The spectra show the presence of atmospheric CO2. Purging your instrument is the best way to remove this. I use compressed air that has been passed through a CO2 and H2O scrubber on my FT-IR system and that works really well.

Can you do the measurements with the sample and then switch to the same measurement with a vacuum reference (keeping all other optical surfaces/windows the same) and take the difference?  If you have high resolution in wavelength, there might be a way to look at individual absorption lines and relate their linewidths to the pressure and temperature, which can be correlated to the particular source gas, but you might have difficulty distinguishing without some kind of modulation (chopping to add modulation).   

Thanks to all Profs and Scholars.. 

Very often - even if the spectrometer's chamber is well purged by nitrogen - the traces of CO2 still can be seen - normally as a doublet of bands. But, suppose you are trying to obtain an IR-spectrum of some organic acid - by employing the KBr technique, moreover - in order to remove the traces of water you use a heat-treatment (usually 20-30 min at 125C) - in such a case - the CO2 as the acid-decomposition product - will be seen in the spectrum as a Single peak - just because of absorption of CO2 molecules on the walls of the pores in KBr. I have many times faced this situation. Some similar cases are also described in Marek Urban's Book 

Thanks Prof. I agree.

Best way to avoid atmospheric CO2 is to purge with nitrogen. If you re run the sample just after background often time CO2 from atmosphere can be avoided. Peak around 2341 cm-1 is from atmospheric CO2 but I am not sure about the peak around 1368 is due to atmospheric CO2 or not. Also some of the FTIR software's has a built in feature to correct atmospheric CO2. Hope this helps.

Please send some reference paper in which it is specified that this 2341 peak is due to background CO2

The CO2 peak around 15 um, or 660cm^(-1) is almost invisible.