We are surrounded every moment by air - a gas of nitrogen, oxygen, argon and other elements. This gas is comprised of molecules that fly around at hundreds of meters per second. Although individual molecules obey the strange laws of quantum mechanics, these quantum effects are “washed out” by the random thermal motion given to the atoms by virtue of the fact that we live in a relatively warm world. However, if one was able to cool atoms to a fraction of a degree above absolute zero – where atoms were slowed down to speeds of less than a millimeter per second – one could strip away the random motion of these atoms and be able to observe behaviour that was fundamentally quantum mechanical.
In the last decade, experimental physicists have been able to realize just that situation. Using a combination of laser cooling and magnetic trapping techniques, physicists are able to cool atoms down to less than a millionth of a degree above absolute zero. These quantum gases are interesting to us because the atoms in the gas interact very strongly. The consequence of this strong interaction is that no atom acts individually, rather the atoms act in concert, always aware of what each other is doing.
and this is just where the research really begins…
Lattice Experiment Research:
-click here for a diagram of how the experiment works-
Physicists, biologists, economists, and sociologists create models, simplified ideas of how the world works. Sometimes the implied behaviour is obvious from the model, or can be solved mathematically. Otherwise, simulation may be necessary. Most mathematical models can be simulated using a computer. Some simple quantum models, however, can be challenging, even to state of the art computers. This difficulty is quite distinct from the difficulties that may arise in defining the model: we are asking what behaviour would occur in a world governed by the model, and neither analytic mathematics nor powerful computers are able to tell us the answer.
The reason quantum models can be hard to simulate is that they require keeping track of an exponentially large set of variables. Of course, nature has no trouble keeping track of these variables — a realization which led Richard Feynman (a Nobel laureate in physics) to propose using quantum systems to simulate quantum models.
In our experiment, we are trying to simulate the quantum Fermi-Hubbard model. This is a model which is thought to explain superconductivity - the phenomenon whereby an electrical current flows through special types of crystals without any resistance. We are simulating these crystals, and the electrons that are super-conducted through them, by making an optical crystal out of interfering laser beams. Our analogue to the electron is the Potassium atom.
By cooling atoms to nano-Kelvin temperatures, we can create an experimental realization of the Fermi-Hubbard model, and be able to see what this model actually predicts. If we do see super-conductive behaviour, then we will have found out the the model is indeed a good one. However, it is likely that we will see unexpected phenomena. In this case, we will have to amend the model to find out the limits in which it is a good description of what actually happens, and the limits in which the theory breaks down.
All this work will hopefully allow people to understand the mechanism of superconductivity enough so that people might construct materials that super-conduct at high enough temperatures (room temperature, for example) to be useful for application.
Chip experiment research:
Fundamental particles can be classified as either bosons or fermions. The two kinds of particles are distinguished by their behaviour towards one another - bosons are the "social" particles and like to be with each other; fermions are "antisocial" and tend to stay apart. At everyday temperatures, particles are generally too far apart from each other, and moving too quickly with respect to each other, for these "quantum statistical" properties to matter. Our goal is to reach temperatures where the quantum statistics matter, and to study the sociology of either bosons or fermions or mixtures of the two.
bosons
fermions
Our current experiment is looking at the behaviour of bosons when we split one cloud into two. The question we're interested in answering is: "How does a boson decide whether to go into the left cloud or the right cloud?" If these were classical particles, that is, the same as those we encounter every day, the answer we get should be no different than if we were to flip a coin many times and count the numbers of heads vs. the number of tails. For 1000 coin tosses, we expect about 500 heads and about 500 tails, though we'd expect some randomness to creep in and such that we shouldn't be surprised if we got 488 heads and 512 tails. The amount by which we can reasonably off from the expected number of heads and tails is described by Poissonian statistics, and tells us that this uncertainty should be on the order of 
, where 
is the number of measurements.
When the particles being split into two groups are ultracold bosons, the situation changes. Bosons do not act as independent, random entities; instead, they are now part of a greater "community," something known as a Bose-Einstein condensate (BEC). We expect that, for a cloud of 1000 atoms, we will get fewer fluctuations from the perfect 500-500 split-ratio. When the atoms work together to maintain a perfect balance, they attain the minimum total energy for the system. In practice, we create a cold cloud of atoms, split it into two, and then very carefully count the numbers of atoms in the right cloud and in the left. We then repeat this experiment many, many times to build up statistics. We are interested in understanding the conditions for which the less-than-random fluctuations are obtained and how we might influence the 'decisions' the bosons make. Ultimately, we hope to gain insight into such "many-body" phenomena - understanding the collective behaviours of quantum particles, with applications ranging from magnetism to superfluidity.
