The separator is a thin polymer membrane, typically around nine micrometers thick, that sits between the cathode and anode. Its job sounds simple: keep the two electrodes from touching each other electrically while allowing lithium ions dissolved in the electrolyte to pass freely through its microscopic pores. But that porous architecture carries consequences. A separator with high porosity offers abundant channels for ion transport, reducing the internal resistance of the cell. A low-porosity separator provides better mechanical stability but throttles the flow of ions. For decades, engineers have treated this as a straightforward trade-off, and most studies have optimized separators in isolation from the rest of the cell.

The Korean team, led by Seungyeop Choi, Yeseo Lim, Jaejin Lim, and corresponding author Yong Min Lee, decided to ask a more subtle question: what happens when you pair separators of different porosities with electrodes whose intrinsic lithium diffusion kinetics differ dramatically? To find out, they built coin cells using a nickel-rich NCM622 cathode paired with an artificial graphite anode, and swapped in two polyethylene separators of identical thickness but very different microstructures. The high-porosity separator, dubbed HPPE, had a porosity of 58.2 percent, while the low-porosity LPPE came in at 37.2 percent. Every other design variable, from electrode composition and loading to electrolyte volume, was held constant, ensuring that any performance differences could be traced back to the separator alone.

The physical measurements told a clear story. Scanning electron microscopy revealed that the LPPE possessed a denser fiber network, which translated into a Gurley air-permeability number of 159.3 seconds per 100 milliliters, nearly three times the 58.9 seconds recorded for HPPE. Impedance measurements on stainless steel/separator/stainless steel cells confirmed the consequence: the LPPE carried roughly 60 percent more internal resistance, 0.855 ohms compared with 0.538 ohms for HPPE. Since both separators are made of the same polyethylene with equivalent wettability, the difference in ionic transport stemmed purely from microstructure, not surface chemistry.

Then came the surprise. When the cells were cycled under fast-discharge conditions, with a 0.5C charge followed by a 3C discharge, the HPPE cell clearly outperformed its low-porosity counterpart, delivering 136.0 mAh per gram of cathode material against 123.0 mAh per gram for the LPPE cell, a gap of 9.6 percent. But when the test was flipped to fast charging, a 3C charge followed by a 0.5C discharge, the two cells behaved almost identically. Rate capability tests reinforced the asymmetry: the LPPE cell’s capacity fell off a cliff above 3C during discharge, yet both cells tracked each other closely through 3C charging, with only a modest divergence appearing at 5C and beyond. Direct-current internal resistance measurements painted the same picture, showing a pronounced separator effect during discharge but only a faint one during charge.

Why would a component that sits symmetrically between the two electrodes behave so differently depending on the direction of current flow? The answer, the researchers found, lies in the electrodes themselves. Using electrochemical impedance spectroscopy combined with galvanostatic intermittent titration technique measurements across the full state-of-charge range, they calculated the chemical lithium-ion diffusion coefficients of both active materials. The result was striking: NCM622 diffused lithium at roughly 3.28 times ten to the minus twelve square centimeters per second, while graphite managed 2.28 times ten to the minus nine, nearly three orders of magnitude faster. During discharge, lithium ions must travel through the separator and then insert into the sluggish NCM622 particles, so any extra resistance imposed by a dense separator compounds the cathode’s inherent slowness. During charging, lithium ions leave the fast-diffusing graphite, and the anode’s rapid internal redistribution masks the separator’s limitations.

To test this coupling hypothesis directly, the team fabricated four cells with bilayer separators, stacking high- and low-porosity films in different orientations: HP-HP, HP-LP, LP-HP, and LP-LP, where the first layer faced the cathode and the second faced the anode. Although the HP-LP and LP-HP configurations should have similar total cell resistance, the HP-LP cell, with its porous layer adjacent to the slow-diffusing cathode, delivered better high-rate discharge capacity than the LP-HP cell. The explanation is interfacial: when a restrictive separator sits next to an electrode that already struggles to redistribute lithium, local ion accumulation and concentration polarization build up, inflating the discharge overpotential. Placing the high-porosity layer against the cathode smooths the lithium flux at precisely the point where the cell is most vulnerable.

The quantitative details sharpen the picture further. Under 3C charging, the constant-current portion of the total charge capacity was highest for the HP-HP cell at 65.9 percent and lowest for the LP-LP cell at 55.1 percent, while the HP-LP and LP-HP cells landed nearly tied at 60.4 and 59.5 percent. Under 3C discharge, however, the discharge capacity contribution ranked strictly by configuration: HP-HP beat HP-LP, which beat LP-HP, which beat LP-LP. In other words, the direction of ion flow through a porosity gradient matters, and the gradient should be arranged to favor the electrode with the weaker diffusion kinetics.

The implications reach well beyond the laboratory coin cell. The authors note that the coupling effect they identified is expected to be even more pronounced in material systems where the diffusion disparity between cathode and anode is large. High-nickel layered oxides such as NCM811 and NCA, which power many of today’s long-range electric vehicles, diffuse lithium even more slowly than NCM622 relative to graphite, making them especially sensitive to separator resistance. Olivine-structured lithium iron phosphate, with its restrictive one-dimensional diffusion channels, is another candidate for strong separator-coupled behavior. For battery designers chasing fast-charging and high-power applications, this means separator selection cannot be treated as a generic materials choice; it must be matched to the specific electrochemical fingerprints of the electrodes it separates.

The study also reframes how the field should think about so-called inactive components. A separator’s properties, the researchers conclude, may either emerge as a limiting factor or remain effectively invisible, depending on the diffusion kinetics of the surrounding electrodes. That duality helps explain why previous studies have reached seemingly conflicting conclusions about how much separators matter: the answer depends on what is on either side of the membrane. As the industry pushes toward extreme fast charging and high-power discharge in electric vehicles, grid storage, and electrified aviation, this work offers a practical design principle, orient and engineer separator porosity gradients to shield the slowest-diffusing electrode, and a conceptual one: in a lithium-ion battery, no component is truly inactive. Every layer participates in the intricate choreography of ion transport, and the weakest link is revealed only when the current flows in the right direction.

Subject of Research: Coupled effects of separator porosity and electrode lithium-ion diffusion kinetics on the fast-charge and fast-discharge performance of lithium-ion batteries

Article Title: Coupled effects of separator microstructure and active material diffusion kinetics on high-power performance of lithium-ion batteries

Article References: Choi, S., Lim, Y., Lim, J., Kang, D., Park, K. T., Hong, R., & Lee, Y. M. (2025). Coupled effects of separator microstructure and active material diffusion kinetics on high-power performance of lithium-ion batteries. Advances in Industrial and Engineering Chemistry, 1(1), Article 32. https://doi.org/10.1007/s44405-025-00033-w

Image Credits: AI Generated

DOI: 10.1007/s44405-025-00033-w

Keywords: lithium-ion batteries, separator, porosity, polyethylene separator, NCM622 cathode, graphite anode, diffusion coefficient, fast charging, fast discharging, ionic resistance, electrochemical impedance spectroscopy, high-power batteries