I am doing quantitative work on carbonyl compounds dissolved in dichloromethane (DCM). I've noticed that on an ATR spectra of DCM that barely any peaks can be seen in the 1500cm-1 to 4000cm-1 range. However, when i switch to using a 3mm KBr transmission cell, many sharp peaks appear in the DCM spectra in the previously mentioned range. Why are these peaks here? Do these "ghost" peaks appear due to the concentration and pathlength being too high?