Are There Exceptions?

Yes, you are right. Unfortunately the range is to small. Thanks.

Not good, it changes phase at less than 100°C if pressure goes down !

What about Kevlar? Many of the fiber materials that we use in composites have negative CTEs.

• They've been around for some time.
  
• 
  
• 
  
• Negative CTE materials
  
• Negative thermal expansion - Wikipedia, the free encyclopedia
  
• Negative-coefficient materials can point the way to positive value ...

Thank you very much. What I am searching for is a practical way to make a reversed thermo-siphon, where the heat to go from top to bottom based on gravitational energy, with no moving parts aside from the fluid.

Well, that's quite a different thing. What else do you have to tell us?

So, what you have in mind is a perpetual motion machine?

It will never happen, no matter what the motive material's properties are.

Not even flubber will work.

I don't want to make a perpetuum mobile. The energy is there where is temperature difference, it does not matter if you use a thermo-siphon or a thermocouple to harvest it. I think that a lot of junction pairs can do the job, but is expensive to install a lot of them, with circulating pumps, etc. to spread the heat around. It will work as long as there is temperature difference, and transfer heat from hot junction to cold junction, it does not matter which one is up. What I have in mind is more like storing the heat in the underground, without big investment. When the temperature above is higher then below, the process starts by itself, working like a one way road. You can do a lot of things with that heat in the winter.

Zirconium tungstate has a fairly stable negative thermal coefficent from about few degrees kelvin to around 1000 degrees kelvin.