Are there alternatives to steam in thermal power stations? 'A thermal power station is a power plant in which heat energy is converted to electric power. In most of the world the prime mover is steam driven. Water is heated, turns into steam and spins a steam turbine which drives an electrical generator.'
Why is H2O chosen? Shouldn't a liquid with a lower boiling point be used?
Also, why can't the following setups be used? Have they all been duly considered?
1) Expanding Liquid:
In this setup, a chosen liquid is heated, expanded, and used to push pistons, without getting boiled. As a liquid is hardly compressible, shouldn't the expansion generate a tremendous pressure? Furthermore, no latent heat is involved.
2) Expanding Empty Spiked Solid in a Liquid:
In this setup, a specially crafted empty solid is heated in a liquid to give rise to a large increase in volume. The liquid then pushes a piston.
3) Strong Balanced Bimetallic Strips:
In this setup, three sets of bimetallic strips/rods are used to push a piston when cooled (and bent when heated).
4) Expansion of Ice:
In this setup, freezing H2O in cold regions, natural or artificial, expands by about 9% and is used to push pistons.
 A: Expanding Liquid: the expansion per energy unit is tiny compared with the expansion at the phase transition from liquid to gas.
Expanding Solid:  even worse than expanding liquid
Bimetallic strips: very little energy transfer capability and long cool-down cycle.
Ice: see "expanding solid" above.
The next best choice after conversion of (any) liquid to gas is direct gaseous expansion as is done in internal combustion engines; the latter is far less efficient.  
As to the choice of water: it's easily available, relatively nontoxic , and boils at a reasonable temperature: not terribly hot, but not so low that it boils at, say, room temperatures. 
A: Typically, modern gas powerplants use combined cycle and are projected to achieve over to 61 % thermal efficiency, which is to my knowledge far highest of any contemporary technology. The first stage relies of thermal gas expansion similar to a jet engine, the second stage operating at lower temperature is the classical steam cycle. So the answer is yes.
Wikipedia also lists alternative technologies to the gas turbine in the first stage: "Other historically successful combined cycles have used hot cycles with mercury vapor turbines, magnetohydrodynamic generators or molten carbonate fuel cells, with steam plants for the low temperature "bottoming" cycle."
A: A liquid with a lower boiling point would be better, however the boiling point still needs to be above ambient temperature, and there just aren't any liquids available with their boiling point at a better temperature and without other disadvantages. 
Lower-temperature-boilingpoint liquids are used in thermal power generation when the high temperature is lower than in regular gas or coal fired plants, e.g. geothermal or solar plants. Here usually a hydrocarbon liquid is used, and this technology is known as the organic rankine cycle. Hydrocarbons are nice because you can engineer them to have any boiling point you want, but they are not a good fit for regular power plants because they start to decompose at high temperatures. 
There is also research into supercritical CO$_2$ powered turbines, however AFAIK that is not about thermal efficiency but about other advantages such as the smaller size or CO$_2$ capture. There is also research into inert gas powered turbines, but there the main application is nuclear power plants.
Regarding your other setups, you seem to think that a lower expansion ratio is preferable. That's not the case, the higher the better. Boiling water has an expansion ration of about 1000, or 100000%, so try to beat that. Steam can also achieve any pressure you want, but the pressure limits are dictated by the construction materials, not the working medium. 
