The classical case is measurement of hydrocarbons extracted from water by tetrachloroethylene. In this case TCE should be purified in florisil to remove polar interference (not HC) and it makes it loosing also ... polar additive of TCE in FTIR. As a consequence if you made your background on "pure TCE" you obtain large negative peaks. To avoid that you should make the background in "pure TCE purified in florisil". In your case, maybe there is such a difference in the treatment of your background solvent versus measurement sovent? Good luck
@S Taleb: infrared transparent optics (including the beamsplitter) for most FTIR instrument use KBr (more rarely, CsI), since it has a relatively wide range of infrared transparency. The KBr plate in contact with outside air needs to be coated with a moisture barrier (since KBr and especially CsI are very hygroscopic). At least in instruments I've used, the manufacturer sales rep said a very thin polyethylene film is used on that optic.
Now, in the perfect situation, that polyethylene signal from the costing should show up only in the raw background "single beam" (~black body radiation) data, and really shouldn't show up in the transmittance spectrum either of the background or the sample. But these features can still crop up in the sample spectrum in practice. Reflection absorption mode in particular seemed to cause this. But also if your IR beam source fluctuates enough, that may also cause these features to pop up.
Please note, I'm just giving an example of what might be the source of your 2800-3000 cm^-1 signal. It could also be contamination in your sample as others suggested, although I suspect you already ruled this out experimentally.
Thank you very much for these explanations. I don't use in this specific case (aqueous solution of ortho-cresol) KBr cell. The analyzed sample is liquid.
@ Safia Taleb: when I mentioned KBr, it was not in reference to the sample cell, but rather the optics on the FTIR spectrometer.
Since you're using a liquid sample cell, there may be the possibility that, when you take a background, and the sample cell without liquid has some scratches, that it would scatter a decent amount of IR radiation when empty compared against when full of liquids (scattering scales with refractive index difference). Although, I assume you are taking background by filling your cell with water (no ortho-cresol)? If so, I would not think the scattering would matter much. If your background is sample cell with air inside (no liquid), then scattering may play a bigger role. Again, without further detail, all this is speculation at this point.
Hello Safia Taleb - I am seeing inverted C-H bands also, which are not due to contamination. This is on a Thermo Scientific iN10 Fourier transform infrared microscope (μ-FTIR) to measure reflectance. Did you ever chase down the solution to this issue?
We also have inverted peak in 2800-3000 cm-1 region (C-H band) in the FTIR-ATR transmittance spectra, but some sample specially pristine Aluminum sample, we saw normal C-H band. Attached sample spectra etched Aluminium sample, after 10 days the peaks have inverted. Other laser treated samples also have inverted peaks only this region.
We cleaned ATR unit by using acetone, remove background and used smoothing filter, however this peak is presented.
How can we explain this inverted transmittance peak in the FTIR-ATR spectra?
I would like to all of your valuable contributions.