0.34 nm),16 the increased micropore volume of Y-F5, Y-F10, and YF20 indicates that additional accessible space is created by fluoride etching (Figure S4). After a longer treatment, Y-F20, the mesoporosity increases as the newly created mesopores reach a size of about 3 nm (Figure S4). Though the micropore volume of Y-F20 is lower than Y-F10, it is still slightly higher than the parent (Table S2). The pore size of zeolites can also be studied by relating the confinement of Xe molecules to their chemical shift; hyperpolarized 129Xe NMR is a very sensitive tool to monitor such minute changes.17 The parent, P-Y, exhibits a single peak (53− 75 ppm) in the 225−370 K temperature range, attributed to xenon atoms confined in supercages (Figure S5). The kinetic diameter of xenon (0.44 nm),18 larger than nitrogen and argon, prevents its access to sodalite cages through their 6MR. All NH4F treated samples (Y-F5, Y-F10, and Y-F20) exhibit such a peak and a new one at higher chemical shifts (70−110 ppm), first as a broad shoulder below 300 K (Figures S6−S8) evolving to a single broad peak at 225 K (Figure 2). This indicates the presence of two types of connected cavities in the treated samples. Based on the structural features of the FAU framework, the second peak can be attributed to xenon atoms located in now open sodalite cages, a more confined environment than the supercages. The collapse of the high (open SOD) and low (super cage) chemical shift peaks in a single one at high temperatures points to a fast exchange of xenon between these two locations; this implies that these two environments are quite close to each other and communicating.
Figure 1. (a) Nitrogen (77 K) isotherms for P-Y, Y-F5, and Y-F10, (b) argon (87 K) isotherms for P-Y, Y-F10, and Y-F20 plotted on a logarithmic x-axis, (c) cumulative pore volume calculated from the argon (87 K) adsorption isotherms using a DFT model assuming spherical micropores and cylindrical mesopores, which takes into account the delay in condensation due to metastable pore fluid.
Figure 2. HP 129Xe NMR at 225 K of the parent (P-Y) and derivatives treated for 5, 10, and 20 min. Inset: HP 129Xe NMR spectra highlighting xenon located in sodalite cages. 17274
DOI: 10.1021/jacs.7b10316 J. Am. Chem. Soc. 2017, 139, 17273−17276
Communication
Journal of the American Chemical Society The high chemical shift peaks could also be attributed to Xe interacting strongly with extra-framework Al; however, 27Al NMR does not detect any such species (Figure S9), ruling out this alternative.19 The accessibility of open sodalite cages can also be probed by ion-exchange with tetramethylammonium cations (TMA, kinetic diameter 0.44 nm), too bulky to enter closed sodalite cages.20 Y-F5 shows a 2.4 wt % higher TMA capacity than its parent (Figure S10a). Moreover, the thermogram (TG/DSC) of the TMA-exchanged Y-F5 displays a new exothermic peak at a higher temperature (733 K) than its parent, due to a more difficult combustion of some TMA, located in a more confined environment than the supercages, i.e., open sodalite cages.21 Furthermore, on Y-F5, the combustion temperature of the TMA located in supercages is lower (Figure S10b), indicative of reduced transport limitations. The FAU-type structure is particularly prone to an opening of its sodalite cages because extracting a single framework (Si or Al) cation leads to the opening of a sodalite cage and a subsequent increase of micropore volume. TEM of the samples shows that the macromorphological features of Y-F5 and Y-F10 are identical. High resolution transmission electron microscopy (HR-TEM) and 3D electron tomography (ET) on Y-F5 do not show any evidence of dissolution, suggesting the process takes place at an atomic scale, impossible to monitor by HR-TEM. The coherent crystalline domains do not contain any traces of structural alteration, further illustrating that, at the onset of the etching process, when the micropore volume increases without the appearance of mesoporosity, the extraction of framework cations is already taking place. First evidence of a longerrange dissolution is observed on Y-F10 at the interface between two, much larger in size, neighboring crystalline domains (Figure 3a−c). This confirms our previous observations that defect zones are dissolved by such a nonselective chemical etching leading ultimately to the formation of extended mesopores.22 Indeed, some mesoporosity appears in the sorption measurements (Figure S4, Table S3) and is clearly visible by TEM in Y-F10 (Figure 3d). Extending the treatment to 20 min results in the formation of a number of randomly distributed, but relatively uniform in size (around 3 nm) mesopores (Figure 1c and Figure S11). TEM tomography analysis showed that some of the mesopores are connected (Video 1, Supporting Information). This TEM study is well in line with the physisorption measurements (Figure S4 and Table S3). The catalytic consequences of this new etching methodology are highlighted in two reactions: (i) dealkylation of 1,3,5 triisopropylbenzene (TiPBz, kinetic diameter 0.95 nm)23 (ii) hydroconversion of n-octane (n-C8, kinetic diameter 0.43 nm)24 on derived bifunctional catalysts (Pt/zeolite with Pt = 0.5 wt %) to minimize deactivation issues. The bulky TiPBz is known to probe specifically the external and mesoporous surfaces of zeolites, including large pore FAU, whereas n-C8 penetrates all microporosity and reacts also on the external surface. Figure 4a shows that TiPBz conversion is identical on samples with similarly low mesoporosity (P-Y, Y-F5, Y-F10) but increases significantly when mesoporosity is added (YF20); this reaction is insensitive to atomic level changes (SOD cages opening) in micropores the reactant cannot enter. On the other hand (Figure 4b,c), the n-C8 conversion on the parent, PY (5 to 44% by varying space time only) increases significantly
Figure 3. HR-TEM images of Y-F10: (a) a coherent crystalline domain; (b) first evidence of dissolution at the interface of two coherent domains; (c) two coherent domains are color-coded and the inset represents two crystalline networks overlapping; (d) low magnification TEM image of Y-F10 highlighting the formation of mesopores along defect zones.
after 5 min of NH4F etching, Y-F5 (11 to 72%). This is due to the now accessible acid sites in opened sodalite cages since fewer Brønsted sites are located in the supercages (IR spectroscopy, Figures S12 and S13). All catalytic results point to conclusions similar to those from the physical and spectroscopic characterizations, i.e., the opening of a portion of the sodalite cages resulting in a better accessibility of their acid/active sites. An important issue related to the commercial use of zeolite catalysts and sorbents is their thermal stability. High temperature X-ray diffraction (Figure S14) indicates that all samples, including the most severely etched (Y-F20), are very stable. High temperature X-ray diffraction indicates that the structure of the treated zeolites collapses at 873 K, i.e., 100 K lower that their parent. Such a thermal stability is sufficiently high to withstand most harsh operational conditions (e.g., hydrocracking in oil refining, aromatics processing in petrochemistry, desorption step in separations, ...). In conclusion, the nonselective chemical etching with NH4F of a cage-containing zeolite increases its micropore volume without or with a minor formation of mesopores. This is achieved by a fast, easily tunable atomic scale extraction of zeolite framework cations. In the particular case of FAU, the resulting increase in accessible microporosity is due to the opening of some sodalite cages. This approach can also be used to increase the micropore volume of other zeolites with cages. This opens new opportunities to increase by a simple postsynthesis treatment, the adsorption capacity, enable adsorption of bulkier molecules and further improve the catalytic performances of the most industrially relevant zeolites, including their nanosized counterparts. 17275
DOI: 10.1021/jacs.7b10316 J. Am. Chem. Soc. 2017, 139, 17273−17276
Communication
Journal of the American Chemical Society Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This study was sponsored by the grant ANR-15-CE06-0004-02 (DirectCatSynBioFuel).
Figure 4. (a) Dealkylation of TiPBz on P-Y, Y-F5, Y-F10, and Y-F20 (T = 453 K, W/F° = 0.022 kg h mol−1, P = 0.1 MPa,). Hydroconversion of n-C8 on (b) P-Y and (c) Y-F5 (T = 573 K, W/F° = 8− 20 kg h mol−1, P = 5 MPa, H2/n-C8 = 6−500).
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b10316. Details for the preparation, characterization, and catalytic evaluation (PDF) Video of mesopores (AVI)
■
REFERENCES
(1) Thommes, M.; Kaneko, K.; Neimark, A. V.; Olivier, J. P.; Rodriguez-Reinoso, F.; Rouquerol, J.; Sing, K. S. W. Pure Appl. Chem. 2015, 87, 1051−1069. (2) Flanigen, E. M.; Broach, R. W.; Wilson, S. T. In Zeolites in Industrial Separation and Catalysis; Kulprathipanja, S., Ed.; WileyVCH: Weinheim, 2010; pp 1−26. (3) Zeolites and Catalysis: Synthesis, Reactions and Applications; Cejka, J., Corma, A., Zones, S., Eds.; Wiley-VCH: Weinheim, 2010. (4) Martínez, C.; Corma, A. Coord. Chem. Rev. 2011, 255, 1558− 1580. (5) Vermeiren, W.; Gilson, J.-P. Top. Catal. 2009, 52, 1131−1161. (6) Ennaert, T.; van Aelst, J.; Dijkmans, J.; de Clercq, R.; Schutyser, W.; Dusselier, M.; Verboekend, D.; Sels, B. F. Chem. Soc. Rev. 2016, 45, 584−611. (7) Sherman, J. D. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 3471− 3478. (8) McCusker, L. B.; Liebau, F.; Engelhardt, G. Pure Appl. Chem. 2001, 73, 381−394. (9) Newsam, J. M. Science 1986, 231, 1093−1099. (10) http://www.iza-structure.org/databases/ (accessed March 23, 2017). (11) Jiang, J.; Yu, J.; Corma, A. Angew. Chem., Int. Ed. 2010, 49, 3120−3145. (12) van Donk, S.; Janssen, A. H.; Bitter, J. H.; de Jong, K. P. Catal. Rev.: Sci. Eng. 2003, 45, 297−319. (13) Valtchev, V.; Mintova, S. MRS Bull. 2016, 41, 689−693. (14) Garcia-Martinez, J.; Xiao, C.; Cychosz, K. A.; Li, K.; Wan, W.; Zou, X.; Thommes, M. ChemCatChem 2014, 6, 3110−3115. (15) Qin, Z.; Gilson, J.-P.; Valtchev, V. Curr. Opin. Chem. Eng. 2015, 8, 1−6. (16) Sing, K. S. W.; Williams, R. T. Part. Part. Syst. Char 2004, 21, 71−79. (17) Fraissard, J.; Wang, L.-Q. In Hyperpolarized Xenon-129 Magnetic Resonance: Concepts, Production, Techniques and Applications; Meersmann, T., Brunner, E., Eds.; The Royal Society of Chemistry: Cambridge, 2015. (18) Chen, F.; Chen, C.-L.; Ding, S.; Yue, Y.; Ye, C.; Deng, F. Chem. Phys. Lett. 2004, 383, 309−313. (19) Chen, Q. J.; Fraissard, J. J. J. Phys. Chem. 1992, 96, 1809−1814. (20) Pérez-Ramírez, J.; Verboekend, D.; Bonilla, A.; Abelló, S. Adv. Funct. Mater. 2009, 19, 3972−3979. (21) Mintova, S.; Valtchev, V. Stud. Surf. Sci. Catal. 1999, 125, 141− 148. (22) Qin, Z.; Melinte, G.; Gilson, J.-P.; Jaber, M.; Bozhilov, K.; Boullay, P.; Mintova, S.; Ersen, O.; Valtchev, V. Angew. Chem., Int. Ed. 2016, 55, 15049−15052. (23) Awala, H.; Gilson, J.-P.; Retoux, R.; Boullay, P.; Goupil, J.-M.; Valtchev, V.; Mintova, S. Nat. Mater. 2015, 14, 447−451. (24) Lugstein, A.; Jentys, A.; Vinek, H. Appl. Catal., A 1999, 176, 119−128.
AUTHOR INFORMATION
Corresponding Author
*
[email protected] ORCID
Zhengxing Qin: 0000-0002-0277-9885 Christian Fernandez: 0000-0002-5476-3148 Svetlana Mintova: 0000-0002-0738-5244 Present Address §
State Key Laboratory of Heavy Oil Processing, College of Chemical Engineering, China University of Petroleum (East China), Qingdao 266580, China 17276
DOI: 10.1021/jacs.7b10316 J. Am. Chem. Soc. 2017, 139, 17273−17276
