Structural analysis of hierarchically organized zeolites

Abstract

Advances in materials synthesis bring about many opportunities for technological applications, but are often accompanied by unprecedented complexity. This is clearly illustrated by the case of hierarchically organized zeolite catalysts, a class of crystalline microporous solids that has been revolutionized by the engineering of multilevel pore architectures, which combine unique chemical functionality with efficient molecular transport. Three key attributes, the crystal, the pore and the active site structure, can be expected to dominate the design process. This review examines the adequacy of the palette of techniques applied to characterize these distinguishing features and their catalytic impact.

Figure 1. Hierarchical organizations in zeolites.

(a–i) Compared with a bulk material (a), TEM micrographs illustrate the distinct ways in which zeolites can be furnished with hierarchical pore structures. Both bottom-up and top-down synthesis approaches can be followed to configure the secondary porosity either within (intracrystalline, b and c) or between (intercrystalline, d–i) the zeolite crystals; a mesoporous USY zeolite attained by demetallation (b), a macroporous MFI-type zeolite prepared by steam-assisted crystallization (c), a nanosized Y-zeolite directly synthesized by a non-templated approach (d), an ITQ-2 zeolite derived by delamination of MCM-22 (e), an intergrown assembly of spherical silicalite-1 nanocrystals attained by confined synthesis in a mesoporous carbon template (f), a self-pillared assembly of ZSM-5 lamellae prepared by repetitive branching (g), silica-pillared ZSM-5 nanosheets synthesized by surfactant templating (h) and an organic–inorganic-layered hybrid with organic linkers covalently bonded to ICP-1P zeolite layers (i). This review examines the state of the art in the structural analysis of these morphologically diverse materials with the aim of establishing directions for their improved design in catalytic applications. Scale bars, 20 nm (a,b,g,i), 200 nm (c,f), 10 nm (d,e,h). (a–i) Reprinted with permission from ref. (a, © 2014 Macmillan Publishers Ltd), 105 (b, © 2011 American Chemical Society), 106 (c, © 2015 John Wiley and Sons Inc.), 49 (d, © 2015 Macmillan Publishers Ltd), 107 (e, © 1998 Macmillan Publishers Ltd), 108 (f, © 2011 American Chemical Society), 44 (g, © 2010 AAAS), 37 (h, © 2010 American Chemical Society) and 109 (i, © 2014 American Chemical Society).

Figure 2. Opportunities and threats for zeolite catalysis.

(a) The linear correlation between the yield of methylenedianiline (MDA) mixtures and the mesopore surface area factored by the concentration of Brønsted acid sites evidenced over FAU-type zeolites in the liquid phase condensation of aniline with formaldehyde to MDA, a key intermediate in polyurethane production, exemplifies the need to balance active site and pore quality to maximize their performance. (b) Compared with other zeolite framework types, the activity of the FAU-type catalysts could be significantly enhanced while retaining the unique shape selective properties by preserving the crystal structure and alleviating the diffusion constraints. (c) Evaluation of nanosized (2D) MFI-type zeolites hydrothermally synthesized for different durations (1–9 d) in the alkylation of toluene with isopropanol reveals that at least two pentasil layers (9 d) are required to substantially increased para-selectivity with respect to that expected thermodynamically (equilibrium, EQ), although this remained inferior to that observed over the bulk zeolite. (d) The key role of both pore and active site quality was also demonstrated by the varying catalyst lifetimes evidenced in the conversion of methanol to hydrocarbons over hierarchically organized MFI-type catalysts of equivalent bulk composition, but synthesized by different approaches. (a–d) Adapted with permission from ref. (a,b, © 2015 American Chemical Society), 19 (c, © 2013 The Royal Society of Chemistry) and 22 (d, © 2014 Macmillan Publishers Ltd).

Figure 3. Mechanism of HOZ synthesis.

(a) Approaches to synthesize HOZs involve either interrupting their crystallization and growth in the synthesis gel (bottom-up) or the post-synthetic removal or rearrangement of framework atoms in a bulk zeolite (top-down). A multilevel mechanistic understanding is vital to effectively control the crystal, active site and pore structure. For top-down methods, (b) a contour plot of the mesopore surface area of demetallated ZSM-5 zeolites versus the base concentration was applied and the starting Si/Al ratio and c, the linear dependence between the average mesopore diameter and the surfactant length in the post-synthetic mesostructuring of zeolite Y, typify the mesoscopic observations attained experimentally. (d) As illustrated by the zero-point energy-corrected energy profiles (E+ZPE), theoretical efforts to improve understanding at the atomic level have so far focused on predicting the likely first steps of demetallation pathways. Regarding direct synthesis methods (e), unparalleled atomic-level insights have been attained by solid state NMR. Here, 2D ²⁹Si {¹H} heteronuclear shift correlation (HECTOR) experiments reveal strong intensity correlations between the head groups of diquarternary ammonium surfactants and the resulting ZSM-5 nanosheets and the nanolayered silicate intermediates. (f) Dissipative particle dynamics simulations of the impact of the shear rate on the self-assembly of an amphiphilic surfactant and silica source in aqueous solution represent one of the first computational contributions to the mesoscopic understanding. (b–f) Adapted with permission from ref. (b, © 2011 The Royal Society of Chemistry), 15 (c, © 2014 John Wiley and Sons Inc.), 33 (d, © 2014 American Chemical Society), 39 (e, © 2014 John Wiley and Sons Inc.) and 43 (f, © 2015 Elsevier).

Figure 4. Crystalline structure of HOZs.

Simulated (-S) and experimental (-E) XRD patterns of (a) a 2D nanosized ZSM-5 zeolite and (b) a USY zeolite with hexagonally ordered mesopores compared with the simulated patterns of their respective bulk forms illustrate the changes that can be expected upon reducing the crystal size in one or more dimensions. In the first case, only the 0kl reflections are distinguished, while in the latter case, low-angle reflections appear, and depending on the relative integration of the micro- and mesopore domains, high-angle reflections arising from the crystalline framework may be significantly altered, making it hard to identify with respect to the bulk phase. (c–e) Electron-based techniques are widely applied to gain complementary information on the crystal structure. (c) TEM images confirm the morphology and crystalline order of a 2D nanosized MFI-type zeolite, revealing the 90° rotational boundaries between the nanosheets (scale bars, 200 nm, top of panel; 20 nm, bottom of panel). (d) Electron crystallography enables the structure determination of a complex intergrown ITQ-39 zeolite (scale bar, 5 nm in TEM image). (e) The reciprocal lattice (right) reconstructed from rotation electron diffraction data confirms the preserved crystalline order of a hierarchically organized FAU zeolite. A 3D reconstruction of the pore topology from TEM tomography data (left) evidences the abundant presence of interconnected mesopores (scale bar 10 nm). (a–e) Adapted with permission from ref. (a, © 2015 Macmillan Publishers Ltd), 51 (b, © 2012 The Royal Society of Chemistry), 42 (c, © 2014 Macmillan Publishers Ltd), 55 (d, © 2012 Macmillan Publishers Ltd) and 56 (e, © 2014 John Wiley and Sons Inc.).

Figure 5. Active site distribution in HOZs.

(a,b) A recent literature survey revealed a prominent reduction in the concentration of Brønsted acid sites (a) and an increase in the concentration of Lewis acid sites (b) quantified by the IR spectroscopy of adsorbed pyridine with the increasing mesopore or external surface area (S_meso) in HOZs independent of the framework type or synthesis route of the zeolite, suggesting a common explanation. (c) Concomitant reductions in the strength of Brønsted acid sites have also been evidenced by the higher temperature required to decompose adsorbed n-propylamine into propene and ammonia with decreasing c_Brønsted in demetallated ZSM-5 zeolites. (d) ²⁷Al multiple quantum NMR spectroscopy clearly distinguishes the underlying changes in the aluminium speciation, revealing that changes in the coordination are only observed upon calcination to the protonic form and not upon mesoporosity introduction by demetallation. (e) The changes in aluminium coordination within a bulk ZSM-5 zeolite upon steaming were spatially resolved with 30 nm resolution within 3D tomographic reconstructions attained by scanning transmission X-ray tomography with fourfold (bulk) or four-/fivefold (steamed) and sixfold coordinate aluminium (coloured blue and red, respectively). (f) Scanning transmission electron microscopy energy-dispersive X-ray spectroscopy maps revealed a 2.5-fold variation in the Si/Al ratio across a zeolite X-nanosheet prepared by repetitive branching (scale bars, 100 nm, top of panel; 200 nm, bottom of panel). (a–f) Adapted from ref. (a,b,d, © 2015 John Wiley and Sons Inc), 18 (c, © 2013 Elsevier), 62 (e, © 2013 John Wiley and Sons Inc.) and 45 (f, © 2014 John Wiley and Sons Inc.).

Figure 6. Toolbox for porosity assessment in HOZs.

(a) The inadequacies of the t-plot method were recently exposed by the underestimation of the micropore volume in a mesostructured USY zeolite. (b) Argon isotherms evidence the hysteresis induced in small mesopores within a mesostructured USY zeolite upon measurement at lower temperatures. (c,d) A unified model connecting the pore size distributions derived from argon sorption and mercury porosimetry in a shaped zeolite catalyst (c) emphasizes the importance of integrative data analysis approaches. A cross-sectional focused ion beam scanning electron tomography image clearly reveals the internal particle organization originating the multimodal porosity (scale bar, 2 μm (d)). (e) The size of the micropore domains in HOZS are quantified by TEM tomography. (f,g) The accessibility of open and constricted mesopores in hierarchical ZSM-5 zeolites can be distinguished from the comparative mesopore volume determined by Hg intrusion and N₂ sorption (f). Identical-location scanning electron and scanning transmission electron micrographs reveal the structural origin (scale bar, 100 nm (g)). (h) The direct correlation and the methanol-to-hydrocarbon (MTH) lifetime of hierarchical ZSM-5 zeolites demonstrates the unique sensitivity of positron annihilation lifetime spectroscopy to the pore connectivity and related function of the auxiliary pore network. As illustrated schematically, the presence of open or constricted mesopores observed by TEM (scale bar, 20 nm) strongly impacts the amount of ortho-positronium escaping from the zeolite. (a-h) Adapted from ref. (a, © 2014 American Chemical Society), 56 (b, © 2014 John Wiley and Sons Inc.), 74 (c,d, © 2015 American Chemical Society), 66 (e, © 2012 John Wiley and Sons Inc.), 22 (f,g, © 2014 Macmillan Publishers Ltd) and 28 (h, © 2015 John Wiley and Sons Inc.).

Figure 7. Enhanced diffusion properties and catalyst effectiveness.

(a) Snapshots determined by dynamic Monte Carlo simulations illustrate the different guest distributions expected within zeolite crystals under the limiting cases of slow and fast exchange between the meso- and micropores. (b) Self-diffusivity of ethane in bulk and mesoporous LTA zeolites determined by pulsed field gradient NMR. Mass transfer strictly adheres to normal diffusion over the space and timescales of the measurements. Even at high temperature, ethane diffusion in the mesoporous zeolite does not significantly exceed that in the bulk counterpart, while at low temperature the similarly reduced values reflect the low amount of molecules in the mesopores. Blockage of the mesopores with cyclohexane (open symbols) is evident from the reduced, but equivalent temperature dependency of the ethane diffusivity compared with the bulk zeolite. (c) Correlation of the effective diffusivity, determined by gravimetry under slow-exchange conditions, of 2,2-dimethylbutane in hierarchical ZSM-5 with the interface area S_meso. The sharp increase evidenced above 150 m² g⁻¹ was ascribed to the attainment of an interconnected mesopore network. (d) Concentration profiles of benzene within MFI-type zeolite crystals are monitored along different crystallographic directions by measuring the change in the refractive index. (e) The catalytic conversion of furfuryl alcohol is mapped within single-needle-shaped ZSM-22 zeolite crystals with nanometre accuracy by fluorescence microscopy. The inset shows the reaction rate as measured along the white line. (b–e) Adapted with permission from ref. (b, © 2012 Elsevier), 89 (c, © 2014 John Wiley and Sons Inc.), 95 (d, © 2014 Macmillan Publishers Ltd) and 103 (e, © 2009 © 2015 John Wiley and Sons Inc.).