I'm trying to understand in a simple, non-mathematical way the basic concept underlying why a particular, very recent experiment turned out the way it did.
The experiment, by Serra and Lutz entitled 'Reversing the thermodynamic arrow of time using quantum correlations', demonstrated an interesting wrinkle in the 'arrow of time' and entropy concepts.
I understand this much: the experimenters took a chloroform molecule and cooled the Carbon atom and warmed the Hydrogen atom, and also created a correlation between the hydrogen and carbon atoms. Because this correlation constituted 'information', and information isn't just something in the mind of a human being but has a physical meaning in the world, this correlation/information represented 'organization' as opposed to 'randomness' and therefore the system in the experiment had-- as an initial value for entropy-- less entropy than would have been the case if there were no correlation between the H and C atoms in the nucleus. (I think I've presented everything up to this point correctly-- if not, please indicate exactly what errors I've made.)
Then, apparently two things happened next: the correlation between the H and C atoms decohered, and 'heat' flowed from the cold Carbon atom to the warm Hydrogen atom in the nucleus of the chloroform. Because the dissolving of the correlation was accompanied by an increase of entropy, heat was released, as is always the case when entropy increases. That much I understand.
What I don't understand (and what wasn't explained in either Serra and Lutz's paper or the article I read about their paper) is this: why, in this case, did the release of heat because of the loss of 'information' contained in the now-decohered correlation not just go into the environment and nothing more-- why was it accompanied by a flow of heat from the cold Carbon atom to the warm Hydrogen atom, which was therefore doing 'work' by creating less randomness and thus we see a lesser net entropy increase (Entropy increased by X amount because of the end of the correlation between H and C atoms, but entropy decreased by Y amount because of the flow of heat from the cold C to the warm H, thus reducing the net Entropy increase in the system.)
Why did that happen?-- and since it was predicted to occur, there must be some physical requirement for it to happen that is well understood by physicists. Could you explain it to me please? I'm wondering if the correlation that existed between the H and C atoms required that some of the heat generated by the dissolving of the correlation 'stay at home' with the correlated pair and be used to offset the decreased organization from the lost correlation by increasing the organization elsewhere if possible, and in this case that meant having heat flow from the cold C to the warm H. (That is sheer speculation on my part based on no knowledge at all!)
Here is a link to the Serra-Lutz paper: https://arxiv.org/abs/1711.03323