All-solid-state batteries utilizing ceramic electrolytes have several advantages compared to the conventional batteries using liquid electrolytes, such as enhanced safety and high temperature operation. Additionally, the compressive pressure required for operation of all-solid-state batteries has been shown to suppress capacity decay which originates from the expansion of high capacity electrode materials such as silicon (Si). Si is an attractive anode active material due to its high theoretical capacity and natural abundance. However, the volume expansion (400%) during lithiation and the following mechanical pulverization and active material isolation leads to fast capacity decay hindering more widespread commercialization of this material.

Recently we reported several articles utilizing Sn as an electrochemically active and mixed conductive matrix to host Si. The highly reversible Si-Sn hybrid anode, which was prepared by simple powder mixing, could maintain 700 mAh/g electrode specific capacity for 50 cycles, and the utilization of both Sn and Si was confirmed through electrochemical analysis and XRD. The ductile nature of Sn was pointed out as one of the important factors as it can adapt well to the elevated pressure originating from volume expansion of the active materials during lithiation and thus act conformally on Si leading to greater reversibility. Especially the fact that Sn is lithiated prior to Si, indicated that Sn would expand first, preventing extensive expansion and the following mechanical degradation of Si particles inside the electrode matrix.

In this study, we present a more advanced method to prepare Si-Sn hybrid electrodes, using sputtering technique. Multilayered electrodes were prepared by sequentially sputtering Sn and Si, which eventually turned out to result in a more homogeneous mixing between the two active materials during cycling, therefore leading to Sn more efficiently stabilizing the electrochemical performance of Si. The morphology of the electrode during cycling was monitored by ex-situ FIB-TEM analyses, which showed cracking of the individual layers and mixing of Sn and Si.

Sn was first sputtered as a single layer anode, and the cycling test showed a voltage profile with clear plateau regions corresponding to the four lithiation phases of Sn, and very good cycling stability showing specific charge and discharge capacities of ~850 mAh /g (at 55^th cycle). On the other hand, Si sputtered as a single layer did not perform well as an anode in terms of cycling stability. The Si-sputtered anode showed fast capacity decay with 33% of irreversible capacity loss in the first cycle, and at the 50^th cycle the charging capacity was only 18% of the initial charging capacity.

When Sn and Si are sputtered sequentially in a layered fashion, the cycling stability significantly improved. Especially, the sputtered Sn layer between the stainless steel substrate and the Si layer appeared to be critical for stable cycling. The Sn-Si-sputtered anode showed much improved cycling stability compared to the Si-only anode, and outperformed the Si-Sn-anode (sputtered Si first, then Sn) as well. The Sn-Si-sputtered anode showed 0.956 mAh total charging capacity at the 40^th cycle, which corresponds to 2926 mAh/g of Si specific capacity, assuming that the Sn layer reached maximum capacity each cycle. The Si-Sn-sputtered anode on the other hand, reached 0.607 mAh total charging capacity at the 58^th cycle, which corresponds to 1453 mAh/g Si specific capacity. The thin layer of Sn (≈0.32 µm thickness, by calculation with mass and density) which is located between the Si layer (≈0.76 µm) and stainless-steel substrate is expected to act as a mixed conductive layer which improves and retains the interface between Si and the substrate during cycling.

The Sn-Si-Sn trilayer-sputtered anode performed with excellent cycling stability, maintaining 1.44 mAh total charging capacity at the 51^st cycle, which corresponds to 3860 mAh/g Si specific capacity. Compared to the Sn-Si bilayer anode, the trilayer anode shows less irreversible capacity loss in the first cycle. This effect was attributed to the top Sn layer which is expected to prevent loss of contact due to Si pulverization which happens during the first lithiation cycle.