Intra-crystalline mesoporous zeolite encapsulation-derived thermally robust metal nanocatalyst in deep oxidation of light alkanes

Zeolite-confined metal nanoparticles (NPs) have attracted much attention owing to their superior sintering resistance and broad applications for thermal and environmental catalytic reactions. However, the pore size of the conventional zeolites is usually below 2 nm, and reactants are easily blocked to access the active sites. Herein, a facile in situ mesoporogen-free strategy is developed to design and synthesize palladium (Pd) NPs enveloped in a single-crystalline zeolite (silicalite-1, S-1) with intra-mesopores (termed Pd@IM-S-1). Pd@IM-S-1 exhibited remarkable light alkanes deep oxidation performances, and it should be attributed to the confinement and guarding effect of the zeolite shell and the improvement in mass-transfer efficiency and active metal sites accessibility. The Pd−PdO interfaces as a new active site can provide active oxygen species to the first C−H cleavage of light alkanes. This work exemplifies a promising strategy to design other high-performance intra-crystalline mesoporous zeolite-confined metal/metal oxide catalysts for high-temperature industrial thermal catalysis.


Catalysts preparation
Synthesis of Pd@SiO2. Pd@SiO2 was synthesized by using a reversed-phase microemulsion method according to our previous report 1,2 with some modifications. Briefly, a calculated amount of 4.32 mL Pd(NH3)4(NO3)2 solution was rapidly added into 960 mL of cyclohexane solution containing NP-5 (polyethylene glycol mono-4-nonylphenyl ether, 40.32 g). After stirring at 25 o C for 15 h, an aqueous ammonia solution (28 wt.%, 4.32 mL) was then added quickly. After stirring for 2 h, tetraethyl orthosilicate (TEOS, 5 mL) was added into the above mixture solution. Next, the solution was stirred for 48 h at 25 o C. The initial sample was collected by centrifugation at 9000 rpm. After drying in a vacuum oven at 40 °C overnight, and then calcining at 550 °C for 4 h in air, the final sample of Pd@SiO2 was obtained.
Synthesis of Pd@IM-S-1. The Pd@IM-S-1 was synthesized by in situ conversion of the SiO2 shell in Pd@SiO2 to silicalite-1 through a continuous heating in situ dry-gel conversion method according to previously reported literature 3,4 with some modifications. For typical synthesis of Pd@IM-S-1, 0.4 g of Pd@SiO2 and 0.5332 g of tetrapropylammonium hydroxide (TPAOH, 25 wt.%) was mixing and then subject to grinding at room temperature for 15 min. The solid was dried under infrared light for 15 min to partly remove the water, and then the solid powder was transferred into a Teflon-lined autoclave with limited amount of water. The material was thermally treated in a conventional oven at 80 o C for 2 days and then at 120 o C for 1 day under static conditions. Finally, the material was calcined at 550 o C for 4 h to obtain the Pd@IM-S-1 sample.

Synthesis of S-1.
Typically, 20 mL TEOS was firstly mixed with 16 mL TPAOH (25 wt.%) and then 44 mL deionized water was added into the above mixture solution at room temperature with fast stirring until hydrolyzed completely, followed transferred into a Teflon-lined autoclave and conducted in a conventional oven at 170 o C for 1 day. The initial sample was collected by centrifugation at 9000 rpm and washed with deionized water for several times. After drying in a conventional oven at 80 °C overnight and calcining at 550 °C for 6 h in air, the final sample of S-1 was obtained.
Synthesis of Pd/S-1. For a comparison study, Pd/S-1 was prepared by Pd nanoparticles supported on the outer surface of S-1 (The S-1 was not calcined.). As a typical run for the synthesis of Pd/S-1 (The content of Pd is consistent with Pd@IM-S-1 sample.) were prepared via wet impregnation method. After impregnation, the samples were dried in a conventional oven at 80 °C overnight and calcined at 550 °C for 4 h in air, the final sample of Pd/S-1 was obtained.

Catalyst characterization
X-ray diffraction (XRD) patterns were recorded on a Bruker AXS D8Focus diffractometer that operated at 40 kV and 30 mA with a Cu target and Kα-ray irradiation (λ = 1.54178 Å). Scans were collected in the 2θ range from 5 to 65 ° with a step of 2 ° min -1 to analyze the phase structure. Raman spectra were collected in the anti-Stokes range of 200-1500 cm -1 using an inVia Reflex-Renishaw spectrometer. The Pd and Ce content of catalysts was determined by inductively coupled plasma optical emission spectrometry (ICP-OES) were performed with an Agilent 5100 ICP OES with Dichroic Spectral Combiner (DSC) technology (Agilent Technologies, Mulgrave, Australia). Scanning electron microscope (SEM), transmission electron microscope (TEM), scanning TEM (STEM), high-resolution TEM (HRTEM), line scan and elemental mapping images were obtained using a JEOL 3000F TEM working at 300 kV equipped with an energy dispersive spectroscopy (EDS) detector. The Tomogram-section TEM images were obtained using a Leica EM UC7 working at FEI Talos F200X 200kV. Samples for TEM investigation were prepared by dispersing samples in ethanol and immersing them in an ultrasonic bath for 30 min, and then dropping a few drops of the resulting suspensions onto a copper grid coated with a layer of amorphous carbon. Nitrogen adsorption-desorption analysis was carried out at 77 K on Micromeritics ASAP2020 instrument. The specific surface areas of the samples were calculated using Brunauer-Emmett-Teller (BET) method. The total pore volume of each catalyst was accumulated at a relative pressure of P/P0 = 0.99. X-ray photoelectron spectrum (XPS) was performed on PHI-5000C ESCA system using a single Mg-K-X-ray source operated at 250 W and 14 kV voltage. The spectra were obtained at ambient temperature with an ultra-high vacuum. The binding energies were calibrated by the standard C1s peak of graphite at 284.8 eV.
To confirm Pd@IM-S-1 has possess abundant intra-mesopores and Pd NPs are confined within the mesopores of the zeolite shell. The focused ion beam (FIB) technique was performed to preparing specimens for aberration-corrected TEM characterization. The FIB specimens were obtained using the FEI Helios Nanolab 600i from ZKKF (BEIJING) Science & Technology Co., Ltd. and working at 500 V-30 kV. Detailed steps are as follows: (1) Sample pretreatment: Attached non-conductive powder sample Pd@IM-S-1 to a carbon film.
(2) Selecting area and depositing Pt protective layer: In the electron beam imaging mode of the double beam electron microscope, select the area of interest, and use the electron beam to deposit a layer of C or Pt with a thickness of about 0.5-1 microns, and then use the ion beam to deposit a layer of Pt with a thickness of 2-3 microns to protect the surface structure of the sample.
(3) Proposal and transfer: Gradually dig out V-shaped grooves on both sides and cut off the root of the sample, extend the manipulator and weld the top of the sample, then cut the supporting parts on both sides of the sheet, take out the sample and transfer it to the support and weld it firmly. Finally, the sample was cut and separated from the tip of the manipulator.
(4) Thinning: The FIB specimens were obtained through the ion beam to thinning the thin sample layer. The actual thinning parameters, including the acceleration voltage required to change the thickness and the tilt angle of the sample during the thinning process, mainly depend on the material system being studied. For general materials, the following parameters: use 30 KV to process the sample to ~1000 nm, then reduce the voltage to 16 KV, continue to thin to ~500 nm, and then reduce the voltage to 8 KV, continue to thin to ~200nm, and finally use 5 KV to thin to ~100 nm. If the sample needs to be characterized with high resolution, it is also necessary to reduce the voltage to 2 KV and gradually reduce the thickness of the sample to ~50 nm. After thinning to the target thickness, the sample is cleaned with a low voltage of 1 KV to remove the amorphous layer introduced by the ion beam.
The simple flowchart was presented in Figure S8. Hereafter, the aberration-corrected TEM (AC-TEM) was applied to characterize FIB specimens via the FEI titan themis working at 300 kV equipped with an energy dispersive spectroscopy (EDS) detector.
The X-ray absorption spectra were collected on the beamline BL01C1 in National Synchrotron Radiation Research Center (NSRRC), with electron energy of 1.5 GeV and a beam current between 100 and 200 mA, and were provided technical support by Ceshigo Research Service "www.ceshigo.com". The radiation was monochromatized by a Si (111) double-crystal monochromator. The XAFS data were recorded under fluorescence mode with standard Lytle ion chambers for the Pd K-edge. Typically, the energy was calibrated according to the absorption edge of a pure Pd foil as appropriate. The Athena software was used to extract and fit the XANES and EXAFS data. The passive electron factors, S0 2 , were determined by fitting the experimental Pd foil data and fixing the Pd-Pd coordination number (CN) to be 12, and then fixed for further analysis of the measured samples. In addition, the parameters including CN, bond distance (R), Debye-Waller factor (σ 2 ), and the inner potential correction (ΔE0) were describing the electronic properties and local structure environment.
The in situ near ambient pressure X-ray photoelectron spectroscopy (In situ NAP-XPS) measurements were carried out on a SPECS NAP-XPS system. All spectra were obtained using monochromatized Al Kα irradiation (1,486.6 eV) generated by an Al anode (SPECS XR-50) and an excitation source power fixed at 50 W. XPS measurements at pressures up to ~25 mbar are possible owing to a differential pumping system, which separates the electron analyser (SPECS Phoibos NAP-150) from the reaction area. The aperture of the nozzle is 0.3 mm. Through a pressure reducing valve kept the reaction pressure was at 1 mbar (C3H8:O2 = 1:5).

Investigation on growth mechanism of Pd@IM-S-1
To understand the growth mechanism of Pd@IM-S-1, the crystallinity and morphological evolution with different crystallization time between 80 to 120 o C has been monitored. First, the Pd@IM-S-1-80-1d, Pd@IM-S-1-80-2d, Pd@IM-S-1-120-1h, Pd@IM-S-1-120-2h, Pd@IM-S-1-120-4h, Pd@IM-S-1-120-12h and Pd@IM-S-1-120-24h samples are obtained by in situ dry-gel conversion method. These samples are thermally treated in a conventional oven as following 80 o C for 1 day→80 o C for 2 days→120 o C for 1 hour→120 o C for 2 hours→120 o C for 4 hours→120 o C for 12 hours→120 o C for 24 hours under static conditions. And then, the XRD and TEM are applied to characterization the crystallinity and morphological evolution of these samples. During this test, all samples are unreduced.

In situ DRIFTS measuring processes
Propane adsorption and oxidation at 325 o C. In situ diffuse reflectance infrared Fourier transform spectroscopy (In situ DRIFT) measurements were performed of Pd@IM-S-1 sample on a Bruker Vertex FT-IR spectrometer with a mercury-cadmium-telluride (MCT) detector. In the DRIFT cell with KBr windows connected with a gas flow system, the sample was pretreated at 400 o C in N2 for 30 min and then cooled to 325 o C in N2.
After the background spectra were recorded at the temperature, N2 gas was replaced by the gas of 0.2% C3H8-99.8% N2 (30 mL min -1 ), and in situ DRIFT spectra of the samples were taken at different times. After 30 min, the gas of 0.2% C3H8-99.8% N2 was replaced by the gas of 10% O2-90% N2 (30 mL min -1 ) flow for 30 min.
Propane temperature-programmed desorption in N2 + O2 atmosphere. The TPD experiments were performed on the same Bruker Vertex FT-IR spectrometer. The sample powders of Pd@IM-S-1 were first purged in situ in a N2 stream (30 mL min −1 ) at 400 o C for 30 min and then cooled to 50 o C in N2. After the background spectra were recorded at 50 o C, N2 gas was replaced by the gas of 0.2% C3H8-99.8% N2 was fed at a flow rate of 30 mL min −1 for 30 min and pure N2 was introduced to remove any physical adsorption molecules. The spectra were recorded from 50 to 400 o C in 10% O2-90% N2 atmosphere at a heating rate of 10 o C min -1 .
Propane temperature-programmed oxidation. The IR experiment of the reaction pathway was also recorded on the Bruker Vertex FT-IR spectrometer. The sample powders of Pd@IM-S-1 were purged in situ in a N2 stream (30 mL min −1 ) at 400 o C for 30 min and then cooled to 50 o C in N2. After the background spectra were recorded at 50 o C, N2 gas was replaced by the gas of 0.2% C3H8-10% O2-89.8% N2 was fed at a flow rate of 30 mL min −1 . The spectra were recorded from 50 to 400 o C at a heating rate of 10 o C min -1 .

Density functional theory calculations
The periodic density function theory (DFT) calculations were carried out using the CP2K package. 5,6 The exchange-correlation potential was described using the Perdew-Burke-Ernzerh (PBE) form of the generalized-gradient approximation (GGA). 7 The wave functions were expanded in a molecularly optimized double-Gaussian basis set, with an auxiliary plane wave basis set with a cutoff energy of 500 Rydberg. The scalar relativistic norm-conserving pseudo-potentials were employed to modelled the core electrons with 18, 6, 4 and 1 valence electrons for Pd, O, C and H, respectively. 8 The only Γ-point in reciprocal space mesh was used for integrating the Brillouin zone. Grimme's third generation DFT-D3 approach was used to describe dispersion corrections. 9 The Pd (111) and PdO (101) surfaces were used to model the Pd and PdO substrates, constructed with cell dimensions of 19.0847×19.0811×29.4223 Å and 12.5225×22.3838×25.7640 Å respectively with 15 Å vacuum space to minimize the interaction between slabs. For the Pd-PdO interface model, we firstly built Pd (111) slab, then generated one layer of PdO (111) on the surface of Pd (111) and finally formed the interface structure shown in Figure S23. For the O2 dissociation on Pd (111), we have also conducted corresponding calculations shown in Figure S24. The convergence criterion used for geometry optimizations was a maximum force of 0.01 eV Å -1 . Spin polarization was considered in all calculations.

Activity test
Methane (CH4) and propane (C3H8) total oxidation. Methane and propane as the typical light alkanes were selected as the model compounds to evaluate the catalytic activity of Pd@IM-S-1 and related catalysts. First, 50 mg of catalyst (40-60 mesh) was put in a quartz tubular fixed-bed reactor and in situ reduction at 400 o C for 2 h under 10% H2/Ar (30 mL min -1 ). Then, the temperature was cooled down to 30 o C and purged for 30 min under ultra-high purity N2. Subsequently, the activity was test from 200-400 o C at a rate of temperature change of 2 °C min −1 under 1% CH4-21% O2-N2 or 0.2% C3H8-10% O2-N2. The total gas flow rate was 30 mL min -1 and the weight hourly space velocity (WHSV) was fixed at 36 000 mL gcat. -1 h -1 . Finally, the reactants and products were monitored through a Fourier transform infrared (FTIR) spectrometer (ANTARIS IGS-Analyzer) with a 2 m gas cell. The CH4 and C3H8 conversion were calculated according to the following equation:

Reaction kinetics measurement
The kinetics data for CH4 and C3H8 combustion of Pd@IM-S-1, Pd/S-1 and Pd@SiO2 samples were tested in a fixed-bed reactor (101.325 kPa) according to previous reported literature. 10 Typically, the feed gas were consisted of 1% CH4-21% O2-N2 and 0.2% C3H8-10% O2-N2 with a WHSV of 180 000 mL gcat. -1 h -1 and 20 mg of catalyst mixed with 80 mg of inert quartz sand (40-60 mesh) was used for each testing while the internal and external diffusion have been eliminated. The methane and propane conversions were controlled within 15%.
The r CH 4 and r C 3 H 8 (mol (gcat·s) -1 ) expressed the reaction rate were calculated by the equation of r CH 4 = X CH 4 • V CH 4 /g cat and r C 3 H 8 = X C 3 H 8 • V C 3 H 8 /g cat , where X CH 4 and X C 3 H 8 are the conversion, V CH 4 and V C 3 H 8 are the gas flow rate (mol s -1 ) of CH 4 and C 3 H 8 , g cat is the weight of the catalyst.
The total turnover frequency (TOF) (s -1 ) of CH4 and C3H8 were calculated by the equation of TOF total = X CH 4 • V CH 4 • N A /N total and TOF total = X C 3 H 8 • V C 3 H 8 • N A /N total , where N A is Avogadro constant and N total is total atom numbers of Pd, which can be calculated by the equation of N total = (m pd /M pd ) • N A , where m pd is the mass and M pd is relative atomic mass of Pd. Absence of mass transport during kinetic measurements.
The absence of mass transport resistances was checked by Weisz-Prater criterion (CWP) for internal diffusion and Mears' criterion (CM) for external diffusion.

Weisz-Prater Criterion for Internal Diffusion
If = 2 < 1, then internal mass transfer effects can be neglected.
Where robs = observed reaction rate, mol/kgcat·s Rp = catalyst particle radius, m ρc = density of catalyst, kg/m 3 Deff = effective diffusivity, m 2 /s Cs = gas concentration of A at the external surface of the catalyst, mol/m 3 The results of internal mass transfer are presented in Table S6 and Table S7.

If
= < 0.15, then external mass transfer effects can be neglected.
Where robs= observed reaction rate, mol/kgcat·s n= reaction order Rp= catalyst particle radius, m ρc = density of catalyst, kg/m 3 ρb = bulk density of catalyst bed, kg/m 3 = (1-Φ) ρc (Φ=porosity) ≈ ρc ≈ ρcat Cs = gas concentration of A at the external surface of the catalyst, mol/m 3 . CAb = bulk gas concentration of A, mol/m 3 . ≈Cs kc = external mass transfer coefficient, m/s The results of external mass transfer are presented in Table S8 and Table S9.

Thermal stability test, water resistance and recycling test
Thermal stability test. The Pd@IM-S-1 and Pd/S-1 catalysts were subjected to thermal stability testing to evaluate the sintering resistance performance of the catalysts. Firstly, these two catalysts were calcined at 800 °C for 8 h in argon before testing (noted as Pd@IM-S-1-800 and Pd/S-1-800). These two catalysts were then

Supplementary Tables
Supplementary Table 1

Supplementary Note 4:
To further study the C3H8 deep oxidation process over Pd@IM-S-1, C3H8-TPO was tested, and the DRIFTS spectra were collected at 50-400 °C via propane and O2 co-adsorption. As presented in Supplementary Fig. 38a and b, the band intensity around bicarbonate (1647 cm -1 ) gradually increases and then decreases. But the band intensity around the aliphatic ester (1750 cm -1 ), acetone (1688 cm -1 ), and formate (1597 cm -1 ) always increases; and that of the acetate (1516 cm -1 , and 1462 cm -1 ) gradually decreases. Significantly, as the temperature increases, the band intensity around gaseous C3H8 (2800-3000 cm -1 ) gradually increases and then decreases, as shown in Supplementary Fig. 38c and d. These results demonstrate that Pd@IM-S-1 has the trapping and concentrating ability of C3H8 at low temperatures and activates it at high temperatures. These results are similar to those shown in Supplementary Fig. 37a and b; and they further confirm that acetate and bicarbonate species should be considered as the active intermediates and the aliphatic ester, acetone, and formate should be considered as the inert intermediates.