Abstract
Fluorine-19 MRI is increasingly being considered as a tool for biomolecular imaging, but the very poor sensitivity of this technique has limited most applications. Previous studies have long established that increasing the sensitivity of (19)F molecular probes requires increasing the number of fluorine nuclei per probe as well as decreasing their longitudinal relaxation time. The latter is easily achieved by positioning the fluorine atoms in close proximity to a paramagnetic metal ion such as a lanthanide(III). Increasing the number of fluorine atoms per molecule, however, is only useful inasmuch as all of the fluorine nuclei are chemically equivalent. Previous attempts to achieve this equivalency have focused on designing highly symmetric and rigid fluorinated macrocyclic ligands. A much simpler approach consists of exploiting highly fluxional lanthanide complexes with open coordination sites that have a high affinity for phosphated and phosphonated species. Computational studies indicate that Ln(III)-TREN-MAM is highly fluxional, rapidly interconverting between at least six distinct isomers. In neutral water at room temperature, Ln(III)-TREN-MAM binds two or three equivalents of fluorinated phosphonates. The close proximity of the (19)F nuclei to the Ln(III) center in the ternary complex decreases the relaxation times of the fluorine nuclei up to 40-fold. Advantageously, the fluorophosphonate-bound lanthanide complex is also highly fluxional such that all (19)F nuclei are chemically equivalent and display a single (19)F signal with a small LIS. Dynamic averaging of fluxional fluorinated supramolecular assemblies thus produces effective (19)F MR systems.