Silicon is often studied as an alternative to graphite for negative electrodes in lithium-ion battery technology due to its high theoretical specific capacity (3579 mAh/g [1]). However, silicon expands during lithiation, which leads to fast degradation and limited practical use of silicon-based anodes. In order to overcome these challenges a deeper understanding of the degradation mechanisms is needed. However, a common problem with electrode testing is the lack of control of factors such as thickness and loading which effectively conceals the variation in degradation behavior of silicon. A change in the silicon treatment is often concealed by more prominent factors such as loading and thickness variations. To better understand the influence of the silicon material and electrode preparation a better control of these factors is needed, i.e. improving the “signal-to-noise ratio” by improving the process control.

All components used in the electrode and the interaction between them will influence the processes towards the resulting battery cell. The current work focuses on optimizing one of these processes to improve the homogeneity of the electrode. The coating method can give severe differences in both thickness and loading properties, and it is of great importance to keep these differences as low as possible. Only cells with similar loading and thickness can give comparable results and yield better understanding of the degradation mechanisms in the silicon anodes.

In the search of a uniform electrode, the coating methods play a large role and can be optimized. For laboratory based research, tape casting is often the preferred method as it requires simple equipment and the material consumption is low. An alternative coating method is screen printing which has been applied in various industries, such as clothing, photovoltaics and printed electronics due to the low cost and reliable quality. This work compares silicon-based electrodes prepared with two coating methods; tape casting and screen printing.

A water-based mixture of 60 % silicon (0.3 µm, produced at IFE [2]), 10 % graphite (KS6L), 15 % carbon black (Super C65) and 15 % binder (CMC) was prepared and deposited on a structured copper foil with the two different coating techniques. Tape casting was used as a reference method, while parameters for the screen printing process, such as mesh size, printing speed, pressure and snap off, were varied to achieve thickness and loading variations. Issues such as slurry preparation and slurry requirements, cleaning procedure and ease of electrode preparation were compared. The electrodes were analyzed in terms of morphology (cross section and surface), resistance, adhesion testing, thickness and loading parameters as well as final cycling properties.

Initial results show a more homogeneous thickness distribution from the screen printed electrode, and the loading variations in terms of active material per area are accordingly decreased. The beneficial homogeneity in thickness distribution for the screen printing can be seen in the attached figure, while the electrode coated via tape casting shows a higher variation in thickness over the entire electrode. The improved control of thickness and loading for the screen printed electrodes led to an improved understanding of the degradation mechanism of the electrodes as the noise in cycleability results was reduced.

[1] U. Kasavajjula et al. J. Power Sources 163 (2007) 1003-1039

[2] H. F. Andersen et al. ECS Trans. 62(1) (2014) 97–105