Does ambient day/night temperature change have significant energy generation potential Is there significant energy to capture in daily day/night temperature change? For instance, I see from posts such as the following that there are materials with varying coefficients of linear thermal expansion. Would not some have enough power to be useful? We know that water, upon freezing, can lift sidewalks (but I presume that the magic range from 4C to 0C, wherein I see it gets this magic power, is an outlier).
Thanks very much for entertaining this layman's question, and I hope this doesn't "generate discussion" because I know how antithetical that is to the Stack philosophy.
What kinds of materials contract the most in cold temperatures?
 A: Significant (and practical) energy capture - probobly not.   Indirect energy capture, like warming water during the day time and then using that water to warm your home at night - sure.  Pretty sure that's already done in a few places.
But electricity generation from day to night time temperatures - that's hard.   The most practical way (that I can think of) is you need something that readily freezes or liquifies at night and readily turns into a gas during day and that can be used to push turbines.   Energy output is limited to how much freezes at night and how fast it expands into gas during the day.   That's the principal behind Ocean Thermal Energy (OTEC), where in the warm ocean you have surface temps of maybe 27 or 28 degrees C and in the deep ocean not too far down, close to 4 degrees C and you run a contained liquid through it that condenses in the 4 degree water and expands back into a gas at the surface and pushes a Turbine to generate electricity.   It's an idea that's been on the drawing board since the 70s but it's not practical or cost efficient.
Ice expansion is also difficult as a form of energy generation.  When Ice forms it expands with great force but very little  volume, so you can lift heavy objects but not lift them very far.   That's not convenient for energy generation.   The expansion of water from 4 to 0 degrees that you mention isn't enough expansion to generate energy from at all.   The expansion from water to ice might be, but that's very difficult.   Significant expansion, like water into steam, is far easier to work with as an energy source, as is consistent movement, like a waterfall.   
It's a cool idea (pun not intended), but hard to do, and not highly practical or cost efficient.
A: I am re-requesting that someone try to answer my question after now seeing that MIT has done exactly what I posited, extracting energy from diurnal temperature gradient, 3 years after my post:
http://news.mit.edu/2018/system-draws-power-daily-temperature-swings-0215
My question might not have been perfect, but it was not about the specific matter of water. It is whether someone who knows the numbers can answer, is it "enough power to be useful?" TLK has stated 'not highly practical or cost efficient' but I think that's in the context of my example (water, oceans, and ice). Since my original post, I've been looking at a vessel with just a little bit of volatile liquid, specifically the polypropylene gasoline jugs in my garden shed. Until someone provides the numbers, I will simply remind myself of the stages of all new ideas:
4 That's impossible.
3 It's too expensive.
2 It will undermine my own personal interests.
1 OK, we'll do it.
