(405f) Formation of Percolating Carbon Networks in Battery Processing and Their Effects on Electrode Performance

The increasing demand for energy storage has led to continual improvements in materials for battery electrodes. However, the overall processing of these materials into devices and the need for composite electrodes has remained essentially the same. Currently, the optimal processing conditions are determined through time-, labor-, and capital-intensive trial and error. Determining the fundamental particle and electrochemical physics that result in high-performance composite electrodes is so far an untapped field of potential. Our work strives to fully understand from a rheological perspective the interactions of components in electrode slurries. These insights will allow the manufacturing process of new materials to be accelerated, resulting in faster, more efficient implementation of new materials into effective electrodes.

The mixing, coating, drying and calendaring of active material, conductive additive, and polymer binder in solvent affect the final battery microstructure. Although it is well-known that the electrode slurry microstructure impacts the final battery performance, there is little agreement on what constitutes a favorable microstructure [1,2]. Slurries may be broadly divided into two rheological categories, fluids and gels. In a fluid-like system, suspended particles typically interact via hydrodynamic forces. In a colloidal gel, particles form a percolating network that extends across the volume of the system through surface or depletion interactions.

In this work, we study a model system of lithium nickel manganese cobalt oxide (NMC), carbon black, polyvinyldifluoride (PVDF), and 1-methyl-2-pyrrolidinone (NMP) to determine the effects of slurry microstructure on battery performance. Slurries are characterized using small-amplitude oscillatory shear (SAOS) in which a constant strain is applied at increasing angular frequencies and the viscous and elastic components of the stress are determined as a function of frequency (frequency sweep). Figure 1 compares the frequency sweeps for a fluid system and a gel system formed with micron-sized and nano-sized active material, while maintaining constant PVDF molecular weight, carbon black, and mixing parameters. A fluid system is depicted by Gâ? larger than Gâ??. Gel systems are typically observed when Gâ?? dominates.

Suspended micron-sized active material does not form a gel in the presence of polymer binder. However, the addition of a critical volume fraction of free nano-sized carbon additive causes the suspension to gel. We find that the critical gelation volume fraction of free carbon black (40nm) is equal to 0.5 vol%. The typical quantity used in commercial battery slurry recipes is 2 vol% [3]. The amount of free carbon is highly dependent on the commonly used dry-mixing step, where active material and conductive additive are blended together. It has been shown that dry â??mixing causes conductive additive to coat the active material reducing the amount of free carbon in the suspension. In other words, in the absence of dry-mixing, all the carbon is above the critical concentration and readily forms a gel, i.e. percolating network. However, depending on the dry-mixing conditions, enough carbon is coated on the active material such that the amount of free carbon is too low to form a colloidal gel. By varying the dry mixing step of slurry preparation, a wide range of slurry microstructures can be achieved without varying the total carbon concentration. Contrary, nano-sized active material readily forms a gel due to its size and large volume fraction in the electrode slurry.

Slurries with characterized microstructures are coated, dried, and calendered into electrodes in order to relate slurry microstructure to electronic conductivity, life cycle and rate capability of the processed electrodes. Measurements of the electronic conductivity and battery performance show that gel slurries form electrodes with higher initial conductivity and lower porosity. Electrodes from gel slurries also cycle longer and have higher capacities than fluid slurries. We hypothesize that the continuous carbon network is necessary to provide sufficient electronic conductivity. Future work will investigate the dependence of performance and electrode tortuosity on slurry microstructure and the effects of drying parameters on battery performance.

References

[1]W. Bauer, D. Noetzel, Ceram. Int. 40 (2014) 4591.

[2]G.-W. Lee, J.H. Ryu, W. Han, K.H. Ahn, S.M. Journal of Power Sources 195 (2010) 6049â??6054.

[3] T. Marks, S. Trussler, A. J. Smith, D. Xiong and J. R. Dahn, J. Electrochem. Soc. 2011, Volume 158, Issue 1, Pages A51-A57.