Dendrites: Zinc and Lithium

I just got back from the spring Electrochemical Society conference in Chicago, where I was presenting in a session on Dendritic Growth and Interface Stability. Several people showed in operando images of dendrites growing in real time. One in particular I liked a lot was by¬†Layla Mehdi at PNNL … she has a specialized test cell to observe lithium dendrites by transmission electron microscopy. I’ve done a lot of work on dendrites, and what I care most about is why dendrites form. While I was watching all these great movies¬†of dendrites growing, I wondered something I often wonder: why do some dendrites look like they grow from the tip, and some from the bottom?

Dendrites can be the result of multiple mechanisms. They’re often a mass transport effect, but sometimes not. If you flow electrolyte over dendrites, you can directly observe the kinetic growth limit at their tips (plug for my paper about about that), but the non-growing locations aren’t really mass-transport starved … rather, they are in locations of low overpotential. So, that all makes sense, but still sometimes I’m baffled. Consider the two movies below …

The first one (this is my data from the paper above btw) shows zinc dendrites growing in a 1 mm wide flow channel, with alkaline electrolyte flowing from top to bottom. Essentially this is a model of a zinc flow battery. Growth here is potentiostatic (2.5 V) with oxygen generation at the electrode on the right. The important thing is watch the tips. That’s where the growth is. The bases of the dendrites are static.

Here is an analogous movie of a lithium dendrite growing in a flow channel. This was taken by Owen Crowther (paper here), who’s now working at Eagle Picher. If you focus on the dendrite tip, it doesn’t look like there’s any action there. Growth seems to come from the base (or middle). But that would be the location of lowest overpotential and lowest mass transport, so there must be something catalytic happening. Perhaps it’s because of a locally thin SEI layer.