Skip to Main content Skip to Navigation


Abstract : At the time of my arrival in Laboratoire Aimé Cotton (LAC) in November 2006, the activity of the experimental group of cold molecules in LAC was, on one hand on the formation and manipulation of cold molecules in a Magneto-Optical Trap (MOT), and on the other, in the realization of a dipole trap for cesium atoms. The preparation of a dipole trap of Cs atoms aimed both in the preparation of Bose-Einstein Condensation (BEC) of atomic Cs, and in the study of preparation and manipulation of Cs dimers in ultra-low temperatures. My enrollment in the activity of the experimental group of cold molecules in LAC was both in the preparation of the dipole trap, and in the study of cold molecule's creation and manipulation. In the first part of my thesis, I describe the studies conducted in the period November 2006 to October 2008. In this period, I focused my efforts on the study of different techniques for the loading of a dipole trap with Cs atoms, using a pre-existing set-up. The target was the creation of a cold and dense trapped atomic sample, in which the evaporation technique could be applied, to further cool the sample down to the critical temperature for the creation of a Cs BEC. At the beginning of the experiment, a Cs BEC had been reported only once [Web03], after years of unsuccessful efforts by many groups [Sod98, Boir98, Thom04]. Despite the difficulty of the subject, the interest in preparing samples of ultra-cold Cs atoms remained high, especially due to experiments related to ultra-cold molecules, as the creation of a molecular BEC [Herb03], or the formation of Cs trimers and the observation of Efimov states [Lee07, Knoo08]. The strategy upon which the dipole trap experiment was based, is theoretically studied in a previous publication of the group [Comp06], which considers rapid evaporation of a dense atomic sample in a crossed, deep dipole trap. This dense dipole trap, is provided by superimposing the dipole laser to a much larger trapped atomic sample, the so-called atomic reservoir. Collisions in this atomic reservoir can thermalize the sample and lead to the transfer of atoms in the dipole trap. Since this process can last for relatively large time intervals, it can result to higher loading efficiencies with comparison to alternative, instantaneous loading methods. This approach is very different to the one used in the only successful BEC experiment at that time [Web03], in which a shallow, very cold, but not so dense dipole trap, is prepared with the use of Raman Sideband Cooling. In the theoretical proposal reported in [Comp06], a magnetic trap was considered for the realization of the atomic reservoir, while the initial experimental study of this approach is the subject of a previous thesis in our group [Stern08]. However, the general ideas considered in this approach, allowed for the substitution of this magnetic reservoir, by several atomic traps. Thus, the experimental studies discussed in the first part of my thesis, are the continuation of the work made during the thesis of G. Stern [Stern08], with whom I collaborated in the beginning of my thesis. In particular, I studied the loading a dipole trap from a magnetic trap reservoir, and compared it to the loading obtained when the magnetic reservoir is replaced by a Dark-SPOT and a Compressed MOT (C-MOT). Furthermore, all these reservoir-loading methods are compared to a simpler, instantaneous-loading method which involves optical molasses. By June 2008, it was made clear that our experimental approach to the dipole trap loading, could not lead to the preparation of a sufficiently cold and dense atomic sample, in which evaporating cooling could be applied for the preparation of cesium atoms in ultra-low temperatures. In the same time, the approach considered in the first successful realization of a Cs BEC [Web03] gained ground, since the experiments reported in [Hung08] showed that it could provide with a Cs BEC with a fast and relatively simple experimental sequence. Thus, we also attempted the preparation of an ultra-cold Cs sample with the use of a shallow, not very dense, but very cold dipole trap provided by Raman Sideband Cooling. Unfortunately, a series of experimental problems related to the old vacuum system used in our experiment, prevented us from creating an ultra-cold atomic cesium sample with such an approach, despite the encouraging preliminary results. On the same period, the studies of the manipulation of cold molecules created in a MOT, conducted by members of the experimental group of cold molecules in LAC, advanced considerably, leading to the demonstration of the vibrational cooling technique reported in [Vit08]. The operating principle of the vibrational cooling technique is similar to the one of optical pumping in atoms [Kast66]. In this process, molecules that initially lie in different vibrational levels, are simultaneously excited by shaped broadband light and are accumulated to a single vibrational level via spontaneous emission. The accumulation to a single vibrational level, is accomplished by choosing to remove from the shaped pulse, all frequencies resonant to transitions from this level and thus turn it to a dark state. The technique enjoys simplicity and generality, and its demonstration opened the way for many interesting extensions, some of which are the subject of the second part of my thesis. More particularly, my activity in the cold molecule experiment which is discussed in the second part of my thesis, considered several extensions and generalizations of the vibrational cooling technique. The first extension to be considered, was the transfer of the molecular population to any pre-selected vibrational level, via optical pumping induced by more sophisticated, shaped femptosecond pulses, and is also discussed in [Sof09]. Another extension, considered the realization of vibrational cooling and molecular population transfer with the use of a broadband, non-coherent, diode light source, instead of a femptosecond laser and is reported in [Sof09b]. Another extension was considered to be the vibrational cooling of Cs molecules in their ground triplet electronic state, in addition to the ground singlet state, that was so far manipulated. Despite the optimistic initial predictions, the experimental study did not led to considerable results. However, this 'failed' experimental study, provides with an opportunity to revisit the various key elements of the vibrational cooling technique, and to consider the possible reasons that can lead to its failure. Such a discussion is particularly useful for the following study of the extension of the vibrational cooling technique to heteronuclear molecules through the example of NaCs. Finally, the generalization of the vibrational cooling technique to include rotation, which is theoretically considered in various publications of the group in which I participated [Vit09, Sof09, Sof09c], is discussed. In addition to these theoretical considerations, I discuss the preliminary experiments considered for rotational cooling, which involve the preparation of rotationally resolved depletion spectroscopy, and which are also discussed in [Fio09].
Mots-clés : lasers atoms molecules optique
Complete list of metadatas

Cited literature [167 references]  Display  Hide  Download
Contributor : Dimitrios Sofikitis <>
Submitted on : Tuesday, January 5, 2010 - 3:42:22 PM
Last modification on : Wednesday, October 14, 2020 - 3:41:35 AM
Long-term archiving on: : Thursday, June 17, 2010 - 9:07:06 PM


  • HAL Id : tel-00444057, version 1



Dimitris Sofikitis. PREPARATION ET MANIPULATION DES ATOMES ET DES MOLECULES DE CESIUM. Physique Atomique [physics.atom-ph]. Université Paris Sud - Paris XI, 2009. Français. ⟨tel-00444057⟩



Record views


Files downloads