If you are analyzing your film by a transmission sampling method, film thickness will be your effective path length. Mostly films with less than 50 micrometer should be used to avoid totally absorbed peaks. In this case your increasing film thickness will increase the concentration of the material under test. For better understanding please refer Beer Lambert Law A =  (absorptivity coefficient) x b (path length) x c (concentration).

If you are analyzing your film by a transmission sampling method, film thickness will be your effective path length. Mostly films with less than 50 micrometer should be used to avoid totally absorbed peaks. In this case your increasing film thickness will increase the concentration of the material under test. For better understanding please refer Beer Lambert Law A =  (absorptivity coefficient) x b (path length) x c (concentration).

In addition depending on the complex refractive index (n+ik) of both the substrate and the thin film and their respective thicknesses (actually the optical thickness d(opt)=n*thickness is of relevance) you might see oscillations in the transmission due to interference from multiple internal reflections within the substrate or the film (depending on the wavelength).

Some polymers show substrate induced crystallinity and because of that the IR spectra from solution-cast thin films may show less lines , as disorder lifts the IR selection rules!

Two points I want to share with you.

1. If, the film thickness increases from nano meter to micro meter, there may be some interference pattern (max and min) in addition with bond absorption peak.

2. If the structural properties is changes with film thickness, that will be affect FTIR spectra. For an example, some films internal stress will increases with thickness, FTIR peak will shift in higher energy region and FWHM will also increased.

Take a look at the thin film interference on Wikipedia

http://en.wikipedia.org/wiki/Thin-film_interference