In many places the temperature difference between day and night is more than 20 degrees C. Max 45 C and min 25C. Can we create a machine which uses a liquid which has boiling point around 25-30 C and generates electricity in the same way the steam engine does? Do we have such liquid which boils at normal temperatures and pressures?
If we fill such liquid in a non insulated but properly sealed capsules, do they have potential to reduce global warming (without the need of being replaced)?
I'm going to answer the implicit question instead of the explicit one. The implicit question is, if we had such a liquid, would this be a good idea?
Let's first talk about power stations in general. Typically, you don't really want to operate them on a small temperature difference. The reason is Carnot's theorem, which tells us that the efficiency of a heat engine depends on the temperature difference that's used to power it. The maximum possible efficiency (work output divided by heat input) is
$$\eta = 1 - \frac{T_\text{cold}}{T_\text{hot}}.$$ For a typical heat engine used for electricity generation, the cold reservoir is around room temperature (since the power station dumps heat into the outside world, e.g. by cooling towers). The hot reservoir is provided by whatever powers the power station, but will be set up to be above the boiling point of the working fluid. (I assume well above it, as you really want it to be as high as possible for maximum efficiency.)
So for a general power plant, lowering the boiling point of the working fluid wouldn't help. It would allow you to use a hot reservoir with a lower temperature, but that's not actually what you want to do, since that would lower the efficiency of the process.
The temperatures in the equation above are in kelvin, so if the hot and cold reservoirs are 45 and 25 C, the maximum efficiency is $1-\frac{298}{318} \approx 0.06$.
Your idea is to generate electricity from the difference between night and day temperatures. With this heat source you don't have the option to use a larger temperature difference, so a working fluid with a boiling point around 25-30 C might in principle be a good idea. The resulting engine would be inefficient but not impossible to build. However, in practice, a better way to generate work from such a small temperature difference is the Stirling engine, which uses a gas as its working fluid, without taking advantage of a phase change.
Stirling engines can operate on temperature differences that small, so why aren't they used to generate power in the way you describe? It's probably largely because of practical and economic matters - the amount of electricity you could generate that way is probably far too small to make it worth the cost of building the generators in the first place. This scheme would be a very inefficient way to take advantage of the sun's energy (which is ultimately what's causing that temperature difference), and it would take up a lot of space, because you would need a large thermal mass that gets heated up during the day and cooled down at night. Because of this, I would expect that in all places where such a plant could be built, a more traditional solar or wind plant would produce a much greater power output for the same or lower cost.
Butane or a similar hydrocarbon would boil at the correct temperature. However, butane is HIGHLY flammable, making it unsuitable as a refrigerant. In addition, even if a non-flammable substance exists that meets your requirements, it probably wouldn't be useful for electricity generation. The amount of work that you can get from a heat engine depends on the temperature difference between the heat source and the heat sink, regardless of the working fluid involved. Because there is such a small temperature difference between day and night temperatures (i.e., 20 degrees C), the efficiency of any process that used this method would be VERY low. For an example of a similar process using the temperature difference between surface ocean water and deep ocean water (OTEC), see this.
Engines working at thermal differences this small actually do exist.
See, e.g. this: https://en.wikipedia.org/wiki/Drinking_bird
Some mechanical clocks are powered like this:
https://en.wikipedia.org/wiki/Atmos_clock
This one even does some practically useful work.
What all these heat machines have in common is their minor power output, matched only by the small power requirement of the mechanism they power.
If you really want, you can produce electricity this way. It will be probably enough to power an LCD clock.
Now, you need to scale up? No problem... except for the scale factor.
The heat exchange is the limiting factor here. If you scale the above drinking bird 2x, your heat exchange will get 4x (as it scales with the surface), but the mass will get 8x the original. So will do the friction forces. I think you cannot make an 1kg bird that works.
In fact, I am not sure you can make a bird that will produce in its whole lifetime the energy used for welding the glass parts.
In regard to "boiling" point in your question: You don't really need boiling for a heat engine. In pure thermodynamic terms, boiling is in fact a source of inefficiency. In a steam engine, one uses boiling in order to achieve a practically useful power density. Otherwise, evaporation from the surface would do (this is how our planetary water cycle works, after all).
What's more, liquids evaporate (and even boil) at wide ranges of temperatures, depending on the pressure. E.g. the ether duck example self-regulates its working pressure and will work in any non-freezing room temperature.
The normal steam cycle is called a Rankine cycle.
What you are describing is called an Organic Rankine Cycle and they have been around for a long time. Butane and propane are sometimes used as a working fluids, despite their flammability. A less flamable alternative is chlorinated or fluorinated hydrocarbons but many of these have been banned due to concerns about depletion of the ozone layer. CO2 is also used because it is safer, and NH3 because of its excellent thermal properties.
In theory it is actually possible to operate a steam cycle below 100C. However the equipment would be huge because it would need to operate below atmospheric pressure which means the volume of vapour processed to get any reasonable amount of power would be very high. Using fluids with lower boiling temperatures means the cycle operates at higher pressure, so the equipment can be kept small and the high pressure means frictional losses are proportionally much lower.
As others have said, the temperature difference between night and day is low which results in low efficiency. You would also need to store heat during the day in order to be able to run the equipment at night, which adds additional cost. It would be better to run the equipment during the day and employ some kind of solar collector to get the temperature of the working fluid up to at least 100C.
The only case where temperature differences as low as you mention in your question are used is ocean thermal energy conversion, where tropical surface seawater is used as the source of heat, and cold seawater is brought from depth to act as the heat sink. The fact that warm/cool water is available at all times allows these systems to run continuously which offsets the high cost resulting from the low efficiency.
Cycles can work work closed cycle (with a working fluid with a lower boiling point than water as mentioned above.) They can also work open cycle, where the warm seawater boils at around 24C and is then condensed; the low pressure means a large and costly turbine is required, but there is the advantage that the plant also produces desalinated water at the condenser as a byproduct.
Another technology that can extract power from low temperature differences is the Stirling engine.
How would you use the temperature difference betwen night and day to run a thermal engine? You need the two sources, cold and hot, at the same time in order to run an engine at a "normal" cycle rate. It does not matter what the temperature difference is, it is not between two reservoirs existing at the same time. Unless you want to run your engine with one cycle in 24 hours. What if the day temperature were 110 Celsius and the night temperature 50 Celsius? Do you think you could run a steam engine with water as fluid? The water will evaporate during the day an will condensate during night. But this won't make the engine produce too much work. The piston will move once per day, maybe.
This technology already exists and is used widely in geothermal power generation all over the world. It is binary type (as opposed to flash steam turbine), commonly using pentane as the motive fluid (boiling point of about $36°\text{C}$). The most successful of these is done by a company called Ormat. It is truly wonderful stuff.
