Laser cooling and classical statistical physics

The art of cooling neutral atoms with laser light was conceived as an idea almost half a century ago, and the technologies that enabled the realization of this broke through in the mid 80ies. Initially the most promoted application was atomic clocks and precision measurements. Ten years later, the first successful production of Bose-Einstein condensates in dilute gases saw the light. Nowadays, cold atom physics is a very active area of research, and it is successfully applied in for example, multibody physics, quantum statistics, atomic clocks and quantum information. Albeit the field of cold atoms is a success story, the precise details of the mechanisms of laser cooling is still not entirely known. To a large extent, this is because the actual cooling has become more of a technology used in order to access more fundamental questions, and the main importance with the cooling process is that it does work. It can also be assumed that many of the unanswered questions about the laser cooling process result rather from a large number of degrees of freedom, than due to any fundamental issues. Nevertheless, it has been argued that laser cooled atoms, under some circumstances, may produce velocity distributions important for fundamental statistical physics. There are indications that non-ergodic ensembles may be achieved, which could make the cooled atoms into a testbed for studies of basic statistical physics. In this colloquium, I will first recapitulate the emergence of laser cooling, and explain the fundamental principles of the cooling mechanisms. I will then address the issues, and existing controversies, about the statistical properties of laser cooled (non-degenerate) atoms.