Simulating Autoconvective Flow to Improve the Rate Performance of High Energy Density Lithium-Ion Electrodes

Abstract

There are many design tradeoffs in lithium-ion batteries, the most notable being the loss of energy density at high power densities. New design requirements, such as low battery cost, further exacerbate these classic tradeoffs because they require thick electrodes with high loading. The power density in these dense electrodes is typically limited by Li+ depletion in the cathode electrolyte, which shuts off the cell before the full capacity can be extracted. In most industrial processes, concentration depletion can be overcome by introducing advection, but the small-scale pores and solid current collectors in batteries prevent the electrolyte from sustaining any form of bulk motion. Strategies to improve power density in thick electrodes, therefore, include increasing the electrode porosity (which reduces energy density) or increasing the electrolyte transference number (which is a difficult materials challenge).This work uses finite-element simulations to understand how battery performance would be influenced by internal advection of liquid electrolyte. We also propose a mechanism to achieve this through electro-osmotic flow of the electrolyte through charged separators. Our simulations demonstrate that the addition of a channel that circulates the electrolyte between the two ends of the cell can improve the battery’s volumetric energy density by 69% when discharged at a 4 C rate. Performance gains can be further amplified to 104% if an advective current can be maintained at 1 um/s.Figures 1(a) and 1(b) depict the simulation layout, which includes a standard lithium-ion battery with an 80-um thick lithium cobalt oxide (LCO) cathode, 50 um separator region, and 152 um LiC6 anode with 1 M LiPF6 in 1:1 EC/DEC. A modification introduces a thin channel, 1.2X the width of the cell (~340 um), that allows for electrolyte flow between the current collector ends of the cathode and anode. The multiphysics model follows a Newman approach and captures Li+ diffusion, migration, and advection in the electrolyte. Lithium diffusion and electron conduction are modeled in the electrodes, and Butler-Volmer kinetics govern the reaction kinetics. In both geometries, the LCO cathode is discharged from 4.3 V to 2.5 V.We perform simulations of the following geometries to understand the isolated effects of electrolyte continuity and advection: (1) a standard cell with no channel, (2) a cell with a channel but no flow, and (3) a cell with bulk motion of electrolyte through a channel. Figure 1(c) shows the electrolyte salt concentration within each battery at the end of discharge. The standard cell geometry notably features depletion of lithium ions ~15 um into the cathode, rationalizing the sharp drop in the cell’s discharge curve at 2.85 V (Figure 2). The addition of a channel results in migration and diffusion of lithium ions along the concentration gradient between the two current collector ends, significantly delaying ion depletion in the cathode during discharge. Introducing bulk motion further minimizes the depletion of electrolyte in the cathode and, in the extreme case of 100 um/s flow, completely eliminates ion depletion limitations.Figure 2 shows the voltage vs. energy density graph of each design, demonstrating a 104% increase in energy density for a channel-added geometry with 1 um/s of continuous bulk motion. Higher flow rates have marginal performance benefits, as a two order of magnitude increase in velocity leads to a <1% increase in energy density. Figure 2 further illustrates that, without bulk motion, the introduction of a channel results in an energy density increase of 69% relative to a normal, channel-less cell. Because the channel’s contribution to the battery size varies depending on the design of the modified cell, the discharge curves in Figure 2 do not account for volumetric changes and thus represent the maximum performance benefit that can be realized. Taking into account the channel’s volume in this particular design, however, only drops the performance gains to 82% for 1 um/s flow and 52% without any flow, demonstrating significant improvements despite the lack of robust design optimization.This simulation shows that advective flow of liquid electrolyte can improve the high rate performance of classic lithium-ion battery designs, as well as future designs with thick and dense electrodes. We believe that this can be rationally done by using porous current collectors and allowing flow through individual cell layers until the electrolyte is recirculated at cylindrical cell walls. To drive the flow internally, a charged membrane can function as an electro-osmotic pump that harnesses the inherent electric field of the electrolyte, allowing for massive gains in energy density within compact energy storage systems. Figure 1