Creating, cooling, and probing a molecule, all in one place

I love all aspects of quantum physics, but the quantum mechanics of ultra-cold atoms and molecules has a special place in my heart. Cooling and controlling molecules is really, really hard work, but you can do some impressive things with the results.

In a recent publication, a team of physicists has outlined a general procedure to cool and control the quantum state of a molecule. The physicists claim that their procedure will (probably) work on any molecule. If true, this is important because molecular spectroscopy is probably going to be one of the keys to finding out where our current theories of quantum physics fail.

Breaking the Standard Model

The Standard Model of physics has come to rule the roost. It describes how the most basic building blocks of the Universe—fundamental particles and forces—talk to each other. From these building blocks, atoms, molecules, and, eventually, French villas are built. Every experiment and observation has confirmed the Standard Model’s awesomeness.

But with the Large Hadron Collider spraying data, we are beginning to see a few more hints that it might be wrong.

Still, there is more to life than new particles and particle decay ratios. The other place to look for cracks in the Standard Model are the fundamental constants. The Standard Model doesn’t always dictate the values of these constants—some are just whatever we measure them to be. Others, however, do have predicted values, and that gives us two avenues with which to approach the standard model without getting slapped into a ditch.

The accuracy of the Standard Model predictions rely upon these values in many different ways. At some level of precision, the Standard Model could be shown to be inconsistent. It could require one value of the constant to explain one set of experiments and a different value to explain a second set of experiments. The second approach is even simpler: measure those constants that the Standard Model predicts a value for. If the measured value of the constant doesn’t agree with predictions, then we have New Physics.

Having a moment

A popular target for this second approach is the “electron dipole moment,” which can be thought of as the spread of charge represented by an electron. Typically, we imagine the electron as a perfect point particle. In this picture, the dipole moment has to be zero. Quantum mechanics, however, actually predicts that the electron has a very tiny dipole moment. Most of the popular proposals for models that extend the Standard Model predict that the electron dipole moment is two to four orders of magnitude larger.

From an experimentalist’s perspective, this is heaven. At the moment, our instruments are not sensitive enough to measure the electron dipole moment, but, as we increase sensitivity, we’ll eventually detect it. And, if the extensions to the Standard Model are correct, we will detect the dipole moment with a less sensitive instrument than would otherwise be required.

The way to approach this sort of problem is not to try to measure the dipole moment directly, but to use our models to predict the dipole moment’s influence on physical systems that we can perform measurements on with high accuracy. It turns out that measuring the energy difference between quantum states of molecules might be one of the better choices in the case of the electron dipole moment.

To do that, you need really accurate spectroscopy on molecules. That means the molecules need to be really cold, and the gas they reside in needs to be diluted so that the molecules don’t collide too often. Even then, the molecule has to be rather well-trained in the sense that, if you put it in a given state, it will stay there. Finding this sweet spot is not a forgone conclusion, but the new work may point the way.

Cool molecules

I’ve written about temperature and molecules a few times. Reading that history will tell you that temperature is not just defined by the translational motion of a molecule, but also the molecule’s internal state. For a gas of atoms, we can use lasers to transfer translational energy to the atom’s internal state and, from there, into a photon that takes energy away, leaving the atomic gas colder.

But, with molecules, the complexity of the internal states defies control schemes. We can suck energy out of the translational motion of the molecule and into the internal structure of the molecule, but we can’t get that energy to leave as a photon. So the molecule remains hot. While it’s possible in principle to cool molecules as it is done with atoms, it’s just not feasible. Each molecule would require a vast set of unique lasers. Luckily, other ways exist. And one of those other ways is called sympathetic cooling.

Playing snooker with molecules

The idea is exactly like playing snooker. I hit the white ball. If I am lucky, the white ball hits a red one (in a one-in-a-million chance, it may even be the ball I was aiming for). The white ball stops, and the red takes off. Effectively, the white ball has been cooled by the red.

In the actual experiment, the physicists start with two calcium ions (an ion is an atom with an electron removed) that are trapped by a cage of electric fields. They then use a set of lasers that lets them efficiently cool the calcium ions to very low temperatures. Next, they introduce a tiny amount of hydrogen into the cage. The hydrogen reacts with a calcium ion to create a calcium hydride ion. The energy of this reaction heats it up considerably. As the molecule shakes about in the trap, it sets its neighbor, a calcium ion, vibrating as well.

If left to themselves, the two ions would eventually end up at the same temperature, vibrating around in the trap rather violently. But the calcium ion is cooled continuously so that it is always colder than the molecule. This means, as they interact, the molecule transfers energy to the calcium ion. Eventually, the molecule ends up almost as cold as the calcium ion.

The molecule is cold, what now?

After cooling the molecule, the researchers want to perform spectroscopy on it. That means measuring the energies at which the molecule absorbs and emits energy. Again, this would normally mean that a different set of lasers is required for each molecule, since they all absorb and emit at different energies.

But, there is a spectroscopic technique called Raman spectroscopy where this is not true. In Raman spectroscopy, the molecule absorbs a photon of light and then immediately emits a photon of light but with a lower energy (so a redder color). The leftover energy is deposited in the molecule, often as vibrational or rotational motion. Here, the color of the light used does not have to be specifically chosen to match the energy landscape of the molecule. Or, put more simply, nearly any laser will do.

Now, the question is, how do we know that the laser has done its job and put the molecule in the state we wanted? To do this, the researchers make use of the calcium ion to spy on the neighboring molecule.

Because the two molecules are trapped in the same set of electric fields and are in a magnetic field, they are coupled. The rotational motion and orientation of the molecule (called a spin state) with respect to the magnetic field changes the orientation (spin state) of the calcium ion. So, an attempt to get the calcium ion to emit light will succeed or fail depending on the rotational state of the neighboring molecule. That is clever, because testing the state of the calcium atom does not change the rotational state of the molecule.

The sheltered life of a trapped molecule

All this is impressive but still not quite enough. For truly precise spectroscopy, the molecule needs to be undisturbed. To check this, the researchers looked for coherence. The basic idea is as follows: the laser drives the molecule into a certain state with a probability of unity if the laser emits a certain amount of energy in a specific amount of time. If the laser is switched on for twice as long, then the probability of driving the molecule to the target state returns to zero. The probability varies smoothly between zero and one for different lengths of laser pulse. These oscillations are called “Rabi flops,” which are a signature of a coherent process.

Understand coherence is not really important in this case. What’s important is knowing that coherence in a quantum state is a delicate thing that can be destroyed by the smallest outside influence. So, the observation of Rabi flops tells us that the molecule is well-sheltered from the outside world.

But will this work for other molecules?

The researchers claim that this technique can be generally applied to pretty much any molecule. Just not quite yet. First of all, they don’t have very good control over rotational states of the molecule. The researchers have plans to install a laser system to control to the rotational state.

But there is more to the story. The cooling method works well in this case for two reasons. First, the mass of the molecular ion is similar to the mass of the atomic ion, so the transfer of energy between the two is very efficient. For more complicated molecules, I’m not sure that this will work so well.

The second problem is control of the internal state of the molecule. The calcium hydride ion is simple. It has only one way to vibrate (the Ca and H separate and come back together again) and two rotational states (the whole molecule rotating about two axes at 90 degrees from each other). The energy separation between these three forms of motion is high; energy isn’t transferred between different states so easily. For more complicated molecules, this is not the case. Indeed, you often end up with combination states that involve both rotations and vibrations.

So, yes, I can see this working well for translation motion of all molecules and for simple molecules with only a few possible ways to vibrate and rotate. But something like water, which has a single extra hydrogen atom (compared to the calcium hydride ion), has six vibrational modes.

The final problem is one of control. In this experiment, the molecule was created in the trap, so the researchers could (through a bit of trial and error) choose to play only when they obtained a trap with a single atomic ion and a single molecular ion. For more complicated molecules, this will not work. Instead, the molecules will have to be actively trapped after they are introduced to the chamber and ionized. And you will still want to balance the number of molecular ions with the number of atomic ions (assuming you want more than one of each in some cases).

Nevertheless, this is still an excellent bit of work, and the ideas outlined here will find extensive use in molecular spectroscopy laboratories around the world.

Nature, DOI: 10.1038/nature22338 (About DOIs).

Related articles

Comments

Share article

Latest articles