It’s spectacular science: the loophole free Bell test. At least, I was told it is. So I went on Wikipedia to find out what that is about. I gave up. ‘But this will become ‘textbook material’!!’. I tried again with Wikipedia. And gave up.
A few Sundays ago, I ground my teeth and really tried. It’s fascinating. Although Quantum Mechanics is generally accepted nowadays as the model to explain reality on the smallest scale, I learned that so far nobody managed to deliver the ultimate proof: an experiment that passes the Bell test, with both ‘loopholes’ closed.
We are in a race: technology is ready to prove that the Great Einstein was actually wrong in one fundamental point. Ronald Hanson’s group is racing for Delft University. Wow.
And somewhere this year, I may have to write a ‘press release’.
Quantum Mechanics explains the behaviour of quanta, the smallest things we know. They behave odd, in a way that beats every form of human imagination.
Particles on the smallest scale (electrons, photons) can be at two different places at the same time, and can have two different states at the same time. Particles can be ‘entangled’, meaning that they effectively form one particle although they are in two different places and in different states simultaneously. Imagine an apple that is on two trees at the same time, being both green and red at the same time. Only when you look, you force reality to choose the tree and the colour. And, amazingly, if you look in one tree, you instantly affect the ‘apple’ in the other tree. This leads to very odd behaviour but also, possibly, to great new technology: a computer that exploits this behaviour to create a mind-boggling calculating power: the quantum computer.
Was Einstein wrong?
If you’ve trouble believing that, you are in good company. Einstein never believed it. “God doesn’t play dice”. In Einstein’s view, a particle can only be influenced by its direct surroundings and not instantly by a spooky entangled particle far away. Influence, his theory of relativity says, is limited by the speed of light: particles nor information can travel faster (a principle called ‘locality’). And Einstein also favoured ‘realism’: the apple is already green before you look at it, and not green and red at the same time, until you look (as could be the case in Quantum Mechanics). He called entanglement ‘spooky action at a distance’ and he didn’t buy it. The greatness of Einstein is of course undisputed, but on this rare occasion he was probably wrong.
Quantum Mechanics is not new, it’s about 100 years old and its origins go back to the famous physicist Max Planck. A first trigger was the realisation that light can be seen both as a wave and a particle, which classical physics can’t explain. One great scientist, Huygens, showed that light is a wave. Another giant, Newton, proved it is a particle. It lead to centuries of debate, as they can’t be both right, can they? Yes, they can, and the answer lies in Quantum Mechanics.
In 1964, physicist John Bell devised an experiment using two entangled particles that could experimentally show that the the world really is not ‘local’ as strongly believed by Einsteing, and that this ‘spooky action at a distance really existed: : the Bell test (see below1)
Over the last decades, an overwhelming amount of experiments has shown that our world is not ‘local’, and Quantum Mechanics explains it really accurately, although very much beyond our imagination2. As the great scientist Richard Feynman said: “Quantum mechanics describes nature as absurd from the point of view of common sense. And yet it fully agrees with experiment. So I hope you can accept nature as She is – absurd”.
Not settled yet
But even today, with all our advanced technology, the old and heated debate is still not fully settled. There is still one catch left: a Bell experiment has ‘loopholes’. What if, said its critics inspired by Einstein, there is a classical phenomenon that we do not yet know that could also explain the results within classical physics? Not such a strange thought, as there where days we didn’t know of magnetic and electrical fields, for instance. There might be ‘something else’. If there would be such a ‘hidden variable’, we wouldn’t need that spooky science.
All previous Bell tests suffer from two ‘loopholes’3: The first is the ‘locality loophole‘. If the detectors (for some reason always named Alice and Bob) or the particles would have an unknown way to ‘whisper’ to each other about their questions and answers, they could pass the Bell test cheating. It would be pretty spectacular of course, but it would not need be ‘instantaneous’. So the world could still be ‘local’. The logical way to close the locality loophole is by placing the detectors and entangled particles far apart, but that is really difficult. It can be done for instance, with photons, particles of light. But then you hit the ‘detection loophole’.
The problem with photons is that many of the entangled photons disappear on the way. Photon experiments normally detect only a tiny fraction of the photons emitted, as may are absorbed on their way by molecules in the air or glass fibers. There is no way to be sure your subset is representative, so an alternative explanation could be hidden there. The obvious way around is to place the detectors very close to each other, but -you guessed it- then you hit the locality loophole.
Both these loopholes could hide a classic explanation for the measured results – and make Einstein (‘post mortem’) a very happy man: no spooky science. And closing them simultaneously is very very hard.
Closing the loopholes
In the recent years, scientists have labored hard to close these loopholes and both where closed individually in the recent decades. A rather spectacular experiment on the Canary Islands in 2009 closed the locality loophole by sending a light particle (photon) from La Palma to Tenerife. The detectors were so far away from each other that both the photons and the detectors could not possibly exchange information without breaking the speed limit set by Einstein himself, the speed of light. But that experiment did not close the ‘detection’ loophole, as only a small percentage of the entangled photons where detected.
No experiment to date closed ALL loopholes in one test. Although the evidence for quantum theory is overwhelming, Einstein has still not been proven wrong. There COULD still be another classical explanation.
Racing for QM history
But probably not for long. Technology has reached the point where it is possible to do a loophole-free Bell test and various groups over the world are racing to be the first to write quantum history.
Among them: Ronald Hanson and his group at Delft University, researching quantum bits based on ‘spin’ in diamond, a tiny directional magnetic effect that electrons have. Ronald already ran successful experiments to entangle two qubits within 3.5 meters before. The ‘detection’ -loophole is not an issue here, as he detects 100% of all the entanglements he creates. But with his 3.5 meter set-up he cannot fully exclude that there is some unknown ‘whisper’ that exchanges information between the qubits or between his detectors, so the locality loophole is wide open. Einstein could still be right.
So they will now move one detector over 1300 m away, to a lab inside the Reactor Institute – a location simply chosen for the distance to Hanson’s own lab. The detectors need to be at least 900 meters away to ensure the speed of light (300.000 km/s!) rules out any exchange of information between the qubits or the detectors, which effectively closes ‘locality’.
The cables are laid, the experiment is about to start somewhere this spring. How long it will actually take, nobody really knows. Entanglement between the separated qubits is created only when exactly the right two photons meet in exactly the right way in the middle, a connecting hub hidden underneath the Faculty of Electrical Engineering, Mathematics and Computer Science. Many photons are absorbed along the way by the glass fiber, so only rarely the entangling will succeed. One successful entanglement may take minutes, our maybe the good part of an hour. And they need hundreds of entanglements to get the proof they want.
But if the setup works, at some point in the near future there might be proof. And not just for the sake of science (and proving Einstein wrong). Entanglement promises a new way of secure communications: a encryption key based on entanglement cannot be intercepted and without that key, the data flowing between ‘Alice’ and ‘Bob’ can’t be decoded by a third party. Except when a very clever spy would exploit one of the open loopholes. That’s why closing the loopholes is more than just a scientific exercise.
Some time from now, there won’t be any open loopholes anymore. And I will start writing the most incomprehensible press release of my career, which is likely to completely fail in telling you something really, really exciting has happened. And our world will have become firmly and definitely a little more weird.
Update 2/2/15: The Story Continues. It’s not easy, and this is why.
1 Bell devised an experiment with two parties ( ‘Alice’ and ‘Bob’). One could compare the experiment with a referee, asking Alice and Bob question x or question y with two possible answers- A and B.
If at least one of them receives question x, they should give the same answer (so A/A or B/B) to score a point. This is the case for questions (x/x, x/y and y/x). However, if they both receive question y, they get a point if they answer differently (A/B or B/A).
If Alice and Bob would answer purely randomly, they will score 50%. The best strategy to get points is to always give the same answer, which will give 75%. Surprisingly, John Bell showed that using entanglement a score of 85% is possible! To be able to prove entanglement, you therefore need to get to a score above 75%.
2 OK, except for toddlers! My youngest daughter has no problem accepting her cuddle toy can be in two different places (bed AND washing machine) and in two states (very snotty AND clean) at the same time. Which proves you lose your fundamental understanding of the universe when you grow up 😉
3 Bell himself apparently identified a third possible loophole, which would be that the detectors would not be free to chose the questions they ask in a Bell test. Experimentally proving that such super-determinism doesn’t exist, is deemed fundamentally impossible, as one has to prove that the questions weren’t already ‘imprinted’ in some way in the Big Bang. This loophole is often seen as not relevant.
(this blog was updated a few times, as my understanding of this topic is still slowly growing – I hope)