Ion Solvation-Driven Liquid–Liquid Phase Separation in Divalent Electrolytes with Miscible Organic Solvents

Abstract

Liquid–liquid phase separation (LLPS) is a common phenomenon, but LLPS of electrolytes prepared in miscible organic solvents is rarely documented. In this study, we report four new cases of LLPS that occur in MgTFSI2 or ZnTFSI2 electrolytes in mixed organic solvents, and we use multimodal NMR methods to investigate the speciation differences between the upper and lower phases of these electrolytes. The conditions for the formation of this LLPS share four common features: cations with high charge density; anions with low charge density; a strongly coordinating solvent (indicated by its Gutmann donor number) with high compressibility; and a relatively weak coordinating solvent with low relative permittivity and high mobility. With these conditions, the cations tend to draw the strongly coordinating solvents together to form a densely packed, energetically favorable ion solvation region, while the cosolvents with low relative permittivity tend to form an ion-depleted dilute phase. Therefore, the driving force for LLPS with miscible solvents is the preferred interaction between cations and the strongly coordinating solvents that overcomes entropy loss due to phase separation. Furthermore, contact ion pairs and large ion aggregates are clearly evidenced in the dilute upper layer due to the incomplete screening of electrostatic interactions by the weaker cosolvent. This LLPS driven by cation solvation may be more common in multivalent electrolytes and is a design consideration that should not be overlooked in this context.

From Bulk to Interface: Solvent Exchange Dynamics and Their Role in Ion Transport and the Interfacial Model of Rechargeable Magnesium Batteries

“…In contrast to these results for the solvents and cosolvents, the TFSI – anions experience the fast-exchange limit in most Mg electrolytes, , so we usually do not observe well-resolved 13 C, 17 O, or 19 F NMR signals arising from free and bound anions, even at the lowest temperatures before the solutions freeze or nucleate. While Raman spectroscopy has been employed as the major tool to distinguish between free TFSI – , solvent-separated ion pairs (SSIPs), and contact ion pairs (CIPs) in these electrolytes, the assignments of these species are still under debate with conflicting reports in the literature. ,,, Therefore, in this study, we use NMR line width and relaxation analysis to qualitatively evaluate the existence of CIPs in a given composition at a given temperature.…”

“…One sulfoneethyl isopropyl sulfone (EiPS)and one phosphatetriethyl phosphate (TEP)were included in this study, and the remaining candidates were nitrogen-containing solvents: ACN, MEA, and trimethylsilyl imidazole (TMSI). From this set of electrolyte permutations, it was determined that MgTFSI 2 -3DME-DPE and MgTFSI 2 -3DME-DBE have a solubility less than 0.2 M, while 0.25–0.75 M MgTFSI 2 -3DME-MeTHF, MgTFSI 2 -3DME-DMC, and MgTFSI 2 -3DME-EMC form a liquid–liquid phase separation boundary, which is described in our previous work . For the remaining 10 solutions, we observe a variety of solvation structures and solvent exchange rates dependent on the physiochemical properties of the cosolvent.…”

“…From this set of electrolyte permutations, it was determined that MgTFSI 2 -3DME-DPE and MgTFSI 2 -3DME-DBE have a solubility less than 0.2 M, while 0.25−0.75 M MgTFSI 2 -3DME-MeTHF, MgTFSI 2 -3DME-DMC, and MgTFSI 2 -3DME-EMC form a liquid−liquid phase separation boundary, which is described in our previous work. 33 For the remaining 10 solutions, we observe a variety of solvation structures and solvent exchange rates dependent on the physiochemical properties of the cosolvent.…”