as you have mentioned your films are having 5-10nm, better choose transmission mode...btw by saying so need to know the substrate where the films are being deposited...
As Dillip kumar Mishra suggested, transmission mode is the best choice when your sample absorbs properly.
When the absorption is too weak (because the sample only 10 nm thick), you could try multiple internal reflection. Another possibility would be to bring you sample on a gold (or another highly conductive, and hence, IR-reflecting) surface and measure at grazing incidence.
ATR means attenuated total reflection which utilizes a so-called evanescent wave that extends beyond the internal interface (i.e. from your ATR crystal to air at which you make the total reflection). This evanescent wave decays over a fraction of the wavelength, so in infrared this is over micrometer(s), and your film is only 5-10 nm thick, and therefore you don't get too much signal. What you can try is something that is more surface sensitive. I would try to make internal reflection at very small angles. But you need a reflection fixture for this. I worked on this about 18 years ago, I remember that the Seagull cell from Harrick Scientific was one of the best reflection setups for this
A thicker film would definitely increase the amount of absorbed light.
But 30 nm is still not much. In my opinion reflection measurements at a small angle with your film on a highly conductive surface would by a better choice.