The coldest place in the universe that we have observed is in the Boomerang Nebula. Similarly to a refrigerator, heat is carried away through a fluid, resulting in a space colder than the surrounding environment. However, in the case of the Boomerang Nebula, the coolant is heated gas being expelled away from the remains of a star - bringing almost all heat energy with it.
This seems like an efficient way of removing heat / cooling a large object - and perhaps it could be applied to a planet rather than a star (or a fridge). For instance, a planet with a runaway greenhouse effect may channel its heat outward relatively quickly, a planet too small to hold an atmosphere may lose gas quickly, or a planet without a protective magnetic field may be stripped by solar winds.
So about how cold can a planet actually be?
- Any feasible, naturally occuring method of cooling is acceptable; if it is speculative, it must be justified.
- The accepted answer will be credible and justify a reasonable surface temperature colder than those provided in other answers.
- Note that this is hard-science so rather than "it could be around x degrees because this cold nebula works in the same way" you must provide calculations or citations.
- The rest is up to you. I have no specifications for this planet other than that it must be colder than the temperatures the cosmic microwave background usually permits; you can choose size, density, composition, location, nearby objects, moons or lack thereof, etc.