Following Schrödinger’s cat to its death and giving it a reprieve

One thing I dislike about quantum mechanics is that it encourages journalists to overuse words like "mysterious" and "spooky." Given that we can model quantum systems to an accuracy that would make a god cry, where is the mystery?

I personally blame Schrödinger and his eponymous cat. That thought experiment, combined with an ever-increasing body of experimental results, shows just how subtle quantum mechanics is. Rather than admitting a lack of understanding, some journalists seem to use Schrödinger’s cat as a get-out-of-explaining-for-free card.

No doubt the words "mysterious" and "spooky" are going to turn up again regarding an experiment that, according to its press release, shows how to predict when Schrödinger’s cat is about to die. The experiment is really only tangentially related to Schrödinger’s cat; instead, it shows how no matter how well you understand quantum mechanics, there is always another layer to peel back and explore.

Forget about cats

Instead of talking about cats, I’m going to talk about atoms and the way they behave. Imagine you have an atom sitting in space. Nothing will disturb the atom except me. I have a trusty light source that emits single photons that I can use to excite the atom.

Since I’ve done nothing (yet), the atom is completely relaxed and in the ground state. I blast the atom with a single photon. In doing so, my knowledge about the atom reduces. There is a possibility that the atom absorbed the photon and is now in a higher energy state, but the photon may have simply gone straight past the atom, leaving it in its current state. To predict the atom’s behavior, I have to take my lack of knowledge into account by describing the atom as being both in an excited state and being in its ground state. This is called a superposition state.

The superposition state sounds like a mathematical trick to get the numbers right, but it isn’t. The atom really is in a mix of two states. The superposition state is described by probabilities—the chance that if I measure the state of the atom, I will find it in the ground state. The balance of probabilities between the excited state and the ground state doesn’t have to be 50/50, as it depends on the probability that the atom absorbs the photon.

I can manipulate that probability: if I shoot the atom with a second photon, the balance of probabilities shifts to increase the likelihood of the atom being found in the excited state. That balance continues to shift with each additional photon.

If I put a photon detector behind the atom, I change my knowledge of the atom. If the photon is not absorbed by the atom, it hits the detector, which clicks. Now I know when a photon was not absorbed, which means the atom is definitely in the ground state. 

In a more subtle variation, I can put the detector to one side. That doesn't tell me if my photons are being absorbed or not. But if my detector goes "click," I'll know the atom was in an excited state and exited it by emitting a photon that happened to go in the direction of my detector. 

An important part of these last two versions is that there is an element of unpredictability involved: I can't know beforehand if a photon will be absorbed, and I can't predict when my detector will go "click." The events of photon absorption and emission represent quantum jumps, where the atom suddenly transitions from one state to another. The timing of these jumps is entirely unpredictable. Well, almost unpredictable—it turns out to depend on the timescale at which you observe the atom.

Artificial atoms let us peek inside

Now we get to the most recent research on this topic. Because the researchers don't have access to our theoretically perfect atom, they used an artificial atom that is nearly perfect. Without going into too many details, the atom has three states. One is the ground state, which has the lowest energy. The next highest energy state is called a dark state. The dark state is shielded as best as possible from all possible forms of interference. Because of the shielding, it is impossible to directly measure if the atom is in the dark state. The final state is a high-energy, short-lived bright state—the atom never stays in the bright state for long. From the ground state, the atom can go to the dark state or the bright state. But from the bright state or the dark state, the only possible transition is back to the ground state.

The bright state is also connected to a read-out circuit. If the atom was in the bright state and relaxes to the ground state, it emits microwave photons that are detected with nearly 100% efficiency. Likewise, if photons suddenly disappear (so the detector falls silent), it can be inferred that the atom went from the ground state to the dark state. Effectively, the bright state is continuously monitored. 

To drive transitions, the atom is bombarded with photons that allow it to transition from the ground state to the bright or dark state. Left to its own devices, the atom rapidly transitions back and forth from the ground state to the bright state, emitting microwaves. Then, at some random moment, the atom goes quiet, signaling that the atom has entered the dark state. Eventually, the bright state lights up again, indicating that the atom has left the dark state. There is no way to predict when these transitions will occur.

Spotting the moment

But it turns out that you can see the on-coming transition shortly beforehand. It all comes down to timing. Every time the photon detector goes "click," it means that the atom has left the bright state and has returned to the ground state. From that moment on, there is a non-zero chance that the atom will enter the dark state: the atom is in a superposition state of the ground and dark states. As time goes on with no further clicks, the probability of the atom being in the dark state increases, until it effectively becomes unity. The time it takes to initiate the superposition state and for it to collapse to the dark state is much shorter than the average time that the atom spends in the dark state.

The trajectory that the probabilities of the superposition state follows is entirely predictable because we know the atomic system so well. Hence, it should be possible to apply a kick (via a pulse of microwaves) to modify the probabilities. The researchers showed this by watching for the bright state to go quiescent for slightly longer than normal, then applying a kick at just the right moment to ensure that the atom entered the dark state. They also showed that they could do the reverse and apply a different kick to prevent the atom from entering the dark state.