Silicon-based materials offer a wide range of benefits for Li-ion battery applications. In comparison to carbon-based anodes, higher energy densities can be reached, and compared to lithium metal, silicon-based materials offer a reduced dendrite formation at similar capacity. Moreover, silicon-based anodes enable fast-charging and significantly easier processing, which is suitable for most existing production lines. Looking at solid state batteries, challenges of current silicon technologies are the stability during cycling due to the volume expansion of the silicon causing loss of electrical contact and potentially inconsistent cell pressure. Also here, rate performance is still a more limiting factor. Furthermore, e.g. for sulfur-based solid electrolytes, a partial loss of the lithium inventory due to reactions at the Si/solid electrolyte interface is limiting the cycle life.
Amorphous SiNx (a-SiNx) is a promising next-generation anode material which tackles some of those issues from a materials perspective. It is a conversion material which irreversibly forms a matrix phase in the first cycle, consisting mainly of Li2SiN2. This means that part of the lithium is converted into an intentionally formed matrix phase, affecting the first cycle Coulombic efficiency (FCE). However, it shows greatly enhanced cycle stability for hundreds of full cycles at capacities around 1400 mAh/g with liquid electrolyte. The matrix phase formed in the first cycle can dampen electrode swelling, act as an intrinsic solid electrolyte and prevent excessive SEI growth. In solid state batteries, 1050 mAh could be shown over 100 cycles, with a loading of 7.2 mAh/cm2.
In this work, the battery performance in respect to nitrogen content and particle size is reported and an optimum in terms of capacity, stability and FCE is proposed. Submicron scale SiNx has been produced from the gas phase in a hot-wall reactor from monosilane (SiH4) and ammonia (NH3). Product stoichiometry, size and morphology e.g. controlling the nitrogen distribution in the particles (almost homogenously or Si/SiNx core/shell type) can be tuned by process conditions. In addition, we have demonstrated the scale up from a lab-scale equipment (10 g/h) to a pilot scale equipment (1 kg/h) while still obtaining the desired product. Amorphous SiNx powders with a nitrogen content up to 30 wt.%, a specific surface area of 5 – 50 m2/g and Si crystallinity below 2 % can be obtained.