This is the second of a multipart series that recaps a technical presentation by Murali Ramasubramanian, Enovix co-founder and vice president of research and development, at the International Battery Seminar & Exhibit. The first part of the series illustrated that our Enovix 3D Silicon Lithium-ion Cells with a typical core energy density of 960 Wh/l have achieved median cycle life of about 600 to 20% fade (80% capacity). Before describing how we’ve achieved this combination of high cycle life and high energy density, Murali provided some general background on the advantage and challenges of a silicon anode.

The vast majority of lithium-ion batteries use a graphite anode. Battery scientists have long known that silicon has a much higher lithium-ion intake capability than graphite. As the table below shows, replacing a graphite anode with one made entirely of silicon can increase the volumetric energy density from 800 milli-amp-hours per cubic centimeter (mAh/cc) to 2,100 milli-amp-hours per cubic centimeter.

Therefore, replacing a graphite anode with one composed of silicon would appear to be an obvious decision. Unfortunately, it’s not that simple. Incorporating a silicon anode into conventional lithium-ion battery architecture is problematic. The table below compares behavior of a graphite and a silicon anode for four critical performance characteristics in a conventional lithium-ion cell architecture.

The stability of a graphite anode in a lithium-ion battery is excellent from first cycle formation through end of cycle life. Formation expansion is minimal at about 10%. Conversely, formation expansion of a silicon anode can exceed 100%.

First cycle formation of a graphite anode is highly efficient (90 – 95%), with no need for pre-lithiation. First cycle formation is extremely inefficient (50 – 60%) with a predominantly silicon anode. While pre-lithiation can mitigate this problem, the process has proved impractical with conventional lithium-ion battery architecture.

Beyond first cycle formation, graphite anode swelling during repeated charge-discharge cycles is low (less than 10%), enabling typical cycle life of 500 to 1,000. Swelling during subsequent charge-discharge cycles of a silicon anode are relatively higher (over 20%), which results in rapid fade and significantly shorter cycle life—typically less than 100).

The greatly reduced cycle life is due to silicon particles disconnecting during discharge, causing non-uniformity. The non-uniformity causes over-discharge and silicon particle pulverization. And this causes new silicon surfaces to open in the pulverized particles which causes solid-electrolyte interphase (SEI) to regenerate leading to lithium-ion loss.

While small amounts of silicon, about 5 to 10 percent, have been added to predominantly graphite anodes in conventional lithium-ion battery architecture, this increases energy density only about 5%. Adding greater amounts of silicon has proved impractical with conventional lithium-ion battery architecture.

In the next post we’ll examine how our patented 3D cell architecture has overcome the problems of conventional lithium-ion cell architecture.