ACS Publications. Most Trusted. Most Cited. Most Read
My Activity
CONTENT TYPES

Lewis Acidic Aluminosilicates: Synthesis, 27Al MQ/MAS NMR, and DFT-Calculated 27Al NMR Parameters

  • Martin Kejik
    Martin Kejik
    Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
    More by Martin Kejik
  • Jiri Brus
    Jiri Brus
    Department of NMR Spectroscopy, Institute of Macromolecular Chemistry, Czech Academy of Sciences, Heyrovskeho nam. 2, Prague CZ-16206, Czech Republic
    More by Jiri Brus
  • Lukas Jeremias
    Lukas Jeremias
    Department of Chemistry and Biochemistry, Mendel University in Brno, Brno CZ-61300, Czech Republic
  • Lucie Simonikova
    Lucie Simonikova
    Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
  • Zdenek Moravec
    Zdenek Moravec
    Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
  • Libor Kobera
    Libor Kobera
    Department of NMR Spectroscopy, Institute of Macromolecular Chemistry, Czech Academy of Sciences, Heyrovskeho nam. 2, Prague CZ-16206, Czech Republic
    More by Libor Kobera
  • Ales Styskalik
    Ales Styskalik
    Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
  • Craig E. Barnes
    Craig E. Barnes
    Department of Chemistry, University of Tennessee, Knoxville, Tennessee 37996-1600, United States
  • , and 
  • Jiri Pinkas*
    Jiri Pinkas
    Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
    *Email: [email protected]
    More by Jiri Pinkas
Cite this: Inorg. Chem. 2024, 63, 5, 2679–2694
Publication Date (Web):January 25, 2024
https://doi.org/10.1021/acs.inorgchem.3c04035

Copyright © 2024 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY 4.0.
  • Open Access

Article Views

804

Altmetric

-

Citations

-
LEARN ABOUT THESE METRICS
PDF (6 MB)
Supporting Info (1)»

Abstract

Porous aluminosilicates are functional materials of paramount importance as Lewis acid catalysts in the synthetic industry, yet the participating aluminum species remain poorly studied. Herein, a series of model aluminosilicate networks containing [L–AlO3] (L = THF, Et3N, pyridine, triethylphosphine oxide (TEPO)) and [AlO4] centers were prepared through nonhydrolytic sol–gel condensation reactions of the spherosilicate building block (Me3Sn)8Si8O20 with L–AlX3 (X = Cl, Me, Et) and [Me4N] [AlCl4] compounds in THF or toluene. The substoichiometric dosage of the Al precursors ensured complete condensation and uniform incorporation, with the bulky spherosilicate forcing a separation between neighboring aluminum centers. The materials were characterized by 1H, 13C, 27Al, 29Si, and 31P MAS NMR and FTIR spectroscopies, ICP-OES, gravimetry, and N2 adsorption porosimetry. The resulting aluminum centers were resolved by 27Al TQ/MAS NMR techniques and assigned based on their spectroscopic parameters obtained by peak fitting (δiso, CQ, η) and their correspondence to the values calculated on model structures by DFT methods. A clear correlation between the decrease in the symmetry of the Al centers and the increase of the observed CQ was established with values spanning from 4.4 MHz for distorted [AlO4] to 15.1 MHz for [THF–AlO3]. Products containing exclusively [TEPO–AlO3] or [AlO4] centers could be obtained (single-site materials). For L = THF, Et3N, and pyridine, the [AlO4] centers were formed together with the expected [L–AlO3] species, and a viable mechanism for the unexpected emergence of [AlO4] was proposed.

This publication is licensed under

CC-BY 4.0.
  • cc licence
  • by licence

Synopsis

Model aluminosilicate networks containing [L−AlO3] (L = THF, Et3N, pyridine, triethylphosphine oxide) and [AlO4] sites were prepared through nonhydrolytic sol−gel condensation reactions of the spherosilicate building block (Me3Sn)8Si8O20 with L−AlX3 (X = Cl, Me, Et) and [Me4N] [AlCl4]. Solid-state 27Al MQ/MAS NMR analysis supported by DFT calculations allowed for the resolution of the individual aluminum species and established the crucial inverse correlation between the site symmetry and the observed quadrupolar coupling constant (CQ).

Introduction

ARTICLE SECTIONS
Jump To

Aluminosilicates are industrially indispensable materials with far-reaching global commercial importance. Besides their more mundane use in construction materials and ceramics, their porous forms, most notably zeolites, (1) represent the backbone of industrial heterogeneous acid catalysis. Zeolites are crystalline microporous aluminosilicate networks where some of the [SiO4] tetrahedra are replaced by [AlO4], necessitating charge compensation by a loosely bound cation in the pore space. Protonation gives rise to well-understood intraframework Bro̷nsted acid sites of the Si(μ (2)–OH)Al type, (2) which were in the past believed to be responsible for all the catalytic properties of these materials. More recently, the presence and importance of Lewis acid sites have been uncovered and intensively studied. (3−5) Such sites are not ordinarily present in freshly synthesized zeolites, but instead, they are formed as extra-framework or framework-associated defects under the harsh conditions of high-temperature steaming and calcination; procedures used to increase the catalytic activities and hydrothermal stabilities of zeolites prior to industrial use. Evidence of synergistic interaction between the Bro̷nsted and the Lewis acid sites has also been presented. (6)
Direct characterization of the Lewis acid sites is challenging due to their extremely low volumetric abundance (near-surface species) and the high spin of the 27Al nucleus (I = 5/2), giving rise to a considerable quadrupolar moment, which, in turn, interacts with the electric field gradient (EFG) to produce potentially very broad and complex line shapes. (7) Although simple one-dimensional (1D) 27Al MAS NMR techniques are routinely used to observe highly symmetric species (e.g., [AlO4] or [AlO6]), which exhibit relatively narrow line shapes, other species of lower symmetry can result in resonances spanning several MHz (even tens of MHz), making them practically “invisible”. (6) The development of multiple-quantum (MQ) NMR techniques has recently allowed for direct observation and resolution of some of the low-symmetry species in zeolites. (8−10) While direct observation of the hypothesized planar [Al(OSi)3] remains elusive (predicted CQ = 34.8 MHz) (11) and requires application of strong magnetic fields and ultrawide-line NMR experimental approaches, (7) its presence in activated zeolites has been inferred through chemisorption of probe ligands (L), most notably pyridine (py), yielding resonances assigned as the corresponding [L–AlO3] species. (8) The small quantity of the species of interest (compared to the total Al content of the materials) hinders more direct investigation of their properties. Therefore, a method of selective preparation of the Lewis acidic Al species in silicate matrices (single-site materials) is sought.
On the synthetic front, Barnes et al. have developed a two-step nonhydrolytic sol–gel procedure for the preparation of microporous metallosilicates under mild conditions. (12−15) The first step introduces metal centers in a nonhydrolytic condensation of the molecular spherosilicate building block (Me3Sn)8Si8O20 (CUBE) with metal precursors MXn (X = halide, alkyl) with the elimination of a Me3SnX species. The oligomeric species are cross-linked in the second step with multifunctional chlorosilanes (SiCl4, HSiCl3, MeSiCl3, and Me2SiCl2). More recently, Styskalik et al. (16) have used the method under milder conditions, starting from L–AlCl3 and [AlCl4], and utilized long, hybrid chlorosilane linkers ClMe2Si–(CH2)n–SiMe2Cl (n = 1–3) to produce single-site Lewis acidic materials with enhanced porosity. Their investigation focused on the catalytic application of the products and did not include any MQ/MAS analysis.
In this work, we concentrated on the first step of the synthetic method described by Styskalik et al. (16) to obtain a series of sparse, amorphous, statistically connected aluminosilicate networks (oligomers) containing [L–AlO3] and [AlO4] sites (L = tetrahydrofuran (THF), pyridine (py), triethylamine (Et3N), or triethylphosphine oxide (TEPO)). The CUBE building block was reacted with a limited amount of L–AlCl3, L–AlMe3, L–AlEt3, or [Me4N] [AlCl4] in toluene and THF (Figure 1). The major emphasis of this investigation was placed on the identification of well-defined structural units in these materials and their characterization by 27Al TQ/MAS NMR techniques in order to extract their spectroscopic parameters─the isotropic chemical shift (δiso), the quadrupole coupling constant (CQ), and the asymmetry parameter (η). The chosen palette of ligands provides a range of site symmetries and coordination strengths to correlate their effect on the synthesis and spectroscopic parameters. The impact of the local disorder can be studied in comparisons of materials with two different Al loadings, leading to varying extents of cross-linking and, thereby, different levels of conformational freedom for the sites. The experimentally determined NMR parameters were compared to the values predicted by DFT calculations on the corresponding fully relaxed, truncated models to provide a sound assignment to the experimental spectra and establish a correlation of the CQ parameter with the site symmetry. The insight gained from this investigation should provide more clarity to the assignment of 27Al solid-state NMR spectra of the more opaque, real-world systems, such as the aforementioned activated zeolites. Finally, a reaction mechanism is proposed for a side-reaction generating [AlO4] sites.

Figure 1

Figure 1. Synthetic strategy leading to statistically linked aluminosilicate oligomers.

Experimental Section

ARTICLE SECTIONS
Jump To

Chemicals and Methods

All procedures were conducted with strict exclusion of oxygen and moisture, using Schlenk techniques with a vacuum-nitrogen manifold. Both precursors and products were stored and manipulated in a drybox (<1 ppm of H2O/<1 ppm of O2). All glass surfaces of reaction vessels and storage vials were silylated by an equimolar mixture of Me3SiCl and Et3N in CH2Cl2 prior to each use to eliminate surface −OH groups and thus prevent the production of HCl in contact with the Al–Cl functionality. The reaction vessels were sealed by PTFE sealing rings. All solvents were dried, deoxygenated, and stored under nitrogen over activated 4 Å molecular sieves. Starting compounds SnMe3Cl, AlMe3 (2 M in toluene), AlCl3, [Me4N]OH (25 wt % in MeOH), Si(OEt)4, Me3SiCl, pyridine, Et3N, and TEPO were purchased from Merck and used without further purification. The building block (Me3Sn)8Si8O20 was prepared according to a previously described procedure (17) from SnMe3Cl, [Me4N]OH, and Si(OEt)4. The compounds py–AlMe3, Et3N–AlMe3, TEPO–AlMe3, and [Me4N] [AlCl4] were synthesized according to the procedures described in the Supporting Information (Section S1) and stored in the drybox. The compounds THF–AlCl3, py–AlCl3, and TEPO–AlCl3 were generated in situ and never isolated to avoid the formation of complex ionic solids with the potential loss of the ligands. (18,19)
Caution! SnMe3Cl, SnMe4, and SnMe3Et are classified as GHS Category 1 acute toxicity (H300, H330, H310) and long-term aquatic toxicity (H400, H410) hazards. The same hazards should be presumed for both (Me3Sn)8Si8O20 and the final products (contain residual─SnMe3).
Caution! AlMe3 and AlEt3 solutions are classified as GHS Category 1 pyrophoric liquids (H250), skin and eye damage (H314, H318), and aspiration (H304) hazards.

Synthesis of Aluminosilicate Oligomers

A Schlenk vessel equipped with a stir bar was sealed, evacuated, weighed, taken into the drybox, and loaded with (Me3Sn)8Si8O20 (∼2 g, CUBE). Then, the vessel was evacuated and weighed again to determine the exact masses of the vessel and the reagent. The vessel was reconnected to the manifold, the solvent (toluene or THF, 20 cm3) was added by a syringe, and the resulting solution was cooled to −80 °C. In the case of compounds L–AlMe3 (L = py, Et3N, TEPO) and [Me4N] [AlCl4], a weighed 20 cm3 screw-top vial with a PTFE septum was loaded with the corresponding Al complex (1.5 or 3.0 equiv of the reactive groups per CUBE) in the drybox, taken out, and weighed and the solvent (5 cm3) was introduced by a syringe to produce a solution. In the case of L–AlCl3 (L = THF, py, and TEPO), the ligand and AlCl3 were loaded into separate weighed vials and the solvent was first used to dissolve the ligand before it was transferred to the AlCl3 vial to generate the complex in situ. Care was taken to ensure the ratio of L/Al ≥1 to avoid uncoordinated AlCl3. The AlCl3 vial was initially cooled by liquid N2 to dampen the highly exothermic contact of the ligand with the strong Lewis acid. The room-temperature solution of an Al complex was then added dropwise to the cooled solution of the CUBE with vigorous stirring over 10 min. Another portion of the solvent (5 cm3) was used in several doses to ensure a quantitative transfer of L–AlCl3. The Schlenk vessel was sealed and allowed to slowly warm up in the cooling bath to room temperature over 12–18 h. Fast warm-ups generally lead to the formation of inhomogeneous precipitates. After 24 h, the solution was heated at 60/100 °C for another 48 h, at which point a clear to hazy solution of oligomers was obtained. All volatiles were removed under dynamic vacuum at 60/100 °C to obtain a glassy residue, which was broken to powder by a magnetic stirrer and outgassed further for 48 h at 60/100 °C. The evacuated Schlenk vessel was then weighed and taken into the drybox for product handling. The solids were analyzed by 1H, 13C, 27Al, 29Si, and 31P MAS NMR and FTIR spectroscopies, as well as ICP-OES and N2 adsorption porosimetry. The volatiles were collected in a Schlenk-type cold trap and analyzed by solution NMR (Supporting Information, Section S2).
The reactions with compounds L–AlCl3 and [Me4N] [AlCl4] could only be conducted in THF due to their low solubility in toluene. Moreover, Et3N is incompatible with THF as a solvent due to observed ligand scrambling; therefore, this ligand was used only as Et3N–AlMe3 and only in toluene.
The mass lost as the volatile byproducts was used to calculate the gravimetric degree of condensation (DCG/%) of the reacting functional groups DCG(SnMe3) and DCG(AlX), defining the average connectivity at the CUBE and Al, respectively. Ideally, only a single byproduct would be generated (SnMe4, SnMe3Et, or SnMe3Cl); however additional SnMe4 was observed in conjunction with each of the expected byproducts. The excess of SnMe4 will be further linked to an unexpected formation of [AlO4] species, and therefore, it is conveniently quantified in mol % with respect to the Al content as DCG(SnMe4). In the case of compounds L–AlCl3, L–AlEt3, and [Me4N] [AlCl4], the byproduct ratio is determined by 1H NMR integration while with L–AlMe3, a full condensation of Al–Me must be assumed in order to quantify the excess of SnMe4. A full description of the calculation procedure and the complete characterization data for all products are provided in the Supporting Information (Sections S2 and S3, respectively).

Solid-State NMR (ssNMR)

The 13C, 29Si, and 31P MAS ssNMR spectra were acquired in 4 mm ZrO2 rotors on a Bruker Avance Neo 700 MHz spectrometer with a MASDVT700S4 BL4 N–P/H probe head at 298 K. 13C CP/TOSS (20) (176.21 MHz) MAS NMR spectra were obtained with 5 kHz MAS frequency, 500 μs contact time, 4 s delays, and 1H CW decoupling. All 13C ssNMR chemical shifts were referenced externally to adamantane (δ = 38.5 ppm). (21) 29Si (139.23 MHz) and 31P (283.69 MHz) MAS NMR spectra were obtained using the standard 90°-pulse experiment, with 12 kHz MAS frequency, and 60 and 4 s delays for 29Si and 31P, respectively. The 29Si and 31P ssNMR chemical shifts were referenced externally to DSS (δ = 1.5 ppm) (21) and NH4H2PO4 (δ = 1.0 ppm), (22) respectively. The 29Si ssNMR spectra were deconvoluted in SpinWorks 4.2.5 using two peaks (Lorentzian–Gaussian linear combination) centered around −101 (narrow) and −107 (broad) ppm, corresponding to [SnOSi] and [AlOSi] moieties of unreacted and reacted corners of the CUBE, respectively. The ratio of the two peaks was then used to independently calculate the spectroscopic degree of condensation DCN of the tin groups DCN(SnMe3) for the purpose of validation of the combined gravimetry/1H NMR approach. The calculation procedure (Table S1), peak fitting data (Table S18), and the resulting DC values (Table S19) are provided in the Supporting Information.
The 1H and 27Al MAS ssNMR spectra were acquired in 3.2 mm ZrO2 rotors on a Bruker Avance Neo 700 MHz spectrometer with a PH MASDVT700S3 BL3.2 N–P/F-H probe head at 298 K. The samples were packed into the rotors in a drybox, sealed, and stored under the inert atmosphere of the drybox until measurement. The 1H spectra were obtained using standard 90°-pulse excitation at 20 kHz MAS frequency with 8 s delays and 1H chemical shifts referenced to crystalline alanine (δ = 1.2 ppm, CH3 signal). Conventional 1D 27Al (182.48 MHz) MAS NMR spectra were obtained using 30°-pulse excitation (1.0 μs) at 20 kHz MAS frequency with 5 s delays and 1H SPINAL64 decoupling. 2D 27Al TQ/MAS NMR experiments (23) were set up using kyanite as a model system and carried out at 20 kHz MAS frequency with 2 s delays and 1H SPINAL64 decoupling. (24) The standard z-filtered three-pulse sequence with excitation, reconversion, and selective pulse lengths of 4.2, 1.5, and 43.0 μs, respectively, was used. The z-filter pulse length was 20 μs. All 27Al ssNMR chemical shifts were referenced externally to aqueous [Al(H2O)6]3+ (δ = 0.0 ppm).
The 27Al TQ/MAS NMR spectra were processed using the TopSpin 4.1.4 software. Initially, iso-shearing transformation was applied by xfshear in order to recover isotropic line shapes in the indirect dimension (F1). Then, individual sites were identified using the maxima of the projections of the spectra into F1 domain and slices were performed along the direct dimension (F2) to obtain 1D spectra of the quadrupolar line shape for each Al site. Finally, with the effects of the chemical shift anisotropy removed by MAS and the pure line shape separated by TQ NMR, the quadrupolar NMR parameters (δiso, CQ, η) were extracted by quadrupolar line shape simulation fitting using the central transition model. The line broadening parameter (LB) was not optimized but carefully enforced in several steps as it strongly interacts with the CQ parameter (both broaden the line shape). Further refinements of the fits were performed based on the assignment of corresponding sites across multiple synthetic products. All TQ/MAS NMR spectra, extracted line shapes, and details of fitting are provided in the Supporting Information (Section S3).
The details of auxiliary characterization techniques (solution NMR, FTIR spectroscopy, ICP-OES, and N2 adsorption porosimetry) are available in the Supporting Information (Section S2).

Computational Methods

The Al sites of interest were modeled by idealized truncated molecular species containing up to four monodentate (OCUBE = (Me3Sn)7Si8O19(O−)) or two bidentate (O2CUBE = (Me3Sn)6Si8O18(O−)2, bound through neighboring corners) CUBEs. The terminal −OSnMe3 groups of the CUBEs were replaced by −OMe for computational efficiency (fewer nuclei and electrons), and the effect of the substitution was benchmarked. The molecular geometries were optimized by the PBE0 functional (25,26) and the def2-TZVPP basis set (27) (the central part of the molecule, e.g., THF–Al(OSiO3)3, up to and including the first [SiO4] tetrahedron) or the def2-SVP basis set (28) (the rest of the molecule) with the corresponding def2-ECP for Sn atoms (29,30) as implemented in the Turbomole 7.5.0 program. (31) The structures were optimized using the DFT-D3 (32,33) dispersion correction with Becke–Johnson damping and the Conductor-like Screening Model (COSMO) solvent model of THF. For the calculations of chemical shielding, the central part of the molecule was taken out (e.g., THF–Al(OSiO3)3), terminated by hydrogen atoms (e.g., 9 H to form THF–Al(OSi(OH)3)3) and reoptimized with frozen coordinates of the central atoms. All calculations using this protocol were performed with the m5 integration grid and the following convergence criteria: 10–6 for the density matrix change and 10–4 Hartree × Bohr–1 for the geometry gradient.
The calculation of 27Al NMR parameters was performed using the zeroth-order regular approximation (ZORA) (34,35) Hamiltonian at the scalar (spin-free) level, as implemented in the ADF program package (version 2019). (36) The CQ and η parameters (the whole molecule was used) were calculated by the PBE0 functional, the TZ2P basis set (the QZ4P basis set for the Al atom), (37,38) and the COSMO solvent model of THF. The chemical shielding (the central part of the molecule was used) was calculated using the PBE0 functional, the QZ4P basis set, and the COSMO (THF). The calculated chemical shifts were referenced relative to the experimental chemical shift of solid AlCl3 as a secondary reference (δref = −1.6 ppm), (7) and they are reported relative to AlCl3 in D2O (primary reference). The Euler angles (α, β, γ) were extracted from ADF output files using the program INFOR. (39)
The full list of the calculated spectroscopic parameters is available in the Supporting Information (Section S4).

Results and Discussion

ARTICLE SECTIONS
Jump To

The first step of the nonhydrolytic sol–gel procedure of Styskalik et al., (16) as described in the experimental details, was used with the aim to produce single-site materials containing exclusively L–Al(OCUBE)3 (L = THF, Et3N, py, TEPO) or [Me4N]+ [Al(OCUBE)4] sites. The molecular spherosilicate building block CUBE was cross-linked by limited amounts of L–AlCl3, L–AlMe3, L–AlEt3, or [Me4N] [AlCl4], affording sparse, statistically linked networks–oligomers (Table 1). Reactions 15 and 716 produced nonporous glassy solids that could be reversibly redissolved into colloidal viscous solutions, indicating the oligomeric character of the formed species. Reaction 6 produced a gel upon heating and subsequently an insoluble porous (126 m2 g–1) solid upon drying, indicating an even higher extent of cross-linking leading to the formation of an extended 3D network. In the following discussion, the same numbering system refers to both reactions and the resulting materials (products).
Table 1. Synthesis and Important Characterization Data for Prepared Oligomers
  stoichiometry   average connectivity
sampleAl precursorAl/CUBEAlX/CUBEsolvent/temperature (°C)primary condensation DCG(Al–X)(%)excess SnMe4DCG(SnMe4) (mol % of Al)AlCUBE
1THF–AlCl30.992.97THF/60105.127.63.283.39
20.5 py–AlCl30.481.45THF/60100.056.23.561.72
3py–AlCl30.972.90THF/60103.17.13.073.06
40.5 py–AlMe30.491.47TOL/60100.031.83.321.63
5py–AlMe31.013.03TOL/60100.06.93.073.10
6apy–AlMe30.992.96TOL/100100.058.93.593.54
70.5 py–AlMe30.491.46THF/60100.062.23.621.76
8py–AlMe30.992.97THF/60100.011.53.123.08
9py–AlEt31.003.00TOL/6096.76.52.972.97
100.5 Et3N–AlMe30.501.51TOL/60100.016.63.171.60
11Et3N–AlMe31.003.00TOL/60100.07.43.073.07
120.5 TEPO–AlCl30.501.49THF/60100.40.03.001.49
13TEPO–AlCl30.992.97THF/6096.11.72.852.82
140.5 TEPO–AlMe30.501.50TOL/60100.00.53.011.50
15TEPO–AlMe30.982.95TOL/60b83.90.02.472.47
160.375 [Me4N][AlCl4]0.371.49THF/6089.041.93.981.48
a

Porous: SABET = 126 m2 g–1.

b

Heated in solution for 72 h and dried under vacuum for 120 h in attempts to achieve full condensation.

The condensation reaction should ideally produce only a single volatile byproduct SnMe3X (X = Cl, Me, or Et), corresponding to the type of reactive functional groups in the particular L–AlX3 precursor. The analysis of the volatiles removed from the reactions, however, revealed that additional SnMe4 was generated even in the case of precursors containing the Al–Et and Al–Cl functional groups, leading to SnMe3Et/SnMe4 and SnMe3Cl/SnMe4 mixtures, respectively. Conveniently, no evidence of mutual interaction or substituent exchange was observed and the two byproducts could be distinguished and their molar ratio determined by 1H NMR integration, allowing for a combined gravimetry/NMR quantification technique. The resulting values of the gravimetric degree of condensation of the −SnMe3 groups (DCG(SnMe3)) were compared to their spectroscopic counterparts calculated from 29Si MAS NMR integration (DCN(SnMe3)) and elemental analysis obtained from ICP-OES(Sn)/gravimetry (DCE(SnMe3)) (Table S19) to make sure the quantification procedure was accurate. The resulting correlation plot (Figure S33) shows generally good agreement among all three sources of information.
Of note is the observation that no free ligand (L) was detected in the volatile byproducts, with the exception of reactions 13 where the ligands and AlCl3 were supplied separately and deliberately with a slight excess of the ligand. This indicates that L/Al = 1 is maintained.
In the case of precursors L–AlX3 (X = Me, Et, Cl) (reactions 115), the gravimetric analysis revealed that full condensation of the AlX groups DCG(AlX) is generally reached, with the exception of reactions 9, 13, and 15. The small deviations of DCG in reactions 1, 3, 9, 12, and 13 (within 5.1%) could easily be attributed to experimental error, while the deviation of 15 (−16.1%) is a strong proof of incomplete condensation.
The quantification of the excess of SnMe4 showed the generated amount to be significant and highly variable, yet always smaller than the quantity of Al. Since a full condensation was nearly always reached, a separate process involving the Al site and the consumption of another −SnMe3 group must be at play. With no reasonable redox options, the transformation of [L–AlO3] into [AlO4] in a metathesis reaction, liberating SnMe4 at the expense of the introduction of additional connectivity, was suspected. Therefore, the amount of SnMe4 was expressed as a fraction of the quantity of the Al sites DCG(SnMe4). The comparison of reactions 1, 3, 5, 8, 9, 11, 13, and 15 indicates that the quantity of SnMe4 depends strongly on the ligand used in the series THF > Et3N ≥ py > TEPO ≈ 0 while it is independent of the substituent X. Further comparison of the reactions 3, 5, 8, and 11 (AlX/CUBE = 3.0) to their less extensively cross-linked counterparts 2, 4, 7, and 10 (AlX/CUBE = 1.5) clearly demonstrates the suppression of the reaction in samples with more cross-linking. This is attributed to the loss of conformational freedom. Moreover, reactions 4 and 5 (L = py) compared to 10 and 11 (L = Et3N) indicate that with the lower extent of cross-linking, the sterically more demanding Et3N also becomes a limiting factor. Similarly, evidence that more SnMe4 is generated in THF could be found in the contrast between reactions 2 and 7 (in THF) and reaction 4 (in toluene). Finally, reaction 6, which was conducted at an increased temperature (100 °C), produced nearly ten times as much SnMe4 as the corresponding reaction at 60 °C (5), leading to the highest overall average connectivity (3.59 on Al/3.54 on CUBE) at the formation of a porous gel structure (BET 126 m2 g–1).
In the case of [Me4N] [AlCl4], only the AlX/CUBE = 1.5 stoichiometry was used to avoid too heavily cross-linked products. To our surprise, a full condensation, as monitored by the evolved SnMe3Cl, was not reached and SnMe4 was also generated. Further analysis revealed that the total average connectivity on Al reached 4 (3.98), pointing at the existence of a competing mechanism of condensation where the Cl atom remains embedded in the material and SnMe4 is evolved instead.

Basic Characterization

The infrared spectra of dried products (Figure 2) are dominated by the strong vibration bands at 1139 and 1035 cm–1 assigned to the intracage νas(Si–O–Si) and the extra-cage νas(Si–O–Al) linkages of the CUBE, respectively. (40,41) The vibration bands at 1402, 2919, and 2990 cm–1 are also common to all the prepared samples and primarily assigned to the bending and valence vibrations δas(CH3), νs(CH3), and νas(CH3) of the residual −SnMe3 groups. (42) Additional bands, corresponding to the valence/bending vibrations of the ethyl substituents of Et3N (1470 cm–1) (43) and TEPO (1460, 2890, and 2942 cm–1), as well as the δas(CH3) vibrations of [Me4N]+ (1488 cm–1), (44) are observable in the spectra of the corresponding products. The spectra of products prepared from py–AlX3 (X = Me, Et, Cl) (Figure S34) are identical and all contain vibration bands at 1455, 1495, and 1622 cm–1 assigned to Lewis acid-bound pyridine (marked L-py in Figures), (45,46) indicating a single, well-defined mode of ligand coordination and the convergent nature of the synthesis from different precursors. In general, the fingerprint area is nearly identical across all ligands/precursors.

Figure 2

Figure 2. FTIR spectra (KBr pellets) of products 1, 3, 11, 13, and 16, demonstrating the general similarity of the products, with differences in the presence of the vibration bands of the individual ligands.

13C CP/TOSS MAS NMR spectra of all products (Figure 3) contain a prominent resonance from −1 to 1 ppm, assigned to the residual −SnMe3 groups, with additional resonances corresponding to the particular ligand/cation used: THF (28 and 75 ppm), Et3N (13 and 50 ppm), py (129 and 151 ppm), TEPO (10 and 21 ppm), and [(CH3)4N]+ (59.3 ppm). (47) In parallel to the FTIR spectra, the 13C CP/TOSS MAS NMR spectra (Figure S35) of products prepared from py–AlX3 (X = Me, Et, Cl) are identical. The comparison of 13C ssNMR spectra of products 1215 (Figure S36) reveals an additional resonance in the spectrum of 15 at −10 ppm, which could be assigned to residual Al–CH3 due to incomplete condensation as indicated by gravimetry. The negative amplitude of the −SnMe3 resonances is indicative of their large chemical shift anisotropy with respect to the MAS frequency. (48,49)

Figure 3

Figure 3. 13C CP/TOSS MAS NMR spectra of products 1, 3, 11, 13, and 16. Spinning side-bands are denoted by asterisks.

1H MAS NMR characterization (Figures S37 and S38) completely mirrors the 13C ssNMR spectra with a common resonance at ∼0 ppm, assigned to the residual −SnMe3, and individual resonances for the corresponding bound ligands and the cation: THF (1.3 and 3.6 ppm), Et3N (1.1 and 2.8 ppm), py (7.5 and 8.8 ppm), TEPO (1.1 and 1.9 ppm), and [(CH3)4N]+ (3.0 ppm). (50) In the case of product 15, the expected resonance of the residual Al–CH3 is observed at −1.3 ppm (Figure S39).
29Si MAS NMR spectra of all samples contain only two [SiO4] resonances at −101 ppm (narrow) and −107 ppm (broad), assigned to Me3SnOSi(O−)3 and AlOSi(O−)3, respectively (Figure 4). (14,51) The intensity ratio of the two resonances reflects the reaction stoichiometry and the degree of condensation, DC (Figure S33).

Figure 4

Figure 4. 29Si MAS NMR spectra of products 1, 3, 11, 13, and 16. Spinning side-bands are denoted by asterisks.

31P MAS NMR spectra of products 1215 (Figure 5) are dominated by a resonance at 75 ppm, assigned to the intended TEPO–Al(O−)3 site. In the spectra of products 12 and 13, a minor resonance at 55 ppm was assigned to uncoordinated TEPO. (52) Additionally, the spectra of 13 and 15 contain low-intensity shoulders at 81 ppm, for the latter the shoulder is more pronounced. Based on their similarity to the main site and the incomplete condensation indicated by gravimetry, these are assigned as TEPO–AlCl(O−)2 and TEPO–AlMe(O−)2, respectively.

Figure 5

Figure 5. 31P MAS NMR spectra of products 1215. Spinning side-bands are denoted by asterisks; a trace rotor contaminant is denoted by c.

1D 27Al MAS NMR Analysis

27Al{1H} MAS NMR characterizations (Figure 6) of products 115 showed broad asymmetric features with increasing line width in the order TEPO < Et3N < py < THF. Single quadrupolar nucleus line shapes did not yield good fits to the observed features, hinting at the presence of multiple overlapped resonances and preventing direct extraction of the spectroscopic parameters. Product 16 showed a much narrower feature with two maxima at 47 and 52 ppm, allowing for a confident assignment as four-coordinated Al species. (53,54) Most notably, the observed line shapes are more complex (16) and much broader (115) than those previously reported (16) for the corresponding products. This is attributed to the broader-band excitation used in this work and demonstrates the trickiness involved in the characterization of low-symmetry Al species by 27Al NMR techniques.

Figure 6

Figure 6. 27Al {1H} MAS NMR spectra of products 1, 3, 11, 13, and 16.

Multidimensional 27Al MAS NMR Analysis

In contrast, the 2D 27Al TQ/MAS NMR spectra revealed a more complex nature of the prepared materials. The resonance assignments presented below were made based on the clustering (grouping) of the fitted resonances in parameter space (δF1, δiso, CQ) across all the prepared products (Figure 7) and the proximity of the clusters to the values predicted by DFT calculations (CUBEs terminated by −OMe). Further refinements of the fits were made once the initial grouping was established. The full list of all the observed resonances (labeled by product number and a letter, e.g., 1/A) and their 27Al NMR parameters and assignments are available in Table S20. Tables S21 and S22 summarize all the model structures investigated by DFT calculations and the predicted parameters. CQ is the most reliably fitted parameter as it depends primarily on the limits of the peak envelope, while η is the least reliable as it subtly changes the peak shape within the limits of the envelope. The fit of δiso is contaminated by the uncertainty of η. The less chemically relevant indirect chemical shift δF1 can be measured directly with the greatest precision from the 1D projections of the 27Al TQ/MAS spectra. Therefore, the assignments were made using plots of δiso/CQ (Figure 7a) and δF1/CQ (Figure 7b), with the latter showing more distinct clustering and serving as the starting point for the assignment. Figure S40 shows the lack of significant clustering in the η/CQ space, demonstrating the unreliability of the fitted η values.

Figure 7

Figure 7. Plots of all observed 27Al TQ/MAS NMR resonances and relevant DFT-calculated models in parameter spaces (a) δiso/CQ and (b) δF1/CQ.

Figure 8 then represents a portfolio of all the observed 27Al TQ/MAS NMR resonances in fully condensed products. All products 115 displayed broad resonances (CQ = THF: 15.1, py: 11.8–14.1, Et3N: 10.7–12.1, and TEPO: 10.1–10.8 MHz) corresponding to the sought L–Al(OCUBE)3 sites; however, only in the case of 1214 (L = TEPO) were they the only sites present (true single-site materials).

Figure 8

Figure 8. 27Al TQ/MAS NMR spectra of (a) product 1, (b) products 3 and 5, (c) product 11, (d) product 13, (e) product 16, and (f) product 8 (20 kHz MAS).

In products 19 (L = THF, py), another series of narrow resonances (CQ = THF: 12.0, py: 8.0–9.4 MHz), assigned as L–Al(O2CUBE)(OCUBE), was also identified, indicating that a bidentate CUBE binding mode is possible with the two sterically relatively undemanding ligands. The site is visibly more dominant in products 7 and 8 (py–AlMe3/THF) (Figure 8f), than in 2, 3 (py–AlCl3/THF), 4, and 5 (py–AlMe3/toluene), while it is significantly suppressed in 9 (py–AlEt3/toluene). This could suggest that py–AlMe3 is sensitive to solvent effects with THF promoting cyclization more than toluene and that the substitution of Al–Me for larger alkyls tends to disfavor the process.
As suspected based on gravimetry, the spectra of products 112 displayed additional narrow resonance(s) (CQ = 4.8–6.3 MHz), assigned to [AlO4] species. The dominance of these signals in the spectra appears to correlate with the amount of excess SnMe4 generated. Similarly, product 16, which was intended to produce solely [AlO4], displays only a single band of narrow resonances (CQ = 4.4–5.2 MHz) with three discernible maxima of increasing δiso and CQ (Figure 8e). The bands of [AlO4] resonances of 112 possess the same oblique characteristic and generally overlap with the one of 16, allowing for fitting of 3 or fewer maxima.
The product 15 (TEPO–AlMe3/toluene) is a special case where full condensation could not be achieved despite increased reaction and drying times (72 and 120 h). The 27Al TQ/MAS spectrum (Figure S41) revealed the presence of a weak broad resonance (CQ = 12.8 MHz) assigned to TEPO–AlMe(OCUBE)2, but also a significant resonance of [AlO4], indicating that the increased heating times did not overcome the barrier to condensation but rather promoted the formation of [AlO4].
In general, the resonances observed in the spectra of the products with decreased Al loading (2, 4, 7, 10, 12, and 14) were systematically narrower (giving lower fitted values of CQ) than the corresponding ones in the higher-loading products (3, 5, 8, 11, 13, and 15), at the expense of a decreased signal-to-noise ratio. This is interpreted primarily as the result of the increased structural disorder in the more heavily cross-linked materials. Notably, a trend can be found in the CQ values of py–Al(OCUBE)3 with a systematic decrease in the series: 59 (py–AlX3/toluene, X = Me, Et) > 3 (py–AlCl3/THF) > 6 (py–AlMe3/toluene/100 °C) > 4 (0.5 py–AlMe3/toluene) > 2 (0.5 py–AlCl3/THF). This may be interpreted as the substitution of Al–Cl precursors for Al–Me/Al–Et and thermal relaxation (6) also act to decrease the site disorder. Resonances 1/A (THF–Al(OCUBE)3) and 15/C (TEPO–AlMe(OCUBE)2) suffer from uneven excitation due to their extremely large CQ (≈ 15 MHz), resulting in very poor fits.
Table 2 summarizes the average spectroscopic properties of all the observed types of sites, along with the corresponding values predicted by the DFT calculations (−OMe-terminated CUBEs). Since the predictions come from fully relaxed model structures of particular, single conformers, the values of δisoDFT and ηDFT are subject to slight variations and the predicted |CQDFT| must be understood as the lower limits in comparisons to the experimental data; thus, it is best related to the minimum observed values CQmin. The average isotropic chemical shifts δisoavg of all the experimentally observed, well-excited resonances as well as all the corresponding DFT predictions δisoDFT fall within the range of 48–60 ppm, establishing a good agreement with the chemical shifts of 4-coordinate Al species reported previously in the literature. (53,54) Most importantly, the observed CQmin values for the sites L–Al(OCUBE)3 match remarkably well to the predicted |CQDFT|, with both data sets following the key trend of a decreasing CQ with L in the series THF > py > Et3N > TEPO, clearly reflecting the increasing symmetry of the coordinated L in the series THF (C1) < py (C2) < Et3N (close C3) ≤ TEPO (distant C3). The series may be expanded with the DFT predictions for the unobserved sites [AlCl(OCUBE)3] (L = Cl, point-like) and [Al(OCUBE)4] (true Td at Al) to span the range of CQ = 2–15 MHz, pointing to the ligand as the dominant symmetry-breaking element and its connection to the largest component V33 of the EFG tensor. In the case of py–Al(OCUBE)3, the observed CQmin = 11.8 MHz maps very well to the previously reported value of 9.4 ± 0.5 MHz for the analogous sites observed after the adsorption of pyridine onto dehydrated zeolites. (8) Similarly, the observed CQmin = 10.1 MHz for TEPO–Al(OCUBE)3 and 4.8 MHz for [AlO4] compare well to the reported values of 11.3 and 4.6 MHz, assigned to highly distorted sites of the corresponding type in synthetic amorphous aluminosilicate materials after the adsorption of TEPO. (55) The larger positive deviations of CQmin compared to |CQDFT| for L = Et3N (2.3 MHz) and TEPO (2.1 MHz) could be explained as an additional disorder due to the conformational freedom of the pendant ethyl substituents of the two ligands. As illustrated above, the experimental fits of the asymmetry parameter ηavg for the L–Al(OCUBE)3 sites are highly unreliable with values generally larger or comparable to the corresponding predicted ηDFT. In agreement with the predictions, the sites L–Al(O2CUBE)(OCUBE) (L = THF, py) exhibited systematically higher δisoavg and lower CQmin than the corresponding L–Al(OCUBE)3 sites. Although the site TEPO–AlMe(OCUBE)2 yields poor experimental fits, the data supports the assignment with a predicted shift toward larger δiso and CQ, as compared to TEPO–Al(OCUBE)3.
Table 2. Overview of the Extracted Spectroscopic Parameters of Observed Sites and the Corresponding Values Predicted by DFT Calculations
site typesourceδiso(ppm)CQavga(MHz)CQmin(MHz)η
THF–Al(OCUBE)3exp38.115.115.10.683
DFT47.914.6 0.244
py–Al(OCUBE)3exp52.313.011.80.682
DFT52.911.5 0.334
Et3N–Al(OCUBE)3exp48.211.410.70.848
DFT53.18.4 0.226
TEPO–Al(OCUBE)3exp50.810.410.10.431
DFT48.58.0 0.453
THF–Al(O2CUBE)(OCUBE)exp49.712.012.00.100
DFT55.411.5 0.950
py–Al(O2CUBE)(OCUBE)exp60.48.78.00.510
DFT60.08.0 0.974
TEPO–AlMe(OCUBE)2exp66.912.812.80.439
DFT84.514.9 0.614
[AlO4]exp: 1–1557.65.84.80.577
exp: 16/A51.74.4 0.477
exp: 16/B55.34.5 0.508
exp: 16/C59.25.2 0.504
[Al(OCUBE)4]DFT51.72.2 0.692
[Al(O2CUBE)(OCUBE)2]DFT52.34.8 0.394
[Al(O2CUBE)2]DFT50.94.8 0.018
[Me4N]+[Al(OCUBE)4]DFT50.72.0 0.884
[AlCl(OCUBE)3]DFT63.26.7 0.535
[AlCl(O2CUBE)(OCUBE)]DFT64.87.5 0.902
a

CQ data obtained by DFT are taken as absolute values |CQDFT|.

The exact nature of the observed [AlO4] resonances is somewhat uncertain. Both the δF1 and δiso NMR shifts of all the [AlO4] resonances in products 115 fall within the range defined by the resonances 16/A–C (Figure 8), with the fitted CQ values comparable or slightly larger to 16/A–C. In many cases, a direct correspondence between the resonances in 115 and those in 16 could be established based on δF1 but the fitted δiso values failed to match the correspondences in δF1, with most [AlO4] resonances in 115 more closely matching to the resonance 16/C. The comparison of products 5 and 6 (Figure 9) provides further insights. The higher reaction temperature in 6 resulted in the reduction of the py–Al(O2CUBE)(OCUBE) resonance and an increase in the resonance assigned to [AlO4], indicating that the former is transformed into the latter, preferentially over the more sterically crowded py–Al(OCUBE)3. Moreover, the [AlO4] resonances in 6 show signs of thermal relaxation into narrower (slightly lower CQ) features with three distinct maxima, establishing a direct match to resonances 16/A–C in δF1. This implies that the formed [AlO4] sites must possess at least one bidentate CUBE and that both the transformation of [L–AlO3] and the condensation of [AlCl4] converge to the same types of pseudotetrahedral sites. A comparison to the DFT predictions affirms the presence of bidentate CUBEs as the best match is found with [Al(O2CUBE)(OCUBE)2] and less so with [Al(O2CUBE)2] (Table 2), both being indistinguishable based on δiso/CQ and most closely matching to the resonance 16/A. The better agreement of η in the case of the former (0.394) over the latter (0.018) is not insignificant as the experimental fits of all the resonances assigned as [AlO4] resulted in η ≈ 0.5 quite consistently. On the other hand, the assignment of the observed resonances as the originally sought [Al(OCUBE)4] is disfavored as the predicted |CQ| is smaller by at least 50%. While the inclusion of the weakly interacting countercation [Me4N]+ does increase η, the larger observed CQ could not be explained this way. Additionally, the assignment of any of the observed sites as [AlCl(OCUBE)3] or [AlCl(O2CUBE)(OCUBE)] (incomplete condensation) is strongly disfavored based on the much larger predicted δiso and |CQ|. The correspondence between the two candidate types of sites and the three observed maxima 16/A–C could not be further clarified; however, the thermal relaxation observed in 6 and its close match to 16/A–C would suggest that 16/A–C may simply represent a band of conformers of a single type of site, with three favored conformations of a decreasing symmetry (increasing CQ).

Figure 9

Figure 9. 27Al TQ/MAS NMR spectra of products 5 (blue) and 6 (red) (20 kHz MAS).

Further Computational Predictions of 27Al MAS NMR Parameters

Since the scope of the NMR parameter calculations went beyond the portfolio of the experimentally observed sites, further supporting correlations can be established based on the computational results only. Figures 10, 11, and 12 illustrate several important trends.

Figure 10

Figure 10. Comparison of predicted δiso for various sites based on the number of bidentate CUBEs (terminated by −OMe).

Figure 11

Figure 11. Comparison of predicted |CQ| for various sites based on the number of bidentate CUBEs (terminated by −OMe).

Figure 12

Figure 12. Comparison of predicted η for various sites based on the number of bidentate CUBEs (terminated by −OMe).

First, the isotropic chemical shifts of structures with [AlO3N] coordination spheres are systematically higher than those of the structures with the [AlO4] environment (Figure 10). This trend has been well documented in the literature and it is generally attributed to the increased p-character of Al–N bonds over Al–O. (53)
Second, the introduction of a bidentate CUBE in the [L–AlO3] type of sites results in a significant increase of δiso and η, along with a decrease of |CQ|, indicating an overall symmetrization of the environment with a shift from a unidirectional (|V33| ≫ |V22| ≈ |V11|) to a bidirectional asymmetry (|V33| ≈ |V22| ≫ |V11|) of the EFG tensor. In the case of the symmetric [Al(OCUBE)4], the presence of bidentate CUBEs does not appreciably affect δiso while |CQ| is doubled by the introduction of the first bidentate CUBE and further unaffected by the introduction of the second one. This is consistent with the idea that the initial approximate Td symmetry is first reduced to C2v by the introduction of a dominant C2 symmetry axis through the first bidentate CUBE while the dominance of the axis is unchanged with the transition to D2d by the introduction of the second bidentate CUBE. In contrast to |CQ|, η decreases monotonically to ∼0, pointing to the increasing dominance of a single direction in the EFG tensor and the significant off-main-axis symmetrization due to the appearance of a S4 axis in the transition from C2v to D2d.
Finally, the effect of the presence of a countercation is illustrated. In the geometry optimizations, the quite inert [Me4N]+ ion was moved away from the Al center to an equilibrium N–Al distance d(N–Al) = 5.025, 4.986, and 4.920 Å for Al in [Al(OCUBE)4], [Al(O2CUBE)(OCUBE)2], and [Al(O2CUBE)2], respectively. Consequently, it bears little effect on δiso and |CQ| while η is systematically increased compared to the naked anions, reflecting the additional symmetry breaking due to the proximity of a point charge. The relevant cations [py–SnMe3]+ and [Me3Sn]+ (vide infra) exhibit progressively stronger interactions with the Al–O–Si bridges at d(Sn–O) = 2.496–2.551 and 2.171–2.201 Å, resulting in d(Sn–Al) = 3.579–3.707 and 3.352–3.443 Å for the two cations, respectively. In the case of [Me3Sn]+, the Sn–O distance approaches the average value (1.987 Å) reported for anhydrous CUBE from the single-crystal X-ray diffraction measurements. (13) As a result, the NMR spectroscopic parameters are progressively more influenced in the general direction of increasing δiso and |CQ|, consistent with the increasing dominance of the cation as a symmetry-breaking element and its deshielding effect.
Additionally, Figures S42–S44 map the expected effect of the presence of residual reactive groups Al–Cl and Al–Me due to incomplete condensation. In general, δiso is increased slightly from 47.9–53.1 to 62.0–73.2 ppm by the presence of Al–Cl but more dramatically to 84.5–99.5 ppm by Al–Me (Figure S42) in consistence with the literature. (7,53) For the [L–AlXO2] sites, |CQ| is slightly decreased (L = THF, py) or unaffected (L = Et3N, TEPO) by Al–Cl while it is increased dramatically by the presence of Al–Me (Figure S43). (7) In both cases, η is increased comparably (Figure S44). Conversely, for the [AlClO3] sites, |CQ| is increased compared to the corresponding [AlO4], reflecting the lowered symmetry.
Finally, Figures S45–S47 benchmark the effect of the functional groups used to terminate the corners of the CUBEs on three selected structures: py–Al(OCUBE)3, Et3N–Al(OCUBE)3, and [Al(OCUBE)4]. The termination by −OMe used in this work was chosen as a compromise between the computationally cheap but problematic −OH (strong electric dipoles, undesirable hydrogen bonding) and the authentic but computationally expensive −OSnMe3. The benchmark data indicate that δiso (Figure S45) and η (Figure S47) are reproduced well with −OMe termination compared to −OSnMe3, with the exception of the δiso of [Al(OCUBE)4], which is significantly overestimated with −OMe compared to −OSnMe3 as well as −OH and −OSiMe3. Most importantly, the |CQ| values predicted with −OMe termination are systematically underestimated by an average of 23% compared to −OSnMe3, providing yet another explanation for the generally observed positive deviation of the experimental CQ from the DFT predictions (Figure S46). The significantly higher Sn4+ cation polarizability in comparison to Si4+ cation (56) strongly affects the EFG tensor and the local geometry of aluminum atoms, which results in the increase of experimental CQ values.

Proposed Mechanism of the Formation of [AlO4]

The experimental results established the presence of unexpected [AlO4] centers in the aluminosilicate products from the reactions of the L–AlX3 precursors with the CUBE building block. Furthermore, there appears to be a correlation between their formation and the observation of excess SnMe4. The process by which the excess SnMe4 is formed is certainly less favored than the primary condensation reactions of Al–Cl, Al–Me, and Al–Et since full condensation was reached for nearly all products (111 and 1315), yet a full transformation of all the [L–AlO3] moieties to [AlO4] was never observed. Also, the process was observed to be strongly suppressed with more extensive cross-linking at higher AlX/CUBE ratios while it showed no sensitivity to the type of the primary condensation reaction (and the type of the condensation byproduct). From this, it stands to reason that the formation of [AlO4] sites takes place in the late stages of the synthetic procedure─during the warm-up and heating─after the precursors have lost their initial identity to invariably become [L–AlO3] centers and after an extensive, hindering oligomer network may or may not have formed. Extensive cross-linking is expected only at the late stages of a step-growth condensation process. The strong sensitivity to the cross-linking extent, the observed evidence of additional connectivity (even to the point of sustaining microporosity in product of reaction 6), and the lack of evidence for the participation of the byproducts from the primary condensation reaction (SnMe3Cl, SnMe4, or SnMe3Et), constitute facts that all point to the unreacted Me3Sn–O–CUBE groups as the sole source of the additional tin removed in the form of excess SnMe4. Since a new negatively charged species is formed, the charge must be compensated by the formation of a countercation, which would represent the natural site for the liberated primary ligand (L = THF, py, Et3N, TEPO) to coordinate, as no free ligand was observed to leave the product.
Based on the aforementioned evidence, a two-step process is proposed (Figure 13). In the first step, a [L–AlO3] center is attacked by an oxygen atom of a proximal Me3Sn–O–CUBE group, acting as a nucleophile. A new Al–O–CUBE bond is established and the expelled primary ligand L is captured by the emerging cation [Me3Sn]+ in order to stabilize it.

Figure 13

Figure 13. Proposed two-step pathways leading to the transformation of [L–AlO3] to [AlO4] with one bidentate CUBE.

In the second step, the relatively free-roaming [L–SnMe3]+ cation undergoes an exchange reaction with yet another Me3Sn–O–CUBE group in the vicinity, exchanging L for Me to become the observed additional SnMe4. As the Al centers do not participate, this step should only be affected by the strength of the ligand (and the solvent) and it should be independent of any structural properties of the oligomer network. The presence of the intermediate cation [L–SnMe3]+ and the final [L–SnMe2O–CUBE]+ group is inferred based on the formation of excess SnMe4 as their presence could not be confirmed spectroscopically. It is suspected that their 1H and 13C NMR resonances are too similar to those of the Al-bound ligands and the remaining Me3Sn–O–CUBE groups to allow for resolution in the MAS NMR spectra.
Since the 27Al TQ/MAS NMR analysis supports the presence of at least one bidentate CUBE at each [AlO4] center, only pathways leading to such arrangements are considered. In the simplest case, the initial L–Al(OCUBE)3 center is attacked by a neighboring Me3SnO– group of one of the directly connected CUBEs, making it bidentate (Figure 13a). Based on a series of structure optimizations using molecular mechanics, only a single mode of bidentate binding was found feasible─along the edge of the CUBE. While this pathway may be partially responsible for the transformation and it may somewhat increase the local rigidity, it cannot explain the observed dramatic increase in the long-range connectivity (even beyond the point of the sol–gel transition in reaction 6) and the preferential disappearance of L–Al(O2CUBE)(OCUBE) over L–Al(OCUBE)3 as observed by 27Al TQ/MAS NMR (Figure 9). Therefore, a second, presumably more preferred pathway (Figure 13b) is proposed where the L–Al(O2CUBE)(OCUBE) center is attacked by a Me3SnO– group of a more distant CUBE, creating new long-range connectivity. Although the first pathway may benefit from the enforced proximity of the Me3SnO– group, the apparent preference for the second pathway can be supported by the more sterically favorable conditions at the aluminum center (only two CUBEs in the coordination sphere) and the fact that, due to its freedom, the incoming Me3SnO– group can approach from a more favorable direction. As the 27Al TQ/MAS NMR analysis could also support the presence of [Al(O2CUBE)2] sites, the participation of a third pathway starting from L–Al(O2CUBE)(OCUBE) cannot be ruled out (Figure 13c).

The Proposed Source of SnMe4 in the Reactions of [AlCl4]

The previous part of the discussion provided an explanation for the combined appearance of the [AlO4] moiety together with the unexpected formation of SnMe4 in the reactions of L–AlX3 compounds. This, however, cannot explain the observation of significant amounts of SnMe4 in reactions with [Me4N] [AlCl4] (Table 1), where no such transformation is possible and no other types of Al sites were observed (Figure 8e). A closer analysis revealed a fundamental difference─the SnMe4 is generated instead of SnMe3Cl and not in addition to it, resulting in exactly 4 eq. of byproducts liberated per Al (average connectivity 3.98). Based on the experimental observation that the missing Cl does not leave the material in any form, a synchronous trimolecular mechanism is proposed as the simplest possible explanation (Figure 14). In an otherwise regular primary condensation reaction, the activated Cl nucleophile attacks a second Me3Sn–O–CUBE group, polarizing it and, in turn, causing it to transfer Me nucleophile to the first, activated Me3Sn–O–CUBE group. It is suspected that the generated Cl–Me2Sn–O–CUBE would be similarly difficult to distinguish from CUBE as the proposed cations [L–SnMe3]+ and [L–Me2Sn–O–CUBE]+.

Figure 14

Figure 14. Proposed trimolecular mechanism explaining the evolution of SnMe4 in the condensation reaction of [AlCl4].

Conclusions

ARTICLE SECTIONS
Jump To

This work targeted the synthesis of four-coordinated aluminosilicate sites of general formulas [L–AlO3] and [AlO4], embedded in loose silicate matrices by utilizing the first step of a previously published two-step nonhydrolytic sol–gel procedure. The synthetic procedure proved to be flexible enough to accommodate a multitude of ligands, reactive groups, and solvents while it showed convergent behavior, leading only to several types of sites with well-defined 27Al NMR signatures. 27Al TQ/MAS NMR spectroscopy allowed for the observation and confident assignment of Lewis acid sites of general formulas L–Al(OCUBE)3 (L = THF, py, Et3N, TEPO) and L–Al(O2CUBE)(OCUBE) (L = THF, py). While the products always contained the targeted Lewis acid species, only TEPO–Al(OCUBE)3 could be obtained as a true single-site material. For weaker ligands, a considerable amount of unexpected [Al(O2CUBE)(OCUBE)2] and possibly [Al(O2CUBE)2] sites was also formed through a side reaction. This is contrary to the conclusions of Styskalik et al. (16) who assigned the singular, observed symmetric 27Al NMR resonances at ∼49 ppm to the Lewis acid species. From experimental and computational results described here, it is now clear that these resonances correspond to the unexpected [AlO4] species while the species of interest remained invisible in the 1D 27Al NMR spectra due to insufficient excitation. Our investigations have successfully characterized these sites and established a good correlation between the Lewis acid site symmetry and its observed CQ (4.4–15.1 MHz). Moreover, the experimentally obtained values of δiso and CQ compare well to the predictions calculated by DFT methods. The experimental CQ values are systematically increased with respect to the predictions, reflecting the additional disorder of the real-world amorphous systems compared to the fully relaxed model structures. The emergence of the synthetically undesirable [AlO4] species is attributed to the nucleophilic attack of the Me3SnO– groups of the spherosilicate building block at the aluminum site due to its Lewis acid properties.

Supporting Information

ARTICLE SECTIONS
Jump To

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.inorgchem.3c04035.

  • (S1) Synthesis and characterization data of precursors, (S2) auxiliary methods and composition calculation procedures, (S3) product synthesis and characterization data, and (S4) 27Al NMR parameter calculation data (PDF)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

ARTICLE SECTIONS
Jump To

  • Corresponding Author
  • Authors
    • Martin Kejik - Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
    • Jiri Brus - Department of NMR Spectroscopy, Institute of Macromolecular Chemistry, Czech Academy of Sciences, Heyrovskeho nam. 2, Prague CZ-16206, Czech RepublicOrcidhttps://orcid.org/0000-0003-2692-612X
    • Lukas Jeremias - Department of Chemistry and Biochemistry, Mendel University in Brno, Brno CZ-61300, Czech Republic
    • Lucie Simonikova - Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
    • Zdenek Moravec - Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech Republic
    • Libor Kobera - Department of NMR Spectroscopy, Institute of Macromolecular Chemistry, Czech Academy of Sciences, Heyrovskeho nam. 2, Prague CZ-16206, Czech RepublicOrcidhttps://orcid.org/0000-0002-8826-948X
    • Ales Styskalik - Department of Chemistry, Faculty of Science, Masaryk University, Kotlarska 2, Brno CZ-61137, Czech RepublicOrcidhttps://orcid.org/0000-0002-9998-6978
    • Craig E. Barnes - Department of Chemistry, University of Tennessee, Knoxville, Tennessee 37996-1600, United States
  • Notes
    The authors declare no competing financial interest.

Acknowledgments

ARTICLE SECTIONS
Jump To

The primary funding of this work was provided by the Czech Science Foundation (GACR Junior 20-03636Y project). We acknowledge CIISB, Instruct-CZ Centre, supported by MEYS CR (LM2023042) for the financial support of the measurements at the Josef Dadok National NMR Centre. This work was supported by the Ministry of Education, Youth and Sports of the Czech Republic through the e-INFRA CZ (ID:90254)─the computational resources provided by the IT4Innovations OPEN-24-63 project are gratefully appreciated. This publication was supported by the project Quantum materials for applications in sustainable technologies (QM4ST), CZ.02.01.01/00/22_008/0004572 by Program Johannes Amos Commenius, call Excellent Research.

References

ARTICLE SECTIONS
Jump To

This article references 56 other publications.

  1. 1
    Zeolites and Catalysis: synthesis, reactions and applications; Čejka, J.; Corma, A.; Zones, S., Eds.; Wiley, 2010.  DOI: 10.1002/9783527630295 .
  2. 2
    Haag, W. O.; Lago, R. M.; Weisz, P. B. The Active Site of Acidic Aluminosilicate Catalysts. Nature 1984, 309 (5969), 589591,  DOI: 10.1038/309589a0
  3. 3
    Ravi, M.; Sushkevich, V. L.; van Bokhoven, J. A. Towards a Better Understanding of Lewis Acidic Aluminium in Zeolites. Nat. Mater. 2020, 19 (10), 10471056,  DOI: 10.1038/s41563-020-0751-3
  4. 4
    Li, G.; Pidko, E. A. The Nature and Catalytic Function of Cation Sites in Zeolites: A Computational Perspective. ChemCatChem. 2019, 11 (1), 134156,  DOI: 10.1002/cctc.201801493
  5. 5
    Xu, J.; Wang, Q.; Deng, F. Metal Active Sites and Their Catalytic Functions in Zeolites: Insights from Solid-State NMR Spectroscopy. Acc. Chem. Res. 2019, 52 (8), 21792189,  DOI: 10.1021/acs.accounts.9b00125
  6. 6
    Xin, S.; Wang, Q.; Xu, J.; Chu, Y.; Wang, P.; Feng, N.; Qi, G.; Trébosc, J.; Lafon, O.; Fan, W.; Deng, F. The Acidic Nature of “NMR-Invisible” Tri-Coordinated Framework Aluminum Species in Zeolites. Chem. Sci. 2019, 10 (43), 1015910169,  DOI: 10.1039/C9SC02634G
  7. 7
    Kobera, L.; Czernek, J.; Abbrent, S.; Mackova, H.; Pavlovec, L.; Rohlicek, J.; Brus, J. The Nature of Chemical Bonding in Lewis Adducts as Reflected by 27Al NMR quadrupolar Coupling Constant: Combined Solid-State NMR and Quantum Chemical Approach. Inorg. Chem. 2018, 57 (12), 74287437,  DOI: 10.1021/acs.inorgchem.8b01009
  8. 8
    Yakimov, A. V.; Ravi, M.; Verel, R.; Sushkevich, V. L.; van Bokhoven, J. A.; Copéret, C. Structure and Framework Association of Lewis Acid Sites in MOR Zeolite. J. Am. Chem. Soc. 2022, 144 (23), 1037710385,  DOI: 10.1021/jacs.2c02212
  9. 9
    Chen, K.; Gan, Z.; Horstmeier, S.; White, J. L. Distribution of Aluminum Species in Zeolite Catalysts: 27Al NMR of Framework, Partially-Coordinated Framework, and Non-Framework Moieties. J. Am. Chem. Soc. 2021, 143 (17), 66696680,  DOI: 10.1021/jacs.1c02361
  10. 10
    Brus, J.; Kobera, L.; Schoefberger, W.; Urbanová, M.; Klein, P.; Sazama, P.; Tabor, E.; Sklenak, S.; Fishchuk, A. V.; Dědeček, J. Structure of Framework Aluminum Lewis Sites and Perturbed Aluminum Atoms in Zeolites as Determined by 27Al{1H} REDOR (3Q) MAS NMR Spectroscopy and DFT/Molecular Mechanics. Angew. Chemie Int. Ed. 2015, 54, 541545,  DOI: 10.1002/anie.201409635
  11. 11
    Lam, E.; Comas-Vives, A.; Copéret, C. Role of Coordination Number, Geometry, and Local Disorder on 27Al NMR Chemical Shifts and quadrupolar Coupling Constants: Case Study with aluminosilicates. J. Phys. Chem. C 2017, 121 (36), 1994619957,  DOI: 10.1021/acs.jpcc.7b07872
  12. 12
    Ghosh, N. N.; Clark, J. C.; Eldridge, G. T.; Barnes, C. E. Building Block Syntheses of Site-Isolated Vanadyl Groups in Silicate Oxides. Chem. Commun. 2004, (7), 856,  DOI: 10.1039/b316184f
  13. 13
    Clark, J. C.; Saengkerdsub, S.; Eldridge, G. T.; Campana, C.; Barnes, C. E. Synthesis and Structure of Functional Spherosilicate Building Block Molecules for Materials Synthesis. J. Organomet. Chem. 2006, 691 (15), 32133222,  DOI: 10.1016/j.jorganchem.2006.03.028
  14. 14
    Clark, J. C.; Barnes, C. E. Reaction of the Si8O20(SnMe3)8 Building Block with Silyl Chlorides: A New Synthetic Methodology for Preparing Nanostructured Building Block Solids. Chem. Mater. 2007, 19 (13), 32123218,  DOI: 10.1021/cm070038b
  15. 15
    Lee, M.-Y.; Jiao, J.; Mayes, R.; Hagaman, E.; Barnes, C. E. The Targeted Synthesis of Single Site Vanadyl Species on the Surface and in the Framework of Silicate Building Block Materials. Catal. Today 2011, 160 (1), 153164,  DOI: 10.1016/j.cattod.2010.06.029
  16. 16
    Styskalik, A.; Abbott, J. G.; Orick, M. C.; Debecker, D. P.; Barnes, C. E. Synthesis, Characterization and Catalytic Activity of Single Site. Lewis Acidic aluminosilicates. Catal. Today 2019, 334, 131139,  DOI: 10.1016/j.cattod.2018.11.079
  17. 17
    Saengkerdsub, S. Group 4 Metallocene and Half-Sandwich Derivatives of Spherosilicate: Preparation From Si8O20(SnMe3)8 and Their Olefin Polymerization Activity; University of Tennessee, 2002. https://trace.tennessee.edu/utk_graddiss/2195.
  18. 18
    Engelhardt, L. M.; Junk, P. C.; Raston, C. L.; Skelton, B. W.; White, A. H. Unidentate Nitrogen Base Adducts of Aluminium Trichloride. J. Chem. Soc. Dalton Trans. 1996, (15), 3297,  DOI: 10.1039/dt9960003297
  19. 19
    Cowley, A. H.; Cushner, M. C.; Davis, R. E.; Riley, P. E. Crystal and Molecular Structure of the 1:2 Aluminum Trichloride-Tetrahydrofuran Complex AlCl3.2THF. Inorg. Chem. 1981, 20 (4), 11791181,  DOI: 10.1021/ic50218a044
  20. 20
    Dixon, W.; Schaefer, J.; Sefcik, M.; Stejskal, E.; McKay, R. Total Suppression of Sidebands in CPMAS C-13 NMR. J. Magn. Reson. 1982, 49 (2), 341345,  DOI: 10.1016/0022-2364(82)90199-8
  21. 21
    Hayashi, S.; Hayamizu, K. Chemical Shift Standards in High-Resolution Solid-State NMR (1) 13C, 29Si, and 1H Nuclei. Bull. Chem. Soc. Jpn. 1991, 64 (2), 685687,  DOI: 10.1246/bcsj.64.685
  22. 22
    Hayashi, S.; Hayamizu, K. Shift References in High-Resolution Solid-State NMR. Bull. Chem. Soc. Jpn. 1989, 62 (7), 24292430,  DOI: 10.1246/bcsj.62.2429
  23. 23
    Frydman, L.; Harwood, J. S. Isotropic Spectra of Half-Integer quadrupolar Spins from Bidimensional Magic-Angle Spinning NMR. J. Am. Chem. Soc. 1995, 117 (19), 53675368,  DOI: 10.1021/ja00124a023
  24. 24
    Bräuniger, T.; Wormald, P.; Hodgkinson, P. Improved Proton Decoupling in NMR Spectroscopy of Crystalline Solids Using the Spinal-64 Sequence. In Current Developments in Solid State NMR Spectroscopy; Springer Vienna: Vienna, 2002; pp 6974.  DOI: 10.1007/978-3-7091-3715-4_4 .
  25. 25
    Adamo, C.; Barone, V. Toward Reliable Density Functional Methods without Adjustable Parameters: The PBE0Model. J. Chem. Phys. 1999, 110 (13), 61586170,  DOI: 10.1063/1.478522
  26. 26
    Adamo, C.; Scuseria, G. E.; Barone, V. Accurate Excitation Energies from Time-Dependent Density Functional Theory: Assessing the PBE0Model. J. Chem. Phys. 1999, 111 (7), 28892899,  DOI: 10.1063/1.479571
  27. 27
    Weigend, F.; Ahlrichs, R. Balanced Basis Sets of Split Valence, Triple Zeta Valence and Quadruple Zeta Valence Quality for H to Rn: Design and Assessment of Accuracy. Phys. Chem. Chem. Phys. 2005, 7 (18), 3297,  DOI: 10.1039/b508541a
  28. 28
    Schäfer, A.; Horn, H.; Ahlrichs, R. Fully Optimized Contracted Gaussian Basis Sets for Atoms Li to Kr. J. Chem. Phys. 1992, 97 (4), 25712577,  DOI: 10.1063/1.463096
  29. 29
    Bergner, A.; Dolg, M.; Küchle, W.; Stoll, H.; Preuß, H. Ab Initio Energy-Adjusted Pseudopotentials for Elements of Groups 13–17. Mol. Phys. 1993, 80 (6), 14311441,  DOI: 10.1080/00268979300103121
  30. 30
    Metz, B.; Stoll, H.; Dolg, M. Small-Core Multiconfiguration-Dirac–Hartree–Fock-Adjusted Pseudopotentials for Post-d Main Group Elements: Application to PbH and PbO. J. Chem. Phys. 2000, 113 (7), 25632569,  DOI: 10.1063/1.1305880
  31. 31
    Turbomole V7.5 2020, a Development of University of Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989–2007; Turbomole GmbH, since 2007. http://www.Turbomole.com/.
  32. 32
    Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A Consistent and Accurate Ab Initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010, 132 (15), 154104,  DOI: 10.1063/1.3382344
  33. 33
    Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the Damping Function in Dispersion Corrected Density Functional Theory. J. Comput. Chem. 2011, 32 (7), 14561465,  DOI: 10.1002/jcc.21759
  34. 34
    van Lenthe, E.; Snijders, J. G.; Baerends, E. J. The Zero-order Regular Approximation for Relativistic Effects: The Effect of Spin–Orbit Coupling in Closed Shell Molecules. J. Chem. Phys. 1996, 105 (15), 65056516,  DOI: 10.1063/1.472460
  35. 35
    Saue, T. Relativistic Hamiltonians for Chemistry: A Primer. ChemPhysChem 2011, 12 (17), 30773094,  DOI: 10.1002/cphc.201100682
  36. 36
    te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. Chemistry with ADF. J. Comput. Chem. 2001, 22 (9), 931967,  DOI: 10.1002/jcc.1056
  37. 37
    Van Lenthe, E.; Baerends, E. J. Optimized Slater-Type Basis Sets for the Elements 1–118. J. Comput. Chem. 2003, 24 (9), 11421156,  DOI: 10.1002/jcc.10255
  38. 38
    Chong, D. P.; Van Lenthe, E.; Van Gisbergen, S.; Baerends, E. J. Even-Tempered Slater-Type Orbitals Revisited: From Hydrogen to Krypton. J. Comput. Chem. 2004, 25 (8), 10301036,  DOI: 10.1002/jcc.20030
  39. 39
    Czernek, J.; Brus, J. On the Predictions of the 11B Solid State NMR Parameters. Chem. Phys. Lett. 2016, 655–656, 6670,  DOI: 10.1016/j.cplett.2016.05.027
  40. 40
    Blackwell, C. S. Investigation of Zeolite Frameworks by Vibrational Properties. 1. The Double-Four-Ring in Group 3 Zeolites. J. Phys. Chem. 1979, 83 (25), 32513257,  DOI: 10.1021/j100488a014
  41. 41
    Marcolli, C.; Lainé, P.; Bühler, R.; Calzaferri, G.; Tomkinson, J. Vibrations of H8Si8O12, D8Si8O12, and H10Si10O15 As Determined by INS, IR, and Raman Experiments. J. Phys. Chem. B 1997, 101 (7), 11711179,  DOI: 10.1021/jp962742d
  42. 42
    Clark, H. C.; O’Brien, R. J. Trimethyltin Perchlorate, Trimethyltin Nitrate, and Their Infrared Spectra. Inorg. Chem. 1963, 2 (4), 740744,  DOI: 10.1021/ic50008a018
  43. 43
    Baker, C.; Gole, J. L.; Brauer, J.; Graham, S.; Hu, J.; Kenvin, J.; D’Amico, A. D.; White, M. G. Activity of Titania and Zeolite Samples Dosed with Triethylamine. Microporous Mesoporous Mater. 2016, 220, 4457,  DOI: 10.1016/j.micromeso.2015.08.022
  44. 44
    Berg, R. W. The Vibrational Spectrum of the Normal and Perdeuterated Tetramethylammonium Ion. Spectrochim. Acta Part A Mol. Spectrosc. 1978, 34 (6), 655659,  DOI: 10.1016/0584-8539(78)80067-1
  45. 45
    Lefrancois, M. The Nature of the Acidic Sites on Mordenite Characterization of Adsorbed Pyridine and Water by Infrared Study. J. Catal. 1971, 20 (3), 350358,  DOI: 10.1016/0021-9517(71)90097-2
  46. 46
    Pieta, I. S.; Ishaq, M.; Wells, R. P. K.; Anderson, J. A. Quantitative Determination of Acid Sites on Silica–Alumina. Appl. Catal. A Gen. 2010, 390 (1–2), 127134,  DOI: 10.1016/j.apcata.2010.10.001
  47. 47
    Hayashi, S.; Suzuki, K.; Shin, S.; Hayamizu, K.; Yamamoto, O. High-Resolution Solid-State 13C NMR Spectra of Tetramethylammonium Ions Trapped in Zeolites. Chem. Phys. Lett. 1985, 113 (4), 368371,  DOI: 10.1016/0009-2614(85)80383-3
  48. 48
    Dixon, W. T. Spinning-Sideband-Free and Spinning-Sideband-Only NMR Spectra in Spinning Samples. J. Chem. Phys. 1982, 77 (4), 18001809,  DOI: 10.1063/1.444076
  49. 49
    Antzutkin, O. N.; Song, Z.; Feng, X.; Levitt, M. H. Suppression of Sidebands in Magic-Angle-Spinning Nuclear Magnetic Resonance: General Principles and Analytical Solutions. J. Chem. Phys. 1994, 100 (1), 130140,  DOI: 10.1063/1.466983
  50. 50
    Sau, A. C.; Carpino, L. A.; Holmes, R. R. Synthesis and 1H NMR Studies of Some Pentacoordinate Tin(IV) Complexes Derived from Triphenyltin Halides. J. Organomet. Chem. 1980, 197 (2), 181197,  DOI: 10.1016/S0022-328X(00)93565-4
  51. 51
    Irwin, A. D.; Holmgren, J. S.; Jonas, J. 27Al and 29Si NMR Study of Sol-Gel Derived aluminosilicates and Sodium aluminosilicates. J. Mater. Sci. 1988, 23 (8), 29082912,  DOI: 10.1007/BF00547467
  52. 52
    Osegovic, J. P.; Drago, R. S. Measurement of the Global Acidity of Solid Acids by 31P MAS NMR of Chemisorbed triethylphosphine Oxide. J. Phys. Chem. B 2000, 104 (1), 147154,  DOI: 10.1021/jp992907t
  53. 53
    Haouas, M.; Taulelle, F.; Martineau, C. Recent Advances in Application of 27Al NMR Spectroscopy to Materials Science A. Prog. Nucl. Magn. Reson. Spectrosc. 2016, 94–95, 1136,  DOI: 10.1016/j.pnmrs.2016.01.003
  54. 54
    Masika, E.; Mokaya, R. Mesoporous aluminosilicates from a Zeolite BEA Recipe. Chem. Mater. 2011, 23 (9), 24912498,  DOI: 10.1021/cm200706n
  55. 55
    Leonova, L.; Moravec, Z.; Sazama, P.; Pastvova, J.; Kobera, L.; Brus, J.; Styskalik, A. Hydrophobicity Boosts Catalytic Activity: The Tailoring of aluminosilicates with Trimethylsilyl Groups. ChemCatChem 2023, 15 (13), e202300449  DOI: 10.1002/cctc.202300449
  56. 56
    Dimitrov, V.; Komatsu, T. Correlation among Electronegativity, Cation Polarizability, Optical Basicity and Single Bond Strength of Simple Oxides. J. Solid State Chem. 2012, 196, 574578,  DOI: 10.1016/j.jssc.2012.07.030

Cited By

ARTICLE SECTIONS
Jump To

This article has not yet been cited by other publications.

  • Abstract

    Figure 1

    Figure 1. Synthetic strategy leading to statistically linked aluminosilicate oligomers.

    Figure 2

    Figure 2. FTIR spectra (KBr pellets) of products 1, 3, 11, 13, and 16, demonstrating the general similarity of the products, with differences in the presence of the vibration bands of the individual ligands.

    Figure 3

    Figure 3. 13C CP/TOSS MAS NMR spectra of products 1, 3, 11, 13, and 16. Spinning side-bands are denoted by asterisks.

    Figure 4

    Figure 4. 29Si MAS NMR spectra of products 1, 3, 11, 13, and 16. Spinning side-bands are denoted by asterisks.

    Figure 5

    Figure 5. 31P MAS NMR spectra of products 1215. Spinning side-bands are denoted by asterisks; a trace rotor contaminant is denoted by c.

    Figure 6

    Figure 6. 27Al {1H} MAS NMR spectra of products 1, 3, 11, 13, and 16.

    Figure 7

    Figure 7. Plots of all observed 27Al TQ/MAS NMR resonances and relevant DFT-calculated models in parameter spaces (a) δiso/CQ and (b) δF1/CQ.

    Figure 8

    Figure 8. 27Al TQ/MAS NMR spectra of (a) product 1, (b) products 3 and 5, (c) product 11, (d) product 13, (e) product 16, and (f) product 8 (20 kHz MAS).

    Figure 9

    Figure 9. 27Al TQ/MAS NMR spectra of products 5 (blue) and 6 (red) (20 kHz MAS).

    Figure 10

    Figure 10. Comparison of predicted δiso for various sites based on the number of bidentate CUBEs (terminated by −OMe).

    Figure 11

    Figure 11. Comparison of predicted |CQ| for various sites based on the number of bidentate CUBEs (terminated by −OMe).

    Figure 12

    Figure 12. Comparison of predicted η for various sites based on the number of bidentate CUBEs (terminated by −OMe).

    Figure 13

    Figure 13. Proposed two-step pathways leading to the transformation of [L–AlO3] to [AlO4] with one bidentate CUBE.

    Figure 14

    Figure 14. Proposed trimolecular mechanism explaining the evolution of SnMe4 in the condensation reaction of [AlCl4].

  • References

    ARTICLE SECTIONS
    Jump To

    This article references 56 other publications.

    1. 1
      Zeolites and Catalysis: synthesis, reactions and applications; Čejka, J.; Corma, A.; Zones, S., Eds.; Wiley, 2010.  DOI: 10.1002/9783527630295 .
    2. 2
      Haag, W. O.; Lago, R. M.; Weisz, P. B. The Active Site of Acidic Aluminosilicate Catalysts. Nature 1984, 309 (5969), 589591,  DOI: 10.1038/309589a0
    3. 3
      Ravi, M.; Sushkevich, V. L.; van Bokhoven, J. A. Towards a Better Understanding of Lewis Acidic Aluminium in Zeolites. Nat. Mater. 2020, 19 (10), 10471056,  DOI: 10.1038/s41563-020-0751-3
    4. 4
      Li, G.; Pidko, E. A. The Nature and Catalytic Function of Cation Sites in Zeolites: A Computational Perspective. ChemCatChem. 2019, 11 (1), 134156,  DOI: 10.1002/cctc.201801493
    5. 5
      Xu, J.; Wang, Q.; Deng, F. Metal Active Sites and Their Catalytic Functions in Zeolites: Insights from Solid-State NMR Spectroscopy. Acc. Chem. Res. 2019, 52 (8), 21792189,  DOI: 10.1021/acs.accounts.9b00125
    6. 6
      Xin, S.; Wang, Q.; Xu, J.; Chu, Y.; Wang, P.; Feng, N.; Qi, G.; Trébosc, J.; Lafon, O.; Fan, W.; Deng, F. The Acidic Nature of “NMR-Invisible” Tri-Coordinated Framework Aluminum Species in Zeolites. Chem. Sci. 2019, 10 (43), 1015910169,  DOI: 10.1039/C9SC02634G
    7. 7
      Kobera, L.; Czernek, J.; Abbrent, S.; Mackova, H.; Pavlovec, L.; Rohlicek, J.; Brus, J. The Nature of Chemical Bonding in Lewis Adducts as Reflected by 27Al NMR quadrupolar Coupling Constant: Combined Solid-State NMR and Quantum Chemical Approach. Inorg. Chem. 2018, 57 (12), 74287437,  DOI: 10.1021/acs.inorgchem.8b01009
    8. 8
      Yakimov, A. V.; Ravi, M.; Verel, R.; Sushkevich, V. L.; van Bokhoven, J. A.; Copéret, C. Structure and Framework Association of Lewis Acid Sites in MOR Zeolite. J. Am. Chem. Soc. 2022, 144 (23), 1037710385,  DOI: 10.1021/jacs.2c02212
    9. 9
      Chen, K.; Gan, Z.; Horstmeier, S.; White, J. L. Distribution of Aluminum Species in Zeolite Catalysts: 27Al NMR of Framework, Partially-Coordinated Framework, and Non-Framework Moieties. J. Am. Chem. Soc. 2021, 143 (17), 66696680,  DOI: 10.1021/jacs.1c02361
    10. 10
      Brus, J.; Kobera, L.; Schoefberger, W.; Urbanová, M.; Klein, P.; Sazama, P.; Tabor, E.; Sklenak, S.; Fishchuk, A. V.; Dědeček, J. Structure of Framework Aluminum Lewis Sites and Perturbed Aluminum Atoms in Zeolites as Determined by 27Al{1H} REDOR (3Q) MAS NMR Spectroscopy and DFT/Molecular Mechanics. Angew. Chemie Int. Ed. 2015, 54, 541545,  DOI: 10.1002/anie.201409635
    11. 11
      Lam, E.; Comas-Vives, A.; Copéret, C. Role of Coordination Number, Geometry, and Local Disorder on 27Al NMR Chemical Shifts and quadrupolar Coupling Constants: Case Study with aluminosilicates. J. Phys. Chem. C 2017, 121 (36), 1994619957,  DOI: 10.1021/acs.jpcc.7b07872
    12. 12
      Ghosh, N. N.; Clark, J. C.; Eldridge, G. T.; Barnes, C. E. Building Block Syntheses of Site-Isolated Vanadyl Groups in Silicate Oxides. Chem. Commun. 2004, (7), 856,  DOI: 10.1039/b316184f
    13. 13
      Clark, J. C.; Saengkerdsub, S.; Eldridge, G. T.; Campana, C.; Barnes, C. E. Synthesis and Structure of Functional Spherosilicate Building Block Molecules for Materials Synthesis. J. Organomet. Chem. 2006, 691 (15), 32133222,  DOI: 10.1016/j.jorganchem.2006.03.028
    14. 14
      Clark, J. C.; Barnes, C. E. Reaction of the Si8O20(SnMe3)8 Building Block with Silyl Chlorides: A New Synthetic Methodology for Preparing Nanostructured Building Block Solids. Chem. Mater. 2007, 19 (13), 32123218,  DOI: 10.1021/cm070038b
    15. 15
      Lee, M.-Y.; Jiao, J.; Mayes, R.; Hagaman, E.; Barnes, C. E. The Targeted Synthesis of Single Site Vanadyl Species on the Surface and in the Framework of Silicate Building Block Materials. Catal. Today 2011, 160 (1), 153164,  DOI: 10.1016/j.cattod.2010.06.029
    16. 16
      Styskalik, A.; Abbott, J. G.; Orick, M. C.; Debecker, D. P.; Barnes, C. E. Synthesis, Characterization and Catalytic Activity of Single Site. Lewis Acidic aluminosilicates. Catal. Today 2019, 334, 131139,  DOI: 10.1016/j.cattod.2018.11.079
    17. 17
      Saengkerdsub, S. Group 4 Metallocene and Half-Sandwich Derivatives of Spherosilicate: Preparation From Si8O20(SnMe3)8 and Their Olefin Polymerization Activity; University of Tennessee, 2002. https://trace.tennessee.edu/utk_graddiss/2195.
    18. 18
      Engelhardt, L. M.; Junk, P. C.; Raston, C. L.; Skelton, B. W.; White, A. H. Unidentate Nitrogen Base Adducts of Aluminium Trichloride. J. Chem. Soc. Dalton Trans. 1996, (15), 3297,  DOI: 10.1039/dt9960003297
    19. 19
      Cowley, A. H.; Cushner, M. C.; Davis, R. E.; Riley, P. E. Crystal and Molecular Structure of the 1:2 Aluminum Trichloride-Tetrahydrofuran Complex AlCl3.2THF. Inorg. Chem. 1981, 20 (4), 11791181,  DOI: 10.1021/ic50218a044
    20. 20
      Dixon, W.; Schaefer, J.; Sefcik, M.; Stejskal, E.; McKay, R. Total Suppression of Sidebands in CPMAS C-13 NMR. J. Magn. Reson. 1982, 49 (2), 341345,  DOI: 10.1016/0022-2364(82)90199-8
    21. 21
      Hayashi, S.; Hayamizu, K. Chemical Shift Standards in High-Resolution Solid-State NMR (1) 13C, 29Si, and 1H Nuclei. Bull. Chem. Soc. Jpn. 1991, 64 (2), 685687,  DOI: 10.1246/bcsj.64.685
    22. 22
      Hayashi, S.; Hayamizu, K. Shift References in High-Resolution Solid-State NMR. Bull. Chem. Soc. Jpn. 1989, 62 (7), 24292430,  DOI: 10.1246/bcsj.62.2429
    23. 23
      Frydman, L.; Harwood, J. S. Isotropic Spectra of Half-Integer quadrupolar Spins from Bidimensional Magic-Angle Spinning NMR. J. Am. Chem. Soc. 1995, 117 (19), 53675368,  DOI: 10.1021/ja00124a023
    24. 24
      Bräuniger, T.; Wormald, P.; Hodgkinson, P. Improved Proton Decoupling in NMR Spectroscopy of Crystalline Solids Using the Spinal-64 Sequence. In Current Developments in Solid State NMR Spectroscopy; Springer Vienna: Vienna, 2002; pp 6974.  DOI: 10.1007/978-3-7091-3715-4_4 .
    25. 25
      Adamo, C.; Barone, V. Toward Reliable Density Functional Methods without Adjustable Parameters: The PBE0Model. J. Chem. Phys. 1999, 110 (13), 61586170,  DOI: 10.1063/1.478522
    26. 26
      Adamo, C.; Scuseria, G. E.; Barone, V. Accurate Excitation Energies from Time-Dependent Density Functional Theory: Assessing the PBE0Model. J. Chem. Phys. 1999, 111 (7), 28892899,  DOI: 10.1063/1.479571
    27. 27
      Weigend, F.; Ahlrichs, R. Balanced Basis Sets of Split Valence, Triple Zeta Valence and Quadruple Zeta Valence Quality for H to Rn: Design and Assessment of Accuracy. Phys. Chem. Chem. Phys. 2005, 7 (18), 3297,  DOI: 10.1039/b508541a
    28. 28
      Schäfer, A.; Horn, H.; Ahlrichs, R. Fully Optimized Contracted Gaussian Basis Sets for Atoms Li to Kr. J. Chem. Phys. 1992, 97 (4), 25712577,  DOI: 10.1063/1.463096
    29. 29
      Bergner, A.; Dolg, M.; Küchle, W.; Stoll, H.; Preuß, H. Ab Initio Energy-Adjusted Pseudopotentials for Elements of Groups 13–17. Mol. Phys. 1993, 80 (6), 14311441,  DOI: 10.1080/00268979300103121
    30. 30
      Metz, B.; Stoll, H.; Dolg, M. Small-Core Multiconfiguration-Dirac–Hartree–Fock-Adjusted Pseudopotentials for Post-d Main Group Elements: Application to PbH and PbO. J. Chem. Phys. 2000, 113 (7), 25632569,  DOI: 10.1063/1.1305880
    31. 31
      Turbomole V7.5 2020, a Development of University of Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989–2007; Turbomole GmbH, since 2007. http://www.Turbomole.com/.
    32. 32
      Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A Consistent and Accurate Ab Initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010, 132 (15), 154104,  DOI: 10.1063/1.3382344
    33. 33
      Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the Damping Function in Dispersion Corrected Density Functional Theory. J. Comput. Chem. 2011, 32 (7), 14561465,  DOI: 10.1002/jcc.21759
    34. 34
      van Lenthe, E.; Snijders, J. G.; Baerends, E. J. The Zero-order Regular Approximation for Relativistic Effects: The Effect of Spin–Orbit Coupling in Closed Shell Molecules. J. Chem. Phys. 1996, 105 (15), 65056516,  DOI: 10.1063/1.472460
    35. 35
      Saue, T. Relativistic Hamiltonians for Chemistry: A Primer. ChemPhysChem 2011, 12 (17), 30773094,  DOI: 10.1002/cphc.201100682
    36. 36
      te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. Chemistry with ADF. J. Comput. Chem. 2001, 22 (9), 931967,  DOI: 10.1002/jcc.1056
    37. 37
      Van Lenthe, E.; Baerends, E. J. Optimized Slater-Type Basis Sets for the Elements 1–118. J. Comput. Chem. 2003, 24 (9), 11421156,  DOI: 10.1002/jcc.10255
    38. 38
      Chong, D. P.; Van Lenthe, E.; Van Gisbergen, S.; Baerends, E. J. Even-Tempered Slater-Type Orbitals Revisited: From Hydrogen to Krypton. J. Comput. Chem. 2004, 25 (8), 10301036,  DOI: 10.1002/jcc.20030
    39. 39
      Czernek, J.; Brus, J. On the Predictions of the 11B Solid State NMR Parameters. Chem. Phys. Lett. 2016, 655–656, 6670,  DOI: 10.1016/j.cplett.2016.05.027
    40. 40
      Blackwell, C. S. Investigation of Zeolite Frameworks by Vibrational Properties. 1. The Double-Four-Ring in Group 3 Zeolites. J. Phys. Chem. 1979, 83 (25), 32513257,  DOI: 10.1021/j100488a014
    41. 41
      Marcolli, C.; Lainé, P.; Bühler, R.; Calzaferri, G.; Tomkinson, J. Vibrations of H8Si8O12, D8Si8O12, and H10Si10O15 As Determined by INS, IR, and Raman Experiments. J. Phys. Chem. B 1997, 101 (7), 11711179,  DOI: 10.1021/jp962742d
    42. 42
      Clark, H. C.; O’Brien, R. J. Trimethyltin Perchlorate, Trimethyltin Nitrate, and Their Infrared Spectra. Inorg. Chem. 1963, 2 (4), 740744,  DOI: 10.1021/ic50008a018
    43. 43
      Baker, C.; Gole, J. L.; Brauer, J.; Graham, S.; Hu, J.; Kenvin, J.; D’Amico, A. D.; White, M. G. Activity of Titania and Zeolite Samples Dosed with Triethylamine. Microporous Mesoporous Mater. 2016, 220, 4457,  DOI: 10.1016/j.micromeso.2015.08.022
    44. 44
      Berg, R. W. The Vibrational Spectrum of the Normal and Perdeuterated Tetramethylammonium Ion. Spectrochim. Acta Part A Mol. Spectrosc. 1978, 34 (6), 655659,  DOI: 10.1016/0584-8539(78)80067-1
    45. 45
      Lefrancois, M. The Nature of the Acidic Sites on Mordenite Characterization of Adsorbed Pyridine and Water by Infrared Study. J. Catal. 1971, 20 (3), 350358,  DOI: 10.1016/0021-9517(71)90097-2
    46. 46
      Pieta, I. S.; Ishaq, M.; Wells, R. P. K.; Anderson, J. A. Quantitative Determination of Acid Sites on Silica–Alumina. Appl. Catal. A Gen. 2010, 390 (1–2), 127134,  DOI: 10.1016/j.apcata.2010.10.001
    47. 47
      Hayashi, S.; Suzuki, K.; Shin, S.; Hayamizu, K.; Yamamoto, O. High-Resolution Solid-State 13C NMR Spectra of Tetramethylammonium Ions Trapped in Zeolites. Chem. Phys. Lett. 1985, 113 (4), 368371,  DOI: 10.1016/0009-2614(85)80383-3
    48. 48
      Dixon, W. T. Spinning-Sideband-Free and Spinning-Sideband-Only NMR Spectra in Spinning Samples. J. Chem. Phys. 1982, 77 (4), 18001809,  DOI: 10.1063/1.444076
    49. 49
      Antzutkin, O. N.; Song, Z.; Feng, X.; Levitt, M. H. Suppression of Sidebands in Magic-Angle-Spinning Nuclear Magnetic Resonance: General Principles and Analytical Solutions. J. Chem. Phys. 1994, 100 (1), 130140,  DOI: 10.1063/1.466983
    50. 50
      Sau, A. C.; Carpino, L. A.; Holmes, R. R. Synthesis and 1H NMR Studies of Some Pentacoordinate Tin(IV) Complexes Derived from Triphenyltin Halides. J. Organomet. Chem. 1980, 197 (2), 181197,  DOI: 10.1016/S0022-328X(00)93565-4
    51. 51
      Irwin, A. D.; Holmgren, J. S.; Jonas, J. 27Al and 29Si NMR Study of Sol-Gel Derived aluminosilicates and Sodium aluminosilicates. J. Mater. Sci. 1988, 23 (8), 29082912,  DOI: 10.1007/BF00547467
    52. 52
      Osegovic, J. P.; Drago, R. S. Measurement of the Global Acidity of Solid Acids by 31P MAS NMR of Chemisorbed triethylphosphine Oxide. J. Phys. Chem. B 2000, 104 (1), 147154,  DOI: 10.1021/jp992907t
    53. 53
      Haouas, M.; Taulelle, F.; Martineau, C. Recent Advances in Application of 27Al NMR Spectroscopy to Materials Science A. Prog. Nucl. Magn. Reson. Spectrosc. 2016, 94–95, 1136,  DOI: 10.1016/j.pnmrs.2016.01.003
    54. 54
      Masika, E.; Mokaya, R. Mesoporous aluminosilicates from a Zeolite BEA Recipe. Chem. Mater. 2011, 23 (9), 24912498,  DOI: 10.1021/cm200706n
    55. 55
      Leonova, L.; Moravec, Z.; Sazama, P.; Pastvova, J.; Kobera, L.; Brus, J.; Styskalik, A. Hydrophobicity Boosts Catalytic Activity: The Tailoring of aluminosilicates with Trimethylsilyl Groups. ChemCatChem 2023, 15 (13), e202300449  DOI: 10.1002/cctc.202300449
    56. 56
      Dimitrov, V.; Komatsu, T. Correlation among Electronegativity, Cation Polarizability, Optical Basicity and Single Bond Strength of Simple Oxides. J. Solid State Chem. 2012, 196, 574578,  DOI: 10.1016/j.jssc.2012.07.030
  • Supporting Information

    Supporting Information

    ARTICLE SECTIONS
    Jump To

    The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.inorgchem.3c04035.

    • (S1) Synthesis and characterization data of precursors, (S2) auxiliary methods and composition calculation procedures, (S3) product synthesis and characterization data, and (S4) 27Al NMR parameter calculation data (PDF)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

Pair your accounts.

Export articles to Mendeley

Get article recommendations from ACS based on references in your Mendeley library.

You’ve supercharged your research process with ACS and Mendeley!

STEP 1:
Click to create an ACS ID

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

Please note: If you switch to a different device, you may be asked to login again with only your ACS ID.

MENDELEY PAIRING EXPIRED
Your Mendeley pairing has expired. Please reconnect