Researchers discover ferroelectricity at the atomic scale
phys.org
phys.org
https://phys.org/news/2018-10-validate-year-old-ferroelectri...
They claim ferroelectricity, not a new form of ferroelectricity, presumably this means the old theory applies.
I’ve been wondering if the ferromagnetic effect could be used to make mid-scale integrated logic devices at home. CMOS devices (and really just anything which depends on PN junctions) are notoriously difficult to manufacture outside of a controlled setting and often require dangerous chemicals. On the other hand, the ferromagnetic effect works on a macro scale (read: hobbyist-friendly) with fewer process steps and friendlier materials for the most basic implementations.
You can make memory devices out of it from just two materials — a toroid-shaped ferromagnetic “core” (e.g. nickel) and a wire running through the core (which can be used to polarize the core, or to read its polarization). That’s core memory, used extensively after the era of vacuum tubes but before the era of flash or silicon-based memories.
To my knowledge it was never used much for logic, but the row + column select of a core memory circuit perform an AND operation, and if you search around you can find at least a few papers proposing combinatorial logic based on those ferromagnetic cores.
It’s two materials (a ferromagnetic material and a conductor), it’s something that was actually used for a decade in industry (core memory), and you can get to 1 MHz even with macro-scale devices. Ferromagnetic devices seem like a hobbyist’s dream! The biggest missing link seems to be a way to automate their manufacturing, since threading wire through a thousand cores to make a multiplier sounds too tedious for me.