WEEBIT NANO LTD (ASX:WBT) - Weebit - 2023 and beyond, page-550 - HotCopper | ASX Share Prices, Stock Market & Share Trading Forum

Originally posted by Flectional: ↑

Secret sauce
weebit made it clear relatively recently they had to stop discussing the 'new' sauce in emails and phone communications to protect the latest development on the Leti / wbt test bench

this paper 'by others' in 2020 I believe is significant as one of wbts' /Leti's recent patents establishes this exact manufacturing process using halogenated plasma etcing - unique in this process with these materials as i understand it
TiN/Ti/TiO2/SiOx/Si Resistive Random Access Memory
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7559005/relevant patent - initially I missed the significant detail in this patent

the halogenated plasma etching process increases the area of the siox contact face by 'roughing' it out and bonds successfully to the titatanium oxide without oxidisation ...underlined significance is because in past manufacturing processes by others under this 'annealing' bond process combined with the etch process...oxidisation occurs and 'shorts out' the cell only discovered under test conditions by electrical verification ( this is the issue the cousin has consistently been unable to overcome with the pcmo manufacture )

those interested can check out the patent description for themselves on wipo as its pretty lengthy and below is just the confirmation of the breakthrough for weebit/Leti partnership

massive breakthrough in neuromorphic reram soc / noc innovation - expect to hear more on this across 2023/24

patent description:

happy to be corrected in this dive into the process

yet another moat for this exceptional company
weebit for the win

Expand

Great post. Thanks for sharing.

I noticed recently, Coby letting slip on a number of occasions that metal atoms were involved in WBTs cell filament. I remember thinking it was a slip of the tongue as I understood WBT's filament to be formed purely from oxygen depletion of SO_x. Switching between 0 & 1 binary states then involving oxygen migration to/from the filament region (differentiating it from most other filamentary ReRAMs that used a metallic filament and metal atom migration to make & break the filament). This patent perhaps explains it, but there is no mention of metallic atoms playing a part in the filament.

The patent talks of 2 oxide layers, the "first oxide" also referred to as the "active layer" of SO_x and the "second oxide", a metal oxide layer located between the metal layer and the active SO_x layer (through which the filament is formed). They mention that the metal layer is still acting as the oxygen reservoir so there must be oxygen migration through the second (metal) oxide layer from the active layer.

It could be that there is some oxygen depletion also occurring in the second oxide, and that Coby was referring to this. However, I wonder also if perhaps they have noticed some metal atom migration into the filament region during the forming process. If this happens it could form a more conductive filament. A little would go long way in this situation, due to the free valence electrons of the metallic atoms. The filament could thus become semiconducting with the metal atoms acting as an N-type dopant. The metal atoms could remain stationary after forming, with the mechanism of binary state change still being purely oxygen migration. Just a thought.