Towards an interferometer with thermal atoms trapped on a chip (Orale)

International audienceSummary form only given. Atom chips [1] have shown to be very successful for trapping and manipulating cold atoms. Atom interferometry in the vicinity of an atom chip has been recently achieved by coherently splitting a Bose-Einstein condensate with radiofrequency [2] and microwave [3] dressing. However, some limitations on the precision of the phase measurement occur in this case due to the phase diffusion process resulting from mean field interactions in the Bose-Einstein condensate.In this communication, we will discuss the possibility of using a thermal (i.e. non-condensed) ensemble of ultra-cold atoms trapped on an atom chip in order to create a trapped-atom interferometer with much weaker atomic interactions. In such a scheme (which can be seen as the atomic equivalent of a white light interferometer in optics), the coherence time will strongly depend on the level of symmetry that can be achieved between the potentials forming the two arms of the interferometer. For example, in a simple model of two harmonic potentials, it can be shown that a slight relative frequency difference Δω/ω can be associated with a limitation on the coherence time T given by the following expression: T≈ 2π1ίω kΘ Δω where Θ is the temperature of the atomic ensemble and k the Boltzmann constant (the latter expression being valid only for kΘ >> ! ω). For example, a relative asymmetry of 10-3 will result, for a temperature of about 1μK, to a limitation on the coherence time of 40ms, which favourably compares to what has been achieved experimentally with Bose-Einstein condensates [2]. We will then discuss how we plan to achieve, with the on-going experiment at Thales TRT [4], a splitting of the atomic cloud with the highest possible level of symmetry. The basic idea is to use internal state labelling on the so-called “clock states” of the 87Rb atom and microwave dressing [3] with two independent coplanar waveguides, each waveguide being preferentially resonant with one of the two internal states involved in the process. This way, we expect to have quasi-independent control over the two arms of the interferometer and thus to be able to achieve a coherent splitting as much symmetrical as possible.We will finally discuss the potential applications of this technique in terms of acceleration and gravity measurement with cold atoms on a chip, which could be of great interest in the future if combined, for example, with high bandwidth microelectromechanical sensors