Insights on the Buffer Effect in Blended Positive Electrodes for Lithium Ion Batteries

Abstract

An approach to improve the performance of positive electrodes for Li-ion batteries that is commonly used by companies is fabricating them using mixtures of different materials (usually 2) with complementary features1. This coexistence, when properly formulated, provides an extra design degree of freedom and gives rise to synergetic effects, allowing blends to perform better than what the rule of mixture predicts. Unfortunately, the internal dynamics of the electrodes are still poorly understood at fundamental level and the optimization of formulations for such blends still mainly depends on empirical studies. One of the important characteristics identified in blends is termed the buffer effect. This occurs when the material exhibiting fastest kinetics can receive an excess of the incoming lithium under high power discharge and subsequently, when the current stops, transfer this excess to the higher capacity material.In this study, the role and extent of this effect is explored for a set of representative materials and the resulting blends resulting from their mixtures. For the measurements we are using a special experimental model-like blended electrode setup reported by Heubner et al.2 slightly adapted to our needs. In this setup, two positive electrodes with different active material, but same active mass, are facing each other, separated by 2 separators with a disk of perforated lithium in the middle. During measurement, current is applied between an electrode and lithium while a potentiostat actively keeps the two positive electrodes at the same potential. Such setup gives us access to the current exchanged between the positive electrodes while keeping them in equal potential at all times, as would happen in a blended electrode with 50% of each material.For the study, 4 cathode active materials were used, LiNi0.5Mn0.3Co0.2O2 (NMC), LiMn2O4 (LMO), LiFePO4 (LFP) and LiFe0.4Mn0.6PO4 (LFMP), and every binary combination of them was measured. The cells after 3 C/10 formation cycles were subjected to a pulsed electrochemical test (1 minute 3C pulses followed by 10 minutes of relaxation). Top figure shows the voltage response of the cell vs. time while the bottom one shows the charge that was transferred between the materials during their relaxations (buffering charge) and while the cell current is equal to zero , for the NMC-LMO blend. We applied the same protocol for all the combinations under study and found out large variations of this buffering ability depending on the blends constituents and their individual electrochemical properties. It is clearly evident that both the magnitude and the direction (sign) of the buffering current varies in the voltage range meaning each material can act as donor or acceptor of the buffering lithium. We believe that this provides a new insight in the internal processes in blended electrodes and brings us one step closer to their rational design. References : Casas‐Cabanas et al. Isr J Chem 61, 26–37 (2021).Heubner et al. J Power Sources 363, 311–316 (2017). Figure 1