Speaker
Description
The mercury co-magnetometer used in neutron EDM experiments relies on optical detection of the atomic spin precession. The probe light itself induces a shift of the precession frequency, proportional to the light power, know as the light shift.
This phenomenon was discovered by Cohen-Tannoudji in the early 1960s, who identified two distinct mechanisms: the virtual light shift, occurring when the probe light frequency is detuned from the atomic resonance, and the real light shift, arising because atoms spend a finite fraction of time in the excited state, where their spin precesses at a faster rate. Although the real light shift has received comparatively little attention, it constitutes an important systematic effect in neutron EDM experiments.
We present a revisited theory, with focus on the real light shift, recasting Cohen-Tannoudji's formulas in a modern framework. The theory provides a unified description of the light shift and the associated depolarization rate, and predicts configurations in which the shift vanishes depending on the polarization of the probe light. The predictions are compared with dedicated measurements performed in Grenoble and with recent measurements obtained with the n2EDM experiment at PSI.