FTIR or fourier transformation infrared red spectroscopy is the name of the special analysis done on various products to compare the chemical composition of products . The question is what is the error you are getting in your graphs. If you are able to compare them with the graohs and oeaks in the library ?

First, those are spectra, not graphs. Second, absorption is overall quite low but still, the absorption bands seem to be OK. Third, which spectrum corresponds to which CNT sample ?

Without having any detail about your spectra, it's acrobatic to try and interpretate them... even more to tell if there's a "problem".

From what I'm seeing, no carbonyl band in any of them, anyway.

Pierre Caulet a spectra is before functionalization and b spectra is after functionalization with acid.

b spectra is actually okay compare to previous work but the problem is a spectra. It supposed to be a somewhat straight line with small ppeak around 3500.

As stated above the spectra are very weak and the ripples on both spectra could be due to interference bands due to internal reflection in the tubes. As stated there is no sign of the usual carboxylate bands in the 1700cm-1 region. However the strong band at about 1550cm-1 in spectrum (a) is similar to some carboxylic acid salts. This is reduced in (b) with a strong OH at about 3400 and 1000 cm-1. The baseline looks very flat and the OH bands appear out of proportion. Has a background correction "flat" been applied which could account for the strange band shapes ? The bands at about 2300 cm-1 are due to in-balance in the CO2 band between the background and sample scans.

You say the band at approx 3500 cm-1 should be small. What does small mean ? The spectra are scale expanded so the band looks strong. On a full 0 - 100% T it is between 100%T-anout 95%T