Precision in Metal–Organic Framework Pores: Implications in Catalytic and Structural Properties

Abstract

In nature, highly evolved enzymes feature active sites that are constructed with a high degree of precision, which allows them to facilitate challenging heterolytic bond transformations. Enzymes achieve perfect selectivity and rate accelerations through their ability to colocalize multiple functional groups into one environment. Metal–organic frameworks (MOFs) are a promising heterogenous platform to mimic these effects due to their well-ordered structures, high degree of tunability, and confined pore environments. There are still major synthetic challenges associated with imparting enzyme-like precision into MOF pores, with progress being limited by the lack of synthetic strategies to control functional group orientation and the interactions in the surrounding microenvironment. This work herein concerns progress towards building precise microenvironments in MOF pores, resulting in improved catalytic and structural properties. Chapter 1 provides an overview of the current landscape of bifunctional MOFs in catalysis. There is a particular emphasis on synthetic strategies towards bifunctional acid–base frameworks, and some perspective on their advantages and shortcomings. The chapter concludes with a brief discussion on strategies towards controlling the pore microenvironment, an area of research that has been much less explored in MOFs. In chapter 2, I establish the importance of functional group proximity in MOF pores to enable catalytic activity. In nature, enzymes use weakly basic residues in conjunction with weakly acidic residues to catalyze challenging reactions. In this work, these cooperative effects are replicated in a MOF containing bifunctional Brønsted acid–base sites. A templating strategy colocalizes acid and base sites, and this colocalization is shown to be essential for catalytic activity. Specifically, a thermolabile crosslinker containing tertiary ester and tertiary carbamate linkages is used to tether carboxylic acid and benzylamine pairs in close proximity during the synthesis of the framework. This templated material is over 4-fold more active towards the aldol condensation reaction compared to its nontemplated analogue containing randomly distributed acid and base sites. Together, this work establishes MOFs as an exciting platform that combines the advantages of heterogenous catalysts with the structural precision of enzymes. In chapter 3, I build upon this work to show that colocalization is not the only important design principle in this bifunctional system; catalytic activity is highly dependent of the structure of the active site and its surrounding pore microenvironment. Here, we show that functionalizing MOF pores with nonpolar alkyl chains mimics the beneficial effect of nonpolar solvents in suppressing the formation of off-cycle charged intermediates. Using our templating strategy, we synthesized a new bifunctional MOF catalyst containing secondary amine–carboxylic acid pairs. Kinetic measurements in acetone:hexanes solvent mixtures confirmed that the activity of our bifunctional secondary amines is highly sensitive to solvent polarity, with initial rates decreasing more than 4-fold upon switching from hexanes mixtures to pure acetone. We show that high activity can be partially recovered by incorporating nonpolar n-hexyl chains into the framework backbone that destabilize charged species. Together, this work illustrates how controlling both the local active site structure in conjunction with the surrounding pore environment can be used to increase the activity of framework catalysts. In chapter 4, I shift focus from the use of crosslinkers as active site templates towards their utility in stabilizing large pores in flexible MOF architectures. A barrier to the isoreticular expansion of flexible metal–organic frameworks is their complex breathing behavior, which can lead to pore closure upon solvent exchange and removal. In this work, chemical cross-linking is shown to stabilize the open form of a flexible aluminum framework with large 17 Å pores. The degree of flexibility and accessible surface area can be tuned by changing the concentration of the crosslinker, with the 100% crosslinked framework showing a nearly three-fold increase in BET surface area compared to the non-crosslinked framework. This drastic improvement is particularity impressive as adding functional groups to MOF pores typically reduces the surface area. This work emphasizes the versatility of crosslinker molecules beyond functional group templating, serving as a generalizable route towards accessing large pore variants of other flexible frameworks. Appendix A returns to the theme of mimicking solvent effects in MOF pores. More specifically, I provide experimental details towards the synthesis of Zr-based frameworks with Brønsted acidity and the incorporation of solvent-like functional groups. Here, I build a complete picture of the synthesis and structural characterization of a Zr-base MOF functionalized with diethylene glycol chains, and briefly cover its application in catalyzing the fructose dehydration reaction. Our results suggest that the pores of the Zr framework are too small to accommodate the addition of functional chains, thereby inhibiting the diffusion of substrates. Appendix B shifts away from applications in catalysis and to a focus on synthetic strategies towards precise Ln3+ dimers in MOFs. Here, I provide experimental details towards a synthetic route that successfully incorporates Ln3+ dimers into the lattice of a Ti-oxo framework. In this work I establish a microwave-assisted route as a milder alternative to solvothermal synthesis conditions with the goal of preserving the structural integrity of the Ln3+ dimers. Our results suggest that even with full Ln3+ incorporation, the microwave-assisted route is a robust strategy towards the formation of the framework. Spectroscopic characterization is currently underway to study the Ln–Ln interactions, and to determine their structural integrity with this synthesis method.