Probing the Design Rationale of a High‐Performing Faujasitic Zeotype Engineered to have Hierarchical Porosity and Moderated Acidity

Abstract Porosity and acidity are influential properties in the rational design of solid‐acid catalysts. Probing the physicochemical characteristics of an acidic zeotype framework at the molecular level can provide valuable insights in understanding intrinsic reaction pathways, for affording structure–activity relationships. Herein, we employ a variety of probe‐based techniques (including positron annihilation lifetime spectroscopy (PALS), FTIR and solid‐state NMR spectroscopy) to demonstrate how a hierarchical design strategy for a faujasitic (FAU) zeotype (synthesized for the first time, via a soft‐templating approach, with high phase‐purity) can be used to simultaneously modify the porosity and modulate the acidity for an industrially significant catalytic process (Beckmann rearrangement). Detailed characterization of hierarchically porous (HP) SAPO‐37 reveals enhanced mass‐transport characteristics and moderated acidity, which leads to superior catalytic performance and increased resistance to deactivation by coking, compared to its microporous counterpart, further vindicating the interplay between porosity and moderated acidity.


Catalyst Synthesis
The synthesis of HP SAPO-37 and MP SAPO-37 followed the same procedure, except that the latter did not include the addition of dimethyloctadecyl[ (3-(trimethoxysilyl)propyl] ammonium chloride.
Solution A: Phosphoric acid (9.25 g, 85 wt. % in water, Sigma Aldrich) and deionized water (20 mL) were stirred together in a Teflon beaker, to which pseudo-boehmite (5.58 g, Condea Vista) was added over the course of 1 hour. Solution A was stirred for 7 hours. Solution B: Tetra-n-propylammonium hydroxide (TPAOH, 38.69 g, 40 wt. % in water, Alfa Aesar), tetramethylammonium hydroxide pentahydrate (TMAOH, 0.37 g, Sigma Aldrich) and fumed silica (1.00 g, Sigma Aldrich) were stirred for 2 hours in a glass beaker. Solution B was added dropwise to Solution A and stirred for 68 hours. Dimethyloctadecyl [3-(trimethoxysilyl)propyl]ammonium chloride solution (DMOD, 3.43 mL, 42 wt. % in methanol) was added dropwise to the gel, which was stirred for 2 hours. The gel was transferred to a Teflon-lined, stainless steel reactor and crystallized at 200 °C for 24 hours. The reactor was allowed to cool to ambient temperature before its contents were removed and distributed between 4 x 100 mL centrifuge tubes. The solid in each tube was washed 3 times with deionized water at 10,000 rpm, and then dried overnight in an oven at 80 °C. The solid (~ 32 g) was calcined in batches (~ 2 g) by heating to 550 °C (ramp rate of 2 °C min -1 ) for 16 hours.

Characterisation
Powder X-ray diffraction patterns were acquired using a Bruker D2 diffractometer with Cu Kα1 radiation. Unit cell refinements were performed using the CelRef software. 1 Low-angle X-ray diffraction patterns were obtained using a Rigaku SmartLab diffractometer with Cu rotating anode source.
Nitrogen adsorption measurements were performed at 77K using the Micromeritics Gemini 2375 Surface Area Analyser. Samples were degassed, under vacuum, at 120 °C for 12 hours prior to measurement.
For ICP-OES elemental analysis, samples were subject to acid digestion before analysis in the Varian Vista MPX CCD Simultaneous Axial ICP-OES.
NH3-TPD measurements were performed using the Quantachrome Autosorb iQ-Chemi apparatus. Catalysts were pre-treated at 150 °C under a flow of helium gas for 2 hours. The samples were then dosed with ammonia gas for three hours at 100 °C, before desorption under a flow of helium gas, with a temperature ramp of 10 °C min -1 up to 600 °C.
Carbon, hydrogen, and nitrogen (CHN) elemental analysis was performed using the Thermo Carlo Erba Flash 2000 Elemental Analyser.
Thermogravimetric analysis was performed on the Netzsch TG 209 F1 Libra by heating under air from 30 -900 °C at a ramp rate of 10 °C min -1 .
TEM images were acquired using the Hitachi HT7700 Transmission Electron Microscope with Morada G3 camera.

Positron annihilation lifetime spectroscopy (PALS)
The catalyst powders were packed to 2 mm either side of a Mylar sealed 22 NaCl positron source. The samples were measured at room temperature, under high vacuum (5 x 10 -6 Torr) and placed between two EG&G Ortec spectrometers for a minimum of 4.5 x 10 6 integrated counts. The final spectra were analyzed using LT (Version 9.0) software 2 using a source correction (1.486 ns and 3.593 %). The spectra were best fitted to 5 components; the first component being fixed to 125 ps due to para-positronium decay, and the second at ~400 ps due to free annihilation within the sample. The remaining three lifetimes were attributed to orthopositronium decay, and associated with a tri-modal pore structure within the materials. The zeolites featured intrinsic micro-porosity (τ3) due to the porous zeolite cages and channels, the inter-particle micropores (τ4) and larger mesopores (τ5). 3 The lifetimes were converted to average pore diameters using the Tao-Eldrup equation for τ3 and the Rectangular Tao-Eldrup equation for τ4 and τ5. [4][5][6] The pore-size distribution was calculated using PAScual (Version 1.4) software. 7

MAS NMR characterization of the framework
The calcined catalyst was packed in a 4 mm zirconium oxide rotor before drying overnight in a fan-assisted oven at 100 °C. Whilst still hot, the rotors were sealed by capping with a turbine. All NMR experiments were performed at a sample rotation frequency of 11 kHz, on a 9.4 WB Ascend Bruker magnet with a Neo console, using a triple resonance 4 mm RevolutionNMR probe. The 1 H NMR spectra were referenced to adamantane at 1.8 ppm. 8 The 29 Si NMR spectra were referenced to tetrakis(thrimethylsiloxy)silane at -9.8 and -135.4 ppm. 8 The 31 P spectra were referenced to phosphoric acid (85 % in H2O) at 0 ppm. The 27 Al spectra were referenced to AlCl3 (1 M in H2O) at 0 ppm. The 29 Si cross-polarization experiments were obtained with ramped cross polarisation 9 are the result of 24576 scans, with a contact time of 3 ms and spinal decoupling at 80 kHz during acquisition. The 31 P NMR experiments were recorded with direct excitation using 2 scans and 180 seconds between scans. Proton decoupling using SPINAL64 was applied during acquisition. The 27 Al experiments were recorded with one pulse excitation and no decoupling with 64 scans.

Probe-based MAS NMR with 15 N-pyridine
Calcined catalyst was transferred to a ceramic boat and sealed inside a Schlenk tube. The Schlenk tube was transferred to a furnace to dry under vacuum at 120 °C for 12 hours. After isolation from the vacuum, the Schlenk tube was cooled to room temperature and transferred to a glove back under N2 atmosphere. 7 μL of 15 N-labelled pyridine (Sigma Aldrich) was added to the catalyst (0.07 g) with mechanical mixing. The catalyst-pyridine sample (0.02 g) was packed into a 3.2 mm thin wall pencil rotor. The experiments were performed on a narrow bore 14.1 T Agilent DDR2 spectrometer equipped with a 3.2mm triple resonance Agilent probe, at a spinning speed of 13 kHz. 1 H NMR spectra were referenced to adamantane at 1.8 ppm. 8 15 N spectra were referenced to NH4Cl at 39.3 ppm. The direct excitation was recorded with 300 scans, and a pulse delay of 300 s between scans.
Probe-based FTIR studies FTIR spectra of self-supporting pellets were collected under vacuum conditions (residual pressure <10 -5 mbar) using a Bruker Equinox 55 spectrometer equipped with a pyroelectric detector (DTGS type) with a resolution of 4 cm -1 . NH3, pyridine, and 2,6-di-tertbutylpyridine were each adsorbed at room temperature using specially designed cells, permanently connected to a vacuum line for in situ adsorption-desorption measurements. FTIR spectra were normalized with respect the pellet weight and, whenever specified, are reported in difference-mode by subtracting the spectrum of the sample in vacuum, from the spectrum with adsorbed molecules. The total number of accessible Brønsted acid sites (N) was estimated using the Lambert-Beer law in the form A = εNρ, where A is the integrated area of the bands of the protonated species, ε is the molar extinction coefficient (cm 2 mmol -1 ), N is the concentration of the vibrating species (mmol g -1 ), and ρ is the density of the disk (mass/area ratio of the pellet, mg cm -2 ). The accessibility factor (AF) is defined as the number of Brønsted sites detected by, 2,6-di-tert-butylpyridine adsorption, divided by the total number of Brønsted acid sites detected by ammonia adsorption.

Catalysis
Pelletised catalyst (0.2 g) was sandwiched between glass beads (1 mm diameter) within a cylindrical, quartz, fixed-bed reactor tube (4 mm diameter). The tube was transferred to within the heater unit of a flow-reactor setup, and the catalyst pre-treated by heating at 400 °C for 1 hour under a flow of He gas (50 mL min -1 ). For catalysis, the temperature and He flow were reduced to 300 °C and 33.3 mL min -1 , respectively. A liquid-feed of 100 g L -1 of cyclohexanone oxime in ethanol was supplied to the reactor via electronic syringe pump, to maintain a weight hourly space velocity (WHSV) of 0.79 hr -1 . Simultaneously, an external standard feed of 100 g L -1 of mesitylene in ethanol was introduced into the exit feed at a WHSV of 0.79 h -1 . Once steady-state was established, samples were collected on an hourly basis, and analyzed using the Clarus 480 gas chromatograph with FID detector and Elite-5 column.     Figure S3. TEM images of HP SAPO-37 showing aggregated, rod-like crystallites that contain striations attributed to mesoporosity. Figure S4. The N2 gas adsorption-desorption isotherms at 77 K of MP SAPO-37 and HP SAPO-37, with their respective BJH adsorption pore-size distributions inset.

SI.7 PALS theory
PALS is a non-destructive technique that monitors the lifetime of the ortho-positronium, o-Ps: a metastable system comprising of the bound state of a positron and an electron of the same spin. For textural characterization, PALS exploits the tendency of o-Ps to localize in electron-deficient defects (e.g. such as pores and voids) in an insulating material. The o-Ps will diffuse into a porous network and, where accessible, move into increasingly larger void spaces until annihilation. 10 If the o-Ps is not destroyed by pickoff annihilation with an electron from the surrounding matrix, it annihilates with an intrinsic vacuum lifetime of 142 ns ( Figure S5). Figure S6. Positron (e + ) formation and annihilation in a porous material. Positrons travelling in an insulating material may capture an electron to form the o-Ps species. o-Ps will tend to localize in the pores, travelling into increasingly larger voids (where accessible).
The o-Ps may undergo pickoff annihilation with an electron (e -) from the surrounding matrix, or escape to the surroundings and annihilate in vacuum. In either case, annihilation occurs with the production of gamma photons (γ).
Therefore, any structural feature that increases the probability of pickoff annihilation (e.g. closed or narrow pores) will increase the probability of o-Ps annihilating with a lifetime < 142 ns. By applying suitable models, [4][5][6]11 it is possible to relate the o-Ps lifetime to the size of the void in which it was annihilated. Significantly, the number of o-Ps annihilating at a particular lifetime is directly related to the contribution of each annihilation site; hence, the PALS intensity reflects the relative quantity and accessibility of the pores in a sample.
The o-Ps is sufficiently long-lived to quantify pore dimensions in the range 0.3 -30 nm, 12 and the resolution of the technique is sufficient to discriminate different pore architectures within same pore-size regime. 13 Where PALS is able to distinguish and quantify porosity over multiple length scales, it has proven well suited to the study of hierarchical porosity. 3,10,12,[14][15][16][17][18] In particular, recent studies have sought to exploit the unique capabilities of the PALS technique to quantify the interconnectivity of micro-and mesopores in hierarchical zeolites. 10,12,[14][15][16][17] Ultimately, the o-Ps lifetime will tend towards the self-annihilation lifetime if the o-Ps can move into increasingly larger void spaces. Therefore, in a hierarchical system, the proportion of o-Ps annihilating in the micro-and mesopores will depend on their relative pore volume and interconnectivity. 17   Table S6. Assignment of the aromatic ring vibrations between 1400 -1700 cm -1 in the FTIR spectra ( Figure 10) of pyridine adsorbed on MP and HP SAPO-37.