Synthesis and Electrochemical Characterization of Magnesium-Ion Battery Electrolytes

Abstract

Magnesium-ion batteries are thought to be a viable successor to the current Li-based systems owing to high elemental abundance (Mg, 2.3%), environmental benignity, and a two- fold volumetric capacity increase over lithium metal. However, moving beyond today’s lithium- ion technology requires advances in fundamental electrolyte science as magnesium-based congeners of the typical salts and solvents used in Li-ion electrolyte solutions result in electrode passivation, which ceases any reversible battery cycling.

Initial magnesium-ion electrolyte solutions were composed of Grignard reagents such as PhMgCl, due to their ability to electrochemically deposit magnesium. Unfortunately the widespread use in battery manufacturing remains hindered by air and moisture sensitivity. With this limitation in mind, researchers began to investigate phenolate-based electrolyte solutions, which show reduced deleterious reactivity with air and moisture. With the systematic substitution of both the nature and number of substituents about the aromatic ring within the standard 0.5 M RPhOMgCl and 0.25 M AlCl3 in tetrahydrofuran (THF) solvent, two design principles were determined: (1) electron withdrawing groups increase anodic stability and (2) increasing steric bulk of the organic moiety improves solution conductivity. While focusing entirely on sterics, an electrolyte solution starting from 2,4,6-trimethylphenol exhibits high solution conductivity (2.56 mS/cm).

Guided by the aforementioned design principles and density functional theory calculations, a series of highly soluble fluorinated alkoxide-based electrolytes were prepared, examined electrochemically, and reversibly cycled. Most notably, the electrolyte composed of 1.2 M ((CF3)2CH3)COMgCl and 0.2 M AlCl3 in THF (F6-t-butoxide) exhibits high anodic stability (3.2 V vs Mg2+/0) and solution conductivity (3.5 mS/cm). Additionally, the F6-t-butoxide solution shows excellent galvanostatic cycling and capacity retention (94%) with more than 300 h of cycle time while employing the standard Chevrel phase-Mo6S8 cathode material. Overall, suggesting that fluorinated alkoxide-based electrolytes are promising candidates for practical high voltage magnesium-ion batteries.

The resulting electrode-electrolyte interface was further examined for three magnesium- ion electrolyte solutions to understand the influence an in situ deposited Mg layer has on continued electrodeposition. Comparing the deposition morphologies of three previously reported electrolyte solutions, the F6-t-butoxide solution shows the most uniform and crystalline deposits growing along [100], normal to the electrode surface. Overall, this work illustrates that sporadic deposition and lower solution conductivities of the phenolate- and Grignard-based electrolytes hinder the deposition rate.

Additionally, a thorough investigation of possible decomposition products formed from a series of solutions containing PhMgCl in THF was performed. Solutions containing no additional salt, as well as those containing added MgCl2 or Al(OPh)3, demonstrate an initial anodic current response between 2 and 4 V (vs Mg2+/0) which results in the adsorption of an aromatic polymer decomposition product on the working electrode surface. Once this adsorbed layer is formed, electrode impedance increases by ~100 Ω with no additional growth of the insulating film. A phenyl radical is deemed the culprit, as adding a phenyl anion source (Ph−) insulates electrodes even for originally non-passivating electrolyte solutions. Overall, this work highlights many of the key advancements in electrolyte design and magnesium electrodeposition as well as presents many of the remaining challenges for continued development of magnesium-ion battery technology.