(De)coding SABRE of [1-13C]pyruvate

Abstract

Hyperpolarized pyruvate is a key molecular probe for biomedical imaging but achieving efficient ¹³C signal amplification by reversible exchange (SABRE) enhancement remains elusive. Here, we report a comprehensive study integrating catalyst design, systematic experimentation, and advanced theoretical modelling. We synthesized and tested seven Ir-NHC catalysts, spanning the main families of carbene ligands, including previously unexplored variants for pyruvate SABRE. IMes remains the benchmark, delivering ∼3% ¹³C polarization at 50% parahydrogen enrichment (extrapolated to ∼10% at 100% parahydrogen), but structurally distinct alternatives such as IPr and SIPr achieve only ∼20% lower performance, allowing detection of natural abundance ¹³C signals in one scan at 1.4 T. DFT calculations indicate that J-couplings between hydrides and ¹³C nuclei are similar across binding geometries and catalysts, indicating that exchange dynamics-rather than coupling strength-govern polarization efficiency. To clarify this, we performed variable-temperature experiments on both free and catalyst-bound pyruvate. To interpret the observed trends, we developed a detailed mechanistic model that incorporates species concentrations, parahydrogen fraction, exchange kinetics, spin couplings, and relaxation. By leveraging molecular symmetry to reduce Liouville space dimensionality, the model serves as an efficient and predictive tool for SABRE systems. Finally, we apply this framework to devise a SABRE protocol based on a temperature jump designed to selectively enhance the free pyruvate signal. This approach yields an ∼30% increase in free pyruvate polarization at the expense of Ir catalyst-bound forms, with potential for further optimization. Altogether, our work bridges molecular design, theoretical modelling, and protocol development, offering a blueprint for the rational optimization of SABRE hyperpolarization of pyruvate and beyond.