The basis of photothermal spectroscopy is the change in thermal state of the sample resulting from the absorption of radiation.
And it is this fact that gives the technique its high sensitivity, because in transmission techniques the absorbance is calculated as the difference between total light impinging on the sample and the transmitted (plus reflected, plus scattered) light, with the usual problems of accuracy when one deals with small differences between large numbers, if the absorption is small.
It works via a sort of "mirage effect"[1] where a refractive index gradient exists adjacent to the test sample surface.
A probe laser beam is refracted or bent in a manner proportional to the temperature gradient of the transparent medium near the surface.
Transverse PDS can be done in Nitrogen, but better performance is gained in a liquid cell: usually an inert, non-absorbing material such as a perfluorocarbon is used.
[citation needed] In both collinear and transverse PDS, the surface is heated using a periodically modulated light source, such as an optical beam passing through a mechanical chopper or regulated with a function generator.
Due to the photoheating effect, the coupling efficiency is changed and characterized to indicate the thin film absorption.