My concern is that a highly rough surface would introduce points of non-contact between the sample and ATR crystal, which may mean the background would not be properly subtracted out. %T may be > 100% for some frequencies.
It depends on the hardness of the sample. ATR FTIR is frequently used with relatively "soft" polymers, where pressing to the ATR crystal smooths out the surface roughness. Even relatively hard polymeric materials can be studied with diamond anvil ATR. If your sample is really hard you may try to use diffuse reflection.
It depends on the hardness of the sample. ATR FTIR is frequently used with relatively "soft" polymers, where pressing to the ATR crystal smooths out the surface roughness. Even relatively hard polymeric materials can be studied with diamond anvil ATR. If your sample is really hard you may try to use diffuse reflection.
You asked a question which is very important for those using FTIR-ATR-techniques In studies of rough high-moduli (hard) materials. Prof. Gyorgy Banhegyi gave you excellent recommendations and I fully agree with him. Here I am going to add just a few words concerning the ATR measurements of such materials. Namely, since the relative intensity of ATR-spectrum strongly depends on penetration depth i.e., on amount of material which comes in contact with IR-radiation In the case where the contact of the rough material with the internal reflection element occurs only in several points - the resulting spectra will be inevitably distorted. At higher frequencies (due to the dependence of the absorbance intensity on the penetration depth) the spectral information can be completely lost and in the spectrum will be seen only a few very weak bands in the frequency range below 800 cm-1.
Indeed, those losses of spectral information appear to have been observed for a chemical image generated by FTIR microscopy. It uses a Ge ATR crystal, which has a penetration depth of 0.66 µm at 45 deg. There were several spots that yielded spectra with structure similar to the background, which suggests very weak contact in those areas. Also, the specimen was HDPE, which can be easily deformed. However, we were analyzing for contamination, which could have had very weak absorbances and very high moduli. These two factors may have coupled to produce spectra marred by steep baselines and %T > 100.