Physics of polarized scattering at multi-level atomic systems

TL;DR

This paper investigates the polarization phenomena in multi-level atomic systems, especially the sodium D1 line, through laboratory experiments and extended quantum scattering theory, revealing complex polarization structures unexplained by standard models.

Contribution

It extends the theory of polarized scattering to include coherences in both initial and final states, successfully modeling experimental results for the sodium D1 line.

Findings

Laboratory results show rich polarization structures in the D1 line.

Extended theory including coherences matches experimental data quantitatively.

Constraints from D2 line data reduce free parameters in modeling D1 line polarization.

Abstract

The symmetric peak observed in linear polarization in the core of the solar sodium D$_1$ line at 5896 \AA\ has remained enigmatic since its discovery nearly two decades ago. One reason is that the theory of polarized scattering has not been experimentally tested for multi-level atomic systems in the relevant parameter domains, although the theory is continually being used for the interpretation of astrophysical observations. A laboratory experiment that was set up a decade ago to find out whether the D$_1$ enigma is a problem of solar physics or quantum physics revealed that the D$_1$ system has a rich polarization structure in situations where standard scattering theory predicts zero polarization, even when optical pumping of the $m$ state populations of the hyperfine-split ground state is accounted for. Here we show that the laboratory results can be modeled in great quantitative detail if the theory is extended to include the coherences in both the initial and final states of the scattering process. Radiative couplings between the allowed dipole transitions generate coherences in the initial state. Corresponding coherences in the final state are then demanded by a phase closure selection rule. The experimental results for the well understood D$_2$ line are used to constrain the two free parameters of the experiment, collision rate and optical depth, to suppress the need for free parameters when fitting the D$_1$ results.