I am getting bands at 2171 and 2124 cm-1 in case of ZnO treated with CO+O2. 2124 cm-1 corresponds to what I am not able to find. Both of these corresponds to physisorbed CO? As these both vanishes during degassing
it is very likely that both peaks correspond to CO. I haven't seen your spectra, but would assume that there are even more than two features (maybe a deconvolution of the bands would help). have you tried exposing the ZnO to CO only in order to exclude any interference effects?
a potentially useful paper on CO on ZnO is http://dx.doi.org/10.1016/S0040-6090(01)01472-9