Quantum Mechanics Braces for the Ultimate Test: The now-celebrated Aspect experiment, along with similar ones, helped to write nonlocality into physics textbooks. But there is another loophole that those experiments did not close. The trouble is that photons are slippery customers: small, fast, and notoriously hard to detect. Typically, if five photons are hurled at a detector, it will register only one. That means that physicists can trust that Bell's bound has been violated only if they assume that the photons caught provide a fair representation of how all the photons in the experiment behaved—much the way exit polls at voting booths predict election results.
Most physicists accept that the fair-sampling assumption is a good one. “It's unlikely that nature is so malicious that it conspires with the apparatus to hold back particular photons just to fool us into thinking that quantum mechanics works,” Gisin says.
Nonetheless, physicists hate loose ends, so the chase to find a perfect, loophole-free test has continued over the past decade. “Until the test is done, we can't honestly say that hidden variables have been ruled out—even if the consensus is they don't make sense—because we haven't proved it,” says Harald Weinfurter of the Ludwig Maximilian University in Munich, Germany...
Building on Wineland's experiment, Weinfurter's group is attempting to tie up both loopholes at once, by weaving photons together with atoms to reap the benefits of both. The idea is to start with two initially unentangled atoms in separate laboratories—ideally more than 100 meters apart, so that the atoms cannot influence each other over the course of the test. Each atom emits a photon; the two photons are captured and transmitted along optical fibers to a third location, where they are entangled. “The magic is that as soon as the photons are entangled, their parent atoms automatically become entangled, too...”