Search for: All records

Abstract Carbon dioxide capture technologies are set to play a vital role in mitigating the current climate crisis. Solid‐state¹⁷O NMR spectroscopy can provide key mechanistic insights that are crucial to effective sorbent development. In this work, we present the fundamental aspects and complexities for the study of hydroxide‐based CO₂capture systems by¹⁷O NMR. We perform static density functional theory (DFT) NMR calculations to assign peaks for general hydroxide CO₂capture products, finding that¹⁷O NMR can readily distinguish bicarbonate, carbonate and water species. However, in application to CO₂binding in two test case hydroxide‐functionalised metal‐organic frameworks (MOFs) – MFU‐4l and KHCO₃‐cyclodextrin‐MOF, we find that a dynamic treatment is necessary to obtain agreement between computational and experimental spectra. We therefore introduce a workflow that leverages machine‐learning force fields to capture dynamics across multiple chemical exchange regimes, providing a significant improvement on static DFT predictions. In MFU‐4l, we parameterise a two‐component dynamic motion of the bicarbonate motif involving a rapid carbonyl seesaw motion and intermediate hydroxyl proton hopping. For KHCO₃‐CD‐MOF, we combined experimental and modelling approaches to propose a new mixed carbonate‐bicarbonate binding mechanism and thus, we open new avenues for the study and modelling of hydroxide‐based CO₂capture materials by¹⁷O NMR. 

A palladium diphosphine pincer complex H3(PNNNP-PdI) has been encapsulated in the benzotriazolate metal-organic framework MFU-4l-OH ([Zn5(OH)4(btdd)3], btdd2− = bis(1,2,3-triazolo)dibenzodioxin), and the resulting materials were investigated as Lewis acid catalysts for cyclization of citronellal to isopulegol. Rapid catalyst immobilization is facilitated by a Brønsted acid–base reaction between the H3(PNNNP-PdI) benzoic acid substituents and Zn–OH groups at the framework nodes. Catalyst loading can be controlled up to a maximum of 0.5 pincer complexes per formula unit [PdI-x, Zn5(OH)4−nx(btdd)3(H3−nPNNNP-PdI)x x = 0.06–0.5, n ≈ 2.75]. Oxidative ligand exchange was used to replace I− with weakly coordinating BF4− anions at the Pd–I sites, generating the activated PdBF4-x catalysts (x = 0.06, 0.10, 0.18, 0.40). The Lewis acid catalytic activity of the PdBF4-x series decreases with increasing catalyst density as a result of the appearance of mass transport limitations. Initial catalytic rates show that the activity of PdBF4-0.06 approaches the intrinsic activity of a homogeneous PNNNP-PdBF4 catalyst analogue. In addition, PdBF4-0.06 exhibits better catalytic activity than the metallolinker-based MOF Zr-PdBF4 and was not subject to leaching or catalyst degradation processes observed for the homogeneous analogue.