I got this question after reading a bit into heat death (https://en.wikipedia.org/wiki/Heat_death_of_the_universe#Time_frame_for_heat_death)
In that section, it is indicated that:
It is suggested that, over vast periods of time, a spontaneous entropy decrease would eventually occur via the Poincaré recurrence theorem, thermal fluctuations, and fluctuation theorem. Through this, another universe could possibly be created by random quantum fluctuations or quantum tunnelling
And it gives several references. Given the usual definition of entropy in cosmology that comes from statistical mechanics (https://en.wikipedia.org/wiki/Entropy#Statistical_mechanics), I understand that all evidence we have now points towards such scenario and that if the universe has a closed topology (which in principle is flat but we cannot completely rule it out) or is isolated from other parts of the universe (by some kind of geometry) or is finite, then, entropy growth will have a limit
However how could fluctuations (be thermal fluctuations, fluctuation theorem or quantum fluctuations as the articles says) affect the universe in the future? I mean, will they be able to form new matter once the universe is emptied and reaches heat death?
Of course, this is assuming that these models apply correctly to those extremely large timescales
