We assess the utility of non-resonant light for rapidly controlling vapor density in glass cells and discuss the diffusion processes involved, revealing the pertinent potentials experienced by atoms moving between the adsorbed and vapor phases.
We review some recent studies on light-induced atomic desorption (LIAD) from dielectric surfaces. Alkali-metal atoms adsorbed either on organic films or on porous glass are released into the vapor phase under illumination. The measurements were performed in Pyrex resonance cells either coated with siloxane films or containing a porous glass sample. In both cases, the experimental results show that LIAD can be used to produce atomic densities suitable for most atomic physics experiments. Moreover, we find that photoinduced effects, correlated with LIAD, produce reversible formation and evaporation of alkali-metal clusters in porous glass. These processes depend on the light frequency, making the porous glass transmittance controllable by light.
Atomic-vapor density change due to light induced atomic desorption (LIAD) is studied in paraffin-coated rubidium, cesium, sodium, and potassium cells. In the present experiment, low-intensity probe light is used to obtain an absorption spectrum and measure the vapor density, while light from an argon-ion laser, array of light emitting diodes, or discharge lamp is used for desorption. Potassium is found to exhibit significantly weaker LIAD from paraffin compared to Rb and Cs, and we were unable to observe LIAD with sodium. A simple LIAD model is applied to describe the observed vapor-density dynamics, and the role of the cell's stem is explored through the use of cells with lockable stems. Stabilization of Cs vapor density above its equilibrium value over 25min is demonstrated. The results of this work could be used to assess the use of LIAD for vapor-density control in magnetometers, clocks, and gyroscopes utilizing coated cells.
The light-induced atomic desorption effect, known as LIAD, has been explained so far by a debatable Atutov's theory [S. N. Atutov et al., Phys. Rev. A 60, 4693 (1999)]. In this Brief Report, we propose an approach in which the difficulties of this theory are removed. The theory yields predictions for the atomic vapor density n(t) in a fully illuminated resonance cell, the time tmax that it takes the vapor density to arrive at a maximum value nmax and the dependence of nmax and dn/dtt=+0 on the desorbing light intensity. Some of our results are at variance with outcomes of Atutov's model (such as saturation effect for nmax and the dependence of tmax on the light intensity) but are in agreement with the experimental data. Additionally, the model presented in this Brief Report satisfactorily explains a reservoir effect recently investigated in experiments where cells equipped with a lockable stem were used.