Group separations of lanthanides from minor actinides is required in the currently considered scenarios for closing of the nuclear fuel cycle. TALSPEAK is a well-known and historically first process suggested for such separations. The process is based on competitive complexation of trivalent f-group ions by an aminopolycarboxylate (such as the base of diethylenetriamine-N,N,N',N '',N ''-pentaacetic acid, DTPA) in an aqueous buffer and a dialkylphosphate (such as the base of bis(2-ethylhexyl)phosphoric acid, HDEHP) in an organic phase. Unfortunately, this method exhibits excessive sensitivity to pH and composition of the aqueous feed. In this study, we "reinvent" TALSPEAK, retaining the competitive ion binding but changing considerably the chemical implementation of the underlying general principles. The DTPA moiety is integrated into a functionalized ionic liquid (IL) that is immiscible with an organic phase containing dialkylphosphate ligands. Choline and betainium bistriflimides double as IL diluents and synthetic reagents. The integration of the aminopolycarboxylate moiety into these ILs is achieved in situ through the reactions of the cyclical dianhydride of DTPA with IL functional groups, either through the formation of a mixed dianhydride (for the betainium cation) or a diester (for the choline cation). The deprotonated DTPA-betainium conjugate forms 1:1 complexes with trivalent f-element cations whereas these metal ions form 1:2 complexes with the DTPA-choline conjugates. Large separation factors for Eu/Am partitioning between the two phases are observed, approaching 120-150 for DTPA-betainium and 250-270 for DTPA-choline. In the latter system, as in the traditional aqueous TALSPEAK, there is a characteristic "parabolic" dependence of the phase distribution ratios as a function of ionic radius that allows separations of the largest lanthanide ions. Group separations of all lanthanides from americium has been demonstrated, and a separation process that is based on this chemistry is suggested.