March 21, 2016

We had a new paper out on the arXiv a couple of days ago (but I was ill so didn’t get around to writing anything before now) https://arxiv.org/abs/1603.03589

Using single photons, we experimentally demonstrated so-called Einstein-Podolsky-Rosen steering (the famous ‘spooky action-at-a-distance’ pointed out by Einstein and colleagues). I say ‘we’, but of course I didn’t turn the knobs myself. Rather, it’s been a fun collaboration between me and a couple of other theorists, and some of the excellent quantum optics experimentalists we share roof with. It’s been great to see them turn our abstract ideas into reality in the lab :).

As I have written about in previous posts, quantum physics allows for correlations which are in a certain sense stronger than any which are possible in classical physics. These correlation come in different strengths, according to which scenario they are obtained in. Consider two parties, Alice and Bob. In a scenario where we trust both Alice and Bob (we know what they are doing in each of their labs), we can certify that they share quantum entanglement. This can be exploited e.g. for precision measurements or for some types of quantum cryptography. In a scenario where we do not trust neither Alice nor Bob (we do not make assumptions about what exactly they are doing in their labs), the experimental requirements become more difficult, but if we work hard then we can certify that they share nonlocal correlations. These are even more powerful than entanglement and can be exploited for ultra-secure cryptography and generation of random numbers, as I’ve explained here and here. Steering is an intermediate scenario in which we trust only one of the parties. For example, we trust Bob (we know what he is doing) but not Alice (we don’t make assumptions about what happens in her lab). The correlations obtained in this case are stronger than entanglement, but weaker than full non-locality. And the experimental requirements are also harder than for entanglement, but not quite as hard as for nonlocality.

Our experiment can be seen as a strong demonstration of single-photon entanglement, in a scenario where one party is untrusted. And also as an important step forward towards implementation of full non-locality with single photons. The results are also interesting from a more theoretical point of view, because we had to develop new steering inequalities which are suited for our particular experimental setup.

Published paper: https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.117.070404