Mimicking the graphene oxide structure in solutions by interaction of Fe(III) and Gd(III) with model small-size ligands. The NMR relaxation study

Abstract

Despite long history and recent progress, the chemical behavior of graphene oxide (GO) is not fully understood. One of the remaining open questions is an extremely high cation exchange capacity, which cannot be explained in the frames of the commonly accepted Lerf-Klinowski structural model. The interaction of GO with metals, termed as sorption, is usually considered as physisorption, or as non-specific electrostatic attraction. Recently, we demonstrated that in fact this interaction has the coordinate-covalent bonding nature. In this work, by the means of the NMR relaxation, we investigated the interaction of Fe(III) and Gd(III) ions with several small molecule size chelators and with linear polymers in order to possibly mimic the GO structure. The experiments were conducted in the broad pH range, and at the different metal/chelator ratios. From the two tested metal ions, Fe(III) has the stronger affinity toward GO due to its higher charge density. GO binds Fe(III) with formation of strong high-relaxivity complexes, which do not undergo hydrolysis even in basic conditions. Numerous model small-size molecules, tested to mimic the GO structure, did not demonstrate the same effect on the relaxation of the Fe(III) and Gd(III) solutions, as does interaction with GO. From the tested linear polymers, only polyacrylic acid (PAA) can bind Fe(III), however, even PAA, consisting from carboxyl groups, cannot prevent Fe(III) from hydrolysis already in neutral solutions. Subsequently, carboxyl groups were excluded from the species in GO structure, responsible for the strong binding of meatal ions. We conclude that there are no preexisting sites or functional group fragments in GO structure, capable to strongly bind transition metal cations. The binding sites are formed from the existing structure upon the reaction with metal cations, to afford formation of the metal/GO coordination compounds. The most probable change in the GO structure is the formation of enolates from tertiary alcohols with rupture of the C[sbnd]C bonds. The newly reported data provide additional experimental evidence for the Dynamic Structural Model of GO and broadens the areas of its applicability.