Abstract :
[en] The aim of this master thesis was to study the problem of crossover resonances in saturation spectra.
In a typical saturation spectroscopy experiment, two counterpropagating laser beams with same frequency interact with a gas of atoms or molecules. As the frequency is in resonance with an atomic transition, both beams interact only with the zero-velocity group of atoms. As one of the beams (called pump beam) is saturating the atomic transition of concern, the second beam (called probe beam) cannot be absorb by these atoms, creating a Doppler-free dip in the absorption spectra at the resonance frequency.
Crossover resonances are spurious resonances which appear when both beams interact with two different transitions corresponding to two nonzero-velocity groups when the frequency of the beams is tuned at the exact average frequency between both transitions. These peculiar resonances can be positive or negative according to the configuration of the atomic levels involved. Predicting the exact relative intensity of such resonances is a difficult task. Indeed, one have to take into account several process such as optical pumping between hyperfine levels or Zeeman levels, atomic alignment, collisional processes, etc. All these phenomena can indeed have an influence on the atomic population of the levels involved and so determine the relative intensity of both true transitions or crossovers. A part of this master thesis was dedicated to an original review on these resonances and on the effects which have an influence on them.
Two models have been used to simulate saturation spectra of the transition at 379.99 nm of iron and the transition at 372.8 nm of ruthenium. The first model uses the so-called “Bordé&Bordé” formula (J. Bordé and C. Bordé, J. Mol. Spectrosc. 7, 353 (1979)) which can be directly applied for generating simulations. The second is an original generalization of the optical Bloch equations for atoms with a hyperfine structure and an arbitrary number of states.
The saturation spectrum of iron at 371.99 nm is well known (S. Krins, S. Oppel, N. Huet, J. von Zanthier, and T. Bastin, Phys. Rev. A 80, 062508 (2009)). However, no study has yet been done on the transition at 372.8 nm of ruthenium. To perform simulations, the specific mass shift coefficient, the field shift coefficient as well as the hyperfine structure constant and the quadrupolar coupling constant for this atom at the frequency of interest have been estimated during this master thesis.
Simulations show good qualitative agreement with the spectrum of iron for both models. Since Bordé&Bordé formula cannot be applied for crossover resonances occurring when both transitions do not share a common level, only one dip is not reproduced by this model. The generalized optical Bloch equations predict correctly the sign of all crossover resonances. Concerning the spectrum of ruthenium, both models show same features, but any qualitative comparison with experimental spectra could have been done because of the lack of spectra taken with optimal parameters. However, simulations seem to show that although the numbers of crossovers is larger in this case, their contribution to the total spectrum is small.
Further studies should be directed towards an improvement of the generalized optical Bloch equations for example to take into account effects of Zeeman pumping or collisional processes. Experimental studies in the case of ruthenium should also allow to a better estimation of spectroscopic parameters for this transition and could allow us to confront our simulations to new data.