Expansion of the ADOR Strategy for the Synthesis of Zeolites: The Synthesis of IPC‐12 from Zeolite UOV

Abstract The assembly–disassembly–organization–reassembly (ADOR) process has been used to disassemble a parent zeolite with the UOV structure type and then reassemble the resulting layers into a novel structure, IPC‐12. The structure of the material has previously been predicted computationally and confirmed in our experiments using X‐ray diffraction and atomic resolution STEM‐HAADF electron microscopy. This is the first successful application of the ADOR process to a material with porous layers.

For more than 60 years,z eolites have been almost exclusively prepared via hydrothermal, [1] solvothermal, [2] and ionothermal [3] synthesis techniques.T he recently discovered ADOR (assembly-disassembly-organization-reassembly) strategy [4] is an alternative way to prepare new zeolite structures.T he method consists of the chemically selective disassembly of aparent zeolite into its constituent layers.This is followed by organization of these units into as uitable relative orientation and the reassembly of the units into new materials.C ontrolled disassembly of the parent zeolite is possible when there is aw eakness engineered into the structure during the initial synthesis. [5] In general, this involves the regioselective substitution of silicon for germanium, which is much more hydrolytically sensitive than the silicon species,a llowing selective dissolution of the germanium out of the material.
TheA DOR process is fundamentally different from hydrothermal synthesis in that the final framework-forming step is an irreversible condensation at high temperature (500-700 8 8C) rather than areversible crystallization step.This leads to new zeolites that have unusual properties that include the possibility of preparing isoreticular families of materials with continuously controllable porosity. [6] Of potential great importance is the possibility of preparing materials that do not obey the energy-density rules [7] associated with hydrothermal synthesis,l eading to new zeolites that in the past would have been thought unfeasible synthetic targets. [8] An important point to note is that the reassembly process is very easy to model, leading to final products that are computationally predictable, [9] something that is very difficult in hydrothermal synthesis.
One crucial issue of the ADOR approach not yet demonstrated is its general applicability.U pt on ow,t he ADOR approach to new materials has concentrated on the use of zeolite UTL as the parent material, although other parent zeolites,s uch as IWW have been successfully disassembled. [10] Germanosilicate UTL is an ideal ADOR starting point because of its chemical composition and because of the stability of the layered units that are formed on disassembly.H erein we report the synthesis of an ew zeolite,w hich we name IPC-12, using the ADOR transformation of ag ermanosilicate parent zeolite with the UOV topology. [11] We have previously predicted that UOV would be agood target, [12] but since it has pores in three dimensions, rather than only in two as is the case for UTL,w ecould not rule out the possibility that the layers (which are porous) would be less stable than in UTL.Indesigning this successful procedure,p articular attention has to be given to the factors controlling the ADOR process,such as appropriate chemical and structural properties of the parent material, optimized conditions for disassembly and reassembly of formed intermediates.
IPC-12 retains the same pore systems as UOV in one direction (viewed perpendicular to the bc crystallographic plane (see Figure 1), but has new structural features when viewed in the other two directions.T he structure of the new material has been confirmed using X-ray diffraction and atomic resolution STEM-HAADF electron microscopy.
It was previously predicted that different zeolites are potentially suitable for ADOR application on the basis of their topologies,and in particular the presence of double four ring (d4r) units in their structures. [4,13] Germanium is well known to site preferentially in the d4r units in many structures. [5] Thep redicted transformations for the UOV parent structure are shown in Figure 1.
As ample of germanosilicate zeolite with the UOV topology was synthesized as described in the experimental section. Theo btained sample was single phase and highly crystalline by X-ray diffraction. Them olar ratio of silicon to germanium used in the reaction mixture was 0.5, but chemical analysis (ICP) of the final UOVp roduct indicated aS i/Ge ratio of 3.1. Theu se of 19 FNMR spectroscopy after postsynthetic fluorination according to the method of Tu el and coworkers [14] revealed two main resonances (Supporting Information, Figure SI-2). One peak at around À10 ppm is typically assigned to F À occluded in Ge 4 Si 4 d4rs [14] and ar esonance at À30 ppm that is typically assigned to F À located in the siliceous layer surrounded only by silicon atoms.T hese results indicate that, as in the case of germanosilicate UTL,t he germanium is preferentially accommodated into the d4r units between the layers,indicating that not only do the prepared germanosilicate UOV materials have asuitable topology for the ADOR process,but also asuitable chemical composition.
Then ext stage in the ADOR process is to complete the disassembly of the parent material into its layered components.T he disassembly of the parent germanosilicate UOV was accomplished by exposure of the parent solid to acidic solutions of 0.1m or 12 m HCl for 24 hours at room temperature.Both reactions gave the same product after calcination at 550 8 8C. TheADOR process can be followed by XRD as is shown in Figure 2. Thed irect way to assess the changes that occur during this process is the evaluation of the intensity and position of the peaks with hkl indexes related to inter-(h ¼ 6 0) and intralayer (h = 0) planes.I nt his regard, the 013 and 004 peaks retain their positions,asthe b and c unit cell dimensions are not affected by the ADOR transformation (Figure 1), while 100, 111, and 102 are expected to change position markedly if disassembly of the initial zeolite takes place.I n the case of UOV,t he shift of the 100 to higher 2q values (smaller unit cell a dimension) is particularly revealing of the structural changes ( Figure 2). This confirms the prediction shown in Figure 1o ft he shortening of the crystallographic a axis as the d4r units are removed from between the layers.
TheADOR behavior of UOV differs from that of UTL,as the latter provides two different materials,w hile for UOV only one material is formed under these conditions.I nt he UTL ADOR process at low acidity the disassembly process dominates,but at very low pH (high acidity) the disassembly also happens quickly but is followed by as ubsequent rearrangement process,w ith extra silicon-containing bridges forming between the layers,r esulting,a fter calcination, in ad ifferent zeolite IPC-2. [15][16][17] This rearrangement process does not occur in UOV,i nstead the higher acidity promotes reconnection of the layers without the intercalation of any extra silicon. Ther econnection of the UOV-derived layers is supported because they cannot be swollen using standard techniques,u nlike the UTL-derived layers,w hich can be swollen when first formed. Thereasons why UTL-and UOVtype zeolites behave so differently is yet to be discovered, but one must remember that the ADOR process is as ubtle balance between several different processes (disassembly, organization, intercalation of species between the layers and reconnection of the layers) and it is not surprising that any one of these processes may be slower or thermodynamically disfavored in certain materials.T his seems to be the case for UOV as the re-intercalation step seen for UTL does not seem to occur.
TheXRD patterns for the IPC-12 materials correspond to each other and match well with theoretically predicted one. Given this information, and the fact that we have apredicted structure from computational work, Rietveld refinement against synchrotron X-ray diffraction data was attempted. Compared to many very highly crystalline solids there is not Figure 1. The predicted ADOR process starting with the disassembly (D) of aparent UOV zeolite into layered intermediates by removal of the d4r units, followed by the organization and reassemblys teps (O/ R) into the final material. Note that the process should not affect the structure of the layers themselves (as is seen in the top view), which means the intralayer unit cell parameters (b and c)r emain constant but the interlayer unit cell parameter (a)d ecreases throughout the process. Figure 2. The XRD patters of the initial UOV and intermediates recovered after 5, 30, and 60 minutes of hydrolysis in 12 m HCl, together with the final material after treatment for 1day.Iti sc lear that the positions of those reflections with h = 0are approximately invariant during the process while those with h ¼ 6 0a re significantly shifted, which is consistent with the predicted ADOR process for UOV shown in Figure 1.
as much data in the diffraction patterns for IPC-12. This is not unusual for ADOR-derived materials as any mistakes in the bonding formed in the irreversible final framework forming step cannot be healed as they can in the reversible crystallization of hydrothermal synthesis.T his means that the successful Rietveld refinement requires rather heavy restraints on the SiÀOa nd OÀOi nteratomic distances.T he fit observed and calculated diffraction patterns ( Figure 3) is however, acceptable and the final structural model matches well to that predicted from the previous computational work.
Thep orous system of the parent UOV zeolite can be described as acombination of parallel 12-and 8-ring channels going through the layers and "interlayer" 10-ring channels. The1 2r ing channels are arranged in ap seudo hexagonal arrangement, six such channels surrounding one of the 8-ring channels.I nt he prediction for the ADOR transformation of UOV the hexagonal arrangement of the 12-and 8-ring channels should remain unchanged as these layers should remain intact, and the product IPC-12 should have exactly the same pseudo hexagonal arrangement. This structural model of IPC-12 was confirmed by atomic resolution spherical aberration (C s )c orrected STEM-HAADF images analysis. As is clearly seen in Figure 4, the pseudohexagonal arrangement of the 12-ring channels predicted by the XRD model (Figure 4a)i sc learly visible in the STEM-HAADF images (Figure 4b).
Them ajor difference between the UOV and IPC-12 topologies is the connectivity perpendicular to these 12-/8ring channels.Inthe parent UOVthere is a10-ring channel in this direction. However,inIPC-12 the prediction is that there should be no channel structure in this direction, the 10-rings being reduced to 6-rings by loss of the d4r units.A gain this model is confirmed by both the XRD and STEM-HAADF. Figure 4c shows the images of the ac projection of the IPC-12 structure,showing no obvious channels and arepeat distance of about 10.5 (Figure 4c), which is consistent with the layers in the structure now being connected by 6-rings,a s predicted from the Rietveld refinement (Figure 4d).
Taking into account the structural transformations during the UOV-to-IPC-12 rearrangement (Figures 1a nd 4), the pore system is changed from two dimensional (12 + 8) 10 to 1D (12 + 8) and should therefore be accompanied by the significant loss in microporosity caused by the disappearance of the interlayer porosity (due to the removal of d4rs the 10ring channels become 6-rings,w hich are too small to be  classed as pores). As expected, as measured using argon adsorption experiments (Supporting Information, Figure SI-4), the micropore volume, V micro ,f or IPC-12 material decreased to 0.052 cm 3 g À1 from 0.111 cm 3 g À1 in the parent UOV zeolite.
Given that the structural model is consistent with the XRD,t he TEM and the adsorption measurements we are extremely confident that the predicted model correctly describes the connectivity in the IPC-12 structure.
Ther esults reported here are significant in that they illustrate that the ADOR process is not limited to one parent zeolite only.T he further development of the ADOR technique,a iming towards the design of new UOV-derived zeolites as analogues of the isoreticular zeolites IPC-2, IPC-6, IPC-7, IPC-9, and IPC-10 that can be prepared from UTL, as well as the use of other parent zeolites is in progress. However,i ti sc lear that each parent zeolite has subtly different behavior in the ADOR process and that there is still much to do to fully understand the key features of ADORable materials.Ofparticular note in this work was the worry that porous layers,s uch as those in the UOV topology (with perpendicular pores) would be unstable under the ADOR conditions.T his worry has proved to be unfounded.