Ion correlations and transference numbers in polyelectrolyte solutions for Li-ion batteries

Nonaqueous polyelectrolytes have been presented in recent years as promising high cation transference number (t₊) alternatives to conventional Li-ion battery electrolytes: it is thought that by slowing the motion of the anions via attachment to a polymer chain, the cation will carry the majority of the electrolyte current. Experimental characterization of t₊ in these polyelectrolytes via the Nernst-Einstein (ideal solution) approximation corroborates this intuition, yielding t₊ values approaching unity. Here, we present the Onsager transport framework as a means of rigorously computing the transference number, without neglecting non-idealities, from molecular dynamics simulations. We demonstrate for a range of polyelectrolyte solutions that the Nernst-Einstein assumption does not yield a physically meaningful approximation for polyelectrolyte transport due the substantial anion-anion and cation-anion correlations in these systems, and that t₊ is generally lower in a polyelectrolyte than in a conventional battery electrolyte. This work has thus demonstrated critical flaws in the standard methods for experimentally characterizing transport in polyelectrolytes and has dispelled the notion that polyelectrolytes may yield promising transport behavior for next-generation batteries.