Loading...
Please wait, while we are loading the content...
Title Deconvoluting the Role of Charge in a Supramolecular Catalyst Permalink
| Content Provider | Semantic Scholar |
|---|---|
| Author | Cynthia M. Morimoto, Mariko Kapustin, Eugene A. Alzakhem, Nicola Bergman, Robert G. Raymond, Kenneth N. Toste, F. Dean |
| Copyright Year | 2018 |
| Abstract | We have demonstrated that the microenvironment of a highly anionic supramolecular catalyst can mimic the active sites of enzymes and impart rate accelerations of a million-fold or more. However, these microenvironments can be challenging to study, especially in the context of understanding which specific features of the catalyst are responsible for its high performance. We report here the development of an experimental mechanistic probe consisting of two isostructural catalysts. When examined in parallel transformations, the behavior of these catalysts provides insight relevant to the importance of anionic host charge on reactivity. These two catalysts exhibit similar host-substrate interactions, but feature a significant difference in overall anionic charge (12− and 8−). Within these systems, we compare the effect of constrictive binding in a net neutral aza-Cope rearrangement. We then demonstrate how the magnitude of anionic host charge has an exceptional influence on the reaction rates for a Nazarov cyclization, evidenced by an impressive 680-fold change in reaction rate as a consequence of a 33% reduction in catalyst charge. ■ INTRODUCTION Starting from humble beginnings, some supramolecular catalysts now rival enzymes in accelerating rates of catalyzed reactions. In contrast to typical homogeneous small molecular catalysts, self-assembled supramolecular catalysts achieve these rate accelerations by emulating the mechanisms of enzymes. Like enzymatic active sites, the microenvironments within these assemblies encapsulate substrate molecules with specificity and utilize noncovalent host−guest interactions to induce significant rate accelerations and impart remarkable selectivities. Thus, supramolecular hosts are important to study not only for the design of improved synthetic catalysts but also to advance our understanding of the governing principles that underlie enzymatic catalysis. The catalytic activity of supramolecular hosts is dictated by multiple parameters, including overall charge, cavity size, and the degree of solvent exclusion. However, it is often challenging to deconvolute the consequences of these specific features of supramolecular catalysts on reactivity, and such studies are lacking. We report here a powerful experimental approach that exploits the modularity of a reported metal− ligand supramolecular assembly, enabling the isolation of host charge to investigate its influences on reaction rates. Supramolecular host Ga-1 (K12Ga426) developed by Raymond and co-workers was chosen as a suitable candidate for the development of a catalyst with varied charge. The unusually large anionic host charge (12−) is due to four trianionic homochiral Ga(III) triscatecholate vertices, which enforce overall T symmetry upon the framework (Figure 1). Host Ga-1 effectively catalyzes a variety of transformations in polar solvents like water. Two notable examples are the 1catalyzed Nazarov cyclization of a dienol substrate as well as aza-Cope rearrangements of cationic enammoniums. Subsequent kinetics and DFT-based investigations have probed the origins of these rate enhancements, proposing several factors such as anionic host charge and constrictive binding. Thus, Received: February 10, 2018 Published: May 16, 2018 Figure 1. [Ga426] −12 tetrahedron Ga-1 and the [Si426] −8 tetrahedron Si-3. Only one ligand 2 is shown of six for clarity. Blue lines represent 2 and spheres represent metal ions. Article pubs.acs.org/JACS Cite This: J. Am. Chem. Soc. 2018, 140, 6591−6595 © 2018 American Chemical Society 6591 DOI: 10.1021/jacs.8b01701 J. Am. Chem. Soc. 2018, 140, 6591−6595 we envisioned an experimental investigation of the specific role of charge in these reactions by changing the metal vertices in Ga-1 to reduce the overall host charge while maintaining the chemical structure and geometry of the microenvironment. ■ RESULTS AND DISCUSSION To this end, we turned to Si(IV) as an alternative vertex to access an isostructural M426 host with overall reduced charge (Figure 1). Triscatecholate Si(IV) complexes are well-reported in literature, wherein the hypervalent metalloid coordination environment is pseudooctahedral, similar to that of Ga(III) in host Ga-1. To test if an M426 host with Si(IV) vertices could form, six equivalents of ligand 2 and four equivalents of tetramethyl orthosilicate were heated in DMF in the presence of tetraethylphosphonium bromide, a strongly binding template. The appearance of six aromatic resonances in the H NMR spectrum in concomitance with a diagnostic upfield shift in the resonances of one equivalent of the tetraethylphosphonium cation confirmed that the host−guest inclusion complex PEt4 + ⊂ Si-3 had formed (see Supporting Information). Conditions were then screened for the template-free synthesis of Si-3. This is often the most challenging step of accessing new cavity-bearing architectures and has previously prohibited access to related octaanionic, catalytically active hosts. In addition to the lack of the thermodynamic driving force of strong guest binding, multiple iterations of self-assembly and self-correction are required for high conversion to the desired host. Ultimately, long reaction times and elevated temperatures (60 h, 150 °C) were necessary for the clean assembly of template-free Si-3, likely due to the low lability of mechanically coupled silicon catecholate bonds. Electrospray mass spectrometry (ESI-MS) confirmed the Si426 stoichiometry and octaanionic charge of Si-3, primarily detected with various countercations in the 3− and 4− charge states (see Supporting Information). Further support for the proposed structure of Si-3 was obtained by single crystal X-ray diffraction analysis. Crystals suitable for X-ray diffraction were grown by vapor diffusion of benzene into a solution of NEt4 + ⊂ Si-3 in DMSO and measured with synchrotron radiation. Despite minor perturbations in ligand conformation, both Si-3 and Ga-1 are ideal tetrahedra and the Si(IV) and Ga(III) coordination environments are remarkably well-preserved (see Supporting Information). More importantly, these data confirm the architecture of Si-3 and demonstrate the isostructural relationship between Si-3 and Ga-1 (Figure 2). With isostructural catalysts Si-3 and Ga-1 in hand, we sought to compare their catalytic activities in order to probe the effects of host charge. Because Si-3 is designed to mirror all the features of catalyst Ga-1 except for the overall anionic character, their catalytic profiles should be identical unless the reaction mechanism is sensitive to a change in substrate charge. Two mechanistically distinct reactions were selected to profile the catalytic abilities of Si-3 and Ga-1: the aza-Cope rearrangement and the Nazarov cyclization. In the aza-Cope rearrangement of enammonium substrates accelerated by host Ga-1, the cationic substrate charge is maintained throughout the reaction, thus resulting in no change in overall charge. In contrast, the Nazarov cyclization involves protonation of a neutral substrate, leading to an increase in cationic charge within the microenvironment. Investigating the Ga-1and Si-3-catalyzed reactions in parallel should thus enable us to determine the extent to which their rate accelerations depend on anionic host charge. The aza-Cope rearrangement features a [3,3] sigmatropic shift of allylor propargyl-enammonium substrates to generate γ,δ-unsaturated iminium species via a cationic cyclic transition state (Figure 3). Under reaction conditions with catalyst Ga1, an enammonium substrate such as 4 is initially encapsulated prior to the rate-limiting sigmatropic shift. We postulated earlier that rate acceleration from Ga-1 derives from a reduction of the entropic barrier of cyclization due to constrictive binding within the host. Specifically, the proposed chairlike conformation of the encapsulated substrate resembles the sigmatropic rearrangement transition state more closely than the preferred extended conformations the substrate adopts in bulk solution. Note that the resulting iminium product 5 is also strongly encapsulated but hydrolyzes upon rapidly reversible egress to the corresponding γ,δ-unsaturated aldehyde 6. This precludes product inhibition and enables catalyst turnover. Because iminium hydrolysis was previously shown to proceed outside the host cavity, the reaction proceeds with a true retention of cationic charge inside Ga-1. One consideration when comparing the kinetics of the Si-3or Ga-1-catalyzed aza-Cope rearrangement is the difference in binding strength of the cationic substrate due to a variation in anionic charge of the catalysts. While it is likely that the association constant of 4 is smaller with octaanionic catalyst Si3, compared to that of its dodecanionic counterpart Ga-1, experiments with a model substrate show that this difference is negligible under our reaction conditions (see Supporting Information). Indeed, treatment of 4 with catalytic Figure 2. (Left) [Ga426] 12− Ga-1. (Center) Overlay of Si-3 in green and Ga-1 in red. (Right) [Si426] 8− Si-3. Counterions, solvents, and guests are removed for clarity. Journal of the American Chemical Society Article DOI: 10.1021/jacs.8b01701 J. Am. Chem. Soc. 2018, 140, 6591−6595 6592 amounts of Si-3 in DMSO-d6 immediately generated the quantitatively encapsulated host-substrate complex 4 ⊂ Si-3, evidenced by upfield resonances in the H NMR spectrum corresponding to one equivalent of 4. The subsequent reaction was tracked, and it was noted that the concentration of 4 ⊂ Si-3 is maintained at early reaction times via catalyst turnover and further encapsulation of free 4. This has been previously observed with Ga-1, where it was also shown that the aza-Cope rearrangement proceeds with first-order dependence on the host-substrate complex [enammonium ⊂ Ga-1]. We thus concluded that any differences in the association constant of 4 with Si-3 or Ga-1 have no effect at early reaction times in which the catal |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://cloudfront.escholarship.org/dist/prd/content/qt9f97x4g3/qt9f97x4g3.pdf?t=pbx6cx |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
