To be or not to be a diode? It depends on temperature gradient

I was curious how this worked, so I looked at the paper.

My first thought was just about how semiconductors worked in general. You heat something up, you get electrons in excited states, and this creates pairs of N carriers and P carriers. The N carriers in common semiconductors used in industry have more mobility than P carriers, is that what’s causing the effect? However, when I started thinking more about this idea, it made less and less sense.

Turns out there’s a phase change here! The phase change is a reversible phase change alters the crystal structure enough to change between majority N and majority P type carriers. The temperature gradient is necessary to put the N type and P type phases close enough together.

>"Then, with appropriate design,

a single device could function as a resistance, a diode, or a transistor,

depending on the temperature gradient and the position of the connections."

This would be very interesting should it ever be created -- a single electronic component -- that could (potentially!) act like all other electronic components...

One question might be how to introduce variable capacitance into such a component, again, should it ever be created...

Normally you want devices to be independent of temperature, so perhaps this effect can play a role in offsetting temperature effects ... somehow ...

This looks like a really interesting finding, probably it can be used in quantum or servers hardware designs where a lot of variables are under control like temperature. I think in quantum computers it could help to minimize the size of them and make those hardware more resistant to certain physical variables to reach it beyond labs.