The physical basis for absorption of light

Abstract

THE effect of low intensity light on the wave functions of gas-phase molecules and atoms is well understood, and solutions of the time-dependent wave equation, or its equivalent, are common. However, two different operational approaches can be used to examine the relationship between the wave function and light absorption. The first approach¹ involves examining the time dependence of the populations of internal energy states. Changes in the average internal energy are then used to infer a corresponding absorption of energy from the radiant field by invoking energy conservation. The second approach² involves computing the polarisation of a medium induced by the radiant field. Absorption is then shown to be proportional to the imaginary part of the susceptibility. The latter is a natural approach for sample geometries which are small compared with the wavelength of the light, but it can also be used for larger geometries and shorter wavelengths. When the sample geometry is large compared with the wavelength, the field from the polarisation lags behind the polarisation itself by 90° (ref. 3). Absorption then cornes from the part of the induced polarisation which lags the incident field by 90°. Absorption of light in this picture is due to stimulated emission which is 180° out of phase with the incident field. Both the population and polarisation approaches have been used to describe experiments in which the absorption coefficient is observed as a function of time and as a function of the frequency of the incident light³. The two approaches are equivalent for the usual case in which a power detector is used to measure the time-dependent absorption coefficient in the sample⁴. In the experiment described here, a Stark-modulated sample is irradiated at a single frequency and radiation emerging from the sample is resolved into frequency components using a spectrum analyser to distinguish between the upper and lower modulation sidebands. It will be shown here that the upper and lower sidebands are not always equal and that this result is only consistent with a modified polarisation approach in which the stimulated emission is explicitly considered.