Direct imaging and location of Pb2+ and K+ in EMT framework-type zeolite

Direct Imaging and Location of Pb2+ and K+in EMT Framework-Type Zeolite Yaping Zhang, Daniel Smith, Jennifer E. Readman,*and Alvaro Mayoral* Cite This: J. Phys. Chem. C 2021, 125, 6461−6470 Read Online ACCESS Metrics & More Article Recommendations * sıSupporting Information ABSTRACT: The understanding of the structural framework and metal locations within zeolites are two critical aspects to tune and further exploit their properties as heterogeneous catalysts, advanced optical devices, or as water remediation materials. The development of electron microscopy has made it possible to observe directly the zeolitic framework and the distribution of extra framework cations within the pores at the atomic level. Here, we have studied the EMT framework providing data with an unprecedented spatial resolution, which have allowed the analysis of the FAU domains present in the framework. Additionally, potassium and lead (introduced by aqueous ion exchange), which were nonperiodically distributed in the pores have been located. An alternative image mode, annular bright field, has been used, proving its usefulness to extend the spatial resolution and increase the sensitivity toward light elements such as bridging oxygen. Finally, the combination of atomic imaging (local information) with the three-dimensional electron diffraction tomography analysis (averaged information) determined the lead-EMT crystal symmetry to be P63mc. 1. INTRODUCTION Zeolites are microporous materials with tunable porosity widely used as ion exchangers or heterogeneous catalysts 1 among many other applications. They are able to confine elements within their pores and also exhibit very good thermal stability. 2−4 Due to the confinement capability that zeolites exhibit, they can protect extra framework species, minimizing their agglomeration, and release back to the environment as well as tuning their physical properties (for example, work function or luminescence). 5−18 There are currently up to 253 different zeolite topologies 19 in the International Zeolite Association (IZA) structure database; among them, TSC,EMT, and FAU exhibit the lowest framework densities of 13.2−13.3 T/1000 Å3of the aluminosilicate family. Lower density zeolites are highly desirable for soil treatment applications 20 or catalysis. 21 EMT was first synthesized in 1971 22 as a mixture of FAU and EMT. The pure EMT framework, named EMC-2 after Elf (or Ecole Supe rieure) Mulhouse Chimie-two, was obtained by Delprato and co-workers 23 in 1990 and the structure was completely solved by Baerlocher et al. 4 years later. 24 EMT is the hexagonal polytype of the FAU-type zeolite (cubic) sharing the same layer units composed of sodalite cages (t-toc) linked by double six-membered rings (d6R, t-hpr). The EMT framework contains large t-wof and t-wou cages with three and five twelvemembered ring (12MR) windows, respectively. In contrast with the zigzag channels of FAU and the eight-membered ring windows of TSC,EMT has straight access along three directions, all with 12MR windows. EMT has previously been used in biomedicine 25 and gas capture and separation. 26 It also shows great potential for industrial applications such as petroleum cracking due to its large channel and cavity size. Besides the use of zeolites as heterogeneous catalysts and water softeners, the capability of zeolites to capture different elements within their pore system is also attractive for various applications such as optoelectronics or water remediation. The interest of occluding lead within the pores of zeolites or metal−organic frameworks (MOFs) has regained much interest due to the possibility of tuning the optical response of lead-loaded zeolites while preventing lead discharge into the environment due to its coordination to the framework. The Received: January 21, 2021 Revised: February 21, 2021 Published: March 15, 2021 Articlepubs.acs.org/JPCC © 2021 American Chemical Society 6461 https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 Downloaded via CSIC on March 30, 2022 at 07:48:21 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. incorporation of lead species can be summarized in three types: (i) unitary: clusters (including single atom) with only one element; 14,27 (ii) binary: species formed by two elements, such as PbS and PbI2; 28−31 and (iii) ternary: lead compounds formed by three elements, CsPbI3and CsPbBr3. 32 To gain structural information, there are several characterization methods available to help explain the physicochemical properties of zeolites. Such properties are strongly affected by the type of metal introduced, its size, and location; therefore, characterization at the atomic level is becoming indispensable to tailor the properties of materials at the nanoscale. Zeolite frameworks present high degree of symmetry, whereas the exchanged forms are usually disordered. Therefore, long-range techniques such as X-ray diffraction and Rietveld analysis do not always give the short-range information needed. In this sense, transmission electron microscopy is the most powerful methodology as it can provide direct visualization of the framework and the guest species at an atomic level together with crystallographic and chemical information. 33−37 All of these results have been achieved with thorough control of the electron dose, as zeolites are strongly beam sensitive. Working in the probe mode offers a quick scan, reducing the sample damage. The possibility of obtaining sub-Angstrom probes together with the use of several detectors simultaneously allows for the collection of signals at a high angle (annular dark field, ADF), providing information regarding the heavy elements. Meanwhile, an annular bright-field detector (ABF) enhances the signal of light compounds, 38,39 which are scattered at low angle. Therefore, by combining the information from both detectors, determining the location and element type of both light and heavy elements is achievable. In this work, EMT framework-type zeolite was ion exchanged with Pb(NO3)2to finally obtain a partially loaded Pb-EMT. The characterization has been primarily carried out by atomic-resolution transmission electron microscopy imaging that has been acquired by combining ADF and ABF detectors to obtain atomically resolved data of the framework and on the extra framework cations. These observations have been correlated with the data obtained by three dimensional electron diffraction tomography (3D-EDT). The existence of FAU intergrowths has also been analyzed based on this technique. 2. METHODS 2.1. Synthetic Procedures. 2.1.1. Pb-EMT. EMT was synthesized using a previously reported method. 23,40 Sodium aluminate (1.45 g) and a 50% NaOH solution (1.21 g) was added to 7.8 g of deionized water, together with 1.76 g of 18crown-6 and stirred vigorously. Next, 15.4 g of a homogeneous silica solution (LUDOX AM-30) was added with stirring and incubated for 24 hours at room temperature to incubate. The mixture was placed in a 45 mL Parr pressure vessel and heated for 12 days at 110 °C. The product was filtered and washed with copious amounts of deionized water and then calcined in air at 540 °C for 16 h. The potassium-ion exchange was performed as follows: 4 g of EMT was added to 500 mL of a 0.1 M KNO3solution and stirred overnight. The powder was recovered and washed with 1 L of distilled water and air-dried overnight. The exchange procedure was repeated further two times. The lead-ion exchange was carried out as follows: 2 g of EMT was added to 500 mL of a 0.1 M Pb(NO3)2solution and stirred overnight. The powder was recovered and washed with 1 L of distilled water and air-dried overnight. The exchange and washing procedures were repeated further two times. 2.2. Characterization. 2.2.1. Scanning Electron Microscopy. STEM analyses were carried out on a JEOL Grand ARM 300, equipped with a cold field-emission gun (cold FEG) and operated at 300 kV. The column was fitted with a JEOL double spherical aberration corrector, which was aligned prior to the analyses to assure a maximum spatial resolution of 0.7 Å in the scanning mode. The microscope was also equipped with a JEOL EDS and a Gatan Quantum Energy Filter for spectroscopic measurements. 2.2.2. 3D-EDT. The 3D-EDT data were collected using JEOL.Shell software in a JEOL JEM2100 Plus electron microscope operated at 200 kV in the transmission electron microscope (TEM) mode. A total collecting angle of 120°was obtained by tilting the sample holder from −60 to +60°. The intensities were extracted with EDT.Process software. Finally, the structure was solved with Sir2014. 41 2.2.3. Energy-Dispersive X-ray Spectroscopy (EDS). EDS data were collected using a JEOL JEM2100 Plus operated in the STEM mode with an accelerating voltage of 200 kV. 2.2.4. Scanning Electron Microscopy (SEM). The scanning electron microscopy (SEM) images were collected on a JEOL Figure 1. (a) Low-magnification SEM image of several EMT particles. (b) SEM image of an isolated EMT crystal with a hexagonal plate morphology. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6462 JEM7800 Prime with a work distance of 7 mm under a landing voltage of 1 kV. 2.2.5. STEM Simulation. The simulated ADF-STEM and ABF-STEM images along [100] were obtained with software QSTEM. 42 A supercell of 150 ×150 ×150 Å3was constructed for pure silica EMT. The probe array was 600 ×600 pixels with 0.032 Å resolution and a slice thickness of 1 Å. The detector geometry inner and outer angles were 50−200 mrad for ADF and 5−40 mrad for ABF. 2.2.6. Image Treatment. To minimize the beam damage, the electron dose was significantly reduced in comparison with standard STEM conditions. For the current experiments, the total time for image acquisition ranged from 6 to 10 s, 1024 × 1024 pixels using an electron dose of 1120−2800 e−/Å2. Under these conditions, images could be directly interpretated and all intensity analyses were performed over raw data. However, for a clearer visualization of the frameworks and of the light cations, the Random Noise 2D difference filter was used 43 from HREM Research Inc. 3. RESULTS AND DISCUSSION The EMT framework type is a zeolite belonging to the hexagonal crystal family and P63/mmc space group. The lattice constants previously reported are a=b= 17.215 Å, c= 28.082 Å, α=β=90°, and γ= 120°. The morphology of the obtained EMT particles was in agreement with the hexagonal symmetry presenting well-defined hexagonal particles with a thickness up to 500 nm and widths that reached up to several microns, see Figure 1. This framework is formed by double six rings (d6Rs) and sodalite (sod) cages that when connected, form faujasite layers (layer unit), with 12-membered rings formed along the band c directions. The stacking sequence of these layers allows the formation of either EMT or FAU.ByAA′stacking of faujasite sheets, EMT is obtained, where the adjacent two layers are related to each other through a mirror plane. Alternatively, in FAU, the stacking sequence is ABC, where the connectivity exhibits inversion symmetry between successive layers. For EMT, along the b-axis, the 12MRs run parallel to the observation axis while the layer unit runs perpendicular along the c-axis in an AA′stacking. The chemical formula for calcined EMC-2 has been reported to be [Na20(H2O)6]- [Si76Al20O192]-EMT, where the 20 Na+cations that compensate the presence of Al tetrahedra sit in three crystallographic sites at X2: 0.5890, 0.1770, and 0.0470; X3: 0.6667, 0.3333, and 0.6270; and X4: 0.3571, 0.1790, and 0.3880 with 0.25, 0.32, and 0.33 site occupancy, respectively; see Figure S1.InFigure S1, the T (T = Si and Al) atoms are presented as blue, brown, purple, and green spheres corresponding to the different crystallographic sites, O in red, while X2appears in yellow, X3 in light blue, and X4in pink, all of them associated to the 6SRs of the sod cages. Figure 2. (a) EMT model along the [110]/[100]EMT orientation. (b) FAU model along [110]. (c) Schematic model of a layer unit (faujasite layer), front view. (d) Top view of the model corresponding to a layer unit. The origins for the subsequent layers are marked by yellow, blue, and green dots. Color code: rose cages correspond to t-toc (sodalite cages); t-hpr (double six-membered ring, D6R) appear in green (between layers) and in purple (within per layer) cages. (e) Cs-corrected STEM-ADF micrograph recorded along [110]/[100]EMT with an FAU intergrowth marked by yellow lines. The Fourier diffractogram (FD) is shown in the inset with yellow arrows pointing at the diffuse streaks. (f) Closer observation of another crystal where two FAU domains are jointed by a mirror plane BB′. The dashed blue circles correspond to the 12MRs, while the yellow ones denote the partially blocked 12MRs within the twin plane. (g) Schematic representation of the structure model of (f) along the [001]EMT direction. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6463 3.1. Analysis of FAU Intergrowths. It has been previously mentioned that due to the structural similarities (the use of the same building units differing only on how the layers stack), FAU and EMT can coexist in the same crystals. 23,37,44 Along the [110] orientation, both hexagonal EMT and cubic FAU exhibit different ways of pillaring the faujasite layers, where for EMT, there is an AA′AA′... stacking (Figure 2a) and for FAU (Figure 2b), it follows ABC... pillaring. These layers are formed by sodalite cages (t-toc), in pink, linked by d6Rs (t-hpr), in purple, which are subsequently also connected by d6Rs (t-hpr) shown in green between different layers. An isolated faujasite layer unit is depicted in Figure 2c. Because of these structural similarities, it could be expected that intergrowths of both phases due to the layer stacking sequence in a defective way would exist. In fact, there are many cases where EMT has been crystallized in the presence of FAU. 45−47 For EMT, A and A′layers are mirror related to each other. 48 While in FAU, layer B is shifted (1/2a, 1/2b, 1/2c) relative to layer A, and layer C is shifted (1/2a, 0, 1/2c) relative to layer A. The origin of the A, B, and C layers is marked in Figure 2d. Large t-wou and t-wof cages for EMT and t-fau cages for FAU are formed between layers. t-wof, t-wou, and t-fau have three, five, and four 12MRs windows, respectively. Figure 2e corresponds to the Cs-corrected STEM ADF image of an EMT crystal with structural defects in the framework evidenced by the diffuse spots (yellow arrows) in the Fourier diffractogram (see inset). This region, of approximately 6 nm along the [110]/[100]EMT zone axis, corresponds to a FAU domain enclosed by EMT regions. In another crystal, the sequence AA′ABB′A′AA′can be observed, as shown in Figure 2f. This region is formed by two FAU domains connected through a mirror plane (defined by BB′ layers) enclosed within the EMT. For direct interpretation, the schematic model has been superimposed displayed in different colors to distinguish the three zones, as shown in Figure 2f. The left (green) and right (yellow/red) regions, corresponding to the EMT framework, allow the clear visualization of the 12MRs, marked by blue dashed circles. While in between, the twin plane is marked in pink color. By observing this structure along the [001] projection, as shown in Figure 2g, these two EMT domains would overlap, where the 12MRs marked with a blue background would be blocked by sodalite cages from the twin plane, BB′, as shown in Figure 2f. The large 12MR windows of the BB′stacking would also be blocked by the EMT sodalite cages from both sides, as shown in Figure 2g. By observing Figure 2f (yellow circles), the existence of a faint signal inside the 12MRs in the twin plane that is associated with the presence of sodalite cages from the EMT region can be appreciated, partially blocking these channels. These 12MRs are fully empty in the EMT area (blue dashed circles). 3.2. Analysis of the EMT Framework and Extra Framework Cations. It is possible to synthesize the EMT framework with different Si/Al ratios, but in general terms, the most common zeolite obtained has been EMC-2 with the chemical formula |Na20(H2O)6|[Si76Al20O192]. 23,24 In this work, energy-dispersive X-ray spectroscopy (EDS) analyses were used to estimate the chemical composition of the current EMT (Table S1). The results obtained were reasonable and in agreement with the data reported. The most relevant differences with previous studies were that K+was also used as a counter ion together with Na+, and that Pb2+ was also detected due to the ion exchange. Based on these results, the electroneutrality would be maintained as the sum of the positive charges from the cationic species would be 18 that would compensate 20 (AlO4 −) of the framework. The small mismatch between the cations and aluminum is related to a slight experimental error of the technique or to the presence of protonic species H+from water. In fact, the excess of O detected could be associated with water adsorbed in the pores. For simplicity, hereafter, we will refer only K+when considering both Na+and K+, as due to their low content and a similar atomic number, it is not possible to distinguish between both metals. Figure 3. Cs-corrected STEM-ABF data of Pb-EMT. (a) Atomic-resolution data along the [100] projection with the FD shown in the inset. The white arrow corresponds to a structural defect. The yellow circle in the FD indicates the maximum transfer information corresponding to the 00022 spot. (b) Symmetry p2gm averaged data with the framework model superimposed. The red arrows point at O bridges, the blue one corresponds to the T atoms, and the yellow and pink ones to extra framework signals. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6464 Annular dark-field (ADF) together with annular bright-field (ABF) imaging modes can reveal unique information, especially at a local level. Using an ADF detector, the electrons scattered at a high angle are used to form the images, allowing the visualization of heavier elements as the signal is proportional to Z1.6−2, 49,50 where Zis the atomic number. On the other hand, ABF provides complementary information of the framework, as it is more sensitive to light compounds, even with a higher spatial resolution. 36,38,39 For the EMT framework type, the most suitable projections for the structural observation are [110]/[100] two equivalent orientations with the minimum atomic column overlapping; hereafter, we will refer only to [100] for simplicity. Figure 3a corresponds to the Cs-corrected STEM-ABF micrograph of Pb-loaded EMT zeolite along the [100] orientation; in this case, the atomic columns appear as black dots and the big pores, which are clearly observable corresponding to 12MRs. The white arrows point to a structural defect resulting in the merging of two faujasite sheets, which breaks the AA′A stacking; the FD shown in the inset proves that the spatial resolution that can be achieved is at least 1.2 Å represented by the 00022 diffraction spot, marked by a yellow circle. The symmetry-averaged image is depicted in Figure 3b; although the ABF data already present a sufficient degree of quality to analyze the structure at an atomic level, the averaged data are displayed to facilitate the observation of some particular aspects. Symmetry averaging image treatment presents certain advantages as it allows discussion about plane symmetry elements and it significantly increases the signal-to-noise ratio. On the other hand, it introduces undesired periodicity, which is disadvantageous for nonperiodic materials such as zeolites with uneven cationic distribution. Here, symmetry averaging has been used to improve the signal-to-noise ratio; for EMT topology, the reported space group is P63/mmc (194); therefore, its plane group along [100] should be p2gm. The averaged data assuming this space group is depicted in Figure 3b, highlighting (i) T atoms (blue arrow) and (ii) oxygen bridges (red arrows) and additional signals that are not part of the framework (pink and yellow arrows). To facilitate the interpretation, the model of the framework has been superimposed where Si appears in blue and O in red; for clarity, the presence of extra framework cationic species has been omitted. Figure S2 compares the experimental data (Figure S2a) with the simulated image (Figure S2b) of the allsilica framework (without cationic species to highlight the differences between the bare framework and the experimental data); in both cases, colored arrows have been used to point the framework atoms and the extra framework sites. Figure 3b exhibits excellent image quality and it corroborates the feasibility of ABF imaging for the observation of light compounds as oxygen bridges. In addition, the two distinct signals that do not belong to the framework can be inferred to be cationic species; however, an aspect that needs to be considered is that, to obtain this image, symmetry restraints have been introduced and the presence of those extra framework species does not necessarily need to be periodic. Figure 4. (a) Cs-corrected STEM-ADF along [100] of Pb-EMT. The three colored rectangles correspond to the areas where the intensity analysis was performed. The dashed yellow rectangle indicates from where figure (b) was obtained from. (b) Closer observation of the framework where the bright spots are marked by yellow arrows. An empty s6R is denoted by a dashed blue circle. The dashed pink circle corresponds to s6R with a cation inside. (c) Enlarged image of s6R where the two arrows indicate where the intensity profiles were extracted from, numbered as 1 and 2. (d) Intensity profiles along 1 and 2. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6465 To evaluate this aspect and to study the possible introduction of Pb2+ in the pores after ion exchange, we turned to Cscorrected STEM-ADF without imposing any symmetry averaging, as shown in Figure 4.Figure 4a depicts the corresponding Cs-corrected STEM-ADF data of the Pb-EMT zeolite on the [100] projection; here, the contrast is reversed with respect to the ABF data, where the atomic columns appear as bright spots and in which the 12MRs are observed together with the d6Rs and the sodalite cages. Interestingly, a strong contrast is observed as bright spots, which appear at specific sites in a nonperiodic manner. The areas marked by rectangles with different colors, numbered 1−3, were enlarged to analyze the contrast variations at different regions of the crystal, while the additional dashed yellow rectangle corresponds to the area that is magnified, as shown in Figure 4b. The intensity profiles along several sodalite cages were extracted (Figure S3), observing some maxima corresponding to the bright spots. In Figure 4a, the 12MRs are clearly visualized as empty pores; meanwhile, the bright spots would be placed at the sites where Na+has been proposed to occupy, corresponding to the X3and X4sites. This observation is in agreement with the data obtained with an ABF detector (see Figure 3b, yellow arrow), which indicates a very strong signal (that appears periodically due to the symmetry-averaging) near the oxygen bridges. Such signal observed in the ABF and in the ADF cannot derive from the O bridges due to their low scattering factor, and therefore, it is owed to the presence of single Pb2+ sites (single atoms, or atomic columns). Due to the low initial occupancy of the extra framework cations (K+and Na+here), it cannot be estimated whether Pb2+ is in the form of single atoms or if there are several Pb2+ on the same column. Nevertheless, X3and X4sites correspond to two distinct atomic sites with different coordinates; however, due to their proximity in the projected image ≈0.5 Å, it cannot be initially distinguished whether Pb2+ replaced X3,X 4, or both. For in-depth investigation of the ion-exchange capability and therefore the location of the cations, the schematic model of a sodalite cage is shown in Figure S4a. The color code is same as in Figure S1, which displays T3(T = Si, Al) crystallographic sites in purple, forming s6R, where the light-blue sphere corresponds to the cationic site at X3.BlueTatoms correspond to T1, brown spheres are the T2sites, purple T3, and green T4. Cations at X4are marked in pink, while cationic sites at X2appear in yellow. If Pb2+ would occupy the X4site, this would correspond to the location of Pb2+ on top of the S6Rs of the t-toc cage; see Figure S4a. On the other hand, if Pb2+ would replace X3, this one would be in a similar position, on top of another S6Rs of the t-toc cage but closer to the framework. Figure S4b presents an entire unit cell with the same color code, along this projection, the proximity between X3and X4. It is important to mention that no experimental evidence was found of Pb2+ occupying the X2sites, which in the projected image would be inside s6Rs very close to T1 atoms. The other extra framework signals which were not as intense as Pb2+ were observed within s6Rs. In this case, the signal marked by a pink arrow in the ABF data (Figure 3b) and by a pink dashed circle in Figure 4b, would correspond to K+ cations (based on the chemical analyses, the K+content is much higher than Na+, although because of their similar atomic numbers, it would not be possible to distinguish between them). These extra framework species were not periodically distributed along the entire framework, as empty s6Rs could also be identified; see dashed blue circles in Figure 4b. This observation could be associated with the presence of aluminum in the framework, suggesting that the aluminum was not periodically distributed as the empty s6Rs would be linked to the absence of Al at those sites; meanwhile, the extra framework cationic presence suggests the existence of Al around them. Figure 4c depicts an enlarged image of 12MR, obtained from the yellow dashed rectangle in Figure 4b, where K+can be located within s6R. Two different intensity profiles were plotted in Figure 4d, along the arrows 1 and 2, where two peaks are present: the first one and more intense corresponds to an atomic column of the framework, Si or Si + Al, while the other belongs to K+. The measured projected distances were d1 = 2.05 Å, corresponding to the distance between a T2cation and d2= 2.28 Å, corresponding to the distance between the T4 cation; these measurements (slightly shorter) are in agreement with the data reported for Na-EMT (for X4sites), where projected distances are 2.20 and 2.42 Å, respectively. Furthermore, no evidence for the presence of K+or Pb2+ at X2sites was found in any of the equivalent sites. Based on these observations, Pb2+ would preferentially replace the cations at the X3sites, as shown in Figure 4b (yellow arrows). Despite that the X3and X4sites are too close to be distinguished, we have not observed evidence for the presence of Pb2+ at s6Rs, denoted by pink dashed circles in Figure 4b, which would correspond to K+at X4sites. Therefore, as K+at X4sites has two equivalent positions if Pb2+ would enter at X4, this stronger signal would be observed at the sites indicated by a pink circle and at the sites pointed by yellow arrows. As this is not the case, it is possible to conclude that Pb2+ would only go into the sites denoted as X3; in fact, as K+is at X4, the total signal detected at X3would be mostly owing to Pb2+ but with a contribution from K+present at X4. Despite that, the Pb2+ signal was not homogeneously observed along the entire crystal; it always occupied a particular crystallographic site. However, due to the low occupancy of the extra framework cations, 0.33 for X4, it does not appear along the entire crystal. From these observations, it can also be inferred that Pb2+ would compensate the electric charge of the cations at X2sites; as seen from the experimental data, no signal was detected at these positions, suggesting that the cations were removed from these sites but Pb2+ did not fill the positions; the electroneutrality was maintained as one Pb2+ would compensate two monoatomic species of either Na+or K+. For comparison, simulated data were compared with the experimental data; see Figure S5.Figure S5a corresponds to the simulated image of the EMT framework with no cations added to the structure. Figure S5b displays the same image with different amounts of K+and Pb2+ added to the system. The blue dashed circle corresponds to s6R where K+is not observed; the number of K+that was placed at that site was six atoms in that column; therefore, for a given thickness of 8 nm (≈4 unit cells), six atoms of K+per column were not detectable. The pink dashed circle corresponds to K+at s6Rs (18 atoms per column, which are clearly visible). Meanwhile, the yellow arrow points at the signal owing to Pb2+ (in this case, the column was filled by six atoms). Figure S5c corresponds to the experimental data obtained. To ignore that the strong signal observed can be the result of the possible interaction of the framework atoms with the cations (due to their proximity on the projection image), additional simulations with different cationic contents are depicted in Figure S5d,e. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6466 3.3. Framework Analysis and Crystallographic Determination of Pb2+ and K+Positions. How image analysis at the atomic level can provide local information of the zeolitic framework and extra framework cations has been proved so far. On the other hand, electron diffraction can deliver more general information as the solution obtained is based on an average over several unit cells. The structure of Pb-EMT zeolite was solved based on the diffraction intensities extracted from the 3D-EDT using direct method. Figure S6 displays the extracted electron diffraction patterns along the main crystallographic orientations. The patterns shown in Figure S6a−c belong to c*,b*, and a*projections, respectively. Along the c*- axis, this zeolite exhibits 6-fold symmetry, while two perpendicular mirror planes are present along the other orientations, running along a*,b*, and c*axes (see Figure S6b,c). From the hhk plane (Figure S6d), the reflection condition 00l:l=2nwas confirmed, marked by a blue rectangle. For hk0, hk1, hk2, and hk3 planes, hhl and h-hl (l= 1, 2, 3, 4) reflections are marked with green and pink rectangles, respectively. Reflections h-hl appear in all of the four layers, while hhl only appears when l=2n(Figure S6e− h). These reflection conditions are in agreement with the extinction symbol P__c. Thus, only three space groups fulfill these conditions P63mc (186), P-62c(190), and P63/mmc (194). These three were used to solve the structure, obtaining reliable results when P63mc and P63/mmc were used. The structure solution assuming the highest symmetry would be P63/mmc (194) in agreement with the EMT topology, 24 (Figure 5a). Based on the data extracted from 3D-EDT, (see Table S2 for atomic coordinates) Pb2+ cations would occupy the t-wof cage with 4 equiv sites, represented as green spheres in Figure 5a. Meanwhile, the light cations Na+/K+situated within the large t-wou cages have 12 equiv sites per cell (rose spheres). These results are in concordance with the experimental atomic-resolution data presented in Figures 3 and 4along the [100] projection (yellow arrows in Figures 3 and 4). Although the averaged solution from 3D-EDT suggested four equivalent positions, the experimental Cscorrected data revealed that Pb2+ did not occupy all of the four sites. Instead, Pb2+ would not fill the sites situated opposite to each other, corresponding to either PbAor PbBin Figure 5a. Based on these observations, if two Pb2+ atoms from either PbA or PbBsites are removed, the space group would lower its symmetry changing from P63/mmc to P63mc (see Figure 5b), where the two Pb2+ atoms (green spheres) are marked by yellow circles. On the other hand, light cations Na+/K+were found at s6Rs of the t-wou cages (rose spheres). At this crystallographic site, some of the Na+/K+symmetric equivalent sites overlap with Pb2+ along [110]/[100] projections, as shown in Figure 5b, but they are in different cages. To reach out to these findings, we turned out to image analyses (Figure 5c). Figure 5c displays the ADF symmetryaveraged data obtained after using the plane groups marked in Figure 5c; p1 is the experimental data with no symmetry operation imposed, and only translation. The other two images were obtained after using pg (plane group from P63mc) and p2gm (plane group from P63/mmc). To clearly highlight the differences between the three images, color circles were superimposed in the figure; green would correspond to the strong signals associated with Pb2+, rose circle indicates the light signals from K+, and the yellow circle indicates equivalent sites for Pb2+, which would overlap the symetrically related sites of K+marked by rose circles in Figure 5c, as shown in Figure 5b. Although Pb2+ and K+, green and rose circles, respectively, are equal for the three cases the major difference is indicated by the yellow circle as a strong signal is observed for p2gm that is absent for pg and for p1 (experimental data). Based on this significant difference, it is possible to conclude Figure 5. (a) Structural solution using P63/mmc (194). (b) Structural model after removing PbAor PbB, leading to a lower symmetry P63mc. (c) Cs-corrected STEM-ADF data after symmetry averaging. The colored circles indicate the same crystallographic sites in each image, green corresponds to Pb2+, rose corresponds to K+, and yellow would be the equivalent site to Pb2+ marked by a green circle. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6467 that Pb2+ did not fill the sites situated opposite to each other occupying two of the four symmetrically related sites and that the correct plane group should be pg, and therefore, the correct space group considering this cationic distribution is P63mc. By comparing the experimental data acquired by Cscorrected imaging at an atomic level, it was possible to elucidate preferential sites for Pb2+ and K+as well as evaluate the ion-exchange capability and the sites that are more suitable to be initially exchanged. The electron diffraction data allowed the characterization over an averaged volume of the structure, and the results obtained were in agreement with that obtained by Cs-corrected STEM, allowing us to conclude that Pb2+ is located in the t-wof cage and K+is located in the t-wou cage. By imaging, it was possible to acquire local information showing the absence of cations at positions that should have been occupied based on the electron diffraction should be occupied. This is because the results obtained by electron diffraction were averaged over a relatively large area (100 nm, limited by the diffraction aperture used). Therefore, the atomic coordinates occupied by Pb2+ and Na+/K+, which were not part of the zeolite framework, were the result of not single atoms but the preferential sites for cationic occupancy that on average tend to go to those sites. These results illustrate the importance of combining electron diffraction with atomicresolution imaging to achieve a complete characterization of the structure of zeolites, especially with the intention of gaining information at a local level. 4. CONCLUSIONS In conclusion, a combination of various electron microscopy techniques has been used to analyze the structure and the ionexchange capability of EMT zeolite, where K+and Pb2+ were both introduced in the pores. By using a multitechnique approach, the particular features of Pb-EMT have been revealed. SEM data allowed the visualization of the morphology, obtaining symmetry elements, which facilitated the structural solution. Cs-corrected STEM observations displayed data with enough quality to obtain atomic-resolution information, even showing the oxygen bridges as well as cations Pb2+ and K+that were atomically distributed over the framework. The atomic coordinates were determined by 3D-EDT. Both Pb2+ and K+ were found to be outside the sodalite cages and above the s6Rs of these cages, while Pb2+ was in the t-wof cages, K+was in twou cages. Overall, atomic-resolution images provided insights of the local structure of Pb-EMT gaining information on the ion-exchange capability and on the location and periodicity of the extra framework cations, which helped achieving a more accurate structural solution. By the combination of 3D-EDT with the image analysis, it was concluded that crystal symmetry had to be reduced from P63/mmc to P63mc. ■ASSOCIATED CONTENT * sıSupporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jpcc.1c00550. Experimental details, including chemicals, synthesis, and characterization techniques (PDF) ■AUTHOR INFORMATION Corresponding Authors Jennifer E. Readman −School of Physical Sciences and Computing, University of Central Lancashire, Preston PR1 2HE, United Kingdom; Email: [email protected] Alvaro Mayoral −Instituto de Nanociencia y Materiales de Aragón (INMA), CSIC-Universidad de Zaragoza, Zaragoza 50009, Spain; Center for High-Resolution Electron Microscopy (CℏEM), School of Physical Science and Technology (SPST), ShanghaiTech University, Shanghai 201210, China; Advanced Microscopy Laboratory (LMA), University of Zaragoza, Zaragoza 50018, Spain; orcid.org/0000-0002-5229-2717; Email: amayoral@ unizar.es Authors Yaping Zhang −Center for High-Resolution Electron Microscopy (CℏEM), School of Physical Science and Technology (SPST), ShanghaiTech University, Shanghai 201210, China Daniel Smith −School of Physical Sciences and Computing, University of Central Lancashire, Preston PR1 2HE, United Kingdom Complete contact information is available at: https://pubs.acs.org/10.1021/acs.jpcc.1c00550 Author Contributions This manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript. Notes The authors declare no competing financial interest. ■ACKNOWLEDGMENTS The authors deeply thank Professor Osamu Terasaki for fruitful discussions. The authors would like to thank the Centre for High-resolution Electron Microscopy (CℏEM), supported by SPST of ShanghaiTech University under contract No. EM02161943 and the National Natural Science Foundation of China NFSC-21850410448 and NSFC21835002. A.M. also acknowledges the Spanish Ministry of Science through the Ramon y Cajal Program (RYC2018-024561-I) and the regional government of Aragon (Spain) (DGA E13_20R). This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement no. 823717ESTEEM3. ■REFERENCES (1) Li, Y.; Li, L.; Yu, J. Applications of Zeolites in Sustainable Chemistry. Chem 2017,3, 928−949. (2) Atienzar, P.; Valencia, S.; Corma, A.; Garcia, H. TitaniumContaining Zeolites and Microporous Molecular Sieves as Photovoltaic Solar Cells. ChemPhysChem 2007,8, 1115−1119. (3) Koeppe, R.; Bossart, O.; Calzaferri, G.; Sariciftci, N. Advanced Photon-Harvesting Concepts for Low-Energy Gap Organic Solar Cells. Sol. Energy Mater. Sol. Cells 2007,91, 986−995. (4) Nishida, K.; Ohnishi, A.; Kitaura, M.; Sasaki, M.; Kuriyama, Y. Optical Absorption of Cdi2 single Molecule and Clusters Incorporated into Zeolite Na-FAU. J. Phys. Soc. Jpn. 2009,78, No. 104704. (5) Xu, D.; Wang, S. Y.; Wu, B. S.; Zhang, B.; Qin, Y.; Huo, C. F.; Huang, L. H.; Wen, X. D.; Yang, Y.; Li, Y. W. Highly Dispersed Single-Atom Pt and Pt Clusters in the Fe-Modified Kl Zeolite with Enhanced Selectivity for N-Heptane Aromatization. ACS Appl. Mater. Interfaces 2019,11, 29858−29867. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6468 (6) Sun, Q. M.; Wang, N.; Zhang, T. J.; Bai, R.; Mayoral, A.; Zhang, P.; Zhang, Q. H.; Terasaki, O.; Yu, J. H. Zeolite-Encaged Single-Atom Rhodium Catalysts: Highly-Efficient Hydrogen Generation and Shape-Selective Tandem Hydrogenation of Nitroarenes. Angew. Chem., Int. Ed. 2019,58, 18570−18576. (7) Qiu, J. Z.; Hu, J. B.; Lan, J. G.; Wang, L. F.; Fu, G. Y.; Xiao, R. J.; Ge, B. H.; Jiang, J. X. Pure Siliceous Zeolite-Supported Ru SingleAtom Active Sites for Ammonia Synthesis. Chem. Mater. 2019,31, 9413−9421. (8) Dong, B.; Retoux, R.; de Waele, V.; Chiodo, S. G.; Mineva, T.; Cardin, J.; Mintova, S. Sodalite Cages of Emt Zeolite Confined Neutral Molecular-Like Silver Clusters. Microporous Mesoporous Mater. 2017,244,74−82. (9) Nozue, Y.; Tang, Z. K.; Goto, T. Excitons in Pbi2 Clusters Incorporated into Zeolite Cages. Solid State Commun. 1990,73, 531− 534. (10) Kang Tang, Z.; Nozue, Y.; Goto, T. Quantum Size Effect on the Excited-State of Hgi2, Pbi2 and Bii3 Clusters and Molecules in Zeolite Lta. J. Phys. Soc. Jpn. 1992,61, 2943−2950. (11) Tang, Z. K.; Nozue, Y.; Terasaki, O.; Goto, T. Frenkel Excitons in Ordered Pbi2 Clusters Incorporated into Zeolite. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 1992,218,61−66. (12) Sarwar, S.; Ko, K.-W.; Han, J.; Han, C.-H.; Jun, Y.; Hong, S. Improved Long-Term Stability of Dye-Sensitized Solar Cell by Zeolite Additive in Electrolyte. Electrochim. Acta 2017,245, 526−530. (13) Lee, H.; Kim, H. S. Work-Function Engineering in Lead Sulfide and Cadmium Sulfide Quantum Dots Incorporated into Zeolite Y Using Ion Exchange. Part. Part. Syst. Charact. 2016,33, 126−131. (14) Baekelant, W.; et al. Confinement of Highly Luminescent Lead Clusters in Zeolite A. J. Phys. Chem. C 2018,122, 13953−13961. (15) Leiggener, C.; Calzaferri, G. Synthesis and Luminescence Properties of Ag2s and Pbs Clusters in Zeolite A. Chem. - Eur. J. 2005, 11, 7191−7198. (16) Fenwick, O.; et al. Tuning the Energetics and Tailoring the Optical Properties of Silver Clusters Confined in Zeolites. Nat. Mater. 2016,15, 1017−1022. (17) Terasaki, O.; Tang, Z. K.; Nozue, Y.; Goto, T. Pbi2 Confined in the Spaces of Lta Zeolites. MRS Online Proc. Libr. 1991,233, 139− 143. (18) De Cremer, G.; et al. Characterization of Fluorescence in HeatTreated Silver-Exchanged Zeolites. J. Am. Chem. Soc. 2009,131, 3049−3056. (19) http://www.iza-online.org. (20) Ming, D.; Allen, E. Use of Natural Zeolites in Agronomy, Horticulture, and Environmental Soil Remediation. Rev. Mineral. Geochem. 2001,45, 619−654. (21) Moliner, M.; Diaz-Cabanas, M. J.; Fornes, V.; Martinez, C.; Corma, A. Synthesis Methodology, Stability, Acidity, and Catalytic Behavior of the 18 ×10 Member Ring Pores Itq-33 Zeolite. J. Catal. 2008,254, 101−109. (22) Kokotailo, G. T.; Ciric, J. Synthesis and Structural Features of Zeolite Zsm-3. Adv. Chem. Ser. 1971,101, 109−121. (23) Delprato, F.; Delmotte, L.; Guth, J. L.; Huve, L. Synthesis of New Silica-Rich Cubic and Hexagonal Faujasites Using Crown-EtherBased Supramolecules as Templates. Zeolites 1990,10, 546−552. (24) Baerlocher, C.; McCusker, L. B.; Chiappetta, R. Location of the 18-Crown-6 Template in Emc-2 (Emt). Rietveld Refinement of the Calcined and as-Synthesized Forms. Microporous Mater. 1994,2, 269−280. (25) Derakhshankhah, H.; Jafari, S.; Sarvari, S.; Barzegari, E.; Moakedi, F.; Ghorbani, M.; Varnamkhasti, B. S.; Jaymand, M.; Izadi, Z.; Tayebi, L. Biomedical Applications of Zeolitic Nanoparticles, with an Emphasis on Medical Interventions. Int. J. Nanomed. 2020,15, 363−386. (26) Chen, X. Y.; Nik, O. G.; Rodrigue, D.; Kaliaguine, S. Mixed Matrix Membranes of Aminosilanes Grafted Fau/Emt Zeolite and Cross-Linked Polyimide for CO2/CH4 Separation. Polymer 2012,53, 3269−3280. (27) Kim, S. H.; Lim, W. T.; Kim, G. H.; Lee, H. S.; Heo, N. H. Synthesis and Crystal Structure of Lead Iodide in the Sodalite Cavities of Zeolite a (Lta). Bull. Korean Chem. Soc. 2006,27, 679−686. (28) Kim, H. S.; Yoon, K. B. Preparation and Characterization of Cds and Pbs Quantum Dots in Zeolite Y and Their Applications for Nonlinear Optical Materials and Solar Cell. Coord. Chem. Rev. 2014, 263−264, 239−256. (29) Chen, W.; Wang, Z. G.; Lin, Z. J.; Qian, J. J.; Lin, L. Y. New Observation on the Formation of Pbs Clusters in Zeolite-Y. Appl. Phys. Lett. 1996,68, 1990−1992. (30) Kim, S. H.; Heo, N. H.; Kim, G. H.; Hong, S. B.; Seff, K. Preparation, Crystal Structure, and Thermal Stability of the Cadmium Sulfide Nanoclusters Cd6s44+ and Cd2na2s4+in the Sodalite Cavities of Zeolite a (Lta). J. Phys. Chem. B 2006,110, 25964−25974. (31) Togashi, N.; Sakamoto, Y.; Ohsuna, T.; Terasaki, O. Arrayed Pbi2 Clusters in the Spaces of Zeolite Lta. Mater. Sci. Eng., A 2001, 312, 267−273. (32) Ye, S.; Sun, J. Y.; Han, Y. H.; Zhou, Y. Y.; Zhang, Q. Y. Confining Mn(2+)-Doped Lead Halide Perovskite in Zeolite-Y as Ultrastable Orange-Red Phosphor Composites for White LightEmitting Diodes. ACS Appl. Mater. Interfaces 2018,10, 24656−24664. (33) Li, C.; Zhang, Q.; Mayoral, A. Ten Years of Aberration Corrected Electron Microscopy for Ordered Nanoporous Materials. ChemCatChem 2020,12, 1248−1269. (34) Mayoral, A.; del Angel, P.; Ramos, M. Electron Microscopy Techniques to Study Structure/Function Relationships in Catalytic Materials. In Advanced Catalytic Materials: Current Status and Future Progress; Domínguez-Esquivel, J.; Ramos, M. M., Eds.; Springer: Cham, 2019; pp 97−128. (35) Li, J.; Zhang, C.; Jiang, J.; Yu, J.; Terasaki, O.; Mayoral, A. Structure Solution and Defect Analysis of an Extra-Large Pore Zeolite with Utl Topology by Electron Microscopy. J. Phys. Chem. Lett. 2020, 11, 3350−3356. (36) Mayoral, A.; et al. Direct Atomic-Level Imaging of Zeolites: Oxygen, Sodium in Na-Lta and Iron in Fe-Mfi. Angew. Chem., Int. Ed. 2020,59, 19361−19721. (37) Zhang, Q.; Mayoral, A.; Li, J.; Ruan, J.; Alfredsson, V.; Ma, Y.; Yu, J.; Terasaki, O. Electron Microscopy Studies of Local Structural Modulations in Zeolite Crystals. Angew. Chem., Int. Ed. 2020, 19403− 19413. (38) Okunishi, E.; Ishikawa, I.; Sawada, H.; Hosokawa, F.; Hori, M.; Kondo, Y. Visualization of Light Elements at Ultrahigh Resolution by Stem Annular Bright Field Microscopy. Microsc. Microanal. 2009,15, 164−165. (39) Okunishi, E.; Sawada, H.; Kondo, Y. Experimental Study of Annular Bright Field (Abf) Imaging Using Aberration-Corrected Scanning Transmission Electron Microscopy (Stem). Micron 2012, 43, 538−544. (40) Weitkamp, J.; Schumacher, R. Proceed. Ninth hit. Zeo. Conf.; von Ballmoos, R.; Higgins, J. B.; Treacy, M. M. J., Eds.; ButterworthHeinemann, 1993; p 353. (41) Burla, M. C.; Caliandro, R.; Carrozzini, B.; Cascarano, G. L.; Cuocci, C.; Giacovazzo, C.; Mallamo, M.; Mazzone, A.; Polidori, G. Crystal Structure Determination and Refinement Via Sir2014. J. Appl. Crystallogr. 2015,48, 306−309. (42) Woo, J.; Guliants, V. V. Qstem-Based Haadf-Stem Image Analysis of Mo/V Distribution in Movtetao M1 Phase and Their Correlations with Surface Reactivity. Appl. Catal., A 2016,512,27− 35. (43) Ishizuka, K.; Eilers, P. H. C.; Kogure, T. Optimal Noise Filters in High-Resolution Electron Microscopy. Microsc. Today 2007,16− 20. (44) Terasaki, O.; Ohsuna, T. What Can We Observe in Zeolite Related Materials by Hrtem? Catal. Today 1995,23, 201−218. (45) Alfredsson, V.; Ohsuna, T.; Terasaki, O.; Bovin, J. O. Investigation of the Surface Structure of the Zeolites Fau and Emt by High-Resolution Transmission Electron Microscopy. Angew. Chem., Int. Ed. 1993,32, 1210−1213. The Journal of Physical Chemistry C pubs.acs.org/JPCC Article https://dx.doi.org/10.1021/acs.jpcc.1c00550 J. Phys. Chem. C 2021, 125, 6461−6470 6469