Model-based characterization for silicon-carbon blended electrodes

Summary:

In recent research for the design of better performing lithium-ion batteries, blended
electrodes have started gaining relevance, combining beneficial characteristics of different
materials. As an example, silicon graphite anodes are good candidates to increase the energy
density of lithium-ion batteries. The complexity of transport phenomena increases in blended
electrodes and therefore accurate modelling and characterization of the cell behavior is crucial
in order to define good electrode design guidelines.
In the particular case of silicon-carbon composite anodes with silicon nanoparticles, (very) low
lithium diffusivity in silicon combined with large variations of lithium diffusivity in graphite,
lead to a complex interplay among several transport mechanisms (including lithium exchange
between carbon matrix and silicon nanoparticles). Hysteresis phenomena and the competition
for lithiation/delithiation between the different active materials, contribute to quite diverse
scenarios along anode utilization rate, hindering the application of standard characterization
protocols.
To overcome this problem, a novel methodology is presented to identify relevant transport
parameters for each component of the electrode. This is accomplished using specific test
protocols adapted to the particular case of the tested materials, in this case, a proprietary
silicon-carbon composite anode.
Firstly, characteristic times of the different transport mechanisms present in silicon-carbon
composite anodes are estimated from available data in literature and experimental
measurements, highlighting the critical role of lithium exchange between the carbon matrix
and silicon nanoparticles, as well as lithium diffusion within the latter.
Then, based on the previous analysis of characteristic times, a multi-scale model (including two
spatial subscales, one corresponding to graphite and carbon-silicon particles and a second one
corresponding to silicon nanoparticles) is derived and its numerical simulation for some
relevant problems is presented. Some distinguished limit models (corresponding to particular
anode operation conditions and electrode utilization stages) are also discussed.
Finally, conclusions from the model analysis are used to propose suitable
experimental tests in order to identify relevant model parameters not easily identifiable in
standard tests.