In a refrigerator you pump in energy and it is used to separate cold (inside) from warm (outside). An RTG used to provide power on spacecrafts runs the same process backwards (I think) by finding a source of hot and a source of cold and extracting energy from the difference.
The RTG's source of hot is obvious - decaying radioactive material. But what is the source of cold? Is the near-vacuum around the spacecraft really sufficient?
Sure, it might be cold if we strictly look at temperature.
I wouldn't put my hand in water at 80 degrees C, but being in a sauna at that temperature is OK since the air has a lower particle density. Outer space certainly has considerably lower particle density than air on earth, so does its low temperature really "count"?

__________________
I have a much better signature in a parallel universe

I don't know the mechanics of how this is actually set up, so I may be wrong in the details.
IIRC, thermoelectric generators use a pair of conductors at different temperatures to generate a current flow from the Seebeck effect. In the case you're talking about, the "hot" conductor sits next to the radioisotope source, and the "cold" conductor sits (presumably) somewhere that is thermally insulated from the radioisotope source. If you keep this cold conductor facing away from the sun, it will cool continuously by radiation (I presume you could help this along with some radiator fins), because it "sees" nothing but starlight and the 3K background radiation.
So it's the balance between incoming and outgoing thermal radiation that does the cooling, rather than access to a source of "cold" molecules.
Or so I think.

Grant Hutchison

Think about the Sun - you can feel its heat across 150,000,000km of 'empty' space.

Conductive heat isn't the only means to get rid of thermal energy - it can also radiate away (such as holding your hands up infront of a fire). The black fins on an RTG are there to radiate away the thermal energy. Radiating it into the depths of space is quite an efficient way to do it. Lots of spacecraft use various techniques to do the same to get rid of any excess heat from inside.

Doug

OK, thank you for the replies. Even if radiating away heat is efficient, actual contact with "cold molecules" ought to be more efficient. I suppose an RTG providing power for some experiment on the south pole could operate more efficiently than the ones in space.

__________________
I have a much better signature in a parallel universe

Remember technically there is no source of cold. Cold is just "less thermal energy".

Heat transfer occurs through conduction, convection, and radiation. In space you only have radiation. Effiency of radiation depends upon the emissivity (at a wavelength >1000 nm), the size of the radiator, and the temperature difference between it and what it is pointed at. This last can be important since a temperature difference of 370 K upon night side Earth orbit can go down to 90 K or less when the radiator is facing the Earth during the day side of orbit.

Suffice it to say that a radiator in space would have to be many times larger than one operating in the Earth's atmosphere at the south pole (for the same thermal transfer rate). A radiator immersed in water would be smaller yet.

__________________
"Those who cannot remember the past are condemned to repeat it" - George Santayana